Engineering artificial casein micelles for future food: Is casein phosphorylation necessary?

Industrial-scale production of recombinant proteins for food products may become economically feasible but correct post-translational modification of proteins by microbial expression systems remains a challenge. For efficient production of hybrid products from bovine casein and recombinant casein, it is therefore of interest to evaluate the necessity of casein post-translational phosphorylation for the preparation of hybrid casein micelles and study their rennet-induced coagulation. Our results show that dephosphorylated casein was hardly incorporated into artificial casein micelles but was capable of stabilising calcium phosphate nanoclusters with an increased size through adsorption on their surface. Thereby, dephosphorylated casein formed larger colloidal particles with a decreased hydration. Furthermore, the presence of increasing amounts of dephosphorylated casein resulted in increasingly poor rennet coagulation behaviour, where dephosphorylated casein disrupted the formation of a coherent gel network by native casein. These results emphasise that post-translational phosphorylation of casein is crucial for their assembly into micelles and thereby for the production of dairy products for which the casein micelle structure is a prerequisite, such as many cheese varieties and yoghurt. Therefore, phosphorylation of future recombinant casein is essential to allow its use in the production of animal-free dairy products.

A quantitative calcium phosphate nanocluster model of the casein micelle: the average size, size distribution and surface properties

Caseins (α_S1, α_S2, β and κ) are the main protein fraction of bovine milk. Together with nanoclusters of amorphous calcium phosphate (CaP) and divalent cations, they combine to form a polydisperse distribution of particles called casein micelles. A casein micelle model is proposed which is consistent with the way in which intrinsically disordered proteins interact through predominantly polar, short, linear, motifs. Using the model, an expression is derived for the size distribution of casein micelles formed when caseins bind to the CaP nanoclusters and the complexes further associate with each other and the remaining mixture of free caseins. The result is a refined coat-core model in which the core is formed mainly by the nanocluster complexes and the coat is formed exclusively by the free caseins. Example calculations of the size distribution and surface composition of an average bovine milk are compared with experiment. The average size, size distribution and surface composition of the micelles is shown to depend on the affinity of the nanocluster complexes for each other in competition with their affinity for free caseins, and on the concentrations of free caseins, calcium ions and other salts in the continuous phase.