Journal of Food Processing & Technology

Olive oil represents the main vegetable oil source used in the Mediterranean area and its use is now spreading rapidly to other areas such as the U.S. and Australia. Olive oil is now recognized as one of the healthiest lipid sources worldwide, due mainly to its high polyphenol content proven to possess antioxidant, anti-inflammatory, antimicrobial and anticancer properties [1-3]. The polyphenol content in commercial olive oil depends on several factors, such as cultivar and genetics, health and degree of fruit maturation, climatic conditions, agricultural methods and the operating conditions of the extraction process. In fact, the quality of the olive oil is strongly linked to the characteristics of the fruit from which it is extracted [4]. To produce good quality virgin olive oil characterized by a reasonable polyphenol concentration, it is necessary to minimize their decay at every stage of olive oil production. One of the olive oil production steps which can lead to a reduction in the quality of the product is the degree of fruit maturation before their processing, since during the maturation phases of olive fruits, biochemical processes occur which result in an accumulation and decay of phenolic compound content in the oily globules originally present in the cells [5]. Sufficient polyphenol content represents an important quality parameter for many virgin olive oils, since, along with the benefits listed above, they have a positive effect on the shelf-life of the commercial product. Reasonable polyphenol content can be obtained in the extracted olive oil if the fruits are collected during an optimum technological harvesting period. Studies made so far of the dynamics of the phenolic content during the ripening phase involved either the whole fruit or the oily fraction extracted, using an analytical strategy that mimics the industrial oil mill process. In fact the commercial olive specific analyser, and the other conventional systems used in the different laboratories, provides for the grinding of the olives with a hammer crusher and the paste obtained is then usually mixed and finally centrifuged without addition of warm water [6-11]. The oil sample composition obtained with these systems is similar to the olive oil obtained with the mill, but is quite different compared to the original composition of the oily fraction of the fruits, because during the extraction process some new phenolic compounds appear, whereas others disappear, in the extracted olive oil samples. More significant information on the technological optimum harvesting period of the olive fruits can be obtained by considering the extent of the polyphenol reduction directly in the oily fraction of the fruits during their maturation. For this purpose however, the fidelity of the polyphenol characteristics of the initial oily fraction of the fruit to that of the extracted olive oil samples, provides the theoretical basis for judging the analytical techniques that allow the evaluation of polyphenol concentration decay due to fruit ripening [12]. To achieve this it is necessary to extract the oily fraction from the fruit by using sample handing strategies that can reduce or eliminate the action of the endogenous enzymes of the fruits and microbiota, both of which are responsible for the rapid qualitative modification and decay of polyphenol concentration in the oil samples during laboratory extraction. A Mild Sample Handling method (MSH) able to minimize polyphenol degradation due to endogenous enzymes during the sample preparation was recently proposed by us in the assessment of polyphenol decay in the oily fraction of the olives during their storage before processing [13]. In this research has been studied the quality improvement of commercial olive oil produced in the mill, through the identification of the technological optimum harvesting period of the fruits during their maturation. To achieve this objective the dynamics of the phenolic compounds content in the oily fraction of the fruits during their maturation using both the recent MSH method and OLM has been assessed.

Olive samples

The laboratory experiments using different oil extraction methods have been conducted using the Leccino olive variety grown at 600 m altitude in a hilly area of the Molise region (Southern Central Italy). The olives were harvested at different periods with a commercial pneumatic olive harvester. They were taken from plants randomly chosen and marked, belonging to an olive grove consisting of 300 twenty-yearold trees. In the laboratory, the olives were carefully blended and some fruits were randomly used to determine the maturation index as described below.

Maturation index

The Maturation Index (MI) was established as a function of fruit colour of both skin and pulp using the method described by Uceda and Hermoso [14]. More specifically, a group of 100 olives were separated into different categories according to skin colour and pulp as follows: 0 - deep green; 1- yellow green; 2 - changing colour with reddish spots; 3 - changing colour with red or light purple present in the whole fruit; 4 - black, without colour under the skin; 5 - black, with colour in less than the half of the pulp; 6 - black, with colour in more than half, but not in the stone; 7 - black, with colour in the whole pulp. Subsequently the MI was calculated as follows:

MI = ax0+bx1+cx2+dx3+ex4+fx5+gx6+hx7/100

where a, b, c, d, e, f, g and h represent the fruit numbers of each category. The analyses were repeated three times. The Weeks After Flowering (WAF) were also established using the last week of May as the initial flowering period.

Olive oil extraction trials

A homogeneous batch of 30 kg of olives was picked weekly in an MI range of between 1.44 and 4.23 (WAF range of between 16 and 23), it was then divided into three groups of 10 kg each and analyzed as described below. The first group of olive samples was subjected to a preliminary de- stoning process using a handle device and then processed using the OLM. The second group was processed using the OLM without any preliminary treatment whereas the third was submitted to an extraction process using the MSH method. The OLM was equipped with a metal hammer crusher, a thermo-mixer maintained at 25°C and a basket centrifuge. The olives from the first and second group described before were subjected to the extraction process. The fruits of the first group without the stones and those of the second group with the stones were crushed by using a metal hammer crusher and then the whole mass obtained was subjected separately to malaxation for 30 min at 25°C and finally centrifuged. After processing, the olive oil obtained was transferred into sterile dark glass bottles without head- space and analysed after 1-2 days. The third group equal to 10 kg of whole olive fruit, was processed using the MSH extractor system which consists of two essential elements: the sample press unit and the centrifuge (OMCN, mod. 155, Bergamo, Italy; FAVS, Bologna, Italy). The sample press unit permitted a rapid extraction of the liquid fraction from the olive samples in about 3 min, without causing the stones to rupture and creating a strong homogenization of the components of the olives. The extracted liquid fraction without malaxation was promptly centrifuged at 9000×g for 5 min using a Universal mod. 32 Hettich centrifuge (Tuttlinger, Germany) and the oil was then transferred into a sterile dark glass bottle without head-space. The trials were repeated three times for each period. The malaxed whole mass of the samples produced with the OLM as well as the unmalaxed liquid fraction from the MSH method were both observed using a light microscopy (mod. BX50F4, Olympus Co., Japan), before the centrifugation.

Microbiological analyses

The microbiological analyses were carried out using the pasta obtained from the whole olives processed in the laboratory with the OLM and the liquid (oil and vegetation water) extracted from the same whole olives with the MSH method. The paste from the OLM was subjected to malaxation for 30 min and analysed before centrifugation. The liquid from the MSH method was not subjected to malaxation and analysed before the centrifugation. 10 g of pasta from the OLM and 10 mL of liquid from the MSH method were transferred into 90 mL of sterile distilled water and homogenized for 5 min with a Stomacher Lab Blender (Seward Ltd, Worthing, UK) and a magnetic stirrer, respectively. After decimal dilution, 200 μL suspensions was plated onto specific growth media using the spread plating technique and the colony forming units (CFU) were recorded. More specifically, the mesophilic bacteria were evaluated after 5 days of incubation at 28°C, using the standard plate count agar (Oxoid, Basingstoke, Hampshire, England) as medium. The total moulds were evaluated after 7 days’ incubation at 28°C using glucose yeast extract agar medium (Oxoid) containing both 100 μg mL^-1 of gentamicin and chloramphenicol. Whereas the total yeasts were assessed after 7 days’ incubation at 30°C in Petri dishes with MYGP agar medium containing 3g yeast extract (Biolife, Milan, Italy), 3 g malt extract (BBL, Cockeysville, MD, USA), 5 g beef extract powder (BBL), 10 g D-glucose (Merck, Darmstadt, Germany), 1000 mL distilled water, pH 7 as described by Kurtzman and Fell [15]. All the microbiological analyses were repeated four times.

Total polyphenol content analysis

The total amount of phenolic compounds was evaluated in the olive oil extracted from the same fruit processed with the OLM, the MSH, and the traditional oil mill method described below. The total polyphenol content of the olive oil was extracted following the method proposed by Gutfinger [16]. An amount of olive oil (2.5 g) was put into a test tube with 5 mL of hexane and 5 mL methanol/water (60:40, vol/vol), the mixture was vortexed vigorously for 2 min. Then 0.2 mL from the phenolic fraction in methanol/water were added to Folin-Ciocalteu’s phenol reagent (0.5 mL) and 4.8 mL of distilled water, then 1 mL of sodium bicarbonate (35%, wt/vol) and a specific amount of water were added to obtain a final volume of 10 mL. The mixture was incubated for 2 h in the dark at room temperature. The total polyphenol content was determined using a spectrophotoeter (Jenway mod. 6300, UK) at 765 nm and the concentration was expressed as mg gallic acid kg^-1 oil (calibration curve with r²=0.997). The analyses were repeated three times.

Bitterness index

The bitterness index (BI) was evaluated by the extraction of the bitter components of 1 g of olive oil sample dissolved in 4 mL hexane, passed over a C18 column (Waters Sep-Pack Cartriges), previously activated with 8 mL of methanol and washed with hexane (8 mL). After elution, 10 mL hexane was passed to eliminate the fat, whereas the retained compounds were eluted with methanol/water (1:1, vol/vol) to 25 mL. The absorbance of the extract was measured at 225 nm against methanol/water (1:1, vol/vol) in a 1 cm cuvette [17]. The analyses were repeated three times.

HPLC profile of the phenolic compounds in olive oil

The profile of the phenolic compounds was assessed using the olive oil samples extracted in the laboratory with the MSH method and OLM. The phenolic compound profile in the oily fraction of the fruits was evaluated throughout the maturation period when the above total polyphenol content and the BI assessed in the oil samples extracted with the MSH method started to decrease. The Mateos and Garcìa method [18] was used with few modifications. Briefly, 2 g of each olive oil sample was enriched with 0.015 mg of syringic acid dissolved in 6 mL of hexane and loaded on a diol-bonded phase cartridge previously conditioned with 6 mL of methanol and 6 mL of hexane. After washing with 6 mL of hexane and 4 mL of hexane:ethyl acetate (90:10, vol/vol), the phenolic compounds were eluted with 10 mL of methanol. After evaporation at room temperature the residue was re-dissolved in 0.5 mL of methanol: water (1:1, vol/vol). HPLC analysis was performed on an Agilent 1200 liquid chromatographic system equipped with a diode array UV detector. A C18 column (4.6 mm i.d. × 250 mm; particle size 5 μm) (Phenomenex, Torrance, CA), coupled to a security guard C18 4×3.0 mm (Phenomenex) was used. Elution was performed at a flow rate of 1.0 mL min-1, using as mobile phase a mixture of water/ acetic acid (97:3, vol/vol), (solvent A) and methanol/acetonitrile (50:50, vol/vol), (solvent B). The solvent gradient changed according to the following conditions: from 95% (A)-5% (B) to 70% (A)-30% (B) in 25 min; 65% (A)-35% (B) in 10 min; 60% (A) -40% (B) in 5 min; 30% (A)-70% (B) in 10 min; and 100% (B) in 5 min, followed by 5 min of maintenance. Chromatograms were acquired at a wavelength of 240 and 280 nm. Identification and quantification of the compounds was performed by taking into consideration retention time and absorption at different wavelengths. The analysis was repeated three times for each olive oil sample.

Enzymatic activity

The oxidoreductase enzymes responsible for the oxidation process of the phenolic compounds, the phenoloxidase and the peroxidase activities have been studied. The activity of these two enzymes was assessed in the pulp as well as in the seeds of the same fruit samples, collected during harvesting time, when the WAF were between 18 and 23 and MI ranged from 3.29 and 4.23. 1000 g of olive samples were divided into two batches, the first batch was de-stoned using laboratory equipment. The stones removed were broken with an appropriate hand tool and the seeds were collected and transferred into beakers. The pulp of the same samples was also collected in separate beakers. Before the enzymatic analysis both the seeds and the pulp of the same samples were crushed separately for 1 min at 11,000 rpm with a Turrax (mod. T25B, IKA Labortechnik, Germany) and immediately analysed.

Phenoloxidase assay

Phenoloxidase activity was evaluated using a spectrophotometric method based on an initial rate of increase in absorbance at 410 nm [19]. The phenoloxidase was assayed by using pyrocatechol (Sigma chemical St. Louis, Missouri) as substrate, whereas the crude extract was prepared with some modifications, as described by Ortega- Garcia et al. [20]. To prepare the acetone power extract, 5 g of frozen (-20°C) olive pulp and seeds were transferred separately into 50 mL plastic test tubes containing 20 mL cold acetone (-20°C) and 5 g of polyethylenglycol (Sigma-Aldrich, Missouri, USA). The resulting mixture was homogenized with a Turrax at 11,000 rpm for 1 min, and then centrifuged at 10000× g for 20 min. The solid fraction of the homogenate was separated from the supernatant and re-extracted twice with the same volume of cold acetone. The combined acetone extracts were dried overnight. Then 0.4 g of dry acetone powder extract was suspended in 10 mL of extraction buffer (phosphate buffer 0.1 M pH 6.2 containing 1 M KCl) and stirred for 1 h at 4°C. The suspension was then centrifuged at 10000×g for 20 min at 4°C, filtered (0.45 μm) and used immediately for the phenoloxidase assay. At the moment of the assay the crude extract was diluted 20 times by transferring 200 μ L of sample into the spectrophotometer cuvette containing 3.8 mL of phosphate buffer 0.1 M pH 6.2 with 100 mM of pyrocatechol. The spectrophotometer dates were collected at a time interval of 30 s for 5 min at 410 nm. The reaction rates were calculated from the initial linear parts of the absorbance-versus-time graphs. In the above experiments, extractions were repeated twice and 2 samples were taken from each extract. One unit of phenoloxidase activity (U) was defined as 0.01 change in absorbance at 410 nm under given conditions per min.

Peroxidase assay

The peroxidase activity was evaluated with a spectrophotometer at 470 nm using pyrogallol (Sigma-Aldrich, Missouri, USA) as a phenolic substrate with hydrogen peroxide [21]. The crude extract was prepared by transferring 5 g of the above olive pulp and the seeds respectively into 50 mL test tubes containing 10 mL of phosphate buffer 0.1 M pH 6.2 with 1M KCl and 0.5 g polyvinylpirrolidone (Sigma-Aldrich, Missouri, USA). The mixture was homogenized with a Turrax at 11000 rpm for 1 min, and then centrifuged at 9000 x g for 20 min, and finally the supernatant was filtered (0.45 μm) and used promptly as crude extract. At the moment of the assay the crude extract was diluted 1:1 (vol/vol) with phosphate buffer 0.1 M pH 6.2, then 60 μ L of each diluted sample was transferred to the spectrophotometer cuvette with 5 mL of reaction mixture containing 0.30 ml of 4% (wt/vol) pyrogallol, 0.30 mL of 1% H₂O₂ and 4.40 mL of 0.1 M potassium phosphate pH 6.2. The spectrophotometric dates were collected at a time interval of 30 s for 5 min at room temperature. The reaction rates were calculated from the initial linear part of the absorbance-versus-time graphs. In the above experiments, extractions were repeated twice and 2 samples were taken from each extract. One unit of peroxidase activity (U) was defined as 0.01 change in absorbance at 470 nm under given conditions per min.

Traditional olive processing with a Pieralisi Simplex oil mill (Jesi, Italy)

Olive fruits of cv. “Leccino” were harvested weekly as described before, in the MI range of between 3.29 and 4.23, they were freed from leaves and other debris using a mechanical device, then washed and crushed with millstones. The paste obtained was subjected to malaxation for 30 min at room temperature, then to squeezing (300×10⁵ pascal). The liquid obtained (aqueous and oily) was separated by a standard centrifuge (Alfa Laval, Monza, Italy). The final products were oil and two waste products (pomace and wastewater), the oil samples were taken from the oil mass immediately after being produced and stored in a dark bottles at 8°C. Approximately 250 Kg of fruit was used for each process, performed in duplicate (n=2), respectively. The total polyphenol content and the BI of the oil samples were assessed as before, whereas acidity (% oleic acid), peroxide value (meq O₂ Kg^-1 oil) and the UV absorption characteristics (K232, K270, ΔK) were evaluated using the analytical methods described in European Union Commission Regulation EEC/2568/91 and its subsequent modifications [22].

Sensory analysis

A fully trained analytical taste panel recognized by the International Olive Council (IOC) performed a sensory analysis on the above oil samples produced through the Pieralisi simplex oil mill. A panel test was established using a standard profile sheet IOC method. Three olive oil samples from each harvesting period were analyzed simultaneously by each taster during three different sessions. The sample sets were randomly distributed among ten assessors. The values of the median sensory data were calculated and the test supervisor chose a significance level of 5%.

Statistical analysis

A priori one-way analysis of variance, using Tukey’s honest significant differences test was performed using a statgraphics computer program (Statgraphics, version 6, Manugistics, Inc. Rockville, MA) and the values different at p<0.05 were recorded.

Comparison of the laboratory extraction systems

The extraction tests performed in the laboratory concerned the OLM and MSH method. The OLM refers to the conventional system of oil extraction that consists of three operational steps: (i) olive crushing – where fruits are crushed to break down the cells and release the oil; (ii) malaxation- where paste (crushed fruits) is slowly mixed to increase oil yield; (iii) oil separation - where oil is separated from the remaining waste by using centrifugation (Figure 1A). The MSH method on the other hand involves the extraction of the liquid fraction from whole fruit and the immediate separation of the oil through centrifugation (Figure 1B). In all the oil samples extracted from whole olives using the OLM a lower concentration of total polyphenols was noticed compared to those obtained using the MSH method. The chemical analyses of oil samples obtained through the OLM of the pitted fruits first showed a slight increase in the phenolic content in correspondence of a WAF equal to 16 and 17, when the MI was between 1.44 and 2.14, and then a gradual linear reduction depending on the degree of maturation. However, when the whole olives were processed using the same method, a gradual reduction of total polyphenol content in the oily fraction of the fruit was observed in correspondence of a WAF of between 16 and 23, when the maturation was characterized by a MI of between 1.44 and 4.23 (Figure 2). An increase in the concentration of total polyphenols in the early stages of ripening was, however, observed from the analyses performed on the oil samples coming from the same whole olive extracts using the MSH method, when the MI of 1.44 increased to 3.29 (Figure 2). Above this limit the decay phase of the phenolic content in the oil began and was very high in the first week (WAF equal to 19), when the MI reached the value of 3.71 from an initial 3.29. In the later stages of maturation however, when the MI was between 3.71 and 4.23, the decay of the total polyphenol content was more gradual (Figure 2). The microbiological analysis performed on the same samples described above showed a higher number of total micro-organisms in the paste from whole olives produced using the OLM, compared to the liquid fraction of the fruit obtained using the MSH method. The total CFU of the microbiota represented by bacteria, yeasts and moulds, was not influenced by the MI of the fruit (Figure 2). The values of the BI during the ripening of the fruit, varied in the same manner as the above phenolic content. In fact the BI values were lower in oil samples obtained using OLM compared to those produced using the MSH method. In detail, in the initial phase of maturation the values found in the samples of oil extracted from the whole fruit using the OLM were constant when the MI ranged from 1.44 to 3.71, while above this limit, they decreased slowly. The dynamics of the BI evaluated in the samples of oil extracted from whole olives using the MSH method, was rather constant during the first stage of maturation (WAF between 16 and 18) when the MI ranged between 1.44 and 3.29, then slowly declined in the later stages (Figure 3). The dynamics of the BI however, assessed in the oil samples from pitted olives and extracted using the OLM showed an intermediate behavior compared to the above dynamics obtained from the whole fruit processed using both the methods.

Figure 1: Ligh microscopy observation (100x) of malaxed whole olive raw material processed with the OLM (A); liquid fraction, containing vegetation water and micro globules of olive oil, extracted from the cells of the whole olive samples with the MSH method (B).

HPLC polyphenol profile

The HPLC analysis of the phenolic compounds has been assessed throughout the maturation period when in the oily fraction of the fruit the beginning of the decay of the concentration of total polyphenols and BI described above was observed. Detailed analyses were carried out in the oil samples obtained from the whole olives harvested at 18, 19 and 20 WAF, when the MI was equal to 3.29, 3.71 and 4.00 and processed using both the MSH method and the OLM. The results obtained from the analysis of the olive oil samples extracted with the MSH method showed a different behavior of the phenolic compounds content in the oily fraction of the fruits during this period of maturation. In fact the dynamic of the total phenolic compounds and some phenolic compounds including oleuropein, oleuropein derivative and ligstroside derivative, fell sharply during the first week of the polyphenol and BI decay period. In detail the content of oleuropein during the MI transition from 3.29 to 3.71, was reduced by 50 % passing, in 7 days, from 24 to 12 mg kg-1 oil. The oleuropein derivative as well as the ligstroside derivative, in the same time period, was reduced respectively by 24% and 29% (Table 1). During the second week of the polyphenol decay period, when the MI went from 3.71 to reach a value equal to 4.00, the oleuropein, the oleuropein derivative and the ligstroside derivative contents were reduced respectively by only 17%, 20% and 12% whereas the reduction of the other simple phenolic compounds increased (Table 1). In contrast to the above results, from the data obtained by the chemical analysis of oil samples extracted with the OLM, was not possible to assess the dynamic of the phenolic compounds content in the oily fraction of the fruits during the same period of maturation, because they were significantly lower than those reported for the MSH method. In fact, a minor concentration of total polyphenols varying between 51 and 61% was observed in the oil samples extracted using the OLM in comparison to those obtained with the MSH method. A lesser content of oleuropein, equal to 90- 91%, oleuropein derivatives in the proportion of 85-89%, ligstroside derivatives (15-39%), hydroxytirosol (25-50%) and tyrosol (34-60%) was observed in the phenolic compounds. The elenolic acid, however, was more concentrated by 110-181% while all the other compounds remained more or less the same (Table 1).

Table 1: Phenolic compounds profile of the oily fraction of the olives, harvested at the beginning of the total polyphenol decay.

Enzymatic activity

The activity of oxidoreductase enzymes in the pulp and olive seeds was evaluated throughout the period of fruit maturation when the MI varied between 3.29 and 4.23, which corresponds to the decay period of total polyphenol content observed in the oily fraction of the fruits extracted using the MSH method. The activity of oxidoreductase enzymes was higher in the seeds than in the pulp. The activity of phenoloxidase in the seeds was constant throughout the fruit maturation when the MI was found to be between 3.29 and 4.23, whereas in the pulp the enzymatic activity increased slowly in accordance with the increase of the MI from 3.29 to 4.00, and then remained constant (Figure 4). Peroxidase activity was almost constant in the first maturation phase when the MI increased from 3.29 to 3.71, then increased rapidly in the seeds, whereas in the pulp it decreased slowly (Figure 4).

Olive processing in the oil mill

The olives processed in the traditional oil mill were harvested when the MI was between 3.29 and 4.23. On the basis of qualitative indices, the batches of oil produced were classified as Extra Virgin Olive Oil (EVOO) commercial category, while that from the olives with MI equal to 4.23 was characterized by an acidity equal to 0.85%, and so can be classified as a Virgin Olive Oil (VOO) (data not shown). The oil yield varied from a minimum of 17% (wt/wt) to a maximum of 18.50% (wt/ wt). The total polyphenols reached the maximum concentration of 250 mg gallic acid kg^-1 oil in the samples from olives characterized by an MI equal to 3.29; then the phenolic compounds decrease in accordance with the ripening of the fruit until they reach a minimum equal to 100 mg gallic acid kg^-1 oil in the samples extracted from olives with a MI equal to 4.23 (Table 2). The phenolic content found in the oil samples extracted with the mill method using olives characterized by a MI between 3.29 and 4.23 was highly correlated (r²=0.99) with the phenolic content of the oil samples of these same olives extracted using the MSH method (Figure 2). The BI showed the same dynamics of the phenolic content, in fact, the highest value of 1.5 was found in the oil coming from the olives with a MI equal to 3.29, while the lowest equal to 0.3 was detected in the oil samples obtained from olives characterized by a MI equal to 4.23 (Table 2). Sensory analysis carried out on the same samples of commercial oil obtained in different harvesting periods showed no negative qualities, whilst the positive qualities were related to the total polyphenol content present in the samples analyzed. It was noticed, however, that all the samples of oil characterized by a total polyphenol content superior to 160 mg of gallic acid per kg of oil and a B1 superior to 1 showed a median of fruitness, bitterness and pungency higher than 2.5 (Figure 5 and Table 2).

Figure 2: Dynamic of the polyphenol concentration and the spontaneous microbiota evaluated throughout the olive fruits maturation (–▲–, total polyphenols assessed in the oil samples extracted from the whole fruits with the OLM; –■–, total polyphenols assessed in the oil samples extracted from the pitted fruits using the OLM; –♦–, total polyphenols evaluated in the oil samples extracted from the whole fruits using the MSH method; --♦--, total microorganisms evaluated in the pasta of the whole fruits obtained using the OLM; --▲--, total microorganisms evaluated in the liquid fraction (oil and vegetation water) of the whole fruits extracted using the MSH method; MI, maturation index; WAF, weeks after flowering).

Table 2: Characteristics of olive oil produced in the mill with the olives harvested in different maturation phases

Phenolic compound content in olive oil is an important factor to consider in the evaluation of the quality of virgin olive oil as an antioxidant, for its health benefits, and for the physical qualities of the product. Some types of oil are not very stable chemically given that they show a phenolic content which is too low and which in some cases is not higher than 80-90 mg of gallic acid per kg of product. One of the systems which permit an increase in the phenolic content of oil is the transformation of the olives collected in an optimum technological period, and which is assessed through an evaluation of the dynamics of the phenolic content of the oily fraction of fruits during their maturation. On the basis of what has been reported it is clear how the knowledge of the initial level and forms of phenolic compounds present in the oily fraction of the olives is of considerable interest since the chemical composition of the oil extracted in the laboratory or in the mill is quite different compared with the original oil present in the fruit. During the first step of the oil extraction process using the OLM, the crushing of the olives produced a notable rupture of the cellular membranes releasing small drops of oil and producing raw material. This is a multiphase system in which phenolic compounds are divided according to their polarities [23]. Malaxation is the step in oil extraction that specifically modifies the oil composition because many chemical and enzymatic reactions occur, producing deep qualitative and quantitative modification in the olive oil polyphenols [24]. It has already been demonstrated by Zullo et al. that a MSH method of oil extraction in the laboratory, compared to the OLM, is able to recover a greater amount of polyphenols in the olive oil samples, reducing or avoiding their oxidation and other chemical and biochemical changes during the extraction process because the oil micro globules are extracted from the cells of the fruits together with the vegetation water and then centrifuged immediately (Figure 1B). Present research confirmed the higher polyphenol content in the oil samples extracted using the MSH method rather than that obtained using the OLM (Figure 2). The lower polyphenol concentration found in the olive oil samples extracted from whole fruit using the OLM can be explained by taking into consideration the fact that, during the extraction of the oil samples, the olives are subjected to a crushing and malaxation that leads to contact between the oily fraction and raw material containing the endogenous enzymes of the fruits and of the microorganisms that are involved in the hydrolysis and oxidation process of the polyphenols (Figure 1A). The data in Figure 2 show that in the paste obtained with the whole olives processed using the OLM, the decay of polyphenol content was greater in extent than in the pulp produced from pitted olives, because, as demonstrated by the results shown in Figure 4, in the seeds the oxidoreductase activity is much higher than that found in the pulp, therefore the paste produced from whole olives is richer in enzymes. These results are quite similar to those reported for Taggiasca and Conservable cultivars [25]. On the contrary, when the olive oil samples were extracted using the MSH method, which is less invasive, when compared to the first method, contact between the endogenous enzymes and the substrates was reduced because during the extraction the whole fruit was subjected to a mild pressure without breaking the stones. Using this procedure it was possible to reduce significantly the detrimental effects of many native endogenous enzymes on the phenolic compounds present in the oily fraction of the fruits during the extraction process. In fact, the data reported in Table 1 show how during the process of extraction, the oil obtained using the OLM, differently from that obtained through the MSH method, loses about half of the total polyphenols originally present in the oily fraction of the fruits. Among the individual phenolic compounds, oleuropein and its derivate are the most penalized compounds. In fact about 85-91% of their content is hydrolyzed during the oil extraction process using the OLM. The reduction of the other phenolic compounds such as tyrosol and hydroxytyrosol can be attributed to the activity of the oxidoreductase enzymes whilst the accumulation of elenolic acid could be attributed to the lesser activity of those same enzymes mentioned above as regards this substrate. These results, in line with what has already been reported by Zullo et al., show that the MSH extraction method, when compared to the OLM, is more suitable for assessing the dynamics of the phenolic content in the oily fraction of the fruits during their ripening and consequently, on the basis of this, establish the technological optimal harvesting period. In fact, from Figure 2 it is possible to see that, in an MI ranging between 1.44 and 4.23, the dynamics of the phenolic content measured using the MSH method shows first an increasing trend and then a decreasing phase, while the second, as revealed in the samples produced using the OLM, shows values substantially lower than antecedents, characterized by a decreasing trend typical of this extraction method which was also observed for other tests performed on monovarietal olive oils [26]. The highest values of BI detected in oil samples extracted using the MSH method (Figure 3), are in agreement with the data shown in Figure 2 and Table 1. In fact, during the ripening of fruits, the reduction of the phenolic compound content responsible for the bitter taste of the oil (BI), can be explained by taking into account the decay of the oleuropein content, which, at the beginning of the period of the total polyphenol decay (Figure 2), was reduced by 50% in just 7 days whereas the decay of the oleuropein derivative and ligstroside derivative was more regular during the entire period studied (Table 1). The dynamics of the total polyphenol decay and the BI observed in the oil extracted using the MSH method from whole olives characterized by an increasing MI, may be attributed for the most part, to the activity of some endogenous oxidoreductase of the fruit. The henoloxidase and peroxidase enzymes, as demonstrated also in other experiments [27], were very active during the whole period of maturation (Figure 4). The chemical data reported in Table 2 obtained from the commercial oil produced in mills with the same olives mentioned above, show that if the fruits are harvested as soon as the decay of the total polyphenols content in the oily fraction of the fruits extracted with the MSH method starts (Figure 2), it is possible to obtain a commercial olive oil in the mill with a reasonable content of phenolic compounds without affecting the yield. In the case studied, the best technological harvest period is between 19 and 22 WAF, which coincides with the period of decay of the phenolic content in the oily fraction of the fruits processed using the MSH method when the MI varied between 3.71 and 4.17. In these conditions it was possible to obtain at the mill level a commercial extra virgin olive oil (EVOO) of good quality characterised by a median of the positive attributes of the sensory test superior to 2 (Figure 5), a total polyphenol content equal to 160-240 mg gallic acid kg-1 oil, and the same yield (Table 2).

Figure 3: Dynamic of the bitterness index (BI) assessed throughout fruit maturation (–▲– , BI assessed in the oil samples extracted from whole fruits using the OLM; –♦– , BI assessed in the oil samples extracted from whole fruit using the MSH method;–■– , BI assessed in the oil samples extracted from the pitted fruits using the OLM; MI, maturation index; WAF, weeks after flowering).

Figure 4: Oxidoreductase activity evaluated in olive pulp and seeds during fruit maturation ( –Δ– , phenoloxidase activity in the pulp; –▲– , phenoloxidase activity in the seed; –□– , peroxidase activity in the pulp; –■–, peroxidase activity in the seed; MI, maturation index; WAF, weeks after flowering).

Figure 5: Sensory Profile (positive attributes) of the olive oil produced in the mill with olives collected at different stages of maturation. 18 WAF; 19 WAF; 20 WAF; □21, 22, 23

The results of the trials showed that the MSH extraction method, in comparison to the OLM, respects the phenolic composition of the original oily fraction of the fruit more. This feature makes the MSH method suitable for monitoring the decay of the phenolic compound content in the oily fraction of the fruits during their ripening in order to establish the technological optimal harvesting period. If the olives are harvested and processed in the mill just after the beginning of the total polyphenol decay in the oily fraction of the fruits assessed with the MSH method it is possible to improve the polyphenol content and the quality of the commercial olive oil.