Design and Synthesis of a Novel Series of Orally Bioavailable, CNS-Penetrant, Isoform Selective Phosphoinositide 3-Kinase γ (PI3Kγ) Inhibitors with Potential for the Treatment of Multiple Sclerosis (MS)

Article abstract of DOI:10.1021/acs.jmedchem.8b00085

The lipid kinase phosphoinositide 3-kinase γ (PI3Kγ) has attracted attention as a potential target to treat a variety of autoimmune disorders, including multiple sclerosis, due to its role in immune modulation and microglial activation. By minimizing the number of hydrogen bond donors while targeting a previously uncovered selectivity pocket adjacent to the ATP binding site of PI3Kγ, we discovered a series of azaisoindolinones as selective, brain penetrant inhibitors of PI3Kγ. This ultimately led to the discovery of 16, an orally bioavailable compound that showed efficacy in murine experimental autoimmune encephalomyelitis (EAE), a preclinical model of multiple sclerosis.

Full text of DOI:10.1021/acs.jmedchem.8b00085

Subscriber access provided by University of Winnipeg Library
Article
Design and Synthesis of a Novel Series of Orally Bioavailable, CNS-
Penetrant, Isoform Selective Phosphoinositide 3-Kinase # Inhibitors
(PI3K#) with Potential for the Treatment of Multiple Sclerosis (MS)
Jon H Come, Philip N Collier, James A Henderson, Albert C Pierce, Robert J Davies, Arnaud Le Tiran,
Hardwin O'Dowd, Upul K. Bandarage, Jingrong Cao, David Deininger, Ron Grey, Elaine B. Krueger, Derek
Bates Lowe, Jianglin Liang, Yusheng Liao, David Messersmith, Suganthi Nanthakumar, Emmanuelle Sizensky,
Jian Wang, Jinwang Xu, Elaine Y Chin, Veronique Damagnez, John D. Doran, Wojciech Dworakowski, James
P Griffith, Marc D. Jacobs, Suvarna Khare-Pandit, Sudipta Mahajan, Cameron S. Moody, and Alex M. Aronov
J. Med. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.jmedchem.8b00085 • Publication Date (Web): 30 May 2018
Downloaded from http://pubs.acs.org on May 30, 2018
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 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 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.
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 42
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 42
1
2
3
4
5
6
7
8
9
Design and Synthesis of a Novel Series of Orally
Bioavailable, CNSꢀPenetrant, Isoform Selective
Phosphoinositide 3ꢀKinase γ Inhibitors (PI3Kγ) with
Potential for the Treatment of Multiple Sclerosis
(MS)
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
Jon H. Come,* Philip N. Collier,* James A. Henderson, Albert C. Pierce, Robert J. Davies,
Arnaud Le Tiran, Hardwin O’Dowd, Upul K. Bandarage, Jingrong Cao, David Deininger, Ron
Grey, Elaine B. Krueger, Derek B. Lowe, Jianglin Liang, Yusheng Liao, David Messersmith,
Suganthi Nanthakumar, Emmanuelle Sizensky, Jian Wang, Jinwang Xu, Elaine Y. Chin,
Veronique Damagnez, John D. Doran, Wojciech Dworakowski, James P. Griffith, Marc D.
Jacobs, Suvarna KhareꢀPandit, Sudipta Mahajan, Cameron S. Moody and Alex M. Aronov*
Vertex Pharmaceuticals Incorporated, 50 Northern Avenue, Boston, Massachusetts 02210,
United States
ABSTRACT
The lipid kinase phosphoinositide 3ꢀkinase γ has attracted attention as a potential target to treat
a variety of autoimmune disorders, including multiple sclerosis, due to its role in immune
ACS Paragon Plus Environment
1

Page 3 of 42
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
modulation and microglial activation. By minimizing the number of hydrogen bond donors while
targeting a previously uncovered selectivity pocket adjacent to the ATP binding site of PI3Kγ,
we discovered a series of azaisoindolinones as selective, central nervous system (CNS) penetrant
inhibitors of PI3Kγ. This ultimately led to the discovery of 16, an orally bioavailable compound
which showed efficacy in murine experimental autoimmune encephalomyelitis (EAE), a
preclinical model of multiple sclerosis.
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
INTRODUCTION
The phosphoinositide 3ꢀkinases (PI3Ks) are a family of lipid kinases that catalyze the
phosphorylation of phosphatidylinositols at the 3ꢀposition to generate phosphatidylinositolꢀ3,4,5ꢀ
triphosphate, which then interacts with effector proteins to impact the regulation of cell growth,
differentiation, proliferation, survival, migration, and intracellular trafficking.1ꢀ3 Based on their
structure and substrate specificity, the PI3Ks have been divided into three classes. Class I has
four members (PI3Kα, PI3Kβ, PI3Kγ, and PI3Kδ) which are heterodimeric complexes that have
highly homologous 110 kDa catalytic subunits and bind to a regulatory subunit. Class Ia PI3Ks
are activated through tyrosine kinase signaling, while the sole Class Ib member, PI3Kγ, is
activated through Gꢀproteinꢀcoupled receptor (GPCR) signaling.1,4,5
PI3Kα and PI3Kβ are ubiquitously expressed and are important in cell growth, proliferation and
survival; deletion of the catalytic subunit of these two isoforms is embryonically lethal in mice.6ꢀ
8 Inhibition of PI3Kα/β has mostly been limited to oncology applications. 9ꢀ11 In contrast, PI3Kγ
and PI3Kδ are mainly expressed in the hematopoietic system,12 and have been the target of
programs in inflammation, autoimmune disease and cardiovascular disease. 13ꢀ16 Dual inhibitors
of PI3Kγ and PI3Kδ are being investigated for cancer treatment in the clinic.17ꢀ19 Selective
ACS Paragon Plus Environment
2

Journal of Medicinal Chemistry
Page 4 of 42
1
2
3
4
5
6
7
8
9
inhibition of PI3Kγ may have useful immuneꢀmodulating activity, by inhibiting the action of
microglia, the brain’s resident immune defense against disease and injury. Microglia are
activated in response to immunologic stimuli, and are important in maintaining tissue
homeostasis and neuronal integrity in the CNS. In neurodegenerative diseases such as MS,
microglia may be exposed to nonꢀphysiological immune activation. The overꢀactivated microglia
contribute to the neurodegeneration by secreting proꢀinflammatory mediators such as cytokines,
chemokines and reactive oxygen species (ROS).20,21 The PI3Kγ pathway has been implicated in
ROS production and chemokine signaling pathways,22,23 and progress has been made towards
understanding the complex mechanisms involved.24 Li et al. have shown that systemic treatment
with selective PI3Kγ inhibitor ASꢀ60485023 reduced the number of infiltrated leukocytes in the
CNS and alleviated the clinical symptoms of EAE.25 The same group saw increased myelination
and axon number in the spinal chord of EAE mice treated with a PI3Kγ inhibitor.25 Hence PI3Kγ
blockade may be beneficial for the treatment of MS.
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
Reports of the discovery of selective inhibitors of PI3Kγ are noteworthy given the high sequence
homology of the ATP binding sites between PI3K isoforms.26ꢀ33 Designing CNSꢀpenetrant kinase
inhibitors is a challenge due to the typical property space characteristic of ATPꢀcompetitive
kinase inhibitors. Hydrogenꢀbond donors that interact with kinase hinge elements are a common
feature of many kinase pharmacophores.34 It is also a critical factor in determining the likelihood
of CNS penetration.35 In this report, we describe the design and optimization of a series of
isoindolinones as potent, selective, CNSꢀpenetrant inhibitors of PI3Kγ, culminating in the
discovery of 16, an orally bioavailable lead compound showing potential for the treatment of
MS.
ACS Paragon Plus Environment
3

Page 5 of 42
Journal of Medicinal Chemistry
1
2
3
4
RESULTS AND DISCUSSION
5
6
7
8
9
Previously, we have reported on the discovery of potent and selective benzothiazole ureaꢀbased
inhibitors of PI3Kγ (Figure 1).36,37 Selectivity in the benzothiazole series was achieved by
accessing a pocket specific to the PI3Kγ isoform. This pocket is made accessible by the
movement of Lys883, allowing placement of the distal urea substituent in the newly formed
binding cleft. The unique geometry of the urea was key to positioning this substituent. However,
the presence of two hydrogenꢀbond donors in both the benzothiazole36 and thiazolopiperidine37
ureaꢀbased scaffold classes resulted in low CNS exposure (efflux ratio = 18 for 1 in MDR1ꢀ
overexpressing MDCK cell line).
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 1. Structures of previously reported potent and selective PI3Kγ inhibitors.
In pursuit of our goal of discovering CNSꢀpenetrant, isoform selective inhibitors of PI3Kγ, the
ureaꢀbased scaffolds were evolved in accordance to the following criteria: (1) maintain access to
the selectivity pocket; (2) keep no more than a single hydrogen bond acceptor to the hinge
region; (3) remove all formal hydrogen bond donors. Of all the designs that were explored, the
isoindolinone scaffold design proved the most compelling. We hypothesized that the
isoindolinone carbonyl group would provide a hydrogen bond acceptor to the backbone NH of
hinge residue Val882, while the N2ꢀsubstituent would allow access to the previously described
selectivity pocket (Figure 2). An aromatic proton would be positioned in place of the urea NH
ACS Paragon Plus Environment
4

Journal of Medicinal Chemistry
Page 6 of 42
1
2
3
4
5
6
7
8
9
donor. While it was uncertain how significant the CHꢀꢀO pseudo hydrogen bond would be, we
felt confident based on literature precedent, that it would be tolerated, and that its introduction
would lead to compounds with lower affinity for PgPꢀmediated efflux pumps.38,39 Finally, the C5
substituent on the isoindolinone scaffold would provide potency, guided by SAR knowledge
gained from our earlier work.
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
Val882
Val882
N
N
H
H
O
H
O
selectivity
pocket
O
H
N
S
selectivity
pocket
N
O
H
R
C5
N
N
Ar
Ar
No formal H-bond donors
Figure 2. Rational design of the isoindolinone scaffold shows key design features: (a) contacts
between inhibitor and hinge region Val882; (b) vector toward selectivity pocket; (c) C5 vector to
aromatic substituent.
We designed and synthesized an initial library of 80 compounds with variation at the
isoindolinone C5 position and at the metaꢀposition of the phenyl ring attached to N2. We biased
the C5 set towards substituents that were shown to possess higher affinity in the benzothiazole
series,36 and focused work on metaꢀsubstituted phenyls at N2, as we felt this position afforded
the best vector to reach the selectivity pocket.
ACS Paragon Plus Environment
5

Page 7 of 42
Journal of Medicinal Chemistry
1
2
3
4
O
R
5
6
N
Ar
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
N
N
N
N
N
F3C
F
F
N
R =
O
O
Ar =
OMe
F
OMe
N
N
N
OMe
CN
SO2Me
HN
SO2Me
O
N
O
H
O
Figure 3. Library design for initial exploration of the isoindolinone scaffold
From the initial set, two hits were identified, including 3 (Table 1), which inhibited PI3Kγ with a
Ki of 160 nM and was greater than 20ꢀfold selective against PI3Kα and PI3Kβ, and 7ꢀfold
selective against PI3Kδ. Encouraged by these results, we sought to improve the potency and
metabolic stability of the series, while capitalizing on prior SAR knowledge at C5.36, 37
Consistent with previous results from the benzothiazole series, conversion of the C5 substituent
from 3,4ꢀdimethoxyphenyl to 3,4ꢀdimethoxyꢀ5ꢀpyridinyl increased PI3Kγ affinity fourfold
(Table 1, 4) while maintaining isoform selectivity. PK studies revealed that compound 4 had
high plasma clearance in rats (see Table 1) so we attempted to improve the metabolic stability by
reducing compound lipohilicity. Replacing the N2 (cyanomethyl)phenyl moiety with an Nꢀ
ACS Paragon Plus Environment
6

Journal of Medicinal Chemistry
Page 8 of 42
1
2
3
4
5
6
7
8
9
(cyanomethyl)pyrazole substituent (Table 1, 5) led to higher PI3Kγ affinity and reduced
clearance in rats, while selectivity against other Class 1 PI3Ks was largely maintained. Finally,
introduction of a nitrogen atom in place of C4 in the isoindolinone core further increased PI3Kγ
affinity and lowered in vitro and in vivo clearance (Table 1, 6). The reduced intrinsic clearance
and Rat IV clearance observed on transitioning from 3 to 6 correlated with a corresponding drop
in lipophilicity (logD @ pH 7.4: 4.1 for 3, 1.7 for 6).
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 1. Initial Optimization of Isoindolinone Library hit 3.
Structure
Cpd PI3Kγa
Ki
Fold Selectivity over:
Hepatocyte
int. Clb
Rat IV
Clc
PI3Kα
PI3Kβ
PI3Kδ
Rat/Human
0.16
0.035
0.006
0.004
>25
40
>25
7
111/26
116
160
56
3
4
5
6
11
27
25
16
17
12
88/30
ND
51
40
7.9/10
27
b
c
aAll enzyme data are expressed in ꢁM. CL expressed in ꢁL/min/million cells. CL expressed in
mL/min/Kg.
ACS Paragon Plus Environment
7

Page 9 of 42
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
The Xꢀray crystal structure of 6 in complex with PI3Kγ confirmed that it bound in a manner
envisioned in our design, with the carbonyl of the azaisoindolinone acting as a hinge binding
motif, while the nitrile substituent pointed towards the selectivity pocket (Figure 4). The proton
at the 5ꢀposition of the pyrazole is directed at the carbonyl of Val882, forming the aromatic
CHꢀꢀO hydrogen bond as designed and mimicking the urea NH donor interaction observed in
the benzothiazole series (Figure 5). PI3Kγ affinity is enhanced by the dimethoxypyridine group
which makes (a) key waterꢀmediated hydrogen bonds between the pyridine nitrogen atom and
Tyr867 and Asp841, residues which are conserved across PI3K isoforms, and (b) favorable
protein contacts with Lys833, similar to those seen in the benzothiazole series.36 An overlay of 6
with one of our more selective benzothiazole ureas is shown in Figure 5. 6 does not occupy the
selectivity pocket in PI3Kγ to the same extent seen in the case of ureas such as 1 and does not
induce a movement of the Lys883 sidechain as seen in the benzothiazole urea scaffold,
suggesting that azaisoindolinones such as 6, derive PI3Kγ selectivity in a somewhat different
manner.
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 4. Xꢀray structure of azaisoindolinone 6 bound to PI3Kγ (PDB code 6C1S). The ligand
makes a bidentate hinge binding interaction with Val882. The dimethoxypyridine group
ACS Paragon Plus Environment
8

Journal of Medicinal Chemistry
Page 10 of 42
1
2
3
4
5
6
7
8
9
enhances affinity by making three favorable interactions to the ATP binding site (Lys833, and
water mediated interactions with Tyr867 and Asp841). Ala885, which drove a degree of isoform
selectivity in the benzothiazole series, is highlighted.
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 5. Overlay of azaisoindolinone 6 and benzothiazole urea 1 as they bind to PI3Kγ. The
lactam carbonyl oxygen atom fulfills the role of the nitrogen atom in the benzothiazole by
making a hydrogen bond to the backbone NH of Val882. The pyrazole aromatic CꢀH donor
mimics the role of the urea Hꢀbond donor of the benzothiazole scaffold.
Compound 6 had an acceptable rat PK profile with a clearance of 27 mL/min/kg. However,
upon further investigation, it was established that dealkylation of the pyrazole to produce
compound 7 was a major route of metabolism, potentially generating cyanide via oxidation of
the methylene of the cyanomethyl substituent (see Supporting Information for details).
Consequently, we surveyed a number of alternative substituents to replace the pyrazole
cyanomethyl group, and the results of this study are shown in Table 2.
Desꢀalkyl compound 7 displayed good selectivity against class Ia PI3Ks but was a substrate for
Pꢀglycoprotein efflux protein (efflux ratio = 9 in MDR1ꢀMDCK), disfavoring entry into the
CNS. Isoform selectivity was generally improved on transitioning from methyl to larger alkyl
ACS Paragon Plus Environment
9

Page 11 of 42
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
groups. This may be due to PI3K isoform residue differences at Ala885 as described in our
previous work.36 In the other PI3K class Ia isoforms, this residue is a serine which, we
hypothesize, may negatively interact with larger lipophilic pyrazole Nꢀalkyl groups. The lower
PI3Kγ affinity observed for 13 is potentially due to steric disruption of the hinge binding
pyrazole CHꢀꢀO hydrogen bond. By replacing the cyanomethyl with a trifluoroethyl group
(compound 10), it was possible to maintain potency and selectivity while eliminating the
potential for cyanide generation. Pꢀgp mediated efflux was reduced in pyrazole Nꢀalkyl
compounds 8 – 13, generally correlating with the lower PSA of these compounds relative to 6.
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 2. Pyrazole Nꢀalkyl group SAR
R
Cpd PI3Kγ Ki
Fold Selectivity
PSA
MDCKꢀ
MDR1
ꢁM
Å2
PI3Kα PI3Kβ PI3Kδ
ER
4.7
8.7
2.5
2.2
CH2CN
H
0.003
0.011
0.06
50
34
5
31
32
9
16
42
3
106
93
6
7
8
Me
82
82
Et
0.008
25
39
14
9
82
82
0.8
2.3
2.6
2.1
ND
CH2CF3
CH2CCH
0.004
0.009
0.009
0.028
0.02
65
11
22
15
11
31
25
12
25
10
13
12
5
10
11
12
13
14
CH2CHF2
C(Me)2CN
CH2CH2CN
82
10
4
106
106
ACS Paragon Plus Environment
10

Journal of Medicinal Chemistry
Page 12 of 42
1
2
3
4
5
6
7
8
9
We also evaluated substitutions at C3, which allowed us to explore interesting vectors out of
the plane of the rest of the molecule. While addition of a (R)ꢀmethyl group at C3 (Table 3, 16)
maintained a biochemical profile similar to before, we observed an improvement in rat IV
clearance from 56 to 20 mL/min/kg and half life from 1.1 to 3.3 h. Substitution of C3 with a
gemꢀdimethyl group lowered clearance still further to 6.3 mL/min/kg, but activity and selectivity
were somewhat eroded.
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
Advanced lead compounds 10 and 15ꢀ17, were tested in various assays to determine their
potency in inhibiting PI3Kγꢀdependent signaling and function. Chemokines such as MCPꢀ1 bind
to their receptors, resulting in activation of the PI3Kγ signaling pathway. MCPꢀ1 activation leads
to PI3Kγꢀmediated Akt phosphorylation, and PꢀAkt following MCPꢀ1 introduction was used as a
marker of PI3Kγ activity. Compounds 10, 15ꢀ17 were found to inhibit MCPꢀ1 induced Akt
phosphorylation in THPꢀ1 cells. Furthermore, compound 16 was found to inhibit MCPꢀ1 induced
Akt phosphorylation in splenocytes with an IC50 of 0.21 ꢁM (+/ꢀ0.15; n=3). PI3Kγ functional
activity was measured in TNFꢀα primed human neutrophils and monocytes via the generation of
ROS. Compound 16 showed reduction in ROS production after stimulation with NꢀformylꢀMetꢀ
LeuꢀPhe (fMLP), with an IC50 of 0.14 ꢁM (0.50 ꢁM in whole blood), giving us greater
confidence in our mechanistic hypothesis, and the potential of our selective PI3Kγ inhibitors in
the treatment of multiple sclerosis.
No cross activity was shown for 16 against a diverse panel of serine/threonine and tyrosine
kinases and against a wider panel of kinases (50/50 < 15% inhibition at 10 ꢁM, see Supporting
Information). Additionally, 16 showed excellent selectivity against a diverse set of 75 nonꢀkinase
targets (all < 25% inhibition at 10 ꢁM). The only substantial cross activity of 16 was against the
PI3Kꢀrelated serine/threonine kinase DNAꢀPK (Ki = 0.26 ꢁM).
ACS Paragon Plus Environment
11

Page 13 of 42
Journal of Medicinal Chemistry
1
2
3
4
Table 3. Azaisoindolinone C3 SAR
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
R, R’
Cpd PI3Kγ Ki
Fold Selectivity
THPꢀ1a,b Rat Hep
Rat IV
PKd
int. Clc
ꢁM
(MCPꢀ1)
PI3Kα
PI3Kβ
PI3Kδ
Cl, T1/2
IC50
H, H
0.004
0.007
0.004
0.011
65
28
60
33
31
18
10
12
13
7
0.28
0.21
0.17
0.34
ND
9.2
7.0
ND
56, 1.1
15, 4.7
20, 3.3
6.3, 6.6
10
15
16
17
(S)ꢀMe, H
(R)ꢀMe, H
Me, Me
14
10
aCell data are expressed in ꢁM; bthe procedure involves monitoring the phosphorylation state of
Serꢀ473; cCl expressed in ꢁL/min/million cells; dCl expressed in mL/min/Kg and T1/2 in h
Additional profiling of compound 16 is shown in Tables 4 and 5. The compound showed good
cross species IV PK profiles, presenting low to moderate clearance values and adequate half
lives. Based on inꢀvitro data and allometric scaling from preclinical species, the predicted human
whole blood clearance, volume of distribution and halfꢀlife were 3.3 mL/min/Kg, 4.2 L/Kg and
15 h, respectively. The predicted human oral bioavailability was 66%. Compound 16 was highly
permeable and showed no efflux in Cacoꢀ2 or MDR1ꢀMDCK assays, thereby indicating potential
for good penetration at the bloodꢀbrain barrier. When taken together, these data suggested that
exposure of 16 in the CNS would be sufficient to modulate PI3Kγ in an in vivo model.
Compound 16 was also assessed for its ability to cause drugꢀdrug interactions, and showed no
inhibition against a panel of recombinant human CYP isoforms, and no potential for timeꢀ
dependent inhibition of CYPs. Finally, compound 16 showed no cardiovascular liabilities (hERG
IC50 of > 30 ꢁM) or genotoxicity (negative in AMES assay).
ACS Paragon Plus Environment
12

Journal of Medicinal Chemistry
Page 14 of 42
1
2
3
4
Table 4. Cross species comparison of PK parameters and hepatocyte intrinsic clearance of 16
5
6
7
8
9
b
Cla
T1/2
(h)
2.8
3.2
2.4
ꢀ
Vss
%F
Hepatocyte
Intrinsic Clc
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
Rat
Dog
8.8
19
4.0
ꢀ
32
5.2
0.8
ꢀ
70
28
39
ꢀ
5
12
7
Monkey
Human
4
aUnits of mL/min/Kg; bunits of L/Kg; cunits of ꢁL/min/M cells
Table 5. Further profiling of compound 16
MW, logD (pH 7.4), PSA
Rat PO PK @ 3 mg/Kg (%F, T1/2, AUC, Cmax)
MDR1ꢀMDCK BꢀA/AꢀB; AꢀB
Cacoꢀ2 BꢀA/AꢀB; AꢀB
433, 3.0, 82
100, 5.1 h, 7.2,a 0.7b
0.7, 50c
0.9, 24c
CYP inhibition 3A4, 2C9, 2D6
P450, TDI
all >30 ꢁM
No issues
5ꢀStrain Ames
No issues
hERG (manual patch) IC50
>30 ꢁM
aUnits of ꢁg*h/mL; bUnits of ꢁg/mL; cUnits of x10ꢀ6 cm/s.
On the basis of its overall favorable profile, and given that it had met our design criteria,
compound 16 was selected as the preferred compound for in vivo pharmacological evaluation. 16
was evaluated in murine EAE, a wellꢀestablished model of multiple sclerosis. It has previously
been shown that treatment with PI3Kγ inhibitor ASꢀ604850, ameliorated the clinical symptoms
ACS Paragon Plus Environment
13

Page 15 of 42
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
of EAE mice.25 PI3Kγ deletion in knockout mice has also mitigated the clinical symptoms of
EAE compared to PI3Kγ+/+ controls.25 In our study, C57BL6 mice were immunized by injection
of myelin oligodendrocyte glycoprotein (MOG) in complete Freund’s adjuvant containing
Mycobacterium tuberculosis (see Supporting Information for full details). Following injection of
pertussis toxin, animals were evaluated for disease symptoms. On reaching a clinical score of
0.5ꢀ1, animals were dosed with vehicle, 16 or positive control fingolimod (FTYꢀ720). At doses
of 3, 10 and 30 mg/Kg BID of 16, we observed significant reduction in clinical disability scores
as shown in Figure 6, with efficacy comparable to FTYꢀ720. Based on the aforementioned Cacoꢀ
2/MDR1ꢀMDCK data, we expected compound 16 to display good brain penetration.
Gratifyingly, we found unbound brain exposure at 4 h post dose was comparable (3 mg/kg) or
exceeded (10 and 30 mg/kg) IC50 (ROS functional assay), indicating ample target coverage at the
selected doses (Table 6). These results give us greater confidence in PI3Kγ inhibition as an
approach for the potential treatment of MS.
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
14

Journal of Medicinal Chemistry
Page 16 of 42
1
2
3
4
5
6
7
8
Figure 6. Treatment with 16 attenuates clinical symptoms of EAE mice. Animals were dosed
orally BID with 16, or QD with FTYꢀ720 on reaching a clinical score of 0.5ꢀ1. Plot indicates
clinical EAE scores as a function of time after EAE onset.
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
Table 6. Unbound brain concentrations of 16 from EAE studya
Concentration (nM)
Time (h)
3 mg/kg
398 (±89)
87 (±45)
10 mg/Kg
1017 (±230)
294 (±128)
30 mg/Kg
1594 (±303)
764 (±96)
4
14
aBrain samples collected 4 and 14 h after the last dose in the EAE study (see Supporting
Information). Unbound concentrations calculated with brain homogenate; fraction unbound
values (fu 16%) obtained from in vitro equilibrium dialysis experiments.
Chemistry. The synthesis of 5ꢀarylisoindolinones 3 ꢀ 5 are depicted in Schemes 1 and 2.
The isoindolinone ring was formed by an alkylation/ring closure oneꢀpot sequence. Subsequent
SuzukiꢀMiyaura coupling provided 3 and 4. The 4ꢀazaisoindolinone core of compounds 6, 7, 10ꢀ
14 was prepared in an overall similar manner to above, albeit using radical halogenation to
provide intermediates 24 and 25 (Schemes 3 and 4). Orthoꢀchlorination of intermediate 25 was
achieved by treatment of its pyridine Nꢀoxide 26 with phosphorus oxychloride.
The pyrazole Nꢀalkyl substituent at position 2 of the azaisoindolinine ring was varied in a
late stage alkylation step on pyrazole 7, yielding compounds 10ꢀ12 and 14 (Scheme 4). The
cyano(dimethyl)methyl group of compound 13 was introduced by alkylation of 4ꢀnitroꢀ1Hꢀ
pyrazole with 2ꢀbromoꢀ2ꢀmethylꢀpropanamide (Scheme 5). Compounds 8 and 9 were
synthesized from fully elaborated 1ꢀalkylꢀ4ꢀaminoprazoles using the oneꢀpot alkylation/ring
ACS Paragon Plus Environment
15

Page 17 of 42
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
closure as the key step (Scheme 6). Substituents were introduced at the azaisoindolinone C3
position by a deprotonation/alkylation sequence (Scheme 7) and preparatory chiral HPLC was
used to separate enantiomers 15 and 16.
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
Scheme 1. Synthesis of isoindolinones 3 and 4a
aReagents and conditions: (a) 2ꢀ(3ꢀaminophenyl)acetonitrile, DMF, 110 ºC; (b) (3,4ꢀ
dimethoxyphenyl)boronic acid (for 3) or 2,3ꢀdimethoxyꢀ5ꢀ(4,4,5,5ꢀtetramethylꢀ1,3,2ꢀ
dioxaborolanꢀ2ꢀyl)pyridine (for 4), Pd(dppf)Cl2, Cs2CO3, DMSO, H2O, 100 ºC.
Scheme 2. Synthesis of isoindolinone 5a
aReagents and conditions: (a) iodoacetonitrile, Cs2CO3, DMF; (b) Fe, NH4Cl, EtOH, 75 ºC; (c)
methyl 4ꢀbromoꢀ2ꢀ(bromomethyl)benzoate, DMF, 100 ºC; (d) 2,3ꢀdimethoxyꢀ5ꢀ(4,4,5,5ꢀ
tetramethylꢀ1,3,2ꢀdioxaborolanꢀ2ꢀyl)pyridine, Pd(dppf)Cl2, Cs2CO3, DMSO, H2O, 100 ºC.
ACS Paragon Plus Environment
16

Journal of Medicinal Chemistry
Page 18 of 42
1
2
3
4
Scheme 3. Synthesis of isoindolinone 6a
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
aReagents and conditions: (a) POCl3, 100 ºC; (b) 2,3ꢀdimethoxyꢀ5ꢀ(4,4,5,5ꢀtetramethylꢀ1,3,2ꢀ
dioxaborolanꢀ2ꢀyl)pyridine, Pd(PPh3)4, Na2CO3, MeCN/H2O (3/1), 90 ºC; (c) NBS, AIBN,
K2CO3, CCl4, 80 ºC; (d) 2ꢀ(4ꢀaminoꢀ1Hꢀpyrazolꢀ1ꢀyl)acetonitrile, Na2CO3, DMF, 90 ºC.
Scheme 4. Synthesis of azaisoindolinones 7, 10ꢀ12, 14a
aReagents and conditions: (a) 1,3,5ꢀtrichloroꢀ1,3,5ꢀtriazinaneꢀ2,4,6ꢀtrione, DCM; (b) mꢀCPBA,
DCM; (c) POCl3; (d) 1Hꢀpyrazolꢀ4ꢀamine, DIPEA, DMF, 80 ºC; (e) 2,3ꢀdimethoxyꢀ5ꢀ(4,4,5,5ꢀ
tetramethylꢀ1,3,2ꢀdioxaborolanꢀ2ꢀyl)pyridine, Pd(PPh3)4, Na2CO3, DMF, 90 ºC; (f) 2,2,2ꢀ
trifluoroethyl trifluoromethanesulfonate (for 10), 3ꢀchloropropꢀ1ꢀyne (for 11), 2ꢀbromoꢀ1,1ꢀ
difluoroꢀethane (for 12), 3ꢀbromopropanenitrile (for 14), Cs2CO3, DMF, microwave or
conventional heating.
ACS Paragon Plus Environment
17

Page 19 of 42
Journal of Medicinal Chemistry
1
2
3
4
Scheme 5. Synthesis of azaisoindolinone 13a
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
aReagents and conditions: (a) 2ꢀbromoꢀ2ꢀmethylꢀpropanamide, K2CO3, DMF; (b) H2 (40 psi),
Pd/C (10%), MeOH; (c) ethyl 6ꢀchloroꢀ2ꢀ(chloromethyl)pyridineꢀ3ꢀcarboxylate, DIPEA, IPA, 80
ºC; (d) 2,3ꢀdimethoxyꢀ5ꢀ(4,4,5,5ꢀtetramethylꢀ1,3,2ꢀdioxaborolanꢀ2ꢀyl)pyridine, Pd(PPh3)4,
Na2CO3, DMF, 80 ºC; (e) Burgess reagent, DCM/MeCN (1:1).
Scheme 6. Synthesis of alkylpyrazoles 8 and 9a
aReagents and conditions: (a) DIPEA, iPrOH (for 35) or MeCN (for 36), reflux; (b) 2,3ꢀ
dimethoxyꢀ5ꢀ(4,4,5,5ꢀtetramethylꢀ1,3,2ꢀdioxaborolanꢀ2ꢀyl)pyridine, Pd(PPh3)4, K2CO3, dioxane,
microwave, 150 ºC.
ACS Paragon Plus Environment
18

Journal of Medicinal Chemistry
Page 20 of 42
1
2
3
4
Scheme 7. Synthesis of 7ꢀsubstitutedꢀazaisoindolinones 15 ꢀ 17a
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
aReagents and conditions: (a) NaH, MeI, THF; (b) chiral HPLC separation; (c) NaH, MeI, DMF.
CONCLUSION
We have designed and synthesized a series of high affinity, isoform selective, CNSꢀpenetrant
and orally bioavailable inhibitors of PI3Kγ. Guided by structural data from our earlier reported
work in this area, our scaffold design strategy was biased towards hingeꢀbinding motifs lacking
formal Hꢀbond donor atoms. This yielded compounds without the efflux liabilities associated
with our earlier scaffolds. Further optimization culminated in the discovery of 16, which
demonstrated efficacy in murine EAE, providing support for the further evaluation of this
mechanism for the treatment of multiple sclerosis.
ACS Paragon Plus Environment
19

Page 21 of 42
Journal of Medicinal Chemistry
1
2
3
4
EXPERIMENTAL SECTION
5
6
7
8
Chemistry. All commercially available reagents and anhydrous solvents were used without
further purification. An inert atmosphere of nitrogen was used for reactions involving air of
moisture sensitive reagents. Purity assessment for final compounds based on analytical HPLC:
column, 4.6 x 50 mm Waters YMC ProꢀC18 column, 5 ꢁM, 120Å. Mobile phases are as follows:
A, water with 0.2% formic acid; B, acetonitrile with 0.2% formic acid; gradient, 10ꢀ90% B in 3
min with 5 min run time. The flow rate is 1.5 mL/min. All final compounds were ≥ 95% purity.
Mass samples were analyzed on a Waters/MicroMass ZQ, ZMD, Quattro LC, or Quatro II mass
spectrometer operated in a single MS mode with electrospray ionization. Samples were
introduced into the mass spectrometer using flow injection (FIA) or chromatography. The mobile
phase for all mass analysis consisted of acetonitrileꢀwater mixtures with either 0.2% formic acid
or ammonium formate. Highꢀresolution mass measurements were performed on a Thermo
QExactive mass spectrometer with a heated electrospray source operated in positive ion mode.
1H NMR spectra were recorded either using a Bruker Avance 400 (400 MHz) or a Bruker
Avance IIꢀ300 (300 MHz) instrument at ambient temperature and are quoted in ppm relative to a
tetramethylsilane internal standard, or by referencing on the chemical shift of the deuterated
solvent. Data are displayed in following format: chemical shift, multiplicity (s = singlet, d =
doublet, t = triplet, q = quartet, br = broad, m = multiplet), coupling constants, and number of
protons. The column chromatography was performed using Teledyne ISCO RediSep Normal
Phase (35ꢀ70 microns) or RediSep Gold Normal Phase (25ꢀ40 microns) silica flash columns
using a Teledyne ISCO Combiflash Companion or Combiflash Rf purification system.
Preparative reversed phase chromatography was carried out using a Gilson 215 liquid handler
coupled to a UV_vis 156 Gilson detector, an Agilent Zorbax SBꢀC18 column, 21.2 x 100 mm, a
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
20

Journal of Medicinal Chemistry
Page 22 of 42
1
2
3
4
5
6
7
8
9
linear gradient from 10ꢀ90% acetonitrile in water over 10 min (0.1% TFA); the flow rate was 20
mL/min. Microwaveꢀassisted reactions were performed using a CEM Discover SꢀClass
microwave instrument (single mode microwave reactor) with 48ꢀposition autosampler.
Temperature was monitored during microwave reactions by a vertically sensored infrared
temperature sensor, which comes as a standard feature of the CEM Discover SꢀClass system.
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
2ꢀ(3ꢀ(5ꢀ(3,4ꢀDimethoxyphenyl)ꢀ1ꢀoxoisoindolinꢀ2ꢀyl)phenyl)acetonitrile (3). Step 1: A 10
mL microwave vial was charged with 2ꢀ(3ꢀaminophenyl)acetonitrile (500 mg, 3.78 mmol),
methyl 4ꢀbromoꢀ2ꢀ(bromomethyl)benzoate (1.2 g, 3.8 mmol) and DMF (3 mL). The vial was
sealed and microwaved at 110 °C for 25 min. The reaction mixture was diluted with MeOH (50
mL) and an offꢀwhite solid filtered off and dried under vacuum to give 18 (703 mg, 57%) as a
white solid. 1H NMR (300 MHz, DMSOꢀd6) δ 7.96 (d, J = 7.3 Hz, 2H), 7.83 ꢀ 7.71 (m, 3H), 7.47
(t, J = 8.0 Hz, 1H), 7.18 (d, J = 7.6 Hz, 1H), 5.03 (s, 2H), 4.10 (s, 2H). Mass spectrum (ESI) m/z
328.3 [M + H]+.
Step 2: A microwave vial was charged with 18 (100 mg, 0.31 mmol), (3,4ꢀ
dimethoxyphenyl)boronic acid (60 mg, 0.34 mmol), Cs2CO3 (200 mg, 0.61 mmol), DMSO (2
mL) and H2O (0.3 mL). The solution was sparged with nitrogen for 10 min and Pd(dppf)Cl2 (12
mg, 0.015 mmol) was added. The vial was sealed and microwaved at 100 °C for 10 min. The
reaction mixture was poured into EtOAc/H2O and the organic layer passed through a plug of
florisil and concentrated to a residue which was triturated with MeOH, filtered and dried under
vacuum to give 3 (80 mg, 68%) as a white solid. 1H NMR (300 MHz, DMSOꢀd6) δ 8.03 (s, 1H),
7.95 (s, 1H), 7.86 ꢀ 7.81 (m, 3H), 7.48 (t, J = 8.0 Hz, 1H), 7.34 ꢀ 7.31 (m, 2H), 7.17 (d, J = 7.6
Hz, 1H), 7.10 (d, J = 9.0 Hz, 1H), 5.07 (s, 2H), 4.11 (s, 2H), 3.88 (s, 3H), 3.82 (s, 3H). HRMS
ACS Paragon Plus Environment
21

Page 23 of 42
Journal of Medicinal Chemistry
1
2
3
4
found for M+ + 1, 385.1547, C24H20N2O3 requires 385.1546.
5
6
7
8
9
2ꢀ(3ꢀ(5ꢀ(5,6ꢀDimethoxypyridinꢀ3ꢀyl)ꢀ1ꢀoxoisoindolinꢀ2ꢀyl)phenyl)acetonitrile (4). A 10 mL
microwave vial was charged with 18 (50 mg, 0.15 mmol), 2,3ꢀdimethoxyꢀ5ꢀ(4,4,5,5ꢀtetramethylꢀ
1,3,2ꢀdioxaborolanꢀ2ꢀyl)pyridine (44 mg, 0.17 mmol), Cs2CO3 (100 mg, 0.31 mmol), DMSO (1
mL) and H2O (150 µL). The solution was sparged with nitrogen for 10 min and Pd (dppf)Cl2 (6
mg, 0.008 mmol) was added. The vial was sealed and microwaved at 100 °C for 10 min. The
reaction mixture was poured into EtOAc/H2O and the organic layer was passed through a plug of
florisil and concentrated to a residue which was triturated with MeOH, filtered and dried under
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
vacuum to give 4 (59 mg, 69%) as a white solid. H NMR (300 MHz, DMSOꢀd6) δ 8.12 (d, J =
1.9 Hz, 1H), 8.02 (d, J = 8.2 Hz, 2H), 7.88 (m, 3H), 7.65 (d, J = 1.9 Hz, 1H), 7.49 (t, J = 8.0 Hz,
1H), 7.18 (d, J = 7.7 Hz, 1H), 5.09 (s, 2H), 4.11 (s, 2H), 3.94 (s, 3H), 3.92 (s, 3H). HRMS found
for M+ + 1, 386.1501, C23H29N3O3 requires 386.1499.
2ꢀ(4ꢀ(5ꢀ(5,6ꢀDimethoxypyridinꢀ3ꢀyl)ꢀ1ꢀoxoisoindolinꢀ2ꢀyl)ꢀ1Hꢀpyrazolꢀ1ꢀyl)acetonitrile
(5). Step 1: To a flask containing 4ꢀnitroꢀ1Hꢀpyrazole (1.0 g, 8.84 mmol) and Cs2CO3 (2.88 g,
8.84 mmol) in DMF (30 mL) was added iodoacetonitrile (1.48 g, 8.84 mmol). After stirring at
room temperature for 30 min, the reaction mixture was poured into H2O/EtOAc and the organic
layer was dried and concentrated to give 19 (1.1 g, 86%) as an oil. 1H NMR (300 MHz, CDCl3) δ
8.38 (s, 1H), 8.17 (s, 1H), 5.16 (s, 2H).
Step 2: A 5L 3ꢀneckꢀflask equipped with a mechanical stirrer was charged with 19 (50 g, 0.32
mol), ethanol (1 L), NH4Cl (106 g, 1.97 mol) and water (250 mL). Iron (73.4 g, 1.3 mol) was
added to this solution and the reaction was sparged with nitrogen and heated to 75 °C internal
temp (took 30 min to reach this temp). After 3 h at 75 °C, the reaction mixture was filtered
ACS Paragon Plus Environment
22

Journal of Medicinal Chemistry
Page 24 of 42
1
2
3
4
5
6
7
8
through Celite, washed with ethanol (3 x 500 mL) and the filtrate was evaporated to dryness on
the rotovap. DCM (1 L) was added followed by sodium sulfate and the resulting suspension was
filtered. The collected solids were washed with DCM (3 x 400 mL). The filtrate was
9
1
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
concentrated to dryness to give 20 as a tan solid (22.8 g, 57%). H NMR (300 MHz, CDCl3) δ
7.24 (s, 1H), 7.13 (s, 1H), 4.97 (s, 2H), 3.02 (br s, 2H). Mass spectrum (ESI) m/z 122.9 [M + H]+.
Step 3: In a 10 mL microwave tube, 20 (200 mg, 1.6 mmol) and methyl 4ꢀbromoꢀ2ꢀ
(bromomethyl)benzoate (250 mg, 0.81 mmol) were combined in DMF (2 mL) and heated to 110
°C for 20 min. The reaction mixture was poured into EtOAc/H2O and the organic layer was
separated, dried and concentrated to provide an oil which was triturated with MeOH. The
1
precipitated solid was filtered off to give 21 (65 mg, 13%). H NMR (300 MHz, DMSOꢀd6) δ
8.35 (s, 1H), 7.91 (d, J = 2.2 Hz, 2H), 7.71 (q, J = 8.1 Hz, 2H), 5.53 (s, 2H), 4.88 (s, 2H). Mass
spectrum (ESI) m/z 316.9 [M + H]+.
Step 4: 21 (65 mg, 0.2 mmol) was combined with 2,3ꢀdimethoxyꢀ5ꢀ(4,4,5,5ꢀtetramethylꢀ1,3,2ꢀ
dioxaborolanꢀ2ꢀyl)pyridine (54 mg, 0.2 mmol), and Cs2CO3 (133 mg, 0.41 mmol), DMSO (2
mL) and H2O (0.2 mL), Pd(dppf)Cl2 (14.6 mg, 0.02 mmol) in a 10 mL microwave tube and
microwaved at 100 °C for 10 min. The reaction mixture was poured in to EtOAc/H2O and the
organic layer passed through a plug of florisil washing with EtOAc. The filtrate was
concentrated to a residue which was triturated with MeOH and filtered to give 5 (32 mg, 40%) as
a white solid. 1H NMR (300 MHz, DMSOꢀd6) δ 8.37 (s, 1H), 8.11 (d, J = 2.0 Hz, 1H), 7.97 (m,
2H), 7.84 (m, 2H), 7.64 (d, J = 2.0 Hz, 1H), 5.55 (s, 2H), 4.93 (s, 2H), 3.93 (s, 3H), 3.91 (s, 3H).
HRMS found for M+ + 1, 376.1407, C20H17N5O3 requires 376.1404.
2ꢀ(4ꢀ(2ꢀ(5,6ꢀDimethoxypyridinꢀ3ꢀyl)ꢀ5ꢀoxoꢀ5,7ꢀdihydroꢀ6Hꢀpyrrolo[3,4ꢀb]pyridinꢀ6ꢀyl)ꢀ
1Hꢀpyrazolꢀ1ꢀyl)acetonitrile (6). Step 1: A mixture of ethyl 2ꢀmethylꢀ6ꢀoxoꢀ1,6ꢀ
ACS Paragon Plus Environment
23

Page 25 of 42
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
dihydropyridineꢀ3ꢀcarboxylate (5.92 g, 32.6 mmol) in phosphorous oxychloride (45 mL) was
heated at 90 °C for 1 h. After cooling, the reaction mixture was concentrated under reduced
pressure and ice water was added to the residue, followed by addition of 28% ammonium
hydroxide to adjust the pH to 7. The resulting white solid was collected by filtration, washed
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
with ice water, and dried under high vacuum to give 22 (6.18 g, 95%). H NMR (300 MHz,
CDCl3) δ 8.18 (d, J = 8.2 Hz, 1H), 7.27 (d, J = 8.2 Hz, 1H), 4.40 (q, J = 7.1 Hz, 2H), 2.84 (s,
3H), 1.42 (t, J = 7.4, 3H). Mass spectrum (ESI) m/z 200.2 [M + H]+.
Step 2: A mixture of 22 (4.0 g, 20.0 mmol), 2,3ꢀdimethoxyꢀ5ꢀ(4,4,5,5ꢀtetramethylꢀ1,3,2ꢀ
dioxaborolanꢀ2ꢀyl)pyridine (5.84 g, 22.0 mmol), Pd(PPh3)4 (1.15 g, 1.00 mmol), and sodium
carbonate (6.37 g, 60.1 mmol) in a mixture of acetonitrile/water (3:1, 90 mL) was heated at 90
°C under an atmosphere of nitrogen for 4 h. After cooling, the volatiles were removed under
reduced pressure and the residue dissolved in DCM. After washing with water, the organic
phase was dried (Na2SO4), filtered, and concentrated under reduced pressure. The residue was
purified by silica gel chromatography (0ꢀ20% EtOAc/hexanes) to give 23 (5.8 g, 96% yield). 1H
NMR (300 MHz, CDCl3) δ 8.27 (d, J = 1.8 Hz, 1H), 8.17 (d, J = 8.2 Hz, 1H), 7.83 (d, J = 1.8
Hz, 1H), 7.52 (d, J = 8.3 Hz, 1H), 4.32 (q, J = 7.1 Hz, 2H), 4.01 (s, 2H), 3.93 (s, 3H), 2.83 (s,
3H), 1.34 (t, J = 7.1 Hz, 3H). ESMS m/z 303.41 (M+ + 1).
Step 3: To a solution of 23 (4.4 g, 14.6 mmol) in CCl4 (75 mL) was added AIBN (239 mg,
1.46 mmol) and NBS (1.7 g, 9.55 mmol). The mixture was stirred at 80 °C for 1.5 h under an
atmosphere of nitrogen. The reaction mixture was filtered and concentrated. The residue was
purified by silica gel chromatography (20% EtOAc/hexanes) to give a mixture of starting
material and 24 (4.44 g, 47%, ca. 60% pure). 1H NMR (300 MHz, CDCl3) δ 8.39 (d, J = 2.0 Hz,
1H), 8.35 (d, J = 8.3 Hz, 1H), 7.95 (d, J = 2.0 Hz, 1H), 7.74 (d, J = 8.3 Hz, 1H), 5.12 (s, 2H),
ACS Paragon Plus Environment
24

Journal of Medicinal Chemistry
Page 26 of 42
1
2
3
4
5
6
4.47 (q, J = 7.1 Hz, 2H), 4.11 (s, 3H), 4.03 (s, 3H), 1.47 (t, J = 7.1 Hz, 3H). Mass spectrum (ESI)
m/z 303.4 [M + H]+.
7
8
9
Step 4: A solution of ethyl 6ꢀ(bromomethyl)ꢀ5',6'ꢀdimethoxyꢀ[2,3'ꢀbipyridine]ꢀ5ꢀcarboxylate
24 (2.2 g, 3.46 mmol) in DMF (40 mL) at 0 °C was added dropwise over 2 h to a suspension of
2ꢀ(4ꢀaminoꢀ1Hꢀpyrazolꢀ1ꢀyl)acetonitrile (844 mg, 6.9 mmol) and sodium carbonate (732 mg, 6.9
mmol) in DMF (20 mL). After addition was complete, the reaction mixture was stirred at 0 °C
for 2 h and heated at 80 °C for 15 h. Additional sodium carbonate (732 mg) was added and the
reaction mixture was heated at 90 °C for an additional 7 h. After cooling, the mixture was
poured into water and a precipitate formed. The solid was collected by filtration, washed with
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
methyl tꢀbutyl ether, and dried under high vacuum to provide 6 (450 mg, 32%). H NMR (300
MHz, DMSOꢀd6) δ 8.56 (d, J = 1.9 Hz, 1H), 8.39 (s, 1H), 8.22 (d, J = 8.2 Hz, 1H), 8.20 – 8.03
(m, 1H), 8.03 (s, 1H), 7.97 (s, 1H), 5.56 (s, 2H), 5.00 (s, 2H), 3.96 (s, 3H), 3.92 (s, 3H). HRMS
found for M+ + 1, 377.1359, C19H16N6O3 requires 377.1356.
2ꢀ(5,6ꢀDimethoxypyridinꢀ3ꢀyl)ꢀ6ꢀ(1Hꢀpyrazolꢀ4ꢀyl)ꢀ6,7ꢀdihydroꢀ5Hꢀpyrrolo[3,4ꢀ
b]pyridinꢀ5ꢀonec (7). Step 1: A round bottom flask was charged with ethyl 2ꢀmethylpyridineꢀ3ꢀ
carboxylate (50 g, 303 mmol), 1,3,5ꢀtrichloroꢀ1,3,5ꢀtriazinaneꢀ2,4,6ꢀtrione (106 g, 454 mmol)
and DCM (250 mL). The solution was stirred for at room temperature for 12 h and then diluted
with saturated Na2CO3 (600 mL) and DCM (600 mL). The organic layer was separated, washed
with saturated Na2CO3 followed by brine, then dried over Na2SO4, filtered and concentrated to
1
provide 25 (59.8 g, 99%) as a pale yellow oil. H NMR (300 MHz, CDCl3) δ 8.64 (dd, J = 4.8,
1.7 Hz, 1H), 8.21 (dd, J = 7.9, 1.7 Hz, 1H), 7.28 (dd, J = 7.9, 4.8 Hz, 1H), 5.04 (s, 2H), 4.36 (q, J
= 7.1 Hz, 2H), 1.34 (q, J = 7.4 Hz, 3H). Mass spectrum (ESI) m/z 200.0 [M + H]+.
ACS Paragon Plus Environment
25

Page 27 of 42
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Step 2: To a round bottom flask containing 25 (25.5 g, 128 mmol) in DCM (250 mL) was
added mCPBA (32.4 g, 141 mmol). The reaction was stirred at room temperature for 12 h. The
reaction mixture was then washed with saturated aqueous Na2CO3, brine, and then dried over
Na2SO4, filtered and concentrated. The crude material was purified through a silica plug (300 g)
while eluting with EtOAc. Product containing fractions were concentrated to yield 26 (20.8 g,
75%). 1H NMR (300 MHz, MeOHꢀd4) δ 8.68 (dd, J = 4.9, 1.7 Hz, 1H), 8.36 (dd, J = 7.9, 1.7 Hz,
1H), 7.51 (dd, J = 7.9, 4.9 Hz, 1H), 5.10 (s, 2H), 4.43 (q, J = 7.1 Hz, 2H), 1.43 (t, J = 7.1 Hz,
3H). Mass spectrum (ESI) m/z 216.7 [M + H]+.
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
Step 3: 26 (14.7 g, 68.3 mmol) was dissolved in POCl3 (209 g, 1.37 mol) and heated to 90 °C
for 36 h under N2. The reaction was concentrated to dryness, diluted with DCM (300 mL) and
washed with water. The organics were dried over sodium sulfate, filtered, and concentrated to
dryness. The crude residue was purified by flash chromatography (25%ꢀ50% EtOAc/hexanes) to
provide 27 (9.4 g, 59%). 1H NMR (300 MHz, CDCl3) δ 8.26 (d, J = 8.3 Hz, 1H), 7.40 (d, J = 8.3
Hz, 1H), 5.08 (s, 2H), 4.45 (q, J = 7.1 Hz, 2H), 1.51 – 1.36 (m, 3H). Mass spectrum (ESI) m/z
234.3 [M + H]+.
Step 4: To a round bottom flask containing 1Hꢀpyrazolꢀ4ꢀamine (2.04 g, 24.6 mmol) in DMF
(23 mL) at room temperature was added DIEA (3.75 g, 5 mL, 29.0 mmol), followed by dropwise
addition of a solution of 27 (4.53 g, 19.4 mmol) in DMF (23 mL) over 10 min. The reaction was
stirred for 2 h at RT and then heated to 80 °C for 12 h. To this solution was added MeOH (100
mL) and the reaction was allowed to cool. The solid precipitate was filtered and washed with
MeOH (2 x 25 mL). The solid was dried under high vacuum for 8 h to yield 28 (3.02 g, 66%). 1H
NMR (300 MHz, DMSOꢀd6) δ 12.91 (s, 1H), 8.17 (d, J = 8.1 Hz, 1H), 8.12 (s, 1H), 7.90 (d, J =
1.8 Hz, 1H), 7.66 (d, J = 8.1 Hz, 1H), 4.92 (s, 2H). Mass spectrum (ESI) m/z 235.0 [M + H]+.
ACS Paragon Plus Environment
26

Journal of Medicinal Chemistry
Page 28 of 42
1
2
3
4
5
6
7
8
9
2ꢀ(5,6ꢀDimethoxypyridinꢀ3ꢀyl)ꢀ6ꢀ(1ꢀmethylꢀ1Hꢀpyrazolꢀ4ꢀyl)ꢀ6,7ꢀdihydroꢀ5Hꢀpyrrolo[3,4ꢀ
b]pyridinꢀ5ꢀone (8). Step 1: 1ꢀMethylpyrazolꢀ4ꢀamine (46 mg, 0.479 mmol), ethyl 6ꢀchloroꢀ2ꢀ
(chloromethyl)pyridineꢀ3ꢀcarboxylate (112 mg, 0.479 mmol), and DIPEA (110 mL, 0.655 mmol)
were combined in isopropanol (4 mL). The reaction was heated to 100 °C under N2 for 12 h,
concentrated to dryness, then MeOH (5 mL) was added and the slurry was filtered and dried
under vacuum to provide 35 (90 mg, 77%). Mass spectrum (ESI) m/z 249.2 [M + H]+.
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
Step 2: A 10 mL microwave vial was charged with 35 (71 mg, 0.285 mmol), 2,3ꢀdimethoxyꢀ5ꢀ
(4,4,5,5ꢀtetramethylꢀ1,3,2ꢀdioxaborolanꢀ2ꢀyl)pyridine (91 mg, 0.343 mmol), K2CO3 (2M, 2
mmol) and dioxane (2 mL). The solution was sparged with nitrogen for 5 min, followed by
addition of Pd(PPh3)4 (16.4 mg, 0.0142 mmol). The reaction was sealed and microwaved at 150
°C for 10 min. After cooling, the reaction was diluted with EtOAc and the organics were
separated and concentrated. To the crude residue was added MeOH (10 mL), and the resulting
precipitate was collected by filtration, washed with MeOH (2 x 10 mL) and dried under vacuum
to provide 8 (42 mg, 42%). 1H NMR (300 MHz, CDCl3) δ 8.40 (d, J = 2.0 Hz, 1H), 8.24 – 8.14
(m, 2H), 7.88 (dd, J = 13.9, 5.0 Hz, 2H), 7.65 (s, 1H), 4.85 (s, 2H), 4.12 (s, 3H), 4.04 (s, 3H),
3.98 (d, J = 4.1 Hz, 3H). HRMS found for M+ + 1, 352.1407, C18H17N5O3 requires 352.1404.
2ꢀ(5,6ꢀDimethoxypyridinꢀ3ꢀyl)ꢀ6ꢀ(1ꢀethylꢀ1Hꢀpyrazolꢀ4ꢀyl)ꢀ6,7ꢀdihydroꢀ5Hꢀpyrrolo[3,4ꢀ
b]pyridinꢀ5ꢀone (9). Step 1: 1ꢀEthylpyrazolꢀ4ꢀamine (1.06 g, 9.49 mmol), ethyl 6ꢀchloroꢀ2ꢀ
(chloromethyl)pyridineꢀ3ꢀcarboxylate (2.00 g, 8.54 mmol), and DIPEA (1.66 g, 2.2 mL, 12.8
mmol) were combined in acetonitrile (21 mL). The reaction was stirred at rt for 5 min, then
heated to 80 °C under N2 for 18 h. MeOH was added and the reaction slurry was filtered, dried
1
under vacuum to give 36 (1.08 g, 48%) as a light grey solid. H NMR (300 MHz, DMSOꢀd6) δ
ACS Paragon Plus Environment
27

Page 29 of 42
Journal of Medicinal Chemistry
1
2
3
4
5
6
8.19 (d, J = 7.8 Hz, 2H), 7.80 (s, 1H), 7.67 (d, J = 8.1 Hz, 1H), 4.91 (s, 2H), 4.17 (q, J = 7.3 Hz,
2H), 1.38 (t, J = 7.3 Hz, 3H). Mass spectrum (ESI) m/z 263.2 [M + H]+.
7
8
9
Step 2: A 10 mL microwave vial was charged with 36 (75 mg, 0.285 mmol), 2,3ꢀdimethoxyꢀ5ꢀ
(4,4,5,5ꢀtetramethylꢀ1,3,2ꢀdioxaborolanꢀ2ꢀyl)pyridine (91 mg, 0.343 mmol), K2CO3 (2M, 1 mL)
and dioxane (2 mL). The solution was sparged with nitrogen for 5 min, followed by addition of
Pd(PPh3)4 (16.4 mg, 0.0142 mmol). The reaction was sealed and microwaved at 150 °C for 10
min. After cooling, the reaction was diluted with EtOAc (1 mL) and the organics were separated
and concentrated. To the crude residue was added MeOH (10 mL), and the resulting precipitate
was collected by filtration, washed with MeOH (2 x 10 mL) and dried under vacuum to provide
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
9 (40 mg, 38%). H NMR (300 MHz, CDCl3) δ 8.40 (d, J = 2.0 Hz, 1H), 8.27 – 8.15 (m, 2H),
7.88 (dd, J = 14.6, 5.0 Hz, 2H), 7.66 (d, J = 0.5 Hz, 1H), 4.86 (s, 2H), 4.25 (q, J = 7.3 Hz, 2H),
4.13 (s, 3H), 4.04 (s, 3H), 1.56 (t, J = 7.3 Hz, 3H). HRMS found for M+ + 1, 366.1561,
C19H19N5O3 requires 366.1560.
2ꢀ(5,6ꢀDimethoxyꢀ3ꢀpyridyl)ꢀ6ꢀ[1ꢀ(2,2,2ꢀtrifluoroethyl)pyrazolꢀ4ꢀyl]ꢀ7Hꢀpyrrolo[3,4ꢀ
b]pyridinꢀ5ꢀone (10) Step 1: A 250 mL round bottom flask fitted with a reflux condenser was
charged with 28 (1.0 g, 4.26 mmol), sodium carbonate (1 M, 8.52 mmol), 2,3ꢀdimethoxyꢀ5ꢀ
(4,4,5,5ꢀtetramethylꢀ1,3,2ꢀdioxaborolanꢀ2ꢀyl)pyridine (1.35 g, 5.11 mmol) and DMF (85 mL).
The solution was sparged with nitrogen for 30 min, and then Pd(PPh3)4 (985 mg, 0.852 mmol)
was added. The reaction was heated to 90 °C under a nitrogen atmosphere. After 18 h, the
reaction mixture was then diluted with MeOH (100 mL), cooled and the resulting solid was
collected by vacuum filtration. The solid was washed with additional MeOH (2 x 50 mL) and
1
dried under high vacuum to yield 7 (1.30 g, 90%) as pale green solid. H NMR (300 MHz,
ACS Paragon Plus Environment
28