Improvement of phenolic antioxidants and quality characteristics of virgin olive oil with the addition of enzymes and nitrogen during olive paste processing

303 an experiment at industrial scale. The addition of enzymes in the olive paste during processing increased the total phenol and ortho-diphenol contents, as well as some simple phenolic compounds (3,4-DHPEA, p-HPEA) and the secoiridoid derivatives (3,4-DHPEA-EDA and 3,4-DHPEAEA) in olive oil and therefore improved its oxidative stability. Furthermore, enzyme treatment ameliorated the quality parameters of the produced olive oil (acidity and peroxide value) and their sensory attributes. The use of additional N2 flush with the enzyme treatments did not improve the quality parameters of olive oil any further; however it did not affect the concentration of individual and total sterols or most of the fatty acid composition. Consequently, olive paste treatment with enzymes not only improved the quality characteristics of olive oil and enhanced the overall ogranoleptic quality, but also increased the olive oil yield.


INTRODUCTION
The world production of olive oil amounts to approximately 2.800.000t, with Greece ranking third in the world in terms of olive oil quantity (331.310t) after Spain and Italy (FAOSTAT, 2007).The virgin olive oil, apart from the oily phase (fatty acids) contains precious substances of high biological value such as phenols, ortho-diphenols and various natural antioxidants, tocopherol (Vitamin E), flavones as well as a high content of oleic acid, alcohols, sterols, chlorophyll, volatile aromatic substances, etc. (Baldioli et al., 1996;Giovannini et al., 1999;Iconomou et al., 2005;De Faveri et al., 2008).This fact leads to health claims, particularly regarding children, concerning the prevention of cardiovascular and gastrointestinal diseases, the deceleration of ageing and cancer prevention.
During malaxation, centrifugation and separation, a small fraction of phenolic compounds (ca 10-15 %) is released into the oily phase.The remaining larger fraction is removed with the wastewater and pomace (Servili et al., 1999).The concentration of phenolic compounds in olive oil is affected by the extraction conditions during processing (Montedoro et al., 1992).However, the exact mechanism that RESUMEN Mejora de las caracterísiticas de calidad y de los antioxidantes fenólicos de aceite de oliva virgen por la adición de enzimas y nitrógeno durante el procesado de la pasta de aceituna.

Improvement of phenolic antioxidants and quality characteristics of virgin olive oil with the addition of enzymes and nitrogen during olive paste processing.
The evolution of phenolic compounds and their contribution to the quality characteristics in virgin olive oil during fruit processing was studied with the addition of a combination of various commercial enzymes containing pectinases, polygalacturonases, cellulase and β-glucanase with or without nitrogen flush.Olive fruits (Olea europaea, L.) of the cultivar Megaritiki, at the semi black pigmentation stage of maturity, were used in a 3-phase extraction system in

Improvement of phenolic antioxidants and quality characteristics of virgin olive oil with the addition of enzymes and nitrogen during olive paste processing
By D. It has been shown that the addition of exogenous enzymes to the olive paste increases the oil yield and the antioxidant content in virgin olive oil, depending on the olive variety and degree of ripening (Garcia et al., 2001;Chiaccheirini et al., 2007;Aliakbarian et al. 2008;Najafian et al., 2009).Exogenous enzymatic preparations (i.e.β-glucanase) aid the hydrolysis of secoiridoid glycosides.The enzymes degrade the polysaccharides and liberate phenolic antioxidants, which, after equilibrium, are distributed among three phases: oil, water and solid.Antioxidants are sensitive to oxidation during malaxation of the olive paste and the nitrogen atmosphere protects them against oxidative destruction (Gutiérrez et al., 1977;Giovannini et al., 1999).In general the native enzymes present in the olive fruit are deactivated during the oil extraction process or crushing step.It was shown that a few native enzymes (such as lipoxygenase: LOX and polyphenol oxidase: PPO), retain some detectable activity even in virgin olive oils (De Faveri et al. 2008).It has been suggested that LOX plays a role in the oxidation of unsaturated fatty acids and pigments of the olive fruits (Georgalaki M.D., et al. 1998).After a series of reactions, LOX, together with other enzymes, end up in the formation of volatile compounds responsible in the development of the most characteristic "green" and "fruity" aroma of olive oil (Kyritsakis and Markakis, 1987).Although the effect of the use of enzymes on the quality of olive oil has been studied, its action has not been fully described in combination with a nitrogen atmosphere in the malaxation process at industrial scale.
This paper deals with the effect of various enzyme combinations during the treatment of the olive paste of cv.Megaritiki, with or without nitrogen flush, on the release of phenolic antioxidants in virgin olive oil and on its qualitative and sensory characteristics using a three-phase extraction system at industrial scale.

Olive variety
4080 kg of healthy olive fruits of c.v. Megaritiki at semi-black maturity index, (M.I. =2.75), were harvested from an olive orchard near Agios Konstantinos, Central Greece.(Garcia et al., 2001;Iconomou et al., 2005).The maturity index (M.I.) was calculated as a subjective evaluation of the skin color and flesh as developed in the Research Station of Venta del Llano (Jaen, Spain) and proposed by Uceda and Frias (1975).The following formula was applied to 100 olives that were randomly selected.
A,B,C,D,E,F,G,H, are the number of fruits in the various types described below and 0,1,2,3,4,5,6,7, the grade of ripeness respectively.(i) type 0: intense green skin; (ii) type 1: yellowish green skin; (iii) type 2: green skin with red spots, in less than half of the fruits; (iv) type 3: reddish or purple skin explains the quantitative modification of secoiridoids in the olive oil during malaxation is unknown.The phenolic compounds (natural antioxidants) of olive fruits are distributed among oil, vegetation waste water and solid phase and some may be linked to the colloid oil droplets, such as phenolic polymeric structures and pectins, hemicelluloses, proteins, etc.The relationship between phenolic compounds in virgin olive oil the organoleptic characteristics and its oxidative stability has been studied (Montedoro et al., 1992;Baldioli et al., 1996;Garcia et al., 2001).Olive fruit phenol concentration ranges from 1.0 -3.0 % (w/w), on a raw fruit basis, and depends on various factors such as variety, cultivation environment and ripening stage during harvesting, storage conditions, processing methods etc. (Solinas et al., 1978;Montedoro et al., 1992;Cert et al., 1999;Iconomou et al., 2005;Garcia et al., 2001).
IMPROVEMENT OF PHENOLIC ANTIOXIDANTS AND QUALITY CHARACTERISTICS OF VIRGIN OLIVE OIL WITH THE ADDITION… of total phenols in the methanolic extract was determined colorimetrically using the Folin-Ciocalteau reagent.The absorbance was measured at 725 nm (in the range 0.01-1.00mg/mL) against a blank, using a UV-VIS spectrophotometer (GBS model 916).Results were expressed in mg/kg of gallic acid (Gutiérrez et al., 1977).Ortho-diphenol content in the methanolic extract was determined (in mg/kg of caffeic acid) according to the procedure described by Gutfinger (1981).

HPLC separation of phenolic compounds
The separation of phenolic compounds was performed according to Montedoro et al. (1992) and Servili et al. (1999).The HPLC system consisted of a Spectra System liquid chromatograph model 2000 (Thermo Separation Product, USA), equipped with a 250mm x 4.6mm C18 NovaPak column coupled with a UV detector.Individual phenolic compounds were detected at 278nm.The flow rate was 1 mL/ min.The mobile phase used was 0.2 % (v/v) acetic acid in water (A) vs. methanol (B) for a total running time of 60 min and the gradient changed as follows: 95 % A / 5 % B for 2 min, 80 % A / 20 % B for 10 min, 70 % A / 30 % B for 10 min., 60 % A / 40 % B for 10 min, 40 % A / 60 % B for 10 min., 100 % A / 0 % B for 10 min until the end of running (Angerosa and Di Giacinto, 1995).Samples were dissolved in methanol; a sample loop of 20-µL capacity was used for the introduction of the sample.
Gallic, protocatechuic, p-hydroxybenzoic acid, vanillic, caffeic, syringic, p-coumaric, ferulic and in more than half of the fruits; (v) type 4: black skin and white pulp; (vi) type 5: black skin and pulp purple; (vii) type 6: black skin in more than half of the pulp purple; (viii) type 7: black skin and totally purple pulp.

Enzyme preparations and industrial scale experiments
Olivex and Glucanex were chosen in industrial scale experiments and were kindly supplied by Novo Nordisk Ferment Ltd. (Dittingen, Switzerland).Olivex is an enzyme preparation (produced by the fungus Aspergillus aculeatus) rich in pectinolytic, hemicellulolytic and cellulolytic side activities.Olivex activity was 26.000 PGU (polygalacturonase units) per mL at pH 3.5.Glucanex is a β-glucanase preparation produced by a selected strain of Trichoderma sp.It contains all the enzymes needed for the complete hydrolysis of β-glucan to glucose.Glucanex activity was 300 BGU (β-glucanase units) per gram.
There were three treatments plus the control with three runs in each treatment as follows: (i) E 1 : addition of 0.25 mL of Olivex and 0.03 g of Glucanex per kg of olive paste at the beginning of malaxation; (ii) E 2 : addition of 0.5 mL of Olivex and 0.06 g of Glucanex per kg of olive paste; (iii) E 2 +N 2 : E 2 enzymes used under nitrogen flush (with 2 L / min in a covered malaxator), and (iv) control treatment without enzymes.340.0 kg of olive fruits were used in each run.
Fig. 1 shows the flow sheet of the extraction system used to obtain virgin olive oil.An Alfa-Laval 3phase olive oil extraction system with 2 parallel malaxators and maximum working capacity of 1.0 tonnes per hour was used.Firstly the leaves were removed from olive lots and then olive fruits were subjected to a milling of drupes by a hammer crusher operating at 2800 rpm with a sieve with 6 mm holes working at 100 rpm in the same direction as the crusher.The olive paste was malaxed for 30 min at 30±2°C with or without nitrogen flush.After malaxation, 40 L of water was added to 100 kg of olive paste before entering the 3-phase decanter.Finally, separation of the oily must into oil and vegetation water took place using a simple centrifugal oil separator operating at 1700 rpm (Angerosa and Di Giacinto, 1995;García et al., 2001;Iconomou et al., 2005;De Faveri et al., 2008).The produced olive oil was stored in 1.0 L plastic bottles at room temperature, without nitrogen addition, for a period of 5 months, in the absence of light in order to test the effect of phenolic compounds on the shelf-life of olive oil.

Extraction and colorimetric determination of total phenols and ortho-diphenols
The extraction of phenolic compounds from olive oil was carried out according to Montedoro et al. (1992) and Servili et al. (1999).The concentration

Quality characteristics
The analyses of free acidity (% oleic acid) and Peroxide value (meq O 2 /kg) as well as sterols (mg/kg olive oil) and the fatty acid (%) composition of olive oil were carried out according to the official methods of the EC Regulation 2568/91.Indices K 270 and K 232 extinction coefficients (absorption of 1 % solution in isooctane at 270 and 232 nm, respectively, with 1 cm of passage length) were measured using a double beam UV/Visible spectrophotometer, model GBC-916 (Scientific Equipment Ltd, Victoria, Australia).Chlorophyll was determined (in mg/kg) according to AOCS (1978).Resistance to oxidation was determined using the Rancimat apparatus (Methrohm, Basel, Switzerland) at 120°C with an air flow of 20 L/h.Results were expressed as induction time in hours: Rancimat stability (Läubli andBruttel, 1986: Kyritsakis andMarkakis, 1987).
A sensory evaluation of the sample was performed by a panel of experts according to the official methods of Annex XII of the EC Regulation: 2568/1991.The descriptive analysis used a fivepoint intensity scale, ranging from 0 (no perception) to 5 (extreme).Overall grading used a nine-point scale, 9 for exceptional quality and 1 for the worst.Ten trained tasters were used (Tous et al., 1997).

Statistical analysis
Samples were taken at random in triplicate runs.The results were calculated as the means of three separate runs.Average values were compared with the least significant difference (LSD) at p=0.05 and with the Students t-test where appropriate.

Quality parameters and resistance to oxidation
The effects of the enzymatic preparations with or without nitrogen flush (E 1 , E 2 and E 2 +N 2 ) during malaxation of the olive oils are presented in Table 1.Regarding the quality parameters of the olive oils, there was a decreasing trend in olive oil acidity and peroxide value, with the addition of enzymes compared to control, which is in agreement with previous studies (García et al., 2001;Iconomou et al., 2005, Chiaccheirini et al., 2007, De Faveri et al. 2008).
The use of nitrogen (E 2 +N 2 ) resulted in a significant increase in peroxide value compared with the E 2 treatment (Table 1).The increase in peroxides was probably due to a reduction in the rate of their degradation in the presence of nitrogen, illustrating their relative increase (E.Stefanoudaki, Greece personal communication).
All enzymatic treatments resulted in the improvement of all parameters tested compared to the control, even after 5 months of storage in plastic bottles in the absence of light.By doubling the added enzymes (E 2 vs. E 1 ) there was a significant decrease in olive oil acidity and peroxide value and an increase in chlorophyll (p<0.05).The decrease in peroxide value (Table 1) is likely due to the total phenol increase by the use of enzymes, as shown in Table 2 (Ranalli and Serraiocco, 1996).
The quality parameters (acidity, peroxide value, K 232 -K 270 indices) of the obtained olive oil in all treatments classify it in the extra-virgin olive oil category according to the E.U.Regulation 2568/91.
The addition of enzymes in all treatments resulted in a significant increase (p<0.05) in olive  et al., 2001;Aliakbarian et al. 2008;Najafian et al., 2009).

Phenolic compounds in various treatments
Table 3 shows that the use of enzymes, resulted in a significant increase (p<0.05) in phenolic compound concentration (3,4-DHPEA, p-HPEA, syringic acid, 3,4-DHPEA-EA and 3,4-DHPEA-EDA) in all treatments compared to the control.The E 2 treatment showed the highest concentration in the above phenolic compounds, while there was no difference (p>0.05), between E 1 and E 2 +N 2 .
The antioxidants may be absorbed by polymers and/or dissolved into the water around the hydrophilic sites and trapped inside and among the poly-oil yield of about 15.0% compared to the control.This corresponds to an oil yield increase of about 2.0 kg olive oil per 100Kg of olives.This was also reported by other researchers (Uceda and Frias, 1975;Ranalli and Serraiocco, 1995;García et al., 2001;Vierhuis et al., 2001;Chiacchierini et al., 2007).The action of enzymes in the olive paste is not fully understood.It could be stipulated that enzymes degrade the olive cell wall and therefore change the rheologic behavior of the paste.The disruption of cell walls by enzymes may explain the increase in olive oil yield (Iconomou et al., 2005;De Faveri et al. 2008).
Table 2 shows resistance to oxidation, total phenols and the ortho-diphenol content of the obtained olive oil.The addition of enzymes E 1 , E 2 and E 2 +N 2 in pastes increased the resistance to oxidation -induction time in Rancimat-(p<0.05) and the content of total phenols compared to the control (p<0.05).There was no significant difference among enzyme treatments E 1 , E 2 and E 2 +N 2 oils in total phenol concentration.Ortho-diphenol content increased in E2 and E2 +N2 treatments (p<0.05) but not in E1.This is in agreement with other studies on olive oils obtained from treated olive pastes with   et al., 1975;Solinas et al., 1978;Montedoro et al., 1992).It has been shown that enzyme formulation degrades the walls of the oil-bearing cells.Also the enzymes breaks up the liquid/solid and the liquid/liquid emulsions mainly caused by crushing and centrifuging the paste and through its endopolygalacturonase activity.These results in the release and merging of the oil droplets into larger ones forming a mass of free oil, which is more easily extracted mechanically (Chiacchierini et al., 2007: De Faveri et al., 2008).Additionally it was noted that exogenous enzymes free more antioxidants in olive oil.For example, β-glycosidase was observed to increase the oleuropein aglycon concentration (García et al., 2001;Vierhuis et al., 2001).
The concentrations of total phenols and 3, 4-DHPEA-EA were also correlated to the resistance in oxidation -Rancimat (Iconomou et al., 2005;De Faveri et al., 2008).The maximum correlation (r=0.897) was found between the total phenols and the resistance to oxidation, while the correlation between 3, 4-DHPEA-EA and resistance to oxidation was found to be lower (r=0.792).

Fatty acid and sterol content
Table 4 shows the effect of the olive paste treatments with the enzyme combination E1 and E 2 in the fatty acid composition of virgin olive oil.There was a significant increase in oleic acid in E 1 and E 2 treatments and a simultaneous decrease in linoleic acid compared to the control (p<0.05).During malaxation, the linoleic acid oxidizes quickly and probably the presence of enzymes accelerates its degradation.The reduction of the percent of linoleic acid might have changed the value of the other fatty acids making the increase in oleic acid significant.LOX may be involved in this procedure (Kyritsakis, A. and Markakis P., 1987: Ranalli et al. 2002: De Faveri et al., 2008).The concentration of the rest of the fatty acids was not affected by the addition of the same enzymes E 1 and E 2 (p>0.05).
In addition, the addition of the enzymes (E 1 , E 2 ) did not influence the sterol content of olive oil (Table 5).The composition of fatty acids and sterols in olive oil remained at levels which are in accordance with the limits mentioned in the EU Regulation 2568/91 for all treatments and the control (Iconomou et al., 1998).

Sensory evaluation of olive oil
Table 6 shows the sensory evaluation of virgin olive oil obtained with enzymes (E 1 and E 2 ), by a panel of trained tasters.Treatments E 1 and E 2 resulted in a significant improvement (p<0.05) of most of the desirable attributes of the sensory characteristics of olive oil, especially the flavor of "olive fruity", "green leafy" and "apple" compared to the control.The overall organoleptic grades from a taste panel for E 1 and E 2 were 7.0 and 6.50, respectively, while the corresponding value for the control was lower than the accepted limit (≥ 6.50) for extra virgin olive oil (E.U.Regulation 2568/91).Overall, E 1 and E 2 did not differ significantly (p>0.05) in every individual sensory characteristic and overall grading, although they scored higher compared to the control.
The use of enzymes improved olive oil quality by enrichment with natural antioxidants and increase in its protection to oxidation and the amelioration of its organoleptic characteristics.the results of this study may have a potential practical application in the production of olive oil.Previous reports on the simultaneous use of enzymes and nitrogen on the olive oil quality are scarce in the literature and do not show a definite trend in all cases (Vierhuis et al., 2001;Garcia et al., 2001and Chiacchierini et al., 2007) and are mixed depending on the olive cultivar used, endogenous enzymes, etc.
According to our results the effect of both enzymes and nitrogen exhibited either no improvement or even E.U.Regulations (2568/91 and theirs amendments) prohibit the use of any external adjuvant (chemicals, enzymes) except the addition of water and/or the use of talk as co-adjuvant, for virgin olive oil extraction during paste malaxation.However there may be amendments to the above E.U. regulations in the future for olive oil production as other enzymes have already been approved and used in the agro-food and juice industry with technological and economic benefits (Milan-Linares et al. 2006;De Faveri et al., 2008).Therefore  a slight decrease in the quality of olive oil compared to the use of one enzyme alone.This may be due to the different behavior of N 2 interference during malaxation on the effect of endogenous and exogenous enzymes in various olive cultivars on olive oil qualitative characteristics, as observed in the case of two cultivars Koroneiki vs. Megaritiki.(Garcia et al., 2001;Ranalli et al., 2003;Iconomou et al., 2005 andChiacchierini et al., 2007).This adverse effect of nitrogen was reported with the use of some other enzymes, like tyrosinase (Zhang et al. 2001).

CONCLUSIONS
The addition of a mixture of exogenous enzymes during the olive paste malaxation of c.v. Megaritiki improves the quality characteristics of the obtained olive oil such as acidity, peroxide value and chlorophyll.Sterols and most of fatty acids were not affected by paste enzymatic treatment, except for an increase in oleic acid and a decrease in linolenic acid.The addition of enzymes increased the amount of total phenols and ortho-diphenols, as well as some simple phenolic compounds (3,4-DHPEA, p-HPEA),especially, the secoiridoid derivatives (3,4-DHPEA-EDA and 3,4-DHPEA-EA) in olive oil.There was also an increase in olive oil yield and an improvement in the resistance to oxidation and shelf life in the produced virgin olive oil.The use of an N 2 flush with the enzyme treatments during paste processing did not improve the quality parameters of olive oil compared to enzyme treatments alone.The addition of enzymes considerably improved the olive oil yield and increased the release of antioxidants and total phenols into the virgin olive oil.It also contributed to the enhancement of its quality characteristics of resistance to oxidation, the improvement of the olive oil aroma and the overall organoleptic quality compared to the control.

Figure 1 .
Figure 1.Flow-sheet of the three-phase extraction system used to obtain virgin olive oil, from the olive paste of the Megaritiki variety at industrial scale.

Table 1 Qualitative characteristics and yield of virgin olive oil of the cv. Megaritiki treated during malaxation with commercial enzyme preparation (E 1 , E 2 and E 2 +N 2 )*
*The results in the table represent the average values of the means of 3 runs ± S.D.. E 1 : Olivex+Glucanex.E 2 : 2 x E 1 (Concentration of enzymes is double the concentration in the E 1 treatment).E 2 +N 2 : Olivex+Glucanex+Nitrogen. a,b,c Different superscripts are statistically signifi cant, p=0.05.IMPROVEMENT OF PHENOLIC ANTIOXIDANTS AND QUALITY CHARACTERISTICS OF VIRGIN OLIVE OIL WITH THE ADDITION… other enzyme combinations (Garcia

Table 2 Colorimetric determination of total phenols, ortho-diphenols and resistance to oxidation (Rancimat stability) in virgin olive oil from the cv. Megaritiki during malaxation with commercial enzyme preparations E 1 , E 2 and E 2 +N 2 *
The results in the table represent the average values of the means of 3 runs ± S.D.. E 1 : Olivex+Glucanex.E 2 : 2 x E 1 (Concentration of enzymes is double the concentration in the E 1 treatment).E 2 +N 2 : Olivex+Glucanex+Nitrogen.

Table 4 Effect of olive paste treatment with enzyme preparations E 1 and E 2 on the fatty acid (%) composition of virgin olive oil of the cv. Megaritiki*
*The results in the table represent the average values of the means of 3 runs ± S.D.. E 1 : Olivex+Glucanex.E 2 : 2 x E 1 (Concentration of enzymes is double the concentration in the E 1 treatment).a,b,cDifferent superscripts are statistically signifi cant, p=0.05.IMPROVEMENT OF PHENOLIC ANTIOXIDANTS AND QUALITY CHARACTERISTICS OF VIRGIN OLIVE OIL WITH THE ADDITION…

Table 6 Sensory evaluation of virgin olive oil treated during malaxation with commercial enzyme preparations (E 1 , E 2 ), performed by a panel of ten trained tasters*
The results in the table represent the average values of the means of 3 runs ± S.D.. **Limit acceptance for extra virgin olive oil: Overall grade ≥ 6.50.E 1 : Olivex+Glucanex.E 2 : 2 x E 1 (Concentration of enzymes is double the concentration in the E 1 treatment).a,b Different superscripts are statistically signifi cant, p=0.05. *