Metabolic analysis of the cinnamate/monolignol pathway in Carthamus tinctorius seeds by a stable-isotope-dilution method

Article abstract of DOI:10.1039/b616705e

The present study established a system for comprehensive metabolic analysis of the cinnamate/monolignol and lignan pathways by the use of a stable-isotope-dilution method. The system was successfully applied to characterization of the pathways in Carthamu

Full text of DOI:10.1039/b616705e

View Online / Journal Homepage / Table of Contents for this issue
PAPER
www.rsc.org/obc | Organic & Biomolecular Chemistry
Metabolic analysis of the cinnamate/monolignol pathway in Carthamus
tinctorius seeds by a stable-isotope-dilution method
a
Norikazu Sakakibara,†a Tomoyuki Nakatsubo,a Shiro Suzuki,‡a Daisuke Shibata,b Mikio Shimada§ and
Toshiaki Umezawa*a
Received 15th November 2006, Accepted 8th January 2007
First published as an Advance Article on the web 29th January 2007
DOI: 10.1039/b616705e
The present study established a system for comprehensive metabolic analysis of the cinnamate/
monolignol and lignan pathways by the use of a stable-isotope-dilution method. The system was
successfully applied to characterization of the pathways in Carthamus tinctorius cv. Round-leaved White
maturing seeds in combination with administration of stable-isotope-labelled precursors. Experimental
results obtained using this technique strongly suggested the intermediacy of ferulic acid in lignan
biosynthesis in the plant.
with CAOMT in the presence of S-adenosyl-L-methionine gave
rise to the corresponding methylation products, ferulic acid (3),
sinapic acid (5), and sinapaldehyde (10), respectively. Ferulic
acid (3) and coniferaldehyde (8) were hydoxylated efficiently to
afford 5-hydroxyferulic acid (4) and 5-hydroxyconiferaldehyde
(9), respectively, when incubated individually with F5H. How-
ever, 5-hydroxyconiferaldehyde (9) was found to inhibit the
CAOMT-catalyzed methylation of caffeic acid (2) and 5-
Introduction
The cinnamate/monolignol pathway, in which monolignols are
biosynthesized from cinnamic acids, supplies precursors for
various phenylpropanoid compounds such as lignins, lignans,
neolignans, norlignans, flavonoids, and stilbenes.1–3 As illustrated
in Fig. 1, many parallel routes can be envisaged for this pathway,
and it is very difficult to determine the physiologically important
routes among the various possibilities.
hydroxyferulic acid (4), while coniferaldehyde (8) inhibited the
For example, the pathway towards sinapyl alcohol (15) via
F5H-catalyzed 5-hydroxylation of ferulic acid (3), indicating that
ferulic acid (3), sinapic acid (5), and sinapaldehyde (10) had
5-hydroxyconiferaldehyde (9) and coniferaldehyde (8) are pre-
long been proposed for angiosperm syringyl lignin biosynthesis.4
This was based mostly on the fact that administration of 14C-
labelled p-hydroxycinnamic acids such as p-coumaric (1), caffeic
(2), ferulic (3), 5-hydroxyferulic (4) and sinapic (5) acids into
some plants, especially wheat (Triticum vulgare var. tharcher),
resulted in the formation of 14C-labelled syringyl lignin.5,6 Recently,
however, this pathway was challenged and a novel pathway via
caffeic acid (2), caffeoyl CoA, feruloyl CoA, coniferaldehyde
(8), 5-hydroxyconiferaldehyde (9), sinapaldehyde (10) and sinapyl
alcohol (15) was proposed for syringyl lignin biosynthesis in
angiosperm trees.7–9 Later, similar pathways were also proposed
for herbaceous model plants, arabidopsis (Arabidopsis thaliana)10
and alfalfa (Medicago sativa).11
ferred substrates of CAOMT and F5H, respectively.7,8 Therefore
CAOMT and F5H were renamed as 5-hydroxyconiferaldehyde
OMT (CAldOMT) and coniferaldehyde 5-hydroxylase (CAld5H),
respectively.7–9 The novel pathway accords in all respects with the
recent findings of lignin characters in transgenic plants where caf-
feoyl CoA O-methyltransferase (CCoAOMT) and/or CAldOMT
(CAOMT) were downregulated. The suppression of CCoAOMT
resulted in reduction of lignin content with unchanged sy-
ringyl/guaiacyl (S/G) ratio, whereas the downregulation of
CAldOMT (CAOMT) caused a significant decrease in S/G ratio
with essentially no change in lignin content and the appearance of
5-hydroxyguaiacyl unit.11–22 Thus, the current view of the cinna-
mate/monolignol pathway involves the 5-hydroxyconiferaldehyde
pathway for syringyl lignin biosynthesis.3,7–11
However, the conventional pathway via sinapic acid (5), sina-
paldehyde (10), and sinapyl alcohol (15) towards syringyl lignin
was re-proposed based on administration of labelled ferulic (3)
and sinapic (5) acids to black locust (Robinia pseudoacacia) and
oleander (Nerium indicum).23,24 Recently, a new and redundant
route for 3-hydroxylation of the p-hydroxyphenyl moiety has been
proposed,25,26 which involved the ester exchange catalyzed by
hydroxycinnamoyl CoA:shikimate hydroxyxinnamoyl transferase
(HCT) and the subsequent 3-hydroxylation of the formed p-
coumaroyl ester giving rise to the corresponding caffeoyl ester,
followed by its conversion to caffeoyl CoA by HCT (Fig. 1).25–28
These results suggest that physiologically important routes in
the cinnamate/monolignol pathway towards lignin biosynthesis
might vary with plant species. In addition, it seems plausible
that the physiologically important routes in the pathway and
The discovery of the novel pathway via 5-hydroxydroxy-
coniferyldehyde (9) was based on kinetic analysis of recombinant
caffeic acid O-methyltransferases (CAOMTs) and ferulic acid 5-
hydroxylase (F5H).7,8 Thus, individual incubation of caffeic acid
(2), 5-hydroxyferulic acid (4), and 5-hydroxyconiferaldehyde (9)
aResearch Institute for Sustainable Humanosphare, Kyoto University, Uji,
Kyoto, 611-0011, Japan. E-mail: tumezawa@rish.kyoto-u.ac.jp; Fax: +81
774 38 3682; Tel: +81 774 38 3625
bKazusa DNA Research Institute, 2-6-7 Kazusa-kamatari, Kisarazu, Chiba,
292-0818, Japan
† Present address: Faculty of Pharmaceutical Sciences at Kagawa Campus,
Tokushima Bunri University, 1314-1 Shido, Sanuki City, Kagawa, 769-
2193, Japan
‡ Present address: Institute of Sustainability Science, Kyoto University,
Uji, Kyoto 611-0011, Japan
§ Present address: Fukui University of Technology, 3-6-1 Gakuen, Fukui
910-8505, Japan
802 | Org. Biomol. Chem., 2007, 5, 802–815
This journal is
The Royal Society of Chemistry 2007
©

View Online
Fig. 1 The cinnamate/monolignol pathway. 1, p-coumaric acid; 2, caffeic acid; 3, ferulic acid; 4, 5-hydroxyferulic acid; 5, sinapic acid; 6,
p-coumaraldehyde; 7, caffealdehyde; 8, coniferaldehyde; 9, 5-hydroxyconiferaldehyde; 10, sinapaldehyde; 11, p-coumaryl alcohol; 12, caffeyl alcohol;
13, coniferyl alcohol; 14, 5-hydroxyconiferyl alcohol; and 15, sinapyl alcohol. Thick solid arrow: a lignan biosynthetic pathway in Carthamus tinctorius
proposed in this study, solid arrow: currently accepted lignin biosynthetic pathways, broken arrow: other routes in the cinnamate/monolignol pathway.
C4H, cinnamate 4-hydroxylase; C3H, p-coumarate 3-hydroxylase; CAldOMT, 5-hydroxyconiferaldehyde O-methyltransferase: 4CL, 4-hydroxycinnamate
CoA ligase; HCT, hydroxycinnmamoyl CoA :shikimate hydroxycinnamoyl transferase; CCoAOMT, caffeoyl CoA O-methyltransferase; CCR, cinnamoyl
CoA reductase; CAld5H, coniferaldehyde 5-hydroxylase; CAD, cinnmamyl alcohol dehydrogenase; SAD, sinapyl alcohol dehydrogenase.
genes for lignin synthesis in normal cell wall formation in
development might be different from those for biosynthesis of
other phenylpropanoid compounds, such as lignan, norlignan,
and neolignan and even from those for defense lignin formation
after fungal attack or physical wounding.
Thus, many questions remain unsolved in this field, and in the
post-genomic era, functional genomics including the identification
of the roles of each gene from different gene families and their
assignment to the appropriate step of metabolism is indispensable
towards answering these questions. One of the major concerns
associated with functional genomics is comprehensive analyses
of expressed mRNAs (transcriptomics), proteins (proteomics)
and metabolites (metabolomics), together with the integration of
omics data.29 However, no single technology for metabolomics,
such as a DNA sequencer for genomics or DNA arrays for
transcriptomics, is available.30,31 Practically, for comprehensive
analysis of metabolites, two strategies can be adopted.
Such a metabolic analysis of the lignan pathway was conducted
in combination with administration of stable-isotope-labelled
precursors.33
We have expanded this technique in this study to exam-
ine the upstream lignan pathway, that is, the formation of
matairesinol and its congeners from coniferyl alcohol and the
cinnamate/monolignol pathway which supplies coniferyl alcohol,
and established a system to comprehensively quantify metabolic
intermediates of such pathways by GC-MS. This new system has
been successfully applied to safflower (Carthamus tinctorius) seeds
to characterize both lignin and lignan biosynthesis (Figs. 1 and 2).
Results
Lignans, p-hydroxycinnamates, p-hydroxycinnamaldehydes and
p-hydroxycinnamyl alcohols in C. tinctorius seeds
One is non-targeted and high-throughput analysis with Fourier-
transform ion cyclotron resonance mass spectrometry (FT-ICR
MS) which can cover a larger number of compounds than any
other type of mass spectrometry. However, this technique is very
weak at quantitation, and is used for metabolic fingerprinting
purposes, but not for quantitative metabolic analysis.30,31 The other
concentrates on rather limited but still significant numbers of
metabolites through the use of gas chromatography-mass spectro-
metry (GC-MS), liquid chromatography-mass spectrometry (LC-
MS) etc., which allow us an accurate and comprehensive quantita-
tive analysis, or metabolic profiling,31,32 for the target compounds.
Recently, taking the latter strategy with GC-MS, we have
established a pathway towards a dibenzylbutyrolactone lignan,
yatein, from matairesinol in cow parsley (Anthriscus sylvestris).33
GC-MS analysis of the methyl alcohol extracts of Carthamus
tinctorius seeds from 12 days after flowering (DAF) indicated
the presence of p-coumaric acid (1), caffeic acid (2), ferulic
acid (3), sinapic acid (5), coniferyl alcohol (13), sinapyl alcohol
(15), secoisolariciresinol (18), matairesinol (19), arctigenin (20),
and trachelogenin (21). These compounds were identified by
comparison of their mass spectra and retention times on GC with
those of authentic samples [p-coumaric acid (1) (TMS ether): MS
m/z (EI) 308 (M+, 41.8%), 293 (56.0), 249 (37.4), 219 (46.7), 179
(15.2), tR = 8.1 min; caffeic acid (2) (TMS ether): MS m/z (EI)
396 (M+, 41.5%), 381 (13.1), 307 (6.9), 219 (53.4), 191 (9.9), tR =
9.4 min; ferulic acid (3) (TMS ether): MS m/z (EI) 338 (M+,
81.3%), 322 (53.5), 308 (48.8), 293 (30.2), 249 (34.9), tR = 9.3 min;
sinapic acid (5) (TMS ether): MS m/z (EI) 368 (M+, 83.7%), 353
This journal is
The Royal Society of Chemistry 2007
Org. Biomol. Chem., 2007, 5, 802–815 | 803
©

View Online
Fig. 2 The lignan pathway in Carthamus tinctorius. 16, pinoresinol; 17, lariciresinol; 18, secoisolariciresinol; 19, matairesinol; 20, arctigenin; and 21,
trachelogenin.
(46.5), 338 (88.4), 323 (27.9), tR = 11.2 min; coniferyl alcohol
(13) (TMS ether): MS m/z (EI) 324 (M+, 41.9%), 309 (14.9), 293
(34.9), 235 (16.3), 204 (25.6), tR = 8.0 min; sinapyl alcohol (15)
(TMS ether): MS m/z (EI) 354 (M+, 56.9%), 323 (32.6), 293 (11.6),
265 (15.1), 234 (18.6), tR = 9.0 min; secoisolariciresinol (18) (TMS
ether): MS m/z (EI) 650 (M+, 12.5%), 560 (27.6), 470 (14.6), 261
(56.1), 209 (100), tR = 11.1 min; matairesinol (19) (TMS ether):
MS m/z (EI) 502 (M+, 51.2%), 209 (100), 179 (30.1), tR = 16.1 min;
arctigenin (20) (TMS ether): MS m/z (EI) 444 (M+, 60.9%), 209
(100), 179 (35.4), 151 (32.2), tR = 16.5 min; and trachelogenin (21)
(TMS ether): MS m/z (EI) 532 (M+, 34.4%), 396 (17.1), 245 (82.9),
209 (100), 151 (80.5), tR = 14.3 min].
the seeds of 5–15 DAF, and the time-course data are summarized
in Fig. 3.
Analysis of stable-isotope incorporation after administration of
labelled precursors
The six stable-isotope-labelled compounds, [U-ring-13C6]phenyl-
alanine, p-[U-ring-13C6]coumaric acid (1-13C6), [2,5-2H2, OC2H3]-
ferulic acid (3-d5), [2,5,7-2H3]caffeic acid (2-d3), [3-OC2H3]-
coniferaldehyde (8-d3), and [3-OC2H3]sinapaldehyde (10-d3), were
administered individually to 7 and 13 DAF seeds of C. tinc-
torius (0.4 mg of each stable-isotope-labelled compound to
200 mg of seeds). The methyl alcohol extracts obtained fol-
lowing the administration were submitted to analysis of stable-
isotope incorporation into the p-hydroxycinnamic acids (1)–(5),
p-hydroxycinnamaldehydes (6)–(10), p-hydroxycinnamyl alcohols
(11)–(15) by GC-MS after b-glucosidase treatment (Table 1). In
addition, the stable-isotope incorporation into matairesinol (19)
was examined (Table 1).
The other possible metabolic intermediates shown in Figs. 1
and 2, namely, 5-hydroxyferulic acid (4), p-coumaraldehyde (6),
caffealdehyde (7), coniferaldehyde (8), 5-hydroxyconiferaldehyde
(9), sinapaldehyde (10), p-coumaryl alcohol (11), caffeyl alcohol
(12), 5-hydroxyconiferyl alcohol (14), pinoresinol (16), and lar-
iciresinol (17) were not detected in GC-MS analysis of methyl
alcohol extracts from the seeds of 5–15 DAF even after searching
using the corresponding authentic samples. The CoA esters (p-
coumaroyl, caffeoyl, feruloyl, 5-hydroxyferuloyl, and synapoyl
CoAs) were not analyzed because of their insufficient volatility
and stability for GC-MS measurement.
In summary, the following conversions shown in Scheme 3
2
were observed, while incorporation of 13C or H into the other
cinnmamic acids, cinnamaldehydes and cinnamyl alcohols shown
in Fig. 1 was not detected.
Simultaneous administration
Quantitation of p-hydroxycinnamic acids, p-hydroxycinnamal-
dehydes, p-hydroxycinnamyl alcohols, and lignans in C. tinctorius
seeds
Three sets of simultaneous administration experiments with
(a) [2,5-2H2, 3-OC2H3]ferulic acid (3-d5) and [U-ring-13C6]phenyl-
alanine (Phe-13C6); (b) [2,5-2H2, 3-OC2H3]ferulic acid (3-d5)
and p-[U-ring-13C6]coumaric acid (1-13C6); and (c) [2,5-2H2, 3-
OC2H3]ferulic acid (3-d5) and [2,5,7-2H3]caffeic acid (2-d3) were
conducted. Stable isotope incorporation into coniferyl alcohol
(13) and matairesinol (19) was examined by GC-MS. As
shown in Table 2, formation of [2H5]coniferyl alcohol (13-d5)
and [2H5]matairesinol (19-d5) was observed in all cases, and
the deuterium incorporation into matairesinol (19) (0.020–
0.0073 nmol mg−1) was not reduced, but slightly increased,
compared with those in the sole administration of [2,5-2H2, 3-
OC2H3]ferulic acid (3-d5) [0.0039 nmol mg−1, (Table 1)].
Quantitation of the compounds of the cinnamate/monolignol
and lignan pathways was conducted by
a stable-isotope-
dilution method with GC-MS. Deuterium-labelled and unla-
belled standards for the fifteen phenylpropanoid monomers [p-
hydroxycinnamic acids (1)–(5), p-hydroxycinnamaldehydes (6)–
(10), p-hydroxycinnamyl alcohols (11)–(15)] and two lignans
[arctigenin (20) and trachelogenin (21)] were synthesized by
organic chemical techniques, as summarized in Schemes 1 and
2. The standards for four lignans (pinoresinol (16), lariciresinol
(17), secoisolariciresinol (18), and matairesinol (19)) were prepared
previously.34–36 Using these standards, calibration curves for the
targets were established, and used for their quantitation. The
phenylpropanoid monomers were analyzed all at once by a single
injection to GC-MS, while the lignan fraction was analyzed with a
shorter GC column. The quantitative analysis was conducted for
Lignin analysis
Wiesner color reaction37,38 with phloroglucinol in hydrochloric
acid indicated clear red–purple color in the seed hulls but not
804 | Org. Biomol. Chem., 2007, 5, 802–815
This journal is
The Royal Society of Chemistry 2007
©

View Online
Scheme 1 Synthetic routes for p-[3,5,7-2H3]coumaric acid (1-d3), [2,5,7-2H3]caffeic acid (2-d3), p-[3,5,7-2H3]coumaraldehyde (6-d3), [2,5,7-2H3]-
caffealdehyde (7-d3), p-[3,5,7-2H3]coumaryl alcohol (11-d3), and [2,5,7-2H3]caffeyl alcohol (12-d3). Reagents and conditions: i, Br2, CHCl3, 60 ◦C, 18 h;
ii, BnBr, K2CO3, KI, DMF, 55 ◦C, 3 h; iii, CH(OMe)3, p-TsOH, MeOH, 0 ◦C, 10 min; iv, n-BuLi, MeO2H, Et2O, −35 ◦C, 0.5 h; v, 1 N HCl, acetone,
0
◦C, 5 min; vi, Ag2O, EtOH, 60 ◦C, 10 min; vii, CH3CHN2, Et2O, rt, 15 min; viii, LiAl2H4, Et2O, −35 ◦C, 0.5 h; ix, MnO2, CH2Cl2, rt, 28 h; x, H2, 10%
Pd–C, THF, MeOH, rt. 6 min; xi, malonic acid, pyridine, piperidine, aniline, 65 ◦C, 12 h; xii, monoethyl malonate, pyridine, piperidine, aniline, 60 ◦C,
11 h; xiii, TBDMSCl, imidazole, DMF, rt, 4 h; xiv, LiAlH4, Et2O, −35 ◦C, 0.5 h; xv, n-Bu4NF, THF, rt, 10 min.
in the inside, of matured C. tinctorius seeds (Fig. 4), suggesting
lignification in the seed hulls. This was confirmed by chemical
degradation analysis. Thus, the seed hulls from commercial
matured seeds were submitted to thioacidolysis.39 GC-MS analysis
of the thioacidolysis products indicated the presence of trithioethyl
phenylpropane compounds (Fig. 5) which are derived specifically
from b-O-4 substructures in lignin polymers. Both guaiacyl and
syringyl monomers derived from guaiacyl and syringyl lignins,
respectively, were detected, indicating the Carthamus seed hull
lignin is guaiacyl-syringyl lignin. This was further confirmed by
nitrobenzene oxidation,40 which gave vanillin and syringaldehyde
from the seed hulls (Fig. 6), whereas p-hydroxybenzaldehyde was
not detected (data not shown).
15 DAF seeds (Fig. 6). The final lignin amounts, 125 lg mg−1,
correspond to ca. 700 nmol of coniferyl alcohol unit mg−1. In
addition, nitrobenzene oxidation showed that lignin accumulation
started at 6 DAF. Taking these results together, the accumulation
of seed lignin was found to start about 6 DAF and terminated at
around 12 DAF.
Discussion
To quantify precisely and comprehensively the compounds
in the cinnamate/monolignol and lignan pathways, we em-
ployed a stable-isotope-dilution method using the corresponding
deuterium-labelled compounds as the internal standards. This
technique has several advantages. First, the loss of the target
compounds during extraction and derivatization can be ignored
due to the presence of the labelled internal standards. Second,
Next, seed lignin amounts were measured by the acetyl bromide
method.41 The seed lignin content increased from 9 DAF to 12
DAF, and there was no significant difference between 12 and
This journal is
The Royal Society of Chemistry 2007
Org. Biomol. Chem., 2007, 5, 802–815 | 805
©

View Online
Scheme 2 Synthetic routes for [3-OC2H3]ferulic acid (3-d3), [2,5-2H2, 3-OC2H3]ferulic acid (3-d5), [3-OC2H3]-5-hydroxyferulic acid (4-d3),
[3-OC2H3]sinapic acid (5-d3), [3-OC2H3]coniferaldehyde (8-d3), [3-OC2H3]-5-hydroxyconiferaldehyde (9-d3), [3-OC2H3]sinapaldehyde (10-d3),
[3-OC2H3]coniferyl alcohol (13-d3), [3-OC2H3]-5-hydroxyconiferyl alcohol (14-d3), and [3-OC2H3]sinapyl alcohol (15-d3). Reagents and conditions:
i, CH(OMe)3, p-TsOH, MeOH, 0 ◦C, 10 min; ii, H2, 10% Pd–C, THF, MeOH, rt, 1 h; iii, 1 N HCl, acetone, 0 ◦C, 3 min; iv, BnBr, K2CO3, KI,
DMF, rt, 12 h; v, C2H3I, DMF, rt, 8 h; vi, malonic acid, pyridine, piperidine, aniline, 65 ◦C, 12 h; vii, monoethyl malonate, pyridine, piperidine, aniline,
65 ◦C, 11 h; viii, TBDMSCl, imidazole, DMF, rt, 4 h; ix, LiAlH4, ether, −35 ◦C, 0.5 h; x, n-Bu4NF, THF, rt, 10 min; xi, MnO2, CH2Cl2, rt, 23 h; xii, 1 N
NaOH, I2, KI, H2O, rt, 8 h; xiii, CuSO4·5H2O, 4 N NaOH, 145 ◦C, 4 h; xiv, Cu, Na, MeOH, 128 ◦C, 3 h.
incomplete chromatographic separation can be overcome by mass
separation. Third, ion suppression in mass spectrometry due to a
significant reduction of the ion intensity of the target compound
by the coexisting compounds can be neglected. Fourth, this
technique, employing mass spectrometry, is highly sensitive. In
fact, when the deuterium-labelled internal standards were added
to the MeOH extracts of C. tinctorius seeds after b-glucosidase
treatment, measured amounts of sinapyl alcohol (15) decreased
about 2 orders of magnitude compared with the value obtained
when the standards were added in the initial stage of the extraction
(data not shown). This was not a surprising result, because it is
well-known among lignin chemists that sinapyl alcohol (15) is a
very unstable compound.42 This result clearly demonstrates the
effectiveness of the stable-isotope-dilution method, especially for
quantitation of unstable metabolites.
However, this technique requires a stable-isotope-labelled inter-
nal standard as well as an unlabelled authentic sample for each
target compound. In most cases, these labelled and unlabelled
compounds are not commercially available, and therefore, need
to be synthesized. We have chemically synthesized the required
labelled and unlabelled compounds. Schemes 1 and 2 outline
the synthesis of the deuterium-labelled compounds. Using the
labelled compounds together with the corresponding unlabelled
ones, we prepared calibration curves for the target compounds
and established a metabolite quantitation system for the cinna-
mate/monolignol and lignan pathways.
dependently varying metabolic flows towards monolignols, lignins,
and lignans are the most suitable systems.
Members of tribe Cynareae of family Asteraceae such as
edible burdock (Arctium lappa) were characterized for their
wide distribution of 4ꢀ-O-methyldibenzylbutyrolactone lignans
(e.g. arctiin = arctigenin glucoside), especially in their seeds.43–49
Recently, when we analyzed the stereochemical properties of
lignans,48,50,51 we found that the accumulation of matairesinol (19)
and arctigenin (20) in A. lappa seeds started at about 6 DAF
(data not shown). In addition, histochemical analysis with Wiesner
reagent (phloroglucinol–HCl) showed the purple coloration in the
seed hulls, strongly suggesting lignin deposition (data not shown).
Therefore, the seeds, where the biosynthesis of lignan and most
probably lignin is taking place during the seed maturation phase,
seem to be an attractive system to test the applicability of the
stable-isotope-dilution-method-based quantitation system to the
biosynthesis of monolignol and lignan. However, flowering in
this species often occurs more than one year after germination,
meaning it is not be feasible for our purpose. Instead, C.
tinctorius which is another member of the tribe and also produces
dibenzylbutyrolactone lignans in the seeds44–47 has been chosen
in the present work, because it flowers at about 2 months after
germination.
Preliminary GC-MS analysis of b-glucosidase-treated methyl
alcohol extracts from C. tinctorius seeds of 12 DAF indicated the
presence of several lignans and the phenylpropanoid monomers of
the cinnamate/monolignol pathway. Thus, p-coumaric (1), caffeic
(2), ferulic (3) and sinapic acids (5), and coniferyl (13) and sinapyl
Next, we tested the applicability of the system to metabolic
analysis of the pathways. For this purpose, plants which have time-
806 | Org. Biomol. Chem., 2007, 5, 802–815
This journal is
The Royal Society of Chemistry 2007
©

View Online
Fig. 3 Time course (from 5 to 15 day after flowering) of accumulation of p-hydroxycinnamates (1–5), p-hydroxycinnamaldehydes (6–10),
p-hydroxycinnamyl alcohols (11–15) and lignans (16–21). Each value is the average of three independent experiments. Standard deviations are indicated
by vertical bars. The panels are laid out according to the biosynthetic pathways (Figs. 1 and 2).
alcohols (15), as well as secoisolariciresinol (18), matairesinol
(19), arctigenin (20), and trachelogenin (21) were detected. In
contrast, other monomeric compounds, such as coniferaldehyde
(8) and sinapaldehyde (10) were not detected. The CoA esters
(p-coumaroyl, caffeoyl, feruloyl, 5-hydroxyferuroyl, and sinapoyl
CoAs), which are not sufficiently volatile and or stable to be
analyzed by GC-MS, were not pursued in the present study. In
addition, semi-quantitative GC-MS analysis of the methyl alcohol
extracts from 5 to 15 DAF C. tinctorius seeds showed rapid
increases in the amounts of dibenzylbutyrolactone lignans after
8 DAF, as in the case of A. lappa, indicating the drastic metabolic
change in the cinnamate/monolignol pathway as seeds mature.
Like the Arctium seeds, histochemical detection of cinnamalde-
hyde groups in lignins with Wiesner reagent showed the red–
purple coloration in C. tinctorius matured seed hulls (Fig. 4),
strongly suggesting lignin deposition. This was also confirmed by
chemical degradation of lignins. Thioacidolysis and nitrobenzene
oxidation of the powdered seed hulls after organic solvent
extraction gave the phenylpropanoid monomers, 1-guaiacyl-1,2,3-
trithioethylpropane and 1-syringyl-1,2,3-trithioethylpropane in
thioacidolysis (Fig. 5) and vanillin and syringaldehyde in nitroben-
zene oxidation, respectively, which are diagnostic of the presence
of guaiacyl and syringyl lignins.
pathway compounds and lignans at various time intervals. As
shown in Fig. 3, the detected compounds from 5 to 15 DAF
seeds were again p-coumaric (1), caffeic (2), ferulic (3) and
sinapic acids (5), and coniferyl (13) and sinapyl alcohols (15),
secoisolariciresinol (18), matairesinol (19), arctigenin (20), and
trachelogenin (21). On the other hand, none of the other com-
pounds, pinoresinol (16), lariciresinol (17), and 5-hydroxyferulic
acid (4), p-coumaraldehyde (6), caffealdehyde (7), coniferaldehyde
(8), 5-hydroxyconiferaldehyde (9), sinapaldehyde (10), p-coumaryl
alcohol (11), caffeyl alcohol (12), 5-hydroxyconiferyl alcohol (14)
were detected throughout the GC-MS measurement of the 5–15
DAF samples.
The amounts of matairesinol (19) reached the maximum at 10
DAF, which is followed by the accumulation peak of arctigenin
(20) (14 DAF) (Fig. 3). This is in good accordance with their
metabolic sequence; O-methyltransferase catalyzing methylation
of matairesinol (19) to afford arctigenin (20) was detected in
the seeds and a cDNA encoding the enzyme has been cloned.52
The accumulation of arctigenin (20) was followed by that of
trachelogenin (21) (Fig. 3), strongly suggesting that trachelogenin
(21) is formed by hydroxylation of arctigenin (20).
Three phenylpropanoid monomers, caffeic (2) and ferulic acids
(3) and coniferyl alcohol (13) exhibited a transient accumulation
around 5–7 DAF (Fig. 3), while the start of matairesinol (19)
accumulation occurs just after this 5–7 DAF period. Lignin depo-
sition also started around 6 DAF, as determined by nitrobenzene
oxidation analysis (Fig. 6). The amounts (2–4 nmol mg−1) of the
Having the preliminary data in hand, the quantitation system
was used for the metabolic analysis of the cinnamate/monolignol
and lignan pathways in C. tinctorius maturing seeds. This was done
by measuring the accumulation of the cinnamate/monolignol
This journal is
The Royal Society of Chemistry 2007
Org. Biomol. Chem., 2007, 5, 802–815 | 807
©

View Online
Scheme 3
Fig. 4 Phloroglucinol staining of C. tinctorius seed section. (a) before
staining, (b) after staining.
Fig. 5 Typical thioacidolysis monomers obtained from Catrhamum
tinctorius seed hulls.
monomers are comparable to those of lignans (e.g. 10 nmol mg−1
for matairesinol (19)) (Fig. 3). In sharp contrast, lignin content
in the 14 DAF seeds was ca. 125 lg mg−1 (ca. 700 nmol of
coniferyl alcohol unit mg−1), which is two orders of magnitude
greater than the amounts of the cinnamate/monolignol pathway
compounds (Fig. 3). Obviously, the metabolic intermediate pools
in the cinnamate/monolignol pathway in C. tinctorius maturing
seeds are much smaller than the amounts of lignins.
808 | Org. Biomol. Chem., 2007, 5, 802–815
This journal is
The Royal Society of Chemistry 2007
©

View Online
Fig. 6 Amounts of lignin and nitrobenzene oxidation products: ᭛:
vanillin; ꢀ: syringaldehyde; ꢁ: lignin content measured by acetyl bromide
method.
Clearly, these results suggest that the monomers [i.e. caffeic
(2) and ferulic acids (3) and coniferyl alcohol (13)] accumulated
transiently during the 5–7 DAF are converted to lignans, but
probably not to lignins. In addition, because there is almost no
doubt that in C. tinctorius seeds guaiacyl lignin is biosynthesized
by the polymerization of coniferyl alcohol (13) as in other
plants,4,53 the biosynthesis of coniferyl alcohol (13) used for lignan
is probably regulated differently from that towards guaiacyl lignin
biosynthesis.
In contrast to coniferyl alcohol (13), the current view of lignin
biosynthesis does not involve ferulic acid (3) as a precursor.3,7–11
Hence, we conducted feeding experiments with labelled precursors
to confirm the intermediacy of ferulic acid (3) in lignan biosyn-
thesis. Administration of [2,5-2H2, 3-OC2H3]ferulic acid (3-d5) to
the seeds followed by comprehensive analysis of deuterium incor-
poration indicated that the deuterium atoms were incorporated
into matairesinol (19) as well as coniferyl alcohol (13), while no
deuterium incorporation into other phenylpropanoid monomers
(1, 2, 4–12, 14, 15) could be detected (Table 1). Similarly, when
p-[U-ring-13C6]coumaric acid (1-13C6) was administered, 13C incor-
poration into caffeic (2) and ferulic (3) acids, coniferyl alcohol
(13) and matairesinol (19) was observed (Table 1). Furthermore,
the administration of [2,5,7-2H3]caffeic acid (2-d3) resulted in
deuterium incorporation into ferulic acid (3), coniferyl alcohol
(13), and matairesinol (19) (Table 1). These results further suggest
that ferulic acid (3) is a precursor to lignans.
However, it should be noted that even compounds that are
not in the physiological metabolic route can be converted to the
pathway compounds, when administered exogenously, as pointed
out in our previous work.33 While this artifactual process has been
shown to take place,33 we have also demonstrated in our previous
study of the biosynthesis of a norlignan, cis-hinokiresinol,54 that
simultaneous or competitive administration of a true pathway
precursor together with the tested compound can inhibit the non-
physiological conversion. Since there is almost no doubt that
phenylalanine is a physiological precursor for lignans, we carried
out the simultaneous administration of [2,5-2H2, 3-OC2H3]ferulic
acid (3-d5) and [U-ring-13C6]phenylalanine (Table 2). In addition,
because p-coumaric (1) and caffeic acids (2) are possible precursors
of ferulic acid (3), and because the sole administration of
p-[U-ring-13C6]coumaric acid (1-13C6) and [2,5,7-2H3]caffeic acid
2
(2-d3) resulted in the incorporation of 13C and H, respectively,
This journal is
The Royal Society of Chemistry 2007
Org. Biomol. Chem., 2007, 5, 802–815 | 809
©

View Online
into ferulic acid (Table 1), two additional simultaneous ad-
ministration experiments were carried out: one with p-[U-ring-
13C6]coumaric acid (1-13C6) and [2,5-2H2, 3-OC2H3]ferulic acid
(3-d5) and the other with [2,5,7-2H3]caffeic acid (2-d3) and [2,5-
2H2, 3-OC2H3]ferulic acid (3-d5) (Table 2). The results showed
that these simultaneous or competitive administrations did not
lead to the reduction of deuterium incorporation from [2,5-2H2,
3-OC2H3]ferulic acid (3-d5) into matairesinol (19) (Table 2). Taken
together, these results reveal that ferulic acid (3) is a precursor of
lignans in C. tinctorius seeds.
Using the described metabolite quantitation system, we demon-
strate in this study the difference between the metabolic flow
towards lignans and that towards lignins. The fact that the
three monomeric precursors for lignans, caffeic (2) and ferulic
acids (3) and coniferyl alcohol (13), show similar accumulation
profiles indicates the fast conversion of caffeic acid (2) to coniferyl
alcohol (13), which is in accordance with no accumulation of
coniferaldehyde (8). However, the conversion of these transiently
accumulated monomers to matairesinol (19) proceeded within a
time span of a few days, which hold for further conversion of
matairesinol (19) to arctigenin (20) and trachelogenin (21) (Fig. 3).
This time span suggests a slow metabolism for lignans relative to
the lignin metabolic rate as follows.
Compared with the amounts of the lignins formed (125 lg mg−1,
correspond to 700 nmol of coniferyl alcohol unit mg−1), the accu-
mulation of metabolic intermediates in the cinnamate/monolignol
pathway is insignificant (2–4 nmol mg−1) and some, such as
sinapaldehyde (10), which is a precursor specific to syringyl lignin
but not lignans, are not even detectable (Fig. 3). In addition, b-
glucosidase treatment of the methyl alcohol extracts from the seeds
eliminates the possibility of the monolignol accumulation as their
glycosides, namely coniferin and syringin. This indicates that once
formed, the monomeric intermediates for lignins are converted
rapidly to their products, lignin polymers. Together, these results
strongly suggest a rapid metabolic flow for the biosynthesis of
lignin.
In our study of the pathway for sinapyl alcohol (15),
the precursor of syringyl lignin, administration of [3-OC2H3]-
coniferaldehyde (8-d3), [3-OC2H3]sinapaldehyde (10-d3), and [U-
ring-13C6]phenylalanine resulted in the formation of [2H3]sinapyl
alcohol (15-d3) and [13C6]sinapyl alcohol (15-13C6), respectively. As
importantly, deuterium incorporation from [3-OC2H3]ferulic acid
(3-d3) into sinapyl alcohol (15) was completely absent, though
it was incorporated into coniferyl alcohol (13) very efficiently.
These results are in excellent agreement with the current view that
syringyl lignin is formed from sinapyl alcohol (15) via conifer-
aldehyde (8), 5-hydroxyconiferaldehyde (9), and sinapaldehyde
(10), but not from ferulic acid (3).7–11 Furthermore, this accords
well with the kinetic properties of recombinant C. tinctorius
CAldOMT; a cDNA encoding CtCAldOMT was functionally
expressed in Escherichia coli and 5-hydroxyconiferaldehyde (9) was
found to be the best substrate of the recombinant CtCAldOMT
among the possible substrates of the cinnamate/monolignol
pathway. In addition, 5-hydroxyconiferaldehyde (9) inhibits
methylation of other substrates, such as caffeic acid (2) and
5-hydroxyferulic acid (4) (T. Nakatsubo et al., unpublished), like
sweetgum (Liquidambar styraciflua) and aspen (Populus tremu-
loides) CAldOMTs.8 Recently, Nair et al. proposed that in A.
thaliana, sinapic acid (5) was formed from sinapaldehyde (10).55
However, our [3-OC2H3]sinapaldehyde (10-d3) administration did
not result in deuterium incorporation into sinapic acid (5), though
it was efficiently reduced to [2H3]sinapyl alcohol (15-d3).
Taken together, these results demonstrate unequivocally that
the metabolite quantitation system coupled with administration
of stable-isotope-labelled precursors is a powerful tool for the
characterization of complex metabolic pathways or metabolic
profiling.
Experimental
Plant materials
Seeds of Carthamus tinctorius L. cv. Round-leaved White (Aster-
aceae) were purchased from Takii Seed, Co. Ltd. The seeds
were germinated and the plants were maintained in a controlled-
environment chamber (1.2 m × 1.2 m × 2.0 m; Koito-tron
KG50HLA; Koito Industries, Co., Ltd.) at 24 ◦C under a 14–10 h
light–dark regime.
Synthesis of monolignols and lignans
Synthesis of 5-hydroxyferulic acid (4), p-hydroxycinnamal-
dehydes (6)–(10), and p-hydroxycinnamyl alcohols (11)–(15). 5-
Hydroxyferulic acid (4),7 p-coumaraldehyde (6),9 caffealdehyde
(7),9 coniferaldehyde (8),56 5-hydroxyconiferaldehyde (9),7 p-
coumaryl alcohol (11),9 caffeyl alcohol (12),9 coniferyl alcohol
(13),34,57 and 5-hydroxyconiferyl alcohol (14)9 were prepared as
previously. Sinapaldehyde (10) and sinapyl alcohol (15) were
prepared from syringaldehyde in a similar manner to the synthesis
of 5-hydroxyconiferaldehyde (9)7 and 5-hydroxyconiferyl alcohol
(14).9
5-Hydroxyferulic acid (4): dH (acetone-d6) 7.52 (1 H, d, J 15.9),
6.90 (1 H, d, J 2.0), 6.82 (1 H, d, J 1.8), 6.32 (1 H, d, J 15.9),
3.88 (3 H, s); m/z (EI) 210.0553 (M+, 100%. C10H10O5 requires
210.0528), 195 (40.9), 168 (33.6), 166 (36.2), 128 (74.5), 114 (59.7).
p-Coumaraldehyde (6): dH (CDCl3) 9.65 (1 H, d, J 7.7), 7.49
(2 H, d, J 7.4), 7.42 (1 H, d, J 15.9), 6.89 (2H, d, J 7.2), 6.61
(1 H, dd, J 15.8 and 7.8); m/z (EI) 148.0538 (M+, 100%. C9H8O2
requires 148.0525), 131 (35.6), 119 (47.0), 107 (13.8), 91 (54.4).
Caffealdehyde (7): dH (acetone-d6) 9.62 (1 H, d, J 7.7), 7.52 (1 H,
d, J 15.8), 7.21 (1 H, d, J 2.1), 7.12 (1 H, dd, J 8.2 and 1.8), 6.90
(1 H, d, J 8.2), 6.53 (1 H, dd, J 15.8 and 7.7); m/z (EI) 164.0448
(M+, 100%. C9H8O3 requires 164.0474), 147 (47.6), 136 (36.9), 110
(36.5), 97 (34.5).
Coniferaldehyde (8): dH 9.64 (1 H, d, J 7.8), 7.40 (1 H, d, J 15.6),
7.13 (1 H, dd, J 7.8, 2.0), 7.06 (1 H, d, J 1.9), 6.97 (1 H, d, J 8.0),
6.59 (1 H, dd, J 15.8 and 7.8), 3.91 (3 H, s); m/z (EI) 178.0629
(M+, 100%. C10H10O3 requires 178.0629), 161 (11.5), 147 (24.4),
135 (33.3), 107 (15.2).
5-Hydroxyconiferaldehyde (9): dH (CDCl3) 9.65 (1 H, d, J 7.6),
7.34 (1 H, d, J 15.6), 6.87 (1 H, d, J 2.0), 6.70 (1 H, d, J 2.0),
6.58 (1 H, dd, J 15.7 and 7.7), 3.90 (3 H, s); m/z (EI) 194.0601
(M+, 100%. C10H10O4 requires 194.0583), 177 (14.3), 166 (52.9),
151 (98.5), 123 (32.3).
Sinapaldehyde (10): dH (CDCl3) 9.65 (1 H, d, J 7.6), 7.38 (1 H,
d, J 15.8), 6.80 (2 H, s), 6.60 (1 H, dd, J 15.8 and 7.8), 3.93 (6 H, s);
m/z (EI) 208.0719 (M+, 100%. C11H12O4 requires 208.0736), 180
(80.0), 165 (98.0), 149 (31.4), 137 (61.1).
810 | Org. Biomol. Chem., 2007, 5, 802–815
This journal is
The Royal Society of Chemistry 2007
©

View Online
p-Coumaryl alcohol (11): dH (CDCl3) 7.28 (2 H, d, J 8.6), 6.78
(2 H, d, J 8.6), 6.55 (1 H, d, J 15.9), 6.23 (1 H, dt, J 15.8 and
6.0), 4.30 (2 H, d, J 6.0); m/z (EI) 150.0670 (M+, 54.3%. C9H10O2
requires 150.0681), 147 (100), 131 (38.5), 121 (22.8), 107 (88.6), 94
(37.7).
Caffeyl alcohol (12): dH (acetone-d6) 6.93 (1 H, d, J 1.8), 6.73–
6.77 (2 H, m), 6.43 (1 H, d, J 15.8), 6.14 (1 H, dt, J 15.8 and
5.6), 4.18 (2 H, d, J 5.6); m/z (EI) 166.0625 (M+, 100%. C9H10O3
requires 166.0630), 148 (34.3), 138 (58.6), 123 (95.6), 110 (100),
101 (65.7).
Coniferyl alcohol (13): dH (CDCl3) 6.83–6.95 (3 H, m), 6.53
(1 H, d, J 15.9), 6.23 (1 H, dt, J 15.8 and 5.8), 4.29 (2 H, d, J 5.7),
3.83 (3 H, s); m/z (EI) 180.0789 (M+, 100%. C10H12O3 requires
180.0787), 162 (10.4), 137 (92.6), 124 (61.1), 119 (37.0), 91 (33.3).
5-Hydroxyconiferyl alcohol (14): dH (acetone-d6) 6.60 (1 H, d,
J 1.9), 6.58 (1 H, d, J 1.9), 6.43 (1 H, d, J 15.8), 6.19 (1 H, dt, J
15.8 and 5.6), 4.18 (2 H, d, J 5.5), 3.82 (3 H, s); m/z (EI) 196.0753
(M+, 100%. C10H12O4 requires 196.0736), 178 (14.3), 167 (49.1),
153 (97.6), 140 (35.1), 107 (18.6).
Sinapyl alcohol (15): dH (CDCl3) 6.63 (2 H, s), 6.53 (1 H, d, J
15.9), 6.24 (1 H, dt, J 15.8 and 5.8), 4.31 (2 H, dd, J 5.8 and 5.1),
3.90 (6 H, s); m/z (EI) 210.0911 (M+, 100%. C11H14O4 requires
210.0892), 194 (22.9), 182 (44.6), 167 (82.9), 154 (35.5), 149 (52.3),
128 (42.6).
[2,5-2H2]-3,4-Dibenzyloxybenzaldehyde dimethyl acetal (24b):
dH (CDCl3) 7.24–7.50 (10 H, m), 6.95 (1 H, s), 5.29 (1 H, s), 5.17
(2 H, s), 5.16 (2 H, s), 3.28 (6 H, s).
[3,5-2H2]-4-Benzyloxybenzaldehyde (25a): dH (CDCl3) 9.89 (1 H,
s), 7.84 (2 H, s), 7.28–7.50 (5 H, m), 5.15 (2 H, s).
[2,5-2H2]-3,4-Dibenzyloxybenzaldehyde (25b): dH (CDCl3) 9.81
(1 H, s), 7.26–7.50 (11 H, m), 5.26 (2 H, s), 5.22 (2 H, s).
Ethyl [3,5-2H2]-4-benzyloxybenzoate (26a): dH (CDCl3) 7.99
(2 H, s), 7.27–7.50 (5 H, m), 5.11 (2 H, s), 4.33 (2 H, q, J 7.1), 1.37
(3 H, t, J 7.2).
Ethyl [2,5-2H2]-3,4-dibenzyloxybenzoate (26b): dH (CDCl3) 7.63
(1 H, s), 7.27–7.50 (10 H, m), 5.22 (2 H, s), 5.20 (2 H, s), 4.43 (2 H,
q, J 7.2), 1.36 (3 H, t, J 7.2).
[3,5-2H2]-4-Benzyloxybenz(2H)aldehyde (27a): dH (CDCl3) 7.84
(2 H, s), 7.32–7.50 (5 H, m), 5.15 (2 H, s).
[2,5-2H2]-3,4-Dibenzyloxybenz(2H)aldehyde (27b): dH (CDCl3)
7.28–7.50 (11 H, m), 5.26 (2 H, s), 5.22 (2 H, s).
[3,5-2H2]-4-Hydroxybenz(2H)aldehyde (28a): dH (CDCl3) 7.81
(2 H, s).
[2,5-2H2]Protocatechu(2H)aldehyde (28b): dH (acetone-d6) 7.33
(1 H, s).
Ethyl p-[3,5,7-2H3]coumarate (29a): dH (CDCl3) 7.42 (2 H, s),
6.28 (1 H, s), 4.27 (2 H, q, J 7.2), 1.33 (3 H, t, J 7.2).
Ethyl [2,5,7-2H3]caffeate (29b): dH (acetone-d6) 7.03 (1 H, s), 6.25
(1 H, s), 4.17 (2 H, q, J 7.1), 1.26 (3 H, t, J 7.1).
Ethyl [3,5,7-2H3]-4-O-tert-butyldimethylsilyl-p-coumarate (30a):
dH (CDCl3) 7.41 (2 H, s), 6.28 (1 H, s), 4.25 (2 H, q, J 7.1), 1.32
(3 H, t, J 7.2), 0.98 (9 H, s), 0.21 (6 H, s).
Synthesis of p-[3,5,7-2H3]coumaric acid (1-d3), [2,5,7-2H3]-
caffeic acid (2-d3), p-[3,5,7-2H3]coumaraldehyde (6-d3), [2,5,7-2H3]-
caffealdehyde (7-d3), p-[3,5,7-2H3]coumaryl alcohol (11-d3), and
[2,5,7-2H3]caffeyl alcohol (12-d3). 2,5-Dibromoprotocatechu-
aldehyde (23b) was prepared by bromination of protocatechu-
aldehyde (22b).58 Phenolic hydroxyl and formyl groups of
3,5-dibromobenzaldehyde (23a) (Tokyo Chemical Industry, Co.,
Ltd.) and (23b) were protected, and resulting derivatives were
deuterated with n-butyllithium and methyl alcohol-d1 to give
dimethyl acetals (24a and b), respectively. The acetals (24a and
b) were hydrolyzed with 1 N HCl to give [2H2]benzyloxybenz-
aldehydes (25a and b). Then, 25a and 25b were oxidized with Ag2O,
and esterified to give ethyl [2H2]benzyloxybenzoates (26a and b).
Ethyl [2,5,7-2H3]-3,4-bis(O-tert-butyldimethylsilyl)caffeate (30b):
dH (CDCl3) 7.01 (1 H, s), 6.22 (1 H, s), 4.25 (2 H, q, J 7.2), 1.33
(3 H, t, J 7.2), 0.99 (9 H, s), 0.98 (9 H, s), 0.22 (6 H, s), 0.21
(6 H, s).
[3,5,7-2H3]-4-O-tert-Butyldimethylsilyl-p-coumaryl alcohol(31a):
dH (CDCl3) 7.26 (2 H, s), 6.23 (1 H, t, J 5.8), 4.28 (2 H, d, J 5.8),
0.97 (9 H, s), 0.19 (6 H, s).
[2,5,7-2H3]-3,4-Bis(O-tert-butyldimethylsilyl)caffeylalcohol(31b):
dH (CDCl3) 6.85 (1 H, s), 6.17 (1 H, t, J 5.9), 4.29 (2 H, dd,
J 5.9 and 5.8), 1.00 (9 H, s), 0.98 (9 H, s), 0.20 (6 H, s), 0.19
(6 H, s).
[3,5,7-2H3]-4-O-tert-Butyldimethylsilyl-p-coumaraldehyde (32a):
dH (CDCl3) 9.64 (1 H, d, J 7.8), 7.46 (2 H, s), 6.59 (1 H, d, J 7.8),
0.98 (9 H, s), 0.21 (6 H, s).
2
The esters were reduced with LiAl2H4 (Aldrich, H: 98 atom%),
and subsequent oxidation afforded [2H3]benzyloxybenzaldehydes
(27a and b). Then, 27a and 27b were debenzylated to give
[2H3]hydroxybenzaldehydes (28a and b). The deuterium-labelled
benzaldehydes 28a and 28b were converted to 1-d3 or 2-d3 by Kno-
evenagel reaction57 with malonic acid. Similarly, condensation of
28a and 28b with monoethyl malonate afforded ethyl cinnamates
(29a and b). The esters were protected with tert-buthyldimethylsilyl
chloride to give 30a and 30b, which were then reduced with LiAlH4
to the corresponding alcohols (31a and b). The cinnamyl alcohols,
31a and 31b, were oxidized and deprotected to afford deuterium
labelled cinnamaldehydes 6-d3 and 7-d3, while 31a and 31b were
deprotected directly to give 11-d3 and 12-d3, respectively. The
overall outline of the synthesis is displayed in Scheme 1.
2,5-Dibromoprotocatechualdehyde (23b): dH (acetone-d6) 10.12
(1 H, s), 7.63 (1 H, s).
[2,5,7-2H3]-3,4-Bis(O-tert-butyldimethylsilyl)caffealdehyde(32b):
dH (CDCl3) 9.65 (1 H, d, J 7.8), 7.06 (1 H, s), 6.54 (1 H, d, J 8.0),
1.00 (9 H, s), 0.99 (9 H, s), 0.23 (6 H, s), 0.22 (6 H, s).
p-[3,5,7-2H3]Coumaric acid (1-d3): dH (acetone-d6) 7.54 (2 H, s),
2
6.32 (1 H, s); m/z (EI) 167.0657 (M+, 100%. C9H5 H3O3 requires
167.0662), 166 (40.3), 165 (8.4), 164 (1.5), 150 (33.4), 122 (23.3),
94 (20.0).
[2,5,7-2H3]Caffeic acid (2-d3): dH (acetone-d6) 7.03 (1 H, s),
2
6.25 (1 H, s); m/z (EI) 183.0613 (M+, 100%. C9H5 H3O4 requires
183.0611), 182 (44.1), 181 (13.6), 180 (1.0), 166 (26.8), 149 (20.1),
136 (33.6), 92 (20.1).
p-[3,5,7-2H3]Coumaraldehyde (6-d3): dH (CDCl3) 9.65 (1 H, d,
J 7.7), 7.48 (2 H, s), 6.60 (1 H, d, J 7.8); m/z (EI) 151.0708 (M+,
2
100%. C9H5 H3O2 requires 151.0710), 150 (79.7), 149 (27.1), 148
[3,5-2H2]-4-Benzyloxybenzaldehyde dimethyl acetal (24a): dH
(CDCl3) 7.36 (2 H, s), 7.28–7.50 (5 H, m), 5.35 (1 H, s), 5.07
(2 H, s), 3.31 (6 H, s).
(1.1), 134 (30.0), 122 (37.3), 94 (42.3).
[2,5,7-2H3]Caffealdehyde (7-d3): dH (acetone-d6) 9.61 (1 H, d, J
7.8), 7.11 (1 H, s), 6.52 (1 H, d, J 7.8); m/z (EI) 167.0654 (M+,
This journal is
The Royal Society of Chemistry 2007
Org. Biomol. Chem., 2007, 5, 802–815 | 811
©

View Online
2
100%. C9H5 H3O3 requires 167.0662), 166 (39.0), 165 (14.2), 164
[3-OC2H3]-5-Hydroxyconiferaldehyde (9-d3): dH (acetone-d6)
9.62 (1 H, d, J 7.7), 7.50 (1 H, d, J 15.8), 6.96 (1 H, d, J 1.6),
6.89 (1 H, d, J 1.7), 6.61 (1 H, dd, J 15.8 and 7.8); m/z (EI)
(1.4), 150 (40.3), 139 (30.2), 112 (23.5), 92 (24.2).
p-[3,5,7-2H3]Coumaryl alcohol (11-d3): dH (CDCl3) 7.27 (2 H,
s), 6.22 (1 H, t, J 5.8), 4.30 (2 H, d, J 5.8); m/z (EI) 153.0855 (M+,
197.0759 (M+, 100%. C10H7 H3O4 requires 197.0768), 196 (7.8),
2
2
59.7%. C9H7 H3O2 requires 153.0869), 152 (13.6), 151 (7.8), 150
195 (1.0), 194 (0.0), 180 (14.1), 169 (37.9), 151 (62.4), 123 (31.5).
[3-OC2H3]Sinapaldehyde (10-d3): dH (CDCl3) 9.66 (1 H, d, J
7.6), 7.38 (1 H, d, J 15.6), 6.81 (2 H, s), 6.61 (1 H, dd, J 15.9
(1.2), 135 (23.5), 122 (10.1), 110 (100), 96 (50.3).
[2,5,7-2H3]Caffeyl alcohol (12-d3): dH (acetone-d6) 6.74 (1 H, s),
6.13 (1 H, t, J 5.6), 4.17 (2 H, d, J 5.6); m/z (EI) 169.0805 (M+,
and 7.8), 3.94 (3 H, s); m/z (EI) 211.0935 (M+, 100%. C11H9 H3O4
2
2
94.0%. C9H7 H3O3 requires 169.0818), 168 (20.3), 167 (7.8), 166
(3.7), 151 (67.1), 126 (100), 112 (53.7), 94 (60.4).
requires 211.0924), 210 (7.9), 209 (0.4), 208 (0.0), 183 (43.6), 168
(33.6), 140 (22.5), 122 (14.4), 105 (14.1).
[3-OC2H3]Coniferyl alcohol (13-d3): dH (CDCl3) 6.85–6.97 (3 H,
m), 6.53 (1 H, d, J 15.9), 6.22 (1 H, dt, J 15.6 and 5.8), 4.30 (2 H,
Synthesis of [3-OC2H3]cinnamic acids (3-d3)–(5-d3), [2,5-2H2,
3-OC2H3]ferulic acid (3-d5), [3-OC2H3]cinnamaldehydes (8-d3)–
(10-d3), and [3-OC2H3]cinnamyl alcohols (13-d3)–(15-d3). [3-
OC2H3]Cinnamic acids, [3-OC2H3]cinnamaldehydes, and [3-
OC2H3]cinnamyl alcohols were prepared from protocatechualde-
hyde (22b), as outlined in Schemes 1 and 2.
d, J 5.6); m/z (EI) 183.0981 (M+, 94.0%. C10H9 H3O3 requires
2
183.0974), 182 (6.7), 181 (2.7), 180 (0.0), 165 (23.5), 140 (100), 127
(53.0), 119 (33.9), 91 (37.3).
[3-OC2H3]-5-Hydroxyconiferyl alcohol (14-d3): dH (CDCl3) 6.65
(1 H, d, J 1.7), 6.50 (1 H, d, J 1.7), 6.46 (1 H, d, J 15.8), 6.22 (1 H,
dt, J 15.8 and 5.9), 4.28 (2 H, d, J 5.8); m/z (EI) 199.0915 (M+,
Briefly, [3-OC2H3]vanillin (33a) was prepared from proto-
catechualdehyde (22b) as described previously,33,57 and then
transformed to [3-OC2H3]-5-hydroxyvanillin (33b)33,59 and [3-
OC2H3]syringaldehyde (33c)33,60 by the methods reported previ-
ously. Similarly, [2,5-2H2, 3-OC2H3]vanillin (33d) was synthesized
from [2H2]dibenzylprotocatechualdehyde (25b) (Scheme 2).
[3-OC2H3]Ferulic acid (3-d3), [2,5-2H2, 3-OC2H3]ferulic acid
(3-d5), [3-OC2H3]-5-hydroxyferulic acid (4-d3), [3-OC2H3]sinapic
acid (5-d3), [3-OC2H3]coniferaldehyde (8-d3), [3-OC2H3]-5-
hydroxyconiferaldehyde (9-d3), [3-OC2H3]sinapaldehyde (10-d3),
[3-OC2H3]coniferyl alcohol (13-d3), [3-OC2H3]-5-hydroxyconiferyl
alcohol (14-d3), and [3-OC2H3]sinapyl alcohol (15-d3), were
prepared from the corresponding labelled benzaldehydes {[3-
OC2H3]vanillin (33a), [3-OC2H3]-5-hydroxyvanillin (33b), [3-
OC2H3]syringaldehyde (33c), and [2,5-2H2, 3-OC2H3]vanillin
(33d)} in the same manner as preparation of the corresponding
unlabelled compounds and labelled p-hydroxyphnenyl and 3,4-
dihydroxyphenyl compounds (Scheme 1).
2
100%. C10H9 H3O4 requires 199.0923), 198 (10.5), 197 (3.4), 196
(0.0), 181 (48.0), 171 (51.0), 156 (86.5), 135 (50.3).
[3-OC2H3]Sinapyl alcohol (15-d3): dH (CDCl3) 6.63 (2 H, s), 6.52
(1 H, d, J 15.6), 6.24 (1 H, dt, J 15.6 and 5.9), 4.32 (2 H, d, J 5.7),
3.90 (3 H, s); m/z (EI) 213.1093 (M+, 100%. C11H112H3O4 requires
213.1080), 212 (6.4), 211 (1.7), 210 (0.0), 195 (39.6), 185 (43.6),
170 (60.4), 157 (35.2), 149 (46.0).
Synthesis of p-[U-ring-13C6]coumaric acid. p-[U-ring-13C6]-
Coumaric acid (1-13C6) was prepared by Knoevenagel reaction
of [U-ring-13C6]-4-hydroxybenzaldehyde (CIL, 99 atom% 13C) and
malonic acid, as previously described.54,57
p-[U-ring-13C6]Coumaric acid (1-13C6): dH (acetone-d6) 7.60
(1 H, ddt, J 16.0, 5.3 and 1.8), 7.29–7.78 (2 H, m), 6.64–7.14
(2 H, m), 6.33 (1 H, dd, J 16.0 and 5.2); m/z (EI) 172 (0.9%),
171 (6.0), 170.0705 (M+, 100, C313C6H8O3, requires 170.0675), 169
(44.0), 168 (2.2), 167 (1.3), 166 (0.7), 165 (0.2), 164 (3.3), 125 (28.3),
124 (20.9), 113 (12.1), 96 (17.8).
[3-OC2H3]Ferulic acid (3-d3): dH (acetone-d6) 7.58 (1 H, d, J
15.8), 7.25 (1 H, d, J 2.0), 7.13 (1 H, dd, J 8.3, 1.9), 6.86 (1 H,
d, J 8.0), 6.36 (1 H, d, J 16.1); m/z (EI) 197.0757 (M+, 100%.
Synthesis of lignans. ( )-Pinoresinols (16),34 ( )-[9,9,9ꢀ,
9ꢀ-2H4]pinoresinols (16-d4)61, ( )-lariciresinols (17),35 ( )-[9,9,9ꢀ,
9ꢀ-2H4]lariciresinols (17-d4),61 ( )-secoisolariciresinols (18),34
( )-[3-OC2H3]secoisolariciresinols (18-d3),33 ( )-matairesinols
(19),36 ( )-[3-OC2H3]matairesinols (19-d3),33 and ( )-arctigenins
(20)36 were prepared previously. ( )-[3-OC2H3]Arctigenins
(20-d3) were prepared in the same manner as the prepara-
tion of ( )-arctigenins (20)36 but with [3-OC2H3]-4-benzyloxy-3-
methoxybenzyl bromide.33 ( )-Trachelogenins (21) were prepared
in the same manner as the synthesis of wikstromol62,63 but
with ( )-b-(3,4-dimethoxybenzyl)-c-butyrolactones prepared as
previously described,64 while ( )-[3-OC2H3]trachelogenins (21-d3)
were prepared exactly as that of the unlabelled (21) but with [3-
OC2H3]-4-benzyloxy-3-methoxybenzyl bromide.33
( )-[3-OC2H3]Arctigenins (20-d3): dH (CDCl3) 6.81 (1 H, d, J
7.8), 6.74 (1 H, d, J 8.3), 6.57–6.66 (2 H, m), 6.54 (1 H, dd, J 8.0
and 2.0), 6.46 (1 H, d, J 2.0), 4.13 (1 H, dd, J 9.2 and 7.2), 3.87
(1 H, dd, J 9.2 and 7.2), 3.84 (3 H, s), 3.80 (3 H, s), 2.95 (1 H, dd,
J 14.2 and 5.2), 2.88 (1 H, dd, J 14.2 and 6.6), 2.42–2.68 (4 H, m);
m/z (EI) 375.1787 (M+, 42.8%. C21H212H3O6 requires 375.1761),
374 (0.0), 373 (0.6), 372 (0.0), 337 (9.1), 224 (10.0), 177 (25.7), 151
(100), 140 (88.6), 107 (15.7).
2
C10H7 H3O4 requires 197.0768), 196 (10.2), 195 (0.0), 194 (0.0),
179 (26.9), 151 (13.4), 133 (19.5), 105 (10.1).
[2,5-2H2, 3-OC2H3]Ferulic acid (3-d5): dH (acetone-d6) 7.59 (1 H,
d, J 15.8), 7.13 (1 H, s), 6.37 (1 H, d, J 15.8); m/z (EI) 199.0883
2
(M+, 72.1%. C10H5 H5O4 requires 199.0892), 198 (41.4), 197 (15.2),
196 (1.5), 195 (0.0), 194 (0.0), 181 (19.2), 165 (13.5), 136 (50.0),
108 (91.3).
[3-OC2H3]-5-Hydroxyferulic acid (4-d3): dH (acetone-d6) 7.53
(1 H, d, J 15.9), 6.88 (1 H, d, J 2.0), 6.81 (1 H, d, J 2.0), 6.32 (1 H,
2
d, J 15.8); m/z (EI) 213.0710 (M+, 100%. C10H7 H3O5 requires
213.0716), 212 (8.4), 211 (0.0), 210 (0.0), 195 (22.8), 167 (16.8),
149 (17.1), 121 (17.8).
[3-OC2H3]Sinapic acid (5-d3): dH (acetone-d6) 7.35 (1 H, d, J
15.9), 6.78 (2 H, s), 6.17 (1 H, d, J 15.8), 3.66 (3 H, s); m/z (EI)
227.0864 (M+, 100%. C11H9 H3O5 requires 227.0873), 226 (3.7),
2
225 (0.0), 224 (0.0), 212 (17.3), 181 (9.7), 149 (12.8), 121 (13.4).
[3-OC2H3]Coniferaldehyde (8-d3): dH (CDCl3) 9.65 (1 H, d, J
7.8), 7.40 (1 H, d, J 15.6), 7.12 (1 H, dd, J 8.0, 2.0), 7.06 (1 H, d, J
1.9), 6.96 (1 H, d, J 8.3), 6.59 (1 H, dd, J 15.8 and 7.8); m/z (EI)
181.0793 (M+, 100%. C10H7 H3O3 requires 181.0818), 180 (31.1),
2
179 (0.0), 178 (0.0), 164 (21.8), 147 (34.2), 135 (38.6), 107 (26.8).
812 | Org. Biomol. Chem., 2007, 5, 802–815
This journal is
The Royal Society of Chemistry 2007
©

View Online
( )-Trachelogenins (21): dH (CDCl3) 6.85 (1 H, d, J 8.0), 6.79
(1 H, d, J 8.0), 6.60–6.74 (4 H, m), 3.98–4.07 (2 H, m), 3.86 (3 H,
s), 3.85 (3 H, s), 3.85 (3 H, s), 3.10 (1 H, d, J 13.9), 2.93 (1 H, d, J
13.6), 2.90–3.03 (1 H, m), 2.46–2.57 (2 H, m); m/z (EI) 388.1497
(M+, 88.6%. C21H24O7 requires 388.1522), 337 (16.0), 356 (8.6),
151 (100), 138 (88.0), 122 (33.4), 107 (28.6).
( )-[3-OC2H3]Trachelogenins (21-d3): dH (CDCl3) 6.85 (1 H, d,
J 8.0), 6.79 (1 H, d, J 8.0), 6.60–6.73 (4 H, m), 3.98–4.07 (2 H, m),
3.86 (3 H, s), 3.85 (3 H, s), 3.10 (1 H, d, J 13.7), 2.93 (1 H, d, J
13.9), 2.90–3.02 (1 H, m), 2.47–2.59 (2 H, m); m/z (EI) 391.1691
(M+, 100%. C21H212H3O7 requires 391.1710), 390 (0.0), 389 (0.0),
388 (0.0), 337 (9.7), 256 (18.6), 151 (78.9), 141 (87.1), 122 (31.4),
108 (40.0).
(11), Y = 0.76X + 0.07, 0.05–0.40 nmol range; caffeyl alcohol
(12), Y = 1.44X + 0.07, 0.05–0.36 nmol range; coniferyl alcohol
(13), Y = 0.26X + 0.03, 0.44–3.32 nmol range; 5-hydroxyconiferyl
alcohol (14), Y = 3.96X + 0.73, 0.04–0.31 nmol range; sinapyl
alcohol (15), Y = 0.72X + 0.07, 0.04–0.29 nmol range; pinoresinol
(16), Y = 2.71X + 0.73, 0.28–2.09 nmol range; lariciresinol (17),
Y = 1.31X + 0.36, 0.28–2.08 nmol range; secoisolariciresinol (18),
Y = 0.87X + 0.14, 2.20–16.56 nmol range; matairesinol (19), Y =
1.09X + 0.15, 2.23–16.75 nmol range; arctigenin (20), Y = 1.22X +
0.16, 2.15–16.12 nmol range; trachelogenin (21), Y = 0.93X +
0.33, 2.15–16.21 nmol range; where X represents the amounts of
each unlabelled compound (2–21) (nmol), and Y represents the ion
current ratio of unlabelled compounds (2–21)/deuterium-labelled
compounds (2-d3–15-d3, 16-d4, 17-d4, 18-d3–21-d3), respectively.
Chemicals
Quantitation of p-hydroxycinnamic acids, p-hydroxycinnamalde-
hydes, p-hydroxycinnamyl alcohols, and lignans. Maturing seeds
of C. tinctorius were harvested and washed successively with tap
and distilled water. Four seeds (ca. 0.2 g, each of 5–15 DAF) were
frozen (liquid N2), freeze-dried, powdered using a mortar and
pestle, and extracted with hot MeOH (3.0 ml). Then deuterium-
labelled internal standards [15 phenylpropanoid monomers (1-d3–
15-d3) and 6 lignans (16-d4, 17-d4, 18-d3–21-d3)] were added to the
MeOH solution. The MeOH solution thus obtained was extracted
with hexane, and the hexane layer was removed. Solvents of the
resultant MeOH solution were evaporated off, and the MeOH
extracts thus obtained were treated with b-glucosidase (from
almonds, SIGMA G-0395, 68 units in 1.0 ml of NaOAc buffer
at pH 5.0) for 24 h at 37 ◦C. The reaction mixture was extracted
with AcOEt (0.5 ml × 3), and the solvent was evaporated off. The
extracts were dissolved in 8 ll of N,O-bis(trimethylsilyl)acetamide
p-Coumaric acid (1), caffeic acid (2), ferulic acid (3), and
sinapic acid (5) were purchased from Tokyo Chemical Industry,
Co., Ltd.; Nakalai Tesque, Co., Ltd.; Nakalai Tesque, Co.,
Ltd.; Tokyo Chemical Industry, Co., Ltd. respectively. [U-ring-
13C6]Phenylalanine (99 atom% 13C), [U-ring-13C6]-4-hydroxybenz-
aldehyde (99 atom% 13C), CH3O2H, LiAl2H4 (98 atom% 2H),
and C2H3I (99.5 atom% H) were purchased from ICON, CIL,
CIL, Aldrich, and Aldrich, respectively. The chemicals used were
reagent grade unless otherwise stated.
2
Quantitative analysis of p-hydroxycinnamic acids,
p-hydroxycinnamaldehydes, p-hydroxycinnamyl alcohols, and
lignans in C. tinctorius
Preparation of standard curves. Standard curves for phenyl-
propanoid monomers (1–15) and lignans (16–21) were prepared
using unlabelled (1–21) and labelled standards (1-d3–15-d3, 16-d4,
17-d4, 18-d3–21-d3). Typically, 1.00 nmol of p-[3,5,7-2H3]coumaric
acid (1-d3) as internal standard was added to each of four 1 ml
glass microtubes (Maruemu Co. Ltd.) containing certain amounts
(0.25, 0.50, 1.25, 1.80 nmol, respectively) of unlabelled p-coumaric
acid (1). After evaporation to dryness, each sample was dissolved
in N,O-bis(trimethylsilyl)acetamide (8.0 ll), heated at 60 ◦C
for 45 min, and an aliquot (0.8 ll) was subjected to GC-MS
measurement with selected monitoring of ions arising from TMS
ethers of p-coumaric acid (1) (m/z 308) and p-[3,5,7-2H3]coumaric
acid (1-d3) (m/z 311). Peaks at the appropriate GC retention times
were integrated, and the ratio of the ion current at m/z 308 was
divided by that at m/z 311. This ratio was a linear function of
the amount of unlabelled p-coumaric acid (1) initially added to
each microtube over the 0.25–1.80 nmol range: Y = 1.30X +
0.31, X = p-coumaric acid (1) (nmol), Y = ion current ratio of
p-coumaric acid (1)/p-[3,5,7-2H3]coumaric acid (1-d3). Standard
curves of the other compounds (2–21) were prepared as above;
caffeic acid (2), Y = 0.38X + 0.07, 0.22–1.66 nmol range; ferulic
acid (3), Y = 1.02X + 0.04, 0.21–1.54 nmol range; 5-hydroxyferulic
acid (4), Y = 2.31X + 0.31, 0.03–0.28 nmol range; sinapic acid (5),
Y = 1.04X + 0.32, 0.03–0.26 nmol range; p-coumaraldehyde (6),
Y = 1.04X + 0.08, 0.05–0.40 nmol range; caffealdehyde (7), Y =
1.09X + 0.11, 0.05–0.37 nmol range; coniferaldehyde (8), Y =
0.92X + 0.03, 0.05–0.34 nmol range; 5-hydroxyconiferaldehyde
(9), Y = 0.75X + 0.04, 0.04–0.31 nmol range; sinapaldehyde (10),
Y = 0.83X + 0.07, 0.08–0.58 nmol range; p-coumaryl alcohol
◦
(BSA), heated for 45 min at 60 C34 and subjected to GC-MS
measurement, so that phenylpropanoid monomers and lignans
present in the extracts were identified and quantified by the use of
the standard curves.
Administration experiments
Administration of 13C- or deuterium-labelled compounds. [U-
ring-13C6]Phenylalanine, p-[U-ring-13C6]coumaric acid (1-13C6),
[2,5,7-2H3]caffeic acid (2-d3), [2,5-2H3, 3-OC2H3]ferulic acid (3-d5),
[3-OC2H3]coniferaldehyde (8-d3), and [3-OC2H3]sinapaldehyde
(10-d3) (0.4 mg each) were dissolved in 0.1 M KPB buffer (pH 7.0,
400.0 ll), and incubated individually with four maturing seeds (7
and 13 DAF) for 24 h. Then, the seeds were freeze-dried, powdered
using a mortar and pestle, and extracted with hot MeOH. The
MeOH extracts were submitted to GC-MS analysis exactly as
above. The peaks on GC corresponding to phenylpropanoid
monomers [(1)–(15)], and matairesinol (19) were analyzed for 13
C
or deuterium incorporation.
Simultaneous administration of two distinct labelled compounds.
Three sets of simultaneous administration of two distinct
compounds {(i) [2,5-2H2, 3-OC2H3]ferulic acid (3-d5) and [U-
ring-13C6]phenylalanine, (ii) [2,5-2H2, 3-OC2H3]ferulic acid (3-d5)
and p-[U-ring-13C6]coumaric acid (1-13C6), and (iii) [2,5-2H2, 3-
OC2H3]ferulic acid (3-d5) and [2,5,7-2H3]caffeic acid (2-d3)} were
conducted. Thus, each mixture of the two compounds (0.2 mg
each) was dissolved in 0.1 M KPB buffer (pH 7.0, 400.0 ll),
and the resultant solution was incubated with four maturing
This journal is
The Royal Society of Chemistry 2007
Org. Biomol. Chem., 2007, 5, 802–815 | 813
©

View Online
seeds of C. tinctorius (7 and 13 DAF) for 24 h. Then, the seeds
were submitted to GC-MS analysis exactly as above.
Professor Kazuhiko Fukushima, Nagoya University, Japan, for
his suggestion in thioacidolysis method.
Lignin analysis
References
Seed hulls of 30 seeds of C. tinctorius cv. Round-leaved White were
pulverized. The powder thus obtained was extracted three times
(1 h × 2, 2 h × 1) with 10 ml of ethyl ether at room temperature,
and then four times (1 h × 2, 2 h × 1 and overnight) with 10 ml of
MeOH at room temperature. The extractive-free seed hull samples
were characterized by acetyl bromide,41 nitrobenzene oxidation,40
and thioacidolysis39 methods. Phloroglucinol-HCl histochemical
lignin staining was done as previously described.37,38
1 T. Umezawa, Regul. Plant Growth Dev., 2001, 36, 57–67.
2 T. Umezawa, Phytochem. Rev., 2003, 2, 371–390.
3 T. Umezawa, Kagaku to Seibutsu, 2005, 43, 461–467.
4 T. Higuchi, Biosynthesis of lignin, in Biosynthesis and Biodegradation
of Wood Components, ed. T. Higuchi, Academic, Orlando, 1985,
pp. 141–160.
5 T. Higuchi and S. A. Brown, Can. J. Biochem. Physiol., 1963, 41, 613–
620.
6 S. A. Brown and A. C. Neish, J. Am. Chem. Soc., 1959, 81, 2419–2424.
7 K. Osakabe, C. C. Tsao, L. Li, J. L. Popko, T. Umezawa, D. T. Carraway,
R. H. Smeltzer, C. P. Joshi and V. L. Chiang, Proc. Natl. Acad. Sci.
U. S. A., 1999, 96, 8955–8960.
Instrumentation
8 L. Li, J. L. Popko, T. Umezawa and V. L. Chiang, J. Biol. Chem., 2000,
275, 6537–6545.
9 L. Li, X. F. Cheng, J. Leshkevich, T. Umezawa, S. A. Harding and V. L.
Chiang, Plant Cell, 2001, 13, 1567–1585.
10 J. M. Humphreys, M. R. Hemm and C. Chapple, Proc. Natl. Acad. Sci.
U. S. A., 1999, 96, 10045–10050.
11 J. M. Marita, J. Ralph, R. D. Hatfield, D. Guo, F. Chen and R. A.
Dixon, Phytochemistry, 2003, 62, 53–65.
12 U. N. Dwivedi, W. H. Campbell, J. Yu, R. S. S. Datla, R. C. Bugos,
V. L. Chiang and G. K. Podila, Plant Mol. Biol., 1994, 26, 61–71.
13 J. Van Doorsselaere, M. Baucher, E. Chognot, B. Chabbert, M.-T.
Tollier, M. Petit-Conil, J.-C. Leple´, G. Pilate, D. Cornu, B. Monties,
M. Van Montagu, D. Inze´, W. Boerjan and L. Jouanin, Plant J., 1995,
8, 855–864.
14 C.-J. Tsai, J. L. Popko, M. R. Mielke, W.-J. Hu, G. K. Podila and V. L.
Chiang, Plant Physiol., 1998, 117, 101–112.
15 R. Atanassova, N. Favet, F. Martz, B. Chabbert, M.-T. Tollier, B.
Monties, B. Fritig and M. Legrand, Plant J., 1995, 8, 465–477.
16 R. Zhong, W. H. Morrison, III, J. Negrel and Z.-H. Ye, Plant Cell,
1998, 10, 2033–2045.
1H NMR spectra were taken with a JAM-LA400MK FT NMR
System (JEOL). Chemical shifts and coupling constants (J) were
given in d and Hz, respectively. GC-MS measurement of lignans
(16–21) was performed on a JMS-DX303HF mass spectrometer
(JEOL) equipped with a Hewlett-Packard 5890J gas chromato-
graph and a JMA-DA5000 mass data system [electron-impact
mode, 70 eV; gas chromatographic column, Shimazu HiCup CBP-
10 M25-025 (5 m × 0.22 mm); temperature, 40 ◦C at t = 0–2 min,
then to 230 ◦C at 30 ◦C min−1; carrier gas, He; splitless injection],
while that of phenylpropanoid monomers (1–15) was done with
a GC-MS QP5050A mass spectrometer (Shimadzu) equipped
with a Shimadzu GC-17A gas chromatograph [electron-impact
mode 70 eV; gas chromatographic column, Shimazu HiCup CBP-
10 M25-025 (20 m × 0.22 mm); temperature, 40 ◦C at t = 0–
◦
2 min, then to 240 ◦C at 40 C min−1; carrier gas, He; splitless
17 R. Zhong, W. H. Morrison, III, D. S. Himmelsbach, F. L. Poole, II and
Z.-H. Ye, Plant Physiol., 2000, 124, 563–577.
18 H. Meyermans, K. Morreel, C. Lapierre, B. Pollet, A. De Bruyn, R.
Busson, P. Herdewijn, B. Devreese, J. Van Beeumen, J. M. Marita, J.
Ralph, C. Chen, B. Burggraeve, M. Van Montagu, E. Messens and W.
Boerjan, J. Biol. Chem., 2000, 275, 36899–36909.
injection]. Ultraviolet (UV) spectra were recorded on a Shimadzu
UV-1600PC spectrometer.
Conclusions
19 G. Pinc¸on, S. Maury, L. Foffmann, P. Geoffroy, C. Lapierre, B. Pollet
and M. Legrand, Phytochemistry, 2001, 57, 1167–1176.
20 J. M. Marita, J. Ralph, C. Lapierre, L. Jouanin and W. Boerjan, J. Chem.
Soc., Perkin Trans. 1, 2001, 2939–2945.
21 F. Chen, M. S. S. Reddy, S. Temple, L. Jackson, G. Shadle and R. A.
Dixon, Plant J., 2006, 48, 113–124.
22 J. Ralph, C. Lapierre, J. M. Marita, H. Kim, F. Lu, R. D. Hatfield,
S. Ralph, C. Chapple, R. Franke, M. R. Hemm, J. Van Doorsserlaere,
R. R. Sederoff, D. M. O’Malley, J. T. Scott, J. J. MacKay, N. Yahiaoui,
A.-M. Boudet, M. Pean, G. Pilate, L. Jouanin and W. Boerjan,
Phytochemistry, 2001, 57, 993–1003.
23 K. Yamauchi, S. Yasuda and K. Fukushima, J. Agric. Food Chem.,
2002, 50, 3222–3227.
In conclusion, the present study established a comprehensive and
quantitative system for the cinnamate/monolignol and lignan
pathways. This system, coupled with the administration of stable-
isotope-labelled pathway intermediates, was successfully applied
to the characterization of these pathways in C. tinctorius seeds,
indicating that ferulic acid (3) is a precursor of lignans in this
species. The results also lead to a proposition that the metabolic
flow in the cinnamate/monolignol pathway towards lignan is
much slower than that for lignin in C. tinctorius and perhaps in
plants in general.
24 K. Yamauchi, S. Yasuda, K. Hamada, Y. Tsutsumi and K. Fukushima,
Planta, 2003, 216, 496–501.
25 G. Schock, S. Goepfert, M. Morant, A. Hehn, D. Neyer, P. Ullmann
and D. Werck-Reichhart, J. Biol. Chem., 2001, 276, 36566–36574.
26 R. Franke, M. R. Hemm, J. W. Denault, M. O. Ruegger, J. M.
Humphreys and C. Chapple, Plant J., 2002, 30, 47–59.
27 R. Niggeweg, A. J. Nichael and C. Martin, Nat. Biotechnol., 2004, 22,
746–754.
28 J. Ralph, T. Akiyama, H. Kim, F. Lu, P. F. Schatz, J. M. Marita, S. A.
Ralph, M. S. S. Reddy, F. Chen and R. A. Dixon, J. Biol. Chem., 2006,
281, 8843–8853.
29 T. Tokimatsu, N. Sakurai, H. Suzuki, H. Ohta, K. Nishitani, T.
Koyama, T. Umezawa, N. Misawa, K. Saito and D. Shibata, Plant
Physiol., 2005, 138, 1289–1300.
30 K. M. Oksman-Caldentey and K. Saito, Curr. Opin. Biotechnol., 2005,
16, 174–179.
31 K. Saito, Tanpakushitsu Kakusan Koso, 2003, 48, 2199–2204.
Acknowledgements
This research was partly supported by Grants-in-Aid for Scientific
Research from the Japan Society for the Promotion of Science (nos.
12660150, 16380116, and 18658069) and by a grant from the New
Energy and Industrial Technology Development Organization
(Development of Fundamental Technologies for Controlling the
Process of Material Production of Plants). The authors thank
Professor Vincent L. Chiang, North Carolina State University,
USA, for discussions. The authors are grateful to Professor
Fumiaki Nakatsubo, Kyoto University, Japan, for his suggestion
in nitrobenzene oxidation and acetyl bromide methods, and to
814 | Org. Biomol. Chem., 2007, 5, 802–815
This journal is
The Royal Society of Chemistry 2007
©

View Online
32 O. Fiehn, Plant Mol. Biol., 2002, 48, 155–171.
33 N. Sakakibara, S. Suzuki, T. Umezawa and M. Shimada, Org. Biomol.
Chem., 2003, 1, 2474–2485.
34 T. Umezawa, H. Kuroda, T. Isohata, T. Higuchi and M. Shimada,
Biosci., Biotechnol., Biochem., 1994, 58, 230–234.
35 T. Umezawa and M. Shimada, Mokuzai Gakkaishi, 1994, 40, 231–
235.
47 R. Palter, R. E. Lundin and W. F. Haddon, Phytochemistry, 1972, 11,
2871–2874.
48 S. Suzuki, T. Umezawa and M. Shimada, Biosci., Biotechnol., Biochem.,
2002, 66, 1262–1269.
49 T. Umezawa, Wood Res. (Kyoto, Jpn.), 2003, 90, 27–110.
50 T. Umezawa and M. Shimada, Biosci., Biotechnol., Biochem., 1996, 60,
736–737.
36 T. Umezawa, T. Isohata, H. Kuroda, T. Higuchi and M. Shimada,
Chiral HPLC and LS-MS analysis of several lignans, in Biotechnology
in pulp and paper industry, ed. M. Kuwahara and M. Shimada, Uni,
Tokyo, 1994, pp. 507–512.
37 T. Umezawa and T. Higuchi, Analysis of lignin degradation interme-
diates by thin layer chromatography and gas chromatography/mass
spectrometry, in Methods in Enzymology, ed. W. A. Wood and
S. T. Kellogg, Academic, San Diego, 1988, vol. 161, pp. 200–211.
38 T. Umezawa and T. Higuchi, Methods used in chemistry of lignin
biodegradation, in Modern Methods of Plant Analysis, New Series, ed.
H. F. Linskens and J. F. Jackson, Springer-Verlag, Berlin, 1989, vol. 10,
pp. 161–185.
51 S. Suzuki, T. Umezawa and M. Shimada, Biosci., Biotechnol., Biochem.,
1998, 62, 1468–1470.
52 T. Umezawa, L. Li, S. Suzuki, N. Sakakibara, T. Nakatsubo and V. L.
Chiang, Proceedings of the 49th Lignin Symposium (Tsukuba, Japan),
2004, pp. 33–36.
53 K. Freudenberg, Science, 1965, 148, 595–600.
54 S. Suzuki, T. Umezawa and M. Shimada, J. Chem. Soc., Perkin Trans.
1, 2001, 24, 3252–3257.
55 R. B. Nair, K. L. Bastress, M. O. Ruegger, J. W. Denault and C. Chapple,
Plant Cell, 2004, 16, 544–554.
56 Y. Nakamura, F. Nakatsubo and T. Higuchi, Wood Res. (Kyoto, Jpn.),
1974, 56, 1–6.
39 K. Hamada, Y. Tsutsumi, K. Yamauchi, K. Fukushima and T. Nishida,
J. Wood Sci., 2003, 49, 333–338.
57 T. Umezawa, L. B. Davin and N. G. Lewis, J. Biol. Chem., 1991, 266,
10210–10217.
40 R. Katahira and F. Nakatsubo, J. Wood Sci., 2001, 47, 378–382.
41 S. Hirosawa, R. Katahira and F. Nakatsubo, J. Wood Sci., 2002, 48,
46–50.
42 H. Kim and J. Ralph, J. Agric. Food Chem., 2005, 53, 3693–3695.
43 von R. Ha¨nsel, H. Schulz and C. Leuckert, Z. Naturforsch., B: Anorg.
Chem., Org. Chem., Biochem., Biophys., 1964, 19, 727–734.
44 R. Palter and R. E. Lundin, Phytochemistry, 1970, 9, 2407–2409.
45 R. Palter, W. F. Haddon and R. E. Lundin, Phytochemistry, 1971, 10,
1587–1589.
58 L. Lundgren, K. Olsson and O. Theander, Acta Chem. Scand., Ser. B,
1979, 33, 105–108.
59 S. K. Banerjee, M. Manolopoulo and J. M. Pepper, Can. J. Chem.,
1962, 40, 2175–2177.
60 J. M. Pepper and J. A. MacDonald, Can. J. Chem., 1952, 31, 476–483.
61 T. Okunishi, T. Umezawa and M. Shimada, J. Wood Sci., 2000, 46,
234–242.
62 T. Umezawa and M. Shimada, Mokuzai Gakkaishi, 1996, 42, 180–185.
63 T. Okunishi, N. Takaku, P. Wattanawikkit, N. Sakakibara, S. Suzuki, F.
Sakai, T. Umezawa and M. Shimada, J. Wood Sci., 2002, 48, 237–241.
64 E. Brown and A. Daugan, Heterocycles, 1987, 26, 1169–1172.
46 S. Nishibe, A. Sakushima, S. Hisada and I. Inagaki, Phytochemistry,
1972, 11, 2629.
This journal is
The Royal Society of Chemistry 2007
Org. Biomol. Chem., 2007, 5, 802–815 | 815
©