Discovery of 3-Cyano- N-(3-(1-isobutyrylpiperidin-4-yl)-1-methyl-4-(trifluoromethyl)-1 H-pyrrolo[2,3- b]pyridin-5-yl)benzamide: A Potent, Selective, and Orally Bioavailable Retinoic Acid Receptor-Related Orphan Receptor C2 Inverse Agonist

Article abstract of DOI:10.1021/acs.jmedchem.8b00392

The nuclear hormone receptor retinoic acid receptor-related orphan C2 (RORC2, also known as RORγt) is a promising target for the treatment of autoimmune diseases. A small molecule, inverse agonist of the receptor is anticipated to reduce production of IL-17, a key proinflammatory cytokine. Through a high-throughput screening approach, we identified a molecule displaying promising binding affinity for RORC2, inhibition of IL-17 production in Th17 cells, and selectivity against the related RORA and RORB receptor isoforms. Lead optimization to improve the potency and metabolic stability of this hit focused on two key design strategies, namely, iterative optimization driven by increasing lipophilic efficiency and structure-guided conformational restriction to achieve optimal ground state energetics and maximize receptor residence time. This approach successfully identified 3-cyano-N-(3-(1-isobutyrylpiperidin-4-yl)-1-methyl-4-(trifluoromethyl)-1H-pyrrolo[2,3-b]pyridin-5-yl)benzamide as a potent and selective RORC2 inverse agonist, demonstrating good metabolic stability, oral bioavailability, and the ability to reduce IL-17 levels and skin inflammation in a preclinical in vivo animal model upon oral administration.

Full text of DOI:10.1021/acs.jmedchem.8b00392

Subscriber access provided by UNIV OF NEW ENGLAND ARMIDALE
Article
Discovery of 3-Cyano-N-(3-(1-isobutyrylpiperidin-4-yl)-1-methyl-4-
(trifluoromethyl)-1H-pyrrolo[2,3-b]pyridin-5-yl)benzamide.
A Potent, Selective, and Orally Bioavailable Retinoic Acid
Receptor-Related Orphan Receptor C2 (RORC2) Inverse Agonist
Mark E. Schnute, Mattias Wennerstål, Jennifer Alley, Martin Bengtsson, James R. Blinn, Charles W. Bolten,
Timothy Braden, Tomas Bonn, Bo Carlsson, Nicole L. Caspers, Ming Z. Chen, Chulho Choi, Leon P. Collis,
Kimberly Crouse, Mathias Färnegårdh, Kimberly F. Fennell, Susan Fish, Andrew Flick, Annika
Goos-Nilsson, Hjalmar Gullberg, Peter K. Harris, Steven E. Heasley, Martin Hegen, Alexander E. Hromockyj,
Xiao Hu, Bolette Husman, Tomasz Janosik, Peter Jones, Neelu Kaila, Elisabet Kallin, Björn Kauppi, James R.
Kiefer, John D. Knafels, Konrad Koehler, Lars Kruger, Ravi G. Kurumbail, Robert E. Kyne, Jr., Wei Li, Joakim
Löfstedt, Scott A. Long, Carol A. Menard, Scot Mente, Dean Messing, Marvin J. Meyers, Lee Napierata, Daniel
Nöteberg, Philippe Nuhant, Matthew J. Pelc, Michael J. Prinsen, Patrik Rhönnstad, Eva Backström-Rydin,
Johnny Sandberg, Maria Sandström, Falgun Shah, Maria Sjöberg, Aron Sundell, Alexandria P. Taylor, Atli
Thorarensen, John I Trujillo, John D Trzupek, Ray Unwalla, Felix F. Vajdos, Robin A. Weinberg, David C.
Wood, Li Xing, Edouard Zamaratski, Christoph W Zapf, Yajuan Zhao, Anna Wilhelmsson, and Gabriel Berstein
J. Med. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.jmedchem.8b00392 • Publication Date (Web): 21 Aug 2018
Downloaded from http://pubs.acs.org on August 21, 2018
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a service to the research community to expedite the dissemination
of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in
full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully
peer reviewed, but should not be considered the official version of record. They are citable by the
Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore,
the “Just Accepted” Web site may not include all articles that will be published in the journal. After
a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web
site and published as an ASAP article. Note that technical editing may introduce minor changes
to the manuscript text and/or graphics which could affect content, and all legal disclaimers and
ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or
consequences arising from the use of information contained in these “Just Accepted” manuscripts.
is published by the American Chemical Society. 1155 Sixteenth Street N.W.,
Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.

Page 1 of 105
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Kurumbail, Ravi; Pfizer, Medicine Design
Kyne, Jr., Robert; Pfizer, Medicine Design
Li, Wei; Pfizer, Inflammation and Immunology
Löfstedt, Joakim; Karo Bio AB
Long, Scott; Pfizer, Worldwide Medicinal Chemistry
Menard, Carol; Pfizer, Medicine Design
Mente, Scot; Pfizer, Medicine Design
Messing, Dean; Pfizer, Medicine Design
Meyers, Marvin; Pfizer, Worldwide Medicinal Chemistry
Napierata, Lee; Pfizer, Inflammation and Immunology
Nöteberg, Daniel; Karo Bio AB
Nuhant, Philippe; Pfizer Inc, Medicine Design
Pelc, Matthew; Pfizer, Worldwide Medicinal Chemistry
Prinsen, Michael; Pfizer, Inflammation and Immunology
Rhönnstad, Patrik; Karo Bio AB
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Backström-Rydin, Eva; Karo Bio AB
Sandberg, Johnny; Karo Bio AB
Sandström, Maria; Karo Bio AB
Shah, Falgun; Pfizer, Medicine Design
Sjöberg, Maria; Karo Bio AB
Sundell, Aron; Karo Bio AB
Taylor, Alexandria; Pfizer Worldwide R&D, Medicine Design
Thorarensen, Atli; Pfizer, Medicine Design
Trujillo, John; Pfizer, Medicine Design
Trzupek, John; Pfizer, Medicine Design
Unwalla, Ray; Pfizer, Medicine Design
Vajdos, Felix; Pfizer, Medicine Design
Weinberg, Robin; Pfizer, Inflammation and Immunology
Wood, David; Pfizer, Inflammation and Immunology
Xing, Li; Pfizer, Medicine Design
Zamaratski, Edouard; Karo Bio AB
Zapf, Christoph; Pfizer, Medicine Design
Zhao, Yajuan; Pfizer, Inflammation and Immunology
Wilhelmsson, Anna; Karo Bio AB
Berstein, Gabriel; Pfizer, Inflammation and Immunology
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 2 of 105
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Discovery of 3ꢀCyanoꢀNꢀ(3ꢀ(1ꢀisobutyrylpiperidinꢀ
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
4
ꢀyl)ꢀ1ꢀmethylꢀ4ꢀ(trifluoromethyl)ꢀ1Hꢀpyrrolo[2,3ꢀ
b]pyridinꢀ5ꢀyl)benzamide. A Potent, Selective, and
Orally Bioavailable Retinoic Acid ReceptorꢀRelated
Orphan Receptor C2 (RORC2) Inverse Agonist
†
#
‡
#
║
Mark E. Schnute*, Mattias Wennerstål, Jennifer Alley, Martin Bengtsson, James R. Blinn,
┴
┴
#
#
§
Charles W. Bolten, Timothy Braden, Tomas Bonn, Bo Carlsson, Nicole Caspers, Ming
§
§
‡
‡
#
Chen, Chulho Choi, Leon P. Collis, Kimberly Crouse, Mathias Färnegårdh, Kimberly F.
§
‡
§
#
#
Fennell, Susan Fish, Andrew C. Flick, Annika GoosꢀNilsson, Hjalmar Gullberg, Peter K.
┴
§
‡
┴
┴
Harris, Steven E. Heasley, Martin Hegen, Alexander E. Hromockyj, Xiao Hu, Bolette
#
#
†
†
#
#
Husman, Tomasz Janosik, Peter Jones, Neelu Kaila, Elisabet Kallin, Björn Kauppi, James
║
§
#
#
§
R. Kiefer, John Knafels, Konrad Koehler, Lars Kruger, Ravi G. Kurumbail, Robert E. Kyne
§
‡
#
║
§
†
†
Jr., Wei Li, Joakim Löfstedt, Scott A. Long, Carol A. Menard, Scot Mente, Dean Messing,
║
‡
#
§
║
Marvin J. Meyers, Lee Napierata, Daniel Nöteberg, Philippe Nuhant, Matthew J. Pelc,
┴
#
#
#
Michael J. Prinsen, Patrik Rhönnstad, Eva BackströmꢀRydin, Johnny Sandberg, Maria
#
†
#
#
§
Sandström, Falgun Shah, Maria Sjöberg, Aron Sundell, Alexandria P. Taylor, Atli
†
§
†
†
§
Thorarensen, John I. Trujillo, John D. Trzupek, Ray Unwalla, Felix F. Vajdos, Robin A.
ACS Paragon Plus Environment
1

Page 3 of 105
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
┴
┴
†
#
†
‡
Weinberg, David C. Wood, Li Xing, Edouard Zamaratski, Christoph W. Zapf, Yajuan Zhao,
#
‡
Anna Wilhelmsson, and Gabriel Berstein
†
‡
Medicine Design and Inflammation and Immunology Research, Pfizer Inc, Cambridge,
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Massachusetts 02139, United States
§
Medicine Design, Pfizer Inc, Groton, Connecticut 06340, United States
║
┴
Worldwide Medicinal Chemistry and Inflammation and Immunology Research, Pfizer Inc, St.
Louis, Missouri 63017, United States
#
Karo Bio AB (now Karo Pharma AB), Stockholm, Sweden
KEYWORDS: RORC2, RORγt, ILꢀ17, inflammation, psoriasis, high throughput screening
ABSTRACT
The nuclear hormone receptor RORC2 (also known as RORγt) is a promising target for the
treatment of autoimmune diseases. A small molecule, inverse agonist of the receptor is
anticipated to reduce production of ILꢀ17, a key proꢀinflammatory cytokine. Through a highꢀ
throughput screening approach, we identified a molecule displaying promising binding affinity
for RORC2, inhibition of ILꢀ17 production in Th17 cells, and selectivity against the related
RORA and RORB receptor isoforms. Lead optimization to improve the potency and metabolic
stability of this hit focused on two key design strategies, namely, iterative optimization driven by
increasing lipophilic efficiency and structureꢀguided conformational restriction to achieve
ACS Paragon Plus Environment
2

Journal of Medicinal Chemistry
Page 4 of 105
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
optimal ground state energetics and maximize receptor residence time. This approach
successfully identified 3ꢀcyanoꢀNꢀ(3ꢀ(1ꢀisobutyrylpiperidinꢀ4ꢀyl)ꢀ1ꢀmethylꢀ4ꢀ(trifluoromethyl)ꢀ
1Hꢀpyrrolo[2,3ꢀb]pyridinꢀ5ꢀyl)benzamide as a potent and selective RORC2 inverse agonist
demonstrating good metabolic stability, oral bioavailability, and the ability to reduce ILꢀ17 levels
and skin inflammation in a preclinical in vivo animal model upon oral administration.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
INTRODUCTION
Interleukinꢀ17 (ILꢀ17) is a proꢀinflammatory cytokine that plays a dominant role in
1
,2
inflammation, autoimmunity and host defense.
ILꢀ17 drives an amplification mechanism in
inflammatory disease. At a site of inflammation, ILꢀ17 stimulates resident cells to secrete
chemokines and other proꢀinflammatory mediators that recruit additional inflammatory cells to
3
the tissue. Monoclonal antibodies directed towards the ILꢀ17 cytokine itself or the ILꢀ17
4
ꢀ6
7ꢀ9
receptor have demonstrated clinical efficacy in the treatment of psoriasis, psoriatic arthritis
10
and ankylosing spondylitis. The ILꢀ17 pathway has also been implicated in other diseases such
1
,11
12,13
as rheumatoid arthritis and multiple sclerosis.
ILꢀ17 is produced by a specific subset of T
cells and other immune cells, of which T helper 17 (Th17) cells are believed to play a dominant
2,14
role in disease pathogenesis.
In these cells, the nuclear hormone receptor RORC2, a member
of the retinoic acid receptorꢀrelated orphan receptor family and the immune cell specific isoform
of the RORC gene, is required for the expression of ILꢀ17 and the differentiation of Th17 cells
15
from naïve CD4+ T cells. Human subjects displaying a homozygous lossꢀofꢀfunction mutation
in the RORC gene lack ILꢀ17 producing cells in their peripheral blood, suggesting a critical role
1
6
for RORC2 in ILꢀ17 production. Although the endogenous ligand of RORC2 remains to be
ACS Paragon Plus Environment
3

Page 5 of 105
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
firmly established, evidence suggests that one or more intermediates along the cholesterol
1
7ꢀ19
biosynthesis pathway may play this role.
Two other members of the same nuclear receptor
family, RORA and RORB, are believed to contribute to the regulation of circadian and motor
2
0
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
functions.
In addition to the large molecule ILꢀ17 (or ILꢀ17R) monoclonal antibody approach to modulate
the cytokine pathway, targeting RORC2 with a small molecule approach is also an attractive
therapeutic strategy. An inverse agonist of the receptor is anticipated to attenuate production of
ILꢀ17 not only in the presence of agonist but also in the case where the receptor has a
2
1
constitutive level of transcriptional activity. Indeed, a diverse range of chemotypes have been
reported in the literature by industry and academia that display pharmacology consistent with
22ꢀ30
inverse agonism of the receptor in vitro and in some cases in vivo.
Most of the RORC2
ligands reported to date have been found to bind to a large, hydrophobic pocket within the ligand
binding domain (LBD), presumably the same binding site of the endogenous ligand. The nature
of this pocket to favor lipophilic ligands has presented a specific challenge towards the
identification of pharmaceutical agents suitable for oral delivery as favorable ligands tend to
have poor metabolic stability and modest oral bioavailability. In spite of the anticipated
challenges in candidate optimization, we initiated a small molecule program targeting RORC2
based on its strong linkage to inflammatory diseases. Herein we describe our approach to lead
optimization of RORC2 inverse agonists originating from the identification of a highꢀthroughput
screening (HTS) hit. The combined approach of structureꢀguided ligand conformational
restriction and lipophilic efficiency focused optimization has led to the identification of a potent,
selective and orally bioavailable inverse agonist of RORC2.
ACS Paragon Plus Environment
4

Journal of Medicinal Chemistry
Page 6 of 105
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
CHEMISTRY
The synthesis of 5ꢀaminoꢀ3ꢀpiperidine substituted indoles has been previously described in the
3
1
literature and intermediate 1 served as the precursor for the preparation of the initial HTS hit 3
and related analogs, Scheme 1. Differentiation of the two amine functionalities was realized
through benzamide formation with the 5ꢀamino substituent of the indole ring to afford compound
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
2
. Subsequent piperidine Nꢀdeprotection followed by acylation with cyclopentane carbonyl
chloride provided compound 3. Access to the corresponding 1ꢀmethylindole analog was possible
by alkylation of the indole nitrogen to provide compound 4, Scheme 1. However, to enable
broader structureꢀactivityꢀrelationship (SAR) studies installation of the methyl group prior to
amide formation was more advantageous, Scheme 2. 1ꢀMethylindole 5 allowed for the
sequential amide bond formation steps suitable for a parallel medicinal chemistry approach to
yield analogs such as 7 and 8a-b. The representative 1ꢀethylindole analog 11 was also prepared
from intermediate 1 by alkylation of the indole nitrogen prior to piperidine amide formation,
Scheme 3.
Synthesis efforts next focused on the preparation of pyrrolopyridine and pyrrolopyrimidine
scaffolds, Scheme 4ꢀ5. In these cases, installation of the piperidine substituent was made
possible through a Suzuki crossꢀcoupling reaction between vinyl boronate 12 and the
corresponding 3ꢀiodoꢀpyrrolopyridine or 3ꢀiodoꢀpyrrolopyrimidine scaffold intermediate. The
pyrrolo[2,3ꢀb]pyridine iodide was prepared from 5ꢀnitroꢀ1Hꢀpyrrolo[2,3ꢀb]pyridine (13) by a
oneꢀpot Nꢀmethylation/iodination reaction to provide compound 14, Scheme 4. Subsequent
crossꢀcoupling with 12 afforded compound 15. Reduction of the olefin and nitro group
concurrently under transfer hydrogenation conditions afforded amine 16 which was then coupled
with the appropriate acid chlorides to provide analogs 17 and 18. Pyrrolo[3,2ꢀb]ꢀ and
ACS Paragon Plus Environment
5

Page 7 of 105
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
pyrrolo[2,3ꢀc]pyridine scaffolds were provided through crossꢀcoupling of 12 with 21a and 21b,
32
respectively, Scheme 5. Each was prepared through iodination of either 19a or 19b followed
by Nꢀmethylation of the iodinated heterocycle. The crossꢀcoupling reaction occurred with
advantageous deprotection of the formamide protecting group during workꢀup to provide
compounds 22a and 22b. Subsequent reduction of the olefin by transfer hydrogenation and
amide formation afforded analogs 23 and 24. The required coupling partner for the pyrrolo[3,2ꢀ
d]pyrimidine scaffold, 28, arose from initial regioselective reductive dehalogenation of 2,4ꢀ
dichloroꢀ5Hꢀpyrrolo[3,2ꢀd]pyrimidine (25) followed by iodination and Nꢀmethylation, Scheme 5.
The crossꢀcoupling of iodoheterocyle 28 with boronate 12 provided compound 29. Installation
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
of the 5ꢀamino functionality was achieved through a catalytic amination reaction between 29 and
33
benzophenone imine followed by transfer hydrogenation.
The resulting amine was then
reacted with 4ꢀcyanopicolinoyl chloride to provide analog 31.
The synthesis of analogs bearing substitution at the 4ꢀposition of either the indole or
pyrrolo[2,3ꢀb]pyridine scaffold was undertaken as follows. The 4ꢀmethylindole analog 36 was
prepared through crossꢀcoupling of 12 with 3ꢀiodoꢀ4ꢀmethylindole 33, which originated from 4ꢀ
methylꢀ5ꢀnitroindole (32) through a oneꢀpot iodination/Nꢀmethylation, Scheme 6. Variation of
the 4ꢀposition substituent to the pyrrolo[2,3ꢀb]pyridine scaffold was challenging to introduce
through parallel synthetic methods. Therefore, the corresponding substituted 3ꢀiodoꢀpyrrolo[2,3ꢀ
b]pyridines bearing either methyl, isopropyl, trifluoromethyl, or methoxy substitution each
required unique synthetic approaches depending on the substituent, Schemes 7–10. The 4ꢀ
34
methyl substituent was introduced through a Kumada coupling between 4ꢀchloroꢀ1Hꢀ
pyrrolo[2,3ꢀb]pyridine (37) and methylmagnesium bromide to provide precursor 38, Scheme 7.
In order to achieve clean nitration at the 5ꢀposition of the ring, it was necessary to convert 38 to
ACS Paragon Plus Environment
6

Journal of Medicinal Chemistry
Page 8 of 105
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
the corresponding benzenesulfonamide derivative 39. In the absence of Nꢀsulfonylation,
competitive nitration at the C3ꢀposition led to complex mixtures of products. Subsequent
nitration of 39 with tetramethyl ammonium nitrate selectively afforded 5ꢀnitroꢀpyrrolo[2,3ꢀ
b]pyridine 40. Removal of the phenylsulfonyl protecting group was accomplished by heating 40
with morpholine under basic conditions to afford compound 41. Subsequent Nꢀmethylation and
C3ꢀiodination provided the desired 4ꢀmethylꢀ3ꢀiodoꢀpyrrolo[2,3ꢀb]pyridine coupling partner 43.
For the corresponding 4ꢀisopropyl intermediate 49, a similar approach was adopted, Scheme 8.
Suzuki crossꢀcoupling with isopropenyl boronic acid pinacol ester proceeded cleanly with
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
3
5
benzenesulfonamide 44 to afford the isoprenyl intermediate 45. Hydrogenation of the
exocyclic olefin then provided the corresponding isopropylpyrrolo[2,3ꢀb]pyridine 46. Following
the same sequence of reactions (nitration, iodination and Nꢀmethylation) as with the methyl
analog, the isopropyl substituted coupling partner 49 was prepared. Preparation of 4ꢀmethoxyꢀ3ꢀ
iodopyrrolo[2,3ꢀb]pyridine 53 relied on a different approach to introduce the 4ꢀposition
substituent, Scheme 9. Compound 50, which was obtained by nitration of chloroheterocycle
35
4
4, was found to be sufficiently electron deficient to facilitate clean introduction of
nucleophiles at the 4ꢀposition. Therefore, condensation of 50 with sodium methoxide afforded
ether 51 which proceeded with concomitant desulfonylation. Subsequent Nꢀmethylation and
iodination afforded coupling partner 53.
The desired 4ꢀtrifluoromethyl substituted intermediate for the coupling reaction was
particularly challenging to prepare, Scheme 10. Starting from 4ꢀchloroꢀ1Hꢀpyrrolo[2,3ꢀ
b]pyridine (37), in situ protection of the N1 nitrogen atom as the acetyl derivative was required
prior to a Finkelstein reaction which yielded iodide 54. Next, Nꢀphenylsulfonylation followed by
ring nitration with trimethylammonium nitrate afforded the 4ꢀiodoꢀ5ꢀnitropyrrolo[2,3ꢀb]pyridine
ACS Paragon Plus Environment
7

Page 9 of 105
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
5
6.
Copperꢀmediated trifluoromethylation of 56 with methyl 2,2ꢀdifluoroꢀ2ꢀ
3
6
(fluorosulfonyl)acetate provided compound 57. Although it was found that the corresponding
bromoꢀ and chloroꢀanalogs of 56 could facilitate introduction of the trifluoromethyl group, the
iodide was found to be the most efficient. Furthermore, it was also found that the nitro and
phenylsulfonyl groups were both critical to the success of the reaction, as all attempts to
introduce a trifluoromethyl group onto substrates lacking either one of these functionalities failed
to yield synthetically useful product. Compound 57 was then subjected to a oneꢀpot
iodination/desulfonylation and pyrrolopyridine Nꢀmethylation to afford coupling partner 58.
Iodides 43, 49, 53 and 58 were subjected to Suzuki crossꢀcoupling conditions with NꢀBocꢀ
tetrahydropyridineꢀ4ꢀboronic acid pinacol ester to provide compounds 59a–d respectively,
Scheme 11. Hydrogenation afforded piperidinyl intermediates 60a–d which were then converted
to the corresponding 3ꢀcyanobenzamides 61a–d. Subsequent deprotection provided piperidine
amines 62a–d. In the case of 4ꢀmethyl substituted intermediate 62a, a variety of amides 63a–l
were prepared to explore the SAR of this substituent. On the other hand, only isobutyrate amides
were prepared from 62b–d resulting in analogs 64 (4ꢀiPr), 65 (4ꢀOMe) and 66 (4ꢀCF3),
respectively. As shown in Scheme 12, compound 59a was also transformed to amine 68 to allow
for diversification of the benzamide substituent as in analogs 69a–d.
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
RESULTS AND DISCUSSION
Screening Strategy
A cellꢀbased GAL4 luciferase reporter assay is uniquely suited as a screening approach to
identify small molecule modulators of nuclear hormone receptors since the intrinsic
ACS Paragon Plus Environment
8

Journal of Medicinal Chemistry
Page 10 of 105
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
transcriptional activity of the receptor is the functional readout. As opposed to a radioligand
binding assay based on an endogenous ligand of the receptor, the reporter assay approach allows
for the identification of pharmacologically diverse hits that are either competitive with the
endogenous ligand or function by an allosteric binding site mechanism. A notable limitation of
the reporter assay, however, is the requirement that the modulator is cell permeable.
Nonetheless, as an initial screening strategy for a new target, an approach that quickly delivers a
cell permeable hit can greatly facilitate validation of a chemotype or mechanism through
functional activity in a diseaseꢀrelevant pharmacological system.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Consequently, a subset of the Pfizer screening collection (ca. 150 000 compounds) was
screened at 5 µM compound concentration using a GAL4ꢀRORC2 luciferase reporter assay.
HEK293 cells were transiently transfected with a mammalian expression vector containing
GAL4ꢀRORC2 LBD and a GAL4ꢀresponsive reporter gene containing firefly luciferase. An
RORC2 inverse agonist would be expected to suppress the constitutively active, transcriptionally
derived luminescence. Compounds demonstrating inhibition of greater than 50% were then
counterꢀscreened in a modification of the reporter assay employing a GAL4ꢀP65 expression
vector instead of RORC2. Undesired activity under these assay conditions would suggest nonꢀ
specificity for RORC2, transcriptional inhibition through a general cytotoxicity mechanism, or
compound specific interference with the luminescence readout. A dose response curve for
compounds demonstrating satisfactory differentiation was next acquired using comparable
luciferase reporter assays in murine Neuro2A cells transfected by either GAL4ꢀRORC2 LBD,
GAL4ꢀRORA LBD, or GAL4ꢀRORB LBD in order to characterize isoform selectivity. As a
result of this screening cascade, indole benzamide 3 was identified as a moderately potent
ACS Paragon Plus Environment
9

Page 11 of 105
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
RORC2 inverse agonist (IC50 = 2.7 µM) which lacked significant potency against RORA or
RORB (IC > 30 µM), Table 1.
5
0
In order to further confirm that the functional changes in gene transcription observed with
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
compound 3 resulted from specific ligand interactions with RORC2, we characterized compound
3
in a radioligand SPA binding assay and a time resolvedꢀfluorescence resonance energy transfer
3
(
TRꢀFRET) assay, Table 1. Compound 3 displaced [ H]ꢀ25ꢀhydroxycholesterol, a putative in
3
7
vitro agonist of RORC2, from purified human RORC2 LBD with a K of 1.2 µM, comparable
i
to the functional response in the reporter assay. The competitive behavior with 25ꢀ
hydroxycholesterol was the first suggestion that this new chemotype might bind within the
endogenous ligand binding site of the receptor. The TRꢀFRET assay monitors the ability of the
histidineꢀtagged RORC2ꢀLBD to bind a biotinylated coꢀactivator peptide (SRC1ꢀ2) by
fluorescence transfer mediated through association of a europium (Eu)ꢀlabelled antiꢀHis
antibody. An RORC2 inverse agonist that displaces the endogenous ligand would be expected to
decrease the affinity of the coꢀactivator peptide towards the RORC2 LBD and thus decrease the
resulting fluorescence intensity. Indeed, compound 3 demonstrated potent suppression of SRC1ꢀ
2
binding in the RORC2 TRꢀFRET assay (IC = 54 nM, 96% maximum suppression) while
50
again not showing significant inverse agonism towards RORA or RORB (IC > 30 µM) when
5
0
profiled in a similar assay format. Most importantly, compound 3 inhibited the production of ILꢀ
7 in human primary Th17 cells (IC = 2.9 µM). Unfortunately, compound 3 demonstrated
1
5
0
poor metabolic stability (HLM CL = 142 µL/min/mg) precluding its utility in in vivo
experiments. Nonetheless, encouraged by the evidence that compound 3 could functionally
inhibit production of ILꢀ17 in a disease relevant human cell, demonstrate competent linkage of
functional activity to binding of the RORC2 receptor, and display promising isoform selectivity,
ACS Paragon Plus Environment
10

Journal of Medicinal Chemistry
Page 12 of 105
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
we undertook efforts to further optimize the chemotype for both potency and pharmacokinetic
properties.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Lead Optimization
The topology of indole benzamide 3 provides three clear vectors for structureꢀactivity
exploration, namely the benzamide, the piperidine amide, and the indole nitrogen substituent.
Diversification of the core heterocycle was also considered in our optimization strategy. Our
first venture to explore the vector derived from the indole nitrogen was immediately productive.
NꢀMethyl analog 4 was found to be 16ꢀfold more potent than compound 3 in the GAL4ꢀRORC2
reporter assay (IC = 0.17 µM), Table 1. A similar significant improvement in potency was
5
0
observed in the RORC2 TRꢀFRET assay (IC = 7 nM) and in binding affinity based on the SPA
5
0
assay (K = 50 nM). High isoform selectivity was maintained as evident from no significant
i
inverse agonism of RORA or RORB in either the TRꢀFRET or reporter assays (IC > 30 µM).
5
0
Compound 4 was also a significantly more potent inhibitor of ILꢀ17 production in Th17 cells
IC = 67 nM, 90% maximum inhibition). Although the lipophilicity of 4 is slightly higher than
(
5
0
compound 3 (LogD = 4.4 and 4.2, respectively), the addition of a single methyl resulted in a net
positive improvement to lipophilic efficiency (ꢁ LIPE = 1.4).
From compound 4, exploration of potency and ADME space for the benzamide substituent and
the piperidine amide was well suited for application of parallel synthetic chemistry technology.
Although this effort did not identify a compound with significantly improved properties, it did
provide insight into the dominant interactions with the receptor and led to a compound suitable
for coꢀcrystallization with the receptor. The region of the receptor occupied by the piperidine
amide appeared to preferentially tolerate large hydrophobic residues; however, this did not lead
ACS Paragon Plus Environment
11

Page 13 of 105
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
to favorable lipophilic efficiency improvements. On the other hand, the cyano group present in
the benzamide substituent was found to play a dominant role in functional potency for the series.
Compound 7, where the cyano was replaced by hydrogen, demonstrated a 10ꢀfold reduction in
RORC2 TRꢀFRET and ILꢀ17 inhibition potency compared to lead compound 4, Table 1.
Replacement of phenyl by pyridine as in 4ꢀcyanopicolinamides 8a or 8b was one of the few
modifications of the benzamide moiety that was tolerated. Although compound 8b neither
demonstrated improved potency (ILꢀ17 IC = 42 nM) nor metabolic stability (HLM CL = 173
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
5
0
µL/min/mg), it was a successful candidate for coꢀcrystallization with the RORC2ꢀLBD.
3
8
The recently reported crystal structure of apoꢀRORC2 receptor LBD and earlier structures of
37
the receptor in complex with hydroxyl cholesterols have revealed that the receptor has a large
endogenous ligand binding pocket. Both of these structures display the receptor in an active
conformation with helixꢀ12 (H12) closely associated with the remainder of the receptor and
forming a cleft allowing for coꢀactivator peptide binding. A key structural element that stabilizes
the positioning of H12 and enabling the active receptor conformation is a triplet residue latch
consisting of His479–Tyr502–Phe506, Figure 1. The H12 residue Tyr502 forms a key hydrogen
bond with His479 which resides on H11, thus bridging the two helixes. In addition, Phe506
forms a favorable edge to face aromatic stacking arrangement with both His479 and Tyr502.
This arrangement of residues allows the charged His479 residue to have a strongly favorable
3
9
cationꢀπ interaction and hold H12 in the active conformation.
The coꢀcrystal structure of compound 8b with the RORC2 LBD confirmed that the ligand
occupies the endogenous ligand binding pocket, Figure 2. Most notably from the structure, H12
is highly disordered along with partial unwinding of H11′. As a result, H12 no longer forms the
coꢀactivator binding cleft or the hydrogen bond between His479 and Tyr502. In place of the key
ACS Paragon Plus Environment
12

Journal of Medicinal Chemistry
Page 14 of 105
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
latch interaction, His479 is now engaged in a hydrogen bond with the piperidine amide carbonyl
of ligand 8b, Figure 3a. The cyclohexyl amide substituent occupies a hydrophobic pocket
formed by H11 and the junction of H6 and H7. Compared to the active conformation of the
receptor, this pocket has significantly expanded in size due to a 110° torsional change in Trp317
with the vacancy only partially replaced by movement of Phe486. The rest of the binding site
pocket is bordered by H3, H5 and H7 and a βꢀsheeting underlying the indole core. The
important indole Nꢀmethyl substituent was found to buttress against H7 making hydrophobic
contacts with Ile400, Phe401, and Tyr369 within a small pocket, Figure 3b. Indeed, a
detrimental impact on inhibition of coactivator binding and functional ILꢀ17 production was
encountered by only increasing the substituent size to ethyl as in compound 11 (ILꢀ17 IC > 10
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
5
0
µM), Table 1. The indole ring itself is situated in a hydrophobic region of the receptor
positioned between Met365 on top and Val376 from the βꢀsheet below, Figure 3c. The
benzamide substituent occupies a noticeably more hydrophilic region of the pocket making two
key hydrogen bond contacts. First, the cyano group of the ligand makes a bifurcated hydrogen
bond to Arg367 and the backbone amide NH of Leu287. The second contact is between the
backbone carbonyl of Phe377 and the benzamide NH of the ligand. In order to accommodate
this hydrogen bond the amide bond torsion (C4–C5–N–C) must deviate from planarity with the
indole by 113°.
An analysis of available crystallographic data for similar aryl amides found in the Cambridge
Crystallographic Database suggests a strong preference for either of the two coꢀplanar rotamers
40
about the carbonꢀnitrogen bond. Therefore, in order to adopt the bound state dihedral angle, a
noteworthy energetic penalty must be paid with respect to the ground state conformation.
Consequently, structural changes to the ligand that would increase the predominance of the
ACS Paragon Plus Environment
13

Page 15 of 105
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
32
bound conformation in solution would be expected to improve potency. Filla and coworkers
have reported a strategy to influence the amide rotamer population in a closely related series of
Nꢀacylimidazole 5ꢀHT receptor agonists through replacement of the indole ring carbon atom by
nitrogen adjacent to the amide substituent. Their analysis of crystallographic data suggested a
preference for an antiꢀconformation between the ring nitrogen and the amide carbonyl although a
planar conformation was still highly favored. To probe this effect in our series, we prepared
pyrrolo[3,2ꢀb]pyridine 23 and pyrrolo[2,3ꢀc]pyridine 24, Table 2. Introduction of nitrogen into
the indole core was also attractive for the potential to lower lipophilicity and potentially improve
metabolic stability. To this effect, we also prepared pyrrolo[2,3ꢀb]pyridine 17 and pyrrolo[3,2ꢀ
d]pyrimidine 31. Unfortunately, compounds 24 and 31 demonstrated significantly reduced
potency in the TRꢀFRET assay compared to the indole isostere 8a. Although compounds 17 and
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
2
3 were potent in the TRꢀFRET assay, they demonstrated a modest reduction in potency for
suppression of ILꢀ17 production in Th17 cells. The most interesting compound of the series was
pyrrolo[2,3ꢀb]pyridine 17. Although its cellular potency was diminished, it afforded a 0.5 unit
reduction in lipophilicity and maintained comparable lipophilic efficiency compared to the
indole isostere 8a. Unfortunately, this reduction in lipophilicity was not sufficient to improve
metabolic stability (HLM CL = 144 µL/mg/min).
An alternative approach considered to alter the amide rotamer population was to substitute the
indole ring adjacent to the amide substituent. For the two possible positions, the coꢀcrystal
structure of 8b suggested that only substitution at the 4ꢀposition of the indole ring would be
tolerated in the binding pocket. To test this hypothesis we pursued 4ꢀmethylindole 36, Table 3.
As shown in the calculated torsional energy profile about the C4–C5–N–C dihedral angle of a
4
1
representative model system (Figure 4), substitution at the 4ꢀposition of the indole ring with
ACS Paragon Plus Environment
14

Journal of Medicinal Chemistry
Page 16 of 105
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
methyl significantly disfavors the proximalꢀorientation of the carbonyl and methyl (θ = 0°). At
the dihedral angle of the observed bound conformation, the calculation suggests that the
unsubstituted analog (R = H) suffers a higher energetic penalty (3.4 kcal/mol) compared to the
methyl analog (2.2 kcal/mol) with respect to the lowest energy conformation (θ = 180° or 0°).
This was indeed reflected in the potency of 4ꢀmethylindole 36, which was 23ꢀfold more potent in
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
the ILꢀ17 suppression assay (IC = 3.3 nM) compared to the unsubstituted analog 4 (IC = 75
50
50
nM), Table 3. A coꢀcrystal structure of compound 36 bound to the RORC2 LBD confirmed the
same binding orientation in the pocket to that observed with compound 8b, Figure 5. The
dihedral angle of the C–N bond (C4–C5–N–C) was maintained at 113° and the methyl
substituent now occupies a previously vacant cavity in the binding pocket bordered by His323
and Phe378. A similar phenomenon was observed for the pyrrolo[2,3ꢀb]pyridine ring (Figure
S1, See Supporting Information) where the 4ꢀmethyl derivative 63h was 13ꢀfold more potent
(IC = 21 nM) compared to the unsubstituted analog 18 (IC = 274 nM). In both cases, the
5
0
50
addition of the methyl substituent led to a reduction in lipophilicity, ꢁLogD = ꢀ0.4 and ꢀ0.9
respectively for compound 36 and 63h. As a result, through increased potency and reduced
lipophilicity, the combination of the pyrrolo[2,3ꢀb]pyridine and the 4ꢀmethyl substituent into
compound 63h led to a significant improvement in the lipophilic efficiency compared to
compound 4 (ꢁLIPE = 2.1).
The methyl substituent at the 4ꢀposition of the indole ring may also have a profound effect on
the torsional dynamics between the piperidine and indole rings. In coꢀcrystal structures of both
8
b and 36, the piperidine ring adopts a dihedral angle of 133° (C2–C3–C–H). The torsional
energy profile of this dihedral was also calculated using a simplified model system, Figure 4. In
order to adopt the observed bound conformation, the calculations suggest that the methyl analog
ACS Paragon Plus Environment
15

Page 17 of 105
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
suffers a minor energetic penalty (0.4 kcal/mol) compared to the unsubstituted analog. However,
the methyl group imparts a significantly higher rotational barrier (9.8 kcal/mol) compared to the
unsubstituted analog (3.6 kcal/mol). Likewise, the methyl analog displays a greater energetic
preference (ꢁ ꢀ2.4 kcal/mol) for the bound conformation than the next higher local minimum (θ
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
=
0°) as compared to the hydrogen analog (ꢁ ꢀ0.3 kcal/mol). Again, a similar profile was
observed for the pyrrolo[2,3ꢀb]pyridine ring (Figure S1, See Supporting Information).
It should be noted that a similar potency enhancement as observed for inhibition of ILꢀ17
production was not observed in the recombinant RORC2 TRꢀFRET assay for compounds 36 and
6
3h, Table 3. In order to further understand this apparent disconnect, we explored the
dissociation kinetics of our inverse agonists from the RORC2 receptor using an adaptation of the
TRꢀFRET assay. The corresponding inverse agonist was incubated with the receptor for 3 h at
the respective concentration to achieve 75% inhibition of response. A high concentration of
agonist was then added and the return of TRꢀFRET signal as a result of SRC1ꢀ2 binding was
monitored over time. Using this method we estimated the effective dissociation halfꢀlife of
compound 36 and 63h from the recombinant receptor to be 21.8 h and 10.5 h, respectively. In
both cases, the dissociation rate decreased by more than an order of magnitude by installing the
4
ꢀmethyl substituent, Table 3. The halfꢀlife of the receptor in human Th17 cells as determined
by pulseꢀchase experiments was found to be 5.0 h and 5.1 h either in the presence or absence of
ligand, respectively (Figure S2, See Supporting Information), and thus supports the potential for
a sustained pharmacological effect dependent on the compound dissociation rate in the case of
42
these compounds. Swinney has attributed one characteristic of antagonists that possess high
biochemical efficiency (BE) as the capacity to induce a nonꢀequilibrium system such as having a
residence time on the target longer than the degradation rate of the receptor. The effective
ACS Paragon Plus Environment
16

Journal of Medicinal Chemistry
Page 18 of 105
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
pseudoꢀirreversible behavior of these compounds contributes to their high efficiency in the
cellular assay. A subsequent screen of a range of inverse agonists demonstrated a strong
correlation of effective dissociation halfꢀlife with ILꢀ17 inhibitory potency (Figure S3, See
Supporting Information). As a consequence, the measurement of ILꢀ17 inhibition in Th17 cells
was viewed as the most appropriate assay to rank order compound potency given the presence of
the endogenous ligand in physiological concentrations and the long assay window (6 days) to
achieve equilibrium.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Unfortunately, even though a significant reduction in lipophilicity had been achieved with
compound 63h, metabolic clearance remained high (HLM CL = 125 µL/min/mg). In order to
further address metabolic stability, we revisited optimization of the piperidine amide substituent
in light of the advancements made elsewhere to the molecule. A range of acyclic, cyclic, and
aromatic amides were prepared spanning size and hydrophobicity, Table 4. The ability of the
compound to inhibit ILꢀ17 production was highly dependent on the size and hydrophobicity of
the substituent. For example, potency increased exponentially as the size of the acyclic group
increased from methyl (63a, IC = 5,270 nM) to ethyl (63b, IC = 625 nM) and eventually
5
0
50
isopropyl (63d, IC = 69 nM). Similarly, for cycloalkyl substituents, potency increased with
50
ring size from cyclopropyl (63c, IC = 242 nM) to cyclohexyl (63i, IC = 13 nM). An example
5
0
50
of an aromatic substituent, 63j, also showed excellent potency. Although isosteric with the
potent analogs 63h and 63j, the polar analogs based on tetrahydrofuran (63k) and 3,5ꢀ
dimethylisoxazole (63l) were significantly less potent. On the other hand, metabolic stability
was inversely proportional to the size of the hydrophobic substituent in 63aꢀj. Overall, the
isopropyl analog 63d demonstrated the best combination of potency, metabolic stability, and
lipophilic efficiency (IC = 69 nM, HLM CL = 8 µL/min/mg).
5
0
ACS Paragon Plus Environment
17

Page 19 of 105
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Now with a better understanding of the chemical space required to achieve good metabolic
stability, we pursued further optimization of potency. First, we considered additional
substitution to the cyanobenzamide. Based on the coꢀcrystal structures of 8b and 36, the 4ꢀ
position of the phenyl ring appeared to provide the greatest opportunity for additional
hydrophobic or hydrogen bond contacts, Table 5. For the analogs prepared (69a–d), all of them
showed good metabolic stability except for the 4ꢀchloro analog 69c, presumably due to the
significant increase in lipophilicity. Unfortunately, only the 4ꢀmethoxy analog 69a was more
potent, although only modestly (2ꢀfold), for inhibition of ILꢀ17 production.
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Finally, we investigated the scope of the pyrrolo[2,3ꢀb]pyridine 4ꢀposition substituent beyond
the initially identified methyl. We surveyed examples in which the substituent was sterically
larger or had a different electronic character compared to methyl (64–66), Table 6. Replacing
the methyl by the sterically larger isopropyl group (64) was predicted through torsional energy
calculations to yield the smallest energetic penalty for the anticipated bound state, Figure 4.
Nonetheless, compound 64 imparted comparable potency and metabolic stability as compound
6
3d, possibly due to counterbalancing steric interactions with His323 and Phe378. Compound
5 bearing an electron donating methoxy group, however, demonstrated reduced potency and
6
higher metabolic clearance. This would be anticipated from the torsional energy profile with
methoxy yielding the largest energetic penalty presumably due to a favorable intramolecular
interaction between the amide hydrogen and the oxygen lone pair when the C4–C5–N–C
dihedral angle is 180°. On the other hand, an electron withdrawing trifluoromethyl substituent
43
(
66) provided improved potency compared to methyl. Compound 66 achieved optimal
inhibition of ILꢀ17 production in Th17 cells (IC = 9.5 nM, 90% maximum inhibition) while
5
0
affording excellent metabolic stability. Favorable metabolic stability was observed for
ACS Paragon Plus Environment
18

Journal of Medicinal Chemistry
Page 20 of 105
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
compound 66 in spite of a significant increase in lipophilicity induced by the trifluoromethyl
group (ꢁLogD = 1.7) which is consistent with the trend observed in reported matched molecular
4
4
pair analyses. Several factors likely contribute to the enhanced potency of the trifluoromethyl
analog 66. Calculations suggest the trifluoromethyl group may further skew the C4–C5–N–C
rotamer population toward the bound conformation through an unfavorable dipole interaction
with the amide carbonyl as well as favor the bound conformation of the piperidine ring, Figure 4.
Secondly, the electron withdrawing nature of the trifluoromethyl group now reduces the
polarization of the pyrrolopyridine nitrogen, which would be better accommodated in the
hydrophobic binding pocket. This is supported by the reduced negative character for the
calculated electrostatic potential (ESP) of the ring nitrogen compared to methyl analog 63d,
Table 6. Lastly, the lipophilicity of compound 66 is now of the same magnitude of earlier leads
with single digit nanomolar potency (e.g. 36, LogD = 4.0), and therefore, the hydrophobic nature
of the binding pocket may be a strong driver of potency.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Encouraged by the initial in vitro profile of compound 66, we carried out additional
pharmacological characterization. The very slow dissociation rate from the RORC2 recombinant
receptor seen previously with 4ꢀsubstituted derivatives was also evident (T½ = 6 h) in compound
6
6. To demonstrate that compound effect was at the level of gene transcription, we showed that
compound 66 inhibited mRNA production of ILꢀ17A and other genes that have previously been
described as being controlled by RORC2 such as ILꢀ17F, ILꢀ22, ILꢀ26 and ILꢀ23R (Table S1,
1
5,45
See Supporting Information).
As expected, RORC expression was not inhibited by
compound 66. In addition, Th1 and Th2 lymphocyte differentiation as well as cellular viability
of lymphocytes were not affected by compound 66 at the highest concentration tested (10 ꢂΜ)
suggesting specificity of the observed effects. High isoform selectivity was demonstrated with
ACS Paragon Plus Environment
19

Page 21 of 105
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
no significant inverse agonism towards RORA or RORB in the TRꢀFRET assay (IC > 25 µM).
5
0
To more broadly explore the receptor selectivity of compound 66, we profiled the compound
using the transꢀFACTORIAL™ platform (Attagene Inc), Figure 6. This multiplexed technology
allows for the measurement of reporter RNA levels upon transfection of ligand binding domain
chimeric constructs from 48 nuclear receptors with GAL4 DNA. Of the 48 receptors, only
RORC2 showed a fold reduction (antagonist effect) of greater than 20% of control at a dose of 1
µM compound. Only the estrogenꢀrelated receptor alpha (ERRα) (1.28ꢀfold, p=0.03) and the
liver X receptor alpha (LXRα) (1.30ꢀfold, p=0.165) showed a modest fold increase (agonist
effect) greater than 20% of control at the same dose. These results clearly demonstrate the high
selectivity of compound 66 for RORC2 compared to all other nuclear receptors.
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
In addition to low in vitro human microsomal and hepatocyte clearance, compound 66 also
ꢀ6
demonstrated high passive permeability (Papp = 13.2×10 cm/sec, RRCK cell monolayer). Our
primary concern with regard to predicted human pharmacokinetics was the relatively low
thermodynamic solubility of compound 66 (1.0 µM, pH 6.5). In a rat pharmacokinetic study,
compound 66 demonstrated low in vivo clearance (CL = 3.4 mL/min/kg) and a moderate halfꢀ
life (6 h) consistent with the low human (8 µL/min/mL) and rat (<14 µL/min/mL) microsomal
clearance, Table 7. Clearance was also very low in dog and mouse (0.46 and 1.2 mL/min/kg,
respectively) with the elimination halfꢀlife in the dog being exceptionally long (39 h). Upon oral
dosing of crystalline solid (3 mg/kg) in rats, oral bioavailability was moderate (21%), consistent
with a highly permeable but low solubility compound. Unfortunately, due to the decreased
basicity of the pyrrolopyridine nitrogen, salt formation was not a viable approach for solubility
enhancement. In hopes of improving the compound dissolution rate and ultimately the oral
bioavailability, we formulated the compound as a sprayꢀdried dispersion (SDD) (25%,
ACS Paragon Plus Environment
20

Journal of Medicinal Chemistry
Page 22 of 105
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
46
HPMCASꢀH polymer).
This approach was highly effective resulting in an increased oral
bioavailability in rat of 83%. Oral bioavailability in dog remained high with the SDD
formulation (73%, 3 mg/kg). Oral bioavailability in mouse was also acceptable at higher doses
using the SDD formulation (30%, 30 mg/kg). Overall, compound 66 demonstrated excellent in
vivo pharmacokinetics consistent with in vitro parameters, and the SDD formulation approach
was very effective in overcoming the low aqueous solubility of the molecule.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Compound 66 also showed potent inhibition of ILꢀ17 production in mouse derived Th17 cells
(
IC = 32 nM, 92% maximum inhibition) which supported assessment of in vivo efficacy in the
50
4
7
mouse imiquimodꢀinduced skin inflammation model. Topical administration of imiquimod can
induce a psoriasisꢀlike skin inflammation resulting in epidermal thickening. This clinical
outcome is at least in part mechanistically dependent on the ILꢀ23/ILꢀ17 pathway. Mice were
challenged with topical imiquimod cream (5%) applied to their back and ear for 3 days (days 1ꢀ
3). Compound 66 was dosed orally at three dose levels (10, 30 and 100 mg/kg) once a day over
five days (days 1ꢀ5). Compound administration was well tolerated without any compoundꢀ
related adverse events. Compound plasma AUC and Cmax increased with dose although not
proportionally, Figure 7. An antiꢀILꢀ17 monoclonal antibody and an antiꢀp40 (ILꢀ23 and ILꢀ12
subunit) antibody were administered by intraperitoneal injection as controls. At the conclusion
of the dosing, ear thickness was measured and ILꢀ17A protein levels in the ear were assessed,
Figure 8. Compound 66 demonstrated a dose dependent inhibition of ear swelling in the model.
A maximum inhibition of 46% (p<0.0001) was achieved at a dose of 100 mg/kg which was
comparable to the reduction observed with the antiꢀILꢀ17A Ab (45%, p=0.0005) compared to the
isotype control. The degree of inhibition correlated with the time period in which compound
concentration exceeded the IC throughout the dosing interval, Figure 7. Consistent with the
8
0
ACS Paragon Plus Environment
21

Page 23 of 105
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
clinical outcome and the role RORC2 plays, a significant reduction in the level of ILꢀ17A protein
in the ear was also observed at all doses of compound 66 compared to the vehicle control, Figure
8. Similar reductions of ILꢀ17A protein in the ear were seen at both the 30 and 100 mg/kg doses
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
(70% and 72%, respectively) (p<0.0001).
CONCLUSION
Through a combination of de novo and structureꢀguided design, we have optimized a novel
highꢀthroughput screening hit against RORC2 with modest potency and poor metabolic stability
into a highly potent and orally bioavailable lead. In order to align the two properties of potency
and metabolic stability into a single molecule, it was important to exploit two key design
strategies, namely, iterative optimization through lipophilic efficiency and conformational
restriction to achieve optimal ground state energetics and maximize receptor residence time. The
4
8,49
introduction of two “magic methyl” substituents
profoundly impacted the potency
characteristics of the chemotype. The first example was transformation of compound 3 to 1ꢀ
methylindole 4 resulting in a 43–fold improvement in potency. As revealed by the subsequent
coꢀcrystal structure of compound 8b, this methyl group occupies a favorable hydrophobic pocket
in the receptor yielding a beneficial lipophilic efficiency (ꢁLIPE = 1.4). The second example
was the transformation of compound 4 to 4ꢀmethylindole 36 resulting in an additional 20–fold
potency improvement and further lipophilic efficiency improvement (ꢁLIPE = 1.7). In this case,
the methyl substituent is believed to restrict the conformational dynamics of the ligand leading to
a lower solution to bound state energetic penalty as well as imparting a profound resonance time
enhancement to the ligand in the binding pocket. Although largely neutral towards lipophilic
efficiency, the next two prominent changes to the molecule provided an important lowering of
ACS Paragon Plus Environment
22

Journal of Medicinal Chemistry
Page 24 of 105
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
the intrinsic lipophilicity. These changes included the introduction of the pyrrolo[2,3ꢀb]pyridine
ring system (63h, ꢁLogD = ꢀ1.1) and replacement of the cyclopentylamide by isopropylamide
(63d, ꢁLogD = ꢀ0.9). Effectively, this series of transformations achieved a comparable level of
potency to the early lead compound 4, but in a substantially lower lipophilic space (ꢁLogD = ꢀ
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
2
.4) and now with excellent metabolic stability. The reduction in lipophilicity was crucial to
allow for the final potency enhancement through introduction of the metabolically stable but
lipophilic 4ꢀtrifluoromethyl group found in compound 66. Indeed, compound 66 was found to be
a potent and selective RORC2 inverse agonist with excellent pharmacokinetic properties in
preclinical species. The ability of compound 66 to reduce levels of ILꢀ17A in vivo after oral
dosing in mice and corresponding reduction in skin inflammation further supports the potential
of small molecule RORC2 modulation as a therapeutic target for the treatment of inflammatory
diseases.
EXPERIMENTAL SECTION
Chemistry. All reagents and solvents were used as purchased without further purification. The
purity of the final compounds was characterized by highꢀperformance liquid chromatography
(HPLC) using a gradient elution program (e.g. C18, acetonitrile:water, 0.1% formic acid, 5:95–
9
5:5) and UVꢀdetection (220 nM). The purity of all final compounds was 95% or greater.
1
Proton ( H) NMR chemical shifts are referenced to a residual solvent peak.
tert-Butyl 4-(5-(3-cyanobenzamido)-1H-indol-3-yl)piperidine-1-carboxylate (2). DIPEA
(0.45 mL, 2.6 mmol) was added to a solution of 3ꢀcyanobenzoic acid (0.367 g, 2.5 mmol) and
TBTU (0.881 g, 2.6 mmol) in DMF (10 mL). After stirring at room temperature for 30 min, a
3
1
solution of tertꢀbutyl 4ꢀ(5ꢀaminoꢀ1Hꢀindolꢀ3ꢀyl)piperidineꢀ1ꢀcarboxylate (1, 0.787 g, 2.5
ACS Paragon Plus Environment
23

Page 25 of 105
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
mmol) in DMF (5 mL) was added and the mixture was stirred at room temperature overnight.
The reaction mixture was concentrated and the residue was purified by silica gel column
1
chromatography (MeOH/CH Cl , 2/98) to afford 0.8 g (72%) of 2. H NMR (300 MHz, CDCl )
2
2
3
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
δ 9.45 (br s, 1H), 9.10 (br s, 1H), 8.25 (s, 1H), 8.17 (d, J = 8.2 Hz, 1H), 7.95 (s, 1H), 7.65 (d, J =
8
4
.0 Hz, 1H), 7.45 (t, J = 7.8, 1H), 7.28 (d, J = 8.2 Hz, 1H), 7.20 (d, J = 8.2 Hz, 1H), 6.80 (s, 1H),
.20–4.05 (m, 2H), 2.80–2.60 (m, 3H), 1.95–1.80 (m, 2H), 1.55–1.28 (s, 11H).
3
-Cyano-N-(3-(1-(cyclopentanecarbonyl)piperidin-4-yl)-1H-indol-5-yl)benzamide (3). A
solution of 4N HCl in dioxane (0.9 mL) was added to a solution of 2 (201 mg, 0.45 mmol) in
methanol (2 mL). The mixture was stirred at room temperature for 2.5 h and then concentrated
under reduced pressure at 30 ºC. The residue was suspended in toluene and concentrated. The
resulting crude amine was dissolved in DMF (1 mL) and DIPEA (0.37 mL, 2.24 mmol) was
added. The solution was stirred at room temperature for 30 min and then a solution of
cyclopentanecarbonyl chloride (86 mg, 0.65 mmol) in dichloroethane (1.5 mL) was added. The
mixture was stirred at room temperature overnight, and then quenched with saturated aqueous
NaHCO solution. The aqueous layer was extracted with EtOAc. The combined organic layers
3
were washed with brine, dried (Na SO ), filtered and then concentrated. The residue was
2
4
purified by silica gel column chromatography (MeOH/CH Cl , 0/100–10/90). The crude product
2
2
was dissolved in EtOAc, washed with water and concentrated to afford 180 mg (91%) of 3 as a
1
tan solid. MS (ES+) m/z 441 (M+H). H NMR (300 MHz, CDCl ) δ 8.40 (s, 1H), 8.35 (s, 1H),
3
8.22 (s, 1H), 8.18 (d, J = 8.0, 1H), 8.00 (s, 1H), 7.80 (d, J = 8.5, 1H), 7.60 (t, J = 8.3, 1H), 7.30–
7
.20 (m, 2H), 6.95 (s, 1H), 4.75–4.65 (m, 2H), 4.10–4.00 (m, 2H), 3.20–2.85 (m, 3H), 2.70–2.60
(m, 2H), 2.15–1.50 (m, 9H).
ACS Paragon Plus Environment
24

Journal of Medicinal Chemistry
Page 26 of 105
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
3
-Cyano-N-(3-(1-(cyclopentanecarbonyl)piperidin-4-yl)-1-methyl-1H-indol-5-yl)-
benzamide (4). NaH (11 mg, 0.27 mmol) was added to a cold (0 ºC) solution of 3 (100 mg, 0.23
mmol) in THF (10 mL). The mixture was stirred at room temperature for 30 min, and then
iodomethane (38 mg, 0.27 mmol) was added dropwise while maintaining 0 ºC. The reaction
mixture was stirred at room temperature for 3 h and then was poured into saturated aqueous
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
NH Cl solution (3 mL). The organic layer was concentrated and the crude product was purified
4
by preparative HPLC to afford 22 mg (21%) of 4 as a light yellow solid. MS (ES+) m/z 455
1
(
M+H). H NMR (400 MHz, CD OD) δ 7.69 (s, 1H), 7.48–7.46 (m, 2H), 7.30–7.25 (m, 3H),
3
7
.04 (s, 1H), 7.00–6.99 (m, 1H), 4.68–4.60 (m, 1H), 4.20–4.13 (m, 1H), 3.53 (s, 3H), 3.26–3.20
(m, 1H), 3.15–3.05 (m, 1H), 3.05–2.95 (m, 1H), 2.80–2.75 (m, 1H), 1.89–1.80 (m, 4H), 1.76–
1
.60 (m, 6H), 1.55–1.40 (m, 2H).
tert-Butyl 4-(5-amino-1-methyl-1H-indol-3-yl)piperidine-1-carboxylate (5). NaH (2.1 g,
52.48 mmol, 60% w/w in mineral oil) was added to a cold (0 ºC) solution of tertꢀbutyl 4ꢀ(5ꢀnitroꢀ
3
1
1
Hꢀindolꢀ3ꢀyl)ꢀ5,6ꢀdihydropyridineꢀ1(2H)ꢀcarboxylate (4.5 g, 13.12 mmol) in THF (80 mL).
The mixture was stirred at room temperature for 1 h and then iodomethane (3.3 mL, 52.48
mmol) was added dropwise at 0 ºC. The reaction mixture was then allowed to stir at room
temperature overnight. The mixture was quenched by addition of ice/water and then extracted
by with ethyl acetate (2×100 mL). The combined organic layers were dried (Na SO ), filtered
2
4
and concentrated. Methanol (50 mL) and Pd/C (0.1 g) were added to the residue, and the
mixture was heated at 40 ºC under a hydrogen atmosphere (Balloon pressure) for 7 h. The
reaction mixture was filtered through Celite® and washed with methanol. The combined filtrate
1
was concentrated to obtain 3.4 g (82%) of 5 as a light brown solid. H NMR (400 MHz, DMSOꢀ
ACS Paragon Plus Environment
25

Page 27 of 105
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
d6) δ 7.04 (d, J = 8.4 Hz, 1H), 6.88 (s, 1H), 6.69 (s, 1H), 6.51 (dd, J = 8.4, 1.6 Hz, 1H), 4.47 (s,
2
H), 4.06–4.03 (m, 2H), 3.60 (s, 3H), 3.09–2.75 (m, 5H), 1.87 (d, J = 12.4 Hz, 2H), 1.41 (s, 9H).
tert-Butyl 4-(5-benzamido-1-methyl-1H-indol-3-yl)piperidine-1-carboxylate (6a). Benzoyl
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
chloride (0.41 mL, 3.52 mmol), triethylamine (0.13 mL, 0.91 mmol) and a catalytic amount of
DMAP were added to a solution of indole 5 (301 mg, 0.91 mmol) in CH Cl (2 mL). The
2
2
mixture was stirred at room temperature for 1 h. The reaction mixture was concentrated and the
residue was purified by silica gel column chromatography (heptane/EtOAc, 100/0–0/100) to
afford 100 mg (25%) of 6a as an offꢀwhite solid.
tert-Butyl
4-(5-(4-cyanopyridine-2-carboxamido)-1-methyl-1H-indol-3-yl)piperidine-1-
carboxylate (6b). To a cold (0 ºC) solution of 5 (3.3 g, 10.0 mmol) in CH Cl (20 mL) was
2
2
added triethylamine (4.2 mL, 30.1 mmol) followed by a solution of 4ꢀcyanopicolinoyl chloride
2.1 g, 12.53 mmol) in CH Cl (10 mL). After 1 h, the reaction mixture was poured into water
(
2
2
and extracted with CH Cl . The organic layers were dried (Na SO ), filtered and concentrated.
2
2
2
4
The crude compound was purified by silica gel column chromatography to afford 3.7 g (80%) of
1
6
8
1
1
b as a light brown solid. H NMR (400 MHz, CDCl ) δ 9.88 (s, 1H), 8.82 (d, J = 4.8 Hz, 1H),
3
.13 (s, 1H), 7.71 (dd, J = 4.8, 1.2 Hz, 1H), 7.47 (dd, J = 8.8, 2.0 Hz, 1H), 7.28 (d, J = 8.8 Hz,
H), 6.84 (s, 1H), 4.23–4.18 (m, 2H), 3.76 (s, 3H), 3.15–2.88 (m, 3H), 2.03 (d, J = 12.8 Hz, 2H),
.65–1.54 (m, 2H), 1.49 (s, 9H).
N-(3-(1-(Cyclopentanecarbonyl)piperidin-4-yl)-1-methyl-1H-indol-5-yl)benzamide (7). A
solution of 4 M HCl in dioxane (3 mL) was added to 6a (100 mg, 0.23 mmol). The reaction
mixture was stirred at room temperature for 1 h and then concentrated. Triethylamine (160 µl,
1
.15 mmol) and cyclopentanecarbonyl chloride (28 µl, 0.23 mmol) were added to a solution of
the residue in CH Cl (1 mL), and the mixture was stirred at room temperature for 1.5 h. The
2
2
ACS Paragon Plus Environment
26

Journal of Medicinal Chemistry
Page 28 of 105
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
reaction mixture was concentrated, and the residue was purified by silica gel column
chromatography (heptane/EtOAc, 100/0–0/100) to afford 39 mg (39%) of 7 as a white solid. MS
1
(ES+) m/z 430 (M+H). H NMR (400 MHz, CDCl ) δ 8.08 (s, 1H), 7.96–7.82 (m, 2H), 7.64–
3
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
7.44 (m, 2H), 7.37–7.26 (m, 4H), 6.83 (s, 1H), 4.80 (d, J = 13.3 Hz, 1H), 4.09 (d, J = 13.7 Hz,
1H), 3.76 (s, 2H), 3.21 (t, J = 12.1 Hz, 1H), 3.11 (t, J=12.1 Hz, 1H), 2.96 (quint, J = 8.1 Hz, 1H),
2.75 (t, J = 13.1 Hz, 1H), 2.27–1.98 (m, 2H), 1.92–1.45 (m, 9H), 1.28 (br. s., 1H), 0.89 (t, J = 6.6
Hz, 1H).
General method for the synthesis of 8a-b. A solution of 4M HCl in dioxane (30 mL) was
added to a cold (0 ºC) solution of 6b (3.7 g, 8.04 mmol) in methanol (15 mL). The mixture was
allowed to warm to room temperature and stir overnight. The reaction mixture was concentrated,
added to a saturated aqueous NaHCO solution, and extracted with 10% methanol in CH Cl .
3
2
2
The combined extracts were dried (Na SO ), filtered and concentrated to afford 2.2 g (78%) of
2
4
the amine. The corresponding carboxylic acid (1.11 mmol) and TBTU (430 mg, 1.34 mmol)
were dissolved in DMF (15 mL) and DIPEA (720 mg, 5.57 mmol) was added. After stirring at
4
0 ºC for 1 h, amine (400 mg, 1.11 mmol) was added and the mixture was stirred overnight at 40
ºC. The crude product was purified by preparative HPLC.
-Cyano-N-(3-(1-(cyclopentanecarbonyl)piperidin-4-yl)-1-methyl-1H-indol-5-
4
yl)picolinamide (8a). Cyclopentanecarboxylic acid; 21 mg (4%), yellow solid. MS (ES+) m/z
1
4
56 (M+H). H NMR (400 MHz, DMSOꢀd ) δ 10.65 (s, 1H), 9.01 (d, J = 4.8 Hz, 1H), 8.51 (s,
6
1H), 8.23 (d, J = 1.6 Hz, 1H), 8.18 (dd, J = 4.8, 1.2 Hz, 1H), 7.65 (dd, J = 8.8, 1.6 Hz, 1H), 7.40
(d, J = 8.8 Hz, 1H), 7.15 (s, 1H), 4.60–4.52 (m, 1H), 4.18–4.09 (m, 1H), 3.75 (s, 3H), 3.28–3.16
(m, 1H), 3.10–2.98 (m, 1H), 2.78–2.68 (m, 2H), 2.10–1.95 (m, 2H), 1.76–1.40 (m, 10H).
ACS Paragon Plus Environment
27

Page 29 of 105
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
4
-Cyano-N-(3-(1-(cyclohexanecarbonyl)piperidin-4-yl)-1-methyl-1H-indol-5-
yl)picolinamide (8b). Cyclohexanecarboxylic acid; 40 mg (8%), yellow solid. MS (ES+) m/z
1
470 (M+H). H NMR (400 MHz, DMSOꢀd ) δ 10.62 (s, 1H), 8.99 (d, J = 4.8 Hz, 1H), 8.49 (s,
6
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
H), 8.21 (d, J = 1.6 Hz, 1H), 8.16 (dd, J = 4.8, 1.6 Hz, 1H), 7.63 (dd, J = 8.8, 1.6 Hz, 1H), 7.38
d, J = 8.8 Hz, 1H), 7.14 (s, 1H), 5.77 (s, 1H), 4.58–4.52 (m, 1H), 4.10–4.04 (m, 1H), 3.87 (s,
H), 3.24–3.15 (m, 1H), 3.06–2.95 (m, 1H), 2.72–2.58 (m, 2H), 2.08–1.95 (m, 2H), 1.74–1.10
(
3
(
m, 12H).
tert-Butyl
4-(5-(4-cyanopicolinamido)-1H-indol-3-yl)piperidine-1-carboxylate
(9).
Triethylamine (14.6 mL, 105 mmol) was added to a suspension of tertꢀbutyl 4ꢀ(5ꢀaminoꢀ1Hꢀ
3
1
indolꢀ3ꢀyl)piperidineꢀ1ꢀcarboxylate (1, 8.25 g, 26.2 mmol) in CH Cl (120 mL) and the mixture
2
2
was cooled to 0 °C. A solution of 4ꢀcyanopicolinoyl chloride (4.49 g, 26.9 mmol) in CH Cl (60
2
2
mL) was slowly added to the mixture. The reaction mixture was stirred at 0 °C for 1 h and then
at room temperature for 20 h. Saturated aqueous sodium chloride and the mixture was extracted
with CH Cl and EtOAc. The combined organic layers were concentrated. The residue was
2
2
triturated with ethanol (150 mL) and filtered. The resulting solid was washed with ethanol (50
mL) and dissolved in a mixture of acetone and CH Cl . The solution was filtered and
2
2
1
concentrated to afford 12.16 g (95%) of 9 as a yellow solid. H NMR (400 MHz, DMSOꢀd ) δ
6
1
0.82 (s, 1H), 10.55 (s, 1H), 8.95 (d, J = 5.2 Hz, 1H), 8.45 (s, 1H), 8.15–8.10 (m, 2H), 7.54 (dd,
J = 9.1, 1.0 Hz, 1H), 7.3 (d, J = 9.1 Hz, 1H), 7.12 (d, J = 1.0 Hz, 1H), 4.10–4.00 (m, 2H), 2.95–
2.85 (m, 3H), 1.90–1.80 (m, 2H), 1.60–1.46 (m, 2H), 1.40 (s, 9H).
tert-Butyl
4-(5-(4-cyanopicolinamido)-1-ethyl-1H-indol-3-yl)piperidine-1-carboxylate
(
10). A solution of sodium hydroxide (2.5 M, 6.28 mL, 15.7 mmol), ethyl iodide (1.08 mL, 13.5
mmol) and a catalytic amount of Aliquat 336 was added to a solution of 9 (200 mg, 0.45 mmol)
ACS Paragon Plus Environment
28