Discovery of the 1,7-diazacarbazole class of inhibitors of checkpoint kinase 1

Article abstract of DOI:10.1016/j.bmcl.2014.10.063

Checkpoint kinase 1 (ChK1) is activated in response to DNA damage, acting to temporarily block cell cycle progression and allow for DNA repair. It is envisaged that inhibition of ChK1 will sensitize tumor cells to treatment with DNA-damaging therapies, an

Full text of DOI:10.1016/j.bmcl.2014.10.063

Bioorganic & Medicinal Chemistry Letters 24 (2014) 5704–5709
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Discovery of the 1,7-diazacarbazole class of inhibitors of checkpoint
kinase 1
b
e
a
c
a
Lewis Gazzard a, , Brent Appleton , Kerry Chapman , Huifen Chen , Kevin Clark , Joy Drobnick ,
Simon Goodacre e, Jason Halladay d, Joseph Lyssikatos a, Stephen Schmidt c, Steve Sideris c,
Christian Wiesmann b, Karen Williams e, Ping Wu b, Ivana Yen c, Shiva Malek c
⇑
a Department of Discovery Chemistry, Genentech Inc., 1 DNA Way, South San Francisco, CA 94080, United States
b Department of Structural Biology, Genentech Inc., 1 DNA Way, South San Francisco, CA 94080, United States
c Department of Biochemical and Cellular Pharmacology, Genentech Inc., 1 DNA Way, South San Francisco, CA 94080, United States
d Department of Drug Metabolism and Pharmacokinetics, Genentech Inc., 1 DNA Way, South San Francisco, CA 94080, United States
e Argenta, A Charles River Company, 8-9 Spire Green Centre, Harlow, Essex CM19 5TR, United Kingdom
a r t i c l e i n f o
a b s t r a c t
Article history:
Checkpoint kinase 1 (ChK1) is activated in response to DNA damage, acting to temporarily block cell cycle
progression and allow for DNA repair. It is envisaged that inhibition of ChK1 will sensitize tumor cells to
treatment with DNA-damaging therapies, and may enhance the therapeutic window. High throughput
screening identified carboxylate-containing diarylpyrazines as a prominent hit series, but with limited
biochemical potency and no cellular activity. Through a series of SAR investigations and X-ray crystallo-
graphic analysis the critical role of polar contacts with conserved waters in the kinase back pocket was
established. Structure-based design, guided by in silico modeling, transformed the series to better satisfy
these contacts and the novel 1,7-diazacarbazole class of inhibitors was discovered. Here we present the
genesis of this novel series and the identification of GNE-783, a potent, selective and orally bioavailable
inhibitor of ChK1.
Received 10 September 2014
Revised 10 October 2014
Accepted 17 October 2014
Available online 27 October 2014
Keywords:
ChK1
Checkpoint
Structure-based design
Diazacarbazole
GNE-783
Ó 2014 Elsevier Ltd. All rights reserved.
Checkpoint kinase 1 (ChK1) is a serine/threonine kinase that
functions as a central mediator of the S and G2/M phase cell cycle
checkpoints. ChK1 is activated in response to DNA damage, such as
that resulting from certain chemotherapies or radiotherapy, and
serves to activate these checkpoints to block progression into
mitosis and allow for repair of damaged DNA.1 As a result of
genetic deficiencies many tumor cells lack a functional G1/S check-
point earlier in the cell cycle. For example, more than half of
human cancers exhibit mutations in the p53 gene, resulting in fail-
ure to achieve p53-mediated G1/S checkpoint arrest.2 These tumor
cells are therefore disproportionately dependent on the later,
ChK1-mediated S and G2/M checkpoints for survival in response
to DNA-damaging therapies. Inhibition of ChK1 is an emerging
strategy for selectively potentiating the cytotoxicity of chemother-
apeutic agents in G1 checkpoint compromised tumor cells while
minimizing toxicity to normal, checkpoint competent cells.1
It is envisaged that an inhibitor of ChK1 would preferentially
sensitize p53-deficient tumor cells to the effects of DNA-damaging
agents and improve the therapeutic window of these agents. A
number of ChK1 inhibitor programs have been reported and
several molecules are undergoing clinical evaluation, with two
programs having progressed into Phase II.3 It is likely that
co-treatment of a sensitizing agent with the cytotoxic agent will
require careful coordination of the treatment regimens.4 We there-
fore embarked on a program to discover an orally dosed inhibitor
of ChK1, intending to maximize utility in the clinic by allowing
the greatest possible flexibility in dosing in combination with
radiotherapy and both current and future standard of care
chemotherapies.
High-throughput screening of our small molecule screening
library utilizing an ATP-competitive Z-lyte assay5 yielded a rela-
tively low hit rate, and few compounds with promising potency
and ligand efficiency.6 The proportion of validated hits having a
biochemical IC50 below 25 lM was 0.01% of the library compounds
screened. Clustering identified 2,6-diarylpyrazines as the most
prominent hit series, of which 40-carboxyphenyl analogs intrigu-
ingly represented the most numerous and potent subseries, exhib-
iting low micromolar potency. This observation was rationalized
using our in silico docking models,7 which predicted binding of
⇑
Corresponding author.
http://dx.doi.org/10.1016/j.bmcl.2014.10.063
0960-894X/Ó 2014 Elsevier Ltd. All rights reserved.

L. Gazzard et al. / Bioorg. Med. Chem. Lett. 24 (2014) 5704–5709
5705
Table 1
the pyrazine moiety to the hinge (Cys87) placing the carboxylate
Aryl substituent SAR in the (4’-carboxyphenyl)azine series
function in proximity to the catalytic lysine (Lys38) (Fig. 1).
After hit confirmation from resynthesized samples using a ChK1
AlphaScreen assay5 we embarked on expansion of this scaffold
with the intention of improving biochemical activity, demonstrat-
ing cellular activity, and identifying novel chemical matter.
Synthesis of a range of 2,6-diarylpyrazines and 3,5-diarylpyridines
was easily accomplished by sequential Suzuki-Miyaura cross-
couplings of 2,6-dichloropyrazine or 3,5-dibromopyridine, respec-
tively, with the appropriate arylboronic acids or arylboronate
esters. Similarly, 2-amino-3,5-diarylpyrazines could be synthe-
sized from 2-amino-3,5-dibromopyrazine with the first-performed
coupling invariably occurring adjacent to the 2-amino function in
the 3-position (Scheme 1).8
Compound
Y
R
ChK1 IC50
1.2
(lM)
1
N
2
3
N
N
1.3
4.1
An initial series of 40-carboxyphenyl analogs (Table 1) quickly
established that improved biochemical potency could be achieved
with the 4-(4-methylpiperizin-1-yl)phenyl substituent (compound
4). Replacement of the hinge-binding pyrazine core with pyridine
resulted in a further increase in potency (compound 5). However,
the 4’-carboxyphenyl substituted analogs consistently demon-
strated no activity in our cellular assays, presumed to result from
an inability of the anionic carboxylate to permeate the lipid mem-
brane. Consequently our focus turned to replacement of the unde-
sirable carboxylate function. Evaluation of a number of carboxylate
replacements and isosteres demonstrated little improvement in
biochemical potency and in no case cellular activity (Table 2).
Docking experiments implicated the involvement of a non-
classical C2-H contact with the Glu85 hinge carbonyl. We postu-
lated that an additional polar contact with this hinge carbonyl
might improve potency as this approach had successfully been
utilized in a 2-aminopyrazine-3-carboxamide series of ChK1 inhib-
itors.9 In this case however, the 2-amino substituted analog of
compound 4 exhibited lower potency in the ChK1 assay (com-
pound 12).
4
5
N
0.21
0.06
CH
Table 2
Carboxylate replacement and modification of the hinge-binding moiety
Compound
R
ChK1 IC50 (lM)
Lys38
4
0.21
0.39
1.2
Leu86
Glu85
6
7
8
>10
Cys87
OH
HO2
C
9
0.47
0.69
N
N
1
10
Figure 1. In silico prediction of binding mode of lead compound 1, solvent
accessible surface is shown.
11
12
0.20
0.82
—
A structural basis for both the preference for a carboxylate and
intolerance for the 2-amino function were provided by X-ray
crystallography (Fig. 2). Examination of a high resolution (2.05 Å)
co-crystal structure of compound 1 with the kinase domain of
ChK1 (Fig. 2a) demonstrated the involvement of both carboxylate
Scheme 1. Synthesis of 2-H and 2-amino diarylazines. Reagents and conditions: (a)
R1B(OH)2, PdCl2(PPh3)2, Na2CO3(aq), MeCN,
l
W; (b) R2B(OH)2, PdCl2(PPh3)2,
Na2CO3(aq), MeCN, lW.

5706
L. Gazzard et al. / Bioorg. Med. Chem. Lett. 24 (2014) 5704–5709
Glu85
Lys38
Lys38
water
triad
Cys87
Cys87
OH
OH
HO2C
HO2C
N
N
N
H2
region
13
(c)
1
H2N
(a)
N
(b)
Figure 2. Comparison of ChK1 kinase domain co-crystal structures of HTS lead compound 1 (a) with 2-amino analog 13 (b); overlay indicates torsion of the phenyl spacer in
13 relative to 1, while carboxylate orientation is maintained (c).
oxygen atoms, one in an electrostatic, hydrogen bonded interaction
with the catalytic lysine (Lys38), and the other in hydrogen-
bonded interactions with two of the three conserved crystallo-
graphic waters in the back-pocket. The carboxylate, phenyl, and
pyrazine moieties are bound in a coplanar arrangement. This
mode of binding had been correctly predicted by the docking
experiments, involving acceptance of a single hydrogen bond from
the kinase hinge (Cys87), with the 5-aryl substituent providing
further binding affinity through hydrophobic interactions in the
‘H2 region’.10 An identical binding mode is observed for the
3,5-diarylpyridine analogs with the carboxylate, phenyl, and pyri-
dine moieties remaining coplanar. The highly conserved nature of
the water triad observed in the back pocket of ChK1 is evident from
the many co-crystal structures reported in the literature and has
Figure 3. Conceptual transformation of the diarylazine series to the 1,7-
diazacarbazole.
previously been noted.11
Introduction of a 2-amino function to compound 1 provided a
second polar contact to a hinge carbonyl (Glu85) (compound
13).12 Comparison of the X-ray co-crystal structure of 2-amino
Conceptual transformation of the 2-aminoazine series to the
compound 13 (Fig. 2b) demonstrates a significant torsion of the
3-phenyl moiety (+43°) versus the desamino compound 1 (ꢀ15°)
(Fig. 2c). This torsion of the phenyl unit relative to the 2-aminopyr-
azine hinge-binder was expected from steric considerations, and is
consistent with comparable examples found in the Cambridge
Structural Database (35–48°).13 This indicates the torsion angle
remains similar to the low energy conformation favored in solution
phase. Crucially however, the carboxylate function maintains a
coplanar arrangement relative to the pyrazine core (ꢀ3°), signifi-
cantly limiting efficient conjugation with the phenyl spacer moiety
(+46°). This represents a high-energy carboxylate conformation,
versus the phenyl-conjugated conformer predicted to be favored
in solution, and a significant energetic penalty for ligand binding.
We therefore examined ways to eliminate this penalty, by achiev-
ing these coplanar hydrogen bonds in a ligand that maintains the
binding conformation in solution.
novel 1,7-diazacarbazoles was accomplished in three steps
(Fig. 3). (i) Cyclization of the 2-amino function to the 3-phenyl
ortho-position yielded a tricyclic alpha-carboline. This was envis-
aged to bring the carboxylate and phenyl spacer into coplanarity
with the pyridine core, and further, to eliminate an unsatisfied
hydrogen bond donor and the consequent desolvation penalty.
This operation is therefore expected to eliminate two energetic
penalties to ligand binding. In silico modeling indicated the vector
presented by the 7-carboxylate function allowed insufficient space
to avoid a deleterious steric clash with the gate-keeper residue
(Leu84), resulting in movement of the tricyclic core away from
the hinge to accommodate (Fig. 4a). Further, the closer approach
of the carboxylate would result in disruption of the highly favored
back-pocket water network. (ii) Migration of the carboxylate to the
6-position relieved steric congestion and presented a suitable
vector to maintain contact with the catalytic lysine (Lys38)
Lys38
4.1 Å
4.1 Å
3.9 Å
3.1 Å
3.2 Å
3.0 Å
4.1 Å
3.4 Å
3.4 Å
3.6 Å
3.3 Å
3.2 Å
Cys87
1.3 Å
Leu84
Glu85
(a)
(b)
(c)
Figure 4. In silico docking models of (a) 7-carboxy-alpha-carboline and (b) 6-carboxy-alpha-carboline intermediates in the conceptual transformation of the diarylazine
series to the 1,7-diazacarbazole series (c). Indicated surface and waters (red spheres) are those of the docking basis structure.7

L. Gazzard et al. / Bioorg. Med. Chem. Lett. 24 (2014) 5704–5709
5707
(Fig. 4b). This operation allows movement of the tricyclic core
closer to the hinge. However the favorable polar contact with the
conserved back-pocket water network is removed. (iii) Migration
of the 7-carboxylate to the 6-position permits introduction of a
7-aza atom to serve as an acceptor, and to reestablish a hydrogen
bond from the back-pocket water (Fig. 4c). Supported by our dock-
ing experiments, we targeted the 1,7-diazacarbazole core directly
as most likely to satisfy polar contacts with both the hinge residues
and conserved waters.
Preparation of the archetype 1,7-diazacarbazole (14) was real-
ized by Pictet–Spengler synthesis in five steps from commercially
available 7-azatryptophan (Scheme 2).14 Esterification with thionyl
chloride in methanol yielded the methyl ester hydrochloride salt,
which underwent Pictet–Spengler cyclization with formalin in
pyridine, and subsequent aromatization with selenium dioxide to
yield methyl 9H-pyrrolo[2,3-b:5,4-c0]dipyridine-6-carboxylate.
The oxidative aromatization step was accelerated by the addition
of pyridine. The preparation of this core tricycle was recently
reported utilizing a comparable route.15 As predicted, regiospecific
bromination upon treatment with bromine and sodium acetate in
acetic acid occurred at the 3-position to yield the key 1,7-diazacar-
bazole intermediate methyl 3-bromo-9H-pyrrolo[2,3-b:5,4-c0]
dipyridine-6-carboxylate. Suzuki-Miyaura coupling with [4-(4-
methylpiperazin-1-yl)phenyl]boronic acid afforded the corre-
sponding 3-aryl-6-carboxy-1,7-diazacarbazole (14). Compound
14 was subsequently converted to 6-carboxamide 15 and upon
dehydration afforded 6-carbonitrile 16.
Scheme 2. Pictet-Spengler synthesis of 1,7-diazacarbazoles 14–16. Reagents and
conditions: (a) SOCl2, MeOH, reflux; (b) CH2O(aq), pyridine, 100 °C; (c) SeO2,
pyridine, 1,4-dioxane, reflux, 80% over 3 steps; (d) Br2, NaOAc, AcOH, 100 °C, sealed
tube, quant.; (e) 4-(4-methylpiperidin-1-yl)phenylboronic acid, PdCl2(PPh3)2, Na2-
CO3(aq), MeCN,
lW, 46%; (f) NH3, PyBop, HOBt, DIPEA, DMF, 45%; (g) (CF3CO)2O,
NEt3, THF, 35 °C, 95%.
Synthesis of regioisomeric 1,5-diazacarbazole-6-carbonitrile
17 was accomplished via Suzuki-Miyaura cross-coupling of
commercially available 5-amino-6-bromopyridine-2-carbonitrile
and 2-fluoropyridine-3-boronic acid (Scheme 3).16 The resulting
bipyridyl intermediate underwent base-mediated cyclization to
form the 1,5-diazacarbazole nucleus. Regiospecific bromination
and Suzuki-Miyaura coupling with [4-(4-methylpiperazin-1-yl)
phenyl]boronic acid yielded compound 17.
Evaluation of the 6-carboxy-1,7-diazacarbazole 14 in our ChK1
biochemical assay demonstrated a significant 50-fold improve-
ment in potency versus the analogous diarylpyrazine (compound
4), and 15-fold versus the diarylpyridine (compound 5). Again, no
detectable activity was observed with compound 14 in our cellular
assays, presumed to result from poor cell membrane permeation of
the charged carboxylate (Table 3). Analysis of the ChK1 co-crystal
structure with compound 14 confirmed our docking predictions
(Fig. 5). The 7-aza atom accepts a hydrogen bond from the one of
Scheme 3. Synthesis of the 1,5-diazacarbazole compound 17. Reagents and
conditions: (a) 2-fluoropyridine-3-boronic acid, PdCl2(PPh3)2, Na2CO3(aq), MeCN,
lW, 60%; (b) NaHMDS, THF, 50 °C, 55%; (c) Br2, NaOAc, AcOH, 80 °C, 89%; (d) [4-(4-
methylpiperidin-1-yl)phenyl]boronic acid, PdCl2(PPh3)2, Na2CO3(aq), MeCN,
lW,
52%.
Table 3
Biochemical & cellular potency of diazacarbazoles
Compound
X
ChK1 IC50
(
lM)
pHH3a IC50
(l
M)
Isobol.b EC10fs
(l
M)
HT-29c GI50
(lM)
14
15
16
17
CO2H
CONH2
CN
0.0041
0.0023
0.0013
0.675
>10
—
0.098
—
>25
>25
—
>25
—
0.080
0.047
—
—
a
b
c
Checkpoint abrogation assay.
Isobologram assay.
Antiproliferative activity.

5708
L. Gazzard et al. / Bioorg. Med. Chem. Lett. 24 (2014) 5704–5709
expected position to accept a hydrogen bond from the back-pocket
water triad, and mimics the position of a crystallographic water in
the apo structure. The importance of this contact was demon-
strated by comparison with the regioisomeric 1,5-diazacarbazole
(compound 17). This isomer is able to maintain all direct polar con-
tacts with protein residues, but unable to satisfy the back pocket
water contact, and as a result suffers a greater than 500-fold loss
in biochemical potency (Table 3).
Cellular activity was evaluated with a checkpoint abrogation
assay in combination with SN-38.18 Treatment of HT-29 cells with
SN-38 results in ChK1-mediated activation of the G2/M checkpoint
and mitotic arrest. Release of checkpoint control in response to
treatment with the ChK1 inhibitor is evidenced by progression into
mitosis and phosphorylation of histone H3 (pHH3). Compound 16
exhibited excellent cellular potency, with a checkpoint abrogation
IC50 of 98 nM (Table 3).
(a)
(b)
(c)
Figure 5. Comparison of contacts with the catalytic lysine (Lys38) and back pocket
waters for compound 1 (a) and compound 14 (b); overlay indicates the displace-
ment of the carboxylate (c).
Lys38
Compound 16 was next evaluated in an isobologram cytotoxic-
ity assay,19 to demonstrate synergy in combination with gemcita-
bine. Co-treatment of HT-29 cells with varying concentrations of
gemcitabine and ChK1 inhibitor permits generation of a two-
dimensional dose response, and measurement of the degree of
potentiation (increase) of gemcitabine induced cytotoxicity. The
N7
concentration of ChK1 inhibitor required to reduce the IC50 of
19
gemcitabine by ten-fold is termed EC10fs
.
Again compound 16
exhibited excellent cellular potency with a EC10fs of 47 nM. Further,
16 demonstrated no single agent antiproliferative activity in HT-29
Cys87
Leu84
cells at concentrations up to 25 lM, representing a greater than
500-fold margin. Compound 16 was therefore selected for in vivo
Tyr86
pharmacokinetic evaluation.
Following intravenous administration of compound 16 to rats
and mice, moderate and high clearances (Cl) were observed,
respectively, (Table 4). Bioavailability (F) values following oral
administration were greater than 50%, and dose-proportional in
mice in the CD1 and NCR nude (Taconic) strains. Importantly, the
exposures achieved following oral dosing in enabling excipients
could be matched in oral dosing of the free base in neutral
suspension. However, suspension dosing consistently lowered the
maximal plasma concentrations (Cmax). The well defined pharma-
cokinetic profiles in mice provide a valuable tool compound for
dose-scheduling investigations.
The kinase selectivity profile of compound 16 was assessed
against a screening panel of 65 kinases,20 chosen for general
pharmaceutical relevance. When compound 16 was screened at
100 nM concentration, representing a 77-fold excess versus the
ChK1 IC50, only one of the 65 kinases demonstrated significant
(>80%) inhibition (MAP4K4). When compound 16 was screened
Glu85
Figure 6. Alignment of ChK1 co-crystal structure with compound 16 (binding site
residues in gray & waters in red) with ChK1 apoprotein structure, PDB code 1IA8
(residues & waters in blue).
the conserved back-pocket waters (Fig. 5b), with the less solvent-
exposed ‘inner’ carboxylate oxygen providing a comparable vector
to the catalytic lysine to lead compound 1 (Fig. 5c). With a single
polar contact we had predicted the carboxylate to be more amena-
ble to functional group replacement. In fact, conversion to the pri-
mary amide (compound 15) and nitrile (compound 16) yielded
analogs with comparable excellent potency (Table 3).
Analysis of the ChK1 co-crystal with compound 16 demon-
strates effective replacement of the carboxylate, with the nitrile
accepting a hydrogen bond from the catalytic lysine (Fig. 6).
Alignment with the published apoprotein structure17 reveals the
positions of the three core nitrogen atoms and cyano substituent
are remarkably consistent with the position of the crystallographic
waters in the unbound structure. In particular, N7 occupies the
at 1 lM concentration, 10 of 65 kinases demonstrated >80% inhibi-
tion. These data indicate a promising kinase selectivity profile.
Notably, no significant inhibition of ChK2 or of the key pro-mitotic
regulators CDK1/cyclinB or CDK2/cyclinA was detected at 1
indicting a greater than 700-fold selectivity.
lM,
Table 4
Pharmacokinetic parameters for compound 16
Species (strain)
Mouse (CD1)
Route (dosing form)
Dose (mg/kg)
AUC (mM h)
AUC/dose (h kg/L)
CL (ml/min/kg)
67
t1/2 (h)
F (%)
Cmax (lM)
IVa
0.5
5
5
25
25
1
0.34
1.85
1.72
8.46
9.37
1.81
5.64
0.25
0.14
0.13
0.13
0.14
0.67
0.42
5.8
1.6
2.9
2.1
4.0
1.8
2.3
PO (solutionb)
PO (suspensionc)
PO (solutionb)
PO (suspensiond)
IVa
54
50
50
55
0.45
0.20
2.51
0.91
Mouse (NCR)
Rat (SD)
27
PO (solutiona)
5
62
0.93
a
b
c
Solution in 60% PEG400, 50 mM sodium citrate, pH 4.5.
Solution in 25% hydroxypropyl-b-cyclodextrin, 50 mM sodium citrate, pH 3.3.
Suspension in MCT, pH 7.0.
d
Suspension in 25% hydroxypropyl-b-cyclodextrin, 50 mM sodium citrate, pH 6.4.

L. Gazzard et al. / Bioorg. Med. Chem. Lett. 24 (2014) 5704–5709
5709
3. Data from www.clinicaltrials.gov: (a) LY-2603618 is under evaluation in
combination with pemetrexed and cisplatin for the treatment of NSCLC:
NCT01139775; (b) MK-8776 (SCH-900776) is under evaluation in combination
with cytarabine for the treatment of relapsed AML: NCT01870596.
4. Blackwood, E.; Epler, J.; Yen, I.; Flagella, M.; O’Brien, T.; Evangelista, M.;
Schmidt, S.; Xiao, Y.; Choi, J.; Kowanetz, K.; Ramiscal, J.; Wong, K.; Jakubiak, D.;
Yee, S.; Cain, G.; Gazzard, L.; Williams, K.; Halladay, J.; Jackson, P. K.; Malek, S.
Mol. Cancer Ther. 1968, 2013, 12.
5. Unless stated otherwise Chk1 biochemical data refers to AlphaScreen assay
data. Details of the ChK1 Z-lyte and AlphaScreen in vitro biochemical assays are
given in Supplemental material.
6. Hopkins, A. L.; Groom, C. R.; Alex, A. Drug Discov. Today 2004, 9, 430.
7. Ligand docking was based on a public domain ChK1 structure, PDB entry 1ZYS.
Details are provided in Supplemental material.
In summary, an HTS-derived lead series was studied through a
series of simple SAR investigations, establishing limited potency
and a dependence on an undesirable carboxylate functionality.
X-ray crystallographic analysis revealed the critical role of struc-
tural waters in binding the carboxylate, enabling the incorporation
of polar water contacts in compound design. Guided by insights
from crystallography and in silico modeling, structure-based
design yielded the novel 1,7-diazacarbazole series and compound
16, which demonstrates high biochemical and cellular potency, a
good pharmacokinetic profile and excellent kinase selectivity.
Compound 16 (GNE-783) represents a selective, orally bioavailable
tool for further investigation of ChK1 as a therapeutic target, and
an attractive lead for further optimization.
8. Examples of diarylazine syntheses as represented in Scheme 1: Jenmalm
Jensen, A. J.; Ringom, R.; Medina, C.; Shilvock, J.; Wiik, M.; Koolmeister, T.;
Angbrant, J.; Henriksson, M.; Sandvall, T.; Sutin, L.; Johansson, L.; Nilsson, B. M.
Patent Application US 2008/0039450.
9. Buhr, C. A.; Baik, T.-G.; Ma, S.; Tesfai, Z.; Wang, L.; Co, E. W.; Epshteyn, S.;
Kennedy, A. R.; Chen, B.; Dubenko, L.; Anand, N. K.; Tsang, T. H.; Nuss, J. M.;
Peto, C. J.; Rice, K. D.; Ibrahim, M. A.; Schnepp, K. L.; Shi, X.; Leahy, J. W.; Chen,
J.; Dalrymple, L. E.; Forsyth, T. P.; Huynh, T. P.; Mann, G.; Mann, L. W.; Takeuchi,
C. S. U.S. Patent 7,704,995, 2010.
Acknowledgments
The authors wish to thank colleagues in the Analytical and
Purification and DMPK groups for their contributions. X-ray crys-
tallographic data for compounds 14 and 16 were collected at the
Advanced Light Source at Lawrence Berkeley National Laboratory.
We would particularly like to thank Dr. Charles Eigenbrot for his
assistance in preparation of the crystallographic data.
10. For an explanation of ‘H2 region’ or ‘hydrophobic region II’: Traxler, P.; Furet, P.
Pharmacol. Ther. 1999, 82, 195.
11. Huang, S.; Garbaccio, R. M.; Fraley, M. E.; Steen, J.; Kreatsoulas, C.; Hartman, G.;
Stirdivant, S.; Drakas, B.; Rickert, K.; Walsh, E.; Hamilton, K.; Buser, C. A.;
Hardwick, J.; Mao, X.; Abrams, M.; Beck, S.; Tao, W.; Lobell, R.; Sepp-Lorenzino,
L.; Yan, Y.; Ikuta, M.; Zugay Murphy, J.; Sardana, V.; Munshi, S.; Kuo, L.; Reilly,
M.; Mahan, E. Bioorg. Med. Chem. Lett. 2006, 16, 5907.
Supplementary data
12. The matched pair compounds 1 & 13 exhibited Chk1 IC50 of 6.2 lM and 6.1 lM
respectively in the HTS Z-lyte assay.
13. The Cambridge Structural Database (CSD) provides only fused-ring examples
incorporating the 2-amino-3-phenylpyrazine moiety, CSD codes: FITZAJ01,
GOLDUG, LICZOM, SEWZOJ. Allen, F. H. Acta Crystallogr., Sect. B: Struct. Sci. 2002,
B58, 380.
14. Syntheses of compounds 14–16 are detailed in the patent literature: Chen, H.;
Dyke, H. J.; Ellwood, C.; Gancia, E.; Gazzard, L. J.; Goodacre, S.; Kintz, S.;
Lyssikatos, J.; Macleod, C.; Williams, K. U.S. Patent 8,501,765.
15. Bahekar, R. H.; Jain, M. R.; Jadav, P. A.; Goel, A.; Patel, D. N.; Prajapati, V. M.;
Gupta, A. A.; Modi, H.; Patel, P. R. Bioorg. Med. Chem. 2007, 15, 5950.
16. Synthesis of compound 17 is detailed in the patent literature: Chen, H.; Dyke,
H. J.; Gancia, E.; Gazzard, L. J.; Goodacre, S.; Lyssikatos, J.; Macleod, C.;
Williams, K. Patent Application US 2011/0124654.
Details of in silico docking procedures, biochemical and cellular
assays, in vivo pharmacokinetic experimental procedures, kinase
selectivity profiling data and crystallographic collection and refine-
ment data are available as supporting information. Coordinates
and structure factors for the ChK1 complexes for compounds 1,
13, 14 and 16 are deposited with the Protein Data Bank (PDB) with
accession codes 4QYE, 4QYF, 4QYG and 4QYH respectively. Supple-
mentary data associated with this article can be found, in the
online version, at http://dx.doi.org/10.1016/j.bmcl.2014.10.063.
These data include MOL files and InChiKeys of the most important
compounds described in this article.
17. Chen, P.; Luo, C.; Deng, Y.; Ryan, K.; Register, J.; Margosiak, S.; Tempczyk-
Russell, A.; Nguyen, B.; Myers, P.; Lundgren, K.; Kan, C. C.; O’Connor, P. M. Cell
2000, 100, 681.
18. Details of the checkpoint abrogation assay are given in Supplemental material.
SN-38 is the active metabolite of irinotecan: Kawato, Y.; Animimi, M.; Hirota,
Y.; Kuga, H.; Sato, K. Cancer Res. 1991, 51, 4187.
References and notes
19. Greco, W. R.; Bravo, G.; Parsons, J. C. Pharmacol. Rev. 1995, 47, 331. EC10fs is the
effective concentration of test compound required for 10-fold sensitization to
gemcitabine. Further details of the isobologram assay are given in
Supplemental material.
1. Carrassa, L.; Damia, G. Cell Cycle 2011, 10, 2121; (b) Thompson, R.; Eastman, A.
Br. J. Clin. Pharmacol. 2013, 76, 358; (c) Ma, C. X.; Janetka, J. W.; Piwnica-Worms,
H. Trends Mol. Med. 2011, 17, 88; (d) Tse, A. N.; Carvajal, R.; Schwartz, G. K. Clin.
Cancer Res. 1955, 2007, 13.
2. Ma, C. X.; Cai, S.; Li, S.; Ryan, C. E.; Guo, Z.; Schaiff, W. T.; Lin, L.; Hoog, J.;
Goiffon, R. J.; Prat, A.; Aft, R. L.; Ellis, M. J.; Piwnica-Worms, H. J. Clin. Invest.
2012, 122, 1541.
20. Selectivity screening was performed by SelectScreenÒ Kinase Profiling Services
(Invitrogen-Life Technologies, Madison, USA) and these data are provided in
Supplemental material.