Synthesis and biological characterization of (Z)-9-heptadecenoic and (Z)-6-methyl-9-heptadecenoic acids: Fatty acids with antibiotic activity produced by Pseudozyma flocculosa

Article abstract of DOI:10.1023/A:1005464326573

Difficulties in isolating and purifying antibiotic fatty acids from culture filtrates of Pseudozyma flocculosa, a biocontrol agent against powdery mildew, have been limiting factors in studying the properties and understanding the mode of action of the biocontrol agent. We report a new protocol for synthesizing (Z)-9-heptadecenoic and for the first time synthesis of (Z)-6-methyl-9-heptadecenoic acids, two antibiotic fatty acids produced by P. flocculosa. This allowed reproducible and quantifiable means of assaying biological activity of the molecules. In these bioassays, both molecules exhibited antifungal activity corresponding to their expected potency. These new developments should facilitate further studies aimed at deciphering the ecological properties of P. flocculosa.

Full text of DOI:10.1023/A:1005464326573

Journal of Chemical Ecology, Vol. 26, No. 4, 2000
SYNTHESIS AND BIOLOGICAL CHARACTERIZATION
OF (Z)-9-HEPTADECENOIC AND
(Z)-6-METHYL-9-HEPTADECENOIC ACIDS:
FATTY ACIDS WITH ANTIBIOTIC ACTIVITY
PRODUCED BY Pseudozyma flocculosa
´
T. J. AVIS, R. R. BOULANGER, and R. R. BELANGER*
De´partement de phytologie
Faculte
´
des sciences de l’agriculture et de l’alimentation
Centre de recherche en horticulture
Universite
´ Laval
Ste-Foy, Quebec, Canada G1K 7P4
´
(Received January 20, 1999; accepted December 7, 1999)
Abstract—Difficulties in isolating and purifying antibiotic fatty acids
from culture filtrates of Pseudozyma flocculosa, a biocontrol agent against
powdery mildew, have been limiting factors in studying the properties and
understanding the mode of action of the biocontrol agent. We report a new
protocol for synthesizing (Z)-9-heptadecenoic and for the first time synthesis
of (Z)-6-methyl-9-heptadecenoic acids, two antibiotic fatty acids produced by
P. flocculosa. This allowed reproducible and quantifiable means of assaying
biological activity of the molecules. In these bioassays, both molecules
exhibited antifungal activity corresponding to their expected potency. These
new developments should facilitate further studies aimed at deciphering the
ecological properties of P. flocculosa.
Key Words—Fatty acids, Pseudozyma flocculosa, powdery mildew, antibiosis,
biocontrol.
INTRODUCTION
Pseudozyma flocculosa (Traquair, L. A. Shaw & Jarvis) Boekhout & Traquair
(syn.: Sporothrix flocculosa Traquair, Shaw & Jarvis) (Boekhout, 1995) is a
yeastlike fungus with antagonistic activity against powdery mildew fungi (Jarvis
*To whom correspondence should be addressed.
987
0098-0331 00 0400-0987$18.00 0  2000 Plenum Publishing Corporation
/
/
/

´
AVIS, BOULANGER, AND BELANGER
988
et al., 1989; Hajlaoui and Belanger, 1991, 1993; Belanger et al., 1994). Micro-
´ ´
scopic and cytochemical observations revealed that the antagonist induces rapid
collapse of powdery mildew conidial chains and cytoplasmic disintegration of
fungal cells without hyphal penetration (Hajlaoui et al., 1992). These findings
suggested that antibiosis was the main mode of action by which P. flocculosa
exerted its biocontrol activity.
Recently, our laboratory isolated extracellular fatty acids with antimicrobial
properties produced by P. flocculosa, two of which were reported for the first time
(Benyagoub et al., 1996a). These molecules, identified as 9-heptadecenoic acid and
6-methyl-9-heptadecenoic acid, exhibited toxic activity against several fungi, but
had little effect on P. flocculosa itself (Benyagoub et al., 1996b). Belanger and Dea-
´
con (1996) found a correlation between fungal sensitivity to the fatty acids and fun-
gal susceptibility to P. flocculosa, suggesting that these fatty acids are the active
principles behind the biocontrol potential of the antagonist.
Isolation and manipulation of these fatty acids from culture filtrates has
proven difficult as extraction and purification procedures resulted in low and
unreliable yields. This has hampered our efforts to learn more about the eco-
logical and molecular properties of these compounds. As a result, we sought a
way to obtain large and stable amounts of the antibiotics. Therefore the objec-
tive of this work was to develop a synthesis method for both (Z)-9-heptadecenoic
and (Z)-6-methyl-9-heptadecenoic acids in order to facilitate biological studies
of the antibiotics.
METHODS AND MATERIALS
General Chemical Procedures and Instruments
All oxygen- or moisture-sensitive reactions were performed in flame- or
oven-dried glassware under pressure of nitrogen or argon. Air-sensitive liquids
and solutions were transferred by syringe or cannula. Ether, tetrahydrofuran, and
1,4-dioxane were dried before use. 1,7-Dibromoheptane, 3-methylglutaric anhy-
dride, and 1-nonyne were obtained from commercial sources (Sigma-Aldrich,
Mississauga, Ontario, Canada). Flash column chromatography was performed
with silica gel EM science 200-400 mesh (40–63 mm). All nuclear magnetic
resonance spectroscopy (1H, 13C) was recorded in deuterium chloroform with a
Bruker 300 MHz. Peaks positions are expressed as downfield shifts in parts per
million (d) from tetramethylsilane internal standards. Split patterns are designed
as singlet (s); doublet (d); doublet–triplet (dt); triplet (t); quartet (q); multiplet
(m); broad signal (br); J is coupling in Hertz. GC-MS data were recorded with
a Hewlett Packard 5890 series II gas chromatography coupled to a Hewlett
Packard 5972 mass selective detector. Mass spectra were obtained with electron
impact (EI) (70 eV, energy beam).

ANTIBIOTIC FATTY ACIDS IN Pseudozyma
989
FIG. 1. Synthetic scheme for the synthesis of (Z)-9-heptadecenoic acid.
Synthesis of (Z)-9-heptadecenoic Acid
Preparation of various homologs and positional isomers of the C-13–C-19
odd-carbon ethenoic acids were synthesized according to Grimmer and Kracht
(1963) and Jacob and Grimmer (1966) (Figure 1). These acids were prepared by
reaction of the appropriate lithium acetylide and dihaloalkane in 1,4-dioxane and
followed by reaction of the resulting acetylenic halide with sodium cynanide to
obtain the corresponding acid 7, which was half-hydrogenated to the ethenoic
according to Jacob and Grimmer (1966).
1,7-Diiodoheptane (3). A solution of 1,7-dibromoheptane 1 (5.17 g, 0.022
mol) in 20 ml of dry acetone was added dropwise to a 250-ml flask equipped with
a condenser, containing a suspension of sodium iodide (13.3 g, 0.09 mol) with
100 ml of dry acetone at room temperature. The mixture was heated to reflux
for 3 hr. After cooling, the reaction mixture was filtered and the solvent was
evaporated in vacuo. The crude residue was purified by flash chromatography
1
with hexane as eluent. The yield was 96%, 7.6 g. C7H14I2: H NMR (CDCl3:
d 1.30 (m, 6H), 1.39 (m, 4H), 3.17 (t, 4H) ppm: 13C NMR (CDCl3): d 6.91
′
(C-4), 27.32 (C-3), 27.81 (C-3 ), 30.14 (2 carbons) (C-2), 33.21 (2 carbons)
(C-1); EI-MS (m z, loss fragment): 352 (M+), 225 (M+-HI), 183 (M+-HI-C3H6),
/
155 (M+-C5H11I), 97 (C7H13).
9-Hexadecyne Iodide (4). Under an atmosphere of argon, 1-nonyne 2 (2.5
g, 0.02 mole) in 500 ml of dry 1,4-dioxane was refluxed with lithium amide
(0.414 g, 0.018 mole) for 3.5 hr in a tricol equipped with a condenser and a
dropping funnel with pressure equalization arm. To the organolithium compound

´
AVIS, BOULANGER, AND BELANGER
990
thus formed, 1,7-diiodoheptane 3 (8 g, 0.022 mol) in 50 ml of dry 1,4-dioxane
was added dropwise. The mixture was refluxed an additional 2 hr, allowed to
cool to room temperature, and then poured into a mixture of ice–water and stirred
15 min. The mixture was extracted with diethylether (3 × 50 ml). The combined
organic fractions were dried over magnesium sulfate, filtered, and the solvent
was evaporated in vacuo. The residue was purified by flash chromatography with
hexane as eluent to give the corresponding iodoalkyne 4 with a 95% yield (6.6
g). C16H29I: 1H NMR (CDCl3): d 0.87(t, 3H), 1.27–1.54 (m, 18H), 1.79–1.92 (m,
2H), 2.13 (t, 4H), 3.18 (t, 2H) ppm; 13C NMR (CDCl3): d 7.11 (CH3CH2), 14.11,
18.72, 18.78, 22.65, 28.07, 28.58, 28.85 (2 carbons), 29.00, 29.19, 30.42, 31.79,
33.52, 80.00 (C, alkyne (80.42 (C, alkyne) ppm; EI-MS (m z, loss fragment):
/
348 (M+), 305 (M+-C3H9), 291 (M+-C4H9), 263 (M+-C6H13), 249 (M3-C7H15).
9-Heptadecyne Cyanide (5). A solution of iodide compound 4 (4 g, 0.0115
mol) with 200 ml of 92% ethanol was added to a 250-ml flask equipped with a
condenser and was refluxed 5 hr with sodium cyanide (2.8 g, 0.06 mol). After
cooling to room temperature the mixture was poured into 50 ml water and
extracted with hexane (3 × 50 ml). The combined organic fraction was dried
over magnesium sulfate, filtered, and the solvent was evaporated in vacuo. The
residue was purified by flash chromatography with hexane as eluent. The yield
1
was 2.5 g (88%). C17H29N: H NMR (CDCl3): d 0.86 (t, 3H), 1.26–1.49 (m,
18H), 1.59–1.69 (m, 2H), 2.11 (t, 4H), 2.31 (t, 2H) ppm; 13C NMR (CDCl3): d
13.92 (CH3CH2), 16.95, 18.52, 18.59, 22.97, 25.19, 28.14, 28.27 (two carbons),
28.42, 28.68, 28.77, 29.01, 31.63, 79.71 (C, alkyne), 80.31 (C, alkyne) 119.59
(C, nitrile) ppm; EI-MS (m z, loss fragment): 247 (M+), 176 (M+-C5H11), 162
/
M+-C6H13), 148 (M+-C7H15), 134 (M+-C8H17), 81 (C5H7N), 67 (C4H5N).
9-Heptadecynoic Acid (6). In a 250-ml flask equipped with a condenser,
the cyanide 5 (2.5 g, 0.0094 mol) was diluted with 50 ml of 92% ethanol.
Sodium hydroxide (2.28 g, 0.06 mol) in 25 ml of water was added in one por-
tion, and the mixture was heated to reflux for 10 hr. The reaction was cooled to
room temperature, diluted with 200 ml of water, acidified with hydrochloric acid
2 N, and extracted with diethylether (3 × 50 ml). The crude residue 3.8 g was
purified by flash chromatography with a gradient of 10–30% ethylacetate–hexane
1
as eluent to yield 2 g (66%) of 9-heptadecynoic acid 6. C17H30O2: H NMR
(CDCl3): d 0.81 (t, 3H), 1.21–1.40 (m, 18H), 1.52–1.60 (m, 2H), 1.98–2.09 (t,
4H), 2.26–2.31 (t, 2H) 8.75 (br, COOH, 1H) ppm; 13C NMR (CDCl3): d 13.93
(CH3CH2), 18.58, 22.48, 24.48, 28.46, 28.62 (two carbons), 28.68 (two carbons),
28.81, 28.92, 29.03, 31.63, 33.81, 79.92 (C, alkyne), 80.22 (C, alkyne) 179.46
(COOH) ppm; EI-MS (m z, loss fragment): 248 (M+), 195 (M+-C5H11), 178
/
(M+-C4H8O2), 164 (M+-C5H10O2), 128 (M+-C7H12O2), 95 (C7H11), 81 (C6H8),
67 (C5H7).
(Z)-9-heptadecenoic Acid (7). 9-Heptadecynoic acid 6 (1.35 g, 0.0051 mol)
was dissolved with 50 ml of methanol pyridine (5 : 1), poured into a hydrogena-
/

ANTIBIOTIC FATTY ACIDS IN Pseudozyma
991
tion tube, and 0.03 g of Raney nickel powder was added to the mixture. The tube
was closed and shaken under hydrogen atmosphere until hydrogen adsorption
ceased (30 min). The reaction mixture was filtered through celite, the solvent
was evaporated in vacuo, and the residue was purified by flash chromatography
with a gradient of 10–30% ethylacetate–hexane as eluent. Pure (Z)-9-heptade-
1
cenoic acid 7 was obtained. The yield was 1.1 g (81%). C17H32O2: H NMR
(CDCl3: d 0.81 (t, 3H), 1.18–1.24 (m, 19H), 1.51–1.59 (m, 2H), 1.91–2.07 (t,
4H), 2.25–2.30 (t, 2H) 5.22–5.33 (m, 2H, J 5.80 Hz) ppm; 13C NMR (CDCl3):
d 13.99 (CH3CH2), 22.56, 24.54, 27.09, 28.94 (two carbons), 29.02, 29.15 (two
carbons), 29.56 (two carbons), 29.65, 31.76, 33.96, 129.61 (C, alkene), 129.90
(C, alkene) 180.20 (COOH) ppm; EI-MS (m z, loss fragment): 268 (M+), 250
/
(M+-H2O), 196 (M+-C3H4O2), 182 (M+-C4H6O2), 138 (C7H14O2), 95 (C7H11),
81 (C6H8), 67 (C5H7).
Synthesis of (Z)-6-Methyl-9-Heptadecenoic Acid
Heptadecenoic acid methyl substituted at the 6 position was synthesized
by a modification of Figure 1. The corresponding a, q-dibromopentane was
prepared by reduction of methyl-3-glutaric anhydride followed by bromation
with phosphorus tribromide or a Vonbraun degradation of substituted piperidine
(Nguyen and Cartledge, 1986) (Figure 2).
3-Methyl-1,5-Pentanediol (10). A solution of 3-methylglutaric anhydride
9 (1.48 g, 0.008 mol) in 50 ml of anhydrous tetrahydrofuran was added drop-
wise to a suspension of lithium aluminum hydride (0.59 g, 0.96 mol) in 200
ml of anhydrous tetrahydrofuran. After 2 hr of stirring at room temperature,
hydrolysis was accomplished by pouring the reaction mixture in an ice-cold sat-
urated ammonium chloride solution. The mixture was extracted with diethylether
(3 × 50 ml). Drying of the combined organic phase was performed over mag-
nesium sulfate, and evaporation of solvent gave pure 3-methyl-1,5-pentanediol
1
10, 0.602 g (44% yield). C6H14O2: H NMR (CDCl3): d 0.92 (d, 3H, J 6.63
Hz), 1.36–1.45 (dt, 2H), 1.53–1.64 (dt, 2H), 1.72–1.83 (m, 1H), 3.39 (br, 2H),
3.59–3.74 (m, 4H) ppm; 13C NMR (CDCl3): d 19.71, 25.69, 39.24 (two car-
bons), 60.18 (two carbons) ppm; EI-MS (m z, loss (fragment): 119 (M++1), 99
/
(M+-H3O), 82 (C6H10), 67 (C5H7).
3-Methyl-1,5-dibromopentane (11). Phosphorus tribromide, 1.41 g (0.0052
mol) was slowly added to a 250-ml flask equipped with a condenser, containing
0.6 g (0.0051 mol) of 3-methyl-1,5-pentanediol 9, at 10 C. The mixture was
brought to room temperature and stirred 20 hr. Ethylacetate (20 ml) was added
to precipitate the oxides, followed by filtration and evaporation of solvent in
vacuo. The crude residue was purified by flash chromatography with hexane as
eluent or by distillation (bp 8 C 5 mm Hg), to give pure dibromide 11. The
/
1
yield was 0.9 g (73%) C6H12Br2: H NMR (CDCl3): d 0.93 (d, 3H, J 6.26

´
AVIS, BOULANGER, AND BELANGER
992
FIG. 2. Synthetic scheme for the synthesis of (Z)-6-methyl-9-heptadecenoic acid.
Hz), 1.65–1.77 (m, 2H), 1.82–1.97 (m, 3H), 3.36–3.49 (m, 4H) ppm; 13C NMR
(CDCl3): d 18.05, 30.47, 31.12 (two carbons), 39.28 (two carbons) ppm; EI-
MS (m z, loss fragment): 244 (M+, 79Br), 246 (M+, 81Br), 162 (M+-81Br), 164
/
(M+-79Br), 135 (M+-C2H481Br), 137 (M+-C2H479Br), 83 (C6H11).
3-Methyl-1,5-diiodopentane (12). A solution of 3-methyl-1,5-dibromopen-
tane 11 (1.15 g, 0.045 mol) in 20 ml of dry acetone was added dropwise to a
250-ml flask equipped with a condenser and containing a suspension of sodium
iodide (1.8 g, 0.012 mol) with 25 ml of dry acetone at room temperature. The
mixture was heated to reflux for 3 hr. After cooling, the reaction mixture was

ANTIBIOTIC FATTY ACIDS IN Pseudozyma
993
filtered and the solvent was evaporated in vacuo. The crude residue was purified
by flash chromatography with hexane to give pure diiodide 12. The yield was
1.65 g (89%). C6H12I2: 1H NMR (CDCl3): d 0.90 (d, 3H, J 5.32 Hz), 1.71 (m,
3H), 1.90 (m, 2H), 3.30 (m, 4H) ppm; 13C NMR (CDCl3): d 3.87, 17.69, 34.91,
40.00 (two carbons) ppm; EI-MS (m z, loss fragment): 338 (M+), 211 (M+-I),
/
169 (M+-C3H6I), 155 (M+-C4H8I), 127 (M+-C6H12I), 83 (C6H11), 55 (C4H7).
3-Methyl-6-tetradecyne Bromide (13) or 3-Methyl-6-tetradecyne Iodide
(14). Under an atmosphere of argon, 1-nonyne 2 (3.72 g, 0.03 mol) in 100 ml
of dry 1,4-dioxane was refluxed with lithium amide (0.762 g, 0.033 mol) for 3.5
hr in a tricol equipped with a condenser and a dropping funnel with pressure
equalization arm. To the organolithium compound thus formed, 0.0047 mol of
the 3-methyl-1,5-halopentane 11 (1.14 g) or 12 (1.59 g) in 50 ml of dry 1,4-diox-
ane was added dropwise, and the mixture was heated to reflux another 2 hr. The
mixture was allowed to cool to room temperature and poured into a mixture of
ice–water and stirred 15 min. The mixture was extracted with diethylether (3 ×
50 ml). The combined organic fractions were dried over magnesium sulfate, fil-
tered, and the solvent was evaporated in vacuo. The residue was purified by flash
chromatography with light petroleum ether as eluent to give the corresponding
1
haloalkyne 13 (0.897 g, 76% yield) or 14 (1.00 g, 64% yield). C15H27Br: H
NMR (CDCl3): d 0.89 (d, J 6.47 Hz, 3H), 0.92 (t, 3H), 1.10–1.49 (m, 11H),
1.52–1.94 (m, 4H), 2.10–2.27 (m, 4H), 3.47 (m, 2H, CH2-Br) ppm; 13C NMR
(CDCl3): d 13.94 (CH3CH2), 16.22, 18.37, 18.61, 22.49, 28.70, 28.99 (2 car-
bons), 30.49, 31.56, 31.64, 35.57, 39.57, 79.46 (C, alkyne), 80.45 (C, alkyne)
ppm; EI-MS (m z, loss fragment): 288 (M+), 207 (M+-81Br), 189 (C8H1281Br),
/
187 (C8H1279Br), 125 (C9H17), 95 (C7H11), 81 (C6H9), 67 (C5H7).
C15H27I: 1H NMR (CDCl3): d 0.89 (d, J 6.10 Hz, 3H), 0.92 (t, 3H), 1.25
(m, 14H), 1.55 (m, 2H), 1.63–1.80 (m, 1H), 1.83–2.20 (m, 2H), 3.20 (t, 2H,
CH2-I) ppm; 13C NMR (CDCl3): d 4.53 (CH3CH2), 13.97, 16.22, 18.18, 18.63,
22.51, 28.71 (two carbons), 29.00, 31.65, 32.98, 35.32, 40.50, 79.48 (C, alkyne),
80.58 (C, alkyne) ppm; EI-MS (m z, loss fragment): 334 (M+), 207 (M+-I), 155
/
(M+-C4H8I), 123 (C6H12I), 55 (C4H7), 41 (C3H5).
5-Methyl-8-hexadecyn-1-ol (5). Under an atmosphere of argon, 0.087 g
(3.6 mmol) of magnesium turnings were introduced into a three-necked 100-
ml round-bottomed flask equipped with a pressure-equalizing addition funnel, a
condenser, and a magnetic stir bar. The magnesium was covered by 10 ml of
anhydrous diethylether (freshly distilled on sodium). A solution of 3-methyl-6-
tetradecyn iodide 14 (0.800 g, 0.0024 mol) in 20 ml of anhydrous tetrahydrofuran
was added dropwise at 20 C. The formation of the Grignard reagent was accom-
plished after 2 hr of stirring. The solution was then cooled to 0 C in an ice water
bath and a flow of ethylene oxide was bubbled into the Grignard solution for a
period of 45 min. The resulting suspension was stirred for an additional 45 min
at 0 C. The reaction was allowed to warm at room temperature and treated with

´
AVIS, BOULANGER, AND BELANGER
994
cold saturated ammonium chloride. The resulting mixture was extracted with
diethylether, and the combined ether extract was washed with brine and dried
over anhydrous magnesium sulfate. Removal of the solvent by distillation under
pressure afforded of the desired product. The alcohol was purified by flash chro-
matography and eluted with light petroleum ether–ethylacetate (70 : 30). Pure
hydroxyalcyne 14 was obtained (0.100 g, 13%). C17H32O: 1H NMR (CDCl3): d
0.89 (t, 3H), 0.92 (d, J 6.57 Hz, 3H), 1.11–1.75 (m, 20H), 2.09–2.27 (m, 4H),
3.62–3.76 (m, 2H, CH2OH) ppm; 13C NMR (CDCl3): d 13.93, 16.32, 18.60,
19.00, 19.21, 22.95, 28.62 (two carbons), 28.69, 29.01, 29.55, 31.63, 36.26,
39.39, 60.91, 79.81 (C, alkyne), 80.26 (C, alkyne) ppm; EI-MS (m z, loss frag-
/
ment): 223 (M+-C2H5), 209 (M+-C3H8), 193 (M+-C3H7O), 179 (C4H9O), 95
(C7H11), 80 (C6H8), 67 (C5H7).
5-Methyl-8-hexadecyne Bromide (16). Freshly recrystallized from hexane,
triphenylphosphine (0.155 mg, 0.00059 mol) was added in one portion to a
cold solution of hydroxyalcyne 15 (0.135 g, 0.00053 mol) with 5 ml of dry
dichloromethane. After a 1-hr reaction period, carbon tetrabromide, 0.197 mg
(0.00059 mol) was added in one portion and the mixture was stirred at 0 C for
5 hr. A mixture of ethylacetate and petroleum ether (20 : 80) was added, and the
mixture was filtered on a bed of silica gel and washed with the same solvent.
After evaporation of the solvent in vacuo, the bromoalkyne 16 was purified by
flash chromatography with light petroleum ether as eluent to give 0.138 mg, 82%
1
yield. C17H31Br: H NMR (CDCl3): d 0.87 (t, 3H), 0.98 (d, J 6.40 Hz, 3H),
1.10–1.96 (m, 19H), 2.11–2.25 (m, 4H), 3.36–3.51 (m, 2H, CH2Br) ppm; 13C
NMR (CDCl3): d 13.94, 16.22, 18.38 (two carbons), 18.61, 22.49, 28.71, 28.99
(three carbons), 30.79, 31.60, 31.64, 35.57, 39.57, 79.48 (C, alkyne), 80.48 (C,
alkyne) ppm; EI-MS (m z, loss fragment): 273 (M+-C3H6), 235 (M+-79Br), 179
/
(M+-C4H7Br), 95 (C7H11), 81 (C6H9).
6-Methyl-9-heptadecynoic Acid (17). A solution of bromide compound 16
(0.120 g, 0.00038 mol) in 20 ml of 92% ethanol was added to a 50-ml flask
equipped with a condenser and was heated to reflux for 5 hr with sodium cyanide
(0.093 g, 0.00191 mol). The cyanide formed was not isolated and was immedi-
ately treated with sodium hydroxide (0.0076 g, 0.00199 mol) in 25 ml of water.
The mixture was heated to reflux for 14 hr. The reaction was cooled to room
temperature, diluted with 200 ml of water, acidified with 2 N hydrochloric acid
and extracted with diethylether (3 × 50 ml). The crude residue was purified
by flash chromatography with a gradient mixture of 10–30% ethylacetate–light
petroleum ether as eluent to give 0.060 g, 57% yield of 6-methyl-9-heptade-
1
cynoic acid 17. C18H32O2: H NMR (CDCl3): d 0.79 (t, 3H), 0.82 (d, J 6.47
Hz), 1.19–1.68 (m, 19H), 2.03–2.17 (t, 4H), 2.26–2.39 (t, 2H) 9.75 (br, COOH,
1H) ppm; 13C NMR (CDCl3): d 13.95 (CH3CH2), 18.59, 18.59, 18.63, 22.49,
28.69 (three carbons), 28.99, 29.57, 31.11 (two carbons), 31.37, 31.64, 35.70,
79.62 (C, alkyne), 80.31 (C, alkyne) 180.25 (COOH) ppm; EI-MS (m z, loss
/

ANTIBIOTIC FATTY ACIDS IN Pseudozyma
995
fragment): 281 (M++1), 236 (M+-CO2), 169 (M+-C6H8O2), 108 (M+-C8H12),
95 (C7H11), 81 (C6H9), 67 (C5H7).
(Z)-6-Methyl-9-heptadecenoic Acid (18). 6-methyl-9-heptadecynoic acid
17 (0.127 g, 0.00045 mol) was dissolved in 50 ml of methanol–pyridine (5 : 1),
poured into a hydrogenation tube, and 0.03 g of Raney nickel powder was added
to the mixture. The tube was closed and shaken under hydrogen atmosphere until
adsorption ceased (15 min). The reaction mixture was filtered through celite, the
solvent was evaporated in vacuo, and the residue was purified by flash chro-
matography with a gradient of 10–30% ethylacetate–light petroleum ether as
eluent. Pure (Z)-6-methyl-9-heptadecenoic acid 18 was obtained (0.117 g, 91%
1
yield). C18H34O2: H NMR (CDCl3): d 0.86 (t, 3H), 0.89 (d, J 6.19 Hz, 3H),
1.12–1.76 (m, 17H), 1.99–2.18 (t, 4H), 2.27–2.45 (m, 2H) 5.28–5.40 (m, 2H,
J
5.80 Hz) 9.81 (CH2COOH) ppm; 13C NMR (CDCl3): d 13.94 (CH3CH2),
18.96, 19.11, 22.53, 24.48, 27.06, 29.08, 29.13, 29.22, 29.56, 29.61, 31.45,
31.73, 31.87, 36.51, 129.39 (C, alkene), 129.97 (C, alkene), 180.08 (COOH)
ppm; EI-MS (m z, loss fragment): 281 (M+-1), 267 (M+-CH3), 254 (M+-C2H4),
/
238 (M+-CO2), 236 (M+-H2O, -C2H4), 221 (M+-C2H5O2), 114 (C6H10O2), 97
(C7H13), 83 (C6H11).
Biological Activity of Synthesized Antibiotics on Fungal Material. Cla-
dosporium cucumerinum Ellis and Arth. was obtained from Biosystemics
Research Centre (Ottawa, Ontario, Canada) and maintained on potato dextrose
agar (PDA). Sphaerotheca fuliginea (Schlechtend.: Fr.) Pollacci was maintained
in axenic conditions on long English cucumber (Cucumis sativus L. cv. Corona).
Two agar disks (15 mm) of C. cucumerinum were suspended in 100 ml of
potato dextrose broth (PDB). The antibiotics were tested at concentrations of
0.1, 0.2, 0.4, 0.6, 0.8, and 1 mg ml. The fatty acids were solubilized in N,N-
/
dimethylformamide (DMF) prior to being added to the medium. DMF alone
(0.5%) was added to the culture media to serve as control. Growth was quanti-
fied after a three-day culture at 25 C on a rotary shaker (150 rpm) by dry weight
measurement following lyophilization. Further concentrations of each antibiotic
were tested to determine minimum inhibitory concentration (MIC) of C. cu-
cumerinum. MIC was defined as the lowest concentration of each antibiotic at
which no macroscopic evidence of growth was observed. For each concentration,
the experiment was repeated three times. In order to determine the dose of each
antibiotic which reduced growth of C. cucumerinum by 50%, probit analysis was
performed with the PROC Probit procedure in SAS System (SAS Institute Inc.,
Cary, North Carolina).
Foliar disks (5 cm) were cut from S. fuliginea-infected cucumber leaves and
placed on 20-20-20 agar containing 2 g liter 20-20-20 and 8 g liter bacto-agar.
/
/
Disks were cut from the leaf portion covered by at least 95% with colonies of S.
fuliginea. The antibiotics were sprayed on leaf disks as aqueous solutions of 0.1
and 0.2 mg ml. The antibiotics were solubilized in DMF prior to adding to ster-
/

´
AVIS, BOULANGER, AND BELANGER
996
ile water. DMF alone (0.5%) was added to the sterile water to serve as control.
Evaluation of antibiotic effects on S. fuliginea was determined following incu-
bation periods of 0, 12, and 24 hr at 25 C. Further evaluation of the antibiotics
was not possible because of their rapid degradation under these conditions. Anti-
fungal activity was determined using an arbitrary scale of 0 to 4, where 0 no
effect on S. fuliginea conidial chains, 1 1 to 25% collapse of conidial chains,
2
26–50%, 3 51–75%, and 4 76–100% collapse. For each treatment, the
experiment was repeated three times with two replicates per experiment.
RESULTS AND DISCUSSION
By following the synthetic route of Figure 1, it was possible to obtain
pure (Z)-9-heptadecenoic acid. Ames and Bowman (1951) and Broughton et al.
(1952) described the synthesis of ethylenic acids starting with an acid chloride
and tribenzyltricarboxylate, which can also lead to synthesis of this fatty acid.
However, that synthesis was not explored because the protocol in this study was
more efficient and reduced the number of steps. Figure 2 gave pure (Z)-6-methyl-
9-heptadecenoic acid. Global yield of a, q-dibromopentane was low when pre-
pared by reduction of 3-methylglutaric anhydride followed by a bromation with
phosphorus tribomide. The Vonbraun degradation of substituted piperidine gave
better results. In order to obtain the organolithium reagent, Branda
¨
nge et al.
(1984) proposed a new technique by using lithium 4,4 -di-tert-butylbiphenylide
′
(LDBB) as an intermediate reagent at 60 C in tetrahydrofuran. However, our
attempts to use this synthetic route were not conclusive (Figure 2). On the other
hand, the yield of 5-methyl-8-hexadecyn-1-ol (15), although not optimal, was
obtained reproducibly by the route in Figure 2. For large-scale endeavors, this
step could possibly be optimized with a modification of the halogenure struc-
ture such as hydrogenation of the haloalcyne to haloalcene prior to the Grignard
reaction.
Biological testing of the synthesized fatty acids was performed to verify
their toxic activity on fungi and to determine the specific characteristics of each
molecule. Both compounds were assayed against C. cucumerinum, a cucumber
pathogen that inhabits the same ecological niche as P. flocculosa and is sen-
sitive to the antifungal fatty acids produced by the antagonist. Growth of C.
cucumerinum was inhibited by both fatty acids, with 6-methyl-9-heptadecenoic
acid demonstrating greater biological activity (Figure 3). Determination of MICs
resulted in no growth of C. cucumerinum at 0.6 and 0.35 mg ml for 9-hep-
/
tadecenoic acid and 6-methyl-9-heptadecenoic acid, respectively. Probit analy-
sis revealed that the dose of each antibiotic expected to reduce growth of C.
cucumerinum by 50% is 0.16 and 0.10 mg ml for 9-heptadecenoic acid and 6-
/
methyl-9-heptadecenoic acid, respectively. These results support previous quali-

ANTIBIOTIC FATTY ACIDS IN Pseudozyma
997
FIG. 3. Growth of three-day liquid cultures of Cladosporium cucumerinum following
treatment with (Z)-9-heptadecenoic and (Z)-6-methyl-9-heptadecenoic acids at concen-
trations ranging from 0 to 1 mg ml. Values are mean SD.
/
tative reports that 6-methyl-9-heptadecenoic acid has the higher biological activ-
ity of the two when assayed against C. cucumerinum (Benyagoub et al., 1996a).
The antifungal fatty acids were also tested against S. fuliginea, the causal
agent of cucumber powdery mildew for which P. flocculosa is under study as a
biocontrol agent. When assayed against S. fuliginea, both antibiotics induced a
rapid collapse of the conidial chains of the pathogen, with once again a seem-
ingly greater effect of 6-methyl-9-heptadecenoic acid (Table 1). This collapse
of conidial chains was similar to the one induced by P. flocculosa itself. This
suggests that these two fatty acids are, at least in part, responsible for the biocon-
trol activity. When assayed at 0.1 mg ml, 6-methyl-9-heptadecenoic acid caused
/
a significantly higher degree of collapse compared to 9-heptadecenoic acid 12
hr following the treatment. At 24 hr, these same treatments no longer showed
differences in biological activity. This indicates that 6-methyl-9-heptadecenoic
acid initially acts more rapidly on S. fuliginea than its nonmethylated counter-
part. The fatty acids used under these conditions lost their activity when assayed
at 36 hr (data not shown). This is a result of their rapid degradation, as previously
demonstrated, when they are not stored under suitable conditions (appropriate
temperature and protection from oxidation) (Kates, 1986). This rapid degrada-

´
AVIS, BOULANGER, AND BELANGER
998
TABLE 1. Sphaerotheca fuliginea CONIDIAL CHAIN COLLAPSE FOLLOWING TREATMENT
WITH AQUEOUS SOLUTIONS OF ANTIBIOTIC FATTY ACIDS FROM Pseudozyma flocculosaa
(Z)-9-Heptadecenoic
(Z)-6-Methyl-9-heptadecenoic
acid
acid
Control
Time (hr)
(0.0 mg ml)
0.1 mg ml
0.2 mg ml
0.1 mg ml
0.2 mg ml
/
/
/
/
/
0
12
24
0.000a
0.000a
0.167a
0.083a
1.250b
2.917b
0.083a
1.917c
3.000b
0.167a
2.083c
3.083b
0.167a
2.167c
3.417b
a
Values are means of conidial collapse where 0 no effect on S. fuliginea conidial chains, 1
1–25% collapse of conidial chains, 2 26–50%, 3 51–75%, and 4 76–100% collapse. Within
a line, means followed by same letters are not significantly different by Fisher’s protected LSD
(P 0.01).
tion indicates that P. flocculosa itself must be applied as the control measure
of S. fuliginea because the fatty acids alone could not provide continuous con-
trol. Because the fatty acids degrade rapidly, there is less selection pressure on
powdery mildew populations as a result of limited exposure to the compounds.
This implies that the development of resistant strains of powdery mildew fungi
to the antibiotics may be reduced or retarded on the leaf surface. Along these
lines, it is apparent that residual activity would not be an issue, which makes P.
flocculosa an environmentally friendly agent of powdery mildew biocontrol.
From an ecological perspective, P. flocculosa is considered a weak com-
petitor for nutrients (Hajlaoui et al., 1992) and would need some other mech-
anisms to survive on the phylloplane. Therefore, these antibiotics would pro-
vide this antagonist with a means to protect its ecological niche against powdery
mildew fungi. This in turn explains its rather selective biological control effi-
ciency against powdery mildew fungi, which compete for the same ecological
niche on the leaf surface. The competition is not based on nutrients because they
are biotrophs, i.e., organisms that develop only on another living organism.
We consider the synthesis of these two fatty acids produced by P. flocculosa
as a major breakthrough in our system because it allows further experimenta-
tion of their antibiotic activity without the underlying difficulties associated with
extraction and purification procedures from culture filtrates. The newly synthe-
sized molecules may be used as standards to follow kinetics and quantify biosyn-
thetic productions by P. flocculosa. This should give a better understanding of
the agent’s antagonistic activity on leaf surfaces and provide a tool for screening
and selecting P. flocculosa strains with superior biocontrol potential.
It has been shown that these antibiotic fatty acids induce leakage of elec-
trolytes and proteins (Hajlaoui et al., 1992). These and other results led to the
hypothesis of a structural mode of action of P. flocculosa antibiotics, where fun-

ANTIBIOTIC FATTY ACIDS IN Pseudozyma
999
gal membranes are the primary target of action. Briefly, the antibiotics would par-
tition into fungal membranes and induce disorder due to their bulkiness caused
by the cis double bond located near the center of these fatty acids. This in turn
would cause changes in the physical properties and function of fungal mem-
branes, which would induce electrolyte leakage and increased fluidity as shown
by Benyagoub et al. (1996b). It is now apparent that the antifungal fatty acids
do not directly or indirectly act upon membrane sterols, factors that influence
membrane fluidity, because they are toxic to pythiaceous fungi that do not con-
tain membrane sterols (unpublished results). Furthermore, the dose-dependent
increase in both toxicity (this study) and fluidity (Benyagoub et al., 1996b) by
the fatty acids would suggest that they interfere with the general membrane lipid
domain rather than a localized effect on one of its components. This model is
also consistent with results in this study, in which the two antibiotics show dif-
ferent activity in exposed fungi. An additional methyl branch would occupy a
larger cross-sectional area in fungal membranes, thereby inducing greater disor-
der and increased fluidity. This would support the higher biological activity of
6-methyl-9-heptadecenoic acid.
The newly synthesized molecules should facilitate elucidation of the spe-
cific mode of action of this biocontrol agent. For example, immunolocalization
and cytochemical observations could pinpoint sites of action of the fatty acids
within fungal cells, thus providing a valuable tool to validate the model and
eventually help in predicting the efficacy of the antagonist by analysis of cellu-
lar membranes of powdery mildew fungi.
Overall, the synthesis of (Z)-9-heptadecenoic and (Z)-6-methyl-9-heptade-
cenoic acids from P. flocculosa will give stable and sufficient quantities of the
products for further testing with regards to specific mode of action of the anti-
biotics. This will be helpful in elucidating the molecular and ecological phenom-
ena behind the antagonism by P. flocculosa and in helping the prediction of its
biocontrol efficiency. Finally, the antibiotics can be used in repeated selection
pressure experiments to determine whether target fungi can eventually develop
resistance.
Acknowledgments—Financial assistance from NSERC and Plant Products Co. Ltd.
REFERENCES
AMES, D. E., and BOWMAN, R. E. 1951. Synthetic long-chain aliphatic compound. Part III. A critical
examination of two methods of synthesis of olefinic acids. J. Chem. Soc. 1951:1079–1086.
BELANGER, R. R., and DEACON, J. W. 1996. Interaction specificity of the biocontrol agent Sporothrix
´
flocculosa: A video microscopy study. Phytopathology 86:1317–1323.
´
´
BELANGER, R. R., LABBE, C., and JARVIS, W. R. 1994. Commercial-scale control of rose powdery
mildew with a fungal antagonist. Plant Dis. 78:420–424.

´
AVIS, BOULANGER, AND BELANGER
1000
´
BENYAGOUB, M., BEL RHLID, R., and BELANGER, R. R. 1996a. Purification and characterization
of new fatty acids with antibiotic activity produced by Sporothrix flocculosa. J. Chem. Ecol.
22:405–413.
BENYAGOUB, M., WILLEMOT, C., and BELANGER, R. R. 1996a. Influence of a subinhibitory dose of
antifungal fatty acids from Sporothrix flocculosa on cellular lipid composition in fungi. Lipids
31:1077–1082.
BOEKHOUT, T. 1995. Pseudozyma Bandoni emend. Boekhout, a genus for yeast-like anamorphs of
Ustilaginales. J. Gen. Appl. Microbiol. 41:359–366.
¨
BRANDANGE, S., DAHLMAN, O., LINDQUIST, B., MAHLEN, A., and MORCH, L. 1984. Absolute config-
uration and enantioselective synthesis of spiculisporic acid. Acta Chem. Scand. B 38:837–844.
BROUGHTON, B. W., BOWMAN, R. E., and AMES, D. E. 1952. Synthetic long-chain aliphatic com-
pound. Part VII. Some mono-olefinic acids. J. Chem. Soc. 1952:671–677.
GRIMMER, G., and KRACHT, J. 1963. The synthesis of odd-numbered unsaturated fatty acids. Berichte.
96:3370–3373.
´
HAJLAOUI, M. R., and BELANGER, R. R. 1991. Comparative effects of temperature and humidity
on the activity of three potential antagonists of rose powdery mildew. Neth. J. Plant Pathol.
97:203–208.
´
HAJLAOUI, M. R., and BELANGER, R. R. 1993. Antagonism of the yeast-like phylloplane fungus
Sporothrix flocculosa against Erysiphe graminis var. tritici. Biocont. Sci. Technol. 3:427–434.
´
HAJLAOUI, M. R., BENHAMOU, N., and BELANGER, R. R. 1992. Cytochemical study of the antagonistic
activity of Sporothrix flocculosa on rose powdery mildew, Sphaerotheca pannosa var. rosae.
Phytopathology 82:583–589.
JACOB, V. J., and GRIMMER, G. 1966. Bildung doppelbindungsisomerer alkencarbonsa
hydrierung von 9-alkincarbonsauren. Tetrahedron Lett. 24:2687–2688.
¨
uren bei der
¨
JARVIS, W. R., SHAW, L. A., and TRAQUAIR, J. A. 1989. Factors affecting antagonism of cucumber
powdery mildew by Stephanoascus flocculosus and S. rugulosus. Mycol. Res. 92:162–165.
KATES, M. 1986. Techniques of lipodology: isolation, analysis and identification of lipids, pp.
100–111, in T. S. Burton and P. H. van Knippenberg (eds.). Laboratory Techniques in Bio-
chemistry and Molecular Biology. Elsevier, Amsterdam.
NGUYEN, B. T., and CARTLEDGE, F. K. 1986. A convenient synthetic route to methylated silacyclo-
hexanes. J. Org. Chem. 51:2206–2210.