A Pyrazolo[3,4-D]pyrimidin-4-amine Derivative Containing an Isoxazole Moiety Is a Selective and Potent Inhibitor of RET Gatekeeper Mutants

Article abstract of DOI:10.1021/acs.jmedchem.5b01522

Aberrant RET kinase signaling plays critical roles in several human cancers such as thyroid carcinoma. The gatekeeper mutants (V804L or V804M) of RET are resistant to currently approved RET inhibitors such as cabozantinib and vandetanib. We, for the first time, report a highly selective and extremely potent RET inhibitor, 6i rationally designed. Compound 6i inhibits strongly RET gatekeeper mutants and other clinically relevant RET mutants as well as wt-RET. This substance also significantly suppresses growth of thyroid cancer-derived TT cell lines and Ba/F3 cells transformed with various RET mutants. Docking studies reveal that the isoxazole moiety in 6i is responsible for binding affinity improvement by providing additional site for H-bonding with Lys758. Also, 6i not only substantially blocks cellular RET autophosphorylation and its downstream pathway, it markedly induces apoptosis and anchorage-independent growth inhibition in TT cell lines while having no effect on normal thyroid Nthy ori-3-1 cells.

Full text of DOI:10.1021/acs.jmedchem.5b01522

Subscriber access provided by The Bodleian Libraries of The University of Oxford
Article
A Pyrazolo[3,4-d]pyrimidin-4-amine Derivative Containing an Isoxazole
Moiety is a Selective and Potent Inhibitor of RET Gatekeeper Mutants
Hojong Yoon, Yeonui Kwak, Seunghye Choi, Hanna Cho, Nam Doo Kim, and Taebo Sim
J. Med. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.jmedchem.5b01522 • Publication Date (Web): 10 Dec 2015
Downloaded from http://pubs.acs.org on December 16, 2015
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 free 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 accessible to all
readers and 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.
Journal of Medicinal Chemistry 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
A Pyrazolo[3,4ꢀd]pyrimidinꢀ4ꢀamine Derivative Containing an Isoxazole Moiety is a Selective
and Potent Inhibitor of RET Gatekeeper Mutants
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
†
,
‡,
‡
‡
§
Hojong Yoon ∥, Yeonui Kwak , Seunghye Choi , Hanna Cho , Nam Doo Kim ,
∥
, †, ‡
Taebo Sim*
†
Chemical Kinomics Research Center, Korea Institute of Science and Technology (KIST), 5,
‡
Hwarangꢀro 14ꢀgil, Seongbukꢀgu, Seoul 02792, Republic of Korea. KUꢀKIST Graduate School
of Converging Science and Technology, Korea University, 145, Anamꢀro, Seongbukꢀgu, Seoul,
§
1
36ꢀ713, Republic of Korea. DaeguꢀGyeongbuk Medical Innovation Foundation, 2387
dalgubeolꢀdaero, Suseongꢀgu, Daegu 706ꢀ010, Republic of Korea.
ABSTRACT
Aberrant RET kinase signaling plays critical roles in several human cancers such as thyroid
carcinoma. The gatekeeper mutants (V804L or V804M) of RET are resistant to currently
approved RET inhibitors such as cabozantinib and vandetanib. We, for the first time, report a
highly selective and extremely potent RET inhibitor, 6i rationally designed. Compound 6i
inhibits strongly RET gatekeeper mutants and other clinically relevant RET mutants as well as
wtꢀRET. This substance also significantly suppresses growth of thyroid cancerꢀderived TT cell
lines and Ba/F3 cells transformed with various RET mutants. Docking studies reveal that the
ACS Paragon Plus Environment
1

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
isoxazole moiety in 6i is responsible for binding affinity improvement by providing additional
site for Hꢀbonding with Lys758. Also, 6i not only substantially blocks cellular RET
autophosphorylation and its downstream pathway, it markedly induces apoptosis and anchorageꢀ
independent growth inhibition in TT cell lines while having no effect on normal thyroid Nthy
oriꢀ3ꢀ1 cells.
0
1
2
3
4
5
6
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
RET (REarranged during Transfection) is a singleꢀpass transmembrane receptor tyrosine kinase
1
that is mainly expressed in both the central and peripheral nervous system. This kinase contains
2
+
Ca ꢀdependent cell adhesion (cadherin) like domains and a cysteine rich region in its
extracellular domain. Moreover, the tyrosine kinase domain of RET activates numerous
2
intracellular signaling cascades in the cytosol. RET is activated by a complex of its coꢀreceptors
(GFRαs distributed within lipid rafts) and their distinct ligands (GDNF, neurturin, artemin,
persephin). The ligandꢀcoꢀreceptor complex recruits monomeric RET and promotes its
dimerization and autophosphorylation. After initial autophosphorylation at Y900 and Y905,
subsequent autophosphorylations at Y981, Y1015, and Y1062 take place to generate centers that
function as specific docking sites for numerous adaptor molecules. Y981 and Y1015 are binding
3
sites for Src and PLCγ, respectively, while Y1062 is a binding site for various adaptor proteins,
including Shc, ShcC, IRS1/2, FRS2, DOK 1/4/5, Enigma, ERK5 and MAPK. In addition, Y1062
engages in Ras/MAPK and PI3K/AKT signaling, which play roles in cell proliferation and
2
survival.
ACS Paragon Plus Environment
2

Page 3 of 64
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
The results of several clinical studies show that deregulation of RET signaling can lead to
thyroid cancers. For example, medullary thyroid carcinoma (MTC) arises either sporadically
(75%) or as a consequence of germline RET mutation (25%). The inherited form of MTC
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
includes multiple endocrine neoplasia type 2A (MEN2A), type 2B (MEN2B) and familial
thyroid carcinoma (FMTC), which are characterized by missense mutations that are different in
each genotype. RET point mutations in MEN2A take place predominantly at six cysteine
4
residues in the extracellular domain, including C609, C611, C618, C620, C630 and C634.
5
Predominant among these is the mutation at C634, which accounts for 85% of the cases.
Mutations of the cysteine residues lead to the formation of intramolecular disulfide bonds and
constitutive activation of RET kinase. MEN2B is principally characterized by the M918T
6
mutation (95%) with 5% of cases involving the A833F mutation, both of which bring about
7
conformational changes that lead to highly increased kinase activities. In addition to these
thyroid carcinomas, which are a consequence of oncogenic point mutations of RET, papillary
thyroid carcinoma (PTC) is associated with specific chromosomal rearrangements of RET
8
including for example CCDC6, PRKAR1A, NCOA and TRIM24. These chimeric products
(RET/PTC) lead to ligand independent dimerization of the kinase domain and constitutive
activation. Additionally, several other chimeric RET proteins, containing fusions such as KIF5Bꢀ
RET, CCDC6ꢀRET, NCOA4ꢀRET and TRIM33ꢀRET identified in lung adenocarcinoma, result
in highly increased RET transcription that leads to tumorigenesis.
To date, various small molecules that inhibit RET have been developed for the treatment of
9
RETꢀderegulated thyroid malignancy and KIF5BꢀRET fusionꢀpositive NSCLC. Quinazolineꢀ
based vandetanib and quinolineꢀbased cabozantinib are clinically approved for treatment of MTC
patients, and are in an ongoing phase II clinical trial for treatment of RETꢀrearranged NSCLC
ACS Paragon Plus Environment
3

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
10
(
NCT 01639508 for cabozantinib, NCT 01823068 for vandetanib). Other small molecule
kinase inhibitors are known to have a high binding affinity toward RET. These substances
include alectinib, ponatinib, sorafenib and several pyrazolopyrimidines such as PP1, PP2 and 1ꢀ
0
1
2
3
4
5
6
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
isopropylꢀ3ꢀ(phenylethynyl)ꢀ1Hꢀpyrazolo[3,4ꢀd]pyrimidinꢀ4ꢀamine (7a reported by Peter Diner
et al.) (Figure 1). Although PP1, PP2 and vandetanib effectively inhibit RET kinase activity in
HEK293 cells driven by various RET/MEN2 mutants, these inhibitors are seldom able to inhibit
1
2
gatekeeper mutants (V804M and V804L) of RET, which are frequently identified in familial
1
3
and sporadic MTC in certain regions. V804 of RET is a gatekeeper residue that corresponds to
specific codons, such as T315 in BcrꢀAbl, V561 in FGFR1, T338 in Src and T790 in EGFR, in
which point mutations are crucial for drug resistance. The BcrꢀAbl T315I is a mutation that leads
1
4
to secondary acquired resistance to imatinib in BcrꢀAbl positive CML patients, and EGFRꢀ
15
T790M causes resistance to gefitinib in NSCLC patients. The observations that the RETꢀ
V804G mutation leads to improved sensitivities to PP1 and PP2 while methionine mutations at
these positions result in resistance to these drugs suggest that substitution of codon 804 with
bulky residues causes steric hindrance with the inhibitors. In addition, the findings indicate that
V804M and V804L mutations cause primary or secondary resistance in vandetanibꢀtreated MTC
1
2
patients.
Recently, a few RET inhibitors that overcome inhibition blocking caused by the V804
mutation have been discovered. For example, the established ALK inhibitor, alectinib, was
demonstrated to be efficacious in RET fusion proteinꢀdriven NSCLC and to have the ability to
1
6
overcome KIF5BꢀRET V804L/V804M. Ponatinib, reported to be a BcrꢀAbl T315I inhibitor,
1
7
also displays a potent activity against RETꢀV804M. Finally, Pzꢀ1, which was reported to be a
18
dual panꢀRET/VEGFR2 inhibitor, also exhibits kinase inhibitory activity on RETꢀV804M. It is
ACS Paragon Plus Environment
4

Page 5 of 64
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
important to note that none of the RET inhibitors described above possess high kinase
selectivities and most, including vandetanib and cabozantinib, display several adverse effects
such as QTc prolongation, proteinuria and high blood pressure, especially a consequence of
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
19
strong VEGFR2 inhibition.
In the study described below, we devised a new strategy for the design of RET inhibitors.
Based on the known crystal structure of the RET kinase domain (KD) complexed with PP1 and
an analysis of the results of studies in which the previously described substance 7a is docked
1
1
with RET KD, we devised a novel strategy for the design of RET KD inhibitors. Specifically,
we envisioned that inhibitor activity would be enhanced by replacement of the alkynyl group of
7
a with a small heterocyclic moiety, which can participate in hydrogen bonding interactions with
RET KD. We selected the isoxazole ring for this purpose based on the hypothesis that this
moiety would form a hydrogen bond with the εꢀamino group of Lys758 that is positioned in the
ATP binding site. In addition, we believed that like the alkyne moiety in 7a, the isoxazole group
would function as a rigid linker. Utilizing this strategy, we identified 6i as a lead compound. The
results of initial studies showed that this substance has strong kinase inhibitory activities against
RETꢀV804M (IC = 4 nM) and V804L (IC = 21 nM) as well as on RETꢀwild type (IC = 1.4
50
50
50
nM) and that it significantly suppresses growth of Ba/F3 cells transformed with RETꢀV804M
(
GI = 320 nM) and RETꢀV804L mutants (GI = 500 nM). The results of this effort are
5
0
50
described below.
ACS Paragon Plus Environment
5

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
0
1
2
3
4
5
6
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 1. (A) Representative RET inhibitors. (B) Rational design of novel isoxazoleꢀcontaining
RET inhibitors.
ACS Paragon Plus Environment
6

Page 7 of 64
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
RESULT AND DISCUSSION
Chemistry
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
The routes used to synthesize the 3ꢀ(isoxazolꢀ3ꢀyl)ꢀ1Hꢀpyrazolo[3,4ꢀd]pyrimidinꢀ4ꢀamines
explored in this investigation are presented in Schemes 1 and 2. In one pathway, the pyrazole
ring in 3ꢀbromoꢀ1Hꢀpyrazolo[3,4ꢀd]pyrimidinꢀ4ꢀamine (1) was Nꢀalkylated using 2ꢀ
bromopropane in the presence of potassium carbonate in DMF. The bromo group in 2, the
product produced in this process, was replaced by a vinyl group to form 3 using a Stille coupling
reaction with tributyl(vinyl)tin and Pd(PPh ) . Dihydroxylation of 3 with OsO and NMO,
3
4
4
followed by oxidative cleavage using NaIO produced the corresponding aldehyde 4. Subsequent
4
independent condensations of the aldehyde group in 4 with hydroxylamine and Oꢀ
benzylhydroxylamine yielded the key oximes 5a and 5b. Oxime 5a was used to optimize the
nitrile oxide cycloaddition (NOC) reaction promoted by either NBS/TEA, NaOCl or tꢀBuOCl.
Reaction of 5a using NBS failed to bring about bromination and treatment of this substance with
either NaOCl or tꢀBuOCl results in formation of only trace amounts of desired isoxazole product.
Jawalekar et al. reported earlier that phenyliodineꢀbis(trifluoroacetate) (PIFA) promotes NOC
reactions that generate isoxazoles or isoxazolines with excellent yields and levels of
2
0
regioselectivity. By utilizing this methodology, the desired NOC reactions of 5a occur in
moderate yields to form the corresponding isoxazole derivatives 6a-l. Treatment of 6l with
cesium fluoride then affords the isoxazole ring unsubstituted analog 7.
Another route developed for preparation of 3ꢀ(isoxazolꢀ3ꢀyl)ꢀ1Hꢀpyrazolo[3,4ꢀd]pyrimidinꢀ4ꢀ
amines began with cyclopropyl ketone 8 (Scheme 2). Reaction of 8 with diethyl oxalate in the
presence of sodium hydride produced diketoꢀester 9, which reacted with hydroxylamine to form
ACS Paragon Plus Environment
7

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
isoxazole 10. Subsequent hydrolysis of the ester moiety in 10 formed the carboxylic acid 11,
which was transformed to the corresponding acid chloride by treatment with oxalyl chloride.
Subsequent addition of malononitrile followed by methylation produced 14 in a 63% yield over 3
steps. Reaction of 14 with hydrazine yielded pyrazole 15, which upon treatment with formamide
0
1
2
3
4
5
6
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
o
at 160 C generated pyrazolo[3,4ꢀd]pyrimidinꢀ4ꢀamine 16. Preparation of the Nꢀalkyl derivatives
17a-f was accomplished by using Mitsunobu reactions with selected alcohols or by alkylation
with alkyl halides in basic media.
a
Scheme 1. Synthesis of inhibitors 5a-b, 6a-l and 7
a
o
Reagent and condition: (a) 2ꢀbromopropane, K CO , DMF, 80 C, 69%; (b) tributyl(vinyl)tin,
2
3
o
Pd(PPh ) , toluene, reflux, 82%; (c) i) OsO , NMO, THF/H O, 0 C to rt; ii) NaIO , THF/H O, 0
3
4
4
2
4
2
o
C to rt; (d) hydroxylamine hydrochloride or benzyloxyamine hydrochloride, NaOAc, EtOH, rt,
72ꢀ89% over 2 steps; (e) styrene or various acetylenes, phenyliodine bis(trifluoroacetate),
MeOH/H O, rt, 17ꢀ32%; (f) 6l, CsF, MeCN/EtOH, rt, 43%.
2
ACS Paragon Plus Environment
8

Page 9 of 64
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
a
Scheme 2. Synthesis of inhibitors 16 and 17a-f
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
a
o
Reagent and condition: (a) diethyl oxalate, NaH, THF, 0 C to rt, 92%; (b) hydroxylamine
hydrochloride, MeOH, reflux, 88%; (c) NaOH, EtOH, rt, 100%; (d) oxalyl chloride, cat.DMF,
o
CH Cl , rt; (e) malononitrile, DIPEA, THF, 0 C; (f) dimethylsulfate, THF, reflux, 63% over 3
2
2
o
steps; (g) hydrazine dihydrochloride, TEA, EtOH, rt, 72%; (h) formamide, 160 C, 82%; (g) a)
o
R ꢀX (X = Br, I) or 2,2ꢀdimethoxyoxirane, Cs CO , DMF, 70 C or b) R ꢀOH, DIAD, PPh , THF,
2
2
3
2
3
rt, 32ꢀ81%
Structure-Activity Relationships
In vitro biochemical kinase assays with recombinant RET kinase were used to assess the
kinaseꢀinhibitory activities of the isoxazole derivatives prepared in the manner described above.
The results show that 6b has an IC value of 0.190 ꢁM against RETꢀwild type (WT) but that it
5
0
displays no inhibitory activity against the RETꢀV804M gatekeeper mutant (Table 1). In addition,
the oxime derivatives 5a and 5b have low inhibitory activities, which is consistent with the
ACS Paragon Plus Environment
9

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
previous report that rigidity of linker is essential for this activity. Moreover, replacement of
isoxazole ring in 6b with an isoxazoline ring (6a) and the phenyl group in 6b with a benzyl (6c)
or 4ꢀfluorophenyl (6d) group have no distinctive effects on inhibitory activity. Introduction of
pyridine ring (6e) leads to a 5ꢀfold decrease in activity (IC = 0.910 ꢁM) compared with that of
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
5
0
6b, which could be a consequence of the existence of a highly hydrophobic back pocket in the
1
1
kinase that accommodates the R group.
1
1
In contrast to the observation that replacements by several aromatic groups at R do not alter
activity, incorporation of small alkyl substituents at the R position was found to lead to
1
enhanced inhibitory activities. For example, introduction of small alkyl groups such as ethyl (6f)
and isopropyl (6g) caused remarkable increases in activity. In contrast, introduction of the bulky
tꢀbutyl group (6h) at this position resulted in poor activity (IC = 0.210 ꢁM). Compared to the
5
0
activity difference between 6g and 6h, the cyclopropyl group containing substance 6i has
excellent kinaseꢀinhibitory activities against both RETꢀWT (IC50 = 0.0014 ꢁM) and V804M
gatekeeper mutant (IC = 0.004 ꢁM). Encouraged by these observations, we explored kinase
5
0
inhibition by cyclopentyl (6j, IC50 = 0.241 ꢁM) and cyclohexyl (6k, IC50 = 0.910 ꢁM)
derivatives related to 6i. The results show that these substances have activities, which decrease as
1
the size of ring increases. Finally, studies with 7, in which the R group is hydrogen, indicate that
it has mediocre activity (IC50 = 0.213 ꢁM). This observation suggests that hydrophobic
1
interaction between the R group and a hydrophobic site in the kinase is essential for RET
inhibitory activity. In summary, the combined results of this effort show that the cyclopropyl
group containing isoxazole derivative, 6i has optimal kinase inhibitory activity.
Our attention next turned to an evaluation of the effects on kinase inhibitory activity of
2
changing the Nꢀisopropyl group in 6i to other Nꢀalkyl groups (R ). The results show that
ACS Paragon Plus Environment
10

Page 11 of 64
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
elimination (16, IC = 0.052 ꢁM) or substitution of the Nꢀisopropyl group with the relatively
5
0
small methyl group (17a, IC50 = 0.027 ꢁM) leads to a lower activity than that of 6i. Also,
introduction of bulkier groups such as benzyl (17b), heptyl (17c) and 2,2ꢀdimethyl propanol (17d)
causes a decrease in activity. Even though substances containing cyclic substituents, like 17e and
17f, display higher activities (IC = 0.002 ꢁM and 0.005 ꢁM, respectively) than their bulky
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
5
0
acyclic analogs and exhibit comparable potency to 6i on RETꢀWT, they are slightly less potent
on RETꢀV804M gatekeeper mutant than 6i which contains an isopropyl group at the R position.
2
Encouraged by the excellent kinase inhibitory activities observed using the biochemical kinase
assay, we next assessed the capability of the 3ꢀ(isoxazolꢀ3ꢀyl)ꢀ1Hꢀpyrazolo[3,4ꢀd]pyrimidinꢀ4ꢀ
amines to inhibit RET in the two humanꢀdriven thyroid cell lines, TT and Nthy oriꢀ3ꢀ1. TT cells
are RETꢀdependent human thyroid cancer cells harboring a MEN2Aꢀtype mutation in the form
21
of a C634W point mutation. Nthy oriꢀ3ꢀ1 cells are a normal thyroid cell line, which are used as
a control to assess the differential cytotoxicity of the RET inhibitors. As the data displayed in
Table 2 shows, GI50 values of the 3ꢀ(isoxazolꢀ3ꢀyl)ꢀ1Hꢀpyrazolo[3,4ꢀd]pyrimidinꢀ4ꢀamines
against TT thyroid cancer cells range between 0.24 ꢁM to over 10 ꢁM. In addition, the trend
observed in growthꢀinhibitory activities of the inhibitors on these cell lines wellꢀmatches with the
trends seen in the enzymatic assay data. For example, 6i, which showed the highest potency in
the enzyme assay is also highly potent against TT cells (GI = 0.24 ꢁM). However, the cellular
5
0
activity of 6i is slightly less than that of cabozantinib (GI = 0. 12 ꢁM). It should be noted that
5
0
cabozantinib blocks the growth of normal thyroid cells, Nthy oriꢀ3ꢀ1, to a greater extent (GI =
5
0
4
.9 ꢁM) than does 6i (GI = 13.7 ꢁM), which suggests that the strong growth inhibitory effect of
50
cabozantinib might be explained in part by offꢀtarget effects of cabozantinib, a multiꢀtargeted
kinase inhibitor. The results show that the optimized oxazole derivative 6i possesses remarkable
ACS Paragon Plus Environment
11

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
potency against TT thyroid cancer cells and an excellent differential cytotoxicity (over 50ꢀfold,
TT GI = 0.24 ꢁM, Nthy oriꢀ3ꢀ1 GI = 13.7 ꢁM). Meanwhile, among other tested substances,
50
50
1
7e exhibited the greatest selectivity for TT thyroid cancer cells (GI = 0.29 ꢁM) over normal
50
0
1
2
3
4
5
6
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
thyroid cells (GI = 32.5 ꢁM).
5
0
In order to estimate kinase activityꢀdependent cellular activities of the new RET inhibitor in an
accurate and efficient fashion, several engineeredꢀBa/F3 cell lines transformed with various RET
mutations along with wildꢀtype RET were constructed. The antiꢀproliferative activities of the
inhibitors on Ba/F3 cell lines transformed with wildꢀtype and gatekeeper mutants (V804M) of
RET were determined first. As expected, the trend in growth inhibitory activities of the inhibitors
against these Ba/F3 cell lines closely matches the kinaseꢀinhibitory trend observed in the
biochemical kinase assay data. Compared with activities against RETꢀWT Ba/F3 cells, all the
inhibitors except 6e and 6k are less potent against RETꢀV804M Ba/F3 cells. This observation is
consistent with the results emanating from the biochemical kinase assays. Using the Ba/F3 cell
based assay, we showed that the optimized substance 6i has excellent potencies against RETꢀWT
Ba/F3 (GI = 0.08 ꢁM) and RETꢀV804M Ba/F3 (GI = 0.32 ꢁM), which demonstrates that this
5
0
50
substance has a remarkable inhibitory effect on RETꢀWT and RETꢀV804M at the cellular level.
The results of studies designed to assess the in vitro microsomal stability and toxicity properties
of 6i (Table S1, S2 and Figure S1 in Supporting Information) show that this substance has a
moderate level of metabolic stability in human liver microsome (T1/2 = 13 min). Also, the four
cytochrome P450 enzymes CYP2C9, CYP2C19, CYP2D6, CYP3A4 are not inhibited by 6i (IC50
>
30 ꢁM) while CYP1A2 is influenced to some extent (IC = 2 ꢁM). Moreover, no inhibition
50
occurs even at 100 ꢁM in the herg patch clamp assay, indicating that 6i is free from cardiac
toxicity caused by QTꢀinterval prolongation.
ACS Paragon Plus Environment
12

Page 13 of 64
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Table 1. Results of biochemical assays of the inhibitory activities of the 3ꢀ(isoxazolꢀ3ꢀyl)ꢀ1Hꢀ
pyrazolo[3,4ꢀd]pyrimidinꢀ4ꢀamines against RETꢀWT and RETꢀV804M.
R1
R1
R1
O
N
O
N
O
N
NH2
NH2
NH2
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
N
N
N
or
or
N
N
N
N
N
N
N
N
N
R2
R2
5
a-b
6a
6b-l, 7, 16, 17a-f
a,b
IC (ꢁM)
5
0
Compound
R1
R2
RETꢀWT RETꢀV804M
5a
H
ꢀ
ꢀ
4.90
8.33
3.68
inactive
5.68
5b
benzyl
phenyl
phenyl
benzyl
6a
isopropyl
isopropyl
isopropyl
isopropyl
isopropyl
isopropyl
isopropyl
isopropyl
isopropyl
isopropyl
isopropyl
isopropyl
H
0.583
0.190
0.241
0.150
0.91
6b
inactive
1.99
6c
6d
4ꢀfluorophenyl
2ꢀpyridinyl
ethyl
inactive
5.88
6e
6f
0.041
0.041
0.210
0.0014
0.241
0.910
0.213
0.052
0.027
0.074
0.096
0.068
0.002
0.005
0.531
0.355
2.10
6g
isopropyl
6h
tꢀbutyl
6
i
cyclopropyl
cyclopentyl
cyclohexyl
H
0.004
1.57
6j
6k
inactive
0.903
0.354
0.182
0.557
0.927
0.217
0.014
0.041
7
16
cyclopropyl
cyclopropyl
cyclopropyl
cyclopropyl
cyclopropyl
cyclopropyl
cyclopropyl
1
7a
7b
7c
7d
7e
17f
methyl
1
benzyl
1
nꢀheptyl
1
2ꢀmethylꢀpropanꢀ2ꢀolꢀ1ꢀyl
cyclohexyl
1
tetrahydropyranꢀ4ꢀyl
a
b
Radiometric biochemical kinase assay. ‘inactive’ means that the kinase activity was inhibited
by less than 50% even at 10 ꢁM concentration of the compound.
ACS Paragon Plus Environment
13

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
Table 2. Antiꢀproliferation activities of 3ꢀ(isoxazolꢀ3ꢀyl)ꢀ1Hꢀpyrazolo[3,4ꢀd]pyrimidinꢀ4ꢀamines
against various cell lines.
a
GI (ꢁM)
5
0
Compound
Ba/F3 cell lines
RETꢀWT
0.05 ± 0.02
0
1
2
3
4
5
6
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
TT
Nthy oriꢀ3ꢀ1
Parental
>10
RETꢀV804M
0.89 ± 0.03
2.60 ± 0.01
>10
Cabozantinib 0.12 ± 0.02
4.93 ± 0.49
3.58 ± 0.88
107 ± 14.1
9.85 ± 1.59
18.9 ± 1.21
5.72 ± 2.06
21.5 ± 3.69
9.77 ± 5.82
7
a
a
2.54 ± 0.89
>10
>10
0.43 ± 0.06
>10
5
>10
5
b
9.27 ± 1.97
>10
>10
7.41 ± 1.31
6.24 ± 0.34
3.46 ± 0.35
>10
>10
6
a
>10
>10
6
b
>10
>10
4.97 ± 1.52
>10
6c
>10
>10
6
d
>10
>10
>10
>10
6e
9.26 ± 1.26
2.61 ± 0.09
5.13 ± 0.07
>10
1.63 ± 0.12 5.30 ± 1.64
1.11 ± 0.35
1.56 ± 0.52
2.40 ± 0.61
6.12 ± 0.62
0.08 ± 0.00
7.12 ± 1.67
9.50 ± 1.49
3.21 ± 0.05
5.34 ± 0.45
2.67 ± 0.69
>10
0.70 ± 0.16
5.77 ± 1.66
5.90 ± 0.17
7.31 ± 0.39
0.32 ± 0.06
>10
6
f
28.6 ± 3.22
23.7 ± 1.81
24.7 ± 1.27
13.7 ± 0.42
11.5 ± 2.13
4.11 ± 0.53
18.6 ± 1.95
0.51 ± 0.04
28.7 ± 6.85
39.6 ± 1.95
31.6 ± 4.01
2.82 ± 0.50
32.5 ± 7.16
32.9 ± 6.30
>10
>10
>10
>10
>10
>10
>10
>10
>10
>10
>10
>10
>10
>10
6g
6
h
6
i
0.24 ± 0.05
>10
6
j
6
k
>10
4.73 ± 1.14
8.94 ± 2.46
>10
7
7.92 ± 1.15
>10
1
6
17a
7b
17c
7d
17e
7f
3.51 ± 0.04
>10
5.56 ± 0.65
>10
1
>10
>10
>10
1
4.25 ± 0.73
0.29 ± 0.06
0.94 ± 0.19
1.77 ± 0.19
0.52 ± 0.11
0.44 ± 0.30
6.75 ± 1.17
2.00 ± 0.18
1.93 ± 0.32
1
a
GI represents the concentration at which a compound causes halfꢀmaximal growth inhibition.
5
0
GI values for Nthy oriꢀ3ꢀ1 and transformed Ba/F3 cell lines were determined with SD values
5
0
arising from three independent data sets, while those for TT cells and parental Ba/F3 cells were
determined in a duplicate mode.
ACS Paragon Plus Environment
14

Page 15 of 64
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Inhibitory activities of 6i against clinically relevant RET mutants.
The next phase of this investigation was designed to gain more detailed information about the
inhibitory activities of 6i and two reference compounds, cabozantinib and 7a, against other
oncogenic RET mutants as well as the V804M gatekeeper mutant. For this purpose, the activities
of these substances against clinically relevant RET mutants (G691S, Y791F, V804L, V804M,
S891A, M918T) were determined using an in vitro biochemical kinase assays. As shown in
Table 3, 6i has excellent kinase inhibitory activities against all of the clinically relevant RET
mutants, with IC values ranging from 1.4 to 21 nM. Moreover, 6i is much more potent than 7a
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
5
0
against all of the RET mutants tested. In particular, 6i has a several hundredꢀfold higher (over
00ꢀfold for V804M and over 100ꢀfold for V804L) potency against RET gatekeeper mutants
6
relative to that of 7a. Furthermore, 6i also displays a 10 to 90ꢀfold greater kinaseꢀinhibitory
activity against V804L/M gatekeeper mutants as compared to cabozantinib. Both 7a and
cabozantinib are much less active against RET gatekeeper mutants compared to RETꢀWT while
6i is highly potent against both RET gatekeeper mutants and RETꢀWT.
ACS Paragon Plus Environment
15

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
Table 3. Inhibitory activities of 6i, cabozantinib, 7a against clinically relevant RET mutants.
6i
Cabozantinib
7a
a
Enzyme assay (IC , nM)
5
0
0
1
2
3
4
5
6
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
RETꢀWT
1.4
10
3
9
14
7
28
75
RETꢀG691S
RETꢀY791F
RETꢀV804L
RETꢀV804M
RETꢀS891A
RETꢀM918T
14
21
4
230
358
5
3,130
2,650
54
2
6
18
57
b
Ba/F3 Cell assay (GI , µM)
5
0
Parental
>50
>50
16.1 ± 5.98
0.61 ± 0.06
5.43 ± 0.35
2.60 ± 0.01
0.35 ± 0.03
RETꢀWT
0.08 ± 0.00
0.50 ± 0.06
0.32 ± 0.06
0.09 ± 0.26
0.05 ± 0.02
0.91 ± 0.03
0.89 ± 0.03
0.01 ± 0.00
RETꢀV804L
RETꢀV804M
RETꢀS891A
a
b
Radiometric kinase assay. All GI values, except those for parental and RETꢀV804M Ba/F3
50
cells, were determined in a duplicate mode.
The growth inhibitory activities of 6i, cabozantinib and 7a on parental Ba/F3 and three different
Ba/F3 cell lines transformed with RET mutations, including gatekeeper mutations, were also
determined (Table 3). None of the three compounds was observed to suppress significantly the
proliferation of parental Ba/F3 cells. Inhibitor 6i exhibits profound antiꢀproliferative activities on
RETꢀWT Ba/F3 and three Ba/F3 cells harboring three RET pointꢀmutations (V804L, V804M, or
S891A) and more significantly it blocks proliferation of Ba/F3 cells harboring gatekeeper
mutations to a greater extent than does cabozantinib. Cabozantinib and 6i have doubleꢀdigit
nanomolar potencies against Ba/F3 cells transformed with RETꢀWT and RETꢀS891A. Finally,
the antiꢀproliferative activities of 7a against Ba/F3 cells transformed with gatekeeper mutations
ACS Paragon Plus Environment
16

Page 17 of 64
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
(
V804L, V804M) is low, a finding that is consistent with its low inhibition activities against
gatekeeper mutants in biochemical kinase assays.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Molecular Docking Study of 6i
To gain a better understanding of the source of the activity differences between 6i, cabozantinib
and 7a against RETꢀWT and its gatekeeper mutants, docking studies were performed using the
Schrödinger docking program, Glide. Analysis of the binding mode predicted using this docking
method shows that 6i forms hydrogen bonds with both Glu805 and Ala807 in the hinge region.
In addition, the cyclopropyl group in 6i occupies a hydrophobic back pocket in the ATP binding
site and the oxygen atom of isoxazole moiety acts as hydrogen bond acceptor for interaction with
the ammonium group of Lys758 (Figure 2A, 2B). The additional hydrogen bonding interaction
caused by the presence of the isoxazole group and Hꢀbonding contacts to the kinase hinge appear
to be the major contributing factors making 6i more potent RET inhibitor than 7a. In addition,
the hydrogen bonding interactions with the isoxazole moiety might also be the reason why 6i has
a high affinity for RET gatekeeper mutants. Furthermore, the relatively small cyclopropyl group
in 6i seems to be sterically less affected by the bulky groups present in the gatekeeper pointꢀ
mutants, M804 and L804.
We also carried out docking studies on complexes between the two reference compounds, 7a
and cabozantinib, and the RETꢀV804M gatekeeper mutant. The goal was to understand why
these substances are much less potent inhibitors of the gatekeeper mutant as compared to RETꢀ
WT. Based analysis of the docking model of the complex, the phenyl group of 7a is located in
the hydrophobic pocket of the kinase where it is involved in a steric interaction with the bulky
group that is present in M804 but not in V804. This interaction might be the cause of the
ACS Paragon Plus Environment
17

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
substantially lower activity of 7a against the V804M mutant (Figure 2C, 2D). The results of the
docking studies show that cabozantinib binds to the DFGꢀout inactive conformation, which is
characteristic of a type II kinase inhibitor and decreased activity of cabozantinib against V804M
might be a consequence of a steric interaction between the bridge phenyl ring and a bulky group
in the V804M mutant (Figure 2E, 2F).
0
1
2
3
4
5
6
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 2. Docking models of (A, B) 6i, (C, D) compound 7a and (E, F) cabozantinib bound to
RET kinase (PDB code 2IVV for 6i and 7a, 3LQ8 for cabozantinib). Hydrogen bonds are
depicted as dashed lines.
ACS Paragon Plus Environment
18

Page 19 of 64
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Kinase Selectivity Profile of 6i
The inhibition selectivity of 6i was assessed using a panel of 332 kinases [Reaction biology,
USA] and an in vitro biochemical autophosphorylation assay at a single concentration of 1 ꢁM.
The results show that 6i displays an excellent kinase selectivity with a 0.02 selectivity score S10
value, defined as the percentage of the kinases showing >90% inhibition out of the total (332)
number of kinases profiled (Figure 3A). The kinases, which are inhibited by 6i (1 ꢁM) to an
extent of >80%, are RET, p70S6K, PKA, FLT4, FGR, MEK2, VEGFR2, MEK1, HCK, RSK1.
IC values of 6i against these 10 kinases were determined in order to assess its selectivity for
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
5
0
RET (Figure 3B). The IC values for the 10 kinases are in the range of 23ꢀ311 nM, showing that
5
0
6i possesses an approximately 16ꢀ200 fold selectivity for RET (IC = 1.4 nM).
50
It has recently been reported that cabozantinib and vandetanib cause VEGFR2ꢀrelated adverse
22
effects such as high blood pressure and proteinuria in patients with thyroid cancer. As a result,
a significant effort is being given to the discovery of inhibitors that are more selective for RET
over other kinases including VEGFR2. From this perspective, 6i is advantageous because it is a
much less potent inhibitor of VEGFR2 (IC = 96 nM) relative to RET (IC = 1.4 nM). We
5
0
50
further demonstrated that this selectivity exists at the cellular level using VEGFR2 Ba/F3 cells
Figure S2). In a manner that is different from the high potency of cabozantinib (GI = 3 nM), 6i
(
5
0
exhibits only a moderate antiꢀproliferative activity on VEGFR2ꢀBa/F3 cells (GI = 3 ꢁM). This
5
0
result suggests that 6i might have an important pharmacological advantage in that it may not
trigger VEGFR2ꢀdriven adverse effects.
ACS Paragon Plus Environment
19

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
A.
B.
Enzyme
Score (%) IC (nM)
5
0
0
1
2
3
4
5
6
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
RET
p70S6K
PKA
1.62
5.94
7.34
7.61
7.96
10.8
11.5
11.5
14.1
14.9
1.4
23
39
72
82
65
96
FLT4
MEK2
FGR
VEGFR2
MEK1
HCK
142
137
311
RSK1
Figure 3. (A) KinomeScanTreeSpot description of the 6i selectivity profile. 6i was screened
against a 332 kinase panel at a concentration of 1 ꢁM. Scores for primary screen hits are reported
as a percent of the DMSO control (% control). (B) Biochemical IC s of 6i measured with
5
0
selected kinases.
Inhibition of RET signaling in Cancer Cell Lines by 6i
The capability of 6i to block RET signaling in a cellular context was evaluated. First, the effects
of 6i, 7a and cabozantinib on RET phosphorylation and its downstream signaling on Ba/F3 cell
lines transformed with several RET variants (WT, V804M, V804L, S891A) were determined. In
its phosphorylated state, Y905 located in the Aꢀloop enables the RET kinase domain to exist in
the active conformation, and Y1062 plays a role as a docking site for multiple adaptor
2
3
molecules. In agreement with in vitro assay data, 6i and cabozantinib effectively inhibit the
level of phosphorylation of RET and its adaptor molecules such as PLCγ and Shc in RETꢀWT
ACS Paragon Plus Environment
20

Page 21 of 64
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Ba/F3 and RETꢀS891A Ba/F3 cells. In contrast, 7a is a less effective inhibitor of
phosphorylation (Figures 4A and 4B). More importantly, 6i and cabozantinib suppress RET
phosphorylation in RETꢀV804L/V804M Ba/F3 cells in a doseꢀdependent manner (Figure 4C and
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
4
D). Notably, 6i more effectively blocks RET phosphorylation in RETꢀV804M Ba/F3 cells than
does cabozantinib, an observation that is consistent with the results of in vitro biochemical
kinase assays. In contrast to 6i and cabozantinib, 7a is unable to block RET signaling in RETꢀ
V804L/V804M Ba/F3 cells even at a high concentration of 10 ꢁM.
In order to investigate further their abilities to abrogate autophosphorylatoin of RET kinase and
its downstream signaling, the effects of 6i, cabozantinib and 7a on TT cells harboring RETꢀ
C634W mutation were evaluated (Figure 4E). Autophosphorylation of RET at Y905 and Y1062
is effectively attenuated by 1 ꢁM 6i and cabozantinib, while 7a does not exhibit significant
inhibition at this concentration. Phosphorylation of RET downstream molecules, such as PLCγ,
AKT and ERKs, is also much more effectively blocked by 6i and cabozantinib than by 7a. In
addition, 6i much more strongly suppresses cellular phosphorylation of S6 than does
cabozantinib. This observation is consistent with the finding that 6i inhibits p70S6K, a direct
upstream regulator of S6, with an IC value of 23 nM in the biochemical kinase assay (Figure
5
0
3
B).
The cellular activity of 6i against GDNFꢀinduced phosphorylation of ERK1/2 in MCFꢀ7 breast
2
4
cancer cells was explored (Figure S3). GDNF (glial cell derived neurotrophic Factor) is known
to induce RETꢀdependent phosphorylation of ERK1/2. In accordance with the previous report,
11
addition of GDNF results in an increase in ERK1/2 phosphorylation. Importantly, 6i effectively
inhibits GDNFꢀinduced phosphorylation of ERK1/2 even at a 100 nM concentration.
ACS Paragon Plus Environment
21

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
0
1
2
3
4
5
6
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. Autophosphorylation of RET and its downstream signaling are blocked by 6i. The
effect of 6i on autophosphorylation of RET in (A) RETꢀWT Ba/F3, (B) RETꢀS891A Ba/F3, (C)
RETꢀV804L Ba/F3 and (D) RETꢀV804M Ba/F3. (AꢀD) Ba/F3 cells, stably transformed with the
indicated constructs, were treated for 1 h with gradually increasing concentrations of 7a,
ACS Paragon Plus Environment
22

Page 23 of 64
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
cabozantinib and 6i (0ꢀ10 ꢁM) and then lysed. Protein extracts were immunoblotted with
antibodies specific for the Y905 and Y1062 phosphorylated forms of RET. Actin was used as
loading control. (E) TT cells were treated with 6i (0.1, 1 and 10 ꢁM), cabozantinib and 7a (1 and
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
1
0 ꢁM) for 4 hrs. Cell lysates were immunoblotted with the indicated antibodies.
Effects of 6i on apoptosis and cell cycle arrest in TT and Nthy ori-3-1 cell lines
The effects of 6i, cabozantinib and 7a on apoptosis induction and cell cycle arrest in both TT
and Nthy oriꢀ3ꢀ1 cell lines were determined. The results show that 6i and cabozantinib do not
induce cell cycle arrest but they do promote increased apoptosis in TT cells (Figures 5A and 5C).
Apoptosis induction caused by 6i was also confirmed by immunoblotting apoptosis markers,
which showed an increase in cleaved PARP1/2 and the increased amount of PUMA, a proꢀ
apoptotic marker (Figure 5E). It is worthwhile to note that cabozantinib at a concentration of 10
ꢁ
M also causes cell cycle arrest in the G2 phase and apoptosis in the normal thyroid cell line,
Nthy oriꢀ3ꢀ1, (Figures 5B and 5D), while 6i at 10 ꢁM does not. This observation indicates that 6i
has appreciable selectivity for thyroid cancer cells over normal cells while cabozantinib does not.
Meanwhile, 7a does not cause significant cell cycle arrest and apoptosis in either TT or Nthy oriꢀ
3ꢀ1 cells.
ACS Paragon Plus Environment
23

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
0
1
2
3
4
5
6
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. 6i induces apoptosis in TT cells but not in Nthy oriꢀ3ꢀ1. Effects of 7a, 6i and
cabozantinib on cell cycle arrest of (A) TT cells and (B) Nthy oriꢀ3ꢀ1 cells. TT cells and Nthy
oriꢀ3ꢀ1 cells were incubated with the indicated compounds for 48 hrs and 24 hrs, respectively.
Cells, fixed with 70% ethanol, were stained with propidium iodide and analyzed by using flow
cytometry. (A, B) The left panel contains a representative image obtained from the flow
cytometric analysis of cell cycle arrest. Bar graphs in the right panel contain means ± SDs for
three independent experiments on Nthy oriꢀ3ꢀ1 cells and for two independent sets on TT cells.
Effects of 7a, 6i and cabozantinib on apoptosis of (C) TT cells and (D) Nthy oriꢀ3ꢀ1 cells. The
ACS Paragon Plus Environment
24

Page 25 of 64
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
indicated concentrations of the compounds were incubated with TT cells for 96 hrs, and Nthy
oriꢀ3ꢀ1 cells for 48 hrs. Cells were harvested and incubated in a binding buffer containing
propidium iodide and annexin V and then analyzed by using flow cytometry. (C, D) the left
panel contains a representative image of the flow cytometric analysis of apoptosis. In the right
panel are bar graphs that indicate the mean ± SDs for three independent experiments. **p < 0.01,
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
***p < 0.001, One way ANOVA tested. (E) Treatment with 6i for 96 hrs induces cleavage of
PARP1/2, while increasing the expression of PUMA in TT cells. Actin was used as the loading
control.
Inhibition of anchorage-independent growth of TT cells by 6i
In the final phase of this effort, we investigated the antiꢀtumor effect of 6i by utilizing the soft
agar assay (Figure 6). Incubation with 7a at two different concentrations (0.3 ꢁΜ, and 1 ꢁΜ) for
30 days does not suppress anchorageꢀindependent growth with a mean value of 150 colonies. On
the contrary, incubation with either 6i or cabozantinib causes TT cells to form in significantly
lower numbers (# < 3) and in smaller colonies in soft agar (p < 0.001). Treatment with 0.3 ꢁΜ 6i
also leads to prevention of RETꢀdriven oncogenic transformation of TT cells. The effect caused
by 0.3 ꢁΜ 6i is comparable to that promoted by 1 ꢁΜ cabozantinib, indicating that 6i is more
effective in inhibiting anchorageꢀindependent growth of TT cells.
ACS Paragon Plus Environment
25

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
0
1
2
3
4
5
6
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. Anchorage independent growth of TT cells is suppressed by treatment with 6i. (AꢀG)
Effects of 7a, 6i and cabozantinib on anchorageꢀindependent growth of TT cells in soft agar. (Aꢀ
F) A representative colony under each indicated condition was photographed after crystal violet
staining at 400x magnification. Bar = 0.2 mm. (G) Soft agar assays was performed three times.
Total colonies in each well were automatically counted using Image J software. The average
number of colonies (n = 3) are presented with ± SD. ***p < 0.001, One way ANOVA tested.
CONCLUSION
In the investigation described above, we rationally designed and synthesized a series of novel 3ꢀ
(isoxazolꢀ3ꢀyl)ꢀ1Hꢀpyrazolo[3,4ꢀd]pyrimidinꢀ4ꢀamines containing an isoxazole moiety, and
carried out a biologically evaluation of their RET kinase inhibitor activities and selectivities. The
strategy employed to design these substances arose from an analysis of important interactions
that exist in the crystal structure of RET kinase domain complexed with PP1 and a model
obtained by docking of 7a (reported by Peter Diner et al.) in the RET kinase domain. The results
ACS Paragon Plus Environment
26

Page 27 of 64
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
of an SAR study, led to the identification of the 3ꢀ(isoxazolꢀ3ꢀyl)ꢀ1Hꢀpyrazolo[3,4ꢀd]pyrimidinꢀ
4ꢀamine 6i as an optimal RET inhibitor. The results of biochemical kinase assays show that this
substance has singleꢀdigit nanomolar IC values against V804M gatekeeper mutant (IC = 4
5
0
50
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
nM) and RETꢀWT (IC50 = 1.4 nM). In addition, 6i has a high potency against Ba/F3 cells
transformed with RETꢀV804M. The results of molecular docking studies suggest that hydrogen
bonding interaction between the isoxazole ring in 6i and the Lys758 ammonium group in the
kinase is important in causing the high inhibitory activities of 6i. Kinomeꢀwide selectivity
profiling, using 332 kinases, revealed that 6i displays an excellent selectivity (five ones of 332
kinases are inhibited over 90% by 1 ꢁΜ of 6i) for RET over other kinases including VEGFR2,
whose inhibition is thought to be the main cause of the adverse effects of vandetanib and
cabozantinib. Considering the fact that no selective RET inhibitor that is capable of blocking
RET gatekeeper mutant has been uncovered to date, observations made in studies with 6i are
significant. Also, we observed that 6i substantially suppresses proliferation of thyroid cancerꢀ
derived TT cells harboring the C634WꢀRET mutant and significantly blocks RET
autophosphorylation and downstream signaling. A remarkable level of apoptosis is promoted by
6i and this substance is a more effective than cabozantinb as a selective inhibitor on TT cell lines
over noncancerous thyroid Nthy oriꢀ3ꢀ1 cells. Moreover, 6i is more effective in inhibiting
anchorageꢀindependent growth of TT cells rather than cabozantinb.
Taken together, the
capability of 6i to inhibit both RETꢀwild type and its gatekeeper mutant in a highly potent and
selective manner highlights its potential therapeutic value for treatment of RET mutationꢀdriven
diseases. Finally, the isoxazole ring containing pyrazolo[3,4ꢀd]pyrimidinꢀ4ꢀamine 6i provides
insight into structural features and functionality that need to be considered when designing
selective and potent RET inhibitors capable of overcoming RET gatekeeper mutations.
ACS Paragon Plus Environment
27

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
EXPERIMENTAL SECTION
Compounds
0
1
2
3
4
5
6
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 the compounds used in this research were synthesized except cabozantinb, which were
purchased from ChemScene LLC. Every compound was diluted in 100% DMSO stock, and
diluted compound was directly treated to cells by adapting final DMSO concentration to 0.5 %
(v/v).
Cell culture
TT cells were grown in RPMI media supplemented with 15% FBS and 1% antibiotics. Parental
Ba/F3 cells were grown in RPMI supplemented with 10% FBS in the presence of ILꢀ3, and all
RETꢀWT, RETꢀV804L, RETꢀV804M, RETꢀS891A and VEGFR2 transformed Ba/F3 cell lines
were cultured in ILꢀ3ꢀfree RPMI media supplemented with 10% FBS and 1% antibiotics. Nthy
oriꢀ3ꢀ1 cells were grown in RPMI1640 media supplemented with 10% FBS, 2 mM glutamine
and 1% antibiotics.
Kinase profiling and kinase assay
Biochemical kinase assay and kinome profiling were performed by Reaction Biology Corp (San
Diego, CA). In biochemical kinase assay, test compounds were tested with ATP (10 ꢁΜ) and
RET enzymes (RETꢀWT : 2 nM, RETꢀG691S : 90 nM, RETꢀY791F : 1.5 nM, RETꢀV804L : 2.5
nM, RETꢀV804M : 60 nM, RETꢀS891A : 3 nM) in a 10ꢀdose IC50 mode with 3ꢀfold serial
dilution starting at 10 ꢁΜ. In kinase profiling, 6i (1 ꢁΜ) was tested at a concentration of 10 ꢁΜ
ATP against 332 recombinant human kinases in duplicate mode.
ACS Paragon Plus Environment
28

Page 29 of 64
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Growth inhibition of thyroid cell lines
TT cells were seeded at a density of 10,000 cells/ well in a 96ꢀwell plate. The next day,
compound (0.5% DMSO) was treated to the cells in a 10ꢀdose GI mode with 1:3 serial dilution
5
0
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
in DMSO. 10 days later, growth of compoundꢀtreated TT cells relative to DMSOꢀtreated one was
measured using MTT assay kit (Promega, USA). Doseꢀresponse curve was fit utilizing Graphpad
prism 5.0 software. GI values on Nthy oriꢀ3ꢀ1 cells were determined 3 days after compound
5
0
treatment in the same way as TT cells and SD values were determined by replicating each assay
three times independently.
Site-directed mutagenesis (SDM)
Point mutations of RET were generated by site direct mutagenesis using QuickChange kit
(Stratagene) and the following primers.
RETꢀV804L
RETꢀV804M
RETꢀS891A
5'ꢀCCGCTCCTCCTCATCTTGGAGTACGCCAAATꢀ3'
5'ꢀATTTGGCGTACTCCAAGATGAGGAGGAGCGGꢀ3'
5'ꢀCCGCTCCTCCTCATCATGGAGTACGCCAAATꢀ3'
5'ꢀATTTGGCGTACTCCATGATGAGGAGGAGCGGꢀ3'
5'ꢀGAGGGGCGGAAGATGAAGATTGCGGATTTCGGCꢀ3'
5'ꢀGCCGAAATCCGCAATCTTCATCTTCCGCCCCTCꢀ3'
After Dpn I digestion, competent E. coli cells were transformed with the pointꢀmutated plasmids
and spreaded on an ampicilinꢀpositive LB agar plate. After inoculation and shaking incubation
(200 rpm, 37 ℃), plasmid DNA was isolated using plasmid DNA miniprep Kit (Qiagen).
Intended mutation was confirmed by sequencing open reading frame of RET.
ACS Paragon Plus Environment
29