Design of a Janus Kinase 3 (JAK3) Specific Inhibitor 1-((2S,5R)-5-((7H-Pyrrolo[2,3-d]pyrimidin-4-yl)amino)-2-methylpiperidin-1-yl)prop-2-en-1-one (PF-06651600) Allowing for the Interrogation of JAK3 Signaling in Humans

Article abstract of DOI:10.1021/acs.jmedchem.6b01694

Significant work has been dedicated to the discovery of JAK kinase inhibitors resulting in several compounds entering clinical development and two FDA approved NMEs. However, despite significant effort during the past 2 decades, identification of highly s

Full text of DOI:10.1021/acs.jmedchem.6b01694

Subscriber access provided by UB + Fachbibliothek Chemie | (FU-Bibliothekssystem)
Article
The design of a Janus Kinase 3 (JAK3) specific inhibitor 1-((2S,5R)-5-((7H-
pyrrolo[2,3-d]pyrimidin-4-yl)amino)-2-methylpiperidin-1-yl)prop-2-en-1-one
(PF-06651600) allowing for the interrogation of JAK3 signaling in humans
Atli Thorarensen, Martin E. Dowty, Mary Ellen Banker, Brian M Juba, Jason Jussif, Tsung
Lin, Fabien Vincent, Robert M. Czerwinski, Agustin Casimiro-Garcia, Ray Unwalla, John I
Trujillo, Sidney Liang, Paul Balbo, Ye Che, Adam M. Gilbert, Matthew F. Brown, Matthew
Hayward, Justin Montgomery, Louis Leung, Xin Yang, Sarah Soucy, Martin Hegen, Jotham
Wadsworth Coe, Jonathan Langille, Felix F. Vajdos, Jill E Chrencik, and Jean-Baptiste Telliez
J. Med. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.jmedchem.6b01694 • Publication Date (Web): 31 Jan 2017
Downloaded from http://pubs.acs.org on January 31, 2017
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a free service to the research community to expedite the
dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts
appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been
fully peer reviewed, but should not be considered the official version of record. They are accessible to all
readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered
to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published
in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just
Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor
changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers
and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors
or consequences arising from the use of information contained in these “Just Accepted” manuscripts.
Journal of Medicinal Chemistry is published by the American Chemical Society. 1155
Sixteenth Street N.W., Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.

Page 1 of 72
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 2 of 72
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
The design of a Janus Kinase 3 (JAK3) specific inhibitor 1-((2S,5R)-5-((7H-pyrrolo[2,3-d]pyrimidin-
4-yl)amino)-2-methylpiperidin-1-yl)prop-2-en-1-one (PF-06651600) allowing for the interrogation
of JAK3 signaling in humans
1
2
3
4
4
4
Atli Thorarensen *, Martin E. Dowty , Mary Ellen Banker , Brian Juba , Jason Jussif , Tsung Lin , Fabien
3
4
1
1
5
5
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Vincent , Robert M Czerwinski , Agustin Casimiro-Garcia , Ray Unwalla , John Trujillo , Sidney Liang ,
4
5
5
5
5
5
Paul Balbo , Ye Che , Adam M. Gilbert , Matthew F. Brown , Matthew Hayward , Justin Montgomery ,
2
2
4
4
5
5
5
Louis Leung , Xin Yang , Sarah Soucy , Martin Hegen , Jotham Coe , Jonathan Langille , Felix Vajdos , Jill
5
4
Chrencik , Jean-Baptiste Telliez *
1
Worldwide Medicinal Chemistry, Pfizer Worldwide R&D, 610 Main Street, Cambridge, MA 02139 (USA).
2
Pharmacokinetics, Dynamics, and Metabolism, Pfizer Worldwide R&D, 1 Burtt Rd, Andover, MA 01810
and Eastern Point Road, Groton, CT 06340 (USA)
3
Primary Pharmacology Group, Pfizer Worldwide R&D, Eastern Point Road, Groton, CT 06340 (USA).
4
Inflammation and Immunology, Pfizer Worldwide R&D, 610 Main Street, Cambridge, MA 02139 (USA).
5
Worldwide Medicinal Chemistry, Pfizer Worldwide R&D, Eastern Point Road, Groton, CT 06340 (USA).
Abstract: Significant work has been dedicated to the discovery of JAK kinase inhibitors resulting in
several compounds entering clinical development and two FDA approved NMEs. However, despite
significant effort during the past two decades, identification of highly selective JAK3 inhibitors has
eluded the scientific community. A significant effort within our research organization has resulted in the
identification of the first orally active JAK3 specific inhibitor, which achieves JAK isoform specificity
through covalent interaction with a unique JAK3 residue Cys-909. The relatively rapid resynthesis rate of
the JAK3 enzyme presented a unique challenge in the design of covalent inhibitors with appropriate
pharmacodynamics properties coupled with limited unwanted off-target reactivity. This effort resulted
in the identification of 11 (PF-06651600), a potent and low clearance compound with demonstrated in
vivo efficacy. The favorable efficacy and safety profile of this specific JAK3 inhibitor 11 led to its
evaluation in several human clinical studies.
ACS Paragon Plus Environment

Page 3 of 72
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Introduction
Janus kinases (JAK) are non-receptor tyrosine kinases required for signaling through type I/II
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
1
cytokine receptors. There are four JAK family members: JAK1, JAK2, JAK3 and TYK2. JAK1, JAK2 and
TYK2 are ubiquitously expressed while JAK3 expression is largely restricted to leukocytes. All four
isoforms are involved in immune related functions while JAK1 and JAK2 also play critical roles related to
hematopoiesis, growth and neuronal functions amongst others. A notable feature of JAK signaling is the
requirement for a dimer of JAK kinases within the cytokine receptor complex, and cytokine receptors
signaling through JAK1/JAK3, JAK1/JAK2, JAK1/TYK2, JAK2/TYK2 and JAK2/JAK2 have been described.
The importance of the individual JAKs in cytokine signaling has resulted in significant efforts to
2,3
identify isoform selective inhibitors.
Several inhibitors have entered clinical evaluation for
autoimmune diseases as well as various cancers and myeloproliferative disorders, resulting in two FDA
4
approvals to date (Tofacitinib (1) and Ruxolitinib (2), (Figure 1). JAK3 was among the first of the JAKs
targeted for therapeutic intervention due to the strong validation provided by human SCID patients
5
displaying JAK3 deficiencies. That said, the majority of inhibitors reported to date lack selectivity within
the JAK family. Therefore, the therapeutic potential of JAK3 selective inhibitors remains to be addressed.
Indeed, in recent years there has been a debate on the role of JAK3 versus JAK1 for the common
6
gamma-chain cytokines utilizing this heterodimeric kinase pair. Recently, we illustrated that a selective
JAK3 inhibitor could effectively inhibit IL-15 mediated signaling in vitro, reaffirming the therapeutic
7
potential of selective JAK3 inhibition.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 4 of 72
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Figure 1. FDA approved JAK inhibitors
Sequence alignment of the ATP binding site for the four JAK family members illustrates that
JAK3 has only two unique residues that could be utilized to design selective inhibitors (4). In addition,
available crystal structures for each JAK isoform highlight very minor conformational differences
8
amongst the set. This structural information, combined with the higher ATP affinity of JAK3 versus
other family members, illustrates the extraordinary challenge associated with the design of highly
7
selective, ATP competitive JAK3 inhibitors. The unique residue Cys-909 in JAK3 provides an opportunity
to engineer selectivity for JAK3 in a covalent manner to overcome the high sequence and shape
similarity within the ATP binding cavity. This idea has been reported in the literature, most notably by
9,10
the disclosure of highly selective JAK3 covalent inhibitors with supporting in vitro characterization.
Targeting JAK3 in a covalent manner provides alternative dimensions in drug design, such as the
potential for an extended pharmacodynamic (PD) effect following target inactivation. Covalent
modification also presents unique liabilities, such as off-target reactivity which could lead to undesired
pharmacology and non-traditional metabolic clearance. We recently disclosed the selective covalent
JAK3 inhibitor 11 (PF-06651600) and summarized efficacy results from multiple pre-clinical in vivo
20
models of inflammatory diseases. This highly optimized covalent inhibitor of JAK3 provides rapid
kinase inactivation while possessing minimal chemical reactivity, resulting in an inhibitor with good
ADME properties and overall selectivity. In this paper, we describe the research efforts which
culminated in the discovery of 11.
ACS Paragon Plus Environment

Page 5 of 72
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Results and discussion
Initial design and impact of reactivity on translation. The crystal structure of 1 in complex with the
kinase domain of JAK3 provided the inspiration for the design of inhibitors that would interact in a
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
1
1
covalent manner with Cys-909. Rotation around the pyrimidyl amine N-C bond would potentially place
the amide substituent in close proximity with Cys-909, where an appropriately designed electrophile
could capture the nucleophilic cysteine (Figure 2a). Placing an acrylamide on the tofacitinib core
resulted in compound 3, which had modest inhibition of JAK3, but poor JAK family selectivity, indicating
that the affinity was mainly due to reversible (Ki) protein interactions (Table 1). This hypothesis was
supported by X-ray crystal structure analysis, which indicated that the pyrolopyrimidine hinge binder
scaffold was well-ordered in the active site, while the acrylamide substituted piperidine was highly
disordered (data not shown). Since JAK3 has the highest affinity for ATP among its family members, the
Ki component of target inhibition must not be the major driver for potency as that would only provide
modest selectivity gains through covalent interaction with Cys-909. In order to increase the selectivity
for JAK3, the inactivation (Kinact) rate had to be improved. Modeling studies indicated that removal of
the N-methyl and piperidine cis methyl groups would result in an analog with a low energy
1
2
conformation, placing the electrophilic warhead in close proximity to Cys-909, Figure 2b. Indeed,
preparation of compound 4 revealed a significant improvement in potency (JAK3 IC = 56 nM, [ATP] =
50
1
mM). An IC can be an excellent measure of potency of irreversible inhibitors when produced under
50
conditions of relatively long assay reaction times, as this enables a good correlation with the time
13,14
independent potency measurement Kinact/Ki.
In fact, the greater ease associated with IC50
determination and the logical translation to functional responses in both in vitro cell-based and in vivo
assays renders IC50 the parameter of choice for SAR determination. The high JAK3 selectivity for 4
relative to 3 indicated that a significant improvement in the Kinact parameter had been achieved. The
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 6 of 72
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
covalent bond forming event of 4 with Cys-909 was further verified in a follow on study where the
compound was shown to have a dissociation rate of > 8 h.
The concept of high JAK3 specificity utilizing a covalent handle was extended via a range of
electrophiles and motifs that would place the electrophile appropriately in the vicinity of Cys-909
resulting in analogs such as the anilinoacrylamide 6, which demonstrated potent JAK3 inhibition (IC =
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
50
2
9 nM). Compounds 4 and 6 represent two different series (i.e. aliphatic and aryl aminoacrylamides)
that were explored in parallel. While the majority of the potency for each series was obtained through
rapid covalent inactivation (i.e. Kinact), each achieved this via significantly different principles. Our
quantification of electrophilicity utilizing glutathione (GSH) as a trapping reagent illustrated that the two
series have very different inherent chemical reactivities (aryl-aminoacrylamides: T1/2 = 0.4 - 4h; aliphatic
15
aminoacrylamides: 4 – 40 h). Aryl aminoacrylamides have been successfully utilized in the design of
9
16
17
irreversible inhibitors for JAK3 ) EGFR and BTK resulting in potent inhibitors, therefore, it was unclear
how differing relative reactivities would translate to off-target effects. The impact of reactivity became
quite apparent when evaluating the translation from a cell-based PBMC assay to one conducted in
human whole blood, a medium that is rich in proteins (Figure 3a). While the aliphatic aminoacrylamides
demonstrated a solid correlation between assays, the aryl aminoacrylamides provided a poor
translation. The reason for the differential translation for aliphatic vs. aryl aminoacrylamides has not
been fully elucidated, however, the enhanced reactivity of aryl aminoacrylamides described above was
likely a major contributor. In addition, blood stability studies conducted with representatives from both
series demonstrated a similar trend, (Figure 3b). Therefore, while the aryl aminoacrylamide class of
compounds generated some of our most potent analogs in enzymatic and PBMC cell-based assays, it
was clear that the level of chemical reactivity inherent to this series needed to be attenuated to provide
optimal potency in a whole blood environment.
ACS Paragon Plus Environment

Page 7 of 72
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
2
a)
N
R
1
R1
N
N
N
180° rotation
N
N
O
O
N
N
N
N
H
H
Conformer A
Conformer B
Conformer A
(Kcal/mol)
Conformer B
(Kcal/mol)
Cmpd#
R1
H
4
5
0.0
0.6
Me
0.0
0.02
2
b)
Figure 2. 2a) Crystal structure of tofacitinib in JAK3 illustrating key movement around the N-C
axis that would place the amide in close proximity to Cys-909. 2b) Illustrates the dependency of the NMe
group on conformational energetics with conformer A being favored by 70:30 for 4.
Table 1. Design of covalent inhibitors of JAK3 with JAK enzyme inhibition data
Ilꢀ15
PBMC
(cell)
ILꢀ15
JAK3
IC50
(
JAK3
IC50
(nM)*
JAK1
IC50
(nM)
Km#
JAK1
IC50
(nM)*
JAK2
IC50
TYK2
IC50
(nM)*
Cmpd
#
HWB
(cell)
IC50
(nM)
Structure
nM)
(nM)*
IC50
Km♦
(
nM)
3
75
1930
274
988
2030
>10000
ND
ND
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 8 of 72
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Ilꢀ15
PBMC
(cell)
ILꢀ15
JAK3
IC50
JAK3
IC50
(nM)*
JAK1
IC50
(nM)
Km#
JAK1
IC50
(nM)*
JAK2
IC50
(nM)*
TYK2
IC50
(nM)*
Cmpd
#
HWB
(cell)
IC50
(nM)
Structure
(
nM)
IC50
Km♦
(
nM)
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
4
5
0.7
56
1400
>10000 >10000
>10000
>10000
107
ND
580
18
3840
1150
7820
8710
> 10000
>10000
ND
6
7
8
0.3
29
47
966
984
863
>10000
>10000
>10000
32
2280
399
ND
0.4
>10000 >10000
116
ND
372
>10000
>10000 >10000
9
1610
>10000
2290
1640
>10000 >10000
>10000
>10000
ND
51
ND
197
1
1
0.3
33
>10000 >10000
ACS Paragon Plus Environment

Page 9 of 72
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
The enzyme assay were performed using the following description: ♦ATP concentration for
JAK3 is 4 uM at Km, # ATP concentration for JAK1 is 40 uM at Km, * these assays utilized 1
mM ATP.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 3. 3a) Correlation between our IL-15 PBMC assay and the corresponding HWB assay.
Anilinoacrylamide = blue, aliphaticacrylamide = green. 3b) The stability of the compounds in HWB.
Legend T1/2 <100 min magenta, 100-200 min tan, 200-300 min yellow, >300 min green (n=107)
Identification of key parameters leading to the discovery of orally active JAK3 specific inhibitors. The
identification of compound 4 illustrated an important guiding principle for future design. The analogs
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 10 of 72
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
would need to have a sufficient Ki component to effectively compete with ATP and initiate binding to
the protein, but that this could only be a minor component of enzyme inhibition. Rapid inactivation
(Kinact) continued to be the major driver for optimization as it was determined to be the key for
achieving sufficient selectivity. Simply increasing chemical reactivity was deemed an unproductive
strategy for Kinact optimization given the results discussed above for the aryl aminoacrylamides. When
considering the individual events involved in covalent adduct formation, the initial bound complex EI
would not necessarily represent the reactive complex EI* (Figure 4). Therefore, increasing the
concentration of the reversibly bound EI* complex should result in acceleration of covalent capture.
The increase in chemical reactivity needed to obtain potent compounds in the series represented by 6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
18
likely indicated that a low concentration of the EI* complex was present. The design principle of
increased EI* complex concentration can be illustrated with compound 4. The crystal structures of 4
and 6 bound to JAK3 provided significant guidance for additional design with a focus on stabilization of
the reactive conformation (Figure 5a-b). Both of the compounds are Cys-909 adducts, but their
conformations are significantly different with regard to acrylamide alignment. For 4, the acrylamide is
observed making two hydrogen bonds; one to a conserved water molecule that is interacting with the
ligand pyrimidine nitrogen and the second to the NH of the Cys-909 amide. While these observations
are post covalent bond formation, our assumption is that they exist in the EI* complex as well,
stabilizing the reactive conformation and providing activation of the acrylamide for addition to the Cys-
9
09. In contrast, the acrylamide in 6 does not retain any of these interactions, thus requiring increased
inherent chemical reactivity to capture the Cys-909. These insights were applied to the design of a
more ring constrained compound 7 providing a potent inhibitor of JAK3. The X-ray crystal structure of 7
bound to Jak3 illustrates that the cyclization of the piperidine ring back to the linker amine constrains
the rotation around the piperidine-N bond, and stabilizes the piperidine in the appropriate chair
conformation to present the acrylamide for nucleophilic attack by Cys-909, Figure 5c.
ACS Paragon Plus Environment

Page 11 of 72
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
1
9
JAK1, JAK2, and TYK2 resynthesis rates are fairly rapid and in the range of 2 - 4 h. The lack of
information for JAK3 prompted measurement of the resynthesis rate in PBMC cells, which was
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
2
0
determined to be 3.5 h. This relatively rapid resynthesis rate indicated that the potential for an
extended pharmacodynamic (PD) effect due to covalent inhibition was out of scope. This is contrary to
21
22
our experience with slowly resynthesized biological targets such as FAAH and BTK . In these cases,
relatively short-term drug exposure could provide long-term target knockdown in vivo. For JAK3,
excellent pharmacokinetics (PK) with low clearance would be required in order to achieve sustained
target inactivation and a high level efficacy. Compound 4 (data not shown) and 7 suffered from
moderate to poor PK in rat despite very reasonable oxidative stability as judged by liver microsomes,
(
Table 2). A set of PK experiments with compound 7 and the corresponding acetate derivative 8
provided key insights into drug clearance mechanisms. For both compounds, very high clearance was
observed in an in vivo rat PK experiment (Table 3). In order to gauge the contribution of oxidative
metabolism to overall clearance, the same experiment was performed following the administration of
the broad cytochrome P450 (CYP450) inhibitor aminobenzotriazole (ABT). In this instance, the acetate 8
illustrated a significant reduction in clearance while acrylamide 7 was still highly cleared, suggesting the
presence of non-CYP450 clearance pathways. Indeed, glutathione-S-transferase (GST)-mediated
23
glutathione (GSH) addition to the acrylamide was found to account for this discrepancy. In order to
build an SAR understanding of GST-mediated clearance, a whole blood stability assay was developed to
provide an in vitro surrogate measure of whole body GST mediated clearance. The validation of this
24
assay as a predictive tool has been described recently.
Despite the poor PK of compound 7, it was evaluated in an in vivo study utilizing the mouse
Delayed Type Hypersensitivity (DTH) which is a T-cell dependent preclinical model that was utilized to
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 12 of 72
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
characterize 7. In the prophylactic mouse DTH model, a Th1 response was induced by priming mice with
sheep red blood cells (sRBC) subcutaneously on Day 0 and a challenge of sRBC was injected to the right
hind footpads on Day 5 to elicit a measurable inflammatory response. Compound was dosed once daily
throughout the study. 7 significantly inhibited the DTH reaction in the right hind footpad in a dosage-
dependent manner, and efficacy was observed with dosages as low as 3 mg/kg twice daily. 1 was also
efficacious with twice-daily dosing at 30 mg/kg in this DTH model (Figure 6). This provided our
preliminary support that selective inhibition of JAK3 would result in in vivo functional consequences.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Figure 4. Illustration of individual binding and reactive events required for covalent modification
of Cys-909 in JAK3. E = enzyme, S = substrate, I = inhibitor
5a)
ACS Paragon Plus Environment

Page 13 of 72
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
5
b)
5
c)
Figure 5. Crystal structures of compounds bound to JAK3 with a covalent bound to Cys-909 5a) 4 PDB
code 5TTS, 5b) 6 PDB code 5TTV. 5c) 7 PDB code 5TTU
Table 2. ADME properties of JAK3 covalent inhibitors
HLM
(uL/min/mg)
8
RLM
(uL/min/mg)
<15
RRCK
(10ꢀ6 cm/sec)
8
Human blood
T1/2 (min)
~332*
Rat blood
T1/2 (min)
98
Cmpd#
4
7
8
12.2
<8
8
70
25
<15
20
16
15
192
NA
>360
164
NA
161
1
1
*
A broad range of stability was observed over multiple experiments (n=3 T1/2 = >260, >360, 303
Table 3. In vivo rat pharmacoketic clearance of compound 7 and 8, in the absence and presence
of the broad based CYP450 inhibitor ABT
Cmpd#
7
109
8
76
CL (mL/min/kg) - ABT
CL (mL/min/kg) + ABT
81
10
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 14 of 72
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
7
Figure 6. Efficacy of 7 Eight week old female BALB/c mice (Taconic) were sensitized with 2 x 10 sheep
red blood cells (sRBC) subcutaneously on day 0. Oral dosing began on day 0 prior to the sensitization and
8
continuing through day 6. The mice were challenged with 1 x 10 sRBC in the right hind foot pad and the
left hind footpad received an injection of sterile PBS. Paw thickness was measured prior to the injection
for a baseline measurement and then again at 24 hours via caliper. Paw thickness was significantly
reduced in a dose-dependent manner compared to the mice that were dosed with the vehicle. All
compounds were dosed twice daily except where otherwise noted.
Design of orally active JAK3 specific inhibitors. There is very limited knowledge regarding design
strategies to modulate GST mediated clearance, and generally speaking, the GSTs are a class of enzymes
25
that can tolerate a wide range of substrates. Early on, we attempted to inhibit chemical reactivity by
adding substituents to the acrylamide, as steric hindrance around the electrophile should reduce the
ability of GSTs to catalyze GSH addition. This approach provided a significant reduction in GST-mediated
clearance, but unfortunately, was not compatible with JAK3 affinity. For example, addition of a beta-
methyl to the acrylamide moiety of 4 resulted in a nearly inactive analog 9 (IC50 > 1000 nM). An
alternative approach was the incorporation of a remote steric handle to modulate GST binding; a
26
strategy utilized in a related series of cyanamide JAK3 inhibitors. This strategy was attempted with
ACS Paragon Plus Environment

Page 15 of 72
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
analog 4 wherein methyl substitution was explored at various piperidine ring positions. A key initial
analog was 10 (Table 5), which showed enhanced blood stability relative to the desmethyl analog 4, but
unfortunately was poorly active against JAK3. Our rationale for the poor activity involved destabilization
of the reactive conformation needed to capture the Cys-909 due to a potential steric clash with
sidechain residue Leu-956 in the ATP pocket. Our modeling efforts indicated that moving the
aminopyrrolopyrimidine group from the 2- to the 4-position would result in a highly favorable ground
state conformation for interaction with Cys-909 without a steric clash to any residue in the ATP pocket
of JAK3. In this conformation, we expected interaction of the acrylamide carbonyl with the NH of Cys-
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
9
09 further stabilizing the reactive conformation and potentially activating the acrylamide for
nucleophilic addition. This concept was realized in compound 11 which was found to be a highly specific
inhibitor of JAK3 (IC50 33 nM), devoid of relevant activity vs. other JAK enzymes at 1 mM ATP
27
concentrations, and demonstrated by X-ray crystallography to be covalently attached to Cys-909. The
high in vitro potency of 11 translated to excellent activity illustrated in both cellular (IL-15 stimulated
PBMCs, IC = 51 nM) and human whole blood assays (IL-15 stimulated HWB assay IC = 197 nM) (Table
50
50
1
). The importance of free drug concentration of 11 driving JAK3 pharmacology was apparent following
conversion of the whole blood IC to a free IC of 106 nM, which is within 2 fold of the PBMC IC . A
50
50
50
comparable profile of inhibition for other common γ-chain cytokine signaling was observed, while
20
exquisitely sparing the inhibition of cytokines which signal through other JAK kinase pairs. The
28
observed cellular functional potency correlated well with occupancy of the JAK3 enzyme by 11.
Importantly, the placement of the 2-methyl moiety resulted in a stable compound as judged by human
liver microsomes, hepatocytes and blood stability assays. In vivo pharmacokinetic studies demonstrated
low clearance of compound 11 in rat and dog, but rapid clearance in monkey (relative to liver blood
flow) and lower oral availability (Table 4). The pharmacokinetic profiles across species had a direct
relationship to both stability in hepatocytes and blood, illustrating the value of applying both in vitro
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 16 of 72
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
tools to predict in vivo pharmacokinetic behavior. The high oral bioavailability of 11 in a number of
species (rat/dog) was consistent with its high in vitro passive permeability (RRCK 15 x 10-6 cm/s), high
aqueous solubility > 2 mg/mL), and low hepatic clearance. The metabolic clearance of 11 was
determined to be mediated through both oxidative metabolism (acrylamide and ring oxidation primarily
through CYP3A4) and glutathione conjugation (on the acrylamide). Renal and biliary clearance of 11 was
low in rat. Through an understanding of clearance mechanisms and in vitro to in vivo correlations,
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
21
human blood clearance of 11 was predicted to be approximately 5.6 mL/min/kg. Human oral
bioavailability was likewise predicted to be approximately 90% with a half-life of approximately 2 hrs.
These results prompted a systematic evaluation of the impact of methyl substitution at other positions
on the piperidine ring (Table 5). Analogs with ring substitution favoring the chair conformation
observed in the 4-JAK3 crystal structure typically place the electrophile in close proximity to Cys-909 and
provide potency (15 and 19). The impact on GST-mediated clearance was more subtle as both
enantiomer configuration (14 vs. 12) and relative stereochemical configuration (11 vs 12) had a dramatic
impact on blood stability. This exercise enabled the identification of an additional analog (19) with both
excellent potency (IC = 56 nM) and blood stability via a surprisingly remote steric effect.
50
Table 4. In vitro and in vivo pharmacokinetic properties of 11 in various species
Species Mouse
Rat
Monkey
Dog
Human
CL Assay
Hepatocyte CLint,u (µl/min/million cells)
Blood t1/2 (min)
53
>360
48
14
161
53
7.6
37
2.1
>360
8.6
2.8
>360
-
Observed In Vivo CL (mL/min/kg)
42
b
a
Vss (L/kg)
0.8
1.3
62
1.4
0.3
85
2.6
0.7
56
1.1
1.3
b
In Vivo Plasma t1/2 (h)
PO Bioavailability (%)
1.1
1.8
109
-
ACS Paragon Plus Environment

Page 17 of 72
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Species Mouse
Rat
Monkey
Dog
Human
CL Assay
CLint,u = Unbound intrinsic clearance; CLb = Blood clearance
predicted from allometric scaling of unbound Vss
a
b
predicted human half-life
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Table 5. Evaluation of remote small steric effect on JAK3 potency and GST stability
Cmpd#
10
11
33
12
13
14
J3 1mM IC50 (nM)
Hblood stab t1/2
>1000
>10,000
>2478
>10,000
>
360
>331
185
>360
>360
(
min)
Cmpd#
15
16
17
18
19
56
20
J3 1mM IC50 (nM)
6
0
2021
>10,000
304
>10,000
>360
7570
Hblood stab t1/2
(min)
>250
103
>300
142
Compound 11 demonstrated a high degree of selectivity vs. a panel of 305 kinases (Invitrogen,
29
Figure 7). Approximately 200 kinases are known to contain a CYS residue within the ATP binding site.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 18 of 72
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Within this set, 11 kinases contain a CYS at the same position as that found in JAK3. When evaluated
against this subset, compound 11 demonstrated measurable activity against 8 of the targets. (Table 6).
IC50 data generated for a subset of these targets utilizing 1mM ATP levels showed generally good
selectivity vs. JAK3. The contribution of Ki and Kinact to activity was also determined for a subset of
these targets. It is interesting to note that the Kinact contribution is the dominant feature with regard
to JAK3 inhibition, illustrating how well-optimized the compound is for alignment of the electrophile,
with rapid kinase inactivation by Cys-909 addition. In contrast, the inactivation of ITK is very slow and
the majority of its affinity is associated with the Ki contribution. The very slow inactivation rate would
require significant residence time for capture in a cellular environment with high ATP levels suggesting
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
that 11 would be an ineffective inhibitor of ITK. In addition the utilization of strictly Kinact/K as a measure
i
of selectivity ignores the differential impact competition with ATP has on functional outcome as seen
when comparing BMX and TXK. We observed very weak affinity for three other kinases (SLK, FGR, and
FLT3) containing CYS in the ATP site, and none of those kinases were inhibited in a covalent manner.
ACS Paragon Plus Environment

Page 19 of 72
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
b)
Figure 7. a) Kinome selectivity of 11 determined at 1 ꢀM for 305 kinases determined at
Invitrogen. The percent inhibition is color coded and ranges from 0% inhibition (darker green) to 100 %
inhibition (darker red) with 50% inhibition displayed in yellow. The selectivity determined by Gini
coefficient is 0.88 b) Kinases tested that contained CYS in the ATP site.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 20 of 72
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Table 6. Inactivation rates of 11 towards kinases with CYS at the same location as JAK3. IC ’s were
50
determined at Carna Bioscience.
-1
-1 -1
kinact/K (M ·s )
IC50 (1mM ATP)
units = nM
K (µM)
kinase
JAK3
BMX
kinact (s )
i
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
i
2.32
6.31
0.545
0.0269
0.131
0.679
62.3
32.4
ND
3.68E+05
1.03E+04
5.35E+03
3.79E+03
2.30E+03
1.99E+03
8.58E+02
ND
90
606
0.00564
0.000144
0.000487
0.00156
0.124
ITK
8510
193
TXK
TEC
592
BTK
607
BLK
0.0278
ND
19000
25200
>50000
>50000
>50000
HER4
EGFR
HER2
MAP2K7
ND
ND
ND
ND
ND
ND
ND
ND
ND
Note: Given the experimental design [see Methods], exact solutions for individual kinetic parameters, K and kinact
,
i
were obtained only for BMX, ITK, and TXK; the solutions reported for all other kinases are lower limits. For all
kinases tested, the ratio (kinact/K ) is accurately determined.
i
The potent inhibition of in vitro cellular signaling demonstrated with 11 clearly illustrates that
selective inhibition of JAK3 can effectively modulate signaling via the common γ-chain cytokine receptor
which utilizes the heterodimeric JAK1/JAK3 pair. Furthermore, the high degree of efficacy reported for
11 and 7 in a variety of in vivo inflammatory disease models (AIA, DTH, CIA, and EAE) driven by gamma-
chain signaling unambiguously demonstrates the therapeutic potential for selective JAK3 inhibition in a
20
number of human diseases. Based on these results, 11 has been selected for clinical evaluation.
ACS Paragon Plus Environment

Page 21 of 72
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
CHEMISTRY. The synthesis of compounds 3, 4, 5, 6 and 9 are described in Scheme 1. Compound 3, the
acrylamide analog of tofactinib, was readily synthesized from the known amine, N-methyl-N-((3R,4R)-4-
methylpiperidin-3-yl)-7H-pyrrolo[2,3-d]pyrimidin-4-amine 21 (30) and acryloyl chloride in
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
dichloromethane/DIPEA in 32% yield. Compound 4 was synthesized beginning from the known 4-
3
1
chloro-7-tosyl-7H-pyrrolo[2,3-d]pyrimidine 22 followed by S displacement of the chlorine with tert-
nAr
32
butyl (R)-3-aminopiperidine-1-carboxylate 23 to give 24 in 42% yield. The tosyl group at N-1 was
subsequently removed under basic conditions (LiOH*H O, THF:MeOH:H O) to give 25 in 70% yield.
2
2
Removal of the Boc group of 25 with 4N HCl/dioxane to give 26 and installation of the acrylamide
acryloyl chloride, TEA/DCM/EtOH) provided 4 in 33% yield. Similarly, compound 9 was synthesized
(
using (E)-but-2-enoyl chloride in 32% yield. Compound 5 was prepared from 2,4-dichloro-7H-
pyrrolo[2,3-d]pyrimidine 27 followed by SnAr displacement of the 4-Cl with (R)-tert-butyl 3-
(
methylamino)piperidine-1-carboxylate to give compound 29 in 60% yield. Subsequent reduction of the
2-Cl gave 29, which was treated with acid to give 30. The secondary amine of 30 was then acylated with
acryloyl chloride to give compound 5. Finally, compound 6 was synthesized via Suzuki coupling (33) of 4-
chloro-7H-pyrrolo[2,3-d]pyrimidine 32 with (3-aminophenyl)boronic acid 33 under standard conditions
(Pd(PPh ) , Na CO , THF:H O) to provide intermediate 34 in 50% yield, followed by acylation of the
3
4
2
3
2
anilino nitrogen with acryloyl chloride (THF/TEA) to give 6 in 20% isolated yield.
Scheme 1. Synthesis of Acrylamides 3, 4, 5, 6 and 9
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 22 of 72
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
The synthesis of compounds 7 and 8 are described in Scheme 2. The racemic tert-butyl octahydro-1H-
pyrrolo[2,3-c]pyridine-1-carboxylate amine 35 was treated with Cbz-Cl to give 36 in 90% yield followed
ACS Paragon Plus Environment

Page 23 of 72
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
by removal of the Boc group to give intermediate 37 in 85% yield. The racemic (cis-fused) compound 37
was separated by chiral super critical fluid chromatography (SFC) to give the enantiomeric pair of (38a
and 38b). The absolute stereochemistry of 38b was determined by X-ray crystallography (see
Supporting Information). Nucleophilic aromatic substitution of the more activated chlorine at the 4-
position of 2,4-dichloro-7H-pyrrolo[2,3-d]pyrimidine 27 with compound 38a (R,R) provided 39 in 80%
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
yield. Reduction of the Cbz group and 2-chloro with 10% Pd/C-H provided compound 40 in 90% yield.
2
Final installation of the acrylamide was accomplished using aqueous basic conditions and acryloyl
chloride to give analog 7 in 81% yield. Compound 8 was prepared in a similar manner, except using
acetyl chloride.
Scheme 2. Synthesis of compound 7 and 8
For compound 10, the synthesis was accomplished as described in Scheme 3. 4-iodo-7-tosyl-7H-
pyrrolo[2,3-d]pyrimidine 41 and racemic benzyl (2R,3R)-3-amino-2-methylpiperidine-1-carboxylate 42
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 24 of 72
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
were reacted under SnAr conditions to give 43, which was then treated with base to effect tosyl group
removal to give 44 in 91% yield. The Cbz group was removed under transfer hydrogenation conditions
(Pd(OH) /cyclohexadiene) to give the free piperidine 45, which was subsequently acylated (acryloyl
2
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
chloride/DCM/DIPEA) to give 10 in 71% yield.
Scheme 3. Synthesis of compound 10
The synthesis of examples 11-14 are described in Scheme 4. Beginning from 6-methylpyridin-3-amine
4
6 the free amino group was protected as the Boc derivative to give pyridine 47, which was then
o
reduced using PtO /H /HOAc at 50 C to give piperidine 48 as a mixture of diastereomers (~2:1 cis/trans)
2
2
. The piperidine nitrogen of the diastereomeric mixture 48 was subsequently protected as the Cbz
carbamate to give 49 as a diastereomeric mixture. Compound 49 was separated into its diasteromeric
pair using chiral SFC to give compound 50a (rac-trans, pk1) and compound 50b (rac-cis, pk2). The Boc
groups of 50a and 50b were removed under standard conditions to give 51a (rac-trans) and 51b (rac-
cis). For compounds 11 and 12, the cis isomer 51b was reacted with 38 to give intermediate 52 (rac-cis),
which was then subjected to reduction conditions to give 53. Compound 53 was then acylated to
provide compound 54 as a rac-cis mixture. The racemic mixture was separated by chiral SFC to provide
ACS Paragon Plus Environment

Page 25 of 72
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
the enantiomeric pair 11 and 13. The preparation of compounds 12 and 14 was similarly accomplished,
except 51b (rac-trans) was utilized via intermediates 55-57 with chiral SFC of 57 (rac-trans) to give the
enantiomeric pair 12 and 14.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Scheme 4. Synthesis of compounds 11-14
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 26 of 72
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
For compounds 15-18 the headpiece amine was synthesized as described in Scheme 5, similar to the
piperidine amine 49 utilized for compounds 11-14 (Scheme 4), but utilizing 5-methylpyridin-3-amine as
the pyridine precursor. Unlike the 2-methyl amines, the 3-methyl amine derivatives 61 and 62 were
inseparable by chromatographic methods and therefore were carried on as a rac-cis/trans mixture 62,
with the potential for separation further along the synthesis. Indeed SnAR reaction of (62, rac-cis/trans)
with 38 provided a separable mixture, providing 63 and 64. Intermediate 63 (rac-cis) was then taken
ACS Paragon Plus Environment

Page 27 of 72
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
forward and the Cbz group and Cl removed to give 65 and the acrylamide installed to give 66 (rac- cis).
Chiral SFC of 66 provided the enantiomeric pair 15 and 17. Similarly, 64 (rac-trans) was taken forward
via intermediates 67 and 68 and final chiral SFC to provide the enantiomeric pair 18 and 16.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Scheme 5. Synthesis of compounds 15-18
Finally, the synthesis of compounds 19 and 20 is described in Scheme 6. The pyrrolopyrimidine 38 and
racemic tert-butyl (3R,4S)-3-amino-4-methylpiperidine-1-carboxylate 69 were reacted under SnAR
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 28 of 72
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
conditions to give compound 70 in 66% yield. The chlorine of 70 was reduced under catalytic
hydrogenation conditions to give 71 (93% yield), which was then subjected to acidic conditions (4N
HCl/dioxane) to effect removal of the Boc group and provide 72. Treatment of 72 with acryloyl chloride
under aqueous basic conditions gave acrylamide 73 in 34% yield. The racemic acrylamide 73 was
separated by chiral SFC to provide the enantiomeric pair, 19 and 20.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Scheme 6. Synthesis of Compounds 19 and 20
Conclusion
In summary we have disclosed the design of the first specific orally active JAK3 inhibitor in this
report. The ultimate selection of 11 illustrated how to improve Kinact without an increase in chemical
reactivity. Potency was coupled with minimizing GST mediated clearance by utilizing a remote steric
effect through appropriate substitution of a methyl group on the piperdine ring. High metabolic stability
resulted in an overall profile that made 11 suitable for oral dosing. The compound has exceptionally low
reactivity for other proteins containing cysteine as judged by kinome, HSA and human liver hepatocyte
profiling illustrating its specific design for inhibition of JAK3. This compound has been utilized to
ACS Paragon Plus Environment

Page 29 of 72
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
illustrate the consequences on JAK3 inhibition in a range of in vitro and in vivo systems once for all
settling the debate that it is sufficient to inhibit JAK3 only as the partner in the heterodimeric pair of
JAK1/JAK3 kinase mediated cytokine signaling though the γ-common chain receptors. 11 is currently in
clinical development for potential applications in inflammatory disease.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Experimental Section
7
The JAK enzyme data and the PBMC data was obtained as previously described.
Stability of JAK3 Covalent Inhibitors in Rat and Human Whole Blood
All animal experiments were performed following protocols reviewed and approved by Pfizer
Institutional Animal Care and Use Committee. Pfizer facilitites that supported this work are accredited
by AAALAC International. Rat blood was collected from 3 male Sprague-Dawley rats (200-250g, Charles
River Laboratories) and pooled for each study. Human blood was collected from one male and one
female healthy subject at the Occupational Health & Wellness Center at Pfizer, Groton, CT and pooled
for each study. Both rat and human blood was collected freshly into K2-EDTA tubes and kept on ice. An
aliquot of the blood was transferred to microtubes and pre-warmed for 10 min at 370C using a heat
block. The test compound was then added (1 µM final concentration) and the incubation was continued
for 180 min at 37 0C in duplicates. An aliquot of the incubation mixture was removed at designated
time points during the course of the incubation, mixed with an aliquot of acetonitrile containing an
internal standard, vortexed and centrifuged. The resulting supernatants were removed and subjected to
LC-MS/MS analyses to determine parent compound concentrations. Peak area ratios of the parent
compound vs the internal standard were used to determine the % of parent compound remaining vs
incubation time.
ACS Paragon Plus Environment