EFFECTS OF HIGH INTENSITY PULSED ELECTRIC FIELDS OR THERMAL PASTEURIZATION AND REFRIGERATED STORAGE ON ANTIOXIDANT COMPOUNDS OF FRUIT JUICE-MILK BEVERAGES. PART I: PHENOLIC ACIDS AND FLAVONOIDS

The effects of High Intensity Pulsed Electric Fields (HIPEF) or thermal pasteurization (TP) over phenolic compounds in mixed beverages were evaluated after processing and during chilled storage, having untreated beverage as reference. Total phenolic (TPC, 57.0 – 58.8 mg of galic acid/100mL) and flavonoid (TFC, 4.14 – 4.33 mg of quercetin/100mL) contents remained constant in fruit juice-skim milk (FJ-SM) and whole milk (FJ-WM) beverages just after HIPEF or TP. Nonetheless, concentration of most individual phenolics augmented. TPC in HIPEF treated beverages remained constant through storage, while in thermally pasteurized beverages tended to decrease (5 – 15%). No significant changes were observed in TFC in untreated and treated beverages over time. The concentration of individual phenolics in fresh and treated beverages remained constant or decrease with time, except hesperidin, which significantly increased (19 – 61%) after 56 days. Hence, HIPEF is a feasible technology to obtain mixed beverages with antioxidant properties.


INTRODUCTION
Phenolic compounds represent a group of secondary metabolites, being the most abundant antioxidants in human diets (Erlund, 2004). With over 8,000 structural variants, phenolic acids and flavonoids are the largest subclasses of polyphenols widely distributed in plant kingdom (Han et al., 2007). It is well known that fruits are a good source of these compounds, which possess antimicrobial, antiviral and antiinflammatory properties (Ignat et al., 2011). Additionally they are considered as multifunctional antioxidants and can act as oxidative enzymes inhibitors, metal chelators or free radical scavengers (Boskou et al., 2006). Furthermore, it has been reported that phenolic acids and flavonoids are able to inhibit lipid peroxidation, protect low-density lipoproteins, reduce platelet aggregation and enhance vasodilatation (Vinson and Dabbagh, 1998;Fuhrman et al., 1995;Renaud andDelorgeril, 1992 andFitzpatrick et al., 1993). In this way, it is believed that these compounds induce health benefits and have an important role in the protection against a number of pathological disturbances (Gordon, 1996).
Nowadays, dietary habits of modern population are changing and consumers are showing a growing interest on health promoting foods. As a result, the development of new fruit-based products such as functional beverages is increasing worldwide, not only to boost their sensory acceptability but also their nourishing properties through the combination of different ingredients such as fruit juices and milk. Mixed beverages represent a potential way to contribute to consumers health due to their elevate concentration of phytochemicals from fruits and other bioactive compounds of milk.
Different studies demonstrate that fruit juice-milk beverages are a good source of vitamins, having at the same time, high antioxidant capacity (Salvia-Trujillo et al., 2011;Zulueta et al., 2010).
Though thermal pasteurization (TP) is the most used method to make this kind of beverages safe and shelf-stable, the high temperatures achieved during processing cause detrimental effects on some desirable food constituents, physicochemical characteristic, flavor attributes and antioxidant properties (Plaza et al., 2006). Thus, overcoming these changes is one of the major challenges in the food industry to fulfil consumers demand.
At present, emerging technologies are being explored to process foods at low temperatures avoiding the negative effects induced by heat. Among them, high intensity pulsed electric fields (HIPEF) has demonstrated to be a gentle food preservation technique capable of providing fresh-like characteristics and shelf-stable products with minimum changes on their nutritional, physicochemical, and sensorial properties (Elez-Martinez and Martín-Belloso, 2007;Soliva-Fortuny et al., 2009).
To the best of the authors' knowledge, limited information is available concerning the effects of HIPEF on health-related compounds in mixed beverages containing fruit juices and milk (Zulueta, et al., 2010;Rivas et al., 2007) and data on phenolic acids and flavonoids content is scarce. Hence, the aim of this research was to evaluate the phenolic acid and flavonoid profile of fruit juice-milk beverages prepared with whole or skim milk and compare the effects of HIPEF and TP over these compounds immediately after processing and during the storage (56 days) at 4 °C.

Beverage preparation
Orange, mango, kiwi and pineapple fruits were purchased in a local supermarket (Lleida, Spain) at commercial maturity stage. Fruit was sanitized in a 200 ppm sodium hypochlorite solution for 2 min and rinsed with a tap water. Then the juice was extracted from each fruit separately and mixed with commercial pasteurized whole (3.5% fat) or skim (0.3% fat) bovine milk with the following proportions: orange (30%), kiwi (25%), mango (10%), pineapple (10%), milk (17.5%), and sugar (7.5%) These proportions, of each ingredient, were selected in basis of previous studies in order to maximize the content of vitamin C in the beverages (Salvia-Trujillo et al., 2011a).
Fruit juice-whole milk (FJ-WM) or -skim (FJ-SM) beverages were filtered through a cheese cloth, and their pH was adjusted to 3.35 with citric acid in order to have an acidified product and avoid microbial growth. Physicochemical characterization of the beverages was previously evaluated in terms of electrical conductivity, pH, total acidity and soluble solids content (Salvia-Trujillo et al., 2011a).

HIPEF processing
A continuous flow bench-scale system OSU-4F (The Ohio State University, Colombus, OH), delivering square-wave pulses, was used for HIPEF processing. According to a previous study, the HIPEF process was conducted at 35 kV/cm electric field strength for 1800 μs, a pulse frequency of 200 Hz, and 4 μs bipolar pulses in order to inactivate Listeria innocua, guaranteeing product safety (Salvia-Trujillo et al., 2011a). Electric field strength, pulse duration and frequency were controlled through a pulse generator (model 9410, Quantum Composers, Inc., Bozeman, MT) and measured with an oscilloscope (TEKScope, Tektronix Inc., Beaverton, OR). The samples were pumped through the system at a flow rate of 760 mL/min with a variable gear pump (model 752210-26, 106 Cole Palmer Instrument Co., Vermon Hills, IL). The system was composed of eight collinear treatment chambers serial connected, each with two stainless steel electrodes separated by 0.292 cm. Each treatment chamber has a diameter of 0.23 cm and a volume of 0.0121 cm 3 . Between each treatment chamber the product was refrigerated in an ice-water bath so that the temperature of the product was always below 40 °C, which was measured with thermocouples at the inlet and outlet of each treatment chamber.

Thermal treatment
Beverages were thermally pasteurized at 90 °C for 1 min to ensure the inactivation of spoilage microorganisms and to simulate a conventional preservation treatment based on literature (Nagy et al., 1993). The samples were pumped with a peristaltic pump (model D-21 V, Dinko, Barcelona, Spain) at a flow rate of 40 mL/min and passed through a tubular stainless steel heat exchanger coil system (0.037 cm 2 section and 1100 cm long) submerged in a hot water bath settled at 90 °C (Universitat de Lleida, Spain).
Then the heated beverages were immediately cooled in a water bath with ice passing through a stainless steel coil.

Packaging and storage
HIPEF and TP fluid systems were disinfected first with 4% of NaOH and then with 10% chlorine and 20% ethanol solutions prior to processing. Polypropylene sterile bottles of 100 mL were used to store the FJ-WM and FJ-SM beverages. Once filled, the receptacle was tightly close and stored at 4 ± 1 °C in the absence of light and with minimal headspace volume. According to a previous study (Salvia-Trujillo et al., 2011a). HIPEF and TP treatments are able to inactivate mesophilic, mould and yeast populations, leading to a shelf-life of 56 days. Non-treated beverages were stored for just 14 days owing to the rapid growth of spoilage microorganisms.

Total flavonoid content
Total flavonoids compounds (TFC) were extracted with 5% NaNO2, 10% AlCl3 and NaOH 1M and measured spectrophotometrically at 415 nm using quercetin as standard (Dae-Ok et al., 2003;Meda et al., 2005). The results were expressed as mg of quercetin equivalents (QE) per 100 mL of beverage.

Total phenolic content
Total phenolic compounds (TPC) were determined by the colorimetric method described by Singleton et al. (1999) using the Folin-Ciocalteu reagent. An aliquot of 0.5 mL of the beverage was mixed with 0.5 mL of Folin-Ciocalteu reagent and 10 mL of saturated Na2CO3 solution. Samples were kept at room temperature for 1 h. After this time, absorbance at 725 nm was measured using a CECIL 2021 spectrophotometer (Cecil Instruments Ltd., Cambridge, UK). Concentrations were determined by comparing the absorbance of the samples with a calibration curve built with solutions of 0, 100, 250, 500 and 1000 mg galic acid/100 mL (Scharlau Chemie, SA, Barcelona, Spain). Results were expressed as mg of galic acid per 100 mL of beverage.

Individual phenolic compounds
Individual phenolic compounds were extracted and quantified by HPLC, following a procedure validated by Hertog et al. (1992).

Extraction and hydrolysis
To perform the phenolic compounds extraction, untreated and treated beverages were first frozen at -30 °C and then, freeze-dried (Telstar Cryodos-80 Freeze-Dryer) at -44.6 ± 5 °C and vacuum at 0.131 mBar. Lyophilized samples were stored at room temperature until analysis. Twenty millilitres of 62.5% aqueous methanol with 2 g/L of tert-butylhydroquinone and 5 mL of hydrochloric acid 6M were carefully mixed with 0.5 g of freeze-dried beverage. After refluxing at 90 °C for 2 h with regular swirling, the extract was cooled and subsequently made up to 50 mL with methanol and sonicated for 5 min. The extract was then passed through a 0.45 μm filter prior to injection.

Chromatography conditions
An aliquot of 20 μL of the extracted samples was injected into the HPLC system, which was equipped with a 600 Controller, a 486 Absorbance Detector programmed to scan from 200 to 350 nm, a thermostatic column compartment, and a 717 Plus Auto Sampler with cooling system (Waters, Milford, MA). Phenolic compounds were separated following the procedure described by Morales-de la Peña et al. (2011) using a reversephase C18 Spherisorb ODS2 (5 μm) stainless steel column (4.6 mm x 250 mm) at room temperature with flow rate of 1 mL/min. A gradient elution was employed with a solvent mixture of 2.5% HCOOH in water (solvent A) and 2.5% HCOOH in methanol (solvent B) as follows: linear gradient from 5% to 13% B, 0-15 min; linear gradient from 13% to 15%B, 15-20 min; linear gradient from 15% to 30%, 28-32 min; isocratic elution 45% B, 32-35 min; linear gradient 45% to 90% B, 35-40 min; isocratic elution 90% B, 40-45 min; linear gradient to reach the initial conditions after 5 min; post-time 10 min before the next injection. Individual phenols were identified by comparison of their UV-vis spectral data and retention times with those of the reference standards (chlorogenic, caffeic, p-coumaric, ferulic, and sinapic acids; hesperidin, rutin, narirutin and quercetin). Quantification of phenolic compounds was carried out by integration of the peak areas. Data were compared to calibration curves of each phenolic compound and results were expressed as mg of phenolic compound per 100 mL of beverage.

Statistical analysis
Treatments were conducted in duplicate and two replicate analyses were carried out for each sample. Analysis of the variance (ANOVA) was performed to compare treatments.
Least significance difference (LSD) test was employed to determine differences between means immediately after processing and throughout the storage. The confidence interval was set at 0.95 for analysis and procedures. Results were analysed using the Statgraphics Plus v.5.1 Windows package (Statistical Graphics Co., Rockville, Md).

Total phenolic compounds
Initial concentration of TPC in FJ-SM and FJ-WM beverages varied from 57.0 to 58.8 mg of galic acid/100 mL. These values were within the range of those reported by Zulueta et al. (2007) in commercial beverages containing a blend of diverse fruit juices and milk (25.5 to 99.8 mg of galic acid/100 mL). Nonetheless, the dissimilarities in total phenolic content between different beverages could be mainly attributed to the ripening degree and environmental growing conditions of the fruits employed for the formulation (Fernández de Simon et al., 1992;Spanos and Wrolstad, 1992). Immediately after processing, no significant changes were observed among untreated FJ-SM (58.3 ± 3.25 mg of galic acid/100 mL) and FJ-WM (55.3 ± 0.96 mg of galic acid/100 mL); HIPEF  (2009) did not find significant changes in TPC among fresh, HIPEF-treated and heatpasteurized tomato and carrot juices, respectively.
As can be seen in Figure 1, HIPEF treated FJ-SM and FJ-WM beverages retained the initial content of TPC for a period of 56 days at 4 °C, although some fluctuations were observed during the storage. Different to our results, Zulueta et al. (2013) reported that storage time was a significant factor for TPC in an orange juice-milk beverage processed by HIPEF and stored at refrigeration conditions. They found a significant increase on TPC on treated samples at the end of the storage. According to Prior et al. (2005) the Folin-Ciocalteu protocol for total phenols determination may also includes the contribution from ascorbic acid, reducing sugars, soluble proteins and other substances. This fact can explain the fluctuation and increase in TPC of HIPEF treated mixed beverages, mainly due to the formation of other compounds during the storage period that can react with Folin-Ciocalteu phenol reagent. Conversely, a significant decrease of the TPC concentration was observed in thermally treated FJ-SM (15%) and FJ-WM (5%) beverages during the storage period. In agreement to our data, Morales-de la Peña et al. (2010) reported that the concentration of TPC in heat treated fruit juicesoymilk beverages diminished after the third day of storage at 4 °C. Consistent with Kumar-Roy et al. (2007), TP affects in a considerable way, phenolic content in vegetables. Therefore, the elevated temperatures achieved during the conventional pasteurization process (90 °C, 60 s) might have affected some phenolic compounds in the mixed beverages, making them easily degradable over time.

Total flavonoid compounds
Initial content of TFC in untreated FJ-SM and FJ-WM beverages was 4.14 ± 0.18 and 4.33 ± 0.09 mg of quercetin/100 mL, respectively. No significant differences in TFC were observed between HIPEF and TP processed FJ-WM beverages just after processing with respect to the untreated sample (Figure 2, day 0). Nonetheless, HIPEF and thermally treated FJ-SM beverages showed a significant decrease in TFC reducing their concentration up to 3.68 ± 0.26 and 3.00 ± 0.35 mg of quercetin/100 mL, respectively (Figure 3, day 0). To the best of the authors' knowledge, there are no studies reporting the concentration TFC of mixed beverages containing fruit juices and whole or skim milk, neither the effects caused by TP nor HIPEF treatments in these compounds contained in complex matrix. However, Gil-Izquierdo et al. (2002) reported that mild and standard pasteurization processes did not influence the flavanone content of an orange juice. But, when the juice was subjected to a concentration process the flavanone content slightly decreased with respect to the content before concentration.
Otherwise, in a previous study, Gil-Izquierdo et al. (2001), suggested that orange juice flavones decreased as a result of the TP process. As can be seen in Figure 2

Individual phenolic profile
The initial phenolic profile, including phenolic acids and flavonoids, of untreated and treated FJ-SM and FJ-WM beverages is shown in Figure 4. Five phenolic acids including caffeic, chlorogenic, coumaric, ferulic and sinapic; and four flavonoids, hesperidin, rutin, narirutin and quercetin, were identified in both beverages regardless of the treatment applied. As can be seen in Figure 4, chlorogenic was the main hydroxycinnamic acid derivative, obtained in concentrations of 38.34 to 51. 42 mg/100 mL, while hesperidin was the most abundant flavonoid (11.23 -15.64 mg/100 mL) present in untreated and treated samples. In accordance to our results, Morales-de la Peña et al. (2011) found a similar phenolic profile in a mixed beverage containing fruit juices and soymilk. However, authors reported that coumaric acid and narirutin were the major phenolic compounds present in the beverages contributing with 32 -46% and 19.5 -27.5% of the total content, respectively. Generally, phenolic concentration in fruits and vegetables is highly variable and is strongly influenced by the maturity stage, growing areas, variety and conditions of storage and processing.
Immediately after HIPEF or TP processing, the concentration of most individual phenolic acids and flavonoids identified in both beverages increased or remained with no significant changes (Tables 1 and 2). Interestingly, HIPEF process better maintained the initial content of most phenolic compounds in both beverages than TP. Similarly, Vallverdú-Queralt et al. (2012) reported that HIPEF treated tomato juices retained higher concentration of polyphenols than those thermally treated. The lower processing temperatures (< 40 °C) reached during HIPEF treatment would explain the higher retention of phenolic acids and flavonoids in the HIPEF-treated beverages compared to those processed by heat. Other studies have been focused on the variation in phenolic compounds of an orange juice thermally treated and a HIPEF-processed tomato juice (Sentandreu et al., 2007;Odriozola-Serrano et al., 2009). Results from both studies agreed that neither TP nor HIPEF caused significant effects on the phenolic concentration identified in both juices. On the other hand, Dawes and Kenee (1991) observed that just after processing, an ultra-pasteurized kiwi juice contained higher levels of phenolic acids in comparison to the fresh juice. In the same way, Morales-de la Peña et al. (2011) reported that the concentration of most individual phenolic compounds of a fruit juice-soymilk beverage increased immediately after HIPEF or TP treatments.
According to Kelebek et al. (2009), individual phenolic concentration in processed fruit juices mainly depends on the preservation treatments, storage conditions and food matrix. Different kind of stress such as extreme temperatures may provoke changes and reactions on the fruit phenolic content (Zobel et al., 1997). It has been reported that during processing, different reactions such as hydroxylation, methylation, isoprenylation, dimerization, and/or glycosylation, which induce modifications between the different phenolic compounds, can occur at various levels (Rice-Evans et al., 1997).
Moreover, the presence of some enzymes such as phenylanine ammonialyase (PAL), which is the key enzyme in phenolic biosynthesis, can alter phenolic composition of fruits (Macheix et al., 1990 nonetheless, some juices, such as orange juice, showed a significant diminution on its flavonoid content after a certain period of time. Therefore, it could be said that the variation observed on the content of phenolic acids and flavonoids in FJ-SM and FJ-WM beverages throughout the time are highly influenced by the treatment applied, the type of reactions that take place within the products along the storage and the enzymatic activity which can induce the degradation or synthesis of each compound.

CONCLUSIONS
The concentration of TPC and TFC, regardless of the treatment applied, did not present Funds. ICREA Academia Award is also acknowledged by Olga Martín-Belloso.