Food Hydrocolloids 79 (2018) 170e178

Contents lists available at ScienceDirect

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

Rheological and structural properties of coagulated milks

reconstituted in D2O: Comparison between rennet and a tamarillo
enzyme (tamarillin)
Zhao Li a, Zhi Yang a, Don Otter b, Christine Rehm c, d, Na Li e, Peng Zhou f,
Yacine Hemar a, g, *
a
School of Chemical Sciences, The University of Auckland, Private Bag, 92019, Auckland, New Zealand
b
Center for Dairy Research, College of Agriculture and Life Sciences, University of Wisconsin-Madison, 1605 Linden Drive, Madison, USA
c
Australian Centre for Neutron Scattering, Australian Nuclear Science and Technology Organisation, Locked Bag 2001, Kirrawee, DC, Australia
d
Guangdong Technion Israel Institute of Technology, 241 Da Xue Road, Shantou, Guangdong Province, 515063, China
e
National Centre for Protein Science Shanghai, Chinese Academy of Sciences, Shanghai, 201204, China
f
State Key Laboratory of Food Science and Technology, School of Food Science and Technology, Jiangnan University, Wuxi, Jiangsu Province, 214122, China
g
The Riddet Institute, Palmerston North, New Zealand

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

Article history: The physicochemical properties and structural characteristics of milk gels in deuterium oxide (D2O)
Received 17 October 2017 induced by rennet and tamarillo enzyme (tamarillin) were investigated. SDS-PAGE showed that both
Received in revised form proteinases hydrolysed k-casein while tamarillin also exhibited a broader specificity on a- and b-casein.
23 November 2017
Small and large deformation rheological measurement showed that rennet-induced milk gels at low milk
Accepted 4 December 2017
Available online 7 December 2017
solid content (10% w/w) displayed higher elasticity than the gels induced by tamarillin. In high con-
centration milk gels (20% w/w), the aggregation time was sharply decreased and the elasticity was
increased for both rennet and tamarillin induced gels. In addition, large deformation experiments
Keywords:
Skim milk gel
indicated that rennet-induced gels are more brittle than those made with tamarillin for both 10 and 20%
Tamarillin (w/w) milk gels. The microstructures of the milk gels displayed more porosity in the gels made with
Rennet tamarillin, compared to those made with rennet, likely due to the broader caseinolytic activity of the
Rheology tamarillin. Ultra-small angle neutron scattering (USANS) showed that there are no differences in the
USANS large-scale structure of tamarillin and rennet induced gels. However, small angle X-ray scattering (SAXS)
SAXS revealed that the milk gel fine structures are different.
© 2017 Elsevier Ltd. All rights reserved.

1. Introduction
Cheese is one of the traditional processed dairy products ob-
tained by first coagulating the milk (Omotosho, Oboh, & Iweala,
2011). The milk clotting enzyme rennet is usually used in the
cheesemaking industry because of its high specificity to cleave
Phe105-Met106 in k-casein. Hence, the stability between casein
micelles is affected resulting in the aggregation of casein micelles
and subsequent milk clotting (Pires et al., 1994; Vishwanatha, Rao,
& Singh, 2010). However, the increase in the demand for cheese
products and the world-wide shortage of rennet result in the
* Corresponding author. School of Chemical Sciences, The University of Auckland,
Private Bag, 92019, Auckland, New Zealand.
E-mail address: y.hemar@auckland.ac.nz (Y. Hemar).
increase of the price of rennet (Jacob, Jaros, & Rohm, 2011). Other
factors such as the restriction of the use of animal rennet for reli-
gious and diet (vegetarianism) reasons, motivated the search for
rennet substitutes (Roseiro, Barbosa, Ames, & Wilbey, 2003).
Although bacterial and genetically engineered rennet have been
proven to be good milk coagulants, recently the interest in plant
proteases has grown (Shah, Mir, & Paray, 2014). Plant-based co-
agulants such as extract from the flower of Cynara cardunculus has
been traditionally used in ovine and caprine cheeses (Pires et al.,
1994; Trujillo, Carretero, & Guamis, 1994; De Sa & Barbosa, 1972).
Plant proteases and rennet share the similar feature of hydro-
lyzing the Phe105-Met106 in k-casein, but plant protease also have
broader proteolytic activity on a-, b- and k-casein (Macedo, Faro, &
Pires, 1993). Ahmed, Babiker, and Mori (2010) investigated the ac-
tion of a milk clotting enzyme from the seeds of Solanum dubium on

https://doi.org/10.1016/j.foodhyd.2017.12.004
0268-005X/© 2017 Elsevier Ltd. All rights reserved.
 Z. Li et al. / Food Hydrocolloids 79 (2018) 170e178
bovine caseins, and reported that all caseins are more sensitive to
the action of this enzyme compared to rennet. They also reported
that b- and k-casein are more susceptible to hydrolysis than a-
casein. Egito et al. (2007) studied the proteolytic activity of extracts
from the seeds of Albizia lebbeck and Helianthus annuus, and also
reported that a- and b-casein are more susceptible to the action of
these plant extracts than rennet. The extract from Helianthus
annuus has the same cleavage site on k-casein than rennet, while
Lys116-Thr117 bond was the preferred target of extract from Albizia
lebbeck. In addition, actinidin extract from kiwi fruit hydrolyses
preferably b-casein than k-casein (Lo Piero, Puglisi & Petrone,
2011).
Mazorra-Manzano et al. (2013) compared milk clotting behav-
iour of extracts from kiwi fruit, melon and ginger with rennet. Their
results showed that the curds induced by extract from kiwi fruit
had similar properties to that produced by rennet. However, the
melon extract-induced curds were more fragile and had a lower
yield than that produced by rennet. The properties of curds were
related to the proteolytic activities of these enzymes, with the
melon extract displaying a higher proteolytic activity compared to
kiwi fruit extract and rennet. Esteves, Lucey, and Pires (2001) also
stated that plant coagulant extracts from the flowers of Cynara
cardunculus and Cynara humilis exhibited higher proteolytic activ-
ities than that of rennet, which resulted in lower milk gel firmness
due to higher casein hydrolysis of plant coagulants.
In this study a serine protease isolated from tamarillo fruit, with
rennet-like properties, is investigated. This protease has never been
reported before, and will be referred to here as tamarillin. The effect
of tamarillin on the physicochemical properties and structural
characteristics of milk gels is compared to those obtained by the
addition of rennet. In addition, the effect of milk concentration on
the gel properties was also considered. The milks were prepared by
reconstituting skim milk powders in D2O to allow fast Ultra-Small
Angle Neutron Scattering (USANS) measurements. Please note
that while the effect of D2O on the casein micelle structure is still
unknown, De Kruif, Huppertz, Urban, and Petukhov (2012) re-
ported that the radius of gyration, measured by SANS, and the
hydrodynamic radius, measured by dynamic light scattering, were
the same for casein micelles suspended in D2O and in H2O. SDS-
PAGE was performed to assess the extent of casein hydrolysis by
rennet and tamarillin, and Confocal Laser Scanning Microscopy
(CLSM) was employed for observation of the gel network. Rheo-
logical measurements, both under small and large deformation,
were performed to determine the mechanical properties of the
resulting milk gels. Small-Angle X-ray Scattering (SAXS) measure-
ments were performed to probe the fine-structure of the gel pro-
tein network, while USANS was used to probe the protein network
on a larger scale on the milk gels. Note that although the milk gel
formation induced by acidification has been studied using SAXS
and USANS before (Van Heijkamp, de Schepper, Strobl, Tromp,
Heringa, & Bouwman, 2010; Moitzi, Menzel, Schurtenberger, &
Stradner, 2010), to the best of our knowledge these methods have
never been previously used to study milk gels obtained by an
enzymatic action.
2. Materials and methods
2.1. Materials
Low heat skim milk powder (SMP) was obtained from Synlait
Milk Ltd., Rakaia, New Zealand. Deuterium oxide (D2O) was pro-
duced in Cambridge Isotope Laboratories, Inc., Andover, USA.
Rennet (280 International Milk Clotting Units) was ordered from
RENCO, Eltham, New Zealand. Capillary tubes (2 mm I.D) were
purchased from Charlessupper Company, Natick, USA. All other
171
chemicals and reagents were purchased from Sigma-Aldrich Ltd.,
Auckland, New Zealand. Laird's Large tamarillo fruits were hand-
picked from the trees in an orchard in Maungatapere, New Zealand.
2.2. Tamarillo protease extraction and purification
Purification of tamarillo enzyme was based on McDowall (1970)
and Aminlari, Shekarforoush, Gheisari, and Golestan (2009).
Tamarillo fruits (300 g) were homogenized for 20 min using a
Polytron PCU2 laboratory homogeniser (Brinkmann Instruments,
Luzern, Switzerland), followed by filtration through two layers of
cheesecloth to remove the insoluble material. Filtrate of tamarillo
(around 85 mL) was mixed with 80 mL of 0.05 M sodium citrate (pH
5.5) buffer by gentle stirring, then the mixture was centrifuged at
15,000
g for 20 min at 4  C in a Sorvall LYNX superspeed
Centrifuge (Thermo Fisher Scientific, USA). The resulting superna-
tant (around 155 mL solution) was precipitated with (NH4)2SO4
(65% saturation) using gentle stirring, overnight at 4  C. The
precipitated proteins were collected after centrifugation at
15,000
g for 20 min at 4  C and removing the supernatant. The
collected fraction was then dissolved in 30 mL 0.05 M sodium cit-
rate (pH 5.5) buffer. The (NH4)2SO4 was removed by dialysis against
0.05 M sodium citrate buffer (pH 5.5), for 24 h at 4  C.
The tamarillo crude extract solution (40 mL, 1.17 ± 0.03 mg/mL),
obtained using the steps above, was loaded on a DEAE-Sepharose
fast flow (GE Healthcare Life Sciences, New Zealand) ion ex-
change column (22 cm
5 cm) pre-equilibrated with 0.05 M so-
dium citrate buffer (pH 5.5) overnight. The proteins were separated
by eluting with 200 mL linear gradient of 0e1.0 M NaCl at a flow
rate of approximately 1.0 mL/min. Each 10 mL eluted solution was
collected as a fraction. The protein content was determined using a
spectrophotometer (Shimadzu Corporation, Australia) at 280 nm.
Fractions corresponding to peak areas were pooled together, and
the pooled solution with the highest activity was dialysed against
Milli-Q water to remove salt for 24 h at 4  C. The protease was
concentrated using a freeze-drier (Labconco Corporation, Missouri,
USA), then stored at 80  C until further use.
2.3. Preparation of reconstituted milks and milk gels
The reconstituted milks in this study were prepared by mixing
skim milk powder with D2O in appropriate amounts. The concen-
trations of milk samples were 1%, 2%, 10%, 20% and 25% (w/w). Milk
powder and D2O were first mixed using a magnetic stirrer (Hei-
dolph Instruments, Germany) for at least 2 h at room temperature
to ensure full dissolution. The mixtures were kept in a 4  C fridge
overnight to ensure complete hydration. The mixtures were
equilibrated to room temperature for at least 4 h before use.
The 10% (w/w) milk gel samples were prepared by mixing either
25 g of 10% (w/w) milk with 25 mL of 10 times-diluted rennet or
12.5 g of 20% (w/w) milk with 50 mg tamarillin pre-dissolved in
12.5 g D2O. In case of the 20% (w/w) milk samples, which were
prepared by mixing either 25 g of 20% (w/w) milk with 50 mL of 10
times-diluted rennet or 20 g of 25% (w/w) milk with 100 mg
tamarillin pre-dissolved in 5 g D2O. In addition, the preparation of
1% (w/w) milk samples was carried by mixing either 5 g 1% (w/w)
milk with 10 mL 100 times-diluted rennet or 2.5 g of 2% (w/w) milk
with 2.5 g 2 mg tamarillin pre-dissolved in 5 g D2O. These milk and
enzyme concentrations were chosen in order to have a desired final
concentration of milk and to keep constant the enzyme to protein
ratio.
To obtain the milk serums, milk gels (10% and 20% w/w) were
incubated at 35  C for 10 h in 40 mL volume glass vials. Then, gels
were cut using a thin spatula into roughly 10 mm3 cubes. The cut
gel was transferred into a 50 mL centrifuge tube and centrifuged at
172 Z. Li et al. / Food Hydrocolloids 79 (2018) 170e178
1500
g for 30 min at room temperature using Eppendorf
centrifuge 5810R (North Ryde, Australia). Centrifugation allowed
obtaining a supernatant serum and a precipitated gel curd. The
serum was collected then further filtered through a 0.22 mm filter
(TISCH Scientific, Ohio, USA) to be measured by USANS and SAXS
for baseline subtraction.
2.4. Analysis of casein hydrolysis by SDS-PAGE
An aliquot (100 ml) of the milks with added rennet or tamarillin,
as described in section 2.3 above, were transferred into glass vials
and incubated at 35  C for 10 h in an incubator (Function Line
incubator, Heraeus, Langenselbold, Germany). After 10 h, the
enzymatic hydrolysis was stopped by heating the glass vials at
100  C for 3 min in a water bath (Ahmed et al., 2010). Urea buffer
(8 M, pH 8) was added into each sample to break non-covalent
bonds between proteins and peptides using vortexing for 5 min.
To achieve the same final protein concentration, 1%, 10% and 20%
(w/w) milk gel samples were added with 100 mL, 400 mL and 900 mL
urea buffer respectively. Aliquot (100 mL) of the mixtures were
mixed with 200 mL SDS-PAGE sample buffer, consisting of 50 mM
Tris-HCl (pH 6.8), 2% (w/w) SDS, 0.1% (w/w) bromophenol blue, 10%
(v/v) glycerol and 5% (v/v) b-mercaptoethanol.
Following Laemmli (1970), SDS-PAGE was performed using 4%
acrylamide stacking gel and 12% acrylamide resolving gel con-
taining 0.1% SDS. Samples (10 mL) were injected into the gel wells.
Electrophoresis was conducted at 150 V for 2 h in 0.025 M Tris-HCl,
0.192 M glycine buffer containing 0.1% (w/v) SDS using Mini-
PROTEAN tetra cell (Bio-Rad Technologies Inc., California, USA).
After running the SDS-PAGE, the gels were stained using 0.1%
Coomassie brilliant blue R-250 overnight. And the stained gel was
distained using 40% (v/v) methanol, 10% (v/v) acetic acid and 50%
(v/v) Milli-Q water.
2.5. Rheological measurements
Rheological characterization was conducted in an Anton Paar
Physica MCR 301 stress-controlled rheometer (Anton-Paar, Graz,
Austria), equipped with a stainless-steel cup and bob geometry
system (26.5 mm diameter and 48 mm length bob, and a cup with
an inner diameter of 27.5 mm). Milk sample (18.5 mL) containing
rennet or tamarillin, was loaded into the cup and bob geometry. The
samples were prepared as described in section 2.3 above, stirred for
2 min on magnetic stirrer to mix the enzyme solution and the
milks. A few drops of soya oil were added to cover the surface of
milk sample to prevent water evaporation during measurement.
The following three rheological characterization protocols were
performed: (1) A time sweep was performed by first equilibrating
the sample temperature at 25  C for 2 min, followed by increasing
temperature to 35  C at a rate of 2  C/min. The elastic modulus G0
(storage) and the viscous (loss) modulus G00 were recorded at a
constant frequency of 1 Hz and a constant applied strain of 1% every
minute for 10 h; (2) At the end of the time sweep, the temperature
was lowered to 25  C for 10 min, and a frequency sweep mea-
surement was performed where G0 and G00 were measured as a
function of frequency from 0.1 to 10 Hz at a constant strain of 1% at
25  C; (3) The frequency sweep measurement, was followed by a
large-deformation sweep performed with a constant shear rate 0.1
s1 for a total interval time of 2000 s at 25  C. During this large-
deformation sweep test the stress was measured, while the strain
(¼shear rate
time) was varied from 0 to 200%.
2.6. Confocal laser scanning microscopy
The microstructures of milk gels were investigated by confocal
laser scanning microscopy (CLSM) following Esteves, Lucey, Wang,
and Pires (2003) with minor modifications. The milk samples were
prepared in the same way as those prepared for rheological mea-
surements. Few drops of 0.25% (w/v) acridine orange were mixed
with the milk samples using a magnetic stirrer for 2 min at room
temperature. Aliquots of mixtures (100 mL) were transferred into a
slide with a cavity. A coverslip was used to cover the milk sample
and was sealed by nail polish to prevent evaporation. Then the
slides were incubated in an oven for 10 h at 35  C. The micro-
structure of gels was observed using an Olympus FV1000 CLSM
microscope (Olympus Corporation, Tokyo, Japan) with an X-Cite
series 120Q mercury halide lamp (Lumen Dynamics, Ontario,
Canada) at a wavelength of 488 nm. Images of gel microstructure
were obtained using a 60
oil immersion objective lens. Each
sample was prepared in duplicate and at least three representative
areas from the same sample were observed.
2.7. Ultra small angle neutron scattering (USANS)
USANS measurements, based on the Bonse-Hart method (Bonse
& Hart, 1965), were performed on the Kookaburra beamline at
ANSTO (Sydney, Australia) (Rehm, Brûle , Freund, & Kennedy, 2013).
It consists of two channel-cut perfect Si single crystals, namely a
monochromator and an analyser. Milk in D2O samples were
measured by USANS at a neutron wavelength l ¼ 4.74 Å and a Cd
aperture with a diameter of 30 mm, at a Q range of
0.0003e0.1 nm1, at room temperature. The milk samples were
filled into demountable sample cells. Then the sample cells were
incubated at 35  C in an oven (Laboratory equipment PTY. Ltd.
Australia) for 10 h. Note that in order to optimise the intensity of
the neutron scattering, cell pathlengths of 1 mm and 0.5 mm were
used for the 10% (w/w) and 20% (w/w) milks and milk gels,
respectively.
2.8. Small angle X-ray scattering (SAXS)
The samples used in SAXS experiments were prepared similarly
to those used in USANS measurements. The main difference is that
the milk gels were prepared by transferring the milks-enzyme
mixtures into 2 mm inner diameter capillary quartz tubes (Char-
lessupper Company, Natick, USA), sealed by para-film, followed by
incubation at 35  C for 10 h in a BPZ-6123 oven (Shanghai Yiheng
Scientific Instruments Co., Ltd. Shanghai, China). The serums of the
milk gels were also measured in the same capillary tubes.
Synchrotron Small-Angle X-ray Scattering (SAXS) measure-
ments were performed on the BL19U2 BioSAXS beamline of Na-
tional Centre for Protein Science Shanghai (NCPSS) at the Shanghai
Synchrotron Radiation Facility (SSRF, Shanghai, China). A mono-
chromatic beam of 1.033 Å wavelength (l) was used and sample-to-
detector distance of 2234.4 mm, yielding a q-range (q ¼ 4psinq/l, q
is the scattering angle) between 0.06 and 4 nm1. The SAXS profiles
of the samples were analysed by subtracting the background of the
corresponding serum using the BioXTAS RAW software version
1.2.1 (Nielsen et al., 2009). All the samples were measured at room
temperature.
3. Results and discussion
3.1. Protein profile and proteolytic activity
Extent of skim milks hydrolysis by rennet or tamarillin after 10 h
incubation was investigated by SDS-PAGE. In the case of rennet,
compared to untreated milk (Fig. 1, line 8) there is a small amount
k-casein remaining for 1% (w/w) milk (Fig. 1, lane 2), and a peptide
band appearing at around 14 kDa. In the case of 10% (w/w) (Fig. 1,
 Z. Li et al. / Food Hydrocolloids 79 (2018) 170e178 173

Fig. 1. SDS-PAGE of skim milk hydrolysed by rennet and tamarillin, after 10 h incu-
bation. Lane 1 is protein marker, lane 2, 3, 4 are rennet-treated 1%, 10% and 20% (w/w)
skim milks, respectively. Lane 5, 6, 7 are for 1% 10% and 20% (w/w) skim milks treated
with the tamarillin, respectively. Lane 8 is 10% (w/w) skim milk alone. Milk protein
bands are indicated by a-casein (a-CN), b-casein (b-CN), k-casein (k-CN), b-lacto-
globulin (b-Lg) and a-lactalbumin (a-Lac).

lane 3) and 20% (w/w) (Fig. 1, lane 4) milks, k-casein was completely
hydrolysed and a clear peptide band can be seen around 14 kDa.
This new peptide band consist of para-k-casein, and similar result
has been reported for the action of rennet on milk by Timotijevi c,
Radovi c, and Maksimovic (2006). Both a- and b-casein in all the
milks treated by rennet (Fig. 1, lane 2, 3 and 4) are not affected
when compared to untreated milk (Fig. 1, lane 8). This is expected,
as previous studies also reported that compared to k-casein, a- and
b-casein are less susceptible to the action of rennet (Fox, Singh, &
McSweeney, 1994). However, in the case of tamarillin all the ca-
seins are hydrolysed (Fig. 1, lane 5, 6 and 7) after 10 h incubation. k-
caseins in all the milks treated by tamarillin has been fully hydro-
lysed, and only small amounts of a- and b-casein remained. The
amounts of the non-hydrolysed a- and b-casein seems to be
dependent on the concentration of milk, with the higher the milk
concentration, the higher the amount of these caseins remaining.
Compared to the action of rennet, there are more peptide bands
(between 25 and 10 kDa) generated through the action of tamar-
illin, particularly peptides with molecular weights close to 18 kDa,
16 kDa and 13 kDa. This clearly indicates the broad hydrolytic ac-
tivity of tamarillin. Similar findings were reported for the effect of
other of plant proteases on caseins, including lettucine extracted
from leaves of Lactuca sativa (Lo Piero, Puglisi, & Petrone, 2002),
and protein extracts from seeds of Albizia lebbeck and Helianthus
annuus (Egito et al., 2007).

3.2. Rheological measurements

The G0 and G00 as a function of time for the 10 and 20% (w/w)
milks containing rennet or tamarillin are shown in Fig. 2A. The
corresponding complex modulus G* (¼ ((G0 )2 þ (G00 )2)1/2), which
take into consideration the contribution of both G0 and G00 is also
reported in the inset of Fig. 2A. Qualitatively, the gelation behav- Fig. 2. (A) G0 , G00 , and G* as an inset figure, as a function of time, (B) frequency sweep,
and (C) large deformation rheological measurements of milks gelled with the addition
iours measured by rheology is similar to those previously reported of rennet or tamarillin. Symbols correspond to: 10% (w/w) rennet-induced milk gel ( );
for milk treated with rennet (Waungana, Singh, & Bennett, 1998). It 10% (w/w) tamarillin-induced milk gel ( ); 20% (w/w) rennet-induced milk gel (:);
can be seen that the kinetics of gelation for 20% (w/w) milks is and 10% (w/w) tamarillin-induced milk gel ( ).
faster than that of the 10% (w/w) milks. Further, at the same milk
concentration, the kinetics of gelation of milks treated with rennet
is also faster than that of the milks treated with tamarillin. Similar reported that ~1 h after enzyme coagulant addition to milk at 32  C,
finding was reported by Esteves, Lucey, and Pires (2002), who the increase rate of G0 of rennet-induced gel were faster than plant
174 Z. Li et al. / Food Hydrocolloids 79 (2018) 170e178
coagulants extracts from Cynara cardunculus and Cynara humilis
induced gels. The gelation time, defined here as the time at which
G0 ¼ G00 , was 52.0 ± 1.4 min and 117.0 ± 2.8 min for 10% (w/w) milks
containing rennet and tamarillin, respectively; and 24.0 ± 0.0 min
and 35.0 ± 1.4 min for 20% (w/w) milk containing rennet and
tamarillin, respectively. After 10 h gelation the value of G* was
222.0 ± 0.0 Pa and 86.4 ± 0.7 Pa, for 10% (w/w) milks treated with
rennet and tamarillin respectively and 1895.0 ± 35.4 Pa and
1645.0 ± 7.0 Pa, for 20% (w/w) milks treated with rennet and
tamarillin, respectively. The fact that the gelation kinetics is faster
for the high concentration milks compared to the low ones, can be
understood in terms of the number of caseins micelles involved,
and the closer they are from each other (Culioli & Sherman, 1978;
McMahon, Yousif, & Kala b, 1993; Waungana, Singh, & Bennett,
1996). After 10 h, G* is also higher for the high concentration
milks and this can be understood by the greater number of bonds
between the milk protein aggregates in the gel network (Waungana
et al., 1996, 1998). Ong, Dagastine, Kentish, and Gras (2013) stated
that rennet-induced milk gels formed with concentrated milks
were much firmer, as the concentration of milk protein increased
and the distance between casein micelles decreased. Green and
Grandison (1993) indicated that the increase in the gel firmness
is due to the increase in number and strength of the casein micelle
network. The lower G* observed for tamarillin treated milks
compared to rennet treated milks, is likely due to the broad hy-
drolysis of tamarillin. In other word, as shown by SDS-PAGE more
caseins are hydrolysed by tamarillin than rennet, resulting in less
intact caseins present in the protein network. Esteves et al. (2002)
reported that the G0 values in rennet-induced gels were higher than
that obtained from plant coagulants (from Cynara cardunculus and
Cynara humilis) induced gels. This was explained by the higher
extent of rearrangements in the milk gels obtained by the action of
plant coagulants, compared to rennet-induced gels (Esteves et al.,
2002).
To investigate the viscoelastic behaviour of the milks gels, a
frequency sweep was carried out at the end of the time sweep
(Fig. 2B). For all gels, G0 is higher than G00 , and both G0 and G00 in-
crease only slightly with the increase in frequency. This confirms
the gel-like behaviour of the milk gels obtained by the addition of
rennet or tamarillin, however, because G0 is larger than G00 by less
than ten-fold, this indicates that these gels are weak gels
(Fernandes, Gonçalves, & Doublier, 1991; Lapasin & Pricl, 1995).
A strain deformation measurement was performed after the
completion of the frequency sweep to investigate the large defor-
mation rheological behaviour of the milk gels (Fig. 2C). All the
samples exhibit a similar behaviour, within a low strain region,
where the stress is proportional to the strain. This linear region
occurs at strain <10%, and confirms that the frequency sweep
measurement, performed at 1% strain, was carried out in the Linear
Viscoelastic Region. This linear region is followed by a non-linear
region at the end of which the stress reaches a maximum. This
non-linear region, where the stress increases with the strain, in-
dicates that these gels behave as strain-hardening systems (Groot,
Bot, & Agterof, 1996). After a maximum stress is reached, the
stress starts to decrease with the increase in the strain, indicating
that the gels start to break. To compare between the different milk
gels the value of maximum stress (smax) and maximum strain (g
max) are analysed. As can be seen in Fig. 2C, for 10% (w/w) milks, the
value of smax is 52.70 ± 8.34 Pa and the value of g max is
44.70 ± 2.12% for milks gelled by rennet, and the value of smax is
74.65 ± 3.04 Pa and the value of g max is 96.00 ± 2.12% for milks
gelled by tamarillin. In the case of 20% (w/w) milks, the value of
smax is 267.50 ± 10.61 Pa and the value of g max is 45.70 ± 0.71% for
milks gelled by rennet, and the value of smax is 560.00 ± 8.49 Pa and
the value of g max is 93.00 ± 0.71% for milks gelled by tamarillin.
Clearly, for a given milk concentration, smax and g max of milks
gelled by tamarillin are higher than those of the milks gelled by
rennet. These results do not contradict those obtained by the time
sweep and the frequency sweep, obtained by small deformation
rheological measurements, where the gels made with rennet were
found to be more elastic than those made with tamarillin. The re-
sults obtained by the large strain experiments indicates that milk
gels made with tamarillin are less brittle than those made with
rennet. This could be a consequence of the extensive hydrolysis of
the casein by tamarillin. In which case the protein network in the
tamarillin-induced milk gels, is made of “weaker” casein aggre-
gates, which will deform more under larger strains before breaking.
3.3. Confocal laser scanning microscopy (CLSM) observations
The confocal micrographs for different milk gels are depicted in
Fig. 3. The protein network, due to the staining of the proteins,
appears white, while the voids made of the milk serums are dark.
After 10 h incubation, the micrograph of the 10% (w/w) milk with
added rennet shows clearly a protein network with small voids
distributed throughout the gel (Fig. 3A). Similarly, the 10% (w/w)
milk made with tamarillin show similar features (Fig. 3B), with the
voids being larger than those observed in the 10% (w/w) milk made
with rennet. The larger porosity observed for tamarillin-induced
gels might be attributed to the extensive hydrolysis of the ca-
seins. This is more evident in the 20% (w/w) milk gels, where the
micrograph of milk gelled with rennet show smaller voids and a
more homogeneous protein network (Fig. 3C), while the micro-
graph of the 20% (w/w) milk gelled by tamarillin show clearly a
protein network with larger voids (Fig. 3D).
Microscopy observation remain qualitative and cannot allow the
determination of the macroscopic behaviour of the milk gels, as
obtained by rheology. However, these CLSM experiments show
clearly that all the milks with added rennet or tamarillin resulted in
a protein network. Further, they indicate that the voids in the milk
gels obtained by tamarillin treatment are bigger than those
observed in the milks gelled by rennet. This might explain the large
deformation behaviour of these gels (Fig. 2C), where a gel with a
denser structure (rennet milk gels) are expected to break at lower
strains, than more flexible gels such as those obtained by tamarillin
with larger voids.
3.4. USANS and SAXS investigations
Small-angle X-ray scattering (SAXS) and small-angle neutron
scattering (SANS) have been extensively employed to characterize
the structure of casein micelles over a wide range of length scales.
SAXS and SANS can determine the nanostructures of casein mi-
celles in their native and unperturbed state avoiding the compli-
cations of sample preparation steps including drying, and oxidative
dying that usually is used in various microscopy (Nakagawa &
Kagemoto, 2013). Here, by combining ultra-small angle neutron
scattering (USANS) and SAXS, the hierarchical milk gel structures,
induced by rennet and tamarillin, in length scales ranging from few
nanometres to several microns can be probed. Firstly, USANS is
employed to examine the large-scale structures of the different
milk gels. The USANS data (Fig. 4A and B) have been obtained for a
wavevector 5
104Å1<q < 102 Å1, corresponding to a probed
length scale of 60 nm to 1.2 mm. The scattering curves of the milk
gels show a scattering intensity, I (q), that follows a power law
decay equation given as:
I ðqÞ  qm (1)
Where m is the power law exponent, from which the nature and
 Z. Li et al. / Food Hydrocolloids 79 (2018) 170e178
Fig. 3. Confocal laser scanning micrographs of 10% (w/w) (A and B) and 20% (w/w) (C and D) skim milk gel induced by rennet (A and C) or tamarillin (B and D) after 10 h gelation.
Scale bar ¼ 20 mm.
geometry of the scattering object can be determined. For example,
0 < m < 3 may refer to mass fractal structures (three dimensional
self-similarity over a large scale lengths); which indicates the
compactness of the object (Martin & Hurd, 1987).
Across the USANS q-range, all milk gels with different milk solid
contents and coagulants (rennet or tamarillin) demonstrate two
power-law regimes as shown in Fig. 4A and B. More specifically, in
the q-range of around 3
104 to 3
103 Å1, a power law
exponent of m~2 can be observed for both rennet and tamarillin
coagulated milk gels. In this q-range, both 10% (w/w) milk gels
formed by rennet and tamarillin may be described as mass fractal
as indicated by a power law exponent of m, 2.10 ± 0.03 and
2.15 ± 0.05. Similarly, the power law exponents of m, 2.20 ± 0.06
and 2.20 ± 0.04 can be observed for both 20% (w/w) milk gels
formed by rennet and tamarillin, respectively. The similarities in
exponent, m, in the low q range that characterize the gel large-scale
structure suggest similar morphologies of the milk gel networks
coagulated by rennet and tamarillin for both 10% and 20% (w/w)
milk. This mass fractal behaviour is also observed in acid induced
skim milk gels (Van Heijkamp et al., 2010) and other colloidal ag-
gregates systems (Schaefer, Martin, Wiltzius, & Cannell, 1984).
In the high-q region, there is a distinct crossover to a second
power law regime with an exponent of ~4. This slope was originally
calculated by Porod for two-phase systems with sharp boundaries
(Guimer, Fournet, & Walker, 1955). For both rennet and tamarillin
175
coagulated milk gels, the crossover point is around q ¼ 0.003 Å1,
corresponding to the scattering size of 200 nm, a typical size of the
casein micelles (Holt, 2016). The fact that the crossover point and
the Porod slope is the same, suggests that the casein micelles
remain intact in the milk gels (Schaefer et al., 1984).
In terms of milk, both 10% and 20% (w/w) milk demonstrate
similar USANS scattering patterns. The I (q) over the low q regime
(around 5
104Å1 to 3
104Å1) can be fitted by the Guinier
function (Hayter & Penfold, 1983), and decays by a q4 power law in
high q regime. The radius of gyration (Rg) of casein micelles
determined by the Guinier analysis for 10% and 20% (w/w) milk is
83 ± 3 nm and 82 nm, respectively. This compares well with
another study where Rg z 100 nm was reported for casein micelles
(Pignon et al., 2004). The scattering profile decays by a q4 power
law in high q regime is a typical Porod-like casein micelles spheres
with sharp interfaces between casein micelles and the surrounding
solvent (Day, Raynes, Leis, Liu, & Williams, 2017).
To extend the structural characterization down to the nm scale
and even smaller scales, SAXS was employed to further characterize
the structure of the milk gels induced by rennet or tamarillin. The
SAXS scattering profiles of milk (Fig. 4C and D) show similar scat-
tering characteristics as those previously reported SAXS studies
(Day et al., 2017; De Kruif, 2014; Liu et al., 2017). Generally, the SAXS
scattering pattern of 10% (w/w) milk samples coagulated with
rennet or tamarillin is similar to the respective 20% (w/w) milk
176 Z. Li et al. / Food Hydrocolloids 79 (2018) 170e178
Fig. 4. (A, B) Ultra-small angle neutron scattering (USANS) patterns and (C, D) small-angle X-ray scattering (SAXS) patterns of milks and milk gels coagulated by rennet or tamarillin
at milk solid content of 10% (w/w) and 20% (w/w).
samples. At low q region, between 0.008 and 0.02 Å1 the scat-
tering intensity was observed to decay nearly following q4 power
law, suggesting a relatively smooth and sharp interface of the
casein micelles (Mata, Udabage, & Gilbert, 2011). This agrees with
previous observations for casein micelles using SAXS experiments
on dilute milk solutions (Day et al., 2017; Liu et al., 2017). A
distinctive shoulder at q~0.08 Å1 is observed in milk samples,
which is commonly attributed to the scattering of colloidal calcium
phosphate (CCP) nanoclusters. However, De Kruif (2014) argued
that the scattering magnitude from CCP nanoclusters is too small to
be visible in the scattering profiles and instead he believes that this
shoulder should be attributed to protein inhomogeneities on
1e3 nm scale. A recent resonant soft X-ray scattering study
confirmed that the shoulder at q~0.08 Å1 was due to presence of
protein inhomogeneities within the casein micelles (Ingham et al.,
2015). In the high q region (0.08e0.18 Å1), all the scattering pro-
files followed a power law q2.5 decay, demonstrating a non-
globular internal structure with a fractal dimension of 2.5
(Marchin, Putaux, Pignon, & Le onil, 2007). The scattering profile
transition from around q4 to q2.5, suggesting characteristics of
Porod-like spheres with sharp interfaces between the casein mi-
celles and surrounding solvent (D2O) in low q region to much less
dense internal structure with a fractal dimension around 2.5 in high
q region (Day et al., 2017).
Upon addition of rennet or purified tamarillin to milk, a peak
appeared around 0.035 and 0.030 Å1 for 20% (w/w) milk coagu-
lated by rennet and tamarillin, respectively. This feature is also
observed in SANS experiments (Holt, Kruif, Tuinier, & Timmins,
2003) and SAXS of dried milk powders (Mata et al., 2011), how-
ever it is usually absent in SAXS of casein micelles in aqueous
solution (Day et al., 2017). In SANS experiments, the peak around
0.035 Å1 is most prominent when the solvent is contrast matched
to protein. While in SAXS experiments, the peak around 0.035 Å1
observed in dried milk powders other than milk solution is sug-
gested to be due to higher packing density of casein micelles found
in dried milk powder (Mata et al., 2011). The recent resonant soft X-
ray scattering study suggests that this feature corresponds to the
average separation distance between CCP nanoclusters (Ingham
et al., 2015), which is consistent with previous SANS studies (Holt
et al., 2003).
Upon addition of rennet or tamarillin to milk and incubation at
35  C for 10 h, the milk gels were formed. Therefore, the appearance
of the peak around 0.035 Å1 in milk gel could be attributed to the
increment of local packing density of casein micelles and hence CCP
nanoclusters in the gel state compared to the solution (Mata et al.,
2011). Indeed, similar features are also observed in SAXS scattering
profiles of casein micelle suspensions following ultrafiltration,
where the casein micelles packing density is greatly enhanced after
ultrafiltration (Pignon et al., 2004). It is also worth noting that 20%
(w/w) milk coagulated with tamarillin demonstrated a lower-q
peak position (0.030 Å1) compared to that of the rennet sample
(0.035 Å1). Hence, the CCP nanoclusters separation distance is
larger in tamarillin coagulated milk (~210 Å1) than that of rennet
sample (~180 Å1), according to Bragg's law d ¼ 2p/q (Windsor,
1988). These results further confirmed the CLSM observations
that rennet-induced denser protein networks (smaller voids) than
tamarillin, which also contributes to higher complex modulus of
milk gel networks made with rennet. In 10% (w/w) milk samples
with rennet and tamarillin addition, this peak is less obvious
compared to that of 20% (w/w) milk samples with rennet and
 Z. Li et al. / Food Hydrocolloids 79 (2018) 170e178
tamarillo extracts addition. Especially for 10% (w/w) milk coagu-
lated with tamarillin, where only a shoulder around ~0.030 Å1 can
be seen (Fig. 4C). This could be due to the packing density of casein
micelles, and hence CCP nanoclusters, in 10% (w/w) milk with
rennet and tamarillin addition being much lower than that of the
corresponding 20% (w/w) milk samples.
4. Conclusion
SDS-PAGE demonstrated that the tamarillo enzyme (tamarillin)
exhibited a broader casein hydrolysis than rennet, when added to
1%, 10% and 20% (w/w) milk samples prepared in D2O. In small
deformation rheological measurements, the gelation kinetics of
high concentration milk (20% w/w) is faster than that of low con-
centration (10% w/w) and the complex modulus G* in high con-
centration milk gels is higher than that of the low concentration
milk gels. This result can be understood by the high number and
density of casein micelles at higher milk concentration, which re-
sults in reduced distance between casein micelles, and the forma-
tion of more protein-protein bonds in the gel network. A typical
strain-hardening behaviour of all milk gels formed by tamarillin
and rennet can be observed under large shear deformation.
Compared to low concentration milk gels, a denser protein network
can be observed in the micrographs of the high concentration milk
gels induced by both rennet and tamarillin. However, tamarillin
induced gels showed more porosity and larger voids than rennet-
induced gels due to the high extent of casein hydrolysis. Rennet-
induced gels were found to have a denser structure than those
treated with tamarillin, which is reflected in the higher elasticity
but more brittle behaviour of these milk gels, compared to those
obtained by tamarillin. Large-scale structure, as probed by USANS,
of both rennet and tamarillin induced gels have little difference,
and similar morphologies of milk gel networks were obtained for
both 10% and 20% (w/w) milk. However, for the 20% (w/w) gels, at
smaller scales, SAXS revealed a peak in the tamarillin-induced gel
spectra at 0.030 Å1 and in the rennet-induced gel spectra at
0.035 Å1. These peaks are associated with the distance of CCP
nanocluster in milk gels, confirming that the protein network of
rennet-induced milk gels (20% w/w) are denser than their tamar-
illin counterparts. Finally, the differences found between rennet
and tamarillin induced milk gels may affect the texture properties
of cheese, thus offering the possibility for the development of new
cheese varieties with different texture characteristics.
Acknowledgements
We would like to thank Ms. Hilary Holloway for her technical
help with the CLSM experiments. USANS beam time has been
allocated by the Australian Centre for Neutron Scattering (ACNS) at
the Australian Nuclear Science and Technology Organisation
(ANSTO). The beam line BL19U2 at NCPSS and SSRF are acknowl-
edged for the provision of synchrotron beam time. We thank G. Liu,
Y. Li and H. Wu from BL19U2 beam line for assistance during SAXS
data collection and processing.
Appendix A. Supplementary data
Supplementary data related to this article can be found at
https://doi.org/10.1016/j.foodhyd.2017.12.004.
References
Ahmed, I. A. M., Babiker, E. E., & Mori, N. (2010). pH stability and influence of salts
on activity of a milk-clotting enzyme from Solanum dubium seeds and its
enzymatic action on bovine caseins. LWT-Food Science and Technology, 43(5),
177
759e764.
Aminlari, M., Shekarforoush, S. S., Gheisari, H. R., & Golestan, L. (2009). Effect of
actinidin on the protein solubility, water holding capacity, texture, electro-
phoretic pattern of beef, and on the quality attributes of a sausage product.
Journal of Food Science, 74(3), C221eC226.
Bonse, U., & Hart, M. (1965). Tailless X-ray single-crystal reflection curves obtained
by multiple reflection. Applied Physics Letters, 7(9), 238e240.
Culioli, J., & Sherman, P. (1978). Rheological aspects of renneting of milk concen-
trated by ultrafiltration. Journal of Texture Studies, 9, 257e281.
Day, L., Raynes, J. K., Leis, A., Liu, L. H., & Williams, R. P. W. (2017). Probing the in-
ternal and external micelle structures of differently sized casein micelles from
individual cows milk by dynamic light and small-angle X-ray scattering. Food
Hydrocolloids, 69, 150e163.
De Kruif, C. G. (2014). The structure of casein micelles: A review of small-angle
scattering data. Journal of Applied Crystallography, 47(5), 1479e1489.
De Kruif, C. G., Huppertz, T., Urban, V. S., & Petukhov, A. V. (2012). Casein micelles
and their internal structure. Advances in Colloid and Interface Science., 171e172,
36e52.
De Sa, F. V., & Barbosa, M. (1972). Cheese-making with a vegetable rennet from
Cardo (Cynara cardunculus). Journal of Dairy Research, 39(3), 335e343.
Egito, A. S., Girardet, J. M., Laguna, L. E., Poirson, C., Molle, D., Miclo, L., et al. (2007).
Milk-clotting activity of enzyme extracts from sunflower and albizia seeds and
specific hydrolysis of bovine k-casein. International Dairy Journal, 17(7),
816e825.
Esteves, C. L., Lucey, J. A., & Pires, E. M. (2001). Mathematical modelling of the
formation of rennet-induced gels by plant coagulants and chymosin. Journal of
Dairy Research, 68(03), 499e510.
Esteves, C. L., Lucey, J. A., & Pires, E. M. (2002). Rheological properties of milk gels
made with coagulants of plant origin and chymosin. International Dairy Journal,
12(5), 427e434.
Esteves, C. L. C., Lucey, J. A., Wang, T., & Pires, E. M. V. (2003). Effect of pH on the
gelation properties of skim milk gels made from plant coagulants and chy-
mosin. Journal of Dairy Science, 86(8), 2558e2567.
Fernandes, P. B., Gonçalves, M. P., & Doublier, J. L. (1991). A rheological character-
ization of kappa-carrageenan/galactomannan mixed gels: A comparison of lo-
cust bean gum samples. Carbohydrate Polymers, 16(3), 253e274.
Fox, P. F., Singh, T. K., & McSweeney, P. L. H. (1994). Proteolysis in cheese during
ripening. Special Publication-Royal Society of Chemistry, 150, 1e4.
Green, M. L., & Grandison, A. S. (1993). Secondary (non-enzymatic) phase of rennet
coagulation and post-coagulation phenomena. In P. F. Fox (Ed.), Cheese: Chem-
istry, physics and microbiology: Volume 1: General aspects (2nd ed., pp. 101e140).
London: Springer ScienceþBusiness Media.
Groot, R. D., Bot, A., & Agterof, W. G. (1996). Molecular theory of strain hardening of
a polymer gel: Application to gelatin. The Journal of Chemical Physics, 104(22),
9202e9219.
Guimer, A., Fournet, G., & Walker, C. (1955). Small angle scattering of X-rays. New
York: J. Wiley & Sons.
Hayter, J., & Penfold, J. (1983). Determination of micelle structure and charge by
neutron small-angle scattering. Colloid and Polymer Science, 261(12),
1022e1030.
Holt, C. (2016). Casein and casein micelle structures, functions and diversity in 20
species. International Dairy Journal, 60, 2e13.
Holt, C., De Kruif, C. G., Tuinier, R., & Timmins, P. (2003). Substructure of bovine
casein micelles by small-angle X-ray and neutron scattering. Colloids and Sur-
faces A: Physicochemical and Engineering Aspects, 213(2), 275e284.
Ingham, B., Erlangga, G. D., Smialowska, A., Kirby, N. M., Wang, C., Matia-Merino, L.,
et al. (2015). Solving the mystery of the internal structure of casein micelles.
Soft Matter, 11(14), 2723e2725.
Jacob, M., Jaros, D., & Rohm, H. (2011). Recent advances in milk clotting enzymes.
Society of Dairy Technology, 64, 14e33.
Laemmli, U. K. (1970). Cleavage of structural proteins during assembly of the head
of bacteriophage T4. Nature, 277, 680e685.
Lapasin, R., & Pricl, S. (1995). Rheology of polysaccharide systems. In R. Lapasin, &
S. Pricl (Eds.), Rheology of industrial Polysaccharides: Theory and applications (1st
ed., pp. 162e249). London: Springer ScienceþBusiness Media.
Liu, D., Yu, Y., Zhang, J., Liu, X., Wang, M., Hemar, Y., et al. (2017). Biochemical and
physico-chemical changes of skim milk during acidification with glucono-d-
lactone and hydrogen chloride. Food Hydrocolloids, 66, 99e109.
Lo Piero, A. R., Puglisi, I., & Petrone, G. (2002). Characterization of “lettucine”, a
serine-like protease from Lactuca sativa leaves, as a novel enzyme for milk
clotting. Journal of Agricultural and Food Chemistry, 50, 2439e2443.
Lo Piero, A. R., Puglisi, I., & Petrone, G. (2011). Characterization of the purified
actinidin as a plant coagulant of bovine milk. European Food Research and
Technology, 233(3), 517.
Macedo, I. M. Q., Faro, C. J. F., & Pires, E. M. V. (1993). Specificity and kinetics of the
milk-clotting enzyme from cardoon (Cynara cardunculus L.) toward bovine k-
casein. Journal of Agricultural and Food Chemistry, 41, 1537e1540.
Marchin, S., Putaux, J.-L., Pignon, F., & Le onil, J. (2007). Effects of the environmental
factors on the casein micelle structure studied by cryo transmission electron
microscopy and small-angle x-ray scattering/ultrasmall-angle x-ray scattering.
The Journal of Chemical Physics, 126(4), 45101.
Martin, J. E., & Hurd, A. (1987). Scattering from fractals. Journal of Applied Crystal-
lography, 20(2), 61e78.
Mata, J. P., Udabage, P., & Gilbert, E. P. (2011). Structure of casein micelles in milk
protein concentrate powders viasmall angle X-ray scattering. Soft Matter, 7(8),
178 Z. Li et al. / Food Hydrocolloids 79 (2018) 170e178
3837e3843.
Mazorra-Manzano, M. A., Perea-Gutie rrez, T. C., Lugo-Sanchez, M. E., Ramirez-
Suarez, J. C., Torres-Llanez, M. J., Gonza lez-Co
rdova, A. F., et al. (2013). Com-
parison of the milk-clotting properties of three plant extracts. Food Chemistry,
141(3), 1902e1907.
McDowall, M. A. (1970). Anionic proteinase from Actinida chinensis. Preparation and
properties of the crystalline enzyme. European Journal of Biochemistry, 14,
214e221.
McMahon, D. J., Yousif, B. H., & Kal
ab, M. (1993). Effect of whey protein denaturation
on structure of casein micelles and their rennetability after ultra-high tem-
perature processing of milk with or without ultrafiltration. International Dairy
Journal, 3(3), 239e256.
Moitzi, C., Menzel, A., Schurtenberger, P., & Stradner, A. (2010). The pH induced sol
gel transition in skim milk revisited. A detailed study using time-resolved light
and X-ray scattering experiments. Langmuir, 27(6), 2195e2203.
Nakagawa, K., & Kagemoto, M. (2013). Characterization of casein-based nano-
particles formed upon freezing by in situ SAXS measurement. Colloids and
Surfaces B: Biointerfaces, 103, 366e374.
Nielsen, S. S., Toft, K. N., Snakenborg, D., Jeppesen, M. G., Jacobsen, J. K.,
Vestergaard, B., et al. (2009). BioXTAS RAW, a software program for high-
throughput automated small-angle X-ray scattering data reduction and pre-
liminary analysis. Journal of Applied Crystallography, 42(5), 959e964.
Omotosho, O. E., Oboh, G., & Iweala, E. E. J. (2011). Comparative effects of local
coagulants on the nutritive value, in vitro multienzyme protein digestibility and
sensory properties of wara cheese. International Journal of Dairy Science, 6,
58e65.
Ong, L., Dagastine, R. R., Kentish, S. E., & Gras, S. L. (2013). Microstructure and
composition of full fat Cheddar cheese made with ultrafiltered milk retentate.
Foods, 2(3), 310e331.
Pignon, F., Belina, G., Narayanan, T., Paubel, X., Magnin, A., & Ge san-Guiziou, G.
(2004). Structure and rheological behavior of casein micelle suspensions during
ultrafiltration process. The Journal of Chemical Physics, 121(16), 8138e8146.
Pires, E., Faro, C., Macedo, I., Esteves, C., Morgado, J., Veríssimo, P., et al. (1994). Flor
do cardo versus quimosina no fabrico de queijos artesanais. Revista da Sociedade
Portuguesa de Química, 54, 66e68.
Rehm, C., Brûle , A., Freund, A. K., & Kennedy, S. J. (2013). Kookaburra: The ultra-
small-angle neutron scattering instrument at OPAL. Journal of Applied Crystal-
lography, 46(6), 1699e1704.
Roseiro, L. B., Barbosa, M., Ames, J. M., & Wilbey, R. A. (2003). Cheesemaking with
vegetable coagulantsdthe use of Cynara L. for the production of ovine cheeses.
International Journal of Dairy Technology, 56, 76e85.
Schaefer, D. W., Martin, J. E., Wiltzius, P., & Cannell, D. S. (1984). Fractal geometry of
colloidal aggregates. Physical Review Letters, 52(26), 2371.
Shah, M. A., Mir, S. A., & Paray, M. A. (2014). Plant proteases as milk-clotting en-
zymes in cheesemaking: A review. Dairy Science & Technology, 94, 5e16.
Timotijevi c, G. S., Radovi
c, S. R., & Maksimovi c, V. R. (2006). Aspartic proteinases
from buckweath (Fagopyrum esculetum Moench) seeds e purification and
properties of the 47 kDa enzyme. Archives of Biological Sciences, 58, 171e177.
Trujillo, A. J., Carretero, C., & Guamis, B. (1994). Los coagulantes en la industria
quesera. Alimentacion, Equipos Y Tecnologia, 13, 91e97.
Van Heijkamp, L. F., de Schepper, I. M., Strobl, M., Tromp, R. H., et al. (2010). Milk
gelation studied with small angle neutron scattering techniques and Monte
Carlo simulations. The Journal of Physical Chemistry A, 114(7), 2412e2426.
Vishwanatha, K. S., Rao, A. A., & Singh, S. A. (2010). Production and characterization
of a milk-clotting enzyme from Aspergillus oryzae MTCC 5341. Applied Micro-
biology and Biotechnology, 85, 1849e1859.
Waungana, A., Singh, H., & Bennett, R. J. (1996). Influence of denaturation and ag-
gregation of b-lactoglobulin on rennet coagulation properties of skim milk and
ultrafiltered milk. Food Research International, 29(8), 715e721.
Waungana, A., Singh, H., & Bennett, R. J. (1998). Rennet coagulation properties of
skim milk concentrated by ultrafiltration: Effects of heat treatment and pH
adjustment. Food Research International, 31(9), 645e651.
Windsor, C. G. (1988). An introduction to small-angle neutron scattering. Journal of
Applied Crystallography, 21(6), 582e588.