These are not extracted by chelating agents or ice-cold Na2CO3, but are by concentrated aqueous alkali. They generally share the property of hydrogen-bonding to cellulose, at least in vitro; and some hemicelluloses probably also do this in muro, tethering adjacent microfibrils (Fry 1989; Hayashi 1989) . For this reason, the term 'crosslinking glycans' was suggested for hemicelluloses. However, this term is not used here because the proposed in-muro tethering role remains largely hypothetical in many cases, and also because some non-hemicellulosic polysaccharides (e.g. rhamnogalacturonan II, RG-II) do crosslink.
Was this article helpful?