Exploring the Impact of Reconstituted Skim Milk Pre-treatments on Partitioning of Caseins and Rennet Activity During Cheese-making

Cheese-making offers the opportunity to exploit the unique flavour and health benefits of fermented milk proteins. The process involves several operations including milk pretreatments, gelation/curdling and transforming curd into intended cheese. Curdling resulting from destabilisation of casein (CN) micelles is a critical step of the process with the key for many cheeses being the hydrolysis of κ-CN by rennet into para-κ-CN and caseino/glycomacropeptide. A review of the literature shows that the extent/degree of κ-CN hydrolysis that is required to induce gelation of bovine milk during renneting under normal cheese-making conditions is <80%, despite partial estimation of total macropeptide release in most studies. Estimating the % fraction of hydrolysed κ-CN in the whole sample by mass balance (initial minus the residual amounts measured e.g., by Reverse-phase high performance liquid chromatography (RP-HPLC)) would be the most accurate approach. The accuracy of measuring the degree of κ-CN hydrolysis has implications on the precision of the data in relation to κ-CN’s partitioning. Understanding and controlling the partitioning of κ- and other CNs (αs1-, αs2- and β- CNs), fat (in full–low fat cheeses) and rennet activity (RA) during cheese-making is essential for preserving the intrinsic cheese quality and improving curd structure, yield and whey quality. Partitioning is mainly influenced by milk pre-treatments and renneting conditions. Regression analysis of existing data on Cheddar cheese suggests that partitioning of CN and fat, cheese yield and whey composition could be improved by standardising milk CN and fat to both optimum levels and ratio. However, while studies on CN partitioning are many, information on partitioning of individual CNs is very limited. What drives partitioning of both CNs and RA is also not well understood, and the clotting assay for quantifying RA partitioning is very subjective. Therefore, experimental studies evaluating the effect of some reconstituted skim milk pre-treatments (standardisation (8–26% total solids (TS), pre-salting (0–3 M, NaCl) and pre-acidification (pH 5–7)) were designed to address some of these gaps. First, a small amplitude oscillatory method was developed to improve the milk clotting assay. This involves the creation of a calibration curve of objective rennet coagulation time (RCT) vs. the inverse of RA, requiring one spike level only to estimate the unknown residual RA in whey. Interestingly, the concentration of residual RA (International milk clotting units, IMCU/mL whey) appeared to be the same as the initial concentration added to milk. A common RP-HPLC method, modified for whey samples, confirmed the finding which became the basis for developing a general model to estimate RA partitioning. The modified RP-HPLC method was used in a second study examining the effect of reconstituted skim milk concentration under both natural and slight acidification conditions aiming to reveal the driving force. For all milk protein levels, findings were consistent with the first study despite linear positive relationships between protein or moisture and RA retentions, indicating that RA partitioning is not controlled by enzymes-CNs associations. In the third and fourth studies, the impact of reconstituted skim milk concentration on partitioning of individual CNs and curd structure under natural and slight acidification conditions was examined. Partitioning was assessed by RP-HPLC and Sodium dodecyl sulphate polyacrylamide gel electrophoresis (SDS-PAGE), whereas curd structure was tested by performing dynamic rheometry and light scattering measurements. Low levels of κ- and αs2- CNs only were found in whey and were attributed to heat-induced complexation of these CNs and whey proteins as confirmed by SDS-PAGE. Their concentrations in whey increased with increasing milk TS whereas partitioning into whey decreased. This could be explained partly by increased hydrophobic interactions and restricted rearrangements during curd aging as revealed in the fourth study. In a fifth study, the impact of pre-acidification and pre-salting of reconstituted skim milk was also assessed by RP-HPLC and SDS-PAGE, and as per study three, small amounts of κ- & αs2- CNs (<11%) were detected in rennet whey from unsalted or salted milk samples with ionic strength ≤ 0.9 M, NaCl. Greatest losses of all CNs occurred at 3 M, NaCl when pH was adjusted to 7 and 5.5, and for this type of milk, pH 6 was found to be optimal. Varying levels of para-κ-CN found in rennet whey from pre-salted milk demonstrate that losses of CNs into rennet whey were caused by an ionic strength- and pH- dependent dissociating effect of NaCl on CN micelles, inducing further κ-/αs2-CNs-whey protein complexation, and inhibitory effect on para-CN aggregation. In conclusion, current research advances the understanding the effect of pretreatments of reconstituted skim milk in relation to partitioning of individual CNs and RA during cheese making. Findings could be particularly useful for cheese yield prediction modelling and quality control of both cheese and whey co-products. While CN-CN interactions appeared the driving force for RA partitioning, their role in relation to individual CNs partitioning was partially examined, thus, a full understanding remains to be established in future studies.