Towards a rational design of commercial maltodextrins : a mechanistic approach

Industrially produced starch is used for various applications mainly in the food, paper and textile industries. A considerable quantity of this starch is modified (chemically, physically, or enzymatically) before use. The most important modifications are polymer degradation, oxidation, cross-linking, and substitution with various groups, and combinations thereof. This thesis deals with one of these modifications, the partial enzymatic degradation of starch to a saccharide mixtures. These partially hydrolyzed starches (maltodextrins) are used in a wide variety of products, primarily in the food industry. A general introduction to starch and maltodextrin production is given in chapter 1. Maltodextrins are normally characterized by the dextrose equivalent (DE), which is a measure for the number-average molecular weight. Chapter 2 describes the determination of the theoretical dextrose equivalent by measuring the osmolality (mol dissolved particles / kg H ₂ O) by freezing-point depression. Relations for DE and increase in dry weight during hydrolysis were derived as a function of molality and amount of dry weight at the start of the hydrolysis. With freezing-point depression it was possible to determine the theoretical DE of oligosaccharides (dextrose to maltoheptaose), whereas a traditional titration method (Luff-Schoorl) overestimates 20-50%. The overestimation of DE by Luff-Schoorl titration was also evident during the hydrolysis of amylopectin potato starch. With freezing-point depression it was possible to determine the degree of hydrolysis of starch in a fast, reliable and above all accurate way. The relation for the increase of dry weight during the hydrolysis of starch with anα-amylase was experimentally validated. Throughout the work presented in this thesis osmometry was used.

The hydrolysis of amylopectin potato starch with Bacillus licheniformis α-amylase was studied (chapter 3) under industrially relevant conditions (i.e. high dry-weight concentrations). The following ranges of process conditions were chosen and investigated by means of an experimental design: pH [5.6-7.6]; calcium addition [0-120μg/g]; temperature [63-97°C]; dry- weight concentration [3 - 37% [w/w]]; enzyme dosage [27.6-372.4μl/kg] and stirring [0-200 rpm]. The rate of hydrolysis was followed as a function of the theoretical DE. The highest rate (at a DE of 10) was observed at high temperature (90°C) and low pH (6). At a higher pH (7.2), the maximum temperature of hydrolysis shifted to a lower value. Also, high levels of calcium resulted in a decrease of the maximum temperature of hydrolysis. The pH, temperature, and the enzyme dosage showed interactive effects on the observed rate of hydrolysis. No product or substrate inhibition was observed. Stirring did not effect the rate of hydrolysis. The temperature at which the starch was hydrolyzed was found to influence the saccharide composition obtained at the end of the hydrolysis. The number-average molecular weight of these hydrolysates was the same, but their saccharide composition differed. The level of maltopentaose [15-24% [w/w]], a major product of starch hydrolysis catalyzed by B. licheniformis α-amylase, was influenced the most by temperature.

The temperature effects were investigated in more detail in chapter 4. B. licheniformis α-amylase was added to 10% [w/w] gelatinised amylopectin potato starch solutions. The hydrolysis experiments were done at 50, 70, and 90°C. Samples, taken at defined DE-values were analysed with respect to their saccharide composition. At the same DE the oligosaccharide composition depended on the temperature of hydrolysis. This implies that at the same net number of bonds hydrolysed by the enzyme, the saccharide composition was different. The temperature at which hydrolysis was performed also influenced the initial overall molecular-weight distribution. Higher temperatures led to a more homogenous molecular weight distribution. Similar effects were observed forα-amylases from other microbial sources such as Bacillus amyloliquefaciens and Bacillus stearothermophilus . The pH of the reaction (5.1, 6.2, and 7.6) at 70 °C did not significantly influence the saccharide composition obtained during B. licheniformis α-amylase hydrolysis.

The underlying mechanisms for B. licheniformis α-amylase were studied using pure linear oligosaccharides, ranging from maltotriose to maltoheptaose as substrates. Activation energies for the hydrolysis of individual oligosaccharides were calculated from Arrhenius plots at 60, 70, 80, and 90°C. Oligosaccharides with a degree of polymerisation exceeding that of the substrate could be detected. The contribution of these oligosaccharides increased as the degree of polymerisation of the substrate decreased and the temperature of hydrolysis increased. The product specificity decreased with increasing temperature of hydrolysis, which led to a more equal distribution between the possible products formed. Calculations with the subsite map as determined for the closely relatedα-amylase from B. amyloliquefaciens reconfirmed this finding of decreased substrate specificity with increased temperature of hydrolysis.

The reaction temperature proved to be a valuable tool for the production of starch hydrolysate products with more defined saccharide compositions. A model, capable of describing the hydrolysis of starch in terms of the saccharide produced, can be a potent tool for the development of tailor-made maltodextrins. Since no suitable models were available, work was initiated on the development of such a model (chapter 5). The branched structure of potato amylopectin (degree of polymerization ~ 200,000) was modeled in a computer matrix. For the amylopectin molecule the chain-length distribution, the length of a cluster and its width were used as input variables in the model. Independent literature values related to the structure of amylopectin (%β-hydrolysis and ratio of A to B chains) were used for evaluation of the branching characteristics (length of branch area and chance of branching) of the modeled amylopectin.

The structural parameters predicted by the model agreed very well with data from literature. The chain-length distribution and values for the % ofβ-hydrolysis were the two most important parameters required to model the structure of amylopectin. This computer-generated model of potato amylopectin in solution can be used to simulate various enzymatic (i.e.α-amylase,β-amylase, glucoamylase, pullunanase) or chemical reactions (i.e. acid hydrolysis, hypochlorite oxidation). The modeling approach is also suitable for a starch structurally different from amylopectin potato starch as obtained from other botanical sources (i.e. corn, wheat, tapioca). When this modeled amylopectin structure was evaluated with independent literature values related to the structure of amylopectin (i.e. %β-hydrolysis), it was recognized that these values were not corrected for the hydrolytic gain during hydrolysis. This overestimation of the actual percentage ofβ-hydrolysis can however be corrected for (chapter 6). Some of the structural parameters of starch (i.e. % beta- or gluco-hydrolysis) were influenced by the increase in mass during the hydrolysis reactions (hydrolytic gain). Procedures were derived to correct this apparent percentage of hydrolysis to actual percentage of hydrolysis. These analytically derived equations are not only valid for the hydrolysis of starch but also for the hydrolysis of lower molecular weight saccharides (e.g.α-limit dextrin). With a minor modification, these equations can be used to correct for hydrolytic gain that occurs during the hydrolysis of other (bio)polymers.

An application of the computer-generated model of amylopectin in solution is given in chapter 7 by evaluation of the saccharides produced upon hydrolysis with anα-amylase, the principle enzyme used in maltodextrin production. The four different subsite maps presented in literature forα-amylase originating from B. amyloliquefaciens were used to describe the hydrolysis reaction in a Monte Carlo simulation. The saccharide composition predicted by the model was evaluated with experimental values. Overall, the model predictions were acceptable, but no single subsite map gave the best predictions for all saccharides produced. The influence of anα(1 - 6) linkage on the rate of hydrolysis of nearbyα(1 - 4) linkages by theα-amylase was evaluated using various inhibition constants. For all subsite maps considered the use of inhibition constants led to an improvement in the predictions (decrease of residual sum of squares), indicating the validity for the use of inhibition constants as such.

As without inhibition constants, no single subsite map gave the best fit for all saccharides. The possibility of generating a hypothetical subsite map by fitting was therefore investigated. With a genetic algorithm it was possible to construct hypothetical subsite maps (with inhibition constants) that gave further improvements in the average prediction for all saccharides. The advantage of this type of modeling over a regular fit is the additional information about all the saccharides produced during hydrolysis, including the ones that are difficult to measure experimentally. The final chapter deals with the design procedure for more tailor-made starch hydrolysate products. The saccharide composition of a maltodextrin determines both its physical and biological functionality. Aspects related to the saccharide composition such as hygroscopicity, fermentability in food products, viscosity, sweetness, stability, gelation, osmolality, and adsorption by humans are discussed.

The translation of basic knowledge on the behavior of saccharides to the formulation of actual products is illustrated by three examples: a sport drink, maltodextrins in liquid beverages with limited solubility, and parental and enteral fluids. Various factors that can be used to influence the saccharide composition produced during starch hydrolysis are different hydrolytic enzymes, source and concentration of the starch, temperature of hydrolysis, addition of organic solvents, pressure, immobilization of the hydrolytic enzymes, downstream processing, extraction of products during hydrolysis, and combinations thereof. Tools in designing the production process for a maltodextrin with a specific saccharide composition are databases with saccharide compositions obtained at various hydrolysis conditions and hydrolysis models, which can be used to predict new saccharide compositions. A production strategy for desired saccharide compositions at acceptable prices is discussed and future research needs are pointed out.