1209

Braz J Med Biol Res 38(8) 2005

Pressure-induced dissociation of casein micellesBrazilian Journal of Medical and Biological Research (2005) 38: 1209-1214ISSN 0100-879X

Pressure-induced dissociation of caseinmicelles: size distribution and effect oftemperature

1Physics Department E13, Technical University Munich, Garching, Germany2Department for Food and Nutrition, Freising-Weihenstephan, Germany

R. Gebhardt1,W. Doster1 and

U. Kulozik2

Abstract

Pressure-induced dissociation of a turbid solution of casein micelleswas studied in situ in static and dynamic light scattering experiments.We show that at high pressure casein micelles decompose into smallfragments comparable in size to casein monomers. At intermediatepressure we observe particles measuring 15 to 20 nm in diameter. Thestability against pressure dissociation increased with temperature,suggesting enhanced hydrophobic contacts. The pressure transitioncurves are biphasic, compatible with a temperature (but not pressure)-dependent conformational equilibrium of two micelle species. Ourthermodynamic model predicts an increase in structural entropy withtemperature.

CorrespondenceW. Doster

Physikdepartment E13

Technische Universität München

D-85748 Garching

Germany

Fax: +49-89-289-12473

E-mail: wdoster@ph.tum.de

Presented at the 3rd International

Conference on High Pressure

Bioscience and Biotechnology,

Rio de Janeiro, RJ, Brazil,

September 27-30, 2004.

Reesearch supported by the Deutsche

Forschungsgemeinschaft, Research

Project “High Pressure Treatment

of Food”.

Received January 21, 2005

Accepted May 10, 2005

Key words• Casein micelles• High-pressure light

scattering• Pressure dissociation

Introduction

Casein micelles are poly-disperse, roughlyspherical aggregates in milk with a meanradius of 150 nm. Their main physiologicaltask is to transport calcium, proteins andphosphorus to neonates. The micelles arecomposed of four different phospho-proteinsαS1-, αS2-, ß-, κ-casein and calcium phos-phate (1). The structure of casein micelles isstill a matter of debate (2-4). Casein sub-micelle models propose smaller units con-sisting mainly of caseins, which are linkedby small calcium phosphate clusters (5), whileother models deny casein sub-micelles andconsider calcium phosphate cluster as grow-ing centers for caseins (6). The dual bindingmodel of Horne (7,8) is based on primarystructures of caseins, which show distinct

hydrophilic and hydrophobic regions alongthe protein chain. According to this model,caseins can be described as block copoly-mers. The hydrophilic regions in αS1-, αS2-,and ß-casein are rich in phospo-serines, towhich the colloidal calcium phosphate par-ticles become attached. The casein micelle isstabilized by a hydrophobic core and hydro-philic fringes on the surface and by electro-static coupling to calcium phosphate par-ticles.

Both hydrophobic and electrostatic inter-actions decrease with pressure, which gener-ally leads to dissociation and unfolding ofproteins. Casein micelles disintegrate irre-versibly under pressure.

As demonstrated by ex-situ measurementsusing static and dynamic light scattering (3,9)and imaging techniques such as electron mi-

1210

Braz J Med Biol Res 38(8) 2005

R. Gebhardt et al.

croscopy (10,11) or atomic force micros-copy (12), high pressure reduces the hydro-phobic interactions between the caseins in-side the micelle. Increased exposure of hy-drophobic groups was observed after pres-sure treatment of milk proteins (13). More-over, pressure affects the mineral equilib-rium of calcium, phosphorus and other com-ponents. Conflicting data from ex-situ meas-urements related to the calcium equilibriumhave been presented: no effect of pressureon the free calcium concentration was foundby Shibauchi et al. (14) while Anema et al.(9) found an increased calcium concentra-tion after a pressure treatment of 200 MPa.

Material and Methods

Casein micelles were extracted from com-mercial-grade skim milk by combined uni-form transmembrane pressure micro-filtra-tion (mean pore diameter: 0.1 µm) and ultra-filtration, concentrated by five washing stepsand dried in a spraying tower. The fraction-ation procedure has been described in detailby Tolkach and Kulozik (15). Casein pow-der was dissolved in filtered 0.1 M Mes/Tris-HCl solution at a concentration of 40mg/ml and the pH was adjusted to 7.3 withHCl. All solutions were equilibrated by thor-oughly stirring for 5 h at 20ºC.

Turbid casein solutions (3%) were stud-ied by photon correlation spectroscopy in aback-scattering geometry to reduce multiplescattering of light. An He-Ne-laser lightsource (λ = 632.4 nm) from an ALV-NIBSSystem (ALV-Laser GmbH, Langen) wasfocused on an optical SITEC high-pressurecell. In dilute solutions the single-scatteredintensity of light is proportional to the squareof the molecular weight, the concentrationand the molecular form factor. The scatteredintensity is thus heavily biased by high-molecular weight particles.

We employed dynamic light scatteringto analyze the molecular size distribution.These experiments showed that the pres-

sure-induced dissociation process essentiallyinvolves a high-molecular weight species,the intact micelle and low-molecular weightfragments. For the present purpose, we thusused the average light intensity as an empiri-cal indicator of the degree of dissociation ofthe intact micelle since the fragments do notcontribute significantly to the scattering sig-nal. After subtraction of a background inten-sity I (P = 400 MPa) at high pressure wedefined a dissociated fraction by normaliz-ing the measured intensity I(P) by I(P = 1bar):

α = {I(P) - I(P = 400 MPa)}/{I(P = 1 bar) -I(P = 400 MPa)} (Eq. 1)

The size distribution was derived by ex-panding the measured scattered field auto-correlation function, g1(t) in terms of expo-nentials:

(Eq. 2)

where G(Γ) is the distribution of relaxationrates, Γ(q) = q2D(R), where q denotes thescattering vector. D(R) is the diffusion coef-ficient depending on the effective hydrody-namic radius according to the Stokes-Einsteinrelation: D = KT/(6πηR), η is the viscosityof the solvent.

G(Γ) was derived using the Laplace-in-version program CONTIN. For a particularvalue of R the distribution is given by Pz(R)= n(R) . M(R)2, where n is the number den-sity of particles with radius R and M is theirmolecular mass. Assuming spherical, com-pact particles, M ∝ R3, one can derive anestimate of the number distribution n(R).

Results

Dynamic and static light scattering ex-periments were performed at 177º with tur-bid casein solutions at various pressures.Figure 1 displays the distribution Pz(R) at0.1 MPa, which is dominated by micelleswith diameters near 300 nm. The micelles

1211

Braz J Med Biol Res 38(8) 2005

Pressure-induced dissociation of casein micelles

were rather poly-disperse, with the half widthof the distribution corresponding to 100 nm.At 100 MPa the turbidity had nearly van-ished, but the z-average of R was still domi-nated by 300-nm particles. At low turbidityit became possible to determine the numberaverage, which is shown in Figure 1. Now thedominant particle size was 15-20 nm. At 300MPa most fragments were reduced to the mono-mer level with diameters close to 3 nm.

Figure 2 displays pressure-temperaturedissociation plots derived from the averagelight scattering intensity in 177º back-scat-tering geometry. The dissociation of caseininto small fragments, M → F, was completeat 350 MPa at all temperatures. The locationof the midpoint of the transition on the pres-sure scale increased with increasing temper-ature. The forces that stabilize the micelle,thus increased with temperature. Moreover,the transition curves were not as smooth asexpected for a simple two-state mechanism.In the temperature range between 30 and60ºC, a plateau emerged, suggesting the ex-istence of an intermediate state.

Since the fragments differed in size andmolecular weight from the intact micelle atleast by a factor of ten, they contributed onlyweakly to the average light scattering inten-sity. The measured dissociated fraction α(Equation 1) is thus dominated by the high-molecular weight micelle fraction. Thismeans that the plateau in α at the 50% levelin Figure 2 cannot be attributed to a partiallydissociated intermediate state with two frag-ments of approximately equal size. Since wedo not observe fragments of the appropriatesize (100 nm), a purely sequential dissocia-tion process is not compatible with the data.Instead we suggest a parallel model, whichassumes two types of casein micelles, whichdiffer in stability.

At low temperatures the dominant formwas the pressure-unstable state ML, whilewith increasing temperature a high-tempera-ture form, MH, was increasingly present andwas more stable.

Thermodynamic analysis of pressure-temperature dissociation of casein micelles

Pressure-induced dissociation of caseinmicelles is not fully reversible due to disso-ciation between colloidal calcium and phos-

1.0

0.8

0.6

0.4

0.2

0.0

Siz

e di

strib

utio

n

1 10 100 1000Diameter [nm]

300 MPa 100 MPa 0.1 MPa

Figure 1. In situ size distribution of casein particles at three pressures derived from theintensity correlation function, z-distribution at 0.1 MPa and number distribution at 100 and300 MPa (see text).

Figure 2. Temperature-pressure profile of the dissociated fraction of casein micelles (pH7.3) and simultaneous fit according to the model of Equations 4-8.

1.0

0.8

0.6

0.4

0.2

0.0

8070

6050

4030

2010

50

100

150200

250

350

T [ C]

Deg

ree

of d

isso

ciat

ion

p[M

Pa]

300

1212

Braz J Med Biol Res 38(8) 2005

R. Gebhardt et al.

phate particles (11). This process, which isassociated with the formation of 15- to 20-nm particles (Figure 1), is slow. Thus, on atime scale of 30 min one can determinestable, time-independent values of scatteredlight. With this restriction, we performed athermodynamic analysis of the transitioncurves. Next we consider an equilibrium oftwo types of micelles, ML and MH, whichdissociated under pressure into fragments FL

and FH:

ML MH

FHFL

L (T)L,H

L (T)L,H

K (P)HK (P)L

(Eq. 3)

The difference in chemical potential be-tween ML and MH follows from the Gibbs-Duhem relation:

(Eq. 4)

∆µ0 is the potential of the reference state,∆SHL denotes the difference in molar en-tropy and ∆VHL denotes the respective vol-ume difference. ∆CP is the difference heatcapacity at constant pressure between the Hand L forms. The difference in chemicalpotential vanishes in equilibrium, ∆µHL = 0.Thus, at the reference temperature and pres-sure one has:

∆µ0 = -RT ln LHL (Eq. 5)

The equilibrium constant, LHL, is definedby the ratio of the molar concentrations if thesolution is diluted: LHL = NH/NL. If the refer-ence temperature and pressure, T0 and P0, arechosen at the midpoint of the HL equilib-rium, NH = NL, and the reference potentialvanishes: ∆µ0 = 0.

The equilibrium is controlled by the en-tropy difference ∆SHL(T), which decreaseswith increasing temperature. This implies a

finite difference in the heat capacities, ∆CHL

≠ 0.We can fit the data ignoring the pressure

dependence of the conformational equilib-rium; thus ∆VHK ≈ 0.

The fraction of the dissociated conformerαH/L H or L is controlled by the pressure-dependent constants KL and KH(P):

(Eq. 6)

and

ln KL/H,F(P) = ln KL/H,F(P = 0) - 1/(RT) × P ×∆VH/L,F + ∆SH/L,F/R (Eq. 7)

∆VH/L,F and ∆SH/L,F are the volume and en-tropy differences between the H or L formand its fragments F. The combined dissocia-tion fraction is then given by:

(Eq. 8)

In its present form the model is over-determined with 8 parameters. The simplestmodel fitting the data was based on theassumption that the H ↔ L equilibrium isonly temperature dependent, ∆VHL ≈ 0 andthat the dissociation of H or L into fragmentsis only pressure but not temperature depend-ent. This implies ∆SH/L,F ≈ 0. Global fits to alltransition curves were performed, and arerepresented by the full lines in Figure 2. Thecorrelation coefficient of the fit was 0.98 andthe residuals did not exhibit any systematicdeviations. The resulting parameters are listedin Table 1. The dissociation constant KL(0)= 0.09 of the low temperature form is small,favoring intact micelles at ambient pressure.The dissociation volume, ∆VLF = -61 ± 7 ml/mol, is negative and surprisingly small. Thedissociation constant of the high tempera-ture form H, KH(0) = 9.5 x 10-10, is negligibleat ambient pressure, while the dissociationvolume, ∆VHF = -204 ± 3 ml/mol, is againnegative and significantly larger in compar-ison with the L form. The positive ∆SHL =150 ± 10 J/(mol . K) implies that the H form

1213

Braz J Med Biol Res 38(8) 2005

Pressure-induced dissociation of casein micelles

is more stable than the L form at high tem-perature and vice versa. Introducing a finiteheat capacity change noticeably improvesthe quality of the fits. The resulting differ-ence, ∆CHL ≈ -0.5 kJ/(mol . K), is smallerthan heat capacity changes found in denatur-ation experiments with small proteins. How-ever, the sign of ∆CHL is negative, implyingthat the H-form exhibits a lower heat capac-ity than the L-form in spite of its largerentropy. Figure 3 shows the resulting tem-perature-dependent populations of H and L.The pressure-induced dissociation transitionis monophasic at high temperatures sinceonly the high-temperature form H is present.At low temperatures both species, L and H,contribute. The midpoint temperature is T0 =307 ± 2 K, independent of the pressure.

In conclusion, in situ light scatteringmeasurements demonstrate that casein mi-celles (d ≈ 200 nm) disintegrate at pressuresabove 300 MPa into 3-nm particles, mostlikely casein monomers or small oligomers.We also observe particles of intermediatesize near 15 nm, which represent either in-termediate states or aggregates of caseinmonomers, which evolve a posteriori to pres-sure dissociation. The pressure required todissociate the micelles increases with tem-perature in accordance with hydrophobicstabilization. A continuous enhancement ofhydrophobic interactions would simply shiftthe midpoint of the transition on the pressurescale. Instead we observe two discrete stepswith a temperature-dependent partition: withincreasing temperature the fraction of a lowtemperature form (L) decreases in favor of ahigh temperature form (H), which is morepressure stable.

The thermodynamic analysis based onEquation 3 indicates that the H-form is moredisordered, T0

. ∆SHL = 46 kJ/mol (T0 = 307K), and exhibits a larger dissociation vol-ume, 204 versus 61 ml/mol, than the L-micelle (Table 1). This could suggest thatthe high-temperature form has an open struc-ture, while the L-form is more compact.

However, the H ↔ L equilibrium is approxi-mately pressure independent, and thus thevolume difference ∆VHL must be small. Thisimplies that the fragments of H and L, FL andFH differ in volume. However, due to thesmall size of the fragments they contributevery little to the observed scattering inten-sity. Thus, no information about the FL ↔ FH

equilibrium is obtained from the light scat-tering data. The apparent volume differenceis derived from the width of the pressuretransition. It is likely that the apparent differ-ence in the dissociation volumes, ∆VHF, ∆VLF,reflects a difference in heterogeneity of thesize and stability of the L and H micelles,which will broaden the transition. The disso-ciation constants of H and L, KH(P) andKL(P), depend only on the pressure but noton the temperature. This implies that theentropic difference between micelle and frag-ments must be small, ∆SH/L,F ≈ 0. Since Hand L differ in entropy, one must conclude

Table 1. Parameters of simultaneous fit to themodel of Equation 3.

Parameters Values at pH 7.3

KL(0) 0.09 ± 0.02∆VLF -61 ± 7 ml/molKH(0) 9.5 x 10-10

∆VHF -204 ± 3 ml/molT0 307 ± 2 K∆SHL(T0) 149 ± 10 J/(mol . K)T0 ∆SHL 46 kJ/mol∆CHL -0.5 ± 0.15 kJ/(mol . K)

T [°C]

0 10 20 30 40 50 60 70 80

Fra

ctio

n

0.00

0.25

0.50

0.75

1.00 MHML

Figure 3. Temperature-depend-ent population the low (L) andhigh (H) temperature conformerof casein micelles at ambientpressure according to the modelof Equation 3.

1214

Braz J Med Biol Res 38(8) 2005

R. Gebhardt et al.

that ∆SHL is stored in the respective frag-ments ∆SHL ≈ ∆SF,H/L. We observe indeedtwo distinct fragments corresponding to par-ticle sizes near 3 nm and 15 nm. But thearguments concerning the fragments givenabove apply again. Moreover, the pressure-induced dissociation process is only par-tially reversible. The main result of this insitu pressure work is the finding of two statesof native-like casein micelles, which differ

in sensitivity to pressure dissociation. The Hform is dominant at high temperatures. Weare presently investigating whether the twostates differ in the manner how calcium phos-phate particles are bound. The model shouldbe considered as rather preliminary, and fur-ther extensions for explicit inclusion of het-erogeneity and size distribution are inprogress.

References

1. Walstra P & Jenness R (1984). On the stability of casein micelles.Journal of Dairy Science, 73: 1965-1979.

2. Dalgleish DG, Spagnuolo PA & Goff HD (2004). A possible structureof the casein micelle based on high-resolution field-emission scan-ning electron microscopy. International Dairy Journal, 14: 1013-1089.

3. Desobry-Banon S, Richard F & Hardy J (1994). Study of acid andrennet coagulation of high pressurized milk. Journal of Dairy Sci-ence, 77: 3267-3274.

4. McMahon DJ & McManus WR (1998). Rethinking casein micellestructure using electron microscopy. Journal of Dairy Science, 81:2985-2993.

5. Walstra P (1999). Casein sub-micelles: do they exist? InternationalDairy Journal, 9: 189-192.

6. Holt C, de Kruif CG, Tuinier R et al. (2003). Substructure of bovinecasein micelles by small-angle X-ray and neutron scattering. Col-loids and Surfaces A, 213: 275-284.

7. Horne DS (1998). Casein interactions: casting light on the blackboxes, the structure in dairy products. International Dairy Journal, 8:171-177.

8. Horne DS (2003). Casein micelles as hard spheres: limitations ofthe model in acidified gel formation. Colloids and Surfaces A: Physi-cochemical and Engineering Aspects, 213: 255-263.

9. Anema SG, Lee SK, Schrader K et al. (1997). Effect of pH on the

turbidity of pressure-treated calcium caseinate suspensions andskim milk. Milchwissenschaft, 52: 141-146.

10. Needs EC, Stenning RA, Gill AL et al. (2000). High pressure treat-ment of milk: Effects on casein micelle structure and on enzymecoagulation. Journal of Dairy Research, 67: 31-42.

11. Keenan RD, Hubbard CD, Mayes DM et al. (2003). Role of CalciumPhosphate in High Pressure-Induced Gelation of Milk. Food Col-loids-Biopolymers and Materials, Royal Society of Chemistry, Cam-bridge, UK, 109-118.

12. Regnault S, Thiebaud M, Dumay E et al. (2004). Pressurisation ofraw skim milk and of a dispersion of phosphocaseinate at 9ºC or20ºC: effects on casein micelle size distribution. International DairyJournal, 14: 55-68.

13. Gaucheron F, Famelart M, Mariette F et al. (1997). Combined ef-fects of temperature and high pressure on physicochemical charac-teristics of skim milk. Food Chemistry, 59: 439-447.

14. Shibauchi Y, Yamamoto H, Sagara Y et al. (1992). Pressure-in-duced conformational changes of casein micelles. In: Heremans K &Masson P (Editors), High Pressure and Biotechnology. ColloqueINSERM. Vol. 224. John Libbey Eurotext Ltd., London, UK.

15. Tolkach A & Kulozik U (2005). Fractionation of whey proteins andcaseino-macropeptide by means of enzymatic cross-linking andmembrane separation techniques. Journal of Food Engineering, 67:13-20.