Quality Changes in Mango Juice Treated by High-Intensity Pulsed Electric Fields Throughout the Storage

The effect of high-intensity pulsed electric fields (HIPEF) processes on Listeria innocua inhibition, physicochemical parameters and activity of oxidative enzymes of mango juice was evaluated to set the optimal HIPEF treatment time. Quality parameters, microbial population and bioactive compounds of HIPEF-treated (35 kV/cm, 1800 μs) and thermally treated (TT) (90 °C, 60 s) mango juices were studied and compared with those non-treated during 75 days of storage at 4 °C. HIPEF treatment for 800 μs ensured 5 log reductions of L. innocua. Polyphenoloxidase (PPO), lipoxygenase (LOX) and peroxidase (POD) residual activities were significantly reduced to 70, 53 and 44%, respectively, at treatment times of 1800 μs. Similar sensory properties compared with fresh mango juice were attained from product treated at 1800 μs. Moreover, fresh mango juice colour (L* = 38.79, h° = 106.57) was preserved after HIPEF treatment throughout storage. Moulds and yeasts and psychrophilic bacteria counts in HIPEF-treated (1800 μs) mango juice remained below 6 log cycles CFU/mL up to 2 months of refrigerated storage. The content of total phenolic compounds in those HIPEF-treated increased from 333 to 683 μg of GAE/mL from day 0 to the end of storage. Hence, the application of HIPEF may be a feasible treatment in order to ensure microbiological stability, high bioactive compound content and fresh-like characteristics of mango juice.


INTRODUCTION
Mango (Mangifera indica L.), one of the most harvested tropical fruits, is widely used to produce juices due to its well-appreciated sensorial attributes (FAO 2003(FAO , 2012;;Nanjundaswamy 1998).Furthermore, this fruit is a rich source of bioactive compounds such as phenolics and carotenoids, hence mango consumption could have health benefits in preventing degenerative diseases (Rawson et al. 2011;Schieber et al. 2000).
Mango juice can undergo quality-degrading reactions triggered by microbial growth population and quality-degrading enzymes, among others.Therefore, preservation treatments are required to ensure its safety and quality stability.On one hand, thermal treatment is commonly used in the juice industry because of its well-known effectiveness in the inactivation of microorganisms and quality-degrading enzymes (Mercadante and Rodriguez-Amaya 1998;Soliva-Fortuny et al. 2009).However, undesired chemical, physical and sensorial changes as well as reduction of bioactive compounds content have been observed in thermally-treated juices (Sánchez-Moreno et al. 2005;Wibowo et al. 2015).On the other hand, non-thermal treatments allow to obtain microbiologically stable fruit juices but also a better preservation of sensorial and nutritional characteristics than conventional treatments (Chen et al. 2013).Hence, highintensity pulsed electric fields (HIPEF) technology has been considered as a feasible non-thermal technique for the preservation of liquid foods.The electric field strength and treatment time are reported as the main parameters of HIPEF treatment to induce an electric potential across cell membrane conducting the cell damage (Morales-de la Peña et al. 2010).
Several studies have proved the efficiency of HIPEF on the inactivation of microorganisms such as Listeria innocua, which is one of the main foodborne microorganisms in fruit juices (Huang et al. 2012;Mosqueda-Melgar et al. 2007;Timmermans et al. 2014).Nevertheless, published data evidenced that the degree of microbial inactivation is strongly dependent on the HIPEF conditions (Jiménez-Sánchez et al. 2017).With regard to enzyme activity, peroxidase (POD), polyphenoloxidase (PPO) and lipoxygenase (LOX) catalyse some reactions affecting sensory and nutritional properties in fruit juices.HIPEF treatments from 20 to 35 kV/cm have halved enzymatic activity in tomato and orange juices (Aguiló-Aguayo et al. 2010;Vervoort et al. 2011).Moreover, HIPEF seems to maintain quality characteristics including colour, soluble solids and viscosity as well as retain bioactive compounds of fruit juices (Buckow et al. 2013;Odriozola-Serrano et al. 2008).
Despite of the noteworthy literature using HIPEF treatment for fruit juices quality preservation, no studies comparing the effects of HIPEF and thermal treatment on quality changes of mango juice have been found.Therefore, the objectives of the present work were firstly to select the HIPEF treatment time capable to inactivate L. innocua and to reduce enzymatic activity in mango juice while preserving its fresh-like sensorial attributes.Secondly, to compare the effect of HIPEF and thermal treatments on microbial stability, activity of oxidative enzymes, total carotenoids and phenolics content, antioxidant capacity and physicochemical properties in mango juice throughout 75 days of refrigerated storage.

HIPEF treatments
HIPEF treatments were performed using a continuous flow bench scale system (OSU- were kept to apply pulses in bipolar quadratic mode, while different treatment times were assayed (50, 100, 200, 400, 800, 1200, 1600, 1800 and 2000 µs).According to the results of microbial and enzymatic inactivation of HIPEF-treated mango juice, the treatment conditions were set for subsequently study of preservation of mango juice along the storage.

Thermal treatment
MJ was heat-treated at 90 °C for 60 s.The juice was pumped with a peristaltic pump (model D-21V, Dinko, Barcelona, Spain) and passed through a tubular stainless steel heat exchange coil system (University of Lleida, Lleida, Spain).Immediately after heating, the tubular stainless steel was immersed in an ice-water bath at 4 °C and thereafter MJ was packaged (Odriozola-Serrano, Soliva-Fortuny, Hernández-Jover, & Martín-Belloso, 2009).

Packaging and storage
Treated MJ was bottled directly from the treatment systems in sterilized 100 mL polypropylene bottles and leaving the minimum headspace volume.Non-treated MJ was bottled thereafter the juice preparation.Once filled, the containers were tightly closed and stored in darkness under refrigeration (4 ± 1 °C) until analysis.Non-treated and treated MJ were analysed twice a week the first 3 weeks and once a week until day 75.

Listeria innocua culture, inoculation and enumeration
L. innocua IPL 1.17 (Institute Pasteur de Lille; Lille, France) was cultured in tryptone soy broth (TSB) with 0.6 % yeasts extract (Bioakar Diagnostic; Beauvais, France) and incubated at 35 °C with continuous agitation at 200 rpm for 15 h to obtain cells in stationary growth phase.The final concentration reached in the culture was 10 8 -10 9 colonies forming unit per mL (CFU/mL).MJ was inoculated with L. innocua to have an initial concentration of 10 7 -10 8 CFU/mL and then HIPEF-treated.

Peroxidase (POD)
POD activity was determined using the method described by Elez-Martínez, Aguiló-Aguayo, Martín-Belloso, 2006) with some modifications.The enzyme extract for POD activity measurement was obtained by the homogenization of 10 mL of MJ and 20 mL of sodium phosphate buffer 0.2 M at pH 6.5.The homogenate was centrifuged at 24000g for 15 min at 4°C (AVANTI TM J-25, Beckman Instruments Inc; Fullerton, CA, USA).The supernatant was filtered throughout a Whatman paper (no. 1) and the resulting liquid constituted the enzymatic extract.POD activity was assayed spectrophotometrically (CECIL CE 2021 spectrophotometer Cecil Instruments Ltd, Cambridge, UK) in a 1 cm path cuvette by adding at 0.1 mL of enzymatic extract 2.7 mL of sodium phosphate buffer (0.05 M, pH 6.5), 0.1 mL phenylenediamine (1 %) and 0.1 mL hydrogen peroxide (1.5 %).The oxidation of p-phenylenediamine was determined at 470 nm measuring the absorbance every 10 seconds during 3 min.The absorbance values were referred to a sample blank containing all reagents except hydrogen peroxide, which was substituted by distilled water.POD activity was obtained from the slope of the linear portion of the curve.One unit of POD activity was defined as the change of absorbance per minute and millilitre of enzymatic extract at 22ºC.

Polyphenoloxidase (PPO)
PPO activity was determined by the method of Vásquez-Caicedo et al, (2007) with some modifications.For the extraction of the enzyme, 5 g of MJ were mixed with 0.5 g polyvinylpolypyrrolidone (PVPP) and 4.5 g McIlvaine buffer solution (pH 6.5) consisting of 35 % of 0.1 M citric acid and 75 % 0.2 M disodium phosphate.The mixture was homogenised and centrifuged at 23000 g for 15 min at 4 °C (Centrifuge AVANTI TM J-25, Beckman Instruments Inc; Fullerton, CA).The supernatant was filtered with Whatman paper (no. 1) to obtain the enzyme extract.PPO activity was measured using a spectrophotometer (CECIL CE 2021; Cecil Instruments Ltd, Cambridge, UK) at 400 nm by adding 100 µL enzyme extract and 3 mL of 0.5 M cathecol solution and obtaining the absorbance every 10 seconds during 3 min.A blank of cathecol without extract was used.The PPO activity was obtained from the slope of the linear portion of the curve; one unit of PPO activity was defined as a change of one unit of absorbance per minute and millilitre of enzyme extract at 22 °C.

Lipoxigenase (LOX)
LOX activity was determined by the method described by Anthon & Barrett (2003) with modifications.The enzyme extract was obtained by mixing 20 mL of MJ with 5 mL of a solution containing 0.5 M phosphate buffer (pH 6.5) and 0.5% Triton X-100 and centrifuging 10 min at 10000 g at 4 °C (Centrifuge AVANTI TM J-25, Beckman Instruments Inc; Fullerton, CA).The pellet was discarded and the supernatant was filtered with Whatman paper (N o .1).The LOX activity of the enzyme was measured by mixing 2 mL phosphate buffer 0.1 M (pH 6.5), 40 µL linoleic acid and adding 100 µL enzymatic extract.The reaction was measured with a spectrophotometer (CECIL CE 2021; Cecil Instruments Ltd, Cambridge, UK) at 234 nm each 10 seconds during 3 min.
The activity was calculated from the slope of the linear portion of the curve.A blank was prepared with 2 mL phosphate buffer 0.1 M mixed with 1 mL linoleic.One unit of LOX activity was defined as a change of one unit of absorbance per minute and per millilitre of enzyme extract at 22 °C.
Enzymatic activity was expressed as percentage of residual activity (RA %) which was calculated by the quotient between the enzyme activity of treated (AEt) and the nontreated (AE O ) MJ.
After one hour in dark storage, absorbance was measured at 765 nm (CECIL CE 2021 spectrophotometer; Cecil Instruments Ltd, Cambridge, UK).TP content was calculated on the basis of a standard curve of gallic acid and expressed as µg of gallic acid equivalent (GAE) per MJ mL.

Antioxidant Capacity
Antioxidant capacity was determined by a radical-scavenging activity (RSA) assay evaluated as bleaching of the stable 1,1-diphenyl-2-picrylhydrazyl (DPPH) radical.MJ (10 mL) was centrifuged at 3500 g, 20 min and 4°C in a Centrifuge AVANTI TM J-25 (Beckman Instruments Inc; Fullerton, CA, USA).The reaction mixture constituted of 10 µL of supernatant, 3.9 mL of methanolic DPPH (0.0025 gL -1 ) and 90 µL of distilled water was carried out.The samples were shaken vigorously and kept in the dark for 30

Sensory evaluation
A total 30 non-trained panellists participated in the sensory test of treated and nontreated MJ at day of processing.A hedonic scale from 0 (dislike) to 10 (extremely like) was used to rate the colour, flavour and overall acceptance.MJ (30 mL) processed by HIPEF (35 kV, 1800 µs, 200Hz, 4µs), heat (90ºC, 60 s) and non-treated (NT) were served at 16 ± 1 °C in transparent cup coded with three digits randomly numbered.
Moreover, a glass containing potable water and a piece of non-salted cracker were provided to panellists to eliminate the residual taste between samples (Mosqueda-Melgar, Raybaudi-Massilia, & Martín-Belloso, 2012)

Statistical analysis
All the treatments were assayed in duplicate and two replicate analyses were carried out for each sample to obtain the mean values and standard deviations (SD) for each analysed parameter.The analysis of variance (ANOVA) and Least Significant Differences (LSD) was performed in order to find statistical differences (p ≤ 0.05).All statistical analyses were conducted with Statgraphics plus Centurion XV software Version 15.1.02(StatPoint Technologies, Inc.).

L. innocua inactivation
A maximal reduction of L. innocua survival of 5.7 log units was achieved after applying HIPEF (35 kV/ cm, 200 Hz, 4µs) for 800 µs to MJ (Figure 1).The longer is the HIPEF treatment time up to 800 µs, the higher the decrease of microbial population.As The studied oxidative enzymes followed similar pattern of inactivation.Nevertheless, differences on the RA between LOX and the other oxidative enzymes in MJ at the longest treatment time were observed (Figure 2).This could indicate a different level of HIPEF effect on the enzymatic structure.PPO and POD structure contains a prosthetic group, thereby, the influence of electric fields on changing the structure of coppercontaining enzyme has been reported scarcely since it can be considered tightly bound organic molecules (Sharma et al. 2013).Otherwise, conformational changes in LOX structure, with no prosthetic group, could occur easily.Moreover, other authors have reported that charges separation of tertiary structure occurred in LOX native conformation leading almost complete inactivation of LOX, when long treatments and high voltage are used in enzymatic solution, but not in PPO (Luo et al. 2010).In agreement with the available scientific literature, electrochemical effect of HIPEF may affect the local electrostatic fields in proteins and disrupt electrostatic interactions of peptide chains leading to conformational changes in enzymes (Buckow et al. 2013).
Therefore, HIPEF treatment had greatest degree of activity reduction on LOX compared with PPO and POD in HIPEF-treated MJ at 2000 µs. Figure 3 shows a non-significant changes of L* value of MJ from 50 to 2000 µs.
Similarly, the h° value was maintained in the range of 74.5 to 73.9 in HIPEF-treated MJ as treatment time increased (Figure 3).Thus, HIPEF treatment preserved characteristic colour of MJ.The significant reduction of enzymatic activity in HIPEF-treated MJ might prevent quality degrading oxidative reactions (Pathare et al. 2012).The present results are aligned with previous studies, where colour of HIPEF-treated orange (Cortés et al. 2008) and carrot juice (Quitão-Teixeira et al. 2007) were preserved as in fresh juices.Carrot, orange and mango juice have similar yellow colour tonality, which could be mainly attributed to carotenoid compounds.Thus, yellow colour might be preserved whether great content of natural pigments such as carotenoids is maintained.Since HIPEF-treated MJ at 1800 µs led to a significant reduction of L. innocua population and enzymatic activity as well as fresh-like physicochemical characteristics, sensory evaluation of MJ treated by HIPEF and TT at day of processing, and further quality analysis along the storage were carried out at 35 kV/cm, 1800 µs, 200 Hz, 4 µs.

Microbial evaluation
Initial counts of moulds and yeasts in non-treated MJ were 4.22 ± 0.58 log 10 CFU/mL, while those of psychrophilic bacteria were 1.74 ± 0.15 log 10 CFU/mL.HIPEF or TT effectively reduced microbial loads of the juice up to the detection limit just after processing (day 0) (Figure 5).During storage, it was observed that moulds and yeasts population increased earlier than psychrophilic bacteria in treated and non-treated MJ.
No microbial growth in HIPEF-treated MJ was detected during the first two weeks of storage, while the TT-MJ did not show microbial growth along the entire storage time.
Microbial counts for HIPEF-treated and thermal-treated MJ was lower than 6 log 10 CFU/mL until day 59 and 75, respectively, whereas non-treated MJ exceed those counts at day 23.
Diverse studies have suggested that microorganisms are inactivated because of electroporation and electrofusion phenomena during the HIPEF treatment (Buckow et al. 2013).Nevertheless, a microbial growth in HIPEF-treated MJ could be attributed to a non-complete inactivation of microorganisms (Mosqueda-Melgar et al., 2007).HIPEF treatment enabled to extend the lag phase of MJ microbial population, hence, the recovery of injured microorganisms and germination of those sporulated was delayed.Timmermans et al., (2011) and Elez-Martínez, Soliva-Fortuny, Martín-Belloso, (2006) observed no growth of moulds and yeasts in HIPEF-treated orange juice at 25 and 35 kV/ cm, respectively, during 20 and 56 days.Although, Timmermans et al., (2011) achieved similar microbial stability compared with the present study, it must be noted that the treatment temperature used was 56 °C, whereas present results were obtained without exceeding 40 °C.

Enzyme activity
At the beginning of storage, RA of HIPEF-treated (35 kV/cm for 1800 µs with bipolar pulses of 4 µs at 200Hz) MJ were 70.0 ± 5.1; 69.9 ± 4.9 and 46.3 ± 10.2 % for PPO, LOX and POD, respectively.The application of thermal treatment to MJ significantly reduced activity of PPO and POD up to 55.5 ± 0.5 and 20.7 ± 1.0 at day 0 (table 1).The PPO and POD molecular structure, which contains a prosthetic group in their structure, has been reported to be specially affected by pH, temperature and electric fields (Luo et al. 2010).Otherwise, RA LOX after thermal treatment increased at day of processing, LOX appeared to be less thermo-sensible.During storage, a severe increase of RA POD in non-treated MJ was observed, whereas PPO and LOX activities were slightly reduced.
Probably, the increase of POD activity might be assigned to the cell release of POD substrate (organic hydroperoxides), which enable the enzyme-substrate contact (Vervoort et al. 2011).
Both electrochemical and thermal effects associated with HIPEF and TT could result in changes in the structure and conformation of enzymes, which may lead to inactivation (Huang et al. 2012;Timmermans et al. 2011).However, the appearing of isoenzymes and uncomplete inactivation might explain the fluctuations of enzymatic activity in TT and HIPEF-treated MJ along the storage.RA of PPO and POD in MJ treated by TT and HIPEF had a drastically decrease from day 16 until the end of storage.Among oxidative enzymes, RA POD of 25.1 ± 3.5 % (day 75) and 17.0 ± 4.4 % (day 49) was the lowest in MJ treated by TT and HIPEF, respectively.Consistently, literature has reported that POD seemed to be more susceptible to HIPEF than other enzymes and is associated with the modification of the α-helix structure (Leong and Oey 2014).These results are inconsistent with the complete POD inactivation during 56 days reported by Elez-Martínez, Soliva-Fortuny, et al. (2006) in orange juice after HIPEF treatment (35 kV/cm for 1000 μs with bipolar pulses of 4 μs at 200 Hz).However, other authors described a progressive decrease of RA POD in HIPEF-treated orange juice (23 kV/cm, 90 Hz, monopolar pulses of 2 μs and 130 L/h) along 58 days (Vervoort et al. 2011).
In contrast, significant RA LOX reduction in treated MJ required long storage time.Both TT and HIPEF treatments reduced significantly more than a 50 % the initial activity of LOX at the end of storage.Similar to other studies a retarded decrease of the RA LOX was observed (Espachs- Barroso et al. 2006;Zhao et al. 2007).According to Aguiló-Aguayo, Soliva-Fortuny & Martín-Belloso ( 2010), LOX protein chain could undergo changes and a development of resistant isoforms in HIPEF-treated fruit juices.Thus, the conformational changes in LOX structure might delay the reduction of the activity throughout storage time.
It is known that HIPEF and thermal enzyme inactivation mechanisms are related to the unfolding of proteins due to changes in their secondary structure (Salvia-Trujillo et al., 2011).Also, a weak affinity of enzyme-substrate complex might describe the decrease of RA in HIPEF-treated MJ during the storage.Another hypothesis for reducing enzymatic activity in HIPEF-treated MJ throughout storage would be the formation of aggregates as a result of a strong polarization of the protein molecules and hydrophobic interactions or covalent bonds (Luo et al. 2010).Therefore, the protein aggregation along the storage could reduce the enzymatic reaction by avoiding the substrate from fitting the active site of the enzyme.

Physicochemical parameters
pH and TSS values remained stable throughout the storage and no statistical differences among treatments were observed.pH average values for non-treated, TT and HIPEFtreated MJ were 3.7 ± 0.1, 3.76 ± 0.04 and 3.7 ± 0.1, respectively.The mean values of TSS for non-treated, TT and HIPEF-treated MJ were 9.4 ± 0.9, 10.72 ± 0.52 and 8.53 ± 1.62, respectively.In contrast to the obtained results, Timmermans et al., (2011) observed a TSS increase in HIPEF-treated orange juice ( 23 According to other studies, the loss of colour in non-treated MJ might be related with the oxidative reactions mostly triggered by residual activity of POD and PPO (Timmermans et al. 2011;Wibowo et al. 2015).In this sense, the increase of RA POD observed in non-treated MJ probably conducted the deterioration of colour.Differently, all treated MJ significantly reduced the activity of POD and PPO; hence, enzymatic browning was avoided.Therefore, treated MJ preserved the yellow colour of fresh mango juice.

Bioactive compounds and antioxidant activity
The effects of processing and storage time on bioactive compounds and antioxidant activity of MJ are shown in Figure 6.Considering total carotenoid content, TT and HIPEF-treated MJ showed a decrease of 17 and 13 %, respectively, compared with nontreated MJ at the beginning of the storage (Figure 6a).Carotenoids compounds are thermo-labile; hence, heat processing leads to significant higher losses in TT MJ than those HIPEF-treated.Differently, an electroporation on the cell membrane, which enable the releasing of carotenoids among other compounds, in HIPEF-treated MJ could occur.Oxidative reactions promoted by enzymes, light or oxygen could affect rapidly the carotenoids released in TT or HIPEF-treated MJ, which could explain the subsequently decline of carotenoids content (Soliva-Fortuny, Balasa, Knorr & Martín-Belloso, 2009).According to Odriozola-Serrano, Soliva-Fortuny, Hernández-Jover, & Martín-Belloso (2009) oxidation may occur by self-oxidation, where alkylperoxyl radicals are formed and these radicals attack the double bonds resulting in formation of epoxides.Thus, the severity of oxidation depends on the structure of carotenoids and the environmental conditions.However, during storage period, HIPEF-treated MJ reached 2.2 times more carotenoids than those heat-treated (Figure 6a).Similarly, other studies described great retention of carotenoids in HIPEF-treated compared to heat-treated orange juice during storage (Buckow et al. 2013).Total phenolic compounds in MJ varied from 560.1 ± 17.9 (non-treated) to 333.8 ± 27.8 (HIPEF-treated) and 529.6 ± 15.4 (TT) µg of gallic acid/ mL at processing day (day 0).Similarly to Santhirasegaram, Razali, George, & Somasundram, (2015), no significant difference in the phenolics concentration after thermal treatment compared with non-treated MJ was observed immediately after processing.Although other authors have also reported that after HIPEF treatment he phenolic content is reduced, the mechanism is not well known (Rawson et al. 2011).The interaction with other compounds such as solutes resulting from the high electric field and long treatment time applied could create aggregations reducing the content of phenolic compounds (Soliva-Fortuny et al. 2009).
Total phenolics decreased in non-treated MJ along the storage (Figure 6b).Otherwise, total phenolic compounds concentration increased in MJ treated by HIPEF throughout the storage.Indeed, HIPEF-treated MJ (683.79 ± 0.50 µg GAE/mL) showed the highest phenolics concentration compared with TT MJ at day 59.Phenolic compounds are formed in plant products via the action of phenylalanine ammonia-lyase (PAL) in the phenylpropanoid metabolism (Patthamakanokporn et al. 2008).This response is initiated when the plant recognizes a stimulus at the cellular level.It could be hypothesized that HIPEF induced PAL activity and may influence the voltage-gated ion channels and increase the membrane permeability for Ca 2+ at the cellular level, followed by a rapid influx of Ca 2+ through cation channels.Through this process, Ca 2+ -dependent protein kinase phosphorylates PAL, which regulates the phenylpropanoid metabolism (Vallverdú-Queralt et al. 2012).On the other hand, the loss of phenolic compounds in non-treated fruit juice was also observed by Patthamakanokporn, Puwastien, Nitithamyong, & Sirichakwal (2008) who attributed the decrease of phenolics during the storage to deleterious enzymes such as PPO.After analyzing the data obtained in this work, it was observed that there was a negative correlation (r= -0.74) between the activity of PPO and the content of phenolics.This result seems to indicate the importance of TT and HIPEF treatment in reducing RA PPO .Thus, decreasing of PPO activity, which uses phenolic compounds for the oxidative processes to trigger on quinones, was mainly associated with increasing in phenolics (Cheema and Sommerhalter 2015).
Initial antioxidant capacity in MJ was 20.4 ± 1.3, 18.7 ± 0.4 and 17.9 ± 0.4 % of DPPH inhibition for HIPEF, TT and non-treated MJ, respectively.The enhancement of radical scavenging activity in HIPEF-treated MJ might be attributed to the stress response of antioxidant compounds.During storage, the antioxidant capacity of MJ depleted irrespective of the treatment applied (Figure 6c).It is remarkable that both total carotenoids content and antioxidant capacity rapidly decreased in TT and non-treated MJ along the storage.Our results for HIPEF-treated MJ were in accordance with Odriozola-Serrano et al. (2008) who observed a significant loss of antioxidant capacity as storage time increased in HIPEF-treated tomato juice (35 kV/ cm, 100 Hz and 1500 µs of treatment time).In many plant species, a good relationship between antioxidant activity and total phenolics was noted.Contrarily, no correlation between total phenolic compounds and antioxidant capacity in treated MJ was observed.Thus, antioxidant capacity in MJ during refrigerated storage could be related to other bioactive compounds such as vitamin C, which could be easily affected by oxidative deleterious reactions (Buckow et al. 2013).

CONCLUSIONS
HIPEF treatment at 35 kV/cm, 4 μs-bipolar pulses, 200 Hz and 1800 µs proved to be feasible in the reduction of L. innocua population to pasteurization levels in mango juice while enzymatic activity of PPO, LOX and POD was reduced up to 70, 53 and 44 % RA, respectively, and fresh-like physicochemical properties maintained.The native flora stability of HIPEF-treated mango juice was assured throughout 59 days at 4 °C.
On the other hand, LOX activity of HIPEF-treated mango juice was halved along the storage.Also, the POD and PPO enzymatic activity in HIPEF-treated mango juice was lower than in those untreated throughout storage.The reduction of PPO enabled a significant increase of the phenolic content in HIPEF-treated mango juice during 59 days.Differently, antioxidant capacity and carotenoid content of all evaluated mango juices decreased gradually throughout storage period.However, bioactive compounds in mango juice were better retained after HIPEF than thermal treatments.The beneficial effect of the HIPEF treatment was noticeable over the storage period with enhanced phenolic content and maintaining fresh-like characteristics of mango juice.NT: Non-treated mango juice Values represent the mean ± standard deviation.Values in a column followed by the same lower case letter and in a row followed by the same upper case letter are not significantly different
min.The absorption of the samples was measured with a spectrophotometer (CECIL CE 2021 Cecil Instruments Ltd, Cambridge, UK) at 515 nm against a blank of methanol without DPPH (Odriozola-Serrano et al. 2008).The results were expressed as percentage of DPPH inhibition as shown in equation 1 where A o is de absorbance of DPPH reagent and As is the absorbance of the MJ sample reaction with DPPH.
described in Figure1, no significant differences in the L. innocua inactivation levels were appreciated at HIPEF treatments from 800 to 2000 µs.According to microbiological criteria proposed byFDA (2004)  for fruit juices, 5 log reductions of target microorganisms should be accomplished for obtaining safe product.Similarly, Mosqueda-Melgar, Raybaudi-Massilia, & Martín-Belloso (2007) achieved 5 log reduction of L. innocua population in melon juice treated by HIPEF (35 kV/cm, bipolar square wave, 4 µs pulses and 200 Hz) as treatment times increased up to 1250 μs.Previous research explained the effect of increasing HIPEF treatment time on microbial inactivation by the formation of membrane pores triggering to membrane destabilization and cell rupture(Vega-Mercado et al. 1997).Although the efficacy of HIPEF (20 kV/cm, 90 Hz and 130 L/h) against L. innocua was also proved in orange juice(Timmermans et al. 2014), less studies have been found to reduce more than 5 log at 800 µs.The low pH (4.1) and conductivity (1.71 mS/cm) of MJ could cause L. innocua cells more sensible to damage.Indeed,Amiali, Ngadi, Raghavan, & Nguyen, (2006) reported that lowering ionic concentration cause an increase of the treatment chamber resistance, which could enhance the microbial inactivation levels.Wouters, Dutreux, Smelt, & Lelieveld (1999) observed better reduction of L. innocua in solutions with low pH (4.0) and conductivity (2.7 mS/cm) than in alkaline solutions.3.1.2.Enzyme activityHIPEF treatment applied for 1800 µs reduced at 70, 53 and 44 % PPO, LOX and POD activity in the MJ, respectively (Figure2).Differently, at treatment times below 1800 µs, when no significant reduction of enzymatic activity was observed, various deleterious reactions affecting loss of nutritive value and yellow colour might occur in MJ.A reduction of the RA as increasing HIPEF treatment time has been also reported by Aguiló-Aguayo, Sobrino-López,Soliva-Fortuny, & Martín-Belloso (2008)  andAguiló- Aguayo et al. (2010), who reached 10 and 30 % of RA for PPO and LOX, respectively, in strawberry treated by HIPEF (35 kV/ cm, 1000 µs, 200 Hz and 4µs).HIPEF treatment, that is known to conduct to cell electroporation, might benefit the contact between enzyme and substrate released from the cell, hence, no complete inactivation was achieved in MJ(Huang et al. 2012).The effect of HIPEF at 1800 µs might cause an irreversible conformational change of the globular protein chain of enzymes in MJ.An enzyme denaturation might be a feasible reason for enzymatic activity reduction(Luo et al. 2010).
no significant effect (p ≥ 0.05) on pH and conductivity of MJ when different treatment times were applied.Average values in pH and conductivity of treated-MJ were 4.1 ± 0.1 and 1.71 ± 0.01 mS/cm, respectively.In a similar way,Zhang, Gao, Zhang, Shi, & Xu (2010) and Aguilar-Rosas, Ballinas-Casarrubias, Nevarez-Moorillon, Martin-Belloso, & Ortega-Rivas, (2007) reported that both HIPEFprocessed longan and apple juice, did not show pH differences with the non-treated products.Other reports indicated no change of conductivity after HIPEF treatment(Mosqueda-Melgar et al., 2012;Vega-Mercado et al., 1997).Although no effect of HIPEF treatment time on TSS content or viscosity of MJ was observed, differences between HIPEF-treated and non-treated MJ were detected.Non-treated MJ (10.8 ± 0.7 °Brix and 4.0 ± 0.3 mPa•s) presented lower average values of TSS and viscosity compared with HIPEF-treated (35 kV/ cm, 200 Hz, 4µs and 2000 µs) MJ (12.9 °Brix ± 0.6 and 5.4 mPa•s ± 1.1).Cserhalmi, Sass-Liss, Tóth-Markus & Lechner (2006)  andFalade, Babalola, Akinyemi, & Ogunlade (2004) reported an increase in TSS and viscosity of citrus juices treated by HIPEF (28 kV/cm, 100 μs, 2 μs-bipolar pulses at 100 Hz), which were attributed to the breakdown cell effect releasing soluble solids from the cell.Moreover, changes in HIPEF-treated MJ compared with the non-treated might be also attributed to a decline of the pectinolitic enzyme activity, which could enable to maintain pectin content in MJ and hence higher TSS and viscosity (Espachs-Barroso et al. 2006).

Figure 4
Figure4shows the influence of HIPEF (35 kV/cm, 1800 µs, 200 Hz, 4 µs) and TT (90 ˚C, 60 s) on sensorial attributes (colour, flavour and overall acceptance) of MJ compared with the non-treated.Similar overall acceptance and flavour between treated and non-treated MJ were observed.Mosqueda-Melgar, Raybaudi-Massilia, Martín-Belloso, (2012) observed no differences in flavour and overall acceptance comparing fresh fruit juices and those treated by HIPEF and TT.On the other hand, colour values in HIPEF and thermally-treated MJ were alike.Nevertheless, significant differences (p ≤ 0.05) in colour perception of non-treated MJ (5.6 ± 1.6) compared with the HIPEFtreated (7.2 ± 1.8) were detected.The reduction of oxidative enzyme activity in HIPEFtreated MJ might avoid the loss of colour.Also, the possible release of natural pigments due to the electroporation effect in HIPEF treatment could explain the great colour score of HIPEF-treated MJ given by the consumers.
kV/cm and 90 Hz) after 58 days of refrigerated storage.Differences might be attributed to the use of lower electric field compared with that of the present study; hence, less reduction of enzymatic activity might lead deleterious quality process as increment of turbidity and TSS.L* values of the non-treated, HIPEF-treated and TT MJ at day 0 were 39.78 ± 0.01, 38.87 ± 0.52 and 40.34 ± 0.25, respectively (table 2).During storage, non-treated MJ rapidly declined L*, whereas a slightly decreased in HIPEF-treated MJ was observed.L* values of thermal-treated MJ were preserved along the storage.On the other hand, initial h° values of non-treated (106.57± 0.26), TT (107.4 ± 0.6) and HIPEF-treated (108.03 ± 0.38) MJ were not significantly different.Along the storage, h° of non-treated MJ decreased; hence loss of yellow colour might occur.TT and HIPEF treatment maintained similar h˚ in MJ throughout the storage.The loss of L* and h˚ could be associated with the formation of dark colour compounds and reduction of yellow colour in beverages due to the non-enzymatic browning reactions (Pathare et al. 2012).