Production of caloxanthin 3′-β-d-glucoside, zeaxanthin 3,3′-β-d-diglucoside, and nostoxanthin in a recombinant Escherichia coli expressing system harboring seven carotenoid biosynthesis genes, including crtX and crtG

Article abstract of DOI:10.1016/j.phytochem.2011.02.017

The aim of this study was to produce rare β-carotene-modified carotenoids possessing 2-O (-H or -glu) and/or 3-O (-H or -glu) functionalities in their β-ionone ring(s) using a recombinant Escherichia coli approach. This involved expressing seven carotenoid biosynthesis genes (crtE, crtB, crtI, crtY, crtZ, crtX and crtG). From the cells of the recombinant E. coli, caloxanthin (β,β-carotene-2,3,2′,3′-tetrol)-3′- β-d-glucose, zeaxanthin (β,β-carotene-3,3′-diol) 3,3′-β-d-diglucoside, and nostoxanthin (β,β-carotene-2,3, 3′-triol) (rare carotenoids) were isolated and identified. Caloxanthin 3′-β-d-glucose displayed potent ¹O₂ quenching activity (IC₅₀ 19 μM).

Full text of DOI:10.1016/j.phytochem.2011.02.017

Phytochemistry 72 (2011) 711–716
Contents lists available at ScienceDirect
Phytochemistry
journal homepage: www.elsevier.com/locate/phytochem
Production of caloxanthin 30-b-D-glucoside, zeaxanthin 3,30-b-D-diglucoside,
and nostoxanthin in a recombinant Escherichia coli expressing system harboring
seven carotenoid biosynthesis genes, including crtX and crtG
Ayako Osawa a, Hisashi Harada b, Seon-Kang Choi c, Norihiko Misawa b,1, Kazutoshi Shindo a,
⇑
a Department of Food and Nutrition, Japan Women’s University, Mejirodai, Bunkyo-ku, Tokyo 112-8681, Japan
b Central Laboratories for Frontier Technology, Kirin Holdings Co. Ltd., i-BIRD, Suematsu, Nonoichi-machi, Ishikawa 921-8836, Japan
c Marine Bio Industry Support Center, Gangnueung Science Industry Park, Daejeon-dong, Gangneung-city 210-340, South Korea
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 16 June 2010
Received in revised form 7 January 2011
Available online 21 March 2011
The aim of this study was to produce rare b-carotene-modified carotenoids possessing 2-O (-H or -glu)
and/or 3-O (-H or -glu) functionalities in their b-ionone ring(s) using a recombinant Escherichia coli
approach. This involved expressing seven carotenoid biosynthesis genes (crtE, crtB, crtI, crtY, crtZ, crtX
and crtG). From the cells of the recombinant E. coli, caloxanthin (b,b-carotene-2,3,20,30-tetrol)-30-b-
D
-glu-
-diglucoside, and nostoxanthin (b,b-carotene-2,3,30-
triol) (rare carotenoids) were isolated and identified. Caloxanthin 30-b- -glucose displayed potent 1O2
quenching activity (IC50 19 M).
cose, zeaxanthin (b,b-carotene-3,30-diol) 3,30-b-
D
Keywords:
D
Caloxanthin 30-b-
D
-glucoside
l
Zeaxanthin 3,30-b-
D-diglucoside
Ó 2011 Elsevier Ltd. All rights reserved.
Nostoxanthin
1O2 quenching activity
CrtX
CrtG
Recombinant Escherichia coli
1. Introduction
supply sufficient amounts of carotenoids (referred to as ‘‘rare
carotenoids’’), which hinder researchers from examining their bio-
logical activities.
More than 750 carotenoids with different molecular structures
have been isolated from natural sources (Britton et al., 2004); how-
ever, a few carotenoid species can be obtained in sufficient
amounts, including dicyclic carotenoids, such as b-carotene, a-car-
otene, b-cryptoxanthin, zeaxanthin, lutein, canthaxanthin,
astaxanthin and fucoxanthin, and an acyclic carotenoid, lycopene.
Among these carotenoids, b-carotene, a-carotene, b-cryptoxanthin,
zeaxanthin, lutein, fucoxanthin and lycopene are present in edible
plant sources, and their beneficial effects on human health such as
cancer prevention have been extensively examined (Hosokawa
et al., 2004; Nishino et al., 2000; Talegawkar et al., 2008).
Astaxanthin, which was isolated from green alga Haematococcus
pluvialis, was shown to inhibit low-density lipoprotein oxidation
in human and prevent diabetic nephropathy for diabetic db/db
mice (Iwamoto et al., 2000; Naito et al., 2004). With the exception
of such carotenoids, it is difficult to find natural sources that can
Engineering biosynthetic pathways in heterologous organisms
(pathway engineering) is one of the most powerful methods of
generating plenty of natural compounds of interest. With this ap-
proach, a variety of rare carotenoids, including structurally novel
ones, have been produced in either Escherichia coli or higher plants,
using various combinations of carotenoid biosynthesis genes that
were isolated from carotenogenic (carotenoids-producing) bacteria
(Lee et al., 2003; Misawa, 2010; Shindo et al., 2008). These carote-
nogenic bacteria belonged to the class Gamma- or Alpha-proteobac-
teria, which can synthesize derivatives of a plant-type cyclic
carotenoid b-carotene (Misawa, 2010; Nishida et al., 2005). A gene
cluster needed for the biosynthesis of a carotenoid glycoside was
first isolated from the soil bacterium Pantoea ananatis (re-classified
from Erwinia uredovora) belonging to Gammaproteobacteria, and
was structurally and functionally analyzed (Misawa et al., 1990).
In the carotenoid gene cluster of Pantoea species, zeaxanthin
(b,b-carotene-3,30-diol) glucosyltransferase, named CrtX, was
found to glucosylate zeaxanthin with UDP-glucose to yield
⇑
Corresponding author. Address: Department of Food and Nutrition, Japan
Women’s University, 2-8-1 Mejirodai, Bunkyo-ku, Tokyo 112-8681, Japan. Tel./fax:
+81 3 5981 3433.
zeaxanthin 3,30-b-
D-diglucoside (2) (structure shown in Fig. 1)
(Misawa et al., 1990; Nakagawa and Misawa, 1991; Hundle et al.,
1992). When genes encoding enzymatic steps from farnesyl
E-mail address: kshindo@fc.jwu.ac.jp (K. Shindo).
Present address: Research Institute for Bioresources and Biotechnology, Ishikawa
1
Prefectural University, Suematsu, Nonoichi-machi, Ishikawa 921-8836, Japan.
0031-9422/$ - see front matter Ó 2011 Elsevier Ltd. All rights reserved.
doi:10.1016/j.phytochem.2011.02.017

712
A. Osawa et al. / Phytochemistry 72 (2011) 711–716
Fig. 1. Structures of caloxanthin 30-b- -glucoside (1), zeaxanthin 3,30-b-
D D-diglucoside (2), and nostoxanthin (3).
diphosphate (FPP) to zeaxanthin (crtE, crtB, crtI, crtY, and crtZ),
which were present in the Pantoea gene cluster, in addition to
the crtX gene, were introduced into E. coli and expressed there,
thin 3,30-b-
D-diglucoside (2), and nostoxanthin (3) (rare
carotenoids) as the main pigments. The antioxidative (1O2 quench-
ing) activity of caloxanthin 30-b-
D-glucoside (1) was also examined.
the resultant bacterial cells synthesized zeaxanthin 3,30-b-
D-dig-
lucoside (2) as its predominant carotenoid product. This was also
the main pigment present in the Pantoea species (Nakagawa and
Misawa, 1991; Hundle et al., 1991). E. coli expressing the P. anana-
tis crtE, crtB, crtY, crtZ and crtX genes in addition to the crtI14 gene
(in vitro evolved gene of crtI) was shown to synthesize torulene-
monoglucoside (Lee et al., 2003). The crtX gene was also found in
a marine bacterium Paracoccus sp. strain N81106 (formerly called
Agrobacterium aurantiacum) that produces astaxanthin 3-b-gluco-
side (Maruyama et al., 2007; Yokoyama et al., 1995). On the other
hand, a marine bacterium, Brevundimonas sp. strain SD212, was
reported to produce trihydroxy-keto-carotenoids, such as
2-hydroxyastaxanthin (Yokoyama et al., 1996). A carotenoid 2,20-
hydroxylase gene, named crtG, was first isolated from this bacte-
rium, and E. coli that expressed the crtG gene in addition to the
crt genes needed for the biosynthesis of zeaxanthin from FPP was
shown to generate nostoxanthin (3) (b,b-carotene-2,3,20,30-tetrol)
by way of caloxanthin (b,b-carotene-2,3,30-triol) (Nishida et al.,
2005). The crtG genes were also identified in a soil bacterium
Brevundimonas vesicularis strain DC263 (Tao et al., 2006) and a cya-
nobacterium Thermosynechococcus elongatus strain BP-1 (Iwai
et al., 2008).
2. Results
2.1. Isolation and identification of carotenoids produced by
recombinant proteins
Recombinant E. coli cells harboring two plasmids pUCBreG-
CAR1 (Fig. 2) and pAC-MeV (Harada et al., 2009) were cultured in
12 L medium, and collected by centrifugation. Carotenoids were
extracted by adding CH2Cl2:MeOH (1:1) solution after sonication
of the cells. The solution was then centrifuged, and the yellow
supernatant was concentrated to a small volume in vacuo, and
partitioned between n-BuOH/H2O without adjusting the pH. The
n-BuOH layer was evaporated to dryness and analyzed by thin-
layer chromatography (TLC) on silica gel (E. Merck 60 F-254
0.25 mm silica gel plates) using CH2Cl2–MeOH (10:1). By TLC anal-
ysis, three yellow spots were observed at Rf 0.4, 0.3, and 0.2, with
compounds at Rf 0.4 and 0.2 identified as nostoxanthin (3) and
zeaxanthin 3,30-b-
D-diglucoside (2) (Fig. 1), respectively, by HPLC
analyses using the method of standard addition. The compound
at Rf 0.3 (1) was not identified. The ratio of each peak area (at
450 nm) was nostoxanthin (3):unidentified carotenoid (1):zeaxan-
The present study shows that co-expression of the P. ananatis
carotenogenic genes (crtE, crtB, crtI, crtY, crtZ, and crtX) and the
Brevundimonas SD212 crtG gene in E. coli results in the production
thin 3,30-b-
D-diglucoside (2) = 2:1:2, approximately.
To determine the structure of 1, the n-BuOH extract was
subjected to silica gel chromatography, and the fractions
of caloxanthin 30-b-
D-glucoside (1) (a novel carotenoid), zeaxan-

A. Osawa et al. / Phytochemistry 72 (2011) 711–716
713
Fig. 2. Depiction of the reconstituted carotenoid biosynthetic gene cluster in plasmid pUCBreG-CAR1.
Table 1
containing 1 were collected and concentrated to give a yellow oil.
1H NMR spectroscopic data for caloxanthin 30-b-D-glucoside (1), and 1H and 13C NMR
data for caloxanthin 30-b-D-glucoside hexa-acetate (4).
The yellow oil was further subjected to preparative silica gel HPLC
and ODS HPLC to give pure 1 (caloxanthin b-D
-30-glucoside,
1
4
3.2 mg).
Position
dH (J in Hz)
dH (J in Hz)
dC
1
2
3
4
40.6 (s)
77.2 (d)
68.8 (d)
37.5 (t)
3.26 (d, 10.3)
3.79 (m)
2.12 (dd, 8.9, 15.9)
2.49 (dd, 3.7, 15.9)
5.04 (d, 10.4)
5.13 (m)
2.23 (dd, 2.5, 17.5)
2.63 (dd, 6.4, 17.5)
2.2. Structural elucidation
Compound 1 was dissolved in MeOH and analyzed by positive
ion HRESI-MS. The (M+Na)+ peak was observed at m/z 769.46664
{Calcd. for 769.46654 (C46H66O8Na)}, and the molecular formula
of 1 was determined to be C46H66O8. Since 1 was constituted from
46 carbons, it was proposed to be a monoglucoside of a C40
carotenoid.
5
6
7
8
124.9 (s)
136.9 (s)
124.3 (d)
139.6 (d)
135.2 (s)
131.9 (d)
124.7 (d)
138.2 (d)
136.5 (s)
132.8 (d)
130.2 (d)
25.6 (q)
22.6 (q)
21.0 (q)
12.7 (q)
12.7 (q)
36.9 (s)
a
–
–
6.01 (d, 15.9)
6.12 (d, 15.9)
a
9
10
11
12
13
14
15
16
17
18
19
20
10
6.16 (d, 11.0)
6.65 (dd, 11.0, 15.0)
6.38 (d, 15.0)
6.15 (d, 11.0)
6.65 (dd, 11.0, 15.0)
6.38 (d, 15.0)
The 1H NMR spectrum for 1 in CD3OD showed the presence of a
b-pyranose {H-100 (d 4.47, J = 7.8 Hz)}, and 1H–1H DQF COSY analy-
sis of 1 indicated two sp3 vicinal spin networks [CH2 {H-40b (d 2.12)
and H-40a (d 2.44)}–CH(O) {H-30 (d 4.11)}–CH2 {H-20b (d 1.57) and
H-20a (d 1.89)}], and [CH(O) {H-2 (d 3.26)}–CH(O) {H-3 (d 3.79)}–
CH2 {H-4b (d 2.12) and H-4a (d 2.49)}], suggesting the attachment
of one OH function at H-3 in an b-ionone ring and two OH func-
tions at H-2 (or H-4) and H-3 in the other ring.
Acetylation of 1 with Ac2O in dry pyridine gave the hexa-acetyl
derivative 4 {ESI-MS m/z 999.6 (M+H)+}. Since 4 was highly soluble
in CDCl3, further structural studies were performed on 4. The 1H
NMR spectroscopic data for 4 (Table 1) showed 10 methyl singlets,
4 sp3 methylenes, 8 sp3 methines, and 14 sp2 methines, apart from
the signals derived from the six acetyl groups. The 13C NMR (Table
1) and DEPT spectroscopic experiments on 4 established presence
of 10 methyls, 4 sp3 methylenes, 8 sp3 methines, 2 sp3 quaternary
carbons, 14 sp2 methines, and 8 sp2 quaternary carbons, apart from
the resonances derived from the six acetyl groups. The 1H and 13C
chemical shifts of sp2 methine signals in 4 were very similar to
those of zeaxanthin (Molnar and Szabolcs, 1993), indicating they
possessed the same structure in the olefin regions.
6.27 (m)
6.66 (m)
1.12 (s)
0.99 (s)
1.75 (s)
1.99 (s)
1.99 (s)
6.26 (m)
6.65 (m)
1.08 (s)
1.00 (s)
1.72 (s)
1.96 (s)
1.97 (s)
20
1.57 (dd, 12.0, 12.3)
1.89 (dd, 5.0, 12.3)
4.11 (m)
2.12 (dd, 8.9, 16.9)
2.44 (dd, 6.3, 16.9)
1.53 (m)
1.86 (dd, 5.0, 12.3)
3.97 (m)
2.01 (m)
2.33 (dd, 6.3, 16.9)
45.8 (t)
30
40
73.4 (d)
39.0 (t)
50
125.2 (s)
138.2 (s)
125.4 (d)
138.6 (d)
135.6 (s)
131.4 (d)
124.9 (d)
138.0 (d)
136.4 (s)
132.6 (d)
130.0 (d)
28.7 (q)
30.1 (q)
21.6 (q)
12.7 (q)
12.7 (q)
99.5 (d)
71.8 (d)
72.9 (d)
68.7 (d)
71.5 (d)
62.3 (t)
60
a
70
–
–
6.11 (d, 15.9)
6.15 (d, 15.9)
a
80
90
100
110
120
130
140
150
160
170
180
190
200
100
6.16 (d, 11.0)
6.65 (dd, 11.0, 15.0)
6.38 (d, 15.0)
6.17 (d, 11.0)
6.64 (dd, 11.0, 15.0)
6.36 (d, 15.0)
6.27 (m)
6.66 (m)
1.09 (s)
1.08 (s)
1.72 (s)
1.99 (s)
1.99 (s)
4.47 (d, 7.8)
3.24 (dd, 7.8, 7.8)
3.42 (m)
3.42 (m)
3.25 (m)
6.26 (m)
6.65 (m)
1.06 (s)
1.06 (s)
1.70 (s)
1.96 (s)
1.97 (s)
4.65 (d, 7.9)
4.98 (dd, 7.9, 9.5)
5.22 (dd, 9.5, 9.5)
5.08 (dd, 9.5, 9.5)
3.72 (ddd, 2.6, 6.2, 9.5)
4.14 (dd, 6.2, 11.6)
4.26 (dd, 2.6, 11.6)
Analyses of the 1H–1H DQF COSY and HMQC spectra of 4 con-
firmed the presence of b-glucose [H-100 {d 4.65 (J = 7.9 Hz)}, H-200
{d 4.98 (J = 7.9, 9.5 Hz)}, H-300 {d 5.22 (J = 9.5, 9.5 Hz)}, H-400 {d
5.08 (J = 9.5, 9.5 Hz)}, H-500 {d 3.72 (J = 2.6, 6.2, 9.5 Hz)}, H-600 {d
4.14 (J = 6.2, 11.6 Hz) and d 4.26 (J = 2.6, 11.6 Hz)}] in 4. The down-
field shifts of H-2, H-3, H-200, H-300, H-400, and H-600 in 4 (Table 1)
indicated the presence of OH functions at each position in 1.
In the HMBC experiment using 4, the 1H–13C long-range cou-
plings from the singlet methyls to the corresponding carbons
(Fig. 3) confirmed that 4 and zeaxanthin possessed the same olefin
structure. In addition, the 1H–13C long-range coupling from H-14 (d
1.70) to C-4 (d 37.5), and from H-16 (d 1.00) and H-17 (d 1.08) to C-
2 (d 77.2) (Fig. 3) proved the presence of a 2,3-dihydroxy b-ionone
ring, and a large coupling constant of J2,3 (10.4 Hz) indicated a trans
relationship between 2-OH and 3-OH. The 1H–13C long range-cou-
pling from H-100 (d 4.65) to C-30 (d 73.4) (Fig. 3) proved the attach-
ment of b-glucose at C-30. All double bonds in the aglycone of 1 and
2 were thus determined to be in the E configuration by the J values
of the sp2 methines and 13C chemical shifts of the singlet methyls
of 4 (Fig. 3). The absolute configurations of C-2, C-3, and C-30 in 1
and 4 were judged to be all R, since the 3-hydroxylated b-ring
and the 2,3-dihydroxylated b-ring in carotenoids, which were pro-
duced in Brevundimonas sp. SD212, were shown to be 2R, and 2R,
3R, respectively (Yokoyama et al., 1996).
200
300
400
500
600
3.76 (dd, 5.3, 11.7)
3.88 (dd, 2.8, 11.7)
CH3 signals derived from acetates were observed at dY 2.01–2.10 in 4.
a
d 6.05–6.16, 4H (H-7, 70, 8, 80).
From these observations, the structure of 1 was determined to
be as shown in Fig. 1. The IUPAC-IUB semisystematic name of 1
is
30-b-glucopyranosyloxy-(2R,3R,300R)-b,b-carotene-2,3,30-triol,
and is a novel carotenoid.
2.3. Antioxidative activity
We examined the 1O2 suppression activity of caloxanthin 3-b-
D-
glucoside (1), and the IC50 value was 19 lM. We also examined the

714
A. Osawa et al. / Phytochemistry 72 (2011) 711–716
OAc
OAc
AcO
21.6
18'
δ
12.7
19
12.7
20
δ
δ
O
25.6
16
22.6
17
δ
δ
O
CH2OAc
3'
15.0 Hz
15.9 Hz
AcO
2
3
15.0 Hz
15.9 Hz
18
17'
16'
28.7
19'
12.7
20'
12.7
AcO
30.1
δ
δ
δ
δ
21.0
δ
Fig. 3. Key 1H–13C long range couplings, J values, and dc values observed in the NMR analyses of caloxanthin 30-glucoside hexa-acetate (4).
1O2 suppressing activity of astaxanthin, and b-carotene as a con-
trol. Their IC50 values were 19 and >100 lM, respectively.
recombinant E. coli. Thus, carotenoid 2,20-hydroxylase (CrtG) likely
does not accept the 3-glucosylated b-ionone ring (b-ring) in calo-
xanthin 3-b-D D-diglucoside (2)
-glucoside (1) or zeaxanthin 3,30-b-
as substrates, and zeaxanthin glucosyltransferase (CrtX) is likely
not to accept the 2,3-dihydroxylated b-ring in the caloxanthin or
nostoxanthin as substrates.
3. Discussion
In the present study, recombinant E. coli cells that harbored
plasmid pUCBreG-CAR1, which carries the P. ananatis crtE, crtB, crtI,
crtY, crtZ, and crtX genes and the Brevundimonas SD212 crtG gene,
On the other hand, when the crtW gene that encodes carotenoid
4,40-ketolase (oxygenase) was introduced into recombinant E. coli,
which expressed the crtE, crtB, crtI, crtY, crtZ, and crtX genes, the
resultant transformant was shown to produce astaxanthin(3,30-
were found to synthesize zeaxanthin 3,30-b-
D
-diglucoside (2) and
nostoxanthin (3), which are naturally rare carotenoids, as well as
the novel carotenoid, caloxanthin 30-b-
-glucoside (1). The biosyn-
dihydroxy-b,b-carotene-4,40-dione) 3,30-b-
D-diglucoside (Yokoy-
D
ama et al., 1998). It was thus thought that CrtX can glucosylate
both the 3-hydroxylated b-ring and 3-hydroxy, 4-ketolated b-ring.
thetic pathway to these carotenoids is proposed in Fig. 4. Nosto-
xanthin glucoside or diglucoside was not detected in the
lycopene
FPP
CrtE, CrtB, Crt I
CrtY
4'
3'
2
2'
3
4
β-carotene
CrtZ
OH
HO
zeaxanthin
CrtX
CrtG
zeaxanthin 3,3'-β-
D-diglucoside (2)
OH
HO
HO
caloxanthin
CrtX
CrtG
caloxanthin 3'-β-
D
-glucoside (1)
OH
HO
HO
OH
nostoxanthin (3)
Fig. 4. Biosynthesis of novel or naturally rare carotenoids with recombinant E. coli cells, into which seven bacterial carotenoid biosynthesis genes, crtE, crtB, crtI, crtY, crtZ, crtX,
and crtG, were introduced.

A. Osawa et al. / Phytochemistry 72 (2011) 711–716
715
The crtG genes have been isolated from three species of the
genus Brevundimonas, i.e., Brevundimonas sp. strain SD212, B. vesic-
ularis strain DC263 and Brevundimonas aurantiaca (ATCC15266),
and complementation experiments with these crtG were per-
formed using recombinant E. coli hosts that synthesized different
carotenoid substrates (Nishida et al., 2005; Tao et al., 2006). Conse-
quently, all the CrtG enzymes were demonstrated to convert zea-
xanthin into nostoxanthin (3). Whereas, the CrtGs from
Brevundimonas sp. SD212 and B. aurantiaca worked well on cantha-
xanthin or astaxanthin as substrates, and the CrtG from DC263 did
not work on either of the ketocarotenoiods (Tao et al., 2006). An-
other crtG gene was also identified in a cyanobacterium T. elongatus
strain BP-1 (Iwai et al., 2008). This gene was shown to have a CrtG
function not only converting zeaxanthin to nostoxanthin (3), but
also converting myxol into 2-hydroxymyxol by analyzing the crtG
disruptant mutant (Iwai et al., 2008). It is therefore reiterated that
CrtG can accept the 3-hydroxylated b-ring in zeaxanthin, caloxan-
thin, or myxol as substrates to form the 2,3-dihydroxylated b-ring.
On the other hand, some CrtGs including Brevundimonas SD212
and B. aurantiaca ones are likely to further accept both the 4-keto-
lated b-ring (in canthaxanthin) and the 3-hydroxy, 4-ketolated b-
ring (in astaxanthin).
crtE, crtX, crtY, crtI, crtB and crtZ, from soil bacterium P. ananatis
(Accession No. D90087) were isolated from pCAR1 (Misawa et al.,
1990), and ligated into the corresponding sites (downstream sites
of crtG) of pUCBre-O11containingthe crtGgene derived frommarine
bacterium Brevundimonas sp. strain SD212 (Accession No.
AB181388; Nishida et al., 2005). For carotenoid production by
E. coli, the plasmid pUCBreG-CAR1 was co-transformed into the
E. coli strain JM109 (ECOS Competent E. coli JM109; Nippon Gene,
Tokyo, Japan) along with plasmid pAC-MeV, which increased the
productivity of the carotenoids by engineering the upstream
isoprenoid biosynthetic pathway in E. coli (Harada et al., 2009).
Escherichia coli harboring two plasmids, pUCBreG-CAR1 and
pAC-MeV, was pre-cultured in 50 mL LB medium (yeast extract
5.0 g, NaCl 5.0 g, tryptone 10.0 g in 1 L) (Sambrook and Russell,
2001) containing ampicillin (Ap) (100
l
g/mL) and chlorampheni-
rotary shaker
col (Cm) (30 g/mL) at 30 °C for 12 h on
l
a
(150 rpm) (seed culture). One percent (v/v) of this culture was
inoculated into 50 mL 2 ꢀ YT (polypeptone 16 g, yeast extract
5.0 g, NaCl 5.0 g in 1 L) (Sambrook and Russell, 2001) medium con-
taining Ap and Cm, and cultivated at 30 °C until the OD600 of the
culture reached approximately 0.5. Then, IPTG (final concentration
1
l
M) was added and further cultivated at 20 °C for 24 h with
shaking at 150 rpm. Since there was a little effect on the amount
of glucosylated carotenoids, mevalonate substrate (e.g. -mevalon-
Plasmid pAC-MeV, which carries the mevalonate pathway
genes that include the isopenteny diphosphate (IPP) isomerase
(idi; type2) gene (Kaneda et al., 2001) from Streptomyces sp. strain
D
olactone) was not supplemented to the medium (data not shown).
The OD600 reached 6.9 after 24 h culture.
CL190 for D-mevalonate (D-mevalonolactone) utilization (Harada
et al., 2009), was used in the present study. The expression of the
idi gene in E. coli was shown to improve the productivity of FPP
and subsequent carotenoids (Harada and Misawa, 2009) even
without supplying mevalonolactone to the culture medium, and
structural identification of the products should be easier. In other
5.2. Isolation of caloxanthin glucoside (1)
Cells of E. coli from 12 L culture (240 flasks) were collected by
centrifugation at 13,000g. After removing the supernatant,
CH2Cl2:MeOH (1:1) solution was added to the cells, and the solu-
tion was sonicated for 5 min to extract the yellow pigments in
the cells. The solution was centrifuged, and the yellow supernatant
was collected. This procedure was performed twice, and the com-
bined yellow supernatant was concentrated to a small volume in
vacuo, and partitioned between n-BuOH/H2O without adjusting
the pH. The n-BuOH layer was next evaporated to dryness
(1.86 g), and the resulting residue was subjected to silica gel col-
umn chromatography (Silica Gel 60; KANTO CHEMICAL,
20 ꢀ 200 mm) using CH2Cl2–MeOH (7:1). The fractions containing
1 were collected and concentrated to give a yellow oil (0.41 g). The
yellow oil was subjected to preparative silica gel HPLC (Senshu Pak
PEGASIL silica 60-5, 10 ꢀ 250 mm, flow rate 3.0 mL/min) and sep-
arated with CH2Cl2–MeOH (10:1) as a solvent. The yellow peak at
Rt 11.0 min was collected and concentrated to give crude 1
(11.0 mg). Finally, crude 1 was subjected to preparative ODS HPLC
(Senshu Pak PEGASIL ODS, 20 ꢀ 250 mm, flow rate 8.0 mL/min)
and separated with CH2Cl2–MeOH (1:20) as a solvent. The yellow
peak at Rt 6.3 min was collected and concentrated to give pure 1
to achieve the efficient production of caloxanthin 30-b-
D-glucoside
(1) along with other naturally rare carotenoids, further study is
needed on the culture conditions of the recombinant E. coli cells.
The antioxidative activity (1O2 quenching activity) of the novel
carotenoid, caloxanthin 30-b-
D
-glucoside (1), was examined.
Compound 1 has a potent 1O2 quenching activity (IC50 19
M) as
astaxanthin (IC50 19
M). In a previous study, zeaxanthin 3,30-b-
-diglucoside (2) was proved to possess superior 1O2 quenching
activity to zeaxanthin (Tatsuzawa et al., 2000). Since b-carotene
did not show 1O2 activity (IC50 > 100
M), the production of carote-
l
l
D
l
noids with CrtG and CrtX may be useful to produce antioxidative
carotenoids.
4. Concluding remarks
The present study showed that a novel carotenoid, caloxanthin
30-b-
D
-glucoside (1), can be synthesized in E. coli cells along with
rare carotenoids [zeaxanthin 3,30-b-
D-diglucoside (2) and nosto-
xanthin (3)], by using the crtG and crtX genes in addition to the
crt genes required for the synthesis from FPP to b-carotene. These
dicyclic carotenoids include two b-ionone rings in their molecules,
which is the same backbone to b-carotene, zeaxanthin and viola-
xanthin, main carotenoids in plant leaves, although all the crt
genes originated from carotenogenic bacteria. Moreover, the func-
tional expression of such crt genes is likely to be facile not only in
E. coli but also in higher plants whose gene-transfer systems are
established, as needed (Misawa, 2009, 2010).
(caloxanthin b-D
-30-glucoside (1); yield 3.2 mg).
5.3. Spectroscopic analysis
The 1H and 13C NMR spectra were measured at 400 MHz with a
Bruker AMX400. The HRESI-MS was recorded with a Jeol JMS-
T100LP. The UV spectrum was recorded with a Hitachi U-3200
spectrophotometer.
5. Experimental
5.1. Plasmid construction, bacterial strains and growth conditions
5.4. Caloxanthin 30-b-
D-glucoside (1)
The plasmid used for production of caloxanthin glucoside (1)
(pUCBreG-CAR1; Fig. 2) was created as follows: A 6.9 kb-XbaI and
Hind III fragment, including the six carotenoid biosynthesis genes,
Yellow powder; UV (EtOH) kmax (log e) 450 (145), 478 (127) nm;
for 1H NMR spectroscopic data, see Table 1; HRESI-MS m/z
769.46664 [M+Na]+ (Calcd. for C46H66O8Na, 769.46654).

716
A. Osawa et al. / Phytochemistry 72 (2011) 711–716
proposed uridine diphosphate binding site. Proc. Natl. Acad. Sci. USA 89, 9321–
9325.
5.5. HPLC analysis of carotenoid pigments accumulated in E. coli
Iwai, M., Maoka, T., Ikeuchi, M., Takaichi, S., 2008. 2, 20-b-Hydroxylase (CrtG) is
involved in carotenogenesis of both nostoxanthin and 2-hydroxymyxol 20-
fucoside in Thermosynechococcus elongatus strain BP-1. Plant Cell Physiol. 49,
1678–1687.
The cultured cells of E. coli JM109 transformants, which were
obtained from a 50 mL culture, were extracted with CH2Cl2–MeOH
(1:1, v/v; 5 mL) solution. The supernatant was analyzed by high-
performance liquid chromatography (HPLC) (alliance; Hitachi)
with a photodiode array detector (PDA) (model L-7455; Hitachi).
This was performed using a Senshu Pak PEGASIL ODS column
(4.6 ꢀ 250 mm), and carotenoids were eluted at a flow rate of
1 mL/min with MeOH–H2O (90:10). Using this system, the reten-
Iwamoto, T., Hosoda, K., Hirano, R., Kurata, H., Matsumoto, A., Miki, W., Kamiyama,
M., Itakura, H., Yamamoto, S., Kondo, K., 2000. Inhibition of low-density
lipoprotein oxidation by astaxanthin. J. Atheroscler. Thromb. 7, 216–222.
Kaneda, K., Kuzuyama, T., Takagi, M., Hayakawa, Y., Seto, H., 2001. An unusual
isopentenyl diphosphate isomerase found in the mevalonate pathway gene
cluster from Streptomyces sp. strain CL190. Proc. Natl. Acad. Sci. USA 98, 932–
937.
Kobayashi, M., Sakamoto, Y., 1999. Singlet oxygen quenching ability of astaxanthin
esters from the green alga Haematococcus pluvialis. Biotechnol. Left 21, 265–
269.
tion times of zeaxanthin 3,30-b-
D
-diglucoside (2), caloxanthin 30-
b- -glucoside (1), and nostoxanthin (3) were 4.2, 5.1, and
D
6.6 min, respectively.
Lee, P.C., Momen, A.Z.R., Mijts, B.N., Schmidt-Dannert, C., 2003. Biosynthesis of
structurally novel carotenoids in Escherichia coli. Chem. Biol. 10, 453–462.
Maruyama, T., Kasai, H., Choi, S.-K., Ramasamy, A.K., Inomata, Y., Misawa, N., 2007.
5.6. 1O2 suppression experiment
Structure of
a complete carotenoid biosynthesis gene cluster of marine
bacterium Paracoccus sp. strain N81106. Carotenoid Sci. 11, 50–55.
Misawa, N., 2009. Pathway engineering of plants toward astaxanthin production.
Plant Biotechnol. 26, 93–99.
Misawa, N., 2010. Carotenoids. In: Mander, L., Lui, H.-W. (Eds.), Comprehensive
Natural Products II Chemistry and Biology, vol. 1. Elsevier, Oxford, pp. 733–753.
Misawa, N., Nakagawa, M., Kobayashi, K., Yamano, S., Izawa, Y., Nakamura, K.,
Harashima, K., 1990. Elucidation of the Erwinia uredovora carotenoid
biosynthetic pathway by functional analysis of gene products expressed in
Escherichia coli. J. Bacteriol. 172, 6704–6712.
The 1O2 quenching activity was examined by measuring meth-
ylene blue-sensitized photooxidation of linoleic acid (Kobayashi
and Sakamoto, 1999). 0.05 mM Methylene blue (40
linoleic acid (10 L) with or without carotenoid (40
centration 1–100 M, each dissolved in EtOH) were added to micro
l
L) and 2.4 M
l
lL) (final con-
l
glass vials (5.0 mL). The vials were tightly closed with a screw cap
with a septum, and mixtures were illuminated at 7000 lux at 22 °C
for 3 h in corrugated cardboard. Then, an aliquot of each reaction
Molnar, P., Szabolcs, J., 1993. (Z/E)-Photoisomerization of
iodine., J. Chem. Soc. Perkin Trans. 2, 261–266.
C40-carotenoids by
Naito, Y., Uchiyama, K., Aoi, W., Hasegawa, G., Nakamura, N., Yoshida, N., Maoka, T.,
Takahashi, J., Yoshikawa, T., 2004. Prevention of diabetic nephropathy by
treatment with astaxanthin in diabetic db/db mice. Biofactors 20, 49–59.
Nakagawa, M., Misawa, N., 1991. Analysis of carotenoid glycosides produced in
gram-negative bacteria by introduction of the Erwinia uredovora carotenoid
biosynthesis genes. Agric. Biol. Chem. 55, 2147–2148.
Nishida, Y., Adachi, K., Kasai, H., Shizuri, Y., Shindo, K., Sawabe, A., Komemushi, S.,
Miki, W., Misawa, N., 2005. Elucidation of a carotenoid biosynthesis gene cluster
encoding a novel enzyme, 2,20-b-hydroxylase, from Brevundimonas sp. strain
SD212 and combinatorial biosynthesis of new or rare xanthophylls. Appl.
Environ. Microbiol. 71, 4286–4296.
mixture (50 lL) was removed and diluted in EtOH (1.5 mL), and
the absorbance at 235 nm was measured to estimate formation
of conjugates dienes (Hirayama et al., 1994). The value in the ab-
sence of carotenoid was determined and the 1O2 repression activity
was calculated relative to this reference value. Activity is indicated
as the IC50 value representing the concentration at which 50% inhi-
bition was observed.
Nishino, H., Tokuda, H., Murakoshi, M., Satomi, Y., Masuda, M., Onozuka, M.,
Yamaguchi, S., Takayasu, J., Tsuruta, J., Okuda, M., Khachik, F., Narisawa, T.,
Takasuka, N., Yano, M., 2000. Cancer prevention by natural carotenoids.
Biofactors 13, 89–94.
Sambrook, J., Russell, D.W., 2001. Molecular Cloning: A Laboratory Manual, third ed.
Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY.
Acknowledgements
This work was partially supported by the New Energy and
Industrial Technology Development Organization (NEDO) of Japan,
and in part by the Research and Development Program for New
Bio-industry Initiatives (2006–2010) from the Bio-oriented Tech-
nology Research Advancement Institution (BRAIN) of Japan.
Shindo, K., Hasunuma, T., Asagi, E., Sano, A., Hotta, E., Minemura, N., Miyake, C.,
Maoka, T., Misawa, N., 2008. 4-Ketoantheraxanthin,
a novel carotenoid
produced by the combination of the bacterial enzyme b-carotene ketolase
CrtW and endogenous carotenoid biosynthetic enzymes in higher plants.
Tetrahedron Lett. 49, 3294–3296.
Talegawkar, S.A., Johnson, E.J., Carithers, T.C., Taylor Jr., H.A., Bogle, M.L., Tucker, K.L.,
2008. Carotenoid intakes, assessed by food-frequency questionnaires (FFQs), are
associated with serum carotenoid concentrations in the Jackson Heart Study:
validation of the Jackson Heart Study Delta NIRI Adult FFQs. Public Health Nutr.
11, 989–997.
Tao, L., Rouvière, P.E., Cheng, Q., 2006. A carotenoid synthesis gene cluster from a
non-marine Brevundimonas that synthesizes hydroxylated astaxanthin. Gene
379, 101–108.
References
Britton, G., Liaaen-Jensen, S., Pfander, H., 2004. Carotenoid Handbook. Birkhauser
Verlag, Basel.
Harada, H., Misawa, N., 2009. Novel approaches and achievements in biosynthesis
of functional isoprenoids in Escherichia coli. Appl. Microbiol. Biotechnol. 84,
1021–1031.
Harada, H., Yu, F., Okamoto, S., Kuzuyama, T., Utsumi, R., Misawa, N., 2009. Efficient
synthesis of functional isoprenoids from acetoacetate through metabolic
pathway-engineered Escherichia coli. Appl. Microbiol. Biotechnol. 81, 915–925.
Hirayama, O., Nakamura, K., Hamada, S., Kobayashi, Y., 1994. Singlet oxygen
quenching ability of naturally occurring carotenoids. Lipids 29, 149–150.
Hosokawa, M., Kudo, M., Maeda, H., Kohno, H., Tanaka, T., Miyashita, K., 2004.
Fucoxanthin induces apoptosis and enhances the antiproliferative effect of the
PPARg ligand, troglitazone, on colon cancer cells. Biochim. Biophys. Acta 1675,
113–119.
Hundle, B.S., Beyer, P., Kleinig, H., Englert, G., Hearst, J.E., 1991. Carotenoids of
Erwinia herbicola and an Escherichia coli HB101 strain carrying the Erwinia
herbicola carotenoid gene cluster. Photochem. Photobiol. 54, 89–93.
Hundle, B.S., O0Brien, D.A., Alberti, M., Beyer, P., Hearst, J.E., 1992. Functional
Tatsuzawa, H., Maruyama, T., Misawa, N., Fujimori, K., Nakano, M., 2000. Quenching
of singlet oxygen by carotenoids produced in Escherichia coli-attenuation of
singlet oxygen-mediated bacterial killing by carotenoids. FEBS Lett. 484, 280–
284.
Yokoyama, A., Adachi, K., Shizuri, Y., 1995. New carotenoid glucosides, astaxanthin
glucoside and adonixanthin glucoside, isolated from the astaxanthin-producing
marine bacterium Agrobacterium aurantiacum. J. Nat. Prod. 58, 1929–1933.
Yokoyama, A., Miki, W., Izumida, H., Shizuri, Y., 1996. New trihydroxy-keto-
carotenoids isolated from an astaxanthin-producing marine bacterium. Biosci.
Biotechnol. Biochem. 60, 200–203.
Yokoyama, A., Shizuri, Y., Misawa, N., 1998. Production of new carotenoids,
astaxanthin glucosides, by Escherichia coli transformants carrying carotenoid
biosynthetic genes. Tetrahedron Lett. 39, 3709–3712.
expression of zeaxanthin glucosyltransferase from Erwinia herbicola and
a