Fragment-Based Discovery of a Potent, Orally Bioavailable Inhibitor That Modulates the Phosphorylation and Catalytic Activity of ERK1/2

Article abstract of DOI:10.1021/acs.jmedchem.8b00421

Aberrant activation of the MAPK pathway drives cell proliferation in multiple cancers. Inhibitors of BRAF and MEK kinases are approved for the treatment of BRAF mutant melanoma, but resistance frequently emerges, often mediated by increased signaling through ERK1/2. Here, we describe the fragment-based generation of ERK1/2 inhibitors that block catalytic phosphorylation of downstream substrates such as RSK but also modulate phosphorylation of ERK1/2 by MEK without directly inhibiting MEK. X-ray crystallographic and biophysical fragment screening followed by structure-guided optimization and growth from the hinge into a pocket proximal to the C-α helix afforded highly potent ERK1/2 inhibitors with excellent kinome selectivity. In BRAF mutant cells, the lead compound suppresses pRSK and pERK levels and inhibits proliferation at low nanomolar concentrations. The lead exhibits tumor regression upon oral dosing in BRAF mutant xenograft models, providing a promising basis for further optimization toward clinical pERK1/2 modulating ERK1/2 inhibitors.

Full text of DOI:10.1021/acs.jmedchem.8b00421

Subscriber access provided by University of Winnipeg Library
Article
Fragment-based discovery of a potent, orally bioavailable inhibitor
which modulates the phosphorylation and catalytic activity of ERK1/2
Tom D. Heightman, Valerio Berdini, Hannah Braithwaite, Ildiko Maria Buck, Megan Cassidy, Juan Castro,
Aurelie Courtin, James E. H. Day, Charlotte East, Lynsey Fazal, Brent Graham, Charlotte M. Griffiths-
Jones, John F Lyons, Vanessa Martins, Sandra Muench, Joanne M Munck, David Norton, Marc O'Reilly,
Nick Palmer, Puja Pathuri, Michael Reader, David C. Rees, Sharna J. Rich, Caroline Richardson,
Harpreet Saini, Neil T Thompson, Nicola G Wallis, Hugh Walton, Nicola E. Wilsher, Alison J.-A. Woolford,
Michael Cooke, David Cousin, Stuart Onions, Jonathan Shannon, John Watts, and Christopher W Murray
J. Med. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.jmedchem.8b00421 • Publication Date (Web): 18 May 2018
Downloaded from http://pubs.acs.org on May 18, 2018
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a service to the research community to expedite the dissemination
of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in
full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully
peer reviewed, but should not be considered the official version of record. They are citable by the
Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore,
the “Just Accepted” Web site may not include all articles that will be published in the journal. After
a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web
site and published as an ASAP article. Note that technical editing may introduce minor changes
to the manuscript text and/or graphics which could affect content, and all legal disclaimers and
ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or
consequences arising from the use of information contained in these “Just Accepted” manuscripts.
is published by the American Chemical Society. 1155 Sixteenth Street N.W.,
Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.

Page 1 of 42
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Cousin, David; Sygnature Discovery Ltd
Onions, Stuart; Sygnature Discovery Ltd
Shannon, Jonathan; Sygnature Discovery Ltd
Watts, John; Sygnature Discovery Ltd
Murray, Christopher; Astex Pharamceuticals,
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
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 2 of 42
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
Fragment-based discovery of a potent, orally bioavailable inhibitor which
modulates the phosphorylation and catalytic activity of ERK1/2
1
,
1
1
1
1
Tom D. Heightman, * Valerio Berdini, Hannah Braithwaite, Ildiko M. Buck, Megan Cassidy, Juan
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
1
1
1
1
1
Castro , Aurélie Courtin, James E.H. Day, Charlotte East, Lynsey Fazal, Brent Graham, Charlotte
1
1
1
1
1
M. GriffithsꢀJones, John F. Lyons, Vanessa Martins, Sandra Muench, Joanne M. Munck, David
1
1
1
1
1
1
Norton, Marc O'Reilly, Nick Palmer, Puja Pathuri, Michael Reader, David C. Rees, Sharna J.
1
1
1
1
1
1
Rich, Caroline Richardson, Harpreet Saini, Neil T. Thompson, Nicola G. Wallis, Hugh Walton,
1
1
2
2
2
Nicola E. Wilsher, Alison J.ꢀA. Woolford, Michael Cooke, David Cousin, Stuart Onions,
2
2
1
Jonathan Shannon, John Watts, and Christopher W. Murray
1
Astex Pharmaceuticals, 436 Cambridge Science Park, Cambridge, CB4 0QA, UK
Sygnature Discovery Ltd, BioCity, Pennyfoot Street, Nottingham, NG1 1GF, UK
Corresponding author
2
*
1
ACS Paragon Plus Environment

Page 3 of 42
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
ABSTRACT
Aberrant activation of the MAPK pathway drives cell proliferation in multiple cancers. Inhibitors of
BRAF and MEK kinases are approved for the treatment of BRAF mutant melanoma, but resistance
frequently emerges, often mediated by increased signalling through ERK1/2. Here we describe the
fragment based generation of ERK1/2 inhibitors which block catalytic phosphorylation of
downstream substrates such as RSK but also modulate phosphorylation of ERK1/2 by MEK without
directly inhibiting MEK. Xꢀray crystallographic and biophysical fragment screening followed by
structureꢀguided optimization and growth from the hinge into a pocket proximal to the Cꢀα helix
afforded highly potent ERK1/2 inhibitors with excellent kinome selectivity. In BRAF mutant cells the
lead compound suppresses pRSK and pERK levels and inhibits proliferation at low nanomolar
concentrations. The lead exhibits tumor regression upon oral dosing in BRAF mutant xenograft
models, providing a promising basis for further optimization towards clinical pERK1/2 modulating
ERK1/2 inhibitors.
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
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 4 of 42
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
INTRODUCTION
The extracellular signalꢀrelated kinases 1 and 2 (ERK1/2) are highly homologous serine/threonine
kinases that form a key node in the transduction of growth factor signals, regulating a range of cellular
1
functions including cell survival, differentiation, adhesion, migration and proliferation. Of particular
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
relevance to oncology, the RASꢀRAFꢀMEKꢀERK signalling cascade (MAPK pathway) is activated
through mutations in RAS or RAF in over 30% of cancers including melanoma, lung, pancreatic and
2
colorectal (CRC) cancer. Although RAS has proven refractory to drug targeting for many years, the
successful development of inhibitors of BRAF and MEK1/2 kinases has led to effective treatment
particularly of melanomas whose tumor growth is driven by activating mutations in BRAF such as
3
V600E. Despite these successes, resistance often emerges after several months leading to increased
4,5
signalling through ERK1/2.
Human ERK1 and ERK2 are 84% identical in sequence and show some redundancy in cellular
function, although some studies have suggested that ERK2 plays the dominant role in driving cell
6
,7,8ꢀ9
proliferation.
ERK1/2 are activated by MEK1/2 through sequential dual phosphorylation of
Tyr187 and Thr185 residues in the ERK1/2 activation loop. Diphosphorylation of ERK1/2 triggers
1
0
two distinct modes of function. Firstly, the activated ERK1/2 catalyse the phosphorylation of
numerous cytosolic downstream substrates, including RSK which is commonly used as a biomarker
for ERK1/2 activity and inhibition. Secondly, the phosphorylation of ERK1/2 leads to their
translocation to the nucleus, where they phosphorylate and activate nuclear substrates including
transcription factors.
Motivated by the importance of the ERK1/2 signaling cascade in cancer and the validation through
approved inhibitors of BRAF and MEK, multiple groups have targeted ERK1/2 inhibitors, including
1
1
12
13
several reaching clinical trials (Figure 1) (ulixertinib (1), GDCꢀ0994 (2), MKꢀ8353, LY3214996
14
15
16
17ꢀ20 21 22 23
(
3), LTT462, KOꢀ947, and others disclosed.
Among the clinical studies completed to
date, ulixertinib reported responses in BRAF V600 and nonꢀV600 mutant solid tumors including
melanoma, glioblastoma multiforme, lung cancers with brain metastases, gallbladder and head & neck
3
ACS Paragon Plus Environment

Page 5 of 42
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
2
4
25
cancers; while MKꢀ8353 showed responses in BRAF V600ꢀmutant melanoma. The majority of
clinical ERK inhibitors are ATP competitive, blocking the catalytic activity of ERK1/2, but do not
prevent its phosphorylation by MEK, and hence do not prevent the nuclear localisation of pERK. The
series of compounds represented by SCH772984 (4, Figure 1) published by Schering/Merck, and
structurally related to MKꢀ8353, showed the ability to inhibit both the catalytic activity of ERK1/2
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
6
and their phosphorylation by MEK. This dual pharmacology might be expected to more completely
suppress MAPK pathway signalling, with potential benefits in terms of efficacy and emergence of
resistance.
We sought to assess the potential benefits of the dual inhibitionꢀpERK modulation profile in in vivo
models, but were unable to achieve good exposure levels following oral or i.p. dosing of 4, prompting
us to seek an alternative chemotype with similar pharmacology. Crystal structural studies performed
27
by us (vide infra) and others on 4 suggested that it binds to an altered conformation of ERK2, in
which Tyr36 in the glycineꢀrich loop (Pꢀloop) becomes tucked under the loop, allowing the extended
structure of the inhibitor to occupy a second pocket. We hypothesized that the binding mode of 4
might underlie its ability to block the phosphorylation of ERK1/2, and initiated a fragmentꢀbased
approach to develop orally bioavailable ERK1/2 inhibitors which bind to a similar conformation of
ERK1/2 and which also modulate the phosphorylation of ERK1/2.
RESULTS AND DISCUSSION
Fragment screen and structural insights
Given the high sequence similarity between ERK1 and 2, we used ERK2 for structural and SAR
studies. In order to obtain crystals of human ERK2 for fragment and inhibitor structural analysis,
18
conditions similar to those published for expression, purification and crystallization were used.
By soaking crystals of ERK2 in a solution of 4, a 2.3Å coꢀstructure was obtained. The structure is
27
similar to that subsequently published. Of particular note is the occupancy of a second pocket in
ERK2 upon binding SCH772984. In structures of ERK2 bound with ATPꢀcompetitive hinge binding
4
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 6 of 42
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
inhibitors such as ulixertinib, Tyr36 on the glycineꢀrich loop is involved in a piꢀpi stacking interaction
with Tyr64 on the Cꢀα helix (“out” conformation, Figure 2a). By contrast, in order to accommodate
the aromatic tail portion of 4, the glycineꢀrich loop undergoes a conformational change in which
Tyr36 tucks under the loop and stacks on the hydrophobic side of the pyrrolidine ring of 4 (“in”
conformation, Figure 2b). The pyrimidine ring of the inhibitor mimics Tyr36 by engaging in a
stacking interaction with Tyr64. The central pyrrolidine nitrogen and two amide carbonyl oxygens of
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
also mediate an unusual tridentate interaction with Lys54.
A recently published NMR study showed the presence of multiple conformational states for ERK2 in
28
solution, depending on the phosphorylation state. This work provided evidence that the distribution
of conformational states could be altered in the presence of the Vertex inhibitor 11e (Supplementary
Figure 1), closely related to ulixertinib, favouring the active conformation of ERK2 when the enzyme
is doubly phosphorylated. These findings supported the notion that changes in ERK1/2
conformational distributions in response to different inhibitor binding modes might influence
molecular recognition by ERK and by extension the phosphorylation state.
A crystallographic screen of our corporate fragment collection was conducted by soaking crystals of
ERK2 in solutions of fragments. In addition, a screen using thermal melt (differential scanning
fluorimetry, DSF) was carried out, together with a virtual library screen. Hits from these methods
were evaluated for inhibitory potency in an ERK2 bioassay. These approaches led to the identification
of several fragments binding to ERK2, including the pyrazolyl aminopyrimidine fragment 5, which
binds to the hinge of ERK2 (Figure 2c); and aminoindazole fragment 6, which binds not only to the
hinge but also in the second pocket between the Pꢀloop and Cꢀα helix (Figure 2d). Interestingly,
binding of fragment 6 elicits a third conformation of Tyr36 in which it is pushed away from Tyr64
towards solvent. The fact that Tyr36 could be displaced to allow such a small, weakly binding
fragment to occupy the second pocket suggests a significant conformational flexibility in the ERK2 Pꢀ
loop. This structure was also useful as it provided additional insights with which to design inhibitors
occupying this pocket.
5
ACS Paragon Plus Environment

Page 7 of 42
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Among the hinge binding fragment hits, fragment 5 showed promising affinity and ligand efficiency
(
Table 1), with attractive synthetic vectors for further optimization and elaboration. The structure of
fragment 5 bound to ERK2 indicated several key interactions with the protein which drive its affinity
Figure 3a). Hydrogen bonds were observed between the pyrimidine N and amino NH of the fragment
(
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
and the ERK2 backbone NH and C=O of hinge residue Met108 respectively. In addition, the pyrazole
nitrogen accepts a hydrogen bond from Lys114. The methyl group of the pyrimidine points towards
the side chain of the gatekeeper residue Gln105.
Prior to growing further into the ATP binding site, we considered two features to guide our initial
optimisation of fragment 5. First, we explored replacing the pyrazole moiety with an oxan (THP), in
3
order to increase the sp content of the fragment and mitigate against potential future solubility
problems in more elaborated lead molecules. The resulting fragment 7 showed a >10ꢀfold increase in
affinity and ligand efficiency (Table 1). Second, given the close proximity of the pyrimidine methyl
group to the side chain of gatekeeper residue Gln105, we explored the effects of larger and smaller
groups in this position. The data for fragments 8-10, containing hydrogen, fluoro or chloro
respectively, indicated a preference for larger substituents (Table 1), such that the chloroꢀsubstituent
gave a modest increase in affinity, attributable to a halogen bonding interaction between the Cl sigma
hole and the Gln105 side chain carbonyl oxygen. The structure of chloropyrimidine oxanyl fragment
10 shows similar interactions with the hinge residues as fragment 5, but an altered conformation of the
side chain of Lys114 in order to achieve a hydrogen bond between the 4ꢀoxanyl oxygen and the Lys εꢀ
amino group (Figure 3b, c).
Growth towards the gatekeeper: identification of the isoindolinone core
Having identified a potent and ligand efficient hinge binder in fragment 10, we next sought to grow
across the ATP binding site towards the second pocket. Prototypical ERK inhibitors such as
ulixertinib avoid the introduction of an additional hydrogen bond donor into the inhibitor to interact
with the gatekeeper carbonyl oxygen by exploiting a waterꢀbridged interaction with a carbonyl
oxygen in the inhibitor. In our series, docking experiments suggested that benzolactams such as 11
6
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 8 of 42
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
and 12 (Table 1) should position their carbonyl oxygen in a similar location. Experimentally, both
benzolactams achieved a dramatic jump in inhibitory potency, with ERK2 biochemical IC values of
50
11 and 24 nM respectively. The crystal structure of 11 in complex with ERK2 indicates minimal
movement of the hinge binding component, while the isoindolinone indeed positions its carbonyl
oxygen to allow a waterꢀbridged interaction with the gatekeeper carbonyl in a similar fashion to
ulixertinib (Figure 3d). The isoindolinone carbonyl oxygen also accepts a hydrogen bond from the
side chain of Lys54, in a manner reminiscent of the binding of 4. The isoquinolinone 12 binds
similarly to ERK2 (structure not shown); however, in view of its favourable ligand efficiency and
attractive vector towards the second pocket we chose to use the less lipophilic isoindolinone 11 as the
basis for further SAR studies.
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
Growth towards the second pocket: interactions with catalytic Lys54
In order to design linkers to grow from the benzolactam core of 11 towards the second pocket, we
superimposed the crystal structure of the ERK2ꢀ11 complex onto the ERK2 structure when bound to 4
(
Figure 3e). With the Tyr36 “in” conformation, a narrow channel exists between the hinge region and
the second pocket, bounded primarily by Tyr36 and Lys54. We therefore sought to incorporate
linking groups that would be sufficiently slim to fit through this channel, and contain at least one
hydrogen bond accepting atom to interact with the Lys54 amino group. The ether analogue 13 showed
a modest increase in potency; the coꢀcrystal structure indicated that the ether oxygen approaches the
Lys54 amino group but is positioned somewhat far for a strong hydrogen bond (OꢀN distance 3.4Å),
and in this coꢀstructure Tyr36 adopts an “out” conformation (Supplementary Figure 2a). This
V600E
analogue was the first in the series to show appreciable antiꢀproliferative activity in BRAF
mutant
cell lines, with IC values of 980 and 870 nM in A375 and Colo205 96 hour proliferation assays
50
respectively. The morpholino analogue 14 was designed to provide an alternative hydrogen bond
acceptor to interact with Lys54; the crystal structure shows that the morpholino group does not
interact with Lys54, but rather tucks under the Pꢀloop. This binding mode also requires a significant
twist about the CꢀC bond linking the pyrimidine and isoindolinone rings, which might be energetically
less favourable, consistent with the lower activity (Supplementary Figure 2b). The next series of
7
ACS Paragon Plus Environment

Page 9 of 42
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
analogues explored the attachment of an acetamide group to the isoindolinone nitrogen. The tꢀbutyl
amide 15 showed similar potencies to ether 13 in the biochemical and antiꢀproliferative assays (Table
1
). In the ERK2 coꢀcrystal structure of 15, the acetamide oxygen is positioned close to the Lys54
amino group but again not close enough to form a strong hydrogen bond (NꢀO distance 3.55 Å)
Figure 2f). However, in this structure Tyr36 adopts an “in” conformation, and the amide nitrogen of
5 provides a synthetically accessible growth point into the second pocket, here partially filled by 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
(
1
tꢀbutyl group. The Nꢀmethyl tertiary amide analogue 16 shows similar biochemical and antiꢀ
proliferative potencies. The ERK2 structure of 16 shows a similar binding mode to 15, with a slight
shift in the positioning of the tꢀbutyl group to accommodate the extra methyl group (Figure 3g).
In vitro and in vivo pharmacology of early leads
Before pursuing inhibitors extending further into the second pocket, we profiled compounds 13-16 to
understand the potential of this chemotype for oral exposure (Table 2). All the compounds showed
promising permeability in a Caco2 monolayer assay. Mouse in vivo pharmacokinetic experiments
identified the chemotype as showing low to moderate clearance with low volumes of distribution, and
a range of oral exposures. While the methoxyethyl derivative 13 showed low oral exposure, this was
significantly improved in the morpholino derivative 14, although surprisingly the addition of the
weakly basic centre did not increase the volume of distribution significantly. Among the two
acetamido derivatives 15 and 16, methylation of the amido NH improved oral absorption, consistent
with improvements in both permeability and clearance.
Since the tertiary amide 16 showed promising antiꢀproliferative activity and oral exposure, we carried
V600E
out additional pharmacological profiling (Figure 4). In BRAF
mutant melanoma A375 cells,
Western blot analysis confirmed that 16 suppresses the phosphorylation of the ERK substrate RSK
Figure 4a). Under these conditions 16 elicits negligible effects on ERK/pERK levels. Quantitation of
pRSK by Mesoscale Discovery (MSD) assay gave a cellular IC value of 630 nM (Figure 4b). A
(
50
single oral 150 mg/kg dose of compound 16 administered to mice bearing subꢀcutaneous colorectal
8
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 10 of 42
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
V600E
BRAF
mutant Colo205 xenografts elicited largely complete tumor pRSK suppression at 2 and 6
hours, which recovered at 24 hours (Figure 4c). At the 2 and 6 hour timepoints, 16 appeared to
somewhat increase tumor pERK levels, possibly arising from activation of MAPK pathway feedback
as a result of ERK inhibition. In mice bearing subꢀcutaneous Colo205 tumors, once daily oral dosing
of 16 at 150 mg/kg conferred significant antiꢀtumor activity (Figure 4d). Compound 16 was wellꢀ
tolerated and no excessive bodyweight loss was observed (Supplementary Figure 3a).
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
Growth into the second pocket: discovery of pERK modulators
With these promising results in hand, we sought to build on the favorable properties of amides 15 and
16 and grow further into the second pocket to explore potential effects on pERK modulation.
Inspection of the ERK2 coꢀstructure of 15 showed two water molecules in the cavity between the
inhibitor tꢀbutyl group and Tyr64 (Figure 5a). Docking experiments suggested that replacement of the
tꢀbutyl group with benzylic groups might occupy this pocket efficiently. Experimentally, the success
of this approach depended on the substitution at the benzylic carbon. The simple benzyl amide 17
showed no potency improvement in the biochemical assay, but a significant leap in potency in the
antiproliferative assays (Table 3). The branched methyl (R)ꢀenantiomer 18 also showed good potency
in both the biochemical and antiproliferative assays, while the (S)ꢀenantiomer 19 was less active. The
dimethyl substituted analogue 20 showed intermediate activity in both biochemical and
antiproliferative assays. A structure of 17 was not obtained, but the structure of 18 shows the phenyl
group occupying the second pocket as intended (Figure 5b), while the methyl group is accommodated
in the space between Thr68 and Glu61 of the Cꢀαꢀhelix. In this structure Tyr64 adopts a slightly
altered conformation in which its phenol ring is twisted to engage in an edgeꢀtoꢀface piꢀpi interaction
with the phenyl group of the inhibitor, forming a lid over the second pocket. Presumably the (S)ꢀ
enantiomer 19 is less active because in order to place its phenyl ring in the same position as 18, the
methyl group would be eclipsed with the amide carbonyl, representing a potentially less favourable
inhibitor conformation. Of note, the α,αꢀdimethyl analogue 20 binds in a different conformation in
9
ACS Paragon Plus Environment

Page 11 of 42
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
which the phenyl group folds back over the benzolactam core, mimicking Tyr36 in the structure of 18,
while Tyr36 reverts to an “out” conformation (Figure 5c).
The ERK2 coꢀstructure of αꢀmethyl benzyl amide 18 suggested the potential for further potency gains.
At this point we used the cellular assay to drive structureꢀactivity relationships, as highly active
compounds had a potency that exceeded the detection limit of the enzyme assay. First, although the
phenyl group is somewhat embedded in the second pocket, the 3ꢀposition is oriented towards solvent,
providing a vector for additional substitution. Introduction of a methoxy group in this position,
compound 21, afforded a 3ꢀ4ꢀfold increase in cellular potency, giving IC50s of 37 and 33 nM in the
A375 and Colo205 antiproliferation assays respectively (Table 3). The ERK2 coꢀstructure of 21
shows increased surface contacts between the inhibitor methoxyphenyl group and Tyr64; in addition
the methoxyphenyl resides over the piꢀcation system of Arg67 (Figure 5d). Second, inspection of the
protein environment in the vicinity of the benzylic methyl group of 18 suggested that introduction of a
hydroxyl group at this position might elicit favourable interactions with the polar sidechain of Asp167
and the backbone NH of Gly169. The corresponding (S)ꢀhydroxymethyl analogue 22 showed an
approximately 3ꢀfold increase in cellular potency compared with compound 18, with IC50s of 34 and
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
49 nM in the A375 and Colo205 antiproliferation assays, while the (R)ꢀenantiomer is less active
(
Table 3). The ERK2 coꢀstructure of 22 shows that the hydroxyl group indeed interacts with the side
chain of Asp167 as intended, but intriguingly, the side chain of Arg67 adopts an altered conformation
to also interact with the inhibitor hydroxyl group (Figure 5e). In this way the hydroxyl group bridges a
chargeꢀcharge interaction between Asp167 and Arg67.
Combination of these two SAR features appeared to be additive, such that incorporation of both the 3ꢀ
methoxy group and benzylic hydroxyl in compound 24 gave high cellular potency with IC s of 7.1
5
0
and 7.3 nM in A375 and Colo205 antiproliferation assays.
We also explored the effects of substitution on the methylene part of the acetamide linker. Inspection
of the ERK2 coꢀstructure of compound 22 suggested that a methyl group might be tolerated in this
position with an (R)ꢀconfiguration to direct the substituent towards solvent. The two diastereomeric
1
0
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 12 of 42
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
analogues of 22 bearing (R) and (S) methyl groups were synthesized: the (R) diastereomer 25 showed
similar biochemical and cellular potency to 22, while the (S)ꢀdiastereomer 26 was less potent,
possibly due to steric crowding of the methyl group with the side chain of Asp167. Although this
methyl group did not modulate the potency, it proved to be useful in increasing oral exposure in mice
as discussed below.
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
Combination of SAR learnings described above, including the hydroxymethyl benzyl group with a
methyl group on the 3ꢀposition of the phenyl ring and an (R)ꢀmethyl group on the acetamide linker,
afforded compound 27 which also showed excellent antiꢀproliferative potency with IC50s of 4.9 and
7.5 nM in A375 and Colo205 cells respectively.
Several of the more potent analogues were profiled in mouse pharmacokinetic studies to assess their
potential for further progression into pharmacodynamic and efficacy studies (Table 2). The
unsubstituted benzyl analogue 17 showed good oral exposure comparable with that of the tꢀbutyl
amide 16. The addition of the benzylic methyl group and 3ꢀmethoxy group in 21 lowered oral
bioavailability, potentially due to the effect of increased lipophilicity on solubility. The
hydroxymethyl analogues 22, 24, 25 and 27 showed moderate to good bioavailability and,
interestingly, somewhat higher volumes of distribution which might be desirable for tumor exposure.
In this grouping the 3ꢀmethoxy substitution in 24 resulted in a significant jump in clearance, with a
correspondingly lower oral AUC and bioavailability than the other analogues. Compound 27
possessed a promising combination of antiproliferative activity against mutant BRAF cells and oral
exposure in mice, and was profiled in selectivity and mechanistic cellular assays.
The in vitro activity of 27 was confirmed against both ERK1/2 isoforms: activity was close to the
limit of the assays, showing 60% inhibition at 1 nM and an IC of 3.0 nM against ERK1 and ERK2
50
respectively. Screening against a panel of 429 kinases revealed excellent selectivity towards ERK1/2,
with only 3 kinases inhibited at >50% at 100 nM and 16 kinases at 1µM (Supplementary Table 1). We
ascribe this selectivity to a number of unique features in the binding mode of compound 27 to ERK:
3
the sp character of the oxanyl group in the solvent exposed channel; the chloro and isoindolinone
1
1
ACS Paragon Plus Environment

Page 13 of 42
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
interactions with the glutamine gatekeeper residue, present in only 2% of kinases; and the shape and
hydrogen bonding motifs designed to closely complement the unusual second pocket of ERK1/2.
In A375 cells 27 suppressed phospho levels of the downstream substrate RSK, with an IC50 of 3.5 nM
as measured by MSD assay (Figure 6a). In addition, 27 suppressed the phosphorylation of ERK itself,
achieving an IC value of 2.3 nM using ELISA (Figure 6b). Pharmacodynamic assays performed in
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
5
0
mice bearing subꢀcutaneous Colo205 xenografts confirmed in vivo suppression of both pRSK and
pERK in tumors following a single 25 mg/kg oral dose of compound 27 (Figure 6c). Maximal
suppression of pRSK reached approximately 90% at 2 hours after dosing, while maximal suppression
of pERK at approximately 70% was also observed at 2 hours, both partially recovering at 6 hours and
completely at 24 hours.
Repeat oral dosing of compound 27 over 15 days elicited tumor regression in mice bearing Colo205
xenografts (Figure 6d). The efficacy observed in response to either once daily dosing at 50 mg/kg or
twice daily dosing at 25 mg/kg was very similar, suggesting that sustained suppression of ERK1/2
signalling may not be required for effective tumor growth inhibition. Compound 27 was wellꢀ
tolerated and no excessive bodyweight loss was observed (Supplementary Figure 3b).
Synthesis
Fragments 5 and 7-10 were synthesized in a single step from the corresponding 5ꢀsubstituted 2ꢀ
chloropyrimidines and aminopyrazole or oxanꢀ4ꢀamine respectively. The isoindolinone 11 was
prepared according to Scheme 1. Commercial 6ꢀbromoꢀ2,3ꢀdihydroꢀ1Hꢀisoindolꢀ1ꢀone 30 was
converted to the pinacol boronate ester 31 by palladium mediated coupling, followed by a second
palladium coupling to furnish the dichloropyrimidine derivative 32, and then heating in the presence
of oxanꢀ4ꢀamine. The homologous isoquinolinone 12 required construction of the heterobicycle as
shown in Scheme 2. First, 2ꢀ(4ꢀchlorophenyl)ethanꢀ1ꢀamine was converted to the methyl carbamate
34, which was then cyclized to give 7ꢀchloroꢀ1,2,3,4ꢀtetrahydroisoquinolinꢀ1ꢀone (35) under acidic
conditions. Elaboration to give 12 then followed a similar process of borylation, Suzuki coupling and
1
2
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 14 of 42
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
chloride displacement with oxanꢀ4ꢀamine. In this case the more active XPhos ligand was required to
couple the aryl chloride 35.
The Nꢀsubstituted isoindolinone derivatives 13 and 14 were prepared according to Scheme 3. Nꢀ
Alkylation of 6ꢀbromoꢀ2,3ꢀdihydroꢀ1Hꢀisoindolꢀ1ꢀone (30) was effected with sodium hydride and the
corresponding alkyl bromide, followed by sequential borylation and Suzuki coupling to give the
dichloropyrimidines 38 and 39. Displacement of the 2ꢀchloro substituent with oxanꢀ4ꢀamine gave the
final compounds 13 and 14. The acetamide derivatives 15ꢀ24 were prepared following the general
procedure illustrated for compound 15 in Scheme 4. Alkylation of 6ꢀbromoꢀ2,3ꢀdihydroꢀ1Hꢀisoindolꢀ
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
1ꢀone (30) with tꢀbutyl bromoacetate followed by palladium coupling gave the boronate ester 40,
which was converted to the dichloropyrimidine 41 by Suzuki coupling. Displacement of the 2ꢀchloro
substituent with oxanꢀ4ꢀamine and cleavage of the tꢀbutyl ester using anhydrous acid afforded the key
intermediate acid 42. Coupling to give the final amide 15 was effected with tꢀbutylamine and HBTU.
For other amides, alternative coupling agents were used, including HATU or T3P.
The synthesis of the 2ꢀpropionamide derivatives 25ꢀ27 required a modified route as outlined in
Scheme 5. In order to install the more bulky methylated linker, cyclization of bromomethyl
bromobenzoic acid 43 with alanine tꢀbutyl ester was used to form the Nꢀsubstituted lactam.
Conversion to the boronate ester and Suzuki coupling with trichloropyrimidine were effected as
previously. Attempts to install the aminoꢀoxanyl group at this stage led to racemization of the methyl
group, and so the steps were reversed by first cleaving the tꢀbutyl ester using TFA to give acid 47,
then carrying out the aminoꢀoxan substitution to give the intermediate 48. Finally, amide couplings
using HATU or TBTU gave the final compounds 25ꢀ27.
CONCLUSION
In recent years, fragmentꢀbased drug discovery has established a track record for providing novel start
points for medicinal chemistry campaigns, in some cases identifying alternative binding modes and/or
1
3
ACS Paragon Plus Environment

Page 15 of 42
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
allosteric sites for ligand binding, and providing a rational basis for the design of drug leads and
2
9 30
candidates with carefully controlled physicochemical properties. ,
Here, a combined approach using Xꢀray crystallographic and biophysical fragment screening together
with structure based optimization led to the discovery of highly potent and selective orally
bioavailable ERK1/2 inhibitors which also block the phosphorylation of ERK1/2. Careful
optimization of the initial hinge binding fragment hit proved important in achieving the overall profile
of the lead compound. Due to the relatively large distance from the hinge to the second pocket,
parsimonious use of molecular weight and hydrogen bond donors was critical to balance permeability
with other properties in the series. The chloro substituent allowed an energetically favourable
interaction with the side chain carbonyl oxygen of gatekeeper Gln105 without adding a hydrogen
bond donor. The design of the isoindolinone core allowed a favourable interaction with the side chain
carbonyl of gatekeeper Gln105 bridged via a water molecule, while providing an optimal spatial and
synthetic vector with which to grow towards the second pocket. Acetamide and 2ꢀpropionamide
linkers provided access to the second pocket while interacting favourably with the catalytic lysine 54.
Finally, substituted hydroxymethyl benzyl groups provided good shape complementarity, stacking
and hydrogen bonding interactions with residues lining the second pocket.
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
Occupancy of the second pocket correlated with modulation of pERK levels, confirming the original
structural hypothesis. Hence, the smaller inhibitors represented by tꢀbutyl amide 16 did not show
modulation of pERK levels in vitro or in vivo; while inhibitors occupying the second pocket with a
benzylic group, represented by compound 27, suppressed pERK levels. The lead compound 27
suppressed both pRSK and pERK1/2 in mutant BRAF cells in vitro and in vivo, and displayed
promising efficacy in Colo205 xenograft models. Further pharmacological characterization of 27 will
be published elsewhere.
1
4
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 16 of 42
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
EXPERIMENTAL PROCEDURES
Protein Expression, Purification & Crystallography
Full length human ERK2 (hERK2) was cloned into pET30a with a nonꢀcleavable MAHHHHHH Nꢀ
terminal tag. hERK2 was expressed in E.coli BL21(DE3) and nonꢀphosphorylated hERK2 (confirmed
by LCMS) was purified using sequential NiꢀHiTRAP, desalt, ResourceꢀQ and S75 26/60 column
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
steps. hERK2 was crystallised under conditions adapted from Aronov et al. and crystals were
soaked in a solution equivalent to the crystallisation solution but also containing 0.1ꢀ100 mM ligand,
10mM DTT and 10% DMSO final. Crystals were cryoꢀprotected using crystallisation solution
containing 35% 2KMPEG final. Xꢀray diffraction data were collected using both inꢀhouse and
synchrotron Xꢀray sources.
ERK2 kinase assay
ERK2 kinase activity was determined using a time resolved fluorescence (TRF) activity assay. ERK2
enzyme (0.25 nM) was incubated with the substrates ATF2ꢀGFP (0.4µM) and ATP (20 µM) in 50
2
mM Tris pH7.5, 10 mM MgCl , 1 mM EGTA, 0.01% Triton X100, 1 mM DTT and 2.5% DMSO,
with shaking at room temperature for 30 mins. Reactions were stopped by the addition of a stop and
detection mix, containing 25 mM EDTA, 2 nM TbꢀpATF2 antibody in TRꢀFRET dilution buffer (Life
Technologies, Paisley, UK) and the plate was incubated with shaking at room temperature for 1 hour.
Upon excitation at 340nm, fluorescence was measured at 520nm (A counts) and 495nm (B counts)
using a Pherastar plate reader (BMG Labtech, Ortenberg, Germany).
Cell culture and reagents
The human cell lines A375 and Colo205 were purchased from the American Type Culture Collection
(
ATCC, Manassas,VA, USA) and the European Collection of Authenticated Cell Cultures (ECACC
Salisbury, UK), respectively. The cells lines were not passaged for more than 6 months after
authentication by the cell banks (DNA fingerprinting and cytogenetic analysis or shortꢀtandem repeat
(
STR) PCR). Both cell lines were grown in DMEM culture medium, supplemented with 10 % FBS
1
5
ACS Paragon Plus Environment

Page 17 of 42
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
(
ThermoFisher Scientific, Waltham, Massachusetts, USA) and maintained at 37 °C in an atmosphere
of 5 % CO2.
Cell Proliferation Assay
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
Cell proliferation assays were carried out using Alamar Blue (ThermoFisher Scientific, Waltham,
3
1
3
Massachusetts, USA) as described previously. Briefly, 5x10 cells were seeded in complete culture
medium into flatꢀbottomed 96ꢀwell plates, one day before the drug treatment. Cells were incubated
with compound in 0.1% (v/v) dimethyl sulfoxide (DMSO) for 96 hours before viability was assessed
using Alamar blue. IC values were calculated using a sigmoidal dose response equation (Prism
5
0
GraphPad software, La Jolla, CA, USA).
In vitro permeability
Permeability of test compounds was assessed using the CacoReadyTM system. Test and control
compounds (propranolol, antipyrine, vinblastine) were incubated at a final concentration of 10 µM in
duplicate to either the apical (180 µL) of the monolayer to measure apical to basolateral transport
(A>B) across the cell barrier or to the basolateral side (750 µL) to measure the basolateral to apical
transport (B>A). For A>B spiking, test and control compounds were diluted from 10 mM DMSO
stocks to 10 µM in HBSS buffer with 100 mM CaCl .2H O, 50 mM MgCl .5H O and 0.5 mg/mL
2
2
2
2
lucifer yellow made to volume with sterile water (lucifer yellow was used to determine the integrity of
the Cacoꢀ2 monolayer). For B>A compound spiking, test and control compounds were prepared as
for the A>B solutions without the addition of lucifer yellow. Test and control compounds were
2
incubated for 1 hour at 37°C in a highly humidified atmosphere of 95% air and 5% CO .
Western Blotting
6
Cells were seeded at 10 cells per well, incubated overnight at 37 °C and then treated with compound
for 16 hours. Cells were harvested and lysed in iceꢀcold Triton lysis buffer and lysates cleared by
centrifugation. Colo205 xenograft tumor lysates were prepared by grinding the frozen tissue to a fine
powder with a mortar/pestle under liquid nitrogen and then adding iceꢀcold Mesoscale Discovery
1
6
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 18 of 42
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
(
MSD) lysis buffer (Rockville, Maryland, USA). Samples were freezeꢀthawed, vortexed and then
lysates cleared by centrifugation. Protein concentrations were determined using a BCA protein assay
ThermoFisher Scientific, Waltham, Massachusetts, USA). Equivalent amounts of protein lysates
(
were resolved by SDS PAGE TrisꢀGlycine gel systems, transferred to nitrocellulose membranes and
incubated with primary antibodies specific for phosphoꢀp90RSK (Thr359), phosphoꢀp42/44 MAPK
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
(
ERK1/2), p44/42 MAPK (ERK1/2) (Cell Signaling Technology, Massachusetts, USA), RSK (R&D
Systems, Minneapolis, USA), and actin (Abcam, Cambridge, UK). Blots were then incubated with
infraredꢀdyeꢀlabeled secondary antibodies and detected on the Odyssey infrared imaging system
(
LiCor Bioscience, Lincoln, USA).
Quantification of pRSK by MSD
4
A375 cells were seeded at 1.5 x 10 cells per well into 96ꢀwell plates and allowed to recover for 16 h,
prior to the addition of compounds (in 0.1% DMSO v/v) and incubation for a further 4 h. Cells were
lysed by adding cell lysis buffer (Cell Signaling Technology, Massachusetts, USA) and incubating at
room temperature for 20 minutes. Custom MSD plates (Meso Scale Discovery, Maryland, USA) preꢀ
coated with antiꢀpRSK antibody (Cell Signaling Technology, Massachusetts, USA) were blocked with
kit blocking buffer for 1 hour at room temperature, prior to washing. Equivalent amounts of protein
lysate were added to the blocked plates and incubated for 3 hours at room temperature. After washing,
plates were incubated for 1 hour at room temperature with sulfoꢀtag conjugated antiꢀRSK detection
antibodies (R&D Systems, Minneapolis, USA). Plates were washed, and read buffer added before
reading the plate on a MESO QuickPlex SQ 120 (Meso Scale Discovery, Maryland, USA). IC50
values were calculated using a sigmoidal dose response equation (Prism GraphPad software, La Jolla,
CA, USA). For pharmacodynamic studies, Colo205 xenograft tumor lysates were prepared as
described above, and transferred directly to Custom MSD pRSK plates (Meso Scale Discovery,
Maryland, USA) for analysis.
Quantification of pERK by ELISA
A375 cells were treated with compounds and lysed as described above. Lysates were then cleared by
centrifugation, transferred to pERK ELISA plates (Cell Signalling technology, Massachusetts, USA)
1
7
ACS Paragon Plus Environment

Page 19 of 42
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
and analysed according to manufacturer’s instructions. IC50 values were calculated using a sigmoidal
dose response equation (Prism GraphPad software, La Jolla, CA, USA). For pharmacodynamic
studies, Colo205 xenograft tumor lysates were prepared as described above, transferred to pERK
ELISA plates (Cell Signalling technology, Massachusetts, USA) and analysed according to
manufacturer’s instructions.
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
In vivo studies
The care and treatment of experimental animals were in accordance with the United Kingdom
32
Coordinating Committee for Cancer Research guidelines and the United Kingdom Animals
33
(
Scientific Procedures) Act 1986. Mouse studies were performed with mice allowed access to food
and water ad libitum.
Pharmacokinetic studies
For intravenous administrations, test compounds were formulated in either 20% DMA/80% saline,
10% DMSO/20% PEG400/70% water or 10% DMSO/90% saline. For oral administrations test
compounds were formulated in either 10% ethanol/10% cremaphor/10% PEG200/70% water, 20%
PEG400/80% HPMC or 10% DMSO/20% cremaphor/70% saline. Intravenous dosing at 0.5 mg/kg
was administered via the lateral tail vein at a dose volume of 5 mL/kg. Test compounds were
administered orally at 5 mg/kg by nasogastric gavage at a dose volume of 10 mL/kg. All doses were
calculated as freebase equivalent per kg of bodyweight. Pharmacokinetic studies were performed in
male Balb/c mice, obtained from Harlan Laboratories Inc. (Shardlow, UK). Following dosing in
mice, blood samples (0.2 mL) were drawn in tubes containing potassium EDTA, via either saphenous
vein bleeding or cardiac puncture at various time points over 24 hours using sparse sampling (n=3 per
time point), prior to centrifugation (2000 g at 4°C, 10 min). The resultant plasma was separated from
the erythrocyte pellets for analysis and stored at ꢀ20°C. Terminal samples were collected via cardiac
puncture and centrifuged (10 000 g at 4°C, 2 min) and the resultant plasma stored at ꢀ20°C. Nonꢀ
compartmental pharmacokinetic (PK) analyses were performed using Phoenix 6.3.0.395® (Certara
USA, Inc.) software. Calculated parameters included clearance, volume of distribution (Vss), time of
1
8
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 20 of 42
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
maximum observed concentration (Tmax), maximum concentration (Cmax), terminal halfꢀlife, area
under the curve (AUC) from the time of dosing to the last measurable concentration (AUClast) and
extrapolated to infinity (AUC0ꢀ∞). Nonꢀcompartmental PK fitting to sparse sampling data allowed
the calculation of standard errors on AUClast and Cmax.
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
Xenograft Studies
6
Colo205 xenografts were prepared by subcutaneously injecting 5 × 10 cells suspended in serumꢀfree
medium mixed 1:1 with Matrigel (BD Biosciences, San Jose, CA, USA) into the right flank of each
male BALB/c nude mouse. Tumors were measured using caliper and tumour volumes calculated by
applying the formula for an ellipsoid. For efficacy studies, when the tumours reached an average of
3
approximately 200 mm , mice were randomised into groups of 8. Mice were dosed orally once a day
at a dose of 150 mg/kg of compound 16 and either once a day with 50 mg/kg or twice a day with 25
mg/kg of compound 27. Body weights were recorded daily and tumor volumes were measured every
3–4 days. For PKPD studies, a single dose of compound was administered orally to mice. Following
dosing, blood samples were drawn in tubes containing potassium EDTA via either saphenous vein
bleeding or cardiac puncture at various time points, prior to centrifugation (2000g at 4 °C, 10 min).
The resultant plasma was separated from the erythrocyte pellets for analysis and stored at −20 °C.
Tumors were immediately excised and flashꢀfrozen in liquid nitrogen.
Bioanalysis
All in vitro and in vivo samples were extracted by protein precipitation with acetonitrile containing
internal standard (1:3 v/v). For quantitative studies, calibration standards and quality controls were
prepared in blank matrix and extracted under the same conditions. All samples were centrifuged at
3700 rpm at 4°C for 20 min. Compound bioanalysis of all in vitro and in vivo samples were
performed using high performance liquid chromatography mass spectrometry (LC/MS). Test
compounds and the internal standard were ionised using positive mode (ESI+) electrospray ionisation.
Analytes were detected using multiple reaction monitoring (MRM). Test compounds typically ran on
a gradient HPLC method over 5 min with a 10 mM ammonium acetate and acetonitrile mobile phase
1
9
ACS Paragon Plus Environment

Page 21 of 42
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
at a flow rate of 0.5 mL/min. Separation was typically achieved using an Acquity HSS T3 1.8µM
50x2.1 mm column maintained at 40°C.
Synthesis of Compound 27 and intermediates 44-48
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
General procedures: All solvents and commercially available reagents were used as
received. All reactions were followed by TLC analysis (TLC plates GF254, Merck) or LCꢀ
MS (liquid chromatography mass spectrometry). LCꢀMS analysis was performed with an
Agilent or Shimadzu LC system with variable wavelength UV detection using reverse phase
chromatography with a CH CN and water gradient with a 0.02 or 0.1% TFA modifier (added
3
to each solvent) and using a reverse phase column, e.g., Thermo Hypersil Gold C18. MS was
determined using either PE Sciex 150EX LCꢀMS, Waters ZQ LCꢀMS, or Agilent 6140 LCꢀ
MS Single Quadrupole instruments. Column chromatography was performed on prepacked
silica gel columns (3090 mesh, IST) using a Biotage SP4 or similar. NMR spectra are
1
referenced as follows: H (400 MHz), internal standard TMS at δ = 0.00. Abbreviations for
multiplicities observed in NMR spectra: s; singlet; br s, broad singlet; d, doublet; t, triplet; q,
quadruplet; p, pentuplet; spt, septuplet; m, multiplet. All compounds reported are of at least
95% purity according to LCꢀMS unless stated otherwise.
(2R)-2-(6-{5-Chloro-2-[(oxan-4-yl)amino]pyrimidin-4-yl}-1-oxo-2,3-dihydro-1H-
isoindol-2-yl)-N-[(1S)-2-hydroxy-1-(3-methylphenyl)ethyl]propanamide (27)
Triethylamine (2.060 mL, 14.78 mmol) was added to a stirred suspension of (R)ꢀ2ꢀ(6ꢀ(5ꢀ
chloroꢀ2ꢀ((oxanꢀ4ꢀyl)amino)pyrimidinꢀ4ꢀyl)ꢀ1ꢀoxoisoindolinꢀ2ꢀyl)propanoic acid (1.54 g,
3.69 mmol) and (S)ꢀ2ꢀaminoꢀ2ꢀ(3ꢀtolyl)ethanol hydrochloride (0.763 g, 4.06 mmol) in DMF
(15 mL, 194 mmol). After 15 minutes, TBTU (1.305 g, 4.06 mmol) was added and the
mixture was stirred for 2 h at room temperature. The reaction was diluted with EtOAc (30
2
0
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 22 of 42
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
mL) and water (100 mL). The phases were separated and the aqueous phase was extracted
with EtOAc (2x30 mL). The combined organic extracts were washed with NH
4
Cl (100 mL),
NaHCO (100 mL), brine (2x100 mL), dried (MgSO ) and concentrated. The crude product
3
4
was purified by chromatography (SiO , 40 g column, 0ꢀ7% MeOH in DCM) to afford the
2
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
product as a colourless gum. The gum was triturated with diethyl ether (100 mL) to give the
title compound as a colourless solid (1.409 g, 68%).
1
H NMR (DMSOꢀd
6
) 8.52 (d, J = 8.1 Hz, 1H), 8.44 (s, 1H), 8.04 (d, J = 1.5 Hz, 1H), 7.97
(dd, J = 7.9, 1.7 Hz, 1H), 7.74 (d, J = 7.9 Hz, 1H), 7.67–7.50 (m, 1H), 7.20 (t, J = 7.5 Hz,
1H), 7.12–7.07 (m, 2H), 7.04 (d, J = 7.5 Hz, 1H), 5.00 (q, J = 7.2 Hz, 1H), 4.84 (t, J = 5.5
Hz, 1H), 4.82–4.71 (m, 2H), 4.59 (d, J = 18.1 Hz, 1H), 3.98–3.81 (m, 3H), 3.60–3.49 (m,
2H), 3.43–3.32 (m, 2H), 2.29 (s, 3H), 1.92–1.80 (m, 2H), 1.61–1.46 (m, 2H), 1.43 (d, J = 7.2
Hz, 3H).
13
6
C NMR (DMSOꢀd ) 170.67, 166.85, 161.40 (broad), 159.99, 158.10 (broad), 143.89,
1
1
40.99, 137.05, 135.92 (broad), 132.06, 131.97, 128.00, 127.44, 127.40, 123.79, 123.39,
23.12, 114.67, 65.99, 64.55, 55.19, 49.58, 47.15, 47.07, 32.27, 21.07, 16.28.
5 4
HRMS (ESIꢀQTOF): m/z [M+H]⁺ Calcd for C29H32ClN O 550.2213; Found 550.2205. ꢁ = ꢀ
1.46 ppm.
tert-Butyl (2R)-2-(6-bromo-1-oxo-2,3-dihydro-1H-isoindol-2-yl)propanoate (44)
DIPEA (28.5 ml, 163 mmol) and (R)ꢀtertꢀbutyl 2ꢀaminopropanoate hydrochloride (14.83 g,
8
2 mmol) were addded to a stirred solution of methyl 5ꢀbromoꢀ2ꢀ(bromomethyl)benzoate
(16.76 g, 54.4 mmol) in MeCN (100 ml) under nitrogen and the mixture stirred and heated to
5 °C for 24 h. The solvent was removed in vacuo and the residue partitioned between Et O
7
2
(500 mL) and 10% aqueous citric acid (200 ml). The organic phase was collected and washed
2
1
ACS Paragon Plus Environment

Page 23 of 42
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
with water (200 ml) followed by sat.NaHCO3 (100 ml). The ethereal solution was dried
(MgSO
4
), filtered and concentrated to a solid. The solid was slurried with isohexane (200 ml)
and collected by filtration to afford the title compound as a cream coloured solid (15.25 g,
4
1
4.8 mmol, 82%). 1H NMR (CDCl ): 8.00 (d, 1H), 7.66 (dd, 1H), 7.34 (dd, 1H), 5.07 (q,
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
+
H), 4.60 (d, 1H), 4.35 (d, 1H), 1.54 (d, 3H), 1.44 (s, 9H). LCꢀMS: [M+HꢀtBu] = 284/286.
tert-Butyl (2R)-2-[1-oxo-6-(4,4,5,5-tetramethyl-1,3,2-dioxaborolan-2-yl)-2,3-dihydro-1H-
isoindol-2-yl]propanoate (45)
A 25 mL round bottom flask was charged with (R)ꢀtertꢀbutyl 2ꢀ(6ꢀbromoꢀ1ꢀoxoisoindolinꢀ2ꢀ
yl)propanoate (9 g, 26.5 mmol), bis(pinacolato)diboron (8.06 g, 31.7 mmol), potassium
acetate (7.79 g, 79 mmol) and XPhos Pd G3 (0.448 g, 0.529 mmol). The system was
evacuated and backꢀfilled with nitrogen (x 4). Dioxane (70 mL) was added and the
evacuation/backꢀfilling procedure repeated (x 4). The reaction was heated to 90 °C and stirred
under nitrogen for 2h. After cooling to room temperature, the reaction was diluted with
EtOAc (10 mL) and filtered through Celite. The filtrate was concentrated in vacuo to give the
crude product as a pale brown solid. The crude product was purified by chromatography
2
(SiO , 80 g column, 0ꢀ50% EtOAc in isoꢀhexane) to afford the title compound as an off white
solid (9.6 g, 94%). 1H NMR (DMSOꢀd6): 7.95 (s, 1H), 7.89 (dd, 1H), 7.64 (dd, 1H), 4.79 (q,
+
1
H), 4.57 (d, 1H), 4.50 (d, 1H), 1.48 (d, 3H), 1.38 (s, 9H), 1.31 (s, 12H). LCMS: [M+H] =
32.
3
tert-Butyl (2R)-2-[6-(2,5-dichloropyrimidin-4-yl)-1-oxo-2,3-dihydro-1H-isoindol-2-
yl]propanoate (46)
2
2
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 24 of 42
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
A mixture of (R)ꢀtertꢀbutyl 2ꢀ(1ꢀoxoꢀ6ꢀ(4,4,5,5ꢀtetramethylꢀ1,3,2ꢀdioxaborolanꢀ2ꢀ
yl)isoindolinꢀ2ꢀyl)propanoate (5.0 g, 12.78 mmol), 2,4,5ꢀtrichloropyrimidine (3.52 g, 19.17
mmol) and 2.0 M Na
2 3
CO (aq.) (12.78 ml, 25.6 mmol) in dioxane (64 mL) was evacuated
and backꢀfilled with nitrogen (x 3). Pd(PPh ) (0.738 g, 0.639 mmol) was added and the
3
4
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
system was evacuated and backꢀfilled with nitrogen (x 3). The mixture was heated to 90 °C
and stirred for 6 h under nitrogen. The reaction was diluted with water (150 mL) and EtOAc
(150 mL). The phases were separated and the aqueous phase was extracted with EtOAc (2 x
120 mL). The combined organic extracts were washed with brine (200 mL), dried (MgSO
and concentrated. The crude product was adsorbed onto silica and purified by
4
)
chromatography (SiO , 220 g column, 0ꢀ30% EtOAc/isohexane) to afford (R)ꢀtertꢀbutyl 2ꢀ(6ꢀ
2
(2,5ꢀdichloropyrimidinꢀ4ꢀyl)ꢀ1ꢀoxoisoindolinꢀ2ꢀyl)propanoate as an off white solid (4.204 g,
77 %). 1H NMR (DMSOꢀd6): 9.03 (s, 1H), 8.14 (d, 1H), 8.06 (dd, 1H), 7.83 (dd, 1H), 4.82
+
(
q, 1H), 4.67 (d, 1H), 4.62 (s, 1H), 1.51 (d, 3H), 1.40 (s, 9H). LCMS: [M+Na] = 430.
(R)-2-(6-(2,5-Dichloropyrimidin-4-yl)-1-oxoisoindolin-2-yl)propanoic acid
trifluoroacetic acid salt (47)
Trifluoroacetic acid (25 ml, 324 mmol) was added dropwise to a solution of tertꢀbutyl (2R)ꢀ2ꢀ
[6ꢀ(2,5ꢀdichloropyrimidinꢀ4ꢀyl)ꢀ1ꢀoxoꢀ2,3ꢀdihydroꢀ1Hꢀisoindolꢀ2ꢀyl]propanoate (4.5 g, 11.02
mmol) in DCM (25 ml) and the mixture was stirred at room temperature for 1 h. The mixture
was concentrated in vacuo and azeotroped with toluene (3x50 ml). The solid was stirred and
2 2
sonicated in Et O and collected by fitration, washing with Et O. The solid was air dried
overnight to afford the title compound as a colourless solid (3.7g, 95%). The product was
used without further purification. 1H NMR (DMSOꢀd6): 12.99 (s, 1H), 9.04 (s, 1H), 8.13 (dd,
2
3
ACS Paragon Plus Environment

Page 25 of 42
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
1H), 8.06 (dd, 1H), 7.83 (dd, 1H), 4.87 (q, 1H), 4.67 (d, 1H), 4.61 (d, 1H), 1.54 (d, 3H). LCꢀ
MS: [M+H]+ = 352.
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
(R)-2-(6-{5-Chloro-2-[(oxan-4-yl)amino]pyrimidin-4-yl}-1-oxoisoindolin-2-yl)propanoic
acid (48)
(R)ꢀ2ꢀ(6ꢀ(2,5ꢀDichloropyrimidinꢀ4ꢀyl)ꢀ1ꢀoxoisoindolinꢀ2ꢀyl)propanoic acid (20g, 56.8
mmol), oxanꢀ4ꢀamine HCl (23.4g, 0.17mol, 3eq) and potassium hydroxide (2M aq., 114ml,
0
1
.227mol, 4eq) were charged to a reaction vessel with stirring. Heat was applied to 90°C for
9hrs. The reaction was cooled and hydrochloric acid (2M aq., 57ml) was added, causing the
immediate formation of a thick colourless precipitate. The slurry was filtered, and the filter
cake washed with water (50ml), then petrol (100ml, 200ml). The resulting solid was dried
under vacuum at 40°C to give the title compound as an off white solid (24.6g, quant.).
1H NMR (DMSOꢀd6): 12.95 (1H, br. s), 8.45 (1H, s), 8.03 (1H, s), 7.99 (1H, dd), 7.91 (1H,
br. s), 7.76 (1H, d), 7.60 (1H, s), 4.87 (1H, q), 4.61 (2H, dd), 3.99ꢀ3.80 (3H, m), 1.85 (2H, d),
1.61ꢀ1.45 (3H, m), 1.53 (3H, d). 1H obscured by solvent peaks.
2
4
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 26 of 42
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
AUTHOR INFORMATION
Corresponding Author
*T.D.H.
Tel: +44 (0) 1223 226270. Fax: +44 (0) 1223 226201. Email:
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
Tom.Heightman@astx.com
ACKNOWLEDGEMENTS
The authors wish to thank Dr Anne Cleasby for assistance in the refinement of Xꢀray crystal
structures, Mr Torren Peakman for confirmation of NMR assignments and Mr Stuart Whibley for
assistance in the LCꢀMS characterization and purification of compounds.
ABBREVIATIONS
BRAF, BꢀRaf protoꢀoncogene; DSF, differential scanning fluorimetry; ERK, extracellular
signalꢀrelated kinase; HATU, 1ꢀ[bis(dimethylamino)methylene]ꢀ1Iꢀ1,2,3ꢀtriazolo[4,5ꢀ
b]pyridinium 3ꢀoxid hexafluorophosphate; HBSS, Hanks’ Balanced Salt Solution; HBTU,
N,N,N′,N′ꢀtetramethylꢀOꢀ(1Hꢀbenzotriazolꢀ1ꢀyl)uronium
hexafluorophosphate;
HPMC,
hydroxypropyl methylcellulose; MAPK, mitogenꢀactivated protein kinase; MEK, mitogen
activated protein kinase kinase; MSD, meso scale discovery; RAS, retrovirus associated
DNA sequence; RSK, ribosomal S6 kinase; T3P, 2,4,6ꢀtripropylꢀ1,3,5,2,4,6ꢀ
trioxatriphosphorinaneꢀ2,4,6ꢀtrioxide;
TBTU,
2ꢀ(1Hꢀbenzotriazoleꢀ1ꢀyl)ꢀ1,1,3,3ꢀ
tetramethyluronium tetrafluoroborate; TRF, timeꢀresolved fluorescence.
2
5
ACS Paragon Plus Environment

Page 27 of 42
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
ASSOCIATED CONTENT
PDB ID Codes
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
Coordinates and structure factors for the ERK2ꢀligand complex have been deposited with
PDB with the following accession codes: 1, 6gdq; 4, 6gdm; 5, 6g92; 6, 6g8x; 10, 6g91; 11,
6g93; 13, 6g97; 14, 6g9a; 15, 6g9d; 16, 6g9h; 18, 6g9j; 20, 6g9m; 21, 6geo; 22, 6g9k; 27,
6g9n. Authors will release the atomic coordinates and experimental data upon article
publication.
SUPPORTING INFORMATION
The following Figures and Tables are provided as Supporting Information: S2 –
Supplementary Figure 1 – Structure of ERK1/2 inhibitor Vertex 11e. S3 – Supplementary
Figure 2 ꢀ ERK2 coꢀstructures of methoxyethyl analogue 13 and morpholinoethyl analogue
14. S4 – Supplementary Figure 3 ꢀ Effects of compound treatment on the body weight of
mice bearing Colo205 tumor xenografts. S5 – Supplementary Table 1 ꢀ Kinome selectivity
profile of Compound 27. S6 – Additional synthetic procedures for compounds 5-26 and
associated intermediates. Molecular formula strings are provided as a csv file.
2
6
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 28 of 42
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
REFERENCES
1.
Roskoski Jr, R. ERK1/2 MAP Kinases: Structure, Function, and Regulation. Pharmacological
Research 2012, 66 (2), 105ꢀ143.
Prior, I. A.; Lewis, P. D.; Mattos, C. A Comprehensive Survey of Ras Mutations in Cancer.
Cancer Research 2012, 72 (10), 2457ꢀ2467.
Caunt, C. J.; Sale, M. J.; Smith, P. D.; Cook, S. J. MEK1 and MEK2 Inhibitors and Cancer
Therapy: the Long and Winding Road. Nat Rev Cancer 2015, 15 (10), 577ꢀ592.
Corcoran, R. B.; Settleman, J.; Engelman, J. A. Potential Therapeutic Strategies to Overcome
Acquired Resistance to BRAF or MEK Inhibitors in BRAF Mutant Cancers. Oncotarget 2011, 2 (4),
2.
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
36ꢀ346.
Little, A. S.; Smith, P. D.; Cook, S. J. Mechanisms of Acquired Resistance to ERK1/2
5.
Pathway Inhibitors. Oncogene 2012, 32, 1207.
6
7
.
.
Lloyd, A. C. Distinct Functions for ERKs? Journal of Biology 2006, 5 (5), 13.
Qin, J.; Xin, H.; Nickoloff, B. J. Specifically Targeting ERK1 or ERK2 Kills Melanoma
Cells. Journal of Translational Medicine 2012, 10 (1), 15.
Shin, M.; Franks, C. E.; Hsu, K.ꢀL. IsoformꢀSelective ActivityꢀBased Profiling of ERK
Signaling. Chemical Science 2018, 9 (9), 2419ꢀ2431.
Lefloch, R.; Pouysségur, J.; Lenormand, P. Single and Combined Silencing of ERK1 and
8.
9.
ERK2 Reveals Their Positive Contribution to Growth Signaling Depending on Their Expression
Levels. Molecular and Cellular Biology 2008, 28 (1), 511ꢀ527.
1
0.
Functions of Extracellular Signal–Regulated Kinases. Science Signaling 2011, 4 (196), re3.
1. Germann, U.; Furey, B.; Roix, J.; Markland, W.; Hoover, R.; Aronov, A.; Hale, M.; Chen, G.;
Rodríguez, J.; Crespo, P. Working Without Kinase Activity: PhosphotransferꢀIndependent
1
MartinezꢀBotella, G.; Alargova, R.; Fan, B.; Sorrell, D.; Meshaw, K.; Shapiro, P.; Wick, M. J.; Benes,
C.; Garnett, M.; DeCrescenzo, G.; Namchuk, M.; Saha, S.; Welsch, D. J. Abstract 4693: The Selective
ERK Inhibitor BVDꢀ523 is Active in Models of MAPK PathwayꢀDependent Cancers, Including
Those with Intrinsic and Acquired Drug Resistance. Cancer Research 2015, 75 (15 Supplement),
4
1
693ꢀ4693.
2. Blake, J. F.; Burkard, M.; Chan, J.; Chen, H.; Chou, K.ꢀJ.; Diaz, D.; Dudley, D. A.; Gaudino,
J. J.; Gould, S. E.; Grina, J.; Hunsaker, T.; Liu, L.; Martinson, M.; Moreno, D.; Mueller, L.; Orr, C.;
Pacheco, P.; Qin, A.; Rasor, K.; Ren, L.; Robarge, K.; ShahidiꢀLatham, S.; Stults, J.; Sullivan, F.;
Wang, W.; Yin, J.; Zhou, A.; Belvin, M.; Merchant, M.; Moffat, J.; Schwarz, J. B. Discovery of (S)ꢀ1ꢀ
(
1ꢀ(4ꢀChloroꢀ3ꢀfluorophenyl)ꢀ2ꢀhydroxyethyl)ꢀ4ꢀ(2ꢀ((1ꢀmethylꢀ1Hꢀpyrazolꢀ5ꢀyl)amino)pyrimidinꢀ4ꢀ
yl)pyridinꢀ2(1H)ꢀone (GDCꢀ0994), an Extracellular SignalꢀRegulated Kinase 1/2 (ERK1/2) Inhibitor
in Early Clinical Development. Journal of Medicinal Chemistry 2016, 59 (12), 5650ꢀ5660.
1
3.
Advanced Malignancies (MKꢀ8353ꢀ013). www.clintrials.gov NCT02972034, April 26 , 2017.
4. A Study of LY3214996 Administered Alone or in Combination With Other Agents in
Participants With Advanced/Metastatic Cancer. www.clintrials.gov NCT02857270, April 26 , 2017.
5. A Phase I Clinical Study With Investigational Compound LTT462 in Adult Patients With
Specific Advanced Cancers. www.clintrials.gov NCT02711345, April 26 , 2017.
6. FirstꢀinꢀHuman Study of KOꢀ947 in NonꢀHematological Malignancies. www.clintrials.gov
NCT03051035, April 26 , 2017.
7. Aronov, A. M.; Baker, C.; Bemis, G. W.; Cao, J.; Chen, G.; Ford, P. J.; Germann, U. A.;
Study of MKꢀ8353 in Combination With Pembrolizumab (MKꢀ3475) in Participants With
th
1
th
1
th
1
th
1
Green, J.; Hale, M. R.; Jacobs, M.; Janetka, J. W.; Maltais, F.; MartinezꢀBotella, G.; Namchuk, M. N.;
Straub, J.; Tang, Q.; Xie, X. Flipped Out:ꢂ StructureꢀGuided Design of Selective Pyrazolylpyrrole
ERK Inhibitors. Journal of Medicinal Chemistry 2007, 50 (6), 1280ꢀ1287.
18.
Aronov, A. M.; Tang, Q.; MartinezꢀBotella, G.; Bemis, G. W.; Cao, J.; Chen, G.; Ewing, N.
P.; Ford, P. J.; Germann, U. A.; Green, J.; Hale, M. R.; Jacobs, M.; Janetka, J. W.; Maltais, F.;
Markland, W.; Namchuk, M. N.; Nanthakumar, S.; Poondru, S.; Straub, J.; ter Haar, E.; Xie, X.
StructureꢀGuided Design of Potent and Selective Pyrimidylpyrrole Inhibitors of Extracellular Signalꢀ
Regulated Kinase (ERK) Using Conformational Control. Journal of Medicinal Chemistry 2009, 52
(
20), 6362ꢀ6368.
2
7
ACS Paragon Plus Environment

Page 29 of 42
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
1
9.
Deng, Y.; Shipps, G. W.; Cooper, A.; English, J. M.; Annis, D. A.; Carr, D.; Nan, Y.; Wang,
T.; Zhu, H. Y.; Chuang, C.ꢀC.; Dayananth, P.; Hruza, A. W.; Xiao, L.; Jin, W.; Kirschmeier, P.;
Windsor, W. T.; Samatar, A. A. Discovery of Novel, Dual Mechanism ERK Inhibitors by Affinity
Selection Screening of an Inactive Kinase. Journal of Medicinal Chemistry 2014, 57 (21), 8817ꢀ8826.
20.
Bagdanoff, J. T.; Jain, R.; Han, W.; Zhu, S.; Madiera, A.ꢀM.; Lee, P. S.; Ma, X.; Poon, D.
Tetrahydropyrroloꢀdiazepenones as Inhibitors of ERK2 Kinase. Bioorganic & Medicinal Chemistry
Letters 2015, 25 (18), 3788ꢀ3792.
21.
Lim, J.; Kelley, E. H.; Methot, J. L.; Zhou, H.; Petrocchi, A.; Chen, H.; Hill, S. E.; Hinton, M.
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.; Hruza, A.; Jung, J. O.; Maclean, J. K. F.; Mansueto, M.; Naumov, G. N.; Philippar, U.; Raut, S.;
Spacciapoli, P.; Sun, D.; Siliphaivanh, P. Discovery of 1ꢀ(1HꢀPyrazolo[4,3ꢀc]pyridinꢀ6ꢀyl)urea
Inhibitors of Extracellular SignalꢀRegulated Kinase (ERK) for the Treatment of Cancers. Journal of
Medicinal Chemistry 2016, 59 (13), 6501ꢀ6511.
22.
Lechtenberg, B. C.; Mace, P. D.; Sessions, E. H.; Williamson, R.; Stalder, R.; Wallez, Y.;
Roth, G. P.; Riedl, S. J.; Pasquale, E. B. StructureꢀGuided Strategy for the Development of Potent
Bivalent ERK Inhibitors. ACS Medicinal Chemistry Letters 2017, 8 (7), 726ꢀ731.
23.
Ward, R. A.; Bethel, P.; Cook, C.; Davies, E.; Debreczeni, J. E.; Fairley, G.; Feron, L.;
Flemington, V.; Graham, M. A.; Greenwood, R.; Griffin, N.; Hanson, L.; Hopcroft, P.; Howard, T.
D.; Hudson, J.; James, M.; Jones, C. D.; Jones, C. R.; Lamont, S.; Lewis, R.; Lindsay, N.; Roberts,
K.; Simpson, I.; StꢀGallay, S.; Swallow, S.; Tang, J.; Tonge, M.; Wang, Z.; Zhai, B. StructureꢀGuided
Discovery of Potent and Selective Inhibitors of ERK1/2 from a Modestly Active and Promiscuous
Chemical Start Point. Journal of Medicinal Chemistry 2017, 60 (8), 3438ꢀ3450.
24.
Li, B. T.; Janku, F.; Patel, M. R.; Sullivan, R. J.; Flaherty, K.; Buchbinder, E. I.; Lacouture,
M. E.; Varghese, A. M.; Wong, D. J. L.; Sznol, M.; Sosman, J. A.; Keedy, V. L.; WangꢀGillam, A.;
Ribas, A.; Tolcher, A. W.; Patel, S. P.; Varterasian, M. L.; Welsch, D.; Hyman, D. M.; Infante, J. R.
FirstꢀinꢀClass Oral ERK1/2 Inhibitor Ulixertinib (BVDꢀ523) in Patients with Advanced Solid Tumors:
Final Results of a Phase I Dose Escalation and Expansion Study. Journal of Clinical Oncology 2017,
35 (15_suppl), 2508ꢀ2508.
25. Moschos, S. J.; Sullivan, R. J.; Hwu, W.ꢀJ.; Ramanathan, R. K.; Adjei, A. A.; Fong, P. C.;
ShapiraꢀFrommer, R.; Tawbi, H. A.; Rubino, J.; Rush, T. S., III; Zhang, D.; Miselis, N. R.; Samatar,
A. A.; Chun, P.; Rubin, E. H.; Schiller, J.; Long, B. J.; Dayananth, P.; Carr, D.; Kirschmeier, P.;
Bishop, W. R.; Deng, Y.; Cooper, A.; Shipps, G. W.; Moreno, B. H.; Robert, L.; Ribas, A.; Flaherty,
K. T. Development of MKꢀ8353, an Orally Administered ERK1/2 Inhibitor, in Patients with
Advanced Solid Tumors. JCI Insight 2018, 3 (4).
26.
Morris, E. J.; Jha, S.; Restaino, C. R.; Dayananth, P.; Zhu, H.; Cooper, A.; Carr, D.; Deng, Y.;
Jin, W.; Black, S.; Long, B.; Liu, J.; DiNunzio, E.; Windsor, W.; Zhang, R.; Zhao, S.; Angagaw, M.
H.; Pinheiro, E. M.; Desai, J.; Xiao, L.; Shipps, G.; Hruza, A.; Wang, J.; Kelly, J.; Paliwal, S.; Gao,
X.; Babu, B. S.; Zhu, L.; Daublain, P.; Zhang, L.; Lutterbach, B. A.; Pelletier, M. R.; Philippar, U.;
Siliphaivanh, P.; Witter, D.; Kirschmeier, P.; Bishop, W. R.; Hicklin, D.; Gilliland, D. G.; Jayaraman,
L.; Zawel, L.; Fawell, S.; Samatar, A. A. Discovery of a Novel ERK Inhibitor with Activity in Models
of Acquired Resistance to BRAF and MEK Inhibitors. Cancer Discovery 2013, 3 (7), 742ꢀ750.
27.
Chaikuad, A.; M C Tacconi, E.; Zimmer, J.; Liang, Y.; Gray, N. S.; Tarsounas, M.; Knapp, S.
A Unique Inhibitor Binding Site in ERK1/2 is Associated with Slow Binding Kinetics. Nat Chem Biol
2014, 10 (10), 853ꢀ860.
28. Rudolph, J.; Xiao, Y.; Pardi, A.; Ahn, N. G. Slow Inhibition and Conformation Selective
Properties of Extracellular SignalꢀRegulated Kinase 1 and 2 Inhibitors. Biochemistry 2015, 54 (1), 22ꢀ
3
2
1.
9.
Erlanson, D. A.; Fesik, S. W.; Hubbard, R. E.; Jahnke, W.; Jhoti, H. Twenty Years On: The
Impact of Fragments on Drug Discovery. Nature Reviews Drug Discovery 2016, 15, 605.
0. Johnson, C. N.; Erlanson, D. A.; Jahnke, W.; Mortenson, P. N.; Rees, D. C. FragmentꢀtoꢀLead
Medicinal Chemistry Publications in 2016. Journal of Medicinal Chemistry 2018, 61 (5), 1774ꢀ1784.
1. Squires, M. S.; Feltell, R. E.; Wallis, N. G.; Lewis, E. J.; Smith, D.ꢀM.; Cross, D. M.; Lyons,
3
3
J. F.; Thompson, N. T. Biological Characterization of AT7519, a SmallꢀMolecule Inhibitor of Cyclinꢀ
Dependent Kinases, in Human Tumor Cell Lines. Molecular Cancer Therapeutics 2009, 8 (2), 324ꢀ
332.
2
8
ACS Paragon Plus Environment