Total synthesis of syringolin A and improvement of its biological activity

Article abstract of DOI:10.1002/anie.201402428

The development process for syringolin A analogues having improved proteasome inhibitory and antitumor activity is described. The strategy was to first establish a convergent synthesis of syringolin A using a rare intramolecular Ugi three-component reaction in the last stage of the synthesis, so as to gain access toa set of structure-based analogues. The inhibitory activity of chymotrypsin-like activity of 20S proteasome was largely improved by targeting the S3 subsite of the β5 subunit. Cytotoxic activity was also improved by installing the membrane-permeable substituent. These biological properties are comparable to those of bortezomib, a clinically used first-line proteasome inhibitor.

Full text of DOI:10.1002/anie.201402428

Angewandte
Chemie
DOI: 10.1002/anie.201402428
Drug Discovery
Total Synthesis of Syringolin A and Improvement of Its Biological
Activity
Takuya Chiba, Hidetaka Hosono, Koji Nakagawa, Masahiro Asaka, Hiroshi Takeda,
Akira Matsuda, and Satoshi Ichikawa*
Abstract: The development process for syringolin A analogues
having improved proteasome inhibitory and antitumor activity
is described. The strategy was to first establish a convergent
synthesis of syringolin A using a rare intramolecular Ugi three-
component reaction in the last stage of the synthesis, so as to
gain access toa set of structure-based analogues. The inhibitory
activity of chymotrypsin-like activity of 20S proteasome was
largely improved by targeting the S3 subsite of the b5 subunit.
Cytotoxic activity was also improved by installing the mem-
brane-permeable substituent. These biological properties are
comparable to those of bortezomib, a clinically used first-line
proteasome inhibitor.
Figure 1. Structures of syrbactins.
A
round 50% of new drug entries between 1981 and 2010
were natural-product-based compounds,[1] and it is no doubt
that natural products are a rich source for drug development.
Improvement of biological activity is a key issue associated
with natural-product-based drug development because it is
difficult to use the natural product itself without modifying its
structure, as found in many cases, and much effort has to be
devoted to adding desired properties to make them into drug
candidates. Therefore it is necessary to develop a synthetic
strategy that allows an efficient route and also provides
a range of analogues to investigate structure–activity relation-
ships so as to improve potency. Syringolins[2] and glidobac-
tins[3] constitute members of the syrbactins, which are 12-
membered macrolactams containing an E-configured a,b-
unsaturated carboxamide moiety with an appended side chain
(Figure 1).[4] In 2008, they were revealed to be irreversible
inhibitors of the 20S proteasome, which is a multicatalytic
proteolysis machinery.[5] Proteasome inhibitors are promising
candidates as anticancer agents,[6] and syrbactins are expected
to be a next-generation proteasome inhibitor in cancer
chemotherapy[7] because of its mode of inhibition is different
from those of the currently used proteasome inhibitors
bortezomib (Velcade) and carfilzomib (Kyprolis), which
have several associated problems.[8] Syringolin A (1), how-
ever, exhibits a moderate proteasome inhibitory activity and
weak cytotoxic activity against cancer cells in vitro. As in an
effort to develop more potent syringolin A analogues by
Kaiser et al., who synthesized hybrid-type analogues of 1 and
2,[9] it is necessary to improve the biological activity in order
to set an initial stage of anticancer drug development. Herein
we describe the process to develop syringolin A analogues
with improved biological activity. Our strategy was to first
establish a synthetic strategy applicable to accessing a set of
structure-based analogues. Upon improvement of the protea-
[*] T. Chiba, K. Nakagawa, H. Takeda, A. Matsuda, S. Ichikawa
Faculty of Pharmaceutical Sciences, Hokkaido University
Kita-12, Nishi-6, Kita-ku, Sapporo 060-0812 (Japan)
E-mail: ichikawa@pharm.hokudai.ac.jp
H. Hosono, M. Asaka
Graduate School of Medicine, Department of Cancer Preventive
Medicine, Hokkaido University
Kita-12, Nishi-7, Kita-ku, Sapporo, 060-0812 (Japan)
S. Ichikawa
Center for Research and Education on Drug Discovery
Hokkaido University
Kita-12, Nishi-6, Kita-ku, Sapporo 060-0812 (Japan)
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/anie.201402428. This research was
supported by JSPS Grant-in-Aid for Challenging Exploratory Research
(SI, grant number 22659020), Scientific Research on Innovative
Areas “Chemical Biology of Natural Products” (SI, grant number
24102502), and Scientific Research (B) (SI, grant number 25293026).
We thank S. Oka and A. Tokumitsu (Center for Instrumental Analysis,
Hokkaido University) for measurement of the mass spectra.
Figure 2. Retrosynthetic analysis of syringolin A.
Angew. Chem. Int. Ed. 2014, 53, 1 – 5
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
1
These are not the final page numbers!

.
Angewandte
Communications
Table 1: Proteasome inhibitory[a] and cytotoxic activity[b] of 1 and its
analogues.
Ki [nm]
isolated cell-based
IC50 [nm]
HCT116 A431
RPMI8226
1
16e
19a
19b
19c
19d
19e
bortezomib
680
44
0.14
1.1
0.6
0.2
0.4
–
–
–
2830
2010
3240
3370
1160
947
2.2
4.7
4.8
5.7
6.2
2.0
5.5
6.0
8.5
11
12
2.1
5.7
4.3
16.5
18.2
9.1
8.1
6.0
20.9
21.1
12.6
10.1
5.3
[a] Chymotrypsin-like activity of 20S proteasome. [b] HCT116: human
colon cancer cells, A431: human epidermal cancer cells, RPMI8226:
human myeloma cells.
the corresponding diastereomer at the dehydrolysine residue
was not isolated in the intra-U3CR, dehydration, and
deprotection steps. It was, therefore, presumed that the
intra-U3CR proceeded stereoselectively.
Synthetic 1 moderately inhibited chymotrypsin-like (CT-
L; b5 subunit) activity of a purified 20S proteasome with an
apparent Ki value of 680 nm as reported. Weak cytotoxic
Scheme 1. Total synthesis of syringolin A. Reagents and conditions.
a) LiOH, H2O/THF/MeOH (60:75:75), 08C to RT, 15 h; b) 8, EDCI,
iPr2NEt, DMAP, CH2Cl2, 08C to RT, 19 h, 89% over 2 steps; c) 80% aq.
TFA, 08C, 2 h; d) HCO2H, iPr2NEt, EDCI, CH2Cl2, 08C, 72 h, 50% over
2 steps; e) TBSCl, imidazole, DMF, 08C to RT, 30 h, f) K2CO3, MeOH,
RT, 1 h, 72% over 2 steps; g) IBX, MeCN, 808C, 1 h; f) triphosgene,
Et3N, CH2Cl2, À788C, 30 min, 86% over 2 steps; i) 2,4-dimethoxyben-
zylamine, 5, CH2Cl2, reflux, 36 h; j) 3HF·Et3N, MeCN, RT, 24 h, 32%
over 2 steps; k) MsCl, Et3N, CH2Cl2, 08C, 10 min; l) DBN, CH2Cl2, RT,
48 h, 64% over 2 steps; m) 80% aq. TFA, Et3SiH, RT, 4 h, 75%.
DBN=1,5-diazabicyclo[4.3.0]non-5-ene, DMAP=N,N-dimethylamino-
pyridine, DMF=N,N-dimethylformamide, EDCI=1-ethyl-3-(3-dimethyl-
aminopropyl)carbodiimide, IBX=2-iodoxybenzoic acid, Ms=methane-
sulfonyl, TBS=tert-butyldimethylsilyl, TFA=trifluoroacetic acid,
THF=tetrahydrofuran.
some inhibitory activity, further modification was pursued to
improve the cytotoxic activity. We planned to assemble 4, 5,
and an ammonia equivalent by a rare intramolecular Ugi
three-component reaction[10] (intra-U3CR) in the last stage of
the synthesis (Figure 2). This reaction constructs the highly
strained 12-membered macrolactam and a nonproteinogenic
amino acid, b,g-dehydrolysine, within the maclolactam, and
links the ureadipeptide side chain 5 simultaneously. This
strategy was completely different from those previously
reported.[11]
The total synthesis of 1 is summarized in Scheme 1. The
known methyl ester 7,[12] prepared from l-valine, was hydro-
lyzed and the resulting carboxylic acid was condensed with
the amine 8 to give the carboxamide 9. Protecting group
manipulations of 9 by way of 10 provided the formamide 11.
Oxidation of the primary alcohol of 11 followed by dehy-
dration of the formamide moiety by triphosgene cleanly
afforded the isocyanoaldehyde 4, a precursor for the intra-
U3CR, in 86% yield over two steps. A mixture of 4, the
ureadipeptide carboxylic acid 5, and 2,4-dimethoxybenzyl-
amine was heated under reflux in CH2Cl2 for 36 hours and
gave the desired 12 in 32% over two steps after removal of the
TBS group.[13] Sequential dehydration of 12 by mesylation and
elimination with DBN (64% over 2 steps) gave a a,b-
unsaturated carboxamide, which was deprotected further by
TFA and Et3SiH (75% yield) to complete the total synthesis
of 1. Synthetic 1 was obtained as a single diastereomer, and
Scheme 2. Synthesis of syringolin A analogues. Reagents and condi-
tions. a) piperidine, DMF, RT, 15 min; b) 14, iPr2NEt, CH2Cl2, RT, 1 h;
c) 5% TFA/CH2Cl2, 10 min; d) 4, 2,4-dimethoxybenzylamine, CH2Cl2,
reflux, 41 h; e) 3HF·Et3N, MeCN, RT, 40 h, 17–29% over 2 steps;
f) MsCl, Et3N, CH2Cl2, 08C, 10 min; g) DBU, CH2Cl2, RT, 17 h, 18–33%
over 2 steps; h) 80% aq. TFA, Et3SiH, RT, 5 h, 40–86%; i) N-Alloc
phenylalanine, 2,4-dimethoxybenzylamine, CH2Cl2, reflux; j) 3HF·Et3N,
MeCN, RT, 39 h, 57% over 2 steps; k) MsCl, Et3N, CH2Cl2, 08C,
10 min; l) DBN, CH2Cl2, RT, 6 h, 20% over 2 steps; m) [Pd(PPh3)4],
morpholine, THF, RT, 30 min, 81%; n) carboxylic acid, EDCI, CH2Cl2,
08C to RT, 12 h; o) 80% aq. TFA, Et3SiH, RT, 30 min, 45–72% over 2
steps.
2
www.angewandte.org
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2014, 53, 1 – 5
These are not the final page numbers!

Angewandte
Chemie
Figure 3. Biological properties of syringolin A analogues. a) X-ray crystal structure of b5-subunit of 20S proteasome bound to 1 (PDB code:
2ZCY). b) Comparison of CT-L inhibition of 1 and 16a–e at 250 nm. c) Cell-based CT-L inhibitory activity of bortezomib, 1, and 19a–e with human
embryonic kidney 293 (HEK293) cells. d) Profiling of HEK293 cell lysates with 19 f. e) Accumulation of poly-ubiquitinated IkBa in HEK293 cells
treated with bortezomib, 1, 19a, and 19e at 100 nm.
activity against a range of human cancer cells (IC50 = 1160–
3240 nm) was observed for synthetic 1 (Table 1).
size of the hydrophobic substituents resulted in increased CT-
L inhibitory activity. The benzyl-substituted analogue 16e
(Ki = 44 nm, Table 1) exhibited a 15-fold stronger CT-L
inhibitory activity than 1. The analogue 16e, however,
showed weak cytotoxic activity (IC50 = 947–3370 nm,
Table 1) and this is because of the lack of cell membrane
permeability of 16e. It was reported that hybrid-type
analogues of 1 and 2 exhibited an increased proteasome
inhibition and cytotoxic activity against human cancer
cells.[9,14] These results prompted us, likewise, to install
a cell-permeable side chain on the core structure of 16e.
The analogues 19a–e, which were prepared as shown in
Scheme 2b, exhibited a strong isolated CT-L inhibitory
activity with apparent Ki values in a range of 0.14–1.1 nm
(Table 1). We also evaluated the ability of 19a–e to inhibit
the proteasome in living cells using a cell-based proteasome
activity with human embryonic kidney 293 (HEK293) cells.
The analogues 19a–e showed a strong CT-L inhibitory activity
(Figure 3c) as well as trypsin-like (T-L, b2 subunit) activity
(see the Supporting Information). A strong cytotoxic activity
was also observed and the analogue 19a, possessing the
decanoyl group, and it was the most potent with an IC50 value
The a,b-unsaturated carboxamide moiety of 1 serves as
a Michael acceptor to covalently bind to the secondary
hydroxy group of the N-terminal threonine (Thr) residue of
the b5 subunit of 20S proteasome. The X-ray crystal structure
of the yeast 20S proteasome covalently bound to 1 (PDB
code: 2ZCY)[5a] revealed that the isopropyl side chain of the
internal valine (Val) residue was recognized by an S3 subsite,
which constitutes a hydrophobic pocket (Figure 3a). More-
over, the hydrophobic pocket could accept a larger hydro-
phobic substituent than the isopropyl group. To improve the
proteasome inhibitory activity by increasing the hydrophobic
interaction, we designed the analogues 16a–e, wherein the
internal isopropyl group was replaced with a range of alkyl
substituents, including a benzyl group (Scheme 2a). The
ureadipeptide carboxylic acids 15a–e were prepared using
solid-phase synthesis to deliver a variety of analogues
efficiently in a short time, and the analogues 16a–e were
synthesized in a manner similar to the synthesis of 1. The CT-
L inhibitory activity of 16a–e at 250 nm was well correlated to
the size of alkyl substituents (Figure 3b), and increasing the
Angew. Chem. Int. Ed. 2014, 53, 1 – 5
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
3
These are not the final page numbers!

.
Angewandte
Communications
[3] a) M. Oka, K. Nishikawa, S. Ohta, H. Kamei, M. Konishi, T.
Miyaki, T. Oki, H. Kawaguchi, J. Antibiot. 1988, 41, 1331 – 1337;
b) M. Oka, K. Yaginuma, K. Numata, M. Konishi, T. Oki, H.
Kawaguchi, J. Antibiot. 1988, 41, 1906 – 1909.
[4] D. Krahn, C. Ottmann, M. Kaiser, Nat. Prod. Rep. 2011, 28,
1854 – 1867.
of 2.2 nm for human myeloma RPMI8226 cells (Table 1). To
confirm its target selectivity, the cell lysate, treated with the
activity-based probe[15] 19 f (see the Supporting Information
for its preparation), was analyzed by gel electrophoresis
(Figure 3d). As a result, only two bands, corresponding to the
b2 and b5 subunits, were fluorescently labeled with 19 f.
Labeling of these bands was inhibited by bortezomib in
a dose-dependent manner. This result clearly indicates that
the syringolin A analogues 19 exhibit a high specificity to the
20S proteasome. The transcription factor NFkB, responsible
for immune and inflammatory responses, is primarily regu-
lated by interaction with inhibitory IkB proteins, which is
further regulated by phosphorylation followed by proteasome
degradation through the poly-ubiquitination. Treatment of
HEK293 cells with 19a,e led to the accumulation of
phosphorylated and poly-ubiquitinated IkBa as a result of
proteasome inhibition, and was similar in behavior to that of
bortezomib (Figure 3e). These results confirm that 19a,e are
promising cell-permeable and selective proteasome inhibitors
as anticancer agents.
In conclusion, the total synthesis of syringolin A was
established by a rare intra-U3CR. The 20S proteasome
inhibitory activity of 1 was improved and guided by the
structure-based drug design targeting the S3 subsite of the
b5 subunit. Cytotoxic activity was also greatly improved by
installing the membrane-permeable substituent. Further
studies to develop a next-generation proteasome inhibitor
as an anticancer drug, including evaluation of in vivo anti-
cancer activity, administration, distribution, metabolism, and
toxicity are currently ongoing.
[5] a) M. Groll, B. Schellenberg, A. S. Bachmann, Nature 2008, 452,
755 – 759; b) A. K. Nussbaum, T. P. Dick, W. Keilholz, M.
´
Schirle, S. Stevanovic, K. Dietz, W. Heinemeyer, M. Groll,
D. H. Wolf, R. Huber, H.-G. Rammensee, H. Schild, Proc. Natl.
Acad. Sci. USA 1998, 95, 12504 – 12509.
[6] a) J. Adams, Cancer Cell 2004, 5, 412 – 417; b) P. Moreau, P. G.
Richardson, M. Cavo, R. Z. Orlowski, J. F. San Miguel, A.
Palumbo, J.-L. Harousseau, Blood 2012, 120, 947 – 959.
[7] See Ref. [4].
[8] C. Naujokat, D. Fuchs, C. Berges, Biochim. Biophys. Acta Mol.
Cell Res. 2007, 1773, 1389 – 1397.
[9] a) J. Clerc, N. Li, D. Krahn, M. Groll, A. S. Bachmann, B. I.
Florea, H. S. Overkleeft, M. Kaiser, Chem. Commun. 2011, 47,
385 – 387; b) C. R. Archer, M. Groll, M. L. Stein, B. Schellen-
berg, J. Clerc, M. Kaiser, T. P. Kondratyuk, J. M. Pezzuto, R.
Dudler, A. S. Bachmann, Biochemistry 2012, 51, 6880 – 6888.
[10] Only one study has been reported in the literature. A. I. Khan,
A. K. Saxene, Tetrahedron 2012, 68, 294 – 299.
[11] For total synthesis, see: a) J. Clerc, M. Groll, D. J. Illich, A. S.
Bachmann, R. Huber, B. Schellenberg, R. Dudler, M. Kaiser,
Proc. Natl. Acad. Sci. USA 2009, 106, 6507 – 6512; b) J. Clerc, B.
Schellenberg, M. Groll, A. S. Bachmann, R. Huber, R. Dudler,
M. Kaiser, Eur. J. Org. Chem. 2010, 3991 – 4003; c) C. Dai, C. R.
Stephenson, Org. Lett. 2010, 12, 3453 – 3455; d) M. C. Pirrung, G.
Biswas, T. R. Ibarra-Rivera, Org. Lett. 2010, 12, 2402 – 2405.
[12] K. M. Lassen, J. Lee, M. M. Joullie, Tetrahedron Lett. 2010, 51,
1635 – 1638.
[13] Initial attempt of the intra-U3CR with the a,b-unsaturated
carboxamide A or the diastereomer at the silyloxy group (B)
failed to give the corresponding macrocycle. Details will be
described elsewhere.
Received: February 14, 2014
Published online: && &&, &&&&
Keywords: antitumor agents · drug discovery · lactams ·
.
medicinal chemistry · natural products
[14] C. R. Archer, D. L. Koomoa, E. M. Mitsunaga, J. Clerc, M.
Shimizu, M. Kaiser, B. Schellenberg, R. Dudler, A. S. Bach-
mann, Biochem. Pharmacol. 2010, 80, 170 – 178.
[15] J. Clerc, B. I. Florea, M. Kraus, M. Groll, R. Huber, A. S.
Bachmann, R. Dudler, C. Driessen, H. S. Overkleeft, M. Kaiser,
ChemBioChem 2009, 10, 2638 – 2643.
[1] a) D. J. Newman, G. M. Cragg, J. Nat. Prod. 2012, 75, 311 – 335;
b) D. J. Newman, G. M. Cragg, J. Nat. Prod. 2007, 70, 461 – 477;
c) R. OꢀShea, H. E. Moser, J. Med. Chem. 2008, 51, 2871 – 2878.
[2] a) U. Wꢁspi, D. Blanc, T. Winkler, P. Ruedi, R. Dudler, Mol.
Plant-Microbe Interact. 1998, 11, 727 – 733; b) U. Wꢁspi, P.
Hassa, A. A. Staempfli, L. P. Molleyres, T. Winkler, R. Dudler,
Microbiol. Res. 1999, 154, 89 – 93.
4
www.angewandte.org
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2014, 53, 1 – 5
These are not the final page numbers!

Angewandte
Chemie
Communications
Drug Discovery
T. Chiba, H. Hosono, K. Nakagawa,
M. Asaka, H. Takeda, A. Matsuda,
S. Ichikawa*
&&&& — &&&&
Total Synthesis of Syringolin A and
Improvement of Its Biological Activity
Fine-tuned activity: The total synthesis of
syringolin A (1) was established by an
intramolecular Ugi three-component
reaction. To improve the moderate bio-
logical activities of 1, analogues having
strong proteasome inhibitory and cyto-
toxic activities comparable to clinically
used bortezomib, were prepared.
Angew. Chem. Int. Ed. 2014, 53, 1 – 5
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
5
These are not the final page numbers!