Molecular design of antifungal agents

Article abstract of DOI:10.1021/jf020088v

In a rational approach to the design of antifungal agents against Saccharomyces cerevisiae, a series of alkyl gallates (3,4,5-trihydroxybenzoates) were synthesized and assayed. Nonyl gallate (1) was found to be the most effective with a minimum fungicidal concentration (MFC) of 12.5 μg/mL (42 μM), followed by octyl gallate (2) with an MFC of 25 μg/mL (89 μM). These MFCs are little influenced by pH values. A time-kill curve study indicates that nonyl gallate exhibits fungicidal activity against S. cerevisiae at any growing stage. The antifungal activity of nonyl gallate is due primarily to its ability to act as a nonionic surface-active agent (surfactant). The length of the alkyl group is not a major contributor but plays a role in eliciting the activity to a large extent. As far as alkyl gallates are concerned, their antimicrobial spectra and potency depend largely on the hydrophobic portion of the molecules.

Full text of DOI:10.1021/jf020088v

3992 J. Agric. Food Chem. 2002, 50, 3992−3998
Molecular Design of Antifungal Agents
ISAO KUBO,*,† PING XIAO,† KEN-ICHI NIHEI,† KEN-ICHI FUJITA,†
‡
YOSHIRO YAMAGIWA,‡ AND TADAO KAMIKAWA
Department of Environmental Science, Policy and Management, University of California,
Berkeley, California 94720-3112, and Department of Chemistry, Faculty of Science and Technology,
Kinki University, Kowakae, Higashi-Osaka-shi, Osaka 577-8502, Japan
In a rational approach to the design of antifungal agents against Saccharomyces cerevisiae, a series
of alkyl gallates (3,4,5-trihydroxybenzoates) were synthesized and assayed. Nonyl gallate (1) was
found to be the most effective with a minimum fungicidal concentration (MFC) of 12.5 µg/mL (42
µM), followed by octyl gallate (2) with an MFC of 25 µg/mL (89 µM). These MFCs are little influenced
by pH values. A time-kill curve study indicates that nonyl gallate exhibits fungicidal activity against
S. cerevisiae at any growing stage. The antifungal activity of nonyl gallate is due primarily to its
ability to act as a nonionic surface-active agent (surfactant). The length of the alkyl group is not a
major contributor but plays a role in eliciting the activity to a large extent. As far as alkyl gallates are
concerned, their antimicrobial spectra and potency depend largely on the hydrophobic portion of the
molecules.
KEYWORDS: Antifungal activity; Saccharomyces cerevisiae; nonyl gallate; surfactant property; antioxi-
dant activity
INTRODUCTION
A possible solution to cross this hurdle is combining two or
more inhibitors to increase the total biological activity or their
synthetic modification. Several examples of the former have
already been reported (3, 4), so the emphasis for this paper was
placed on an example for the latter case, because the ideal
antimicrobial agent was suggested to be produced by a rational
design (5). To facilitate it, our previous structure-antimicrobial
activity relationship (SAR) study with a series of alkanols was
selected for this study as a model. Briefly, the maximum
antimicrobial activity depends on the hydrophobic alkyl (tail)
chain length from the hydrophilic hydroxyl group (head) and
also on the microorganisms being tested (6). A series of alkyl
benzoate derivatives were synthesized and tested for their
antifungal activity against S. cereVisiae for comparison.
Yeast fermentations are involved in the manufacturing of
foods such as bread, beer, wines, vinegar, and surface-ripened
cheese. Most yeasts of industrial importance are of the genus
Saccharomyces and mostly of the species S. cereVisiae. These
ascospore-forming yeasts are readily bred for desired charac-
teristics. However, yeasts are undesirable when they cause
spoilage to sauerkraut, fruit juices, syrups, molasses, honey,
jellies, meats, wine, beer, and other foods (1). The finishing
process of the fermentation is usually through either filtration
or pasteurization. However, the use of the latter is limited to
certain foods because it is a heat treatment and hence denatural-
izes proteins, and the former is also limited to clear liquids (2).
Neither process is applicable to some foods such as sauerkraut
and “miso” (soybean pastes). Therefore, safe and effective
fungicides are still needed to control yeasts with new modes of
action.
Safety is a primary consideration for antifungal agents,
especially for those in food products, which may be utilized in
unregulated quantities on a regular basis. In our continuing
search for antimicrobial agents, a number of phytochemicals
have been characterized as active principles against S. cereVisiae
from edible plants, spices, and beverages. However, the
individual activity of antifungal agents characterized in plants
is usually not potent enough to be considered for practical use.
MATERIALS AND METHODS
General Methods. The procedures used for antimicrobial assay were
the same as previously described (3, 4, 6). For the synthesis, the melting
points were measured on a micro hot-stage Yanaco MP-S2 and are
uncorrected. IR spectra were recorded on a Shimadzu FTIR-8100 or
Hitachi 100-80 spectrometer, and 1H NMR spectra were recorded on a
Varian Mercury-300BB, JEOL/JNM-GX500, or JEOL-GSX270 spec-
trometer in CDCl3 using TMS as the internal standard unless otherwise
noted. Mass spectra were measured on a JEOL LMS-HX 100 or JEOL-
JMS700 spectrometer. Silica gel TLC and column chromatography were
done on a Merck precoated TLC 5715 and Merck kieselgel 60 7734,
respectively. Air- and/or moisture-sensitive reactions were carried out
under an argon or nitrogen atmosphere. The organic solvents were
purified and dried using appropriate procedures.
* Author to whom correspondence should be addressed [telephone (510)
643-6303; fax (510) 643-0215; e-mail ikubo@uclink4.berkeley.edu].
† University of California.
Chemicals. Alkanols were available from our previous work (3, 6,
7). Gallic acid and its methyl and propyl esters used for the assay were
‡ Kinki University.
10.1021/jf020088v CCC: $22.00 © 2002 American Chemical Society
Published on Web 06/06/2002

Molecular Design of Antifungal Agents
J. Agric. Food Chem., Vol. 50, No. 14, 2002 3993
(t, J ) 6.5 Hz, 3H, -CH3), 4.24 (t, J ) 6.5 Hz, 2H, -OCH2), 7.16 (s,
2H, ArH); IR (KBr) 3520, 3495, 3010, 2960, 1686, 1615, 1480, 1395,
1270, 1140 cm-1
.
Dodecyl gallate (dodecyl 3,4,5-trihydroxybenzoate) (9) was obtained
in 65% yield as a colorless solid: 1H NMR (270 MHz, CDCl3) δ 0.88
(t, J ) 6.5 Hz, 3H, -CH3), 4.24 (t, J ) 6.5 Hz, 2H, -OCH2), 7.16 (s,
2H, ArH); IR (KBr) 3515, 3490, 3000, 2960, 1680, 1610, 1480, 1395,
1270, 1145 cm-1
.
Undecyl gallate (undecyl 3,4,5-trihydroxybenzoate) (8) was obtained
in 63% yield as a colorless solid: 1H NMR (270 MHz, CDCl3) δ 0.88
(t, J ) 6.5 Hz, 3H, -CH3), 4.24 (t, J ) 6.5 Hz, 2H, -OCH2), 7.16 (s,
2H, ArH); IR (KBr) 3500, 3450, 2960, 2890, 1685, 1620, 1480, 1395,
1280, 1160 cm-1
.
Decyl gallate (decyl 3,4,5-trihydroxybenzoate) (7) was obtained in
74% yield as a colorless solid: 1H NMR (300 MHz, CDCl3) δ 0.88 (t,
J ) 6.5 Hz, 3H, -CH3), 4.24 (t, J ) 6.5 Hz, 2H, -OCH2), 7.16 (s,
2H, ArH); IR (KBr) 3391, 2924, 1686, 1615, 1448, 1312, 1258, 1028
cm-1; EI-MS, m/z 310 (M+), 282, 226, 170 (base peak), 153, 125, 83,
55; HRMS-EI, m/z [M]+ calcd for C17H26O5 310.1778, found 310.1788.
Nonyl gallate (nonyl 3,4,5-trihydroxybenzoate) (1) was obtained in
86% yield as colorless needles from benzene/n-pentane: mp 96.7-
1
97.3 °C; H NMR (300 MHz, CDCl3/CD3OD) δ 0.88 (t, J ) 6.6 Hz,
3H, -CH3), 4.25 (t, J ) 6.6 Hz, 2H, -OCH2), 7.21 (s, 2H, ArH); IR
(KBr) 3489, 3439, 3333, 1670, 1625, 1604, 1531, 1300, 1256, 1024
cm-1; HRMS-EI, m/z [M]+ calcd for C16H24O5 296.1630, found
296.1624.
Nonyl protocatechuate (nonyl 3,4-dihydroxybenzoate) (13) was
obtained in 64% yield as a colorless solid: 1H NMR (300 MHz, CDCl3)
δ 0.89 (t, J ) 6.6 Hz, 3H, -CH3), 4.24 (t, J ) 6.6 Hz, 2H, -OCH2),
6.66 (d, J ) 7.8 Hz, 1H, ArH), 7.40 (dd, J ) 2.3, 7.8 Hz, 1H, ArH),
7.53 (d, J ) 2.3 Hz, 1H, ArH); IR (KBr) 3530, 3360, 2960, 2890,
1700, 1625, 1550, 1480, 1395, 1290, 1180 cm-1
.
Nonyl 3,5-dihydroxybenzoate (15) was obtained in 68% yield as a
colorless solid: 1H NMR (270 MHz, CDCl3) δ 0.88 (t, J ) 6.6 Hz,
3H, -CH3), 4.21 (t, J ) 6.6 Hz, 2H, -OCH2), 6.34 (t, J ) 2.3 Hz,
1H, ArH), 6.45 (d, J ) 2.3 Hz, 2H, ArH); IR (KBr) 3340, 2960, 2890,
1690, 1610, 1480, 1395, 1255, 1165 cm-1; HRMS-EI, m/z [M]+ calcd
for C16H24O4 280.1673, found 280.1675.
Nonyl 3-hydroxy-4-methoxybenzoate (19) was obtained in 76% yield
as a colorless solid: 1H NMR (270 MHz, CDCl3) δ 0.88 (t, J ) 6.6
Hz, 3H, -CH3), 4.20 (s, 3H, -OCH3), 4.26 (t, J ) 6.6 Hz, 2H,
-OCH2), 6.73 (d, J ) 8.7 Hz, 1H, ArH), 7.43 (dd, J ) 2.3, 8.7 Hz,
1H, ArH), 7.65 (d, J ) 2.3 Hz, 1H, ArH); IR (KBr) 3290, 2960, 2890,
1690, 1630, 1600, 1480, 1395, 1250, 1180 cm-1
.
Nonyl 4-hydroxy-3-methoxybenzoate (20) was obtained in 61% yield
as a colorless solid: 1H NMR (270 MHz, CDCl3) δ 0.88 (t, J ) 6.6
Hz, 3H, -CH3), 3.89 (s, 3H, -OCH3), 4.20 (t, J ) 6.6 Hz, 2H,
-OCH2), 6.78 (d, J ) 8.7 Hz, 1H, ArH), 7.47 (dd, J ) 2.3, 8.7 Hz,
1H, ArH), 7.53 (d, J ) 2.3 Hz, 1H, ArH); IR (KBr) 3360, 2960, 2890,
Figure 1. Chemical structures of gallates and related compounds.
1640, 1595, 1525, 1480, 1395, 1250, 1165 cm-1
.
Nonyl 4-hydroxybenzoate (12) was obtained in 64% yield as a
colorless solid: 1H NMR (270 MHz, CDCl3) δ 0.88 (t, J ) 6.6 Hz,
3H, -CH3), 3.66 (m, 2H, -CH2), 4.26 (t, J ) 6.6 Hz, 1H, -OCH2),
6.85 (d, J ) 8.7 Hz, 2H, ArH), 7.92 (d, J ) 8.7 Hz, 2H, ArH); IR
(KBr) 3360, 2960, 2890, 1640, 1605, 1550, 1480, 1395, 1252, 1175
purchased from Aldrich Chemical Co. (Milwaukee, WI). Benzoic acid,
geraniol, 4-hydroxybenzoic acid, 3-hydroxybenzoic acid, protocatechuic
acid (3,4-dihydroxybenzoic acid), 3,5-dihydroxybenzoic acid, 3-hy-
droxy-4-methoxybenzoic acid, 4-hydroxy-3-methoxybenzoic acid, and
1,1-diphenyl-2-p-picryhydrazyl (DPPH) were obtained from Sigma
Chemical Co. (St. Louis, MO). N,N-Dimethylformamide (DMF) was
purchased from EM Science (Gibbstown, NJ).
Synthesis. To a solution of the corresponding phenolic acid (2.00
mM) and alcohol (2.00 mM) in THF (6 mL) cooled at 0 °C was added
a solution of N,N′-dicyclohexylcarbodiimide (DCC) (4.2 mM) in THF
(6 mL). After the solution had been allowed to stir for 20 h, the solvent
was removed under reduced pressure. The residue was extracted with
ethyl acetate several times and filtered. The filtrate was washed
successively with diluted aqueous citric acid solution, saturated aqueous
sodium hydrogen carbonate solution, and water, dried over MgSO4,
and evaporated. The crude products were purified by chromatography
(SiO2; elution with CHCl3/MeOH, 98:2). Structures of the synthesized
esters were established by spectroscopic methods (UV, IR, MS, and
NMR; Figure 1).
cm-1
.
Nonyl 3-hydroxybenzoate (14) was obtained in 72% yield as a
colorless solid: 1H NMR (270 MHz, CDCl3) δ 0.82 (t, J ) 6.6 Hz,
3H, -CH3), 3.23 (m, 2H, -CH2), 4.24 (t, J ) 6.6 Hz, 2H, -OCH2),
6.87 (d, J ) 7.5 Hz, 1H, ArH), 6.95 (s, 1H, ArH), 6.98 (d, J ) 7.5 Hz,
1H, ArH), 7.18 (t, J ) 7.5 Hz, 1H, ArH); IR (KBr) 3360, 2960, 2890,
1640, 1590, 1551, 1480, 1395, 1252, 1170 cm-1; HRMS-EI, m/z [M]+
calcd for C16H24O3 264.1721, found 264.1725.
Nonyl benzoate (11) was obtained in 57% yield as a colorless solid:
1H NMR (270 MHz, CDCl3) δ 0.88 (t, J ) 6.6 Hz, 3H, -CH3), 4.20
(t, J ) 6.6 Hz, 2H, -OCH2), 7.32 (t, J ) 7.8 Hz, 2H, ArH), 7.45 (t,
J ) 7.8 Hz, 1H, ArH), 8.06 (d, J ) 7.8 Hz, 2H, ArH); IR (KBr) 2960,
2890, 1720, 1650, 1550, 1480, 1395, 1250, 1175 cm-1
.
Octyl gallate (octyl 3,4,5-trihydroxybenzoate) (2) was obtained in
89% yield as a colorless solid: 1H NMR (270 MHz, CDCl3) δ 0.88 (t,
J ) 6.5 Hz, 3H, -CH3), 4.24 (t, J ) 6.5 Hz, 2H, -OCH2), 7.15 (s,
Tridecyl gallate (tridecyl 3,4,5-trihydroxybenzoate) (10) was obtained
in 77% yield as a colorless solid: 1H NMR (270 MHz, CDCl3) δ 0.88

3994 J. Agric. Food Chem., Vol. 50, No. 14, 2002
Kubo et al.
2H, ArH); IR (KBr) 3325, 2926, 1686, 1614, 1468, 1314, 1246, 1028
cm-1; EI-MS, m/z 282 (M]+), 224, 170, 143, 99, 56 (base peak) 41;
HRMS-EI, m/z [M]+ calcd for C15H22O5 282.1469, found 282.1467.
Heptyl gallate (heptyl 3,4,5-trihydroxybenzoate) (6) was obtained
in 71% yield as a colorless solid: 1H NMR (270 MHz, CDCl3) δ 0.88
(t, J ) 6.5 Hz, 3H, -CH3), 4.24 (t, J ) 6.5 Hz, 2H, -OCH2), 7.16 (s,
2H, ArH); IR (KBr) 3389, 2920, 1685, 1620, 1448, 1320, 1258, 1140
Table 1. Antifungal Activity of Alkanols against S. cerevisiae ATCC
7754
µg/mL
alkanol tested
MIC
MFC
C
>1600
800
200
100
50
>1600
>1600
800
200
50
6
C
7
cm-1
.
C
8
Hexyl gallate (hexyl 3,4,5-trihydroxybenzoate) (5) was obtained in
58% yield as a colorless solid: 1H NMR (270 MHz, CDCl3) δ 0.88 (t,
J ) 6.5 Hz, 3H, -CH3), 4.18-4.28 (br, 2H, -OCH2), 6.4-5.8 (br,
3H, exchangeable with D2O, -ArOH), 7.20 (s, 2H, ArH); IR (KBr)
C
9
C
10
C
25
25
11
C
12.5a
>1600
12
b
3389, 2932, 1688, 1615, 1449, 1314, 1258, 1026 cm-1
.
C
13
>1600
−
Hexyl protocatechuate (hexyl 3,4-dihydroxybenzoate) (16) was
obtained in 63% yield as a colorless solid: 1H NMR (270 MHz, CDCl3)
δ 0.89 (t, J ) 6.6 Hz, 3H, -CH3), 4.24 (t, J ) 6.6 Hz, 2H, -OCH2),
6.66 (d, J ) 7.8 Hz, 1H, ArH), 7.40 (dd, J ) 2.3, 7.8 Hz, 1H, ArH),
7.53 (d, J ) 2.3 Hz, 1H, ArH); IR (KBr) 3400, 3310, 2960, 2890,
a This value is variable. b −, not tested.
demonstrated no visible growth. The MFCs were examined as follows.
After the MIC had been determined, a 30-µL aliquot was taken from
each clear tube and added into 3 mL of drug-free fresh medium. After
48 h of incubation, the MFC was determined as the lowest concentration
of the test compound in which no recovery of microorganism was
observed.
Time-kill studies were performed to examine the effects of
combinations of compounds in more detail. The culture tubes were
prepared as described above and incubated at 30 °C for 5 h. A 30-µL
aliquot of the culture was inoculated into 3 mL of ME broth containing
appropriate concentrations of the test compounds. The initial population
size for S. cereVisiae was 5.8 × 105 CFU/mL. Samples were taken at
selected times during 48 h of exposure, and serial dilutions were made
in sterile saline before the samples were plated onto YPD agar plates.
The plates were incubated at 30 °C for 24 h before the number of
CFU was determined.
Adsorption Test. The test strain was cultured with shaking in YPD
broth overnight at 30 °C and washed twice with 50 mM MOPS buffer
(pH 6.0). After each gallate ester was mixed with or without yeast cells
(108 cells/mL) in the above buffer at 30 °C, the suspension was vortexed
for 5 s. Absorbance of the supernatants obtained by centrifugation for
2 min was measured at 272 nm.
Radical Scavenging Activity on DPPH. The reaction mixture
consisted of 1 mL of 100 mM acetate buffer (pH 5.5), 1 mL of ethanol,
0.5 mL of ethanolic solution of DPPH, and 0.5 mL of the sample
solution. After the mixture was allowed to stand at room temperature
for 20 min, the absorbance of the remaining DPPH was determined
colorimetrically at 517 nm. The scavenging activity was measured as
the decrease in absorbance of the DPPH expressed as a percentage of
the absorbance of a control DPPH solution (8). Inhibitory activity was
expressed as the mean 50% inhibitory concentration of triplicate
determinations, obtained by interpolation of concentration-inhibition
curves.
1720, 1625, 1550, 1480, 1395, 1290, 1185 cm-1
.
Hexyl 3,5-dihydroxybenzoate) (17) was obtained in 74% yield as a
colorless solid: 1H NMR (270 MHz, CDCl3) δ 0.89 (t, J ) 6.6 Hz,
3H, -CH3), 4.21 (t, J ) 6.6 Hz, 2H, -OCH2), 6.35 (d, J ) 2.3 Hz,
1H, ArH), 6.45 (d, J ) 2.3 Hz, 2H, ArH); IR (KBr) 3341, 2960, 2890,
1685, 1610, 1480, 1395, 1250, 1160 cm-1
.
Hexyl 4-hydroxybenzoate (18) was obtained in 66% yield as a
colorless solid: 1H NMR (270 MHz, CDCl3) δ 0.88 (t, J ) 6.6 Hz,
3H, -CH3), 3.66 (m, 2H, -CH2), 4.26 (t, J ) 6.6 Hz, 2H, -OCH2),
6.85 (d, J ) 8.7 Hz, 2H, ArH), 7.92 (d, J ) 8.7 Hz, 2H, ArH); IR
(KBr) 3585, 3020, 2930, 2859, 1705, 1600, 1450, 1275, 1270, 1225
cm-1
.
Decahydro-2-naphthyl gallate (23) was obtained in 66% yield as a
colorless solid: 1H NMR (500 MHz, CDCl3) δ 0.80-2.20 (m, 16H,
-CH2 and -CH), 4.90-5.20 (m, 1H, -OCH), 6.10 (br s, 3H, -OH),
7.30-7.35 (m, 2H, ArH); IR (Nujol) 3440, 1660, 1600, 1540, 1440,
1330, 1250, 1240, 1030 cm-1; EI-MS, m/z 306 (M+), 224, 170 (base
peak), 153, 136, 121, 95, 41; HRMS-EI, m/z [M]+ calcd for C17H22O5
306.1467, found 306.1444.
Geranyl gallate (3,7-dimethylocta-2,6-dienyl 3,4,5-trihydroxyben-
zoate) (21) was obtained in 57% yield as colorless powder: mp 67.0-
1
68.5 °C; H NMR (300 MHz, CDCl3) δ 1.60 (s, 3H, -CH3), 1.67 (d,
J ) 0.6 Hz, 3H, -CH3), 1.75 (d, J ) 1.2 Hz, 3H, -CH3), 2.08 (m,
4H, -CH2), 4.79 (d, J ) 7.2 Hz, 2H, -OCH2), 5.09 (m, 1H, -CH),
5.43 (qt, J ) 7.2, 1.2 Hz, 1H, -CH), 6.12 (br s, 3H, -OH), 7.31 (s,
2H, ArH); IR (KBr) 3549, 3406, 3288, 1686, 1614, 1537, 1312, 1244,
1229, 1198, 1022 cm-1; EI-MS, m/z 306 (M+), 237, 170, 153 (base
peak), 136, 121, 93, 69, 41; HRMS-FAB, m/z [M + H]+ calcd for
C17H23O5 307.11546, found 307.1561.
Test Strain. The test strain S. cereVisiae ATCC 7754 used for this
study was purchased from the American Type Culture Collection
(Rockville, MD).
Medium. S. cereVisiae was maintained at -80 °C in yeast nutrient
broth (YNB; Difco Laboratories, Detroit, MI) containing 25% glycerol
and subcultured at 30 °C in Sabouraud’s dextrose agar (SDA) medium
(Bactopeptone 1%, dextrose 4%, Bacto-agar 1.8%). A fresh culture
was preincubated with shaking for 5 h at 30 °C in 2.5% malt extract
(ME) broth (BBL) medium.
Antifungal Assay. The maximum extent and rate of activity are
known to vary with the seed culture mediums, the physiological age
of the culture, and the type of culture medium. All antifungal
susceptibility tests in this study were performed under a standard
condition using fresh inoculum from a 5-h shaking culture in ME
medium, final inoculum size of 105 colony forming units (CFU)/mL,
and a 48-h stationary incubation in ME medium, unless otherwise
specified.
Broth macrodilution minimum inhibitory concentrations (MICs) were
determined as previously described (7). Briefly, serial 2-fold dilutions
of the test compounds were made in DMF, and 30 µL of 100×
concentrated solution was added to 3 mL of ME medium. These were
inoculated with 30 µL of seed culture to give the final inoculum of 105
CFU/mL. The assay tubes were incubated without shaking at 30 °C
for 48 h. The MIC is the lowest concentration of test compound that
RESULTS
Primary alkanols from C6 to C13 were first tested for their
antifungal activity against S. cereVisiae, and the results are listed
in Table 1. Among them, undecanol (C11) was found to be the
most potent with the MFC of 25 µg/mL (145 µM), whereas
dodecanol (C12) was completely inactive. Interestingly, fungi-
cidal activity against S. cereVisiae completely disappears after
dodecanol (C12), and this is known as the so-called cutoff
phenomenon. In the case of undecanol, no differences in MIC
and MFC were noted, suggesting that no residual fungistatic
activity is involved. In contrast, dodecanol still showed fungi-
static activity with an MIC of 12.5 µg/mL but did not exhibit
any fungicidal activity up to 1600 µg/mL. Yeast cells appeared
to adapt to dodecanol stress, eventually recovering and growing
normally. It appears that the length of the alkyl group is
associated with the activity to a large extent. On the other hand,
the similar series of antimicrobial (2E)-alkenals characterized
from olive oil (6) and anacardic acids identified in the cashew

Molecular Design of Antifungal Agents
J. Agric. Food Chem., Vol. 50, No. 14, 2002 3995
Table 2. Antifungal Activity of Phenolic Acids against S. cerevisiae
ATCC 7754
µg/mL
acid tested
benzoic acid
4-hydroxybenzoic acid
3-hydroxybenzoic acid
3,5-dihydroxybenzoic acid
3,4-dihydroxybenzoic acid
3,4,5-trihydroxybenzoic acid
MIC
800
1600
3200
3200
3200
>3200
MFC
1600
>3200
>3200
>3200
>3200
a
−
a −, not tested.
Table 3. Antifungal Activity of Alkyl Gallates against S. cerevisiae
ATCC 7754
Figure 2. Effect of nonyl gallate (1) on the growth of S. cerevisiae ATCC
7754: 0 (b) (control), 3.13 (O), 6.25 (4), and 12.5 (0) µg/mL nonyl
gallate. Exponentially growing cells were inoculated into ME broth and
then cultured at 30 °C without shaking.
µg/mL
gallate tested
MIC
MFC
a
C
>3200
3200
400
−
>3200
400
1
did not exhibit any antifungal activity against S. cereVisiae,
because they may no longer possess the head and tail structure
balance due to their increased hydrophobicity, but still showed
antibacterial activity against Gram-positive bacteria (data not
listed). The maximum activity of the carbon chain lengths in
the gallates differed between the microorganisms tested, similar
to those found for alkanols (6). The maximum activity seems
to depend on the carbon chain lengths, reflecting differences in
their cell-envelope structures. On the basis of our previous SAR
study with alcohols, geranyl gallate (21) can be expected to
show activity similar to that of nonyl gallate. The rationale for
this idea is that the hydrophobic alkyl chain length from the
hydrophilic hydroxy group as well as its volume is known to
be a key factor in eliciting the activity (6). It seems that geraniol
(22) nicely fits this assumption as a superior example. The
introduction of branching or unsaturation into the hydrophobic
group is known to increase the solubility of the surfactant in
water (11). Hence, geranyl gallate was synthesized and assayed.
As expected, this gallate exhibited antifungal activity against
S. cereVisiae with an MFC of 50 µg/mL, as potent as nonyl
gallate. No differences in MIC and MFC were noted, suggesting
that no residual fungistatic activity is involved. In addition to
its structural appropriateness, an additional reason of selection
of geraniol as an alcohol moiety of this ester should be
mentioned. This monoterpene alcohol is known to increase
glutathione S-transferase activity, which is believed to be a major
mechanism for chemical carcinogen detoxification (12).
The fungicidal effect of both nonyl and geranyl gallates was
confirmed by the time-kill curve experiment. Cultures of S.
cereVisiae, with a cell density of 6 × 105 CFU/mL, were
exposed to two different concentrations of nonyl gallate. The
number of viable cells was determined following different
periods of incubation with nonyl gallate. Figure 2 verifies that
the MIC and MFC of nonyl gallate are the same. It shows that
half-MIC slows growth but that the final cell count is not
significantly different from the control. The result shows that
lethality occurs remarkably quickly, within the first 1 h. A
similar result was also obtained with geranyl gallate as shown
in Figure 3. Noticeably, its lethality also occurs fairly quickly,
similar to that of nonyl gallate. The results obtained likely
indicate that both gallates are common in possessing a mem-
brane disruptive effect.
C
3
C
6
C
100
12.5
100
25
12.5
25
>1600
>1600
50
7
C
8
C
6.25
12.5
>1600
>1600
50
9
C
10
C
11
C
12
geranyl
decahydro-2-naphthyl
50
50
a −, not tested.
apple (9) were observed to show similar SAR. These results
indicate that the hydrophilic hydroxyl group can be replaced
by any hydrophilic groups as long as the head and tail structure
is balanced.
To create effective antifungal agents, phenolic acids were
selected as the alternative hydrophilic (head) portion. The
antifungal activities of the selected phenolic acids themselves
against S. cereVisiae are listed in Table 2. They possessed little
or no activity against S. cereVisiae. For example, gallic acid
(3,4,5-trihydroxybenzoic acid; 3) did not exhibit any fungicidal
activity up to 3200 µg/mL. However, this phenolic acid was
first selected as the hydrophilic (head) part because gallic acid
is a common natural product isolated from many edible plants
and, more importantly, its three estersspropyl, octyl, and
dodecyl (lauryl)sare currently permitted for use as antioxidant
additives in food (10). Furthermore, synthesis of a series of the
esters and their related analogues is essential for SAR study,
and this can be easily carried out. All of the esters used for the
assay were synthesized as described under Materials and
Methods. It should be noted that synthesis was achieved up to
eicosanyl gallate (C20), but the assay data were obtained only
up to tridecyl gallate (C13) (10), because an alkyl chain length
of >11 carbon atoms did not exhibit any antifungal activity
against S. cereVisiae in our preliminary assay. In addition,
several nonyl benzoate analogues such as 4-hydroxybenzoate,
3-hydroxybenzoate, 3,4-dihydroxybenzoate (protocatechuate),
and 3,5-dihydroxybenzoate were synthesized according to the
same method for comparison.
The antifungal activities of the alkyl gallates synthesized
against S. cereVisiae are listed in Table 3. The maximum
fungicidal activity against this yeast was found in nonyl gallate
(C9) with an MFC of 12.5 µg/mL (42 µM), whereas decyl gallate
(C10) (7) no longer showed any activity up to 1600 µg/mL. It is
worth noting that undecyl (C11) (8) and dodecyl (C12) (9) gallates
Further evidence for this surfactant postulate was obtained
in experiments that showed nonyl gallate to be fungicidal at
any growing culture. Nonyl gallate at 1× MFC rapidly reduced
the number of viable cells (106 CFU/mL within the first 2 h)

3996 J. Agric. Food Chem., Vol. 50, No. 14, 2002
Kubo et al.
Table 4. pH Effect of Antifungal Activity of Nonyl Gallate and Sorbic
Acid against S. cerevisiae ATCC 7754
nonyl gallate
MIC MFC
sorbic acid
MIC MFC
400
benzoic acid
MIC MFC
pH
3
5
7
9
6.25
6.25
12.5
25
12.5
12.5
25
1600
3200
−
800
800
>1600
>1600
1600
1600
−
800
>1600
>1600
50
−
−
Table 5. Antifungal Activity of Nonyl and Hexyl Esters of Various
Phenolic Acids against S. cerevisiae ATCC 7754
µg/mL
Figure 3. Effect of geranyl gallate (21) on the growth of S. cerevisiae
ATCC 7754: 0 (b) (control), 12.5 (O), 25 (4), and 50 (0) µg/mL geranyl
gallate. Exponentially growing cells were inoculated into ME broth and
then cultured at 30 °C without shaking.
ester tested
MIC
>400
>400
>400
>400
>400
>400
MFC
nonyl
benzoate
4-hydroxybenzoate
a
−
−
>400
>400
>400
>400
12.5
12.5
3-hydroxybenzoate
3,5-dihydroxybenzoate
3-hydroxy-4-methoxybenzoate
4-hydroxy-3-methoxybenzoate
3,4-dihydroxybenzoate
3,4,5-trihydroxybenzoate
6.25
6.25
hexyl
3,4-dihydroxybenzoate
4-hydroxybenzoate
3,5-dihydroxybenzoate
400
>800
>800
400
−
−
a −, not tested.
to this activity. Therefore, all of the gallate esters can be used
as antioxidants in food. It appears that the potency of the
antifungal activity of the gallates against microorganisms
depends on their alkyl chain length, as described above, but
not their scavenging activity. The amount required for antioxi-
dant activity (IC50) is slightly less compared to the amounts
needed to control microorganisms.
Figure 4. Fungicidal effect of nonyl gallate (1) on the growth stage of S.
cerevisiae ATCC 7754. Yeast cells (105 cells/mL) were inoculated to ME
broth medium. Nonyl gallate (concentrations: 0, b; MFC, O; 2 × MFC,
4; and 4 × MFC, 0) was added after 0 (b), 4 (4), or 24 h (0) of
incubation.
The gallates were designed on the basis of the consideration
of the head and tail concept similar to alkanols (6). Therefore,
their mode of antifungal action can be expected to act as a
surfactant. However, additional functions may need to be
considered for newly synthesized phenolic esters. For example,
the ester group did not exist in the original alkanol structure
and may be related to eliciting the additional activity. As can
be seen in Table 1, nonanol itself exhibits antifungal activity,
so the possibility of S. cereVisiae exuding an esterase that
hydrolyzes nonyl gallate (C9) to the original gallic acid and
antifungal nonanol was first taken into account. This possibility
can be readily ruled out because neither nonyl benzoate (11)
nor nonyl 4-hydroxybenzoate (12) exhibited any antifungal
activity against S. cereVisiae up to 800 µg/mL as shown in Table
5. If these benzoates were hydrolyzed by esterase, the freed
nonanol should show the activity. In connection with the ester
group, this group is essentially not related to the activity.
Both nonyl 3,4,5-trihydroxybenzoate and nonyl 3,4-dihy-
droxybenzoate (13) showed both antifungal and scavenging
activity on the DPPH radical. Their potencies of both activities
are almost comparable. On the other hand, nonyl benzoate, nonyl
3-hydroxybenzoate (14), nonyl 4-hydroxybenzoate, and nonyl
3,5-dihydroxybenzoate (15) exhibited neither antifungal nor
scavenging activity on the DPPH radical, indicating that catechol
or pyrogallol moieties are associated with the activities tested.
Both hexyl 3,5-dihydroxybenzoate (17) and hexyl 4-hydroxy-
benzoate (18) were synthesized and tested for their antifungal
activity against S. cereVisiae. Neither benzoate showed any
activity up to 800 µg/mL. In addition, nonyl 3-hydroxy-4-
when added to the culture at any stage of growth. Remarkably,
nonyl gallate rapidly reduced the number of viable cells even
when it was added to the cultures in the late-stationary growth
phase (108 CFU/mL) as shown in Figure 4.
Among the food preservatives currently used, sorbic acid and
its salts are some of the most commonly used as yeast inhibitors
(13). Hence, the activity of nonyl gallate was compared with
that of sorbic acid. Despite their wide application in various
foods, especially as yeast and mold inhibitors, sorbates are
generally static (14). In our preliminaly screening, sorbic acid
needs 3200 µg/mL to exhibit lethal activity (MFC) against S.
cereVisiae. This lack of potency limits their use, although sorbic
acid is considered to be one of the least harmful preservatives
in use. The fungicidal activity of nonyl gallate (C9), 50 µg/mL,
is 64-fold more potent than that of sorbic acid. More importantly,
nonyl gallate has another superior property. As a weak acid
antifungal agent, the activity of sorbic acid is pH dependent
and increases as the pH of the substrate decreases (13), as shown
in Table 4. At higher pH values (>5), sorbic acid did not show
any antifungal activity up to 1600 µg/mL due to a higher degree
of dissociated molecules. In contrast, nonyl gallate was little
influenced by pH values.
Propyl, octyl, and dodecyl gallates are currently permitted
for use as antioxidant additives in food, so that the gallates
synthesized were tested for their antioxidant activity (15). All
of the gallates tested, regardless of their alkyl chain length,
showed potent scavenging activity on the DPPH radical,
indicating that the alkyl chain length was not directly related

Molecular Design of Antifungal Agents
J. Agric. Food Chem., Vol. 50, No. 14, 2002 3997
volume of the hydrophobic alkyl group are also associated with
its antifungal activity to a large extent. As long as the catechol
moiety of “a” exists, the hydrophobic portion seems to be
flexible. For example, the alcohol portion of alkyl gallate can
be replaced by a bicyclic alcohol, decahydro-2-naphthol (23),
as well as a monoterpene alcohol, geraniol (Table 3). The result
may provide a hint to design more appropriate antimicrobial
agents because branched-chain or ring-containing surfactants
are generally more soluble in aqueous media than straight-chain
materials with the same number of carbon atoms (11). In
addition, the straight-chain surfactants are much more biode-
gradable than the branched-chain or ring-containing materials.
The hydrophobic portion is not a major contributor but obviously
related to the activity.
Figure 5. Adsorption of nonyl (1) and dodecyl (9) gallates to the cells of
S. cerevisiae ATCC 7754. After each gallate was mixed with (9) or (b)
without yeast cells (108 cells/mL), the suspension was vortexed for 5 s.
Absorbance of the supernatant obtained by centrifugation for 2 min was
measured.
Although the current study was focused against S. cereVisiae,
the gallates synthesized showed a broad antimicrobial spectrum.
Their antimicrobial activity is nonspecific (17), and hence
binding sites seem to be nonspecific as surfactants. The results
obtained indicate that microorganisms having different mem-
brane structures showed different susceptibilities to gallates
having different chain lengths. For example, dodecyl gallate did
not inhibit the growth of S. cereVisiae but still shows antibacte-
rial activity against Gram-positive bacteria and scavenging
activity on the DPPH radical, indicating possibly different
inhibition mechanisms between yeasts and bacteria.
methoxybenzoate (19) and nonyl 4-hydroxy-3-methoxybenzoate
(20) did not show any antifungal activity, although both
exhibited scavenging activity on the DPPH radical. On the basis
of the data obtained, it appears that at least a catechol moiety
seems to be essential to elicit the antifungal activity against S.
cereVisiae and that the alkyl chain length also plays an important
role. The results obtained demonstrate that the structural criteria
of antifungal hydroxybenzoates differ from those of aliphatic
alkanols. For instance, alkanols do not have any scavenging
activity, but both 3,4,5-trihydroxybenzoates and 3,4-dihydroxy-
benzoates possess potent antioxidant activity. The antifungal
mechanism of these benzoates seems to be associated with their
specific structural features. The length of the hydrophobic alkyl
group is associated with their antifungal activity against S.
cereVisiae to a large extent but not antioxidant activity.
Further support for the surfactant concept was obtained in
an additional experiment that indicates antifungal nonyl gallate
rapidly adsorbed onto the surface of S. cereVisiae cells but
dodecyl gallate did much less, as shown in Figure 5. The
pyrogallol moiety adsorbed by an intermolecular hydrogen bond
in attaching itself to hydrophilic portion of the membrane
surface. This creates, as a surfactant, disorder in the fluid bilayer
of the membrane. It seems that amphiphilic alkyl gallate unlikely
disrupts specific target proteins such as cell-surface receptors
or signal transduction proteins. Their adsorbing sites are unlikely
specific but need to be clarified. On the other hand, most of the
dodecyl gallate molecules did not adsorb onto the cell surface
and remained in the water-based medium, probably in the form
of insoluble monolayer or spread film (16). This may reveal
why dodecyl gallate did not show any effects on eukaryotic
microorganisms such as S. cereVisiae. The length of the alkyl
chain is not a major contributor but plays an important role in
eliciting the activity.
The aforementioned primary alkanols were recently found
to inhibit the succinate-supported respiration of intact mito-
chondria isolated from rat liver (18). The potency increased with
increasing chain length up to undecanol. Given each alkanol’s
nearly identical effect on state 3 and uncoupled respiration,
action is not directly on ATP synthetase but earlier in the
respiratory process. The data obtained with freeze-thaw
(broken) mitochondria distinguish effects on the mitochondrial
substrate carrier from that on the electron transport chain,
suggesting that inhibition originates from interference with the
dicarboxylate carrier, which must transport succinate across the
mitochondrial membranes. It seems that alkanols inhibit this
plasma membrane transporter and other membrane-bound
proteins as nonionic surfactants (18). Antifungal hydroxyben-
zoates such as nonyl 3,4,5-trihydroxybenzoate and nonyl 3,4-
dihydroxybenzoate can be expected in part to inhibit the same
respiration as nonionic surfactants. However, it cannot be
inferred that membrane damage is the only cause of the lethal
effect of these hydroxybenzoates. It appears that the possibility
of additional mechanisms may need to be taken into consider-
ation. Further studies are currently underway and will be
reported separately.
It should be noted that the hydrolyzable ester group can be
an ideal food additive to prevent undesired side effects,
particularly endocrine-disrupting activity such as alkylphenolic
compounds (19), which are environmentally persistent estrogen
mimics (20). In general, the gallatesssuch as octyl and nonyls
act as multifunctional agents in foodssat least as antifungal
and antioxidant agents. After the gallates are consumed together
with the food to which they are added as additives, these esters
are hydrolyzed to the original gallic acid and the corresponding
alcohols. The former still acts as a potent antioxidant, and the
latter is a common plant component. More specifically, the freed
gallic acid acts as an antioxidant; for instance, it scavenges
superoxide anions generated enzymatically and nonenzymat-
icallyseven more potent than its ester forms. In addition, nonyl
alcohol likely prevents the generation of superoxide anion and
hydrogen peroxide by mitochondria in the resting state, as an
uncoupler similar to fatty acids (2). The rationale of this
DISCUSSION
In view of the present investigation it appears that nonyl
gallate’s antifungal activity is due primarily to its surfactant
property, similar to alkanols (6). Evidence for this postulate can
be provided by the time-kill curve experiment, which showed
that nonyl gallate is fungicidal against S. cereVisiae. It killed S.
cereVisiae cells notably quickly, and no viable cells were
detected after exposure to 50 µg/mL (177 µM) for 2 h. In
addition, this fungicidal activity against S. cereVisiae was
observed at any stage of growth. In addition, the length and

3998 J. Agric. Food Chem., Vol. 50, No. 14, 2002
Kubo et al.
(7) Kubo, I.; Himejima, M. Anethole, a synergist of polygodial
against filamentous microorganisms. J. Agric. Food Chem. 1991,
39, 2290-2292.
preventing mechanism at molecular levels is not yet clear, but
the similar concept of the fatty acids cycle (22, 23) is also very
likely applicable to this alcohol. Gallic acid is found in many
plants such as blackberry bark, henna, tea, and uva ursi. Geraniol
is reported in a large number (>160) of essential oilsssuch as
lemon grass, coriander, lavender, carrot, and geraniumsand used
as food flavor for baked goods, soft and hard candies, gelatin,
pudding, and chewing gum. As mentioned, the three gallatess
propyl, octyl, and dodecylsare currently permitted for use as
antioxidant additives in foods; therefore, the above gallates may
also be safe to use as food preservatives. In addition, it is evident
that the volume of the hydrophobic portion also contributes to
the activity. This suggests that even more optimization is
possible through the synthetic approach. However, the most
pertinent volume to elicit the maximum activity needs to be
established. Antioxidant, gallic acid, and a glutathione S-
transferase inducer, geraniol, may contribute to reduce cancer
risk as well as oxidative damage related diseases. The complete
scavenging activity of gallic acid was observed in the range of
10-30 µM, which is much less than the amounts needed to
control S. cereVisiae.
(8) Blois, M. S. Antioxidant determination by the use of a stable
free radical. Nature 1958, 181, 1199-1200.
(9) Kubo, I.; Muroi, H.; Himejima, M.; Yamagiwa, Y.; Mera, H.;
Tokushima, K.; Ohta, S.; Kamikawa, T. Structure-antibacterial
activity relationships of anacardic acids. J. Agric. Food Chem.
1993, 41, 1016-1019.
(10) Aruoma, O. I.; Murcia, A.; Butler, J.; Halliwell, B. Evaluation
of the antioxidant and prooxidant actions of gallic acid and its
derivatives. J. Agric. Food Chem. 1993, 41, 1880-1885.
(11) Rosen, M. J. Surfactants and Interfacial Phenomena, 2nd ed.;
Wiley-Interscience: New York, 1989; pp 1-32.
(12) Zheng, G. Q.; Kenney, P. M.; Lam, L. K. Potential anticarci-
nogenic natural products isolated from lemongrass oil and
galanga root oil. J. Agric. Food Chem. 1993, 41, 153-156.
(13) Sofos, J. N.; Busta, F. F. Sorbates. In Antimicrobials in Food;
Branen, A. L., Davidson, P. M., Eds.; Dekker: New York, 1983;
pp 141-175.
(14) Robach, M. C.; Sofos, J. N. Use of sorbates in meat products,
fresh poultry and poultry products: A review. J. Food Prot. 1982,
45, 374-383.
Compared to sorbic acid, nonyl gallate is odorless, as well
as colorless and tasteless, with superior antifungal activity, and
this antifungal activity is not influenced by the pH values in
contrast to sorbic acid. In addition to this fungicidal activity,
the gallates have potent antioxidant activity, which is also an
important property to protect food, but sorbic acid does not
possess this activity. Therefore, nonyl gallate should have a
wider application potential as a food additive, particularly to
control S. cereVisiae (1). It should be noted that this antifungal
agentsas a nonionic surfactantstargets the extracytoplasmic
region and thus does not need to enter the cell, thereby avoiding
most cellular pump-based resistance mechanisms.
(15) Kubo, I. Molecular design of antioxidative and antimicrobial
agents. CHEMTECH 1999, 29, 37-42.
(16) Jones, M. N.; Chapman, D. Micelles, Monolayers, and Biomem-
branes; Wiley-Liss: New York; 1994.
(17) Kubo, I.; Xiao, P.; Fujita, K. Antifungal activity of octyl
gallate: Structural criteria and mode of action. Bioorg. Med.
Chem. Lett. 2001, 11, 347-350.
(18) Hammond, D. G.; Kubo, I. Alkanols inhibit respiration of intact
mitochondria and display cutoff similar to that measured in ViVo.
J. Pharm. Exp. Ther. 2000, 293, 822-828.
(19) Soto, A. M.; Juticia, H.; Wray, J. W.; Sonnenschein, C. p-Nonyl-
phenol: an estrogenic xenobiotic released from “modified”
polystyrene. EnViron. Health Perspect. 1991, 92, 167-173.
(20) White, R.; Jobling, S.; Hoare, S. A.; Sumpter, J. P.; Parker, M.
G. Environmentally persistent alkylphenolic compounds are
estrogenic. Endocrinology 1994, 135, 175-182.
(21) Korshunov, S. S.; Korkina, O. V.; Ruuge, E. K.; Skulachev, V.
P.; Starkov, A. A. Fatty acids as natural uncouplers preventing
generation of O2•- and H2O2 by mitochondria in the resting state.
FEBS Lett. 1998, 435, 215-218.
ACKNOWLEDGMENT
We are grateful to Dr. N. Masuoka for performing the DPPH
antioxidation assay and R. J. Lee for obtaining preliminary
antimicrobial activity data of some alkyl gallates at an earlier
stage of the work.
(22) Jezek, P.; Engstova, H.; Zackova, M.; Vercesi, A. E.; Costa, A.
D. T.; Arruda, P.; Garlid, K. Fatty acid cycling mechanism and
mitochondrial uncoupling proteins. Biochim. Biophys. Acta 1998,
1365, 319-327.
LITERATURE CITED
(1) Fleet, G. Spoilage yeasts. Crit. ReV. Biotechnol. 1992, 12, 1-44.
(2) Frazier, W. C.; Westhoff, D. C. Food Microbiology, 4th ed.;
McGraw-Hill: New York, 1988; pp 410-439.
(3) Kubo, I.; Himejima, M. Potentiation of antifungal activity of
sesquiterpene dialdehydes against Candida albicans and two
other fungi. Experientia 1992, 48, 1162-1164.
(4) Kubo, I.; Lee, S. H. Potentiation of antifungal activity of sorbic
acid. J. Agric. Food Chem. 1998, 46, 4052-4055.
(5) Spratt, B. G. Resistance to antibiotics mediated by target
alterations. Science 1994, 264, 388-393.
(23) Skulachev, V. P. Uncoupling: new approaches to an old problem
of bioenergertics. Biochim. Biophys. Acta 1998, 1363, 100-124.
Received for review January 25, 2002. Revised manuscript received
April 16, 2002. Accepted April 17, 2002. This work was presented in
part at the Symposium of Chemistry of Antimicrobials in the Division
of Agricultural and Food Chemistry at the 220th National Meeting of
the American Chemical Society in Washington, DC. K.F. thanks Osaka
City University for financial support during his study at UCB.
(6) Kubo, I.; Muroi, H.; Kubo, A. Structural functions of antimi-
crobial long-chain alcohols and phenolics. Bioorg. Med. Chem.
1995, 3, 873-880.
JF020088V