Structure-based Design of Pyridone-Aminal eFT508 Targeting Dysregulated Translation by Selective Mitogen-activated Protein Kinase Interacting Kinases 1 and 2 (MNK1/2) Inhibition

Article abstract of DOI:10.1021/acs.jmedchem.7b01795

Dysregulated translation of mRNA plays a major role in tumorigenesis. Mitogen-activated protein kinase interacting kinases (MNK)1/2 are key regulators of mRNA translation integrating signals from oncogenic and immune signaling pathways through phosphoryla

Full text of DOI:10.1021/acs.jmedchem.7b01795

Subscriber access provided by - Access paid by the | UCSB Libraries
Structure-based Design of Pyridone-aminal eFT508 Targeting
Dysregulated Translation by Selective Mitogen-activated
Protein Kinase Interacting Kinases 1 and 2 (MNK1/2) Inhibition
Siegfried H. Reich, Paul A. Sprengeler, Gary G. Chiang, Jim R. Appleman, Joan Chen, Jeff Clarine,
Boreth Eam, Justin T. Ernst, Qing Han, Vikas K. Goel, Edward ZR Han, Vera Huang, Ivy NJ Hung,
Adrianna Jemison, Katti A. Jessen, Jolene Molter, Douglas Murphy, Melissa Neal, Gregory S.
Parker, Michael Shaghafi, Samuel Sperry, Jocelyn Staunton, Craig R. Stumpf, Peggy A. Thompson,
Chinh Tran, Stephen E. Webber, Christopher J. Wegerski, Hong Zheng, and Kevin R. Webster
J. Med. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.jmedchem.7b01795 • Publication Date (Web): 10 Mar 2018
Downloaded from http://pubs.acs.org on March 11, 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 83
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Title
Structure-based Design of Pyridone-aminal eFT508 Targeting Dysregulated Translation by
Selective Mitogen-activated Protein Kinase Interacting Kinases 1 and 2 (MNK1/2)
Inhibition
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Authors
*,a
a
a
b
a
Siegfried H. Reich, Paul A. Sprengeler , Gary G. Chiang , Jim R. Appleman , Joan Chen , Jeff
a
a
a
c
a
c
Clarine , Boreth Eam , Justin T. Ernst , Qing Han , Vikas K Goel , Edward ZR Han ,
d
a
e
f
a
Vera Huang , Ivy NJ Hung , Adrianna Jemison , Katti A. Jessen , Jolene Molter , Douglas
g
a
a
h
a
Murphy , Melissa Neal , Gregory S. Parker , Michael Shaghafi , Samuel Sperry ,
a
a
a
a
i
Jocelyn Staunton , Craig R. Stumpf , Peggy A. Thompson , Chinh Tran , Stephen E. Webber ,
a
c
a
Christopher J. Wegerski , Hong Zheng , and Kevin R. Webster
Author Affiliation
a
b
eFFECTOR Therapeutics, 11180 Roselle St., San Diego, CA 92121; JRA Consulting;
c
d
StructureꢀBased Design Inc., 6048 Cornerstone Ct W # D, San Diego, CA 92121; Molecular
e
Stethoscope, 10835 Rd to the Cure #100, San Diego, CA 92121; Department of Chemistry,
f
University of Pennsylvania, 231 South 34th Street, Philadelphia, PA 19104ꢀ632; Oncternal
g
Therapeutics, 3525 Del Mar Heights Rd #821, San Diego, CA 92130; Molcentrics Inc., 11835
h
Carmel Mtn. Rd. #1304ꢀ110, San Diego, CA 92128; Abide Therapeutics, 10835 Road to the
i
Cure, Suite 250, San Diego, CA 92121; Polaris Pharmaceuticals, 9373 Towne Centre Dr # 150,
San Diego, CA 92121
Email: sreich@effector.com
ACS Paragon Plus Environment

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
Keywords
StructureꢀBased Design, MNK, Inhibitor, Translation, eFT508, Lymphoma, Cancer
Abstract
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
Dysregulated translation of mRNA plays a major role in tumorigenesis. MNK1/2 kinases are key
regulators of mRNA translation integrating signals from oncogenic and immune signaling
pathways through phosphorylation of eIF4E and other mRNA binding proteins. Modulation of
these key effector proteins regulates mRNA which control tumor/stromal cell signaling.
Compound 23 (eFT508), an exquisitely selective, potent dual MNK1/2 inhibitor, was designed to
assess the potential for control of oncogene signaling at the level of mRNA translation. The
crystal structureꢀguided design leverages stereoelectronic interactions unique to MNK
culminating in a novel pyridoneꢀaminal structure described for the first time in the kinase
literature. Compound 23 has potent in vivo antiꢀtumor activity in models of diffuse large cell Bꢀ
cell lymphoma (DLBCL) and solid tumors suggesting that controlling dysregulated translation
has real therapeutic potential. Compound 23 is currently being evaluated in Phase 2 clinical trials
in solid tumors and lymphoma. Compound 23 is the first highly selective dual MNK inhibitor
targeting dysregulated translation being assessed clinically.
Introduction
Translation, the most energy consuming process in the cell, plays a significant role in gene
regulation and ultimately the control of protein levels. The precept that there is a direct
correspondence between the abundance of mRNA and that of its corresponding protein product
is an oversimplification. Translation is a tightly controlled process for a select set of mRNAs and
dysregulation of this process drives aberrant proliferation, angiogenesis, survival, and alterations
ACS Paragon Plus Environment

Page 3 of 83
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
1ꢀ5
in immune function, all hallmarks of cancer (Figure 1). The key step of capꢀdependent
translation initiation relies upon the availability and activity of eukaryotic initiation factor eIF4E,
which is in turn regulated by the mitogenꢀactivated protein kinase interacting kinases
6
ꢀ11
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
(
MNK)1/2. The MNKs are activated by upstream Ras/Raf/Erk and MyD88/p38 signaling
pathways, resulting in regulation of RNA translation through eIF4E and other key effector
proteins such as heterogeneous nuclear ribonucleoprotein A1 (hnRNPA1) and proteinꢀassociated
1
, 2, 4, 12ꢀ16
splicing factor (PSF).
Phosphorylation of these mRNA binding proteins selectively
regulates the translation and stability of a subset of cellular mRNA that control both tumor and
stromal cell signaling. MNK, a Ser/Thr kinase, is the only kinase known to phosphorylate eIF4E
at serine 209. This modification has been shown to be essential for eIF4E’s role in tumorigenesis
but not for normal development and cell homeostasis. MNK1/2 double knockout studies in mice
1
7ꢀ19
further demonstrated that these kinases are not required for normal growth and development.
Thus, the underlying genetics of MNK lead to an expectation that a potent and selective MNK
inhibitor would not necessarily have a strong antiꢀproliferative phenotype but would limit
cellular processes necessary for oncogene driven signaling and survival, sparing normal tissues
and offering the potential for a good therapeutic window. This distinction between transformed
versus normal cells makes the MNKs particularly exciting as a therapeutic opportunity.
Interestingly the literature has examples of less selective MNK inhibitors which do have a broad
2
0ꢀ24
antiꢀproliferative phenotype which may be a result of offꢀtarget effects unrelated to MNK.
2
5ꢀ28
Other MNK inhibitors have demonstrated improved selectivity.
While the biology of MNK
and its effects on tumorigenesis has been the subject of many studies, the breadth of these
effects, particularly through pharmacological intervention, has not been demonstrated for a
ACS Paragon Plus Environment

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
potent, selective dual MNK inhibitor to date. As a consequence, the full potential for control of
1
1
mRNA translation via MNK in cancer therapy has not been revealed.
As a result we sought to identify a dual MNK1/2 inhibitor with both exceptional kinome
selectivity and cellular potency so that the pharmacologic phenotype of MNK1/2 inhibition could
be fully explored. We have designed a highly selective and potent MNK1/2 inhibitor, compound
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
3 (eFT508), leveraging the unique active site of this kinase in an iterative structureꢀbased
2
9
30,
approach. The pyridoneꢀaminal chemotype has not been reported in kinase inhibitors to date.
3
1
Compound 23 activity in models of DLBCL is associated with potent pꢀeIF4E knockdown and
selective destabilization of proꢀinflammatory cytokine mRNA. Proꢀinflammatory cytokines,
including interleukinꢀ6 (ILꢀ6), interleukinꢀ8 (ILꢀ8) and TNFα are drivers of many hallmarks of
cancer, e.g., tumor cell survival, migration and invasion, angiogenesis, immune evasion, and
32, 33
stress response, while affecting drug resistance.
Importantly, this pꢀeIF4E and cytokine
knockdown translates into potent in vivo antiꢀtumor activity in DLBCL models harboring
1
3, 32
activating MyD88 mutations, consistent with the MNKs being activated by TLR signaling.
In addition, efficacy has been demonstrated in solid tumor settings which may reflect the MNKs
impact on mRNA translation that controls tumor and stromal cell signaling. This breadth of
antitumor activity highlights the significant potential for control of dysregulated translation by a
3
3
small molecule inhibitor. Compound 23 is currently in Phase 2 clinical trials for the treatment
of both solid and hematologic cancers and has been granted Orphan Drug status by the FDA for
the treatment of DLBCL.
Results
ACS Paragon Plus Environment

Page 5 of 83
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Molecular Design Strategy Leading to the Pyridone-aminal Chemotype. A common
approach to selective kinase inhibition is to begin with a molecule which contains a potent hinge
binding motif and whose shape complements the targeted kinase. Additional groups are then
added that are tolerated by the kinase of interest but make negative interactions with the antiꢀ
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
34
targets. Our approach to MNK inhibitor design focused on incorporating favorable
stereoelectronic interactions with the MNK active site residues that differentiate it from other
kinases. Stereoelectronic interactions are highly sensitive to distance and orientation which can
35ꢀ37
improve selectivity.
These interactions are often less affected by solvation which can
enhance potency and permeability. There are several atypical residues in the ATP binding site of
MNK1/2, two of which offered the greatest opportunity for stereoelectronic interactions, the
38
gatekeeper, Phe159, and the preꢀDFG residue, Cys225, and these were the focus of our design.
Across the kinome, only Flt3 and 4, cyclin dependent kinaseꢀlike (CDKL)1ꢀ5, PRP4
(serine/threonineꢀprotein kinaseꢀPRP4), a dual specificity tyrosineꢀphosphorylationꢀregulated
kinase family kinase, mitogenꢀactivated protein kinaseꢀ7, and MNK have this combination
suggesting that interacting with these residues in a ligand efficient manner could facilitate the
design of MNKꢀselective inhibitors (Figure 2A). A unique sequence element in the MNK1/2
kinases is a DFD motif in place of the canonical DFG, however, these residues are more distant
39
from the hinge. The active sites of MNK1 and MNK2 are nearly identical (90%), so there was
2
5
an expectation that achieving similar affinity for both isoforms should be feasible. Dual
inhibition was thought to be necessary as the isoforms can compensate for one another in the
4
phosphorylation of eIF4E.
Our early strategy was to identify a chemical starting point with high ligand efficiency.
Time spent on identifying the best starting scaffold and then maintaining its attributes can greatly
ACS Paragon Plus Environment

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
facilitate medicinal chemistry success later; this is the underlying principle of fragmentꢀbased
drug discovery wherein a priority is placed on ligand efficiency and physicochemical properties
at the outset of design. We identified six initial fragment or fragmentꢀlike scaffolds (15ꢀ20 heavy
atoms, Figure 2AꢀF). Compounds 1, 2, 3 and 6 were in the public domain and 4 and 5 were
designed using common kinase inhibitor hinge binding motifs modeled into the MNK2 crystal
structure. We utilized a mutated MNK protein for crystallization as it has been shown that
mutation of the WT MNK from DFD to the more canonical DFG results in a more readily
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
38
crystallized protein. All had good affinity for MNK as measured by their MNK1 IC and
5
0
ligand efficiency. We attempted to obtain crystal structures of each compound in complex with
mutated MNK2 to understand their potential for making protein interactions and facilitate the
design of a cohort of initial follow on molecules; structures were obtained for four of the six.
Figure 2A shows the coꢀcrystal structure of carboxamide 1 in the MNK2 active site and
highlights the electrostatic potential of the gatekeeper residue Phe159 and the cone representing
the nucleophilic directionality of the sulfur of the preꢀDFG Cys225. The negative potential of
the Phe159 face suggested that ligand moieties having positive potential (i.e., the edge of an
aromatic ring, a methyl group, or a polarizable atom, etc.) would be complimentary. This
structure indicated potential interactions were possible with both key residues, Phe159 and
Cys225, for compound 1 optimization. It also revealed that the Nꢀ9 nitrogen of the imidazole
ring was engaged with the hinge through a 3.5 Å hydrogen bond to the Met162 carbonyl oxygen,
a weak interaction. The binding mode of compound 2 (Figure 2B) also showed good potential for
interactions with Phe159 and Cys225 (compound 2 shares the imidazo[1,2ꢀb]pyridazine core of
2
7
BAY1143269). Compound 3, on the other hand, did not make intimate interactions with either
the gatekeeper or the hinge and its suitability for engaging Cys225 was deemed low. Compound
ACS Paragon Plus Environment

Page 7 of 83
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
4
crystalized in two conformations (only the pose most representative of the series shown) and,
similar to 3, the ability of compound 4 to interact with the gatekeeper and Cys225 was
considered to be marginal. The crystal structures for 5 and 6 could not be determined,
nevertheless, a small cohort of compounds was designed using all six fragments as starting
points and the potency trajectory of each scaffold was assessed (Figure 3). The carboxamide
scaffold 1 emerged as the preferred structure, having one of the highest ligand efficiencies and
good vectors allowing rapid optimization to low nanomolar binding to MNK with less than 20
compounds. The purine was deconstructed to pyrimidine 7 resulting in a 10ꢀfold loss in potency;
however, the molecular weight and logP were reduced, and ligand efficiency was maintained
(Table 1). A loss of potency in exchange for improved physicochemical properties is an often
overlooked yet powerful optimization strategy in medicinal chemistry. Lactam compound 8 was
designed through analysis of the cocrystal structure of compound 1 and was 2.5ꢀfold more active.
This lactam was expected to have improved permeability due to loss of one hydrogen bond
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
3
donor. Furthermore, introduction of the lactam sp carbon afforded vectors out of the aryl plane.
We thought substitution of this buried methylene might force the bicyclic ring system to flip 180
degrees accommodating the increased size of this substitution and accessing the pꢀloop, a rich
source of kinase binding affinity. Compound 9 was 25ꢀfold more potent than the original
carboxamide 7 and the coꢀcrystal structure showed that the ring flip had indeed occurred relative
to its unsubstituted counterpart compound 8 (Figure 4). Linkers between the pyrimidine and
benzene rings other than nitrogen (i.e., C, O, or S) and replacement of the pyrimidine with
pyridine consistently resulted in less active molecules likely due to reduced planarity.
As described previously, the sulfur of Cys225 can make favorable stereoelectronic
3
7
interactions affording the potential to both increase potency and selectivity (Figure 2A).
ACS Paragon Plus Environment

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
Heterocycles in general can interact positively with cysteines and several tested were effective
binders. A single attempt at covalent interaction with Cys225 with a α,βꢀunsaturated amide was
unsuccessful suggesting that Cys225 is not particularly reactive. Stereoelectronic interactions of
atoms having free lone pairs of electrons with carbonyl groups in a manner similar to the Dunitz
attack for amide hydrolysis are observed quite often in crystal structures both in the PDB and
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
4
0, 41
CSD databases.
Highly polarizable chlorine and sulfur atoms are especially capable of
engaging in this type of interaction but it is also observed for fluorine, oxygen, and nitrogen.
The pyridone ring system of 11 was designed to provide a stereoelectronic (Dunitz) interaction
with Cys225 (Figure 5A). While 11 was 4ꢀfold less active in terms of binding affinity,
importantly, its clogP was almost two orders of magnitude lower (XlogP = ꢀ0.99, Dotmatics)
than its benzene counterpart 7 as reflected in the 16ꢀfold improvement in lipophilic ligand
1
.2
42
efficiency (LLE) (Table 1, 6.5 – 5.3 = 1.2, 10 = 16). This modest reduction in binding
affinity commensurate with a 72ꢀfold reduction in lipophilicity was a critical result in the overall
optimization process. The gemꢀdimethyl pyridone, compound 12, was equipotent with its
benzene counterpart 10, maintaining the 17ꢀfold improvement in LLE. The cocrystal structure of
1
2 confirmed that atom distances and angles between the sulfur atom and carbonyl were
consistent with a Dunitz interaction (Figure 5B). Comparison of the kinome profiles of 10 and
1
2 (>400 kinases) showed that the pyridone imparts selectivity over the benzene ring (Figure
5
C). This trend for greater selectivity was observed across additional pyridoneꢀbenzene pairs.
Another advantage of the pyridone scaffold was its considerably more facile synthesis
and exceptional chemical stability. The pyridone aminal could be readily substituted by simply
heating the pyridone carboxamide with a ketone in strong acid (Figure 6A and Scheme 5), a big
advantage over the generally low yielding alkylation strategy required for the benzolactam. As a
ACS Paragon Plus Environment

Page 9 of 83
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
consequence of this simple condensation, greater diversity could now be incorporated at the
aminal carbon as outlined in Figure 6A, facilitating the exploration of potential interactions with
the pꢀloop.
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Pyridone Cellular eIF4E-Ser209 Phosphorylation Potency. A focused protein structureꢀ
guided library was pursued and the properties of this library are depicted in Figure 6B. The
compounds are segregated by potency and substitution pattern and cover a spectrum of both
4
2
clogP (XlogP) and MW. Potential substituents off of the aminal carbon with MW < 200 were
modeled into the MNK crystal structure, taking into account that the pꢀloop is quite flexible.
Compounds that were exceptionally potent fell into the area of unfavorably high logP and high
molecular weight (shaded area, Figure 6B). Highly substituted spirocyclic compounds were
preferred, with 6ꢀmembered rings showing the best kinome selectivity. Chiral compounds at the
aminal carbon had no advantage in terms of potency or selectivity over achiral molecules so the
latter became the focus of further design. The potencies of the compounds were routinely at the
lower limit of the ATP competitive biochemical assay (ca. 1.5 nM) so the inhibition of eIF4E
phosphorylation in cells was used to track potency. Cellular potency was improved 30ꢀ300ꢀfold
(cf. 12 to 17) by incorporation of a cyclopropylamide in the 4ꢀposition of the pyrimidine 16
(Table 1 and 2). This moiety has led to improved potency in inhibitors targeting other kinases,
e.g. JAK2, TYK2, bRAF, Abl, VEGFR2, etc.
Antiꢀproliferative effects were correlated with lack of kinase selectivity. Compound 17,
having single digit nanomolar cell potency for eIF4E phosphorylation and good selectivity for
MNK1/2, showed no antiproliferative effects in three solid tumor cell lines (PC3, HCTꢀ116, SWꢀ
6
20). In contrast, compounds 18 and 19 with less MNK1/2 selectivity demonstrated antiꢀ
proliferative effects across multiple cell lines (Table 2). Compound 18 is an example from the
ACS Paragon Plus Environment

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
structureꢀguided library described above which is potent, however, it is chiral, higher MW and
has a higher logP. Potent cellular inhibition of pꢀeIF4E along with the absence of an antiꢀ
proliferative signal was used as a measure of onꢀtarget selectivity not usually possible in an
oncology drug discovery program. Compound 17, demonstrated target engagement over the
dosing period as measured by pꢀeIF4E reduction, and when dosed in vivo showed antiꢀtumor
activity.
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
As described earlier, the ring flip observed for substituted lactams and aminals created an
unfilled pocket adjacent to gatekeeper residue Phe159 which was our second residue of focus in
the initial analysis. We hypothesized that this space might accommodate a single heavy atom
positioned in the 5ꢀposition of the pyridone ring. A greater than 50ꢀfold improvement in potency
was observed with a 5ꢀchlorine followed closely by a 5ꢀmethyl whereas a 5ꢀfluorine had
significantly less benefit (compounds 15, 14 and 13 vs 12, Table 1 and Figure 8). The enhanced
affinity of the chlorine and methyl group is consistent with their ability to interact with the
negative potential of the Phe159 aryl face, as revealed by the electrostatic potential analysis
(Figure 2A). The methyl substituent provided the best balance of kinome selectivity and potency
enhancement overall.
Final Optimization and Identification of Compound 23. It was found that an unsubstituted 4ꢀ
amino pyrimidine was able to maintain the hydrogen bond to Met162 backbone carbonyl as was
observed with the cyclopropyl amide (cf. compound 20 vs 21) but importantly, the free amino
group abrogated metabolism observed for molecules with spirocarbocycles off the pyridone
aminal. While the cyclopropylamide improved potency, it was also slightly less selective, adding
five additional heavy atoms with an increased log P, therefore we focused on 4ꢀ
aminopyrimidines. A chlorine in the 5ꢀposition of the pyrimidine provided up to a log of
ACS Paragon Plus Environment

Page 11 of 83
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
additional affinity likely through a favorable interaction with the carbonyl of Leu90 (compound
2
2, 84 pM cell activity, Table 1).
Starting with and maintaining good overall physicochemical properties along with the
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
above potency enhancements provided a cohort of optimized molecules having low nanomolar
cellular pꢀeIF4E potency, in vivo efficacy, good metabolic stability and kinome selectivity. These
were narrowed further based on multispecies pharmacokinetics (mouse, rat, dog, and monkey;
Table 3). Compound 23 had the best overall profile and was consistently the top performer in
tumor growth inhibition (TGI) and pharmacokinetic/pharmacodynamic (PK/PD) studies and was
therefore selected as the development candidate (Figure 9). The crystal structure of compound
2
3 is shown in Figure 7A and confirms the pyridone’s involvement in a Dunitz interaction with
the Cys225 sulfur.
Compound 23 is a Potent Inhibitor of MNK1 and MNK2 Signaling and Tumor
Growth. We examined the effect of compound 23 on eIF4E phosphorylation in an expanded set
of tumor cell lines using the pꢀeIF4E Ser209 homogeneous time resolved fluorescence (HTRF)
assay. In all cell lines tested, compound 23 inhibited Ser209 phosphorylation of eIF4E with IC50
values ranging from 1.4 to 22 nM (Table S1). The effect of compound 23 on eIF4E Ser209
phosphorylation is specific to MNK1/2 inhibition, as overꢀexpression of wildꢀtype or a 23ꢀ
resistant allele of MNK2 (C225L) is sufficient to increase the 23 IC 100ꢀ to 860ꢀfold,
5
0
respectively (Figure S2A). Consistent with the HTRF results, potent doseꢀdependent inhibition
of eIF4E Ser209 phosphorylation was observed by immunoblot analysis (Figure 10A and Figure
S2B). In addition, compound 23 did not affect 4EꢀBP1 phosphorylation atThr37 or Thr46 or 4Eꢀ
BP1 expression. Importantly, phosphorylation of MAPK at Thr202/Tyr204 was also unaffected
by compound 23, demonstrating that 23 did not impact the activation status of signaling
ACS Paragon Plus Environment

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
pathways that lie upstream of MNK1/2. We next expanded our analysis to growth factor and
cytokine signaling pathways following compound 23 treatment. Treatment of TMD8 cells with
23 led to a doseꢀdependent reduction in secreted, ILꢀ6, ILꢀ8, and TNFα (Figure 10BꢀD).
Mechanistically, the decreased cytokine production arising from 23 treatment of TMD8 cells
corresponded with reduced mRNA stability (Figure 10E), which is consistent with previous
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
2ꢀ14
reports implicating MNK kinases in regulating the phosphorylation of RNA binding proteins.
Other small molecule inhibitors possessing activity against MNK1/2, such as
cercosporamide, CGP57380, merestinib, and cabozatinib, show antiꢀproliferative activity in cellꢀ
based assays, which could potentially be attributed to their broader kinome activity as we also
2
0, 21, 23, 24
observed similar effects with our lessꢀselective MNK1/2 compounds.
In contrast,
compound 23 did not show antiꢀproliferative activity in a panel of solid and hematological tumor
cell lines (IC > 30 ꢁM), although modest sensitivity (IC < 10 ꢁM) was observed in a subset of
5
0
50
DLBCL and a multiple myeloma cell line (Figure S3). This finding is not surprising given that
genetic studies demonstrate that MNK1/2 is dispensable for normal growth but required for
oncogeneꢀinduced transformation as assessed by anchorageꢀindependent growth in vitro or
tumorigenesis in vivo. The sensitivity observed in the DLBCL cell lines is consistent with
previous studies demonstrating a role for MNK kinases in integrating signals from TLRs to
4
3, 44
regulate proꢀinflammatory cytokines.
In particular, TMD8 cells harbor activating mutations
in myeloid differentiation primary response gene 88 and CD79 and exhibit constitutive TLR
pathway signaling, consistent with their sensitivity to 23. In addition, elevated levels of eIF4E
4
5
have been observed in DLBCL patient samples.
Compound 23 potently inhibited eIF4E phosphorylation, select mRNA stability and proꢀ
inflammatory cytokine production in DLBCL cells; we therefore examined the efficacy of 23 in
ACS Paragon Plus Environment

Page 13 of 83
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
a TMD8 xenograft model. Compound 23 was wellꢀtolerated at doses of 1 and 10 mg/kg QD as
measured by lack of body weight loss (Figure S4) which corresponds to a minimal therapeutic
index of >/= 10 in this model. Furthermore, 23 treatment produced significant TGI over a 10ꢀfold
dose range, achieving an average TGI of 71% and 75% when dosed orally at 1 and 10 mg/kg
QD, respectively (Figure 11A). We also observed similar activity in HBLꢀ1 xenografts, an
additional model of MyD88/CD79 mutant DLBCL (Figure 11B). Again, compound 23 was
wellꢀtolerated in the animals and treatment resulted in an average TGI of 66% and 96%, when
dosed at 1 or 10 mg/kg QD, respectively. This activity is consistent with the hypothesis that
MNK plays a significant role in mediating TLRꢀMyD88 signaling in DLBCL. Finally, we tested
compound 23 in the MDAꢀMBꢀ231 breast cancer xenograft model. Increased phosphorylation
of eIF4E has been shown in breast cancer patients and has been linked to poorer clinical
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
4
6
outcome. The growth of MDAꢀMBꢀ231 xenografts in vivo was significantly inhibited (TGI
>100%) when animals were dosed with 1 mg/kg or 10 mg/kg 23 QD (Figure 11C). Consistent
with the MNK phenotype previously observed with compound 23, MDAꢀMBꢀ231 cells did not
show growth inhibition in vitro, underscoring the fact that tumor cell/tumor microenvironment
(TME) interactions important for tumorigenesis may be regulated by MNK.
In conjunction with the efficacy experiments, we assessed MNK inhibition through
PK/PD measurement of pꢀeIF4E (Ser209) in the TMD8 xenografts at 3 dose levels (0.3, 1 and 10
mg/kg). In general, an oral dose of 1 mg/kg of 23 QD produced maximal efficacy and exhibited
over 80% reduction in pꢀeIF4E for 8 hours (Figure 11D). Plotting of the aggregated data from
the three dose groups to generate an exposureꢀresponse curve resulted in calculated IC and IC
5
0
90
values of 15.8 and 376 nM respectfully, which are consistent with the pꢀeIF4E inhibition values
ACS Paragon Plus Environment

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
obtained in vitro (Figure 11E, Table S1). Importantly, these results demonstrate that pꢀeIF4E
can be used as a PD marker for MNK inhibition in vivo and that ≥ 80% inhibition is associated
with efficacy.
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
Synthetic Chemistry
All compounds described herein were prepared as outlined in Schemes 1ꢀ6. Compound 4 was
obtained in the following manner (Scheme 1). Heating of 2ꢀaminonicotinonitrile 27 with aqueous
2
ꢀchloroacetaldehyde afforded imidazocarbonitrile 28. Subsequent bromination followed by
Suzuki coupling with pyridinylꢀ3ꢀylboronic acid yielded compound 30, which was hydrolyzed
with basic hydrogen peroxide to give final compound 4. Final compound 5 was obtained in three
steps by Boc protection of pyrrolidine 31, followed by Suzuki coupling with a Boc protected
pyrazole boronic ester to form the protected pyrazole which was bisꢀdeprotected to afford final
compound 5 (Scheme 2). Pyrimidines 7 and 8 were prepared using conventional acid catalyzed
SNAr chemistry, from 4,6ꢀdichloropyrimidine and 6ꢀchloroꢀ9Hꢀpurine 35, respectively (Scheme
3
). Monochloro pyrimidine 34 was dehalogenated via hydrogenolysis to produce final compound
. Final isoindolinone compounds 9 and 10 were prepared as outlined from the same
7
bromomethylbenzoate 37 (Scheme 4). Coupling of pꢀmethoxybenzylamine with 37 followed by
ring closure produced PMB protected lactam 38. Bisꢀmethylation of 38 by heating with excess
MeI in the presence of NaH, followed by Buchwald coupling with 4ꢀaminopyrimidine and PMB
deprotection gave final lactam compound 10. Monomethylation of 38 was affected by treating
with 1.5 equivalents of MeI at room temperature to give 39 which was subjected to the same two
step sequence of Buchwald coupling/PMB deprotection to yield final lactam compound 9.
Pyridone 11 was prepared beginning with esterification of 5ꢀchloroꢀ2ꢀpyridine carboxylic acid
40 followed by oxidation to the Nꢀoxide 41. Nꢀoxide 41 was converted to the pyridone with
ACS Paragon Plus Environment

Page 15 of 83
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
trifluoroacetic anhydride followed by benzylic protection to give compound 42. Buchwald
coupling of 42 with 4ꢀaminopyrimidine yielded 43 which was debenzylated with triflic acid
followed by aminolysis to afford final compound 11.
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Pyridones 12-15, 17,18, and 20-26 were all prepared using the general procedure outlined in
Scheme 5. The suitably substituted pyridone 44, prepared in a similar fashion outlined for 42,
was subjected to aminolysis to give the key intermediate carboxamide 45. Formation of the
lactam structure 46 was carried out by simply heating the respective carboxamide 45 with the
desired ketone (R R CO) in H SO /1,4ꢀdioxane to afford good yields of the desired cyclized
1
2
2
4
material. Buchwald coupling of the corresponding haloꢀlactams 46 with the required substituted
4ꢀamino pyrimidine (R4), yielded the final compounds in good yields after either basic or acidic
deprotection.
The remaining two compounds 16 and 19 were prepared as outlined in Scheme 6. 6ꢀChloroꢀ2ꢀ
aminopyrimidine 47 was bisꢀprotected with Bocꢀanhydride and then coupled with
cyclopropanecarboxamide under palladium catalysis to give amide 48. Boc deprotection with
acid followed by Buchwald coupling with PMB protected lactam 38 and final deprotection
yielded the final cyclopropylamide 16. Suzuki coupling of bromide 50 with isoquinolinꢀ6ꢀ
ylboronic acid gave compound 52. SNAr coupling of tꢀbutylꢀ(3ꢀhydroxycyclobutyl)carbamate
and 52 and sodium hydride yielded intermediate ether 53 which was deprotected with aqueous
HCl to provide final compound 19.
Discussion and Conclusions
Recruitment of mRNA to the ribosome is a fundamental component of mRNA translation
and is regulated through translation initiation by eIF4E. Importantly, eIF4E has been shown to
play a key role in the translation of proteins involved in driving tumorigenesis. While the MNKs
ACS Paragon Plus Environment

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
are key modulators of eIF4E activity, the eIF4E protein has been targeted directly with an
4
7
antisense oligonucleotide, ISIS 183750. In addition, small molecule inhibitors such as EGIꢀ1,
briciclib, and 4E1Rꢀcat that block the binding of eIF4E to eIF4G have been pursued, yet only
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
modest antiꢀtumor activity has been observed with these agents and little progress has been made
48ꢀ51
clinically.
In addition, a number of MNK small molecule inhibitors have been described that
2
0ꢀ23, 25, 26, 52
are either less potent or lack selectivity against MNK1/2.
BAY1143269, which is
2
7
more selective for MNK1, entered the clinic but enrollment is currently suspended. There may
be limitations with an MNK1 selective approach as MNK2 can function in a compensatory
5
3
fashion. A dual MNK inhibitor has been reported to be in Phase 1 clinical trial in Singapore. A
selective and potent dual MNK1/2 inhibitor represents a new therapeutic opportunity, yet to date,
compounds to definitively assess this potential have remained elusive.
We have outlined the design of a very potent and selective dual MNK1/2 inhibitor, 23,
having a novel pyridoneꢀaminal chemical structure described for the first time in the kinase
literature, and the first dual MNK inhibitor to be tested clinically in the treatment of both solid
and hematological malignancies. The design process leveraged specific stereoelectronic
interactions with the unique active site of this kinase in an iterative structureꢀbased approach
employing 30 coꢀcrystal structures. Potency was deliberately traded for drugꢀlike properties
(lower logP and MW) during optimization (Figure 8). Importantly, only 110 pyridone
containing compounds were required to identify 23, highlighting the efficient optimization
approach employed and the value of beginning with and maintaining good physicochemical
properties. A road map for the key molecules in the optimization process is outlined in Figure 8
highlighting strategic potency loss (1ꢀ3, red vertical lines) and key contenders for candidate
ACS Paragon Plus Environment

Page 17 of 83
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
selection. During the course of optimization cell potency closely tracked enzyme potency, LE
was maintained finishing slightly higher (0.6), and final LLEs improved to >8.
Compound 23 was designed to be equally effective against both MNK1 and MNK2
isoforms as they are generally coꢀexpressed and can both serve to phosphorylate eIF4E at
Ser209, implying that dual inhibition would be necessary to avoid compensatory signaling by
either isoform. As such, inhibition of eIF4E phosphorylation was always commensurate with the
activity against the less sensitive MNK isoform suggesting that dual MNK1/2 engagement was
critical for maximal reduction in pꢀeIF4E levels and robust efficacy. Importantly, 23 potently
inhibits pꢀeIF4E in a range of solid and hematological tumor cell lines. In contrast to other small
molecule inhibitors of MNK1/2, 23 does not cause inhibition of proliferation across a broad
panel of tumor cell lines. This biology might have been masked by a less selective kinome
profile and broad antiꢀproliferative effects and this understanding facilitated the design of 23.
Sufficient selectivity against the broader protein and lipid kinome is essential to understanding
MNK pharmacology. Only two protein kinases outside of MNK were significantly inhibited by
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
2
3 in biochemical assays (Figure 7B): DRAK1/STK17A, a member of the deathꢀassociated
protein family of serine/threonine kinases, has been shown to have a proꢀapoptotic role in certain
5
4ꢀ57
contexts, but a proꢀtumorigenic effect in others
; and CLK4, a member of the cdc2/cdc28ꢀlike
5
8, 59
kinase family, plays a role in alternative mRNA splicing and in cytokinesis.
It is unclear
whether 23 significantly inhibits DRAK1 and CLK4 in intact cells, and we do not see any
cellular phenotypes consistent with inhibition of either kinase in tumor cells.
DLBCL is currently defined by three main subtypes: Activated Bꢀcell (ABC), germinal
6
0
center Bꢀcell and primary mediastinal Bꢀcell lymphoma. Patients with ABCꢀDLBCL have the
worst prognosis with < 40% cure rate. ABCꢀDLBCL has been characterized by constitutive
ACS Paragon Plus Environment

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
activation of NFkB signaling due to oncogenic mutations that activate B cell receptor signaling
6
1
(
e.g. CD79, CARD11 and A20). In addition, activating mutations in MyD88 occur in ~39% of
ABCꢀDLBCL leading to NFkB activation that drives increased cytokine and chemokine
production. MNK1/2 are recognized regulators of cytokine signaling and production. Proꢀ
inflammatory cytokines are key survival factors for ABCꢀDLBCL tumors and cytokine
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
3
2, 33
expression is associated with poor prognosis.
Compound 23 is effective at blocking proꢀ
inflammatory cytokine production in MyD88 mutant models of ABCꢀDLBCL. Postꢀ
transcriptional regulation of cytokine production is known to be regulated at the level of mRNA
stability through adenylateꢀuridylateꢀrich elements in the 3’ꢀuntranslated regions of their
6
2
mRNA. Compound 23 treatment results in a 1.5ꢀ2 fold decrease in cytokine mRNA halfꢀlife
consistent with the observed reduction in secreted cytokine proteins. Further testing of 23 in vivo
demonstrated significant efficacy in multiple MyD88 mutant DLBCL tumor models illustrating
its potential as a novel therapeutic strategy for this disease (Figure 11). Compound 23 is well
tolerated at doses that deliver maximal efficacy and target engagement in vivo as measured by
inhibition of pꢀeIF4E, consistent with normal viability and development of MNK1/2 double
knockꢀout mice.
6
3
Tumorꢀpromoting inflammation is recognized as an enabling characteristic of cancer.
Proꢀinflammatory cytokines and chemokines are key mediators of this effect impacting both
tumor cell survival signaling and the composition and signaling of the TME. Compound 23
selectively regulates the production of proꢀinflammatory cytokines and chemokines and has the
potential to reꢀshape the TME. This mechanism of action is likely contributing to the antiꢀtumor
efficacy in both the DLBCL and MDAꢀMBꢀ231 models and is the focus of continuing
evaluation.
ACS Paragon Plus Environment

Page 19 of 83
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
This work has highlighted the strong potential for agents which modulate dysregulated
translation acting through eIF4E and also affecting mRNA stability. We have shown that potent
MNK inhibition has significant effects on multiple proꢀtumorigenic cytokines, such as TNFα, ILꢀ
6
4, 65
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
6
and ILꢀ8, which are in turn, important mediators of oncogenesis and tumor progression.
These profound effects, both intrinsic and extrinsic to the tumor, are particularly interesting
given that a potent and selective dual MNK inhibitor does not display broad antiꢀproliferative
activity in vitro, yet demonstrates potent antiꢀtumor activity in vivo against models utilizing the
same cell lines. These data are consistent with the phenotype of the MNK1/2 double knockout
mouse, which develops normally. This also highlights the potential limitation of screening for
antiꢀcancer activity in an in vitro antiꢀproliferative setting alone, particularly for agents that
might act via mechanisms extrinsic to the tumor. Based on the preclinical profile of 23 and
demonstrated in vivo activity in multiple tumor models, selective inhibition of MNK1 and
MNK2 has potential in the treatment of cancer. Compound 23 is currently under evaluation in
Phase 2 clinical trials in patients with advanced solid tumors and lymphoma.
Experimental Section
General. All reagents and solvents were used as purchased from commercial sources. Flash
column chromatography was performed with a Teledyne ISCO CombiFlash Rf system using
1
normalꢀphase silica columns (230−400 mesh). H NMR spectra were recorded on a Bruker
Advanceꢀ400 spectrometer at 400 MHz or a BrukerꢀBiospin AVANCE 500 MHz NMR
spectrometer. Coupling constants (J) are expressed in hertz (Hz). Chemical shifts (δ) of NMR are
reported in parts per million (ppm) units relative to an internal control (TMS). Microwave
reactions were performed with a Biotage Initiator focused beam microwave reactor (300 W).
HPLC purification was performed on a Waters automated purification system with 2767 sample
ACS Paragon Plus Environment

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
controller and 2545 binary pump using Mass Lynx software and a Waters Sunfire Cꢀ18 (19 X
2
50mm, 10 µm) / Waters XꢀSelect Phenyl Hexyl (19 X 250mm,5 µm) column . Analytical purity
was assessed on a Waters Acquity Ultra Performance UPLC with 3100 SQD equipped with
Acquity BEH Cꢀ18 (2.1 X 50mm, 1.7 µm ) column and all compounds tested were determined to
be >95% pure using this method. High resolution mass spectroscopy (HRMS) was performed
using a Triple TOF 5600+ mass spectrometer (hybrid quadrupole timeꢀofꢀflight platform; AB
Sciex) connected to a 1290 UHPLC system (Agilent). The mass spectrometer was operated in
electrospray positive ionization mode (ESI +). Acquisition was a full scan from m/z 100 to 1000
with a pulser frequency of 18.092 KHz and accumulation time of 75 ms. All animal studies were
carried out in accordance with the guidelines established by the Institutional Animal Care and
Use Committee at Explora BioLabs (San Diego, CA; Animal Care and Use Protocol (ACUP)
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
#
EB15ꢀ053).
4
-((7H-Purin-6-yl)amino)benzamide (1).
2
8
1
See Oyarzabal J, et al. Purchased from Ryan Scientific, Inc. Purity >95%; H NMR (400 MHz,
DMSOꢀd ) δ 13.30 (br s, 1 H), 10.07 (s, 1 H), 8.48 (s, 1 H), 8.36 (s, 1 H), 8.10 (d, J = 22 Hz, 1 H), 7.92ꢀ
6
7
.84 (m, 3 H), 7.30 (s, 1 H).
3
-(Pyridin-4-yl)imidazo[1,2-b]pyridazine (2).
28
1
See Oyarzabal J, et al. Purity 95.7%; H NMR (400 MHz, DMSOꢀd ) δ 8.76 (dd, J = 8, 3 Hz, 1 H),
6
8
.72 (d, J = 15 Hz, 2 H), 8.61 (s, 1 H), 8.34ꢀ8.26 (m, 3 H), 7.42 (dd, J = 23, 11 Hz, 1 H).
3
-Phenyl-5-(pyridin-4-yl)-1H-indazole (3).
6
6
See Reich et al. Purity >99%: The spectral data of the compound matched those in the
reference.
ACS Paragon Plus Environment

Page 21 of 83
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
3
-(Pyridin-4-yl)imidazo[1,2-a]pyridine-8-carboxamide (4).
67
Synthesized via the methodology described in Yang et al. (Scheme 1)
A mixture of 2ꢀaminonicotinonitrile 27 (2.5 g, 21 mmol) and 40% aqueous 2ꢀchloroacetaldehyde
(18.8 mL, 94.88 mmol, 4.5 eq.) in ethanol (100 mL) was refluxed overnight. The volatiles were
evaporated and the residue was treated with ethyl acetate (20 mL). The solid was collected by
filtration, washed with EtOAc (5 mL), and dried under vacuum to give the desired product as the
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
HCl salt. The salt was treated with 2N aqueous Na CO solution (20 mL), and extracted with
2
3
DCM (50 mL x2). The combined DCM solution was washed with brine, dried (Na SO ) and
2
4
concentrated to give imidazo[1,2ꢀa]pyridineꢀ8ꢀcarbonitrile 28 as a light brown solid (2.6 g, yield
1
8
6%). H NMR (400 MHz, DMSOꢀd ) δ 8.88 (d, J = 16.9 Hz, 1 H), 8.15 (s, 1 H), 7.96 (d, J = 17.6 Hz,
6
+
1 H), 7.73 (s, 1 H), 7.06 (t, J = 17.3 Hz, 1 H); MS (ESI) m/z 144.17 [M + H] .To a stirred solution of
2
ꢀmethylimidazo[1,2ꢀa]pyridineꢀ8ꢀcarbonitrile 28 (1.5 g, 9.5 mmol) in DMF (15 mL) at rt, was
added NBS (1.8 g, 10 mmol). The resulting mixture was stirred for 5 minutes and was diluted
with water (150 mL). The mixture was extracted with EtOAc (100 mL x 2). The combined
extracts were washed with brine, dried (Na SO ) and concentrated to give the title compound 29
2
4
1
as a white solid (1.73 g, yield 77%). The compound was pure enough for the next reaction. H
NMR (400 MHz, DMSOꢀd ) δ 8.69 (d, J = 17.2 Hz, 1 H), 8.09 (d, J = 17.7 Hz, 1 H), 7.91 (s, 1
6
+
H), 7.23 (t, J = 17.6 Hz, 1 H); MS (ESI) m/z 223.90 [M + H] . To a mixture of 3ꢀ
bromoimidazo[1,2ꢀa]pyridineꢀ8ꢀcarbonitrile 29 (0.5 g, 1 eq.), pyridinꢀ3ꢀylboronic acid (1.1 eq.)
and K CO (5 eq.) in 1,4ꢀdioxane (10 mL) and water (2.5 mL) under nitrogen atmosphere, was
2
3
added Pd(dppf)Cl (0.1 eq.). The resulting reaction mixture was heated at 100 °C overnight. The
2
reaction was cooled to rt and diluted with water (30 mL), and extracted with EtOAc (30 mL x 2).
The combined organic extracts were washed with brine, dried (Na SO ) and evaporated to give
2
4
ACS Paragon Plus Environment

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
the crude product, which was purified on a silica gel column to provide 3ꢀ(pyridinꢀ3ꢀ
1
yl)imidazo[1,2ꢀa]pyridineꢀ8ꢀcarbonitrile 30 in 40% yield. H NMR (400 MHz, DMSOꢀd ) δ
6
8.90 (d, J = 4.6 Hz, 1H), 8.88 (d, J = 17.4 Hz, 1H), 8.68 (dd, J = 11.7, 2.9 Hz, 1H), 8.16 (dt, J =
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
5.4, 4.2 Hz, 1H), 8.06 (d, J =17.6 Hz, 1H), 8.04 (s, 1H), 7.60 (dd, J = 19.6, 12.1 Hz, 1H), 7.12
+
(
t, J = 17.6 Hz, 1H); MS (ESI) m/z 221.06 [M + H] .To a solution of 3ꢀ(pyridinꢀ3ꢀ
yl)imidazo[1,2ꢀa]pyridineꢀ8ꢀcarbonitrile 30 (0.15 g) in DMSO (3 mL) cooled with an ice water
bath, was added 30% H O (1.2 mL) and anhydrous K CO (0.2 g). The reaction mixture was
2
2
2
3
allowed to warm to rt and stirred for 10 minutes. The reaction mixture was diluted with water (10
mL) and extracted with EtOAc. The ethyl acetate extracts were concentrated and the residue was
purified on a silica gel column to provide 3ꢀ(pyridinꢀ3ꢀyl)imidazo[1,2ꢀa]pyridineꢀ8ꢀcarboxamide
1
4
as a light yellow solid in 28% yield. HPLC purity: 96.25%; H NMR (400MHz, DMSOꢀd ), δ
6
9
.54 (br, 1H), 8.91 (d, J = 1.6 Hz, 1H), 8.78 (dd, J = 6.8, 0.8 Hz, 1H), 8.68 (dd, J = 4.8, 1.2 Hz,
1H), 8.17 (dt, J = 8.0, 2.0 Hz, 1H), 8.11 (dd, J = 7.2, 1.2 Hz, 1H), 8.05 (br, 1H), 8.01 (s, 1H),
1
3
7
1
.61 (dd, J = 7.6, 5.2 Hz, 1H), 7.17 (t, J = 7.2 Hz, 1H); C NMR (125 MHz, DMSOꢀd ) δ 163.6,
6
49.2, 148.7, 144.0, 135.3, 132.4, 128.2, 127.6, 124.6, 124.0, 123.0, 121.0, 112.8; mass
+
spectrometry atmospheric pressure ionization: m/z 239 [M + H] . HRMS: measured m/z [M +
+
H] 239.0927 (calcd. for C H N O: 239.0928).
13
11
4
3
,5-Dimethyl-4-(4-(pyrrolidin-2-yl)phenyl)-1H-pyrazole (5). To a solution of 2ꢀ(4ꢀ
bromophenyl)pyrrolidine 31 (0.5 g, 2.20 mmol) in 1,4ꢀdioxane (12.5 mL), 10% sodium
hydroxide solution (2 mL) was added at 0 °C and the mixture was stirred for 10 min (Scheme 2).
Diꢀtꢀbutyl pyrocarbonate (0.3 mL, 1.2 mmol) was added and the reaction mixture was allowed to
stir at rt for 3 h. The reaction mixture was diluted with water and extracted with ethyl acetate.
The organic layer was separated, washed with brine, dried over Na SO and concentrated under
2
4
ACS Paragon Plus Environment

Page 23 of 83
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
reduced pressure to obtain a residue which was purified by silica gel column chromatography to
1
afford tꢀbutyl 2ꢀ(4ꢀbromophenyl)pyrrolidineꢀ1ꢀcarboxylate 32. Yield: 0.62 g, 86%; H NMR
(400 MHz, Chloroformꢀd ) δ 7.41 (d, J = 7.92 Hz, 2 H) 7.05 (d, J = 7.92 Hz, 2 H) 4.67 ꢀ 4.79
1
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
(m, 1 H) 3.60ꢀ3.62 (m, 2 H) 2.29ꢀ2.32 (m, 1 H) 1.83 ꢀ 1.91 (m, 2 H) 1.75ꢀ1.78 (m, 1 H) 1.22 (s, 9
+
H); MS (ESI) m/z 333 [M + H] . A solution of tertꢀbutyl 2ꢀ(4ꢀbromophenyl)pyrrolidineꢀ1ꢀ
carboxylate 32 (0.3 g, 0.93 mmol), tertꢀbutylꢀ3,5ꢀdimethylꢀ4ꢀ(4,4,5,5ꢀtetramethylꢀ1,3,2ꢀ
dioxaborolanꢀ2ꢀyl)ꢀ1Hꢀpyrazoleꢀ1ꢀcarboxylate (0.303 g, 0.93 mmol) and 2M sodium carbonate
(0.3 g, 2.79 mmol) in 1,4ꢀdioxane (12 mL) was degassed with argon for 30 min.
Tetrakis(triphenylphosphine)palladium(0) (0.054 g, 0.046 mmol) was added and the reaction
mixture was further degassed for 15 min. The reaction mixture was heated at 100 °C for 16 h.
The reaction mixture was filtered through Celite and washed with ethyl acetate. The filtrate was
concentrated under reduced pressure and the residue was purified by silica gel column
chromatography to afford tꢀbutyl 4ꢀ(4ꢀ(1ꢀ(tertꢀbutoxycarbonyl)pyrrolidinꢀ2ꢀyl)phenyl)ꢀ3,5ꢀ
+
dimethylꢀ1Hꢀpyrazoleꢀ1ꢀcarboxylate. Yield: 0.1 g, 24%; MS (ESI) m/z 442 [M + H] .
To a stirred solution of tertꢀbutyl 4ꢀ(4ꢀ(1ꢀ(tertꢀbutoxycarbonyl)pyrrolidinꢀ2ꢀyl)phenyl)ꢀ3,5ꢀ
dimethylꢀ1Hꢀpyrazoleꢀ1ꢀcarboxylate (0.1 g, 0.22 mmol) in 1,4ꢀdioxane (8mL), 4M 1,4ꢀdioxane:
HCl (8 mL) was added at 0 °C and the reaction mixture was stirred at rt for 15 h. After complete
consumption of starting material, the solvent was removed under reduced pressure and the
residue which was neutralized with sodium carbonate resin. The reaction mixture was filtered
and concentrated under reduced pressure and purified by repeated washing with ether and
pentane to obtain 3,5ꢀdimethylꢀ4ꢀ(4ꢀ(pyrrolidinꢀ2ꢀyl)phenyl)ꢀ1Hꢀpyrazole 5. Yield: 0.03 g, 42%;
1
HPLC purity: 98.37%; H NMR (400 MHz, Methanolꢀd ) δ 7.52 (d, J = 7.9 Hz, 2H), 7.39 (d, J =
4
7
.9 Hz, 2H), 4.59 (dd, J = 9.5, 6.5 Hz, 1H), 3.49 – 3.29 (m, 2H), 2.56 – 2.41 (m, 1H), 2.32 – 2.12
ACS Paragon Plus Environment

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
13
(
m, 9H); C NMR (125 MHz, DMSOꢀd ) δ 142.8, 131.9, 128.2, 128.1, 126.5, 125.5, 116.7, 61.5,
6
+
+
5
4.3, 46.3, 34.5, 25.2, 11.3. MS (ESI) m/z 242 [M + H] HRMS: measured m/z [M + H]
242.1651 (calcd. for C H N : 242.1654).
1
5
20
3
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
4
-(3-(Piperidin-4-yl)-1H-pyrazol-5-yl)pyridine (6).
28, 68
Prepared as reported by Bilodeau M. T., et al.
Purity 83%; The spectral data of the
compound matched those in the reference.
4
-(Pyrimidin-4-ylamino)benzamide (7).To a stirred suspension of 4,6ꢀdichloropyrimidine (2 g,
13.51 mmol) and 4ꢀaminobenzamide 33 (1.83 g, 13.51 mmol) in tolueneꢀ1,4ꢀdioxane (20 mL,
:1), pꢀtoluenesulfonic acid (pTSA) (2.82 g, 14.86 mmol) was added and the reaction mixture
1
was heated at 140 °C for 5 h in a sealed tube (Scheme 3). The reaction mixture was neutralized
with saturated aqueous NaHCO solution and the compound was extracted with 10% methanolꢀ
3
DCM. The combined organic layer was dried over anhydrous Na SO , filtered and concentrated
2
4
under reduced pressure. The resulting solid was washed with ethyl acetate to afford 4ꢀ((6ꢀ
1
chloropyrimidinꢀ4ꢀyl)amino)benzamide 34 as a white solid. Yield: 0.68 g, 22.7%; H NMR (400
MHz, DMSOꢀd ) δ 10.09 (s, 1 H) 8.55 (s, 1 H) 7.85ꢀ7.88 (m, 3 H) 7.72 (d, J = 8.47 Hz, 2 H)
6
+
7
.24 (br s, 1 H) 6.88 (s, 1 H); MS (ESI) m/z 249 [M + H] . To a stirred suspension of 4ꢀ((6ꢀ
chloropyrimidinꢀ4ꢀyl)amino)benzamide 34 (0.3 g, 1.21 mmol) in 1,4ꢀdioxane:DCM:methanol
(10 mL, 2:1:1) was added 10% PdꢀC and the reaction mixture was hydrogenated under 1 atm
pressure at rt for 2 h. The reaction mixture was filtered through a Celite pad and the pad was
washed with methanol. The filtrate was concentrated under reduced pressure, taken up in a
minimum amount of methanol and refluxed to give a clear solution which was cooled to rt. The
obtained solid was filtered and dried to afford 4ꢀ(pyrimidinꢀ4ꢀylamino)benzamide (7) as a white
ACS Paragon Plus Environment

Page 25 of 83
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
1
solid. Yield: 0.039 g, 15%; HPLC purity: 97.50%; H NMR (400 MHz, DMSOꢀd ) δ 11.55 (s,
6
1
H), 8.97 (s, 1H), 8.41 (dd, J = 7.0, 1.4 Hz, 1H), 7.99ꢀ7.90 (m, 3H), 7.79 (d, J = 8.5 Hz, 2H),
1
3
7.35 (s, 1H), 7.18 (d, J = 7.0 Hz, 1H); C NMR (125 MHz, DMSOꢀd ) δ 167.0, 161.5, 152.6,
6
+
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
44.3, 139.8, 130.6, 128.4, 120.9, 107.6. MS (ESI) m/z 215.09 [M + H] ; HRMS: measured m/z
+
[
M + H] 215.0927 (calcd, for C H N O: 215.0926).
1
1
11
4
5
-((9H-purin-6-yl)amino)isoindolin-1-one (8)
6
9
Synthesized as reported in U.S. Patent Application No. WO2017075394.
A mixture of 6ꢀchloroꢀ9Hꢀpurine 35 (0.15 g, 0.97 mmol), 5ꢀaminoisoindolineꢀ1ꢀone 36 (0.14 g,
.97 mmol) and (1S)ꢀ(+)ꢀcamphorꢀ10ꢀsulfonic acid (0.27 g, 1.16 mmol) in isopropanol (10 mL)
0
was heated in a sealed tube at 100 °C for 4 h. After completion of the reaction, the mixture was
concentrated and the obtained solid was filtered and reꢀcrystallized from ethanol and isopropanol
to afford 5ꢀ((9Hꢀpurinꢀ6ꢀyl)amino)isoindolinꢀ1ꢀone (8) as offꢀwhite solid. Yield: 0.15 g, 58%;
1
HPLC purity: 99.92%; H NMR (400 MHz, DMSOꢀd ) δ 11.01 (s, 1H), 8.66 (s, 2H), 8.44 (s,
6
1
H), 8.31 (d, J = 1.9 Hz, 1H), 7.99 (dd, J = 8.3, 1.8 Hz, 1H), 7.67 (d, J = 8.3 Hz, 1H), 4.40 (s,
1
3
2H); C NMR (125 MHz, DMSOꢀd ) δ 169.5, 149.8, 149.7, 149.5, 141.2, 141.4, 127.9, 123.3,
6
+
+
1
2
20.1, 114.9, 114.3, 44.8. MS (ESI) m/z 267 [M + H] . HRMS: measured m/z [M + H]
67.0989 (calcd. for C H N O: 267.0989).
1
3
11
6
3
-Methyl-5-(pyrimidin-4-ylamino)isoindolin-1-one (9)
6
9
Synthesized as reported in U.S. Patent Application No. WO2017075394. To a stirred solution
of ethyl 4ꢀbromoꢀ2ꢀ(bromomethyl)benzoate 37 (16 g, 52.1 mmol) in dimethylformamide (150
mL) at 0 °C, triethylamine (22 mL, 156.3 mmol) was added and stirred for 15 min and 4ꢀ
methoxybenzylamine (8.84 mL, 67.7 mmol) was added and the reaction mixture was stirred at rt
ACS Paragon Plus Environment

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
for 30 min and then heated at 65 °C for 5 h (Scheme 4). The reaction mixture was cooled,
quenched with ice and acidified with 1N hydrochloric acid to pH = 3. The precipitated solid was
filtered, dried and purified by silica gel column chromatography using 30% ethyl acetate in
1
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
hexane to afford 5ꢀbromoꢀ2ꢀ(4ꢀmethoxybenzyl)isoindolinꢀ1ꢀone 38. Yield: 13.04 g, 75.4%; H
NMR (400 MHz, Chloroformꢀd ) δ 7.74 (d, J = 8.0 Hz, 1H), 7.60 (d, J = 8.4 Hz, 1H), 7.53 (s,
1
1
H), 7.22 (d, J = 8.4 Hz, 2H), 6.86 (d, J = 8.4 Hz, 2H), 4.72 (s, 2H), 4.22 (s, 2H), 3.79 (s, 3H)MS
+
(
ESI) m/z 332[M + H] . To an ice cooled suspension of sodium hydride (0.072 g, 60% dispersion
in mineral oil, 1.8 mmol) in dry THF (5 mL) a solution of 5ꢀbromoꢀ2ꢀ(4ꢀ
methoxybenzyl)isoindolinꢀ1ꢀone 38 (0.5 g, 1.5 mmol) in dry THF (5 mL) was added dropwise
under nitrogen atmosphere. The reaction mixture was stirred for 1 h and iodomethane (0.14 mL,
2
.2 mmol) was added. The reaction mixture was stirred at rt for 2 h and then quenched with iceꢀ
water and extracted with ethyl acetate. The combined organic layer was dried over anhydrous
sodium sulfate, filtered and concentrated under reduced pressure. The residue was purified by
silica gel column chromatography using 0ꢀ8% of ethyl acetate in hexane to afford 5ꢀbromoꢀ2ꢀ(4ꢀ
1
methoxybenzyl)ꢀ3ꢀmethylisoindolinꢀ1ꢀone 39. Yield: 0.29 g, 57%; H NMR (400 MHz, DMSOꢀ
d6) δ 7.88 (s, 1H), 7.68 (m, 2H), 7.22 (d, J = 8.4 Hz, 2H), 6.89 (d, J = 8.8 Hz, 2H), 4.94 (d, J =
1
5.6 Hz, 1H), 4.38 (m, 1H), 4.31 (d, J = 15.2 Hz, 1H), 3.72 (s, 3H), 1.38 (d, J = 6.8 Hz, 3H); MS
+
(
ESI) m/z 246, 248[M + H] . A mixture of 5ꢀbromoꢀ2ꢀ(4ꢀmethoxybenzyl)ꢀ3ꢀmethylisoindolinꢀ1ꢀ
one 39 (0.2 g, 0.578 mmol), 4ꢀaminopyrimidine (0.066 g, 0.693 mmol), sodium tertiary butoxide
(0.111 g, 1.15 mmol) and XꢀPhos (0.027 g, 0.057 mmol) in toluene (2 mL) was degassed with
argon for 30 min. Pd (dba) (0.053 g, 0.057 mmol) was added under argon atmosphere and
2
3
degassing was continued for another 10 min. The reaction mixture was heated at 110 °C for 4 h.
After completion of the reaction (monitored by TLC), it was diluted with ethyl acetate and
ACS Paragon Plus Environment

Page 27 of 83
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
filtered through a Celite pad and the filtrate was concentrated under reduced pressure. The
residue was purified by silica gel column chromatography to afford 2ꢀ(4ꢀmethoxybenzyl)ꢀ3ꢀ
1
methylꢀ5ꢀ(pyrimidinꢀ4ꢀylamino)isoindolinꢀ1ꢀone. Yield: 0.12 g, 58%; H NMR (400 MHz,
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Chloroformꢀd ) δ 8.73 ꢀ 8.79 (m, 1 H) 8.34 (d, J = 6.10 Hz, 1 H) 7.84 (d, J = 8.28 Hz, 1 H) 7.70
1
(s, 1 H) 7.35 ꢀ 7.43 (m, 2 H) 7.22 (d, J = 8.72 Hz, 2 H) 6.84 (d, J = 8.72 Hz, 2 H) 6.76 (dd, J =
6
.10, 1.31 Hz, 1 H) 5.30 (s, 2H) 4.38 (q, J = 6.54 Hz, 1 H) 3.78 (s, 3 H) 1.45 (d, J = 6.54 Hz,
+
3
H); MS (ESI) m/z 361[M + H] . A solution of 2ꢀ(4ꢀmethoxybenzyl)ꢀ3ꢀmethylꢀ5ꢀ(pyrimidinꢀ4ꢀ
ylamino)isoindolinꢀ1ꢀone (0.12 g, 0.332 mmol) in trifluoroacetic acid (2 mL) was heated at 100
C for 18 h. After completion of reaction (monitored by TLC), the mixture was concentrated
°
under reduced pressure. The residue was dissolved in methanol and carbonate supported polymer
was added and stirred for 1 h. The mixture was filtered and the filtrate was concentrated under
reduced pressure. The crude material was purified by prep. HPLC to afford compound 9 as an
1
offꢀwhite solid; Yield: 0.015 g, 19%; HPLC purity: 96.68%; H NMR (400 MHz, Methanolꢀd ) δ
4
8
4
.67 (s, 1H), 8.27 (d, J = 6.1 Hz, 1H), 8.13 (s, 1H), 7.74ꢀ7.61 (m, 2H), 6.86 (d, J = 6.0 Hz, 1H),
13
.69 (q, J = 6.8 Hz, 1H), 1.48 (d, J = 6.7 Hz, 3H), 1.15 (d, J = 6.2 Hz, 1H); C NMR (125 MHz,
DMSOꢀd ) δ 168.7, 159.6, 157.9, 155.5, 150.4, 142.9, 125.6, 123.3, 119.0, 112.6, 107.8, 51.4,
6
+
+
2
0.4. MS (ESI) m/z 241[M + H] ; HRMS: measured m/z [M + H] 241.1083 (calcd. for
C H N O: 241.1085).
13
13
4
3
,3-Dimethyl-5-(pyrimidin-4-ylamino)isoindolin-1-one (10).
6
9
Synthesized as reported in U.S. Patent Application No. WO2017075394. To a solution of 5ꢀ
bromoꢀ2ꢀ(4ꢀmethoxybenzyl)isoindolinꢀ1ꢀone 38 (1 g, 3 mmol) in THF (10 mL) at 0 °C, sodium
hydride (0.3 g, 7.5 mmol) was added portion wise and the reaction mixture was allowed to stir at
rt for 30 min. Methyl iodide (0.57 mL, 9 mmol) was added and the reaction mixture was stirred
ACS Paragon Plus Environment

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
at 70 °C for 16 h. The reaction mixture was quenched with water and extracted with ethyl
acetate. The organic layer was separated, dried over sodium sulfate and concentrated under
reduced pressure. The residue was purified by silica gel column chromatography using 30%
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
ethyl acetate in hexane to afford 5ꢀbromoꢀ2ꢀ(4ꢀmethoxybenzyl)ꢀ3,3ꢀdimethylisoindolinꢀ1ꢀone.
1
Yield: 0.4 g, 36.9%; H NMR (400 MHz, DMSOꢀd ) δ 8.05 (s, 1H) 7.64ꢀ7.67 (m, 2H) 7.36 (d, J
6
=
8.72 Hz, 2 H) 6.84 (d, J = 8.72 Hz, 2 H) 4.62 (s, 2 H) 3.78 (s, 3 H) 1.38 (s, 6H); MS (ESI) m/z
+
3
60[M + H] .To a solution of 2ꢀ(4ꢀmethoxybenzyl)ꢀ3,3ꢀdimethylꢀ5ꢀ(pyrimidinꢀ4ꢀ
ylamino)isoindolinꢀ1ꢀone (0.3 g, 0.8 mmol) in toluene (6 mL), 4ꢀaminopyrimidine (0.095 g, 0.99
mmol) and then sodium tꢀbutoxide (0.16 g, 1.66 mmol) were added and the reaction mixture was
degassed with argon for 15 min. Pd (dba) (0.076 g, 0.08 mmol) and Xꢀphos (0.039 g, 0.08
2
3
mmol) were added and the reaction mixture was heated at 110 °C for 16 h. The reaction mixture
was diluted with water and extracted with ethyl acetate. The organic layer was separated, dried
over sodium sulfate, and concentrated under reduced pressure. The residue was purified by silica
gel chromatography using 5% methanol in DCM to afford 2ꢀ(4ꢀmethoxybenzyl)ꢀ3,3ꢀdimethylꢀ5ꢀ
1
(
pyrimidinꢀ4ꢀylamino)isoindolinꢀ1ꢀone. Yield: 0.2 g, 65%; H NMR (400 MHz, DMSOꢀd ) δ
6
9
.98 (brs, 1H) 8.72 (s, 1H) 8.32ꢀ8.38 (m, 1H) 7.98 (s, 1H) 7.64ꢀ7.76 (m, 2H) 7.36 (d, J = 8.72
+
Hz, 2 H) 6.83ꢀ6.87 (m, 3 H) 4.62 (s, 2 H) 3.78 (s, 3 H) 1.38 (s, 6H); MS (ESI) m/z 375[M + H] .
A solution of 2ꢀ(4ꢀmethoxybenzyl)ꢀ3,3ꢀdimethylꢀ5ꢀ(pyrimidinꢀ4ꢀylamino)isoindolinꢀ1ꢀone (0.26
g, 0.69 mmol) in trifluoroacetic acid (8 mL), was heated at 95 °C for 16 h. The reaction mixture
was concentrated under reduced pressure and the residue was basified with a saturated solution
of sodium bicarbonate. The compound was extracted with 10% methanol in DCM. The organic
layer was separated, dried over sodium sulfate, concentrated under reduced pressure and the
residue was purified by silica gel column chromatography with 5% methanol in DCM to afford
ACS Paragon Plus Environment

Page 29 of 83
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
1
compound 10. Yield: 0.038 g, 22%; HPLC purity: 97.84%; H NMR (400 MHz, DMSOꢀd ) δ
6
9
.89 (s, 1H), 8.66 (s, 1H), 8.42 (s, 1H), 8.30 (d, J = 5.9 Hz, 1H), 7.87 (d, J = 1.9 Hz, 1H), 7.69
(dd, J = 8.3, 1.9 Hz, 1H), 7.52 (d, J = 8.3 Hz, 1H), 6.83 (dd, J = 5.9, 1.3 Hz, 1H), 1.40 (s, 6H);
1
3
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
C NMR (125 MHz, DMSOꢀd ) δ 167.5, 159.6, 157.9, 155.5, 154.4, 143.0, 124.8, 123.5, 119.0,
6
+
+
1
11.3, 107.9, 57.8, 27.6; MS (ESI) m/z 255[M + H] . HRMS: measured m/z [M + H] 255.1240
(
calcd. for C H N O: 255.1240).
14 15
4
6-Oxo-5-(pyrimidin-4-ylamino)-1,6-dihydropyridine-2-carboxamide. (11)To a solution of 5ꢀ
chloroꢀ2ꢀpyridinecarboxylic acid 40 (5 g, 31.74 mmol) in ethanol (50 mL), conc. H SO (0.5 ml)
2
4
was added and the mixture was heated at 80 °C for 12 h. The reaction mixture was cooled to rt
and concentrated under reduced pressure. The residue was dissolved in ethyl acetate and washed
with saturated aqueous NaHCO solution and brine, dried over anhydrous Na SO , filtered and
3
2
4
concentrated under reduced pressure to afford ethyl 5ꢀchloropicolinate as white crystalline solid.
1
Yield: 5 g, 86%; H NMR (400 MHz, DMSOꢀd ) δ 8.69 (m, 1H), 8.08 (m, 1H), 7.81 (m,1H),
6
+
4.47 (m,2H), 1.43 (t, J = 7.2 Hz, 3H); MS (ESI) m/z 186 [M + H] . To a solution of ethyl 5ꢀ
chloropicolinate (5 g, 27 mmol) in chloroform (60 mL), urea hydrogen peroxide (5.08 g, 54.05
mmol) was added under ice cooling. Trifluoroacetic anhydride (7.5 mL, 54.0 mmol in CHCl3
(60 mL) was added dropwise over 30 min and the mixture was stirred for 2 h at rt. A saturated
aqueous sodium thiosulfate solution was added dropwise to the reaction mixture followed by
extraction with chloroform and the organic layer was dried over anhydrous Na SO , filtered and
2
4
concentrated under reduced pressure. The residue was purified by silica gel column
chromatography using 0ꢀ50% ethyl acetate in hexanes to afford 5ꢀchloroꢀ2ꢀ
1
(
ethoxycarbonyl)pyridine 1ꢀoxide 41. Yield: 3.5 g, 64.8%; H NMR (400 MHz, DMSOꢀd ) δ
6
+
8
.65 (s,1H),7.76 (m,1H), 7.58 (m,1H), 4.33 (m,2H), 1.29 (m,3H); MS (ESI) m/z 202 [M + H] .
ACS Paragon Plus Environment