Discovery of 2-(4-Chloro-3-(trifluoromethyl)phenyl)- N-(4-((6,7-dimethoxyquinolin-4-yl)oxy)phenyl)acetamide (CHMFL-KIT-64) as a Novel Orally Available Potent Inhibitor against Broad-Spectrum Mutants of c-KIT Kinase for Gastrointestinal Stromal Tumors

Article abstract of DOI:10.1021/acs.jmedchem.9b00280

Starting from our previously developed c-KIT kinase inhibitor CHMFL-KIT-8140, through a type II kinase inhibitor binding element hybrid design approach, we discovered a novel c-KIT kinase inhibitor compound 18 (CHMFL-KIT-64), which is potent against c-KIT wt and a broad spectrum of drug-resistant mutants with improved bioavailability. 18 exhibits single-digit nM potency against c-KIT kinase and c-KIT T670I mutants in the biochemical assay and displays great potencies against most of the gain-of-function mutations in the juxtamembrane domain, drug-resistant mutations in the ATP binding pocket (except V654A), and activation loops (except D816V). In addition, 18 exhibits a good in vivo pharmacokinetic (PK) profile in different species including mice, rats, and dogs. It also displays good in vivo antitumor efficacy in the c-KIT T670I, D820G, and Y823D mutant-mediated mice models as well as in the c-KIT wt patient primary cells which are known to be imatinib-resistant. The potent activity against a broad spectrum of clinically important c-KIT mutants combining the good in vivo PK/pharmacodynamic properties of 18 indicates that it might be a new potential therapeutic candidate for gastrointestinal stromal tumors.

Full text of DOI:10.1021/acs.jmedchem.9b00280

Subscriber access provided by ALBRIGHT COLLEGE
Article
Discovery of 2-(4-chloro-3-(trifluoromethyl)phenyl)-N-(4-((6,7-
dimethoxyquinolin-4-yl)oxy)phenyl)acetamide (CHMFL-KIT-64) as
a Novel Orally Available Potent Inhibitor against Broad Spectrum
Mutants of c-KIT Kinase for Gastrointestinal Stromal Tumors (GISTs)
Yun Wu, Beilei Wang, Junjie Wang, Shuang Qi, Fengming Zou, Ziping Qi, Feiyang
Liu, Qingwang Liu, Cheng Chen, Chen Hu, Zhenquan Hu, Aoli Wang, Li Wang,
Wenchao Wang, Tao Ren, Yujiao Cai, Mingfeng Bai, Qingsong Liu, and Jing Liu
J. Med. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.jmedchem.9b00280 • Publication Date (Web): 17 Jun 2019
Downloaded from http://pubs.acs.org on June 17, 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 64
Journal of Medicinal Chemistry
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
0
1
2
3
4
5
6
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 64
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
Discovery
trifluoromethyl)phenyl)-N-(4-((6,7-
dimethoxyquinolin-4-yl)oxy)phenyl)acetamide
CHMFL-KIT-64) as a Novel Orally Available
of
2-(4-chloro-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
(
Potent Inhibitor against Broad Spectrum Mutants of
c-KIT Kinase for Gastrointestinal Stromal Tumors
(GISTs)
Yun Wu , Beilei Wang1,2,9, Junjie Wang1,2,9, Shuang Qi , Fengming Zou1,3,4, Ziping Qi1,3,4
1
,9
1,9
,
Feiyang Liu1,3,4, Qingwang Liu , Cheng Chen , Chen Hu , Zhenquan Hu , Aoli Wang1,3,4, Li
3,4
1,2
1
1
Wang , Wenchao Wang1,3,4, Tao Ren3,4,5, Yujiao Cai , Mingfeng Bai , Qingsong Liu1,2,3,4,8*, Jing
1
,2
6
7
Liu1,3,4*
1
. High Magnetic Field Laboratory, Key Laboratory of High Magnetic Field and Ion Beam
Physical Biology, Hefei Institutes of Physical Science, Chinese Academy of Sciences, Hefei,
Anhui, 230031, China
2
3
. University of Science and Technology of China, Hefei, Anhui 230026, P. R. China
. Precision Medicine Research Laboratory of Anhui Province, Hefei, Anhui 230088, P. R.
China
ACS Paragon Plus Environment

Page 3 of 64
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
4
. Precision Targeted Therapy Discovery Center, Institute of Technology Innovation, Hefei
Institutes of Physical Science, Chinese Academy of Sciences, Hefei, Anhui 230088, P. R.
China
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
. Precedo Pharmaceuticals Inc, Hefei, Anhui, 230088, P. R. China
6
.
Department of General Surgery, Second Hospital Affiliated to Army Medical University,
Xinqiao Road, Chongqing 400037, P. R. China
7
8
9
.
Molecular Imaging Laboratory, Department of Radiology, University of Pittsburgh Cancer
Institute, University of Pittsburgh, Pittsburgh, Pennsylvania 15219, United States
. Institutes of Physical Science and Information Technology, Anhui University, Hefei, Anhui
230601, P. R. China
. These authors contribute equally
ABSTRACT
Starting from our previously developed c-KIT kinase inhibitor CHMFL-KIT-8140, through a
type II kinase inhibitor binding elements hybrid design approach, we discovered a novel c-KIT
kinase inhibitor compound 18 (CHMFL-KIT-64), which is potent against c-KIT wt and a broad
spectrum of drug resistant mutants with improved bioavailability. 18 exhibits single digit nM
potency against c-KIT kinase and c-KIT T670I mutant in the biochemical assay and displays
great potencies against most of the gain-of-function mutations in the juxtamembrane domain,
drug resistant mutations in the ATP binding pocket (except V654A) and activation loops (except
D816V). In addition, 18 exhibits good in vivo PK profile in different species including mice, rats
and dogs. It also displays good in vivo antitumor efficacy in the c-KIT T670I, D820G and
Y823D mutants mediated mice models as well as in the c-KIT wt patient primary cells which are
known to be imatinib-resistant. The potent activity against a broad spectrum of clinically
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 4 of 64
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
important c-KIT mutants combining the good in vivo PK/PD properties of 18 indicate that it
might be a new potential therapeutic candidate for GISTs.
INTRODUCTION
0
1
2
3
4
5
6
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
Gastrointestinal stromal tumors (GISTs) are the most common mesenchymal tumors in the
1
-2
human digestive tract . Approximately 20 individuals in a million are diagnosed this disease
3
each year in the United States . Identification of the c-KIT mutant was a major breakthrough in
the pathology of GISTs, which provided a therapeutic target for the treatment of GISTs. 95% of
GISTs express c-KIT and 80-85% GISTs carry activation mutations such as V559A/D/G,
V560D, L576P etc, which mainly located in the juxtamembrane region of c-KIT encoded by
4
exon 11 . c-KIT kinase belongs to type III receptor tyrosine kinase family which also includes
FLT3, CSF1R and PDGFR/.It plays critical roles in cell proliferation, migration and survival
upon activation in different physiological and pathological context. Drug discovery targeting c-
KIT has attracted extensive attentions in the past decades.
Imatinib (1), an inhibitor of c-KIT, PDGFRs and ABL kinases, was approved as the first-line
5
treatment for advanced GISTs . However, except 20% of the patients with GISTs exhibit
primary resistance to 1, most patients will eventually relapse due to secondary mutations upon
6
chronical treatment . Most of the secondary mutations including those occurred in the ATP
binding pocket such as T670I/E, V654A, outside the ATP binding pocket S709F and activation
loops such as D816E/H/V, D820E/G/Y, D822K, Y823D, A829P etc. Multi-targeted kinase
inhibitor sunitinib (2) was then approved as the second line treatment for the patients with GISTs
who are refractory to or intolerant of imatinib therapy. Although 2 could overcome some of
7
imatinib-resistant mutations such as V654A and T670I , certain imatinib-resistant mutants
8
located in the activation loop remain resistant to it . Later, multi-targeted kinase inhibitor
ACS Paragon Plus Environment

Page 5 of 64
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
regorafenib (3) was approved as the third line therapy for the patients with GISTs who are not
9
amenable to surgery and refractory to both 1 and 2 . Similarly, 3 also could not overcome some
of the secondary resistance mutations in the c-KIT activation loops. Furthermore, both 2 and 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
10
have limited overall clinical response and exhibit severe side effects . In addition, ponatinib (4),
which was clinically approved for CML, could also potently inhibit KIT exon 11 primary
1
1
mutants and a range of secondary mutants, including those within the activation loop . However,
it was suspended by FDA due to serious side effects in 2013 and later was listed in black box
warning status for clinical usage. Cabozantinib (5) (a TKI targeting c-KIT, MET, VEGFR2 and
AXL) also displays potency against different c-KIT mutations in the patient-derived xenograft
1
2
(PDX) models of GISTs . In addition, BLU285 (6) which is under clinical investigation is a
potent and highly selective inhibitor of mutated c-KIT and PDGFRα. It exhibits good antitumor
1
3
14
effect in the PDX models of GISTs , but it could not overcome the c-KIT T670I mutant .
Furthermore, DCC-2618 (7)15 and AZD3229 (8), which are under clinical and preclinical
14
development respectively, exhibited different broad mutants inhibition profile . However, given
the fact that there are over a dozen of different c-KIT mutants emerged in the clinic, there is still
a great demand to develop new therapeutics which bears different potency profiles against the
broad spectrum of the c-KIT mutants.
Starting from our previously developed quinoline based type II c-KIT kinase inhibitor
1
6
CHMFL-KIT-8140 (9) , which could also overcome a variety of imatinib-resistant secondary
mutants including T670I but with poor PK properties. Through type II kinase inhibitor binding
17
elements hybrid drug design approach , we designed a series of compounds and screened them
with transformed BaF3 cells which led to the disovery of a novel highly potent c-KIT kinase
inhibitor CHMFL-KIT-64 (18). It was potent against a broad spectrum of c-KIT primary gain-of-
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 6 of 64
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 mutants, secondary mutants and A-loop mutants. In addition, 18 also displayed good in
vivo PK properties and antitumor efficacy in different preclinical mice models of GISTs.
N
O
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
F
N
H
H
H
N
N
CF3
Cl
N
N
N
H
N
HN
O
F
HN
N
N
H
O
O
O
O
N
H
3
(Regorafenib)
1
(Imatinib)
2 (Sunitinib)
F
H N
2
H
H
N
N
N
N
N
CF3
N
N
O
O
O
N
N
O
N
F
N
N
N
H
O
O
N
N
N
N
N
4
(Ponatinib)
5 (XL-184)
6
(BLU-285)
H
N
N
N
Br
F
N
O
O
F
HN
N
N
N
N
H
H
N
HN
N
O
O
O
7
(DCC-2618)
8 (AZD3299)
Figure 1. Chemical structures of representative inhibitors that bear c-KIT kinase activity.
Tail (R1)
Linker 2
(L2)
O
Linker 1
(L1)
Linker 1 Linker 2
Tail (R1)
CF3
N
H
(
L1)
(L2)
H
N
Potency improved
PK improved
N
O
O
O
N
Cl
O
N
N
H
H
O
O
CF3
O
O
PD improved
Selectivity improved
9
18
(CHMFL-KIT-64)
N
(CHMFL-KIT-8140)
Head
Hinge binding
Head
Hinge binding
Figure 2. Schematic illustration of discovery of compound 18 (CHMFL-KIT-64) from 9.
RESULTS AND DISCUSSION
ACS Paragon Plus Environment

Page 7 of 64
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Structure-Activity Relationships (SAR) Exploration.
Compound 9 is a highly potent type II c-KIT kinase inhibitor which is capable of overcoming
the T670I gatekeeper mutant; however the poor PK profiles limit its further development. In
order to address this deficiency, we initiated a medicinal chemistry study that resulted in the
identification of a highly potent inhibitor which could overcome a broad spectrum of c-KIT
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
17
mutants. Using the type II kinase inhibitor binding elements hybrid design approach , we started
to explore the SAR by varying the “head”, “linker” and “tail” moieties. We first fixed the hinge
1
binding part as 6,7-dimethoxyquinolin and explored the “linker2” (L2) and “tail” (R ) moieties.
The growth inhibition of IL-3 independent BaF3 cells (GI ) expressing the c-KIT wt (BaF3-tel-
5
0
c-KIT) and c-KIT T670I (BaF3-tel-c-KIT-T670I), which has been used for the development of 9
and shows a correlation with the biochemical IC5016, was used as the primary readout to evaluate
the potency and selectivity of the new compounds (Table 1). Replacement of the urea of L2 in 9
1
with acetamide and the “tail” moiety R with simple phenyl ring (10) or pyridine (11) resulted in
great activity loss. Switching the L2 in 10 to cyclopropanecarboxamide (12) did not achieve
1
apparent potency. Interestingly, introduction of the CF group in the para-position of R (13)
3
significantly increased the activity both to the c-KIT wt (GI = 0.022 μM) and the c-KIT T670I
5
0
(
GI = 0.011 μM), meanwhile it showed good selectivity against parental BaF3 cells (GI =
5
0
5
0
2.16 μM). Analogous compound with the m-CF substituent (14) also achieved good potency
3
against the c-KIT wt and c-KIT T670I (GI = 0.020 μM and 0.001 μM, respectively). However,
5
0
1
changing the substituent of R from m-CF to m-F (15) or m-OMe (16) all resulted in significant
3
1
activity loss. Addition of the p-F substituent to the R moiety of 14 (17) lost about 13-fold
activity against c-KIT T670I in comparison with 14, although it still displayed similar potency
toward c-KIT wt. Interestingly, changing the F atom to a larger Cl atom (18) not only remarkably
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 8 of 64
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
increased the potency against c-KIT wt (GI = 0.001 μM) and c-KIT T670I (GI = 0.004 μM),
5
0
50
but also improved the selectivity window toward the parental BaF3 cells (GI = 5.97 μM).
5
0
However, further increasing the size of the substituent with a methyl group (19) slightly
0
1
2
3
4
5
6
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
decreased the activity toward the c-KIT wt and c-KIT T670I, even though it still kept good
selectivity against the parental BaF3 cells. Replacement of the CF group in 19 with a methyl
3
group (20) displayed similar trends to 19. Further investigating the effects of other substituents at
the meta- and para-positions of the benzene ring, such as 3, 4-dimethoxy (21), 3, 4-difluoro (22)
and 3, 4-dichloro (23), all resulted in significant activity loss against c-KIT wt. Installing a F or
CF group at the ortho-position of the benzene ring (24-26) also led to activity loss. Relatively
3
1
larger groups at R , e.g., benzodioxole (27) and naphthyl (28), both resulted in activity loss.
1
Additionally, aliphatic rings such as cyclohexyl (29) and N-methyl piperazine (30) at R both
significantly decreased the potency. Furthermore, reversed amide counterpart of the L2 in 13
(31) and 14 (32) lost the activity to both c-KIT wt and c-KIT T670I as well.
L2
R1
O
O
O
N
Table 1. SAR Exploration Focused on the L2/R Moitiesa
1
Parental BaF3 BaF3-tel-c-KIT BaF3-tel-c-KIT-
Compd.
L2
-
R1
(
GI : μM)
(GI : μM)
T670I (GI : μM)
5
0
50
50
9
-
>10
0.057
0.063
ACS Paragon Plus Environment

Page 9 of 64
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
O
O
O
O
O
O
O
1
0
>10
5.13
>10
2.16
1.89
1.0
>10
>10
N
H
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
11
4.71
6.99
N
H
1
1
1
2
3
4
>10
>10
N
H
CF3
0.022
0.020
0.923
0.293
0.024
0.001
0.017
0.027
0.011
0.001
0.342
0.339
0.013
0.004
0.014
0.015
N
H
N
H
CF3
15
N
H
F
1
1
1
6
7
8
2.78
1.03
5.97
7.26
6.42
N
H
OMe
O
O
O
O
F
N
H
CF3
Cl
N
H
CF3
19
N
H
CF3
2
0
N
H
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 10 of 64
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
O
O
O
O
O
O
O
O
O
O
OMe
OMe
2
1
4.13
0.67
2.03
6.70
1.93
0.704
3.10
6.23
2.47
8.70
6.18
0.319
0.312
0.342
0.424
2.78
0.179
0.309
0.037
0.309
1.70
N
H
F
F
0
1
2
3
4
5
6
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
22
N
H
Cl
Cl
2
2
2
3
4
5
N
H
N
H
CF3
F
N
H
F
F
26
0.040
0.11
0.285
0.167
0.034
0.362
8.52
N
H
F
O
O
2
2
2
7
8
9
N
H
0.112
0.635
7.33
N
H
N
H
N
30
N
N
H
O
CF3
3
1
2.90
0.668
N
H
ACS Paragon Plus Environment

Page 11 of 64
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
O
3
2
10
2.35
4.61
N
CF3
H
a
All GI values were obtained by triplet testing.
5
0
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
Since 18 exhibited potent antiproliferative efficacy in the BaF3-tel-c-KIT and BaF3-tel-c-KIT-
1
T670I cells, we then fixed the L2 moiety as acetamide and R moiety as 4-chloro-3-
(trifluoromethyl)phenyl and further explored the “linker 1” (L1) moiety (Table 2). Replacement
of the oxygen atom linkage in 18 with –NH (33), –NMe (34) or S atom (35) all resulted in
significant activity loss against c-KIT wt and c-KIT T670I. Therefore, we kept the phenoxy
group and explored a series of substituents of the benzene ring in L1 moiety. Addition of one F
atom at the ortho-position of the acetamide (36) increased the activity against c-KIT wt (GI =
5
0
0
.049 M) and c-KIT T670I (GI = 0.018 M). Changing the F atom to Cl atom (37) or methyl
5
0
group (38) remained similar activities compared with 36. Switching the F atom from the ortho-
to meta-position (39) led to decrease of the activity against c-KIT wt (GI50 = 0.116 M).
Variation of different substituents at this position such as Cl (40), methyl (41), methoxyl (42),
trifluoromethyl (43) and nitrile (44) groups all resulted in activity loss in comparison with 18.
H
N
L1
O
Cl
O
O
CF3
N
Table 2. SAR Exploration Focused on the Substitution (L1) Moietya
Parental BaF3
GI : μM)
BaF3-tel-c-KIT BaF3-tel-c-KIT-T670I
(GI : μM) (GI : μM)
Compd.
L1
(
50
50
50
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 12 of 64
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
3
3
3
4
2.21
2.67
3.09
3.21
8.22
4.86
1.99
9.45
7.05
8.48
0.742
4.28
0.4
HN
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
8.46
35
0.154
0.049
0.081
0.061
0.116
0.078
0.105
0.034
0.11
S
F
3
3
3
6
7
8
0.018
0.070
0.075
0.021
0.056
0.009
1.56
O
O
Cl
O
F
39
40
O
Cl
O
4
4
1
2
O
MeO
O
ACS Paragon Plus Environment

Page 13 of 64
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
F3C
4
4
3
4
2.92
9.97
3.0
2.84
0.21
O
NC
O
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
0.989
a
All GI values were obtained by triplet testing.
5
0
Selectivity Profiling of Compound 18
Since 18 displayed the best potency against both c-KIT wt and T670I and meanwhile
exhibited the best selectivity window to parental BaF3 cells, we then further screened it in a
panel of broad spectrum mutants with BaF3 cell lines expressing a variety of clinically relevant
secondary mutants of c-KIT kinase, including those in the juxtamembrane domain (V559D/A/G,
V560D, L576P), ATP binding pocket (V654A, V654A/V559D, T670E/I, T670I/V559D, S709F)
and the activation-loop (D816E/H/V, D820G/Y, N822K, Y823D and A829P) (Table 3). The
results showed that 18 displayed the best potency against c-KIT wt and selectivity window to
parental BaF3 cells compared with 1 and 2. In addition, 18 exhibited similar or better activities
compared with 1 and 2 against all of the primary gain-of-function mutations of c-KIT in exon 11
including V559A/D/G, V560D and L576P. Furthermore, 18 showed better potencies than 2
against thecompound 1-resistant c-KIT gatekeeper residue T670I/E mutation. 18 and 2 could
also overcome the compound 1-resistant secondary mixed mutation T670I/V559D and displayed
better potencies against S709F mutation which is located in exon 14 but outside the ATP binding
pocket. Only 2 but not 1 or 18 was resistant to V654A mutant which located in exon 13 and
inside the ATP binding pocket. Interestingly, 18 exhibited much better activities than 1 and 2
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 14 of 64
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
against activation loop mutants including D816E/H, D820E/G/Y, D822K, Y823D and A829P
except D816V which was resistant to all of these three compounds.
Table 3. Antiproliferative Effects of Compounds 1, 2 and 18 against a Panel of c-KIT
0
1
2
3
4
5
6
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
wt/Mutants Transformed BaF3 Cells.
Compd. 1
Compd. 2
Compd. 18
Cells
c-KIT Status
(GI : M)
(GI : M)
(GI : M)
5
0
50
50
BaF3
Parental
>10
0.35
1.01
0.003
0.02
5.97
0.001
0.021
0.006
<0.0003
0.004
0.001
0.042
0.32
BaF3-tel-c-KIT
WT
Exon 11/Juxtamembrane
domain
BaF3-tel-c-KIT-V559A
BaF3-tel-c-KIT-V559D
BaF3-tel-c-KIT-V559G
BaF3-tel-c-KIT-V560D
BaF3-tel-c-KIT-L576P
BaF3-tel-c-KIT-V654A/V559D
BaF3-tel-c-KIT-V654A
BaF3-tel-c-KIT-T670E
BaF3-tel-c-KIT-T670I
BaF3-tel-c-KIT-T670I/V559D
BaF3-tel-c-KIT-S709F
BaF3-tel-c-KIT-D816E
0.905
0.01
Exon 11/Juxtamembrane
domain
0.001
0.004
0.035
0.006
0.019
0.005
0.089
0.01
Exon 11/Juxtamembrane
domain
0.008
0.003
0.085
0.608
1.41
Exon 11/Juxtamembrane
domain
Exon 11/Juxtamembrane
domain
Exon13/11/ATP binding
pocket
Exon 13/ATP binding
pocket
Exon14/ATP binding
pocket
4.08
0.071
0.004
0.045
0.017
0.016
Exon 14/ATP binding
pocket
>10
Exon14/11/ATP binding
pocket
>10
0.012
0.028
0.059
Exon14/outside the ATP
binding pocket
0.115
0.174
Exon17/Activation loop
ACS Paragon Plus Environment

Page 15 of 64
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
BaF3-tel-c-KIT-D816H
BaF3-tel-c-KIT-D816V
BaF3-tel-c-KIT-D820E
BaF3-tel-c-KIT-D820G
BaF3-tel-c-KIT-D820Y
BaF3-tel-c-KIT-N822K
BaF3-tel-c-KIT-Y823D
BaF3-tel-c-KIT-A829P
Exon17/Activation loop
0.651
>10
0.315
1.72
0.097
1.73
Exon17/Activation loop
Exon17/Activation loop
Exon 17/Activation loop
Exon 17/Activation loop
Exon17/Activation loop
Exon17/Activation loop
Exon18/Activation loop
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
0.035
0.337
0.435
1.47
0.093
0.389
0.172
0.384
0.704
0.018
0.001
0.009
0.003
0.026
0.035
0.022
5.87
0.58
a
All GI values were obtained by triplet testing.
5
0
We then moved forward to investigate the selectivity profile of 18 against other protein
1
8
kinases with DiscoverX’s KINOMEscan assay. The data demonstrated that 18 bears a good
selectivity (S score (1) = 0.02) in a panel of 468 kinases and mutants at 1 μM concentration
(
Figure 3A and Supplementary Table 1). Besides high binding affinity to c-KIT and c-KIT
mutant kinases, 18 also displayed strong binding to CSF1R, DDR1, FLT4, PDGFRα, PDGFRβ,
LOK and RET kinases (percent activity remaining less than 1% at 1 μM of 18). Given the fact
that KINOMEscan is a binding assay and sometimes it cannot really reflect the inhibitory
activity of a compound, we then used Invitrogen’s Z’lyte and Promega’s ADP-Glo based
biochemical activity assays with the purified kinase proteins to confirm the inhibition activity of
1
8 against those kinase targets (Figure 3B). The results showed that 18 could potently inhibit the
activity of c-KIT wt (IC : 4 nM), c-KIT T670I (IC : 8 nM), CSF1R (IC : 18 nM), PDGFR
50
50
50
(
IC : 25 nM) and RET (IC : 34 nM), and it relatively less potently inhibited DDR1 (IC : 135
50 50 50
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 16 of 64
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
nM), FLT4 (IC : 121 nM) and PDGFR (IC : 97 nM). We also tested the antiproliferative
5
0
50
effects of 18 against CSF1R, DDR1, FLT4, PDGFRα, PDGFRβ and RET kinases in the
transformed BaF3 cells. The results showed that 18 exhibited strong inhibition to CSF1R,
PDGFRα and PDGFRβ kinases. This is not surprising because c-KIT, CSF1R, PDGFRα and
PDGFRβ kinases all belong to the type III receptor tyrosine kinase family and the ATP binding
pockets of these kinases are highly conserved. These data indicated that c-KIT, PDGFRα,
PDGFRβ and CSF1R kinases are the main targets of 18.
0
1
2
3
4
5
6
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 3. Biochemical examination of compound 18. (A) Treespot demonstration of selectivity
profiling of 18 with DiscoverX KINOMEscan technology. Measurements were performed at a
concentration of 1 μM of the inhibitor. The affinity was defined with respect to a DMSO control.
(B) The biochemical tests of 18 with Invitrogen SelectScreen platform or Promega’s ADP-Glo
assay (IC ) and the antiproliferative effects of 18 on a panel of transformed BaF3 cells (GI )
50
50
ACS Paragon Plus Environment

Page 17 of 64
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
against the kinase targets that demonstrated strong binding to 18 with a percent control number
less than 1% in the KINOMEscan profiling.
In order to better understand the sensitivity of 18 among different c-KIT mutants and the SAR,
we then compared the binding modes of 18 with 1 and 2. As shown in the X-ray crystal structure
of 1 with c-KIT wt, 1 fitted well with c-KIT wt via four hydrogen bonds, including the N atom of
pyridine ring with Cys673 in the hinge binding area, the NH of aminopyrimidine with Thr670,
and two canonical type II inhibitor hydrogen bonds between the amide and Glu640/Asp810
(Figure 4A). Compound 1 was resistant to c-KIT V654A because the conformational change
affected the binding of 1 to c-KIT19 (Figure 4B). The hydrogen bond of the NH of
aminopyrimidine with Thr670 was lost when Thr670 mutated to Ile670, which resulted in the
resistance to 1 (Figure 4C). As shown in the X-ray crystal structure of 2 with c-KIT wt, 2 bound
to inactivated c-KIT wt through two hydrogen bonds with Cys673 and Glu671 in the hinge
binding area with the N and O atom of hydroxy indole (Figure 4D). The mutants of Val654 and
Thr670 did not affect the binding of 2, therefore they were still sensitive to 2 (Figure 4E-F).
Docking of 18 into c-KIT wt (PDB ID: 6GQK) showed that 18 adopted a canonical type II
binding mode (Figure 4G). As expected, the quinoline nitrogen formed a hydrogen bond with
Cys673 in the hinge binding area. The amide in the linker formed two hydrogen bonds with
Asp810 in the DFG motif and Glu640 in the c-Helix. The “tail” part occupied the hydrophobic
pocket generated by the DFG-out shifting. In the homology model of c-KIT T670I mutant, 18
adopted a similar type II binding mode (Figure 4H). The V654A mutant might weaken the van
der Waals interactions of 18 or the conformational change resulted in a similar effect to 1 which
made it insensitive to 18. The O-bridged phenyl moiety in 18 oriented to an angle that provided
enough space for the bulky isoleucine residue, which could explain its potency against T670I
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 18 of 64
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
mutant (Figure 4I). The D816V mutant was also resistant to 18 but it was located outside of the
ATP binding pocket and hence very hard to model.
The modeling study also helps to understand the SAR observed in Table 1. The residues of
Val643, Leu644, His790 and Cys809 formed the hydrophobic pocket generated by DFG-out
0
1
2
3
4
5
6
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
shifting. The trifluoromethyl in the “tail” (R ) pointed to Val643 and Leu644, which helped to
gain better potency due to the hydrophobic interaction. In addition, the substituent p-Cl of phenyl
in the “tail” moiety could achieve higher binding affinity due to the halogen bond interaction
2
0
with His790 . Loss of hydrophobicity (29, 30) or halogen bond resulted in the activity loss. The
amide moiety (L2) which could form two typical hydrogen bonds with Glu640 and Asp810 were
critical for the binding. Elongation of L2 (31, 32) would result in the loss of these bonds. In
addition, π-π stack between the phenyl in the L1 moiety and Phe811 might help to stabilize the
binding conformation. Smaller substituents in L1 moiety (36-40) would not affect the rotation of
phenyl. However, larger substituents in L1 moiety (43-44) might distort the conformation and
lead to decreased activity to c-KIT wt and T670I mutant.
ACS Paragon Plus Environment

Page 19 of 64
Journal of Medicinal Chemistry
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
0
1
2
3
4
5
6
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 4. Molecular modeling analysis of the binding modes of c-KIT wt/V654A/T670I with
compounds 1, 2 and 18. (A) X-ray of 1 with c-KIT wt (PDB code:1T46); (B) Binding mode of 1
with c-KIT V654A homology model (generated based on PDB code:1T46); (C) Binding mode of
1
with c-KIT T670I homology model (generated based on PDB code:1T46); (D) X-ray of 2 with
c-KIT wt (PDB code:3G0E); (E) Binding mode of 2 with c-KIT V654A homology model
generated based on PDB code: 3G0E); (F) Binding mode of 2 with c-KIT T670I homology
(
model (generated based on PDB code:3G0E); (G) Binding mode of 18 with c-KIT wt (PDB
code:6GQK); (H) Binding mode of 18 with c-KIT V654A homology model (generated based on
PDB code:6GQK); (I) Binding mode of 18 with c-KIT T670I homology model (generated based
on PDB code:6GQK).
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 20 of 64
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
Cellular evaluation of compound 18
Next, we examined the antiproliferative effects of 18 against a panel of established GISTs
cancer cell lines. Not surprisingly, it did potently inhibit the growth of c-KIT dependent GISTs
0
1
2
3
4
5
6
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
cancer cells such as GIST-T1 (GI : 0.006 μM) and GIST-882 (GI : 0.013 μM) which are also
50
50
sensitive to 1 and 2 (Table 4). In the compound 1-resistant cell line GIST-T1-T670I (generated
by CRISPR knock-in method) and GIST-5R (generated by continuous treatment of 1) cell lines,
both 18 and 2 exhibited strong growth inhibition effects. In addition, 18 also displayed good
selectivity window over GIST-48B (GI : 1.37 μM), whose growth is c-KIT independent. This
5
0
further confirmed that 18 exerted its antiproliferative effect against GISTs cancer cells through
the on-target inhibition of c-KIT but not general toxicity.
Table 4. Antiproliferation Effects of Compounds 1, 2 and 18 against a Variety of GISTs Cancer
Cell Linesa
Compd.1
GI :M)
Compd.2
Compd.18
Cell line
GIST-T1
c-KIT Status
∆560-578
(
(GI :M)
(GI :M)
5
0
50
50
0.015
0.048
>10
0.011
0.014
0.004
0.035
0.74
0.006
0.013
0.011
0.073
1.37
GIST-882
K642E
GIST-T1-T670I
GIST-5R
∆560-578/T670I
∆560-578/T670I
KIT negative
>10
GIST-48B
>10
a
All GI values were obtained by triplet testing.
5
0
We also investigated the effects of 18 on the c-KIT mediated signaling pathways in the c-KIT
wt and c-KIT T670I driven GISTs cell lines GIST-T1, GIST-T1-T670I and GIST-5R
ACS Paragon Plus Environment

Page 21 of 64
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
respectively (Figure 5A). 18 blocked the autophosphorylation of c-KIT pY703, pY719, pY823 in
GIST-T1, GIST-T1-T670I and GIST-5R cells at the concentration of 30 nM and also remarkably
inhibited the downstream signaling mediators pAKT (T308/S473), pS6 (S235/236) and pERK
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
(
T202/204) (EC less than 100 nM). 1 and 2 displayed similar effects on the signaling pathways.
50
In the compound 1-resistant cell lines GIST-T1-T670I and GIST-5R (c-KIT T670I), 18 also
significantly affected the phosphorylation of c-KIT pY703, pY719, pY823 and downstream
mediators (EC less than 100 nM). Similar effects were observed from 2 but not 1. It is worthy
5
0
to note that both 18 and 2 strongly inhibited the phosphorylation of STAT3 in GIST-T1-T670I
and GIST-5R cells but moderately in GIST-T1 cells. Furthermore, 18 and 2 affected the
autophosphorylation of AKT308 in GIST-T1 and GIST-5R cells but not GIST-T1-T670I cells.
These discrepancies reflected the different genetic backgrounds of these cells. In addition, 18
could induce dose-dependent cell apoptotic death (by examining the cleaved PARP and cleaved
caspase 3) and arrest the cell cycle into G0/G1 phase in all of these three cell lines (Figure 5B-
C).
In order to exclude the off-target effects (such as PDGFRs and CSF1R) on the signaling
pathway, we also detected the expression and phosphorylation of PDGFRα/β and CSF1R kinases
in GIST-T1 cells (Supplementary Figure 1). The results showed that only PDGFRα expressed
and phosphorylated in GIST-T1 cells and 18 could inhibit its autophosphorylation at the Y849
site. CSF1R did not express in GIST-T1 cells and phosphorylated PDGFRβ was not detected
either. In order to further exclude the effects of downstream mediators caused by the inhibition
21
of PDGFRα, we also used a highly selective PDGFRα inhibitor CHMFL-PDGFR-159 to test
the PDGFR signaling pathway and the antiproliferative effect in GIST-T1 cells (Supplementary
Figure 2). The data demonstrated that although p-PDGFR was potently inhibited, the
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 22 of 64
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
downstream mediators such as AKT, STAT3 and ERK were less potently inhibited. What is
more, the antiproliferative effect of selective PDGFRα inhibitor was much less than 1, which
further proved that the PDGFRα inhibition did not play essential roles in the inhibition of 18 to
c-KIT mediated signaling pathway. All these data suggested that 18 exerted its inhibitory effect
to the GISTs cancer cells through the on-target inhibition of c-KIT.
0
1
2
3
4
5
6
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

Page 23 of 64
Journal of Medicinal Chemistry
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
0
1
2
3
4
5
6
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. Cellular effects of compound 18 in GISTs cell lines GIST-T1, GIST-T1-T670I and
GIST-5R. (A) Effects of 18 on c-KIT mediated signaling pathways. (B) Effects of 18 on
apoptosis. (C) Effects of 18 on cell cycle progression.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 24 of 64
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
Inhibition of the proliferation of primary GISTs patient cells
We then investigated the effects of 18 in a more pathologically relevant setting of ex vivo
culture of primary cells derived from GISTs patients (PDC) (Figure 6). In cells from patient A
which harbor the c-KIT V559D gain-of-function mutation, 18 displayed better growth inhibitory
effect than 1 although both of them exhibited dose-dependent antiproliferative effects.
Interestingly, 18 also exhibited potent and dose-dependent inhibitory effect against cells from
patient B harboring c-KIT wt which are known to be compound 1-resistent22-24. These data
suggested that 18 might be more potent and effective than 1 in the genetically more complicated
GISTs cells and therefore indicated potential clinical benefits.
0
1
2
3
4
5
6
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. Antiproliferative effects of compounds 18 and 1 against primary cells derived from
GISTs patients upon 6 days of drug treatment with the Cell Titer-Glo assay.
In vivo PK/PD and efficacy evaluation of compound 18
The in vivo PK properties of 18 were evaluated in different species including mice, Sprague
Dawley rats and beagle dogs following intravenous injection and oral administration (Table 5
and Supplementary Figure 3). 18 exhibited good drug exposure (AUC ) in all of these three
0-t
.
.
species, i.e., 25054 ng/mL h at 10 mg/kg in rats, 12608 ng/mL h at 10 mg/kg in mice and 19065
ACS Paragon Plus Environment

Page 25 of 64
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
.
ng/mL h at 5mg/kg in dogs). This compound displayed different half-life (T ) in these three
1
/2
different species after oral administration, i.e., 4.5 h in mice and relative longer time in rats (6.4
h) and in dogs (19.4 h). These results indicated that 18 was metabolized very slowly in dogs
compared with in mice and rats. 18 possessed acceptable bioavailability in mice (F = 43%), rats
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
(
F = 50%) and dogs (F = 81%). These PK properties indicated that 18 would be suitable for oral
administration.
We also evaluated the pharmacodynamic profile of 18 in the GIST-T1-T670I mouse model at
the 80 mg/kg/day dosage (p.o.) for the impact on phosphorylation levels of c-KIT and
downstream signaling mediators such as ERK and AKT etc (Figure 7 and Supplementary Figure
4). Meanwhile, we also monitored the drug concentration in the blood plasma. Following a
single dose treatment, in 4 h, 8 h and 12 h time courses the phosphorylation levels of c-
KIT/ERK/AKT were decreased time-dependently and then started to get back at 24 h. The drug
concentration in the plasma reached highest level in 4 h and then gradually decreased; even up to
2
4 h it did not reduce to the basal level. These data indicated that the drug concentration in the
plasma was well correlated to the phosphorylation levels of the signaling mediators. In addition,
it is worthy to note that the plasma protein binding rate of 18 was over 90%.
Table 5. Pharmacokinetic Study of Compound 18 in Mice, Sprague Dawley Rats and Beagle
Dogsa
1
8
Rats
Mice
Beagle Dogs
i.v.
p.o.
i.v.
p.o.
i.v.
p.o.
Parameter
AUC(0-t)
(1 mg/kg)
(10 mg/kg)
(1mg/kg)
(10 mg/kg)
(1mg/kg)
(5 mg/kg)
2
553±155.1
0±0
25054±7060
9±0
2989
0.03
12608
5323±569
19065±5696
5.33±1.15
(
ng/mL·h)
Tmax (h)
4
0.033±0
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 26 of 64
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
T1/2 (h)
15.85±3.72 6.41±0.34
5.4
520
--
4.5
1390
43
20.61±4.82
1107±117
--
19.4±4.02
1163±288
81
Cmax
660±17
--
1643±462
50
(ng/mL)
F (%)
0
1
2
3
4
5
6
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
aAll testing data were obtained from three independent mice (±SD).
Figure 7. Pharmacodynamic effect of 18 on tumor p-c-KIT, p-ERK and p-AKT in GIST-T1-
T670I mouse xenograft model with a single dose of 80 mg/kg/day (p.o.).
We finally evaluated the in vivo antitumor efficacy of 18 in three transformed BaF3 cells, i.e.,
BaF3-tel-c-KIT-T670I (1 resistant, 2 and 18 sensitive), BaF3-tel-c-KIT-D820G (18 sensitive, 1
and 2 resistant) and BaF3-tel-c-KIT-Y823D (18 sensitive, 1 and 2 resistant) and three GISTs
intact cells, i.e., GIST-T1 (1, 2 and 18 sensitive), GIST-T1-T670I (1 resistant, 2 and 18 sensitive)
and GIST-5R (1 resistant, 2 and 18 sensitive) inoculated xenograft mouse models by oral
administration (Figures 8 and 9). All dosages of 18 (10 mg/kg/day, 20 mg/kg/day, 40 mg/kg/day
and 80 mg/kg/day) did not affect the animal weights which indicated that there was no general
ACS Paragon Plus Environment

Page 27 of 64
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
cytotoxicity at those dosages (Figures 8A-C, 9A-C left lane). During 11 days of continuous
treatment, 18 dose-dependently inhibited the BaF3-tel-c-KIT-T670I tumor progression and
exhibited almost 100% TGI (tumor growth inhibition) at a dosage of 100 mg/kg/day (Figure 8A).
After 9 days of continuous treatment, 18 dose-dependently inhibited the BaF3-tel-c-KIT-D820G
and BaF3-tel-c-KIT-Y823D tumor progression and exhibited 92.9% and 90.2% TGI respectively
at a dosage of 80 mg/kg/day (Figure 8B-C). Comparably, 2 did not show apparent efficacy in the
BaF3-tel-c-KIT-D820G or BaF3-tel-c-KIT-Y823D xenograft mouse model. In the GIST-T1 cell-
inoculated xenograft mouse models, 18 dose-dependently inhibited the tumor progression and a
dosage of 20 mg/kg/day exhibited 95.6% TGI (Figure 9A) which was much better than the
efficacy of 1 at 100 mg/kg/day. In addition, for the 1-resistant cell lines GIST-T1-T670I and
GIST-5R, 18 dose-dependently inhibited the tumor progression and a dosage of 40 mg/kg/day
exhibited 76.4% and 96.9% TGI, respectively (Figure 9B-C). It is worth noting that the mice
treated with 2 started to die at the third day in the GIST-T1-T670I xenograft models and there
was only one mouse alive at the 28th day. However, the mice treated with 18 were in good
condition. All these in vivo data suggested that 18 bore suitable anti-GISTs efficacy in different
c-KIT mutant spectra from 1 and 2 and it might have a potential usage in the clinic.
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 28 of 64
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
0
1
2
3
4
5
6
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 8. (A) The antitumor efficacy of 18 in the BaF3-tel-c-KIT-T670I xenograft mouse model.
Female nu/nu mice bearing established BaF3-tel-c-KIT-T670I tumor xenografts were treated
with 18 at 40 and 100 mg/kg/day dosage or vehicle. (B) The antitumor efficacy of 18 in the
BaF3-tel-c-KIT-D820G xenograft mouse model. Female nu/nu mice bearing established BaF3-
tel-c-KIT-D820G tumor xenografts were treated with 18 at 40 and 80 mg/kg/day dosage or
vehicle. Daily oral administration was initiated when BaF3-tel-c-KIT-D820G tumors had
3
reached a size of 200−400 mm . (C) The antitumor efficacy of 18 in the BaF3-tel-c-KIT-Y823D
ACS Paragon Plus Environment

Page 29 of 64
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
xenograft mouse model. Female nu/nu mice bearing established BaF3-tel-c-KIT-Y823D tumor
xenografts were treated with 18 at 40 and 80 mg/kg/day dosage or vehicle.
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 9. (A) The antitumor efficacy of 18 in the GIST-T1 xenograft mouse model. Female
nu/nu mice bearing established GIST-T1 tumor xenografts were treated with 18 at 20 and 40
mg/kg/day dosage or vehicle. (B) The antitumor efficacy of 18 in the GIST-T1-T670I xenograft
mouse model. Female nu/nu mice bearing established GIST-T1-T670I tumor xenografts were
treated with 18 at 20 and 40 mg/kg/day dosage or vehicle. (C) The antitumor efficacy of 18 in
the GIST-5R xenograft mouse model. Female nu/nu mice bearing established GIST-5R tumor
xenografts were treated with 18 at 20 and 40 mg/kg/day dosage or vehicle.
ACS Paragon Plus Environment