The discovery of PLK4 inhibitors: (E)-3-((1H-indazol-6-yl)methylene) indolin-2-ones as novel antiproliferative agents

Article abstract of DOI:10.1021/jm400380m

The family of Polo-like kinases is important in the regulation of mitotic progression; this work keys on one member, namely Polo-like kinase 4 (PLK4). PLK4 has been identified as a candidate anticancer target which prompted a search for potent and selective inhibitors of PLK4. The body of the paper describes lead generation and optimization work which yielded nanomolar PLK4 inhibitors. Lead generation began with directed virtual screening, using a ligand-based focused library and a PLK4 homology model. Validated hits were used as starting points for the design and discovery of PLK4 inhibitors of novel structure, namely (E)-3-((1H-indazol-6-yl)methylene)indolin-2-ones. Computational models, based on a published X-ray structure (PLK4 kinase domain), were used to understand and optimize the in vitro activity of the series; potent antiproliferative activity was obtained. The kinase selectivity profile and cell cycle analysis of selected inhibitors are described. The results of a xenograft study with an optimized compound 50 (designated CFI-400437) support the potential of these novel PLK4 inhibitors for cancer therapy.

Full text of DOI:10.1021/jm400380m

Subscriber access provided by Columbia Univ Libraries
Article
The Discovery of PLK4 Inhibitors: (E)-3-((1H-indazol-6-
yl)methylene)indolin-2-ones as Novel Anti-Proliferative Agents
Radoslaw Laufer, Bryan Forrest, Sze-wan Li, Yong Liu, Peter Sampson, Louise Edwards,
Yunhui Lang, Don Awrey, Guodong Mao, Olga Plotnikova, Genie Leung, Richard Hodgson,
Irina Beletskaya, Jacqueline M. Mason, Xunyi Luo, Xin Wei, Yi Yao, Miklos Feher,
Fuqiang Ban, Reza Kiarash, Erin Green, Tak W. Mak, Guohua Pan, and Henry W Pauls
J. Med. Chem., Just Accepted Manuscript • DOI: 10.1021/jm400380m • Publication Date (Web): 06 Jul 2013
Downloaded from http://pubs.acs.org on July 13, 2013
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 60
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
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
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 2 of 60
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
The Discovery of PLK4 Inhibitors: (E)ꢀ3ꢀ((1Hꢀindazolꢀ
6
ꢀyl)methylene)indolinꢀ2ꢀones as Novel Antiꢀ
Proliferative Agents
0
1
2
3
4
5
6
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
Radoslaw Laufer, Bryan Forrest, SzeꢀWan Li, Yong Liu, Peter Sampson, Louise Edwards, Yunhui Lang,
Donald E. Awrey, Guodong Mao, Olga Plotnikova, Genie Leung, Richard Hodgson, Irina Beletskaya,
Jacqueline M. Mason, Xunyi Luo, Xin Wei, Yi Yao, Miklos Feher, Fuqiang Ban, Reza Kiarash, Erin
Green, Tak W. Mak, Guohua Pan, Henry W. Pauls*
Campbell Family Institute for Breast Cancer Research, TMDT East Tower, MaRS Centre, 101 College
Street, Toronto, Ontario, MG5 1L7, CANADA
hpauls@uhnresearch.ca
RECEIVED DATE
RUNNING TITLE: The Discovery of Novel PLK4 Inhibitors as AntiꢀProliferative Agents.
CORRESPONDING AUTHOR FOOTNOTE: hpauls@uhnresearch.ca; 416ꢀ581ꢀ7620
ABSTRACT: The family of Poloꢀlike Kinases is important in the regulation of mitotic
progression; this work keys on one member, namely Poloꢀlike Kinase 4 (PLK4). PLK4 has been
identified as a candidate antiꢀcancer target which prompted a search for potent and selective inhibitors
of PLK4. The body of the paper describes lead generation and optimization work which yielded
nanomolar PLK4 inhibitors. Lead generation began with directed virtual screening, using a ligand based
focused library and a PLK4 homology model. Validated hits were used as starting points for the design
and discovery of PLK4 inhibitors of novel structure, namely (E)ꢀ3ꢀ((1Hꢀindazolꢀ6ꢀ
ACS Paragon Plus Environment
1

Page 3 of 60
Journal of Medicinal Chemistry
yl)methylene)indolinꢀ2ꢀones. Computational models, based on a published Xꢀray structure (PLK4
kinase domain), were used to understand and optimize the in vitro activity of the series; potent antiꢀ
proliferative activity was obtained. The kinase selectivity profile and cell cycle analysis of selected
inhibitors are described. The results of a xenograft study with an optimized compound 50 (designated
CFIꢀ400437) support the potential of these novel PLK4 inhibitors for cancer therapy.
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
KEYWORDS: Poloꢀlike Kinase 4; PLK4 inhibitors; Sak; (E)ꢀ3ꢀ((1Hꢀindazolꢀ6ꢀyl)methylene)indolinꢀ2ꢀ
ones; antimitotic agents; breast cancer.
BRIEFS: Molecular modeling guided lead optimization which yielded nanomolar PLK4 inhibitors with
potent antiꢀproliferative activity on breast cancer cells and the results of a mouse xenograft study with
an optimized inhibitor are described.
INTRODUCTION:
A number of therapeutic agents used in the treatment of cancer target tubulin assembly during
1
mitosis. These include the tubulin deꢀstabilizers, such as vinca alkaloids, and newer agents, such as the
taxols, which stabilize microtubules. These successes have encouraged research into nonꢀstructural
mechanisms of inducing aberrant mitosis in tumor cells. As our understanding of mitosis has matured
and the discrete morphological processes have been linked to biochemical reactions, attention has
2
3
focused on the mitotic kinesins and kinases . The coordinated action of the mitotic kinases from the
4
5
6
Cyclinꢀdependent kinase (CDK), Aurora kinase (AURK) and Poloꢀlike kinase (PLK) families is
required for normal mitosis. A priori, any one of the members of the mitotic kinases can be considered
7
as a legitimate target for therapeutic intervention. Much attention has focused on the Aurora kinase
family in recent years, in particular Auroras A and B. Pan Aurora inhibitors have been described as
well as compounds selective for one or other of this pair; a number of these agents have entered the
8
clinic.
ACS Paragon Plus Environment
2

Journal of Medicinal Chemistry
Page 4 of 60
PLKs are a conserved family of serine/threonine kinases that function at the intersection of mitosis
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
9
,10
and DNA damage pathways, and are often deregulated in cancer.
Mammalian cells contain five
proteins: PLK1, Snk/PLK2, Prk/Fnk/PLK3, Sak/PLK4 and PLK5 of which PLK1 is the most
extensively characterized. A number of PLK1 inhibitors have been reported and work has progressed to
1
1
the stage that some have been taken into the clinic. By contrast, PLK4 has not, as yet, received much
0
1
2
3
4
5
6
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
2
attention; this enzyme is structurally the most distinct member of the family. Unlike PLK1, 2, and 3,
PLK4 has only one poloꢀbox and an active site with high homology to the Aurora kinases. PLK4
1
3ꢀ
localizes to the centriole, and is essential for controlling centriole duplication and mitotic progression.
1
7
18ꢀ22
It is a low abundance protein in normal adult tissues, and autoregulates its own stability.
13ꢀ16
Overexpression of PLK4 causes centriole overduplication,
which directly results in chromosome
2
3
segregation errors that may lead to the evolution of more malignant phenotypes. We found that PLK4
2
4,25
depletion caused the death of breast cancer cell lines, but not of normal breast cells.
Examination of
microarray datasets demonstrated that PLK4 is up regulated in breast tumors and specifically the tripleꢀ
negative/basalꢀlike (TNBC) subtype, and that high PLK4 levels in breast tumors are associated with
2
6ꢀ28
poor survival.
Together, these observations identified PLK4 as a potential target for the
development of therapeutics for cancer, and we validated this hypothesis through a series of in vitro and
in vivo experiments. RNAiꢀmediated depletion of PLK4 in breast cancer cells was found to prevent
centriole duplication causing mitotic defects and cell death, and growth suppression of breast cancer
2
4,25
xenografts in vivo.
The potential importance of this enzyme in the development of drugs acting on the human kinome
29
was underscored by Johnson’s publication, which screened a number of known kinase inhibitors
against the PLK family. A striking result was the propensity of this collection to inhibit PLK4, the rest
of the Poloꢀlike kinases were more or less resistant. Three compounds in particular were potent
inhibitors of PLK4, namely 1, 2 and 3 which have been developed as an antiangiogenic, an antimitotic
agent and PKC inhibitor respectively. This result raises the question of the potential mechanistic
ACS Paragon Plus Environment
3

Page 5 of 60
Journal of Medicinal Chemistry
importance of PLK4 inhibition when interpreting the in vitro and in vivo effects of these molecules. For
example a multikinase inhibitor 4 of very different structure has been shown to be potent against
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
0
PLK4 and the authors have speculated that the antiꢀmitotic effect of this compound may be due, at
least in part, to this activity. Isolated reports on PLK4 inhibitors have appeared in the patent literature;
compounds 5 and 6 are representative of the macrocyclic inhibitors described by Janssen
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
3
1,32
Pharmaceuticals.
These compounds are relatively selective for PLK4 vis a vis other members of the
PLK family and inhibit cancer cell growth but little is known about their general properties or in vivo
activity.
The discovery of potent and selective PLK4 inhibitors would help to further understanding of the role
of this enzyme in cancer cell proliferation. Our efforts in the identification of novel PLK4 inhibitors
began with a focused screening approach described herein. One series that emerged from our compound
screens were the indolinones, a well described class of ATPꢀcompetitive kinase inhibitors. These
compounds offered several advantages for lead series generation including acceptable ligand
efficiencies, a testable binding hypothesis and ready synthetic accessibility. Given the relatively large
number of commercial indolinones available, we had the additional advantage of generating SAR with
little to no synthetic effort. The results of our lead generation and subsequent optimization efforts have
3
3
been preliminarily disclosed and are fully described in the body of this paper. The initial synthetic
work was driven by several goals, namely, a desire to identify new chemotypes for subsequent
elaboration, verifying binding hypotheses and ultimately the identification of drug like inhibitors.
RESULTS and DISCUSSION:
Lead finding. As a starting point in our leadꢀfinding effort, we built a focused kinase library using
two different approaches: 1. a ligandꢀbased library of ~6,000 potential kinase inhibitors using molecular
3
4
fingerprint consensus scoring and 2. a targetꢀbased virtual screening set of ~1,700 molecules identified
through an inꢀhouse PLK4 homology model. This homology model was built following a well
3
5
described method, using multiple stages of tethered refinement. This process is highly conservative; it
ACS Paragon Plus Environment
4

Journal of Medicinal Chemistry
Page 6 of 60
ensures that initial clashes are resolved and do not lead to sudden large structural changes during
molecular dynamics. The homology model generated in this manner was validated by docking available
PLK4 ligands and inspecting the consistency of their binding modes. When the first Xꢀray structure of
PLK4 became available during the course of this work (PDB code: 3COK), an overlay of the PLK4
homology model to this structure was performed. This exercise revealed that the two were in good
agreement, with the heavy atom RMS distance of the two active site structures being a mere 0.8Å.
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
36
37
Another retrospective validation of the homology model was obtained using GOLD and Glide XP
docking: the binding modes of the PLK4 ligands, when docked into the homology model and the Xꢀray
structure, were identical. This model also provided an understanding of the interactions between
screening hits and the kinase active site and was used in subsequent inhibitor design.
The focused kinase libraries generated above were augmented with a commercial collection of ~280
known kinase inhibitors. The merged compound collection was screened against PLK4 using an indirect
ELISA assay. Compounds inhibiting >75% at the screening dose of 10 ꢀM were defined as hits and
subjected to a validation exercise. Hits were clustered into chemical series using the SAR report within
MOE; IC s were determined for selected inhibitors and their chemical integrity checked by LCMS.
5
0
Representative compounds from series of further interest were subjected to kinetic analysis in the
biochemical assay and their potential binding modes were determined in a PLK4 homology model.
Of the several series identified by the focused screening exercise, the 4ꢀhydroxybenzylideneindolinꢀ2ꢀ
ones held the most promise. The indolinone hits were present in all three sources of molecules, i.e. the
ligand and homology modelꢀbased libraries as well as the known kinase inhibitors. A kinetic
characterization of a representative 4ꢀhydroxybenzylideneindolinꢀ2ꢀone 7 indicated that this compound
was an ATP competitive inhibitor of PLK4 (Ki = 0.7 ꢀM). The initial screening hits were augmented by
the acquisition of additional commercially available indolinones. SAR relationships (for PLK4 and
PLK1) could be discerned; the key points are illustrated in Table 1. It appears that the indolinone NH is
important for activity against both PLK4 and PLK1 (compounds 8 vs 9), as is the phenolic hydroxyl
(compound 10). A flat SAR was observed for PLK1 and PLK4 with respect to changes in substitution
ACS Paragon Plus Environment
5

Page 7 of 60
Journal of Medicinal Chemistry
patterns of the phenolic ring (11ꢀ13); PLK4 was tolerant of an additional hydroxyl on the ring whereas
PLK1 was not (compound 14).
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
Inhibitor Design. Compound 11 was used as the starting point for new inhibitor design. Indolinones
38
of this sort exist as Z or E isomers and an equilibrium mixture, generally 1:1, is attained in aqueous
solution. It was not possible to unequivocally determine the favored binding configuration of the 4ꢀ
hydroxybenzylideneindolinꢀ2ꢀones, however, our working model (Figure 1) was consistent with
preferential binding of the Z geometric isomer. The essential features of the SAR were in agreement
with our PLK4 binding model (Figure 1, left panel) which ascribed a role for the indolinone amido
function (i.e hydrogen bonding to Glu90 and Cys92 of the hinge region) and the phenolic hydroxyl
0
1
2
3
4
5
6
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
(weak interaction with His93). Our design idea was to morph the phenol into bioisosteric heterocycles
15ꢀ17 (Figure 1, right panel) which could retain the Hꢀbonding capability of the parent. A model of
triazole 15 bound into the active site of PLK4 overlays well with the binding mode of compound 11
(Figure 1, left panel). These new compounds embody additional improvements in that the undesirable
phenol is replaced and they represent unexplored IP space. The synthesis of these molecules proved to
be straight forward (Scheme 1) and it was possible to further derivatize this core in the pursuit of more
potent congeners (vida infra).
Chemical Synthesis. Triazole 15, (1Hꢀindazolꢀ5ꢀylꢀmethylene)indolinꢀ2ꢀone 16 and the simple (1Hꢀ
indazolꢀ6ꢀylꢀmethylene)indolinꢀ2ꢀones 17 and 20-27 can be readily obtained by a oneꢀstep Knoevenagel
condensation between the requisite aldehydes 18 and indolinones 19 (Scheme 1). In all instances, the E
geometric isomer predominates; occasionally a significant amount of Z is also present. For compounds
2
0
and 21 the stereoisomers were separated chromatographically. Stereoisomeric
3
(arylmethylene)indolinꢀ2ꢀones can be characterized by their J(C=O, H) couplings, measured through
39
protonꢀdetected heteronuclear experiments such as HMBC. EXSIDE (ExcitationꢀSculptured Indirectꢀ
Detection Experiment) NMR can also be employed to determine long range CH coupling constant
4
0
3
measurements; this technique was used to determine the J(C=O, H) couplings for 20 and 21 as 6.4
and 13.2 Hz respectively. Alternatively, and more generally, the Zꢀisomer was identified by a diagnostic
ACS Paragon Plus Environment
6

Journal of Medicinal Chemistry
Page 8 of 60
peak for proton 7 of the indazole ring at 8.9ꢀ9.0 ppm. The downfield shift of proton 7 is attributed to the
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
4
1
proximity of the carbonyl of the indolinone in the Zꢀisomer; this deshielding effect is not observed in
the Eꢀisomer and hence the chemical shift of indazole proton 7 in this configuration is observed up field
of 8.0 ppm.
Several pathways (Scheme 2) were employed to synthesize compounds with aryl vinyl substitution on
the indazole ring. For example, 3ꢀiodoꢀ1Hꢀindazoleꢀ6ꢀcarbaldehyde 28 was protected with a SEM group
to give a mixture of regioisomers 29 that could be used in subsequent Heck couplings without prior
separation. The protected aryl vinyl indazoles 30, were subjected to Knoevenagel condensation with
indolinones 31 to yield intermediates 32. Deprotection of the SEM group was effected by various
treatments (e.g. Lewis acid; HCl or TBAF/THF) to yield the desired compounds 41ꢀ45 and 48ꢀ50. In a
further variation of this route, deprotection of intermediate 30 to 33 (Ar = pyridinꢀ2ꢀyl, Ph and pꢀ
0
1
2
3
4
5
6
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
(
Me NCH )Ph) preceded the final condensation to give inhibitors 37, 46 and 47 respectively,.
2 2
An alternate route began with HornerꢀWadsworthꢀEmmons olefination of aldehyde 34 to give
intermediates 35. In this procedure, subsequent reduction of the pendant nitrile yielded aldehydes 33.
The standard Knoevenagel condensation gave final products 38ꢀ40. In the case of inhibitor 51 (Ar = 4ꢀ
carboxyphenyl) the desired carboxylic acid 36 was revealed by base hydrolysis of the precursor ester,
prior to the Knoevenagel condensation.
Lead Optimization. We were gratified to see that these new compounds e.g. 15ꢀ17 (Table 2)
inhibited PLK4 in the single digit micromolar range or better. Our working models (Figures 1 and 2A)
suggested that 5 or 6 substitution of the indolinone with Hꢀbond acceptors could result in additional
interactions with Lys and compounds were prepared to explore this hypothesis. Analogs of the most
potent indazole 17, were rapidly synthesized to yield a collection of inhibitors 20ꢀ27 with IC s ranging
5
0
from double digit nanomolar to micromolar (Table 2). The selectivity of the inhibitors against the Poloꢀ
like kinase family members, namely PLK1, 2 and 3, increased with increasing PLK4 potency.
ACS Paragon Plus Environment
7

Page 9 of 60
Journal of Medicinal Chemistry
These encouraging results were tempered by the observation that the discernible SAR appeared to be
1
2
at odds with our working models on several points. First, the predominance of the E isomer negates any
possible interaction between His93 and the heterocyclic Hꢀbond donor. Yet, in a head to head
comparison, the E isomer, compound 20 was somewhat more potent than the Z isomer, 21. The double
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
8,42
bond of (arylmethylene)indolinꢀ2ꢀones has the potential to isomerize in aqueous and alcoholic media
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
and we have observed the conversion of homogenous compound 20, to 3:2 E/Z mixture within 6 h in
DMSOꢀd at rt. Despite the potential for isomeric lability in solution, it appears that the geometric
6
configuration of the starting solid informs in vitro results. This is consistent with the observation that no
isomerization occurs upon incubation of compound 20 under the buffer condition mimicking the PLK4
kinase assay.
Hꢀbond acceptors/donors at position 5 and 6 of the indolinone resulted in a loss of activity
(compounds 25ꢀ27) whereas hydrophobes and halogens were favored at the 5 position (compounds 20,
2
2 and 24). These observations were also inconsistent with the original binding hypothesis. To begin to
understand the nature of the binding of the 3ꢀ((1Hꢀindazolꢀ6ꢀyl)methylene)indolinꢀ2ꢀones, a kinetic
analysis of compound 17 was performed (Supporting Material). This study showed that compound 17
displays kinetics consistent with direct competition for ATP binding to the active site of PLK4 with an
apparent K of 22 nM. These results, coupled with an understanding that indazoles are capable of
i
making hinge region interactions in their own right, prompted a reexamination of the binding models.
The two possibilities are illustrated in Figure 2: the left panel shows the indolinone making two Hꢀbonds
to the hinge region (Glu90 and Cys92); the right panel shows the indazole interacting with the hinge
region and also ascribes a role for the indolinone i.e Hꢀbond to Asp154. Using the homology model, it
was not possible to distinguish between these two modes by computational docking alone.
29
The publication of Johnson’s seminal work on the inhibition of the PLK family by kinase inhibitors
suggested a way to test the binding model corresponding to Figure 2B. This work reported that Axitinib,
an RTK inhibitor and antiꢀangiogenic agent, was also a potent inhibitor of PLK4 (IC = 46 nM in our
5
0
hands) and it docked and scored highly in our binding model. Overlay of Axitinib (an indazole with
ACS Paragon Plus Environment
8

Journal of Medicinal Chemistry
Page 10 of 60
putative hinge region interactions) with our own inhibitors generates a hybrid target molecule, namely
7. This pyridinꢀ2ꢀyl vinyl analog was docked into the 3COK model using GlideXP and was predicted
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
to potently inhibit PLK4 (Figure 3A). The 3COK model, generated from the PLK4 Xꢀray structure, was
optimized to improve quantitative predictivity as measured by the correlation of docking scores and
pIC50 values (see computational methods in the experimental section). Whereas the position of the
indazole and indolinone groups are unchanged from the parent 17 (i.e. the Hꢀbonds to the hinge are
retained), the vinyl linker places the pyridyl group in a good position for hydrophobic interactions (e.g.
with Leu18). In the event, compound 37 inhibited PLK4 with an IC of 13 nM (Table 3). This tenfold
0
1
2
3
4
5
6
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
improvement in activity is in good agreement with the log unit increase predicted by the model. It
should be noted that the PLK4 IC of the styryl compound 46 suffers in comparison to 37 by 3 fold, but
5
0
still represents a significant improvement over the parent, 17.
This result prompted the synthesis of the other pyridine positional isomers. The pyridinꢀ(3 & 4)ꢀyl
isomers 38 and 41 respectively, were more active than the pyridinꢀ2ꢀyl isomer (Table 3). Presumably,
the polar nitrogen is more readily accommodated in the aqueous milieu at the protein surface. An
additional increase in potency was realized by incorporation of the 5ꢀmethoxy substitution on the
indolinone ring (compounds 40 and 44); alkyl groups (39, 43), chlorine (42) and hydroxy (45) were
somewhat less effective. Once again, our model correctly predicted the improvement observed upon
substitution of a methoxy group at the 5ꢀposition of the indolinone. However, the measured IC s for
5
0
inhibitors 40 and 44 must be viewed as apparent values since they approach the nominal concentration
of enzyme in the assay i.e. 0.4 nM. Figure 3B displays compound 44 docked into the active site of the
PLK4 enzyme. Apart from the usual hingeꢀbinding to the backbones of Glu90 and Cys92, there is also a
hydrogen bond to the side chain of Asp154. There is extensive hydrophobic contact between the enzyme
and the ligand, especially at the aromatic rings and the methoxy group which nestles in a groove
bounded by the protein backbone and the side chains of Leu143 and Glu96. In contrast, the Zꢀisomers of
the aryl vinyl indazoles given in Table 3 generally did not dock into the PLK4 active site and those that
did had docking scores up to 3 log units lower than the Eꢀisomers.
ACS Paragon Plus Environment
9

Page 11 of 60
Journal of Medicinal Chemistry
Compounds 37ꢀ45 represented a significant advance; however, their potent PLK4 inhibition did not
result in a uniformly potent antiꢀproliferative effect on breast cancer cell lines. As a group these
compounds had variable effects on cell growth inhibition; GI s were in the subꢀmicromolar range for
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
5
0
the MCFꢀ7 and MDAꢀMBꢀ468 cell lines, only marginally better than Axitinib. A correlation between
PLK4 potency and GI s was discernible in the MCFꢀ7 and MDAꢀMBꢀ468 cell lines, but the MDAꢀ
0
1
2
3
4
5
6
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
50
MBꢀ231, SKBrꢀ3 and HCCꢀ1954 lines appeared to be less responsive. The effects of the potent PLK4
inhibitor 41 on cell cycling of MDAꢀMBꢀ468 cells are shown in Figure 4. Concentrations of 41 at 0.5
ꢀM and above cause mitotic checkpoint loss, polyploidy and death, underscored by the accumulation of
8
N and 16N peaks.
Another drawback to the pyridinyl inhibitors is captured in Table 4. Compound 41, for example,
demonstrated significant inhibitory effects on CYP450’s 2C9 and 2C19, which is ascribed to the
aromatic ring nitrogen. With a view to attenuating this activity, and understanding that biochemical
potency does not always translate into cellular activity, the physicochemical properties of the PLK4
inhibitors were calculated. Data on representative compounds is given in Table 4. The physicochemical
properties of the series were modified by incorporation of waterꢀsolubilizing substituents on the phenyl
ring distal from the indazole core. The beneficial effects of these substitutions on solubility are captured
by the ACD calculated values given in Table 4 and that this is achieved without sacrificing lipophilicity
is reflected by the calculated logP. Molecular modeling indicated that this region of the inhibitor was
projected into solvent and would have no deleterious effect on PLK4 potency (see Figure 3B). For
example, the addition of tertiary amines at this position yielded potent PLK4 inhibitors 47ꢀ50,
moreover, their inhibition of cell growth was significantly better than pyridinyl compounds 37ꢀ45
(Table 3). In contrast, appending a carboxylic acid function at the para position of the phenyl ring
yielded a subꢀnanomolar inhibitor 51, however, a decrease in cellular activity was observed. Once again,
not all breast cancer cell lines were responsive to compounds 47ꢀ50 to the same degree. MCFꢀ7, MDAꢀ
MBꢀ468 and MDAꢀMBꢀ231 had GI s in the double digit nanomolar range; growth inhibition of SKBRꢀ
5
0
3
and HCCꢀ1954 was observed only at micromolar concentrations. No relationship was observed
ACS Paragon Plus Environment
10

Journal of Medicinal Chemistry
Page 12 of 60
between response to inhibitors and PLK4 mRNA levels, which are comparable across the cell lines
tested (see Experimental Methods). A lack of correlation between target expression and responsiveness
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
4
3
of cancer cells to pharmacologic inhibition of the target has been reported for other mitotic kinases.
Gratifyingly, compounds 47 and 50 demonstrated not only enhanced growth inhibitory activity but also,
attenuated CYP450 inhibition (Table 4).
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Across a broad range of chemotypes and biochemical activities, we were able to show a correlation
between the measured GI in MDAꢀMBꢀ468 cells and pGI calc (Figure 4). QSAR analysis established
5
0
50
that calculated pGI was best described by three independent variables, namely PLK4 pIC , molecular
5
0
50
2
volume and logP with an R of 0.52 and RMSE of 0.62. The level of variance observed indicates that
additional unrecognized factors may be important. According to the predictive equation given in the
computational methods, PLK4 inhibition was the most important contributor. However, it’s expected
that the ability of a PLK4 inhibitor to cross cellular membranes (as reflected by its logP) would factor
into antiꢀproliferative activity.
To further illustrate this point, PAMPA values were obtained on representative compounds given in
Table 4. All six are potent PLK4 inhibitors; however the compounds with low permeability (hydroxy
substituted indolinone 45 and carboxylate 51) showed correspondingly poor activity in the cell growth
assays. Pyridine 41 is 2 orders of magnitude more permeable but this did not translate into potent cell
growth inhibition. Some of the most potent inhibitors of breast cancer cell growth, tertiary amines 47,
49 and 50 also had high PAMPA values; it would appear that permeability and potent PLK4 inhibition
are necessary but not sufficient for robust cell activity.
Kinase Selectivity. The PLK4 inhibitor, 50 and was profiled against a panel of 274 protein
kinases at a screening concentration of 1 ꢁM. This data is presented as a heat map (Figure 6) which
indicates that compound 50 displayed better selectivity than Sunitinib and somewhat less than Axitinib,
two ATP competitive kinase inhibitors used clinically. Of the 274 kinases tested, 30 were inhibited at
greater than 90 %, at the screening concentration of 1 ꢁM (Supporting Information). For a subset of the
more potently inhibited kinases of interest, IC s were determined using inꢀhouse assays. Compound 50
5
0
ACS Paragon Plus Environment
11

Page 13 of 60
Journal of Medicinal Chemistry
inhibited Aurora A, Aurora B, KDR and FLTꢀ3 with IC s of 0.37, 0.21, 0.48, and 0.18 ꢁM,
5
0
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
respectively, i.e. two orders of magnitude higher than its IC of 0.6 nM against PLK4.
5
0
In vivo Efficacy. Given its potent antiꢀproliferative activity against the MDAꢀMBꢀ468 breast
cancer cell line, compound 50 was selected for testing in the corresponding mouse xenograft model. In
preparation for this in vivo experiment, we determined the plasma levels of compound 50 in mice upon
IP dosing at 50 mg/kg; a Cmax of 92 ng/mL and AUC of 190 ng.h/mL were observed. The mouse plasma
protein binding measurement for compound 50 is 99% i.e. unbound compound in plasma is 1%. Thus,
the extrapolated Cmax for the xenograft IP dose (25 mg/kg) would equate to the single digit nanomolar
range, well below the offꢀtarget kinase activities measured (>100 nM). At this point, we can say nothing
definitive about target engagement by the PLK4 inhibitor 50 since compound enrichment in tumors
and/or kinase active sites is possible. Figure 7 shows the in vivo efficacy results using an MDAꢀMBꢀ468
mouse xenograft model: at 50 mg/kg daily IP dose, VX680, a literature standard, achieved a substantial
reduction in tumor volume compared to the negative control (70 % TGI). Upon dosing the PLK4
inhibitor 50 at 25 mg/kg IP, no increase in tumor volume was observed, in fact, the tumor regressed in
size (120 % TGI). There was no significant weight loss in any of the arms of the study. Clearly, the
PLK4 inhibitor 50 compared favorably with VX680 in this xenograft model.
0
1
2
3
4
5
6
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
CONCLUSIONS:
PLK4, identified by an inꢀhouse siRNA screen and supporting validation efforts as a tractable, novel
target for drug discovery, was subjected to a lead generation effort. A two pronged focused screening
campaign identified 3ꢀ(hydroxybenzylidene)indolinꢀ2ꢀone hits for subsequent validation work.
Bioisosteric replacement of the phenolic function of the validated hits resulted in the identification of
(E)ꢀ3ꢀ((1Hꢀindazolꢀ6ꢀyl)methylene)indolinꢀ2ꢀones as a novel class of kinase inhibitors. Systematic
optimization of this series yielded inhibitors with nanomolar activity against PLK4 and concomitant
antiꢀproliferative effects against a panel of breast cancer cell lines. Models of the inhibitor series in a
3
COK derived structure help explain and predict the observed PLK4 activity.
ACS Paragon Plus Environment
12

Journal of Medicinal Chemistry
Page 14 of 60
One inhibitor was selected for detailed study; compound 50 (designated CFIꢀ400437) is potent against
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
PLK4 (IC = 0.6 nM) and selective against other members of the PLK family (> 10 ꢀM). Against a
50
broad panel of kinases, only a small subset is significantly inhibited. However, like other ATP
competitive kinase inhibitors, the cell activity and in vivo effects seen with 50 cannot be ascribed solely
to the inhibition of PLK4. Compound 50 is a potent inhibitor of MCFꢀ7, MDAꢀMBꢀ468 and MDAꢀMBꢀ
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
2
31 cell growth and very effective in a mouse xenograft model of tumor growth. The data reported
herein on this new class of antimitotic agents, culminating in the in vivo efficacy for compound 50,
illustrates the potential of these novel PLK4 inhibitors for cancer therapy. Further optimization work
towards the identification of a development candidate will be disclosed in due course.
EXPERIMENTAL:
Computational methods:
Most of the computational processes, including ligand preparation, ligandꢀbased virtual screening,
4
4
homology model building and QSARꢀtype analyses were performed using the MOE software.
3
4,45
A focused kinase library was generated using a ligandꢀbased virtual screening approach
with the
aim of identifying a diverse set of readily available compounds with reasonable physicochemical
properties that are ‘kinase inhibitorꢀlike’. This approach began with selecting a training set of 379
diverse compounds that have better than 100 nM activities in different protein kinase assays, drawn
4
6,47
from a collection of ~2700 known kinase inhibitors in the Wombat database.
This training set was
utilized in molecular fingerprint based consensus scoring using the MACCSF, TGTFO and MP61
45
fingerprints. The screened dataset comprised of ~3 million commercially available compounds, put
together from 12 leading compound vendors. The initial hitꢀset of ~50,000 compounds was obtained by
selecting the top scoring ~30,000 hits each from the MACCSF and TGTFO hit lists as well as the
consensus list obtained using the sum rank of all three. This set further was reduced utilizing property
filters (250 < MW < 600, number of Hꢀbond donors <5, number of Hꢀbond acceptors <10, number of
chiral centres <5, total number of rotatable bonds <11 and total number of contiguous rotatable bonds
<
7, as well as a substructureꢀbased filter to remove reactive molecules, based on the MOE descriptor
ACS Paragon Plus Environment
13

Page 15 of 60
Journal of Medicinal Chemistry
‘
reactive’ with some inꢀhouse modifications). The remaining compounds were clustered using the
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
MACCS fingerprint into ~6500 clusters using the top scoring hits from each, and then some compounds
were removed after manual selection, yielding ~6000 compounds, purchased from commercial sources.
A homology model of the PLK4 kinase domain was generated. The PLK4 homology model was built
4
8
using the xꢀray structure of AuroraꢀA coꢀcrystallized with PHAꢀ680632 (PDB code: 2BMC). This
structure has 37% residue identity in the kinase domain with PLK4, using the blosum35 substitution
matrix. Primary sequence alignment was followed by threading the PLK4 residues to the AuroraꢀA
structure, fully preserving the main chain secondary structure (i.e. no minimizations were performed on
the intermediate and final structures). The solution with the best packing score was selected. After this,
a threeꢀstage refinement process was performed using tethered molecular dynamics (MD) and energy
minimizations using the MMFFx forcefield in MOE. These tethers were added to all dihedral
connections within the protein structure in order to prevent significant perturbations and movements in
the positions of heavy atoms. First, strong dihedral tethers were used (for the backbone a force constant
of 100 kcal/mol/deg, applied on torsional angle changes over 0.2º, whereas for the sidechains and ligand
they were 10 kcal/mol/deg and 0.5º, respectively), then somewhat weaker ones (backbone 10
kcal/mol/deg and 0.5º, sidechains and ligand 1 kcal/mol/deg and 1º) and finally no tethers at all. The
dynamics runs included 100 ps heating to 300K, followed by a simulation carried on until the energy
seemed to stay constant, followed by energy minimization to a gradient of 0.05 kcal/mol/Å. In the first
stage the simulation took 150 ps, in the second it was 900 ps and in the third 2340 ps. Docking was
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
3
6
37
performed using the GOLD and Glide software using standard settings. In all cases constraints were
applied for the hinge binding contacts (at least one contact from the list of Glu90 carbonyl, Cys92 NH
or carbonyl), to ensure that all solutions have the expected hinge binding mode. An overlay of the
homology model to the first Xꢀray structure of PLK4 (3COK: 2.25Å resolution, coꢀcrystallized with an
ATP analog) was performed. The RMS distance of two active sites was 0.8Å when considering heavy
atoms for residues within 12Å of the cognate ligand.
ACS Paragon Plus Environment
14

Journal of Medicinal Chemistry
Page 16 of 60
The homology model described above was used for virtual screening purposes. Approximately 3
million commercial compounds were initially considered. Using somewhat different property filters
than above (250 < MW < 600, number of chiral centres <3 and the same substructureꢀbased reactivity
filter), 2 million molecules were further considered. These were protonated/deprotonated using the
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
4
9
Wash process, built into 3D using stereoisomer generation with the stergen setting in Corina and
minimized using the MMFF94x forcefield in MOE. These molecules were then docked with GlideSP
using the hinge binding constraints on the carbonyl of Glu90 and the carbonyl and NH of Cys92 (at least
one of these to be satisfied). The obtained hit list was filtered using the GlideSP score, corresponding to
0
1
2
3
4
5
6
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
5ꢀM activity on the training set, yielding ~158 thousand compounds. This set was then docked using
two methods: GlideXP and GOLD (using default settings), again using hingeꢀbinding constraints. In
order to ensure that a balanced set of hits is identified, the following selection process was applied: the
top 500 molecules were selected using the top scoring GOLD poses rescored using the dockpKi
5
0
function in MOE , the top 1000 using the GlideXP score, the top 500 using calculated ligand efficiency
in GlideXP and the top 2,500 using sum rank consensus score the GOLD/dockpKi and GlideXP. The
5
1
hits were then filtered to select molecules with hinge binding interactions precedented in the literature
yielding ~3100 ligands. After removing duplicate molecules, clustering using the MACCS fingerprints
and manual selection, 1956 compounds were purchased. Compounds of Table 1 were purchased from
the following sources: 7, 9, 11 and 14, ChemDiv, San Diego, CA, USA; 13, Life Chemicals, Burlington,
ON, Canada; 12, Enamine Ltd., Kiev, Ukraine; 8, Ambintner, c/o Greenpharma, Orléans, FRANCE; 10,
ChemBridge Corporation, San Diego, CA, USA. LCMS data (i.e. purities and parent masses) can be
found in the supporting materials.
A second model was generated using the PLK4 3COK Xꢀray structure deposited in the PDB. The Xꢀ
ray structure differed significantly from the homology model above at two residues namely, Lys41 and
Asp154. The 3COK structure was modified by generating rotamers of these sideꢀchains with the MOE
rotamer explorer, minimized in place with the rest of the structure fixed and the structures obtained were
then used to generate docking models. These docking models were compared using the correlation of
ACS Paragon Plus Environment
15

Page 17 of 60
Journal of Medicinal Chemistry
docking scores and pIC ’s on 270 ligands. The best quantitative predictivity was obtained when the
5
0
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
Lys41 rotamer was similar to that in the homology model and the Asp154 was left in its original
position. With this modification the predictive performance of the xꢀray structure based model using
2
GlideXP improved dramatically (R of 0.54 and RMSE of 0.82 on the same set). The performance of
this model was monitored throughout the course of this work and the correlation coefficient did not
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
2
deviate significantly from an R of 0.5. This compares favourably to the initial homology model which
2
had modest predictive power on the ligand set available at the time (R of 0.3 and RMSE of 0.93 on 270
ligands).
QSAR models for GI data were generated using standard procedures, only using a small pool of ~15
5
0
descriptors from 4 categories that were expected to contribute to cell growth inhibition. These included
the biochemical IC for PLK4, a measure of size (e.g. molecular weight, volume, number of rotatable
5
0
bonds), a measure of lipophilicity (logP, logD) and calculated solubility. The model was initially built
52
using a training set of 110 compounds and was validated using forward prediction . The best model on
the training set included PLK4 activity, as well as the McGowan characteristic molecular volume and
53
logP. The predictive equation is given below:
−5
pGI =1.98 + 0.38⋅ pIC
+ 0.28⋅log P + 6⋅10 ⋅V
5
0
50,PLK 4
McGowan
Figure 5 displays the performance of the same model on the full set of 181 molecules, obtained at a
later time point and includes all molecules available at the time. Note that the performance of the model
remained substantially unchanged at later time points (n>400, not shown).
Synthesis:
General Experimental Methods. Commercially available starting materials, reagents, and
solvents were used as received, with the exception of N,Nꢀdimethylꢀ1ꢀ(4ꢀvinylphenyl)methanamine
which was purified by chromatography on silica gel prior to use in Heck reactions. In general,
anhydrous reactions were performed under an inert atmosphere such as nitrogen or argon. Microwave
reactions were performed with a Biotage Initiator microwave reactor. Yields were not optimized.
ACS Paragon Plus Environment
16

Journal of Medicinal Chemistry
Page 18 of 60
Reaction progress was generally monitored by TLC using Merck silica gel plates with visualization by
UV at 254 nm, by analytical HPLC or by LCMS (Bruker Exquire 4000). Flash column chromatographic
purification of intermediates or final products was performed using 230ꢀ400 mesh silica gel 60 from
EMD chemicals. Final products were sometimes purified by preparative reverseꢀphase HPLC.
Purification was performed on a Varian PrepStar model SDꢀ1 HPLC system with a Varian Monochrom
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
1
1
0u Cꢀ18 reverseꢀphase column using a gradient of about 5ꢀ30% acetonitrile/ 0.05% TFA water to 70ꢀ
00% acetonitrile/0.05% TFA water over a 20ꢀ40ꢀmin period at a flow rate of 30ꢀ50 mL/min. Fractions
containing the desired material were concentrated and/or lyophilized to obtain the final products.
1
Proton nuclear magnetic resonance ( H NMR) spectra were recorded on a Bruker AV 400 MHz
or Agilent DD2ꢀ500 spectrometers. Chemical shifts were reported in parts per million (ppm) from
tetramethylsilane. Proton coupling patterns were described as singlet (s), doublet (d), triplet (t), quartet
(q), multiplet (m), and broad (b). The E:Z configuration of (indazolyl)methylene indolinone linker was
1
3
determined by H NMR spectroscopy. Stereoisomers were characterized by their J(C=O, H) couplings,
3
3
measured through EXSIDE NMR experiments: Jcis = 7ꢀ8 Hz for the Z isomer and Jtrans = 10ꢀ12 Hz
3
9,40
for the E isomer.
The configuration of the alkene linker could also be assigned by the shift of a
diagnostic peak for proton 7 of the indazole ring: 8.9ꢀ9.1 ppm for the Zꢀisomer and 7.8ꢀ8.1 ppm for the
4
1
Eꢀisomer.
The purity of compounds described within was determined by one of two HPLC methods: (A)
using a Phenomenex Gemini C18 column (50 mm x 4.6 mm, 3ꢁm) with a gradient of 10ꢀ90%
acetonitrile in water (with 0.1% formic acid in each mobile phases), with UV detection at λ = 254 or
2
14 on an Agilent 1100 Series; and/or (B) using a Varian MonoChrom C18 column (100 mm x 4.6 mm,
0 ꢁm) with a gradient of 10ꢀ90% acetonitrile in water (with 0.1% TFA in each mobile phases), with
1
UV detection at λ = 254 and 214 on a Varian ProStar 335 LC detector. The purity of all final
compounds was shown by HPLC to be ≥95%.
ACS Paragon Plus Environment
17

Page 19 of 60
Journal of Medicinal Chemistry
Low resolution mass spectra (MS ESI) were obtained using a Bruker Esquire 4000 spectrometer
1
2
under electron spray ionization conditions. High resolution mass spectra (HRMS) were obtained by a
Waters Acquity UPLC coupled with a Waters Synapt G2 timeꢀofꢀflight mass spec, using an ESI source.
Compounds in DMSO solutions (5 mg/mL) were diluted to 0.01 mg/mL; 5 uL was injected on to a
Waters BEH C18 1.7um 2.1x100 mm column equilibrated at 10% acetonitrile in water, spiked with 0.1
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
%
formic acid at a flow rate of 0.5 mL/min. Bound analytes were eluted by a linear gradient of 10 to
9
5% acetonitrile in water, between 1 and 8 minutes. The mass spectrum was calibrated using formate
clusters; leucine enkephalin was used as a "lock mass" for additional, realꢀtime, accurateꢀmass
correction.
(E)-3-((1H-benzo[d][1,2,3]triazol-5-yl)methylene)indolin-2-one (15). An EtOH (4 mL) solution of
indolinꢀ2ꢀone (101 mg, 0.76 mmol), 1Hꢀbenzo[d][1,2,3]triazoleꢀ5ꢀcarbaldehyde (123 mg, 0.84 mmol),
o
piperidine (7.5 ꢀL, 0.08 mmol) in a sealed scintillation vial was heated at 90 C for 2 h. The resulting
precipitate was washed with EtOH (20 mL) and dried to yield the title compound as a yellow solid (149
1
mg, 75 %). H NMR (400 MHz, DMSOꢀd ) δ 15.94 (s, 1H), 10.65 (s, 1H), 8.27 (bs, 1H), 8.05 (bs, 1H),
6
7
.80 (s, 1H), 7.77 (bs, 1H), 7.51 (d, J = 7.6 Hz, 1H), 7.24 (t, J = 7.6 Hz, 1H), 6.88 (d, = 7.6 Hz, 1H),
+
6
.84 (d, J = 7.6 Hz, 1H); HRMS (ESI) m/z calcd for [C H N O + H] 263.0933, found 263.0931;
23 16
4
HPLC A: 98A% at 254 nm.
(E&Z)-3-((1H-indazol-5-yl)methylene)indolin-2-one (16). An EtOH (2 mL) solution of indolinꢀ2ꢀ
one (67 mg, 0.50 mmol), 1Hꢀindazoleꢀ5ꢀcarbaldehyde (73 mg, 0.50 mmol) and piperidine (5.0 ꢀL, 0.05
o
mmol) was heated at 90 C in a sealed scintillation vial for 2 h. The EtOH was removed under reduced
pressure and the crude mixture was purified by preparative reverse phase HPLC to give the title
1
compound as a yellow solid (10 mg, 7.6%) with a 12:1 E to Z ratio. H NMR (400 MHz, CD OD) δ
3
8
.20ꢀ8.18 (m, 2H), 7.89 (s, 1H), 7.76 (d, J = 7.9 Hz, 1H), 7.71ꢀ7.67 (m, 2H), 7.24 (t, J = 7.4 Hz, 1H),
+
6
.93 (d, J = 7.7 Hz, 1H), 6.90 (t, J = 7.5 Hz, 1H); MS (ESI) m/z calcd for [C H N O + H] 262.1,
16 11
3
found 262.0; HPLC A: >95A% at 214 nm.
ACS Paragon Plus Environment
18

Journal of Medicinal Chemistry
Page 20 of 60
(
E)-3-((1H-indazol-6-yl)methylene)indolin-2-one (17). According to the method described for
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
1
compound 15, the title compound was prepared as a yellow solid (35 mg, 27 %). H NMR (400 MHz,
CD OD) δ 8.14 (s, 1H), 7.91 (d, J = 8.5 Hz, 1H), 7.89 (s, 2H), 7.65 (d, J = 7.6 Hz, 1H), 7.47 (d, J = 8.4
3
Hz, 1H), 7.25 (t, J = 7.7 Hz, 1H), 6.93 (d, J = 7.8 Hz, 1H), 6.87 (t, J = 7.6 Hz, 1H); HRMS (EI) m/z
+
calcd for [C H N O] 261.0902, found 261.0908; HPLC A: >95A% at 214 nm.
0
1
2
3
4
5
6
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
16 11
3
(E)-3-((1H-indazol-6-yl)methylene)-5-fluoroindolin-2-one (20) and (Z)-3-((1H-indazol-6-yl)-
methylene)-5-fluoroindolin-2-one (21). A scintillation vial was charged with 5ꢀfluoroindolinꢀ2ꢀone (40
mg, 0.26 mmol) in MeOH (2 mL), 1Hꢀindazoleꢀ6ꢀcarbaldehyde (43 mg, 0.292 mmol) and piperidine (3
o
uL, 0.03 mmol) The reaction was heated to 60 C for 2 h, cooled to rt and filtered (LCMS analysis of the
yellow solid that formed indicated a 9:1 mixture of geometric isomers). The mixture was separated by
1
prep HPLC to give of the E isomer, 20 as yellow solid (11 mg, 15 %). H NMR (400 MHz, DMSOꢀd ) δ
6
1
3.31 (s, 1H), 10.66 (s, 1H), 8.18 (s, 1H), 7.92 (d, J = 8.7 Hz, 1H), 7.91 (s, 1H), 7.86 (s, 1H), 7.42 (d, J
=
8.5 Hz, 1H), 7.29 (d, J = 9.5 Hz, 1H), 7.10 (t, J = 6.8 Hz, 1H), 6.88 (dd, J1 = 8.8 Hz, J2 = 4.7, 1H);
+
MS (ESI) m/z calcd for [C H FN O + H] 280.1, found 280.0; HPLC A: 95A% at 214nm. The minor
16
10
3
fraction was also isolated to give predominantly the Z isomer, 21 as a yellow solid; a trace of the E
1
isomer was evident (6.1 mg, 8.0 %). H NMR (400 MHz, CD OD) δ 8.94 (s, 1H), 8.07 (s, 1H), 7.94ꢀ
3
7.85 (m, 2H), 7.80 (d, J = 8.6 Hz, 1H), 7.49 (d, J = 8.8 Hz, 1H), 6.97 (t, J = 9.2 Hz, 1H), 6.87ꢀ6.81 (m,
+
1
H); MS (ESI) calcd for [C H FN O+ H] 280.09, found 280.0; HPLC A: >95A% at 214 nm.
1
6
10
3
Alternatively, compound 20 was prepared by heating a solution of 5ꢀfluoroindolinꢀ2ꢀone (0.60g,
4.0 mmol), 1Hꢀindazoleꢀ6ꢀcarbaldehyde (0.58 g, 4.0 mmol) and piperidine (0.04 mL, 0.4 mmol) in
o
MeOH (50 mL) at 60 C overnight. The precipitate obtained upon cooling to rt was collected by
filtration, rinsed with excess MeOH and dried to yield a yellow solid, (0.66 g, 58 %). The material was
+
identical to that described above for the E isomer. HRMS (ESI) m/z calcd for [C H N O + H]
1
6
11
3
2
80.0886, found 280.0889; HPLC A: 96A% at 254 nm. Incubation of compound 20 (0.2 mM) under
ACS Paragon Plus Environment
19

Page 21 of 60
Journal of Medicinal Chemistry
PLK4 assay conditions (2% DMSO, 50 mM Hepes pH 7.5; 1 mM EGTA, 10 mM MgCl , 0.01 % Brijꢀ
2
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
3
5, 0.1 mM ATP) at 29 C for 4 h was monitored: 96A% at 254 nm.
The filtrate was concentrated and triturated (Et O) multiple times with sonication to afford an orange
2
solid (0.37 g, 33 %) as a 3:2 mixture of 20 and 21 (peaks 1 and 2 respectively by HPLC 98A% at 254
0
1
2
3
4
5
6
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
+
nm). For peak 2, HRMS (ESI) m/z calcd for [C H N O + H] 280.0886, found 280.0875. EXSIDE
16
11
3
NMR experiment on the 3:2 mixture (125.66 MHz, 499.67 MHz, DMSOꢀd ): Compound 20: δ 7.87,
6
3
3
1
1
69.08 (d, J(C=O, H) = 6.4 Hz); HPLC Rt: 1.58 min. Compound 21: δ 8.04, 167.69 (d, J(C=O, H)
3.2 Hz); HPLC Rt: 1.73 min.
(E)-3-((1H-indazol-6-yl)methylene)-5-chloroindolin-2-one (22). According to the method for
compound 15, using 5ꢀchlorooxindole (21 mg, 0.13 mmol), the title compound was obtained as a yellow
1
solid (6.3 mg, 29 %). H NMR (400 MHz, CD OD) δ 8.17 (s, 1H), 7.96 (s, 1H), 7.94 (d, J = 8.0 Hz,
3
1H), 7.88 (s, 1H), 7.59 (d, J = 4.0 Hz, 1H), 7.45 (d, J = 8.0 Hz, 1H), 7.25 (dd, J = 8.0, 4.0 Hz, 1H), 6.92
+
(
d, J = 4.0 Hz, 1H); MS (ESI) calcd for [C H ClN O + H] 296.1, found 296.0; HPLC B: > 99A% at
1
6
10
3
2
14 nm.
(E)-3-((1H-indazol-6-yl)methylene)-6-chloroindolin-2-one (23). According to the method for
1
compound 20, the title compound was obtained as a yellow solid (22 mg, 42 %). H NMR (400 MHz,
DMSOꢀd ) δ 13.3 (s, 1H), 10.8 (s, 1H), 8.16 (s, 1H), 7.90ꢀ7.87 (m, 2H), 7.83 (s, 1H), 7.56 (d, J = 8.0
6
+
Hz, 1H), 7.41 (d, J = 8.8 Hz, 1H), 6.93ꢀ6.88 (m, 2H); MS (ESI) calcd for [C H ClN O + H] 296.1,
16
10
3
found 296.0; HPLC A: 95A% at 214 nm.
(E and Z)-3-((1H-indazol-6-yl)methylene)-5-methoxyindolin-2-one (24). According to the method
for compound 15, except substituting 5ꢀmethoxyoxindole (22 mg, 0.14 mmol), the title compound was
obtained as a yellow solid (12.3 mg, 31 %), determined to be a mixture of E and Z isomers (8.5:1.5) by
1
NMR. H NMR for the major isomer (400 MHz, CD OD) δ 8.14 (s, 1H), 7.95ꢀ7.87 (m, 3H), 7.46 (d, J
3
=
8.8 Hz, 1H), 7.23 (s, 1H), 6.84 (s, 2H), 3.62 (s, 3H); a signal at δ 9.0 (s, 1H) indicated the presence of
ACS Paragon Plus Environment
20

Journal of Medicinal Chemistry
Page 22 of 60
+
the Z isomer. HRMS (ESI) m/z calcd for [C H N O ] 291.1008, found 291.1009; HPLC B: 98A% at
17
13
3
2
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
2
14 nm.
Synthesis of (E)-3-((1H-indazol-6-yl)methylene)-N-methyl-2-oxoindoline-5-carboxamide (25). Nꢀ
methylꢀ2ꢀoxoindolineꢀ5ꢀcarboxamide. An ovenꢀdried roundꢀbottom flask was charged with 2ꢀ
0
1
2
3
4
5
6
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
oxoindolineꢀ5ꢀcarboxylic acid (89 mg, 0.50 mmol), TBTU (193 mg, 0.60 mmol), HOBt hydrate (81 mg,
0.60 mmol), DIPEA (0.13 mL, 0.75 mmol) and DMF (3.0 mL) and stirred for 5 min. MeNH (0.75 mL
2
of a 2.0 M solution in THF) was then added and reaction was stirred for 4 h. Saturated NaHCO3
solution (2 mL) was added, solvents were removed in vacuo and the residue was purified by column
chromatography (SiO , 92:8 CH Cl /MeOH) to afford Nꢀmethylꢀ2ꢀoxoindolineꢀ5ꢀcarboxamide (11 mg,
2
2
2
1
1
2 %). H NMR (400 MHz, CD OD)
δ
ppm 7.69 ꢀ 7.75 (m, 2 H), 6.94 (d, J=8.1 Hz, 1 H), 3.58 (s, 2 H),
3
+
2.97 (s, 3 H); MS (ESI) m/z calcd for [C H N O + H] 191.1, found 191.0.
1
0
10
2
2
(E)ꢀ3ꢀ((1Hꢀindazolꢀ6ꢀyl)methylene)ꢀNꢀmethylꢀ2ꢀoxoindolineꢀ5ꢀcarboxamide (25). According to the
method for compound 20, using Nꢀmethylꢀ2ꢀoxoindolineꢀ5ꢀcarboxamide (11 mg, 0.0578 mmol) and 1Hꢀ
indazoleꢀ6ꢀcarbaldehyde (10 mg, 0.0636 mmol), the title compound was obtained as an orange solid
1
(3.0 mg, 17 %). H NMR (400 MHz, DMSOꢀd ) δ 13.33 (s, 1H), 10.88 (s, 1H), 8.24ꢀ8.22 (m, 1H), 8.18
6
(s, 2H), 7.93ꢀ7.88 (m, 2H), 7.84 (s, 1H), 7.72 (d, J = 7.9 Hz, 1H), 7.46 (d, J = 8.0 Hz, 1H), 6.91 (d, J =
+
8.0 Hz, 1H), 2.68 (d, J = 4.0 Hz, 3H); MS (ESI) m/z calcd for [C H N O + H] 319.1, found 319.1;
18
14
4
2
HPLC A: 98A% at 214 nm.
(E)-N-(3-((1H-indazol-6-yl)methylene)-2-oxoindolin-5-yl)acetamide (26). A solution of Nꢀ(2ꢀ
oxoindolinꢀ5ꢀyl)acetamide (19 mg, 0.1 mmol) was treated with 1Hꢀindazoleꢀ6ꢀcarbaldehyde (16 mg,
o
0
.12 mmol), piperidine (2 uL, 0.011 mmol) and MeOH (2 mL). The reaction was then heated to 85 C
for 30 min. The MeOH was removed in vacuo and the solid triturated with ether to give 11 mg, 34 % of
1
the title compound as an orange solid with a 13:1 E to Z ratio. H NMR (400 MHz, CD OD) δ 8.13 (s,
3
1
H), 8.06 (s, 1H), 7.92ꢀ7.88 (m, 2H), 7.84 (s, 1H), 7.49 (d, J = 8.6 Hz, 1H), 7.35 (d, 1H), 6.88 (d, J =
ACS Paragon Plus Environment
21

Page 23 of 60
Journal of Medicinal Chemistry
+
8
9
.3 Hz, 1H), 2.04 (s, 3H); MS (ESI) m/z calcd for [C H N O S + H] 319.1 found 319.1. HPLC B:
18 17 5 3
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
7A% at 214 nm.
(E)-N-(3-((1H-indazol-6-yl)methylene)-2-oxoindolin-6-yl)acetamide (27). According to the method
for compound 20 except substituting Nꢀ(2ꢀoxoindolinꢀ6ꢀyl)acetamide (25 mg, 0.131 mmol), the title
0
1
2
3
4
5
6
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
compound was obtained (12 mg, 29 %). H NMR (400 MHz, DMSOꢀd ) δ 13.25 (s, 1H), 10.59 (s, 1H),
6
10.08 (s, 1H), 8.15 (s, 1H), 7.88 (s, 2H), 7.62 (s, 1H), 7.54 (d, J = 6.9 Hz, 1H), 7.50 (s, 1H), 7.41 (d, J =
+
7
.9 Hz, 1H), 6.87 (d, J = 7.0 Hz, 1H), 2.04 (s, 3H); MS (ESI) m/z calcd for [C H N O + H] 319.1,
1
8
14
4
2
found 319.1; HPLC A: 95A% at 214nm.
Synthesis of (E)-3-((3-((E)-2-(pyridin-2-yl)vinyl)-1H-indazol-6-yl)methylene)indolin-2-one (37).
ꢀiodoꢀ(1 and 2)ꢀ(2ꢀ(trimethylsilyl)ethyl)ꢀ2Hꢀindazoleꢀ6ꢀcarbaldehyde (29). To a stirred suspension of
ꢀiodoꢀ1Hꢀindazoleꢀ6ꢀcarbaldehyde 28 (3.01 g, 11.1 mmol) and TBAB (36 mg, 0.11 mmol) in CH Cl
3
3
2
2
(70 mL) and 50 % aq. KOH (20 mL) was added, dropwise, (2ꢀ(chloromethoxy)ethyl)trimethylsilane
o
o
(
2.3 mL, 13.3 mmol) at 0 C. The stirring was continued at 0 C for 3 h. The solution was then diluted
with CH Cl (200 mL), the layers were separated and the organic phase was washed with brine (2 x 100
2
2
mL), dried (MgSO ) and concentrated under reduced pressure. The resulting residue was purified by
4
column chromatography (SiO , 100 % CH Cl ) to give of the Nꢀ1 isomer (higher eluting spot, 2.88 g, 65
2
2
2
%) and the Nꢀ2 isomer (lower eluting spot, 757 mg, 17 %). Typically, the crude regioisomeric mixture
was purified by chromatography and used in subsequent reactions without separation into its constituent
1
isomers. H NMR Nꢀ1 isomer: (400 MHz, CDCl ) δ 10.18 (s, 1H), 8.11 (s, 1H), 7.81 (d, J = 8.4 Hz,
3
1H), 7.65 (d, J = 8.4 Hz, 1H), 5.82 (s, 2H), 3.60 (m, 2H), 0.91 (m, 2H), ꢀ0.042 (s, 9H); MS ESI 425.0
+
+
[
M + Na] , calcd for [C H IN O Si + Na] 425.02. Nꢀ2 isomer : (400 MHz, CD OD) 10.09 (s, 1H),
14 19 2 2 3
8
.31 (s, 1H), 7.62 (m, 2H), 5.91 (s, 2H), 3.71 (m, 2H), 0.92 (m, 2H), ꢀ0.039 (s, 9H); MS (ESI) m/z calcd
+
for [C H IN O Si + Na] , 425.0, found 425.0.
14
19
2
2
(E)ꢀ3ꢀ(2ꢀ(pyridinꢀ2ꢀyl)vinyl)ꢀ1Hꢀindazoleꢀ6ꢀcarbaldehyde. A 4:1 mixture of 3ꢀiodoꢀ(1 and 2)ꢀ((2ꢀ
(
trimethylsilyl)ethoxy)methyl)ꢀ1Hꢀindazoleꢀ6ꢀcarbaldehyde (0.42 g, 1.0 mmol, 29), 2ꢀvinylpyridine
ACS Paragon Plus Environment
22

Journal of Medicinal Chemistry
Page 24 of 60
(
0.143 g, 1.3 mmol), DIPEA (0.36 mL, 2.1 mmol) in DMF (3 mL) was purged with a slow stream of Ar
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
for 20 min. Pd(OAc) (23 mg, 0.10 mmol) and P(oꢀtol) (95 mg, 0.31 mmol) were added and the
2
3
o
reaction was heated to 100 C for 5.5 h. The reaction was cooled to rt, concentrated under reduced
pressure and purified by column chromatography (SiO , 0ꢀ2 % MeOH in CH Cl ) to give the title
2
2
2
compound as a 4:1 mixture of (1:4) as light yellow solid (4:1 mixture of regioisomers) which was used
0
1
2
3
4
5
6
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
+
+
+
without further purification. MS ESI 380.2 [M + H] , calcd for [C H N O Si H] 380.2. This
2
1
25
3
2
o
material was heated overnight with TBAF (1.0 M in THF, 4.5 mL, 4.5 mmol) at 60 C under Argon in a
sealed vial. The reaction mixture was concentrated under reduced pressure and purified by flash
chromatography (SiO , 0ꢀ4 % MeOH in DCM) to afford (E)ꢀ3ꢀ(2ꢀ(pyridinꢀ2ꢀyl)vinyl)ꢀ1Hꢀindazoleꢀ6ꢀ
2
1
carbaldehyde (98 mg, 44 %); H NMR (400 MHz, CD OD)
δ ppm 10.1 (s, 1 H), 8.54 (m, J = 4.8 Hz, 1
3
H), 8.19 (d, J = 8.5 Hz, 1 H), 8.08 (s, 1 H), 7.86 (d, J = 16.6 Hz, 1 H), 7.78ꢀ7.83 (m, 1 H), 7.76 (d, J =
.5 Hz, 1 H), 7.64 (d, J = 7.3 Hz, 1 H), 7.58 (d, J = 16.6 Hz, 1 H), 7.21ꢀ7.30 (m, 1 H); MS (ESI) m/z
calcd for [C H N O+ H]+ 250.1, found250.1.
8
15
11
3
(E)ꢀ3ꢀ((3ꢀ((E)ꢀ2ꢀ(pyridinꢀ2ꢀyl)vinyl)ꢀ1Hꢀindazolꢀ6ꢀyl)methylene)indolinꢀ2ꢀone. An EtOH (10 mL)
solution of (E)ꢀ3ꢀ(2ꢀ(pyridinꢀ2ꢀyl)vinyl)ꢀ1Hꢀindazoleꢀ6ꢀcarbaldehyde (98 mg, 0.39 mmol), indolinꢀ2ꢀ
o
one (52 mg, 0.39 mmol) and piperidine (3 drops) was heated in a sealed vial under Ar at 70 C
overnight. Upon cooling to rt, the precipitated solid was collected, rinsed with EtOH then triturated in
sequence with MeCN, DCM, MeOHꢀDCM (1:1 v/v), MeOH, PhMe, hexanes, Et O and DCM to
2
1
provide the title compound as a yellow solid (62.9 mg, 44 %). H NMR (400 MHz, DMSOꢀd ) δ ppm
6
1
3.52 (bs., 1 H), 10.65 (s, 1 H), 8.62 (d, J = 4.3 Hz, 1 H), 8.34 (d, J = 8.5 Hz, 1 H), 8.00 (d, J = 16.3
Hz, 1 H), 7.93 (s, 1 H), 7.84 (d, J = 7.8 Hz, 1 H), 7.81 (s, 1 H), 7.69 (d, J = 7.8 Hz, 1 H), 7.64 (d, J =
8.0 Hz, 1 H), 7.62 (s, 1 H), 7.55 (d, J = 7.8 Hz, 1 H), 7.21 ꢀ 7.31 (m, 2 H), 6.83 ꢀ 6.92 (m, 2 H); HRMS
(EI) m/z calcd for C H N O (M+) 364.1324, found 364.1323. HPLC A: 95A% at 254 nm.
23 16 4
Synthesis of (E)-3-((3-((E)-2-(pyridin-3-yl)vinyl)-1H-indazol-6-yl)methylene)-indolin-2-one (38).
(E)ꢀ3ꢀ(2ꢀ(pyridinꢀ3ꢀyl)vinyl)ꢀ1Hꢀindazoleꢀ6ꢀcarbonitrile. A suspension of 3ꢀ(chloromethyl)pyridine
ACS Paragon Plus Environment
23

Page 25 of 60
Journal of Medicinal Chemistry
hydrochloride (6.54 g, 40 mmol) in benzene (75 mL) was treated with 40% NaOH (2.7 mL); the
resulting mixture was sonicated for 10 min and filtered. The residue was treated with additional benzene
(25 ml), sonicated and filtered; the combined benzene layers were dried (Na SO ) to give a solution of
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
2
4
3
ꢀ(chloromethyl)pyridine in benzene. The solution was added drop wise (over 15 min) to a solution of
sodium diethyl phosphate (prepared from diethyl phosphate (6.06 g, 44 mmol) in benzene (50 mL) and
0
1
2
3
4
5
6
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
o
1
.02 g (44 mmol) of freshly cut Na), refluxed (oil temp. 95 C) for 30 min then cooled to 0 C. The
o
resulting mixture was refluxed for 2 h (oil temp. 100 C) until LCꢀMS indicated the completion of
reaction. After cooling to rt, the resulting white precipitate (NaCl) was removed by filtration and rinsed
with benzene (50 mL). The filtrate was concentrated and dried under high vacuum to give 6.30 g of
diethyl pyridinꢀ3ꢀylmethylphosphonate as a light yellow liquid. The liquid was dissolved in DMF (50
o
t
mL), cooled to 0 C and treated with BuOK (6.72 g, 60 mmol) portion wise over 3 min; the reaction
o
turned a dark reddish brown. After stirring for 3 min at 0 C, a solution of 3ꢀformylꢀ1Hꢀindazoleꢀ6ꢀ
5
4
carbonitrile (34, 3.42 g, 20 mmol) in DMF (25 mL) was added drop wise by pipette over 5 min. The
o
resulting mixture was stirred for 1 h at 0 C before quenching with ice (100 mL). The reaction mixture
o
was cooled to 0 C and slowly acidified with 2M HCl to pH 5. During this addition, a copious amount
of precipitate was formed. After stirring for 2 min at this temperature, sat NaHCO was slowly added to
3
attain a pH of 8; the mixture was stirred for additional 2 min. Water was added until the total volume
reached 600 mL. After stirring for 10 min, the resulting precipitate was collected, rinsed thoroughly
1
with water, then dried under high vacuum to give the title compound (3.30 g, 67%) as a beige solid. H
NMR (400 MHz, DMSOꢀd ) δ 13.80 (s, 1H, NH), 8.89 (s, 1H), 8.48 (d, J = 4.4 Hz, 1H), 4.42 (d, J =
6
8
.4 Hz, 1H), 8.22ꢀ8.17 (m, 2H), 7.74 (d, J = 16.8 Hz, 1H), 7.59 (d, J = 18 Hz, 1H, partially overlapping
with the doublet at 7.55 ppm), 7.55 (d, J = 10 Hz, 1H, partially overlapping with the doublet at 7.59
+
ppm), 7.43 (dd, J = 8.0 Hz, 5.6 Hz, 1H); MS (ESI) m/z calcd for [C H N + H] 247.1, found 247.0.
1
5
10
4
(E)ꢀ3ꢀ(2ꢀ(pyridinꢀ3ꢀyl)vinyl)ꢀ1Hꢀindazoleꢀ6ꢀcarbaldehyde. A suspension of (E)ꢀ3ꢀ(2ꢀ(pyridinꢀ4ꢀ
yl)vinyl)ꢀ1Hꢀindazoleꢀ6ꢀcarbonitrile (984 mg, 3 mmol) pyridine (30 mL) was treated with HOAc (8
ACS Paragon Plus Environment
24

Journal of Medicinal Chemistry
Page 26 of 60
mL) and DMF (30 mL). The resulting mixture was heated and sonicated to make a clear solution. After
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
cooling to 0 C, a solution of sodium hypophosphite (1.41 g, 16 mmol) in H O (8 mL) was added,
2
o
followed by RaneyꢀNickel 2400 (slurry in H O, 0.8 mL). The resulting mixture was heated at 60 C (oil
2
temp.) for 1 h before cooling to rt. Water (50 mL) was added and the mixture was extracted with EtOAc
(
100 mL + 50 mL x 2). The combined organic extracts were dried (Na SO ). And concentrated to give a
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
2 4
yellow solution in DMF (about 30 mL); H O (500 mL) was added with swirling and a yellow precipitate
2
formed. After standing 10 min, the resulting precipitate was collected by suction filtration, rinsed with
1
H O and dried under high vacuum to give the title compound (500 mg, 50 %) as a yellow solid. H
2
NMR (400 MHz, DMSOꢀd ) δ 13.79 (s, 1H, NH), 10.14 (s, 1H, CHO), 8.48 (d, J = 4.8 Hz, 1H), 8.39
6
(
d, J = 8.0 Hz, 1H), 8.19 (d, J = 9.2 Hz, 1H, partially overlapping with the singlet at 8.18 ppm), 8.18 (s,
1H, partially overlapping with the doublet at 8.19 ppm), 7.75 (d, J = 16.4 Hz, 1H), 7.69 (d, J = 8.4 Hz,
H), 7.59 (d, J = 16.8 Hz, 1H), 7.43 (dd, J = 8.0 Hz, 5.2 Hz, 1H); MS (ESI) m/z calcd for [C H N O +
1
1
5
11
3
+
H] 250.1, found 250.0.
(E)ꢀ3ꢀ((3ꢀ((E)ꢀ2ꢀ(pyridinꢀ3ꢀyl)vinyl)ꢀ1Hꢀindazolꢀ6ꢀyl)methylene)ꢀindolinꢀ2ꢀone (38). A mixture of
oxindole (26.6 mg, 0.2 mmol) and (E)ꢀ3ꢀ(2ꢀ(pyridinꢀ3ꢀyl)vinyl)ꢀ1Hꢀindazoleꢀ6ꢀcarbaldehyde (49.8 mg,
0
.2 mmol) was treated with a drop of piperidine. The resulting mixture was refluxed for 2 h (oil temp.
o
70 C) and then cooled to rt. The resulting precipitate was collected by suction filtration to give the title
1
compound as a yellow solid (61 mg, 84 %). H NMR (400 MHz, DMSOꢀd ) δ 13.46 (s, 1H, NH), 10.64
6
(s, 1H, NH), 8.91 (s, 1H), 8.48 (d, J = 4.4 Hz, 1H), 8.36 (d, J = 8.4 Hz, 1H), 8.20 (d, J = 8.4 Hz, 1H),
7
.90 (s, 1H), 7.80 (s, 1H), 7.73 (d, J = 17 Hz, 1H), 7.61 (d, J = 7.2 Hz, 1H, partially overlapping with
the peak at 7.60 ppm), 7.60 (d, J = 17.2 Hz, 1H, partially overlapping with the peak at 7.61 ppm), 7.54
(d, J = 8.4 Hz, 1H), 7.43 (dd, J = 7.8 Hz, 4.8 Hz, 1H), 7.24 (t, J = 7.8 Hz, 1H), 6.89 (d, J = 7.6 Hz,
1
H), 6.85 (t, J = 7.8 Hz, 1H); HRMS (ESI) m/z calcd for C H N O (M+) 365.1402, found 365.1399.
23 16 4
HPLC B: 99A% at 214 nm
ACS Paragon Plus Environment
25

Page 27 of 60
Journal of Medicinal Chemistry
Synthesis of (E)-5-ethyl-3-((3-((E)-2-(pyridin-3-yl)vinyl)-1H-indazol-6-yl)methylene)indolin-2-
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
one (39). According to the method for compound 38, part 3 (oil temp 70 C, reflux 2 h), using 5ꢀ
ethylindolinꢀ2ꢀone (77 mg, 0.48 mmol) and (E)ꢀ3ꢀ(2ꢀ(pyridinꢀ3ꢀyl)vinyl)ꢀ1Hꢀindazoleꢀ6ꢀcarbaldehyde
(
100 mg, 0.4 mmol) the title compound (35 mg, 67 %) was synthesized as yellow solid (only E isomer
1
1
was detected by H NMR). H NMR (400 MHz, DMSOꢀd ) δ 13.50 (bs, 1H, NH), 10.55 (s, 1H, NH),
0
1
2
3
4
5
6
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
6
8
.91 (d, J = 1.6 Hz, 1H), 8.48 (dd, J = 4.8, 1.2 Hz, 1H), 8.35 (d, J = 8.4 Hz, 1H), 8.20 (d, J = 8.0 Hz,
H), 7.91 (s, 1H), 7.76 (s, 1H, overlapping with peak at 7.74), 7.74 (d, J = 15.2 Hz, 1H, overlapping
1
with peak at 7.76), 7.59 (d, J = 16.8 Hz, 1H), 7.55 (d, J = 8.4 Hz, 1H), ), 7.44 (s, 1H), 7.43 (dd, J = 7.8,
4
.2 Hz, 1H), 7.09 (d, J = 7.2 Hz, 1H), 6.81 (d, J = 8.0 Hz, 1H), 2.43 (q, J = 7.6 Hz, 2H), 1.06 (t, J = 7.6
+
Hz, 3H); MS (ESI) m/z calcd for [C H N O + H] 393.2, found 393.1; HPLC B: 95 % at 254 nm.
2
5
20
4
(E and Z)-5-methoxy-3-((3-((E)-2-(pyridin-3-yl)vinyl)-1H-indazol-6-yl)methylene)indolin-2-one
o
(
40). According to the method for compound 39, part 3 (90 min at 75 C) using 5ꢀmethyoxyoxindole
(
32.6 mg, 0.2 mmol) and (E)ꢀ3ꢀ(2ꢀ(pyridinꢀ3ꢀyl)vinyl)ꢀ1Hꢀindazoleꢀ6ꢀcarbaldehyde (49.8 mg, 0.2
mmol), the title compound was synthesized as a 4:1 mixture of E/Z isomers (orange solid, 53 mg, 67
1
%
). For the E isomer: H NMR (400 MHz, DMSOꢀd ) δ 13.50 (bs, 1H, NH), 10.44 (s, 1H, NH), 8.91
6
(
d, 1H), 8.48 (d, J = 4.8 Hz, 1H), 8.36 (d, J = 8.4 Hz, 1H), 8.20 (d, J = 7.6 Hz, 1H), 7.92 (s, 1H), 7.79
(s, 1H), 7.73 (d, J = 16.8 Hz, 1H), 7.59 (d, J = 16.4 Hz, 1H), 7.54 (d, J = 8.8 Hz, 1H), 7.43 (dd, J = 8.0
Hz, 5.2 Hz, 1H), 7.19 (d, 1H), 6.86 (dd, J = 8.4 Hz, 2.0 Hz, 1H), 6.81 (d, J = 8.4 Hz, 1H), 3.78 (s, 3H);
A diagnostic signal at δ 9.01 (s, 1H) indicated the presence of the Z isomer. HRMS (ESI) m/z calcd for
C H N O (M+) 394.1430, found 394.1432; HPLC B: 99 % at 214 nm.
24
18
4
2
Synthesis of (E)-3-((3-((E)-2-(pyridin-4-yl)vinyl)-1H-indazol-6-yl)methylene)-indolin-2-one (41).
E)ꢀ3ꢀ(2ꢀ(pyridinꢀ4ꢀyl)vinyl)ꢀ1ꢀ((2ꢀ(trimethylsilyl)ethoxy)methyl)ꢀ1Hꢀindazoleꢀ6ꢀcarbaldehyde. A soluꢀ
(
tion of 3ꢀiodoꢀ1ꢀ((2ꢀ(trimethylsilyl)ethoxy)methyl)ꢀ1Hꢀindazoleꢀ6ꢀcarbaldehyde (2.0 g, 4.98 mmol), 4ꢀ
vinylpyridine (0.69 mL, 6.47 mmol), DIPEA (1.74 mL, 9.96 mmol) and DMF (50 mL) was purged with
Ar gas for 20 min at which time Pd(OAc) (112 mg, 0.500 mmol) and P(oꢀtol) (457 mg, 1.50 mmol)
2
3
ACS Paragon Plus Environment
26

Journal of Medicinal Chemistry
Page 28 of 60
o
were added and the reaction heated to 100 C for 20 h. The reaction was cooled to room temperature
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
and partitioned between EtOAc and NH Cl. The aqueous layer was extracted with EtOAc (3X) and the
4
combined organics washed with water (2x) and brine (2c). The organic layer was dried over MgSO4
and the EtOAc removed in vacuo. The residue was purified by column chromatography (silica gel, 97:3
CH Cl / MeOH) to give the title compound (0.93g, 49 %) which was used without further purification.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
2
2
+
MS (ESI) m/z calcd for [C H N O Si +H] 380.2, found 380.2.
21
25
3
2
(E)ꢀ3ꢀ((3ꢀ((E)ꢀ2ꢀ(pyridinꢀ4ꢀyl)vinyl)ꢀ1ꢀ((2ꢀ(trimethylsilyl)ethoxy)methyl)ꢀ1Hꢀindazolꢀ6ꢀyl)methylene)ꢀ
indolinꢀ2ꢀone. A solution of (E)ꢀ3ꢀ(2ꢀ(pyridinꢀ4ꢀyl)vinyl)ꢀ1ꢀ((2ꢀ(trimethylsilyl)ethoxy)methyl)ꢀ1Hꢀ
indazoleꢀ6ꢀcarbaldehyde (927 mg, 2.44 mmol), oxindole (325 mg, 2.44 mmol) and MeOH (20 mL) was
o
treated with piperidine (24 uL, 0.244 mmol) and heated to 60 C for 4 h. A yellow precipitate formed
which was filtered and washed with MeOH to give the title compound (960 mg, 79 %) as a yellow solid.
+
MS (ESI) m/z calcd for [C H N O Si +H] 495.2, found 495.3.
29
30
4
2
(E)ꢀ3ꢀ((3ꢀ((E)ꢀ2ꢀ(pyridinꢀ4ꢀyl)vinyl)ꢀ1Hꢀindazolꢀ6ꢀyl)methylene)ꢀindolinꢀ2ꢀone (41). A dry round
bottom was charged with (E)ꢀ3ꢀ((3ꢀ((E)ꢀ2ꢀ(pyridinꢀ4ꢀyl)vinyl)ꢀ1ꢀ((2ꢀ(trimethylsilyl)ethoxy)methyl)ꢀ1Hꢀ
indazolꢀ6ꢀyl)methylene)indolinꢀ2ꢀone (906 mg, 1.83 mmol), and CH Cl (100 mL) under an atmosphere
2
2
of N . Boron trifluoride etherate (2.3 mL, 18.3 mmol) was added dropwise and the reaction was stirred
2
for 4 h at which time LCꢀMS indicated complete conversion to the partially deꢀprotected material (E)ꢀ3ꢀ
((1ꢀ(hydroxymethyl)ꢀ3ꢀ((E)ꢀ2ꢀ(pyridinꢀ4ꢀyl)vinyl)ꢀ1Hꢀindazolꢀ6ꢀyl)methylene)ꢀindolinꢀ2ꢀone.
The
CH Cl was removed in vacuo and 150 mL of a 2:1 mixture of EtOH / 2M HCl was added to the
2
2
o
residue. The resulting solution was heated to 50 C for 3 h at which time LCꢀMS indicated complete
conversion to product. The reaction was cooled with an iceꢀbath and neutralized with NH OH to pH~8;
4
the EtOH was removed to yield a suspension. The solid was collected by filtration, washed with water,
dried and triturated several times with 90:10 CH Cl /MeOH. Drying under vacuum yielded the title
2
2
1
compound as an orange solid (353 mg, 53 %). H NMR (400 MHz, DMSOꢀd ) δ 13.57 (s, 1H), 10.64 (s,
6
1
H), 8.57 (d, J = 5.2 Hz, 2H), 8.38 (d, J = 8.1 Hz, 1H), 7.92 (s, 1H), 7.87 (d, J = 16.7 Hz, 1H), 7.80 (s,
ACS Paragon Plus Environment
27

Page 29 of 60
Journal of Medicinal Chemistry
1
H), 7.71 (d, J = 5.2 Hz, 2H), 7.53ꢀ7.62 (m, 3H), 7.23 (t, J = 8.3 Hz, 1H), 6.82ꢀ6.91 (m, 2H); HRMS
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
+
(ESI) m/z calcd for [C H N O + H] 364.1324, found 364.1320; HPLC A: 95A% at 214 nm.
2
3
16
4
(E)-5-chloro-3-((3-((E)-2-(pyridin-4-yl)vinyl)-1H-indazol-6-yl)methylene)indolin-2-one.
(42).
According to the method for compound 41 using 5ꢀchlorooxindole (14 mg, 0.08 mmol), the title
0
1
2
3
4
5
6
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
compound was prepared as an orange solid (13.5 mg, 42 %). H NMR (400 MHz, DMSOꢀd ) δ 13.63
6
(bs, 1H), 10.78 (bs, 1H), 8.56 (d, J = 4.0 Hz, 2H), 8.40 ( d, J = 10 Hz, 1H), 7.91 (s, 1H), 7.89 (d, J = 13
Hz, 2 H), 7.72 (d, J = 3.5 Hz, 2H), 7.57 (d, J = 12 Hz, 1H), 7.52 (d, J = 12 Hz, 2H), 7.30 (d, J = 10.0
+
Hz, 1H), 6.90 (d, J = 11.5 Hz, 1H); MS (ESI) m/z calcd for [C H ClN O + H] 399.1, found 399.2;
23 15
4
HPLC B: 99A% at 214 nm.
(E)-5-methyl-3-((3-((E)-2-(pyridin-4-yl)vinyl)-1H-indazol-6-yl)-methylene)indolin-2-one
(43).
According to the method for compound 41, using 5ꢀmethyloxindole (9 mg, 0.060 mmol) the title
1
compound was prepared as a yellow powder (5.9 mg, 29 %). H NMR (400 MHz, DMSOꢀd ) δ 13.59
6
(bs, 1H), 10.54 (bs, 1H), 8.57 (bs, 2H), 8.3ꢀ8.4 (m, 1H), 7.95 – 7.32 (m, 8H), 7.12 ꢀ7.01 (m, 1 H), 6.73ꢀ
+
6
.84 (m, 1H), 2.16 (s, 3H); MS (ESI) m/z calcd for [C H N O + H] 379.4, found 379.2; HPLC B:
24 18
4
>
99A% at 214 nm.
(E)-5-methoxy-3-((3-((E)-2-(pyridin-4-yl)vinyl)-1H-indazol-6-yl)methylene)indolin-2-one (44).
According to the method for compound 41, using 5ꢀmethoxyoxindole (11 mg, 0.068 mmol) the title
1
compound was obtained as an orange solid (13.6 mg, 51 %). H NMR (400 MHz, DMSOꢀd ) δ 13.58
6
(bs, 1H), 10.45 (bs, 1H), 8.57 (d, J = 6.0 Hz, 2H), 8.37 ( d, J = 8.5 Hz, 1H), 7.93 (s, 1H), 7.87 (d, J =
16.6 Hz, 1 H), 7.79 (s, 1H), 7.71 (d, J = 6.0 Hz, 2H), 7.57 (d, J = 2.3 Hz, 1H), 7.54 (d, J = 6.3 Hz, 2H),
7
.18 (d, J = 10.0 Hz, 1H), 6.86 (dd, J = 8.8, 2.5 Hz, 1H) 6.80 (d, J = 8.3 Hz, 1H), 3.6 (s, 3H); MS (ESI)
+
m/z calcd for [C H N O + H] 395.1, found 395.2; HPLC B: >99A% at 214 nm.
24 18
4
2
(E)-5-hydroxy-3-((3-((E)-2-(pyridin-4-yl)vinyl)-1H-indazol-6-yl)methylene)indolin-2-one
(45)
The title compound was synthesized according to the method described for compound 41, parts 2 and 3
ACS Paragon Plus Environment
28