Discovery of SK-575 as a Highly Potent and Efficacious Proteolysis-Targeting Chimera Degrader of PARP1 for Treating Cancers

Article abstract of DOI:10.1021/acs.jmedchem.0c00821

The nuclear protein poly(ADP-ribose) polymerase-1 (PARP1) has a well-established role in the signaling and repair of DNA and is a validated therapeutic target for cancers and other human diseases. Here, we have designed, synthesized, and evaluated a series of small-molecule PARP1 degraders based on the proteolysis-targeting chimera (PROTAC) concept. Our efforts have led to the discovery of highly potent PARP1 degraders, as exemplified by compound 18 (SK-575). SK-575 potently inhibits the growth of cancer cells bearing BRCA1/2 mutations and induces potent and specific degradation of PARP1 in various human cancer cells even at low picomolar concentrations. SK-575 achieves durable tumor growth inhibition in mice when used as a single agent or in combination with cytotoxic agents, such as temozolomide and cisplatin. These data demonstrate that SK-575 is a highly potent and efficacious PARP1 degrader.

Full text of DOI:10.1021/acs.jmedchem.0c00821

Subscriber access provided by CORNELL UNIVERSITY LIBRARY
Article
Discovery of SK-575 as a Highly Potent and Efficacious Proteolysis
Targeting Chimera (PROTAC) Degrader of PARP1 for Treating Cancers
Chao-Guo Cao, Jie Yang, Yong Chen, Pei-Ting Zhou, Ying-
Wei Wang, Wu Du, Li-Feng Zhao, and Yuanwei Chen
J. Med. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.jmedchem.0c00821 • Publication Date (Web): 14 Sep 2020
Downloaded from pubs.acs.org on September 14, 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 68
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
Discovery of SK-575 as a Highly Potent and Efficacious Proteolysis Targeting
Chimera (PROTAC) Degrader of PARP1 for Treating Cancers
Chaoguo Cao,† Jie Yang,†,∇ Yong Chen,†,∇ Peiting Zhou, Yingwei Wang, Wu Du, Lifeng Zhao,* and
,∇
†
†
§
,‡
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
Yuanwei Chen*,†,§
†State Key Laboratory of Biotherapy, Collaborative Innovation Center of Biotherapy and Cancer Center,
West China Hospital of Sichuan University, Chengdu, 610041, China
‡Sichuan Industrial Institute of Antibiotics, Chengdu University, Chengdu 610052, China
§Hinova Pharmaceuticals Inc, 4th Floor, Building RongYao A, No. 5, Keyuan South Road, Chengdu,
610041, China
ABSTRACT:
The nuclear protein poly(ADP-ribose) polymerase-1 (PARP1) has a well-established role in the
signaling and repair of DNA, and is a validated therapeutic target for cancers and other human diseases.
Here, we have designed, synthesized, and evaluated a series of small-molecule PARP1 degraders based
on the proteolysis targeting chimera (PROTAC) concept. Our efforts have led to the discovery of highly
potent PARP1 degraders, as exemplified by compound 18 (SK-575). SK-575 potently inhibits the growth
of cancer cells bearing BRCA1/2 mutations and induces potent and specific degradation of PARP1 in
various human cancer cells even at low picomolar concentrations. SK-575 achieves durable tumor growth
1
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 2 of 68
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
inhibition in mice when used as a single-agent or in combination with cytotoxic agents, such as
temozolomide (TMZ) and cisplatin. These data demonstrate that SK-575 is a highly potent and efficacious
PARP1 degrader.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
2
ACS Paragon Plus Environment

Page 3 of 68
1
Journal of Medicinal Chemistry
INTRODUCTION
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
Poly(ADP-ribose) polymerase-1 (PARP1) is the most abundant and well-characterized member of
poly(ADP-ribose) polymerase (PARP) family of enzymes.1-3 As a DNA-dependent nuclear enzyme,
2
PARP1 plays a pivotal role in the signaling and repair of DNA damage. Following DNA damage caused
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
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
by the cell’s own metabolic, chemical poisons or ionizing radiation, PARP1 is rapidly activated, then
recognizes and binds to DNA single-strand breaks (SSBs), where it catalyzes the fracture of nicotinamide
+
adenine dinucleotide (NAD ) to ADP-ribose and nicotinamide, and then transfers the ADP-ribose units
to synthesize poly(ADP-ribose) (PAR) chains either on itself or on several acceptor proteins.5-7 These
PAR chains with more negative charges and larger steric hindrance can make PARP1 dissociate from
DNA fracture by reducing the affinity between PARP1 and DNA, and then guide DNA repair factors to
combine with DNA gap to repair the damaged part.2,8,9 This is the key process known as base excision
repair (BRE) pathway for DNA repair.2,5
The PARP1 knockout cancer cells and animals exhibit a high degree of sensitivity to radiation or
cytotoxic agents.10,11 Initially, PARP1 inhibitors (PARPi) were developed as promising co-adjuvant
components of standard chemo- and radiotherapy for tumor therapy.12-14 Cancer cells bearing mutations
in BRCA-1 or -2 are exquisitely sensitive to PARPi, which is called ‘synthetic lethality’.15-17 In this regard,
1
8
19
20
21
many small-molecule PARPi, such as olaparib, rucarparib, niraparib, and talazoparib, have been
22
developed in different stages of clinical trials. For instance, olaparib as the first oral potent PARP1/2
inhibitor, have been approved for the treatment of BRCA-mutation cancer in the clinic, especially for
2
3
ovarian and breast cancers. However, there are still challenges in terms of PARPi usage that limit their
therapeutic utility, including the acquisition of drug resistance, the low proportion of BRCA-1 or -2
3
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 4 of 68
1
5
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
mutations in cancer cells. In addition to the inhibition of PARP1 enzyme activity, the mechanism of
action of PARPi as antitumor agents is via PARP1 trapping.24,25 In short, PARPi prevent PARP1 protein
from dissociating from DNA breaks to form PARP1-PARPi-DNA complexes, which could block and
damage DNA replication, and is therefore more toxic than merely unrepaired SSBs due to the inactivation
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
2
6
of PARP1. Thus, PARPi can kill tumor cells and are also toxic to normal cells. However, we are unable
to assess the relative contribution of these two pathways (PARP1 catalytic inhibition and trapping) to
PARP inhibitor mediated cytotoxicity.
Besides the role in cancer cells, the overactivation of PARP1 has been associated with several
pathophysiological conditions, such as neurological diseases (e.g., stroke, neurodegeneration),27,28
2
9
ischemia-reperfusion injury (e.g., myocardial infarction), severe acute critical illness (e.g., shock, acute
lung injury, acute liver failure)28,30,31 and inflammatory diseases . In these non-oncological diseases, the
32
+
overstimulation of PARP1 with the catalytic activity results in the cellular depletion of NAD and forces
+
the activation of the salvage synthesis pathway. NAD depletion leads to ATP depletion, resulting in
3
3
cellular energetic deficit and cell necrotic death. Moreover, the overactivation of PARP1 leads to the
accumulation of free PAR polymers to reach the cytotoxic level, which induces mitochondria release of
3
4
apoptosis-inducing factors rapidly, leading to cell death termed parthanatos. Several kinds of PARPi
have been proved to be effective in non-oncological indications. However, their usage is still cautious,
mainly due to their PARP1 trapping activity and mechanism-independent cytotoxicity.24,35 Consequently,
developing a practical strategy, by blocking both the catalytic and scaffolding functions of PARP1 without
causing PARP1 trapping, is urgent and meaningful for PARP1-related diseases.
3
6
Proteolysis targeting chimera (PROTAC) concept can be dated back to 2001. PROTAC molecules
4
ACS Paragon Plus Environment

Page 5 of 68
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
are heterobifunctional small molecule containing two ligands, which specifically bind to the target protein
and E3 ubiquitin ligase, respectively, tethered together by a linker. These molecules are capable of driving
the target proteins to E3 ubiquitin ligase, which results in the formation of the ternary complex and leads
3
7
to ubiquitination of the target proteins and consequent proteasome-mediated degradation. In the past
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
decade, PROTAC strategy has gained great progress in inducing target protein degradation in vitro and
in vivo, such as BET,38-42 AR, ER, ALK, and CDK9,
43
44
45
46-48
which are attributed to the discovery of
several highly potent small-molecule ligands for E3 ubiquitin ligase. Several PARP1/2 degraders have
49
50
50
been reported, including compound 1, 2 (iRucaparib-AP5), 3 (iRucaparib-AP6), and 4 (iVeliparib-
5
0
AP6) (Figure 1). Compound 1, based on the niraparib derivative and nutlin-3 derivative, specifically
induced the PARP1 cleavage and cell apoptosis in the MDA-MB-231 cells. In comparison, two of
compounds, 2 and 3, prepared by using different lengths of all polyethylene glycol (PEG) linkers, were
reported as potent and efficacious PARP1 degraders with half-maximal degrading concentration (DC )
5
0
of 36 nM and 82 nM in primary rat neonatal cardiomyocytes, respectively. However, these compounds
could not completely degrade the intracellular PARP1 protein, and the remaining part can still play a part
of catalytic role, thus limiting their therapeutic efficacy. Besides, compound 4 emerged as a potent PARP2
degrader, which was designed using veliparib as the PARP2 ligand and thalidomide as a cereblon (CRBN)
ligand.
5
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 6 of 68
Cl
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
Cl
NH2
O
N
N
H
N
N
N
N
N
( O
)4
N
N
N
O
O
O
O
1
O
O
H
N
O
O
NH2
HN
N
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
F
O
O
N
NH
N
(
)6
O
O
O
N
H
NH
H
N
n
N
H
N
O
N
O
(O
)
O
2
3
(iRucaparib-AP5) n=5
(iRucaparib-AP6) n=6
4
(iVeliparib-AP6)
Figure 1. Chemical structure of three previously reported PARP1 degraders 1, 2, and 3, and one previously reported
PARP2 degraders 4.
In this study, we describe the design, synthesis, and evaluation of PARP1-PROTAC degraders based
upon a PAPR1 inhibitor olaparib and thalidomide/lenalidomide as ligands for cereblon/Cullin 4A. Our
study has resulted in the discovery of highly potent PARP1-PROTAC degraders (hereafter called PARP1
degraders), exemplified by compound 18 (SK-575). SK-575 is capable of effectively inducing PARP1
protein degradation with the DC at the scale of picomolar (pM) and maximum degradation (Dmax) > 95%
5
0
in PARP1-positive cancer cell lines, and achieving durable tumor growth inhibition in mice at well-
tolerated dose schedules. SK-575 is a highly potent and specific PARP1 degrader and warrants extensive
evaluation for the treatment of cancers and other diseases.
RESULTS AND DISCUSSION
6
ACS Paragon Plus Environment

Page 7 of 68
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 2. Design of PROTACs for PARP1. (A) Chemical structures of PARP1/2 inhibitor 5 (olaparib). (B) Crystal
structure of 5 bound to the PARP1 catalytic domain (PDB ID: 5DS3). The possible linker tethering positions are shown
in red arrows.
Design of PARP1 PROTAC Based on Olaparib. To explore the potential PARP1 degraders, we
chose olaparib as the PARP1 ligand in our design for its excellent activity against PARP1/2 (Figure 2A).18
According to the cocrystal structure of olaparib in the complex with PARP1 (Figure 2B), the amide moiety
5
1
on olaparib is essential for binding to PARP1. The cyclopropyl ring locates close to the opening of the
ligand-binding pocket and thus may represent a suitable position to link with CRBN ligands without losing
binding affinity dramatically. Based on the above design principles, we designed and synthesized a series
of PARP1 targeting PROTACs using olaparib as the PARP1 ligand, diverse linkers, and
thalidomide/lenalidomide as the CRBN ligands.
Determination of the Optimal Linker Length and Composition. It has been confirmed that a series
of PARP1 degraders that conjugate olaparib to thalidomide with PEG unit fails to induce PARP1
degradation (see in Figure S1 in Supporting information).49,50 We speculated that the length and the
composition of the linker may be underexplored, and an optimized linker may induce favorable interaction
between PARP1 and CRBN E3 ligase. Thus we have performed extensive optimization of the chemical
7
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 8 of 68
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
composition of the linker with different lengths. A series of PARP1 degraders, compounds 6-11 (Table
), with different amino acid (e.g., using 4-aminobutyric acid and/or 6-aminocaproic acid) linkers were
1
synthesized by a condensation reaction. Their growth inhibitory effect against BRCA1-deficient MDA-
MB-436 and BRCA2-deficient Capan-1 cell lines, and the IC values (half-maximal growth inhibitory
5
0
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
concentration) were then evaluated in a cell growth assay. Compound 11 has IC values of 0.95 μM in
5
0
MDA-MB-436 and 1.5 μM in Capan-1, which were 52 and 56 times less potent than olaparib with IC50
values of 18 nM and 56 nM in these two cell lines, respectively. Next, changing the NH group in 10 to
oxygen atom yielded 12 (Table 1), which showed IC values of 2.2 μM in MDA-MB-436 and 7.1 μM in
5
0
Capan-1, and is less potent than 10. The similar conversion of the NH group in 11 to an oxygen atom
generated 13, which has IC values of 1.7 μM and 2.2 μM in the inhibition of cell growth in MDA-MB-
5
0
436 and Capan-1 cells, respectively, and is less potent than 11. These results demonstrate that the amino
acid linkers have an adverse influence on the cellular potency of the resulting PROTAC molecules. Thus,
we identified that PROTAC molecules conjugating olaparib and CRBN ligand with amino acid linkers
could not benefit for protein-protein interactions (PPIs) between CRBN E3 ligase and PARP1, cell
penetration, and so on.
We next sought to investigate the optimal linkers and synthesized compounds 14-16, 18, and 20 that
incorporated alkyl dicarboxylic acid of different lengths and ethylenediamine (Table 1). Surprisingly,
compound 18, which contained dodecarboxylic acid, displayed similar potency to olaparib in the cell
growth inhibition, achieved IC values of 19 nM in MDA-MB-436 cells, and 56 nM in Capan-1 cells.
5
0
Furthermore, changing the linker in 18 by decreasing or increasing one additional methylene group
resulted in 17 or 19, respectively, which are less potent than 18 in MDA-MB-436 and Capan-1 cells.
8
ACS Paragon Plus Environment

Page 9 of 68
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
Table 1. Optimization of Linker Length and Composition of PARP1 Degraders
O
O
HN
NH
N
O
O
O
N
N
Linker
O
N
F
O
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
a
IC (μM) in cell growth inhibition
5
0
Compound
Linker
MDA-MB-436
0.018 ± 0.003
> 10
Capan-1
olaparib
-
0.027 ± 0.005
> 10
O
O
6
7
N
H
O
O
N
9.4 ± 3.1
> 10
H
H
N
O
O
8
9
N
5.3 ± 1.9
1.9 ± 1.1
> 10
O
H
H
N
O
O
N
H
8.7 ± 2.9
3.1 ± 1.2
1.5 ± 0.4
O
H
N
O
O
1
1
0
1
N
H
1.2 ± 0.7
O
H
N
O
O
N
H
0.95 ± 0.22
O
H
N
O
H
N
1
1
1
1
1
1
2
3
4
5
6
7
N
H
2.2 ± 1.3
1.7 ± 0.7
7.1 ± 2.7
2.2 ± 0.9
O
H
N
O
H
N
N
H
O
O
H
N
N
> 10
> 10
H
O
H
N
N
0.83 ± 0.26
0.123 ± 0.071
0.029 ± 0.008
0.019 ± 0.006
0.021 ± 0.003
0.025 ± 0.004
4.1 ± 1.8
H
O
H
N
N
H
0.495 ± 0.089
0.264 ± 0.046
0.056 ± 0.012
0.087 ± 0.019
0.118 ± 0.021
O
H
N
N
H
O
H
N
1
8 (SK-575)
N
H
O
H
N
1
2
9
0
N
H
O
H
N
N
H
a
IC values were obtained from three independent experiments (Mean ± SD).
5
0
Encouraged by the high potency of compound 18 in the inhibition of cell growth, we next synthesized
9
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 10 of 68
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
compounds 21 and 22 converting the -NH(CH ) NH- group in 18 to a -NH(CH ) NH- or -NH(CH ) NH-
2 2 2 4 2 6
group to further improve potency (Table 2). Compounds 21 and 22, however, which were ~2 times less
potent than compound 18 in cell growth inhibition, achieved IC values of 39 nM and 42 nM in MDA-
5
0
MB-436 cells, and 131 nM and 110 nM in Capan-1 cells, respectively. Furthermore, considering the linker
length and chemical composition were important for PARP1 degraders to effectively inhibit cell growth,
we replaced the -NH(CH ) NH- group in 18 with a different length linker contained oxygen atom, and
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
2
2
generated compounds 23-25 (Table 2). These compounds turned out to be less potent than compound 18
in cell growth inhibition. Consequently, these data thus established that a linker length of 15-16 atoms, as
in compound 18 or 19, is optimal for these PAPR1 degraders for inhibition of cell growth.
Table 2. Further Optimization of Linker Length and Composition
O
O
HN
NH
O
N
O
N
O
O
N
O
N
Linker
F
O
a
IC (nM) in cell growth inhibition
5
0
Compound
Linker
MDA-MB-436
19 ± 6
Capan-1
56 ± 12
H
N
N
1
8 (SK-575)
H
H
N
2
2
2
1
2
3
N
39 ± 12
131 ± 34
110 ± 13
120 ± 19
134 ± 27
154 ± 39
H
H
N
N
H
42 ± 12
O
N
H
N
H
47 ± 18
H
O
N
24
25
N
O
64 ± 21
H
O
N
H
O
O
N
H
65 ± 23
a
IC values were obtained from three independent experiments (Mean ± SD).
5
0
1
0
ACS Paragon Plus Environment

Page 11 of 68
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
We further investigated the protein degradation activity of the resulting PARP1 degraders with an IC50
200 nM in MDA-MD-436 cells, and the protein level of PARP1 was analyzed through western blotting
<
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
assay at four concentrations (1000, 100, 10 and 1 nM) in MDA-MB-436 and Capan-1 cell lines with 24
h treatment. As shown in Figure 3, compound 18 showed the best degradation activity at four
concentrations in these two cancer cell lines, which is consistent with its cell growth inhibitory activity.
In comparison, the corresponding PARP1 inhibitor, olaparib, fails to reduce the protein level of PARP1
at all concentrations tested (1-1000 nM). Compounds 20 and 21 with the same length linker have similar
IC values in inhibition of cell growth, but they reveal different degradation activities. It may be that
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
5
0
changes the position of the amide bond in linker influences the formation of special conformations. It is
noteworthy that compound 21 effectively reduces the PARP1 protein by 70% at 10 nM and by 90.5% at
1
00 nM in MDA-MB-436 cell lines. Interestingly, compound 21 reduces PARP1 protein by 45% at 1 μM,
less than that at 100 nM, probably due to the “hook” effect that has been observed previously for PROTAC
degraders.43,44,52
1
1
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 12 of 68
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. Western blotting analysis of PARP1 protein in MDA-MB-436 (A) and Capan-1 (B) cells treated with
PARP1 inhibitor OLA (olaparib), PARP1 degraders 16-27, and a control compound 28. Cells were treated with each
compound at 1, 10, 100, and 1000 nM for 24 h, and β-Tubulin was used as the loading control. The fold under the strip
represents the PARP1/β-Tubulin ratio normalized with the DMSO control at 100.
Examination of the CRBN Ligand Portion in Compound 18. On the basis of 18, we designed and
synthesized compound 26 (Table 3) that tethered the phenyl ring of thalidomide from different positions.53
1
2
ACS Paragon Plus Environment

Page 13 of 68
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
Compound 26 has IC values of 35 nM and 159 nM in inhibition of cell growth in MDA-MB-436 and
50
Capan-1 cell lines, respectively, which is 2-3 times less potent than 18. In addition, converting the 3-
carbonyl group of thalidomide moiety in 18 to a methylene group generated 27, which inhibits cell growth
in MDA-MB-436 and Capan-1 cells with IC values of 92 nM and 194 nM, respectively, is 4-5 times
5
0
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
less potent than compound 18. Our western blotting analysis showed that 26 and 27 effectively induced
PARP1 degradation, but their activities were weaker than 18 in these two cell lines mentioned above.
Table 3. Further Optimization of CRBN Ligand Portion in Compound 18
O
NH
N
O
N
O
H
N
N
F
N
Cereblon Ligand
H
O
a
IC (nM) in cell growth inhibition
5
0
Compound
Cereblon Ligand
MDA-MB-436
19 ± 6
Capan-1
56 ± 12
O
O
NH
1
8 (SK-575)
N
O
O
O
O
NH
2
2
6
7
N
O
35 ± 15
92 ± 23
159 ± 35
194 ± 31
O
O
NH
N
O
O
a
IC values were obtained from three independent experiments (Mean ± SD).
5
0
Design and Synthesis of a Control Compound for 18. Revealing by the cocrystal structures, the
amino group of the piperidine-2,6-dione in thalidomide forms a strong hydrogen bond with CRBN,54,55
and methylation of the amino group of the piperidine-2,6-dione in thalidomide completely abrogates the
1
3
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 14 of 68
4
2
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
binding to CRBN. We, therefore, synthesized compound 28 by methylation of the amino group of the
piperidine-2,6-dione in compound 18 as an additional control compound (Figure 4). Compound 28 is
completely ineffective in inducing PARP1 degradation in MDA-MB-436 and Capan-1 cells at
concentrations up to 1 μM (Figure 3). These data clearly suggest that the induced degradation of PARP1
by compound 18 is cereblon/cullin-4A ligase-dependent, consistent with the PROTAC design.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
O
O
O
O
NH
N
NH
O
NH
N
NH
O
O
F
O
O
F
O
NH
N
Me
NH
N
NH
N
N
N
O
O
N
O
N
O
O
O
1
8 (SK-575)
28
Figure 4. Design of 28 as a control compound.
Examination of the VHL Ligand Portion in Compound 18. Like CRBN ligands (such as thalidomide
and its analogues), VHL ligands have been successfully used for the design of PROTAC degraders.39,41,
4
3-44
Therefore, on the basis of 18, we designed and synthesized compounds 29-32 using a VHL ligands
(
Table 4). Compound 29 has IC values of 54 nM and 87 nM in inhibition of cell growth in MDA-MB-
5
0
4
36 and Capan-1 cell lines, respectively, which were 2-3 times less potent than 18. After changing the
linker length, the cell proliferation inhibitory activity of these compounds decreased when compared with
29. Our western blotting analysis showed that compounds 29-32 have only a modest effect in reducing
the level of PARP1 protein at 1 μM concentrations in the MDA-MB-436 cell lines (Figure S2). Taken
together, these data indicated that these compounds mainly work as inhibitors, especially compound 29,
but not as PARP1 degraders. More studies are needed to determine if the VHL/cullin 2 E3 ligase system
can be used for the design of highly potent and effective PARP1 degraders.
1
4
ACS Paragon Plus Environment

Page 15 of 68
Journal of Medicinal Chemistry
1
2
3
Table 4. Further Optimization of VHL Ligand Portion in Compound 18
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
NH
N
OH
O
O
F
N
O
S
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
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
N
Linker
NH
N
O
O
a
IC (μM) in cell growth inhibition
50
Compound
Linker
MDA-MB-436
0.019 ± 0.006
0.054 ± 0.011
Capan-1
18 (SK-575)
-
0.056 ± 0.012
0.087 ± 0.023
2
9
N
H
H
N
30
N
H
1.2 ± 0.34
0.68 ± 0.11
0.12 ± 0.021
1.6 ± 0.39
0.87 ± 0.13
0.17 ± 0.033
O
H
N
3
3
2
2
N
H
O
H
N
N
H
O
3
(iRucaparib-
AP6)
-
0.71 ± 0.12
0.95 ± 0.22
a
IC values were obtained from three independent experiments (Mean ± SD).
5
0
In direct comparison, the previously reported PARP1 degrader, compound 3 (iRucaparib-AP6), has
IC values of 0.71 μM and 0.95 μM in inhibition of cell growth in MDA-MB-436 and Capan-1 cell lines,
50
respectively, which are 36 and 17 times less potent than 18 (Table 4). Western blotting data showed that
compound 18 are highly potent and effective in reducing PARP1 protein at 0.01-10 μM with a 12 h
treatment in MDA-MB-436 and SW620 cell lines (Figure 5). However, compound 3 exhibited only
modest effect in reducing the level of PARP1 protein at the same conditions.
1
5
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 16 of 68
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Figure 5. Western blotting analysis of PARP1 protein in MDA-MB-436 (A) and SW620 (B) cells treated with
compounds 3 and 18. Cells were treated with each compound at 1, 10, 100, 1000 nM, and 10 μM for 12 h, and β-Tubulin
was used as the loading control.
Although compound 18 effectively induces degradation of the PARP1 protein at concentrations as low
as 1-10 nM with 24 h treatment in MDA-MB-436 and Capan-1 cell lines, it has little or no effect on
inducing degradation of the PARP2 protein in these two cell lines at concentrations up to 1 μM (Figure
S3). Similarly, olaparib and compound 28 have no effect on the level of PARP2 protein. Hence, our data
demonstrate that compound 18 is an efficient and specific PARP1 degrader. Based on the data of cell
growth inhibition activity and degradation activity of PARP1, compound 18 appears to be the most potent
PARP1 degrader among these resulting compounds. Consequently, we pursued an extensive investigation
of compound 18 (SK-575).
Evaluation of Degradation Activity of SK-575 in Cancer Cells. We further evaluated SK-575 in
MDA-MB-436, Capan-1, and SW620 cells for its potency in inducing PARP1 protein degradation with a
24 h treatment time in a wide range of concentrations (Figure 6A and S4). Quantification of our western
1
6
ACS Paragon Plus Environment

Page 17 of 68
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
blotting data showed that SK-575 achieves DC values of 1.26 nM and 0.509 nM in the MDA-MB-436
50
and SW620 cells, respectively, and > 99% PARP1 degradation at 10 nM in these cell lines. SK-575, with
a DC value of 6.72 nM, achieves near-complete degradation at 100 nM in Capan-1 cells.
5
0
We next evaluated the kinetics of SK-575 in the induction of PARP1 degradation at 30 nM in MDA-
MB-436 and SW620 cell lines. Our data showed that SK-575 effectively reduces the PARP1 protein level
within 1 h and achieves near-complete PARP1 depletion at the 2 h time-point, indicating fast kinetics
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
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 6B). The degradation activity of SK-575 in Capan-1 cell lines is weaker than that in MDA-MB-
4
36 and SW620 cell lines, and it can reach the maximum degradation in 6 h at 100 nM.
To assess the durability of PARP1 protein degradation, we treated MDA-MB-436 cells with SK-575 at
00 nM for 5 days, illustrating that it induced sustained depletion of PARP1(Figure 6C). In another
1
experiment, we treated MDA-MB-436 cells with SK-575 for 2 h at 100 nM and collected cells at various
time points after compound washout. Our data showed that the level of PARP1 protein degrader-treated
cells remained below the level in vehicle-treated cells, even 24 h after washout (Figure 6D).
To further assess the efficiency and broad applicability of SK-575, six different cancer cell lines were
used to evaluate its potency in inducing PARP1 protein degradation with a 24 h treatment time.
Surprisingly, besides the LNCaP cells achieved DC value at the nanomolar level (DC = 1.08 nM), the
5
0
50
DC of all the tested cells were at the scale of picomolar after 24 h treatment (Figure 8A and Figure S4
5
0
in Supporting Information). The DC showed 384 pM in HCC1937, 512 pM in 22Rv1, 263 pM in MDA-
5
0
MB-468, 578 pM in PC-3, and 228 pM in MDA-MB-231. It is noteworthy that SK-575 can reduce the
PARP1 protein level by > 99% with a 24 h treatment time. At the same time, we also observed the rapid
degradation of PARP1 protein in all of these cells (Figure S5 in Supporting Information). Taken together,
1
7
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 18 of 68
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
our data demonstrated that SK-575 not only displays the fast and efficient degradation of PARP1 but also
has a broad application.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
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. Further characterization of SK-575. (A) PARP1 degradation at the indicated dose of SK-575 with a 24 h
treatment time in MDA-MB-436, Capan-1, and SW620 cells (left). PARP1 protein was examined by western blotting,
and the PARP1 protein level was quantified and normalized to the corresponding density of β-Tubulin protein (right).
1
8
ACS Paragon Plus Environment

Page 19 of 68
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
DC are calculated by two biologically independent experiments. (B) Time course of PARP1 degradation by SK-575 in
50
MDA-MB-436 (30 nM), Capan-1 (100 nM), and SW620 (30 nM) cells. (C) Time course of PARP1 degradation by SK-
75 in MDA-MB-436 (100 nM) cells. (D) Cellular PARP1 recovery levels after the washout of SK-575. Cells were pre-
5
treated with SK-575 for 2 h and washout the compound for the indicated times.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Study on the Mechanism of SK-575. We investigated the mechanism of PARP1 degradation induced
by SK-575 in MDA-MB-436 and SW620 cell lines, obtaining the results shown in Figure 7. The data
showed that the degradation of PARP1 protein induced by SK-575 could be effectively blocked by
pretreatment with a PARP1 inhibitor (olaparib) or a CRBN ligand (pomalidomide), in a dose-dependent
manner for 24 treatment. These results supported that the high potency of SK-575 as a PARP1 degrader
is dependent on binding to PARP1 and CRBN. Furthermore, a combination of proteasome inhibitor
(MG132 or carfilzomib) or a NEDD8-activating enzyme (NAE) inhibitor (MLN4924) and SK-575
treatment completely blocked PARP1 degradation induced by SK-575, indicating that the fast and
efficient PARP1 protein degradation by SK-575 was based on a NAE-mediated and proteasome-
dependent mechanism. Taken together, these mechanistic data provided clear evidence that SK-575 is a
bona fide PROTAC PARP1 degrader.
1
9
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 20 of 68
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 7. Western blotting analysis of PARP1 protein in MDA-MB-436 (A) and SW620 (B) cells after a 2 h pretreatment
with PARP1 inhibitor OLA (olaparib), CRBN ligand POM (pomalidomide), proteasome inhibitor (MG132 or
carfilzomib) or NEDD8-activating E1 enzyme inhibitor (MLN4924). Cells were treated with individual compound at
indicated concentrations, following by a 24 h treatment with SK-575 at 30 nM, and β-Tubulin was used as the loading
control.
Evaluation Activity of SK-575 in vitro. Using a PARP1 activity assay, we confirmed that SK-575
and its methylation counterpart 28 still strongly bound to PARP1, with IC values of 2.30 nM and 2.64
50
nM, respectively, which are equivalent to their parental inhibitor olaparib with an IC value of 3.35 nM
5
0
(Figure 8B). Next, we further examined the ability of SK-575 and olaparib to suppress cell growth
inhibitory of HR-deficient cell lines containing BRCA1/2 or PTEN, and HR-proficient cancer cell lines
(
Figure 8C and Table S1 in Supporting Information).21,56 Both SK-575 and olaparib potently inhibit HR-
deficient and HR-proficient cell growth with IC values at the micromolar level. However, SK-575 is
50
more potent than olaparib，indicating that the potency of cell growth inhibition is due in part to their
2
0
ACS Paragon Plus Environment

Page 21 of 68
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
ability to degrade PARP1. In addition, we evaluated the ability to potentiate the cytotoxic effect of SK-
75 by alkylating agent methyl methanesulphonate (MMS). After a growth-inhibition curve for SW620
5
cells was established, the effectiveness of MMS was potentiated in a dose-dependent manner by adding
PARP1 degrader at the concentration of nanomolar (Figure 8D). Moreover, we evaluated the preliminary
safety of SK-575 with six normal cell lines (L-O2, HEK-293, NCM460, HUV-EC, MCF-10A, and H9c2),
and the results were shown in Figure S6. Our data showed that the toxicity of SK-575 to these cells is
acceptable, similar to olaparib.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
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 highly potent and broad applicability of SK-575. (A) DC of SK-575 for PARP1 in different cancer cells.
5
0
Data are the average of two biologically independent experiments. (B) PARP1 activity assay. (C) The IC valves of SK-
5
0
5
75 and olaparib. Cells were treated with indicated doses for 7 days. Cell viability was determined by CCK-8 assay, and
the data were calculated using the GraphPad Prism 7 software. (D) Potentiation of MMS cell killing of cultured SW620
cells in combination with SK-575. Increasing concentrations of MMS were co-incubated with or without the PARP1
2
1
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 22 of 68
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
degrader at single concentrations ranging from 3 to 300 nM.
Because DNA double-strand breaks (DSBs) induce the phosphorylation of histone H2AX (γH2AX) in
the chromatin flanking the break site, the DNA damage levels can be monitored by estimating the
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
5
7
formation of γH2AX. We evaluated the ability of SK-575 and olaparib to induce the formation of
-/-
-/-
γH2AX foci in the MDA-MB-436 (BRCA1 ) and Capan-1 (BRCA2 ) cell lines by western blotting
assay. We found that SK-575 effectively induced the formation of γH2AX in these two cells in a dose-
dependent manner. In contrast, olaparib exhibited a weaker ability than SK-575 at the same concentration
(Figure 9).
Figure 9. Western blotting analysis of γH2AX proteins in MDA-MB-436 (A) and Capan-1 (B) cells, with β-Tubulin
used as the loading control. Cells were treated with PARP1 inhibitor (olaparib) or PARP1 degrader (SK-575) at indicated
concentrations, following by a 24 h treatment.
Pharmacokinetics Evaluation of SK-575 in Mice. To assess the antitumor activity of SK-575 in vivo,
we first performed a pharmacokinetics (PK) analysis in mice, and obtained the data shown in Figure 10A.
2
2
ACS Paragon Plus Environment

Page 23 of 68
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
In this experiment, a single dose of SK-575 at 25 mg/kg intraperitoneally followed by plasma level
quantitation revealed a maximum concentration of over 2.1 μM at 0.25 h postinjection, indicating that
SK-575 is absorbed rapidly. In addition, our data showed that SK-575 with a half-life of 3.08 h (n = 3)
has a concentration of 1.83 μM, 182 nM, and 95 nM in plasma at 1, 4, 8 h time-points, respectively.
Importantly, the plasma concentration remains above 10 nM, a concentration needed for effective PARP1
degradation shown in our in vitro experiments, for 24 h postinjection. Taken together, our PK data indicate
that a single dose of SK-575 at 25 mg/kg achieves a sufficient exposure in plasma for over 24 h.
Pharmacodynamics Evaluation of SK-575 in SW620 Xenograft Tumor Tissues in Mice. Based on
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
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 favorable pharmacokinetic properties, we evaluated the ability to degrade the PARP1 protein of SK-
3
5
75 in the SW620 xenograft tumor tissue in mice. Once the tumors had reached 200-300 mm , the mice
were randomly assigned into the vehicle control group and PROTAC group. The mice were administered
a single dose of SK-575 at 25 mg/kg intraperitoneal injection and were sacrificed at 1, 3, 6, and 24 h time-
points after treatment. The mice treated with vehicle control were sacrificed at the 3 h time-point. Our
western blotting data (Figure 10B) showed that a single dose of SK-575 at 25 mg/kg is very effective in
reducing the level of PARP1 protein in xenograft tumor tissue. The degradation of PARP1 protein started
at 1 h after treatment and reached the maximum effect at 6 h, indicating a rapid degradation of PARP1 in
vivo when compared to the vehicle. Our pharmacodynamics (PD) data thus demonstrates that a single
dose of SK-575 at 25 mg/kg is highly effective in inducing near-complete elimination of PARP1 protein
in tumor tissue in mice with the effect persisting for 24 h.
2
3
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 24 of 68
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 10. In vivo evaluation of PARP1 degrader SK-575. (A) PK study of SK-575 in mice (n = 3 per group). (B) PD
study of SK-575 in SW620 tumor tissue in mice. The treatment groups mice bearing xenograft SW620 tumors were
treated with a single dose of SK-575 (25 mg/kg, ip). Tumor tissues were harvested at the indicated time-points for western
blotting assay.
Antitumor Effect of SK-575 Administration in Xenograft Tumor Models. Upon the basis of its
impressive PK and PD data, we evaluated the efficacy of SK-575 in three xenograft models in mice when
used as a single agent or in combination with cytotoxic agents at different doses and schedules.
In the beginning, to assess the in vivo antitumor effects of SK-575 used as a single agent, we treated
nude mice bearing established subcutaneous Capan-1 tumor xenograft with SK-575 and obtained the
2
4
ACS Paragon Plus Environment

Page 25 of 68
Journal of Medicinal Chemistry
results shown in Figure 11A. Capan-1 is a human pancreatic cancer cell line with BRCA2 deficient.58
Olaparib effectively retarded tumor growth over the control, but SK-575 can achieve much similar
antitumor activity than olaparib at a lower dosage and fewer times of administration. At the lower dose
of 25 mg/kg once daily, the administration of SK-575 for five times a week for 3 weeks intraperitoneally
has a similar effect on tumor growth with olaparib (100 mg/kg, oral, daily dosing). SK-575 at these doses
(25 and 50 mg/kg) were well tolerated, with no mice lethality or significant weight loss observed during
the treatment time (Figure S7 in Supporting Information).
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
We observed a clear additive effect of combining PARP1 degrader SK-575 with MMS in SW620 cell
lines in vitro (Figure 8D). Thus, to assess the antitumor of SK-575 in combination with TMZ in vivo, we
used a nude mice colorectal SW620 xenograft model to compare the antitumor activity of TMZ (at 50
mg/kg for 5 days, po, once daily) either alone or in combination with SK-575 (at 5 and 10 mg/kg for 5
days, ip, once daily) or olaparib (at 10 mg/kg for 5 days, po, once daily) as the reference standard (Figure
1
1B). This dose and schedule of olaparib had previously demonstrated potentiation of TMZ antitumor
1
8
activity. Our data showed that TMZ alone can suppress tumor growth and the efficacy improved when
combined with SK-575 in dose-dependent manners. At a high dose of 10 mg/kg combination with TMZ
for 5 consecutive days, SK-575 significantly inhibited tumor regrowth compared with TMZ alone (P <
0
.01), similarly to the results of olaparib (P < 0.01). We examined combinatorial toxicity through body
weight loss and the results were shown in Figure S7 in supporting information. Our data indicated that
TMZ alone group was well-tolerated with a maximum mean body weight loss of 9% at the 6-day time-
point and recovered quickly after treatment. However, body weight loss was greater in the two
combination groups of the 10 mg/kg of SK-575 and olaparib, with a maximum mean body weight loss of
2
5
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 26 of 68
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
15% at the 9-day time-point in these two groups and full recovery within a week. There was no animal
lethality during the whole experiment. Overall, our combination therapy of PARP1 degrader was well-
tolerated under these doses and schedules. It is worth noting that this is the first preclinical example of
PROTAC compounds combined with other drugs for tumor treatment in vivo, which would provide
favorable evidence for the clinical application of PROTAC compounds.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
We next examined the additive effect of SK-575 when combined with cisplatin in vivo. The nude mice
bearing Capan-1 tumor xenografts were treated with SK-575 (at 25 and 50 mg/kg for 5 days, ip, once
daily) or olaparib (at 10 mg/kg for 5 days, po, once daily) in combination with a 6 mg/kg intraperitoneal
injection of cisplatin at the 2-day time-point (Figure 11C). The nude mice treated with cisplatin alone had
antitumor activity，and the efficacy improved when combined with SK-575 in dose-dependent manners.
Significant potentiation was observed as early as day 16 (P = 0.027) for the combination groups of SK-
5
75 at 50 mg/kg and continued to differentiate from the cisplatin alone group out to day 32 (P < 0.01).
However, we did note that the combination regimens involving cisplatin resulted in a maximum average
body weight loss of 9% and 12% at the 10-day time-point (Figure S7 in Supporting Information), which
were observed for groups that contained SK-575 doses of 25 mg/kg and 50 mg/kg, respectively. In the
same experiment, the nude mice treated with olaparib and cisplatin also showed combinatorial toxicity,
with maximum body weight loss of 11%. Mice recovered their body weight after a week of maximum
body weight losing, and no mice lethality was observed.
2
6
ACS Paragon Plus Environment

Page 27 of 68
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 11. (A) Antitumor efficacy of SK-575 in BRCA2-mutated Capan-1 xenografts (n = 6). (B) Antitumor efficacy
of SK-575 in combination with temozolomide (TMZ) in SW620 xenografts (n = 6). Nude mice were dosed once daily
for 5 consecutive days (days 0-4 are indicated by a short line). (C) Antitumor efficacy of SK-575 in combination with
cisplatin in BRCA2-mutated Capan-1 xenografts (n = 6). Nude mice were dosed SK-575 once daily for 5 consecutive
days (days 0-4 are indicated by a short line), and cisplatin was dosed intraperitoneally at 6 mg/kg on the 2-day time-
2
7
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 28 of 68
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
point. The efficacy of the treatment is expressed as mean ± SEM.
CONCLUSIONS
In conclusion, by linking known PARP1 inhibitor olaparib to E3 ligase ligand (for example,
thalidomide and lenalidomide), we have designed and synthesized a series of PARP1 degraders. Through
extensive optimization of the length and composition of the linker, we identified compound 18 (SK-575)
as a promising PARP1 degrader that exhibits rapid, highly potent and specific degradation of PARP1
protein in various cancer cells with DC value at concentrations as low as picomolar with a 24 h 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
5
0
time. Furthermore, SK-575 can cause persistent depletion of PARP1 protein in MDA-MB-436 cells at
00 nM for 5 days treatment, and the degraded PARP1 protein induced by the pretreatment of SK-575 is
1
rarely recovered after washout it for 24 h. SK-575 displays good potency against PARP1. Our data show
that SK-575 can not only effectively inhibit the growth of a variety of tumor cells as a single agent, as
exemplified by the MDA-MB-436 and Capan-1 cells, but also potentiate the cell killing by alkylating
agent MMS. Significantly, our PK and PD data indicate that a single dose of SK-575 achieves a sufficient
exposure in plasma for over 24 h, and effectively induces PARP1 degradation in the SW620 xenograft
tumor tissue with the effect persisting for >24 h. Furthermore, SK-575 significantly inhibited the tumor
growth in vivo as a single-agent in HR-deficient xenograft models, and synergistic sensitization in
combination with cytotoxic agents (TMZ or cisplatin) at well-tolerated dose schedules. Collectively, SK-
575 is the most potent and efficacious PARP1 degrader reported to date and warrants further evaluation
as a potential new therapy for the treatment of diseases caused by PARP1 hyperactivation.
2
8
ACS Paragon Plus Environment

Page 29 of 68
1
Journal of Medicinal Chemistry
CHEMISTRY
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
Compounds 6-11 were synthesized according to the route shown in Scheme 1. Compound 31 was
obtained by condensation of 3-hydroxyphthalic anhydride (29) and 3-aminopiperidine-2,6-dione
hydrochloride (30). Compound 31 was alkylated with tert-butyl bromate to obtain 32, which was
hydrolyzed in trifluoroacetic acid (TFA) to result in intermediate 33. Condensation of 34 obtained from
commercial source with N-Boc-piperazine in the presence of 1-[bis(dimethylamino)methylene]-1H-1,2,3-
triazolo[4,5-b]-pyridinium 3-oxide hexafluorophosphate (HATU) and triethylamine in
dimethylformamide (DMF) afforded the compound 35. Compounds 36a,b were prepared with 35 and N-
Boc-butylamine or N-Boc-hexamine using a similar procedure described above. After Bco-deprotection
of 36a,b with HCl in DCM, Compounds 6 and 7 were obtained by condensation of 36a,b with 33 using a
similar method as was used for 35. Compounds 8-11 were prepared from 36a,b by the same procedure as
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
was used for the synthesis of 6.
Scheme 1. Synthesis of Compounds 6-11a
2
9
ACS Paragon Plus Environment