α-(1–4) Chain distributions of three-dimensional, randomly generated models, amylopectin and mammalian glycogen: comparisons of chromatograms of debranched chains of these polysaccharides and models with random dendrimeric models with the same chain lengths (CL, ICL, ECL) and fractions of A chains

Abstract

Previously published chromatographic profiles of amounts of debranched α(1-4) chains versus d.p. of 10 amylopectins ( w-, n- and ae-maize, w- and n-rice, w- and n-potato and w-, n- and HA-barley) have been compared with convolved plots of randomly extended and branched, spatially unrestricted models (random dendrimeric) with similar average chain lengths (CL, ECL and ICL) and fractions of A chains: the number of B k chains in the model was Tb 2 a k−1 where k is the number of branches. Eight of these comparisons had an undulating, tri-modal pattern of differences (experimental minus model) in the numbers of chains. This was positive at the calculated d.p. of low branched ( B 1) chains, negative at medium branched ( B 2–5) chains and positive at highly branched (>B 5) chains. Two gave a bimodal pattern. Plots of glucan versus d.p. emphasised differences at high branching. Convolved profiles of numbers of types of chains for constructs generated on a three-dimensional cubic grid, in which the positions occupied by chains were exclusive, with extension and branching random—as well as random with a preference for branching into one plane—were compared with spatially unrestricted models with the same average chain lengths and fraction of A chains. Some had a similar undulating tri-modal pattern of differences in chain numbers versus d.p. and these occurred at the same ranges of types of chains to those between debranched amylopectin and the spatially unrestricted models. These differences appeared when the A chain content was equal to or higher than that of amylopectin and were much more marked when the structure of the three-dimensional models was limited along one axis, producing an oblate ellipsoidal shape. These three-dimensional models had more B chains with higher numbers of branches and more chains with longer chain lengths than the spatially unrestricted model with the same number of A chains. In a printout of a two-dimensional construct based on a square grid, with a total of 500 chains (0.58 of which were A chains) a significant proportion of A chains were inside the radius of gyration, and the number of tiers was twice that of the spatially unrestricted model with the same total number of chains and fraction of A chains. The results suggest a model for the distribution between chains with differing numbers of branches in amylopectin, which is affected by steric restrictions. When the A chain content was near to that of mammalian glycogen and expansion not limited along any axis, giving three-dimensional models with a spheroidal shape, the B chain distributions for the three-dimensional and spatially unrestricted models were similar. The resemblance of convolved plots of these to chromatograms of debranched glycogen-also with a spheroidal shape-indicates that its B chain distribution is similar to a randomly generated model in which steric restrictions do not apply.