Lead optimization of a 4-aminopyridine benzamide scaffold to identify potent, selective, and orally bioavailable TYK2 inhibitors

Article abstract of DOI:10.1021/jm400266t

Herein we report our lead optimization effort to identify potent, selective, and orally bioavailable TYK2 inhibitors, starting with lead molecule 3. We used structure-based design to discover 2,6-dichloro-4-cyanophenyl and (1R,2R)-2-fluorocyclopropylamide modifications, each of which exhibited improved TYK2 potency and JAK1 and JAK2 selectivity relative to 3. Further optimization eventually led to compound 37 that showed good TYK2 enzyme and interleukin-12 (IL-12) cell potency, as well as acceptable cellular JAK1 and JAK2 selectivity and excellent oral exposure in mice. When tested in a mouse IL-12 PK/PD model, compound 37 showed statistically significant knockdown of cytokine interferon-γ (IFNγ), suggesting that selective inhibition of TYK2 kinase activity might be sufficient to block the IL-12 pathway in vivo.

Full text of DOI:10.1021/jm400266t

Article
pubs.acs.org/jmc
Lead Optimization of a 4‑Aminopyridine Benzamide Scaffold To
Identify Potent, Selective, and Orally Bioavailable TYK2 Inhibitors
Jun Liang,*,† Anne van Abbema,‡,● Mercedesz Balazs,∞ Kathy Barrett,‡ Leo Berezhkovsky,§
Wade Blair,‡,△ Christine Chang,‡ Donnie Delarosa,‡ Jason DeVoss,∞ Jim Driscoll,§,▽
Charles Eigenbrot,# Nico Ghilardi,∥ Paul Gibbons,† Jason Halladay,§ Adam Johnson,‡ Pawan Bir Kohli,‡
Yingjie Lai,† Yanzhou Liu,† Joseph Lyssikatos,† Priscilla Mantik,⊥ Kapil Menghrajani,§ Jeremy Murray,#
Ivan Peng,∞ Amy Sambrone,⊥ Steven Shia,# Young Shin,§ Jan Smith,‡ Sue Sohn,∥ Vickie Tsui,†
Mark Ultsch,# Lawren C. Wu,∥ Yisong Xiao,○ Wenqian Yang,× Judy Young,‡ Birong Zhang,†
Bing-yan Zhu,† and Steven Magnuson†
†
‡
§
Departments of Discovery Chemistry, Biochemical and Cellular Pharmacology, Drug Metabolism and Pharmacokinetics,
⊥
#
∥Immunology, Formulation, Structural Biology, and ∞Translational Immunology, Genentech, Inc., 1 DNA Way, South San
Francisco, California 94080, United States
×ChemPartner Co. Ltd., 998 Halei Road, Zhangjiang Hi-Tech Park, Shanghai, 201203, P. R. China
○WuXi AppTec Co., Ltd., 288 Fute Zhong Road, Waigaoqiao Free Trade Zone, Shanghai, 200131, P. R. China
S
* Supporting Information
ABSTRACT: Herein we report our lead optimization effort to
identify potent, selective, and orally bioavailable TYK2
inhibitors, starting with lead molecule 3. We used structure-
based design to discover 2,6-dichloro-4-cyanophenyl and
(1R,2R)-2-fluorocyclopropylamide modifications, each of which
exhibited improved TYK2 potency and JAK1 and JAK2
selectivity relative to 3. Further optimization eventually led to
compound 37 that showed good TYK2 enzyme and interleukin-
12 (IL-12) cell potency, as well as acceptable cellular JAK1 and
JAK2 selectivity and excellent oral exposure in mice. When
tested in a mouse IL-12 PK/PD model, compound 37 showed
statistically significant knockdown of cytokine interferon-γ
(IFNγ), suggesting that selective inhibition of TYK2 kinase
activity might be sufficient to block the IL-12 pathway in vivo.
antidrug antibodies.7 Therefore, significant unmet medical
needs still exist for both psoriasis and IBD.
Although clinically distinct, psoriasis and IBD were reported
to be associated in some patient populations and show
INTRODUCTION
■
Psoriasis and inflammatory bowel diseases (IBD), which
include Crohn’s disease and ulcerative colitis, are both chronic
inflammatory disorders of epithelial cells triggered by an overly
overlapping genetic predispositions, suggesting a shared
activated immune system.1,2 Psoriasis is a chronic skin disease
etiological basis.8 Extensive research confirmed that aberrant
characterized by red and scaly skin plaques, which is estimated
T cells and their associated extracellular signaling molecules
to affect ∼2% of the general population.3 In its moderate to
called cytokines play significant roles in the pathogenesis of
both diseases.9 Of particular importance are T helper type 1
severe form where topical application of corticosteroid is no
longer effective, psoriasis can present a substantial burden on
the well-being of patients.4 IBD is a chronic inflammation of
epithelial cells lining the gastrointestinal tract and colon that
affects approximately 1.4 million patients in United States
alone.5 Recent approvals of several antitumor necrosis factor
(anti-TNF) antibodies represent a major step forward in terms
of treatment options for IBD.6 However, it is reported that one-
third of IBD patients do not respond to anti-TNF therapy. For
those who do respond initially, loss of response occurs in 23−
46% of patients annually, most often because of development of
(Th1) and T helper type 17 (Th17) cells. Th1 cells produce
cytokine IFNγ, which in turn stimulates innate and T cell-
mediated immune response.10 Overstimulated Th1 cells,
however, cause tissue damage and result in chronic
inflammation. The more recently discovered Th17 cells,
which secrete cytokine IL-17, have been quickly recognized
as key players in many autoimmune diseases including psoriasis
Received: February 20, 2013
Published: May 14, 2013
© 2013 American Chemical Society
4521
dx.doi.org/10.1021/jm400266t | J. Med. Chem. 2013, 56, 4521−4536

Journal of Medicinal Chemistry
Article
and IBD.11 Th1 and Th17 cell differentiations are driven by IL-
12 and IL-23, respectively.12 Both IL-12 and IL-23 are
heterodimeric proteins that share a common subunit designated
p40. The receptor for IL-12 consists of two subunits IL-12Rβ1
and IL-12Rβ2,13 whereas the receptor for IL-23 shares the IL-
12Rβ1 subunit and has a unique subunit called IL-23R.14 The
IL-12 and IL-23 receptors lack intrinsic enzymatic activity.
Instead, IL-12Rβ1 is known to be constitutively associated with
the TYK2 enzyme,15 while IL-12Rβ2 and IL-23R are
constitutively associated with the JAK2 enzyme.16 Both TYK2
and JAK2 belong to the Janus family of tyrosine kinases (JAK)
which also includes JAK117 and JAK3.18
In multiple studies with psoriatic patients, significant
elevation of IL-12 and IL-23 was observed in lesional skin
compared to normal skin.19 Consistent with this observation,
IL-12 and IL-23 expression decreased after psoriatic treat-
ment.20 Similar results were observed in studies in animal
models of psoriasis. Transgenic mice that overexpressed the
p40 subunit or nontransgenic mice injected with recombinant
IL-23 developed inflammatory skin disease, showing compara-
ble phenotype to psoriatic patients.21 Inhibition of both IL-12
and IL-23 pathways by anti-p40 antibodies proved to be
therapeutically beneficial in both animal models,22 as well as in
patients.23 Ustekinumab, an anti-p40 antibody, was recently
approved by the FDA for treating psoriasis.24
and JAK2, respectively.33 Thus, cellular assays with pSTAT4
and pSTAT5 as proximal readouts of these respective pathways
were established to assess TYK2 and JAK2 inhibitions in cells.
Required JAK1 and JAK3 selectivity was more difficult to
define at the beginning of our program. JAK1 is essential for
signaling of a wide variety of physiologically important
cytokines such as IL-6, and its deficiency is perinatally lethal
in mice.34 JAK3 expression is mostly restricted to hema-
topoietic cells and is known to be exclusively associated with
the γ common (γc) receptor of type I cytokines.35 Inactivating
mutation or deletion of JAK3 gene in humans results in severe
combined immunodeficiency (SCID).36 Despite the deleterious
phenotypes of JAK1- and JAK3-deficient mice and humans,
JAK kinases inhibitors, most notably tofacitinib (1)37 and
ruxolitinib (2),38 are reported to be tolerated at efficacious
doses during clinical trials at concentrations that only partially
inhibit the activity of JAK1 and JAK3. In order to test our
hypothesis with a TYK2-selective inhibitor, we decided to
tentatively target the JAK1- and JAK3-selectivity criteria of at
least 10-fold as measured in cell-based assays as well.
Given the high degree of sequence homology between the
JAK family kinases, especially in the ATP binding site where
our ATP-competitive inhibitors reside, it was recognized that
achieving a high degree of TYK2 selectivity vs other JAKs
represented a significant challenge.39 To the best of our
knowledge, we are not aware of any published report of
selective TYK2 inhibitors except several recent patent
applications that appeared after completion of our work.40
Toward our goal of discovering potent and selective TYK2
inhibitors, we had identified, through a hit to lead campaign
following a high throughput screen, a 4-aminopyridine
benzamide lead 3 (Figure 1) with many attractive properties.41
In biochemical assays, compound 3 showed moderate potency
inhibiting the TYK2 enzyme, with modest biochemical
Human genetics studies, especially the recent genome-wide
association study, have identified a strong association of IL-23R
and IL-12B (encoding the p40 subunit) polymorphisms with
IBD.25 Once again, studies with transgenic mice provided
further evidence for the importance of the IL-12 and IL-23
pathways in IBD. In well-established T-cell-mediated colitis
models, p40 knockout mice were protected from development
of intestinal inflammation.26 Finally, the anti-p40 antibody
ustekinumab also showed indications of efficacy in certain
patients with Crohn’s disease.27
We were attracted to the opportunity of blocking the IL-12
and IL-23 signaling pathways in vivo by an orally bioavailable
small molecule that inhibits the kinase activity of TYK2 or
JAK2 or both. From the outset, however, we believed that
inhibition of the JAK2 kinase activity might be highly
problematic. JAK2 is associated with signaling of several
hematopoietic growth factors, including the erythropoietin
(EPO) receptor.28 JAK2 deficiency is embryonically lethal in
mice,29 and its inhibition leads to anemia in humans.30 A couple
of reports, however, suggested that inhibition of the TYK2
enzyme alone might be tolerated and sufficient to block IL-12
and IL-23 signaling in vivo. TYK2-deficient mice were viable
with no gross abnormal phenotype and showed severely
impaired signaling and biological response to IL-12 and IL-23
while displaying marked resistance to development of skin
inflammation and experimental colitis.31 A human patient
carrying a TYK2-inactivating mutation was also reported32 and
was clinically diagnosed with hyper-IgE syndromes (HIES).
Consistent with observations with the TYK2-deficient mice, the
human patient’s T cells showed almost complete defects in IL-
12 and IL-23 signaling, with apparently normal embryogenesis,
postnatal growth, and hematopoiesis. Therefore, we were
cautiously optimistic that inhibiting the TYK2 kinase activity,
while sparing JAK2, might be adequate to block IL-12 and IL-
23 signaling in vivo. On the basis of safety considerations, we
decided to target a JAK2 selectivity criterion of at least 10-fold
in cell-based assays. We have previously shown that the
predominant drivers of the IL-12 and EPO pathways are TYK2
Figure 1. Biochemical and cell-based assay data of 1, 2, and 3.
aBiochemical assays. bArithmetic mean of at least four separate runs (n
≥ 4). On average, the coefficients of variation were less than 0.3 times
c
the mean for biochemical assays. Cell-based assays. Arithmetic mean
of at least four separate runs (n ≥ 4). On average, the coefficients of
variation were less than 0.5 times the mean for cell-based assays. dCell-
e
based assay of TYK2 inhibition. Cell-based assay of JAK1 inhibition.
f Cell-based assay of JAK2 inhibition.
4522
dx.doi.org/10.1021/jm400266t | J. Med. Chem. 2013, 56, 4521−4536

Journal of Medicinal Chemistry
Article
selectivity against JAK2 and JAK1 and good selectivity against
JAK3.42
position of the phenyl ring might force the P-loop to move up44
so that the C4-substituent could interact with a very polar
region containing charged residues such as Arg1027. A second
region to potentially capitalize on is where the cyclo-
propylamide group was bound, for which there is a residue
difference between the TYK2 and JAK2 enzymes. In the TYK2
enzyme, this residue is a positively charged arginine (Arg901),
while it is a neutral glutamine (Gln853) in JAK2. Even though
both residues are on the protein surface and solvent-exposed,
we were eager to determine if this difference could be exploited
to improve the JAK2 selectivity of our analogues. Initial
modeling suggested that substitution at the C2-position of the
cyclopropylamide group could provide a vector to access this
region.
We started SAR exploration by first focusing on
modifications at the C4-position of the 2,6-dichlorophenyl
group in compound 3. In addition to the structural rationale
outlined above, incubation of compound 3 with human liver
microsomes showed formation of two major metabolites
resulting from hydrolysis of the cyclopropylamide and
oxidation on the 2,6-dichlorophenyl ring, respectively. Meta-
site45 (Molecular Discovery) analysis predicted that the phenyl
C4-position was likely the site of this oxidative metabolic
transformation. Therefore, introduction of a functional group,
preferably a polar and electron-withdrawing one, might be a
viable approach to block such metabolism while simultaneously
lowering the associated LogD. Previous structure−activity
relationship (SAR) of 4-aminopyridine benzamides from a
library of commercially available 2-chlorobenzoic acids had
demonstrated that small groups were tolerated at the C4-
position of the phenyl ring,41 so all indications pointed to this
being an important region to explore.
In cell-based assays,33 3 demonstrated reasonable potency in
blocking the IL-12 pathway (IL-12 pSTAT4 EC50 = 380 nM)
while displaying less activity in the EPO (JAK2) and IL-6
(JAK1) pathways. Even though the biochemical and cellular
potency as well as JAK1 and JAK2 selectivity required
improvement, compound 3 represented an important milestone
for our program, demonstrating that despite the high sequence
homology of ATP-binding pockets between TYK2 and other
JAK isoforms, reasonably selective TYK2 inhibitors could be
identified. In terms of physiochemical properties, lipophilicy for
3 was somewhat high with measured LogD of 3.4 at pH 7.4.43
We intended to improve potency and selectivity of our
analogues without further increase in lipophilicity. In mice,
compound 3 exhibited relatively high clearance (65 mL/min/
kg) when dosed intravenously (iv 1 mg/kg) and exhibited
modest oral exposure (AUC = 2.6 μM·h at po 5 mg/kg)
despite excellent cell permeability (MDCK A/B Papp = 30 ×
10−6 cm/s) and aqueous solubility (170 μg/mL at pH 7.4).
These results suggested that oral exposure in mice for
compound 3 might be limited by first-pass metabolism. Since
our IL-12 PK/PD experiments would be conducted in mice, it
was important to achieve good oral exposure in mice while
maintaining low predicted clearance in other species. Herein we
describe in detail our lead optimization efforts of compound 3.
RESULTS AND DISCUSSION
■
To enable design of more potent and selective TYK2 inhibitors,
we obtained cocrystal structures of lead molecule 3 with the
JH1 domain of the TYK2 and JAK2 enzymes at 2.0 and 2.4 Å
resolution, respectively (Figure 2).41 Significant difference in P-
Representative analogues from the C4-position SAR and
their biological data are listed in Table 1. Enzyme data were
used to calculate the biochemical JAK1 and JAK2 indexes,
defined as (JAK1 Ki)/(TYK2 Ki) and (JAK2 Ki)/(TYK2 Ki),
respectively. For analogues that showed TYK2 enzyme Ki < 5.0
nM, cellular inhibitory activities (EC50) of the IL-12, IL-6, and
EPO pathways using pSTAT4, pSTAT3, and pSTAT5
readouts, respectively, were also measured. The cell potency
data were then used to calculate the associated cell JAK1 and
JAK2 indexes, defined as (IL-6 pSTAT3 EC50)/(IL-12 pSTAT4
EC50) and (EPO pSTAT5 EC50)/(IL-12 pSTAT4 EC50),
respectively.
Among compounds with an H-bond donor directly
connected to the phenyl ring (compounds 4−7), aniline
analogue 4 was found to be the most potent TYK2 inhibitor,
with slightly better JAK1 and JAK2 indexes than 3. Phenol 7
was also well-tolerated. When the phenol was converted to a
methyl ether (compound 8), a slight loss of potency was
observed. Given the potential liabilities of both phenol and
aniline, we investigated whether benzyl alcohols or benzyl-
amines, with the H-bond donor removed one atom from the
ring, would be suitable replacements. Primary, secondary,
tertiary, and propargylic alcohols (compounds 9−12) were
introduced at the C4-position. None, however, was as potent as
the phenol analogue 7. We also systematically examined
primary, secondary, and tertiary benzylamines (compounds
13−16). Likely protonated at physiological pH, the amino
groups could not only serve as H-bond donors but also result in
significantly lowered lipophilicity. Disappointingly, biochemical
data showed that all the benzylamino analogues were less
potent than the aniline analogue 4.
Figure 2. Overlay of the TYK2 and JAK2 crystal structures with
compound 3 (TYK2 sandy brown, JAK2 blue), highlighting the
different P-loop conformations between the TYK2 and JAK2 enzymes
and the residue difference (Arg901 vs Gln853) in the area where
cyclopropylamide group was bound. The resolutions of the X-ray
structures are 2.0 Å (TYK2) and 2.4 Å (JAK2), respectively.
loop conformation was noted between the two enzymes: in the
crystal structure of 3 with the TYK2 enzyme, the P-loop
reached down toward the C-terminal lobe, and as a result, the
4-position of the phenyl group is only 3.6 Å from the backbone
carbonyl of Glu905 on the tip of the P-loop. On the contrary,
the P-loop in the JAK2 enzyme was significantly more “open”.
We envisioned that introduction of hydrogen bond (H-bond)
donor at the C4-position on phenyl might lead to favorable and
specific H-bond interaction with the carbonyl of Glu905 on the
TYK2 P-loop, resulting in improved potency and possibly
selectivity. It also seemed plausible that substitution at the C4-
4523
dx.doi.org/10.1021/jm400266t | J. Med. Chem. 2013, 56, 4521−4536

Journal of Medicinal Chemistry
Article
Table 1. SAR of Modifications at C4-Position on the Phenyl Ring
a
Biochemical assays. Arithmetic mean of at least three separate runs (n ≥ 3). On average, the coefficients of variation were less than 0.3 times the
b
c
d
mean for biochemical assays. (JAK1 Ki)/(TYK2 Ki). (JAK2 Ki)/(TYK2 Ki). Cell-based assay of TYK2 inhibition. Arithmetic mean of at least
three separate runs (n ≥ 3). On average, the coefficients of variation were less than 0.5 times the mean for cell-based assays. (IL-6 pSTAT3 EC50)/
(IL-12 pSTAT4 EC50). (EPO pSTAT5 EC50)/(IL-12 pSTAT4 EC50).
e
f
Among other functional groups surveyed at the C4-position,
nitrile 19 stood out because of its improved TYK2 Ki (1.8 vs
4.8 nM for 3) and slightly better biochemical JAK1 and JAK2
indexes. Nitrile isostere-acetylene 20 was approximately 4-fold
less potent against TYK2 than nitrile 19. Heterocycles, in the
form of pyrazole and triazole, were also evaluated at the C4-
position. Pyrazole 21 was tolerated, although with poor JAK1
and JAK2 selectivities as measured in both biochemical and
cellular assays. On the other hand, loss of TYK2 inhibitory
activity was noted for triazole 22. Additionally, aqueous
solubility at pH 7.4 for both 21 and 22 was poor (1 and 2
μg/mL, respectively), possibly because of the addition of one
more aromatic ring.46 Further exploration of heterocycles at the
C4-position was not pursued.
The SAR described above is consistent with the design
hypothesis that an H-bond donor such as a primary amino
group in compound 4 could have a favorable interaction with
the TYK2 enzyme. What we had not predicted, however, was
4524
dx.doi.org/10.1021/jm400266t | J. Med. Chem. 2013, 56, 4521−4536

Journal of Medicinal Chemistry
Article
that compound 19 with a C4-nitrile, typically an H-bond
acceptor, would also show improved TYK2 potency. To
understand the interaction of nitrile 19 with the TYK2 enzyme,
we obtained a cocrystal structure at 2.3 Å resolution (Figure 3).
interaction is likely absent with a more “open” JAK2 P-loop.
This difference in P-loop conformations between the TYK2
and JAK2 enzymes had been reported previously with other
classes of JAK inhibitors.49
Despite its better biochemical potency, aniline 4 was less
potent in the IL-12 cell assay than nitrile 19 (pSTAT4 EC50 of
248 vs 157 nM). It was encouraging that for both analogues 4
and 19, the slightly improved biochemical JAK1 and JAK2
indexes translated into slightly better cellular JAK1 and JAK2
indexes than lead molecule 3. It is also worth pointing out that
good cellular JAK1 selectivity was more difficult to achieve than
cellular JAK2 selectivity. We suggest this observation is a
reflection of the weaker ATP Km for JAK1 compared to JAK2
(55 vs 15 μM).50 Compound 19 was profiled against a broad
panel of kinases at 180 nM (100-fold in excess of its TYK2 Ki)
and was found to be reasonably selective. The Gini coefficient
of its kinase selectivity was calculated to be 0.65.51 Among the
187 kinases tested, compound 19 demonstrated >50%
inhibition against only six kinases: 80% inhibition against
RSK3 and RSK4, 73% against RSK2, 68% against LRRK, and
61% against CDK5 and LCK, respectively.
After determining that a nitrile was the optimal substituent at
the C4-position on the 2,6-dichlorophenyl fragment, we then
turned our attention to optimization of the cyclopropylamide
moiety. As highlighted in Figure 2, there is a residue difference
between TYK2 (Arg901) and JAK2 (Gln853) that we hoped to
probe through substitution at the C2-position on the
cyclopropylamide group. For any given C2-substituent on the
cyclopropylamide, cis- and trans-isomers are possible, with each
mixture consisting of a pair of enantiomers. Since the preferred
stereochemistry of this substituent could not be predicted with
confidence from modeling, we proceeded by preparing and
testing both the cis- and trans-isomers, each as a racemic
mixture. If the initial enzyme inhibition data for the racemic
mixture were promising, the cis- or trans-mixture was subject to
further purification by chiral HPLC or SFC to give single
enantiomers for testing. A general trend quickly emerged from
our SAR exploration, where for every C2-substituent evaluated,
the racemic cis-isomeric mixture was always more potent than
the corresponding trans-isomeric mixture. Furthermore, when
the racemic cis-isomers were separated, there was usually a
significant difference in the TYK2 potency between the two
enantiomers. To illustrate this point, data for four diaster-
eomers of the C2-fluoro analogue (compound 23−26) are
shown in Table 2. For other C2-substituents, data for only the
more active cis-enantiomer are included.
When compared to the unsubstituted cyclopropylamide 3,
cis-C2-fluoro analogue 23 showed a small but statistically
significant improvement in TYK2 enzyme potency (2.5 vs 4.8
nM) and slightly better biochemical JAK1 and JAK2 indexes.
The TYK2 enzyme pocket where the cyclopropylamide was
bound seemed sensitive to steric bulk. As the size of the C2-
substituent increased from F to Cl to CH3 (compounds 23, 27,
and 28), the TYK2 potency gradually decreased. Polar
substituents at the C2-position were also examined in hopes
of achieving favorable interactions with the polar side chain of
Arg901 in the TYK2 enzyme. We expected that polar
substituents would also offer an opportunity to reduce the
lipophilicity of compound 3. Contrary to expectations, most of
the polar C2-substituents (29−31) resulted in reduced TYK2
potency when compared to compound 3. An exception to this
trend was one of the cis-enantiomers of carboxylic acid
(analogue 32). It was especially interesting to note that
Figure 3. X-ray crystal structure of compound 19 in complex with the
TYK2 enzyme. The resolution of the X-ray structure is 2.3 Å.
The distance between the nitrile nitrogen and the Arg1027 NH
is 3.5 Å, slightly outside the range of a reasonable H-bond but
not excluding the possibility of a long-range electrostatic
interaction. Instead, what we noticed was that the P-loop
conformation was almost identical to that observed in the
TYK2 cocrystal structure with compound 3. The carbonyl
oxygen of Glu905 on the P-loop is 3.2 Å away from the carbon
of the nitrile group and likely makes an orthogonal multipolar
interaction. This less conventional type of interaction has been
observed in the Cambridge small-molecule database (CSD)47
and is visually identified by Proasis48 (Desert Scientific
Software) to be a favorable interaction. We supplemented
these empirical observations using ab initio calculations of a
model system, restraining the distance and angles between the
carbonyl and the nitrile carbon atom, and also found this to be
an energetically favorable interaction (details in the Exper-
imental Section). One may argue that the potency boost
stemmed instead from a nonclassical H-bond between C5-
hydrogen of the phenyl and the backbone carbonyl of Glu905,
but even though this C5···O distance is 3.4 Å, the angle was not
ideal. From an SAR standpoint, the carbon of the C4-nitrile and
the nitrogen of the C4-aniline occupied similar space. Both
substitutions provided potency gain relative to the C4-
hydrogen, with the C4-aniline being approximately 2-fold
more potent than the C4-nitrile, supporting our interpretation
that the CO···CN is a favorable polar interaction but is not as
strong as an H-bond.
It is remarkable that the P-loop maintained its “closed down”
conformation in the cocrystal structures with compounds 3 and
19, leading us to hypothesize that the TYK2 P-loop is relatively
rigid. We believe that the observed TYK2 potency SAR in
Table 1 could be rationalized by the “inflexible” P-loop
hypothesis. Larger substitutions at the C4-position were
tolerated but did not lead to improved potency as designed
likely because any favorable interactions to the TYK2 enzyme
that such groups facilitated were offset by the energy cost of
moving the P-loop. By contrast, aniline 4 could be readily
modeled using the TYK2 bound crystal structure of compound
3:41 the primary amino group could be accommodated at the 4-
position with no movement, and thus no entropic cost, of the
P-loop while making a favorable H-bond with the carbonyl of
Glu905. This hypothesis could also explain the improved JAK2
index for analogues 4 and 19 as the result of better TYK2
enzyme potency, due to favorable H-bond or orthogonal
multipolar interaction with the adjacent P-loop in TYK2. Such
4525
dx.doi.org/10.1021/jm400266t | J. Med. Chem. 2013, 56, 4521−4536

Journal of Medicinal Chemistry
Article
charged carboxylate in analogue 32 might partially explain the
excellent JAK2 selectivity observed (Figure 4a). This
hypothesis, however, does not explain the high degree of
JAK1 selectivity for compound 32. JAK1 and TYK2 share the
same residue (arginine) at this position, implying that there
might be other significant differences in structure or dynamics
between the two enzymes in this area.
Table 2. SAR of C2-Substituted Cyclopropylamide
As a follow-up to this interesting discovery, we prepared the
cis enantiomers of the corresponding C2-tetrazole and triazole.
The tetrazole 33 was much less selective, although slightly more
potent, than the carboxylic acid analogue 32. Replacement of
the acid with triazole (34), however, resulted in a 5-fold drop in
potency against the TYK2 enzyme. This result is consistent
with our hypothesis that the good TYK2 potency observed for
acid 32 and tetrazole 33 is partially derived from favorable
electrostatic interactions with Arg901. Despite their excellent
TYK2 enzyme potencies, acid 32 and tetrazole 33 were both
inactive in the TYK2 cell assay (IL-12 pSTAT4 EC50 > 5.0
μM). Both the acid and tetrazole, with calculated pKa of 3.7 and
4.6, respectively, should be ionized at physiological pH and
likely have poor passive cell permeability.
The lack of cell potency for acid 32 and tetrazole 33
underscored the importance of identifying a C2-substituent that
was not only potent in biochemical assays but was also
cellularly permeable. Fortunately, the C2-fluoro analogue 23
showed excellent cell potency (IL-12 pSTAT4 EC50 = 193
nM). Compound 23 displayed cell JAK1 and JAK2 indexes of
17× and 18×, respectively, meeting our criteria for cellular
JAK1 and JAK2 selectivity. Excited by its cell potency and
selectivity, we decided to use this particular cis-2-fluorocyclo-
propylamide for future SAR investigation. To define its
stereochemistry, we obtained a single crystal X-ray structure
of the compound 23 (see Supporting Information) and found
that this molecule possessed (1R,2R) stereochemistry. We also
solved the crystal structure of compound 23 in complex with
the TYK2 enzyme at 2.4 Å resolution (Figure 4b). In this
cocrystal structure, the fluorine atom is 3.1 Å away from the
terminal amino of Arg901. The electronegative fluorine might
be slightly favored at the C2-position because of proximity to
the positively charged arginine. Since fluorine atoms usually do
not act as H-bond acceptors within protein−ligand com-
plexes,52 we propose that the favorable dipolar interaction
between the C2-fluorine and Arg901 is responsible for the small
TYK2 potency improvement of analogue 23. It was also
noteworthy that the distance between the fluorine atom and the
a
Unless otherwise indicated, compounds are single enantiomers of
b
unknown absolute configuration. Biochemical assays. Arithmetic
mean of at least three separate runs (n ≥ 3). On average, the
coefficients of variation were less than 0.3 times the mean for
c
d
biochemical assays. (JAK1 Ki)/(TYK2 Ki). (JAK2 Ki)/(TYK2 Ki).
compound 32 showed remarkably high biochemical JAK1 and
JAK2 indexes. A single crystal structure of compound 32 was
obtained, and its absolute stereochemistry was determined to
be 1R,2S (see Supporting Information). On the basis of the
residue difference between the TYK2 and JAK2 enzyme in this
area of the binding pocket (Arg901 vs Gln853), we propose
that the electrostatic interaction between the positively charged
arginine (Arg901) in the TYK2 enzyme and the negatively
Figure 4. (a) Modeling of acid 32 binding with TYK2 enzyme. The C2-carboxylic acid is proposed to have favorable electrostatic interaction with the
positively charged arginine (Arg901) in the TYK2 enzyme. (b) X-ray crystal structure of compound 23 in complex with the TYK2 enzyme. The
resolution of the X-ray structure is 2.4 Å.
4526
dx.doi.org/10.1021/jm400266t | J. Med. Chem. 2013, 56, 4521−4536

Journal of Medicinal Chemistry
Article
Table 3. SAR of Combining (1R,2R)-2-F-Cyclopropylamide with Various Benzamides, in Comparison to Ruxolitinib (2)
a
b
c
d
f
g
TYK2 Ki
(nM)
JAK1 index
JAK2 index
JAK3 index
IL-12 pSTAT4
e
JAK1 index
(cell)
JAK2 index
(cell)
compd R1
R2
(biochemical)
(biochemical)
(biochemical)
EC50 (nM)
35
36
37
2
Cl
F
CN
H
1.4
1.6
1.6
0.5
69×
76×
60×
0.4×
33×
43×
39×
0.2×
178×
194×
239×
6×
145
211
224
10
16×
24×
10×
2×
20×
22×
21×
1×
F
CN
a
Biochemical assays. Arithmetic mean of at least three separate runs (n ≥ 3). On average, the coefficients of variation were less than 0.3 times the
b
c
d
e
mean for biochemical assays. (JAK1 Ki)/(TYK2 Ki). (JAK2 Ki)/(TYK2 Ki). (JAK3 Ki)/(TYK2 Ki). Cell-based assay of TYK2 inhibition.
Arithmetic mean of at least three separate runs (n ≥ 3). On average, the coefficients of variation were less than 0.5 times the mean for cell-based
assays. (IL-6 pSTAT3 EC50)/(IL-12 pSTAT4 EC50). (EPO pSTAT5 EC50)/(IL-12 pSTAT4 EC50).
f
g
methyl carbon of Leu903 is 3.2 Å, which is approximately the
sum of van der Waals radii of both atoms. This compact
packing between the C2-fluorine and the methyl group of
Leu903 might explain why other groups with larger van der
Waals radii, such as chloride or methyl, are not as well-tolerated
at this position.
The next logical step in the SAR progression was to combine
the (1R,2R)-2-F-cyclopropylamide moiety with our preferred
2,6-dichloro-4-cyanophenyl group (Table 3). We were pleased
to find that the potency and selectivity improvement attributed
to these two substitutions seemed to be “additive”: analogue 35
was more potent and more selective than either of the
predecessor molecules (19 or 23). To better understand the
origins of the improved potency and selectivity of analogue 35,
we obtained a crystal structure of the inhibitor in complex with
the TYK2 enzyme at 2.5 Å resolution (Figure 5). In this crystal
the SAR observed for compound 35 could be attributed to the
presence of both favorable interactions from the fluorine atom
as well as the nitrile group with the TYK2 enzyme.
In cell assays, compound 35 was found to be a potent
inhibitor of the IL-12 pathway (pSTAT4 EC50 = 145 nM) and
displayed excellent cell JAK1 and JAK2 indexes of 16× and
20×, respectively. In order to capitalize on this exciting
discovery, we expanded the SAR by preparing analogues with 2-
chloro-6-fluorophenyl substitution. In our previous publica-
tion,41 comparable TYK2 potency was observed for 2-chloro-6-
fluorophenyl analogue and the corresponding 2,6-dichloro-
phenyl analogue. We reasoned that replacement of chlorine
with fluorine could lower the lipophilicity of our molecules.
When the 2-chloro-6-fluorophenyl analogue 36 was prepared
and tested in biochemical assays, we were encouraged that it
was slightly more active and more selective than analogue 23.
Therefore, compound 37, with a C4-nitrile on the phenyl ring,
was synthesized. Compound 37 was equipotent to 36, with
comparable JAK1 and JAK2 indexes. In cell-based assays,
compound 37 showed IL-12 pSTAT4 EC50 of 224 nM, with its
cell JAK1 and JAK2 indexes meeting our selectivity criteria.
Furthermore, we tested compound 37 in both human
peripheral blood mononuclear cells (PBMC) and human
whole blood, stimulated with IL-12 in combination with IL-
18 and using downstream IFNγ production as a readout.33 In
the PBMC assay, compound 37 showed an EC50 of 168 nM,
comparable to its IL-12 pSTAT4 EC50 of 224 nM. In human
whole blood, an EC50 of 737 nM was obtained for compound
37.53 Compounds 35, 36, and 37 were also profiled against a
broad panel of kinases at concentrations 100 times their
corresponding TYK2 Ki values, resulting in Gini coefficients of
0.68, 0.59, and 0.61, respectively. The kinases against which the
three analogues showed >50% inhibition were identical to those
identified for compound 19.
Figure 5. X-ray crystal structure of compound 35 in complex with the
TYK2 enzyme. The resolution of the X-ray structure is 2.5 Å.
The physiochemical and in vitro ADME properties of key
analogues described in this work are summarized in Table 4,
along with data of the initial lead 3 for comparison. Compound
19, which incorporated the 4-nitrile on 2,6-dichlorophenyl ring,
was slightly less ligand efficient than 3 (LE of 0.49 vs 0.51), but
it was also slightly less lipophilic (LogD of 3.2 vs 3.4). As a
result, the calculated lipophilic ligand efficiency (LLE) was
higher for 19 than for 3 (5.5 vs 4.9). It was interesting that
structure, the fluorine atom is located 2.8 Å away from Arg901,
possibly engaged in the same favorable dipolar interaction as
observed in the previously described crystal structure of 23 with
the TYK2 enzyme. In addition, the carbon atom of the C4-
nitrile and the carbonyl oxygen of Glu905 are almost
orthogonal to each other with a measured distance of 3.2 Å.
This positioning is well within the favorable angle and distance
range of an orthogonal multipolar interaction.47 We believe that
4527
dx.doi.org/10.1021/jm400266t | J. Med. Chem. 2013, 56, 4521−4536

Journal of Medicinal Chemistry
Article
Table 4. Physiochemical and in Vitro DMPK Properties of Key Analogues
physiochemical properties
aqueous solubility
in vitro DMPK properties
plasma protein binding in
human, rat, mouse (%)
a
b
c
d
e
compd LE
LLE log D
tPSA
(μg/mL)
MDCK A−B Papp (×10−6 cm/s) (ratio (B−A)/(A−B))
3
0.51
0.49
0.50
0.48
0.51
0.47
4.9
5.5
5.7
6.0
ND
6.3
3.4
3.2
2.9
2.8
ND
2.5
71
94
71
94
71
94
170
64
30 (0.4)
15 (1.0)
20 (0.8)
12 (1.5)
30 (0.4)
95.1, 95.9, 95.6
97.8, 94.1, 96.8
97.5, 95.2, 98.2
93.2, 91.8, 98.3
93.4, 87.8, 96.4
86.6, 91.0, 95.0
19
23
35
36
37
126
74
28
53
11 (1.5)
a
b
c
Ligand efficiency = (−1.4 log Ki)/(n heavy atoms). Ligand−lipophilicity efficiency = (−log Ki) − cLogP. Measured log of distribution coefficient
d
e
between octanol and aqueous pH 7.4 buffer. Aqueous solubility was determined in pH 7.4 PBS buffer. Apparent permeability in MDCK Transwell
culture. A−B: apical-to-basolateral. Ratio is determined by the apparent permeability of B−A basolateral-to-apical divided by A−B apical-to-
basolateral.
Table 5. In Vitro Metabolic Stability and in Vivo PK Parameters of Key Analogues
in vivo PK
in vitro metabolic
a
stability, LM
b
b
c
(mL/min/kg)
rat
mouse
high dose mouse
AUC
Clp
Clp
compd human rat mouse
(mL/min/kg)
t1/2 (h) Vd (L/kg) F (%)
(mL/min/kg)
t1/2 (h) Vd (L/kg) F (%)
(μM·h)
Cmax (μM)
3
7
8
16
12
21
10
20
11
64
38
55
36
64
34
21
1.2
4.2
1.8
4.1
0.5
40
1.1
0.8
1.0
1.0
1.2
2.3
100
86
65
13
0.3
1.0
0.6
0.9
73
19
23
35
36
37
3.1
15
100
818
50
119
47
11
11
4
100
100
100
70
4.6
39
1090
143
8
1.0
917
108
a
b
Hepatic clearance predicted from liver microsome stability. Male Sprague rats or CD-1 mice were dosed with each compound iv 1 mg/kg and po 5
c
mg/kg as a solution in 60% PEG. CD-1 mice were dosed po 100 mg/kg as a MCT suspension.
analogue 23, with the (1R,2R)-2-F-cyclopropylamide, had an
experimental LogD that was 0.5 units lower than that of 3 (2.9
vs 3.4). It has been reported that fluorine substitutions on
alkanes, especially in the presence of adjacent oxygen atom,
usually result in lowered lipophilicity.52a Through the
combination of improved potency and lower LogD, compound
23 has higher LLE than 3 (5.7 vs 4.9). The same trends were
observed when the (1R,2R)-2-F-cyclopropylamide and 4-nitrile
phenyl group were combined to give compound 35 (LLE =
6.0). Replacement of one of the chlorines on the phenyl ring
present in 35 with fluorine yielded compound 37, which
showed an experimental LogD of 2.5, a drop of 0.3 units.
Consequently, compound 37 had the highest LLE (6.3) of all
analogues in this scaffold. Aqueous solubility of these analogues
at pH 7.4 was decent. The cell permeability, as measured in
MDCK cells, was high. The plasma protein binding (PPB) of
these compounds was examined in human, rat, and mouse.
Most of them are highly plasma protein bound. Consistent with
its lower lipophilicity, compound 37 showed the lowest plasma
protein binding in all three species. It is also interesting to note
that for 37, its plasma protein binding in mouse is considerably
higher than in human.
The in vitro metabolic stability and in vivo pharmacokinetic
(PK) data of the key inhibitors are summarized in Table 5.
Analogue 19 contained the 4-nitrile functionality on the phenyl
group, and this substitution was designed to block oxidative
metabolism that was observed for compound 3. No significant
difference in human liver microsomal stability, however, was
observed between analogues 19 and 3. Incubation of
compound 19 with human liver microsomes showed only a
small amount (<2%) of the corresponding cyclopropylamide
hydrolysis product, and no metabolites related to oxidation on
the phenyl ring were detected. Comparison of in vivo metabolic
stability in mouse and rat between compounds 3 and 19 reveals
that addition of C4-nitrile on the phenyl ring clearly had
significant impact, as clearance was reduced from 21 to 3.1 mL/
min/kg in rat and from 65 to 13 mL/min/kg in mouse,
respectively. When compound 19 was dosed orally in mouse at
100 mg/kg, excellent oral exposure (AUC = 818 μM·h) was
observed.
For analogue 23, we were initially concerned that fluorine
substitution on the cyclopropylamide would reduce metabolic
stability. Cyclopropylamide hydrolysis was one of the major
metabolic pathways observed for compound 3, and the strongly
electron-withdrawing fluorine in compound 23 could con-
ceivably, through its inductive effect, render the cyclo-
propylamide carbonyl even more susceptible to nucleophilic
attack. The liver microsomes stability of compound 23 was
comparable to 3 in all three species. To confirm these stability
trends, we evaluated the clearance of analogue 23 in vivo and
obtained a clearance value of 15 mL/min/kg in rat, similar to
that of 3 (21 mL/min/kg). When compound 23 was dosed
orally at 100 mg/kg in mouse, however, modest exposure
(AUC = 50 μM·h) was observed.
When the (1R,2R)-2-F-cyclopropylamide fragment was
combined with the 4-cyano-substituted phenyl groups (com-
pound 35 and 37), attractive PK profiles were obtained. For
compound 35 with the 2,6-dichloro-4-cyanophenyl group, the
in vitro metabolic stability in rat and mouse was comparable to
that of 19. In vivo, a clearance value of 4.6 mL/min/kg was
observed in rat. In a high dose mouse PK experiment with
compound 35, excellent oral exposure (AUC = 1090 μM·h)
4528
dx.doi.org/10.1021/jm400266t | J. Med. Chem. 2013, 56, 4521−4536

Journal of Medicinal Chemistry
Article
Figure 6. High dose (po 100 mg/kg) mouse PK of compound 35 and 37. Three CD-1 mice were dosed orally with 35 or 37 at 100 mg/kg as a
MCT suspension. At different time intervals, a blood sample was taken from each animal and plasma concentration was determined. Reported values
were the average of three animals. The unbound concentration was calculated by multiplying the plasma concentration with the corresponding
unbound fraction in mouse.
was observed, with maximal total concentration (Cmax) reaching
149 μM. However, 2-chloro-6-fluoro analogue 36, lacking a 4-
nitrile on its phenyl group, showed a relatively high clearance
(39 mL/min/kg) in rat. When nitrile was introduced at the C4-
position of the phenyl group, resulting in analogue 37, in vitro
metabolic stability was comparable to that of compounds 19
and 35. Once again, low clearance (1.0 mL/min/kg) was
observed in rat for compound 37, with excellent oral exposure
achieved in mouse.
It was gratifying to see that the changes introduced into the
lead molecule 3, leading to optimized compounds 35 and 37,
improve not only potency and selectivity but also clearance in
rat and oral exposure in mouse. With these compounds in
hand, we proceeded to in vivo PK/PD studies. While
compound 35 was more potent in the IL-12 pSTAT4 assay
than 37, it also had higher plasma protein binding in mouse
(98.3% vs 95.0%). As a result, unbound drug level relative to its
cell EC50 for analogue 37 was slightly higher than that for 35
(24× vs 17×) (Figure 6). On the basis of these calculations,
compound 37 was selected for in vivo PK/PD study.
To determine if a TYK2-selective inhibitor could inhibit the
IL-12 pathway in vivo, a mouse PK/PD model was developed.
Administration of recombinant mouse IL-12 in combination
with mouse IL-18 resulted in a time- and dose-dependent
increase in systemic IFNγ production (data not shown). After
optimization of the time and dose required for induction,
Figure 7. (a) Mouse IL-12 PD response with different doses of
compound 37, showing inhibition of IFNγ production by compound
compound 37 was tested for its ability to inhibit IFNγ
production in this model. Anti-p40 antibody, which neutralizes
the IL-12 cytokine, was used as a positive control and resulted
in near complete (94%) inhibition of IFNγ production (Figure
7a). Thirty minutes prior to cytokine stimulation, ascending
doses of 37 were administered orally via gavage (see Supporting
Information for the time course of mouse IL-12 PK/PD
experiment). Three hours after cytokine stimulation, systemic
levels of IFNγ were measured by cytokine ELISA, and total
plasma concentration of compound 37 was also determined for
this time point. Statistically significant inhibition of the IL-12
pathway was observed for doses of ≥10 mg/kg (Figure 7a),
with maximal PD response at the 70 mg/kg dose. The lowest
dose of 3 mg/kg, however, failed to produce a meaningful
pharmacological effect. The 100 mg/kg dose did not give
significantly higher unbound drug levels than the 70 mg/kg
37. Values shown are the average of eight animals at each dose (n = 8).
Anti-p40 antibody (aP40), which neutralizes IL-12, was used as a
positive control. Anti-ragweed antibody (aRW) was used as a negative
control. An asterisk denotes p < 0.05 compared to control. MCT was
used as the vehicle for oral dosing of compound 37. (b) Measured
unbound concentration of compound 37 at 3.5 h after oral dosing:
plot of the unbound concentration of compound 37 at the 3.5 h time
point for different doses. Values shown are the average of eight animals
at each dose (n = 8). The ratio of unbound concentration of 37 over
its IL-12 pSTAT4 EC50 was also calculated for all doses.
dose (Figure 7b), and thus, the PD response was comparable.
Overall, the PK/PD relationship suggested that a significant
attenuation of cytokine IFNγ production could be achieved at
doses of 100, 70, 30, and 10 mg/kg when unbound drug
4529
dx.doi.org/10.1021/jm400266t | J. Med. Chem. 2013, 56, 4521−4536

Journal of Medicinal Chemistry
Article
a
Scheme 1. Preparation of 2-Cyclopropylamides−Pyridyl-4-benzamides
a
Reagents and conditions: (a) thionyl chloride, toluene, 100 °C; then NaH, DMF, 0 °C; (b) Pd2(dba)3, XantPhos, Cs2CO3, 1,4-dioxane, 130 °C,
microwave; (c) Pd2(dba)3, DavePhos, LHMDS, 1,4-dioxane, 100 °C; then 1 N HCl; (d) HATU, DIPEA, DMF, 70 °C; (e) NEt3, dichloromethane, 0
°C.
concentration is ≥2-fold of the corresponding IL-12 pSTAT4
sequent exploration of various C2-substituent on the cyclo-
propylamide present in 3 led to identification of the (1R,2R)-2-
fluorocyclopropylamide moiety that further improved TYK2
potency and JAK1 and JAK2 indexes, without compromising
metabolic stability. When the two separately optimized moieties
were combined, the resulting compound 35 and the closely
related analogue 37 displayed excellent cell potency and
selectivity profiles. Both analogues also exhibited good oral
exposures when dosed in mouse. Accordingly, compound 37
was selected for a mouse PK/PD experiment because of it
having the best ratio of unbound drug level in mouse over its
cell IL-12 pSTAT4 EC50. Significant suppression of IFNγ
production was achieved when compound 37 was dosed at
greater than 3 mg/kg in a mouse IL-12 PK/PD model. This
result supports our hypothesis that inhibition of the TYK2
kinase alone could lead to blocking of the IL-12 signaling
pathway in vivo.
EC50 value.
CHEMISTRY
■
Preparation of compounds reported in this work is summarized
in Scheme 1, utilizing chemistry similar to what was reported in
our previous publication.41 The syntheses of various benzoic
acids are described in the Supporting Information. Briefly, a
2,6-dichlorobenzoic acid 38 was converted to its corresponding
acid chloride, which was coupled with 2-bromo 4-amino-
pyridine 39 to give amide 40. With 2-bromopyridine 40 in
hand, we developed two methods to prepare cyclopropylamide
42 (methods A and B in Scheme 1). A palladium-catalyzed
Buchwald coupling reaction of bromide 40 with primary amide
41 (method A) could be carried out smoothly to give
cyclopropylamide 42. Alternatively, the 2-bromopyridine 40
could be first transformed to 2-aminopyridine 43. Standard
amide coupling reactions of 2-aminopyridine 43 with a
cyclopropylcarboxylic acid 44, or an acid chloride 45, also led
to cyclopropylamide 42.
EXPERIMENTAL SECTION
■
General. Unless otherwise indicated, all reagents and solvents were
purchased from commercial sources and used without further
purification. Moisture or oxygen sensitive reactions were conducted
under an atmosphere nitrogen gas. Unless otherwise stated, 1H spectra
were recorded at room temperature using Varian Unity Inova Bruker
AVANCE III UltraShield-Plus Digital NMR spectrometer at indicated
frequencies. Chemical shifts are expressed in ppm relative to an
internal standard: tetramethylsilane (=0.00 ppm). The following
abbreviations are used: br = broad signal, s = singlet, d = doublet, dd =
doublet of doublets, t = triplet, q = quartet, m = multiplet. Purification
by silica gel chromatography was carried out using a CombiFlash by
Teledyne ISCO system with prepacked cartridges. Purification by
reverse phase high performance liquid chromatography (HPLC) was
also used, and the conditions were described as indicated. Racemic
mixtures of final compounds were separated into individual
enantiomers by chiral supercritical fluid chromatography (SFC)
under the indicated conditions. All final compounds were purified to
CONCLUSION
■
Starting with lead molecule 3, we carried out comprehensive
and systematic SAR investigation of the 2,6-dichlorophenyl and
cyclopropylamide moieties, aided by structure-based design.
Optimization of the benzamide portion of 3 led to
identification of the 2,6-dichloro-4-cyanophenyl group, which
showed improved TYK2 potency as well as JAK1 and JAK2
selectivity. This illustrates that selectivity can be obtained not
only by directly targeting residue differences but also by
utilizing the overall difference in P-loop conformation or
flexibility. Furthermore, incorporation of the C4-nitrile
substituent reduced clearance in vivo, likely as a result of
blocking oxidative metabolism on the phenyl group. Sub-
4530
dx.doi.org/10.1021/jm400266t | J. Med. Chem. 2013, 56, 4521−4536

Journal of Medicinal Chemistry
Article
>95% chemical purity, as assessed by HPLC (column, Agilent SD-C18,
2.1 mm × 30 mm, 1.8 μm; mobile phases, water with 0.5% TFA,
CH3CN with 0.5% TFA; flow rate of 0.4 mL/min; run time of 8.5
min; oven temperature of 40 °C; UV detection at 254 and 210 nm. In
cases where enantiomers or diastereomers were separated by chiral
stationary phase HPLC or chiral SFC, the final products were >98% ee
or de.
2,6-Dichloro-4-cyano-N-(2-(cyclopropanecarboxamido)-
pyridin-4-yl)benzamide (19). Step 1. A solution of 2,6-dichloro-4-
cyanobenzoic acid (1.15 g, 5.32 mmol) in thionyl chloride (5.0 mL)
was heated to reflux under nitrogen overnight. The next day, the
mixture was concentrated via rotavap to give the desired acid chloride
as a white solid, which was used in the next step without further
purification.
2,6-Dichloro-N-(2-(trans-2-fluorocyclopropane-
carboxamido)pyridin-4-yl)benzamide Enantiomer 1 (25) and
Enantiomer 2 (26). A 10 mL round-bottom flask was charged with
N-(2-aminopyridin-4-yl)-2,6-dichlorobenzamide (28 mg, 0.10 mmol),
followed by anhydrous DMF (1 mL), trans-2-fluorocyclopropanecar-
boxylic acid54 (31 mg, 0.30 mmol), DIPEA (52 mg, 0.40 mmol), and
HATU (114 mg, 0.30 mmol). The reaction mixture was stirred at
room temperature under nitrogen for 18 h. The reaction mixture was
diluted with a saturated NH4Cl solution (20 mL) and extracted with
EtOAc (2 × 20 mL). The combined organics were dried over Na2SO4,
filtered, and concentrated via rotavap. The crude product was purified
by chiral SFC (Chiralpak IC, 20% MeOH + 0.1% NEt3, 5 min) to give
two desired products as the following:
25: first eluting peak, 3.4 mg (9.2% yield), >99% ee (retention time
1
0.65 min, Chiralpak IC, 20% MeOH + 0.1% NEt3, 5 min). H NMR
To a solution of 2-bromo-4-aminopyridine (0.912 g, 5.32 mmol) in
anhydrous DMF (5.0 mL) at 0 °C was carefully added NaH (0.424 g,
10.6 mmol, 60% dispersion in mineral oil). Ice bath was removed, and
the suspension was warmed to room temperature and stirred for 30
min before being cooled to 0 °C again. To this reaction mixture was
slowly added a solution of the above acid chloride dissolved in
anhydrous DMF (4.0 mL). Ice bath was removed, and the mixture was
warmed to room temperature and stirred for 1 h under nitrogen. Ice−
water (20 mL) was added, and the reaction mixture was extracted with
ethyl acetate (3 × 50 mL). Combined organics were dried over
Na2SO4, concentrated via rotavap, and purified by flash column
chromatography on silica gel (EtOAc/petroleum ether = 1:10) to
afford N-(2-bromopyridin-4-yl)-2,6-dichloro-4-cyanobenzamide as a
white solid (1.18 g, 59% yield). LCMS (ESI) m/z: 369.9 [M + H]+.
Step 2. A 10 mL microwave tube equipped with a magnetic stirring
bar was charged with N-(2-bromopyridin-4-yl)-2,6-dichloro-4-cyano-
benzamide (74 mg, 0.20 mmol), followed by cyclopropanecarbox-
amide (34 mg, 0.40 mmol), Pd2(dba)3 (18 mg, 0.020 mmol),
XantPhos (23 mg, 0.040 mmol), Cs2CO3 (131 mg, 0.40 mmol), and
1,4-dioxane (1.2 mL). The tube was flushed with nitrogen for 5 min
before being sealed and heated at 140 °C for 1 h in a microwave
reactor. The tube was cooled to room temperature, and the reaction
mixture was filtered through Celite. The filter cake was washed with
additional EtOAc (2 × 5 mL). Combined filtrate was evaporated to
dryness and purified by preparative HPLC (Gilson GX 281; column,
Shim-pack PRC-ODS 250 mm × 20 mm 15 μm; mobile phase
CH3CN/10 mM NH4HCO3; flow rate of 30 mL/min; run time of 17
min) to afford the desired product (30 mg, 40% yield). 1H NMR (500
MHz, DMSO-d6): δ 11.28 (s, 1H), 10.79 (s, 1H), 8.36−8.24 (m, 4H),
7.47 (m, 1H), 2.00 (m, 1H), 0.81−0.79 (m, 4H). LCMS (ESI) m/z:
375.1 [M + H]+.
2,6-Dichloro-N-(2-((1R,2R)-2-fluorocyclopropane-
carboxamido)pyridin-4-yl)benzamide (23) and 2,6-Dichloro-N-
(2-((1S,2S)-2-fluorocyclopropanecarboxamido)pyridin-4-yl)-
benzamide (24). To a 25 mL round-bottom flask was added N-(2-
aminopyridin-4-yl)-2,6-dichlorobenzamide (60 mg, 0.213 mmol),
followed by anhydrous DMF (2.0 mL), cis-2-fluorocyclopropanecar-
boxylic acid54 (111 mg, 1.06 mmol), DIPEA (137 mg, 1.06 mmol),
and HATU (404 mg, 1.06 mmol). The reaction mixture was heated at
60 °C under nitrogen for 12 h when monitoring the reaction by LCMS
showed complete conversion. The reaction mixture was cooled to
room temperature, diluted with saturated NH4Cl solution (20 mL),
and then extracted with EtOAc (3 × 20 mL). The combined organics
were dried over Na2SO4, filtered, and concentrated via rotavap. The
crude product was purified by chiral SFC (Chiralpak IC, 35% IPA +
0.1% NEt3, 5 min) to give two desired products as the following:
23: second eluting peak, 21 mg (27% yield), >99% ee (retention
time 0.87 min, Chiralpak IC, 35% IPA + 0.1% NEt3, 5 min). 1H NMR
(400 MHz, DMSO-d6): δ 11.15 (s, 1H), 10.77 (s, 1H), 8.39 (s, 1H),
8.24 (d, J = 5.6 Hz, 1H), 7.62−7.55 (m, 2H), 7.55−7.43 (m, 2H), 4.91
(dddd, J = 66.3, 6.3, 6.2, 3.6 Hz, 1H), 2.21 (dddd, J = 9.0, 6.8, 6.3, 5.9
Hz, 1H), 1.65 (dddd, J = 23.4, 6.8, 6.7, 3.6 Hz, 1H), 1.16 (dddd, J =
12.4, 9.0, 6.8, 6.3 Hz, 1H). LCMS (ESI) m/z: 368.0 [M + H]+.
24: first eluting peak, 22 mg (28% yield), >99% ee (retention time
0.61 min, Chiralpak IC, 35% IPA + 0.1% NEt3, 5 min). 1H NMR and
LCMS match 23.
(400 MHz, DMSO-d6): δ 11.14 (s, 1H), 10.88 (s, 1H), 8.38 (d, J = 1.6
Hz, 1H), 8.24 (d, J = 5.6 Hz, 1H), 7.62−7.55 (m, 2H), 7.51 (dd, J =
9.2, 6.8 Hz, 1H), 7.46 (dd, J = 5.6, 1.9 Hz, 1H), 4.86 (dddd, J = 64.7,
5.8, 3.2, 1.4 Hz, 1H), 2.56 (dddd, J = 22.3, 10.1, 6.4, 1.4 Hz, 1H), 1.51
(dddd, J = 22.3, 10.1, 6.4, 3.2 Hz, 1H), 1.23 (dddd, J = 12.9, 6.5, 6.4,
6.3 Hz, 1H). LCMS (ESI) m/z: 368.0 [M + H]+.
26: second eluting peak, 3.3 mg (9.0% yield), >99% ee (retention
1
time 0.74 min, Chiralpak IC, 20% MeOH + 0.1% NEt3, 5 min). H
NMR and LCMS match 25.
(1R,2S)-2-(4-(2,6-Dichlorobenzamido)pyridin-2-
ylcarbamoyl)cyclopropanecarboxylic Acid (32). Step 1. A
solution of N-(2-aminopyridin-4-yl)-2,6-dichlorobenzamide (1.41 g,
5 mmol) and 3-oxabicyclo[3.1.0]hexane-2,4-dione (2.24 g, 20 mmol)
in 1,4-dioxane (25 mL) was heated at 90 °C under nitrogen for 4 h.
The reaction mixture was cooled to room temperature and the
precipitated white solid was collected by filtration to give 2,6-dichloro-
N-(2-(-2,4-dioxo-3-azabicyclo[3.1.0]hexan-3-yl)pyridin-4-yl)-
1
benzamide (1.28 g, 68% yield). H NMR (400 MHz, DMSO-d6): δ
11.42 (s, 1H), 8.47 (d, J = 5.6 Hz, 1H), 7.69 (d, J = 1.8 Hz, 1H), 7.67−
7.59 (m, 3H), 7.55 (dd, J = 9.2, 6.8 Hz, 1H), 2.74 (dd, J = 8.0, 3.6 Hz,
2H), 1.69 (ddt, J = 20.6, 8.1, 4.2 Hz, 2H). LCMS (ESI) m/z: 376.0 [M
+ H]+.
Step 2. To a suspension of 2,6-dichloro-N-(2-(-2,4-dioxo-3-
azabicyclo[3.1.0]hexan-3-yl)pyridin-4-yl)benzamide (136 mg, 0.36
mmol) in THF (4.0 mL) was added a solution of LiOH (1 N, 2
mL, 2 mmol). The reaction mixture was stirred at room temperature
for 2 h. The mixture was then extracted with EtOAc (10 mL). The
aqueous layer was acidified with 1 N HCl to pH 2.0. The precipitated
white solid was collected by filtration. The crude product was
subsequently purified by chiral SFC (Chiralpak AD-H, 35% EtOH +
0.1% NEt3, 6 min) to give two pure enantiomers as the following:
32: second eluting peak, 41 mg (29% yield), >99% ee (retention
time 0.66 min, Chiralpak AD-H, 35% EtOH + 0.1% NEt3, 3 min). 1H
NMR (400 MHz, DMSO-d6): δ 12.10 (s, 1H), 11.13 (s, 1H), 10.85 (s,
1H), 8.36 (d, J = 1.5 Hz, 1H), 8.22 (d, J = 5.6 Hz, 1H), 7.61−7.55 (m,
2H), 7.54−7.50 (m, 1H), 7.48 (dd, J = 5.5, 1.8 Hz, 1H), 2.27 (dd, J =
14.6, 7.5 Hz, 1H), 2.01 (dd, J = 15.9, 8.5 Hz, 1H), 1.39 (td, J = 6.6, 4.2
Hz, 1H), 1.16 (td, J = 8.3, 4.1 Hz, 1H). LCMS (ESI) m/z: 394.0 [M +
H]+.
Enantiomer of 32: first eluting peak, 43 mg (30% yield), >99% ee
(retention time 0.56 min, Chiralpak AD-H, 35% EtOH + 0.1% NEt3, 3
1
min). H NMR and LCMS match 32.
2,6-Dichloro-4-cyano-N-(2-((1R,2R)-(2-fluorocyclopropane-
carboxamido)pyridin-4-yl)benzamide (35). Step 1. A solution of
2,6-dichloro-4-cyanobenzoic acid (324 mg, 1.5 mmol) in thionyl
chloride (10 mL) was heated to reflux under nitrogen overnight. The
next day, the reaction mixture was cooled to room temperature and
volatiles were removed via rotavap. The corresponding acid chloride
was obtained in quantitative yield as a yellow solid, which was used in
the next step without purification.
Step 2. A 50 mL round-bottom flask was charged with 2-bromo-4-
aminopyridine (295 mg, 1.5 mmol), followed by anhydrous DMF (5.0
mL). The resulting solution was cooled to 0 °C, and NaH (120 mg,
3.0 mmol, 60% dispersion in mineral oil) was added carefully under
nitrogen. Ice bath was removed, and the mixture was warmed to room
4531
dx.doi.org/10.1021/jm400266t | J. Med. Chem. 2013, 56, 4521−4536

Journal of Medicinal Chemistry
Article
temperature and stirred for 30 min before being cooled to 0 °C again.
To this reaction mixture was slowly added a solution of the above acid
chloride dissolved in anhydrous DMF (5.0 mL). The resulting mixture
was stirred for another 1 h at room temperature. Then ice−water (20
mL) was added and the resulting mixture was extracted with ethyl
acetate (3 × 50 mL). The combined organic layers were washed with
brine (20 mL), dried over MgSO4, and filtered. After removal of
solvent under vacuum, the residue was purified by flash chromatog-
raphy (ethyl acetate/petroleum ether = 1:10) to give N-(2-
bromopyridin-4-yl)-2,6-dichloro-4-cyanobenzamide (389 mg, 70%
yield). LCMS (ESI) m/z: 369.9 [M + H]+.
Step 3. To a 25 mL microwave tube was added cis-2-
fluorocyclopropanecarboxamide (0.045 g, 0.44 mmol), followed by
N-(2-bromopyridin-4-yl)-2,6-dichloro-4-cyanobenzamide (0.16 g, 0.40
mmol), Pd2(dba)3 (0.045 g, 0.049 mmol), XantPhos (0.035 g, 0.061
mmol), Cs2CO3 (0.51 g, 1.6 mmol), and 1,4-dioxane (6.0 mL). The
mixture was degassed with nitrogen for 10 min. The tube was sealed
and irradiated in a microwave reactor at 140 °C for 1 h before being
cooled to room temperature. The mixture was filtered with Celite, and
the filter cake was washed with additional EtOAc (2 × 10 mL).
Combined filtrate was concentrated via rotavap, and the crude mixture
was purified by preparative HPLC (Gilson GX 281; column, Shim-
pack PRC-ODS 250 mm × 20 mm 15 μm; mobile phase CH3CN/10
mM NH4HCO3; flow rate of 30 mL/min; run time of 17 min) to give
the racemic mixture as a white solid (40 mg). This racemic mixture
was then purified by chiral SFC (Chiralpak IC, 15% MeOH, 5 min) to
give two pure enantiomers as following:
AS-H, 15% MeOH, 12 min) to give two pure enantiomers as the
following:
37: second eluting peak, 288 mg (40% yield), >99% ee (retention
time 1.23 min, Chiralpak AS-H, 15% MeOH, 2.5 min). 1H NMR (400
MHz, DMSO-d6): δ 11.33 (s, 1H), 10.82 (s, 1H), 8.35 (d, J = 1.6 Hz,
1H), 8.26 (d, J = 5.6 Hz, 1H), 8.17−8.08 (m, 2H), 7.49 (dd, J = 5.6,
1.9 Hz, 1H), 4.94 (dddd, J = 66.2, 6.3, 6.2, 3.7 Hz, 1H), 2.21 (dddd, J
= 12.4, 9.2, 6.9, 6.3 Hz, 1H), 1.64 (dddd, J = 23.2, 6.9, 6.8, 3.7 Hz,
1H), 1.16 (dddd, J = 12.4, 9.2, 6.3, 6.3 Hz, 1H). LCMS (ESI) m/z:
377.1 [M + H]+.
The other enantiomer was also obtained (290 mg, 40% yield),
>99% ee (retention time 1.03 min, Chiralpak AS-H, 15% MeOH, 2.5
1
min). H NMR and LCMS data match 37.
Computational Chemistry. To orthogonally validate the multi-
polar interaction described in ref 47 and depicted by Proasis for the
TYK2 X-ray structures with 19 (Figure 3) and 35 (Figure 5), quantum
calculations were carried out on model systems. The carbonyl was
represented by a propan-2-one molecule, and the nitrile was
represented by an acetonitile molecule, both starting in the same
coordinates as in the crystal structure of 19. The program Jaguar
(Schrodinger, Inc.) was used to carry out geometry optimization runs
at the B3LYP/6-31G** level with PBF aqueous solvation for the
propan-2-one alone, the acetonitrile alone, and the propan-2-one/
acetonitrile complex with distance and angle constraints between the
carbonyl and the nitrile to maintain the starting geometry of the
complex. At the end of the geometry optimization runs, the interaction
energy was calculated via subtraction of the sum of the solution phase
energies for each separate molecule from the solution phase energy of
the complex system. A similar method was used by Pierce and co-
workers55 to predict CH···O hydrogen bond strengths. Our calculation
yielded an interaction energy of −0.9 kcal/mol, indicating a favorable
interaction. To further verify the method and compare results with
SAR, we used the same method on a system in which the acetylnitrile
was switched to a prop-1-yne. This system mimics compound 20, and
the resulting interaction energy was −0.1 kcal/mol, too small to
conclude whether the interaction is favorable. This set of results is
consistent with compound 19 having TYK2 Ki of 1.8 nM and
compound 20 having TYK2 Ki of 9.5 nM.
35: second eluting peak, 10 mg (6.4% yield), >99% ee (retention
1
time 1.31 min, Chiralpak IC, 15% MeOH, 5 min). H NMR (400
MHz, DMSO-d6): δ 11.32 (s, 1H), 10.86 (s, 1H), 8.37 (d, J = 1.7 Hz,
1H), 8.32−8.24 (m, 3H), 7.49 (dd, J = 5.6, 1.9, 25 Hz, 1H), 4.95
(dddd, J = 66.2, 6.2, 6.2, 3.7 Hz, 1H), 2.21 (dddd, J = 9.1, 6.7, 6.2, 3.7
Hz, 1H), 1.63 (dddd, J = 23.2, 6.7, 6.2, 3.7 Hz, 1H), 1.16 (dddd, J =
12.4, 9.1, 6.2, 6.2 Hz, 1H). LCMS (ESI) m/z: 393.0 [M + H]+.
The other enantiomer was also obtained (12.5 mg, 8.0% yield),
>99% ee (retention time 1.12 min, Chiralpak IC, 15% MeOH, 5 min).
1H NMR and LCMS match 35.
2-Chloro-4-cyano-6-fluoro-N-(2-((1R,2R)-2-fluoro-
cyclopropanecarboxamido)pyridin-4-yl)benzamide (37). Step
1. A 10 mL round-bottom flask containing a suspension of 2-chloro-4-
cyano-6-fluorobenzoic acid (443 mg, 2.22 mmol) in thionyl chloride
(2.0 mL) was heated to reflux for 3 h under nitrogen. The mixture was
cooled to room temperature, concentrated to dryness, and azeotroped
with anhydrous toluene (2 × 5 mL) to give crude 2-chloro-4-cyano-6-
fluorobenzoyl chloride as a white solid which was used in the next step
without purification.
Step 2. A 25 mL round-bottom flask was charged with 4-amino-2-
bromopyridine (346 mg, 2.0 mmol) and DIPEA (461 mg, 4.0 mmol),
followed by anhydrous DCM (5.0 mL) and CH3CN (0.5 mL). The
resulting clear solution was cooled to 0 °C under nitrogen, and a
solution of 2-chloro-4-cyano-6-fluorobenzoyl chloride (484 mg, 2.2
mmol) in DCM (2.0 mL) was added dropwise. Ice bath was removed,
and the reaction mixture was warmed to room temperature and stirred
for 16 h. The reaction mixture was diluted with a saturated NH4Cl
solution (40 mL) and extracted with EtOAc (3 × 50 mL). The
combined organics were dried over Na2SO4, filtered, and concentrated
to dryness. The crude product was purified by flash column
chromatography on silica gel (0−50% EtOAc/hexane) to give N-(2-
bromopyridin-4-yl)-2-chloro-4-cyano-6-fluorobenzamide as an off-
white solid (379 mg, 53% yield).
Step 3. To a 25 mL microwave tube was added N-(2-bromopyridin-
4-yl)-2-chloro-4-cyano-6-fluorobenzamide (681 mg, 1.92 mmol),
followed by cis-2-fluorocyclopropanecarboxamide (218 mg, 2.11
mmol), Pd2(dba)3 (88 mg, 0.096 mmol), XantPhos (0.111 g, 192
mmol), Cs2CO3 (1.25 g, 3.84 mmol), 1,4-dioxane (10.0 mL), and 1,2-
dimethoxyethane (3.0 mL). The mixture was degassed with nitrogen
for 10 min. The tube was sealed and irradiated in a microwave reactor
at 130 °C for 0.5 h. After being cooled to room temperature, the
mixture was filtered with Celite and the filtrate was concentrated under
vacuum. The resulting residue was purified by chiral SFC (Chiralpak
ASSOCIATED CONTENT
■
S
* Supporting Information
Detailed experimental procedures for different benzoic acids
and the rest of compounds shown in this work; single X-ray
crystal structures of compounds 23 and 32; crystallographic
methods and procedures for 19 in complex with TYK2, 23 in
complex with TYK2, and 35 in complex with TYK2; time
course of mouse IL-12 PK/PD experiment. This material is
available free of charge via the Internet at http://pubs.acs.org.
Accession Codes
PDB codes: 4GI1 for 19 complexed with TYK2, 4GJ2 for 23
complexed with TYK2, and 4GJ3 for 35 complexed with TYK2.
AUTHOR INFORMATION
■
Corresponding Author
*Phone: 650-467-3567,. E-mail: liang.jun@gene.com.
Present Addresses
●A.v.A: Department of Cell Biology, Novartis, Emeryville,
California.
△W.B: Department of Infectious Disease, MedImmune,
Gaithersburg, Maryland.
▽J.D.: Department of Phamcokinetics, Theravance, South San
Francisco, California.
Notes
The authors declare no competing financial interest.
4532
dx.doi.org/10.1021/jm400266t | J. Med. Chem. 2013, 56, 4521−4536

Journal of Medicinal Chemistry
Article
drofuran; TLC, thin layer chromatography; tPSA, topological
polar surface area; UV, ultraviolet; Vd, volume of distribution;
XantPhos, 4,5-bis(diphenylphosphino)-9,9-dimthylxanthene
ACKNOWLEDGMENTS
■
We thank, Mengling Wong, Chris Hamman, Michael Hayes,
Baiwei Lin, Deven Wang, Yutao Jiang for purification 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; Pete
Dragovich, Steve Staben, and Kirk Robarge for many helpful
comments and suggestions. Diffraction data were collected at
the Stanford Synchrotron Radiation Lightsource, operated for
the U.S. Department of Energy Office of Science by Stanford
University, CA. The SSRL Structural Molecular Biology
Program is supported by the DOE Office of Biological and
Environmental Research, National Institute of General Medical
Sciences (NIH) (including Grant P41GM103393), and the
National Center for Research Resources (Grant
P41RR001209).
REFERENCES
■
(1) (a) Christophers, E. Psoriasis-epidemiology and clinical spectrum.
Clin. Exp. Dermatol. 2001, 26, 314−320. (b) Lowes, M. A.; Bowcock,
A. M.; Krueger, J. G. Pathogenesis and therapy of psoriasis. Nature
2007, 445, 866−873.
(2) (a) Abraham, C.; Cho, J. H. N. Engl. J. Med. 2009, 361, 2066−
2078. (b) Garrette, W. S.; Gordon, J. I.; Glimcher, L. H. Homeostasis
and inflammation in the intestine. Cell 2010, 140, 859−870. (c) Kaser,
A.; Zeissig, S.; Blumberg, R. S. Inflammatory bowel disease. Annu. Rev.
Immunol. 2010, 28, 573−621.
(3) (a) Gelfand, J. M.; Weinstein, R.; Porter, S. B.; Neimann, A. L.;
Berlin, J. A.; Margolis, D. J. Prevalence and treatment of psoriasis in
the United Kingdoma population study. Arch. Dermatol. 2005, 141,
1537−1541. (b) Kurd, S. K.; Gelfand, J. M. The prevalence of
previously diagnosed and undiagnosed psoriasis in US adults: results
from NHANES 2003−2004. J. Am Acad. Dermatol. 2009, 60, 218−
224.
(4) (a) Mikhail, M.; Scheinfeld, N. S. Therapy treatment options for
psoriasis: topical and systemic. Therapy 2004, 1, 319−330.
(b) Nograles, K. E.; Krueger, J. G. Anti-cytokine therapies for
psoriasis. Exp. Cell Res. 2011, 317, 1293−1300.
(5) Loftus, E. V. Clinical epidemiology of inflammatory bowel
disease: incidence, prevalence, and environmental influences. Gastro-
enterology 2004, 126, 1504−1517.
ABBREVIATIONS USED
■
AIBN, 2,2′-azobis(2-methylpropionitrile); Arg, arginine; ATP,
adenosine triphosphate; AUC, area under the curve; Boc2O, di-
tert-butyl dicarbonate; Boc, tert-butyl carbamate; brine, a
saturated aqueous solution of sodium chloride; CHCl3,
chloroform; CH3CN, acetonitrile; CL, clearance; Cmax
,
(6) (a) Hanauer, S. B.; Feagan, B. G.; Lichtenstein, G. R.; Mayer, L.
F.; Schreiber, S.; Colombel, J. F.; Rachmilewitz, D.; Wolf, D. C.;
Olson, A.; Bao, W.; Rutgeerts, P. Maintenance infliximab for Crohn’s
disease: the ACCENT I randomized trial. Lancet 2002, 359, 1541−
1549. (b) Rutgeerts, P.; Sandborn, W. J.; Feagan, B. G.; Reinisch, W.;
Olson, A.; Johanns, J.; Travers, S.; Rachmilewitz, D.; Hanauer, S. B.;
Lichtenstein, G.; de Villiers, W. J. S.; Present, D.; Sands, B. E.;
Colombel, J. F. Infliximab for induction and maintenance therapy for
ulcerative colitis. N. Engl. J. Med. 2005, 353, 2462−2476. (c) Ngo, B.;
Farrell, C. P.; Barr, M.; Wolvo, K.; Bailey, R.; Mullin, J. M.; Thornton,
J. J. Tumor necrosis factor blockade for treatment of inflammatory
bowel disease: efficacy and safety. Curr. Mol. Pharmacol. 2010, 3, 145−
152. (d) Caprioli, F.; Pallone, F.; Monteleone, G. Cytokine therapies
in Crohn’s disease: Where are we now and where should we go?
Inflammation Allergy 2011, 10, 47−53.
(7) (a) Baer, F.; Noman, M.; Vermeire, S.; Assche, G. V.; Haens, G.
D.; Cabonez, A.; Rutgeerts, P. Influence of immunogenicity on the
long-term efficacy of infliximab in Crohn’s disease. N. Engl. J. Med.
2003, 348, 601−608. (b) Ben-Horin, S.; Chowers, Y. Loss of response
to anti-TNF treatments in Crohn’s disease. Aliment. Pharmacol. Ther.
2011, 33, 987−995.
(8) (a) Bernstein, C. N.; Wajda, A.; Blanchard, J. F. The clustering of
other chronic inflammatory disease in inflammatory bowel disease: a
population-based study. Gastroenterology 2005, 129, 827−836. (b) de
Mesquita, M. B.; Ferrante, M.; Henckaerts, L.; Joossens, M.; Jaessens,
V.; Hlavaty, T.; Pierik, M.; Joossens, S.; Schuerbeek, M. V.; Assche, G.
V.; Rutgeerts, P.; Vermeire, S.; Hoffman, I. Clustering of (auto)-
immune diseases with early-onset and complicated inflammatory
bowel disease. Eur. J. Pediatr. 2009, 168, 575−583. (c) Ellinghaus, D.;
Ellinghaus, E.; Nair, R. P.; Stuart, P. E.; Esko, T.; Metspalu, A.; Debrus,
S.; Raelson, J. V.; Tejasvi, T.; Belouchi, M.; West, S. L.; Barker, J. N.;
Koks, S.; Kingo, K.; Balschun, T.; Palmieri, O.; Annese, V.; Gieger, C.;
Wichmann, H. E.; Kabesch, M.; Trembath, R. C.; Mathew, C. G.;
Abecasis, G. R.; Weidinger, S.; Nikolaus, S.; Schreiber, S.; Elder, J. T.;
Weichenthal, M.; Nothnagel, M.; Franke, A. Combined analysis of
genome-wide association studies for Crohn’s disease and psoriasis
identifies seven shared susceptibility loci. Am. J. Hum. Genet. 2012, 90,
636−645. (d) Lees, C. W.; Barrett, J. C.; Parkes, M.; Satsangi, J. New
IBD genetics: common pathways with other diseases. Gut 2011, 60,
1739−1753.
maximum concentration; CCl4, carbon tetrachloride; Clp,
clearance; Cs2CO3, cesium carbonate; CO, carbon monoxide;
DavePhos, 2-dicyclohexylphosphino-2′-(N,N-dimethylamino)-
biphenyl; DIPEA, N,N-diisopropylethylamine; DCM, dichloro-
methane; DMF, N,N-dimethylformamide; dppp, 1,3-bis-
(diphenylphosphino)propane; EC50, half maximal effective
concentration; EPO, erythropoietin; EtOAc, ethyl acetate;
EtOH, ethanol; Ex, example; F, oral bioavailability; Gln,
glutamine; Glu, glutamic acid; HATU, O-(7-azabenzotriazol-
1-yl)-N,N,N′,N′-tetramethyluronium hexafluorophosphate; h,
hour; H, hydrogen; HCl, hydrochloric acid; H2O, water;
HPLC, high performance liquid chromatography; H2SO4,
sulfuric acid; i-BuNO2, isobutyl nitrite; IBX, 2-iodoxybenzoic
acid; IL-6, interleukin-6; IL-12, interleukin-12; IL-23, inter-
leukin-23; IFNγ, interferon-γ; iv, intravenous; ip, intra-
peritoneal; IPA, isopropyl alcohol; JAK, Janus kinase;
biochemcal JAK1 index, (JAK1 Ki)/(TYK2 Ki); cellular JAK1
index, (IL-6 pSTAT3 EC50)/(IL-12 pSTAT4 EC50); bio-
chemcal JAK2 index, (JAK2 Ki)/(TYK2 Ki); cellular JAK2
index, (EPO pSTAT5 EC50)/(IL-12 pSTAT4 EC50); K2CO3,
potassium carbonate; Ki, inhibition constant; LCMS, liquid
chromatography−mass spectrometry; LHDMS, lithium hexam-
ethyldisilazide; LiI, lithium iodide; log D, log of partition
coefficient between octanol and aqueous buffer; Leu, leucine;
LM, liver microsome; MCT, methylcellulose/Tween; MDCK,
Madin−Darby canine kidney; MeOH, methanol; min, minute;
NaH, sodium hydride; NaHCO3, sodium bicarbonate; NaNO2,
sodium nitrite; NCS, N-chlorosuccinimide; NEt3, triethyl-
amine; Papp, apparent permeability; Pd2(dba)3, tris-
(dibenzylideneacetone)dipalladium; Pd(dppf)Cl2, 1,1′-bis-
(diphenylphosphino)ferrocenepalladium(II) dichloride; PK,
pharmacokinetics; PK/PD, pharmacokinetics/pharmakody-
namics; po, by mouth; PPB, plasma protein binding; SFC,
supercritical fluid chromatography; STAT, signal transducer
and activator of transcription; t1/2, half-life; TFA, trifluoroacetic
acid; Tf2O, trifluoromethanesulfonic anhydride; THF, tetrahy-
4533
dx.doi.org/10.1021/jm400266t | J. Med. Chem. 2013, 56, 4521−4536

Journal of Medicinal Chemistry
Article
(9) (a) Sweeney, C.; Tobin, A.-M.; Kirby, B. Innate immunity in the
pathogenesis of psoriasis. Arch. Dermatol. Res 2011, 303, 691−705.
(b) Monteleone, G.; Pallone, F.; MacDonald, T. Emerging
immunological targets in inflammatory bowel disease. Curr. Opin.
Pharmacol. 2011, 11, 640−645.
(10) (a) Mucida, D.; Cheroutre, H. The many face-lifts of CD4 T
helper cells. Adv. Immunol. 2010, 107, 139−152. (b) O’Garra, A.;
Robinson, D. Development and function of T-helper 1 cells. Adv.
Immunol. 2004, 83, 133−162. (c) Murphy, K. M.; Reiner, S. The
lineage decisions of helper T cells. Nat. Rev. Immunol 2002, 2, 933−
944.
(11) (a) Skokos, D.; Waite, J. C. Th17 response and inflammatory
autoimmune diseases. Int. J. Inflammation 2012, 819467. (b) Miossec,
P.; Korn, T.; Kuchroo, V. K. Interleukin-17 and type 17 helper cells. N.
Engl. J. Med. 2009, 361, 888−898. (c) Louten, J.; Boniface, K.; Malefyt,
R. d. W. Development and function of Th17 cells in health and
disease. J. Allergy Clin. Immunol. 2009, 123, 1004−1011. (d) Boniface,
K.; Blom, B.; Liu, Y.-J.; Maleft, R. d. W. From interleukin-23 to T-
helper 17 cells: human T-helper cell differentiation revisited. Immunol.
Rev. 2008, 226, 132−146.
(12) (a) Trinchieri, G. Interleukin-12: a proinflammatory cytokine
with immunoregulatory functions that bridge innate resistance and
antigen-specific adaptive immunity. Annu. Rev. Immunol. 1995, 13,
251−276. (b) Chung, Y.; Dong, C. Don’t leave home without it: the
IL-23 visa to Th17 cells. Nat. Immunol. 2009, 10, 236−238.
(13) (a) Chua, A. O.; Chizzonite, R.; Desai, B. B.; Truitt, T. P.;
Nunes, P.; Minetti, L. J.; Warrier, R. R.; Presky, D. H.; Levine, J. F.;
Gately, M. K.; Gubler, U. Expression cloning of a human IL-12
receptor component. J. Immunol. 1994, 153, 128−136. (b) Presky, D.
H.; Yang, H.; Minetti, L. J.; Chua, A. O.; Nabavi, N.; Wu, C.-Y.;
Gately, M. K.; Gubler, U. A functional interleukin 12 receptor complex
is composed of two β-type cytokine receptor subunits. Proc. Natl. Acad.
Sci. U.S.A. 1996, 93, 14002−14007.
(14) Parham, C.; Chirica, M.; Timans, J.; Vaisberg, E.; Travis, M.;
Cheung, J.; Pflanz, S.; Zhang, R.; Singh, K.; Vega, F.; To, W.; Wagner,
J.; O’Farrell, A.-M.; McClanahan, T.; Zurawski, S.; Hannum, C.;
Gorman, D.; Rennick, D.; Kastlein, R. A.; Maleft, R. d. W.; Moore, K.
W. A receptor for the heterodimeric cytokine IL-23 is composed of IL-
12β1 and a novel cytokine receptor subunit, IL-23R. J. Immunol. 2002,
168, 5699−5708.
(15) (a) Firmbach-Kraft, I.; Byers, M.; Shows, T.; Dalla-Favera, R.;
Krolewski, J. J. TYK2, prototype of a novel class of non-receptor
tyrosine kinase genes. Oncogene 1990, 5, 1329−1336. (b) Zou, J.;
Presky, D. H.; Wu, C.-Y.; Gubler, U. Differential associations between
the cytoplasmic regions of the interleukin-12 receptor subunits β1 and
β2 and JAK kinases. J. Biol. Chem. 1997, 272, 6073−6077.
(16) (a) Harpur, A. G.; Andres, A. C.; Ziemiecki, A.; Aston, R. R.;
Wilks, A. F. JAK2, a third member of the JAK family of protein
tyrosine kinases. Oncogene 1992, 7, 1347−1353. (b) Yamamoto, K.;
Shibata, F.; Miura, O.; Kamiyama, R.; Hirosawa, S.; Miyasaka, N.
Physical interaction between interleukin-12 receptor β2 subunit and
JAK2 tyrosine kinase: JAK2 associates with cytoplasmic membrane-
proximal region of interleukin-12 receptor β2 via amino terminus.
Biochem. Biophys. Res. Commun. 1999, 257, 400−404.
(17) Wilks, A. F.; Harpur, A. G.; Kurban, R. R.; Ralph, S. J.; Zurcher,
G.; Ziemiecki, A. Two novel protein-tyrosine kinases, each with a
second phosphotransferase-related catalytic domain, define a new class
of protein kinase. Mol. Cell. Biol. 1991, 11, 2057−2065.
1908−1915. (c) Lee, E.; Trepicchio, W. L.; Oestreicher, J.; Pittman,
D.; Wang, F.; Chamian, F.; Dhodapkar, M.; Krueger, J. G. Increased
expression of interleukin 23p19 and p40 in lesional skin of patients
with psoriasis vulgris. J. Exp. Med. 2004, 199, 125−130. (d) Yawalkar,
N.; Karlen, S.; Hunger, R.; Brand, C. U.; Braathen, L. R. Expression of
interleukin-12 is increased in psoriatic skin. J. Invest. Dermatol. 1998,
111, 1053−1057.
(20) (a) Tada, Y.; Asahina, A.; Takekoshi, T.; Kishimoto, E.; Mitsui,
H.; Saeki, H.; Komine, M.; Tamaki, K. Interleukin 12 production by
monocytes from patients with psoriasis and its inhibition by cyclosprin
A. Br. J. Dermatol. 2006, 154, 1180−1183. (b) Chamian, F.; Lowes, M.
A.; Lin, S.-L.; Lee, E.; Kikuchi, T.; Gilleaudeau, P.; Sullivan-Whalen,
M.; Cardinale, I.; Khatcherian, A.; Novitskaya, I.; Wittkowski, K. M.;
Krueger, J. G. Alefacept reduces infiltrating T cells, activated dendritic
cells, and inflammatory genes in psoriasis vulgaris. Proc. Natl. Acad Sci.
U.S.A. 2005, 102, 2075−2080. (c) Gottlieb, A. B.; Chamian, F.;
Masud, S.; Cardinale, I.; Abello, M. V.; Lowes, M. A.; Chen, F.;
Magliocco, M.; Krueger, J. G. TNF inhibition rapidly down-regulates
multiple proinflammatory pathways in psoriasis plaques. J. Immunol.
2005, 175, 2721−2729. (d) Piskin, G.; Tursen, U.; Sylva-Steenland, R.
M.; Bos, J. D.; Teunissen, M. B. M. Clinical improvement in chronic
plaque-type psoriasis lesions after narrow-band UVB therapy is
accompanied by a decrease in the expression of IFNg inducersIL-
12, IL-18 and IL-23. Exp. Dermatol. 2004, 13, 764−772.
(21) (a) Kopp, T.; Lenz, P.; Bello-Fernandez, C.; Kastelein, R.;
Kupper, T. S.; Stingl, G. IL-23 production by co-secretion of
endogenous p19 and transgenic p40 in keratin 14/p40 transgenic
mice: evidence for enhanced cutaneous immunity. J. Immunol. 2003,
170, 5438−5444. (b) Kopp, T.; Kieffer, J. D.; Rot, A.; Strommer, S.;
Stingl, G.; Kupper, T. S. Inflammatory skin disease in K14/p40
transgenic mice: evidence for interleukin-12-like activities of p40. J.
Invest. Dermatol. 2001, 117, 618−626.
(22) Clarke, A. W.; Poulton, L.; Wai, H. Y.; Walker, S. A.; Victor, S.
D.; Domagala, T.; Mraovic, D.; Butt, D.; Shewmaker, N.; Hennings, P.;
Doylce, A. G. A novel class of anti-IL-12p40 antibodies. mAbs 2010, 2,
539−549.
(23) (a) Kimball, A. B.; Gordon, K. B.; Fakharzadeh, S.; Yeilding, N.;
Szapary, P. O.; Schenkel, B.; Guzzo, C.; Li, S.; Papp, K. A. Long-term
efficacy of ustekinumab in moderate-to-severe psoriasis: results from
the PHOENIX 1 trials through up to 3 years. Br. J. Dermatol. 2012,
166, 861−872. (b) Papp, K. A.; Langley, R. G.; Lebwohl, M.; Krueger,
G. G.; Szapary, P.; Yeilding, N.; Guzzo, C.; Hsu, M.-C.; Wang, Y.; Li,
S.; Dooley, L. T.; Reich, K. Efficacy and safety of ustekinumab, a
human interleukin-12/23 monoclonal antibody, in patients with
psoriasis: 52-week results from a randomized, double-blind, placebo-
controlled trial (PHOENIX 2). Lancet 2008, 371, 1675−1684.
(c) Leonardi, G. L.; Kimball, A. B.; Papp, K. A.; Yeilding, N.;
Guzzo, C.; Wang, Y.; Li, S.; Dooley, L. T.; Gordon, K. B. Efficacy and
safety of ustekinumab, a human interleukin-12/23 monoclonal
antibody, in patients with psoriasis: 76-week results from a
randomized, double-blind, placebo-controlled trial (PHOENIX 1).
Lancet 2008, 371, 1665−1674.
(24) Benson, J. M.; Sachs, C. W.; Treacy, G.; Zhou, H.; Pendley, C.;
Brodmerkel, C. M.; Shankar, G.; Mascelli, M. A. Therapeutic targeting
of the IL-12/23 pathways: generation and characterization of
ustekinumab. Nat. Biotechnol. 2011, 29, 615−624.
(25) (a) Barrett, J. C.; Hansoul, S.; Nicolae, D. L.; Cho, J. H.; Duerr,
R. H.; Rioux, J. D.; Brant, S. R.; Silverberg, M. K.; Taylor, K. D.;
Barmada, M. M.; Bitton, A.; Dassopoulos, T.; Datta, L. W.; Green, T.;
Griffiths, A. M.; Kistner, E. O.; Murtha, M. T.; Regueiro, M. D.;
Rotter, J. T.; Schumm, L. P.; Steinhart, A. H.; Targan, S. T.; Xavier, R.
J.; Libioulle, C.; Sandor, C.; Lathrop, M.; Belaiche, J.; Dewit, O.; Gut,
I.; Heath, S.; Laukens, D.; Mni, M.; Rutgeerts, P.; Gossum, A. v.;
Zelenika, D.; Franchimont, D.; Hugot, J.-P.; de Vos, M.; Vermeire, S.;
Louis, E.; Cardon, L. R.; Anderson, C. A.; Drummond, H.; Nimmo, E.;
Ahmad, T.; Prescott, N. J.; Onnie, C. M.; Fisher, S. A.; Marchini, J.;
Ghori, J.; Bumpstead, S.; Gwilliam, R.; Tremelling, M.; Deloukas, P.;
Mansfield, J.; Jewel, D.; Satsangi, J.; Mathew, C. G.; Parkes, M.;
Georges, M.; Daly, M. J. Genome-wide association defines more than
(18) (a) Rane, S. G.; Reddy, P. E. JAK3: a novel JAK kinase
associated with terminal differentiation of hematopoietic cells.
Oncogene 1994, 9, 2045−2052. (b) Takahashi, T.; Shirasaw, T.
Molecular cloning of rat JAK3, a novel member of the JAK family of
protein tyrosine kinases. FEBS Lett. 1994, 342, 124−128.
(19) (a) Shaker, O. G.; Moustafa, W.; Essmat, S.; Abdel-Halim, M.;
El-Komy, M. The role of interleukin-12 in the pathogenesis of
psoriasis. Clin. Biochem. 2006, 39, 119−125. (b) Piskin, G.; Sylva-
Steenland, R. M. R.; Bos, J. D.; Teunissen, M. B. M. In vitro and in situ
expression of IL-23 by keratinocytes in healthy skin and psoriasis
lesions: enhanced expression in psoriatic skin. J. Immunol. 2006, 176,
4534
dx.doi.org/10.1021/jm400266t | J. Med. Chem. 2013, 56, 4521−4536

Journal of Medicinal Chemistry
Article
30 distinct susceptibility loci for Crohn’s disease. Nat. Genet. 2008, 40,
955−962. (b) Duerr, R. H.; Taylor, K. D.; Brant, S. R.; Rioux, J. D.;
Silverberg, M. S.; Daly, M. J.; Steinhart, A. H.; Abraham, C.; Regueiro,
M.; Griffiths, A.; Dassopoulos, T.; Bitton, A.; Yang, H.; Targan, S.;
Datta, L. W.; Kistner, E. O.; Schumm, L. P.; Lee, A. T.; Gregersen, P.
K.; Barmada, M. M.; Rotter, J. I.; Nicolae, D. L.; Cho, J. H. A genome-
wide association study identifies IL23R as an inflammatory bowel
disease gene. Science 2006, 314, 1461−1463.
(26) (a) Hue, S.; Ahern, P.; Bounocore, S.; Kullberg, M. C.; Cua, D.
J.; McKenzie, B. S.; Pwrie, F.; Maloy, K. J. Interleukin-23 drives innate
and T cell-mediated intestinal inflammation. J. Exp. Med. 2006, 203,
2473−2483. (b) Kullberg, M. C.; Jankovic, D.; Feng, C. G.; Hue, S.;
Gorelick, P. L.; McKenzie, B. S.; Cua, D. J.; Powrie, F.; Cheever, A. W.;
Maloy, K, J.; Sher, A. IL-23 plays a key role in Helicobacter hepaticus-
induced T cell-dependent colitis. J. Exp. Med. 2006, 203, 2485−2494.
(27) Sandborn, W. J.; Feagan, B.; Fedorak, R. N.; Scherl, E.; Fleisher,
M. R.; Katz, S.; Johanns, J.; Blank, M.; Rutgeerts, P. A randomized trial
of ustekinumab, a human interleukin-12/23 monoclonal antibody, in
patients with moderate-to-severe Crohn’s disease. Gastroenterology
2008, 135, 1130−1141.
(28) (a) Kralovics, R.; Passamonti, F.; Buser, A. S.; Teo, S.-S.; Tiedt,
R.; Passweg, J. R.; Tichelli, A.; Cazzola, M.; Skoda, R. A gain-of-
function mutation of JAK2 in myeloproliferative disorders. N. Engl. J.
Med. 2005, 352, 1779−1790. (b) James, C.; Ugo, V.; Couedic, J.-P.;
Staerk, J.; Delhommeau, F.; Lacout, C.; Garcon, L.; Raslova, H.;
Berger, R.; Bennaceur-Griscelli, A.; Villeval, J. L.; Constantinescu, S.
N.; Casadevall, N.; Vainchenker, W. A unique clonal JAK2 mutation
leading to constitutive signaling causes polytheamia vera. Nature 2005,
434, 1144−1148.
Meraz, M. A.; White, J. M.; Lampe, P. A.; Riley, J. K.; Arthur, C. A.;
King, K. L.; Sheehan, K. C. F.; Yin, L.; Pennica, D.; Johnson, E. M.;
Schreiber, R. D. Disruption of the JAK1 gene demonstrates obligatory
and nonredundant roles of the JAKs in cytokine-induced biologic
response. Cell 1998, 93, 373−383.
(35) (a) Comejo, M. G.; Boggon, T. J.; Mercher, T. JAK3: a two-
faced player in hematological disorders. Int. J. Biochem. Cell Biol. 2009,
41, 2376−2379. (b) Pesu, M.; Candotti, F.; Husa, M.; Hofmann, S. R.;
Notarangelo, L. D.; O’Shea, J. J. JAK3, severe combined
immunodeficiency, and a new class of immunosuppressive drugs.
Immunol. Rev. 2005, 203, 127−142.
(36) (a) Candotti, F.; Oakes, S. A.; Johnson, J. A.; Giliani, S.;
Schumacher, R. F.; Mella, P.; Fioini, M.; Ugazio, A. G.; Badolato, R.;
Notarangelo, L. D.; Bozzi, F.; Macchi, P.; Strina, D.; Vezzoni, P.;
Blaese, R. M.; O’Shea, J. J.; Villa, A. Structural and functional basis for
JAK3-deficient severe combined immunodeficiency. Blood 1997, 90,
3996−4003. (b) Russell, S. M.; Tayebi, N.; Nakajima, H.; Riedy, M.
C.; Roberts, J.; Aman, M. J.; Migone, T.-S.; Noguchi, M.; Markert, M.
L.; Buckley, R. H.; O’Shea, J. J.; Leonard, W. J. Mutation of JAK3 in a
patient with SCID: essential role of JAK3 in lymphoid development.
Science 1995, 270, 797−800.
(37) Flanagan, M. E.; Blumenkopf, T. A.; Brissette, W. H.; Brown, M.
F.; Casavant, J. M.; Shang-Poa, C.; Doty, J. L.; Elliott, E. A.; Fisher, M.
B.; Hines, M.; Kent, C.; Kudlacz, E. M.; Littie, B. M.; Magnusson, K.
S.; McCurdy, S. P.; Munchhof, M. J.; Perry, B. D.; Sawyer, P. S.;
Strelevitz, T. J.; Subramanyam, C.; Sun, J.; Whipple, D. A.; Changelian,
P. S. Discovery of CP-690,550: a potent and selective Janus kinase
(JAK) inhibitor for the treatment of autoimmune diseases and organ
transplant rejection. J. Med. Chem. 2010, 53, 8468−8484.
(29) (a) Parganas, E.; Wang, D.; Stravopodis, D.; Topham, D. J.;
Marine, J. C.; Teglund, S.; Vanin, E. F.; Bodner, S.; Colamonici, O. R.;
van Deursen, J. M.; Grosveld, G.; Ihle, J. N. JAK2 is essential for
signaling through a variety of cytokine receptors. Cell 1998, 93, 385−
395. (b) Neubauer, H.; Cumano, A.; Muller, M.; Wu, H.; Huffstadt,
U.; Pfeffer, K. JAK2 deficiency defines an essential developmental
checkpoint in definitive hematopoiesis. Cell 1998, 93, 397−409.
(30) (a) Tefferi, A. JAK inhibitors for myeloproliferative neoplasms:
clarifying facts from myths. Blood 2012, 119, 2721−2730.
(b) Pardanani, A.; Gotlib, J.; Jamieson, C.; Cortes, J. E.; Talpaz, M.;
Stone, R. M.; Silverman, M. H.; Gilliland, D. G.; Shorr, J.; Tefferi, A.
Safety and efficacy of TG101348, a selective JAK2 inhibitor, in
myelofibrosis. J. Clin. Oncol. 2011, 29, 789−796. (c) Verstovsek, S.;
Kantarjian, H.; Mesa, R. A.; Pardanani, A.; Cortes-Franco, J.; Thomas,
D. A.; Estrov, Z.; Fridman, J. S.; Bradley, E. C.; Erickson-Viitanen, S.;
Vaddi, K.; Levy, R.; Tefferi, A. Safety and efficacy of INCB018424, a
JAK1 and JAK2 inhibitor, in myelofibrosis. N. Eng. J. Med. 2010, 363,
1117−1127.
(31) Ishizaki, M.; Akimoto, T.; Muromoto, R.; Yokoyama, M.;
Ohshira, Y.; Sekine, Y.; Maeda, H.; Shimoda, K.; Oritani, K.; Matsuda,
T. Involvement of tyrosine kinase-2 in both the IL-12/Th1 and IL-23/
Th17 axes in vivo. J. Immunol. 2011, 187, 181−89.
(32) Minegishi, Y.; Saito, M.; Morio, T.; Watanabe, K.; Agematsu, K.;
Tsuchiya, S.; Takada, H.; Hara, T.; Kawamura, N.; Ariga, T.; Kaneko,
H.; Kondo, N.; Tsuge, I.; Yachie, A.; Sakiyama, Y.; Iwata, T.; Bessho,
F.; Ohishi, T.; Joh, K.; Imai, K.; Kogawa, K.; Shinohara, M.; Fujieda,
M.; Wakiguchi, H.; Pasic, S.; Abinun, M.; Ochs, H. D.; Renner, E. D.;
Jansson, A.; Belohradsky, B. H.; Melin, A.; Shimizu, N.; Mizutani, S.;
Miyawaki, T.; Nonoyama, S.; Karasuyama, H. Human tyrosine kinase
2 deficiency reveals its requisite roles in multiple cytokine signals
involved in innate and acquired immunity. Immunity 2006, 25, 745−
755.
(38) Quintas-Cardama, A.; Vaddi, K.; Liu, P.; Manshouri, T.; Li, J.;
Scherle, P. A.; Caulder, E.; Wen, X.; Li, Y.; Waeltz, P.; Rupar, M.;
Burn, T.; Lo, Y.; Kelley, J.; Covington, M.; Shepard, S.; Rodgers, J. D.;
Haley, P.; Kantarijian, H.; Fridman, J. S.; Verstovsek, S. Preclinical
characterization of the selective JAK1/2 inhibitor INCB018424:
therapeutic implications for the treatment of myeloproliferative
neoplasms. Blood 2010, 115, 3109−3117.
(39) Alicea-Velazquez, N. L.; Boggon, T. J. The use of structural
biology in Janus kinase targeted drug discovery. Curr. Drug Targets
2011, 12, 546−555.
(40) Goodacre, S. C.; Lai, Y.; Liang, J.; Magnuson, S. R.; Robarge, K.
D.; Stanley, M. S.; Tsui, V.; Williams, K.; Zhang, B.; Zhou, A. Janus
Kinase Inhibitor Compounds and Methods. WO2010142752, 2010.
The following patent applications describing TYK2 inhibitors were
also published: (a) Buchmann, B.; Eis, K., Bothe, U.; Bonin, A.;
Bomer, U.; Carbonylamino-Substituierte Aniline-Pyromidinderivative
als TYK-Inhibitoren, deren Herstellung und Verwendug als
Arzneimittel. DE2009001438, 2009. (b) Boys, M, L.; Burgess, L. E.;
Groneberg, R. D.; Harvey, D. M.; Huang, L.; Kercher, T.; Kraser, C.
F.; Laird, E.; Tarlton, E.; Zhao, Q. 5,7-Substituted-imidazo[1,2-
c]pyrimidines as Inhibitors of JAK Kinases. WO2011130146, 2011.
(c) Ellard, K.; Major, J.; Jones, A.; Lynch, R.; Ramsden, N. Pyridine
Compounds and Aza Analogs Thereof as TYK2 Inhibitors.
WO2012062704, 2012.
(41) Liang, J.; Tsui, V.; van Abbema, A.; Bao, L.; Barrett, K.; Beresini,
M.; Berezhkovsky, B.; Blair, W.; Chang, C.; Driscoll, J.; Eigenbrot, C.;
Ghilardi, N.; Gibbons, P.; Halladay, J.; Johnson, A.; Kohli, P. B.; Lai,
Y.; Liimatta, M.; Mantik, P.; Menghrajani, K.; Murray, J.; Sambrone,
A.; Shao, Y.; Shia, S.; Shin, Y.; Smith, J.; Sohn, S.; Stanley, M.; Ultsch,
M.; Zhang, B.; Magnuson, S. Lead identification of novel and selective
TYK2 inhibitors. Eur. J. Med. Chem., accepted.
(42) Our compounds were found to be consistently much less active
against the JAK3 enzyme than TYK2 (>50-fold difference between the
JAK3 Ki and TYK2 Ki). JAK3 inhibitory activities were not reported
for the majority of compounds described in this work except for
compound 3 (Figure 1) and compounds 35−37 (Table 3). Recent
reports also suggested that JAK3 inhibition is not sufficient in blocking
effects of type I cytokines and that JAK1 is the dominant kinase:
(a) Haan, C.; Rovlering, C.; Rault, F.; Kapp, M.; Druckes, P.; Thoma,
G.; Behrmann, I.; Zerwes, H.-G. JAK1 has a dominant role over JAK3
(33) Sohn, S. J.; Barrett, K.; van Abbema, A.; Chang, C.; Kohli, P. B.;
Kanda, H.; Smith, J.; Lai, Y.; Zhou, A.; Williams, K.; Macleod, C.;
Hurley, C. A.; Kulagowski, J. K.; Lewis-Koh, N.; Johnson, A. R.;
Ghilardi, N.; Zak, M.; Liang, J.; Blair, W. S.; Magnuson, S.; Wu, L. C. A
restricted role for TYK2 catalytic activity in human cytokine response.
J. Immunol., submitted.
(34) (a) Tanaka, T.; Kishimoto, T. Immunotherapeutic implication
of IL-6 blockade. Immunotherapy 2012, 4, 87−105. (b) Rodig, S. J.;
4535
dx.doi.org/10.1021/jm400266t | J. Med. Chem. 2013, 56, 4521−4536

Journal of Medicinal Chemistry
Article
in signal transduction through γc-containing cytokine receptors. Chem.
Biol. 2011, 18, 314−323. (b) Thoma, G.; Nuninger, F.; Falchetto, R.;
Hermes, E.; Tavares, G. A.; Vangrevelinghe, E.; Zerwes, H.-G.
Identification of a potent Janus kinase 3 inhibitor with high selectivity
within the Janus kinase family. J. Med. Chem. 2011, 54, 284−288.
(43) (a) Hughes, J. D.; Blagg, J.; Price, D. A.; Bailey, S.;
DeCrescenzo, G. A.; Devraj, R. V.; Ellworth, E.; Fobian, Y. M.;
Gibbs, M. E.; Gilles, R. W.; Greene, N.; Huang, E.; Krieger-Burke, T.;
Loesel, J.; Wager, T.; Whiteley, L.; Zhang, Y. Physiochemical drug
properties associated with in vivo toxicological outcomes. Bioorg. Med.
Chem. Lett. 2008, 18, 4872−4875. (b) Leeson, P. D.; Spingthorpe, B.
The influence of drug-like concepts on decision-making in medicinal
chemistry. Nat. Rev. Drug Discovery 2007, 6, 881−890.
fluorocyclopropyl)quinolone antibacterial agents. J. Med. Chem.
1994, 37, 3344−3352.
(55) Pierce, A. C.; Sandretto, K. L.; Bemis, G. W. Kinase inhibitors
and the case for CH...O hydrogen bonds in protein-ligand binding.
Proteins 2002, 49, 567−576.
(44) Kinase P-loop is known to be flexible: (a) Taylor, S. S.; Kornev,
A. P. Protein kinases: evolution of dynamic regulatory protein. Trends
Biochem. Sci. 2011, 36, 65−77. (b) Kornev, A. P.; Taylor, S. S. Defining
the conserved internal architecture of a protein kinase. Biochim.
Biophys. Acta 2010, 1804, 440−444.
(45) Cruciani, G.; Carosati, E.; De Boeck, B.; Ethirajulu, K.; Mackie,
C.; Howe, T.; Vianello, R. Metasite: understanding metabolism in
human cytochromes from the perspective of the chemist. J. Med. Chem.
2005, 48, 6970−6979.
(46) (a) Hill, A. P.; Young, R. J. Getting physical in drug discovery: a
contemporary perspective on solubility and hydrophobicity. Drug
Discovery Today 2010, 15, 648−655. (b) Richie, T. J.; MacDonald, S. J.
F. The impact of aromatic ring count on compound developability
Are too many aromatic rings a liability in drug design? Drug Discovery
Today 2009, 14, 1011−1020.
(47) Paulini, R.; Muller, K.; Diederich, F. Orthogonal multipolar
̈
interactions in structural chemistry and biology. Angew. Chem., Int. Ed.
2005, 44, 1788−1805.
(48) Kuhn, B.; Fuchs, J. E.; Reutlinger, M.; Stahl, M.; Taylor, N. R.
Rationalizing tight ligand binding through cooperative interaction
networks. J. Chem. Inf. Model. 2011, 51, 3180−3198.
(49) (a) Tsui, V.; Gibbons, P.; Ultsch, M.; Mortara, K.; Chang, C.;
Blair, W.; Pulk, R.; Stanley, M.; Starovasnik, M.; Williams, D.; Lamers,
M.; Leonard, P.; Magnuson, S.; Liang, J.; Eigenbrot, C. A new
regulatory switch in a JAK protein kinase. Proteins: Struct., Funct.,
Bioinf. 2011, 79, 393−401. (b) Chrencik, J. E.; Patny, A.; Leung, I. K.;
Korniski, B.; Emmons, T. L.; Hall, T.; Weinberg, R. A.; Gormley, J. A.;
Williams, J. M.; Day, J. E.; Hirsch, J. L.; Kiefer, J. R.; Leone, J. W.;
Fischer, H. D.; Sommers, C. D.; Huang, H.-C.; Jacobsen, E. J.;
Tenbrink, R. E.; Tomasselli, A. G.; Benson, T. E. Structural and
thermodynamic characterization of the TYK2 and JAK3 kinase
domains in complex with CP-690550 and CMP-6. J. Mol. Biol. 2010,
400, 413−433.
(50) Knight, Z. A.; Shokat, K. M. Features of selective kinase
inhibitors. Chem. Biol. 2005, 12, 621−637.
(51) Graczyk, P. P. Gini coefficient: a new way to express selectivity
of kinase inhibitors against a family of kinases. J. Med. Chem. 2007, 50,
5773−5779.
(52) (a) Bohm, H.-J.; Banner, D.; Bendels, S.; Kansy, M.; Kuhn, B.;
̈
Muller, K.; Obst-Sander, U.; Stahl, M. Fluorine in medicinal chemistry.
̈
ChemBioChem 2004, 5, 637−643. (b) Purser, S.; Moore, P. R.;
Swallow, S.; Gouverneur, V. Fluorine in medicinal chemistry. Chem.
Soc. Rev. 2008, 37, 320−330. (c) Hagmann, W. K. The many roles for
fluorine in medicinal chemistry. J. Med. Chem. 2008, 51, 4359−4369.
(53) The relatively small shift (4.4-fold) of potency between PBMC
and human whole blood for compound 37 is likely due to its modest
(86.6%) human plasma protein binding (Table 4). A nice review of
“free drug principle” with many recent examples could be found:
Trainor, G. L. Expert Opin. Drug Discovery 2007, 2, 51−64.
(54) Synthesis and chiral resolution of 2-fluorocyclopropanecarbox-
ylic acid have been reported: (a) Akiba, T.; Ikeya, T.; Kawanishi, H.;
Yukimoto, Y.; Kamihara, S.; Ebata, T. Selective Dehalogenation
Process. U.S. Patent 5,780,669, 1998. (b) Kimura, Y.; Atarashi, S.;
Kawakami, K.; Sato, K.; Hayakawa, I. (Fluorocyclopropyl)quinolones.
2. Synthesis and stereochemical structure−activity relationship of
chiral 7-(7-amino-5-azaspiro[2.4]heptan-5-yl)-1-(2-
4536
dx.doi.org/10.1021/jm400266t | J. Med. Chem. 2013, 56, 4521−4536