The temperature-dependent dissociation of /?-casein from the casein micelles of milk and from the soluble casein complexes of colloidal phosphate-free (CPF) milk was investigated by high-speed centrifugation and gel-filtration. The percentage of the total casein in supernatants prepared by high-speed centrifugation of mid-lactation milks increased from approximately 6 to 15% on cooling the milks from 30 to 5 °C; y?-casein accounted for about 46 % of this increase, while a 8 -and /c-casein constituted 30 and 23%, respectively. On gel-filtration both of skimmilk and CPF milk on Sepharose 2B at 0, 2, 5, 10 and 25 C C, maximum amounts of free /?-casein (c. 60 % of total) were obtained at 5 °C. The remainder of the yff-casein appeared to be more strongly bound to the a s -and /c-casein and may be involved in the internal cohesion of casein micelles. The free /?-casein of both milk preparations appeared to be in equilibrium with the bound /?-casein. On Sephadex G-200 columns at 5 °C, approximately 5 and 60 % of the yS-casein of skim-milk and CPF milk, respectively, was eluted in the free form in the expected position for a globular protein of molecular weight about 200000. At low temperatures, particularly at 5 °C, colloidal phosphate appeared to play an integrating role in the association of over half the total y?-casein with the other casein components of native micelles. However, when the equilibrium between micellar and free /?-casein was disturbed by gel-filtration on Sepharose 2B, the presence of colloidal phosphate did not prevent the release of most of the /?-casein from casein micelles. Some problems encountered in the use of densitometry for the estimation of individual caseins on electropherograms are described.It has been suggested (Sullivan et al. 1955) that /?-casein is removed from casein micelles when skim-milk is chilled. More recently Rose (1968) demonstrated that /?-casein constituted about 55 % of the total increase in serum casein obtained by storing milk overnight at 4 °C.In the present investigation the relative amounts of free and bound /?-casein present in milk and in CPF milk at various temperatures between 0 and 30 °C were determined by high-speed centrifugation and gel-filtration. The hypothesis (cf. Rose, 1969) that /?-casein constitutes the internal lattice or framework of casein micelles was further investigated by studying the effect of 23-2

Introduction Extent of lipolysis Lipolysis in milk Protection of milk fat globules Induced lipolysis-kinetic aspects Spontaneous lipolysis Causes of induced lipolysis Causes of spontaneous lipolysis Blood constituents Late lactation Nutrition Low milk yield, hormonal control CONTENTS

1. Sephadex G-100 and G-200 gel-filtration columns were calibrated for molecular-weight estimation with proteins of known molecular weights, and used to study the composition of several lipase or esterase preparations. 2. Enzymes from cow's milk, rat adipose tissue and pig pancreas were detected in the column effluents by their ability to liberate free acid from emulsified tributyrin at pH 8.5. 3. Four tributyrinases were detected in preparations from individual cow's milks. Molecular weights 62000, 75000 and 112000 were estimated for three of them, but although the fourth may be of unusually low molecular weight an estimate was not possible. 4. Extracts of rat adipose tissue apparently contained six tributyrinases (molecular weights 39000, 47000, 55000, 68000, 75000 and 200000) but the relative amounts of these enzymes varied widely from rat to rat. 5. Tributyrinase activity in juice expressed from pig pancreatic tissue was due mainly to one enzyme (molecular weight 42000). On the other hand, activity in extracts of acetone-dried pancreas was confined to material of molecular weight > 10(6), which may be an aggregated form of the lower-molecular-weight enzyme. 6. Activity in fractionated wheat-germ extracts was assayed with emulsified triacetin substrate, and was evidently due to one enzyme (molecular weight 51000). 7. Some problems arising in the application of gel filtration to the study of lipase-esterase systems were indicated.

Skim milks containing sodium chloride (0.75m) were centrifuged at 80000g for 2hr. and portions of the supernatants were submitted to gel filtration on columns of Sephadex G-200. Enzymes in the effluent fractions were assayed titrimetrically for their hydrolytic activities towards tributyrin, triolein and milk-fat emulsions, and triacetin solution. Summation of the measurements gave ratios of activities towards the various substrates similar to those of the original skim milks. Although only partial separation was obtained, five enzymes appeared to be present. They showed some differences in substrate specificity, but all appeared to be lipases in that they hydrolysed the emulsified substrates more rapidly than the dissolved triacetin.

1. The tributyrinases in milk are mainly associated with casein micelles. Dilution or addition of sodium chloride increases the enzyme activity, probably by dissociating the micelle-tributyrinase complexes. 2. Tributyrinase activities of milks activated by dilution and sodium chloride addition were in the range 0.2-1.7muequiv. of acid liberated/ml. of milk/min. from tributyrin emulsion at pH8.5 and 25 degrees . The enzymes have a bivalent-cation requirement for full activity and are rather unstable when separated from casein. 3. Ultracentrifugation of skim milks containing sodium chloride (0.75m) gave preparations low in casein but containing about 70% of the milk tributyrinases. The tributyrinases in such preparations appear to be bound in complexes of molecular weight about 350000. Dilution may result in dissociation to give the free enzymes. 4. Pancreatic lipase also formed complexes with casein micelles, but wheat-germ esterase, xanthine oxidase, milk alkaline phosphatase and other enzymes did not.