On the metabolically active form of metaglidasen: Improved synthesis and investigation of its peculiar activity on peroxisome proliferator-activated receptors and skeletal muscles

Article abstract of DOI:10.1002/cmdc.201402462

Metaglidasen is a fibrate-like drug reported as a selective modulator of peroxisome proliferator-activated receptor γ (PPARγ), able to lower plasma glucose levels in the absence of the side effects typically observed with thiazolidinedione antidiabetic agents in current use. Herein we report an improved synthesis of metaglidasen's metabolically active form halofenic acid (R)-2 and that of its enantiomer (S)-2. The activity of the two stereoisomers was carefully examined on PPARα and PPARγ subtypes. As expected, both showed partial agonist activity toward PPARγ; the investigation of PPARα activity, however, led to unexpected results. In particular, (S)-2 was found to act as a partial agonist, whereas (R)-2 behaved as an antagonist. X-ray crystallographic studies with PPARγ were carried out to gain more insight on the molecular-level interactions and to propose a binding mode. Given the adverse effects provoked by fibrate drugs on skeletal muscle function, we also investigated the capacity of (R)-2 and (S)-2 to block conductance of the skeletal muscle membrane chloride channel. The results showed a more beneficial profile for (R)-2, the activity of which on skeletal muscle function, however, should not be overlooked in the ongoing clinical trials studying its long-term effects.

Full text of DOI:10.1002/cmdc.201402462

DOI: 10.1002/cmdc.201402462
Full Papers
On the Metabolically Active Form of Metaglidasen:
Improved Synthesis and Investigation of Its Peculiar
Activity on Peroxisome Proliferator-Activated Receptors
and Skeletal Muscles
Antonio Laghezza,[a] Roberta Montanari,[b] Antonio Lavecchia,[c] Luca Piemontese,[a]
Giorgio Pochetti,[b] Vito Iacobazzi,[d, e] Vittoria Infantino,[d, f] Davide Capelli,[b]
Michela De Bellis,[a] Antonella Liantonio,[a] Sabata Pierno,[a] Paolo Tortorella,[a]
Diana Conte Camerino,[a] and Fulvio Loiodice*[a]
Metaglidasen is a fibrate-like drug reported as a selective mod-
ulator of peroxisome proliferator-activated receptor g (PPARg),
able to lower plasma glucose levels in the absence of the side
effects typically observed with thiazolidinedione antidiabetic
agents in current use. Herein we report an improved synthesis
of metaglidasen’s metabolically active form halofenic acid (R)-2
and that of its enantiomer (S)-2. The activity of the two stereo-
isomers was carefully examined on PPARa and PPARg sub-
types. As expected, both showed partial agonist activity
toward PPARg; the investigation of PPARa activity, however,
led to unexpected results. In particular, (S)-2 was found to act
as a partial agonist, whereas (R)-2 behaved as an antagonist. X-
ray crystallographic studies with PPARg were carried out to
gain more insight on the molecular-level interactions and to
propose a binding mode. Given the adverse effects provoked
by fibrate drugs on skeletal muscle function, we also investi-
gated the capacity of (R)-2 and (S)-2 to block conductance of
the skeletal muscle membrane chloride channel. The results
showed a more beneficial profile for (R)-2, the activity of which
on skeletal muscle function, however, should not be over-
looked in the ongoing clinical trials studying its long-term ef-
fects.
Introduction
Peroxisome proliferator-activated receptors (PPARs) are ligand-
dependent transcription factors that belong to the nuclear re-
ceptor superfamily. They control the expression of genes in-
volved in fatty acid and glucose metabolism and function as
cellular lipid sensors that activate transcription in response to
the binding of cognate ligands, generally fatty acids and their
eicosanoid metabolites.[1–3] As ligand-dependent receptors,
PPARs form heterodimers with the retinoid X receptor (RXR)
and adopt an active conformation in the presence of a ligand.
Additional co-regulator proteins are recruited to create a com-
plex that binds to peroxisome proliferator response elements
(PPRE) in target genes, thus regulating their expression.[4–6]
There are three PPAR subtypes, commonly designated as
PPARa, PPARg, and PPARd (b) expressed in different tissues.[7]
Agonists of the g subtype have been extensively studied for
their role in regulating glucose metabolism and insulin sensi-
tivity.[8] Full agonists of PPARg such as rosiglitazone and piogli-
tazone have been developed and marketed for the treatment
of type 2 diabetes.[9] However, mechanism-based side effects
including weight gain, edema, congestive heart failure, and
the recently reported increased risk of bone fracture following
treatment with rosiglitazone or pioglitazone are major unde-
sired effects associated with the use of PPARg full agonists.[10,11]
As a result of the clinical observations mentioned above,
emphasis has shifted to the development of partial agonists or
selective PPARg modulators (SPPARgMs). The SPPARM ap-
proach has recently attracted considerable attention because it
proposes that diverse PPAR ligands, depending on their struc-
tures, would bind in a distinct manner to PPARs, inducing vari-
ous levels of activation and distinct conformational changes of
the receptor, leading to differential interactions with co-activa-
[a] Dr. A. Laghezza,+ Dr. L. Piemontese, Dr. M. De Bellis, Dr. A. Liantonio,
Dr. S. Pierno, Prof. P. Tortorella, Prof. D. Conte Camerino, Prof. F. Loiodice
Dipartimento di Farmacia-Scienze del Farmaco
Universitꢀ degli Studi di Bari “Aldo Moro”, 70126 Bari (Italy)
E-mail: fulvio.loiodice@uniba.it
[b] Dr. R. Montanari,+ Dr. G. Pochetti, Dr. D. Capelli
Istituto di Cristallografia, Consiglio Nazionale delle Ricerche
Montelibretti, 00015 Monterotondo Stazione, Roma (Italy)
[c] Prof. A. Lavecchia
Dipartimento di Farmacia, “Drug Discovery” Laboratory
Universitꢀ degli Studi di Napoli “Federico II”, 80131 Napoli (Italy)
[d] Prof. V. Iacobazzi, Dr. V. Infantino
Dipartimento di Bioscienze, Biotecnologie e Biofarmaceutica
Laboratorio di Biochimica e Biologia Molecolare
Universitꢀ degli Studi di Bari “Aldo Moro”, 70126 Bari (Italy)
[e] Prof. V. Iacobazzi
Istituto di Biomembrane e Bioenergetica
Consiglio Nazionale delle Ricerche, 70126 Bari (Italy)
[f] Dr. V. Infantino
Dipartimento di Chimica, Universitꢀ della Basilicata, 85100 Potenza (Italy)
[+] These authors contributed equally to this work.
Supporting information for this article is available on the WWW under
http://dx.doi.org/10.1002/cmdc.201402462.
ChemMedChem 0000, 00, 0 – 0
1
ꢀ 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
These are not the final page numbers! ÞÞ

Full Papers
tors and co-repressors. Therefore, structurally diverse modula-
tors or partial agonists are likely to elicit different pharmaco-
logical and toxicological effects depending on the context of
the tissue, i.e., abundance of cofactor proteins and target
gene. This may enable uncoupling of the benefits of PPAR acti-
vation from the adverse effects associated with full agonism. In
agreement with the SPPARgM concept, a number of these
modulators have already demonstrated desirable pharmaco-
logical profiles in various rodent models with significantly de-
creased side effects relative to those generally observed with
existing full agonists.[12–24] One key representative is metaglida-
sen, a selective PPARg partial agonist that structurally, mecha-
nistically, and preclinically differs from the glitazones.[25] Meta-
glidasen is the R enantiomer of halofenate (1), a drug that was
tested clinically in the 1970s as a hypolipidemic and hypourice-
mic agent.[26,27] Both halofenate and metaglidasen are prodrug
esters that are rapidly and completely modified in vivo by non-
specific serum esterases to give the corresponding free acid
form 2 (Figure 1).
tainty its absolute configuration, which, in a 2007 patent,[32]
was attributed in a somewhat unclear manner. The transcrip-
tional activity of the free acid form (R)-2 was carefully exam-
ined with PPARa and PPARg; the same biological evaluation
was also performed on its enantiomer (S)-2, never investigated
before. As expected, both stereoisomers showed partial ago-
nist activity toward PPARg; however, the investigation of
PPARa activity led to unexpected results. In particular, (S)-2
acted as a partial agonist, whereas its enantiomer behaved as
an antagonist. These results were confirmed by the different
effects on the PPARa-mediated gene expression of mitochon-
drial carnitine palmitoyltransferase 1 (CPT1), which is a molecu-
lar component of the carnitine shuttle system essential for the
mitochondrial oxidation of fatty acids. X-ray crystallography on
the PPARg subtype was performed for both stereoisomers to
gain greater insight into the interactions at a molecular level
and to propose a binding mode that explains the lack of ste-
reoselectivity at this receptor. Finally, because skeletal muscle
is a target for PPAR agonists, the effects of both stereoisomers
on the function of this tissue were evaluated by measuring the
resting chloride conductance (gCl) sustained by the voltage-
gated chloride channel ClC-1. Previous studies have shown
that a decrease in gCl function is one of the mechanisms of
action responsible for myopathies, the most significant of the
complications of lipid-lowering treatment with statins and fi-
brates.[33–36]
Figure 1. Structure of halofenate and its metabolically active form, halofenic
Results and Discussion
acid.
Chemistry
Phase 2a clinical trial data indicate that metaglidasen signifi-
cantly lowers plasma glucose levels in the absence of side ef-
fects such as weight gain and edema, which are observed with
pharmacological agents in current use. During this develop-
ment program, researchers at Metabolex discovered that meta-
glidasen is an effective uricosuric agent with unique proper-
ties; they repurposed the drug to treat gout with excellent
safety and tolerability. Nevertheless, metaglidasen is still inves-
tigated as a useful agent for the treatment of type 2 diabetes
and hyperglycemia as demonstrated by some recent pat-
ents.[28,29] Interestingly, in preclinical rodent models, metaglida-
sen also displays pronounced triglyceride lowering, which is
often considered a hallmark of PPARa activation. However, in
vivo and in vitro studies conducted with (R)-1 and its metabol-
ically active form (R)-2, respectively, indicate that this drug has
no PPARa activity, suggesting that its in vivo lipid-lowering
ability is therefore mediated by an alternative mechanism that
has yet to be determined.[30]
As described in a 2007 patent, the synthesis of racemic halo-
fenic acid 2, the active metabolite of halofenate, is carried out
in five steps.[37] The corresponding levo isomer, the free acid
form of metaglidasen, is obtained either by the same proce-
dure using a chiral auxiliary[32] or by resolution of the race-
mate.[37] The authors assigned the R configuration to this
isomer, even though the procedure used for this attribution re-
mains unclear. In an attempt to find a rapid and mild method
to prepare (Æ)-halofenic acid, we applied a modified form of
the procedure reported by Job, Buchwald, and co-workers,[38,39]
which allowed us to obtain the target compound in a single
step as shown in Scheme 1.
Racemic 4-chloromandelic acid was condensed with 1-iodo-
3-trifluoromethylbenzene in the presence of a catalyst system
consisting of copper(I) iodide and cesium carbonate to afford
(Æ)-halofenic acid in 40% yield. The yield of this reaction is not
remarkable, yet is fairly good considering that this procedure is
Metaglidasen and halofenate show the same activity both in
vitro and in vivo, suggesting a lack of stereoselectivity for
PPARg.[21,25] The mainspring for the clinical development of (R)-
1 in place of the more easily accessible halofenate seems to be
its decreased inhibitory activity toward cytochrome P450 2C9
(CYP2C9) relative to the corresponding dextro isomer (S)-1.[31]
Given the promising therapeutic potential of metaglidasen,
herein we describe a more convenient synthesis to obtain the
active metabolite of this drug; moreover, we assign with cer-
Scheme 1. Preparation of (Æ)-halofenic acid. Reagents and conditions:
a) Cs2CO3, CuI, BuCN, 1108C, argon, 72 h, 40% yield.
&
ChemMedChem 0000, 00, 0 – 0
www.chemmedchem.org
2
ꢀ 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Full Papers
much less expensive and faster than that reported.[37] The only
drawback of this reaction concerns the pH in its workup;
indeed, pH values only slightly outside the range of 5–6 lead
to drastic decreases in yield. The resolution of (Æ)-2 was car-
ried out by fractional crystallization from ethanol/water of the
diastereomeric salts obtained with (S)- or (R)-1-(2-naphthyl)-
ethylamine.[37] As an alternative to resolution, we tried to pre-
pare (+)-2 and (À)-2 by the same procedure used for the race-
mate by starting from optically active 4-chloromandelic acid.
Unfortunately, the reaction occurred with partial racemization;
however, this synthetic method was useful for unambiguously
assigning the absolute configuration to both isomers, given
that the stereogenic center of optically active 4-chloromandelic
acid was not involved in the reaction. So, starting from (R)-4-
chloromandelic acid,[40] we obtained the partially enriched levo
isomer of halofenic acid; this allowed us to confirm that the
absolute configurations of (À)-2 and (+)-2 are R and S, respec-
tively.
PPAR activity
(R)-2 and (S)-2 were evaluated for their agonist activity on the
human PPARa (hPPARa) and PPARg (hPPARg) subtypes. For
this purpose, GAL4–PPAR chimeric receptors were expressed in
transiently transfected HepG2 cells according to a previously
reported procedure.[41] The results obtained were compared
with corresponding data for Wy14,643 and rosiglitazone used
as reference compounds in the PPARa and PPARg transactiva-
tion assays, respectively. The maximum induction obtained
with the reference agonist was defined as 100%.
Figure 2. Gal4 reporter assay data for A) human PPARg and B) human
PPARa. A Gal4–hPPARg or Gal4–hPPARa ligand binding domain expression
plasmid was co-transfected with a luciferase reporter plasmid in HepG2
cells. The insert in panel A) is a magnification of the curves that relate only
to compounds (R)-2 and (S)-2. Data are normalized with respect to control,
and values are the meanÆSEM of n=3 experiments performed in duplicate.
As regards PPARg, the transactivation assay showed this re-
ceptor to lack any stereoselectivity between (R)-2 and (S)-2; in
fact, both isomers acted as partial agonists with similar poten-
cy (EC50: 4.8 and 7.6 mm, respectively) and efficacy (Emax: ~10%)
confirming previously reported data[25] (Figure 2A).
We also decided to determine the thermodynamic parame-
ters relating to formation of the complexes of both stereoiso-
mers with the PPARg ligand binding domain (LBD) by isother-
mal titration calorimetry (ITC). ITC is a very useful alternative to
conventional PPAR binding assays, which need specific radioli-
gands for labeling of receptors. This technique, which we have
successfully applied to some PPAR agonists,[42–45] measures the
heat absorbed or released by titrating the protein with
a ligand at constant temperature, allowing one to obtain the
affinity constant. The results obtained for (R)-2 and (S)-2 were
1.36ꢁ105 mÀ1 and 1.72ꢁ105 mÀ1, respectively, showing that
these ligands bind to PPARg LBD with similar affinity in accord-
ance with their functional activity. Figure 1 of the Supporting
Information shows the calorimetric data (raw and integration
data) obtained in the titration of PPARg with both isomers.
Afterward we evaluated the activity of both stereoisomers
toward PPARa. Previous experiments reported no activity on
this receptor subtype from racemic halofenic acid and its levo
isomer, so it was clear to assume that the dextro isomer was in-
effective as well.[21,30] In spite of this, however, we decided to
investigate the effects on PPARa of (R)-2 and (S)-2 because of
their great similarity with some chiral a-aryloxy-a-arylacetic
acids previously claimed as highly stereoselective PPARa ago-
nists.[46] The investigation of PPARa activity led to unexpected
results for the two stereoisomers. In fact, (S)-2 acted as a partial
agonist (EC50: 11 mm, Emax: 31%), whereas its enantiomer be-
haved as an antagonist (Figure 2B). The antagonist activity of
(R)-2 was confirmed by conducting a competitive binding
assay in which PPARa activity at a fixed concentration of the
reference agonist Wy14,643 was measured in cells treated
with increasing concentrations of (R)-2. As shown in Figure 3,
(R)-2 displayed a dose-dependent inhibition of Wy14,643-
mediated PPARa activity, with a half-maximal inhibitory con-
centration of ~21 mm. These data suggest that (R)-2 is able to
interact with PPARa in such a way to displace Wy14,643 and
inhibit its activity in a cellular context. It is reasonable to
assume that (R)-2 is able to similarly displace (S)-2, thus ex-
plaining why racemic halofenic acid does not display PPARa
activity.
To further corroborate the differing results between (R)-2
and (S)-2 on PPARa activity, we investigated the effects of both
stereoisomers on mitochondrial carnitine/acylcarnitine carrier
(CAC) and carnitine palmitoyltransferase 1 (CPT1) PPARa-medi-
ated gene expression. These two molecular components of the
ChemMedChem 0000, 00, 0 – 0
www.chemmedchem.org
3
ꢀ 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
These are not the final page numbers! ÞÞ

Full Papers
Figure 3. PPARa antagonist effect of (R)-2. PPARa activity was measured in
Gal4–hPPARa LBD transfected cells treated with Wy14,643 at 10 mm and in-
creasing concentrations of (R)-2. Data are normalized with respect to control,
and values are the meanÆSEM of n=2 experiments performed in triplicate.
carnitine shuttle system are essential for the mitochondrial oxi-
dation of fatty acids, because they catalyze the entry of fatty
acid acyl groups into the mitochondrial matrix, where the en-
zymes of fatty acid b-oxidation are located. Recent studies
demonstrated that transcription of the CAC gene is enhanced
by statins and fibrates, providing a novel contribution to the
understanding of their hypolipidemic action.[42,47,48] Therefore,
we decided to test (S)-2 and (R)-2 with the aim to determine if
these fibrate-like drugs are able to similarly up-regulate mito-
chondrial CAC and CPT1 gene expression. The effects of in
vitro application of these compounds were investigated on
HepG2 cells, which were incubated for 48 h with (S)-2 and (R)-
2 at 50 mm. The PPARa agonist Wy14,643 (50 mm) was used as
a reference compound. After incubation, total mRNA was ex-
tracted and used to determine CAC and CPT1 transcript levels.
As shown in Figures 4A and 5A, (R)-2 did not induce any in-
crease of CAC and CPT1 mRNA, as expected; in contrast, (S)-2
caused a ~1.5-fold increase in CPT1 mRNA, whereas no effect
was detected for CAC mRNA. Consistently, western blotting
analysis showed no effects on CAC and CPT1 protein levels
from (R)-2 (Figure 4B,C and 5B,C) whereas it confirmed a sig-
nificant, albeit weak, increase in CPT1 protein level (Fig-
ure 5B,C) in cells treated with (S)-2. The observed effects of
(S)-2 on CPT1 gene expression level, however, turned out to be
lower than that obtained with Wy14,643 at the same concen-
tration.
Figure 4. A) Total RNA extracted from HepG2 cells, treated with or without
50 mm (R)-2, 50 mm (S)-2, or 50 mm Wy14,643 for 24 h, was used to quantify
CAC mRNA by real-time PCR; data are the meanÆSD of three independent
duplicate experiments. B) CAC and b-actin of HepG2 cells, treated for 48 h as
described for panel A), were immunolabeled with specific antibodies. C) The
intensities of immunolabeled protein bands depicted in panel B) were quan-
tified by densitometric scanning. Quantitation represents fold change of
CAC protein signals relative to control cells; data are the meanÆSD from
three independent experiments (*p<0.05; one-way ANOVA, Bonferroni test).
was accommodated between helix 3 and the b-sheet, where
one of its carboxylic oxygen atoms was found to form a hydro-
gen bond (2.7 ꢂ) with the NH group of S342 belonging to the
b-sheet (Figure 7A). This interaction has often been observed
in other PPARg complexes with partial agonists. Upon binding,
the ligand forced the side chains of R288 and E291 to assume
different conformations. In this new orientation, the charged
guanidinium group of R288 interacts with the electronegative
trifluoromethyl group of the ligand (3.5 ꢂ) and made van der
Waals (vdW) interactions with the residue A292 belonging to
helix 3. The chlorine atom, at the opposite extremity of the
ligand, is directed toward a hydrophobic core of the LBD
formed by the side chains of I281, L353, F363, and M364 (Fig-
ure S2, Supporting Information). The shortest distances of the
chlorine atom were with one methyl group of I281 (3.7 ꢂ) and
a carbon atom of the aromatic ring of F363 (3.7 ꢂ). A LigPlot
analysis[49] of the vdW interactions of (R)-2 is shown in Fig-
ure S3A in the Supporting Information.
Given that no crystal structures of PPARa complexed with
carboxylic antagonists are available as a starting point for
docking, X-ray crystal studies of the complexes PPARa–(S)-2
and PPARa–(R)-2 will be crucial to confirm the obtained results.
These studies are still in progress for PPARa, whereas they
have already been achieved for PPARg with the aim to provide
a molecular explanation for the lack of stereoselectivity of this
receptor toward (R)-2 and (S)-2.
Binding of (R)-2 in the PPARg LBD
The final omit map (Figure 6A) showed clear electron density
where one molecule of (R)-2 can be easily fitted. This isomer
&
ChemMedChem 0000, 00, 0 – 0
www.chemmedchem.org
4
ꢀ 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Full Papers
entrance of the LBD (Figure S4, Supporting Information). (S)-2
faced the b-sheet and both its carboxylic oxygens were found
to be engaged in two hydrogen bonds with the NH group of
S342 and the side chain OH of S342, respectively (Figure 7B).
The side chain of R288 maintained its typical position (t confor-
mation), and its guanidinium group made two hydrogen
bonds with one of the ligand carboxylic oxygen atoms and the
NH group of E343, respectively. The ligand electronegative CF3
group interacted with the positively charged side chain of
R288 through a water molecule also bridged with the CO of
L340 (Figure 7B). One of the fluorine atoms also contacted the
residue C285 at distances of ~3.0 ꢂ. (S)-2 formed more and
stronger interactions than (R)-2 with H3 and the b-sheet,
better stabilizing these regions of the LBD (Figure S3B, Sup-
porting Information). The largest number of hydrogen bonds
and vdW interactions realized by this ligand, in comparison
with (R)-2, was in accordance with the more favorable enthalp-
ic contribution observed in the ITC experiment upon binding
to PPARg LBD (À6.5 vs. À3.8 kcalmolÀ1, respectively). The omit
map around (S)-2 is shown in Figure 6B.
Superimposition of (R)-2 and (S)-2 in the PPARg LBD (Fig-
ure 8A) shows that the two enantiomers, although both inter-
acting with the NH group of S342 through their carboxylic
oxygen atoms, occupy slightly different positions between H3
and the b-sheet, with the S enantiomer less inserted into the
LBD cavity. The superimposition with other known PPARg par-
tial agonists showed that (R)-2 occupied the same position of
3 (MRL24)[50] (PDB code 2Q5P) and (S)-2, that of the partial ago-
nist 4 (FS214)[42] (PDB code 4E4Q) (Figure 8B,C). Both stereoiso-
mers seem to be able to stabilize, even though in different
ways, the same region of the receptor (H3 and b-sheet), thus
explaining their similar affinity and activity toward PPARg.
The binding mode of (R)-2 would also explain its ability to
inhibit the CDK5-mediated phosphorylation of PPARg. Recently,
in fact, Choi et al. reported this biochemical effect at the level
of S273 of PPARg2 (corresponding to S245 of PPARg1) for rosi-
glitazone and some PPARg partial agonists including (R)-1.[50]
This study suggests that both thiazolidinediones (TZDs) and
partial agonists with antidiabetic effects improve insulin sensi-
tivity primarily by this mechanism. Inhibition of S273 phos-
phorylation seems to be distinct from classical transcriptional
activation, which involves the stabilization of the highly dy-
namic activation of helix 12 and
Figure 5. A) Total RNA extracted from HepG2 cells, treated with or without
50 mm (R)-2, 50 mm (S)-2, or 50 mm Wy14,643 for 24 h, was used to quantify
CPT1 mRNA by real-time PCR; data are the meanÆSD of three independent
duplicate experiments. B) CPT1 and b-actin of HepG2 cells, treated for 48 h
as described for panel A), were immunolabeled with specific antibodies.
C) The intensities of immunolabeled protein bands depicted in panel B)
were quantified by densitometric scanning. Quantitation represents fold
change of CPT1 protein signals relative to control cells; data are the
meanÆSD from three independent experiments (*p<0.01, **p<0.001; one-
way ANOVA, Bonferroni test).
Binding of (S)-2 in the PPARg LBD
Enantiomer (S)-2 also bound to PPARg assuming the typical
position of partial agonists, although more shifted toward the
appears to mediate at least
some of the undesirable side ef-
fects of chronic PPARg activation.
In this study, using amide hydro-
gen/deuterium exchange kinet-
ics, the authors showed that
binding of the PPARg partial ag-
onist 3 significantly decreases
flexibility of the loop region
around S273 and that both rosi-
glitazone and 3 decrease the
mobility of helix 3 and the adja-
cent b-sheet, which are sites for
Figure 6. 2FoÀFc omit map calculated around A) (R)-2 (green) and B) (S)-2 (yellow) in complex with PPARg.
potential interaction with co-reg-
ChemMedChem 0000, 00, 0 – 0
www.chemmedchem.org
5
ꢀ 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
These are not the final page numbers! ÞÞ

Full Papers
Figure 7. Hydrogen bond network of A) (R)-2 (green) and B) (S)-2 (yellow) in complex with PPARg.
PPARg full agonists. A similar behavior could also be suggested
for (S)-2, which binds to the same region of PPARg LBD, al-
though slightly shifted with respect to (R)-2.
Effects of (R)-2 and (S)-2 on skeletal muscle chloride
conductance
Statins and PPAR agonists such as fibrates are widely used in
therapy to treat hyperlipidemia. However, clinical evidence has
shown drug-associated skeletal muscle disorders ranging from
myalgia to severe myopathy.[35] This type of myopathy has
been characterized biochemically and morphologically, and
there are a few mechanistic hypotheses.[35] Our previous stud-
ies allowed us to identify the target of the skeletal muscle
damage; in fact, both statins and fenofibrate are able to
reduce resting chloride conductance (gCl), sustained by the
voltage-gated chloride channel ClC-1, a parameter that con-
trols sarcolemma electrical stability and functional processes
such as excitation and contraction. A reduced gCl causes disor-
ders related to abnormal action potential firing.[36,51] On this
basis, we decided to test (R)-2 and (S)-2 to assess the possibili-
ty that these fibrate-like drugs were able to interfere with the
chloride ion flux of skeletal muscle membrane, as previously
reported for similar compounds.[36,51,52]
The effects of in vitro application of these compounds were
evaluated on extensor digitorum longus (EDL) muscle in terms
of resting membrane ionic conductances and excitability pa-
rameters. As can be seen in Figure 9, both enantiomers signifi-
cantly decreased gCl in a concentration-dependent manner,
with (S)-2 more potent than (R)-2. In particular, the application
of (S)-2 produced a significant (p<0.001, Student’s t test) 32
and 43% block of gCl at 10 and 100 mm, respectively, relative
to the control value. (R)-2, instead, produced a lower, although
significant (p<0.01, Student’s t test), decrease in gCl by 11 and
30% at 10 and 100 mm, respectively. The resting potassium con-
ductance (gK) was slightly and not significantly increased from
366Æ61 to 435Æ35 mScmÀ2 when a 100 mm dose of the less
potent R enantiomer was applied to the muscle bath solution.
The effects of the two enantiomers were completely reversible,
Figure 8. Superpositions of A) (R)-2 (green) and (S)-2 (yellow); B) (R)-2
(green) and 3 (cyan); C) (S)-2 (yellow) and 4 (magenta) into the LBD of
PPARg1. S245 of PPARg1 corresponds to S273 of PPARg2.
ulatory proteins. These findings suggest that ligand-induced
decreases in the dynamic nature of helix 3, the b-sheet, and
the CDK5 site, may “freeze” this region into a conformation
less favorable to CDK5-mediated phosphorylation. According
to this hypothesis, it is likely that (R)-2 decreases the dynamics
of helix 3, b-sheet, and CDK5 site, explaining the consequent
decrease of PPARg phosphorylation at S273. This could be the
principal molecular mechanism responsible for its antidiabetic
activity and lack of undesirable side effects typical of classical
&
ChemMedChem 0000, 00, 0 – 0
www.chemmedchem.org
6
ꢀ 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Full Papers
15) in the absence and presence of 100 mm (R)-2, respectively;
similar results, respectively 0.30Æ0.005 (n=15) and 0.29Æ
0.008 (n=14), were obtained in the absence and presence of
(S)-2 at 100 mm. Both compounds, therefore, did not interfere
with calcium homeostasis, thus indicating a lack of potential
toxicity at this level. Because intracellular calcium levels can
affect muscle gCl through modulation of the calcium-depen-
dent protein kinase C,[54] these results allow us to hypothesize
that the decrease in gCl induced by both molecules is likely
due to direct inhibition of the ClC-1 chloride channel.
Conclusions
In the present study we report an improved synthesis of the
metabolically active form (R)-2 of metaglidasen, a selective
PPARg modulator, which, even though repurposed as a drug
to treat gout due to its unique uricosuric properties, is still in-
vestigated as a useful agent for the treatment of type 2 diabe-
tes and hyperglycemia. Its activity and that of its enantiomer
on both PPARa and PPARg were carefully investigated. As pre-
viously reported, they showed similar activity at PPARg; there-
fore, with the aim to provide a molecular explanation for this
unusual lack of stereoselectivity, X-ray crystallography studies
with the PPARg LBD were performed for both enantiomers.
Both (R)-2 and (S)-2 bound to the same region of the LBD, al-
though slightly shifted with respect to each other, contributing
to a reduction in the dynamics of H3, b-sheet, and CDK5 site.
Interestingly, this binding mode seems to be common to all ef-
fective antidiabetic PPARg ligands able to inhibit CDK5-mediat-
ed phosphorylation of S273, a mechanism that is suggested to
be responsible, at least in part, for the improved insulin sensi-
tivity by these drugs.
Figure 9. Resting gCl measured in EDL muscle before and after in vitro appli-
cation of either (R)-2 or (S)-2. Data represent the meanÆSEM from 17–49
fibers (*p<0.01, Student’s t test).
because after the application of the normal solution, the gCl
value was similar to that recorded in the absence of drugs.
We also measured the excitability parameters of EDL muscle
fibers upon exposure to (R)-2 at 100 mm (Figure 10). Concur-
rently with the decrease in gCl we found an increase in mem-
brane excitability. Interestingly, we observed a significant de-
crease (15%) of the current (Ith) required to elicit the first
action potential (AP) together with a significant twofold in-
crease in the action potential latency (Lat) (p<0.01). However,
no significant differences were observed in the amplitude of
the AP and the maximum number (N) of spikes (data not
shown).
At this point, given that intracellular calcium homeostasis is
a critical determinant of muscle function,[53] in order to evalu-
ate the potential ability of the drug to induce muscle damage
by interfering with Ca2+ signaling machinery, we decided to
evaluate the effects of (R)-2 and (S)-2 on calcium homeostasis
of native rat skeletal muscle fibers. These effects were evaluat-
ed by fura-2 microspectrofluorimetric analysis. Both drugs,
tested in a concentration range of 10–100 mm, did not produce
any change in resting cytosolic calcium. Particularly, the ratio
value (I340/I380) was 0.31Æ0.004 (n=16) and 0.29Æ0.004 (n=
In contrast, unexpected results were obtained at PPARa. In
this case, (S)-2 exhibited a partial agonist activity, whereas (R)-2
displayed an antagonism profile. These results were confirmed
by the different effects on the PPARa-mediated gene expres-
sion of mitochondrial carnitine palmitoyltransferase 1 (CPT1),
which is a molecular component of the carnitine shuttle
system essential for the mitochondrial oxidation of fatty acids.
Given that no crystal structures of PPARa complexed with car-
boxylic antagonists are available
as a starting point for docking
analysis, X-ray structure determi-
nation of the crystal complexes
involving PPARa and each of the
two stereoisomers will be crucial
to confirm the obtained results.
These studies are still in progress
for PPARa.
Finally, we evaluated the
blocking activity of both (R)-2
and (S)-2 on skeletal muscle
chloride conductance, which is
related to myopathies frequently
Figure 10. Excitability parameters measured in EDL muscle after in vitro application of (R)-2 (100 mm). A) The cur-
rent (Ith) required to induce the first action potential (AP) was decreased by 15%, and B) the latency to obtain
the AP was significantly increased with respect to control. Data represent the meanÆSEM from 10–19 fibers
(*p<0.01, Student’s t test).
observed in treatment with lipid-
lowering drugs. The results
showed a more beneficial profile
ChemMedChem 0000, 00, 0 – 0
www.chemmedchem.org
7
ꢀ 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
These are not the final page numbers! ÞÞ

Full Papers
diethyl ether and washed with 2n HCl and brine. The organic layer
was dried over Na2SO4 and filtered. The solvent was distilled off to
afford a white solid that was crystallized from n-hexane to give (R)-
2-(4-chlorophenyl)-2-(3-trifluoromethyl)phenoxyacetic acid in 59%
yield (0.30 mmol). The mother liquors obtained from the first crys-
tallization of the diastereomeric salts were evaporated to dryness
and the residue was suspended in diethyl ether and washed with
2n HCl and brine. The organic layer was dried over Na2SO4, filtered
and evaporated to dryness. The white residue (0.4 mmol) was
poured into a boiling mixture of water (1.4 mL) and 96% ethanol
(2 mL) together with (R)-1-(2-naphthyl)ethylamine (0.2 mmol) and
KOH (0.2 mmol). The solution was cooled to room temperature;
after 12 h the resulting crystals were collected by filtration and
treated as described above to afford a white solid that was crystal-
lized from n-hexane to give (S)-2-(4-chlorophenyl)-2-(3-trifluorome-
thyl)phenoxyacetic acid in 67% yield (0.34 mmol).
for (R)-2; nevertheless, the activity of this isomer on skeletal
muscle function should not be overlooked in the current clini-
cal trials, especially when studying the long-term effects.
In conclusion, it appears that a more careful investigation of
the structural determinants of the metabolically active form of
metaglidasen could allow the identification of new analogues
with improved pharmacological properties, that is, a dual
PPARa/g activity associated with low inhibition of cytochrome
P450 and decreased blocking activity on skeletal muscle chlo-
ride conductance. In this perspective, particular attention must
be paid to the stereochemistry, which, as widely reported for
PPAR ligands, plays a key role in the biological activity of these
compounds.
Experimental Section
(S)-2: mp: 98–998C; [a]D = +91 (c=0.1, MeOH); ee=99% (methyl
ester, Chiralcel OD column, n-hexane/iPrOH 95:5 as a mobile
phase, flow rate: 0.5 mLminÀ1, detection: 254 nm; tR =10.2 min);
Anal. calcd for C15H10ClO3F3: C 54.48, H 3.05, found: C 54.55, H 3.31.
Chemistry
Melting points were determined in open capillaries on a Gallen-
kamp electrothermal apparatus and are uncorrected. Mass spectro-
metric data were recorded on an HP GC–MS 6890–5973 MSD spec-
trometer, electron impact 70 eV, equipped with HP chemstation.
1H NMR spectra were recorded in CDCl3 on a Varian Mercury 300
(300 MHz) spectrometer at room temperature. Chemical shifts (d)
are expressed as parts per million. The purity of all tested com-
pounds was >95%, as confirmed by combustion analysis carried
out with an Eurovector Euro EA 3000 model analyzer. Optical rota-
tions were measured with a PerkinElmer 341 polarimeter at room
(R)-2: mp: 99–1008C; [a]D =À90.5 (c=0.1, MeOH); ee=98%
(methyl ester, Chiralcel OD column, n-hexane/iPrOH 95:5 as
a mobile phase, flow rate: 0.5 mLminÀ1, detection: 254 nm; tR =
11,4 min); Anal. calcd for C15H10ClO3F3: C 54.48, H 3.05, found: C
54.35, H 3.11.
Resolution of 4-chloromandelic acid. (R)-4-Chloromandelic acid
was obtained as reported.[40] Briefly, (R)-1-phenylethylamine (2.0 g,
16.5 mmol) was added to a solution of racemic 4-chloromandelic
acid (3.0 g, 16.1 mmol) in 20 mL of 95% ethanol. The resulting mix-
ture was held at reflux for 15 min and then slowly cooled to room
temperature. A crystalline salt (2.4 g) with [a]D =À33 (c=1.0, EtOH)
was collected by filtration and recrystallized from 95% ethanol
(15 mL). The salt obtained (1.8 g), having [a]D =À44 (c=1.0, EtOH),
was recrystallized again from 95% ethanol (10 mL) to give the (R,R)
diastereomeric salt (1.0 g) with [a]D =À49 (c=1.0, EtOH). This salt
was treated with 2n HCl and extracted (3ꢁ75 mL) with diethyl
ether. The organic phase was washed with brine, dried over
Na2SO4 and filtered. The solvent was evaporated to dryness to
afford (R)-4-chloromandelic acid as a white solid (0.6 g). (S)-4-Chlor-
omandelic acid was obtained in the same manner starting from
racemic 4-chloromandelic acid and (S)-1-phenylethylamine. GC–MS,
(methyl ester) m/z (%)=202 (3) [M+ +2], 200 (10) [M+], 141 (100),
temperature (208C): concentrations are expressed as g(100 mL)À1
.
The enantiomeric excesses of the final compounds were deter-
mined by HPLC analysis on Chiralcel AD or OD column (4.6 mm
i.d.ꢁ250 mm, Daicel Chemical Industries Ltd., Tokyo, Japan). Ana-
lytical liquid chromatography was performed on a PE chromato-
graph 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) and Alfa
Aesar (Karlsruhe, Germany) and were used without further purifica-
tion.
Synthesis of (R,S)-2-(4-chlorophenyl)-2-(3-trifluoromethyl)phe-
noxyacetic
acid
[(R,S)-2].
1-Iodo-3-trifluoromethylbenzene
(1.3 mmol), cesium carbonate (2 mmol) and copper(I) iodide
(0.05 mmol) were added to a solution of (R,S)-4-chloromandelic
acid (1 mmol) in butyronitrile (2 mL) under argon atmosphere. The
resulting mixture was stirred at 1108C for 72 h. Then the solvent
was distilled off and the solid was poured into water and washed
twice with ethyl acetate. The aqueous phase was acidified up to
pH 5–6 with 10% citric acid and extracted (3ꢁ10 mL) with ethyl
acetate. The organic layer was washed with brine, dried over
Na2SO4 and filtered. The solvent was evaporated to dryness to
afford a solid that was crystallized from n-hexane. The title com-
pound was obtained as a white solid in 40% yield; mp: 99–1008C;
GC–MS, (methyl ester) m/z (%)=346 (1) [M+ +2], 344 (2) [M+], 285
1
77 (72); H NMR: d=7.33–7.40 (4H, m, aromatics), 5.21 (1H, s, CH),
4.67 ppm (2H, b, OH+COOH, D2O exchanged).
(S)-4-Chloromandelic acid. [a]D = +135 (c=1.0, MeOH); ee=99%
(methyl ester, Chiralcel AD column, n-hexane/iPrOH 95:5 as
a mobile phase, flow rate: 0.5 mLminÀ1, detection: 254 nm; tR =
16,5 min).
(R)-4-Chloromandelic acid. [a]D =À135 (c=1.0, MeOH); ee=99%
(methyl ester, Chiralcel AD column, n-hexane/iPrOH 95:5 as
a mobile phase, flow rate: 0.5 mLminÀ1, detection: 254 nm; tR =
17,8 min).
1
(25), 183 (100), 155 (83); H NMR: d=7.04–7.54 (9H, m, 8 aromat-
ics+COOH, D2O exchanged), 5.65 ppm (1H, s, CH).
Synthesis of partially enriched (S)-2 and (R)-2. These compounds
were synthesized by using the method described above for (R,S)-2,
but starting from (S)- and (R)-4-chloromandelic acids, respectively.
The reaction was carried out at 758C instead of 1108C.
Resolution of (R,S)-2. The racemate (1 mmol) was poured into
a boiling mixture of water (3.5 mL) and 96% ethanol (5 mL) togeth-
er with (S)-1-(2-naphthyl)ethylamine (0.5 mmol) and KOH
(0.5 mmol). The solution was cooled to room temperature; after
12 h the resulting crystals were collected by filtration and recrystal-
lized using a mixture of water (1.75 mL) and ethanol 968 (3.75 mL).
The so obtained crystalline diastereomeric salt was suspended in
(S)-2: mp: 99–1008C; [a]D = +45 (c=0.1, MeOH); ee=77% (methyl
ester, Chiralcel OD column, n-hexane/iPrOH 95:5 as a mobile
phase, flow rate: 0.5 mLminÀ1, detection: 254 nm; tR =10,2 min).
&
ChemMedChem 0000, 00, 0 – 0
www.chemmedchem.org
8
ꢀ 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Full Papers
(R)-2: mp: 98–1008C; [a]D =À16 (c=0.1, MeOH); ee=30% (methyl
ester, Chiralcel OD column, n-hexane/iPrOH 95:5 as a mobile
phase, flow rate: 0.5 mLminÀ1, detection: 254 nm; tR =11,4 min).
a minimal thymidine kinase promoter that is adjacent to the luci-
ferase gene were described previously.[59]
Cell culture and transfections. Human hepatoblastoma cell line
HepG2 (Interlab Cell Line Collection, Genoa, Italy) was cultured in
minimum essential medium (MEM) containing 10% heat-inactivat-
ed fetal bovine serum, penicillin G (100 UmLÀ1), and streptomycin
sulfate (100 mgmLÀ1) at 378C in a humidified atmosphere of 5%
CO2. For transactivation assays, 105 cells per well were seeded in
a 24-well plate and transfections were performed after 24 h with
CAPHOS, a calcium-phosphate method, according to the manufac-
turer’s guidelines. Cells were transfected with expression plasmids
encoding the fusion protein Gal4–PPARa LBD or Gal4–PPARg LBD
(30 ng), pGal5TKpGL3 (100 ng), pCMVbgal (250 ng). Four hours
after transfection, cells were treated for 20 h with the indicated li-
gands in triplicate. Luciferase activity in cell extracts was then de-
termined by a luminometer (VICTOR3 V Multilabel Plate Reader, Per-
kinElmer). b-Galactosidase activity was determined using ortho-ni-
trophenyl-b-d-galactopyranoside as described previously.[60] All
transfection experiments were repeated at least twice.
Protein expression, purification and crystallization
The LBD of PPARg was expressed in BL21 DE3 cells as N-terminal
His-tagged protein using a peT28 vector. The protein was purified
onto a Ni2+-nitriloacetic acid column (GE Healthcare) as previously
described.[55] Crystals of apo-PPARg were obtained by the vapor dif-
fusion method at 188C using a sitting drop made by mixing 2 mL
of protein solution (10 mgmLÀ1, in 20 mm Tris, pH 8.0, 1 mm TCEP
and 0.5 mm EDTA) with 2 mL of reservoir solution (0.8m sodium cit-
rate and 0.15m Tris, pH 8.0). Crystals were soaked for two weeks in
a storage solution (1.2m sodium citrate and 0.15m Tris, pH 8.0)
containing the ligand (0.25 mm). The ligand was dissolved in
DMSO and added to the storage solution so that the final concen-
tration of DMSO was 0.5%. The storage solution with glycerol
[20% (v/v)] was used as a cryoprotectant. Crystals belong to the
space group C2 with cell parameters shown in Table S1, Supporting
Information.
Real-time PCR. HepG2 cells were incubated for 24 h with Wy14,643
(50 mm), (R)-2 (50 mm), or (S)-2 (50 mm), starting 24 h after having
been depleted of serum. Total RNA was extracted from 1ꢁ106
cells; reverse transcription and real-time PCR were performed as
previously reported.[61,62] Assays-on-demand for human CPT1, CAC,
and human actin (cat. nos. Hs00912671m1, Hs01088810g1, and
Hs00357333g1, respectively) were purchased from Applied Biosys-
tems. All transcript levels were normalized against the b-actin ex-
pression levels.
Structure determination
X-ray crystallographic data were collected at 100 K under a nitrogen
stream using synchrotron radiation (beamline ID29 at ESRF, Greno-
ble, France). The diffracted intensities were processed using the
programs MOSFLM and SCALA.[56] Structure solution was per-
formed with AMoRe,[57] using the coordinates of PPARg/(S)-2-(4-
phenylphenoxy)-3-phenylpropanoic acid (LT175; PDB code 3B3K)
as the starting model. The coordinates were then refined with the
Phenix 1.8.4 package.[58] All data between 58 and 2.6 ꢂ were in-
cluded for both structures. The statistics of crystallographic data
and refinement are summarized in Table S1, Supporting Informa-
tion. The coordinates of PPARg/(R)-2 and PPARg/(S)-2 have been
deposited in the Brookhaven Protein Data Bank (PDB) with access
codes 4PVU and 4PWL, respectively.
Western blot analysis. HepG2 cells were incubated for 48 h with
Wy14,643 (50 mm), (R)-2 (50 mm), or (S)-2 (50 mm), starting 24 h
after having been depleted of serum. Proteins were electroblotted
onto nitrocellulose membranes (Bio-Rad) and subsequently treated
with anti-CPT1 (ARP44796_P050, Aviva Systems Biology), anti-
CAC,[47] and anti-b-actin (sc-58619, Santa Cruz) antibodies. The im-
munoreaction was detected by the ECL plus system (Amersham).
Electrophysiological recordings of resting membrane ionic conductan-
ces and excitability parameters. All experiments were performed in
accordance with the Italian Guidelines for the use of laboratory ani-
mals, which conform with the European Community Directive pub-
lished in 1986 (86/609/EEC). The electrophysiological recordings
were done in vitro on the extensor digitorum longus (EDL) muscle
dissected from adult male Wistar rats (Charles River Laboratories,
Calco, Italy) under urethane anesthesia (1.2 gkgÀ1 i.p.). Soon after
the dissection, rats still anaesthetized were euthanized with a ure-
thane overdose. The EDL muscles were immediately placed in
a 25 mL bath chamber, maintained at 308C, and perfused with
normal or chloride-free physiological solution (gassed with 95% O2
and 5% CO2; pH 7.2–7.3).[51] The normal (chloride containing) phys-
iological solution had the following composition: 148 mm NaCl,
4.5 mm KCl, 2 mm CaCl2, 1 mm MgCl2, 12 mm NaHCO3, 0.44 mm
NaH2PO4, 5.5 mm glucose. The chloride-free solution was made by
equimolar substitution of methylsulfate salts for NaCl and KCl and
nitrate salts for CaCl2 and MgCl2. Using the two intracellular micro-
electrodes technique, in current clamp mode, the membrane resist-
ance (Rm) and the fiber diameter were calculated. These parame-
ters were obtained by injecting a hyperpolarizing constant square-
wave current pulse (100 ms duration) into the muscle fiber
through the current electrode and by recording the resulting volt-
age deflection with a second microelectrode inserted at two dis-
tances from the current electrode. The current pulse generation,
the acquisition of the voltage records and the calculation of the
Isothermal titration calorimetry
ITC experiments were performed at 258C using a MicroCal ITC200
microcalorimeter (MicroCal Inc., Northampton, MA, USA) as previ-
ously described.[42] Briefly, the protein solution (50 mm) was placed
in the sample cell, and the ligand solution (500 mm) was loaded
into the syringe injector. The titrations involved 19 injections of
2 mL at 180 s intervals. The syringe stirring speed was set at
1000 rpm. Reference titrations of ligands into buffer were used to
correct for heats of dilution. All ITC measurements were carried
out at 298 K in 20 mm HEPES buffer at pH 8. The thermodynamic
data were processed with Origin 7.0 software provided by Micro-
Cal.
Biological methods
Reference compounds, media, and other cell culture reagents were
purchased from Sigma–Aldrich (Milan, Italy).
Plasmids. The expression vectors expressing the chimeric receptors
containing the yeast Gal4 DNA binding domain fused to the
human PPARa or PPARg ligand binding domain (LBD), and the re-
porter plasmid for these Gal4 chimeric receptors (pGal5TKpGL3)
containing five repeats of the Gal4 response elements upstream of
ChemMedChem 0000, 00, 0 – 0
www.chemmedchem.org
9
ꢀ 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
These are not the final page numbers! ÞÞ

Full Papers
[2] J. Berger, D. E. Moller, Annu. Rev. Med. 2002, 53, 409–435.
[3] S. A. Kliewer, S. S. Sundseth, S. A. Jones, P. J. Brown, G. B. Wisely, C. S.
Koble, P. Devchand, W. Wahli, T. M. Willson, J. M. Lenhard, J. M. Leh-
mann, Proc. Natl. Acad. Sci. USA 1997, 94, 4318–4323.
[4] R. T. Nolte, G. B. Wisely, S. Westin, J. E. Cobb, M. H. Lambert, R. Kuroka-
wa, M. G. Rosenfeld, T. M. Willson, C. K. Glass, M. V. Milburn, Nature
1998, 395, 137–143.
[5] R. T. Gampe, Jr., V. G. Montana, M. H. Lambert, A. B. Miller, R. K. Bledsoe,
M. V. Milburn, S. A. Kliewer, T. M. Willson, H. E. Xu, Mol. Cell 2000, 5,
545–555.
[6] T. M. Willson, P. J. Brown, D. D. Sternbach, B. R. Henke, J. Med. Chem.
2000, 43, 527–550.
fiber constants were done in real time under computer control as
described elsewhere.[51] The reciprocal of Rm from each fiber in
normal physiological solution was the total membrane conduc-
tance (gm), and the same parameter measured in chloride-free so-
lution was the potassium conductance (gK). The mean chloride
conductance (gCl) was estimated as the mean gm minus the mean
gK value. The tested compounds were dissolved in DMSO and ap-
plied in vitro on muscle bath. Resting gm and gK were measured
before and 30 min after addition of increasing concentrations of
each compound. The maximal concentration of DMSO used (0.5%)
was without effect on the parameters studied.
[7] A Unified Nomenclature System for the Nuclear Receptor Superfamily:
Nuclear Receptors Nomenclature Committee, Cell 1999, 97, 161–163.
[8] J. Berger, P. Bailey, C. Biswas, C. A. Cullinan, T. W. Doebber, N. S. Hayes, R.
Saperstein, R. G. Smith, M. D. Leibowitz, Endocrinology 1996, 137, 4189–
4195.
[9] N. D. Oakes, C. J. Kennedy, A. B. Jenkins, D. R. Laybutt, D. J. Chisholm,
E. W. Kraegen, Diabetes 1994, 43, 1203–1210.
[10] R. W. Nesto, D. Bell, R. O. Bonow, V. Fonseca, S. M. Grundy, E. S. Horton,
M. L. Le Winter, D. Porte, C. F. Semenkovich, S. Smith, L. H. Young, R.
Kahn, Diabetes Care 2004, 27, 256–263.
[11] S. E. Kahn, B. Zinman, J. M. Lachin, S. M. Haffner, W. H. Herman, R. R.
Holman, B. G. Kravitz, D. Yu, M. A. Heise, R. P. Aftring, G. Viberti, Diabetes
Care 2008, 31, 845–851.
[12] S. Rocchi, F. Picard, J. Vamecq, L. Gelman, N. Potier, D. Zeyer, L. Dubu-
quoy, P. Bac, M. F. Champy, K. D. Plunket, L. M. Leesnitzer, S. G. Blan-
chard, P. Desreumaux, D. Moras, J. P. Renaud, J. Auwerx, Mol. Cell 2001,
8, 737–747.
The excitability parameters of sampled fibers were determined by
recording the intracellular membrane potential response to
square-wave depolarizing constant current pulses. In each fiber the
membrane potential was set by a steady holding current to
À80 mV, before passing the depolarizing pulse. The current intensi-
ty was gradually increased until the depolarization was just suffi-
cient to elicit a single action potential, and then further increased
to generate two or more action potentials. In this way, it was possi-
ble to record and measure these parameters: the minimum current
intensity that would elicit a single action potential (Ith), the mem-
brane potential at which a single action potential could be elicited
(Th), the amplitude of action potential (AP), the latency, that is, the
maximal delay from the beginning of the current pulse to the
onset of the spike (Lat) and the maximum number of action poten-
tials that the muscle fibers could generate (N spikes).
[13] J. P. Berger, A. E. Petro, K. L. Macnaul, L. J. Kelly, B. B. Zhang, K. Richards,
A. Elbrecht, B. A. Johnson, G. Zhou, T. W. Doebber, C. Biswas, M. Parikh,
N. Sharma, M. R. Tanen, G. M. Thompson, J. Ventre, A. D. Adams, R.
Mosley, R. S. Surwit, D. E. Moller, Mol. Endocrinol. 2003, 17, 662–676.
[14] P. Misra, R. Chakrabarti, R. K. Vikramadithyan, G. Bolusu, S. Juluri, J. Hiriy-
an, C. Gershome, A. Rajjak, P. Kashireddy, S. Yu, S. Surapureddi, C. Qi, Y.-
J. Zhu, M. S. Rao, J. K. Reddy, R. Ramanujam, J. Pharmacol. Exp. Ther.
2003, 306, 763–771.
[15] H. Minoura, S. Takeshita, M. Ita, J. Hirosumi, M. Mabuchi, I. Kawamura, S.
Nakajima, O. Nakayama, H. Kayakiri, T. Oku, A. Ohkubo-Suzuki, M. Fuka-
gawa, H. Kojo, K. Hanioka, N. Yamasaki, T. Imoto, Y. Kobayashi, S. Mutoh,
Eur. J. Pharmacol. 2004, 494, 273–281.
Calcium microspectrofluorimetric analysis. Calcium fluorescence
measurements were performed on tendon-to-tendon isolated EDL
muscle fibers using a QuantiCell 900 integrated imaging system
(VisiTech International Ltd., Sunderland, UK) as previously de-
scribed.[63] Fura-2-loaded muscle fibers were mounted in a modified
glass-bottomed RC-27NE experimental chamber (Warner Instru-
ment Corp., Hamden, USA), setting the sarcomere length at 2.4–
2.5 mm, and then placed on the stage of an inverted Eclipse TE300
microscope with a 40ꢁ Plan-Fluor objective (Nikon, Japan). Pairs of
background-subtracted images of the fura-2 fluorescence (510 nm)
excited at 340 and 380 nm were acquired at rest, and pixel-to-pixel
ratiometric images were calculated for each muscle fiber. Resting
ratio values were used to monitor cytosolic calcium level.
[16] J. J. Acton III, R. M. Black, A. B. Jones, D. E. Moller, L. Colwell, T. W. Doeb-
ber, K. L. MacNaul, J. Berger, H. B. Wood, Bioorg. Med. Chem. Lett. 2005,
15, 357–362.
[17] J. F. Dropinski, T. Akiyama, M. Einstein, B. Habulihaz, T. Doebber, J. P.
Berger, P. T. Meinke, G. Q. Shi, Bioorg. Med. Chem. Lett. 2005, 15, 5035–
5038.
Glossary
[18] A. Reifel-Miller, K. Otto, E. Hawkins, R. Barr, W. R. Bensch, C. Bull, S. Dana,
K. Klausing, J.-A. Martꢃn, R. Rafaeloff-Phail, C. Rafizadeh-Montrose, G.
Rhodes, R. Robey, I. Rojo, D. Rungta, D. Snyder, K. Wilbur, T. Zhang, R.
Zink, A. Warshawsky, J. T. Brozinick, Mol. Endocrinol. 2005, 19, 1593–
1605.
[19] J. A. Martꢃn, D. A. Brooks, L. Prieto, R. Gonzꢄlez, A. Torrado, I. Rojo, B.
Lꢅpez de Uralde, C. Lamas, R. Ferritto, M. D. Martꢃn-Ortega, J. Agejas, F.
Parra, J. R. Rizzo, G. A. Rhodes, R. L. Robey, C. A. Alt, S. R. Wendel, T. Y.
Zhang, A. Reifel-Miller, C. Montrose-Rafizadeh, J. T. Brozinick, E. Hawkins,
E. A. Misener, D. A. Briere, R. Ardecky, J. D. Fraser, A. M. Warshawsky,
Bioorg. Med. Chem. Lett. 2005, 15, 51–55.
PPAR: peroxisome proliferator-activated receptor; LBD: ligand bind-
ing domain; TZDs: thiazolidinediones; CAC: carnitine/acylcarnitine
carrier; CPT1: carnitine palmitoyltransferase 1; ITC: isothermal titra-
tion calorimetry; CDK5: cyclin-dependent kinase 5; ClC-1: chloride
channel 1.
Acknowledgements
This work was accomplished thanks to financial support from
the Cariplo Foundation (file 2009-2727) and the Ministero dell’Is-
truzione, dell’Universitꢀ e della Ricerca [MIUR 2009K7R7NA (F.L.),
MIUR 2010W7YRLZ_003 (A.L.)].
[20] E. Burgermeister, A. Schnoebelen, A. Flament, J. Benz, M. Stihle, B. Gsell,
A. Rufer, A. Ruf, B. Kuhn, H. P. Mꢆrki, J. Mizrahi, E. Sebokova, E. Niesor,
M. Meyer, Mol. Endocrinol. 2006, 20, 809–830.
[21] T. Allen, F. Zhang, S. A. Moodie, L. E. Clemens, A. Smith, F. Gregoire, A.
Bell, G. E. Muscat, T. A. Gustafson, Diabetes 2006, 55, 2523–2533.
[22] M. C. Carmona, K. Louche, B. Lefebvre, A. Pilon, N. Hennuyer, V. Audi-
not-Bouchez, C. Fievet, G. Torpier, P. Formstecher, P. Renard, P. Lefebvre,
C. Dacquet, B. Staels, L. Casteilla, L. Pꢇnicaud, Consortium of the French
Ministry of Research and Technology, Diabetes 2007, 56, 2797–2808.
[23] M. K. Kim, Y. N. Chae, H. S. Kim, S.-h. Choi, M. H. Son, S. H. Kim, J. K. Kim,
H. S. Moon, S. K. Park, Y. A. Shin, J. G. Kim, C. H. Lee, J. I. Lim, C. Y. Shin,
Arch. Pharmacal Res. 2009, 32, 721–727.
Keywords: X-ray crystallography · molecular recognition ·
PPAR modulators · receptors
[1] J. P. Berger, T. E. Akiyama, P. T. Meinke, Trends Pharmacol. Sci. 2005, 26,
244–251.
&
ChemMedChem 0000, 00, 0 – 0
www.chemmedchem.org
10
ꢀ 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Full Papers
[24] A. Motani, Z. Wang, J. Weiszmann, L. R. McGee, G. Lee, Q. Liu, J. Staun-
ton, Z. Fang, H. Fuentes, M. Lindstrom, J. Liu, D. H. Biermann, J. Jaen,
N. P. Walker, R. M. Learned, J. L. Chen, Y. Li, J. Mol. Biol. 2009, 386, 1301–
1311.
[25] F. M. Gregoire, F. Zhang, H. J. Clarke, T. A. Gustafson, D. D. Sears, S. Fave-
lyukis, J. Lenhard, D. Rentzeperis, L. E. Clemens, Y. Mu, B. E. Lavan, Mol.
Endocrinol. 2009, 23, 975–988.
Montanari, C. Di Giovanni, A. Paradiso, G. Pochetti, G. M. Simone, P. Tor-
torella, M. Crestani, F. Loiodice, J. Med. Chem. 2012, 55, 37–54.
[45] E. Calleri, G. Fracchiolla, R. Montanari, G. Pochetti, A. Lavecchia, F. Loio-
dice, A. Laghezza, L. Piemontese, G. Massolini, C. Temporini, J. Chroma-
togr. A 2012, 1232, 84–92.
[46] S. M. Rangwala, M. L. O’Brien, V. Tortorella, A. Longo, F. Loiodice, D. J.
Noonan, D. R. Feller, Chirality 1997, 9, 37–47.
[47] V. Iacobazzi, P. Convertini, V. Infantino, P. Scarcia, S. Todisco, F. Palmieri,
Biochem. Biophys. Res. Commun. 2009, 388, 643–647.
[48] C. Indiveri, V. Iacobazzi, A. Tonazzi, N. Giangregorio, V. Infantino, P. Con-
vertini, L. Console, F. Palmieri, Mol. Aspects Med. 2011, 32, 223–233.
[49] R. A. Laskowski, M. B. Swindells, J. Chem. Inf. Model. 2011, 51, 2778–
2786.
[50] J. H. Choi, A. S. Banks, J. L. Estall, S. Kajimura, P. Bostrçm, D. Laznik, J. L.
Ruas, M. J. Chalmers, T. M. Kamenecka, M. Blꢈher, P. R. Griffin, B. M. Spie-
gelman, Nature 2010, 466, 451–456.
[51] S. Pierno, G. M. Camerino, V. Cippone, J.-F. Rolland, J.-F. Desaphy, A.
De Luca, A. Liantonio, G. Bianco, J. D. Kunic, A. L. George, Jr., D. Conte
Camerino, Br. J. Pharmacol. 2009, 156, 1206–1215.
[26] W. S. Aronow, P. R. Harding, M. Khursheed, J. S. Vangrow, N. P. Papa-
george’s, J. Mays, Clin. Pharmacol. Ther. 1973, 14, 358–365.
[27] W. S. Aronow, P. R. Harding, M. Khursheed, J. S. Vangrow, N. P. Papa-
george’s, Clin. Pharmacol. Ther. 1973, 14, 371–373.
[28] T. Rauch, B. S. Hamilton, M. Tsutsumi (Boehringer Ingelheim Internation-
al GmbH), Int. PCT Pub. No. WO2013167554A1 20131114, 2013.
[29] B. S. Hamilton, T. Rauch, M. Tsutsumi (Boehringer Ingelheim Internation-
al GmbH), Int. PCT Pub. No. WO2012059416A1 20120510, 2012.
[30] A. Chandalia, H. J. Clarke, L. E. Clemens, B. Pandey, V. Vicena, P. Lee, B. E.
Lavan, F. M. Gregoire, PPAR Res. 2009, 706852.
[31] K. L. Luskey, J. Luo (Metabolex Inc., Hayward, USA), US Pat. No.
6.262.118B1, 2001.
[52] G. Fracchiolla, A. Laghezza, L. Piemontese, P. Tortorella, F. Mazza, R.
Montanari, G. Pochetti, A. Lavecchia, E. Novellino, S. Pierno, D. Conte
Camerino, F. Loiodice, J. Med. Chem. 2009, 52, 6382–6393.
[53] M. W. Berchtold, H. Brinkmeier, M. Mꢈntener, Physiol. Rev. 2000, 80,
1215–1265.
[32] Y. Zhu, P. Cheng, X. Chen, J. Ma, Z. Zhao (Metabolex Inc., Hayward,
USA), US Pat. No. 2007/0072858A1, 2007.
[33] D. R. Feller, V. S. Kamanna, H. A. Newman, K. J. Romstedt, D. T. Witiak, G.
Bettoni, S. H. Bryant, D. Conte-Camerino, F. Loiodice, V. Tortorella, J.
Med. Chem. 1987, 30, 1265–1267.
[54] S. Pierno, J. F. Desaphy, A. Liantonio, A. De Luca, A. Zarrilli, L. Mastrofran-
cesco, G. Procino, G. Valenti, D. Conte Camerino, J. Physiol. 2007, 584,
983–995.
[34] G. Bettoni, F. Loiodice, V. Tortorella, D. Conte-Camerino, M. Mambrini, E.
Ferrannini, S. H. Bryant, J. Med. Chem. 1987, 30, 1267–1270.
[35] C. Hodel, Toxicol. Lett. 2002, 128, 159–168.
[55] G. Pochetti, C. Godio, N. Mitro, D. Caruso, A. Galmozzi, S. Scurati, F. Loio-
dice, G. Fracchiolla, P. Tortorella, A. Laghezza, A. Lavecchia, E. Novellino,
F. Mazza, M. Crestani, J. Biol. Chem. 2007, 282, 17314–17324.
[56] A. G. W. Leslie, Newsl. Protein Crystallogr. 1992, 26, 27–33.
[57] J. Navaza, Acta Crystallogr. Sect. A 1994, 50, 157–163.
[58] P. D. Adams, P. V. Afonine, G. Bunkꢅczi, V. B. Chen, I. W. Davis, N. Echols,
J. J. Headd, L.-W. Hung, G. J. Kapral, R. W. Grosse-Kunstleve, A. J. McCoy,
N. W. Moriarty, R. Oeffner, R. J. Read, D. C. Richardson, J. S. Richardson,
T. C. Terwilliger, P. H. Zwart, Acta Crystallogr. Sect. D 2010, 66, 213–221.
[59] E. Raspꢇ, L. Madsen, A. M. Lefebvre, I. Leitersdorf, L. Gelman, J. Peinado-
Onsurbe, J. Dallongeville, J.-C. Fruchart, R. Berge, B. Staels, J. Lipid Res.
1999, 40, 2099–2110.
[60] T. Hollon, F. K. Yoshimura, Anal. Biochem. 1989, 182, 411–418.
[61] V. Iacobazzi, V. Infantino, P. Convertini, A. Vozza, G. Agrimi, F. Palmieri,
Biochem. Biophys. Res. Commun. 2009, 386, 186–191.
[62] P. Convertini, V. Infantino, F. Bisaccia, F. Palmieri, V. Iacobazzi, Biochem.
Biophys. Res. Commun. 2011, 404, 376–381.
[36] S. Pierno, M. P. Didonna, V. Cippone, A. De Luca, M. Pisoni, A. Frigeri,
G. P. Nicchia, M. Svelto, G. Chiesa, C. Sirtori, E. Scanziani, C. Rizzo, D.
De Vito, D. Conte Camerino, Br. J. Pharmacol. 2006, 149, 909–919.
[37] P. Cheng, J. Ma, X. Chen, Y. Zhu, Z. Zhao (Metabolex Inc., Hayward,
USA), Int. PCT Pub. No. WO2007/038277A2, 2007.
[38] M. Wolter, G. Nordmann, G. E. Job, S. L. Buchwald, Org. Lett. 2002, 4,
973–976.
[39] M. M. Cavalluzzi, C. Bruno, G. Lentini, A. Lovece, A. Catalano, A. Carocci,
C. Franchini, Tetrahedron: Asymmetry 2009, 20, 1984–1991.
[40] Q. He, Y.-F. Peng, S. Rohani, Chirality 2010, 22, 16–23.
[41] A. Pinelli, C. Godio, A. Laghezza, N. Mitro, G. Fracchiolla, V. Tortorella, A.
Lavecchia, E. Novellino, J.-C. Fruchart, B. Staels, M. Crestani, F. Loiodice,
J. Med. Chem. 2005, 48, 5509–5519.
[42] A. Laghezza, G. Pochetti, A. Lavecchia, G. Fracchiolla, S. Faliti, L. Piemon-
tese, C. Di Giovanni, V. Iacobazzi, V. Infantino, R. Montanari, D. Capelli, P.
Tortorella, F. Loiodice, J. Med. Chem. 2013, 56, 60–72.
[43] C. Temporini, G. Pochetti, G. Fracchiolla, L. Piemontese, R. Montanari, R.
Moaddel, A. Laghezza, F. Altieri, L. Cervoni, D. Ubiali, E. Prada, F. Loio-
dice, G. Massolini, E. Calleri, J. Chromatogr. A 2013, 1284, 36–43.
[44] L. Porcelli, F. Gilardi, A. Laghezza, L. Piemontese, N. Mitro, A. Azzariti, F.
Altieri, L. Cervoni, G. Fracchiolla, M. Giudici, U. Guerrini, A. Lavecchia, R.
[63] B. Fraysse, J. F. Desaphy, S. Pierno, A. De Luca, A. Liantonio, C. I. Mitolo,
D. Conte Camerino, FASEB J. 2003, 17, 1916–1918.
Received: October 24, 2014Published online on && &&, 0000
ChemMedChem 0000, 00, 0 – 0
www.chemmedchem.org
11
ꢀ 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
These are not the final page numbers! ÞÞ

FULL PAPERS
A. Laghezza, R. Montanari, A. Lavecchia,
L. Piemontese, G. Pochetti, V. Iacobazzi,
V. Infantino, D. Capelli, M. De Bellis,
A. Liantonio, S. Pierno, P. Tortorella,
D. Conte Camerino, F. Loiodice*
&& – &&
Unexpected activity: The metabolically
active forms of the selective PPARg
modulator metaglidasen and its enan-
tiomer were carefully investigated
toward PPARa. Unexpectedly, the
former showed antagonist activity,
whereas the latter behaved as a partial
agonist. Crystallographic studies on
PPARg were performed to gain more in-
sight on molecular-level interactions.
Both stereoisomers also showed block-
ing activity on skeletal muscle chloride
conductance.
On the Metabolically Active Form of
Metaglidasen: Improved Synthesis
and Investigation of Its Peculiar
Activity on Peroxisome Proliferator-
Activated Receptors and Skeletal
Muscles
&
ChemMedChem 0000, 00, 0 – 0
www.chemmedchem.org
12
ꢀ 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!