Cellulose Nanocrystals Gels to

Dec 2, 2011 - University of Reims Champagne-Ardenne, UMR614 Fractionnement des AgroRessources et Environnement, Reims, France. •S Supporting ...
0 downloads 0 Views 4MB Size

Article pubs.acs.org/Biomac

Characterization of Arabinoxylan/Cellulose Nanocrystals Gels to Investigate Fluorescent Probes Mobility in Bioinspired Models of Plant Secondary Cell Wall Gabriel Paes̈ *,†,‡ and Brigitte Chabbert†,‡ †

INRA, UMR614 Fractionnement des AgroRessources et Environnement, Reims, France University of Reims Champagne-Ardenne, UMR614 Fractionnement des AgroRessources et Environnement, Reims, France

S Supporting Information *

ABSTRACT: Biomass from lignocellulose (LC) is a highly complex network of cellulose, hemicellulose, and lignin, which is considered to be a sustainable source of fuels, chemicals and materials. To achieve an environmental friendly and efficient LC upgrading, a better understanding of the LC architecture is necessary. We have devised some LC bioinspired model systems, based on arabinoxylan gels, in which mobility of dextrans and BSA grafted with FITC has been studied by FRAP. Our results indicate that the probes diffusion is more influenced by their hydrodynamic radius than by the gel mesh size. The addition of some cellulose nanocrystals (CNCs) decreases polymer chain mobility and has low effect on the probes diffusion, suggesting that the gels are better organized in the presence of CNCs, as shown by rheological measurements and scanning electronic microscopy observations. This demonstrates that the FRAP analysis can be a powerful tool to screen the architecture of LC model systems.

INTRODUCTION Biomass from lignocellulose (LC) is potentially a renewable and sustainable source of fuels, chemicals, and other products that could replace their fossil-derived counterparts whose availability is limited, but controlled destructuration and degradation processes of lignocellulosic materials are still challenging because of their high complexity and recalcitrance.1 Producing efficient and cheap lignocelluloytic enzymes is considered as the bottleneck to achieve selective and sustainable destructuration processes, but enzymes diffusion, adhesion,2,3 and catalysis in complex substrates and often at solid−liquid interface are not fully understood, whereas enzyme accessibility to substrate is a central feature to consider4 and thus requires more investigation, in particular, at the molecular scale. Comprehensive overview of the architecture of plant cell walls has been greatly improved thanks to the large panel of immunocytochemical tools, mapping polymers in situ is now achieved mainly by using specific antibodies directed against some particular epitopes of the cell wall,5,6 but all chemical structures are far from being recognized. More recently, use of CBMs (carbohydrate binding modules) has gained attention because these small proteins produced by microorganisms naturally bind to various plant polymers,7 and their number is still increasing. They can be localized in cellulose or heteroxylans by measuring directly the fluorescence of a fluorochrome grafted onto the CBMs8−12 or by using antibodies (generally directed against the CBM His-tag).13,14 Interestingly, the use of mutagenically inactivated enzymes (endoxylanases in particular) © 2011 American Chemical Society

has prompted, in the last years, the probes being localized by atomic force microscopy,15 immunolabeling,16 or directly by measuring the fluorescence of a chimeric enzyme bearing a fluorochrome.17 These techniques are quite useful for mapping plant cell wall polymers, but because probes are directly soaked onto whole thin sections of plant cell walls, they do not bring valuable information on probes mobility. Indeed, few techniques exist for gaining insight into dynamics of some molecular probes inside LC, although structural entanglements and chemical interactions are known as limiting factors for probe mobility. Studying larger plant cell wall samples and probing dynamics inside is a great challenge but requires setting up new protocols, which is not straightforward. Besides, computational approaches can give complementary information on the molecular level but require that the structural model has been validated, which is the case for cellulose but not for LC, mainly because of hemicellulose/ lignin variety and complexity.18 That is why a relevant alternative is to design some model systems, which contain some elements isolated from plant cell walls of interest and then to describe the parameters that influence the behavior of some probes in these systems. In grasses (in particular cereals like wheat, maize, rice), secondary Received: October 20, 2011 Revised: November 25, 2011 Published: December 2, 2011 206

dx.doi.org/10.1021/bm201475a | Biomacromolecules 2012, 13, 206−214



Table 1. Properties of the Fluorescent Probes


molecular weight MW (Da)

hydrodynamic radius RH (nm)

degree of labeling (number of FITC molecules per probe molecule)


10 000 70 000 250 000

2.3 6.0 11.1

0.2 to 1.2 1.3 to 8.6 11.9

commercial commercial commercial

69 700




cell walls constitute at least 50% of the cell wall mass,19 so they represent the largest source of LC. They are composed of cellulose microfibrils (35−45% in dry weight) embedded in an amorphous matrix phase consisting predominantly of lignins (20%) and hemicelluloses (40−50%).19 These hemicelluloses are essentially arabinoxylans (AXs), which are decorated with glucuronic and ferulic acids (FAs) or acetyl groups and covalently bonded to each other or to lignin to make lignin carbohydrate complexes (LCCs). Because of the importance of the secondary cell wall of grass in the biorefinery concept, we have decided to base our model system on feruloylated arabinoxylan (FAX) gels. FAs are of primary chemical importance in the organization of unlignified cell walls, and they possibly act as nucleation sites when lignification begins.20 In our case, FAs can react between each other to make covalent bonds under the action of a laccase, leading to the FAX reticulation and the formation of gels.21 FAX gels have been extensively studied in food applications for their water absorption and their transport properties of drugs and proteins.21−24 Besides, the gels can be complexified by adding some cellulose nanocrystals (CNCs) before triggering gelation. These gels can therefore be considered to be a good template of grass secondary cell wall. For investigation of the mobility of probes inside FAX gels, fluorescence microscopy has proved to be a valuable and practical method. Therefore, we have embedded selected fluorescent probes into our model systems and used fluorescence recovery after photobleaching (FRAP) technique and confocal laser scanning microscopy (CLSM) to calculate mobility of selected probes. FRAP measurements have already been done for example in gelatin25 and poly(ethylene glycol)-based gels26 and to study surface mobility and interactions of cellulases27,28 or xylanases.29 Here we propose first to characterize bioinspired lignocellulosic gels by rheological measurements and scanning electronic microscopy (SEM), then to study the mobility of the fluorescent probes to define the most critical parameters that limit their diffusion.

reference supplier’s data supplier’s data supplier’s data, RH estimated according to the formula: RH = 0.488 × MW0.437 49 36

by measuring the absorbances at 280 and 495 nm using the following formula ε prot A 495· × 195 MW

molar ratio =


A280 − 0.35 × A 495


where MWFITC is FITC molecular weight (389 Da), εprot is molar extinction coefficient of BSA (43824 L·mol−1·cm−1), and A280 and A495 are absorbances at 280 and 495 nm, respectively. To get a final molar ratio comprised between 1 and 2, we prepared different mixes. Finally, BSA-FITC was obtained using an initial molar ratio of 1.5 between FITC and BSA to gel a final one of 1.5. For subsequent uses, BSA-FITC probe was extensively dialyzed against double-distilled water then lyophilized and kept at +4 °C. Probes properties are gathered in Table 1. According to manufacturer’s information, dextran-FITC have hydrodynamic radii ranging from 2.3 to 11.0 nm for DF-10 and DF-250, respectively, and their FITC labeling is comprised between 0.003 and 0.02 mol FITC per glucose residue. This means that a FITC molecule is branched every 50 to 333 Glc residue. Preparation of FAXs, CNCs, and Gels. Water-extractable FAXs from wheat were provided by the UMR Ingénierie des Agropolymères et Technologies Emergentes (INRA, University of Montpellier and CIRAD, Montpellier, France) and isolated and purified as previously described.21 FA content was 1.25 μg/mg AX, that is, one molecule of FA bound every 670 xylose residues, which is above the threshold of one FA per 2000 xylose residues required for the gelation to happen.21 Laccase (EC from Pycnoporus cinnabarinus30 (Uniprot ID: Q9UVQ2) was provided by the UMR Biotechnologie des Champignons Filamenteux (INRA and Universities of Marseille, Marseille, France). CNCs were prepared from ramie fibers by controlled acidic hydrolysis, resulting in cellulose rod-like nanocrystals 130 ± 5 nm long and between 5 and 6 nm in diameter.31 CNCs suspension was concentrated at 18 mg/mL (1.8%) by a rotary evaporator and stored at +4 °C. Before use, solution was sonicated for 10 min to disperse CNCs. FAX solutions were prepared by adding FAX powder to 0.05 M citrate phosphate buffer at pH 5.5 previously filtered through a 0.22 μm filter. Gels were obtained by mixing FAX solution poured in an eight-chamber coverglass system (Lab-Tek II, Thermo Scientific) with the other necessary components (buffer, fluorescent probe, CNCs) so that the final volume was completed to 500 μL with citrate phosphate buffer. Gelation was triggered by adding 25 μL of laccase at 40 IU·mL−1 at 25 °C and was achieved within 1 h. Xylanase (EC used for FAX gel hydrolysis was a glycoside hydrolase family 11 endoxylanase (Uniprot ID: Q14RS0) from Thermobacillus xylanilyticus, recombinantly expressed, purified, and characterized as previously described.32 For hydrolysis purpose, a droplet of 20 μL of an enzymatic solution at 1800 UI·mL−1 (activity measured at 60 °C on birchwood xylan) was poured onto the gels. Rheological Measurements. A cone−plate geometry (5.0 cm in diameter, 0.04 rad in cone angle) rheometer (ARES, Rheometric Scientific, Champ sur Marne, France) equipped with a Peltier temperature-control system maintained at 25 °C was used for all measurements. FAX solutions were placed onto the plate and intrinsic viscosity [η] measurements were performed. Then, dynamic strain sweep tests were used to determine the range where solutions showed a linear behavior, between 0.0125 and 100%, with a frequency of 10 rad·s−1. For all solutions, 10% strain was used for the


Preparation and Characterization of Fluorescent Probes. Dextran-FITC of 10, 70, and 250 kDa (DF10, DF70, and DF250), bovine serum albumin (BSA), and fluorescein isothiocyanate (FITC) were purchased from Sigma-Aldrich (Saint-Quentin Fallavier, France). For BSA labeling, isothiocyanate function of FITC was able to react with −NH2 amine function of Lys residues of BSA (60 Lys residues in BSA, several of them on the surface) at pH of 9.0 or above to make a covalent bond. BSA was grafted with FITC by mixing 5 mg BSA with 42 μg FITC in 0.1 M carbonatebicarbonate buffer at pH 9.0. Reaction was let 2 h at room temperature under agitation in the dark. Then, the mix was applied to a Sephadex G-25 M column (Sigma-Aldrich) equilibrated with phosphate buffer saline (PBS) buffer, which was used for elution. The first yellow fraction eluted corresponding to the BSA-FITC conjugate was collected. Determination of FITC/BSA molar ratio in the conjugate was performed according to manufacturer instructions 207

dx.doi.org/10.1021/bm201475a | Biomacromolecules 2012, 13, 206−214



measurements, in particular, for recording of G′ (storage modulus) and G″ (loss modulus). Rapid addition of laccase was done directly onto the plate, followed by a short mix of the solution with the cone. G′ and G″ were monitored before, during, and after gelation until they were constant, typically after 60 min. Finally, frequency sweep tests were used to obtain mechanical spectra of gels from 1 to 100 rad·s−1, which are representative of the stability of 3D cross-linked networks. CLSM-FRAP Analysis. The CLSM system was a Leica TCS SP2 (Manheim, Germany) with a 63× oil-immersion objective, equipped with a 488 nm argon laser, placed in a controlled-temperature room (22 ± 2 °C). Samples in the eight-well chamber were put directly onto the objective. Acousto-optical tunable filters (AOTFs) selected light from 493 to 600 nm. Images were collected using the following parameters: 1× zoom factor, 512 × 512 pixels size at a frequency of 400 Hz, and one acquisition. FRAP experiments were conducted with a square region of interest (ROI) of 150 μm size as follows:

lyophilized gels were then observed in a XL 30 SEM (Philips) at an accelerating voltage of 10 kV.

RESULTS AND DISCUSSION FAX Gel Characterization. During gelation process, covalent linkages are catalytically created between close AX chains under the action of a laccase.21,34 Bridges involving two or even three FA can be created, which makes the gel network more or less dense depending on FA content and AX concentration. Gels were prepared using 0.5, 1.0, and 1.5% FAX solutions (w/v). They are classified as hydrogels because their water content is between 98.5 and 99.5% water (w/v). According to previous measurements of swelling experiments, the mesh sizes (ζ) of analogous FAX gels approximately range from 220 to 320 nm.24 These values are at least one order of magnitude higher than the hydrodynamic diameter of the probes analyzed, which range from 4.6 to 22.0 nm (Table 1). Morphological Analysis by SEM. Sections of lyophilized FAX gels of different concentrations (0.5, 1.0, or 1.5% (w/v)) were observed by SEM (Figure 1). At 100× magnification (Figure 1A−C), structures are highly porous and heterogeneous, with pore size clearly decreasing with increasing FAX concentration. Statistical analysis performed on the distribution frequency of the pore diameters for each SEM image shows that (Supporting Information): - Mean pore diameter decreases from 41.4, 34.4, to 21.9 μm in Figure 1A−C, respectively. - More importantly, distribution becomes narrower, which is characterized by the ranges in the middle 50% of the distribution: 20−53 μm in 0.5% FAX gel, 23−39 μm in 1.0% FAX gel, and only 12−23 μm in 1.5% FAX gel. At 500× magnification (Figure 1D−F), gels appear as a mix of sheets and fine strings, whose number also goes up in more concentrated FAX gels. In particular, the network is pretty open in 0.5% FAX gel, whereas it shows many connections in 1.5% FAX gel, but the FAX gel network can be compared with an irregular honeycomb structure in which diffusion pathways of free molecules are not straightforward. Rheology. To better characterize the different FAX gels, the mechanical spectra of the FAX solutions (before gelation) and of the FAX gels (after gelation) were recorded. Whatever the FAX concentration was (0.5, 1.0, or 1.5% (w/v)), the storage modulus G′ and the loss modulus G″ of the FAX solutions were dependent on frequency and G″ > G′: this is typical of a viscous solution (Figure 2A). After gelation, G′ became independent of frequency with G′ > G″, and the ratio G″/G′ (= tan δ) was below 0.1 (Figure 2B): this is typical of an elastic behavior. Indeed, at high frequencies, FAX polymer chains fail to rearrange rapidly upon the imposed stress and therefore stiffen up. This unambiguously demonstrates the transition from a viscous system to an elastic system, where FAX chain mobility is largely decreased, indicating the formation of the gel. FRAP Measurement of Probes in Citrate Buffer. FRAP data were obtained from DF10, D70, DF250, and BSA-FITC at 0.2% in citrate buffer. Because diffusion is supposed to be free in a first approach, data were fitted to a simple exponential equation,35 giving a high correlation coefficient (r2 = 0.99). Experimental diffusion constant Dexp was obtained from eq 2 using τ values from FRAP and compared with theoretical diffusion constant Dth using eq 3, with T = 293 K, η = 1 mPa·s (water viscosity), and R depending on probes from Table 1. Results are presented in Table 2 and show that for the four

- Prebleaching: laser at 10% of its power, acquisition of 10 reference images

- Bleaching: laser at 100%, 10 images - Postbleaching: laser at 10%, acquisition of images until the ROI intensity was constant

- FRAP experiments were repeated four or five times for each sample at different XY positions FRAP analysis was performed by selecting the ROI intensity, I, of each image through ImageJ software (http://rsbweb.nih.gov/ij/) tools. A recovery curve I = f(t) was then drawn and normalized by taking into account the intensities I0 (immediately after photobleaching) and I∞ (at the end of images acquisition). Mobile fraction (MF) is defined as I∞, whereas immobile fraction (IF) is the difference between I0 and I∞. Considering that diffusion of the probe mainly occurs in the two lateral XY dimensions33 with no binding interaction between the probes and the gel structure, fluorescence recovery data were fitted to a simple exponential equation I(t) = I∞·(1 − e−τt), where τ is the recovery time constant. If some interactions exist, then the equation does not properly fit the experimental data. Experimental diffusion, Dexp was derived from33

Dexp =

r2 4τ

(2) where r is the radius of the bleach spot. Theoretical diffusion Dth was computed using the Stokes−Einstein equation

Dth =

k·T 6π·η·R


where k is the Boltzmann constant (1.38 × 10−23 J·K−1), T is the absolute temperature (in kelvins), η is the viscosity of the solution (in pascals per second), and R is the hydrodynamic radius of the probe (in meters). Before evaluating diffusion in FAX gels, we performed FRAP experiments on probes in buffered solutions (citrate buffer used for FAX gels) to evaluate the concentration of probe necessary to obtain good quality images. Considering that the DF probes have variable MW, the amount of probe mixed to the buffer was expressed as a percentage of the buffer volume, and tests were performed on the range 0.05 to 1.0% (w/v). As a result, a concentration of 0.2% was selected, displaying good intensity fluorescence and low noise level. To analyze statistical significance between diffusion coefficient values and between MF values, we performed a one-way ANOVA, followed by a Tukey test (SigmaPlot 11.0, Systat Software) so that values could be sorted in Tukey groups. Scanning Electronic Microscopy. Gels were frozen at −40 °C and lyophilized at −40 °C/0.7 mbar overnight in a Lyovac GT2 freezedrier (GEA Lyophil). Sections of the dry gels were prepared and mounted on aluminum pin stubs using conductive adhesive carbon labels. Samples were then coated with a layer of carbon in a sputter coater MED 10 (Balzers) during a few seconds. Carbonated 208

dx.doi.org/10.1021/bm201475a | Biomacromolecules 2012, 13, 206−214



Figure 1. SEM micrographs of lyophilized FAX gels sections at concentrations of (A,D) 0.5%, (B,E) 1.0%, and (C,F) 1.5%, at 100× magnification (left column) and 500× magnification (right column).

probes analyzed, Dexp and Dth are highly similar with no deviation above 20%. Despite the number of measurement replicates for each probe (four or five), the standard deviation values remained pretty low ( G′ indicating viscous solutions (Figure 9A), but in solutions where CNC concentration was 0.5 or 1.0%, G′ and G″ appear much less dependent on frequency, and more importantly, G′ > G″; solutions could therefore be considered to be more elastic than viscous, not still gels because G″/G′ values were still above 0.1 (Figure 9A). Considering the four solutions, G′ frequency-dependency decreased with the increase in CNC concentration. The addition of a quantity of CNCs in FAX solution equal to that of FAX provoked a dramatic change in the nature of the FAX solution. The presence of CNCs seemed to limit the rearrangement of the AX chains, maybe through some interactions with FAX, which was characterized by high G′ values at high frequencies.

formation of oriented sheet-like structures, which seem to close partially the pore structure without altering the pore size. A similar effect of CNCs addition in hydrogels has already been mentioned.45 Because free CNCs are roughly 130 nm long, which is much less that the size of the visible pores, CNCs could associate between each other or to AX chains to give an orientation to the network. Associations between CNCs and very low-substituted AXs like FAX are indeed favored by the numerous hydrogen bonds that can be created.46 To get more insights into the FAX-CNCs gel properties, we measured diffusion of fluorescent probes DF10, DF70, DF250, and BSA-FITC (Figure 11). Depending on probes, results show 212

dx.doi.org/10.1021/bm201475a | Biomacromolecules 2012, 13, 206−214



gels may be more organized, resulting in fewer entanglements, thus not affecting diffusion coefficients; on the other hand, they might create more compartments, which decrease the MFs of the probes. SEM observations (Figure 10) are in accordance with these results; CNCs seem to create some sheets, giving an orientation to the network that could limit the openings in the network. This observation is corroborated by a previous analysis of a polyacrylamide gel, which was more homogeneous in the presence of rod-like molecules than without,47 which is exactly the case of CNCs. Moreover, it is therefore possible that FAXs and CNCs are able to chemically associate, but the type of the linkages and nature (covalent or noncovalent) remain unknown.




LC plant cell walls are complex materials whose complete destructuration requires gaining more knowledge on their architecture. The strategy we have set up has consisted of measuring the mobility of various fluorescent probes by FRAP in simple and original model systems consisting of CNCs embedded in FAX gels, which represent two of the most important polymers in LC. We have been able to demonstrate that probes diffusion was more dependent on their hydrodynamic radius than on the mesh size of the gels, but anomalous diffusion of some probes could occur and was probably caused by some interactions with the gel polymers. When some CNCs were trapped into the gels, they were able to increase the organization of the gels but to make in the same time some compartments, which resulted in an increase in the diffusion coefficients but a decrease in the MFs. More importantly, FRAP technique can be considered to be a reliable and powerful tool to assay the specificity and efficiency of various deconstructing enzymes (cellulases, hemicellulases, and lignin oxidases) in LC model systems in which a fluorescent probe has been embedded. Measuring the diffusion coefficient and MF of the probe directly gives information on the impact of the deconstructing agent on the LC architecture. In this context, the mobility of the probe could be assayed by varying environmental conditions, like mild temperature that is known to “open” LC network and to facilitate enzyme accessibility to substrate or agitation, which has been shown to make crystalline cellulose more easily degradable by cellulases.48 It could give access to information on synergy between enzymes, type, and concentration of enzymes necessary to obtain a certain degree of mobility. Compared with more classical cartography of lignocellulosic plant cell walls components using CBMs or antibodies, it would bring invaluable new data on the dynamics of the systems. Therefore, investigation of the mobility of fluorescent lignocellulolytic enzymes in more complex systems is under progress and should bring crucial insights for LC destructuration.

Figure 11. Diffusion coefficients Dexp of DF10, DF70, DF250, and BSA-FITC probes in 0.5% FAX gels containing variable concentrations of CNCs. For each probe in each gel, diffusion coefficient values with the same letter above are not significantly different and belong to the same Tukey group (P value < 0.05).

that the addition of CNCs in the 0.5% FAX gels has different effects: for DF10 and DF250, there is no statistical difference between diffusion coefficient values, whereas a slight increase was measured for DF70 and BSA-FITC. If one considers all gels for a given probe, then an increase in CNC concentration has no or a slight positive effect on diffusion but no negative effect, taking into account the uncertainties on measurements. An increase in CNCs concentration has thus less impact on diffusion than an increase in FAX concentration (Figures 4 and 11). Examination of the MFs brings complementary information (Figure 12). First, MFs are dependent on probe size, like

Figure 12. Mobile fractions MFs of DF10, DF70, DF250, and BSAFITC probes in 0.5% FAX gels containing variable concentrations of CNCs. For each probe, the presence of the same letter above the bars indicates that values are not significantly different and belong to the same Tukey group (P value < 0.05).

S Supporting Information * Distribution and statistical parameters of pore size in SEM images. This material is available free of charge via the Internet at http://pubs.acs.org.

diffusions are: MF was 100% for DF10 in all gels, whereas it was ∼90% for the larger DF250 probe, with an obvious statistical difference. In the same gels, the network entanglements reduced by 10 points the space that could be explored by DF250 in comparison with DF10. Contrarily to diffusion coefficients that were stable or increased when CNCs concentrations went up, MFs diminished. This means that the presence of CNCs in FAX gels has a double impact: on the one hand, the structure of the

Corresponding Author *E-mail: [email protected] 213

dx.doi.org/10.1021/bm201475a | Biomacromolecules 2012, 13, 206−214



(26) Vermonden, T.; Jena, S. S.; Barriet, D.; Censi, R.; van der Gucht, J.; Hennink, W. E.; Siegel, R. A. Macromolecules 2010, 43, 782−789. (27) Jervis, E. J.; Haynes, C. A.; Kilburn, D. G. J. Biol. Chem. 1997, 272, 24016−24023. (28) Moran-Mirabal, J. M.; Bolewski, J. C.; Walker, L. P. Biophys. Chem. 2011, 155, 20−28. (29) Cuyvers, S.; Hendrix, J.; Dornez, E.; Engelborghs, Y.; Delcour, J. A.; Courtin, C. M. J. Phys. Chem. B 2011, 115, 4810−4817. (30) Record, E.; Punt, P. J.; Chamkha, M.; Labat, M.; van den Hondel, C.; Asther, M. Eur. J. Biochem. 2002, 269, 602−609. (31) Aguié-Béghin, V.; Molinari, M.; Hambarzymyan, A.; Foulon, L.; Habibi, Y.; Heim, T.; Blossey, R.; Douillard, R. In Model Cellulosic Surfaces; Roman, M., Ed.; ACS Symposium Series 1019; American Chemical Society: Washington, DC, 2009; pp 115−136. (32) Paës, G.; O’Donohue, M. J. J. Biotechnol. 2006, 125, 338−350. (33) Axelrod, D.; Koppel, D. E.; Schlessinger, J.; Elson, E.; Webb, W. W. Biophys. J. 1976, 16, 1055−1069. (34) Dervilly-Pinel, G.; Rimsten, L.; Saulnier, L.; Andersson, R.; Aman, P. J. Cereal. Sci. 2001, 34, 207−214. (35) Sprague, B. L.; McNally, J. G. Trends Cell Biol. 2005, 15, 84−91. (36) Johnson, E. M.; Berk, D. A.; Jain, R. K.; Deen, W. M. Biophys. J. 1996, 70, 1017−1023. (37) Vaiana, S. M.; Emanuele, A.; Palma-Vittorelli, M. B.; Palma, M. U. Proteins 2004, 55, 1053−1062. (38) Fadda, G. C.; Lairez, D.; Arrio, B.; Carton, J. P.; Larreta-Garde, V. Biophys. J. 2003, 85, 2808−2817. (39) Biely, P.; Vršanská, M.; Tenkanen, M.; Kluepfel, D. J. Biotechnol. 1997, 57, 151−166. (40) Burke, M. D.; Park, J. O.; Srinivasarao, M.; Khan, S. A. Macromolecules 2000, 33, 7500−7507. (41) Westbye, P.; Svanberg, C.; Gatenholm, P. Holzforschung 2006, 60, 143−148. (42) Liu, D. G.; Chen, X. Y.; Yue, Y. Y.; Chen, M. D.; Wu, Q. L. Carbohydr. Polym. 2011, 84, 316−322. (43) Zhou, C. J.; Wu, Q. L.; Yue, Y. Y.; Zhang, Q. G. J. Colloid Interface Sci. 2011, 353, 116−123. (44) Wang, Y. X.; Chen, L. Y. Carbohydr. Polym. 2011, 83, 1937− 1946. (45) Aouada, F. A.; de Moura, M. R.; Orts, W. J.; Mattoso, L. H. C. J. Agric. Food Chem. 2011, 59, 9433−9442. (46) Linder, A.; Bergman, R.; Bodin, A.; Gatenholm, P. Langmuir 2003, 19, 5072−5077. (47) Zaroslov, Y. D.; Gordeliy, V. I.; Kuklin, A. I.; Islamov, A. H.; Philippova, O. E.; Khokhlov, A. R.; Wegner, G. Macromolecules 2002, 35, 4466−4471. (48) Kent, M. S.; Cheng, G.; Murton, J. K.; Carles, E. L.; Dibble, D. C.; Zendejas, F.; Rodriquez, M. A.; Tran, H.; Holmes, B.; Simmons, B. A.; Knierim, B.; Auer, M.; Banuelos, J. L.; Urquidi, J.; Hjelm, R. P. Biomacromolecules 2010, 11, 357−368. (49) Venturoli, D.; Rippe, B. Am. J. Physiol.: Renal, Fluid Electrolyte Physiol. 2005, 288, F605−F613.

ACKNOWLEDGMENTS We are grateful to V. Aguié-Béghin, L. Foulon, C. Rémond and N. Aubry (INRA/Reims Champagne-Ardenne University, Reims) for the preparation of CNCs and xylanase; to V. Micard (Montpellier University/INRA, Montpellier) for providing purified FAX; to J.-C. Sigoillot (Marseille Universities/INRA, Marseille) for providing laccase; and to B. Rogé (Reims Champagne-Ardenne University) for her help in rheology measurements. SEM observations were performed at the Cellular and Tissular Imaging Platform with the assistance of C. Terryn. This work was supported by a grant from INRA.


(1) Chundawat, S. P. S; Donohoe, B. S.; Sousa, L. D.; Elder, T.; Agarwal, U. P.; Lu, F. C.; Ralph, J.; Himmel, M. E.; Balan, V.; Dale, B. E. Energy Environ. Sci. 2011, 4, 973−984. (2) Eriksson, T.; Borjesson, J.; Tjerneld, F. Enzyme Microb. Technol. 2002, 31, 353−364. (3) Ximenes, E.; Kim, Y.; Mosier, N.; Dien, B.; Ladisch, M. Enzyme Microb. Technol. 2011, 48, 54−60. (4) Rollin, J. A.; Zhu, Z. G.; Sathitsuksanoh, N.; Zhang, Y. H. P. Biotechnol. Bioeng. 2011, 108, 22−30. (5) Guillon, F.; Tranquet, O.; Quillien, L.; Utille, J. P.; Ortiz, J. J. O.; Saulnier, L. J. Cereal. Sci. 2004, 40, 167−182. (6) McCartney, L.; Marcus, S. E.; Knox, J. P. J. Histochem. Cytochem. 2005, 53, 543−546. (7) Boratson, A. B.; Bolam, D. N.; Gilbert, H. J.; Davies, G. J. Biochem. J. 2004, 382, 769−781. (8) Hilden, L.; Daniel, G.; Johansson, G. Biotechnol. Lett. 2003, 25, 553−558. (9) Porter, S. E.; Donohoe, B. S.; Beery, K. E.; Xu, Q.; Ding, S. Y.; Vinzant, T. B.; Abbas, C. A.; Himmel, M. E. Biotechnol. Bioeng. 2007, 98, 123−31. (10) Filonova, L.; Gunnarsson, L. C.; Daniel, G.; Ohlin, M. BMC Plant Biol. 2007, 7, Article no. 54. (11) Moran-Mirabal, J. M.; Corgie, S. C.; Bolewski, J. C.; Smith, H. M.; Cipriany, B. R.; Craighead, H. G.; Walker, L. P. Anal. Chem. 2009, 81, 7981−7987. (12) Kawakubo, T.; Karita, S.; Araki, Y.; Watanabe, S.; Oyadomari, M.; Takada, R.; Tanaka, F.; Abe, K.; Watanabe, T.; Honda, Y. Biotechnol. Bioeng. 2010, 105, 499−508. (13) McCartney, L.; Gilbert, H. J.; Bolam, D. N.; Boraston, A. B.; Knox, J. P. Anal. Biochem. 2004, 326, 49−54. (14) McCartney, L.; Blake, A. W.; Flint, J.; Bolam, D. N.; Boraston, A. B.; Gilbert, H. J.; Knox, J. P. Proc. Natl. Acad. Sci. U.S.A. 2006, 103, 4765−70. (15) Adams, E. L.; Kroon, P. A.; Williamson, G.; Gilbert, H. J.; Morris, V. J. Carbohydr. Res. 2004, 339, 579−90. (16) Beaugrand, J.; Paës, G.; Reis, D.; Takahashi, M.; Debeire, P.; O’Donohue, M.; Chabbert, B. Planta 2005, 222, 246−257. (17) Dornez, E.; Cuyvers, S.; Holopainen, U.; Nordlund, E.; Poutanen, K.; Delcour, J. A.; Courtin, C. M. J. Agric. Food Chem. 2011, 59, 6369−6375. (18) Beckham, G. T.; Bomble, Y. J.; Bayer, E. A.; Himmel, M. E.; Crowley, M. F. Curr. Opin. Biotechnol. 2011, 22, 231−238. (19) Vogel, J. Curr. Opin. Plant Biol. 2008, 11, 301−307. (20) Ralph, J. Phytochem. Rev. 2010, 9, 65−83. (21) Carvajal-Millan, E.; Landillon, V.; Morel, M. H.; Rouau, X.; Doublier, J. L.; Micard, V. Biomacromolecules 2005, 6, 309−317. (22) Vansteenkiste, E.; Babot, C.; Rouau, X.; Micard, V. Food Hydrocolloids 2004, 18, 557−564. (23) Carvajal-Millan, E.; Guigliarelli, B.; Belle, V.; Rouau, X.; Micard, V. Carbohydr. Polym. 2005, 59, 181−188. (24) Carvajal-Millan, E.; Guilbert, S.; Morel, M. H.; Micard, V. Carbohydr. Polym. 2005, 60, 431−438. (25) Hagman, J.; Loren, N.; Hermansson, A. M. Biomacromolecules 2010, 11, 3359−3366. 214

dx.doi.org/10.1021/bm201475a | Biomacromolecules 2012, 13, 206−214