Food Hydrocolloids 96 (2019) 246–258

Contents lists available at ScienceDirect

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

Gel properties of potato protein and the isolated fractions of patatins and T
protease inhibitors – Impact of drying method, protein concentration, pH
and ionic strength
Jesper Malling Schmidta, Henriette Damgaardb, Mathias Greve-Poulsenc, Anne Vuholm Sundsa,
Lotte Bach Larsena, Marianne Hammershøja,∗
a
Department of Food Science, Aarhus University, Blichers Allé 20, DK-8830, Tjele, Denmark
b
AKV Langholt, Gravsholtvej 92, 9310, Vodskov, Denmark
c
KMC, Herningvej 60, 7330, Brande, Denmark

ARTICLE INFO ABSTRACT

Keywords: The use of plant protein in food products is increasing. Potato protein gelation is scarcely documented, therefore
Potato protein potato protein gel properties were tested for two whole protein isolates (freeze dried or spray dried) and the
Patatin purified sub-fractions, protease inhibitors and patatins. The effects of protein concentrations, 8 or 15% (w/w),
Protease inhibitor pH-range of 3.0–7.5, and 2 different levels of ionic strength; low (15 mM NaCl) or high (200 mM NaCl) on heat
Gelation
induced gel properties was studied by oscillatory rheology and by uniaxial compression. Spray drying did not
Rheology
impair gel properties of the whole protein isolates, with the spray dried isolate actually showing higher gel
strength at most conditions.
For whole protein isolates, gel strength expressed as the storage modulus (G′) had a minimum at pH 4.7 at
both high and low salt. The isolated patatin fraction showed a similar trend at low salt, but at high salt the gel
strength was very low at pH 3, with progressively higher values at pH 5 and pH 7.5. The protease inhibitor
fraction gelled best at low salt and pH values below pH 3.3, while higher pH resulted in very weak gels with a 10-
fold decrease in G’. Gel appearance was affected by protein sample, pH, and ionic strength. Both whole protein
isolates and the protease inhibitor fraction formed translucent gels at pH 3.0 and ionic strength 15 mM, while
patatin formed translucent gels at both pH 3.0 and pH 7.5 at low salt concentration. All other conditions resulted
in opaque gels.

1. Introduction patterns (Baier & Knorr, 2015; Barta, Bartova, Zdrahal, & Sedo, 2012).
The second group accounting for ∼50% of total protein fraction con-
The world population is rapidly growing, which puts a challenge on tains the protease inhibitors (PI) (Pouvreau et al., 2001). The molecular
food security with increasing demand for food protein. This has in re- weights of PIs range from 4.3 to 20.6 kDa, with pI values of pH 5.1–9.0
cent years lead to exploration of new, alternative and sustainable pro- (Pouvreau et al., 2001). The third group is mainly composed of oxi-
tein sources from e.g. plants, algae and insects. One such source may be dative and other enzymes, like polyphenol oxidase, lipoxygenase and
potato proteins, which are found in a liquid sub-fraction named potato enzymes associated with starch synthesis (Jorgensen, Stensballe, &
fruit juice (PFJ) from potato starch production. Typically, PFJ contains Welinder, 2011).
2–5% solid matter, of which 35% is N-containing substances, protein, Like other proteins, the potato proteins possess, besides nutritional
peptides and amino acids, corresponding to 0.7–1,75% (Knorr, Kohler, quality, also promising food functional properties in their ability to
& Betschart, 1977). form structural networks such as gels, foams and emulsions (Holm &
The proteins in PFJ can be classified into three groups. The first Eriksen, 1980; Ralet & Gueguen, 2000; Schmidt, Damgaard, Greve-
group accounting for up to 40% of the total potato protein is the pa- Poulsen, Larsen, & Hammershoj, 2018). In food gels, the proteins are
tatins. These are 39–43 kDa glycoproteins (existing as native 80 kDa important for both gel building by binding water in a continuous pro-
dimers) with differing isolectric points (pI, 4.5–5.2) and glycosylation tein network as well as for textural properties of the final set gel, which

∗
Corresponding author.
E-mail address: [email protected] (M. Hammershøj).

https://doi.org/10.1016/j.foodhyd.2019.05.022
Received 13 December 2018; Received in revised form 23 April 2019; Accepted 13 May 2019
Available online 14 May 2019
0268-005X/ © 2019 Elsevier Ltd. All rights reserved.
J.M. Schmidt, et al. Food Hydrocolloids 96 (2019) 246–258

is detrimental for the sensory perception of the food gel. result in synergistic effects relative to gel strength, as seen when mixing
Even though there has been increased interest in potato proteins in specific ratios of whey and rapeseed protein (Ainis et al., 2019), but
recent years, only a limited number of studies deal with aspects of purified proteins may also show enhanced gel strength compared to
gelling and pre-processing of the potato proteins. Lokra, Helland, mixtures depending on pH and mixing ratios (Arntfield, 1996).
Claussen, Straetkvern, and Egelandsdal (2008), found poor gelation In addition to protein composition, also the heat treatment pro-
properties of a spray dried whole protein isolate (outlet temperature cesses may affect the gel properties. Heating processes like pasteurisa-
165 °C), but a freeze dried isolate could form gels. tion and spray drying can affect heat labile proteins, such as patatin,
Gelation by patatin and protease inhibitor fractions was studied by whey proteins or ovalbumin, to unfold and aggregate and consequently
(Giuseppin, van der Sluis, & Laus, 2008), who found patatin to have a their ability to form gel networks, which may either promote or nega-
pH optimum for gelation at pH 4.8–5.5, and the protease inhibitors tively affect gel formation properties (Hammershoj, Peters, & Andersen,
have a lower optimum pH, at pH 3.2–4.3. 2004; Lechevalier et al., 2005; Mishra, Govindasamy-Lucey, & Lucey,
Patatin gelation properties at pH 7 at different ionic strength values 2005). Alternatives to heat processing technologies may be of interest
studied by Creusot, Wierenga, Laus, Giuseppin, and Gruppen (2011), to ensure the protein quality and functional properties, which for potato
revealed that patatin readily gels by heat treatment, exhibiting a rela- proteins has been reported in usage of high pressure treatment instead
tively low denaturation temperature (Td) of 59–60 °C, approximately of traditional heat treatment to ensure high solubility and improved
20 °C lower than other common food grade proteins, such as β-lacto- foam stability (Baier & Knorr, 2015). As potato protein typically will be
globulin, ovalbumin and glycinin. Furthermore, patatins appears to spray dried into a powder for food applications, the impact of spray
obtain gel structure at a much lower protein concentrations, i.e. 6% in drying on the potato protein for gelling and gel properties is very im-
water (pH 7), compared with these other protein representatives, where portant to obtain knowledge on.
concentrations of 8–11% protein is required for heat induced gel for- Our hypothesis is that potato protein gelation and the gel properties
mation. It was ascribed to the high level of exposed hydrophobic areas depend on protein pre-processing, protein denaturation, protein purity
in patatin, which at increased ionic strength may decrease its solubility as well as pH and ionic strength at gel formation. Hence, the aim of the
and increase the tendency of the protein to self-aggregate (Creusot present study was to evaluate the effect of drying method for whole
et al., 2011). Previously, we have reported potato protein to be able to potato protein and the further impact of physicochemical conditions;
gel at concentrations as low as 3%, although with a less firm gel texture pH and ionic strength, on protein solubility, protein gelation and final
(Schmidt, Larsen, & Hammershoj, 2017). Another recent study ex- protein gel properties by small- and large-deformation rheology and gel
amining the aggregation behaviour of ovalbumin, β-lactoglobulin and appearance.
patatin concludes that patatin has a higher tendency to form large and/ This was studied by two processing technologies for producing
or dense aggregates compared to the other proteins, but it was not di- dried, whole potato protein powder from PFJ; freeze drying = native
rectly correlated with the charge and exposed hydrophobicity proper- protein, and spray drying = denatured protein, in combination with
ties of these proteins (Delahaije, Wierenga, Giuseppin, & Gruppen, values of pH 3 (≠ pI, i.e. highly soluble patatin and PI), pH 5 (= pI, i.e.
2015). Furthermore, patatin forms di-and trimeric structures upon insoluble proteins, which may aggregate), and pH 7 (≠ pI of patatin,
heating, which dissociate into monomers in the presence of β-mercap- i.e. soluble; however near pI of some PI proteins, which may aggregate),
toethanol, indicating formation of disulphide bridges in the heat in- and low ionic strength (I: 15 mM NaCl) versus high ionic strength (I:
duced gels (Pots, ten Grotenhuis, Gruppen, Voragen, & de Kruif, 1999). 200 mM NaCl).
The impact of disulphide bridges in patatin gelation in relation to other
protein bonding and interaction mechanisms is, however, not presently
2. Materials and methods
clear.
Research in other food proteins illustrates that the conditions, e.g.
2.1. Potato protein samples
pH, ionic strength, concentration, temperature during gelation have
significant impact on both gelation process and final gel properties
Two different types of whole protein isolate powders produced by
(Hammershoj, 2001; Rickert, Johnson, & Murphy, 2004; Weijers, Sagis,
ultrafiltration of potato fruit juice (PFJ) were provided by KMC, Brande,
Veerman, Sperber, & Linden, 2002).
Denmark. One protein powder isolate was spray dried on an Anhydro
The final appearance of gels may also be affected by the conditions
PSD 55 pilot scale spray drier (SPX FLOW, Soeborg, Denmark) with an
used during gelation. Globular proteins tend to form particulate
inlet temperature of 200 °C and outlet temperature of 70 °C and the
(opaque) gels when pH = pI or at high ionic strength, and fine-stranded
other was prepared by freeze drying in a lab-scale freeze dryer
(translucent) gels, when pI < pH < pI or at low ionic strength
(Scanvac, Lynge, Denmark). The spray and freeze dried protein isolate
(Foegeding, 2006). As an example, egg albumen proteins cover a range
powders contained, on dry matter basis, 88.5 or 91.4 g/100 g protein
of pI from pH 4.1–6.6 for the proteins representing > 80% of the
(determined as described in section 2.3) on dry matter basis, respec-
quantity, and gels prepared at pH 5, i.e. near the pI are white and
tively. One additional protease inhibitor (PI) protein isolate fraction
opaque, which become more grey at neutral pH 7, and are completely
was provided by KMC. This fraction was spray dried and contained
transparent at highly alkaline conditions at pH 11, i.e. pH ≫ pI
86.9 g/100 g protein. A patatin fraction was purified in laboratory scale
(Hammershoj, 2001). Urbonaite et al. (2016) analysed whey protein
from spray dried potato protein isolate, as described previously
isolate gels at pH 7.2 at varying salt concentration from 0 to 300 mM
(Schmidt, Greve-Poulsen et al., 2017). The purified patatin fraction was
NaCl, and found significant difference opacity, going from translucent
subsequently freeze dried and had a protein content of 94.9 g/100 g.
to opaque, already when increasing salt from 0 to 50 mM. Doi (1993)
studied egg ovalbumin gels at different pH and salt conditions. At
10–20 mM Nacl turbid gels was found near pI of ovalbumin, but 2.2. Dry matter determination
transparent gels above and below pI. At 30–80 mM Nacl turbid gels was
also found at pH values above and below pI of ovalbumin. Dry matter (DM) of the protein powders was determined by drying
Different food applications may require either translucent or opaque ∼1 g sample in a HR73 halogen moisture analyzer (Mettler Toledo,
gels and it is therefore important to know which pH and salt conditions Schwerzenbach, Switzerland). Drying was conducted at 125 °C for up to
that determines the appearance. 2 min until a stable weight was achieved and recorded as DM (g/
Furthermore will both protein composition and purity of individual 100 g) = dry sample weight (g) *100/wet sample weight (g).
protein isolated play a role. Blending of different types of proteins may Determination was done in duplicates.

247
J.M. Schmidt, et al. Food Hydrocolloids 96 (2019) 246–258

2.3. Protein determination Table 1

pH values of potato protein (spray dried and freeze dried) and patatin in protein
The total protein content (g/100 g) of powders were determined by concentrations of 8% and 15% in respective buffers, n = 2.
Dumas combustion by a DUMATHERM® N PRO (Gerhardt, Used buffer Freeze dried Spray dried PI 8%/15% Patatin 8%/
Königswinter, Germany), using a nitrogen (N) conversion factor of 8%/15% 8%/15% solutions 15% solutions
protein = N x 6.25 for all samples. Determination was carried out in solutions solutions
duplicates.
pH 3 I: 15 mM 4.99/5.01 4.69/4.97 3.30/3.44 4.70/4.89
pH 3 I: 200 mM 4.85/4.97 4.67/5.16 3.44/3.53 4.90/4.90
2.4. Protein composition by SDS-PAGE pH 5 I: 15 mM 6.02/6.29 6.14/6.30 4.05/3.87 Not used
pH 5 I: 200 mM 5.93/6.30 6.02/6.22 4.06/3.91 Not used
pH 7 I: 15 mM 6.84/6.75 6.59/6.57 Not used 7.52/7.57
Protein dispersions with a concentration of ∼1 g/L was prepared in
pH 7 I: 200 mM 6.74/6.57 6.56/6.43 Not used 7.55/7.58
MIlli-Q water and used for SDS-PAGE analysis using Criterion TGX™
8–16% precast gels (Bio-Rad, Richmond, CA, USA). Reduced, (10 mM
dithioerythritol), samples were prepared as described earlier (Laemmli, Table 2
1970) and stained by Coomassie Brilliant blue G-250. A broad range Conductivity of 8% (w/w) potato protein dispersions measured at 20 °C. For the
molecular weight (Mw) marker was included for size estimation PI sample protein was dispersed in water with addition of NaCl to an ionic
(Thermo Scientific™ Spectra™ Multicolor Broad Range Protein Ladders). strength of 15 or 200 mM, and pH adjusted to 7.
Buffer pH, I: Freeze dried Spray dried PI Patatin
2.5. Relative protein solubility
pH 7, I: 15 mM 2.13 mS/cm 3.43 mS/cm 4.16 mS/cm 2.40 mS/cm
The powders were suspended in buffers to achieve a protein con- pH 7, 1: 200 mM 17.49 mS/cm 18.17 mS/cm 16.84 mS/cm 18.18 mS/cm
centration of 1 g/L corrected for dry matter content. In total, five buf-
fers representing pH values of pH 3, pH 4, pH 5, pH 6, and pH 7 were
tested at a temperature of 20 °C. Only small differences between sam-
used. These were 30 mM trisodium citrate dehydrate/citric acid (target
ples were observed (Table 2).
pH 3), 12 mM trisodium citrate dehydrate/citric acid (target pH 4),
22 mM sodium acetate/acetic acid (target pH 5), 21 mM Bis-Tris/HCL
(target pH 6) and 7.5 mM disoudium hydrogenphosphate/dihy- 2.7. Thermal analysis by differential scanning calorimetry (DSC)
drogenphosphate (target pH 7) all adjusted to an ionic strength of
15 mM or 200 mM with NaCl. The solutions were stirred 30 min, pH In order to determine the thermal denaturation profile of 15% (w/
was measured and readjusted if necessary, stirred for additional 30 min w) protein suspensions adjusted to pH 7.5 and I: 15 mM thermograms
and then subjected to centrifugation at 15300g for 10 min at 4 °C in a were run using a differential scanning calorimeter (Q1000 DSC, TA
table top centrifuge (Eppendorf 5417 R, Hamburg, Germany). The su- Instruments, New Castle, DE, USA). Samples of 10–15 mg protein dis-
pernatants were used for protein determination by the bicinchoninic persion were weighed into aluminium pans and sealed. Samples were
acid assay (BCA, Thermo Scientific™ Pierce™), where bovine serum heated at a rate of 5 °C/min from 20 to 90 °C. TA Universial Analysis
albumin (2 g/L) as reference protein was used for protein quantification software was used to analyse the thermograms and calculate dena-
(Smith et al., 1985). Two suspensions were prepared for each pH and turation onset (Tonset, °C), denaturation temperature (Td, °C) and en-
ionic strength condition and the BCA assay conducted in triplicates. The thalpy (J/g) of denaturation. Samples were analysed in duplicates.
relative protein content (%, w/v) was calculated based on the amount
of protein powder added initially to the buffer. 2.8. Dynamic rheological measurements

2.6. Preparation of protein suspension for gels To follow the gelation process and record the gel strength, suspen-
sions of 8% (w/w) were prepared as described in section 2.5above. A
The protein powders were suspended in given concentrations of 8 or sample volume of 20 mL was transferred to the rheometer (AR G2, TA
15% (w/w) in various buffers for 1 h at r.t. by stirring before use. The instruments, New Castle, DE, USA) fitted with a cup-and-bob geometry
used buffers were 30 mM trisodium citrate dehydrate/citric acid (target (cup radius 15 mm, bob radius 14 mm), which was connected to a
pH 3), 22 mM sodium acetate/acetic acid (target pH 5) and 7.5 mM water-bath providing temperature control during the thermal program.
disoudium hydrogenphosphate dehydrate/dihydrogenphosphate An oil-layer was pipetted on top of the solution to prevent evaporation
(target pH 7) all adjusted to an ionic strength of I: 15 mM or I: 200 mM during heating. Strain (0.01–99%) and frequency (0.01–100 Hz) sweeps
with NaCl. Protein suspensions were used for gelation experiments the were performed to determine the linear viscoelastic region. The vis-
same day. coelastic properties as storage (G′) and loss (G″) moduli during heating
The pH of protein solutions was measured by a PHM 92 pH-meter and cooling were measured by the rheometer using oscillatory mode at
(Radiometer, Copenhagen, Denmark) and the measured pH summar- a frequency of 0.5 Hz and 0.2% strain. The thermal program started
ized in Table 1. Differences were observed between the buffer pH and with 5 min equilibration at 20 °C, heating to 85 °C (1 °C/min), holding
the actual pH measured, therefore additional samples were prepared at 85 °C for 20 min, cooling from 85 °C to 20 °C (1 °C/min), holding at
with protein suspended in Milli-Q water with I: 15 or I: 200 mM NaCl 20 °C for 20 min and finally a strain sweep from 0.1 to 99%. Each
and adjusted to pH 3.0, pH 5.0 and pH 7.0 with 1 M HCL or 1 M NaOH. sample was tested at least in triplicates, however, patatin was tested in
In the following sections, the pH values referred to are the measured duplicate.
values (displayed with one decimal). Cold-gelation at 20 °C were tested for suspensions of 15% (w/w)
Conductivity was measured by a CDM210 Conductivity Meter spray dried protein, pH 3.0 at I: 15 and 200 mM. Gelation was followed
(Radiometer, Copenhagen, Denmark) for all samples at the highest pH after 5 min equilibration at 20 °C and further measurement for 6 h using

248
J.M. Schmidt, et al. Food Hydrocolloids 96 (2019) 246–258

oscillatory mode at a frequency of 0.5 Hz and deformation of 0.2%

strain. Measurements were conducted in duplicates.

2.9. Preparation of gels for uniaxial compression

Suspensions of 15% (w/w) protein were prepared as described

previously. The suspension was transferred to a 10 mL plastic syringe
(BD Plastipak, REF 302188) with an internal diameter of 14 mm. The
inside of the syringe was coated with a thin layer of oil before transfer
of the protein solution. The syringe was heated in a water-bath at 85 °C
for 20 min, then cooled in water and stored at 5 °C overnight.

2.10. Gel texture analysis by uniaxial compression

The gels prepared (section 2.9) were equilibrated for 4 h to ∼20 °C

before careful removal of the gel-cylinder by application of air pressure
to one end of the syringe tube. The gel cylinders were cut with an oiled
knife to a height of 15 mm. Gel texture was analysed by uniaxial
compression test until fracture of the gel on a FTC TMS-Touch texture
analyzer (Food Technology Corporation, Sterling, USA) with a 25 N
load cell, 75 mm diameter flat stainless steel plate and a compression
speed of 48 mm/min. Recordings of force (N) and displacement (m)
were used in calculation of true axial stress (equation (1)) and Hencky Fig. 1. SDS-PAGE gel of potato protein samples at a concentration of ∼1 g/L.
strain (equation (2)) where F = force (N), A = initial end area (m2), Lane 1, molecular weight (Mw) marker with masses indicated in kDa; Lane 2,
Hi = initial gel height (m), and H = height (m) at fracture. Each sample spray dried protein; Lane 3, freeze dried protein; Lane 4, Patatin fraction; Lane
was tested at least in triplicates. An example of a compression test force- 5, Protease inhibitor (PI) fraction.
displacement curve is enclosed in Supplementary Fig. 1.
3.2. Potato protein solubility
F H
= [Pa]
A Hi (1)
The solubilities of the whole potato protein isolates were tested at
pH 3, pH 4, pH 5, pH 6, and pH 7 and at ionic strengths of I: 15 or I:
H
= ln [ ] 200 mM (Fig. 2). The freeze dried (Fig. 2A) and spray dried (Fig. 2B)
Hi (2) whole protein isolates showed similar solubility profiles at low ionic
strength, with high solubilities at pH 3 and pH 7 and a minimum at pH
5. Increase in ionic strength from 15 mM to 200 mM at pH 5 resulted in
2.11. Gel colour an increase in solubility from ∼60 to ∼77% for both powders, but a
decrease for spray dried powder of 18% point at pH 3 and 5% at pH 4.
The cut gel cylinders (section 2.10) were used for colour measure- Overall, the highest solubility for both powders was obtained at pH 7
ment with a Minolta Chroma Meter CR-400 (Konica Minolta, Osaka, and 200 mM NaCl, where > 90% solubility was observed. The PI pro-
Japan) using the Hunter Lab scale calibrated against standardized teins had high solubility (83–86%) at high ionic strength with very
daylight (D65). The L-, a- and b-values reflect lightness (0: black; 100: limited impact of pH (Fig. 2C). At low ionic strength, PI proteins dis-
white), redness (−100: green; 100: red) and yellowness (−100: blue; played a solubility minimum at pH 5. The patatin protein had very high
100: yellow), respectively. Measurements were conducted in triplicates. solubility at pH 6 and 7 at both ionic strength conditions (94.8–100%).
Minimum solubility was observed at pH 5 at low ionic strength. By
increasing the ionic strength to 200 mM the solubility of patatin at pH 5
2.12. Statistical analysis
was improved by 18.6% point. Under acidic conditions, pH 3 and pH 4,
the low ionic strength resulted in the highest solubility of patatin.
The results are presented as means with standard deviations.
Significant differences between treatments were determined by one-
way or two-way ANOVA analysis with parameters pH, ion strength and
3.3. Thermal unfolding of proteins
interactions hereof using GraphPad Prism 6 (GraphPad Software Inc, La
Jolla, USA, version 6.01). Differences were regarded to be significant at
Differential scanning calorimetry analysis was used to map the un-
minimum 95 %-level (P < 0.05).
folding properties between the purified PI and patatin fractions and
further to assess the impact of drying method for the two potato whole
3. Results protein isolates. Based on the DSC thermogram with displayed values of
Tonset, Td and enthalpy (Fig. 3), patatin had a significantly lower Td of
3.1. Protein composition ∼9 °C than the PI fraction and the drying method had only minor, but
significant impact on Tonset and Td of the two whole protein isolates.
The protein compositions of the four tested samples by SDS-PAGE The thermograms of the two whole protein isolates displayed less dis-
are visualised in Fig. 1. Both the spray dried (lane 2) and the freeze crete and broader peaks that the isolated fractions (Fig. 3), which
dried sample (lane 3) are whole protein isolates consisting of all the corresponds to the various proteins contained in the protein isolates.
proteins present in PFJ. The patatin fraction (lane 4) has a high content Furthermore, the endothermic enthalpy was highest for the PI fraction,
of patatin (molecular weight ∼40 kDa) with the presence of some PI indicating a higher degree of folding structure affected by thermal de-
proteins. The PI fraction (lane 5) is consisting of PI proteins within naturation. No significant difference was found in enthalpy between
molecular weights of 10–25 kDa. drying methods.

249
J.M. Schmidt, et al. Food Hydrocolloids 96 (2019) 246–258

Fig. 2. Relative protein solubility as func-

tion of pH at two ionic strength of I: 15 mM
(white bars) or 200 mM (dark bars), re-
spectively, of A) Freeze-dried total potato
protein isolate; B) Spray-dried total potato
protein isolate; C) protease inhibitor (PI)
fraction; D) patatin fraction, n = 6.
a-h
bars in the same panel with different
letter superscripts are significantly different
(P < 0.05).

3.4. Gelation profile by dynamic rheological measurements In the following, the results of heating profiles on G′ are shown in
Fig. 4, the G′ values after end of heating and holding at 85 °C are shown
Small deformation dynamic rheology was applied to monitor the in Table 3, the cold setting effect on G′ and strain at 20 °C is given in
heat-induced gel point during the 20–85 °C temperature ramp of the 8% Fig. 5, and hereof the relative increase in G′ from heat-set gel to cold-set
(w/w) dispersions and to map the value of G′ after 20 min of holding gel is given in Table 3 as the ratio, G′ 20 °C/G’ 85 °C.
time at 85 °C (G′85°C).

Fig. 3. DSC thermograms of 15% (w/w) protein dispersions of the four potato protein samples at pH 7.5 at ionic strength I: 15 mM NaCl. Black arrows indicate Tonset,
grey arrows indicate Td, the dashed line indicate the area under the curve and the calculated enthalpy, ΔH.

250
J.M. Schmidt, et al. Food Hydrocolloids 96 (2019) 246–258

Fig. 4. Gelation profiles storage modulus G’ (Pa) measured by oscillatory rheometry during heating from 20 to 85 °C at various pH-values of 8% (w/w) freeze dried
potato protein isolate at ionic strength A) 15 mM; B) 200 mM; spray dried potato protein isolate at ionic strength C) 15 mM; D) 200 mM; protease inhibitor (PI)
fraction at ionic strength E) 15 mM; F) 200 mM; and patatin fraction at ionic strength G) 15 mM; H) 200 mM.

251
J.M. Schmidt, et al. Food Hydrocolloids 96 (2019) 246–258

Table 3
Gelation profile data for 8% (w/w) potato protein dispersions recorded by oscillatory rheology with values of storage modulus (G′) at 85 °C after 20 min holding time
(G′85°C), a calculated ratio between the storage modulus G′20°C after cooling to 20 °C compared to G′85°C, i.e. cold-setting effect, gelation temperature (˚C), and the
observed initial structure of protein dispersions.
Sample pH and ionic strength I: (mM) G′85˚C(Pa) G′20 ˚C/G′85 ˚C(-) Gelation temp(˚C) Initial structure

b a
Freeze dried 3.0 I: 15 678 ± 95 3.4 56.2 ± 1.9 liquid
3.0 I: 200 839 ± 71 a 3.2 24.5 ± 3.3 d gel
5.0 I: 15 111 ± 4f 10.6 40.6 ± 0.6 c viscous
4.9 I: 200 214 ± 26 f 10.3 27.0 ± 3.9 d liquid
6.0 I: 15 186 ± 7 ef 11.2 25.7 ± 0.9 d liquid
5.9 I: 200 287 ± 5 de 10.5 42.9 ± 0.9 c liquid
6.8 I: 15 536 ± 35 c 9.5 43.6 ± 1.2 bc liquid
6.7 I: 200 313 ± 7d 10.2 48.7 ± 0.5 b liquid

Spray dried 3.0 I: 15 820 ± 72 b 2.9 53.9 ± 0.3 a liquid

a
3.0 I: 200 1501 ± 103 3.1 21.0 ± 0.5 f gel
4.7 I: 15 144 ± 6 f 10.7 39.4 ± 0.4 c viscous
4.7 I: 200 279 ± 13 e 10.8 30.0 ± 0.8 e liquid
6.1 I: 15 293 ± 5 d e 10.6 35.2 ± 1.4 d liquid
6.0 I: 200 466 ± 16 d 10.0 41.1 ± 0.3 c liquid
6.6 I: 15 610 ± 23 c 10.0 40.8 ± 1.4 c liquid
6.6 I: 200 522 ± 11 cd 10.0 47.7 ± 0.3 b liquid

PI 3.0 I: 15 1198 ± 19 a 3.9 60.9 ± 0.1 d liquid

3.0 I: 200 265 ± 35 c 5.9 55.3 ± 0.3 e liquid
3.3 I: 15 384 ± 24 b 8.8 60.1 ± 0.1 d liquid
3.4 I: 200 197 ± 17 c 4.7 59.4 ± 0.7 d liquid
4.1 I: 15 77.7 ± 17.8 d 10.5 64.6 ± 0.7 ab liquid
4.1 I: 200 66.0 ± 1.0 d 10.6 61.6 ± 0.8 cd liquid
5.0 I: 15 4.4 ± 0.3 d 14.4 66.5 ± 0.9 a liquid
5.0 I: 200 22.6 ± 1.2 d 14.8 64.2 ± 0.1 b liquid
7.0 I: 15 18.4 ± 0.9 d 13.0 63.8 ± 0.2 bc liquid
7.0 I: 200 24.9 ± 0.4 d 13.2 63.6 ± 1.2 bc liquid

Patatin 3.0 I: 15 1337 ± 35 a 1.9 47.5 ± 1.5 c liquid

3.0 I: 200 488.5 ± 31 bc 1.0 30.2 ± 0.3 d viscous
4.7 I: 15 176 ± 42 c 4.7 49.5 ± 0.4 c viscous
4.9 I: 200 679 ± 46 b 6.2 50.0 ± 0.0 c viscous
7.5 I: 15 1076 ± 145 a 3.5 61.6 ± 0.6 a liquid
7.5 I: 200 1244 ± 58 a 6.3 55.5 ± 0.1 b liquid

a,b,d
For each sample different letters in the same column indicate significant different values at P < 0.05.

Fig. 5. Storage modulus of heat induced and cold-set gels at 20 °C (G′20°C) measured at frequency = 0.5 Hz and strain = 0.2% (A–D); and from strain sweep at
0.1–99% strain, the strain (%) value causing a 10% reduction of the G′ after cold setting at 20 °C (E–H) of 8% (w/w) protein gels at different pH values at ionic
strength I: 15 or I: 200 mM. A, E) freeze dried potato protein; B, F) spray dried potato protein; C, G) protease inhibitor fraction; D, H) patatin fraction, n = 3.

252
J.M. Schmidt, et al. Food Hydrocolloids 96 (2019) 246–258

Liquid protein dispersions have low G′ values initially until a certain withstand much deformation before structural breakdown occurs, i.e. a
temperature, where a rapid increase in G′ is observed i.e. the gel point brittle structure, while a high strain value indicates an elastic structure.
(Frydenberg, Hammershoj, Andersen, Greve, & Wiking, 2016). Some of Due to small differences in the measured pH of the various samples
the tested samples did, however, show behaviour of a viscoelastic gel (Table 1), a direct comparison between samples must take this into
(G’ > G″) already at 20 °C. Upon heating G′ dropped, and then increase account. The storage modulus (G′20°C) of freeze- and spray dried whole
again, corresponding to a drop in the phase angle. Other samples dis- potato protein isolate was affected by pH and ionic strength changes
played a gel-like rheological response in the whole temperature regime. (Fig. 5A and B). At an ionic strength of 15 mM, the G′ had a minimum
For samples marked “viscous” in Table 3, the gel point is determined value at pH 4.7–5.0 with higher values at pH 3, and increasingly higher
based on the drop in phase angle. An example of a gelation profile of a values at pH 6 and above, with the highest pH resulting in the strongest
viscous sample with display of G’ and phase angle is enclosed in gel. At I: 200 mM, G′ was lowest at pH 4.7–4.9 and less effect was seen
Supplementary Fig. 2. upon pH increase, especially for the freeze dried sample. For pH 3.0, the
Results for freeze dried protein isolate at an ionic strength of 15 and spray dried sample had a markedly higher G′ than the freeze dried
200 mM are presented in Fig. 4A and B, respectively. At an ionic sample and at pH 3.0 I: 200 mM markedly higher G′ that the patatin and
strength of 15 mM, the pH 5.0 sample had a markedly different profile PI fraction. Generally, an increase in ionic strength did result in higher
showing viscous behaviour and an initial 10-fold higher G′ at 20 °C G′, except at the highest pH (Fig. 5A and B).
compared to the other tested pH-values (Fig. 4A). An increase in G′ at The PI fraction was highly affected by pH, with pH values below pH
pH 3.0, pH 6.02 and pH 6.84 was seen at ∼56 °C, ∼26 °C and ∼44 °C, 3.5 resulting in the strongest gels, especially at low ionic strength. Very
respectively (Table 3). The final G′-values at 85 °C was lowest at pH 5.0 weak gels were produced at both ionic strength conditions at pH of 5 to
with a value of 111 Pa and highest at pH 3.0 with a value of 678 Pa. pH 7 (Fig. 5C).
Increase in ionic strength from 15 mM to 200 mM resulted in markedly Ionic strength and pH had a different impact on purified patatin
different profiles for the pH 3.0 and pH 4.85 samples, for pH 3.0 it compared to the whole protein isolate, with a marked decrease in G′ at
resulted in initial gelation, while at pH 4.85 it resulted in an initial high ionic strength at pH 3.0, while high ionic strength resulted in an
liquid structure. At an ionic strength of 200 mM the pH 3.0 sample had increased gel strength at pH 4.9 and above (Fig. 5D). The highest G′ of
a substantial higher G′ value of ∼839 Pa than the other samples. all samples was recorded for patatin at pH 7.5 and I: 200 mM. For all
The gelation profiles for the spray dried samples (Fig. 4C and D) samples except the freeze dried total protein isolate, G’ was sig-
were comparable to the freeze dried samples (Fig. 4A and B), except at nificantly affected by pH, ionic strength and the interactions hereof
pH ∼6 at I: 15 mM, which had an initial increase in G′ at 35.2 °C, i.e. (P < 0.05), For the freeze dried isolate only pH and the interaction of
9.5 °C, higher than the freeze dried sample. Also, the spray dried potato factors was significant.
protein produced a G′85°C at pH 3.0 and I: 200 mM, which was almost For both drying methods of whole protein isolates, pH 3.0 at I:
twice as high, 1501 Pa, as for the freeze dried potato protein, 839 Pa 200 mM resulted in the lowest strain values and pH 3.0 at I: 15 mM the
(Table 3). highest (Fig. 5E and F), i.e. more brittle gels at high salt. A trend was
When looking into the G′20 ˚C/G′85 ˚C both whole protein isolates observed at the other pH values, with high ionic strength resulting in
depicts similar trends with low values of ∼3 at pH 3.0 at both ionic the largest strain values, and thereby increasing elasticity of the gels.
strength conditions, and higher values of ∼10 at all other conditions. The PI fraction had a high strain value at pH 3.0 at I: 15 mM, but a
The gelation profiles of the PI samples at I: 15 mM all behaved as very low value when the ionic strength was increased to 200 mM
liquids, until a temperature of 60–65 °C was reached (Fig. 4E). The (Fig. 5G). A similar trend was observed for both ionic strength condi-
G′85°C was highly affected by pH, where a pH of 3.0 resulted in a value tions at all pH values above pH 3.3, with increasing pH resulting in
of 1198 Pa, while at pH 5.0 the value decreased by > 200-fold to larger strain values. For patatin, markedly higher strain values were
4.4 Pa (Table 3). High ionic strength resulted in a ∼6 °C lower gelation observed at pH ∼7.5 at both ionic strength conditions (Fig. 5H). At pH
temperature at pH 3.0, and a significantly (P < 0.05) lower G′85°C 3.0 and pH ∼4.8, the patatin sample showed similar properties as the
value of 265 Pa, i.e. a 5-fold reduction compared to PI gelation at low whole protein isolate samples, but an even higher strain at pH 3.0 at I:
ionic strength. The G′20 ˚C/G′85 ˚C ratio had high values above 10 at pH 15 mM was found.
values above 4 at both ionic strength conditions. A two-fold increase in For all samples strain was significantly affected by pH, ionic
the ratio was found when going from pH 3.0 to pH 3.3 I: 15 mM. strength and the interactions heroff (P < 0.05).
The heating profile of patatin resembled that of the total potato In summary, whole protein isolate produced the strongest gels at pH
protein isolates at I: 15 mM, i.e. showing viscous behaviour at pH 4.7 values below or above pH 5 and the most elastic gels at pH 3.0 at low
and liquid behaviour above and below this pH (Fig. 4G). At high ionic ionic strength. The PI protein required low pH and low ionic strength to
strength, both the sample at pH 3.0 and pH 4.9 showed viscous beha- form strong and elastic gels. PI protein can produce very weak, but
viour at 20 °C, while the sample at pH ∼7.5 had a very clear gelation elastic, gels at pH above 5.0 at both low and high ionic strength. Patatin
point at 55 °C. The G′85°C of the pH 3.0 at I: 200 mM sample and pH gels are hardest and most elastic at pH values above pH 7, and very
4.7 at I: 15 mM displayed relatively low values of 30 Pa and 176 Pa, brittle at pH 5. Strong and elastic gels can be made at pH 3.0 at low
respectively, while high values > 1000 Pa was obtained at pH 3.0 at I: ionic strength.
15 mM and pH 7.5 at I: 15 and 200 mM (Table 3). Patatin at pH 3.0 I:
200 mM yielded the lowest G′20 ˚C/G′85 ˚C ratio of all the tested samples 3.6. Cold gelation of spray dried protein isolate
with a value of 1, while pH 3.0 I: 15 mM had a value of 1.9.
During preparation of 15% (w/w) protein dispersions for uniaxial
3.5. Gel properties by dynamic rheological measurements compression studies, it was observed that certain conditions resulted in
gelation without any heating, i.e. gelation at 20 °C. For both spray dried
Following the heating ramp from 20 to 85 °C, samples were cooled and freeze dried protein isolate did gelation occur, when the dispersions
to 20 °C and the gel strength expressed as the storage modulus (G′20°C) were adjusted to pH 3.0 at both low and high ionic strength. Fig. 6
(Fig. 5A–D). A strain sweep was conducted from 0.1 to 99%, and the displays the gelation profile of a dispersion of spray dried potato whole
resulting loss in G′ was monitored. When strain is increased from 0.1 to protein isolate. The sample at high ionic strength formed a gel already
99%, the linear viscoelastic region (LWR) will eventually be surpassed. within the initial 5 min of equilibration before the measurement was
Hence, the strain causing a 10% reduction of the initial G′20°C was re- started, as indicated by the low value of phase angle. A firm gel was
corded and used to characterise the rigidity of the cooled gels formed within the 6 h measurement period. The sample at low ionic
(Fig. 5E–H). A low strain value indicates that the sample could not strength was initially in a liquid state as indicated by the phase angle

253
J.M. Schmidt, et al. Food Hydrocolloids 96 (2019) 246–258

Fig. 6. Storage modulus (G′) (circles) and phase angle (squares) during cold gelation at 20 °C of 15% (w/w) spray dried total potato protein dispersions at pH 3.0 at
ionic strength of I: 15 mM (black) and I: 200 mM (white).

value of ∼90°. After 38 min did the phase angle drop to 45° indicating a axial stress (σ) of 15% (w/w) freeze dried whole protein gels were
transition from liquid to gel. significantly affected by pH, ionic strength and the interaction hereof,
In the preparation of samples for large deformation textural analysis while for the spray dried only pH and interaction was significant
(section 3.7), the patatin dispersion at 15% (w/w) was also observed to (Fig. 7). At I: 15 mM, gel strength was lowest at pH close to 5, with
gel under cold conditions at a very high rate at pH 3.0 at I: 200 mM and increasing gel strength at pH 3.0 and pH values above 6 (Fig. 7A and B).
pH 4.7 at I: 15 mM, and at a slower rate at pH 3.0 at I: 15 mM and pH At I: 200 mM, the lowest stress was found at pH 3.0, with progressively
4.9 at I: 200 mM. higher values as pH approached more alkaline conditions. The PI
fraction produced the strongest gel at pH 3.0 at I: 15 mM. Weak gels
3.7. Gel textural properties by large deformation uniaxial compression were formed at the other pH and ionic strength conditions, with pH
5.0 at I: 15 producing a gel that was not self-supporting, and with
The 15% (w/w) gels were prepared for uniaxial compression ana- presence of a gravity induced precipitate (Fig. 7C and G). Stress was
lysis until fracture, to assess gel texture at large deformation. The true significantly affected by pH, Ionic strength and interaction hereof.

Fig. 7. Textural properties of 15% (w/w) potato protein isolate gels measured by uniaxial compression until fracture expressed as axial stress σ (A–D) and Hencky
strain ε (E–H). A, E) freeze dried potato protein isolate; B, F) spray dried potato protein isolate; C, G) PI fraction; D, H) patatin fraction.
*This sample point was not possible to measure due to very weak gel.

254
J.M. Schmidt, et al. Food Hydrocolloids 96 (2019) 246–258

Table 4
Colour of 15% (w/w) whole potato protein isolate, PI and patatin gels evaluated by lightness L (0 = black, 100 = white), redness a (−100 = green, 100 = red), and
yellowness b (−100 = blue, 100 = yellow).
Sample pH and ionic strength (mM) L a b Appearance

Freeze dried 3.0 I: 15 21.59 ± 0.04 g −0.57 ± 0.01 e

1.04 ± 0.02 f transparant
3.0 I: 200 61.97 ± 0.06 c −0.64 ± 0.03 f
7.19 ± 0.01 a opaque
5.0 I: 15 64.13 ± 0.03 a 0.64 ± 0.01 c 5.53 ± 0.01 b opaque
5.0 I: 200 63.92 ± 0.14 b 0.56 ± 0.01 d 5.20 ± 0.01 c opaque
6.3 I: 15 52.45 ± 0.23 e 2.70 ± 0.01 b 2.96 ± 0.03 d opaque
6.3 I: 200 54.68 ± 0.05 d 2.68 ± 0.03 b 2.46 ± 0.01 e opaque
6.8 I: 15 50.38 ± 0.18 f 2.94 ± 0.01 a 2.85 ± 0.04 d opaque
6.6 I: 200 52.65 ± 0.01 e 2.98 ± 0.02 a 2.56 ± 0.08 e opaque

Spray dried 3.0 I: 15 21.25 ± 0.02 g −0.57 ± 0.02 d 1.00 ± 0.01 h transparant
3.0 I: 200 64.47 ± 0.01 a −0.56 ± 0.03 d 6.99 ± 0.01 a opaque
5.0 I: 15 62.20 ± 0.13 c 1.27 ± 0.01 c 5.94 ± 0.01 b opaque
5.2 I: 200 63.57 ± 0.09 b 1.21 ± 0.02 c 5.54 ± 0.02 c opaque
6.3 I: 15 54.14 ± 0.05 e 4.28 ± 0.02 b 4.32 ± 0.01 d opaque
6.2 I: 200 55.29 ± 0.05 d 4.26 ± 0.03 b 3.75 ± 0.01 f opaque
6.6 I: 15 53.50 ± 0.02 f 4.47 ± 0.02 a 3.82 ± 0.01 e opaque
6.4 I: 200 54.27 ± 0.01 e 4.52 ± 0.01 a 3.60 ± 0.01 g opaque

PI 3.0 I: 15 53.05 ± 0.14 f −0.10 ± 0.05 f 12.22 ± 0.05 h transparant

3.0 I: 200 79.73 ± 0.24 de 1.54 ± 0.09 a 14.40 ± 0.01 d opaque
3.4 I: 15 79.33 ± 0.09 e 1.58 ± 0.01 a 14.91 ± 0.01 c opaque
3.5 I: 200 81.74 ± 0.11 b 1.27 ± 0.01 b 14.28 ± 0.03 e opaque
3.9 I: 15 81.35 ± 0.08 b 0.81 ± 0.01c 14.37 ± 0.03 de opaque
3.9 I: 200 82.26 ± 0.08 a 0.80 ± 0.01 c 14.14 ± 0.02 f opaque
5.0 I: 15 nd nd nd nd
5.0 I: 200 82.55 ± 0.09 a 0.68 ± 0.01 de 13.96 ± 0.05 g opaque
7.0 I: 15 80.53 ± 0.04 c 0.61 ± 0.02 e 16.01 ± 0.01 b opaque
7.0 I: 200 80.12 ± 0.07 cd 0.74 ± 0.03 cd 16.42 ± 0.02 a opaque

Patatin 3.0 I: 15 21.61 ± 0.14 d −1.04 ± 0.03 e

−0.93 ± 0.02 e transparant
3.0 I: 200 nd nd nd nd
4.8 I: 15 53.32 ± 0.07 b 0.51 ± 0.01 c 2.01 ± 0.01 b opaque
4.9 I: 200 60.98 ± 0.02 a 0.74 ± 0.01 b 2.72 ± 0.01 a opaque
7.6 I: 15 14.74 ± 0.02 e 0.11 ± 0.02 d −0.60 ± 0.02 d transparant
7.6 I: 200 46.71 ± 0.05 c 2.64 ± 0.01 a 1.60 ± 0.09 c opaque

a-h
For each sample different letters in the same column indicate significant different values at P < 0.05. nd not determined.

At an ionic strength of 15 mM patatin produced a weak gel at pH 3.8. Colour and visual appearance of gels
4.9, but very strong gels at both pH 3.0 and pH 7.5 (Fig. 7D). Gel
strength at pH 4.9 increased markedly at high ionic strength, while the The visual appearance of 15% (w/w) gels were highly affected by
sample at pH 7.5 decreased. The sample at pH 7.5 behaved opposite in pH, salt and sample type (Table 4). The freeze dried and the spray dried
respect to ionic strength when compared to the oscillatory measure- total isolates produced gels with a transparent appearance at pH 3.0 at
ments, where I: 200 produced the strongest gel (Fig. 5D). No sample I: 15 mM, while all other conditions resulted in opaque gels. Generally,
could be prepared at pH 3.0 I: at 200 mM due to cold gelation of the gels became darker, when pH increased (lower L), more reddish (higher
sample as described in section 3.6. a) and more yellowish (higher b) (Table 4). Changes in ionic strength
Hencky strain (ε) of 15% (w/w) freeze dried and spray dried whole mainly had an effect on the gel colour at pH 3.0.
protein isolate gels were significantly affected by pH, ionic strength and The PI protein produced transparent gels at pH at 3.0 I: 15 mM and
interaction hereof (Fig. 7E and F). Both samples showed a similar trend opaque at all other conditions. The pH had minor effect on lightness,
at I: 200 mM, with a low strain value at pH 3.0 and a progressively but the most white gels were found in the range of pH 3.5 to pH 5.
higher strain as pH increased. At I: 15 mM, the pH 3.0 gels did yield Increasing pH to 7 resulted in higher b-values, i.e. in more yellowish
higher strain values compared to high salt conditions. At low ionic gels. For patatin, at pH 3.0 and I: 15 mM and at pH 7.5 and I: 15 mM
strength, the freeze dried sample showed a minimum in strain at pH 5 transparent gels were observed, but the remaining conditions resulted
and a maximum at pH 6.3, while the spray dried sample showed pro- in opaque gels. The whitest gel was found at pH 4.9 at I: 200 mM. Since
gressively higher strain as pH was increased from 3.0 to pH 6.2. The PI the pH of the patatin fraction itself was higher than the freeze and the
fraction had similar strains at both ionic strength conditions in the pH spray dried samples, an additional gel of spray dried isolate was made
interval of pH 3.0 to pH 4.0, and at pH 7, with pH 7 yielding the highest at pH 7.5 at I: 15 to check if this pH could produce a transparent gel, but
values. At pH 5, no self-supporting gel could be produced at I: 15 mM, also at this high pH an opaque gel was found (results not shown).
but an increase in ionic strength to 200 mM resulted in a gel with an
intermediate strain value when compared to the samples at pH 3.9 and
pH 7 (Fig. 7G), strain was significantly affected by pH, ionic strength 4. Discussion
and interaction hereof. Patatin at I: 15 mM had a minimum in strain at
pH 4.9, with progressively higher values at pH 3.0 and especially at pH 4.1. Protein solubility, protein denaturation and gelation profiles
7.5. At pH 4.9, the strain increased when ionic strength was increased
from 15 mM to 200 mM. The highest strain was found at I: 200 mM at Spray drying of the whole potato protein powders did generally nor
pH 7.5 (Fig. 7H). result in inferior solubility compared to freeze drying except at pH 3 I:

255
J.M. Schmidt, et al. Food Hydrocolloids 96 (2019) 246–258

200 mM. It is proposed that spray drying resulted in a change in the Nieuwland, & de Jong, 2014). The results from whole protein isolates
secondary structure and/or tertiary structure, and that high salt con- shows that a different aggregation mechanism or rate is taking place at
ditions promoted protein-protein interactions and loss of solubility. The pH 3 (ratio of ∼3), compared to the other pH values tested (ratio ∼10),
drying method can, however, result in very different solubilities, as possible due to the partly pH induced denaturation of patatin at this low
reported previously (Lokra et al., 2008) with 10% solubility at pH 6, pH. The results from the isolated PI and patatin fractions are less con-
when spray drying outlet temperature of 165 °C was applied, compared sistent, and highly dependent on salt and pH conditions, but low values
to ∼70% solubility of a vacuum freeze dried protein fraction. In an- are generally seen at low pH.
other study, spray drying potato protein at an outlet temperature of When comparing G′85 ˚C and gelation temperature of the two whole
90 °C did not result in significantly different solubility, when compared protein isolates, some differences were found. The gelation temperature
to freeze drying (Claussen, Strommen, Egelandsdal, & Stretkvern, showed high similarity between the two samples at the different pH and
2007). The drying conditions had minor, although significant, effect on salt conditions except at pH 6 I: 15 mM, where the spray dried sample
the onset of denaturation but no significant effect on enthalpy of un- had a 10 °C higher gelation temperature. This may be attributed to the
folding, with the freeze dried sample having a higher onset temperature difference in pH of the two samples (Table 1), with the spray dried
and lower enthalpy. The study by Claussen et al. (2007) shows no sample having a 0.12 point higher pH value. The spray dried sample
significant difference in enthalpy, but onset of denaturation was sig- had higher G′85 ˚C at all pH and salt conditions, with a 1.8-fold higher
nificantly higher for the freeze dried protein. In contrast Lokra et al. value at pH 3 I: 200 mM. It is proposed that slight changes in protein
(2008) find decreasing enthalpy and increasing onset temperature, structure during drying resulted in increased protein-protein interac-
when comparing freeze drying to spray drying. It is assumed that the tions, and hence resulted in the higher values.
native structure of a freeze dried protein would result in higher en- DSC denaturation temperature (Td) of patatin and PI were in
thalpy because more energy is required to unfold the structure com- agreement with the temperatures found by rheological measurements
pared to a spray dried protein, where the structure has already been of the gel point determination. The values for whole protein isolates
affected in some degree by the drying process. Anandharamakrishnan, were, however, much lower when determined by rheology. One would
Rielly, and Stapley (2007) found this to be true for a native whey assume that denaturation should precede aggregation, but other au-
protein isolate compared to spray dried isolated dried at progressively thors have reported similar behaviour i.e. an increase in G’ at a lower
higher outlet temperatures of 80–120 °C, which resulted in lower de- temperature than (Td) (Stading & Hermansson, 1990).
naturation enthalpies. However when comparing samples dried at 60 °C
to samples dried at 80 °C, the 80 °C samples had enthalpies either higher 4.2. Gel properties by small- and large-scale deformation and visual
or similar to 60 °C. Haque, Chen, Aldred, and Adhikari (2015) did not appearance of gels
find a significantly different enthalpy between a native whey protein
isolate and one spray dried at an outlet temperature of 80 °C. There was, As stated in the introduction globular proteins will tend to form
however, a difference in secondary structure with a decrease of β-sheet particulate (opaque) gels when pH = pI or at high ionic strength and
and an increase in random coil structure. A full explanation of our re- fine-stranded (translucent) gels, when pI < pH < pI or at low ionic.
sults would also benefit from knowledge about possible changes in Patatins have pI's of 4.5–5.2, while the protease inhibitors have pI's
secondary structure. 5.1–9.0. Hence, a fine-stranded gel would therefore be expected at pH 3
The PI protein had the highest enthalpy of all samples, while the and I: 15 mM, and the translucent gels of both patatin and spray dried
freeze dried isolate the lowest. van Koningsveld et al. (2001) also found powder agrees with this theory. Patatin could also form a translucent
the PI proteins to yield highest enthalpy of denaturation, while patatin gel at pH 7.5, while whole protein isolate powder formed an opaque gel
gave the lowest values and a whole protein isolate an intermediate at high pH, most likely due to the composition of several proteins,
value. Lokra, Schuller, Egelandsdal, Engebretsen, and Straetkvern among others enzymes and protease inhibitor proteins with pI in this
(2009) found a PI fraction to yield similar enthalpy to a total protein region. Particulate and fine-stranded gels differ in gel elasticity and gel
isolate, while a fraction with higher proportion of patatin gave sig- strength. This was clearly seen at pH 3.0 at I: 15 mM for all samples,
nificantly lower values. The observed difference between studies may where translucent fine-stranded gels had markedly higher elasticity
be related to sensitivity of instruments, heat ramp and protein con- than turbid gels at pH 3 at I: 200 mM. Tranlucent gels was found at
centration. Our studies gave very broad peaks in the DSC thermogram conditions that also resulted in high protein solubility (Fig. 2). It was
for both total protein isolates, and sharp peaks for PI and patatin not expected that low solubility conditions would result in translucent
fractions, while van Koningsveld et al. (2001) had sharp peaks for all gels, since a proportion of the proteins would exist as undissolved
samples. particles able to diffract light. High solubility conditions can however
For patatin van Koningsveld et al. (2001) reported close to zero also yield turbid gels as seen for the patatin sample at high pH and high
solubility at pH 3.5 at I: 200 mM and around 40% at pH 4 at I: 200 mM. salt, again highlighting the effect of aggregate-type formed during
Ralet and Gueguen (2000) reported a value of 60% at pH 4 at I: heating.
172 mM, while values around 68% was found in our study at pH 4 at I: Texture measurements by rheology or uniaxial compression did
200 mM. This difference can be explained, a.o. by differences in start generally show the same trends in regards to effect of pH and ionic
material, purification method, final purity and eventual presence of strength, when comparing measurements of strain. Differences were,
polyphenols. however, found for G′20C and stress at pH 3.0 for the whole protein
Viscous or gelled solutions, resulting in G’ values > 1 Pa before isolates and at pH 7.5 for the patatin sample. This may be due to the use
heat treatment, were observed at T < 25 °C for total isolate powders of different protein concentrations (8 or 15% w/w) and/or the fact that
and patatin at pH 4.5–5.0 at I: 15 mM and pH 3.0 at I: 200 mM. This oscillatory rheology is a nondestructive technique compared to the
behaviour was caused either by a high concentration of suspended large deformation until fracture situation in the uniaxial compression
undissolved protein particles, since the observed conditions fit to the analysis. Other authors have also reported very different mechanical
sample solubility minima or due to the fact that patatin unfolds at properties when the same gel was tested by small or large deformation
pH < 4 and that high salt content promotes interactions between (Stading, Langton, & Hermansson, 1995).
newly exposed hydrophobic patches. Optimum pH conditions for gelling as cited by (Giuseppin et al.,
The ratio between G′ after cooling relative to G′ at high temperature 2008) was pH 4.8–5.5 for patatin and pH 3.2–4.3 for the PI fraction.
(G′20 ˚C/G′85 ˚C) have previously been used to determine the role of Other optimum conditions was found in the present study, with the PI
covalent and noncovalent forces during protein gelation, with gels fraction having optimum at pH 3.0 and patatin at pH 7.5. This differ-
dominated by noncovalent forces yielding higher numbers (Martin, ence might related to ionic strength conditions, purity of sample

256
J.M. Schmidt, et al. Food Hydrocolloids 96 (2019) 246–258

material and the heat treatment used to induce gelation. Anandharamakrishnan, C., Rielly, C. D., & Stapley, A. G. F. (2007). Effects of process
The spray dried protein isolate had higher G′20 ˚C than the freeze variables on the denaturation of whey proteins during spray drying. Drying
Technology, 25(4–6), 799–807.
dried protein isolate at all tested conditions and also a higher stress Arntfield, S. D. (1996). Effects of divalent cations, phytic acid, and phenolic compounds
except at pH 3.0 at I: 15 mM. It is believed that spray drying results in on the gelation of ovalbumin and canola proteins. In N. Parris, A. Kato, L. K. Creamer,
limited denaturation, partial protein unfolding and thus a higher sur- & J. Pearce (Vol. Eds.), Macromolecular interactions in food Technology: Vol. 650, (pp.
82–92). Washington: Amer Chemical Soc.
face hydrophobicity, which may increase the protein gel firmness by Baier, A. K., & Knorr, D. (2015). Influence of high isostatic pressure on structural and
increased hydrophobic interaction in the gel network (Hammershoj, functional characteristics of potato protein. Food Research International, 77, 753–761.
Rasmussen, Carstens, & Pedersen, 2006). It can be difficult to predict Barta, J., Bartova, V., Zdrahal, Z., & Sedo, O. (2012). Cultivar variability of patatin bio-
chemical characteristics: Table versus processing potatoes (Solanum tuberosum L.).
the effect of drying conditions on protein functionality for two overall Journal of Agricultural and Food Chemistry, 60(17), 4369–4378.
reasons. Firstly, different proteins differ in sensitivity to drying heat Claussen, I. C., Strommen, I., Egelandsdal, B., & Stretkvern, K. O. (2007). Effects of drying
load (differ in denaturation temperature), and secondly, depending on methods on functionality of a native potato protein concentrate. Drying Technology,
25(4–6), 1091–1098.
the nature and magnitude of the change in protein folding structure
Creusot, N., Wierenga, P. A., Laus, M. C., Giuseppin, M. L., & Gruppen, H. (2011).
upon drying it can provide a change in one functional property, though Rheological properties of patatin gels compared with beta-lactoglobulin, ovalbumin,
not in another (Luck et al., 2013). The observed effect on increased gel and glycinin. Journal of the Science of Food and Agriculture, 91(2), 253–261.
strength by spray drying indicated a change in the potato protein Delahaije, R., Wierenga, P. A., Giuseppin, M. L. F., & Gruppen, H. (2015). Comparison of
heat-induced aggregation of globular proteins. Journal of Agricultural and Food
conformation, which do require a certain heat load. In contrast, for the Chemistry, 63(21), 5257–5265.
soy-protein fractions p-conglycinin and glycinin, the opposite relation is Doi, E. (1993). Gels and gelling of globular proteins. Trends in Food Science & Technology,
found, as spray drying with increased temperature results in protein 4(1), 1–5.
Foegeding, E. A. (2006). Food biophysics of protein gels: A challenge of nano and mac-
gels with lower G′ values and being more fragile (Tarone, Fasolin, roscopic proportions. Food Biophysics, 1(1), 41–50.
Perrechil, Hubinger, & da Cunha, 2013). Frydenberg, R. P., Hammershoj, M., Andersen, U., Greve, M. T., & Wiking, L. (2016).
Protein denaturation of whey protein isolates (WPIs) induced by high intensity ul-
trasound during heat gelation. Food Chemistry, 192, 415–423.
5. Conclusion Giuseppin, M.L.F., van der Sluis, C., Laus, M.C. (2008) Native potato protein isolates,
European Patent 1920662. https://patents.google.com/patent/EP1920662A1 as-
In this study, gels were prepared from either spray dried or freeze sesed 16.05.2019.
Hammershoj, M. (2001). In A. Friis, & S. L. Mason (Vol. Eds.), Physicochemical effects on
dried whole protein isolate powder or from patatin and protease in- egg albumen gel textural properties evaluated by large-deformation studies: Vol. 9, (pp.
hibitor isolates. Oscillatory rheology was used to map gelation tem- 107–109). The Nordic Rheology Society.
perature, storage modulus and deformation properties, while uniaxial Hammershoj, M., Peters, L. V., & Andersen, H. J. (2004). The significance of critical
processing steps in the production of dried egg albumen powder on gel textural and
compression was used to map stress and strain. For a few samples, gel
foaming properties. Journal of the Science of Food and Agriculture, 84(9), 1039–1048.
structure was initiated in the protein dispersions before heating. These Hammershoj, M., Rasmussen, H. C., Carstens, J. H., & Pedersen, H. (2006). Dry-pas-
viscous samples were associated to pH and ionic strength conditions teurization of egg albumen powder in a fluidized bed. II. Effect on functional prop-
that favoured low protein solubility or patatin denaturation. Spray erties: Gelation and foaming. International Journal of Food Science and Technology,
41(3), 263–274.
drying did generally not impair solubility or gel properties of whole Haque, M. A., Chen, J., Aldred, P., & Adhikari, B. (2015). Denaturation and physical
protein isolates as compared to freeze drying. Gels made of purified characteristics of spray-dried whey protein isolate powders produced in the presence
patatin had optimal solubility and gel properties at pH 7, whereas and absence of lactose, trehalose, and polysorbate-80. Drying Technology, 33(10),
1243–1254.
protease inhibitor showed optimum conditions at pH 3.0 and at I: Holm, F., & Eriksen, S. (1980). Emulsifying properties of undenatured potato protein-
15 mM. Gels made of total isolate protein powders was less affected by concentrate. Journal of Food Technology, 15(1), 71–83.
pH and salt conditions, but had a minimum at pH 5 I: 15 mM. Gels with Jorgensen, M., Stensballe, A., & Welinder, K. G. (2011). Extensive post-translational
processing of potato tuber storage proteins and vacuolar targeting. FEBS Journal,
opaque appearance were observed at most pH and salt conditions ex- 278(21), 4070–4087.
cept at pH 3.0 at I: 15 mM for total isolate powder and for the protease Knorr, D., Kohler, G. O., & Betschart, A. A. (1977). Potato protein concentrates: The in-
inhibitors, as well as pH 3.0 and pH 7.5 at I: 15 mM for patatin, since fluence of various methods of recovery upon yield, compositional and functional
characteristics. Journal of Food Processing and Preservation, 1(3), 235–247.
these conditions far from pI of patatin resulted in translucent gels.
van Koningsveld, G. A., Gruppen, H., de Jongh, H. H. J., Wijngaards, G., van Boekel, M.,
Hence, the protein structural unfolding and surface charge in relation to Walstra, P., et al. (2001). Effects of pH and heat treatments on the structure and
isolectric point of the proteins correlated well with the gel network solubility of potato proteins in different preparations. Journal of Agricultural and Food
Chemistry, 49(10), 4889–4897.
appearance, especially for the more pure patatin fraction.
Laemmli, U. K. (1970). Clevage of structural proteins during assembly of head of bac-
This work gives further insight into the impact of physicochemical teriophage-T4. Nature, 227(5259), 680–685.
conditions modified by pH, salt and protein purity on potato protein Lechevalier, V., Perinel, E., Jeantet, R., Lesaffre, C., Croguennec, T., Guerin-Dubiard, C.,
gelation and the resulting gels properties. Hereby, we enable control of et al. (2005). Statistical analysis of effects of industrial processing steps on functional
properties of pasteurised liquid egg white. Journal of the Science of Food and
mechanical properties and appearance of potato protein that are im- Agriculture, 85(5), 757–769.
portant for different food applications. Lokra, S., Helland, M. H., Claussen, I. C., Straetkvern, K. O., & Egelandsdal, B. (2008).
Chemical characterization and functional properties of a potato protein concentrate
prepared by large-scale expanded bed adsorption chromatography. Lwt-Food Science
Acknowledgments and Technology, 41(6), 1089–1099.
Lokra, S., Schuller, R. B., Egelandsdal, B., Engebretsen, B., & Straetkvern, K. O. (2009).
The authors thank the Future Food Innovation of region Mid- Comparison of composition, enzyme activity and selected functional properties of
potato proteins isolated from potato juice with two different expanded bed resins.
Jutland, Denmark, KMC, AKV Langholt and Aarhus University for fi- Lwt-Food Science and Technology, 42(4), 906–913.
nancial support of the present work. Luck, P. J., Vardhanabhuti, B., Yong, Y. H., Laundon, T., Barbano, D. M., & Foegeding, E.
A. (2013). Comparison of functional properties of 34% and 80% whey protein and
milk serum protein concentrates. Journal of Dairy Science, 96(9), 5522–5531.
Appendix A. Supplementary data
Martin, A. H., Nieuwland, M., & de Jong, G. A. H. (2014). Characterization of heat-set gels
from RuBisCO in comparison to those from other proteins. Journal of Agricultural and
Supplementary data to this article can be found online at https:// Food Chemistry, 62(44), 10783–10791.
Mishra, R., Govindasamy-Lucey, S., & Lucey, J. A. (2005). Rheological properties of re-
doi.org/10.1016/j.foodhyd.2019.05.022.
nnet-induced gels during the coagulation and cutting process: Impact of processing
conditions. Journal of Texture Studies, 36(2), 190–212.
References Pots, A. M., ten Grotenhuis, E., Gruppen, H., Voragen, A. G. J., & de Kruif, K. G. (1999).
Thermal aggregation of patatin studied in situ. Journal of Agricultural and Food
Chemistry, 47(11), 4600–4605.
Ainis, W. N., Ersch, C., Farinet, C., Yang, Q. H. Z., Glover, Z. J., & Ipsen, R. (2019). Pouvreau, L., Gruppen, H., Piersma, S. R., van den Broek, L. A. M., van Koningsveld, G. A.,
Rheological and water holding alterations in mixed gels prepared from whey proteins & Voragen, A. G. J. (2001). Relative abundance and inhibitory distribution of pro-
and rapeseed proteins. Food Hydrocolloids, 87, 723–733. tease inhibitors in potato juice from cv. Elkana. Journal of Agricultural and Food

257
J.M. Schmidt, et al. Food Hydrocolloids 96 (2019) 246–258

Chemistry, 49(6), 2864–2874. Smith, P. K., Krohn, R. I., Hermanson, G. T., Mallia, A. K., Gartner, F. H., Provenzano, M.
Ralet, M. C., & Gueguen, J. (2000). Fractionation of potato proteins: Solubility, thermal D., Fujimoto, E. K., Goeke, N. M., Olson, B. J., & Klenk, D. C. (1985). Measurement of
coagulation and emulsifying properties. Lebensmittel-Wissenschaft Und-Technologie- protein using bicinchoninic acid. Analytical Biochemistry, 150(1), 76–85.
Food Science And Technology, 33(5), 380–387. Stading, M., & Hermansson, A. M. (1990). Viscoelastic behaviour of beta-lactoglobulin gel
Rickert, D. A., Johnson, L. A., & Murphy, P. A. (2004). Functional properties of improved structures. Food Hydrocolloids, 4(2), 121–135.
glycinin and beta-conglycinin fractions. Journal of Food Science, 69(4), 303–311. Stading, M., Langton, M., & Hermansson, A. M. (1995). Small and large-deformation
Schmidt, J. M., Damgaard, H., Greve-Poulsen, M., Larsen, L. B., & Hammershoj, M. studies of protein gels. Journal of Rheology, 39(6), 1445–1450.
(2018). Foam and emulsion properties of potato protein isolate and purified fractions. Tarone, A. G., Fasolin, L. H., Perrechil, F. D., Hubinger, M. D., & da Cunha, R. L. (2013).
Food Hydrocolloids, 74, 367–378. Influence of drying conditions on the gelling properties of the 7S and 11S soy protein
Schmidt, J. M., Greve-Poulsen, M., Damgaard, H., Sunds, A. V., Zdrahal, Z., Hammershoj, fractions. Food and Bioproducts Processing, 91(C2), 111–120.
M., et al. (2017a). A new two-step chromatographic procedure for fractionation of Urbonaite, V., van der Kaaij, S., de Jongh, H. H. J., Scholten, E., Ako, K., van der Linden,
potato proteins with potato fruit juice and spray-dried protein as source materials. E., et al. (2016). Relation between gel stiffness and water holding for coarse and fine-
Food and Bioprocess Technology, 10(11), 1946–1958. stranded protein gels. Food Hydrocolloids, 56, 334–343.
Schmidt, J. M., Larsen, L. B., & Hammershoj, M. (2017b). Appearance and textural Weijers, M., Sagis, L. M. C., Veerman, C., Sperber, B., & Linden, E. v.-d. (2002). Rheology
properties of sheared low concentration potato protein gels-impact of drying method, and structure of ovalbumin gels at low pH and low ionic strength. Food Hydrocolloids,
pH, and ionic strength. Journal of Food Science, 82(9), 2056–2061. 16(3), 269–276.

258