Discovery of 3-Benzyl-1-(trans-4-((5-cyanopyridin-2-yl)amino)cyclohexyl)-1-arylurea Derivatives as Novel and Selective Cyclin-Dependent Kinase 12 (CDK12) Inhibitors

Article abstract of DOI:10.1021/acs.jmedchem.8b00683

Cyclin-dependent kinase 12 (CDK12) plays a key role in the coordination of transcription with elongation and mRNA processing. CDK12 mutations found in tumors and CDK12 inhibition sensitize cancer cells to DNA-damaging reagents and DNA-repair inhibitors. This suggests that CDK12 inhibitors are potential therapeutics for cancer that may cause synthetic lethality. Here, we report the discovery of 3-benzyl-1-(trans-4-((5-cyanopyridin-2-yl)amino)cyclohexyl)-1-arylurea derivatives as novel and selective CDK12 inhibitors. Structure-activity relationship studies of a HTS hit, structure-based drug design, and conformation-oriented design using the Cambridge Structural Database afforded the optimized compound 2, which exhibited not only potent CDK12 (and CDK13) inhibitory activity and excellent selectivity but also good physicochemical properties. Furthermore, 2 inhibited the phosphorylation of Ser2 in the C-terminal domain of RNA polymerase II and induced growth inhibition in SK-BR-3 cells. Therefore, 2 represents an excellent chemical probe for functional studies of CDK12 and could be a promising lead compound for drug discovery.

Full text of DOI:10.1021/acs.jmedchem.8b00683

Subscriber access provided by READING UNIV
Article
Discovery of 3-benzyl-1-(trans-4-((5-cyanopyridin-2-
yl)amino)cyclohexyl)-1-arylurea derivatives as novel and
selective cyclin-dependent kinase 12 (CDK12) inhibitors
Masahiro Ito, Toshio Tanaka, Akinori Toita, Noriko Uchiyama, Hironori Kokubo, Nao
Morishita, Michael G Klein, Hua Zou, Morio Murakami, Mitsuyo Kondo, Tomoya
Sameshima, Shinsuke Araki, Satoshi Endo, Tomohiro Kawamoto, Gregg B. Morin,
Samuel A. Aparicio, Atsushi Nakanishi, Hironobu Maezaki, and Yasuhiro Imaeda
J. Med. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.jmedchem.8b00683 • Publication Date (Web): 01 Aug 2018
Downloaded from http://pubs.acs.org on August 1, 2018
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a service to the research community to expedite the dissemination
of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in
full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully
peer reviewed, but should not be considered the official version of record. They are citable by the
Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore,
the “Just Accepted” Web site may not include all articles that will be published in the journal. After
a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web
site and published as an ASAP article. Note that technical editing may introduce minor changes
to the manuscript text and/or graphics which could affect content, and all legal disclaimers and
ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or
consequences arising from the use of information contained in these “Just Accepted” manuscripts.
is published by the American Chemical Society. 1155 Sixteenth Street N.W.,
Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.

Page 1 of 74
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 2 of 74
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Discovery of
ꢀbenzylꢀ1ꢀ(transꢀ4ꢀ((5ꢀcyanopyridinꢀ2ꢀyl)amino)cy
3
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
clohexyl)ꢀ1ꢀarylurea derivatives as novel and
selective cyclinꢀdependent kinase 12 (CDK12)
inhibitors
,
†
†,┴
†
†
†,┴
Masahiro Ito,* Toshio Tanaka, Akinori Toita, Noriko Uchiyama, Hironori Kokubo, Nao
†
‡
‡
†
†,┴
Morishita, Michael G. Klein, Hua Zou, Morio Murakami, Mitsuyo Kondo, Tomoya
†
†
†,#
†,┴
§,@
Sameshima, Shinsuke Araki, Satoshi Endo, Tomohiro Kawamoto, Gregg B. Morin,
ǁ
†
†,┴
,†
Samuel A. Aparicio, Atsushi Nakanishi, Hironobu Maezaki, and Yasuhiro Imaeda*
†
Pharmaceutical Research Division, Takeda Pharmaceutical Company Limited, 26ꢀ1,
MuraokaꢀHigashi 2ꢀchome, Fujisawa, Kanagawa 251ꢀ8555, Japan
‡
Department of Structural Biology, Takeda California Inc., 10410 Science Center Dr., San
Diego, CA 92121, United States
§
Genome Sciences Centre, British Columbia Cancer Agency, 675 W. 10th Avenue, Vancouver,
BC, V5Z 1L3, Canada
@
Department of Medical Genetics, University of British Columbia, Vancouver, BC, V6H 3N1,
Canada
ACS Paragon Plus Environment
1

Page 3 of 74
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
ǁ
Department of Molecular Oncology, British Columbia Cancer Agency, 675 W. 10th Avenue,
Vancouver, BC, V5Z 1L3, Canada
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
ACS Paragon Plus Environment
2

Journal of Medicinal Chemistry
Page 4 of 74
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
Abstract:
Cyclinꢀdependent kinase 12 (CDK12) plays a key role in the coordination of transcription with
elongation and mRNA processing. CDK12 mutations found in tumors and CDK12 inhibition
sensitize cancer cells to DNAꢀdamaging reagents and DNAꢀrepair inhibitors. This suggests that
CDK12 inhibitors are potential therapeutics for cancer that may cause synthetic lethality. Here,
0
1
2
3
4
5
6
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
we
report
the
discovery
of
3
ꢀbenzylꢀ1ꢀ(transꢀ4ꢀ((5ꢀcyanopyridinꢀ2ꢀyl)amino)cyclohexyl)ꢀ1ꢀarylurea derivatives as novel
and selective CDK12 inhibitors. Structureꢀactivity relationship studies of a HTS hit,
structureꢀbased drug design, and conformationꢀoriented design using the Cambridge Structural
Database afforded the optimized compound 2, which exhibited not only potent CDK12 (and
CDK13) inhibitory activity and excellent selectivity, but also good physicochemical properties.
Furthermore, 2 inhibited the phosphorylation of Ser2 in the Cꢀterminal domain of RNA
polymerase II, and induced growth inhibition in SKꢀBRꢀ3 cells. Therefore, 2 represents an
excellent chemical probe for functional studies of CDK12 and could be a promising lead
compound for drug discovery.
Keywords: Cyclinꢀdependent kinase 12 (CDK12), CDK12 inhibitor, RNA polymerase II (Pol II),
Cꢀterminal domain (CTD) of Pol II, conformationꢀoriented design, Cambridge Structural
Database (CSD)
ACS Paragon Plus Environment
3

Page 5 of 74
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Introduction
In eukaryotes, three distinct RNA polymerases (I–III) coordinate transcription: RNA
polymerase I (Pol I) catalyzes the transcription of ribosomal RNA (rRNA) genes, RNA
polymerase II (Pol II) is responsible for the transcription of proteinꢀcoding genes into messenger
RNA (mRNA), and RNA polymerase III (Pol III) synthesizes small nonꢀcoding RNAs such as
transfer RNAs (tRNAs). Pol II is the most studied of these RNA polymerases. It is
wellꢀestablished that the modulation of Pol II activity is regulated by the phosphorylation of a
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
repeated motif (Y1ꢀS2ꢀP3ꢀT4ꢀS5ꢀP6ꢀS7) in the C terminal domain (CTD). The number of Pol II
CTD repeats vary from 26 in yeast to 52 in mammals, whereas this motif is completely absent in
Pol I and Pol III. The number of CTD repeats tends to increase depending on the complexity of
organisms. Truncation studies of Pol II CTD revealed a certain length of CTD repeats was
2
, 3
necessary for yeast cell viability or mammalian cell growth. The main function of the CTD is
to serve as a flexible binding scaffold for a variety of nuclear factors involved in the different
phases of transcription. The regulation of the phosphorylation pattern on the CTD of Pol II helps
to coordinate the assembly of complexes with transcription factors, chromatin modifiers, and
RNA processing enzymes. Although phosphorylation of the CTD can occur on tyrosine,
4
threonine, and serine residues, phosphorylation at position 2 serines (Ser2) and position 5
5
serines (Ser5) are far more abundant than at other sites. In general, Ser5 phosphorylation is
essential for transcriptional initiation, whereas the phosphorylation of Ser2 contributes to
transcriptional elongation and the recruitment of 3' RNAꢀprocessing factors, which play a major
role in transcription termination. Five cyclinꢀdependent kinases (CDKs: 7, 8, 9, 12, and 13) are
1
reported to phosphorylate the CTD of Pol II; thus, they are grouped together as “CTD kinases”.
All CTD kinases can phosphorylate Ser5, but only CDK9, CDK12, and CDK13 are responsible
ACS Paragon Plus Environment
4

Journal of Medicinal Chemistry
Page 6 of 74
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
6–8
for the phosphorylation of Ser2. CDK12 and its close homolog CDK13 are unusually large
proteins (1,490 and 1,512 amino acids, respectively, in humans) compared with other CDKs (e.g.,
9
–13
CDK9, 372 amino acids).
Both CDK12 and CDK13 contain a central kinase domain and a
1
4–16
0
1
2
3
4
5
6
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
Nꢀterminal arginineꢀserineꢀrich (RS) domain,
which are frequently found in proteins
involved in preꢀmRNA splicing and RNA processing. As suggested by their structural features,
CDK12 and CDK13 regulate coꢀtranscriptional processing, including splicing and 3'ꢀend RNA
8
,17–19
processing.
Recently, several differences between CDK12 and CDK13 were revealed by
7
genetic depletion studies. CDK12 is essential for the expression of longer genes with a high
2
0,21
number of exons which are involved in DNA damage response pathway.
On the other hand,
the loss of CDK13 does not affect the expression of such genes, but instead affect genes related
to protein translation. Several mutations that affect the function of CDK12 have been identified
in a variety of tumors. Consistent with the results of the genetic depletion experiments, these
mutations in CDK12 make tumors sensitive to cytostatic DNAꢀdamaging reagents, such as
platinum analogs and alkylating agents, and the inhibitors of DNAꢀrepair enzymes (PARP
22–24
25,26
inhibitors).
This demonstrates the potential of CDK12 as a therapeutic target
that may
2
7
induce synthetic lethality in cancers.
In order to elucidate the detailed functions and biological roles of CDK12 and validate its
potential as a therapeutic target, we initiated the exploration of molecular probes targeting
CDK12. A highꢀthroughput screen (HTS) of an inꢀhouse compound library was performed and
hit compound 1 was identified as a modestly potent CDK12 inhibitor (Figure 1). Compound 1
showed optimal selectivity over other CTD kinases (CDK7, 8, and 9) and also showed potent
2
8
inhibition of another CDK subtype, CDK2. As CDK2 is also known to regulate the cell cycle, a
significant reduction of the CDK2 inhibitory activity was necessary for the use of this chemical
ACS Paragon Plus Environment
5

Page 7 of 74
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
series as molecular probes for CDK12 in cells. To improve the selectivity of 1 for CDK12 over
CDK2, chemical optimization was conducted.
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. Profile of hit compound 1 identified from inꢀhouse HTS campaign.
This report presents the discovery process of the structurally novel, potent, and selective
nonꢀcovalent CDK12 inhibitor 2 by using a conformationꢀoriented design inspired by
2
9,30
experimentally defined real conformations in the Cambridge Structural Database (CSD)
(Figure 2). In the process of our research, several selective CDK12 inhibitors 3 and 4 were
31,32
reported by other groups.
As it is generally preferred to use structurally distinct chemical
3
3,34
series for target validation to exclude the possibility of unknown offꢀtarget effects,
our novel
compound 2 will provide another valuable option for molecular biologists and the
pharmaceutical industry.
Figure 2. Newly discovered selective CDK12 inhibitor 2 and reported inhibitors.
ACS Paragon Plus Environment
6

Journal of Medicinal Chemistry
Page 8 of 74
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
Chemistry
The synthesis of compounds 1 and 7–13 is illustrated in Scheme 1. Commercially available
ꢀaminopiperidine derivative 5 was converted to intermediate 6 by sulfonylation, Suzuki
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
coupling, and removal of the Boc group. Subsequent Nꢀarylation of 6 under several conditions
afforded target compounds 1 and 7–12, whereas compound 13 was synthesized by Nꢀmethylation
of 1.
a
Scheme 1. Synthesis of 1 and 7–13
a
Reagents
and
conditions:
(a)(i)
4ꢀbromobenzenesulfonyl
chloride,
Et N,
THF,
rt,
96%;
(ii)
3
1
ꢀmethylꢀ4ꢀ(4,4,5,5ꢀtetramethylꢀ1,3,2ꢀdioxaborolanꢀ2ꢀyl)ꢀ1H-pyrazole, Pd(dppf)Cl ·CH Cl K CO DMF, H O, 100 °C, 83%;
2
2
2,
2
3,
2
(iii) HCl/MeOH, rt, 99%.; (b) 6ꢀfluoronicotinonitrile (for 1) or 6ꢀchloroꢀ2ꢀpyridinecarbonitrile (for 7) or
2
2
1
ꢀchloroisonicotinonitrile (for 8), DIPEA, NMP, 120–200 °C, 5–73%; (c) 2ꢀbromopyridine (for 9) or
ꢀbromoꢀ5ꢀmethoxypyridine (for 10) or 2ꢀbromoꢀ5ꢀ(trifluoromethyl)pyridine (for 11), Pd (dba) , Davephos, tertꢀBuONa, toluene,
2
3
00 °C, 7–29%; (d) 4ꢀfluorobenzonitrile, K CO , DMSO, 120 °C, 9%; (e) MeI, NaH, THF, DMF, 0 °C to rt, 85%.
2
3
The synthesis of compounds 16 and 17 is depicted in Scheme 2. Coupling of 14 with
ꢀchloronicotinonitrile, followed by removal of Boc group, yielded intermediate 15. Then,
6
acylation of 15 using 4ꢀbromobenzoyl chloride and subsequent Suzuki coupling afforded
compound 16. Alkylation of 15 using 4ꢀbromobenzyl bromide, followed by Suzuki coupling,
yielded compound 17.
ACS Paragon Plus Environment
7

Page 9 of 74
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
a
Scheme 2. Synthesis of 16 and 17
a
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
Reagents and conditions: (a)(i) 6ꢀchloronicotinonitrile, DIPEA, DMF, 120 °C, 66%; (ii) HCl/EtOAc, EtOH, rt, 98%; (b)
4
ꢀbromobenzoyl chloride, DIPEA, THF, rt, 89%; (c) 1ꢀmethylꢀ4ꢀ(4,4,5,5ꢀtetramethylꢀ1,3,2ꢀdioxaborolanꢀ2ꢀyl)ꢀ1H-pyrazole,
3
5
PdCl (Amphos) , Cs CO , DMF, 100 °C (microwave), 61–82%; (d) 4ꢀbromobenzyl bromide, K CO , DMF, 80 °C, 96%.
2
2
2
3
2
3
Compounds 21–30, 33, and 34 were synthesized as shown in Scheme 3. Commercially
available transꢀ1,4ꢀcyclohexanediamine derivative 18 was converted to intermediate 20 using
BuchwaldꢀHartwig coupling and Suzuki coupling sequentially. Amidation, methylsulfonylation,
alkoxycarbonylation, or alkylaminocarbonylation of the arylamino group in 20, followed by
removal of the Boc group, and finally coupling with 2ꢀhalopyridine derivatives or
2
ꢀchloriquinazoline afforded compounds 21–29. Compound 30 was obtained from 20 by
deprotection of Boc group and coupling with 6ꢀfluoronicotinonitrile. The cisꢀisomers 33 and 34
were synthesized from 31 in an analogous manner to those of 21 and 27.
a
Scheme 3. Synthesis of 21–30, 33, and 34
a
Reagents and conditions: (a) 1ꢀbromoꢀ4ꢀiodobenzene, tertꢀBuONa, Pd (dba) , Xantphos, toluene, 80 °C, 26–45%; (b)
2
3
1
ꢀmethylꢀ4ꢀ(4,4,5,5ꢀtetramethylꢀ1,3,2ꢀdioxaborolanꢀ2ꢀyl)ꢀ1H-pyrazole, Cs CO , PdCl (Amphos) , DMF, H O, 100 °C, 84–96%;
2 3 2 2 2
ACS Paragon Plus Environment
8

Journal of Medicinal Chemistry
Page 10 of 74
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
(c) acetyl chloride (for 21, 29, and 33) or phenylpropionyl chloride (for 23), DIPEA, THF, 0 °C, 79–89%; (d) methanesulfonyl
chloride, pyridine, rt, 81%; (e) methyl chloroformate (for 24) or benzyl chloroformate (for 25), DIPEA, CH Cl , rt, 62–85%; (f)
2
2
ethyl isocyanate, THF, rt, 53%; (g) benzyl isocyanate (for 27 and 34) or cyclohexylmethyl isocyanate (for 28), THF, rt, 68–90%;
(h) (i) TFA, rt (ii) 6ꢀchloronicotinonitrile, DIPEA, DMF, 120 °C, 19–23%; (i) (i) HCl/EtOAc, rt (ii) 6ꢀfluoronicotinonitrile,
DIPEA, NMP, 80–100 °C, 33–82%; (j) (i) HCl/dioxane, CH Cl , rt (ii) 6ꢀbromonicotinonitrile, Cs CO , DMF, rt, 15%; (k) (i)
2
2
2
3
TFA, rt (ii) 2ꢀchloroquinazoline, DBU, DMF, 120 °C, 37%.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Compounds 2, 39, 40, 44, 46, and 47 were synthesized as depicted in Scheme 4. Intermediate 19
was converted to benzylurea 35 followed by reduction of the bromo group afforded 36.
Compound 37 was synthesized by Suzuki coupling of 35 and 3ꢀpyridinylboronic acid.
Compounds 35–37 were converted to aminopyridines 38–40, respectively, by removal of the Boc
group and subsequent coupling with 6ꢀfluoronicotinonitrile. Suzuki coupling of bromide 38 with
borates 41 and 42 yielded compounds 2 and 44, respectively, whereas 38 was converted to borate
45
by
Miyaura
reaction
followed
by
Suzuki
coupling
with
6ꢀ
or
4
ꢀbromoꢀ1ꢀmethylpyridinꢀ2(1H)ꢀone to give the designed compounds 46 and 47.
a
Scheme 4. Synthesis of 2, 39, 40, 44, 46 and 47
ACS Paragon Plus Environment
9

Page 11 of 74
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
a
Reagents and conditions: (a) benzyl isocyanate, Et N, THF, rt, 93%; (b) H , Pd/C, Et N, EtOH, THF, rt, quant.; (c)
3
2
3
3
ꢀpyridinylboronic acid, Cs CO , PdCl (Amphos) , DMF, H O, 100 °C, 95%; (d) (i) HCl/EtOAc, THF, rt (ii)
2 3 2 2 2
6ꢀfluoronicotinonitrile, DIPEA, NMP, 100 °C, 53 80%; (e) 41 (for 2), 42 (for 44), AcOK, Pd(dppf)Cl ·CH Cl , NMP, 80–
–
2 2 2
1
4
00 °C, 35–69%; (f) 43, AcOK, Pd(dppf)Cl , DMSO, 80 °C, 88%; (g) 6ꢀbromoꢀ1ꢀmethylpyridinꢀ2(1H)ꢀone (for 46) or
2
ꢀbromoꢀ1ꢀmethylpyridinꢀ2(1H)ꢀone (for 47), Cs CO , PdCl (Amphos) , DMF, H O, 80 °C, 44–48%.
2
3
2
2
2
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Results & Discussion
Biological evaluation of CDK12, 13, and 9 inhibitory activity was conducted using the
TRꢀFRET assay system (PerkinElmer), which detects the inhibition of the phosphorylation of an
acceptorꢀlabeled peptide substrate and reports the results as IC values. The inhibitory activity
5
0
of compounds against CDK7 and CDK8 was assessed by the binding assay using the TRꢀFRET
assay system, in which an acceptorꢀlabeled ligand was used as a probe against GSTꢀtagged
CDK7 or GSTꢀtagged CDK8. The displacement of the probe was detected by the decrease in
FRET signal, and the results were expressed as IC values. The inhibitory activity of CDK2 was
5
0
measured using a radiometric assay system with Histone H1 as the substrate and radiolabeled
33
[
γꢀ P] ATP. The radiolabeled substrate was detected and the results were expressed as IC50
values.
The aminopyridyl moiety of the hit compound 1 is a prevalent substructure in hinge
3
6–38
binding scaffolds for kinase inhibitors.
Thus, we hypothesized that the
6
ꢀaminonicotinonitrile moiety in compound 1 could bind to the hinge region of CDK12 and first
addressed the modification of this moiety (Table 1). The analogs of 1, such as benzonitrile (12)
and Nꢀmethylamino derivative (13), showed decreased activity, which indicates that the
aminopyridyl moiety is essential for activity, thereby supporting our assumption. It was also
revealed that the 3ꢀcyano group of 1 is critical for activity (1 vs 7–9). Considering the results of
ACS Paragon Plus Environment
10

Journal of Medicinal Chemistry
Page 12 of 74
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 methoxy (10) and trifluoromethyl (11) derivatives, an electron withdrawing functional group
was suggested to be required at the 3ꢀposition of the pyridine ring.
Table 1. SAR around the amino pyridyl moiety
0
1
2
3
4
5
6
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
CDK12
CDK2
Compound
R
a
a
IC (µM)
IC (µM)
5
0
50
0.36
0.01
1
(
0.18–0.69) (0.0088–0.012)
12
>30
>30
NT
68
1
3
(15–322)
0
.032
7
>30
(
0.028–0.036)
46
32–66)
8
9
>30
>30
(
1.2
(
0.97–1.4)
2
1
1
0
>30
1.8
(
17–26)
0.0058
1
1
(
0.87–3.7) (0.0047–0.0071)
a
n = 2, 95% confidence intervals shown in parentheses. NT = not tested.
To identify key structures to improve selectivity, preliminary docking models of compound 1
with CDK2 and CDK12 were constructed, respectively. From these, plausible models in which
the aminopyridyl moiety interacts with the hinge region were selected (Figure 3). The
ACS Paragon Plus Environment
11

Page 13 of 74
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
comparison of these models suggests that an oxygen atom from the sulfonyl moiety in 1 could
interact with Lys89 of CDK2, while that interaction is not present in the model of CDK12
because the corresponding amino acid is not conserved in CDK12 and is replaced by Gly822.
The assumption based on the model prompted us to modify the sulfonamide moiety to weaken or
disrupt the putative interaction with Lys89 in CDK2.
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 3. Docking models of compound 1 with CDK2 (left) and CDK12 (right). The
sulfonamide moiety of 1 is suggested to interact with the side chain of Lys89 in CDK2. As the
corresponding amino acid in CDK12 is Gly822, the hydrogen bonding could not be formed.
Replacement of the sulfonyl moiety in 1 with a carbonyl (16) or alkyl (17) moiety was not
tolerated (Table 2). As sulfonamides are known to have strong conformational preferences in
which the nitrogen lone pair bisects the O=S=O angle in the Newman projection (Figure 4,
2
8,39
left),
we speculated that the biaryl moiety in 1 would be oriented perpendicularly against the
piperidine ring. This notion was supported by the single crystal Xꢀray structure of 1 (Figure 4,
right). Hence, it was inferred the orthogonal conformation of the sulfonamide should be
mimicked to allow compounds to exhibit inhibitory activity. To explore the candidate
substructures for the replacement of sulfonamide, we referred to the experimentally determined
ACS Paragon Plus Environment
12

Journal of Medicinal Chemistry
Page 14 of 74
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
40
conformations in the CSD. The conformational search of the CSD suggested that tertꢀamides
and tertꢀsulfonamides bearing a cyclohexyl moiety would show the desired conformation (Figure
5), in which the substituent on the amino moiety is perpendicular to the cyclohexyl ring. As the
total number of useful examples in the CSD was too small to draw conclusions, we tested our
speculation by direct experimentation. To that end, compounds 21, 22, and 33 were designed and
synthesized (Figure 6 and Table 2). Although these replacements for sulfonamide are not
common in medicinal chemistry, the trans isomers 21 and 22 exhibited similar levels of potency
to compound 1. In contrast, the decreased activity of the cis isomer 33 implies that the assumed
diꢀequatorial conformation of the cyclohexyl ring in trans isomers might be necessary for the
proper orientation of substituents (Supporting information 5, supplementary figures 5 and 10).
Intriguingly, the selectivity of compounds 21 and 22 was better than that of 1 for CDK12 over
CDK2. Especially 21 showed 3ꢀfold selectivity for CDK12 over CDK2, whereas 22 still showed
3ꢀfold stronger inhibitory activity for CDK2 rather than CDK12. This is probably attributable to
the methylsulfonyl group of 22, which might interact with Lys72 of CDK2 in a similar manner to
the sulfonamide moiety of 1. An overlay of the single crystal Xꢀray structures of 1, 21, and 22,
shown in Figure 7, revealed that these compounds adopt similar conformations, at least in their
solid state. In contrast, the amide derivative 16 does not adopt the desired conformation.
Contrary to 21 and 22, the transꢀcyclohexylamino derivative (30) resulted in reduced activity,
which indicates that the acetyl group (21) or methylsulfonyl group (22) are essential for activity.
0
1
2
3
4
5
6
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
4
1
42
The coꢀcrystal structure of the analog of 21 (compound 29) with CDK12 was solved (Figure
8
). The resulting crystallographic model revealed the expected bioactive conformation described
above. In addition, the oxygen atom of the acetyl group and the backbone of Asp819 are within
hydrogen bonding distance. These observations indicate a dual role for the acetyl group,
ACS Paragon Plus Environment
13

Page 15 of 74
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
comprising the adjustment of ligand conformation and formation of a hydrogen bonding with the
protein.
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
ACS Paragon Plus Environment
14

Journal of Medicinal Chemistry
Page 16 of 74
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. SAR of compounds 16, 17, 21, 22, 30, and 33
0
1
2
3
4
5
6
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
CDK12
CDK2
Compound
X
a
a
IC (µM)
IC (µM)
5
0
50
2
3
1
1
6
7
NT
(
18–28)
1
6
NT
(
14–19)
0.13
0.47
2
1
3
(
0.090–0.18) (0.35–0.63)
3
>30
>100
0
.11
0.031
2
2
0
(0.083–0.15) (0.028–0.036)
8.6
4.9–15)
9.8
(4.0–24)
3
(
a
n = 2, 95% confidence intervals shown in parentheses. NT = not tested.
ACS Paragon Plus Environment
15

Page 17 of 74
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Figure 4. Preferences in the conformation of the sulfonamide moiety. Sulfonamides generally
have strong conformational preferences, as shown in the Newman projection (left), which is
preserved in the single crystal Xꢀray structure of 1 (right).
Figure 5. Representative examples of tertiary amides and tertiary sulfonamides bearing a
cyclohexyl amino moiety in the Cambridge Structural Database.
ACS Paragon Plus Environment
16

Journal of Medicinal Chemistry
Page 18 of 74
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. Schematic design of alternatives to the sulfonamide moiety. The designs are based on
comparisons with experimentally determined conformations in the Cambridge Structural
Database.
Figure 7. Overlay of single crystal Xꢀray structures of various ligands. Left: 1 (blue, CCDC
1
835992), 21 (green, CCDC 1835994), and 22 (red, CCDC 1835995), with similar potencies,
ACS Paragon Plus Environment
17

Page 19 of 74
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
overlap well with each other. Right: less potent 16 (amber, CCDC 1835993) exhibited a different
conformation to 1 (blue).
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Figure 8. The Xꢀray coꢀcrystal structure of 29 with CDK12 and cyclin K (PDB:6CKX, 2.75Å).
Compound 29 occludes the ATPꢀbinding site of CDK12 and the aminoquinazoline moiety
interacts with Met816 in the hinge region. The carbonyl oxygen in the center of the compound is
within hydrogen bonding distance of the backbone Asp819. The pyrazole moiety of 29 extends
out towards the solvent.
ACS Paragon Plus Environment
18

Journal of Medicinal Chemistry
Page 20 of 74
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
Based on the in vitro results, further derivatization of compound 21 was conducted to identify
more selective analogues. Replacement of the acetyl group with a variety of substituents was
examined, and representative results are shown in Table 3. The methyl carbamate (24) and
ethylurea derivatives (26) showed similar levels of CDK12 inhibitory activity, whereas the
selectivity over CDK2 was slightly improved. Larger substituents, such as phenylpropanamide
0
1
2
3
4
5
6
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
(23) and benzyl carbamate (25), showed similar results to 24 and 26. In contrast, benzylurea
derivative (27) exhibited >800ꢀfold selectivity for CDK12 over CDK2. The overlay of the
docking model of 27 with CDK12 and the reported crystal structure of CDK2 suggests that this
dramatic improvement in selectivity was derived from the steric clash of the benzyl moiety with
Lys72 in CDK2 (Figure 9). It is noteworthy that there are apparent differences in selectivity,
despite the high similarity of their 2D structures (27 vs 23 and 25). We speculated that this
difference is dependent on their distinct 3D conformational preferences. The overlay of the
single crystal Xꢀray structures demonstrates that only urea derivative 27 adopts the bent
conformation of the benzyl moiety (Figure 10); moreover, this bent conformation is thought to
be essential to cause the steric clash with Lys72 of CDK2 (Figure 9). This propensity was
confirmed by global conformational preferences in the CSD (Figure 11); that is, only the
“C(O)ꢀNHꢀCꢀC” substructure (Figure 11 left) tends to adopt a bent conformation, whereas other
substructures “C(O)ꢀCꢀCꢀC” (Figure 11 middle) and “C(O)ꢀOꢀCꢀC” (Figure 11 right) strongly
prefer extended conformations. The cyclohexylmethyl analog (28) showed similar levels of
activity and selectivity to 27, which implies that it is not the aromaticity or planarity of the
benzyl moiety, but the bulkiness that is necessary to weaken CDK2 activity. This is consistent
with the assumptions for CDK12 selectivity discussed above (Figure 9). In accordance with the
ACS Paragon Plus Environment
19

Page 21 of 74
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
results for compound 33, up to 100 µM of the cisꢀisomer (34) did not show any CDK12
inhibitory activity.
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
ACS Paragon Plus Environment
20

Journal of Medicinal Chemistry
Page 22 of 74
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. SAR of compounds 23–28, and 34
0
1
2
3
4
5
6
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
CDK12
CDK2
Compound
R
a
a
IC (µM)
IC (µM)
5
0
50
0.20
1.6
(1.2–2.2)
24
(
0.16–0.27)
0.46
4.8
(4.1–5.7)
26
(
0.36–0.60)
0.16
0.086–0.30)
2.9
(2.5–3.4)
2
3
(
0.64
2.5
(1.9–3.4)
25
(
0.38–1.1)
0.063
2
7
8
>100
>100
NT
(
0.032–0.12)
0.099
0.064–0.15)
2
(
34
>30
a
n = 2, 95% confidence intervals shown in parentheses. NT = not tested.
ACS Paragon Plus Environment
21

Page 23 of 74
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Figure 9. Overlay of docking model of 27/CDK12 (ligand: white, protein: green) and crystal
4
3
structure of CDK2 (2FVD, ligand: omitted, protein: cyan). Lys72 in CDK2 is highlighted.
Figure 10. Overlay of single crystal Xꢀray structures of ligands 23 (magenta, CCDC 1835996),
2
5 (cyan, CCDC 1835997), and 27 (green, CCDC 1835998).
ACS Paragon Plus Environment
22

Journal of Medicinal Chemistry
Page 24 of 74
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 11. Torsion histograms for each substructure in the Cambridge Structural Database. The
xꢀaxis represents torsion angle bins for the highlighted bonds and the yꢀaxis represents
occurrence. Substructure queries are depicted above each histogram (cyclic structures around the
highlighted bonds were omitted for precise estimation). The histogram for “C(O)ꢀNHꢀCꢀC” is
illustrated on the left, the histogram for “C(O)ꢀCꢀCꢀC” is illustrated in the middle, and the
histogram for “C(O)ꢀOꢀCꢀC” is illustrated on the right.
ACS Paragon Plus Environment
23

Page 25 of 74
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Finally, the distal pyrazole ring in 27 was optimized to adjust the physicochemical properties (e.g.
solubility) while retaining activity and selectivity. Surprisingly, the truncation of the pyrazole
ring (39) led to diminished activity (Table 4), although the pyrazole ring of 27 is expected to be
solventꢀexposed (Figures 8 and 9). The 3ꢀpyridyl derivative (40) showed a similar level of
potency, selectivity, and solubility as 27, which also supports the contribution of the distal
aromatic ring for activity. These observations imply two possibilities: the pyrazole ring of 27 (or
pyridine ring of 40) might interact with a solvent exposed residue of CDK12 via a water
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
9
molecule; or, these heterocycles may interact with the Cꢀterminal kinase extension, which is
disordered and not defined in the template crystal structure (Supporting information 1). To
improve aqueous solubility further, the more hydrophilic methylpyridone ring was employed as a
distal ring (2, 44, 46, and 47). Among them, compound 2 exhibited the highest aqueous solubility
and good permeability with double digit nanomolar CDK12 inhibitory activity and excellent
selectivity over CDK2. Although compound 2 is not the best one for selectivity over CDK9, it
was considered to be the most suitable molecule for cellꢀbased studies. The selectivity data for
known CTD kinases are summarized in Table 4. Compound 2 showed excellent CDK12
44
selectivity over CDK9, 8, and 7 compared with the nonꢀselective kinase inhibitor staurosporine
Table 5). Meanwhile, 2 turned out to be a strong CDK13 inhibitor, which is reasonable because
(
CDK13 is the closest homolog of CDK12 (92% identity in kinase domain). In contrast, up to 100
µM of compound 34 did not inhibit any of the five CTD kinases tested, including CDK13, thus
3
4 could be useful as a negative control. The global kinase selectivity of 2 was also investigated
Supporting information 2), and only three out of 441 kinase assays showed ≥80% inhibition at 1
µM compound concentration. As these three kinases (TAOK1, DYRK2, and STK39) are not
(
4
5
related to splicing or RNA processing, their inhibition is not likely to disturb the investigation
ACS Paragon Plus Environment
24

Journal of Medicinal Chemistry
Page 26 of 74
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
of RNA biology using 2. The phosphorylation of CTD Ser2 was evaluated using a
ERBB2/HER2 amplified human breast cancer cell line, SKꢀBRꢀ3. According to the Human
4
6
Protein Atlas database, CDK12 mRNA is highly expressed in SKꢀBRꢀ3 cells compared to other
cell lines, whereas CDK9 mRNA is expressed at a level similar to other cell lines. Furthermore,
0
1
2
3
4
5
6
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
24
the CDK12 gene is coꢀamplified with the HER2 oncogene in this cell line; therefore, SKꢀBRꢀ3
is a plausible pathophysiological model for characterization of CDK12 inhibitors as therapeutic
2
5
agents targeting HER2ꢀpositive breast cancer. Table 6 shows that compound 2 inhibited the
phosphorylation of the CTD Ser2 in SKꢀBRꢀ3 cells at low submicromolar concentrations and its
4
7
degree of inhibition was comparable to the known CDK inhibitor, ibulocydine. In our assay
system, ibulocydine showed inhibitory activity for CDK9 and CDK7 consistent with previous
reports; however, we noticed that ibulocydine also showed inhibitory activity against CDK12
(
IC50 values for CDK9, CDK7, and CDK12 were 220, 1200, and 110 nM, respectively).
4
8
Meanwhile, the negative control 34 (up to 100 µM) did not show any inhibition. In addition to
showing inhibitory activity towards Ser2 phosphorylation, compound 2 inhibited the growth of
32
SKꢀBRꢀ3 cells at submicromolar concentrations. To clarify the mode of inhibition of
compound 2, the time dependency of CDK12 inhibition was investigated by using the known
covalent CDK12 inhibitor 3 (THZꢀ531) as a control (Supporting information 3). Compound 2
did not show any timeꢀdependent inhibition for CDK12, whereas 3 showed clear time
3
0
dependency as reported. This result suggests compound 2 is a nonꢀcovalent CDK12 inhibitor,
as expected from its structural features.
ACS Paragon Plus Environment
25

Page 27 of 74
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Table 4. Activity, selectivity, and physicochemical properties of compounds 2, 27, 39, 40, 44, 46
and 47
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
CDK2
CDK9 Solubility logD PAMPA
CDK12
Compound
R
H
IC50a
IC50a
at pH 6.8 at pH at pH 7.4
a
IC (µM)
5
0
(
µM)
(µM)
(µg/mL)
7.4
(nm/s)
0
.063
2
7
>100
>30
>30
0.33
2.8
288
(
0.032–0.12)
0
.75
56
(33–96)
39
4.0
0.67
3.3
3.0
2.7
2.6
2.3
2.4
NT
280
NT
NT
247
273
(
0.36–1.6)
0
.059
40
44
46
47
>100
18
>30
15
(0.014–0.25)
0
.25
4.7
(
(
0.11–0.56)
(10–32) (11–20)
0
.31
27
>30
<0.090
<0.080
7.7
0.12–0.80)
(15–49)
0.044
(0.023–0.084)
>100
>100
>30
0
.052
16
(8.4–30)
2
(
0.023–0.12)
a
n = 2, 95% confidence intervals shown in parentheses. NT = not tested.
ACS Paragon Plus Environment
26

Journal of Medicinal Chemistry
Page 28 of 74
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 5. Selectivity over Cꢀterminal domain kinases
CDK12
CDK9
CDK8
CDK7
CDK13
Compound
a
a
a
a
a
IC (µM)
IC (µM)
IC (µM)
IC (µM)
IC (µM)
5
0
50
50
50
50
0.052
0.023–0.12)
16
(8.4–30)
0.010
(0.0030–0.034)
2
>10
>10
0
1
2
3
4
5
6
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
(
34
>30
>30
>10
>10
>10
0
.053
0.014
0.015
0.12
0.057
Staurosporine
(
0.033–0.086) (0.010–0.021) (0.010–0.022) (0.059–0.23) (0.023–0.14)
n = 2, 95% confidence intervals shown in parentheses.
a
Table 6. Inhibition of Cꢀterminal domain Ser2 phosphorylation and growth inhibition in
SKꢀBRꢀ3 cell
CTD Ser2 phosphorylation growth inhibition
Compound
a
IC (nM)
GI (µM)
5
0
50
2
195 (± 2.6)
0.8 (± 0.36)
b
c
3
4
ND
NT
c
Ibulocydine
259 (± 2.3)
NT
a
b
c
n = 4, SD values shown in parentheses. Not detected up to 100 µM. Not tested.
Conclusion
We discovered novel and selective CDK12 inhibitors by utilizing experimentally determined
conformations in the CSD. The optimized compound 2 showed double digit nanomolar
inhibitory activity for CDK12 (and CDK13), good selectivity over other CTD kinases, and
excellent physicochemical properties. Furthermore, 2 exhibited high kinomeꢀwide selectivity,
and inhibited CTD Ser2 phosphorylation in SKꢀBRꢀ3 cells and cell growth. Compound 2 and the
negative control 34 represent valuable chemical probes to study the functions of CDK12 in
ACS Paragon Plus Environment
27

Page 29 of 74
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
splicing, RNA processing, and its other biological roles. In addition, these compounds could
serve as a foundation for the development of CDK12 inhibitors as chemotherapeutic agents alone,
or in combination with other therapies.
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
ACS Paragon Plus Environment
28