Studies on methodology in dietary fibre analysis : a neutral detergent fibre method using glucoamylase : a thesis presented in partial fulfilment of the requirements for the degree of Master of Philosophy in Food Technology at Massey University
The dietary fibre content of foods is conveniently and rapidly determined by the neutral and acid detergent methods devised originally by Van Soest and associates. A serious disadvantage of the neutral detergent method relates to the interference caused by starch during filtration when the method is applied to cereals and cereal products. In these circumstances the results of neutral detergent fibre (NDF) measurements are variable and often over-estimated . A study of the starch-lipid reaction which takes place when cereal products are heated in Van Soest's neutral detergent solution showed that although the precipitate derived from pure wheat starch and lipid is soluble in hot water this action is often far from complete when much fibrous cereal matter is present. Much of the starch appears to be occluded in the NDF residue which then takes on a gummy-like character and tends to clog the filter. Southgate recently recommended purified amyloglucosidase from Aspergillus niger (Boehringer) for the purpose of hydrolysing starch in cereal samples before starting the neutral detergent extraction. Present studies have been concerned with the development of this enzymatic procedure with the aim of devising improved methodology and enhancing existing knowledge of the behavioural characteristics of amyloglucosidases from A. niger and from an alternative source, Rhizopus spp. Preliminary investigations showed that amyloglucosidase from A. niger (Boehringer) was completely effective as a starch hydrolysing agent in the pretreatment of a cereal substrate but that in order to use the enzyme economically it was necessary to use a semi micro version of Van Soest's neutral detergent extraction procedure. The main features of the new method are as follows: preparation of a subsample of lipid-free food sample of fine particle size; gelatinization of starch before enzyme treatment; treatment with the minimum quantity of enzyme (2 mg); extraction with neutral detergent at half the normal rate; separation of detergent solution from the residue by means of centrifugation; dehydration of the residue with acetone before filtration; special techniques for filtration, drying and weighing procedures. A table of NDF values for various cereal products determined by the semi micro procedure is presented. The results agree, for the most part, with the results of other workers in this field, the exceptions being for cornflakes, rolled oats and puffed wheat. The coefficients of variation for the NDF values compare favourably with those of other workers. A semi micro version of Van Soest's acid detergent method of evaluating dietary fibre was devised and is described with supporting analytical data. Tests performed with a low cost preparation of amyloglucosidase from Rhizopus spp (Sigma) showed that the crude enzyme was capable of fully hydrolysing the starch component of cereal products before commencing the neutral detergent extraction procedure but that it also seriously reduced the NDF values. In order to establish the cause of the discrepancies two approaches were made: an attempt was made to analyse the products of enzymatic hydrolysis; and a study of the effect of enzyme concentration on the yield of neutral detergent fibre was undertaken. The former approach proved impracticable, the latter suggested that either impurities in the crude enzyme preparation were responsible or the amyloglucosidase itself was active towards one or more components of dietary fibre. In order to determine which of the alternative explanations was correct small amounts of the crude enzyme preparation were purified by means of anion exchange chromatography using DEAE cellulose and one of two buffer systems, one based on citrate-phosphate, the other on tris-HCl. The citrate-phosphate conditions reported by Pazur and Lineback et al for the column separation of amyloglucosidase of A. niger were found to be quite unsuitable for the enzyme from Rhizopus spp. and a new set of conditions had to be determined for this enzyme. The activity of small amounts of the purified enzyme (< 1mg) was estimated by an improvised visual method using buffered 1% wheat starch, and the effect of the enzyme on cereal fibre was determined by means of the semi micro neutral detergent procedure using 0.08-0.2 g wholemeal flour as a substrate. It was found that both crude and purified forms of the enzyme caused a loss of ca 30% NDF from wholemeal flour, from which it was concluded that amyloglucosidase from Rhizopus spp was not a suitable enzyme for use in the neutral detergent method of measuring fibre. A literature review of the known chemistry of the amyloglucosidases of A. niger and R. delemar showed that differences in molecular structure reported by Pazur and others could account for their different electrophoretic properties. In the light of the present work it appears that another important biochemical difference between these enzymes relates to the activity of the Rhizopus enzyme towards the dietary fibre component of cereals.