Design, synthesis and docking studies of novel dipeptidyl boronic acid proteasome inhibitors constructed from αα- and αβ-amino acids

Article abstract of DOI:10.1016/j.bmcl.2016.03.007

A series of novel dipeptidyl boronic acid proteasome inhibitors constructed from αα- and αβ-amino acids were designed and synthesized. Their structures were elucidated by ¹H NMR, ¹³C NMR, LC-MS and HRMS. These compounds were evaluated for their β5 subunit inhibitory activities of human proteasome. The results showed that dipeptidyl boronic acid inhibitors composed of αα-amino acids were as active as bortezomib. Interestingly, the activities of those derived from αβ-amino acids lost completely. Of all the inhibitors, compound 22 (IC₅₀ = 4.82 nM) was the most potent for the inhibition of proteasome activity. Compound 22 was also the most active against three MM cell lines with IC₅₀ values less than 5 nM in inhibiting cell growth assays. Molecular docking studies displayed that 22 fitted very well in the β5 subunit active pocket of proteasome.

Full text of DOI:10.1016/j.bmcl.2016.03.007

Bioorganic & Medicinal Chemistry Letters xxx (2016) xxx–xxx
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Design, synthesis and docking studies of novel dipeptidyl boronic
acid proteasome inhibitors constructed from aa- and ab-amino acids
Jingmiao Shi a,b,y, Meng Lei c,y, Wenkui Wu d, Huayun Feng c, Jia Wang e, Shanshan Chen e,
Yongqiang Zhu d, , Shihe Hu e, Zhaogang Liu e, Cheng Jiang a,b,
⇑
⇑
a Jiang Su Key Laboratory of Drug Design and Optimization, China Pharmaceutical University, Nanjing 210009, China
b Department of Medicinal Chemistry, School of Pharmacy, China Pharmaceutical University, Nanjing 210009, China
c College of Science, Nanjing Forestry University, Nanjing 210037, China
d College of Life Science, Nanjing Normal University, Nanjing 210023, China
e Jiangsu Chia Tai Fenghai Pharmaceutical Co. Ltd, Nanjing 210037, China
a r t i c l e i n f o
a b s t r a c t
Article history:
A series of novel dipeptidyl boronic acid proteasome inhibitors constructed from aa- and ab-amino acids
were designed and synthesized. Their structures were elucidated by 1H NMR, 13C NMR, LC–MS and HRMS.
These compounds were evaluated for their b5 subunit inhibitory activities of human proteasome. The
results showed that dipeptidyl boronic acid inhibitors composed of aa-amino acids were as active as
Received 16 January 2016
Revised 1 March 2016
Accepted 3 March 2016
Available online xxxx
bortezomib. Interestingly, the activities of those derived from
ab-amino acids lost completely. Of all
the inhibitors, compound 22 (IC50 = 4.82 nM) was the most potent for the inhibition of proteasome activ-
ity. Compound 22 was also the most active against three MM cell lines with IC50 values less than 5 nM in
inhibiting cell growth assays. Molecular docking studies displayed that 22 fitted very well in the b5 sub-
unit active pocket of proteasome.
Keywords:
Proteasome inhibitor
Dipeptidyl boronic acid
Structure–activity relationship
Cytotoxicity
Ó 2016 Published by Elsevier Ltd.
Docking study
The ubiquitin-proteasome pathway (UPP) plays a critical role in
recognizing and degrading abnormal and misfolded proteins.1,2 In
this pathway, the 26S proteasome is the main proteolytic compo-
nent, which contains two ATP-dependent 19S regulatory particles
(RPs) and one 20S core particle (CP). The catalytic 20S CP consists
of 28 protein subunits which arrange in
a1-7b1-7b1-7a1-7 four
stacked rings.3,4 In eukaryotic proteasomes, three potent prote-
olytic activities are harbored in b-subunits and are classified as cas-
pase-like (PGPH, b1 subunit), trypsin-like (T-L, b2 subunit), and
chymotrypsin-like (CT-L, b5 subunit), respectively. All these active
c
centers are related to N-terminal threonine residue (O -Thr1),
which acts as a nucleophile in peptide bond hydrolysis.5
Bortezomib (Fig. 1), a dipeptidyl boronic acid proteasome inhi-
bitor, covalently interact with the nucleophilic oxygen lone pair of
Figure 1. Structures of bortezomib and drug-like candidates A, B, and C.
c
the residue O -Thr1 of 20S proteasome. It is noteworthy to point
out that bortezomib prefers to target b5 active site rather than
b2 and b1 active sites (b5 > b2>>b1),6 showing reversible inhibition
of CT-L activity. Although bortezomib is now used in clinics for the
treatment of relapsed and refractory multiple myeloma (MM)
patients, it can cause some severe side effects in peripheral nerve,
cardiovascular system and gastrointestinal tract.7–10 Therefore, our
effort has been devoted to design and synthesis of a series of
dipeptidyl boronic acid proteasome inhibitors containing aa- and
ba-amino acid building blocks in the past few years, hoping to
overcome the reported side effects. Three potent drug-like candi-
dates A, B and C were screened (Fig. 1).
In our previous study, compounds constructed from b
(such as candidate C) showed longer half-life and less toxicity than
⇑
Corresponding authors. Tel.: +86 25 85891591.
E-mail addresses: zhyqscu@hotmail.com (Y. Zhu), jc@cpu.edu.cn (C. Jiang).
These two authors contributed equally to this work.
a-peptides
y
http://dx.doi.org/10.1016/j.bmcl.2016.03.007
0960-894X/Ó 2016 Published by Elsevier Ltd.
Please cite this article in press as: Shi, J.; et al. Bioorg. Med. Chem. Lett. (2016), http://dx.doi.org/10.1016/j.bmcl.2016.03.007

2
J. Shi et al. / Bioorg. Med. Chem. Lett. xxx (2016) xxx–xxx
saponification and acidification of methyl esters 2a–2g, acids
3a–3g were obtained in good to excellent yields.
The synthesis of b-amino boronates 10a–10d hydrochlorides
was critical steps in the preparation of target compounds
(Scheme 2). Corresponding boronic acids reacted with (+)-pinane-
diol to form the corresponding borate ester 5a–5c in 92–98%
yields. Then 5a–5c underwent a Matteson homologation–alkyla-
tion reaction14,15 to give the desired intermediates 6a–6c in excel-
Figure 2. General structure of final compounds (n = 1 or 0).
bortezomib because of the employment of b-unnatural amino acid
in the backbone.11,12 To further explore the structure–activity
relationship (SAR) of such kind of compounds, we designed a series
of dipeptidyl boronic acid proteasome inhibitors containing
lent yields. Sodium azide was used to build
a-azide containing
borate esters 7a and 7b in 92% and 91% yields, respectively. With
7a and 7b in hand, b-azide containing borate esters 8a and 8b were
obtained through a second Matteson homologation–alkylation
reaction in 81% and 70%, respectively, which were then reduced
to 9a and 9b in high yields. After reduction of azide, b-amino
boronates 10a and 10b were obtained in low to moderate yields.
However, for phenyl substituted intermediate 10c, an attempt to
prepare it according to the same method as 10a and 10b was not
successful. So we employed another synthetic route (Scheme 2).
Treatment of monochlorosubstituted boronate 6c with lithium
bis(trimethylsilyl) amide gave protected amine 7c, which was
directly used for the next step without further purification. Inser-
tion of a methylene in compound 7c was performed at ꢀ78 °C with
a
b-amino acids (Fig. 2, n = 1). Most of our previous work focused
on the aromatic-substituted analogues for R1 position (Fig. 2);
but in one case, the substitution of pyrazinyl group on bortezomib
by a methyl group maintained inhibitory activity.13 Encouraged by
this, we designed a series of dipeptidyl boronic acid proteasome
inhibitors constructed from aa-amino acids with aliphatic substi-
tutions at R1 position (Fig. 2, n = 0). Herein we report the synthesis
and in vitro inhibitory activity of these new compounds. In order to
understand ligand–protein interaction mode, molecular docking
study was carried out.
As shown in Scheme 1, compounds with aromatic substitutions
at R1 position (2a–2b) were readily synthesized by common pep-
tide synthesis methods using 1-ethyl-3-(3-dimethyllaminopropyl)
carbodiimide hydrochloride (EDCI) as a coupling agent and N,N-
diisopropylethylamine (DIPEA) as a base. Compounds substituted
with aliphatic groups at R1 position (2c–2g) were prepared from
corresponding acyl chloride with triethylamine as a base. After
chloroiodomethane to produce 8c according to
a reported
method.16 Deprotection of the trimethylsilyl group in 8c afforded
amino boronate 10c. And
a-amino boronate 10d was obtained
from commercial source.
Coupling of amino boronates 10a–10d with various acids 3a–3g
in the presence of EDCI, HOBt and DIPEA gave dipeptidyl boronates
1a-3a: R1=Phenyl; 1b-3b: R1=Pyrazin-2-yl; 1c-3c: R1=Methoxy methyl; 1d-3d: R1=Methoxy ethyl; 1e-3e: R1=n-Propyl; 1f-3f:
R1=Neopentyl; 1g-3g: R1=Cyclopentyl.
Scheme 1. Synthesis of N-terminal-protected
ꢀ15 °C to rt, 4 h; (ii) (1) LiOHꢁH2O, MeOH/H2O, rt, 14 h; (2) 1 M HCl, rt.
a-amino acids 3a–3g. Reagents and conditions: (i) 2a–2b: EDCI, HOBt, DIPEA, DCM, ꢀ15 °C to rt, 10 h; 2c–2g: SOCl2, Et3N, DCM,
4a-10a: R2 = Isobutyl; 4b-10b: R2 = Phenethyl; 4c-6c: R2= Phenyl.
Scheme 2. Synthesis of b-amino boronates 10a–10c hydrochlorides. Reagents and conditions: (i) (+)-pinanediol, EA, rt, 7 h; (ii) n-BuLi, anhydrous DCM, anhydrous ZnCl2,
THF, ꢀ78 °C to rt, 2 h; (iii) 7a–7b: NaN3, (Bu)4N+Brꢀ, DCM/H2O, rt, 10 h; 7c: LiN(SiMe3)2, THF, ꢀ78 °C to rt, 20 h; (iv) n-BuLi, anhydrous DCM, anhydrous ZnCl2, THF, ꢀ78 °C to
rt, 2 h; (v) LiBH(C2H5)3, THF, 0 °C to rt, 7 h; (vi) (1) LiAlH4, THF, ꢀ78 °C to rt, 20 h. (2) 4.5 M HCl in Et2O, ꢀ20 °C to rt, 3 h; (vii) n-BuLi, ICH2Cl, THF, ꢀ78 °C to rt, 5 h; (viii) 4.5 M
HCl in Et2O, ꢀ78 °C to rt, 5 h.
Please cite this article in press as: Shi, J.; et al. Bioorg. Med. Chem. Lett. (2016), http://dx.doi.org/10.1016/j.bmcl.2016.03.007

J. Shi et al. / Bioorg. Med. Chem. Lett. xxx (2016) xxx–xxx
3
11a–11k, which were used directly for acid-catalyzed transesteri-
fication with isobutyl boronic acid to provide target compounds
12–22 in good yields (Scheme 3).
The CT-L inhibitory activities of 20S human proteasome of tar-
get compounds were evaluated and bortezomib was used as a con-
trol (results shown in Table 1). To our great surprise, dipeptides
R2 position. Some groups, such as alkoxys (methoxymethyl in com-
pound 18 and methoxyethyl in compound 19), linear alkyl (n-pro-
pyl in compound 20), branched alkyl (neopentyl in compound 21),
naphthenic hydrocarbon (cyclopentyl in compound 22) were
selected to investigate SAR. The data in Table 1 revealed that com-
pounds with alkyl groups at R1 position (18–22) were as active as
bortezomib (all IC50 values less than 10 nM). These results demon-
strated that aliphatic substituents were also active building blocks
for dipeptidyl boronic acid proteasome inhibitors.
constructed from
ab-amino acids (n = 1) exhibited no inhibitory
activities whether substituents at R2 position with aliphatic groups
such as isopropyl (12, 13), phenylethyl (14, 15) or aromatic phenyl
groups (16, 17). So an
a
-amino acid substituent (generally leucine)
Next the effects of compounds 18–22 in inhibiting cells growth
were tested in three multiple myeloma cell lines U266, RPMI8226
and ARH77 (Table 2). The result showed that the compounds
exhibited effective cytotoxicities with IC50 less than 10 nM and
at R2 position was critical for the activities, which was consistent
with previous report.13 So we next investigated the activities of
the aliphatic substituents at R1 position with isopropyl group at
Scheme 3. Synthesis of dipeptidyl boronic acids 12–22. Reagents and conditions: (i) EDCI, HOBt, DIPEA, DCM, ꢀ15 °C to rt, 10 h; (ii) isobutylboronic acid, 1 M HCl, MeOH/
hexane, rt, 6 h.
Table 1
CT-L inhibitory activities of compounds 12–22 and cytotoxicities of compounds 18–22
Compds
n
R1
R2
Enzymatic assays
IC50 (nM)a
Cellular assaysb (IC50, nM)
RPMI8226
U266
ARH77
12
13
14
15
16
17
1
1
1
1
1
1
NAc
NA
NA
NA
NA
NA
18
19
20
21
22
0
0
0
0
0
6.31
8.47
6.94
4.69
4.82
6.04
11.5
8.51
8.01
2.48
4.56
7.32
5.61
4.54
2.29
5.72
9.03
7.87
8.31
3.87
Bortezomibd
0
7.09
5.73
3.88
6.07
a
b
c
Each enzymatic IC50 determination was performed with eight concentrations, and each assay point was determined in duplicate.
Each cellular IC50 determination was performed with ten concentrations, and each assay point was determined in triplicate.
NA, not active.
d
IC50 value obtained for bortezomib under our experimental conditions.
Please cite this article in press as: Shi, J.; et al. Bioorg. Med. Chem. Lett. (2016), http://dx.doi.org/10.1016/j.bmcl.2016.03.007

4
J. Shi et al. / Bioorg. Med. Chem. Lett. xxx (2016) xxx–xxx
Table 2
Compared with bortezomib and compound 22, inhibitor 13 had
the lowest Cdock affinity (5.54) and longest B–O bond distance
c
Results of covalent docking between proteasome and bortezomib, compounds 13 and
22
(3.39 Å), which was not a covalent bond. Furthermore, compound
13 had an additional methylene group between boronic acid phar-
macophore and R2 substituent, which made the peptide backbone
much longer and changed the interaction mode of phenyl and pyr-
azine-2-carboxyl moieties with the residues in the b5 subunit
(Fig. 3A), while the two moieties of compound 22 almost took
the same interaction mode as bortezomib (Fig. 3B). Moreover, as
shown in Figure 3C, five hydrogen bonds were formed between
proteasome and compound 13, four of which were formed
between pharmacophore boronic acid and residues (Thr1, Thr168
and Lys33) and only one hydrogen bond was bridged between pep-
tide backbone moiety and residue Thr21O, which could not fix the
backbone in an effective interaction mode.
c
Compds
Glide Gscore
(absolute value)
Cdock affinity
(absolute value)
B–O bond
distance (Å)
13
22
4.78
8.03
5.54
6.62
7.07
3.39
2.26
2.17
Bortezomib 8.20
correlated well with enzymatic results, suggesting that these were
probably not off-target effects. Due to its excellent enzymatic and
cellular activities, compound 22 was selected for further devalua-
tion for its in vivo efficacy in animal models.
In order to better understand the binding mode of these two
types of compounds with proteasome, molecular docking was per-
formed for compounds 13 and 22 using Covalent Dock module in
Schrodinger.17 The structure of proteasome was downloaded from
the PDB (code: 2F16). In covalent docking calculation, the reaction
residue and reaction type were set as ‘Thr1’ and ‘boronic acid addi-
tion reaction’, respectively. The output parameters were shown in
Table 2.
For compound 22, the boron atom covalently interacted with
c
the nucleophilic oxygen lone pair of the residue O -Thr1 to form
c
a tetrahedral adduct (B–O 2.26 Å). And the two hydroxyl groups
of boronic acid formed hydrogen bonds with residues Arg19,
Thr1 and Lys33 to strengthen the tetrahedral adduct. Figure 3D
showed that the backbone of compound 22 was also stabilized
by other four hydrogen bonds with the conserved residues
Figure 3. Binging models of bortezomib (green), compound 13 (slate) and compound 22 (pink) in the active site of b5 subunit of proteasome. For A and B: Hydrophobic
surface was shown in pink, hydrophilic surface in blue. For C and D: Thr1 was shown in orange, and other key residues were shown in cyan, hydrogen bonds in yellow;
compounds and key residues were shown in stick, protein in cartoon. (A) binding of bortezomib and compound 13; (B) binding of bortezomib and compound 22; (C)
interactions between compound 13 and residues; (D) interactions between compound 22 and residues.
Please cite this article in press as: Shi, J.; et al. Bioorg. Med. Chem. Lett. (2016), http://dx.doi.org/10.1016/j.bmcl.2016.03.007

J. Shi et al. / Bioorg. Med. Chem. Lett. xxx (2016) xxx–xxx
5
(Gly47, Thr21N, Thr21O and Ala49). Compared with bortezomib,
compound 22 adopted cyclopentyl instead of the pyrazine-2-
carboxyl at R1 position, and this variation did maintain the
hydrophobicity, which made compound 22 form strong hydrophobic
interactions with the residues Ala20, Ala22, Val26, and Ala27 of the
deep hydrophobic pocket. All these interactions made compound
22 fit very well in the active pocket, which was consistent with
the results of inhibitory activity. All the theoretical results
displayed that the interactions between proteasome and com-
pound 13 were much weaker than 22, explaining the fact that
compound 13 was less active than bortezomib and compound 22.
In conclusion, a series of dipeptidyl boronic acid derivatives
molecular design and drug discovery from China Pharmaceutical
University for Schrodinger software support.
Supplementary data
Supplementary data associated with this article can be found, in
the online version, at http://dx.doi.org/10.1016/j.bmcl.2016.03.
007.
References and notes
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, H. D.; Huber, R.
Nature 1997, 386, 463.
4. Lowe, J.; Stock, D.; Jap, B.; Zwickl, P.; Baumeister, W.; Huber, R. Science 1995,
268, 533.
constructed from aa- and
evaluated for their proteasome inhibition. The inhibitors contain-
ing b-amino acids were inactive, while those with aa-ones were
ab-amino acids were synthesized and
a
as potent as bortezomib. Molecular docking studies demonstrated
that pharmacophore boronic acid of proteasome inhibitors con-
5. Brannigan, J. A.; Dodson, G.; Duggleby, H. J.; Moody, P. C.; Smith, J. L.; Tomchick,
D. R.; Murzin, A. G. Nature 1995, 378, 416.
6. Berkers, C. R.; Verdoes, M.; Lichtman, E.; Fiebiger, E.; Kessler, B. M.; Anderson,
K. C.; Ploegh, H. L.; Ovaa, H.; Galardy, P. J. Nat. Methods 2005, 2, 357.
7. Orlowski, R. Z.; Zeger, E. L. Expert Opin. Invest. Drugs. 2006, 15, 117.
8. Dorsey, B. D.; Iqbal, M.; Chatterjee, S.; Menta, E.; Bernardini, R.; Bernareggi, A.;
Cassara, P. G.; D’Arasmo, G.; Ferretti, E.; De Munari, S.; Oliva, A.; Pezzoni, G.;
Allievi, C.; Strepponi, I.; Ruggeri, B.; Ator, M. A.; Williams, M.; Mallamo, J. P. J.
Med. Chem. 2008, 51, 1068.
9. Richardson, P. G.; Barlogie, B.; Berenson, J.; Singhal, S.; Jagannath, S.; Irwin, D.;
Rajkumar, S. V.; Srkalovic, G.; Alsina, M.; Alexanian, R.; Siegel, D.; Orlowski, R.
Z.; Kuter, D.; Limentani, S. A.; Lee, S.; Hideshima, T.; Esseltine, D. L.; Kauffman,
M.; Adams, J.; Schenkein, D. P.; Anderson, K. C. N. Engl. J. Med. 2003, 348, 2609.
10. Ludwig, H.; Khayat, D.; Giaccone, G.; Facon, T. Cancer 2005, 104, 1794.
11. Zhu, Y.; Zhu, X.; Wu, G.; Ma, Y.; Li, Y.; Zhao, X.; Yuan, Y.; Yang, J.; Yu, S.; Shao, F.;
Li, R.; Ke, Y.; Lu, A.; Liu, Z.; Zhang, L. J. Med. Chem. 2010, 53, 1990.
12. Zhu, Y.; Wu, G.; Zhu, X.; Ma, Y.; Zhao, X.; Li, Y.; Yuan, Y.; Yang, J.; Yu, S.; Shao, F.;
Lei, M. J. Med. Chem. 2010, 53, 8619.
structed from
a
b-amino acids could not form covalent bond with
c
O -Thr1 and the backbone of b-dipeptidyl boronic acid inhibitors
a
was prolonged and could not form more effective hydrogen bonds
than those derived from aa-amino acids. Among all of the new
compounds, inhibitor 22 with aliphatic substituent at R1 group
showed excellent activities in both proteasome inhibition and
cytotoxicities against three MM tumor cell lines, deserving further
evaluation for its in vivo efficacy in animal models.
Acknowledgments
Part of this work was supported by the following grants: the
National Natural Science Foundation of China (21302097), Natural
Science Fund for Colleges and Universities in Jiangsu Province
(14KJD350002), Technology Support Program of Science and Tech-
nology Department of Jiangsu Province (BE2014720), National and
Local Joint Engineering Research Center of Biomedical Functional
Materials. Authors would like to acknowledge the laboratory of
13. 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.
14. Matteson, D. S.; Ray, R. J. Am. Chem. Soc. 1980, 102, 7590.
15. Matteson, D. S.; Sadhu, K. M. J. Am. Chem. Soc. 1983, 105, 2077.
16. Gorovoy, A. S.; Gozhina, O.; Svendsen, J. S. J. Pept. Sci. 2013, 19, 613.
17. Zhu, K.; Borrelli, K. W.; Greenwood, J. R. J. Chem. Inf. Model. 2014, 54, 1932.
Please cite this article in press as: Shi, J.; et al. Bioorg. Med. Chem. Lett. (2016), http://dx.doi.org/10.1016/j.bmcl.2016.03.007