Chemically Diverse Group i p21-Activated Kinase (PAK) Inhibitors Impart Acute Cardiovascular Toxicity with a Narrow Therapeutic Window

Article abstract of DOI:10.1021/acs.jmedchem.6b00638

p21-activated kinase 1 (PAK1) has an important role in transducing signals in several oncogenic pathways. The concept of inhibiting this kinase has garnered significant interest over the past decade, particularly for targeting cancers associated with PAK1 amplification. Animal studies with the selective group I PAK (pan-PAK1, 2, 3) inhibitor G-5555 from the pyrido[2,3-d]pyrimidin-7-one class uncovered acute toxicity with a narrow therapeutic window. To attempt mitigating the toxicity, we introduced significant structural changes, culminating in the discovery of the potent pyridone side chain analogue G-9791. Mouse tolerability studies with this compound, other members of this series, and compounds from two structurally distinct classes revealed persistent toxicity and a correlation of minimum toxic concentrations and PAK1/2 mediated cellular potencies. Broad screening of selected PAK inhibitors revealed PAK1, 2, and 3 as the only overlapping targets. Our data suggest acute cardiovascular toxicity resulting from the inhibition of PAK2, which may be enhanced by PAK1 inhibition, and cautions against continued pursuit of pan-group I PAK inhibitors in drug discovery.

Full text of DOI:10.1021/acs.jmedchem.6b00638

Subscriber access provided by UNIV OF PITTSBURGH
Article
Chemically diverse Group I p21-activated kinase (PAK) inhibitors
impart acute cardiovascular toxicity with a narrow therapeutic window
Joachim Rudolph, Lesley J Murray, Chudi O Ndubaku, Thomas O'Brien, Elizabeth M. Blackwood,
Weiru Wang, Ignacio Aliagas, Lewis Gazzard, James J. Crawford, Joy Drobnick, Wendy Lee,
Xianrui Zhao, Klaus P Hoeflich, David A Favor, Ping Dong, Haiming Zhang, Christopher E Heise,
Angela Oh, Christy C. Ong, Hank La, Paroma Chakravarty, Connie Chan, Diana Jakubiak,
Jennifer Epler, Sreemathy Ramaswamy, Roxanne Vega, Gary Cain, Dolores Diaz, and Yu Zhong
J. Med. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.jmedchem.6b00638 • Publication Date (Web): 11 May 2016
Downloaded from http://pubs.acs.org on May 12, 2016
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 83
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Chemically diverse Group I p21ꢀactivated kinase
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
(
PAK) inhibitors impart acute cardiovascular
toxicity with a narrow therapeutic window
a,*
b
a,†
c
Joachim Rudolph, Lesley J. Murray, Chudi O. Ndubaku, Thomas O’Brien, Elizabeth
c
d
a
a
a
Blackwood, Weiru Wang, Ignacio Aliagas, Lewis Gazzard, James J. Crawford, Joy
a
a
a
c,§
e,#
e
Drobnick, Wendy Lee, Xianrui Zhao, Klaus P. Hoeflich, David A. Favor, Ping Dong,
f
g
d
c
b
Haiming Zhang, Christopher E. Heise, Angela Oh, Christy C. Ong, Hank La, Paroma
h
h
c
c
g
Chakravarty, Connie Chan, Diana Jakubiak, Jennifer Epler, Sreemathy Ramaswamy,
i
i
i
i
Roxanne Vega, Gary Cain, Dolores Diaz, Yu Zhong
a
b
c
Departments of Discovery Chemistry, Drug Metabolism and Pharmacokinetics, Translational
d
f
g
Oncology, Structural Biology, Process Chemistry, Biochemical and Cellular Pharmacology,
h
i
Pharmaceutical Sciences, Safety Assessment, Genentech, Inc., 1 DNA Way, South San
e
Francisco, CA 94080, USA, Shanghai Chempartner Inc., 998 Halei Road, Zhangjiang HiꢀTech
Park, Pudong New Area, Shanghai 201203, People’s Republic of China
ACS Paragon Plus Environment
1

Journal of Medicinal Chemistry
Page 2 of 83
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
ABSTRACT
p21ꢀactivated kinase 1 (PAK1) has an important role in transducing signals in several oncogenic
pathways. The concept of inhibiting this kinase has garnered significant interest over the past
decade, particularly for targeting cancers associated with PAK1 amplification. Animal studies
with the selective Group I PAK (panꢀPAK1, 2, 3) inhibitor Gꢀ5555 from the pyrido[2,3ꢀ
d]pyrimidinꢀ7ꢀone class uncovered acute toxicity with a narrow therapeutic window. To attempt
mitigating the toxicity we introduced significant structural changes, culminating in the discovery
of the potent pyridone side chain analog Gꢀ9791. Mouse tolerability studies with this compound,
other members of this series, and compounds from two structurally distinct classes revealed
persistent toxicity and a correlation of minimum toxic concentrations and PAK1/2 mediated
cellular potencies. Broad screening of selected PAK inhibitors revealed PAK1, 2, and 3 as the
only overlapping targets. Our data suggest acute cardiovascular toxicity resulting from the
inhibition of PAK2, which may be enhanced by PAK1 inhibition, and cautions against continued
pursuit of panꢀGroup I PAK inhibitors in drug discovery.
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
ACS Paragon Plus Environment
2

Page 3 of 83
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Introduction
The p21ꢀactivated kinases (PAKs) are serine/threonine kinases in the STE20 kinase family that
1
ꢀ3
have generated significant interest as therapeutic targets in cancer. The six PAK family
members are divided into two subgroups, group I (PAK1ꢀ3) and group II (PAK4ꢀ6), based on
sequence homology and distinct autoinhibitory regions in group I and II PAKs. PAKs are key
signal transducers in several oncogenic and survival pathways, including Ras, Raf, NFκB, Akt,
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
Bad, and p53. As downstream effectors of the Rho family GTPases Cdcꢀ42, Rac1 and Rho A,
they govern the reorganization of the actin cytoskeleton induced by growth factors and thus,
4
have fundamental roles in regulating cytoskeletal dynamics, cell shape, adhesion and migration.
Moreover, PAK1 is implicated in cellular processes that directly contribute to tumorigenesis,
5
including growth factor pathways, cell proliferation, and proꢀsurvival signaling. Genomic
2
amplification and overꢀexpression of PAK1 are prevalent in luminal breast cancer and other
malignancies, and data from us and others demonstrated a functional dependence of PAK1ꢀ
6
,7
amplified cell lines on PAK1 expression and activity for cell survival and transformation. Not
only is PAK function increased in many human cancers, but occurrence of PAK amplification
1
,8
also correlates with advanced tumor grade and, inversely, patient survival. As a result, PAKs,
3
, 9,10
and PAK1 in particular, have been the subject of significant drug discovery efforts.
We recently described the discovery of a potent and selective inhibitor of the group I PAKs
1
1
(PAK1, 2, and 3), Gꢀ5555 (1). Compound 1 possesses a favorable in vitro pharmacology and in
vivo DMPK profile, rendering it a useful tool for target validation studies. In an array of 23
breast cancer cell lines, 1 had significantly greater growth inhibitory activity in cell lines that
1
2
were PAKꢀamplified compared to nonꢀamplified lines.
ACS Paragon Plus Environment
3

Journal of Medicinal Chemistry
Page 4 of 83
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
In an H292 nonꢀsmall cell lunger cancer (NSCLC) xenograft study in mice, 1 inhibited
phosphorylation of the PAK1/2 downstream substrate mitogenꢀactivated protein kinase 1
(
MEK1) S298 and, when administered at an oral dose of 25 mg/kg b.i.d., imparted 60% tumor
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
13
growth inhibition in this model and a PAK1 amplified breast cancer xenograft model, MDAꢀ
1
2
MBꢀ175. Unfortunately, doses greater than 25 mg/kg b.i.d. and 30 mg/kg q.d. were not
tolerated, and doses of 40 and 50 mg/kg q.d. were associated with death of the majority of study
animals within 2ꢀ4 hours after dosing. Death in these mice was typically preceded by hunched
posture, hypoactivity, lowered body temperature, and blood that was dark and difficult to draw,
indicative of decreased blood flow and cardiac output. Concomitant pharmacokinetic studies
revealed a Cmax of approximately 50 µM, corresponding to a Cmax,u (Cmax unbound) of ~150 nM
as minimum toxic concentration.
Although screening data from a kinase and receptor pharmacology panel revealed no
1
1
immediate explanation or hypothesis, our initial speculation for the cause of the toxicity was
offꢀtarget activity. We decided to address this possibility through chemical modification, and
specifically, variation of the head and tail groups in 1. Compound 1 and analogs in this series
possess a lipophilic head group, and it was deemed advantageous to replace this moiety with a
more polar group to improve drug properties and better balance solubility, clearance, and
permeability requirements. To strike this balance in compound 1 we had to resort to the
introduction of an unorthodox 5ꢀaminoꢀ1,3ꢀdioxanyl tail, a product of considerable fineꢀtuning.
A more polar head group was expected to broaden the range of options for the tail group and
thus, address the goal of introducing structural diversity.
Chemistry
ACS Paragon Plus Environment
4

Page 5 of 83
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
The central pyrido[2,3ꢀd]pyrimidinꢀ7ꢀone core of compounds 1 to 26 was accessed through
cyclocondensation of pyrimidine amino aldehyde building blocks I-4 or I-25 with arylacetic acid
esters I-6, I-9 or I-11. Pyrimidine amino aldehydes I-4a and I-4b were synthesized by reacting
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
4
ꢀchloroꢀ2ꢀ(methylthio)pyrimidineꢀ5ꢀcarboxylic acid ester I-1 with ammonia or ethylamine to
form I-2a and I-2b, ester reduction to a methylalcohol (I-3a, I-3b) in the penultimate and
conversion to an aldehyde functionality in the final step (Scheme 1).
1
4
A chemoselective ReformatskyꢀNegishi reaction provided the foundation to conveniently
access the aryl acetic esters I-9 and I-11. Ethyl 2ꢀbromoacetate I-5 was converted into the
corresponding 2ꢀbromozincacetate, which was further reacted with 4ꢀbromoꢀ2ꢀchloroꢀ1ꢀ
iodobenzene using palladiumꢀcatalyzed coupling to form I-6a. Cꢀlinked analogs were accessed
by converting I-6a into boronic ester I-7 and subsequent Suzuki reaction to form I-9a and I-9b.
Ullmann coupling with 3ꢀmethylpyridinꢀ2ꢀone (I-10a) or 3ꢀmethylpyridinꢀ2ꢀone (I-10b) yielded
aryl acetic esters I-11a and I-11b (Scheme 2).
Compound 2 was synthesized starting with cyclocondensation between I-9a and I-4a followed
by Nꢀalkylation to provide I-13. Oxidation of the thiomethyl group, subsequent SnAr reaction
with methylamine, and deprotection yielded compound 2 (Scheme 3).
Nꢀlinked head group analogs 4 and 7-10 were accessed using a sequence that was similar to the
previously described one; however for the cycloꢀcondensation reaction with pyrimidine amino
aldehyde I-4a, the derivatizable bromo head group aryl acetic acid ester I-6a was used in the first
step to generate I-16. Nꢀalkylation to append the tail moiety (I-17) was followed by Buchwaldꢀ
Hartwig or Ullmann coupling to install several Nꢀlinked head groups. The final steps d, e, and f
paralleled steps c, d, and e in the sequence described in Scheme 3 (Scheme 4a). As illustrated in
ACS Paragon Plus Environment
5

Journal of Medicinal Chemistry
Page 6 of 83
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 synthetic sequence for compounds 5 and 6, Ullmann coupling also succeeded with the leftꢀ
handꢀside methylamino group already installed (Scheme 4b).
Nꢀethyl substituted core analogs 11 and 12 were synthesized by cycloꢀcondensation reaction
between pyrimidine aldehyde I-4b containing a preꢀformed Nꢀethyl group and aryl acetic ester I-
6a. The subsequent transformations b through d paralleled the steps c through e described in
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 3. For compound 12, oxetanꢀ3ꢀamine was used in the final S Ar step (Scheme 5).
N
Compounds 13ꢀ15, 18, 19, 23, 25, and 26 were made analogously to compound 8 using
appropriate tail group precursors, while compounds 16, 17, and 22 required a different method
because of difficulties introducing the amine chain via nucleophilic substitution. The sequence to
access the latter compounds started with an S Ar reaction of pyrimidine aldehyde I-28 with fully
N
elaborated tail group amino building blocks to form I-29a and I-29b. Subsequent steps were
similar to the transformations described in the previous schemes (Scheme 6).
Results and Discussion
We began our head group SAR explorations using the pyrido[2,3ꢀd]pyrimidinꢀ7ꢀone core
substituted with a simple aminopropyl tail group, as earlier studies had found this group to
1
1
15
impart robust potency and solubility (Table 1). In addition to PAK1 potency and associated
ligand lipophilic efficiency (LLE), we monitored selectivity versus PAK4 as representative of a
Group II PAK, and Lymphocyteꢀspecific proteinꢀtyrosine kinase (LCK), a persistent offꢀtarget of
compounds in this series. The primary goals of this exploration were to increase LLE and
decrease human liver microsome (HLM) turnover of reference compound 2, an analog
substituted with the default 6ꢀmethylꢀ2ꢀpyridyl head group.
Our initial design idea was
replacement of this head group with Nꢀlinked heterocyclic systems containing a flanking
ACS Paragon Plus Environment
6

Page 7 of 83
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
carbonyl group. For such systems we predicted a similar dihedral angle between the middle aryl
and terminal ring compared to analogs with terminal pyridine rings and, most importantly, an
opportunity for the carbonyl oxygen to engage in a cationꢀdipole interaction with the catalytic
lysine side chain.
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
First, we established that presence of a head group is necessary for both potency and selectivity
(
R = H; compound 3). Introduction of a simple Nꢀlinked pyrrolidone group (compound 4), while
reducing potency versus 2, maintained LLE and, likely as result of greater polarity, increased
metabolic stability. As we knew that an appropriately placed methyl substituent can boost
11
activity by filling a lipophilic pocket in the head group binding area, we prepared the Nꢀmethyl
substituted imidazolinone 5 and imidazolone 6, and observed further increased potency/LLE. We
were particularly pleased with the data of the unsaturated imidazolone system, with an LLE of
6
.7, significant selectivity vs. both PAK4 and LCK, and moderateꢀtoꢀlow HLM clearance. As
this SAR data suggested advantages of unsaturated versus saturated ring systems we turned to
exploring unsaturated 6ꢀmembered ring systems, such as pyridones. Consistent with the SAR for
the 5ꢀmembered ring systems, introduction of a methyl group at the 3ꢀposition of pyridone
analog 7 increased potency and selectivity (compound 8). We were further heartened by the high
HLM stability of this compound. Encouraged by the fact that a significant amount of polarity
was tolerated in this area of the molecule, we introduced an additional nitrogen atom at two
different ring positions, resulting in the pyrazinone analog 9 and pyrimidinone analog 10. Both
compounds showed comparable PAK1 potency to their pyridone counterpart 8, combined with
exquisite selectivity and low HLMꢀ predicted clearance.
An Xꢀray coꢀcrystal structure of compound 9 bound to PAK1 (PDB: 5IME; resolution = 2.2 Å)
11
and superposition of this structure with the coꢀcrystal structure of compound 1 bound to PAK1
ACS Paragon Plus Environment
7

Journal of Medicinal Chemistry
Page 8 of 83
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
is shown in Figure 2. The most striking difference between these structures is the orientation of
the tail groups; while the 5ꢀaminoꢀ1,3ꢀdioxanyl motif of compound 1 forms a hydrogen bond
interaction with the backbone carbonyl of Asp393 (not shown), the aminopropyl group of 9
forms a hydrogen bond with the DFGꢀaspartate side chain (Asp407). The pyrazinone head group
of 9 is tightly packed against the αꢀCꢀhelix in the pocket surrounded by Lys299, Glu315, Ile316
and Val342, and the general shape of the protein residues lining the deep pocket is qualitatively
similar in both structures. The main difference is the carbonyl group in the pyrazinone group
with the oxygen atom projecting a 3.0 Å cation dipole interaction toward the catalytic lysine 299
side chain, which had been our original design objective. Furthermore, the crystal structure of
compound 9 shows an outꢀofꢀplane (OOP) angle of 65° for the biaryl system (pyrazinoneꢀ
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
1
0
phenyl), which is very close to the calculated ground state and similar to the observed and
calculated OOP angle for compound 8 (see quantumꢀmechanical calculations in the supplemental
material). Having identified less lipophilic head groups than pyridyl, we turned our attention to
replacing the strongly basic amine tail (cpKa ~ 11) with less basic or neutral tail groups to
balance polarity and permeability (Table 2). Among the head groups identified above, the
pyridone group was deemed best suited to reach a desirable logD range and hence was kept
constant. A particular focus in this effort was optimization of cellular potency, as determined
from the inhibition of phosphorylation of the PAK1/2 downstream substrate MEK1 S298.
Introduction of a simple ethyl tail group (compound 11) imparted single digit nanomolar
PAK1 potency and high permeability in MadinꢀDarby canine kidney (MDCK) cells, however
less desirable cellular potency (pMEK IC50 = 340 nM) and high HLMꢀpredicted clearance.
1
Earlier SAR studies had found the methyl group as hinge binder substituent, R , to impart
10
optimal properties but oxetanyl was found to be an alternative option with the potential to
ACS Paragon Plus Environment
8

Page 9 of 83
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
increase cellular potency and selectivity. We tried oxetanyl as replacement for methyl in 11
(compound 12), and indeed selectivity vs. LCK and cell potency were both increased, however
HLMꢀpredicted clearance remained high. Furthermore, both compounds showed poor aqueous
solubility. To address these shortcomings, we turned to tail groups that possessed a low level of
polarity. The hydroxyethyl analog 13 exhibited insufficient potency and strikingly lower MDCK
permeability compared to its ethyl tail counterpart 11. Both liabilities were addressed with the
corresponding methoxyethyl analog 14, but cell potency was still poor. We successfully replaced
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
the methyl hinge binder group R with oxetanyl, leading again to a gain in selectivity versus
LCK and cell potency (compound 15). Both compounds showed improved aqueous solubility
versus their ethyl tail counterparts (80 and 38 µM, respectively, at pH 7.4).
Our earlier work had demonstrated that amine containing tail groups can have a profound
11
effect on potency in this series, even if basicity is substantially attenuated, and we decided to
reꢀexplore this direction. Introduction of two fluorine atoms to the tail group of 8, as realized in
compound 16, led to a drop of pKa from ~11 to 7.3, resulting in a large gain in permeability from
ꢀ6
16
0
.1 to 8.4 10 cm/s (MDCK A to B). This compound was not sufficiently potent but we were
encouraged by the compatibility of the pyridone head group with a low pKa amine tail group
with respect to cell permeability. Further variation of the tail group led to compounds with
double digit nanomolar cellular potency. While the Nꢀmethyl ethoxyazetidine analog 18 was
hampered by low MDCK permeability and high efflux, presumably because of high basicity, the
less basic tail analogs 17 and 19 (G-9791) possessed good permeability. A clear standout was the
3ꢀfluoroazetidine tail analog 19, a weakly basic compound with high biochemical and cellular
potency, and good solubility.
ACS Paragon Plus Environment
9

Journal of Medicinal Chemistry
Page 10 of 83
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
Based on its favorable LLE and strong cell potency, we moved forward with evaluating 19 in a
mouse pharmacokinetic study using a standard oral dose of 25 mg/kg. Disappointingly, the
animal with the highest exposure in this study group (n=3) (plasma concentration at the 15 min
time point was 2.75 µM), died at ~ 2 hours post dosing. The other two study mice survived
(plasma concentrations were 2.24 and 1.71 µM, respectively), however, their blood was
darkened and difficult to draw, similar to what was observed in the efficacy study with PAK
inhibitor 1. A total drug concentration of 2.75 µM corresponds to an unbound concentration of
88 nM, a value close to the IC50 in the phosphoꢀMEK cellular assay (33 nM). This close
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
relationship between minimum adverse free Cmax and cellular IC paralleled the observations for
50
compound 1. Considering the two major structural differences between 19 and 1, the possibility
of onꢀtarget toxicity became a growing concern, and hence we selected a structurally diverse set
of PAK inhibitors for singleꢀdose tolerability studies in mice (Tables 3). All compounds were
ATP competitive inhibitors and, as a result of high homology in the ATP binding site, showed
similar inhibition of PAK1 and PAK2. Tolerability studies were conducted at multiple doses,
using either oral (p.o.) or intraperitoneal (i.p.) dosing as appropriate, and minimum unbound drug
concentrations associated with adverse effects (Cmax,adverse,u) were compared to pMEK IC50
values.
As shown in Table 3, none of the compounds investigated, except an inactive control
compound (26) with close structural similarity to one of the active compounds tested (21) were
tolerated, and in many instances dosing resulted in lethality. Since all compounds shown in
Table 3 stem from the same class (pyrido[2,3ꢀd]pyrimidinꢀ7ꢀones), we conducted similar
tolerability studies with molecules from two structurally distinct series, 27, a PAK inhibitor from
13
the aminopyridine hinge binder class previously pursued by us, and PFꢀ3758309 (28), a panꢀ
ACS Paragon Plus Environment
10

Page 11 of 83
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
17
PAK inhibitor that had entered Phase I clinical studies, but was later discontinued (Table 4). As
it was challenging to achieve sufficient drug exposures with these compounds in mice by oral
dosing, both compounds were administered intraperitoneally (i.p.). Both compounds, at 25 and
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
0 mg/kg i.p., for 27 and 28, respectively, displayed intolerability similar to the compounds
shown in Table 3. Analyzing the entire data set in more detail, we noticed a striking correlation
between Cmax,adverse,u and pMEK IC values (Figure 3). Compounds with higher cell potency
50
caused adverse effects at lower concentrations while compounds with reduced cell potency were
better tolerated. This relationship was chemical class independent, and the value for Cmax, adverse,u
was consistently in a range of 1 to 4ꢀfold over the pMEK IC value. This clearly supported our
5
0
onꢀtarget toxicity hypothesis.
To rule out coincidental overlap of offꢀtarget activities we performed broader offꢀtarget
profiling of three structurally distinct compounds, 1, 23, a compound with a polar head group
and devoid of a basic amine tail, and 28, a panꢀPAK inhibitor from an entirely different series.
This included testing in a large kinase panel as well as a secondary pharmacology screening
panel. As illustrated in the Venn diagrams in Figure 4, the only overlapping kinase activities
were PAK1, 2, and 3, and there was no overlap in activities within the secondary pharmacology
panel. In addition, 1 was tested in a specific panel of pharmacology assays related to ion channel
+
2+
targets with known cardiac functions (hERG, hNav1.1, hNav1.5, hCav3.2, Na /Ca ATPase,
+
+
Na /K ATPase), and no significant activities were observed (see supporting information for
details). Thus, no apparent shared biological targets outside of Group I PAK isoforms could be
identified for this group of compounds.
Inꢀlife observations in the described mouse tolerability experiments suggested implication of
the cardiovascular system and, to further investigate the nature of the toxicity, we examined the
ACS Paragon Plus Environment
11

Journal of Medicinal Chemistry
Page 12 of 83
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
effects of tool compound 1 on a Langendorff heart preparation isolated from rats. The
Langendorff heart assay is a widely used ex vivo model that allows the examination of cardiac
hemodynamic and electrocardiographic parameters without the confounding reflex components
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
that are present in an intact animal (such as neuronal and hormonal effects), and we expected to
gain clarity on the mode of toxicity from such a study. Hearts from Sprague Dawley rats were
paced via the right atrium at 10ꢀ15% above the intrinsic heart rate and exposed to increasing
concentrations of compound 1 (0, 0.5, 1.5, 5 and 15 µM, n=5) for 15 minutes at each
concentration. Exposure to 1.5, 5 and 15 µM of 1 resulted in a significant increase of 64ꢀ83% in
coronary perfusion pressure (CPP). Notable decreases in left ventricular developed pressure
(
LVDP), dP/dt maximum, dP/dt minimum, dP/dtmax/P, and dP/dtmax/P at 40 mmHg as well as
increases in left ventricular systolic pressure (LVSP) and left ventricular end diastolic pressure
LVEDP) were observed. In addition, drug concentrations of 5 and 15 µM led to significant
(
increases of 37 and 34%, respectively, in PR interval, and loss of 1:1 atrioventricular conduction
occurred in one heart at these concentrations. Ventricular fibrillation (VF) occurred in another
heart after perfusion with 5 µM of 1. There were no substantive effects at concentrations of ≤ 15
µM of 1 on HR or QRS duration. In summary, compound 1 caused a decline of the left ventricle
function, arrhythmia, and heart failure in isolated rat hearts, consistent with the acute adverse
effects observed in mice. In addition, the lowest concentration associated with effects (0.5 – 1.5
µM) lays within only 3ꢀ10ꢀfold of Cmax,adverse,u (150 nM). While we did not perform systematic
19
rat tolerability studies, one of the PAK1 inhibitors shown in Table 3, FRAX1036 (20), was
investigated in a rat singleꢀdose experiment and found to cause lethality at a Cmax of 4 µM,
corresponding to a Cmax,adverse,u of 0.076 µM. This is again in close proximity to the pMEK S298
IC of this compound, 0.22 µM. Furthermore, rats dosed with 20 displayed similar clinical
5
0
ACS Paragon Plus Environment
12

Page 13 of 83
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
signs as the mice (slow blood flow slow and darkened blood, hypoactivity, and hunched
posture). This suggests that the acute cardiovascular toxicity imparted by PAK1/2 inhibitors is
not limited to mice. The interꢀspecies homology of PAK1 and 2 is high (species conservation vs.
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
3
human PAK1: mouse 99.26%, rat 99.08%; PAK2: mouse 97.14%, rat 96.56%), and
pharmacological translatability was therefore expected.
PAK tissue expression is also consistent with the onꢀtarget toxicity hypothesis. PAK1 and
2
0
PAK2 are broadly expressed, with PAK1 levels being high in brain, muscle, and heart, and
2
1
PAK2 levels high in endothelial cells. Conversely, PAK3 is almost exclusively found in the
21
brain, and based on this and the fact that our compounds are unlikely to be brain exposed due
−
6
to considerable efflux (e.g., MDCKꢀMDR1 AꢀB flux is 0.3 10 cm/s and the efflux ratio 68.9
for compound 1) we ruled out its association with the toxicity findings. PAK1 knockout mice
are viable, but recent studies with PAK1ꢀdeficient mice revealed vulnerability to cardiac
hypertrophy and fast progression to heart failure under sustained pressure overload, in addition to
2
0
susceptibility to ischemia and reperfusion injury. This is consistent with the recently discovered
role of PAK1 in cardiac physiology, which includes the regulation of cardiac ion channel and
2
0
1
actomyosin function. PAK2 has an important role in angiogenesis during development and
PAK2 knockout mice die early in embryogenesis due to multiple developmental abnormalities,
2
1, 22
most prominently those involving defective vascularization.
In adult mice, PAK2 plays
important roles in maintaining the integrity of blood vessels, and ubiquitous PAK2 deletion in
23
adult conditional knockout mice also resulted in lethality. Given these known functions of
PAK1 and 2 in heart and vasculature, toxicity as a consequence of inhibition of both targets was
plausible.
ACS Paragon Plus Environment
13

Journal of Medicinal Chemistry
Page 14 of 83
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
Conclusions
The question of whether toxicity imparted by small molecules is caused by onꢀtarget or offꢀ
target pharmacological effects is critical for stop/go decisions of drug discovery projects. Onꢀ
target toxicity is related to an exaggerated pharmacological effect on the primary target and
typically not possible to eliminate. Offꢀtarget toxicities may be related to the chemical, structural,
or physicochemical properties and therefore are considered solvable with chemical
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
2
4
modifications. Common key characteristics of onꢀtarget toxicities are outlined in Table 5. For
onꢀtarget toxicity, the target is usually expressed in the tissue associated with toxicity, and the
toxicity is oftentimes consistent with the phenotype observed in knockout models. As discussed
above, both PAK1 and PAK2 are broadly expressed and modulate cardiac and vasculature
function and/or development. In addition, if toxicity is onꢀtarget, the therapeutic window is
25
typically expected to be narrow, as it was the case with the PAK1/2 inhibitors, with ratios
between Cmax,adverse,u and pMEK IC50 values within 1ꢀ4. Furthermore, the ratios remained
constant across 11 compounds with PAK1 potency spanning > 2 log units. In the example of a
structurally closely related inactive/active pair (26 and 21), only the active analog 21 was toxic in
vivo. Another indication of toxicity being onꢀtarget is its presence with two or more chemical
series. Here we have shown that compounds from three chemotypes with no apparent common
offꢀtargets were all acutely toxic in mice. The weight of evidence thus suggests that the toxicity
caused by panꢀGroup I PAK inhibitors is onꢀtarget.
One hypothesis is that combined inhibition of PAK1 and PAK2 cause the observed toxicity. It
is unlikely that inhibition of PAK1 alone suffices in driving the observed toxicity, as PAK1
knockout mice, in contrast to PAK2 knockout mice, are viable. However, we cannot rule out the
possibility that developmental compensation could mask the effects of PAK1 deletion on
ACS Paragon Plus Environment
14

Page 15 of 83
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
26
cardiovascular function. Another hypothesis is that PAK2 inhibition alone is driving the
2
1, 22
toxicity. PAK2 is not only critical in embryonic vasculature development,
but is also
important for the integrity and function of blood vessels in adult mice, and inducible ubiquitous
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
23
PAK2 deletion was lethal in adult mice.
Both hypotheses lead to the possibility that selective PAK1 inhibition may not result in
cardiovascular toxicity and might therefore offer a viable therapeutic strategy. In fact, allosteric
PAK1 isoform selective inhibitors with biochemical selectivity versus PAK2 of > 50ꢀfold were
2
7
recently reported, although no appreciable inhibition of phosphorylation of the downstream
substrate MEK1 S298 was observed. The authors concluded that for cell lines that are dependent
on and express both PAK1 and PAK2, a dual PAK1/2 inhibitor would likely be necessary for
driving cellular efficacy. Data from us and others indicate that PAK1 amplified tumors are
6,28
highly dependent on PAK1 signaling
suggesting that this may be one possibility where a
PAK1 isoform specific inhibitor could be therapeutically viable, but additional work is needed to
validate this proposition.
In summary, this study describes the structural diversification of our previous lead compound
1, a Group I PAK inhibitor from the pyrido[2,3ꢀd]pyrimidinꢀ7ꢀone class, in an attempt to dial out
acute toxicity that was initially believed to be offꢀtarget. Considerable chemical variation of both
head and tail groups yielded compound 19, a potent Group I PAK inhibitor with improved
cellular potency and solubility, but with a similar toxicity profile. Mouse tolerability studies with
nine PAK1/2 inhibitors from the pyrido[2,3ꢀd]pyrimidinꢀ7ꢀone series and two compounds from
structurally distinct chemical classes revealed consistent acute toxicity, and we found that
unbound Cmax associated with adverse effects in mice correlated with cellular inhibition of
phosphoꢀMEK1 S298. A mechanistic study in isolated rat hearts with compound 1 uncovered a
ACS Paragon Plus Environment
15

Journal of Medicinal Chemistry
Page 16 of 83
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
doseꢀdependent decline of function in the left ventricle, arrhythmia and heart failure, in
alignment with the acute lethality in mouse tolerability studies. PAK1 and 2 are expressed in
heart and vasculature tissues; PAK1 is known to be involved in cardiac function and PAK2 has a
role in maintaining vascular integrity. As PAK1 knockꢀout mice are viable and PAK2 knockꢀout
mice lethal it is likely that PAK2 inhibition is the major contributor to the toxicity. The only
potential path forward for Group I PAK inhibitors might be with a selective PAK1 inhibitor in
the context of PAK1 amplified tumors. However, given that PAK1 has been implicated in
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
2
0
cardiac function, a compound with such a profile would need to be carefully evaluated to
ensure it has an acceptable safety profile. Although kinases have been implicated in
2
9
cardiovascular functions, we are not aware of kinase inhibitors that elicited cardiovascular onꢀ
target toxicity of similar acuteness and severity as observed with panꢀGroup I PAK inhibitors.
This suggests fundamentally important functions of PAK2, either alone or in combination with
PAK1, in cardiovascular physiology.
ACS Paragon Plus Environment
16

Page 17 of 83
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
EXPERIMENTAL SECTION
All chemicals were purchased from commercial suppliers and used as received. Flash
chromatography was carried out with prepacked silica cartridges from either ISCO or SiliCycle
on an ISCO Companion chromatography system using gradient elution. NMR spectra were
recorded on a Bruker AV III 400 NMR spectrometer at 300K or 350K and referenced to
tetramethylsilane. Preparative HPLC was performed on a Polaris C18 5 ꢁm column (50 mm × 21
mm), eluting with mixtures of water/acetonitrile. All final compounds were purified to >95%
chemical and optical purity, as assayed by LC/MS. This analysis was performed on an Agilent
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
1100 HPLC coupled with Agilent MSD mass spectrometer using ESI as the ionization source.
The LC separation was performed using an Agilent ZORBAX SBꢀC18, 1.8 mm, 30 × 2.1 mm
column with a flow rate of 0.4 mL/min. The gradient consisted of 3−97% acetonitrile in water,
each containing 0.05% TFA, over 7 min; 97% acetonitrile was then maintained for 1.5 min,
followed by reꢀequilibration at 3% acetonitrile for 1.5 min. The LC column temperature was
maintained at 40 °C. UV absorbance was measured at 220 and 254 nm along with a full scan
mass spectrum. For select compounds, a longer gradient LC was used (method B). In these cases,
the gradient consisted of 2−98% acetonitrile in water, each containing 0.05% TFA, over 25.5
min; 98% acetonitrile was then maintained for 2.5 min, followed by reꢀequilibration at 2%
acetonitrile for 1.5 min. UV and MS detection was performed as before. The synthesis of
1
1
compound 1 was described in an earlier publication, and compound 28 was prepared according
17
to a procedure described in the literature. Compound 20 was obtained from Afraxis, Inc.
Ethyl 4-amino-2-(methylthio)pyrimidine-5-carboxylate (I-2a). To a 2 L threeꢀnecked flask
was added ethyl 4ꢀchloroꢀ2ꢀ(methylthio)pyrimidineꢀ5ꢀcarboxylate (I-1, 100 g, 0.43 mol), THF
ACS Paragon Plus Environment
17

Journal of Medicinal Chemistry
Page 18 of 83
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
(
500 mL), and triethylamine (186 mL, 1.29 mol). The ammonium hydroxide (28% in water, 400
mL) was added by portions to keep the internal temperature below 30 °C. After 2 h, water (1000
mL) was added and THF was distillated out under vacuum. The resulted solid was filtrated and
dried under vacuum at 50 °C to afford the title compound (90 g, 99% yield). LCMS (ESI): m/z =
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
+
1
213.9 [M+1] ; H NMR (400 MHz, DMSOꢀd ) δ 8.57 (s, 1H), 8.03 (br, 1H), 7.65 (br, 1H), 4.27
6
(q, J = 7.5 Hz, 2H), 2.46 (s, 3H), 1.29 (t, J = 7.2 Hz, 3H).
Ethyl 4-(ethylamino)-2-(methylthio)pyrimidine-5-carboxylate (I-2b). To a 3 L round bottom
flask equipped with a mechanical stirrer under N atmosphere was charged 4ꢀchloroꢀ2ꢀ
2
methylthioꢀ5ꢀpyrimidinecarboxalate (I-1, 200 g, 0.86 mol), trimethylamine (3.0 equiv, 352 mL),
then NH .HCl (140.2 g, 2.0 equiv) in DCM (5 mL/g, 1000 mL) was added into the solution at 0ꢀ
2
o
o
5 C. After the addition, the mixture was warmed to 20ꢀ25 C and the mixture was stirred at this
temperature for the completion. The reaction was quenched with water, washed with aq. NH Cl
4
and then the DCM layer was concentrated under reduced pressure to give a light yellow liquid
1
(191 g, 92% yield) which was used for the next step without further purification, H NMR (400
MHz, DMSOꢀd ): δ 8.51 (s, 1H), 8.29 (t, 1H, J = 5.4 Hz), 4.27 (q, 2H, J = 7.2 Hz), 3.49ꢀ3.54 (m,
6
2H), 2.47 (s, 3H), 1.30 (t, 3H J = 7 Hz), 1.16 (q, 3H, J = 7.2 Hz).
(
4-Amino-2-(methylthio)pyrimidin-5-yl)methanol (I-3a). To a 3 L threeꢀnecked flask was
added lithium aluminum hydride (I-2a, 25.7 g, 0.68 mol) and THF (1 L) and the suspension was
cooled to −10−0 °C. A solution of ethyl 4ꢀaminoꢀ2ꢀ(methylthio)pyrimidineꢀ 5ꢀcarboxylate (120
g, 0.56 mol) in THF (1 L) was slowly added through an addition funnel to the stirring mixture.
The reaction mixture was then allowed to warm to 20−25 °C. After 2 h, the mixture was cooled
to 0 °C, and slowly quenched by water (26 mL). Then 10% sodium hydroxide (26 mL) was
added and agitated for 2 h, which was followed by the addition of 78 mL (water). After filtration,
ACS Paragon Plus Environment
18

Page 19 of 83
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
the solvent was removed to dryness to afford 80 g crude product. This was used in the next step
+
without further purification. LCMS (ESI): m/z = 171.8 [M+1] .
(4-(Ethylamino)-2-(methylthio)pyrimidin-5-yl)methanol (I-3b). To a 3 L round bottom flask
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
equipped with a mechanical stirrer under N atmosphere was charged lithium alumimum hydride
2
o
(1.5 equiv, 28.4 g) and THF (960 mL) and the mixture was cooled down at below 10 C. A
solution of ethyl 4ꢀ(ethylamino)ꢀ2ꢀ(methylthio)pyrimidineꢀ5ꢀcarboxylate (I-2b) in THF (8
o
mL/g, 960 mL) was added into the solution while maintaining the inner temperature at 0ꢀ10 C.
o
The reaction mixture was then warmed to 20 C and stirred at this temperature for 1 h. The
o
reaction was determined complete by HPLC. The reaction mixture was cooled to 0ꢀ10 C and
quenched with 20 mL water and 10% aq sodium hydroxide (20 mL). The mixture was filtered
and the wet cake was washed with THF (300 mL × 2). The filtrate was concentrated to remove
1
the solvent to afford the title compound as white solid (85.2 g, 86% yield). H NMR (400 MHz,
DMSOꢀd6): δ 7.82 (s, 1H), 6.82 (t, 1H, J = 5.4 Hz), 5.12 (t, 1H, J = 5.4 Hz), 4.30 (d, 2H, J = 5.2
Hz), 3.39 (q, 2H, J = 6Hz), 2.41 (s, 3H), 1.14 (t, 3H, J = 7.2 Hz).
4-Amino-2-(methylthio)pyrimidine-5-carbaldehyde (I-4a). To a 3 L threeꢀnecked flask was
added (4ꢀaminoꢀ2ꢀ(methylthio)pyrimidinꢀ5ꢀyl)methanol (I-3a, 100 g, 0.58 mol), activated
manganese dioxide (152 g, 1.75 mol), and THF (2000 mL). The mixture was heated to 40−45 °C
for 16 h. The reaction mixture was cooled to 20−25 °C and filtered through celite pad and rinsed
with THF (400 mL × 3). The combined filtrate was concentrated to dryness. Reslurrying in
200 mL EtOAc/nꢀheptane (1:5) at 20−25 °C yielded pure product (51 g, 51.6% yield). LCMS
+
1
(
ESI): m/z =169.8 [M+1] ; H NMR (400 MHz, CDCl ): δ 9.80 (s, 1H), 8.44 (s, 1H), 8.22 (br,
3
1H), 5.77 (br, 1H), 2.57 (s, 3H).
4-(Ethylamino)-2-(methylthio)pyrimidine-5-carbaldehyde (I-4b). To a 3 L round bottom flask
ACS Paragon Plus Environment
19

Journal of Medicinal Chemistry
Page 20 of 83
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
equipped with a mechanical stirrer under N atmosphere was charged (4ꢀ(ethylamino)ꢀ2ꢀ
2
(methylthio)pyrimidinꢀ5ꢀyl)methanol (I-3b, 150 g, 1.0 equiv), activated manganese dioxide (5.0
o
equiv, 327 g) and THF (15 mL/g, 2250 mL). The mixture was heated to 40 C for 16 h for
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
o
completion. The mixture was then cooled to 20 C and filtered through a pad of celite. The cake
was rinsed with 500 mL (2x) of THF. The combined filtrate was concentrated to remove solvent
to give a light yellow solid. The solid was reꢀslurried in EtOAc to obtain a white solid (123 g,
1
8
2% yield). H NMR (400 MHz, CDCl ): δ 9.71 (s, 1H), 8.30 (s, 1H), 3.49ꢀ3.53 (m, 2H), 2.57 (s,
3
3H), 1.30 (t, 3H, J = 7.4 Hz).
Ethyl 2-(4-bromo-2-chlorophenyl)acetate (I-6a). To a 3 L threeꢀnecked flask under nitrogen
was added zinc dust (117 g, 1.79 mol), THF (0.3 L). Chlorotrimethylsilane (10.4 mL) was added
dropwise to control the internal temperature below 27 °C. The mixture was allowed to stir at
25−27 °C for 30 min and then heated to 30 °C. A solution of ethyl bromoacetate (150 g, 0.90
mol) in THF (1.2 L) was slowly added dropwise to the reaction mixture (the internal temperature
should be below 50 °C). After the addition was complete, the reaction mixture was allowed to
cool back to 25−30 °C. The mixture was filtered through celite pad under nitrogen to afford a
yellow (occasionally also greenish or orange) solution of bromoꢀ(2ꢀethoxyꢀ2ꢀoxoethyl)zinc. The
concentration was titrated to be 0.50 M.
To a 3 L threeꢀnecked flask under nitrogen were added 4ꢀbromoꢀ2ꢀchloroꢀ1ꢀiodoꢀbenzene (80 g,
0.25 mol), bis(dibenzylideneacetone)palladium
(7.2 g, 0.0125 mol), 4,5ꢀbis(diphenylꢀ
phosphino)ꢀ9,9ꢀdimethylxanthene (Xantphos, 7.2 g, 0.0125 mol), and THF (800 mL). The
mixture was degassed and backfilled with nitrogen three times. Bromoꢀ(2ꢀethoxyꢀ2ꢀ
oxoethyl)zinc (960 mL, 0.48 mol) was added to the reaction mixture, followed by heating to
65 °C. The reaction was complete after 1 h. The reaction mixture was cooled to 30 °C and
ACS Paragon Plus Environment
20

Page 21 of 83
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
quenched with aq. 1N HCl (400 mL) and 25% brine (400 mL). The organic layer was separated,
filtered through a celite pad, and concentrated under vacuum. The crude material was purified by
silica column chromatography (0−30% EtOAc/nꢀheptane) to afford the product as light yellow
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
oil (60 g, 86% yield). H NMR (400 MHz, CDCl ) δ 7.48 (d, J = 2.0 Hz, 1H), 7.29 (dd, J = 8.0
3
1
Hz, J = 2.0 Hz, 1H), 7.08 (d, J = 8 Hz, 1H), 4.10 (q, J = 6.8 Hz, 2H), 3.64 (s, 2H), 1.18 (t, J =
2
6.8 Hz, 3H).
Ethyl 2-(2-chloro-4-(4,4,5,5-tetramethyl-1,3,2-dioxaborolan-2-yl)phenyl)acetate (I-7). To a
500 mL threeꢀnecked flask was added ethyl 2ꢀ(4ꢀbromoꢀ2ꢀchlorophenyl)acetate (I-6a, 50 g, 0.18
mol), 4,4,4',4',5,5,5',5'ꢀoctamethylꢀ2,2'ꢀbi(1,3,2ꢀdioxaborolane) (62.5 g, 0.24 mol), potassium
acetate (33 g, 0.34 mol), 1,1'ꢀbis(diphenylphosphino)ferroceneꢀpalladium(II)dichloride DCM
complex (10.4 g, 0.012 mol) and 1,4ꢀdioxane (250 mL). The mixture was degassed and
backfilled with nitrogen three times. It was then heated at refluxing for 20 h. This mixture was
filtered. 25% brine (200 mL) was added, and the resulted mixture was extracted with EtOAc
(
300 mL × 3). The organic layer was dried over anhydrous sodium sulfate, filtered and
concentrated to give crude product. The crude material was then purified by silica column
chromatography (5%−25% EtOAc/nꢀheptane) to afford the title compound (54 g, 90% yield).
+
LCMS (ESI): m/z =324.6 [M+1] .
Ethyl 2-(2-methyl-4-(6-methylpyridin-2-yl)phenyl)acetate (I-9a). To a 2000mL jacket reactor
equipped with a mechanical stirrer under an atmosphere of N was charged ethyl 2ꢀ(2ꢀmethylꢀ4ꢀ
2
(
4,4,5,5ꢀtetramethylꢀ1,3,2ꢀdioxaborolanꢀ2ꢀyl)phenyl)acetate (I-7, 60 g, 1.0 equiv) and 2ꢀchloroꢀ
6
ꢀmethylpyridine (I-8a, 25 g, 1.0 equiv) followed by 1200 mL of toluene, 120 mL of EtOH and
2
40 mL of water. To the mixture was charged Pd(dppf)Cl (12 g, 20 wt%) in one portion. The
2
reactor was filled with N and subsequently heated until complete of consumption of starting
2
ACS Paragon Plus Environment
21

Journal of Medicinal Chemistry
Page 22 of 83
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
material was observed (deemed to complete when s.m./product <10%). The reaction mixture was
o
cooled to 20 C and undissolved substances filtered out. The filtrate was concentrated to remove
the solvent, and the residue purified by silica gel chromatograpgy to give the title compound in
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
1
9
7.7% purity (54 g, 51% yield). H NMR (400 MHz, DMSOꢀd6): δ 7.89 (s, 1H), 7.81ꢀ7.84 (m,
1
2
1H), 7.73 (t, 2H, J = 4.6 Hz), 7.29 (d, 1H, J = 8 Hz), 7.20 (dd, 1H, J = 1.4 Hz, J = 7 Hz), 4.10
(q, 2H, J = 7.2 Hz), 3.73 (s, 2H), 2.53 (s, 3H), 2.31 (s, 3H), 1.19 (t, 3H, J = 7.2 Hz). LCMS
+
(
ESI): m/z = 270.2 [M+1] .
Ethyl 2-(2-chloro-4-(6-methylpyrazin-2-yl)phenyl)acetate (I-9b). To a 2 L threeꢀnecked flask
was added ethyl 2ꢀ(2ꢀchloroꢀ4ꢀ(4,4,5,5ꢀtetramethylꢀ1,3,2ꢀ dioxaborolanꢀ2ꢀyl)phenyl)acetate (I-7,
5
4 g, 0.17 mol), potassium acetate (33 g, 0.34 mol), 1,1'ꢀbis(diphenylphosphino)ferroceneꢀ
palladium(II)dichloride DCM complex (10.4 g, 0.012 mol), 2ꢀchloroꢀ6 methylpyrazine (I-8b,
1.3 g, 0.17 mol) and 1,4ꢀdioxane/water (2:1, 750 mL). The mixture was degassed and backfilled
2
with nitrogen (3×). Then it was heated to 100 °C for 15 h. After cooling to 20−25 °C the mixture
was diluted with water (200 mL) and extracted with EtOAc (300 mL × 3). The combined organic
layers were washed with 25% brine (100 mL × 2), and dried over anhydrous sodium sulfate. The
crude material was purified by silica gel column chromatography (5%−25% EtOAc/nꢀheptane),
followed by slurrying in nꢀheptane. The title compound was obtained as white solid (18 g, 37.5%
1
yield). H NMR (400 MHz, CDCl ): δ 8.81 (s, 1H), 8.42 (s, 1H), 8.10(d, J = 1.6 Hz, 1H), 7.85
3
(d, J = 8 Hz, 1H), 7.43 (d, J = 8 Hz, 1H), 4.20 (q, J = 7.2 Hz, 2H), 3.84 (s, 2H), 2.64 (s, 3H), 1.28
(t, J = 7.2 Hz, 3H).
Methyl 2-(2-chloro-4-(3-methyl-2-oxopyridin-1(2H)-yl)phenyl)acetate (I-11a). A mixture of
methyl 2ꢀ(4ꢀbromoꢀ2ꢀchlorophenyl)acetate (I-6b, prepared analogously to I-6a, 100 mg,
1
2
0
.38 mmol), 3ꢀmethylpyridinꢀ2(1H)ꢀone (I-10a, 62 mg, 0.57 mmol), N ,N ꢀdimethylethaneꢀ1,2ꢀ
ACS Paragon Plus Environment
22

Page 23 of 83
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
diamine (18 mg, 0.20 mmol), copper(I) iodide (20 mg, 0.10 mmol) and potassium phosphate
(550 mg, 1.71 mmol) in 1,4ꢀdioxane (1.5 mL) was stirred in a sealed tube at 110 °C under a
nitrogen atmosphere for 48 h. After cooling to rt and filtration, the filtrate was concentrated and
purified by silica gel chromatography using DCM/methanol (100:1) as eluting solvents to afford
the title compound as yellow solid (60 mg, 54% yield). LCMS (ESI): m/z = 292.1 [M+1].
Methyl 2-(2-chloro-4-(3-methyl-2-oxopyrazin-1(2H)-yl)phenyl)acetate (I-11b). A mixture of
methyl 2ꢀ(4ꢀbromoꢀ2ꢀchlorophenyl)acetate (I-6b, 792 mg, 3.0 mmol), 3ꢀmethylpyrazinꢀ2(1H)ꢀ
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
2
one (I-10b, 660 mg, 6.0 mmol), N ,N ꢀdimethylethaneꢀ1,2ꢀdiamine (105.6 mg, 1.2 mmol),
copper(I) iodide (114.6 mg, 0.60 mmol) and potassium phosphate (2.89 g, 9.0 mmol) in 1,4ꢀ
dioxane (10 mL) was stirred at 110 °C under nitrogen atmosphere for 2 d. After cooling to rt and
filtration, the filtrate was concentrated and purified by silica gel chromatography using
DCM/methanol (100:1) as eluting solvents to afford the title compound as yellow solid (560 mg,
64 % yield). LCMS (ESI): m/z = 293.1 [M+1].
6-(2-Chloro-4-(6-methylpyridin-2-yl)phenyl)-2-(methylthio)pyrido[2,3-d]pyrimidin-7(8H)-
one (I-12). To a 3L round bottom reactor was charged 4ꢀaminoꢀ2ꢀ(methylthio)pyrimidineꢀ5ꢀ
carbaldehyde (I-4a, 115.8 g, 684.5 mmol), ethyl 2ꢀ(2ꢀchloroꢀ4ꢀ(6ꢀmethylpyridinꢀ2ꢀyl)ꢀ
phenyl)acetate (I-9a, 238 g, 1.2 equiv) and K CO (283.8 g, 3 equiv) followed DMF (600 mL).
2
3
0
The mixture was stirred at 110 C for 21 h under a nitrogen atmosphere and progression
monitored by HPLC. After completion, DMF was partially removed in vacuo. HCl (12 M,
1
70 mL) and H O (1 L) were added to the mixture, and the reaction mixture was filtered. The
2
cake was reslurried in sat. aq. NaHCO (1 L) and water (2 L), respectively and filtered. The cake
3
1
was dried at 50°C in vacuum to give the title compound as brown solid (260 g, 96.0% yield). H
NMR (400 MHz, DMSOꢀd6): δ 8.90 (s, 1H), 8.24 (d, 1H, J = 1.6 Hz), 8.10 (t, 2H, J = 4.8 Hz),
ACS Paragon Plus Environment
23

Journal of Medicinal Chemistry
Page 24 of 83
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
7
.99 (s, 1H), 7.821ꢀ7.89 (m, 3H), 7.53 (d, 1H, J = 8 Hz), 7.29 (d, 1H, J = 7.6 Hz), 2.51 (s, 3H),
2
.38 (s, 3H).
tert-Butyl 3-(6-(2-chloro-4-(6-methylpyridin-2-yl)phenyl)-2-(methylthio)-7-oxopyrido[2,3-
d]pyrimidin-8(7H)-yl)propylcarbamate (I-13). A mixture 6ꢀ(2ꢀchloroꢀ4ꢀ(6ꢀmethylpyridinꢀ2ꢀ
yl)phenyl)ꢀ2ꢀ(methylthio)pyrido[2,3ꢀd]pyrimidinꢀ7(8H)ꢀone (I-12, 193 mg, 0.49 mmol), cesium
carbonate (478 mg, 1.47 mmol) and 3ꢀ(tertꢀbutoxycarbonylamino)propyl 4ꢀmethylbenzeneꢀ
sulfonate (200 mg, 0.61 mmol) in 1,4ꢀdioxane (5 mL) was stirred at 85 °C overnight. After
filtration and concentration, the residue was purified by silica gel chromatography using
DCM/MeOH (40:1 to 20:1) as eluting solvents to afford the title compound as a brown solid
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
+
(300 mg, 90% yield). LCMS (ESI): m/z = 496.0 [Mꢀ55] .
tert-Butyl 3-(6-(2-chloro-4-(6-methylpyridin-2-yl)phenyl)-2-(methylsulfinyl)-7-oxopyrido-
[2,3-d]pyrimidin-8(7H)-yl)propylcarbamate (I-14). To a solution of tertꢀbutyl 3ꢀ(6ꢀ(2ꢀchloroꢀ
4ꢀ(6ꢀmethylpyridinꢀ2ꢀyl)phenyl)ꢀ2ꢀ(methylthio)ꢀ7ꢀoxopyrido[2,3ꢀd]pyrimidinꢀ8(7H)ꢀyl)propylꢀ
carbamate (300 mg, 0.54 mmol) in dry DCM (10 mL) was added in portions mꢀCPBA (164 mg,
0
.80 mmol, 85% wt). The reaction mixture was stirred at rt for 30 min. Water (50 mL) was added
and the solution was extracted with DCM (2×50 mL). The combined organic layers were dried
over anhydrous sodium sulfate and concentrated under reduced pressure to afford the title
compound as a white solid (280 mg, crude), which was used in the next step without further
+
purification. LCMS (ESI): m/z = 512.0 [Mꢀ55] .
tert-Butyl 3-(6-(2-chloro-4-(6-methylpyridin-2-yl)phenyl)-2-(methylamino)-7-oxopyrido[2,3-
d]pyrimidin-8(7H)-yl)propylcarbamate (I-15). A mixture of 6ꢀ(2ꢀchloroꢀ4ꢀ(3ꢀmethylꢀ2ꢀ
oxopyridinꢀ1(2H)ꢀyl)phenyl)ꢀ8ꢀethylꢀ2ꢀ(methylsulfinyl)pyrido[2,3ꢀd]pyrimidinꢀ7(8H)ꢀone (280
mg, 0.49 mmol) and methyl amine (30% wt in EtOH, 5 mL) in THF (20 mL) was stirred at 50°C
ACS Paragon Plus Environment
24

Page 25 of 83
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
overnight. After concentration under reduced pressure, the residue was purified by preparative
HPLC to afford the title compound as a yellow oil (150 mg, crude), which was used in the next
+
step without further purification. LCMS (ESI): m/z = 480.1 [Mꢀ55] .
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
6
-(4-Bromo-2-chlorophenyl)-2-(methylthio)pyrido[2,3-d]pyrimidin-7(8H)-one (I-16). To a
2 L round bottom flask equipped with a mechanical stirrer under N atmosphere was charged 4ꢀ
2
aminoꢀ2ꢀmethylsulfanylꢀpyrimidineꢀ5ꢀcarbaldehyde (80.00 g, 1.0 equiv), potassium carbonate
(196 g, 3.0 equiv), DMF (5 mL/g, 400 mL), and methyl 2ꢀ(4ꢀbromoꢀ2ꢀchloroꢀphenyl)acetate (1.0
equiv, 131 g). The mixture was heated to 120 °C. The reaction was complete in 8 h based on
HPLC analysis and then cooled to ambient temperature. Water (2400 mL, 30 vol) was added,
and the mixture was stirred for 30 min. The mixture was stirred for an additional 1 h, filtered and
the wet cake was rinsed with water (150 mL × 2). The solids were recrystallized in THF/heptane
and dried at 50 °C under vacuum to afford the product as a light orange solid (151 g, 83% yield).
1
H NMR (400 MHz, DMSOꢀd6): δ 12.70 (s, 1H), 8.90 (s, 1H), 7.98 (s, 1H), 7.85 (s, 1H), 7.65 (d,
1
H, J = 8 Hz), 7.38 (d, 1H, J =8 Hz).
tert-Butyl (3-(6-(4-bromo-2-chlorophenyl)-2-(methylthio)7-oxopyrido[2,3-d]pyrimidin-
8(7H)-yl)propyl)carbamate (I-17). To a 1L 3ꢀneck roundꢀbottom flask was added 6ꢀ(4ꢀbromoꢀ
ꢀchlorophenyl)ꢀ2(methylthio)pyrido[2,3ꢀd]pyrimidinꢀ7(8H)ꢀone (I-16; 26.0 g, 67.9 mmol),
2
cesium carbonate powder (66.4 g, 3.00 equiv), tertꢀbutyl Nꢀ(3ꢀbromopropyl)carbamate (24.3 g,
1.50 equiv) and DMF (390 mL). The mixture was stirred at rt for 17 h and quenched with H O
2
(390 mL). The mixture was then extracted with DCM (300 mL) and phases were separated. The
aqueous layer was extracted with DCM (100 mL) and the combined organic layers were washed
with H O (100 mL × 3) and brine (100 mL). The organic layer was filtered through a 2ꢀinch
2
silica gel pad and washed with DCM (100 mL × 2), 2.5% MeOH in DCM (100 mL × 2) and 5%
ACS Paragon Plus Environment
25

Journal of Medicinal Chemistry
Page 26 of 83
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
MeOH in DCM (100 mL). The filtrate was washed with H O (200 mL × 2) followed by H O
2
2
(
300 mL × 2) and dried over anhydrous Na SO . The solution was concentrated to dryness to
2 4
afford the title compound as an orange oil (51.0 g, 74% yield). LCMS (ESI): m/z = 441.0 [Mꢀ
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
+ 1
Boc+H] ; H NMR (300 MHz, CDCl ): δ 8.67 (s, 1H), 7.67 (d, J = 1.9 Hz, 1H), 7.65 (s, 1H),
3
7.49 (dd, J = 8.2, 1.9 Hz, 1H), 7.23 (d, J = 8.2 Hz, 1H), 5.26 (s, 1H), 4.56 (t, J = 6.4 Hz, 2H),
3.13 (q, J = 6.3 Hz, 2H), 2.65 (s, 3H), 2.03–1.92 (m, 3H), 1.43 (s, 9H).
tert-Butyl 3-(6-(2-chloro-4-(2-oxopyrrolidin-1-yl)phenyl)-2-(methylthio)-7-oxopyrido[2,3-
d]pyrimidin-8(7H)-yl)propylcarbamate (I-19a). A mixture of tertꢀbutyl 3ꢀ(6ꢀ(4ꢀbromoꢀ2ꢀ
chlorophenyl)ꢀ2ꢀ(methylthio)ꢀ7ꢀoxopyrido[2,3ꢀd]pyrimidinꢀ8(7H)ꢀyl)propylcarbamate (I-17, 590
mg, 1.10 mmol), pyrrolidinꢀ2ꢀone (187.2 mg, 2.2 mmol), tris(dibenzylideneacetone)diꢀ
palladium(0) (100.7 mg, 0.11 mmol), dimethylbisdiphenylphosphinoxanthene (127.1 mg,
0.22 mmol) and cesium carbonate (1.08 g, 3.3 mmol) in toluene (10 mL) was stirred in a sealed
at 100 °C overnight. After filtration and concentration under reduced pressure, the residue was
purified by silica gel chromatography using DCM/MeOH (20:1) as eluting solvents to afford the
+
title compound (600 mg, 99% yield) as a yellow solid. LCMS (ESI): m/z = 488 [Mꢀ55] .
tert-Butyl 3-(6-(2-chloro-4-(2-oxopyridin-1(2H)-yl)phenyl)-2-(methylthio)-7-oxopyrido[2,3-
d]pyrimidin-8(7H)-yl)propylcarbamate (I-19b). A mixture of tertꢀbutyl 3ꢀ(6ꢀ(4ꢀbromoꢀ2ꢀ
chlorophenyl)ꢀ2ꢀ(methylthio)ꢀ7ꢀoxopyrido[2,3ꢀd]pyrimidinꢀ8(7H)ꢀyl)propylcarbamate
(I-17,
212 mg, 0.40 mmol), pyridinꢀ2(1H)ꢀone (I-18b, 76 mg, 0.80 mmol), cuprous iodide (14.8 mg,
1
2
0
.08 mmol), N ,N ꢀdimethylethaneꢀ1,2ꢀdiamine (14.3 mg, 0.16 mmol) and potassium phosphate
(254 mg, 1.2 mmol) in dioxane (2 mL) was stirred in a sealed tube at 110 °C overnight. After
filtration and concentration under reduced pressure, the residue was purified by silica gel
chromatography using DCM: MeOH (20:1) as eluting solvents to afford the title compound as a
ACS Paragon Plus Environment
26

Page 27 of 83
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
+
yellow solid (157 mg, 71% yield). LCMS (ESI): m/z = 570.1 [M+17] .
tert-Butyl 3-(6-(2-chloro-4-(2-oxopyrrolidin-1-yl)phenyl)-2-(methylsulfinyl)-7-oxopyrido-
2,3-d]pyrimidin-8(7H)-yl)propylcarbamate (I-20a). A mixture of tertꢀbutyl 3ꢀ(6ꢀ(2ꢀchloroꢀ4ꢀ
2ꢀoxopyrrolidinꢀ1ꢀyl)phenyl)ꢀ2ꢀ(methylthio)ꢀ7ꢀoxopyrido[2,3ꢀd]pyrimidinꢀ8(7H)ꢀyl)propylꢀ
[
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
(
carbamate (I-19a, 200 mg, 0.37 mmol) and mꢀCPBA (112 mg, 0.56 mmol, 85% wt) in DCM (10
mL) was stirred at r.t. overnight. Saturated sodium bicarbonate and DCM (40 mL) were added.
The separated organic layer was washed with brine and dried over sodium sulfate. After
concentration under reduced pressure, tertꢀbutyl 3ꢀ(6ꢀ(2ꢀchloroꢀ4ꢀ(2ꢀoxopyrrolidinꢀ1ꢀyl)phenyl)ꢀ
2ꢀ(methylsulfinyl)ꢀ7ꢀoxopyrido[2,3ꢀd]pyrimidinꢀ8(7H)ꢀyl)propylcarbamate (250 mg, crude) was
obtained as a white solid, which was used in the next step without further purification. LCMS
+
(ESI): m/z = 505.1 [Mꢀ55] .
tert-Butyl
3-(6-(2-chloro-4-(2-oxopyrrolidin-1-yl)phenyl)-2-(methylsulfinyl)-7-oxo-
pyrido[2,3-d]pyrimidin-8(7H)-yl)propylcarbamate (I-20b). A mixture of tertꢀbutyl 3ꢀ(6ꢀ(2ꢀ
chloroꢀ4ꢀ(2ꢀoxopyrrolidinꢀ1ꢀyl)phenyl)ꢀ2ꢀ(methylthio)ꢀ7ꢀoxopyrido[2,3ꢀd]pyrimidinꢀ8(7H)ꢀyl)ꢀ
propylcarbamate (I-19b, 80 mg, 0.14 mmol) and mꢀCPBA (42 mg, 0.21 mmol, 85% wt) in DCM
(5 mL) was stirred at r.t. overnight. Saturated sodium bicarbonate and DCM (20 mL) was added.
The separated organic layer was washed with brine and dried over sodium sulfate. After
concentration under reduced pressure, the title compound (100 mg, crude) was obtained as a
white solid, which was used in the next step without further purification. LCMS (ESI): m/z =
+
5
15.1 [Mꢀ55] .
tert-Butyl
3-(6-(2-chloro-4-(2-oxopyrrolidin-1-yl)phenyl)-2-(methylsulfinyl)-7-oxopyrido-
[2,3-d]pyrimidin-8(7H)-yl)propylcarbamate (I-21a). A mixture of tertꢀbutyl 3ꢀ(6ꢀ(2ꢀchloroꢀ4ꢀ
(2ꢀoxopyrrolidinꢀ1ꢀyl)phenyl)ꢀ2ꢀ(methylsulfinyl)ꢀ7ꢀoxopyrido[2,3ꢀd]pyrimidinꢀ8(7H)ꢀ
ACS Paragon Plus Environment
27

Journal of Medicinal Chemistry
Page 28 of 83
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
yl)propylcarbamate (I-20a, 250 mg, crude) and methyl amine (30% wt in EtOH, 5 mL) in MeOH
at was stirred overnight at r.t. after concentration under reduced pressure, the title compound
(150 mg, crude) was obtained as yellow oil, which was used in the next step without further
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
+
purification. LCMS (ESI): m/z = 472.0 [Mꢀ55] .
tert-Butyl 3-(6-(2-chloro-4-(2-oxopyridin-1(2H)-yl)phenyl)-2-(methylamino)-7-oxopyrido-
[2,3-d]pyrimidin-8(7H)-yl)propylcarbamate (I-21b). A mixture of tertꢀbutyl 3ꢀ(6ꢀ(2ꢀchloroꢀ4ꢀ
(2ꢀoxopyrrolidinꢀ1ꢀyl)phenyl)ꢀ2ꢀ(methylsulfinyl)ꢀ7ꢀoxopyrido[2,3ꢀd]pyrimidinꢀ8(7H)ꢀ
yl)propylcarbamate (I-20b, 100 mg, crude) and methyl amine (30% wt in EtOH, 2 mL) in MeOH
at was stirred overnight at r.t. After concentration under reduced pressure, it was afforded the title
compound (110 mg, crude) as yellow oil, which was used in the next step without further
+
purification. LCMS (ESI): m/z = 482.0 [Mꢀ55] .
tert-butyl (3-(6-(4-bromo-2-chlorophenyl)-2-(methylsulfinyl)-7-oxopyrido[2,3-d]pyrimidin-
8(7H)-yl)propyl)carbamate (I-22). To a 1L 3ꢀneck roundꢀbottom flask was added tertꢀbutyl Nꢀ
[
3ꢀ[6ꢀ(4ꢀbromoꢀ2ꢀchlorophenyl)ꢀ2ꢀmethylsulfanylꢀ7ꢀoxoꢀpyrido[2,3ꢀd]pyrimidinꢀ8ꢀyl]propyl]ꢀ
carbamate (I-17, 23.4 g, 74 wt%, 32.1 mmol) and CH CN (350 mL). The mixture was cooled to
3
0−5 °C and mꢀCPBA (9.62 g, 70 wt%, 1.20 equiv) was added. The mixture was stirred for 30
min and quenched with a mixture of satd. Na SO (75 mL) and H O (75 mL). The organic layer
2
3
2
was separated and washed with a mixture of satd. Na SO (50 mL) and H O (50 mL), then dried
2
3
2
over anhydrous MgSO (10.0 g) and filtered. The filtrate was concentrated to ca. 275 mL and
4
used directly in the next step.
6-(2-chloro-4-(2-oxopyrrolidin-1-yl)phenyl)-2-(methylamino)pyrido[2,3-d]pyrimidin7(8H)-
one (I-23). To the above CH CN solution of crude I-22 was added MeNH (12.1 mL, 33 wt% in
3
2
EtOH, 3.0 equiv) at 0−10 °C and the mixture was stirred at 10−20 °C for 30 min. The mixture
ACS Paragon Plus Environment
28

Page 29 of 83
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
was then concentrated to a red oil, diluted with DCM (50 mL) and passed through a 2ꢀinch silica
gel pad eluting with DCM (100 mL × 2) followed by 2.5% MeOH in DCM (100 mL x5). The
combined organic fractions were concentrated to dryness and triturated in heptane/EtOAc (50:1).
The solids were filtered and dried under vacuum at 50 °C to afford the title compound as 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
yellow solid (16.0 g, 94% yield). LCMS (ESI): m/z = 424.0 [MꢀBoc+H] ; H NMR (300 MHz,
CDCl ) δ 8.47 (br s, 1H), 7.65 (d, J = 1.9 Hz, 1H), 7.53 (s, 1H), 7.46 (dd, J = 8.2, 2.0 Hz, 1H),
3
7
.23 (d, J = 8.2 Hz, 1H), 5.56 (br s, 1H), 4.51 (br s, 2H), 3.18−3.08 (m, 4H), 1.98 (br s, 3H), 1.44
(s, 9H).
tert-Butyl (3-(6-(2-chloro-4-(3-methyl-2-oxoimidazolidin-1-yl)phenyl)-2-(methylamino)-7-
oxopyrido[2,3-d]pyrimidin-8(7H)-yl)propyl)carbamate (I-24a). A suspension of tertꢀbutyl Nꢀ
[3ꢀ[6ꢀ(4ꢀbromoꢀ2ꢀchloroꢀphenyl)ꢀ2ꢀ(methylamino)ꢀ7ꢀoxoꢀpyrido[2,3ꢀd]pyrimidinꢀ8ꢀyl]propyl]ꢀ
carbamate (I-23, 108 mg, 207 µmol), 1ꢀmethylimidazolidinꢀ2ꢀone (43.1 mg, 430 µmol, 2.1
equiv), tris(dibenzylideneacetone)ꢀdipalladium(0) (17.4 mg, 18.8 µmol, 0.09 equiv), 4,5ꢀbis(diꢀ
phenylphosphino)ꢀ9,9ꢀdimethylxanthene (18.1 mg, 31.3 µmol, 0.15 equiv) and cesium carbonate
(
204 mg, 627 µmol, 3.0 equiv) in degassed toluene (4.0 mL) was heated to 100 °C for 6 h and
allowed to cool. The mixture was diluted with EtOAc, washed with water and brine, dried over
sodium sulfate, filtered and concentrated in vacuo, and used directly in the next step.
6
-(4-Bromo-2-chlorophenyl)-8-ethyl-2-(methylthio)pyrido[2,3-d]pyrimidin-7(8H)-one
(I-
25). To a 2 L roundꢀbottom flask was added 4ꢀ(ethylamino)ꢀ2ꢀ(methylthio)pyrimidineꢀ5ꢀ
carbaldehyde (62.0 g, 0.314 mmol), potassium carbonate (3.0 equiv, 130 g), DMF (400 mL), and
ethyl 2ꢀ(4ꢀbromoꢀ2ꢀchloroꢀphenyl) acetate (I-6b, 1.0 equiv, 87 g). The mixture was heated to
1
20 °C. The reaction was completed in 2 h based on HPLC analysis and was cooled to ambient
temperature. Water (2400 mL, 30 vol) was added and the mixture was stirred for 30 min, solid
ACS Paragon Plus Environment
29