O-N Intramolecular acyl migration strategy in water-soluble prodrugs of taxoids

Article abstract of DOI:10.1016/j.bmcl.2003.09.020

We synthesized a highly water-soluble canadensol prodrug 6 that formed canadensol 3 by a simple pH-dependent chemical mechanism via the O-N intramolecular acyl migration of the isobutyryl group. This prodrug, a 2′-O-isobutyryl isoform of 3, has no additio

Full text of DOI:10.1016/j.bmcl.2003.09.020

Bioorganic & Medicinal Chemistry Letters 13(2003) 4441–4444
O–N Intramolecular Acyl Migration Strategy in Water-Soluble
Prodrugs of Taxoids
Mariusz Skwarczynski, Youhei Sohma, Maiko Kimura, Yoshio Hayashi,* Tooru Kimura
and Yoshiaki Kiso*
Department of Medicinal Chemistry, Center for Frontier Research in Medicinal Science, Kyoto Pharmaceutical University,
Yamashina-Ku, Kyoto 607-8412, Japan
Received 20 August 2003; accepted 6 September 2003
Abstract—We synthesized a highly water-soluble canadensol prodrug 6 that formed canadensol 3 by a simple pH-dependent
chemical mechanism via the O–N intramolecular acyl migration of the isobutyryl group. This prodrug, a 20-O-isobutyryl isoform of
3, has no additional functional auxiliaries released during the conversion to 3. This is a significant advantage in toxicology and
medical economics, since the potential side effects of reported water-soluble auxiliaries and the use of detergent for solubilization
can be avoided. The solubility of 6 was 2.26 mg mLꢀ1 and only the parent drug 3 was released under physiological conditions
(pH=7.4) while, in acidic medium, the release of 3 slowed until migration was completely obstructed at pH=2. In further con-
sideration of this strategy, we elucidated the use of an ‘O–N acyl-like’ migration reaction of the Boc group in the design of a doc-
etaxel prodrug. Both O–N migration and undesired hydrolysis of the Boc group occurred under physiological conditions, although
no oxazolidinone formation was observed, suggesting the limitation of our water-soluble prodrug strategy to docetaxel.
# 2003Elsevier Ltd. All rights reserved.
The introduction of anti-cancer agents paclitaxel
(Taxol1, 1)1 and docetaxel (Taxotere1, 2)2 has revolu-
tionized the treatment of cancer and markedly improved
the survival time of patients. More recently, other tax-
oids, such as canadensol (3),3 have been developed, with
improved potency against cancer showing multi-drug
resistance (MDR).4 However, despite the hope and
promise that these agents have engendered, their poor
water-solubility, based on their common taxane ring
structure, is a serious problem in intravenous adminis-
tration. Although this has prompted many researchers
to develop water-soluble paclitaxel prodrugs through
the introduction of hydrophilic moieties to C20 or/and
C7 positions,5 the released auxiliary moieties may have
negative effects in vivo, suggesting a need for novel
approaches to water-soluble prodrugs.
parent drug, paclitaxel (0.00025 mg mLꢀ1), and the
formation of paclitaxel through a simple pH-dependent
chemical mechanism via the O–N intramolecular acyl
migration of the benzoyl group.6 This prodrug, a 20-O-
benzoyl isoform of paclitaxel, releases no additional
functional auxiliaries during its conversion to paclitaxel.
Isotaxel had a practical conversion time under physi-
ological conditions (t1/2=15 min) and sufficient stability
in 0.035% citric acid saline (pH 4.0), one of the possible
injection conditions for practical clinical use. These
characteristics therefore promise a significant advantage
in toxicology and medical economics and the diversified
application of the O–N acyl migration prodrug to other
important water-insoluble taxoids such as 2 and 3 is
highly valuable.
In this paper, we studied the viability of the water-solu-
ble prodrugs of docetaxel (2) and canadensol (3) based
on O–N acyl migration. Both taxoids include an a-
hydroxy-b-amino acid structure in the phenylisoserine
part, as does paclitaxel (Fig. 1), but the functional
groups at the 30-position differed, with t-butyloxy-
carbonyl (Boc) and isobutyryl groups in 2 and 3,
respectively, instead of the benzoyl group in 1. Hence,
the water-soluble prodrugs 5 and 6, which are 20-O-Boc
Along the same lines, we previously designed and syn-
thesized a novel water-soluble paclitaxel prodrug, iso-
taxel 4 that realized a higher water-solubility (0.45 mg
mLꢀl), due to the ionized 30-amino group, than the
*Corresponding authors. Tel.: +81-75-595-4635; fax: +81-75-591-
9900; e-mail: kiso@mb.kyoto-phu.ac.jp
0960-894X/$ - see front matter # 2003Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2003.09.020

4442
M. Skwarczynski et al. / Bioorg. Med. Chem. Lett. 13 (2003) 4441–4444
and 20-O-isobutyryl isoforms of 2 and 3, respectively,
were designed to increase water solubility and allow
conversion to their corresponding parent drugs via O–N
acyl migration under physiological conditions.
Next, the effect of the isobutyryl group on the kinetics
of the O–N acyl migration was examined using a model
compound 15 which has a cyclohexyl structure instead
of the taxane ring in 6. Compound 15 was synthesized
through the N-Boc-protection of (2R,3S)-phenylisoser-
ine, followed by acylation with isobutyryl chloride,
coupling with cyclohexanol by the DCC/DMAP
method and final deprotection of the Boc group with
4 M HCl in dioxane (Scheme 2). The conversion time of
15 to the corresponding N-acyl compound was analyzed
in a similar manner as for 10. Compound 15 was com-
pletely and promptly converted to its N-acyl isomer
with a t1/2 value of 3.7 min with no disruptions, such as
hydrolysis. This faster migration in 15 could be
explained by the relatively higher electrophilic nature of
the carbonyl carbon of the isobutyryl group compared
to the benzoyl group in isotaxel.6 However, this is com-
patible with our previous study on prodrugs of HIV-1
protease inhibitors.9,10
Firstly, to examine the capability of O–N migration of
the Boc group, we synthesized a model compound 10,
which includes the Boc group at the 2-position in the
phenylisoserine part and a bulky cyclohexyl structure
instead of the taxane ring in 5 (Scheme 1). An amino
group
of
commercially
available
(2R,3S)-
.
phenylisoserine HCl 7 was protected by the benzylox-
ycarbonyl (Z) group in a conventional manner, then the
O-Boc group was introduced by coupling with Boc2O.
The resulting compound 8 was coupled with cyclohex-
anol using a DCC/DMAP method to afford 9. Finally,
the Z group of 9 was deprotected by hydrogenation,
followed by ion-exchange HPLC to give compound 10
as a hydrochloride salt.7 This compound was dissolved
in phosphate buffered saline (PBS, pH 7.4), incubated at
37 ꢁC and the kinetic profile was examined. As Figure 2
shows, 10 was consumed completely and the time-
dependent formation of parent compound 11 with N-
Boc was observed, indicating that ‘O–N acyl-like’
migration of Boc occurred through the formation of a
five-membered ring transition state. To the best of our
knowledge, the formation of t-butyloxycarbamate from
corresponding carbonate through intramolecular
migration is unprecedented. Moreover, we did not
observed any oxazolidinone formation, which was
reported in other carbonates under basic conditions.8
However, compound 12 was also produced by simple
hydrolysis of the carbonate bond in 10. Since the anti-
tumor activity of 2 without the Boc group was drasti-
cally lower than that of 2,2 further application of our
prodrug strategy to docetaxel 5 was abandoned.
This promising result prompted us to apply the strategy
to the water-soluble prodrug of canadensol 3. The pro-
drug 6 was synthesized according to Scheme 3. Com-
pound 16, prepared according to the procedures
described previously,6 was coupled with isobutyryl
Scheme 1. Synthesis of model compound 10: (a) benzyloxycarbonyl
chloride, NaHCO3, H2O/Et2O (1:1), 0 ꢁC to rt; (b) Boc2O, DMAP,
THF, rt; (c) cyclohexanol, DCC, DMAP, CH2Cl2, rt; (d) Pd/C, H2,
EtOAc, then ion exchange HPLC with 12 mM HCl.
Scheme 2. Synthesis of model compound 15: (a) Boc2O, Et3N, THF/
H2O, 0 ꢁC to rt; (b) isobutyryl chloride, pyridine, CH2Cl2, 0 ꢁC to rt;
(c) cyclohexanol, DCC, DMAP, CH2Cl2, rt; (d) 4 M HCl in dioxane,
anisole, 0 ꢁC to rt.
Figure 1. The O–N acyl migration reaction of taxoid prodrugs to their
corresponding parent drugs under physiological conditions (pH 7.4).

M. Skwarczynski et al. / Bioorg. Med. Chem. Lett. 13 (2003) 4441–4444
4443
Figure 2. The HPLC profiles of model compound 10 with the Boc group in PBS (pH 7.4, 37 ꢁC).
Scheme 3. Syntheses of canadensol prodrug 6: (a) isobutyryl chloride,
pyridine, CH2Cl2, 0 ꢁC; (b) Zn, AcOH, AcOEt, rt, then ion exchange
HPLC with 12 mM HCl.
chloride to give ester 17. The deprotection of both the
Figure 3. (A) HPLC profiles of prodrug 6 in PBS (pH 7.4, 37 ꢁC);
(B) migration of prodrug 6 in different pH at 37 ꢁC.
2,2,2-trichloroethyloxycarbonyl (Troc) groups of this
ester and the following purification with ion exchange
by HPLC gave prodrug 6 as an HCl salt.11 The water-
solubility and migration rate of 6 at pH 7.4 (37 ꢁC) were
examined (Fig. 3). The water-solubility was determined
as 2.26 mg mLꢀ1, 10-fold higher than that of cana-
densol (0.22 mg mLꢀ1).12 Complete migration was
observed at pH 7.4 with a t1/2 value of 4.3min similar to
that in model compound 15, indicating that the highly
bulky taxane ring did not influence the migration rate.
In addition, slower migration was observed at pH 4.9
with a value of 82.0 min and no migration at pH 2.0
after 6 h incubation. These findings suggested that the
kinetics of migration from 6 to parent drug 3 was clearly
pH-dependent. The prodrug 6 was stable in acidic
aqueous media (pH=2.0) for at least 6 h. In addition, a
.
solid of 6 HCI was also stable for at least 1 month at 4
ꢁC. Only slight migration of 6 (1% of canadensol was
released after 1 h) was observed during incubation in
0.035% citric acid saline (pH 4.0) at room temperature,
suggesting a possible injection condition for practical
clinical use.
In conclusion, the O-acyl isoform of canadensol 6, hav-
ing no auxiliary and no by-product during conversion
to parent drug 3, was synthesized and the water-solubi-
lity and conversion time at various pH conditions were

4444
M. Skwarczynski et al. / Bioorg. Med. Chem. Lett. 13 (2003) 4441–4444
examined. Prodrug 6 had a 10-fold higher water solubi-
lity than 3 and was converted to parent 3 through O–N
acyl migration with clear pH-dependency. This finding
supported the premise that this prodrug strategy based
on the O–N intramolecular acyl migration reaction has
significant advantages in toxicity and medical econom-
ics. Summarized, our studies of water-soluble prodrugs
in both paclitaxel and canadensol suggest that this
strategy can be applied to other 30-N-acyl taxoids with
similar promising results. However, it is important to
note that ‘O–N acyl-like’ migration for the docetaxel-
type prodrug is difficult to apply using our present pro-
drug strategy.
Damen, E. W. P.; Wiegerinck, P. H. G.; Braamer, L.; Sperling,
D.; de Vos, D.; Scheeren, H. W. Bioorg. Med. Chem. 2000, 8,
427. (d) Seligson, A. L.; Terry, R. C.; Bressi, J. C.; Douglass,
J. G., III; Sovak, M. Anti-Cancer Drugs 2001, 12, 305. (e)
Wrasidlo, W.; Gaedicke, G.; Guy, R. K.; Renaud, J.; Pitsinos,
E.; Nicolaou, K. C.; Reisfeld, R. A.; Lode, H. N. Bioconjugate
Chem. 2002, 13, 1093.
6. Hayashi, Y.; Skwarczynski, M.; Hamada, Y.; Sohma, Y.;
Kimura, T.; Kiso, Y. J. Med. Chem. 2003, 46, 3782.
7. Selected spectroscopic data for 10, mp 156–159 ꢁC, 1H
NMR (CD3OD, 400 MHz): d 7.50–7.43(m, 5H), 5.16 (d,
J=8.3Hz, 1H), 4.68 (d, J=8.4 Hz, 1H), 4.69–4.62 (m, 1H),
1.76–1.58 (m, 2H), 1.54–1.18 (m, 7H), 1.48 (s, 9H), 1.06–0.96
(m, 1H). HRMS (FAB+): calcd For C20H30NO5 [M++H]:
364.2124, found: 364.2128.
8. Bartnik, R.; Cebulska, Z.; Laurent, A. Tetrahedron Lett.
1983, 24, 4197.
9. Hamada, Y.; Ohtake, J.; Sohma, Y.; Kimura, T.; Hayashi,
Y.; Kiso, Y. Bioorg. Med. Chem. 2002, 10, 4155.
Acknowledgements
This research was supported in part by the Frontier
Research Program of the Ministry of Education,
Science and Culture of Japan, and grants from the
Ministry of Education, Science and Culture of Japan.
10. Hamada, Y.; Matsumoto, H.; Kimura, T.; Hayashi, Y.;
Kiso, Y. Bioorg. Med. Chem. Lett. 2003, 13, 2727.
11. Selected spectroscopic data for 6, mp 156–160 ꢁC, 1H
NMR (CD3OD, 400 MHz): d 8.08–8.06 (m, 2H), 7.76–7.71 (m,
1H), 7.66–7.50 (m, 6H), 7.38–7.34 (m, 1H), 6.41 (s, 1H), 5.93
(t, J=8.0 Hz, 1H), 5.59 (d, J=7.3Hz, 1H), 5.33(d, J=9.0 Hz,
1H), 4.97 (dd, J=9.6, 1.9 Hz, 1H), 4.82 (d, J=9.0 Hz, 1H),
4.31 (dd, J=6.5, 11.0 Hz, 1H), 4.17, 4.14 (2d, J=8.2 Hz, 2H),
3.72 (d, J=7.1 Hz, 1H), 2.84 (septet, J=7.0 Hz, 1H), 2.48–
2.41 (m, 1H), 2.27 (s, 3H), 2.16 (s, 3H), 1.92–1.85 (m, 1H), 1.85
(d, J=1.3 Hz, 3H), 1.81–1.74 (m, 1H), 1.62 (s, 3H), 1.46–1.40
(m, 1H), 1.28 (d, J=7.0 Hz, 3H), 1.25 (d, J=6.8 Hz, 3H), 1.12
(s, 3H), 1.10 (s, 3H). HRMS (FAB+): calcd For C44H54NO14
[M++H]: 820.3544, found: 820.3542. Purity was higher than
98% (HPLC analysis at 230 nm).
References and Notes
1. Wani, M. C.; Taylor, H. L.; Wall, M. E.; Coggon, P.;
McPhail, A. T. J. Am. Chem. Soc. 1971, 93, 2325.
2. Gueritte-Voegelein, F.; Guenard, D.; Lavelle, F.; Le Goff,
M. T.; Mangatal, L.; Potier, P. J. Med. Chem. 1991, 34, 992.
3. Zamir, L.; Caron, G.; Zheng, Y. F. US Patent, 6,410,756,
1997; Chem. Abstr. 1998, 128, 321780.
4. Lin, S.; Ojima, I. Exp. Opin. Ther. Pat 2000, 10, 869.
5. (a) For example: Nicolau, K. C.; Riemer, C.; Kerr, M. A.;
Rideout, D.; Wrasidlo, W. Nature 1993, 364, 464. (b) Khmel-
nitsky, Y. L.; Budde, C.; Arnold, M. J.; Usyatinsky, A.; Clark,
D. S.; Dordick, J. S. J. Am. Chem. Soc. 1997, 119, 11554. (c)
12. Canadensol 3 was synthesized via simple acylation (i-
0
.
PrCOOH, HOBt, EDC HCl) of 3 -amine group of Troc-
deprotected (Zn, AcOH, AcOEt) compound 16. Spectroscopic
data are identical with data from ref 3.