Elucidation of lignin and polysaccharide linkages in wheat straw by 2H/13C isotopic tracer

Abstract

To elucidate chemical linkages between lignin and polysaccharides, the aqueous mixed solutions of coniferin-[α-13C], syringin-[α-13C], D-glucose-[6-2H], and phenylalanine ammonia-lyase inhibitor were injected into a living wheat stalk. Internode tissues with high abundance of 2H-13C were collected. The milled wood lignin, lignin-carbohydrate complex (LCC), and residual LCC (R-LCC) with enrichment of 2H-13C were isolated. The 13C and 2H abundances showed that the lignin and polysaccharides of internode tissues were labeled by 13C and 2H, respectively. Analysis with carbon-13 nuclear magnetic resonance (13C-NMR) showed that ketal and benzyl ether bonds were formed between α-C of lignin and carbohydrates. The R-LCC and LCC were further treated with enzymes to obtain enzymatic degraded R-LCC (ED-R-LCC) and enzymatic degraded LCC (ED-LCC). 13C-NMR spectra of ED-LCC showed that the α-C of lignin side chain was combined with 6-C of carbohydrates by ether, ester, and ketal linkages. 1H-NMR differential spectra of ED-LCCs revealed an LC linkage of benzyl ether bond. Glucan-lignin (En-R-GL) and xylan-lignin (En-R-XL) complexes were separated from ED-R-LCC by ionic liquid. A part of lignin α-C was linked to cellulose 6-C by benzyl ether and α-ketal linkages. 13C-NMR spectra of En-R-XL showed there were α-benzyl ether and α-ketal bonds between lignin and xylan.

INTRODUCTION

The cell wall of lignocellulosic biomass is mainly composed of cellulose, hemicellulose, and lignin. Lignin-carbohydrate complexes (LCC) have chemical bonds between carbohydrates and lignin moieties, and these can seriously inhibit the separation of cellulose from lignin and hemicellulose (Björkman 1956). In the process of biorefinery and pulping, cellulose and lignin in plant raw materials are difficult to separate efficiently, which results in huge resource consumption and energy waste (Tribot et al. 2019).

LCC linkage plays a crucial role in wood structure, since all lignin moieties in softwoods and 47 to 66% of lignin fragments in hardwoods (Henriksson et al. 2007) are bound to carbohydrates, mainly to hemicellulose (Lawoko et al. 2005). Lignin and xylose in softwood are mainly linked by ether bonds, while lignin and mannose are mainly linked by phenyl glycosidic bond. Xyloses in hardwood LCC are mainly linked with lignin by benzyl ether bonds (Koshijima and Watanabe 2003), while lignin structures are linked to 30% glucuronic acid through the benzyl ester bond (Takahahsi et al. 1988). LCC from grass is structurally different from that in woods due to the incorporation of hydroxycinnamates into the cell wall (You et al. 2015). It is well established that ferulic acid is ester-linked to carbohydrates and ether-linked to lignin in the cell wall of grass, forming a “lignin–ferulate–polysaccharide” (LFP) complex (Buranov and Mazza 2008). The LCC in wheat straw were particularly enriched with xylan and exhibited narrow polydispersity (Yao et al. 2015). Although previous work has studied many wood LCC and wheat LCC, the chemical linkage between cellulose and lignin in wheat LCC is still worthy of in-depth study, which may help develop the appropriate process to break the lignin-carbohydrate bond for efficient and selective separation of cellulose from biomass.

Many studies have shown that the chemical linkages of LCC belong to the following types: benzyl ether bond, benzyl ester bond, ketal bond, and phenyl glycosidic bond, etc. (Xie et al. 2000). The quinone methide intermediates in lignin biosynthesis couples with carbohydrate to form benzyl ether and benzyl ester bonds. The hydroxyl groups of alcohol and phenol on lignin moieties can be easily combined with the hydroxyl groups of glycosides on carbohydrates to form phenyl glycosidic bonds. Acidification (Eriksson et al. 1980), alkaline hydrolysis (Takahashi and Koshijima 1988), and other degradative methods can be used to study the structure of LCC. However, during its treatment process, it is inevitable to cause damage to the structure of lignin and carbohydrates. Therefore, non-destructive methods are required to better understand LCC structure without cleavage of lignin-carbohydrate (LC) bonds, and thus to extract lignocelluloses effectively and selectively (Tarasov et al. 2018).

Isotope labeling technique has a wide application in exploring the structure of LCC. Xie et al. (1991, 1993, 1994a,b) synthesized the lignin precursors of coniferin-[α-¹³C], coniferin-[β-¹³C], and coniferin-[γ-¹³C] and successively carried out selective ¹³C enrichments of ginkgo and rice straw lignin side chains (Cα, Cβ, and Cγ). The ¹³C abundance of Cα of newly-formed xylem of ginkgo wood was 3.5 times more than that of natural abundance. Milled wood lignin (MWL) was prepared from the xylem of ginkgo, and the ¹³C-NMR spectrum of lignin was analyzed. The results showed that the structure contained α-carbonyl, α-aldehyde, Cγ-carbonyl, Cγ-carboxyl, methylene, and phenylcoumaran. The connection between lignin and carbohydrates was also found. Xiang et al. (2013) studied the cellulose precursor, i.e., uridine diphosphate glucose-[6-¹³C], which was synthesized and injected into a living ginkgo tree together with a lignin inhibitor L-α-aminooxy-β-phenylpropionic acid (AOPP), and an exogenous lignin precursor. The ¹³C-enriched LCC was isolated from the newly-formed xylem of ginkgo shoots. Then, it was degraded by cellulase and hemicellulase. Thus, the lignin-rich fractions, which were called enzyme-degraded LCC (ED-LCC), were obtained. Through determining their carbon-13 nuclear magnetic resonance (¹³C NMR) spectra, the bond formation between C6 position of glucose units in cellulose and carbons of lignin side chain were confirmed to be benzyl ether linkage. However, Xiang’s study was able to prove the existence of LC bonds from the cellulose side only.

Enzymatic hydrolysis cuts the long chains of carbohydrates without changing the structure of the LC bonds due to its high selectivity and mild conditions (Karlsson et al. 2001). At present, researchers have proposed a general classification process for extracting LCC from lignocellulosic biomass, i.e., the non-extractable biomass is first ground by ball milling and dissolved in DMSO/Tetra-Butyl-Ammonium Hydroxide (TBAH) solution, and three LCC fractions are further extracted as follows: a glucan-enriched fraction (glucan-lignin, GL), a mannan-enriched fraction (GML), and a xylan-enriched fraction (xylan-lignin, XL) (Du et al. 2013).

The inhibitor of phenylalanine ammonia lyase (PAL), i.e., (carboxymethyl methoxy amine hydrochloride, AOA) was applied. The AOA can inhibit the transformation of D-glucose-[6-²H₂] to lignin as shown in Fig. 1. The coniferin-[α-¹³C] was degraded into the corresponding coniferyl alcohol by β-glucosidase in cells and deposited in the biosynthesis of lignin macromolecules.

Fig. 1. Inhibition of the transformation of D-glucose-[6,6-2H2] to lignin and metabolism of coniferin-[α-13C]

This research used ¹³C-enriched lignin and ²H-enriched carbohydrate precursors, respectively. The internode tissues with high abundances of ²H-¹³C from wheat straw were collected. As compared with previous studies, this research applied the Björkman’s method (1957) and enzymatic hydrolysis to remove the fragments rich in hemicellulose, thereby enriching the fragments with LC bonds between cellulose and lignin. Then, the linkages between lignin and cellulose were elucidated clearly by ¹³C-NMR and ¹H-NMR differential spectral analysis.

EXPERIMENTAL

Materials

Wheat Emai 596 was provided by Hubei Academy of Agricultural Sciences (Wuhan, China). Sodium acetate-1-¹³C and D-glucose- [6-²H₂] were purchased from Sigma-Aldrich (Saint Louis, MO, USA). The AOA was purchased from McLin (Shanghai, China). All other chemicals used were of analytical grade.

Fig. 2. Chemical structures of coniferin-[α-13C] (I) and syringin-[α-13C] (II)

Synthesis of Isotope Labeled Lignin Precursors

According to the method of Xie et al. (1991, 1993, 1994a, and 1994b), Sodium acetate-1-¹³C was used as the starting material in the synthesis of coniferin-[α-¹³C] and syringin-[α-¹³C].

Administration of Lignin and Carbohydrate Precursors

Two batches of growing wheat Emai 596 were selected as the experimental plant because of its wide internode cavities. Before wheat heading in early March, the aqueous solutions of D-glucose-[6-²H₂] (5 mg/mL) mixed with coniferin-[α-¹³C] (5 mg/mL) or syringin-[α-¹³C] (5 mg/mL) were injected into the internodal cavities of the first, second, and third sections from the root to the top of two batches of wheat, as shown in Table 1,, Fig. 3 (a) and Fig. 3 (b). To inhibit the conversion of deuterium-labeled glucose to lignin, a solution of AOA (0.35 mg/mL), which was the inhibitor of phenylalanine ammonia-lyase, was also injected. After the injection, the plant was allowed to grow for 30 days (Imai and Terashima 1992).

Fig. 3. Administration of Lignin and Carbohydrate Precursors: (a) Flow chart showing the injections of lignin precursors, carbohydrate precursor, and AOA into wheat internode cavities and the subsequent processing; (b) Stalk internodes from root to top

Table 1. Concentration of the Solution Injected to Wheat Internode

Preparation of Extractive-free Wheat Straw Mill

Wheat stalks fully absorbed the two types of lignin precursors and D-glucose-[6-²H₂] samples and were harvested after their complete maturity. The internode tissues of group C, group G, and group S were collected. The wheat straw samples were then fully air-dried and ground using a Wiley mill to pass through 60-mesh screen. The milled straw was extracted by ethanol-benzene mixture (1/2, v/v) and hot water, and then air-dried.

Determination of ¹³C and Deuterium Abundances

The ¹³C and ²H isotope abundances in 1.0 mg wheat tissue samples were determined by an elemental analyzer – isotope ratio mass spectrometer (EA-IRMS) equipped with FLASH2000 elemental analyzer (Thermo Fisher Scientific GmbH, Dreieich, Germany) and Delta V isotope mass spectrometer (Thermo Fisher Scientific GmbH, Dreieich, Germany). Then, the value of ¹³Cα/¹²Cα was calculated using Eqs.1 to 3, while D6/H6 was calculated by Eqs. 4 to 5:

¹³C/¹²C = 1.11802% × (1 + δ¹³C ÷ 1000) (1)

(¹³Cα/¹²Cα)_G= 1.07252% + (¹³C/¹²C－1.07252%) ÷ 0.2073 × 10 (2)

(¹³Cα/¹²Cα)_S= 1.07252% + (¹³C/¹²C－1.07252%) ÷ 0.2073 × 11 (3)

In Eqs. 1 to 3, ¹³C/¹²C is the ratio of ¹³C and ¹²C abundances in the sample; δ¹³C is the relative ¹³C isotope abundance value of the sample (Vienna Pee Dee Belemnite, VPDB‰); 1.11802% is the ¹³C isotope abundance of the standard sample of VPDB; ¹³Cα/¹²Cα is the ¹³C and ¹²C isotopic ratios of Cα position in the lignin structural units of the sample; 1.07252% is the natural abundance of ¹³C isotope of wheat straw; 0.2073 is the lignin content in wheat straw; 10 is the ratio of total ¹³C content to α-¹³C content in guaiacyl propane structural unit, 11 is the ratio of total ¹³C content to α-¹³C content in syringyl propane structural unit. The relative abundances of D are shown in Eqs. 4 and 5:

D/H = 0.015575% × (1 + δD ÷ 1000) (4)

D6/H6 = (D/H – 0.01317%) ÷ 0.445 × 5 + 0.01317% (5)

In Eqs. 4 and 5: D/H is the ratio of D and H abundances in the sample; δD is the relative D isotope abundance value of the sample (Vienna Standard Mean Ocean Water, VSMOW‰); 0.015575% is the D isotope abundance of the standard sample VSMOW, D6/H6 is the abundance ratio of D to H on glucose 6-C position in the sample; 0.01317% is the natural abundance of D isotope of wheat straw; 0.445 is the cellulose content in wheat straw; 5 is the ratio of total D content in glucose unit to D content on 6-C.

Determination of CP/MAS ¹³C-NMR Spectrum of Milled Wheat Straw

An Avance III HD 600 MHz wide-cavity solid-state NMR spectrometer (Bruker, Billerica, MA, USA) was used. The samples were continuously scanned at 150.6 MHz to obtain ¹³C-NMR by conventional cross polarization (CP) and magic angle spinning (MAS) methods. Experimental conditions were as follows: temperature 25 °C, pulse delay 3 s, acquisition time of 0.05 s, pulse width 75 kHz, and 5000 scans.

Preparation of LCC and R-LCC

As shown in Fig. 4，the extractives in milled wheat straw (20 g ground sample with a Wiley mill and sieved using 100-mesh) were removed. The extracted sample was dried in vacuo with P₂O₅ for 2 weeks. The extractive-free milled straw was further ground by a water-cooling vibration ball mill for 72 h, and it was extracted three times with aqueous dioxane (96/4, v/v).

The dioxane solution was subjected to rotary evaporation and freeze-drying to obtain crude MWL. A filtrate was obtained by dissolving in acetic acid aqueous (9/1, v/v) and followed by filtration. The filtrate was added dropwise to deionized water and followed by centrifugation. The obtained precipitate was dissolved in dichloroethane-ethanol (2/1, v/v) and centrifugated. The filtrate was added dropwise to absolute ether and then centrifugated. The precipitate was washed twice with absolute ether, and then it was washed once with petroleum ether. After vacuum drying, MWL 0.8 g was obtained with a yield of 4 %.

The Residue Ι was extracted with acetic acid-water (1/1, v/v) for three times, and the resulting precipitate was R-LCC 15.6 g with a yield of 78%.

The obtained acetic acid-water solution was freeze-dried by rotary evaporation, extracted with DMF, and added dropwise to dichloroethane-ethanol (2/1, v/v). After centrifugation, the precipitate was washed once with dichloroethane-ethanol (2/1, v/v) and washed three times with absolute ether. After vacuum drying, the product was dissolved in acetic acid-water (1/1, v/v), and added dropwise to acetone. The precipitate was obtained by centrifugation and washed once with acetone-acetic acid (96/1, v/v), three times with absolute ether, and once with petroleum ether. After vacuum drying, 2.66 g LCC was obtained with a yield of 13.3%.

Fig. 4. Preparation processes of MWL, LCC and R-LCC

Enzymatic Hydrolysis of LCC and R-LCC

Cellulase (Onozuka RS, Yakult Pharmaceutical Industry Co., Nishinomiya, Japan, 16,000 units/g), hemicellulase (from Aspergillus niger, sigma, ≥ 1500 units/g), and xylanase (from Thermomyces canuginosus, sigma, ≥ 2500 units/g) were completely dissolved in 120 mL 0.5 M acetic acid/sodium acetate buffer (pH=4.5). The enzyme solution was filtered by a G4 glass filter and stored at 5 °C.

Approximately 2 g of LCC was added to 20 mL of the above enzyme solution and 80 mL of the acetic acid/sodium acetate buffer, and a little toluene was then dropped as a protective agent. The mixture was shaken in a water bath at 50 ℃ for 48 h, and centrifuged. After washing with deionized water 4 times, and then freeze -dried, ED-LCC 0.2 g was obtained with a yield of 10% (Lin et al. 1992).

The R-LCC (10 g) was added to 100 mL of the above enzyme solution and 400 mL of the acetic acid/sodium acetate buffer. Then the above enzymatic hydrolysis steps were repeated. After centrifugation and freeze-drying, 1.2 g En-R-LCC was obtained with a yield of 12%.

Classification of En-R-LCC Components by Ionic Liquid

As shown in Fig. 5, En-R-LCC (1.0 g) was added to the ionic liquid composed of DMSO/TBAH (5 mL/5 mL). The mixture was stirred continuously for 12 h until the sample was completely dissolved. The solution was then dropped into 100 mL deionized water with stirring and centrifuged to obtain a precipitate and supernatant. The precipitate was fully washed with deionized water to neutral. After freeze drying, the glucan-lignin fraction (En-R-GL) 150 mg was obtained with a yield of 15%. The supernatant was neutralized with dilute HCl, dialyzed (1000 Da) and freeze-dried to obtain 250 mg of xylan-lignin fraction (En-R-XL) with a yield of 25 %.

Fig. 5. Classification of wheat straw En-R-LCC with ionic liquid

Acetylation of Samples

A total of 150 mg of En-R-LCC or En-R-GL was added to the mixture of DMSO/N-methylimidazole (3 mL/1 mL). The mixture was stirred for 5 h until the sample was completely dissolved. Then, 0.6 mL of acetic anhydride was added and allowed to react for 1 h. The solution was dropped into 50 mL deionized water with stirring. After complete precipitation, the mixture was centrifuged, washed 4 times with water, and freeze dried to give acetylated products (Ac-En-R-LCC and Ac-En-R-GL).

Determination of Solution NMR Spectrum

A total of 90 mg of sample was completely dissolved in 0.6 mL DMSO-d₆ using a φ5-mm NMR tube. The NMR spectrum was recorded by a Bruker Avance III 600 MHz spectrometer (Bruker BioSpin GmbH, Rheinstetten, Germany) at 294 K and scanned at 150 MHz for ¹³C-NMR. Analysis conditions were as follows: pulse delay 2 s, acquisition time 0.9 s, and 20,000 scans. Spectrum with 2 Hz line broadening was used for processing.

¹H-NMR spectrum recording parameters included scanning at 600 MHz, pulse delay of 1.0 s, acquisition time of 2.7 s, and 500 scans with 12k acquisition data points.

RESULTS AND DISCUSSION

Analysis of ¹³C/²H Abundances in Wheat Internode Tissues

As shown in Tables 2 and 3, the abundances of ¹³C and D (²H) in the labeling groups were noticeably increased. The ¹³Cα/¹²Cα ratio of milled wheat straw sample labeled with ¹³C and D was about three times that of natural isotope ratio. The D6/H6 ratio was about eight times that of natural isotope ratio, indicating that exogenous coniferin-[α-¹³C], syringin-[α-¹³C], and D-glucose-[6-²H₂] were effectively absorbed and metabolized during the growth of wheat internode tissues. This resulted in lignin in the cell wall of wheat straw that was labeled with ¹³C, while the polysaccharides were labeled with deuterium.

Table 2. 13C Abundance of Milled Wheat Straw Administered with Coniferin-[α-13C], Syringin-[α-13C], and D-Glucose-[6,6-2H2]

Table 3. 2H Abundance of Milled Wheat Straw Administered with Coniferin-[α-13C], Syringin-[α-13C], and D-Glucose-[6,6-D2]

Solid-State CP/MAS ¹³C-NMR Analysis of Wheat Straw Mill

As shown in Fig. 6, the CP/MAS ¹³C-NMR spectra of milled wheat straw administered with syringin-[α-¹³C], wheat straw mill administered with coniferin-[α-¹³C], and natural wheat straw were analyzed. The results showed that there was no remarkable difference in the aromatic regions (δ 110 ppm to 160 ppm).

Fig. 6. CP/MAS 13C-NMR spectra of milled wheat straw: a: administered with syringin-[α-13C], b: administered with coniferin-[α-13C], c: control group, d: difference spectrum obtained by subtracting spectrum (c) from spectrum (a), and e: difference spectrum obtained by subtracting spectrum (c) from spectrum (b)

To distinguish the signals of lignin side chain carbons labeled with ¹³C isotope from that in unlabeled samples, the CP/MAS ¹³C-NMR spectra of milled wheat straw were differentiated. In the differential spectra as shown in Fig. 6, all the other carbon signals were eliminated except for the ¹³C enhanced signal. Therefore, the signals could be assigned more accurately to different types of α-C according to the chemical shifts. The peak of -OCH₃ at δ 56.0 ppm (No. 11) with stable content was used as the internal reference, and the tentative assignment of each signal is shown in Table 4. An enhanced signal at δ 100.5 to 110.2 ppm (No.1’) is the signal of α-C with ketal linkages to carbohydrates (Xie et al. 2000). The peak at δ 93.1 to 80.7 ppm (No. 2’) primarily arises from α-C in β-5, β-β, and benzyl ether linkage to carbohydrates (Xiang et al. 2014). The strong signal at δ 80.1 to 67.9 ppm (No. 3’) are assigned to the α-C in the β-O-4 structure and α-C with ester linkage to carbohydrates (Xie et al. 1991), while the signal at δ 67.9 to 58.0 ppm (No. 4’) arises from α-C in the β-1 structure between lignin moieties (Hafrén et al. 2002). Therefore, the connection between lignin structural units and LC bonds can be elucidated.

Fig. 7. Linkages between lignin structural units and LCC

Fig. 8. 13C-NMR spectra of ED-LCCs prepared from wheat straw

Legend: a: Sample administered with syringin-[α-¹³C] and D-glucose-[6,6-²H₂].; b: sample administered with coniferin-[α-¹³C] and D-glucose-[6,6-²H₂].; and c: control sample

Table 4. Tentative Assignment of Signals from CP/MAS 13C-NMR Difference Spectrum of Wheat Straw Mill Administered with Syringin-[α-13C] and Coniferin-[α-13C], and D-Glucose-[6,6-2H2]

Analyses of ¹³C-NMR Spectra of ED-LCCs

The ¹³C-NMR spectra of wheat straw ED-LCCs are shown in Fig. 8, and the tentative assignments of the signals are shown in Table 5.

Table 5. Assignments of 13C-NMR Signals of ED-LCCs from Wheat Straw

The -OCH₃ peak with stable content at δ 55.8 ppm (No. 25) was used as the internal reference. An enhanced signal at δ 191.5 ppm (No. 1) is from α-CHO. After α-¹³C labeling, an enhancement signal can be observed at δ 101.8 ppm (No. 17) and assigned to α-C of lignin linked to the carbohydrates by ketal bond (Zhang et al. 2021). The enhanced signal at δ 87.2 ppm (No. 18) is from α-C in phenylcoumaran. The signal at δ 85.2 ppm (No. 20) is from α-C in pinoresinol. After labeling with α-¹³C, a weak signal at δ 82.5 ppm (No. 21) is enhanced and assigned to the α-C of the lignin side chain linked to the carbohydrates via benzyl ether bond (Gu et al. 2001). An enhanced signal at δ 73.9 ppm (No. 22) is from α-C connected with carbohydrates through ester bond (Xie et al. 2020). The resonance signal at δ 72.8 ppm (No. 23) is the α-C in the structure of β-O-4 linkage. A peak at δ 60.1 ppm (No. 25) comes from the γ-C in the β-O-4 substructure (Xiang et al. 2014).

Analyses of ¹H-NMR Differential Spectra of ED-LCCs

To further understand the connection between lignin side chain α-C and glucan C6 from polysaccharide side, ¹H-NMR differential spectra of the ED-LCCs administered with D-glucose-[6-²H₂] together with coniferin-[α-¹³C] or syringin-[α-¹³C] were analyzed. A proton on the lignin aromatic ring (δ 7.2 ppm) was used as the internal standard. The ¹H-NMR differential spectrum of ED-LCC was obtained by subtracting the ¹H-NMR spectrum of the D-labeled group from that of control group, as shown in Fig. 9. The tentative signal assignment is shown in Table 6.

Fig. 9 1H-NMR differential spectra of EDLCCs

Legend: a: Control group subtracted wheat straw labeled with syringin-[α-¹³C] and D-glucose-[6-²H₂]; b: Control group minus wheat straw labeled with coniferin-[α-¹³C] and D-glucose-[6-²H₂]

A pair of peaks appeared at δ 4.65 ppm (No. 2) and 4.53 ppm (No. 3), which arise from a pair of H signals on the 6-C of the polysaccharide connected with the lignin side chain α-C by benzyl ether bond. The characteristic signals at δ 4.22 ppm (No. 4) and 3.97 ppm (No. 5) were assigned to H on 6-C of carbohydrates connected with ester bond to lignin side chain (Nishida et al. 1984). A signal at δ 3.63 ppm arises from carbohydrates 6-H,H’ without linkage with lignin, which indicates that some carbohydrates were linked with lignin not through C-6 position.

Table 6. Assignments of signals of 1H-NMR differential spectra of ED-LCCs from Wheat Straw

Analyses of ¹³C-NMR Spectra of Ac-En-R-GL Fractions

The ¹³C-NMR spectra of Ac-En-R-GL fractions of wheat straw are shown in Fig. 10. A stable -OCH₃ peak at δ 55.5 ppm (No. 20) is used as an internal reference. The enhanced signals at δ 194.5 ppm (No. 1) and 191.5 ppm (No. 2) are primarily assigned to α-CHO. The signal at δ 170.8 ppm (No. 3) arises from -COO- of ferulic acid structure. The signal at δ 134.2 ppm (No. 8) is assigned to C1 on the aromatic nucleus. The signal at δ 128.1 ppm (No. 9) is assigned to α-C in lignin side chain -CαH =CH- structure. The peak at δ 105.8 ppm is significantly enhanced after α-¹³C labelling due to the formation of α-ketal bond between lignin side chain α-C and cellulose.

Fig. 10. 13C-NMR spectra of Ac-En-R-GL prepared from En-R-LCC fractions of wheat straw

Legend: a: Sample administered with syringin-[α-¹³C] and D-glucose-[6-²H₂]; b: sample administered with coniferin-[α-¹³C] and D-glucose-[6-²H₂]; and c: control sample

The enhanced signal at δ 87.6 ppm (No. 14) is assigned to α-C of β-5. The signal at δ 84.0 ppm (No. 15) could be assigned to enhanced α-C of β-β. Through ¹³C labeling, the signal at δ 82.2 ppm (No. 16) is enhanced and assigned to α-C linked to cellulose with benzyl ether bond. The peak at δ 72.8 ppm (No. 17) can be assigned to enhanced α-C of β-O-4.

Analysis of ¹H-NMR Differential Spectrum of Ac-En-R-GL Fractions

Figure 11 shows the ¹H-NMR differential spectra of Ac-En-R-GLs. The peaks at δ 4.71 ppm (No. 2) and 4.59 ppm (No. 3) arise from a pair of hydrogen on cellulose6-C connected with lignin side chain α-C. The H of glycosyl 6-C has an benzyl ester bond signals in the 4.20 ppm~4.00 ppm region of the spectra, and ether bond signals in the 4.65 ppm~4.40 ppm region (Nishida et al. 1984). The 2D-NMR analysis of birch LCC shows that there was an ether bond between lignin side chain α-C and glucan 6-C (Balakshin et al. 2011). Combined with the ¹³C-NMR analysis of Ac-En-R-GL, it is considered that the signals at 4.71 ppm (No.2) and 4.59 ppm (No.3) are from cellulose 6-C linked to lignin side chains α-C, which was connected by benzyl ether bond. Because the acetylation of the sample will lead to the formation of ester bonds, the resonance signals at δ 4.28 ppm (No. 4) and 4.03 ppm (No. 5) in the spectra cannot prove the ester bond connection between cellulose 6-C and lignin side chain α-C (Hikichi et al. 1995).

Fig. 11. 1H-NMR differential spectra of Ac-En-R-GLs prepared from En-R-LCC fractions of wheat straw

Legend: a: Control group subtracting wheat straw labeled with syringin-[α-¹³C] and D-glucose-[6-²H₂]; b: Control group minus wheat straw labeled with coniferin-[α-¹³C] and D-glucose-[6-²H₂]

Analyses of ¹³C-NMR Spectra of En-R-XL Fractions

As shown in Fig. 12, in the ¹³C-NMR spectra of xylan-lignin fractions, the peak due to -OCH₃ at δ 55.9 ppm (No. 24) with stable content was used as the internal reference. The enhanced signal at δ 191.2 ppm (No. 1) was primarily assigned to α-CHO. After labeling, a signal at δ 130.1ppm (No. 10) was enhanced and assigned to α-C of coniferyl alcohol subunits. The signal intensities of the ¹³C-enriched samples and the control group did not exhibit remarkable difference in the range of δ 110 to 120 ppm (No. 12 to No. 14), which primarily arise from aromatic carbons (C2, C5, and C6). The signal at δ 101.9 ppm (No. 16) was enhanced after ¹³C labeling, which was assigned to the α-C of the lignin side chain linked to the xylan subunit by ketal bond (Xie et al. 2000). It was found that there were two enhanced signals No.17 and No.18 at δ 86 ppm and 88 ppm in the ¹³C labeled samples and assigned to α-C in phenylcoumaran and pinoresinol, indicating the low contents of these two structures. The enhanced peak at δ 81.5 ppm (No. 20) after ¹³C labeling arises from the α-C of lignin linked to xylan by an ether bond. The signal intensity shows that the content of this kind of ether bond in wheat straw En-R-XL was low (Gu et al. 2001). The enhanced signal at δ 71.5 to 72.3 ppm (No. 21) was from the α-C of β-O-4. From the peak intensity, it can be concluded that β-O-4 is the main connection between lignin moieties of En-R-XL (Besombes et al. 2003). The signal at δ 62.7 ppm (No. 22) comes from α-C of β-1, while that at δ 60.2 ppm (No. 23) is γ-C of β-O-4.

Fig. 12. 13C-NMR Spectra of En-R-XL fractions prepared from R-LCC fractions of wheat straw

Legend: a: Sample administered with syringin-[α-¹³C] and D-glucose-[6-²H₂]; b: sample administered with coniferin-[α-¹³C] and D-glucose-[6-²H₂]; and c: control sample

Fig. 13. Linkages between lignin structural units, lignin-glucan, and lignin-xylan

Neish (1958) confirmed that the first step of xylan biosynthesis in plant cell wall was enzymatic oxidation of D-glucose to form glucuronic acid. Then, a series of metabolisms were conducted to form xylan. In this process, the ²H isotope on C6 of D-glucose-[6-²H₂] in plants will be eliminated due to enzymatic hydrolysis. Therefore, no difference is found in the ¹H-NMR differential spectra of xylan-lignin fractions.

In summary, there is an LCC structure linked by benzyl ether bond and ketal bond between lignin and xylan in the En-R-XL fraction of wheat straw. Figure 13 shows the linkage between lignin in wheat straw and the LC bond between lignin, xylan, and glucan.

CONCLUSIONS

1. Through the analyses of the abundances of ¹³C and D atoms in the labeling and control groups, it was found that the ¹³Cα/¹²Cα ratio of wheat straw labeled with ²H-¹³C was about three times that of natural isotope ratio, while the D6/H6 ratio was about eight times that of natural isotope ratio, indicating that exogenous coniferin-[α-¹³C], syringin-[α-¹³C], and D-glucose-[6-²H₂] were effectively absorbed and metabolized during the growth of wheat internode tissues. These results indicate that lignin in the cell wall of wheat straw was labeled with ¹³C, while the polysaccharides were labeled with deuterium isotopes.

2. Analysis of CP/MAS ¹³C-NMR spectra and their differential spectra of extractive-free milled wheat straw showed that enrichment of stable isotope (²H-¹³C) had no obvious interference on the normal lignification process of wheat cell wall. The lignin structural units are mainly composed of β-O-4, β-5, β-β, and β-1 subunits, and also contain a small amount of coniferyl alcohols substructure. The α-C of lignin phenylpropane side chain is connected with carbohydrates by ketal bond and benzyl ether linkages.

3. To further understand the structure of LC linkages in wheat straw, LCC and R-LCC were isolated and treated by enzymatic hydrolysis. It was found that the side chain α-C of lignin in ¹³C/²H-enriched ED-LCC of wheat straw was combined with 6-C of carbohydrates by benzyl ether, ester, and ketal bonds. It was further confirmed that the lignin was primarily composed of β-O-4, β-β, β-5, and β-1 subunits, and also contained a small amount of coniferyl alcohol subunits.

4. The ¹³C-NMR spectra of En-R-GL showed that a part of lignin α-C was connected with cellulose by benzyl ether and α-ketal linkages. The ¹H-NMR differential spectrum of Ac-En-R-GL fraction showed that a part of lignin side chain α-C was bound with cellulose 6-C via benzyl ether bond. The ¹³C-NMR spectra of En-R-XL showed that there was a small amount of lignin α-C was connected with xylose by α-benzyl ether and α-ketal linkages. The lignin structural units were primarily composed of β-O-4, and also included a small amount of β-β, β-5, and β-1 linkages.

ACKNOWLEDGMENTS

The authors are grateful for the support of the National Natural Science Foundation of China (Grant No. 21878070).

References

Balakshin, M., Capanema, E., Gracz, H., Chang, H., and Jameel, H. (2011). “Quantification of lignin–carbohydrate linkages with high-resolution NMR spectroscopy,” Planta 233, 1097-1110. DOI: 10.1007/s00425-011-1359-2

Besombes, S., Robert, D., Utille, J. P., Taravel, F. R., and Mazeau, K. (2003). “Molecular modeling of syringyl and p-hydroxyphenyl β-O-4 dimers. Comparative study of the computed and experimental conformational properties of lignin β-O-4 model compounds,” Journal of Agricultural and Food Chemistry 51(1), 34-42. DOI: 10.1021/jf0206668

Björkman, A. (1956). “Studies on finely divided wood. Part 1. Extraction of lignin with neutral solvents,” Svensk Papperstidn 59(13), 477-485.

Björkman, A. (1957). “Lignin and lignin-carbohydrate complexes,” Industrial & Engineering Chemistry 49(9), 1395-1398. DOI: 10.1021/ie50573a040

Buranov, A. U., and Mazza, G. (2008). “Lignin in straw of herbaceous crops,” Industrial Crops and Products 28(3), 237-259. DOI: 10.1016/j.indcrop.2008.03.008

Du, X., Gellerstedt, G., and Li, J. (2013). “Universal fractionation of lignin–carbohydrate complexes (LCCs) from lignocellulosic biomass: an example using spruce wood,” The Plant Journal 74: 328-338. DOI: 10.1111/tpj.12124

Eriksson, Ö., Goring, D. A. I., and Lindgren, B.O. (1980). “Structural studies on the chemical bonds between lignins and carbohydrates in spruce wood,” Wood Science and Technology 14, 267-279. DOI: 10.1007/BF00383454

Gu, R., and Xie, Y. (2001). “Formation mechanism and chemical structure of lignin-carbohydrate complex (I) -synthesis of DHP in the presence of holocellulose,” Paper Science and Technology 20(5), 1-6. DOI: 10.3969/j.issn.1671-4571.2001.05.001

Hafrén, J., Westermark, U., Lennholm, H., and Terashima, N. (2002). “Formation of 13C-enriched cell-wall DHP using isolated soft xylem from Picea abies,” Holzforschung 56(6), 585-591. DOI: 10.1515/HF.2002.089

Henriksson, G., Lawoko, M., Martin, M. E. E., and Gellerstedt, G. (2007). “Lignin-carbohydrate network in wood and pulps: A determinant for reactivity,” Holzforschung 61(6), 668-674. DOI: 10.1515/HF.2007.097

Hikichi, K., Kakuta, Y., and Katoh, T. (1995). “1H NMR study on substituent distribution of cellulose diacetate,” Polymer Journal 27, 659-663. DOI: 10.1295/polymj.27.659

Imai, T., and Terashima, N. (1992). “Determination of the distribution and reaction of polysaccharides in wood cell walls by the isotope tracer technique. IV. Selective radio-labeling of xylan in magnolia (Magnolia kobus) and visualization of its distribution in differentiating xylem by microautoradiography,” Mokuzai Gakkaishi 38, 693-699.

Karlsson, O., Pettersson, B., and Westermark, U. (2001). “The use of cellulases and hemicellulases to study lignin-cellulose as well as lignin-hemicellulose bonds in kraft pulps,” Journal of Pulp and Paper Science 27(6), 196-201.

Koshijima, T., and Watanabe, T. (2003). “Association between lignin and carbohydrates in wood and other plant tissues,” Springer-Verlag, Berlin, Heidelberg. DOI: 10.1007/978-3-662-05191-7

Lawoko, M., Henriksson, G., and Gellerstedt, G. (2005). “Structural differences between the lignin-carbohydrate complexes present in wood and in chemical pulps,” Biomacromolecules 6(6), 3467-3473. DOI: 10.1021/bm058014q

Lin, S. Y., and Dence, C. W. (1992). “Methods in lignin chemistry,” in: Springer Series in Wood Science, S. Y. Lin, and C. W. Dence (eds.), Springer-Verlag. Berlin, Heidelderg. DOI: 10.1007/978-3-642-74065-7_3

Neish, A. C. (1958). “The biosynthesis of cell wall carbohydrates: IV. Further studies on cellulose and xylan in wheat,” Canadian Journal of Biochemistry and Physiology 36(1), 187-193. DOI: 10.1139/y58-021

Nishida, Y., Ohrui, H., and Meguro, H. (1984). “1H-NMR studies of (6r)-and (6s)-deuterated d-hexoses: assignment of the preferred rotamers about C5-C6 bond of D-glucose and D-galactose derivatives in solutions,” Tetrahedron Letters 25(15), 1575-1578. DOI: 10.1016/S0040-4039(01)90014-0

Takahashi, N., and Koshijima, T. (1988). “Ester linkages between lignin and glucuronoxylan in a lignin-carbohydrate complex from beech (Fagus crenata) wood,” Wood Science and Technology 22, 231-241. DOI: 10.1007/BF00386018

Tarasov, D., Leitch, M., and Fatehi, P. (2018). “Lignin–carbohydrate complexes: Properties, applications, analyses, and methods of extraction: A review,” Biotechnology for Biofuels 11, article ID 269. DOI: 10.1186/s13068-018-1262-1

Tribot, A., Amer, G., Alio, M. A., Baynast, H. D., Delattre, C., Pons, A., Callois, J. M., Vial, C., Michaud, P., and Dussap, C. G. (2019). “Wood-lignin: Supply, extraction processes and use as bio-based material,” European Polymer Journal 112, 228-240. DOI: 10.1016/j.eurpolymj.2019.01.007

Xiang, S., Xie, Y., Yang, H., and Yao, L. (2013). “Cellulose ~ (13) C isotope tracer method to study the connection between cellulose and lignin,” Spectroscopy and Spectral Analysis 33(9), 2488-2491. DOI: 10.3964/j.issn.1000-0593(2013)09-2488-04

Xiang, S., Xie, Y., Yang, H., and Yao, L. (2014). “Analysis of the association between cellulose and lignin by carbon 13 tracer method with cellulose precursor,” Chemistry and Industry of Forest Products 34(1), 37-42. DOI: 10.3969/j.issn.0253 2417.2014.01.007

Xie, Y., and Terashima, N. (1991). “Selective carbon 13-enrichment of side chain carbons of ginkgo lignin traced by carbon 13 nuclear magnetic resonance,” Mokuzai Gakkaishi 37(10), 935-941.

Xie, Y., and Terashima, N. (1993). “Selective carbon 13-enrichment of side chain carbons of rice stalk lignin traced by carbon 13 nuclear magnetic resonance,” Mokuzai Gakkaishi 39(1), 91-97.

Xie, Y., and Terashima, N. (1994a). “Selective carbon 13-enrichment of side chain carbons of oleander lignin traced by carbon 13 nuclear magnetic resonance,” Mokuzai Gakkaishi 40(2), 191-198.

Xie, Y., Liu, Y., Jiang, C., Wu, H., and Bi, S. (2020). “The existence of cellulose and lignin chemical connections in ginkgo traced by 2H-13C dual isotopes,” BioResources 15(4), 9028-9044. DOI:10.15376/biores.15.4.9028-9044

Xie, Y., Robert, D, R., and Terashima, N. (1994b). “Selective carbon 13-enrichment of side chain carbons of ginkgo lignin traced by carbon 13 Nuclear Magnetic resonance,” Plant physiology and Biochemistry 32(2), 234-249.

Xie, Y., Yasuda, S., Wu, H., and Liu, H. (2000). “Analysis of the structure of lignin-carbohydrate complexes by the specific 13C tracer method,” Journal of Wood Science 46(2), 130-136. DOI: 10.1007/BF00777359

Yao, L., Yang, H., Xu, J., Zhang, N., Chen, Y., and Xie, Y. (2015). “Elucidation of the structure of lignin and lignin carbohydrate complex of Gramineae,” Cellulose Chemistry and Technology 49(3-4), 259-266.

You, T. T., Zhang, L. M., Zhou, S. K., and Xu, F. (2015). “Structural elucidation of lignin-carbohydrate complex (LCC) preparations and lignin from Arundo donax Linn,” Industrial Crops and Products 71, 65-74. DOI: 10.1016/j.indcrop.2015.03.070

Zhang, K., Liu, Y., Cui, S., and Xie, Y. (2021). “Elucidation of the structure of lignin–carbohydrate complexes in ginkgo CW-DHP by 13C-2H dual isotope tracer,” Molecules 26(19), 5740. DOI: 10.3390/molecules26195740