Food Hydrocolloids 55 (2016) 65e76

Contents lists available at ScienceDirect

Food Hydrocolloids
journal homepage: www.elsevier.com/locate/foodhyd

Calcium binding and calcium-induced gelation of sodium alginate

modified by low molecular-weight polyuronate
Makoto Nakauma a, *, Takahiro Funami a, Yapeng Fang b, **, Katsuyoshi Nishinari b,
Kurt I. Draget c, Glyn O. Phillips d
a
Texture Design Laboratory, San-Ei Gen F.F.I., Inc., 1-1-11, Sanwa-cho, Toyonaka, Osaka 561-8588, Japan
b
Glyn O. Phillips Hydrocolloid Research Centre at HUT, School of Food and Pharmaceutical Engineering, Faculty of Light Industry, Hubei University of
Technology, Wuchang, Wuhan 430068, China
c
Norwegian Biopolymer Laboratory, Department of Biotechnology, Norwegian University of Science and Technology, N-7491 Trondheim, Norway
d
Phillips Hydrocolloids Research Ltd., 45 Old Bond Street, London W1S 4AQ, UK

a r t i c l e i n f o a b s t r a c t

Article history: The functions of low molecular-weight Mw polyuronate on the calcium binding and calcium-induced
Received 23 April 2015 gelation of normal sodium alginate (ALG) have been investigated. Mannuronate- and guluronate-rich
Received in revised form fractions were prepared from ALG at two different Mw for each. In the mixtures of ALG and each algi-
25 October 2015
nate fraction, changes in the relative viscosity of dilute solutions and rheological properties of the gels
Accepted 26 October 2015
were examined after calcium addition. In dilute solutions, the mannuronate-rich fractions did not sub-
Available online 18 November 2015
stantially alter the calcium binding behavior of ALG regardless of Mw. On the contrary, the guluronate-
rich fractions changed the calcium binding behavior of ALG, and more calcium was required for in-
Keywords:
Alginate
crease in the relative viscosity relating to the formation of egg-box dimers and subsequent aggregations.
Calcium binding These results were more evident when Mw of guluronate-rich fractions was lower. Gel rheology was also
Egg-box dimer different between the mannuronate- and the guluronate-rich fractions. In the gels, both fractions
Gelation decreased the storage modulus in the linear viscoelastic regime with increased yield strain, but these
Guluronate effects of the guluronate-rich fractions were greater than the mannuronate-rich fractions when
Pectin compared at equivalent Mw. These functions of the guluronate-rich fractions were quite different from
those of low-methoxyl pectin fraction. By using well-characterized polyuronate samples, calcium-
induced gelation for the mixture of ALG and each low Mw polyuronate was compared on the molecu-
lar level.
© 2015 Elsevier Ltd. All rights reserved.

1. Introduction
Alginate is a collective term for a family of exopolysaccharides
produced mainly from the brown seaweeds such as Laminaria
hyperborea and Macrocystis pyrifera. Alginate can be produced in a
variety of counterion forms, but sodium alginate is preferentially
used as a gelling agent in the food industry mainly due to its high
water-solubility. Alginate is a natural polyelectrolyte and made up
of a linear copolymer of (1e4)-linked b-D-mannuronic acid (M) and
a-L-guluronic acid (G) units (Draget, Moe, Skjåk-Bræk, & Smidsrød,
* Corresponding author.
** Corresponding author.
E-mail addresses: m-nakauma@saneigenffi.co.jp (M. Nakauma), fangypphrc@
163.com (Y. Fang).
2006). These units form M-, G-, and MG-block structures depend-
ing on the sequence (Moe, Draget, Skjåk-Bræk, & Smidsrød, 1995).
G-blocks influence alginate gelation, and in the presence of some
multivalent cations, gelation occurs instantly forming ordered
structures by cation-bridged intermolecular associations (Gant,
Morris, Rees, Smith, & Thom, 1973; Li, Fang, Vreeker, &
Appelqvist, 2007). The ordered structures schematically resemble
an egg-box, which is now generally accepted as the gelation model
of alginate (Djabourov, Nishinari, & Ross-Murphy, 2013; Morris,
Rees, Thom, & Boyd, 1978; Thom, Grant, Morris, & Rees, 1982).
Formation of egg-box dimers is greatly affected by G-block length
in the alginate molecules (Aarstad, Strand, Klepp-Andersen, &
Skjåk-Bræk, 2013). Added cations also have effects with calcium the
most common among divalent cations for increasing the gel
strength (Mørch, Donati, Strand, & Skjåk-Bræk, 2006) mainly
because of structural compatibility with G residues. Compared with

http://dx.doi.org/10.1016/j.foodhyd.2015.10.021
0268-005X/© 2015 Elsevier Ltd. All rights reserved.
66 M. Nakauma et al. / Food Hydrocolloids 55 (2016) 65e76
G-blocks, M- and MG-blocks are less sensitive to cations, and the
gel strength is less dependent on cation variations (Mørch et al.,
2006).
The effects of the macromolecular characteristics of alginate on
calcium binding have been investigated (Draget et al., 2001;
Funami et al., 2009; Stokke, Smidsrød, Bruheim, & Skjåk-Bræk,
1991). As a molecular parameter, average molecular weight of
alginate and content/sequentiality of G residues have been mainly
studied (Draget, Skjåk-Bræk, & Smidsrød, 1994; Kong, Kaigler, Kim,
& Mooney, 2004; Sikorski, Mo, Skjåk-Bræk, & Stokke, 2007).
Recently, the effects of G-block length on calcium binding of algi-
nate were studied using samples which had been enzymatically
converted from M to G in the alginate molecules (Aarstad et al.,
2013). Furthermore, multiple steps in binding of calcium to algi-
nate were demonstrated using dilute solutions (Borgogna, Skjåk-
Bræk, Paoletti, & Donati, 2013; Fang et al., 2007). Gelation and
molecular associations of alginate have been extensively of
fundamental and practical interest.
Pectin is another representative polyuronate widely used in the
food industry. Calcium binding of pectin has been investigated
using samples with enzymatic or chemical treatment for modified
degree of esterification and average molecular weight. These
characteristics influence pectin gelation, including gel strength and
the kinetics of gel formation, and have been studied in relation to
affinity and sensitivity to calcium (Hotchkiss et al., 2002; Luzio &
Cameron, 2008; Ralet, Dronnet, Buchholt, & Thibault, 2001;
Thibault & Rinaudo, 1985).
Here we seek to understand the calcium binding and calcium-
induced gelation of sodium alginate in the presence of low
molecular-weight alginate fractions. Both the M-rich and the G-rich
fractions were prepared from normal sodium alginate after acid
hydrolysis followed by fractionation by pH. By controlling the hy-
drolysis condition, two different average molecular-weights were
prepared for each fraction. The macromolecular and physico-
chemical characteristics of these fractions were identified. In the
mixtures of normal sodium alginate and each fraction, changes in
the relative viscosity of dilute solutions and rheological properties
of the gels were examined after calcium addition. Low molecular-
weight low-methoxyl pectin fraction was also prepared by two-
step enzyme treatment, and the effects were compared to the G-
rich fractions at equivalent molecular weight. The objective of the
present study is to clarify the functions of low molecular-weight
alginate fractions for calcium binding and calcium-induced gela-
tion of normal sodium alginate on a molecular level compared to
those of low molecular-weight low-methoxyl pectin fraction. Use of
well characterized samples in terms of macromolecular and phys-
icochemical properties must be an essential approach to obtain
clear-cut conclusions. However, in previous studies, macromolec-
ular and physicochemical characteristics of polyuronate samples,
including average molecular-weight, polydispersity, composition
and sequence of monomers etc, are sometimes insufficient, and
thus key factors that govern phenomena may be obscure. This point
was clarified in the present study. Usefulness of the alginate frac-
tions was also testified as a functional material for increasing the
water holding capacity of the gel system as one of the requirements
from the industry.
2. Materials and methods
2.1. Materials
Sodium alginate (ALG) SAN-SUPPORT® P-80 and high-methoxyl
pectin (HMP) from citrus SAN-SUPPORT® P-160 were provided by
San-Ei Gen F.F.I., Inc. (Osaka, Japan) as commercial products. Re-
agent grade of glucono-d-lactone (GDL) and calcium carbonate
were purchased from Kishida Chemical (Osaka, Japan) as an
acidulant and as a water-insoluble calcium source, respectively. The
median particle diameter of calcium carbonate was 18.1 mm in the
original form, which was pulverized to approx. 10 mm before use.
Pectinase (Pectinex® Yield MASH) was purchased from Novozyme
(Bagsværd, Denmark) with one unit defined as a capability to
liberate 1.0 mmol galacturonic acid per minute at pH 4.0 at 25
C.
Pectin methyl-esterase (Novo shape® XL) was also purchased from
Novozyme (Bagsværd, Denmark) with one unit defined as a capa-
bility to liberate 1.0 mmol methanol per minute at pH 7.5 at 30
C.
2.2. Preparation of low molecular-weight alginate fractions
Twenty grams (on a dry base, hereafter the same unless other-
wise specified) of ALG was dispersed in 200 ml of 0.3 M HCl and
stirred for 17 h at 25
C. The media were replaced with 50 ml of
0.3 M HCl and was heated at 95
C for 5 h for hydrolysis. After
centrifugation at 750g for 15 min, precipitate obtained was
dispersed in 50 ml of distilled de-ionized water, and this was
repeated twice. The dispersions were adjusted at pH 3.5 using
0.5 M NaOH and stirred at 25
C for 17 h to recover the G-rich
fractions. After centrifugation at 750g for 15 min, precipitate ob-
tained was dispersed in 100 ml of distilled de-ionized water, and
pH was adjusted at 7.0 using 4.0 M NaOH for solubilization. The
solutions were filtered through GF/A glass filters of 1.6 mm pore size
and freeze-dried to obtain a sample identified as LMw-GUL1. On the
other hand, supernatant obtained was adjusted at pH 2.6 using
1.0 M HCl and left standing at 25
C for at least 1 h to recover the M-
rich fractions. After centrifugation at 750g for 15 min, precipitate
obtained was dispersed in 100 ml of distilled de-ionized water, and
pH was adjusted at 7.0 using 4.0 M NaOH for solubilization. The
solutions were filtered through the glass filters and freeze-dried to
obtain a sample identified as LMw-MAN1. In the procedure above,
different heating condition for hydrolysis; for 1 h (in place of 5 h)
and different pH conditions for recovery of the G-rich fractions; 3.8
(in place of 3.5) and the M-rich fractions; 2.4 (in place of 2.6) were
used to obtain samples identified as LMw-GUL2 and LMw-MAN2,
respectively.
2.3. Preparation of low molecular-weight low-methoxyl pectin
fraction
Thirty grams of HMP was dispersed in 970 ml of distilled de-
ionized water and stirred for 30 min at 25
C. One ml of the pec-
tinase (200 unit/ml) was added to the dispersion, incubated at
40
C for 2 h for hydrolysis, and heated at 90
C for 30 min to stop
the enzymatic reaction. One ml of the pectin methyl-esterase
(45 unit/ml) was then added to the dispersions, incubated at
40
C for 15 h for de-esterification, and heated at 90
C for 30 min to
stop the enzymatic reaction. The solutions were filtered through
the glass filters and freeze-dried to obtain a sample identified as
LMw-LMP.
2.4. Macromolecular and physicochemical characteristics of ALG
and alginate fractions
Macromolecular characteristics of ALG and each alginate frac-
tion, including weight-average molecular weight Mw, number-
average molecular weight Mn, radius of gyration Rg, and poly-
dispersity index defined by Mw/Mn, were determined by size-
exclusion chromatography coupled with a multiangle laser light
scattering photometer (SEC-MALS) in basically the same manner as
reported (Funami et al., 2009). The Flory exponent n was also
determined based on the relationship between Mw and Rg:
 M. Nakauma et al. / Food Hydrocolloids 55 (2016) 65e76
Rg ¼ KMw
n
:
G content of ALG and each alginate fraction was determined by a
nuclear magnetic resonance NMR spectrometry on a 600-MHz
apparatus (type ECA 600, JOEL, Tokyo, Japan). Respective sample
was treated with D2O to exchange the labile protons with deu-
terons before NMR. On the 1H spectrum, peak areas of the signal
from guluronate anomers and mannuronate anomers were detec-
ted at 5.00e5.15 ppm and at 4.60e4.75 ppm, respectively, as a
chemical shift in reference to an internal standard, sodium
trimethylsilyl-2, 2, 3, 3,-d4-propionate. G content was determined
as a percentage of the peak area from guluronate to that from the
sum of guluronate and mannuronate. Based on the sequential in-
formation, the length of G-block larger than 1 (G-block length) was
determined (Funami et al., 2009):
G  block length ¼ ðFG  FMGM Þ=FGGM
Here FG, FMGM, and FGGM represent a fraction consisting of
guluronic acid, a fraction consisting of two mannuronic acids
interspaced with guluronic acid, and a fraction starting or ending
with a block of guluronic acid.
2.5. Macromolecular and physicochemical characteristics of HMP
and LMw-LMP
Macromolecular characteristics of HMP and LMw-LMP were
determined in the same manner as Section 2.4. Constitutional
sugars of HMP and LMw-LMP were identified by high-performance
anion-exchange chromatography coupled with pulsed ampero-
metric detection (HPAEC-PAD) in basically the same manner as
reported (Funami et al., 2007). Degree of methylation DM of HMP
and LMw-LMP was determined spectrophotometrically in combi-
nation with alcohol oxidase and 2, 4-pentanedione treatment
(Klavons & Bennett, 1986). Absorbance at 412 nm was read to
determine the methanol content, and DM was calculated as a
percentage of the molar concentration of methanol to that of
galacturonate.
2.6. Relative viscosity measurement of dilute solutions
Changes in relative viscosity of dilute solutions after calcium
addition were measured using an Ubbelohde type capillary
viscometer isothermally at 25 ± 0.2
C (Fang et al., 2008). Test so-
lutions for the viscometry were prepared by dissolving the mixture
of ALG and each alginate fraction or LMw-LMP in 20 mM acetate
buffer (pH 5.0). Concentration of ALG in the mixture solutions was
0.05%, whereas those of each alginate fraction or LMw-LMP were
0.01%, 0.02%, and 0.05%. Two hundred ml of 7.5 mM CaCl2 (in 20 mM
acetate buffer) was titrated to each mixture solution (10 ml) in a
stepwise manner (till 4000 ml at the maximum). After titration, the
Ubbelohde viscometer was shaken manually for approx. 20 s for
mixing. Control experiments were carried out using the acetate
buffer in place of CaCO3 solution. The relative viscosity hr was
determined from ts/t0, where ts is the flow time for test solutions,
and t0 is the flow time for the solvent (i.e. acetate buffer). To
eliminate the dilution effect by the addition of CaCO3 solutions
during titration, the relative viscosity was normalized:
. and 0.2% each alginate fraction or LMw-LMP, which was the same as
hN Ca
r ¼ hr hCr
Here hCar is the relative viscosity in calcium titration, and hr is
C Darmstadt, Germany) with a metal grating support, placed in a
the relative viscosity in buffer titration (Fang et al., 2008). Data were
presented as a mean ± SD of triplicate.
67
2.7. Rheological measurements of gels
Rheological properties of gels were measured using a strain-
controlled rheometer (type ARES LS-1, TA Instruments, DE, USA)
in an oscillation shear mode when equipped with a temperature-
controlled water bath. Test solutions for the rheology were pre-
pared by dissolving the mixture of ALG and each alginate fraction or
LMw-LMP in 20 mM acetate buffer (pH 5.0), into which CaCO3 and
GDL were added. Concentration of ALG in the test solutions was
fixed at 0.8%, whereas those of each alginate fraction or LMw-LMP
were varied at 0.2%, 0.4%, and 0.8%. Concentrations of CaCO3 and
GDL were fixed at 20 mM. One point five gram of each test solution
was placed between a parallel-plate geometry (25 mm in diameter
and 1.0 mm in gap), which was set at 25
C beforehand. Test so-
lutions were incubated isothermally at the same temperature for in
situ gelation, and at least 20 min was required to reach to pseudo
saturation since we confirmed preliminarily that viscoelastic
modulus did not increase after 20 min during incubation for
60 min. Dynamic frequency and strain sweep tests were applied to
the gelled systems. Dynamic frequency sweep tests were carried
out from 0.1 to 100 rad/s at 25
C, where strain was fixed at 1.0%
with confirmation of the linearity between stress and strain. The
relationship between frequency u and complex viscosity h* was
described by a power-law regression to characterize rheological
properties (Keogh & O’ Kennedy, 1998):
 
h* ðuÞ ¼ Kf unf 0 < nf < 1
Here h* is defined by {(G0 2 þ G00 2)1/2}/iu. Constant Kf represents
the dynamic consistency index, whereas exponent nf the dynamic
power-law factor. After the frequency sweep tests, using the same
systems, dynamic strain sweep tests were carried out from 0.1 to
100% at 25
C, where frequency was fixed at 6.28 rad/s. Resultant
strain (g)-dependence curve of G0 was reduced empirically to an
exponential equation composed of two reactions with different
elastic moduli to characterize rheological properties:
G0 ðgÞ ¼ Ks1 expðns1  gÞ þ Ks2 expðns2  gÞ:
Ks1 and Ks2 represent a constant for the higher modulus
component and for the lower modulus component, respectively.
The sum of Ks1 and Ks2, which corresponds to the equilibrium G0 in
the linear viscoelastic regime, was compared among treatments.
ns1 and ns2 represent an exponent for the higher modulus
component and for the lower modulus component, respectively. In
addition, the elastic stress (G0 multiplied by strain) was plotted as a
function of strain, and the yield strain was estimated from the peak
(Walls, Caines, Sanchez, & Khan, 2003). Data were presented as a
mean ± SD of triplicate for each rheological parameter.
2.8. Water holding capacity of gels
The mixture of ALG (0.4 g) and each alginate fraction or LMw-
LMP (0.1 g for each) was dispersed in 20 mM acetate buffer (pH 5.0,
40.0 ml), into which 200 mM CaCO3 (5.0 ml) and 200 mM GDL
(5.0 ml) were added. The preparation was poured into a glass
container (20.0 mm in diameter and 10.0 mm in height), which was
enclosed with plastic boards on both sides, and cured at 20
C for at
least 2 h with still standing for gelation. The gel contained 0.8% ALG
one of the treatments in Table 3. After weighing the gel (A g), the gel
was put on a PTFE membrane filter (Merck Millipore corporation,
50 mL polypropylene centrifugation tube (cylinder of 25 mm in
diameter and 100 mm in height with cone to the top, Asahi Glass
68 M. Nakauma et al. / Food Hydrocolloids 55 (2016) 65e76

Table 1
Macromolecular characteristics and componential/sequential information of alginate samples.

Mwa (kg/mol) Rga (nm) Polydispersity indexb na Gcontc (%) G-block lengthc

ALG 111.0 24.5 1.44 0.585 74.1 13.9

LMw-MAN1 5.4 8.3 1.08 0.878 13.0 2.1
LMw-MAN2 52.0 19.5 1.20 0.654 29.5 3.9
LMw-GUL1 12.1 14.3 1.12 0.662 92.0 17.6
LMw-GUL2 44.5 18.7 1.29 0.628 85.1 15.5

See the text for experimental detail.

a
Weight-average molecular weight Mw, radius of gyration Rg, and polydispersity index were determined by SEC-MALS. The Flory exponent n was determined from the
relationship Rg ¼ Mwn .
b
Polydispersity index was calculated as the ration of number-average molecular weight to weight-average molecular weight.
c
Guluronate content Gcont and G-block length were determined by 1H NMR.

Table 2
Macromolecular characteristics and componential information of pectin samples.

Mwa (kg/mol) Rga (nm) Polydipersity indexb na Constituent sugarsc (%, w/w) DMd (%)

Rha Ara Gal Glu GalA

HMP 205.2 27.2 1.50 0.552 2.2 6.5 9.8 2.4 72.5 72.3
LMw-LMP 66.2 18.0 1.29 0.765 3.5 2.3 4.7 1.3 81.3 5.2

Rha: rhamnose, Ara: arabinose, Gal: galactose, Glc: glucose, GalA: galacturonic acid.
See the text for experimental detail.
a
Weight-average molecular weight Mw, radius of gyration Rg, and polydispersity index were determined by SEC-MALS. The Flory exponent n was determined from the
relationship Rg ¼ Mwv .
b
Polydispersity index was calculated as the ration of number-average molecular weight to weight-average molecular weight.
c
Constituent sugars were determined by HPAEC-PAD.
d
Degree of methylation DM was determined by spectrophotometry in combination with alcohol oxidase and 2, 4-pentanedione treatment.

Table 3
Rheological parameters for the mixture of normal sodium alginate (ALG) and each alginate fraction.

Total G (mM) Rtotal G Yield straina (%) Ks1 þ Ks2b (Pa) nfc

0.8% ALG (control) 33.7 0.59 12.6 ± 1.5 422.3 ± 1.1 0.963 ± 0.003
þ0.2% MAN1 34.2 0.58 12.6 ± 1.9 390.9 ± 16.8 0.952 ± 0.001
þ0.4% MAN1 35.7 0.56 15.9 ± 5.3 198.4 ± 7.5 0.919 ± 0.009
þ0.8% MAN1 38.6 0.52 25.1 ± 2.9 177.6 ± 36.0 0.887 ± 0.006
þ0.2% MAN2 36.1 0.55 12.6 ± 0.0 373.0 ± 16.6 0.961 ± 0.002
þ0.4% MAN2 39.4 0.51 15.9 ± 2.4 286.2 ± 16.0 0.938 ± 0.008
þ0.8% MAN2 46.1 0.43 15.9 ± 5.3 205.9 ± 16.4 0.915 ± 0.001
þ0.2% GUL1 43.2 0.46 25.1 ± 3.8 174.2 ± 20.9 0.925 ± 0.012
þ0.4% GUL1 53.6 0.37 63.2 ± 18.2 57.2 ± 8.0 0.887 ± 0.018
þ0.8% GUL1 74.5 0.27 ND 31.4 ± 10.2 0.824 ± 0.035
þ0.2% GUL2 42.4 0.47 15.9 ± 2.4 284.5 ± 13.4 0.950 ± 0.006
þ0.4% GUL2 52.1 0.38 20.0 ± 4.3 98.5 ± 10.8 0.927 ± 0.005
þ0.8% GUL2 71.4 0.28 25.1 ± 7.2 64.6 ± 8.7 0.916 ± 0.015

ND: not determined.

See the text for experimental detail. Data are presented as a mean ± SD of triplicate.
a
Obtained from strain-elastic stress (G0 multiplied by strain) response curve in the strain-sweep test of dynamic storage modulus.
b
Calculated using the equation of G0 (g) ¼ Ks1 ens1$g þ Ks2 ens2$g in the strain-sweep test of dynamic storage modulus.
c
Calculated using the equation of h* ðuÞ ¼ Kf unf in the frequency-sweep test of dynamic storage modulus.

Co. Ltd, Shizuoka, Japan), and centrifuged at 210g for from 30 to
120 min in 30 min increments. After centrifugation, syneresis
generated was separated from the gel to the bottom of the tube
through the filter, and the gel was reweighted (B g). Water holding
capacity was determined by weight percentage of B/A. That is, 100%
means no syneresis and higher water holding capacity. Data were
presented as a mean ± SD of triplicate.
3. Results and discussion
3.1. Macromolecular and physicochemical characteristics of
polyuronate samples
Data for ALG and each alginate fraction were shown in Table 1.
Decreases in Mw and Rg with increased hydrolysis time were
confirmed for each alginate fraction. Both Mw and Rg for LMw-GUL1
were larger than those of LMw-MAN1. This is explicable since G-
blocks predominate over M-blocks in ALG. Decrease in the poly-
dispersity index with increased hydrolysis time was also confirmed
for each alginate fraction. Polydispersity index for each alginate
fraction was in the range of 1.08e1.29, which was smaller than the
Flory distribution; 2.0 (Picout, Ross-Murphy, Errington, & Harding,
2001). The Flory exponent n for ALG is between 0.5 (the Gaussian
chain value) and 0.6 (excluded volume chain value) (Picout et al.,
2001). This indicates that ALG should behave as flexible to semi-
flexible linear polymer chains in the aqueous media used. Higher
exponents for each alginate fraction than that for ALG indicates that
the conformation should be more rod-like, and this is the most
evident for LMw-MAN1 due to its equatorial orientation. Decreases
in the G content and the G-block length were confirmed for the M-
rich fractions, whereas increases in the G content and the G-block
length were confirmed for the G-rich fractions compared to the
 M. Nakauma et al. / Food Hydrocolloids 55 (2016) 65e76
corresponding data for ALG.
Data for HMP and LMw-LMP were shown in Table 2. Decreases in
Mw and Rg after the enzymatic (pectinase) treatment were
confirmed. Polydispersity index for LMw-LMP was comparable to
those for the G-rich fractions, particularly LMw-GUL2. Regarding
the Flory exponent, similar discussion to alginate is possible, and
results indicate that HMP should behave as flexible to semi-flexible
linear polymer chains in the aqueous media used, whereas the
conformation of LMw-LMP should be more rod-like compared to
that of HMP. Decrease in DM after the enzymatic (pectin-methyl-
esterase) treatment was also confirmed. No difference in Mw before
and after the methyl-esterase treatment (68.0 kg/mol and 66.2 kg/
mol, respectively) ensures that this enzyme hardly affected the
molecular weight of pectin.
3.2. Relative viscosity measurement of dilute solutions
3.2.1. Mixture of ALG and each alginate fraction
In the absence of any alginate fraction, hN r of ALG decreased
slightly until 0.42 mM, then increased rapidly, and peaked at
approx. 1.0 mM when hN r was plotted as a function of calcium
content in the system (closed circles in Figs. 1 and 2 upper). In the
present study, the boundaries at 0.42 mM and 1.0 mM are iden-
tified as the initial and the second critical thresholds, respectively.
Below the initial critical threshold, ALG can be nucleated by the
calcium binding to a single guluronate unit, which reduces elec-
trostatic repulsion within one molecule and promotes intra-
molecular contraction. This has been considered as the
monocomplex formation of ALG (Fang et al., 2007) and should be a
cause of decreased hN r . On the other hand, the rapid increase in hr
N the initial critical threshold, and each monocomplex should not
above the initial critical threshold should be attributed to the
intermolecular associations of ALG. When hN r was plotted as a
function of RALG; the molar ratio of fed calcium to the G residues
from ALG, 0.42 mM corresponded to RALG 0.25 (closed circles in
Figs. 1 and 2 middle). This is consistent with previous study (Fang
et al. 2007), proposing the scheme of the multiple-step calcium
binding to alginate, and the initial critical threshold is assigned to
the initiation of egg-box dimer formation. Calcium content that
gave hN r peak corresponded to RALG 0.64 (closed circles in Figs. 1
and 2 middle), and this second critical threshold is assigned to
the initiation of lateral associations of egg-box dimers (Fang et al.,
2007). RALG 0.64 is larger than reported value by Fang et al. (2007);
0.55. The second critical threshold depends on Mw, G content, and
G-block length of alginate samples used, and the deviation from
Fang's study can be explained by the difference in the multimer
structure; more extended in the present study mainly due to
larger Mw.
The mixture of ALG and each M-rich fraction showed compa-
rable initial critical threshold to ALG alone but sloped more gently
above the initial critical threshold when hN r was plotted as a
function of either calcium content (Fig. 1 upper) or RALG (Fig. 1
middle). For the mixture of ALG and LMw-MAN1, when the
molar ratio of fed calcium to the G residues was determined by
incorporating the effects of the G residues from the alginate
fractions (Rtotal G), the viscosity profile was almost identical to ALG
alone (Fig. 1e) except for 0.05% LMw-MAN1 (open circles) with
decreased hN r in higher regime. For the mixture of ALG and LMw-
MAN2, the second critical threshold was shifted to higher with the
peak increase when hN r was plotted as a function of either calcium
content (Fig. 1b) or RALG (Fig. 1d). These changes were more
evident with increased concentration. In contrast, only the peak
increase was evident when hN r was plotted as a function of Rtotal G.
This can be simply due to increased concentration of poly-
saccharide in the system. In total, the M-rich fractions, regardless
of Mw, should not substantially alter the calcium binding behavior
69
of ALG although they can slow down the calcium ion accessibility
due to a reversible and non-cooperative binding. Also, egg-box
dimers and subsequent multimers which the M-rich fractions
form could not be structurally incompatible with those which ALG
forms.
In the mixture of ALG and each G-rich fraction, the initial
critical threshold was shifted to higher when hN r was plotted as a
function of either calcium content (Fig. 2 upper) or RALG (Fig. 2
middle). This change was more evident with increased concen-
tration, and the initial threshold shifted from 0.42 mM to 0.70 mM
when hN r was plotted as a function of calcium content and from
0.24 to 0.40 when plotted as a function of RALG for each G-rich
fraction. This indicates decreased amount of calcium which ALG is
accessible for egg-box dimer formation. For the mixture of ALG
and LMw-GUL2, decrease in hN r before the initial critical threshold
was more evident with increased concentration of LMw-GUL2
(Fig. 2a and b). This can be due to structural incompatibility be-
tween LMw-GUL monocomplex and ALG monocomplex. When
Rtotal G was used instead of RALG, the initial critical threshold was
almost identical among treatments for each G-rich fraction (Fig. 2e
and f). The slope above the initial critical threshold was gentler
with increased concentration of each G-rich fraction, and LMw-
GUL1 was the more effective. Accordingly, the second critical
threshold shifted to higher or disappeared with increased con-
centration of LMw-GUL1 (Fig. 2e). In the case of LMw-GUL2, the
second critical threshold was also shifted to higher with peak
increase with increased concentration (Fig. 2f). For the mixture of
ALG and each G-rich fraction, it is suggested that the formation of
intramolecular monocomplex should occur independently below
contribute to the viscosity increase due to the reduced molecular
size. Above the initial critical threshold, the formation of inter-
molecular egg-box dimers should occur also independently. The
dimer formation for ALG can result in dramatic growth of the
molecular size and thus contribute to the increase in hN r , whereas
that for the G-rich fractions cannot and thus hardly contribute to
the increase in hN r , particularly for LMw-GUL1. This can be a cause
of repressed viscosity enhancement at increased concentration of
the G-rich fractions as a result of competitive calcium binding,
inhibiting the formations of egg-box dimers and their aggregates
of ALG. As in the case of the M-rich fractions, an increase in
polysaccharide concentration can be a cause of increased viscosity
at the second critical threshold, particularly for LMw-GUL2. In
addition, intermolecular associations between ALG and LMw-GUL2
through cooperative binding via calcium is possible.
3.2.2. Mixture of ALG and LMw-LMP
The mixture of ALG and LMw-LMP showed a similar hN r profile
to ALG alone when hN r was plotted as a function of calcium content
(Fig. 3a) although the viscosity above the second critical threshold
tended to increase with increased concentration of LMw-LMP. On
the other hand, when hN r was plotted as a function of RALG, the
initial critical threshold shifted to lower by the addition of LMw-
LMP, and this effect was independent of concentration (Fig. 3b).
This trend was the most evident when hN r was plotted as a func-
tion of RGulþfGal (Fig. 3c). Here RGulþfGal represents the molar ratio
of fed calcium to sum of guluronate from ALG and free gal-
acturonate from LMw-LMP; both of which are capable of binding to
calcium. In this plot, the initial critical threshold was shifted to
lower from RGulþfGal 0.24 to 0.08 with increased concentration of
LMw-LMP. This indicates that LMw-LMP should form egg-box di-
mers at a lower calcium feed than ALG. Calcium-binding behavior
of pectin is less critical than that of alginate due to sequential ir-
regularity of the calcium binding site, starting even when the
stoichiometry of egg-box dimers is not achieved (Fang et al.,
70 M. Nakauma et al. / Food Hydrocolloids 55 (2016) 65e76

Fig. 1. Changes in normalized relative viscosity hN Ca C

r (hr =hr ) during titration of 7.5 mM CaCl2 for the mixture of 0.05% normal sodium alginate (ALG) and mannuronate-rich alginate
fraction at 0% (closed circle), 0.01% (open triangle), 0.02% (closed square), and 0.05% (open circle) for LMw-MAN1 (a, c, e) and LMw-MAN2 (b, d, f). Data are plotted as a function of
calcium concentration (a & b), the molar ratio RALG (Ca/G residue from ALG) (c & d), and the molar ratio Rtotal G (Ca/total G residue from ALG and alginate fractions) (e & f). See the
text for experimental detail. Data are presented as a mean ± SD of triplicate.

2008). The size of egg-box dimers from LMw-LMP would be larger
than those from ALG if the distribution of free carboxyl groups on
the galacturonate backbone in LMw-LMP is less blockwise than the
G-blocks in ALG. This can explain the rapid increase in hN r above
the initial critical threshold. The second critical threshold was
shifted to lower from RGulþfGal 0.55 to 0.27 with increased con-
centration of LMw-LMP although the decreasing degree of hN r
above the second critical threshold was comparable among
treatments (Fig. 3c). This indicates that LMw-LMP should promote
the lateral associations of egg-box dimers of ALG without chang-
ing the nature of those inter-clusters, which may relate to previous
observation that LMP itself does not undergo significant lateral
associations (Fang et al., 2008). Or if LMw-LMP does not compete
with ALG for calcium, observed decrease in the second critical
threshold can be just a consequence of an increased denominator
without any mechanistic significance, and this hypothesis may fit
better with rheological results mentioned below.
3.3. Rheological measurements of gels
3.3.1. Mixtures of ALG and each alginate fraction
Data were shown in Table 3. For the mixture of ALG and LMw-
 M. Nakauma et al. / Food Hydrocolloids 55 (2016) 65e76 71

Fig. 2. Changes in normalized relative viscosity hN Ca C

r (hr =hr ) during titration of 7.5 mM CaCl2 for the mixture of 0.05% normal sodium alginate (ALG) and guluronate-rich alginate
fraction at 0% (closed circle), 0.01% (open triangle), 0.02% (closed square), and 0.05% (open circle) for LMw-GUL1 (a, c, e) and LMw-GUL2 (b, d, f). Data are plotted as a function of
calcium concentration (a & b), the molar ratio RALG (Ca/G residue from ALG) (c & d), and the molar ratio Rtotal G (Ca/total G residue from ALG and alginate fraction) (e & f). See the text
for experimental detail. Data are presented as a mean ± SD of triplicate.

MAN1, the yield strain increased, the sum of Ks1 and Ks2 decreased,
and the power-law exponent nf decreased with increased concen-
tration of LMw-MAN1. The system is completely elastic when the
exponent nf is equal to 1, while the system is completely viscous
when the exponent nf is 0. Similar results were found in the
mixture of ALG and LMw-MAN2 although the effects of LMw-MAN2
were smaller than those of LMw-MAN1 when compared at the same
concentration. In the frequency-dependence of G0 , decrease in G0
was detected in the low frequency regime for the mixture with 0.8%
LMw-MAN1 (Fig. 4a) due to stress relaxation. These results indicate
that the addition of the M-rich fractions, particularly LMw-MAN1,
should inhibit the gelation of ALG in terms of both the strength and
the number of junction zones. For the mixture of ALG and LMw-
GUL1, the yield strain increased, the sum of Ks1 and Ks2 decreased,
and the nf decreased with increased concentration of LMw-GUL1
(Table 3). The mixture with 0.8% LMw-GUL1 did not form gels. In the
frequency-dependence of G0 , marked decrease in G0 was detected in
the low frequency regime for the mixture with 0.8% LMw-GUL1
(Fig. 5a) due to stress relaxation. Similar results were found in the
mixture of ALG and LMw-GUL2 although the effects of LMw-GUL2
were smaller than those of LMw-GUL1 when compared at the same
concentration. Also, the effects of the G-rich fractions were much
72 M. Nakauma et al. / Food Hydrocolloids 55 (2016) 65e76

larger than those of the M-rich fractions when compared at

equivalent Mw. These results indicate that the addition of the G-rich
fractions, particularly LMw-GUL1, should inhibit the gelation of ALG
in terms of both the strength and the number of junction zones
with greater effects than those of the M-rich fractions when
compared at equivalent Mw.
Concentration of calcium fed to the system is 20 mM in theory,
which corresponds to Rtotal G 0.59 for 0.8% ALG alone with G content
33.7 mM. Therefore, the calcium feeding should be sufficient for
ALG to form egg-box dimers and gels. The addition of each alginate
fraction increases total G content in the system and decreases Rtotal G
when the calcium concentration is constant. For example, the
addition of 0.8% LMw-GUL1 increases the total G content to 74.5 mM
and decreases Rtotal G to 0.27. The G-rich fractions can form egg-box
dimers more easily than ALG, particularly LMw-GUL1, due to higher
mobility and more homogeneous molecular structures. However,
the G-rich fractions cannot form continuous gel network. Compet-
itive calcium binding can occur between ALG and each G-rich
fraction, inhibiting the formation of junction zones of ALG and
decreasing the elasticity of the system. Egg-box dimers of smaller
molecular size and discontinuous network structure which the G-
rich fractions form can be entrapped within egg-box dimers of
larger molecular size and continuous network structure which ALG
forms. This should be a cause of increased flexibility of the ALG
system. Again, in this experiment, since calcium concentration is
fixed, the decrease in Rtotal G is greater when the G content in the
alginate fractions is higher. Therefore, the alginate fraction with the
highest G content, namely LMw-GUL1, shows the greatest effect,
whereas the alginate fraction with the lowest G content, namely
LMw-MAL1, shows the least effect on the gelation of the system.

3.3.2. Mixture of ALG and LMw-LMP

The yield strain for the mixture was comparable to that for ALG
alone (control) regardless of the concentration of LMw-LMP,
whereas the sum of Ks1 and Ks2 for the mixture increased with
increased concentration of LMw-LMP (Table 4). The frequency-
dependence of G0 for the mixture showed a similar pattern to
ALG alone regardless of the concentration of LMw-LMP although G0
itself increased with increased concentration of LMw-LMP (Fig. 6).
Also, the power-law exponent nf for the mixture was almost com-
parable to that for ALG alone (Table 4). These results indicate that
the addition of LMw-LMP should not inhibit the gelation of ALG and
rather strengthens the gel structure without changing the number
of junction zones. The effect of LMw-LMP to strengthen the gel
structure is almost comparable to that of ALG because for example,
addition of 0.4% LMw-LMP to 0.8% ALG gives rise to equivalent
elasticity to 1.2% ALG (data not shown). This indicates the impor-
tance of enthalpic contribution to elasticity enhancement of the
ALG system rather than entropic contribution. However, it is un-
clear whether or not and how LMw-LMP should be involved in the
intermolecular associations of ALG. LMw-LMP binds to calcium for
egg-box dimer formation similar to the alginate fractions, and in
the mixture of ALG and LMw-LMP, competitive calcium binding
occurs. However, the elasticity of the system increases even when
RGulþfGal decrease by increased mixing ratio of LMw-LMP. Inherent
calcium content in LMw-LMP was 0.43 mg/g, and the effect as a
calcium provider was almost negligible relative to the effect of
CaCO3. Egg-box dimers of LMw-LMP could be sandwiched between
egg-box dimers of ALG and could work as a linker during lateral

and 0.05% (open circle). Data are plotted as a function of calcium concentration (a), the
r (hr =hr ) during titration of 7.5 mM
Fig. 3. Changes in normalized relative viscosity hN Ca C

CaCl2 for the mixture of 0.05% normal sodium alginate (ALG) and low-methoxyl pectin molar ratio RALG (Ca/G residue from ALG) (b), and the molar ratio RGulþfGal (Ca/the sum
fraction (LMw-LMP) at 0% (closed circle), 0.01% (open triangle), 0.02% (closed square), of G residue from ALG and free galacturonate from LMw-LMP) (c). See the text for
experimental detail. Data are presented as a mean ± SD of triplicate.
 M. Nakauma et al. / Food Hydrocolloids 55 (2016) 65e76 73

a a
10000 10000

1000 1000

G' (Pa)
G' (Pa)

100 100

10 10

1 1
0.1 1 10 100 0.1 1 10 100
Freqency (rad/s) Freqency (rad/s)

b b

10000 10000

1000 1000
G' (Pa)
G' (Pa)

100 100

10 10

1 1
0.1 1 10 100
Freqency (rad/s)
Fig. 4. Frequency-dependence of dynamic storage modulus for the mixture of 0.8%
normal sodium alginate (ALG) and mannuronate-rich alginate fraction at 0% (closed
circle), 0.2% (open triangle), 0.4% (closed square), and 0.8% (open circle) for LMw-MAN1
(a) and LMw-MAN2 (b). Concentrations of both CaCO3 and glucono-d-lactone were
fixed at 20 mM. See the text for experimental detail. Measurements were carried out in
triplicate, and one representative data are shown.
0.1 1 10 100
Freqency (rad/s)
Fig. 5. Frequency-dependence of dynamic storage modulus for the mixture of 0.8%
normal sodium alginate (ALG) and guluronate-rich alginate fraction at 0% (closed
circle), 0.2% (open triangle), 0.4% (closed square), and 0.8% (open circle) for LMw-GUL1
(a) and LMw-GUL2 (b). Concentrations of both CaCO3 and glucono-d-lactone were fixed
at 20 mM. See the text for experimental detail. Measurements were carried out in
triplicate, and one representative data are shown.

aggregation. This may be a cause of increased elasticity of the
system.
3.4. Water holding capacity of gels
Water holding capacity for 0.8% ALG (control) decreased with
increased centrifugation time, and centrifugation for 120 min
resulted in 29.6% (Fig. 7). The addition of the M-rich fractions did
not change markedly the water holding capacity of the system
compared to the control, and centrifugation for 120 min resulted in
32.3% for LMw-MAN1 and 34.2% for LMw-MAN2. On the other hand,
the addition of the G-rich fractions increased the water holding
capacity of the system compared to the control, and centrifugation
for 120 min resulted in 47.9% for LMw-GUL1 and 43.3% for LMw-
GUL2. The addition of LMw-LMP did not change markedly the water
holding capacity of the system compared to the control when the
centrifugation time is longer (i.e. 120 min) but rather decreased it
when centrifugation time is shorter (i.e. 30 min). Either each M-rich
fraction or LMw-LMP was not effective in increasing the water
holding capacity, while the G-rich fractions were effective with
greater effect of LMw-GUL1 than that of LMw-GUL2.
Increased water holding capacity by the addition of the G-rich
fractions is in accordance with changes in the rheological behavior;
‘more flexible’ gel presented by lower G0 and higher yield strain
compared to the ALG single gel. In contrast, decreased water
holding capacity by the addition of LMw-LMP is associated with
’more brittle’ gel presented by higher G0 and lower yield strain
compared to the ALG single gel. For calcium-induced gelation of
gellan gum, it has been reported that flexible gels presented by
higher fracture strain, which form at relatively low calcium con-
centration, have less-packed structure with large pore size in the
network and show increased water holding capacity when it is
subjected to external forces such as centrifugal force (Mao, Tang, &
Swanson, 2001). This is also the case of the ALG system. Results are
supportive to our discussion mentioned in 3.3 that egg-box dimers
of the G-rich fractions provide the network with enlarged spaces
and make the gel more flexible by being entrapped within
continuous network structure of ALG. On the other hand, LMw-LMP
should diminish the network size due to some reasons explained in
Section 3.5.
74 M. Nakauma et al. / Food Hydrocolloids 55 (2016) 65e76

Table 4
Rheological parameters for the mixture of normal sodium alginate (ALG) and low-methoxyl pectin fraction.

Gul þ fGal (mM) RGulþfGal Yield straina(%) Ks1 þ Ks2b nfc

0.8% ALG (control) 33.7 0.59 12.6 ± 1.5 422.3 ± 1.1 0.963 ± 0.003
þ0.2% LMw-LMP 44.5 0.45 10.0 ± 3.8 587.5 ± 25.7 0.972 ± 0.009
þ0.4% LMw-LMP 55.2 0.36 12.6 ± 5.2 902.6 ± 107.3 0.947 ± 0.003
þ0.8% LMw-LMP 76.8 0.26 10.0 ± 2.8 1171.5 ± 53.8 0.962 ± 0.011

See the text for experimental detail. Data are presented as a mean ± SD of triplicate.
a
Obtained from strain-elastic stress (G0 multiplied by strain) response curve in the strain-sweep test of dynamic storage modulus.
b
Calculated using the equation of G0 (g) ¼ Ks1 ens1$g þ Ks2 ens2$g in the strain-sweep test of dynamic storage modulus.
c
Calculated using the equation of h*(u) ¼ Kf unf in the frequency-sweep test of dynamic storage modulus.

methoxyl pectin) on the ALG gelation were compared and pre-

10000
sented schematically (Fig. 8). It is anticipated from rheological data
that the G-rich fractions can change the textural properties of
calcium-induced ALG gels from firm brittle to soft deformable.
1000 These can be explained on the molecular level. In the mixture of
ALG and each G-rich fraction, formations of egg-box dimers should
occur independently, and in this process, alginate molecules of
G' (Pa)

100
10
1 mers by ALG, which contributes greatly to the rheological proper-
0.1 1 10 100
Freqency (rad/s)
Fig. 6. Frequency-dependence of dynamic storage modulus for the mixture of 0.8%
normal sodium alginate (ALG) and low-methoxyl pectin fraction (LMw-LMP) at 0%
(closed circle), 0.2% (open triangle), 0.4% (closed square), and 0.8% (open circle).
Concentrations of both CaCO3 and glucono-d-lactone were fixed at 20 mM. See the text
for experimental detail. Measurements were carried out in triplicate, and one repre-
sentative data are shown.
3.5. Molecular mechanism of the functions of polyuronate fractions
and potential use as a quality improving agent for foods
Functions of polyuronate fractions (from either alginate or low-
Fig. 7. Water holding capacity for the mixture of normal sodium alginate (ALG) and
each polyuronate fraction. 0.8% ALG (closed circle), 0.8% ALG þ 0.2% LMw-MAN1 (open
triangle), 0.8% ALG þ 0.2% LMw-MAN2 (open square), 0.8% ALG þ 0.2% LMw-GUL1
(closed triangle), 0.8% ALG þ0.2% LMw-GUL2 (closed square), and 0.8% ALG þ 0.2%
LMw-LMP (crosses). Concentrations of both CaCO3 and glucono-d-lactone were fixed at
20 mM. See the text for experimental detail. Data are presented as mean ± SD of
triplicate.
similar size would prefer to couple with each other. For the mixture
of ALG and the G-rich fractions, the initial critical threshold is not
RALG 0.25 but Rtotal G 0.25, and this supports independent dimer
formations. Egg-box dimer formation for the G-rich fractions
should occur more preferentially than that for ALG due to higher
mobility and lower molecular flexibility of the G-rich fractions
compared to ALG. Under this situation, formation of egg-box di-
ties of the system, can follow the formation of egg-box dimers by
the G-rich fractions. As a result, egg-box dimers of the G-rich
fractions are dispersed within continuous network of ALG. This can
provide the ALG network with larger spaces and can loosen the
ordered structure of ALG. Intermolecular associations between ALG
and the alginate fractions through cooperative binding of calcium
ions in between long G-blocks is unlikely to occur for LMw-GUL1
(Liao et al., 2015) but may be possible for LMw-GUL2 because LMw-
GUL2 can be more structurally compatible to ALG than LMw-GUL1
due to the similarity of the molecular conformation. Cooperative
binding of calcium ions between ALG and LMw-GUL2 within long G-
blocks can result in decreased number of crosslinks per unit volume
and may explain the lower G0 and higher yield strain with increased
addition level of LMw-GUL2. Under the present experimental con-
ditions, when the total amount of G is increased, there will be a
reduced amount of calcium ions available relative to the number of
G-blocks. Cooperative binding between G-blocks more than DP 20
has been indicated (Kohn, 1975; Kohn & Larsen, 1972), and it can be
anticipated that the available calcium ions should bind first irre-
versibly between long G-blocks leading to reduced number of
effective junction zones per unit volume (no calcium ions left for
the shorter G-blocks) and reduced G'. In the case of the M-rich
fractions, formation of egg-box dimers is limited, and the effect to
loosen the ordered structure of ALG is smaller than that of the G-
rich fractions.
It is anticipated from rheological data that LMw-LMP can change
the textural properties of calcium-induced ALG gels to stiffer. This
effect of LMw-LMP is in contrast to that of the G-rich fractions and
can be explained from electrostatic point of view. In calcium-
induced pectin gels, it has been reported that 14 to 20 consecu-
tive free carboxyl groups are necessary for formation of egg-box
dimers (Axelos & Thibault, 1991; da Silva & Rao, 2006; Rees,
1982). For alginate, 6e8 G residues are required for formation of
egg-box dimers (Stokke et al., 1991), and LMw-LMP can restrict
topologically active binding site to calcium in the G-blocks and can
increase the accessibility of ALG to calcium. As a result, LMw-LMP
can help ALG form egg-box dimers and subsequent lateral
aggregations.
 M. Nakauma et al. / Food Hydrocolloids 55 (2016) 65e76
ALG Initial critical
thresholds
M-block
G-block
Monocomplex
ALG + LMw-GUL
ALG + LMw-MAN
ALG + LMw-LMP
Methyl
esterified Non esterified
region region
Fig. 8. Schematic presentation of calcium-induced gelation for the mixture of normal sodium alginate (ALG) and each polyuronate fraction.
Regarding food application, it was found in the industry that use
of low molecular-weight alginate, preferably in sodium form and
prepared using an enzyme lyase, is effective in noodles as a texture
modifier (JP2006-56A). Furthermore, physiological effects of low-
molecular weight alginate have been reported, including preven-
tion of rapid increases in blood pressure and blood sugar after
ingestion by incorporating high G content fractions in foods
(US2010/0256090). These functions of polyuronate fractions used
in this study should be investigated separately in the future.
4. Conclusions
The alginate fractions modify the calcium binding and calcium-
induced gelation behaviors of normal sodium alginate. Both G
content and Mw determine the functions of the alginate fractions.
As another calcium-binding polyuronate, low-methoxyl pectin
fraction shows different functions from the alginate fractions when
compared at equivalent Mw. Differences in the functions are
explained on molecular level. Observation in this study can point to
better applications of these polyuronate fractions as a quality
improving agent for foods.
References
Aarstad, O., Strand, B. L., Klepp-Andersen, L. M., & Skjåk-Bræk, G. (2013). Analysis of
G-block distributions and their impact on gel properties of in vitro epimerized
mannuronan. Biomacromolecules, 14, 3409e3416.
Axelos, M. A. V., & Thibault, J. F. (1991). Chapter 6: the chemistry of low methoxyl
pectin gelation. In R. H. Walter (Ed.), The chemistry and technology of pectin (pp.
109e118). New York: Academic Press, Inc.
75
Second critical
thresholds
Egg-box dimers Multimers
Calcium
ALG
LMw-GUL
LMw-MAN
LMw-LMP
Borgogna, M., Skjåk-Bræk, G., Paoletti, S., & Donati, I. (2013). On the initial binding of
alginate by calcium ions. The tilted egg-box hypothesis. Journal of Physical
Chemistry B, 117, 7277e7282.
Djabourov, M., Nishinari, K., & Ross-Murphy, B. S. (2013). In M. Djabourov,
K. Nishinari, & B. S. Ross-Murphy (Eds.), Physical gels from biological and syn-
thetic polymers (pp. 144e148). New York: Cambridge University Press.
Draget, K. I., Gåserød, O., Aune, I., Andersen, P. O., Stobakken, B., Stokke, B. T., et al.
(2001). Effect of molecular weight and elastic segment flexibility on syneresis in
Ca-alginate gels. Food Hydrocolloids, 15, 485e490.
Draget, K. I., Moe, S. T., Skjåk-Bræk, G., & Smidsrød, O. (2006). Alginates. In
A. M. Stephen, G. O. Phillips, & P. A. Williams (Eds.), Food polysaccharides and
their applications (2nd ed., pp. 289e334). London: Taylor & Francis.
Draget, K. I., Skjåk-Bræk, G., & Smidsrød, O. (1994). Alginic acid gels: the effect of
alginate chemical composition and molecular weight. Carbohydrate Polymers,
25, 31e38.
Fang, Y., Al-Assaf, S., Phillips, G. O., Nishinari, K., Funami, T., & Williams, P. A. (2008).
Binding behavior of calcium to polyuronates: comparison of pectin with algi-
nate. Carbohydrate Polymers, 72, 334e341.
Fang, Y., Al-Assaf, S., Phillips, G. O., Nishinari, K., Funami, T., Williams, P. A., et al.
(2007). Multiple steps and critical behaviors of the binding of calcium to algi-
nate. Journal of Physical Chemistry B, 111, 2456e2462.
Funami, T., Fang, Y., Noda, S., Ishihara, S., Nakauma, M., Draget, K. I., et al. (2009).
Rheological properties of sodium alginate in an aqueous system during gelation
in relation to supermolecular structures and Ca2þ binding. Food Hydrocolloids,
23, 1746e1755.
Funami, T., Zhang, G., Hiroe, M., Noda, S., Nakauma, M., Asai, I., et al. (2007). Effects
of the proteinaceous moiety on the emulsifying properties of sugar beet pectin.
Food Hydrocolloids, 21, 1319e1329.
Gant, G. T., Morris, E. R., Rees, D. A., Smith, P. J. C., & Thom, D. (1973). Biological
interactions between polysaccharides and divalent cations: the egg-box model.
FEBS Letters, 32, 195e198.
Hotchkiss, A. T. J. R., Savary, B. J., Cameron, R. G., Chau, H. K., Brouillette, J.,
Luzio, G. A., et al. (2002). Modification of pectin to increase its calcium sensi-
tivity while preserving its molecular weight. Journal of Agricultural and Food
Chemistry, 50, 2931e2937.
JP Patent Application Publication JP2006-56A.
Keogh, M. K., & O’ Kennedy, B. T. (1998). Rheology of stirred yogurt as affected by
76 M. Nakauma et al. / Food Hydrocolloids 55 (2016) 65e76
added milk fat, protein and hydrocolloids. Journal of Food Science, 63, 108e112.
Klavons, J. A., & Bennett, R. B. (1986). Determination of methanol using alcohol
oxidase and its application to methyl ester content of pectins. Journal of Agri-
cultural and Food Chemistry, 34, 597e599.
Kohn, R. (1975). Ion binding on polyuronates e alginate and pectin. Pure and Applied
Chemistry, 42, 371e397.
Kohn, R., & Larsen, B. (1972). Preparation of water-soluble polyuronic acids and
their calcium salts, and the determination of calcium ion activity in relation to
the degree of polymerization. Acta Chemica Scandinavica, 26, 2455e2468.
Kong, H. J., Kaigler, D., Kim, K., & Mooney, D. J. (2004). Controlling rigidity and
degradation of alginate hydrogels via molecular weight distribution. Bio-
macromolecules, 5, 1720e1727.
Liao, H., Ai, W., Zhang, K., Nakauma, M., Funami, T., Fang, Y., et al. (2015). Mecha-
nisms of oligoguluronate modulating the calcium-induced gelation of alginate.
Polymer, 74, 166e175.
Li, L., Fang, Y., Vreeker, R., & Appelqvist, I. (2007). Reexamining the egg-box model in
calcium-alginate gels with X-ray diffraction. Biomacromolecules, 8, 464e468.
Luzio, G. A., & Cameron, R. G. (2008). Demethylation of a model homogalacturonan
with the salt-independent pectin methylesterase from citrus: Part II. Structur-
eefunction analysis. Carbohydrate Polymers, 71, 300e309.
Mao, R., Tang, J., & Swanson, B. G. (2001). Water holding capacity and microstruc-
ture of gellan gels. Carbohydrate Polymers, 46, 365e371.
Moe, S. T., Draget, K. I., Skjak-Braek, G., & Smidsrød, O. (1995). Chapter 9: alginates.
In A. M. Stephen (Ed.), Food polysaccharides and their applications (pp.
245e286). New York: Marcel Dekker.
Mørch, Y. A., Donati, I., Strand, B. L., & Skjåk-Bræk, G. (2006). Effect of Ca2þ, Ba2þ,
and Sr2þ on alginate microbeads. Biomacromolecules, 7, 1471e1480.
Morris, E. R., Rees, D. A., Thom, D., & Boyd, J. (1978). Chiroptical stoichiometric
evidence of a specific, primary dimerization process in alginate gelation. Car-
bohydrate Research, 66, 145e154.
Picout, D. R., Ross-Murphy, S. B., Errington, N., & Harding, S. E. (2001). Pressure cell
assisted solution characterization of polysaccharides. 1. Guar gum. Bio-
macromolecules, 2, 1301e1309.
Ralet, M. C., Dronnet, V., Buchholt, H. C., & Thibault, J. F. (2001). Enzymatically and
chemically de-esterified lime pectins: characterisation, polyelectrolyte behav-
iour and calcium binding properties. Carbohydrate Research, 336, 117e125.
Rees, D. A. (1982). Polysaccharide conformation in solutions and gels e recent re-
sults on pectins. Carbohydrate Polymers, 2, 254e263.
Sikorski, P., Mo, F., Skjåk-Bræk, G., & Stokke, B. T. (2007). Evidence for egg-box-
compatible interactions in calcium-alginate gels from fiber X-ray diffraction.
Biomacromolecules, 8, 2098e2103.
da Silva, J. A. L., & Rao, M. A. (2006). Pectins: structure, functionality, and uses. In
A. M. Stephen, G. O. Phillips, & P. A. Williams (Eds.), Food polysaccharides and
their applications (2nd ed., pp. 353e411). London: Taylor & Francis.
Stokke, B. T., Smidsrød, O., Bruheim, P., & Skjåk-Bræk, G. (1991). Distribution of
uronate residues in alginate chains in relation to alginate gelling properties.
Macromolecules, 24, 4637e4645.
Thibault, J. F., & Rinaudo, M. (1985). Interactions of mono- and divalent counterions
with alkali- and enzyme-deesterified pectins in salt-free solutions. Biopolymers,
24, 2131e2143.
Thom, D., Grant, G. T., Morris, E. R., & Rees, D. A. (1982). Characterisation of cation
binding and gelation of polysaccharides by circular dichroism. Carbohydrate
Research, 100, 29e42.
US Patent Application Publication US2010/0256090.
Walls, H. J., Caines, S. B., Sanchez, A. M., & Khan, S. A. (2003). Yield stress and wall
slip phenomena in colloidal silica gels. Journal of Rheology, 47, 847e868.