Heterologous expression of geraniol dehydrogenase for identifying the metabolic pathways involved in the biotransformation of citral by Acinetobacter sp. Tol 5

Article abstract of DOI:10.1080/09168451.2018.1501263

The biotransformation of citral, an industrially important monoterpenoid, has been extensively studied using many microbial biocatalysts. However, the metabolic pathways involved in its biotransformation are still unclear, because citral is a mixture of the trans-isomer geranial and the cis-isomer neral. Here, we applied the heterologous expression of geoA, a gene encoding geraniol dehydrogenase that specifically converts geraniol to geranial and nerol to neral, to identify the metabolic pathways involved in the biotransformation of citral. Acinetobacter sp. Tol 5 was employed in order to demonstrate the utility of this methodology. Tol 5 transformed citral to (1R,3R,4R)-1-methyl-4-(1-methylethenyl)-1,3-cyclohexanediol and geranic acid. Biotransformation of citral precursors (geraniol and nerol) by Tol 5 transformant cells expressing geoA revealed that these compounds were transformed specifically from geranial. Our methodology is expected to facilitate a better understanding of the metabolic pathways involved in the biotransformation of substrates that are unstable and include geometric isomers.

Full text of DOI:10.1080/09168451.2018.1501263

Bioscience, Biotechnology, and Biochemistry
ISSN: 0916-8451 (Print) 1347-6947 (Online) Journal homepage: http://www.tandfonline.com/loi/tbbb20
Heterologous expression of geraniol
dehydrogenase for identifying the metabolic
pathways involved in the biotransformation of
citral by Acinetobacter sp. Tol 5
Atsushi Usami, Masahito Ishikawa & Katsutoshi Hori
To cite this article: Atsushi Usami, Masahito Ishikawa & Katsutoshi Hori (2018): Heterologous
expression of geraniol dehydrogenase for identifying the metabolic pathways involved in the
biotransformation of citral by Acinetobacter sp. Tol 5, Bioscience, Biotechnology, and Biochemistry,
DOI: 10.1080/09168451.2018.1501263
To link to this article: https://doi.org/10.1080/09168451.2018.1501263
View supplementary material
Published online: 10 Aug 2018.
Submit your article to this journal
Article views: 3
View Crossmark data
Full Terms & Conditions of access and use can be found at
http://www.tandfonline.com/action/journalInformation?journalCode=tbbb20

BIOSCIENCE, BIOTECHNOLOGY, AND BIOCHEMISTRY
https://doi.org/10.1080/09168451.2018.1501263
Heterologous expression of geraniol dehydrogenase for identifying the
metabolic pathways involved in the biotransformation of citral by
Acinetobacter sp. Tol 5
Atsushi Usami, Masahito Ishikawa and Katsutoshi Hori
Dept. Biomolecular Engineering, Grad. Sch. Engineering, Nagoya University, Nagoya, Japan
ABSTRACT
ARTICLE HISTORY
Received 5 April 2018
Accepted 6 July 2018
The biotransformation of citral, an industrially important monoterpenoid, has been exten-
sively studied using many microbial biocatalysts. However, the metabolic pathways involved
in its biotransformation are still unclear, because citral is a mixture of the trans-isomer
geranial and the cis-isomer neral. Here, we applied the heterologous expression of geoA, a
gene encoding geraniol dehydrogenase that specifically converts geraniol to geranial and
nerol to neral, to identify the metabolic pathways involved in the biotransformation of citral.
Acinetobacter sp. Tol 5 was employed in order to demonstrate the utility of this methodology.
Tol 5 transformed citral to (1R,3R,4R)-1-methyl-4-(1-methylethenyl)-1,3-cyclohexanediol and
geranic acid. Biotransformation of citral precursors (geraniol and nerol) by Tol 5 transformant
cells expressing geoA revealed that these compounds were transformed specifically from
geranial. Our methodology is expected to facilitate a better understanding of the metabolic
pathways involved in the biotransformation of substrates that are unstable and include
geometric isomers.
KEYWORDS
Biotransformation; citral;
whole cell biocatalyst;
heterologous expression;
geraniol dehydrogenase
Terpenoids are naturally occurring compounds that
are widely used as industrially important chemicals
such as pharmaceuticals, flavors, fragrances, and as
large-volume feedstocks for chemical industries [1,2].
According to the number of isoprene units (C5H8),
terpenoids are classified into numerous groups
including monoterpenoids (two isoprene units), ses-
quiterpenoids (three isoprene units), and diterpe-
noids (four isoprene units). Monoterpenoids are the
main constituents of essential oils and are used
industrially as raw materials for flavor and fragrance
compounds [3,4]. Citral, a volatile α,β-unsaturated
aldehyde in the monoterpenoid series, is obtained
from the essential oil of lemongrass (Cymbopogon
citratus) and is composed of geometric isomers
(trans-: geranial; cis-: neral). Due to its intense
lemon aroma and flavor, citral has been used exten-
sively in the food, cosmetic, and detergent industries
since the early 1900s [5]. In addition, derivatives of
monoterpenoids including citral have the valuable
potential to serve as biologically active compounds
or as lead molecules for their development [6].
yeast, and bacteria has been investigated [11–17].
However, previous studies have not determined which
biotransformation products are derived from which of
the two citral isomers, that is, geranial and neral. This is
due to their instability; although each pure isomer can be
obtained from chemical synthesis, these are easily iso-
merized and autoxidized under ambient conditions [18].
Hence, it is difficult to provide a biocatalyst with each
isomer as the sole substrate to determine each metabolic
pathway involved in the biotransformation of geranial or
neral.
Geraniol dehydrogenase (GeoA) from Castellaniella
defragrans is an enzyme that specifically converts geraniol
to geranial and nerol to neral, respectively (Figure 1)[19].
We hypothesized that the introduction of geoA into a
host strain might be useful for identifying the metabolic
pathways involved in the biotransformation of citral,
because it can provide geranial or neral as the sole sub-
strate by intracellularly converting their respective pre-
cursors, geraniol and nerol. For proof of this concept, a
desirable host strain requires four criteria: (1) citral bio-
transformation activity, (2) tolerance to the cytotoxicity
of citral and its derivative products, (3) absence of the
original capability for the biotransformation of citral
precursors, and (4) established tools for genetic manip-
ulation. Acinetobacter sp. Tol 5, a toluene-degrading
Gram-negative bacterium isolated from a biofiltration
process [20], seemed to satisfy these criteria. Since strain
Tol 5 can metabolize diverse aromatic hydrocarbons and
Biotransformation using whole cells or isolated
enzymes as catalysts is one of the ways to produce high-
value oxygenated derivatives from monoterpenoids,
which makes it possible to realize environmentally safe,
energy-saving, and regio- and stereo-selective production
under mild conditions [7–11]. To date, the biotransfor-
mation of citral by microorganisms including fungi,
CONTACT Katsutoshi Hori
khori@chembio.nagoya-u.ac.jp
Supplemental data for this article can be accessed here.
© 2018 Japan Society for Bioscience, Biotechnology, and Agrochemistry

2
A. USAMI ET AL.
Figure 1. Reaction schemes of geraniol and nerol oxidation catalyzed by geraniol dehydrogenase (GeoA).
organic solvents [20,21], it was expected to be capable of
citral biotransformation and to show high tolerance to
the cytotoxicity of citral and its derivatives. A sequence
similarity search did not identify orthologs of geoA in the
genome of Tol 5. Tools for the genetic manipulation of
Tol 5 have already been established [22,23]. Here, we
generated and utilized the transformant of Tol 5 into
which geoA was introduced in order to demonstrate
that the heterologous expression of geoA is useful for
identifying the metabolic pathways involved in the bio-
transformation of citral.
harvested during the exponential growth phase every
hour and OD660 was measured. The maximum growth
rate at a given citral concentration (μmax) was determined
at the exponential growth phase immediately after citral
addition. The maximum growth rate without citral
(μmax0) at the exponential growth phase was also deter-
mined. Growth inhibition was evaluated using the ratio
of μmax and μmax0 (μmax/μmax0).
Construction of a transformant of Acinetobacter
sp. Tol 5 expressing geraniol dehydrogenase
The geoA gene optimized for codon usage of Tol 5
was artificially synthesized (Genewiz, Kawaguchi,
Japan) and cloned into the EcoRI-XbaI site of
pARP3, an Escherichia coli – Acinetobacter shuttle
plasmid [22], generating pGeoA. Genetic manipula-
tion of Tol 5 with pGeoA was performed as described
previously [22]. The expression of geoA in the resul-
tant transformant, Tol 5 (pGeoA), was confirmed by
SDS-PAGE. Whole cell lysates of the samples were
prepared as described previously [21] and loaded on a
10.0% polyacrylamide gel. After electrophoresis, the
gel was stained with Coomassie brilliant blue.
Materials and methods
Substrates, microorganisms, and cell growth
Citral was purchased from Tokyo Chemical Industry
Co., Ltd (Tokyo, Japan). Geraniol and nerol were
purchased from Wako Pure Chemical Industries,
Ltd. (Osaka, Japan). (R)-(–)- and S-(+)-α-methoxy-
α-(trifluoromethyl)phenylacetyl chloride were pur-
chased from Sigma-Aldrich (Tokyo, Japan).
Acinetobacter sp. Tol 5 and its transformant were
grown on basal salt (BS) medium [21] supplemented
with sodium lactate (0.2%, wt/vol) or in Luria-Bertani
(LB) medium at 28°C, with shaking. Ampicillin (500 μg/
mL) and gentamicin (10 μg/mL) were added to the
medium when required. Arabinose was added to a final
concentration of 0.5% (wt/vol) for the induction of geoA.
The inhibitory effect of citral on the cell growth of Tol
5 was investigated by using the growth inhibition assay of
Mi et al. [24]. To avoid the volatilization of citral into the
air, 15-mL conical centrifuge tubes with screw caps were
used. Tol 5 cells were precultured in LB medium at 28°C
with shaking for 8 h. Twenty microliters of the preculture
was inoculated into 2 mL LB medium in a screw-capped
test tube. Citral was added from a dimethyl sulfoxide
(DMSO) stock solution at the beginning of the exponen-
tial growth phase, when the optical density at 660 nm
(OD660) reached 0.2. Although a control with the highest
applied DMSO concentration was also tested, DMSO
had no effect on the cell growth of Tol 5. Samples were
Analytical methods
1H- and 13C-NMR spectra were recorded at 500 MHz
and 125 MHz, respectively, on a spectrometer (Advance
III 500; Bruker Biospin Ltd.). Tetramethylsilane was used
as the internal standard (1H-, δ 0.00; 13C-, δ 77.00) for
NMR spectra measured in CDCl3. Multiplicities were
determined by using the DEPT pulse sequence. Optical
rotations were measured with a digital polarimeter (P-
1010-GT; JASCO). The infrared (IR) spectra were
recorded on a Fourier transform IR spectrometer (FT/
IR-4100; JASCO). Thin-layer chromatography (TLC)
was performed on precoated plates (Silica gel 60 F254,
0.25 mm; Merck). The mobile phase was hexane/ethyl
acetate (EtOAc). Compounds were visualized by spraying
the plates with 0.5% vanillin in 96% H2SO4 followed by

HETEROLOGOUS EXPRESSION FOR IDENTIFYING METABOLIC PATHWAYS
3
brief heating. Gas chromatography-mass spectrometry
(GC-MS) analysis was performed on a model 7820A gas
chromatograph (Agilent Technologies, Wilmington, DE)
equipped with a model MSD 5977E mass spectrometer
(Agilent Technologies). A fused-silica capillary Rtx-200
column (30 m length, 0.32 mm i.d.; Restex, Bellefonte,
PA) was used and helium was supplied through the
column as the carrier gas at 10 mL/min. The oven tem-
perature was programmed from 60 to 250°C at an
increase rate of 4°C/min. The injection port temperature
was maintained at 245°C. The GC column effluent was
introduced directly into the ion source via a transfer line
at 280°C. The ion source temperature was set at 230°C.
The split ratio for the injection and the electron impact
ionization voltage were set to 50:1 and 70 eV, respectively.
1.90 (1H, ddd, J = 12.7, 3.9, 3.9 Hz, H-4), 2.07 (1H, td,
J = 13.9, 2.7 Hz, H-2), 3.82 (1H, td, J = 3.7, 2.7 Hz, H-3),
4.98 (2H, m, H-10); 13C NMR (CDCl3, 125 MHz) δ
146.0 (C, C-8), 113.1 (CH2, C-10), 71.4 (C, C-1), 67.3
(CH, C-3), 53.9 (C, C-4), 46.1 (CH2, C-2), 38.1 (CH2,
C-6), 31.7 (CH3, C-7), 25.3 (CH2, C-5), 19.3 (CH3, C-9).
The assigned compound was confirmed by comparison
to the literature [30].
Preparation of esters with (R)-(–) - and (S)-(+)-2-
methoxy-2-phenyl-3,3,3-trifluoropropanoic acids
Esters of 1-methyl-4-(1-methylethenyl)-1,3-cyclohexa-
nediol (MMC) with (R)-(–)- and (S)-(+)-2-methoxy-2-
phenyl-3,3,3-trifluoropropanoic acids (MTPA esters)
were prepared by the method of Ohtani et al. [31],
respectively. MMC (0.023 mmol) was dissolved in
2.0 ml dichloromethane and stirred with (R)-(–)- or
(S)-(+)-α-methoxy-α-(trifluoromethyl)phenylacetyl
chloride (0.041 mmol), trimethylamine (0.090 mmol),
and 4-dimethylaminopyridine (0.004 mmol) at room
temperature for 30 min under N2. The reaction mixture
was concentrated in vacuo, and then (R)-(–)- or (S)-
(+)-MTPA esters of MMC was purified by silica gel
open-column chromatography.
Time courses of biotransformation and
autoxidation
One hundred and fifty microliters of precultured Tol 5
or Tol 5 (pGeoA) cells was inoculated into 15 mL BS
medium supplemented with 8 μL toluene in a 50-mL
conical centrifuge tube. When the OD660 of the culture
reached 1.0, 80 mM citral, geraniol, or nerol dissolved in
20 mM DMSO was added to the culture for their
biotransformation. During incubation at 28°C for
14 days with shaking at 115 rpm, 1 mL culture was
harvested every other day, saturated with NaCl, and
extracted with EtOAc, followed by evaporation of the
solvent. The crude extracts were analyzed by TLC, GC,
and GC-MS. To isolate a biotransformation product of
citral by Tol 5, the crude extract obtained from a culture
of the biotransformation for 14 days was subjected to
silica gel open-column chromatography (silica gel,
40–66 µm; Wako) with gradient elution (hexane/
EtOAc: 9:1 – 1:9). Relative abundances of the substrates
and their metabolites were determined on the basis of
GC peak area. Citral autoxidation was investigated at
the same condition in the absence of microbial cells.
The identities of the autoxidation products of citral
were confirmed by comparison of their mass spectra
to data reported in the literature [25–29], computa-
tional searches of a commercial mass spectral library
(NIST 14, which includes 276,248 EI spectra for 242,466
compounds), and of authentic samples.
Results
Tolerance of Acinetobacter sp. Tol 5 to citral
Since citral and some its biotransformation products
show antimicrobial activity [2], the tolerance of a host
strain to citral is an important factor for its biotrans-
formation. The tolerance of Tol 5 cells to citral was
investigated using the growth inhibition assay.
Maximum growth rates (μmax) at different concentra-
tions of citral (10–100 mM) and µmax0 (in the absence
of citral) were determined. The ratios of µmax and
µmax0 were plotted against different concentrations of
citral (Figure 2). The cell growth of Tol 5 was signifi-
cantly affected at 10 mM citral, distinctly impaired at
90 mM, and completely ceased at 100 mM. Since
6.6 μM – 29 mM citral was subjected to biotransfor-
mation in previous studies [11–17]. Tol 5 has sufficient
tolerance to be used for citral biotransformation.
Structural analysis of 1-methyl-4-(1-
methylethenyl)-1,3-cyclohexanediol as an
isolated biotransformation product
Capability of citral biotransformation in
Acinetobacter sp. Tol 5
From the metabolic versatility of Tol 5 [20,21], we
expected it would possess the capability for citral bio-
transformation. Biotransformation was investigated by
incubating Tol 5 cells at 28°C with 80 mM citral for
14 days. To distinguish biotransformation products
from autoxidation products, citral was also incubated in
the same condition except for the absence of Tol 5 cells.
The proportion of geranial and neral was 55:45 in the
Pale yellow oil: [α]D21.9 − 19.1° (CHCl3, c 0.70); EIMS,
m/z (rel. intensity) 170 [M]+ (1), 152 (13), 134 (19), 119
(21), 108 (55), 87 (40), 82 (31), 68 (100), 58 (27), 43 (47);
IR (KBr, νmax, cm−1) 3396; 1H NMR (CDCl3, 500 MHz)
δ 1.28 (3H, s, H-7), 1.41 (1H, dd, J = 11.6, 2.7 Hz, H-6),
1.43 (1H, m, H-5), 1.50 (1H, dd, J = 13.9, 2.7, Hz, H-2),
1.68 (1H, m, H-5), 1.75 (3H, s, H-9), 1.85 (1H, m, H-6),

4
A. USAMI ET AL.
abundance of geranial and neral also decreased
(Figure 3B) and 8 products were detected (Figure 3D).
On the basis of previous reports on the autoxidation and
biotransformation of citral [11–17,25–29], 7 autoxidation
products were identified by the mass spectral fragmenta-
tion pattern and retention time on the GC chromato-
gram; 3 aliphatic compounds were geranic acid, neric
acid, and 3,7-dimethyl-6,7-epoxy-2-octenal, the other
compounds possessed cyclic skeleton such as p-mentha-
1,5-diene-8-ol, trans-p-menth-2-en-1,8-diol, p-cymen-8-
ol, and 8-hydroperoxy-p-cymene. Although 7 of the 8
products in the presence of Tol 5 cells were identified as
the same compounds produced by autoxidation in the
absence of the cells, one product (cross marker in
Figure 3D) could not be identified. To identify this pro-
duct, the large-scale biotransformation of citral in the
presence of Tol 5 cells was performed, and then this
product was isolated from the EtOAc extract by silica
gel open-column chromatography. The molecular for-
mula of this product was determined as C10H18O2 from
its mass spectral fragmentation pattern. The IR spectrum
contained a characteristic absorption band for an alcohol
Figure 2. Growth inhibition of Acinetobacter sp. Tol 5 by
citral.
µmax, maximum growth rate at a given citral concentration; and µmax0
maximum growth rate without citral.
,
The experiments were conducted using three different conical centrifuge
tubes with screw caps. Data are expressed as the mean
division (n = 3).
standard
citral used in this study (Figure 3AB). Even in the absence
of Tol 5 cells, the relative abundance of geranial and neral
decreased (Figure 3A) and 7 products were detected
(Figure 3C), indicating that autoxidation reactions pro-
gressed. In the presence of Tol 5 cells, the relative
1
(3396 cm−1) compared with the citral. The H and 13C
NMR spectra of the isolated product were assigned using
the two-dimensional techniques: heteronuclear single-
quantum coherence (HSQC), correlation spectroscopy
Figure 3. Time courses of the autoxidation and biotransformation of citral by Acinetobacter sp. Tol 5.
(A) Consumption of citral during autoxidation. (B) Consumption of citral during biotransformation by Tol 5 cells. (C) Products from citral autoxidation. (D)
Products from biotransformation by Tol 5 cells.

HETEROLOGOUS EXPRESSION FOR IDENTIFYING METABOLIC PATHWAYS
5
(COSY), and heteronuclear multiple bond correlation
(HMBC). The 1H and 13C NMR spectra of the isolated
product indicated the existence of two new methylene
groups and the disappearance of one methyl group and
one quaternary carbon group. The characteristic COSY
spectrum indicated the correlation cross-peaks of the
three signals at δ3.82–2.07, 1.90, and 1.50 (Figure S1).
The characteristic HMBC spectrum showed the key cor-
relation of H-9 (1.75 ppm) to C-4 (53.9 ppm), C-8 (146.0
ppm), and C-10 (113.1 ppm); H-5 (1.43 ppm) to C-3 67.3
ppm), C-4 (53.9 ppm), and C-8 (146.0 ppm) (Figure S1).
These data indicated that two hydroxyl groups should be
located at C-1 and C-3 of the isolated product. Specific
rotation shows the (−)-form (−19.1°). Finally, the isolated
product was determined as 1-methyl-4-(1-methylethe-
nyl)-1,3-cyclohexanediol (MMC), which was reported
as a major product from geraniol in chemical synthesis
[30], but has not been reported as a biotransformation
product to date. The structural evidence was confirmed
by comparing it with the spectroscopic data reported in
the literature [30]. The absolute configuration of MMC
was determined by the modified Mosher’s method [31].
This method only esterified the secondary hydroxyl
group in position 3 of MMC. The differences of proton
chemical shifts (Δδvalues, δS − δR) between (S)-MTPA
ester of MMC and (R)-MTPA ester of MMC indicated
that the configuration of C-1, C-3, and C-4 of MMC were
R ((1R,3R,4R)-MMC, Figure S2). The products from the
autoxidation and biotransformation of citral are sum-
marized in Table 1. The present study is the first to
demonstrate the biotransformation of citral into
(1R,3R,4R)-MMC; however, it was not determined
whether this compound was derived from geranial or
neral.
occurred (Figure S3), indicating that Tol 5 cells are
unable to biotransform citral precursors. Although
geraniol and nerol were also incubated in the absence
of Tol 5 cells, autoxidation did not occur (Figure S4),
indicating that geraniol and nerol are more stable in
the medium than citral.
As Tol 5 is tolerant to citral and can transform it but
not its precursors (geraniol and nerol), it was consid-
ered to be a suitable strain into which geoA could be
introduced to identify the metabolic pathways involved
in the biotransformation of citral. Therefore, Tol 5 was
transformed with the pGeoA plasmid. The resultant
transformant, Tol 5 (pGeoA), was confirmed to express
geoA by SDS-PAGE (Figure 4). Arabinose was added
for the induction of geoA under the control of an
arabinose-inducible promoter. A remarkable band was
detected at approximately 38 kDa from Tol 5 (pGeoA)
cells cultured in the presence of arabinose. Since the
molecular weight of GeoA is 38,272 Da, the expression
of geoA in the Tol 5 (pGeoA) cells could be confirmed.
Identification of the pathway for the synthesis of
(1R,3R,4R)-1-methyl-4-(1-methylethenyl)-1,3-
cyclohexanediol in Tol 5 (pGeoA) cells
To identify the pathway for the synthesis of MMC, the
time courses of the biotransformation of geraniol and
nerol by Tol 5 (pGeoA) cells were investigated. In the
biotransformation of geraniol by Tol 5 (pGeoA) cells,
3 compounds ((1R,3R,4R)-MMC, geranic acid, and
Expression of geoA in Acinetobacter sp. Tol 5
The biotransformation of citral precursors (geraniol
and nerol) was investigated to confirm the absence of
the capability for the biotransformation by the wild-
type cells of Tol 5. Geraniol and nerol were incubated
with Tol 5 cells at 28°C for 14 days, but no reaction
Table 1. Production ratios of citral autoxidation and biotrans-
formation by Acinetobacter sp. Tol 5.
Product ratio (%)a
Compound
Autoxidation Biotransformation
Geranial
2.89c
3.23c
–
9.85c
9.09c
9.53
5.98
2.52
37.58
9.85
9.09
7.16
2.24
Neral
(1R,3R,4R)-MMCb
Geranic acid
6.45
3.98
Neric acid
3,7-Dimethyl-6,7-epoxy-2-octenal
p-Mentha-1,5-diene-8-ol
trans-p-Menth-2-en-1,8-diol
p-Cymen-8-ol
43.64
15.96
10.33
16.90
4.54
Figure 4. Confirmation of GeoA expression in a transformant
of Acinetobacter sp. Tol 5.
Whole cell lysate from Tol 5 transformant, Tol 5 (pGeoA), cells was analyzed by
SDS-PAGE. An arrowhead indicates a protein induced remarkably by the
addition of 0.5% arabinose. M, protein marker (AE-1440, EzStandard; ATTO);
(+), Tol 5 (pGeoA) cells grown with 0.5% arabinose; (–), Tol 5 (pGeoA) cells
grown without 0.5% arabinose.
8-Hydroperoxy-p-cymene
a Percentage was calculated from the peak area in the gas chromatogram.
b
(1R,3R,4R)-1-methyl-4-(1-methylethenyl)-1,3-cyclohexanediol
c
Recovered substrates

6
A. USAMI ET AL.
geranial) were detected during 14 days of incubation
(Figure 5A). The relative abundance of the produced
(1R,3R,4R)-MMC gradually increased and finally
reached 8.6% at 10 days. The relative abundance of
geranic acid and geranial increased to 2.0% and 1.2%
in the first two days, but decreased to 1.5% and 0.1%
on the fourth day, respectively. After that, the relative
abundance of geranic acid finally increased to 1.7%,
whereas that of geranial did not change. Since GeoA
converts geraniol into geranial, but not into
(1R,3R,4R)-MMC and geranic acid, this result indi-
cates that (1R,3R,4R)-MMC and geranic acid were
produced from the geranial generated by the oxidation
of geraniol by GeoA (Figure 6A). In the biotransfor-
mation of nerol by Tol 5 (pGeoA) cells, 4 compounds
were identified as neral, geraniol, geranial, and
(1R,3R,4R)-MMC (Figure 5B). Unlike the biotransfor-
mation of geraniol, the level of geranic acid was below
the detection limit of GC-MS. GeoA intracellularly
oxidized nerol into neral, but the relative abundance
of the produced neral (2.1%) did not change after the
fourth day, implying no further reaction. The relative
abundance of geraniol and geranial gradually
increased to 2.2% and 2.0% until the fourth and sixth
days, respectively, and thereafter it decreased to 0.1%
and 0.2% in two further days of incubation, respec-
tively. The sequential detection and disappearance of
these two chemicals show that geraniol was produced
by the isomerization of nerol by Tol 5 (pGeoA) and
then oxidized into geranial. In contrast to geraniol and
geranial, (1R,3R,4R)-MMC was not detected at the
beginning of the incubation period, but it was detected
gradually from the fourth day and its relative abun-
dance finally increased to 3.2%. Since the relative
abundance of (1R,3R,4R)-MMC increased with the
decrease of geranial and geraniol, (1R,3R,4R)-MMC
was considered to be produced from geranial oxidized
from geraniol by GeoA.
Discussion
In the present study, the biotransformation of citral by
Tol 5 cells and the autoxidation of citral in the absence
of Tol 5 cells were investigated. The resultant products
are listed in Table 1. Seven of the 8 biotransformation
products detected were also produced by citral autoxi-
dation in the absence of Tol 5 cells. Taking all of the
analyses in the present study into account, the biotrans-
formation products of citral by Tol 5 cells were geranic
acid and (1R,3R,4R)-MMC, and the other compounds
were derived from the autoxidation of citral (Figure 6B).
Geranic acid was generated from both the biotransfor-
mation and autoxidation of citral. Although geranic
acid was detected at a low level as the biotransformation
product of geraniol by Tol 5 (pGeoA) cells (Figure 5A),
it was below the detection limit as the biotransforma-
tion product of nerol by Tol 5 (pGeoA) cells
(Figure 5B). Only a very small amount of geraniol was
produced from nerol by Tol 5 (pGeoA) cells, resulting
in the production of a small amount of geranial. Since
the production ratio of (1R,3R,4R)-MMC and geranic
acid was considered to be constant (5:1) in the biotrans-
formation of geranial, the amount of geranic acid pro-
duced in the biotransformation of nerol was lower than
the detection limit of GC-MS. The enzymatic activity
involved in the production of (1R,3R,4R)-MMC in Tol
5 cells might be higher than that of geranic acid.
Geranial dehydrogenase is known as an enzyme that
coverts geranial into geranic acid [19]. Sequence simi-
larity search found 12 homologs of geranial dehydro-
genase that show more than 30% amino acid sequence
identity in the genome of Tol 5. Probably, some of them
are involved in the biotransformation of geranial into
geranic acid. On the other hand, no enzyme that con-
verts geranial into (1R,3R,4R)-MMC has been known,
because the present study is the first to detect it as the
biotransformation product. Further study is needed to
identify and characterize these enzymes in Tol 5 cells.
Figure 5. Time courses of the biotransformation of (A) geraniol and (B) nerol by a transformant of Acinetobacter sp. Tol 5
expressing geraniol dehydrogenase.
(1R,3R,4R)-1-methyl-4-(1-methylethenyl)-1,3-cyclohexanediol ((1R,3R,4R)-MMC) is indicated by the cross marker.

HETEROLOGOUS EXPRESSION FOR IDENTIFYING METABOLIC PATHWAYS
7
Figure 6. Reaction schemes of citral and its precursors in this study.
(A) Biotransformation of citral precursors (geraniol and nerol) by Acinetobacter sp. Tol 5 and its transformant expressing geraniol dehydrogenase. (B)
Biotransformation products of citral by Acinetobacter sp. Tol 5 and autoxidation products of citral in the cultivation condition of Tol 5 cells.
The autoxidation [25–29] and biotransformation
of citral, one of the most valuable and widespread
monoterpenoids, by bacteria, yeast, and fungi have
been studied extensively [11–17]. The autoxidation of
citral in the cultivation condition of Tol 5 cells started
immediately, and cyclization and chain oxidation
reactions occurred (Figure 3AB). This autoxidation
pathway of citral is similar to that observed in acidic
conditions [25–29], but the number of products and
the formation rates were different. Probably, such
differences were attributed to the temperature and
pH of the incubation conditions. In the biotransfor-
mation of citral by microorganisms, different pro-
ducts have been reported. For example, sporulated
surface cultures of Aspergillus niger strains and
Penicillium species mainly produced linalool [12,13],
Penicillium digitatum transformed citral into 6-
methylhept-5-en-2-one [14]. The major products of
A. niger PTCC5011 were citronellol and hydroxy
citronellal [15]. Zymomonas mobilis, Citrobacter
freundii, Candida rugosa, and Candida parapsilosis
produced citronellal or citronellol [16,17]. The pre-
sent study is the first to demonstrate that (1R,3R,4R)-
MMC was transformed from citral. From the results
of the biotransformation of citral precursors (geraniol
and nerol) by Tol 5 (pGeoA) cells, we finally con-
cluded that (1R,3R,4R)-MMC was specifically trans-
formed from geranial in Tol 5 cells.
Although our results indicated that autoxidation of
citral progresses readily, even in the cultivation con-
dition used in the present study, previous studies of
citral biotransformation have not paid close attention
to it. In the biotransformation of citral precursors by
Tol 5 (pGeoA) cells, no autoxidation products of
citral were detected (Figure 5), indicating that the
intracellularly produced geranial and neral were not
affected by autoxidation. Since it is generally consid-
ered that the intracellular environments of microor-
ganisms are reductive, it might be difficult for citral
autoxidation to progress inside cells. Therefore, the
biotransformation of more stable citral precursors by
Tol 5 (pGeoA) cells could clearly distinguish the
biotransformation products of citral from the auto-
xidation products of citral. In the field of microbiol-
ogy, heterologous expression of an enzyme has been
used extensively for the mass production of value-
added chemicals. In contrast, we used the heterolo-
gous expression of geoA to distinguish biotransfor-
mation products from autoxidation products and to
identify the metabolic pathways involved in the

8
A. USAMI ET AL.
completely revised and enlarged edition. Weinheim:
Wiley – VCH; 2006.
[6] Salakhutdinov NF, Volcho KP, Yarovaya OI.
Monoterpenes as a renewable source of biologi-
cally active compounds. Pure Appl Chem.
2017;89:1105–1117.
[7] Lye GJ, Dalby PA, Woodley JM. Better biocatalytic
processes faster: new tools for the implementation of
biocatalysis in organic synthesis. Org Process Res
Dev. 2002;6:434–440.
[8] Doig SD, Simpson H, Alphand V, et al.
Characterization of a recombinant Escherichia coli
TOP10 [pQR239] whole-cell biocatalyst for stereose-
lective Baeyer–villiger oxidations. Enzyme Microb
Technol. 2003;32:347–355.
biotransformation of citral in Tol 5 cells. Since this
methodology can be applied for genes encoding other
enzymes, it is expected to facilitate a better under-
standing of the metabolic pathways involved in the
biotransformation of substrates that are unstable and
include geometric isomers.
Acknowledgments
This work was supported by the Advanced Low Carbon
Technology Research and Development Program (ALCA)
of the Japan Science and Technology Agency (JST) and by
the Program for Leading Graduate Schools “Integrative
Graduate Education and Research in Green Natural
Sciences,” Ministry of Education, Culture, Sports, Science
and Technology (MEXT), Japan. The authors have declared
that no competing interests exist and they have no conflicts
of interest with the contents of this article.
[9] Xu F. Applications of oxidoreductases: recent pro-
gress. Ind Biotechnol. 2005;1:38–50.
[10] Pinheiro L, Marsaioli AJ. Microbial monooxygenases
applied to fragrance compounds. J Mol Catal B
Enzym. 2007;44:78–86.
[11] Demyttenaere
J.
and
De
Kimpe
N,
Biotransformation of terpenes by fungi - study of
the pathways involved. J Mol Catal B Enzym.
2001;11:265–270.
Author contribution
[12] Demyttenaere
JCR.
and
De
Pooter
HL,
AU designed and performed the experiments, analyzed the
data, and wrote the manuscript. MI and KH discussed the
results and completed the manuscript. All authors have
read and approved the final manuscript.
Biotransformation of citral and nerol by spores of
Penicillium digitatum. Flavour Frag J. 1998;13:173–176.
[13] Demyttenaere JC, Del Carmen Herrera M. and De
Kimpe N, Biotransformation of geraniol, nerol and citral
by sporulated surface cultures of Aspergillus niger and
Penicillium sp. Phytochemistry. 2000;55:363–373.
[14] Esmaeili A, Tavassoli A. Microbial transformation of
citral by Penicillium sp. Acta Biochim Pol.
2010;57:265–268.
[15] Esmaeili A, Rohany S, Safaiyan S, et al. Microbial
transformation of citral by Aspergillus niger-PTCC
5011 and study of the pathways involved. Czech J
Food Sci. 2011;29:610–615.
Disclosure statement
No potential conflict of interest was reported by the
authors.
Funding
[16] Müller A, Hauer B, Rosche B. Enzymatic reduction
of the α, β-unsaturated carbon bond in citral. J Mol
Catal B Enzym. 2006;38:126–130.
[17] Hall M, Hauer B, Stuermer R, et al. Asymmetric
whole-cell bioreduction of an α,β-unsaturated alde-
hyde (citral): competing prim-alcohol dehydrogenase
and C–C lyase activities. Tetrahedron: Asymmetry.
2006;17:3058–3062.
This work was supported by the Advanced Low Carbon
Technology Research and Development Program of the
Japan Science and Technology Agency [grant number
JPMJAL1402].
ORCID
[18] Cannon RJ, Curto NL, Esposito CM, et al. The dis-
covery of citral-like thiophenes in fried chicken. J
Agric Food Chem. 2017;65:5690–5699.
Katsutoshi Hori
http://orcid.org/0000-0003-1994-3124
[19] Lüddeke F, Wulfing A, Timke M, et al. Geraniol and
geranial dehydrogenases induced in anaerobic mono-
terpene degradation by Castellaniella defragrans. Appl
Environ Microbiol. 2012;78:2128–2136.
[20] Hori K, Yamashita S, Ishii S, et al. Isolation, char-
acterization and application to off-gas treatment of
toluene-degrading bacteria. J Chem Eng Japan.
2001;34:1120–1126.
[21] Hori K, Ishikawa M, Yamada M, et al. Production of
peritrichate bacterionanofibers and their proteinac-
eous components by Acinetobacter sp. Tol 5 cells
affected by growth substrates. J Biosci Bioeng.
2011;111:31–36.
[22] Ishikawa M, Nakatani H, Hori K. AtaA, a new
member of the trimeric autotransporter adhesins
from Acinetobacter sp. Tol 5 mediating high adhe-
siveness to various abiotic surfaces. PLoS One.
2012;7:e48830.
References
[1] Joshi S, Chanotiya CS, Agarwal G, et al. Terpenoid
compositions, and antioxidant and antimicrobial prop-
erties of the rhizome essential oils of different
Hedychium species. Chem Biodivers. 2008;5:299–309.
[2] Vespermann KA, Paulino BN, Barcelos MC, et al.
Biotransformation of alpha- and beta-pinene into
flavor compounds. Appl Microbiol Biotechnol.
2017;101:1805–1817.
[3] Chisowa EH, Hall DR, Farman DI. Volatile constitu-
ents of the essential oil of Cymbopogon citratus Stapf
grown in Zambia. Flavour Frag J. 1998;13:29–30.
[4] Van Der Werf MJ, De Bont JA, Leak DJ. Opportunities
in microbial biotransformation of monoterpenes. Adv
Biochem Eng Biotechnol. 1997;55:147–177.
[5] Surburg H, Panten J. Common fragrance and flavor
materials: preparation, properties and uses, 5th

HETEROLOGOUS EXPRESSION FOR IDENTIFYING METABOLIC PATHWAYS
9
[23] Ishikawa M, Hori K. A new simple method for
introducing an unmarked mutation into a large
gene of non-competent Gram-negative bacteria by
alkoxy radical intermediate. J Agric Food Chem.
2004;52:5677–5684.
[28] Schieberle P. Garching, and Grosch W, Quantitative
analysis of important volatile flavour compounds in
fresh and stored lemon oil/citric acid emulsions.
Lebensm Wiss Technol. 1988;21:158−162.
[29] Hagvall L, Baron JM, Borje A, et al. Cytochrome
P450-mediated activation of the fragrance com-
pound geraniol forms potent contact allergens.
Toxicol Appl Pharmacol. 2008;233:308–313.
[30] Fkyerat A, Tabacchi R. Enantioselective preparation
of 1-hydroxy neoisopulegol and 1-hydroxy neoi-
somenthol. Tetrahedron: Asymmetry. 1997;8:2231–
2236.
[31] Ohtani I, Kusumi T, Kashman Y, et al. A new aspect
of the high-field NMR application of Mosher’s
method. The absolute configuration of marine triter-
pene sipholenol A. J Org Chem. 1991;56:1296–1298.
FLP/FRT
recombination.
BMC
Microbiol.
2013;13:86–95.
[24] Mi J, Becher D, Lubuta P, et al. De novo production
of the monoterpenoid geranic acid by metabolically
engineered Pseudomonas putida. Microb Cell Fact.
2014;13:170–180.
[25] Kimura K, Nishimura H, Iwata I, et al. Deterioration
mechanism of lemon flavor. 2. Formation mechan-
ism of off-odor substances arising from citral. J Agric
Food Chem. 1983;31:801–804.
[26] Schieberle P, Grosch W. Identification of potent flavor
compounds formed in an aqueous lemon oil citric-acid
emulsion. J Agric Food Chem. 1988;36:797–800.
[27] Ueno T, Masuda H, Ho CT. Formation mechanism
of p-methylacetophenone from citral via a tert-