Binding characterization, synthesis and biological evaluation of RXRα antagonists targeting the coactivator binding site

Article abstract of DOI:10.1016/j.bmcl.2016.07.027

Previously we identified the first retinoid X receptor-alpha (RXRα) modulators that regulate the RXRα biological function via binding to the coregulator-binding site. Here we report the characterization of the interactions between the hit molecule and RXRα through computational modeling, mutagenesis, SAR and biological evaluation. In addition, we reported studies of additional new compounds and identified a molecule that mediated the NF-κB pathway by inhibiting the TNFα-induced IκBα degradation and p65 nuclear translocation.

Full text of DOI:10.1016/j.bmcl.2016.07.027

Bioorganic & Medicinal Chemistry Letters 26 (2016) 3846–3849
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Binding characterization, synthesis and biological evaluation
of RXRa antagonists targeting the coactivator binding site
Dingyu Xu a,y, Shangjie Guo a,y, Ziwen Chen a, Yuzhou Bao a, Fengyu Huang a, Dan Xu a, Xindao Zhang a,
Zhiping Zeng a, Hu Zhou a, Xiaokun Zhang a,b, Ying Su a,b,
⇑
a School of Pharmaceutical Sciences, Xiamen University, Xiamen 361005, China
b Sanford Burnham Prebys Medical Discovery Institute, 10901 N. Torrey Pines Road, La Jolla, CA 92037, USA
a r t i c l e i n f o
a b s t r a c t
Article history:
Previously we identified the first retinoid X receptor-alpha (RXRa) modulators that regulate the RXRa
Received 13 May 2016
Revised 10 July 2016
Accepted 11 July 2016
Available online 14 July 2016
biological function via binding to the coregulator-binding site. Here we report the characterization of
the interactions between the hit molecule and RXR through computational modeling, mutagenesis,
SAR and biological evaluation. In addition, we reported studies of additional new compounds and iden-
a
tified a molecule that mediated the NF-
p65 nuclear translocation.
j
B pathway by inhibiting the TNF
a-induced IjBa degradation and
Keywords:
Nuclear receptors
Ó 2016 Elsevier Ltd. All rights reserved.
RXRa modulators
Coregulator-binding site
Coactivator-binding site ligand
TNFa/NF-jB pathway
Nuclear receptors (NRs) are a superfamily of transcription fac-
NRs,10–12 suggesting that targeting alternate binding areas on the
nuclear receptor surface may offer opportunities to mitigate side
effects and to discover new therapeutic stratregies.13–15 Among
these alternate sites, the coregulator-binding site has attracted
increasing attention. Compounds that bind to the coregulator-
binding site of some nuclear receptors,15–17 including estrogen
receptor, androgen receptor, vitamin D receptor and thyroid hor-
mone receptor, have been reported. Recently we reported the first
tors of which many function via a ligand-mediated mechanism.1,2
As nuclear receptors are essential players in various biological pro-
cesses such as differentiation, apoptosis, metabolism, and inflam-
mation and NRs are implicated in many diseases including
cancer, diabetes and obesity, NRs have become important drug tar-
gets.3,4 Members of the nuclear receptor superfamily share con-
served domains, including a N-terminal domain, a DNA-binding
domain (DBD) and a ligand-binding domain (LBD).5–7 The LBD
plays a crucial role in ligand-regulated nuclear receptor activities.
The LBD consists of a canonical ligand-binding pocket (LBP) for
the binding of small molecule ligands, a transactivation function
domain termed AF-2 composed of helix 12 of the LBD, a coregula-
tor-binding surface groove, and a dimerization surface. A well-
accepted mechanism for ligand-mediated nuclear receptor activi-
ties is that ligand binds to the LBP to induce a major conforma-
tional change, converting the corepressor-binding site into a
coactivator-binding site and triggering a cascade of events that
lead to biological activities. Therefore, many nuclear receptor drugs
are developed to target the LBP.8,9 However, drugs acting by bind-
ing to the LBP are associated with undesirable side effects. Protein
crystallographic studies have revealed various alternate sites on
example of an RXRa modulator that acts via the coregulator-bind-
ing site.18 The reported binder, 23 (Fig. 1A) was identified through
employing a docking-based virtual screening approach. Various
RXRa mutants were studied to demonstrate that the identified bin-
der, 23 does not bind to the LBP. Modeling and mutagenesis studies
further show that 23 binds directly to the coregulator-binding sur-
face. 23 could regulate the biological functions of tRXR
minally truncated form of RXR that is overexpressed in many
cancer cells and is implicated in diseases.11,19,20 23 inhibits tumor
necrosis factor-alpha (TNF )-induced interaction of tRXR with
the p85 subunit of phosphatidylinositol-4,5-bisphosphate 3-
a, an N-ter-
a
a
a
a
kinase (PI3K), resulting in the inhibition of AKT activation
in vitro and the induction of apoptosis. These results demonstrate
the feasibility of targeting the alternate binding sites on the surface
of RXRa
for therapeutic intervention.21
Here we describe the further characterization of the binding
nature of 23 and some important features of the structure–activity
relationships (SAR) resulting from molecular modeling, biological
⇑
Corresponding author.
E-mail address: ysu@sbpdiscovery.org (Y. Su).
These authors contribute equally to this work.
y
http://dx.doi.org/10.1016/j.bmcl.2016.07.027
0960-894X/Ó 2016 Elsevier Ltd. All rights reserved.

D. Xu et al. / Bioorg. Med. Chem. Lett. 26 (2016) 3846–3849
3847
A
B
Figure 2. The antagonist effect of 23 and 24 on the transactivation activity of RXR
a
or RXR
a
-E453A. HEK-293T cells cotransfected with pG5-Luc, RXR
a
or mutant
E453A expression vector were treated with 9-cis-RA (10ꢀ7 M), and the indicated
concentration of 23 or 24 for 12 h.
Figure 1. (A) Structure 23. (B) Binding model of 23 in the coactivator-binding site of
RXRa. Protein is shown in ribbon diagram and 23 and the interacting side chains are
in shown in stick representation.
testing and chemical synthesis. Based on the obtained structural
insights, we designed and synthesized a series of new molecules.
Biological testing of the new molecules led to the identification
of a novel compound with new biological function.
To characterize the binding nature of 23 to the protein, we first
investigated the potential binding mode of 23 in the coactivator-
binding region using the Glide docking program from
Schrodinger.22 The 10 top-scored docking modes were visually
evaluated and one docking mode was intuitively selected as the
binding mode shown in Figure 1B. In this mode, 23 sits in the co-
activator binding groove consisting of Phe277 and Val280 of H3,
Phe289 of L2, Val298 and Leu301 of H4 and Phe450 of H12
(Fig. 1B). 23 interacts with RXRa through both hydrophobic inter-
Figure 3. Structure of 24 and the binding of 24 to RXR
sensorgrams were obtained from injection of series of concentration of 24 over the
immobilized RXR -LBD chip.
a-LBD by SPR assay. The
actions and H-bond. The 7-OH-4-Me-2-oxo-2H-Chromen-8-yl por-
tion of the compound is located near H4 with the ring system
making hydrophobic interactions with the side chains of Phe289,
Val298 and Leu301, and the para AOH group forming a H-bond
with Glu453. The contribution of Val298 to the ligand–protein
interaction has proved to be critical.18 Here, to evaluate the
involvement of the para AOH group, compound 24 is synthesized
where the AOH group in 23 was methylated (Fig. 3 and scheme
in the Supplementary data) and became incapable of acting as an
H-bond donor. As anticipated, 24 showed a weaker inhibitory
a
involved in the protein/ligand interaction. This data, together with
data from 24, supports the binding mode proposed by the docking
study (Fig. 1B).
We then asked if the binding of molecule 23 prefers the coacti-
vator-binding site to the corepressor-binding site as the coactiva-
tor-binding region and the corepressor binding region overlap.2
The role played by E453 in the binding of 23 supports that 23 binds
to the coactivator-binding region. E453 is located in H12 which is
part of the coactivator-binding site, whereas H12 does not con-
tribute to the formation of the corepressor-binding region and
E453 is likely not available for interacting with ligand. Further-
more, classical ligand like 9-cis-retinoic acid (9-cis-RA) binds to
effect on the transactivation of RXR
of the AOH group in the ligand–protein interaction. Binding of 24
to the RXR -LBD was also evaluated by the surface plasma reso-
nance (SPR) method. In consistency with the transactivation result,
24 binds weaker to RXR (Fig. 3). In the proposed binding mode,
a (Fig. 2), demonstrating a role
a
a
this AOH group forms an H-bond with side chain of Glu453. Thus,
mutant E543A could have an impact on 23 binding and its activity.
However mutating this residue can preclude our evaluation of 23
binding from using the reporter gene assay that depends on the
binding of coactivator. This is because Glu453 plays a key role in
the recruitment of the coactivator, an essential step leading to
transactivation after the binding of an agonist.5 Indeed, mutant
E453A is inactive (Fig. 2). Therefore, in order to confirm the
involvement of E453 in the binding of 23, we performed an SPR
experiment to directly measure the binding of compound 23 to
E453A mutant. Our SPR result showed that 23 bound to the
the LBD of RXRa, which stabilizes the coactivator-binding region
and can augment the binding of 23 if 23 binds to the coactivator-
binding site. Indeed we observed that in the presence of the 9-
cis-RA, 23 binds tighter to RXR
the Supplementary data). Therefore, 23 binds to the coactivator-
binding site of RXR
a in the SPR experiment (Fig. S2 in
a
.
We then examined the binding nature of compound 23 in the
coactivator-binding site to identify a strategy to optimize its bind-
ing property. First we were interested in the region where the ring
system of 2-oxo-2H-Chromen-8-yl binds. The binding mode shows
that there is limited space around 2-oxo-2H-Chromen-8-yl to
accommodate substituents on 2-oxo-2H-Chromen-8-yl. In
E453A mutant protein 10 fold weaker than to the wild type RXR
a
(Fig. S1 in the Supplementary data), suggesting that Glu453 is

3848
D. Xu et al. / Bioorg. Med. Chem. Lett. 26 (2016) 3846–3849
addition, as the lactone motif is susceptible to hydrolysis, we
decided to replace 2-oxo-2H-Chromen-8-yl with benzene, which
affords room to introduce substituents to optimize the protein/
ligand interaction. Scheme 1 was used to synthesize a series of
new compounds (Table 1).
Compounds in Table 1 were first evaluated in a single-concen-
tration reporter gene assay for their antagonist effect. Among the
tested compounds, compounds 4d and 4m displayed an inhibition
of >50% (Table 1) and were further assessed in a dose-dependent
reporter gene assay. Compound 4m showed better antagonist
activity than 4d, with an IC50 of 2.81 lM (Figs. 4 and S3 in the
Supplementary data), which is also more potent than 23. To under-
stand why compounds 4d is more active, we modeled the binding
of 4m to the coactivator-binding groove. Although 4m interacts
with RXR
a in a similar fashion to 23, in which the 2-OH forms
Figure 4. Dose-dependent effect of 4m on inhibiting the RXR
HEK-293T cells cotransfected with pG5-Luc, RXR expression vector were treated
with 9-cis-RA (10ꢀ7 M) and the indicated concentrations of 4m.
a transactivation.
an H-bond with side chain of E453, the diethyl amino group makes
stronger hydrophobic interactions with the protein (Fig. S4 in the
Supplementary data). Results for compounds 4a–4o demonstrated
the critical role of 2-OH. When 2-OH was missing (4n) or methy-
lated (4o), compounds exhibited much weaker antagonist effect
(Table 1). At the para position, substituents of bulky groups such
as benzyloxy (4k) were not well tolerated as they could cause
steric hindrance with the protein. SPR-based experiment showed
a
that 4m bound to RXR
in the Supplementary data).
The TNF /NF- B pathway plays an important role in the regu-
lation of inflammation, cell proliferation, differentiation and apop-
tosis through activation of I B kinase, and subsequent I B protein
degradation and NF- can medi-
B nuclear translocation.23,24 RXR
ate the TNF /NF- ligands
B signaling activation25 and some RXR
a-LBD with a Kd value of 0.91 lM (Fig. S5
a
j
j
j
j
a
a
j
a
Figure 5. The effect of 4m on TNF
pretreated with 4m for 6 hour before being exposed to TNF
additional 30 min. I expression were analyzed by immunoblotting, b-Actin was
used as a loading control.
a
-induced I
j
Ba
degradation. HepG2 cells were
a
(20 ng/mL) for an
jBa
DAPI
p65
Merged
Scheme 1. Synthesis of
a series of new compound (Table 1). Reagents and
conditions: (a) Ethyl chloroacetate, K2CO3, acetone, 70 °C. (b) Hydrazine hydrate,
EtOH, 80 °C (c) EtOH, room temperature.
Vehicle
Table 1
List of new compounds and their antagonist effect
TNF
α
Compds
R
Yield (%)
%Inhibition (at 10 lM)
4a
4b
4c
4d
4e
4f
4g
4h
4i
4j
4k
4l
4m
4n
4o
2,3-Dihydroxy
2,4-Dihydroxy
2,3,4-Trihydroxy
2-Hydroxy-4-OMe
2-hydroxy-4-F
2-F
3-F
3-Me
3-hydroxy
4-NO2
4-(Benzyloxy)-2-hydroxy
4-Benzyloxy
4-Diethylamino-2-hydroxy
4-Diethylamino
4-Diethylamino-2-OCH3
57
40
54
80
36
42
90
50
63
66
40
88
87
30
41
39.6
25.2
24.3
57.3
33.2
13.0
13.8
47.9
16.2
12.7
43.9
3.7
TNFα+ 4m
74.1
19.3
17.9
Figure 6. Effect of 4m on TNF
a-induced p65 nuclear translocation. HepG2 cells
pretreated with 4m (20 lm/L) for 6 h were exposed to TNFa (20 ng/mL) for 30 min.

D. Xu et al. / Bioorg. Med. Chem. Lett. 26 (2016) 3846–3849
3849
have been reported to inhibit the TNF
We therefore speculated if compound 4m which binds to RXR
with a novel binding mechanism could act to regulate the TNF
NF- B pathway. Thus, the effect of both 4d and 4m on TNF
induced I degradation was examined in HepG2 cells (Fig. S6
a-induced NF-j
B pathway.26
References and notes
a
1. Huang, P.; Chandra, V.; Rastinejad, F. Annu. Rev. Physiol. 2010, 72, 247.
2. Rastinejad, F.; Huang, P.; Chandra, V.; Khorasanizadeh, S. J. Mol. Endocrinol.
2013, 51, T1.
a/
j
a-
jBa
3. Overington, J. P.; Al-Lazikani, B.; Hopkins, A. L. Nat. Rev. Drug Disc. 2006, 5, 993.
4. Burris, T. P.; Solt, L. A.; Wang, Y.; Crumbley, C.; Banerjee, S.; Griffett, K.;
Lundasen, T.; Hughes, T.; Kojetin, D. J. Pharmacol. Rev. 2013, 65, 710.
5. Moras, D.; Gronemeyer, H. Curr. Opin. Cell Biol. 1998, 10, 384.
6. Bain, D. L.; Heneghan, A. F.; Connaghan-Jones, K. D.; Miura, M. T. Annu. Rev.
Physiol. 2007, 69, 201.
7. McEwan, I. J.; Nardulli, A. M. Nucl. Recept. Signal. 2009, 7, e011.
8. Dawson, M. I.; Zhang, X. K. Curr. Med. Chem. 2002, 9, 623.
9. Ahuja, H. S.; Szanto, A.; Nagy, L.; Davies, P. J. J. Biol. Regul. Homeost. Agents 2003,
17, 29.
in the Supplementary data). 4m displayed stronger inhibition of
the degradation than 4d. Further analysis demonstrated that 4m
could dose-dependently inhibit the TNFa-induced I
jBa
degrada-
tion (Fig. 5). In addition, 4m could inhibit the TNF
a
-induced p65
(RelA) nuclear translocation (Fig. 6). Together these data suggest
that ligands targeting the coactivator-binding site in a LBP-inde-
pendent manner can be applied to mediate the TNF
NF- B signaling pathway.
In summary, we have verified that a previously reported LBP-
independent RXR ligand binds to the coactivator-binding site of
RXR by interacting with the hydrophobic side chains of the pro-
a-induced
10. Wang, Y.; Chirgadze, N. Y.; Briggs, S. L.; Khan, S.; Jensen, E. V.; Burris, T. P. Proc.
Natl. Acad. Sci. U.S.A. 2006, 103, 9908.
j
11. Chen, L.; Wang, Z. G.; Aleshin, A. E.; Chen, F.; Chen, J.; Jiang, F.; Alitongbieke, G.;
Zeng, Z.; Ma, Y.; Huang, M.; Zhou, H.; Cadwell, G.; Zheng, J. F.; Huang, P. Q.;
Liddington, R. C.; Zhang, X. K.; Su, Y. Chem. Biol. 2014, 21, 596.
12. Scheepstra, M.; Leysen, S.; van Almen, G. C.; Miller, J. R.; Piesvaux, J.; Kutilek,
V.; van Eenennaam, H.; Zhang, H.; Barr, K.; Nagpal, S.; Soisson, S. M.; Kornienko,
M.; Wiley, K.; Elsen, N.; Sharma, S.; Correll, C. C.; Trotter, B. W.; van der Stelt,
M.; Oubrie, A.; Ottmann, C.; Parthasarathy, G.; Brunsveld, L. Nat. Commun.
2015, 6, 8833.
13. Munuganti, R. S.; Hassona, M. D.; Leblanc, E.; Frewin, K.; Singh, K.; Ma, D.; Ban,
F.; Hsing, M.; Adomat, H.; Lallous, N.; Andre, C.; Jonadass, J. P.; Zoubeidi, A.;
Young, R. N.; Guns, E. T.; Rennie, P. S.; Cherkasov, A. Chem. Biol. 2014, 21, 1476.
14. Moore, T. W.; Mayne, C. G.; Katzenellenbogen, J. A. Mol. Endocrinol. 2010, 24,
683.
a
a
tein and forming an H-bond with H12. In addition, we developed
a series of new LBP-independent ligands. Among them, compound
4m exhibited submicromolar affinity in the SPR experiment and
acted as a transcriptional antagonist of RXR
2.81 M. Furthermore, 4m could mediate the NF-
inhibiting the TNF -induced I degradation and p65 nuclear
translocation, implying that ligands that bind to the coactivator-
binding site of RXR could offer a new strategy to target the NF-
B pathway for therapeutic purpose. Further optimization of 4m
a
with an IC50 of
l
jB pathway by
a
jBa
a
15. Caboni, L.; Lloyd, D. G. Med. Res. Rev. 2013, 33, 1081.
16. Buzon, V.; Carbo, L. R.; Estruch, S. B.; Fletterick, R. J.; Estebanez-Perpina, E. Mol.
Cell. Endocrinol. 2012, 348, 394.
17. Moore, T. W.; Mayne, C. G.; Katzenellenbogen, J. A. Mol. Endocrinol. 2010, 24,
683.
j
is currently in progress.
18. Chen, F.; Liu, J.; Huang, M.; Hu, M.; Su, Y.; Zhang, X. K. ACS Med. Chem. Lett.
2014, 5, 736.
Acknowledgments
19. Zhou, H.; Liu, W.; Su, Y.; Wei, Z.; Liu, J.; Kolluri, S. K.; Wu, H.; Cao, Y.; Chen, J.;
Wu, Y.; Yan, T.; Cao, X.; Gao, W.; Molotkov, A.; Jiang, F.; Li, W. G.; Lin, B.; Zhang,
H. P.; Yu, J.; Luo, S. P.; Zeng, J. Z.; Duester, G.; Huang, P. Q.; Zhang, X. K. Cancer
Cell 2010, 17, 560.
20. Wang, Z.-G.; Chen, L.; Chen, J.; Zheng, J.-F.; Gao, W.; Zeng, Z.; Zhou, H.; Zhang,
X.-K.; Huang, P.-Q.; Su, Y. Eur. J. Med. Chem. 2013, 62, 632.
21. Zhang, X. K.; Su, Y.; Chen, L.; Chen, F.; Liu, J.; Zhou, H. Acta Pharmacol. Sin. 2015,
36, 102.
22. Halgren, T. A.; Murphy, R. B.; Friesner, R. A.; Beard, H. S.; Frye, L. L.; Pollard, W.
T.; Banks, J. L. J. Med. Chem. 2004, 47, 1750.
23. Hayden, M. S.; Ghosh, S. Semin. Immunol. 2014, 26, 253.
24. Hinz, M.; Scheidereit, C. EMBO Rep. 2014, 15, 46.
This work was supported by grants from the National Natural
Science Foundation of China (Nos. 91129302, 91429306, and
U1405229), the Xiamen Science and Technology Project (No.
3502Z20123015), the funding from Xiamen Oceanic Administra-
tion (No. 14PYY051SF04), the Research Foundation from Ministry
of Education of China (313050), the National Institutes of Health
(No. CA179379) and the California Breast Cancer Research Program
(No. 20IB-0138).
25. Zhang, X.; Zhou, H.; Su, Y. Acta Biochim. Biophys. Sin. 2016, 48, 49.
26. Zeng, Z.; Sun, Z.; Huang, M.; Zhang, W.; Liu, J.; Chen, L.; Chen, F.; Zhou, Y.; Lin,
J.; Huang, F.; Xu, L.; Zhuang, Z.; Guo, S.; Alitongbieke, G.; Xie, G.; Xu, Y.; Lin, B.;
Cao, X.; Su, Y.; Zhang, X. K.; Zhou, H. Cancer Res. 2015, 75, 2049.
Supplementary data
Supplementary data associated with this article can be found, in
the online version, at http://dx.doi.org/10.1016/j.bmcl.2016.07.
027.