Sensopeptidomic Kinetic Approach Combined with Decision Trees and Random Forests to Study the Bitterness during Enzymatic Hydrolysis Kinetics of Micellar Caseins

Abstract

Protein hydrolysates are, in general, mixtures of amino acids and small peptides able to supply the body with the constituent elements of proteins in a directly assimilable form. They are therefore characterised as products with high nutritional value. However, hydrolysed proteins display an unpleasant bitter taste and possible off-flavours which limit the field of their nutrition applications. The successful identification and characterisation of bitter protein hydrolysates and, more precisely, the peptides responsible for this unpleasant taste are essential for nutritional research. Due to the large number of peptides generated during hydrolysis, there is an urgent need to develop methods in order to rapidly characterise the bitterness of protein hydrolysates. In this article, two enzymatic hydrolysis kinetics of micellar milk caseins were performed for 9 h. For both kinetics, the optimal time to obtain a hydrolysate with appreciable organoleptic qualities is 5 h. Then, the influence of the presence or absence of peptides and their intensity over time compared to the different sensory characteristics of hydrolysates was studied using heat maps, random forests and regression trees. A total of 22 peptides formed during the enzymatic proteolysis of micellar caseins and influencing the bitterness the most were identified. These methods represent simple and efficient tools to identify the peptides susceptibly responsible for bitterness intensity and predict the main sensory feature of micellar casein enzymatic hydrolysates.

Enzymatic preparation of casein hydrolysates with high digestibility and low bitterness studied by peptidomics and random forests analysis

“…A 2% hydrolysate-water solution (w/w) was prepared and coded with 3-digit random numbers. Eight trained panelists (3 males and 5 females, aged [22][23][24][25] were given samples at room temperature in a randomized order. Bitterness was rated using a continuous line scale with labeled end points ranging from no intensity (0, casein as reference), to high intensity (5, papain casein hydrolysates at the DH of 13% as reference).…”

“…However, the release pattern of bitter peptides by different enzymes remains unknown. Daher et al 24 established a method to screen bitter peptides using a heat map and random forests. The random forests model is a tree-based approach, which was used to model the bitterness descriptor as a function of a number of predictors ( presence, absence or normalised abundance of peptides) and quantify the contribution of each predictor in the study above.…”

Review on the release mechanism and debittering technology of bitter peptides from protein hydrolysates

“…However, the debittering mechanism of biological debittering has not been systematically studied and reviewed. Understanding the mechanism of biological debittering can help inform the selection of a certain type of enzyme preparation and determine the degree of hydrolysis to generate hydrolysates with adequate sensory properties (Daher et al., 2021). Further advancements will lead to the designs combining the physicochemical and biological technologies for the purpose of producing high‐quality products (Su et al., 2021).…”

Proteomics and Food Analysis: Principles, Techniques, and Applications

“…In this context, the manuscript published by Daher et al, "Sensopeptidomic kinetic approach combined to decision trees and random forest to study the bitterness during enzymatic hydrolysis kinetics of micellar caseins", reports the proteomic characterization of protein hydrolysates responsible for the unpleasant taste in milk [1]. Milk protein hydrolysates have significant advantages in sport nutrition and elderly and infant nutrition, as the use of these hydrolysates induces a very rapid release of amino acids in the blood, which maximizes muscle protein anabolism and facilitates body recovery and nutrition.…”