Xylan chain length and plant cell wall assembly

Abstract

The plant cell wall is a structure surrounding plant cells to serve signalling, developmental, and structural functions. In woody tissues, the cell wall is composed of cellulose (40-60%), lignin (20-30%), and hemicelluloses (20-30%). The plant cell wall is the largest source of renewable carbon (>75% of the biosphere carbon), appearing in numerous products. However, many applications are hindered by our poor understanding of the cell wall. Here, I present my studies on the plant cell wall assembly and the synthesis of xylan, a hemicellulose, with a focus on its chain length. Xylan is a polysaccharide of β-1,4 linked xylosyl residues, some with short branches of acetyls or glucuronic acid. Xylan coats cellulose, but is this coating important for the cell wall assembly? I used solid-state NMR to analyse cell walls of mutants with scarce and short xylan, i.e., with cellulose more uncoated. I found that shorter xylan interacted less with cellulose and more with lignin, and that cellulose glucosyl residues underwent conformational changes upon uncoating. Xylan binds cellulose, and xylan glucuronic acid branches crosslink xylan and lignin. Then, does xylan indirectly connect cellulose and lignin? I found that plants with short and scarce xylan were more severely stunted when their xylan lacked glucuronidation. This phenomenon was not explained by compositional changes, and solid-state NMR revealed that scarce, short, and non-glucuronidated xylan had unusual hydration levels, suggesting a different cell wall arrangement. Xylan has been hypothesised to crosslink microfibrils and to prevent their coalescence, and the partially uncoated cellulose I had analysed also showed conformational changes in its glucosyl residues. Then, does cellulose undergo major structural changes in the complete absence of xylan? I studied plants devoid of xylan with solid-state NMR, unveiling highly pronounced conformational changes in cellulose glucosyl residues. Moreover, X-ray diffraction experiments suggested potential alterations in the arrangement of cellulose chains, and extraction and imaging of these cellulose fibrils seemed to reveal that they indeed had a different width. Three putative glycosyltransferases are required to extend the xylan backbone, as well as two more proteins: IRX15 and IRX15L. Do these serve a structural or a catalytic function? I found that IRX15 and IRX15L keep structural features of methyltransferases needed for their function, and in vitro experiments seemed to support methyltransferase activity. I also found another protein serving the same function in Arabidopsis, IRX15L2, apparently, acting mostly in seed and leaf. Altogether, this work shows a substantial role for xylan coating and chain length in influencing the conformation of glucosyl residues in cellulose and the overall cell wall arrangement. This study also proposes IRX15 proteins to be methyltransferases assisting xylan backbone synthesis.