Discovery and Optimization of Small Molecules Targeting the Protein-Protein Interaction of Heat Shock Protein 90 (Hsp90) and Cell Division Cycle 37 as Orally Active Inhibitors for the Treatment of Colorectal Cancer

Article abstract of DOI:10.1021/acs.jmedchem.9b01659

Cell division cycle 37 (Cdc37) is known to work as a kinase-specific cochaperone, which selectively regulates the maturation of kinases through protein-protein interaction (PPI) with Hsp90. Directly disrupting the Hsp90-Cdc37 PPI is emerging as an alternative strategy to develop anticancer agents through a specific inhibition manner of kinase clients of Hsp90. Based on a first specific small-molecule inhibitor targeting Hsp90-Cdc37 PPI (DDO-5936), which was previously reported by our group, we conducted a preliminary investigation of the structure-Activity relationships and pharmacodynamic evaluations to improve the potency and drug-like properties. Here, our efforts resulted in the currently best inhibitor 18h with improved binding affinity (K_d = 0.5 μM) and cellular inhibitory activity (IC₅₀ = 1.73 μM). Both in vitro and in vivo assays revealed that 18h could efficiently block the Hsp90-Cdc37 interaction to specifically inhibit kinase clients of Hsp90. Furthermore, 18h showed ideal physiochemical properties with favorable stability, leading to an oral efficacy in vivo.

Full text of DOI:10.1021/acs.jmedchem.9b01659

Subscriber access provided by UNIVERSITY OF TECHNOLOGY SYDNEY
Article
Discovery and Optimization of Small Molecules Targeting the Protein-
Protein Interaction of Heat Shock Protein 90 (Hsp90) and Cell Division Cycle
37 (Cdc37) As Orally Active Inhibitors for the treatment of colorectal cancer
Lei Wang, Jingsheng Jiang, Lixiao Zhang, qiuyue zhang, Jianrui Zhou, Li Li, Xiao-Li Xu, and Qi-Dong You
J. Med. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.jmedchem.9b01659 • Publication Date (Web): 14 Jan 2020
Downloaded from pubs.acs.org on January 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 50
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Discovery and Optimization of Small Molecules Targeting the
Protein-Protein Interaction of Heat Shock Protein 90 (Hsp90) and
Cell Division Cycle 37 (Cdc37) As Orally Active Inhibitors for the
Treatment of Colorectal Cancer
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
,2
1,2
1,2
1,2
1,2
Lei Wang , Jingsheng Jiang , Lixiao Zhang , Qiuyue Zhang , Jianrui Zhou ,
Li Li , Xiaoli Xu 1,2,*, Qidong You 1,2,*
1,2
1State Key Laboratory of Natural Medicines and Jiang Su Key Laboratory of Drug
Design and Optimization, China Pharmaceutical University, Nanjing 210009, China.
2Department of Medicinal Chemistry, School of Pharmacy, China Pharmaceutical
University, Nanjing 210009, China.
*
To whom correspondence should be addressed:
Xiaoli Xu, Ph.D.
Phone: +86-15261483858
E-mail: xuxiao_li@126.com
Prof. Qidong You, Ph.D.
Phone: +86 025 83271351
E-mail: youqd@163.com
1
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 2 of 50
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Abstract
Cdc37 is known to work as a kinase-specific cochaperone, which selectively
regulates the maturation of kinases through protein-protein interaction (PPI) with
Hsp90. Directly disrupting the Hsp90-Cdc37 PPI is emerging as an alternative
strategy to develop anticancer agents through a specific inhibition manner of kinase
clients of Hsp90. Based on a first specific small molecule inhibitor targeting
Hsp90-Cdc37 PPI (DDO-5936), which was previously reported by our group, we
conducted a preliminary investigation of the structure-activity relationships and
pharmacodynamic evaluations to improve the potency and drug-like properties. Here,
our efforts resulted in a currently best inhibitor 18h with improved binding affinity
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
(
K = 0.5 μM) and cellular inhibitory activity (IC = 1.73 μM). Both in vitro and in
d
5
0
vivo assays revealed 18h could efficiently block the Hsp90-Cdc37 interaction to
specifically inhibit kinase clients of Hsp90. Furthermore, 18h showed ideal
physiochemical properties with favorable stability, leading to an oral efficacy in vivo.
2
ACS Paragon Plus Environment

Page 3 of 50
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Introduction
Molecular chaperone heat shock protein 90 (Hsp90) is one of the most abundant
intracellular protein, which plays a significant role for its client proteins in correct
1
-3
folding, maturation and stability . A considerable number of client proteins of Hsp90
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
4
are fundamental effectors in the growth and proliferation of cancer cells . In the past
decade, directly targeting the Hsp90 ATPase pocket was considered as a promising
5
-7
anti-cancer target . Although several small molecule inhibitors have entered into
clinical trials, inevitable heat shock response, limited efficacy and diverse toxicities
become the formidable obstacles for them to be approved to the market8-9.
Apparently, directly inhibiting the activity of Hsp90 ATPase totally blocked the ATP
hydrolysis, leading to a nonselective degradation of all the clients of Hsp90 as well as
a feedback of heat shock response (HSR)1, 10. In fact, during the chaperone cycle,
Hsp90 works with the help of many co-cochaperones through diverse protein-protein
interactions (PPIs)11-12. Different co-chaperones exhibit disparate functions and play
specific roles in the dynamic cycle5, 13. Among them, cell division cycle 37 (Cdc37) is
responsible for selectively recognizing and recruiting the kinase clients, revealing
1
2
direct and high correlations with diverse cancers . In the Hsp90-Cdc37-kinase
recruitment cycle, Cdc37 selectively recognizes the C lobe of unfolded kinase clients
1
4
(eg. CDK4) under a phosphorylated condition by CK2 . Then, the complex of
Cdc37-CDK4 binds to an open state of Hsp90. At this stage, Cdc37 interacts with
Hsp90 N-terminus via its M domain, as shown in the co-crystal structure (PDB:
1
5
1
US7) . Subsequently, Cdc37-CDK4 complex shifts to Hsp90 M domain before ATP
is captured by Hsp90, making Hsp90 turns to a closed state (PDB: 5FWL). Finally,
the captured ATP is fully hydrolyzed and Hsp90 opens to provide a chance for correct
folding of CDK4, after which all the components release with the help of PP5 to
1
6
de-phosphorylate p-Cdc37 (Figure 1) . Thus, a dynamic recruitment cycle of
Hsp90-Cdc37-kinases reminds us that inhibition of ATP hydrolyzation is only a
rate-limiting step of Hsp90. Different with directly inhibiting ATPase, targeting
Hsp90-Cdc37 PPI reveals as an alternative strategy to block the maturation of kinases
3
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 4 of 50
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
1
7
to achieve cancer therapy with specificity and safety . In addition, elevated levels of
Cdc37 were observed in proliferating cancer cells while normal tissues required little
1
8
Cdc37, which provided a potential therapeutic window . Recently, many structural
studies combined with computational methods of Hsp90-Cdc37-kinase interaction
were reported, indicating a more rational way to understand the recognition
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
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
mechanism . Using NMR and computational studies, the structure-based network of
protein kinase clients to the Hsp90-Cdc37 chaperone is emerging and allosteric
regulation manner is becoming feasible20-21. Structural characterization of
Hsp90-Cdc37 also provide important evidence including critical interactions and
binding sites22.
Figure 1. Hsp90-Cdc37 chaperone cycle for kinase recruitment. (Ⅰ) With the
help of Casein kinase 2 (CK2), Cdc37 is phosphorylated to bind the C lobe of the
unfolded kinase client (such as CDK4). (Ⅱ) The complex of Cdc37-kinase
subsequently binds to the open state of Hsp90. At this time, M domain of Cdc37
interacts with the N-terminus of Hsp90 (PDB: 1US7). (Ⅲ) Hsp90 captures ATP and
turns to a closed state, resulting in Cdc37-kinase complex transferred to M domain of
4
ACS Paragon Plus Environment

Page 5 of 50
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Hsp90 (PDB: 5FWL). (Ⅳ) When ATP is hydrolyzed, N terminus of Hsp90 opens to
provide a chance for kinase folding in a Cdc37-kinase bound state. (Ⅴ) After kinase
is totally folded, PP5 helps the de-phosphorylated process of Cdc37. Finally, kinases
and Cdc37 are displaced and released.
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
Currently, small molecular inhibitors with the ability to directly target
Hsp90-Cdc37 PPI are rarely reported. At the very beginning, several natural products
with anti-cancer potency were found to disrupt Hsp90-Cdc37 PPI in cells by co-IP
assays. Although most of them were identified with multiple intracellular targets, they
were the first evidence to demonstrate the feasibility of targeting Hsp90-Cdc37 PPI to
achieve potential cancer therapy (Figure 2). Among them, celastrol (extract of
celastranceae plants) was one of the most studied natural products for its explicit
anti-cancer potency23-24. However, the direct binding site of celastrol to disrupt
Hsp90-Cdc37 PPI remains controversial, including Hsp90 C-terminus (determined by
2
5
fingerprinting assay) and Hsp90 N-terminus (predicted by molecular docking) .
Another more convincing result illustrated that celastrol covalently bound to Cdc37,
2
6
leading to vanishing peaks in NMR assay determined by HSQC spectrum .
Withaferin A (WA), a bioactive compound from extract of Withania somnifera, and
FW-04-804, an extract from the culture broth of soil Streptomyces, were also reported
to reveal their anti-proliferative activity through a mechanism involving Hsp90-Cdc37
PPI inhibition27-28. In addition, Kongensin A (KA), a natural product isolated from
Croton kongensis, was reported to covalently bind to a previously undiscovered site
involving cysteine 420 in the middle domain of Hsp90, resulting in dissociation of
Hsp90 from Cdc37 and providing an original site for the design of covalent
2
9
inhibitors . Different with the above natural products, VS-8 was a non-natural
compound targeting Hsp90-Cdc37 PPI discovered by our group through virtual
screening based on the peptides of Hsp90-Cdc37 binding interface. Although VS-8
could bind to Hsp90 with a moderate K value of 80.4 μM, its accurate binding site
D
remained unclear and the inhibitory function of Hsp90-Cdc37 PPI in cells was
5
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 6 of 50
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
3
0
unkonwn . In 2018, another small molecular inhibitor (DCZ3112) was reported to
exert anti-tumor activity against HER-2 positive breast cancer through inhibition of
Hsp90-Cdc37 PPI with a K value of 4.98 μM. In fact, although the direct binding site
d
of DCZ3112 was unclear, it provided important clues for potential therapeutic effects
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
3
1
by pharmacologically targeting Hsp90-Cdc37 PPI .
Recently, our group reported that specifically disrupting the Hsp90-Cdc37 PPI with
small molecules was feasible to achieve potential therapeutic effects for colorectal
cancer by selectively down-regulating the kinase clients of Hsp90 and arresting the
32
cell cycle without heat shock response . Based on the discovery of hit compound
1
1(Figure 2), we identified DDO-5936 as a novel Hsp90-Cdc37 PPI inhibitor with
potency and specificity through binding to a critical site on Hsp90 involving Glu47
the most important binding determinants for Hsp90-Cdc37 PPI discovered by our
(
group). Here, on the basis of our previous work, we reported the structure-activity
(SAR) and structure-property (SPR) relationships of compound 11 and DDO-5936.
The optimized compound 18h showed ~10 folds enhancement of binding affinity (Kd
= 0.5 μM) and cellular inhibitory activity against HCT116 cells (IC = 1.73 μM).
5
0
Physicochemical properties were also improved, especially the cell membrane
permeability, which resulted in enhanced activities in cells. Both in vitro and in vivo
assays confirmed that 18h efficiently disrupted Hsp90-Cdc37 PPI, selectively
down-regulated kinase clients of Hsp90 and arrested cell cycle in phase G /G to
0
1
achieve anti-proliferation effects. Notably, compared with DDO-5936, 18h exhibited
much better stability in plasma and liver microsome, leading to favorable anti-cancer
effects in the HCT116 tumor xenograft model by oral administration, while
DDO-5936 showed limited oral efficiency. These results suggested that Hsp90-Cdc37
PPI inhibitors can achieve potential anti-cancer therapeutic effects with preferable
selectivity and safety, providing a novel insight for the design of inhibitors by
targeting molecular chaperone pathways. 18h represented as a class of promising
inhibitors targeting Hsp90-Cdc37 PPI.
6
ACS Paragon Plus Environment

Page 7 of 50
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
H3C COOH
H C
H3C H3C
3
H
H
CH3
H
O
H3C
H
3
H C
O
OAc
O O
H
H
OH
H3C
O
H
H
O
O
O
N
N
O
O
Me
H3C
CH3
O
H
O
H
Me Me
OH
H
O
CH3
H
HO
H
CH3 CH3
OH
2 Withaferin A (WA)
Binding site unknown
no Kd value
3 FW-04-804
Binding site unknown
4 Kongensin A (KA)
Covalently binding to Hsp90 M
1
Celastrol
Covalently binding to Cdc37
Binding site unknown
no Kd value
no Kd value
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
Cl
H
HN
CF3
H
N
N
N
N
H
N
H
N
O O
S
N
N
O
O
N
H
H
O
S
N
N
N
N
O
O
N
N
NH2
N
S
S
N
N
N
H
H
OH
O
O
O
O
N
N
N
O
O
O
5 VS-8
6 DCZ3112
7 Compound 11
Binding to Hsp90 N, Kd = 21.1 M
8 DDO-5936
Binding to Hsp90 N, Kd = 3.86 M
Interacting with Glu47 on Hsp90
Binding to Hsp90 N, Kd = 80.4 M
Binding to Hsp90, Kd = 4.98 M
Binding site unknown
Bind site unclear
Figure 2. Previously reported natural products involving a Hsp90-Cdc37 PPI
inhibition mechanism and small molecule inhibitors targeting Hsp90-Cdc37 PPI.
Results and Discussion
Preliminary SAR Study Revealed DDO-5936 with Potent Activity. On the basis of
the discovery of 7 (compound 11, chemical structure shown in Figure 3A), which
exhibited moderate potency to disrupt Hsp90-Cdc37 PPI (K = 21.1 μM, inhibition
D
rate ~56% at 100 μM), our efforts were firstly focused on the improvement of the
binding affinity and physical chemical properties. Although 7 emerged as a potential
start compound, its binding affinity could not be determined by ITC or NMR assays,
which might due to its poor solubility and low activity. Based on the previous results
and guided by the potential binding mode of 7 (Figure 3B), we conducted a most
important hydrogen bond formed between Glu47 of Hsp90 and imino group of 7.
Then, we split the chemical structure of 7 into three parts, including A ring
substituents (which is surrounded by hydrophobic residues involving Phe213),
sulfamide part (exhibiting potential space reaching to both protein surface and
solvent) and pyrimidine substituents (close to a polar pocket involving Ser50, Asn51,
Ser53 and Asp54), shown in Figure 3A. As shown in Table 1, A ring of 7 was
replaced with diverse substituted benzene (Scheme 1). Among all the derivatives,
compound 12c (2,4,6-trimethy phenyl substituted) exhibited potent inhibitory ability
with 66.0% inhibition rate, leading to an ideal activity at current stage. However, the
7
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 8 of 50
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 rate of the compounds with mono-substituted benzene were dramatically
decreased to below 20%, including 12a (4-trifluoromethyl phenyl substituted), 12d
(4-methoxyphenyl substituted) and 12e (4-tertbutyl phenyl substituted). Compounds
with di-substituted benzene were not ideal either, both 12f (2,4-dimethoxyphenyl
substituted) and 12g (2,5-dimethoxyphenyl substituted) showing low inhibition rates
in HTRF assay. In addition, compound 12b demonstrated that larger occupation with
naphthalene was not suitable for this site either. Thus, 12c was the most suitable
group for A ring optimization for its favorable inhibitory activity.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Figure 3. Binding mode of compound 11 and the optimization strategies. (A)
Chemical structure of compound 11 and the three sites to be optimized according to
the binding mode. (B) Compound 11 binds to a surface site next to the Hsp90
ATPase pocket (PDB code 2K5B), interacting with the critical residues (forming
hydrogen bond) of Hsp90 including Glu47 (binding determinant for Hsp90-Cdc37
PPI). ATPase site on Hsp90 is enclosed and highlighted. Hsp90 is shown in white
ribbon and the compound 11 is shown as cyan stick, respectively. Hydrogen bonds
are indicated with red dashed lines.
Table. 1 Structure-Activity Relationships of different substitution on A ring.
H
N
N
N
O
O
S
N
R
N
H
A Ring
Inhibition rate (%)a
Inhibition rate (%) a
Comp·
R
Comp.
R
8
ACS Paragon Plus Environment

Page 9 of 50
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
O
Comp.11
56.3 ± 9.4
12d
12e
＜
20
O
1
2a
＜
20
＜20
F3C
1
2b
46.7 ± 5.2
66.0 ± 8.7
12f
36.2 ± 4.9
O
O
O
O
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
2c
12g
41.8 ± 5.0
a
All inhibition rates were determined using 100 μM compound.
Considering the tri-methyl phenyl was important for activity, we further tried to
investigate the substituents on sulfamide with tri-methyl phenyl remaining on A ring.
Based on the chemical structure of 12c, multiple side chains were employed to obtain
1
5b (previously reported as DDO-5936) with improved activity. As shown in Table 2,
benzyl groups were not appropriate in reducing the inhibition rates below 50%,
including 15d (4-benzoic acid substituent), 15e (4-isopropyl phenyl substituent), 15f
(
4-chloro phenyl substituent) and 15g (4-methoxy phenyl substituent). Both 15a and
1
5b showed ideal inhibition rates of 65.4% and 72.6% respectively, indicating a
tolerable modification by employing a side chain of acetic acid. Although 15b only
slightly improved the in vitro activity, including a 72.6% inhibition rate by HTRF
assay and a moderate binding affinity by BLI (K = 7.41 μM), it could be detected by
D
ITC (K = 7.81 μM, Figure S2) and NMR assays (previously determined by CPMG,
d
STD and HSQC) 32, which illustrated an explicit binding affinity and a pivotal
binding site to disrupt Hsp90-Cdc37 PPI (Table 4). 15b exhibited anti-proliferative
potency with an IC50 value of 10.24 μM against HCT116 cells (Cdc37 high
expression), a lower activity with an IC50 value of 55.74 μM against A549 cells
(
Cdc37 low expression) and almost no activity against normal hepatocyte (L02).
Although 15b exhibited as an initial compound with moderate potency, it was still
urgent to enhance its binding affinity and in vivo efficiency. In addition, 15b showed
non-ideal stability in plasma or liver microsome (Table 6 & Table 7), as well as an
oral inefficacy in in vivo models. Thus, our next task was to optimize its binding
9
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 10 of 50
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
affinity, anti-proliferation activity, stability and physicochemical properties to achieve
a more potent Hsp90-Cdc37 PPI inhibitor both in vitro and in vivo.
Table.2 Structure-Activity Relationships of Different Substituents on Sulfamide.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
H
N
N
N
O
O
S
N
N
R
Sulfamide
Comp.
2c
R
Inhibition rate (%)a Comp.
R
Inhibition rate (%)a
1
-
66.0 ± 10.1
15d
52.8 ± 4.8
OH
O
O
1
5a
65.4 ± 7.4
72.6 ± 3.5
15e
15f
＜
＜
50
50
O
O
OH
1
5b
DDO-5936)
5c
Cl
(
NH2
1
15g
＜
50
＜50
O
O
a
All inhibition rates were determined using 100 μM compound.
Further SARs Led to 18h with Improved Binding Affinity and Anti-proliferation
Activity. On the basis of 15b, we further investigated the pyrimidine substituents to
explore the pocket with the aim to improve activity on this site. As shown in Table 3,
we firstly replaced the pyrrolidine group with methylamine and ethylamine, resulting
in compound 17a, 18a and 17b, 18b with dramatic activity loss (less than 50%).
Similarly, substituents including cyclopropylamine (compound 17c and 18c) and
pentan-3-amine (compound 17d and 18d) exhibited low inhibition rates (Table 3),
indicating an unfavorable flexibility site on this region. Interestingly, 17e with
diethylamine substituent exhibited low activity while 18e showed a considerable
inhibition rate of 79.58%. Understanding the region of pyrimidine-substituents was
not suitable for the flexible chains, we next employed piperidine and azepane with
10
ACS Paragon Plus Environment

Page 11 of 50
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
larger rings, to compare with pyrrolidine-substituted DDO-5936. Notably, 17f and 18f
maintained inhibitory activity with inhibition rates of 79.11% and 86.17%,
respectively. However, 17g and 18g reduced the inhibitory activity to below 50%,
demonstrating an unfavorable employment of larger groups. Encouraged by 17f and
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
8f, we subsequently employed more polar groups including piperazine and N-methyl
piperazine to obtain compound 17h, 18h and 17i, 18i. Although 17i and 18i
maintained activity, 17h and 18h exhibited best inhibition rates (95.48% and 97.41%,
respectively) which were worthy of further evaluation.
H
H
N
N
R
N
N
R
O
O
O
O
S
N
S
N
N
N
Pyrimidine-Substituents
Pyrimidine-Substituents
O
O
O
OH
1
7 a-i
18 a-i
Table. 3 Structure-Activity Relationship of Pyrimidine-Substituted compounds.
R
Comp. Inhibition rate (%)a
17a
Comp.
18a
Inhibition rate (%)a
H
N
＜
50
＜50
H
N
1
7b
18b
＜
50
＜50
＜50
H
N
17c
17d
17e
18c
18d
18e
18f
18g
18h
18i
＜50
50
H
N
＜
＜50
79.58 ± 4.1
N
＜50
1
7f
79.11 ± 3.3
86.17 ± 6.0
N
1
7g
N
＜50
＜50
NH
N
1
7h
95.48 ± 5.7
97.41 ± 6.2
N
N
1
7i
70.13 ± 2.9
71.09 ± 4.8
a
All inhibition rates were determined using 100 μM compound.
11
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 12 of 50
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
Considering the results that different side chains of 17h and 18h (methyl acetate
and acetic acid respectively) showed similar inhibitory activity, we next explored
more substituents and bioisosteres of acetic acid of 18h. As shown in Table 4,
inhibitors including 17h, 18h, 19 and 21 all showed parallel inhibitory activities
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
(inhibition rates exceeded 95%), while 20 exhibited an activity loss to only 64.31%.
We subsequently conducted thermal shift assay to determine the potential binding
o
affinity, leading to 17h and 18h with maximum △T values of 3.53 and 4.75 C,
m
o
respectively, which showed superior results compared with 2.13 C of DDO-5936. In
order to comprehensively investigate the binding processes and kinetic profiles,
binding assays including both BLI (binding kinetics) and ITC (binding
thermodynamics) were carried out to determine the precise binding affinities.
Consistent with thermal shift assay, 17h and 18h exhibited the most potent binding
affinity to Hsp90, leading to an improved in vitro activity comparing to DDO-5936.
Especially for 18h, which could be regarded as a first compound with binding affinity
below one micromolar (K = 0.50 μM by BLI assay and K = 0.63 μM by ITC assay,
D
d
Figure 4A and 4B).Although Kon value of 18h slightly decreased to 10600 Ms-1
-1
(compared to DDO-5936, K = 16800 Ms ), a comparably low dissociation rate was
on
-1
observed than other compounds (Koff = 0.005 s ), which contributed most to the
increased binding affinity. In ITC assay, both entropy change (-T△S = -4.52
kcal/mol) and enthalpy change (△H = -3.96 kcal/mol) favorably contributed to gibbs
free energy of 18h. Unfortunately, 17h, 19 and 20 all showed similar binding affinity
with DDO-5936 in thermal shift assay, BLI assay and ITC assay (Table 4). In
addition, ΔG values from the computational tools (MM-GBSA) also showed
consistent results with ITC experiments (Table S2).
Consistent results were observed in the T1ρ NMR spectra assays (Figure S3),
indicating that the WT Hsp90 protein significantly interfered with 18h. Meanwhile,
three mutants (R46A, E47A and Q133A) exhibited less effects on 18h, especially
mutant E47A, showing a little effect on CPMG spectrum of 18h which demonstrated
12
ACS Paragon Plus Environment

Page 13 of 50
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
that Glu47 of Hsp90 was one of the binding determinants for 18h. In addition,
positive saturation transfer difference (STD) spectra was also employed to detect STD
signals, which reconfirmed that 18h directly bound to Hsp90 (Figure 4C). We further
compared the potential binding modes between 18h and DDO-5936. As shown in
Figure 5, 18h occupied a similar binding site of DDO-5936, a cleft next to the
ATPase pocket, forming a critical hydrogen bond with Glu47 on Hsp90. Piperazine
substituent might exhibit a better adaption to the pocket involving Ser50, Asn51 and
Asp54. In addition, all the compounds including 17h, 18h, 19, 20, 21 and DDO-5936
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
exhibited no inhibition on Hsp90 ATPase activity (IC ＞100 μM), using AT13387,
50
a phase Ⅱ Hsp90 inhibitor targeting ATPase site, as reference (IC = 150 nM, Table
5
0
S1). Together, all the results demonstrated an explicit binding mechanism of 18h with
a ~10 folds improvement on in vitro binding affinity comparing to DDO-5936.
Figure 4. 18h directly bound to Hsp90. (A) Binding affinity of 18h with Hsp90
protein, determined by binding kinetic assay of biolayer interferometry (ForteBio
Octet). The K value was 0.5 μM. (B) Direct binding experiments of Hsp90 and 18h,
D
determined by binding thermodynamic assay of ITC. The K value was 0.63 μM. (C)
d
1H saturation transfer difference (STD) NMR spectrum of 100 μM 18h binding to 5
μM Hsp90. ppm, parts per million. All data are representative of three independent
experiments.
We next evaluated these Hsp90-Cdc37 PPI inhibitors for their cell growth
inhibitory activity against two cancer cell lines including HCT116 (Cdc37 high
expression) and A549 (Cdc37 low expression) and a non-cancer cell line (L02,
13
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 14 of 50
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
normal human liver cells). Although all the derivatives showed efficiency in cells
(Table 4), 18h exhibited both the preferable antiproliferative activity against HCT116
cells (IC = 1.73 μM) and the best selectivity (IC ＞50 μM against A549 and L02).
5
0
50
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Table 4. Determination of in vitro inhibition activity and binding affinities of
Hsp90-Cdc37 PPI inhibitors using HTRF, thermal shift, BLI, ITC and
antiproliferative activities by CCK8 assay.
O
R=
O
17h
NH
H
H
N
N
N
OH
N
O O
S
OH
N
O
O
N
18h
19
R
N
CN
NH
N
N
20
21
Binding Kinetics
(BLI) b
Binding Thermo Dynamics
(ITC) b
Antiproliferative activities
Comp.
Inhibition
△T
m
rate (%)a
(℃) b
IC50 (μM)
K
D
K
on
K
off
K
d
△H
-T△S
HCT116
A549
L02
-1
-1
(μM)
(Ms )
20300
(s )
(μM)
(kcal/mol)
(kcal/mol)
-5.90
1
1
7h
8h
95.48 ± 4.7
97.41 ± 6.2
96.24 ± 4.9
64.31 ± 2.3
98.58 ± 3.4
+ 3.53
+ 4.75
+ 1.59
+ 1.10
+ 1.36
1.97
0.04
0.005
0.02
1.73
-1.96
-3.96
-0.56
-
6.25 ± 1.2
1.73 ± 0.6
3.94 ± 0.8
-
43.26 ± 8.9
＞50
＞50
＞50
0.50
2.66
10600
6350
-
0.63
5.49
-4.52
-6.62
1
9
0
1
16.23 ± 2.2
-
21.94 ± 2.3
-
2
2
5.50
7.41
7010
0.04
0.13
2.40
7.81
-1.61
-6.06
-2.80
5.42 ± 1.7
37.94 ± 2.7
＞50
1
5b
7
2.60 ± 3.5
+ 2.13
16800
-4.17
10.24 ± 3.9
55.74 ± 9.8
＞50
(
DDO-5936)
a
All inhibition rates were determined using 100 μM compound.
b
Determined against Hsp90 alone.
14
ACS Paragon Plus Environment

Page 15 of 50
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 5. Predicted binding mode of DDO-5936 and 18h in Hsp90-Cdc37 binding
sites (PDB ID: 1US7). (A) Binding pattern of DDO-5936 to Hsp90 (PDB code
2
K5B). The compound is shown in yellow stick and Hsp90 is shown in white ribbon.
(B) Binding pattern of 18h to Hsp90 (PDB code 2K5B). The compound is shown in
green stick and Hsp90 is shown in white ribbon. Hydrogen bonds are shown in red
dashed lines and the residues in the binding site are labeled. ATP binding site is
circled in red dashed lines and highlighted.
Physicochemical Properties of 18h and Its Derivatives. To evaluate the drug-like
properties of 18h and its derivatives, lipid−water distribution coefficients (ClogP)
were computationally calculated through ACD lab and experimentally determined on
a Gemini Profiler instrument (pION) by the potentiometric titration method, similar as
3
3
previously reported . The permeability coefficients (Pe) were determined by a
standard parallel artificial membrane permeability assay (PAMPA) on a PAMPA
Explore instrument (pION). As shown in Table 5, 18h displayed improved
permeability at pH 7.4, compared with DDO-5936 and compound 11. All the
15
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 16 of 50
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 showed reasonable lipid−water distribution coefficients, while compound
1
9, 20, 21 showed limited or decreased permeability.
Table 5. Physicochemical Properties of the compounds 17h, 18h, 19, 20, 21 and
DDO-5936.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Comp.
ClogP
log D, pH 7.4
Pe, pH 7.4
-6
(
10 cm/s)
101.22
104.35
44.10
1
1
7h
8h
3.98
3.52
2.21
3.36
2.63
1.93
0.82
2.01
1
9
0
1
2
61.49
2
2.62
4.54
-
55.37
70.42
1
5b
DDO-5936)
Compound 11
1.67
(
5.00
4.82
65.87
1
8h Showed Preferable Stability in Plasma and Liver Microsomes. As 18h showed
ideal physicochemical properties and improved activities on cellular level, we further
evaluated the stability of 18h in mouse and human plasma as well as in vitro
metabolic stability in human and mouse liver microsomes, using DDO-5936 as
reference. As shown in Table 6, 18h showed preferable stability both in mouse
plasma and human plasma, exhibiting an approximately 100% compound retention
after 90 minutes incubation. However, DDO-5936 showed only 59% remain in mouse
plasma and 62% in human plasma after 90 minutes incubation, indicating a feeblish
stability which might limit its further in vivo efficacy. The in vitro metabolic stability
of the two compounds was assessed both in human liver microsomes and mouse liver
microsomes (Table 7). 18h possessed better metabolic stability with half-life values
of 124.5 minutes in human liver microsomes and 96.3 minutes in mouse liver
microsomes, compared with 30.4 minutes and 16.82 minutes of DDO-5936
respectively. Meanwhile, 18h also revealed lower clearance rates. Thus, 18h
16
ACS Paragon Plus Environment

Page 17 of 50
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
exhibited promising stability and was worthy for further development.
Table 6. Stability of DDO-5936 and 18h in mouse and human plasma.
Incubation
Time (min)
DDO-5936 remained
18h remained
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
Mouse Plasma Human Plasma
Mouse Plasma
Human Plasma
100%
0
100%
100%
86%
100%
101%
92%
100%
100%
102%
101%
105%
1
3
6
9
0
0
0
0
98%
104%
71%
76%
102%
59%
62%
103%
Table 7. In vitro metabolic stability in human and mouse liver microsomes.
Compound
Human liver microsomes
Mouse liver microsomes
T
1/2 (min)
30.4
CL (μl/min/mg)
T
1/2 (min)
16.82
96.3
CL(μl/min/mg)
DDO-5936
32.6
8.97
59
1
8h
124.5
10.3
MD Simulation Revealed 18h With Favorable Stability. In order to explore the
binding process and simulate the dynamic course of DDO-5936 and 18h, we
3
2
performed a 50 ns MD simulation based on the binding modes described previously .
Based on the simulation trajectory, we extracted five binding modes of each
compound at 0, 10, 20, 30, 40 and 50 ns, respectively and conducted them to
superimposition modes. As shown in Figure 6, DDO-5936 started to exhibit an
unstable binding mode from 40 ns simulation, exerting an unexpected increase of
RMSD values. The binding mode of DDO-5936 at 50 ns also showed a departure
pattern from the binding pocket. In contrast, 18h showed consistent binding modes
from 0 to 50 ns, revealing a preferable stability in this simulation system.
Additionally, we calculated the changes of Gibbs free energy of both DDO-5936 and
1
8h at different time point of simulation trajectory (Figure S4). Consistent with other
17
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 18 of 50
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
experimental results, the ΔG values of DDO-5936 exhibited a markedly decrease at
4
0 ns simulation (less than -5 kcal/mol remained at 50 ns) while the ΔG values of 18h
showed a constant level during the whole simulation, demonstrating a more stable
binding pattern of 18h compared to DDO-5936.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
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. Molecular dynamic simulation (50ns) revealed 18h with preferable
stability. The superimposition of binding modes of DDO-5936 (A) and 18h (B) on 0,
1
0, 20, 30, 40 and 50 ns. RMSD values of backbone atoms of Hsp90 and the
compound were shown to represent the stability. Orange curve represents RMSD
values of DDO-5936 and green curve for 18h. RMSD values of Hsp90 protein are all
displayed with gray curves. Data are representative of three independent experiments.
1
8h Efficiently Disrupted Hsp90-Cdc37 PPI in Cells and Selectively Induced
Kinase Degradation. To confirm the detailed functions and interactions between 18h
and the Hsp90 protein in cells, we further performed a cellular thermal shift assay
(CETSA) to demonstrate the binding affinity, cellular uptake and target engagement
of 18h in cells. As shown in Figure 7A, 18h exhibited ability to thermally stabilize
Hsp90 in HCT116 cells by Western blots analysis. Using β-actin as a reference, a
detectable phenomenon revealed that Hsp90 showed an explicit distinction between
being untreated or treated with 20 μM 18h for 12 hours, with an alternation of
18
ACS Paragon Plus Environment

Page 19 of 50
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
o
temperature ranging from 37 to 75 C, demonstrating a direct binding pattern of 18h
to Hsp90 protein and a preferable cellular uptake. Then, quantitative analysis of these
o
o
results was calculated to indicate a T shift from 51.75 C to 58.34 C. ΔT value was
m
m
o
+
6.39 C which showed a consistent result with thermal shift assay (ΔT = + 4.75
m
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
oC). Subsequently, we carried out coimmunoprecipitation (co-IP) experiments to
determine the Hsp90-Cdc37 PPI inhibition efficacy of 18h in cells. In order to
evaluate the different regulation mechanisms of 18h compared with Hsp90 ATPase
inhibitors, we used AT13387 as a reference. After treatment with 18h (0, 1, 5, 10 μM)
and AT13387 (0, 0.2, 0.5, 1 μM) at different concentrations, respectively, we
immunoprecipitated Hsp90 to determine the bait proteins. We distinctly observed that
1
8h disrupted the Hsp90-Cdc37 PPI in HCT116 cells with a concentration-dependent
manner, while AT13387 showed no inhibition of Hsp90-Cdc37 PPI (Figure 7B). In
addition, with Hsp90-Cdc37 being blocked, CDK4 was also detected to be decreased
in 18h treated groups, showing a reasonable and consistent result as previously
3
2
reported . Thus, both CETSA and co-IP assay indicated 18h as an efficient small
molecule inhibitor targeting Hsp90-Cdc37 PPI in cellular level.
To further identify the specific inhibition effects of 18h on Hsp90-Cdc37 pathway,
we performed Western-blots assays to determine the representative biomarkers of
Hsp90-Cdc37 in both concentration-dependent and time-dependent manner. As
shown in Figure 7C, 18h specifically inhibited the kinase clients of Hsp90 (CDK4
and CDK6), leading to downregulation of cyclin D1/cyclin D3 and upregulation of
p21/p27. Treatment of 18h showed no effects on either the glucocorticoid receptor
(
GR) or its phosphorylated form (p-GR), which was a nonkinase client protein of
Hsp90. This result indicated a favorable specificity of selective inhibition of Hsp90
clients as previously reported. Notably, Hsp70 showed no increase after treatment of
1
8h to avoid potential heat shock response, which was different with the reactions
after treated with Hsp90 ATPase inhibitors. In addition, Hsp90, Cdc37 and p-Cdc37
all showed consistent expression. Therefore, 18h disrupted Hsp90-Cdc37 PPI through
a direct binding mode in cellular lever to achieve a specifically inhibiting manner of
19
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 20 of 50
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
kinase clients of Hsp90.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
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. 18h disrupted the Hsp90-Cdc37 PPI in cells through directly binding to
Hsp90, and selectively down-regulated kinase clients of Hsp90 without HSR. (A)
co-IP assay of 18h (concentrations of the compound increasing from 0, 1, 5 to 10 μM)
and AT13387 (0, 0.2, 0.5, 1 μM) in HCT116 cells. 18h efficiently blocked
Hsp90-Cdc37 PPI through a dose-dependent manner, leading to the dissociation of
CDK4 from the Hsp90-Cdc37 complex. AT13387 showed no effect on Hsp90-Cdc37
PPI in cells. Western blots were performed with anti-Hsp90, anti-Cdc37, or
anti-CDK4 in each experiment. (B) Cellular thermal shift assay of HCT116 cells
treated with DMSO or 20 μM of 18h. β-actin was determined as a loading control.
Data are representative of three independent experiments. The gray levels were used
o
to calculate the T value of 18h and DMSO, indicating a ΔT value of + 6.39 C. (C)
m
m
Dose-dependent and time-dependent western blot analysis of Hsp90, Hsp70, Cdc37,
p–Cdc37, GR, p-GR, CDK4, CDK6, Cyclin D1, Cyclin D3, p21 and p27 protein
expression levels in HCT116 cells after treatment with 0, 1, 5 and 10 μM 18h for 24
hours or treatment with 10 μM 18h for 0, 4, 12, 24 and 48 hours. β-Actin was used as
a loading control. Data are representative of three independent experiments.
18h Induced Cell Cycle Arrest. As previously reported, Hsp90-Cdc37 PPI played a
20
ACS Paragon Plus Environment

Page 21 of 50
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
significant role in the correct folding of CDK4 and exhibited high correlation with
cell cycle process and related regulatory effectors16, 32. Treatment with 18h efficiently
inhibited CDK4/6 and its downstream effectors (Figure 7C). Based on this result, we
subsequently performed a cell cycle distribution analysis following a standard flow
cytometry method. HCT116 cells were treated with 18h (0, 5, 10 μM) for 24 hours or
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
4
8 hours, respectively, by staining DNA with propidium iodide (PI). 18h
concentration-dependently and time-dependently enhanced the percentages of
HCT116 cells in G -G phase, demonstrating a mechanism of G -G phase arrest to
0
1
0
1
inhibit cell cycle progression (Figure 8).
Figure 8. Cell cycle analysis of compound 18h. (A) Cell cycle distribution
determined by propidium iodide (PI) staining of HCT116 cells after treatment with 0,
5 and 10 μM of 18h for 24 h or 48 h, respectively. (B) A graphic histogram of cell
cycle distribution. All data are representative of three independent experiments as the
means ± SD (*P < 0.05, **P < 0.01, ***P < 0.001).
Pharmacokinetic Profiles of 18h and In Vivo Antitumor Efficacy. As shown in
Table 8, we determined the pharmacokinetic profiles of 18h using a single-dose (80
21
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 22 of 50
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
mg/kg, po) in mice bearing HCT116 xenografts tumors. The concentrations of 18h
both in plasma and tumor tissues were measured to indicate favorable properties
including the exposure of 18h (area under the curve) as 533.4 hours·g/ml in plasma
and 132.3 hours·g/ml in tumor tissue, moderate clearances (186.5 ml/h/kg in plasma
and 167.9 ml/h/kg in tumor tissue) and reasonable half-life values (t1/2 = 1.46 hours in
plasma and 4.66 hours in tumor tissue). Thus, the ideal pharmacokinetic profiles of
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
8h led to favorable druglike properties for further efficacy studies in vivo.
The in vivo antitumor efficacy of 18h (DDO-5636 was used for comparison) was
evaluated in the same model used in pharmacokinetic assay. In this study, the
hydrochloride salt of 18h and sodium salt of DDO-5936 were daily treated by
intragastric administration at 40 and 80 mg/kg, respectively. As shown in Figure 9A,
compound 18h caused inhibition of tumor growth significantly in a dose-dependent
manner. DDO-5936 exhibited no inhibition effects at low dose (40 mg/kg), indicating
an obviously weak potency compared with 18h, although it showed a moderate effect
at high dose group (80 mg/kg). Both DDO-5936 and 18h were well tolerated for the
absence of serious weight loss during the whole assay (Figure 9B). The results of the
in vivo antitumor efficacy revealed consistence with previous data.
Hematoxylin-eosin (H&E) staining of 18h-treated and DDO-5936 high dose groups
showed that tumor cells in the xenografts decreased significantly, while DDO-5936
low dose group exhibited similar results of control group with no inhibition effect
(Figure 9C). We also observed a significant decrease of CDK4 in 18h-treated and
DDO-5936 high dose groups by immunohistochemical staining, while DDO-5936
low dose group remained no inhibition of CDK4. In addition, both DDO-5936 and
1
8h exhibited little influence on different organs of mice (Figure S4). Thus, these
results demonstrated that 18h showed improved oral availability, compared with
DDO-5936, to suppress HCT116 xenografted tumor growth in vivo.
22
ACS Paragon Plus Environment

Page 23 of 50
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Figure 9. 18h inhibited xenograft tumor growth in immunocompromised mice.
(A) In vivo antitumor efficacy of DDO-5936 and 18h in HCT116 xenograft model.
Mice were administered DDO-5936 and 18h daily (0, 40, 80 mg/kg, respectively) by
gavage starting 7 days after cell implantation. Tumor volumes were determined by
*
calipers. The tumor volumes are plotted as the mean ± s.e.m. ( P<0.05; ** P <0.01; ***
P <0.001, n = 6). (B) The effects on body weight changes of the mice treated with
DDO-5936 and 18h. (C) H&E staining showed the tumor tissue morphology of
different groups with DDO-5936 and 18h. Scale bars were represented as 50 μm. The
expression levels of CDK4 were analyzed in tumor tissues by immunohistochemistry.
Table. 8 Pharmacokinetic profiles of 18h.
Parameters
18h
Plasma
0.5
Tumor
1
T
max (h)
1/2 (h)
t
1.46
4.66
50.0
C
max(μg/ml)
307.5
23
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 24 of 50
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
AUC0-t(h*μg/ml)
533.4
536.2
391.6
186.5
1.46
132.3
167.9
-
AUC0-∞(h*μg/ml)
Vz_F(ml/kg)
Cl_F(ml/h/kg)
MRT (h)
595.7
6.02
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
CONCLUSIONS
Although Hsp90 ATPase inhibitors have been considered effective in anti-cancer
treatment, such side effects as inevitable heat shock response, limited efficacy and
diverse toxicities are unavoidable, which limit their clinical usage. In the Hsp90
chaperone cycle, the specific functions are achieved by diverse protein-protein
interactions with co-chaperones. Cdc37 has been regarded as a kinase-specific
co-chaperone with important roles for both Hsp90 and kinases. Different with directly
inhibiting ATPase, targeting Hsp90-Cdc37 PPI reveals as an alternative strategy to
block the maturation of kinases to achieve cancer therapy with specificity and safety.
Development of small molecules targeting Hsp90-Cdc37 PPI shows promising
opportunities to understand the Hsp90-Cdc37 chaperone cycle for kinase recruitment.
Considering the totally different regulation mechanism, these modulators might be
used in combination with other kinase inhibitors to explore more potential utility. In
this study, the integration of medicinal chemistry and multiple biochemical tests along
with computational methods was used to develop more potent Hsp90-Cdc37 PPI
inhibitors (Figure 10). Based on an identified binding mode of compound 11, the
structure-activity and structure-property relationships of Hsp90-Cdc37 PPI inhibitors
were explored to obtain a series of compound exhibiting Hsp90-Cdc37 inhibition
potency, resulting in lead compound 18h with best in vitro binding affinity, improved
anti-proliferative activity, preferable stability in plasma and microsome, as well as
oral efficacy in vivo, compared with our previously reported compound DDO-5936.
In this section, MD simulation and docking models were also employed to guide and
explain the results of the compounds. The results of both in vitro and in vivo assays
24
ACS Paragon Plus Environment

Page 25 of 50
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
confirmed that compound 18h could efficiently block the Hsp90-Cdc37 PPI through a
direct binding mode, leading to a specific down-regulation of kinase clients of Hsp90
and cell cycle arrest to achieve potent anti-proliferative activities. In HCT116
xenografts, 18h showed promising tumor growth inhibition with no weight loss by
oral administration. In summary, 18h could be regarded as an orally active
Hsp90-Cdc37 PPI inhibitor and a promising candidate agent for the treatment of
cancer through an alternative way to regulate the function of the molecular chaperone.
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. Schematic overview of the workflow, including the combination of
computational methods, medicinal chemistry and therapeutic efficacy
evaluation.
EXPERIMENTAL SECTION
General Information. Unless otherwise mentioned, all of the commercial reagents
were directly used without further purification. All the organic solutions were
o
concentrated using an evaporator (BüchiRotavapor) below 60 C under reduced
pressure. All the reactions were visualized using UV light and monitored by thin layer
chromatography (TLC) on 0.25 mm silica gel plates (GF-2.5). Melting points of all
the compounds were detected by a MP50 Melting Point System. A Bruker AV-300
25
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 26 of 50
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
1
13
instrument was used to determine the H-NMR and C-NMR, deuterated solvents
with tetramethysilane (TMS) as internal standard. ESI-MS and high-resolution mass
spectra (HRMS) results of all the compounds were recorded on a Water Q-Tof micro
mass spectrometer (all data were within 0.4 % of the theoretical values). The purity
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
(
≥95%) of the compounds was determined by the HPLC on an Agilent C18 column
(
4.6 mm × 150 mm, 3.5 μm) using a mixture of solvent containing methanol/water at
a flow rate of 0.5 mL/min and the peaks were determined at 254 nm under UV.
Chemistry
Scheme 1. Synthetic Route of Compounds 12 a-i.
NH2
H
N
H
N
H
N
H
N
N
Cl
N
N
N
N
N
N
(a)
(b)
(c)
(d)
O
O
N
N
S
N
HN
Boc
N
R
N
HN
Boc
H2N
H
HN
Boc
9
10
11
12 a-i
o
Reagents and conditions: (a) 2,4-dichloropyrimidine, i-PrOH, DIPEA, 80 C, 10 h,
o
7
2%; (b) tetrahydropyrrole, 1,4-dioxane, DIPEA, 60 C, 6 h, 93%; (c) CF COOH,
3
DCM, rt, 70%; (d) different substituted sulfonyl chloride, toluene or DCM, TEA, rt,
5%.
6
General procedures for the synthesis of 9-11. 2,4-dichloropyrimidine (3g, 20
mmol) and tert-butyl (4-aminophenyl) carbamate were dissolved in the solution of
o
i-PrOH (20 mL) and DIPEA (1mL) to react for 10 h at 80 C. The reaction solvent
was concentrated to 2 ml by rotary evaporator below 40 °C under reduced pressure
before adding 50 mL water to make product precipitated. The product was filtered
and washed with 50 mL water, then dried overnight in a vacuum desiccator, yielding
9 as a white solid. Then, 9 (1g, 3 mmol) was dissolved in the solution of 1,4-dioxane
(
15 mL) and DIPEA (0.5 mL) before tetrahydropyrrole (350 mg, 5 mmol) was added.
o
The reaction was carried out at 60 C for 6 h and then the solvent was removed by
rotary evaporator below 60 °C under reduced pressure. The oily crude product was
the dissolved in 1 mL DMF before adding 50 mL water to make product precipitated.
26
ACS Paragon Plus Environment

Page 27 of 50
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Crude product of 10 was filtered and washed with 50 mL water, then dried overnight
in a vacuum desiccator. Finally, 10 (700 mg, 2 mmol) was dissolved in DCM (15 mL)
and followed by adding trifluoroacetic acid (1 mL) to react for 4 h at room
temperature. Then the solvent in the reaction mixture was removed by rotary
evaporator below 40 °C under reduced pressure to obtain white solid as the final
product 11 and dried overnight in a vacuum desiccator.
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
tert-butyl (4-((2-chloropyrimidin-4-yl)amino)phenyl) carbamate (9): Yield:
1
7
7
2% as white solid. Mp 142-143 °C. H-NMR (300 MHz, DMSO, δ) 9.58 (s, 1H),
.88 (s, 1H), 7.41 (d, 2H, J = 6.0Hz), 7.20 (d, 2H, J = 8.36Hz), 6.10 (d, 1H, J = 4.52
+
Hz), 1.62 (s, 9H). HRMS (ESI): found 321.7863 (C H ClN O , [M+H] , requires
1
5
17
4
2
3
21.7819).
tert-butyl (4-((2-(pyrrolidin-1-yl)pyrimidin-4-yl)amino)phenyl)carbamate (10):
1
Yield 93% as light yellow solid; Mp 162-163 °C; H-NMR (300 MHz, DMSO, δ)
9.63 (s, 1H), 7.78 (d, 1H, J = 4.52 Hz), 7.47 (d, 2H, J = 4.36 Hz), 7.26 (d, 2H, J =
6.06Hz), 5.93 (d, 1H, J = 3.12 Hz), 3.18 (m, 4H), 2.01 (m, 4H), 1.42 (s, 9H). HRMS
+
(ESI): found 356.4432 (C H N O , [M+H] , requires 356.4498).
1
9
25
5
2
1
N -(2-(pyrrolidin-1-yl)pyrimidin-4-yl)benzene-1,4-diamine (11): Yield 70% as
1
white solid; Mp 151-153 °C; H-NMR (300 MHz, DMSO, δ) 7.83(d, 1H, J = 3.64
Hz), 7.07 (d, 2H, J = 8.42 Hz), 6.26 (d, 2H, J = 7.22 Hz), 5.91 (d, 1H, J = 3.2 Hz),
4
.38 (s, 2H), 3.12 (m, 4H), 1.88 (m, 4H). HRMS (ESI): found 256.3282 (C H N ,
14 17 5
+
[
M+H] , requires 256.3277).
General procedures for preparation of 12 a-g. 11 (500 mg, 2 mmol) was
dissolved in the solution of toluene (10 mL) and TEA (0.5 mL) before different
benzensulfonyl chloride was added, respectively, including
-(trifluoromethyl)-benzenesulfonyl chloride (488 mg, 2 mmol) to obtain 12a,
naphthalene-2-sulfonyl chloride (453 mg, mmol) to obtain 12b,
,4,6-trimethylbenzenesulfonyl chloride (440 mg, 2 mmol) to obtain 12c,
4
2
2
4
4
-methoxybenzenesulfonyl chloride (412 mg, 2 mmol) to obtain 12d,
-(tert-butyl)benzenesulfonyl chloride (464 mg, 2 mmol) to obtain 12e,
27
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 28 of 50
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
2
2
,4-dimethoxybenzenesulfonyl chloride (472 mg, 2 mmol) to obtain 12f and
,5-dimethoxybenzenesulfonyl chloride (472 mg, 2 mmol) to obtain 12g. The reaction
was carried out at room temperature until the crude solid was precipitated. The final
product was filtered and washed with 50 mL water, then dried overnight in a vacuum
desiccator, yielding 12 a-g (55%-88%).
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
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-(4-((2-(pyrrolidin-1-yl)pyrimidin-4-yl)amino)phenyl)-4-(trifluoromethyl)
1
benzenesulfonamide (12a): Yield 64% as light yellow solid; Mp 236-237 °C; H
NMR (300 MHz, DMSO) δ 10.20 (s, 2H), 8.86 (s, 1H), 7.96 (d, J = 8.6 Hz, 2H), 7.90
(d, J = 8.7 Hz, 2H), 7.47 (d, J = 8.7 Hz, 2H), 6.95 (d, J = 8.7 Hz, 2H), 5.45 (s, 1H),
+
2
.25 (s, 4H), 1.89 (s, 4H). HRMS (ESI): found 464.4856 (C H F N O S, [M+H] ,
2
1
20
3
5
2
requires 464.4852). HPLC (90:10 methanol:water with 1‰ TFA): t = 3.87 min,
R
9
7.85%.
N-(4-((2-(pyrrolidin-1-yl)pyrimidin-4-yl)amino)phenyl)naphthalene-2-sulfona
1
mide (12b): Yield 55% as yellow solid; Mp 252-254 °C; H NMR (300 MHz,
DMSO) δ 10.07 (s, 1H), 8.87 (s, 1H), 8.36 (s, 1H), 8.10 (t, J = 8.0 Hz, 2H), 8.00 (d, J
= 7.6 Hz, 1H), 7.74 (m, 4H), 7.38 (m, 2H), 6.98 (d, J = 8.7 Hz, 2H), 5.54 (s, 1H), 3.24
+
(m, 4H), 1.89 (s, 4H). HRMS (ESI): found 446.1578 (C H N O S [M+H] , requires
2
4
23
5
2
446.1572). HPLC (90:10 methanol:water with 1‰ TFA): t = 4.21 min, 96.43%.
R
2
,4,6-trimethyl-N-(4-((2-(pyrrolidin-1-yl)pyrimidin-4-yl)amino)phenyl)
1
benzenesulfonamide (12c): Yield 55% as yellow solid; Mp 189-191 °C; H NMR
(300 MHz, DMSO) δ 10.10 (s, 1H), 7.88 (d, J = 3.52 Hz, 1H), 7.23 (s, 2H), 7.08 (d, J
=
3.0 Hz, 2H), 6.88 (d, J = 5.84 Hz, 2H), 6.04 (d, J = 4.32 Hz, 1H), 3.27 (m, 4H), 1.74
+
(m, 4H). HRMS (ESI): found 438.1898 (C H N O S [M+H] , requires 438.1885).
2
4
27
5
2
HPLC (90:10 methanol:water with 1‰ TFA): t = 3.64 min, 96.12%.
R
4
-methoxy-N-(4-((2-(pyrrolidin-1-yl)pyrimidin-4-yl)amino)phenyl)
benzenesulfonamide (12d): Yield 88% as light grey solid; Mp 213-215 °C; H NMR
300 MHz, DMSO) δ 9.80 (s, 1H), 8.89 (s, 1H), 8.08 (s, 1H), 7.63 (d, J = 9.0 Hz, 2H),
1
(
7
1
.40 (d, J = 9.0 Hz, 2H), 7.03 (d, J = 9.0 Hz, 2H), 6.93 (d, J = 9.0 Hz, 2H), 5.57 (s,
H), 3.78 (s, 3H), 1.89 (m, 4H). HRMS (ESI): found 426.1527 (C H N O S
2
1
23
5
3
28
ACS Paragon Plus Environment

Page 29 of 50
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
+
[
M+H] , requires 426.1522). HPLC (90:10 methanol:water with 1‰ TFA): t = 3.87
R
min, 97.85%.
-(tert-butyl)-N-(4-((2-(pyrrolidin-1-yl)pyrimidin-4-yl)amino)phenyl)
4
1
benzenesulfonamide (12e): Yield 62% as white solid; Mp 217-219 °C; H NMR
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
(
300 MHz, DMSO) δ 9.95 (s, 1H), 8.91 (s, 1H), 8.08 (s, 1H), 7.66 (d, J = 8.4 Hz, 2H),
7
.56 (d, J = 8.4 Hz, 2H), 7.41 (d, J = 8.8 Hz, 2H), 6.99 (d, J = 8.7 Hz, 2H), 5.58 (s,
+
1
H), 1.90 (s, 4H), 1.26 (s, 9H). HRMS (ESI): found 452.2049 (C H N O S [M+H] ,
2
4
29
5
2
requires 452.2042). HPLC (90:10 methanol:water with 1‰ TFA): t = 4.13 min,
R
9
8.42%.
,4-dimethoxy-N-(4-((2-(pyrrolidin-1-yl)pyrimidin-4-yl)amino)phenyl)
2
1
benzenesulfonamide (12f): Yield 61% as white solid; Mp 225-226 °C; H NMR (300
MHz, DMSO) δ 9.53 (s, 1H), 8.84 (s, 1H), 8.07 (s, 1H), 7.59 (d, J = 8.6 Hz, 1H), 7.35
(
d, J = 8.8 Hz, 2H), 6.95 (d, J = 8.7 Hz, 2H), 6.64 (s, 1H), 6.54 (d, J = 8.8 Hz, 1H),
5.55 (s, 1H), 3.90 (s, 3H), 3.79 (s, 3H), 1.90 (s, 4H). HRMS (ESI): found 456.1641
+
(C H N O S [M+H] , requires 456.1627). HPLC (90:10 methanol:water with 1‰
2
2
25
5
4
TFA): t = 4.03 min, 97.66%.
R
2,5-dimethoxy-N-(4-((2-(pyrrolidin-1-yl)pyrimidin-4-yl)amino)phenyl)
1
benzenesulfonamide (12g): Yield 58% as white solid; Mp 228-229 °C; H NMR
300 MHz, DMSO) δ 9.68 (s, 1H), 8.87 (s, 1H), 8.07 (s, 1H), 7.38 (d, J = 8.3 Hz, 2H),
(
7
.17 (s, 1H), 7.12 (s, 1H), 6.98 (d, J = 8.2 Hz, 2H), 5.56 (s, 1H), 3.86 (s, 3H), 3.69 (s,
H), 3.34 – 3.23 (m, 4H), 1.90 (s, 4H). HRMS (ESI): found 456.1630 (C H N O S
3
22 25
5
4
+
[
M+H] , requires 456.1627). HPLC (90:10 methanol:water with 1‰ TFA): t = 4.09
R
min, 96.85%.
Scheme 2. Synthetic Route of Compounds 15a-g.
H
H
N
H
N
NH2
NH2
N
N
Cl
N
N
N
N
O
O
O
O
O
O
O
O
S
(c)
S
N
S
N
(
e)
S
N
(
(
a)
b)
(d)
N
N
N
N
H
H
H
(
f)
R
HN
Boc
13
14
12c
15 a-g
29
ACS Paragon Plus Environment