Fundamental Studies on the Nature of Protein/Phenolic Interactions, their Role in Diffusion Kinetics and their Binding Characteristics Following Ultra-high Temperature Processing

Interactions amongst food ingredients that have been subject to high temperature treatments have only recently gained attention as a major contributor to both the technological and nutritional functionality of food systems. Ultra-high temperature (UHT) treated liquid food systems are particularly sensitive to changes in their organoleptic properties due to their long shelf life and requisite severe heat treatment. Furthermore, the food industry is looking to satisfy consumer demand for healthy convenience products which contain high levels of protein and an ever-increasing proportion of functional plant based ingredients, most commonly dietary fibres, which has to date, mainly involved the inclusion of soluble dietary fibres (SDF) (due to their ease of incorporation) in the formulation, however, insoluble dietary fibres (IDF) have commonly become desirable inclusions to reduce production waste streams and take advantage of their relatively high proportion of bioactive microconstituents, for example cereal based IDF contain double the proportion of phenolic compounds when compared to their SDF counterparts. The inclusion of IDF presents issues to the textural properties of the beverage, however, this can be overcome, often via the milling of particles to a level that is imperceptible to the human palate and the addition of thickeners/stabilisers to increase liquid phase viscosity, thereby reducing sedimentation of insoluble particles and increasing shelf stability. However, subsequent UHT treatment of formulations containing insoluble fibre also presents other unintended consequences, such as the breaking of ester linkages that bind phenolic compounds such as ferulic acid to the lignin backbone of many insoluble fibres originating from cereal grains. Breaking of these bonds and subsequent phenolic extraction is facilitated by the high pressure, high temperature (usually around 140C) water which is ubiquitous in UHT treated liquid beverages. Although the extraction of previously IDF bound phenolic compounds does not present a challenge in itself (in fact phenolic compounds are suspected to have a positive impact on health due to their antioxidant properties), the now free phenolic compounds become available in solution for a variety of (usually detrimental) interactions. For example, free ferulic acid can be thermally or enzymatically decarboxylated to p-vinylguaiacol (PVG), and undesirable aroma compound. More interestingly, recent research indicates that free phenolic compounds can be thermally oxidised to their corresponding quinone/semiquinone, leading to the nucleophilic attack of electrophilic amine groups residing on residues such as lysine, culminating to the covalent attachment of the phenolic compound to an amino residue. This is the focus of the first and second experimental chapters, which propose that the phenolic compounds, ferulic acid and 4-hydroxybenzoic acid, may covalently attach to a glutamine and a lysine residue respectively. The implications of such attachments are not yet agreed upon, with some research indicating that covalent attachment of phenolic compounds leads to an increase in antioxidant activity, and other research finding a decrease in protein digestibility or reduced techno-functionality. To shed further light on the impact that covalent attachment of a phenolic compound may have on protein structure, the third experimental chapter involves the in silico creation of a covalently linked 4-hydroxybenzoic acid - β-casein complex. This complex is subjected to a dynamic analysis and compared to an unmodified β-casein molecule. The covalent complex was shown to have more disruptive (more frequent, but less stable) intramolecular hydrogen bond formation between the affected residue and the protein. This significantly destabilised the protein in vicinity of the covalent attachment, leading to increased solvent accessibility. Benchtop solubility experiments supported this in silico observation, with the covalent adduct showing a substantially increased solubility when compared to its unmodified counterpart. The fourth experimental chapter of this thesis examines the role of interactions between the diffusant and the release medium in modulating the diffusional behaviour of phenolic compounds. The diffusion behaviours of two structurally distinct phenolic compounds, ferulic acid and epicatechin, were examined in the presence and absence of β-casein. The rate of diffusion in both cases was decreased when in the presence of the protein, with the magnitude of this decrease being proportional to the interaction strength between the protein and the phenolic, which were determined and visualised using in silico pulling simulations. This work highlights the need to consider interactions between the delivery medium and the diffusant in predicting diffusional behaviours of bioactive compounds. Throughout this thesis, increasingly complex molecular modelling techniques are utilised, providing an example of how to meaningfully apply this type of analysis in the fields of food chemistry and biophysics. This work extends the knowledge of protein-phenolic covalent interactions and proposes binding constants, stoichiometry, and locations on the β-casein molecule for a variety of phenolic compounds. It also demonstrates the use of appropriate techniques, prerequisites and equations for this type of analysis, as opposed to still commonly used, but inappropriate linearised methods. The common pitfalls of this type of analysis, and more accurate/appropriate techniques that can be used to correctly determine binding parameters are outlined in the final chapter of this thesis.