Discovery of MFH290: A Potent and Highly Selective Covalent Inhibitor for Cyclin-Dependent Kinase 12/13

Article abstract of DOI:10.1021/acs.jmedchem.9b01929

Genetic depletion of cyclin-dependent kinase 12 (CDK12) or selective inhibition of an analog-sensitive CDK12 reduces DNA damage repair gene expression, but selective inhibition of endogenous CDK12 is difficult. Here, we report the development of MFH290, a novel cysteine (Cys)-directed covalent inhibitor of CDK12/13. MFH290 forms a covalent bond with Cys-1039 of CDK12, exhibits excellent kinome selectivity, inhibits the phosphorylation of serine-2 in the C-terminal domain (CTD) of RNA-polymerase II (Pol II), and reduces the expression of key DNA damage repair genes. Importantly, these effects were demonstrated to be CDK12-dependent as mutation of Cys-1039 rendered the kinase refractory to MFH290 and restored Pol II CTD phosphorylation and DNA damage repair gene expression. Consistent with its effect on DNA damage repair gene expression, MFH290 augments the antiproliferative effect of the PARP inhibitor olaparib.

Full text of DOI:10.1021/acs.jmedchem.9b01929

Subscriber access provided by Uppsala universitetsbibliotek
Article
Discovery of MFH290: A Potent and Highly Selective
Covalent Inhibitor for Cyclin-Dependent Kinase 12/13
Yao Liu, Mingfeng Hao, Alan Leggett, Yang Gao, Scott B Ficarro, Jianwei Che, Zhixiang He,
Calla Olson, Jarrod A. Marto, nicholas Kwiatkowski, Tinghu Zhang, and Nathanael S Gray
J. Med. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.jmedchem.9b01929 • Publication Date (Web): 05 Jun 2020
Downloaded from pubs.acs.org on June 5, 2020
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 59
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Discovery of MFH290: A Potent and Highly Selective Covalent Inhibitor for Cyclin-
Dependent Kinase 12/13
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
Yao Liu⊥,†,‡, Mingfeng Hao⊥,†,‡, Alan L. Leggett⊥,†, Yang Gao⊥,†, Scott B. Ficarro⊥,†,§, Jianwei Che
⊥,†,
Zhixiang He⊥,†, Calla M. Olson⊥,†, Jarrod A. Marto⊥,†,§, Nicholas P. Kwiatkowski⊥,†,*, Tinghu
Zhang⊥,†,*, Nathanael S. Gray⊥,†,
*
⊥Department of Biological Chemistry and Molecular Pharmacology, Harvard Medical
†
School, Boston, MA 02115, USA; Department of Cancer Biology, Dana Farber Cancer
§
Institute, 450 Brookline Avenue, Boston, MA 02215, USA; Blais Proteomics Center,
Dana Farber Cancer Institute, Harvard Medical School, Boston, MA 02215, USA;
KEYWORDS: CDK12/13 inhibitor, covalent inhibitor, cysteine 1039.
ABSTRACT: Genetic depletion of cyclin-dependent kinase 12 (CDK12) or selective
inhibition of an analog-sensitive CDK12 reduces DNA damage repair gene expression,
but selective inhibition of endogenous CDK12 is difficult. Here, we report the
development of MFH290, a novel cysteine (Cys)-directed covalent inhibitor of CDK12/13.
MFH290 forms a covalent bond with Cys-1039 of CDK12, exhibits excellent kinome
selectivity, inhibits the phosphorylation of serine-2 in the C-terminal domain (CTD) of
RNA-polymerase II (Pol II) and reduces the expression of key DNA damage repair genes.
Importantly, these effects were demonstrated to be CDK12-dependent as mutation of
Cys-1039 rendered the kinase refractory to MFH290 and restored Pol II CTD
phosphorylation and DNA damage repair gene expression. Consistent with its effect on
1
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 2 of 59
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
DNA damage repair gene expression, MFH290 augments the anti-proliferative effect of
the PARP inhibitor, Olaparib.
INTRODUCTION
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Cyclin Dependent Kinases, or CDKs, are an intensively investigated family of
evolutionarily conserved protein kinases that orchestrate the cell cycle and gene
1, 2
transcription . Human cells encode 20 distinct CDKs which possess highly similar ATP-
binding pockets thereby making the development of selective CDK inhibitors challenging3.
Despite these difficulties, some CDKs can be targeted quite selectively, such as CDK4/6
by the FDA approved drug palbociclib, based on recognition of unique combinations of
residues in the ATP-pocket. A more recently explored strategy has been to exploit
uniquely positioned cysteine residues with reactive, electrophile-modified inhibitors as
exemplified by the development of the CDK7/12/13 inhibitor THZ1 (Figure 1) which
targets an unusual cysteine 312 located on a loop that traverses the inhibitor binding
4
pocket . Further structure-guided optimization of THZ1 resulted in the development of the
5
selective CDK12/13 inhibitor THZ531 (Figure 1), which targets a spatially distinct
cysteine residue (Cys1039 for CDK12 and Cys1017 for CDK13).
CDK12 and the highly homologous protein CDK13 are both activated upon binding to
cyclin K to regulate transcription through phosphorylating Ser2 residues located on the
C-terminal domain (CTD) of RNA-polymerase II (RNA-Pol2) which contains 52 repeats of
6
the YSPTSPS motif . Selective inhibition of CDK12 kinase activity preferentially
downregulates a small set of genes that respond to and mediate the repair of DNA-
damage including homologous recombination (HR) repair genes BRAC1 and BRCA27, 8
.
2
ACS Paragon Plus Environment

Page 3 of 59
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Inhibitors of CDK12 increase DNA-damage and have been demonstrated to potentiate
the anti-proliferative effects of poly (ADP-ribose) polymerase (PARP) inhibitors9.
Recently, an in vivo combination treatment of Dinaciclib (Figure 1), a clinical-stage non-
selective CDK12 inhibitor, with Olaparib, a poly (ADP-ribose) polymerase (PARP)
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
10
inhibitor, has demonstrated a significant tumor growth inhibition in a PDX ovarian model ,
spurring a wave of medicinal chemistry efforts to discover more potent and selective
CDK12 inhibitors.
H
HO
N
N
Cl
N
Cl
O
N
N
N
N
O
N
O
N
N
H
N
N
N
O
H
H
N
O
N
N
NH
NH
H
N
N
H
THZ1
THZ531
Dinaciclib
O
N
HO
N
O
NH
N
N
N
2
N
N
4
H
N
NC
N
H
F
N
N
N
N
N
N
F
H
H
N
N
N
N
N
N
N
F
Cl
H
F
O
Cl
3
Figure 1. Structure of THZ1, THZ531, Dinaciclib and reported CDK12 inhibitors 2 to 4
Amongst the reported inhibitors of CDK12 and CDK13 are the acrylamide-modified
covalent inhibitors: THZ1 and THZ531 (Figure 1) in cancer cells, THZ531 was shown to
irreversibly inhibit phosphorylation of ser2 of RNA pol II CTD, represses transcription of
DNA damage response (DDR) genes and results in potent anti-proliferative effects5.
Although validated as a cellular probe, THZ531 is not suitable for in vivo murine efficacy
5
studies due to its poor metabolic stability resulting in high clearance in mice . In addition
11
to the covalent inhibitors, Dinaciclib , a clinical-stage pan-CDK inhibitor, exhibits activity
in numerous models but is not suitable as a selective pharmacological probe of CDK12.
3
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 4 of 59
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
Recently, 2,6,9-trisubstituted purines (2 and 4) and a trisubstituted urea (3, Figure 1)
have been reported as reversible CDK12 inhibitors that exhibit high kinase selectivity for
CDK1212-14. Here we report our efforts that culminated in the development of MFH290, a
highly-selective covalent inhibitor of CDK12/13 using a structure-guided drug design
strategy by combining structural features of THZ531 with a previously reported non-
0
1
2
3
4
5
6
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
15
covalent, pan-CDK inhibitor SNS032 (Figure 2A).
CHEMISTRY
The synthesis of compounds 1a, 1b and 1c is depicted in Scheme 1. Commercially
available intermediate 4a reacts with carboxylic acids 5, 6 or 7 through a HATU-mediated
condensation reaction under basic conditions to provide the final compound 1a or the
intermediate 8 and 9 after deprotection of the Boc group. While 8 was then subjected to
a second amide condensation reaction with 3-acrylamidobenzoic acid 5 to afford 1b,
compound 9 was treated with acryloyl chloride under basic conditions to introduce the
acrylamide warhead to generate 1c.
Scheme 1. Synthesis of compounds 1a-c
O
H
N
HO
N
O
N
O
H
O
S
N
5
a
S
N
H
O
1a
O
HO
HO
NBoc
O
N
N
N
O
O
N
N
N
N
O
O
O
S
S
S
N
O
S
NH2
O
S
S
c
O
S
S
N
H
6
b
S
8
N
H
NH
N
H
4a
1b
N
O
N
N
NBoc
O
S
9
N
H
O
N
H
d
7
b
O
NH
N
1c
4
ACS Paragon Plus Environment

Page 5 of 59
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Reagents and conditions: (a) HATU, DIEA, DMF, 8%; (b) (i) DMAP, EDCI,
triethylamine, DCM/DMF; (ii) TFA, DCM, 16-83%; (c) HATU, 4-acrylamidobenzoic acid,
o
DIEA, DMF, 12%; (d) Acryloyl chloride, DIEA, THF, 0 C, 51%.
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
Further analogs including 1d-1i, 1k and 1l were synthesized by reacting 2-(((2-
bromothiazol-5-yl)thio)methyl)-5-(tert-butyl)oxazole (10) with Boc protected-3-amino
piperidine or its variants or with 1,3-cyclohexane diamine to provide the intermediate 11
(Scheme 2). Free amino compound, derived from intermediate 11 after deprotection, was
subjected to either a one-step amide coupling reaction with compound 12 or a three-step
reaction as amide coupling (compound 13), reduction (compound 14) and acrylamide
formation with acryl chloride, leading to the final products 1d-1i, 1k and 1l.
Scheme 2. Synthesis of compounds 1d-1i, 1k and 1l.
Br
H
O
S
H
Boc
N
S
N
S
S
N
O
a
L
S
N
b
L
S
N
N
N
N
O
2
O N
O
10
11
13
d
COOH
O
1
H
c
R
N
S
R
L
N
S
HN
O
N
H2N
O
4
e
12
O
H
R
N
S
O
L
O
S
N
N
H
N
H
N
N
N
N
N
N
N
N
L =
R =
L =
R =
F
F
H
H
H
H
H
H
F
CF3
1l
1d
1e
1f
1g
1h
1i
1k
o
Reagents and conditions: (a) NMP, DIEA, 140 C, 6-67%; (b) (i) 4N HCl/dioxane, MeOH,
(ii) 4-nitrobenzoyl chloride, Py, 27-55%; (c) (i) 4N HCl/dioxane, MeOH, (ii) T P, DIEA,
3
5
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 6 of 59
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
DCM, 19-60%; (d) SnCl , ethyl acetate/MeOH, 80 C, 24-76%; (e) acryloyl chloride, DIEA,
2
o
MeCN, 0 C, 16-41%.
Scheme 3. Synthesis of compounds 1j, 1m-1o and 1u-1v
0
1
2
3
4
5
6
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
S
SCN
S
S
R2
c
S
S
R2
d
S
Br
a
b
H N
2
H2N
H N
2
Br
N A R1
18
N
A
N
A
A
N
R1
R1
R1
15
16
17
H
N
S
H
H
N
N
S
e
S
S
f
S
N
R2
S
A
R1
N
N
R2
R2
A
N
N
A
N
Boc
R1
O
O
H
R1
19
20
X
N
H
N
O
O
O
R2
=
N
O
CF3
N
N
N
N
A
R1
X
=
=
=
C
H
N
C
H
C
C
H
C
C
H
C
N
--
C
C
Me
C
1j
1m
1n
1o
1u
1v
o
Reagents and conditions: (a) KSCN, MeOH, 41-93%; (b) (i) NaBH , ethanol, 0 C to
4
o
room temperature, (ii) R CH X, ethanol, 65 C, 41-57%; (c) CuBr , t-BuONO, MeCN, 0
2
2
2
o
o
C, 37-94%; (d) 1-Boc-(R)-3-aminopiperidine, NMP, DIEA, 140 C, 53-67%; (e) 4N
HCl/dioxane, MeOH, 51-95% ; (f) 4-acrylamidobenzoic acid, T P, DIEA, DCM, 53% or (i)
3
benzoyl chloride, Py or benzoic acid, HATU, DIEA, DMF, (ii) SnCl , ethyl acetate/MeOH,
2
o
o
80 C, (iii) acryloyl chloride, DIEA, MeCN, 0 C, 7-17%.
Scheme 3 presents the strategy to synthesize compounds 1j, 1m-1o and 1u-1v. The
commercially available starting material: 2,4,5-trisubstituted thiazole, or 2,5-disubstituted
thiadiazole were reacted with KSCN to afford thiocyanate intermediate 16. In a one-pot
reaction thiocyanate 16 was reduced to corresponding thiol using NaBH , and then
4
alkylated with the appropriate halide. The intermediate 17 was subjected to Sandmeyer
reaction to convert the amine group to the bromide, which was then replace by a Boc-
6
ACS Paragon Plus Environment

Page 7 of 59
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
protected-3-amino-piperidine. Deprotection of Boc from intermediate 19 led to compound
20, which was then processed using the same procedure as shown in Scheme 2 to
generate the final products. The compound 1t was synthesized with a similar reaction
route with 2,4,5-trichloropyrimidine as the starting material shown in Scheme 4.
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
Scheme 4. Synthesis of compound 1t
Cl
Cl
Cl
Cl
S
Cl
S
d
a
b
c
N
N
N
N
N
O
O
O
Cl
N
Cl
N
SCN
Cl
N
N
HN
N
N
N
N
21
22
23
24
BocN
Cl
S
Cl
S
Cl
S
N
N
O
O
HN
N
HN
HN
N
N
N
e
f
O
N
O
N
O
N
NO2
25
NH2
HN
O
26
1t
o
Reagents and conditions: (a) KSCN, MeOH, 55%; (b) (i) NaBH4, ethanol, 0 C to RT,
(
ii) 5-(tert-butyl)-2-(chloromethyl)oxazole, ethanol, 65 oC, 17%; (c) 1-Boc-(R)-3-
o
aminopiperidine, NMP, DIEA, 140 C, 92%; (d) (i) 4N HCl/dioxane, MeOH, (ii) 4-
o
nitrobenzoyl chloride, Py, 89%; (e) SnCl , ethyl acetate/MeOH, 80 C, 20%; (f) acryloyl
2
o
chloride, DIEA, MeCN, 0 C, 65%.
The synthesis of compound 1p, 1q and 1r that contain different linkers between the
oxazole and thiazolyl ring was carried out as shown in Scheme 5 and 6. First, the Wittig
reaction was performed using aldehyde 29 and phosphonate intermediate 28, prepared
from 5-(tert-butyl)-2-(chloromethyl)oxazole using an Arbuzov reaction. Reduction of the
double bond with hydrogen/Pd provided intermediate 31, which was converted to
7
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 8 of 59
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
compound 32 by a Sandmeyer reaction. Installation of piperidine and acrylamide warhead
was followed by the same procedure as shown in Scheme 2.
Scheme 5. Synthesis of compound 1p
0
1
2
3
4
5
6
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
N
O
BocHN
S
N
N
O
N
N
O
a
29
b
c
N
d
O
O
O
O
P
BocHN
S
H2N
S
Cl
N
O
27
28
30
31
N
NH2
N
HN
S
N
N
O
N
N
N
BocN
S
f
HN
S
g
N
HN
O
N
O
O
O
S
e
Br
BocN
HN
32
33
34
1p
HN
O
o
Reagents and conditions: (a) triethyl phosphite, 120 C, 39%; (b) t-BuOK, THF, 84%;
o
(
c) H , Pd/C, MeOH, then 4N HCl/dioxane, 90%; (d) CuBr , t-BuONO, MeCN, 0 C, 53%;
2
2
o
(e) NMP, DIEA, 140 C, 63%; (f) 4N HCl/dioxane, MeOH, 88%; (g) 4-acrylamidobenzoic
acid, HOBT, EDCI, DIEA, DMF, 16%.
For 1q and 1r, the synthesis of 36 is straightforward as 35 and 27 are all commercially
available reagents and thus one-step substitution reaction led to the key intermediate
41a. Intermediate 41b was synthesized using a 3-step reaction sequence involving
coupling 2-bromothiazole-5-carbonitrile with 3-aminopiperidine followed by reduction of
cyano group to a primary amine and further reaction with 5-(tert-butyl)-2-
(chloromethyl)oxazole. The rest of reactions to install the acrylamide is same as shown
in Scheme 2.
Scheme 6. Synthesis of compounds 1q and 1r
8
ACS Paragon Plus Environment

Page 9 of 59
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
NH2
N
O
N
BocN
37
Cl
N
O
N
OH
2
7
O
N
N
A
N
O
b
S
a
O
Br
S
HN
N
A
Br
S
36
35
f
HN
S
O
N
NH2
N
BocN
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
NH2
A = O, 41a
A = N, 41b
A = O, 1q
A = N, 1r
N
N
O
27
e
BocN
37
CN
Cl
HN
S
HN
S
N
d
HN
CN
S
c
Br
O
BocN
BocN
40
3
8
39
o
o
Reagents and conditions: (a) NaOH, DMF, 120 C, 45%; (b) NMP, DIEA, 140 C, 22%;
o
(c) THF, DIEA, 80 C, 97%; (d) CoCl -6H O, NaBH , ethanol, 28%; (e) NMP, DIEA, 80
2
2
4
oC, 58%; (f) (i) 4N HCl/dioxane, MeOH, (ii) 4-acrylamidobenzoic acid, HOBT, EDCI, DIEA,
DMF, 22-38%.
RESULT
Development of MFH290 as covalent CDK12 inhibitor
High-throughput biochemical screening of kinase-focused chemical libraries is a common
method to find starting points for inhibitor optimization campaigns. However our efforts to
develop a kinase assay suitable for screening CDK12 reversible inhibitors were
unsuccessful. Therefore, we changed our strategy to use a competition-based pull-down
assay to screen for reversible CDK12 binders. Our aim was to identify compounds that
bound CDK12 that could be optimized and converted to covalent inhibitors targeting the
uniquely positioned cysteine 1039 (Cys1039) residue that lies in the C-terminal domain
of CDK12. Competitive pull-down assays were performed by utilizing biotinylated
(bio)THZ1, which has been shown to covalently bind and pulldown both CDK12 and
5
CDK13 (CDK12/13) . As bioTHZ1 can be pulled down using streptavidin-coated beads,
we gauged the ability of bioTHZ1 to bind free CDK12/13 by monitoring the extent to which
9
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 10 of 59
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
CDK12, CDK13, or their obligate binding partner, cyclin K, were pulled down. We
screened a focused collection of reported CDK inhibitors for their ability to protect CDK12
and 13 from being pulled down by bioTHZ1 (Supplementary figure S1A). Amongst all
known CDK inhibitors we screened, pretreatment with SNS032, a known inhibitor of
CDK2, 7 and 9 (Figure 2A), resulted in a significant reduction of cyclin K pulldown from
Jurkat cells by bioTHZ1 (Supplementary figure S1B), providing a starting point for
medicinal chemistry efforts.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Figure 2. (A) Structure and biochemical IC of SNS032 (B) Overlay of SNS032/CDK2
50
(PDB code: 5D1J, beige) and CDK12 (PDB code: 5ACB, blue) reveals that the piperidine
ring could be extended towards Cys1039.
The co-crystal structure of SNS032 with CDK2 (PDB: 5D1J) clearly illustrates the binding
mode where the: (1) 3-aminothiazole group forms two key hydrogen bonds with the hinge
region; (2) 5-(tert-butyl)oxazole ring occupies the back hydrophobic pocket II; and (3) the
piperidine faces the solvent exposed region. Structural overlay of CDK12 crystal structure
(PDB: 5ACB) and SNS032/CDK2 co-crystal structure indicated that the solvent-exposed
10
ACS Paragon Plus Environment

Page 11 of 59
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
piperidine ring would lie in close proximity to Cys1039 (Figure 2B). Subsequent chemistry
was performed to explore the ability to form a covalent bond with Cys1039 and to gain
additional selectivity for CDK12 relative to other kinases. All newly synthesized
compounds were assayed for CDK12 binding using the bioTHZ1 competition assay
Supplementary figure S2). To quickly assess the selectivity for CDK12 versus the other
CDKs targeted by SNS032, we used commercially available biochemical assays from Life
Technologies for CDK2/Cyclin A, CDK7/Cyclin H/MNAT1 and CDK9/Cyclin T1. In
addition, we engineered HAP1 cells by mutating Cys1039 of CDK12 to a non-reactive
serine to generate a chemical-genetic system to allow dissection of covalent-dependent
inhibitor pharmacology.
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
To explore advantageous modifications to SNS032 we started by installing the requisite
acrylamide warhead on the solvent exposed region facing the cysteine residue. For
example, compounds 1a, 1b and 1c were all designed with an acrylamide group attached
to either phenyl or piperidine ring which structural alignment suggested would position
this warhead in close proximity to Cys1039. All warhead modifications conserved the
ability to block cyclin K pulldown with bioTHZ1, indicating that CDK12/13 binding was
preserved (Table 1). In regard to selectivity, compound 1b maintained inhibitory activity
against all the CDKs tested, whereas 1a and 1c exhibited dramatically reduced inhibition
of CDK7 （Table 1 and Supplementary figure S2）. In contrast, 3-aminopiperidine,
which based on our work on THZ531 emerged as a privileged scaffold for selective
CDK12 inhibition, led to a dramatic reduction in inhibition of CDK2/7/9 activity while
CDK12/13 binding was retained (1d, hereafter named MFH290, Table 1). Moreover, the
S-enantiomer of MFH290 (1e) lost the ability to target CDK12, meaning that R
11
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 12 of 59
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
configuration of 3-aminopiperidine is required to provide the proper trajectory to target
Cys1039. Removing the carbonyl and installing an amino group through 3-amino-
cyclohexane (1f) slightly reduced CDK12 binding while also significantly restoring activity
on CDK2 and CDK9. Therefore, further modifications focused on retaining the 3-
aminopiperidine 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
Introduction of either a methyl group to the 2-position (1g), or two fluorines to 5-postion
of piperidine ring (1h), abrogated the inhibitory activity on all CDKs, whereas introduction
of a methyl group to the 6-position (1i) resulted in comparable, but less selective, activity
to MFH290. Moreover, a phenyl ring extending from the piperidine was found to be more
favored than a pyridine ring (1j) and a small group such as a fluorine (1k) was better
tolerated compared to a larger CF group at the 4-position (1l) (Table 1). Replacement
3
of the tert-butyloxazole group with pyridine (1m), pyrimidine (1n), or tetrahydropyran (1o)
rings resulted in compete loss of activity. Additionally, the thioether linkage to the core
thiazole was necessary for CDK12/13 activity as all non-thioether linked compounds such
as 1p, 1q and 1r were inactive (Table 2). Likewise, oxidation of the thiol to sulfoxide (1s)
diminished inhibition and binding to CDK2/7/9 and CDK12. Finally, a non-substituted
thiazole core was most favored as replacement with a pyrimidine (1t), thiadiazole (1u), or
even methylated thiazole (1v) significantly reduced CDK12 binding activity (Table 3).
Taken all together, MFH290 emerged as our best-in-class inhibitor from this series.
Table 1. IC50 value or pulldown competition percentage with bioTHZ1 of
compounds SNS032, 1a-1l against CDK2,7,9 and CDK12
12
ACS Paragon Plus Environment

Page 13 of 59
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
CDK12c
IC (nM)
50
Compound
I
CDK2a
CDK7b
CDK9b
(percentage)
O
SNS032
21
184
6
34.6
0.3
N
H
1
a
b
58.8
11
1140
99
24.6
1.79
O
O
O
1
N
0.3
2.6
O
1c
10.8
120
457
3.16
114
N
O
O
de
N
6020
0
1
1e
N
>10000
15.7
>10000
557
>10000
30.9
78.2
0.8
H
N
1f
13
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 14 of 59
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
O
O
CH3
1
g
h
N
N
1420
1480
7790
9600
257
3.1
3.5
1
1100
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
F
F
O
1
i
N
N
73.2
609
14.7
0.1
H C
3
O
1
j
561
211
3890
5540
514
138
1.2
0
N
O
F
1k
1l
N
O
N
142
3490
87.6
1.0
F
F
F
a
b
c
tested using Z’-Lyte assays with ATP concentrations at the apparent KM.
tested using Adapta assays with ATP concentrations at the apparent KM.
tested using CDK12 pulldown assay. The remaining CDK12 protein in the presence of
each compound was pulled down using bioTHZ1, and quantified by densitometry of the
bands of CDK12 (normalized with DMSO-treated).
e
“1d” is MFH290.
Table 2. IC50 value or pulldown competition percentage with bioTHZ1 of 1m-1s
against CDK2, 7,9 and CDK12
14
ACS Paragon Plus Environment

Page 15 of 59
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
O
H
N
S
O
N
R
N
N
H
CDK12c
IC (nM)
5
0
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Compound
R
CDK2a
>10000
>10000
>10000
CDK7b
>10000
>10000
>10000
CDK9b
>10000
>10000
>10000
(
percentage)
S
N
1m
100
F
N
S
F
1n
74.1
100
F
N
S
O
1o
1p
1q
O
>10000
>10000
>10000
41.0
N
N
>10000
3170
>10000
8870
>10000
7140
81.4
81.3
O
O
O
N
H
1r
N
O
O
S
1s
>10000
>10000
9220
85.7
N
a, b, c
defined in Table 1
Table 3. IC value or pulldown competition percentage with bioTHZ1 of compounds1t-1v
50
against CDK2, 7,9 and CDK12
O
H
N
O
N
O
II
S
N
N
H
15
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 16 of 59
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
CDK12c
IC50 (nM)
Compound
II
CDK2a
CDK7b
CDK9b
(percentage)
0
1
2
3
4
5
6
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
N
1t
>10000
5120
>10000
5.4
N
Cl
S
1
u
v
>10000
>10000
>10000
>10000
>10000
2870
100
N
N
S
1
N
91.2
CH3
a, b, c
defined in Table 1
To confirm the covalent bond formation of MFH290 with Cys1039 on CDK12, we analyzed
purified CDK12 protein by LC-MS after incubation with a 10-fold molar excess of MFH290
for 2 hours at 37 °C, and observed a mass shift consistent with addition of a single
molecule of MFH290 to CDK12 (Figure 3A and 3B). To confirm the site of modification,
labeled protein was digested with trypsin, and peptides analyzed by capillary
electrophoresis-MS/MS. Database search revealed exclusive modification of CDK12 by
MFH290 at Cys1039 (Figure 3C). Taken together, our mass spectrometry-based analysis
showed that MFH290 forms a covalent adduct with CDK12 by reacting with a single
cysteine, Cys1039. We therefore carried MFH290 forward into all subsequent cellular
experiments to characterize the inhibitor and evaluate CDK12/13 function.
16
ACS Paragon Plus Environment

Page 17 of 59
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 3. CDK12 is exclusively labeled by MFH290 at Cys1039. (A) Mass spectra and
(
B) zero-charge mass spectra of CDK12/Cyclin K complex treated with (top) DMSO or
bottom) a 10-fold excess of MFH290 for 1 hour at room temperature. (A) Mass spectral
(
peaks corresponding to Cyclin K, unmodified CDK12, and CDK12 covalently labeled with
MFH290 are indicated with black, green, or red circles, respectively. (B, left) Cyclin K is
unaffected by compound treatment while (B, right) CDK12 protein exhibits an increase in
mass consistent with modification by a single molecule of MFH290. (C) MS/MS spectrum
17
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 18 of 59
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 CDK12 peptide 1025LSKMAPPDLPHWQDC HE1041 modified with MFH290 indicates
covalent labeling of Cys1039. Ions of type b and y are indicated with blue and red glyphs
16
respectively. Inhibitor related ions 1 are marked with green glyphs.
0
1
2
3
4
5
6
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
To characterize the kinome-wide selectivity of MFH290, we employed DiscoverX’s
KinomeScan binding assays against a panel of 468 kinases at a concentration of 1 µM17,
18. As shown in Figure 4A and Supplementary table S1, MFH290 exhibits exquisite
selectivity with CDK13 as the only potently inhibited target (CDK12 is not included in the
panel). We also tested MFH290 on an extensive enzymatic CDKs panel which included
CDKs 1-9, 12-14 and 16. The measured IC s across all other CDKs are higher than 100
50
nM, indicating that MFH290 did not potently inhibit any of these CDKs (Figure 4B).
Figure 4. (A) Kinase profiling of MFH290. The levels of binding were measured at
concentrations 1 µM against a panel of 468 human kinases. The results for the binding
interactions are reported as blue circles in the KiNOME tree. (B) IC values for MFH290
50
18
ACS Paragon Plus Environment

Page 19 of 59
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
a
against CDK1-9, 12-14 and 16. tested in Z’-Lyte assays with ATP concentrations of K .
M
b
tested in LanthaScreen™ Eu Kinase Binding Assay with ATP concentrations of KM. c
tested using Adapta assays with ATP concentrations at the apparent KM. d tested in
radioisotope filter binding assay.
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
To define the working concentration range of MFH290, we employed the same
competition-based target engagement assay described above (Supplementary figure
S1A). Concentrations of MFH290 that block bioTHZ1-mediated pulldown of
CDK12/13/cyclin K identify doses at which MFH290 covalently binds CDK12/13 in cells.
Experiments in both Jurkat T-ALL and HAP1 (a near haploid) CML lines indicate that
MFH290 saturates CDK12/13 binding at a concentration of approximately 40 nM (Figure
5
A, Supplementary figure S3).
Figure 5. (A) MFH290 targets CDK12 and CDK13 in cells. Treatment of HAP1 cells with
MFH290 for 6 hours blocks pull down of CDK12 and associated cyclin K by biotinylated
19
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 20 of 59
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
THZ1. (B) Antiproliferation assay using WT, CDK12 C1039S, CDK13 C1017S, and
CDK12 C1039S/ CDK13 C1017S doubly-mutated HAP1 cells. Output is CellTiterGlo®
luminescent signal following 72-hour treatment with MFH290. (C) MFH290 reduces Pol II
CTD phosphorylation in a concentration-dependent manner. Shown are effects of
MFH290 on CTD phosphorylation in WT, CDK12 C1039S, CDK13 C1017S, and CDK12
C1039S/ CDK13 C1017S doubly-mutated HAP1 cells after 6 hours of treatment.
0
1
2
3
4
5
6
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
To determine the cellular selectivity of MFH290, we used CRISPR/Cas9-mediated allelic
8
replacement to establish isogenic cell lines expressing covalent inhibitor refractory
cysteine-to-serine mutant forms of CDK12 (C1039S), CDK13 (C1017S), and
CDK12/CDK13 double mutants (Supplementary figure S4A). These isogenic lines
enable us to assess the relative contribution of CDK12, CDK13, and CDK12/13 inhibition
to MFH290 pharmacology. Western blotting demonstrated that WT and mutant forms of
CDK12 and CDK13 were all expressed at similar levels (Supplementary figure S4B).
We previously reported that THZ531, our first-generation CDK12/13 inhibitor, led to a
5
pronounced and rapid decrease in cell proliferation , therefore we tested MFH290 across
its effective dose range for antiproliferative activity in the HAP1 isogenic lines. In WT
HAP1 cells, MFH290 displayed low nanomolar anti-proliferative activity. Mutation of the
reactive cysteine within CDK12, but not CDK13, significantly blunted these effects, with
CDK12 mutant cells exhibiting 11-fold resistance to MFH290 to WT CDK12 cells, which
was further enhanced to 25-fold as compared to WT when both CDK12 and CDK13 were
mutated (Figure 5B).
20
ACS Paragon Plus Environment

Page 21 of 59
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
CDK12 and CDK13 have been implicated in the regulation of RNA polymerase II (Pol II)-
mediated gene transcription by phosphorylating Pol II at conserved residues within the
19
CTD . Therefore, we asked whether MFH290 affected Pol II CTD phosphorylation levels
in our HAP1 isogenic lines. Indeed, MFH290 treatment reduced Pol II CTD
phosphorylation at serine residues 2, 5, 7 (Ser2, Ser5, Ser7), as well as overall Pol II
levels (Figure 5C). Mutation of CDK12 greatly restored levels of both total and
phosphorylated Pol II to near WT levels in the presence of MFH290, while mutation of
CDK13 exhibited a significantly reduced rescue. Complete rescue of Pol II CTD
phosphorylation was achieved by expressing both mutant CDK12 and CDK13 in HAP1
cells (Figure 5C). The difference in rescue demonstrated by mutation of CDK12
compared to that of CDK13 likely stems from the fact that CDK12 is the more highly
expressed and predominant homolog in HAP1 cells.
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
As CDK12 has been shown to specifically regulate the expression of DNA damage
20
response (DDR) genes , we treated HAP1 cells with MFH290 at 40 and 200 nM and
analyzed expression of DDR RNA and protein expression by RT qPCR and western blot.
We found that treatment with MFH290 reduced expression of RAD51 and RAD51C at
both the RNA and protein level (Figure 6A and 6B). These effects were almost completely
rescued at the protein level and largely rescued at the RNA level by expression of mutant
CDK12 alone, or in combination with C1017S CDK13.
21
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 22 of 59
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. (A) MFH290 treatment leads to a reduction in DNA damage response proteins.
Western blots of WT, CDK12 C1039S, CDK13 C1017S, and CDK12 C1039S/ CDK13
C1017S doubly-mutated HAP1 cells following treatment with MFH290 for 24 hours. (B)
MFH290 treatment leads to reduced expression of DNA damage response genes. RT-
qPCR of WT, CDK12 C1039S, CDK13 C1017S, and CDK12 C1039S/ CDK13 C1017S
doubly-mutated HAP1 cells following treatment with MFH290 for 6 hours. (C) Excess
over Bliss synergy plots for serial dilutions of MFH290 in combination with olaparib in
HAP1 cells. Excess over Bliss scores > 0 indicate drug synergy, whereas negative scores
indicate antagonism.
Impaired expression of RAD51 and other homologous recombination (HR)-regulating
21
proteins have been shown to sensitize cancer cells to PARP inhibitors . To ask if
MFH290 could do the same, we treated HAP1 cells with increasing concentrations of both
Olaparib, an FDA-approved PARP inhibitor, and MFH290. We examined the proliferation
22
ACS Paragon Plus Environment

Page 23 of 59
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
data for synergistic effects as evaluated by extrapolated Excess over Bliss score (0-1)22,
23.
Indeed MFH290 (25 nM) exhibited maximum synergy with Olaparib (1.2 M) with
excess over Bliss scores = 0.75 (Figure 6C).
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
These data indicate that CDK12 is the primary target of MFH290 in HAP1 cells, that
covalent targeting of CDK12 is obligatory for MFH290 activity, and that CDK12 inhibition
is largely responsible for the observed anti-proliferative effects. Consistent with
biochemical kinase profiling, these results indicate that MFH290 has exceptional cellular
selectivity for CDK12/13 (Table 1-3), supporting the notion that CDK12/13 inhibition is
responsible for inhibitor-induced effects on DDR gene expression.
In order to know whether MFH290 is suitable for an in vivo efficacy study, we then took
the compound for a pharmacokinetic study. As shown in the Table 4，MFH290 showed
a short half-life in mouse plasma and a high clearance, indicating further optimization will
be needed to improve its in vivo stability.
Table 4. Pharmacokinetic Parameters of MFH290
a
Fb
(%)
-
Route Dose
Tmax
C /C
AUClast
T1/2
CL
Vss
(mL/kg)
730
0
max
(mg/kg) (hr)
(ng/mL) (hr*ng/mL) (hr)
(mL/min/kg)
i.v.
i.p.
1
5
-
1312.89
162.22
393.63
412.44
0.17
0.15
-
102.03
0.08 885.05
0.25 556.46
-
-
-
-
49
25
p.o.
10
a
b
-
-
C back extrapolated conc. for i.v. group
0,
AUClast was considered for calculating bioavailability
Discussion and Conclusions
23
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 24 of 59
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
In conclusion, we have converted a pan-CDK inhibitor SNS032 to a selective CDK12/13
inhibitor MFH290 by hybridizing it with structural elements of the first selective CDK12/13
covalent inhibitor THZ531. Extensive structure-guided medicinal chemistry modifications
delivered a narrow SAR, with (R)-3-aminopiperidine emerging as a privileged scaffold,
enabling high binding affinity for CDK12/13 to the exclusion of other CDKs. MFH290
formed a covalent bond with Cys1039 on CDK12 and thus demonstrated a sustained
CDK12 inhibition-induced phenotype including strong anti-proliferative effect on cancer
cells, a transcriptional defect for DDR genes and a combinatorial effect with PARP
inhibition. Compared to first generation covalent inhibitor THZ531, a higher degree of
rescue from MFH290-induced phenotypes in HAP1 cells engineered with cysteine-to-
serine CDK12 and CDK13 mutations, implies that MFH290 possesses a superior
selectivity for CDK12/13. Further optimization of this scaffold to improve in vivo properties
will enable in vivo investigation of drug efficacy.
0
1
2
3
4
5
6
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
EXPERIMENTAL SECTION
Chemistry.
Unless otherwise noted, reagents and solvents were obtained from commercial suppliers
1
and were used without further purification. H NMR spectra were recorded on 500 MHz
(Bruker A500), and chemical shifts are reported in parts per million (ppm, δ) downfield
from tetramethylsilane (TMS). Coupling constants (J) are reported in Hz. Spin
multiplicities are described as s (singlet), br (broad singlet), d (doublet), t (triplet), q
(quartet), and m (multiplet). Mass spectra were obtained on a Waters Micromass ZQ
instrument. Preparative HPLC was performed on a Waters Sunfire C18 column (19 x 50
24
ACS Paragon Plus Environment

Page 25 of 59
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
mm, 5µM) using a gradient of 15-95% methanol in water or acetonitrile in water containing
0.05% trifluoroacetic acid (TFA) over 22 min (28 min run time) or 35 min (45 min run time)
at a flow rate of 20 mL/min. All of the final compounds reported had purities greater than
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
9
5% based on H nuclear magnetic resonance (NMR) and liquid chromatography−mass
spectrometry (LC−MS).
Synthesis of compound 1a.
To a solution of 5-(((5-(tert-butyl)oxazol-2-yl)methyl)thio)thiazol-2-amine (54 mg, 0.2
mmol) in DMF (1 mL) was added HATU (160 mg, 0.42 mmol), 3-acrylamidobenzoic acid
(60 mg, 0.31mmol) and DIEA (52 mg, 0.4 mmol). The resulting mixture was stirred at
room temperature for 1 hour, and then diluted with ethyl acetate (100 mL). The mixture
was washed with water and brine, dried over Na SO and concentrated. The residue was
2
4
purified by silica gel column (MeOH/DCM, 0-15%) to afford the compound 1a (7 mg, yield
8%) as a white solid.
1
3
-acrylamido-N-(5-(((5-(tert-butyl)oxazol-2-yl)methyl)thio)thiazol-2-yl)benzamide (1a). H
NMR (400 MHz, CDCl ) δ 12.36 (s, 1H), 9.08 (s, 1H), 8.24-7.99 (m, 1H), 7.90 (s, 1H),
3
7.63 (d, J = 7.6 Hz, 1H), 7.41 (t, J = 7.9 Hz, 1H), 6.95 (s, 1H), 6.55 (s, 1H), 6.44 (d, J =
16.8 Hz, 1H), 6.32 (dd, J = 16.8, 10.0 Hz, 1H), 5.76 (d, J = 10.0 Hz, 1H), 3.95 (s, 2H),
1
.23 (s, 9H). MS (ESI) m/z 443.12 [M+H]+.
Synthesis of compounds 1b and 1c.
To a solution of 5-(((5-(tert-butyl)oxazol-2-yl)methyl)thio)thiazol-2-amine (4) (0.69 mmol,
1 equiv), appropriate acid (0.94 mmol, 1.4 equiv), DMAP (0.35mmol, 0.5 equiv) and
triethylamine (1.4 mmol, 2 equiv) in DMF (1 mL) and DCM (2 ml) was added EDCI (1.4
25
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 26 of 59
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
mmol, 2 equiv) at room temperature. The reaction mixture was stirred for 1.5 hour, and
then diluted with ethyl acetate. The solution was washed with water and brine, dried over
Na SO and concentrated. The residue was purified by silica gel column and then treated
2
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
with 4M HCl ethyl acetate (3mL). The reaction was stirred at room temperature for 30
min. The mixture was concentrated under vacuum to afford the compounds 8 and 9 as
HCl salt.
To a solution of 8 (50 mg, 0.12 mmol) in DMF (1 mL) was added HATU (92 mg, 0.24
mmol), 4-acrylamidobenzoic acid (46 mg, 0.24 mmol) and DIEA (0.22 ml, 1.2 mmol). The
resulting mixture was stirred at room temperature for 1 h, and then diluted with ethyl
acetate (100 mL). The solution was washed with water and brine, dried over Na SO and
2
4
concentrated. The residue was purified by silica gel column (MeOH/DCM, 0-20%) to
afford the compound 1b as a white solid.
1-(4-acrylamidobenzoyl)-N-(5-(((5-(tert-butyl)oxazol-2-yl)methyl)thio)thiazol-2-
1
yl)piperidine-3-carboxamide (1b). H NMR (400 MHz, DMSO) δ 12.37 (s, 1H), 10.32 (s,
1H), 7.73 (d, J = 7.8 Hz, 2H), 7.55-7.22 (m, 3H), 6.71 (s, 1H), 6.44 (dd, J = 16.9, 10.3 Hz,
1H), 6.28 (d, J = 16.8 Hz, 1H), 5.78 (d, J = 9.9 Hz, 1H), 4.05 (s, 2H), 3.67-2.69 (m, 4H),
1
.97-1.23 (m, 5H), 1.12 (s, 9H). MS (ESI) m/ 554.21 [M+H]+.
To a solution of 9 (20 mg, 0.05 mmol) in THF (5 mL) was added DIEA (17 mg, 0.15 mmol)
o
and acryloyl chloride (43 mg, 0.05 mmol) at 0 C for 1 hour. The resulting mixture was
stirred at room temperature for 2 hours. The solution was concentrated and purified by
silica gel column (MeOH/DCM, 0-20%) to afford the compound 1c as a slightly white solid.
1-acryloyl-N-(5-(((5-(tert-butyl)oxazol-2-yl)methyl)thio)thiazol-2-yl)piperidine-4-
1
carboxamide (1c). H NMR (400 MHz, CDCl ) δ 11.64 (s, 1H), 7.29 (s, 1H), 6.61-6.55 (m,
3
26
ACS Paragon Plus Environment

Page 27 of 59
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
2
H), 6.28 (d, J = 16.7 Hz, 1H), 5.71 (d, J = 10.6 Hz, 1H), 4.60-4.57 (m, 1H), 4.12-3.96 (m,
H), 3.96 (s, 2H), 3.21-2.90 (m, 2H), 2.68 (t, J = 10.4 Hz, 1H), 2.01-1.82 (m, 4H), 1.24 (s,
1
9H). MS (ESI) m/z 435.15 [M+H]+.
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
Synthesis of compounds 1d-1f.
24
A mixture of 2-(((2-bromothiazol-5-yl)thio)methyl)-5-(tert-butyl)oxazole (10) (0.45 mmol,
1 equiv), appropriate amine (0.9 mmol, 2 equiv) and DIEA (1.35 mmol, 3 equiv) in NMP
(
3 mL) was heated to 140 °C overnight. The solution was diluted with water (20 mL) and
extracted with chloroform and iso-propanol (4:1). The organic phase was washed with
brine (50 mL × 2) and dried over Na SO , filtered and concentrated. The residue was
2
4
purified by silica gel column (MeOH/DCM, 0-20%) to give compound 11.
To a mixture of 11 (0.33 mmol, 1 equiv) in methanol (2 mL) was added 4N HCl/dioxane
(2 mL). The resulted solution was stirred at room temperature for 3 h. The mixture was
concentrated under reduced pressure to give the corresponding HCl salt. To the solution
of HCl salt (0.33 mmol, 1 equiv) in pyridine (2 mL) was added 4-nitrobenzoyl chloride
(0.66 mmol, 2 equiv), the reaction mixture was stirred at room temperature overnight. The
resulting mixture was concentrated under reduced pressure and the residue was purified
by silica gel column (MeOH/DCM, 0-20%) to give compound 13.
To a solution of 13 (0.54 mmol, 1 equiv) in ethyl acetate and methanol (1:1) were added
Tin (II) chloride dehydrate (5.4 mmol, 10 equiv) and conc. HCl (0.2 mL). After stirring for
3 h at 80 ˚C, the reaction mixture was diluted with chloroform and iso-propanol (4:1),
neutralized with saturated NaHCO and filtered. The filtrate was extracted with chloroform
3
and iso-propanol (4:1), concentrated under reduced pressure and the resulting residue
27
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 28 of 59
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
was purified by silica gel column chromatography (MeOH/DCM = 0-20%) to give the
compound 14.
To a solution of 14 (0.034 mmol, 1 equiv) and DIEA (0.1 mmol, 3 equiv) in MeCN (2 mL)
0
1
2
3
4
5
6
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
was added acryloyl chloride (0.044 mmol, 1.3 equiv) dropwise at 0 C. When the reaction
was completed, Na CO aq.(sat.) was added, the resulting mixture was extracted with
2
3
chloroform and iso-propanol (4:1) (20 mL x 2). The combined organic layer was
concentrated and purified with Prep-HPLC (MeOH/H O, 0.05% TFA) to give compounds
2
1d-1f.
(R)-N-(4-(3-((5-(((5-(tert-butyl)oxazol-2-yl)methyl)thio)thiazol-2-yl)amino)piperidine-1-
1
carbonyl)phenyl)acrylamide (1d, MFH290). H NMR (500 MHz, DMSO-d6) δ 10.27 (s,
1
H), 7.98 (s, 1H), 7.85 – 7.57 (m, 2H), 7.49 – 7.27 (m, 2H), 7.15 – 6.81 (m, 1H), 6.71 (s,
H), 6.44 (dd, J = 17.0, 10.2 Hz, 1H), 6.28 (dd, J = 17.0, 2.0 Hz, 1H), 5.78 (dd, J = 10.1,
1
2.0 Hz, 1H), 3.94 (s, 2H), 3.86 – 3.61 (m, 2H), 3.53 – 2.77 (m, 3H), 2.02 – 1.40 (m, 4H),
.18 (s, 9H). MS (ESI) m/z 526.50 [M + H]+;
1
(S)-N-(4-(3-((5-(((5-(tert-butyl)oxazol-2-yl)methyl)thio)thiazol-2-yl)amino)piperidine-1-
1
carbonyl)phenyl)acrylamide (1e). H NMR (500 MHz, DMSO-d ) δ 10.31 (s, 1H), 8.08 (s,
6
1H), 7.80 – 7.60 (m, 2H), 7.46 – 7.28 (m, 2H), 7.00 – 6.82 (m, 1H), 6.71 (s, 1H), 6.43 (dd,
J = 17.0, 10.1 Hz, 1H), 6.27 (dd, J = 17.0, 1.9 Hz, 1H), 5.78 (dd, J = 10.1, 1.9 Hz, 1H),
3.94 (s, 2H), 3.73 – 3.63 (m, 2H), 3.34 – 3.21 (m, 1H), 3.15 – 2.70 (m, 2H), 2.03 – 1.41
(
m, 4H), 1.17 (s, 9H). MS (ESI) m/z 526.50 [M + H]+;
N-(4-((3-((5-(((5-(tert-butyl)oxazol-2-yl)methyl)thio)thiazol-2-
1
yl)amino)cyclohexyl)amino)phenyl)acrylamide (1f). H NMR (500 MHz, DMSO-d6) δ
10.37 – 10.01 (m, 1H), 8.19 – 7.96 (m, 1H), 7.79 – 7.56 (m, 2H), 7.30 – 7.02 (m, 2H), 6.98
28
ACS Paragon Plus Environment

Page 29 of 59
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
–
6.88 (m, 1H), 6.75 – 6.66 (m, 1H), 6.49 – 6.36 (m, 1H), 6.30 – 6.19 (m, 1H), 5.81 – 5.68
(m, 1H), 4.03 (s, 1H), 3.96 – 3.92 (m, 2H), 2.28 – 2.17 (m, 1H), 2.00 – 1.85 (m, 2H), 1.85
– 1.74 (m, 1H), 1.74 – 1.49 (m, 2H), 1.44 – 1.22 (m, 3H), 1.22 – 1.16 (m, 9H), 1.15 – 1.00
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
m, 1H). MS (ESI) m/z 511.96 [M + H]+;
(
Synthesis of compounds 1g-1i, 1k and 1l.
To a mixture of compound 11 (0.024 mmol, 1 equiv) in methanol (2 mL) was added 4N
HCl/dioxane (2 mL). The resulted solution was stirred at rt for 3 h. The mixture was
concentrated under reduced pressure to give the corresponding HCl salt. To a mixture of
the HCl salt (0.024 mmol, 1 equiv), 12 (0.03 mmol, 1.3 equiv) and DIEA (0.12 mmol, 5
equiv) in DCM (2 mL) was added T P (50% solution in ethyl acetate, 46 mg, 0.07 mmol)
3
dropwise at room temperature. The reaction mixture was stirred at room temperature
overnight. The solution was diluted with DCM (20 mL) and washed with brine (50 mL × 2)
and dried over Na SO , filtered and concentrated. The residue was purified by Prep-
2
4
HPLC (MeOH/H O, 0.05% TFA) to give the compounds 1g-1i, 1k and 1l.
2
N-(4-((2R,3R)-3-((5-(((5-(tert-butyl)oxazol-2-yl)methyl)thio)thiazol-2-yl)amino)-2-
1
methylpiperidine-1-carbonyl)phenyl)acrylamide (1g). H NMR (500 MHz, DMSO-d ) δ
6
10.30 (s, 1H), 8.02 (s, 1H), 7.72 (d, J = 8.7 Hz, 2H), 7.38 (d, J = 8.2 Hz, 2H), 7.00 – 6.62
(m, 2H), 6.45 (dd, J = 17.0, 10.1 Hz, 1H), 6.28 (dd, J = 17.0, 2.0 Hz, 1H), 5.79 (dd, J =
10.1, 2.0 Hz, 1H), 3.94 (s, 2H), 3.86 – 3.70 (m, 2H), 3.14 – 2.57 (m, 2H), 1.81 – 1.44 (m,
4
H), 1.17 (s, 9H), 1.04 (d, J = 6.9 Hz, 3H). MS (ESI) m/z 540.58 [M + H]+;
N-(4-(5-((5-(((5-(tert-butyl)oxazol-2-yl)methyl)thio)thiazol-2-yl)amino)-3,3-
1
difluoropiperidine-1-carbonyl)phenyl)acrylamide (1h). H NMR (500 MHz, DMSO-d ) δ
6
10.35 (s, 1H), 8.20 – 8.04 (m, 1H), 7.74 (d, J = 8.3 Hz, 2H), 7.42 (d, J = 8.5 Hz, 2H), 7.05
29
ACS Paragon Plus Environment