Novel galbonolide derivatives as IPC synthase inhibitors: Design, synthesis and in vitro antifungal activities

Article abstract of DOI:10.1016/j.bmcl.2003.09.072

A series of novel galbonolide derivatives having a modified methyl enol ether moiety were prepared in total synthetic procedures and evaluated for their in vitro antifungal activities. The antifungal activity was labile to modification of the enol ether functionality and almost all of the modified compounds lacked the activity except for the analogue with an introduction of a methylthio group at the C-6 position, which retained a modest antifungal potency against Cryptococcus neoformans.

Full text of DOI:10.1016/j.bmcl.2003.09.072

Bioorganic & Medicinal Chemistry Letters 14 (2004) 143–145
Novel galbonolide derivatives as IPC synthase inhibitors:
design, synthesis and in vitro antifungal activities
Hiroki Sakoh, Yuichi Sugimoto,* Hideaki Imamura, Shunji Sakuraba, Hideki Jona,
Rie Bamba-Nagano, Koji Yamada, Terutaka Hashizume and Hajime Morishima
Banyu Tsukuba Research Institute, Okubo-3, Tsukuba 300-2611, Ibaraki, Japan
Received 15 August 2003; accepted 30 September 2003
Abstract—A series of novel galbonolide derivatives having a modified methyl enol ether moiety were prepared in total synthetic
procedures and evaluated for their in vitro antifungal activities. The antifungal activity was labile to modification of the enol ether
functionality and almost all of the modified compounds lacked the activity except for the analogue with an introduction of a
methylthio group at the C-6 position, which retained a modest antifungal potency against Cryptococcus neoformans.
# 2003 Elsevier Ltd. All rights reserved.
Galbonolide A (rustmicin, 1, Fig. 1)1 is a 14-membered
macrolide antibiotic with antifungal activity against
clinically important strains such as Candida albicans and
Cryptococcus neoformans. It was originally isolated as
an inhibitor of wheat rust fungus by an unknown mode
of action, however, researchers at Merck research
laboratories recently discovered that fungicidal activities
of galbonolide A acted through an inhibitory process
against inositol phosphoceramide (IPC) synthase.2 The
inhibition of this essential biosynthetic step is reversible
and leads to the accumulation of ceramide in growing
cells, leading to cell death. In addition, selective toxicity
could be expected because no corresponding enzyme in
mammalian sphingolipid biosynthesis pathway exists.3
Therefore, it seemed an attractive compound possessing
a novel antifungal mechanism. Total synthesis of gal-
bonolide B (neorustmicin, 2)4 and determination of its
absolute stereochemistry have been also reported by
Tse.5
compound. Upon hydrolysis of the enol ether function
under acidic condition, an inactive ketone analogue was
formed.
At the beginning, our effort focused on constructing
more stable functionality as an alternative to the acid-
sensitive methylenolether. As a part of the total syn-
thetic approach to galbonolide A analogues, we recently
reported the synthesis of ‘benzene analogues’ that con-
tain benzene ring instead of diene part and more stabi-
lized enolether-equivalent groups.7 The resulting
‘benzene analogues’ did not show any appreciable anti-
fungal activities. Therefore, we designed several pro-
mising candidates by replacing methoxy group at the C-
6 position with hydrogen (23), fluoroalkoxy (25),
methylthio (26) and halogen (24) and varying the posi-
tion of methoxy group to C-5 (27, 28). Herein, we
describe the synthesis of novel galbonolide A analogues
that have the original carbon skeleton (diene part) and
their in vitro antifungal activities.8
Several degradation pathways are known, even under
physiological conditions, due to the presence of several
unstable functional groups.6 Under basic to neutral
conditions, rapid epimerization of the C-2 methyl group
to a significantly less potent diastereomer and sub-
sequent translactonization leads to a completely inactive
Keywords: 14-membered macrolide; IPC synthase; antifungal agent;
methylenolether; Macro–Dieckmann cyclization.
* Corresponding author. Tel.: +81-298-77-2000; fax: +81-298-77-
2027; e-mail: sugmtoyt@banyu.co.jp
Figure 1. Structure of galbonolide A and B.
0960-894X/$ - see front matter # 2003 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2003.09.072

144
H. Sakoh et al. / Bioorg. Med. Chem. Lett. 14 (2004) 143–145
The synthetic route to target compounds 23–28 is
shown in Scheme 1. According to the reported synthetic
strategy,5 we prepared the key intermediate 4 from
commercially available (S)-benzylglycidyl ether (3,
>99% ee).9 The construction of a,b-unsaturated esters
6–9 was accomplished by Horner–Wadsworth–Emmons
reaction or aldol reaction followed by dehydration
using methanesulfonyl chloride (MsCl) in the presence
of 1,8-diazabicyclo[5.4.0]undec-7-ene (DBU). After
reduction of ester groups with diisobutylaluminum
hydride (DIBAL-H), the resultant alcohols were con-
verted to allyl iodides by the action of PPh3, imidazole
and I2 combination, and subjected to condensation
reaction with lithium enolate derived from benzylidene
acetal 5 to give adducts 11–14.
methyl group to b-form, if necessary, afforded a pro-
tected macrolide. Finally, cleavage of 2,4,6-tri-
methylbenzylidene acetal under hydrolysis condition
(aqueous acetic acid) were conducted to successfully
furnish galbonolide A analogues 23–28.
All the new compounds were evaluated for their in vitro
antifungal activity against C. albicans, Cr. neoformans,
Aspergillus fumigatus and Saccharomyces cerevisiae.
Table 1 summarizes the minimum inhibitory concentra-
tions (MICs) and the concentration required for the
80% inhibition of fungal growth (IC80), in which
Amphotericin B, fluconazole, galbonolide A, C-2 epi-
mer of galbonolide A and galbonolide B were listed as
the reference compounds.
Methoxy analogues 15–16 were prepared starting from
a,b-unsaturated ester 6. Reduction of the ester moiety
of 6 with DIBAL-H followed by oxidation by means of
tetrapropylammonium perruthenate (TPAP) afforded
aldehyde 10. Aldol reaction with lithium enolate of 5
gave a secondary alcohol as a 1:1 mixture of diaster-
eomers. The resultant alcohols were converted to methyl
ethers 15–16 by MeI in the presence of potassium hexa-
methyldisilazide (KHMDS).
The introduction of functional groups including hydro-
gen (23), chlorine (24) and 2,2,2-trifluoroethoxy (25) at
the C-6 position resulted in a loss of potency. Other
analogues (27, 28) with methoxy groups at the C-5
position also lacked activity. Among all derivatives
modified at the C-6 position instead of methoxy group,
only the methylthio analogue (26)10 showed moderate
activity against Cr. neoformans ATCC9011 strain with
the MIC of 0.5 mg/mL and IC80 of 0.063 mg/mL, and
weak activity against C. albicans ATCC90028 strain and
S. cerevisia Y9 strain with the MICs of 64 mg/mL,
respectively. Although the antifungal activity was
reduced as compared with the parent galbonolide A,
this is the first antifungal analogue of galbonolide A
produced by using total synthetic methodology.
Deprotection of the 2-(trimethylsilyl)ethoxymethyl
(SEM) groups of 11–16 with tetraethylammonium
fluoride at 90 ꢀC, followed by esterification of the car-
boxylic acids resulting from hydrolysis with trimethylsi-
lyl (TMS)-diazomethane provided methyl esters.
Acetates, precursor of macrocycles, were prepared by
acetic
anhydride
using
dimethylaminopyridine
In conclusion, acid-labile methyl enol ether was mod-
ified through a total synthetic approach to afford sev-
eral galbonolide A (rustmicin) analogues, which were
evaluated for antifungal activities against a panel of
fungi in comparison with galbonolide A. The antifungal
(DMAP). Macro–Dieckmann cyclization proceeded
smoothly to afford the desired macrocycles 17–22 in
good yield (up to 83%). Introduction of a methyl group
to the cyclized product followed by epimerization of the
Scheme 1. Reagents: (a) (EtO)2P(O)CH2COOEt, LiCl, DBU, 80% (E only, for 6), (EtO)2P(O)CH2COOEt, NaH, NCS, 34% (E/Z=5:3, for 7),
CF3CH2OCH2COOMe, LHMDS, then MsCl, Et3N, DBU, 57% (E/Z=1:4, for 8), MeSCH2COOEt, LHMDS, then MsCl, Et3N, DBU, 75% (Z
only, for 9); (b) DIBAL-H, 50–94%; (c) I2, PPh3, imidazole; (d) 5, LHMDS, HMPA, 32–52%; (e) Et4NF, then TMSCHN2, 40–77%; (f) Ac2O,
DMAP, pyridine 91–97%; (g) LHMDS, HMPA, 66–83%; (h) MeI, t-BuOK, 35–86%; (i) t-BuOK; (j) AcOH–H2O, 61–80%; (k) DIBAL-H, 93%;
(l) TPAP, NMO, MS4A, 82%; (m) 5, LHMDS, HMPA; (n) MeI, KHMDS, 60% (m+n, two steps); (o) Et4NF, then TMSCHN2, 57%; (p) Ac2O,
DMAP, pyridine, 95–97%; (q) LHMDS, HMPA, 49–65%; (r) MeI, t-BuOK, 41–85%; (s) t-BuOK, 63–72%; (t) AcOH–H2O, 74%-quant.

H. Sakoh et al. / Bioorg. Med. Chem. Lett. 14 (2004) 143–145
Table 1. In vitro antifungal activities of novel rustmicin analogues
145
a
Strain
MIC (IC80)
AMPHb
FLCZc
C. albicans ATCC90028
Cr. neoformans ATCC9011
A. fumigatus TIMM1776
S. cerevisiae Y9
0.25
0.25
1
(0.5)
(8)
(>64)
(8)
4 (1)
0.063 (<0.031)
>64
1 (0.5)
>64
>64
>64
>64
>64
>64
>64
>64
NTd
>64 (32)
>64
>64
0.25
C. albicans ATCC90028
Cr. neoformans ATCC9011
A. fumigatus TIMM1776
S. cerevisiae Y9
>64 (64)
16 (0.15)
>64
NTd
>64
64 (16)
0.5 (0.063)
>64
64 (32)
>64
>64
>64
NTd
>64
>64
>64
NTd
32
>64
NTd
a MIC (mg/mL) determined by the broth microdilution method and numbers in parentheses indicate IC80 values (mg/mL).
b Amphotericin B.
c Fluconazole.
d Not tested.
14942. Aureobasidin A and Khafrefungin were also
known as IPC synthase inhibitors. See reviews: Fostel, J.-
M.; Lartey, P.-A. Drug Discov. Today 2000, 5, 25.
3. (a) Georgopapadakou, N.-H.; Walsh, T.-J. Antimicrob.
Agents Chemother. 1996, 40, 279. (b) Nagiec, M.-M.;
Nagiec, E.-E.; Baltisberger, J.-A.; Wells, G.-B.; Lester, R.-
L.; Dickson, R.-C. J. Biol. Chem. 1997, 272, 9809. (c)
Groll, A.-H.; De Lucca, A.-J.; Walsh, T.-J. Trends
Microbiol. 1998, 6, 117.
activity of synthetic analogues proved to be strictly
affected by the substituents at the C-6 position and most
of the synthetic analogues lost the antifungal activities.
However, the substitution of a methylthio group for the
methoxy group was tolerated and showed modest anti-
fungal activity against the Cr. neoformans strain.
Acknowledgements
4. Abe, Y.; Nakayama, H.; Shimizu, A.; Furihata, K.;
Ikeda, K.; Furihata, K.; Seto, H.; Ohtake, N. J. Antibiot.
1985, 38, 1810.
We are indebted to Mr. Shinnosuke Abe for FAB-
HRMS analyses and to Dr. Shigeru Nakajima for
NMR measurements. We are also grateful to Dr. Bruno
Tse, Merck & Co. Inc., for helpful discussions.
5. Tse, B. J. Am. Chem. Soc. 1996, 118, 7094.
6. Stability of rustmicin was analyzed and reported that
optimal pH was 5.5 and half-lives were 80 min at pH 7.0.
and 40 min in mouse plasma. See ref 2.
7. Sakoh, H.; Sakuraba, S.; Sugimoto, Y.; Imamura, H.; Jona,
H.; Yamada, K.; Bamba-Nagano, R.; Hashizume, T.; Mor-
ishima, H. Chem. Pharm. Bull., accepted for publication.
8. A part of this work was presented by the authors at the
22nd Symposium on Medicinal Chemistry, The Pharma-
ceutical Society of Japan, Poster No. 1P-17, Nov. 27–29,
2002, Shizuoka, Japan.
References and notes
1. (a) Takatsu, T.; Nakayama, H.; Shimizu, A.; Furihata,
K.; Ikeda, K.; Furihata, K.; Seto, H.; Ohtake, N. J. Anti-
biot. 1985, 38, 1806. (b) Achenbach, H.; Muhlenfeld, A.;
Fauth, U.; Zahner, H. Tetrahedron Lett. 1985, 26, 6167.
(c) Fauth, U.; Zahner, H.; Muhlenfeld, A.; Achenbath, H.
J. Antibiot. 1986, 39, 1760. (d) Tse, B.; Blazey, C.-M.; Tu,
B.; Balkovec, J. J. Org. Chem. 1997, 62, 3236. (e) Sig-
mund, J.-M.; Hirsch, C.-F. J. Antibiot. 1998, 51, 829. (f)
Harris, G.-H.; Shafiee, A.; Cabello, M.-A.; Curotto, J.-E.;
Genilloud, O.; Goklen, K.-E.; Kurtz, M.-B.; Rosenbach,
M.; Salmon, P.-M.; Thornton, R.-A.; Zink, D.-L.; Man-
dala, S.-M. J. Antibiot. 1998, 51, 837.
9. Purchased from Daiso Co., Ltd, Japan.
10. Selected data for compound 26: IR (KBr) nmax 2960,
2920, 1709, 1456 cmÀ1 1H NMR (300 MHz, CDCl3) d
;
0.79 (3H, t, J=7.4 Hz), 0.94 (3H, d, J=6.9 Hz), 1.43 (3H,
d, J=7.0 Hz), 1.50 (2H, m), 1.69 (3H, s), 1.89 (3H, s), 2.24
(2H, m), 2.66 (2H, AB q, J=14.8 Hz), 3.15 (1H, m), 3.66
(2H, AB q, J=11.9 Hz), 3.80 (1H, q, J=7.0 Hz), 4.85
(1H, s), 4.93 (1H, t, J=6.1 Hz), 4.99 (1H, s), 5.71 (1H, d,
J=9.3 Hz), 5.92 (1H, s); 13C NMR (125 MHz, C6D6) d
9.9, 14.4, 15.2, 15.6, 18.9, 26.2, 34.5, 37.2, 45.6, 50.8, 67.5,
80.9, 82.9, 117.2, 128.0, 135.4, 142.0, 143.6, 168.8, 208.4;
FAB-HRMS calcd for C21H32O5SNa+: 419.1868. Found
419.1873.
2. Mandala, S.-M.; Thornton, R.-A.; Milligan, J.; Rosen-
bach, M.; Garcia-Calvo, M.; Bull, H.-G.; Harris, G.;
Abruzzo, G.-K.; Flattery, A.-M.; Gill, C.-J.; Bartizal, K.;
Dreikorn, S.; Kurtz, M.-B. J. Biol. Chem. 1998, 273,