Lead identification of novel and selective TYK2 inhibitors

Article abstract of DOI:10.1016/j.ejmech.2013.03.070

A therapeutic rationale is proposed for the treatment of inflammatory diseases, such as psoriasis and inflammatory bowel diseases (IBD), by selective targeting of TYK2. Hit triage, following a high-throughput screen for TYK2 inhibitors, revealed pyridine 1 as a promising starting point for lead identification. Initial expansion of 3 separate regions of the molecule led to eventual identification of cyclopropyl amide 46, a potent lead analog with good kinase selectivity, physicochemical properties, and pharmacokinetic profile. Analysis of the binding modes of the series in TYK2 and JAK2 crystal structures revealed key interactions leading to good TYK2 potency and design options for future optimization of selectivity.

Full text of DOI:10.1016/j.ejmech.2013.03.070

European Journal of Medicinal Chemistry 67 (2013) 175e187
Contents lists available at SciVerse ScienceDirect
European Journal of Medicinal Chemistry
journal homepage: http://www.elsevier.com/locate/ejmech
Original article
Lead identification of novel and selective TYK2 inhibitors
,
,
Jun Liang b, Vickie Tsui b, Anne Van Abbema a 1, Liang Bao b 2, Kathy Barrett a,
Maureen Beresini a, Leo Berezhkovskiy c, Wade S. Blair a 3, Christine Chang a,
,
James Driscoll c 4, Charles Eigenbrot f, Nico Ghilardi d, Paul Gibbons b, Jason Halladayc,
,
Adam Johnson a, Pawan Bir Kohli a, Yingjie Lai b, Marya Liimatta a, Priscilla Mantik e,
Kapil Menghrajani c, Jeremy Murray f, Amy Sambrone e, Yisong Xiao g, Steven Shia f,
Young Shin c, Jan Smith a, Sue Sohn d, Mark Stanley b, Mark Ultsch f, Birong Zhang b,
,
Lawren C. Wu d, Steven Magnuson b
*
a Department of Biochemical and Cellular Pharmacology, Genentech, Inc., 1 DNA Way, South San Francisco, CA 94080, United States
b Department of Discovery Chemistry, Genentech, Inc., 1 DNA Way, South San Francisco, CA 94080, United States
c Department of Drug Metabolism and Pharmacokinetics, Genentech, Inc., 1 DNA Way, South San Francisco, CA 94080, United States
d Department of Immunology, Genentech, Inc., 1 DNA Way, South San Francisco, CA 94080, United States
e Department of Pharmaceutical Sciences, Genentech, Inc., 1 DNA Way, South San Francisco, CA 94080, United States
f Department of Structural Biology, Genentech, Inc., 1 DNA Way, South San Francisco, CA 94080, United States
g Department of Chemistry, WuXi Apptec, Waigaoqiao Free Trade Zone, Shanghai, PR China
a r t i c l e i n f o
a b s t r a c t
Article history:
A therapeutic rationale is proposed for the treatment of inflammatory diseases, such as psoriasis and
inflammatory bowel diseases (IBD), by selective targeting of TYK2. Hit triage, following a high-
throughput screen for TYK2 inhibitors, revealed pyridine 1 as a promising starting point for lead iden-
tification. Initial expansion of 3 separate regions of the molecule led to eventual identification of
cyclopropyl amide 46, a potent lead analog with good kinase selectivity, physicochemical properties, and
pharmacokinetic profile. Analysis of the binding modes of the series in TYK2 and JAK2 crystal structures
revealed key interactions leading to good TYK2 potency and design options for future optimization of
selectivity.
Received 8 December 2012
Received in revised form
28 March 2013
Accepted 31 March 2013
Available online 14 May 2013
Keywords:
Kinase inhibitor
Janus family kinase
Structureeactivity relationship
Crystal structure
Hit to lead
Ó 2013 Elsevier Masson SAS. All rights reserved.
Immunology
Abbreviations used: AcOH, acetic acid; ADME, absorption, distribution, metabolism, and excretion; aq., aqueous; TYK2BME, b-mercaptoethanol; Boc, tert-butyl carbamate;
CDCl3, deuterated chloroform; DCE, dichloroethane; DCM, dichloromethane; DMF, dimethylformamide; DMAP, 4-(N,N-dimethylamino)pyridine; DMSO, dimethylsulfoxide;
DMSO-d6, deuterated DMSO; DTT, dithiothreitol; EDCI, 1-ethyl-3-(3-dimethylaminopropyl) carbodiimide hydrochloride; ESI, electrospray ionization; h, hour; HATU, 2-(7-aza-
1H-benzotriazole-1-yl)-1,1,3,3-tetramethyluronium hexafluorophosphate; HOBt, hydroxybenzotriazole; IV, intravenous; LCMS, liquid chromatographyemass spectrometry;
LogD7.4, log of partition coefficient between octanol and pH7.4 aqueous buffer; MCT, methyl cellulose/tween; MDCK, MadineDarby canine kidney cells; Papp, apparent
permeability; Ph, phenyl; p-TsOH, para-toluenesulfonic acid; PO, by mouth; SCX-2, pre-packed IsoluteÒ silica-based sorbent with a chemically bonded propylsulfonic acid
functional group; SFC, supercritical fluid chromatography; TCEP, tris(2-carboxyethyl)phosphine; Tf, trifluorosulfonate; TFA, trifluoroacetic acid; THF, tetrahydrofuran; TLC,
thin layer chromatography.
* Corresponding author. Tel.: þ1 650 467 5325.
E-mail addresses: magnuson.steven@gene.com, stevenrm@gene.com (S. Magnuson).
Current address: Department of Cell Biology, Novartis, Emeryville, CA, United States.
Current address: ChemShuttle, Wuxi, PR China.
Current address: Department of Infectious Disease, MedImmune, Gaithersburg, MD, United States.
1
2
3
4
Current address: Department of Phamcokinetics, Theravance, South San Francisco, CA, United States.
0223-5234/$ e see front matter Ó 2013 Elsevier Masson SAS. All rights reserved.
http://dx.doi.org/10.1016/j.ejmech.2013.03.070

176
J. Liang et al. / European Journal of Medicinal Chemistry 67 (2013) 175e187
1. Introduction
[14]. Neutralization or genetic deficiency of IL-12 and IL-23 is effi-
cacious in mouse models of psoriasis and IBD [15]. Moreover, an-
tibodies against the p40 subunit of IL-12 and IL-23 are highly
efficacious and approved for the treatment of psoriasis, and have
shown efficacy in clinical studies in Crohn’s disease [16]. Thus, in-
hibition of TYK2 kinase activity may be an effective therapeutic
strategy for the treatment of psoriasis and IBD.
Therapies for psoriasis and IBDs must be safe. Therefore, in-
hibitors of TYK2 must have some degree of selectivity against JAK1,
JAK2, and JAK3. JAK1 and JAK3 mediate the signaling of multiple
receptors of the immune system. Deficiency of JAK1 results in
perinatal lethality in mice with significant impairments in lym-
phopoiesis [17], whereas deficiency of JAK3 results in severe com-
bined immunodeficiency in both mice and humans [18]. Although
complete inhibition of the kinase activity of JAK1 and JAK3 is likely
to be too broadly immunosuppressive, partial inhibition is tolerated
in humans and is in fact efficacious for the treatment of rheumatoid
arthritis and psoriasis [19]. Furthermore, it was recently reported
that inhibition of JAK3 is not sufficient in blocking effects of
1.1. TYK2 target biology
The Janus kinases (JAKs) are a family of four protein tyrosine
kinases that associate with the intracellular domains of a wide
range of cytokine and growth factor receptors that mediate
inflammation and hematopoiesis [1]. Each of the four Janus kinases
(JAK1, JAK2, JAK3, and TYK2) associates with a distinct set of re-
ceptors and induces intracellular signaling upon ligand binding by
phosphorylating the receptor intracellular domains to generate
docking sites for signal transducers and activators of transcription
(STAT) proteins. STAT proteins are subsequently phosphorylated by
the JAKs, form homodimers and heterodimers, and mediate gene
transcription upon translocation to the cell nucleus.
The JAK family kinase TYK2 associates with several cytokine
receptors, including those of the IL-12/IL-23 family, the IL-6 family,
the IL-10 family, and the type I IFN family [1b,c,2]. Studies of TYK2-
deficient mice [3] and a human patient lacking TYK2 protein [4]
indicate that TYK2 protein is essential for the signaling of IL-12
and IL-23, which are associated with T helper type 1 (Th1) and T
helper type 17 (Th17) inflammation, respectively [5]. IL-12 and IL-
23 are heterodimeric proteins consisting of a common p40 sub-
unit and a specific subunit of p35 for IL-12 and p19 for IL-23 (Fig. 1).
The IL-12 and IL-23 receptors are heterodimeric receptors that
immunologically relevant gc cytokines and that JAK1 is the domi-
nant kinase [20]. JAK2 is important for the signaling of several he-
matopoietic growth factors, including the erythropoietin receptor
that is responsible for the generation of red blood cells. Deficiency of
JAK2 results in embryonic lethality in mice due to defective eryth-
ropoiesis [21], and therefore complete inhibition of the kinase ac-
tivity of JAK2 will lead to anemia and other unacceptable effects on
hematopoietic cells. On the basis of such safety considerations, we
set forth to identify a TYK2 selective chemical series. In order to
precisely measure TYK2 potency and JAK1- and JAK2-selectivity at
both the enzyme and cell level, we had established biochemical and
cell-based assays and determined that in cells, the predominant
driver of IL-12, IL-6, and EPO pathway was TYK2, JAK1, and JAK2,
respectively [22]. At the outset of the project, we were uncertain
what the optimal level of JAK-family selectivity would be. We
initially set a target of at least 10-fold in cell-based assays with the
goal of obtaining compounds with a range of selectivity profiles that
we could use to answer this important question.
share a common IL-12R
chain for IL-12 [6] or the IL-23R for IL-23 [7]. The p40 subunit of IL-
12 and IL-23 binds to IL-12R 1, whose intracellular domain asso-
ciates with TYK2; whereas the p35 and p19 subunits of IL-12 and IL-
23 bind to IL-12R 2 and IL-23R respectively, whose intracellular
b1 chain that pairs with either the IL-12Rb2
b
b
domains associate with JAK2. In response to stimulation with IL-12
or IL-23, the TYK2 and JAK2 enzymes phosphorylate the intracel-
lular domains of the cytokine receptors, which in turn serve as
docking sites for STAT proteins [8]. Subsequently, the TYK2 and
JAK2 enzymes phosphorylate the STAT proteins, which then
translocate into the nucleus where they regulate gene expression
and transcription [1b,9]. Specifically, in the case of the IL-12
pathway, phosphorylation of STAT4 leads to expression of IFN
g
[10] while in the IL-23 pathway, phosphorylation of STAT3 leads to
increased IL-17 levels [7,11].
1.2. Lead matter and chemical plan
Th1 and Th17 immune responses and the IL-12/IL-23 pathways
have been implicated in the pathogenesis of psoriasis and the in-
flammatory bowel diseases (IBD; Crohn’s disease and ulcerative
colitis), which are chronic inflammatory disorders of the skin
(psoriasis) and gastrointestinal tract and colon (IBD) [12]. There is a
strong genetic association of the IL-12 and IL-23 pathways with
psoriasis and IBD [13] and substantial evidence for Th1 and Th17
inflammation in the skin and gut of patients with active disease
A survey of the literature at the time we started the project
revealed no TYK2 selective molecules, but there were a number of
pan-JAK inhibitors [23] that would serve as control compounds as
we identified new chemical matter. In order to identify potential
chemistry starting points for a TYK2 small molecule inhibitor
program, a high throughput screen (HTS) of the Genentech
corporate collection was performed using an ATP-consumption
assay [24]. Over 5000 primary hits [25] were identified and, after
Fig. 1. Cytokines IL-12, IL-23, EPO, and IL-6 signaling pathways.

J. Liang et al. / European Journal of Medicinal Chemistry 67 (2013) 175e187
177
hit we tested it against a panel of 50 kinases at 1
mM [30] and found
no other kinases with greater than 50% inhibition. Additionally,
DMPK-related properties and thermodynamic solubility were very
promising for this lead-like molecule.
At the time we began hit expansion of pyridine 1, we did not
have a crystal structure of compounds related to this scaffold bound
to TYK2. We assumed that pyridine 1 was interacting with the TYK2
protein in the ATP binding pocket and determined that the hit was
indeed competitive with ATP by doing the appropriate kinetic ex-
periments [31]. For the docking of compound 1 (Fig. 3), we built a
TYK2 homology model using a co-crystal structure of a literature
pan-JAK inhibitor [28], whose structure on its own was very flat and
had little three-dimensionality (see Supplemental data for details).
The modeling predicted hydrogen bonding of the pyridine nitrogen
to the backbone NH of Val981 in the hinge region. The dichlor-
ophenyl group, which was connected to the core pyridine via an
amide linker, was believed to extend back toward the P-loop. Even
though the docked ligand was strained, we believed the general
binding mode was correct, and small changes in the protein would
relieve the ligand strain by allowing the 2,6-dichlorophenyl to be
more orthogonal to the pyridine-amide moiety.
Our initial chemical plan was designed to understand the SAR of
three different areas of the molecule. One area was the aforemen-
tioned lipophilic dichlorophenyl ring, where we intended to
determine if alternate, less lipophilic substitutions were tolerated.
We also planned to investigate whether the amide provided the
best spatial and electronic linkage of the pyridine to the dichlor-
ophenyl group. Lastly, we wanted to investigate the hinge binding
pyridine and determine whether installation of an H-bond donor at
the 2-position of the pyridine, thus creating an additional hinge
interaction, would lead to a significant increase in potency. Based
on modeling, there did not appear to be enough room to install an
H-bond donor, other than perhaps eNH2, toward the gatekeeper
residue Met978 to interact with the backbone carbonyl of Glu979.
Instead, it seemed more likely that a hydrogen bond donor would
form an interaction with the backbone carbonyl of Val981, and
extended functional groups linked with this H-bond donor would
orient toward the solvent-exposed region.
Fig. 2. Structure and properties of screening hit pyridine 1.
clustering and a visual inspection of the hits, TYK2 IC50’s, along
with JAK1, 2 and 3 IC50’s, were obtained on approximately 900 hits.
To triage these 900 hits, chemical series and singletons were
prioritized based on their potency, ligand efficiency, physico-
chemical properties, JAK kinase selectivity and general kinase
selectivity. In choosing chemical series for initial hit expansion,
priority was given to chemical series with good properties in all
categories while compounds with poor data in multiple categories
were set aside. In cases where a scaffold had a significant liability,
but otherwise an attractive profile, the initial chemical plan, while
establishing SAR, was designed to determine if the liability was
addressable. Other properties, such as chemical tractability,
possible binding modes, and novelty were considered, but more so
factored into the initial chemical plan that was undertaken for hit
expansion.
One hit that was of high interest to us was pyridine 1 (Fig. 2).
This small, lead-like compound had very good potency (TYK2
Ki ¼ 0.23
mM) for its size and thus excellent ligand efficiency
(LE ¼ 0.55) [26]. Physicochemical property space was generally
good for a lead molecule [27], leaving ample opportunity to grow
the molecule or install certain functional groups as we optimized
potency. One concern with the physicochemical properties was the
cLogP of 2.9. While this value in and of itself was not alarming, we
recognized that we would not want the logP to climb much higher
during optimization. The dichlorophenyl moiety of the molecule
was the major contributor to the lipophilicity and we wanted to
find less lipophilic alternatives early in the scaffold expansion.
Pyridine 1 had been screened previously 13 times against other
targets, including 9 other kinase targets and had only been iden-
tified as a hit once before. The compound was also a weak hit in a
JAK2 HTS and during the TYK2 hit confirmation process, pyridine 1
was determined to have a 5-fold preference for TYK2 over closely-
related JAK2, a 9-fold preference over JAK1, and greater than 20-
fold selectivity over JAK3 [28,29]. This modest, but discernible
selectivity over the other JAK kinases, was a key attribute that led to
the selection of pyridine 1 for hit expansion. After resynthesis of the
2. Structure activity relationships
2.1. Dichlorophenyl region
While straightforward to investigate from a synthetic perspec-
tive, it proved difficult to identify less lipophilic substitutions than
the 2,6-dichloro pattern on the phenyl ring while maintaining
potency. A distinct preference for a 2,6-dihalo pattern (compounds
1, 6, 7) found in the original hit was clear after analysis of the initial
SAR (Table 1). Removal of both chlorine atoms (compound 2)
reduced activity by 10-fold while the analog with a single chlorine
atom (compound 3) was 5-fold less potent than the lead. Replacing
the chlorine of compound 3 with a methyl (compound 4), or
Fig. 3. Initial binding hypothesis of pyridine 1 and 3 regions for early scaffold expansion.

178
J. Liang et al. / European Journal of Medicinal Chemistry 67 (2013) 175e187
Table 1
Variation of dichlorophenyl ring.
Ex
R1
R2
TYK2 Ki (
m
M)a
JAK1 Ki (
m
M)a
JAK2 Ki (m
M)a
1
2
3
4
5
6
7
8
Cl
H
6-Cl
6-H
6-H
6-H
6-H
6-F
6-F
6-CH3
6-NH2
6-CF3
3-Cl
3-NH2
3-OCH3
5-Cl
5-CF3
5-OCH3
4-Cl
4-CF3
4-NH2
4-OCH3
0.23
3.1
1.1
2.2
>3.5
>3.5
>3.5
>3.5
1.7
3.1
>3.5
1.7
>3.5
>3.5
>3.5
>3.5
2.1
2.0
>3.5
4.6
1.2
3.2
1.7
>3.5
>3.5
2.0
Cl
CH3
CF3
Cl
F
2.4
>3.5
0.23
0.54
2.8
>3.5
1.9
>3.5
>3.5
>3.5
0.79
0.90
2.8
2.8
Cl
Cl
Cl
Cl
Cl
Cl
Cl
Cl
Cl
Cl
Cl
Cl
Cl
>3.5
>3.5
>3.5
3.2
>3.5
3.2
1.7
1.8
>3.5
3.0
>3.5
3.2
Fig. 4. Model of 19 interacting with the backbone carbonyl of Glu905 at the tip of P-
loop.
9
10
11
12
13
14
15
16
17
18
19
20
dichlorophenyl ring (Table 2). The reverse amide (compound 21)
was only 5-fold less potent than the lead molecule indicating that
there was some tolerance for altering the electronic nature of the
linker. Conversely, the sulfonamide (compound 22), reduced amide
(compound 23) and N-methyl amide (compound 24) analogs were
less active, indicating that the steric requirements were more sig-
nificant in this region. Similarly, inserting a one atom extension into
the linker (compound 25) or installing heteroaromatic linkers
(compounds 26e28) were not acceptable substitutions for the
amide. Based on this initial assessment, we continued to use the
amide that was present in the HTS hit for further lead identification
efforts.
1.2
>3.5
0.38
3.2
>3.5
>3.5
>3.5
3.1
a
Biochemical assays. Arithmetic mean of at least 3 separate determinations
(n ꢀ 3). On average, the coefficients of variation were less than 0.3 times the mean
for these assays.
trifluoromethyl (compound 5) reduced activity. Setting chlorine as
the preferred substitution at the 2-position, we investigated
different substitutions around the phenyl ring. At the 6-position,
introduction of a methyl, amino, or trifluoromethyl group (com-
pounds 8e10) was not as good as the 2,6-dichloro lead. Substitution
at the 3-position was not beneficial (compounds 11e13), but at the
2.3. Addition of hydrogen bond donor to hinge binding region
As anticipated, modifications to the pyridine ring, particularly
installation of a hydrogen bond donor at the 2-position, had a more
positive effect on TYK2 potency compared to changes to other re-
gions of the molecule. Interestingly, two of the first analogs made,
the unsubstituted 2-amino analog (compound 29) and closely
related N-methyl compound (analog 30) were less potent than the
parent analog lacking the hydrogen bond donor. In both cases, the
decreased potency is likely due to increased basicity of the pyridine
nitrogen when the hydrogen bond donor was added. The calculated
pKa [32] for the conjugate acid of the pyridine nitrogen in com-
pound 1 is 5.8. Conversely, when the pyridine is substituted with a
2-NH2 (compound 29) or a 2-NHCH3 (compound 30), the calculated
pKa’s for the conjugate acids of the pyridine nitrogen are 7.1 and 7.0,
respectively. This indicates that a significant portion of these more
basic pyridines may be protonated under the assay conditions
(pH ¼ 7.2), and this would result in weaker binding. When an
NH-acetyl group was installed (compound 31), the calculated pKa
for the conjugate acid of the pyridine nitrogen was 5.3, and in this
5-position
a chlorine substitution (compound 14) and tri-
fluoromethyl group (compound 15) were well-tolerated with
comparable or even slightly better potency than the unsubstituted
comparator, compound 3. Nevertheless, they were not more potent
than the lead molecule, and did not accomplish the goal of reducing
lipophilicity. Along these lines, the most promising vector on this
ring was the 4-position, where an amino substitution (compound
19) afforded good potency (TYK2 Ki ¼ 0.38
mM) and was over a full
log unit less lipophilic (cLogP ¼ 1.6) than the 2,6-dichloro lead.
Modeling, using the same homology model employed to predict the
binding mode of compound 1 (Fig. 3), indicated that the phenyl
group of compound 19 may be sitting under the tip of the P-loop
with the 4-amino substitution pointed toward a polar region and
specifically interacting with the backbone carbonyl of Glu905
(Fig. 4). While the aniline has potential liabilities, this result did
encourage us to further explore this vector with other polar sub-
stituents during the lead optimization of this series. There was no
notable improvement in JAK1 and JAK2 selectivity for the com-
pounds in Table 1, but given the small modifications that were
being made, this was perhaps not surprising.
case we observed
a
15-fold improvement in TYK2 potency
M).
(Ki ¼ 0.096 M) compared to the N-methyl analog 30 (Ki ¼ 1.5
m
m
The conjugate acid of the pyridine nitrogen of sulfonamide 32 has a
favorable calculated pKa of 4.3, but this substitution was not
tolerated in TYK2, presumably due to steric congestion of the
bulkier sulfonamide in the hinge region. A number of very potent
analogs (34, 36e38) were obtained when different 6-membered
heteroaromatics were added to the hydrogen bond donor. It is
2.2. Amide replacements
The second area of the molecule that we explored was the amide
linker between the hinge-binding pyridine and the lipophilic
interesting to compare NH-4-pyrimdine 37 (TYK2 Ki ¼ 0.0048
mM)

J. Liang et al. / European Journal of Medicinal Chemistry 67 (2013) 175e187
179
Table 2
Amide bond modifications.
Ex
1
XeY
TYK2 Ki (
m
M)a
JAK1 Ki (
m
M)a
JAK2 Ki (m
M)a
0.23
2.2
1.2
21
1.1
>3.5
>3.5
22
23
24
>3.5
>3.5
>3.5
>3.5
>3.5
>3.5
>3.5
>3.5
>3.5
Fig. 5. Modeled binding mode for compound 37.
the binding pocket (Fig. 5). Compound 37 was the most potent
analog that we obtained in the initial expansion of the HTS hit but
suffered from extremely high clearance in a rat PK study. An initial
survey of 5-membered heterocycles, such as pyrazole 39, was also
undertaken but no significant benefit in terms of selectivity or
clearance was obtained. Other functional groups were also inves-
tigated in our initial exploration of the hinge region. Inclusion of
the hinge-binding pyridine in a bicyclic 7-azaindole bidentate
hinge binder (compound 40) was also well-tolerated, but use of this
well-studied group [33] did not reveal any clear advantage to the
other hinge binder options and was not investigated further. Pyr-
25
26
>3.5
>3.5
>3.5
>3.5
>3.5
>3.5
idone 41 was equipotent (TYK2 Ki ¼ 0.23
pyridine 1, but this new analog was also 4-fold more potent against
JAK2 (Ki ¼ 0.060 M). This compound was co-crystallized with the
mM) to unsubstituted
m
27
>3.5
>3.5
>3.5
>3.5
1.4
2.0
JH1 kinase domains of both TYK2 and JAK2 (Fig. 6) [34]. These
crystal structures revealed that the interactions of the pyridone
core were optimal with the hinge residues of JAK2, whereas such
interactions were less optimal in the case of TYK2, due to a shift of
the core away from the deep pocket. For example, the hydrogen
bond donating pyridone nitrogen sat an ideal 2.8 Å away from the
backbone carbonyl of GLU930 in JAK2, whereas the pyridone ni-
trogen sat 3.4 Å away from the backbone carbonyl of GLU979 in
TYK2. With the shift away from the back pocket in TYK2, the
hydrogen bond accepting pyridone oxygen is positioned closer to
the backbone carbonyl of VAL981 than it is to backbone NH (3.1 vs.
3.4 Å), indicating the suboptimal positioning of the pyridone in
TYK2. This shift may be due to different conformations between
TYK2 and JAK2’s gatekeeper methionine residues (MET978 and
MET929, respectively). However, this residue is known to be flex-
ible in both proteins as well as in other kinases with methionine
gatekeepers, and specific causes of its differential conformation and
flexibility is unknown.
28
a
Biochemical assays. Arithmetic mean of at least 3 separate determinations
(n ꢀ 3). On average, the coefficients of variation were less than 0.3 times the mean
for these assays.
to the NH-2-pyrimidine 35 (TYK2 Ki ¼ 1.6
mM). The modeled
binding mode of compound 37 (Fig. 5) places a favorable CeH in
proximity to the backbone carbonyls of Val981 and P982 in the
hinge region. Conversely, compound 35 is forced to place and un-
favorable hydrogen bond accepting nitrogen in this region, thus
explaining the 300-fold difference in potency of these compounds.
It is also noteworthy to compare pyrimidine 37 to NH-phenyl
analog 33 (TYK2 Ki ¼ 0.14
mM). In this case, the 27-fold difference
in potency can be attributed to N(3) of the pyrimidine in 37 elim-
inating a steric clash with the pyridine C(3)eH, that would be
present with analog 33. Avoiding this steric clash allows the rings of
analog 37 to sit co-planar to one another and achieve a good fit in
2.4. NH-acetyl variation
In surveying all of these hinge binder modifications we felt that
NH-acetyl 31 presented a unique opportunity in terms of ligand

180
J. Liang et al. / European Journal of Medicinal Chemistry 67 (2013) 175e187
Table 3
Hinge binder modifications.
Ex
R
TYK2 Ki (
m
M)a
JAK1 Ki (
m
M)a
JAK2 Ki (m
M)a
1
H
NH2
NHCH3
NHC(O)CH3
NHS(O)2CH3
NH-phenyl
0.23
0.93
1.5
0.096
>3.5
0.14
2.2
>3.5
>3.5
2.9
>3.5
>3.5
1.2
2.9
29
30
31
32
33
0.31
>3.5
0.86
Fig. 6. Co-crystal structures of compound 41 with JAK2 (ligand in green, protein in
blue) and with TYK2 (ligand in orange, protein in salmon). Selected distances between
core atoms and hinge are shown, as well as names and numbers of relevant residues
(TYK2 in deep orange, JAK2 in deep blue). (For interpretation of the references to color
in this figure legend, the reader is referred to the web version of this article.)
34
35
36
0.029
1.2
1.0
>3.5
0.49
0.13
2.5
efficiency (0.47) and physicochemical properties, and a more
detailed investigation of NH-carbonyl substitutions was under-
taken (Table 4). While an ethyl group (compound 42) was well-
tolerated, branched acyclic alkyls such as isopropyl and t-butyl
(compounds 43 and 44) were not favored. In contrast to the iso-
propyl substitution though, small cycloalkyls (compounds 46e48)
were favored, with cyclopropyl analog 46 being 19-fold more
potent than methyl analog 31. When heteroatoms were introduced
into these small alkyl groups (compounds 50e53), in an attempt to
lower cLogP, a sharp drop-off in potency was observed. In addition
to exploring different amides, we also investigated whether the
amide functional group itself could be modified to related func-
tional groups such as ureas and carbamates. These changes were
indeed tolerated, as exemplified by the urea 54 and carbamate 55,
and provided another avenue for future lead optimization.
0.014
0.055
37
38
0.0049
0.010
0.20
1.0
0.020
0.063
3. Synthesis
39
0.019
0.36
0.035
One obvious benefit of this lead scaffold was the straightforward
synthesis of analogs. Exploration of the first area of interest e the
dichlorophenyl ring e was easily accomplished through simple
amide bond formation of 4-aminopyridine and a number of
substituted benzoyl chlorides. Synthesis of HTS hit 1 is shown in
Scheme 1, and specific conditions can be found in the Experimental
section of the paper, while spectral data for compounds 2e20 can
be found in the Supplemental information. In a similar fashion,
analogs 21 and 25 (Table 2) were made by the appropriate coupling
of an amine and an acid chloride (Supplemental data) while com-
pound 22 was made by coupling 4-aminopyridine with 2,6-
dichlorobenzene-1-sulfonyl chloride. Compound 23 was made by
C]O reduction of the amide found in compound 1 using boron
dimethyl sulfide complex, while compound 24 was made through
N-methylation of compound 1 with sodium hydride and methyl
iodide.
40b
0.10
0.24
1.1
0.23
41c
OH
0.10
0.060
a
Biochemical assays. Arithmetic mean of at least 3 separate determinations
(n ꢀ 3). On average, the coefficients of variation were less than 0.3 times the mean
for these assays.
b
5-Member ring fused to pyridine to form 7-azaindole hinge binder.
Compound 41 adopts the pyridone tautomer.
c
Compounds 26e28 (Table 2), where the linker amide was
replaced by an isosteric heteroaromatic ring were made as shown
in Scheme 2. Regioisomeric imidazoles 26 and 27 were
easily formed in 1 step reactions by coupling of the appropriate
amidine and bromomethyl ketone. Thus, compound 26 was
made by coupling isonicotinimidamide with 2-bromo-1-(2,6-
dichlorophenyl)ethanone, while analog 27 was synthesized
through reaction of 2,6-dichlorobenenzenecarboximidamide and
2-bromo-1-(4-pyridinyl)ethanone. Compound 28 was made in a
two-step route by first reacting 2,6-dichlorobenzoyl chloride with
hydrazine, and then coupling the resulting hydrazide to 4-pyridine
carbonitrile to afford the desired triazole 28.

J. Liang et al. / European Journal of Medicinal Chemistry 67 (2013) 175e187
181
Table 4
N-Acetyl variation.
31, and 42e53 were made from amine 29 by reacting with the
appropriate acid chloride, as exemplified for cyclopropyl amide 46
or through other standard amide bond forming reaction [37]. Other
analogs found in Tables 3 and 4 such as sulfonamide 32, pyridone
41, and urea 54, were made using standard transformations similar
to the ones discussed above and synthetic details can be found in
the Supplemental material.
4. Lead compound properties
4.1. In vitro pharmacology
Ex
R
TYK2 Ki (
m
M)a
JAK1 Ki (
m
M)a
JAK2 Ki (m
M)a
From all of the SAR generated in this initial survey of HTS hit
pyridine 1, cyclopropyl amide 46 clearly stood out and was profiled
more extensively (Table 5). With a TYK2 Ki of 4.8 nM, its ligand
efficiency (0.51) was comparable to the screening hit, and logD and
solubility were still in acceptable range. In terms of biochemical
selectivity, cyclopropyl amide 46 is 17-fold more potent at TYK2
than JAK1 and 6-fold more potent at TYK2 than JAK2 [28,38]. This
level of selectivity is slightly better than the HTS hit 1, and while the
selectivity is admittedly subtle, the scaffold preferentially binds to
TYK2 compared to other JAK kinases. Given the high homology of
the JAK ATP binding sites, this was an important finding on which to
build during lead optimization. To assess broader kinase selectivity,
compound 46 was tested in a panel of 185 different kinases at a
single point concentration of 500 nM (100ꢁ the TYK2 Ki) at Invi-
trogen. Only 7 other kinases (besides TYK2, JAK1 and JAK2) had
>70% inhibition at this concentration, and after Ki determination,
only CDK9 (111 nM) had a selectivity ratio of less than 50-fold (23-
fold for CDK9 Ki/TYK2 Ki). The general kinase selectivity, while
promising, would need to be continually monitored during lead
optimization with particular attention to CDK9.
31
42
43
44
45
46
47
48
49
50
51
CH3
CH2CH3
CH(CH3)2
C(CH3)3
CH2ecyclopropyl
Cyclopropyl
Cyclobutyl
Cyclopentyl
Phenyl
0.096
0.064
0.84
>3.5
0.094
0.0048
0.074
0.24
2.9
2.4
0.31
0.25
>3.5
>3.5
0.38
0.028
0.51
2.4
>3.5
>3.5
>3.5
0.084
1.5
>3.5
>3.5
>3.5
>3.5
2
0.25
1.2
2.7
1.4
2.4
CH2OH
CH2NH2
52
53
>3.5
>3.5
>3.5
>3.5
1.5
1.3
54
55
NHCH2CH3
OCH3
0.018
0.17
0.43
2.6
0.070
0.61
a
Biochemical assays. Arithmetic mean of at least 3 separate determinations
(n ꢀ 3). On average, the coefficients of variation were less than 0.3 times the mean
for these assays.
With compound 46 we had obtained sufficient biochemical
potency to assess cell potency and see how cellular selectivity
correlated with biochemical selectivity (Table 5). In NK92 cells, IL-
Synthesis of many of the analogs that probed addition of a
second hinge binder (Tables 3 and 4) made use of bromopyridine
intermediate 56 (Scheme 3) and aminopyridine 29. Bromopyridine
56 itself was made by coupling 4-amino-2-bromopyridine to 2,6-
dichlorobenzoyl chloride in the same way that compound 1 was
made. Aminopyridine 29 could be made by either direct displace-
ment of the bromide with ammonia under forcing conditions [35],
or by reacting with lithium hexamethyldisilazide under Hartwig
conditions [36], followed by acidic work-up. Buchwald coupling
from bromide 56 was used to make NH-aryl and heteroaryl analogs
33e39, as specifically exemplified for aminopyrimidine 37. Amides
12 stimulation leads to STAT4 phosphorylation in
a TYK2-
dependent fashion [22] and this was used to assess TYK2-
dependent cell activity (see Supplemental data for experimental
details). Compound 46 was approximately 100-fold less potent in
the TYK2-dependent cell assay than it was in the TYK2 biochemical
assay. The Km for ATP in the TYK2 biochemical assay is 10 mM so the
100-fold shift to the cell assay, where the intracellular ATP con-
centration is approximately 1 mM, was not unexpected [39]. To
assess JAK1 and JAK2-dependent cell activity, IL-6 mediated STAT3
phosphorylation and Epo-mediated STAT5 phosphorylation assays,
23
R=H, X=H,H
NH
Cl
R
Cl
2
Cl
Cl
iii.
O
ii.
i.
S
O
N
N
X
b
HN
N
1
R=H, X=O
iv.
N
22
24
R=CH , X=O
3
aReagents and conditions: (i) 2,6-dichlorobenzoyl chloride, DMF, NEt3, 95%;
(ii) 2,6-dichlorobenzene-1-sulfonyl chloride, N-methylmorpholine, DMF, 15%;
b
(iii) BH3-SMe2, THF, 44%; (iv) NaH, DMF; MeI, 67%. Analogs 2-21 and 25
were made by analogous procedure to compound 1, coupling appropriate amine
and acid chloride.
Scheme 1. Synthesis of Analogs 1e25.

182
J. Liang et al. / European Journal of Medicinal Chemistry 67 (2013) 175e187
Scheme 2. Synthesis of analogs 26e28.
Table 5
Physicochemical and in vitro PD properties of compound 46.
in TF-1 cells, were used respectively [22]. For cyclopropyl amide 46
there was reasonable agreement between biochemical selectivity
and cell selectivity, with a small degradation in selectivity ratios in
the cell assays. The JAK1 Ki/TYK2 Ki biochemical selectivity index of
cyclopropyl amide 46 is 17-fold, but the ratio of IL6-mediated
pSTAT3 (JAK1) EC50 over IL12-mediated pSTAT4 (TYK2) EC50 indi-
cated only 5-fold cellular JAK1 selectivity. Similarly, the JAK2 Ki/
TYK2 Ki biochemical selectivity index of compound 46 is 6-fold, but
the ratio of Epo-mediated pSTAT5 (JAK2) EC50 over IL12-mediated
pSTAT4 (TYK2) EC50 indicated only 4-fold cellular JAK2 selectivity.
As mentioned in the introduction, we were not certain, at the
outset of the project, what the optimal level of cellular selectivity
was, but had set an initial target of 10-fold for TYK2 vs. both JAK1
and JAK2. While compound 46 had demonstrated meaningful
cellular activity and selectivity, a major task during lead optimi-
zation would be to increase selectivity and deliver a set of
Physicochemical
properties
In vitro potency
and selectivity
TYK 2 LE
TYK2 LLE
0.51
4.9
TYK2 Ki
JAK1 Ki
0.0048
0.084
(17-fold)
0.028
(6-fold)
0.38
mM
mM
(selectivity index)
JAK2 Ki
MW
350.2
mM
(selectivity index)
IL12 e pSTAT4 EC50
a,b
a,c
tPSA
logD (pH 7.4)
71
3.4
m
M
IL6 e pSTAT3 EC50
(cell selectivity index)
2.0
mM (5-fold)
a,d
Aqueous
0.17 mg/mL Epo e pSTAT5 EC50
1.7
mM (4-fold)
solubility (pH 7.4)
(cell selectivity index)
a
Cell-based assays. Arithmetic mean of at least 4 separate runs (n ꢀ 4). On
average, the coefficients of variation were less than 0.5 times the mean for cell-
based assays.
b
Cell-based assay of TYK2 inhibition.
Cell-based assay of JAK1 inhibition.
Cell-based assay of JAK2 inhibition.
c
d
Scheme 3. Synthesis of analogs 31, 33e39, and 42e53.

J. Liang et al. / European Journal of Medicinal Chemistry 67 (2013) 175e187
183
Fig. 7. a) Compound 46 co-crystallized with TYK2 (ligand in orange, protein in salmon) and JAK2 (ligand in green, protein in deep blue); b) Compound 46 with TYK2 e expanded
view of dichlorophenyl region; c) Compound 46 with TYK2 e expanded view of cyclopropyl amide region. (For interpretation of the references to color in this figure legend, the
reader is referred to the web version of this article.)
compounds with a range of selectivity to help determine, more
definitively, the optimal level of TYK2 cellular selectivity over JAK1
and JAK2.
larger cycloalkyl functional groups. Unlike the ethyl analog, the
branched cyclopropyl group in this conformation is able to make
van der Waals interactions with the N- and C-termini of the protein.
4.3. ADME properties
4.2. Structural biology
The ADME properties of analog 46 were also investigated and a
summary of key data are included in Table 6. Predicted clearance
from microsomes was moderate in human and rat with clearance at
38% and 29% of liver blood flow respectively. In mouse, clearance
was higher (71% of liver blood flow), although clearance from
mouse hepatocytes was slightly lower (56% of liver blood flow).
In vivo clearance of Compound 46 was also assessed in rat and
mouse PK through dosing at 1 mg/kg i.v. The measured clearance
values in vivo (rat 21 mL/min/kg and mouse 65 mL/min/kg) were in
good agreement with predicted in vitro clearance suggesting that
extrahepatic mechanisms were not major routes of clearance for
our compound. Compound 46 was also dosed at 5 mg/kg orally
using a 5% methylcellulose formulation. The compound was
absorbed well and this was reflected in the excellent bioavailability
values of 105% and 73% for rat and mouse respectively. These in vivo
PK parameters were also consistent with the high MDCK perme-
ability and plasma protein binding that were observed in vitro.
Taken in total, this was a very promising PK profile from which to
launch a lead optimization effort, where reducing clearance was
perhaps the most pressing concern.
We were able to obtain both TYK2 and JAK2 crystal structures of
compound 46 (Fig. 7a). A significant difference in these structures is
the shape of the P-loop. In TYK2, the P-loop is more closed down
whereas that of JAK2 is significantly more open. We believe this is
not due to crystallographic artifacts because this has recently been
observed by other investigators using different constructs as well
[28,40]. This region is also not close to any asymmetric subunits. As
a result of the exquisite shape of TYK2’s P-loop, the dichlorophenyl
ring is enclosed, fitting snugly in a pocket formed by the P-loop on
the N-terminus and a small cavity in the C-terminus where the
chlorine resides (Fig. 7b). The shape of the P-loop likely contributes
to the TYK2 selectivity over JAK2 for compound 46, and the small
cavity in the C-terminus explains the tight SAR observed at the 2-
position of the phenyl ring. On the other side of the molecule, the
cyclopropyl group also fits snugly in a region that approaches the
bulk solvent but does not pass the opening of the pocket (Fig. 7c).
The dihedral angles from C2 and C3 on the cyclopropyl ring to the
amide carbonyl oxygen are 40ꢂ and ꢃ30ꢂ, respectively. The center
of the cyclopropyl ring has a dihedral angle with the carbonyl ox-
ygen that is close to zero. This is the most frequently observed
conformation of the cyclopropyl-amide in the Cambridge Structural
Database [41] likely due to the special electronic characteristics of a
cyclopropane [42]. Its unique conformation and small size explain
the improved potency of 46 relative to analogs with isopropyl or
5. Conclusions
While a number of patent applications describing identification
of TYK2 inhibitors have published [43], this manuscript is the first
detailed publication, to our knowledge, disclosing efforts to identify
selective TYK2 inhibitors. Lead identification of HTS hit 1 proceeded
efficiently to afford a number of potent TYK2 inhibitors with
modest selectivity over JAK1 and JAK2. Cyclopropyl amide 46
demonstrated very good pharmacodynamic and pharmacokinetic
properties and was a promising lead from which to launch a lead
optimization effort. Efforts to improve potency and selectivity over
JAK1 and JAK2, while maintaining good ADME properties will be
the subject of a future publication.
Table 6
In vitro and in vivo ADME properties of compound 46.
In vitro ADME
In vivo rat and mouse PKe
Micromsomal Clhep (mL/min/kg) 8, 16, 64
(human, rat, mouse)a
Clhep (mL/min/kg)
21, 65
(rat, mouse)f
Hepatocyte Clhep (mL/min/kg)
7, 14, 51
All > 10
30
T1/2 (h) (rat, mouse) 1.2, 0.3
(human, rat, mouse)b
Cytochrome p450 inhibition
(3A4, 2C9, 2C19, 1A4, 2D6)c
MDCK permeability A:B Papp
(ꢁ10ꢃ6 cm/s)d
m
M
Vss (L/h) (rat, mouse) 1.1, 1.2
Cmax
(
mM)
2.5, 1.6
105, 73
(rat, mouse)g
Bioavailability (%)
(rat, mouse)
6. Experimental protocols
Plasma protein binding (%)
(human, rat, mouse)
95, 96, 96
6.1. General conditions
a
Predicted clearance from in vitro microsome preparations.
b
c
d
e
f
Predicted clearance from cryopreserved rat hepatocytes.
Unless otherwise indicated, all reagents and solvents were
purchased from commercial sources and used without further
purification. Moisture or oxygen sensitive reactions were con-
ducted under an atmosphere nitrogen gas. Unless otherwise stated,
1H spectra were recorded at room temperature using Varian Unity
IC50 for inhibition of cytochrome P450
.
Apparent permeability in MDCK transwell culture, A:B apical to basolateral.
Animals dosed as 1 mg/kg IV and 5 mg/kg PO.
Plasma clearance following iv dosing.
g
Maximal plasma concentration following po dosing.

184
J. Liang et al. / European Journal of Medicinal Chemistry 67 (2013) 175e187
Inova Bruker AVANCE III UltraShield-PlusTM Digital NMR spec-
trometer at indicated frequencies. Chemical shifts are expressed
in ppm relative to an internal standard: tetramethylsilane
(ppm ¼ 0.00). The following abbreviations are used: br ¼ broad
signal, s ¼ singlet, d ¼ doublet, dd ¼ doublet of doublets, t ¼ triplet,
q ¼ quartet, m ¼ multiplet. Purification by silica gel chromatog-
raphy was carried out using a CombiFlash by Teledyne ISCO sys-
tems with pre-packed cartridges. Purification by reverse phase
High Performance Liquid Chromatography (HPLC) was also used
and the conditions were described as indicated. All final com-
pounds were purified to >95% chemical purity, as assessed by HPLC
the reaction mixture was stirred for an additional 4 h. The mixture
was poured into the ice-water and the precipitate was collected to
afford the desired product (70 mg, 67% yield). 1H NMR (DMSO-d6,
400 MHz)
d
(2:1 mixture amide rotamers) 8.61 (dd, J ¼ 1.6, 4.8 Hz,
1H), 8.42 (dd, J ¼ 1.6, 4.8 Hz,1H), 7.62e7.29 (m, 4H), 7.19 (dd, J ¼ 4.8,
1.6 Hz, 1H), 3.40 (s, 2H), 3.17 (s, 1H). LCMS (ESI) m/z ¼ 280.7
[M þ H]þ.
6.2.5. 4-(5-(2,6-Dichlorophenyl)-1H-imidazol-2-yl)pyridine (26)
To a 10 mL round-bottom flask was added 2-bromo-1-(2,6-
dichlorophenyl)ethanone (110 mg, 0.41 mmol) and 1,4-dioxane
(1.0 mL). The white suspension was cooled to 0 ꢂC under nitrogen
and sodium carbonate (87 mg, 0.82 mmol) and iso-
nicotinimidamide (50 mg, 0.40 mmol) were added sequentially.
The reaction was heated to 100 ꢂC and stirring was continued for
two days when LCMS showed complete conversion. The mixture
was cooled to room temperature, diluted with H2O (10 mL) and
extracted with EtOAc (2 ꢁ 10 mL). Combined organics were dried
over Na2SO4, concentrated under reduced pressure and purified by
(column: Agilent SD-C18, 2.1 ꢁ 30 mm, 1.8
mm; mobile phases:
water with 0.5% TFA, CH3CN with 0.5% TFA; flow rate: 0.4 mL/min;
run time: 8.5 min; oven temperature: 40 ꢂC: UV detection: 254 and
210 nm).
6.2. Representative procedures
6.2.1. 2,6-Dichloro-N-(pyridin-4-yl)benzamide (1)
To a round-bottom flask was added 4-aminopyridine (0.24 g,
2.55 mmol), followed by N,N-dimethylformamide (10 mL) and
triethylamine (0.69 mL, 4.3 mmol). 2,6-dichlorobenzoyl chloride
(0.5 g, 2.4 mmol) was added dropwise and the reaction mixture was
stirred at 23 ꢂC for 4 h. The mixture was poured into ice-water. The
resulting precipitate was collected via filtration to give crude
product which was purified by rpHPLC (Gilson 281; column: ASB
rpHPLC (Gilson 215; column: Waters XBridge 30 ꢁ 100 mm, 10
mm;
mobile phases: CH3CN/water with 0.1% ammonia; flow rate: 60 mL/
min) to give desired product (0.021 g, 18% yield). 1H NMR (DMSO-
d6, 500 MHz)
d
13.21 (s, 1H), 8.66 (d, J ¼ 5.1 Hz, 2H), 7.88 (d,
J ¼ 5.1 Hz, 2H), 7.65e7.40 (m, 4H). LCMS (ESI) m/z ¼ 290.0 [M þ H]þ.
6.2.6. 4-(2-(2,6-Dichlorophenyl)-1H-imidazol-5-yl)pyridine (27)
The title compound was made (17% yield) by coupling 2,6-
dichlorobenenzenecarboximidamide and 2-bromo-1-(4-pyridinyl)
ethanone using the same procedure that was used to make com-
C18, 150 ꢁ 25 mm, 5
mm; mobile phases: CH3CN/aq. 0.01 M
ammonium bicarbonate solution; flow rate: 25 mL/min) to afford
the desired product (0.65 g, 95% yield). 1H NMR (DMSO-d6,
400 MHz)
d
11.21 (s, 1H), 8.47 (d, J ¼ 6.0 Hz, 2H), 7.63 (d, J ¼ 6.0 Hz,
pound 26. 1H NMR (DMSO-d6, 500 MHz)
d 12.78 (s, 1H), 8.51 (d,
2H), 7.57e7.47 (m, 3H). LCMS (ESI) m/z ¼ 267.1 [M þ H]þ.
J ¼ 5.0 Hz, 2H), 8.10 (s,1H), 7.75 (d, J ¼ 5.0 Hz, 2H), 7.64 (d, J ¼ 9.0 Hz,
2H), 7.57 (dd, J ¼ 9.0, 7.1 Hz, 1H). LCMS (ESI) m/z ¼ 290.1 [M þ H]þ.
6.2.2. 2,6-Dichloro-N-(pyridin-4-yl)benzenesulfonamide (22)
To a round-bottom flask was added 4-aminopyridine (0.15 g,
1.6 mmol), N,N-dimethylformamide (10 mL) and N-methylmorpho-
line (0.35 mL, 3.2 mmol) sequentially. 2,6-Dichlorophenylsulfonyl
chloride (0.39 g, 1.6 mmol) was added dropwise and the reaction
mixture was stirred at 23 ꢂC for 4 h. The reaction mixture was poured
into ice-water and was extracted with EtOAc (2 ꢁ 50 mL). Combined
organics were dried over sodium sulfate and evaporated to dryness.
The residue was purified by preparative TLC (10% MeOH/DCM) to
afford the desired product (73.2 mg, 15% yield). 1H NMR (DMSO-d6,
6.2.7. 4-(5-(2,6-Dichlorophenyl)-4H-1,2,4-triazol-3-yl)pyridine (28)
Step 1. To a 100 mL round-bottom flask was added hydrazine
(330
mL, 10 mmol), dichloromethane (20 mL) and triethylamine
(1.2 mL, 8.9 mmol). The resulting solution was cooled to ꢃ78 ꢂC and
2,6-dichlorobenzoyl chloride (1.7 g, 8.1 mmol) was added dropwise
under nitrogen. The reaction mixture was stirred at ꢃ78 ꢂC for one
more hour and slowly warmed to room temperature over 2 h. The
mixture was diluted with H2O (50 mL) and extracted with
dichloromethane (2 ꢁ 30 mL). The combined organic extracts were
dried over sodium sulfate and concentrated under reduced pres-
sure to give crude 2,6-dichlorobenzohydrazide which was directly
carried to the next step.
400 MHz)
d
12.35 (s, 1H), 7.99 (d, J ¼ 6.4 Hz, 2H), 7.49 (d, J ¼ 7.6 Hz,
2H), 7.39 (t, J ¼ 7.6 Hz, 1H), 6.85 (d, J ¼ 7.2 Hz, 2H). LCMS (ESI) m/
z ¼ 302.8 [M þ H]þ.
Step 2. A 100 mL round-bottom flask was charged with crude
2,6-dichlorobenzohydrazide, followed by 2-methoxyethanol
(10 mL), 4-pyridine carbonitrile (0.56 g, 5.4 mmol) and 25% so-
dium methoxide solution in methanol (0.56 mL, 2.7 mmol). The
resulting brown suspension was heated at 130 ꢂC under nitrogen
for three days when LCMS showed complete conversion. The
mixture was quenched with acetic acid (1 mL), diluted with 50 mL
saturated sodium chloride solution, and extracted with EtOAc
(2 ꢁ 50 mL). The combined organic extracts were dried over
Na2SO4, concentrated under reduced pressure, and purified by
rpHPLC to give desired product (0.46 g, 29% yield). 1H NMR (DMSO-
6.2.3. N-(2,6-Dichlorobenzyl)pyridin-4-amine (23)
2,6-Dichloro-N-(pyridin-4-yl)benzamide (0.5 g, 1.87 mmol) was
dissolved in anhydrous. THF (20 mL) and borane dimethyl sulfide
complex (0.4 mL, 6.8 mmol) was added. The reaction mixture was
heated to reflux under nitrogen overnight. The next day, the reac-
tion mixture was quenched by addition of methanol and extracted
with EtOAc (2 ꢁ 20 mL). The combined organics were dried over
sodium sulfate and evaporated to dryness. The residue was purified
by silica gel column chromatography (0e10% MeOH/DCM) to afford
the desired product (0.21 g, 44% yield). 1H NMR (DMSO-d6,
400 MHz)
d
8.07 (d, J ¼ 6.4 Hz, 2H), 7.58 (bt, J ¼ 4.4 Hz, 1H), 7.51 (d,
d6, 500 MHz)
d
8.72 (d, J ¼ 5.0 Hz, 2H), 7.98 (d, J ¼ 5.0 Hz, 2H), 7.72e
J ¼ 8.0 Hz, 2H), 7.39 (t, J ¼ 8.0 Hz, 1H), 6.76 (d, J ¼ 6.4 Hz, 2H), 4.49
7.60 (m, 3H). LCMS (ESI) m/z ¼ 291.0 [M þ H]þ.
(d, J ¼ 4.4 Hz, 2H). LCMS (ESI) m/z ¼ 252.9 [M þ H]þ.
6.2.8. N-(2-Bromopyridin-4-yl)-2,6-dichlorobenzamide (56)
6.2.4. 2,6-Dichloro-N-methyl-N-(pyridin-4-yl)benzamide (24)
To a suspension of NaH (45 mg, 1.11 mmol) in anhydrous N,N-
dimethylformamide (5 mL) was added 2,6-dichloro-N-(pyridin-4-
yl)benzamide (0.1 g, 0.37 mmol) at 0 ꢂC under nitrogen. After
30 min, methyl iodide (68 mg, 0.48 mmol) was added dropwise and
To a round-bottom flask was added sodium hydride (0.31 g,
7.82 mmol, 60% in mineral oil) and N,N-dimethylformamide
(15 mL). The resulting suspension was cooled to 0 ꢂC under ni-
trogen and then 2-bromopyridin-4-amine (0.74 g, 4.3 mmol) was
added. The mixture was stirred at 0 ꢂC for 20 min and then

J. Liang et al. / European Journal of Medicinal Chemistry 67 (2013) 175e187
185
2,6-dichlorobenzoyl chloride (0.82 g, 3.91 mmol) was added
dropwise. The reaction was warmed to room temperature, stirred
for 12 h, and then poured into ice-water (20 mL). The mixture was
extracted with ethyl acetate (2 ꢁ 50 mL). The combined organic
extracts were dried over Na2SO4 and concentrated under reduced
pressure to afford crude desired product (1.0 g, 74% yield) that
could be used without further purification. 1H NMR (DMSO-d6,
column: Phenomenx FA 150 ꢁ 30 mm; mobile phases: CH3CN/H2O
with 0.1% TFA; flow rate: 25 mL/min; run time: 15 min) to afford the
desired product (311 mg, 50% yield). 1H NMR (DMSO-d6, 400 MHz)
d
11.18 (s,1H),10.79 (s,1H), 8.40 (d, J ¼ 1.6 Hz,1H), 8.23 (d, J ¼ 5.6 Hz,
1H), 7.60e7.50 (m, 3H), 7.47 (dd, J ¼ 5.6, 1.6 Hz, 1H), 2.02e1.98 (m,
1H), 0.81e0.79 (m, 4H). LCMS (ESI) m/z ¼ 350.0 [M þ H]þ.
400 MHz):
d
11.41 (s, 1H), 8.33 (d, J ¼ 5.6 Hz, 1H), 7.99 (d,
Accession codes
J ¼ 1.6 Hz, 1H), 7.64e7.54 (m, 4H). LCMS (ESI) m/z ¼ 345.0
[M þ H]þ.
PDB nos.: 4GFO for 41 complexed with TYK2D1023N, 4GFM for
41 complexed with JAK2, 4GIH for 46 complexed with TYK2-
combo, 4GMY for 46 complexed with JAK2.
6.2.9. N-(2-Aminopyridin-4-yl)-2,6-dichlorobenzamide (29)
Method 1: N-(2-bromopyridin-4-yl)-2,6-dichlorobenzamide
(0.6 g, 1.7 mmol) was dissolved in an aq. ammonia solution
(20 mL) and the mixture was heated to 200 ꢂC in an autoclave for
16 h. After cooling to room temperature, the mixture was concen-
trated under reduced pressure. The residue was purified by silica
gel column chromatography (0e10% MeOH/DCM) to afford the
desired product (0.3 g, 61% yield). 1H NMR (DMSO-d6, 400 MHz)
Acknowledgment
We thank, Mengling Wong, Chris Hamman, Michael Hayes,
Baiwei Lin, Deven Wang, Yutao Jiang, and Yanzhou Liu for purifi-
cation and analytical support; Daniel Hascall, Grady Howes, Gigi
Yuen, and Garima Porwal for compound management support; Hoa
Le and Qin Yue for ADME support; Emile Plise and Jonathan Cheng
for MDCK data; Ning Liu for microsomal stability; Jasleen Sodhi for
reversible CYP inhibition measurement; Quynh Ho for plasma
protein binding measurement.
Diffraction data were collected at the Stanford Synchrotron
Radiation Lightsource, operated for the U.S. Department of Energy
Office of Science by Stanford University. The SSRL Structural Mo-
lecular Biology Program is supported by the DOE Office of Biological
and Environmental Research, National Institute of General Medical
Sciences (NIH) (including P41GM103393) and the National Center
for Research Resources (P41RR001209).
d
10.71 (s, 1H), 7.79 (d, J ¼ 5.6 Hz, 1H), 7.59e7.49 (m, 3H), 6.95 (s,
1H), 6.61 (d, J ¼ 5.6 Hz, 1H), 5.91 (s, 1H). LCMS (ESI) m/z ¼ 282.1
[M þ H]þ.
Method 2: A 20 mL microwave tube was charged with N-(2-
bromopyridin-4-yl)-2,6-dichlorobenzamide (346 mg, 1.00 mmol),
Pd2(dba)3 (91.6 mg, 0.1 mmol) and 2-dicyclohexylphosphino-20-
(N,N-dimethylamino)biphenyl (DavePhos, 78.7 mg, 0.2 mmol). The
tube was evacuated and back-filled with nitrogen three times
before 1,4-dioxane (6 mL) and lithium bis(trimethylsilyl)amide
(4.0 mL of 1.0 M solution in THF, 4.0 mmol) were added. The tube
was sealed and heated at 100 ꢂC for 5 h. The reaction mixture was
cooled to room temperature, diluted with 1 N HCl (5 mL), and
stirred for 10 min. The reaction mixture was adjusted to pH 12
using 1 N sodium hydroxide solution and then extracted with
dichloromethane (3 ꢁ 30 mL). The combined organic extracts were
dried over Na2SO4, and concentrated under reduced pressure. The
residue was purified by silica gel column chromatography (0e10%
MeOH/DCM) to afford the title compound (282 mg, 65% yield).
Appendix A. Supplementary data
Supplementary data related to this article can be found at http://
dx.doi.org/10.1016/j.ejmech.2013.03.070.
References
6.2.10. 2,6-Dichloro-N-(2-(pyrimidin-4-ylamino)pyridin-4-yl)
benzamide (37)
[1] (a) K. Ghoreschi, A. Laurence, J.J. O’Shea, Janus kinases in immune cell
signaling, Immunological Reviews 228 (2009) 273;
(b) W.J. Leonard, J.J. O’Shea, Jaks and STATs: biological implications, Annual
Review of Immunology 16 (1998) 293;
(c) K.D. Liu, S.L. Gaffen, M.A. Goldsmith, Jak/STAT signaling by cytokines,
Current Opinion in Immunology 10 (1998) 271.
A 20 mL microwave tube was charged with pyrimidin-4-amine
(220 mg, 2.3 mmol), N-(2-bromopyridin-4-yl)-2,6-dichloro benza-
mide (1.2 g, 0.35 mmol), Pd(OAc)2 (52 mg, 0.23 mmol), 4,5-
bis(diphenylphosphino)-9,9-dimethylxantene (Xantphos, 0.27 g,
0.46 mmol), cesium carbonate (1.5 g, 4.6 mmol) and 1,4-dioxane
(6 mL). The mixture was degassed with nitrogen. The tube was
sealed, and heated by microwave at 150 ꢂC for 20 min. The mixture
was purified by silica gel column chromatography (0e15% DCM/
MeOH) to afford the desired product (70 mg, 8.4% yield). 1H NMR
[2] J.N. Ihle, B.A. Witthuhn, F.W. Quelle, K. Yamamoto, O. Silvennoinen, Signaling
through the hematopoietic cytokine receptors, Annual Review of Immunology
13 (1995) 369.
[3] (a) M. Karaghiosoff, H. Neubauer, C. Lassnig, P. Kovarik, H. Schindler,
H. Pircher, B. McCoy, C. Bogdan, T. Decker, G. Brem, K. Pfeffer, M. Muller,
Partial impairment of cytokine responses in Tyk2-deficeint mice, Immunity 13
(2000) 549;
(b) K. Shimoda, K. Kato, K. Aoki, T. Matsuda, A. Miyamoto, M. Shibamori,
M. Yamashita, A. Numata, K. Takase, S. Kobayashi, S. Shibata, Y. Asano,
K. Sekiguchi, K. Nakayama, T. Nakayama, T. Okamura, S. Okamara, Y. Niho,
(DMSO-d6, 400 MHz)
d 11.17 (s, 1H), 10.23 (s, 1H), 8.67 (s, 1H), 8.39
(d, J ¼ 6.0 Hz, 1H), 8.21 (d, J ¼ 5.6 Hz, 1H), 8.04 (s, 1H), 7.72 (d,
J ¼ 6.0 Hz, 1H), 7.59e7.51 (m, 3H), 7.37 (dd, J ¼ 5.6, 1.6 Hz, 1H). LCMS
(ESI) m/z ¼ 360.0 [M þ H]þ.
K. Nakayama, Tyk2 plays a restricted role in IFN
a signaling, although it is
required for IL-12-mediated T cell function, Immunity 13 (2000) 561;
(c) M.H. Shaw, V. Boyartchuck, S. Wong, M. Karaghiosoff, J. Ragimbeau,
S. Pellegrini, M. Muller, W.F. Dietrich, G.S. Yap, A natural mutation in the Tyk2
pseudokinase domain underlies altered susceptibility of B10.Q/J mice to
infection and autoimmunity, Proceedings of the National Academy of Sciences
of the United States of America 100 (2003) 11594.
6.2.11. 2,6-Dichloro-N-(2-(cyclopropanecarboxamido)pyridin-4-yl)
benzamide (46)
[4] Y. Minegishi, M. Saito, T. Morio, K. Watanabe, K. Agematsu, S. Tsuchiya,
H. Takada, T. Hara, N. Kawamura, T. Ariga, H. Kaneko, N. Kondo, I. Tsuge,
A. Yachie, Y. Sakiyama, T. Iwata, F. Bessho, T. Ohishi, K. Joh, K. Imai, K. Kogawa,
M. Shinohara, M. Fujieda, H. Wakiguchi, S. Pasic, M. Abinun, H.D. Ochs,
E.D. Renner, A. Jansson, B.H. Belohradsky, A. Melin, N. Shimizu, S. Mizutani,
T. Miyawaki, S. Nonoyama, H. Karasuyama, Human tyrosine kinase 2 defi-
ciency reveals its requisite roles in multiple cytokine signals involved in
innate and acquired immunity, Immunity 25 (2006) 745.
N-(2-Aminopyridin-4-yl)-2,6-dichlorobenzamide
(0.5
g,
1.77 mmol) was dissolved in anhydrous pyridine (15 mL), and the
mixture was cooled to 0 ꢂC under nitrogen. Cyclopropanecarbonyl
chloride (0.24 g, 2.3 mmol) was added dropwise and the reaction
was warmed to room temperature. Stirring was continued for 4 h,
and then the reaction was quenched with ice water (50 mL) and
extracted dichloromethane (3 ꢁ 50 mL). The combined organic
extracts were dried over Na2SO4 and concentrated under reduced
pressure. The residue was purified by rpHPLC (Gilson GX 281;
[5] (a) D. Mucida, H. Cheroutre, The many face-lifts of CD4 T helper cells, Ad-
vances Immunology 107 (2010) 139;
(b) M.K. Gately, L.M. Renzetti, J. Magram, A.S. Stern, L. Adorini, U. Guber,
D.H. Presky, The interleukin-12/interleukin-12-receptor system: role in

186
J. Liang et al. / European Journal of Medicinal Chemistry 67 (2013) 175e187
normal and pathologic immune responses, Annual Review of Immunology 16 wide association study identifies IL23R as an inflammatory bowel disease
(1998) 495;
gene, Science 314 (2006) 1461.
(c) W.T. Watford, B.D. Hissong, J.H. Bream, Y. Kanno, L. Muul, J.J. O’Shea,
Signaling by IL-12 and IL-23 and the immunoregulatory roles of STAT4,
Immunological Reviews 202 (2004) 139;
[14] (a) N. Yawalkar, S. Karlen, R. Hunger, C.U. Brand, L.R. Braathen, Expression of
interleukin-12 is increased in psoriatic skin, Journal of Investigative Derma-
tology 111 (1998) 1053;
(d) K. Boniface, B. Blom, Y.-J. Liu, R.D.W. Maleft, From interleukin-23 to T-
helper 17 cells: human T-helper cell differentiation revisited, Immunological
Reviews 226 (2008) 132;
(e) C.A. Hunter, New IL-2 family members: IL-23 and IL-27, cytokines with
divergent functions, Nature Reviews Immunology 5 (2005) 521;
(f) C.L. Langrish, B.S. McKenzie, N.J. Wilson, R.D.W. Malefyt, R.A. Kastelein,
D.J. Cua, IL-12 and IL-23: master regulators of innate and adaptive immunity,
Immunological Reviews 202 (2004) 96;
(b) E. Lee, W.L. Trepiccio, J.L. Oestreicher, D. Pittman, F. Wang, F. Chamian,
M. Dhodapkar, J.G. Krueger, Increased expression of interleukin 23 p19 and
p40 in lesional skin of patients with psoriasis vulgaris, Journal of Experimental
Medicine 199 (2004) 125;
(c) G. Piskin, R.M.R. Sylva-Steenland, J.B. Bos, M.B.M. Teunissen, In vitro and in
situ expression of IL-23 by keratinocytes in healthy skin and psoriasis lesions:
enhanced expression in psoriatic skin, Journal of Immunology 176 (2006)
1908;
(g) G. Trinchieri, S. Pflanz, R.A. Kastelein, The IL-12 family of heterodimeric
cytokines: new players in the regulation of T cell responses, Immunity 19
(2003) 641.
(d) G. Monteleone, L. Biancone, R. Marasco, G. Morrone, O. Marasco, F. Luzza,
F. Pallone, Interleukin 12 is expressed and actively released by Crohn’s disease
intestinal lamina propria mononuclear cells, Gastroenterology 112 (1997)
1169;
(e) D. Berrebi, M. Besnard, G. Fromont-Hankard, R. Paris, J.F. Mougenot, P. De
Lagausie, D. Emilie, J.P. Cezard, J. Navarro, M. Peuchmaur, Interleukin-12
expression is focally enhanced in the gastric muscosa of pediatric patients
with Crohn’s disease, American Journal of Pathology 152 (1998) 667;
(f) V. Holtta, P. Klemetti, T. Sipponen, M. Westerholm-Ormio, G. Kociubinski,
H. Salo, L. Rasanen, K.-L. Kolho, M. Farkkila, E. Savilahti, O. Vaaraia, IL-23/IL-17
immunity as a hallmark of Crohn’s disease, Inflammatory Bowel Disease 14
(2008) 1175;
[6] (a) A.O. Chua, R. Chizzonite, B.B. Desai, T.P. Truitt, P. Nunes, L.J. Minetti,
R.R. Warrier, D.H. Presky, J.F. Levine, M.K. Gately, U. Gubler, Expression cloning
of a human IL-12 receptor component, Journal of Immunology 153 (1994)
128;
(b) D.H. Presky, H. Yang, L.J. Minetti, A.O. Chua, N. Nabavi, C.-Y. Wu,
M.K. Gately, U. Gubler,
A functional interleukin 12 receptor complex is
composed of two -type cytokine receptor subunits, Proceedings of the Na-
b
tional Academy of Sciences of the United States of America 93 (1996) 14002.
[7] C. Parham, M. Chirica, J. Timans, E. Vaisberg, M. Travis, J. Cheung, S. Pflanz,
R. Zhang, K. Singh, F. Vega, W. To, J. Wagner, A.-M. O’Farrell, T. McClanahan,
S. Zurawski, C. Hannum, D. Gorman, D. Rennick, R.A. Kastlein, R.D.W. Maleft,
K.W. Moore, A receptor for the heterodimeric cytokine IL-23 is composed of
(g) J. Pene, S. Chevalier, L. Preisser, E. Venereau, M.-H. Guilleux, S. Ghannam, J.-
P. Moles, Y. Danger, E. Ravon, S. Lesaux, H. Yssel, H. Gascan, Chronically
inflamed human tissues are infiltrated by highly differentiated Th17 lym-
phocytes, Journal of Immunology 180 (2008) 7423.
IL-12b1 and a novel cytokine receptor subunit, IL-23R, Journal of Immunology
168 (2002) 5699.
[15] (a) L. van der Fits, S. Mourits, J.S.A. Voerman, M. Kant, L. Boon, J.D. Laman,
F. Cornelissen, A.-M. Mus, E. Florencia, E. Prens, E. Lubberts, Imiquimod-
induced psoriasis-like skin inflammation in mice is mediated via the IL-23/IL-
17 axis, Journal of Immunology 182 (2009) 5836;
[8] (a) C.M. Bacon, E.F. Petricoin III, J.R. Ortaldo, R.C. Rees, A.C. Larner, J.A. Johnson,
J.J. O’Shea, Interleukin 12 induces tyrosine phosphorylation and activation of
STAT4 in human lymphocytes, Proceedings of the National Academy of Sci-
ences of the United States of America 92 (1995) 7307;
(b) H.-L. Ma, S. Liang, J. Li, L. Napierata, T. Brown, S. Benoit, M. Senices, D. Gill,
K. Dunussi-Joannopoulos, M. Collins, C. Nickerson-Nutter, L.A. Fouser,
D.A. Yong, IL-22 is required for Th17 cell-mediated pathology in a mouse
model of psoriasis-like skin inflammation, Journal of Clinical Investigation 118
(2008) 597;
(c) C.O. Elson, Y. Cong, C. Weaver, T.R. Schoeb, T.K. McClanahan, R.B. Fick,
R.A. Kastelein, Monoclonal anti-interleukin 23 reverses active colitis in a T
cell-mediated model in mice, Gastroenterology 132 (2007) 2359;
(d) S. Hue, P. Ahern, S. Buonocore, M.C. Kullberg, D.J. Cua, B.S. McKenzie,
F. Powrie, K.J. Maloy, Interleukin-23 drives innate and T cell-mediated
intestinal inflammation, Journal of Experimental Medicine 203 (2006)
2473;
(b) B. Oppmann, R. Lesley, B. Blom, J.C. Timans, Y. Xu, B. Hunte, F. Vega, N. Yu,
J. Wang, K. Singh, F. Zonin, E. Vaisberg, T. Churakova, M. Liu, D. Gorman,
J. Wagner, S. Zurawski, Y.-J. Liu, J.S. Abrams, K.W. Moore, D. Rennick,
R.D.W. Maleft, C. Hannum, J.F. Bazan, R.A. Kastelein, Novel p19 protein en-
gages IL-12p40 to form a cytokine, IL-23, with biological activities similar as
well as distinct from IL-12, Immunity 13 (2000) 715.
[9] M.M. Brierley, E.N. Fish, STATs: multifaceted regulators of transcription,
Journal of Interferon & Cytokine Research 23 (2005) 733.
[10] (a) W.E. Thierfelder, J.M. van Deursen, K. Yamamoto, R.A. Tripp, S.R. Sarawar,
R.T. Carson, M.Y. Sangster, D.A.A. Vignali, P.C. Doherty, G.C. Grosveld, J.N. Ihle,
Requirement for STAT4 in interleukin-12-mediated response of natural killer
and T cells, Nature 382 (1996) 171;
(e) M.C. Kullberg, D. Jankovic, C.G. Feng, S. Hue, P. Gorelick,
B.S. McKenzie, D.J. Cua, F. Powrie, A.W. Cheever, K.J. Maloy, A. Sher, IL-
23 plays a key role in Helicobbacter-induced T cell-dependent colitis,
Journal of Experimental Medicine 203 (2006) 2485;
(b) M.H. Kaplan, Y.-L. Sun, T. Hoey, M.J. Grusby, Impaired IL-12 responses and
enhanced development of Th2 cells in STAT4-deficient mice, Nature 382
(1996) 174.
[11] S. Aggarwal, N. Ghilardi, M.-H. Xie, F.J. de Sauvage, A.L. Gurney, Interleukin-23
promotes a distinct CD4 T cell activation state characterized by the production
of interleukin-17, Journal of Biological Chemistry 278 (2003) 1910.
[12] (a) P. Miossec, T. Korn, V.K. Kuchroo, Interleukin-17 and type 17 helper cells,
The New England Journal of Medicine 361 (2009) 888;
(f) H.H. Uhlig, B.S. McKenzie, S. Hue, C. Thompson, B. Joyce-Shaikh,
R. Stephankova, N. Robinson, S. Buonocore, H. Tlaskalova-Hogenova,
D.J. Cua, F. Powrie, Differential activity of IL-12 and IL-23 in mucosal
and systemic innate immune pathology, Immunity 25 (2006) 309;
(g) D. Yen, J. Cheung, H. Scheerens, F. Poulet, T. McClanahan, B.S. McKenzie,
M.A. Kleinschek, A. Owyang, J. Mattson, W. Blumenschein, E. Murphy,
M. Sathe, D.J. Cua, R.A. Kastelein, D. Rennick, IL-23 is essential for T cell-
mediated colitis and promotes inflammation via IL-17 and IL-6, Journal of
Clinical Investigation 116 (2006) 1310;
(b) J. Louten, K. Boniface, R.D.W. Malefyt, Development and function of Th17
cells in health and disease, Journal of Allergy Clinical Immunology 123 (2009)
1004;
(c) W. Ouyang, J.K. Kolls, Y. Zheng, The biological functions of T helper 17 cell
effector cytokines in inflammation, Immunity 28 (2008) 454;
(h) N.J. Davidson, S.A. Hudak, R.E. Lesley, S. Mennon, M.W. Leach,
(d) P.P. Ahem, A. Izcue, K.J. Maloy, F. Powrie, The interleukine-23 axis in in-
testinal inflammation, Immunological Reviews 226 (2008) 147.
[13] (a) J.H. Cho, The genetics and immunopathogenesis of inflammatory bowel
disease, Nature Reviews Immunology 8 (2008) 458;
D.M. Rennick, IL-12, but not IFNg, plays a major role in sustaining the
chronic phase of colitis in IL-10-deficient mice, Journal of Immunology 161
(1998) 3143.
[16] (a) J.M. Benson, C.W. Sachs, G. Treacy, H. Zhou, C.E. Pendley, C.M. Brodmerkel,
G. Shankar, M.A. Mascelli, Therapeutic targeting of IL-12/23 pathways: gen-
eration and characterization of ustekinumab, Nature Biotechnology 29 (2011)
615;
(b) L. Steinman, Mixed results with modulation of Th-17 cells in human
autoimmune diseases, Nature Immunology 11 (2010) 41.
[17] S.J. Rodig, M.A. Meraz, J.M. White, P.A. Lampe, J.K. Riley, C.A. Arthur, K.L. King,
K.C.F. Sheehan, L. Yin, D. Pennica, E.M. Johnson, R.D. Schreiber, Disruption of
the Jak1 gene demonstrates obligatory and nonredundant roles of the Jaks in
cytokine-induced biologic response, Cell 93 (1998) 373.
[18] (a) F. Candotti, S.A. Oakes, J.A. Johnson, S. Giliani, R.F. Schumacher, P. Mella,
M. Fioini, A.G. Ugazio, R. Badolato, L.D. Notarangelo, F. Bozzi, P. Macchi,
D. Strina, P. Vezzoni, R.M. Blaese, J.J. O’Shea, A. Villa, Structural and functional
basis for Jak3-deficient severe combined immunodeficiency, Blood 90 (1997)
3996;
(b) R.P. Nair, K.C. Duffin, C. Helms, J. Ding, P.E. Stuart, D. Goldgar,
J.E. Gudjonsson, Y. Li, T. Tejasvi, B.-J. Feng, A. Ruether, S. Schreiber,
M. Weichenthal, D. Gladman, P. Rahman, S.J. Schrodi, S. Prahalad, S.L. Guthery,
J. Fischer, W. Liao, P.-Y. Kwok, A. Menter, G.M. Lathrop, A. Carol, G.A. Wise,
A.B. Begovich, J.J. Voorhees, J.T. Elder, G.G. Krueger, A.M. Bowcock,
G.R. Abecasis, Genome-wide scan reveals association of psoriasis with IL-23
and NF-kB pathways, Nature Genetics 41 (2009) 199;
(c) M. Cargill, S.J. Schrodi, M. Chang, V.E. Garcia, R. Brandon, K.P. Callis,
N. Matsunami, K.G. Ardlie, D. Civello, J.J. Catanese, D.U. Leong, J.M. Panko,
L.B. McAllister, C.B. Hansen, J. Papenfuss, S.M. Prescott, T.J. White, M.F. Leppert,
G.G. Krueger, A.B. Begovich, A large-scale genetic association study confirmed
IL12b and leads to the identification of IL23R as psoriasis-risk gene, American
Journal of Human Genetics 80 (2007) 273;
(d) F. Capon, P. Di Meglio, N.J. Szaub, N.J. Prescott, C. Dunster, L. Baumber,
K. Timms, A. Gutin, V. Abkevic, A.D. Burden, J. Lanchbury, J.N. Barker,
R.C. Trembath, F.D. Nestle, Sequence variants in the genes for the interleukin-
23 receptor (IL23R) and its ligand (IL12b) confer protection against psoriasis,
Human Genetics 122 (2007) 201;
(b) S.M. Russell, N. Tayebi, H. Nakajima, M.C. Riedy, J. Roberts, M.J. Aman, T.-
S. Migone, M. Noguchi, M.L. Markert, R.H. Buckley, J.J. O’Shea, W.J. Leonard,
Mutation of Jak3 in a patient with SCID: essential role of Jak3 in lymphoid
development, Science 270 (1995) 797;
(e) R.J. Duerr, K.D. Taylor, S.R. Brant, J.D. Rioux, M.S. Silverberg, M.J. Daly,
A.H. Steinhart, C. Abraham, M. Regueiro, A. Griffiths, T. Dassopoulos, A. Bitton,
H. Yang, S. Targan, L.W. Datta, E.O. Kistner, L.P. Schumm, A.T. Lee,
P.K. Gregersen, M.M. Barmada, J.I. Rotter, D.L. Nicolae, J.H. Cho, A genome-
(c) P. Macchi, A. Villa, S. Giliani, M.G. Sacco, A. Frattini, F. Porta, A.G. Ugazio,
J.A. Johnson, F. Candotti, J.J. O’Shea, P. Vezzoni, L.D. Notarangelo, Mutations of
Jak-3 gene in patients with autosomal severe combined immune deficiency,
Nature 377 (1995) 65;

J. Liang et al. / European Journal of Medicinal Chemistry 67 (2013) 175e187
187
(d) D.C. Thomis, C.B. Gurniak, E. Tivol, A.H. Sharpe, L.J. Berg, Defects in B
lymphocyte maturation and T lymphocyte activation in mice lacking Jak3,
Science 270 (1995) 794;
(e) T. Nosaka, J.A. van Deursen, R.A. Tripp, W.E. Thierfelder, B.A. Witthuhn,
A.P. McMickie, P.C. Doherty, G.C. Grosveld, J.N. Ihle, Defective lymphoid
development in mice lacking Jak3, Science 270 (1995) 800;
[28] In vitro kinase activities of purified JAK1, JAK2, and TYK2 JH1 kinase domains
and the Ki values for ATP-competitive inhibition were determined as described
previously: V. Tsui, P. Gibbons, M. Ultsch, K. Mortara, C. Chang, W. Blair, R. Pulk,
R.M. Stanley, M. Starovasnik, D. Williams, M. Lamers, P. Leonard, S. Magnuson,
J. Liang, C. Eigenbrot, A new regulatory switch in a Jak protein kinase Proteins:
Structure, Function and Bioinformatics 79 (2011) 393.
(f) S.Y. Park, K. Saijo, T. Takahashi, M. Osawa, H. Arase, N. Hirayama, K. Miyake,
H. Nakauchi, T. Shirasawa, T. Saito, Developmental defects of lymphoid cells in
Jak3 kinase-deficient mice, Immunity 3 (1995) 371.
[29] TYK2 biochemical selectivity is expressed as the ratio of JAK1 (Ki)/TYK2 (Ki)
and JAK2 (Ki)/TYK2 (Ki). JAK3 Ki was also obtained for all analogs, but com-
pounds were universally less potent against this isoform and so JAK3 Ki data
were not included in the following SAR tables for the purpose of brevity.
[30] (a) Invitrogen, http://www.invitrogen.com/site/us/en/home.html?s_kwcid
¼TCj12295jinvitrogenjjSjej14171840346&kw¼slivgn (accessed 22.07.12).
(b) S.M. Rodems, B.D. Hamman, C. Lin, J. Zhao, S. Shah, D. Heidary,
L. Makings, J.H. Stack, B.A. Pollok, Assay and Drug Development Technol-
ogies 1 (2002) 9.
[19] a) M. Pesu, A. Laurence, N. Kishore, S.H. Zwiilich, G. Chan, J.J. O’Shea, Thera-
peutic targeting of Janus kinases, Immunological Reviews 223 (2008) 132;
(b) J.M. Kremer, B.J. Bloom, F.C. Breedveld, J.H. Coombs, M.P. Fletcher,
D. Gruben, S. Krishnaswami, R. Burgos-Vargas, B. Wilkinson, C.A. Zerbini,
S.H. Zwillich, The safety and efficacy of a JAK inhibitor in patients with cative
rheumatoid arthritis: results of a double-blind, placebo-controlled phase IIa
trial of three dosage levels of CP-690,550 versus placebo, Arthritis and
Rheumatism 60 (2009) 1895;
[31] The steady state kinetics mechanism of inhibition for compound
determined. See Supplemental data.
1 was
(c) M.G. Boy, C. Wang, B.E. Wilkinson, V.F.-S. Chow, A.T. Clucas, J.G. Krueger,
A.S. Gaweco, S.H. Zwillich, P.S. Changelian, G. Chan, Double-blind, placebo-
controlled, dose-escalation study to evaluate the pharmacologic effect of
CP-690,550 in patients with psoriasis, Journal of Investigative Dermatology
129 (2009) 2299.
[32] pKa values were calculated using the MoKa sofware(www.moldiscovery.com)
and a customized pKa model. (a) F. Milletti, L. Storchi, G. Sforna, G. Cruciani,
New and original pKa prediction method using grid molecular interaction
fields, Journal of Chemical Information and Modeling 47 (2007) 2172;
(b) F. Milletti, L. Storchi, L. Goracci, S. Bendels, B. Wagner, M. Kansy,
G. Cruciani, Extending pKa prediction accuracy: high-throughput pKa mea-
surements to understand pKa modulation of new chemical series, European
Journal of Medicinal Chemistry 45 (2010) 4270.
[20] (a) C. Haan, C. Rovlering, F. Rault, M. Kapp, P. Druckes, G. Thoma, I. Behrmann,
H.-G. Zerwes, Jak1 has a dominant role over JAK3 in signal transduction
through gc-containing cytokine receptors, Chemistry & Biology 18 (2011) 314;
(b) G. Thoma, F. Nuninger, F. Falchetto, E. Hermes, G.A. Tavares,
E. Vangrevelinghe, H.-G. Zerwes, Identification of a potent Janus kinase 3 in-
hibitor with high selectivity within the Janus kinase family, Journal of Medic-
inal Chemistry 54 (2011) 284.
[33] P. Bamborough, M.J. Brown, J.A. Christopher, C. Chung, G.W. Mellor, Selectivity
of kinase inhibitor fragments, Journal of Medicinal Chemistry 54 (2011) 5131.
and references cited therein.
[34] The structure of TYK2 with ADP was also determined and details can be found
in the Supplementary data.
[35] E.-X. Zhang, D.-X. Wang, Z.-T. Huang, M.-X. Wang, Synthesis of (NH)m(NMe)
4-m-bridged calix[4]pyridines and the effects of NH bridge on structure and
properties, Journal of Organic Chemistry 74 (2009) 8595.
[36] S. Lee, M. Jorgensen, J.F. Hartwig, Palladium-catalyzed synthesis of arylamines
from aryl halides and lithium bis(trimethylsilyl)amide as an ammonia
equivalent, Organic Letters 3 (2001) 2729.
[21] (a) H. Neubauer, A. Cumano, M. Muller, H. Wu, U. Huffstadt, K. Pfeffer, Jak2
deficiency defines an essential development checkpoint in definitive hema-
topoiesis, Cell 93 (1998) 397;
(b) E. Parganas, D. Wang, D. Stravopodis, D.J. Topharm, J.C. Marine, S. Teqlund,
E.F. Vanin, S. Bodner, O.R. Colamonici, J.M. van Deursen, G. Grosveld, J.N. Ihle,
Jak2 is essential for signaling through a variety of cytokine receptors, Cell 93
(1998) 385;
(c) B.A. Witthuhn, F.W. Quelle, O. Silvennoinen, T. Yi, B. Tang, O. Miura,
J.N. Ihle, Jak2 associates with the erythropoietin receptor and is tyrosine
phosphorylated and activated following stimulation with erythropoietin, Cell
74 (1993) 227.
[37] Isopropyl amide 43 was made through Buchwald coupling of bromine 56 and
2-methyl propanamide. Details can be found in the Supplementary data.
[38] Consistent with other members of this structural class, compound 46 was not
biochemically potent against JAK3 (Ki ¼ 253 nM).
[22] S.J. Sohn, K. Barrett, A. Van Abbema, C. Chang, P. Bir Kohli, H. Kanda, J. Smith, Y.
Lai, A. Zhou, B. Zhang, W. Yang, K. Williams, C. Macleod, C.A. Hurley, J.J.
Kulagowski, N. Lewin-Koh, H.S. Dengler, A.R. Johnson, N. Ghilardi, M. Zak, J.
Liang, W.S. Blair, S. Magnuson, L.C. Wu, A restricted role for TYK2 catalytic
activity in human cytokine responses, Journal of Immunology, in press.
[23] (a) M.E. Flanagan, T.A. Blumenkopf, W.H. Brissette, M.F. Brown, J.M. Casavant,
C. Shang-Poa, J.L. Doty, E.A. Elliott, M.B. Fisher, M. Hines, C. Kent, E.M. Kudlacz,
B.M. Littie, K.S. Magnusson, S.P. McCurdy, M.J. Munchhof, B.D. Perry,
P.S. Sawyer, T.J. Strelevitz, C. Subramanyam, J. Sun, D.A. Whipple,
P.S. Changelian, Discovery of CP-690,550: a potent and selective Janus kinase
(JAK) inhibitor for the treatment of autoimmune diseases and organ trans-
plant rejection, Journal of Medicinal Chemistry 53 (2010) 8468;
(b) A. Quintas-Cardama, K. Vaddi, P. Liu, T. Manshouri, J. Li, P.A. Scherle,
E. Caulder, X. Wen, Y. Li, P. Waeltz, M. Rupar, T. Burn, Y. Lo, J. Kelley,
M. Covington, S. Shepard, J.D. Rodgers, P. Haley, H. Kantarijian, J.S. Fridman,
S. Verstovsek, Preclinical characterization of the selective Jak1/2 inhibitor
INCB018424: therapeutic implications for the treatment of myeloproliferative
neoplasms, Blood 115 (2010) 3109;
[39] Z.A. Knight, K.M. Shokat, Features of selective kinase inhibitors, Chemistry &
Biology 12 (2005) 621.
[40] J.E. Chrencik, A. Patny, I.K. Leung, T.L. Emmons, T. Hall, R.A. Weinberg,
J.A. Gormley, J.M. Williams, J.E. Day, J.L. Hirsch, J.R. Kiefer, J.W. Leone,
H.D. Fischer, C.D. Sommers, H.-C. Huang, E.J. Jacobsen, R.E. Tenbrink,
A.G. Tomasselli, T.E. Benson, Structural and thermodynamic characterization
of the Tyk2 and Jak3 kinase domains in complex with CP-690550 and CMP-6,
Journal of Molecular Biology 400 (2010) 413.
[41] F.H. Allen, The Cambridge structural database: a quarter of a million crystal
structures and rising, Acta Crystallographica B 58 (2002) 380.
[42] (a) T. Clark, G.W. Spitznagel, R. Klose, P.V.R. Schleyer, Substituent effects on
the geometries and energies of cyclopropanes and the corresponding 2-
propyl derivatives, Journal of the American Chemical Society 106 (1984)
4412;
(b) R. Fuchs, J.H. Hallman, M.O. Perlman, Thermochemistry of conjugation of
simple cyclopropane derivatives, Canadian Journal of Chemistry 60 (1982)
1832.
[43] (a) B. Buchmann, K. Eis, U. Bothe, A. Bonin, U. Bomer, Carbonylamino-sub-
stituierte aniline-pyromidinderivative als Tyk-inhibitoren, deren Herstellung
und Verwendug als arzneimittel. DE2009001438 (2009);
(c) J.E. Thompson, R.M. Cubbon, R.T. Cummings, R.F. Wicker, P.M. Cunningham,
P.M. Cameron, P.T. Meinke, N. Liverton, Y. Weng, J.A. DeMartino, Photochemical
preparation of a pyridone containing tetracycle: a JAK protein kinase inhibitor,
Bioorganic & Medicinal Chemistry Letters 12 (2002) 1219.
(b) M.L. Boys, L.E. Burgess, R.D. Groneberg, D.M. Harvey, L. Huang, T. Kercher,
C.F. Kraser, E. Laird, E. Tarlton, Q. Zhao, 5,7-Substituted-imidazo[1,2-c]py-
rimidines as Inhibitors of JAK Kinases. WO2011130146 (2011);
(c) K. Ellard, J. Major, A. Jones, R. Lynch, N. Ramsden, Pyridine Compounds and
Aza Analogs Thereof as TYK2 Inhibitors (2012). WO2012062704;
(d) T. Blench, S. Goodacre, Y. Lai, J. Liang, S. Magnuson, V.H. Tsui, K. Williams,
B. Zhang, Pyrazolopyridines and Pyrazolopyridines and Their Use as Tyk2
Inhibitors (2012). WO2012066061;
(e) T. Blench, C. Ellwood, Y. Lai, J. Liang, C. MacLeod, S. Magnuson, V. Tsui,
K. Williams, B. Zhang, Azabenzothiazole Compounds, Compositions and
Methods of Use (2012). WO2012035039;
(f) Y. Lai, J. Liang, S. Magnuson, V.H. Tsui, B. Zhang, K.D. Robarge, Imidazo-
pyridine Compounds, Compositions and Methods of Use (2011).
WO2011113802;
[24] M. Koresawa, T. Okabe, High-throughput screening with quantitation of ATP
consumption: a universal non-radioisotope, homogeneous assay for protein
kinase, Assay and Drug Development Technologies 2 (2004) 153.
[25] Primary hits were defined as compounds giving > 50% inhibition at a test
concentration of 5 mM in the HTS assay.
[26] (a) C. Abad-Zapatero, J.T. Metz, Ligand efficiency indices as guideposts for drug
discovery, Drug Discovery Today 10 (2005) 464;
(b) A.L. Hopkins, C.R. Groom, A. Alex, Drug Discovery Today 9 (2004) 430;
(c) I.D. Kuntz, K. Chen, K.A. Sharp, P.A. Kollman, The maximal affinity of li-
gands, Proceedings of the National Academy of Sciences of the United States
of America 96 (1999) 9997.
[27] (a) M.C. Wenlock, R.P. Austin, P. Barton, A.M. Davis, P.D. Leeson, A comparison
of physiochemical property profiles of development and marketed oral drugs,
Journal of Medicinal Chemistry 46 (2003) 1250;
(g) S.C. Goodacre, Y. Lai, J. Liang, S. Magnuson, K.D. Robarge, M.S. Stanley,
V.H. Tsui, K. Williams, B. Zhang, A. Zhou, Janus Kinase Inhibitor Compounds
and Methods. WO2010142752 (2010);
(h) K. Ellard, N. Ramsden, Triazolopyridines as TYK2 Inhibitors.
WO2012000970 (2012).
(b) T.I. Opera, A.M. Davis, S.J. Teague, P.D. Leeson, Is there a difference between
leads and drugs? A historical perspective, Journal of Chemical Information and
Computer Sciences 41 (2001) 1308.