Journal of Functional Foods 42 (2018) 58–74

Contents lists available at ScienceDirect

Journal of Functional Foods

journal homepage: www.elsevier.com/locate/jff

Whey protein hydrolysates as a source of bioactive peptides for functional T

foods – Biotechnological facilitation of industrial scale-up
⁎
Anja Dulliusa, Márcia Inês Goettertb, Claucia Fernanda Volken de Souzaa,
a
Laboratório de Biotecnologia de Alimentos, Programa de Pós-Graduação em Biotecnologia, Universidade do Vale do Taquari - Univates, Lajeado, RS, Brazil
b
Laboratório de Cultura de Células, Programa de Pós-Graduação em Biotecnologia, Universidade do Vale do Taquari - Univates, Lajeado, RS, Brazil

A R T I C L E I N F O A B S T R A C T

Keywords: Whey proteins, which possess the highest nutritional quality of all food proteins, are an optimal source of
Functional foods functional food ingredients. Enzymatic hydrolysis of whey proteins liberates fragments that can promote health
Bioactive peptides benefits in the immune, cardiovascular, nervous and gastrointestinal systems. Industrial production of peptide-
Whey proteins based food ingredients requires overcoming several challenges in product development to achieve economically
Downstream process
viable downstream processes. We suggest hybrid strategies currently utilized in biopharmacy that are based on
Industrial scale-up
computational modelling combined with heuristics and mechanistic modelling, thus minimizing the time- and
Peptide enrichment techniques
cost-intensive trial and error approach. Additionally, we propose an application of a cost performance indicator
during early decisions at the laboratory scale, which can lead to optimal downstream processing of bioactive,
peptide-based food ingredients. This review summarizes the requirements of industrial processes regarding
peptide release and stability, depending on several process parameters, and considers some enrichment tech-
niques for whey-derived peptides that are potentially applicable to industry.

1. Introduction Maeda-Yamamoto (in press) provided circumstantial information about

Functional foods (also called nutraceuticals, pharmafoods and de-
signer foods) are made from natural ingredients and offer, in addition to
their nutritional value, a specific health advantage due to the functional
in vivo activity of a supplemented food additive (McIntosh et al., 1998).
Japan was the first country to introduce so-called Foods for Specified
Health Uses (FOSHU) into the market in 1991 (Yamada, Sato-Mito,
Nagata, & Umegaki, 2008). These types of functional foods are divided
into three categories: (1) standardized FOSHU, (2) reduction of disease
risk FOSHU, and (3) qualified FOSHU. Currently, these categories are
based on evidence of eleven different health claims, e.g., gastro-
intestinal regulation, lower blood cholesterol, glucose level reduction or
dental health (Maeda-Yamamoto, in press). Yamada et al. (2008) and
the necessary documentation for the approval of products as FOSHU by
the Japanese Office of Health Policy on Newly Developed Foods,
highlighting that legal recognition as FOSHU requires a detailed review
of scientific evidence, including human clinical studies, animal studies,
and in vitro metabolic and biochemical data.
Food proteins are suggested to be a high-quality source of protein
because they are a natural ingredient in functional foods. Typical food
proteins derived from plant and animals are meat, milk, egg, fish, soy,
and wheat, amongst others. Whey proteins have the highest nutritional
quality (Protein Digestibility-Corrected Amino Acid Score
(PDCAAS) = 1) of the dietary protein sources and even greater biolo-
gical value than the milk protein casein. Whey proteins are rich in
human-essential amino acids (AAs), branched-chain amino acids

Abbreviations: FOSHU, Foods for Specified Health Uses; PDCAAS, protein digestibility corrected amino acid score; BCAAs, branched-chain amino acids; β-LG, β-lactoglobulin; TMP, total
milk protein; α-LA, α-lactalbumin; BSA, bovine serum albumin; BLF, bovine lactoferrin; GMP, glycomacropeptides; IgG, immunoglobulin G; BOD, biological oxygen demand; COD,
chemical oxygen demand; ACE, angiotensin-converting enzyme; DH, degree of hydrolysis; WPI, whey protein isolate; WPC, whey protein concentrate; BLP, Bacillus licheniformis protease;
UHPH, ultrahigh-pressure homogenization; STTT, short-time thermal treatment; US, ultrasonication; MW, microwave; VIP, virtual intermediate peptide; GIT, gastrointestinal tract; CPP,
critical process parameters; DPP-IV, dipeptidylpeptidase IV; BU, binding unit; SU, stimulating unit; SUMO, small ubiquitin-related modifier; LAB, lactic acid bacteria; NSLAB, non-starter
lactic acid bacteria; CCP, corn cob powder; PES, polyethersulfone; HPLC, high-pressure liquid chromatography; MS, mass spectrometry; FDA, Food and Drug Administration; AAs, amino
acids; ACPs, anticancer peptides; NK cells, natural killer cells; DoE, Design of Experiments; QSAR, quantitative structureactivity relationships; AMP, antimicrobial peptide; SVM, support
vector machines; RF, random forest; ANN, artificial neuronal networks; DA, discriminant analysis; MBPD, Milk Bioactive Peptide Database; CAMP, Collection of Anti-Microbial Peptides;
DPP-4, dipeptidyl peptidase-4; PPI, purification performance index; SCI, separation cost indicator; QSPR, quantitative structureproperty relations; SDS, sodium dodecyl sulphate; AC,
activated carbon; EDUF, electrodialysis-ultrafiltration; CFEMF, cross-flow electro-membrane filtration; MC, membrane chromatography; MAC, membrane adsorption chromatography;
MWCO, molecular weight cut-off; IMER, immobilized enzyme reactors; EOPO Co-polymer (UCON)/Phosphate, poly(ethylene glycol-ran-propylene glycol) monobutyl ether/phosphate
⁎
Corresponding author.
E-mail address: claucia@univates.br (C.F.V. de Souza).

https://doi.org/10.1016/j.jff.2017.12.063
Received 16 August 2017; Received in revised form 27 December 2017; Accepted 28 December 2017
Available online 10 January 2018
1756-4646/ © 2017 Elsevier Ltd. All rights reserved.
A. Dullius et al.
(BCAAs) and sulfur-containing AAs, and they promote metabolic reg-
ulation and protein folding (Smithers, 2015).
The health benefits of whey proteins may have been known for
hundreds of years or even longer, as whey was part of treatments for
illness in folk medicine (Hoffmann, 1961) and is still used by various
societies for food preservation and illness treatment (Gupta, 2012;
Pieroni & Giusti, 2008; Pieroni & Gray, 2008; Sikarwar & Kaushik,
1993). In the period from the 17th to the early 19th century, whey
became popular as a medicinal drink and an inexpensive alternative
bath ingredient (Smithers, 2015).
The precise molecular components of whey are known. In addition
to lactose, minerals and vitamins, the health benefits of whey are re-
lated to whey proteins, including enzymes and glycomacropeptide
(GMP) (Anand, Som Nath, & Chenchaiah, 2013). The main components
of bovine whey proteins are β-lactoglobulin (β-LG, 3.3 g/L total milk
protein (TMP)) and α-lactalbumin (α-LA, 1.2 g/L TMP), and compo-
nents in minor concentrations include bovine serum albumin (BSA,
0.3 g/L TMP), bovine lactoferrin (BLF, 0.1 g/L TMP), immunoglobulins
(0.5–1 g/L TMP), lactoperoxidase (0.03 g/L TMP), and proteinaceous
GMP (1.2 g/L TMP) (Korhonen, 2009a; Korhonen, 2012; Madureira,
Tavares, Gomes, Pintado, & Malcata, 2010). The enzymatic action of
chymosin releases proteinaceous GMP from milk κ-casein during
cheese-making, cleaving the C-terminal region of κ-casein between AA
residues 105 and 106 and therefore leading to the precipitation of
caseins, thus forming cheese (Thomä-Worringer, Sørensen, & López-
Fandiño, 2006). GMP is a glycophosphopeptide that is 64 AAs in length
and lacks the aromatic AAs Phe, Tyr and Trp; for this reason, it is re-
commended in medical food design (Ney & Etzel, 2017) and showed
prebiotic and anti-inflammatory effects in mice (Sawin et al., 2015).
Yadav et al. (2015) noted that in addition to their chemical function
(e.g., aiding emulsification or gelling) and nutraceutical characteristics,
whey proteins have known physiological benefits, including vitamin
binding, antibacterial activity and immunomodulatory effects, amongst
others.
Whey proteins with in vivo biological functions, such as positive
influences on the cardiovascular, digestive, endocrine, immune, and
nervous systems, are therefore optimal for use as a functional food
additive. However, many of the bioactivities of whey proteins are en-
crypted within their native protein sequences and can thus be liberated
only by protein fragmentation, e.g., through enzymatic hydrolysis
(Pihlanto-Leppälä, 2000). The natural production of these bioactive
whey peptides occurs via (I) gastrointestinal digestion of milk proteins
by digestive enzymes and (II) food processes, such as the fermentation
of milk or the ripening of cheese by lactic acid bacteria (LAB)
(Korhonen, 2009b).
A problematic aspect of whey is its increasing volume as a by-pro-
duct, which became detrimental to the environment with in-
dustrialization and expansion of the manufacture of dairy products in
the 20th century (Smithers, 2015). For the production of 1 kg of cheese,
approximately 9 kg of cheese whey accrues as an effluent (Prazeres,
Carvalho, & Rivas, 2012). Bovine whey contains 20% of the milk pro-
tein content, the entire milk lactose content (∼5%), and traces of fat
(0.1%) and mineral salts (0.46–10%). Approximately 99% of this or-
ganic matter is biodegradable, resulting in a high biochemical oxygen
demand (BOD; > 30.000 mg O2/L for sweet whey and 35.000 mg O2/L
for acid whey) and chemical oxygen demand (COD; > 60.000 mg O2/L
for sweet whey and ∼80.000 mg O2/L for acid whey). Biological
treatment of whey is difficult because of its acidic pH (3.8–6.5), low
alkalinity, and sodium, free ammonia, potassium and volatile fatty acid
contents (Prazeres et al., 2012; Smithers, 2015). Yadav et al. (2015)
reviewed possible methods to transform the proteinaceous part of ef-
fluent whey into bioprotein, functional proteins and bioactive peptides;
however, as mentioned before, nearly the entire lactose content of whey
remains in the whey permeate and thus requires further effluent post-
treatment steps. Innovative treatment technologies, such as the pro-
duction of D-tagatose from lactose as an alternate sweetener (D-tagatose
59
Journal of Functional Foods 42 (2018) 58–74
contains only 30% of the energy content of sucrose), could result in the
high-grade transformation of the effluent whey (Jayamuthunagai,
Srisowmeya, Chakravarthy, & Gautam, 2017).
The elevated nutritional value of bovine whey proteins, together
with their technological properties and distinguished health benefits in
the form of hydrolysed peptides, have transformed whey into a high-
value raw material for the food industry with particular applications in
the production of functional foods (Yadav et al., 2015). Several recent
studies have confirmed the applicability of health-promoting peptides
to functional foodstuffs: Chatterjee, Kanawjia, and Khetra (2016)
showed that the addition of a tryptic whey protein hydrolysate to In-
dian sweetened yogurt significantly increased its angiotensin-con-
verting enzyme (ACE) inhibition and antioxidant activity, indicating
the maintenance of peptide stability during food processing and within
the food matrix. Yu, Amorim, Marques, Calhau, and Pintado (2016)
reported as a lead for peptide functionality and stability that a low-
molecular-weight (< 1 kDa) whey peptide extract of a mixture of cow,
goat and ewe whey stimulated the growth of probiotic bacteria (Lac-
tobacillus acidophilus and Bifidobacterium animalis) in Wistar rats. Graves
et al. (2016) found that an anticancer pentapeptide derived from rice
bran was stable for 6 months when stored in spray-dried orange juice
and received acceptance from consumers regarding its sensory prop-
erties. Hafeez et al. (2014) proposed the use of bioactive peptides from
milk proteins, including whey proteins, to increase the functionality of
fermented milk products and reviewed the three known production
process strategies: (a) peptide release directly into fermented milk
products, (b) supplementation of fermented milk products with pep-
tides produced outside of the product and (c) production of bioactive
peptides using recombinant DNA technology.
Generally, natural digestion of whey proteins in the gastrointestinal
tract via the consumption of whey protein-containing products is not
sufficient to achieve a positive health effect; as a consequence, the
bioactive peptides responsible for the health benefits must be enriched
by industrial processing of hydrolysed whey (Gauthier, Pouliot, & Saint-
Sauveur, 2006). The ongoing development in the past two decades of
technologies for the production of whey bioactive peptide-containing
foods on the industrial scale has resulted in several products on the
international market (Hettiarachchy, Sato, Marshall, & Kannan, 2011;
Korhonen, 2009b). Table 1 lists bioactive peptides that are derived from
whey proteins and are used as ingredients in functional foods. Re-
garding the industrial production of peptide-based food ingredients,
membrane filtration techniques (ultrafiltration and nanofiltration) are
the only methods that enrich food ingredients, such as whey hydro-
lysates, with bioactive peptides in an economically viable way (Agyei,
Ongkudon, Wei, Chan, & Danquah, 2016; Korhonen, 2009b). More ef-
fective and specific enrichment of bioactive peptides, including pur-
ification, within downstream processes requires chromatography steps,
which are cost intensive and time consuming when applied using
conventional laboratory methods. These processes are expensive, dif-
ficult to apply or affect the secondary structure of peptides, resulting in
the possible alteration or elimination of their bioactive characteristics;
therefore, industrial scale-up is not economically viable for the food
industry. Thus, the lack of economically profitable production processes
decelerates the development of products that contain bioactive peptides
(Agyei & Danquah, 2011; Agyei et al., 2016). Furthermore, basic re-
search purification processes must be adapted to meet the requirements
of an economically viable process at the industrial scale (Korhonen &
Marnila, 2013). New purification technologies in the food industry are
less focused on products with high-grade, pure bioactive peptides and
are more focused on the development of economically viable enrich-
ment techniques (Agyei et al., 2016).
This review article provides a brief overview of the conventional
downstream process steps of bioactive whey peptide research devel-
oped under laboratory conditions compared to the actual flow path of
industrial production. Furthermore, we discuss current critical process
steps that can affect the release of bioactive peptide sequences and
A. Dullius et al.
Table 1
Commercially sold functional food ingredients based on bioactive peptides derived from whey proteins (Data from Ballard et al., 2009; Korhonen, 2009b; Lacroix et al., 2016).
Product name Manufacturer Type of product
BioZate® product line Davisco, USA Hydrolysed whey protein isolate
BioPureGMP™ Davisco, USA Whey protein isolate (WPI)
Vivinal® ALPHA Borculo Domo Ingredients (BDI), Ingredient/hydrolysate
the Netherlands
Praventin™ DMV International, the Food supplement/capsule
Netherlands
Dermylex™ Advitec, Inc., Canada Food supplement/ tablet
Hilmar™8390 Hilmar Ingredients, USA Food supplement
NOP-47™ Glanbia Nutritionals, USA Food supplement and
pharmaceutical applications
Abbreviations: GMP: glycomacropeptide, ACE: angiotensin-converting enzyme, DPP-IV: dipeptidylpeptidase IV.
product quality and thus remain a challenge to industrial bioactive
peptide production. Additionally, we briefly review previously pro-
posed modified basic research processes that include in silico biopros-
pecting of peptide release and activity and have thus already led to a
more systematic research process. Within this review, we propose an
integration of complementary processes enabling the simultaneous
production of bioactive peptide products, thus enhancing the efficiency
of the industrial process to optimize the value of whey as a raw mate-
rial. This technique requires a more methodical design of the processing
pathway based on hybrid methods, including in silico methods, me-
chanistic modelling and heuristics, as currently applied in the bio-
pharmaceutical industry. Finally, we review enrichment methods with
a particular emphasis on the AA profiles of whey-derived bioactive
peptides, which are promising for economically viable industrial scale-
up and the consequent more targeted production of whey hydrolysate
products.
2. Conventional downstream process and its challenges in
bioactive peptide production
The common in vitro bottom-up proteomics approach is divided into
several steps (Fig. 1A): (I) isolation of proteins or a protein mixture
from a food matrix; (II) hydrolysis of proteins by different enzymatic
proteases, namely, (a) gastrointestinal enzymes such as pepsin, trypsin
and chymotrypsin, (b) microorganism or plant proteases such as alca-
lase, bromelain and ficin, and (c) the enzymes present in microbial
fermentation by starter LAB, non-starter LAB (NSLAB), or several yeast
strains used in cheese-making; (III) preliminary bioactivity screening;
(IV) separation and purification of bioactive peptides, combined with
intermediate bioactivity tests; (V) identification of interesting bioactive
peptides; and (VI) confirmation of bioactivity by de novo peptide
synthesis and bioactivity tests in vitro or/and in vivo (Agyei et al., 2016;
Capriotti, Cavaliere, Piovesana, Samperi, & Laganà, 2016; Li-Chan,
2015).
The industrial production of whey protein hydrolysates enriched
with bioactive peptides (Fig. 1B) includes the following processing
steps: (1) fractionation of total solids to separate whey proteins from
fat, lactose and minerals; (2) enzymatic hydrolysis of whey proteins by
(a) the use of commercial enzymes from animal sources, such as pan-
creatin, trypsin, and pepsin, (b) bacterial and fungal proteases from
microbial fermentation, or (c) plant enzymes such as papain and bro-
melain; (3) post-enzymatic treatment, e.g., filtration, centrifugation and
chromatographic steps; (4) spray-drying of the product; and (5) product
packaging (Pasupuleti & Braun, 2008). These workflows (A and B in
Fig. 1) include crucial industrial processing steps that can alter or, in
the worst case, destroy the beneficial bioactivity of the peptide. These
steps are (a) the pre-treatment steps before protein hydrolysis, (b) the
protein hydrolysis per se, (c) the purification of bioactive peptides, and
(d) the spray-drying step (Agyei et al., 2016).
60
Journal of Functional Foods 42 (2018) 58–74
Peptide bioactivity Health claim
β-lactoglobulin fragments Reduction of blood pressure
GMP f(106–169) Prevention of dental caries, blood clotting,
antibacterial, antivirus
Whey-derived peptide Relaxation and sleep
Lactoferrin enriched whey Acne reduction
protein hydrolysate
Whey protein extract XP-828L Reduces symptoms of psoriasis
Whey protein hydrolysate ACE inhibition, DPP-IV enzyme inhibition
Whey derived peptide Anti-inflammatory properties
2.1. Effects of pre-treatment on peptide bioactivity
The main issue in protein hydrolysis is that well-defined secondary
structures in proteins, such as α-helices, are less susceptible to pro-
teolysis and thus lead to a low degree of hydrolysis (DH) (Tsai, de
Laureto, Fontana, & Nussinov, 2002). Pre-treatment of proteins with
heat or high pressure led α-helices to refold into β-sheets or random
coils (Ngarize, Herman, Adams, & Howell, 2004), which improves en-
doprotease access and hydrolysis (Fernández & Riera, 2013). It is well
known that alterations in environmental conditions applied for pre-
treatment, such as pH, salts, solvents, pressure and temperature, cause
protein denaturation; moreover, temperature is a favoured tool for
protein unfolding (Tavano, 2013). Cheison and Kulozik (2017) de-
scribed the heat-denaturation of whey proteins as a two-step process,
wherein unfolding firstly occurs at low denaturing temperatures
(characterized as the molten globule state), followed by the aggregation
step. Therefore, thermal protein treatment can enhance the exposure of
a peptide to the solvent, thus facilitating peptide release on the one
hand. On the other hand, peptide release can also be hindered, de-
pending on the position of the peptide within the protein, in-
tramolecular disulfide exchange reactions due to conformational
changes in non-native monomer formation, or positioning of the pep-
tide within aggregation sites. For example, the release of the ACE-in-
hibiting, bioactive peptide f (Lys9-Lys14) derived from β-LG by trypsin
(E.C. 3.4.21.4) was hindered by the formation of aggregates at acidic
pH (pH 5.1) and non-native monomers at neutral and alkaline pH (pH
6.8 and 8.0) until a DH of 5% was reached (Leeb, Götz, Letzel, Cheison,
& Kulozik, 2015). Generally, the denaturation pattern is influenced by
the composition of whey proteins (e.g., whey protein isolate (WPI),
whey protein concentrate (WPC) or pure protein) and environmental
factors, as ionic strength, pH or lactose concentration, as described and
reviewed in detail by Cheison and Kulozik (2017).
Proteolysis may also be enhanced by high pressure. β-LG, which is
resistant to peptic proteolysis under atmospheric pressure (0.1 MPa),
released high levels of peptides under high-pressure treatment
(300 MPa) (Chobert et al., 1996). Blayo, Vidcoq, Lazennec, and Dumay
(2016) showed that high pressurization (UHPH, ultrahigh-pressure
homogenization, i.e., 300 MPa of pressure followed by immediate
cooling) or thermal treatment (STTT, short-time thermal treatment,
exposure to 75 °C for 10 s) of whey proteins that were enriched in β-LG
and α-LA (∼68.5 and ∼21.5%, respectively, with other whey con-
stituents present in trace amounts) led to the highest proteolysis yields.
In contrast, Bamdad, Bark, Kwon, Suh, and Sunwoo (2017) showed that
high pressurization at 100 MPa led to a decrease in the DH to β-LG
hydrolysis using trypsin, whereas the DH increased by the use of the
enzymes neutrase, alcalase, savinase, elastase and thermolysin com-
pared to hydrolysis under atmospheric pressure. Moreover, the anti-
oxidant and immunomodulatory activities of alcalase-hydrolysed β-LG
were improved. Hydrolysis of BSA with trypsin and chymotrypsin under
high pressure at 400 MPa for 25 min also resulted in an improved DH of
A. Dullius et al.
Fig. 1. (A) Conventional downstream process in bioactive peptide research with the “trial and error” approach. (B) Industrial process scheme of whey protein hydrolysates that are
enriched with bioactive peptides.
7% and 10%, respectively, whereas high pressurization above 400 MPa
led to the formation of aggregates and thus a decrease in DH (De Maria,
Ferrari, & Maresca, 2017). These results show that beyond improving
proteolysis activity, the use of high pressurization with simultaneous
enzymatic hydrolysis can reduce industrial production costs as it
eliminates protein pre-treatment as a separate process step.
Other pre-treatment methods that do not employ heat or high
pressure include extrusion, sonication, microwave and pulsed electric
field, as mentioned by Agyei et al. (2016) and Uluko et al. (2015), but
are not further discussed in this review.
Pre-treatment of proteins before their exposure to enzymatic hy-
drolysis does not necessarily destroy their functional activity. When α-
lactalbumin was heated, its antibacterial activity increased (Expósito &
Recio, 2006). Arrutia, Puente, Riera, Menéndez, and González (2016)
showed that heat pre-treatment of BSA for 1 h over a temperature range
of 65–75 °C increased the DH and thus resulted in a greater number of
bioactive peptides. The degree of BSA denaturation was not measured
in that study, but Leeb et al. (2015) demonstrated the contrary results
that the hydrolysis of β-LG that had been denaturated close to 100% led
to reduced DH values, indicating that the degree of denaturation seems
to be an important factor influencing the hydrolysis rate and reaction
time. Uluko et al. (2015) compared the antioxidant activity of untreated
milk protein hydrolysates derived from milk protein concentrates to
that of samples pre-treated with ultrasonication (US), microwave (MW)
or heat and combinations thereof (US + MW; US + heat and
US + MW + heat). US led to a higher release of antioxidant peptides
compared to that of the control, whereas MW and heat pre-treatment
lowered the radical-scavenging activity of the hydrolysates. Combina-
tions of the pre-treatment methods did not significantly improve the
antioxidant activity compared to the single pre-treatments. Adjonu,
Doran, Torley, and Agboola (2013) also reported that heat pre-treat-
ment did not improve the antioxidant activity of whey protein hydro-
lysates. Wali et al. (2017) reported higher ACE-inhibition activity using
multifrequency ultrasound pre-treatment of rapeseed protein before
61
Journal of Functional Foods 42 (2018) 58–74
enzymatic hydrolysis. These examples indicate that different pre-
treatment methods have different effects on protein structure during
denaturation. Pre-treatment methods can favour desired protein release
or prolong it, as demonstrated by Leeb, Kulozik, and Cheison (2011)
and Leeb et al. (2015), who applied thermal heat at 80 °C as a pre-
treatment tool. In these works, different pH values caused different
particle sizes of β-LG, which signalled diverging alterations of its ter-
tiary structure due to intramolecular disulfide bridge exchange reac-
tions or aggregation, resulting in decreased variability of the peptide
pattern due to lower trypsin (E.C. 3.4.21.4) susceptibility of some
cleavage sites. These findings enable the design of a more controlled
process to release desired peptide sequences. Moreover, Delahaije,
Gruppen, van Eijk- van Boxtel, Cornacchia, and Wierenga (2016) re-
ported that the extent of aggregation could be directed by controlling
the level of protein unfolding. While increased protein concentration
led to a higher concentration of aggregates, a shift from pH 7 to pH 8
increased the native-like protein content after heat pre-treatment at
80 °C. It was also found that non-native β-LG consists of two types of
conformations, one of which has an identical structure to native β-LG.
Increased temperature led to a shift from a native-like to the non-native
conformation, whereas pH, ionic strength and protein concentration
had no influence on the ratio between native-like and non-native-like β-
LG states.
As a disadvantage of protein pre-treatments, when the secondary
structure changes from α-helix to β-sheet, aggregates of whey peptides
are commonly expected to be formed because of their hydrophobic AA
content, leading to an abrogation of bioactivity (Bulone, Martorana, &
San Biagio, 2001). Hydrolysis of whey proteins can produce amphi-
philic peptides, whose self-assembly into micelles is guided by β-sheet
interactions (Cui, Webber, & Stupp, 2010). Creusot and Gruppen (2008)
showed that peptide aggregation within a WPI that had been hydro-
lysed with a Bacillus licheniformis protease (BLP, E.C. 3.4.21.19) in-
creased as DH increased. Under certain pH conditions, self-assembled
peptide aggregation can be reversible as a result of different charge
A. Dullius et al.
distributions within the peptide (Aggeli et al., 2003). However, the
extensive hydrolysis (> 18%) of whey proteins with Alcalase® 2.4L
(E.C. 3.4.21.62) led to irreversible peptide aggregation and gel forma-
tion. Peptide aggregates were generally formed by low-molecular-
weight peptides (below 2 kDa) with a net zero charge in the pH range
from 4 to 7 (Doucet, Otter, Gauthier, & Foegeding, 2003). Cheison,
Wang, and Xu (2007) reported that WPI hydrolysates generated with
Alcalase® 2.4L and Proteinase N (IUB 3.4.24.28) were mostly composed
of peptides with a molecular mass below 1 kDa that reached a DH of
15% without any peptide aggregation.
2.2. Roles of degree of hydrolysis and peptide distribution in the protein
hydrolysis step
The DH refers to the percentage of peptide bonds cleaved during the
hydrolysis process (Adler-Nissen, 1979). Consequently, the DH de-
termines the desired end product (specifically, the peptide chain
length), which in turn influences peptide function. Thus, in the food
industry, a low DH (< 10%) is used to produce peptides with useful
physical functions, such as foaming, gelation or emulsification, whereas
peptides with biological functions require conditions with a high DH
(> 10%) (Agyei et al., 2016). In addition to DH, critical process para-
meters (CPP) that influence product quality include the starting mate-
rial quality (such as protein content, traces of other constituents, and
pH), the properties of the enzyme reaction mix (substrate specificity,
purity, number of reacting enzymes and their stability and optimal
reaction conditions) and the process conditions (concentrations of re-
action constituents, enzyme/substrate ratio, reaction conditions and
duration) (Li-Chan, 2015). Cheison and Kulozik (2017) thoroughly
described the enzymatic reactions related to whey protein hydrolysis
and reviewed the environmental influences on DH, with a special focus
on the trypsinisation of α-LA and β-LG. Under defined reaction condi-
tions in WPC hydrolysis, enzymes show a preference for certain whey
proteins, whereas others are attacked only partially or not at all.
Cheison, Leeb, Toro-Sierra, and Kulozik (2011) showed that WPI hy-
drolysis beyond the optimal conditions for trypsin (37 °C and pH 7.8)
led to the recovery of α-LA (67.87% at 25 °C and pH 8.5), while β-LG
was completely hydrolysed after 2 h. In contrast, using acid protease A
(37.5 °C and pH 3.0 for 120 min) allowed β-LG to be retained and α-LA
to be depleted (Cheison, Bor, Faraj, & Kulozik, 2012). Hence, tem-
perature and pH are powerful tools to control the hydrolysis process
and to enable targeted peptide release. For example, the use of pH >
7.5 or < 2.0 or temperature above 40 °C caused the β-LG dimer to
dissociate to the monomeric form and the protein to undergo the
Tanford transition, improving its hydrolysis. In contrast, pH and tem-
perature values out of these ranges led β-LG to self-associate to the
dimeric form, and neutral pH led to the octameric form (Cheison, Lai,
Leeb, & Kulozik, 2011). This study showed that different hydrolysis pH
and temperature values have effects beyond enzyme performance; they
change the enzyme susceptibility of the protein, which leads to an al-
teration of the peptide release pattern as a result of cleavage-resistant
peptide bonds (Cheison, Brand, Leeb, & Kulozik, 2011; Cheison, Leeb,
Letzel, & Kulozik, 2011). In this regard, selectivity as the rate at which a
protease cleaves each individual cleavage site compared to the total
rate of all cleavage sites of a protein was introduced (Butré, Sforza,
Wierenga, & Gruppen, 2015), and its dependence on pH was evidenced
using Bacillus licheniformis protease for WPI hydrolysis at 40 °C. Similar
to the findings of Cheison, Leeb, Toro-Sierra et al. (2011), the results
showed that the susceptibility of β-LG within the WPI increased with
increasing pH and independently of DH, whereas the remaining intact
α-LA decreased almost linearly with increasing DH. These results
showed that the stability of α-LA was independent of pH. The pre-
ference of β-LG was explained by its higher concentration in the WPI
and a higher Glu content, belonging to the BLP specificity, compared to
α-LA. As the selectivity of the enzyme for specific cleavage sites de-
termines the final composition of the hydrolysate, it could be shown
62
Journal of Functional Foods 42 (2018) 58–74
that the mechanism of hydrolysis strongly depended on pH, probably as
a result of changes in the charge state at the active site of the enzyme or
the presence of negative charges in the P1′ position. Therefore, de-
pending on the pH, the selectivity for some cleavage sites varied from
0% to 27%. Vorob’ev, Butré, Sforza, Wierenga, and Gruppen (2017)
verified that the release of hardly accessible cleavage sites for the en-
zyme BLP within intact β-LG could be described by a two-step de-
masking model following a non-monotonous increasing function, which
means the formation of a virtual intermediate peptide (VIP) from which
the final peptides can be formed. Thus, once a cleavage site is de-
masked, its hydrolysis rate is similar to other cleavage sites and follows
first-order reaction kinetics.
Cheison, Leeb, Letzel et al. (2011) showed that buffer type and
concentration also influenced peptide release: Tris-HCl- or NaOH-buf-
fered water resulted in similar peptide compositions derived from β-LG,
whereas potassium phosphate-buffered hydrolysis released a different
peptide pattern at higher buffer concentrations, causing dimerisation of
the β-LG and peptide bond resistance in specific regions. Le Maux,
Nongonierma, and FitzGerald (2017) tried to reach defined DH values
of 9% and 13% by hydrolysing β-LG with elastase at a constant tem-
perature and pH (50 °C and pH 7) using different enzyme:substrate
ratios (0.5, 1 and 1.5%). Despite minimal differences in the molecular
mass distributions of the peptides in all DH 9% samples (62% common
peptides) and, accordingly, in all DH 13% samples (84% common
peptides), there was no significant difference in the inhibitory activity
of DPP-IV (dipeptidylpeptidase IV, an enzyme involved in type 2 dia-
betes), indicating that reaction kinetics can be the key advantage to
reduce production costs in this special case.
The bioactivity of a protein hydrolysate also depends on the applied
enzyme, which means that enzymes differ in peptide distributions and
the chemical characteristics of the protein hydrolysates that they pro-
duce (Hudson, 1995). For instance, when fish protein from Selaroides
leptolepis was hydrolysed with alcalase, peptides with antioxidant ac-
tivities were produced at a low DH of 5%, whereas Flavourzyme 500L
hydrolysates worked better at a DH of 25% (Klompong, Benjakul,
Kantachote, & Shahidi, 2007). Additionally, bioactive peptides may
behave differently in a peptide mixture than when they are pure due to
antagonistic and/or synergistic effects. As an example, tests of the in-
hibition of individual hydrolysed WPIs showed that the α-LA hydro-
lysate had the highest DPP-IV-inhibiting activity, whereas after peptide
purification, the more active peptides originated from β-LG (Lacroix &
Li-Chan, 2013; Lacroix & Li-Chan, 2014). Li-Chan (2015) also men-
tioned that the coexistence of different proteins in WPIs and con-
formational changes in proteins during the industrial hydrolysis process
could cause different peptide activity profiles by comparing commer-
cially produced β-LG hydrolysates and research samples. Therefore, it is
critical to use industrial production conditions in the discovery of new
bioactive peptides in applied laboratory research.
2.2.1. Protein hydrolysis at the industrial scale
Several methods make the industrial production of bioactive pep-
tides technically feasible. The pharmaceutical production of bioactive
peptides can occur via the use of aggregation-promoting carriers, acidic
peptide-mediated production, de novo chemical synthesis, or the het-
erologous expression of the peptides fused to thioredoxin or a small
ubiquitin-related modifier (SUMO) in microbial strains (Aoki & Ueda,
2013). Application of the above-mentioned large-scale production
methods to the food industry is too cost-intensive and therefore not
economically viable (Aoki & Ueda, 2013; Capriotti et al., 2016; Lafarga
& Hayes, 2017).
As a result, the common method of production is large-scale enzy-
matic hydrolysis using food-grade enzymes (Lafarga & Hayes, 2017).
Agyei and Danquah (2011) suggested the use of proteases from selected
LAB strains in protein hydrolysis because of their advantages, which
include cheap cultivation and less laborious enzyme purification. The
proteolytic systems of LAB are composed of a high diversity of
A. Dullius et al.
proteases, thus offering a variety of enzymatic specificities to produce
specific bioactive peptides (Hati, Patel, Sakure, & Mandal, 2017; Liu,
Bayjanov, Renckens, Nauta, & Siezen, 2010).
The simplest method to perform large-scale hydrolysis is to use free
enzymes in a continuous process that is coupled to an ultrafiltration
membrane separation module (Korhonen & Pihlanto, 2006). The filter
system in the membrane reactor hinders the enzyme’s removal and si-
multaneously functions as the first purification step in the downstream
process. The use of immobilized enzymes offers various advantages
over conventional soluble hydrolysis, e.g., better hydrolysis control,
avoidance of enzyme autolysis activity and recycling of the im-
mobilized enzyme after the hydrolysis step to make the process eco-
nomically viable (Chmiel, 2011). A recent innovation in bioactive
peptide production is on-column-digestion of proteins on continuous-
flow immobilized enzyme reactors (IMER) as they enable an effective,
low cost and fast cleavage process, compared to conventional digestion
processes. An efficient and rapid protein digestion, together with a low
backpressure of this technique, while operating with a high flow rate,
subsequently enhance the productivity and yield of the bioactive pep-
tide production system, rendering it suitable for large-scale operations.
Siliceous monoliths possessing a tortuous tri-modal flow-through
channel structure with macropores of 20–30 µm pore size and meso-
pores of ∼20 nm and ∼3 nm are ideal to function as an enzyme support
and to guarantee the access of bulky proteinaceous substrates to the
immobilized enzymes. Thus a back-pressure of 64 kPa/cm2 could be
maintained with a flow rate of 750 mL/cm2 ∗ min, which eliminated the
necessity of a HPLC pump, making the process safer and cost-efficiently
(Szymańska et al., 2016). Mao, Černigoj, Zalokar, Štrancar, and Kulozik
(2017) compared two flow-through systems with immobilized trypsin
on a polymethacrylate monolith with a pore size of 2.1 µm and 6 µm to
free trypsin hydrolysis. The reactor system could be reused 18 times at a
flow rate of 15 mL/min and reached a higher DH (∼10%) in 4 h than
the stirred solution (∼6%), although the contact time of the im-
mobilized enzyme to its substrate was much shorter due to the autolysis
of trypsin in the free enzyme system. The peptide release pattern was
very similar, hence only the two peptides f(125–135) and f(61–69) were
missed, whereas the peptides f(41–70), f(41–69), f(101–124) and f
(92–101) were only present in the hydrolysates from IMER. However,
the high costs of common immobilization supports often decrease the
economic advantage of this method, whereas the use of agricultural
wastes can be an attractive alternative (Ittrat, Chacho, Pholprayoon,
Suttiwarayanon, & Charoenpanich, 2014). Bassan et al. (2016) im-
mobilized the enzyme trypsin (E.C. 3.4.21.4) on the agro-industrial
lignocellulose-rich waste corn cob powder (CCP) to generate bioactive
peptides from whey proteins. The use of trypsin-glyoxyl-CCP resulted in
an average DH of 23%, and more than 80% of the enzyme activity
remained after five reuse cycles.
A more controlled enzymatic hydrolysis reaction requires a constant
pH value, which is achieved by the addition of an alkali or acidic so-
lution (usually sodium hydroxide or hydrochloric acid). This so-called
pH-stat technique also controls the DH (Adler-Nissen, 1986). Non-pH-
controlled methods are interesting in industrial applications because
they avoid the addition of external additives to the reaction mixture.
Fernández and Kelly (2016) compared the pH-stat technique with free-
fall pH hydrolysis and determined that the hydrolysis of WPI with the
Bacillus protease preparation Protamex® (containing subtilisin, E.C.
3.4.21.62) via the free-fall pH technique is faster, whereas the pH-stat
technique is more suitable for WPI hydrolysis with trypsin (E.C.
3.4.21.4). The hydrolysis of WPI with papain comparing the pH-stat
and pH free-fall methods resulted in a similar DH, but the peptide
profile and physicochemical properties differed. At a constant pH of
7.0, the antioxidant activity of the whey protein hydrolysate was
higher, whereas the DPP-IV activity was not influenced by pH regula-
tion. This finding indicates that bioactive peptide production should be
controlled not only on a temporal basis but also by bioactivity tests (Le
Maux, Nongonierma, Barre, & FitzGerald, 2016), as pH has a decisive
63
Journal of Functional Foods 42 (2018) 58–74
influence on the peptide profile (Carvalho, Pessato, Fernandes, de Lima
Zollner, & Netto, 2017; Cheison, Brand et al., 2011; Cheison, Leeb,
Toro-Sierra et al., 2011).
To achieve a high DH, which is favoured for the production of
physiologically active peptides, hydrolysis times of 4–100 h are re-
quired, raising the issue of microbial contamination. While bacterio-
static or bactericidal preservatives were previously a common tech-
nique to prevent microbial contamination in industrial production
processes, newer technologies such as UV treatment, pulsed electric
field applications, or ozone are currently used. New enzyme technolo-
gies that withstand high temperatures or acidic conditions can help
eliminate optimal growth conditions for microbes, preventing process
contamination (Pasupuleti & Braun, 2008). Microfiltration steps using a
MWCO of 20 kDa reduced microbial load in WPC before hydrolysis,
whereas the peptide fractions after nanofiltration (3 kDa) were free
from bacteria, yeasts and molds (Tavares et al., 2012). Bamdad et al.
(2017) noted enzyme hydrolysis under high pressurization as a method
that has a reduced risk of microbial contamination while preserving the
shelf-life stability of the product.
The DH during the industrial production of peptides is monitored by
in-process sampling and determination of the ratio of amine nitrogen to
total nitrogen. Once the desired DH is reached, the process is stopped by
deactivating the enzymes with heat (Pasteurization). The free and total
AA contents and the molecular weight distribution of peptides in the
hydrolysate are also determined (Pasupuleti & Braun, 2008). Conesa
and FitzGerald (2013) investigated the impacts of DH and total solid
content (5–30% (w/v)) of hydrolysed WPC on the thermal inactivation
process of Corolase® PP at 60–75 °C. They found that the enzyme was
more heat sensitive at a higher DH (∼21%) than at a lower DH
(∼10%), indicating that the type of protein or peptide species present
in the solution is important. Additionally, Corolase PP® was more stable
at a higher total solid content, which was probably related to the
viscosity of the solution and the decreasing heat transfer coefficient in
solutions with increasing viscosity. It was also conjectured that the
bond of an enzyme to a cleavage site of intact protein has a thermo-
protective effect. This theory implies that a low total solid content
combined with a high DH results in more effective enzyme inactivation
by the application of heat, which is an interesting factor for the food
industry regarding energy costs and product stability. Mohan,
Udechukwu, Rajendran, and Udenigwe (2015) mentioned that in-
activation of the protease is likely more efficient by application of heat
than cold, whereas heat inactivation can result in N-terminal cyclisation
of the peptide. Nonetheless, cold-inactivated papain hydrolysate
showed higher ferric-reducing capacity than heat-inactivated hydro-
lysates after 5 h.
2.3. Further demands on product quality
The promised physiological effects of valuable peptide-based pro-
ducts must be guaranteed; however, more challenges regarding peptide
characteristics must still be overcome to deliver these effects. Thus,
another challenge in the large-scale production of bioactive peptides is
the stability of the peptides during downstream processing because
peptides must survive high pressures and temperatures as well as pH
variations (Lafarga & Hayes, 2017). A change in processing conditions
can lead to structural modifications and alter or abrogate the physio-
logical function of the peptide. Acidic treatments target Gln and Asn,
whereas basic conditions affect Cys, Ser and Thr, leading to the pro-
duction of lysinoalanine and D-AAs and decreasing the bioavailability
of Lys, which undergoes a Maillard reaction (Korhonen, Pihlanto-
Leppäla, Rantamäki, & Tupasela, 1998). Mohan, Udechukwu et al.
(2015) observed that 1% remaining lactose and other sugars under
hydrolysis conditions led to a Maillard reaction in papain and alcalase
reactions, which influenced the ferric-reducing capacity and SH content
of peptides, whereas the latter had no influence on the antioxidant
bioactivity. The increase in ferric-reducing capacity was related not
A. Dullius et al.
only to whey protein hydrolysis but also to Maillard reaction products,
the occurrence of which depends on protease specificity and tempera-
ture/pH conditions during the reaction process. Deng, Wierenga,
Schols, Sforza, and Gruppen (2017) found that lysine/arginine-specific
proteases such as trypsin or α-chymotrypsin were unable to hydrolyse
glycated cleavage sites or cleavage sites near glycation sites in α-LA,
whereas hydrolysis with BLP and subtilisin A was independent of gly-
cation. This means that for Lys/Arg- dependent proteases, a modifica-
tion of the protein primary structure has a direct influence on selectivity
and consequently on the extent of hydrolysis, but not on the rate of
hydrolysis.
A further problem for peptide stability occurs when bioactive pep-
tides have destinations inside the human body. The passage through the
gastrointestinal tract (GIT) can strongly hinder their effectiveness,
particularly long-chain peptides with a certain conformation, because
of the harsh pH conditions and presence of the GIT enzymes (Bouglé &
Bouhallab, 2017). Regarding short peptides, García-Mora et al. (2017)
reported that the simulated gastrointestinal digestion of lentil-derived
peptides generally enhanced antioxidant and ACE-inhibiting activity,
resulting in the additional release of smaller peptide fragments with
synergistic biological effects. A study by García-Tejedor et al. (2014)
showed that lactoferrin-derived peptides were unstable under in vitro
GIT conditions, but ACE-inhibiting activity in vivo remained in spon-
taneously hypertensive rats. For the long bovine peptide lactoferricin,
stabilisation of its tertiary structure by a sulfur bridge is crucial for its
antitumoral activity but not for its antibacterial effect (Eliassen et al.,
2002). Naturally phosphorylated or glycosylated peptides such as
casein-derived peptides or GMP are protected from hydrolysis by di-
gestive proteases and have an improved binding capacity, thus pre-
serving their in vivo activity (Bouglé & Bouhallab, 2017). Recently,
Sanchón et al. (2018) found that in vitro GIT-simulated digestion of
casein and whey proteins could be considered a good approximation of
in vivo digestion, although the peptide release in vitro was much lower
than in vivo, which was probably related to suboptimal enzyme con-
centration or a lack of enzyme adjuvants in the in vitro digestion. It was
also highlighted that whey peptides containing proline or negative
charges were resistant to gastrointestinal enzyme digestion. Further-
more, the hypocholesterolemic peptides I71IAEK75 and G9LDIQK15 and
the DPP-IV-inhibiting peptide I78PAVF82 derived from β-LG were
identified. These findings are an interesting basis for peptide ex-
amination using simulated GIT conditions. Peptides with Pro or hy-
droxyproline at their C-terminus are generally resistant to digestive
degradation (Vermeirssen, Van Camp, & Verstraete, 2004). PEGylation
of the peptide sequence is an inexpensive and FDA (Food and Drug
Administration)-approved modification that does not influence peptide
function and may be used to overcome these stability challenges in
bioactive peptide production. Additionally, PEGylation reduces the
antigenicity of α-LA and β-LG sequences (Roberts, Bentley, & Harris,
2012; Zhong et al., 2016).
Another challenge is bitterness, which relates to a high DH as a
desired tool for the production of bioactive peptides that provide health
benefits. Ney (1971) related peptide bitterness to peptide hydro-
phobicity, indicating that peptides with a Q value (a measure of the free
energy (ΔG) needed to transfer peptides from ethanol to water) greater
than 1400 cal/mole were bitter, whereas peptides with a Q value lower
than 1300 cal/mole were predicted to not be bitter. However, hydro-
phobicity alone cannot be used to predict peptide bitterness, as there
are peptides with a Q value above 1400 cal/mole that not taste bitter,
thus Lafarga, Rai, O'Connor, and Hayes (2016) suggest the use of sen-
sory evaluation panels. Maehashi and Huang (2009) reported that bitter
peptides have up to eight AA residues that form a spherical shape and
contain a BU (binding unit for bitter taste receptor binding) site and an
SU (stimulating unit) site. AAs that can occupy the BU site are Arg, Phe,
His and Pro. The SU site needs a bulky, basic group or a hydrophobic
group at the N-terminus. Aluko (2017) reviewed in detail the structural
characteristics of bitter peptides derived from food.
64
Journal of Functional Foods 42 (2018) 58–74
To address the challenge of bitterness, the encapsulation of whey
protein hydrolysates comprising bioactive peptides by spray-drying
utilizing a carrier system of proteins, polysaccharides or lipids is a
widely applied method. Through this method, other product quality
challenges such as hygroscopicity, stability within the food matrix or
during GIT passage, or bioavailability can be resolved at the same time
(Mohan, Rajendran, He, Bazinet, & Udenigwe, 2015). In bioactive
peptide research or product development, it is crucial to demonstrate
biostability, which can be affected by the spray-drying process. How-
ever, studies regarding the bioactivity of spray-dried bioactive whey
peptides are still scarce. Microencapsulation of an immunomodulatory
whey protein hydrolysate with WPC or a WPC/sodium alginate mixture
by spray-drying produced a significantly lower bitter taste. The results
of this study evidenced that spray-drying preserved the im-
munomodulatory activity of the whey protein hydrolysate and there-
fore is a good strategy for the industrial production of encapsulated
non-bitter peptides due to its low processing costs (Ma et al., 2014).
3. Purification of bioactive peptides at the industrial scale and its
challenges
After the hydrolysis step, peptides with different functionalities
must be separated and enriched to achieve high-yield products that may
be saleable as bioactive food ingredients. The main challenges in that
process step are the elimination of so-called contaminant peptides,
which can have antagonistic effects, and the enrichment of all the de-
sired peptides with special physiological features and synergistic ef-
fects. Peptides can be distinguished by selective criteria, such as charge,
size, solubility, and hydropathicity (Fernández, Zhu, FitzGerald, &
Riera, 2014). Whey proteins contain higher proportions of essential and
sulfur-containing AAs than caseins (whey proteins: 609 mg/g essential
AAs, 52 mg/g sulfur-containing AAs; caseins: 511 mg/g essential AAs,
32 mg/g sulfur-containing AAs) (Anand et al., 2013). Generally, caseins
have a higher content of Gln, Pro and Arg, whereas whey proteins are
rich in Leu, Ile and Val (BCAAs) (Da Silva, Bigo, Barbier, & Rudkowska,
2017). In recent decades, it has become clear how primary and sec-
ondary structural characteristics are related to peptide bioactivity
(Eliassen et al., 2002; Udechukwu, Downey, & Udenigwe, 2018), al-
lowing bioactive peptides to be classified into species based on their
physiological activity and specific physicochemical characteristics.
3.1. Bioactive whey peptide primary and secondary structural
characteristics
Many studies have shown that peptides from enzymatic hydrolysed
whey proteins possess a variety of physiological functions, such as an-
tioxidant, antimicrobial (Sah, Vasiljevic, McKechnie, & Donkor, 2017),
antihypertensive (Bhat, Kumar, & Bhat, 2017), anticancer (Hernández-
Ledesma & Hsieh, 2017), opioid (Garg, Nurgali, & Kumar Mishra,
2016), and immunomodulatory functions (Santiago-López, Hernández-
Mendoza, Vallejo-Cordoba, Mata-Haro, & González-Córdova, 2016),
which makes them valuable for use in functional food production
(Table 2). The bioactivity of a given peptide sequence can also have
dual or more functions as it is known for the well-researched multi-
functional peptide lactoferricin (Aoki & Ueda, 2013), and other food
derived peptides (García-Mora et al., 2017; Lacroix, Meng, Cheung, &
Li-Chan, 2016; Pandey, Kapila, & Kapila, 2018).
Bioactive peptides are typically 2–20 AAs in length and are classi-
fied as (a) short peptides, which have fewer than 7 AAs; (b) moderate-
length peptides (7–25 AAs) and (c) large peptides, which have more
than 25 AAs (Panchaud, Affolter, & Kussmann, 2012). Pihlanto (2006)
reported that dairy-derived peptides with antioxidant activity usually
contain 5–11 AAs, including hydrophobic residues such as Pro, His, Tyr
or Trp. Whey protein fractions containing aromatic AAs have stronger
antioxidant potential than simple AAs (Peña-Ramos, Xiong, & Arteaga,
2004). Proper AA content and correct positioning within the sequence
A. Dullius et al. Journal of Functional Foods 42 (2018) 58–74

Table 2
Bioactive peptide sequences derived from bovine whey proteins with antioxidant, antimicrobial, antihypertensive, anticancer, opioid or immunomodulatory activity.

Substrate Enzyme Amino acid sequence Activity Reference

β-LG Corolase PP W19YSLAMAASDI29 Antioxidant Hernández-Ledesma, Dávalos,

Bartolomé, and Amigo (2005)
β-LG Thermo-lysin L58QKW61 Antioxidant ACE inhibition del Mar Contreras, Hernández-
L59DTDYKK101 Ledesma, Amigo, Martín-Álvarez, and
Recio (2011)
β-LG Trypsin V15AGTWY20 Antibacterial against Gram- Pellegrini, Dettling, Thomas, and
A25ASDISLLDAQSAPLR40 positive bacteria Hunziker (2001)
I78PAVFK83
V92LVLDTDYK100
β-LG Chymo-trypsin H146IRL149 Opioid activity Pihlanto-Leppälä, Paakkari, Rinta-
Koski, and Antila (1997)
β-LG Trypsin L22AMA25 ACE inhibition Pihlanto-Leppälä et al. (2000) and
L32DAQSAPLR40 Mullally, Meisel, and FitzGerald
V81FK83 (1997)
A142LPMIHR148
β-LG Corolase® PP MAA ACE inhibition O'Keeffe et al. (2017)
VAGT
Lacto-ferrin Pepsin Lactoferricin: F17KCRRWQWRMKKLGAPSITCVRRAF41 Antimicrobial, Antitumoral Eliassen et al. (2002)
α-LA Protease from flower of K16GYGGVSLPEW26 ACE inhibition Tavares et al. (2011)
β-LG Cynara cardunculus D97KVGINYW104
D33AQSAPLRVY42
α-LA Synthetic f(18–20) Tyr-Gly-Tyr Lymphozyte proliferation Gauthier et al. (2006)
f(50–51) Tyr-Gly
BSA Proteinase K F221P222 ACE inhibition Hernández-Ledesma et al. (2011)
Whey protein Alcalase® 2.4L Trp-Tyr-Ser-Leu Antioxidant Zhang, Wu, Ling, and Lu (2013)
concentrate

Abbreviations: β-LG: β-lactoglobulin, ACE: angiotensin-converting enzyme, α-LA: α-lactalbumin, BSA: bovine serum albumin.

can be crucial to the antioxidant activity of peptides. Thus, Chen,
Muramoto, and Yamauchi (1995) found a general sequence that pos-
sesses high antioxidant characteristics; an N-terminal Val or Leu and
Pro, His and Tyr are included within the sequence. Similarly, Elias,
Kellerby, and Decker (2008) concluded that antioxidant peptides often
contain hydrophobic AAs, including Val or Leu, at the N-terminus.
Sarmadi and Ismail (2010) also reported that antioxidant properties are
related to composition, hydrophobicity and whether AA sequences
contain Tyr, Trp, Met, Lys, Cys, and His. In whey protein derived
peptides, the Cys, His, Tyr, Lys, Val and Phe content was significantly
higher as in the WPI hydrolysate, indicating that these AA are crucial
for antioxidant activity (Jiang, Zhang, Yuan, Qu, & Feng, 2017).
Antibacterial peptide sequences are very diverse, although most
sequences have similar AA composition, net charge and hydrophobicity
(Akalin, 2014). Most antimicrobial peptides display physicochemical
similarities, including small molecular size (10–50 AAs) and cationic
and amphipathic characteristics, that are essential for activity
(Madureira et al., 2010; Théolier, Fliss, Jean, & Hammami, 2014). In
addition, anionic peptides can contribute to the amphiphilic nature of a
molecule (Li et al., 2016). Mohanty et al. (2016) reported that anti-
bacterial peptides can have a length up to 100 AA with amphipathic
character, possessing a positive net charge of +2 to +9 and exist in
cationic or anionic forms. Dziuba and Dziuba (2014) and Sah et al.
(2017) described three subclasses of cationic antibacterial peptides: (1)
linear cationic α-helical peptides, (2) cationic peptides that contain Pro
and Arg and are mainly linear, although some may show extended coils
and (3) peptides containing Cys residues that form disulfide linkages
and stable β-sheets. Milk protein-derived antimicrobial peptides con-
tain higher Lys, Leu, Val, and Pro contents (Dziuba & Dziuba, 2014),
whereas Arg, Lys, Ala, Cys, and Leu dominate various positions in an-
tibacterial peptides obtained from whey proteins (Sah et al., 2017).
Anticancer peptides (ACPs) have a length of 5–30 AAs and a cationic
charge (Chen, Ding, Feng, Lin, & Chou, 2016). Specifically, membra-
nolytic ACPs are very similar to antimicrobial peptides and thus share a
related mode of action. They may act using a variety of models: the
barrel-stave pore, the toroidal pore, the carpet, the detergent-like and
the Shai-Huang-Matzusaki models have been discussed. The most
prevalent membranolytic ACPs (∼20%) are α-helical with a kink in
their helical axis. They generally have a length of 10–30 AA residues,
are amphipathic, have a cationic net charge between +2 and +9 and
have 40–60% hydrophobic AA content. ACPs are abundant in Arg and
Lys, whereas Lys is overrepresented in toxic peptides. The high hy-
drophobicity on the apolar face of ACPs leads to greater peptide effi-
cacy and haemolytic activity. Active α-ACPs are rich in Ala, Gly, Ile, Lys
and Leu residues, whereas Lys is the preferred positively charged AA
(Gabernet, Müller, Hiss, & Schneider, 2016). A well-described anti-
tumoural whey-derived peptide is lactoferricin B (LfcinB, 25 AAs in
length, residues 17–41), which is obtained from bovine lactoferrin. The
secondary structure consists of an amphipathic α-helix (residues
17–28), a turn (residues 29 and 30) and a β-strand (residues 31–41).
Together, these form a cyclic structure that is fixed by a disulfide bridge
between Cys19 and Cys36 (Eliassen et al., 2002).
ACE (peptidyl dipeptide hydrolase, E.C. 3.4.15.1)-inhibiting pep-
tides are antihypertensive peptides obtained from whey proteins α-LA
and β-LG and are known as lactokinins (Madureira et al., 2010). Gen-
erally, ACE-inhibiting peptides have sequences between 2 and 12 AAs
in length (Hernández-Ledesma, del Mar Contreras, & Recio, 2011).
Pihlanto-Leppälä, Koskinen, Piilola, Tupasela, and Korhonen (2000)
reported that the low-molecular-weight (< 1 kDa) fraction from pepsin-
hydrolysed α-LA and β-LG inhibited ACE more strongly than fractions
containing peptides with higher molecular weights. When aromatic
AAs, such as Trp, Tyr, Phe and Pro, are located in the C-terminus, they
can interact with many subsites of ACE and thus inhibit the enzyme
activity (García-Mora et al., 2017). AA residues, such as Leu, Ile, Val,
Lys and Arg, have also been evidenced to play an important role in
significantly increasing ACE inhibition (Pan, Cao, Guo, & Zhao, 2012;
Pina & Roque, 2009). Peptide conformation is expected to play a sig-
nificant role in ACE inhibition by long-chain peptides (Hernández-
Ledesma et al., 2011).
Peptides with opioid activities, have an affinity for opiate receptors
because the presence of the aromatic AA Tyr at the N-terminus and a
Phe at the third or fourth position leads to a structural motif that fits
into the binding site of opioid receptors (Meisel, 1997). Well-known
opioid-like sequences generated from bovine whey proteins include the

65
A. Dullius et al.
α- and β-lactorphins from α-LA f(50–53) and β-LG f(102–105), stimu-
lating gene expression and mucin secretion in intestinal cells (Martínez-
Maqueda, Miralles, Ramos, & Recio, 2013; Martínez-Maqueda et al.,
2012). Typical opioid peptides possess the N-terminal sequence Tyr-
Gly-Gly-Phe, whereas atypical opioid peptides have only a conserved N-
terminal Tyr (sequences: Tyr-X-Phe; Tyr-X1-X2-Phe) (Pihlanto-Leppälä,
2000).
Casein and whey proteins are the most important protein sources of
immunomodulatory proteins (Santiago-López et al., 2016). The im-
munomodulatory effects whey derived peptides are related to the AA
content, sequence and length of the peptides. Thus, the im-
munomodulatory effect of the peptide lactoferricin is related to its
positively charged region (Vogel et al., 2002). Acidic and large peptides
derived from a tryptic β-LG hydrolysate increased the secretion of IFN-γ
(a TH1 cytonkine), whereas short peptides of the hydrolysate enhanced
TNF-α production of monocytes (Rodríguez-Carrio, Fernández, Riera, &
Suárez, 2014). It was reported that GMP also has immunomodulatory
effects that can be enhanced by digestion with pepsin (E.C. 3.4.23.1) (Li
& Mine, 2004). Short, hydrophobic peptide sequences of 2–12 AA
containing aromatic and charged AA are able to interact with the im-
mune system; however the exact mechanism of interaction is still un-
clear, but may be by binding to receptors on the surface of immune
cells, thus causing immune responses (Wu et al., 2016). Bamdad et al.
(2017) found that 38% of anti-inflammatory peptide sequences derived
from β-LG hydrolysis with alcalase consisted of aromatic or hydro-
phobic residues. All of these sequences had a Leu, Phe, or both at the C-
terminus.
3.2. Process development challenges: up-scaling downstream processes of
whey peptides
Current industrial peptide separation techniques are membrane se-
paration methods, such as ultrafiltration and nanofiltration.
Chromatography methods are mainly applied in the pharmaceutical
industry (Agyei & Danquah, 2011; Capriotti et al., 2016; Korhonen,
2012; Lafarga & Hayes, 2017; Li-Chan, 2015). Peptide separation via
membrane technologies is a low-cost method that preserves the func-
tional properties of the peptide and is easy to scale-up. Thus it could be
shown at pilot-scale, that hydrolysis of a WPC with less conventional
Cynara cardunculus enzymes and subsequent fractionation using ul-
trafiltration (20 kDa) and further fractionation of the permeate with
nanofiltration (3 kDa) led to a peptide concentrate below 3 kDa (43%
w/w protein), which possessed ACE inhibition activity (Tavares et al.,
2012). Mundi and Aluko (2014) recommended the use of a series-
connected membrane ultrafiltration and nanofiltration system, e.g.,
fractionating the hydrolysate with a series of 1–10 kDa cut-off mem-
branes. O'Loughlin et al. (2014) tested fractionation of heat-pre-treated
WPI (80 °C for 10 min) to separate ACE inhibition and ferrous chelating
activity peptides after enzymatic hydrolysis with Corolase® PP on a
pilot scale and compared the process to fractionation without pre-
treated, hydrolysed WPI. Therefore, a series of membrane filtration
with a molecular weight cut-off (MWCO) range of 1 kDa, 5 kDa, 10 kDa
and 30 kDa was used, showing the highest bioactivity in the 1-kDa
permeate fraction of pre-heated WPI. However, to fractionate the heat
pre-treated hydrolysate, the 30-kDa membrane had to be changed to a
0.14-µm membrane due to the high level of insoluble aggregated ma-
terial, but in spite of everything, the work findings of laboratory ex-
periments could be validated at pilot scale. Further investigations of the
1-kDa permeate of heat pre-treated WPI hydrolysate and the 30-kDa
retentate of unheated WPI hydrolysate resulted in an increased iron
solubility of 72% in the presence of the 1-kDa permeate (high content of
the hydrophilic amino acids Arg, Lys, and His), which was maintained
to about 98% during simulated GIT digestion (O’Loughlin, Kelly,
Murray, FitzGerald, & Brodkorb, 2015).
At a semi-pilot scale ACE inhibiting and antioxidant peptides from
WPC hydrolysed with Corolase® PP could be obtained after
66
Journal of Functional Foods 42 (2018) 58–74
ultrafiltration with cut-off values of 0.14 µm and a 5 kDa. Further la-
boratory-scale membrane processing using a 1-kDa and, subsequently, a
0.65-kDa membrane resulted in a 4.4-fold increase in antioxidant ac-
tivity and a 1.3-fold increase in ACE-inhibitory activity. As the most
potent novel peptides could be identified the sequences MAA and VAGY
showing the highest ACE inhibition and MPI with the highest anti-
oxidant activity (O'Keeffe et al., 2017).
High levels of peptide transmission can be achieved by operating at
a pH near the peptide’s isoelectric point, highlighting the importance of
the net charge and charge distribution of the peptide during filtration
process (Arrutia, Rubio, & Riera, 2016). However, the use of ultra-
filtration is not strongly recommended for the purification of small
peptides because some apolar peptides, even though their passage is
theoretically possible, are unable to cross the membrane and thus are
lost (Capriotti et al., 2015). Arrutia, Rubio et al. (2016) reported that
polyethersulfone (PES) membranes with the smallest exclusion pores
(1–5 kDa) can have very similar sieve effects by forming a polarized
layer over the membrane surface. Thus a flux reduction could be ob-
served over time and almost the same peptides crossed both mem-
branes, whereas the whey peptide sequences ALPMHIR, TKIPAVFK and
SLAMAASDISLLDAQSAPLR were absent in both permeates.
To prevent a deposite layer over the membrane, and to overcome the
low selectivity of pressure-driven membrane filtration, charged and low-
molecular weight peptides can be separated using electro membrane
technologies, as for instance electrodialysis-ultrafiltration (EDUF), which
uses electric charge and molecular mass to effect peptide separation in
one step. Compared with conventional ultrafiltration, EDUF uses an
electric field rather than pressure to drive peptides through the mem-
brane, thus the peptide separation occurs by different electrophoretic
mobility or different molecular size of the peptides. Stacking ultrafiltra-
tion membranes of different sizes (20 and 50 kDa) enabled peptide
fractionation with a higher resolution and grouping into the specific
bioactivities ACE inhibition and glucose uptake stimulation (Doyen et al.,
2014). In a recent study, the tryptic hydrolysation process of WPI rea-
lized in separately to the EDUF step (ex situ) was compared to a hydro-
lysis process with simultaneous EDUF processing (in situ) (Suwal, Rozoy,
Manenda, Doyen, & Bazinet, 2017). It could be observed, that in situ
digestions resulted in a higher migration of anionic peptides to their
recovery compartment, while ex situ hydrolysis led to a higher cationic
peptide migration. Nonetheless, independent of the digestion process,
bioactive sequences such as IDALNENK (antimicrobial) and VY-
VEELKPTPEGDLEILLQK (hypocholesterolemic) could be recovered at the
anionic recovery compartment and ALPMHIR (antihypertensive), TKI-
PAVFK (hypocholesterolemic), VLVLDTDYKK (antimicrobial), and
VAGTWY (antimicrobial) at the cationic recovery compartment. With a
regard to membrane fouling in electrodialysis methods, Persico et al.
(2016) showed, that this process is driven by charge distribution of the
peptides, which means that only specific sequences are responsible for
this undesirable effects.
A further method that applies an electric field to membrane filtra-
tion to minimize particle deposition on the membrane is cross-flow
electro-membrane filtration (CFEMF). For CFEMF, an electrical field is
superimposed upon pressure driven membrane filtration causing se-
paration effects as electroosmosis, electrophoresis, and electrolysis.
Holder, Scholz, Kulozik, and Hinrichs (2013) verified, that cathode or
anode located on the permeate side has an influence on fractionation
efficiency, which was tested using six casein derived peptides, which
are known to possess antihypertensive activity, but differ in charge,
length and hydrophobicity. With the anode on the permeate side, a
higher permeation rate for three of these six peptides was observed,
whereas those, for which a retention was expected, also permeated.
Further studies showed that in this case negatively charged peptides
were mainly affected by the electrical field, while higher transmem-
brane pressures increase the convective flow and therefore permeation
is increased. Positively charged peptides were retained only when a
lower transmembrane pressure was applied, but utilization of high
A. Dullius et al.
transmembrane pressures led the drag forces to overcome the electrical
effects resulting in an increased peptide transmission to the permeate
(Holder, Merath, Kulozik, & Hinrichs, 2015).
Compared to filtration methods, chromatographic methods separate
peptides at higher resolution, guaranteeing high-quality (very pure)
products. On the other hand, the costs of chromatographic technologies
at large scales are exorbitant and are therefore a limiting factor in the
commercialization of food ingredients based on bioactive peptides.
Purification processes resulting in high-purity products, as in the
pharmaceutical industry, account for 90% of the downstream process
costs, and even biologically produced chemicals can require more than
50% of the production costs (Agyei & Danquah, 2011). To scale la-
boratorial and pilot processes to industrial levels, parameters such as
resin material, particle size, linear flow rate, bed height and buffer
composition must remain constant. In the case of gel filtration methods,
zone broadening relative to the column length must not change,
whereas adsorptive chromatography techniques require a constant ratio
of product mass to adsorber mass. Furthermore, step gradients, rather
than linear gradients, are usually used in the laboratory to facilitate
elution. Thus, an upscaling of the developed laboratory process is often
economically not viable, and subsequent modifications of the operating
process generate highly out-of-scale costs (Chmiel, 2011). From an
economic point of view, manufacture of biotechnological products for
the food industry require different chromatography processes than the
conventional using bead-based columns due to the fact, that low flow
rates and a high backpressure lead to suboptimal enrichment processes
(Leeb et al., 2014). Table 3 gives several examples of enrichment
methods (explained in the text below) that are applicable for whey-
derived peptides and may also operate at the industrial scale in the
future.
Thus, membrane chromatography (MC) seems to be an interesting
purification process, as it is already used in industry for protein se-
paration. This technique combines ultrafiltration with chromatography
by functionalizing the filtration membrane with grafted ligands that
have special chemical features (e.g. ion exchangers) that select for the
desired proteins (Bhut, Weaver, Carter, Wickramasinghe, & Husson,
2011). Advantages of MC over conventional chromatographic methods
are its low backpressure, high flow rates and ease of scale-up (Chik &
Saufi, 2010). To this effect the use of adsorbent materials as alternatives
in peptide purification is mentioned in the literature. Membrane ad-
sorption chromatography (MAC) can be applied as a high throughput
ion exchange chromatography method, which is suited for selection
between different net charges of proteins or peptides (Voswinkel &
Kulozik, 2011). The application of a radial flow MAC, which has a spiral
wound membrane instead of stacked layers with axial flow, resulted in
more homogenous flow properties. At 50-fold pilot scale, Voswinkel
and Kulozik (2014) showed that a linear scale up by maintaining the
bed height is possible as the dynamic binding capacity was close to that
of lab scale and high flow rates of 5 bed volumes min−1 could be
maintained without cleaning at 5 repeated cycles. The scalability of
Table 3
Methods to enrich bioactive whey peptides that may be amenable to industrial scale-up.
Enrichment step Enriched peptide class
Molecular weight cut-off membrane Depends on molecular weight of peptides
filtration
Cross-flow electro membrane filtration Electrostatically charged peptides with
different molecular weight
Ion exchange adsorber membrane Discriminates between charge
chromatography
Adsorption with activated carbon Short Trp-containing peptides, BCAA-rich
peptides
Extraction with EOPO Co-polymer Peptides with hydrophobic and
(UCON)/Phosphate antioxidative AA
Abbreviations: SDS: sodium dodecyl sulphate, BCAA: branched-chain amino acid, AAs: amino acids.
67
Journal of Functional Foods 42 (2018) 58–74
radial flow MAC is interesting by the fact, that yields and purities are
independent of the flow rate, which means that maximal possible flow
rates can be used to achieve a reduction of process time and costs
(Brand, Dachmann, Pichler, Lotz, & Kulozik, 2016). Brand, Voigt,
Zochowski, and Kulozik (2016) showed that loading in recirculation
mode instead of a single pass mode increased the binding capacity of
the sample. Furthermore this fluid contacting mode prevents pore
plugging, due to a mix of convective flow close to the flow channel and
diffusive mass transport in the membrane pores compared to convective
flow alone in radial flow MAC. Additionally, mass transfer of tangential
flow MAC was tenfold higher than for batch adsorption, as it was found
by comparing the effect of temperature to the binding capacity of β-LG
in batch adsorption, radial and tangential flow using anion exchange
MAC. Increasing the temperature up to 50 °C led to higher mass transfer
but lowered the binding capacity. Thus, working at lower temperatures,
for instance, 10 °C was recommended to additionally prevent microbial
growth during protein processing (Voswinkel, Etzel, & Kulozik, 2017).
In this way, Leeb et al. (2014) obtained the ACE inhibitor peptide β-LG
(9–14) with a yield of 52% from a complex tryptic β-LG hydrolysate by
ion exchange MAC in a two-step fractionation process using a strong
anion exchange and cation exchange adsorber module in series. While
CFEMF enables the selective fractionation of a complex peptides mix-
ture into two fractions, the fractionation capacity of ion exchange MAC
is much higher, thus it produces a high amount of fractions, carrying
different properties. However, in contrast to CFEMF the ion exchange
MAC requires further treatment after peptide enrichment because of the
high salt content in the elution buffer.
As a further peptide enrichment process, column adsorption chro-
matography using activated carbon (AC) as an adsorbent enables an
enrichment of Trp-containing peptides such as Ile-Trp or Ala-Trp, which
show ACE-inhibiting activity (Hippauf, Lunow, Borchardt, Henle, &
Kaskel, 2014). For an economic up-scale and to allow higher through-
puts in a continuous system Hippauf et al. (2016) used adsorbent col-
umns loaded with granulated AC instead of powdered AC. This column
arrangement can replace expensive preparative HPLC equipment and
therefore enables large-scale production. AC is a non-toxic material that
can be produced from natural bioresources, including algae, wood
sawdust or agricultural waste such as cow manure or pulp-mill sludge.
Hydrophobic, π-π and other intermolecular interactions with the AC are
responsible for peptide adsorption.
As a chromatography free, recently developed separation method,
an aqueous two-phase system which consist of poly(ethylene glycol-
ran-propylene glycol) monobutyl ether/phosphate (EOPO Co-polymer
(UCON)/Phosphate) was applied for enrichment of antioxidant pep-
tides derived from a pepsin hydrolysate of WPI. This technique was
developed to facilitate and accelerate the enrichment step for bioactive
peptides, as antioxidant and hydrophobic AA moved to the top phase of
the two-phase system at acidic conditions (pH 4), which consequently
showed a higher antioxidant activity as peptides in the bottom phase.
The antioxidant sequences DIQKVAGTWYSL, NENKVLVLDTDYKKY and
Advantages of the method Method source
Simple and rapid method with large-scale application Mundi and Aluko
(2014)
Reduce membrane fouling High permeation flux can be maintained Holder et al.
Low costs and easy to scale up (2013)
High separation selectivity High flow rates applicable, while Leeb et al. (2014)
backpressure remains low
Non-toxic material, can be produced from bioresources, enables Hippauf et al.
large-scale production (2016)
Rapid and inexpensive method Can be specifically separate Jiang et al. (2017)
peptides with different properties Polymer can be reused Up-scaling
possible
A. Dullius et al. Journal of Functional Foods 42 (2018) 58–74

CAQKKIIAEKTKIPAVF could be identified and under optimum condi-
tions (pH 4, 4:4 vol ratio of 40% (w/w) UCON solution and 15.5% (w/
w) KH2PO4 solution, 2 mL of 49.35 mg/mL WPI hydrolysate and 0.40 g/
10 mL NaCl) the extraction efficiency was 38.95%. Because of structural
changes of EOPO, due to high sensitivity to external conditions, the
polymer can be separated from the peptides and reused, which ad-
ditionally turns the technique viable for industrial application (Jiang
et al., 2017).
4. In silico strategies to develop economically feasible
downstream processing of bioactive peptides
Experimental trial and error approaches, which reflect the conven-
tional process stream at the laboratory scale (described above), are cost-
intensive and time-consuming (Agyei et al., 2016). Application of dif-
ferent proteases results in hydrolysates with diverse peptide profiles
and subsequent bioactivities (Welsh et al., 2017). Processing conditions
in pre-treatment and hydrolysis for instance: time, temperature, pH,
pressure, buffer type or enzyme-substrate-ratio also have a crucial in-
fluence on the enzyme specificity resulting in varying peptide profiles
(Adjonu et al., 2013; Cheison, Brand et al., 2011; Cheison, Lai et al.,
2011; Cheison, Leeb, Letzel et al., 2011; Cheison, Leeb, Toro-Sierra
et al., 2011; Leeb, Kulozik, & Cheison, 2011). Thus, as mentioned by Li-
Chan (2015) it is important to keep process parameters constant during
up-scaling to guarantee a stable peptide profile and underlines the
importance to use industrial production conditions in the discovery of
new bioactive peptides in applied laboratory research.
Bioinformatic approaches are innovative and helpful tools to over-
come expensive and time-consuming experiment trials, which can lead
to a more systematic process design and invention of new
methodologies (Tong, Li, Bai, & Li, 2017; Udenigwe, 2014). Fig. 2 lists
the possible applications of bioinformatics to support the classical ap-
proach and shows research tendencies in the field of bioactive peptides
derived from whey protein hydrolysates, which are explained in the
text.
A food protein source can be selected using random sequence si-
milarity. As sequences up to 20 AA in length are too short to be eval-
uated without risk of error, the programs must be carefully chosen. As
recommended by Minkiewicz, Dziuba, Iwaniak, Dziuba, and Darewicz
(2008), MS BLAST can find short peptide precursors. Recently, a novel
platform for bioactive peptides from milk accessible under the URL
http://mbpdb.nws.oregonstate.edu, was published (Nielsen, Beverly,
Qu, & Dallas, 2017). With the help of this database, named Milk
Bioactive Peptide Database (MBPD), it will be possible to screen protein
sources for bioactive peptides, based on peptide structure activity re-
lationship data. Only peptides whose sequence, related to the bioac-
tivity, is exactly known were included to the database. Additionally, as
a new feature, these bioactive peptides were visually mapped on pro-
tein sequences derived from human or cow proteins.
Bioinformatic approaches furthermore enable the prediction of
peptide sequences by simulation of hydrolysis in silico. The bioactive
behaviour of the resulting peptide sequences can be subsequently ex-
amined using physicochemical or structural properties, such as electric
charge or hydrophobicity, searching for sequence motifs and predicting
allergenic, sensory, toxic properties or the ability to permeate the in-
testines (Fu & Lin, 2017; Iwaniak, Minkiewicz, Darewicz, Sieniawski, &
Starowicz, 2016; Jung et al., 2007; Minkiewicz et al., 2008). Today, it is
well known, that peptides possessing certain length, and physico-
chemical properties, defined sterical conformation, specific AA pat-
terns, or given AAs at defined positions within the peptide, can be

Fig. 2. Initial approach for predictive application of bioinformatics (marked with blue stars) to the conventional research downstream process. The design of the downstream process
involves hybrid methods, like algorithmic methods, knowledge-based approaches (heuristics) and mechanistic models. These important tools, commonly used in biopharmaceutical
industry, can be crucial for peptide enrichment process development, with the possibility of industrial upscaling. CQA, critical quality attribute, for example measurable product
properties within product quality requirements. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

68
A. Dullius et al.
grouped into various bioactivities, as treated above. To explore the
bioactive profiles and physicochemical properties of peptides,
Minkiewicz et al. (2008) reviewed the potential application of internet-
available computational peptide science tools and still provides a
comprehensive list of currently 62 bioactive peptide databases, avail-
able online with URL http://www.uwm.edu.pl/biochemia/index.php/
pl/biopep/32-bioactive-peptide-databases.
A platform that is widely used to implement the above-mentioned
computational approaches is the BIOPEP database (http://www.uw-
m.edu.pl/biochemia), which was developed at the Chair of Food
Biochemistry at the University of Warmia and Mazury in Olsztyn,
Poland (Minkiewicz et al., 2008; Nongonierma, Le Maux, Hamayon, &
FitzGerald, 2016; Tulipano, Faggi, Nardone, Cocchi, & Caroli, 2015).
The platform can currently simulate an in silico hydrolysis of the desired
protein with a library of 31 proteolytic enzymes, from which two en-
zymes can be tested at a different pH. Various strategies for enzyme
selection were applied until today. Recent studies tried to overcome this
hurdle using a ranking of A-values (defined as a ratio of the number of
peptides with a given activity to the number of AA residues in a protein
sequence) or an additional appliance of the PeptideRanker (http://
bioware.ucd.ie/∼compass/biowareweb/Server_pages/peptider-
anker.php), developed by Mooney, Haslam, Pollastri, and Shields
(2012), beside a comparison with literature data (Agirbasli & Cavas,
2017; Lafarga, Aluko, Rai, O'Connor, & Hayes, 2016; Wang et al.,
2017).
Evaluation of potential biological and physiological activities, al-
lergenicity, sensory activity, and peptide release also can be performed
with the BIOPEP platform (Iwaniak et al., 2016; Minkiewicz et al.,
2008). Minkiewicz et al. (2008) mentioned that predicting the sec-
ondary structures and physicochemical properties of short sequences is
more promising than searching for sequence homology and re-
commended the parallel use of other databases and programs that
process and evaluate peptide sequences, such as CutDB for protein
hydrolysis, ProtParam for physicochemical properties or REMUS to
evaluate differences in peptide fragments (Gasteiger et al., 2005;
Igarashi et al., 2007; Pai et al., 2006). A useful program to predict
secondary structures for linear peptides is PEP-FOLD3 or can be found
at the UniProt server (Lamiable et al., 2016; UniProt Consortium,
2008).
Dziuba and Dziuba (2014) indicated, as is the case for antimicrobial
peptides (AMP), that the simple prediction of physicochemical prop-
erties and values of biological activity indicators is inadequate to effect
a final selection, and must be complemented with multidimensional
statistical analysis using machine learning algorithms, such as support
vector machines (SVM), random forest (RF), artificial neuronal net-
works (ANN) and discriminant analysis (DA), which are available in the
CAMP (Collection of Anti-Microbial Peptides) database. Quantitative
structure-activity relationship (QSAR) modelling has been recently
applied to investigate DPP-IV activity (Nongonierma & FitzGerald,
2016). Gabere and Noble (2017) verified on the basis of AMP prediction
tools that the quality of the predictions is difficult to quantify and tested
two sets of AMP (antimicrobial, antibacterial and bacteriocins) and
non-AMP on ten publicly available prediction tools. The CAMPR3
(random forest) model was evaluated to have the best prediction per-
formance for AMPs.
Capriotti et al. (2016) advised that the use of computational tools
strongly depends on their ability to recognize bioactive sequences.
However, Singh, Sharma, Kumari, and Korpole (2014) reported, that a
new bacteriocin, recognized by the prediction tools for antibacterial
activity CAMPR3 and BACTIBASE, had no significant similarity with
known sequences available in that databases. Chatterjee, Kanawjia,
Khetra, and Saini (2015) criticized that many peptides obtained in vitro
were not predicted in silico, and thus proposed prediction tools that are
more flexible, including variables such as pH, time, temperature, and
enzyme:substrate ratio. The fact that in vitro production of peptides that
were predicted in in silico digestion may not occur as a result of the
69
Journal of Functional Foods 42 (2018) 58–74
complex nature of protein-enzyme interactions was also mentioned by
Udenigwe (2014). Gu and Wu (2013) named factors such as pH, tem-
perature, the formation of complex protein structures, and the catalytic
efficiency of proteases. However, in silico analysis is time-saving and,
compared to the traditional in vitro experimental approaches, a more
economical strategy for bioactive peptide investigations (Fu et al.,
2016).
During peptide processing it is most important not to affect the
functional regions of the peptides. Molecular docking studies are useful
to explain the mode of action of involved AAs in certain peptide
bioactivities (Tu et al., 2017). García-Mora et al. (2017) identified the
molecular mechanism of dual antioxidant/ACE inhibiting activity, de-
rived from lentil proteins, which were hydrolysed with Savinase®. Thus,
the C-terminal tripeptide RYL of lentil derived peptides plays a key role
in ACE inhibiting activity, while interacting strongly through a network
of hydrogen bonds with the enzyme. Furthermore, other ACE inhibiting
peptides listed in the BIOPEP database, also possess this C-terminal
tripeptide, which underlines its importance. These results imply the
necessity to preserve such key regions during downstream processing,
to prevent the loss of the peptides bioactivity.
Regarding the enzymatic hydrolysis step, several working groups
tried to embody the complex reaction kinetics into mathematical
modelling. Valencia, Espinoza, Ceballos, Pinto, and Almonacid (2015)
proposed a new methodology to predict hydrolysis time under different
operation conditions for protein concentration, enzyme concentration
and temperature, usable for different protein sources and enzymes.
Maresca and Ferrari (2017) worked out a model for hydrolysis of BSA
α-chymotrypsin and trypsin under high pressure conditions
(100–600 MPa, 37 °C, hydrolysis time: 0–25 min), which, after inclusion
in future enzyme-substrate ratios, different processing conditions and
types of enzyme is thought to be usable for hydrolysis process devel-
opment in the food, ingredient or pharmaceutical industry. To predict
the impact of Maillard reaction in protein structure to enzyme specifi-
city, selectivity and binding site sensitivity in vitro reaction kinetics
were modelled (Deng et al., 2017).
To use the full capacity of whey derived proteins, in a bioeconomic
sense, it is fundamental to achieve an optimal design of the enrichment
process in order to separate and enrich peptide groups with different
biofunctions. This can be carried out in series by concatenate unit op-
erations or by simultaneous fractionation, as it was done by Suwal et al.
(2017). Therefore, computational tools, as mentioned above, facilitate
the investigation of a peptide’s bioactivity and its physicochemical
properties. Combining possible hydrolysis results with in silico bio-
prospecting can enable organizing the peptides into bioactivity cate-
gories whose members share special physicochemical properties. For
example, hydrophobic sequences with sulfhydryl groups can act as
antioxidant peptides, whereas hydrophobic peptides with cationic
charges often have antimicrobial properties (Dziuba & Dziuba, 2014;
Tian et al., 2015).
Additionally, the use of a hybrid methods model that incorporates
experimental, heuristic and in silico processes can lead to a successful
downstream operating process that can be industrially upscaled (Li-
Chan, 2015; Nfor, Verhaert, van der Wielen, Hubbuch, & Ottens, 2009).
As an example from the biopharmaceutical industry, Nfor et al. (2009)
proposed an in silico process design to model a protein purification
system based on the simulated behaviour of biomolecules. Transferring
this computational modelling approach to the field of peptide enrich-
ment and combining it with knowledge-based procedures could help to
select an economically effective and time-saving operating process,
which would enable upscaling to an industrial level. Hanke and Ottens
(2014) listed the commonly required parameters that are needed to
describe a mechanistic model. Appropriate amino acid descriptors from
QSAR can be used to rate a peptide’s hydrophilicity/hydrophobicity,
molecular size/bulkiness and electronic properties/charge (Kim & Li-
Chan, 2006). Notably, VHSE (a set of vectors describing hydrophobic,
steric and electronic properties) has better predictive power than other
A. Dullius et al.
named descriptors because it uses a broader range of physicochemical
properties (Xie et al., 2013). 3D characterization (molecular weight,
isoelectric point and hydrophobicity) coupled to a random forest cali-
bration or quantitative structure-property relations (QSPR) are helpful
to evaluate peptide separation and column performance to optimize an
enrichment step. Further important parameters are fluid flow and mass
transfer in the column, as well as solute mixture composition, its mo-
lecular properties, possible sample interaction, and the resin- in the
form of absorption isotherms. These in silico based predictions can also
be applied for similar solutes or to determine characteristics for con-
taminants to eliminate them, namely heuristic flowsheeting (Hanke &
Ottens, 2014).
The economic feasibility of an industrial, large-scale process de-
pends strongly on early planning-phase decisions that are based on the
laboratorial scale process (Cziner, Virkki-Hatakka, Hurme, & Turunen,
2005). Therefore, the decision to introduce alternative methods to
purification downstream processes always requires a cost evaluation.
Winkelnkemper and Schembecker (2010) proposed a new performance
indicator for rating purifications and their cost-efficiency, based on
purity improvements, yields and specific costs from single steps, which
can be extracted from the beginning of the experimental investigation.
Purification processes can be divided into (A) an initial purification
step, which results in moderate purity; (B) an intermediate purification
step, which generates a moderately pure to high purity sample; and (C)
a final purification step, which requires a great deal of effort but
reaches very high purity.
High to very high purity in food industry peptide-based ingredients
is not an essential target (Lafarga & Hayes, 2017). The purification
processes can be rated by using the classical parameters of purity dif-
ferences, purification factors and clearance factors (a measure of the
reduction in the level of impurities), as well as by normalizing indices,
such as the purification index, the logarithmic purification index, and
the purification performance index (PPI). The PPI, defined as the ratio
of the logarithmized clearance factors of the step to those of the total
process, is the factor that is most balanced in describing the purification
process and can thus be used to link a purification performance with its
expected effort. The separation cost indicator (SCI) depends on the
normalized purification rating, yield and specific cost of each single
step, and allows a quantitative rating of cost-efficiency to be made.
Furthermore, SCI can be used at the beginning of an experiment-based
process to direct purification development towards an economically
viable procedure (Winkelnkemper & Schembecker, 2010).
5. Conclusion
There is an increasing interest to integrate bioactive whey peptides
as ingredients in functional foods because of their physiological activ-
ities such as providing health-promoting benefits to the corporal im-
mune, cardiovascular, nervous and gastrointestinal system. However,
production at the lab scale and a subsequent scale-up to the industrial
scale requires surmounting many challenges. Production processes,
such as pre-treatment and controlled hydrolysis, increase the specific
release of particular bioactive peptides. Hydrolysis and enrichment
must be performed under mild conditions to avoid affecting the struc-
ture and function of the peptides. Overcoming the bitterness and
maintaining the stability of these bioactive peptides are other chal-
lenges, and they must be resolved while maintaining cost-friendly
production conditions. Achieving a more systematic process flow of
product development that can be cost-efficiently up-scaled to an in-
dustrial production are the main challenges in this field, and passing
these obstacles has led to a reorganization of conventional downstream
processes that requires the integration of hybrid methods. In contrast to
bioactive peptide production in the pharmaceutical industry, purifica-
tion of bioactive peptides by chromatographic methods is cost-prohi-
bitive for peptide enrichment in the food industry, preventing their use
in large-scale production. In silico approaches like peptide activity
70
Journal of Functional Foods 42 (2018) 58–74
prospection, determination of physicochemical properties, mathema-
tical simulation of hydrolysis and “omics”-integrated techniques, such
as molecular docking studies can help to overcome the challenges of
time-consuming and cost-intensive lab-scale peptide discoveries and
characterization. The use of new viable peptide separation methods
(e.g. EDUF, CFEMF or simultaneous hydrolysis/enrichment) in labora-
tory research, facilitated by mechanistic modelling can bridge up-
scaling challenges in an economical way, as they relate to their in-
dustrial application in the functional food industry. Early monitoring of
the downstream process costs during development at the lab scale fa-
cilitates deciding which methods apply and contributes to the afford-
ability of the scale-up. Considering these systematic process design
factors (such a proposed processing pathway), applicable for whey
derived peptides, is likely to produce various targeted whey hydrolysate
products, with a high content of the desired bioactive peptides.
Acknowledgements
We would like to thank Conselho Nacional de Desenvolvimento
Científico e Tecnológico (CNPq) and Coordenação de Aperfeiçoamento
de Pessoal de Nível Superior (CAPES) for scholarships. We would also
like to thank Universidade do Vale do Taquari - Univates and Secretaria
do Desenvolvimento Econômico, Ciência e Tecnologia (SDECT) of Rio
Grande do Sul state for financial support.
Conflicts of interest
The authors declare they do not have any conflicts of interest.
References
Adjonu, R., Doran, G., Torley, P., & Agboola, S. (2013). Screening of whey protein isolate
hydrolysates for their dual functionality: Influence of heat pre-treatment and enzyme
specificity. Food Chemistry, 136(3), 1435–1443.
Adler-Nissen, J. (1979). Determination of the degree of hydrolysis of food protein hy-
drolysates by trinitrobenzenesulfonic acid. Journal of Agricultural and Food Chemistry,
27(6), 1256–1262.
Adler-Nissen, J. (1986). Enzymic hydrolysis of food proteins. Elsevier Applied Science
Publishers.
Aggeli, A., Bell, M., Carrick, L. M., Fishwick, C. W., Harding, R., Mawer, P. J., ... Boden, N.
(2003). PH as a trigger of peptide β-sheet self-assembly and reversible switching
between nematic and isotropic phases. Journal of the American Chemical Society,
125(32), 9619–9628.
Agirbasli, Z., & Cavas, L. (2017). In silico evaluation of bioactive peptides from the green
algae Caulerpa. Journal of Applied Phycology, 29(3), 1635–1646.
Agyei, D., & Danquah, M. K. (2011). Industrial-scale manufacturing of pharmaceutical-
grade bioactive peptides. Biotechnology Advances, 29(3), 272–277.
Agyei, D., Ongkudon, C. M., Wei, C. Y., Chan, A. S., & Danquah, M. K. (2016). Bioprocess
challenges to the isolation and purification of bioactive peptides. Food and
Bioproducts Processing, 98, 244–256.
Akalın, A. S. (2014). Dairy-derived antimicrobial peptides: Action mechanisms, phar-
maceutical uses and production proposals. Trends in Food Science & Technology, 36(2),
79–95.
Aluko, R. E. (2017). Structural characteristics of food protein-derived bitter peptides.
Bitterness: Perception, Chemistry and Food Processing105–129.
Anand, S., Som Nath, K., & Chenchaiah, M. (2013). Whey and whey products. Milk and
Dairy Products in Human Nutrition: Production, Composition and Health477–497.
Aoki, W., & Ueda, M. (2013). Characterization of antimicrobial peptides toward the de-
velopment of novel antibiotics. Pharmaceuticals, 6(8), 1055–1081.
Arrutia, F., Puente, Á., Riera, F. A., Menéndez, C., & González, U. A. (2016). Influence of
heat pre-treatment on BSA tryptic hydrolysis and peptide release. Food Chemistry,
202, 40–48.
Arrutia, F., Rubio, R., & Riera, F. A. (2016). Production and membrane fractionation of
bioactive peptides from a whey protein concentrate. Journal of Food Engineering,
184, 1–9.
Ballard, K. D., Bruno, R. S., Seip, R. L., Quann, E. E., Volk, B. M., Freidenreich, D. J., ...
Volek, JS. (2009). Acute ingestion of a novel whey-derived peptide improves vascular
endothelial responses in healthy individuals: A randomized, placebo controlled trial.
Nutrition Journal, 8(1), 34.
Bamdad, F., Bark, S., Kwon, C. H., Suh, J. W., & Sunwoo, H. (2017). Anti-inflammatory
and antioxidant properties of peptides released from β-lactoglobulin by high hy-
drostatic pressure-assisted enzymatic hydrolysis. Molecules, 22(6), 949.
Bassan, J. C., de Souza Bezerra, T. M., Peixoto, G., da Cruz, C. Z. P., Galán, J. P. M., Vaz,
A. B. D. S., ... Monti, R. (2016). Immobilization of trypsin in lignocellulosic waste
material to produce peptides with bioactive potential from whey protein. Materials,
9(5), 357.
A. Dullius et al.
Bhat, Z. F., Kumar, S., & Bhat, H. F. (2017). Antihypertensive peptides of animal origin: A
review. Critical reviews in food science and nutrition, 57(3), 566–578.
Bhut, B. V., Weaver, J., Carter, A. R., Wickramasinghe, S. R., & Husson, S. M. (2011). The
role of polymer nanolayer architecture on the separation performance of anion-ex-
change membrane adsorbers: I. Protein separations. Biotechnology and Bioengineering,
108(11), 2645–2653.
Blayo, C., Vidcoq, O., Lazennec, F., & Dumay, E. (2016). Effects of high pressure pro-
cessing (hydrostatic high pressure and ultra-high pressure homogenisation) on whey
protein native state and susceptibility to tryptic hydrolysis at atmospheric pressure.
Food Research International, 79, 40–53.
Bouglé, D., & Bouhallab, S. (2017). Dietary bioactive peptides: Human studies. Critical
Reviews in Food Science and Nutrition, 57(2), 335–343.
Brand, J., Dachmann, E., Pichler, M., Lotz, S., & Kulozik, U. (2016). A novel approach for
lysozyme and ovotransferrin fractionation from egg white by radial flow membrane
adsorption chromatography: Impact of product and process variables. Separation and
Purification Technology, 161, 44–52.
Brand, J., Voigt, K., Zochowski, B., & Kulozik, U. (2016). Lysozyme fractionation from egg
white at pilot scale by means of tangential flow membrane adsorbers: Investigation of
the flow conditions. Journal of Chromatography A, 1438, 143–149.
Bulone, D., Martorana, V., & San Biagio, P. L. (2001). Effects of intermediates on ag-
gregation of native bovine serum albumin. Biophysical Chemistry, 91(1), 61–69.
Butré, C. I., Sforza, S., Wierenga, P. A., & Gruppen, H. (2015). Determination of the in-
fluence of the pH of hydrolysis on enzyme selectivity of Bacillus licheniformis protease
towards whey protein isolate. International Dairy Journal, 44, 44–53.
Capriotti, A. L., Cavaliere, C., Foglia, P., Piovesana, S., Samperi, R., Chiozzi, R. Z., &
Laganà, A. (2015). Development of an analytical strategy for the identification of
potential bioactive peptides generated by in vitro tryptic digestion of fish muscle
proteins. Analytical and Bioanalytical Chemistry, 407(3), 845–854.
Capriotti, A. L., Cavaliere, C., Piovesana, S., Samperi, R., & Laganà, A. (2016). Recent
trends in the analysis of bioactive peptides in milk and dairy products. Analytical and
Bioanalytical Chemistry, 408(11), 2677–2685.
Carvalho, N. C.d., Pessato, TB., Fernandes, LGR., de Lima Zollner, R., & Netto, FM.
(2017). Physicochemical characteristics and antigenicity of whey protein hydro-
lysates obtained with and without pH control. International Dairy Journal, 71, 24–34.
Chatterjee, A., Kanawjia, S. K., & Khetra, Y. (2016). Properties of sweetened Indian yogurt
(mishti dohi) as affected by added tryptic whey protein hydrolysate. Journal of Food
Science and Technology, 53(1), 824–831.
Chatterjee, A., Kanawjia, S. K., Khetra, Y., & Saini, P. (2015). Discordance between in
silico & in vitro analyses of ACE inhibitory & antioxidative peptides from mixed milk
tryptic whey protein hydrolysate. Journal of Food Science and Technology, 52(9),
5621–5630.
Cheison, S. C., Bor, E. K., Faraj, A. K., & Kulozik, U. (2012). Selective hydrolysis of α-
lactalbumin by Acid Protease A offers potential for β-lactoglobulin purification in
whey proteins. LWT-Food Science and Technology, 49(1), 117–122.
Cheison, S. C., Brand, J., Leeb, E., & Kulozik, U. (2011). Analysis of the effect of tem-
perature changes combined with different alkaline pH on the β-lactoglobulin trypsin
hydrolysis pattern using MALDI-TOF-MS/MS. Journal of Agricultural and Food
Chemistry, 59(5), 1572–1581.
Cheison, S. C., & Kulozik, U. (2017). Impact of the environmental conditions and sub-
strate pre-treatment on whey protein hydrolysis: A review. Critical Reviews in Food
Science and Nutrition, 57(2), 418–453.
Cheison, S. C., Lai, M. Y., Leeb, E., & Kulozik, U. (2011). Hydrolysis of β-lactoglobulin by
trypsin under acidic pH and analysis of the hydrolysates with MALDI–TOF–MS/MS.
Food Chemistry, 125(4), 1241–1248.
Cheison, S. C., Leeb, E., Letzel, T., & Kulozik, U. (2011). Influence of buffer type and
concentration on the peptide composition of trypsin hydrolysates of β-lactoglobulin.
Food Chemistry, 125(1), 121–127.
Cheison, S. C., Leeb, E., Toro-Sierra, J., & Kulozik, U. (2011). Influence of hydrolysis
temperature and pH on the selective hydrolysis of whey proteins by trypsin and
potential recovery of native alpha-lactalbumin. International Dairy Journal, 21(3),
166–171.
Cheison, S. C., Wang, Z., & Xu, S. Y. (2007). Multivariate strategy in screening of enzymes
to be used for whey protein hydrolysis in an enzymatic membrane reactor.
International Dairy Journal, 17(4), 393–402.
Chen, W., Ding, H., Feng, P., Lin, H., & Chou, K. C. (2016). IACP: A sequence-based tool
for identifying anticancer peptides. Oncotarget, 7(13), 16895.
Chen, H. M., Muramoto, K., & Yamauchi, F. (1995). Structural analysis of antioxidative
peptides from soybean β-conglycinin. Journal of Agricultural and Food Chemistry,
43(3), 574–578.
Chik, T., & Saufi, S. M. (2010). Mixed matrix membrane chromatography for bovine whey
protein fractionation. Doctoral Thesis available at < http://hdl.handle.net/10092/
3647 > . Accessed June, 2017.
Chmiel, H. (Ed.). (2011). Bioprozesstechnik (pp. 356–372). Springer-Verlag.
Chobert, J. M., Briand, L., Dufour, E., Dib, R., Dalgalarrondo, M., & Haertle, T. (1996).
How to increase β-lactoglobulin susceptibility to peptic hydrolysis. Journal of Food
Biochemistry, 20(4), 439–462.
Conesa, C., & FitzGerald, R. J. (2013). Total solids content and degree of hydrolysis in-
fluence proteolytic inactivation kinetics following whey protein hydrolysate manu-
facture. Journal of Agricultural and Food Chemistry, 61(42), 10135–10144.
Creusot, N., & Gruppen, H. (2008). Hydrolysis of whey protein isolate with Bacillus li-
cheniformis protease: Aggregating capacities of peptide fractions. Journal of
Agricultural and Food Chemistry, 56(21), 10332–10339.
Cui, H., Webber, M. J., & Stupp, S. I. (2010). Self-assembly of peptide amphiphiles: From
molecules to nanostructures to biomaterials. Peptide Science, 94(1), 1–18.
Cziner, K., Virkki-Hatakka, T., Hurme, M., & Turunen, I. (2005). Evaluative approach for
process development. Chemical Engineering & Technology, 28(12), 1490–1499.
71
Journal of Functional Foods 42 (2018) 58–74
Da Silva, M. S., Bigo, C., Barbier, O., & Rudkowska, I. (2017). Whey protein hydrolysate
and branched-chain amino acids downregulate inflammation-related genes in vas-
cular endothelial cells. Nutrition Research, 38, 43–51.
De Maria, S., Ferrari, G., & Maresca, P. (2017). Effect of high hydrostatic pressure on the
enzymatic hydrolysis of bovine serum albumin. Journal of the Science of Food and
Agriculture, 97(10), 3151–3158.
del Mar Contreras, M., Hernández-Ledesma, B., Amigo, L., Martín-Álvarez, P. J., & Recio,
I. (2011). Production of antioxidant hydrolyzates from a whey protein concentrate
with thermolysin: Optimization by response surface methodology. LWT-Food Science
and Technology, 44(1), 9–15.
Delahaije, R. J., Gruppen, H., van Eijk- van Boxtel, E. L., Cornacchia, L., & Wierenga, P. A.
(2016). Controlling the ratio between native-like, non-native-like, and aggregated β-
lactoglobulin after heat treatment. Journal of Agricultural and Food Chemistry, 64(21),
4362–4370.
Deng, Y., Wierenga, P. A., Schols, H. A., Sforza, S., & Gruppen, H. (2017). Effect of
Maillard induced glycation on protein hydrolysis by lysine/arginine and non-lysine/
arginine specific proteases. Food Hydrocolloids, 69, 210–219.
Doucet, D., Otter, D. E., Gauthier, S. F., & Foegeding, E. A. (2003). Enzyme-induced ge-
lation of extensively hydrolyzed whey proteins by Alcalase: Peptide identification
and determination of enzyme specificity. Journal of Agricultural and Food Chemistry,
51(21), 6300–6308.
Doyen, A., Udenigwe, C. C., Mitchell, P. L., Marette, A., Aluko, R. E., & Bazinet, L. (2014).
Anti-diabetic and antihypertensive activities of two flaxseed protein hydrolysate
fractions revealed following their simultaneous separation by electrodialysis with
ultrafiltration membranes. Food Chemistry, 145, 66–76.
Dziuba, B., & Dziuba, M. (2014). New milk protein-derived peptides with potential an-
timicrobial activity: An approach based on bioinformatic studies. International Journal
of Molecular Sciences, 15(8), 14531–14545.
Elias, R. J., Kellerby, S. S., & Decker, E. A. (2008). Antioxidant activity of proteins and
peptides. Critical Reviews in Food Science and Nutrition, 48(5), 430–441.
Eliassen, L. T., Berge, G., Sveinbjørnsson, B., Svendsen, J. S., Vorland, L. H., & Rekdal, Ø.
(2002). Evidence for a direct antitumor mechanism of action of bovine lactoferricin.
Anticancer Research, 22, 2703–2710.
Expósito, I. L., & Recio, I. (2006). Antibacterial activity of peptides and folding variants
from milk proteins. International Dairy Journal, 16(11), 1294–1305.
Fernández, A., & Kelly, P. (2016). PH-stat vs. free-fall pH techniques in the enzymatic
hydrolysis of whey proteins. Food Chemistry, 199, 409–415.
Fernández, A., & Riera, F. (2013). Β-Lactoglobulin tryptic digestion: A model approach
for peptide release. Biochemical Engineering Journal, 70, 88–96.
Fernández, A., Zhu, Y., FitzGerald, R. J., & Riera, F. A. (2014). Membrane fractionation of
a β-lactoglobulin tryptic digest: Effect of the membrane characteristics. Journal of
Chemical Technology and Biotechnology, 89(4), 508–515.
Fu, Z., & Lin, J. (2017). An overview of bioinformatics tools and resources in allergy. Food
Allergens: Methods and Protocols223–245.
Fu, Y., Young, J. F., Løkke, M. M., Lametsch, R., Aluko, R. E., & Therkildsen, M. (2016).
Revalorisation of bovine collagen as a potential precursor of angiotensin I-converting
enzyme (ACE) inhibitory peptides based on in silico and in vitro protein digestions.
Journal of Functional Foods, 24, 196–206.
Gabere, M. N., & Noble, W. S. (2017). Empirical comparison of web-based antimicrobial
peptide prediction tools. Bioinformatics, 33(13), 1921–1929.
Gabernet, G., Müller, A. T., Hiss, J. A., & Schneider, G. (2016). Membranolytic anticancer
peptides. Medicinal Chemistry Communications, 7(12), 2232–2245.
García-Mora, P., Martín-Martínez, M., Bonache, M. A., González-Múniz, R., Peñas, E.,
Frias, J., & Martinez-Villaluenga, C. (2017). Identification, functional gastrointestinal
stability and molecular docking studies of Lentil peptides with dual antioxidant and
Angiotensin I converting enzyme inhibitory activities. Food Chemistry, 221, 464–472.
García-Tejedor, A., Sánchez-Rivera, L., Castelló-Ruiz, M., Recio, I., Salom, J. B., &
Manzanares, P. (2014). Novel antihypertensive lactoferrin-derived peptides produced
by Kluyveromyces marxianus: Gastrointestinal stability profile and in vivo angio-
tensin I-converting enzyme (ACE) inhibition. Journal of Agricultural and Food
Chemistry, 62(7), 1609–1616.
Garg, S., Nurgali, K., & Kumar Mishra, V. (2016). Food proteins as source of opioid
peptides-a review. Current Medicinal Chemistry, 23(9), 893–910.
Gasteiger, E., Hoogland, C., Gattiker, A., Duvaud, S., Wilkins, M. R., Appel, R. D., &
Bairoch, A. (2005). The proteomics protocols handbook. In J.M. Walker.
Gauthier, S. F., Pouliot, Y., & Saint-Sauveur, D. (2006). Immunomodulatory peptides
obtained by the enzymatic hydrolysis of whey proteins. International Dairy Journal,
16(11), 1315–1323.
Graves, A. M., Hettiarachchy, N., Rayaprolu, S., Li, R., Horax, R., & Seo, H. S. (2016).
Bioactivity of a rice bran–derived peptide and its sensory evaluation and storage
stability in orange juice. Journal of Food Science, 81(4), H1010–H1015.
Gu, Y., & Wu, J. (2013). LC–MS/MS coupled with QSAR modeling in characterising of
angiotensin I-converting enzyme inhibitory peptides from soybean proteins. Food
Chemistry, 141(3), 2682–2690.
Gupta, C. (2012). Whey proteins: A novel source of bioceuticals. Middle-East Journal of
Scientific Research, 12(3), 365–375.
Hafeez, Z., Cakir-Kiefer, C., Roux, E., Perrin, C., Miclo, L., & Dary-Mourot, A. (2014).
Strategies of producing bioactive peptides from milk proteins to functionalize fer-
mented milk products. Food Research International, 63, 71–80.
Hanke, A. T., & Ottens, M. (2014). Purifying biopharmaceuticals: Knowledge-based
chromatographic process development. Trends in Biotechnology, 32(4), 210–220.
Hati, S., Patel, N., Sakure, A., & Mandal, S. (2017). Influence of whey protein concentrate
on the production of antibacterial peptides derived from fermented milk by lactic
acid bacteria. International Journal of Peptide Research and Therapeutics, 1–12.
Hernández-Ledesma, B., Dávalos, A., Bartolomé, B., & Amigo, L. (2005). Preparation of
antioxidant enzymatic hydrolysates from α-lactalbumin and β-lactoglobulin.
A. Dullius et al.
Identification of active peptides by HPLC-MS/MS. Journal of Agricultural and Food
Chemistry, 53(3), 588–593.
Hernández-Ledesma, B., del Mar Contreras, M., & Recio, I. (2011). Antihypertensive
peptides: Production, bioavailability and incorporation into foods. Advances in Colloid
and Interface Science, 165(1), 23–35.
Hernández-Ledesma, B., & Hsieh, C. C. (2017). Chemopreventive role of food-derived
proteins and peptides: A review. Critical reviews in Food Science and Nutrition, 57(11),
2358–2376.
Hettiarachchy, N. S., Sato, K., Marshall, M. R., & Kannan, A. (Eds.). (2011). Bioactive food
proteins and peptides: Applications in human health. CRC Press.
Hippauf, F., Huettner, C., Lunow, D., Borchardt, L., Henle, T., & Kaskel, S. (2016).
Towards a continuous adsorption process for the enrichment of ACE-inhibiting
peptides from food protein hydrolysates. Carbon, 107, 116–123.
Hippauf, F., Lunow, D., Borchardt, L., Henle, T., & Kaskel, S. (2014). Extraction of ACE-
inhibiting dipeptides from protein hydrolysates using porous carbon materials.
Carbon, 77, 191–198.
Hoffmann, K. F. (1961). On the history of whey cures, especially in the 17th, 18th, and
19th centuries. Medizinische Monatsschrift, 15, 411.
Holder, A., Merath, C., Kulozik, U., & Hinrichs, J. (2015). Impact of diffusion, trans-
membrane pressure and the electrical field on peptide fractionation using cross-flow
electro membrane filtration. International Dairy Journal, 46, 31–38.
Holder, A., Scholz, S., Kulozik, U., & Hinrichs, J. (2013). Cross-flow electro membrane
filtration: Theory and application in the dairy industry. Chemie Ingenieur Technik,
85(8), 1193–1200.
Hudson, M. J. (1995). Product development horizons-a view from industry. European
Journal of Clinical Nutrition, 49, S64–S70.
Igarashi, Y., Eroshkin, A., Gramatikova, S., Gramatikoff, K., Zhang, Y., Smith, J. W., ...
Godzik, A. (2007). CutDB: A proteolytic event database. Nucleic Acids Research,
35(suppl_1), D546–D549.
Ittrat, P., Chacho, T., Pholprayoon, J., Suttiwarayanon, N., & Charoenpanich, J. (2014).
Application of agriculture waste as a support for lipase immobilization. Biocatalysis
and Agricultural Biotechnology, 3(3), 77–82.
Iwaniak, A., Minkiewicz, P., Darewicz, M., Sieniawski, K., & Starowicz, P. (2016). BIOPEP
database of sensory peptides and amino acids. Food Research International, 85,
155–161.
Jayamuthunagai, J., Srisowmeya, G., Chakravarthy, M., & Gautam, P. (2017). D-Tagatose
production by permeabilized and immobilized Lactobacillus plantarum using whey
permeate. Bioresource Technology, 235, 250–255.
Jiang, B., Zhang, X., Yuan, Y., Qu, Y., & Feng, Z. (2017). Separation of antioxidant
peptides from pepsin hydrolysate of whey protein isolate by ATPS of EOPO co-
polymer (UCON)/Phosphate. Scientific Reports, 7(1), 13320.
Jung, E., Kim, J., Kim, M., Jung, D. H., Rhee, H., Shin, J. M., ... Choi, YJ. (2007). Artificial
neural network models for prediction of intestinal permeability of oligopeptides.
BMC Bioinformatics, 8(1), 245.
Kim, H. O., & Li-Chan, E. C. (2006). Quantitative structure-activity relationship study of
bitter peptides. Journal of Agricultural and Food Chemistry, 54(26), 10102–10111.
Klompong, V., Benjakul, S., Kantachote, D., & Shahidi, F. (2007). Antioxidative activity
and functional properties of protein hydrolysate of yellow stripe trevally (Selaroides
leptolepis) as influenced by the degree of hydrolysis and enzyme type. Food Chemistry,
102(4), 1317–1327.
Korhonen, H. J. (2009a). Bioactive components in bovine milk. Bioactive Components in
Milk and Dairy Products, 15–42.
Korhonen, H. (2009b). Milk-derived bioactive peptides: From science to applications.
Journal of Functional Foods, 1(2), 177–187.
Korhonen, H. J. (2012). Production and properties of health-promoting proteins and
peptides from bovine colostrum and milk. Cellular & Molecular Biology, 58(1), 26–38.
Korhonen, H. J., & Marnila, P. (2013). Milk bioactive proteins and peptides. Milk and Dairy
Products in Human Nutrition: Production, Composition and Health148–171.
Korhonen, H., & Pihlanto, A. (2006). Bioactive peptides: Production and functionality.
International Dairy Journal, 16(9), 945–960.
Korhonen, H., Pihlanto-Leppäla, A., Rantamäki, P., & Tupasela, T. (1998). Impact of
processing on bioactive proteins and peptides. Trends in Food Science & Technology,
9(8), 307–319.
Lacroix, I. M., & Li-Chan, E. C. (2013). Inhibition of dipeptidyl peptidase (DPP)-IV and α-
glucosidase activities by pepsin-treated whey proteins. Journal of Agricultural and
Food Chemistry, 61(31), 7500–7506.
Lacroix, I. M., & Li-Chan, E. C. (2014). Isolation and characterization of peptides with
dipeptidyl peptidase-IV inhibitory activity from pepsin-treated bovine whey proteins.
Peptides, 54, 39–48.
Lacroix, I. M., Meng, G., Cheung, I. W., & Li-Chan, E. C. (2016). Do whey protein-derived
peptides have dual dipeptidyl-peptidase IV and angiotensin I-converting enzyme in-
hibitory activities? Journal of Functional Foods, 21, 87–96.
Lafarga, T., Aluko, R. E., Rai, D. K., O'Connor, P., & Hayes, M. (2016). Identification of
bioactive peptides from a papain hydrolysate of bovine serum albumin and assess-
ment of an antihypertensive effect in spontaneously hypertensive rats. Food Research
International, 81, 91–99.
Lafarga, T., & Hayes, M. (2017). Bioactive protein hydrolysates in the functional food
ingredient industry: Overcoming current challenges. Food Reviews International,
33(3), 217–246.
Lafarga, T., Rai, D. K., O'Connor, P., & Hayes, M. (2016). Generation of bioactive hy-
drolysates and peptides from bovine hemoglobin with in vitro renin, angiotensin-I-
converting enzyme and dipeptidyl peptidase-IV inhibitory activities. Journal of Food
Biochemistry, 40(5), 673–685.
Lamiable, A., Thévenet, P., Rey, J., Vavrusa, M., Derreumaux, P., & Tufféry, P. (2016).
PEP-FOLD3: Faster de novo structure prediction for linear peptides in solution and in
complex. Nucleic Acids Research, 44(W1), W449–W454.
72
Journal of Functional Foods 42 (2018) 58–74
Le Maux, S., Nongonierma, A. B., Barre, C., & FitzGerald, R. J. (2016). Enzymatic gen-
eration of whey protein hydrolysates under pH-controlled and non pH-controlled
conditions: Impact on physicochemical and bioactive properties. Food Chemistry, 199,
246–251.
Le Maux, S., Nongonierma, A. B., & FitzGerald, R. J. (2017). Peptide composition and
dipeptidyl peptidase IV inhibitory properties of β-lactoglobulin hydrolysates having
similar extents of hydrolysis while generated using different enzyme-to-substrate
ratios. Food Research International. http://dx.doi.org/10.1016/j.foodres.2017.05.012.
Leeb, E., Götz, A., Letzel, T., Cheison, S. C., & Kulozik, U. (2015). Influence of dena-
turation and aggregation of β-lactoglobulin on its tryptic hydrolysis and the release of
functional peptides. Food Chemistry, 187, 545–554.
Leeb, E., Holder, A., Letzel, T., Cheison, S. C., Kulozik, U., & Hinrichs, J. (2014).
Fractionation of dairy based functional peptides using ion-exchange membrane ad-
sorption chromatography and cross-flow electro membrane filtration. International
Dairy Journal, 38(2), 116–123.
Leeb, E., Kulozik, U., & Cheison, S. (2011). Thermal pre-treatment of β-Lactoglobulin as a
tool to steer enzymatic hydrolysis and control the release of peptides. Procedia Food
Science, 1, 1540–1546.
Li, S., Hao, L., Bao, W., Zhang, P., Su, D., Cheng, Y., ... Yang, Y. (2016). A novel short
anionic antibacterial peptide isolated from the skin of Xenopus laevis. Archives of
Microbiology, 198(5), 473–482.
Li, E. W., & Mine, Y. (2004). Immunoenhancing effects of bovine glycomacropeptide and
its derivatives on the proliferative response and phagocytic activities of human
macrophagelike cells, U937. Journal of Agricultural and Food Chemistry, 52(9),
2704–2708.
Li-Chan, E. C. (2015). Bioactive peptides and protein hydrolysates: Research trends and
challenges for application as nutraceuticals and functional food ingredients. Current
Opinion in Food Science, 1, 28–37.
Liu, M., Bayjanov, J. R., Renckens, B., Nauta, A., & Siezen, R. J. (2010). The proteolytic
system of lactic acid bacteria revisited: A genomic comparison. BMC Genomics,
11(1), 36.
Ma, J. J., Mao, X. Y., Wang, Q., Yang, S., Zhang, D., Chen, S. W., & Li, Y. H. (2014). Effect
of spray drying and freeze drying on the immunomodulatory activity, bitter taste and
hygroscopicity of hydrolysate derived from whey protein concentrate. LWT-Food
Science and Technology, 56(2), 296–302.
Madureira, A. R., Tavares, T., Gomes, A. M. P., Pintado, M. E., & Malcata, F. X. (2010).
Invited review: Physiological properties of bioactive peptides obtained from whey
proteins. Journal of Dairy Science, 93(2), 437–455.
Maeda-Yamamoto, M. (in press). Development of functional agricultural products and use
of a new health claim system in Japan. Trends in Food Science & Technology. http://dx.
doi.org/10.1016/j.tifs.2017.08.011, available online 22 August 2017.
Maehashi, K., & Huang, L. (2009). Bitter peptides and bitter taste receptors. Cellular and
Molecular Life Sciences, 66(10), 1661–1671.
Mao, Y., Černigoj, U., Zalokar, V., Štrancar, A., & Kulozik, U. (2017). Production of β-
lactoglobulin hydrolysates by monolith based immobilized trypsin reactors.
Electrophoresis. http://dx.doi.org/10.1002/elps.201700188.
Maresca, P., & Ferrari, G. (2017). Modelling of the kinetics of bovine serum albumin
enzymatic hydrolysis assisted by high hydrostatic pressure. Food and Bioproducts
Processing, 105, 1–11.
Martínez-Maqueda, D., Miralles, B., De Pascual-Teresa, S., Reverón, I., Muñoz, R., &
Recio, I. (2012). Food-derived peptides stimulate mucin secretion and gene expres-
sion in intestinal cells. Journal of Agricultural and Food Chemistry, 60(35), 8600–8605.
Martínez-Maqueda, D., Miralles, B., Ramos, M., & Recio, I. (2013). Effect of β-lactoglo-
bulin hydrolysate and β-lactorphin on intestinal mucin secretion and gene expression
in human goblet cells. Food Research International, 54(1), 1287–1291.
McIntosh, G. H., Royle, P. J., Le Leu, R. K., Regester, G. O., Johnson, M. A., Grinsted, R. L.,
... Smithers, GW. (1998). Whey proteins as functional food ingredients? International
Dairy Journal, 8(5–6), 425–434.
Meisel, H. (1997). Biochemical properties of regulatory peptides derived from mil pro-
teins. Peptide Science, 43(2), 119–128.
Minkiewicz, P., Dziuba, J., Iwaniak, A., Dziuba, M., & Darewicz, M. (2008). BIOPEP
database and other programs for processing bioactive peptide sequences. Journal of
AOAC International, 91(4), 965–980.
Mohan, A., Rajendran, S. R., He, Q. S., Bazinet, L., & Udenigwe, C. C. (2015).
Encapsulation of food protein hydrolysates and peptides: A review. RSC Advances,
5(97), 79270–79278.
Mohan, A., Udechukwu, M. C., Rajendran, S. R., & Udenigwe, C. C. (2015). Modification
of peptide functionality during enzymatic hydrolysis of whey proteins. RSC Advances,
5(118), 97400–97407.
Mohanty, D., Jena, R., Choudhury, P. K., Pattnaik, R., Mohapatra, S., & Saini, M. R.
(2016). Milk derived antimicrobial bioactive peptides: A review. International Journal
of Food Properties, 19(4), 837–846.
Mooney, C., Haslam, N. J., Pollastri, G., & Shields, D. C. (2012). Towards the improved
discovery and design of functional peptides: Common features of diverse classes
permit generalized prediction of bioactivity. PLoS ONE, 7(10), e45012.
Mullally, M. M., Meisel, H., & FitzGerald, R. J. (1997). Identification of a novel angio-
tensin-I-converting enzyme inhibitory peptide corresponding to a tryptic fragment of
bovine β-lactoglobulin. FEBS Letters, 402(2–3), 99–101.
Mundi, S., & Aluko, R. E. (2014). Inhibitory properties of kidney bean protein hydrolysate
and its membrane fractions against renin, angiotensin converting enzyme, and free
radicals. Austin Journal of Nutrition and Food Sciences, 2, 1008.
Ney, K. H. (1971). Prediction of bitterness of peptides from their amino acid composition.
Zeitschrift für Lebensmittel-Untersuchung und -Forschung, 147, 64–68.
Ney, D. M., & Etzel, M. R. (2017). Designing medical foods for inherited metabolic dis-
orders: Why intact protein is superior to amino acids. Current Opinion in Biotechnology,
44, 39–45.
A. Dullius et al.
Nfor, B. K., Verhaert, P. D., van der Wielen, L. A., Hubbuch, J., & Ottens, M. (2009).
Rational and systematic protein purification process development: The next genera-
tion. Trends in Biotechnology, 27(12), 673–679.
Ngarize, S., Herman, H., Adams, A., & Howell, N. (2004). Comparison of changes in the
secondary structure of unheated, heated, and high-pressure-treated β-lactoglobulin
and ovalbumin proteins using Fourier transform Raman spectroscopy and self-de-
convolution. Journal of Agricultural and Food Chemistry, 52(21), 6470–6477.
Nielsen, S. D., Beverly, R. L., Qu, Y., & Dallas, D. C. (2017). Milk bioactive peptide da-
tabase: A comprehensive database of milk protein-derived bioactive peptides and
novel visualization. Food Chemistry, 232, 673–682.
Nongonierma, A. B., & FitzGerald, R. J. (2016). Structure activity relationship modelling
of milk protein-derived peptides with dipeptidyl peptidase IV (DPP-IV) inhibitory
activity. Peptides, 79, 1–7.
Nongonierma, A. B., Le Maux, S., Hamayon, J., & FitzGerald, R. J. (2016). Strategies for
the release of dipeptidyl peptidase IV (DPP-IV) inhibitory peptides in an enzymatic
hydrolyzate of α-lactalbumin. Food & Function, 7(8), 3437–3443.
O’Loughlin, I. B., Kelly, P. M., Murray, B. A., FitzGerald, R. J., & Brodkorb, A. (2015).
Molecular characterization of whey protein hydrolysate fractions with ferrous che-
lating and enhanced iron solubility capabilities. Journal of Agricultural and Food
Cchemistry, 63(10), 2708–2714.
O'Keeffe, M. B., Conesa, C., & FitzGerald, R. J. (2017). Identification of angiotensin
converting enzyme inhibitory and antioxidant peptides in a whey protein concentrate
hydrolysate produced at semi-pilot scale. International Journal of Food Science &
Technology, 52(8), 1751–1759.
O'Loughlin, I. B., Murray, B. A., FitzGerald, R. J., Brodkorb, A., & Kelly, P. M. (2014).
Pilot-scale production of hydrolysates with altered bio-functionalities based on
thermally-denatured whey protein isolate. International Dairy Journal, 34(1),
146–152.
Pai, T. W., Chang, M. D. T., Tzou, W. S., Su, B. H., Wu, P. C., Chang, H. T., & Chou, W. I.
(2006). RE MUS: A tool for identification of unique peptide segments as epitopes.
Nucleic Acids Research, 34(suppl_2), W198–W201.
Pan, D., Cao, J., Guo, H., & Zhao, B. (2012). Studies on purification and the molecular
mechanism of a novel ACE inhibitory peptide from whey protein hydrolysate. Food
Chemistry, 130(1), 121–126.
Panchaud, A., Affolter, M., & Kussmann, M. (2012). Mass spectrometry for nutritional
peptidomics: How to analyze food bioactives and their health effects. Journal of
Proteomics, 75(12), 3546–3559.
Pandey, M., Kapila, R., & Kapila, S. (2018). Osteoanabolic activity of whey-derived anti-
oxidative (MHIRL and YVEEL) and angiotensin-converting enzyme inhibitory (YLLF,
ALPMHIR, IPA and WLAHK) bioactive peptides. Peptides, 99, 1–7.
Pasupuleti, V. K., & Braun, S. (2008). State of the art manufacturing of protein hydro-
lysates. Protein hydrolysates in biotechnology (pp. 1–32). Netherlands: Springer.
Pellegrini, A., Dettling, C., Thomas, U., & Hunziker, P. (2001). Isolation and character-
ization of four bactericidal domains in the bovine β-lactoglobulin. Biochimica et
Biophysica Acta (BBA)-General Subjects, 1526(2), 131–140.
Peña-Ramos, E. A., Xiong, Y. L., & Arteaga, G. E. (2004). Fractionation and character-
isation for antioxidant activity of hydrolysed whey protein. Journal of the Science of
Food and Agriculture, 84(14), 1908–1918.
Persico, M., Mikhaylin, S., Doyen, A., Firdaous, L., Hammami, R., & Bazinet, L. (2016).
How peptide physicochemical and structural characteristics affect anion-exchange
membranes fouling by a tryptic whey protein hydrolysate. Journal of Membrane
Science, 520, 914–923.
Pieroni, A., & Giusti, M. E. (2008). The remedies of the folk medicine of the Croatians
living in Ćićarija, northern Istria. Collegium Antropologicum, 32(2), 623–627.
Pieroni, A., & Gray, C. (2008). Herbal and food folk medicines of the Russlanddeutschen
living in Künzelsau/Taläcker, South-Western Germany. Phytotherapy Research, 22(7),
889–901.
Pihlanto, A. (2006). Antioxidative peptides derived from milk proteins. International Dairy
Journal, 16(11), 1306–1314.
Pihlanto-Leppälä, A. (2000). Bioactive peptides derived from bovine whey proteins:
Opioid and ace-inhibitory peptides. Trends in Food Science & Technology, 11(9),
347–356.
Pihlanto-Leppälä, A., Koskinen, P., Piilola, K., Tupasela, T., & Korhonen, H. (2000).
Angiotensin I-converting enzyme inhibitory properties of whey protein digests:
Concentration and characterization of active peptides. Journal of Dairy Research,
67(1), 53–64.
Pihlanto-Leppälä, A., Paakkari, I., Rinta-Koski, M., & Antila, P. (1997). Bioactive peptide
derived from in vitro proteolysis of bovine β-lactoglobulin and its effect on smooth
muscle. Journal of Dairy Research, 64(1), 149–155.
Pina, A. S., & Roque, A. C. A. (2009). Studies on the molecular recognition between
bioactive peptides and angiotensin-converting enzyme. Journal of Molecular
Recognition, 22(2), 162–168.
Prazeres, A. R., Carvalho, F., & Rivas, J. (2012). Cheese whey management: A review.
Journal of Environmental Management, 110, 48–68.
Roberts, M. J., Bentley, M. D., & Harris, J. M. (2012). Chemistry for peptide and protein
PEGylation. Advanced Drug Delivery Reviews, 64, 116–127.
Rodríguez-Carrio, J., Fernández, A., Riera, F. A., & Suárez, A. (2014). Immunomodulatory
activities of whey β-lactoglobulin tryptic-digested fractions. International Dairy
Journal, 34(1), 65–73.
Sah, B. N. P., Vasiljevic, T., McKechnie, S., & Donkor, O. N. (2017). Antioxidative and
antibacterial peptides derived from bovine milk proteins. Critical Reviews in Food
Science and Nutrition, 1–15.
Sanchón, J., Fernández-Tomé, S., Miralles, B., Hernández-Ledesma, B., Tomé, D.,
Gaudichon, C., & Recio, I. (2018). Protein degradation and peptide release from milk
proteins in human jejunum. Comparison with in vitro gastrointestinal simulation.
Food Chemistry, 239, 486–494.
73
Journal of Functional Foods 42 (2018) 58–74
Santiago-López, L., Hernández-Mendoza, A., Vallejo-Cordoba, B., Mata-Haro, V., &
González-Córdova, A. F. (2016). Food-derived immunomodulatory peptides. Journal
of the Science of Food and Agriculture, 96(11), 3631–3641.
Sarmadi, B. H., & Ismail, A. (2010). Antioxidative peptides from food proteins: A review.
Peptides, 31(10), 1949–1956.
Sawin, E. A., De Wolfe, T. J., Aktas, B., Stroup, B. M., Murali, S. G., Steele, J. L., & Ney, D.
M. (2015). Glycomacropeptide is a prebiotic that reduces Desulfovibrio bacteria,
increases cecal short-chain fatty acids, and is anti-inflammatory in mice. American
Journal of Physiology-Gastrointestinal and Liver Physiology, 309(7), G590–G601.
Sikarwar, R. L. S., & Kaushik, J. P. (1993). Folk medicines of the Morena district, Madhya
Pradesh, India. International Journal of Pharmacognosy, 31(4), 283–287.
Singh, P. K., Sharma, S., Kumari, A., & Korpole, S. (2014). A non-pediocin low molecular
weight antimicrobial peptide produced by Pediococcus pentosaceus strain IE-3 shows
increased activity under reducing environment. BMC Microbiology, 14(1), 226.
Smithers, G. W. (2015). Whey-ing up the options - Yesterday, today and tomorrow.
International Dairy Journal, 48, 2–14.
Suwal, S., Rozoy, É., Manenda, M., Doyen, A., & Bazinet, L. (2017). Comparative study of
in-situ and ex-situ enzymatic hydrolysis of milk protein and separation of bioactive
peptides in an electromembrane reactor. ACS Sustainable Chemistry & Engineering,
5(6), 5330–5340.
Szymańska, K., Pietrowska, M., Kocurek, J., Maresz, K., Koreniuk, A., Mrowiec-Białoń, J.,
... Jarzębski, A. (2016). Low back-pressure hierarchically structured multichannel
microfluidic bioreactors for rapid protein digestion–proof of concept. Chemical
Engineering Journal, 287, 148–154.
Tavano, O. L. (2013). Protein hydrolysis using proteases: An important tool for food
biotechnology. Journal of Molecular Catalysis B: Enzymatic, 90, 1–11.
Tavares, T. G., Amorim, M., Gomes, D., Pintado, M. E., Pereira, C. D., & Malcata, F. X.
(2012). Manufacture of bioactive peptide-rich concentrates from whey:
Characterization of pilot process. Journal of Food Engineering, 110(4), 547–552.
Tavares, T., del Mar Contreras, M., Amorim, M., Pintado, M., Recio, I., & Malcata, F. X.
(2011). Novel whey-derived peptides with inhibitory effect against angiotensin-
converting enzyme: In vitro effect and stability to gastrointestinal enzymes. Peptides,
32(5), 1013–1019.
Théolier, J., Fliss, I., Jean, J., & Hammami, R. (2014). Antimicrobial peptides of dairy
proteins: From fundamental to applications. Food Reviews International, 30(2),
134–154.
Thomä-Worringer, C., Sørensen, J., & López-Fandiño, R. (2006). Health effects and
technological features of caseinomacropeptide. International Dairy Journal, 16(11),
1324–1333.
Tian, M., Fang, B., Jiang, L., Guo, H., Cui, J., & Ren, F. (2015). Structure-activity re-
lationship of a series of antioxidant tripeptides derived from β-Lactoglobulin using
QSAR modeling. Dairy Science & Technology, 95(4), 451–463.
Tong, J., Li, L., Bai, M., & Li, K. (2017). A new descriptor of amino acids-SVGER and its
applications in peptide QSAR. Molecular Informatics, 36(5–6).
Tsai, C. J., de Laureto, P. P., Fontana, A., & Nussinov, R. (2002). Comparison of protein
fragments identified by limited proteolysis and by computational cutting of proteins.
Protein Science, 11(7), 1753–1770.
Tu, M., Feng, L., Wang, Z., Qiao, M., Shahidi, F., Lu, W., & Du, M. (2017). Sequence
analysis and molecular docking of antithrombotic peptides from casein hydrolysate
by trypsin digestion. Journal of Functional Foods, 32, 313–323.
Tulipano, G., Faggi, L., Nardone, A., Cocchi, D., & Caroli, A. M. (2015). Characterisation
of the potential of β-lactoglobulin and α-lactalbumin as sources of bioactive peptides
affecting incretin function: In silico and in vitro comparative studies. International
Dairy Journal, 48, 66–72.
Udechukwu, M. C., Downey, B., & Udenigwe, C. C. (2018). Influence of structural and
surface properties of whey-derived peptides on zinc-chelating capacity, and in vitro
gastric stability and bioaccessibility of the zinc-peptide complexes. Food Chemistry,
240, 1227–1232.
Udenigwe, C. C. (2014). Bioinformatics approaches, prospects and challenges of food
bioactive peptide research. Trends in Food Science & Technology, 36(2), 137–143.
Uluko, H., Zhang, S., Liu, L., Tsakama, M., Lu, J., & Lv, J. (2015). Effects of thermal,
microwave, and ultrasound pretreatments on antioxidative capacity of enzymatic
milk protein concentrate hydrolysates. Journal of Functional Foods, 18, 1138–1146.
UniProt Consortium (2008). The universal protein resource (UniProt). Nucleic Acids
Research, 36(suppl 1), D190–D195.
Valencia, P., Espinoza, K., Ceballos, A., Pinto, M., & Almonacid, S. (2015). Novel mod-
eling methodology for the characterization of enzymatic hydrolysis of proteins.
Process Biochemistry, 50(4), 589–597.
Vermeirssen, V., Van Camp, J., & Verstraete, W. (2004). Bioavailability of angiotensin I
converting enzyme inhibitory peptides. British Journal of Nutrition, 92(3), 357–366.
Vogel, H. J., Schibli, D. J., Jing, W., Lohmeier-Vogel, E. M., Epand, R. F., & Epand, R. M.
(2002). Towards a structure-function analysis of bovine lactoferricin and related
tryptophan-and arginine-containing peptides. Biochemistry and Cell Biology, 80(1),
49–63.
Vorob’ev, M. M., Butré, C. I., Sforza, S., Wierenga, P. A., & Gruppen, H. (2017).
Demasking kinetics of peptide bond cleavage for whey protein isolate hydrolysed by
Bacillus licheniformis protease. Journal of Molecular Catalysis B: Enzymatic, http://dx.
doi.org/10.1016/j.molcatb.2017.03.005.
Voswinkel, L., Etzel, M. R., & Kulozik, U. (2017). Adsorption of beta-lactoglobulin in
anion exchange membrane chromatography versus the contacting mode and tem-
perature. LWT-Food Science and Technology, 79, 78–83.
Voswinkel, L., & Kulozik, U. (2011). Fractionation of whey proteins by means of mem-
brane adsorption chromatography. Procedia Food Science, 1, 900–907.
Voswinkel, L., & Kulozik, U. (2014). Fractionation of all major and minor whey proteins
with radial flow membrane adsorption chromatography at lab and pilot scale.
International Dairy Journal, 39(1), 209–214.
A. Dullius et al.
Wali, A., Ma, H., Aadil, R. M., Zhou, C., Rashid, M. T., & Liu, X. (2017). Effects of mul-
tifrequency ultrasound pretreatment on the enzymolysis, ACE inhibitory activity, and
the structure characterization of rapeseed protein. Journal of Food Processing and
Preservation. http://dx.doi.org/10.1111/jfpp.13413.
Wang, T. Y., Hsieh, C. H., Hung, C. C., Jao, C. L., Lin, P. Y., Hsieh, Y. L., & Hsu, K. C.
(2017). A study to evaluate the potential of an in silico approach for predicting di-
peptidyl peptidase-IV inhibitory activity in vitro of protein hydrolysates. Food
Chemistry, 234, 431–438.
Welsh, G., Ryder, K., Brewster, J., Walker, C., Mros, S., Bekhit, A. E. D. A., ... Carne, A.
(2017). Comparison of bioactive peptides prepared from sheep cheese whey using a
food-grade bacterial and a fungal protease preparation. International Journal of Food
Science & Technology, 52(5), 1252–1259.
Winkelnkemper, T., & Schembecker, G. (2010). Purification performance index and se-
paration cost indicator for experimentally based systematic downstream process
development. Separation and Purification Technology, 72(1), 34–39.
Wu, W., Zhang, M., Sun, C., Brennan, M., Li, H., Wang, G., ... Wu, H. (2016). Enzymatic
preparation of immunomodulatory hydrolysates from defatted wheat germ (Triticum
Vulgare) globulin. International Journal of Food Science & Technology, 51(12),
74
Journal of Functional Foods 42 (2018) 58–74
2556–2566.
Xie, J., Xu, Z., Zhou, S., Pan, X., Cai, S., Yang, L., & Mei, H. (2013). The VHSE-based
prediction of proteasomal cleavage sites. PLoS ONE, 8(9), e74506.
Yadav, J. S. S., Yan, S., Pilli, S., Kumar, L., Tyagi, R. D., & Surampalli, R. Y. (2015). Cheese
whey: A potential resource to transform into bioprotein, functional/nutritional pro-
teins and bioactive peptides. Biotechnology Advances, 33(6), 756–774.
Yamada, K., Sato-Mito, N., Nagata, J., & Umegaki, K. (2008). Health claim evidence re-
quirements in Japan. The Journal of Nutrition, 138(6), 1192S–1198S.
Yu, Y. J., Amorim, M., Marques, C., Calhau, C., & Pintado, M. (2016). Effects of whey
peptide extract on the growth of probiotics and gut microbiota. Journal of Functional
Foods, 21, 507–516.
Zhang, Q. X., Wu, H., Ling, Y. F., & Lu, R. R. (2013). Isolation and identification of an-
tioxidant peptides derived from whey protein enzymatic hydrolysate by consecutive
chromatography and Q-TOF MS. Journal of Dairy Research, 80(3), 367–373.
Zhong, J., Cai, X., Liu, C., Liu, W., Xu, Y., & Luo, S. (2016). Purification and conforma-
tional changes of bovine PEGylated β-lactoglobulin related to antigenicity. Food
Chemistry, 199, 387–392.