Multivariate model to characterise relations between maize mutant starches and hydrolysis kinetics

Determining whether granule structural or surface features govern the wheat starch digestion, a kinetic analysis

Deciphering the determinants of starch digestion from multiple interrelated properties is a challenge that can benefit from multifactorial data analysis. The present study investigated the digestion kinetic parameters (rate, final extent) of size-fractions from four commercial wheat starches with different amylose contents. Each size-fraction was isolated and characterized comprehensively using a large range of analytic techniques (FACE, XRD, CP-MAS NMR, time-domain NMR, DSC…). A statistical clustering analysis applied on the results revealed that the mobility of water and starch protons measured by time-domain NMR was consistently related to the macromolecular composition of the glucan chains and to the ultrastructure of the granule. The final extent of starch digestion was determined by the granule structural features. The digestion rate coefficient dependencies, on the other hand, changed significantly with the range of granule size, i.e. the accessible surface for initial binding of α-amylase. The study particularly showed the molecular order and the chains mobility predominantly limiting or accelerating the digestion rate depending on the accessible surface. This result confirmed the need to differentiate between the surface and the inner-granule related mechanisms in starch digestion studies.

How Does Starch Structure Impact Amylolysis? Review of Current Strategies for Starch Digestibility Study

In vitro digestibility of starch is a common analysis in human nutrition research, and generally consists of performing the hydrolysis of starch by α-amylase in specific conditions. Similar in vitro assays are also used in other research fields, where different methods can be used. Overall, the in vitro hydrolysis of native starch is a bridge between all of these methods. In this literature review, we examine the use of amylolysis assays in recent publications investigating the complex starch structure-amylolysis relation. This review is divided in two parts: (1) a brief review of the factors influencing the hydrolysis of starch and (2) a systematic review of the experimental designs and methods used in publications for the period 2016–2020. The latter reports on starch materials, factors investigated, characterization of the starch hydrolysis kinetics and data analysis techniques. This review shows that the dominant research strategy favors the comparison between a few starch samples most frequently described through crystallinity, granule type, amylose and chain length distribution with marked characteristics. This strategy aims at circumventing the multifactorial aspect of the starch digestion mechanism by focusing on specific features. An alternative strategy relies on computational approaches such as multivariate statistical analysis and machine learning techniques to decipher the role of each factor on amylolysis. While promising to address complexity, the limited use of a computational approach can be explained by the small size of the experimental datasets in most publications. This review shows that key steps towards the production of larger datasets are already available, in particular the generalization of rapid hydrolysis assays and the development of quantification approaches for most analytical results.

Digestion kinetics of native and modified cardaba banana starch: A biphasic approach

To understand the mechanism through which cardaba banana starch is hydrolysed, the starch digestion kinetics of native and modified cardaba banana starch samples of Nigeria origin were investigated using an in vitro procedure. The digestion kinetics of the starch samples revealed the samples exhibited a biphasic digestogram. A second-order polynomial with an average coefficient of determinant (r²) of 0.7732 (p < 0.005) was used to segment the biphasic digestogram into two monophasic digestograms. The digestion kinetics parameters (average) obtained using a modified first-order model suggested the accuracy of the model in describing the digestogram. The values obtained for the initial and final digestion rate constant (initial, k_i = 3-4 × 10^-2 min^-1; final k_f = 6-8.3 × 10^-2 min^-1) revealed that the final monophasic segment had a faster rate of digestion after the initial resistant to digestion had been overcome. The logistic model approach in which the digestogram was carried out in a single process also accurately predicted the biphasic behaviours of the cardaba banana (average r² = 0.9736, p < 0.05; root mean square of error, RMSE = 1.588). Weibull model was used for the first time to describe the biphasic approach of cardaba banana starch and according to the digestogram parameters obtained (average r² = 0.9954, p < 0.05; root mean square of error, RMSE = 0.578), the model accurately predicts the biphasic digestogram. In comparison among the models, the Weibull model best described the biphasic digestogram of the cardaba banana starch. The maximum starch digestion obtained in each of the digestion approaches was less than 100% which is an indication of the presence of resistant starch.

Modeling Starch Digestograms: Computational Characteristics of Kinetic Models for in vitro Starch Digestion in Food Research

Starch digestion is mostly investigated with in vitro techniques, and time-course measurements are common. These yield digestograms that are modeled by theoretical, semitheoretical, and empirical kinetic equations, many of which are reviewed here. The Duggleby model has Michaelis-Menten functions, and its dependent variable is on both sides of the equation with no apparent parameter for maximum digestible starch (D_∞ ). The Gaouar and Peleg models are equivalent. They predict both the initial digestible starch (D₀ ) and D_∞ , and an average digestion rate, but they can reveal "biratial" digestions. The first-order kinetic model exhibits diverse predictabilities and, when linearized, D_∞ is sometimes equated to 100 g/100 g dry starch (100%), it yields an average rate of digestion and can predict negative D₀ . The log of slope (LOS) model is unique in revealing the rapid-to-slow digestion rate phenomenon, but without guidelines to identify such. The LOS model does not sometimes use all the digestogram data, can predict D_∞ greater than 100%, and returns zero digestion rate for some digestograms. However, some starchy materials exhibit a slow-to-rapid digestion rate phenomenon, as demonstrated with an example. The modified first-order kinetic model uses all the digestogram data with practical constraints (D₀  ≥ 0 g/100 g dry starch; D_∞  ≤ 100 g/100 g dry starch), describes all digestograms, and yields an average digestion rate, but it can also be used for "biratial" digestions. In addition, the logistic and Weibull models are discussed. Using some published data, the computational characteristics of these commonly used models are presented with objective parameters to guide choices.