Abstract
Milk fat plays a vital role in determining the flavour of cheese. The milk fat globule (MFG) consists of a triglyceride (TG) core and a three-layer membrane termed the milk fat globule membrane (MFGM). The TGs within the MFG are the major lipid constituents in milk and they have been the subject of numerous investigations on the role of milk fat in flavour development in cheese. In cheese, the MFGM is known to be an additional source of lipids, mainly polar lipids (PLs) and specific proteins and enzymes like xanthine oxidase (XO). However, research investigating the influence of MFGM components, other than TG, on the flavour profile of cheese is limited. XO, the main enzyme in the MFGM, is capable of catalysing the oxidation and reduction of a wide range of substrates (aldehydes) and therefore it has been speculated that it may play a role in determining the oxidation-reduction potential (Eh) of cheese. This is of interest as it has been hypothesised that a negative Eh is necessary for the correct formation and stability of flavour in cheese. Thus, the activity and availability of the redox enzymes (such as XO) in the MFGM, may change the volatile compound profile of ripened cheese through its effect on the oxidation-reduction potential (Eh) of cheese products. This thesis firstly aimed to understand the relationship between the structure and composition of the MFGM and XO activity, and their impact on the Eh in recombined emulsion systems. Secondly, the thesis explored the role of MFGM composition, XO activity and Eh on the volatile profile of ripened model Cheddar cheese samples.
Buttermilk, was produced using various churning conditions at different temperatures (10, 15 or 20°C) and pH values (5.5 or 6.6) and used for MFGM isolation (Chapter 3). The isolated MFGM fractions were characterised for protein composition, and XO content. Milkfat (5%) emulsions (containing 2% MFGM) were prepared from the MFGM isolates and examined for XO activity and Eh. A greater proportion of non-membrane proteins, e.g., caseins and whey proteins, were found at higher temperatures and cream pH, and the XO content decreased significantly under these conditions. XO activity was greater at lower temperatures and pH, yielding a more positive Eh value for the emulsions. Storage at 4°C for 14 days significantly reduced the Eh of the emulsions to between -320 mV and -580 mV for low pH/low temperature, and high pH/high temperature emulsions, respectively.
Industrial sources of MFGM material, namely buttermilk powder (BMP), β-serum, and α-serum, were also used for MFGM isolation and emulsion preparation (Chapter 4). Caseins were the dominant proteins in the MFGM fractions isolated from BMP, whereas fractions from α- and β-serum contained higher amounts of the MFGM proteins. The XO content and activity was >70-fold and >700-fold higher in α- and β-serum samples, respectively, compared to the BMP fraction. The Eh values of the recombined emulsions were highest for α-serum (196 mV), and β-serum (169 mV), followed by BMP (131 mV). These positive values contrasted to the highly negative Eh of the emulsions characterised in Chapter 3 (-580 mV). From this study, no consistent trend was observed between the XO content, activity and the Eh of the emulsion system.
The polar lipid (PL) profiles of the MFGM fractions isolated from fresh raw cream and the commercial sources were subsequently investigated (Chapter 5). Decreasing the temperature and pH during churning increased the total PLs in the MFGM fractions, with temperature having a greater effect at pH 5.5 compared to pH 6.6, and pH have a greater effect as the temperature decreased from 20 to 15 to 10°C. The optimal preparative conditions to maximize the total PL content was determined to be churning at pH 5.5 and 10°C. BMP contained the lowest amount of total PL amongst all fractions. Phosphatidylinositol was the primary polar lipid component in all MFGM fractions, followed by phosphatidylethanolamine, phosphatidylserine, phosphatidylcholine, and sphingomyelin. The results provide evidence to support a trilayer MFGM model, where phosphatidylinositol and phosphatidylserine are located in the inner monolayer.
Finally the impact of MFGM composition and structure, and XO activity on Eh and the volatile profile of ripened model Cheddar cheese samples was investigated (Chapter 6). Recombined emulsions produced using the MFGM fractions isolated in Chapter 3 (at pH 6.6, and 20°C) and Chapter 4 (all fractions) were combined with reconstituted skim milk for Cheddar cheese manufacturing. Two reference cheese samples, one made using cream and skim milk (native MFGM) and one made with Tween 80 were used to evaluate the effect of the MFGM and its structure and composition on flavour of cheese. All six cheeses were ripened for 6 months at 8°C. The composition and structure of the recombined MFGs did not affect the redox potential of the cheese medium. A higher XO activity favoured the production of carboxylic acids, particularly short-chain fatty acids such as acetic, butanoic and hexanoic acid, whereas rearrangement of the MFGM layer (removing MFGM from the lipid interface and re-emulsifying it back onto the surface of reformed fat globules) negatively affected the acetoin / diacetyl production pathway. The rearrangement of the MFGM layer, or the XO activity did not affect the production of none of the detected esters during ripening.
The MFGM had a significant effect on development of volatile compounds in cheese, as the Ref 2-cheese containing recombined MFGs with Tween 80 had significantly lower concentration of the important volatile groups of mainly carboxylic acids, alcohol, and ketones.
In conclusion, the results from this work have provided new insights into the influence of processing on the MFGM structure and its subsequent effect on the flavour profile of cheese. The knowledge taken from this study can be used as a platform to manufacture food products or ingredients with beneficial health effects or low-fat cheese products with enhanced flavor profiles.