Novel Functional Whey-Based Drinks with Great Potential in the Dairy Industry

About 33 % of the milk production worldwide is used in cheese manufacture, averaging about 19·106 tonnes per year (1). Whey is a by-product obtained during the production of cheese that can cause environmental contamination. It is basically composed of lactose and proteins of high biological value. Because of their nutritional importance, whey powder, whey proteins and whey protein fractions represent a global market value of approx. 6.3 billion euros (2). However, the industrial production of whey powder, whey protein concentrates, whey protein isolates, microparticulated whey proteins or polymerized whey proteins normally involves a fi nal drying step, generally carried out by spray drying. The ISSN 1330-9862 original scientifi c paper


Introduction
About 33 % of the milk production worldwide is used in cheese manufacture, averaging about 19 •10 6  tonnes per year (1).Whey is a by-product obtained during the production of cheese that can cause environmental contamination.It is basically composed of lactose and proteins of high biological value.Because of their nutri-tional importance, whey powder, whey proteins and whey protein fractions represent a global market value of approx.6.3 billion euros (2).However, the industrial production of whey powder, whey protein concentrates, whey protein isolates, microparticulated whey proteins or polymerized whey proteins normally involves a fi nal drying step, generally carried out by spray drying.The

Novel Functional Whey-Based Drinks with Great Potential in the Dairy Industry
Carlos Pereira 1 , Marta Henriques 1 *, David Gomes 1 , Andrea Gomez-Zavaglia 2 and Graciela de Antoni 2,3  application of this operation is usually limited to large dairy industries.In small and medium scale cheese plants, the volume of whey to be processed does not justify the costs associated with the installation of such equipment.For these plants, whey disposal represents a major problem and, in most cases, a high operational cost.For this reason, the development of novel strategies that avoid the spray drying step contributes to the value of this by-product in small and medium industries.
A group of novel products obtained from whey proteins has been developed in the last years aiming to replace fats in food.These products include the use of microparticulated whey proteins manufactured from whey protein concentrates in a process involving heating and shearing (3,4), or the thermal denaturation of polymerized whey proteins to obtain soluble whey protein that does not aggregate because of the absence of salts (5)(6)(7).The use of β-lactoglobulin and whey protein isolate aggregates to obtain cold-set gels, foams and emulsions, encapsulation and fi lms/coatings has recently been reviewed (8).
In general, whey proteins are used in the form of dry whey proteins, whey protein concentrates, microparticulated whey proteins or polymerized whey proteins (9)(10)(11)(12) as ingredients in mixtures for yoghurt production.The production and direct incorporation of liquid whey protein concentrates in food products is seldom reported.Recently, liquid whey protein concentrates have been successfully produced by ultrafi ltration, and then incorporated into fresh cheese and set yoghurt (13,14), providing a solution for the immediate reutilization of whey and avoiding evaporation and drying.However, the use of liquid whey protein concentrates as a substrate for the fermentation of probiotic bacteria provides an extra added value and has never been addressed.Kefi r grains include up to 400 diff erent species of probiotic bacteria and yeasts and are able to grow in whey permeate (15).Several benefi cial health eff ects have been ascribed to kefi r, namely modulation of the immune response as well as the antibacterial and antifungal activities, among others (16)(17)(18)(19)(20).
The present approach envisages the utilization of liquid whey protein concentrates or ultrafi ltrated permeate concentrated by reverse osmosis as substrates for the production of fermented drinks.To this aim, liquid whey protein concentrates and concentrated ultrafi ltrated permeates were fermented with: (i) well characterized kefi r grains (16) containing microorganisms with inhibitory properties against pathogenic bacteria (17)(18)(19) or parasites (20), (ii) a commercial mix of probiotic bacteria and, (iii) their mixture, in order to obtain novel fermented dairy products.

Liquid whey protein concentrate and concentrated ultrafi ltrated permeate production
Cheese whey was obtained immediately aft er the production of semi-hard bovine cheese at the pilot plant of Escola Superior Agrária de Coimbra (ESAC, Coimbra, Portugal).Cheese coagulation (45 min) was performed in the presence of 0.04 % (by mass per volume) of CaCl 2 solution (Betelgeux S.A., Gandía, Valencia, Spain), 10 ppm of Mesófi lo Plus Starter (Abiasa, Tui, Spain), 25 ppm of Lysozyme and 20 ppm of rennet (> 92 % by mass of chimosin, Tecnilac, Mundão, Portugal) at 33 °C.Aft er production, whey was fi ltered, analysed and processed.The production of liquid whey protein concentrates (Fig. 1) consisted of whey concentration at 45-50 °C in a batch ultrafi ltration pilot plant (Proquiga, Biotech S.A., La Coruña, Spain), using an organic polyvinylidene fl uoride membrane UF 3838, with 10-kDa cut-off (Iberlact, Madrid, Spain), aiming for a volumetric concentration factor (volume feed/volume retentate) of 15.Aft er ultrafi ltration, the concentrate was thermally treated (at 90-95 °C for 5 min) under batch conditions in order to denature whey proteins.The heat treatment also pasteurized the product.The mixture was then homogenized at 10 MPa using a homogenizer Rannie™ model Bluetop (Copenhagen, Denmark) to achieve a particle diameter of whey protein aggregates lower than 10 μm to avoid graininess and increase smoothness of the concentrate.The ultrafi ltrated permeate was concentrated by reverse osmosis using a pilot module (ORM -Tecnologia e Ciência na Indústria Lda, Belas, Portugal) equipped with a membrane model 2.5 S Seawater (Advanced Structures Inc., Los Angeles, CA, USA) aiming at a volumetric concentration factor of 2, in order to obtain the concentrated ultrafi ltration permeate, which was subsequently pasteurized under the same conditions as the liquid whey protein concentrates.Aft er processing, both liquid whey protein concentrates and concentrated ultrafi ltrated permeates were refrigerated (<5 °C) or frozen (-18 °C) for further utilization.

Production of fermented drinks
Liquid whey protein concentrates or concentrated ultrafi ltrated permeates (fresh or thawed) were used as culture media for the production of fermented drinks.Mango fruit pulp (10 %, by mass; Continente, Coimbra, Portugal), sucrose (5 %, by mass) and mango fl avouring (500 ppm; Aromas & Sabores, Póvoa de Santo Adrião, Portugal) were added to liquid whey protein concentrates and concentrated ultrafi ltrated permeates and the mixtures were acclimated at 25 °C before the addition of the inoculum.Three types of inocula were used: (i) fresh or thawed kefi r grains CIDCA AGK1 (2.5 %, by mass per volume; Centro de Investigación y Desarollo en Criotecnología de Alimentos, La Plata, Argentina), (ii) a commercial mix of probiotic bacteria containing Lactobacillus acidophilus, Lactobacillus casei and Lactobacillus rhamnosus in a ratio of 1:1:1 (0.3 %, by mass per volume; Sacco™, Lyofast ACR, Milan, Italy), and (iii) a mixture of (i) and (ii) in a 1:1 ratio.
The kefi r grains used as inoculum for the production of fermented drinks were added, either fresh (incubated overnight at 25 °C in milk as culture media), or aft er thawing frozen grains.In both cases, the kefi r grains were added to the formulations into a mesh bag during the fermentation period and removed before storage.
The inoculated liquid whey protein concentrates and concentrated ultrafi ltrated permeates were incubated at 25 °C for 24 h.Subsequently, the products were refrigerated to below 5 °C and kept under these conditions for 14 days.Fig. 1 presents the production process diagram of the fermented drinks.

Physicochemical analyses
The pH, titratable acidity, and the content of total solids, ash and total proteins of dairy products were determined according to the Offi cial Methods of Analysis (21).The pH was determined directly with a pH meter (HI 9025; HANNA Instruments, Ronchi di Villafranca, Padova, Italy).Titratable acidity, expressed as a percentage of lactic acid, was determined aft er titration using 0.1 M NaOH solution.Total solids were determined aft er oven drying the samples at 105 °C.Ash content was obtained by incineration of dry samples in an electric muffl e furnace (model LE 4/11/R6; Nabertherm, Lilienthal, Germany).Total nitrogen content was determined using the Kjeldahl method (21).Total protein content was estimated by multiplying the total nitrogen content of the samples by a factor of 6.38.The fat content of liquid whey protein concentrate and concentrated ultrafi ltrated permeate products was determined by the Gerber method (22).Each product sample was collected using the Portuguese specifi c standard procedure for dairy products (23).
Viscosity was evaluated with a Brookfi eld™ viscometer (Brookfi eld Engineering Laboratories, Madison, WI, USA) and expressed in mPa•s.Viscosity measurements of drinks with liquid whey protein concentrates were made at 8 °C and 1 rpm for 30 s with spindle 01 and of drinks with concentrated ultrafi ltrated permeates with the device UL Adapter (Brookfi eld Engineering Laboratories) at 100 rpm.

Microbiological analysis
Enumeration of viable microorganisms was carried out immediately aft er inoculation (at 0 h), during the fermentation period (at 12 and 24 h) and during refrigerated storage at 5 °C (at 48, 168 and 336 h).Counts of presumptive lactococci and lactobacilli were carried out on M17 agar BK088 and on MRS agar BK089 (Biokar Diagnostics, Beauvais, France), respectively, aft er aerobic incubation at 37 °C for 48 h, according to the standard method specifications (24).Yeasts were enumerated on Rose Bengal chloramphenicol agar BK151 (Biokar Diagnostics) aft er incubation at 25 °C for 72 h.

Sensorial analysis
Consumer preference tests were performed with a panel of 35 persons with experience in sensorial evaluation.Products were tested on the 7th day of storage and panellists were asked to grade the diff erent products on a scale from 1 (dislike) to 5 (like a lot).The sensory acceptability of the products was evaluated by a group of six trained members of the staff on an unacceptable/acceptable basis on the 14th day of storage.

Statistical analysis
Experiments were done in duplicate using three independent cultures of bacteria in each experiment.Analysis of variance (ANOVA) corresponding to the diff erent assays was carried out using the soft ware STATISTICA v. 8 (StatSoft Inc., Tulsa, OK, USA) (25).Comparison of mean values by Tukey's method was performed, and if p<0.05, the diff erences were considered statistically signifi cant.

Chemical composition of fermented drinks
Membrane technologies applied to cheese whey allowed for the production of a protein-enriched fraction and a lactose-enriched fraction.The use of such fractions as substrates for the production of fermented drinks as carried out in this work can be a reasonable alternative for in-plant valorization of cheese whey.
Table 1 shows the chemical composition of the fermented drinks.As expected, the levels of total solids, proteins and fat are higher in products based on liquid whey protein concentrates than in those obtained with concentrated ultrafi ltrated permeates, in which lactose is the main solid component (approx.80 % of total solids).The presence of proteins and fat in concentrated ultrafi ltrated permeates results from the permeate concentration by reverse osmosis, which doubled the residual levels of these components in the ultrafi ltrated permeate, as well as from the low levels of proteins and fat in the added fruit pulp (approx. 1 %, by mass).

pH and acidity during fermentation and storage
Table 2 shows the changes in the pH and titratable acidity values of the products during fermentation (at 0 and 24 h) and aft er 7 days of storage (at 168 h).Products with liquid whey protein concentrates had an initial pH in the range of 5.37-5.50,and aft er fermentation the pH was reduced to values lower than 4.5.During refrigerated storage, these values decreased to 4.2-4.3.In these products the titratable acidity increased from approx.0.039 to 0.077-0.087% of lactic acid.It can also be observed that the products with liquid whey protein concentrates containing the commercial probiotic cultures had lower acidity than the ones containing kefi r grains or the mixture of kefi r grains and commercial probiotic bacteria.
The initial pH of drinks with concentrated ultrafi ltrated permeates was signifi cantly lower than of products with liquid whey protein concentrates, most probably as a result of the absence of whey proteins that have a buff er capacity, and also of the acidifi cation that occurred during the concentration by reverse osmosis (18-25 °C, 2 h).
These values decreased to less than 4.5 aft er fermentation (t=24 h), with the exception of the products containing the commercial probiotic mix, in which it did not change in the fi rst 24 h.This may be explained by the fact that the composition of these products does not favour the growth of the probiotic starter culture.However, aft er 7 days of storage, the pH of the products was similar to the ones obtained using kefi r grains as inoculum.It appears that mixing of kefi r grains with probiotic bacteria had a synergistic eff ect on the growth of both cultures (Table 2) at high lactose mass fractions such as in products with concentrated ultrafi ltrated permeates.In these products, differences in acidity were not signifi cant (p>0.05).As it can be observed in Table 2, drinks with concentrated ultrafi ltrated permeates had much lower acidity than those with liquid whey protein concentrates.Despite the high acidity developed in the latt er, the pH remained in the range of 4.2-4.5 as a result of the buff er eff ect of whey proteins.

Viscosity of fermented drinks
As expected, products with liquid whey protein concentrates had higher viscosity than the products with concentrated ultrafi ltrated permeates (Table 3).This refl ects the higher levels of solids, proteins and fat in the former.A signifi cant (p<0.05)decrease of viscosity of drinks with liquid whey protein concentrates from day 1 to day 7 (at 168 h) was observed especially when fermented with kefi r grains.Viscosity of the products with concentrated ultrafi ltrated permeates slightly increased during storage.
Due to their low molecular mass and spherical shape, native whey proteins are poor thickening agents.The thermal aggregation of whey proteins increases their hydrodynamic volumes, thus favouring their thickening properties.In this case, the thermal treatment of the liquid whey protein concentrates allowed for the formation of such aggregates, while the homogenization step eliminated graininess and increased smoothness, being determinant for the fi nal viscosity of the product.Products containing liquid whey protein concentrates fermented with commercial probiotic bacteria or with the mixture of kefi r grains and commercial probiotic bacteria had significantly higher viscosity than the ones containing just kefi r Table 2.The pH and titratable acidity during fermentation and storage of whey drinks produced using liquid whey protein concentrates and ultrafi ltrated and concentrated permeates with kefi r grains, commercial mix of probiotics, and kefi r grains combined with commercial mix of probiotics grains.This fact can be explained by the lower solid contents of the latt er (Table 1).
Viscosity of the products with liquid whey protein concentrates was almost three times higher than the ones reported for cow's and goat's milk yoghurts of approx.3.4 % protein, in which polymerized whey proteins and pectin were used as gelling agents (7).This can be explained considering the high level of protein, which is nearly twice as high in products with liquid whey protein concentrates (>6 % protein).The viscosity of products with liquid whey protein concentrates is 2-3 times lower than the one reported for yoghurts in which diff erent commercial whey protein concentrates were used to enrich the protein content of milk up to 4.5 % (by mass) (26), but it is of the same order of magnitude as the one obtained for low-fat yoghurts, in which microparticulated whey proteins were used as fat replacers, and with protein levels between 4.25 and 5.0 % (by mass) (12).

Microbiological profi le of fermented products
Concerning microbial enumeration (Fig. 2), it was observed that freezing and then thawing kefi r inoculum negatively aff ected the counts of lactobacilli, lactococci and yeasts in drinks with liquid whey protein concentrates and concentrated ultrafi ltrated permeates (Figs.2c  and d).As expected, when thawed kefi r grains were used, microbial counts were in the order of 2 log below the ones observed in products obtained aft er inoculation with fresh kefi r grains (Figs.2a and b).It is therefore concluded that freezing and thawing the inoculum is a harmful process for microorganisms, and if kefi r grains are to be kept frozen for storage, they require a prior activation at 25 °C for 24 h before inoculating the products.
On the other hand, as it can be observed in Fig. 3, freezing liquid whey protein concentrates and keeping them frozen for one month did not aff ect its ability as culture medium to produce fermented drinks with kefi r grains, commercial probiotic bacteria, or both.Hence, this procedure may be used to preserve liquid whey protein concentrates in small/medium size dairy factories which do not possess evaporation/drying equipment.Fig. 2. Microbiological profi le of the fermented drinks: a) liquid whey protein concentrates fermented with fresh kefi r grain inoculum, b) concentrated ultrafi ltrated permeates with fresh kefi r grain inoculum, c) liquid whey protein concentrates with thawed kefi r grain inoculum, d) concentrated ultrafi ltrated permeates with thawed kefi r grain inoculum.Fresh kefi r grains were obtained by overnight fermentation of milk at 25 °C and thawed kefi r grains were added aft er thawing frozen kefi r grains previously prepared under the same conditions Table 3. Viscosity of the drinks obtained by fermentation of liquid whey protein concentrates and ultrafi ltrated and concentrated permeates with kefi r grains, commercial mix of probiotics, and kefi r grains combined with commercial mix of probiotics The number of yeasts was above 6 log CFU/mL, except for the products with solely commercial probiotic bacteria.Counts of lactococci and lactobacilli were above 7 log CFU/mL in all the fermented drinks (Fig. 3), being similar to those reported for yoghurts produced with milk (12 % of total solids) and milk and whey bases (8 and 10 % of total solids) (27) or to whey-based fermented beverages (28).The highest microbial counts of lactobacilli and lactococci were observed in products containing the mixture of kefi r grains and probiotic bacteria (Figs.3e and  f), which is consistent with the values obtained for lactic acid production (Table 2).Microbial counts of viable lactobacilli, lactococci and yeasts were still high aft er 14 days of storage, showing the resistance of the strains towards storage conditions.It has been reported that the benefi ts of probiotic consumption can be obtained when the food products contain at least 6-7 log CFU/mL of viable microorganisms per g of product at the end of shelf life (29,30).The probiotic properties of kefi r microorgan-isms are well known (16)(17)(18)(19)(20), and fermented drinks are usually consumed within 15 days of storage.Therefore, the use of fresh kefi r grains and/or commercial probiotics as inocula of liquid whey protein concentrates or concentrated ultrafi ltrated permeates appears as an excellent choice for the production of functional fermented drinks.
If particular probiotic bacteria are used in combination with kefi r microorganisms in fermented drinks based on liquid whey protein concentrates, most of the β-lactoglobulin and lactose can be removed, reducing to a minimum the potential allergenicity or lactose intolerance of such products, which have high concentrations of essential and sulphur-containing amino acids (28,31).

Sensory evaluation
Fermented drinks with liquid whey protein concentrates obtained using kefi r grains or the mixture of kefi r grains and commercial probiotic bacteria as inocula reb) a) N/(log CFU/mL) Fig. 3. Microbiological profi le of the fermented drinks produced from liquid whey protein concentrates: a) kefi r with fresh liquid whey protein concentrates, b) kefi r with thawed liquid whey protein concentrates, c) commercial mix of probiotics with fresh liquid whey protein concentrates, d) commercial mix of probiotics with thawed liquid whey protein concentrates, e) kefi r combined with commercial mix of probiotics with fresh liquid whey protein concentrates, f) kefi r combined with commercial mix of probiotics with thawed liquid whey protein concentrates ceived bett er scores in sensory evaluation (Fig. 4).Indeed, almost 60 % of the consumers gave them 4 or 5 points (on a fi ve-grade scale).Products with concentrated ultrafi ltrated permeates were less appreciated, as less than 25 % of consumers gave them 4 or more points (Fig. 5).
Among products with liquid whey protein concentrates, preferred was the one fermented with kefi r grains, while among products with concentrated ultrafi ltrated permeates the one fermented with the commercial probiotic culture received the highest score.Consumers referred that products with liquid whey protein concentrates had the characteristics of yoghurt but with a palatable carbonated taste associated with gas production by kefi r microorganisms.A general comment about products with concentrated ultrafi ltared permeates was that they did not have the typical characteristics of a dairy product, being associated with low quality fruit juices.Aft er 14 days of storage the products still kept acceptable sensory characteristics when evaluated by an experienced panel.Further work is needed in the development of these products in order to introduce diff erent fruits and to improve the consumer acceptability of the products.In the developed products, the use of fruit pulp masked the possible development of turbidity during their shelf life.If the intention is to produce the same type of fermented drinks without fruit pulp this aspect should be endorsed.

Conclusion
The impact of this work comes from two main aspects: (i) the use of liquid whey protein concentrates and ultrafi ltrated and concentrated permeates, and (ii) the production of palatable fermented drinks.It must be underlined that obtaining liquid whey protein concentrates and ultrafi ltrated and concentrated permeates resulted in the complete use of whey, that is, only water remained as whey by-product.From this perspective, the use of liquid whey protein concentrates and ultrafi ltrated and concentrated permeates certainly contributes to the protection of environment.This strategy may also be extremely att ractive for medium/small producers, who intend to develop integrated solutions for the cheese whey problem by viewing its great potential as a source of added value to incorporate into own products.
On the other hand, the counts of probiotic microorganisms aft er 14 days of storage at refrigeration temperatures are compatible with those required by international organisations like EFSA or FDA to provide the benefi ts of consuming probiotics.
On the whole, the strategy developed in this work can be considered a valuable contribution to expand the applications of products derived from whey fractionation in medium/small cheese industries for the design of new functional foods.

Fig. 4 .Fig. 5 .
Fig. 4. Sensory evaluation of fermented drinks produced from liquid whey protein concentrates.Number of the consumers that classifi ed products in the indicated range is expressed in percentage

Table 1 .
Overall chemical composition of fermented products obtained by inoculation of liquid whey protein concentrates and ultrafi ltrated and concentrated permeates with kefi r grains, commercial mix of probiotics, and kefi r grains combined with commercial mix of probiotic bDiff erent lett ers in superscript indicate statistically signifi cant diff erences (p<0.05).LWPC=liquid whey protein concentrate, UFCP=ultrafi ltrated and concentrated permeate, k=kefi r grains, p=commercial mix of probiotics, k+p=kefi r grains with commercial mix of probiotics, TA=titratable acidity as percentage of lactic acid