Exploring extraction/dissolution procedures for analysis of starch chain-length distributions

The analysis of starch chain-length distributions (CLDs) is important for understanding starch biosythesis-structure-property relations. It is obtained by analyzing the number distribution of the linear glucan chains released by enzymatic debranching of starch α-(1→6) glycosidic bonds for subsequent characterization by techniques such as fluorophore-assisted carbohydrate electrophoresis (FACE) or size-exclusion chromatography (SEC). Current literature pretreatments for debranching prior to CLD determination involve varying protocols, which might yield artifactual results. This paper examines the two widely used starch dissolution treatments with dimethyl sulfoxide (DMSO) containing 0.5% (w/w) lithium bromide (DMSO-LiBr) at 80 °C and with aqueous alkaline (i.e. NaOH) solvents at 100 °C. Analyses by FACE with a very high range of degree of polymerization, and by SEC, of the CLD of barley starches with different structures show the following. (1) The NaOH treatment, even at a dilute concentration, causes significant degradation at higher degrees of polymerization, leading to quantitatively incorrect CLD results in longer amylopectin and in amylose chains. (2) Certain features in both amylopectin and amylose fractions of the CLD reduced to bumps or are missing with NaOH treatment. (3) Overestimation of amylose chains in starch CLD due to incomplete amylopectin dissolution with dilute NaOH concentration. These results indicate starch dissolution with DMSO-LiBr is the method of choice for minimizing artifacts. An improved pretreatment protocol is presented for starch CLD analysis by FACE and SEC