Synthesis of novel water soluble benzylazolium prodrugs of lipophilic azole antifungals

Article abstract of DOI:10.1016/S0960-894X(02)00557-7

Water soluble N-benzyltriazolium or N-benzylimidazolium salt type prodrugs of several highly lipophilic triazole or imidazole antifungals have been synthesized. They were designed to undergo an enzymatic activation followed by a self-cleavage to release a

Full text of DOI:10.1016/S0960-894X(02)00557-7

Bioorganic & Medicinal Chemistry Letters 12 (2002) 2775–2780
Synthesis of Novel Water Soluble Benzylazolium Prodrugs of
Lipophilic Azole Antifungals
Jun Ohwada,a Chikako Murasaki,a Toshikazu Yamazaki,b Shigeyasu Ichihara,c
Isao Umedaa,* and Nobuo Shimmaa
aDepartment of Chemistry, Nippon Roche Research Center, 200 Kajiwara, Kamakura, Kanagawa 247-8530, Japan
bDepartment of Mycology, Nippon Roche Research Center, 200 Kajiwara, Kamakura, Kanagawa 247-8530, Japan
cDepartment of Preclinical Science, Nippon Roche Research Center, 200 Kajiwara, Kamakura, Kanagawa 247-8530, Japan
Received 5 June 2002; accepted 5 June 2002
Abstract—Water soluble N-benzyltriazolium or N-benzylimidazolium salt type prodrugs of several highly lipophilic triazole or
imidazole antifungals have been synthesized. They were designed to undergo an enzymatic activation followed by a self-cleavage to
release a parent drug. The prodrugs such as 16 had enough chemical stability and water solubility for parenteral use and were
rapidly and quantitatively converted to the active substance in human plasma.
# 2002 Elsevier Science Ltd. All rights reserved.
3
Introduction
and several groups modified the hydroxyl group for
preparing a prodrug though there is a considerable
steric hindrance around the hydroxyl group. To avoid
such a synthetic difficulty, we designed a benzylazolium
salt type prodrug in reference to the work of Bogardus
and Higuchi.4 The general concept of prodrugs is
illustrated in Scheme 1.
Systemic fungal infection is increasing, but no anti-
fungal agent is satisfactory with respect to either activ-
ity, spectrum or safety. Although fluconazole (FCZ)
and amphotericin B are currently used for the treatment
of systemic mycoses, fluconazole has problems in anti-
fungal spectrum and emergence of resistant strains,
whereas amphotericin B has a problem of nephrotoxic.1
Itraconazole (ITCZ) and the azoles under development
such as BMS-207147 have a drawback of low oral
absorption due to their high lipophilicity despite having
the broad spectrum.2 Consequently, there still exist high
medical needs for an injectable agent with a broad
spectrum for the treatment of severe deep mycoses of
hospitalized patients. There are some reports to try to
overcome this problem by a prodrug approach.3 For
example, Ichikawa et al have prepared TAK-457 as a
water soluble antifungal prodrug by quaternizing of tri-
azole ring. Because there was room for improvement in
stability and/or water solubility of antifungal prodrugs,
we also started to develop a new prodrug approach
which could be applicable for the existing triazole and
imidazole antifungals.
The prodrugs employ 4-hydroxybenzyl moiety5 as a
self-cleavable linker and an amino acid ester moiety as a
solubilizing part. First, the ester moiety of the prodrug
can be rapidly hydrolyzed by a nonspecific enzyme,
serum esterase, to generate a phenolic intermediate, and
then it undergoes a spontaneous degradation to release
the active substance and a benzylalcohol derivative.
Quaternary salt accelerates the chemical degradation
speed of the phenolic intermediate. In the previous art, a
phosphate of 4-hydroxybenzyl moiety5a was used as a
prodrug part. However, its bioconversion is still slow
because a phosphate is generally more stable than an
ester group in plasma. In this paper, we describe the
design and synthesis of water soluble prodrugs of azole
type antifungals, and their physicochemical and
biological properties.
As shown in Figure 1, some of the potent azole anti-
fungals such as BMS-207147 have a tertiary alcohol,
Chemistry
We used ITCZ, ketoconazole (KCZ) and BMS-207147
as parent drugs to pursue feasibility studies. First, we
synthesized a benzyltriazolium derivative of ITCZ as
*Corresponding author. Tel.: +81-467-47-2244; fax: +81-467-45-
6824; e-mail: isao.umeda@roche.com
0960-894X/02/$ - see front matter # 2002 Elsevier Science Ltd. All rights reserved.
PII: S0960-894X(02)00557-7

2776
J. Ohwada et al. / Bioorg. Med. Chem. Lett. 12 (2002) 2775–2780
Figure 1.
Scheme 1.
shown in Scheme 2. 3,5-Dimethyl-4-hydroxybenzyl
bromide 1, prepared from 3,5-dimethyl-4-hydroxy-
benzyl alcohol and hydrobromic acid, was reacted
with ITCZ in chloroform-acetonitrile to give a phenolic
intermediate 2. Acetylation of 2 in acetic anhydride-
pyridine gave 3 in high yield. Compound 4 was pre-
pared from KCZ according to a manner analogous to
that of 3.
with N-tert-butoxycarbonyl sarcosine and WSCI in di-
chloromethane in the presence of 4-dimethylaminopyr-
idine (DMAP) to give 13. Reduction of compound 13
followed by bromination with carbon tetrabromide and
triphenylphosphine gave 14. BMS-207147 was reacted
with 14 in acetonitrile to give compound 15 as a
quaternary benzyltriazolium salt. Deprotection of 15
by hydrochloric acid in ethyl acetate afforded 16
6
8
quantitatively.
Compound 6 was prepared from 4-acetoxybenzyl bro-
mide and KCZ in chloroform. Compound 7 and 8 were
also prepared from 4-acetoxy-3-methylbenzyl bromide
and 4-pivaloyloxybenzyl bromide, respectively, in a
same manner (Scheme 3). Prodrugs of BMS-207147
having amino acid esters were synthesized by two
methods described in Schemes 4 and 5. 1 was treated
with BMS-207147 in acetonitrile to give an intermediate
9 as a white powder, 7 which was successively condensed
with N-tert-butoxycarbonyl proline using 1-(3-
dimethylaminopropyl)-3-ethylcarbodiimide hydrochlo-
ride (WSCI) to give 10.
Results and Discussion
All the reactions above smoothly proceeded and gave
sufficiently stable quaternary benzylazolium prodrugs of
a variety of triazole and imdazole antifungals that are
currently used clinically or under development. Com-
pounds 3 and 4, 4-acetoxy-3,5-dimethylbenzylazolium
derivatives, were prepared for proof of the prodrug
concept. In addition, 6, 7 and 8 were synthesized to
evaluate the effect of methyl substituent in the benzyl
moiety on the stability of prodrugs using KCZ as a
parent drug. Then we prepared 11 and 16 to make water
soluble prodrugs of BMS-207147, which is one of the
most potent azole antifungal agents. We initially mea-
sured the conversion rate9 of the derivatives 3, 4, 6, 7, 8,
11 and 16 in human plasma and their solubility in water.
Proline ester 11 was obtained by removal of Boc
group with trifluoroacetic acid in dichloromethane as
a trifluoroacetic acid salt. Sarcosine ester prodrug of
BMS-207147, 16, was prepared from 3,5-dimethyl-4-
hydroxybenzaldehyde 12 as follows. 12 was treated

J. Ohwada et al. / Bioorg. Med. Chem. Lett. 12 (2002) 2775–2780
2777
As shown in Table 1, all the prodrugs had high water
solubility compared to parent drugs. Among them,
compound 16, a prodrug of BMS-207147 having a sar-
cosine ester moiety, was over 4Â104 times more soluble
in water than the parent drug. The conversion rate was
indicated as half life (T1/2: min) of prodrug dis-
appearance in plasma by hydrolysis of an ester moiety
(Table 1). Among the prodrugs of KCZ (4, 6, and 7)
having a defferent linker group, 4 was the most stable
(T1/2:8 min). These results suggested that the methyl
substituent in the benzyl moiety apparently affected the
stability of prodrugs against hydrolysis by plasma
esterase and 3,5-dimethyl substituent seemed to be
appropriate from a view point of chemical stability of a
prodrug.
Compounds 3, a prodrug of ITCZ having an acetate,
and 11, a prodrug of BMS-207147 having a proline
ester, showed relatively long half life (T1/2: 53 min and
126 min, respectively).
Figure 2 shows the conversion of 4 into KCZ in human
plasma. As 4 was hydrolyzed by esterase to give a phe-
nolic intermediate, ketoconazole was observed almost in
a same extent, suggesting a rapid degradation of the
intermediate and high conversion rate to KCZ.
Scheme 2. (a) ITCZ, CHCl3–CH3CN (7:3), rt; (b) Ac2O, pyridine, rt (93%, two steps).
Scheme 3. (a) KCZ, CHCl3, rt (76%).

2778
J. Ohwada et al. / Bioorg. Med. Chem. Lett. 12 (2002) 2775–2780
Compound 11 having a proline ester and 16 having
sarcosine ester exhibited quite different half life in
human plasma (T1/2: 126 and <2 min, respectively).
Both 11 and 16 were highly water soluble and chemi-
cally stable: 97.6% of 16 retained in 0.1% buffer at pH 3
(25 ꢀC) after 7 days.
Because compound 11 and 16 showed contrastive
plasma stability and were applicable for intravenous (iv)
formulation, 11 and 16 were evaluated in vivo using
fungal infectious models in male Fisher rat. For sys-
temic candidiasis model due to Candida albicans
ATCC48130, rats were treated with the test compounds
Scheme 4. (a) BMS-207147, CH3CN, rt (81%); (b) Boc-l-Pro, WSCI, DMAP, CH2Cl2, rt (79%); (c) 10% TFA–CH2Cl2, rt (quant).
Scheme 5. (a) Boc-Sar, WSCI, DMAP, CH2Cl2, rt (92%); (b) (1) NaBH4, THF, rt (quant); (2) CBr4, PPh3, CH2Cl2, rt (89%); (c) BMS-207147,
CH3CN, reflux (70%); (d) HCl, EtOAc, rt (quant).

J. Ohwada et al. / Bioorg. Med. Chem. Lett. 12 (2002) 2775–2780
2779
Table 1. Solubility in water and conversion rate in human plasma of
the prodrugs
Table 2. In vivo efficacy of compound 11, 16 and BMS-207147
against systemic candidiasis and systemic and pulmonary aspergillosis
in rat (11 and 16 were dissolved in saline, and fine powder of BMS-
207147 was suspended in 0.5%CMC solution)
Compd
Solubility in
water (mg/mL)
Conversion ratea in
human plasma
(T1/2) (min)
ED50 (mmol/kg) on day 14
3
4
6
7
8
11
1
>2
>2
6
>2
>5
49
53
8
<2
<2
4
126
<2
—
Systemic
Pulmonary
C. albicansa
A. fumigatusb
A. fumigatusb
11
11
16
16
iv3.5
po
iv4.6
po
po
>35
>55
17
18
<2.1
16
BMS-207147
17
8
<0.001
4.7
1.6
19
11
17
19
BMS-207147
aHydrolysis of ester moiety.
aStrain: ATCC48130, treatments: 1Âbid+2Âqd.
bStrain: CF437, treatments: 1Âbid+4Âqd.
Figure 2. Conversion of 4 in human plasma and formation of ketoconazole.
at 0, 4, 24 and 48 h after infection. For systemic and
pulmonary aspergillosis model due to Aspergillus fumi-
gatus CF437, rats were treated twice on the first day and
once daily on following 4 days. For pulmonary asper-
gillosis model, rats were immunosuppressed with corti-
sone acetate prior to infection. The in vivo efficacy of 11
and 16 is summarized in Table 2, and expressed as the
ED50 values on day 14.
and 16 as mentioned above. Thus, the prodrug 16
showed highly potent antifungal activities 10 in the three
infection models mentioned above in both iv and po
administration.
In conclusion, we designed and synthesized the new
water-soluble quaternary benzylazolium prodrugs of
several triazole and imidazole type antifungal agents.
Especially, 16 showed potent antifungal activity against
both systemic candidiasis and aspergillosis as well as pul-
monary aspergillosis in rat in both ivand oral adminis-
tration. Because most of antifungal azoles with a broad
spectrum are too lipophilic for parenteral formulation
and can not attain reliable blood levels after oral admin-
istration, this prodrug approach could be a useful and
promising method to make lipophilic antifungal azoles
injectable for chemotherapy of systemic fungal infections.
In the systemic candidiasis model in rat, prodrugs 11
and 16 exhibited strong activities against C. albicans
ATCC48130 in both ivand oral (po) administration. 11
and 16 were also highly active in the pulmonary asper-
gillosis model in rat in both administrations. However,
in the systemic aspergillosis model in rat, sarcosine ester
16 was more active than proline ester 11. It might come
from the big difference of T-half in plasma between 11

2780
J. Ohwada et al. / Bioorg. Med. Chem. Lett. 12 (2002) 2775–2780
0813, 1998 (d) Murtiashaw, C. W.; Stephenson, P. T. WO 97/
28169, 1997.
References and Notes
4. Bogardus, J. B.; Higuchi, T. J. Pharm. Sci. 1982, 71, 729.
5. (a) Li, J.; Luo, X.; Wang, Q.; Zheng, L.-M.; King, I.; Doyle,
T. W.; Chen, S.-H. Bioorg. Med. Chem. Lett. 1998, 8, 3159. (b)
Papot, S.; Combaud, D.; Bosslet, K.; Gerken, M.; Czech, J.;
Gesson, J.-P. Bioorg. Med. Chem. Lett. 2000, 10, 1835.
6. Compound 1 was prepared as follows. 3,5-Dimethyl-4-
hydroxybenzyl alcohol (2.5 g) in 5 mL of 48% HBr and 45 mL
of acetic acid was stirred for 40 min at 15 ꢀC and the mixture
was poured onto a mixture of chloroform and ice-water. The
organic layer was dried and concentrated to give 1 as a white
solid, which could be used without further purification.
7. Compound 9 was obtained as a stable powder and quickly
released BMS-207147 in an alkaline medium.
8. Compound 16 was gradually precipitated and could be
isolated by filtration. Anal. calcd for C34H34BrClF2N6O3S.5/
3H2O: C, 51.68; H, 4.76; N, 10.64; S, 4.06. Found: C, 51.64;
H, 4.93; N, 10.31; S, 3.94.
9. The compounds were incubated with human plasma at a
concentration of 10 mg/mL at 37 ꢀC for up to 120 min.
10. In the systemic candidiasis and aspergillosis model in rat,
the efficacy of 16 was slightly reduced in spite of its high con-
version rate in plasma. The reason is unclear so far.
1. (a) Maesaki, S.; Kohno, S. Saishin Igaku, 1992, 47. (b)
Ohno, R.; Mizoguchi, H.; Kume, H. Nikkei Medical Aug 10,
1993. (c) Saxena, S.; Khan, J. A.; Ghosh, P. C. J. Antimicrob.
Chemother. 1998, 42, 635. (d) Petit, C.; Cheron, M.; Joly, V.;
Rodrigues, J. M.; Bolard, J.; Gaboriau, F. J. Antimicrob.
Chemother. 1998, 42, 779.
2. (a) Fung-Tomc, J. C.; White, T. C.; Minassian, B.; Huczko,
E.; Bonner, D. P. Diagn. Microbiol. Infect. Dis. 1999, 35, 163.
(b) Kitazaki, T.; Ichikawa, T.; Tasaka, A.; Hosono, H.; Mat-
sushita, Y.; Hayashi, R.; Okonogi, K.; Itoh, K. Chem. Pharm.
Bull. 2000, 48, 1935. (c) Van Cutsem, J. Mycoses 1989, 32, 14.
(d) Nakamura, H.; Yamada, A.; Ohki, H.; Maki, K.; Kitano,
Y.; Takasugi, H.; Tanaka, H. Abstracts, the 17th Symposium
on Medicinal Chemistry, 6th Annual Meeting of Division of
Medicinal Chemistry, Abstract No. 1P-09, Nov. 19–21, 1997.
The Pharmaceutical Society of Japan, Tsukuba, Japan.
3. (a) Ichikawa, T.; Kitazaki, T.; Matsushita, Y.; Yamada,
M.; Hayashi, R.; Yamaguchi, M.; Kiyota, Y.; Okonogi, K.;
Itoh, K. Chem. Pharm. Bull. 2001, 49, 1102. (b) Ueda, Y;
Matiskella, J. D.; Golik, J; Hudyma, T. W.; Chen, C. WO 01/
52852, 2001 (c) Lieven, M.; Jan, H.; Jozef Maria, R. WO 98/