ARTICLE IN PRESS

International Dairy Journal 16 (2006) 361–369

www.elsevier.com/locate/idairyj

The stabilizing behaviour of soybean soluble polysaccharide and

pectin in acidified milk beverages
Akihiro Nakamuraa,b, Ryuji Yoshidab, Hirokazu Maedab, Milena Corrediga,
a
Department of Food Science, University of Guelph, Guelph, Ontario, Canada N1G 2W1
b
Food Science Research Institute, Tsukuba R&D Center, Fuji Oil Co. Ltd., 4-3 Kinunodai, Yawara-mura, Tsukuba-gun, Ibaraki 300-2497, Japan
Received 9 September 2004; accepted 8 January 2005

Abstract

The mechanisms of stabilization of soybean soluble polysaccharide (SSPS) and high methoxyl pectin (HMP) in acidified milk
drinks were studied focusing on the differences in behaviour between the two polysaccharides. The changes in casein micelles size
during acidification with glucono-d-lactone or by direct acidification were measured using light scattering. When HMP was added to
skim milk before acidification, pectin adsorbed on the surface of the casein micelles via electrostatic interactions and prevented
casein aggregation. Results suggested that adsorption of pectin occurred from the beginning of acidification and somewhat affected
the rearrangement of casein micelles in the pH range between 5.8 and 5.0. On the other hand, SSPS, at concentrations up to 2%
(w/w), did not interact with caseins at pH 44.6. At pH o4.2 SSPS showed better stabilizing properties than HMP. In addition,
between pH 4.2 and 3.2, SSPS-stabilized acid dispersions were not affected by pH, while dispersions homogenized with pectin
showed a size distribution that depended on pH. The differences in structure between SSPS and HMP account for the unique
functionalities of the two polysaccharides in acid milk systems.
r 2005 Published by Elsevier Ltd.

Keywords: Soybean soluble polysaccharide; Casein micelles; Stabilization; Acidified milk

1. Introduction When charged polysaccharides are added to protein

solutions, complexes may be formed depending on pH
Acidified milk drinks comprise a large range of and ionic strength (Imeson, Ledward, & Mitchell, 1977).
products, from those usually prepared from fermented The interactions between proteins and polysaccharides
milk with stabilizers and sugar to those prepared by are typically electrostatic in nature (Burgess, 1990;
direct acidification with fruit juices and/or acids. The pH Galazka, Smith, Ledward, & Dickinson, 1999; Schmitt
of these products ranges from 3.4 to 4.6, and because of et al., 2000; Syrbe, Bauer, & Klostermeyer, 1998).
the instability of caseins in this pH range, a stabilizer Caseins at neutral pH in milk are present in the form of
needs to be added to prevent protein aggregation and micelles, and these protein particles are stabilized by
achieve optimal mouth feel. Stabilizers are important to steric repulsive interactions (De Kruif, 1998). During
control properties such as texture, viscosity or mouth acidification, at a pH around the isoelectric point (pH
feel; correct ingredient formulation and appropriate 4.6) the micelles aggregate, mainly because of the collapse
processing ultimately affect consumer’s acceptability of the extended layer formed by k-casein on the surface of
(De Kruif & Tuinier, 2001; Syrbe, Fernandes, Dannen- the micelle (De Kruif, 1998; Holt, 1982). At a pH below
berg, Bauer, & Klostermeyer, 1995; Tostoguzov, 1986). their isoelectric points, milk proteins are positively
charged; negatively charged pectins are employed as
Corresponding author. Tel.: +1 519 824 4120; stabilizers in products such as acidified milk drinks and
fax: +1 519 824 6631. yoghurt (Cai & Arntfield, 1997; Glahn, 1982; Mishra,
E-mail address: [email protected] (M. Corredig). Mann, & Joshi, 2001; Pereyra, Schmidt, & Wicker, 1997).

0958-6946/$ - see front matter r 2005 Published by Elsevier Ltd.

doi:10.1016/j.idairyj.2005.01.014
 ARTICLE IN PRESS
362 A. Nakamura et al. / International Dairy Journal 16 (2006) 361–369

Pectin, an anionic polysaccharide generally extracted skim milk powder was donated by Parmalat (Toronto,
from citrus fruits or apple pomace, is mainly composed Ont., Canada). SSPS was obtained from Fuji Oil Co. Ltd.,
of a backbone of galacturonic acid, partly methyl (Osaka, Japan); it was prepared from the residue of protein
esterified, and by hairy regions containing rhamnose, extraction, as previously described (Nakamura et al., 2001;
galactose, arabinose, xylose, and glucose (Voragen, Nakamura, Takahashi, Yoshida, Maeda, & Corredig,
Pilnik, Thibault, Axelon, & Renard, 1995). It has been 2004). HMP was donated by CpKelco (Genu Pectin YM-
demonstrated that, below pH 5.0, the polysaccharide 100L, CpKelco, San Diego, CA).
chains of high methoxyl pectin (HMP) adsorb on the
surface of the casein micelles, via the charged blocks of 2.2. Acidification of skim milk solution by GDL
the pectin chains and the uncharged blocks of HMP
extend into solution contributing to stabilization via Skim milk powder was dissolved in MilliQ water,
steric repulsion (Pereyra et al., 1997; Tuinier, Rolin, & containing 0.02% (w/v) sodium azide as an antimicro-
De Kruif, 2002). bial to obtain a final concentration of 17% total solids
Soybean soluble polysaccharide (SSPS) extracted (w/w). Milk was heat treated at 80 1C for 10 min and
from soybean cotyledons is often used to stabilize cooled to 20 1C. To reach a certain end-pH after
acidic beverages (Asai et al., 1994). SSPS is an acidic overnight fermentation, different amounts of GDL from
polysaccharide containing 18% galacturonic acid 1.5% to 4.0% were added to heated skim milk and
(Nakamura, Furuta, Maeda, Nagamatsu, & Yoshimoto, incubated at 30 1C. The pH during acidification was
2001; Nakamura, Furuta, Maeda, Takao, & Nagamat- measured at room temperature at appropriate intervals.
su, 2002). The main polysaccharide backbone consists of
galacturonan and rhamnogalacturonan, organized in 2.3. Acidification with GDL in the presence of stabilizer
diglycosyl repeating units, -4)-a-D-GalA-(1-2)-a-L-
Rha-(1-. The stabilizing mechanism of SSPS in acidified SSPS was dissolved in MilliQ water at various
milk has been previously investigated after digestion of concentrations in the range from 0.2% to 2.0% (w/w),
the sugar chains with purified pectinase or hemicellulase heated at 70 1C for 10 min and cooled to 20 1C. Pectin
(Nakamura, Furuta, Kato, Maeda, & Nagamatsu, was also dissolved in MilliQ water at concentrations
2003): in addition to the negatively charged galactur- ranging from 0.2% to 0.6% (w/w), heated at 80 1C for
onan chains, the branched portions, composed of 10 min and cooled to 20 1C. The stabilizer solution was
galactan and arabinan, are also important, contributing added to the heat-treated skim milk to obtain a final
to the formation of a thick hydrated layer on the surface skim milk solution of 8.5% total solids (skim milk and
of the casein micelles. However, the mechanism of stabilizer solution 1:1 diluted). GDL (1.5%) was added
stabilization of SSPS in fermented milks is not to lower the pH gradually, to mimic a lactic fermenta-
completely understood, and although SSPS stabilization tion and the mixture was kept at 30 1C in flasks
is often compared with that of HMP, differences exist in throughout the experiment. Milk with stabilizer (2 kg
their behaviour. total sample) was circulated continuously through a
The objective of the present work was to study the laboratory scale ultrafiltration cartridge (Millex-HV
mechanism of stabilization of SSPS in acidified milk CDUF001LG, Millipore Co. Ltd., Billerica, MA,
systems and to compare the behavior of SSPS with that USA) with a nominal molecular mass cut off of
of HMP. The particle size distribution of acidified milk 10,000 Da and a nominal area of 0.095 m2. The milk
drinks was measured as a function of the pH and of the was circulated without applying back pressure, using a
stabilizer concentration in various model systems. In peristaltic pump, and permeate and retentate were
addition, the changes in particle size of SSPS-stabilized continuously re-circulated, to avoid concentration. The
caseins after homogenization were measured using experiments were performed until the first signs of
pectinase to evaluate the thickness of the stabilizing aggregation, to avoid blockage of the membrane filter.
layer. The particle diameter of the casein micelles was
measured as a function of pH by dynamic laser light
scattering (DLS) using a Malvern 4700 optical system
2. Materials and methods (Malvern Instruments, Malvern, England) attached to a
7032 correlator at 25 1C and at a scattering angle of 901.
2.1. Materials Measurements were carried out after diluting the milk in
the corresponding permeate collected through the UF
Glucono-d-lactone (GDL) was obtained from Sigma- filter. To collect the permeate, the back pressure of the
Aldrich Chemical Co. Ltd. (St. Louis, MO, USA). Pectinase filter cartridge was increased by applying resistance on
(from Aspergillus aculeatus) and other analytical grade the flexible tubing on the retentate side. For each
reagents were also purchased from Sigma-Aldrich. MilliQ permeate sample a volume of 20 mL (corresponding to
water was used in the preparation of all solutions. Low heat the dead volume of the filter) was collected, returned to
 ARTICLE IN PRESS
A. Nakamura et al. / International Dairy Journal 16 (2006) 361–369 363

the circulating milk and then 2 mL were collected for the The apparent diameter of the casein micelles under
DLS experiment. Samples were immediately measured acidic conditions was also measured by DLS (Malvern
and the pH of the milk was measured in parallel so 4700, Malvern Instruments). All acidified milk samples
that the particle size data could be related to the were measured after dilution with freshly prepared
decrease in pH. permeate.

2.4. Addition of stabilizer after milk acidification with 2.7. Enzyme treatment of the acidified milk
GDL
The effect of hydrolysis by pectinase on the particle size
SSPS or HMP were dissolved in MilliQ water at of caseins stabilized by SSPS was measured by DLS as
concentrations of 0.4%, 0.8%, and 1.2% (w/w), heated previously described for SSPS-stabilized emulsions
at 70 1C for 10 min, cooled to 20 1C and then filtered (Nakamura, Yoshida, Maeda, Furuta, & Corredig, 2004).
through 0.45 mm filters (Millex-HV, Millipore Co. Ltd., Aliquots of pectinase solution (3–10 mL, 12 units mL 1 as
Billerica, MA, USA). GDL (2.5%) was added to 1 kg of a-polygalacturonase) were added to a clean cuvette contain-
17% (w/w) heated skim milk and the milk was incubated ing acidified milk previously diluted at a rate of 0.5 mL
at 30 1C overnight until a final pH of 4.2. To determine acidified milk per 4.0 mL of filtered serum. The permeates
the effect of pH on the stabilizing properties of SSPS were prepared from 200 g of the acidified milk solutions
and HMP, aliquots (50 g) of acidified skim milk were (pH 3.2–4.2) by circulation through the laboratory scale
mixed with 50 g of stabilizer solution and the pH was ultrafiltration cartridge (Millex, Millipore Co.) as previously
adjusted to 4.2, 4.0, 3.8, 3.6, 3.4, and 3.2 by adding 0.5 N described. The changes in particle diameter during hydro-
HCl. The mixtures were pre-homogenized with a shear lysis were measured at 35 1C with the DLS at 1 min intervals
mixer (Power Gen 125, Fisher Scientific, Nepean, Ont., for 30 min.
Canada) and then homogenized with a laboratory
homogenizer (Emulsiflex-C5, Avestin Inc. Ottawa,
Canada) with two passes at 40 MPa. 3. Results

3.1. Changes in particle size during acidification in the

2.5. Direct acidification of skim milk presence of stabilizer
Skim milk was reconstituted from low heat skim milk The interactions of SSPS and HMP with casein
powder to 17% total solids and heat treated as described micelles during milk acidification by GDL were studied
above. SSPS or HMP were dissolved in MilliQ water to by observing the changes in average diameter. The
a final concentration of 0.8% (w/w), heated at 80 1C for casein micelles were dispersed in permeate (collected by
10 min and cooled to 20 1C. Stabilizer and skim milk ultrafiltration) and their apparent diameter was mea-
were mixed together in a 1:1 ratio to obtain 1 kg of 8.5% sured by DLS as shown in Fig. 1. The values of pH were
skim milk containing 0.4% stabilizer. The mixture was also measured over time. Skim milk acidified in the
directly acidified to pH 4.2 by adding concentrated HCl absence of HMP or SSPS showed an average diameter
at 20 1C with continuous stirring using a magnetic of about 250 nm at the beginning of the acidification. In
stirrer. The pH was then decreased to 3.2 and 100 g of agreement with previous research, at a pH between 5.8
samples were collected at 0.2 intervals of pH. Aliquots and 5.1 a decrease in size of the casein micelles was
(100 g) of samples at the various pH values from 4.2 to shown. This decrease of about 25 nm in diameter with
3.2 were pre-homogenized with the shear homogenizer lowering of pH has been attributed to the collapse of the
(Power Gen 125, Fisher Scientific) for 5 min, and then k-casein brush (De Kruif, 1998; Tromp, De Kruif, Van
homogenized (Emulsiflex-C5, Avestin Inc.) with two Eijk, & Rolin, 2004; Tuinier et al., 2002). In the absence
passes at 40 MPa. of stabilizer, at pH o5.2, the apparent diameter of the
casein micelles showed a dramatic increase, indicating
2.6. Evaluation of acidified milk particle casein aggregation.
When skim milk was acidified in the presence of SSPS
The particle-size distribution of the casein micelles (Fig. 1A) (up to 2% SSPS), no difference was shown in
was measured using a Mastersizer X (Malvern Instru- the initial particle diameter compared with the skim
ments Ltd.) using the following parameters: a relative milk control. The amount of SSPS added to skim milk
refractive index of the particles of 1.06, a sample did not affect the apparent particle diameter of the
absorption of 0.001 and a refractive index of the solvent caseins during acidification, and the decrease in dia-
of 1.33. The stability of acidified milk was determined by meter with decreasing pH was comparable to that of the
comparing the shape of the distributions and the values unstabilized control samples. In the control samples, a
of the average droplet size (D[3,2] and D[4,3]). decrease in the particle diameter was observed at pH
 ARTICLE IN PRESS
364 A. Nakamura et al. / International Dairy Journal 16 (2006) 361–369

7.0
650
6.5
Particle Diameter (nm)

550
6.0
450

pH
5.5

350
5.0

250
4.5

150 4.0
0 20 40 60 80 100 120 140 160 180
(A) Time (min)

7.0
650
6.5
Particle Diameter (nm)

550
6.0

450

pH
5.5

350 5.0

250 4.5

150 4.0
0 20 40 60 80 100 120 140 160 180
(B) Time (min)

Fig. 1. Diameter of casein micelles as a function of time during acidification. Acidified milk solutions were diluted with permeates immediately before
measuring the particle diameter with DLS. Skim milk was acidified with GDL in the presence of SSPS (A) or HMP (B). () pH; (K) 0%; (J) 0.2%;
(n) 0.4%; (m) 0.6%; (&) 1.0%; (’) 2.0% stabilizer added.

45.2, and at lower pH aggregation occurred. At the casein aggregation was prevented with HMP at con-
concentrations used in this study, no interactions centrations 40.2%. The particle diameter before
seemed to occur between SSPS and casein micelles aggregation decreased in control samples of about
during acidification. This was in contrast with the 25 nm, and this decrease with acidification increased
behaviour observed with HMP. with pectin concentration, from 25 nm at low pectin
Fig. 1B illustrates the apparent diameter of casein concentration to 60 nm in the presence of 0.6% pectin.
micelles during acidification in the presence of HMP. This change in diameter with acidification corresponds
The initial diameter of the skim milk increased with to a much larger change than that attributable to the
increasing pectin concentration. This indicated an collapse of k-casein, indicating a possible rearrangement
association of pectin to the casein micelles at the initial of the casein micelles in the presence of pectin at pH
pH, possibly because of the presence of calcium ions. All 45.0. These results bring new evidence of non-specific
samples were extensively diluted in permeate immedi- interactions occurring between pectins and caseins at pH
ately before the DLS measurements and it is assumed values higher than the isoelectric point of the proteins.
that the measurement of particle diameter was not
affected by the bulk viscosity of the pectin. When the 3.2. Stabilizer behaviour in homogenized acid milk
concentration of pectin was sufficient, acid-induced dispersions
aggregation was inhibited. While the initial stages of
acidification showed no differences compared with skim Acidified milk was homogenized in the presence of
milk with no stabilizer or with SSPS added, at pH o5.2 SSPS or HMP to determine their behaviour at low pH.
 ARTICLE IN PRESS
A. Nakamura et al. / International Dairy Journal 16 (2006) 361–369 365

Skim milk was acidified with GDL until pH 4.2, and the independent of pH (Fig. 2B). These results suggested that
pH was then further adjusted to values from 4.2 to 3.2 homogenization of acidified milk with 40.4% SSPS
by direct acidification. The effect of varying the resulted in complete coverage of the positively charged
concentration of SSPS on the apparent diameter of the casein micelles with negatively charged SSPS molecules,
homogenized samples as a function of pH is summarized which prevented acid-induced casein aggregation.
in Fig. 2. In addition, the effect of pH on the particle size A very different stabilizing behaviour was shown for
distribution of the acidified suspensions containing acid milk homogenized with HMP in the same pH range
0.4% SSPS is shown in Fig. 2. Homogenization of (Fig. 3). The apparent diameter of pectin-covered
acidified milk in the presence of SSPS resulted in stable acidified casein aggregates increased with decreasing
dispersions with a monomodal distribution of sizes. In pH (Fig. 3A), confirming the importance of electrostatic
this range of pH, from pH 3.2 to 4.2, at a concentration interactions in the association of HMP with caseins.
of SSPS higher than a critical amount (0.3% SSPS), The dependence of particle size with pH was more
there was no effect of pH on the apparent diameter. pronounced in milk containing low amounts of pectin
Results summarized in Fig. 2A demonstrated that 0.2% (0.2% w/w), but it was evident at all the concentrations
SSPS was not sufficient to stabilize the dispersions at pH tested. At pH 4.0 the apparent diameter of acid milk
44.0. The differences at pH 44.0 depending on SSPS stabilized with HMP was larger than that of the
concentration may be caused by the weak electrostatic dispersions stabilized with SSPS (Figs. 2A and 3A).
interactions between SSPS and caseins. At concentra- Fig. 3B illustrates the effect of pH on the particle size
tions 40.4% SSPS the dispersions had a monomodal distribution of acid milk containing 0.4% HMP. While
distribution of particle sizes and the distribution was the particle size distribution of acidified milk was

750
700
Apparent Diameter (nm)

650

600

550
500

450

400

350
300
3.0 3.2 3.4 3.6 3.8 4.0 4.2 4.4
(A) pH

20
18
16
Relative Abundance (%)

14
12
10
8
6
4
2
0
0.01 0.1 1 10 100
(B) Particle size (µm)

Fig. 2. Effect of SSPS and pH on the particle size of acidified milk dispersions. (A) Apparent diameter measured by DLS of dispersions containing
(K) 0.2%; (m) 0.4%; (’) 0.6% SSPS as a function of pH. Error bars indicate standard deviation. (B) Particle size distribution measured by
integrated light scattering of dispersions stabilized with 0.4% SSPS prepared at (J) pH 4.2; (K) pH 4.0; (m) pH 3.8; (n) pH 3.6; (’) pH 3.2.
 ARTICLE IN PRESS
366 A. Nakamura et al. / International Dairy Journal 16 (2006) 361–369

1900

1700

1500
Apparent Diameter (nm)

1300

1100

900

700

500

300
3.0 3.2 3.4 3.6 3.8 4.0 4.2 4.4
(A) pH

20
18
16
Relative Abundance (%)

14
12
10
8
6
4
2
0
0.01 0.1 1 10 100
(B) Particle Size (µm)

Fig. 3. Effect of HMP and pH on the particle diameter of acidified skim dispersions. (A) Apparent diameter by DLS of dispersions containing
(K) 0.2%; (m) 0.4%; (’) 0.6% HMP as a function of pH. Error bars indicate standard deviation. (B) Particle size distribution measured by
integrated light scattering of dispersions stabilized with 0.4% HMP prepared at (J) pH 4.2; (K) pH 4.0; (m) pH 3.8; (n) pH 3.6; (’) pH 3.2.

monomodal for milks prepared at pH 4.2 and 4.0, at a tion with pectinase. Fig. 4 summarizes the effect of
pH o3.8 all acidified milk samples homogenized with pectinase on skim milk homogenized with 0.4% SSPS at
HMP showed a bimodal distribution of sizes, indicating pH 3.8. The hydrolysis of SSPS by pectinase caused a
that at these pH values the interactions between the decrease in the particle diameter of the acid dispersions
caseins and HMP were weak. containing SSPS, followed by a rapid particle aggregation.
Treatment with pectinase decreased the apparent diameter
3.3. Hydrolysis of SSPS adsorbed on the casein particles of 83710 nm, indicating that SSPS covered the surface of
casein particles with a thickness of about 40 nm. The
While the adsorption of HMP on acidified caseins has surface coverage of SSPS was not affected by pH: at all pH
been previously described in detail (Tromp et al., 2004), values tested, a thickness of 40 nm was measured (Fig. 5).
the mechanism of adsorption of SSPS has not been fully In homogenized acid milk samples in the pH range from
characterized. To confirm that SSPS was adsorbed onto 4.2 to 3.2, SSPS adsorbed on the casein particles and the
the surface of casein micelles and contributed to the elongated polysaccharide chains stabilized the caseins from
stability of the dispersion, the effect of addition of aggregation via steric repulsion.
pectinase on the diameter of the SSPS-milk dispersions
was studied. Commercially available pectinase has been 3.4. Stabilization of casein micelles by SSPS or HMP in
shown to hydrolyze SSPS in solution into oligosaccharides directly acidified dispersions
or monosaccharides (Nakamura et al., 2001, 2002).
Acidified milk containing SSPS was diluted in permeate Differences in the stabilizing behaviour between
and the apparent diameter was measured during incuba- SSPS and HMP were also evaluated in model systems
 ARTICLE IN PRESS
A. Nakamura et al. / International Dairy Journal 16 (2006) 361–369 367

1100

1000
Apparent Diameter (nm)

900

800

700

600

500

400

300
0 5 10 15 20 25 30
Time (min)

Fig. 4. Change in average particle diameter of acidified milk homogenized with 0.4% SSPS at pH 3.8 after addition of pectinase (&) 84 units;
(n) 60 units; (J) 36 units of pectinase; (K) no-pectinase control.

700 700
Particle Diameter (nm)

Particle Diameter (nm)

650 650
600 600
550 550
500 500
450 450
400 400
350 350
300 300
0 5 10 15 20 25 30 0 5 10 15 20 25 30
(A) Time (min) (B) Time (min)

700 700
Particle Diameter (nm)

Particle Diameter (nm)

650 650
600 600
550 550
500 500
450 450
400 400
350 350
300 300
0 5 10 15 20 25 30 0 5 10 15 20 25 30
(C) Time (min) (D) Time (min)

Fig. 5. Change in average particle diameter of acidified milk homogenized with 0.4% SSPS at different pH values after addition of 36 units of
pectinase (J); no pectinase added (K). (A) pH 4.2; (B) pH 4.0; (C) pH 3.6; (D) pH 3.4.

prepared by direct acidification. Skim milk was mixed while pectin showed the ability to stabilize milk during
with 0.4% SSPS or HMP and then directly acidified acidification with GDL at pH 44.6, direct acidification
with HCl with continuous stirring. Acid dispersions of milk to pH o4.2 in the presence of pectin showed
were homogenized at various pH and the particle size aggregation in all samples analyzed. These results
distribution was evaluated before and after homogeni- indicated that pectin interactions occur at higher pH
zation as summarized in Fig. 6. and at pH o4.2 electrostatic interactions are quite
Direct acidification to pH o4.0 resulted in a weak. After homogenization all samples containing
monomodal distribution of particle sizes in acidified SSPS showed a monomodal size distribution in agree-
milk containing SSPS. Only samples acidified to pH 4.2 ment with the results shown in Fig. 2. Pectin stabilized
showed the presence of a second population of large the acid dispersions and the samples showed a mono-
aggregates in the 5–10 mm range. These results indicated modal size distribution only at pH 44.0, indicating that
that homogenization may not be needed to form stable homogenization was necessary to aid in the adsorption
dispersions with SSPS, as long as a pH o4.0 is reached of pectin. These results were also in agreement with the
rapidly and with continuous stirring. On the other hand, findings reported in Fig. 3.
 ARTICLE IN PRESS
368 A. Nakamura et al. / International Dairy Journal 16 (2006) 361–369
Relative Abundance (%)

Relative Abundance (%)

20 20
18 18
16 16
14 14
12 12
10 10
8 8
6 6
4 4
2 2
0 0
0.01 0.1 1 10 100 0.01 0.1 1 10 100
(A) Diameter (µm) (B) Diameter (µm)

Relative Abundance (%)

Relative Abundance (%)

20 20
18 18
16 16
14 14
12 12
10 10
8 8
6 6
4 4
2 2
0 0
0.01 0.1 1 10 100 0.01 0.1 1 10 100
(C) Diameter (µm) (D) Diameter (µm)

Fig. 6. Particle size distribution of the directly acidified skim milk containing SSPS (A, C) or HMP (B, D). Non-homogenized (A, B) and
homogenized (C, D) dispersions at various pH: (’) pH 4.2; (K) pH 4.0; (&) pH 3.6; (J) pH 3.4; (n) pH 3.2.

4. Discussion When SSPS was added to skim milk before the

addition of GDL, no interactions were observed
The present studies demonstrated that the two between SSPS and casein micelles with decreasing pH.
polysaccharides being compared are characterized by SSPS contains much less uronic acid (approximately
two distinct stabilizing mechanisms. Tuinier et al. (2002) 18%) compared with the HMP used in this study
investigated the adsorption of HMP onto the surface of (approximately 80%). The charge interactions between
casein micelles under experimental conditions compar- SSPS and caseins are weaker than those between HMP
able to those used in this work. In their case, 1% skim and caseins, and SSPS does not interact with caseins at
milk was diluted in simulated milk ultrafiltrate and pH 44.6. The difference in behaviour at high pH
HMP was added at a concentration range between (44.6) between HMP and SSPS may also be caused by a
0.0075% and 0.0375%. The change in particle size after difference in calcium sensitivity between the two
addition of GDL was measured in situ by DLS. In this polysaccharides.
study skim milk (8.5% solids) was acidified in the Low pH values seem to be necessary for SSPS to
presence of SSPS or pectin at various concentrations, interact with casein micelles. At pH o4.2, SSPS showed
and the milk was diluted with the corresponding better stabilizing properties than HMP. Low pH (less
permeate only when a certain pH was reached. Our than 4.0) and shear may be important for SSPS to
results on HMP stabilization were in agreement with interact with caseins in acidified milk. Direct acidification
those obtained by Tuinier et al. (2002): during acidifica- to pH 4.0 or below resulted in stable dispersions, even
tion, pectin molecules with a particular charge and before homogenization, while at higher pH, homogeniza-
charge distribution adsorb onto the surface of casein tion was necessary to prevent aggregation. A higher
micelles via electrostatic interactions and prevent acid- distribution of positive charges on casein micelles may be
induced aggregation of the micelles. However, we also necessary for interactions with SSPS to occur.
found evidence of an increase in particle diameter with SSPS stabilized caseins in acid milk forming a thick
increasing concentration of pectin in non-acidified milk. layer of about 40 nm. The addition of pectinase to SSPS-
In addition, during acidification the particle diameter stabilized acid milk caused particle aggregation, and
decreased with decreasing pH as previously reported; demonstrated that hydrolysis of the carbohydrate
however, the size difference increased as a function of moieties of SSPS affected the stability of the acid milk
pectin concentration, from 25 nm with no pectin to samples. The layer thickness was not affected by
about 60 nm with 0.6% pectin. These results may changes in pH in the range between 3.2 and 4.2. The
indicate adsorption of pectin at pH values higher than calculated thickness of 40 nm is in agreement with the
those previously reported. It appears that pectin affected values of thickness calculated when SSPS is adsorbed
the casein micelles during acidification prior to the pH on the surface of oil emulsion droplets (Nakamura,
of aggregation (between 5.8 and 5.0). Takahashi et al., 2004).
 ARTICLE IN PRESS
A. Nakamura et al. / International Dairy Journal 16 (2006) 361–369 369

5. Conclusions Mishra, S., Mann, B., & Joshi, V. K. (2001). Functional improvement
of whey protein concentrate on interaction with pectin. Food
Hydrocolloids, 15, 9–15.
SSPS and HMP both can stabilize casein particles in
Nakamura, A., Furuta, H., Kato, M., Maeda, H., & Nagamatsu, N.
low pH milk. SSPS functionality seems to complement (2003). Effect of soluble soybean polysaccharides on the stability of
pectin, as at pH 44.6 pectin stabilizes acid dispersions, milk protein under acidic conditions. Food Hydrocolloids, 17,
while under acidic conditions (pH 3.2–4.0) SSPS adsorbs 333–343.
onto the surface of casein micelles and prevents Nakamura, A., Furuta, H., Maeda, H., Nagamatsu, Y., & Yoshimoto,
aggregation. The differences in functionality are caused A. (2001). Analysis of structural components and molecular
construction of soybean soluble polysaccharides by stepwise
by the different molecular structures of the two enzymatic degradation. Bioscience, Biotechnology, and Biochemis-
polysaccharides. In addition to the lower amount try, 65, 2249–2258.
of galacturonic acid, SSPS has galactan portions Nakamura, A., Furuta, H., Maeda, H., Takao, T., & Nagamatsu, N.
and highly branched arabinan extending from the (2002). Structural studies by stepwise enzymatic degradation of the
galacturonan backbone (Nakamura et al., 2001, 2002); main backbone of soybean soluble polysaccharides consisting of
galacturonan and rhamnogalacturonan. Bioscience, Biotechnology,
the carbohydrate moieties contribute to the steric and Biochemistry, 66, 1301–1313.
repulsion between the particles once SSPS molecules Nakamura, A., Takahashi, T., Yoshida, R., Maeda, H., & Corredig,
are adsorbed onto the caseins. M. (2004). Emulsifying properties of soybean soluble polysacchar-
ide. Food Hydrocolloids, 18, 795–803.
Nakamura, A., Yoshida, R., Maeda, H., Furuta, H., & Corredig, M.
(2004). A study of the role of the carbohydrate and protein moieties
of soy soluble polysaccharides in their emulsifying properties.
References Journal of Agricultural and Food Chemistry, 52, 5506–5512.
Pereyra, R., Schmidt, K. A., & Wicker, L. (1997). Interaction and
Asai, I., Watari, Y., Iida, H., Masutake, K., Ochi, T., Ohashi, S., stabilization of acidified casein dispersions with low and high
Furuta, H., & Maeda, H. (1994). Effect of soluble soybean methoxyl pectins. Journal of Agricultural and Food Chemistry, 45,
polysaccharide on dispersion stability of acidified milk protein. In 3448–3451.
K. Nishinari, & E. Doi (Eds.), Food hydrocolloids: Structure, Schmitt, C., Sanchez, C., Despond, D., Renard, D., Thomas, F., &
properties and functions (pp. 151–156). New York: Plenum Press. Hardy, J. (2000). Effect of protein aggregates on the complex
Burgess, D. J. (1990). Practical analysis of complex coacervate systems. coacervation between b-lactoglobulin and acacia gum at pH 4.2.
Journal of Colloid and Interface Science, 144, 227–238. Food Hydrocolloids, 14, 403–413.
Cai, R., & Arntfield, S. D. (1997). Thermal gelation in relation to Syrbe, A., Fernandes, P. B., Dannenberg, F., Bauer, W., & Klostermeyer,
binding of bovine serum albumin–polysaccharide systems. Journal H. (1995). Whey protein and polysaccharide mixtures: Polymer
of Food Science, 62, 1129–1134. incompatibility and its application. In E. Dickinson, & D. Lorient
De Kruif, C. G. (1998). Supra-aggregates of casein micelles as a (Eds.), Food macromolecules and colloids (pp. 328–339). Cambridge:
prelude to coagulation. Journal of Dairy Science, 81, 3019–3028. The Royal Society of Chemistry.
De Kruif, C. G., & Tuinier, R. (2001). Polysaccharide protein Syrbe, A., Bauer, W. J., & Klostermeyer, H. (1998). Polymer science
interactions. Food Hydrocolloids, 15, 555–563. concept in dairy system: An overview of milk protein and food
Galazka, V. G., Smith, D., Ledward, D. A., & Dickinson, E. (1999). hydrocolloid interaction. International Dairy Journal, 8, 179–193.
Complexes of bovine serum albumin with sulphated polysacchar- Tostoguzov, V. B. (1986). Functional properties of protein–polysac-
ides: Effects of pH, ionic strength, heat and high-pressure charide mixtures. In J. R. Mitchell, & D. A. Ledward (Eds.),
treatment. Food Chemistry, 64, 303–310. Functional properties of food macromolecules (pp. 385–415).
Glahn, P. E. (1982). Hydrocolloid stabilization of protein suspensions London: Elsevier.
at low pH. Progress in Food and Nutrition Science, 6, 171–177. Tromp, R. H., De Kruif, C. G., Van Eijk, M., & Rolin, C. (2004). On
Holt, C. (1982). Structure and stability of the bovine casein micelle. In the mechanism of stabilisation of acidified milk drink by pectin.
C. B. Afinsen, J. D. E. D. Sall, F. K. Richards, & D. S. Eisenberg Food Hydrocolloids, 18, 565–572.
(Eds.), Advances in protein chemistry (pp. 63–151). New York: Tuinier, R., Rolin, C., & De Kruif, C. G. (2002). Electrosorption of
Academic Press. pectin into casein micelles. Biomacromolecules, 3, 632–638.
Imeson, A. P., Ledward, D. A., & Mitchell, J. R. (1977). On the nature Voragen, A. G. J., Pilnik, W., Thibault, J. F., Axelon, M. A. V., &
of the interaction between some anionic polysaccharides and Renard, C. M. G. C. (1995). Pectins. In A. M. Stephan (Ed.), Food
proteins (food texture). Journal of the Science of Food and polysaccharides and their functional applications (pp. 287–339).
Agriculture, 28, 661–668. New York: Marcel Dekker.