Electrophoresis 2018, 39, 1975–1983 1975

Ioannis Mourtzinos1 Research Article

Prodromos Prodromidis1
Spyros Grigorakis2
Dimitris P. Makris3 Natural food colorants derived from onion
Costas G. Biliaderis1
Thomas Moschakis1
wastes: Application in a yoghurt product
1 Department of Food Science The valorization of onion (Allium cepa) solid wastes, a 450 000 tonnes/year waste in
and Technology, School of
Agriculture, Aristotle University Europe, by a green extraction method is presented. Polyphenols of onion solid wastes were
of Thessaloniki, Thessaloniki, extracted using eco-friendly solvents, such as water and glycerol. The 2-hydroxypropyl-␤-
Greece cyclodextrin was also used as a co-solvent for the augmentation of the extraction yield. The
2 Food Quality & Chemistry of
process has been optimized by implementing a central composite face centered design
Natural Products,
Mediterranean Agronomic of experiments, with two replicates in the central point, taking into consideration the
Institute of Chania (M.A.I.Ch.), following independent variables: glycerol concentration, cyclodextrin concentration and
Chania, Greece temperature. The assessment of the extraction model was based on two responses: the
3 School of Environment,
University of the Aegean, total pigment yield and the antiradical capacity. LC-MS analysis was also employed in
Myrina, Lemnos, Greece order to identify polyphenols and colorants of the obtained extracts. The main polyphenols
found were quercetin and quercetin derivatives and the main colorant was cyanidin 3-O-
glucoside. The extract was also tested as a food colorant in a yoghurt matrix. The onion
Received February 12, 2018
leaf extract was found to be a stable natural colorant and could be utilized as an alternative
Revised May 19, 2018
Accepted May 21, 2018 ingredient to synthetic coloring agents.

Keywords:
Cyclodextrins / Extraction / Inclusion complex formation / LC-MS analysis /
Polyphenols / Yoghurt DOI 10.1002/elps.201800073

 Additional supporting information may be found online in the Supporting Infor-

mation section at the end of the article.

1 Introduction food colorants, fulfilling the increased demand for clean la-
bels in food products [5]. OSW extracts have been already
Onion (Allium cepa) is the most widely cultivated horti- incorporated in food matrices exploiting their antiradical
cultural crop in Europe, with a reported annual produc- (phenolics) and antibrowning (thiol containing compounds)
tion of almost 47 million tons. In year 2000, more than characteristics [6].
450 000 tonnes of onion solid waste (OSW) were produced Different categories of high-added-value constituents
only in Europe [1].OSWs is composed of the non-edible part could be recovered from agro-food by-products, such as
of the onion bulb that is the outer dry and semi-dry layers, as carbohydrates, sugars, pectins, proteins and phenolic com-
well as the apical trimmings. The outer dry layers of onion pounds. Their exploitation in the food industry is gain-
bulb, which constitute the main waste of onion, is a source of ing more ground due to their low-cost and also due to
valuable polyphenols such as flavonoids and anthocyanins [2]. their functional properties [7]. There are numerous meth-
Flavonoids from various extracts/fractions of red onion peel ods used for the extraction of polyphenols from plant
possess antioxidant and antimutagenic activities [3, 4]. An- derived wastes. Conventional methods include: solid-liquid
thocyanins, beyond their bioactivity linked to certain health extraction, in which plant tissues are extracted with liquid
benefits, are natural pigments that can be used as natural solvents by maceration or Soxhlet extraction, as well as liquid-
liquid extraction. Basic disadvantages of the traditional extrac-
tion techniques are the usage of petroleum or other organic
Correspondence: Dr. Thomas Moschakis, Department of Food Sci- solvents, the high energy consumption, the thermal degrada-
ence and Technology, School of Agriculture, Aristotle University tion of thermo-sensitive polyphenols, and the long extraction
of Thessaloniki, P.O. Box 235, Thessaloniki GR-541 24, Greece
times [8].
E-mail: [email protected]
Recently, the principles of green engineering and
green chemistry have been introduced in various extraction
Abbreviations: AAR , antiradical capacity; CD, cyclodextrin;
ESI, electrosprayionization; LC-MS, liquid chromatography-
schemes aiming to promote sustainable extraction processes,
mass spectroscopy; OSW, onion solid waste; RSM, response basically by using non-petroleum derived solvents and renew-
surface methodology; YTPm , Total pigment yield able plant residues as raw materials [9]. Glycerol could be


C 2018 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.electrophoresis-journal.com
1976 I. Mourtzinos et al. Electrophoresis 2018, 39, 1975–1983

considered as an ideal green non-conventional solvent for the Several foods, like yoghurt, are considered healthy but
extraction of polyphenols as it is abundant and of low cost, they lack of phenolic compounds which are well known for
because it is a by-product of the bio-diesel industry. Moreover, their multi-functional positive role in human health. There-
it is non- toxic and non-flammable. Glycerol mixtures with fore, a possible incorporation of plant extracts rich in antho-
other compatible solvents (e.g. water) might constitute a very cyanins in such fermented products might impart a desirable
suitable medium for polyphenol extraction, since they can red color, as well as enhance their health status due to the
possess a relatively low dielectric constant, which has been presence of phenolic compounds [22].
claimed to be a key characteristic regarding their solvency The aim of this study was to perform an optimization of
towards polyphenols that are otherwise sparingly soluble in an extraction process for the efficient recovery of flavonoid
pure water [10]. and colorants from OSWs, using ‘green’’ water/glycerol/2-
Molecular encapsulation with cyclodextrins (CDs) in- hydroxypropyl-␤-cyclodextrin ternary mixtures. The opti-
volves the incorporation of a guest molecule or group in the mization was based on a central composite face centered
CD cavity. The most common native CDs are composed of 6, design of experiments, with two replicates in the center
7 and 8 glucosyl units and named as ␣-, ␤- and ␥ -CD, respec- point, targeting at two responses, i.e. the total pigment
tively. ␤-CD has limited aqueous solubility, thus, derivatives yield (YTPm ) and the antiradical capacity (AAR ). Moreover,
like 2-hydroxypropylated-␤-CD (hp-␤-CD), with improved wa- the potential use of the liquid extract enriched in antho-
ter solubility have been synthesized [11]. Cyclodextrins have cyanins as a natural colorant in a yoghurt preparation was
a hydrophobic interior and a hydrophilic exterior. The hy- examined.
drophobic interior of the torus-like molecular structure can
vary in size by varying the number of anhydroglucose units
in the cyclodextrin molecules. Polyphenols, from fruits, veg-
2 Materials and methods
etables and other plant tissues can be encapsulated in cy-
clodextrins as the benzene ring (hydrophobic residue) of 2.1 Materials
polyphenols is an ideal guest for the cyclodextrin cavity. Sev-
eral examples are the encapsulation of gallic acid [12], hy- 2-Hydroxypropyl-␤-cyclodextrin (hp-␤-CD, average MW
droxycinnamic acids, such as caffeic acid, p-coumaric acid 1460), ␤-cyclodextrin (␤-CD), trolox and the 2,2-diphenyl-
and ferulic acid [13], ellagic acid (an important antioxidant picrylhydrazyl (DPPH) stable radical were obtained from
polyphenol of many fruits like pomegranates, raspberries, Sigma Chemical Co. (St. Louis, MO, USA.). Glycerol (⬎99%)
strawberries, blackberries), as well as olive leaf extract (rich in was from Fisher Scientific (New Jersey, NJ, USA.). Red
oleuropein). Inclusion complex formation with polyphenols onions of the variety Red Cross F1, originated from cen-
increases their antioxidant activity which may be attributed tral Greece (Thiva), without apparent damages and infec-
to an increase of their solubility [14]. Aqueous solutions of cy- tions, were bought from a local supermarket (Thessaloniki,
clodextrins could be considered as alternative, green solvents, Greece). The bulbs were peeled using a sharp cutter to re-
since upon formation of complexes involving the hydropho- move the onion parts considered as waste material, such
bic CD cavities and non-polar compounds, a reduction of the as the outer dry and semi-dry layers, and also the apical
system’s free energy can be achieved [15]. Moreover, CDs trimmings. The obtained materials were dried at 65°C for
are non-toxic ingredients that can be used in pharmaceutical, 48 h and ground using a domestic blender (Bosch MMB
cosmetic and food applications [16]. 11R). The particle size of the obtained material was less than
Plant extracts rich in anthocyanins have been used as a 0.3 mm.
substitute for synthetic food colorants, such as FD&C Red 40 Dried and ground OSW extract was stored at −20°C until
and carmine, in food products like jams, dairy products, bev- further use. Full fat milk (3.5%) was obtained from a local
erages, or confectioneries [17]. Common natural sources of supermarket in Greece.
anthocyanins are red grape marc, elderberry, black carrot, red
cabbage, and purple potato [18]. The increasing demand for
natural food colorants cannot currently be fulfilled. Recently, 2.2 Extraction procedure
new natural coloring agents and/or new natural sources of
colorants, or new formulations, are intensively studied. Nat- An amount of plant tissue powder was mixed with solvent
ural pigment applications can be limited by lower stability, (liquid-to-solid ratio of 50 mL/g), composed of varying con-
interactions with food ingredients, and inability to match de- centrations of water (40–100% w/v), glycerol (0–60% w/v) and
sired hues. Therefore, no single naturally derived colorant CD (0–13% w/v) in a stoppered glass bottle. The material was
can serve as a universal alternative for a specified synthetic subjected to extraction under continuous stirring at 600 rpm
colorant in all applications [19]. As an example, anthocyanins for 240 min in a water bath at three different temperatures
from Hibiscus sabdariffa L. (roselle) plant have been studied as (40, 60, and 80 ± 1°C). Following extraction, the samples were
natural colorants [20]. However, the potential use of agricul- centrifuged in a table centrifuge (Hermle Z300K, Wehingen,
tural by-products/residues (agro-wastes) has not been largely Germany) at 5000 rpm for 10 min. The clear supernatant was
explored as a source of natural colorants [21]. stored at −20°C until used for further analysis.


C 2018 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.electrophoresis-journal.com
Electrophoresis 2018, 39, 1975–1983 Liquid Phase Separations 1977

2.3 Determination of antiradical capacity (AAR ) and described the response values as a function of the indepen-
total pigment yield (YTPm ) dent variables. Analysis of variance (ANOVA) was used to
assess the statistical significance of the model. Insignificant
For AAR determination, a previously described protocol was dependent terms (p ⬎ 0.05) were omitted from the regression
essentially used [23]. Briefly, an aliquot of 25 ␮L of the sample models derived, through a “backward elimination” process.
was added to 975 ␮L DPPH solution (100 ␮M in MeOH), Contour plots were subsequently obtained using the fitted
and the absorbance was recorded at t = 0 and t = 30 min. model. The JMP 10 software was used for statistical analysis.
TroloxTM equivalents (mM TE) were determined from linear
regression, after plotting %⌬ A515nm of known solutions of
troloxTM against concentration, where 2.5 Liquid chromatography–mass spectrometry
At=0− At=30
%⌬ A515 = 515 515
× 100 (1) A previously described method [27] has been employed for
At=0
515 LC-MS, using a Finnigan MAT Spectra System P4000 pump
The results were expressed as ␮mol TE per g of dry OSW coupled with a UV6000LP diode array detector and a Finni-
weight. gan AQA mass spectrometer. Analyses were carried out on a
For total pigment yield (YTPm ) a method described before Superspher RP-18, 125 × 2 mm, 4 ␮m, column (Macherey–
was employed [24]. A volume of 100 ␮L of sample was mixed Nagel, Germany), coupled with a guard column packed with
with 0.9 mL of a HCl solution (0.25 M in ethanol) and the the same material, and maintained at 40 ̊C. Analyses were
mixture was left to equilibrate for 10 min. The absorbance at carried out by employing electrospray ionization (ESI) at the
520 nm (A520) was obtained and the total pigment yield was positive ion mode, with acquisition set at 12 and 50 eV, cap-
determined as mg cyanidin 3-O-glucoside equivalents (CyE) illary voltage 4 kV, source voltage 4.9 kV, detector voltage
per g dry OSW, using as ε = 26,900 and MW = 449.2, as 650 V and probe temperature 400 ̊C. Eluent (A) and eluent
follows: (B) were 2.5% acetic acid and methanol, respectively. The flow
rate was 0.33 mL/min, and the elution programme used was
YT Pm = 16.7 × A520 × V × F D/m (2)
as follows: 0–5 min, 0% B; 5–30 min, 100% B; 30–35 min,
where V is the volume of the extraction medium, m the dry 100% B. The samples were diluted with methanol prior to
weight of OSW (g) and FD the dilution factor. analysis.

2.4 Experimental design 2.6 Preparation of yoghurt

A central composite face centered design was employed to Milk (3.5% fat) was heat-treated at 90°C for 30 min, followed
identify the relationship between the response functions by cooling at 42°C and then inoculated under aseptic con-
and process variables, as well as to determine those condi- ditions with a freeze-dried starter culture YF-L811 (Yoflex
tions that optimized extraction. The independent variables or Chr. Hansen Ltd), as recommended by the culture manu-
factors studied were 2-hydroxypropyl-␤-cyclodextrin concen- facturer. Extraction of phenolic compounds from OSW was
tration (CCD ), varying between 1 and 13% (w/v), glycerol con- carried out using a constant ratio of dried OSW powder and
centration (Cgl ), varying between 0 and 60% (w/v), and tem- aqueous solutions (5 g OSW/100 g solution) with or with-
perature (T), varying between 40 and 80°C. Value ranges were out addition of ␤-CD (1.84% w/w) or hp-␤-CD (13% w/w).
chosen on the basis of preliminary experimentation and pre- The freeze-dried material obtained from the natural extracts
vious studies [25, 26]. The independent variables were coded and originating from 50 up to 350 mg OSW was initially
at three levels, −1, 0 and 1 (1, 7, 13 w/v for CCD , 0, 30, 60 w/v dissolved in 100 g of milk. Subsequently, the milk (with or
for Cgl and 40, 60, 80°C for T, respectively) according to the without the colorant containing extract) was heated for 30 min
following equation: at 90°C, followed by cooling at 42°C and inoculation with the
starter culture, and then, the samples were kept quiescent for
Xi − X0
xi = , xi = 1, 2, 3 (3) 200 min. Afterwards, the samples were immediately stored
⌬ Xi at 5°C.
where xi and Xi are dimensionless, and the actual value of the
independent variable i, X0 , is the actual value of the indepen-
dent variable i at the central point, and ⌬ Xi is the step change 2.7 Colour measurements of yoghurt
of Xi, corresponding to a unit variation of the dimensionless
value. The total pigment yield (YTPm ) and the antiradical ca- The change in tristimulus color coordinate (⌬ E*) was mea-
pacity (AAR ) were chosen as the dependent variables since sured with a colorimeter (HunterLab, Reston, Virginia, USA)
they are related with the phenolic content of the obtained over a month’s storage period. Here: ⌬ E* = [(⌬ L*)2 + (⌬ a*)2
extracts under the extraction process conditions adopted. + (⌬ b*)2 ]1/2 , where ⌬ L*, ⌬ a*, and ⌬ b* are the differences
The data obtained were subjected to regression analy- between the colour coordinates [28]. The CIE L*, a* and b*
sis using least square methodology to obtain equations that values were determined as indicators of lightness, redness


C 2018 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.electrophoresis-journal.com
1978 I. Mourtzinos et al. Electrophoresis 2018, 39, 1975–1983

and yellowness, respectively. Five yoghurts from every treat- 2.10 Microbiological analysis
ment were randomly selected and instrumental colour mea-
surements were taken on the product’s surface. Colour values Following milk inoculation with the starter culture, the sam-
reported herein represent the mean of ten measurements per ples were incubated at the required temperature for 200 min,
treatment. fast cooled by immersing them into ice water and subse-
quently stored at 5°C. Enumeration of bacteria, using the
pour plate technique, was carried out at the end of fer-
2.8 Rheological measurements of yoghurt mentation and after 7 and 14 days storage of the products
at 5°C. Plates containing 30–300 colonies were enumer-
Rheological measurements were performed by a Physica ated and recorded as colony forming units per milliliter of
MCR 300 (PhysicaMesstechnik GmbH, Stuttgart, Germany) culture (cfu/mL).
rheometer using a concentric cylinder geometry (CC 27, cup
and bob diameters, 28.92 and 26.66 mm, respectively). The
gelation of milk samples was assessed by applying small
deformation oscillatory rheological testing. The viscoelastic 3 Results and discussion
properties, storage (G ), loss (G ) moduli and loss tangent
(tan␦ = G /G ), were monitored at a constant small target 3.1 Optimization of the extraction
shear strain of ␥ = 0.005 (i.e. 0.5%) and at a fixed frequency
of 1 Hz. These tests were carried out for a time interval of Optimization of the extraction was carried out by evaluating
5 h. During the first 200 min, the temperature was adjusted at the effect of the three selected independent variables (see
42°C, while for the last period, the temperature was dropped Table S1); CD concentration (CCD , X1), glycerol concentra-
rapidly and kept at 5°C, simulating the commercial yoghurt tion (Cgl , X2) and temperature (T, X3) that directly affects the
manufacturing process. The temperature was regulated by a extraction. The time of the extraction for onion solid waste
Paar Physica circulating bath and a controlled peltier system using polar solvent has been optimized previously and was 4
(TEZ 150P/MCR) with an accuracy of ±0.1°C. hours [1, 30]. The liquid-to-solid ratio (L/S) is also crucial for
The structure of the yoghurts was characterized by the the extraction process and generally the higher the ratio the
following rheological parameters: higher the yield. A value of 50 mL/g was selected as a mean
Maximum storage modulus (G max) (Pa) was deter- level of previously proposed values that varied between 10 and
mined as the value of G at 5°C; Minimum loss tangent 100 mL/g [31]. The values of the responses (AAR and YTPm )
(tan␦min ) was the value of tan␦ corresponding to G max; On- obtained experimentally were analyzed by multiple regres-
set of gelation was determined as the time (min) at which sion analysis and after removal of the non-significant terms
tan␦ = 1; and Elasticity Increment IE [(dlogG /dt)max], de- (p ⬎ 0.05), the theoretical models (second-order polynomial
fined by the slope of log G (t) at the inflection point. The equations) were as follows:
latter parameter is used as a measure of gelation rate (i.e. a
high IE value implies rapid gelation) and indicates the num- Y TPm = 2.65 + 0.38X1 + 0.31X2 + 0.15X3 − 0.21X1X2 (4)
ber of decades of G increases at maximum per unit time; its
dimension is reciprocal time. AAR = 245.87+13.17X1+36.49X2+14.80X3+21.22X2X3
(5)
− 46.46X12 +32.75X22

2.9 Acidification activity measurements The quality of fit was checked using the coefficients of de-
termination (R2 ), which were 0.94 and 0.95 for AAR and YTPm ,
Changes in pH during fermentation were monitored con- respectively. The results of ANOVA are given in the Sup-
tinuously by means of a glass electrode pH meter (Hanna porting Information Table 2. This outcome clearly pointed
Instruments, Padova, Italy). The electrode was disinfected to a statistically significant match between observed and pre-
with 70% (v/v) ethanol solution, rinsed with sterile distilled dicted responses, and that the models given by the above two
water and placed in the milk bottles under aseptic condi- equations can predict the optimal experimental conditions
tions after inoculation of milk with the culture. Fermentation with high reliability. The values of all responses (measured
was carried out in a circulating water bath (±0.2°C) preset at and predicted) are analytically provided in the supplementary
42°C and the pH was automatically recorded at 15-min reg- material.
ular intervals. The following acidification parameters were The polynomial equations of both variables AAR and YTPm
calculated to characterize the fermentation kinetics: Maxi- showed that there was a clear positive effect of all three inde-
mum acidification rate (Vm), calculated from the pH curves pendent variables. The influence of CCD was not only clearly
according to the equation Vm = (dpH /dt)max and expressed shown in the polynomial equation of AAR and YTPm as a di-
in absolute values (mUnit pH/min); the time and pH at rect contributing factor, but also by the significant interaction
which the maximum acidification rate was observed, (tm ) and effect between CCD and Cgl .
(pHm), respectively; (te ) is the time at which a pH of 4.7 was Increases in all independent variables showed increased
reached [29]. recovery of polyphenols. These trends were recorded in the


C 2018 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.electrophoresis-journal.com
Electrophoresis 2018, 39, 1975–1983 Liquid Phase Separations 1979

form of contour plots. As can be seen in Supporting Informa- which provided the optimal values for all process variables
tion Fig. 1, maximization of YTPm levels was observed when considered Supporting Information Fig. 2. The maximum
glycerol content was 60% (Supporting Information Fig. 1A), desirability (0.87) was achieved at optimum conditions of
whereas CD content and temperature showed optimum val- Cgl , CCD , and T. Under optimum conditions, the maxi-
ues for the extraction of onion polyphenols at 13% (w/v) and mum values estimated were YTPm = 3.13 mg CyE/g (d.w.)
80 ̊C, respectively (Supporting Information Fig. 1B, C). A [Cgl = 60% (w/v), CCD = 13% (w/v) and T = 80 ̊C] and
similar pattern but slightly different has been also noted for AAR = 353.63␮mol TE / g (d.w.) [Cgl = 60% (w/v), CCD = 6.5 %
AAR (Supporting Information Fig. 1C, D, F), where optimum (w/v) and T = 80 ̊C], respectively. These values fall within
value for CD content was 6.5% (Supporting Information ±1 standard deviation calculated by the model (desirability
Fig. 1E). plot).
The fact that both values of YTPm and AAR were not de-
creased above 70°C may be attributed to the stability offered
by the cyclodextrin to polyphenols during heating. The pro- 3.2 Phytochemical profiling
tection of anthocyanins in aqueous solutions, being the main
colorants of the extract, from thermal degradation in the pres- The polyphenolic composition of extracts is crucial in
ence CD has been evaluated before by thermal stability stud- order to promote the usage of natural extracts in food ap-
ies [32]. plications. In general, cyclodextrins seem to alter the solubil-
Regarding the glycerol content impact, a recent study [33] ity of bioactive ingredients in aqueous media leading to in-
showed similar results, where by increasing the amount of creased extractability of polyphenols. Moreover, less soluble
glycerol in the extraction medium there was an increase in or even non-polar ingredients can be solubilized in aque-
the poyphenolic content up to a maximum glycerol level of ous solutions by including cyclodextrins in the extraction
90% (w/v). In the present study, due to the parallel usage medium. In order to examine the selectivity of the mixture
of cyclodextrin, glycerol cannot be used in amounts greater water/glycerol/2-hydroxypropyl-␤-cyclodextrin, LC-MS anal-
than 60% as the extraction solvent mixture becomes very ysis of the extract was performed. In Fig. 1 and in Supporting
viscous and a certain amount of cyclodextrin becomes non- Information Table 3 the polyphenolic profile of onion waste
dissolvable. extract is presented. The LC-MS analysis revealed the exis-
By using both glycerol and CD as the extraction medium tence of at least 12 main phenolics in the extracts. Phenolic
higher extractions yields of polyphenols can be achieved constituents identified were the main colorants cyanidin 3-O-
(75% higher than ethanol extraction) [15]. Consequently, glucoside and malonylated cyanidin and the main flavonoid
other effective, but in some cases undesirable or adverse quercetin as well as their derivatives. Cylodextrin incorpora-
extraction solvents, such us ethanol or water/ethanol mix- tion into the solvent medium seems to increase the solubility
tures [30], can be avoided. In addition, both glycerol and CD of the main polyphenols without altering the phenolic pro-
are food grade materials and thus the extracts can be directly file of a typical phenolic extract derived from onion solid
incorporated into food formulations, without compromising waste [26].
safety, since there is no issue of toxicity. Overall, it appears
that glycerol can be effective as an alternative green solvent
for extraction of polyphenols with improved efficiency, com- 3.3 Incorporation and stability of colorants in
pared to water, when methanol or ethanol as co-solvents are yoghurt
not desired.
The optimization of CD content in the extraction medium Even though yoghurt does not naturally contain polyphenols
is also crucial as the total amount of ingredients (of varying it appears as an excellent delivery vehicle for plant-derived
composition) that can form inclusion complexes with CD and phenolic compounds to further improve its health promoting
thereby influence positively the extraction process cannot be properties as a fermented dairy product. The stability of phe-
estimated theoretically. Ligand inclusion in the CD cavity is nolic compounds, particularly anthocyanins,was not found
a stoichiometric phenomenon and molecular inclusion be- to be largely altered during storage of yoghurt, due to its low
tween common polyphenols and CD has a stoichiometry of pH [34]. In another study the authors suggested a reduction
1:1. On the other hand, other ingredients of the onion ex- of polyphenols in fortified yoghurt due to the polyphenol-
tract may compete or promote inclusion complex formation protein interaction [35]. It is also known that polyphenols are
between CD and the onion main bioactives. In the present thermo-sensitive ingredients and might be partially decom-
study, the highly soluble hp-␤-CD was selected in order to posed during thermal treatment. The process for production
investigate the CD impact, as the unmodified ␤-CD, the most of yoghurt fortified with polyphenols often involves the ad-
common cyclodextrin used in foods, has limited solubility in dition of polyphenols in a form of solid extract before heat
water, up to a concentration of 16 mM. treatment followed by cooling [35, 36]. In the current pro-
The mathematical models enable the determination of posed methodology natural extracts in liquid form were in-
the optimal set of conditions and the maximum predicted stead added both prior to and post thermal treatment. The
YTPm and AAR values. Such determination was based on addition of the natural colorants (anthocyanin extracts from
the simultaneous maximization of the desirability function, OSW) before the heat treatment has led to enhanced and


C 2018 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.electrophoresis-journal.com
1980 I. Mourtzinos et al. Electrophoresis 2018, 39, 1975–1983

Figure 1. Polyphenolic profile of the extract obtained with 7% w/v CD content, 60% (w/v) glycerol content and temperature at 80°C.
The eluted compounds were monitored at 260 nm. (1) Unknown compound, (2) Unknown compound, (3) 2-(3,4-dihydroxybenzoyl)-2,4,6-
trihydroxybenzofuran-3(2H)-one, (4) Cyanidin 3-O-glucoside, (5) Malonylated cyanidin 3-O-glucoside derivative, (6) Unknown compound,
(7) Quercetin glucoside, (8) Quercetin, (9) Quercetin glucoside/quercetin dehydrate adduct, (10) Quercetin glucoside/quercetin dehydrate
adduct, (11) Quercetin dimer, (12) Quercetin dimer, (13) Quercetin dehydrodimer.

3.3.1 Microbial characterization and acidification

kinetics of fermented milk

Viable count measurements for both bacteria of the yogurt

culture were made at the end of fermentation. In all samples,
the viable count numbers were not significantly different,
with the mean numbers of the yoghurt starter culture be-
ing 5–10 108 cfu/g. Therefore, it can be concluded that the
viable count cells of fresh produced yoghurt were not signif-
icantly affected by the addition of the natural extract and the
cyclodextrins employed to facilitate phenolic extraction.
The acidification kinetics of milk gels were characterized
by the maximum acidification rate (Vm), the time and pH at
which Vm was reached (tm and pHm), ⌬ pH4h and the time
Figure 2. Time dependence of storage modulus (G ) and tan␦ for necessary to reach pH 4.8 (te ) (Table 1). The acidification pa-
model yoghurt systems supplemented with onion waste extracts rameters were not significantly affected by the addition of any
at a frequency of 1 Hz (␥ = 0.5%) at 42°C (200 min), followed by of the natural extracts derived from 50 mg dried OSW/100 g
storage at 5°C for 1 h. milk. However, at higher levels of polyphenol content (in-
volving the addition of freeze-dried extracts, obtained from
150 to 350 mg dried OSW, to 100 g of milk) the acidification
more vivid colours in the final product and therefore the re- kinetics were modified, as indicated by the recorded acidi-
sults presented hereinafter are referred to the addition of the fication parameters (Vm, pHm, tm ); i.e. the tm and te were
extract powders derived from OSW before heating the milk. more extended time periods, indicating lower culture activ-
Stability tests were also performed in the fortified product by ity, as the concentration increased from 50 to 350 mg dried
measuring different physicochemical parameters over a pe- OSW/100 g milk. These findings are in accordance with the
riod of one month. As can be seen in Supporting Information lower ⌬ pH4h and pHm recorded values, and can be ascribed
Table 4, the color parameters and YTPm remained relatively to a reduction of the starter culture activity by incorporation
constant upon storage for 4 weeks. This implies that the nat- of the natural extract containing ␤-CD or hp-␤-CD, possi-
ural colorants were not substantially affected by the low pH bly due to different amounts of polyphenols extracted in the
(4.5) and storage temperature (5°C) of the product; the L*, presence of cyclodextrins. These observations require further
a* and b* values were determined as indicators of lightness, investigation to unravel the origin of the hindering effects on
redness and yellowness, respectively. yoghurt culture activity during fermentation.


C 2018 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.electrophoresis-journal.com

Table 1. Effect of adding feeeze-dried onion natural extracts on the acidification parameters of fermented yoghurt-type products

Control With dried extracts With dried extracts With dried extracts With dried extracts With dried extracts With dried extracts With dried extracts
from 50 mg OSW from 50 mg OSW from 50 mg OSW from 150 mg OSW from 150 mg OSW from 150 mg OSW from 350 mg OSW
using ␤-CD using hp-␤-CD using ␤-CD using hp-␤-CD using ␤-CD

Vmax (mUpH/min) 0.0167 0.0157 0.0155 0.0163 0.0145 0.0220 0.0142 0.0104
tm (min) 120 120 120 120 120 141 120 200
te (min) (pH = 4.7) 210 211 212 214 212 222 217 250
Electrophoresis 2018, 39, 1975–1983

⌬pH 4h 1.95 1.93 1.90 1.92 1.90 1.79 1.91 1.69

pH 4h 4.69 4.69 4.68 4.65 4.64 4.72 4.61 4.79
pHm 5.58 5.61 5.55 5.55 5.50 5.46 5.48 5.10

Vmax : Maximum acidification rate; tm : the time at which the maximum acidification rate was observed; pHm: the pH at which the maximum acidification rate was observed; te : the time at
which a pH of 4.7 was reached.

C 2018 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

Table 2. Effect of adding feeeze-dried onion natural extracts on the rheological parameters of fermented yoghurt-type products. tg : the time (min) at which tan␦ = 1; Elasticity Increment
IE = [(dlogG /dt)max ], defined by the slope of log G (t) at the inflection point

Control With dried extracts With dried extracts With dried extracts from With dried extracts With dried extracts from With dried extracts from With dried extracts
from 50 mg OSW / from 50 mg OSW 50 mg OSW using from 150 mg OSW / 150 mg OSW using 150 mg OSW using from 350 mg OSW
100g milk using ␤-CD / 100g milk hp-␤-CD / 100g milk 100g milk hp-␤-CD / 100g milk hp-␤-CD / 100g milk using ␤-CD / 100g milk

tan␦24h 0.27 0.27 0.27 0.27 0.28 0.27 0.265 0.26

(5°C)
0.22 0.31 0.19 0.40 0.56 0.28 0.24 0.29
IE(min−1 )
tg (min) 109 116 114 120 132 140 135 146
pH (at tg ) 5.85 5.72 5.74 5.69 5.30 5.49 5.27 5.25
Liquid Phase Separations

www.electrophoresis-journal.com
1981
1982 I. Mourtzinos et al. Electrophoresis 2018, 39, 1975–1983

3.3.2 Rheological properties of the yoghurts processes is presented in this study. The findings indicate that
it is possible to develop convenient extraction techniques of
The natural colorant extracts, incorporated into yoghurt, were
bioactive plant polyphenols, including natural pigments (e.g.
also examined for their effect on the mechanical properties
anthocyanins), with the adoption of green-extraction tech-
of the fortified fermented products. The evolution of stor-
niques and thus minimize the use of synthetic colorants de-
age modulus, tan␦ (control sample) and pH of a typical yo-
rived from petrochemicals. The onion solid waste extracts
ghurt formulation (4.3% total fat content, 3.9 % total proteins)
obtained were also successfully employed for the production
are shown in Fig. 2. All samples tested gave the characteris-
of a red coloured stable fermented dairy product of yoghurt
tic gelation kinetic profile of acidified yoghurt products. At
type.
the onset of acidification, the G value was low and the tan␦
(damping factor) was high, all indicative of a liquid-like be- The authors would like to thank Ignantiadis, A., Topalidou,
haviour. As the acidification proceeded further, an abrupt K., Koutsos, G., Karakostas, A. and Papastathis, S. for prepar-
increase of the G (and decrease of tan␦ at ⬍ 1) was noted im- ing several batches of onion extracts and final yoghurt products
plying the establishment of a three dimensional gel network and also presenting a yoghurt product prototype in the 6th Greek
structure. At the early stages of gelation, a local maximum competition of eco-innovative food products ECOTROPHELIA
in tan␦ (tan␦ peak1) was noted which has been previously organized by SEVT (Federation of Hellenic Food Industries).
attributed to a loosening of the network structure due to sol-
ubilization of the colloidal calcium phosphate, resulting in a
partial micellar disintegration, and/or transition of the milk The authors have declared no conflict of interest.
gel from a network prevailed by denatured whey protein-
casein associations to a network dominated by casein-casein
5 References
interactions [29, 37, 38]. The onset of the local increase of the
tan␦ coincided with a change in the slope of G curve, as typ- [1] Kiassos, E., Mylonaki, S., Makris, D. P., Kefalas, P., Innov.
ically observed for acidified milk products. At longer times, Food Sci. Emerg. Technol. 2009, 10, 246–252.
and as the isoelectric point of caseins is approached, the repul- [2] Slimestad, R., Fossen T., Vågen, I. M., J. Agric. Food
sions between casein particles decrease and the hydrophobic Chem. 2007, 55, 10067–10080.
interactions increase, resulting in enhanced aggregation and [3] Singh, N., Singh, B. R., Singh, R. L., Prakash, D., Singh, D.
in the participation of additional protein molecules in the P., Sarma, B. K., Upadhyay, G., Singh, H. B., Food Chem.
gel network structure. As a result, an increase of the G to Toxicol. 2009, 47, 1161–1167.
a greater extent is noted. Finally, a pseudo plateau of the G [4] Dini, I., Tenore G. C., Dini, A., Food Chem. 2008, 107,
was reached when the pH was close to the isoelectric point 613–621.
of the caseins. Upon storage at 5ο C, a further increase of [5] Chung, C., Rojanasasithara, T., Mutilangi, W., Mc-
the G (low temperature induced enhancement of gel rigid- Clements, D. J., Food Chem. 2016, 212, 596–
ity) and another local maximum in tan␦ (tan␦peak2) was ob- 603.
served (Fig. 2) which can be attributed to a loosening of the [6] Roldan, E., Sanchez-Moreno, C., de Ancos, B., Cano, M.
network structure through further solubilization of the col- P., Food Chem. 2008, 108, 907–916.
loidal calcium phosphate and dissolution of mainly ␤-casein [7] Galanakis, C. M., Trends Food Sci. Technol. 2015, 42,
due to the weaker hydrophobic interactions occurring at low 44–63.
temperature [39]. [8] Conidi, C., Cassano, A., Garcia-Castello, E., Water Res.
The incorporation of the natural colorant extracts may 2014, 48, 363–374.
also modify the mechanical properties of the formed yoghurts [9] Li, Y., Fabiano-Tixier, A. S., Tomao, V., Cravotto,
(see also Table 2). That is, at extracts derived from 50 mg dried G., Chemat, F., Ultrason. Sonochem. 2013, 20,
12–18.
OSW/100 g milk, no considerable changes have been noted
in the evolution of the storage modulus (G ) during fermen- [10] Philippi, K., Tsamandouras, N., Grigorakis, S., Makris, D.
P., Environ. Process 2016, 3, 369.
tation. However, at higher natural extract concentrations, the
gelation time was shifted to longer times and the maximum [11] Kurkov, S. V., Loftsson, T., Int. J. Pharm. 2013, 453,
167–180.
storage modulus values reached were substantially lower, in-
dicating a weaker milk protein network. Such responses in [12] Aytac, Z., Kusku, S. I., Durgun, E., Uyar, T., Mater. Sci.
Eng. C 2016, 63, 231–239.
the mechanical properties can be attributed to the effect of
the natural extract containing polyphenols on the starter cul- [13] Mourtzinos, I., Salta, F., Yannakopoulou, K., Chiou,
A., Karathanos, V. T., J. Agric. Food Chem. 2007, 55,
ture activity. As a result, the final yoghurt pH was lower after
8088–8094.
5 hours leading to lower storage modulus values (Table 2).
[14] Fang, Z., Bhandari, B., Trends Food Sci. Tech. 2010, 21,
510–523.
[15] Kyriakidou, K., Mourtzinos, I., Biliaderis, C. G., Makris, D.
4 Concluding remarks
P., Environments 2016, 3, 3: 1–11.
A novel approach for more efficient extraction of polyphe- [16] Del Valle, E. M. M., Process Biochem. 2004, 39,
nols from onion wastes, leading to eco-friendly extracts and 1033–1046.


C 2018 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.electrophoresis-journal.com
Electrophoresis 2018, 39, 1975–1983 Liquid Phase Separations 1983

[17] Helal, A., Tagliazucchi, D., Food Sci. Technol. 2018, 89, [29] Lazaridou, A., Serafeimidou, A., Biliaderis, C. G.,
164–170. Moschakis, T., Tzanetakis, N., Food Hydrocoll. 2014, 39,
[18] Bilek, S.E., Yılmaz, F. M., Özkan, G., Food Bioprod. Pro- 204–214.
cess. 2017, 102, 72–80. [30] Makris, D., J. Food Technol. 2010, 8, 183–186.
[19] Sigurdson, G.T., Tang, P., Giusti, M. M., Annu. Rev. Food [31] Jang, M., Asnin, L., Nile, S. H., Keum, Y. S., Kim, H. Y.,
Sci. Technol. 2017, 8, 261–280. Park, S. W., Inter. J. Food Sci. Technol. 2012, 48, 24,6–252.
[20] Mourtzinos, I., Makris, D. P., Yannakopoulou, K., [32] Mourtzinos, I., Makris, D. P., Yannakopoulou, K.,
Kalogeropoulos, N., Michali, I., Karathanos V.T., J. Agric. Kalogeropoulos, N., Michali, I., Karathanos, V. T., Agric.
Food Chem. 2008, 56, 10303–10310. Food Chem. 2008, 56, 10303–10310.
[21] Balasundram, N., Sundram, K., Samman, S., Food [33] Mourtzinos, I., Anastasopoulou, E., Petrou, A., Grigo-
Chem. 2006, 99, 191–203. rakis, S., Makris, D., Biliaderis, C. G., J. Food Sci. Technol.
[22] Kırca, A., Özkan, M., Cemeroglu, B., Food Chem. 2006, 2016, 53, 3939–3947.
97, 598–605. [34] Chouchouli, V., Kalogeropoulos, N., Konteles, S. J.,
[23] Bassil, D., Makris, D. P., Kefalas P., Food Res. Int. 2005, Karvela, E., Makris, D. P., Karathanos V. T., LWT-Food
38, 395–402 Sci. Technol. 2013, 53, 522–529.
[24] Makris, D. P., Kefalas P., Int. J. Waste Resour. 2013, 3, 126. [35] Oliveira, A., Alexandre, E. M. C., Coelho, M., Lopes, C.,
Almeida, D. P. F., Pintado M., Food Chem. 2015, 171,
[25] Khiari, Z., Makris, D. P., Kefalas, P., Food Sci. Technol. Int.
370–378.
2008, 14, 497–502.
[36] Illupapalayam V. V., Smith, S. C., Gamlath, S., Food Sci.
[26] Katsampa, P., Valsamedou, E., Grigorakis, S., Makris, D.
Technol. 2014, 55, 255–262.
P., Ind. Crops Prod. 2015, 77, 535–543.
[37] Biliaderis, C. G., Khan, M. M., Blank, G., Int. Dairy J. 1992,
[27] Shehata, E., Grigorakis S., Loupassaki, S., Makris, D. P.,
2, 311–323.
Sep. Purif. Technol. 2015, 149, 462–469.
[38] Lucey, J. A. J. Dairy Sci. 2002, 85, 281–294.
[28] Dimakopoulou-Papazoglou, D., Katsanidis, E., Food Bio-
process Technol. 2017, 10, 2034–2045. [39] Moschakis, T., Dergiade, I., Lazaridou, A., Biliaderis, C.
G., Katsanidis, E., J. Funct. Foods 2017, 33, 386–395.


C 2018 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.electrophoresis-journal.com