Massey Documents by Type

Permanent URI for this communityhttps://mro.massey.ac.nz/handle/10179/294

Browse

Has files: YesSubject: search.filters.subject.Casein

Search Results

Now showing 1 - 10 of 23
  • Thumbnail Image
    Item
    Heat-induced interactions between microfluidized hemp protein particles and caseins or whey proteins
    (Elsevier Ltd, 2025-01) Ma S; Ye A; Singh H; Acevedo-Fani A
    The rising demand for sustainable proteins leads to increased interest in plant proteins like hemp protein (HP). However, commercial HP's poor functionality, including heat aggregation, limit its use. This study explored the heat-induced interactions of hemp protein particles (HPPs) with milk proteins, specifically whey proteins and caseins. Using various analysis techniques-static light scattering, TEM, SDS electrophoresis, surface hydrophobicity, and free sulfhydryl content-results showed that co-heating HPPs with whey protein isolate (WPI) or sodium caseinate (NaCN) at 95 °C for 20 min reduced HPPs aggregation. HPPs/WPI particles had a d4,3 of ∼3.8 μm, while HPPs/NaCN were ∼1.9 μm, compared to ∼27.5 μm for HPPs alone. SDS-PAGE indicated that whey proteins irreversibly bound to HPPs, through disulfide bonds, whereas casein bound reversibly, possibly involving the chaperone-like property of casein. This study proposes possible mechanisms by which HPPs interact with milk proteins and impact protein aggregation. This may provide opportunities for developing hybrid protein microparticles
  • Thumbnail Image
    Item
    Interactions between hemp globulins and dairy proteins : a thesis presented in partial fulfilment of the requirements for the degree of Doctor of Philosophy in Food Technology at Riddet Institute, Massey University, Palmerston North, New Zealand
    (Massey University, 2021) Chuang, Chih-Chieh
    Industrial Hemp (Cannabis sativa L.) is a sustainable protein source and is easily digested. However, the hemp globulins (HG), which constitute around 70% of total hemp seed protein, have low solubility in water at neutral pH. The insolubility of HG limits its usage in many food systems. This work explored the interactions between hemp globulins (HG) and dairy proteins, and aimed to increase the functionality of HG. HG was extracted with a mild salt-extraction and heat treatment was avoided, so the native structure of HG was preserved. The composition of HG was studied and a phase diagram of HG solubilisation was obtained regarding pH and ionic strength. Two methods were used to increase the solubility/colloidal stability of HG by introducing interactions between HG and sodium caseinate (SC). The first method is by heating HG and SC together at 90 °C and the ionic strength of 0.5 M. SC exhibited the chaperone-like activity and inhibited the formation of large aggregates of HG. The addition of SC did not change the denaturation kinetics of HG, but rather changed its aggregation pattern. The second method is by pH-cycling. HG and SC formed colloidally stable nanoparticles (Z-average diameter ≈ 130 nm) after adjusting the pH to 12, reacted for 1 hour and neutralised back to pH 7. The solubility of HG increased from ~20% to > 80% after the pH-cycling. The mechanisms and molecular interactions of both processing methods (heating and pH-cycling) were proposed. During heating, SC interacted with HG via hydrophobic interactions and the aggregation regime of HG changed from diffusion-limited cluster aggregation to reaction-limited cluster aggregation, while the kinetics of HG denaturation was unaffected. During the pH-cycling, hydrogen bond was one of the driving forces for assembly of HG|SC nanoparticles. HG partially unfolded at pH 12 and interacted with caseins during the neutralisation and the stable HG–SC nanoparticles were formed. The pH-cycled HG-SC nanoparticles can be used to make Pickering emulsions. Concentrated emulsions were prepared, and the rheological properties of emulsions during storage can be tuned by controlling HG:SC ratio in the HG|SC nanoparticles, i.e. emulsions became more solid-like when there was more HG in the HG-SC nanoparticles. The internal structure and interactions within the emulsions were evaluated by fitting frequency sweep test data according to a co-operative theory of flow. The results suggested that the solid-like emulsion resulted from stronger short-range interactions between flocs of oil droplets, which developed during storage when there was more HG in the HG–SC nanoparticles. In conclusion, the findings in this study advanced our understanding of the interactions between plant seed globulins and caseins during processing. Such knowledge will help to increase the functionalities of plant proteins by mixing plant and dairy proteins.
  • Thumbnail Image
    Item
    The effects of pH-stat long-term lactic acid bacterial activity prior to curd formation on the development of cheese structure in a fat free model cheese : a thesis presented in partial fulfilment of the requirements for the degree of Doctor of Philosophy (PhD) in Food Technology at Massey University, Manawatu, New Zealand
    (Massey University, 2018) Khodamoradi Dashtaki, Abbas
    Cheese ripening is an important step in most cheese production practices during which a tasteless fresh cheese is converted to a tasty and flavourful product with specific textural attributes. However, the complexity of its composition (pH, solubilisation of calcium from the colloidal casein proteins, salt concentration, acid production rate, indigenous and added enzymes, residual activity of the enzymes etc.) coupled with the length of time associated with manufacturing which in some cases can be in excess of two years has made it a complicated area of study. The influences of the composition and process contributors are confounded and it is impossible to connect the impact of one particular parameter on cheese making steps and quality attributes. pH has proven to be an important influential factor to influence the extent of effects of other parameters with significant influence on other ruling parameters in milk, curd and cheese. Proteolysis during ripening is the most important physicochemical pathway to define the quality of cheese. One of the major factors governing cheese ripening reactions is the starter bacteria. This study has aimed to characterise the effects of starter bacteria activity on curd formation and resultant cheese textural attributes of the long fermented cheesemilk. By developing a pH-stat system, long fermentations carried out to assess the proteolytic activity of selected starter lactic acid bacteria (LAB) on a milk based medium before rennet addition. It was attempted to assess the degree of hydrolysis of cheesemilk through extended bacterial fermentation, conducted under pH-stat conditions, prior to curd formation. The effects of the bacterial activity on casein proteins during pH-stat long term (PSLT) fermentations were evaluated by assessing proteolysis index from pH4.6 soluble nitrogen as a fraction of total nitrogen (pH4.6SN/TN). The proteolysis of proteins during PSLT was further assessed by doing reverse-phase high performance liquid chromatography (RP-HPLC) on 70%Ethanol soluble (70%EtOHS) and insoluble (70%EtOHI) fractions of pH4.6 soluble fraction of the samples. The effects of PLST fermentation on formation of small-size peptides were assessed by quadrupole time-of-flight mass spectrometry (Q-ToF MS) on the 70%EtOHS fraction. The effect of PLST fermentation on ‘depth of proteolysis’ during cheese ripening were assessed by analysing the quantity of free amino acid (FAA) formed in resultant cheese after 12 months storage at 4°C. The impact of PLST fermentation on gel formation attributes were assessed by doing dynamic law amplitude oscillatory rheometry (DLAOR). The consequent effects on resultant cheese texture were evaluated using texture profile analysis (TPA). The impact of PSLT on microstructure were assessed by confocal laser scanning microscopy (CLSM). The results provided evidence for the adequacy of developed fermentation to conduct PSLT with reproducible results. High correlation between the parameters of the PSLT fermentation system were obtained. The proteolysis index measured from the PSLT fermentations with different durations showed evidences on the significance of LAB proteolytic system on cheese milk prior to curd formation. The proteolysis index for the longest fermentation prior to curd formation was 5% which was comparable to day one cheese proteolysis index, in presence of rennet, in most cheeses varieties. Peptide profiling of the 70%EtOHS and 70%EtOHI sub-fractions of pH4.6S showed significant (p<0.05) effects arising from PSLT fermentations. Analysis of FAA of ripened cheese also showed a significant increase (p<0.05) in the samples with longer PSLT (20 times increase in total free amino acids compared to non-fermented treatment) fermentations. The differences in gelation behaviour of the sample and textural attributes of cheese and microstructure of final cheese were connected to the extent of proteolytic activity of LAB during PSLT fermentations. The hardness of cheese significantly (p<0.05) decreased (up to ~60%) by increasing fermentation duration over the studied timescale.
  • Thumbnail Image
    Item
    Application of lysozyme in formation of multilayer emulsion containing caseinate : Master of Food Technology, Massey University, Palmerston North, New Zealand
    (Massey University, 2018) Chen, Zhenghao
    The properties of oil-in-water emulsions stabilized by caseinate-lysozyme complexes were investigated. Complexes were prepared by the mixture of 0.8 wt.% sodium caseinate with various amounts of lysozyme (0-0.5% W/W) at neutral pH. The emulsions formed by the serum solutions were not stable at low lysozyme concentrations (0-0.2 % W/W), but were stable at high concentrations (0.3-0.5 % W/W). The effects of lysozyme on caseinate-stabilized O/W emulsions were studied. Multilayer emulsions (containing 0.4 wt.% caseinate and 0-0.5 wt.% lysozyme) were created by mixing a primary caseinate stabilized emulsion with lysozyme solutions. The emulsions were evaluated at pH 3.3 and 6.8. At neutral pH, the emulsions were stable in the initial presence of 0.1 wt.% lysozyme due to bridging flocculation, but unstable at high lysozyme concentration of 0.1-0.5 wt.% . In acidified emulsions, lysozyme had no effect on caseinate-stabilized emulsions. Therefore, at neutral pH, complexes formed with high caseinate-lysozyme ratio could not create stable emulsions. On the other hand, caseinate-stabilized emulsions could only stay stable when the lysozyme to caseinate weight ratio was 1:2.
  • Thumbnail Image
    Item
    Dairy trade between New Zealand and the Republic of Korea, with special reference to casein : thesis presented in partial fulfilment of the requirements of Master of Business Studies at Massey University
    (Massey University, 1991) Nixon, Christopher Gerard
    The purposes of this thesis were firstly, to identify and describe Korean non-tariff and tariff barriers for casein and other dairy products and secondly, to quantify how much New Zealand could gain from a liberalisation of the Korean casein trade. To do this a one-product five-nation quadratic programming model was formulated. How Korea has become a major trading power through industrialisation, while heavily protecting agriculture is described. Measures of protection, the pressure to liberalise and the Japanese liberalisation experience are discussed. Casein was chosen because it is the single biggest dairy commodity exported to Korea from New Zealand. The model consists of demand for casein from the major consuming countries (America, Korea and Japan) and fixed supply from the two major suppliers (New Zealand and the European Community). Various scenarios are run to gauge the effect of a drop in tariff rates in Korea and Japan and at various levels of European production. The study concludes with the recommendation to continue pushing for liberalisation in multilateral and bilateral negotiations particularly with the European Community.
  • Thumbnail Image
    Item
    Development of a lactic casein based savoury flavour product : a thesis presented in partial fulfilment of the requirement for the degree of Master in Food Technology at Massey University, Manawatū, New Zealand
    (Massey University, 2016) Chen, Ting
    The aim of this project was to develop a casein-based hydrolysate formulation that has higher savoury flavour and is more cost effective than an existing commercial savoury hydrolysate. From the literature review, bovine casein protein has the most savoury flavour potential of all proteins due to its high glutamic acid and glutamine content. The symbol of savoury flavour is cheese which is made from casein protein in the western world. The main reaction resulting in cheese savoury flavour development is proteolysis where casein protein breaks down to peptides by protease and free amino acids by peptidases. Two different systems were designed to be based on those reactions in order to generate maximum free glutamic acid during the experiments. The enzyme substrate was a 10% lactic sodium caseinate slurry, which is the foundation of the experiments. With the first system, an enzyme preparation with protease functions was added first and followed by an enzyme preparation with peptidase functions. With the second system only one enzyme preparation with both protease and peptidase activity was added for each trial. From the results, it was found that none of the enzyme combinations from either system were able to achieve the same amount of free glutamic acid as the existing commercial product (31.95 mg/g of protein) within 24 hours. However, multiple options would have had equivalent free glutamic acid if the free glutamine content could be converted to L-glutamic acid using a glutaminase. Flavorzyme 1000L from system one was selected to be the option combining with glutaminase based on its cost, microbiology and chemistry process results. Two different dosages (0.25% and 0.5%) of Glutaminase C100SD were trialled with 2% of Flavorzyme 1000L. From the degree of hydrolysis, free amino acid content, molecular weight profile and residual glutamine results, there were almost no difference between the two trials. The final formulation of Flavorzyme 2% and 0.25% Glutaminase C100SD had 48% more free glutamic acid than the existing commercial control. It also achieved a 33% ingredient cost reduction. Most importantly, the final formulation resulted in a 22.5% final ingredient cost reduction per kilogram based on the same commercial cost model as the control. An informal sensory panel indicated that the new savoury hydrolysate was more savoury and less bitter than the existing commercial control.
  • Thumbnail Image
    Item
    Goat and cow casein derived ingredients and their interactions with iron : a thesis presented in partial fulfilment of the requirements for the degree of Doctor of Philosophy in Food Technology, Massey University, Palmerston North, New Zealand
    (Massey University, 2017) Smialowska, Alice Małgorzata
    The objective of this study was to gain a fundamental understanding of how goat casein micelles and the products of casein proteins behave when fortified with iron. Iron fortified skim milk was characterised by analysing the mass balance of micellar/non micellar fractions, chemical changes, micellar size changes and internal structure. Two treatments were examined to determine where in the processing line the addition of iron might best be added to a milk system. On average, at least 72% of the iron is bound to the micellar phase across the treatments and iron concentrations. Small angle X-ray scattering (SAXS) indicated that internal changes, mainly at the location of the colloidal calcium phosphate, occurred with iron addition. Casein was extracted from goat milk using isoelectric precipitation however the extraction was more difficult than using cow milk. Iron fortification of the caseinates resulted in a tendency for oxidation and precipitation of the proteins to occur causing the formation of large aggregates. The caseinates could not stabilise the same amounts of iron to that of an intact casein micelle. Phosphopeptides were isolated by adding calcium and ethanol to caseinate digests. There was an increase in serine, glutamic acid and isoleucine residues compared to caseinate. There was an increase in phosphorus from 7.8 ± 0.3 mg P/ g solids to 45.4 ± 2.4 mg P/ g solids in the isolate. The phosphopeptides were composed of smaller, more hydrophilic peptides compared to the full digest prior to precipitation. Ferrous sulfate was then investigated for use as the precipitant, instead of calcium. The peptides produced similar trends in terms of amino acid profile changes, phosphorus concentration increase and yield. Immobilised metal affinity chromatography was also investigated however this had a low throughput that may not be effective at process scale. The effect of heating, cooling, ionic strength of the solution, holding time, iron loading, processing order and use in a model milk system were investigated to simulate potential industrial processing conditions using the calcium - extracted phosphopeptides. It was found that goat peptide isolates were able to bind 54.4 ± 0.5 mg Fe/ g protein compared to goat milk of 4.3 ± 0.1 mg Fe/ g protein. The optimal conditions for binding were found to be at pH 6.7 in a low ionic strength solution, around 37 oC. There was a potential synergistic effect of adding the peptides to milk in terms of iron binding capacity. There were few differences in the amount of iron that could be bound comparing cow and goat derived phosphopeptides under the tested conditions. The oxidation potential of ingredients was determined using malondialdehyde (MDA) as an oxidation product marker. There was a reduction in oxidation when iron was bound to milk or peptides compared to free ferrous sulfate in solution with intact goat milk performing the best producing 0.46 ± 0.04 μg MDA/mL after 3 days at 30 oC compared to the blank of 1.25 ± 0.16 μg MDA/mL. The goat peptides produced non-significantly different levels of MDA compared to the blank containing no ferrous sulfate. Caco-2 cell lines are a way of approximating how systems may function in an intestine in terms of nutrient absorption. Iron absorption was improved in the order of casein hydrolysates > caseinate > skim milk for goat milk. In contrast, cow milk appeared to perform better without any modifications to the proteins. On an equal iron filtrate basis after the digestion and intestinal phase, calcium- precipitated goat phosphopeptides produced a response of 9.64 ± 0.94 ng ferritin/ nM iron. This response was greater than all other treatments with the exception of goat milk fortified with 5 mM iron and ascorbic acid with 12.30 ± 1.23 ng ferritin/ nM iron. This work covers a wide range of milk products and iron interactions and has helped to build a fundamental understanding of goat milk protein functionality. The underpinning considerations to a manufacturing setting may allow further development of large scale ingredient production for the improved stability of iron fortified systems.
  • Thumbnail Image
    Item
    The action of rennin on B-casein : a thesis presented in partial fulfilment of the requirements for the degree of Master of Science in Chemistry at Massey University
    (Massey University, 1970) Mills, Owen Edmund
    A study was made of the action of the enzyme rennin on β-casein. Hydrolysis of β-casein initially at a single sensitive bond under controlled conditions of temperature, pH and relative enzyme and substrate concentrations, formed the basis of the investigation. Information on the hydrolysis of this sensitive bond was gained from the isolation of a small peptide produced and from a study of the effect of several parameters on the rate of hydrolysis. Evidence obtained from electrophoresis and gel filtration allowed the assumption that attack on the sensitive bond resulted in a macropeptide and a small peptide of molecular weight about 2000. The small peptide was isolated and partially characterised. As a result it appears that the small peptide is derived from the C-terminal end of the β-casein molecule. A polyacrylamide electrophoresis technique was used to study the effect of ionic strength and calcium ions on the rate of hydrolysis and the rate of appearance and disappearance of degradation products at 10°, 25° and 37°C. It was found that an increase in ionic strength retarded the reaction and the addition of calcium ions at a constant ionic strength further retarded the reaction. Also, the rate of appearance and disappearance of degradation products was found to increase with increasing temperature. A development of the polyacrylamide technique into a quantitative one enabled the determination of the Michaelis constant at pH 6.50 and 37°C for the rennin hydrolysis of β-casein as 9-59 g/1. This technique was also used to study the rate of hydrolysis at pH 6.12, 6.50 and 6.94 where an optimum rate occurred at pH 6.50. Finally, assuming that the small peptide is derived from the C-terminal end of the β-casein molecule and allowing for the sequential degradation elucidated by the temperature studies, alternative courses of rennin degradation of the β-casein molecule have been proposed.
  • Thumbnail Image
    Item
    Prevention of plasmin-induced hydrolysis of caseins : a thesis presented in partial fulfilment of the requirements of the degree of Doctor of Philosophy in Food Technology at Massey University, Palmerston North, New Zealand
    (Massey University, 2014) Bhatt, Hemang
    Bovine plasmin is a proteolytic enzyme that is naturally present in milk. Plasmin can have a detrimental impact on product quality including proteolysis, age-gelation and bitterness. The activity of plasmin is difficult to control as its precursor, plasminogen, and its activators can survive severe heat treatments such as ultra-high-temperature processing. The aim of this work was to understand and control the plasmin-induced hydrolysis of caseins in milk systems. A sequential approach was used. In the first stage, the effect of substrate modification on plasmin-induced hydrolysis in a pure ß-casein model system was studied; this allowed us to propose a control mechanism to limit the availability of the substrate by protein modification. In the second stage, different protein modifications were applied to a real milk system. In the analysis of this system, the casein micelle structure, whey protein denaturation and whey protein association with the casein micelle were considered. The final stage investigated plasmin-induced dissociation of casein micelles in real milk systems to understand the effect of plasmin activity on gelation and sedimentation in heat-treated milks. Modification of lysine residues on the protein decreased plasmin-induced hydrolysis. Lactosylation had a greater effect than succinylation and transglutamination at the same level of lysine modification. A mechanism for this phenomenon was proposed. Lactosylation involves the attachment of lactose and, in advanced stages, cross-linking, thus modifying lysine and making it unrecognisable to plasmin; in addition, the cross-linking may affect the release of plasmin-generated peptides. Transglutamination also modifies lysine by cross-linking and has a similar effect to lactosylation, but to a lesser extent. In contrast, succinylation modifies the charge associated with lysine, making it unrecognisable to plasmin. Collectively, this knowledge can be used to make protein resistant to plasmin activity. The combined effect of micellar structure and protein modification on plasmin activity was also studied. Calcium chelation and dissociation of the casein micelle increased plasmin activity because of reduced steric hindrance, which made the protein more readily available to plasmin. In contrast, succinylation decreased plasmin activity, which could be attributed to the formation of succinyl-lysine rendering ß-casein unrecognisable to the substrate-binding pocket of plasmin, resulting in a decrease in hydrolysis with an increase in modification. These results indicated the importance of the casein micelle structure as a tool for controlling the activity of plasmin on milk proteins in food systems. The effect of high heat treatment on plasmin-induced hydrolysis was also investigated. A high-heat-treated skim milk (120°C/15 min) was found to have greater resistance to plasmin activity than non-heated skim milk. Both whey protein association with the casein micelles and lactosylation decreased the availability of protein to plasmin. Whey-protein-free milk was the most plasmin resistant, followed by skim milk and lactose-free milk. Collectively, these results suggest that lactosylation plays a more significant role than whey protein association with the casein micelles in making protein resistant to plasmin activity. The plasmin-induced dissociation of the casein micelle was explored by identifying peptide release from the micelle. Upon plasmin-induced hydrolysis of the casein micelle, hydrophilic peptides, i.e. proteose peptones, were the first to dissociate from the casein micelle, followed by hydrophobic peptides, which had dissociation patterns that were identical to those of ?-casein. This suggests that the release of ?-casein from the micelle is too slow to cause gelation. Extensive plasmin-induced hydrolysis of the casein micelle leads to sedimentation in heat-treated milk because of the formation of ß-lactoglobulin–?-casein complexes and their aggregation with hydrolysed hydrophobic peptides. Overall, the results of the present study showed that casein modification can be useful in controlling plasmin activity and has developed our understanding of the plasmin-induced dissociation of casein micelles. Further research work is needed to understand the mechanism of plasmin’s selective hydrolysis pattern and the structural aspects of the substrate-binding pocket of plasmin. Studies on casein micelle dissociation separately and in conjunction with physicochemical changes during storage could be useful in further understanding the phenomenon of age gelation.
  • Thumbnail Image
    Item
    Heat-induced interactions of [beta]-lactoglobulin, [alpha]-lactalbumin and casein micelles : a thesis presented in partial fulfilment of the requirements for the degree of Masters of Education in Food Technology at Massey University
    (Massey University, 1996) Chiweshe, Martha Chogugudza
    The denaturation and aggregation of β-lactoglobulin and α-lactalbumin were studied in the following mixtures, designed to simulate the protein concentrations and ionic environment in normal skim milk. 1. β-lactoglobulin (0.4% w/v), 2. α-lactalbumin (0.15% w/v), 3. β-lactoglobulin (0.4% w/v) and casein micelles (~ 2.6% w/v), 4. α-lactalbumin (0.15% w/v) and casein micelles (~ 2.6% w/v), 5. β-lactoglobulin (0.4% w/v) and α-lactalbumin (0.15% w/v) and 6. β-lactoglobulin (0.4% w/v), α-lactalbumin (0.15% w/v) and casein micelles (~ 2.6% w/v) Proteins were dissolved in SMUF, pH 6.7, and heated at 80 and 95°C for various times and centrifuged at 100,000 g for 60 min. The supernatants and pellets obtained were analysed using gel electrophoresis under non-dissociating (Native-PAGE in the absence of dissociating and reducing agents), dissociating but non-reducing (SDSNR-PAGE) and dissociating and reducing conditions (SDSR-PAGE). When β-lactoglobulin was heated alone and examined by native-PAGE, the quantity of native protein decreased with increasing heating time at 80°C. Addition of α-lactalbumin to the β-lactoglobulin solution increased the loss of β-lactoglobulin during the initial stages of heating. Addition of casein micelles to the β-lactoglobulin solution markedly increased the loss of native β-lactoglobulin throughout the heating period. The loss of β-lactoglobulin from the mixture containing β-lactoglobulin, α-lactalbumin and casein micelles was similar to that from the mixture of β-lactoglobulin and casein micelles. The loss of β-lactoglobulin from these protein mixtures could be described by second-order reaction kinetics. Heating these mixtures at 95°C caused very rapid loss of native β-lactoglobulin, but the effects of the addition of casein micelles and α-lactalbumin were generally similar to those observed at 80°C. When α-lactalbumin was heated at 80°C either alone or in the presence of casein micelles, there was only a slight loss of the native α-lactalbumin. However the corresponding losses of native α-lactalbumin were considerable greater on heating at 95°C. At both temperatures, the addition of β-lactoglobulin increased the rate of loss of α-lactalbumin substantially. The addition of casein micelles to the mixture of α-lactalbumin and β-lactoglobulin had little further effect on the loss of native α-lactalbumin. The rates of loss of α-lactalbumin at 95°C in all mixtures could be adequately described by first-order kinetics. When β-lactoglobulin was heated either alone or in the presence of casein micelles and examined by SDSNR-PAGE, the loss of SDS-monomeric β-lactoglobulin was less than the loss of native β-lactoglobulin. In contrast, when α-lactalbumin was added to β-lactoglobulin or β-lactoglobulin and casein micelles mixture, the loss of SDS-monomeric β-lactoglobulin was comparable to that of native β-lactoglobulin. The difference between native and SDS-monomeric β-lactoglobulin represents aggregates that are linked by non-covalent (hydrophobic) interactions. Thus the protein mixtures containing α-lactalbumin, contain no or little non-covalently linked β-lactoglobulin aggregates, and consequently, all the β-lactoglobulin aggregates would be disulphide linked. The results for the loss of SDS-monomeric and native α-lactalbumin at 95°C showed that both non-covalent and disulphide-linked aggregates of α-lactalbumin were present in all the protein mixtures studied. When β-lactoglobulin solution was heated at 95°C, large aggregates were formed which could be sedimented at 100,000 g for 60 min. Addition of casein micelles to β-lactoglobulin solution caused greater sedimentation of β-lactoglobulin. Similar results were obtained when the mixture containing β-lactoglobulin, α-lactalbumin and casein micelles was heated at 95°C. In contrast, the mixture containing β-lactoglobulin and α-lactalbumin behaved in a similar manner to β-lactoglobulin alone. When α-lactalbumin was heated at 95°C alone or in the presence of casein micelles, it did not interact to form large sedimentable aggregates. However when β-lactoglobulin was added to the above protein solutions, there was a considerable increase in sedimentation of α-lactalbumin.