Are lignins optically active?

Article abstract of DOI:10.1021/jf9901136

The accepted derivation of lignins from non(enzymatically)-controlled radical coupling reactions has been recently challenged, and it is relevant to ascertain unequivocally whether lignins are or are not (as normally assumed) optically active. Two approaches were used. First, DFRC (derivatization followed by reductive cleavage) dimers derived from β-5- and β-β-units in pine lignins, which certainly retain unaltered chiral centers (as well as β-1- and β-O-4-units where the intactness may be debated), were shown to be optically inactive by circular dichroism (CD) and chiral high- performance liquid chromatography. CD of β-5-derived dimers following enantiomeric separation readily demonstrated the sensitivity of the method. Second, no optical activity could be detected (above 250 nm to avoid carbohydrate contributions) by CD of lignin isolates from pine, kenaf, maize, or a CAD-deficient pine mutant. Representative lignins are therefore not, within limits of detection by these methods, optically active.

Full text of DOI:10.1021/jf9901136

J. Agric. Food Chem. 1999, 47, 2991−2996
2991
Ar e Lign in s Op tica lly Active?
J ohn Ralph,*,†,‡ J unpeng Peng,† Fachuang Lu,†,‡ Ronald D. Hatfield,† and Richard F. Helm§
U.S. Dairy Forage Research Center, Agricultural Research Service, U.S. Department of Agriculture,
Madison, Wisconsin 53706, Department of Forestry, University of WisconsinsMadison,
Madison, Wisconsin 53706, and Fralin Biotechnology Center, Virginia Technical Institute,
Blacksburg, Virginia 24061
The accepted derivation of lignins from non(enzymatically)-controlled radical coupling reactions
has been recently challenged, and it is relevant to ascertain unequivocally whether lignins are or
are not (as normally assumed) optically active. Two approaches were used. First, DFRC (derivati-
zation followed by reductive cleavage) dimers derived from â-5- and â-â-units in pine lignins, which
certainly retain unaltered chiral centers (as well as â-1- and â-O-4-units where the intactness may
be debated), were shown to be optically inactive by circular dichroism (CD) and chiral high-
performance liquid chromatography. CD of â-5-derived dimers following enantiomeric separation
readily demonstrated the sensitivity of the method. Second, no optical activity could be detected
(above 250 nm to avoid carbohydrate contributions) by CD of lignin isolates from pine, kenaf, maize,
or a CAD-deficient pine mutant. Representative lignins are therefore not, within limits of detection
by these methods, optically active.
Keyw or d s: Lignin; lignin dimer; DFRC; pine, kenaf; maize; CAD-deficient pine mutant; circular
dichroism; optical activity; chirality; racemic
INTRODUCTION
Sidhu, 1961), and various acid treatments of syringares-
inol glycosides release optically active phenolics (Dickey,
1958). The potential for polymerized lignans (monolignol
dimers that are frequently optically active) to contami-
nate or contribute to lignins offers a pathway to partial
optical activity (Lundquist, 1973). Freudenberg and
Neish clearly distinguish resin components (such as
hydroxymatairesinol) from lignin precursors or inter-
mediates (Freudenberg and Neish, 1968). Today, lignins
are generally concluded to be optically inactive, but
evidence appears limited to the above-mentioned ob-
servations concerning â-â-units that arise from mono-
mer-monomer coupling and are consequently only
minor components. The assertion has not been rigor-
ously proven for the polymer or for lignins’ major units.
Recently, Brunow’s group showed that enantiomeri-
cally enriched dimers can result from radical coupling
of chiral conjugates of lignin-related monomers (followed
by removal of the conjugate), using otherwise normal
peroxidase/H2O2 conditions (Bolzacchini et al., 1998).
Such an observation opens up new possibilities for
enantiomerically enriched products from radical cou-
pling reactions in the cell wall, using chiral carbohy-
drate or uronic acid auxiliaries, for example (Helm et
al., 1997). Lignans, dimers of monolignols, are also
produced by many plants and are often in optically
active forms. For example, the lignan pinoresinol and
its derived matairesinol, were found only as their (-)-
isomers in suspension-cultured Pinus taeda (Eberhardt
et al., 1993).
Lignins are phenolic polymers produced primarily
from hydroxycinnamyl alcohols (monolignols) by radical
coupling reactions that are generally considered to be
independent of direct enzyme control (Harkin, 1967;
Freudenberg and Neish, 1968). Freudenberg and Die-
trich discovered that oxidase-mediated dehydrogenation
of coniferyl alcohol produced dimers that were not
optically active (Freudenberg and Dietrich, 1953).
Lundquist examined aspects of lignins’ optical activity
in some of his early acidolysis papers (Lundquist, 1970,
1973, and references cited therein). Lignin-derived â-â-
dimers released from lignins by acidolytic methods are
racemic. For example, pinoresinol, isolated following
methanolysis of (solvent-extracted) spruce wood, was
racemic (Freudenberg et al., 1965), as was syringares-
inol and its derived episyringaresinol isolated following
acidolysis (Lundquist, 1973). Lundquist (K. Lundquist,
Chalmers University, personal communication, 1999)
found other released lignin dimers (e.g., divanillyltet-
rahydrofuran) without optical activity. It has been
convincingly demonstrated that acidolytic conditions do
not scramble the stereochemistry of the important
â-carbons. For example, acidolysis of (+)-pinoresinol
yielded (+)-pinoresinol and (+)-epipinoresinol (Lindberg,
1950; Lundquist, 1970), (-)-2,3-divanillyl-1,4-butanediol
gave (-)-3,4-divanillyltetrahydrofuran (Lundquist, 1970),
acid hydrolysis of (+)-syringaresinol gave (+)-syrin-
garesinol and (+)-episyringaresinol (Freudenberg and
Recent isolation of a “dirigent” protein from forsythia,
which facilitates coniferyl alcohol radical coupling to
pinoresinol in a regio- and stereoselective manner, has
led the discoverers to extrapolate their lignan observa-
tions to lignins (Davin et al., 1997). In a newly proposed
paradigm (Lewis and Davin, 1998), as we interpret it,
lignins form from template arrays of dirigent proteins
* Address correspondence to this author at the U.S. Dairy
Forage Research Center, USDA-ARS, 1925 Linden Dr. W.,
Madison, WI 53706-1108 [telephone (608) 264-5407; fax (608)
264-5147l e-mail jralph@facstaff.wisc.edu].
† U.S. Dairy Forage Research Center.
‡ Department of Forestry.
§ Fralin Biotechnology Center.
10.1021/jf9901136 CCC: $18.00 © 1999 American Chemical Society
Published on Web 07/07/1999

2992 J. Agric. Food Chem., Vol. 47, No. 8, 1999
Ralph et al.
F igu r e 2. CD spectra of isolated DFRC dimers 1-4 along
with the chiral-HPLC-separated â-5-dimer 1. UV spectra are
shown with dashed lines; the CD spectra are more noisy at
high UV absorptions, but the CD rotations are also strongest
in the region of strong UV absorption. Chiral column HPLC
traces are shown to the right of the figure; the â-â-enantiomers
3 were not resolved, nor were the â-O-4-enantiomers 4,
although the threo- and erythro- (syn- and anti-) diastereomers
were. The resolved â-5-enantiomers 1 showed strong CD
spectra; the region from 190 to 245 nm was run using one-
fourth the concentration used for the 240-350 nm region.
F igu r e 1. Dimeric DFRC products from various structural
units in softwood lignins. Optical centers at â-carbons formed
by radical coupling reactions during lignification are denoted
with a f. The â-O-4-dimer 4 is produced only in small amounts
from incomplete ether cleavage.
and are synthesized with exquisite structural control.
With the possibility that the synthesized lignin chain
then structurally dictates the next chain by a template-
polymerization process (Sarkanen, 1998), the paradigm
has the apparent necessary ingredients to bring lignin
more into line with other carefully synthesized biological
polymers such as cellulose and proteins. The idea is not
without significant obstacles, some of which are alluded
to below, and is currently without structural, biochemi-
cal, or enzymatic evidence, but it has sparked valuable
debate. Although the paradigm may not require the
resultant lignin to be optically active, the finding of
chiral lignin would significantly strengthen the idea.
The question of lignin optical activity needs to be
addressed. This paper describes a two-pronged approach
to resolve the issue as unequivocally as possible for
various lignins.
isolated from pine (Peng et al., 1998), using system II (see
below), provided the two enantiomers. Their CD spectra,
Figures 2 and 3, were from 130 µg/mL for the 230-350 nm
range 30 µg/mL for the 190-245 range (Figure 2).
CD. CD spectra were run on an AVIV model 62A-DS CD
spectrometer (Lakewood, NJ ) driven by an Apple Macintosh
computer running IGOR-Pro 3 software (Wavemetrics, Lake
Oswego, OR). The lamp current was set at 26 A; temperature
was 25 °C; all conditions were standard. For lignin dimers,
the solvent was acetonitrile (HPLC grade, Baker); for isolated
lignins, the solvent was 10:3 acetonitrile/water. Sample con-
centrations were adjusted so that optical densities measured
by the dynode were just below the maximum of 600 V at
maximal absorptions in the spectral range. Spectra were
typically obtained using 10 s averaging in 1 nm steps and were
baseline-subtracted (using scans of the blank solvents).
Smoothed and raw spectra are shown in Figures 2 and 3.
Isola ted Lign in s. Pine (Ralph et al., 1997), kenaf (Ralph,
1996), maize (Ralph et al., 1994), and mutant pine (Ralph et
al., 1997) lignins were isolated as previously described. Briefly,
samples of Pinus taeda clear sapwood, Tainung kenaf bast
fibers, mature Zea mays stems, and a 12-year-old CAD-
deficient P. taeda mutant (containing knotty regions) (as well
as another control pine of the same age from the same areas
data not shown) were ground in a Wiley mill to pass through
a 1 mm mesh screen. The ground materials were exhaustively
Soxhlet-extracted with water, methanol, acetone, and chloro-
EXPERIMENTAL PROCEDURES
DF RC Dim er s. DFRC (derivatization followed by reductive
cleavage) dimers 1-4 (Figure 1) were isolated from pine clear
sapwood as previously described (Peng et al., 1998) and
repurified by HPLC (using system I, see HPLC Conditions).
The compounds, ∼0.2 mg, were each dissolved in 1 mL of
acetonitrile and dilutions performed until satisfactory circular
dichroism (CD) spectra could be obtained without exceeding
a dynode voltage of 600 V on the CD spectrometer. Concentra-
tions for 1-4 used for the spectra in Figure 2 were ∼40-50
µg/mL. Chiral separation of 1.5 mg of 1, from DFRC dimers

Are Lignins Optically Active?
J. Agric. Food Chem., Vol. 47, No. 8, 1999 2993
return to initial conditions in 3 min and a 7 min hold before
the next injectionstotal run time of 50 min per injection.
P r oof Th a t P h en ylcou m a r a n s Su r vive DF RC w ith
Th eir â-Ca r bon s Un scr a m bled . The two enantiomers of the
phenylcoumaran model 5 were separated by preparative chiral
HPLC (Figure 4). The fastest-eluting enantiomer [it is not
known whether this was the (R)- or the (S)-isomer] of 5 (2.3
mg) in CH2Cl2 (5 mL) was brominated (Br2, 50 mg; CH2Cl2,
0.1 mL) for 3 min to saturate the double bond and produce
the 1,2-dibromo derivative 6, which would subsequently
regenerate the double bond during the reductive elimination
step of the DFRC procedure (Lu and Ralph, 1997a,b). After
evaporation of the solvent under reduced pressure, the bro-
minated compound 6 was dissolved in 10% (v/v) AcBr in acetic
acid solution. 2-Propanol (0.08 mL) was added to produce some
HBr required for effective reaction. (HBr is required for
effective reactionsfully acetylated models cannot generate any,
whereas hydroxyl groups present in unacetylated models,
isolated lignins, or whole cell wall samples generate sufficient
HBr for reaction.) This mixture was allowed to stand for 16 h
and the solvent evaporated at 45 °C under reduced pressure.
The residue was dissolved in dioxane/acetic acid/water (5:4:1,
3 mL) and treated with powdered Zn (50 mg) for 30 min.
Extraction with CH2Cl2 produced crude product (2.2 mg),
which was acetylated. The desired product 1 was isolated as
the major product by HPLC, 0.9 mg, ∼38%. This product had
GC retention time and NMR and mass spectra identical with
those of the racemic dimer 1 isolated from pine wood following
DFRC but gave only a single peak, the later eluting peak
(Figure 4) on chiral HPLC (Chiralcel-OD, system II, see HPLC
Conditions). When racemic 5 was subjected to the same
bromination and DFRC reactions, chiral HPLC this time
produced two peaks as expected. Because a single enantiomer
of the phenylcoumaran 5 produced a single enantiomer of the
product 1, it is clear that the optical center (the â-carbon)
retained in the product is unchanged from its state in the
original phenylcoumaran, that is, that no scrambling had
occurred.
F igu r e 3. CD spectra of various lignin isolates, illustrating
no detectable optical activity. CD spectra of the separated â-5-
enantiomers 1 are shown again for comparison. To test the
sensitivity of detection, a resolved enantiomer of 1 was spiked
into the pine lignin sample at 5 and 10% levels; the raw data
and the curve smoothed by the CD software indicate that a
5% component is reasonably easily detected (although the
required level may be higher in the complex lignin sample
where various units might have opposing contributions).
RESULTS AND DISCUSSION
Of the various methods for determining optical activ-
ity, CD (Kagan, 1977) was chosen for its sensitivity and
because the association of CD phenomena only with UV-
active absorptions allows optical activity in lignins to
be determined even in the presence of contaminating
(optically active) polysaccharides.
form or diethyl ether. The dried cell wall residues were then
ball-milled for 1.5 h (three 0.5 h intervals separated by 0.5 h)
using a vibratory ball mill. For the kenaf and the maize
samples, excess polysaccharides were degraded using crude
cellulases (Obst and Kirk, 1988; Ralph et al., 1994; Ralph,
1996). The finely divided cell wall materials were then
extracted twice with 96:4 dioxane/water (Bjo¨rkman, 1956).
After lyophilization, the crude lignin powders were washed
with distilled water and again lyophilized. Lignins were not
further purified. The final yields on lignins (based on original
Klason lignin) were pine, 15%; kenaf, 26%; maize, 67%; and
CAD-deficient pine, 17% (and normal control pine, 12.5%). The
samples contained some carbohydrates and uronics but were
>90% lignin in each case. Samples, 1.6-5.5 mg, were dissolved
in 0.26 mL of 10:3 acetonitrile/water and diluted again for CD.
Approximate lignin concentrations used for the 250-345 nm
spectra of Figure 3 were pine, 2.0 mg/mL; CAD-deficient pine
mutant, 1.7 mg/mL; maize, 0.9 mg/mL; and kenaf, 2.8 mg/mL.
Lign in DF RC Dim er s. Dimeric lignin fragments
were isolated from pine wood following DFRC as de-
scribed previously (Peng et al., 1998) and repurified by
reverse-phase HPLC (system I, see Experimental Pro-
cedures). As seen in Figure 1, the bonds originally
formed in the radical coupling step remain unaltered
from their native states following this degradation.
Regardless of what non-stereospecific chemical reactions
have occurred at other sites, the compounds retain
chirality at the sites of original coupling (as is verified
for phenylcoumarans, below). Determination of chirality
in these compounds therefore reflects chirality in the
original lignin. This is particularly clear-cut with â-5-
and â-â-compounds (where the crucial bonds do not
cleave) but less so for â-1’s (where there are uncertain-
ties regarding how they arise) (Ralph et al., 1998b;
Seta¨la¨ et al., 1999) and, perhaps, â-O-4’s (where it is
less clear that the â-O-4 bond has not been broken and
re-formed in the production of the fragment dimers).
HP LC Con d ition s. HPLC was carried out on a Shimadzu
10A HPLC equipped with a diode array detector (detector
range set at 190-350 nm), an autosampler, and a fraction
collector. Two separation systems were used. System I used a
Supelco (Bellefonte, PA) reverse-phase Supelcosil SPLC-8
column (5 µm, 250 mm × 10.0 mm i.d.): mobile phase, an
isocratic elution with acetonitrile/water/THF (45:50:3); flow
rate, 1.6 mL/min. System II used a Chiralcel-OD chiral column
(250 mm × 4.6 mm i.d., Daicel Chemical Industries, Exton,
PA): mobile phase, a gradient elution with hexane (solvent
A) and 2-propanol (solvent B); flow rate, 1.0 mL/min. Initial
conditions were 90% A and 10% B, which were held for 8 min
followed by an increase to 20% B over 2 min, a 10 min hold,
an increase to 35% B in 12 min, an 8 min hold, and then a
Figure 2 shows CD spectra of â-5-, â-1-, â-â-, and â-O-
4-dimers 1-4 isolated from pine lignin following DFRC,
along with spectra from both enantiomers of the â-5-
dimers 1 (which were separated by chiral HPLC). The
spectra are somewhat noisy, particularly near 200 nm,

2994 J. Agric. Food Chem., Vol. 47, No. 8, 1999
Ralph et al.
where the absorption is high and the total optical
density of the sample limits the sample concentration
that can be used for CD. However, high CD rotations
are expected at high absorption wavelengths, so the
sensitivity is ample. This may be seen from the 190-
245 nm scans of the â-5-enantiomers (R)- and (S)-1,
where the concentrations were reduced 4-fold and still
produced pronounced CD rotations. At 280 nm, a
maximum in lignin UV spectra, the noise level is
sufficiently low to allow relatively sensitive determina-
tion of optical activity. Note that the CD spectra shown
here are for the entire wavelength range where the
sample concentration was chosen to allow the recording
of spectra down to 210 nm. Subspectra from more
concentrated samples were also run at higher wave-
lengths to improve the capability of determining activity
throughout the UV spectrum (not shown). No hint of
such activity was found.
Of the various linkage types, â-â-units are perhaps
most likely to show optical activity. The (â-â) lignans
pinoresinol and matairesinol from P. taeda are reported
to be optically active (Eberhardt et al., 1993). There is
the possibility that even carefully pre-extracted wood
meal contains lignans or polymerized lignans (Sarkanen
and Hergert, 1971; Hergert, 1977). One group of re-
searchers has claimed that our observations on lignins
from P. taeda and a CAD-deficient mutant (Ralph et
al., 1997) are almost wholly due to lignan-derived
components (Gang et al., 1998). The lack of any optical
activity in the â-â-dimer, which was obtained from
whole wood, appears to provide evidence in addition to
that already at hand (Sederoff et al., 1999) that the
dimers arise from true racemic lignin, rather than some
poly-lignan (that would likely be optically active).
Alternatively, they would have to derive from some
other (nonoptically active) lignan pool. Similar conclu-
sions arise from examining CD spectra of isolated
lignins (see below).
F igu r e 4. Proof that the DFRC method does not scramble
the crucial â-carbon stereochemistry in â-5-units. Dimeric
model 5 was first brominated so that the unsaturated side
chain would be generated, giving the same â-5-product 1 as
isolated from DFRC of lignin. The potential for scrambling the
center by HBr elimination from the bromo intermediate 7 is
shown but, as noted in the text, is not anticipated. Chiral
HPLC (bottom traces) indicates that a single enantiomer of 5
produces a single enantiomer of 1 following DFRC, proving
that the optical center is not scrambled.
P r oof Th a t Op tica l Cen ter s Ar e Not Scr a m bled
d u r in g DF RC. One of the assumptions in the above
analysis of DFRC dimers is that the DFRC process itself
does not scramble the optical centers that were created
during the radical coupling processes of lignification.
The most likely to scramble are the â-5 (phenylcouma-
ran) moieties, because they have the potential to revers-
ibly eliminate HBr, producing stilbenes (Figure 4).
However, stilbenes 8 produce their own diagnostic
DFRC products, and significant products from HBr
(re-)addition (to 7) have not been observed in our
laboratory (F. Lu, unpublished data).
A starting model compound that could be easily
separated into its component enantiomers was required.
The â-5-dimer 5 was a logical choice. However, it was
also advantageous to produce the same product 1 as is
released from lignins following DFRC; optical isomers
of this compound were readily separated as described
above. In lignin reactions, the side-chain double bond
in 1 arises during the reductive elimination from a
2-bromo ether. This reaction pathway was not feasible
to fully model; the required trimeric â-5/â-O-4 model
compound has too many isomers to allow ready separa-
tion of a single enantiomer. It was recognized that the
reductive elimination could be effectively simulated by
simply brominating the dimeric model 5. The phenyl-
coumaran moiety would then react under DFRC condi-
tions as would such units in lignin, whereas the 1,2-
dibromide would generate the requisite double bond in
the Zn reductive elimination step, producing the same
compound 1 as formed from lignins. Incidentally, DFRC
dimer 1 has been shown to arise from the free phenolic
â-5-units in lignins; etherified units produce different
products (F. Lu and J . Peng, unpublished results).
When racemic 5 was taken through the bromination
and DFRC steps (Lu and Ralph, 1997a,b), product 1 was
indeed the major product. Chiral column HPLC showed
that the two enantiomers of 5 produced two enantiomers
of 1 in equal amounts (Figure 4) as expected. When the
separated faster-eluting enantiomer of 5 was treated the
same way, only the slower-eluting enantiomer of 1
resulted. Clearly, the optical center had been retained
throughout the process; insignificant scrambling had
occurred. Therefore, the assumption made above that
the bond formed during the radical coupling step of
lignification remained intact is valid. This is logically
also true for the chiral centers in â-â-units but, as noted
above, less clear for â-1-units, because their origin in
lignins is still unclear, and perhaps â-O-4-units. Because
â-O-4-units are almost completely cleaved during DFRC
and because the optical isomers of products 4 were not
resolved by our HPLC conditions, their optical activity
retention through DFRC was not examined. Neverthe-
less, it is clear from at least the â-â- and â-5-dimers that
no optical activity is present in these lignin units in the
P. taeda examined.

Are Lignins Optically Active?
J. Agric. Food Chem., Vol. 47, No. 8, 1999 2995
Isola ted Lign in s. A valuable and exploitable feature
of CD spectra is that CD is associated only with UV
bands (Kagan, 1977). Unlike with other optical mea-
surements (direct optical rotation or optical rotatory
dispersion), CD spectra can therefore be run on samples
containing chiral impurities, providing the UV spectrum
of the component of interest has diagnostic absorptions
well separated from those of the chiral contaminants.
Isolated lignins always contain contaminating carbo-
hydrates, which are optically active. However, the UV
spectra of carbohydrates do not extend much above 200
nm. A major lignin peak of interest is at 280 nm, with
shoulders out at around 300+ nm. Acetylated lignins
were dismissed from consideration as acetylated car-
bohydrate absorptions can reach up to 250 nm. The
range 250-375 nm was sufficiently diagnostic for un-
derivatized lignins while devoid of carbohydrate inter-
ference.
Isolated lignins were dissolved in 10:3 acetonitrile/
water, a particularly good solvent for UV studies. CD
spectra of maize, kenaf, pine, and CAD-deficient mutant
pine lignins all showed no detectable optical activity
(Figure 3). Improved signal-to-noise ratio was obtained
by longer averaging in noisy subregions (not shown).
An indication of the level of a chiral component
required to observe optical activity can be gained from
spiking studies. Figure 3 shows CD spectra from pine
lignin with ∼5 and ∼10% levels of an isolated enanti-
omer of the â-5 dimer 1 added. The optical activity can
be readily seen at the 5% level. Admittedly, with several
potential chiral centers in lignin, and the possibility of
non-reinforcing optical CD spectra, the ability to deter-
mine chirality at levels as low as 5% may still be difficult
under our experimental conditions. With unlimited
access to long-term-averaging, improved signal-to-noise
ratios should allow an improvement of these determina-
tions.
or in stress, demand nothing more than changes in
monomer supply (Sederoff et al., 1999).
Producing a variety of structures/stereochemistries is
an asset in plant defense responses (Denton, 1998).
Unrestricted free radical coupling maximizes the struc-
tural diversity while possibly optimizing the functional-
ity of the lignin macromolecule. In light of the above,
the recently advocated bold new paradigm (Lewis and
Davin, 1998) for absolute structural and stereochemical
control of lignification may be an overextrapolation of
the researchers’ lignan findings (Davin et al., 1997). The
proposal cites two convoluted explanations for the
“perceived lack of optical activity of lignins” (Lewis and
Davin, 1998). One explanation, that two types of comple-
mentary proteins encode opposite optical isomers, re-
quires the plant to adopt an energetically excessive
mode of creating an optically active lignin polymer only
to carefully negate that structural feature via a comple-
mentary set of proteins for which additional comple-
mentary biochemical pathways must also be supported.
This effort is to produce two complementary lignin
polymers when each has identical physical properties,
identical to those of the racemic mixture. The template
argument, by which replication of the lignin chain forms
complementary mirror images, has less serious detrac-
tions, although no evidence of any structural replication
ability has yet appeared. Because lignins are found
intimately associated with hemicelluloses, it is not clear
how discontinuities in the alignment of one lignin chain
might contribute to excess optical activity (of the
incompletely replicated section). It appears to us that
nature has already solved the problem more elegantly
by using “uncontrolled” radical coupling reactions. Until
experimental evidence is provided, the new (anti-Ock-
ham’s razor) paradigm for lignification is an interesting
discourse currently without compositional, structural,
or enzymatic evidence.
Con clu sion s. Our inability to detect optical activity
in lignins by two approaches, as presented here, seems
to strengthen the view that lignin is indeed synthesized
by the plant without direct (regio- or) stereocontrol over
the exact course of radical coupling events. A plant’s
ability to produce a cell wall polymer with appropriate
properties for water transport, defense, and other tasks,
with a compositional (and structural) flexibility that is
determined by monomer supply, is presumably well
suited to surviving environmental, gravitropic, and
biological stresses. Indeed, the recently noted abilities
of plants to remain viable by circumventing biogenetic
obstacles placed upon their lignin biosynthetic path-
ways, in a single generation, without the benefit of
evolution, appears to be an endorsement of their flexible
strategy (Ralph et al., 1997, 1998a; Boudet, 1998;
Sederoff et al., 1999).
The maize lignin has high UV absorption centered at
310 nm, resulting in considerable noise in the CD
spectrum in that region. The aldehyde-rich CAD-
deficient pine lignin also has absorptions extending out
to ∼350 nm. The (normal) pine and kenaf lignins have
the more normal distinct 280 maximum. The CAD-
deficient pine mutant (Ralph et al., 1997) was of
particular interest because it has been suggested that
this isolated lignin is merely a polymerized lignan
artifact (Gang et al., 1998), an issue that we address
more fully elsewhere (Sederoff et al., 1999). If this were
so, the mutant’s lignans must be (unexpectedly?) opti-
cally inactive because the isolated lignin/poly-lignan
contains no hint of optical activity.
Im p lica tion s for Lign ifica tion . As suspected by
Freudenberg and others many years ago, lignin appears
to be racemic, fully in accord with its generation from
radical coupling reactions that are without direct en-
zymatic control. We have seen no experimental data
that require a precisely controlled synthesis of structur-
ally defined lignin, that is, data that cannot be sup-
ported by simply recognizing that the plant does ex-
quisitely, temporally and spatially, control the supply
of monolignols, oxidizing enzymes, and oxidative species
(H2O2 for example). The matrix environment of the
polymerization affects the interunit linkage composition,
as has been well demonstrated in synthetic lignification
experiments (Brunow et al., 1993; Terashima et al.,
1996). The regulated differences in lignification in
various cells and various regions of the cell, in wounding
ACKNOWLEDGMENT
We are grateful to Professors K. Lundquist (Chalmers
University, Go¨teborg, Sweden), G. Brunow (University
of Helsinki, Finland), and J . M. Harkin (Univeristy of
WisconsinsMadison) for their help with early references
to lignin optical activity issues and for valuable discus-
sion and to Darrell McCaslin (Department of Biochem-
istry, University of WisconsinsMadison) for help with
CD. Lignin samples were from other studies already
published; J ohn MacKay and Ron Sederoff provided the
CAD-deficient pine mutant and the control.

2996 J. Agric. Food Chem., Vol. 47, No. 8, 1999
Ralph et al.
LITERATURE CITED
for analysis of DFRC monomers. J . Agric. Food Chem.
1997a , 45, 2590-2592.
Bjo¨rkman, A. Studies on finely divided wood. Part I. Extraction
of lignin with neutral solvents. Sven. Papperstidn. 1956, 59,
477-485.
Bolzacchini, E.; Brunow, G.; Meinardi, S.; Orlandi, M.; Rin-
done, B.; Rummakko, P.; Setala, H. Enantioselective syn-
thesis of a benzofuranic neolignan by oxidative coupling.
Tetrahedron Lett. 1998, 39, 3291-3294.
Boudet, A.-M. A new view of lignification. Trends Plant Sci.
1998, 3, 67-71.
Lu, F.; Ralph, J . The DFRC method for lignin analysis. Part
1. A new method for â-aryl ether cleavage: lignin model
studies. J . Agric. Food Chem. 1997b, 45, 4655-4660.
Lundquist, K. Acid degradation of lignin. II. Separation and
identification of low-molecular weight phenols. Acta Chem.
Scand. 1970, 24, 889-907.
Lundquist, K. Acid degradation of lignin. Part VIII. Low
molecular weight phenols from acidolysis of birch lignin.
Acta Chem. Scand. 1973, 27, 2597-2606.
Obst, J . R.; Kirk, T. K. Isolation of lignin. Methods Enzymol.
1988, 161, 87-101.
Peng, J .; Lu, F.; Ralph, J . The DFRC method for lignin
analysis. Part 4. Lignin dimers isolated from DFRC-
degraded loblolly pine wood. J . Agric. Food Chem. 1998, 46,
553-560.
Ralph, J . An unusual lignin from Kenaf. J . Nat. Prod. 1996,
59, 341-342.
Ralph, J .; Hatfield, R. D.; Quideau, S.; Helm, R. F.; Grabber,
J . H.; J ung, H.-J . G. Pathway of p-coumaric acid incorpora-
tion into maize lignin as revealed by NMR. J . Am. Chem.
Soc. 1994, 116, 9448-9456.
Ralph, J .; MacKay, J . J .; Hatfield, R. D.; O’Malley, D. M.;
Whetten, R. W.; Sederoff, R. R. Abnormal lignin in a loblolly
pine mutant. Science 1997, 277, 235-239.
Ralph, J .; Hatfield, R. D.; Piquemal, J .; Yahiaoui, N.; Pean,
M.; Lapierre, C.; Boudet, A.-M. NMR characterization of
altered lignins extracted from tobacco plants down-regulated
for lignification enzymes cinnamyl-alcohol dehydrogenase
and cinnamoyl-CoA reductase. Proc. Natl. Acad. Sci. U.S.A.
1998a , 95, 12803-12808.
Ralph, J .; Peng, J .; Lu, F. Isochroman structures in lignin: a
new â-1 pathway. Tetrahedron Lett. 1998b, 39, 4963-4964.
Sarkanen, S. Template polymerization in lignin biosynthesis.
In Lignin and Lignan Biosynthesis; Lewis, N. G., Sarkanen,
S., Eds.; American Chemical Society: Washington, DC,
1998; pp 194-208.
Sarkanen, K. V.; Hergert, H. L. Classification and Distribution.
In Lignins. Occurrence, Formation, Structure and Reactions;
Sarkanen, K. V., Ludwig, C. H., Eds.; Wiley-Interscience:
New York, 1971; pp 43-94.
Sederoff, R. R.; MacKay, J . J .; Ralph, J .; Hatfield, R. D.
Unexpected variation in lignin. Curr. Opin. Plant Biol. 1999,
2, 145-152.
Seta¨la¨, H.; Pajunen, A.; Rummakko, P.; Sipila¨, J .; Brunow, G.
A novel type of spiro compound formed by oxidative cross-
coupling of methyl sinapate with a syringyl lignin model
compound. A model system for the â-1 pathway in lignin
biosynthesis. J . Chem. Soc., Perkin Trans. 1 1999, 461-
464.
Terashima, N.; Atalla, R. H.; Ralph, S. A.; Landucci, L. L.;
Lapierre, C.; Monties, B. New preparations of lignin polymer
models under conditions that approximate cell wall lignifi-
cation. II. Structural characterization of the models by
thioacidolysis. Holzforschung 1996, 50, 9-14.
Brunow, G.; Karlsson, O.; Lundquist, K.; Sipila, J . On the
distribution of the diastereomers of the structural elements
in lignins: the steric course of reactions mimicking lignin
biosynthesis. Wood Sci. Technol. 1993, 27, 281-286.
Davin, L. B.; Wang, H.-B.; Crowell, A. L.; Bedgar, D. L.;
Martin, D. M.; Sarkanen, S.; Lewis, N. G. Stereoselective
biomolecular phenoxy radical coupling by an auxiliary
(Dirigent) protein without an active center. Science 1997,
275, 362-366.
Denton, F. R. Beetle J uice. Science 1998, 281, 1285.
Dickey, E. E. Liriodendrin, a new lignan diglucoside from the
inner bark of yellow poplar (Liriodendron tulipifera L.). J .
Org. Chem. 1958, 23, 179.
Eberhardt, T. L.; Bernards, M. A.; He, L.; Davin, L. B.; Wooten,
J . B.; Lewis, N. G. Lignification in Cell Suspension Cultures
of Pinus taeda: In Situ Characterization of a Gymnosperm
Lignin. J .Biol. Chem. 1993, 268, 21088-21096.
Freudenberg, K.; Dietrich, R. d,l-Pinoresinol, ein weiteres
Zwischenprodukt der Ligninbildung. Chem. Ber. 1953, 86,
755-758.
Freudenberg, K.; Neish, A. C. Constitution and Biosynthesis
of Lignin; Springer-Verlag: Berlin, Germany, 1968.
Freudenberg, K.; Sidhu, G. S. Die absolute Konfiguration der
Gruppe des Sesamins und Pinoresinols. Che. Ber. 1961, 94,
851.
Freudenberg, K.; Hen, C.-L.; Harkin, J . M.; Nimz, H.; Rener,
H. Observations on lignin. Chem. Commun. 1965, 224-225.
Gang, D. R.; Fujita, M.; Davin, L. D.; Lewis, N. G. The
‘abnormal lignins’: mapping heartwood formation through
the lignan biosynthetic pathway. In Lignin and Lignan
Biosynthesis; Lewis, N. G., Sarkanen, S., Eds.; American
Chemical Society: Washington, DC, 1998; pp 389-421.
Harkin, J . M. Ligninsa natural polymeric product of phenol
oxidation. In Oxidative Coupling of Phenols; Taylor, W. I.,
Battersby, A. R., Eds.; Dekker: New York, 1967; pp 243-
321.
Helm, R. F.; Toikka, M.; Li, K. C.; Brunow, G. Lignin model
glycosides: Preparation and optical resolution. J . Chem.
Soc., Perkin Trans. 1 1997, 533-537.
Hergert, H. L. Secondary lignification in conifer trees. In
Cellulose Chemistry and Technology; ACS Symposium Se-
ries 48; Arthur, J . C., Ed.; American Chemical Society:
Washington, DC, 1977; pp 227-243.
Kagan, H. B., Ed. Determination of configurations by dipole
moments, CD or ORD; Georg Thieme: Stuttgart, Germany,
1977; Vol. 2, p 198.
Lewis, N. G.; Davin, L. B. The biochemical control of monoli-
gnol coupling and structure during lignan and lignin bio-
synthesis. In Lignin and Lignan Biosynthesis; Lewis, N. G.,
Sarkanen, S., Eds.; American Chemical Society: Washing-
ton, DC, 1998; pp 334-361.
Received for review February 12, 1999. Revised manuscript
received April 30, 1999. Accepted May 3, 1999. This work was
supported in part by USDA-National Research Initiatives
Competitive Grants 97-02208 (Improved Utilization of Wood
and Wood Fiber) and 96-35304-3864 (Plant Growth and
Development).
Lindberg, B. Epi-pinoresinol. Acta Chem. Scand. 1950, 4, 391-
392.
Lu, F.; Ralph, J . Derivatization followed by reductive cleavage
(DFRC method), a new method for lignin analysis: protocol
J F9901136