Original article

Rennet-induced aggregation of homogenizedmilk:Impact of the presence of fat globules

on the structure of casein gels

Gisèle ION TITAPICCOLO, Marcela ALEXANDER, Milena CORREDIG*

Department of Food Science, University of Guelph, Guelph, Ontario, N1G 2W1, Canada

Received 21 September 2009 – Revised 18 February 2010 – Accepted 1st March 2010

Published online 18 June 2010

Abstract – The present study investigated the renneting behaviour of casein micelles as influencedby the presence of fat globules. The gelation of skim milk, homogenized milk and homogenizedmilk with Tween 20 added was observed using diffusing wave spectroscopy and rheology. Byadding Tween 20 to homogenized milk, it was possible to displace most of the milk protein fromthe oil/water interface and create a system with the same colloidal attributes as homogenized milkbut with oil droplets with a very different interfacial composition, and relatively inert during rennet-induced aggregation. The primary phase of gelation was faster in homogenized milk compared toskim milk. Casein micelles were altered by the homogenization process which spreads them at theinterface, making κ-casein more readily available. The addition of Tween 20 did not further affect theenzymatic activity on the micelles. However, the onset of gelation occurred earlier in homogenizedmilk with Tween 20 added, as the micelles aggregated at a lower level of proteolysis compared tohomogenized milk. This work clearly showed that the gels formed in homogenized milk with andwithout Tween 20 have different physico-chemical properties, because of the different colloidal stateof the filler particles.

homogenized milk / polysorbate / rennet coagulation / diffusing wave spectroscopy

摘要 – 均质奶的酶凝乳: 脂肪球对酪蛋白胶束结构的影响○ 本文考察了脂肪球的存在对酪蛋白胶束凝乳特性的影响○ 通过运用散射波谱和流变仪观察了脱脂奶、均质奶和添加吐温 20 的均质奶的凝乳过程○ 吐温 20 添加到均质奶后，可能会替代水包油界面中的大多数乳蛋白，形成了与均质奶具有相同胶体特性的新体系，但是油滴界面成分会因此发生变化，并且皱胃酶凝乳过程会相对缓慢○ 均质奶中凝胶相的形成比脱脂奶中要快○ 均质过程使酪蛋白胶束向胶体界面扩散，从而 κ-酪蛋白更易获得○ 虽然添加吐温 20 不能进一步影响胶束中酶活性，但是，添加吐温 20 的均质奶凝胶的形成比未添加吐温 20 的要快，这是由于添加吐温 20 的均质奶可以在较低的水解蛋白浓度下使酪蛋白胶体聚合所致○ 该研究清楚的表明了由于填料颗粒具有不同的胶体状态导致了均质奶和添加吐温 20 的均质奶形成的凝胶具有不同的物化特性○

均质奶 / 聚山梨醇酯 / 皱胃酶凝乳 / 散射波谱

Résumé – Agrégation induite par la présure de lait homogénéisé : impact de la présence deglobules gras dans la structure des gels de caséine. Cette étude porte sur l’influence de laprésence des globules gras sur le comportement des micelles de caséine après ajout de présure.

*Corresponding author (通讯作者): mcorredi@uoguelph.ca

Dairy Sci. Technol. 90 (2010) 623–639© INRA, EDP Sciences, 2010DOI: 10.1051/dst/2010023

Available online at:www.dairy-journal.org

Article published by EDP Sciences

La gélification de lait écrémé, de lait homogénéisé et de lait homogénéisé additionné de Tween 20 aété observée par diffusion dynamique de la lumière et par rhéologie. En ajoutant du Tween 20 aulait homogénéisé, il était possible de déplacer la plupart des protéines adsorbées à l’interface huile/eau et de créer un système avec les mêmes éléments colloïdaux que le lait homogénéisé mais avecdes gouttelettes lipidiques ayant une composition à l’interface très différente et relativement inertesdurant l’agrégation induite par la présure. La phase primaire de la gélification était plus rapide dansle lait homogénéisé en comparaison au lait écrémé. Les micelles de caséines étaient altérées parl’homogénéisation qui les disperse à l’interface, rendant les caséines kappa plus disponibles.L’addition de Tween 20 ne provoquait pas de modification supplémentaire de l’activité enzymatiquesur les micelles. Cependant, le démarrage de la gélification apparaissait plus précocement dans lelait homogénéisé avec Tween 20, puisque les micelles s’agrégeaient à un niveau plus faible deprotéolyse que celui du lait homogénéisé. Ces travaux montrent que les gels formés dans du laithomogénéisé additionné ou non de Tween 20 ont des propriétés physico-chimiques différentes, enraison de l’état colloïdal différent des particules de remplissage.

lait homogénéisé / polysorbate / coagulation présure / spectroscopie en diffusion dynamiquede la lumière

1. INTRODUCTION

Fat globules play an important role in thetexture of many food products, impartingdesirable mouthfeel and carrying flavours.Many dairy products are mixed gels inwhich proteins are a constituent part of thenetwork. When fat globules are dispersedin gelling milk, the microstructure and frac-ture behaviour of the gels are altered, withimportant consequences to texture. Thedynamics of the assembly depend on thecomponents present in the system, as wellas the environmental conditions (pH,temperature and ionic strength). The rennet-induced gelation of homogenized milk, inparticular with regard to the effect of fatglobules on the retention of moisture incheese curd [13], on free oil [20, 28] andon yield [20], has been widely studied.However, very little is known about thedynamics of the interactions between themilk fat globules and the protein matrix inrennet-induced gels.

Fresh whole milk contains fat globulesdistributed in a range of diameters from0.1 to 10 μm [19]. The majority of theglobules has a diameter < 0.15 μm butrepresents a small volume of the total fat,while few of the globules are very large(> 10 μm) and are responsible for containing

the majority of the mass of the milk fat.Homogenization of whole milk reduces thesize of the fat globules by breaking themwithstrong shearing forces. This has the effect ofincreasing the total surface area of the glob-ules by 4–10 times [19], which, in turn, gen-erates the adsorption of proteins onto theinterface to decrease the interfacial energyof the newly created surfaces. Caseinmicelles adsorb preferentially overwheypro-teins [1, 26], and these casein-coated fat glob-ules have the ability to participate in therennet-coagulation process of milk.

Enzymatic coagulation by rennet is thebasis of the cheese-making process and isgenerally considered to be composed oftwo stages [17, 30]. During the first stage,the proteolytic enzyme chymosin cleavesκ-casein, which mostly resides on thesurface of the casein micelles. Becauseκ-casein is the main contributor to the stabil-ity of the micelles, its breakdown affectsmicellar charge and steric stabilization[5, 11], and, when a sufficient amount iscleaved, unstable casein micelles begin toaggregate [32]. Aggregation is the secondstage of the rennet-coagulation process andleads to curd formation.

Studies have focused on the effect ofhomogenized fat globules on the propertiesof rennet gels. Fat globules in homogenized

624 G. Ion Titapiccolo et al.

milk have different rennet-coagulationbehaviour than native globules, due to thedifferences in size and interfacial composi-tion [10, 14, 18, 21]. Homogenized globulesparticipate in the filled gel as “active fillers”,i.e. they contribute, together with caseinmicelles, to the development of the gelmatrix; on the contrary, native fat globulesparticipate as “weakly interactive” or “inertfillers”, occupying the voids within thecasein strands because their membrane isnot able to interact with casein micelles dur-ing coagulation [29].

It has been previously demonstrated,using model systems containing oil dropletsof known size and composition in recom-bined milk, that the rheological and struc-tural properties of milk gels depend on thecomposition of the fat globule surface[6, 9, 14] as well as the colloidal state ofthe droplets [8, 23]. To distinguish betweenthe behaviour of “active” fat globules and“relatively inert” fat globules, homogenizedmilk with polysorbates (Tween 20) was alsostudied. It is known that small molecule sur-factants are able to displace proteins fromthe surface of fat globules in emulsions[15], and it has been previously reportedthat during acid coagulation of recombinedmilk, fat globules with Tween 20 adsorbedat the interface behave as passive, “inert”fillers [2]. By comparing homogenized milkwith and without Tween 20 it is possible tostudy systems with similar colloids charac-teristics (size and flocculation state) duringrennet-induced coagulation of homogenizedmilk. Although Tween 20 has been usedbefore to stabilize model emulsion dropletsin recombined milk [2, 6, 9], the use ofTween 20 to displace protein in homoge-nized milk has yet to be reported as a wayto study the impact of fat globule duringthe formation of structure of rennet caseingels.

This study analyses the behaviour ofhomogenized fat globules in milk duringrennet coagulation. In recent years diffusingwave spectroscopy (DWS) has been

employed to observe sol-gel transitions inmilk without the need for dilution [7, 8, 25,27]. By better understanding the beginningstages of aggregation, we may be able toapply approaches such as selective homog-enization to increase curd yields, improvebody and texture of reduced-fat productsand design novel dairy gels.

2. MATERIALS AND METHODS

2.1. Materials

Fresh whole milk was collected from theElora Dairy Research (Elora, ON, Canada),and sodium azide was added at a concentra-tion of 0.2 g·L−1 to act as a bacteriostaticagent as soon as it was received.

Milk was then passed through a one-stage high pressure homogenizer (Emulsi-flex C5, Avestin, Ottawa, ON, Canada) ata pressure of 34.5 MPa. The homogeniza-tion was carried out for three consecutivepasses. All the experiments on homoge-nized milk were performed the sameday of homogenization unless otherwiseindicated.

For skimming, milk was centrifuged at4000� g for 20 min at 4 °C using aBeckman J2-21 centrifuge with JA-10rotor (Beckman Coulter, Mississauga, ON,Canada) with subsequent filtration throughWhatman glass fibrefilters (Fisher Scientific,Whitby, ON, Canada); the filtration wasrepeated four times. Skimmilk sampleswerestored at 4 °C and analysed within a week ofpreparation.

Polyoxyethylene sorbitan monolaurate(Tween 20, 1227 g·mol−1, Sigma ChemicalCo, St. Louis, USA) was added to homoge-nized milk at a concentration of 2 g in100 mL. The milk was stirred for 6 h atroom temperature before further analysis.

Chymostar Single Strength rennet(Rhodia, Cranbury, USA) was used at a con-centration of 0.018 IMCU·mL−1, at a tem-perature of 30 °C for gelation experiments.

Renneting of homogenized milk 625

The samples were stirred for 30 s after ren-net addition and immediately analysed byDWS or rheology.

2.2. Transmission electronmicroscopy

Transmission electron microscopy(TEM) was performed on homogenizedmilk prior to gelation. Initially, the fluidmilk was prepared for thin sectioning byencapsulation in 20 g·L−1 agar tubes. Fol-lowing primary fixation with a 20 mL·L−1

glutaraldehyde solution in Sorenson phos-phate buffer 0.07 mol·L−1, pH 6.8, the sam-ples were post-fixed in 10 mL·L−1 osmiumtetroxide solution in the same phosphatebuffer. After dehydration in ethanol fol-lowed by propylene oxide using a gradientwith higher alcohol concentration, sampleswere embedded in Spurr’s low viscosityembedding media. The resin blocks werecut into sections 70–90 nm thick using aReichert Ultracut E Ultramicrotome (Leica).The sections were stained with saturateduranyl acetate in water followed byReynolds lead citrate solution. The sectionswere mounted on grids and examined with aTEM (Philips CM 10, Eindhoven, Holland)with an accelerating voltage of 60 kV.Typical magnification factors of 25 000–46 000 were used.

2.3. Gel electrophoresis

Sodium dodecyl sulphate polyacryl-amide gel electrophoresis (SDS-PAGE)was performed using a Bio-Rad electropho-resis unit (Bio-Rad Laboratories Inc.,Hercules, CA) to analyse the protein com-position of the cream and serum phases inthe homogenized milk samples. Thecream was separated from the serum bycentrifugation at 5000� g for 20 min,using an Eppendorf 5415D centrifuge(Brinkmann Instruments Ltd., Mississauga,ON, Canada), dried on a filter paper for30 min and resuspended in Milli-Q water

maintaining the same fat concentration asin milk (35 g·L−1). Each sample was diluted1:2 in reducing sample buffer (containing1 mol·L−1 Tris HCl, pH 6.8, 100 g·L−1 SDS,750 mL·L−1 glycerol, β-mercaptoethanol,10 mL·L−1 bromophenol blue) and heatedfor 5 min at 95 °C.After heating, the sampleswere centrifuged (see above) before loadingin the gel.

The resolving gel for the electrophoresisanalysis contained 150 mL·L−1 acrylamidein 1.5 mol·L−1 Tris HCl at pH 8.9, whilethe stacking gel was 40 mL·L−1 acrylamidein 0.1 mol·L−1 Tris PO4 buffer at pH 6.7.The electrophoresis buffer was 30 g·L−1

Tris HCl, 144 g·L−1 glycine and 10 g·L−1

SDS at pH 8.3. Aliquots of 7 μL of the pre-pared samples were loaded into the gels andthe electrophoresis separation was per-formed at 200 V for 40 min. Gels were thenstained with Coomassie blue in 500 mL·L−1

methanol and 100 mL·L−1 acetic acid for30 min while shaking and destained witha solution of 450 mL·L−1 methanol and100 mL·L−1 acetic acid for 1 h. Sampleswere left overnight in a fresh destain-ing solution diluted 1:1 in Milli-Q waterand scanned with a Sharp JX-330 scanner(Pharmacia Biotech).

2.4. Light scattering

The apparent average diameter of thedifferent milk samples was measured usingdynamic light scattering (DLS) (ZetasizerNano, Malvern Instruments, Worcestershire,UK). The hydrodynamic size of the caseinmicelles was obtained from the average ofthree separate readings, on replicate sam-ples. Samples were diluted ~ 2000 timesin permeate (milk serum) and placed inthe spectrometer right after dilution. Perme-ate was collected during the concentrationof skim milk to a 2x volume fraction byultrafiltration (PLGC 10k regenerated cellu-lose cartridge, Millipore Corp., Bedford,MA), and it was filtered before use througha 0.22 μm filter (Millipore Canada Ltd.,

626 G. Ion Titapiccolo et al.

Mississauga, ON, Canada). The dilution inpermeate was performed to avoid multiplescattering while preserving the environmen-tal conditions of the casein micelles.

The particle size distribution of the sam-ples was also measured using integratedlight scattering (Mastersizer 2000, MalvernSouthborough, MA). The refractive indicesused were 1.46 for milk fat and 1.33 forthe dispersant (water). Each measurementwas obtained from the average of threereadings.

Sampleswere also analysed byDWS dur-ing renneting experiments. A volume of~ 1.5 mL of undiluted milk sample wasplaced into a flat-faced, 5 mm path lengthoptical glass cuvette (Hellma Canada Ltd.,Concord, Canada) and the temperature ofthe equipment was maintained at 30 °C withawater bath.The light sourcewasa solid statediode pumped Nd:YAG laser (Coherent,Santa Clara, CA), of wavelength 532 nmand power of 100 mW and the transmittedscattered light was collected by a single fibreoptic that was then bifurcated and fed totwo matched photomultipliers (HC120-03,Hamamatsu, Loveland, OH) and a corre-lator (FLEX2K-12 � 2, Bridgewater, NJ).Standard latex spheres of 260 nm diameter(Portland Duke Scientific, Palo Alto, CA)were used to calibrate the laser intensitydaily. Correlation functions and intensityof the transmitted scattered light weremeasured at intervals of 2 min for 80 min.Data were analysed using specializedsoftware (DWS-Fit, Mediavention Inc.,Guelph, ON, Canada). Each experimentwas replicated at least three times, startingfrom different batches of fresh milk.

DWS is based on the measurement oftemporal fluctuations of light that has beenmultiple-scattered by particles in a sample[16]. In recent years, it has been employedin the investigation of destabilization mech-anisms in food systems [3, 27]. An exhaus-tive description of the equipment and theorycan be found elsewhere [31]. Briefly, DWSrelies on many scattering events happening

as a photon of light traverses a colloidal dis-persion and the light propagation is approx-imated by a random path distribution. DWScan yield information on the static proper-ties (positional correlations) of a systemvia the photon transport mean free path,l*, as well as dynamic properties via thedecay time, τ, and the mean square displace-ment (MSD). The value of l*, function ofthe physical properties of the scatterers aswell as their spatial correlation, can bedetermined experimentally by dividing theaverage intensity of the transmitted lightof the sample, T, by that of a calibratingsample with well-determined l* [31]. TheMSD is the average of the square of the dis-tance travelled by the particles at a giventime. In a free-diffusing regime, the motionis described as a random walk, for whichthe MSD increases linearly with time andthe radii of the scatterers can be calculatedvia the Stokes-Einstein relation. In a gel,the particles will only be able to move a cer-tain distance from their average positionbefore being restricted in motion by theirinteractions with others. Therefore, at longtimes, their average displacement will reacha constant value [22].

2.5. Rheology

Rheological experiments were per-formed with a stress-controlled rheometer(AR 1000, TA Instrument Ltd., New Castle,USA), using a conical concentric cylindergeometry (5920 μm fixed gap, 15 mmradius and 42 mm cylinder immersedheight). An external water bath (Isotemp3016, Fisher Scientific, Whitby, Canada)was connected to the rheometer to keep thetemperature of the sample at 30 °C for theduration of the experiment. After rennetaddition, each sample was immediatelyplaced in the rheometer and four successivetests were performed. First a time sweep wasrun at 0.5 Pa controlled stress, 1.0 Hzfrequency and 0.1 μN·m initial torque. Thefirst time sweep was ended at the gelation

Renneting of homogenized milk 627

point of each sample, which was the time ofthe crossover between G′ (storage modulus)and G″ (loss modulus). A second time sweepwas performed immediately after the firstone to follow gel development. This sweepalways had a duration of 45 min and wasperformed using the same initial parameters.Subsequently, a frequency sweep was runfrom 10 to 0.01 Hz at a controlled stressof 0.5 Pa. At last, a strain sweep wasperformed (at 1 Hz) to ensure that theparameters applied during the previous testswere within the linear viscoelastic region ofthe sample.

2.6. Rennet activity

The release of the casein macropeptide(CMP) from κ-casein by enzyme actionwas measured with RP-HPLC. After addi-tion of rennet, each sample was divided inaliquots in different test tubes. Every5 min, trichloroacetic acid (TCA) at a con-centration of 40 g·L−1 (final concentrationin the sample of 20 g·L−1) was added inthe tubes to stop the enzymatic reaction.The experiment was performed at constanttemperature of 30 °C for 45 min. AfterTCA addition, each sample was mixedusing a vortex mixer and stored overnightat 4 °C. The next day, all samples wereequilibrated at room temperature and centri-fuged at 4500� g for 15 min, with anEppendorf 5415D centrifuge (BrinkmannInstruments Ltd.,Mississauga, ON,Canada).The supernatants were separated with asyringe, filtered through 0.45 μm Millex-GV filters and analysed by chromatography.

RP-HPLC was carried out with aFinnigan SpectraSystem LC unit compris-ing a degasser, pump, autosampler with a100 μL sample loop and UV detector (at210 nm). The sample was loaded into aVydac C4 guard column (Mandel, USA),then into the column, a Pharmacia BiotechμRPC C2/C18 ST 4.6/100 (Piscataway,NJ). Using a flow rate of 1 mL·min−1,a nonlinear gradient was run between

solvent A (1 mL·L−1 TFA in water) and sol-vent B (1 mL·L−1 TFA in 900 mL·L−1 aceto-nitrile). The elution started at 18% solvent B,increasing to 39% in 35 min and to 100% inthe next 6 min; 100% solvent B was thenkept constant for 8 min before decreasingback to 18% in the last 4 min of run. For eachsample, the peak area was integrated usingChromQuest (version 4.1, ThermoFinnigan,Burlington, ON, Canada).

2.7. Statistical analysis

All the experiments were carried out intriplicate and mean values are reportedunless otherwise indicated. All the figurespresented in this work contain graphs thatare the most representative of the three rep-licates. Analysis of variance and least signif-icant difference computations were carriedout in order to determine eventual signifi-cant differences between treatments. Statis-tical analyses were conducted using SPlus8.0 and differences were considered at95% confidence level.

3. RESULTS

3.1. Characterization ofhomogenized milk systems

TEM, SDS-PAGE and light scatteringtechniques were used to characterize the dif-ferent milk samples: homogenized milk(HM), homogenized milk with Tween 20(HM-Tw) and skim milk (SM); the latterused as a reference to the behaviour of case-ins in isolation.

Transmission electron micrographs ofthe homogenized milk are shown inFigure 1. Homogenization of whole milkreduces the size of the fat globules and cre-ates a heterogeneous fat globule distribu-tion. Two main populations of fat globulescan be identified; the largest globules, of300–400 nm in diameter, are covered withthe milk fat globule membrane and partially

628 G. Ion Titapiccolo et al.

spread casein micelles (Fig. 1A). The sec-ond population has very small globules,with diameters < 100 nm, which in somecases appear to be clustered with themicelles either by direct interaction or byincorporation into them (Fig. 1B). Theseobservations confirm those of previousreports [4].

The proteins in the serum and creamphases of the homogenized milk samples

were analysed with SDS-PAGE (Fig. 2) toconfirm that Tween 20 caused protein dis-placement from the interface. While inhomogenized milk there is a clear presenceof casein proteins in the cream phase(Lane 3), very little protein is left in thecream phase of the same sample afterincubation with Tween 20 (Lane 4). Theaddition of Tween 20 to homogenizedmilk leads to the interfacial replacement

CLUSTERS

FAT

MICELLES

A

B

Figure 1. Transmission electron micrographs of homogenized milk. (A) Magnification of 46 000(scale bar = 1 μm); (B) magnification of 25 000 (scale bar = 1 μm), showing fat globules (F) andcasein micelles (M, heavily stained). Clusters of casein micelles and small globules are alsoindicated in the images.

Renneting of homogenized milk 629

of the proteins by the surfactant and resultsin the removal of casein micelles from thesurface of the globules into solution.

Figure 3 summarizes the results on sizedistribution analysis of SM, HM andHM-Tw, carried out using integrated lightscattering. Skim milk shows a monomodaldistribution of sizes with an average diame-ter around 120 nm. This is expected, asonly casein micelles are present in skimmilk. Both the HM and HM-Tw samplesshow a bimodal distribution of sizes, withfat globules sizes of about 1.2 μm of diam-eter. While integrated light scattering doesnot show significant differences in thesize of the fat globules in homogenizedmilk with or without Tween, there is a sig-nificant reduction in the average apparentdiameter of the particles (from 271 ± 7.1to 216 ± 7.3 nm) in the presence of Tween,when the size is measured by DLS. Thesmaller average size measured by DLScan be attributed to the displacement of

the casein micelles from the interface aswell as the disruption of clusters.

3.2. Renneting

The first phase of the enzymatic coagula-tion by rennet can be monitored by measur-ing the release of CMP from the surface ofcasein micelles. Figure 4 shows the percent-age of CMP released over time after additionof rennet. The maximum CMP released(maximum area) in skim milk control wastaken as 100% release, and all measurementswere compared to this value. All the samplesreached themaximumvalueofCMPreleased(100%).HMandHM-Tw showa statisticallysignificant difference in the release of CMPcompared to control SM, showing a slowerCMP release compared to the HM andHM-Tw. Table I indicates the difference inthe time necessary to reach 95% CMPrelease. On the other hand, no differenceswere measured between homogenized milk

αs casβ casκ cas

β-lg

α-la

Figure 2. SDS-PAGE conducted under reducing conditions of homogenized milk with or without2% Tween 20. Lane 1: homogenized milk serum; Lane 2: serum of homogenized milk with Tween;Lane 3: homogenized milk cream; and Lane 4: cream of homogenized milk with Tween. Thearrow represents the direction of migration (cas = casein, β-lg = β-lactoglobulin and α-la =α-lactalbumin).

630 G. Ion Titapiccolo et al.

with or without Tween. These results werealso confirmed by using half of the rennetconcentration (not shown).

To follow the kinetics of aggregation ofthe caseins and fat globules during gelation,DWS and rheology experiments wereperformed. Figure 5A shows the behaviourof the storage modulus, G′, measured by

rheology, as a function of renneting time.The gelation point of milk is clearly acceler-ated by the presence of fat globules, as skimmilk shows the slowest gelation point com-pared to homogenized milk. In addition,HM-Tw gelled significantly earlier thanHM (see Tab. I for statistical differences).The elastic modulus of the gel was also

Diameter (μm)0.01 0.1 1 10 100

Vol

ume

(%)

0

1

2

3

4

5

6

Figure 3. Particle size distribution obtained by integrated light scattering of skim milk (□),homogenized milk (●) and homogenized milk with Tween (N).

Time (min)0 10 20 30 40 50

% C

MP

rele

ase

0

20

40

60

80

100

Figure 4. % CMP release versus time after addition of rennet for skim milk (■), homogenizedmilk (●) and homogenized milk with Tween (N).

Renneting of homogenized milk 631

modified by the presence of fat; in fact therate of growth of G′ is faster for HM thanskim milk and even faster for HM-Tw.

Figure 5B shows the development of1/l* as a function of time after addition ofrennet. The parameter is normalized ashomogenized milk does not only containcasein micelles but also fat globules, andtherefore contains two scatterers with differ-ent scattering properties. This results inabsolute values of 1/l* quite different com-pared to those of skim milk control. Sincewe are interested in the kinetics of gel for-mation with time, and not necessarily inthe absolute values of turbidity, normaliza-tion can be justified. The presence of thefat globules in homogenized milk samplesheavily skews the light signal [7, 8]. Theirlarger size as well as their higher refractiveindex contrast compared to the caseinsdominates the scattering of light and theresults shown by DWS for homogenizedmilk reflect the behaviour of the fat globulesas affected by the formation of a casein gelaround them [7, 8].

All the samples show a delay phasebefore manifesting a change in 1/l*. Afterthis delay, 1/l* starts to increase, reflectingthe beginning of interparticle interactionsdue to the action of rennet. The overallbehaviour of 1/l* of all the samples isrelated to the development and later forma-tion of a gel [25]. It appears that 1/l*increases earlier in homogenized milk sam-ples than in skim milk, and HM-Tw shows achange before HM. These differences are

statistically significant (see Tab. I). Afterthe initial 1/l* increase, there is a secondchange in slope which corresponds directlyto the beginning of the aggregation ofcasein micelles (see Fig. 5C); this is moreevident for SM and HM-Tw samples thanfor HM. In skim milk samples, the increaseof 1/l* is continuous and does not seem toreach a limit within the experimental time.On the contrary, homogenized milk andHM-Tw show a plateau, with some delayfor HM with Tween, and almost immediatefor HM.

The particle’s aggregation can be fol-lowed also by measuring the apparentradius of the scatterers over time(Fig. 5C). Although an increase in apparentradius after gelation has no physical mean-ing, this change reflects the decrease in dif-fusion (or mobility) of the scatteringparticles. Confirming what was alreadyobserved by rheology (Fig. 5A) and withthe 1/l* parameter measurements (Fig. 5B)also in the case of the radius as a functionof time, it is clear that the coagulation timeis accelerated by the presence of fat globules(HM and HM-Tw aggregate earlier thanSM). In addition, there is a significantly fas-ter aggregation in the sample containingTween. The size change (or more correctly,the rate of decrease of diffusion) is quite fastin SM and HM samples, suggesting in thelatter the interaction of the fat globules withthe casein network. On the other hand, inHM-Tw, although the coagulation time issignificantly faster than for HM and SM

Table I. Rennet-coagulation parameters obtained for skim milk (SM), homogenized milk (HM) andhomogenized milk with Tween (HM-Tw). The values are the means of three replicates. Means inthe same column with no common superscript are statistically different (P < 0.05).

95% CMP(min)

Gel point(min)

1/l*(min)

Radius(min)

G′ after 45′ (Pa) % Strain atbreak

SM 28.9a 37.5a 16.5a 36.3a 43.2a 20.22a

HM 17.2b 25.6b 9.4b 16.6b 52.7a 33.66a

HM-Tw 15.6b 14.7c 7.3c 10.3c 106.2b 33.37a

632 G. Ion Titapiccolo et al.

samples, the apparent radius does notincrease continuously, but rather reaches aplateau at about 1 μm.

Figure 6 shows the difference in thedevelopment of the MSD for SM andHM with or without Tween. Each line

G' (

Pa)

0

20

40

60

80

100

120

Nor

mal

ized

1/l*

0.95

1.00

1.05

1.10

1.15

1.20

Time (min)0 10 20 30 40 50 60

Rad

ius

(μm

)

0

1

2

3

A

B

C

Figure 5. Rennet-coagulation process monitored using rheological and DWS parameters: G′ (A),1/l* (B) and radius (C) as a function of time after addition of rennet. Skim milk (■); homogenizedmilk (●); homogenized milk with Tween (N).

Renneting of homogenized milk 633

corresponds to a different time during thegelation process (with the reaction timesincreasing from left to right). A linear

dependence of the MSD with correla-tion time indicates free diffusive motion[16]. During the initial stages after rennet

Time (s)0 2e-5 4e-5 6e-5 8e-5

MS

D (μ

m2 )

MS

D (μ

m2 )

MS

D (μ

m2 )

0.0

5.0e-6

1.0e-5

1.5e-5

0.0

5.0e-6

1.0e-5

1.5e-5

0.0

5.0e-6

1.0e-5

1.5e-5

A

C

B

Figure 6. MSD as a function of time for skim milk (A), homogenized milk (B) and homogenizedmilk with Tween (C). Process develops from left to right.

634 G. Ion Titapiccolo et al.

addition, all samples show a linear depen-dence of the MSD, as the scatterers are freediffusing. Before the enzyme has a chanceto cleave enough κ-casein off the surfaceof the micelles, both fat globules andmicelles are free to probe the full accessiblespace range, resulting in a linear increase ofMSD with time. As the micelles initiallydevelop into small clusters, the diffusionof these aggregates slows down (comparedto the initially free casein micelles and fatglobules) and this results in a decrease ofthe slope of the linear graph. This is alsoevidenced in Figure 6, in the lines corre-sponding to the intermediate time after ren-net addition. However, differences in theMSD slopes can be seen between the sam-ples. At the gelation point, the time depen-dence of MSD evolves from linear toasymptotic for SM and HM. Homogenizedmilk shows a drastic decrease in the slopeof the MSD curve when compared to SM.When Tween is present in the homogenizedmilk, there is a very little change in theMSD slope over time.

To better understand if the differencesobserved in the rheological and light scatter-ing parameters are related to the amount of

CMP released during the renneting reaction,the change of 1/l* and G′ are plotted as afunction of CMP release in Figure 7. Inthe case of SM, 1/l* starts to increase afteraround 80% of CMP has been released;however, the G′ value increases at about95% of CMP release. Homogenized milkhas a similar behaviour to skim milk. Onthe other hand, in HM-Tw, 1/l* and G′increase at lower levels of CMP released(75% and 90%, respectively) than the othertwo samples.

All rheology and DWS experimentswere also carried out with half the amountof rennet (results not shown). Althoughthe kinetics of coagulation were obviouslyslower due to the decreased amount of ren-net, the exact same behaviours wereobserved for the SM, HM and HM-Tw sys-tems, demonstrating that the effects shownin this work are not dependent on the kinet-ics of aggregation of casein micelles but arerather related to interaction effects.

Table I summarizes the differences inthe key parameters described above withthe corresponding statistical significance.The presence of casein-covered fatglobules significantly reduces the beginning

% CMP release30 40 50 60 70 80 90 100 110

G' (

Pa)

0

5

10

15

20

25

30

Nor

mal

ized

1/l*

0.98

1.00

1.02

1.04

1.06

1.08

1.10

Figure 7. Development of G′ (full symbols) and 1/l* (open symbols) as a function of the amount ofCMP released: skim milk (□,■); homogenized milk (○,●) and homogenized milk with Tween (M,N).

Renneting of homogenized milk 635

of interparticle interaction (initial increase in1/l*) as well as the coagulation point mea-sured by DWS and rheology. When theinterface is replaced by Tween, the reduc-tion is amplified. Table I also shows G′values of the milk gels 45 min after thegelation point. Statistically, there seems tobe no difference in the value of G′ 45 minafter the gelation point between the SMand HM (Tab. I). When Tween replacesthe milk proteins at the interface, the finalG′ value more than doubles compared tothat of SM.

The % strain at break of the gelsobtained from all the samples was calcu-lated as a change of 5% in G′ during thestrain sweeps (Tab. I). The presence of fatglobules does not alter the brittleness ofmilk gels, the gels show a break at about30% strain. However, the presence ofTween in HM induces a biphasic behaviourin the strain sweep; when an increased strainis applied to these gels, there is an initialchange in G′ (about 2%) at around 3.5%strain, meaning that the structure is some-how modified, but the linear viscoelasticrange still extends to 30% strain, as in SMand HM. Again, these tests were also per-formed on gels prepared with a loweramount of rennet, and the nature of theresults was fully conserved.

4. DISCUSSION

Homogenization reduces the size of thefat globules, creating a heterogeneous fatglobules distribution with casein-coatedglobules either free in the serum or clusteredwith casein micelles. Similar results wereobtained homogenizing whole milk with aMicrofluidizer [4].

It is established that when small molecu-lar weight surfactants are added to aprotein-stabilized emulsion, they migrateto the fat/water interface and displace thestabilizing proteins [15]. In this work, usingTween 20, it was possible to create a colloi-

dal system with a similar distribution of fatglobule sizes as in homogenized milk butwith a different interface, as Tween 20displaces most of the proteins present atthe interface (Figs. 2 and 3).

The primary phase of renneting was fas-ter in HM and HM-Tw than in SM, and inall samples, all of the κ-casein was stillavailable for hydrolysis (Fig. 4). Thehomogenization process forces caseinmicelles to adsorb and spread at the oil/water interface. It has been previouslyhypothesized that spreading of caseinmicelles at the interface increases the sus-ceptibility of κ-casein to enzymatic hydroly-sis by rennet [21]. The similar kineticsbetween HM and HM-Tw samples suggestthat even after displacement of the micellesfrom the interface by the surfactant, the sus-ceptibility to hydrolysis does not change.

It is important to note that it has previ-ously been shown that addition of Tween20 to skim milk increases the rate of CMPrelease [12]. Although the molecular mech-anism origin of this effect is not yet clear, ithas been hypothesized that Tween mole-cules modify the supramolecular organiza-tion of the casein micelles, most likely onthe surface. The results shown in this worksuggest that once the supramolecular struc-ture of the casein micelles has been modi-fied by surface adsorption, the addition ofTween 20 no longer affects the primaryphase of renneting.

Both rheology and light scattering mea-surements showed a decrease of the gelationpoint of homogenized milk compared toskim milk. This can be attributed to the fas-ter removal of κ-casein in the first stages ofthe enzymatic reaction and is in full agree-ment with the results obtained by measuringthe CMP release (Figs. 4 and 7). As themicelles are stripped of their stabilizinglayer, they are able to come into close con-tact and aggregate earlier.

The addition of Tween 20 to homoge-nized milk significantly accelerates thesecondary stage of gelation. There was also

636 G. Ion Titapiccolo et al.

a significant difference in the elastic modu-lus for the gels prepared from homogenizedmilk with and without Tween 20. This waspreviously observed for skim milk samples(with no oil droplets) in the presence ofTween 20 [12]. These results are in contrastwith the understanding that Tween-coveredoil droplets would act as “inert” fillers inthe renneted casein network [2, 9]. The oildroplets in the HM-Tw samples are rela-tively inert, as shown by the behaviour ofthe MSD (Fig. 6, see below), and the differ-ences between HM-Tw and HM can beattributed to a number of factors. First ofall, the presence of Tween 20 affects boththe primary and secondary stages of theaggregation in skim milk samples [12]; inaddition, the HM-Tw, because of the proteindisplacement, has a higher amount of pro-tein present in solution compared to HMmilk, and finally, the displaced caseinmicelles may be more susceptible to aggre-gation because of surface denaturation.

The overall behaviour of 1/l* (Fig. 5B) forSM and all the homogenized milk sampleswas in full agreement with that reported inprevious studies on the rennet coagulationof casein micelles in skim milk [25]. Theonset of change of 1/l* is much earlier forHM-Tw than for homogenized and skimmilk, confirmingwhat is shown in Figure 5A(rheology results). Considering the inert roleof the fat globules (mostly covered withTween), it may be hypothesized that thecaseinmicelles aremoreprone to aggregationeither because of a direct effect of Tween 20on the structure of the casein micelles, orbecause of the presence of more protein insolution or, more likely, because of thedisplaced surface-denatured casein micelles.

The differences noticed in the develop-ment of the turbidity parameter, 1/l*, inthe two homogenized milk systems reflectthe different colloidal state of the filler par-ticles (Fig. 5B). In HM, the fat globules arean active part of the network, however,because of the increased volume fraction,the overall spatial distribution reaches its

final state (plateau value) much sooner thanin SM. On the other hand, when Tween 20is added to HM, the fat globules shall notinteract with the caseins and contribute tothe formation of a gel, and they will simplyreside in the pockets created by the gel net-work. Only once the casein gel formed andthe fat globules are “in place”, their averagepositions will not change with time. Thisforcing of the droplets into pockets alsoexplains the much larger change in the rela-tive value of 1/l*: the spatial distribution ofthe oil droplets after the formation of a gel ismore different from the liquid state than thespatial distribution of the oil droplets inHM gels.

By monitoring with DWS the increase inradius (or decrease in diffusion) over time itcan be noticed that skim milk and HM had asimilar gel-driving process, as the rate of“decrease of diffusion” was high and rela-tively similar (Fig. 5C). This is in agreementwith the current understanding that homog-enized milk fat globules are fully incorpo-rated in the casein network, and as such,slow down as the network formation pro-gresses. On the other hand, in HM-Tw, theapparent radius reaches a plateau quitequickly, revealing the presence of a systemin dynamic steady-state. This is in fullagreement with the results shown inFigure 5B and can be interpreted by theenclosure of the fat droplets, un-aggregated,inside the gel protein network. This hasbeen previously shown also for rennetingof recombined milk [7, 8].

The particle dynamics can be also mon-itored with the MSD parameter as a func-tion of time. At the gelation point, thescattering particles in SM and HM becometrapped in a network; their arrest in motionis shown in the deviation from linearity ofthe MSD curves (Fig. 6). However, as theTween-covered fat globules do not interactwith the network, HM-Tw samples do notshow this transition. The MSD dependencewith time remained linear throughoutthe duration of the experiment, in spite of

Renneting of homogenized milk 637

the formation of a gel (as shown by theincrease of the elastic modulus, G′, shownin Fig. 3). This behaviour is consistent withthe notion that Tween-covered fat globulesdo not take active part in the formation ofthe casein gel. The fat globules are trappedin the network cages, but within the lengthscale probed (< 10 nm, see Fig. 6) they takeunimpeded excursions (they are not an inte-gral part of the gel strand).

It is therefore possible to conclude thathomogenized milk showed a coagulationbehaviour similar to that of skim milk; thepresence of casein micelles at the fat inter-face did not modify the secondary phaseof renneting. In fact, the micelles in homog-enized milk aggregate earlier than in SM asa direct consequence of the acceleration ofthe first phase of renneting (see Figs. 5and 7). On the other hand, in HM-Tw lessCMP has to be removed to destabilize thesystem. This lower level of proteolysisrequired for coagulation leads to a fasteraggregation of HM-Tw casein micellescompared to HM. The earlier aggregationcan be attributed to an effect of Tween 20on the casein micelles and not to an effectof non-interacting fat globules. It has beenpreviously shown [12] that the addition ofTween to skim milk not only affects the pri-mary stages of renneting but also acceleratesthe second phase of the coagulation process,reducing the amount of CMP that needsto be released to destabilize the systemtowards aggregation. In the HM-Tw, it isalso important to consider that there wasan increased presence of surface-denaturedcasein micelles in solution during renneting.

The increase in G′ of HM-Tw gels can-not be related only to the increase in volumefraction of the system, as HM has the samecolloidal content as HM-Tw, and theTween-stabilized oil droplets are not show-ing significant aggregation (see Fig. 6) norinteractions with the caseins. Previous workshowed a significant increase in the stiffnessof the gels in the presence of Tween whenno fat globules are present [12], therefore

it is possible to conclude that the increasein stiffness does not depend on the Tween-stabilized fat globules – but from the effectof Tween molecules on the structure of thecasein micelles.

This analysis of the physico-chemicalproperties of the gels and their structuringproperties suggests that, by fine-tuning theinteractions occurring during the formationof the renneted casein network, it may bepossible to design gels with different tex-tural and sensorial properties.

Acknowledgements: This work was partlyfunded by the Ontario Dairy Council and theNatural Sciences and Engineering Council ofCanada. In addition, funding by Kraft FoodsR&D (Chicago, IL) is gratefully acknowledged.

REFERENCES

[1] Cano-Ruiz M.E., Richter R.L., Effect ofhomogenization pressure on the milk fatglobule membrane proteins, J. Dairy Sci. 80(1997) 2732–2739.

[2] Cho Y.H., Lucey J.A., Singh H., Rheologicalproperties of acid milk gels as affected bythe nature of the fat globule material andheat treatment of milk, Int. Dairy J. 9 (1999)537–545.

[3] Dalgleish D.G., Alexander M., Corredig M.,Studies of the acid gelation of milk usingultrasonic spectroscopy and diffusing wavespectroscopy, Food Hydrocoll. 18 (2004)747–755.

[4] Dalgleish D.G., Tosh S.M., West S., Beyondhomogenization: the formation of very smallemulsion droplets during the processing ofmilk by a Microfluidizer, Neth. Milk Dairy J.50 (1996) 135–148.

[5] de Kruif C.G., Casein micelles interactions,Int. Dairy J. 9 (1999) 183–188.

[6] Everett D.W., Olson N.F., Dynamic rheologyof renneted milk gels containing fat globulesstabilized with different surfactants, J. DairySci. 83 (2000) 1203–1209.

[7] Gaygadzhiev Z., Alexander M., CorredigM., Sodium caseinate stabilized fat globulesinhibition of the rennet-induced gelation

638 G. Ion Titapiccolo et al.

of casein micelles studied by diffusing wavespectroscopy, Food Hydrocoll. 23 (2009)1134–1138.

[8] Gaygadzhiev Z., Corredig M., AlexanderM., The impact of the concentration ofcasein micelles and whey protein-stabilizedfat globules on the rennet-induced gelationof milk, Colloid Surf. B 68 (2009) 154–162.

[9] Gaygadzhiev Z., Hill A., Corredig M.,Influence of emulsion droplet type on therheological characteristics and microstruc-ture of rennet gels from reconstituted milk,J. Dairy Res. 76 (2009) 1–7.

[10] Guinee T.P., Gorry C.B., O’Callaghan D.J.,O’Kennedy B.T., O’Brien N., Fenelon M.A.,The effects of composition on the rennetcoagulation properties of milk, Int. J. DairyTechnol. 50 (1997) 99–106.

[11] Horne D.S., Steric stabilization and caseinmicelle stability, J. Colloid Interface Sci. 111(1986) 250–260.

[12] Ion Titapiccolo G., Corredig M., AlexanderM., Modification to the renneting function-ality of casein micelles caused by non-ionicsurfactants, J. Dairy Sci. 93 (2010) 506–514.

[13] Jana A.H., Upadhyay K.G., A comparativestudy of the quality of Mozzarella cheeseobtained from unhomogenized and homog-enized buffalo milks, Cult. Dairy Prod. J. 28(1993) 16–22.

[14] Lopez C., Dufour E., The composition of themilk fat globule surface alters the structuralcharacteristics of the coagulum, J. ColloidInterface Sci. 233 (2001) 241–249.

[15] Mackie A.R., Gunning A.P., Wilde P.J.,Morris V.J., Orogenic displacement of pro-teins from the air/water interface by com-petitive adsorption, J. Colloid Interface Sci.210 (1999) 157–166.

[16] Maret G., Wolf P.E., Multiple light scatteringfrom disordered media. The effect ofBrownian motion of scatterers, Z. Phys.B – Condensed Matter 65 (1987) 409–413.

[17] McMahon D.J., Brown R.J., Enzymaticcoagulation of casein micelles: a review,J. Dairy Sci. 67 (1984) 919–929.

[18] Michalski M.C., Cariou R., Michel F.,Garnier C., Native vs. damaged milk fatglobules: membrane properties affect theviscoelasticity of milk gels, J. Dairy Sci. 85(2002) 2451–2461.

[19] Mulder H., Walstra P., The milk fat globule-emulsion science as applied to milk productsand comparable foods, Commonwealth

Agricultural Bureaux, Farnham Royal,Bucks, UK, 1974.

[20] Peters I.I., Cheddar cheese made frompasteurized milk homogenized at variouspressure, J. Dairy Sci. 39 (1956) 1083–1088.

[21] Robson E.W., Dalgleish D.G., Coagulationof homogenized milk particles by rennet,J. Dairy Sci. 80 (1984) 1901–1907.

[22] Romer S., Sheffold S., Shurtenberger P.,Sol-gel transition of concentrated colloidalsuspensions, Phys. Rev. Lett. 85 (2000)4980–4983.

[23] Rosa P., Sala G., van Vliet T., Van De VeldeF., Cold gelation of whey protein emulsions,J. Text. Stud. 37 (2006) 516–537.

[24] Sandra S., AlexanderM., DalgleishD.G., Therennet coagulation mechanism of skim milkobserved by diffusing wave spectroscopy,J. Colloid Interface Sci. 308 (2007) 364–373.

[25] Sandra S., Dalgleish D.G., The effect of ultrahigh-pressure homogenization (UHPH) onrennet coagulation properties of unheatedand heated fresh skim milk, Int. Dairy J. 17(2007) 1043–1052.

[26] Sharma R., Dalgleish D.G., Interactionsbetween milk serum proteins and syntheticfat globule membrane during heating ofhomogenized whole milk, J. Agric. FoodChem. 41 (1993) 1407–1412.

[27] tenGrotenhuis E., PaquesM., vanAkenG.A.,The application of diffusing wave spectros-copy to monitor the phase behaviour ofemulsion–polysaccharide systems, J. ColloidInterface Sci. 227 (2000) 495–504.

[28] Tunick M.H., Effects of homogenization andproteolysis on free oil in Mozzarella cheese,J. Dairy Sci. 77 (1994) 2487–2493.

[29] van Vliet T., Dentener-Kikkert A., Influenceof the composition of the milk fat globulemembrane on the rheological properties ofacid milk gels, Neth. Milk Dairy J. 36 (1982)261–265.

[30] Walstra P., Bloomfield V.A., Jason Wei G.,Jenness R., Effect of chymosin action on thehydrodynamic diameter of casein micelles,Biochim. Biophys. Acta 669 (1981) 258–259.

[31] Weitz D.A., Zhu J.X., Durian D.J., Gang H.,Pine D.J., Diffusing-wave spectroscopy: Thetechnique and some applications, Phys. Scr.T49B (1993) 610–621.

[32] Zoon P., van Vliet T., Walstra P., Rheolog-ical properties of rennet-induced skim milkgels. 3. The effect of calcium and phosphate,Neth. Milk Dairy J. 42 (1988) 295–312.

Renneting of homogenized milk 639