Different structures of the two peroxisome proliferator-activated receptor gamma (PPARγ) ligand-binding domains in homodimeric complex with partial agonist, but not full agonist

Article abstract of DOI:10.1016/j.bmcl.2015.04.076

We designed and synthesized acylsulfonamide derivative (3) as a human peroxisome proliferator-activated receptor gamma (hPPARγ) partial agonist by structural modification of hPPARγ full agonist 1. Co-crystallization of 3 with hPPARγ LBD afforded a homodim

Full text of DOI:10.1016/j.bmcl.2015.04.076

Bioorganic & Medicinal Chemistry Letters 25 (2015) 2639–2644
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Different structures of the two peroxisome proliferator-activated
receptor gamma (PPARc) ligand-binding domains in homodimeric
complex with partial agonist, but not full agonist
Masao Ohashi a, Takuji Oyama b, Hiroyuki Miyachi a,
⇑
a Graduate School of Medicine, Dentistry and Pharmaceutical Sciences, Okayama University, 1-1-1, Tsushima-Naka, Kita-ku, Okayama 700-8530, Japan
b Department of Biotechnology, Faculty of Life and Environmental Sciences, University of Yamanashi, 4-3-37 Takeda, Yamanashi 400-8510, Japan
a r t i c l e i n f o
a b s t r a c t
Article history:
We designed and synthesized acylsulfonamide derivative (3) as a human peroxisome proliferator-
Received 6 April 2015
Revised 22 April 2015
Accepted 24 April 2015
Available online 1 May 2015
activated receptor gamma (hPPAR
Co-crystallization of 3 with hPPAR
analysis at 2.1 Å resolution showed that one of the LBDs adopts a fully active structure identical with that
in the complex of rosiglitazone, a full agonist; however, the other LBD in the complex of 3 exhibits a dif-
ferent (non-fully active) structure. Interestingly, the apo-homodimer contained similar LBD structures.
Intrigued by these results, we surveyed reported X-ray crystal structures of partial agonists complexed
c
c
) partial agonist by structural modification of hPPAR
LBD afforded a homodimeric complex, and X-ray crystallographic
c full agonist 1.
Keywords:
PPAR
X-ray crystallographic analysis
Acylsulfonamide derivative
Homodimeric complex
with hPPAR
structure. In contrast, both LBDs in the rosiglitazone complex have the fully active structure. These results
suggest hPPAR partial agonists lack the ability to induce fully active LBD. The presence of at least one
c LBD homodimer, and identified several types of LBD structures distinct from the fully active
c
non-fully active LBD in the agonist complex may be a useful criterion to distinguish hPPAR
agonists from full agonists.
c partial
Ó 2015 Elsevier Ltd. All rights reserved.
Human peroxisome proliferator-activated receptors (hPPARs)
are ligand-dependent transcription factors belonging to the
nuclear receptor (NR) superfamily.1 Three subtypes of hPPARs, that
activation of hPPARc 7
partial agonists, which activate hPPAR
In this Letter, we present the design and synthesis of a hPPAR
.
Therefore, attention has turned to hPPAR
c
c
less than maximally.
c
is, hPPAR
a
, hPPARb/d, and hPPAR
c
have been identified to date.
partial agonist (3), as well as the results of X-ray structure determi-
nation of its complex with hPPAR LBD homodimer. The resulting
structure is compared with that of the apo-homodimer and with
the reported structures of hPPAR LBD homodimer complexes of
other hPPAR partial agonists, as well as full agonists, in order to
identify the structural basis of hPPAR partial agonist activity.
These subtypes are expressed differentially in a tissue-specific
c
manner,2 and contribute to pivotal biological responses.3 For
example, hPPAR
c
is mainly expressed in adipocytes and
c
macrophages,2 but is also found in various cancer cells, including
breast cancer, gastric cancer and colorectal cancer.4 It is a master
regulator of adipocyte differentiation,5 and also plays important
roles in insulin sensitivity, cell cycle regulation, differentiation,
inflammation, and immune responses.3 Therefore, modulators of
c
c
For the design of 3, we noted that the formation of a tight
hydrogen-bonding network is critical for full agonistic nature of
hPPAR
hPPAR
c
c
agonists, and so replacement of the carboxyl group of
agonist with other functional groups, such as acylsulfon-
hPPAR
c
are candidates for treatment of various diseases. For exam-
full agonists are widely
ple, thiazolidinedione (TZD) class hPPAR
c
amide, is one option to obtain hPPAR
Therefore, we focused on modification of the structure of the
hPPAR
Àselective agonist MEKT-21 (2), which is structurally
derived from hPPAR
c
partial agonists.8–10
used for the treatment of type 2 diabetes, and are also under clin-
ical trial for the treatment of Alzheimer’s disease.6 However, TZDs
possess a number of adverse effects, including significant weight
gain, peripheral edema, bone loss and increased risk of congestive
heart failure, which are considered to be associated with over-
c
c
Àpan agonist TIPP-703 (1), to remove the car-
boxyl group and introduce a sulfonyl group, as shown in Figure 1.
The synthetic route to 3 is shown in Figure 2(A). Salicylaldehyde
10 was n-propylated to give 5. Compound 5 was treated with
hydroxylamine HCl, and subsequent reduction with 10% Pd on
carbon afforded aminomethylbenzene derivative
6 as the
⇑
Corresponding author. Tel.: +81 086 251 7930.
E-mail address: miyachi@pharm.okayama-u.ac.jp (H. Miyachi).
hydrochloric acid salt. Compound was condensed with
6
http://dx.doi.org/10.1016/j.bmcl.2015.04.076
0960-894X/Ó 2015 Elsevier Ltd. All rights reserved.

2640
M. Ohashi et al. / Bioorg. Med. Chem. Lett. 25 (2015) 2639–2644
O
O
N
H
OH
O
O
O
N
H
OH
1: TIPP-703
N
O
N
O
O
O
O
2: MEKT-21
S
N
H
N
H
N
O
N
3
Figure 1. Structural development of hPPAR
c
pan agonist (1) to hPPAR
c
-selective partial agonist (3).
RLU
(B)
25000
(A)
O
O
O
O
H N
2
HCl
c
a
d
H
b
N
H
H
20000
15000
10000
5000
0
O
O
N
O
HO
N
4
5
6
7
O
O
O
O
O
O
O
S
O
N
H
N
H
OH
N
N
H
e
H
f
H
N
N
N
O
O
N
N
N
8
9
3
μM
-
0.01
0.1
1
10
30
30
10
3
pio
ros
Figure 2. Synthetic routes to 3, and dose-response relation for hPPARc agonistic activity of 3. Reagents and conditions: (a) n-PrI, K2CO3, DMF, 80 °C, 24 h, 89%; (b) (1)
NH2OHÁHCl, pyridine, EtOH, reflux, 12 h, (2) H2, 10% Pd-C, c.HCl, EtOH, rt, 2.5 h, 89% (2 steps); (c) 4-(pyrimidine-2-yl)benzoic acid, DEPC, TEA, DMF, rt, overnight, 64%; (d)
TiCl4, CH3OCHCl2, CH2Cl2, 0 °C to rt, overnight, 71%; (e) NaClO2, NaHPO4Á2H2O, 2-methyl-2-butene, t-BuOH, THF, H2O, rt, overnight, 90%; (f) benzenesulfonamide, EDC, DMAP,
DMF, rt, overnight, 27%.
Table 1
4-(2-pyrimidinyl)benzoic acid and the resulting product 7 was
regioselectively formylated (8), followed by Pinnick oxidation11
to afford benzoic acid 9. Compound 9 was condensed with ben-
zenesulfonamaide in the presence of EDC and DMAP to afford the
target compound 3,12 albeit in low yield.
Crystallographic data and refinement statistics of hPPARc-3
Complex
hPPAR
c
-3
Complex
hPPARc-3
Data collection
Space group
Refinement
Resolution range
(Å)
Rwork/Rfree
Number of atoms
Protein
Water
Ligand
Average B-factor
(Å)
Protein
C2
50.0–2.10
21.5/25.8
To characterize the hPPARc agonistic profile of 3, we examined
transactivation activity.13 Compound 3 exhibited apparent transac-
Unit cell constants
a (Å)
b (Å)
c (Å)
b (deg)
93.249
61.113
119.28
103.115
1
tivation activity and its maximal activity was about 50% of the
4165
72
76
maximal response of the positive controls (30
and 10 M rosiglitazone; both are well-established hPPAR
agonists), as shown in Figure 2(B). These results are consistent with
the idea that 3 is a hPPAR partial agonist.
To understand the structural basis of the hPPAR
character of 3, we solved the X-ray crystallographic structure of the
complex of 3 with the hPPAR LBD homodimer at 2.1 Å resolution.
l
M pioglitazone
l
c
full
Wavelength (Å)
c
Resolution (Å)
50.0–2.10
(2.18–2.10)
187868
(10921)
37883 (3523)
54.5
60.1
46.4
c
partial agonist
No. of total reflections
Water
Ligand
c
No. of unique
reflections
Completeness (%)
The crystal of the complex was obtained by soaking a crystal of the
homodimer in ligand solution. The crystallographic data are sum-
marized in Table 1, and the crystallographic structure is depicted
98.8 (93.2)
19.7 (2.7)
5.0 (3.1)
0.59–0.90
5.3(36.2)
r.m.s.d
I/
r
(I)
Bond lengths (Å)
Angles (deg)
PDB code
0.008
0.99
3WMH
in Figure 3(A–G). Each hPPARc LBD in the homodimer binds one
Redundancy
Mosaicity
Rmerge (%)
molecule of 3 (Fig. 3(A)). It is noteworthy that the structural fold-
ing of the LBD remains almost unchanged, except for the region
from the end of helix 11 (H11) to the C-terminal H12 region
(Fig. 3(B, D and E)); the dotted circle in Fig. 3(B) indicates the chan-
ged region). The bound 3 ligands in the two LBDs exhibit almost
the same three-dimensional structure (Fig. 3(C)), and are posi-
tioned in the Y2 and Y3 arms of the binding pocket.14 However,
the structural folds in one part of the homodimer structure
homodimer structure are different (Fig. 3(E)). Therefore, we tenta-
tively designated the former structure as the fully active form of
the hPPARc LBD.
To understand the nature and significance of the two forms of
the LBD, the binding mode of each form was investigated in detail.
Figure 3(F and H) shows the fully active form of the LBD, while
(Fig. 3(D)) are almost identical to those in the hPPAR
c LBD–rosigli-
tazone complex (Fig. 4(G)),15 whereas those in the other part of the

M. Ohashi et al. / Bioorg. Med. Chem. Lett. 25 (2015) 2639–2644
2641
H1
H9
(A)
(C)
(B)
H8
H10
H4
H11
H5
H2
H12
H7
H3
β2
H6
β3
β1
fully active form
non-fully active form
H2í
(D)
(E)
fully active form
non-fully active form
(F)
(G)
(H)
(I)
Figure 3. (A)–(I) Crystal structures of hPPARc LBD-3 homodimer (PDB: 3WMH). (A) Whole structure of hPPARc LBD-3 homodimer. Protein is represented as a red ribbon
model and 3 is depicted as a Wan Der Waals model. (B) The superimposed structures of both homodimer partners. The numbering of the second structure is also depicted. The
nomenclature of the helices is based on the RXR- crystal structure. (C) Zoomed view of aligned 3 in both LBDs. Compound 3 is depicted as a cylinder model. (D) Whole
structure of the full-agonist form LBD of hPPAR LBD–3 crystal. (E) Whole structure of the non-full-agonist form LBD of hPPAR LBD-3 crystal. (F) and (H) Zoomed views of
LBD-3 crystal. (G) and (I) Zoomed views of the binding mode of 3 in the non-full-agonist form LBD of hPPAR
a
c
c
the binding mode of 3 in the full-agonist form LBD of hPPAR
c
c
LBD-3 crystal.
Figure 3(G and I) shows the other form, which we tentatively des-
ignate as the non-fully active form.
Ser289, His323, Tyr327, His449 and Tyr473, is reported to be
formed in the interaction of full agonists with the LDB.16 Thus,
three of these five amino acids interactions are conserved at one
LDB in the complex with 3, and this appears to be sufficient to sup-
port the fully active LDB structure. However, in the non-fully active
LDB (Fig. 3(G and I)), another interaction was noted: the side chain
imidazole group of His266 is located close to the (pyrimidin-2-
yl)phenyl group of 3, resulting in hydrophobic interaction between
the imidazole side chain of His266 and phenyl group, and between
In the fully active form, compound 3 takes a U-shaped structure
with hydrogen-bonding interactions between (1) the sulfony-
lamino group hydrogen of 3 and sulfur atom of Cys285, and (2)
the amide group hydrogen of 3 and sulfur atom of Cys285. The
acylsulfonamide group of 3 is positioned near H11, and interacts
with the side-chains of Ser289, Tyr327, Lys367, His449 and
Phe363. A hydrogen-bonding network involving five amino acids,

2642
M. Ohashi et al. / Bioorg. Med. Chem. Lett. 25 (2015) 2639–2644
(A)
(B)
(C)
fully active form
non-fully active form
fully active form
non-fully active form
(D)
(E)
PDB:2PRG
(F)
(G)
(H)
fully active form
fully active form
Figure 4. (A)–(H) Crystal structures of apo form of hPPAR
model. (B) Whole structure of full-agonist form LBD of apo-hPPAR
superimposed structures of the full-agonist form LBDs of apo-hPPAR
LBD of apo-hPPAR and non-full-agonist form LBD of hPPAR LBD-3 homodimer crystals. (F) Whole structure of apo-hPPAR
The superimposed structures of both LBDs of hPPAR LBD-rosiglitazone homodimer. (H) Zoomed view of the binding mode of rosiglitazone in the full-agonist form LBD in
hPPAR LBD-rosiglitazone homodimer crystal.
c
LBD homodimer. (A) Whole structure of apo form of hPPAR
homodimer. (C) Whole structure of the non-full-agonist form LBD of apo-hPPAR
and hPPAR LBD-3 homodimer crystals. (E) The superimposed structures of the non-full-agonist form
LBD-rosiglitazone homodimer (PDB: 2PRG). (G)
c
LBD homodimer. Protein is represented as a ribbon
c
c
homodimer. (D) The
c
c
c
c
c
c
c
the imidazole side chain of His266 and the pyrimidinyl group.
These interactions cause the bound 3 to move to the right, so that
the distance from the phenylsulfonylaminocarbonyl group of 3 to
the side chains of Ser289, Tyr327, Lys367, His449 and Phe363
becomes longer. As a result, the hydrogen-bonding network is
weaker in the case of the non-fully active LBD, and this may mean
that the H12 region is not restricted to the appropriate location for
full activity.
each type of LDB (Fig. 4(D and E)), suggesting that partial agonists
lack the ability to induce fully active LBD.
On the other hand, the hPPARc LBD-rosiglitazone (a hPPARc full
agonist) complex also formed a homodimer in the crystal, but each
LBD was present in fully active form (Fig. 4(F) and (G)).14 This
result indicates that full agonists do induce structural change of
non-fully active hPPARc LBD to fully active LBD, presumably by
facilitating a tight hydrogen-bonding network of the LBD, espe-
Next, we considered whether these structural differences in
LBD folding might be induced by binding of 3. To examine this
question, we determined the X-ray crystallographic structure of
cially to the C-terminal region (H12) (Fig. 4(H)).
In order to ascertain the generality of this conclusion, we sur-
veyed the PDB database. Representative hPPAR
c LBD homod-
hPPAR
hPPAR
c
c
LBD apo-form (without ligand). As depicted in Figure 4,
LBD apo-form also forms a homodimer (Fig. 4(A)) in which
imer–partial agonist structures are depicted in Figure 5((A and
B): PDB: 2I4Z,17 (C and D): PDB: 2Q5S,18 (E and F): PDB: 2Q6R18).
In the case of PDB: 2I4Z, the homodimer contains a fully active
LBD and a non-fully active LBD, and the phenoxyacetic acid-type
ligand is bound to the fully active LBD. In the case of PDB: 2Q5S,
the homodimer also contains the two forms of LBD, and the
indole-2-carboxylic acid-type ligands are bound to both LBDs, as
in our case. In the case of PDB: 2Q6R, the homodimer contains
one LDB is in the fully active form (Fig. 4(B)), and the other is in a
non-fully active form (Fig. 4(C)). These results indicate that two
types of hPPAR
irrespective of the presence or absence of agonist. Indeed, the
apo-form hPPAR LBD homodimer and ligand-bound hPPAR
LBD homodimer did not show apparent structural differences in
c LBD structures are present in the crystal state,
c
c

M. Ohashi et al. / Bioorg. Med. Chem. Lett. 25 (2015) 2639–2644
2643
PDB:2I4Z
(A)
(B)
fully active form
non-fully active form
(C)
(D)
PDB:2Q5S
fully active form
non-fully active form
(E)
(F)
PDB:2Q6R
non-fully active form
non-fully active form
Figure 5. (A)-(F) Crystal structures of hPPARc LBD-hPPARc partial-agonist homodimers (PDB: 2I4Z, 2Q5S. 2Q6R). (A) Whole structure of PDB: 2I4Z. The bound ligand
structure is also depicted. (B) The superimposed structures of each LBD of PDB: 2I4Z. (C) Whole structure of PDB: 2Q5S. The bound ligand structure is also depicted. (D) The
superimposed structures of each LBD of PDB: 2Q5S. (E) Whole structure of PDB: 2Q6R. The bound ligand structure is also depicted. (F) The superimposed structures of each
LBD of PDB: 2Q6R.
two non-fully active LBDs, and the indole-2-carboxylic acid-type
ligands are bound to both LBDs. Based on these data, we hypothe-
Technology (MEXT). We also thank the SC-NMR Laboratory of
Okayama University for NMR measurements.
size that hPPARc partial agonists lack the ability to induce the fully
active LBD. At the cellular level, binding of an agonist to hPPAR
leads to heterodimerization with another nuclear receptor, reti-
c
References and notes
noid X receptor (RXR), not homodimerization. Therefore, the
1. Nuclear Receptors Nomenclature Committee Cell 1999, 97, 161.
2. Braissant, O.; Foufelle, F.; Scotto, C.; Dauca, M.; Wahli, W. Endocrinology 1996,
137, 354.
3. Tontonoz, P.; Spiegelman, B. M. Annu. Rev. Biochem. 2008, 77, 289.
4. Tachibana, K.; Yamasaki, D.; Ishimoto, K.; Doi, T. PPAR Res. 2008, 15 (Article ID
102737).
5. Rosen, E. D.; Hsu, C.-H.; Wang, X.; Sakai, S.; Freeman, M. W.; Gonzalez, F. J.;
Spiegelman, B. M. Genes Dev. 2002, 16, 22.
6. Geldmacher, D. S.; Fritsch, T.; McClendon, M. J.; Landreth, G. Arch. Neurol. 2011,
68, 45.
hPPAR
c
LBD homodimer structure might not necessarily reflect
, and to establish whether
the real functional structure of hPPAR
c
a similar situation exists in the heterodimer. Nevertheless, our
results at least suggest that the presence of a non-fully active
LBD in the agonist complex may be a useful criterion to distinguish
hPPARc partial agonists from full agonists.
7. Kawai, M.; Rosen, C. J. Nat. Rev. Endocrinol. 2010, 6, 629.
8. Taygerly, J. P.; McGee, L. R.; Rubenstein, S. M.; Houze, J. B.; Cushing, T. D.; Li, Y.;
Motani, A.; Chen, J. L.; Frankmoelle, W.; Ye, G.; Learned, M. R.; Jaen, J.; Miao, S.;
Timmermans, P. B.; Thoolen, M.; Kearney, P.; Flygare, J.; Beckmann, H.;
Weiszmann, J.; Lindstrom, M.; Walker, N.; Liu, J.; Biermann, D.; Wang, Z.;
Hagiwara, A.; Iida, T.; Aramaki, H.; Kitao, Y.; Shinkai, H.; Furukawa, N.; Nishiu,
J.; Nakamura, M. Bioorg. Med. Chem. 2013, 21, 979.
9. Liu, W.; Lau, F.; Liu, K.; Wood, H. B.; Zhou, G.; Chen, Y.; Li, Y.; Akiyama, T. E.;
Castriota, G.; Einstein, M.; Wang, C.; McCann, M. E.; Doebber, T. W.; Wu, M.;
Chang, C. H.; McNamara, L.; McKeever, B.; Mosley, R. T.; Berger, J. P.; Meinke, P.
T. J. Med. Chem. 2011, 54, 8541.
Acknowledgements
This work was supported in part by a grant for Platform for
Drug Discovery, Informatics, and Structural Life Science from the
Ministry of Education, Culture, Sports, Science and Technology,
Japan, Terumo Foundation and
Research <KAKENHI> (B) (grant number 26293027 to H.M.)
from the Ministry of Education, Culture, Sports, Science and
a Grant in-aid for Scientific

2644
M. Ohashi et al. / Bioorg. Med. Chem. Lett. 25 (2015) 2639–2644
10. Tsujimura, T.; Nagamine, J.; Sugaru, E.; Ono-Kishino, M.; Tokunaga, T.; Kitoh,
M.; Nagata, R.; Nakagawa, T.; Taiji, M. Biol. Pharm. Bull. 1991, 2009, 32.
11. George, A. K.; Michael, J. T. J. Org. Chem. 1980, 45, 1175.
14. Kuwabara, N.; Oyama, T.; Tomioka, D.; Ohashi, M.; Yanagisawa, J.; Shimizu, T.;
Miyachi, H. J. Med. Chem. 2012, 55, 893.
15. Nolte, R. T.; Wisely, G. B.; Westin, S.; Cobb, J. E.; Lambert, M. H.; Kurokawa, R.;
Rosenfeld, M. G.; Willson, T. M.; Glass, C. K.; Milburn, M. V. Nature 1998, 395,
137.
16. Einstein, M.; Akiyama, T. E.; Castriota, G. A.; Wang, C. F.; McKeever, B.; Mosley,
R. T.; Becker, J. W.; Moller, D. E.; Meinke, P. T.; Wood, H. B.; Berger, J. P. Mol.
Pharmacol. 2008, 73, 62.
17. Pochetti, G.; Godio, C.; Mitro, N.; Caruso, D.; Galmozzi, A.; Scurati, S.; Loiodice,
F.; Fracchiolla, G.; Tortorella, P.; Laghezza, A.; Lavecchia, A.; Novellino, E.;
Mazza, F.; Crestani, M. J. Biol. Chem. 2007, 282, 17314.
12. Physico-chemical properties of compound 3: 1H NMR (400 MHz, CDCl3) d 8.82 (d,
J = 5.2 Hz, 2H), 8.45 (d, J = 8.4 Hz, 2H), 8.10 (d, J = 8.4 Hz, 2H), 7.87 (d, J = 2.4 Hz,
1H), 7.82–7.78 (m, 3H), 7.58 (t, J = 7.4 Hz, 1H), 7.49 (t, J = 7.6 Hz, 2H), 7.23 (t,
J = 4.8 Hz, 1H), 6.90 (t, J = 6.0 Hz, 1H), 6.87 (d, J = 8.4 Hz, 1H), 4.63 (d, J = 6.0 Hz,
2H), 4.02 (t, J = 6.6 Hz, 2H), 1.91–1.82 (m, 2H), 1.06 (t, J = 7.4 Hz, 3H). 13C NMR
(100 MHz, DMSO-d6) d 165.87, 162.57, 160.04, 157.87, 139.64, 136.14, 133.42,
129.28, 129.02, 128.20, 127.78, 127.55, 127.20, 120.38, 110.82, 69.60, 37.75,
21.96, 10.43. HRMS (FAB, MH+) Calcd for C28H27N4O5S: 531.1702, Found:
531.1717. HPLC purity was estimated to be 97.9% by means of reversed-phase
HPLC, using a Pegasil ODS sp100 column (4.6 Â 250 mm, Senshu Chemical,
Japan) fitted on a JASCO HPLC system, with CH3CN: 0.1% TFA = 2:1 v/v as the
eluent and detection at 254 nm.
18. Bruning, J. B.; Chalmers, M. J.; Prasad, S.; Busby, S. A.; Kamenecka, T. M.; He, Y.;
Nettles, K. W.; Griffin, P. R. Structure 2007, 15, 1258.
13. Ohashi, M.; Gamo, K.; Tanaka, Y.; Waki, M.; Beniyama, Y.; Matsuno, K.; Wada,
J.; Tenta, M.; Eguchi, J.; Makishima, M.; Matsuura, N.; Oyama, T.; Miyachi, H.
Eur. J. Med. Chem. 2015, 27, 53.