Biosynthesis of (R)-γ-decanolactone in the yeast Sporobolomyces odorus

Article abstract of DOI:10.1021/jf950570w

Addition of [9,10-²H₂]oleic acid to cultures of the yeast Sporobolomyces odorus led to the formation of labeled (R)-γ-decanolactone and both enantiomers of (Z)-6-γ-dodecenolactone and γ-dodecanolactone. The labeling patterns of these lactones were estimated by a quantitative gas chromatography/mass spectrometry method, suitable for ring-labeled lactones. For the first time, there is strong evidence of oleic acid being a genuine precursor of the important aroma compound (R)-γ-decanolactone. On the basis of the presented and former results, a new biosynthetic pathway of this compound is proposed. Starting with a strict enantioselective (R)-12-hydroxylation of oleic acid as an initial step, the following β-oxidation led to the lactone. In addition, biosynthetic aspects of the formation of (Z)-6-γ-dodecenolactone and γ-dodecanolactone in S. odorus are discussed.

Full text of DOI:10.1021/jf950570w

1218
J. Agric. Food Chem. 1996, 44, 1218−1223
Biosyn th esis of (R)-γ-Deca n ola cton e in th e Yea st Spor obolom yces
od or u s
Thomas Haffner* and Roland Tressl
Institut fu¨r Biotechnologie, Fachgebiet Chemisch-Technische Analyse, Technische Universita¨t Berlin,
Seestrasse 13, D-13353 Berlin, Germany
Addition of [9,10-2H2]oleic acid to cultures of the yeast Sporobolomyces odorus led to the formation
of labeled (R)-γ-decanolactone and both enantiomers of (Z)-6-γ-dodecenolactone and γ-dodecano-
lactone. The labeling patterns of these lactones were estimated by a quantitative gas chromatog-
raphy/mass spectrometry method, suitable for ring-labeled lactones. For the first time, there is
strong evidence of oleic acid being a genuine precursor of the important aroma compound (R)-γ-
decanolactone. On the basis of the presented and former results, a new biosynthetic pathway of
this compound is proposed. Starting with a strict enantioselective (R)-12-hydroxylation of oleic
acid as an initial step, the following â-oxidation led to the lactone. In addition, biosynthetic aspects
of the formation of (Z)-6-γ-dodecenolactone and γ-dodecanolactone in S. odorus are discussed.
Keyw or d s: γ-Lactones; aroma compound; (R)-γ-decanolactone; [9,10-2H2]oleic acid; Sporobolomyces
odorus; biosynthesis
INTRODUCTION
serve as precursor either for 1 or for (Z)-6-γ-dodeceno-
lactone (2) (Tressl et al., 1978). In further experiments,
the addition of deuterium-labeled, racemic ethyl (E)-3,4-
epoxydecanoate to cultures of the yeast led to the
formation of (R)-1 with a purity of 38% enantiomeric
excess (ee). In strawberry tissue homogenate, the same
precursor was transformed into (R)-1 with less optical
purity than found in the original fruit. In contrast, the
racemic (Z)-stereoisomer of ethyl 3,4-epoxydecanoate
was not a substrate for the biosynthesis of 1 either in
S. odorus or in strawberry homogenate (Tressl et al.,
1988; Albrecht and Tressl, 1990).
γ-Decanolactone (1), having a distinct fruity flavor,
is a key aroma compound in many fruits. Therefore, 1
is of great interest as a flavor additive used in food and
aroma industries (Gatfield et al., 1993). The enantio-
meric composition of this chiral compound is charac-
teristic for each kind of fruit. In most cases, (R)-1 is
found as the dominating enantiomer (Dufosse et al.,
1994). Enantiomeric compositions are used as one
criterion in analysis of the origin of foodstuff and flavor
additives. Therefore, understanding the basic biosyn-
thetic principles, leading to the different enantiomers
of 1 should be of great interest.
Many cultured microorganisms, including Sporobo-
lomyces odorus, are able to biosynthesize and ac-
cumulate 1 and other γ-lactones (Welsh et al., 1989;
Tahara et al., 1975). These organisms can serve as
suitable models for investigations concerning the γ-lac-
tone biosynthesis in general. In the past, the precursors
of lactones were proven to be hydroxylated fatty acids
of C18 chain length, which were shortened by â-oxidation
steps, followed by lactonization (Albrecht et al., 1992;
Cardillo et al., 1989a,b, 1991; Ercoli et al., 1992). The
enantiomeric compositions of γ-lactones are different in
many microorganisms, too. Under different growing
conditions, S. odorus produces only (R)-1 (>99%); how-
ever, the relative amount of (R)-1 in the fungus Fusar-
ium poae varies from 30 to 85% purity (Albrecht and
Tressl, 1990; Latrasse et al., 1993). Until now, the
natural endogenous C18 precursor of 1 could not be
detected in γ-lactone-producing microorganisms, but the
bioconversion of ricinoleic acid [(R)-12-hydroxyoleic acid]
into pure (R)-1 has been well-known for more than 30
years and patented for various organisms, including S.
odorus. This process serves for the industrial produc-
tion of (R)-1 (Okui et al., 1963; Cheetham et al., 1987;
Cardillo et al., 1989a,b; Farbood and Willis, 1983).
The first successful investigations to elucidate the
biosynthetic pathway to 1 in S. odorus were made by
using labeled C10 precursors. It could be shown that S.
odorus converted radioactively labeled [1-14C]decanoic
acid into 1. In contrast, [1-14C]dodecanoic acid did not
Our intentions were to carry out further investiga-
tions concerning the unsatisfactory enantiomeric purity
of (R)-1 synthesized from the racemic ethyl epoxy-
decanoate by the yeast. Additionally, it must be as-
sumed that the intermediate compound (E)-3,4-epoxy-
decanoate is formed from (E)-3-decenoic acid by the
activity of a monooxygenase. Therefore, linoleic acid is
supposed to be the natural precursor of (E)-3-decenoic
acid via â-oxidative degradation. But linoleic acid itself
did not serve as a potent precursor of 1, as results of
previous experiments showed (Albrecht, 1991). Only
trace amounts of 1 were found labeled after administra-
tion of [9,10,12,13-2H4]linoleic acid to cultures of S.
odorus (Albrecht et al., 1992). Furthermore, the bio-
synthesis of ricinoleic acid in seeds of castor bean
(Ricinus Communis L.) was confirmed to base on oleic
acid as fatty acid precursor (Yamada and Stumpf, 1964;
J ames et al., 1965; Galliard and Stumpf, 1966). This
led us to the idea to test the ability of [9,10-2H2]oleic
acid to serve as a substrate in the biosynthesis of 1 via
ricinoleic acid as an intermediate in S. odorus.
In this paper, a new biosynthetic pathway of (R)-γ-
decanolactone is proposed, on the basis of the gas
chromatography-mass spectroscopy (GC-MS) and chiral
GC analysis of the metabolites, obtained after admin-
istration of [9,10-2H2]oleic acid to cultures of S. odorus.
In addition, the biotransformation of [9,10-2H2]oleic acid
into (Z)-6-γ-dodecenolactone and γ-dodecanolactone is
discussed.
S0021-8561(95)00570-X CCC: $12.00
© 1996 American Chemical Society

Biosynthesis of (R)-
γ
-Decanolactone
J. Agric. Food Chem., Vol. 44, No. 5, 1996 1219
MATERIALS AND METHODS
[9,10-2H]Oleic Acid . 9-Octadecynoic acid (4.9 g, 17.47
mmol) (98%, Lancaster, U.K.), dissolved in 60 mL of hexane,
was hydrogenated under a D2 atmosphere in the presence of
200 mg of 5% Pd/BaSO4 and 400 µL of freshly destilled
quinoline. After the consumption of the required volume of
D2, the quinoline was removed by acidic extraction and the
remaining product was washed with brine, dried over NaSO4,
and purified on silica gel. A total of 4.46 g (15.67 mmol, 90%)
of the product was obtained. The isotopic purity was >97%
(1H-NMR).
Mass spectrum as methyl ester (70 eV) m/z (intensity,
fragment) 298 (1, M+), 267 (10, M+ - 31), 266 (16), 224 (8),
182 (7), 168 (4), 153 (5), 139 (10), 125 (11), 111 (22), 98 (40),
84 (55), 74 (85), 69 (75), 55 (100), 43 (97). 1H-NMR (500 MHz)
(in parts per million): 0.88 (t, 3 H, J ) 7.5 Hz, CH3), 1.20-
1.39 (m, 20 H, CH2), 1.63 (quint, 2 H, J ) 7.5 Hz, CH2CH2-
COOH), 1.98-2.04 (m, 4 H, CH2CDdCDCH2), 2.35 (t, 2 H, J
) 7.5 Hz, CH2COOH).
F eed in g Exp er im en ts. S. odorus (ATCC 24259) was
cultivated in a 1 L Erlenmeyer flask containing 200 mL of
medium [45 g/L sucrose, 5 g/L lactose, 3 g/L MgSO4‚7H2O, 2.5
g/L (NH4)2SO4, 2.5 g/L KH2PO4, 2.5 g/L L-alanine, and 0.1 g/L
CaCl2‚2H2O] at 22 °C and 80 rpm (Albrecht, 1991). After 96
h of incubation, 50 mg (176 µmol) of [9,10-2H2]oleic acid
dissolved in 150 µL of ethanol was added. At different time
points, 10 mL aliquots were removed from the culture broth
and extracted with diethyl ether (experiment I). Following
treatment with diazomethane, the extracts were analyzed by
GC and GC/MS. In a second experiment (II) and under the
same cultivation conditions, 200 mg (704 µmol) of [9,10-2H2]-
oleic acid, dissolved in 300 µL of ethanol, was added to 200
mL of sterile medium before inoculation. After 220 h, the
whole broth was extracted with diethyl ether, and after
concentration to 0.2 mL, the extract was separated on a LSC
column (6.5 g of silica gel) into five 40 mL fractions [I, pentane,
II, pentane/CH2Cl2 (9:1); III, pentane/CH2Cl2 (2:1); IV, pentane/
diethyl ether (9:1); V, diethyl ether]. Fraction IV contained
all of the lactones (Albrecht, 1991).
Ga s Ch r om a togr a p h y (GC)/Ma ss Sp ectr om etr y (MS)/
NMR. If necessary, the extracts were saturated with diazo-
methane, and following concentration to 0.2 mL, 1 µL was
subjected to GC analysis on a capillary column (DB-Wax 60
M, 60 m × 0.32 mm inside diameter, J &W, Folsom, CA) and
a chiral column (Lipodex E, 50 m, Machery & Nagel, Ger-
many), both coupled to a FID detector or for the DB wax
column additionally connected to the mass spectrometer. The
mass spectra were recorded on a Varian CH 5-DF (Varian,
Germany) double-focusing mass spectrometer at 70 eV ioniza-
tion energy, combined with a MS data system DP 10 (AMD,
Germany). 1H-NMR spectra were recorded on a Bruker AMX
500 NMR instrument with CDCl3 as solvent.
F igu r e 1. EI-mass spectra of partially labeled γ-decanolac-
tone (I), (Z)-6-γ-dodecenolide (II), and γ-dodecanolactone (III)
found in cultures of S. odorus 147 h after the addition of 250
ppm [9,10-2H2]oleic acid to growing cells of the yeast (condi-
tions, experiment I).
m/ z 85, 86, and 87 over all MS scans representing the
corresponding lactone. Therefore, the single ion moni-
toring mode of the MS software was used. Before
calculation of the labeling pattern, the naturally occur-
ring isotopic pattern of the mass fragment m/ z 85 was
taken into account.
RESULTS
[9,10-2H2]Oleic acid was administered to S. odorus
cultures. At different time points, aliquots of the culture
broth were taken and the formation of volatiles was
monitored by GC/MS analysis. In Figure 1, mass
spectra of 1, (Z)-6-γ-dodecenolactone (2), and γ-dodec-
anolatone (3) are presented, 147 h after addition of 250
ppm labeled substrate, representing the three labeled
γ-lactones formed during the cultivation. The fragment
ion m/ z 85 (tetrahydropyranonyl cation ) base peak)
is the most important in the mass spectral analysis of
γ-lactones with saturated ring moieties. In addition, the
prominent ions m/ z 86 and 87 were detected, thus
demonstrating the content of the deuterium-labeled
lactones, containing one or two deuterium atoms in the
ring part. In our experiments, isotopomere lactones
showed a partial resolution on capillary GC colums.
Consequently, an isolated MS scan was not representa-
tive for the quantification of the label. More reliable
results were obtained by integrating the fragment ions
In Figures 2 and 3, the time dependent formation of
labeled and unlabeled 1 and 2 after addition of 250 ppm
[9,10-2H2]oleic acid is shown (experiment I). In contrast
to the lactones 1 and 2, labeled 3 was detected for the
first time 96 h after the substrate addition. Hence, a
quantification of 3 compared to 1 and 2 was not feasible.
Until 147 h, the amounts of labeled 1 and 2 increased.
At the same time, no labeled oleic acid could be detected
anymore, but linoleic acid found to be labeled in a
relative amount of about 10-20% (Figure 4). After this
time, only 5-25 ppm oleic acid, biosynthesized endog-
enously, was found, which served as substrate for the
unlabeled lactones furthermore. Overall, less than 3%
of the precursor was transformed into the three γ-lac-
tones, with the remaining amount suggested to be
metabolized via â-oxidation or incorporated into the
triglycerides and phospholipids, respectively.

1220 J. Agric. Food Chem., Vol. 44, No. 5, 1996
Haffner and Tressl
Ta ble 1. Con cen tr a tion a n d Rela tive Am ou n ts of 1- a n d
2-F old -La beled γ-Deca n ola cton e, (Z)-6-γ-Dod ecen o-
la cton e, a n d γ-Dod eca n ola cton e F ou n d a fter Ad d ition of
250 a n d 1000 p p m [9,10-2H2]Oleic Acid to Cu ltu r es of S.
od or u sa
[9,10-2H2]oleic acid
exp I
exp II
250 ppm (147 h)
1000 ppm (220 h)
total
concn [2H1] [2H2] label concn [2H1] [2H2] label
lactone (ppm) (%) (%) (%) (ppm) (%) (%) (%)
total
1
2
3
4.1
28.7
<0.01
16.8 <0.5 17.3
10.8
12.7
1.1
47.5
2.8 50.3
F igu r e 2. Concentration of [2-2H1]-γ-decanolactone and γ-
decanolactone during the fermentation after addition of 250
ppm [9,10-2H2]oleic acid to growing cells of S. odorus.
1.6 19.5 21.1
2.8 26.8 29.6
15.8 33.4 49.2
nd
nd
nd
a
Experiments I and II, see Materials and Methods. nd is not
determined.
F igu r e 3. Concentration of [3,4-2H2]-(Z)-6-γ-dodecenolactone
and unlabeled (Z)-6-γ-dodecenolactone during the fermentation
after addition of 250 ppm [9,10-2H2]oleic acid to growing cells
of S. odorus.
F igu r e 5. Capillary gas chromatogram (reconstructed ion
detection) of the lactone fraction (conditions, experiment II)
operating in the single ion monitoring mode and quantification
of the mass fragments m/ z 85, 86, and 87 [DB-Wax 60 M, 60
m × 0.32 mm inside diameter, 70 °C (4 min isothermal) to
220 °C at 4 °C/min].
fermentation time. In reference cultures, the ratio was
constant at 1 to 3-4. In contrast, addition of 1000 ppm
[9,10-2H2]oleic acid to the medium before inoculation
(experiment II) caused an opposite effect. The relative
ratio of 1 to 2 was 1 to 1.2, 220 h after fermentation
started. At this time, the concentration of 1 was 10.8
ppm and that of 2 was 12.7 ppm (Table 1).
The enhanced metabolization of labeled oleic acid to
1 can also be demonstrated by the relative content of
the labeled lactones (Figure 5). The whole labeling
content of 1 was 50.3%; only a small amount of 2.8%
was labeled 2-fold. Lactone 2 was found labeled in a
relative quantity of 29.6% with an 2.8% amount of 1-fold
labeling. The biosynthesis of lactone 3 increased, too.
After 220 h, a concentration of 1.1 ppm was detected. A
total of 49.2% of 3 was found labeled, but different from
2, 15.8% of 3 was only labeled once and 33.4% contained
a second deuterium atom.
In Figure 6, the chiral analysis of the lactone fraction
147 h after the addition of [9,10-2H2]oleic acid is shown.
The enantiomeric excess (ee) of the (R)-enantiomers of
1 and 2 was found to be >99 and 84%, respectively.
These values are identical to those found in original
cultures of S. odorus before (Albrecht, 1991; Albrecht
et al., 1992). On the basis of these results, an isotope
effect on the enzymes involved in the formation of the
chirality center, caused by the deuterium label, can be
excluded. For the first time, the enantiomeric composi-
tion of γ-dodecanolactone in S. odorus could be analyzed.
F igu r e 4. (I) EI-mass spectrum of [9,10-2H2]methyl linoleate
mixed with unlabeled methyl linoleate found in cultures of S.
odorus 147 h after the addition of 250 ppm [9,10-2H2]oleic acid
to growing cells of the yeast (conditions, experiment I) and
(II) EI-mass spectrum of unlabeled methyl linoleate as refer-
ence.
Under the above conditions (experiment I) and com-
pared to original reference cultures, the addition of 250
ppm [9,10-2H2]oleic acid influenced neither the forma-
tion rate nor the amount of accumulated lactones. The
ratio of 1 to 2 was found to be 1 to 6-7 over the whole

Biosynthesis of (R)-
γ
-Decanolactone
J. Agric. Food Chem., Vol. 44, No. 5, 1996 1221
Sch em e 1. P r op osed Tr a n sfor m a tion of [9,10-2H2]-
Oleic Acid to [2-2H1]-(R)-γ-Deca n ola cton e by Wh ole
Cells of S. od or u s
F igu r e 6. Chiral capillary gas chromatogram of the lactone
fraction isolated from growing cultures of S. odorus 147 h after
addition of [9,10-2H2]oleic acid [conditions, experiment I;
Lipodex E, 50 m × 0.25 mm inside diameter, 150 °C (20 min
isothermal) to 220 °C at 2 °C/min; carrier gas, He].
In contrast to the excess of the (R)-enantiomers of 1 and
2, lactone 3 contained the (S)-enantiomer in 59% ee.
DISCUSSION
For the first time, there is strong evidence of oleic acid
being a genuine precursor of the important aroma
compounds (R)-γ-decanolactone and γ-dodecanolactone,
(Z)-6-γ-dodecenolactone in a microorganism. Along with
linoleic acid, now oleic acid must be seen as a potent
natural substrate in the biosynthesis of γ-lactones in
the yeast S. odorus.
Concerning the biosynthesis of γ-decanolactone, we
suggest a strictly stereospecific, enzymatic introduction
of the hydroxyl group to oleic acid at carbon-12 by the
action of a hydroxylase. This kind of microbial mono-
oxygenase activity, introducing a (R)-12-hydroxy group
to oleic acid, was characterized first in Bacillus pumilis.
Cultures of this organism convert 72% of added oleic
acid to ricinoleic acid after 96 h without further meta-
bolic degradation (Kenji and Kido, 1989). A second
parallel can be drawn to the biosynthesis of ricinoleic
acid in castor bean seeds, where oleic acid had been
proven to be the precursor of the enantiomerical pure
ricinoleic acid [(R)-12-hydroxyoleic acid] (Yamada and
Stumpf, 1964; J ames et al., 1965; Galliard and Stumpf,
1966). Our idea is supported by the ability of S. odorus
to metabolize ricinoleic acid to (R)-γ-decanolactone
rapidly in high yield (Albrecht et al., 1992). Scheme 1
summarizes the proposed enzymatic steps from [9,10-
2H2]oleic acid to [2-2H1]-(R)-1 on the basis of known
biochemical pathways of fatty acid degradation. (R)-
12-Hydroxyoleic acid is activated by the conversion to
its CoA ester and shortened by two â-oxidation cycles
leading to (R)-8-hydroxy-(Z)-5-tetradecenoyl-CoA. This
acid CoA possesses a (Z)-configured ∆5-double bond,
which cannot be metabolized without isomerizing the
double bond to position trans-2 first as Smeland et al.
(1992) and Chen et al. (1995) demonstrated. The next
step in this reaction cascade will be the introduction of
an additional trans-2 double bond to the molecule by
the action of an acyl-CoA dehydrogenase, leading to (R)-
8-hydroxy-(E,Z)-2,5-tetradecadienoyl-CoA. Further en-
zymatic transformation steps, including the key activity
of a ∆3,5,∆2,4-dienoyl-CoA isomerase, led finally to (R)-
8-hydroxy-(E)-2-tetradecenoyl-CoA which serves as sub-
strate for two further â-oxidation steps, resulting in (R)-
4-hydroxydecanoyl-CoA, which lactonizes to 1 after
hydrolysis of the CoA ester. To consolidate this postu-
lated pathway in S. odorus, more detailed investigation
of this enzymatic system will be necessary, including
isolation and characterization of intermediates. Espe-
cially the occurrence of oleic acid, either as free acid or
esterified to triglycerides or phospholipids, as well as
the elucidation of the key step, the enzymatic 12-
hydroxylation remains very interesting. In microsomes
of castor bean seeds, there was evidence of a cytochrome
b5 acting as the electron donor in this reaction. Cyto-
chrome P-450 was not involved here. Additionally,
Bafor et al. (1991) found that only oleic acid esterified
to phosphatidylcholine can serve as substrate for the
hydroxylation reaction. Neither the free acid nor oleoyl-
CoA is a suitable substrate for this reaction catalyzed
by microsomal enzymes in castor bean.
Former results concerning the lactone formation in
the yeast showed that [9,10,12,13-2H4]linoleic acid was
not converted into 1 significantly but served as an
excellent substrate for the biosynthesis of 2 and few
other γ- and δ-lactones (Albrecht, 1991; Albrecht et al.,
1992). These results and the findings that ethyl (E)-
3,4-epoxydecanoate and its corresponding acid, which
are supposed to originate from linoleic acid by â-oxida-

1222 J. Agric. Food Chem., Vol. 44, No. 5, 1996
Haffner and Tressl
Sch em e 2. P ostu la ted Meta bolic Step s of [9,10-2H2]Oleic Acid to γ-La cton es in S. od or u s
tion and epoxidation, are precursors of 1 are inconsis-
tent. More experimental data are needed to solve the
discrepancy.
tabolite of oleic acid in S. odorus. Obviously, the
biosynthesis of 3 requires the introduction of a hydroxyl
function to oleic acid at carbon-10. Microbial hydrations
of oleic acid to 10-hydroxystearic acid performed by
bacteria and yeasts are well-known and occur with
different enantioselectivities, but always with an excess
of the (R)-isomer (Wallen et al., 1962; El-Sharkawy et
al., 1992; Yang et al., 1993). The enantioselectivity of
the hydration is often modified by a following oxidation/
reduction step of the 10-hydroxystearic acid catalyzed
by dehydrogenases (W. Albrecht and R. Tressl, unpub-
lished results). However, the occurrence of the two
different labeled isotopomers of 3 requires an explana-
tion. For [2H1]-3 we assume an enzymatic inversion of
the stereocenter at carbon-10, including a loss of one
deuterium atom. To confirm this hypothesis, a mass
spectral analysis of the two separated enantiomers of 3
would be helpful. In our investigations, this analytical
procedure failed because of the very low concentration
of this compound produced by the yeast. Further
detailed investigations are in progress.
The activities of the involved enzymes can be modified
by exposing the organism to different concentrations of
oleic acid. Addition of 250 ppm oleic acid to 96 h old
cultures of the yeast enhanced the production of 2
compared to that of 1, via the desaturase pathway.
Administration of 1000 ppm oleic acid to the culture
broth before inoculation favors the formation of 1 and
led to a decreased production of 2. Under these condi-
tions, 3 was synthesized in an about 100-fold concentra-
tion, compared to reference cultures where it could be
detected only in trace amounts (<10 ppb). This finding
can be explained by a positive induction of the proposed
hydroxylation activity leading to (R)-12-hydroxyoleic
acid.
The presented results demonstrate the ability of S.
odorus to synthesize two additional lactones (2 and 3)
from oleic acid as a natural precursor. Taking into
account the successful transformation of linoleic acid
into 2, the activity of a ∆12-desaturase as the initial
enzymatic step in the metabolic pathway of oleic acid
leading to 2 must be assumed. There is evidence for
this step, because 147 h after the addition of [9,10-2H2]-
oleic acid (experiment I) about 15-20% of the linoleic
acid was found to be deuterium-labeled (Figure 4)
(quantification of the label in linoleic acid could not be
as exact as for the lactones, because of the loss of
deuterium during fragmentation of the molecule in the
mass spectrometer). Other linoleic acid-derived lactones
could not be detected in labeled form, because their
concentration in the culture broth was too low under
the chosen experimental conditions (δ-decanolactone) or
the label was separated during the metabolic degrada-
tion of the linoleic acid derivatives (γ-nonanolactone).
The enzymatic introduction of the required 10-hy-
droxyl group to linoleic acid as the initial step in the
presented pathway to 2 still remains unclear. In view
of these results, there was neither direct evidence for
the activity of a 10-lipoxygenase, comparable to mush-
rooms (Wurzenberger and Grosch, 1984), nor the action
of a 10-hydratase, or alternatively an activity of a 9,10-
epoxygenase/hydrolase. In contrast to our results where
(R)-2 was the dominant product (application of the
Cahn-Ingold-Prelog rules causes a formal inversion
of the stereocenter), the stereospecificity of microbial 10-
hydration of the ∆9-double bond in linoleic acid always
led to an excess of the (R)-10-hydroxy enantiomer, which
would lead to (S)-2 after metabolic degradation (Hou,
1994). Compared with that, all known lipoxygenases
catalyze in a (S)-stereospecific mode, but there is no
specific information about the stereochemical course of
10-lipoxygenases in microorganisms until now. There-
fore, more detailed investigation of this initial enzymatic
step is necessary to identify the involved enzyme(s) in
S. odorus.
Our results, summarized in proposed pathways
(Scheme 2), indicate that the biotransformation of oleic
acid to γ-lactones in S. odorus is complex and seems to
be based on at least three different metabolic pathways.
For (R)-γ-decanolactone, a strict (R)-12-enantioselective
hydroxylation of oleic acid, catalyzed by a hydroxylase,
should be the first enzymatic step, followed by four
â-oxidation cycles and dehydration, leading to the
lactone. Parallel to this pathway, a ∆12-desaturase
activity leading to linoleic acid was confirmed by the
For the first time, γ-dodecanolactone, containing an
excess of the (S)-enantiomer, was identified as a me-

Biosynthesis of (R)-
γ
-Decanolactone
J. Agric. Food Chem., Vol. 44, No. 5, 1996 1223
Galliard, T.; Stumpf, P. K. Fat metabolism in higher plants.
XXX. Enzymatic synthesis of ricinoleic acid by a microsomal
preparation from developing Ricinus communis seeds. J .
Biol. Chem. 1966, 241, 5806-5812.
Gatfield, I. L.; Gu¨ntert, M.; Sommer, H.; Werkhoff, P. Some
aspects of the microbiological production of flavor-active
lactones with particular reference to γ-decalactone. Chem.,
Mikrobiol., Technol. Lebensm. 1993, 15, 165-170.
Hou, C. T. Conversion of linoleic acid to 10-hydroxy-12(Z)-
octadecenoic acid by Flavobacterium sp. (NRRL B-14859).
J . Am. Oil Chem. Soc. 1994, 71, 975.
J ames, A. T.; Hadaway, H. C.; Webb, J . P. W. The biosynthesis
of ricinoleic acid. Biochem. J . 1965, 95, 448-452.
Kenji, S.; Kido, T. Manufacture of hydroxy unsaturated fatty
acids with Bacillus pumilis. J pn. Kokai Tokkyo Koho J P
01 51.092, 1989.
characterization of the labeled double-unsaturated fatty
acid. The alternative first steps in the subsequent
metabolic degradation of linoleic acid and oleic acid still
remain unclear. They could be catalyzed by a 10-lipoxy-
genase and/or a 10-hydratase or 9,10-epoxygenase for
(Z)-6-γ-dodecanolactone or by the last two enzymes for
the saturated γ-dodecanolactone. The identification of
these enyzmatic activities in S. odorus is the subject of
our current research and needs further investigation,
especially in the isolation of significant intermediates
and in the characterization of the involved enzymes in
cell free systems.
ACKNOWLEDGMENT
Latrasse, A.; Guichard, E.; Piffaut, C.; Fournier, N.; Dufosse,
L. Chirality of the gamma-lactones formed by Fusarium
poae INRA 45. Chirality 1993, 5, 379-384.
We are very grateful to Hans Ko¨ppler for supporting
us in mass spectrometric analysis.
Luo, M. J .; Smeland, T. E.; Shoukry, K.; Schulz, H. ∆-3,5,∆-
2,4-Dienoyl-CoA isomerase from rat liver mitochondria. J .
Biol. Chem. 1994, 269, 2384-2388.
Morris, L. J .; Hall, S. W.; J ames, A. T. The biosynthesis of
ricinoleic acid by Claviceps purpurea. Biochem. J . 1966, 100,
29c-32c.
Okui, S.; Uchiyama, M.; Mizugaki, M. Intermediates of the
oxidative breakdown of ricinoleic acid by Candida. Biochim.
Biophys. Acta 1963, 70, 346-368.
LITERATURE CITED
Albrecht, W. Untersuchungen zur Biosynthese chiraler ali-
phatischer Lactone in Fru¨chten und Mikroorganismen.
Ph.D. Thesis, Technische Universita¨t Berlin, Germany,
1991.
Albrecht, W.; Tressl, R. Studies on the biosynthesis of γ-deca-
lactone in sporobolomyces odorus. Z. Naturforsch. 1990, 45C,
207-216.
Smeland, T. E.; Nada, M.; Cuebas, D.; Schulz, H. NADPH-
dependent beta-oxidation of unsaturated fatty acids with
double bonds extending from odd-numbered carbon atoms.
Proc. Natl. Acad. Sci. U.S.A. 1992, 89, 6673-6677.
Smith, M. A.; J onsson, L.; Stymne, S.; Stobart, K. Evidence
for cytochrome b5 as an electron donor in ricinoleic acid
biosynthesis in microsomal preparations from developing
castor bean (Ricinus communis L.). Biochem. J . 1992, 287,
141-144.
Tahara, S.; Fujiwara, K.; Ishizaka, H.; Mizutani, J .; Obata,
Y. γ-Decalactone - one constituent of volatiles in cultured
broth of Sporobolomyces odorus. Agric. Biol. Chem. 1972,
36, 2585-2587.
Tressl, R.; Apetz, M.; Arrieta, R.; Gru¨newald, K. G. Formation
of lactones and terpenoids by microorganisms. In Flavors
of Foods and Beverages; Charalambous, G., Inglett, G. E.,
Eds.; Academic Press: London, 1978; pp 145-168.
Tressl, R.; Heidlas, J .; Albrecht, W.; Engel, K.-H. Biogenesis
of chiral hydroxyacid esters. In Bioflavor ’87; Schreier, P.,
Ed.; de Gruyter: Berlin, 1988; pp 221-236.
Wallen, L. L.; Benedict, R. G.; J ackson, R. W. The microbio-
logical production of 10-hydroxystearic acid from oleic acid.
Arch. Biochem. Biophys. 1962, 99, 249-253.
Welsh, F. W.; Murray, W. D.; Williams, R. E. Lactones in foods.
Crit. Rev. Biotechnol. 1989, 9, 105-114.
Wurzenberger, M.; Grosch, W. The formation of 1-octen-3-ol
from the 10-hydroperoxide in linoleic acid by a hydroper-
oxide lyase in mushrooms (Psalliota bispora). Biochim.
Biophys. Acta 1984, 794, 25-30.
Yamada, M.; Stumpf, P. K. Enzymic synthesis of ricinoleic acid
by extracts of developing Ricinus communis L. seeds.
Biochem. Biophys. Res. Commun. 1964, 14, 165-171.
Yang, W.; Dostal, L.; Rosazza, J . P. N. Stereospecificity of
microbial hydrations of oleic acid to 10-hydroxystearic acid.
Appl. Environ. Microbiol. 1993, 59, 281-284.
Albrecht, W.; Heidlas, J .; Schwarz, M.; Tressl, R. Biosynthesis
and biotechnological production of aliphatic γ- and δ-lac-
tones. In Flavor Precursors, Thermal and Enzymatic Con-
versions; Teranishi, R., Takeoka, G. R., Gu¨ntert, M., Eds.;
ACS Symposium Series 490; American Chemical Society:
Washington, 1992; pp 46-58.
Bafor, M.; Smith, M. A.; J onsson, L.; Stobart, K.; Stymne, S.
Ricinoleic acid biosynthesis and triacylglycerol assembly in
microsomal preparations from developing castor-bean (Rici-
nus communis) endosperm. Biochem. J . 1991, 280, 507-
514.
Cardillo, R.; Fuganti, C.; Sacerdote, G.; Barbeni, M.; Cabella,
P.; Squarcia, F. Microbial production of (R)-γ-decalactone
and (R)-γ-octalactone. Eur. Pat. Appl. EP 356291, 1989a.
Cardillo, R.; Fronza, G.; Fuganti, C.; Grasselli, P.; Nepoti, V.
On the mode of conversion of racemic, C14-C19, γ-hydroxy
alkene fatty acids into C7-C11, optically active γ- and
δ-lactones in Cladosporium suaveolens. J . Org. Chem.
1989b, 54, 4979-4980.
Cardillo, R.; Fronza, G.; Fuganti, C.; Grasselli, P.; Mele, A.;
Pizzi, D. Stereochemistry of the microbial generation of
δ-decanolide, γ-dodecanolide, and γ-nonanolide from C18 13-
hydroxy, C18 10-hydroxy, and C19 14-hydroxy unsaturated
fatty acids. J . Org. Chem. 1991, 56, 5237-5239.
Cheetham, P. S. J .; Maume, K. A.; de Rooij, J . F. M. Microbial
manufacture of γ-decalactone from ricinoleic acid. Eur. Pat.
App. EP 258993, 1987.
Chen, L. S.; J in, S.-J .; Dejak, I.; Tserng, K.-Y. Isomerization
of trans-2,∆5-dienoyl-CoA’s to ∆3,∆5-dienoyl-CoA’s in the
â-oxidation of ∆5-unsaturated fatty acids. Biochemistry
1995, 34, 442-450.
Dufosse, L.; Latrasse, A.; Spinnler, H.-E. Importance of lac-
tones in food flavors: structure, distribution, sensory prop-
erties and biosynthesis. Sci. Aliments 1994, 14, 17-50.
El-Sharkawy, S. H.; Yang, W.; Dostal, L.; Rosazza, J . P. N.
Microbial oxidation of oleic acid. Appl. Environ. Microbiol.
1992, 58, 1226-2122.
Received for review August 22, 1995. Revised manuscript
received February 28, 1996. Accepted March 18, 1996.X
Ercoli, B.; Fuganti, C.; Grasselli, P.; Servi, S.; Allegrone, G.;
Barbeni, M.; Pisciotta, A. Stereochemistry of the biogen-
eration of C-10 and C-12 gamma lactones in Yarrowia
lipolytica and Pichia ohmeri. Biotechnol. Lett. 1992, 14,
665-668.
J F950570W
X Abstract published in Advance ACS Abstracts, May
1, 1996.
Farbood, R.; Willis, B. J . γ-Decalactone. PCT Int. Appl. WO
83 01,072, 1983.