Chapter4 Isolation and Analysis of Lignin–Carbohydrate Complexes Preparations with Traditional and Advanced Methods:
A Review
Mikhail Balakshin*,Ewellyn Capanema*and Alex Berlin{
*Renmatix Inc.,King of Prussia,Pennsylvania,USA
{Department of Protein Chemistry,Novozymes,Inc.,Davis,California,USA Chapter Outline
Introduction84 Isolation of LCC Preparations
from Wood and Pulps87 Isolation of LCC from Wood87
Isolation of LCC from Pulps93 LCC Analysis95 Model Compound Studies95
Wet Chemistry Methods96
Spectroscopic Methods100 Quantification of LCC Linkages102 Pitfalls to be Overcome in
2D Quantitative Analysis of
LCC Linkages104General Discussion105 What Is the Best LCC
Preparation and Analytical
Method?105 Current Understanding of
the LCC Structure108 Implementation of Enzymes in LCC Studies110 Conclusions111 References111
Studies in Natural Products Chemistry,Vol.42.https://www.doczj.com/doc/8911399946.html,/10.1016/B978-0-444-63281-4.00004-5
?2014Elsevier B.V.All rights reserved.83
INTRODUCTION
Polysaccharides and lignin are the major plant polymers.Carbohydrates make
up the majority of the components present in wood.In wood species,approx-
imately 40–50%of the biomass is cellulose and 20–30%is hemicellulose
[1–3].Lignin is the most abundant aromatic plant polymer on Earth.The lig-
nin content in typical wood species ranges from ca.20%to 35%[1,2,4–7].
Cellulose,a structural component of the cell wall in green plants and certain
microorganisms,is a long linear homopolymer composed of anhydro-D -
glucopyranose units linked by b -(1!4)-glycosidic bonds [1,2].The degree of
polymerization (DP)of cellulose ranges from about 300to about 10,000
(2500–5000nm in length).In woody species,the cellulose DP typically ranges
from about 1500to about 5000[8].Both covalent and hydrogen bonding play
an important role in the formation of the cellulose crystalline structure
providing the polymer with stiff and linear characteristics.Bundling of lateral
cellulose molecules is possible due to a myriad of hydrogen bonds.This large
number of hydrogen bonds enables a relatively strong lateral association of cel-
lulose molecules,resulting in the formation of crystalline regions in the cell
wall.Cellulose contains both crystalline and amorphous regions.The relative
percent content of these regions depends on the plant species.For instance,
the cellulose degree of crystallinity in wood ranges from 67%to 90%[1].
Galactoglucomannan (Fig.1)is the major hemicellulose constituent of
softwood (SW)species,representing approximately 16%of the wood [1–3].
Galactoglucomannans can be divided into two fractions.One fraction has a
low galactosyl unit content with a galactosyl/glucosyl/mannosyl ratio of
0.1:1:4.The galactosyl-poor fraction is usually referred to as glucomannan.
The other,the galactosyl-rich fraction,has a corresponding ratio of 1:3:4.
Galactoglucomannans have a main chain consisting of (1!4)-linked b -D -
glucopyranosyl units and b -D -mannopyranosyl units.The D -glucosyl and
D -mannosyl units in the main chain are arranged randomly [9].The a -D -
galactopyranosyl units are linked as a single-unit side chain to both D -glucosyl
and D -mannosyl units by (1!6)bonds.In addition,about half of the manno-
syl units are O-acetylated at C-2or C-3in equal proportion [10].The weight-
average molecular weight (M w )of a galactoglucomannan isolated from
Scotch pine (Pinus sylvestris )is 23,300Da [11].
Another major hemicellulose component in softwood is arabino-(4-O -
methylglucurono)-xylan (Fig.1),which represents about 7–15%of the
wood [3].Arabinoglucuronoxylans are composed of approximately 200
(1!4)-linked b -D -xylopyranosyl units that are partially substituted by 4-O -
methyl-a -D -glucuronic acid moieties at the C-2position,with an average
of one glucuronic acid unit per 5–6xylopyranosyl units [12].In addition,
a -D -arabinofuranosyl units are bonded at the C-3position of the main
xylan chain,with an average of 1.3arabinofuranosyl units per 10xylopyrano-
syl units.
Studies in Natural Products Chemistry
84
3C O
O
β-D-Mannopyranose β-D-Glucopyranose α-D-Galactopyranose
α-D-Galactopyranose α(1-6)
α
(1-6)Acetyl group α-L-Arabinofuranose 4-O-Methyl-α-D-glucuronopyranose
β-D-Xylopyranose
α(1-2)O 3Xylan is the main component of hardwood (HW)hemicelluloses (10–35%
of the wood).In contrast to softwood Xylans,hardwood xylans do not contain
a -D -arabinofuranosyl units and,therefore,they are glucuronoxylans.The con-
tent of 4-O -methyl-a -D -glucuronic acid moieties in HW xylans is lower than
in the softwood ones,about 1unit per 10xylopyranosyl units [1–3].In addition,
xylopyranosyl units in HWs are partially acetylated at the C-2and/or C-3posi-
tion with total amounts of 3.5–7acetyl groups per 10xylopyranosyl units [13].
Lignin is a high molecular weight polymeric substance produced in vivo
by an enzyme-initiated dehydrogenative polymerization of coniferyl alcohol
and its derivatives [2,4–7,14,15].Lignin also can be described as an irregular
aromatic biopolymer composed by phenylpropane (or C 9)units of the
p -hydroxyphenyl (H),guaiacyl (G),and syringyl (S)types (Fig.1).Softwood
lignins are predominantly of G-type.HW lignins contain both S-and G-units
in different ratios depending on the wood species.The amounts of H-units in
wood lignins are usually small.However,significant amounts of H-units can
be found in grass lignins.The monomeric lignin units are linked in the mac-
romolecule by various ether and C d C bonds.The major types of interunit
Chapter 4Isolation and Analysis of LCC Preparations 85
linkages in native lignins are shown in Fig.1[2,4–7,14–16].The main interunit
linkage in native lignin is the arylglycerol-b -aryl ether (b -O-40)bond.The
amount of b -O-40structures is about 45per 100monomeric lignin units (or
mol.%)in softwood lignins and up to 60–65mol.%in hardwood lignins.The
main lignin functional groups are the phenolic and aliphatic hydroxyl groups,
as well as carbonyl and carboxyl groups.It should be noted that the structure
of the native lignin is dramatically modified during chemical processing of
wood,including different types of pulping and biorefinery processes [7,16].
There is numerous evidence that lignin and polysaccharides are covalently
linked,forming manifestations of the so-called “lignin–carbohydrate com-
plex”(LCC)[2,7,17–19].In spite of there being relatively low amounts of
these linkages in wood [20,21],they play a very important role,as almost
all wood lignin is covalently linked to polysaccharides,mainly to hemicellu-
loses [22].There are three types of LCC linkages:benzyl ether,
phenyl
3
O Carb C 2: R = C2 or C3 in Xyl, Glc,
Man, Gal, Ara
A B C C 1-carb E H 3D : R = H
D': R=Ac
33
H 3G S
3F I
H 333
3
L K FIGURE 1Major structural moieties of hemicelluloses,lignin,and LCC linkages.(A)Phenyl
glycoside,(B)g -ester,(C)benzyl ether LCC linkages,(D)b -O-4/a -OH,(D 0)b -O-4/a -OH
g -acetylated,(E)spirodienone,(F)phenylcoumaran (b -5),(I)resinol (b –b ),(K)cinnamyl alcohol
type;(I)dibenzodioxocinlignin structures;G,guaiacyl;S,syringyl lignin units.
Studies in Natural Products Chemistry 86
Chapter4Isolation and Analysis of LCC Preparations87 glycoside,and esters(Fig.1).The occurrence of stable lignin–carbohydrate bonds creates significant problems in selective separation and isolation of lig-nin and carbohydrate preparations from lignocellulosics.Furthermore,lin-kages between lignin and carbohydrates also impede efficient selective separation of the wood components in biorefining processes.Therefore, understanding the LCC structure is of great fundamental and practical importance.
In addition to the chemical structure,the arrangement of the main compo-nents in the cell wall and their interaction is of primary importance to the understanding of the ultrastructure of the cell wall and of the physico-mechanical and other properties of wood materials[2].However,we have left the ultrastructure of the cell wall out of the scope of the present review to avoid unnecessary complication and have decided to focus on the types and quantity of chemical linkages between lignin and carbohydrates in lignocellulosics.
In spite of extensive research studies on LCC chemical structures,our knowledge in this field is still insufficient.The complex nature of the cell wall structure as well as the heterogeneous and unstable nature of LCC requires development of efficient preparative methodologies for intact and selective isolation of desired LCC moieties and their analysis.The aim of this chapter is a critical review of traditional and advanced methods for isolation and anal-ysis of LCC preparations.
ISOLATION OF LCC PREPARATIONS FROM WOOD
AND PULPS
Most methods for LCC analysis require isolation of LCC preparations from lignocellulosic materials.The main problems in LCC isolation are associated with the complex structure of the cell wall and the interaction of its compo-nents.An appropriate isolation procedure should produce a representative LCC preparation and minimize structural changes during isolation. Isolation of LCC from Wood
It is important to mention that according to the classical view,LCC should be isolated from wood after extraction of“free”lignin,such as the milled wood lignin(MWL)preparation[2,7,17–19,23].However,this view can now be challenged due to the recently demonstrated fact that almost all lignin is linked to polysaccharides[22].This important finding implies that intact LCC in situ could be a large macromolecule encompassing an entire fiber, or even an entire tree due to multiple redundant crosslinking among carbohy-drates and lignin macromolecules.Therefore,it is important to distinguish between the intact LCC in situ and LCC preparations.By LCC preparations,
we define preparations isolated from lignocellulosics to analyze specific struc-
tural features of the native LCC (in other words,preparations useful to pro-
vide information on LCC).Obviously,an LCC preparation is only a part of
the whole intact LCC,but its study makes possible a better understanding
of the specific structural characteristics of the large LCC macromolecule.
Most of LCC preparations discussed in this review are considered to be useful
predominantly for the analysis of specific linkages between lignin and carbo-
hydrates rather than to the elucidation of the detailed structure of the entire
LCC macromolecule.
The “classical”wood LCC preparation is considered to be the Bjo
¨rkman LCC (or its analogs)isolated from milled wood after preextraction of
MWL [24].Such preparations as MWL and cellulolytic enzyme lignin
(CEL)were considered experimental preparations used exclusively to study
lignin structure.However,more recently,it has been demonstrated that these
types of preparations also can be very valuable in the analysis of LCC
linkages [21].
A schematic description of different methods for the isolation of LCC pre-
parations from wood is summarized in Fig.2.Certain fractions of hardwood
and annual plant LCCs can be extracted with diluted NaOH (0.3%)under
mild conditions (1h reflux under nitrogen atmosphere)[16–19,25–28].The
alkali-soluble material contained mostly lignin and up to 50%of carbohy-
drates and therefore was referred to as alkali-soluble LCC (Alk-LCC)
(Fig.2).The Alk-LCC is enriched with H-and G-units and phenolic OH as
compared to the bulk lignin in the same species.In addition,significant differ-
ences in the number of interunit linkages are observed in Alk-LCC and bulk
lignin [25–28].
Most methods for LCC isolation from wood require ball milling,which is
carried out so that the cell wall matrix can be degraded and the lignin and
LCC fragments can be released and extracted.Extraction of the milled wood
with 96%dioxane produces crude milled wood lignin (MWLc)[24].
O
FIGURE 2Isolation of LCC preparations from wood.
Studies in Natural Products Chemistry
88
Chapter4Isolation and Analysis of LCC Preparations89 Fractionation of the pine crude MWL with90%acetic acid(AcOH)results in a preparation called LCC–AcOH[21,29],and purified MWL.Purified pine MWL(PMWLp)contains a very small amount of sugars,indicating that most of lignin–carbohydrate-linked structures contained in the original crude MWL are concentrated in the PLCC–AcOH preparation.The same is not true for the fractionation of birch MWLc,as birch purified MWL(BMWLp)contains sig-nificant amounts of sugars[21].Earlier,using a more complex protocol,a preparation similar to the LCC–AcOH(called LCC–W)was obtained[30].
The enzymatic hydrolysis of milled wood with an enzymatic cellulase/hemi-cellulase complex is very useful for isolation of LCC preparations[31–33].Par-tial enzymatic hydrolysis of the milled wood with an endoglucanase followed by wet chemistry separation allows for the isolation of various LCC fractions from wood and pulps[22].The complete enzymatic hydrolysis of milled wood results in a preparation called milled wood enzymatic lignin(MWEL)[20](Fig.2).The extraction of MWEL with aqueous dioxane produces soluble CEL prepara-tions[33].Another variant of CEL preparations are those isolated by enzymatic hydrolysis of the residue after extraction of MWL[21].
Interestingly,MWLs and CELs isolated from HWs after mild alkali preex-traction of wood sawdust were found to contain very small amounts of sugars, even without the purification typically carried out for MWL[25–28,34].This finding implies that a significant amount of carbohydrates in HWs is attached to lignin via alkali-labile linkages,probably of ester type.However,this same conclusion is not applicable to a softwood material(loblolly pine).
The extraction of the wood residue with dimethylsulfoxide(DMSO)after extraction of MWL produces Bjo¨rkman’s LCC[23]which has been consid-ered for a long time as the“classical”and unique LCC preparation.Other sol-vents and water were also used to extract LCC preparations from the residue obtained after MWL isolation[17–19,35].The concentration of lignin–carbohydrate linkages by enzymatic hydrolysis of the carbohydrate compo-nents of the DMSO-extracted LCC preparations[36]is a useful approach for LCC linkages analysis.
In summary,LCC preparations can be classified as“carbohydrate-rich LCC”(Bjo¨rkman’s LCC and similar ones,enzymatic LCC fractions[22]), and“lignin-rich LCC”(MWEL,CEL,MWLc).
Degradation of LCC Linkages During Isolation Procedure
The degradation of LCC linkages during LCC isolation should be taken into account when extrapolating the results of the composition of isolated LCC preparations to the analysis of wood LCC in situ.
Mechanical Degradation
Since most of the LCC isolation methods require wood ball milling(Fig.2),it is important to understand how this procedure affects the lignin/LCC
structure.It is known that some changes in lignin structure occur during ball
milling,particularly the increases of carbonyl and phenolic OH groups and the
decrease of molecular mass [24,33].Recent studies showed that ball milling
did not cause changes in the aromatic ring of lignin units,but resulted in some
cleavage of b -O-4structures in the whole wood lignin [37,38].
Importantly,the degree of lignin degradation during ball milling is inde-
pendent of the milling intensity and apparatus used [16,27,28,37,38,39],but
it correlates well with the “extracted lignin”yield [37],a value equivalent
to the MWLc yield.Therefore,it is very important to report the lignin yield
for the preparations when ball milling is used.Unfortunately,this information
is often omitted in scientific reports.
Our studies [21]showed that the amount of lignin b -O-4units in LCC pre-
parations decreased in the row:CEL >MWLp (purified milled wood lig-
nin)>MWLc >LCC–AcOH,indicating that the CEL preparation is the least
mechanically degraded lignin/LCC fraction and that the LCC–AcOH is the
most degraded one.
It is very likely that a part of benzyl ether LCC linkages of the type C
(Fig.1)is also degraded during ball milling.This can explain the lowest
and the highest amount of benzyl ethers observed in the birch LCC–AcOH
and CEL preparations,correspondingly (Table 1).The phenyl glycoside lin-
kages are apparently stable during milling,as their amount in LCC–AcOH
preparation is significantly increased with extended ball milling [21].
Enzymatic Degradation of LCC Linkages
The CEL preparations contain low amounts of phenyl glycoside and g -ester
LCC structures [21].The cleavage of phenyl glycoside linkages by b -glycosi-
dases,present in cellulase preparations,has been discussed previously
[19–21,39,40,41].Under certain conditions,b -glycosidases and other glycosi-
dases can present transglycosylating activity of OH groups in LCC carbohy-
drates leading to formation of novel bonds originally absent in wood LCC
in situ .If the transglycosylation were confirmed in LCC isolation,some
reports could require reassessment or revised studies with enzymes lacking
transglycosylating activity.In any case,the transglycosylating rates of this
type of hydrolytic enzymes tend to be much lower than the hydrolysis rates
under typical enzymatic hydrolysis conditions.
Low amounts of ester LCC structures as well as g -acetylated lignin moi-
eties in the CELs [21]indicate that cellulase preparations should also possess
esterase activity [41].Esterase activity in commercial cellulase preparations is
commonly found,since many industrial fungal hosts such as Trichoderma sp.
or Aspergillus sp.cosecrete with cellulases a range of esterases,the so-called
xylan “debranching enzymes”,including acetylxylan esterase (AXE)(EC
3.1.1.72),feruloylesterases (EC 3.1.1.73),and p -coumaroylesterases [42].
Studies in Natural Products Chemistry
90
TABLE1Sugar Compositions of LCC Preparations and the Amounts of LCC Linkages evaluated from NMR Spectra Acquired on a950MHz Spectrometer Equipped with a BrukerCryoProbe TM[21]
Preparation Sugars(%)Sample Sugar Composition(%)Neutral Sugars
LCC Linkages,per100Ar a
Benzyl Ethers
PhGly g-Esters Rha Ara Xyl Man Gal Glc C1C2
PMWLc9.8 1.518.517.116.614.532.6 3.9–1.6b nd 2.0 1.1 PLCC–AcOH28.0 1.920.517.616.615.028.4 3.9–1.6b nd7.2 2.6 PCEL13.50.9 6.319.634.019.419.9 5.3–3.0b nd0.7 1.5 BMWLc22.2 1.9 1.590.5 1.0 1.4 3.7 1.2nd 3.2 2.8 BCEL10.2 3.5 2.773.1 2.6 5.412.7 1.80.4c0.5 1.1
a Per100“Aromatic rings.”As each lignin monomeric unit possesses an aromatic ring,these values are equivalent to mol percentage.
b Shows maximal and minimal value calculated by subtraction of the amount of benzyl ethers in a“sugar-free”preparation(PMWLp)[21].
c Semiquantitative estimation;nd,not detected.
Increasing the Yields of LCC Preparations
Significant efforts have been made recently to increase the yield of the
isolated lignin/LCC preparations and to obtain more representative substrates.
Lu and Ralph [43]have suggested a method that results in complete solubili-
zation of wood by dissolving acetylated milled wood in NMI/DMSO to obtain
a preparation called “acetylated cell walls”(Ac-CW).This approach allows
for the characterization of wood components by solution-state NMR spectros-
copy,which provides much higher resolution than solid-state NMR.Matsu-
moto’s group suggested later an alternative approach consisting in
dissolving all wood material after very mild milling in the DMSO/LiCl
(6%)system [44].The advantage of this method over the Ac-CW protocol
is a better solubility of the preparation obtained under much milder ball mill-
ing conditions and,correspondingly,lower degradation of the wood compo-
nents.Another advantage of this method is that the prepared sample can be
dissolved without modifications such as acetylation and,therefore,more valu-
able information on the structure of native wood components can be obtained.
We have recently attempted to maximize the yield of the CEL prepara-
tion [45]using a protocol for wood dissolution developed at Tokyo Univer-
sity [44](Fig.3).We expected that enzymatic hydrolysis of the milled
wood,after dissolution and regeneration,should have been more successful,
as this procedure would have decomposed the ultramolecular structure of cel-
lulose and therefore would have made the material more accessible to enzy-
matic hydrolysis.Indeed,wood regeneration significantly increased the
yield of the CEL preparation (Table 2).Furthermore,a dramatic increase in
the yield was obtained by using 80%(v/v)dioxane extraction instead of the
classical 96%(v/v)dioxane extraction.Eventually,a CEL from the regener-
ated wood (RCEL-80)was obtained under very mild ball milling conditions
with an 83%yield (the yield of MWL was 15%,which means very little lig-
nin degradation occurred [44]).When the milling time was increased to 3h,
almost all lignin (93%of the original wood lignin)was obtained as a
RCEL-80preparation.Thus,the isolation of whole wood lignin and asso-
ciated LCC moieties was possible with minimal degradation.This result
allowed for detailed analysis of whole wood lignin and LCC moieties by
FIGURE 3Isolation of regenerated CEL (RCEL)[45].
Studies in Natural Products Chemistry
92
high-resolution NMR methods.The significantly lower carbohydrate content
(about 5%)made the NMR analysis of LCC linkages in the RCEL prepara-
tions easier and more accurate than that if the Ac-CW and “dissolved wood”
preparations were used.
Isolation of LCC from Pulps
Similarly to the isolation of LCC from wood,isolation of LCC from pulps can
be performed using enzymatic hydrolysis.These preparations are usually
called “enzymatic residual lignins”[46–49].However,similarly to CEL,they
contain significant amounts of LCC linkages proven by two-dimensional
nuclear magnetic resonance (2D NMR)methods and therefore can be consid-
ered as LCC preparations [48,50,51].Ball milling is not needed for the isola-
tion of LCC from most chemical pulps,and as a result mechanical
degradation can be avoided.
The isolation of enzymatic residual LCCs from unbleached softwood Kraft
pulps is a well-established procedure producing preparations with high yields
(estimated as lignin in isolated preparations per lignin in the original pulp)
and relatively low enzyme impurities [46–48].In contrast,significant pro-
blems are encountered when similar procedures are applied to the isolation
of enzymatic residual LCCs from hardwood Kraft pulps and semi-bleached
pulps [47].Low yields (25–30%)make these preparations nonrepresentative
for the whole pulp LCC.A very large amount of protein impurities
TABLE 2Isolation and Analysis of Red Alder RCEL Preparations [45]
Preparation Milling
Time (h)
Dioxane Concentration (v/v)Yield (%)a Amounts,per 100Ar b
b -O-4/a -OH S/G Ratio Benzyl Ether LCC MWL
c 2
9615CEL-962
9629CEL-802
8060RCEL-962
9635RCEL-802
8083MWLc 3
9621CEL-963
963649 1.19 1.3CEL-803
807049 1.20 1.4RCEL-963
964648 1.14 1.2RCEL-80
3809350 1.21 1.4a
Corrected for the sugar contents in the preparations.b Per 100“Aromatic rings.”As each lignin monomeric unit possesses an aromatic ring,these values
are equivalent to mol percentage.Chapter 4Isolation and Analysis of LCC Preparations 93
(15–35%)leads to significant problems in lignin analysis,particularly where
spectroscopic methods are involved [49].We were able to dramatically
decrease protein contaminations in hardwood residual lignin preparations
(to 1–6%in nonpurified lignins)and to increase the yields to 40–50%by
applying cellulase preparations with high specific activity and by optimizing
the enzyme loading [26,49,50].The highest optimal enzyme loading required
was that for LCC isolation from an Eucalyptus globulus pulp and the lowest
for a birch pulp with similar lignin content [26,49].In the future,enzymes
with lower binding affinity toward lignin/LCC [52]could be used to further
improve this methodology.
In 2005,Berlin et al.[52]showed that certain lignocellulose hydrolytic
enzymes,such as endoglucanases,in particular those lacking cellulose
binding domains,do present lower affinity toward lignin/LCC.In that same
publication,the authors postulated that if LCC linkages in certain lignocellu-
losic biomass were present in significant amounts they could be partially
responsible for limiting biomass enzymatic hydrolysis rate due to LCC or
LCC-like linkages (formed during biomass pretreatment)nonproductive bind-
ing observed for many cellulases and hemicellulases [53].
The “LCC nonproductive binding”can be defined as an interaction,
reversible or irreversible,between an enzyme and a fraction of the biomass
substrate containing recalcitrant lignin–carbohydrate moieties,which does
not result in LCC bond cleavage and translates into a temporarily or perma-
nent reduction of the enzyme active concentration.Nonproductive binding
in enzymatic degradation of biomass can occur between biomass-degrading
enzymes and lignin or LCC,between enzymes and new chemical entities gen-
erated during biomass pretreatment,and even between these enzymes and
recalcitrant cellulose.In addition,other specific enzyme-related factors such
as strong protein–protein interactions,especially at medium and high protein
loadings,could be another important limiting factor contributing to the bio-
mass hydrolysis rate reduction observed,in particular,for processive glyca-
nases such as the Trichoderma reesei CBH I (Cel7A),but this is something
yet to be proven.
During biomass physicochemical pretreatment,a prerequisite for efficient
enzymatic hydrolysis,the structure of plant biomass changes dramatically
yielding often new chemical entities absent in the original raw biomass.
Therefore,we cannot exclude here the possibility of seeing an increase in
the number of LCCs or LCC-like moieties in biomass resulting from the pre-
treatment process and leading to the documented increased recalcitrance of
residual biomass,particularly true at late hydrolysis stage in high solids reac-
tions.We do believe that in some cases,nonproductive binding could result
from binding of enzymes to recalcitrant lignin–carbohydrate moieties or
related structures which are present in increased concentrations toward the
end of the biomass enzymatic hydrolysis.An indirect evidence of the latter
is that the addition of fresh enzymes during the late stages of biomass
Studies in Natural Products Chemistry
94
Chapter4Isolation and Analysis of LCC Preparations95 enzymatic hydrolysis often leads to a significant recovery of the initial reac-tion rate of cellulose conversion[52].There is no documented evidence that the nonproductive binding in pretreated biomass can happen namely at the linkage of the lignin and the carbohydrate moiety,but it is something one can hypothesize about.
Interesting that,recently,the Canadian company,Iogen Energy Corpora-tion(Ottawa,Canada)claimed in a patent certain structural moieties in lignocellulose-degrading enzymes that seem to be responsible for their non-productive binding to lignin[54].
It is important to mention that the low yields obtained for hardwood pulp LCC preparations,which is lower with the decrease in the lignin content in pulps,are an indication of the fact that a significant portion of lignin remains soluble in aqueous solution after enzymatic hydrolysis of the pulps[49]. Apparently,this fraction consists of low molecular weight lignin fragments, likely linked to carbohydrates,which makes it more hydrophilic and therefore more water soluble.The frequency of LCC bonds(per lignin unit)in this frac-tion should be much higher than that in the fraction of higher molecular mass residual lignin.Therefore,it would be of interest to isolate the water-soluble LCC fraction and characterize it[49].
LCC ANALYSIS
The majority of methods in LCC analysis can be divided into model studies, wet chemistry methods,and spectroscopic techniques.
Model Compound Studies
Model compound experiments consist in mimicking reactions of formation and transformation of lignin–carbohydrate linkages with individual chemical compounds under appropriate reaction conditions.The benzyl ether and ben-zyl ester LCC bonds were originally suggested by Freidenberg based on his quinone methide theory of lignification and demonstrated by model experi-ments with the dehydrogenation polymer(DHP)[14,55,56].Model experi-ments were used later to mimic the formation of phenyl glycoside linkages[57]and a lignin–xylan complex[58].Model compound experiments were also useful to study LCC linkages in forages[59,60].However,one must be very careful with the application of these results to LCC in plants,since it is very difficult to properly model the reaction conditions of lignin biosynthe-sis in living cells.As an example,significant efforts to model lignin structure with DHP experiments have been still unsuccessful in mimicking the exact structure of a real lignin polymer.Therefore,model compound experiments could be a valuable tool in predicting various LCC linkages.However,these results must be verified with data obtained from real lignocellulosic substrates.
Another valuable role of model compound studies is the generation of
NMR databases that have been extremely important for NMR studies on
isolated LCC preparations [59–65].
Wet Chemistry Methods
Sugar Analysis
A routine sugar analysis of LCC preparations is very informative.It provides
valuable information on the involvement of different polysaccharides in the
LCCs structures,when the degradation of the polysaccharide component is
minimal (carbohydrate-rich LCC preparations)[18,22,23,28].In the case of
lignin-rich LCC preparations,when most of the carbohydrates are decom-
posed by enzymatic hydrolysis and/or ball milling,the carbohydrate composi-
tions of the residual sugars provide information on which specific
carbohydrate units can be linked to lignin.For example,a study of a pine
LCC–AcOH preparation showed that it contained significant amounts of arab-
inose and galactose (Table 1).The arabinose:xylose and galactose:mannose
ratios indicated that considerable fractions of arabinose and galactose
stemmed from pectins rather than from arabinoglucuronoxylan and galacto-
glucomannan [29].The composition of sugars in pine CEL is appreciably dif-
ferent from that in pine LCC–AcOH.The former contains a much lower
(ca.one-third)amount of arabinose than the latter.The amount of galactose
in the pine CEL is still relatively high.However,there is also a significantly
higher amount of mannose in pine CEL than that in the LCC–AcOH,indicat-
ing that a portion of the galactose originated from galactoglucomannan (rather
than from pectin galactan)is higher in the pine CEL as it compares to the pine
LCC–AcOH preparation.In contrast to the pine preparations,the birch pre-
parations contain xylan as the main sugar constituent.Its amount is about
90%of total sugars in the birch MWLc and LCC–AcOH preparations and
about 70%in the CEL sample (Table 1)[21].
Degradation Techniques
Different degradation techniques are commonly used for the analysis of LCC
isolated from plant tissues.This approach includes cleavage of lignin–
carbohydrate bonds and identification of the resulting products by way of
alkaline hydrolysis (saponification)[20],acid hydrolysis [36,66],Smith deg-
radation [67],ozonolysis [68],methylation analysis [69–72],and DDQ (2,3-
dichloro-5,6-dicyano-1,4-benzoquinone)oxidation [18,73–76].The two last
techniques are the most common for the analysis of carbohydrate linkage sites
in LCC preparations.
The alkaline hydrolysis (saponification)of LCC preparations under mild
conditions results mostly in cleavage of ester bonds between benzyl moieties
in lignin and glucuronic acid moieties in carbohydrates.The structure of the
Studies in Natural Products Chemistry
96
ester LC bonds in the original LCC is postulated based on the analysis of
sugar composition and liberation of carboxyl and hydroxyl groups in different
fractions after alkali hydrolysis [20].Selective acid and alkaline hydrolysis of
an LCC preparation followed by detection of newly formed phenolic and ben-
zyl alcohol hydroxyl was used to evaluate the amount of benzyl ether LCC
linkages in the preparation [66].
In the methylation analysis [69–72],an LCC preparation is methylated
with CH 3I in DMSO under strong alkaline conditions.The resulting methy-
lated LCC sample is then subjected to acid-catalyzed hydrolysis to obtain par-
tially methylated sugar monomers.The hydrolysate is then reduced and
acetylated and the resulting alditol acetate mixture is analyzed by GC and
GC–MS.The nature of the carbohydrate bonding sites is then elucidated from
the unmethylated (acetylated)sites of the monomeric saccharides which are
identified and quantified (Fig.4).
The DDQ oxidation technique identifies only the linkage sites of carbohy-
drates involved in benzyl ether-type LCC bonds (Fig.5)[18,73–76].Briefly,
an LCC preparation is thoroughly acetylated,then subjected to DDQ oxida-
tion to induce the oxidative cleavage of LCC bonds of the benzyl ether type
into the corresponding a -carbonyl in acetylated lignin and terminal hydroxyl
groups in acetylated carbohydrate moieties,respectively.The resulting carbo-
hydrate mixture is then methylated by Prehm’s procedure and
hydrolyzed
O O
R
O O R
AcO 2
3I(OH -)
-
2
22FIGURE 4Analysis of LCC linkages by a methylation technique.
Chapter 4Isolation and Analysis of LCC Preparations 97
with 2M trifluoroacetic acid.The hydrolysates are reduced with sodium boro-
hydride,and then acetylated.The constituents of the resulting partially methy-
lated alditol acetate mixtures are finally analyzed by GC and GC–MS.The
position of the methoxyl group indicates the position of the carbohydrate
bonded to the lignin.Thus,the DDQ oxidation technique shows the relative
proportion of various carbohydrate sites linked specifically to the benzyl car-
bon of the lignin units.
One important issue to be addressed in the DDQ oxidation method is its selec-
tivity.Good selectivity was obtained with individual compounds modeling ben-
zyl ether LCC linkages [18,73].However,the results obtained with DDQ
oxidation of an acetylated MWL preparation followed by 13C NMR analysis [75]
were inconclusive,as 13C NMR cannot provide solid evidence for the disappear-
ance of the benzyl ethers.Recently,it has been demonstrated with a 2D hetero-
nuclear single quantum coherence (HSQC)NMR technique that,in contrast to
the model compound oxidation,the DDQ reaction is not selective for oxidation
of the benzylic position of lignin [77]since significant amounts of side products
were generated in addition to the target a -carbonyl group.Moreover,the reaction
was not complete,as appreciable amounts of the original b -O-4/a -OH and the
ether moieties remained in the lignin after oxidation.It should be mentioned that
the reaction conditions used to analyze lignin with DDQ oxidation are slightly
different from those used for LCC studies [73–75].However,further studies will
require confirmation of the reaction selectivity and its completeness.The latter
can be achieved by applying the DDQ technique to an LCC preparation and
studying the resulting products by multidimensional NMR methods (Fig.6).
HC CH
CH 2OH
O Lignin O
2HC CH CH 2OAc O Lignin O 2C CH CH 2OAc O Lignin O 22OMe O MeO
Lignin
O MeO
Lignin O MeO Lignin
H, Ac H, OH H, Ac AcO –DDQ +1-Prehm's methylation
2-Hydrolysis FIGURE 5Analysis LCC linkages with the DDQ method.
Studies in Natural Products Chemistry 98
FIGURE6Expanded aliphatic region of the HSQC spectra acquired on950MHz spectrometer equipped with a CryoProbe TM:(A)pine LCC–AcOH(insert demonstrates separation of LCC and acetyl g-esters),(B)pine CEL,(C)birch MWLc,and(D)birch CEL.Specific lignin–carbohydrate and lignin structures are labeled according to Fig.1.Carbohydrate signals were labeled as follows:Glc,M,and X are b-D-glucopyranoside,b-D-mannopyranoside,and b-D-xylopyranoside units,correspondingly.M2and M3are b-D-mannopyranoside units acetylated at C-2and C-3positions,correspondingly.X2,X3,and X23are b-D-xylopyranoside units acetylated at C-2,C-3,and at both positions,correspondingly.R a and R b are a-and b-reducing end carbohydrate units,correspondingly.
GlcA and GlcAE are nonesterified and esterified4-O-methyl-a-D-glucuronic acid units,correspondingly.
Both the methylation and DDQ oxidation degradation methods are very
useful for the identification and quantification of specific carbohydrate sites
involved in ether LCC linkages.The main advantage of the DDQ oxidation
technique over other methods is in the specific focus on benzyl ether LCC lin-
kages,whereas the methylation technique does not show differences between
carbohydrate linkage sites involved in LCC bonds and those belonging to the
carbohydrate polymer structure.The conclusions on the LCC linkage sites
derived from the methylation method are made based on the information on
the structures of lignin-free carbohydrates in a given wood species.On the
other hand,the DDQ oxidation method is not able to access any ether LCC
linkages different from the benzyl ether bonds,in contrast to the methylation
technique.Unfortunately,both methods report only relative substitution num-
bers in different monomeric sugar units and do not provide data on the abso-
lute amounts of these centers per the whole carbohydrate fraction and/or LCC
in general.Introduction of an appropriate internal reference might be helpful
in overcoming this method shortcoming.
Spectroscopic Methods
The application of spectroscopic methods to the structural elucidation of LCC
faces significant difficulties because of the heavily overlapping of signals ori-
ginating from different LCC functionalities.IR spectroscopy is useful only in
the study of ester LC bonds [17,20,78].The 13C NMR provides information
on the structure of the carbohydrate or lignin part of the LCCs [74],but not
on the types of LCC bonds due to the fact that their signals are overlapped
with lignin and/or carbohydrate signals [21,29].
An elegant method based on selective enrichment of specific positions of
the lignin side chain with 13C followed by 13C NMR studies [79]was applied
to the study of LCC isolated from labeled wood [80].The author claimed the
presence in this preparation of LCC linkages of acetal,ether,and ester types
at the a -position of the side chain and the absence of LCC bonds at the b -and
g -positions of the side chain.However,a comprehensive discussion revealed
that these conclusions were not properly supported [29].These same problems
did not allow reliable NMR characterization of LCC linkages in a model
Xylan–DHP substrate [58].The main conclusion drawn from these studies
is that 1D 13C NMR is not a reliable tool to investigate LCC linkages even
when using labeled preparations.
Application of 2D NMR methods overcame this obstacle and elucidated
various LCC linkages in preparations isolated from softwoods,HWs,and
pulps [21,25,29,48,50,51].A HSQC correlation 2D NMR technique allowed
for the first time direct detection of phenyl glycoside (Structure A ,Fig.1)
and benzyl ether (Structure C ,Fig.1)LCC linkages [25,29].In contrast to
the common belief [2,14,17,18],no benzyl ester (a -ester)LCC linkages were
detected in these studies.However,g -ester LCC moieties (Structure B ,Fig.1)
Studies in Natural Products Chemistry
100
Chapter4Isolation and Analysis of LCC Preparations101 have been found in the HSQC spectra of the LCC preparations instead [25,29].These linkages were confirmed using a long-range correlation HMBC technique[29].
Although the idea of having g-ether LCC linkages present in wood is not well accepted,this possibility has been discussed[19].Up until now,there has been no experimental evidence to prove or deny this hypothesis.Recently,we suggested the presence of significant amounts of lignin–lignin g-ethers in soft-wood and HW lignins[34,81].Therefore,linkages between g-position of lig-nin and carbohydrates could also be formed by a similar mechanism. Unfortunately,the HSQC spectra of the preparations studied did not allow for any firm statements on the presence or absence of g-ether LCC bonds as the region of their possible resonance,ca.65–75/3.0–4.5ppm,is heavily over-lapped.A3D NMR method should be more informative in verifying the pres-ence of g-ether LCC linkages.
Carbohydrate Analysis in LCC Preparations by2D NMR
The resonance signals of internal anomeric carbons of a-L-arabinofuranoside, b-D-galactopyranoside,b-D-glucopyranoside,b-D-xylopyranoside,and b-D-mannopyranoside were detected in the anomeric carbohydrate region of the softwood LCC preparations[21,75].However,it is important to note that strong signals of xylan anomeric carbons,aroused from various acetylation modes of Xyl units in hardwood preparations,are located very close to the signals assigned earlier[82]to other types of carbohydrates(Fig.6).There-fore,certain considerations are required in order to properly assign carbohy-drate signals in various LCC preparations.In addition to internal carbohydrate anomeric carbons,a group of reducing end anomeric carbons of the a-and b-types can be observed.For the same reason,a detailed analysis of these groups was not performed[21].
Although NMR analysis applied to the general carbohydrate composition in LCCs presents some fundamental interest,from a practical point of view a routine wet chemistry monosaccharide analysis is much simpler and it pro-vides more reliable and,more importantly,quantitative information compared to NMR methods(see,e.g.,Table1).NMR analysis of carbohydrate units in LCCs should be focused on detailed structural characteristics,a piece of infor-mation that routine monosaccharide compositional analysis methods cannot provide,similarly to the NMR investigations conducted on isolated hemicel-luloses(see,for instance,[83]).
The detailed carbohydrate assignment in NMR analysis of LCCs is rather complicated due to heavily overlapping signals.Nevertheless,a few charac-teristic cross peaks,such as mannan units acetylated at the C-2and C-3posi-tion as well as the signals of C-2and C-3acetylated xylan can be observed in softwood and hardwood preparations,correspondingly[21,82](Fig.6).How-ever,this information is not very valuable as it is a well-known fact[3].
An interesting observation was made related to 4-O -methyl-a -D -
glucuronic acid moieties in pine and birch LCC preparations [21].The char-
acteristic signals of nonesterified 4-O -methyl-a -D -glucuronic acid were much
stronger in the spectra of birch preparations in comparison with the pine sam-
ples (Fig.6).This implies that most of the glucuronic acid moieties in the pine
LCC preparations are esterified,whereas the birch LCC preparations still con-
tain significant amounts of nonesterified glucuronic acid moieties [21].
Multidimensional NMR techniques can potentially provide valuable infor-
mation on the carbohydrate sites involved in ether LCC linkages.However,
these efforts were so far unsuccessful,as the region where these signals
should be located (ca.65–75/3.0–4.0ppm)is heavily overlapped and a very
accurate NMR database is required to unambiguously assign the signals in this
region.
Thus,a precise and unbiased assignment of carbohydrate moieties in com-
plex LCC preparations (as well as in dissolved cell wall preparations)is not pos-
sible without a detailed NMR database for sugar moieties acquired in DMSO-d 6
as it has been indicated earlier [21,48,82].Most of the current carbohydrates
databases were acquired in D 2O,while the best NMR solvent for the study of
LCC preparations is DMSO-d 6.The difference in the chemical shift values
between D 2O and DMSO-d 6is often comparable with the difference between
different types of carbohydrate units and this makes their reliable assignment
difficult.Therefore,attempts of detailed assignment of various carbohydrate
moieties are unreasonable before that kind of database is generated.
QUANTIFICATION OF LCC LINKAGES
As it has been shown above,the quantification of LCC linkages with tradi-
tional wet chemistry methods is limited mostly to relative quantification of
carbohydrate sites linked to lignin [18,67–74].
Quantitative information on various types of linkages between lignin and
carbohydrates is very scarce.In fact,we have found only two reports evaluat-
ing the amounts of lignin–carbohydrate linkages in LCC preparations isolated
from wood [20,66].Kos ˇ?′kova
′et al.[66]used mild alkaline and acidic treat-ments of a spruce LCC preparation followed by wet chemistry analysis of
the released phenolic and benzylic OH groups.In Kos ˇ?′kova
′’s study,it was assumed that benzylic OH groups were formed in the cleavage of both
benzyl–aryl lignin–lignin,and benzyl ether lignin–carbohydrate linkages
whereas the formation of phenolic OH occurred in the cleavage of benzyl–aryl
linkages https://www.doczj.com/doc/8911399946.html,ing these assumptions,along with other minor ones,the
authors evaluated the number of benzyl ether LCC linkages as 1per 100mono-
meric lignin units (aromatic rings,Ar)in both p -hydroxy and p -alkoxy units
(free and etherified phenolic position,correspondingly)or 2/100Ar in total.
This approach was well justified by the fundamental knowledge in lignin chem-
istry available at that time.However,a few sources of errors can be revealed
Studies in Natural Products Chemistry
102