Identification of the First PPARα/γ Dual Agonist Able to Bind to Canonical and Alternative Sites of PPARγ and to Inhibit Its Cdk5-Mediated Phosphorylation

Article abstract of DOI:10.1021/acs.jmedchem.8b00835

A new series of derivatives of the PPARα/γ dual agonist 1 allowed us to identify the ligand (S)-6 as a potent partial agonist of both PPARα and γ subtypes. X-ray studies in PPARγ revealed two different binding modes of (S)-6 to the canonical site. However

Full text of DOI:10.1021/acs.jmedchem.8b00835

Subscriber access provided by University of South Dakota
Article
Identification of the First PPAR#/# Dual Agonist Able to Bind to Canonical and
Alternate Sites of PPAR# and to Inhibit its Cdk5-Mediated Phosphorylation
Antonio Laghezza, Luca Piemontese, Carmen Cerchia, Roberta Montanari, Davide Capelli, Marco Giudici,
Maurizio Crestani, Paolo Tortorella, Franck Peiretti, Giorgio Pochetti, Antonio Lavecchia, and Fulvio Loiodice
J. Med. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.jmedchem.8b00835 • Publication Date (Web): 10 Sep 2018
Downloaded from http://pubs.acs.org on September 11, 2018
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a service to the research community to expedite the dissemination
of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in
full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully
peer reviewed, but should not be considered the official version of record. They are citable by the
Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore,
the “Just Accepted” Web site may not include all articles that will be published in the journal. After
a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web
site and published as an ASAP article. Note that technical editing may introduce minor changes
to the manuscript text and/or graphics which could affect content, and all legal disclaimers and
ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or
consequences arising from the use of information contained in these “Just Accepted” manuscripts.
is published by the American Chemical Society. 1155 Sixteenth Street N.W.,
Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.

Page 1 of 74
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Identification of the First PPARα/γ Dual Agonist Able to Bind to Canonical and Alternate
Sites of PPARγ and to Inhibit its Cdk5ꢀMediated Phosphorylation
1
1
2
3
3
Antonio Laghezza, Luca Piemontese, Carmen Cerchia, Roberta Montanari, Davide Capelli,
4
4
1
5
3
Marco Giudici, Maurizio Crestani, Paolo Tortorella, Franck Peiretti, Giorgio Pochetti, Antonio
Lavecchia, Fulvio Loiodice*
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
2
1
1
Dipartimento Farmacia-Scienze del Farmaco, Università degli Studi di Bari “Aldo Moro”, via
Orabona 4, 70125 Bari, Italy
2
Dipartimento di Farmacia, “Drug Discovery” Laboratory, Università degli Studi di Napoli
“Federico II”, via D. Montesano 49, 80131 Napoli, Italy
3
Istituto di Cristallografia, Consiglio Nazionale delle Ricerche, Montelibretti, 00015 Monterotondo
Stazione, Roma, Italy
4
Dipartimento di Scienze Farmacologiche e Biomolecolari, Università degli Studi di Milano, via
Balzaretti 9, 20133 Milano, Italy
5
Aix Marseille Université, INSERM 1263, INRA 1260, C2VN, 13005 Marseille, France
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 2 of 74
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Abstract
A new series of derivatives of the PPARα/γ dual agonist 1 allowed to identify the ligand (S)ꢀ6 as a
potent partial agonist of both PPARα and γ subtypes. Xꢀray studies in PPARγ revealed two different
binding modes of (S)ꢀ6 to canonical site. However, (S)ꢀ6 was also able to bind an alternate site as
demonstrated by transactivation assay in the presence of a canonical PPARγ antagonist and
supported from docking experiments. This compound did not activate the PPARγꢀdependent
program of adipocyte differentiation inducing a very less severe lipid accumulation compared to
rosiglitazone, but increased the insulinꢀstimulated glucose uptake in 3T3ꢀL1 adipocytes. Finally,
(S)ꢀ6 inhibited the Cdk5ꢀmediated phosphorylation of PPARγ at serine 273 that is currently
considered the mechanism by which some PPARγ partial agonists exert antidiabetic effects similar
to thiazolidinediones, without showing their typical side effects. This is the first PPARα/γ dual
agonist reported to show this inhibitory effect representing the potential lead of a new class of drugs
for treatment of dyslipidemic type 2 diabetes.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
ACS Paragon Plus Environment

Page 3 of 74
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
INTRODUCTION
Type 2 diabetes mellitus (T2DM) strikes nearly 300 million people worldwide and represents a
1
major cause of morbidity and mortality. This chronic metabolic disorder results in a combination
of dyslipidemia, glucose intolerance, and inflammation. Peroxisome proliferatorꢀactivated receptors
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
(PPARs) are ligandꢀactivated transcription factors that play a key role in the regulation of a large
number of genes whose products are directly or indirectly involved in glucose homeostasis and lipid
metabolism. For this reason, they have been considered suitable targets for the treatment of
metabolic disorders.
There are three distinct PPAR subtypes, PPARα, PPARγ, and PPARδ, with different tissue
distributions and physiological functions. PPARα is important for the uptake and oxidation of fatty
acids and lipoprotein metabolism and is the molecular target of fibrates, a class of drugs that lower
2
ꢀ4
plasma triglycerides and increase HDL cholesterol levels in humans. Furthermore, PPARαꢀ
agonists can increase the stability of atherosclerotic plaques, reduce the risk of atherothrombosis,
and reduce hepatic fat accumulation leading to nonꢀalcoholic steatohepatitis/fatty liver disease
5
(NASH/NAFLD). Fibrates are weak PPARαꢀagonists whose efficacy is partly limited by doseꢀ
dependent side effects. Commonly reported adverse events include elevations in markers for
3,4,6
cardiovascular disease, renal disease, and liver dysfunction.
To address this problem, a new
generation of more potent PPARα agonists with a lower potential for adverse effects is being
developed. In some cases, a greater lipidꢀmodifying efficacy associated with an improved
7
,8
safety/tolerability profile has been obtained. Thiazolidinediones (TZDs), including rosiglitazone
and pioglitazone, are the only clinically available agents controlling hyperglycemia by improving
insulin resistance, which is regarded as hallmark and early etiologic basis for T2DM. TZDs have
been demonstrated to behave as PPARγ full agonists, which activate PPARγ leading to a wide
9
spectrum of genes expression changes involved in insulin sensitizing activity. However, recent
studies have demonstrated that the use of these compounds is associated with unwanted effects,
such as weight gain, fluid retention, increased incidence of cardiovascular events, and bone
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 4 of 74
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
1
0ꢀ13
fractures.
Different strategies have been conceived to obtain compounds with good therapeutic
potential, trying to avoid the known side effects.
In particular, PPARα/γ dual agonists represent a promising solution given that they exhibit the
insulin sensitizing potential of PPARγ agonists along with the beneficial lipid modulating activities
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
14
of PPARα agonists. Dualꢀacting PPARα/γ agonists, therefore, could simultaneously correct
insulin resistance and lipid imbalance, thereby overcoming the side effects of selective PPARγ
15ꢀ19
agonists. In the past, several dual PPARα/γ agonists have been developed and evaluated
and
recently, one such agent, saroglitazar, has been approved in India for the treatment of diabetic
20
dyslipidemia and hypertriglyceridemia with type 2 diabetes not controlled by statin therapy.
Saroglitazar is predominantly a PPARα agonist, with modest PPARγ agonist actions. This is in
contrast to previously developed dual PPARα/γ agonists, which had either predominantly PPARγ
agonistic activity or similar agonistic activity on both PPARα and PPARγ receptors. In particular,
an appropriate modulation of PPARγ activity appears the key challenge for the development of a
dual agonist able to uncouple the adverse effects of the ligand from therapeutic effects. In this
regard, recent observations indicate that the improved therapeutic profile of selective PPARγ
modulators (SPPARγMs) compared to PPARγ full agonists is a consequence of peculiar ligandꢀ
receptor interactions resulting in a different structural conformation of the complex that gives rise to
21ꢀ26
21,22
different PPAR transcriptional signatures.
Recently, studies from Choi et al.
have shown
that cyclinꢀdependent kinase 5 (Cdk5)ꢀmediated phosphorylation of PPARγ at serine 273 may be
involved in the pathogenesis of insulin resistance. S273 phosphorylation leads to dysregulation of a
large number of PPARγ target genes that are known to be associated with insulin sensitization.
Inhibition of the S273ꢀphosphorylation was thus proposed as a plausible mechanism by which antiꢀ
diabetic PPARγ partial agonists or modulators increase insulin sensitivity, without activating the
full range of PPARγ target genes, and thereby avoiding known side effects such as fluid retention,
bone fractures, and weight gain.
In our previous studies, we have reported the synthesis and biological activity of a number of
ACS Paragon Plus Environment

Page 5 of 74
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
carboxylic acid derivatives showing an interesting dual activity toward PPARα and PPARγ
25ꢀ34
receptors.
In particular, we identified the lead compound LT175 ((S)ꢀ1, Fig. 1), a dual PPARα/γ
35
ligand with a partial agonism profile toward the γ subtype. The Xꢀray structure of the complex
with PPARγ showed that this compound interacts with the receptor in a newly identified
hydrophobic region called “diphenyl pocket” inducing a conformation of the ligand binding domain
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
(LBD) less favorable to the recruitment of coactivators required for full activation of PPARγ. This
new interaction influences the expression of PPARγ target genes, leading to an improved
therapeutic profile compared with rosiglitazone. In fact, (S)ꢀ1 showed potent insulinꢀsensitizing
3
6
effects with reduced adipogenic properties. The corresponding R enantiomer of (S)ꢀ1 ((R)ꢀ1, Fig.
) is still active on PPARγ even though with lower potency and slightly reduced efficacy.
1
Interestingly, the two stereoisomers occupy different regions of the receptor as shown from the Xꢀ
ray structure of their complex with PPARγꢀLBD (Fig. 1): (S)ꢀ1 occupies a branch named “diphenyl
pocket” between helices H11 and H3, whereas (R)ꢀ1 accommodates in the region usually occupied
3
5
by other known partial agonists, between helix H3 and the βꢀsheet. With the aim to evaluate the
effects resulting from the combination of both binding modes, we designed the two novel molecular
scaffolds shown in Fig. 2: compound 2 originates from the fusion of the diphenyl systems of two
molecules of (R,S)ꢀ1, whereas compound 3 derives from the fusion of the benzyl ring of one
molecule of (R,S)ꢀ1 with the diphenyl nucleus of a second molecule. For this compound a gradual
structural simplification of the distal part of the molecule was also realized obtaining derivatives 4ꢀ
6. All the racemates of this series were evaluated for their PPARα/γ activity by the transactivation
assay. For the most active compounds 3 and 6 we prepared all the corresponding stereoisomers to
explore the influence of absolute configuration on receptor activation. (S)ꢀ6 turned out to be the
ligand with the most interesting pharmacological profile able to potently activate both PPARα and γ
subtypes as a partial agonist. Further in vitro experiments confirmed that (S)ꢀ6 is devoid of some of
the known side effects of PPARγ full agonists, but significantly increased the insulinꢀstimulated
glucose uptake. Xꢀray studies and docking experiments were performed on the complex PPARγ/(S)ꢀ
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 6 of 74
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
6
to gain more insights into the molecular mechanism of PPARγ activation and revealed different
binding modes with both canonical and alternate sites. Finally, (S)ꢀ6 was able to inhibit the Cdk5ꢀ
mediated phosphorylation of PPARγ at serine 273. As far as we know, this ligand is the first
PPARα/γ dual agonist reported to show this inhibitory effect. A compound with these properties is
expected to exhibit the beneficial lipid modulating activities of PPARα agonists along with the
insulin sensitizing potential of PPARγ agonists but without the typical side effects of selective
PPARγ full agonists. Therefore, (S)ꢀ6 may represent the potential lead of a new class of drugs with
better pharmacological properties in the treatment of dyslipidemic type 2 diabetes.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
CHEMISTRY
The synthesis of compound 2 is depicted in Scheme 1 and was carried out starting from the ethyl
ester of racemic phenyllactic acid and 4,4’ꢀdiphenol, which were condensed, under Mitsunobu
conditions, to obtain the diester intermediate 2b. This twoꢀstep procedure ensured better yields
compared to the singleꢀstep one utilizing a molar ratio 2:1 of the starting materials. The hydrolysis
of 2b allowed to obtain the desired acid. The synthesis of racemic compounds 3ꢀ5 is reported in
scheme 2 and started from the ethyl esters of phenyllactic, 2ꢀbromopropanoic or bromoacetic acids,
respectively. These esters were alternatively reacted with 4ꢀhydroxyphenylboronic acid pinacol
ester, under Mitsunobu conditions, to give the ester intermediates 3aꢀ5a. These underwent Suzuki
coupling by condensation with ethyl 2ꢀ(4ꢀphenylꢀphenoxy)ꢀ3ꢀ(4ꢀbromophenyl)propanoate 6e (see
scheme 3) affording diesters 3b−5b, which were hydrolyzed to the desired acids. The racemic acid
6 was prepared through the procedure represented in scheme 3. Diethyl 2ꢀchloroꢀmalonate was
reacted with sodium 4ꢀphenylꢀphenate to give diester 6a which was condensed with 4ꢀbromoꢀ
benzylbromide in the presence of NaH affording 6b. The subsequent Suzuki coupling of this
intermediate with benzeneboronic acid gave the diester 6c whose saponification and
decarboxylation led to the target acid. Alternatively, 6b was hydrolyzed and decarboxylated to 6d
which was transformed into the corresponding ethyl ester 6e to be used as reported in scheme 2.
ACS Paragon Plus Environment

Page 7 of 74
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
The synthesis of the four stereoisomers of compound 3 was carried out similarly to the
corresponding racemate (see scheme 2) starting from the readily available (S)ꢀ or (R)ꢀethyl
phenyllactates. The Mitsunobu reaction of these starting materials with 4ꢀhydroxyphenylboronic
acid pinacol ester occurred with inversion of configuration affording the esters (R)ꢀ3a and (S)ꢀ3a,
respectively (scheme 4). These intermediates were condensed alternatively with (S)ꢀ or (R)ꢀethyl 2ꢀ
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
(4ꢀphenylꢀphenoxy)ꢀ3ꢀ(4ꢀiodophenyl)propanoate 6g (see scheme 5) obtaining the four
stereoisomers of 3b, which were hydrolyzed to the corresponding optically active acids. Finally,
(R)ꢀ6 and (S)ꢀ6 were prepared as depicted in scheme 5. The commercially available (R)ꢀ or (S)ꢀ4ꢀ
iodoꢀphenylalanines were converted into the corresponding hydroxyꢀesters (R)ꢀ6f and (S)ꢀ6f
3
7
through a procedure known to occur with retention of configuration. These intermediates were
condensed with 4ꢀphenylphenol under Mitsunobu conditions affording compounds (S)ꢀ6g and (R)ꢀ
6g, respectively, with inverted configuration. The Suzuki crossꢀcoupling reaction with
benzeneboronic acid and the final hydrolysis led to the target acids.
The diastereomeric (de) and/or enantiomeric excesses (ee) of all optically active isomers were
determined by HPLC analysis on chiral stationary phase (see the Experimental Section).
RESULTS AND DISCUSSION
Compounds 2ꢀ6 were evaluated in vitro for their agonist activity toward the human PPARα
(hPPARα) and PPARγ (hPPARγ) subtypes by employing GAL4ꢀPPAR transactivation assay. For
this purpose, GAL4ꢀPPAR chimeric receptors were expressed in transiently transfected HepG2
27
cells according to a previously reported procedure. In particular, the results obtained were
compared with corresponding data for Wyꢀ14,643 and rosiglitazone used as reference compounds
in the PPARα and PPARγ transactivation assays, respectively. Maximum obtained fold induction
with the reference agonist was defined as 100%. The activity of 2ꢀ6 was also compared with the
lead compound (S)ꢀ1.
As a preliminary test, we evaluated the efficacy of racemates 2ꢀ6 on PPARα and PPARγ at
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 8 of 74
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
concentrations increasing up to 100 ꢀM. As shown in Fig. 3, compound 2 was completely inactive,
whereas 4 and 5 showed only a low activation of both subtypes. As a whole, only 3 and 6 showed a
moderate activity and, for this reason, we prepared all the stereoisomers of these two compounds
and determined their potency and efficacy (Table 1). Compared to (S)ꢀ1 and reference compounds,
the four stereoisomers of 3 exhibited lower activity in terms of efficacy and/or potency on PPARα
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
I
II
and PPARγ with (R ,R )ꢀ3 being completely inactive on the former. Both receptors displayed a
significant stereoselectivity toward the four stereoisomers whose activity showed the following rank
I
II
I
II
I
II
I
II
order: (S ,S ) > (S ,R ) > (R ,S ) > (R ,R ). With regard to compound 6, the S enantiomer was more
potent and effective than the R enantiomer on PPARα; however, both acted as partial agonists
compared to (S)ꢀ1 and reference compound. The behavior of these two enantiomers was slightly
different on PPARγ; in fact, they were less active than rosiglitazone but showed an efficacy
comparable with (S)ꢀ1. In addition, (S)ꢀ6 displayed potency about 7ꢀ and 5ꢀfold higher than (R)ꢀ6
(0.12 vs. 0.89 ꢀM) and lead compound (S)ꢀ1 (0.12 vs. 0.55 ꢀM), respectively. Even though (S)ꢀ1
35
and (R)ꢀ1 were completely inactive on PPARδ, we decided to test (S)ꢀ6 on this subtype, but also
in this case no activation was observed (data not shown). Therefore, given that (S)ꢀ1 and (R)ꢀ1 did
35
not even affect the transcriptional activity of other nuclear receptors, it is reasonable to assert that
(S)ꢀ6 is a PPARα/γ selective ligand.
On the basis of these results, we decided to investigate the interaction mode of (S)ꢀ6 (hereafter
LJ570) with both PPARα and PPARγ and, for this reason, we tried to resolve the Xꢀray structures
of the corresponding complexes.
Binding of LJ570 to PPARα and PPARγ
Unfortunately, we could not resolve the crystal structure of LJ570 in the complex with PPARα,
therefore, in order to gain insight into the possible binding mode, we undertook docking
experiments whose results are reported in Supporting Information.
ACS Paragon Plus Environment

Page 9 of 74
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
On the contrary, two structures of PPARγꢀLBD with the ligand LJ570 were collected and solved
from crystals obtained with different soaking times and ligand concentrations. In the first structure
(shortꢀterm soaking time and low ligand concentration) the ligand occupies the region between the
helix 3 (H3) and the βꢀsheet, in the canonical way as other known partial agonists and (R)ꢀ1 (Fig.
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
21,23,24,26,35
4A).
The carboxylate group interacts with the NH of S342 via Hꢀbonds mediated by one
molecule of water. In the second structure (longꢀterm soaking time and high ligand concentration),
LJ570 interacts directly with Y473 of helix 12 (H12) and the two histidines of the canonical triad
(H323 and H449) through Hꢀbonds made by its carboxylate and occupies the diphenyl pocket,
shifting the sideꢀchain of F282, with one of its diphenyl groups in the same position assumed by the
35
ligand (S)ꢀ1 (Fig. 4B). The two different positions of the ligand in the PPARγꢀLBD obtained from
different soaking times allow to suppose the existence of an equilibrium between two binding
modes, both belonging to the canonical site, which resemble those of the two simultaneously bound
3
0
molecules of an isoꢀpropyl derivative of (S)ꢀ1 (PDB code 4E4K). Similar binding modes were
also observed in the crystal structure of PPARγ with the partial agonist 2ꢀchloroꢀNꢀ(3ꢀchloroꢀ4ꢀ(5ꢀ
chlorobenzo[d]thiazolꢀ2ꢀylthio)phenyl)ꢀ4ꢀ(trifluoromethyl)benzenesulfonamide (T2384, PDB code
38,39
3K8S), also bound with two molecules in the canonical site of the LBD.
This latter compound,
which displays unique biochemical and cellular activities together with an attractive in vivo profile,
is considered a prototype for a novel class of PPARγ ligands with better therapeutic properties for
treatment of type 2 diabetes. This prompted us to further investigate the biological activity of
LJ570. To this end, we first determined whether this ligand could induce the differentiation of 3T3ꢀ
L1 adipocytes, as a read out predictive of one of the major side effects of PPARγ full agonists like
rosiglitazone (i.e., adipogenesisꢀmediated weight gain).
Effects of LJ570 on gene expression
To determine whether the particular interaction between LJ570 and PPARγ ligandꢀbinding domain
could affect its adipogenic potential in vitro, 3T3ꢀL1 fibroblasts were induced to differentiate to
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 10 of 74
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
adipocytes with medium containing both LJ570 and insulin. Rosiglitazone was used as reference
compound. Interestingly, LJ570 did not activate the PPARγ dependent program of adipocyte
differentiation in 3T3ꢀL1 cells. We did not observe any significant increased expression of fatty
acid binding protein 4 (Fabp4), adiponectin (Adipoq), insulinꢀsensitive glucose transporter 4
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
(
Glut4), class B scavenger receptor (Cd36), phosphoenolpyruvate carboxykinase 1 (Pck1), glycerol
kinase (Gk), isocitrate dehydrogenase subunit alpha (Idh3a), and succinylꢀCoA ligase subunit alpha
Suclg1) (Fig. 1A of Supporting Information). In accordance with these results, treatment of 3T3ꢀ
(
L1 cells with LJ570 induced only very little lipid accumulation even at the highest tested
concentration as compared to rosiglitazone, further indicating that LJ570 features much lower
adipogenic activity than a full PPARγ agonist (Fig. 5, A and B). The activity of LJ570 on PPARα
was determined in vitro by treating Hepa1ꢀ6 cells with three different concentrations of the ligand,
using Wyꢀ14,643 as reference compound. In this case, as expected, LJ570 increased the expression
of three different PPARαꢀtarget genes: acylꢀCoenzyme A oxidase 1 (Acox1), carnitine
palmitoyltransferase 1a (Cpt1a) and acylꢀCoenzyme A dehydrogenase, medium chain (Acadm)
(Fig. 1B of Supporting Information). The effects on the expression of these three genes, which are
essential for the peroxisomal and mitochondrial βꢀoxidation of fatty acids, allow to hypothesize a
beneficial pharmacological profile of this compound on lipid metabolism.
Coꢀregulator recruitment by Fluorescence Resonance Energy Transfer assay
To further confirm the ability of LJ570 to interact with PPARγ, in spite of the lacking activation of
some PPARγ target genes, we investigated the effects of its interaction on the coactivator
recruitment from the ligandꢀactivated receptor. To this end, we measured coactivator recruitment by
FRET assay both in the absence and in the presence of rosiglitazone. The assay confirmed that
LJ570 is able to recruit the PGC1α coactivator peptide on PPARγ (Fig. 6), although only at
concentration higher than 1 ꢁM, thus providing an additional evidence of its ability to interact with
the receptor. Interestingly, in the presence of three different concentrations of rosiglitazone (0.37, 1,
ACS Paragon Plus Environment

Page 11 of 74
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
and 10 microM), LJ570 induced a synergistic recruitment of PGC1α on the receptor (Fig. 6). This
result prompted us to investigate the possibility that this ligand could bind, in the presence of
rosiglitazone, to an alternate binding site. In fact, recently, Hughes et al. demonstrated that some
synthetic PPARγ ligands not only bind to the canonical pocket at high affinity but can also bind to
an alternate binding site at lower affinity, which is a surface exposed region capped by the
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
40
conformationally flexible ꢂꢀloop located between helix 2′ and helix 3, and affect PPARγ function.
Ligand binding to this latter site does not compete with endogenous ligand binding to the canonical
site, indicating it is an alternate site. To determine if the origin of the synergistic recruitment of
PGC1α on PPARγ was due to the simultaneous binding of rosiglitazone and LJ570 to the canonical
and alternate sites, respectively, we performed a cotransfection assay by preincubating cells with 7
(GW9662, Fig. 7, upper part) before treating cells with ligand. 7 is a commercially available PPARγ
antagonist that binds to the canonical site of PPARγ through a covalent thioether bond with C285
4
1
(
Fig. 4C). It was previously showed that 7 blocks the binding of many PPARγ ligands, including
40
rosiglitazone, to the canonical site, whereas no effect has been reported on the alternate site. In our
experiment, as shown in Fig. 7A, the presence of 7 induced only a slight decrease in the potency of
LJ570 allowing to suppose its binding to the alternate site with affinity similar to that for the
canonical site. On the contrary, 7 blocked the action of rosiglitazone at concentrations up to 100 nM
and only higher concentrations elicited an increase in PPARγ transactivation (Fig. 7B). In this case,
the activation from rosiglitazone occurs through the alternate site too even though by a lower
40
affinity binding as previously reported. It is important to highlight that the two different positions
which LJ570 can assume, as reported above, in the PPARγꢀLBD are almost symmetrical, with a
diphenyl group occupying, in both cases, the central region of LBD in front of C285 of H3 (Fig.
4C). Both positions show a common binding region with the antagonist 7 (Fig. 4C), covalently
bound to C285 in the canonical site. Therefore, it is logical to hypothesize that the residual activity
of LJ570, observed in the presence of 7, might be possible only if the ligand occupies the alternate
binding site. On the other hand, this interaction would also explain the effects showed in Fig. 6: the
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 12 of 74
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
higher the concentration of rosiglitazone, the lower the synergistic recruitment of PGC1α on the
receptor; that is, at higher concentration, rosiglitazone would also occupy the alternate site limiting
or preventing the accommodation of LJ570.
As a whole, it seems reasonable to assert LJ570 is able to interact with PPARγ through binding two
molar equivalents: one to the canonical binding pocket at high affinity, where the ligand is in
equilibrium between two different positions (Fig. 4C), and a second to the alternate site at a lower
but potentially pharmacologically relevant affinity. However, even though the interaction with the
alternate site occurs only at higher concentrations, its functional effects should not be overlooked.
In fact, for in vivo animal model studies, PPARγ ligands are often administered several times a day
over many months. In these cases, it is possible that ligand concentrations could be high enough to
elicit an alternate site functional effect through binding a second molecule of the ligand. Although
the pharmacological utility of the PPARγ alternate site remains unclear, it is possible that ligands
binding this site might favor a correct balance of coactivator and corepressor binding or synergize
with endogenous canonical ligands to provide a unique functional activity profile. As a matter of
fact, it is logical to presume that the slight decrease on PPARγ activation shown by LJ570 after
pretreatment with 7 in the luciferase reporter assay is the result of the occupation of the alternate
site. In order to support this hypothesis, we decided to perform docking experiments to investigate
on putative binding mode of LJ570 in the 7ꢀbound PPARγꢀLBD.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Docking experiments
To further examine the possibility that LJ570 can bind to the alternate site in 7ꢀbound PPARγꢀLBD,
42
a molecular docking simulation was performed using the GOLD Suite docking package. The
crystal structure of human 7ꢀbound PPARγꢀLBD was retrieved from the RCSB Protein Data Bank
(
PDB code 3B0R). However, this structure lacks the ꢂꢀloop, a flexible region between H2′ and H3
residues 262ꢀ274), which comprises part of the alternate site and synergistically affects PPARγ
(
43
43e,44
structure and function. Moreover, recent studies
suggest that the ꢂꢀloop may adopt different
ACS Paragon Plus Environment

Page 13 of 74
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
“
closed” conformations according to the ligands that occupy the alternate binding site to achieve an
optimal fit of the PPARγꢀLBD with the ligand and to inhibit S273 phosphorylation by blocking
Cdk5 from binding to the PPARγ LBD. For these reasons, we systematically scrutinized the PDB
structures of PPARγ complexed with ligands occupying the alternate site in order to assess which
conformation of the ꢂꢀloop should be used to assemble the final receptor model. Finally, the
PPARγ/S35 complex (PDB code 5GTO) was chosen as the best structure available in which the
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
43e
electron density for the flexible ꢂꢀloop was clearly observed. So, the starting 7ꢀbound PPARγ
crystal structure was modified to include the 12 missing residues (residues 262ꢀ274) in a surface
loop structure (see Experimental procedures).
From results of our docking simulation, LJ570 in the presence of 7 is predicted to occupy the
39
alternate ligandꢀbinding site (Fig. 8A), which was previously identified by Hughes et al. and
43e,f
visualized by Xꢀray crystallography.
No steric clash is observed between LJ570 in the alternate
site and 7 and, thus, both ligands can be accommodated simultaneously into the PPARγꢀLBD. In
Fig. 8B the overlap between the two positions of LJ570, in the canonical site, and the docked LJ570
occupying the alternate site is shown.
The interactions between LJ570 and the PPARγꢀLBD involve one Hꢀbond and several hydrophobic
interactions. Specifically, one of the carboxylate oxygens of LJ570 forms a Hꢀbond with the
backbone nitrogen of S342 on the βꢀsheet with a distance of 3.3 Å. The hydrophobic interactions of
LJ570 with the PPARγꢀLBD can be divided in two groups. The first group occurs between the oxyꢀ
diphenyl group of LJ570 and V339, I341, M348 on the βꢀsheet, L330 on helix 5, and M364 on
helix 7 in the binding pocket of known partial agonists. The second group of hydrophobic
interactions involves the remaining diphenyl moiety of LJ570 and residues F264, I267, P269 (part
of the ꢂꢀloop), R280, G284, and F287 (helix 3) in the alternate site. Interestingly, the terminal
phenyl ring of LJ570 forms stacking interactions with F287 from helix H3 and P269 of the ꢂ loop.
Notably, when we superimposed our LJ570ꢀbound PPARγꢀLBD complex onto the structures of S35
(PDB: 5GTO) and SR1664 (PDB: 5DWL) that bind to the PPARγ alternate site (Fig. 8C), we
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 14 of 74
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
observed that the three compounds occupy a similar position in the LBD, contacting the helix H3
and the ꢂꢀloop residues. While this work was in preparation, Kojetin and coworkers released a
crystal structure of PPARγ bound to edaglitazone at the canonical pocket and nonanoic acid coꢀ
45
bound to the alternate site (PDB code: 5UGM). Interestingly, the diphenyl moiety of LJ570
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
overlaps with the crystallized binding mode of nonanoic acid, whereas the carboxylate groups of
both ligands are engaged in Hꢀbonding to the S342 backbone (Fig. 8D).
Effects of LJ570 on Cdk5ꢀmediated phosphorylation at S273 of PPARγ
The different PPARγ binding modes of the full and partial agonists do not fully explain their ability
21
to exert antidiabetic properties. In 2010, Choi et al. described a new antidiabetic mechanism of
PPARγꢀtargeted drugs which depends upon the inhibition of PPARγ S273 phosphorylation by
cyclin dependent kinase 5 (Cdk5). S273 phosphorylation results in the dysregulation of a subset of
PPARγ target genes that are known to be associated with tissue insulin sensitization, whereas its
inhibition maintains unchanged the transcription of several of these genes and the correlated insulin
sensitization. However, the Cdk5ꢀmediated phosphorylation of PPARγ does not have any influence
on its transactivation activity and its role in adipocyte differentiation, which means that any ligand
acting as inhibitor of this phosphorylation mechanism, would be able to exert antidiabetic effects.
Therefore, this new mechanistic scenario might explain why some PPARγ partial agonists or
modulators can exhibit antidiabetic effects similar to TZDs and other full agonists, without showing
their typical side effects; that is, it has been proposed that the antidiabetic activity of PPARγ ligands
occurs through the inhibitory effect on Cdk5ꢀmediated phosphorylation of PPARγ, whereas the
different potency and efficacy as agonists could be related to the different sideꢀeffect profiles. For
this reason, we assessed the ability of LJ570 to inhibit Cdk5ꢀmediated phosphorylation of PPARγ in
vitro. For detecting and quantifying the possible inhibition rate, an ELISA (enzymeꢀlinked
immunosorbent assay) protocol was optimized. As shown in Fig. 9, LJ570 effectively inhibited
S273 phosphorylation with higher concentrations being more efficacious than rosiglitazone. This
ACS Paragon Plus Environment

Page 15 of 74
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
could be explained by taking into account the capability of LJ570 to assume two different positions:
the former in the region between H3 and the βꢀsheet, in the canonical way as other known partial
21,23,24,26
agonists and selective PPARγ modulators,
the latter in the alternate site by accommodating
in a hydrophobic region between helices 2’ and 3, the ꢂ ꢀloop and the β3–β4 strands. In both cases,
LJ570 would stabilize the βꢀsheet region of the PPARγꢀLBD thus shielding S273 from Cdk5ꢀ
mediated phosphorylation due to the location of S273 adjacent to the first βꢀstrand of PPARγ,
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
21
which is part of the Cdk5 recognition site.
Effect of LJ570 on adipocyte glucose uptake
We finally tested the insulinꢀsensitizing effect of LJ570 by investigating its ability to increase
glucose uptake in 3T3ꢀL1 adipocytes. In fact, for some selective PPARγ modulators it has been
demonstrated the possibility to dissociate the effects of PPARγ activation on adipogenesis from
46
glucose uptake stimulation. Therefore, we treated differentiated 3T3ꢀL1 adipocytes with LJ570
and measured glucose uptake. As shown in Fig. 10A, LJ570 significantly increased the insulinꢀ
stimulated glucose uptake suggesting that its binding to PPARγ results in a biological effect that
may improve insulin sensitivity. In spite of its potency as a PPARγ agonist, LJ570 stimulated
glucose uptake only at concentration as high as 10 ꢀM. At this concentration it is reasonable to
suppose the interaction of PPARγ with two molecules of the ligand, the former in the canonical site
and the latter in the alternate one, which cooperatively produce the resulting effect. However, to
further confirm that this effect is also due to the interaction of LJ570 with the alternate site, we
repeated the glucose uptake assay in the presence of 0.3 ꢁM rosiglitazone and 10 ꢁM LJ570. The
concentration of both ligands was chosen on the basis of the additive effect shown, at these doses,
on the recruitment of the PGC1α coactivator peptide (Fig. 6). As expected, the results of this new
experiment demonstrated that the combined treatment significantly increased the insulinꢀstimulated
glucose uptake compared to the treatment with either drug administered alone (Fig. 10B).
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 16 of 74
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
At this point, we decided to verify the presence of a potential weak effect of LJ570 on the
expression of the PPARγ target genes involved in the PPARγꢀdependent program of adipocyte
differentiation also at 5 and 10 ꢀM given that previous experiments were conducted at
concentration up to 2.5 ꢀM. Again, even at this concentration, LJ570 did not induce any significant
increased expression of Fabp4, Adipoq, Glut4, and Pck1 (Fig. 1C of Supporting Information). In
accordance with these results, treatment of 3T3ꢀL1 cells with 10 ꢀM LJ570 still induced much
lower lipid accumulation as compared to rosiglitazone (Fig. 1D of Supporting Information).
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Conclusion
In the present study, we investigated the effects resulting from the chemical modification of the
previously reported PPAR ligand (S)ꢀ1. This allowed the identification of the new compound LJ570
able to potently activate both PPARα and γ subtypes as a partial agonist. However, in vitro
experiments in 3T3ꢀL1 preꢀadipocyte cell line showed that LJ570 did not activate any of the PPARγ
target genes involved in the PPARγꢀdependent program of adipocyte differentiation. In accordance
with these results, treatment of 3T3ꢀL1 cells with LJ570 induced only very little lipid accumulation
which is one of the major side effects of full agonists like rosiglitazone. However, an acute
treatment of 3T3ꢀL1 adipocytes with LJ570 increased the insulinꢀstimulated glucose uptake. In
addition, LJ570 effectively inhibited Cdk5ꢀmediated phosphorylation of PPARγ in vitro which has
been recently described as a new mechanism through which some PPARγꢀtargeted partial agonists
or antagonists are able to maintain their antidiabetic activity without showing the typical side
effects of full agonists. On the contrary, LJ570 increased the expression of three typical PPARα
target genes which are essential for the hepatic peroxisomal and mitochondrial βꢀoxidation of fatty
acids, allowing to hypothesize a beneficial pharmacological profile of this compound on lipid
metabolism.
As far as we know, this ligand is the first PPARα/γ dual agonist reported to inhibit S273ꢀ
phosphorylation of PPARγ and, therefore, may represent the potential lead of a new class of drugs
ACS Paragon Plus Environment

Page 17 of 74
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
with better pharmacological properties in the treatment of dyslipidemic type 2 diabetes.
Interestingly, Xꢀray studies of the complex PPARγ/LJ570 allowed to collect and solve two different
structures from crystals obtained with different soaking times and ligand concentrations: the former,
typical of the partial agonists and characterized by accommodation into the region between helix 3
and the βꢀsheet; the latter, in which the ligand is situated more deeply in the PPARγꢀLBD assuming
a position more similar to full agonists. It is reasonable to suppose that the activity of the ligand is
the result of an equilibrium between these two different binding modes, both belonging to the
canonical pocket. However, LJ570 seems also to be able to bind an alternate site as demonstrated
from its activity on PPARγ even in the presence of a specific irreversible antagonist and the
capability to induce a synergistic recruitment of PGC1α on PPARγ and an additive effect on
glucose uptake in the presence of rosiglitazone. Even though the interaction of LJ570 with this
alternate site occurs only at high, but pharmacologically relevant, concentration, it could account
for its activity profile. In fact, although the pharmacological utility of the PPARγ alternate site
remains unclear, it is possible that ligands bound to this site contribute to regulate receptor activity
allowing to fine tune the functional response of PPARγ without competing, but rather synergizing,
with endogenous canonical ligands. Thus, the development of a new generation of PPARγ drugs as
alternate modulators targeting nonꢀcanonical sites of the ligand binding domain could be a
promising approach to address selectivity concerns and therapeutic profile.
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 18 of 74
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
EXPERIMENTAL SECTION
Chemical Methods. Column chromatography was performed on ICN Silica Gel 60 Å (63−200 ꢁm)
as a stationary phase. Melting points were determined in open capillaries on a Gallenkamp
electrothermal apparatus and are uncorrected. Mass spectra were recorded on an HP GC/MS 6890ꢀ
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
5973 MSD spectrometer, electron impact 70 eV, equipped with an HP chemstation or an Agilent
1
LC/MS 1100 series LC/MSD Trap System VL spectrometer, electrospray ionization (ESI). Hꢀ
NMR spectra were recorded in CDCl (the use of other solvents is specified) on a VarianꢀMercury
3
300 (300 MHz) spectrometer at room temperature. Chemical shifts are expressed as parts per
million (δ). The purity of all tested compounds was > 95%, as confirmed by combustion analysis
carried out with a Eurovector Euro EA 3000 model analyzer. For optical isomers, MS and NMR
spectra are reported only for the racemate or one of the stereoisomers. Optical rotations were
measured with a PerkinꢀElmer 341 polarimeter at room temperature (20 °C): concentrations are
expressed as grams per 100 mL. The enantiomeric excesses of the final acids were determined by
HPLC analysis on a Chiralcel OD or AD column (4.6 mm i.d. × 250 mm, Daicel Chemical
Industries, Ltd., Tokyo, Japan). Analytical liquid chromatography was performed on a PE
chromatograph equipped with a Rheodyne 7725i model injector, a 785A model UV/vis detector, a
series 200 model pump, and an NCI 900 model interface. Chemicals were from Aldrich (Milan,
Italy), AlfaAesar (Karlsruhe, Germany), or Acros (Milan, Italy) and were used without any further
purification.
Ethyl 2ꢀ(1,1'ꢀdiphenylꢀ4'ꢀolꢀ4ꢀoxy)ꢀ3ꢀphenylꢀpropanoate (2a).
A solution of diisopropylazodicarboxylate (DIAD, 5 mmol) in anhydrous THF (10 mL) was added
dropwise to an iceꢀbath cooled mixture of (R,S)ꢀethyl phenyllactate (5 mmol), 4,4'ꢀ
dihydroxydiphenyl (5 mmol), and triphenylphosphine (5 mmol) in anhydrous THF (20 mL). The
2
reaction mixture was stirred at room temperature overnight, under N atmosphere. THF was
evaporated in vacuo, and the residue was chromatographed on silica gel column (petroleum
ether/ethyl acetate 9:1 as eluent) affording the desired compound as a white solid in 41% yields.
ACS Paragon Plus Environment

Page 19 of 74
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
+
1
GCꢀMS m/z (%): 362 (M , 100), 289 (9), 186 (65). HꢀNMR (CDCl
3
): δ 1.20 (t, 3H, J=7.15, CH
ꢀO), 4.80ꢀ4.84 (dd, 1H, CH), 5.03 (s, 1H, OH,
O exchanged), 6.82ꢀ6.88 (m, 4H, aromatics), 7.24ꢀ7.40 (m, 9H, aromatics).
Ethyl 2ꢀ[4ꢀ[4ꢀ(1ꢀethoxyꢀ1ꢀoxopropanꢀ3ꢀphenylꢀ2ꢀyl)oxyphenyl]phenoxy]ꢀ3ꢀphenylꢀpropanoate
2b).
3
),
3
2 2
.21ꢀ3.30 (m, 2H, CH ꢀPh), 4.19 (q, 2H, J=7.15, CH
D
2
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
(
The title compound was obtained as a white solid in 30% yield by using the same procedure
reported above starting from 2a and another equivalent amount of (R,S)ꢀethyl phenyllactate. GCꢀ
+
1
MS m/z (%): 538 (M , 100) 465 (8), 362 (6). HꢀNMR (CDCl ): δ 1.20 (t, 6H, J=7.15, 2 CH ), 3.21ꢀ
3
3
3
4
2
.30 (m, 4H, 2 CH
H, aromatics), 7.24ꢀ7.40 (m, 14H, aromatics).
ꢀ[4ꢀ[4ꢀ(1ꢀCarboxyꢀ2ꢀphenylethoxy)phenyl]phenoxy]ꢀ3ꢀphenylꢀpropanoic acid (2).
2 2
ꢀPh), 4.19 (q, 4H, J=7.15, 2 CH ꢀO), 4.80ꢀ4.84 (m, 2H, 2 CH), 6.82ꢀ6.88 (m,
A solution of 2b (1 mmol) in THF (10 mL) and 1 N NaOH (10 mL) was stirred at room temperature
for 3 h. The organic layer was removed under reduced pressure, and the residue was acidified with
2
N HCl and extracted with Et
2 2 4
O. The organic layer was dried over Na SO and evaporated to
dryness affording the final acid as a white solid which was crystallized from AcOEt/hexane (55%
+
1
yields). Mp: 138ꢀ140 °C; GCꢀMS, (methyl ester) m/z (%): 510 (M , 49) 451 (4), 121 (100). Hꢀ
NMR (CDCl ): δ 3.28 (d, 4H, J=6.32, 2 CH ), 4.83 (t, 2H, J=6.32, 2 CH), 5.60ꢀ6.60 (bb, 2H,
COOH, D O exchanged), 6.84ꢀ6.87 (m, 4H, aromatics), 7.22ꢀ7.36 (m, 14H, aromatics). Anal. Calcd
: C, 74.67; H, 5.43. Found: C, 75.02; H, 5.81.
3
2
2
26 6
for C30H O
General Procedure for Synthesis of (R,S)ꢀ, (R)ꢀ or (S)ꢀethyl 2ꢀ(4ꢀ(4,4,5,5ꢀtetramethylꢀ1,3,2ꢀ
dioxaborolanꢀ2ꢀyl)phenoxy)ꢀ3ꢀphenylꢀpropanoate ((R,S)ꢀ3a, (R)ꢀ3a, (S)ꢀ3a).
A solution of diisopropylazodicarboxylate (DIAD, 2 mmol) in anhydrous THF (7 mL) was added
dropwise to an iceꢀbath cooled mixture of (R,S)ꢀ, (S)ꢀor (R)ꢀethyl phenyllactate (2 mmol), 4ꢀ
(4,4,5,5ꢀtetramethylꢀ1,3,2ꢀdioxaborolanꢀ2ꢀyl)phenol (2 mmol), and triphenylphosphine (3 mmol) in
anhydrous THF (7 mL). The reaction mixture was stirred at room temperature for 48 h, under N2
atmosphere. Then, the solvent was evaporated in vacuo and the residue was chromatographed on
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 20 of 74
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
silica gel column (nꢀhexane/ethyl acetate 90:10 as eluent), affording a colorless oil. GCꢀMS m/z
+
1
(%): 396 (M , 100), 223 (56), 177 (69), 135 (94). HꢀNMR (CDCl
CH ꢀO), 1.31 (s, 12H, 4 CH CꢀO), 3.23ꢀ3.29 (m, 2H, ArꢀCH ꢀCH); 4.15 (q, 2H, J=7.15, CH
O), 4.81ꢀ4.86 (m, 1H, CH); 6.81ꢀ6.85, 7.19ꢀ7.30 and 7.68ꢀ7.71 (m, 9H, aromatics).
R,S)ꢀ3a. Yield: 37%. (R)ꢀ3a. Yield: 17%. (S)ꢀ3a. Yield: 20%.
3
): δ 1.16 (t, 3Η, J=7.15, CH
3
2
ꢀ
ꢀ
2
3
2
3
ꢀCH
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
(
I
I
II II
(
R S , R S )ꢀ Ethyl 2ꢀ(diphenylꢀ4ꢀylꢀoxy)ꢀ3ꢀ[4’ꢀ(1ꢀethoxycarbonylꢀ2ꢀphenylꢀethoxy)diphenylꢀ
I
I
II II
4ꢀyl]propanoate ((R S , R S )ꢀ3b).
2 3
Compound (R,S)ꢀ3a (0.25 mmol), Cs CO (0.25 mmol), and ethyl 2ꢀ(4ꢀphenylphenoxy)ꢀ3ꢀ(4ꢀ
bromophenyl)propanoate 6e (0.16 mmol) were dissolved in anhydrous toluene (10 mL) at 95 °C,
under N atmosphere. After 0.5 h, Pd(PPh (0.0075 mmol) was added. The reaction mixture was
2
3 4
)
stirred for 24 h at 95 °C and then quenched with 1 N HCl (2 mL) and ethyl acetate (2 mL). The
suspension was filtered through a Celite pad to remove the catalyst, and the filtrate was washed
with saturated NaHCO
dissolved in ethyl acetate, washed with brine, dried over Na
to give an oily residue which was chromatographed on a silica gel column (nꢀhexane/ethyl acetate,
0:10, as eluent), affording the desired compound as a colorless oil. Yield: 39%. ESIꢀMS (positive)
3
solution and brine. The filtrate was concentrated under reduced pressure,
2
SO4, filtered, and evaporated to dryness
9
+
1
m/z (%): 637 (M+Na , 27), 569 (34), 391 (100). HꢀNMR (CDCl
.24ꢀ3.30 (m, 4H, 2 CH ꢀAr), 4.15ꢀ4.24 (m, 4H, 2 CH ꢀO), 4.81ꢀ4.86 (m, 2H, 2 CH), 6.87ꢀ6.93 and
7.29ꢀ7.52 (m, 22H, aromatics).
Synthesis of the Stereoisomers of (R S , R S )ꢀ3b.
The title compounds were prepared by the same procedure reported above starting from (R)ꢀ3a or
3
): δ 1.16ꢀ1.23 (m, 6H, 2 CH
3
),
3
2
2
I
I
II II
I
II
(
S)ꢀ3a and (R)ꢀ or (S)ꢀethyl 2ꢀ(4ꢀphenylphenoxy)ꢀ3ꢀ(4ꢀiodophenyl)propanoate 6g. (S R )ꢀ3b: 39%
I
II
I
II
I II
yield; (R S )ꢀ3b: 10% yield; (S S )ꢀ3b: 46% yield; (R R )ꢀ3b: 5% yield.
I
I
II II
General Procedure for Synthesis of (R S , R S )ꢀ2ꢀ(diphenylꢀ4ꢀylꢀoxy)ꢀ3ꢀ[4’ꢀ(1ꢀcarboxyꢀ2ꢀ
I
I
II II
phenylethoxy)diphenylꢀ4ꢀyl]propanoic acid ((R S , R S )ꢀ3) and its Stereoisomers.
ACS Paragon Plus Environment

Page 21 of 74
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
To a stirred solution of the corresponding diethyl ester (0.1 mmol) in THF (5 mL), 1 N NaOH (5
mL) was added. The reaction mixture was stirred overnight at room temperature. Then, the solvent
was evaporated in vacuo and the aqueous phase was acidified with 2 N HCl and extracted with
2 2 4
Et O. The combined organic layers were washed with brine, dried over Na SO and evaporated to
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
dryness affording the title diacids as white solids, which were recrystallized from CHCl
3
/nꢀhexane.
): δ 3.28ꢀ
1
ESIꢀMS (negative) m/z (%): 557 (MꢀH, 1), 409 (8), 387 (100), 239 (51). HꢀNMR (CDCl
3
2
3.34 (m, 4H, 2 CH ꢀAr), 4.84ꢀ4.97 (m, 2H, 2 CH), 6.87ꢀ6.97 and 7.27ꢀ7.52 (m, 22H, aromatics).
The enantiomeric (ee) and diastereomeric excesses (de) of the stereoisomers were determined by
®
using a CSPꢀHPLC column CHIRALCEL AD; lambda = 254 nm; 1 mL/min; nꢀhexane/isoꢀ
I
I
II II
propanol/TFA 80:20:0.2. (R S , R S )ꢀ3: 89% yield, mp 163ꢀ165 °C. Anal. Calcd for C36
H
30
O
6
: C,
7.40; H, 5.41. Found: C, 77.08; H, 5.42. (S R )ꢀ3: 88% yield, mp 198ꢀ200 °C, ee > 99%, de 87%;
α] = + 31.5 (c 1.0, MeOH). Anal. Calcd for C36 : C, 77.40; H, 5.41. Found: C, 77.17; H,
5.40. (R S )ꢀ3: 26% yield, mp 169ꢀ170 °C, ee > 99%, de 90%; [α] = ꢀ 32.4 (c 0.7, MeOH). Anal.
I
II
7
[
D
30 6
H O
I
II
D
I
II
Calcd for C36
H
30
O
6
: C, 77.40; H, 5.41. Found: C, 77.88; H, 5.62. (S S )ꢀ3: 46% yield, mp 205ꢀ207
= + 37.5 (c 1.0, MeOH). Anal. Calcd for C36 : C, 77.40; H, 5.41.
Found: C, 77.33; H, 5.52. (R R )ꢀ3: 88% yield, mp 164ꢀ167 °C, ee > 99%, de 90%; [α] = ꢀ 38.7 (c
0.8, MeOH). Anal. Calcd for C36 : C, 77.40; H, 5.41. Found: C, 77.11; H, 5.75.
°C, ee > 99%, de 92%; [α]
D
30 6
H O
I
II
D
30 6
H O
General Procedure for Synthesis of ethyl 2ꢀ(4ꢀ(4,4,5,5ꢀtetramethylꢀ1,3,2ꢀdioxaborolanꢀ2ꢀ
yl)phenoxy)propanoate (4a) and ethyl (4ꢀ(4,4,5,5ꢀtetramethylꢀ1,3,2ꢀdioxaborolanꢀ2ꢀ
yl)phenoxy)acetate (5a).
4ꢀ(4,4,5,5ꢀTetramethylꢀ1,3,2ꢀdioxaborolanꢀ2ꢀyl)phenol (0.9 mmol) was added to a suspension of
cesium carbonate (1.8 mmol) in anhydrous THF (10 mL). After 0.5 h a solution of ethyl
bromopropanoate or ethyl bromoacetate in anhydrous THF (5 mL) was added dropwise and the
reaction mixture was refluxed for 6 h, under N atmosphere. Then, the solvent was evaporated in
2
vacuo and the residue dissolved in diethyl ether and washed twice with 1 N HCl and brine, dried
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 22 of 74
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
over Na SO and evaporated to dryness. The oily residue was chromatographed on a silica gel
2
4
column (nꢀhexane/ethyl acetate 90:10 as eluent) affording a colorless oil.
Ethyl 2ꢀ(4ꢀ(4,4,5,5ꢀtetramethylꢀ1,3,2ꢀdioxaborolanꢀ2ꢀyl)phenoxy)propanoate (4a).
+
1
Yield: 37%. GCꢀMS m/z (%): 320 (M , 41), 247 (100). HꢀNMR (CDCl ): δ 1.23 (t, 3Η, J=7.15,
3
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
CH
q, 1H, J=6.97, CH), 6.84ꢀ6.88 and 7.71ꢀ7.74 (m, 4H, aromatics).
Ethyl (4ꢀ(4,4,5,5ꢀtetramethylꢀ1,3,2ꢀdioxaborolanꢀ2ꢀyl)phenoxy)acetate (5a).
3 2 3 3 2
ꢀCH ꢀO); 1.32 (s, 12H, 4 CH C); 1.61 (d, 3H, J=6.97, CH CH); 4.20 (q, 2H, J=7.15, CH ); 4.81
(
+
1
Yield: 56%. GCꢀMS m/z (%): 306 (M , 100). HꢀNMR (CDCl ): δ 1.28 (t, 3Η, J=7.15, CH ꢀCH ꢀ
3
3
2
O); 1.33 (s, 12H, 4 CH
3
C); 4.26 (q, 2H, J=7.15, CH
3
ꢀCH
2
2
ꢀO); 4.64 (s, 2H, OꢀCH ꢀCO), 6.88ꢀ6.91
and 7.74ꢀ7.77 (m, 4H, aromatics).
General Procedure for Synthesis of ethyl 2ꢀ(diphenylꢀ4ꢀylꢀoxy)ꢀ3ꢀ[4’ꢀ(1ꢀethoxycarbonylꢀ
ethoxy)diphenylꢀ4ꢀyl]propanoate (4b) and 2ꢀ(diphenylꢀ4ꢀylꢀoxy)ꢀ3ꢀ[4’ꢀ(ethoxycarbonylꢀ
methoxy)diphenylꢀ4ꢀyl]propanoate (5b).
2 3
Compound 4a or 5a (0.34 mmol), Cs CO (0.51 mmol), and 6e (0.34 mmol) were dissolved in
anhydrous toluene (8 mL) at 95 °C, under N atmosphere. After 0.5 h, Pd(PPh ) (0.01 mmol) was
2
3 4
added. The reaction mixture was stirred for 20 h at 95 °C and then quenched with 1 N HCl (3 mL)
and ethyl acetate (3 mL). The suspension was filtered through a Celite pad to remove the catalyst,
and the filtrate was washed with saturated NaHCO
3
solution and brine. The filtrate was
concentrated under reduced pressure, dissolved in ethyl acetate, washed with brine, dried over
Na SO filtered, and evaporated to dryness to give an oily residue which was chromatographed on a
2
4,
silica gel column (nꢀhexane/ethyl acetate, 90:10, as eluent), affording the desired compounds as
colorless oils.
Ethyl 2ꢀ(diphenylꢀ4ꢀylꢀoxy)ꢀ3ꢀ[4’ꢀ(1ꢀethoxycarbonylꢀethoxy)diphenylꢀ4ꢀyl]propanoate (4b).
+
1
Yield: 41%. ESIꢀMS (positive) m/z (%): 561 (M+Na , 100); HꢀNMR (CDCl ): δ 1.19ꢀ1.28 (m,
3
6
H, 2 CH ꢀCH ), 1.64 (d, 3H, J=6.88, CHꢀCH ), 3.23ꢀ3.36 (m, 2H, CH ꢀAr), 4.18ꢀ4.27 (m, 4H, 2
3
2
3
2
ACS Paragon Plus Environment

Page 23 of 74
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
CH ꢀO), 4.77 (q, 1H, J=6.88, CHꢀCH ), 4.83ꢀ4.87 (dd, 1H, CH-CH ), 6.91ꢀ6.94 and 7.29ꢀ7.53 (m,
2
3
2
17H, aromatics).
Ethyl 2ꢀ(diphenylꢀ4ꢀylꢀoxy)ꢀ3ꢀ[4’ꢀ(ethoxycarbonylꢀmethoxy)diphenylꢀ4ꢀyl]propanoate (5b).
+
1
Yield: 56%. ESIꢀMS (positive) m/z (%): 547 (M+Na , 100); HꢀNMR (CDCl ): δ 1.22 and 1.31
3
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
(2t, 6H, J=7.15, 2 CH
3 2 2 2
ꢀCH ), 3.23ꢀ3.36 (m, 2H, CH ꢀAr), 4.66 (s, 2H, CH ꢀO), 4.15 and 4.23 (2q,
4
H, J=7.15, 2 CH ꢀCH ), 4.83ꢀ4.88 (dd, 1H, CH), 6.91ꢀ6.98 and 7.29ꢀ7.52 (m, 17H, aromatics).
3
2
General Procedure for Synthesis of 2ꢀ(diphenylꢀ4ꢀylꢀoxy)ꢀ3ꢀ[4’ꢀ(1ꢀcarboxyꢀethoxy)diphenylꢀ4ꢀ
yl]propanoic acid (4) and 2ꢀ(diphenylꢀ4ꢀylꢀoxy)ꢀ3ꢀ[4’ꢀ(carboxyꢀmethoxy)diphenylꢀ4ꢀ
yl]propanoic acid (5).
To a stirred solution of the corresponding diethyl ester (0.15 mmol) in THF (10 mL), 1 N NaOH
(
10 mL) was added. The reaction mixture was stirred for 3 h at room temperature. Then, the solvent
was evaporated in vacuo and the aqueous phase was acidified with 6 N HCl (2 ml) and extracted
with diethyl ether. The combined organic layers were washed with brine, dried over Na SO and
evaporated to dryness affording the title acids as white solids, which was crystallized from
CHCl /nꢀhexane.
ꢀ(Diphenylꢀ4ꢀylꢀoxy)ꢀ3ꢀ[4’ꢀ(1ꢀcarboxyꢀethoxy)diphenylꢀ4ꢀyl]propanoic acid (4).
2
4
3
2
1
Yield = 30%. Mp = 163ꢀ169 °C. ESIꢀMS (negative) m/z (%): 481 (MꢀH, 100); HꢀNMR (DMSOꢀ
d ): δ 1.49 (d, 3H, J=6.60, CHꢀCH ), 3.12ꢀ3.48 (m, 2H, CH ꢀAr), 4.84 (q, 1H, J=6.60, CHꢀCH ),
6
3
2
3
2
4.99ꢀ5.03 (dd, 1H, CH-CH ), 6.89ꢀ6.94 and 7.25ꢀ7.57 (m, 17H, aromatics). Anal. Calcd for
C H O : C, 76.67; H, 5.43. Found: C, 76.98; H, 5.37.
30
26
6
2ꢀ(Diphenylꢀ4ꢀylꢀoxy)ꢀ3ꢀ[4’ꢀ(carboxyꢀmethoxy)diphenylꢀ4ꢀyl]propanoic acid (5).
1
Yield = 36%. Mp = 188ꢀ194 °C. ESIꢀMS (negative) m/z (%): 467 (MꢀH, 100); HꢀNMR (DMSOꢀ
d
7
5
6
): δ 3.13ꢀ3.31 (m, 2H, CHꢀCH
2 2
), 4.68 (s, 2H, CH ꢀO), 4.98ꢀ5.03 (m, 1H, CH), 6.92ꢀ6.97 and
.25ꢀ7.57 (m, 17H, aromatics). Anal. Calcd for C H O : C, 74.35; H, 5.16. Found: C, 73.74; H,
29
24
6
.13.
Diethyl 2ꢀ(4ꢀphenylꢀphenoxy)malonate (6a).
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 24 of 74
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
A mixture of diethyl chloromalonate (14.4 mmol) and sodium 4ꢀphenylphenate (15.8 mmol),
prepared from an equivalent amount of 4ꢀphenylphenol and sodium in absolute EtOH, was stirred
and heated under reflux in acetone (115 mL) for 6 h. The solvent was evaporated under reduced
pressure, and the residue was taken up with ethyl acetate, washed twice with 0.5 N NaOH and
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
2 4
brine, dried over Na SO and concentrated to give an oily residue which was chromatographed on
silica gel column (nꢀhexane/ethyl acetate 95:5 as eluent). The title compound was obtained as a
+
1
white solid in 78% yield. GCꢀMS m/z (%): 328 (M , 100), 169 (47), 152 (38). HꢀNMR (CDCl
3
):
δ 1.32 (t, 6Η, J=7.15, 2 CH ), 4.33 (q, 4H, J=7.15, 2 CH ), 5.24 (s, 1H, CH), 7.02ꢀ7.06 and 7.25ꢀ
3
2
7.55 (m, 9H, aromatics).
Diethyl 2ꢀ(4ꢀphenylꢀphenoxy)ꢀ2ꢀ(4ꢀbromoꢀbenzyl)malonate (6b).
A solution of 6a (11.4 mmol) in anhydrous DMF (15 mL) was added dropwise to a suspension of
NaH (95% powder, 34.0 mmol) in the same solvent (30 mL) at 0 °C. After stirring for 20 min at rt,
a solution of 4ꢀbromoꢀbenzylbromide (22.8 mmol) in anhydrous DMF (15 mL) was added dropwise
and the resulting reaction mixture stirred at 55 °C for 24 h. DMF was removed under reduced
pressure and the residue was dissolved in ethyl acetate, washed twice with saturated ammonium
chloride solution and brine. The organic phase was dried over Na SO and the solvent evaporated in
2
4
vacuo to give a white solid which was chromatographed on silica gel column (nꢀhexane/ethyl
acetate 98:2 as eluent). The title compound was obtained as a colorless oil in 63% yield. GCꢀMS
+
+
1
m/z (%): 498 (M +2, 1), 496 (M , 1), 170 (100). HꢀNMR (CDCl
3
3
): δ 1.15 (t, 6Η, J=7.15, 2 CH ),
3
.61 (s, 2H, CH ꢀPh), 4.19 (q, 4H, J=7.15, 2 CH ꢀO), 7.02ꢀ7.12 and 7.25ꢀ7.56 (m, 13H, aromatics).
2
2
Diethyl 2ꢀ(4ꢀphenylꢀphenoxy)ꢀ2ꢀ(4ꢀphenylꢀbenzyl)malonate (6c).
Phenylboronic acid (2.0 mmol) and Cs CO (1.5 mmol) were added, under N
3 4
stirred solution of 6b (1 mmol) in anhydrous toluene (18 mL). After 0.5 h, Pd(PPh )
2
3
2
atmosphere, to a
(0.03 mmol)
was added, and the reaction mixture was stirred for 24 h at 95 °C. Then, it was quenched with 1 N
HCl (5 mL) and ethyl acetate (5 mL) and filtered through a Celite pad to remove the catalyst. The
3 2 4
filtrate was washed twice with saturated NaHCO solution and brine, dried over Na SO , and
ACS Paragon Plus Environment

Page 25 of 74
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
concentrated to dryness to give a yellow solid residue which was chromatographed on a silica gel
column, (nꢀhexane/diethyl ether 90:10 as eluent), affording the title compound as a white solid.
1
Yield: 40%. ESIꢀMS (positive) m/z (%): 517 (M+Na, 100), 348 (2). HꢀNMR (CDCl
3
):
δ 1.16 (t, 6Η, J=7.15, 2 CH
.29ꢀ7.58 (m, 18H, aromatics).
(R,S)ꢀ2ꢀ(4ꢀPhenylꢀphenoxy)ꢀ3ꢀ(4ꢀdiphenyl)propanoic acid (6).
3
), 3.70 (s, 2H, CH
2
ꢀPh), 4.21 (q, 4H, J=7.15, 2 CH
2
ꢀO), 7.05ꢀ7.10 and
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
7
The diester 6c (7.0 mmol) was stirred and refluxed with 1 N NaOH (21 mL) in 95% EtOH (70 mL)
for 18 h. The organic solvent was distilled off under reduced pressure and the resulting aqueous
phase was acidified to pH 2 with 6 N HCl and extracted with Et
2
O (5x50 ml). The combined
organic extracts were washed with brine, dried over Na SO and the solvent removed under reduced
2
4
pressure. The resulting mixture was heated at 160 °C for 2 h affording the desired acid as a white
solid which was crystallized from CHCl
3
/nꢀhexane (61% yield). Mp > 200°C. ESIꢀMS (negative)
1
m/z (%): 393 (MꢀH, 100), 223 (32), 169 (24). HꢀNMR (DMSOꢀd ): δ 3.15ꢀ3.48 (m, 2H, CH ),
6
2
5.00ꢀ5.04 (dd, 1H, CH), 6.92ꢀ6.95 and 7.25ꢀ7.64 (m, 18H, aromatics). Anal. Calcd for C27
H
22
O :
3
C, 82.21; H, 5.62. Found: C, 81.97; H, 5.50.
R,S)ꢀ2ꢀ(4ꢀPhenylꢀphenoxy)ꢀ3ꢀ(4ꢀbromophenyl)propanoic acid (6d).
(
The title compound was obtained as a white solid in 35% yield by using the procedure described
1
above for 6. ESIꢀMS (negative) m/z (%): 395 (MꢀH, 100), 225 (43), 169 (59). HꢀNMR (CDCl
3
):
2
δ 3.20ꢀ3.28 (m, 2H, CH ), 4.84ꢀ4.97 (dd, 1H, CH), 6.90ꢀ6.94 and 7.15ꢀ7.55 (m, 13H, aromatics).
Ethyl 2ꢀ(4ꢀphenylꢀphenoxy)ꢀ3ꢀ(4ꢀbromophenyl)propanoate (6e).
The acid 6d (1 mmol) was dissolved in 96° ethanol (20 mL) at 80 °C and added with 3 drops of
concentrated sulfuric acid. The reaction mixture was refluxed for 5 h. Then, the solvent was
evaporated, and the residue was dissolved in ethyl acetate, washed with saturated NaHCO
3
solution
and brine. The organic phase was dried over Na SO , filtered and concentrated under reduced
2
4
+
pressure affording a white solid (0.67 mmol). Yield: 67%. GCꢀMS m/z (%): 426 (M +2, 60), 424
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 26 of 74
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
+
1
(M , 58), 170 (100). HꢀNMR (CDCl
.20 (q, 2H, J=7.15, CH ꢀO), 4.68ꢀ4.81 (dd, 1H, CH), 6.87ꢀ6.92 and 7.19ꢀ7.53 (m, 13H, aromatics).
General Procedure for Synthesis of (R)ꢀ or (S)ꢀethyl 4ꢀiodoꢀphenyllactate ((R)ꢀ6f, (S)ꢀ6f).
N HCl (6 mL) and glacial acetic acid (9 mL) were added to a suspension of commercially
available (R)ꢀ or (S)ꢀ(4ꢀiodo)phenylalanine (5 mmol) in H O (17 mL). The resulting solution was
stirred and cooled to 0 °C and carefully added, during 0.5 h, of NaNO (42 mmol) solution in H O
3
): δ 1.22 (t, 3Η, J=7.15, CH
3 2
), 3.18ꢀ3.28 (m, 2H, CH Ph),
4
2
1
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
2
2
2
(9 mL). The solution was stirred at room temperature for 22 h and carefully added, during 0.5 h, of
concentrated HCl (5 mL). Water was distilled off and the residue was extracted six times with
boiling acetone. Acetone was distilled off and the oily residue was dissolved in Et
Na SO . The solvent was evaporated to dryness affording the corresponding (R)ꢀ or (S)ꢀ(4ꢀ
iodo)phenyllactic acids as white solids. This solid was dissolved in absolute EtOH (30 mL), added
of two drops of concentrated H SO and refluxed with stirring for 5 h. The solvent was removed
under reduced pressure, the residue was dissolved in Et O and the resulting solution was washed
twice with saturated NaHCO solution and brine. After drying over Na SO , the evaporation of the
2
O and dried over
2
4
2
4
2
3
2
4
solvent afforded the corresponding ethyl esters as yellow oils which were used in the next step
+
without any further purification. GCꢀMS m/z(%): 320 (M , 15), 302 (85), 217 (100), 91 (37).
General Procedure for Synthesis of (R)ꢀ or (S)ꢀethyl 2ꢀ(4ꢀphenylphenoxy)ꢀ3ꢀ(4ꢀ
iodophenyl)propanoate ((R)ꢀ6g, (S)ꢀ6g).
A solution of diisopropylazodicarboxylate (DIAD, 5 mmol) in anhydrous THF (8 mL) was added
dropwise to an iceꢀbath cooled mixture of (S)ꢀ6f or (R)ꢀ6f (5 mmol), 4ꢀphenylꢀphenol (5 mmol),
and triphenylphosphine (5 mmol) in anhydrous THF (8 mL). The reaction mixture was stirred at
room temperature overnight, under N
2
atmosphere. Then, the solvent was evaporated in vacuo and
the residue was chromatographed on silica gel column, (nꢀhexane/ethyl acetate 90:10 as eluent),
+
1
affording a yellow solid (0.87 mmol). GCꢀMS m/z (%): 471 (M , 100), 170 (69). HꢀNMR
(CDCl ): 1.21 (t, 3Η, J=6.90 CH ); 3.19ꢀ3.21 (m, 2H, ArꢀCH ); 4.20 (q, 2H, J=6.90, CH ꢀCH ),
.78ꢀ4.80 (m, 1H, CH), 6.87ꢀ7.64 (m. 13H, aromatics).
3
3
2
3
2
4
ACS Paragon Plus Environment

Page 27 of 74
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
(
R)ꢀ6g. Yield: 30%. (S)ꢀ6g. Yield: 37%.
General Procedure for Synthesis of (R)ꢀ or (S)ꢀethyl 2ꢀ(4ꢀphenylphenoxy)ꢀ3ꢀ(4ꢀ
diphenyl)propanoate ((R)ꢀ6h, (S)ꢀ6h).
2 3 2
Phenylboronic acid (0.84 mmol) and Cs CO (0.63 mmol) were added, under N atmosphere, to a
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
stirred solution of (R)ꢀ6g or (S)ꢀ6g (0.42 mmol) in anhydrous toluene (10 mL). After 0.5 h,
Pd(PPh ) (0.013 mmol) was added, and the reaction mixture was stirred for 35 h at 95 °C. Then, it
3
4
was quenched with 1 N HCl (3 mL) and ethyl acetate (3 mL) and filtered through a Celite pad to
remove the catalyst. The filtrate was washed twice with saturated NaHCO solution and brine, dried
over Na SO , and concentrated to dryness to give a dark yellow oily residue which was
chromatographed on a silica gel column (nꢀhexane/ethyl acetate 90:10 as eluent) affording the title
3
2
4
+
1
compounds as white solids. GCꢀMS m/z (%): 422 (M , 60), 253 (64), 167 (100). HꢀNMR (CDCl
.22 (t, 3Η, J=7.03, CH ); 3.30ꢀ3.33 (m, 2H, ArꢀCH ); 4.22 (q, 2H, J=7.03, CH ꢀCH ), 4.86ꢀ4.90
(m, 1H, CH), 6.91ꢀ7.60 (m. 18H, aromatics).
R)ꢀ6h. Yield: 40%. (S)ꢀ6h. Yield: 35%.
3
):
1
3
2
3
2
(
General Procedure for Synthesis of (R)ꢀ or (S)ꢀ2ꢀ(4ꢀphenylꢀphenoxy)ꢀ3ꢀ(4ꢀdiphenyl)propanoic
acid ((R)ꢀ6, (S)ꢀ6).
To a stirred solution of (R)ꢀ6h or (S)ꢀ6h (0.15 mmol) in THF (7.5 mL), 1 N NaOH was added (7.5
mL). The reaction mixture was stirred overnight at room temperature. Then, the solvent was
evaporated in vacuo and the aqueous phase was acidified with 2 N HCl and extracted with ethyl
acetate. The combined organic layers were washed with brine, dried over Na SO and evaporated to
2
4
dryness affording the title acids as white solids, which were crystallized from CHCl /nꢀhexane. ESIꢀ
3
1
MS (negative) m/z (%): 393 (M–H, 100), 223 (32), 169 (24). HꢀNMR (DMSOꢀd
6
): δ 3.15ꢀ3.48 (m,
2
2
H, CH ), 5.00ꢀ5.04 (dd, 1H, CH), 6.92ꢀ6.95 and 7.25ꢀ7.64 (m, 18H, aromatics).
®
The enantiomeric excesses (ee) were determined by using a CSPꢀHPLC column CHIRALCEL
OD; lambda = 254 nm; 1 mL/min; nꢀhexane/isoꢀpropanol/TFA 95:5:0.1.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 28 of 74
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
(
R)ꢀ2ꢀ(4ꢀPhenylꢀphenoxy)ꢀ3ꢀ(4ꢀdiphenyl)propanoic acid ((R)ꢀ6).
Yield: 52%; mp = > 200 °C; ee = > 90%; [α]
C, 82.21; H, 5.62. Found: C, 82.20; H, 5.66.
S)ꢀ2ꢀ(4ꢀPhenylꢀphenoxy)ꢀ3ꢀ(4ꢀdiphenyl)propanoic acid ((S)ꢀ6).
D 22 3
= ꢀ 37.5 (c 1.0, MeOH). Anal. Calcd for C27H O :
(
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
D 22 3
Yield: 65%; mp = 197ꢀ9 °C; ee = 90%; [α] = + 36.7 (c 1.0, MeOH). Anal. Calcd for C27H O : C,
82.21; H, 5.62. Found: C, 82.60; H, 5.67.
Biological Methods. Reference compounds, the medium, and other cell culture reagents were
purchased from SigmaꢀAldrich (Milan, Italy).
Plasmids. The expression vectors expressing the chimeric receptor containing the yeast Gal4ꢀDNA
binding domain fused to the human PPARαꢀ or PPARγꢀLBD and the reporter plasmid for these
Gal4 chimeric receptors (pGal5TKpGL3) containing five repeats of the Gal4 response elements
upstream of a minimal thymidine kinase promoter that is adjacent to the luciferase gene were
27
described previously.
Cell Culture and Transfections. Human hepatoblastoma cell line HepG2 (Interlab Cell Line
Collection, Genoa, Italy) was cultured in minimum essential medium (MEM) containing 10% heatꢀ
−
1
inactivated fetal bovine serum, 100 U of penicillin G mL , and 100 ꢁg of streptomycin sulfate
−
1
5
mL at 37 °C in a humidified atmosphere of 5% CO . For transactivation assays, 10 cells per well
2
were seeded in a 24ꢀwell plate and transfections were performed after 24 h with CAPHOS, a
calcium phosphate method, according to the manufacturer’s guidelines. Cells were transfected with
expression plasmids encoding the fusion protein Gal4−PPARαꢀLBD or Gal4−PPARγꢀLBD (30 ng),
pGal5TKpGL3 (100 ng), and pCMVβgal (250 ng). Four hours after transfection, cells were treated
for 20 h with the indicated ligands in triplicate. For studies involving the covalent antagonist 7,
before incubation with PPARγ ligands LJ570 and rosiglitazone, cells were first preꢀincubated with
ACS Paragon Plus Environment

Page 29 of 74
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
5 ꢀM 7 for 3 h. Luciferase activity in cell extracts was then determined by a luminometer
3
(
VICTOR V Multilabel Plate Reader, PerkinElmer). βꢀGalactosidase activity was determined using
47
orthoꢀnitroꢀphenylꢀβꢀDꢀgalactopyranoside as previously described. All transfection experiments
were repeated at least twice.
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Co-regulator recruitment by Fluorescence Resonance Energy Transfer assay. Fluorescence
resonance energy transfer assays to evaluate PGC1α recruitment were performed in 384ꢀwell plates
in a final volume of 10 ꢁl. A mixture of 8 ng of human PPARγꢀligandꢀbinding domain, 0.8 ng of
europiumꢀlabeled antiꢀHis antibody (PerkinElmer Life Sciences), 86 ng of allophycocyaninꢀlabeled
streptavidin (PerkinElmer Life Sciences) and 8 ng of biotinylated peptide (PRIMM, Milano, Italia)
in a FRET buffer containing 50 mM Tris (pH 7.5), 50 mM KCl, 1 mM DTT, and 0.1% fatty acid
free BSA was prepared. LJ570 was added in 5ꢀpoint dose response curve starting at 10 ꢁM as the
highest concentration. The reactions were equilibrated for 1 h at room temperature and then
measured in an Envision multiplate reader (PerkinElmer Life Sciences) using 340 nm as excitation
and 615 and 665 nm as emission wavelengths. The ratio between 665 (allophycocyanin signal) and
615 (europium signal) was used to evaluate the peptide recruitment on the receptor. The peptide
sequence used for PGCꢀ1α was BiotinꢀDGTPPPQEAEEPSLLKKLLLAPANTꢀCOOH (NR box 1)
spanning amino acids 130 –154 (reference no. NP_037393).
Cell Cultures. Hepa1ꢀ6 were cultured in DMEM supplemented by 10% fetal bovine serum and 1%
penicillinꢀstreptomycin at 37 °C in a humidified atmosphere of 5% CO . Cells were treated once
2
with LJ570 or Wyꢀ14,643. RNA was extracted 6 hours after treatments. Adipogenesis assays were
36
performed as previously described. Briefly, 3T3ꢀL1 cells (ATCC) were cultured in DMEM
supplemented by 10% bovine calf serum; 2 days after reaching confluence, cells were differentiated
−1
in the presence of rosiglitazone or LJ570 at the indicated concentration and 5 ꢁg insulin mL .
Culture medium was replenished with ligands three times every other day. Lipid accumulation was
ACS Paragon Plus Environment