Design, synthesis, and evaluation of novel kazusamycin A derivatives as potent antitumor agents

Article abstract of DOI:10.1016/j.bmcl.2006.03.056

Novel kazusamycin A derivatives were designed in the viewpoint of decrease of reactivity at the α,β-unsaturated δ-lactone moiety against Michael-type addition. Although 25-30 steps were required for the synthesis of each compound, their syntheses were ach

Full text of DOI:10.1016/j.bmcl.2006.03.056

Bioorganic & Medicinal Chemistry Letters 16 (2006) 3315–3318
Design, synthesis, and evaluation of novel kazusamycin A
derivatives as potent antitumor agents
Ryoichi Ando,a,* Yusaku Amano,b Hideo Nakamura,b
Noriyoshi Araic and Isao Kuwajimac,d
aDiscovery Technology Laboratory II, Mitsubishi Pharma Corporation, 1000 Kamoshida-cho, Aoba-ku, Yokohama 227-0033, Japan
bResearch Laboratory IV, Mitsubishi Pharma Corporation, 1000 Kamoshida-cho, Aoba-ku, Yokohama 227-0033, Japan
cCore Research for Evolutional Science and Technology (CREST), Japan Science and Technology Agency (JST),
1-15-1 Kitasato, Sagamihara 228-8555, Japan
dThe Kitasato Institute, 1-15-1 Kitasato, Sagamihara 228-8555, Japan
Received 6 January 2006; revised 27 February 2006; accepted 9 March 2006
Available online 17 April 2006
Abstract—Novel kazusamycin A derivatives were designed in the viewpoint of decrease of reactivity at the a,b-unsaturated d-lactone
moiety against Michael-type addition. Although 25–30 steps were required for the synthesis of each compound, their syntheses were
achieved. Cytotoxicity against HPAC cell line was evaluated, and two of them exhibited comparable potency to kazusamycin A.
Hepatic toxicity of these designed compounds was much lower than that of kazusamycin A.
ꢀ 2006 Elsevier Ltd. All rights reserved.
Kazusamycin A was first isolated from the culture broth
of Streptomyces sp. No. 81-484 in 1984,1 and its absolute
structure was determined by total synthesis in 2004.2
Kazusamycin A exhibited potent antitumor activity
against many kinds of leukemia and cell line of solid tu-
mors both in vitro and in vivo,3,4 but it has not been
used in clinical setting by now5 because of its hepatic
and gastrointestinal toxicity.6,7
group of Cys-529 in the target protein CRM1/exportin 1
causes irreversible Michael-type addition to this a,b-
unsaturated d-lactone moiety.8
OH
O
O
callystatin A
OH
O
O
HOOC
X
A study on structure–activity relationship of callystatin
A derivatives, which have an a,b-unsaturated d-lactone
ring and a lipophilic side chain, also gave us useful infor-
mation.9 Various transformation of the lipophilic side
chain was investigated, and all compounds synthesized
decreased antitumor activity. This result indicates that
the lipophilic side chains of natural products are one
of optimized structures. Accordingly, irreversible
Michael-type addition of CRM1/exportin 1 is the main
reason for extremely potent antitumor activity along
with high affinity of lipophilic side chain to the target
protein.
O
X = OH : kazusamycin A
X = H : leptomycin B
O
Kazusamycin A has a unique a,b-unsaturated d-lactone
ring. The mechanism of action of leptomycin B whose
structure was very similar to that of kazusamycin A
was studied, and it has become clear that the sulfhydryl
However, this reactive lactone moiety would also lead
to high toxicity. It may react with side chains of
cysteine, serine, lysine, or arginine moiety of proteins,
Keywords: Kazusamycin; Antitumor activity.
*
Corresponding author. Tel.: +81 45 963 3440; fax: +81 45 963
3977; e-mail: Ando.Ryoichi@mk.m-pharma.co.jp
0960-894X/$ - see front matter ꢀ 2006 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2006.03.056

3316
R. Ando et al. / Bioorg. Med. Chem. Lett. 16 (2006) 3315–3318
OH
O
or nucleophilic components of a living body. Then, this
reaction would cause unfavorable effects.
HOOC
OH
In order to overcome this drawback of kazusamycin A,
we have designed more stable a,b-unsaturated d-lactone
ring systems. They will decrease the reactivity against
nucleophilic addition of CRM1/exportin 1, but its toxic-
ity will also decrease because of its less reactivity. The
decrease of antitumor activity and toxicity might be par-
allel in general, but we have expected that the safety
margin would become wider since the lipophilic side
chain showed high affinity to the target protein.9
R1
A
R2
O
R3
O
O
O
O
TBSO
AllocO
H
+
R1
R2
O
B
R3
O
Br
There are two strategies to decrease the reactivity of a,b-
unsaturated d-lactone moiety to nucleophilic addition.
One is to increase steric hindrance around the b-position
at which nucleophiles attack, and the other is to deacti-
vate the carbon–carbon double bond by adding elec-
tron-donating substituent at the a- or b-position.
C
CHO
O
TBDPSO
R1
+
R2
We designed four compounds to confirm this hypothesis
(Fig. 1). Compound 1a is an example of more hindered
compound by adding a methyl group at the c-position,
and 1b is an example of deactivation of the carbon–car-
bon double bond by adding an electron-donating methyl
group at the b-position. This methyl group would also
affect steric hindrance around the b-position. Com-
pound 1c is an example of sterically less hindered but
electrically deactivated compound by eliminating the
c-methyl group and adding a methyl group at the b-po-
sition of kazusamycin A. Compound 1d is the least reac-
tive compound in these four because of both steric and
electrical factors arising from the addition of two methyl
groups at the b- and c-position of kazusamycin A.
D
Br
R3
Oi-Pr
2a: R1 = R2 = CH3, R3 = H
2b: R1 = R3 = CH3, R2 = H
2c: R1 = R2 = H, R3 = CH3
2d: R1 = R2 = R3 = CH3
Scheme 1. Retrosynthetic analysis of the designed compounds.
fragment B via Wittig reaction. Then, compound A
would be synthesized in further eight-step reactions.10,11
As shown in Scheme 2, syntheses of 2a–2c were analo-
gous. Aldehydes 3a–3c12 were converted to the dichloro-
ethylene derivatives 4a–4c by Wittig reaction using
CCl3PO(OEt)2. Treatment of 4a–4c with BuLi afforded
the lithiated acetylenes, which were allowed to react with
methyl chloroformate to give the acetylenes 5a–5c.
Hydrogenation of 5a by Lindlar catalyst poisoned by
quinoline in toluene afforded cis-olefin 6a in 87% yield.
Trisubstituted olefins 6b and 6c were synthesized by
addition of Me2CuLi in THF at ꢀ78 ꢁC (92% and
82% yield, respectively). Compounds 6a–6c were con-
verted to 2a–2c in six steps each without any difficulty.
Syntheses of 1a, 1b, and 1c were achieved by similar
reactions described in the literature.2
Syntheses of the designed compounds should be
achieved by using a similar route of total synthesis by
Kuwajima.2 Key intermediates of this project are com-
pounds 2a–2d, which would be transformed to the de-
signed compounds easily by using similar reactions of
total synthesis as shown in Scheme 1.
Coupling of aldehydes 2a–2d with fragment D via Wittig
reaction followed by two-step reactions would afford
fragment C, which would be allowed to react with
For synthesis of compound 2d we developed a new syn-
thetic route which is shown in Scheme 3. The starting
alcohol 714 was oxidized by sulfur trioxide–pyridine
complex at rt in 97% yield, and methylmagnesium bro-
mide was added to the resulting aldehyde at 0 ꢁC. Then,
the secondary alcohol formed was oxidized by sulfur tri-
oxide–pyridine complex at rt to give the methylketone 8.
Treatment of 8 with 6 N HCl in THF furnished hydro-
lysis of the acetal moiety followed by cyclization to
afford the hemiacetal 9. The primary alcohol moiety
which existed as a ring-opening form of 9 was protected
by the reaction with TBDPSCl in the presence of
DMAP. The total yield of these successive four steps
was 26%. Then, the secondary alcohol moiety of 10
was allowed to react with bromoacetyl bromide in the
presence of triethylamine to afford compound 11 in
98% yield. Treatment of 11 with samarium iodide gave
OH
O
HOOC
OH
O
kazusamycin A
O
O
O
O
O
O
O
O
O
1a
1d
1c
1b
Figure 1. Design of kazusamycin A derivatives.

R. Ando et al. / Bioorg. Med. Chem. Lett. 16 (2006) 3315–3318
Table 1. Cytotoxicity against HPAC cell line
3317
R
R
R
R
Compound
R1
R2
R3
a
IC50 (nM)
O
O
a
b
O
O
Cl
CHO
R2
Kazusamycin A
CH3
CH3
CH3
H
H
H
0.0747
0.0382
179
R2
R1
R1
Cl
1a
1b
1c
1d
CH3
H
H
3a: R1 = R2 = CH3, R = CH2CH3
3b: R1 = CH3, R2 = H, R = CH3
3c: R1 = R2 = H, R = CH2CH3
CH3
CH3
CH3
4a-4c
H
CH3
0.309
16100
CH3
a IC50 values were calculated from the concentration–response curves,
as the concentration of the test compounds elicited a decrease in the
measured absorbance, equivalent to 50% of the vehicle group.
O
c
e-j
O
2a
COOMe
the hydroxylactone 12, which was dehydrated via sulf-
inylation by thionyl chloride. Compound 12 could be
converted to 2d in five steps by referring to the reported
method.2
6a
R
R
COOMe
O
O
R2
R1
5a-5c
Cytotoxicity of kazusamycin A15 and four derivatives16
against human HPAC cell line (ATCC CRL-2119) was
tested,17 and IC50 values of them are summarized in
Table 1.
R
R
O
d
e-j
O
2b. 2c
R1
R2
COOMe
6b: R1 = CH3, R2 = H, R = CH3
6c: R1 = R2 = H, R = CH2CH3
Compound 1a exhibited similar cytotoxicity to kazusa-
mycin A, and 1c was slightly less potent than kazusamy-
cin A. IC50 value of compound1b was much greater than
that of kazusamycin A, and 1d showed least potency.
Scheme 2. Syntheses of the intermediates 2a, 2b, and 2c. Reagents and
conditions: (a) BuLi, CCl3PO(OEt)2, THF–Et2O, ꢀ100 ꢁC to ꢀ50 ꢁC;
(b) BuLi, THF, ꢀ78 ꢁC to 0 ꢁC, then ClCOOMe, ꢀ78 ꢁC to 0 ꢁC; (c)
Lindlar catalyst, H2, quinoline, toluene, rt, 87%; (d) Me2CuLi, THF,
ꢀ78 ꢁC, 92% (6b), 82% (6c); (e) Dowex 50 W X8, MeOH, rt, then
Amberlyst 15, CH2Cl2, rt; (f) TBDPSCl, imidazole, DMF, 0 ꢁC to rt;
(g) DIBAL, CH2Cl2, ꢀ78 ꢁC; (h) i-PrOH, PPTS, benzene, rt; (i) TBAF,
THF, rt; (j) (COCl)2, DMSO, CH2Cl2, ꢀ78 ꢁC, then Et3N, ꢀ78 ꢁC to
0 ꢁC.
This result is well explained by steric and electric effects.
The high potency of 1a indicates that the upper face of
the lactone ring would not be important for both bind-
ing to CRM1/exportin 1 and addition of sulfhydryl
group of Cys-529 in the target protein to the a,b-unsat-
urated d-lactone moiety. Comparable potency of 1c to
kazusamycin A could be explained by balance of steric
and electric factors. Namely, addition of a methyl group
at the b-position gave unfavorable effect because of both
deactivation of the carbon–carbon double bond and ste-
ric hindrance around the b-position, but removal of the
methyl group at the c-position gave quite favorable
effect. Indeed, addition of a methyl group to the b-posi-
tion of kazusamycin A caused marked decrease of
potency (1b). The least potency of compound 1d is
reasonable since it is most hindered and deactivated
compound of the four.
O
a-c
O
O
d
O
O
OH
7
8
HO
OH
O
f
e
TBDPSO
10
O
OH
O
9
O
Finally, toxicity was evaluated to confirm our hypothe-
sis. Compounds 1a, 1c, and kazusamycin A were intra-
peritoneally administered once daily to female BALB/c
mice at dose levels of 0.125, 0.25, and 0.5 mg/kg for four
days. Saline was injected to vehicle control mice.
Br
TBDPSO
11
O
O
O
g
OH
TBDPSO
12
O
h
O
i-l
As regards kazusamycin A, all mice given 0.5 mg/kg
died, and higher value of ALT (GPT) was observed at
0.125 and 0.25 mg/kg in serum biochemistry. As regards
compounds 1a and 1c, all mice survived, and value of
ALT was a similar level to that of the control mice even
at 0.5 mg/kg.18,19
2d
TBDPSO
13
Scheme 3. Synthesis of the intermediate 2d. Reagents and condi-
tions: (a) SO3Æpy, DMSO, rt, 97%; (b) MeMgBr, THF, 0 ꢁC to rt;
(c) SO3Æpy, DMSO, rt; (d) 6 N HCl, THF, 0 ꢁC to rt; (e)
TBDPSCl, Et3N, DMAP, CH2Cl2, 0 ꢁC to rt, 26% (four steps);
(f) bromoacetyl bromide, pyridine, CH2Cl2, 0 ꢁC, 98%; (g) SmI2,
THF, ꢀ78 ꢁC; (h) SOCl2, pyridine, 0 ꢁC, 86% (two steps); (i)
DIBAL, CH2Cl2, ꢀ78 ꢁC; (j) i-PrOH, PPTS, benzene, rt; (k)
TBAF, THF, rt; (l) (COCl)2, DMSO, CH2Cl2, ꢀ78 ꢁC, then Et3N,
ꢀ78 ꢁC to 0 ꢁC.
Kazusamycin A exhibited fatal toxicity at higher dose
and hepatic toxicity even at 0.125 mg/kg, whereas com-
pounds 1a and 1c did not exhibit hepatic toxicity even
at 0.5 mg/kg. This result indicates that our designed
compounds with comparable potency to kazusamycin

3318
R. Ando et al. / Bioorg. Med. Chem. Lett. 16 (2006) 3315–3318
11. Practical synthesis of fragments B and D has been
investigated: Zhou, S.; Chen, H.; Liao, W.; Chen, S.-H.;
Li, G.; Ando, R.; Kuwajima, I. Tetrahedron Lett. 2005, 46,
6341.
12. Synthesis of 3a is described in Scheme 3. Compound 3b is
an intermediate for the total synthesis of kazusamycin A.2
Dimethylacetal derivative of 3c was described in a
literature13.
A are much less toxic than kazusamycin A. So it would
be possible to say that our hypothesis has been proved.
In summary, novel kazusamycin A derivatives were de-
signed in the viewpoint of decrease of reactivity at the
a,b-unsaturated d-lactone moiety against Michael-type
addition. Although 25–30 steps were required for the
synthesis of each compound, their syntheses were
achieved. Cytotoxicity against HPAC cell line was eval-
uated, and 1a and 1c exhibited comparable potency to
kazusamycin A. Hepatic toxicity of these designed com-
pounds was much lower than that of kazusamycin A.
13. Labelle, M.; Morton, H. E.; Guindon, Y.; Springer, J. P.
J. Am. Chem. Soc. 1988, 110, 4533.
´
14. Lavallee, P.; Ruel, R.; Grenier, L.; Bissonnette, M.
Tetrahedron Lett. 1986, 27, 679.
15. This sample was synthesized by the method in Ref. 2.
16. These new four compounds were fully characterized. For
27
example, 1a: ½aꢁD ꢀ135.6 (c 0.106, MeOH); IR (KBr, neat,
cmꢀ1) 3421, 2965, 1707, 1643, 1456, 1375, 1251, 1121, 967,
1
Acknowledgments
825; HNMR (500 MHz, CDCl3) d = 0.78 (d, J = 6.9 Hz,
3H), 0.98 (d, J = 6.4 Hz, 3H), 1.05 (t, J = 7.3 Hz, 3H), 1.07
(s, 3H), 1.09 (s, 3H), 1.19 (d, J = 7.3 Hz, 3H), 1.73–1.78
(m, 1H), 1.86 (s, 3H), 1.93 (dd, J = 13.3, 9.2 Hz, 1H), 2.05–
2.15 (m, 2H), 2.13 (s, 3H), 2.19–2.23 (m, 3H), 2.62–2.69
(m, 1H), 2.77–2.82 (m, 1H), 3.58–3.64 (m, 2H), 3.85–3.90
(m, 2H), 4.61 (d, J = 7.8 Hz, 1H), 5.05 (d, J = 9.2 Hz, 1H),
5.23 (d, J= 9.6 Hz, 1H), 5.61–5.72 (m, 3H), 5.92 (d,
J = 9.6 Hz, 1H), 6.00 (d, J = 15.6 Hz, 1H), 6.59 (d,
J = 15.6 Hz, 1H), 6.69 (d, J = 9.6 Hz, 1H); 13CNMR
(125 MHz, CDCl3) d = 12.96, 13.31, 18.37, 24.54, 26.61,
32.12, 33.46, 35.60, 40.77, 45.29, 48.58, 54.15, 60.39, 62.14,
72.06, 86.98, 118.54, 121.27, 122.06, 128.63, 131.77,
135.05, 135.57, 136.94, 139.26, 156.97, 164.46, 172.52,
178.50, 215.26; HRMS (FAB) m/z (M+Na+): calcd for
C34H50O7Na: 593.3454. Found: 593.3453.
The authors thank Dr. Ge Li, Dr. Shu-Hui Chen, and
Dr. Wensheng Liao of WuXi PharmaTech in Shanghai,
China for the development of practical synthetic meth-
ods of intermediates.
References and notes
1. Umezawa, I.; Komiyama, K.; Oka, H.; Okada, K.;
Tomisaka, S.; Miyano, T.; Takano, S. J. Antibiot. 1984,
37, 706.
2. Arai, N.; Chikaraishi, N.; Omura, S.; Kuwajima, I. Org.
Lett. 2004, 6, 2845.
3. Roberts, B. J.; Hamelehle, K. L.; Sebolt, J. S.; Leopold,
W. R. Cancer Chemother. Pharmacol. 1986, 16, 95.
4. Yoshida, E.; Nishimuta, Y.; Naito, K.; Watanabe, Y.;
Tomisaka, S.; Okura, A. S.; Komiyama, K.; Umezawa, I.
J. Antibiot. 1987, 40, 391.
5. Phase I clinical trial of elactocin (leptomycin B) was
performed in UK: Newlands, E. S.; Rustin, G. J.
S.; Brampton, M. H. British J. Cancer 1996, 74,
648.
6. Private communication from Dr. Kanki Komiyama, The
Kitasato Institute.
7. Kazusamycin A also exhibited potent antimicrobial activ-
ity: Tunac, J. B.; Graham, B. D.; Dobson, W. E.; Lenzini,
M. D. J. Antibiot. 1985, 38, 460.
17. A human pancreatic ductal carcinoma cell line HPAC
CRL-2119 was obtained from the American Type Culture
Collection (ATCC) and cultured in 1:1 mixture of
Dulbecco’s modified Eagle’s medium and Ham’s F-12
medium supplemented with 10% fetal bovine serum in a
humidified atmosphere of 95% O2 and 5% CO2 at 37 ꢁC.
Cells were seeded at a concentration of 2 · 103 cells/well in
culture medium into 96-well microtiter plates. Kazusamy-
cin A and four derivatives were dissolved in DMSO. After
a 24 h incubation period, treatment of test compounds was
started. The final concentration of DMSO was 0.1%. After
72 h, 10 lL WST-1 reagent was added to each well, and
the plate was incubated at 37 ꢁC for 3 h. Absorbance of
the culture supernatant at 450 nm was measured using a
microplate reader.
8. Kudo, N.; Matsumori, N.; Taoka, H.; Fujiwara, D.;
Schreiner, E. P.; Wolff, B.; Yoshida, M.; Horinouchi, S.
Proc. Natl. Acad. Sci. USA 1999, 96, 9112.
9. Murakami, N.; Sugimoto, M.; Kobayashi, M. Bioorg.
Med. Chem. 2001, 9, 57.
18. ALT values (means SEM) of kazusamycin A (0.25 mg/
kg), 1a (0.5 mg /kg), 1c (0.5 mg/kg), and vehicle control
were 89.7 9.9, 27.3 7.2, 22.3 2.0, and 27.7 3.9,
respectively.
10. Furthermore, 33 and 10 steps were needed for the
syntheses of fragments B and D, respectively.
19. As regards values of AST (GOT) and LDH, similar
evidence was observed.