Discovery of novel 20S proteasome inhibitors by rational topology-based scaffold hopping of bortezomib

Article abstract of DOI:10.1016/j.bmcl.2018.05.018

A series of structurally novel proteasome inhibitors 1–12 have been developed based rational topology-based scaffold hopping of bortezomib. Among these novel proteasome inhibitors, compound 10 represents an important advance due to the comparable proteasome-inhibitory activity (IC₅₀ = 9.7 nM) to bortezomib (IC₅₀ = 8.3 nM), the remarkably higher BEI and SEI values and the effectiveness in metabolic stability. Therefore, compound 10 provides an excellent lead suitable for further optimization.

Full text of DOI:10.1016/j.bmcl.2018.05.018

Accepted Manuscript
Discovery of novel 20S proteasome inhibitors by rational topology-based scaf-
fold hopping of bortezomib
Yulong Xu, Xicheng Yang, Yiyi Chen, Hao Chen, Huijiao Sun, Wei Li, Qiong
Xie, Linqian Yu, Liming Shao
PII:
S0960-894X(18)30417-7
https://doi.org/10.1016/j.bmcl.2018.05.018
BMCL 25835
DOI:
Reference:
Bioorganic & Medicinal Chemistry Letters
To appear in:
Received Date:
Revised Date:
Accepted Date:
26 March 2018
25 April 2018
8 May 2018
Please cite this article as: Xu, Y., Yang, X., Chen, Y., Chen, H., Sun, H., Li, W., Xie, Q., Yu, L., Shao, L., Discovery
of novel 20S proteasome inhibitors by rational topology-based scaffold hopping of bortezomib, Bioorganic &
Medicinal Chemistry Letters (2018), doi: https://doi.org/10.1016/j.bmcl.2018.05.018
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers
we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and
review of the resulting proof before it is published in its final form. Please note that during the production process
errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Discovery of novel 20S proteasome inhibitors by rational
topology-based scaffold hopping of bortezomib
Yulong Xua, Xicheng Yanga, Yiyi Chena, Hao Chena, Huijiao Suna, Wei Lia, Qiong
Xiea, Linqian Yua and Liming Shao*,a,b
a
School of Pharmacy, Fudan University, 826 Zhangheng Road, Zhangjiang Hi-tech
Park, Pudong, Shanghai 201203, China
b
State Key Laboratory of Medical Neurobiology, Fudan University, 138 Yixueyuan
Road, Shanghai 200032, China
 Corresponding author. Tel.: +86 021 5198 0201; fax: +86 021 5198 0201; e-mail:
limingshao@fudan.edu.cn (L. Shao)
Abstract: A series of structurally novel proteasome inhibitors 1-12 have been
developed based rational topology-based scaffold hopping of bortezomib. Among
these novel proteasome inhibitors, compound 10 represents an important advance due
to the comparable proteasome-inhibitory activity (IC50 = 9.7 nM) to bortezomib (IC50
= 8.3 nM), the remarkably higher BEI and SEI values and the effectiveness in
metabolic stability. Therefore, compound 10 provides an excellent lead suitable for
further optimization.
Keywords: Proteasome inhibitors; Topology-based; Scaffold hopping; Bortezomib

In eukaryotic cells, there are two different pathways for protein degradation: the
lysosomal pathway and the ubiquitin-proteasome pathway (UPP). The lysosomal
pathway can provide amino acids as building materials for protein synthesis by
breaking down endogenous and exogenous proteins in a relatively nonspecific manner.
In contrast, the UPP is the major pathway of endogenous proteins degradation. The
UPP, which plays a critical role in recognizing and degrading abnormal and misfolded
proteins, is responsible for maintenance of an intracellular protein expression.1,2 In
this pathway, the 26S proteasome is the main proteolytic component that consists of
two subcomplexes with different functions: the 19S regulatory particles (RPs) and the
20S core particle (CP). The 19S RPs consists of multiple subunits and caps the 20S
CP at one or both ends. The barrel-shaped catalytic 20S CP is composed of 28 protein
subunits in α1-7β1-7β1-7α1-7 arrangement (Fig. 1A).3,4 In 20S proteasome, three different
catalytic subunits are harbored in β-subunits and are classified on the basis of the
amino acid after which they cleave the peptide bond: caspase-like (C-L, β1 subunit),
which cleaves mainly after acidic amino acids; trypsin-like (T-L, β2 subunit), which
cleaves after basic amino acids; chymotrypsin-like (CT-L, β5 subunit), which cleaves
after hydrophobic residues.5 In vertebrate cells, there exist three different types of
proteasome subtype:
constitutive proteasome,
immunoproteasome,
and
thymoproteasome.6 These three proteasome subtypes differ by the incorporation of
three distinct sets of catalytic subunits: the constitutive proteasome incorporates
subunits β1, β2, and β5, the immunoproteasome incorporates subunits β1i, β2i, and
β5i, and the thymoproteasome incorporates subunits β1i, β2i, and β5t (Fig. 1B).7 All

of these hydrolytic activities are manifested through a threonine residue (Oγ-Thr1),
which acts as a nucleophile in peptide bond hydrolysis.8 Protein degradation by the
UPP is initiated by the labeling of targeted proteins with polyubiquitin chains. Then,
polyubiquitin-labeled proteins are taken up to a hole inside the 20S proteasome and
were hydrolyzed by proteasomal catalytic active sites.9
Fig. 1. (A) The 20S proteasome. (B) Schematic representation of subunit composition
of three proteasome subtypes.
The 20S proteasome is validated as a valuable anticancer target because
proteasome inhibition causes an accumulation of unnecessary proteins and induces
apoptosis.10 Proteasome inhibitor bortezomib (Velcade), a dipeptide boronic acid, is a
frontline drug approved by the U.S. Food and Drug Administration (FDA) for the
treatment of relapsed multiple myeloma (MM).11 The FDA approval of tetrapeptidyl
epoxyketone carfilzomib (Kyprolis) for the treatment of relapsed MM patients who
have received at least two prior therapies further confirmed that targeting the UPP is a
viable route for the treatment of human cancers.12 Besides the bortezomib and
carfilzomib, several structurally diverse proteasome inhibitors have been discovered
and developed by screening chemical libraries containing synthetic and natural small

molecules and by chemical modifications of lead compounds (Fig. 2).13 However,
despite initial clinical successes of bortezomib and carfilzomib, these two drugs have
several shortcomings such as side effects, drug resistance and unsatisfactory
pharmacokinetic profiles.14 Therefore, there is an urgent need for novel proteasome
inhibitors as potential chemotherapeutic agents.
Fig. 2. Structures of approved and currently developed proteasome inhibitors.
As we all know, scaffold hopping has been identified as a promising strategy for
discovering structurally novel compounds.15 Recently, Sun and co-workers classified
scaffold hopping approaches into four major categories by focusing on the degree of
change associated with the original parent molecule.16 However, there are very few
success examples by using topology-based scaffold hopping approach, which is the
highest hop level of scaffold hopping approaches. Therefore, our effort has been
devoted to design and synthesis of a series of structurally novel highly potent 20S
proteasome inhibitors based topology-based scaffold hopping of bortezomib in the
past few years, hoping to overcome the reported shortcomings.

The X-ray crystal structure of 20S proteasome in complex with bortezomib is
depicted in Fig. 3 (PDB ID: 2F16), which clearly shows that bortezomib only binds to
the nonprimed binding site of proteasome.17 Moreover, boronic acid warhead of
bortezomib covalently binds to Thr hydroxyl group of β5 subunit, which is the most
important binding site in its interaction to keep the biological activities. The
P1-leucine side chain is found to occupy the S1 pocket deeply and fully. The
phenylalanine moiety in P2 position points into the vacant primed binding site. The
pyrazine-2-carboxyl group in P3 position makes bortezomib form strong hydrophobic
interactions with the S3 deep hydrophobic pocket. This structural information
suggests that the antiparallel β sheet conformation and the boronic acid warhead of
bortezomib are essential for the strong binding to the β5 subunit of 20S proteasome.
Furthermore, according to the literature, the boronic acid warhead is very important to
reversibly interact with the hydroxyl group of N-terminal Thr in β5 subunit. On this
point, reversible inhibition of endogenous proteases is advantageous for cancer
treatment, while irreversible blockage of parasitic proteases is desirable.18
Fig. 3. X-ray crystal structures of 20S proteasome in complex with bortezomib (green
tube, PDB ID: 2F16).

Based on the structural information and previous studies, we designed and
synthesized the structurally novel proteasome inhibitors 1-12, in which the
pyrazine-2-carboxyl group in P3 position of bortezomib was replaced with aryl
ether/amine group (short-linker scaffold, 1-6) and aryl methyl ether/amine group
(long-linker scaffold, 7-12) (Fig. 4A). The results of alignment were depicted in Fig.
4B and 4C, these compounds can be superimposed onto the binding conformation of
bortezomib and almost took the same interaction mode as bortezomib. Furthermore,
the boronic acid warhead, the P1-leucine side chain and the P2-phenylalanine moiety
of the newly designed compounds were well superimposed on each other and had no
steric repulsion with the β5 subunit of 20S proteasome. Compared with bortezomib,
these compounds adopted aryl ether/amine group and aryl methyl ether/amine group
instead of the pyrazine-2-carboxyl group at P3 position, and these variations did
maintain the strong hydrophobic interactions of P3 moiety with the S3 deep
hydrophobic pocket. All the theoretical results displayed that these newly designed
compounds can retain the strong inhibitory activity of bortezomib.

Fig. 4. (A) Design of structurally novel proteasome inhibitors 1-12 by rational
topology-based scaffold hopping of bortezomib. (B) Results of alignment of the
designed compounds 1-6 (gray tube) onto the binding conformation of bortezomib
(orange tube). (C) Results of alignment of the designed compounds 7-12 (gray tube)
onto the binding conformation of bortezomib (orange tube).
The designed compounds 1-12 were synthesized following the method described
in the literature with modifications,19-24 and the synthetic routes are summarized in
Scheme 1-5. The amino boronate hydrochloride 19 was the key intermediate in the
total synthetic route (Scheme 1). The highly optical (+)-pinanediol 13 was synthesized
from (+)-α-pinene with good yield (83%). The dimethyl dichloromethylboronate 15
was prepared by addition of trimethoxyborane to the solution of
(dichloromethyl)lithium 14 in THF generated in situ at -100 °C. Transesterification of
compound 15 with (+)-pinanediol 13 gave dichlorosubstituted borate ester 16 in 91%
yield. Monochlorosubstituted boronate 17 was obtained in 79% yield with the reaction
of compound 16 and isobutylmagnesium bromide in the presence of anhydrous zinc
chloride as catalyst. Treatment of compound 17 with lithium bis(trimethylsilyl) amide

gave protected amine 18, which was directly used for the next step without further
purification. Deprotection of the trimethylsilyl group in compound 18 afforded
desired amino boronate hydrochloride 19 in 44% yield.
Scheme 1. Synthesis of amino boronate hydrochloride 19. Reagents and conditions: (i)
OsO4, t-butyl alcohol, trimethylamine N-oxide dihydrate, pyridine, water, reflux, 24 h;
(ii) THF, n-BuLi, -100 °C, 1 h; (iii) B(OCH3)3, -100 °C to rt, 1 h; (iv) THF, rt, 24 h; (v)
(CH3)2CHCH2MgBr, Et2O, anhydrous ZnCl2, -78 °C to rt, overnight; (vi)
LiN(SiMe3)2, THF, -78 °C to rt, overnight; (vii) petroleum ether, dry HCl in Et2O,
-78 °C to rt, overnight.
The intermediates 23a-23c with short O-linker were synthesized via a Mitsunobu
reaction of the corresponding phenols 20a-20c with (R)-methyl phenyllactate 21,
followed by saponification and acidification (Scheme 2). The intermediates 26a-26c
with short N-linker were readily prepared by coupling of L-phenylalanine 25 with
corresponding aryl bromides 24a-24c under the catalysis of CuI (Scheme 3). The
intermediates 31a-31c with long O-linker were obtained by the use of corresponding
arylmethyl trichloroacetimidate 28a-28c as the arylmethylating agent to react with

methyl L-3-phenyllactate 29 and subsequent acid-catalyzed hydrolysis of the methyl
ester (Scheme 4). As depicted in Scheme 5, L-phenylalanine 25 was monobenzylated
using arylmethyl chlorides 32a-32c in water to form the desired intermediates
33a-33c with long N-linker.
Scheme 2. Synthesis of intermediates 23a-23c. Reagents and conditions: (i) PPh3,
DIAD, toluene, 0 °C to rt, overnight; (ii) 1 N NaOH, THF, rt, 3-5 h.
Scheme 3. Synthesis of intermediates 26a-26c. Reagents and conditions: (i) CuI,
K2CO3, DMA, 90 °C, 36-48 h.
Scheme 4. Synthesis of intermediates 31a-31c. Reagents and conditions: (i) NaH,
trichloroacetonitrile, Et2O, rt, 2-5 h; (ii) CF3SO3H, CH2Cl2/hexane (1:1), rt, 8-12 h;
(iii) 2 N HCl, dioxane, 70 °C, 5-8 h.

Scheme 5. Synthesis of intermediates 33a-33c. Reagents and conditions: (i) K2CO3, 6
N HCl, H2O, rt, 20-24 h.
General synthetic pathway to obtain the boronic acids 1-12 was presented in
Scheme 6. Coupling of amino boronate hydrochloride 19 with intermediates 23, 26,
31, 33 in the presence of EDC, HOBt and DIPEA afforded the boronates 34a-34l,
which were used directly for acid-catalyzed transesterification with
2-methylpropylboronic acid to provide designed compounds 1-12 in moderate yields.
Scheme 6. Synthesis of designed compounds 1-12. Reagents and conditions: (i)
EDC·HCl, HOBt, DIPEA, CH2Cl2, -15 °C to rt, 18-20 h; (ii) 2-methylpropylboronic
acid, 1 N HCl, MeOH/hexane (1:1), rt, 4-5 h.
The inhibitory effects of the synthesized compounds 1-12 and bortezomib on the
CT-L activity of human 20S proteasome were measured using the fluorogenic
substrate Suc-LLVY-AMC, and the results are summarized in Table 1. All of these
compounds strongly inhibited 20S proteasome CT-L activity, with IC50 values ranging

from 9.7 to 335 nM. The inhibitory potency of 7-12 on CT-L activity (IC50 = 9.7-109
nM) was comparable to bortezomib (IC50 = 8.3 nM), but 1-6 exhibited a diminished
capacity to inhibit the proteasome (IC50 = 126-335 nM), suggesting that the
long-linker P3 group was more favored than the short-linker P3 group to be
accommodated in the S3 deep hydrophobic pocket of the β5 subunit. In these
compounds, the types of aryl group and linked atom had only weak impacts on their
proteasome inhibitory activities.
Subsequently, the binding efficiency index (BEI) and the surface-binding
efficiency index (SEI) were calculated in order to intensively evaluate the influences
of the topology-based scaffold hopping on the molecular properties (Table 1). BEI and
SEI are defined as pIC50 per molecular weight (kDa) and pIC50 per van der Waals
polar surface area (where 100 Å2 is used as a normalizing factor for PSA values),
respectively.25 The importance of molecular weight and PSA to membrane
permeability and oral bioavailability has been described in several publications.26-28 In
general, it is very useful to examine the BEI and SEI values simultaneously during the
drug optimization process, and the desirable lead compounds generally have high BEI
and SEI values.25 As shown in Fig. 5, which was the mapping of BEI-SEI for
structurally novel proteasome inhibitors 1-12 and bortezomib, the BEI values of these
newly proteasome inhibitors were slightly lower than bortezomib except 10, but their
SEI values were comparable with or superior to bortezomib. These results indicated
that the approach of topology-based scaffold hopping made these newly proteasome

inhibitors exhibit a smaller PSA, which kept the SEI values high. Furthermore, 10
showed remarkably higher BEI and SEI values compared with bortezomib,
demonstrating that this structurally novel proteasome inhibitor was clearly superior to
bortezomib, and this rational hopping may change the unsatisfactory pharmacokinetic
profiles of bortezomib.
Table 1. CT-L inhibitory activities and BEI and SEI values of structurally novel
proteasome inhibitors 1-12 and bortezomib.
Compound
IC50 (nM)a
BEIb
SEIc
126
275
155
335
166
192
19
16
19
18
17
19
9
8
7
8
8
6
1
2
3
4
5
6
36
109
53
20
17
20
22
17
19
9
9
7
8
7
9
10
11
12
9.7
92
10
9
65
7
Bortezomibd
8.3
21
7
a
b
IC50 Values represent the mean of at least two independent determinations. pIC50
c
per molecular weight (kDa). pIC50 per van der Waals polar surface area (PSA,
d
Calculated by ChemBioDraw Ultra 12) normalized to 100 Å2. IC50 value obtained

for bortezomib under our experimental conditions.
24
22
B o rte z o m ib
20
18
16
14
C o m p o u n d 1 0
S h o rt-lin k e r sc a ffo ld
L o n g -lin k e r sc a ffo ld
6
8
10
12
14
16
S u r fa c e -b in d in g e ffic ie n c y in d e x (S E I)
Fig. 5. Mapping of the surface-binding and binding efficiency indices for structurally
novel proteasome inhibitors 1-12 and bortezomib.
Given that bortezomib is prone to rapid in vivo inactivation,29 we next examined
whether the structurally novel proteasome inhibitor 10 was metabolically more stable
than bortezomib. Excitingly, compound 10 demonstrated excellent metabolic stability
profiles in mouse and human liver microsomes, as compared to bortezomib (Fig. 6).
These results indicated that compound 10 was likely to have much improved in vivo
stability compared to bortezomib.
M o u se
H um an
1 2 0
1 0 0
8 0
6 0
4 0
2 0
0
1 2 0
1 0 0
8 0
6 0
4 0
2 0
0
C o m p o u n d 1 0
C o m p o u n d 1 0
B o rte z o m ib
B o rte z o m ib
0
1 0
2 0
3 0
4 0
0
1 0
2 0
3 0
4 0
T im e (m in )
T im e (m in )
Fig. 6. Metabolic stability of compound 10 and bortezomib tested using mouse and

human liver microsomes.
In conclusion, we have successfully developed the structurally novel proteasome
inhibitors 1-12 based rational topology-based scaffold hopping of bortezomib. In
addition to their novel scaffold, these newly inhibitors exhibit high 20S proteasome
inhibitory activities. Among these structurally novel proteasome inhibitors, compound
10 represents an important advance due to the comparable proteasome-inhibitory
activity to bortezomib, the remarkably higher BEI and SEI values and the
effectiveness in metabolic stability. Therefore, the metabolically stable 20S
proteasome inhibitor 10 provides an excellent lead suitable for further optimization.
Additional studies regarding the pharmacokinetics and pharmacodynamics of
compound 10 are ongoing.
Acknowledgments
This work was financially supported by the National Basic Research Program of
China (973 Program, 2015CB931804).
Supplementary data
Supplementary data associated with this article can be found, in the online
version.
References

1. Groll M, Huber R. Methods Enzymol. 2005;398:329.
2. Hershko A, Ciechanover A. Annu. Rev. Biochem. 1998;67:425.
3. Groll M, Ditzel L, Lowe J, Stock D, Bochtler M, Bartunik HD, Huber R. Nature.
1997;386:463.
4. Lowe J, Stock D, Jap B, Zwickl P, Baumeister W, Huber R. Science. 1995;268:533.
5. Groll M, Heinemeyer W, Jager S, Ulrich T, Bochtler M, Wolf DH, Huber R. Proc.
Natl. Sci. USA 1999;96:10976.
6. Groll M, Clausen T. Curr. Opin. Struct. Biol. 2003;13:665.
7. Huber EM, Basler M, Schwab R, Heinemeyer W, Kirk CJ, Groettrup M, Groll M.
Cell. 2012;148:727.
8. Brannigan JA, Dodson G, Duggleby HJ, Moody PC, Smith JL, Tomchick DR,
Murzin AG. Nature. 1995;378:416.
9. Navon A, Ciechanover A. J. Biol. Chem. 2009;284:33713.
10. Adams J. Nature Rev. Cancer. 2004;4:349.
11. Adams J. Cancer Cell. 2004;5:417.
12. Vij R, Wang M, Kaufman JL, Lonial S, Jakubowiak AJ, Stewart AK, Kukreti V,
Jagannath S, McDonagh KT, Alsina M, Bahlis NJ, Reu FJ, Gabrail NY, Belch A,
Matous JV, Lee P, Rosen P, Sebag M, Vesole DH, Kunkel LA, Wear SM, Wong AF,
Orlowski RZ, Siegel DS. Blood. 2012;119:5661.
13. Génin E, Reboud-Ravaux M, Vidal J. Curr. Top. Med. Chem. 2010;10:232.
14. Alsina M, Trudel S, Furman RR, Rosen PJ, O’Connor OA, Comenzo RL, Wong A,
Kunkel LA, Molineaux CJ, Goy A. Clin. Cancer Res. 2012;18:4830.

15. Schneider G, Schneider P, Renner S. QSAR Comb. Sci. 2006;25:1162.
16. Sun H, Tawa G, Wallqvist A. Drug Discovery Today. 2012;17:310.
17. Kawamura S, Unno Y, List A, Mizuno A, Tanaka M, Sasaki T, Arisawa M, Asai A,
Groll M, Shuto S. J. Med. Chem. 2013;56:3689.
18. Ettari R, Tamborini L, Angelo IC, Micale N, Pinto A, De Micheli C, Conti P. J.
Med. Chem. 2013;56:5637.
19. Zhu Y, Zhao X, Zhu X, Wu G, Li Y, Ma Y, Yuan Y, Yang J, Hu Y, Ai L, Gao Q.
J. Med. Chem. 2009;52:4192.
20. Pinelli A, Godio C, Laghezza A, Mitro N, Fracchiolla G, Tortorella V, Lavecchia A,
Novellino E, Fruchart J, Staels B, Crestani M, Loiodice F. J. Med. Chem.
2005;48:5509.
21. Ma D, Zhang Y, Yao J, Wu S, Tao F. J. Am. Chem. Soc. 1998;120:12459.
22. Oscar B, Christian S, Vladimir I, Daniel W, Christer S, Bertil S, Asa R, Fredrik W,
Ingemar K. W.O.2008107365, 2008.
23. Eckenberg P, Groth U, Huhn T, Richter N, Schmeck C. Tetrahedron.
1993;49:1619.
24. Shao H, Wu Y, Li R. Synthetic Communications: An International Journal for
Rapid Communication of Synthetic Organic Chemistry. 2000;30:1911.
25. Abad-Zapatero C, Metz JT. Drug Discovery Today. 2005;10:464.
26. Egan WJ, Merz KM, Baldwin JJ. J. Med. Chem. 2000;43:3867.
27. Palm K, Stenberg P, Luthman K, Artursson P. Pharm. Res. 1997;14:568.
28. Ertl P, Rohde B, Selzer P. J. Med. Chem. 2000;43:3714.

29. Lansdell TA, Hurchla MA, Xiang J, Hovde S, Weilbaecher KN, Henry RW, Tepe
JJ. ACS Chem. Biol. 2013;8:578.
Highlights
1. Twelve structurally novel proteasome inhibitors have been developed.
2. Rational topology-based scaffold hopping of bortezomib has been employed.
3. One of the compounds represents an important advance.

Discovery of novel 20S proteasome inhibitors by rational
topology-based scaffold hopping of bortezomib
Yulong Xu a, Xicheng Yang a, Yiyi Chen a, Hao Chen a, Huijiao Sun a, Wei Li a, Qiong
Xie a, Linqian Yu a and Liming Shao *,a,b
a
School of Pharmacy, Fudan University, 826 Zhangheng Road, Zhangjiang Hi-tech
Park, Pudong, Shanghai 201203, China
b
State Key Laboratory of Medical Neurobiology, Fudan University, 138 Yixueyuan
Road, Shanghai 200032, China
Graphical abstract:
 Corresponding author. Tel.: +86 021 5198 0201; fax: +86 021 5198 0201; e-mail:
limingshao@fudan.edu.cn (L. Shao)