Receptor-interacting protein kinase 2 (RIPK2) and nucleotide-binding oligomerization domain (NOD) cell signaling inhibitors based on a 3,5-diphenyl-2-aminopyridine scaffold

Article abstract of DOI:10.1016/j.ejmech.2020.112417

Receptor-interacting protein kinase 2 (RIPK2) is a key mediator of nucleotide-binding oligomerization domain (NOD) cell signaling that has been implicated in various chronic inflammatory conditions. A new class of RIPK2 kinase/NOD signaling inhibitors bas

Full text of DOI:10.1016/j.ejmech.2020.112417

Journal Pre-proof
Receptor-interacting protein kinase 2 (RIPK2) and nucleotide-binding oligomerization
domain (NOD) cell signaling inhibitors based on a 3,5-diphenyl-2-aminopyridine
scaffold
Chalada Suebsuwong, Bing Dai, Daniel M. Pinkas, Anantha Lakshmi Duddupudi, Li
Li, Joshua C. Bufton, Lisa Schlicher, Mads Gyrd-Hansen, Ming Hu, Alex N. Bullock,
Alexei Degterev, Gregory D. Cuny
PII:
S0223-5234(20)30388-3
DOI:
https://doi.org/10.1016/j.ejmech.2020.112417
EJMECH 112417
Reference:
To appear in: European Journal of Medicinal Chemistry
Received Date: 31 March 2020
Revised Date: 27 April 2020
Accepted Date: 30 April 2020
Please cite this article as: C. Suebsuwong, B. Dai, D.M. Pinkas, A.L. Duddupudi, L. Li, J.C. Bufton, L.
Schlicher, M. Gyrd-Hansen, M. Hu, A.N. Bullock, A. Degterev, G.D. Cuny, Receptor-interacting protein
kinase 2 (RIPK2) and nucleotide-binding oligomerization domain (NOD) cell signaling inhibitors based
on a 3,5-diphenyl-2-aminopyridine scaffold, European Journal of Medicinal Chemistry (2020), doi:
https://doi.org/10.1016/j.ejmech.2020.112417.
This is a PDF file of an article that has undergone enhancements after acceptance, such as the addition
of a cover page and metadata, and formatting for readability, but it is not yet the definitive version of
record. This version will undergo additional copyediting, typesetting and review before it is published
in its final form, but we are providing this version to give early visibility of the article. Please note that,
during the production process, errors may be discovered which could affect the content, and all legal
disclaimers that apply to the journal pertain.
©
2020 Published by Elsevier Masson SAS.

Receptor-Interacting Protein Kinase 2 (RIPK2) and
Nucleotide-Binding Oligomerization Domain
(NOD) Cell Signaling Inhibitors Based on a 3,5-
Diphenyl-2-aminopyridine Scaffold
#
,§
ǁ
†
∞
∞
Chalada Suebsuwong , Bing Dai , Daniel M. Pinkas , Anantha Lakshmi Duddupudi , Li Li ,
†
,¥
^
^
∞
†
Joshua C. Bufton , Lisa Schlicher , Mads Gyrd-Hansen , Ming Hu , Alex N. Bullock , Alexei
ǁ,*
∞,**
Degterev , Gregory D. Cuny
#
∞
Department of Chemistry and Department of Pharmacological and Pharmaceutical Sciences,
University of Houston, Health Building 2, Houston, Texas 77204, USA
Structural Genomics Consortium, University of Oxford, Old Road Campus, Roosevelt Drive,
Oxford OX3 7DQ, UK
†
ǁ
Department of Developmental, Molecular & Chemical Biology, Tufts University School of
Medicine, 136 Harrison Avenue, Boston, Massachusetts 02111, USA
^
Nuffield Department of Medicine, Ludwig Institute for Cancer Research, University of Oxford,
Old Road Campus, Roosevelt Drive, OX3 7DQ, UK
1

§
Present address: Department of Pharmacological Sciences, Icahn School of Medicine at Mount
Sinai, New York, New York 10029, USA
¥
Department of Biochemistry, University of Bristol, Biomedical Sciences Building, University
Walk, Bristol BS8 1TH, UK
*
Corresponding author.
* Corresponding author.
*
E-mail addresses: alexei.degterev@tufts.edu (A. Degterev), gdcuny@central.uh.edu (G.D.
Cuny)
2

ABSTRACT
Receptor-interacting protein kinase 2 (RIPK2) is a key mediator of nucleotide-binding
oligomerization domain (NOD) cell signaling that has been implicated in various chronic
inflammatory conditions. A new class of RIPK2 kinase/NOD signaling inhibitors based on a 3,5-
diphenyl-2-aminopyridine scaffold was developed.
Several co-crystal structures of
RIPK2•inhibitor complexes were analyzed to provide insights into inhibitor selectivity versus the
structurally related activin receptor-like kinase 2 (ALK2) demonstrating that the inhibitor sits
deeper in the hydrophobic binding pocket of RIPK2 perturbing the orientation of the DFG motif.
In addition, the structure-activity relationship study revealed that in addition to anchoring to the
hinge and DFG via the 2-aminopyridine and 3-phenylsulfonamide, respectively, appropriate
occupancy of the region between the gatekeeper and the αC-helix provided by substituents in the
4
- and 5-positions of the 3-phenylsulfonamide were necessary to achieve potent NOD cell
signaling inhibition. For example, compound 18t (e.g. CSLP37) displayed potent biochemical
RIPK2 kinase inhibition (IC = 16 ± 5 nM), > 20-fold selectivity versus ALK2 and potent NOD
5
0
cell signaling inhibition (IC = 26 ± 4 nM) in the HEKBlue assay. Finally, in vitro ADME and
5
0
pharmacokinetic characterization of 18t further supports the prospects of the 3,5-diphenyl-2-
aminopyridine scaffold for the generation of in vivo pharmacology probes of RIPK2 kinase and
NOD cell signaling functions.
KEYWORDS
inhibitor, kinase, NOD, nucleotide-binding oligomerization domain, receptor-interacting protein
kinase 2, RIPK2
3

1
. Introduction
The nucleotide-binding oligomerization domain (NOD)-containing proteins 1 and 2 are
members of the NOD-like receptor (NLR) family that are involved in the innate immune
system’s detection of bacterial peptidoglycan (PG) derivatives. NOD1 is stimulated by bacterial
PG fragments containing diaminopimelic acid (DAP), while NOD2 senses muramyl dipeptide
(MDP). NOD1/2 then initiates assembly of signaling complexes by oligomerization through the
nucleotide-binding oligomerization domains (NBD), which triggers the recruitment of interacting
proteins through homotypic caspase-activated recruitment domain (CARD)-mediated
interactions [1-4]. Receptor-interacting protein kinase 2 (RIPK2) is one of the key molecules in
NOD-dependent signaling as it plays an essential role in the activation of NF-κB pathway and
mitogen-activated protein kinase (MAPK) pathways that ultimately lead to synthesis of pro-
inflammatory cytokines and antimicrobial molecules [2, 5].
RIPK2, which has dual serine/threonine and tyrosine kinase function, was identified as an
essential mediator of signaling in both innate and adaptive immune systems [6]. In addition, non-
kinase activity, particularly X-linked inhibitor of apoptosis (XIAP)-mediated ubiquitination of
RIPK2, significantly contributes to the functions of this protein [7-14]. Aberrant RIPK2–NOD
signaling pathways appear to play key roles in various inflammatory diseases. Most importantly,
positive or negative dysregulation of the NOD2-dependent signaling pathway has been shown to
facilitate several chronic inflammatory disorders such as Crohn’s disease [15-17], Blau
syndrome [18-19], early-onset sarcoidosis [19] and multiple sclerosis [20]. Therefore, RIPK2–
NOD has emerged as a crucial mediator of inflammatory pathways and thus a potential
therapeutic target for treating an array of inflammatory and autoimmune diseases.
4

O
S
Cl
N
N
F
N
N
N
N
O
HN
N
HN
HN
N
NH
N
Cl
N
O
N
N
O
O
O
NH
O
F
Gefitinib (2)
OD36 (3a)
OD38 (3b)
SB203580 (1)
NH
N
H
F3C
N
N
N
O O
HN
N
O
N
NH2
S
N
O
tBu
N
F
N
N
N
H
N
N
N
Ponatinib (5)
GSK583 (6a)
WEHI-345 (4)
H N
S
N
2
N
N
N
X
H
N
HN
N
O O
NH
O
O O
S
tBu
S
O
N
R1
N
RO
O
N
R O
2
N
i
6
b R = H
7a X = F, R = Pr, R = Et
8
1
2
N
t
6
c R = P(=O)(OH)2
7b X = Cl, R = Bu, R =
1 2
Figure 1. Reported RIPK2 kinase inhibitors.
With growing appreciation of RIPK2–NOD involvement in inflammatory and
autoimmune cellular pathways and diseases, interest in RIPK2 small molecule ligands has
accelerated. Several compound classes have been reported to inhibit RIPK2’s kinase activity
(
Figure 1), including SB203580 (1) [21], gefitinib (2) [22], OD36 (3a) and OD38 (3b) [23],
WEHI-345 (4) [24], ponatinib (5) [25], GSK583 (6a) [26], 6b and its prodrug 6c [27], 7a and
b [28], and 8 [29]. Several of these compounds (e.g. 1, 2, 4, 6a, 6b and 7b) have been shown to
7
bind the active (DFG-in/αC-helix-in) conformation of RIPK2 via a type I manner, while 5
5

demonstrated binding to the inactive (DFG-out/αC-helix-in) conformation via a type II binding
mode. These molecules have been shown to inhibit various downstream processes in cells
associated with NOD signaling, such as NF-κB activation and MDP stimulated IL-8 or TNFα
release. In addition, 3a and 3b [23] have demonstrated in vivo efficacy in an MDP-induced
peritonitis model, 4 [24] and 8 [29] have shown activity in the experimental autoimmune
encephalomyelitis model of multiple sclerosis and intestinal and lung inflammation models,
respectively, while 7a [28] inhibited inflammatory cytokine release in the ex vivo experiments
using human inflammatory bowel disease biopsy samples. Herein we report the structure-
activity relationship (SAR) of a new structure class of RIPK2 kinase and NOD cell signaling
inhibitors that will be useful as probes for further understanding this protein, its role in NOD cell
signaling and potentially as lead compounds for therapeutic development.
2
. Design and synthesis
In contemplating a new class of RIPK2 kinase inhibitors we considered two scaffolds (9a
and 9b) that had previously been used for activin receptor-like kinase 2 (ALK2) inhibitors. In
both cases, RIPK2 was a prevailing off-target revealed by kinome-wide selectivity profiling [30-
3
1]. The 3,5-diphenyl-2-aminopyridine scaffold, exemplified by LDN-214117 (9a) [30] that
displayed a modest IC of 100 nM in the RIPK2 ADPGlo kinase assay, was chosen to pursue
5
0
(
Figure 2A, Table I).
To improve potency, we desired to introduce a functional group into the 3,5-diaryl-2-
aminopyridine scaffold that would ensure an interaction with the DFG segment of RIPK2. In
order to identify a candidate moiety, the Harvard Medical School Library of Integrated Network-
based Cellular Signatures (http://lincs.hms.harvard.edu) database was reviewed seeking to find
inhibitors with non-overlapping kinase profiles except for RIPK2. Over 160 compounds in the
6

database displayed RIPK2 inhibition and among them PLX4032 (10, vemurafenib, Figure 2A), a
V600E
V600E
B-Raf
kinase inhibitor, showed ~96% inhibition at 10 μM. A B-Raf
•10 co-crystal
structure reveals a DFG-in type I½ binding mode, a hydrogen bond between the sulfonamide and
the backbone NH of Phe595 and Gly596 in the DFG motif and displacement of the αC-
helix [32-34].
A
OMe
NH
MeO
MeO
N
R
N
9
a : R = Me; 9b : R = H
F
O
Cl
O
O
S
N
H
F
N
N
H
10
B
• Pocket binding • Solvent exposed
region region
HN
R
R S
O
O
R
N
•
DFG binding
region
•
Hinge binding
region
V600E
Figure 2. A) Structures of ALK2 inhibitors 9a–b and B-Raf
inhibitor 10, all of which block
RIPK2 kinase activity. B) RIPK2 inhibitor design incorporating the sulfonamide DFG binding
fragment of 10 into scaffold 9.
7

Next, 9a was docked into the DFG-in/αC-helix-in (PDB 5AR4) and DFG-in/αC-helix-
out (PDB 5AR8) structures of RIPK2 using AutodockTool-1.5.6. The co-crystal structures of
V600E
ALK2•9b (PDB 4BGG) and B-Raf
•10 (PDB 3OG7) as well as the RIPK2•9a docked
structures were compared to identify potential positions on 9a for installation of the sulfonamide
from 10 (Figure 3) in order to engage the DFG of RIPK2. The overlay results, with either the
DFG-in/αC-helix-in or DFG-in/αC-helix-out conformations, suggested that the sulfonamide
should be installed into the 4-position of the 3-phenyl to generate 9a/10-hybrids as a potential
new class of RIPK2 inhibitors (Figure 2B).
A
B
C
: ALK2 (PDB ID: 4BGG)
:
:
:
B-RafV600E (PDB ID: 3OG7)
RIPK2 DFG-in/αC-helix-in (PDB ID: 5AR4)
RIPK2 DFG-in/αC-helix-out (PDB ID: 5AR8)
:
9a
9b
10
:
:
Figure 3. Overlaid structures of 10 vs 9b and 10 vs docked 9a suggested the position to install
the sulfonamide moiety on the 3,5-diphenyl-2-aminopyridine scaffold. A) Superimposed crystal
V600E
structures of 10 in B-Raf
and 9b in ALK2; B) and C) Docking result of 9a in RIPK2 DFG-
in/αC-helix-in and DFG-in/αC-helix-out structures, respectively, overlaid with co-crystal
V600E
structure of 10 in B-Raf
.
8

Various 1-(4-(5-phenylpyridin-3-yl)phenyl)piperazines utilized in this study were
prepared using modifications of the reported syntheses of 9a and 9b [30]. Primary or secondary
bromoanilines 11 and 12, many of which were commercially available, were used as starting
materials. However, several non-commercial anilines were prepared by either nucleophilic
aromatic substitution or nitration, followed by iron-mediated nitro reduction. The bromoanilines
1
1 and 12 were allowed to react with sulfonyl chlorides in the presence of pyridine using slightly
modified conditions to provide 13 (Scheme 1). Miyaura borylation reaction of 13 with
bis(pinacolato)diboron using either PdCl (dppf) in THF or PdCl (dppf) in 1,4-dioxane furnished
2
2
arylboronic esters 14. Palladium-catalyzed Suzuki–Miyaura coupling of 14 with 3-bromo-5-
iodopyridines 15a–c delivered 3-bromo-5-(phenylalkylsulfonamide)pyridines 16.
Scheme 1. Synthesis of 3-bromo-5-arylpyridine intermediates 16.*
9

*
1
5
Reagents and Conditions: (a) alkyl or phenyl sulfonyl chloride, Method A: pyridine, CH
6–48 h, Method B: pyridine, cat. DMAP, CH Cl , rt, 16–48 h, Method C: cat. DMAP, pyridine,
0 ºC, 3 h, Method D: pyridine, 50 ºC, 2.5 h or Method E: pyridine, rt, 16–48 h (30–99%); (b)
2 2
Cl , rt
2
2
bis(pinacolato)diboron, Method F: 3 mol% PdCl (dppf), KOAc, THF, reflux, 16 h or Method G:
2
1
0 mol% PdCl (dppf), KOAc, 1,4-dioxane, 80 ºC, 40 h (24–90%); (c) 15a, 15b or 15c, 10 mol%
2
Pd(PPh ) , 1 M Na CO , MeCN, DMF, 90 ºC, 16 h (38–96%).
3
4
2
3
10

2
4
-Chloropyridine intermediates 16 (R = Cl) were subjected to a second Suzuki reaction
with 4-(4-tert-butoxycarbonylpiperazinyl)phenylboronic acid pinacol ester to deliver 17, which
were treated with trimethylboroxine in the presence of a palladium catalyst followed by Boc
deprotection using trifluoroacetic acid (TFA) in dichloromethane at room temperature to furnish
1
8a–e and 18aa (Scheme 2). Similarly, 2-aminopyridine or pyridine intermediates 16 (R = NH
4 2
or H) were coupled with 4-(4-tert-butoxycarbonylpiperazinyl)phenylboronic acid pinacol ester
and deprotected to give 18f z.
–
Scheme 2. Synthesis of 1-(4-(5-phenylpyridin-3-yl)phenyl)piperazine derivatives 18a
–z and
1
8aa.*
11

*
Reagents and Conditions: (a) 4-(4-tert-butoxycarbonylpiperazinyl)phenylboronic acid pinacol
ester, 15 mol% Pd(PPh ) , 1 M Na CO , DME, 90 ºC, 16 h (79–99%); (b) 1) trimethylboroxine,
3
4
2
3
2
0 mol% Pd(PPh ) , K CO , 1,4-dioxane, 110 ºC, 16 h; 2) 10% TFA in CH Cl , rt, 16 h (62–
3 4 2 3 2 2
9
6% over two steps); (c) 1) 4-tert-butoxycarbonylpiperazinyl)phenylboronic acid pinacol ester,
1
5 mol% Pd(PPh ) , 1 M Na CO , DME, 90 ºC, 16 h; 2) 10% TFA in CH Cl , rt, 16 h (18–62%
3
4
2
3
2
2
over two steps).
Finally, a set of compounds was prepared with modifications to the solvent exposed
region. In this case, intermediate 16j was coupled with 2-, 3-, and 4-
(methanesulfonyl)phenylboronic acid to give 18ab, 18ac and 18ad, respectively (Scheme 3).
Similarly, 16m was coupled with (3-(4-(tert-butoxycarbonyl)piperazin-1-yl)phenyl)boronic acid
pinacol ester and deprotected to provide 18ae.
Scheme 3. Synthesis of 18ab–18ae.*
*
Reagents and Conditions: (a) 2-, 3-, or 4-(methanesulfonyl)phenylboronic acid or (3-(4-(tert-
butoxycarbonyl)piperazin-1-yl)phenyl)boronic acid pinacol ester, 10 mol% Pd(PPh ) , 1 M
3
4
12

2 3
Na CO , MeCN, DMF, 90 ºC, 16 h (22–83%); (b) In the case of 3-(4-(tert-
butoxycarbonyl)piperazin-1-yl)phenyl: 10% TFA in CH Cl , rt, 16 h (48% over two steps).
2
2
3
. Results and discussion
As previously mentioned, structure analyses using either the DFG-in/αC-helix-in or
DFG-in/αC-helix-out conformations suggested that the sulfonamide should be installed into the
para-position of the 3-phenyl group occupying the binding pocket region. To test this initial
model, a derivative of 9a containing a 1-propanesulfonamide in this position (e.g. 18a) was
evaluated for RIPK2 inhibitory activity. However, it was not active (Table 1). Therefore, the 1-
propanesulfonamide was changed to a methylsulfonamide and moved to the meta-position (18b)
to further test the strategy. Gratifyingly, this compound proved to be a moderately potent
inhibitor of RIPK2 (IC = 296 nM). Since the primary component of these compounds was 9a,
5
0
ALK2 kinase activity was chosen as the counter-screen for assessing selectivity [30]. Indeed,
8b proved to be more potent against ALK2 (IC = 31 nM). However, these preliminary results
1
5
0
indicated that placement of the sulfonamide in the meta-position was preferred. Several
additional derivatives (18c–e) examined homologation and branching of the sulfonamide. In
general, RIPK2 kinase inhibitory activity was modestly improved, but selectivity versus ALK2
was not significantly altered. Next, an amino group was introduced in place of the methyl at the
2
-position of the pyridine to provide better hinge binding interactions. As expected, this set of
compounds (18f–h) was more potent against both RIPK2 and ALK2. In addition,
phenylsulfonamide 18i was found to also be tolerated. Interestingly, increasing the size of the
ether to an ethyl or isopropyl (18j or 18k) was allowed by RIPK2, but less so for ALK2.
13

Table 1. RIPK2 and ALK2 kinases, and NOD2 cell signaling inhibitory activities of 9a and 18a–
k.
Kinase
RIPK2/NOD2
Compounds
R1
R2
R3
R4
IC50 (nM)
Cell Assay
IC (nM)
5
0
RIPK2 ALK2
9
a
OMe
OMe
OMe
H
Me 100 ± 23 24 [30]
NI
n
2
1
1
8a
8b
OMe NHSO Pr
Me
NI
NI
2944 ± 420
3705 ± 85
5239 ± 861
2669 ± 138
OMe
OMe
OMe
H
H
H
NHSO
Me
2
Me 296 ± 10 31 ± 34
Me 127 ± 18 86 ± 45
1
8c
2
NHSO Et
n
2
1
8d
NHSO
Pr
Me
Me
164 ± 7 108 ± 92
239 ±
i
2
1
8e
OMe
H
NHSO
Pr
76 ± 12
523 ± 18
170
1
8f
OMe
OMe
OMe
H
H
H
NHSO
NHSO
NHSO
2
Me NH
2
2
2
51 ± 26
14 ± 0.6
5 ± 6
9 ± 3
390 ± 22
243 ± 14
121 ± 27
n
2
1
8g
8h
Pr NH
Pr NH
i
2
1
11 ± 10
1
3 ±
14

0
.04
1
8i
OMe
OEt
H
H
H
NHSO
2
Ph NH
2
2
2
23 ± 0.5
21 ± 2
18 ± 8
5 ± 7
53 ± 36
46 ± 0.5
323 ± 12
299 ±
i
18j
NHSO
2
Pr NH
180
i
i
1
8k
O Pr
NHSO
2
Pr NH
NI
*
Values are shown as the mean of two or three determinations ± standard deviation; NI: No
inhibition up to 10 µM
Given the previously demonstrated importance of methoxy groups on the 3-phenyl ring
of 3,5-diphenyl-2-aminopyridine ALK2 inhibitors, a series of compounds was evaluated that
offered changes to this region, while maintaining a 1-propanesulfonamide in the meta-position
(Table 2) [30]. Installing a methyl in the para-position (18l) was detrimental for both RIPK2
and ALK2 inhibitory activities. However, moving the methoxy to the para-position (18m)
proved advantageous for improving selectivity (to 86-fold) for RIPK2 versus ALK2. Replacing
this group with hydroxyl (18n) or methyl (18o) resulted in reduced RIPK2 inhibitory activity and
selectivity over ALK2.
Table 2. RIPK2 and ALK2 kinase, and NOD2 cell signaling inhibitory activities of 18g and 18l–
o.
15

Kinase
RIPK2/NOD2
Cell Assay
Compounds R1
R2
IC50 (nM)
IC50 (nM)
RIPK2
ALK2
9 ± 4
1
8g
OMe
OMe
H
H
Me
14 ± 0.8
NI
243 ± 14
3347 ± 495
420 ± 34
1
8l
355 ± 318
3545 ± 2114
387 ± 42
2120 ± 509
18m
OMe
OH
Me
41 ± 19
54 ± 6
102 ± 22
1
8n
H
> 5000
1
8o
H
1064 ± 109
*
Values are shown as the mean of two or three determinations ± standard deviation; NI: No
inhibition up to 10 µM
Based on the encouraging results with a methoxy group in the ortho-position relative to
the 3-phenylsulfonamide, additional changes were examined in this region (Table 3).
Reintroducing a meta-methoxy (18p) retained potent RIPK2 inhibitory activity, but predictably
resulted in reduced selectivity versus ALK2. Increasing the size to ethoxy (18q) or tethering the
ethers in an ethylenedioxy (18r) resulted in loss of inhibition for both RIPK2 and ALK2.
However, replacement with methyl (18s), fluoro (18t) or chloro (18u) was well tolerated with 10
to 20-fold selectivity versus ALK2. Not surprisingly, replacement of the sulfonamide with a
methoxy (18v) completely abolished selectivity. Installing a methylene between the sulfonamide
and phenyl (18w) eroded both enzyme inhibitory activity and selectivity. In addition,
incorporation of the sulfonamide in a ring (18x–y) was also not well tolerated. Interestingly,
16

removal of the 2-amino group on the pyridine of 18t (e.g. 18z) did not erode kinase inhibitory
activity. However, replacing the amino group with methyl (18aa) decreased both RIPK2 and
ALK2 inhibitory activities, as expected.
Table 3. RIPK2 and ALK2 kinase, and NOD2 cell signaling inhibitory activities of 18m, 18p–z
and 18aa.
Kinase
RIPK2/NOD
Compounds
R1
R2
R3
R4
IC50 (nM)
2 Cell Assay
IC (nM)
5
0
RIPK2
ALK2
3
545 ±
114
n
1
8m
H
OMe
NHSO
2
Pr
NH
2
32 ± 9
476 ± 97
2
n
n
n
n
1
8p
8q
OMe OMe
OEt OMe
NHSO
NHSO
NHSO
NHSO
NHSO
2
2
2
2
Pr
Pr
Pr
Pr
Pr
NH
NH
NH
NH
NH
2
2
2
2
20 ± 0.8
NI
57 ± 13
NI
1 ± 0.4
NI
1
1
8r
8s
OCH
Me
F
2
CH
2
O
520 ± 13
41 ± 1
16 ± 5
NI
2458 ± 309
129 ± 38
26 ± 4
1
OMe
OMe
395 ± 245
348 ± 107
n
1
8t
2
2
17

n
1
8u
Cl
F
OMe
NHSO
2
Pr
NH
NH
2
2
2
2
2
21 ± 4
39 ± 4
99 ± 20
202 ± 88
29 ± 14
46 ± 0.4
595 ± 70
1290 ± 94
1
8v
OMe
OMe
n
2
1
8w
H
OMe CH
2
NHSO
Pr NH
145 ± 57
n
1
1
8x
8y
H
H
F
CH
2
CH
N(SO
2
2
Pr)
NH
NH
H
185 ± 16
78 ± 20
NI
NI
NT
NI
2426 ± 116
1011 ± 27
352 ± 83
n
2
OCH
OMe
OMe
2
CH
2
N(SO
Pr)
n
n
1
8z
NHSO
NHSO
2
2
Pr
27 ± 2
18aa
F
Pr
Me
1414 ± 312
2556 ± 253
*
Values are shown as the mean of two or three determinations ± standard deviation; NI: No
inhibition up to 10 µM; NT: Not tested
Finally, the solvent exposed region of the inhibitor was briefly examined. Previous
studies have demonstrated that selectivity for RIPK2 can be improved by introducing functional
groups in this area that engage Ser25 in the glycine-rich loop, which is fairly unique to this
particular kinase [26]. For example, a co-crystal structure of 6 with RIPK2 reported by Haile
and co-worker showed a hydrogen bonding interaction of a methylsulfone with this residue [26].
The potency of the methylsulfone derivative was > 200-fold greater compared with the
unsubstituted analog. Based on these observations, (methylsulfonyl)phenyl moieties were
introduced to the solvent exposed region of the hybrid series (Table 4). However, the 2-, 3-, and
4
-(methylsulfonyl) derivatives (18ab, 18ac and 18ad) demonstrated significantly less potency.
On the other hand, moving the piperazine from the para- to meta-position (18ae) retained potent
RIPK2 kinase inhibition and improved selectivity versus ALK2 raising the possible that this
group provides a similar interaction with Ser25.
18

Table 4. RIPK2 and ALK2 kinase, and NOD2 cell signaling inhibitory activities of 18t and
1
8ab–18ae.
Kinase
RIPK2/NOD2 Cell
Assay IC (nM)
Compounds
R1
R2
IC50 (nM)
5
0
RIPK2
16 ± 5
ALK2
1
8t
F
H
H
H
F
A
B
C
D
E
348 ± 76
NI
26 ± 4
NI
1
8ab
NI
1
8ac
103 ± 16
264 ± 38
11 ± 0.6
NI
2655 ± 49
NI
1
8ad
NI
1
8ae
522 ± 317
16 ± 0.6
*
Values is are shown as the mean of two or three determinations ± standard deviation; NI: No
inhibition up to 10 µM
19

Figure 4. Co-crystal structure of 18f•RIPK2 (PDB 6S1F). The inhibitor (green) forms hydrogen
bonds to backbone NH and carbonyl of Met98 and Glu96, respectively, in the hinge region,
while the methylsulfonamide is oriented towards the Asp164 in the DFG motif. Lys47 forms an
ionic-ionic interaction with Glu66, in the αC-helix, stabilizing the DFG-in/Glu-in conformation.
In order to understand inhibitor interactions with RIPK2 and to gain insights for
preferential binding for a subset of compounds versus ALK2, compound 18f was co-crystallized
with RIPK2 (PDB 6S1F). The 18f•RIPK2 structure was resolved at a resolution of 3.1 Å (see
Table S1 for data collection and refinement statistics) and displayed RIPK2 in an inactive DFG-
in/Glu-in conformation with several critical residues misplaced from their active positions,
including misorientation of the catalytic residues, perturbation of the R-spine and P-loop, as well
20

as a disordered activation segment (Figure 4) similar to the 18m•RIPK2 complex [7]. The 2-
aminopyridine formed hydrogen bonds with the backbone NH and carbonyl of hinge residues
Met98 and Glu96, respectively. In addition, the 3-phenylsulfonamide of 18f projected towards
the hydrophobic pocket comprised of Val32, Leu70, Leu79, Ile93, Leu153 and Ala163, with the
sulfonamide oriented towards Asp164 in the DFG motif. Finally, an ionic-ionic interaction was
evident between Lys47 and Glu66 maintains the αC-helix in the Glu-in conformation.
21

22

Figure 5. (A) Alignment of 18f•RIPK2 (green; PDB 6S1F) and 9b•ALK2 (light purple; PDB
BGG) co-crystal structures. Several residues, the DFG and DLG motifs, as well as the αC-
4
helices are labeled. (B) Alignment of 18m•RIPK2 (dark purple; PDB 6FU5) and 9b•ALK2 (light
purple; PDB 4BGG) co-crystal structures. Several residues, the DFG and DLG motifs, as well as
the αC-helices are labeled. (C) Alignment of 18f•RIPK2 (green) and 18m•RIPK2 (dark purple)
co-crystal structures. The DFG motifs are labeled. (D) Comparison of the DFG and DLG of
1
8f•RIPK2 (green) and 9b•ALK2 (light purple) with the DFG of 18m•RIPK2 (dark purple).
A number of interactions observed in the 18f•RIPK2 co-crystal structure were seen for
inhibitor 9b bound to ALK2 with the two inhibitors occupying similar space in their respective
kinases (Figure 5A) [30]. Also the DFG (in RIPK2) and DLG (in ALK2) motifs, as well as the
αC-helix, were comparable. However, juxtaposition of the 9b•ALK2 and 18m•RIPK2 co-crystal
structures (Figure 5B) demonstrated that the latter inhibitor, which has the methoxy at the 4-
position on the 3-phenylsulfonamide and is more selective for RIPK2 versus ALK2, sits deeper
in the hydrophobic pocket. In addition, it perturbs the orientation of the DFG relative to the
DLG in ALK2. Interestingly, the DLG in ALK2 appears to be less flexible as evidenced by a
lack of disclosed DLG-out ALK2 inhibitors, whereas adoption of a DFG-out conformation in
RIPK2 has been previously observed with 5 [25]. The ability of the DFG in RIPK2 to move, in
the case of 18m relative to 18f in RIPK2 and 9b in ALK2 (Figures 5C and D), may play a role in
its enhanced selectivity for RIPK2 versus ALK2. One additional difference noted for 18m versus
1
8f is rotation of the sulfonamide (Figure 5C) potentially induced by the presence of the
methoxy in the 4-position.
23

Although kinase activity is a function of RIPK2, this protein is also an ubiquitination
substrate in response to NOD1/2 stimulation [8-9]. Several ubiquitin (Ub) ligases participate in
the conjugation of Ub chains to RIPK2. Particularly critical is XIAP-mediated ubiquitination of
RIPK2, which results in recruitment [10] and linear ubiquitin chain assembly complex (LUBAC)
mediated Met1-Ub conjugation of RIPK2 [11]. These latter processes are intimately involved in
the downstream signaling step leading to the assembly of the NF-κB-activating IKK kinase
complex [12]. Disruption of XIAP-RIPK2 interactions through genetic mutations in XIAP [14]
or RIPK2 inhibitors mitigates NOD2 inflammatory signaling [7, 13]. Given this array of
functional roles of RIPK2, the newly developed kinase inhibitors were assessed in a HEKBlue
assay that measures NF-κB activation, downstream of XIAP-mediated RIPK2 ubiquitination, in
response to L18-MDP stimulation.
A significant portion of the compounds, despite demonstrating potent RIPK2 kinase
inhibition (IC < 100 nM), either modestly blocked NOD2 signaling in the HEKBlue assay (100
5
0
nM < IC50 < 1 µM) or were comparatively inactive (IC50 > 1 µM), suggesting that inhibition of
RIPK2 catalytic activity was not sufficient for inhibition of NOD signaling in cells. In a recently
published report, Hrdinka, et al. demonstrated that a subset of these 3,5-diaryl-2-aminopyridine
based RIPK2 inhibitors (e.g. 18p and 18t) engaged RIPK2 to impede XIAP-mediated RIPK2
ubiquitination, leading to efficient inhibition of L18-MDP-induced NOD-dependent signaling
responses, e.g. CXCL8 production in U2OS/NOD2 cells and TNFα release from RAW264.7
macrophages. Both 18p and 18t interfered with the interaction between GST-BIR2-cIAP1 and
RIPK2 blocked XIAP-mediated RIPK2 ubiquitination resulting in disruption of NOD2 signaling
similar to ponatinib and GSK583 [7]. Furthermore, 18t did not block lipopolysaccharide and
pan-caspase inhibitor induced RIPK1 and RIPK3 dependent necroptosis [35] in RAW264.7
24

macrophages at 1 µM (data not shown) demonstrating the compound’s selectivity for
RIPK2/NOD signaling. The mechanism of how allosteric binding of compounds to the ATP
pocket disrupts XIAP-mediated RIPK2 ubiquitination is currently not known. However, the
structural features that appear to achieve NOD cell signaling inhibitory function for a subset of
compounds in Tables 1-4 as assessed in the NF-κB activation HEKBlue assay were the 2-
aminopyridine and 3-phenylsulfonamide anchoring to the hinge and DFG, respectively, as well
as appropriate occupancy of the region between the gatekeeper and the αC-helix provided by
substituents (e.g. H or OMe) in the 4-position and relatively small substituent (e.g. F, Cl, or
OMe) in the 5-positions of the 3-phenylsulfonamide. Absence of this last structural feature was
sufficient to eliminate NOD cell signaling, but not kinase inhibitory activity. Further
interrogation of this portion of the molecule will likely provide additional insights into the SAR
for disruption of NOD cell signaling.
Table 5. Physicochemical properties and pharmacokinetic parameters of 18t.
Aqueous Solubility
at pH 7.4
Permeability
Mouse Microsome Stability
-6
(×10 cm/s)
t1/2 (min)
CL (µL/min/mg)
int
(µΜ)
≥100
2.9
25.6
27
In vivo Pharmacokinetics*
Cmax (µM)
T
max (h)
t1/2 (h)
AUC
V
d
(L/kg)
CL
(mL/min/kg)
(
min•µg/mL)
25

0.55 ± 0.2
1.2
4.7 ± 0.2
175.2 ± 59.9
24.0 ± 6.2
59.7 ± 15.1
*
Single 10 mg/kg intraperitoneal administration of 18t to female C57BL/6 mice.
Since 18t was among a set of compounds that displayed potent biochemical RIPK2 kinase
and NOD cell signaling inhibitory activities, it was selected as a representative compound for
further in vitro characterization and pharmacokinetic study (Table 5). Its solubility in pH 7.4
PBS buffer was
≥100
μM. The compound also demonstrated both moderate permeability of 2.9 ×
-6
1
0 cm/s in the PAMPA and mouse liver microsome stability (CLint = 27 µL/min/mg; t1/2 = 25.6
®
min) assays. After a single 10 mg/kg intraperitoneal administration [6% Captisol in water
formulation] to 8-week old female C57BL/6 mice (n = 5 mice/per group), the compound reached
a maximum plasma concentration (Cmax) of 0.55 ± 0.2 µM in 1.2 h (Tmax) with a plasma
elimination half-life (t1/2) of 4.7 ± 0.2 h, an AUC of 175.2 ± 59.9 min•µg/mL, a volume of
distribution of 24.0 ± 6.2 L/kg and clearance (CL) of 59.7 ± 15.1 mL/min/kg, well below mouse
hepatic blood flow. Overall, these properties support 18t as a potential useful in vivo probe for
RIPK2 kinase and NOD cell signaling inhibition.
4
. Conclusion
A new class of RIPK2 kinase/NOD cell signaling inhibitors based on a 3,5-diphenyl-2-
aminopyridine scaffold was developed. Several co-crystal structures of RIPK2•inhibitor
complexes were analyzed to provide insights into inhibitor selectivity versus a related kinase
(e.g. ALK2). The inhibitors occupy space further into the hydrophobic binding pocket of RIPK2
altering the position of the DFG motif. The SAR studies also revealed that achieving potent
NOD cell signaling inhibition depends on three structural features of the inhibitors: anchoring to
the hinge and DFG via the 2-aminopyridine and 3-phenylsulfonamide, respectively, as well as
26

appropriate occupancy of the region between the gatekeeper and the αC-helix provided by
substituents in the 4- and 5-positions of the 3-phenylsulfonamide. For example, a representative
compound from the series 18t demonstrated both potent inhibition of recombinant RIPK2 kinase
(
IC50 = 16 ± 5 nM), > 20-fold selectivity versus ALK2 and NOD cell signaling inhibition (IC50
=
2
6 ± 4 nM). In addition, this compound was found to have in vitro ADME and pharmacokinetic
characteristics further supporting the use of 3,5-diphenyl-2-aminopyridines (e.g. 18t – CSLP37
and 18ae – CSLP58) as the basis for generating in vivo pharmacology probes of RIPK2 kinase
and NOD cell signaling functions.
5
. Experimental section
Unless otherwise stated, all reagents and solvents were obtained from commercial sources and
used directly without further purification. All reactions involving air-sensitive reagents were
carried out with magnetic stirring and in oven-dried glassware with rubber septa under argon
unless otherwise noted. Reactions were monitored by thin-layer chromatography on Baker-flex®
silica gel plates (IB2-F) using UV-light (254 and 365 nm) detection or visualizing agent
(
phosphomolybdic acid stain). Flash chromatography was performed on silica gel (230−400
®
mesh) utilizing Teledyne ISCO CombiFlash Rf. The NMR spectra were recorded at 25 ºC using
1
13
a JEOL ECA ( H NMR at 400, 500, or 600 MHz, and C NMR at 100, 125, or 150 MHz).
1
Chemical shifts (δ) are given in parts per million (ppm) with reference to solvent signals [ H-
1
3
NMR: CDCl (7.26 ppm), CD OD (3.30 ppm), DMSO-d6 (2.49 ppm); C-NMR: CDCl (77.0
3
3
3
ppm), CD OD (49.0 ppm), DMSO-d6 (39.5 ppm)]. Coupling constants (
J) are given in Hz. The
3
1
3
C-F coupling patterns labeled in C-NMR indicate visible patterns in spectra. High resolution
mass spectra (HRMS) were carried out using AccuTOF by the Department of Chemistry, The
University of Texas at Austin. The spectra were measured using TOF-MS with an ESI ionization
27

source and reported as m/z (relative intensity) for the molecular ion [M]. All tested compounds
had a purity 95% as determined by high-performance liquid chromatography (HPLC) analyses
≥
using a Waters 1525 instrument equipped with a quaternary pump and a Proteo-C12 column (250
mm × 1 mm, 4 µm). UV absorption was monitored at λ = 220 nm. HPLC gradient went from 0%
(method A) or 2% (method B) MeCN in H
2 2
O to 90% MeCN in H O (both solvents contain 0.1%
trifluoroacetic acid) with a total run time of 30 min and a flow rate of 0.5 mL/min.
General Procedure for the Preparation of
4-bromo-2-methoxyphenyl)propane-1-sulfonamide (13a). To a solution of 4-bromo-2-
methoxyaniline (11a) (300 mg, 1.48 mmol) in anhydrous CH Cl (15 mL) was added pyridine
N-(Bromophenyl)sulfonamides (Method A); N-
(
2
2
(0.24 mL, 2.97 mmol) and 1-propanesulfonyl chloride (0.18 mL, 1.63 mmol) under an argon
atmosphere. The mixture was stirred at room temperature for 16 h. After being quenched with 1
N HCl(aq) (1.0 mL), water and CH Cl were added and then the layers were separated. The
2
2
combined organic phases were washed with brine, dried over anhydrous Na SO , filtered and
2
4
concentrated. The residue was purified by column chromatography on silica gel (EtOAc/hexane,
1
1
7
,
0:90 to 15:85) to give 13a (490.1 mg, 99%) as a light yellow oil; H NMR (CDCl
3
, 500 MHz)
.41 (1 H, d,
J
= 8.6 Hz), 7.08 (1 H, dd,
J
= 8.3, 2.3 Hz), 7.03 (1 H, d,
J
= 2.5 Hz), 6.74 (1 H, br)
1
3
3.88 (3 H, s), 3.02–2.98 (2 H, m), 1.83–1.79 (2 H, m), 1.00 (3 H, t,
J
= 7.4 Hz); C NMR
(
CDCl , 125 MHz) 149.5, 125.5, 124.2, 121.0, 117.5, 114.3, 56.1, 53.1, 17.1, 12.8.
3
N-(3-Bromo-5-methoxyphenyl)methanesulfonamide (13b). Prepared from 11b and purified by
column chromatography on silica gel (EtOAc/hexane, 10:90 to 20:80) to give 13b (82%) as a
1
yellow solid; H NMR (CDCl , 400 MHz) 6.96–6.95 (2 H, m), 6.86 (1 H, s), 6.75–6.74 (1 H, m),
3
1
3
3
1
.79 (3 H, s), 3.05 (3 H, s); C NMR (CDCl
3
, 100 MHz) 161.1, 137.8, 123.5, 115.1, 113.8,
05.0, 55.7, 39.5.
28