The influence of lignin–carbohydrate complexes on the cellulase-mediated saccharification I: Transgenic black cottonwood (western balsam poplar, California poplar) P. trichocarpa

Fuel 119 (2014) 207–213

Contents lists available at ScienceDirect

Fuel

journal homepage: www.elsevier .com/locate / fuel

The influence of lignin–carbohydrate complexes on thecellulase-mediated saccharification I: Transgenic black cottonwood(western balsam poplar, California poplar) P. trichocarpa including thexylan down-regulated and the lignin down-regulated lines

0016-2361/$ - see front matter � 2013 Elsevier Ltd. All rights reserved.http://dx.doi.org/10.1016/j.fuel.2013.11.047

⇑ Corresponding author at: Lignocellulose Chemistry Group, Department of Pulpand Paper Science, Nanjing Forestry University, Nanjing, Jiangsu, China. Tel.: 86 258542 7118.

E-mail address: [email protected] (D.-y. Min).

Dou-yong Min a,b,⇑, Quanzi Li c, Vincent Chiang b,c, Hasan Jameel b, Hou-min Chang b, Lucian Lucia b,d

a Lignocellulose Chemistry Group, Department of Pulp and Paper Science, Nanjing Forestry University, Nanjing, Jiangsu, Chinab Department of Forest Biomaterials, North Carolina State University, Raleigh, NC, USAc Forest Biotechnology Group, Department of Forestry, North Carolina State University, Raleigh, NC, USAd Department of Chemistry, North Carolina State University, Raleigh, NC, USA

h i g h l i g h t s

� The novelty lies in two lines genetic engineering of samples.� The levels of LCCs were quantified by a combination of 13C and 1H–13C HSQC NMR.� LCCs accounted for additional recalcitrance of biomass biodegradation.� The xylan down-regulated samples revealed the effect of hemicellulose on LCCs.� The effect of lignin on LCCs was elucidated by the lignin down-regulated samples.

a r t i c l e i n f o

Article history:Received 19 August 2013Received in revised form 20 November 2013Accepted 20 November 2013Available online 3 December 2013

Keywords:Transgenic P. trichocarpaLignin–carbohydrate complexes (LCCs)Enzymatic saccharificationFermentable sugars1H–13C HSQC NMR

a b s t r a c t

The influence of the putative lignin–carbohydrate complexes (LCCs) on enzymatic saccharification waselucidated for the first time by examining two groups of transgenic black cottonwood (P. trichocarpa)comprised of the lignin down-regulated and the xylan down-regulated lines. Any adventitious contami-nants that could affect the characterization of LCCs and the enzymatic saccharification were removed byperforming a thorough extraction on the samples. The crude milled wood lignin was subsequently iso-lated from which the phenyl glycoside, benzyl ether and c-ester linkages representative of the LCCs wereidentified and quantified with the combination of 13C and 1H–13C Heteronuclear Single Quantum Coher-ence (HSQC) NMR. The result indicated that the samples showed different levels of the three LCC linkages,depending on the xylan and/or lignin content. The correlation between the LCCs and enzymatic sacchar-ification nearly conclusively demonstrated that the LCCs accounting for the recalcitrance of lignocellu-losic biodegradation.

� 2013 Elsevier Ltd. All rights reserved.

1. Introduction

The present petro-economic climate is languishing under theever dwindling supply and hence increasing cost of a barrel of petro-leum, while there appears no real solution to maintaining economicgrowth and providing sound environmental stewardship outside ofa bio-based economy. Thus, the biofuels industry that has emergedwithin the framework of a bio-based economy is expanding at anextraordinary rate [1–3]. The replacement of petroleum and fossil

fuels (natural gas, coal) by biofuels such as bioethanol and biodieselwill decrease the dependence on fossil fuels and minimize all con-comitant environmental impacts. More importantly, the abundant,inexpensive lignocellulosic biomass are more adopted by the bur-geoning bioethanol industry as feedstocks (mainly derived fromthe plant cell wall) [4]. Another benefit of lignocellulosics is that theyhave native characteristics (such as lignin itself) that make themresistant to microbial attack [5,6]. Yet, despite these attractive qual-ities, the development of the conversion process of biofuels from lig-nocellulosic biomass (such as agricultural residues, forestry wastesand thinning, waste paper, and energy crops) is still in the earlystages of research and advancement [7]. One of the chief reasonsaccounting for the relative lack of progress is the development offeasible, efficient, and economic pretreatment methods (chemical,

208 D.-y. Min et al. / Fuel 119 (2014) 207–213

biological, thermal, etc.) that facilitate the biofuel conversion pro-cess [8–11]. Technically, most of the recent studies have examinedthe recalcitrance of biomass biodegradation from chemical, physicaland biological perspectives. Technically, lignin is the major recalci-trant component of the plant cell wall. Lignin intransigence to pre-treatments for ultimate biofuel conversion platforms requiresincreasingly stronger chemical and biological severity in a mannerproportional to its amount.

Recently, it is proposed that lignin–carbohydrate complexes(LCCs) act to shield cellulose-like hemicellulose from enzymatic sac-charification, as well [12,13]. There are numerous evidences that lig-nin and polysaccharides are covalently bonded to form so calledlignin–carbohydrate complexes [10,14]. In spite of relatively lowamounts of LCCs in wood [15], they play a very important role as al-most all wood lignin is covalently bonded to polysaccharides, mainlyhemicelluloses [16]. Because LCCs hinders the enzymatic saccharifi-cation during the biorefinery process, therefore, understanding LCCsstructure was of great importance to address its effect on enzymaticsaccharification. Advantages and disadvantages of different analyti-cal methods on LCCs were reviewed recently [10,17]. Although infor-mation obtained from wet chemistry techniques and modelexperiments is very valuable, development of methods allowing di-rect observation of LCCs is of primary importance. However, directanalysis of LCCs with spectroscopic techniques is very difficult dueto strong signal overlapping in various spectra, including 1H and13C NMR spectroscopy, which have been widely used for the charac-terization of lignin and carbohydrates.

1H–13C Heteronuclear Single Quantum Coherence (HSQC) NMRcan overcome the shortage of 1D NMR such as the over-lappingsignals and allow direct detection of LCCs, however, it cannot pro-vide quantitative information. Thus, an approach to characterizeLCCs linkages using a combination of quantitative 13C NMR and1H–13C HSQC NMR techniques has been developed, recently. In thisresearch, the crude milled wood lignin (MWLc) was isolated fromtwo groups of transgenic black cottonwood (P. trichocapa) includ-ing the lignin down-regulated and the xylan down-regulated.Then, three major linkages such as phenyl glycoside, benzyl ether,and c-ester in LCCs were identified and quantified on a Bruker300 MHz NMR spectrometer. All enzymatic saccharifications ofthe samples had been previously accomplished and were used asthe basis of comparison for the elucidation of LCCs on the enzy-matic saccharification, eventually [18,19].

2. Materials and methods

2.1. Raw materials

Two groups of transgenic black cottonwood (western balsampoplar, California poplar) Populus trichocarpa including the xylandown-regulated and the lignin down-regulated lines were collectedfrom an in-house greenhouse (Forest Biotechnology Group, NorthCarolina State University). All of samples were approximately eigh-teen (18) months old. The gene expression of GT8D (glycosyltrans-ferase) was silenced for the xylan down-regulated line ptrGT8D-RNAi-3, ptrGT8D-RNAi-5 (8Di3 and 8Di5) and ptrGT8D-RNAi-2,ptrGT8D-RNAi-7 (8Di2 and 8Di7). Two lignin down-regulated trans-genic substrates (4CL1-1 and 4CL1-4) were obtained by down-regulating 4-coumarate: coenzyme A ligase (4CL). Another lignindown-regulated transgenic substrate (CH8-1-4) was obtained byover expressing Cald5H and down-regulating 4CL.

All air-dried and debarked samples were ground to pass 40mesh sieves using a Wiley mill (General Electric, USA). The fractionbetween 40 and 60 meshes was collected, and all adventitious con-taminants were extracted by a mixture of benzene and ethanol(2:1 v/v) for 8 h.

2.2. Sugar compositional analysis of substrates

The sugar composition of samples was determined according tothe TAPPI Standard Method T222 om-98. Briefly, an air dried sam-ple (about 0.1 g) was reacted with 1.5 ml H2SO4 (72% wt.) at 20 �Cfor 2 h with stirring every 15 min. Then the slurry was diluted with56 ml of de-ionized water (H2SO4 concentration: 3% wt.) and trans-ferred to a serum bottle. The slurry was subjected to autoclaving at122 �C for 1.5 h. The sugar content (arabinose, rhaminose, galact-ose, glucose, xylose and mannose) in the filtrate was quantifiedby Dionex-IC (Dionex IC-3000; Dionex, USA). The Dionex-IC systemwas equipped with a guard column (carboPac PA1 2 � 50 mm) andan ion-exchange column (carboPac PA1 2 � 250 mm) in tandem, apulsed amperometric detector with a gold electrode, and a SpectraAS 300 autosampler. Before injection, samples were filteredthrough 0.2 lm Nylon filters (Millipore) on which a volume of5 lL was ultimately loaded. The column was pre-equilibrated with250 mM NaOH and eluted with Milli-Q water at a flow rate of0.3 mL/min.

2.3. Cellulase-mediated saccharification

The batch cellulase-mediated saccharification procedure had al-ready been described elsewhere, but was briefly summarized here[19,20]. The saccharification was carried out in a 40 mL sampleflask containing 5% substrate (w/v) in 10 mL of 0.02 M acetate buf-fer (pH 4.8) supplemented with 40 lg/mL of tetracycline. Two en-zyme charges were used including 5 and 20 FPU/g based on theextracted sample. The slurry was incubated at 48 �C in a water bathshaker maintaining at 150 rpm for 48 h.

2.4. Isolation of crude milled mood lignin (MWLc)

The extractive-free sample (around 2 g) was subjected to 2 hmilling at 600 rpm using ZrO2 bowls and 17 ZrO2 balls in a plane-tary ball milling apparatus (Pulverisette 7, Fritsch, Germany). Thewood meal was extracted with 20 mL 1,4-Dioxane (96% v/v) for24 h at 25 �C. After centrifugation, the residue was carried out an-other extraction and the liquor was collected. The extraction wasperformed three times. The combined extraction liquor was fil-tered to remove fine particles. 1,4-Dioxane was evaporated withrotary-evaporator (under reduced pressure) at 35 �C. Several dropsof deionized water were added to the residue and evaporated againto remove traces of 1,4-Dioxane, a process that was repeated sev-eral times as necessary. The final product (MWLc) was dried in avacuum oven at 35 �C. The yield of MWLc was approximately40% based on lignin.

2.5. NMR spectroscopy

The crude milled wood lignin (about 40 mg) was dissolved in200 lL DMSO-d6. Then the solution was transfer to the Shigemimicrotube and characterized at 25 �C. Quantitative 13C spectrawere acquired on a Bruker AVANCE 300 MHz spectrometerequipped with a 5 mm BBO probe using an inverse gated protondecoupling sequence [21]. Chromium (III) acetylacetonate(0.01 M) was added to provide complete relaxation of all nuclei.The acquisition parameters were: 90� pulse width, a relaxation de-lay of 1.7 s, and an acquisition time of 1.2 s. A total of 20,000 scanswere collected. 1H–13C Heteronuclear Single Quantum Coherence(HSQC) NMR was recorded also on the same spectrometerequipped with a BBI probe. The acquisition parameters used wereas follows: 160 transients (scans per block) acquired using 1000data points in the F2 (1H) dimension with an acquisition time of151 ms and 256 data points in the F1 (13C) dimension with anacquisition time of 7.68 ms. The total running time was about

Table 2Composition of the crude milled wood lignin (Mean ± SD) in which the top halfrepresents the lignin down-regulated studies and the bottom half represents thexylan down-regulated studies.

Sample Glucan Xylan TC⁄ TL⁄ Balance

4CL1-1 16.0 ± 0.08 3.5 ± 0.02 19.4 ± 0.09 70.7 ± 0.56 90.1 ± 0.874CL1-4 16.3 ± 0.04 3.1 ± 0.01 19.4 ± 0.12 73.1 ± 0.44 92.6 ± 0.94CH8-1-4 16.1 ± 0.10 3.2 ± 0.00 19.3 ± 0.11 74.2 ± 0.23 93.5 ± 1.01

WT 16.4 ± 0.12 3.4 ± 0.04 19.7 ± 0.13 74.4 ± 0.54 94.1 ± 0.648Di3 11.1 ± 0.05 1.6 ± 0.01 12.7 ± 0.07 78.3 ± 0.45 91.0 ± 0.828Di5 10.9 ± 0.01 1.6 ± 0.00 12.5 ± 0.05 80.0 ± 0.56 92.5 ± 0.91WT1 15.5 ± 0.07 1.9 ± 0.02 17.4 ± 0.14 74.5 ± 0.63 91.9 ± 0.878Di2 10.1 ± 0.02 1.5 ± 0.01 11.5 ± 0.08 82.7 ± 0.77 94.2 ± 1.038Di7 9.4 ± 0.03 1.4 ± 0.03 10.8 ± 0.11 76.8 ± 0.62 87.6 ± 0.67WT2 14.0 ± 0.08 1.8 ± 0.05 15.9 ± 0.09 76.0 ± 0.49 91.9 ± 0.72

Note: Components is expressed as% (w/w) of the crude milled wood lignin. TC⁄:Total carbohydrates. TL⁄: Total lignin including acid insoluble lignin and acid sol-uble lignin.

D.-y. Min et al. / Fuel 119 (2014) 207–213 209

20 h. A coupling constant 1J C–H of 147 Hz was used. The 2D dataset was processed with 1000 and 91,000 data points using Qsinefunction in both dimensions.

3. Results and discussion

3.1. Polymer composition of the extractives-free substrates

The polymer composition of the transgenic samples was sum-marized in Table 1. From the upper half of the table, comparedto the wild type (WT), the transgenic substrates (4CL1-1, 4CL1-4,CH8-1-4) not surprisingly had a relatively lower level of lignin con-tent in the lignin down-regulated group. Lignin content of the sub-strates was of the following order: WT, CH8-1-4, 4CL1-4, 4CL1-1,whereas the carbohydrate content of the samples were not appre-ciably different. As for the xylan down-regulated samples (lowerhalf of Table 1), there were two sub-groups. The first sub-grouptransgenics (8Di3, 8Di5) had a little higher xylan and carbohydratecontent, yet lower lignin content compared to the second sub-group (8Di2, 8Di7). In all, the transgenic substrates had about30% less xylan compared to the wild type; however, they alsohad an approximately higher (12–25%) level of lignin.

3.2. Polymer composition of the crude milled wood lignin

The composition of the crude milled wood lignin (MWLc) wasdetermined and demonstrated in Table 2. Lignin was the maincomponent that displayed a range of 70–83% in MWLc. And, thesignificant amount of carbohydrate (11–20%) was demonstratedin MWLc. The lignin down-regulated group (top half) had a highercarbohydrate content (both glucan and xylan), compared to the xy-lan down-regulated group. Even after the purification, a smallamount of carbohydrate (about 2–3%, data not shown) remainedin the purified milled wood lignin that became a high priority forthe studies at this point. This remaining sugar after lignin purifica-tion almost certainly indicated the existence of covalent bonds be-tween lignin and carbohydrate (the lignin–carbohydratecomplexes or LCCs).

3.3. Cellulase-mediated saccharification

The effect of LCCs on enzymatic saccharification was easily elu-cidated by a correlation between LCCs identification and quantifica-tion and the sugar recovery. Two different generations of enzymeswere supplied from Novozymes North America, Inc. Both of Cellic

�

CTec2 and Cellic�

CTec are a blend of aggressive cellulases, highlevel of ß-glucosidase and hemicellulase. However, Cellic

�CTec2

has the following advantages: higher conversion yield; more

Table 1Composition of extractives-free substrates (Mean ± SD) in which the top halfrepresents the lignin down-regulated studies and the bottom half represents thexylan down-regulated studies.

Sample Glucan Xylan TC⁄ TL⁄ Balance

4CL1-1 47.1 ± 0.24 16.6 ± 0.08 66.6 ± 0.51 15.0 ± 0.05 82.6 ± 0.614CL1-4 47.2 ± 0.27 17.4 ± 0.12 69.1 ± 0.36 16.7 ± 0.10 85.8 ± 0.46CH8-1-4 46.6 ± 0.21 16.5 ± 0.15 68.7 ± 0.45 19.3 ± 0.12 87.1 ± 0.35

WT 45.2 ± 0.32 16.3 ± 0.11 65.9 ± 0.42 21.4 ± 0.09 87.2 ± 0.528Di3 45.1 ± 0.25 11.4 ± 0.08 62.2 ± 0.42 23.1 ± 0.07 85.3 ± 0.538Di5 45.1 ± 0.37 11.7 ± 0.02 62.2 ± 0.38 24.5 ± 0.01 86.7 ± 0.76WT1 45.4 ± 0.12 16.0 ± 0.02 64.6 ± 0.10 20.8 ± 0.16 86.4 ± 0.258Di2 40.3 ± 0.22 10.6 ± 0.03 56.7 ± 0.29 30.0 ± 0.50 86.7 ± 0.618Di7 42.0 ± 0.57 9.5 ± 0.07 57.8 ± 0.71 28.2 ± 0.21 86.6 ± 0.43WT2 41.8 ± 0.26 15.6 ± 0.09 61.8 ± 0.35 23.9 ± 0.15 85.7 ± 0.54

Note: Composition is expressed as% (w/w) of original extracted free substrate. TC⁄:Total carbohydrates including arabinan, rhamnan, galactan, glucan, xylan andmannan. TL⁄: Total lignin including acid insoluble lignin and acid soluble lignin.

effective at high solids concentrations; more inhibitor-tolerant,compatible with multiple feedstocks and high concentration andmore stable. Thus, the Cellic

�CTec2 can save up to 50% lower

enzyme dosage. 5 FPU/g of Cellic�

CTec2 was applied on the lignindown-regulated substrates, whereas 20 FPU/g of Cellic

�CTec was

added to the xylan down-regulated substrates.

3.4. Isolation of the crude milled wood lignin

Several methods had already been discussed and applied to iso-late lignin from wood [21,22]. So far, ball milled wood lignin hasbeen thought to be the most representative of native lignin.Although ball-milling is known to induce minimal bond cleavage,methods that involve ball-milling are the basis of several cell wallcharacterization procedures that are well established and acceptedby the scientific community. The bond cleavage is not sufficientlyintense to significantly perturb the native structure of the lignin.In terms of the technique, the intensity of milling had a largerinfluence on the particle size than prolonging the milling timewhile the cell corners and middle lamella were the most resistantlayers to milling and the S1 and S2 layers were clearly fibrillatedearly in the milling process. Some studies suggested that most ifnot all extractable cell wall components were obtained from thesesecondary cell wall layers, and not from the whole cell wall. How-ever, the isolated lignin involving ball-milling was still applied forthe structural characterization of lignin and LCCs. Three major link-ages between carbohydrates and lignin including c-ester, benzylether, and phenyl glycoside could be identified and quantified fromthe crude milled wood lignin (MWLc). In contrast, the purifiedmilled wood lignin (MWLp) and the cellulolytic enzyme lignin(CEL) were technically used to characterize the lignin structure be-cause of little interference from carbohydrates (less than 3% carbo-hydrates; data not shown). In this study, the milling speed wasfixed to 600 rpm over a milling time that was adjusted to provideMWLc with a yield of approximately 40% of lignin.

3.5. Identification and quantification of the lignin–carbohydratecomplexes

The milled wood lignin was characterized by the combinationof 13C and 1H–13C HSQC NMR. The major representative bonds rep-resenting lignin–carbohydrate complexes and structures of ligninunits were demonstrated in Fig. 1.

The proposed formulae were applied to quantify the varyingamounts of identified LCCs moieties in the samples [23]. In orderto convert the relative value obtained from 1H–13C HSQC NMR intothe absolute value, the integrations of three clusters atdC103–96 rpm, dC90–78 ppm and dC65–58 ppm in 13C spectra

Fig. 2. 13C NMR spectra of 8Di3 MWLc.

210 D.-y. Min et al. / Fuel 119 (2014) 207–213

(Fig. 2) were used as the internal references, assigned as phenylglycoside, benzyl ether and c-esters, respectively. The value inthe 13C spectra assigned to the resonance of the aromatic carbons(chemical shift from 160–103 ppm in Fig. 2) was set as 612 (includ-ing the a-carbonyl). Thus, the number of specific LCCs moieties per100 aromatic rings could be calculated. The contours that rangedfrom dC90–78 ppm/ dH5.7–3.0 ppm, dC103–96 ppm/dH5.5–3.8 ppmand dC65–58 ppm/dH5.0–2.5 ppm in the 2D spectrum were the to-tal resonance of the corresponding clusters used as internal refer-ences in 13C spectra. The chemical contours of benzyl ether, phenylglycoside and c-ester of LCCs moieties were shown in Fig. 3. Thequantitative amount of benzyl ether linkages in LCCs was acquiredfrom the signal of CH-a in the structure (B). Benzyl ether LCCsstructures were subdivided into two parts: (a) the signal atdC80–81/dH4.5–4.7 ppm was assigned to the B1-linkage betweenthe a-position of lignin and primary OH groups of carbohydrates(at C-6 of glucose, galactose, and mannose, and C-5 of arabinose)(b) another signal at dC80–81/dH4.9–5.1 ppm was assigned toB2-linkages between the a-position of lignin and secondary OHgroups of carbohydrates, mainly of lignin-xylan type [10,24,25].The signal of B1 overlapped with the signal of spirodienone lignin

Fig. 1. Typical lignin units and lignin–carbohydrate complexes: S: syringe unit; G: guC: c-ester; D: Spirodienone; E: b-O-4/a-OH c-acetylated; F:phenylcoumaran; I: resinol.

moieties (F) at approximately dC81.2/dH5.1 ppm. However, theamount of spirodienone structures was quantified from the signalof CH-b at dC79–80/dH4.0–4.1 ppm [26,27]. The amount of signals

aiucyl unit; H: para-hydroxyl benzyl unit; A: phenyl glycoside; B: benzyl ether;

MethoxylMethoxyl

(a)

Benzyl ether

Phenyl glycoside

Phenyl glycoside

(b) Fig. 3. 1H–13C HSQC NMR spectra of 8Di3 MWLc: (a) full scale; (b) enlarged scale.

Table 3Quantification of linkages of LCCs in MWLc.

Units 4CL1 4CL4 CH8 WT 8Di3 8Di5 WT1 8Di2 8Di7 WT2

BE 0.5 0.5 0.8 0.9 1.0 1.0 1.3 0.8 0.7 1.1PhGly 1.5 1.3 2.0 2.1 1.0 1.5 1.3 0.6 1.1 1.4c-Ester 2.4 2.8 2.9 3.1 2.8 3.1 4.2 3.5 3.4 4.1LCCs 4.4 4.7 5.6 6.1 4.9 5.6 6.8 4.8 5.2 6.6

Note: Values based on 100C9, LCCs is the total of the three different linkages of the lignin carbohydrate complex.

D.-y. Min et al. / Fuel 119 (2014) 207–213 211

of CH2 c-ester of LCCs was quantified in two parts: chemical shiftat dC62–63/dH4.05–4.15 ppm and dC63–64/dH4.15–4.2 ppm,respectively [23]. The signals of c-ester (C) overlapped the signalsof c-acylated lignin moieties (E), although it was presented in woo-dy lignin as a small amount [28–30]. Little occurrence of p-coum-arates and ferulates in the transgenic sample was demonstrated ataromatic region in 1H–13C HSQC NMR. Phenyl glycoside (A) gave agroup of signals of carbohydrates C1 at dC99–104/dH4.8–5.2 ppm.The values of LCCs (4.4/100C9–6.8/100C9) in samples investigatedwere summarized in Table 3.

Firstly, it was an advantage to more fully elucidate the influenceof xylan content on LCCs by examining the xylan down-regulatedsamples because these samples varied in xylan cotent with thesimilar lignin content. As a result, it was demonstrated that a de-crease of xylan induced a reducing amount of LCCs in the samples(Fig. 4(a)). For example, the level of LCCs increased from 5.2/100C9

to 6.6/100C9 with the increase of xylan from 9.5% to 15.6%. Sec-ondly, the influence of lignin on LCCs was elucidated by the lignindown-regulated samples directly. It also was shown that theamount of LCCs increased with an increase of lignin in the samples(Fig. 4(b)). For instance, the amount of LCCs increased from 4.4/100C9 to 6.1/100C9 while lignin increased from 15.0% to 21.4%. Inaddition, it was demonstrated that the level of LCCs correlatedwith lignin content more linearly than xylan content. Generally,

(a) y = 0.296x + 1.9556R² = 0.8604

LC

C/1

00C

9

% xylan content

Fig. 4. (a) Correlation between xylan content and lignin carbohydrate complexes

a reducing content of lignin/xylan indicated a deceasing level ofLCCs in the samples. Again, it was verified in this study that thecovalent bonds between carbohydrate and lignin were resultingin lignin–carbohydrate complexes in the plant cell wall.

3.6. Effect of lignin–carbohydrate complexes on cellulase-mediatedsaccharification

The relationship between the cellulase-mediated saccharifica-tion with respect to the quantity of the lignin–carbohydrate com-plexes was elucidated in Fig. 5. It was demonstrated that anincreasing level of LCCs hindered the enzymatic saccharificationmore severely. For example, with an increase of LCCs from 4.4/100C9 to 6.0/100C9, the sugar recovery decreased from 40% to28% in the lignin down-regulated samples. Thus, LCCs was impli-cated for an additional recalcitrance to biodegradation of lignocel-lulosic biomass besides the lignin and hemicelluloses. Twoexplanations were proposed: first of all, lignin–carbohydrate com-plexes (LCCs) played the similar role as hemicellulose by protectingthe cell wall from enzyme attack through shielding effect [31].Thus, a reducing level of LCCs therefore was allowing an improvingaccessibility to cellulose because there were inherently biggerpores which the enzymes were able to penetrate (the size of cellu-lase is about 4–13 nm). Second of all, the morphological and/or sol-

y = 0.2802x + 0.1144R² = 0.9849

LC

C/1

00C

9

%Lignin content

(b)

, (b) correlation between lignin content and lignin carbohydrate complexes.

(a) y = -5.57x + 53.776 R² = 0.5873

10

15

20

25

30

35

4.5 5.0 5.5 6.0 6.5 7.0

% S

ugar

rec

over

y

LCC/100C9

y = -6.5218x + 65.568R² = 0.7325

20

25

30

35

40

45

4.0 4.5 5.0 5.5 6.0 6.5

%Su

gar

reco

nver

y

LCC/100C9

(b)

Fig. 5. Relationship between the amount of LCC and the sugar recovery: (a) the down-regulated xylan samples; (b) the down-regulated lignin samples.

212 D.-y. Min et al. / Fuel 119 (2014) 207–213

ubility characteristics of LCCs network was another significant fac-tor affecting the synergystic adsorption between substrate and en-zymes. Technically, three major fragments such as xylan-lignincomplexes, glucomannan-lignin complexes and glucan-lignin com-plexes were generated from LCCs with the enzymatic saccharifica-tion or other treatments [32]. As a result, which part of the matrix(carbohydrate, the hydrophilic part or lignin, the hydrophobic part)was exposing to the enzymes or media and consequently affectedthe adsorption between the enzymes and substrates, significantly.For instance, if the hydrophilic part (carbohydrate) was exposed tothe enzymes, then an improving sugar recovery could be acquiredfrom the enzymatic saccharification. Otherwise, a reducing sugarrecovery was obtained because more enzymes were adsorbedand denatured by lignin (the hydrophobic part).

4. Conclusion

The milled wood lignin was isolated from both the lignin down-regulated and the xylan down-regulated P. Tricocarpa. Then, thelevels of three major linkages (benzyl ether, c-ester and phenylglycoside) representing LCCs in the samples were identified andquantified by the combination of 13C and 1H–13C HSQC NMR. Theresults indicated: (1) an increasing content of lignin/xylan indi-cated an improving level of LCCs in the sample; (2).a shielding ef-fect of LCCs on cellulose which was embedded in the plant cellmatrix was accounting for an additional recalcitrance of lignocellu-losic biomass; (3) the properties of LCCs such as hydrophobic orhydrophilic also exhibited an effect on the synergetic interactionbetween enzymes and substrates. A significant amount of enzymeswas adsorbed and denatured once the hydrophobic lignin part ofLCCs exposed to enzymes that hindered the enzymatic saccharifi-cation. Otherwise, the exposure of the hydrophilic part to enzymescould facilitate the enzymatic saccharification.

Acknowledgements

Then authors are grateful to the Southeastern Sun Grant Centerof USA for the financial support of this study and to NovezymesNorth America, Inc. for providing the enzymes used in this study.The transgenic P. trichocarpa plants were produced under supportsfrom the NSF Plant Genome Research Program (DBI-0922391) andthe Office of Science (BER, DE-SC0006691), US Department of En-ergy to VLC.

References

[1] Kheshgi HS, Prince RC, Marland G. The potential of biomass fuels in the contextof global climate change: focus on transportation fuels. Annu. Rev. EnergyEnviron 2000;25:199–244.

[2] Pu YQ, Zhang DC, Singh PM, Ragauskas AJ. The new forestry biofuels sector.Biofuels, Bioprod. Biorefin 2008;2:58–73.

[3] Zheng Y, Pan ZL, Zhang RH. Overview of biomass pretreatment for cellulosicethanol production. Int. J. Agric. Biol. Eng 2009;2:51–68.

[4] Lionetti V, Francoccia F, Ferraria S, Volpi C, Bellincampia D, Gallettia R, et al.Engineering the cell wall by reducing de-methyl-esterified homogalacturonanimproves saccharification of plant tissues for bioconversion. Proc Natl Acad SciU.S.A 2010;107:616–21.

[5] Himmel ME, Ding SY, Johnson DK, Adney WS, Nimlos MR, Brady JW, et al.Biomass recalcitrance: engineering plants and enzymes for biofuelsproduction. Science 2007;315:805–7.

[6] Sarkar P, Bosneaga E, Auer M. Plant cell walls throughout evolution: towards amolecular understanding of their principles. J Exp Bot 2009;60:3615–35.

[7] Stocker M. Biofuel and biomass-to-liquid fuels in the biorefinery: catalyticconversion of lignocellulosic biomass using porus materials. Angew Chem, IntEd 2008;47:2–14.

[8] Alizadeh H, Farzaneh T, Thomas G, Bruce D. Pretreatment of switchgrass byammonia fiber explosion (AFEX). Appl Biochem Biotechnol 2005;124:1133–41.

[9] Berlin A, Balakshin M, Gilkes N, Kadla J, Maximenko V, Kubo S, et al. Inhibitionof cellulase, xylanase and beta-glucosidase activities by softwood ligninpreparations. J Biotechnol 2006;125:198–209.

[10] Koshijima T, Watanabe T. Association between lignin and carbohydrates inwood and other plant tissues. Berlin: Springer; 2003.

[11] Sanchez G, Pilcher L, Roslander C, Modig T, Galbe M, Liden G. Dilute-acidhydrolysis for fermentation of the Bolivian straw material Paja Brava.Bioresour Technol 2004;93:249–56.

[12] Wyman CE. Handbook on bioethanol: production and utilization. WashingtonDC (USA): Taylor & Francis; 1996.

[13] Sassner P, Mårtensson CG, Galbe M, Zacchi G. Steam pretreatment of Salix withand without SO2 impregnation for production of bioethanol. Appl BiochemBiotechnol 2005;121:1101–17.

[14] N D, Wegener G. Wood: chemistry ultrastructure reactions. Berlin: Walter deGruyrer; 1984.

[15] Obst J. Frequency and alkali resistance of lignin–carbohydrate bonds in wood.Tappi 1982;65:109–12.

[16] Lawoko M, Henriksson G, Gellerstedt G. Structural differences between thelignin carbohydrate complexes in wood and in chemical pulps.Biomacromolecules 2005;6:3467–73.

[17] Balakshin MY, Capanema EA, Chang H-M. A fraction of MWL with highconcentration of lignin–carbohydrate linkages: isolation and analysis with 2DNMR spectroscopic techniques. Holzforschung 2007;61:1–7.

[18] Taherzadeh MJ. Ethanol from lignocellulose: physiological effects of inhibitorsand fermentation strategies. Ph.D. thesis in Biotechnology, Chemical ReactionEngineering, Chalmers University of Technology;1999.

[19] Min DY, Li QZ, Jameel H, Chiang V, Chang H-m. Enzymatic Saccharification oflow lignin transgenic cottonwood by different pretreatment methods. 2011;Proceedings of the 16th ISWFPC.

[20] Min DY, Li QZ, Jameel H, Chiang V, Chang H-m. Comparison of pretreatmentprotocols for cellulase-mediated saccharification of wood derived fromtransgenic low-xylan lines of cottonwood (P. trichocarpa). Biomass Bioenerg2011;35:3514–21.

[21] Capanema EA, Balakshin MY, Kadla JF. A comprehensive approach forquantitative lignin characterization by NMR spectroscopy. J Agric. FoodChem 2004;52:1850–60.

[22] Karimi K, Kheradmandiniaa S, Taherzadeh MJ. Conversion of rice straw tosugars by diluteacid hydrolysis. Biomass Bioenerg 2006;30:247–53.

[23] Balakshin MY, Capanema E, Gracz H, Chang H-m, Jameel H. Quantification oflignin–carbohydrate linkages with high-resolution NMR spectroscopy. Planta2011;6:1097–110.

[24] Toikka M, Brunow G. Lignin–carbohydrate model compounds. Reactivity ofmethyl 3-O-(a-L-arabinofuranosyl)-b-D-xylopyranoside and methyl b-D-xylopyranoside towards a b-O-4 quinone methide. J. Chem. Soc. Perkin Trans1999;1:1877–83.

[25] Tokimatsu T, Umezawa T, Shimada M. Synthesis of four diastereomericlignin carbohydrate complexes (LCCs) model compounds composed of a b-O-4 lignin model linked to methyl-D-glucose. Holzforschung 1996;50:156–60.

D.-y. Min et al. / Fuel 119 (2014) 207–213 213

[26] Capanema EA, Balakshin MY, Kadla JF. Comprehensive NMR studies onhardwood MWL. J Agric Food Chem 2005;53:9639–49.

[27] Zhang LM, Gellerstedt G, Lu FC, Ralph J. NMR studies on the occurrence ofspirodienone structures in lignins. J Wood Chem Technol 2006;26:65–79.

[28] Martínez Ferrer, Ángel Tomás, Rencoret Jorge, Marques Gisela, GutiérrezSuárez Ana, Ibarra David, et al. Monolignol acylation and lignin structure insome nonwoody plants: a 2D NMR study. Phytochemical 2008;69:2831–43.

[29] Ralph J, Lundquist K, Brunow G, Lu F, Kim H, Schatz PF, et al. Lignins: naturalpolymers from oxidative coupling of 4-hydroxyphenylpropanoids. PhytochemRev 2004;3:29–60.

[30] Yelle DJ, Ralph J, Frihart CR. Characterization of nonderivatized plant cell wallsusing high-resolution solution-state NMR spectroscopy. Magn Res Chem2008;46:508–17.

[31] Adani F, Papa G, Schievano A, Cardinale G, D’Imporzano G, Tambone F.Nanoscale structure of the cell wall protecting cellulose from enzyme attackenviron. Sci Technol 2011;45:1107–13.

[32] Martin L, Rickard B, Fredrik B, Gunnar H, Goran G. Changes in the lignin–carbohydrate complex in softwood kraft pulp during kraft and oxygendelignification. Holzforschung 2004;58:603–10.