Regio- and stereoselective allylic hydroxylation of D-limonene to (+)-trans-carveol with cellulosimicrobium cellulans EB-8-4

Article abstract of DOI:10.1002/adsc.200900210

Cellulosimicrobium cellulans EB-8-4 was discovered by screening of microorganisms as a pow-erful catalyst for the regio- and stereoselective allylic hydroxylation of D-limonene to (+)-trans-carveol that is a useful and valuable fragrance and flavour compound. Cells of strain EB-8-4 were easily obtained, demonstrated more than 99% regio- and stereoselectivity, showed a specific hydroxylation activity of 4.0 U/g cdw (cell dry weight), and accepted 62 mM D-limonene without inhibition. The hydroxylation was possibly catalyzed by an nicotinamide adenine dinucleotide (NADH)-dependent oxygenase involved in the degradation of aromatic ring during cell growth. 13.4 mM of (+)-trans-carveol were obtained by biohydroxylation of D-limonene with resting cells of C. cellulans EB-8-4, thus being 11 times higher than that obtained with the best biocatalyst known thus far. High conversion and high yield were obtained in the biohydroxylation of 11.6 mM of D-limonene with the resting cells as catalyst in a closed shaking flask, giving 10 mM of (+)-irans-carveol, and 0.30 mM of carvone as the only by-product. Thus, a unique biocatalyst for the regio- and stereoselective allylic hydroxylation of D-limonene and an efficient synthesis of natural identical (.+)-trans-carveol by biohydroxylation have been developed.

Full text of DOI:10.1002/adsc.200900210

FULL PAPERS
DOI: 10.1002/adsc.200900210
Regio- and Stereoselective Allylic Hydroxylation of d-Limonene
to (+)-trans-Carveol with Cellulosimicrobium cellulans EB-8-4
a
a
a
a
a,
Zunsheng Wang, Felicia Lie, Estella Lim, Keyang Li, and Zhi Li *
a
Department of Chemical and Biomolecular Engineering, National University of Singapore, 4 Engineering Drive 4,
Singapore 117576
Fax : (+65)-6779-1936; e-mail: chelz@nus.edu.sg
Received: April 1, 2009; Revised: June 5, 2009; Published online: July 16, 2009
Abstract: Cellulosimicrobium cellulans EB-8-4 was ing cells of C. cellulans EB-8-4, thus being 11 times
discovered by screening of microorganisms as a pow- higher than that obtained with the best biocatalyst
erful catalyst for the regio- and stereoselective allylic known thus far. High conversion and high yield were
hydroxylation of d-limonene to (+)-trans-carveol obtained in the biohydroxylation of 11.6 mM of d-li-
that is a useful and valuable fragrance and flavour monene with the resting cells as catalyst in a closed
compound. Cells of strain EB-8-4 were easily ob- shaking flask, giving 10 mM of (+)-trans-carveol, and
tained, demonstrated more than 99% regio- and ste- 0.30 mM of carvone as the only by-product. Thus, a
reoselectivity, showed a specific hydroxylation activi- unique biocatalyst for the regio- and stereoselective
ty of 4.0 U/g cdw (cell dry weight), and accepted allylic hydroxylation of d-limonene and an efficient
6
2 mM d-limonene without inhibition. The hydroxy- synthesis of natural identical (+)-trans-carveol by
ACHTUNGTRENNUNGl ation was possibly catalyzed by an nicotinamide ad- biohydroxylation have been developed.
enine dinucleotide (NADH)-dependent oxygenase
involved in the degradation of aromatic ring during Keywords:
biotransformations;
trans-carveol;
cell growth. 13.4 mM of (+)-trans-carveol were ob- enzyme catalysis; hydroxylation; d-limonene; terpe-
tained by biohydroxylation of d-limonene with rest- noids
Introduction
carvone which is chemically produced from d-limone-
[
5a,7d]
ne,
and the biohydroxylation of d-limonene could
Terpenes are very attractive renewable feedstock for provide with a direct access to and efficient synthesis
[
1]
sustainable chemical synthesis, due to their easy of carveol in the pure trans-form. Many microorgan-
[8–11]
availability, well-defined chirality, and high reactivity. isms have been examined for this hydroxylation,
The key challenge is the development of highly selec- but most of them gave mixtures of several pro-
[
8b–h]
tive and productive transformations of terpenes. ducts.
Among the best, the Basidiomycete Pleuro-
While chemical methods suffer from poor selectivity, tus sapidus transformed d-limonene to give 0.39 mM
biotransformation is a useful alternative and provides of cis- and trans-carveol in a ratio of 2:3 and 0.26 mM
[
9]
with additional advantage of producing natural identi- of carvone after 12 days; Rhodococcus opacus
cal aroma and flavour compounds. Over the years, PWD4 and Rhodococcus erythropolis PWD 8 showed
biotransformations of terpenes have been intensively excellent regio- and stereoselectivity for the allylic hy-
[
2,3]
investigated.
droxylation of d-limonene to produce 1.21 mM of
[
10]
We are interested in developing regio- and stereo- trans-carveol and 15 mM of carvone; However, the
selective allylic hydroxylations for terpene transfor- product concentration is low and could not be in-
mations. The biohydroxylation of d-limonene to creased by changing growth substrate, increasing sub-
trans-carveol was selected as the target reaction, since strate concentration, and applying two-liquid phase
a) d-limonene is an ideal renewable feedstock with an system. Here we report the discovery of Cellulosi-
annual production of 50,000 tons and a price of 0.66– microbium cellulans EB-8-4 as a highly selective and
.45 $ per kg in year 2000; b) carveol is a useful and productive biocatalyst for the regio- and stereoselec-
valuable fragrance and flavour additive and exhibits tive allylic hydroxylation of d-limonene and the de-
chemoprevention of mammary carcinogenesis; and velopment of the biotransformation of d-limonene
c) carveol is prepared in the cis-form
[
11]
[
4]
1
[
5]
[
6]
[
7a,b]
or as a mix- with resting cells of this strain to produce (+)-trans-
[
7c]
ture of trans- and cis-forms by reduction of (S)-(+)- carveol at 10–13.4 mM in a simple system.
Adv. Synth. Catal. 2009, 351, 1849 – 1856
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
1849

Zunsheng Wang et al.
FULL PAPERS
Results and Discussion
Isolation and Screening of Bacterial Strains for the
Biohydroxylation of d-Limonene to trans-Carveol
We have previously discovered enzymes for the regio-
and stereoselective hydroxylation of the non-activated
carbon atom of alicyclic compounds by screening of
[
12]
n-alkane-degrading strains. To find appropriate en-
zymes for the desired allylic hydroxylation, we started
with the isolation of strains that degrade toluene, eth-
ylbenzene, a-pinene, and n-octane. 40 samples from
Scheme 1. Allylic hydroxylation of d-(+)-limonene.
Among the 6 strains producing trans-carveol, the
soil, sea water, and sewage sludge in Singapore were isolate EB-8-4 showed the highest productivity, giving
used for the enrichment in M9 medium containing 2.1 mM of trans-carveol. The isolate EB 2–1 produced
0.2% (v/v) of the individual compound mentioned 138 mM of trans-carveol, while other four strains gen-
above as the sole carbon source, respectively. After erated only 20–56 mM of trans-carveol. The taxonomy
inoculation, growth, and separation of the obtained of these strains was established based on partial 16S
mixed strains on M9 agar plates, 65 pure bacterial RNA sequencing and is given in Table 1. The best
strains were isolated. They are 4 ethylbenzene-, 5a- strain EB-8-4 is an ethylbenzene-degrading strain iso-
pinene-, 22 toluene-, and 34 n-octane-degrading lated from a topsoil sample taken from an electronic
strains.
plant. This strain showed a 99.6% sequence identity
To examine the desired hydroxylation activity, the with Cellulosimicrobium cellulans CCM 2813, thus
isolates were grown in M9 medium containing the in- being named as Cellulosimicrobium cellulans EB-8-4.
dividual carbon source, the cells were harvested and In contrast, all other 5 strains belong to Pseudomonas
resuspended into 1 mL KP-buffer (pH 7.0) to a cell species.
density of 5.0 g cdw/L (cdw: cell dry weight), and d-li-
monene was added to 6.2 mM as substrate. The prod-
ucts after 2 h biotransformation were analyzed by Growth and Biohydroxylation Activity of C. cellulans
GC. As shown in Table 1, 3 ethylbenzene-degrading EB-8-4
strains and 3 toluene-degrading strains catalyzed the
regio- and stereoselective allylic hydroxylation of d-li- With Ethylbenzene as Growth Substrate
monene at position 6 to give trans-carveol. None of
the 5a-pinene-degrading strains could transform d-li- Even during the initial screening experiments, biohy-
monene. On the other hand, 3 n-octane-degrading droxylation with C. cellulans EB-8-4 gave a higher
strains were able to hydroxylate d-limonene at allylic concentration of trans-carveol than that produced
position 7 to afford perillyl alcohol (Scheme 1). This with any other known biocatalyst. To further explore
position is also a terminal position. These results fur- the potential of using this strain for the production of
ther confirm our previous observation: the monooxy- trans-carveol, growth conditions were optimized. As
genases from n-octane-degrading strains prefer hy- shown in Figure 1, C. cellulans EB-8-4 grew fast in
[
12]
droxylation at the terminal position.
M9 medium on vapour of ethylbenzene with a specific
growth rate of 0.91 h , and the cell density reached
À1
Table 1. Isolation and screening of strains for biotransformation of d-limonene.
Growing
substrates
No. of
isolated
strains
No. of strains
producing
trans-carveol
Name of strain
trans-Carveol
produced
[mM]
No. of strains
producing perillyl
alcohol
Ethylbenzene
Toluene
4
3
3
Cellulosimicrobium cellulans EB-8-4
Pseudomonas monteilii EB-2-1
Pseudomonas monteilii EB-2-3
Pseudomonas monteilii TA-1
Pseudomonas monteilii TB-3
Pseudomonas monteilii TC-1
2142
138
56
49
45
0
22
0
20
n-Octane
a-Pinene
34
5
0
0
3
0
1850
asc.wiley-vch.de
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2009, 351, 1849 – 1856

Regio- and Stereoselective Allylic Hydroxylation of d-Limonene to (+)-trans-Carveol
FULL PAPERS
activity of CFE in the presence of NADH was 1.3 U/g
protein, which was much higher than that in the pres-
ence of NADPH (lower than 0.1 U/g protein). No
CO difference absorption at 450 nm was observed for
the CFE, indicating that the hydroxylation enzyme is
not a P450 enzyme. No hydroxylation happened with
CFE in the presence of hydrogen peroxide, suggesting
that the enzyme responsible for the hydroxylation is
not a peroxygenase.
Further studies were carried out with the cells
grown on ethylbenzene for bioconversion of benzene,
toluene, and ethyl benzene. As listed in Table 2, the
bioconversion of benzene with such cells gave phenol
as the major product, which indicates that the enzyme
responsible for the hydroxylation of d-limonene to
trans-carveol might be an oxygenase involved in the
degradation of aromatic ring.
Figure 1. Growth of C. cellulans EB-8-4 on the vapour of
ethylbenzene in M9 medium and the specific activity of hy-
droxylation of d-limonene with the resting cells.
With Other Growth Substrates
The growth of C. cellulans EB-8-4 was also examined
2.9 g cdw/L at 24 h. The specific hydroxylation activity using other aromatic compounds such as benzene, tol-
was determined by performing the biotransformation uene, o-xylene, m-xylene, or p-xylene as carbon sour-
of d-limonene (12.4 mM) at 308C and 250 rpm for ces. As given in Table 2, the strain grew also quite fast
À1
3
0 min with cells harvested at different time points on toluene, with a specific growth rate of 0.89 h .
and resuspended (10 g cdw/L) in KP-buffer (pH 7.0). Cells at the exponential phase showed a specific activ-
The cells produced at later exponential phase are ity of 3.7 U/g cdw for the allylic hydroxylation of d-li-
quite active and reached the highest activity at 24 h: monene. In contrast, very slow growth was observed
4
.1 U/g cdw. Thus, active cells of C. cellulans can be with benzene and o-xylene as carbon sources, giving a
À1
À1
easily obtained in a large quantity within a short time. specific growth rate of 0.09 h and 0.02 h , respec-
The cells in the stationary phase gradually lost the hy- tively. No significant growth was observed on m-
droxylation activity.
xylene or p-xylene. Interestingly, the cells grown on
Hydroxylation of d-limonene to trans-carveol was benzene demonstrated also hydroxylation activity for
also studied with the soluble cell-free extract (CFE) d-limonene, giving a specific activity of 0.8 U/g cdw
of C. cellulans EB-8-4. The enzyme responsible for for the production of trans-carveol. This suggested
the hydroxylation was found to be cofactor depen- again that the enzyme responsible for such a hy-
dent. Nicotinamide adenine dinucleotide (NADH) is
ACHTUNGTRENNGNUd roxylation could be an oxygenase involved in the
much more preferred than nicotinamide adenine di- degradation of the aromatic ring.
nucleotide phosphate (NADPH): the hydroxylation
Table 2. Growth and biotransformation with C. cellulans EB-8-4 on different substrates.
Growth
substrate
Specific growth
rate [h ]
Biotransformation
substrate
Biotransformation
product
Activity
ACHTUNGTRENN[UNG U/g cdw]
À1
[a]
Benzene
Toluene
o-Xylene
Ethylbenzene
0.09
0.89
0.02
0.91
d-limonene
d-limonene
trans-carveol
trans-carveol
0.8
3.7
d-limonene
trans-carveol
4.0
benzene
phenol [2.3 mM]
catechol [0.7 mM]
toluene
ethylbenzene
benzyl alcohol [55 mM]
1-phenyl-1-ethanol [83 mM]
1
-phenyl-1-ethanone [14 mM]
[
a]
Biotransformation was performed with 12.4 mM substrate in 5 mL cell suspension (10 g cdw/L) in 50 mMK HPO -
2
4
KH PO buffer (pH 7.0) at 300 rpm and 308C for 30 min.
2
4
Adv. Synth. Catal. 2009, 351, 1849 – 1856
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
1851

Zunsheng Wang et al.
FULL PAPERS
Regio- and Stereoselective Hydroxylation of d-
Limonene with Resting Cells of C. cellulans EB-8-4
a system as closed as possible. On the other hand, the
hydroxylation reaction needs oxygen as the oxidant,
thus the availability of oxygen in the closed flask
should have a significant influence on hydroxylation.
Substrate Concentration and Toxicity
The resting cells of C. cellulans EB-8-4 prepared by To explore the best conditions for a high-yielding bio-
growing on ethylbenzene were used for the further transformation, two experiments with the same con-
development of an efficient allylic biohydroxylation centration of d-limonene (40 mM) were designed and
of d-limonene to trans-carveol. The addition of glu- compared: System A, with 15 mL cell suspension
cose (100 mM) was found to improve the conversion (10 g cdw/L) in a 125-mL closed conical flask; and
rate, possibly through the regeneration of NADH System B, with 28 mL suspension (10 g cdw/L) in a
during the metabolization of glucose by the cells.
125-mL closed conical flask. System A was proven to
d-Limonene is often a toxic substrate for microbial be much more efficient due to the larger availability
hydroxylations, thus resulting in low product concen- of oxygen: specific hydroxylation activity within the
[10,11]
trations.
concentrations of d-limonene for the C. cellulans EB- was achieved for System B. The concentration of
-4-catalyzed biohydroxylation, 6.2 mM to 154 mM d- trans-carveol reached 13.4 mM at 11 h for System A,
To explore the potential of using high first hour reached 4.0 U/g cdw, while only 1.3 U/g cdw
8
limonene, calculated from the amount added and while only 4 mM product were produced at 11 h for
volume of buffer solution, were used for the biotrans- System B. In both cases, the regio- and stereoselectivi-
formation with resting cells (10 g cdw/L). Although d- ties were excellent, and only a small amount of car-
limonene has very low solubility in aqueous systems vone was formed as by-product. It was found that the
[
13]
(
0.15 mM in water at 258C),
the conversion rate substrate concentration decreased quickly and linearly
was found to be dependent on the amount of d-limo- during the biotransformation for both cases. This is
nene added. As shown in Figure 2, there is a nearly possibly due to the evaporation of the volatile sub-
linear increase of the conversion rate with the in- strate during operation: the analytical samples were
crease of substrate concentration from 6.2 to 62 mM. taken at several time points with a syringe through
When the substrate concentration increased further the rubber stopper.
from 62 mM to 93 mM, the conversion rate decreased
The GC chromatograms of samples taken at 0 min,
dramatically from 1.1 to 0.2 mM/h. These results indi- 1 h, and 11 h for biotransformation of System A are
cate strong inhibition at higher substrate concentra- shown in Figure 3. The reaction gave trans-carveol as
tions. Nevertheless, it is possible to use substrate con- the major product, together with a very small amount
centrations as high as 62 mM for efficient biotransfor- of carvone which was produced from trans-carveol via
mations.
the alcohol dehydrogenase in this strain. No detecta-
ble formation of alcohols and ketones at other posi-
tions and cis-carveol indicates more than 99% regio-
and stereoselectivity of the allylic hydroxylation.
High-Yielding Biotransformation
To avoid the loss of substrate and achieve high con-
version, the biohydroxylation of d-limonene
(
(
11.6 mM) was performed with 15 mL cell suspension
10 g/L) in a 125-mL conical flask with a glass stopper
sealed by parafilm without taking samples during the
reaction (System C). The same transformations were
carried out with 8 different flasks, and the whole mix-
Figure 2. Effect of the concentration of d-limonene on the ture in the individual flask was worked up for analysis
conversion rate of hydroxylation of d-limonene to trans-car- at 30 min, 1 h, 2 h, 3 h, 4 h, 6 h, 8 h, and 10 h, respec-
veol with resting cells of C. cellulans EB-8-4 (10 g cdw/L). tively. As shown in Figure 4, the total concentrations
The conversion rate was calculated over the first three
hours.
of product and substrate were balanced at each of the
8
time points. Moreover, 8 control experiments with
the same mixture but no cells were also carried out
and worked up for analysis at the 8 different time
points. In these experiments, no disappearance of sub-
strate was observed during 10 h. Thus, such systems
Biotransformation with Volatile Substrate
Like other terpenes, d-limonene is volatile, thus high can effectively avoid the loss of d-limonene during
yielding biotransformations should be performed with biotransformation, thus allowing for high conversion
1852
asc.wiley-vch.de
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2009, 351, 1849 – 1856

Regio- and Stereoselective Allylic Hydroxylation of d-Limonene to (+)-trans-Carveol
FULL PAPERS
Figure 3. The GC chromatograms of samples taken from biotransformation of d-limonene (42 mM) in 15 mL cell suspension
in a 125-mL conical flask with rubber stopper; A: at 0 min, peak 1, impurity from substrate; n-hexadecane as internal stan-
dard; B: at 1 h; C: at 11 h.
of the substrate to the product. As shown in Figure 4,
trans-carveol was formed at 10.0 mM in 86% conver-
sion, in addition to the formation of 0.3 mM carvone
as the only by-product.
Preparation of (+)-trans-Carveol by Regio- and
Stereoselective Allylic Hydroxylation of d-Limonene
with C. cellulans EB-8-4
The preparative biotransformation was performed by
scaling the above described system to 60 mL cell sus-
pension in 500-mL conical flask with 115 mL d-limo-
nene (93.7 mg, 0.69 mol, purity 97%). After 12 h of
incubation, the conversion of d-limonene was more
than 90%. Extraction of the product with organic sol-
vent followed by evaporation gave 95.8 mg of crude
product. Further purification by chromatography on
silica gel afforded 70.8 mg (+)-trans-carveol in 69%
Figure 4. Biohydroxylation of d-limonene with resting cells
of C. cellulans EB-8-4 in a 125 mL conical flask with glass
stopper sealed with Parafilm. 15 mL cell suspension: (&) for
carvone, (~
yield. The purity was determined by GC as >99%.
1
) for d-limonene, (*) for trans-carveol; (ꢁ) for The structure was confirmed by H NMR and GC-
d-limonene in a control experiment omitting C. cellulans MS. The specific optical rotation of the isolated prod-
2
D
0
EB-8-4.
uct was measured to be [a] : +220 (c 0.87, CHCl ),
3
Adv. Synth. Catal. 2009, 351, 1849 – 1856
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
1853

Zunsheng Wang et al.
FULL PAPERS
À1
which is comparable to the reported value of +2108 1.0 mLmin . Inlet and detector temperatures were 260 and
[
14]
2
808C, respectively. The temperature was programmed to
(
c 2.0, CHCl ).
3
À1
increase from 608C to 2808C at a rate of 108Cmin and
then held at 2808C for 3 min. Retention times: 11.25, 11.80,
and 6.38 min for trans-carveol, carvone, and d-limonene, re-
Conclusions
spectively.
1
The H NMR spectrum was recorded in CDCl with TMS
3
The ethylbenzene-degrader Cellulosimicrobium cellu-
lans EB-8-4 was discovered by screening of microor-
ganisms as a powerful biocatalyst for allylic hydroxy- polarimeter. The optical density of cell cultures was mea-
as internal standard at 500 MHz on a Bruker machine. The
specific optical rotation was measured on a Jascoꢂ Spectro-
A
C
H
T
U
N
G
T
R
E
N
N
U
N
G
lation of d-limonene to (+)-trans-carveol with high sured at 450 nm (OD450) with a Hitachi U-1900 spectropho-
regio- and stereoselectivity. The cells of this strain can tometer. A cell density of 1 g cdw/L corresponds to OD450 of
be easily prepared and used for the biotransformation
0
.26.
of d-limonene to produce trans-(+)-carveol at
Isolation of Microorganisms with Enrichment Culture
Technology
13.4 mM, an 11-fold increase of the product concen-
tration in comparison with the best biocatalyst known
thus far. High conversion of substrate and high yield
of product were achieved with a simple shaking flask
system.
4
0 Samples of soil (30), sea water (5), and sewage sludge (5)
were collected from different places in Singapore. 1 g solid
sample or 1 mL liquid sample was added into 10 mLM9
[15]
[16]
mineral medium
containing trace element
and 0.2%
AHCTUNGTR(EUNNNG v/v) carbon source (a-pinene, n-octane, toluene, or ethyl-
benzene) in a 28-mL glass bottle with a metal screw cap.
Experimental Section
After incubation at 248C and 200 rpm for 2 weeks, the en-
3
5
richment culture was diluted 10 –10 times, and 0.1 mL of
the diluted culture was dispensed onto M9 agar plates. The
resulting plates were incubated in desiccators with the
vapour of a-pinene, n-octane, toluene, or ethylbenzene as
carbon source. Each single colony was picked up and then
inoculated once again in M9 agar plate with the correspond-
ing carbon source. A total of 65 strains were isolated: 4, 5,
22, and 34 strains degrading ethylbenzene, a-pinene, tolu-
ene, and n-octane, respectively. The isolates were stored in
glycerol at À808C.
Chemicals
(
À)-Carveol [97%; mixture of 58% (À)-trans-carveol and
2% (À) cis-carveol; the only commercial available carveol
with high purity], (+)-(S)-carvone (>97%), perillyl alcohol
>98%), perillic acid (>98%), d-limonene (>97%), a-
pinene (>95%), phenol (>98%), benzyl alocohol (>97%),
-phenyl-1-ethanol (>98%), 1-phenyl-1-ethanone (>97%),
4
(
1
catechol (>98%), o-xylene (>99%), m-xylene (>99%), p-
xylene (>99%), nicotinamide adenine dinucleotide
(NADH) (>99%), and nicotinamide adenine dinucleotide
phosphate (NADPH) (>99%) were purchased from Sigma–
Aldrich. Benzene (>99%), toluene (>99%), and ethylben-
zene (>99%) were obtained from Fluka. n-Octane (>97%)
was bought from Acros Organics. n-Hexadecane (>99%)
and silica gel 60 (200–300 mesh) were obtained from Merck.
Screening of Microorganisms for Allylic
Hydroxylation of d-Limonene
Each isolate was inoculated from M9 agar plate into 5 mL
Luria-Bertani (LB) broth medium and incubated at 308C
and 250 rpm for 14–16 h. 3 mL LB seed culture were trans-
formed into 100 mLM9 medium in a 250-mL conical shak-
ing flask containing a tube with 1 mL n-octane, a-pinene,
toluene, or ethylbenzene. The culture was shaken at
250 rpm and 308C for 26–30 h, and the growth was moni-
tored by measuring OD at 450 nm. Cells were harvested at
early stationary phase by centrifugation at 12,000 g and 48C
for 20 min, washed twice with K HPO -KH PO buffer
Analytical Methods
GC analysis was performed by using an Agilent 6890N in-
strument with a flame ionization detector on a HP-5 capilla-
ry column (30.0 mꢁ322 mmꢁ0.25 mm). Helium was used as
À1
the carrier gas at a flow rate of 1.0 mLmin . Inlet and de-
tector temperatures were 260 and 2808C, respectively. The
2
4
2
4
temperature was programmed to increase from 608C to
(50 mM; pH 7.0) (K-Buffer), and re-suspended to 5 g cdw/L
in 1 mLK-buffer containing 100 mM glucose and 6.2 mM d-
limonene in a 2-mL culture tube. The mixture was shaken at
1000 rpm and 308C for 2 h. 1 mL sample was taken, mixed
with 0.4 mL chloroform containing 1 mM n-hexadecane, and
centrifuged to remove the cells. The organic phase was sepa-
À1
1
008C at a rate of 58Cmin , hold at 1008C for 2 min, in-
À1
crease from 1008C to 1168C at a rate of 28Cmin , increase
from 1168C to 2808C at a rate of 308Cmin , and hold at
À1
2
2
808C for 3 min. Retention times: 16.93, 17.68, 18.20, 19.86,
2.15, 9.29, 7.00, and 23.25 min for trans-carveol, cis-carveol,
carvone, perillyl alcohol, perillic acid, d-limonene, ethyl ben-
zene, and n-hexadecane, respectively; 7,74, 9.14, 9.92, 10.14,
and 13.35 min for phenol, benzyl alocohol, 1-phenyl-1-etha-
nol, 1-phenyl-1-ethanone, and catechol, respectively. 1 mM
n-hexadecane was used as internal standard.
rated, dried over MgSO , and analyzed by GC.
4
Growth and Hydroxylation Activity of C. cellulans
EB-8-4
GC-MS analysis was performed with a Hewlett Packard
HP6890 GC system coupled with a 5973 mass selective de-
tector (EI) on the same HP-5 column as described above.
Helium was used as the carrier gas at a flow rate of
C. cellulans EB-8-4 was grown in 5 mL LB medium at 308C
and 250 rpm for 14–16 h, and 3 mL LB culture were then
transformed into 100 mLM9 medium in a 250-mL shaking
flask with a tube containing 1 mL ethylbenzene. The culture
1854
asc.wiley-vch.de
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2009, 351, 1849 – 1856

Regio- and Stereoselective Allylic Hydroxylation of d-Limonene to (+)-trans-Carveol
FULL PAPERS
was incubated at 308C and 250 rpm. At different time Preparation of (+)-trans-Carveol by
points, samples were taken for the OD measurement at
Biohydroxylation of d-Limonene with Resting Cells
of C. cellulans EB-8-4
450 nm and activity tests. To determine the hydroxylation
activity, cells were suspended to 10 g cdw/L in K-buffer con-
taining 12.4 mM d-limonene, the biotransformation was per-
formed for 30 min at 300 rpm and 308C, and the product
formation was determined by GC analysis. Typical growth
and activity curves are shown in Figure 1. Cells were har-
vested at early stationary phase at 24 h with a cell density of
To a 60 mL of cell suspension (10 g cdw/L) of C. cellulans
EB-8-4 in a 500-mL conical flask were added 115 mL d-limo-
nene (93.0 mg, 11.4 mM, 97% purity). The flask was closed
with a glass stopper, sealed with parafilm and shaken at
3
08C and 250 rpm for 12 h. After cooling at 48C for 20 min,
cells were separated from the supernatant by centrifugation
and washed twice with a total of 40 mL water. The com-
bined supernatants were extracted three times with chloro-
form. The organic phase was separated, washed with 5%
sodium carbonate aqueous solution, and dried with MgSO4
followed by filtration. The solvent was removed by evapora-
tion at reduced pressure and at 48C, giving 95.8 mg crude
product containing 90% trans-carveol as analyzed by GC.
Purification by chromatography with a short column on
silica gel using n-hexane/ethyl acetate 30/1 (v/v) as eluent
2
.9 g cdw/L and a specific activity of 4.0 U/g cdw.
Biohydroxylation of d-Limonene to trans-Carveol
with Resting Cells of C. cellulans EB-8-4
Cells of C. cellulans EB-8-4 prepared above were resuspend-
ed to 10 g cdw/L in K-buffer containing 100 mM glucose
and 42 mM d-limonene. Biotransformation was performed
with 15 mL cell suspensions (System A) or 28 mL suspen-
sions (System B) in a 125-mL conical flask with a rubber
stopper sealed by parafilm. The mixtures were shaken at
gave 70.8 mg trans-carveol (R =0.4) with an isolated yield
f
3
08C and 250 rpm. At regular time intervals, a 1 mL sample
of 69%. GC-MS: retention time, 11.25 min; MW, 152;
1
was taken using a syringe through the rubber stopper and
mixed with 0.4 mL chloroform containing 1 mM n-hexade-
cane. The organic phase was separated, dried over MgSO4,
and analyzed by GC. GC chromatograms of samples taken
from System A are given in Figure 3.
H NMR (CDCl ): d=5.61 (m, 1H), 4.77 (m, 1H), 4.75 (s,
3
1H), 4.04 (s, 1H), 2.33 (m, 1H), 2.16 (m, 1H), 1.95 (m, 1H),
1.89 (m, 1H), 1.82 (m, 3H), 1.76 (s, 3H), 1.62 (m, 1H);
2
D
5
[14]
25
D
[a] : + 220 (c 0.87, CHCl ) [lit. : [a] : + 210, (c 2.0,
3
CHCl ).
3
In System C, biotransformations were performed in eight
150-mL conical flasks with glass stoppers sealed with paraf-
ilm. To each flask were added 8 mL cell suspension (10 g
cdw/L) in K-buffer containing 100 mM glucose and 11.6 mM
d-limonene, and the mixture was shaken at 308C and
Acknowledgements
2
8
50 rpm. At each time point (30 min, 1 h, 2 h, 3 h, 4 h, 6 h,
h, and 10 h), one of the eight flasks was removed from the
This work was financially supported by Science and Engi-
neering Research Council of Agency for Science, Technology,
and Research (A-star) of Singapore through a research grant
shaker, put into ice for 20 min, opened, and mixed with
mL chloroform containing 1 mM n-hexadecane. The flask
8
(
project No. 0621010024).
was kept on ice for another 10 min, shaken by hand, and the
organic phase was separated, dried over MgSO , and ana-
4
lyzed by GC. Similarly, 8 control experiments were per-
formed under the same conditions without cells. The results
are given in Figure 4.
References
[
1] a) R. A. Sheldon, I. Arends, U. Hanefeld, Green
Chemistry and Catalysis, Wiley-VCH, Weinheim, 2007,
pp 1–48; b) R. G. Berfer, Aroma Biotechnology,
Springer, Berlin, 1995, pp 78–91; c) H. van Bekkum,
L. Maat, in: Catalysis for Renewables: From Feedstock
to Energy Production, (Eds.: G. Centi, R. A. van Sant-
en), Wiley, New York, 2007, pp 101–118.
Biohydroxylation of d-Limonene to trans-Carveol
with Cell-Free Extract of C. cellulans EB-8-4
The cells of C. cellulans EB-8-4 were resuspended into
60 mLK-buffer to 10 g cdw/L, and the suspension was
passed through a homogenizer three times. The cell debris
was removed by centrifugation at 120,000 g for 30 min. The
resulting cell-free extract contained 3.1 g/L protein, deter-
[2] a) K. Kieslich. Microbial Transformation of Non-ste-
roid Cyclic Compounds, Thieme, Stuttgart, 1976,
pp 56–84; b) H. L. Holland, Organic Synthesis with Ox-
idative Enzymes, VCH, Weinheim, 1992, pp 95–102;
c) A. Archelas, R. Furstoss, in: Enzyme Catalysis in Or-
ganic Synthesis, (Ed.: K. Drauz, H. Waldmann), VCH,
Weinheim, 1995, pp 667–704.
[3] For reviews, see: a) C. Ponzoni, C. Gasparetti, M. Gor-
etti, B. Turchetti, U. M. Pagnoni, M. R. Cramarossa, L.
Forti, P. Buzzini, Chem. Biodiversity 2008, 5, 471–483;
b) C. C. de Carvalho, M. M. da Fonseca, Biotechnol.
Adv. 2006, 24, 134–142; c) S. Das, J. P. Rosazza, J. Nat.
Prod. 2006, 69, 499–508; d) T. Ishida, Chem. Biodivers-
ity 2005, 2, 569–590; e) S. Maicas, J. J. Mateo, Appl.
Microbiol. Biotechnol. 2005, 67, 322–335; f) U. Krings,
R. G. Berger, Appl. Microbiol. Biotechnol. 1998, 49, 1–
[17]
mined by using the Bradford assay.
Biotransformations
were carried out with 5 mL cell-free extract containing
2
1
mM NADH or NADPH, 1.2 mM d-limonene, and
00 mM glucose in a 50-mL shaking flask. The mixture was
shaken at 308C and 250 rpm for 30 min, followed by mixing
with 2 mL chloroform containing 1 mM n-hexadecane, the
organic phase was separated, dried, and analyzed by GC
analysis.
Adv. Synth. Catal. 2009, 351, 1849 – 1856
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
asc.wiley-vch.de
1855

Zunsheng Wang et al.
FULL PAPERS
8
2
; g) Y. Mikami, Biotechnol. Gen. Engin. Rev. 1988, 6,
71–320.
Y. Asakawa, Phytochemistry 1992, 31, 2725–2727;
g) E. R. Bowen, Proc. Fla. State Hortic. Soc. 1975, 88,
304–308; h) R. S. Dhavalikar, P. K. Bhattacharyya,
Indian J. Biochem. 1966, 3, 144–157.
[
[
4] D. Mazzaro, Chem. Marketing Rep. 2000, 258, 18.
5] a) C. S. Sell, in: The Chemistry of Fragrances, (Eds.:
D. H. Pybus, C. S. Sell), The Royal Society of Chemis-
try, Cambridge, 1999, pp 51–124; b) S. P. Bhatia, D.
McGinty, C. S. Letizia, A. M. Api, Food Chem. Toxicol-
ogy, 2008, 46, S85-S87.
[9] J. Onken, J. , R. G. Berger, J. Biotechnol. 1999, 69,
163–168;.
10] W. A. Duetz, A. H. M. Fjallman, S. Ren, C. Jourdat, B.
[
Witholt, Appl. Environ. Microbiol. 2001, 67, 2829–
[
6] P. L. Crowell, W. S. Kennan, J. D. Haag, S. Ahmad, E.
2
932.
Vedejs, M. N. Gould, Carcinogenesis 1992, 13, 1261–
[
11] C. C. C. R. De Carvalho, M. M. R. da Fonseca, Tetrahe-
dron: Asymmetry 2003, 14, 3925–3931.
12] a) Z. Li, H.-J. Feiten, J. B. van Beilen, W. Duetz, B.
Witholt, Tetrahedron: Asymmetry 1999, 10, 1323–1333;
b) D. Chang, B. Witholt, Z. Li, Org. Lett. 2000, 2,
1
264.
7] a) Z. Rafinski, J. Scianowski, Tetrahedron: Asymmetry
008, 19, 1237–1244; b) R. Rico, F. Bermejo, Tetrahe-
[
[
2
dron Lett. 1995, 36, 7889–7892; c) J. D. Carballeira, E.
Alvarez, M. Campillo, L. Pardoc, J. V. Sinisterraa, Tet-
rahedron: Asymmetry 2004, 15, 951–962; d) K.-G. Fahl-
busch, F.-J. Hammerschmidt, J. Panten, W. Pickenha-
gen, D. Schatkowski, K. Bauer, D. Garbe, H. Surburg,
in: Ullmannꢀs Encyclopedia of Industrial Chemistry,
Wiley-VCH, Weinheim, 2002, 14, pp 73–200.
3
949–3952; c) Z. Li, H. J. Feiten, D. Chang, W. A.
Duetz, J. B. van Beilen, B. Witholt, J. Org. Chem. 2001,
6, 8424–8430.
13] I. Fichan, C. Larroche, J. B. Gros, J. Chem. Eng. Data
999, 44, 56–62.
6
[
1
[
[
14] H. Hamada, Bull. Chem. Soc. Jpn. 1988, 61, 869–878.
15] J. Sambrook, E. F. Fritsch, T. Maniatis, Molecular clon-
ing: a laboratory manual, 2nd edn., Cold Spring
Harbor Laboratory Press, New York, 1989.
16] H. Preusting, R. van Houten, A. Hoefs, E. K. Van Lan-
genberghe, O. Favre-Bulle, B. Witholt, Biotechnol.
Bioeng. 1993, 41, 550–556.
[
8] a) W. A. Duetz, H. Bouwmeester, J. B. van Beilen, B.
Witholt, Appl. Microbiol. Biotechnol. 2003, 61, 269–
2
technol. 2000, 84, 903–915; c) T. Vanek, I. Valterova,
R. Vankova, T. Vaisar, Biotechnol. Lett. 1999, 21, 625–
6
Sci. 1995, 60, 551–553; e) H. C. Chang, P. J. Oriel, J.
Food Sci. 1994, 59, 660–662; f) Y. Noma, S. Yamasaki,
77; b) T. K. Cheong, P. J. Oriel, Appl. Biochem. Bio-
[
[
28; d) H. C. Chang, D. A. Gage, P. J. Oriel, J. Food
17] M. M. Bradford, Anal. Biochem. 1976, 72, 248–254.
1856
asc.wiley-vch.de
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Adv. Synth. Catal. 2009, 351, 1849 – 1856