Investigation of interaction between cellulase and residual lignin fractions in acid-pretreated bamboo residues and their effect on enzymatic digestibility

Background: During dilute acid pretreatment, pseudo lignin and lignin form droplets which deposit on the surface of lignocellulose, and further inhibit its enzymatic hydrolysis. However, how this lignin interacts with cellulase enzymes and then affects enzymatic hydrolysis is still unknown. In this work, different fractions of surface lignin (SL) obtained from dilute acid pretreated bamboo residues (DAP-BR) were extracted by various organic reagents and the residual lignin in extracted DAP-BR was obtained by milled wood lignin (MWL) method. All the obtained lignin fractions from DAP-BR were used to investigate the interaction mechanism between lignin and cellulase using surface plasmon resonance (SPR) technology in order to understand how they affect enzymatic hydrolysis Results: Results showed that removing surface lignin significantly decrease the enzymatic hydrolysis of DAP-BR from 36.5% to 18.6%. The addition of MWL samples to Avicel decreased enzymatic hydrolysis of Avicel, while different SL samples showed a slight increase to its enzymatic digestibility. Due to the higher molecular weight and hydrophobicity of MWL samples versus the SL samples, stronger affinity for MWL (KD = 6.8-24.7 nM) was found versus that of SL (KD = 39.4-52.6 nM) by SPR analysis. The affinity constant of all tested lignin had good correlations (R 2 >0.6) with their effects on enzymatic digestibility of extracted DAP-BR and Avicel. Conclusions: This work reveals that the surface lignin on DAP-BR is necessary towards negative impact on the substrate’s digestibility.

conversion of residual bamboo is a chemical one-its lignin is particularly recalcitrant to the biorefinery processing schemes currently demonstrated upon other biomass resources. [7] Thus, it is imperative to develop tailored pretreatment techniques capable of overcoming bamboo lignin's recalcitrance and allow for efficient production of platform monosaccharides. [8] Many different kinds of pretreatments have been studied upon bamboo substrates to date. [9] Among these pretreatment methods, dilute acid pretreatment appears to be most logical because it has been successfully applied pretreatment to bamboo residues, with the added benefit of low process costs and ease of operation. [10] However, the enzymatic digestibility of bamboo residues could not reached the expected extent and usually lower than 40 %. [11] The explanations for the low enzymatic saccharification efficiency have been proposed: 1) lignin-like compounds (pseudo-lignin) is be formed and problematically deposited on the surface of fibers. [12][13] 2) lignin undergoes repolymerization reactions which also block available surface area, and 3) water-soluble phenolics produced by depolymerization of lignin serve as soluble enzyme inhibitors. [14] However, the reported work mainly focus on the macroscopical effects of these lignin on enzymatic hydrolysis of pretreated biomass, the microcosmic effects of these lignin on cellulase adsorption that can influence the enzymatic hydrolysis is still scarce.
The interaction between lignin and cellulase is one of the important factors influencing enzymatic digestibility of pretreated biomass. [15] Techniques to study this 5 interaction have been extensively explored. For instance, Langmuir adsorption isotherm was used to characterize enzyme affinity to lignin from the macro-scale. [16] Sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) was applied to investigate cellulase-lignin affinity, but the results obtained were not quantitative. [17] In recent years, Quartz crystal microbalance with dissipation (QCM-D) has been relied upon to investigate the kinetic adsorption behavior in real time. [18][19] Surface plasma resonance (SPR), a novel method for measuring interactions between various compounds, has also been applied. [20] Like QCM technology, SPR also achieves real-time monitoring and analysis of interactions between a molecule and solids, allowing for acquisition of highly detailed information such as the kinetics of association and dissociation. [21][22] To our knowledge, there is no current work where SPR technology was applied to investigate the interactions between lignin and cellulase over the course of an enzymatic hydrolysis process.
In order to be able to evaluate the interactions between lignin and cellulase, a representative lignin preparation must be extracted from pretreated biomass (bamboo residues in this case). There are many different methods for extract lignin from pretreated-materials, but most are based around extraction using organic solvent such as 1,4-dioxane, tetrahydrofuran, ethanol, acetone, and more. [23][24] Hansen solubility parameter (HSP) theory typically drives the decisions for lignin solvent selection, taking into consideration the inter-molecular forces between solvents and lignin (dispersion, 6 polarity and hydrogen bonding). [25] According to HSP theory, solvents with different solubility parameter have the ability to extract different lignin fractions from the same pretreated biomass substrate. Importantly, the lignin on the surface and cell wall of acid pretreated biomass, which is mostly likely to be engaging with cellulolytic enzymes, is quite homogeneous from different extractions. However, this material would certainly be the ideal lignin to serve as substrate for SPR analysis. Therefore, a variety of different solvents have been applied in this work to extract different lignin fractions in/on pretreated bamboo residues, in effort to evaluate the range of interactions that SPR can detect.
In this work, three different organic solvents (1,4-dioxane, ethanol, and tetrahydrofuran) with different HSP were used to extract different surface lignin fractions on DAP-BR. A classical lignin preparation (milled wood lignin, "MWL") was also extracted from the extracted residual solids, referred to as E-DAP-BR. Following extraction, enzymatic hydrolysis of DAP-BR, E-DAP-BR and mixtures of pure cellulose with different extracted lignin fractions were performed to evaluate the effects of different lignin fractions on enzymatic hydrolysis. The physicochemical properties of lignin (molecular weight and hydrophobicity) and DAP-BR (crystallinity indices, cellulose accessibility and hydrophobicity) were determined and used to probe for relationships with enzymatic digestibility. Furthermore, SPR technology was utilized to characterize the interaction between extracted lignin fractions and cellulase, with the 7 kinetic constants accounting for their interactions being calculated. It is our hope that this work will provide a more detailed look at the potential mechanisms hindering digestion of pretreated bamboo by enzymes.

Influence of organic reagent extraction on chemical composition of dilute acid pretreated bamboo residues
Recently, Hansen solubility parameters (HSP) theory have become a popular tool for screening different organic solvent systems able to efficiently separate lignin from biomaterial matrices. [25] Based on HSP theory, if a given RED value (Ra/R0) is < 1, the analyzed solvent will exhibit greater lignin affinity. Alternatively, when RED is >1, that means the chosen solvent system will behave as a poor lignin solvent. [26] Thus, in this work, three organic reagent systems (1,4-dioxane, ethanol, and tetrahydrofuran) were selected to extract different surface lignin fractions based on the difference in their RED values. The chemical compositions of extracted materials were determined and shown in Table 1. Table 1 shows that the glucan and xylan fractions remained in DAP-BR after lignin extraction processes, meaning tested system was indeed lignin-selective.
According to Table 1  a: DAP-BR:dilute acid pretreated bamboo residues. Dio-BR, Eth-BR, THF-BR: extracted by 1,4-dioxane extraction using concentration of 96% (v/v, water/1,4-dioxane), ethanol extraction using concentration of 95% (v/v, water/ethanal) and etrahydrofuran extraction using concentration of 100% (v/v); MWL: milled wood lignin, SL: surface lignin. b: Analyzed by dye absorption c: Analyzed by water contact angle To further investigate the reason why enzymatic digestibility of DAP-BR was decreased after being extracted using organic solvents, several physicochemical properties of DAP-BR and E-DAP-BR were evaluated to probe for stark changes (Table 2). It can be seen that the cellulose accessibility of DAP-BR increased from 218 mg/g to 236 mg/g, 236 mg/g and 251 mg/g for Dio-BR, Eth-BR, and THF-BR, respectively. A slightly correlated relationship (R 2 =0.68) between delignification and cellulose accessibility was observed (Fig. 1b). The increase noted for cellulose accessibility for cellulases is likely owed to the removal of lignin covered on the surface of cellulose. Generally, the improvement of cellulose accessibility tended to be the crucial factor to improve its enzymatic digestibility. [30] However, in this work, our results seem to indicate that cellulose accessibility was not an important factor for enzymatic hydrolysis under these specific details of sample workup. One possible hypothesis may be that after removing the surface lignin, more residual lignin is exposed on substrate's surface, foster a greater frequency of non-productive adsorption with cellulases. In support of this hypothesis, hydrophobicity analysis results showed that solvent extracted samples (Dio-BR, Eth-BR and THF-BR) had higher hydrophobicity (3.6, 3.1, and 3.0 mg/mL, respectively) compared to DAP-BR (0.7 mg/mL). This speculation is also supported by the linear fitting shown in Fig. 1c, where a negative correlation (R 2 =0.68) can be found for the enzymatic hydrolysis yield of extracted DAP-BR and their hydrophobicity. Huang et al, 2019 also reported that substrates with higher hydrophobicity values induce more of non-productive binding by cellulase, reducing enzymatic hydrolysis yield. [31] Cellulase is mainly consisted of Cellobiohydrolase I (CBH I or Cel7A), which is negatively charged during enzymatic hydrolysis. So, the higher negative zeta potential of lignin brought stronger electrostatic repulsion between lignin and substrates and reduced their non-productive binding. To further probe into our findings, zeta potentials of all samples were also analyzed ( Table 2). It can be seen that all extracted DAP-BR have a much lower zeta potential (absolute value) than DAP-BR. Higher zeta potential (absolute value) values for substrates represents higher surface charge and more hydrophilic substrates. [32][33] Thus, more hydrophilic substrates also undergo stronger electrostatic repulsion between lignin and cellulase, reducing non-productive binding between lignin and cellulase to improve enzymatic hydrolysis.
As shown in Fig. 2d, a negative correlation (R 2 =0.83) was linearly fitted between enzymatic hydrolysis yields for each sample and their respective zeta potential. Results in Table 2 also show that there were no obvious differences for crystallinity index (CrI) and crystallite size (Bhkl, Dhkl) of cellulose for extracted DAP-BR and DAP-BR. These results illustrate that organic solvent extraction did not damage or change the cellulose, therefore these factors do not affect the enzymatic digestibility.
Based on the aforementioned results, it seems that the main driver between the enzymatic digestibility decrease after organic solvent extraction was exposure of highly hydrophobic lignin on the surface. To further probe this effect, residual lignin in extracted DAP-BR samples were further isolated and characterized to understand if 13 it may play a more critical role on inhibition of enzymatic hydrolysis.

Characterization of lignin fractions
It is widely agreed that some physicochemical properties of lignin, such as molecular weight, hydrophobicity and surface charge, exert varying influence on enzymatic hydrolysis of cellulose. [34] Hence, certain physicochemical properties (molecular weight, contact angel, and zeta potential) of isolated SL samples on DAP-BR and MWL sample in E-DAP-BR were evaluated and shown in Table 2. From

Effects of surface lignin and milled wood lignin on enzymatic hydrolysis of Avicel
In order to further investigate the different effects of surface lignin and the residual lignin fractions on enzymatic digestibility of cellulose, all MWL and SL samples were dosed into enzymatic hydrolysis systems of Avicel (pure cellulose). Resultant enzymatic digestibility are shown in Fig. 2. It is observed that three SL fractions did 15 not inhibit the enzymatic hydrolysis at 20% addition (Fig. 2a). Unexpectedly, a slight increase to enzymatic hydrolysis yield for Avicel was actually achieved, from 77.0% (control) to 81.2% (Dio-SL), 79.7% (Eth-SL), and 77.1% (THF-SL) at 40% lignin addition (Fig. 2b). This phenomenon was similarly reported in the work of Wu et al, 2017. [38] In contrast, the enzymatic hydrolysis yield of Avicel were decreased from 77.0% (control) to 58.4%, 39.7%, 45.3%, and 39.0% with 40% addition of DAP-MWL, Dio-MWL, Eth-MWL and THF-MWL (respectively) (Fig. 2c). These As for the effects of lignin molecular weight, Zhao et al, (2020) recently reported that lignin with greater molecular weights have more non-productive adsorption with enzymes compared to those of lower molecular weights. [39] According to table 2, The Mw of SL (1185-1980 g/mol) was lower than these of MWL (9135-13200 g/mol), which indicated that SL have less non-productive adsorption with enzymes than MWL.
Hence, it can be speculated that the residual lignin in DAP-BR and E-DAP-BR had higher hydrophobicity and molecular weight might showed more non-productive absorption between lignin and cellulase by enhancing the hydrophobic interactions, which can be attributed to their inhibitions on enzymatic hydrolysis yield of Avicel.
To further investigate the how cellulases may be adsorbing to the different lignin fractions investigated in this work, Chrastil's approach (Eq. 1) was used to study the diffusion-limited kinetic model of enzymatic hydrolysis system of Avicel doped with our lignin fractions. [40] The equation is as follows: where P (g/L) and Pσ (g/L) are the products which diffuse at every considered time t and at equilibrium, respectively. A rate constant, k (g.L -1 h -1 ) proportional to the diffusion coefficient as is defined by Fick's law. [41] E0 (g/L) is the initial enzyme concentration, and n is the structural diffusion resistance constant. The resistance constant is dependent on the steric structures of studied system. Generally, the more negative effects takes place in enzymatic hydrolysis system, the constant n is lower.
Meanwhile, a lower k indicates that the substrate is more resistant to enzymatic hydrolysis, translating to decreased engagement between the enzyme and substrate. In this work, the model parameters for enzymatic hydrolysis of Avicel-lignin system were determined and shown in Table 3

Enzyme-lignin surface interaction determined by SPR technology
SPR technology has been widely applied to study the adsorption between proteins and biomacromolecules. To evaluate the interaction and affinity between enzymes After 440s, citric acid buffer was injected again, lignin and cellulase were disintegrated, resulting in the decrease of RU value. Specific enzyme kinetics adsorption parameters on lignin films were calculated and shown in Table 4. It can be observed that the values of RU increased when enzyme bonded to the lignin and decreased when the enzyme dissociated from lignin after addition of buffer solution into the instrument. These observations showed that the decrease of RU in SL-cellulase systems were more obvious than these in MWL-cellulase system during enzyme dissociation process, indicating the enzymes were much more easily dissociated from SL compared MWL.  Similar conclusions are also drawn in the work of Li et al, (2016). [42] Overall, these SPR results indicate that the method can be used as a good technology to investigate the interaction mechanism between lignin and enzymes.

Conclusions
In this work, removal of different surface lignin fractions on DAP-BR showed a reduction effect on the resultant solid's enzymatic digestibility due to the exposure of residual lignin with higher hydrophobicity and molecular weight. The residual lignins in extracted DAP-BR had higher diffusion resistance to enzymes in a cellulose-lignin system, while SL fractions did not show any substrate resistance. SPR results showed that residual lignin in E-DAP-BR with greater hydrophobicity had higher affinity with cellulase than that of SL on DAP-BR, which is linearly correlated to their negative effects on enzymatic hydrolysis.

Materials
Bamboo residues were sampled from a bamboo processing factory in Sichuan, China. The chemical contents of glucan, xylan, and lignin in bamboo residues were 40.1%, 22.0%, and 27.2%, respectively. Cellulase (Cellic CTec2) was provided by Novozymes NA, Franklinton, USA with a filter paper activity of 250.0 FPU/ml.

Dilute acid pretreatment for bamboo residues
For pretreatment, 1.0 kg bamboo residues were massed and then mixed with 6.0 L of 1% (w/v) dilute sulfuric acid inside of a 15 L reactor at 160 o C for 1 h. After pretreatment, the resultant liquor (prehydrolyzate) was collected via vacuum filtration.
Captured solids (DAP-BR) were washed using distilled water until the wash filtrate pH was 7.0. Some of the washed DAP-BR was kept at 4 o C for enzymatic hydrolysis, while another portion was air-dried before further lignin extraction.

Organic extraction for surface lignin from dilute acid pretreated bamboo residues
Air-dried DAP-BR was respectively extracted by a 1,4-dioxane extraction/water solution (96:4, v/v), a 95% ethanol aqueous solution, and a tetrahydrofuran (100%) to prepare three unique surface lignin preparations. All extraction processes were carried out in 2.0 L conical flasks with 10% solids suspension at 150 rpm for 24 h. After 24 h, the liquor was separated from residual solids by filter paper and fresh organic solution was again added to once-extracted solids. This process was totally repeated for three times. All of the extract liquids were combined and then subjected to evaporation via a rotary evaporator and a vacuum freeze dryer to obtain solids. Lignin solid preparations from solutions of 96% 1,4-dioxane extraction/water, 95% ethanol, and tetrahydrofuran were termed as Dio-SL, Eth-SL and THF-SL, respectively.

Extraction of milled wood lignin from DAP-BR and E-DAP-BR
The un-extracted/residual lignin in DAP-BR, Dio-BR, Eth-BR and THF-BR were next obtained using the milled wood lignin (MWL) preparation technique according to Bjorkman, (1954). Hydrophobicity analysis for all lignin samples was performed by water contact angle measurement using automatic contact angle meter (Attention Theta, Biolin Scienfic, Inc. Stockholm, Sweden). Each sample was tested in duplicate.
The zeta potentials of lignin samples were measured by a Zetasizer (Nano-ZS, Malvern Instruments Ltd, Worcestershire, UK) with laser Doppler. For analysis, 2.5 mg of lignin was blended with 5 mL of 0.05 M citrate buffer and dispersed using an disperser to get the homogeneous suspension. All samples were tested in triplicate.

Enzyme-lignin surface interaction studied by SPR
To prepare the lignin film needed as a SPR substrate, either SL or MWL samples were dissolved into neat DMSO at 0.5% solids loading (w/w). The prepared solutions were coated on a SPR gold sensor using a spin coater (KW-4A, Shanghai Daojing Instrument Plant, China) at 5000 rpm for 1 min. The coating process was repeated 3 times. Resultant films were vacuum dried at 40 °C for 4 h, and then soaked in the deionized water for 24 h. The deionized soak water was replaced every 2 h to ensure complete removal of DMSO. The soaked films were vacuum dried at 40 °C for 12 h.
Interaction between lignin films and enzyme was measured by a SPR device minutes. The recorded SPR curves were processed using SPR Navi™ Data Viewer software. Kinetics constants were determined using Scrubber (version 2.0) software.

Analytical methods
The composition of pretreated and organosolv-extracted bamboo residues were determined according to the procedure developed by the National Renewable Energy Laboratory (NREL). [47] Monosaccharide concentrations for enzymatic hydrolysate and compositional analysis acid hydrolysate were measured using a high performance liquid chromatography (HPLC) system equipped with an Aminex HPX-87H column