Heat Stability of Milk (2004)

Vol 57, No 2/3 May/August 2004 International Journal of Dairy Technology

111

SYMPOSIUM CONTRIBUTION

*Author for correspondence. E-mail: [email protected]

© 2004 Society of Dairy Technology

Blackwell Publishing, Ltd.

Heat stability of milk

HARJINDER SINGH

Riddet Centre, Massey University, Palmerston North, New Zealand

The heat stability of milk has been the subject of a considerable amount of research for about a century.This research has been aimed mainly at understanding the effects of compositional and processingfactors on heat stability and elucidating the mechanisms of protein coagulation. This paper provides anoverview of the factors that influence the pH dependence of the heat stability of normal and concentratedmilks. The principal heat-induced changes in the milk system that contribute to coagulation are discussed.Current knowledge of the mechanisms of heat coagulation in normal and concentrated milks is alsoreviewed.

Keywords

Casein micelle stability, Concentrated milk, Heat coagulation, Heat stability, Milk proteins,

Milk.

*Author for correspondence. E-mail: [email protected]

B A C K G R O U N D

The problems of regulating heat stability (that isthe relative resistance of milk to coagulation uponsterilization) appeared over a century ago in themanufacture of evaporated (condensed) milk. Theidea of commercially preserving milk by steriliz-ing dates back to 1856 when Gail Borden wasgranted patents in the United States and Englandfor ‘producing concentrated milk by evaporationin vacuum without addition of sugars and otherpreservatives’.

The commercial production of condensed milkincreased gradually during the First and SecondWorld Wars, and condensed milk became one ofthe major dairy products in the 1920s, because ofeasy transport and a long shelf life. The usual prob-lems faced in those days were that the milk gelledor coagulated during the heat treatment and exces-sive thickening of the product occurred duringstorage. These problems were controlled by carry-ing out various heat-stability tests on the raw milk,and by running pilot sterilization trials on samplesfrom each batch after the addition of variousamounts of sodium bicarbonate.

Between 1900 and 1960, most of the scientificresearch focused on solving heat coagulation prob-lems in concentrated (condensed) milk. Studies offactors affecting heat stability were considered tobe important because they could be used to predictwhether a given milk sample would coagulate afterit had been processed into product. Sommer andHart (1919, 1922) showed that mineral balancewas important, and that if a milk sample was tooacid (insufficient calcium and magnesium) or toobasic (insufficient phosphate and citrate), it wouldbe unstable. Heat stability could, according to the

theory, be restored to these milks by appropriatelybalancing them by adding ‘acids’ (CaCl

2

or HCl)to milks that were too ‘basic’ and vice versa. Thesalt balance theory was criticized by Rogers

et al

.(1921), who showed that the heat stability of con-densed milk could not be predicted from the pHand salt balance of the milk or indeed from theheat stability of the raw milk. There was certainlyno relationship between the heat stability of rawmilk and that of the condensed milk made from it.The effect of preheat treatment (i.e. forewarming)

−

heat treatment given to the milk prior toevaporation

−

was studied in detail in the 1940s(Webb and Bell 1942). Preheating is now a stand-ard commercial practice for the manufacture ofcondensed or evaporated milk.

From the 1960s, studies on heat stability shiftedfrom condensed milk to normal (unconcentrated)milk. This was around the time when the relevanceof pH to heat stability was fully revealed by thework of Rose (1961). In the 1970s, research con-centrated on the effects of processing and compo-sitional factors on the pH dependence of the heatcoagulation time of unconcentrated milk. Some ofthe more important findings include understandingthe roles of

β

-lactoglobulin and

κ

-casein, milksalts and urea in heat coagulation. Advances inanalytical methods (e.g. light scattering) and elec-tron microscopy in the 1980s allowed heat-inducedinteractions in milk proteins to be explored ingreater detail and the development of models forheat coagulation in normal and concentrated milk.However, some aspects of the mechanism of heatcoagulation have not been completely explained ata molecular level.

The extensive literature on the heat stability ofmilk has been reviewed regularly over the past

Vol 57, No 2/3 May/August 2004


112

40 years (Rose 1963; Fox and Morrissey 1977; Fox1981a, 1982; Singh 1988, 1995; Van Boekel

et al

.1989a,b; Singh and Creamer 1992; O’Connell andFox 2003).

A S S E S S M E N T O F H E A T S T A B I L I T Y

The heat stability of milk refers to the ability ofmilk to withstand high processing temperatureswithout visible coagulation or gelation. The mostwidely used method to assess heat stability, at leastfor research purposes, involves sealing a milksample in a glass tube, which is clipped onto a plat-form and placed in a temperature-controlled oil bath,usually at 140

°

C for normal milk and at 120

°

Cfor concentrated milk. The platform is rocked at agiven rate until a coagulum can be observed visu-ally. The heat coagulation time (HCT) is definedas the time that elapses between placing a sampleof milk in an oil bath and the onset of visiblecoagulation. Other methods for determining heatstability include the ethanol test, a whitening test,a protein sedimentation test and a viscosity determi-nation. However, the correlations between differ-ent test methods are generally unsatisfactory andthe HCTs determined by these methods oftencorrelate very poorly with the stability of milk oncommercial sterilization. From an industry pointof view, the use of a pilot-scale or laboratory-scalesterilizer provides more reliable results and predic-tion of behaviour of milk in commercial plants.

H E AT S TA B I L I T Y – pH P RO F I L E

The HCT of milk is affected by a number of factors,of which pH is the most important. The HCT ofmost milks shows a sharp maximum at pH valuesaround 6.7 followed by a minimum at pH 6.9; thestability increases again at higher pH value, asshown in Fig. 1. These milks are classified asType A milks. In Type B milks, the HCT increasesas a function of pH, being lower in the region of themaximum and higher in the region of the minimumthan for Type A milks. The HCT of concentratedmilk (20% non-fat solids) is much lower than that

of unconcentrated milk, with the maximum occur-ring in the pH range 6.4–6.6; the HCT on eitherside of the maximum remains very low. The mini-mum in the HCT vs. pH profile of normal milk canbe eliminated by altering a number of compositionaland processing parameters (Table 1). Artificialmodification of various milk salts influences theHCT–pH profile; a small reduction in the totalcalcium and magnesium ion concentration (from 13to 11 m

m

) eliminates the minimum in the HCT–pHprofile whereas addition of these cations decreasesthe stability throughout the pH range (Morrissey1969). The addition of phosphate to milk increasesthe HCT, whereas reducing the soluble phosphateshifts the HCT–pH profile to more alkaline values.The addition of citrate shifts the maximum tomore acid pH values, and the HCT does not recoveron the alkaline side of the maximum. Removal of40% of the colloidal calcium phosphate (CCP) in-creases the HCT in the pH range 6.4–7.4, whereasremoval of 60–100% of the CCP increases theHCT in the pH range 6.4–7.0 but has a destabiliz-ing effect at higher pH values (Fox and Hoynes1975). Increasing the concentration of lactose, toapproximately 150% of its normal level, destabi-lizes a Type A milk throughout the pH range 6.4–7.4 and shifts the minimum to more alkaline pHvalues (Sweetsur and White 1974).

Urea is the only indigenous constituent of milkthat has been shown to correlate strongly withnatural variations in heat stability. Addition of ureaat low concentrations does not affect the HCT inthe region of the maximum, but at high concentra-tions increases the HCT of milk. By contrast, addi-tion of urea to concentrated milk does not affect itsHCT–pH profile (Muir and Sweetsur 1976).

Of the milk proteins,

β

-lactoglobulin and

κ

-caseinhave the greatest impact on the HCT–pH profile(Fox and Hoynes 1975; Singh and Fox 1987a).

Figure 1 Heat coagulation (HCT) vs. pH profile for normal skim milk heated at 140°C. Type A milk (�), Type B milk (�), serum protein-free casein micelle dispersions (�) or concentrated milk (20% total solids) (�).

Table 1

Methods for eliminating the minimum from the HCT–pH profiles of Type A milks

Conversion of Type A to Type B

Decrease in the assay temperature (150

°

C to 120

°

C)Addition of

κ

-caseinRemoval of whey proteinsReduction in the levels of soluble saltsSeveral additives (e.g. aldehydes, oxidizing agents, polyphenols)Treatment with transglutaminase


113


β

-Lactoglobulin is required for the developmentof a Type A HCT–pH profile. The HCT of serumprotein-free casein micelles (SPFCMs), dispersedin milk ultrafiltrate, increases continuously withincreasing pH. Addition of

β

-lactoglobulin to anSPFCM dispersion introduces a maximum and aminimum into the HCT–pH profile (Fig. 1). Unlikenormal milk, the addition of

β

-lactoglobulin toconcentrated milk has a destabilizing effect overthe entire pH range. Enrichment of the milk with

κ

-casein increases the stability in the region of theminimum and converts a Type A milk to a Type Bmilk (Rose 1961).

In addition to the above factors, the HCT–pH profileof milk can be modified by numerous additives.Addition of thiol-blocking agents, such as

N

-ethylmaleimide (NEM) or iodoacetamide, markedlyreduces the HCT in the region of the maximum(Singh and Fox 1987b). Reducing agents such as

β

-mercaptoethanol destabilize milk over theentire pH range whereas oxidizing agents such asKBrO

4

and iodobenzoate eliminate the minimumin the HCT–pH profile (Singh and Fox 1985b).Formaldehyde increases the stability throughoutthe pH range 6.4–7.4, particularly in the region ofthe minimum, which is eliminated (Singh andFox 1985a). Addition of anionic detergents, such assodium dodecyl sulphate (SDS), to milk increasesthe stability in the region of the maximum andshifts the HCT–pH curve to alkaline pH valueswhereas cationic detergents, such as cetylmethyl-ammonium bromide, shift the HCT–pH profile tomore alkaline values while causing a moderateincrease in the maximum HCT (Fox and Hearn1978). Polyphenol-rich extracts of tea, coffee, wine,oak leaves and bark increase the HCT, particularlyin the region of the minimum (O’Connell

et al

.1988; O’Connell and Fox 1996). Caffeic acid isthe most effective of the polyphenols examined.

H E A T - I N D U C E D C H A N G E S I N M I L K R E L A T E D T O C O A G U L A T I O N

The coagulation of milk on extended heating athigh temperatures (120–140

°

C) is a consequence

of loss of casein micelle stability, as a result ofnumerous physical and chemical changes in itscomponents. When we consider the stability/insta-bility of casein micelles, the surface propertiesrather than the interiors of the micelles are likelyto be more important. The surface of the micellehas a number of dissociated carboxyl and someester phosphate groups, providing a high negativecharge (the zeta potential at 20

°

C is

−

13 mV) andthus electrostatic stabilization. Then there is adiffuse surface layer of flexible, hydrophilic poly-peptide chains consisting mostly of C-terminalsegments of

κ

-casein, providing steric stabilization(Holt 1992). This hairy layer of

κ

-casein providesa barrier against aggregation unless the hairs areremoved by chymosin treatment or the solvent qualityis reduced (for example by addition of ethanol).Inside the micelle, the individual casein moleculesare associated by hydrophobic and electrostaticbonds in which CCP also plays an important role.

Several factors influence the colloidal stabilityof milk. Important factors are calcium ions and pH,both of which diminish electrostatic repulsionsand possibly alter the conformation of

κ

-casein atthe micelle surface (indirectly reducing steric repul-sions). Heat treatment markedly changes the serumphase environment around the casein micelles (e.g.change in pH and soluble minerals, in particularcalcium ions, breakdown of lactose and urea) aswell as the casein micelles themselves (association ofwhey proteins, changes in CCP, dephosphorylation,casein dissociation). Some of these changes are listedin Table 2. It is not known exactly which particularchanges are directly responsible for coagulation,predispose the milk to coagulation or are a conse-quence of the coagulation process. The initial stagesof the heat coagulation process must involve a changein colloidal interactions that allows micelles toapproach each other and stay together long enoughfor chemical reactions to take place.

C H A N G E S I N C A S E I N M I C E L L E S

Heating milk at the heat stability assay tempera-tures causes denaturation of whey proteins andtheir interactions with casein micelles.

κ

-Casein on

Table 2 Changes in milk during heating and their possible impact on heat stability

Changes that promote instability Changes that enhance stability

Decrease in pH Reduction in calcium ion activityDeposition of calcium phosphate onto micelles Association of whey proteins with casein micellesAssociation of whey proteins with casein micelles Reduced sensitivity of casein to calcium ionsDephosphorylation of casein Thermal degradation products of lactoseDissociation and hydrolysis of caseins, in particular κ-caseinReduction in zeta potential and hydrationCovalent bond formation



114

the surface of casein micelles is involved in theformation of a specific disulphide-linked complexwith

β

-lactoglobulin (Singh and Fox 1987a;Jang and Swaisgood 1990). As the

β

-lactoglobulinaggregates or monomers are considered to formdisulphide bonds with

κ

-casein, the cystine resi-dues that are located in the

para

-

κ

-casein part ofthe protein must be relatively accessible to incom-ing proteins. A question still remains as to howwhey protein aggregates penetrate the hairy layerto find the disulphide bonds of

κ

-casein. Perhapsthere are changes in the

κ

-casein hairy layer duringheating, or heating causes rearrangement of themicelle surface to allow this interaction to takeplace.

The pH of heating has a large effect on theextent of the association of whey proteins with thecasein micelles (Smits and van Brouwershaven1980; Singh and Fox 1985a,b). At pH values < 6.8,a majority of whey protein complexes remainassociated with the casein micelle surface whereas,at higher pH values, these complexes remain in theserum. On heating at pH values > 6.8, not only dothe whey protein aggregates remain in the serumbut also micellar

κ

-casein dissociates in the serum.It is unclear whether the interaction between

κ

-casein and whey proteins occurs in the micellesand this complex then dissociates into the serumphase or whether the complex is in fact formed inthe serum. Nevertheless, the presence of wheyprotein markedly enhances the dissociation ofmicellar

κ

-casein. Other caseins,

α

s1

-,

α

s2

- and

β

-caseins, also dissociate from the micelles uponheating, but to a much lesser extent.

Data obtained by Singh and Fox (1985b, 1986)on milk and SPFCMs, heated at 140

°

C for 1 min atdifferent pH values, are shown in Fig. 2. At pH valueslower than 6.7, the amounts of soluble

κ

-casein(nonsedimentable at 100 000

g

for 60 min), repre-sented as 12% TCA-insoluble

N

-acetylneuraminicacid (NANA) and total soluble protein from heatedmilk, are lower than in a corresponding unheatedmilk, whereas at pH values > 6.7, these amountsare greater than in the unheated milk and increasewith increasing pH. It appears that

β

-lactoglobulinprevents the dissociation of micellar

κ

-casein onheating at pH values < 6.7 but enhances the releaseof micellar

κ

-casein at higher pH values (> 6.9).The effect of heat treatment on the zeta potential

is interesting because of similarities between theeffects of

β

-lactoglobulin on the HCT–pH profileand the zeta potential–pH profile of SPFCM dis-persions. Schmidt and Poll (1986) showed thatheating SPFCM dispersions at pH 6.7 for 10 minat 120

°

C had only a slight effect on their zetapotentials at room temperature; additions of

β

-lactoglobulin before heating led to an increase inthe zeta potential in the pH range 6.6–6.9 butreduced it at higher pH values (7.1–7.2). Anemaand Klostermeyer (1996, 1997) showed thatheating milk at 140

°

C at pH 6.5 caused an initialincrease in the zeta potential, which then decreasedon further heating. However, heating at pH 7.1markedly reduced the zeta potential initially, whichremained relatively constant thereafter (Fig. 3).They proposed that initial changes in the zetapotential are due to association of whey proteinswith casein micelles and precipitation of calciumphosphate on to the micelles whereas subsequentchanges in the zeta potential are caused by

κ

-caseindissociation and dephosphorylation of caseins.

Figure 3 Effect of heating time on the zeta potential of particles in skim milk that had been adjusted to pH 6.5 (�) or 7.1 (�) and then heated at 140°C. (Data from Anema and Klostermeyer 1997).

Figure 2 Influence of pH on the formation of nonsedimentable (100 000 g for 60 min) nitrogen (N) and nonsedimentable 12% TCA-insoluble N-acetylneuraminic acid (NANA) (indicative of κ-casein) on heating skim milk at 140°C for 2 min. Unheated milk: N (�) or NANA (�). Heated milk: N (�) or NANA (�). (Data from Singh and Fox 1985b, 1986).


115


Clearly, depending on the pH at heating, at leasttwo different kinds of casein particles with differentstructures and stability are produced. This associa-tion of whey proteins with the micelles at pH valuesbelow 6.8 would certainly modify the surface of thecasein particles, although it is uncertain how thisnew, complex surface is stabilized. It is possiblethat whey protein aggregates attached to the caseinmicelle surface protrude into the serum and thusact as super-steric stabilizers. It has been shownrecently that the association of denatured wheyprotein with the casein micelles increases their size(Anema and Li 2003) and that the zeta potential ofthe whey protein-coated micelles is greater thanthat of the native micelles, which would indeedcontribute to the stability of these particles. Conse-quently, the whey protein-coated micelles are morestable to heat, calcium ions, ethanol or rennet thanthe native casein micelles (Singh and Fox 1986).

The

κ

-casein-depleted micelles formed by heat-ing milk at pH > 6.8 have a reduced zeta potentialand show increased sensitivity to calcium ions,ethanol and heat compared with the native micelles.

κ

-Casein in the micelles has been shown to bepresent as disulphide-linked polymers of approxi-mate molecular weight 300–600 kDa. The deter-mination of the molecular state of the

κ

-caseinthat dissociates from the micelles is complicatedby the fact that, on heating, whey proteins interactwith it through thiol–disulphide interchangereactions. It is unknown whether or not

κ

-caseinpolymers are broken into oligomers and monomersduring the high heat treatment of milk. Reductionof disulphide bonds by treatment with mercaptoeth-anol tends to promote heat-induced dissociationof micellar

κ

-casein, indicating that the thermalbreakdown of disulphide bonds (if it occurs) is likelyto enhance

κ

-casein dissociation from the micelles.Singh and Latham (1993) analysed the dissociatedprotein material by size exclusion chromatographyand showed that the soluble protein material formedon heating milk at high pH is composed of small-sized whey protein/

κ

-casein aggregates and somemonomeric proteins. No monomeric

κ

-caseincould be detected. It is interesting to note that thedissociated protein material can generally be sepa-rated on SDS gel electrophoresis and gel filtration,but that in electrophoresis containing urea buffersystems, these proteins do not separate as discretebands, indicating modification of charged groupsand/or more structural changes to the proteins.

The reason why

κ

-casein dissociates from themicelles at slightly alkaline pH is not clear. It islikely that, when the surface charge reaches certaincritical values, hydrophobic bonds are insufficientto hold

κ

-casein on the micelle surface. The disso-ciation probably occurs as a result of electrostaticrepulsions between

κ

-casein and other micellecomponents. Modification of the charge distribu-

tion on

κ

-casein may also occur as a result ofother heat-induced changes, such as Maillard-typereactions. Alternatively, dissociation may involveconversion of CCP to an alternative form that isless capable of binding casein molecules andmaintaining the micelle structure (Aoki

et al

. 1990;Anema and Klostermeyer 1997). As the bindingof

κ

-casein to the micelles does not involve CCP,it is likely that such a change in the CCP wouldinfluence the dissociation of other caseins but haverelatively little effect on

κ

-casein.Concentration of milk prior to heating has a

marked effect on the dissociation of

κ

-casein; theextent of dissociation of

κ

-casein at any particularpH increases, with the dissociation–pH curve shift-ing towards lower pH values (Nieuwenhuijse

et al

.1991; Singh and Creamer 1991a) (Fig. 4). Thus, inconcentrated milk, considerable dissociation of

κ

-casein occurs on heating at normal pH, i.e.

∼

6.5–6.6. In a comprehensive study on casein micelle

Figure 4 (a) Influence of pH on the dissociation of micellar κ-casein on heating milks of 10% (�), 15% (�), 20% (�) or 25% (×) total solids at 120°C for 4 min. (b) Influence of heating time at 120°C on the dissociation of micellar κ-casein on heating milks of 10% (�), 15% (�), 20% (�) or 25% (×) total solids at pH 6.55. (From Singh and Creamer 1991a, reproduced with permission from Cambridge University Press).



116

dissociation, Singh and Creamer (1991b) showedthat the casein composition of the dissociated pro-tein was dependent on the heating time at 120

°

C;only κ-casein dissociated during the initial stagesof heating (up to 6 min) but other caseins alsodissociated with further heating; after heating for10 min, the dissociated protein was composed of70% κ-casein, 20% β-casein and 10% αS-caseins.The dissociated caseins existed as aggregates ofvarious sizes, including some monomers. Most ofthe dissociated κ-casein was covalently linked towhey proteins. The dissociated κ-casein appearedto have a charge distribution different from that ofnative κ-casein, but α- and β-caseins were essen-tially the same as in their native state, as indicated bytheir separation on urea-containing polyacrylamidegels. On SDS-containing gels, the dissociated caseinsshowed clear, distinct bands, similar to those of nativeproteins, indicating that the molecular weights ofthe proteins were not affected by the heat treat-ments, at least during the initial stages of heating.

Another important change in casein micelles isthat the dephosphorylation of caseins would beexpected to influence the casein micelle structure,as the casein phosphate groups are involved ininteractions of calcium ions and with CCP andprovide negative charges. Loss of phosphate groupsmay decrease the binding of caseins with CCP,resulting in dissociation of caseins. On the otherhand, dephosphorylation of phosphoserine residuesmay generate reactive intermediate dehydroalanine,which may promote casein cross-linking. Althoughthe real significance of thermal dephosphorylationis yet to be elucidated, the rate of dephosphoryla-tion does not appear to correlate directly with therate of heat coagulation.

C H A N G E S I N T H E S E R U M P H A S E

It has long been recognized that a crucial change inthe serum is the decrease in pH, which plays a keyrole in creating an environment that favours coag-ulation of the milk proteins. This is supported bythe observation that, if the pH of milk is readjustedoccasionally to its original value, coagulation ofthe milk may be prolonged for at least 3 h (Fox1981b). The pH of normal milk decreases gradu-ally with increasing heating time at 140°C; the pHat coagulation is usually between 5.5 and 6.0 (aftercooling the milk to 20°C). The decrease in pH iscaused by three reactions:

1 thermal oxidation of lactose to organic acids,which accounts for 50% of the pH decrease;2 hydrolysis of organic phosphate (from phospho-serine in casein), which contributes up to 30% ofthe decrease in pH;3 precipitation of tertiary calcium phosphate witha concomitant release of H+.

Although the pH of milk at the point of coagulation(estimated to be about 4.9) approaches that of acidcoagulation, heat-induced coagulation is not duemerely to an indirect acid-induced coagulationmechanism. This is illustrated by the fact thatthere is no relationship between the initial pHand the rate of pH decrease during heating; theapparent activation energies for acid productionand heat coagulation are very different, and thecoagulum formed on heating cannot be redispersedby increasing the pH (Van Boekel et al. 1989a,b;O’Connell and Fox 2003). Nevertheless, thedecrease in pH is likely to reduce electrostaticrepulsions, thus gradually destabilizing the caseinmicelle system.

Heat treatment has been shown to reduce theconcentration of soluble phosphate and of bothsoluble and ionic calcium. The transfer of calcium andsoluble phosphate from the serum to the micelleswould be expected to shield some of the negativecharges on the micelles, reducing the zeta poten-tial and decreasing electrostatic repulsions. Thecalcium ion activity, which depends on the initialpH of the milk, also decreases upon heating; thisdecrease is reversible and some or all of the cal-cium ion activity is recovered after heating. How-ever, this occurs upon cooling, not during heating,and takes at least 24 h. It is interesting that thedecrease in pH during heating does not result inan increase in calcium ion activity, but that thecalcium ion activity remains more or less constantafter heating for a few minutes (Van Boekel et al.1989a,b). Heating has no effect on monovalent ions,sodium, potassium and chloride, but its effect oncitrate is unclear. Another factor of importancein the serum phase is the whey proteins. Theseproteins are easily denatured by heat treatmentsabove 70°C and then react with each other or caseinmicelles, as discussed earlier.

M E C H A N I S M F O R C O A G U L A T I O N O F M I L K P R O T E I N

Based on many studies, a unified mechanism toexplain the HCT–pH profiles of both normal andconcentrated milks has been developed and largelyaccepted, and fits well with most of the experimen-tal observations (O’Connell and Fox 2003).

The HCT–pH curve is divided into two regions(Fig. 5); region I (pH values below 6.8) is con-cerned with the stability of whey protein-coatedmicelles, although the amount of whey proteinsthat associate with the micelles decreases with pHin this region. At a pH well below the maximum,milk coagulates rapidly because of low pH andhigh calcium ion activity (decreased electrostaticrepulsions). In addition, relatively high amounts ofwhey proteins associated with the micelles at lowpH may promote aggregation of casein particles

© 2004 Society of Dairy Technology 117


through cross-linking of whey proteins bound ontodifferent micelles. The occurrence of a maximum(at pH 6.7–6.8) in the HCT–pH profile of normalmilk is essentially due to greater stability of wheyprotein-coated micelles, as the formation of a com-plex between κ-casein and β-lactoglobulin on thesurface of the casein micelles alters the steric andelectrostatic interactions and prevents the dissoci-ation of micellar κ-casein. The details of the coag-ulation pathway for whey protein-coated micellesare by no means fully understood, but it is likelythat further heating causes lowering of the pH,dephosphorylation, covalent bond formation andother reactions. Consequently, the altered micellescan make more frequent contact and probablyform covalent cross-links within and between themicelles, resulting in an irreversibly aggregatedprotein material.

At pH values above 6.9 (region II), the stabilitydecreases due to the dissociation of micellar κ-casein, thus reducing the stabilizing effect ofthis protein. The κ-casein-depleted micelles aresensitive to calcium ion concentrations. Therefore,the minimum in the HCT–pH profile is a result ofcoagulation of κ-casein-depleted micelles; coagu-

lation is essentially salt induced, caused by calciumions. At higher pH, although dissociation of micel-lar κ-casein increases, the HCT increases due tothe increase in protein charge and low calciumion activity. It is also possible that the dissociatedκ-casein may reassociate during extended heatingbecause of a heat-induced decrease in pH.

An alternative hypothesis to explain the pHdependence of the heat stability of milk has beenproposed recently by O’Connell and Fox (2003).From pH about 6.3 to the pH of maximum stability,β-lactoglobulin denatures through a calcium-mediatedmechanism and enhances the thermal stability of thecasein micelles by chelating calcium. At pH values> 6.9, the disulphide bonds of β-lactoglobulin andκ-casein are reduced on heating, which facilitatescomplex formation. Concurrently, the hydrophobicβ-barrel of β-lactoglobulin is exposed, which resultsin an increase in the surface hydrophobicity of theβ-lactoglobulin–casein micelle complexes, therebysensitizing the casein micelles to the destabilizingeffect of heat-induced calcium phosphate precipi-tation. At pH values on the alkaline side of theminimum, stability increases as a function of pH,possibly because of an increase in protein stabilitywith increasing pH (micellar zeta potential andhydration increase as a function of pH) and a decreasein calcium ion activity with increasing pH. How-ever, this hypothesis does not explain the markeddissociation of κ-casein from the micelles at pH >6.9, as observed by many researchers. In the regionof the minimum, most of the β-lactoglobulin is infact found in the serum and is not associated withthe casein micelles. In addition, there is no evidenceto suggest that the disulphide bonds of nativeβ-lactoglobulin and κ-casein are reduced at hightemperatures, and it is unclear how this reductionwould facilitate complex formation. In such asituation, no disulphide–sulphydryl interchangereactions between κ-casein and β-lactoglobulinwould be expected to occur, which clearly is notthe case.

Concentrated milks are considerably less heatstable than unconcentrated milks. Because of thelower assay temperatures (120°C), many of the heat-induced changes in concentrated milk do not pro-ceed to the same extent as in unconcentrated milk.For example, the degree of dephosphorylation andcovalent bond formation, and the decrease in pH,are far smaller than in unconcentrated milk. Thecoagulum formed on either side of the maximumstability is soluble in dissociating buffers, althoughsome covalent bonds appear to form in the regionof the maximum (Nieuwenhuijse et al. 1991). Inthe pH region 6.5–6.7, the dissociation of κ-casein,the extent of which increases with an increase inpH, induces coagulation by altering the surface ofthe casein micelles, a situation somewhat com-parable with that existing for the coagulation of

Figure 5 Diagrammatic representation of heat-induced interactions of proteins in normal milk in relation to pH-dependence of heat stability. In region I, whey proteins are associated with the casein micelle. The stability of these whey protein-coated micelles increases with pH within this region. In region II, κ-casein/whey protein complexes are largely in the serum, and the stability of κ-casein-depleted micelles increases with pH in this region. Coagulation is mainly caused by aggregation of κ-casein-depleted micelles by calcium.


© 2004 Society of Dairy Technology118

normal milk within the region of minimum stability.Some of the similarities between the coagulationbehaviour of concentrated milk in the region of themaximum and that of unconcentrated milk in theregion of the minimum are shown in Table 3. AtpH values below 6.4, the denatured whey proteins,those in the serum as well as those attached tocasein micelles, are susceptible to heat-inducedaggregation; high calcium ion activity in this pHrange probably promotes the formation of largewhey protein aggregates and network structures.At higher pH (∼ 7.0), the coagulum appears toform a gel-like matrix of protein that is probablyderived from whey protein aggregates and thedissociated protein material (Singh et al. 1995).

P R A C T I C A L S I G N I F I C A N C E O F H E A T S T A B I L I T Y

The ability of milk to withstand high-temperaturetreatments without loss of its stability is fairlyunique and makes the production possible of manysterilized milk products with a long shelf life.These products include UHT milks and creams,in-can sterilized milks, evaporated milk, sweetenedcondensed milk and milk powders, especially thoseintended for reconstitution and recombination intosterilized products (heat-stable powders). Con-siderable knowledge gained on the heat stabilityof milk has allowed most of the practical problemsto be solved relatively easily by manipulating pro-cessing and compositional variables.

From an industrial viewpoint, the heat stabilityof milk of normal concentration has rarely beena problem. However, in recent years, many newliquid milks, fortified with high amounts of calcium,magnesium and zinc, cocoa and tea extracts, havebeen introduced into the market. As many of theseadditives have a negative impact on heat stability,these products require very careful manipulation ofthe formulation to achieve the desired heat stabilityand it is often difficult to achieve the desired shelfstability. Certain problems regarding the heatstability of concentrated milk, especially full-fathomogenized recombined evaporated milk, remainunsolved. These problems relate largely to seasonalvariations and batch-to-batch variations in heat

stability. Practical solutions to the heat stabilityproblems include the following:• manipulation of preheat treatments• matching the natural pH of milk to that of the

heat stability maximum• addition of different levels of phosphate• addition of buttermilk and phospholipids at

appropriate levels• combination of the above treatments.

R E F E R E N C E S

Anema S G and Klostermeyer H (1996) ζ-Potential of caseinmicelles from reconstituted skim milk heated at 120°C.International Dairy Journal 6 73–687.

Anema S G and Klostermeyer H (1997) The effect of pH andheat treatment on the κ-casein content and ζ-potential ofparticles in reconstituted skim milk. Milchwissenschaft 52217–223.

Anema S G and Li Y M (2003) Association of denaturedwhey proteins with casein micelles in heated reconstitutedskim milk and its effect on casein micelle size. Journal ofDairy Research 70 73–83.

Aoki T, Umeda T and Kako Y (1990) Cleavage of the linkagebetween colloidal calcium phosphate and casein on heat-ing milk at high temperature. Journal of Dairy Research57 349–354.

Fox P F (1981a) Heat-induced changes in milk precedingcoagulation. Journal of Dairy Science 64 2117–2137.

Fox P F (1981b) Heat stability of milk: significance of heat-induced acid formation in coagulation. Irish Journal ofFood Science Technology 5 1–11.

Fox P F (1982) Heat-induced coagulation of milk. InDevelopments in Dairy Chemistry. I. Proteins, pp 189–228. Fox P F, ed. London: Applied Science Publishers.

Fox P F and Hearn C M (1978) Heat stability of milk:influence of denaturable proteins and detergents on pHsensitivity. Journal of Dairy Research 45 159–172.

Fox P F and Hoynes MCT (1975) Heat stability of milk: influ-ence of colloidal calcium phosphate and β-lactoglobulin.Journal of Dairy Research 42 427–435.

Fox P F and Morrissey P A (1977) Reviews of the progress ofdairy science: the heat stability of milk. Journal of DairyResearch 44 627–646.

Holt C (1992) Structure and stability of bovine caseinmicelles. Advances in Protein Chemistry 43 64–151.

Jang H D and Swaisgood H E (1990) Disulphide bondformation between thermally denatured β-lactoglobulinand κ-casein in casein micelles. Journal of Dairy Science73 900–904.

Morrissey P A (1969) The heat stability of milk as affected byvariations in pH and milk salts. Journal of Dairy Research36 343–351.

Muir D D and Sweetsur A W M (1976) The influence ofnaturally occurring levels of urea on the heat stability ofbulk milk. Journal of Dairy Research 43 495–499.

Nieuwenhuijse J A, Sjollema A, van Boekel M A J S, vanVliet T and Walstra P (1991) The heat stability of con-centrated milk. Netherlands Milk Dairy Journal 45193–224.

O’Connell J E and Fox P F (1996) Effect of phenolic com-pounds on the heat stability of milk and concentratedmilk. Journal of Dairy Research 66 399–407.

Table 3 Apparent similarities between the coagulation behaviours of concentrated milk at pH ∼ 6.6 (region of maximum stability) and normal milk at pH ∼ 6.9 (region of minimum stability)

Addition of whey proteins causes destabilizationUrea addition up to about 6 mm has no effectThe coagulum is dispersible in 6 m urea bufferThe coagulation is a two-stage process, as revealed by nitrogen-depletion curvesExtensive dissociation of caseins from the micelles, especially κ-casein, occurs

© 2004 Society of Dairy Technology 119


O’Connell J E and Fox P F (2003) Heat-induced coagulationof milk. In Advanced Dairy Chemistry. I. Proteins,3rd edn, pp 879–930. Fox P F and McSweeney P L H, eds.London: Kluwer Academic.

O’Connell J E, Fox P D, Tan-Kintia R and Fox P F (1988)Effects of extracts of tea, coffee and cocoa on the colloidalstability of milk. International Dairy Journal 8 689–693.

Rogers L A, Deysher E F and Evans F R (1921) The relationof acidity to the coagulation temperature of evaporatedmilk. Journal of Dairy Science 4 294–309.

Rose D (1961) Factors affecting the pH-sensitivity of the heatstability of milk from individual cows. Journal of DairyScience 44 1405–1413.

Rose D (1963) Heat stability of bovine milk: a review. DairyScience Abstracts 25 45–52.

Schmidt D G and Poll J K (1986) Electrokinetic measurementson unheated and heated casein micelle systems. Nether-lands Milk Dairy Journal 31 342–357.

Singh H (1988) Effects of high temperatures on caseinmicelles. New Zealand Journal of Dairy Science Tech-nology 23 257–273.

Singh H (1995) Heat-induced changes in caseins includinginteractions with whey proteins. In Heat-Induced Changesin Milk, pp 86–99. Fox P F, ed. Special Issue 9501. Brus-sels: International Dairy Federation.

Singh H and Creamer L K (1991a) Influence of concentrationof milk solids on the dissociation of micellar κ-caseinon heating reconstituted skim milk at 120°C. Journal ofDairy Research 58 99–105.

Singh H and Creamer L K (1991b) Aggregation and dissoci-ation of milk protein complexes in reconstituted skimmilk at 120°C. Journal of Food Science 56 671–677.

Singh H and Creamer L K (1992) Heat stability of milk. InAdvanced Dairy Chemistry. I. Proteins, pp 621–656. FoxP F, ed. London: Elsevier.

Singh H and Fox P F (1985a) Heat stability of milk: the mech-anism of stabilisation by formaldehyde. Journal of DairyResearch 52 65–76.

Singh H and Fox P F (1985b) Heat stability of milk: pH-dependent dissociation of micellar κ-casein on heating

milk at ultrahigh temperatures. Journal of Dairy Research52 529–538.

Singh H and Fox P F (1986) Heat stability of milk: furtherstudies on the pH-dependent dissociation of micellarκ-casein. Journal of Dairy Research 53 237–248.

Singh H and Fox P F (1987a) Heat stability of milk: role of β-lactoglobulin in the pH-dependent dissociation of micellarκ-casein. Journal of Dairy Research 54 509–521.

Singh H and Fox P F (1987b) Heat stability of milk: influenceof modifying sulphydryl–disulphide interactions onthe heat coagulation time–pH profile. Journal of DairyResearch 54 347–359.

Singh H and Latham J M (1993) Heat stability of milk: aggre-gation and dissociation of protein at ultra-high tempera-tures. International Dairy Journal 3 225–237.

Singh H, Creamer L K and Newstead D F (1995) Heatstability of concentrated milk. In Heat-Induced Changesin Milk. Fox P F, ed. Special Issue 9501. Brussels: Interna-tional Dairy Federation.

Smits P and van Brouwershaven J H (1980) Heat-inducedassociation of β-lactoglobulin and casein micelles. Jour-nal of Dairy Research 47 313–325.

Sommer H H and Hart E B (1919) The heat coagulation ofmilk. Journal of Biological Chemistry 40 137–151.

Sommer H H and Hart E B (1922) The heat coagulation ofmilk. Journal of Dairy Science 5 525–543.

Sweetsur A W M and White J C D (1974) Studies on the heatstability of milk protein. I. Interconversion of type Aand type B milk heat stability curves. Journal of DairyResearch 41 349–358.

Van Boekel M A J S, Nieuwenhuijse J A and Walstra P(1989a) The heat coagulation of milk: mechanisms.Netherlands Milk Dairy Journal 43 97–127.

Van Boekel M A J S, Nieuwenhuijse J A and Walstra P(1989b) The heat coagulation of milk: 3. Comparison oftheory and experiment. Netherlands Milk Dairy Journal43 147–162.

Webb B H and Bell R W (1942) The effect of high-temperatureshort-time forewarming of milk upon the heat stability ofits evaporated product. Journal of Dairy Science 25 301–311.

Heat Stability of Milk (2004)

Documents

Transcript of Heat Stability of Milk (2004)