unit can only arise as a modified metabolic product following
dimerization of traditional lignin monomers,11 we have
presented ample evidence that it is formed as a monomer
and that the monomer is incorporated into the lignin via
radical coupling processes.1,6,9 Additionally, DHCA monomer
is found in solvent extracts from these plants.6 It is however
still not known whether it derives from coniferyl alcohol or
can come via other pathways from coniferaldehyde. Since
all indications in the pine mutant are that coniferaldehyde
builds up and is not efficiently reduced to the alcohol, the
significant DHCA component was conjectured to arise from
coniferaldehyde rather than coniferyl alcohol,1 but this has
not yet been elucidated.
Here we identify another significant component of the
mutant pine’s lignin, arylpropane-1,3-diols, whose source
was originally puzzling. It is now clear that these units derive
directly from DHCA by peroxidase-mediated reactions.
Figure 1a shows a subplot of the side chain region from
a 2D HMQC-TOCSY NMR experiment on acetylated lignin
isolated from the mutant pine,12 highlighting the new
arylpropane-1,3-diacetates (red) along with the previously
identified DHCA units (green). Data from a model com-
pound, 1,3-diacetoxy-1-(4-O-benzyl-3-methoxyphenyl)pro-
pane, 5-Ac,13 are at the center of the yellow circles and
obviously match well.
correlations from other units with protons resonating in the
same regions. Thus, resonances at 2.2 ppm arise only from
the new unit (â-proton) and DHCA (â-proton) (as well as
the strong acetone solvent signal). Figure 1e is therefore a
composite HSQC of those units, namely, the acetone, as well
as DHCA (green) and APD (red). The plane through HR (5.8
ppm, Figure 1d) is quite clean but the TOCSY transfer
between HR and Hâ is poor (under the chosen acquisition
conditions), giving rise to only weak Hâ/Câ and Hγ/Cγ cross-
peaks. γ-Protons in lignin seriously overlap; consequently
the plane through an Hγ (4.1 ppm, Figure 1f) shows the nice
correlations for the APD unit, but also with DHCA (green),
lignin’s â-aryl ether units (blue), and the intense methoxyl.
The observation of complete HSQC spectra for the ADP unit
in all three planes corresponding to the APD side chain
protons (Figures 1d-f), along with the evidence from the
2D-HMQC-TOCSY (Figure 1a), provides sufficiently com-
pelling proof of the structure in the isolated lignin from the
CAD-deficient pine mutant. A model compound for etheri-
fied APD structures, 5,13 provides data for comparison,
Figure 1g; note however that there are a large number of
different bonding environments in lignin so the peaks are
broader and more disperse in lignin spectra.9
As with other novel units found in lignins from transgenic
or mutant plants, traces of the same components can be found
in lignins from control plants. Figure 1b shows the R-C/H
region of HSQC spectra of the mutant’s lignin, where the
â-aryl ether units (blue) and the strong new ADP unit (red)
appear most cleanly. Figure 1c shows the same region in
lignin from a normal pine control. Although the peak is weak,
it is diagnostic and, with the other correlations evident (not
shown), well authenticated. The peak is also in lignins
isolated from mature pine clear sapwood.
Perhaps more diagnostic is the data from a 3D TOCSY-
HSQC experiment.14 2D F2-F3 planes from this experiment
show a complete HSQC spectrum of any units bearing
protons resonating at the frequency of the taken slice.9 In
spectra as complex as those from lignins, finding unique
resonances in the proton spectrum is often difficult. However,
slices through each of the three side chain protons in APD
units show rather clear HSQC spectra of that unit, along with
Where do APD units come from? They derive from DHCA
2 via the action of peroxidase and hydrogen peroxide,
Scheme 1. The mechanism, via a vinylogous quinone
(8) Lignin is the term used for the phenylpropanoid polymer in the cell
walls of normal plants and will be used here although there remain issues
regarding the biochemical and functional roles of “nontraditional” lignins
in mutants and transgenics (see refs 6 and 11).
(9) Ralph, J.; Marita, J. M.; Ralph, R. A.; Hatfield, R. D.; Lu, F.; Ede,
R. M.; Peng, J.; Quideau, S.; Helm, R. F.; Grabber, J. H.; Kim, H.; MacKay,
J. J.; Sederoff, R. R.; Chapple, C.; Boudet, A. M. In Progress in
Lignocellulosics Characterization; Argyropoulos, D. S., Rials, T., Eds.;
TAPPI Press: Atlanta, GA, 1999; in press.
Scheme 1
(10) (a) Lundquist, K.; Stern, K. Nordic Pulp Paper Res. J. 1989, 210-
213. (b) Fukagawa, N.; Meshitsuka, G.; Ishizu, A. J. Wood Chem. Technol.
1991, 11, 373-396. (c) Brunow, G.; A¨ mmålahti, E.; Niemi, T.; Sipila¨, J.;
Simola, L. K.; Kilpela¨inen, I. Phytochem. 1998, 47, 1495-1500.
(11) (a) Gang, D. R.; Fujita, M.; Davin, L. D.; Lewis, N. G. In Lignin
and Lignan Biosynthesis; Lewis, N. G., Sarkanen, S., Eds.; American
Chemical Society: Washington, DC, 1998; ACS Symposium Series 697,
pp 389-421. (b) Lewis, N. G.; Davin, L. B.; Sarkanen, S., ref 11a, pp
1-27. (c) Lewis, N. G. In Abstracts of Papers, 215th National Meeting of
the American Chemical Society, Dallas, TX; American Chemical Society:
Washington, DC, 1998; Vol. 1, p Cell-09.
(12) NMR spectra were taken on a Bruker DRX-360 instrument fitted
1
with a 5 mm H/broadband gradient probe with inverse geometry (proton
coils closest to the sample). The conditions used for all samples were ∼80
mg of acetylated isolated lignin in 0.4 mL of acetone-d6, with the central
solvent peak as internal reference (δH 2.04, δC 29.80). Experiments used
were standard Bruker implementations of inverse (1H-detected) 2D-gradient-
HSQC, 2D-HMQC-TOCSY, and 3D-gradient-TOCSY-HSQC experiments.
The TOCSY spin lock period was 100 ms. Carbon/proton designations are
based on conventional lignin numbering. Lignin 2D and 3D spectra have
been extensively detailed in a recent book chapter.9 The CAD-deficient
pine mutant and the isolation of lignin fractions have been described
previously.1 For this study, the lignins were acetylated, extracted into EtOAc,
and washed with 6 mM EDTA to improve NMR relaxation properties (cf.
ref 2).
methide 4, involves two H-radical abstractions. Abstraction
from a benzylic CH2 to produce quinone methides from
phenoxy radicals has been noted previously.15 When DHCA
is subjected to peroxidase-H2O2, monomeric APD 1 as well
as the range of homo dimers and crossed dimers 7-11
(Scheme 2) involving DHCA and APD are found, as will
be detailed elsewhere.
324
Org. Lett., Vol. 1, No. 2, 1999