Discovery of tertiary amine and indole derivatives as potent RORγt inverse agonists

Article abstract of DOI:10.1021/ml4003875

A novel series of tertiary amines as retinoid-related orphan receptor gamma-t (RORγt) inverse agonists was discovered through agonist/inverse agonist conversion. The level of RORγt inhibition can be enhanced by modulating the conformational disruption of H12 in RORγt LBD. Linker exploration and rational design led to the discovery of more potent indole-based RORγt inverse agonists.

Full text of DOI:10.1021/ml4003875

Letter
pubs.acs.org/acsmedchemlett
Discovery of Tertiary Amine and Indole Derivatives as Potent RORγt
Inverse Agonists
Ting Yang,† Qian Liu,† Yaobang Cheng,† Wei Cai,† Yingli Ma,† Liuqing Yang,† Qianqian Wu,†
Lisa A. Orband-Miller,‡ Ling Zhou,† Zhijun Xiang,† Melanie Huxdorf,† Wei Zhang,† Jing Zhang,†
Jia-Ning Xiang,† Stewart Leung,† Yang Qiu,† Zhong Zhong,† John D. Elliott,† Xichen Lin,†
and Yonghui Wang*,†
†Research and Development, GlaxoSmithKline, No. 3 Building, 898 Halei Road, Pudong, Shanghai 201203, China
‡Research and Development, GlaxoSmithKline, 5 Moore Drive, Research Triangle Park, North Carolina 27709, United States
S
* Supporting Information
ABSTRACT: A novel series of tertiary amines as retinoid-related
orphan receptor gamma-t (RORγt) inverse agonists was
discovered through agonist/inverse agonist conversion. The level
of RORγt inhibition can be enhanced by modulating the
conformational disruption of H12 in RORγt LBD. Linker
exploration and rational design led to the discovery of more
potent indole-based RORγt inverse agonists.
KEYWORDS: RORγt, agonists, inverse agonists, Th17 cell differentiation, cocrystal structure, structure-based design
Scheme 1. From HTS Hit (1) to Tertiary Amine RORγt
Agonist (2)
etinoid-related orphan receptor gamma-t (RORγt) is a
Rmember of the nuclear receptor superfamily. RORγt is a
key regulator of the development and functions of T-helper 17
(Th17) cells which are implicated in the pathology of a variety
of human inflammatory and autoimmune disorders.1,2 The
RORγt inhibitors have potential utility in controlling the
activity of Th17 cells and can be developed as therapeutic
agents for treatment of Th17-related autoimmune diseases. A
few small molecule inhibitors of RORγt have been reported in
the literature.3−10 In this paper, we report the discovery of
tertiary amines and indoles as potent RORγt inverse agonists
using structure- and knowledge-based compound design.
A high-throughput screen (HTS) of the GSK in-house
compound collection using a RORγ fluorescence resonance
energy transfer (FRET) assay11 resulted in identification of
thiazole amide 1 as a RORγt inverse agonist with IC50 of 1.0
μM. The binding of 1 to the RORγt ligand binding domain
(LBD) was confirmed with a thermal shift of 7.1 °C in a
thermal shift assay.11 SAR exploration on the left-hand side
(LHS) of 1 led to the identification of tertiary amine 2 as a
potent RORγt agonist with a EC50 of 0.02 μM in dual FRET
assay (Scheme 1).12 Dual FRET assay, using the same
technology as the FRET assay but without adding a surrogate
agonist, only relies on the basal level of RORγt activity and is
able to measure both agonists and inverse agonists. Peptide
profiling study using dual FRET assay showed that coactivator
peptide (e.g., steroid receptor coactivator 1 (SRC1)) was
recruited upon binding of 2 to RORγt LBD whereas
corepressor peptide (e.g., nuclear receptor corepressor 2
(NCOR2)) was not.12 Given the structure similarity of 1 and
2, we assume that they adopt a similar binding mode within
RORγt LBD despite their difference as agonist and inverse
agonist. To understand the binding mode of the chemical
series, an in-silico docking study for 2 based on a reported
RORγt crystal structure13 was conducted.
A RORγt LBD crystal structure (PDB accession code:
3KYT) was selected and processed for the docking study. A
total of 40 poses with the best scores were obtained and visually
inspected after docking with Surflex-Dock v2.314−16 in Sybyl
8.1,17 among which the top 10 poses were found to be
representative and thus further ranked using MM/GBSA18−20
affinity scores based on the VSGB2.0 solvent model.21,22 As a
Received: September 29, 2013
Accepted: November 22, 2013
Published: November 22, 2013
© 2013 American Chemical Society
65
dx.doi.org/10.1021/ml4003875 | ACS Med. Chem. Lett. 2014, 5, 65−68

ACS Medicinal Chemistry Letters
Letter
result, the binding mode with the most favored MM/GBSA
binding energy was selected and illustrated in Figure 1.12 In this
Table 1. SAR of 4-Substituents on LHS Phenyl of Amines
FRET
IC50 (nM)
dual FRET
a
compd
R
max %
XC50 (nM)
max %
b
b
b
2
3
4
5
6
7
8
H
>25 000
>25 000
13
20
25
25
125
89
Me
Et
38
84
65
t-Bu
40
>10 000
CF3
25
92
6
25
85
94
O-c-Pen
Ph
20
115
115
32
20
50
a
b
IC50 (nM) unless indicated. EC50 (nM).
of the structural interference of the substituent to H12 based on
the predicted binding mode. The severity of the structural
interference is translated to the magnitude of conformational
change of H12 and thus the ability of coactivator and
corepressor recruitment. In contrast to the notable change in
the level of inhibition, potency of the compounds was merely
affected by bulk of the substituents. Compound 7 was further
characterized by peptide recruitment profiling based on the
dual FRET assay.12 As a result, neither coactivator peptide nor
corepressor peptide tested was recruited. This was the first time
to observe that a corepressor peptide was not recruited by a
type of inverse agonists. Further biological investigation is
needed to interpret the results.
A crystal structure of RORγt LBD in a complex with 2 was
later resolved at a resolution of 2.01 Å (PDB accession code:
4NIE).12 To our delight, the overall binding mode of 2 in
RORγt LBD is essentially the same as the predicted one. With
the established binding mode, we explored SAR of the amine
containing linker with the LHS fixed as 4-CF3−Ph (Table 2).
Removal of the propyl group from the tertiary amine lowered
the potency by 50-fold (9), indicating that the hydrophobic
interaction between the propyl group and RORγt LBD has
significant impact on compound potency. Interestingly, shifting
the NH by one position (10) regained RORγt potency by more
Figure 1. Predicted binding mode of 2 in RORγt LBD is illustrated,
where 2 is in cyan stick and RORγt LBD in ribbon expression.
Residues involved in key molecular interactions with 2 are highlighted
and labeled.
binding mode, two H-bondings were observed: One from the
sulfone moiety with Arg367 and a backbone amide, and the
other between the linker amide and a backbone carbonyl. In
addition, two π−π stacking interactions are formed: one
between the sulfone-substituted phenyl ring and Phe377 and
the other between the middle phenyl ring and Phe378. The
LHS benzyl group of the ligand occupies the hydrophobic site
near Tyr502 and Trp316, which is believed to be important for
activating RORγt by stabilizing the activation function 2 (AF2)
domain (H12) essential for the downstream peptide recruit-
ment.23
On the basis of literature evidence of agonist/inverse agonist
conversion reported for estrogen receptor ligands,24 we decided
to design RORγt inverse agonists based on 2 and its predicted
binding mode by introducing substituents to the para-position
of the LHS phenyl ring near the AF2 domain to interfere H12’s
packing and thus affect the downstream peptide recruitment.
A series of compounds with different size and/or shape of 4-
substituents on the LHS phenyl ring were designed and
synthesized.12 Potency (IC50 or XC50) and maximum response
(max %) were measured in FRET and dual FRET assays (Table
1). As expected, the compounds showed a range of maximum
responses depending on the properties (size, shape, and
electrostatics) of the 4-substituents. The maximum inhibition
in the FRET assay improved in an order of Me < Et <t-Bu <
CF3 < O-c-Pen ∼ Ph. In the dual FRET assay, maximum %
activation of compounds 2−4 decreased from 125% to 65%
with the increase in size of the substituents. With a larger
substituent than 4, compound 5 showed no effect in
modulating RORγt basal activity and is considered as a neutral
antagonist. Enlarging 4-substituents further from that of 5
resulted in inverse agonists with an increased level of inhibition.
Shape and electrostatic effect also play a role for the
substituents with similar size (5 vs 6; 7 vs 8). The results of
dual FRET assay together with the structural insight from the
docking study revealed the molecular mechanism of action for
the compounds. The strong correlation between bulk of the
substituent and level of inhibition can be interpreted as severity
Table 2. SAR of Amine Linkers
66
dx.doi.org/10.1021/ml4003875 | ACS Med. Chem. Lett. 2014, 5, 65−68

ACS Medicinal Chemistry Letters
Letter
than 12-fold from that of 9. Attaching a propyl moiety to the
amine did not improve much of the potency (11 vs 10). While
introducing a Cl group on the ortho-position of the middle
phenyl ring led to the RORγt potency increased by more than
6-fold (12 vs 10). The potency improvement of 12 can be
attributed to both molecular interaction and conformational
effect. Compounds, 6, 9, 11, and 12, were also tested in mouse
Th17 cell differentiation assay,11 and their IC50 values were
more than 100 nM (Table 2). Therefore, further optimization
was needed for achieving better Th17 potency.
A series of indole analogues were designed and prepared
subsequently to explore SAR of the LHS phenyl substitution.
Synthetic procedures for preparation of indole analogues 13−
21 and 22 were shown in Scheme 2. Reaction of bromides
Scheme 2. Synthetic Procedures for Indole Analogues 13−21
a
and 22
To mimic the aniline geometry of 11 and 12, we designed a
more rigid indole compound 13 as a potential RORγt inverse
agonist. Figure 2 shows the predicted binding mode of 13,
a
Reagents and conditions: (a) for X = Br: 5-nitro-1H-indole, Cs2CO3,
Figure 2. Structural overlay of 13 (in brown stick) and 6 (in cyan
stick) in RORγt LBD.
KI, DMF, 90 °C. For X = I: 5-nitro-1H-indole, K2CO3 (or Cs2CO3),
DMF, RT; or 5-nitro-1H-indole, NaH, DMF, 0−90 °C. (b) SnCl2·
2H2O, ethanol, reflux. (c) For acid A, HATU, DIPEA, DCM, RT; for
perfluorophenyl ester B, DIPEA, DCM, RT; (d) HATU, DIPEA,
DCM; (e) NaBH4, BF3·Et2O, THF, RT; (f) toluene, 4,5-dichloro-3,6-
dioxocyclohexa-1,4-diene-1,2-dicarbonitrile, 80 °C; (g) SnCl2·2H2O,
ethanol, reflux; (h) perfluorophenyl ester B, DIPEA, DCM, RT.
compared with that of 6 in RORγt crystal structure. The 4-
CF3−Ph moiety of 13 extends naturally toward H12 in a low
energy conformation, where CF3 is in close contact with the
side chain of Tyr502 on H12. Comparing to phenyl, indole
seems to be more efficient for the design of inverse agonists
with AF2 domain disrupting mechanism. With this assessment,
13 was synthesized and confirmed to be a potent RORγt
inverse agonist with FRET IC50 of 5 nM (max % = 104) (Table
3).
13a−21a and 5-nitro-1H-indole gave 1-substituted-5-nitro-
indoles 13b−21b. Reduction of the nitro-indoles to amino-
indoles 13c−21c, followed by amide formation with acid A or
activated ester B produced the targeted compounds 13−21.
Synthesis of compound 22 started with reaction of 2-(3,4-
dichlorophenyl)acetic acid and 5-nitro-indoline, which led to
formation of amide 22a. Reduction of the amide, followed by
oxidation of indoline gave the nitro-indole 22b, which upon
nitro reduction and amide formation with B afforded the
targeted compound 22.
Table 3. SAR of Substituents on LHS Phenyl of Indoles
All indole-containing compounds tested showed strong
RORγt inhibition with FRET IC50 values below 10 nM. Similar
to the tertiary amine, the bulk of the para-substitution clearly
impacts the level of RORγt inhibition. The nonsubstituted
analogue 14 showed only 65% of inhibition, and the level of
inhibition generally improved with the increase in size of the
substituted moieties (14−18), provided that shape and
electrostatic effect of the substituents are taken into
consideration when comparing compounds with similar size
of the substituents. Substitution on other positions of the LHS
phenyl was also explored. The reduced maximum response of
19 and 20 suggests that 4-substitution is more effective on H12
disruption than 2- or 3-substitution. The RORγt potencies of
the disubstituted compounds 21 and 22 are similar to that of
16 but the Th17 potencies are higher. This suggests that
lipophilicity might have played a role in cellular potency. With
FRET
IC50 (nM)
Th17
IC50 (nM)
31
compd
R
max %
13
14
15
16
17
18
19
20
21
22
4-CF3
4-H
5
3
8
4
6
6
6
4
6
6
104
65
4-F
92
4-Cl
99
164
457
70
4-CN
4-OiPr
2-CF3
3-CF3
2-Cl, 4-Cl
3-Cl, 4-Cl
108
119
83
57
100
98
9
40
67
dx.doi.org/10.1021/ml4003875 | ACS Med. Chem. Lett. 2014, 5, 65−68

ACS Medicinal Chemistry Letters
Letter
(7) Kumar, N.; Lyda, B.; Chang, M. R.; Lauer, J. L.; Solt, L. A.; Burris,
T. P.; Kamenecka, T. M.; Griffin, P. R. Identification of SR2211: a
potent synthetic RORγ-selective modulator. ACS Chem. Biol. 2012, 7,
672−677.
(8) Huh, J. R.; Englund, E. E.; Wang, H.; Huang, R.; Huang, P.;
Rastinejad, F.; Inglese, J.; Austin, C. P.; Johnson, R. L.; Huang, W.;
Littman, D. R. Identification of potent and selective diphenylpropa-
namide RORγ inhibitors. ACS Med. Chem. Lett. 2013, 4, 79−84.
(9) Kamenecka, T. M.; Lyda, B.; Chang, M. R.; Griffin, P. R.
Synthetic modulators of the retinoic acid receptor-related orphan
receptors. Med. Chem. Commun. 2013, 4, 764−776.
(10) Murali Dhar, T. G.; Zhao, Q.; Markby, D. W. Targeting the
nclear hormone receptor RORγt for the treatment of autoimmune and
inflammatory disorders. Annu. Rep. Med. Chem. 2013, 48, 169−182.
(11) Zhang, W.; Zhang, J.; Fang, L.; Zhou, L.; Wang, S.; Xiang, Z.; Li,
Y.; Wisely, B.; Zhang, G.; An, G.; Wang, Y.; Leung, S.; Zhong, Z.
Increasing human Th17 differentiation through activation of orphan
nuclear receptor retinoid acid-related orphan receptor γ (RORγ) by a
class of aryl amide compounds. Mol. Pharmacol. 2012, 82, 583−590.
(12) See the Supporting Information for details.
proper substituents on LHS phenyl of the indoles, we were able
to achieve high Th17 potency with IC50s equal or below 100
nM (13, 18, 21, and 22).
In summary, we identified a novel series of tertiary amines as
RORγt inverse agonists using structure- and knowledge-based
design. Relationship between ligand/H12 structural disruption
and the level of RORγt inhibition was established for the first
time. Linker exploration and rational design led to a series of
indole-based analogues as more potent RORγt inverse agonists.
Compound 21 was discovered as a potent RORγt lead with
both FRET and Th17 IC50s lower than 10 nM. Further
optimization of the PK profile of the aryl amide series is
ongoing and will be reported in due course.
ASSOCIATED CONTENT
* Supporting Information
■
S
RORγt FRET, dual FRET and Th17 assays description; peptide
profiling studies; modeling studies; cocrystal structure data;
synthetic procedures; and compound characterization. This
material is available free of charge via the Internet at http://
pubs.acs.org.
(13) Jin, L.; Martynowski, D.; Zheng, S.; Wada, T.; Xie, W.; Li, Y.
Structural Basis for Hydroxycholesterols as Natural Ligands of Orphan
Nuclear Receptor RORg. Mol. Endocrinol. 2010, 24, 923−929.
(14) Pham, T. A.; Jain, A. N. Parameter estimation for scoring
protein ligand interactions using negative training data. J. Med. Chem.
2006, 49, 5856−5868.
AUTHOR INFORMATION
Corresponding Author
*Phone: +8621-61590761. E-mail: wang.2.yonghui@gmail.
com.
■
(15) Jain, A. N. Virtual screening in lead discovery and optimization.
Curr. Opin. Drug Discovery Dev. 2004, 7, 396−403.
(16) Jain, A. N. Surflex: Fully automatic flexible molecular docking
using a molecular similarity-based search engine. J. Med. Chem. 2003,
46, 499−511.
Author Contributions
The manuscript was written through contributions of all
authors. All authors have given approval to the final version of
the manuscript.
(17) SYBYL 8.1; Tripos International: St. Louis, MO.
(18) Greenidge, P. A.; Kramer, C.; Mozziconacci, J.-C.; Wolf, R. M.
MM/GBSA Binding Energy Prediction on the PDB bind Data Set:
Successes, Failures, and Directions for Further Improvement. J. Chem.
Inf. Model. 2013, 53, 201−209.
(19) Kollman, P. A.; Massova, I.; Reyes, C.; Kuhn, B.; Huo, S. H.;
Chong, L.; Lee, M.; Lee, T.; Duan, Y.; Wang, W.; Donini, O.; Cieplak,
P.; Srinivasan, J.; Case, D. A.; Cheatham, T. E. Calculating structures
and free energies of complex molecules: Combining molecular
mechanics and continuum models. Acc. Chem. Res. 2000, 33 (12),
889−897.
(20) Wang, J. M.; Hou, T. J.; Xu, X. J. Recent advances in free energy
calculations with a combination of molecular mechanics and
continuum models. Curr. Comput.-Aided Drug Des. 2006, 2 (3),
287−306.
(21) Li, J.; Abel, R.; Zhu, K.; Cao, Y.; Zhao, S.; Friesner, R. A. The
VSGB 2.0 model: a next generation energy model for high resolution
protein structure modeling. Proteins 2011, 79, 2794−2812.
(22) Maestro, ver. 9.2.109; Schrodinger, LLC: New York.
(23) Huang, Z.; Xie, H.; Wang, R.; Sun, Z. Retinoid-related orphan
receptor gt is a potential therapeutic target for controlling
inflammatory autoimmunity. Expert Opin. Ther. Targets 2007, 11
(6), 737−743.
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
We thank Guifeng Zhang, Shuai Wang, Gang An, Bruce Wisely,
Tom Consler, Kevin Meng, Hui Lei, Feng Ren and Hongtao Lu
for their helpful discussions.
REFERENCES
■
(1) Huh, J. R.; Littman, D. R. Small molecule inhibitors of RORγt:
targeting Th17 cells and other applications. Eur. J. Immunol. 2012, 42,
2232−2237.
(2) Burris, T. P.; Busby, S. A.; Griffin, P. R. Targeting orphan nuclear
receptors for treatment of metabolic diseases and autoimmunity.
Chem. Biol. 2012, 19, 51−59.
(3) Huh, J. R.; Leung, M. W. L.; Huang, P.; Ryan, D. A.; Krout, M.
R.; Malapaka, R. R. V.; Chow, J.; Manel, N.; Ciofani, M.; Kim, S. V.;
Cuesta, A.; Santori, F. R.; Lafaille, J. J.; Xu, H. E.; Gin, D. Y.;
Rastinejad, F.; Littman, D. R. Digoxin and its derivatives suppress
TH17 cell differentiation by antagonizing RORγt activity. Nature
2011, 472, 486−490.
(24) Shiau, A. K.; Barstad, D.; Loria, P. M.; Cheng, L.; Kushner, P. J.;
Agard, D. A.; Greene, G. L. The Structural Basis of Estrogen
Receptor/Coactivator Recognition and the Antagonism of This
Interaction by Tamoxifen. Cell 1998, 95, 927−937.
(4) Solt, L. A.; Kumar, N.; Nuhant, P.; Wang, Y.; Lauer, J. L.; Liu, J.;
Istrate, M. A.; Kamenecka, T. M.; Roush, W. R.; Vidovic, D.; Schurer,
̈
S. C.; Xu, J.; Wagoner, G.; Drew, P. D.; Griffin, P. R.; Burris, T. P.
Suppression of Th17 Differentiation and autoimmunity by a synthetic
ROR ligand. Nature 2011, 472, 491−494.
(5) Xu, T.; Wang, X.; Zhong, B.; Nurieva, R. I.; Ding, S.; Dong, C.
Ursolic acid suppresses interleukin-17 (IL-17) production by
selectively antagonizing the function of RORγt protein. J. Biol. Chem.
2011, 286, 22707−22710.
(6) Solt, L. A.; Kumar, N.; He, Y.; Kamenecka, T. M.; Griffin, P. R.;
Burris, T. P. Identification of a selective RORγ ligand that suppresses
Th17 cells and stimulates T regulatory cells. ACS Chem. Biol. 2012, 7,
1515−1519.
68
dx.doi.org/10.1021/ml4003875 | ACS Med. Chem. Lett. 2014, 5, 65−68