Structure of a Precursor to the Blue Components Produced in the Blue Discoloration in Japanese Radish (Raphanus sativus) Roots

Article abstract of DOI:10.1021/acs.jnatprod.6b00121

The internal blue discoloration in Japanese radish (Raphanus sativus L.) roots has been reported to be a physiological phenomenon after harvest and poses a significant problem for farmers. To avoid this discoloration, the fundamental development of new radish cultivars that do not undergo discoloration and/or improved cultivation methods is required. Elucidating the chemical mechanism leading to this discoloration could help overcome these difficulties. To determine the mechanism underlying this discoloration, this study was designed to probe the structure of a precursor to the blue components generated during the discoloration process. Soaking fresh roots in aqueous H₂O₂ resulted in rapid blue discoloration, similar to the natural discoloration. Using a H₂O₂-based blue discoloration assay, the precursor was extracted and isolated from the fresh roots and identified as the glucosinolate, 4-hydroxyglucobrassicin, via spectroscopy and chemical synthesis.

Full text of DOI:10.1021/acs.jnatprod.6b00121

Article
pubs.acs.org/jnp
Structure of a Precursor to the Blue Components Produced in the
Blue Discoloration in Japanese Radish (Raphanus sativus) Roots
Katsunori Teranishi*,† and Nagata Masayasu‡
†Graduate School of Bioresources, Mie University, 1577 Kurimamachiya, Tsu, Mie 514-8507, Japan
‡Food Research Institute, National Agriculture and Food Research Organization, 2-1-12 Kannondai, Tsukuba, Ibaraki 305-8642,
Japan
S
* Supporting Information
ABSTRACT: The internal blue discoloration in Japanese
radish (Raphanus sativus L.) roots has been reported to be a
physiological phenomenon after harvest and poses a significant
problem for farmers. To avoid this discoloration, the
fundamental development of new radish cultivars that do not
undergo discoloration and/or improved cultivation methods is
required. Elucidating the chemical mechanism leading to this
discoloration could help overcome these difficulties. To
determine the mechanism underlying this discoloration, this
study was designed to probe the structure of a precursor to the
blue components generated during the discoloration process.
Soaking fresh roots in aqueous H2O2 resulted in rapid blue
discoloration, similar to the natural discoloration. Using a H2O2-based blue discoloration assay, the precursor was extracted and
isolated from the fresh roots and identified as the glucosinolate, 4-hydroxyglucobrassicin, via spectroscopy and chemical
synthesis.
discoloration in some radish cultivars has been frequently
reported. Because the abnormal blue color decreases the
commercial value of the affected radish, this discoloration is a
serious problem for shopkeepers and farmers.
he internal blue discoloration in Japanese radish
T(Raphanus sativus L.) roots is a physiological phenomen-
on that occurs when harvested radish roots are stored at
approximately 20 °C for several days (Figure 1). Blue
components are generated in the roots, and the onset of this
A substantial volume of research on the physiological
phenomenon of the internal brown discoloration of radish
roots has been reported.1−4 However, few studies have
investigated blue discoloration, and the mechanism underlying
this phenomenon remains unknown. One reason for the lack of
research on blue discoloration is the belief that the blue
components comprise anthocyanins. In our preliminary
research, the blue components did not show the characteristics
of anthocyanins, including pH-dependent color changes and
resistance to reduction by ascorbic acid. It has been reported
that the incidence of blue discoloration depends on the radish
cultivar and the cultivation conditions, similar to brown
discoloration.5
To avoid blue discoloration after harvest, the fundamental
development of new radish cultivars that do not experience
discoloration and/or improved cultivation methods is needed.
Elucidating the chemical mechanism of the discoloration can
contribute to overcoming these difficulties. Additionally,
because this is the first time this discoloration has been
observed in white root vegetables, determining the mechanism
is chemically and biologically important. This study reports the
Figure 1. Time course for the discoloration of freshly harvested
Hukuhomare roots during storage at 20 °C.
Received: February 7, 2016
© XXXX American Chemical Society and
American Society of Pharmacognosy
A
DOI: 10.1021/acs.jnatprod.6b00121
J. Nat. Prod. XXXX, XXX, XXX−XXX

Journal of Natural Products
Article
development of a rapid and simple discoloration assay and the
isolation and chemical structural determination of the
precursors of blue components in Hukuhomare radish roots.
homogenization to partially extract the blue components. For
the latter method, it was hypothesized that ascorbic acid could
lead to discoloration of the blue components. Therefore,
ascorbate oxidase was added during homogenization to remove
the ascorbic acid by mild oxidation and thus suppress the color
fading. Figure 3 compares the visible absorption spectra of the
RESULTS AND DISCUSSION
■
Artificial Blue Discoloration. Rapid and simple methods
to artificially induce the discoloration of fresh Hukuhomare
root sections were investigated and showed that soaking
Hukuhomare root sections in aqueous H2O2 induced blue
discoloration. The discoloration increased with time as the root
sections were soaked in 0.29 M aqueous H2O2 at 20 °C, and
the degree of discoloration reached a steady state within 5 to 60
min (Figure 2).
Figure 3. Visible absorption spectra of the blue components. The
extracts were prepared from Hukuhomare root soaked in 0.29 M
aqueous H2O2 or stored at 20 °C for 4 days.
extract prepared from Hukuhomare root treated with 0.29 M
aqueous H2O2 according to the procedure described in the
Experimental Section and the extract prepared from Hukuho-
mare root stored at 20 °C according to the procedure reported
in the Experimental Section. The maximum absorptions in
these spectra were observed at ca. 628 nm, and HPLC with
PDA detection revealed that the extracts contain many blue
components and that their HPLC−PDA characteristics are
similar (Figure 4). In addition, the location of the discoloration
resulting from the treatment with H2O2 was the same as that
resulting from storage at 20 °C, as shown in Figures 1 and 2.
Figure 2. Time course for the discoloration of freshly harvested
Hukuhomare root sections by soaking in 0.29 M aqueous H2O2 at 20
°C.
To spectroscopically characterize the resulting blue compo-
nents, attempts were made to extract the blue components at 0
°C from roots treated with aqueous H2O2 and stored at 20 °C.
However, the extraction failed because the blue color dissipated
immediately after extraction. For the roots soaked in aqueous
H2O2, the blue color turned brown, and for the roots stored at
20 °C, the blue color disappeared. The instability of the blue
components is not surprising because the components resulting
from both treatment with H2O2 and storage at 20 °C
disappeared. Two different partial extraction methods for the
blue components generated by soaking in aqueous H2O2 and
storage at 20 °C were investigated. For the first method, it was
hypothesized that the decomposition of the blue components
could be promoted by enzymatic oxidation, such as peroxidase
and excess H2O2. Therefore, 2 mL of MeOH, which generally
inactivates enzymes, was added to 4 g of root during
Figure 4. HPLC chromatograms of the extracted components
prepared from Hukuhomare root stored at 20 °C for 4 days or
soaked in 0.29 M aqueous H2O2.
B
DOI: 10.1021/acs.jnatprod.6b00121
J. Nat. Prod. XXXX, XXX, XXX−XXX

Journal of Natural Products
Article
These results indicate that the blue discoloration can be
mimicked by H2O2 treatment.
Freshly harvested Hukuhomare roots were vertically cut
down the middle, and the xylem sections (10 × 10 × 10 mm)
at 50 mm from the root tip were extracted. The addition of 0.29
M aqueous H2O2 to the homogenate and filtrate of the sections
resulted in immediate discoloration (Figure 5), whereas the
Extraction and Isolation of the Precursor to the Blue
Components. To extract the precursor from a large amount of
root sections, the available techniques were severely limited
because of the lengthy time required, the instability of the
precursor, and the difficulty of isolating the root precipitate.
Thus, a method for large-scale extraction was devised.
Homogenization in the presence of acetone in a 2:1 ratio
relative to the root sections (v/w) prevented the enzymatic
decomposition of the precursor, decreased the viscosity of the
homogenate, and excluded components that were insoluble in
aqueous acetone. When equal amounts of acetone and root
sections (v/w) were used, the extract exhibited more gradual
blue discoloration in aqueous H2O2 in the absence of HRP
(Figure S2a, Supporting Information) than in the case of
extraction without acetone. Thus, using equivalent amounts was
less effective at inactivating the peroxidases. The addition of
aqueous H2O2 to the extract prepared using the 2:1 (v/w) ratio
generated no blue components (Figure S2b, Supporting
Information), and the extract exhibited discoloration of the
blue color in the presence of HRP and H2O2 (Figure S2c,
Supporting Information). These results indicated that the
peroxidases were inactivated and/or not contained in the
extract, and the precursor(s) were extracted as expected.
Therefore, twice the amount of acetone was employed for the
homogenization. In order to effectively perform the subsequent
column chromatography of the extract, unnecessary compo-
nents were removed as EtOH-insoluble precipitates by the
addition of excess EtOH after washing with EtOAc.
Figure 5. Time course for the discoloration of Hukuhomare root
filtrate (0.15 mL) by the addition of 0.29 M aqueous H2O2 (7.5 μL) at
20 °C.
addition of aqueous H2O2 to the precipitate caused no
discoloration, indicating that the component(s) involved in
the generation of the blue components exist in the filtrate.
Assay Method. The isolation of the precursor(s) from the
filtrate of freshly harvested Hukuhomare roots was next
attempted. First, the filtrate was adsorbed on a reversed-phase
column with ODS gel and then eluted by a stepwise increase
(10% steps) of the MeOH concentration from 0 to 100%. The
fractions were evaporated to the same volume as the filtrate.
The blue discoloration activities in the fractions (0.135 mL)
were assayed by adding 0.29 M aqueous H2O2 (7.5 μL), but no
fraction showed activity. However, a mixture reconstructed
from all of the fractions and the filtrate exhibited activity.
Combination experiments using each fraction were subse-
quently performed, and the addition of a small amount of the
0% MeOH fraction to the 10 and 20% MeOH fractions was
found to show activity. These results indicated that several
components in the 0, 10, and 20% MeOH fractions are
necessary to produce the blue discoloration using H2O2 and
that the 0% MeOH fraction must contain component(s) that
catalyze the discoloration. Adding EtOH, which generally
decreases enzyme activity, inhibited the discoloration in a dose-
dependent manner, as shown in Figure S1 (Supporting
Information). Thus, the component in the 0% MeOH fraction
must be an oxidation enzyme, such as peroxidase. Based on
these findings, commercially available horseradish peroxidase
(HRP) (15 and 150 units/mL, 0.015 mL) was added to the
assay solutions of the 10 and 20% MeOH fractions (0.135 mL)
with 0.29 M aqueous H2O2 (7.5 μL), and blue discoloration
was observed.
The isolation of the final extract solution was achieved using
HPLC with a reversed-phase column (Figure S3a, Supporting
Information). The blue discoloration activities of the fractions
were measured using the method outlined in the Experimental
Section (Figure S4, Supporting Information). HPLC analysis of
the fraction that exhibited the blue discoloration showed one
peak (Figure S3b, Supporting Information). The final yield of
purified precursor from 2 kg of root sections was approximately
40 mg, corresponding to approximately 95% of the precursor in
the final extract solution based on the HPLC analysis.
Chemical Structure of the Isolated Precursor to the
Blue Components. The isolated precursor to the blue
components was colorless and soluble in H2O, MeOH, and
EtOH but not in EtOAc. The UV−vis absorption spectrum in
H2O showed a peak at 268 nm (Figure S5, Supporting
Information). 13C NMR data revealed the presence of 16
different carbon atoms in the precursor molecule (Figure S6 in
the Supporting Information and Table 1). The HRMS data
revealed the molecular formula to be C16H19N2O10S2 (calcd m/
z 463.04866, obs m/z 463.04861 [M − H]−) (Figure S7a,
Supporting Information), and the HRMS/MS data indicated
the presence of a SO3H group (Figure S7b, Supporting
Information). COSY NMR data revealed the presence of one
glucose residue (Table 1 and Figure S8 in the Supporting
3
Information). The JG‑1,G‑2 value (9.8 Hz) of the glucosyl
moiety indicated the presence of a β-glucoside bond. The 13C
NMR chemical shift of the G-1 carbon was 22 ppm lower than
that of methyl β-D-glucopyranoside (Table S1, Supporting
Information), suggesting the existence of a thio- or amino-β-
glucoside bond. In addition, the 1H NMR chemical shift of the
G-1 proton was 0.66 ppm higher than that of methyl β-D-
glucopyranoside. This result indicated that an sp2 or sp atom
should be bound to the sulfur or nitrogen atom in the above β-
glucoside bond. The UV spectrum and the HMBC (Figure S9,
Supporting Information) and NOESY (Figure S9, Supporting
Various different experiments were subsequently devised to
optimize the amounts of H2O2 and HRP added to the
discoloration assay reaction solution using a microplate
absorbance reader. Finally, addition of 0.029 M aqueous
H2O2 (7.5 μL) and 150 units/mL HRP (15 μL) to the sample
solution (0.15 mL) was found to be optimum to explore the
precursor(s).
C
DOI: 10.1021/acs.jnatprod.6b00121
J. Nat. Prod. XXXX, XXX, XXX−XXX

Journal of Natural Products
Article
Cassel et al.12 (Scheme 1). The 1H and 13C NMR data,
Table 1. 1H and 13C NMR Data of the Isolated Precursor to
the Blue Components
a
25
HPLC−PDA−MS analyses, [α]D value, and UV spectrum of
the synthesized 1 were consistent with those of the isolated
precursor to the blue components, confirming that the isolated
precursor to the blue components is definitely 1.
b
position
δC, type
82.1, CH
δH (J in Hz)
4.67, d (9.8)
G-1
G-2
G-3
G-4
G-5
G-6
A
72.6, CH
78.0, CH
68.8, CH
80.4, CH
60.1, CH2
156.6, C
29.3, CH2
3.06, dd (8.5, 9.8)
2.99, dd (8.5, 8.5)
3.26, dd (8.5, 9.8)
3.12, m
Blue Discoloration Reaction of Compound 1. As seen
in Figure 6A, the blue discoloration of 1 occurred in the
reaction with H2O2 and HRP, and the generated blue
components were stable under the reaction conditions used.
However, the blue color afforded by soaking the root sections
in aqueous H2O2 or storing them at 20 °C turned brown over
time. Thus, it is assumed that the components in the roots
would affect the stability of the blue components. The HPLC−
PDA analyses of the discoloration over time revealed that many
types of blue components were produced during the initial
reaction period and after 18 h (Figure 6B). Given the
complicated root system, it seems reasonable to assume that
many types of blue components could be produced by soaking
the roots in aqueous H2O2 or storing them at 20 °C.
3.56, dd (4.3, 12.0), 3.67, d (12.0)
B
4.03, d (16.5), 4.50, d (16.5)
10.76, br s
I-1-NH
I-2
121.2, CH
6.96, br s
I-3
109.8, C
I-4
151.8, C
I-5
103.06 or 103.10, CH
121.9, CH
6.39, d (7.3)
I-6
6.87, dd (7.3, 7.9)
6.83, d (7.9)
I-7
103.06 or 103.1, CH
138.3, C
I-8
Compound 1 is known to be susceptible to oxidation and is
readily converted to blue compound(s).9,13 The oxidation
mechanism was investigated by Truscott and Manthey using 4-
hydroxyindole as a model.14 They reported that oxidation
proceeds via the formation of indol-(4,7)-p-quinone. Sub-
sequently, Latxague and Gardrat found evidence supporting
Truscott’s results.15 The UV−vis absorption spectrum of the
oxidation products reported in their study was similar to that of
the crude blue components obtained here (Figures 3 and 6A).
Therefore, glucosinolates and their derivatives possessing an
indol-(4,7)-p-quinone skeleton were likely produced via the
blue discoloration reaction in the roots.
Using the Hukuhomare radish, which is susceptible to blue
discoloration, an assay employing H2O2 and HRP was
developed to observe the discoloration precursor and identified
it as compound 1 by spectroscopy and chemical synthesis. This
is the first report indicating that 1 is the precursor to the blue
components in plants. These results should contribute to
elucidating the molecular physiological mechanism underlying
the blue discoloration.
I-9
115.9, C
I-4-OH
9.4, br s
a
Assignments are based on 2D NMR methods. The 1H and 13C NMR
data of the isolated precursor to the blue components were obtained in
b
DMSO-d6 at 30 °C. The positions of the atoms corresponding to the
1H and 13C NMR peaks are presented in the structure of 1.
Information) NMR spectra suggested that the precursor might
contain an indole moiety. Based on these spectroscopic data, it
was speculated that the precursor might be 4-hydroxygluco-
brassicin (1), which is a natural glucosinolate found in plants of
the Brassicaceae family.6,7 However, the 1H and 13C NMR data
of 1 have not been reported thus far, and therefore, we used the
data of desulfo-4-hydroxyglucobrassicin for comparison (Table
6
S1, Supporting Information). The H and 13C NMR data of
the isolated precursor to the blue components and those of
desulfo 4-hydroxyglucobrassicin were similar, strongly support-
ing the speculation regarding the structure of the precursor to
the blue components. The UV spectrum was also similar to that
of desulfo-4-hydroxyglucobrassicin.8
1
EXPERIMENTAL SECTION
■
General Experimental Procedures. Melting points were
determined using an ASONE ATM-01 apparatus and are uncorrected.
Optical rotation data were obtained using a P-2200 instrument
(JASCO Corp. Tokyo, Japan). UV−vis spectra were obtained with a
V-530 DS spectrometer (JASCO Corp.). IR data were obtained using
an FT/IR 410 spectrometer (JASCO Corp.). H, 13C, H−1H COSY,
13C−1H COSY, HMBC, and NOESY NMR spectra were measured on
a JNM-A500 spectrometer (JEOL, Tokyo, Japan) operating at 500
1
1
1
MHz for H NMR and 125 MHz for 13C NMR. TMS and acetone
were used as the internal standards for DMSO-d6 and D2O,
respectively. Chemical shift values and coupling constants are reported
as δ (ppm) and J (Hz), respectively. HRMS analysis was performed
using an LTQ Orbitrap Velos EDT Ultimate 3000 system and
Xcalibur version 2.2 SPI (Thermo Fisher Scientific Inc., MA., USA) in
ESI mode. HPLC−PDA−MS analyses were performed at 25 °C on a
JASCO Gulliver HPLC system equipped with an MD-910 detector
and a ZQ 4000 mass spectrometer (Waters Corporation, Milford, MA,
USA). Mass analyses were carried out in positive ESI mode. Thin-layer
chromatography was performed on Merck Kieselgel 60 F254 plates
(layer thickness, 0.25 mm; Merck KGaA, Darmstadt, Germany), and
the spots were visualized with a UV lamp (254 nm). Preparative
chromatography was performed on BW-200 silica gel columns (Fuji
Silysia, Aichi, Japan). Elemental analyses were achieved with a Yanaco
CHNCORDER MT-3 instrument. HRP and 2,3,4,6-tetra-O-acetyl-1-
Glucosinolates, including 1, are found in plants of the
Brassicaceae family, and in 2011, the total number of
documented natural glucosinolates exceeded 130.8 Compound
1, isolated from cabbage seed and rapeseed, was identified using
1
desulfonylation and subsequent GC/MS spectrometry and H
NMR data in 1982.9 Although many glucosinolates have been
synthesized,10 the synthesis of 1 has not been reported. Gardrat
et al. described the preparation of 4-benzyloxydesulfogluco-
brassicin from 4-benzyloxyindole for the synthesis of 1 in
1999,11 but since then, a synthesis of 1 has not been reported.
In the present study, the synthesis of 1 was achieved from 4-
hydroxyindole and 2,3,4,6-tetra-O-acetyl-1-thio-β-D-glucopyra-
nose using methods adapted from the procedures described by
D
DOI: 10.1021/acs.jnatprod.6b00121
J. Nat. Prod. XXXX, XXX, XXX−XXX

Journal of Natural Products
Article
Scheme 1. Chemical Synthesis of 1
Figure 6. Monitoring the reaction of 1 with H2O2 and HRP. (A) UV−vis absorption spectra and (B) HPLC chromatograms at 2 and 18 h after the
addition of H2O2 and HRP to the solution of 1. The reaction and HPLC conditions are described in the Experimental Section.
thio-β-D-glucopyranose were purchased from Sigma-Aldrich Co. (St.
Louis, MO, USA). Other chemicals were purchased from Wako Pure
Chemical Industries Ltd., Osaka, Japan.
Plant Material and Cultivation Conditions. R. sativus L. cv.
Hukuhomare radish seeds were purchased from the Mikado Kyowa
Seed Co., Ltd., in Japan. The radish cultivar was grown in the field of
the Ishikawa Sand Dune Agricultural Research Center in 2013 using
conventional methods. The sowing date was the September 5, and the
roots were harvested approximately 60 days after sowing.
Radish Root Internal Blue Discoloration at 20 °C and Using
H2O2. Freshly harvested radish roots were stored at 20 °C and
approximately 80% relative moisture content for 3−6 days in the dark.
Freshly harvested radish roots were vertically cut down the middle and
soaked in 0.29 M aqueous H2O2 at approximately 20 °C, and the
resulting discoloration was visually observed for 180 min.
Extraction of the Blue Components from Radish Roots
Treated with H2O2. Freshly harvested Hukuhomare roots were
vertically cut down the middle, and the sections were soaked in 0.29 M
aqueous H2O2 at 20 °C for 5 min. The roots were lightly rinsed with
distilled H2O. The blue portion (4 g) of the root was homogenized
with MeOH (2 mL) in an ice bath, and the homogenate was filtered
through absorbent cotton. The resulting filtrate was centrifuged at
10 000g and 0 °C for 5 min. The supernatant was filtered through a
0.45 μm membrane filter (Millipore Millex-LH 13NL, Merck
Millipore, and Darmstadt, Germany). UV−vis absorption spectra of
the filtrate were obtained using a V-530 DS spectrometer, and HPLC−
PDA analysis of the filtrate was performed immediately.
Extraction of the Blue Components from Radish Roots
Stored at 20 °C. Freshly harvested Hukuhomare roots were stored at
20 °C and approximately 80% relative moisture content for 4 days in
the dark. The roots were vertically cut down the middle, and the blue
portion (4 g) was homogenized with a mixture of 0.4 mL of 100 units/
mL ascorbate oxidase and MeOH (0.8 mL) in an ice bath. The
homogenate was filtered through absorbent cotton, and the resultant
filtrate was centrifuged at 10 000g and 0 °C for 5 min. The supernatant
was filtered through a 0.45 μm membrane filter. UV−vis absorption
spectra of the filtrate were obtained using a UV−vis spectrometer, and
HPLC−PDA analysis of the filtrate was performed immediately, as
outlined in the next section.
HPLC-PDA Analysis of Root Extracts. The HPLC−PDA
instrument was equipped with a Cosmosil 5C18-PAQ column. The
mobile phase was a mixture of aqueous 10 mM phosphate buffer (pH
7.0) (A) and MeOH (B). The flow rate was 0.8 mL/min in a linear
gradient starting with 0% B and reaching 100% B at 30 min.
E
DOI: 10.1021/acs.jnatprod.6b00121
J. Nat. Prod. XXXX, XXX, XXX−XXX

Journal of Natural Products
Article
Blue Discoloration Using the Filtrate. Freshly harvested
Hukuhomare roots were vertically cut down the middle, and xylem
sections (10 × 10 × 10 mm) at 50 mm from the root tip were picked
out. Each section was homogenized with 0.1 mL of MeOH in an ice
bath, and the homogenate was filtered through absorbent cotton. The
resulting filtrate was centrifuged at 10 000g and 0 °C for 10 min. The
supernatant was filtered through a 0.45 μm membrane filter. The
filtrate (0.15 mL) was added to a microquartz cell (Hellma Ultramicro
HRESIMS m/z 463.04861 [M − H]− (calcd for C16H19N2O10S2,
463.04866).
(E)-3-(2-Nitroethenyl)-1H-indole-4-ol (3). To a mixture of 4-
hydroxyindole (3 g, 0.023 mol), 1-(dimethylamino)-2-nitroethylene
(2.5 g, 0.022 mol), and CH2Cl2 (60 mL) was added TFA (3.4 mL,
0.046 mol) at 0 °C with stirring for 30 min. EtOAc (50 mL) was
added and the solution neutralized by addition of aqueous saturated
NaHCO3 (90 mL). The mixture was extracted with EtOAc (2 × 100
mL), and the combined EtOAc layers were dried with anhydrous
Na2SO4 and evaporated under reduced pressure. The resulting residue
was adsorbed on a silica gel column in EtOAc/n-hexane (1:1 v/v) and
eluted with the same solvent. The eluent was evaporated under
reduced pressure to afford crude compound 3 (3.6 g), which was used
in the next step without further purification. ESI-MS m/z 205.13 [M +
H]+ (calcd for C10H8N2O3 + H, 205.06).
Cell, Hellema Analytics, Mullheim, Germany), and the absorbance at
̈
628 nm was measured at 20 °C using a UV−vis spectrometer, with the
absorbance at 628 nm calibrated to 0. Aqueous H2O2 (0.29 M, 7.5 μL)
was added to the quartz cell at 20 °C, and the absorbance at 628 nm
was immediately recorded every 1 s for 5 min.
Measuring the Blue Discoloration Activity. Sample solutions
(0.15 mL) were placed in the wells of a 96-well plate, and 0.029 M
aqueous H2O2 (7.5 μL) and 150 units/mL HRP (15 μL) in 0.1 M
phosphate buffer (pH 7.0) were added. The absorbance at 620 nm was
measured at 20 °C every 2 min for 2 h using a microplate absorbance
reader (Sunrise Rainbow, Tecan Japan Co., Ltd., Kanagawa, Japan).
Extraction and Isolation of the Precursor to the Blue
Components. The precursor to the blue components was extracted
and isolated using the following sequence.
(E)-1-Acetyl-3-(2-nitroethynyl)-1H-indol-4-yl acetate (4). A mix-
ture of crude compound 3 (3.6 g), pyridine (30 mL), and Ac2O (30
mL) was stirred at room temperature for 1 h and the solvent
evaporated under reduced pressure. MeOH (5 mL) was added to the
residue, and the mixture was evaporated to dryness under reduced
pressure. The resulting solid was dissolved in CH2Cl2 and purified by
silica gel column chromatography using CH2Cl2 as the eluent. The
fraction containing compound 4 was evaporated under reduced
pressure and crystallized from a mixture of EtOAc and Et2O to give
yellow prisms of 4 (0.9 g, 14% yield from 4-hydroxyindole): mp 210−
212 °C; λmax/nm (log ε) (MeOH) 208 (4.25), 226 (4.24), 266 (4.06),
360 (4.11); IR (KBr) νmax 1742, 1632, 1544, 1509, 1432, 1375, 1336,
Xylem sections (150 g, at 50 to 150 mm from the root tip) of
freshly harvested Hukuhomare radish roots were homogenized with
300 mL of acetone in an ice bath using a Hsiangtai HG-200
homogenizer and then filtered through a glass filter. The filtrate was
evaporated under reduced pressure to approximately 100 mL. The
resultant concentrate was washed with EtOAc (2 × 150 mL). The
water extract was evaporated under reduced pressure to approximately
70 mL, and the concentrate was filtered through a 0.45 μm membrane
filter. The filtrate was evaporated to approximately 15 mL, to which 90
mL of EtOH was added with stirring at room temperature for 2 h. The
resulting supernatant was separated from the precipitate and rinsed
with EtOH. The EtOH solutions were combined and evaporated to
near dryness, and the residue was dissolved in H2O (30 mL). This
aqueous solution was designated as the “extract solution”. This
extraction process was also performed for 2 kg of radish root sections.
Isolation of the extract solution (0.4 mL) was achieved using a
JASCO Gulliver HPLC system with an MD-910 detector and a
Cosmosil 5C18-PAQ column (20 mm × 250 mm; Nacalai Tesque,
Inc., Kyoto, Japan). The mobile phase was a mixture of aqueous 10
mM phosphate buffer (pH 7.0) (A) and MeOH (B). A linear gradient
was used, starting with 5% B and reaching 30% B in 60 min at room
temperature with a flow rate of 5 mL/min. The precursor to the blue
components eluted between 30 and 35 min as a single peak. The
fraction containing the precursor to the blue components was
evaporated under reduced pressure to near dryness. HPLC purification
of the 2 kg of radish root sections was conducted. The residue
containing the precursor to the blue components was dissolved in
H2O (25 mL), and the resulting solution (1 mL) was purified with
HPLC using the aforementioned conditions. The fractions of
precursor to the blue components were pooled, evaporated under
reduced pressure to near dryness, and then dissolved in H2O (1 mL).
MeOH (15 mL) and EtOH (15 mL) were added to the solution. The
resulting suspension was centrifuged at 10 000g and 4 °C for 10 min,
and the supernatant was separated, which was evaporated to complete
dryness under reduced pressure to afford the pure precursor (40 mg)
to the blue components containing a small amount of phosphate solid.
The purity of the precursor was confirmed using HPLC−PDA−MS at
25 °C on a JASCO Gulliver HPLC system equipped with an MD-910
detector and a ZQ 4000 mass spectrometer. The HPLC elution was
conducted on a Cosmosil 5C18-PAQ column (4.6 mm × 250 mm) at
an elution rate of 0.8 mL/min with a linear gradient system of aqueous
10 mM phosphate buffer (pH 7.0) and MeOH (changing the ratio
from 100:0 to 0:100 in 30 min). Mass analysis was carried out in
negative ESI mode.
1
1301, 1234, 1197 cm−1; H NMR (DMSO-d6, 27 °C) δ/ppm 2.41
(3H, s, CH3), 2.69 (3H, s, CH3), 7.19 (2H, d, J = 7.9 Hz, indol-H5 or
7), 7.42 (1H, t, J = 7.9 Hz, indol-H6), 8.18 (1H, d, J = 13.4 Hz, C
CH), 8.23 (1H, d, J = 13.4 Hz, CCH), 8.25 (1H, d, J = 7.9 Hz,
indol-H5 or 7), 8.79 (1H, s, indol-H2) (see Figure S10, Supporting
Information); 13C NMR (DMSO-d6, 27 °C) δ/ppm 20.72 (CH3),
23.86 (CH3), 110.31 (C), 114.07 (CH, indol-C5 or 7), 118.16 (CH,
indol-C5 or 7), 120.07 (C), 126.25 (CH, indol-C6), 130.85 (CH, C
C), 131.32 (CH, indol-C2), 136.93 (CH, CC), 137.01 (C), 143.24
(C), 168.72 (C, CO), 169.80 (C, CO) (see Figure S11,
Supporting Information). Anal. Calcd for C14H12N2O5: C, 58.33; H,
4.20; N, 9.72%. Found: C, 58.39; H, 4.16; N, 9.68%.
2,3,4,6-Tetra-O-acetyl-β-D-glucopyranosyl-[(4-acetoxy-1-acetyl-
1H-indol-3-yl)methyl]thiohydroxymate (6). Compound 4 (0.5 g, 1.7
mmol) was dissolved in dry CH2Cl2 (25 mL). To the solution, were
added Et3SiH (0.36 mL, 2.3 mmol) and 0.91 M TiCl4 in dry CH2Cl2
(2.9 mL, 2.6 mmol) at 0 °C. After being stirred at 25 °C for 1 h, cold
H2O (50 mL) was added and the solution was extracted with CH2Cl2
(2 × 25 mL). The combined CH2Cl2 layers were dried with anhydrous
Na2SO4 and evaporated under reduced pressure to give crude (4-
acetoxy-1-acetyl-1H-indol-3-yl)ethanimidoyl chloride (5), which was
used in the next step without purification. To crude compound 5 in
CH2Cl2 (25 mL) were added 2,3,4,6-tetra-O-acetyl-1-thio-β-D-
glucopyranose (0.32 g, 0.88 mmol) and Et3N (0.72 mL, 5.2 mmol)
at 25 °C. After being stirred at 25 °C for 1 h, cold H2O (25 mL) and
aqueous 1 M HCl (5 mL) were added and the solution was extracted
with CH2Cl2 (2 × 25 mL). The CH2Cl2 layer was dried with
anhydrous Na2SO4 and evaporated under reduced pressure. The
residue was dissolved in CH2Cl2 and purified by silica gel column
chromatography using EtOAc/CH2Cl2 (1:4, subsequently 1:2 v/v).
The fraction containing compound 6 was evaporated under reduced
pressure, and compound 6 was crystallized from MeOH to afford pure
6 (0.25 g, 23% yield from compound 4): mp 235−237 °C; [α]D25 −25
(c 0.5, CHCl3); λmax/nm (log ε) (MeOH) 208 (4.24), 236 (4.30), 302
1
(3.85); IR (KBr) νmax 1756, 1433, 1370, 1219, 1047 cm−1; H NMR
(DMSO-d6, 27 °C) δ/ppm 1.92 (3H, s, CH3), 1.97 (3H, s, CH3), 2.00
(3H, s, CH3), 2.01 (3H, s, CH3), 2.28 (3H, s, CH3), 2.61 (3H, s,
CH3), 3.97 (1H, d, J = 17.1 Hz, CH2), 4.07 (1H, d, J = 10.4 Hz,
glucose-H6), 4.10 (1H, d, J = 17.1 Hz, CH2), 4.13 (1H, m, glucose-
H6), 4.23 (1H, m, glucose-H5), 4.93 (1H, t, J = 9.8 Hz, glucose-H2),
4.97 (1H, t, J = 9.8 Hz, glucose-H4), 5.45 (1H, t, J = 9.8 Hz, glucose-
H3), 5.76 (1H, d, J = 9.8 Hz, glucose-H1), 7.01 (1H, d, J = 7.9 Hz,
indol-H5 or 7), 7.29 (1H, t, J = 7.9 Hz, indol-H6), 7.73 (1H, s, indol-
Isolated Precursor to the Blue Components: White, amorphous
solid; [α]D25 −7 (c 0.4, H2O); UV (H2O) λmax 268 nm; IR (KBr) νmax
3410, 1638, 1252, 1060 cm−1; for 1H and 13C NMR data, see Table 1;
F
DOI: 10.1021/acs.jnatprod.6b00121
J. Nat. Prod. XXXX, XXX, XXX−XXX

Journal of Natural Products
Article
H2), 8.18 (1H, d, J = 7.9 Hz, indol-H5 or 7), 11.27 (1H, s, OH) (see
Figure S12, Supporting Information); 13C NMR (DMSO-d6, 27 °C)
δ/ppm 20.27 (CH3), 20.29 (CH3), 20.33 (CH3), 20.35 (CH3), 20.84
(CH3), 23.79 (CH3), 28.65 (CH2), 62.20 (CH2, glucose-C6), 68.04
(CH, glucose-C4), 69.73 (CH, glucose-C2), 72.85 (CH, glucose-C3),
74.27 (CH, glucose-C5), 77.85 (CH, glucose-C1), 113.27 (CH, indol-
C5 or 7), 114.72 (C), 117.00 (CH, indol-C5 or 7), 122.62 (C), 124.88
(CH, indol-C6), 125.84 (CH, indol-C2), 136.97 (C), 143.41(C),
147.69 (C), 168.77 (C, CO), 169.07 (C, CO), 169.10 (C, C
O), 169.31 (C, CO), 169.58 (C, CO), 170.07 (C, CO) (see
Figure S13, Supporting Information). Anal. Calcd for C28H32N2O13S:
C, 52.83; H, 5.07; N, 4.40%. Found: C, 52.87; H, 5.01; N, 4.35%.
4-Hydroxyglucobrassicin (1). Compound 6 (0.1 g, 0.16 mmol) was
dissolved in dry pyridine (3 mL), and chlorosulfonic acid (0.21 mL,
3.1 mmol) was added to the solution at −50 °C. After being stirred at
25 °C for 20 h, aqueous saturated NaHCO3 (0.12 mL) was added and
the solution was extracted with CH2Cl2 (2 × 25 mL). The CH2Cl2
layer was dried with anhydrous Na2SO4 and evaporated under reduced
pressure. The residue was dissolved in MeOH (3 mL), to which 0.5 M
NaOMe in MeOH (0.56 mL, 0.28 mmol) was added at 25 °C. After 1
h, HOAc (0.032 mL) and 0.1 M phosphate buffer (pH 7.6) (1 mL)
were added to neutralize the solution, which was then evaporated
under reduced pressure. The residue was dissolved in H2O and filtered
through a 0.45 μm membrane filter. The filtrate was purified using a
JASCO Gulliver HPLC system with an MD-910 detector and a
Cosmosil 5C18-PAQ column (20 mm × 250 mm) using the elution
conditions described above. The fractions containing compound 1
were pooled, evaporated under reduced pressure to near dryness, and
dissolved in H2O (1 mL). MeOH (10 mL) was added, and the
resulting suspension was centrifuged at 10 000g and 4 °C for 10 min.
The supernatant was separated and evaporated under reduced pressure
to complete dryness to afford pure 1 (15 mg, 20% yield from 6)
research. We thank Daisuke Masuda for cultivating the
Hukuhomare radish roots. K.T. gratefully acknowledges
financial support from the Adaptable and Seamless Technology
Transfer Program through target-driven R&D, JST of Japan
(AS251Z00176N), and The Towa Foundation for Food
Science & Research.
REFERENCES
■
(1) Kawashiro, H.; Takeda, H. Bull. Chiba. Agric. Exp. Stn. 1983, 29,
63−70.
(2) Fukuoka, N.; Kano, Y. Bull. Ishikawa Agr. Coll. 1990, 20, 23−28.
(3) Kawai, T.; Hikawa, M.; Fujisawa, T. J. Japan Soc. Hort. Sci. 1992,
61, 339−346.
(4) Kawai, T.; Hikawa, M.; Fujisawa, T.; Ono, Y.; Ishibashi, E. J.
Japan Soc. Hort. Sci. 1993, 62, 165−172.
(5) Ikeshita, Y.; Ishibata, K.; Kanamori, Y. Annual Research Report of
Agricultrue, Forestry and Fisheries in Ishikawa Pref. 2010; Vol. 12, pp
10−11 (in Japanese).
(6) Agerbirk, N.; Petersen, B. L.; Olsen, C. E.; Halkier, B. A.; Nielsen,
J. K. J. Agric. Food Chem. 2001, 49, 1502−1507.
(7) Agerbirk, N.; Olsen, C. E. Phytochemistry 2012, 77, 16−45.
(8) Truscott, R. J.; Johnstone, P. K.; Minchinton, I. R.; Sang, J. P. J.
Agric. Food Chem. 1983, 31, 863−867.
(9) Truscott, R. J.; Burke, D. G.; Minchinton, I. R. Biochem. Biophys.
Res. Commun. 1982, 107, 1258−1264.
(10) Rollin, P.; Tatibouet, A. C. R. Chim. 2011, 14, 194−210.
̈
(11) Gardrat, C.; Darriet, K.; Jes
of the 10th International Rapeseed Congress, Canberra, Australia, 1999.
(12) Cassel, S.; Casenave, B.; Deleris, G.; Latxague, L.; Rollin, P.
́
o, B. Abstracts of Papers. Proceedings
́
́
Tetrahedron 1998, 54, 8515−8524.
(13) Sang, J. P.; Truscott, R. J. W. J. Assoc. Off. Anal. Chem. 1984, 67,
1
containing a small amount of phosphate solid. H NMR, 13C NMR,
829−833.
UV−vis, IR, and HPLC−PDA−MS analyses were performed, and the
(14) Truscott, R. J. W.; Manthey, M. K. J. Sci. Food Agric. 1989, 47,
191−195.
data were found to be consistent with those of the isolated precursor
25
to the blue components: [α]D −7 (c 0.4, H2O).
(15) Latxague, L.; Gardrat, C. Revue Franca
39, 9−13.
̧
ise des Corps Gras 1992,
Blue Discoloration Reaction of Compound 1. An aqueous
solution (0.5 mL) of 1 mM compound 1 was added to a microquartz
cell, and the absorption spectrum was measured at 20 °C using the V-
530 DS spectrometer. Aqueous H2O2 (0.29 M, 7 μL) and 1500 units/
mL HRP solution (10 μL) were added to the solution at 20 °C, and
UV−vis absorption spectra were collected. Simultaneously, a sample
(20 μL) of the solution was removed from the quartz cell for HPLC−
PDA analysis. HPLC−PDA analysis was performed using a JASCO
Gulliver HPLC system with the MD-910 detector. The HPLC was
equipped with a Cosmosil 5C18-PAQ column (4.6 mm × 250 mm),
and the mobile phase was a mixture of aqueous 10 mM phosphate
buffer (pH 7.0) (A) and MeOH (B). The flow rate was 0.8 mL/min in
a linear gradient starting with 0% B and reaching 100% B in 30 min.
ASSOCIATED CONTENT
* Supporting Information
■
S
The Supporting Information is available free of charge on the
ACS Publications website at DOI: 10.1021/acs.jnat-
prod.6b00121.
Figures S1−S13 and Table S1 (PDF)
AUTHOR INFORMATION
Corresponding Author
*Tel: +81-59-231-9615. Fax: +81-59-231-9615. E-mail:
teranisi@bio.mie-u.ac.jp.
■
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
We are grateful to Dr. H. Imaseki, Professor Emeritus, Nagoya
University, for his valuable suggestions at the start of this
■
G
DOI: 10.1021/acs.jnatprod.6b00121
J. Nat. Prod. XXXX, XXX, XXX−XXX