The pattern of xylan acetylation suggests xylan may interact with cellulose microfibrils as a twofold helical screw in the secondary plant cell wall of Arabidopsis thaliana

Abstract

The interaction between xylan and cellulose microfibrils is important for secondary cell wall properties in vascular plants; however, the molecular arrangement of xylan in the cell wall and the nature of the molecular bonding between the polysaccharides are unknown. In dicots, the xylan backbone of β-(1,4)-linked xylosyl residues is decorated by occasional glucuronic acid, and approximately one-half of the xylosyl residues are O-acetylated at C-2 or C-3. We recently proposed that the even, periodic spacing of GlcA residues in the major domain of dicot xylan might allow the xylan backbone to fold as a twofold helical screw to facilitate alignment along, and stable interaction with, cellulose fibrils; however, such an interaction might be adversely impacted by random acetylation of the xylan backbone. Here, we investigated the arrangement of acetyl residues in Arabidopsis xylan using mass spectrometry and NMR. Alternate xylosyl residues along the backbone are acetylated. Using molecular dynamics simulation, we found that a twofold helical screw conformation of xylan is stable in interactions with both hydrophilic and hydrophobic cellulose faces. Tight docking of xylan on the hydrophilic faces is feasible only for xylan decorated on alternate residues and folded as a twofold helical screw. The findings suggest an explanation for the importance of acetylation for xylan-cellulose interactions, and also have implications for our understanding of cell wall molecular architecture and properties, and biological degradation by pathogens and fungi. They will also impact strategies to improve lignocellulose processing for biorefining and bioenergy.

Keywords: Arabidopsis thaliana; acetylation; cellulose interaction; plant cell wall molecular architecture; xylan.

Figure 1

Model of acetylxylan interactions with a 24-chain cellulose microfibril. The end view and side view are shown. DP10 xylan chains with evenly spaced 2-O-Ac decorations at every two xylosyl residues were modelled as a twofold helical screw (2₁). (a) Xylan chains placed on hydrophilic (010) and (020) faces. (b) Xylan chains placed on hydrophobic (100) and (200) faces.

Figure 2

Digestion of gux1 gux2 mutant Arabidopsis acetylated xylan with CmXyn10B and EcXyn30 analysed by PACE. Major oligosaccharides have degrees of polymerization (DPs) of multiples of two xylosyl residues. Digestion was carried out with low (L) or high (H) enzyme loads. After deacetylation with NaOH, the predominantly even DP of products is apparent by comparison with xylo-oligosaccharide markers DP1–DP6 (M). No E, no enzyme digestion acetylxylan control.

Figure 3

MALDI-ToF-MS of xylanase-digested gux1 gux2 acetylated xylan. (a) CmXyn10B digestion of acetylated xylan extracted from delignified gux1 gux2 stem cell walls. (b) CmXyn10B digestion of gux1 gux2 stem cell wall alcohol-insoluble residue. (c) 2-Aminobenzoic acid (2-AA)-labelled EcXyn30 digestion of acetylated xylan extracted from delignified gux1 gux2 stem cell walls. (d) CmXyn10B digestion of acetylated xylan in Golgi membrane vesicles prepared from gux1 gux2 stems.

Figure 4

High-energy MALDI-CID MS/MS of the Xyl₄Ac₂ oligosaccharide released by CmXyn10B from gux1 gux2 acetylated xylan, labelled with 2-AA and separated by HILIC.

Figure 5

NMR analysis showing relationship of acetylated and non-acetylated xylosyl residues. (a) Two-dimensional ¹³C HSQC spectrum showing the assignment of ¹H attached to acetylated ¹³C (top panel) and the anomeric region (below). (b) Two-dimensional ¹H–¹H NOESY and TOCSY spectra, shown in blue and red, respectively. Detailed analysis of the NOE cross peaks allows the various non-acetylated xylosyl residues to be distinguished. Key NOEs, connecting inter-residue H1 and H4/H5_eq in the 50-ms mixing-time experiment, and H1–H1 in the 200-ms mixing-time experiment, are shown by arrows. Dotted circles highlight the absence of NOEs between: (i) X2 and X3; (ii) non-acetylated Xyl with itself, and (iii) X3 and X3. (c) Anomeric region of two-dimensional ¹³C HSQC spectra showing the close similarity of chemical shifts in WT and gux1 gux2 acetylated xylan.

Figure 6

MALDI-ToF-MS of xylanase-digested wild-type (WT) acetylated xylan. (a) CmXyn10B digest of acetylated xylan extracted from delignified WT stem cell walls. (b) CmXyn10B digestion of WT stem cell wall alcohol-insoluble residue.

Figure 7

Molecular dynamics simulation of xylan (a, unsubstituted; b, acetylated; c, glucuronosylated) interacting with hydrophilic (010) and (020), or hydrophobic (100) and (200), surfaces of cellulose. Occupancy-level isosurfaces are shown. Dark- and light-coloured isosurfaces represent spatial regions where substituted xylan is present 40–50% and 20% of the simulation time, respectively.

Figure 8

Xylan in 3₁- and 2₁-fold screw conformations. Histograms showing the distribution of glycosidic dihedral angles Φ + Ψ between adjacent xylose residues of unsubstituted xylan of DP10. Numbers refer to xylose residues. (a) Xylan in water: Φ + Ψ ∼ 190° in water indicates threefold (3₁) helical conformation. (b) xylan on cellulose face 010. Φ + Ψ ∼ 120° indicates twofold (2₁) conformation.

Figure 9

Neighboring xylan and glucan chains for xylan (DP10) adsorbed on the hydrophilic faces of cellulose. (a) Scattered plot of the interchain distance against the sum of dihedrals Φ + Ψ, showing that when the chains are close to each other (xylan adsorbed onto cellulose) only xylosyl twofold screw, 2₁, conformations occur, whereas when the chains are farther apart, xylan assumes a threefold screw, 3₁, conformation. (b) The predominant xylan–glucan hydrogen-bonding mode has xylosyl O2 as the proton donor and glucosyl O6 as the proton acceptor.

Figure 10

Hypothetical model of cellulose–xylan interactions in the secondary cell walls of dicots. A single plane of a cellulose crystallite with a partial xylan shell in a 2₁-fold screw is shown. The major domain is accommodated on vacancies on the hydrophilic surface. The minor domain cannot bind to the hydrophilic surface as a 2₁-fold screw.