Discovery of 6-Oxo-4-phenyl-hexanoic acid derivatives as RORγt inverse agonists showing favorable ADME profile

Article abstract of DOI:10.1016/j.bmcl.2021.127786

The retinoic acid receptor-related orphan nuclear receptor gamma t (RORγt), which is a promising therapeutic target for immune diseases, is a major transcription factor of genes related to psoriasis pathogenesis, such as interleukin (IL)-17A, IL-22, and IL-23R. Inspired by the co-crystal structure of RORγt, a 6-oxo-4-phenyl-hexanoic acid derivative 6a was designed, synthesized, and identified as a ligand of RORγt. The structure–activity relationship (SAR) studies in 6a, which focus on the improvement of its membrane permeability profile by introducing chlorine atoms, led to finding 12a, which has a potent RORγt inhibitory activity and a favorable pharmacokinetic profile.

Full text of DOI:10.1016/j.bmcl.2021.127786

Bioorg. Med. Chem. Lett. 36 (2021) 127786
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Discovery of 6-Oxo-4-phenyl-hexanoic acid derivatives as RORγt inverse
agonists showing favorable ADME profile
,1,
Ryota Nakajimaa *, Hiroyuki Oonoa, Keiko Kumazawaa, Tomohide Idaa, Jun Hirataa,
Ryan D. Whiteb, Xiaoshan Minc, Angel Guzman-Perezb, Zhulun Wangc, Antony Symonsd,
,
,
,
,5
Sanjay K. Singhe 2, Srinivasa Reddy Mothee 3, Sergei Belyakove 4, Anjan Chakrabartie
,
,g,*
Satoshi Shutof
a Teijin Institute for Bio-medical Research, Teijin Pharma Limited, 4-3-2 Asahigaoka, Hino, Tokyo 191-8512, Japan
b Department of Medicinal Chemistry, Amgen Discovery Research, Amgen Inc., 360 Binney Street, Cambridge, MA 02142, United States
c Departments of Molecular Engineering, Amgen Discovery Research, Amgen Inc., 1120 Veterans Boulevard, South San Francisco, CA 94080, United States
d Departments of Inflammation & Oncology Research Amgen Inc., 1120 Veterans Boulevard, South San Francisco, CA 94080, United States
e AMRI Singapore Research Centre, Pte. Ltd., 61 Science Park Road, #05-01 The Galen, Science Park III, Singapore 117525, Singapore
f Faculty of Pharmaceutical Sciences, Hokkaido University, Kita-12, Nishi-6, Kita-ku, Sapporo 060-0812, Japan
g Center for Research and Education on Drug Discovery, Hokkaido University, Kita-12, Nishi-6, Kita-ku, Sapporo 060-0812, Japan
A R T I C L E I N F O
A B S T R A C T
Keywords:
The retinoic acid receptor-related orphan nuclear receptor gamma t (RORγt), which is a promising therapeutic
target for immune diseases, is a major transcription factor of genes related to psoriasis pathogenesis, such as
interleukin (IL)-17A, IL-22, and IL-23R. Inspired by the co-crystal structure of RORγt, a 6-oxo-4-phenyl-hexanoic
acid derivative 6a was designed, synthesized, and identified as a ligand of RORγt. The structure–activity rela-
tionship (SAR) studies in 6a, which focus on the improvement of its membrane permeability profile by intro-
ducing chlorine atoms, led to finding 12a, which has a potent RORγt inhibitory activity and a favorable
pharmacokinetic profile.
Nuclear receptor
RORγt
Inverse agonist
The retinoic acid receptor-related orphan nuclear receptor gamma t
(RORγt) is a major transcription factor of genes related to psoriasis
pathogenesis, such as interleukin (IL)-17A, IL-22, and IL-23R.1–2 Ther-
apies that block IL-17A or IL-23R have successfully improved the skin
lesions of patients with moderate-to-severe psoriasis,3–7 thus rendering
RORγt inhibition as a promising therapeutic target. Many RORγt an-
tagonists (inverse agonists) have been developed,8–23 and some of them
are now being clinically investigated for the treatment of autoimmune
diseases.24 Generally, nuclear receptors have high conserved ligand-
binding domains (LBDs), which are structurally composed of alpha
helices that form large lipophilic pockets responsible for the binding of
small lipophilic ligands, such as retinoid derivatives, fatty acids,
cholesterol, and other lipophilic hormones and vitamins.25 Thus, hy-
drophobicity balance is one of the main challenges for drug delivery in
this target class, as hydrophobicity is essential for strong LBD binding
potency. However, it should also be minimized to avoid undesirable
metabolisms for the favorable drug-like properties in small-molecule
ligands (Fig 1.).
In 2014, a novel RORγt inverse agonist 1 with a basic piperazine ring
was disclosed by GSK.26 Using 1 as a lead, we reported triazolopyridine
Abbreviations: BA, bioavailability; RORγt, retinoic acid receptor-related orphan receptor γt.
* Corresponding authors at: Teijin Limited, 2-1, Kasumigaseki 3-chome, Chiyoda-ku, Tokyo 100-8585, Japan (R. Nakajima). Faculty of Pharmaceutical Sciences,
Hokkaido University, Kita-12, Nishi-6, Kita-ku, Sapporo 060-0812, Japan (S. Shuto).
E-mail addresses: r.nakajima@teijin.co.jp (R. Nakajima), shu@pharm.hokudai.ac.jp (S. Shuto).
1
Present address: Teijin Limited, 2-1, Kasumigaseki 3-chome, Chiyoda-ku, Tokyo 100-8585, Japan.
2
Present address: AWAK Technologies, 3 Tuas Lane, Singapore 638612, Singapore.
3
Present address: Polymer Engineering Characterization, Institute of Chemical and Engineering Sciences, Agency of Science, Technology and Research, 1 Pesek
Road, Jurong Island, Singapore 627833, Singapore.
4
Present address: Wintershine Health and Skincare Pte Ltd., 23 Goldhill Pl., Singapore 308931, Singapore.
5
Present address: Syngene International Limited, Biocon Park SEZ, Bommasandra IV Phase, Jigani Link Road, Bangalore 560099, India.
https://doi.org/10.1016/j.bmcl.2021.127786
Received 11 October 2020; Received in revised form 25 December 2020; Accepted 9 January 2021
Available online 23 January 2021
0960-894X/© 2021 Elsevier Ltd. All rights reserved.

R. Nakajima et al.
Bioorganic & Medicinal Chemistry Letters 36 (2021) 127786
Table 1
Structure-activity relationships (SAR) of the benzamide moiety modifications in
compound 1
Fig. 1. An RORγt ligand used as a reference, and triazolopyridine analog re-
Ra
cLogD7.4
3.8
Binding
IC50
Luc
Human
LM CLint
(mL/
b
Compound
ported in the former study.
IC50
(nM)c
(nM)d
min/mg)e
analog 2 exhibiting in vivo RORγt inhibitory activity after its oral
dosing.27 Herein, we present a separate effort from compound 1 toward
6-oxo-4-phenyl-hexanoic acid analogs as a novel RORγt inverse agonist,
which improves the PK profile compared with 1.
1
59
17
0.088
5a
5b
2.2
3.6
63
1400
1100
110
According to the x-ray co-crystal structure of compound 3 (an analog
of compound 1) complexed with RORγt (PDB ID: 6O3Z),27 a hydrogen
bond between the N-phenylamide NH and the backbone carbonyl of
Phe377 was observed. On the other hand, carboxylic acid 4 was reported
by Japan Tabaco, which is a RORγt inverse agonist that has similar N-
phenylamide moiety and hydrogen bonding with Phe377 (PDB ID:
6IVX).28 Considering the hydrophobicity balance, the carboxylic acid
6a
6b
44
52
120
160
<0.010
0.021
7a
7b
3.1
2100
1800
15,000
6500
8
9
0.82
5.1
330
67
2500
99
moiety was desirable to decrease hydrophobicity, suggesting
a
replacement of the benzamide moiety of 1 or 3 by the 4-aryl-hexanoic
acid moiety of 4 (Fig. 2B).
1.9
Therefore, we synthesized several analogs of 1 bearing 6-oxo-4-aryl-
hexanoic acid moiety, which were evaluated using a RORγt binding
assay (cell-free system) and a RORγt luciferase reporter gene assay (cell-
based system). The SAR results of the benzamide moiety modifications
in compound 1 are summarized in Table 1.
a
The absolute stereochemistry at the branched tertiary carbon (*) in 5a, 5b,
6a, 6b, 7a and 7b remained unknown.
b
Calculated by Pipeline Pilot 17.2.
c
Binding assay.
As outlined in Table 1, the cLogD7.4 values of the novel analogs were
decreased compared with 1 by introducing carboxylic acid moieties (5a,
5b, 6a, 6b, 7a, 7b and 8). The thiazole-hexanoic acid analogs 5a and 5b
showed an excellent RORγt binding activity, but its relatively lower
transcriptional inhibitory activity was observed (IC50 = 63 nM vs 1400
d
Luciferase reporter gene assay used for the evaluation of the RORγt tran-
scriptional activity inhibition (The employed HEK293T cells were transfected
with GAL4-NR-luciferase plasmids, and the activity was evaluated using the
Dual-GLOTM luciferase assay system.).
e
Metabolic stability in the liver microsome (LM).
nM for the diastereomer 5a, IC50 = 110 nM vs 1100 nM for the diaste-
reomer 5b). The phenyl-hexanoic acid analogs 6a and 6b exhibited an
increased transcriptional inhibitory activity in addition to an improved
liver microsomal stability, but a similar discrepancy between the bind-
ing affinity for RORγt and the inhibitory effect in the cells was observed.
The truncated analogs 7a, 7b and 8 showed a decreased binding affinity
and a transcriptional inhibitory activity, which indicated that the
lengths of the alkyl chain and aryl ring at the 4-position of alkyl acid
moiety are important for the RORγt inhibitory effect. Thus, a similar
discrepancy between the binding affinity and transcriptional inhibitory
effects was observed for these carboxylic acid analogs (6a, 6b, 7a, 7b
and 8). However, the 3-phenylpropanamide analog 9 did not show such
a discrepancy although its microsomal stability was decreased. The
phenyl-hexanoic acid analog 6a showed an improved binding affinity
and microsomal stability compared with the parent compound 1, and it
served as a new starting point for further optimization to improve
inhibitory activity. The relatively lower activity in the cells suggested
that the alkyl carboxylic acid analogs had a low membrane permeability.
Therefore, several analogs of 6 were designed and prepared to focus on
improving the permeability profile, and the results of their biological
evaluations were summarized, as shown in Table 2.
The x-ray co-crystal structure of compound 3 complexed with RORγt
described a space around the fluorobenzene of 327, which motivated us
to replace fluorine atom to chlorine atom for balancing hydrophobicity
and permeability. Furthermore, the piperazine ring of 6 was replaced by
a piperidine ring to avoid the formation of a zwitterionic structure,
resulting in the derivatives 10a and 10b. Though cLogD7.4 values for
10a and 10b were increased (cLogD7.4 = 4.1 for 10a and 10b), the
transcriptional inhibitory activity were not improved compared with 6a
(Luc IC50 = 280 nM for 10a and Luc IC50 = 410 nM for 10b). Then, we
focused on the amide NH proton, as the hydrogen bonding donors in the
Fig. 2. (A) Superimposed RORγt binding conformations of 3 (PDB ID: 6O3Z)
and 4 (PDB ID: 6IVX) (3: orange color by atom; 4: green color by atom); Phe377
from the corresponding RORγt crystal structures are shown, colored as 3 and 4;
The overlay is based on the superposition of the RORγt LBD in the crystal
structures of the two complexes; (B) Compound design inspired from the
superimposed conformations of 3 and 4.
drug molecule can have
a deleterious effect on the membrane
2

R. Nakajima et al.
Bioorganic & Medicinal Chemistry Letters 36 (2021) 127786
structure of RORγt LBD in complex with 11a was obtained at 1.9 A
resolution (PDB ID: 7JTW), as shown in Figure 3. The NH of the amide
moiety formed a hydrogen bond with the backbone carbonyl of Phe377.
The phenyl ring of the hexanoic acid occupied a pocket surrounded by
His323, Met365, and Ala327. The carboxylic acid portion extended out
toward the solvent-exposed region, forming a hydrogen bond interac-
tion with the backbone amide of Glu379.
Table 2
SAR of the 6a derivatives.
R2
cLogD7.4
Binding
IC50
Luc
Human
LM CLint
(mL/
b
Compound
a
IC50
(nM)c
(nM)d
The synthetic route to the representative piperazine derivatives 5a,
5b, 6a and 6b is shown in Scheme 1. Compounds 17 and 18 were syn-
thesized using the Horner-Emmons reaction of 13 and 14 followed by
the deprotection of 15 and 16. Then, the olefin moiety was hydroge-
nated by Pd/C under an H2 atmosphere to obtain 19 and 20. The syn-
thesis of 5 and 6 was achieved by subjecting 19 and 20 to amidation,
followed by deprotection and then chiral separation, to provide di-
astereomers a (early eluting) and diastereomers b (late eluting),
respectively.
min/
mg)e
10a
10b
Me
Cl
4.1
4.1
100
170
280
410
11a
11b
4.3
4.3
54
51
31
45
0.016
12a
12b
Cl
4.2
4.2
35
34
19
25
0.031
<0.010
Scheme 2 depicts the synthetic route to 10. The Mitsunobu reaction
of the phenol derivative 23 followed by the saponification of 24 afforded
25. Then, 26 was prepared by the Curtius reaction of 25. The final
compound 10 was synthesized using amidation, subsequent hydroxyl-
ation, and chiral SFC separation in a similar manner to Scheme 1 to
obtain the diastereomers a (early eluting) and diastereomers b (late
eluting). The preparation of the other piperazine and piperidine analogs
is detailed in the Supplementary Information section.
a
The absolute stereochemistry at the branched tertiary carbon (*) in 10a,
10b, 11a, 11b, 12a, and 12b remained unknown.
b
c
Calculated by Pipeline Pilot 17.2.
Binding assay.
d
Luciferase reporter gene assay used for the evaluation of the RORγt tran-
scriptional activity inhibition (The employed HEK293T cells were transfected
with GAL4-NR-luciferase plasmids, and the activity was evaluated using the
Dual-GLOTM luciferase assay system.).
In summary, a new 6-oxo-4-phenyl-hexanoic acid analog 6a was
identified as a potent RORγt inhibitor. The optimization of the mem-
brane permeability profile led to the novel analogs 11a and 12a, which
exhibited a potent RORγt inhibitory activity in the luciferase reporter
gene assay as well as a robust in vitro/in vivo ADME profile. Furthermore,
the binding mode of 11a was clarified using x-ray analysis with the co-
crystal with RORγt LBD.
e
Metabolic stability in the liver microsome (LM).
penetration. Although the intramolecular hydrogen bond in the 2-chloro
aniline structure was not reported, introducing chloride to 6 might affect
the properties of amide NH, which potentially improves the membrane
permeability. Thus, the compounds 11a, 11b, 12a and 12b, in which the
methyl group on the aniline moiety of 6a was replaced by a chlorine
atom, were designed and prepared. To our delight, these analogs
exhibited a retained binding activity and improved the transcriptional
inhibitory activity with a good liver microsomal stability compared with
6a.
The representative compounds 1, 6a, 11a and 12a in the series were
selected for ADME evaluation, and the results were recorded, as shown
in Table 3. Compared with compound 1, all the newly evaluated com-
pounds showed an excellent liver microsomal stability and a similar high
plasma protein binding (PPB) feature. Notably, a low caco-2 perme-
ability in the apical-to-basolateral (A to B) direction and a high efflux
ratio were observed for the compound 6a (Caco-2 efflux ratio = 30).
However, the compounds 11a and 12a exhibited an improved perme-
ability profile in the same evaluation system (Caco-2 efflux ratio = 2.6
for 11a and 3.5 for 12a). The in vivo PK profile of the compounds was
investigated in mice using cassette dosing. Compounds 11a and 12a
showed improved bioavailability (>86%), clearance (<0.52 L/h/kg),
area under the curve (AUC) (>4000 nM*h), and t1/2 (>1.5 h) compared
with compound 6a.
Fig. 3. Crystal structure of 11a in RORγt LBD (PDB ID: 7JTW) (The absolute
stereochemistry of 11a obtained from this x-ray structure indicated that the
absolute configuration of the chiral center at the 4-position of hexanoic acid
was R.)
To understand the binding mode of 11a and 12a in RORγt LBD, we
carried out co-crystallization studies of 11a with RORγt LBD. The crystal
Table 3
ADME profiles of the compounds 1, 5a, 11a and 12a.
Compound
LM CLint (mL/min/mg)a
PPB (%)b
Caco-2 Papp (×10ꢀ 6cm/s)c
d Mice cassette dosing study
human
mouse
human
mouse
A to B
B to A
Efflux ratio
BA (%)
AUC p.o. (nM*h)
CL(L/h/kg)
t1/2 (h)i.v. / p.o.
1
0.088
<0.010
0.016
0.11
0.017
99.7
99.4
99.4
99.7
95.7
97.2
99.5
99.4
40
1.3
16
13
55
39
41
45
1.4
30
48
490
2.0
0.7/0.7
1.2/1.5
2.3/1.9
1.5/1.5
6a
46
200
4.6
11a
<0.010
<0.010
2.6
3.5
86
4800
4000
0.36
0.52
12a
0.031
100
a
Metabolic stability in the liver microsome.
b
c
d
Plasma protein binding using human or mouse plasma.
Permeability in the apical-to-basolateral (A to B) direction and vice versa (B to A) and the efflux ratio.
2 μmol/kg p.o. dose and 1 μmol/kg i.v. dose.
3

R. Nakajima et al.
Bioorganic & Medicinal Chemistry Letters 36 (2021) 127786
interests or personal relationships that could have appeared to influence
the work reported in this paper.
Acknowledgments
We thank Ryo Matsuyama, Mariko Hirano, Atsushi Baba, Emi
Yokoyama and Akiko Takeuchi for the pharmacological discussion. We
thank Steve Thibault and Haoda Xu for the protein supply.
Appendix A. Supplementary data
Supplementary data (Synthetic experimental, compound character-
ization, crystallography studies and assay descriptions) to this article
can be found online at https://doi.org/10.1016/j.bmcl.2021.127786.
These data include MOL files and InChiKeys of the most important
compounds described in this article.
Scheme 1. Synthesis of the representative piperazine derivativesa. aReaction
conditions: a) NaH, tert-butyl 2-(diethoxyphosphoryl)acetate, THF, 0 ◦C; b) 4 N
HCl in dioxane, 0 ◦C, rt; c) H2, 10% Pd/C, THF, MeOH; d) Oxalyl chloride, DMF,
CH2Cl2, 0 ◦C; then (S)-(4-(3-amino-5-fluoro-2-methylbenzyl)-2-methylpiper-
azin-1-yl)(cyclopentyl)methanone26, TEA, CH2Cl2, 0 ◦C; e) LiOH, THF, MeOH,
H2O, chiral column SFC separation (AD-H column for 5, IA column for 6).
References
1
2
3
4
5
6
7
8
9
Ivanov II, McKenzie BS, Zhou L, et al. Cell. 2006;126:1121–1133.
Annunziato F, Cosmi L, Santarlasci V, et al. J Exp Med. 2007;204:1849–1861.
Hueber, W., Patel, D. D., Dryja, T., et al., Sci Transl Med 2010, 2, 52ra72.
Leonardi C, Matheson R, Zachariae C, et al. N Engl J Med. 2012;366:1190–1199.
Langley, R. G., Elewski, B. E., Lebwohl, M., et al., N Engl J Med 2014, 371, 326-338.
Papp KA, Reich K, Paul C, et al. Br J Dermatol. 2016;175:273–286.
Farahnik B, Beroukhim K, Zhu TH, et al. Dermatol Ther. 2016;6:25–37.
Kumar N, Solt LA, Conkright JJ, et al. Mol Pharmacol. 2010;77:228–236.
Cyr P, Bronner SM, Crawford JJ. Bioorg Med Chem Lett. 2016;26:4387–4393.
10 Qiu R, Wang Y. J Med Chem. 2018;61:5794–5804.
11 Pandya VB, Kumar S, Sachchidanand, Sharma R, Desai RC. J Med Chem. 2018;61:
10976–10995.
12 van Niel MB, Fauber BP, Cartwright M, et al. Bioorg Med Chem Lett. 2014;24:
5769–5776.
13 Chao J, Enyedy I, Van Vloten K, et al. Bioorg Med Chem Lett. 2015;25:2991–2997.
14 Wang Y, Cai W, Cheng Y, et al. ACS Med Chem Lett. 2015;6:787–792.
15 Hirata K, Kotoku M, Seki N, et al. ACS Med Chem Lett. 2016;7:23–27.
16 Olsson RI, Xue Y, von Berg S, et al. ChemMedChem. 2016;11:207–216.
17 Rene O, Fauber BP, Barnard A, et al. Bioorg Med Chem Lett. 2016;26:4455–4461.
18 Ouvry G, Bouix-Peter C, Ciesielski F, et al. Bioorg Med Chem Lett. 2016;26:
5802–5808.
19 Kummer DA, Cummings MD, Abad M, et al. Bioorg Med Chem Lett. 2017;27:
2047–2057.
Scheme 2. Synthesis of the representative piperidine derivatives 10a and 10ba,
aReaction conditions: a) Tsunoda reagent, cyclopentyl((2S)-4-hydroxy-2-
methylpiperidin-1-yl)methanone, toluene, 110 ◦C, chiral column SFC separa-
tion (AD-H column); b) LiOH, THF, MeOH, H2O; c) DPPA, DIPEA, tert-BuOH,
100 ◦C; d) 4 N HCl in dioxane; e) Oxalyl chloride, DMF, 20, CH2Cl2, 0 ◦C; then
added to 26, TEA, CH2Cl2, 0 ◦C; f) LiOH, MeOH, H2O, chiral column SFC
separation (AD-H column).
20 Barbay JK, Cummings MD, Abad M, et al. Bioorg Med Chem Lett. 2017;27:5277–5283.
21 Kono M, Ochida A, Oda T, et al. J Med Chem. 2018;61:2973–2988.
22 Gege C, Cummings MD, Albers M, et al. Bioorg Med Chem Lett. 2018;28:1446–1455.
23 Tian J, Sun N, Yu M, et al. Eur. J. Med. Chem. 2019;167:37–48.
24 Sun N, Guo H, Wang Y. Expert Opin Ther Pat. 2019;29:663–674.
25 Burris TP, Solt LA, Wang Y, et al. Pharmacol Rev. 2013;65:710–778.
26 Han F, Lei H, Lin X, Meng Q, Wang Y. WO Patent WO2014/086894A1. 2014.
27 Nakajima R, Oono H, Sugiyama S, et al. ACS Med Chem Lett. 2020;11:528–534.
28 Kotoku M, Maeba T, Fujioka S, et al. J Med Chem. 2019;62:2837–2842.
Declaration of Competing Interest
The authors declare that they have no known competing financial
4