A reversed sulfonamide series of selective RORc inverse agonists

Article abstract of DOI:10.1016/j.bmcl.2014.10.037

The identification of a new series of RORc inverse agonists is described. Comprehensive structure-activity relationship studies of this reversed sulfonamide series identified potent RORc inverse agonists in biochemical and cellular assays which were also selective against a panel of nuclear receptors. Our work has contributed a compound that may serve as a useful in vitro tool to delineate the complex biological pathways involved in signalling through RORc. An X-ray co-crystal structure of an analogue with RORc has also provided useful insights into the binding interactions of the new series.

Full text of DOI:10.1016/j.bmcl.2014.10.037

Bioorganic & Medicinal Chemistry Letters 24 (2014) 5769–5776
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
A reversed sulfonamide series of selective RORc inverse agonists
b
a
a
a
Monique B. van Niel a, , Benjamin P. Fauber , Matthew Cartwright , Simon Gaines , Jonathan C. Killen ,
Olivier René b, Stuart I. Ward a, Gladys de Leon Boenig b, Yuzhong Deng b, Céline Eidenschenk b,
Christine Everett b, Emanuela Gancia a, Arunima Ganguli a, Alberto Gobbi b, Julie Hawkins a,
Adam R. Johnson b, James R. Kiefer b, Hank La b, Peter Lockey a, Maxine Norman a, Wenjun Ouyang b,
Ann Qin b, Nicole Wakes a, Bohdan Waszkowycz a, Harvey Wong b
⇑
a Argenta, Early Discovery, Charles River, 7-9 Spire Green Centre, Flex Meadow, Harlow, Essex CM19 5TR, United Kingdom
b Genentech, Inc., 1 DNA Way, South San Francisco, CA 94080, USA
a r t i c l e i n f o
a b s t r a c t
Article history:
The identification of a new series of RORc inverse agonists is described. Comprehensive structure-activity
relationship studies of this reversed sulfonamide series identified potent RORc inverse agonists in
biochemical and cellular assays which were also selective against a panel of nuclear receptors. Our work
has contributed a compound that may serve as a useful in vitro tool to delineate the complex biological
pathways involved in signalling through RORc. An X-ray co-crystal structure of an analogue with RORc
has also provided useful insights into the binding interactions of the new series.
Received 5 September 2014
Revised 9 October 2014
Accepted 14 October 2014
Available online 25 October 2014
Keywords:
RORc
Ó 2014 Elsevier Ltd. All rights reserved.
ROR
ROR
c
c
t
IL-17
Inflammation
X-ray structure
A large family of 48 nuclear receptors (NRs) has been character-
agonists increase transcription but do so to a lesser extent than full
agonists, and partial inverse agonists elicit reduced efficacy relative
to the response of a full inverse agonist.
Retinoic acid receptor-related orphan receptors (RORs) are new
members of the NR family which show constitutive activity in the
absence of ligands.3 Isoforms of ROR (human isoforms RORa, b and
ised in humans and natural ligands are known for approximately
half of the members.1 The architecture of NRs forms five common
domains including an amino-terminal activation domain (AF-1), a
second activation domain (AF-2), a DNA binding domain (DBD), a
hinge region, and a hydrophobic ligand binding domain (LBD)
located at the C-terminus.2 Binding of ligands in the LBD causes a
conformational change which increases or decreases the recruit-
ment of large proteins, a key process that is required to influence
downstream gene regulation.3 Ligands are classified according to
their role in increasing or decreasing the protein recruitment.
The term inverse agonist is used when ligands induce conforma-
tional changes in the protein which reduce the recruitment of
co-activator proteins and leads to the silencing of target gene
expression.4 Conversely, the term agonist is used for ligands which
increase co-activator recruitment and results in enhanced gene
expression. There is a continuum between these two extreme
states. Ligands can also be termed antagonists if they reverse the
activity of an agonist-induced system or silent ligands if they bind
but induce no functional effect on the system. In addition, partial
c are also known as ROR
physiological processes and have distinct patterns of tissue distri-
bution.5 Murine ROR
exists as two isoforms, termed ROR and
ROR t, which have played a critical role in developing an under-
standing of the pathway and functioning of this NR. Tissue
distribution of the ROR isoforms show high levels of expression
a, b and c) are involved in a variety of
c
c
c
c
in thymus and lower levels in immune tissue, liver, skeletal
muscle, adipose tissue and kidney. Of particular interest is the
involvement of RORct in the differentiation and function of
interleukin (IL)-17 producing helper T (TH17) cells. Animal studies
suggest that the modulation of IL-17 levels in autoimmune dis-
eases including multiple sclerosis, psoriasis, inflammatory bowel
disease and rheumatoid arthritis (RA) may be beneficial in terms
of disease progression.6
Although RORc is classed as an orphan receptor since no natural
ligand has been unambiguously identified,7 some structurally
complex natural products derived from cholesterol, sterol ligands,
⇑
Corresponding author. Tel.: +44 1279 645645.
E-mail address: Bodil.vanniel@crl.com (M.B. van Niel).
http://dx.doi.org/10.1016/j.bmcl.2014.10.037
0960-894X/Ó 2014 Elsevier Ltd. All rights reserved.

5770
M. B. van Niel et al. / Bioorg. Med. Chem. Lett. 24 (2014) 5769–5776
cholanoic acid and digoxin are documented to bind with high
affinities.8 The putative endogenous ligands 20
-, 22(R)- and
a
25-hydroxycholesterol (1, 2, 3, Fig. 1) have high affinity for RORc
(EC50 = 20–40 nM)9 and have crystallographically characterised
binding modes similar to known agonists of other NRs.10
A
co-crystal structure with digoxin (4, Fig. 1) demonstrated that
the ligand-induced disruption of the ordered structure of
helices 11, 110 and 12 resulted in an inverse agonist profile
(Kd = 109 nM).11–13 These potent, lipophilic ligands form predomi-
nantly van der Waals contacts with the largely hydrophobic
ligand-binding pocket of RORc. This information gave an early indi-
cation of the future challenges that would be faced in identifying
small molecule ligands for RORc with appropriate physicochemical
properties commensurate with drug development. Whilst 1–4
have served as important tool compounds, these molecules were
not viewed as optimal starting points for further optimisation
due to their high molecular weight, lipophilicity and complex
synthesis required for scaffold modifications.
Figure 2. Examples of small molecule RORc inhibitors.
observe evolving structure-activity relationships (SAR) which have
successfully identified less promiscuous, but low affinity inverse
agonists including SR1001 (6, Fig. 2).17,24 A diverse range of small
molecule chemotypes have been disclosed as the interest in RORc
has intensified. A comprehensive review covering the literature up
to 2014 discusses RORc as a target and has usefully collated all
available information on small molecule RORc ligands from
patents and publications.25 The considerable structural diversity
of reported ligands suggests that RORc functional response
requires few specific ligand–protein interactions. There is a high
prevalence of functional groups such as amides, sulfonamides
and carboxylic acids whilst multiple aryl rings are commonly
encountered in the ligands reported to date. These structurally
diverse ligands, with lower molecular weight but generally high
lipophilicities, continue to be useful tools in delineating the com-
plex biological pathways involved in signalling through RORc.
The Genentech RORc program targeted the identification of
selective, small molecule inverse agonists which could be pro-
gressed in vivo, and a new series of tertiary sulfonamide compounds
derived from a NR compound subset of the Genentech/Roche corpo-
rate compound collection has recently been reported.26 Compounds
7–9 (Table 1) containing a bi-aryl core and H-bond donor/acceptor
functional groups, were demonstrated to be potent RORc inverse
agonists with appreciable selectivity over ROR isoforms and other
NRs. A co-crystal structure of 7 in the RORc binding pocket revealed
that disruption of protein H-bonds seen in agonist structures
resulted in the inverse agonist character of the series (vide infra).
The overall profile of these compounds was sufficiently interesting
to expand SAR exploration and a further publication has described
the outcomes of this work.27
Since the first disclosure of RORc in the 1990s,14 efforts to iden-
tify small molecule modulators to characterise the role of RORc in
inflammatory processes using in vivo models have gained momen-
tum.15 The experimental autoimmune encephalomyelitis model of
multiple sclerosis,12,13,16,17 mouse psoriasis model,18,19 and mouse
collagen-induced arthritis models20 involve TH17 cells and have
been used to probe disease progression using small molecule RORc
modulators. Although clinical trials using human monoclonal anti-
bodies which target IL-17 have shown efficacy in RA, psoriasis and
uveitis,21 the validation of the RORc pathway with small molecules
in clinical trials has only recently started,22 some 20 years after
RORc was discovered. The continued progression and evaluation
of selective, orally bioavailable small molecules in clinical trials
will help to define the role of RORc as a target to modulate IL-17
in human diseases and offers potential to identify new treatments
for debilitating autoimmune diseases in man.
Following the disclosure of the first small molecule RORa/c
inverse agonist T0901317 (5, Fig. 2),23 it has been interesting to
Herein we report our findings aimed at identifying a second
series by transposing the sulfonamide group to generate a parallel
series with the linker atoms reversed.28 Our screening cascade also
used the previously reported biochemical assays.29 A radiometric
binding assay monitored the displacement of [3H]25-hydroxycho-
lesterol from the RORc-LBD. A separate time-resolved fluorescence
biochemical assay monitored the ability of the human RORc-LBD to
bind to a co-activator peptide derived from steroid receptor
co-activator-1 (SRC1).29 In this assay, a compound that disrupted
the recruitment of the SRC1 co-activator peptide to RORc was
termed a RORc inverse agonist.
Early SAR showed that transposing the sulfonamide linker of 9 to
give tertiary phenethylamine analogue 10 (Table 1) retained an
equivalent potency in the competitive binding assay. However, 10
showed a maximum inhibition of only 40% in the SRC1 co-activator
peptide recruitment assay which contrasted with the high potency
and strong RORc inverse agonist efficacy seen with 9 (EC50 = 21 nM,
ꢀ94% efficacy). The methylene spaced analogue (11) which main-
tained an equivalent four linking atoms between the aryl rings
maintained a potent RORc binding (IC50 = 66 nM) and inverse ago-
nist profile (EC50 = 0.32
l
M, ꢀ79% efficacy) compared to 10. It
was anticipated that synthesis of further analogues based on 10
would be more straightforward compared to 11 and we therefore
focused on this option during subsequent SAR campaigns. The
ligand-lipophilicity efficiencies (LLE = pIC50-cLogP)30 for 10 and
Figure 1. Natural product RORc ligands.

M. B. van Niel et al. / Bioorg. Med. Chem. Lett. 24 (2014) 5769–5776
5771
Table 1
Structures of lead sulfonamide 7 and new reversed sulfonamide analogues
a
c
Compd
Structure
RORc IC50
0.13
(
lM)
RORc LLEb
RORc SRC1 EC50 (lM) [%eff.]
7
2.6
0.46 [ꢀ72%]
8
0.033
0.032
0.037
2.9
3.0
2.2
0.040 [ꢀ90%]
0.021 [ꢀ94%]
1.0 [ꢀ40%]
9
10
11
0.066
2.7
0.32 [ꢀ79%]
12
0.16
1.9
>20 [ꢀ20%]
See Ref. 29 Supplementary data for experimental details of all assays. All assay results are reported as the geometric mean of at least two separate runs.
a
Biochemical inhibition of the RORc LBD and [3H]25-hydroxycholesterol interaction.
b
Lipophilic ligand efficiency (LLE) was calculated using the RORc biochemical IC50 value and calculated logP (cLogP).26
c
Inhibition of RORc LBD recruitment of the SRC1 co-activator peptide; negative percent efficacy ‘%eff.’ denotes inverse agonism relative to the basal activity of apo-RORc
M in this assay except for compound 12 which was tested up to 20 M.
LBD in this assay format. The maximum concentration tested was 10
l
l
11 were lower when compared to 9. Whilst LLE was an important
consideration, this metric is reported using data from the biochem-
ical binding assay and we also had to consider the wider implica-
tions of the data from the SRC1 co-activator peptide recruitment
assay on the overall profile when deciding which leads should be
further investigated. A truncated tertiary benzylamine analogue
12 had reasonable binding potency (IC50 = 160 nM), but it was
nearly a silent ligand in the SRC1 co-activator peptide recruitment
that would require significant improvement if the reversed sulfon-
amides were to become a viable alternative series.
From this preliminary SAR, it was clear that optimisation of the
profile in the SRC1 co-activator peptide recruitment assay together
with improvements of physicochemical properties were required
to advance the series. To progress the new series at a fast pace
we split exploration into three areas. Interactions on the left-hand
side (LHS) were explored using a small selection of polar groups to
conserve potential H-bonding interactions with the RORc protein.
In addition, we noted that changes to the bi-aryl core which occu-
pied the central portion of the binding site should in principle be
tolerated due to the available space and the flexibility of the ali-
phatic side-chains lining the pocket. Finally, the right-hand side
(RHS) of the ligand was the focus of modifications aimed at
increasing the level of the inverse agonism. Due to the large num-
ber of aromatic residues on the RHS of the pocket and the potential
assay, showing a maximum inhibition of only 20% at 20 lM, the
highest test concentration. This co-activator recruitment result
suggested that the relative location of the sulfonamide linker could
modulate inverse agonism. Additional SAR extensively explored
substitution on the aryl rings of the phenethylamine (10) and ben-
zylamine (12) reversed sulfonamides in an attempt to disrupt the
intact C-terminus, as had been observed in the co-crystal of 7.26
Although strong binding affinity in the human RORc-LBD could be
achieved by other analogues (data not shown), they were still only
partial inverse agonists or silent ligands in the SRC1 co-activator
peptide recruitment assay. These data continued to highlight areas
for
p–p interactions, we decided to maintain an aromatic ring in
that region and concentrated our efforts on modifying the linker
length and investigating substitution around the ring.

5772
M. B. van Niel et al. / Bioorg. Med. Chem. Lett. 24 (2014) 5769–5776
The lipophilicity of reversed sulfonamide leads 10–12 was high
combined with the tertiary benzylamine (20, IC50 = 93 nM; 21,
IC50 = 51 nM). The combination of fluorine substitution with a
longer tertiary phenethylamine resulted in equipotent binding
affinities for 22 and 23 (IC50 = 41 nM). However, there was again
a clear preference for higher potencies in the SRC1 co-activator
peptide recruitment assay for the longer side chains of 22 and 23
as had previously been observed when comparing 10 with 12.
Identification of 22 demonstrated for the first time that potent
binding and a high affinity and high efficacy inverse agonist profile
could be achieved in this reversed sulfonamide series. We also
replaced the central aryl ring with pyridine isomers (24–27) as a
way to moderate the overall lipophilicity of the chemical series.
Although pyridines 24–27 successfully lowered the cLogP by
0.5–0.8 units compared to 18 and 19, these heterocycles were also
less potent in both biochemical assays. A range of other six-
membered ring heterocycles including pyrimidine, pyridazine
and pyrazine showed large reductions in binding potencies (data
not shown) and these observations are in line with our previously
reported data in the parent sulfonamide series.27
We next targeted aryl ring substitution of the tertiary benzyl-
amine (18) and phenethylamine (19) since these bind in the region
of the RORc-LBD likely to cause disordering of the H-bond between
H479 and Y502.11 A scan of chloro substitution showed a modest
preference for ortho substitution (28) in the RORc biochemical
binding assay. Whilst meta (29) and para-chloro benzylamines
(30) were equipotent in the binding assay, slightly stronger inverse
agonist potency was obtained for 30. A large body of follow-up
work showed ortho-trifluoromethylbenzyl amine (31) achieved
high binding potency and considerably higher efficacy than had
been observed with the corresponding ortho-chloro benzylamine
analogue (28). Obtrusion into this region of the binding site by
31 presumably resulted in disruption of helices 110–12, which
(cLogP26 4.5–5.2), and we designed compounds such that parame-
ters were in a range considered to be desirable property space for
oral drugs (cLogP <4 was preferred).31–33 The aqueous solubility of
the early reversed sulfonamide leads 10–12 was low (<1 lM at pH
7.4) and there is mounting literature evidence that multiple aryl
rings can be detrimental to this property.34–36 Encouraged by the
reported SAR in the parent sulfonamide series,27 replacing the aryl
ring of 12 located in the hydrophilic region of the binding site with
non-aromatic heterocycles was investigated (Table 2). Morpholine
(13) and 4-hydroxypiperidine (14) analogues were substantially
less potent in the binding assay, indicating that these H-bond
acceptor/donor groups were not optimal at the terminal region of
the molecules. The basic N-methylpiperazine (15) showed a fur-
ther significant reduction in binding potency. Modest affinity for
RORc could be regained by introduction of N-acetyl- (16), N-sulfa-
moyl– (17) and N-methanesulfonyl– (18) substituents on the
piperazine to reduce basicity and these analogues had reduced
lipophilicity (cLogP 2.6–3.9). We noted that 17 and 18 had
improved LLE values (4.2 and 3.1, respectively), but disappoint-
ingly all N-linked heterocyclic ring analogues (13–18) had less than
50% inhibition in the SRC1 peptide co-activator recruitment assay
at the highest test concentration. The homologue (19) had marked
potency improvements in RORc binding (IC50 = 51 nM) when
compared to 18 (IC50 = 0.37 lM).
Subsequent SAR was generated using the N-methanesulfonyl-
piperazine as a ligand-anchor since this N-linked heterocyclic ring
was straightforward to introduce, chemically benign during
synthesis and the overall lipophilicity of 18 was reasonable
(cLogP = 3.3). We next investigated modifications of the central
phenyl ring. Fluorine ortho or meta to the sulfonamide linker of
18 resulted in a large improvement in binding potencies when
Table 2
Piperidine and piperazine analogues
R1-group
X
Y
R2-group
RORc IC50
(lM)
RORc LLEb
RORc SRC1 EC50 (lM) [%eff.]
cLogPb
a
c
Compd
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
O
CAH
CAH
CAH
CAH
CAH
CAH
CAH
CAF
CAH
CAF
CAH
N
CAH
CAH
CAH
CAH
CAH
CAH
CAH
CAH
CAF
CAH
CAF
CAH
CAH
N
PhCH2–
PhCH2–
PhCH2–
PhCH2–
PhCH2–
PhCH2–
PhCH2CH2–
PhCH2–
PhCH2–
PhCH2CH2–
PhCH2CH2–
PhCH2–
PhCH2CH2–
PhCH2–
PhCH2CH2–
(2-Cl) PhCH2–
(3-Cl) PhCH2–
(4-Cl) PhCH2–
(2-CF3) PhCH2–
(2-CF3) PhCH2–
0.51
0.44
5.6
0.42
0.17
0.37
0.051
0.093
0.051
0.042
0.041
0.43
0.19
0.34
0.16
2.1
1.8
0.85
2.5
4.2
3.1
3.6
3.4
3.9
3.5
3.6
3.6
3.5
3.9
3.9
3.8
3.1
3.1
3.3
3.2
>6.7 [ꢀ12%]
>20 [ꢀ29%]
>20 [ꢀ16%]
>10 [ꢀ29%]
>10 [ꢀ3%]
4.2
4.6
3.8
3.9
2.6
3.3
3.7
3.6
3.4
3.9
3.8
2.8
3.2
2.6
2.9
3.9
4.1
4.1
4.2
4.8
CHAOH
NACH3
NACOCH3
NASO2N(CH3)2
NASO2CH3
NASO2CH3
NASO2CH3
NASO2CH3
NASO2CH3
NASO2CH3
NASO2CH3
NASO2CH3
NASO2CH3
NASO2CH3
NASO2CH3
NASO2CH3
NASO2CH3
NASO2CH3
NACOCH3
0.53 [ꢀ29%]
0.34 [ꢀ55%]
>10 [ꢀ42%]
>10 [ꢀ20%]
0.05 [ꢀ85%]
0.15 [ꢀ48%]
>10 [ꢀ36%]
4.5 [ꢀ55%]
>10 [ꢀ24%]
0.34 [ꢀ38%]
>10 [ꢀ35%]
>10 [ꢀ13%]
1.3 [ꢀ72%]
0.10 [ꢀ63%]
0.085 [ꢀ81%]
N
CAH
CAH
CAH
CAH
CAH
CAH
CAH
N
CAH
CAH
CAH
CAH
CAH
0.021
0.070
0.061
0.034
0.0092
See Ref. 29 Supplementary data for experimental details of all assays. All assay results are reported as the geometric mean of at least two separate runs.
a
Biochemical inhibition of the RORc LBD and [3H]25-hydroxycholesterol interaction.
b
c
26
Lipophilic ligand efficiency (LLE) was calculated using the RORc biochemical IC50 value and calculated logP (cLogP)
Inhibition of RORc LBD recruitment of the SRC1 co-activator peptide; negative percent efficacy ‘%eff.’ denotes inverse agonism relative to the basal activity of apo-RORc
M in this assay except for compound 19 which was tested up to 200 M.
.
LBD in this assay format. The maximum concentration tested was 10
l
l

M. B. van Niel et al. / Bioorg. Med. Chem. Lett. 24 (2014) 5769–5776
5773
contrasts with substituents like para-chloro benzylamine (30) that
may adopt an appropriate vector for the para substituent to cause a
similar disruption. At this stage of our SAR, we re-evaluated the
N-acetyl piperazine (32) which confirmed our earlier observations
with 16 and 17 that a number of H-bond acceptors are able to
maintain equivalent profiles in the biochemical assays.
The overall lipophilicity of these new analogues containing a
spacer group was still high (cLogP >4) and our aim was to
re-optimise substitution of the terminal aryl ring to manipulate
the overall physicochemical properties of the series. The ether lin-
ker was conserved during this work since it was facile to introduce
4-hydroxy-N-methanesulfonyl piperidine as the last step through
simple displacement of a para-fluoro substituent on the appropri-
The identification of reversed aryl sulfonamides with greater
inverse agonism efficacy encouraged us to investigate the intro-
duction of linker atoms at alternative locations in the scaffold to
act as a spacer between the ligand–anchoring interactions of the
N-methanesulfonyl piperazine group and the LHS terminal aryl
ring (Table 3). This strategy could be considered complementary
to our initial efforts to use phenethylamine compared to benzyl-
amine. We modified the N-methanesulfonyl piperazine (31) to
the corresponding 4-amino-N-methanesulfonyl piperidine (33).
This strategy also relaxed some conformational rigidity which
may have harmonised the fit of reversed sulfonamides in the
binding site. Identification of 33 (IC50 = 6 nM; EC50 12 nM, ꢀ91%
efficacy) resulted in the first example from the new reversed sul-
fonamide series with single digit nanomolar binding affinity while
also displaying potent and high efficacy RORc inverse agonism. The
homologue (34, IC50 = 33 nM; EC50 = 28 nM, ꢀ91% efficacy) gave a
similar overall profile in the biochemical assays as the correspond-
ing piperazine (31, IC50 = 34 nM; EC50 = 100 nM, ꢀ63% efficacy).
Opportunities to introduce polar functionality had been limited
in the reversed sulfonamide series and additional linker groups
were investigated starting with the ether (35) which had a six-fold
improvement in binding potency when compared to 31. The syn-
thetically versatile carbonyl linker (36) was also prepared and
reduced to the racemic secondary alcohol (37) and eliminated to
the alkene (38) to give analogues which retained a high binding
potency and similar overall profile to 35. This suggested that this
region of the scaffold could usefully be exploited to introduce polar
functionality which may be able to interact with polar residues in
the RORc binding pocket or interact with the water architecture in
the RORc-LBD.
ate benzene sulfonamide fragment.28
A scan of ortho-(39),
meta-(40) and para-fluorine (41) substituents on the terminal aryl
ring confirmed that the meta- and para- substitution were most
potent. A nitrile (42–44) substitution scan showed that analogues
which retained high binding potencies and strong inverse agonist
profiles with reduced lipophilicity (cLogP 3.8) could be identified
through modifications of the terminal aryl ring.
Replacement of the terminal aryl ring with polar heterocycles
was also investigated as part of our ongoing strategy to reduce
lipophilicity of the ether linked series. However, pyridines 45
and 46 gave disappointing potency in both biochemical assays
and this region of the RORc-LBD appears to be unable to tolerate
polar functionality. We also replaced the N-isobutyl group and
investigated a small set of alternative alkyl groups. Although
reasonable binding potency could be maintained, disappointingly,
alternative substituents including N-tert-butyl, N-methylcyclobu-
tyl, N-cyclobutyl, N-trifluoroethyl and N-isopentyl amines all
had low inverse agonist potency (data not shown) and we did
not further investigate other substituents with more polar overall
properties. The data confirmed that there was an optimal steric
contribution of the N-isobutyl group in maintaining the overall
biochemical profile of the reversed sulfonamide series. Our
observations contrast with the parent sulfonamide series where
a
selection of small aliphatic or halogenated groups could
improve RORc inverse agonist potency, reduce lipophilicity and
increase aqueous solubility.27
Finally, we investigated a combination of all beneficial changes
and a number of analogues were prepared including the 4-amino
or 4-hydroxy N-methanesulfonyl piperidine scaffold and a central
Table 3
Piperidine analogues
a
c
Compd
A-B
X
Y
R1-group
RORc IC50
(l
M)
RORc LLEb
RORc SRC1 EC50
(lM) [%eff.]
cLogPb
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
CHANH
CHANH
CHAO
CHAC(O)
CHACH(OH)
C@CH
CHAO
CHAO
CHAO
CHAO
CHAO
CHAO
CHAO
CHAO
CHAO
CHANH
CAH
CAH
CAH
CAH
CAH
CAH
CAH
CAH
CAH
CAH
CAH
CAH
CAH
CAH
N
CAH
CAH
CAH
CAH
CAH
CAH
CAH
CAH
CAH
CAH
CAH
CAH
CAH
CAH
CAH
N
(2-CF3) PhCH2–
(2-CF3) Ph–
(2-CF3) PhCH2–
(2-CF3) PhCH2–
(2-CF3) PhCH2–
(2-CF3) PhCH2–
(2-F) Ph–
(3-F) Ph–
(4-F) Ph–
(2-CN) Ph–
(3-CN) Ph–
(4-CN) Ph–
2-Pyridyl–
3-Pyridyl–
(2-CF3) PhCH2–
(4-F) PhCH2–
0.006
0.033
0.006
0.007
0.011
0.014
0.056
0.008
0.016
0.021
0.014
0.015
0.17
3.3
2.5
3.4
3.6
3.9
2.2
3.1
3.8
3.5
3.9
4.0
4.0
3.6
3.8
3.6
3.2
0.012 [ꢀ91%]
0.028 [ꢀ68%]
0.022 [ꢀ90%]
0.056 [ꢀ82%]
0.022 [ꢀ88%]
0.075 [ꢀ84%]
0.06 [ꢀ79%]
0.021 [ꢀ91%]
0.022 [ꢀ97%]
0.068 [ꢀ52%]
0.059 [ꢀ85%]
0.027 [ꢀ96%]
0.24 [ꢀ63%]
0.49 [ꢀ75%]
0.015 [ꢀ89%]
0.52 [ꢀ64%]
4.9
5.0
4.8
4.5
4.0
5.6
4.1
4.3
4.3
3.8
3.8
3.8
3.2
3.0
4.3
3.6
0.18
0.012
0.18
CAH
See Ref. 29 Supplementary data for experimental details of all assays. All assay results are reported as the geometric mean of at least two separate runs.
a
Biochemical inhibition of the RORc LBD and [3H]25-hydroxycholesterol interaction.
b
Lipophilic ligand efficiency (LLE) was calculated using the RORc biochemical IC50 value and calculated logP (cLogP).26
c
Inhibition of RORc LBD recruitment of the SRC1 co-activator peptide; negative percent efficacy ‘%eff.’ denotes inverse agonism relative to the basal activity of apo-RORc
M in this assay.
LBD in this assay format. The maximum concentration tested was 10
l

5774
M. B. van Niel et al. / Bioorg. Med. Chem. Lett. 24 (2014) 5769–5776
pyridine ring in combination with a range of tertiary sulfonamide
substituents. One highlight from this work was that introduction
of pyridine into the scaffold of 35 to give 47 resulted in a similar
overall profile with a reduction of lipophilicity (cLogP 4.8 to 4.3).
Although only low inverse agonism (ꢀ64% efficacy) was obtained
with a second pyridine analogue (48), a further highlight was a
1.7 Å resolution co-crystal structure of this compound at the RORc
binding site (Fig. 3).37 In this structure, the compound extended
nearly the entire length of the binding site from the hydrophobic
region near helices 11 and 12 to the hydrophilic opposite end.
Van der Waals interactions dominated the interaction between
RORc and 48 (Fig. S1), but two hydrogen bonds linked the terminal
methyl sulfone to the protein (side chain of R367 and main chain of
L287). Several water molecules bordered the compound and filled
the breadth of the pocket; however, only a single solvent H-bond
was observed, linking the pyridyl nitrogen to a portion of this sol-
vent shell. A second water molecule resided 3.1 Å from the central
amine but was out of geometric alignment for the interaction to be
significant. Interestingly, the amino piperidine adopted an atypical
axial geometry37 (A-value = 1.9 kcal/mol) that caused the com-
pound to follow a serpentine path through the binding site, in stark
contrast to most other ligands which adopted an ‘L’ shape.25 This
conformation was definitively supported by the unbiased electron
density (Fig. S2).
substituent positioned such that it would clash with the agonist
mode position of W317, preventing the protein from adopting an
agonist conformation of the C-terminus of the protein. Although
compounds 7 and 48 differed by a single heavy atom in length
(including the fluorine in 48), the paths that the two molecules tra-
versed through the binding site diverged and gave rise to a 1.0 Å
offset of sulfur atom positions at the hydrophilic end and a 2.7 Å
separation of equivalent positions of the phenyl rings at the lipo-
philic end of the pocket. In addition, the phenyl rings were oriented
offset from each other by approximately 23°, as measured from the
most apical position. The reversing of the sulfonamide connectivity
and removal of one bridging atom in the linker moving from 7 to
48 significantly altered the positioning of the hydrophobic region
of the ligands (Fig. S3), explaining the separation in SAR for the
two series.37 Additional X-ray data for analogues which had
improved binding potency and inverse agonist efficacy may offer
insight into the interactions which successfully disrupt the C-ter-
minus. Whilst this was highly desirable and actively pursued, suc-
cessful co-crystallisation of additional reversed sulfonamide
examples has remained elusive.
Next we profiled several of the more potent compounds into
HEK293 cell based assays using Gal4-NR constructs to assess selec-
tivity profiles in the three isoforms of ROR (Table 4). In addition, we
also investigated the selectivity profiles for these analogues against
farnesoid X receptor (FXR), pregnane X receptor (PXR) and liver X
At the hydrophobic end of the pocket, the para-fluoro-phenyl
substituent packed within the lipophilic pocket defined by residues
C320, L324, W317, I400, and L396. In the RORc crystal structure, 48
did not perturb the hydrogen bond between H479 and Y502 that
anchored helix 12 in an agonist conformation. As a result, this crys-
tal was obtained in the presence of co-activator peptide and repre-
sented the ‘agonist mode’ of the protein conformation, consistent
receptor alpha (LXRa) and liver X receptor beta (LXRb) using the
same cell assay system.29 The parent sulfonamide 9 was a weak
inhibitor in the RORc Gal4 cellular assay (EC50 = 453 nM) and was
only ꢁ five-fold selective over the RORa and RORb isoforms and
at least 21-fold selective over FXR, LXRa, LXRb and PXR. Pleasingly,
progression of 33, 34, 41, 44 and 47 through the Gal4 cellular
assays clearly highlighted that excellent RORc potency and selec-
tivity over the RORa and RORb isoforms could be obtained with
with 48 only showing low efficacy (EC50 = 0.52
l
M, ꢀ64% efficacy)
to displace the SRC1 peptide in the co-activator recruitment assay.
In contrast, the previously reported structure of 7 from the sulfon-
amide series bound to RORc (PDB 4QM0) had its terminal phenyl
these reversed sulfonamides. Weak activity at FXR, LXR
a, LXRb
and PXR was observed, which together with the improved RORc
Figure 3. Co-crystal structure of 48 (gold) in complex with RORc in an agonist conformation (violet) and co-activator peptide (cyan).38 Residues within van der Waals’ contact
of the inhibitor are shown (except where removed for clarity). Amino acids are labelled with single letter code, and key protein helices are annotated numerically in yellow
font.

M. B. van Niel et al. / Bioorg. Med. Chem. Lett. 24 (2014) 5769–5776
5775
Table 4
RORc potency and selectivity profiles in Gal4 human transcription and human IL-17 and cytokine biosynthesis assays
Compd RORc Cell
RORa Cell
RORb Cell
FXR Cell
LXR
a
Cell
LXRb Cell
PXR Cell
IL-17A hPBMC
EC50 (lM)
IL-17A
hPBMCb
%max.
IFN
EC50
M)
c
CTG
EC50
(lM)
a
a
a
a
a
a
a
b
EC50
(
lM)
EC50
(lM)
EC50
(
lM)
EC50
(l
M) EC50
(
lM)
EC50
(lM)
EC50
(
lM)
(
l
9
0.453
0.006
0.007
0.019
0.032
0.039
2.4
4.4
>10
>10
8.8
2.5
3.8
>10
>10
>8
>10
>10
>10
>10
>10
>10
>10
>10
>10
>10
>10
>10
>10
>10
>10
>10
>10
>10
>10
>10
>10
>10
>10
>10
n.d.
n.d.
73
67
67
77
68
n.d.
>20
>20
>20
>20
>18
n.d.
>20
>20
>20
>20
>20
33
34
41
44
47
0.27
0.044
0.42
0.76
0.33
>7.1
>3.1
See Ref. 29 Supplementary data for experimental details of all assays. All assay results are reported as the geometric mean of at least two separate runs.
a
All assays were conducted in HEK293 cells using Gal4-NR constructs. All ROR and NR assays monitored the suppression of their respective basal transcriptional activities,
an outcome consistent with inverse agonist activity of ligands with these receptors.
b
All assays were conducted using human peripheral blood mononuclear cells (hPBMCs) isolated from human whole blood. Interferon gamma (INFc) biosynthesis and
CellTiter-GloÒ (CTG) cellular ATP measurement were used as controls to monitor for off-target cytokine activity and aberrant cytotoxicity, respectively. n.d. = not determined.
cellular potency, now gave a significantly improved selectivity
window over the NRs.
maintaining an agonist conformation. This contrasts with reported
inverse agonist ligands, including 7, that successfully disrupt this
key H-bond. Reversed sulfonamides, including 34, continue to serve
as useful in vitro tools whilst our efforts to identify series which can
be progressed to in vivo studies continue. The information from our
SAR and co-crystal structure with the RORc-LBD has been applied to
other related scaffolds and results from these efforts will be dis-
closed in future publications.
Finally, these potent and selective RORc inverse agonists were
progressed into a cytokine biosynthesis assay that monitored their
ability to inhibit IL-17A production in anti-CD3/CD28-stimulated
human peripheral blood mononuclear cells (PBMCs).29 Interferon
gamma (INFc
) biosynthesis and CellTiter-GloÒ (CTG) cellular ATP
measurement were also assessed to monitor for off-target cytokine
activity and aberrant cytotoxicity, respectively. Compound 44 dis-
played modest inhibition of IL-17 production in the human PBMC
Acknowledgments
assay (EC50 = 0.76
lM) whilst 33, 41 and 47 were slightly more
potent (EC50 = 0.27, 0.42 and 0.33
particular highlight was the identification of 34 as the most potent
inhibitor of IL-17 biosynthesis (EC50 = 44 nM) which also did not
l
M, respectively) (Table 4). A
We thank Russell Scammell, Neil Bayliss and Michael Podmore
for analytical LCMS and NMR data and Steve McLachlan, Sarah G.
Hymowitz and Robert Hall for their important contributions to
the project team. We also thank the Genentech Structural Biology
Expression group for cloning and expression experiments and staff
at beamline 5.0.2 for their assistance with data collection setup.
The Berkeley Center for Structural Biology is supported in part by
the National Institutes of Health, National Institute of General
Medical Sciences, and the Howard Hughes Medical Institute. The
Advanced Light Source is supported by the Director, Office of
Science, Office of Basic Energy Sciences, of the U.S. Department of
Energy under Contract No. DE-AC02-05CH11231.
show inhibition of INFc or CTG assays, demonstrating that this
compound was not broadly inhibiting cytokine production and
was not overtly cytotoxic. A six-fold improvement in potency
was seen for 34 in the human PBMC assay when compared to
homologue 33. This observation contrasts with their equivalent
potencies in the HEK293 cell based assay but was not un-expected
since different end points are measured in this primary cell based
assay compared to the native human PBMCs.29 The reversed sul-
fonamide 34 had excellent selectivity for RORc over RORa and
RORb (>1000-fold) and the NRs in our cell assays panel (>1000-
fold), together with high potency in the IL-17 biosynthesis assay,
which represented a useful profile for further progression. When
incubated in human and rat liver microsomes, 34 displayed mod-
erate hepatic clearance (Clhep) values in both species (14 and
21 mL/min/kg, respectively) and the compound was progressed
to rat pharmacokinetic studies.38 Disappointingly, 34 displayed
clearance at greater than rat liver blood flow in vivo and thus the
utility of this reversed sulfonamide was limited to in vitro studies.
Small molecule RORc inverse agonists, with profiles suitable for
progression to in vitro and in vivo screens, continue to be of consid-
erable interest to help delineate the complex role of RORc in normal
biology and autoimmune disease. We have described the SAR of a
new series which incorporated a reversed tertiary sulfonamide lin-
ker. Potent RORc inverse agonists with high cellular potencies and
excellent selectivity over ROR isoforms and other NRs have been
identified. The hydrophobic nature of the RORc binding pocket
has made it challenging to incorporate polar functionality and
reduce lipophilicity of the reversed sulfonamide series. Identifying
strategies which successfully lower lipophilicity may provide com-
pounds with reduced hepatic clearance and improved solubility
which could be progressed to in vivo studies. A co-crystal structure
of 48 with the RORc-LBD supported previous observations that
silent ligands, agonists, and partial inverse agonists do not disrupt
a key H479-Y502 H-bond which results in the RORc protein
Supplementary data
Supplementary data associated with this article can be found,
in the online version, at http://dx.doi.org/10.1016/j.bmcl.2014.
10.037.
References and notes
1. Shi, Y. Drug Discovery Today 2007, 12, 440.
2. Wurtz, J. M.; Bourguet, W.; Renaud, J. P.; Vivat, V.; Chambon, P.; Moras, D.;
Gronemeyer, H. Nat. Struct. Biol. 1996, 3, 87.
3. Isono, F.; Fujita-Sato, S.; Ito, S. Drug Discovery Today 2014, 19, 1205.
4. Solt, L. A.; Burris, T. P. Trends Endocrinol. Metab. 2012, 23, 619.
5. Jetten, A. M.; Joo, J. H. Adv. Dev. Biol. 2006, 16, 313.
6. Ouyang, W.; Kolls, J. K.; Zheng, Y. Immunity 2008, 28, 454.
7. Ivanov, I. I.; McKenzie, B. S.; Zhou, L.; Tadokoro, C. E.; Lepelley, A.; Lafaille, J. J.;
Cua, D. J.; Littman, D. R. Cell 2006, 126, 1121.
8. Wang, Y.; Kumar, N.; Solt, L. A.; Richardson, T. I.; Helvering, L. M.; Crumbley, C.;
Garcia-Ordonez, R. D.; Stayrook, K. R.; Zhang, X.; Novick, S.; Chalmers, M. J.;
Griffin, P. R.; Burris, T. P. J. Biol. Chem. 2010, 285, 5013.
9. Jin, L.; Martynowski, D.; Zheng, S.; Wada, T.; Xie, W.; Li, Y. Mol. Endocrinol. 2010,
24, 923.
10. Ingraham, H. A.; Redinbo, M. R. Curr. Opin. Struct. Biol. 2005, 15, 708.
11. Fujita-Sato, S.; Ito, S.; Isobe, T.; Ohyama, T.; Wakabayashi, K.; Morishita, K.;
Ando, O.; Isono, F. J. Biol. Chem. 2011, 286, 31409.
12. Huh, J. R.; Leung, M. W.; Huang, P.; Ryan, D. A.; Krout, M. R.; Malapaka, R. R.;
Chow, J.; Manel, N.; Ciofani, M.; Kim, S. V.; Cuesta, A.; Santori, F.; Lafaille, J.; Xu,
H.; Gin, D.; Rastinejad, F.; Littman, D. Nature 2011, 472, 486.

5776
M. B. van Niel et al. / Bioorg. Med. Chem. Lett. 24 (2014) 5769–5776
13. Xu, T.; Wang, X.; Zhong, B.; Nurieva, R.; Ding, S.; Dong, C. J. Biol. Chem. 2011,
286, 22707.
14. Hirose, T.; Smith, R. J.; Jetten, A. M. Biochem. Biophys. Res. Commun. 1976, 1994,
205.
27. Fauber, B. P.; René, O.; Boenig, G.; Burton, B.; Deng, Y.; Eidenschenk, C.; Everett,
C.; Gobbi, A.; Hymowitz, S. G.; Johnson, A. R.; La, H.; Liimatta, M.; Lockey, P.;
Norman, M.; Ouyang, W.; Wang, W.; Wong, H. Bioorg. Med. Chem. Lett. 2014, 24,
3891.
15. Marciano, D. P.; Chang, M. R.; Corzo, C. A.; Goswani, D.; Lam, V. Q.; Pascal, B. D.;
Griffin, P. R. Cell Metab. 2014, 19, 193.
28. See the Supplementary data for representative synthetic schemes. For
representative protocols and characterisation of the analogues, see (a) van
Niel, M. B.; Fauber, B. P.; René, O.; Ward, S. WO 2014 090710; (b) van Niel, M.
B.; Fauber, B. P.; Gaines, S.; Killen, J.; René, O.; Ward, S. WO 2014 090712.
29. Fauber, B. P.; Boenig, G.; Burton, B.; Eidenschenk, C.; Everett, C.; Gobbi, A.;
Hymowitz, S. G.; Johnson, A. R.; Liimatta, M.; Lockey, P.; Norman, M.; Ouyang,
W.; René, O.; Wong, H. Bioorg. Med. Chem. Lett. 2013, 23, 6604.
30. Leach, A. R.; Hann, M. M.; Burrows, J. N.; Griffen, E. J. Mol. Biosyst. 2006, 2, 429.
31. Leeson, P. D.; Springthorpe, B. Nat. Rev. Drug Disc. 2007, 6, 881.
32. Waring, M. J. Expert Opin. Drug Discov. 2010, 5, 235.
16. Wang, Y.; Cai, W.; Zhang, G.; Yang, T.; Liu, Q.; Cheng, Y.; Zhou, L.; Ma, Y.; Cheng,
Z.; Lu, S.; Zhao, Y.-G.; Zhang, W.; Xiang, Z.; Wang, S.; Yang, L.; Wu, Q.; Orband-
Miller, L. A.; Xu, Y.; Zhang, J.; Gao, R.; Huxdorf, M.; Xiang, J.-N.; Zhong, Z.;
Elliott, J. D.; Leung, S.; Lin, X. Bioorg. Med. Chem. 2014, 22, 692.
17. 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.; Schürer, S. C.; Xu, J.; Wagoner, G.;
Drew, P. D.; Griffin, P. R.; Burris, T. P. Nature 2011, 472, 491.
18. Skepner, J.; Ramesh, R.; Trocha, M.; Schmidt, D.; Baloglu, E.; Lobera, M.;
Carlson, T.; Hill, J.; Orband-Miller, L. A.; Barnes, A.; Boudjelal, M.; Sundrud, M.;
Ghosh, S.; Yang, J. J. Immunol. 2014, 192, 2564.
19. Pantelyushin, S.; Haak, S.; Ingold, B.; Kulig, P.; Heppner, F. L.; Navarini, A. A.;
Becher, B. J. Clin. Invest. 2012, 122, 2252.
33. Hann, M. M. Med. Chem. Commun. 2011, 2, 349.
34. Ritchie, T. J.; Macdonald, S. J. F. Drug Discovery Today 2009, 14, 1011.
35. Ritchie, T. J.; Macdonald, S. J. F.; Young, R. J.; Pickett, S. D. Drug Discovery Today
2011, 16, 164.
20. Chang, M. R.; Lyda, B.; Kamenecka, T. M.; Griffin, P. R. Arthritis Rheum. 2014, 66,
579.
21. Patel, D. D.; Lee, D. M.; Kolbinger, F.; Antoni, C. Ann. Rheum. Dis. 2013, 72, 116.
22. http://www.jt.com/investors/results/S_information/pharmaceuticals/pdf/P.L.
20140424_E.pdf (article accessed 23 June 2014).
23. Houck, K. A.; Borchert, K. M.; Hepler, C. D.; Thomas, J. S.; Bramlett, K. S.;
Michael, L. F.; Burris, T. P. Mol. Gen. Metab. 2004, 83, 184.
24. Kojetin, D. J.; Burris, T. P. Nat. Rev. Drug Disc. 2014, 13, 197.
25. Fauber, B. P.; Magnuson, S. J. Med. Chem. 2014, 57, 5871.
26. Fauber, B. P.; René, O.; Burton, B.; Everett, C.; Gobbi, A.; Hawkins, J.; Johnson, A.
R.; Liimatta, M.; Lockey, P.; Norman, M.; Wong, H. Bioorg. Med. Chem. Lett. 2014,
24, 2182.
36. Ishikawa, M.; Hashimoto, Y. J. Med. Chem. 2011, 54, 1539.
37. See the Supplementary data for the experimental details associated
with the co-crystal structure of the RORc-LBD with compound 48. The
coordinates and structure factors for the complex between RORc and
compound 48 have been deposited with the Protein Data Bank and assigned
the accession code 4WLB. A binding site interaction map of 7 and 48 with RORc
is also shown.
38. See the Supplementary data for the experimental details of the
pharmacokinetic study.