Rational Design of 5-(4-(Isopropylsulfonyl)phenyl)-3-(3-(4-((methylamino)methyl)phenyl)isoxazol-5-yl)pyrazin-2-amine (VX-970, M6620): Optimization of Intra- and Intermolecular Polar Interactions of a New Ataxia Telangiectasia Mutated and Rad3-Related (ATR) Kinase Inhibitor

Article abstract of DOI:10.1021/acs.jmedchem.9b00426

The DNA damage response (DDR) is a DNA damage surveillance and repair mechanism that can limit the effectiveness of radiotherapy and DNA-damaging chemotherapy, commonly used treatment modalities in cancer. Two related kinases, ataxia telangiectasia mutated (ATM) and ATM and Rad3-related kinase (ATR), work together as apical proteins in the DDR to maintain genome stability and cell survival in the face of potentially lethal forms of DNA damage. However, compromised ATM signaling is a common characteristic of tumor cells, which places greater reliance on ATR to mediate the DDR. In such circumstances, ATR inhibition has been shown to enhance the toxicity of DNA damaging chemotherapy to many cancer cells in multiple preclinical studies, while healthy tissue with functional ATM can tolerate ATR inhibition. ATR therefore represents a very attractive anticancer target. Herein we describe the discovery of VX-970/M6620, the first ATR inhibitor to enter clinical studies, which is based on a 2-aminopyrazine core first reported by Charrier et al. (J. Med. Chem. 2011, 54, 2320-2330, DOI: 10.1021/jm101488z).

Full text of DOI:10.1021/acs.jmedchem.9b00426

Subscriber access provided by UNIV AUTONOMA DE COAHUILA UADEC
Article
Rational design of 5-(4-(isopropylsulfonyl)phenyl)-3-(3-(4-
(methylamino)methyl)phenyl)isoxazol-5-yl)pyrazin-2-amine (VX-970,
(
M6620): Optimization of intra- and inter-molecular polar interactions of a
new ataxia telangiectasia mutated and Rad3-related (ATR) kinase inhibitor
Ronald M. A. Knegtel, Jean-Damien Charrier, Steven John Durrant, Chris Davis, Michael O'Donnell,
Pierre Storck, Somhairle MacCormick, David Kay, Joanne Pinder, Anisa Virani, Heather Twin,
Matthew Griffiths, Philip Reaper, Peter Littlewood, Steve Young, Julian M Golec, and John Pollard
J. Med. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.jmedchem.9b00426 • Publication Date (Web): 10 May 2019
Downloaded from http://pubs.acs.org on May 10, 2019
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 73
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Rational design of 5-(4-(isopropylsulfonyl)phenyl)-3-(3-(4-
((methylamino)methyl)phenyl)isoxazol-5-yl)pyrazin-2-amine
(VX-970, M6620): Optimization of intra- and inter-molecular
polar interactions of a new ataxia telangiectasia mutated and
Rad3-related (ATR) kinase inhibitor
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
*
Ronald Knegtel , Jean-Damien Charrier, Steven Durrant, Chris Davis, Michael
†
O’Donnell, Pierre Storck, Somhairle MacCormick , David Kay, Joanne Pinder,
Anisa Virani, Heather Twin, Matthew Griffiths, Philip Reaper, Peter Littlewood,
‡
Steve Young , Julian Golec, John Pollard
Vertex Pharmaceuticals (Europe) Ltd, 86-88 Jubilee Avenue, Milton Park,
Abingdon, Oxfordshire OX14 4RW, United Kingdom
†Present Address: PharmaCytics Operations BV, Novio Tech Campus, Office 3.12,
Transistorweg 5v, 6534 AT Nijmegen, The Netherlands
‡Present Address: Sygnature Discovery Limited, BioCity, Pennyfoot Street,
Nottingham, NG1 1GF, United Kingdom
ABSTRACT
The DNA damage response (DDR) is a DNA damage surveillance and repair
mechanism that can limit the effectiveness of radiotherapy and DNA-damaging
chemotherapy, commonly used treatment modalities in cancer. Two related
1
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 2 of 73
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
kinases, ataxia telangiectasia mutated (ATM) and ATM and Rad3-related kinase
(
ATR), work together as apical proteins in the DDR to maintain genome stability
and cell survival in the face of potentially lethal forms of DNA damage. However,
compromised ATM signaling is a common characteristic of tumor cells, which
places greater reliance on ATR to mediate the DDR. In such circumstances, ATR
inhibition has been shown to enhance the toxicity of DNA damaging
chemotherapy to many cancer cells in multiple pre-clinical studies, while healthy
tissue with functional ATM can tolerate ATR inhibition. ATR therefore,
represents a very attractive anti-cancer target. Herein we describe the discovery
of VX-970/M6620, the first ATR inhibitor to enter clinical studies, which is based
on a 2-aminopyrazine core first reported by Charrier et al. 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
2
ACS Paragon Plus Environment

Page 3 of 73
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
INTRODUCTION
DNA-damaging agents, such as cisplatin, etoposide, gemcitabine and ionizing
radiation (IR) are important and commonly used agents in cancer therapy. Their
mechanism of action is to induce potentially lethal DNA damage including double
strand breaks and replication stress (RS). RS commonly arises when the
replication machinery encounters a lesion in the DNA, resulting in a stalled and
unstable replication fork that in turn can collapse to form a double strand break.
The effectiveness of DNA damaging chemotherapy can be limited by the cell’s
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
ability to detect and repair the DNA damage. The related kinases ataxia
telangiectasia mutated (ATM) and Rad3-related (ATR) kinase are apical
regulators of the DNA damage response (DDR), a surveillance and repair process
that enables cells to survive double strand breaks and replication stress.
Although each kinase recognizes different DNA damage structures (double
strand breaks in the case of ATM and single stranded DNA within double
stranded DNA in the case of ATR) they share many downstream substrates and
have overlapping roles in maintaining cell survival. ATR and ATM share a high
degree of sequence homology and are members of the phosphoinositol 3-kinase-
3
like kinase (PIKK) family of serine/threonine protein kinases , which includes
other important DNA repair kinases such as DNA-dependent protein kinase
(
DNA-PK). Despite the importance of the DDR for cell survival following DNA
damage, ATM function is often found to be impaired in tumor cells (from defects
in activation, expression or downstream signaling proteins such as p53). Loss of
ATM or its effectors is thought to enable a genomically unstable environment,
4
which is advantageous to the developing tumor. However, impaired ATM
3
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 4 of 73
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
function places additional burden on remaining repair proteins, such as ATR, to
maintain survival. This Achilles’ heel reliance on ATR may be exploited with
drugs. Consistent with this hypothesis, a series of preclinical studies have
demonstrated that cancer cells treated with DNA damaging drugs are especially
sensitive to ATR inhibition when they harbor defects in ATM signalling.5-7 In
contrast, normal tissue can tolerate ATR inhibition by activating an ATM-
0
1
2
3
4
5
6
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
mediated compensatory DDR. Furthermore, several studies have shown that
some tumors are sensitive to ATR inhibition as a single agent, and to
combinations of ATR inhibitors with inhibitors of other repair proteins such as
poly (ADP-ribose) polymerase (PARP) and Chk1. As such, ATR inhibition
represents an attractive anti-cancer target.5,8-20
The first reported series of potent and selective ATR inhibitors, typified by
compound 1 (VE-821, Figure 1) is based on a 2-aminopyrazine amide hinge
1
binding motif. Compound 1 has proven to be a useful tool to assess the biology
of ATR in cells and the potentially beneficial profile for ATR inhibition. However,
certain properties of 1 were not fully optimized in the initial chemistry
campaign. These include enzyme and cell potency, physical properties such as
solubility and the metabolic profile, specifically the potential to form aniline
1
metabolites. Here we present the optimization of the 2-aminopyrazine series to
a more potent and stable hetero-aromatic scaffold from which compound 2 (VX-
70, M6620, Figure 1) the first ATR inhibitor to enter clinical studies21 is
9
derived.
4
ACS Paragon Plus Environment

Page 5 of 73
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
N
NH O
NH2
O
2
N
N
H
N
NHMe
N
N
Compound 1
VE-821
Compound 2
VX-970/M6620
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
SO Me
SO iPr
2
2
Figure 1. Chemical structures of compound 1 (VE-821) and compound 2 (VX-
70/M6620).
9
In our initial optimization efforts of 1, replacement of the anilide group with a
range of fused bicyclics such as benzimidazole, benzoxazole, benzothiazole and
1
indole was well tolerated by ATR but resulted in reduced selectivity for ATR
against ATM. The loss of selectivity over ATM was attributed to a rotation of the
bound inhibitor (Figure 2A) caused by a steric clash between the bulkier fused
bicyclic aromatic systems with the Tyrosine gatekeeper residue of the PIKK
family. This movement in turn resolves a second steric clash between Pro2775 of
ATM (which is a smaller Glycine residue in ATR) and the isopropylsulfone of the
inhibitor, leading to tighter binding to ATM.
5
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 6 of 73
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
Figure 2a-b. (A) Comparison of anilide 1 (in green) versus a benzimidazole
isostere (in cyan) in a model of the active site of ATM based on PDB entry 1E7V.
The benzimidazole is thought to experience steric hindrance from Tyr2755 in
ATM. In response it rotates away from Tyr2755, which allows it to escape a
steric clash with Pro2775 (Gly in ATR). (B) Replacing the benzimidazole with a
phenyl substituted monocyclic heteroaryl (in cyan) yields a better mimic of the
original anilide and avoids movement of the inhibitor due to steric hindrance
with Tyr2755.
0
1
2
3
4
5
6
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
Based on homology models (derived from a crystal structure of the homologous
1
kinase PI3K, PDB entry 1E7V) we hypothesized that phenyl substituted 5-
membered heteroaromatic groups can bind without causing steric clashes with
the gatekeeper residue as it mimics the shape of the original anilide more closely.
Thus we expected selectivity for ATR over ATM to be retained (Figure 2B).
Furthermore, we speculated that functionalization of the phenyl ring would
enable exploitation of a highly negatively charged area of the ATP binding site.
This would have potential to increase potency and selectivity and also provide an
opportunity to modulate physical properties to improve solubility. In keeping
with this, substitution with a 4-benzylamine on the 5-membered heteroaryl core
yielded compounds with picomolar ATR inhibition, good retention of selectivity
and potent cellular activity, culminating in the discovery of the clinical candidate
2 (Figure 1).
6
ACS Paragon Plus Environment

Page 7 of 73
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
RESULTS
The commercially available methyl 3-amino-6-bromopyrazine-2-carboxylate 21
was readily converted into the 4-isopropylsulfonylphenyl using Suzuki cross-
coupling conditions to provide 22.1,22,23 The carboxylic acid in 22 was then
transformed into the anilide 324 or used in a sequence leading to 5-
phenylisoxazole 7, via formation of a Weinreb amide, methyl ketone or
benzylidene ketone intermediate.25 Acid 22 was also engaged in a
cyclodehydration with N-hydroxybenzamidine to furnish 3-phenyl-1,2,4-
oxadiazole 10.26
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
Compound 25 was used as a versatile intermediate in the preparation of 1-
phenyl-1,2,3-triazole 6 and 3-phenylisoxazole 4 (as well as analogues 2, 14-20)
(
Scheme 2). Commercially available 5-bromopyrazin-2-amine 23 was subjected
to Suzuki palladium cross-coupling with 4-isopropylsulfonylphenyl boronic acid
and then brominated at position 3 using NBS to provide intermediate 24 in high
2
7
28
yield. Copper assisted Sonogashira coupling with TMS-acetylene , followed by
di-BOC protection of the amino group and TMS deprotection afforded advanced
intermediate 25. The acetylene functional group in 25 was then used as a
coupling partner for click chemistry with azidobenzene to lead to the 1,2,3-
triazole 6 after BOC removal.29 [3+2] Cycloaddition between 25 and
phenylnitrile oxides, generated in situ from the corresponding N-hydroxyimidoyl
chlorides, delivered the 3-aryl isoxazoles 4 (as well as 2 and 14-20 after
additional steps, see experimental).30
7
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 8 of 73
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
Compound 24 was also used as a coupling partner to 2- and 5-phenyloxazoles
under CH insertion conditions to produce inhibitors 11 and 12 respectively
(Scheme 3).31-33 The 2-vinyl intermediate, 26, prepared from the bromo
0
1
2
3
4
5
6
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
precursor, 24, and vinyl potassium trifluoroborate under palladium catalysis34,
underwent [3+2] cycloaddition with phenylnitrile oxide, generated in situ from
the corresponding N-hydroxyimidoyl chloride, to provide the 3-phenyl-4,5-
dihydro isoxazole 5.35
The carboxylic acid group in the commercial starting point, 27, was used to build
the 1,3,4-oxadiazole and the 1,3,4-thiadiazole rings in intermediates 28 and 29,
via cyclodehydration with benzohydrazide and benzothiohydrazide, respectively
36
(
Scheme 4). These intermediates were then functionalized using Suzuki cross-
couplings to afford 2-phenyl-1,3,4-oxadiazole 8 and 2-phenyl-1,3,4-thiadiazole
3. Using a similar strategy, building block 30 was converted into the 5-phenyl-
,2,4-oxadiazole, 9. The nitrile at position 2 in 30 was transformed into a
1
1
hydroxyimidamide with hydroxylamine and subsequently O-benzoylated prior
to a cyclodehydration under acidic conditions.37
8
ACS Paragon Plus Environment

Page 9 of 73
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Scheme 1a
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
a
Reagents and conditions: (a) (4-isopropylsulfonylphenyl)boronic acid, K
O (4:1), 60 C, 2 h, 97%; (b) LiOH, THF, rt, 2 h, 96%; (c) diethoxyphosphorylformonitrile,
aniline, Et N, DME, 120 C, 18 h, 96%; (d) N-methoxymethanamine hydrochloride, HOBt, DIPEA, 3-
ethyliminomethylene amino)-N,N-dimethyl-propan-1-amine, THF, RT, 18 h, 77%; (e) MeMgBr 3 M
3 4 3 2
PO , Pd[P(tBu) ] ,
MeCN:H
2
3
(
in THF, THF, -20 C, 1 h, 80%; (f) benzaldehyde, NaOH, MeOH:DCM (1:1), rt, 18 h, 70%; (g)
hydroxylamine hydrochloride, NaOAc, EtOH/AcOH (10/3), 130 C, 1 h, 17%; (h) CDI, N-
hydroxybenzamidine, RT to 120 C, 5 h, 67%.
Scheme 2a
a
Reagents and conditions: (a) (4-isopropylsulfonylphenyl)boronic acid, K
MeCN/H O (4/1), 60 C, 1 h, 87%; (b) NBS, DMF, rt, 3 h, 84 %; (c) (trimethylsilyl)acetylene, CuI,
Et N, Pd(PPh , DMF, 0 C to rt, 1 h, 81%; (d) Boc O, DMAP, Et N, DCM, RT, 18 h, 100%; (e)
3 4 3 2
PO , Pd[P(tBu) ] ,
2
3
3
)
4
2
3
9
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 10 of 73
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
Na
CuSO
2
CO
3
, MeOH, RT, 1 h, 97%; (f) azidobenzene (0.5M in MTBE), (+)-sodium L-ascorbate,
4
.5H
2 2
O, t-BuOH: H O (1:1), rt to 40 C, 12 h, 41%; (g) TFA/DCM (1/3), rt, 1h, 80%; (h) N-
hydroxybenzimidoylchloride, Et N, DMF, 60 C, 45 min then TFA/DCM, rt, 1 h, 52%.
3
Scheme 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
a
Reagents and conditions: (a) 2-phenyloxazole, Ag
, dioxane, 140 °C in wave, 3 h, 46%; (c) vinyl
N, n-PrOH, 100 °C in wave, 40 min, 57%; (d) N-
2 3 3
CO , PPh , Pd(dppf)Cl2.DCM, water, 70°C, 24
h, 29%; (b) 5-phenyloxazole, t-BuOLi, Pd(PPh
potassium trifluoroborate, Pd(dppf)Cl2, Et
3 4
)
3
hydroxybenzimidoyl chloride, Et
3
N, DMF, 65 °C, 1h, 58%.
1
0
ACS Paragon Plus Environment

Page 11 of 73
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Scheme 4a
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
a
Reagents and conditions: (a) benzohydrazide, PPh
Br , MeCN, RT, 3 h, 39%; (c) hydroxylamine hydrochloride, Et
MeOH, 0 C to RT, 3 h, 76%; (d) benzoyl chloride, Et N, DCM, RT, 16 h, 93%; (e) PPA, 70 C, 7 h,
88%; (f) (4-isopropylsulfonylphenyl) boronic acid, Na CO , Pd(dppf)Cl .DCM, dioxane, 100 C in
wave, 25 min, 34-69%.
3
Br
2
, MeCN, rt, 1 h, then DIPEA, 0 °C, 1 h, 80%;
(b) benzothiohydrazide, PPh
3
2
3
N,
3
3
3
2

The preparation of intermediates 22, 24-26, 28-29, 31 is described in the
Supporting Information.
DISCUSSION
The main aim of the optimization of the 2-aminopyrazine series was to identify a
more potent and water soluble analog of 1 without the potential to form an
aniline metabolite. The primary approach was to mimic the anilide of 1 with an
optimally functionalized saturated or unsaturated ring.
1
1
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 12 of 73
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
In our earlier work, a 3-fold gain in ATR inhibitory potency was achieved by
1
replacing the methylsulfone of 1 with an isopropyl sulfone. Therefore
compound 3 (Table 1) was used as a more potent starting point for exploring
biaryl replacements of the anilide. Docking analogues of 3 with 5- and 6-
membered heteroaromatic replacements for the anilide into an ATR homology
0
1
2
3
4
5
6
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
model suggested that phenyl substituted 5-membered heteroaromatics would
fit well in the ATP binding site and mimic the shape and planar nature of the
original anilide (Figure 2B). Previously reported initial studies had established
that the amide of 1 linking the phenyl and amino pyrazine was not making any
1
specific H-bonding interactions with the ATP binding site. However, it had not
been determined whether replacement linkers needed to retain a flat profile or
whether a degree of 3-dimensionality could be accommodated. To test this we
prepared a matched pair with either an unsaturated and planar isoxazole (4) or
the partially saturated 4,5-dihydroisoxazole analog (5, Table 1). Compound 4
was approximately 90-fold more potent than 5, indicating that a flat linker is
required for an optimal steric fit in the ATP binding pocket. The energy required
to flatten the 4,5-dihydroisoxazole ring system of 5 (R-isomer) from its
minimized, more 3-dimensional conformation in vacuum to its docked
conformation is estimated to be approximately 2 kcal/mol in the gas phase. This
would equate to about a 30-fold reduction of its inhibition constant, which is
consistent with the observed data.
1
2
ACS Paragon Plus Environment

Page 13 of 73
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Table 1. Enzyme inhibition constants (Ki) and cell assay IC50 values for
aminopyrazine ATR inhibitors. Compound 5 is a racemic mixture.
NH2
R
N
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
N
SO iPr
2
Cellular
Enzyme inhibition Ki (M) potency
Compounds
R
IC50 (M)
ATR
ATM DNA-PK pH2AX
O
3
0.004
4
2.8
1.03
N
H
N
O
4
5
0.001
0.090
1.4
>4
>4
>4
1.1
N
O
>2.5
A selection of diverse 5-membered aromatic heterocycles were considered as
possible amide replacements and were prioritized by assessing their
conformational energies (as outlined below), followed by docking into our ATR
1
kinase homology model. As a consequence of conjugation, the 5-membered
heteroayl ring can assume only two orientations with respect to the 2-
aminopyrazine hinge binder, as illustrated in Figure 3. We will refer to the
orientation that best mimics the anilide of 1 as the bioactive conformation.
1
3
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 14 of 73
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
H
N
H
N
H
N
H
N
Z
N
X
N
Y
Z
Y
X
0
1
2
3
4
5
6
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
Bioactive
conformation
Alternative
conformation
SO iPr
SO iPr
2
2
Figure 3. Intra-molecular polar interactions within heteroaryl ATR inhibitors.
Polar interactions between the central 5-membered heteroaryl ring atoms X and
Y and the 2-aminopyrazine hinge binder determine whether the bioactive
conformation (left) is energetically preferred.
Complementary intra-molecular interactions between the heteroaryl rings of the
inhibitor would assure that the proposed bioactive conformation is energetically
favored. Atoms X and Y in the central ring are either opposite the 2-amino group
2
or the 4-sp nitrogen atom of the pyrazine ring and the nature of their mutual
interactions determines the preferred orientation of the 5-membered heteroaryl
ring. While hydrogen bond acceptors (atom X) are preferred opposite the
pyrazine NH and an aromatic CH (atom Y) opposite the pyrazine ring sp2
2
nitrogen for reasons of electrostatic complementarity, it is less clear which
conformation is preferred when both X and Y are hydrogen bond acceptors.
Aromatic nitrogen, oxygen and sulfur atoms vary in their effectiveness as
38
hydrogen bond acceptors and will be tolerated to different degrees opposite
2
the sp nitrogen at position 4 of the pyrazine ring. Therefore the nature and
1
4
ACS Paragon Plus Environment

Page 15 of 73
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
shape of the 5-membered aryl ring determine how readily a compound can
assume the proposed bioactive conformation. In order to predict which 5-
membered heteroaryl rings favor the proposed bioactive conformation, the gas
phase energies of the two accessible conformations shown in Figure 3 from a
library of putative heteroaryl rings were calculated using Density Functional
Theory (DFT) using the B3LYP functional and a 6-31G** basis set.38-41 Although
calculated rotational barriers can give insight into how readily a certain
conformation can be assumed, they provide no information on which
conformation is more stable and tracked poorly with the binding data when
calculated for compounds 8, 9 and 10 (results not shown). Calculating the
enthalpy difference between two rigid conformers is both computationally less
demanding and a better estimator of the free energy of the binding state
function. Nine phenyl-substituted 5-membered heteroaryl replacements of
anilide 3 were prioritized and synthesized. Table 2 shows the calculated energy
differences between the two accessible conformations for each of these
compounds in kcal/mol (a negative energy signifying a preference for the
proposed bioactive conformation), along with inhibition constants (Ki) against
recombinant full-length ATR, ATM and DNA-PK and where available IC values
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
from an ATR-mediated cell based biomarker assay (measuring phosphorylation
of the ATR substrate variant histone H2A (H2AX) at Serine 139 in human
colorectal carcinoma HCT116 cells).1
1
5
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 16 of 73
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Table 2. Enzyme inhibition constants (Ki), pH2AX cell assay IC values and
5
0
conformational energy differences for aminopyrazine ATR inhibitors.
0
1
2
3
4
5
6
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
NH2
R
N
N
SO iPr
2
Cellular
potency
IC50 (M) difference
Enzyme inhibition Ki (M)
Energy
Compounds
R
ATR
ATM
DNA-PK pH2AX
(kcal/mol)
O
3
0.004
4
2.8
1.03
N
H
N
O
4
6
7
8
0.001
0.006
0.002
0.001
1.4
>4
>4
>4
8.9
1.1
>2.5
0.8
-4.59
-8.94
-8.74
-7.76
N
O
N
N
N
N
N
1.3
1.3
N
>10
0.41
O
O
9
0.005
0.4
4
2.3
-0.87
3.80
N
N
O
10
0.026
0.021
1.35
>10
>1.6
>4
>2.5
N
N
11
12
1.65
-6.73
O
N
0.004
>10
>4
>2.5
O
N N
S
0.001
>4
>4
1.7
-12.47
13
1
6
ACS Paragon Plus Environment

Page 17 of 73
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Table 2 shows that aromatic 5-membered rings are generally well tolerated as an
amide replacement with many compounds showing similar ATR inhibition
potency to the original amide 3. The isoxazole 4, oxadiazole 8 and thiadiazole 13
show improvements in potency (Ki ~1 nM vs ~4 nM for the starting compound,
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). Encouragingly, these analogues retained the >1000-fold selectivity for ATR
over the homologous kinases ATM and DNA-PK, and the cell activity of 3.
The improvement in affinity for ATR observed for some compounds cannot be
attributed to new hydrogen bonding interactions formed between the 5-
membered heteroaryl rings and the ATR active site. For example, the O-1 and the
N-2 atoms of compound 10 that are accessible to the enzyme are the same as
those of compound 4 yet these compounds have substantially different inhibition
constants (26 nM vs 1 nM respectively). Instead, the observed differences in
affinity can be predicted by the conformational energy differences between the
bioactive and alternative conformations listed in Table 2. Compound 4 favors the
bioactive conformation as this aligns the isoxazole oxygen and CH with the 2-
aminopyrazine NH and ring nitrogen, respectively. For compounds 10 and 11,
2
the alternative conformation is favored by 3.8 and 1.7 kcal/mol respectively over
the bioactive conformation and these energetic penalties are reflected in their
elevated Ki values. For compound 11 to assume the bioactive conformation the
aromatic proton of the oxazole ring would need to sit opposite a proton of the
hinge binder amine. For compound 10, the electrostatic misalignment is less
obvious since in both conformations a hydrogen bond acceptor, either the O-1 or
N-4 of the 1,2,4-oxadiazole, sits opposite the 2-aminopyrazine 4-nitrogen or
amine. The calculated preference for the alternative conformation is in
agreement with the observation that aromatic oxygen atoms are weaker
1
7
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 18 of 73
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
hydrogen bond acceptors than aromatic nitrogens.38 The alternative
conformation, which places the oxadiazole O-1 opposite the pyrazine ring
nitrogen, is thus preferred as a consequence of the relatively lower electrostatic
repulsion between these two acceptors compared to the repulsion between two
nitrogen atoms in the bioactive conformation.
0
1
2
3
4
5
6
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
Compounds 4, 8 and 12 maintain low nanomolar potency against ATR as
predicted by the negative relative conformational energies of -4.6, -7.8 and -6.7
kcal/mol, respectively, favoring the bioactive conformation. In compound 9, an
aromatic nitrogen atom of the 1,2,4-oxadiazole is always positioned opposite the
pyrazine ring acceptor. The energy difference between the two conformations is
small but the bioactive conformation is still preferred as confirmed by its low
nanomolar Ki. Aromatic sulfur atoms carry a small positive charge and are
known to interact favorably with hydrogen bond acceptors.42 This favorable
interaction is reflected in the largest negative energy difference calculated for
the thiadiazole 13, indicating a strong preference for the bioactive conformation.
Although ligand pre-organization is only one of several factors contributing to
the free energy of binding, the calculated conformational energy differences for
fully aromatic 5-membered rings still show a reasonable trend when plotted
2
against the measured inhibition constants (R = 0.72). Overall, the calculated
energy differences provided useful guidance for the design of heteroaryl amide
replacements.
The isoxazole 4 was selected as a starting point for further optimization because
of its superior selectivity profile against unrelated kinases, for example GSK3 (4
Ki = 650 nM, 7 Ki = 530 nM, 8 Ki = 22 nM). The phenyl ring of the isoxazole 4 is
predicted to extend deep under the P-loop of ATR near a negatively charged area
1
8
ACS Paragon Plus Environment

Page 19 of 73
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
in our ATR kinase homology model (Figure 4). We speculated that this negatively
charged environment could be exploited by amines projected from the para
position to improve both binding affinity and aqueous solubility. Table 3 shows
enzyme inhibition constants and cell based IC50 values for a selection of
compounds containing para amine substituted benzyl rings.
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Figure 4a-b. (A) Compound 4 docked in a homology model of the ATR active site
based on PDB entry 1E7V with ATR represented by its electrostatic potential
map. An area of high negative charge (in red) is located to the right of the phenyl
ring of compound 4. (B) Compound 2 docked into the active site of ATR.
Hydrogen bonds to the hinge, the unique residue Gly2385 as well as Asn2480
and Asp2494 that are part of the conserved magnesium binding site and salt
bridge are indicated. Polar residues around the phenyl ring are shown as sticks.
1
9
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 20 of 73
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Table 3. Enzyme inhibition constants, pH2AX cell assay IC s and aqueous
5
0
solubility for 3-isoxazole-2-aminopyrazine ATR inhibitors.
N
NH2
O
R
N
N
0
1
2
3
4
5
6
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
SO2iPr
Cellular
Enzyme inhibition Ki (nM) potency
Aqueous
solubility
(M)
Compounds
R
IC50 (M)
ATR
ATM
DNA-PK
pH2AX
4
1.3
1400
>4000
1.1
<0.3
12.5
NH2
14
0.27
115
>4000
>25
N
H
2
0.17
0.21
44
47
>4000
3350
0.019
0.076
52
72
N
H
1
5
6
O
1
N
0.14
1.24
189
82
>4000
785
0.013
0.3
0.8
H
NH2
17
27.9
NH2
18
19
20
0.53
2.40
0.75
105
190
93
1700
>4000
2610
0.13
0.13
0.21
45.1
1.36
4.4
N
H
Cl
N
H
Cl
Introduction of a benzylamine motif, 14, resulted in improved ATR affinity (Ki
270 pM cf 1.3 nM for 3), however cell activity was compromised. This may be
attributed to poor cell penetration associated with a highly solvated primary
2
0
ACS Paragon Plus Environment

Page 21 of 73
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
amine with a low calculated logD of 0.76. Small amine substituents, to form a
secondary amine such as in 2, improved cell activity substantially (e.g. 0.019 µM
for 2 vs 1.1 µM for 4). The introduction of secondary amines also led to a
significant reduction in calculated logD (from 3.6 for compound 4 to 1.1 for 2)
with a concomitant increase in aqueous solubility (cf. Table 3). Larger amine
substituents such as the ethyl in 15 and pyranyl in 16 provided only limited
improvement in terms of affinity and cell potency over the methylamine 2.
As an alternative approach to modulate the properties of the primary amine we
considered alkylation of the benzylic position and prepared the two
stereoisomers of a methyl substituent on the benzylamine methylene. This was
most effective for the R-stereoisomer, which retained good ATR affinity and a
substantial improvement in cell potency compared with 4. Although introduction
of the -methyl group markedly improved the potency in the pH2AX cell assay
compared to 14, N-alkylation, as in 2, yielded lower cell assay IC s (19 nM for 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
5
0
vs 130 nM for 18). Further hydrophobic substitutions on the ortho and meta
positions of the phenyl ring in 19 and 20 lost potency in both the ATR enzyme
and pH2AX cell assays. We speculate that this may be due to the polar local
environment and limited available space around the phenyl ring in the ATP
binding site.
In general little correlation was observed between measured and calculated logD
and pKa values and activity in cellular assays. For example, although compound
16 had a markedly lower measured pKa value of 8.2 versus 9.7 for compound 2
both compounds are essentially equipotent in the pH2AX cell assay. Despite its
relatively high pKa, compound 2 showed good permeability in a Caco2 assay (A-
B 7.2 10-6 cm/s with an efflux ratio of 2.9). From the set of hetero-alkyl
2
1
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 22 of 73
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
secondary amines explored on the isoxazole scaffold, only compounds 15 and 16
showed activity comparable to that of compound 2 in the pH2AX cell assay and
out of these only 2 and 15 combined this with good aqueous solubility.
0
1
2
3
4
5
6
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
Introduction of primary or secondary amines had no impact on affinity for DNA-
PK but did increase affinity for ATM kinase (Ki = 44 nM for 2, cf 1400 nM for 4)
as the amino acids near the benzyl amine are conserved between ATR and ATM.
For DNAPK, any favorable effects of introducing the benzyl amine on binding
affinity may be cancelled out by the proximity of the positive charge of the
unique Arg3733 residue in the P-loop of DNAPK. Despite this increase in ATM
enzyme affinity, selectivity for ATR over ATM is maintained at >200-fold.
The overall selectivity profile of 2 was established by testing the compound
against a large panel of unrelated kinases at a concentration of 400 nM. 278 out
of 291 kinases tested showed <50% inhibition at 400 nM (assuming a
competitive mechanism of action and under the conditions of the experiment
this corresponds to an estimated Ki of > 200 nM, and > 1000 fold selectivity for
ATR). Inhibition constants for the 13 remaining kinases were determined and
are listed in Table 4. Of these only Flt4 kinase showed less than 100-fold
selectivity, with an estimated Ki of 8 nM, which corresponds to a selectivity of
~50-fold).
2
2
ACS Paragon Plus Environment

Page 23 of 73
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Table 4. Human kinases with enzyme inhibition constants Ki < 200 nM for
compound 2.
Kinase
ATR
Ki (nM)
0.165
59a
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
Abl(T315I)
cKit(D816H)
DYRK2
Flt3(D835Y)
Flt3
39a
26a
18a
130
Flt4
8a
33a
GSK3a
GSK3b
140
150
30a
15a
140
JAK2
Mer
MLK1
PI3Kinase (p110/p85)
SYK
190
a
Ki estimated based on the ATP concentration in the assay being equal to the
Michaelis constant (Km) for ATP and inhibition is through an ATP-competitive
mechanism of action. All data was obtained from a minimum of 2 replicates.
Compound 2 was selected as the clinical candidate based on its overall profile in
terms of potency, selectivity and solubility. A ligand interaction diagram
detailing nearby residues and key hydrogen bonds involving compound 2 as
docked into our ATR homology model is shown in Figure 5. This compound
showed >200-fold selectivity over Cytochrome P450’s (IC s against Cyp3A4,
5
0
2
3
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 24 of 73
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
2
C9 and 2D6 all > 30 M) and the hErg potassium ion channel (IC = 3 M).
5
0
Compound 2 had a favorable whole blood IV pharmacokinetic profile in Male
Sprague-Dawley rats (Cl = 26 mL/min/kg; Vss = 21 L/kg; T1/2 = 11.6 h) and oral
bioavailability to support in vivo efficacy studies.8, 12
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Figure 5. Ligand interaction diagram showing nearby residues, aromatic
stacking (green lines) and key hydrogen bonds (purple lines) involving
compound 2 as docked into our ATR homology model derived from PDB entry
1E7V.
In cell experiments, compound 2 markedly sensitized the HCT116 colorectal cell
line to the DNA cross-linking agent cisplatin, at concentrations of 2 <50 nM,
which compares with far higher concentrations of 1 required to achieve the
same levels of cisplatin sensitisation (~500 nM). Figure 6 compares the synergy
2
4
ACS Paragon Plus Environment

Page 25 of 73
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
profiles observed for combinations of cisplatin with compound 1 or compound 2
across a range of concentrations. The potential for 2 to sensitize cancer cells to
other DNA damaging agents or radiation has been extensively reported in a
range of in vitro and in vivo tumor models.8,11,12 IV administration of compound 2
is currently being evaluated in Phase I/II clinical studies as monotherapy and in
combination with several DNA damaging chemotherapies and radiation therapy
for the treatment of solid tumours.21
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Figure 6. Macsynergy plots for compounds 1 and 2 dosed in combination with
cisplatin in an HCT116 MTS cell assay. Synergistic or antagonistic interactions
between ATR inhibitors and cisplatin are represented as deviations above or
below the plane (at z=0) that corresponds to additivity. Compound 2 achieves
similar synergy with cisplatin as 1 at a ~10-fold lower concentration.
2
5
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 26 of 73
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
CONCLUSIONS
Multiple pre-clinical studies have shown that ATR inhibition renders many
cancer cells, but not non-cancer cells, highly sensitive to certain DNA damaging
drugs that are widely used in cancer therapy. This differential has been
attributed to the ability of non-cancer cells to utilize compensatory DNA repair
processes, which are commonly impaired in cancer. Furthermore, ATR inhibition
has been shown to sensitize cancer cells to other inhibitors of DNA repair such as
PARP and Chk1 inhibitors, and can have single agent activity in specific cancer
cells. Compound 1 was the first potent and selective ATR inhibitor to be reported
and is widely used as an in vitro tool, however, it is limited by poor physical
properties and pharmacokinetics. In this manuscript we have described the
optimization of 1 to afford 2 (VX-970, also known as M6620), the first ATR
kinase inhibitor to enter clinical trials. Substantial improvements in ATR affinity,
physical properties and PK were achieved primarily by optimizing the anilide
moiety of 1. An isoxazole ring substituted with a secondary benzylamine was
found to be the optimal replacement for the anilide moiety of 1 leading to
improved ATR affinity, selectivity, aqueous solubility and activity in cell assays.
Compound 2 maintained > 100 fold selectivity over related kinases and
unrelated proteins such as hErg and P450 enzymes and had PK properties
suitable for IV administration.
0
1
2
3
4
5
6
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
Three factors helped to optimize the potency and physical properties of
compound 1. The first was to maximize steric fit to the active site of ATR. While
flat, 5-membered heteroaryl linkers maintained good ATR affinity, a significant
loss of potency was observed for the partially saturated 4,5-dihydroisoxazole 5
2
6
ACS Paragon Plus Environment

Page 27 of 73
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
compared to its fully aromatic isoxazole counterpart 4. The more 3-dimensional
3
shape of 5, due to its sp carbon linking the 4,5-dihydroisoxazole to the 2-
aminopyrazine, results in the compound incurring an energy penalty when
binding to the ATR binding site in a flattened conformation.
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
A second factor was the degree of electrostatic complementarity between the 5-
membered heteroaryl ring and the 2-aminopyrazine hinge binder. In all cases,
except for the 1,2,4-oxadiazole 10 and the oxazole 11, the interactions between
the two rings favor the bound conformation as predicted by quantum chemical
calculations. Furthermore, the magnitude of the gas phase energy differences
between the two possible conformations for each of the inhibitors correlated
with ATR enzyme affinity and as such this calculation provided a convenient tool
for predicting optimal heterocyclic scaffold modifications.
Thirdly, we optimized polar interactions between the inhibitor and the ATP
binding site of ATR by substituting the phenyl isoxazole of 4 at the para position
with amines. Compound 2, with a simple N-methyl benzylamine motif, achieved
an 8-fold improvement in affinity compared to its unsubstituted counterpart 4
along with greatly improved cell potency, aqueous solubility and pharmaco-
kinetics.
The 5-heteroaryl containing inhibitors reported in this study present a new
chemical motif for potent ATR kinase inhibition. The resulting compound, 2 (VX-
9
70/M6620), is currently being evaluated in Phase I/II clinical studies in
combination with several DNA damaging chemotherapies and PARP inhibitors,
and as monotherapy for the treatment of solid tumours.
2
7
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 28 of 73
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 SECTION
General synthetic methods
0
1
2
3
4
5
6
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
All commercially available solvents and reagents were used as received.
Microwave reactions were carried out using a CEM Discovery microwave.
Analytical thin layer chromatography was carried out using glass-backed plates
coated with Merck Kieselgel 60 GF . Plates were visualized using UV light (254
2
40
nm or 366 nm) and/or by staining with potassium permanganate followed by
heating. Flash chromatography was carried out on an ISCO© CombiflashR
CompanionTM system eluting with a 0 to 100% EtOAc/petroleum ether gradient.
Samples were applied pre-absorbed on silica. Semi-preparative reverse phase
HPLC was carried out on a Waters autopurification HPLC-MS system equipped
with a waters C-18 sunfire reverse phase column (19 mm x 150 mm, 5 mm). The
mobile phases were acetonitrile (0.1 % trifluoroacetic acid) and water (0.1 %
1
13
trifluoroacetic acid). H-NMR and C-NMR spectra were recorded at 400 MHz or
5
00 MHz and 125 MHz or 100 MHz, using a Bruker DPX 400 instrument or a
Bruker DPX 500 instrument. MS samples were analyzed on a MicroMass Quattro
Micro mass spectrometer operated in single MS mode with electrospray
ionization. Samples were introduced into the mass spectrometer using
chromatography. All final products had a purity ≥95%, unless specified
otherwise in the experimental details. The purity of the final compounds was
determined using HPLC method: analytical reverse phase HPLC-MS was carried
out on an agilent 100 HPLC with a waters Quattro MS system equipped with an
ACE 5mm C-18 reverse phase column (4.6 mm x 150 mm, 5 mm). The mobile
phases were acetonitrile/methanol (1:1) and water (10 mm ammonium acetate).
2
8
ACS Paragon Plus Environment

Page 29 of 73
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
5
-(4-(Isopropylsulfonyl)phenyl)-3-(3-(4-
(
(methylamino)methyl)phenyl)isoxazol-5-yl)pyrazin-2-amine (2). To a
solution of N,N-di(tert-butoxycarbonyl)-5-(4-(isopropylsulfonyl)phenyl)-3-
ethynylpyrazin-2-amine 25 (6.8 g, 13.56 mmol) and N-hydroxy-4-methyl-
benzimidoyl chloride (2.71 g, 13.56 mmol) in THF (140 mL) at room
temperature was added triethylamine (1.65 g, 2.27 mL, 16.27 mmol) dropwise
over l0 min. The mixture was stirred at room temperature overnight then at 60
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 for 2 h. The reaction mixture was concentrated under reduced pressure and
the residue was partitioned between CH Cl (30 mL) and brine (50 mL). The
2
2
organic layer was washed with an aqueous saturated solution of NaHCO (50
3
mL), dried over MgSO , filtered and evaporated under reduced pressure. The
4
residue was purified by chromatography on silica gel, eluting with 20%
EtOAc/petroleum
ether
to
give
N,N-di(tert-butoxycarbonyl)-5-(4-
(
isopropylsulfonyl)phenyl)-3-(3-(4-methyl)phenylisoxazol-5-yl)pyrazin-2-amine
1
as an off-white solid (7.1 g, 82% Yield); H NMR (400 MHz, DMSO-d )  9.51 (s,
6
IH), 8.66 (m, 2H), 8.07 (m, 2H), 8.01 (s, 1H), 7.92 (m, 2H), 7.39 (m, 2H), 3.55 (m,
1H), 3.34 (s, 3H), 1.33 (s, 18H), 1.21 (m, 6H); MS (ES+) m/z 635.3 (M + H)+.
To a solution of N,N-di(tert-butoxycarbonyl)-5-(4-(isopropylsulfonyl)phenyl)-
3
-(3-(4-methyl)phenylisoxazol-5-yl)pyrazin-2-amine (3.5 g, 5.51 mmol) in ethyl
acetate (35 mL) was added N-bromosuccinimide (1.28 g, 7.17 mmol), followed
by AIBN (180 mg, 1.1 mmol). The resulting mixture was placed under a bright
lamp and heated to 85 °C for l h. The reaction mixture was allowed to cool to
ambient temperature, washed sequentially with brine (50 mL) and an aqueous
saturated solution of NaHCO (50 mL), then filtered, dried over MgSO and
3
4
concentrated
in
vacuo
to
afford
N,N-di(tert-butoxycarbonyl)-5-(4-
2
9
ACS Paragon Plus Environment