Fungal biotransformation of (±)-linalool

Article abstract of DOI:10.1021/jf800099h

The biotransformation of (±)-linalool was investigated by screening 19 fungi. Product accumulation was enhanced by substrate feeding and, for the first time, lilac aldehydes and lilac alcohols were identified as fungal biotransformation byproduct using SPME-GC-MS headspace analysis. Aspergillus niger DSM 821, Botrytis cinerea 5901/02, and B. cinerea 02/FBII/2.1 produced different isomers of lilac aldehyde and lilac alcohol from linalool via 8-hydroxylinalool as postulated intermediate. Linalool oxides and 8-hydroxylinalool were the major products of fungal (±)-linalool biotransformations. Furanoid trans-(2R,5R)- and cis-(2S,5R)-linalool oxide as well as pyranoid trans-(2R,5S)- and cis-(2S, 5S)-linalool oxide were identified as the main stereoisomers with (3S,6S)-6,7-epoxylinalool and (3R,6S)-6,7-epoxylinalool as postulated key intermediates of fungal (±)-linalool oxyfunctionalization, respectively. With a conversion yield close to 100% and a productivity of 120 mg/L·day linalool oxides, Corynespora cassiicola DSM 62485 was identified as a novel highly stereoselective linalool transforming biocatalyst showing the highest productivity reported so far.

Full text of DOI:10.1021/jf800099h

J. Agric. Food Chem. 2008, 56, 3287–3296 3287
Fungal Biotransformation of (()-Linalool
†
§
#
¨
MARCO-ANTONIO MIRATA, MATTHIAS WUST, ARMIN MOSANDL, AND
JENS SCHRADER*,†
DECHEMA e.V., Karl-Winnacker-Institut, Biochemical Engineering, P.O. Box 150104,
D-60061 Frankfurt/Main, Germany; Department of Life Technologies, University of Applied Sciences
Valais, Route du Rawyl 47, CH-1950 Sion 2, Switzerland; and Institut für Lebensmittelchemie,
Johann Wolfgang Goethe-Universität, Max-von-Laue-Strasse 9, D-60438 Frankfurt/Main, Germany
The biotransformation of (()-linalool was investigated by screening 19 fungi. Product accumulation
was enhanced by substrate feeding and, for the first time, lilac aldehydes and lilac alcohols were
identified as fungal biotransformation byproduct using SPME-GC-MS headspace analysis. Aspergillus
niger DSM 821, Botrytis cinerea 5901/02, and B. cinerea 02/FBII/2.1 produced different isomers of
lilac aldehyde and lilac alcohol from linalool via 8-hydroxylinalool as postulated intermediate. Linalool
oxides and 8-hydroxylinalool were the major products of fungal (()-linalool biotransformations.
Furanoid trans-(2R,5R)- and cis-(2S,5R)-linalool oxide as well as pyranoid trans-(2R,5S)- and cis-
(2S, 5S)-linalool oxide were identified as the main stereoisomers with (3S,6S)-6,7-epoxylinalool and
(3R,6S)-6,7-epoxylinalool as postulated key intermediates of fungal (()-linalool oxyfunctionalization,
respectively. With a conversion yield close to 100% and a productivity of 120 mg/L ·day linalool oxides,
Corynespora cassiicola DSM 62485 was identified as a novel highly stereoselective linalool
transforming biocatalyst showing the highest productivity reported so far.
KEYWORDS: Biotransformation; fungi; lilac aldehyde; lilac alcohol; linalool; 8-hydroxylinalool; 6,7-
epoxylinalool; linalool oxide; SPME
INTRODUCTION
biotransformation, which has been well documented during the
past decades (6–9). The motivation for these research activities
was not only to better understand the biochemical pathways of
microbial terpene transformation and degradation but also to
find alternative synthetic routes toward highly valuable industri-
ally relevant flavor and fragrance compounds from cheap natural
precursors. Linalool occurs as one of its enantiomers in many
essential oils, for example, 80–85% (-)-linalool in Cinnamo-
mum camphora oil or 60–70% (+)-linalool in coriander oil (10),
thus representing a precursor that is abundantly available in
nature.
Lilac aldehydes and lilac alcohols have been described as
characteristic monoterpenoids in Syringa Vulgaris L. (Oleraceae)
flowers positively influencing the lilac odor quality (1). Lilac-
type fragrance compounds are in high demand by the perfume
industry because there are no natural lilac flower oils or
concentrates commercially available and synthetic fragrance
compounds are used to imitate the desired odor. Lilac aldehydes
and lilac alcohols are very powerful fragrance compounds due
to their exceptionally low odor thresholds of about 0.2–0.4 and
2–4 ng, respectively (2). Recently, the biogenetic pathway in
S.Vulgariswaselucidatedusing2H-and18O-labeledprecursors(3,4).
It was shown that the formation of the genuine lilac aldehydes
and alcohols proceeds via linalool (the S-enantiomer) as key
intermediate. By the same approach the elucidation of the
chirality and biosynthesis of lilac compounds in Actinidia arguta
flowers have delivered comparable conclusions (5). This raised
the question of whether other biological systems, namely,
microorganisms, were able to convert linalool into the desired
fragrance compounds as well. Among microorganisms fungi
especially are well-known for their versatility in monoterpene
Aspergillus niger DSM 821 and ATCC 9142 were shown to
convert (()-linalool into a mixture of cis- and trans-furanoid
linalool oxide and cis- and trans-pyranoid linalool oxide (8, 11).
Linalool oxide isomers are also valuable aroma compounds
constituting the flavor of tea and being used in the perfume
industry for lavender notes (12, 13).
Another research group found 8-hydroxylinalool as the main
linalool derivative after biotransformation of linalyl acetate by
A. niger (14). The biotransformation of linalool by the plant
pathogenic fungus Botrytis cinerea, which is responsible for
the noble rot of wine grapes, also yielded 90% 8-hydroxylinalool
as the main metabolite, whereas linalool oxides were only minor
products (15). Corynespora cassiicola DSM 62475 turned out
to be a very efficient and selective monoterpene transforming
microorganism, for example, producing pure (1S,2S,4R)-p-
menth-8-ene-1,2-diol from (+)-limonene in high concentrations
* Corresponding author (telephone 0049-69-7564 422; fax 0049-
69-7564 388; e-mail schrader@dechema.de).
† DECHEMA e.V.
§ University of Applied Sciences Valais.
# Johann Wolfgang Goethe-Universität.
10.1021/jf800099h CCC: $40.75  2008 American Chemical Society
Published on Web 04/22/2008

3288 J. Agric. Food Chem., Vol. 56, No. 9, 2008
Mirata et al.
an agar plate with a loop. The increasing concentrations of linalool in
each vial, adjusted with a solution of 0.3% (w/v) (()-linalool in EtOH,
were (in mg/L) 0, 25, 50, 75, 100, 150, 200, 300, 400, 500, 600, and
1000. The vials were covered with cotton stoppers and incubated for
4 days at 220 rpm and 25 °C (higher fungi) or 30 °C (yeasts). To
determine the concentration-dependent toxicity of linalool, the final dry
biomass was analyzed for each vial.
Screening of 19 Fungi in Small-Scale Vials. The screening
experiments were performed in 40 mL SPME vials filled with 15 mL
of MYB medium, pH 6.4, as described above. After autoclaving and
inoculation, 150 µL of 0.3% (w/v) (()-linalool (in ethanol) were added
(30 mg/L), the vials were covered with cotton stoppers and the cultures
were incubated for 14 days at 220 rpm and 25 °C (filamentous fungi)
or 30 °C (yeasts).
Biotransformation with Selected Strains in Erlenmeyer Flasks:
Batch Mode. To study the biotransformation kinetics, the selected
strains were cultivated in 2000 mL Erlenmeyer flasks for 12 days at
130 rpm and 25 °C (filamentous fungi) or 30 °C (yeasts). The flasks
were filled with 450 mL of MYB medium. Inoculation was done with
50 mL of a preculture grown in MYB medium with 500 µL of spore
suspension for the filamentous fungi and by loop for the yeast cultures.
At the beginning of the biotransformations, the precursor was added
according to the maximum concentration tolerated by the respective
strain identified by the toxicity determination experiments. (()-Linalool
was added as ethanolic solution [3% (w/v)]. Concentrations of linalool,
glucose, and dry biomass were determined every day from 5 mL
samples aseptically withdrawn from the culture. The pH values were
measured at the beginning and at the end of the experiments.
Biotransformation with Selected Strains in Erlenmeyer Flasks:
Feed-Batch Mode. The experimental setup and the analytics were
similar to those of the batch mode. To improve the final product
concentrations a feed-batch approach was chosen by feeding additional
linalool and glucose at intervals according to the strain-specific
consumption rates determined in the preceding batch-mode studies.
Glucose was dosed as a 1.1 kg/L aqueous solution. Dependent on the
metabolic activities of the strains, cultivations lasted 3-9 days. Product
analysis was done by headspace SPME-GC-MS after transferring 15
mL into 40 mL SPME vials at the end of cultivation.
(10); surprisingly, its use for linalool biotransformation has not
been reported so far. Besides Aspergillus, Botrytis, and Corynespo-
ra, many other genera, such as Saccharomyces, Penicillium, and
Geotrichum, are also known to include strains efficient in
monoterpene biotransformation (8, 11, 16–18).
Although the biotransformation of linalool has been exten-
sively studied in the past, neither lilac aldehyde nor lilac alcohol
formation has been reported so far.
In this work we systematically investigated whether certain
fungi, most of them preselected from the literature due to their
reported versatility to metabolize monoterpenes, are capable of
converting linalool into the desired lilac aldehydes and alcohols.
In a first screening run 19 strains were tested with a low linalool
concentration to avoid any toxic effects of the precursor. Seven
potentially positive strains were identified on the basis of
headspace solid-phase microextraction and GC-MS analysis
(SPME-GC-MS) after batch cultivation. Of these selected strains
the linalool tolerance thresholds and the glucose/linalool
consumption kinetics were determined to enhance biomass and
product formation during subsequent feed-batch cultivation. For
three strains, SPME-GC-MS analysis revealed the unprecedented
formation of lilac aldehydes and lilac alcohols as metabolic
byproducts of fungal linalool biotransformation. The target
products were identified by comparing their mass spectra and
retention indices with those of chemically synthesized reference
substances. Moreover, linalool oxides and 8-hydroxylinalool
turned out to be the major biotransformation products, and
analysis of their stereochemistry revealed high selectivities of
the biocatalytic reactions. The feed-batch strategy led to linalool
biotransformations with the highest product concentrations
reported so far.
MATERIALS AND METHODS
Microorganisms and Strain Maintenance. B. cinerea 5901/2, 5909/
1, 92/lic/1, 97/4, 99/16/3, 00/II10.1, 02/FBII/2.1, and P10 were kindly
provided by LWG, Bayerische Landesanstalt für Weinbau and Gar-
tenbau, Veitshöchheim, Germany. A. niger ATCC 16404, DSM 821,
C. cassiicola DSM 62475, Penicillium digitatum DSM 62840, and
Penicillium italicum DSM 62846 were purchased from DSMZ, Braun-
schweig, Germany. P. digitatum NRRL 1202 was obtained from ARS
Culture Collection, Illinois, USA, Geotrichum candidum was obtained
from HEVs, Hochschule Wallis, Sitten, Switzerland. Saccharomyces
cereVisiae Ceppo 20, Zymaflor VL1, Uvaferm 228, and SIHA Riesling
7 were a gift from E. Begerow GmbH & Co., Langenlonsheim,
Germany. The strains were grown on agar plates with malt extract agar
(MEA) consisting in w/v of malt extract 3%, soy peptone 0.3%, and
agar 1.7%, adjusted to pH 5.6 (10). Filamentous fungi were grown at
25 °C and yeasts at 30 °C.
Chemicals. (()-Linalool [>97% (v/v)], (-)-linalool [>98.5% (v/
v)], 1-octanol [>99.5% (v/v)], cis- and trans-furanoid linalool oxide
[>97% (v/v), mixture of isomers], and tert-butyl methyl ether (MTBE)
[>99.8% (v/v)[ were purchased from Fluka, Ulm, Germany. Lilac
alcohol and lilac aldehyde were prepared as previously described (2)
and were used as a mixture of stereoisomers [0.01% (v/v) in MTBE].
8-Hydroxylinalool was synthesized from (()-linalool according to the
literature (19) and was used as 0.02% (v/v) in MTBE. Standards of
cis- and trans-furanoid and pyranoid linalool oxide isomers [used as 0.01%
(v/v) in MTBE] were prepared as described elsewhere (13, 20).
Solid-Phase Microextraction Method. For SPME analysis, the pH
of a sample was adjusted to 4.0 with 1 M HCl if necessary, and 25%
of NaCl (3.75 g in 15 mL) was added. The SPME vials were covered
with PTFE silicone and “open top phenolics closures” (Supelco) and
incubated at 40 °C and 400 rpm until headspace SPME using a 75 µm
CAR-PDMS (carboxene/polydimethylsiloxane) coated fiber and a
manual holder (Supelco, Germany) was started. Extraction was carried
out for 20 min at 40 °C and 400 rpm. For GC-MS analysis a desorption
time of 5 min in a GC injector at 250 °C was used.
Analysis of the Sample with GC-MS and Enantioselective GC.
GC-MS analyses were performed with a GC-17 A Shimadzu gas
chromatograph, equipped with a VB-5 Valcobond column (30 m ×
0.25 mm i.d.; coating thickness ) 0.25 µm) and a QP5050 Shimadzu
mass spectrometer (quadrupole type). The working conditions were as
follows: injector, 250 °C; detector, 280 °C; oven temperature, start at
40 °C, hold for 7 min, programmed from 40 to 280 °C at 10 °C/min,
hold for 2 min; carrier gas flow (He), 1.1 mL/min; injection mode,
splitless for SPME samples and 1/30 for liquid samples (injection
volume ) 2 µL); EI, 70 eV; acquisition parameters, scanned m/z,
35–250 (10–25 min). Enantioselective GC analysis was performed with
a GC-17 A Shimadzu gas chromatograph with FID analyzer, equipped
with a Chiraldex B-DM chiral column (30 m × 0.32 mm i.d.; coating
thickness ) 0.12 µm). The working conditions were as follows: injector,
250 °C; detector, 280 °C; oven temperature, isothermal for 30 min at
95 °C; carrier gas flow (He), 1.1 mL/min; injection mode, split 1/20
for liquid samples (injection volume ) 1 µL). Relevant substances were
identified by comparison of their mass spectra and retention indexes
(Kovats indices) (11, 13, 16, 20–23) with those of reference substances
(when possible) and by comparison with MS library data (NIST mass
spectral library V 2.0). Retention indices (for VB-5 Valcobond column)
were as follows: trans-furanoid linalool oxide, 1082; cis-furanoid
linalool oxide, 1096; linalool, 1115; lilac aldehyde isomer a, 1150; lilac
aldehyde isomer b, 1159; lilac aldehyde isomer c, 1173; trans-pyranoid
Determination of Linalool Toxicity. To determine the toxicity of
linalool, each of the seven fungi selected by the screening was cultivated
in small liquid cultures with increasing concentrations of linalool. For
each strain, 12 40 mL SPME vials were filled with 15 mL of MYB
medium (in w/v, malt extract 3%, glucose 1%, peptone 1%, and yeast
extract 0.3%; pH 6.4), autoclaved, and inoculated with 500 µL of a
spore suspension consisting of (in w/v) 0.85% NaCl, 1% peptone, 0.1%
Tween 80, and approximately 2.5 × 107 spores in distilled water; yeast
cultures of the same volume were inoculated by transferring cells from

Linalool Biotransformation
J. Agric. Food Chem., Vol. 56, No. 9, 2008 3289
linalool oxide, 1188; cis-pyranoid linalool oxide, 1195; lilac alcohol
isomer d, 1213; lilac alcohol isomer e, 1230; lilac alcohol isomer f,
1247; 10-hydroxylinalool, 1385; 8-hydroxylinalool, 1399. Retention
indices (for Chiraldex B-DM chiral column) were as follows: furanoid
linalool oxides, trans-(2R,5R), 1149; trans-(2S,5S), 1160; cis-(2R,5S),
1163; cis-(2S,5R), 1169; R-(-)-linalool, 1209; S-(+)-linalool, 1216;
pyranoid linalool oxides, trans-(2S,5R), 1271; trans-(2R,5S), 1284; cis-
(2S,5S), 1290; cis-(2R,5R), 1293. Response factors in relation to
1-octanol were as follows: cis- and trans-furanoid linalool oxide, 1.27;
cis- and trans-pyranoid linalool oxide, 1.32; (()-linalool, 1.19; 8-hy-
droxylinalool, 2.80.
Mass Spectra. trans-Furanoid linalool oxide m/z (relative intensity):
41 (34), 43 (71), 55 (38), 59 (100), 67 (24), 68 (25), 81 (12), 93 (21),
94 (27), 111 (13), 137 (3), 155 (2). cis-Furanoid linalool oxide m/z
(relative intensity): 41 (35), 43 (63), 55 (43), 59 (100), 67 (26), 68
(31), 81 (15), 93 (28), 94 (39), 111 (19), 137 (3), 155 (3). trans-Pyranoid
linalool oxide m/z (relative intensity): 41 (51), 43 (88), 53 (24), 55
(38), 57 (15), 59 (99), 67 (70), 68 (100), 79 (27), 94 (66), 137 (1), 155
(2). cis-Pyranoid linalool oxide m/z (relative intensity): 41 (34), 43 (64),
53 (16), 55 (24), 57 (10), 59 (89), 67 (54), 68 (100), 79 (20), 94 (53),
137 (2), 155 (3). Linalool m/z (relative intensity): 41 (97), 43 (100),
53 (15), 55 (64), 67 (20), 69 (38), 71 (89), 80 (31), 93 (55), 121 (12),
136 (5), 154 (<1, M+). Lilac aldehyde isomer a m/z (relative intensity):
41 (63), 43 (99), 55 (100), 67 (40), 69 (28), 71 (33), 81 (16), 91 (9),
93 (29), 111 (23), 125 (3), 153 (6). Lilac aldehyde isomer b m/z (relative
intensity): 41 (58), 43 (99), 55 (100), 67 (35), 69 (24) 71 (37), 81 (17),
91 (8), 93 (28), 111 (20), 125 (2), 153 (10). Lilac aldehyde isomer c
m/z (relative intensity): 41 (47), 43 (69), 55 (100), 67 (27), 69 (20), 71
(39), 81 (13), 91 (6), 93 (24), 111 (15), 125 (4), 153 (7). Lilac alcohol
isomer d m/z (relative intensity): 41 (58), 43 (100), 55 (90), 67 (40),
69 (23), 71 (28), 81 (15), 91 (11), 93 (56), 111 (69), 125 (5), 155 (6).
Lilac alcohol isomer e m/z (relative intensity): 41 (53), 43 (100), 55
(98), 67 (45), 69 (24), 71 (29), 81 (20), 91 (9), 93 (52), 111 (58), 125
(3), 155 (7). Lilac alcohol isomer f m/z (relative intensity): 41 (53), 43
(85), 55 (100), 67 (40), 69 (22), 71 (28), 81 (17), 91 (9), 93 (53), 111
(55), 125 (2), 155 (8). 10-Hydroxylinalool m/z (relative intensity): 41
(36), 43 (100), 53 (10), 55 (32), 67 (40), 68 (19), 71 (60), 79 (12), 81
(9), 93 (7), 137 (3), 152 (1). 8-Hydroxylinalool m/z (relative intensity):
41 (34), 43 (100), 53 (10), 55 (29), 67 (40), 68 (18), 71 (47), 79 (11),
81 (9), 93 (8), 137 (3), 152 (<1).
Sample Preparation. To determine the concentrations of linalool,
cis- and trans-furanoid linalool oxide, cis- and trans-pyranoid linalool
oxide, and 8-hydroxylinalool, 2 mL of liquid culture was filtered through
a 0.45 µm/25 mm nylon filter (Macherey-Nagel) and extracted with 2
mL of tert-butyl methyl ether (MTBE) prior to GC-MS. 1-Octanol was
used as internal standard for quantification. To determine the enantio-
meric distribution of linalool oxides, samples were analyzed via
enantioselective GC instead of GC-MS. All measurements were done
in triplicate.
Target Compound Identification. For the initial screening of 19
fungi, the occurrence of one fragment ion (m/z 153 for lilac aldehyde
or m/z 155 for lilac alcohol or m/z 111 for lilac aldehyde/alcohol) in a
time window of (0.5 min to the retention times of the chemically
synthesized standards was used as the criterion for transferring a strain
into the second, more detailed, screening run. For this second screening,
the target compounds, lilac aldehyde and lilac alcohol, were identified
by comparison with chemically synthesized standards on the basis of
their mass spectra and retention indexes. Selected ion monitoring (SIM)
using the following masses was applied: lilac aldehyde, m/z 153 (120×);
lilac alcohol, m/z 155 (120×); and lilac aldehyde/alcohol, m/z 111
(30×).
Figure 1. Linalool concentration-toxicity curve for B. cinerea 5901/2.
The next lowest concentration above the threshold concentration, above
which a sharp decrease in cell growth occurred, was chosen as maximum
precursor concentration, here 150 mg/L. The other strains showed
essentially the same toxicity profiles but varied concerning the maximum
concentration values, which are listed in Table 1.
initial linalool concentration of 200 mg/L. The pH of the MYB broth
was adjusted to pH 5 with HOAc. The concentration of the substrate
was analyzed every 2 days and finally after 9 days. To exclude de novo
biosynthesis of the target compounds, the selected strains were
cultivated under biotransformation conditions but without linalool.
RESULTS AND DISCUSSION
Initial Screening of Fungi, Batch Cultivation. Nineteen
fungal strains were screened for their capability to convert
linalool particularly into lilac aldehyde and lilac alcohol using
SPME-GC-MS as the monitoring technique. The strains were
grown as 15 mL liquid cultures in 40 mL SPME vials enabling
direct probing after 14 days of cultivation. The addition of 25%
(w/v) NaCl increased the analytical sensitivity by a factor of
10 due to the enhanced volatility of the target compounds (data
not shown). Potentially lilac aldehyde/alcohol-positive strains
were selected by searching time windows of (0.5 min around
the retention times of the reference substances for the occurrence
of at least one of the characteristic fragmentation ions m/z 155
(alcohol), m/z 153 (aldehyde), and m/z 111 (alcohol and
aldehyde). Due to the fact that only a relatively low concentra-
tion of the toxic precursor was chosen for this orienting
screening (30 mg/L), only low product concentrations near the
detection limit occurred, and the SPME-GC-MS analyses were
evaluated as indicative results: even those strains not giving
unambiguously positive results were chosen for a second
screening run under improved cultivation conditions. The seven
potentially positive strains were A. niger ATCC 16404, A. niger
DSM 821, B. cinerea 5901/2, B. cinerea 02/FBII/2.1, C.
cassiicola DSM 62475, S. cereVisiae Zymalor VL1, and S.
cereVisiae Uvaferm 228. To verify the preliminary results, we
aimed at developing optimized feed-batch cultivation in Erlen-
meyer flasks sequentially providing additional precursor and
glucose while avoiding toxic precursor concentrations. By this
means the final concentrations of linalool biotransformation
products were to raised, thus enabling us to identify not only
the major biotransformation products but also the lilac com-
pounds as metabolic byproduct.
Dry Biomass and Glucose Determination. The dry biomass was
determined gravimetrically using an infrared moisture analyzer (MA
100, Sartorius, Germany) by filtering 2 mL of homogenized liquid
culture through a dried, preweighed 0.45 µm/45 mm cellulose acetate
membrane filter (Schleicher & Schuell, Germany). The concentration
of glucose was determined enzymatically (YSI 2700 Biochemistry
Analyzer, Yellow Springs Instruments, Yellow Springs, OH).
Chemical and Biological Control Experiments. The chemical
stability of the substrate (()-linalool was verified under cultivation
conditions during a period of 9 days without inoculum and with an
Linalool Toxicity. Strain-specific linalool concentration-
toxicity profiles were determined by measuring the final cell
dry weight after incubation of the fungi for 4 days in the
presence of increasing concentrations of linalool. Figure 1
exemplarily illustrates the toxic effects of linalool on B. cinerea
5901/2. Due to an observed sharp decrease in viability at a

3290 J. Agric. Food Chem., Vol. 56, No. 9, 2008
Mirata et al.
Table 1. Feeding Strategies for Improved Biomass/Product Formation
during Cultivation of the Preselected Strains under Feed-Batch Conditions
feeding strategya
start concn
of linaloolb
(mg/L)
feed
feed
glucose
(g/L)
∆t
feed (h)
linalool
(mg/L)
strain
A. niger ATCC 16404
A. niger DSM 821
B. cinerea 5901/2
B. cinerea 02/FB II/ 2.1
C. cassiicola DSM 62475
S. cerevisiae Zymaflor VL1
S. cerevisiae Uvaferm 228
100
150
150
100
150
200
50
72
48
72
24
24
48
72
50
100
80
80
100
0
5
8
2
0
5
10
10
0
a The feed-batch experiments were carried out in a MYB medium over 12 days.
The start concentration of glucose was 10 g/L. Feeding intervals were derived
from the studies of the glucose and precursor consumption kinetics, cf. Figure 2.
b Derived from the toxicity studies, cf. Figure 1.
Figure 2. Linalool and glucose consumption kinetics: glucose (9) and
linalool (2) consumption by B. cinerea 5901/2; glucose (0) and linalool
(4) consumption by C. cassiicola DSM 62475.
medium were favored for initial biomass growth. Total glucose
depletion occurred after 8 days. In contrast, C. cassiicola DSM
62475 showed fast glucose catabolism starting immediately after
inoculation and leading to total consumption in just 2 days
(Figure 2). This was accompanied by stringent linalool con-
sumption resulting in almost total depletion after about 1 day.
B. cinerea 02/FBII/2.1 and A. niger ATCC 16404 showed
glucose consumption profiles comparable to that of B. cinerea
5901/2, whereas the linalool consumption kinetics of C.
cassiicola DSM 62475 was 2-fold faster than those of A. niger
DSM 821 and A. niger ATCC 16404. B. cinerea 02/FBII/2.1
depleted linalool at a rate comparable with that of C. cassiicola
DSM 62475 (data not shown). Both yeast strains tested also
quickly consumed the glucose within 1 day, but after 6 days S.
cereVisiae Uvaferm 228 had consumed only about half of the
50 mg/L linalool added, whereas S. cereVisiae Zymaflor VL1
had metabolized about 110 mg/L of the 200 mg/L added (data
not shown). These data clearly indicate that glucose and
precursor feeding strategies have to be carefully adapted strain-
specifically to ensure maximum product formation under feed-
batch conditions. The strain-specific feeding parameters derived
from these batch studies are summarized in Table 1.
Screening of Selected Fungi, Feed-Batch Cultivation. On
the basis of the data obtained during the toxicity and glucose/
linalool consumption studies (cf. Table 1) the selected micro-
organisms were grown with strain-specific glucose and linalool
feedings. By this means extended biomass growth and thus
higher biotransformation product concentrations were targeted
to facilitate the identification of the metabolites, particularly the
lilac fragrance compounds, by headspace SPME-GC-MS analy-
sis. Table 2 exemplarily illustrates the relative contribution to
the total peak area in percent of the biotransformation products
obtained from linalool as well as other fungal metabolites and
nonconverted linalool. The chemical control test was carried
out without inoculum under otherwise identical conditions.
Among the seven strains tested A. niger DSM 821, B. cinerea
5901/02, and B. cinerea 02/FII/2.1 were undoubtedly able to
convert linalool into minor quantities of lilac aldehyde isomers
and lilac alcohol isomers, whereas the other strains did not
produce detectable amounts of the target compounds (Table
2). The identification of the lilac compounds is described in
the next chapter. The furanoid cis/trans-linalool oxide stereoi-
somers and the pyranoid cis/trans-linalool oxide stereoisomers
were the major products of linalool biotransformation by A.
niger DSM 821, B. cinerea 02/FB II/2.1, and C. cassiicola DSM
62475. The capacity of A. niger and B. cinerea to oxidize
linalool to linalool oxide stereoisomers as well as to stereose-
lectively form the 8-hydroxylinalool diastereoisomer has already
certain threshold concentration, the next lower concentration
tested was chosen as the maximum precursor concentration for
the following feed-batch cultivation experiments to ensure a
reasonable precursor supply while avoiding any toxic effects.
In the case of B. cinerea 5901/2 a linalool concentration of 150
mg/L was identified as the maximum value at which growth
was still unhampered by the presence of the monoterpene. The
other strains showed essentially the same toxicity profiles but
varied with regard to the maximum precursor concentration
values, which are listed in Table 1. The toxic effects of
monoterpenoids are preferentially attributed to their intercalation
into the cell membrane, thereby affecting its physiological
function, rather than by specific interactions with defined targets
(24). This general mechanism may explain why all threshold
concentrations found in our studies were of the same order of
magnitude independent of the fungal strain investigated. The
low threshold concentrations are in good agreement with the
log P concept (logarithm of the partition coefficient of an organic
compound in the two-phase system 1-octanol/water) whereby
lipophilic compounds with a log P between 1 and 5 are usually
toxic to microorganisms (25). The log P of linalool can be
calculated to be approximately 2.8 (http://www.logp.com),
which is lower than those of monoterpene hydrocarbons due to
the hydrophilic character introduced into the molecule by the
alcohol group. On the basis of an empirical equation (26), the
linalool concentration in the cell membrane, which corresponds
to an aqueous solution of 1 mM linalool (154 mg/L), can be
estimated to be about 120 mM (approximately 18 g/L). This
illustrates the enormous linalool accumulation in the membrane
and makes its detrimental effect on the cell physiology
understandable.
Glucose and Precursor Consumption Kinetics. The glucose
and linalool consumption kinetics were investigated under batch
conditions to define a specific protocol for a subsequent feed-
batch cultivation for each of the seven preselected strains. Figure
2 exemplarily illustrates the results with B. cinerea 5901/2 and
C. cassiicola DSM 62475, indicating that glucose and precursor
consumption kinetics are highly strain-dependent. B. cinerea
5901/2 quickly consumed the linalool initially added to the
culture. After 4 days of cultivation in Erlenmeyer flasks, almost
85% of the linalool was metabolized, followed by a less
pronounced decline resulting in almost complete linalool
consumption after 9 days. The amount of linalool that evaporated
or was chemically degraded during the same time was e3.2%,
as determined by a control experiment (data not shown). The
consumption of glucose did not begin until 2 days after
inoculation; obviously other C sources from the complex

Linalool Biotransformation
J. Agric. Food Chem., Vol. 56, No. 9, 2008 3291
Table 2. Relative Contribution (Percent) of Linalool Biotransformation Products, Nonconverted Linalool, and Other Fungal Metabolites to the Total Peak Area
after GC-MS Analysis of Headspace SPME Extracts of Liquid Culturesa
A. niger
B. cinerea
S. cerevisiae
C. cassiicola
no.
compound
AN1
AN2
BC1
BC2
SC1
SC2
CC1
control
1
2
3
4
5
6
7
8
methylheptenone
ꢀ-myrcene
methyheptenol
2,6-dimethyl-1,6-octadiene
limonene
ocimene
trans-linalool oxide furanoid
cis-linalool oxide furanoid
(()-linalool
2-phenylethanol
dihydrolinalool
lilac aldehyde (isomer a)
lilac aldehyde (isomer b)
lilac aldehyde (isomer c)
dihydrocarvone
trans-linalool oxide pyranoid
cis-linalool oxide pyranoid
R-terpineol
lilac alcohol (isomer d)
lilac alcohol (isomer e)
lilac alcohol (isomer f)
1-p-menthen-9-al
nerol
2.9
7.2
nd
nd
0.93
0.22
1.3
1.1
70
2.0
5.6
nd
nd
<0.1
nd
nd
0.46
nd
nd
32
26
23
2.7
0.65
0.12
<0.1
<0.1
nd
4.8
8.5
0.40
<0.1
nd
<0.1
0.29
nd
nd
nd
nd
2.9
nd
nd
0.83
1.2
0.67
0.87
80
nd
3.0
nd
nd
2.6
nd
18
23
48
0.16
0.34
nd
<0.1
nd
nd
6.1
nd
nd
4.1
nd
nd
0.19
72
6.6
8.8
nd
nd
nd
nd
<0.1
nd
0.24
nd
nd
nd
nd
4.9
nd
5.4
nd
0.28
nd
nd
nd
44
34
7.7
0.87
0.70
nd
nd
nd
nd
7.1
nd
nd
nd
2.0
0.46
0.25
0.36
67
15
6.6
nd
nd
nd
nd
nd
nd
0.37
nd
nd
nd
nd
nd
0.21
nd
nd
nd
nd
nd
nd
1.3
0.65
1.7
0.91
nd
0.20
76
nd
10
nd
nd
nd
nd
nd
0.31
<0.1
nd
nd
nd
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
3.0
7.3
nd
0.11
<0.1
nd
0.25
0.10
0.30
<0.1
<0.1
nd
<0.1
<0.1
nd
0.13
nd
0.17
nd
<0.1
0.73
0.19
0.84
nd
<0.1
0.68
<0.1
4.5
nd
nd
nd
nd
1.0
3.6
0.14
nd
<0.1
nd
nd
nd
nd
nd
nd
nd
nd
nd
0.30
nd
0.22
1.8
4.8
nd
nd
nd
nd
nd
nd
nd
nd
nd
nd
nd
nd
nd
nd
0.31
nd
0.33
0.32
nd
0.33
<0.1
<0.1
<0.1
<0.1
0.44
1.7
nd
nd
nd
0.34
1.2
0.87
nd
nd
nd
nd
nd
nd
0.65
citronellol
geraniol
citral
geranylacetone
dihydrocarveol
10-hydroxylinalool
8-hydroxylinalool
other fungal metabolites
SPME artifact
0.24
0.27
0.10
<0.1
nd
nd
nd
0.47
0.48
nd
<0.1
<0.1
nd
0.10
0.45
0.55
0.53
a AN1, Aspergillus niger ATCC 16404; AN2, A. niger DSM 821; BC1, Botrytis cinerea 5901/2; BC2, B. cinerea 02/FB II/2.1; SC1, Saccharomyces cerevisiae Zymaflor
VL1; SC2, S. cerevisiae Uvaferm 228; CC1, Corynespora cassiicola DSM 62475; control, chemical control experiment; nd, not detected.
been reported (9, 11, 14, 27). However, with this study we
demonstrate for the first time that C. cassiicola DSM 62475,
too, is capable of producing linalool oxides from linalool;
furthermore, this proceeds in a highly selective and efficient
way (more details are given under Quantification of the Major
Biotransformation Products). The presence of methylheptenone
and methylheptenol in cultures of A. niger ATCC 16404, A.
niger DSM 821, and C. cassiicola DSM 62475 most probably
resulted from the biotransformation of citronellol, geraniol, and
nerol, which were found as impurities in the substrate (17, 28).
Small amounts of R-terpineol in some cultivation broths resulted
from either chemical conversion of linalool and citronellol (15, 28)
or biotransformation by A. niger and S. cereVisiae (11, 18). The
chemical conversion of (()-linalool (purity g 95%) in acidified
MYB liquid broth (pH 5.0) was checked by SPME-GC-MS (see
Table 2, control). After 9 days of shaking >95% of the
precursor, linalool was recovered in the control experiment, as
determined by organic solvent extraction with MTBE and
1-octanol as internal standard, indicating low chemical reactivity
under the biotransformation conditions. Several compounds,
such as dihydrolinalool, ꢀ-myrcene, limonene, ocimene, linalool
oxide, R-terpineol, nerol, citronellol, geraniol, and citral were
found in the chemical control experiments (and also as impurities
of the precursor), indicating nonbiological conversion reac-
tions (11, 15). Nevertheless, the analytical data (Table 2) clearly
indicate that chemical linalool oxide formation was negligible
compared to the microbial transformation (cf. Table 3). Neither
lilac aldehyde, lilac alcohol, nor 8-hydroxylinalool was detected
in the chemical control test.
Identification of the Target Compounds Lilac Aldehyde
and Lilac Alcohol. Figure 3 exemplarily illustrates the product
peak identification after biotransformations with B. cinerea 5901/
02 and A. niger DSM 821 based on fragment ion chromato-
grams. The mixture of standard compounds contained all four
diastereoisomers of lilac aldehyde and lilac alcohol. For each
compound only three peaks were obtained after nonchiral gas
chromatographic separation due to a coelution of two diaste-
reoisomers (Figure 3C). By comparing the fragment ion
chromatograms and corresponding mass spectra at least two
isomers of both lilac aldehyde and lilac alcohol were detectable
at the positions indicated in the B. cinerea chromatogram
(Figure 3A) and at least three isomers of each compound in
the A. niger chromatogram (Figure 3B). The differences
between the retention times of the biologically produced lilac
aldehydes and alcohols and the standards were e0.08 min.
Besides, the authentication of the biological diastereoisomers
of lilac aldehyde and lilac alcohol was verified by comparing
the retention indices with those of the analogous synthetic
reference substances. The mass spectra of the target compounds
produced, exemplarily illustrated in Figure 4A,B, by one lilac
aldehyde and one lilac alcohol diastereoisomer showed the
characteristic fragmentation pattern and high agreement with
the reference compounds given in Figure 4C,D. The target
compounds accumulated as only minor components in the
medium (Table 2). In the biological control experiment under
the same conditions (with inoculum but without linalool) no
lilac compound was found.

3292 J. Agric. Food Chem., Vol. 56, No. 9, 2008
Mirata et al.
Table 3. Final Concentrations and Conversion Yields of Furanoid cis- and trans-Linalool Oxide, Pyranoid cis- and trans-Linalool Oxide, and 8-Hydroxylinalool
after Feed-Batch Biotransformation of (()-Linalool by Liquid Cultures of A. niger ATCC 16404 (AN1) (after 9 Days of Cultivation), A. niger DSM 821 (AN2)
(after 6 Days of Cultivation), B. cinerea 5901/2 (BC1) (after 9 Days fo Cultivation), B. cinerea 02/FB II/2.1 (BC2) (after 3 Days of Cultivation), and C.
cassiicola DSM 62475 (CC1) (after 3 Days of Cultivation)a
A. niger
B. cinerea
C. cassiicola
AN1
AN2
BC1
BC2
CC1
compound
concn (mg/L) yield (%) concn (mg/L) yield (%) concn (mg/L) yield (%) concn (mg/L) yield (%) concn (mg/L) yield (%)
furanoid trans-linalool oxide nd
furanoid cis-linalool oxide 9.3 ( 0.2
pyranoid trans-linalool oxide nd
nd
77.0 ( 0.9 24.0 ( 2.7
76.1 ( 1.8 23.7 ( 2.9
nd
nd
nd
nd
nd
nd
nd
nd
30.7 ( 0.4 13.4 ( 1.5 157 ( 3
26.3 ( 0.6 11.5 ( 1.4 148 ( 5
42.1 ( 5.6
39.5 ( 6.4
5.3 ( 0.7
nd
nd
22.4 ( 0.3
7.0 ( 0.8
5.7 ( 0.1
2.5 ( 0.3 18.7 ( 0.4 5.0 ( 0.7
6.1 ( 0.7 33.1 ( 0.5 8.9 ( 1.1
pyranoid cis-linalool oxide
8-hydroxylinalool
(()-linalool (recovered %)
nd
33.4 ( 0.5 10.4 ( 1.2
14.0 ( 0.2
37.2 ( 1.1 21.1 ( 2.5 45.7 ( 1.4 14.2 ( 1.9
56.3 ( 9.3 1.9 ( 0.4
167 ( 5
59.9 ( 7.7 39.3 ( 1.2 17.2 ( 2.3 nd
nd
1.4 ( 0.3
21.9 ( 3.5
12.8 ( 2.7
a Analyses were done in triplicate with 1-octanol as internal standard. The molar conversion yields were calculated from the total amount of linalool added to the culture.
The residual linalool not converted by the fungi is given as (()-linalool (recovered in percent).
trans-, cis-. This was determined by injection of the enantiopure
standards on a nonchiral column (VB-5, 5% phenyl/95%
dimethylsiloxane). A. niger DSM 821 (AN2) converted almost
80% of the administered substrate into 153 mg/L furanoid cis-
and trans-linalool oxides, 56 mg/L pyranoid cis- and trans-
linalool oxides, and 46 mg/L 8-hydroxylinalool after 6 days of
feed-batch cultivation (Table 3). We chose a specific glucose
and linalool feeding to avoid any inhibitory effect by excessive
substrate addition. This may be the main reason we were able
to at least duplicate the conversion yield by A. niger DSM 821
compared to the literature (16). Biotransformation by A. niger
ATCC 16404 (AN1) was significantly less pronounced (despite
comparable biomass growth). B. cinerea 5901/2 (BC1) con-
verted about 60% of the dosed linalool to 8-hydroxylinalool as
the main product (Table 3), whereas linalool oxide isomers and
10-hydroxylinalool were detected by SPME analysis as minor
products (cf. Table 2). This is in accordance with the observa-
tions by Bock et al., who found hydroxylated linalools and
linalool oxides among an array of metabolites of B. cinerea
5901/2, whereas linalool was predominately metabolized to
8-hydroxylinalool (>90%) (15). Interestingly, B. cinerea 02/
FB II/2.1 (BC2) produced a greater variety of metabolites with
39 mg/L 8-hydroxylinalool, 57 mg/L furanoid cis- and trans-
linalool oxides, and 20 mg/L pyranoid cis- and trans-linalool
oxides. Rapp and Mandry also identified a comparable mixture
of ω-hydroxylated and oxidized compounds after biotransfor-
mation of linalool by a self-isolated Botrytis strain (27). In
addition, 10-hydroxylinalool was also found as a byproduct of
linalool biotransformation in minor concentrations by both
groups (15, 27). Surprisingly, C. cassiicola DSM 62475 (CC1),
which had not yet been reported to be active toward linalool,
turned out to be the most actively transforming strain: it
consumed >96% of the 340 mg/L (()-linalool added and
produced 357 mg/L linalool oxide isomers in just 3 days of
feed-batch cultivation corresponding to a molar conversion yield
close to 100% when compared with the amount of linalool
consumed. The product formation occurred in a highly stereo-
selective way as described in the next section. Neither 8-hy-
droxylinalool nor other significant byproducts were detected
Figure 3. Fragment ion chromatograms (m/z 111) after SPME-GC-MS
analysis of the medium after 9 days of feed-batch cultivation (cf. Table
1): (A) B. cinerea 5901/2; (B) A. niger DSM 821; (C) chemically
synthesized lilac aldehyde and lilac alcohol reference substances.
(Table 3). Due to the fact that for all strains almost equivalent
diastereoisomeric ratios between cis- and trans-linalool occurred,
the substrate epoxidation does obviously not depend on the (()-
linalool configuration at the C-3 position (29). Figure 5
exemplarily illustrates the kinetics of main product formation
by C. cassiicola DSM 62475, A. niger DSM 821, and B. cinerea
5901/2. The productivity of C. cassiicola DSM 62475, expressed
as the sum of linalool conversion products accumulated during
the first 3 days, was 120 mg/L·day and, thus, almost 3 times
Quantification of the Major Biotransformation Products.
To determine the molar conversion yields for the major products,
linalool oxide diastereoisomers, 8-hydroxylinalool, and the
recovered linalool were quantified after feed-batch cultivation
of A. niger ATCC 16404, A. niger DSM 821, B. cinerea 5901/
2, B. cinerea 02/FB II/2.1, and C. cassiicola DSM 62475 (Table
3). The elution order of the linalool oxide diastereoisomers on
the nonchiral column was furanoid trans-, cis- and pyranoid

Linalool Biotransformation
J. Agric. Food Chem., Vol. 56, No. 9, 2008 3293
Figure 4. SPME-GC-MS pattern of target molecules lilac aldehyde and lilac alcohol: (A, B) Examples of target isomers produced by B. cinerea 5901/2
[lilac aldehyde at tr ) 15.692 min (peak b in Figure 3A); lilac alcohol at tr ) 16.667 min (peak e in Figure 3A)]; (C, D) mass spectra of the chemically
synthesized reference compounds [lilac aldehyde at tr ) 15.642 min (peak b in Figure 3C); lilac alcohol at tr ) 16.662 min (peak e in Figure 3C)].
cultivation of C. cassiicola on a bioreactor scale may thus be
an interesting option for the technical production of natural
linalool oxides, which are of interest to the fragrance industry
as lavender notes (30).
Analysis of the Stereoisomeric Distribution of Furanoid
and Pyranoid Linalool Oxides by Enantioselective GC. To
analyze the enantiomeric and diastereoisomeric distribution of
linalool oxides produced by A. niger DSM 821 (AN2), B.
cinerea 02/FB II/2.1 (BC2), and C. cassiicola DSM 62475
(CC1), the elution order on a DM-ꢀ-CD [(2,6-di-O-methyl)-ꢀ-
cyclodextrin chiral stationary phase] column was determined
by co-injection of enantiopure and racemic standards. The
elution order of furanoid linalool oxides was trans-(2R,5R),
trans-(2S,5S), cis-(2R,5S), cis-(2S,5R), as exemplarily illustrated
Figure 5. Formation kinetics of the main linalool biotransformation products
in Figure 6 for the analysis of the organic fraction of the A.
of C. cassiicola DSM 62475 (sum of linalool oxide diastereoisomers) (0),
niger DSM 821 culture. For pyranoid linalool oxides, the trans-
A. niger DSM 821 (sum of linalool oxide diastereoisomers and 8-hydroxy-
(2R,5S) and cis-(2S,5S) diastereoiomers coeluted on this chiral
linalool) (O), and B. cinerea 5901/2 (8-hydroxylinalool) (4). The arrows
phase (Figure 6, pyranoid linalool oxide, peaks a, b + c, d);
nevertheless, the correct elution order, which is trans-(2S,5R),
trans-(2R,5S), cis-(2S,5S), cis-(2R,5R), was solved by separate
injection of enantiopure standards. The stereoisomeric distribu-
tion of the coeluted isomers could be calculated from the
concentrations of the cis/trans diastereoisomers, which had been
determined beforehand (cf. Table 3). The same elution order
of linalool oxide stereoisomers was found with PME-ꢀ-CD
(permethyl-ꢀ-cyclodextrine) and PET-ꢀ-CD (perethyl-ꢀ-cyclo-
dextrine) as chiral stationary phases (11, 31). The pyranoid
trans-(2R,5S) and cis-(2S,5S) linalool oxide enantiomers were
presumably not separable due to a lower apolarity of DM-ꢀ-
CD chiral stationary phase compared to PME-ꢀ-CD and PET-
ꢀ-CD chiral stationary phases. It has been demonstrated that
the cyclization of linalool into the furanoid and pyranoid linalool
oxides can proceed by two distinct pathways. The diastereoi-
somers of 6,7-epoxylinalool, which are formed by epoxidation
of linalool at C6-C7 position, have been isolated from Carica
papaya fruit and proposed as biogenetic precursors for the
indicate time points of supplementation with linalool and glucose. QP )
average productivity.
higher than that of A. niger DSM 821. This observation can be
explained by a faster growth: at the end of the cultivation C.
cassiicola DSM 62475 had consumed >88% of the added
glucose (17 g/L) and built up a cell dry weight of 8.8 g/L; the
pH of the broth was 5.4. In contrast, A. niger DSM 821
consumed 8.6 g/L glucose accompanied by a cell dry weight
formation of 4.5 g/L and a pH of 4.3 at the end of the cultivation
period (data not shown). Compared with A. niger and C.
cassiicola, the productivity of B. cinerea remained low (20 mg/
L·day, Figure 5). For all strains, even higher product concen-
trations seemed to be achievable because stationary growth
phases had not yet been reached. Compared to other microbial
linalool biotransformations (11, 15, 16, 27, 29), C. cassiicola
DSM 62485 investigated in the present study showed a
remarkable capability of producing linalool oxides in high
conversion yields and with high selectivities. Submerged

3294 J. Agric. Food Chem., Vol. 56, No. 9, 2008
Mirata et al.
Figure 6. Enantioselective GC analysis of the furanoid and pyranoid linalool oxide stereoisomers from the organic fraction of A. niger DSM 821 (AN2)
culture after biotransformation of linalool.
Figure 7. Stereoisomeric distribution in percentage of furanoid and pyranoid linalool oxides after analysis of organic solvent extracts of the feed-batch
biotransformation broths of A. niger DSM 821 (AN2), B. cinerea 02/FB II/2.1 (BC2), and C. cassiicola DSM 62475 (CC1). One hundred percent furanoid
linalool oxides corresponds to a total product concentration of 153, 57, and 305 mg/L for AN2, BC2, and CC1, respectively. One hundred percent
pyranoid linalool oxides corresponds to a total product concentration of 56, 20, and 52 mg/L for AN2, BC2, and CC1, respectively. Above each series
of columns the corresponding linalool oxide structure is depicted with its postulated biosynthetic pathway from (3R)- and (3S)-linalool via the corresponding
6,7-epoxylinalool. The pathways predominantly followed by the fungi investigated are highlighted in bold. The key intermediates, the 6S-configured
6,7-epoxylinalools, are framed. Mean values of duplicate measurements, indicated by error bars, are given.
formation of furanoid and pyranoid linalool oxides (32). In juices
of Muscat grapes, it was shown that there was an in vitro acid-
catalyzed formation of 3,7-dimethyloct-1-ene-3,6,7-triol from
6,7-epoxylinalool, which led to furanoid linalool oxides at pH
<3.5 and/or upon heat treatment (33). More recently, Luan et
al. proved that both mechanisms are responsible for the
formation of linalool oxides in intact grape berries of Vitis
Vinifera L. cv. Morio Muscat by in vivo feeding experiments
using mixed labeled d2,18O-linalool (34, 35). Figure 7 illustrates
the stereoisomeric distribution of the linalool oxides produced
during biotransformation of a racemic mixture of (R)-(-)-
linalool and (S)-(+)-linalool by A. niger DSM 821, B. cinerea
02/FB II/2.1, and C. cassiicola DSM 62475. The diastereomers
furanoid trans-(2R,5R) and cis-(2S,5R) linalool oxide (from 37

Linalool Biotransformation
J. Agric. Food Chem., Vol. 56, No. 9, 2008 3295
Figure 8. Postulated biotransformation of linalool to furanoid, pyranoid linalool oxide, and 8-hydroxylinalool as main products and lilac aldehyde and lilac
alcohol as minor products by A. niger ATCC 16404 (AN1), A. niger DSM 821 (AN2), B. cinerea 5901/02 (BC1), B. cinerea 02/FBII/2.1 (BC2), and C.
cassiicola DSM 62475 (CC1). For each biotransformation pathway the strains are classified in ascending order of catalytic activity.
to 55% per isomer) and pyranoid trans-(2R,5S) and cis-(2S,5S)
linalool oxide (from 35 to 60% per isomer) occurred as main
products and were roughly found as mixtures of equivalent
concentrations. The absolute configuration of these compounds
suggested that they had been stereoselectively formed via 6S-
configured 6,7-epoxylinalool, as highlighted in Figure 7, which
occurred as a pair of diastereomers due to the fact that their
synthesis was independent of the C-3 configuration of (()-
linalool. Therefore, it can be postulated that (3S,6S)- and
(3R,6S)-6,7-epoxylinalool are the dominating key intermediates
in these microbial (()-linalool transformations, because the
minor furanoid and pyranoid linalool oxides isomers did not
exceed 12% of the respective total product concentration.
Moreover, a chemical linalool oxide formation can be excluded
because a pH of e3.5, necessary for an acid-catalyzed reaction
(11), was not given. A comparable biosynthetic pathway was
postulated for D. gossypina ATCC 10936 (29). To our knowl-
edge, this is also the first report on the microbial formation of
lilac aldehyde and lilac alcohol. Previously, biogenetic inves-
tigation in S. Vulgaris using deuterium-labeled precursors has
shown that lilac aldehydes and lilac alcohols are synthesized
from linalool via 8-hydroxylinalool and 8-oxolinalool (3, 4).
More recently, Matich et al. have elucidated an analogous
pathway leading to lilac compounds in Actinidia arguta flowers
by the same approach (5). Because we also found 8-hydroxy-
linalool as the main linalool metabolite in the cultures where
lilac compounds were detected, the same reaction sequence as
in plants may be postulated for the biotransformation of linalool
to lilac compounds in fungi (Figure 8). However, a rough
estimation by comparing the peak areas of the products with
those of the standards revealed that not more than 100-200
µg/L lilac compounds had been produced. It is not surprising
that during previous investigations of linalool biotransformations
neither lilac alcohol nor lilac aldehyde was found (8, 11, 15),
although they are structurally very similar to furanoid linalool
oxide, because these molecules were only produced at very low
concentrations under conventional conditions, as illustrated by
our initial investigations. Not until we carried out a direct
comparison based on retention times and mass spectra with
chemically synthesized reference compounds and applied an
optimized, strain-specific feed-batch cultivation strategy were
we able to undoubtedly identify the desired target compounds
as minor biotransformation products. Figure 8 summarizes the
pathways discussed in this paper leading to the major products,
linalool oxides and 8-hydroxylinalool, and the lilac compounds
as novel metabolic byproduct found in fungal biotransformations
of linalool. The low concentration of the latter compounds is
most probably the result of an inherent low expression and/or
(side) activity of the respective enzymes involved. The con-
centrations of the lilac compounds are obviously too low to
make a molecular biological or process engineering approach
a realistic option for obtaining economically relevant product
concentrations. Nevertheless, the discovery of a biosynthetic
pathway to lilac compounds from linalool in fungi, such as
Botrytis, will surely complement our knowledge of aroma
formation mechanisms during winemaking, where the role of
fungal biotransformation products from linalool, such as linalool
oxide and 8-hydroxylinalool, have already been described (16, 17).
ACKNOWLEDGMENT
We thank Bayerische Landesanstalt für Weinbau and Gar-
tenbau (Veitshöchheim, Germany) for the kind supply of the
Botrytis cinerea strains.
LITERATURE CITED
(1) Wakayama, S.; Namba, S.; Hosoi, K.; Ohno, M. The synthesis
and absolute configurations of lilac alcohols, new naturally
occurring odours ingredients of lilac flower. Bull. Chem. Soc. Jpn.
1973, 46, 3183–3187.
(2) Kreck, M.; Mosandl, A. Synthesis, structure elucidation, and
olfactometric analysis of lilac aldehyde and lilac alcohol stere-
oisomers. J. Agric. Food Chem. 2003, 51 (2), 2722–2726.
(3) Kreck, M.; Püschel, S.; Wüst, M.; Mosandl, A. Biogenetic studies
in Syringa Vulgaris L.: synthesis and biotransformation of
deuterium-labeled precursors into lilac aldehydes and lilac alco-
hols. J. Agric. Food Chem. 2003, 51 (2), 463–469.
(4) Burkhardt, D.; Mosandl, A. Biogenetic studies in Syringa Vulgaris
L.: biotransformation of 18O (2H)-labeled precursors into lilac
aldehydes and lilac alcohols. J. Agric. Food Chem. 2003, 51,
7391–7395.

3296 J. Agric. Food Chem., Vol. 56, No. 9, 2008
Mirata et al.
(5) Matich, A. J.; Bunn, B. J.; Comeskey, D. J.; Hunt, M. B.; Rowan,
D. D. Chirality and biosynthesis of lilac compounds in Actinidia
arguta flowers. Phytochemistry 2007, 68, 1746–1751.
(6) Krasnobajew, V. Terpenoids. In Biotechnology; Rehm, H.-J., Reed,
G., Eds.; Verlag Chemie: Weinheim, Germany, 1984; Vol. 6a,
pp 97–125.
(7) Kieslich, K.; Abraham, W.-R.; Stumpf, B.; Thede, B.; Washausen,
P. Transformations of terpenoids. In Progress in Essential Oil
Research; Brunke, H. J., Ed.; de Gruyter: Berlin, Germany, 1986;
367-394.
(8) Demyttenaere, J. C. R. Biotransformation of terpenoids by
microorganisms. In Studies in Natural Products Chemistry; Atta-
ur-Rahman, Ed.; Elsevier: Amsterdam, The Netherlands, 2001;
Vol. 25, pp 125–178.
(9) Schrader, J.; Berger, R. G. Biotechnological production of
terpenoid flavor and fragrance compounds. In Biotechnology, 2nd
ed.;Rehm, H.-J., Reed, G., Eds.; Wiley-VCH: Weinheim, Ger-
many, 2001; Vol. 10, pp 377–383.
(10) Bauer, K.; Garbe, D.; Surburg, H. Common Fragrance and FlaVor
Materials, 2nd ed.; Wiley-VCH: Weinheim, Germany, 2001; p
28.
(11) Demyttenaere, J. C. R.; Willemen, H. M. Biotransformation of
linalool to furanoid and pyranoid linalool oxides by Aspergillus
niger. Phytochemystry 1998, 47, 1029–1036.
(12) Wüst, M.; Mosandl, A. Important chiral monoterpenoid ethers in
flavours and essential oils - enantioselective analysis and biogen-
esis. Eur. Food Res. Technol. 1999, 209, 3–11.
(13) Weinert, B.; Wüst, M.; Mosandl, A.; Hanssum, H. Stereoisomeric
flavour compounds. LXXVIII. Separation and structure elucidation
of the pyranoid linalool oxide stereoisomers using common gas
chromatographic phases, modified cyclodextrin phases and nuclear
magnetic resonance spectroscopy. Phytochem. Anal. 1998, 9, 10–
13.
(14) Madyastha, K. M.; Murthy, N. S. R. K. Regiospecific hydroxy-
lation of acyclic monoterpene alcohols by Aspergillus niger.
Tetrahedron Lett. 1988, 99 (5), 579–580.
(15) Bock, G.; Benda, I.; Schreier, P. Biotransformation of linalool
by Botrytis cinerea. J. Food Sci. 1986, 51, 659–662.
(16) Demyttenaere, J. C. R.; Adams, A.; Vanoverschelde, J.; De Kimpe,
N. Biotransformation of (S)-(+)-linalool by Aspergillus niger: an
investigation of the culture conditions. J. Agric. Food Chem. 2001,
49, 5895–5901.
(17) Demyttenaere, J. C. R.; del Carmen Herrera, M.; De Kimpe, N.
Biotransformation of geraniol, nerol and citral by sporulated
surface culture of Aspergillus niger and Penicillium sp. Phy-
tochemistry 2000, 55, 363–373.
(18) King, A. J.; Dickinson, J. R. Biotransformation of hop aroma
terpenoids by ale and lager yeasts. FEMS Yeast Res. 2003, 3, 53–
62.
(21) Ettre, L. S. Das Retentionsindex-System, seine Anwendung zur
Identifizierung von Substranzen und Charakterisierung von flüs-
sigen Phasen. Chromatographia 1973, 6 (12), 525–532.
(22) Saroglou, V.; Dorizas, N.; Kypriotakis, Z.; Skaltsa, H. D. Analysis
of the essential oil composition of eight Anthemis species from
Greece. J. Chromatogr., A 2006, 1104, 313–322.
(23) Alissandrakis, E.; Tarantilis, P. A.; Paschalis, C. H.; Polissiou,
M. Aroma investigation of unifloral Greek citrus honey using
solid-phase microextraction coupled to gas chromatographic-mass
spectrometric analysis. Food Chem. 2007, 100, 396–404.
(24) Sikkema, J.; de Bont, J. A. M.; Poolman, B. Mechanisms of
membrane toxicity of hydrocarbons. Microbiol. ReV. 1995, 59 (2),
201–202.
(25) Werf van der, M. J.; de Bont, J. A. M.; Leak, D. J. Opportunities
in microbial biotransformation of monoterpenes. In Biotechnology
of Aroma Compounds; Advances in Biochemical Engineering/
Biotechnology 55; Scheper, T., Berger, R. G., Eds.; Springer:
Berlin. Germany, 1997; pp 147–177.
(26) Sikkema, J.; de Bont, J. A. M.; Poolman, B. Interactions of cyclic
hydrocarbons with biological membranes. J. Biol. Chem. 1994,
269 (11), 8022–8028.
(27) Rapp, A.; Mandrey, H. Influence of Botrytis cinerea on the
monoterpene fraction of wine aroma. In BioflaVour ’87; Schreier,
P., Ed.; de Gruyter: Berlin, Germany, 1988; pp 445–452.
(28) Demyttenaere, J. C. R.; Vanoverschelde, J.; De Kimpe, N.
Biotransformation of (R)-(+)- and (S)-(-)-citronellol by Aspergil-
lus sp. and Penicillium sp., and the use of solid-phase microex-
traction for screening. J. Chromatogr., A 2004, 1027, 137–146.
(29) Abraham, W.-R.; Stumpf, B.; Arfmann, H.-A. Chiral intermediates
by microbial epoxidations. J. Essent. Oil Res. 1990, 2, 251–257.
(30) Schrader, J. Microbial flavour production. In FlaVours and
Fragrances; Berger, R. G., Ed.; Springer-Verlag: New York, 2007;
pp 507–574.
(31) Mosandl, A. Enantioselective and isotope analysisskey steps to
flavour authentication. In FlaVours and Fragrances; Berger, R. G.,
Ed.; Springer-Verlag: New York, 2007; pp 379–408.
(32) Winterhalter, P.; Katzenberger, D.; Schreier, P. 6,7-Epoxy-linalool
and related oxygenated terpenoids from Carica papaya fruit.
Phytochemistry 1986, 25, 1347–1350.
(33) Williams, P. J.; Strauss, C. R.; Wilson, B. Hydroxylated linalool
derivatives as precursors of volatiles monoterpenes of muscat
grape. J. Agric. Food Chem. 1980, 28, 766–761.
(34) Luan, F.; Mosandl, A.; Degenhardt, A.; Gubesch, M.; Wüst, M.
Metabolism of linalool and substrate analogs in grape berry
mesocarp of Vitis Vinifera L. cv. Morio Muscat: demonstration
of stereoselective oxygenation and glycosylation. Anal. Chim. Acta
2006, 563, 353–364.
(35) Luan, F.; Mosandl, A.; Gubesch, M.; Wüst, M. Enantioselective
analysis of monoterpenes in different grape varieties during berry
ripening using stir bar sorptive extraction- and solid phase
extraction-enantioselective multidimensional gas chromato-
graphy-mass spectroscopy. J. Chromatogr., A 2006, 1112, 369–
374.
(19) Behr, D.; Wahlberg, I.; Nishida, T.; Enzell, C. R. (2E,6S)-2,6-
dimethyl-2,7-octadiene-1,6-diol, a new monoterpenoid from Greek
tobacco. Acta Chem. Scand. 1978, B32, 228–229.
(20) Askari, C.; Hener, U.; Schmarr, H.-G.; Rapp, A.; Mosandl, A.
Stereodifferentiation of some chiral monoterpenes using multi-
dimensional gas chromatography. Fresenius’ J. Anal. Chem. 1991,
340, 768–772.
Received for review January 11, 2008. Revised manuscript received
February 25, 2008. Accepted March 1, 2008.
JF800099H