Impact of different aqueous phases on casein micelles: Kinetics of physicochemical changes under variation of water hardness and diafiltration conditions

Abstract

Besides deionised water, tap or softened water could be used as diafiltration media for separation and purification of the casein fraction from milk by means of membrane technology. The effect of compositional differences on key properties of casein micelles was investigated with these media and simulated milk serum as reference at laboratory scale under conditions mimicking typical industrially applied diafiltration processes. Evaluation of zeta-potential, particle size distribution, voluminosity as well as casein and calcium composition revealed marked differences of micellar attributes at temperatures of 10 and 50 °C depending mainly on the medium hardness. While tap water induced a slight micellar contraction, dehardened water types differently fostered a swelling that was followed by partial disintegration over contact time especially at 50 °C. Results demonstrate that water composition impacts micellar properties in typical DF conditions, which can, in consequence, be crucial for processing efficiency and product functionality of obtained casein products.

Introduction

The background of this study on casein micelle structural changes is that ultrafiltration (UF) and microfiltration (MF) operated in crossflow mode are established processing unit operations at dairies for concentration, fractionation and purification of milk proteins. Based on the pore size of the chosen membranes certain separation effects are at first expected, but despite crossflow conditions deposit formation on the membrane surfaces, mainly by casein micelles, is known to have significant effects on flux and the actual separation effect rendering the membrane operation results often unpredictable (Jimenez-Lopez et al., 2008; Vetier, Bennasar, & De La Fuente, 1988). When MF is applied for the fractionation of casein micelles and whey proteins, a process mode of operation known as diafiltration (DF), water is used as DF media to completely wash out the whey proteins and to obtain a whey and casein micelle isolate. Depending on the DF medium used, for instance waters with different levels in hardness, partially dehardened or deionised water, evaporation condensate or reserve osmosis permeate, the casein micelles will react by establishing a new equilibrium in ion distribution between the calcium bound in the micelle structures and in the surrounding medium (Gaucheron, 2005; Lewis, 2011). Also changes in size, charge, composition and overall resulting voluminosity of the casein micelles can be expected. Besides a change in product functionality, this, in turn, is highly likely to lead to modifications of the structure of the layer of casein micelles as building blocks, deposited on the membrane surface at high local concentration.

The impact of such changes related to the use of different DF media on the casein micelle has not been studied in detail so far. However, before studying the MF-DF process in this context and the effects of DF medium composition on flux, deposited layer structure and permeation of whey proteins, the related changes on the casein micelle itself as the building block of deposited layer on membrane surfaces were assessed in this study as a base for considering different water qualities of DF media in the fractionation of milk proteins.

As a base for this study, recent reviews on the composition, inner structure and ionic equilibrium upon modification of milieu conditions published by Broyard and Gaucheron (2015), Gaucheron (2005), Huppertz et al. (2017), are referred to without repeating knowledge established in the public domain. Regarding composition and interaction potential it is well-known that casein micelles are affected by changes of ionic milieu and pH of the surrounding medium. Due to a chemical equilibrium between the aqueous and micellar phase, shifts of serum composition especially regarding different forms of phosphate, citrate and calcium salts are crucial for the degree of colloidal calcium phosphate (CCP) saturation and, in consequence, for the casein micelle integrity (Gaucheron, 2005). Shifts to lower pH values decrease the amount of micellar bound CCP and also leads to the destabilisation of the κ-casein shielding against aggregation of casein micelles which is finally achieved at pH values close to the isoelectric point of 4.6 at ambient temperatures (Broyard & Gaucheron, 2015). Alkalisation also results in a release of calcium ions, but mainly based on a different mechanism and not in the form of calcium phosphate. Furthermore, a loosening or even disruption of the micellar structure can be observed, which is induced by increased repulsive protein–protein interactions deriving from an increase of the negative charge of protein groups with rising pH (Ahmad, Piot, Rousseau, Grongnet, & Gaucheron, 2009; Dombrowski, Dechau, & Kulozik, 2016; Vaia, Smiddy, Kelly, & Huppertz, 2006). Milieu changes can in sum modify apparent micellar size, voluminosity and interaction between single micelles, which in consequence influences techno-functional properties (Dombrowski et al., 2016; Huppertz & Fox, 2006; Le Berre & Daufin, 1998; Marchin, Putaux, Pignon, & Léonil, 2007). Different alterations of micellar properties have also been shown to impact deposit formation and, therefore, filtration performance (Bouzid et al., 2008; Jimenez-Lopez et al., 2011; Kühnl et al., 2010; Rabiller-Baudry, Gesan-Guiziou, Roldan-Calbo, Beaulieu, & Michel, 2005).

During industrially applied filtration processes, an alteration of the natural ionic equilibrium state is also induced by changes to processing temperatures of 10 °C or 50 °C, typically applied for bacteriostatic reasons, as well as the by the concentration of the casein fraction. Such conditions have already been studied and were reported to impact the micellar structure and functionality (Crowley et al., 2018; Ferrer, Alexander, & Corredig, 2011; McCarthy, Wijayanti, Crowley, O'Mahony, & Fenelon, 2017; Srilaorkul, Ozimek, Ooraikul, Hadziyev, & Wolfe, 1991). To completely separate the main protein fractions, namely micellar casein, often also referred to as phospho-caseinate, and whey proteins, a microfiltration process is operated in DF mode (Pierre, Fauquant, Le Graet, Piot, & Maubois, 1992). The MF filtrate or permeate takes with it parts of the whey proteins in amounts according to the volume of filtrate. The remaining aqueous phase on the retentate side thereby still contains whey proteins, which have to be gradually removed by a suitable washing liquid, which replaces the natural milk serum.

Deionised water (DW) is commonly applied for DF in the dairy industry (de Boer, 2014) as well as for isolation of micellar casein for scientific purposes (Schuck et al., 1994). The application of tap water (TW) or softened water (SW) in DF instead of DW could reduce costs and increase sustainability of this processing step due to the avoidance of a high purification process. These media differ markedly from DW in ionic strength, calcium content and pH, but can also be provided in a microbial quality that is safe for processing over several hours. Using DW as DF medium implies that, next to the whey proteins, the majority of low molecular solutes like lactose and minerals present in the serum phase are washed out of the casein-rich retentate fraction and increase product purity (Pierre et al., 1992). Several further studies report on alterations of casein micelles under these conditions impacting also the functionality of the protein (Alexander, Nieh, Ferrer, & Corredig, 2011; Boiani, McLoughlin, Auty, FitzGerald, & Kelly, 2017; Famelart, Lepesant, Gaucheron, Le Graet, & Schuck, 1996; Ferrer, Alexander, & Corredig, 2014; Li & Corredig, 2014; Liu et al., 2017; McKenna, 2000; Renhe, Zhao, & Corredig, 2019; Sachdeva & Buchheim, 1997). However, the authors observed partially contrary analytical results regarding effects on size, zeta-potential and composition of casein micelles. In addition to differences in casein source material, applied analytical method and casein concentration, discrepancies can also be attributed to deviant DF and sample handling conditions.

In summary, much is known with regard to the impact of the milieu composition on the casein micelle properties. None of these studies mentioned above, however, reported on the effects and kinetics of a gradual change from the composition of the natural milk serum to a complete removal of all low molecular weight constituents by DW accounting for extent as well as duration of milieu change and temperature at DF. Also, no report, to our knowledge, published insights regarding alterations to casein micelles induced by different industrially available water types. Our hypothesis was that new insights would be gained from studying these effects in combination and as function of time when gradually replacing the natural milk serum by DF medium. Based on the decisive role of calcium activity for casein micelle integrity and functionality (Dumpler, 2018) and the difference in ionic calcium upon dilution of milk with soft and hard water (Lewis, 2011) we particularly thought that a stabilising effect on the micellar structure during DF could possibly be achieved by an increase of water hardness.

The aim of this work was to shed light on the question how different compositional attributes of water affect casein micelle properties as a function of contact time in dependence of high and low temperature (50/10 °C), casein concentrations and grades of milieu exchange that occur along any DF process during milk protein fractionation by MF. Therefore, we studied the kinetics of casein micelle structural and compositional changes following the gradual and simultaneous changes when replacing the natural milk serum during different DF conditions using various aqueous media. Our approach, in brief, was as follows: isolated casein micelles were exposed to DW, SW and TW under different conditions (speed and degree of milieu exchange, temperature and concentration) at laboratory scale using the natural milk serum as control and resulting effects on the micellar composition, size, zeta-potential and voluminosity were investigated.