Food Research International 65 (2014) 301–310
Contents lists available at ScienceDirect
Food Research International journal homepage: www.elsevier.com/locate/foodres
Polyphenolic characterization of olive mill wastewaters, coming from Italian and Greek olive cultivars, after membrane technology Isabella D'Antuono a,1, Vassiliki G. Kontogianni b,1, Kali Kotsiou b, Vito Linsalata a, Antonio F. Logrieco a, Maria Tasioula-Margari b,⁎, Angela Cardinali a a b
Institute of Sciences of Food Production, ISPA National Research Council of Italy, Via G. Amendola, 122/O, 70126 Bari, Italy Section of Industrial and Food Chemistry, Department of Chemistry, University of Ioannina, Ioannina 45110, Greece
a r t i c l e
i n f o
Article history: Received 26 June 2014 Received in revised form 19 September 2014 Accepted 26 September 2014 Available online 18 October 2014 Keywords: Olive mill wastewaters Olive wastewater membrane fractionation Olive wastewaters biophenols Olive wastewaters chemical composition
a b s t r a c t The aim of this work was to recover and identify the phenolic compounds from olive mill wastewater (OMWW) samples belonging to two Italian (Cellina and Coratina) and three Greek (Asprolia, Lianolia and Koroneiki) olive cultivars. The OMWWs were processed using membrane technologies to obtain three fractions: microfiltrate (MF), ultrafiltrate (UF) and nanofiltrate (NF). These steps allowed purifying the OMWWs in order to achieve fractions with different profile and concentrations of polyphenols. In particular, the amount of polyphenols ranged from 2456 μg/mL to 5284 μg/mL in MF; from 1404 μg/mL to 3065 μg/mL in UF and from 373 μg/mL to 1583 μg/mL in NF. Among the cultivars analyzed Coratina followed by Lianolia showed the highest amount of verbascoside (VB) (308 μg/mL in Coratina versus 145 μg/mL in Lianolia, respectively) in UF fractions. Furthermore, UF fractions that showed adequate purification degree and polyphenol enrichments, were used for the identification of the phenolic compounds by liquid chromatography/diode array detection/electrospray ion trap tandem mass spectrometry (LC/DAD/ESI–MSn) analysis. Twenty three compounds, belonging to the following classes of constituents: secoiridoids and their derivatives, phenyl alcohols, phenolic acid and derivatives, and flavonoids, were identified in almost all the UF fractions of the different cultivars. Finally, differences were observed among the cultivars regarding the presence of elenolic acid derivatives, hydroxytyrosol glucoside, and β-hydroxyverbascoside diastereoisomers. The results obtained showed that OMWW can be considered as raw material for the isolation of valuable bioactive compounds able to be used in food, cosmetic and pharmaceutical industry. © 2014 Elsevier Ltd. All rights reserved.
1. Introduction Olive mill wastewaters (OMWWs) are seasonally generated effluents in the olive oil extraction industry operating in three-phase mode. This agro-industrial waste is produced in huge amounts (6–7 million tons/ year) and it is characterized by a strong undesirable smell, an intense brown to dark color, a pH between 3 and 6 and a highly diverse organic pollutant load (Ginos, Manios, & Mantzavinos, 2006). OMWW, a complex medium containing polyphenols of different molecular masses, is produced in Mediterranean countries. This waste is claimed to be one of the most polluting effluents among those produced by the agro-food industries because of its high polluting load and high toxicity to plants, bacteria, and aquatic organisms, owing to its contents (14–15%) of organic substances and phenols (up to 10 g/L). These latter compounds,
⁎ Corresponding author at: Department of Chemistry, University of Ioannina, Ioannina 45110, Greece. Fax: +30 2651008197. E-mail address:
[email protected] (M. Tasioula-Margari). 1 These authors contributed equally to this work.
http://dx.doi.org/10.1016/j.foodres.2014.09.033 0963-9969/© 2014 Elsevier Ltd. All rights reserved.
characterized by high specific chemical oxygen demand (COD) and resistance to biodegradation, are responsible for its black color, depending on their state of degradation and the olives they come from (Capasso, Cristinzio, Evidente, & Scognamiglio, 1992). For a long time, OMWW has been regarded as hazardous waste with negative impact on the environment and an economic burden on the olive oil industry. Their phytotoxicity is mainly attributed to the high phenolic content (0.5–24 g/L), that, on the other hand, are antioxidant compounds with potential health-benefits (Obied et al., 2005a). In light of these findings, the OMWWs are recognized as a potential low-cost starting material rich in bioactive compounds, that can be extracted and applied as natural antioxidants for the food and pharmaceutical industries. A typical phenolic substance identified in olive fruit is oleuropein, a secoiridoid glucoside that is absent in OMWW due to enzymatic hydrolysis during olive oil extraction, resulting in the formation of side products such as hydroxytyrosol and elenolic acid. Other phenolics identified in OMWW are verbascoside, tyrosol, catechol, 4-methylcatechol, p-hydroxybenzoic acid, vanillic acid, syringic acid, and gallic acid (Capasso et al., 1992; Visioli et al., 1999).
302
I. D'Antuono et al. / Food Research International 65 (2014) 301–310
The most abundant biophenols occurring in OMWW are hydroxytyrosol followed by tyrosol. In particular, hydroxytyrosol is the most potent antioxidant phenolic compound occurring in olive oil (Nissiotis & Tasioula-Margari, 2002), and numerous studies have focused on its many other health-beneficial effects (Obied, Allen, Bedgood, Prenzler, Robards, et al., 2005a) among them in inhibition of lowdensity lipoprotein oxidation (EFSA, 2011). Moreover, the good solubility of hydroxytyrosol in oil and aqueous media and its high bioavailability allow its useful application in multi-component foods, encouraging prospects in commercialization of it in functional foods and natural cosmetics (Bouzid et al., 2005). However, the phenolic composition of OMWW varies strongly between studies, as it is characterized by a significant complexity (Bianco et al., 2003; Obied, Allen, Bedgood, Prenzler, & Robards, 2005b; Obied, Bedgood, Prenzler, & Robards, 2007) and many compounds are recently identified (Cardoso, Falcão, Peres, & Domingues, 2011). Indeed, hydroxytyrosol acyclodihydroelenolate and p-coumaroyl-6′-secologanoside (comselogoside) were recently identified in OMWW and were examined for their antioxidant and antiproliferative activities (Obied, Karuso, Prenzler, & Robards, 2007; Obied, Prenzler, Konczak, Rehman, & Robards, 2009). Traditionally, to isolate and recover polyphenols from matrix such as OMWW, liquid–liquid extraction is employed. This method utilized a large amount of solvent that has a negative impact for both health and environment. Membrane separation has become a promising technology with several advantages: low power consumption, water-reuse and by-products recovery, stabilization of effluent, absence of organic solvents. Some studies are already carried on, and the OMWW may be treated efficiently by using microfiltration (MF), ultrafiltration (UF), nanofiltration (NF) and/or reverse osmosis (RO), to obtain a permeate fraction which can be discharged in aquatic systems according to national or EU regulations or to be used for irrigation (Paraskeva, Papadakis, Tsarouchi, Kanellopoulou, & Koutsoukos, 2007). A membrane process for the selective fractionation and total recovery of polyphenols, water and organic substances from OMWW was also proposed by Russo (2007). It was based on the preliminary MF of the OMWW, followed by two UF steps realized with 6 kDa and 1 kDa membranes, respectively, and a final RO treatment. The RO retentate, containing enriched and purified low molecular weight polyphenols, was proposed for food, pharmaceutical or cosmetic industries, while MF and UF retentates can be used as fertilizers or in the production of biogas in anaerobic reactors (Garcia-Castello, Cassano, Criscuoli, Conidi, & Orioli, 2010). The current investigation aimed at identifying the phenolic compounds in the OMWW samples belonging to different Italian and Greek olive cultivars. The OMWWs were processed using membrane technologies in order to investigate the impact of different cultivar in the phenolic profile of the fractions. The fractions were quantified regarding the main polyphenols present by HPLC analysis and were deeper studied, using LC/DAD/ESI–MSn analysis, in order to elucidate the identity of phenolic components that could also be a characteristic of each OMWW coming from different olive variety.
2. Material and methods 2.1. Chemicals Extraction and chromatography solvents, methanol (MeOH), acetonitrile (MeCN), glacial acetic acid (AcOH), and ethanol (EtOH), were of certified high-performance liquid chromatography (HPLC) grade, and pure standard of hydroxytyrosol (HT), tyrosol (Tyr), caffeic acid (CAA), coumaric acid (CUA), verbascoside (VB), isoverbascoside (IsoVB), were obtained from PhytoLab GmbH & Co. KG (Vestenbergsgreuth, Germany). Folin–Ciocalteu reagent was purchased from Sigma-Aldrich (Milano, Italy).
2.2. OMWW samples All the OMWWs utilized in this study raised from mills that used a three-phase system. Five fresh OMWW samples (~30 L), among them two Italian cultivars: Cellina and Coratina obtained respectively from: Cooperativa Agricola Nuova Generazione Srl (Martano, Lecce, Italy) and Oleificio Di Molfetta (Bisceglie, Bari, Italy) from Apulia region mills and three Greek cultivars, Koroneiki, Lianolia and Asprolia (all organic), collected from Greek mills. Italian samples were processed within 4 days after olive oil production, Greek samples were collected and shipped 2 days later, so were processed within 7 days after olive oil production. Acetic acid was added to pH 5, in the samples, to avoid phenolic compounds oxidation. The raw material was firstly sieved through a test sieve with 425 μm as porosity. This process allows the removal of large particles and colloids from the OMWW before the microfiltration process. With this procedure, all traces of oil, leaves, seeds, which could then cause problems of clogging of the membranes, are eliminated. 2.3. Filtration units (MF, UF and NF) The raw OMWWs coming from the five different cultivars were processed with a laboratory-scale system (Permeare s.r.l., Milano, Italy) present in the laboratory of ISPA-CNR of Bari. This system, that utilizes a continuous parallel flow, is consisting of two different units: Pilot Plant N022/N256C and N021/N256C (Figs. 1 and 2). The first (Fig. 1), filled by external tank, performed microfiltration (MF) process with continuous recirculation of the sample and by using a ceramic membranes, PERMAPORE EOV 1046, with cut-off of about ~ 100,000 Da (membrane porosity 0.1 μm). The volume which can be processed daily will be limited by the substances contained in the fluid. In general, it is possible to treat from 200 up to 2000 L per day. This unit can support transmembrane pressure (differences between the inlet pressure and outlet pressure) up to 6 bars, at 25 °C. The Pilot Plant N021/N256C unit (Fig. 2) is composed of the following sections: process tank (5–10 L), high pressure pump, pressure vessel for membrane housing. The unit can support operating pressure up to 75 bar and operating temperatures that ranges between 5 °C as minimum and 60 °C as maximum. Cooling device is supplied for eventually cooling the solution in order to maintain an acceptable temperature during process. Furthermore, the maximum size of eventual suspended solids in the feed solution should be less than 3 μm. Pressure vessel for membrane is composed of three AISI316 stainless steel parts: testate, end cup and body for membrane housing with 5 cm of diameter and 30.5 cm of length. This unit performed ultrafiltration (UF) and nanofiltration (NF) processes, utilizing polymeric membranes at different porosities. For UF, was employed a polyethersulfone membrane, PERMAPORE DGU 1812 BS EM with cut-off of 5000 Da, instead for NF, the filtration was performed using a polyamide membrane, PERMAPORE AEN 1812 BS with cut-off of 200 Da. After utilization, the membranes were washed with alkaline detergent following the manufacture instructions because the OMWW can provoke the membrane clogging during the process. All the membrane utilized act as a molecular sieve without any chemical interactions with the matrix. In addition, the ceramic membrane was utilized for MF because they are chemically stable and mechanically and biologically inert. In addition, they are available only as limited range of porosity and, for this reason, mainly utilized for microfiltration. 2.4. Determination of total phenolic content The total phenolic content of the OMWW and fractions was determined using a modified Folin–Ciocalteu spectrophotometric method 100 μL of properly diluted samples, calibration solutions or blank were pipetted into separate test tubes and 100 μL of F–C reagent were added to each. The mixture was mixed well and was allowed to equilibrate.
I. D'Antuono et al. / Food Research International 65 (2014) 301–310
303
Fig. 1. Microfiltration: A: Feed, B: Microfiltration membrane, C: Permeate, D: Drain, E: Switchboard.
After exactly 2 min, 800 μL of a 5% (w/v) sodium carbonate solution were added. The mixture was swirled and put in a temperature bath at 40 °C for 20 min. Then, the tubes were rapidly cooled on the rocks and the color generated was read at its maximum absorption (750 nm). The absorbance was measured in 1-cm cuvette by the Varian Cary 50 Scan UV/visible spectrophotometer. For calibration solutions and blank preparation, a methanolic solution at the same concentration of samples was used (Cicco, Lanorte, Paraggio, Viggiano, & Lattanzio, 2009). Results were expressed as micrograms per milliliter (μg/mL) of HT equivalents.
2.6. LC–MS analysis
2.5. HPLC-DAD analysis
2.6.2. Analysis A 10 μL aliquot of ultrafiltrate fraction of OMWW samples was filtered (0.45 μm) and injected into the LC–MS instrument. Separation was achieved on a 25 cm × 4.6 mm i.d., 5 μm Zorbax Eclipse XDB-C18 analytical column (Alltech, Deerfield, USA), at a flow rate of 1.0 mL/min, using as solvent A (water/acetic acid, 99.9: 0.1 v/v) and solvent B (acetonitrile/ methanol 1:1). The gradient used for the analysis of OMWW UF fractions was: 0–20 min, 95–70% A; 20–25 min, 70–65% A; 25–45 min, 65–60% A; 45–60 min, 60–30% A; 60–65 min, 30–0% A; 65–75 min, 0% A; and 75–80 min, 0–95% A. The UV/vis spectra were recorded in the range of 200–700 nm and chromatograms were acquired at 240, 280 and 340 nm. Both precursor and product (MS2 and MS3) ions scanning of the phenolic compounds were monitored between m/z 50 and m/z 1000 in negative polarity. The ionization source conditions were as follows: capillary voltage, 3.5 kV; drying gas temperature, 350 °C; nitrogen flow and pressure, 11 L/min and 50 psi, respectively. Maximum
Analytical-scale HPLC analyses of the OMWW and fractions were performed with an Agilent Technologies series 1100 liquid chromatography (Waldbronn, Germany) equipped with a binary gradient pump G1312A, a G1315A photodiode array detector, and a G1316A column thermostat set at 45 °C. ChemStation for LC3D (Rev. A. 10.02) software was used for spectra and data processing. An analytical Phenomenex (Torrance, CA) Luna C18 (5 μm) column (4.6 × 250 mm) was used throughout this work. The solvent system consisted of (A) methanol and (B) acetic acid/water (5:95, v/v). For low molecular weight phenolics two solvents were used: A, methanol and B, acetic acid–water (5:95 v/v). The elution profile of the linear gradient was: 0–21 min, 15–40% A; 21–30 min, 40% A (isocratic); 30–45 min, 40–63% A; 45–47 min, 63% A (isocratic); and 47–51 min, 63–100% A (Lattanzio, 1982). The flow rate was 1 mL/min and the injection volume was 20 μL.
2.6.1. Instrumentation All LC–MSn experiments were performed on a quadrupole ion trap mass analyzer (Agilent Technologies, model MSD trap SL) retrofitted to a 1100 binary HPLC system equipped with a degasser, an autosampler, a diode array detector and an electrospray ionization source (Agilent Technologies, Karlsruhe, Germany). All hardware components were controlled by Agilent ChemStation software.
304
I. D'Antuono et al. / Food Research International 65 (2014) 301–310
Fig. 2. Ultra and nanofiltration: A: Feed tank, B: Membrane, C: Permeate, D: Concentrate, E: Switchboard.
accumulation time of ion trap and the number of MS repetitions to obtain the MS average spectra were set at 30 and 3 ms, respectively. 2.7. Statistical analysis For total phenolic amount, statistical differences were determined by analysis of variance (ANOVA) followed by multiple comparison procedure on ranks with Student–Newman–Keuls Method, at 5% significance level, using the software SigmaPlot for Windows (Ver. 12, Systat Software Inc., San Jose, CA 95110 USA). 3. Results and discussion 3.1. Filtration units During the filtration steps three different fractions were obtained MF, UF and NF. The MF fraction was recovered using a ceramic membrane with the aim to stabilize and clarify the OMWW and to give a filtrate free from high molecular weight proteins, enzymes and bacterial component. The yield of the process was from 2 to 4 L/h of OMWW, it was depending from the wastewaters quality. The UF fraction, instead, was obtained by a polymeric membrane at 5000 Da of cut-off, with yield ranging from 1.5 to 2.4 L/h. This step gives a fraction free from colloidal substances and separates, from the water solution, organic macromolecules at molecular weight lower than membrane cut-off. Finally, the NF step was performed by a membrane with a molecular weight cut-off of 200 Da. The yield of the process was of about 6 L/h. This
step is mainly used for separating the polymeric fractions of the polyphenols from mineral salts. 3.2. OMWW total phenolic contents The obtained fractions were assayed by a modified Folin–Ciocalteu (FC) method to determine the total phenols content expressed as HT equivalent (Cicco et al., 2009). The results are showed in Table 1. Among Italian cultivars, Cellina showed the higher phenolic content, instead among the Greek, Lianolia was the most abundant. Considering the polyphenol contents in MF as 100%, the percentage of polyphenols recovered in UF fractions, ranged from 55% to 65%, and in NF fraction from 15% to 33% (Table 1). In particular, the amount of polyphenols ranged from 2456 μg/mL to 5284 μg/mL in MF; from 1404 μg/mL to 3065 μg/mL in UF and from 373 μg/mL to 1583 μg/mL in NF. The fractions were also characterized by HPLC analysis, to quantify the main polyphenols present. As shown in Fig. 3(a, b, c), the main Table 1 Total phenolic content and relative percentage of five OMWW cultivars after filtration process. The data are expressed as HT equivalent (μg/mL). OMWW CVs
MF fraction
UF fraction
μg/mL Asprolia Koroneiki Lianolia Coratina Cellina
3488 2456 4623 4768 5284
± ± ± ± ±
% a
500 328b 236c 289c 265c
100 100 100 100 100
NF fraction
μg/mL 2270 1404 2559 2755 3065
± ± ± ± ±
% a
475 221b 353c 321c 431c
65 57 55 58 58
μg/mL 976 373 1402 1583 1423
± ± ± ± ±
% a
110 43b 135c 121c 134c
28 15 30 33 27
Means with a common letter within a column are not significantly different (p ≤ 0.05).
I. D'Antuono et al. / Food Research International 65 (2014) 301–310
305
Fig. 3. Phenolic composition of OMWW fractions, expressed as relative percentage of each compound respect to the total phenolics quantified by HPLC-DAD analysis. MF fraction a); UF fraction b); NF fraction c).
phenolics identified were: HT, Tyr, CAA, CUA, VB, IsoVB, caffeoyl-6secologanoside (SEC), and comselogoside (COM). In all the fractions analyzed and in all CVs (Fig. 3a, b, c), the most abundant compound is HT that represents about 70–80% of the total phenolic concentration, followed by Tyr. On the contrary in MF of Coratina the main phenolic was VB (Fig. 3a), which in the following step (UF fraction Fig. 3b) is reduced with a simultaneous increase of HT, as hydrolysis product of VB molecule. VB was lower and in some cases absent, in the other cultivars. The hydrolytic process affects oleuropein and demethyloleuropein, hydrolyzing them to hydroxytyrosol, while verbascoside, depending on the storage time, was affected to a lower extent. Coratina followed by Lianolia had the higher content of these compounds (308 μg/mL in Coratina versus 145 μg/mL in Lianolia, respectively). β-
Hydroxyverbascoside diastereoisomers were detected in traces in the rest of cultivars. Even though the OMWWs for the Italian cultivars were stored for 4 days before the extraction and the respective OMWW from Greek cultivars were stored for 7 days before the extraction a considerable amount of verbascoside was found for Coratina and Lianolia cultivars. The different distribution of VB could be considered as a distinguishing element among them. Moreover, besides HT, the most studied polyphenol for its biological properties is VB. Recent studies have assessed the biological activities of VB, consistent with disease prevention including antioxidant and anti-inflammatory activity (Cardinali et al., 2010, 2012; Esposito et al., 2010; Korkina, 2007; Speranza et al., 2010). Interesting is the presence of comselogoside and caffeoyl-6secologanoside, already identified in OMWW (Obied, Bedgood et al.,
306
I. D'Antuono et al. / Food Research International 65 (2014) 301–310
2007) and olive fruits (Kanakis et al., 2013; Obied, Karuso et al., 2007) that exhibit antioxidant activity comparable to other compounds (Obied, Bedgood, Prenzler, & Robards, 2007). The last step of filtration, NF (Fig. 3c), showed only low molecular weight phenolic compounds, such as HT, Tyr, CAA and CUA. Caffeic acid, although present at very low concentration (from 3 to 5 μg/mL) it is reported to have high antioxidant activity (Obied, Prenzler et al., 2008). The occurrence of specific biophenols in OMWW depends on the fruit cultivar and maturity (Mulinacci et al., 2001; Obied, Bedgood, Mailer, Prenzler, & Robards, 2008), and storage time (Obied, Bedgood, Prenzler, & Robards, 2008) in addition to the processing extraction technique (De Marco, Savarese, Paduano, & Sacchi, 2007). The endogenous enzymes and microbial activities cause a loss of the recovered phenols. According to Obied, Bedgood, Prenzler et al. (2008) none of storage conditions studied (storage at 4 °C, preserve by 40% w/w ethanol and 1% w/w acetic acid and storage at 4 °C) could prevent the rapid decrease in phenolic concentrations and antioxidant capacity, which happened within the first 24 h. Among the three fractions obtained, the UF was the best balance between the purification degree and the polyphenol enrichments (Garcia-Castello et al., 2010). For this reason the UF fraction was deeper characterized using LC/DAD/ESI–MSn analysis in order to identify the phenolic profile of each cultivar. 3.3. LC–MS analysis The LC/DAD/ESI–MSn analysis of the UF fractions from the five OMWW cultivars, led to the separation and identification of the majority of the constituents. Most of the compounds that were identified belonged to the following classes of constituents: secoiridoids and their derivatives, phenyl alcohols, phenolic acids and derivatives and
flavonoids. MS data were acquired in negative ionization mode, because polyphenols contain one or more hydroxyl and/or carboxylic acid groups. Identification was based on accurate mass measurements of the pseudomolecular [M–H]− ions and their fragmentation pattern, as it has been documented in the literature. In Fig. 4, the total ion current (TIC) chromatogram and UV chromatograms at 280, 240 and 340 nm of UF fraction of Coratina is presented (for the rest of the UF fractions of different cultivars the respective chromatograms are given in Supplementary data Supplements 1–4). Data obtained from the ESI-MSn analysis of the UF fractions are summarized in Table 2. Peak 1 exhibited a base peak [M–H]− at m/z 191, the MS2 spectrum obtained by fragmentation of the ion m/z 191 presented fragment ions at m/z 127 ([M–CO–2H2O]−) and m/z 173 ([M–H2O]−) which correspond to literature reports for quinic acid (Gouveia & Castilho, 2010). Peak 2 exhibited a molecular ion at m/z 169, the fragmentation of this pseudomolecular ion yielded a fragment at m/z 151 probably produced by the loss [M–H–H2O]− and was tentatively identified as 3,4-dihydroxyphenylglycol, that has previously identified in OMWW (Obied, Bedgood et al., 2007). Peaks 3 and 4 showed similar MS spectra, with a molecular ion at m/z ratio 315, and similar MS/MS spectra, thus suggesting the presence of two stereoisomers that were not distinguishable by mass spectrometry. The fragmentation pattern revealed two main fragment ions at the following m/z ratios: 153, which is formed by the loss of a glucose group and 123 corresponding to loss of the CH2OH group. The two isomers were attributed to HT glucoside. Three isomers of HT glucoside have been identified in Olea europaea, namely hydroxytyrosol-1-O-glucoside, hydroxytyrosol-3′-O-glucoside and hydroxytyrosol-4′-O-glucoside (Obied, Bedgood et al., 2007). Romero, Brenes, Garcia, and Garrido (2002), found that hydroxytyrosol-4-β-Dglucoside was the most abundant isomer in olive fruits and derived products.
Fig. 4. Total ion chromatogram and UV chromatograms at 280, 240 and 340 nm of OMWW Coratina UF fraction. Main compounds of extract: 1: quinic acid, 1′: cornoside, 2′: hydroxylated product of decarboxymethyl elenolic acid, 2: 3,4-dihydroxyphenylglycol, 5: HT, 4′: decarboxymethyl-elenolic acid derivative, 6: hydroxylated product of dialdehydic form of decarboxymethyl elenolic acid, 7: Tyr, 9: CAA, 12–13: β-hydroxyverbascoside diastereoisomer, 15: Ver, 18–19: elenolic acid derivative, 21: SEC, 23: COM.
I. D'Antuono et al. / Food Research International 65 (2014) 301–310
307
Table 2 Main ions identified by HPLC-DAD–MSn in the OMWW extracts and their proposed structures. Rt (min)
[M–H]− (m/z)
1 2 1' 2' 3 4 5 3' 4' 6
2.5 4.1 5.5 7.8 9.8 10.1 10.5 11.2 12.1 12.8
191 169 315 199 315 315 153 407 183 199
7 5' 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23
14.3 15.2 16.3 18.2 18.6 19.6 20.9 21.3 22.9 25.6 26.1 27.4 27.7 28.2 29.9 30.7 32.6 35.7
137 353 389 179 183 377 639 639 163 623 609 623 241 241 381 551 539 535
Peak
−
-MS3 [base peak]− (m/z) (%)
Compounds
111 (100), 173 (26), 127 (13) 151 (100) 151 (100), 255 (76) – 153 (100), 123 (20) 153 (100), 123 (20) 123 (100) 389 (100), 375 (88), 357 (72), 313 (58) 111 (100), 165 (14) 111 (100), 155 (78), 181 (51)
– – – – 123(100) – – 313 (100), 357 (74), 161 (6) – –
– 191 (100) 345 (100), 209 (47), 165 (26) 135 (100) 139 (100), 95 (58) 197 (100), 153 (16) 621 (100), 529 (12), 459 (11), 179 (3) 621(100), 529(5), 459(9) 119(100) 421(100) 301(100), 271(6), 343(5) 421(100) 139(100), 127(58), 95(52), 165(26) 139(100), 127(67), 95(43), 165(30), 207(18) 331 (100), 349 (96), 363 (92), 151 (44) 507(100), 281(36), 179(26), 389(25) 275 (100), 307 (95), 377 (68) 491 (100), 265 (28), 389 (27)
– – – – – 153 (100) 459 (100), 469 (13), 179 (10) 459 (100), 469 (13), 179 (10) – 135(100), 315(92), 297(16), 161(16) – 135 (100), 315 (92), 297 (16), 161 (16) 95 (100) 95 (100) 151 (100), 195 (9), 287 (8) 161 (100), 345 (44), 393 (25), 281 (20) 139 (100), 95 (68), 245 (11) 145 (100), 345 (77), 265 (36), 307 (26)
Quinic acid 3,4-Dihydroxyphenylglycol Cornoside Hydroxylated product of decarboxymethyl elenolic acid Hydroxytyrosol glucoside Hydroxytyrosol glucoside HT Unknown Decarboxymethyl-elenolic acid derivative Hydroxylated product of daldehydic form of decarboxymethyl elenolic acid Tyr Unknown Oleoside CAA Decarboxymethyl elenolic acid (HyEDA) Oleuropein aglycon derivative β-Hydroxyverbascoside diastereoisomer β-Hydroxyverbascoside diastereoisomer CUA Ver Rutin IsoVB Elenolic acid derivative Elenolic acid derivative Hydroxytyrosol acyclodihydroelenolate SEC Oleuropein COM
MS2 [M–H]- (m/z) (%)
Peak 4′ exhibited base peaks at m/z 185 (100%) and at m/z 183 (70%), with fragments in its MS2 at m/z 111 (100%) and at m/z 165 (14%). The ion at m/z 183 could probably be assigned to the de(carboxymethyl) elenolic acid derivative ion (De La Torre-Carbot et al., 2005) and the product ion at m/z 111 might be caused by the loss of CO and COO of the elenolic derivative fragment (m/z 183) in aldehyde forms (Ramos et al., 2013). So this peak can be tentatively identified as another elenolic acid derivative. The spectrum generated for peak 6 yielded a deprotonated molecule at m/z 199, which could be attributed to a derivative of the dialdehydic form of decarboxymethyl elenolic acid. Peak 6 was tentatively identified as hydroxylated product of dialdehydic form of decarboxymethyl elenolic acid, as it presented a fragment at m/z 155 corresponding to a loss of 44 units, which is consistent with the fragmentation pattern of its nonderivative form (acid group decarboxylation) (Lozano-Sanchez et al., 2011). The mass spectrum of peak 8, displayed an intense peak at m/z 389 which formed two major fragments in the MS2 spectrum, one at m/z 345 and the other at m/z 209. The former corresponded to the loss of 44 Da, which can be justified by the elimination of a CO2 molecule of a carboxylic group, and the latter can be attributed to the Z fragment of a hexose (loss of 180 Da) (the hexose residue was attributed to glucose). These results are in agreement with the presence of oleoside (Cardoso et al., 2005). Peak 10 showed an intense pseudomolecular peak at m/z 183 in the ESI-MS spectrum which presented a fragment at m/z 139, corresponding to a loss of 44 units. This compound was tentatively identified as the dialdehydic form of decarboxymethyl elenolic acid, that has been detected before in table olives (Medina, Brenes, Romero, García, & De Castro, 2007) and in olive oil and wastes generated during the storage of extra virgin olive oil (Lozano-Sanchez et al., 2011). The mass spectrum of peak 11 exhibited a base peak at m/z 377 and in its MS2 spectrum showed fragments at m/z 197 and 153, which has been identified before as oleuropein aglycon derivative in various tautomeric forms (Bouaziz, Jemai, Khabou, & Sayadi, 2010). The abundant peak at m/z 639, that exhibited peaks 12 and 13, is due to the molecular ion [M–H]− of two diastereoisomers of the molecule
β-hydroxyacteoside or β-hydroxyverbascoside (MW 640) (Innocenti et al., 2006). MS/MS fragmentation of molecular ions at m/z 639 yielded to the main daughter ion at m/z 621, corresponding to the water loss, and three minor fragments at m/z 529, corresponding to the loss of a catechol unit, m/z 459 corresponding to the loss of a caffeyol group or rhamnose, and m/z 179, assigned to caffeic acid ion. Two diastereoisomeric structures of the β-hydroxyl derivative of verbascoside and two diastereoisomeric structures of the β-hydroxyl derivative of isoverbascoside have been recently identified in olive mill wastewater by Cardinali et al. (2012). All of these stereoisomers have the same fragmentation scheme and, therefore, are not distinguishable by mass spectrometry. The mass spectrum of peak 16 exhibited a base peak at m/z 609. Its MS2 spectrum showed fragments at m/z 301, an aglycon ion, and at m/z 271 confirm the presence of rutin (Savarese, De Marco, & Sacchi, 2007). Peaks 18 and 19 showed mass spectra with the [M–H]− molecular ion species at m/z 241. The MS/MS spectra obtained for the precursor ion at m/z 241 gave fragment ions with molecular weights of 139 [M– COOH–COOCH3]−, 127 [M–COOH–C4H5O]−, 95 [M–COOH–COOCH3– CHCHO]− and 165 [M–OH–COOCH3]−, which among others have been referred for dialdehydic form of carboxymethyl elenolic acid (Di Maio et al., 2013). These peaks can be tentatively identified as elenolic acid derivatives. The mass spectrum of peak 20 gave a molecular ion at m/z 381, attributed to hydroxytyrosol acyclodihydroelenolate, identified in OMWW extracts for the first time by Obied, Karuso et al. (2007). In its MS2 spectrum, it formed the following fragments, at m/z 363 (−18 Da) due to the loss of a H2O unit, at m/z 349 (− 32 Da), an ion related to the cleavage of the secoiridoid function of the molecule, occurring due to the respective loss of its CH3OH moiety, at m/z 331 (−50 Da) [M–H–H2O–CH3OH]− and at m/z 151. Peak 21, with a molecular weight at m/z 551, was attributed to caffeoyl-6-secologanoside. Fragmentation of this ion originated an intense signal at m/z 507 from the loss of 44 Da, [M–H–CO2]−, the species at m/z 389 representing the oleoside structure and the ion at m/z 179 is related to the caffeoyl group (Innocenti et al., 2006). The ESI-MS spectrum of peak 22 showed a pseudomolecular ion at m/z 539 with fragments in its MS2 spectrum
308
I. D'Antuono et al. / Food Research International 65 (2014) 301–310
consistent with the reported fragmentation scheme: the ion at m/z 377 arises from cleavage of the glycosyl bond; the ion at m/z 307 is justified by the loss of a C4H6O fragment, while the fragment at m/z 275 may derive from rearranged fragments. These results confirm the presence of oleuropein. The spectrum generated for peak 23 yielded a deprotonated molecule at m/z at 535 was attributed to 6′-p-coumaroyl secologanoside (comselogoside). The ESI-MS2 spectrum of that ion showed the main fragment at m/z 491 from the loss of 44 Da, [M–H–CO2]− along with the ionic species at m/z 389 corresponding to the oleoside ion (Obied, Karuso et al., 2007). Finally peaks 5, 7, 9, 14 15 and 17 were identified as HT, Tyr, CAA, CUA, VB and IsoVB respectively, by direct comparison of their retention times and UV spectra with those of the standards and by using their MS, MS2 and MS3 spectra. Overall twenty three compounds, most of them detected in trace amounts, were identified in almost all the UF fractions of the different cultivars. Results showed no significant qualitative differences in the HPLC–ESI MS phenolic profile among UF fractions from the OMWWs from different cultivars (Table 3). Among the UF fractions from the five OMWW cultivars the one coming from Coratina showed two more compounds indicated with 1′ and 2′. Peak 1′ exhibited a base peak at m/z 315. Its MS2 spectrum showed fragments at m/z 151 and at m/z 255. This compound could be tentatively identified as cornoside, a quinol glucoside identified in the vegetation water of olives (Limiroli, Consonni, Ranalli, Bianchi, & Zetta, 1996). Cornoside was thought to derive from oxidation of the glucoside of tyrosol. Peak 2′ exhibited a base peak at m/z 199 that could be tentatively identified as another hydroxylated product of decarboxymethyl elenolic acid (Kanakis et al., 2013).
In this UF fraction was additionally found and quantified, IsoVB, which was also detected in traces in Lianolia cultivar. However, quantitative differences were observed in some of the main phenolic compounds. An attempt was made to compare the integral-areas of the main peaks of the chromatograms in the three wavelengths, as the compounds identified were not available as standards. In Table 3 integral-areas are given for eight compounds. Differences were observed among the cultivars analyzed for peaks 4′ and 6. In the UF fractions of Lianolia and Cellina the above elenolic acid derivative (P.4′) and the hydroxylated product of dialdehydic form of decarboxymethyl elenolic acid (P.6) were most abundant in comparison with the rest of the cultivars. Also in the UF fraction of Cellina, followed by Lianolia, hydroxytyrosol glucoside (P.3) was most abundant. Moreover differences were observed in the content of the UF fractions regarding 3,4-dihydroxyphenylglycol (P.2) and β-hydroxyverbascoside diastereoisomers (P.12 and P.13). Coratina followed by Lianolia had the higher content of these compounds. It is noteworthy to mention that β-hydroxyverbascoside diastereoisomers were detected in traces in the rest of cultivars. Elenolic acid and derivatives constitute the iridoid part of several important secondary olive metabolites, such as oleuropein and ligstroside, and are often mentioned as one of their hydrolysis products. 3,4-Dihydroxyphenyl glycol is a hydroxylated derivative of hydroxytyrosol. This substance may be of interest in the fields of nutrition and pharmacology due to its powerful antioxidant properties. It is the main metabolite produced by deamination of the human neurotransmitter noradrenaline (norepinephrine). Rodríguez,
Table 3 Compounds identified in the OMWW extracts in each different variety. In parenthesis are given the integral-areas in the chromatographic profiles in three wavelengths (280, 240 and 340 nm) of UF fractions of OMWW of different cultivars for eight compounds. Rt (min)
Peak
1 (240 2 (280 1′ 2′ 3 (280 4 5 3′ 4′ (240 6 (240 7 5′ 8 9 10 11 12 (340 13 (340 14 15 16 17 18b (240 19b (240 20 21 22 23 a b
Compounds
2.5
Quinic acid
4.1
3,4-Dihydroxyphenylglycol
5.5 7.8 9.8
Cornoside Hydroxylated product of decarboxymethyl elenolic acid Hydroxytyrosol glucoside
nm) nm)
nm) 10.1 10.5 11.2 12.1
Hydroxytyrosol glucoside HT Unknown Decarboxymethyl-elenolic acid derivative
12.8 14.3 15.2 16.3 18.2 18.6 19.6 20.9
Hydroxylated product of dialdehydic form of decarboxymethyl elenolic acid Tyr Unknown Oleoside CAA Decarboxymethyl elenolic acid (HyEDA) Oleuropein aglycon derivative β-Hydroxyverbascoside diastereoisomer
21.3
β-Hydroxyverbascoside diastereoisomer
22.9 25.6 26.1 27.4 27.7
CUA Ver Rutin IsoVB Elenolic acid derivative
28.2
Elenolic acid derivative
29.9 30.7 32.6 35.7
Hydroxytyrosol acyclodihydroelenolate SEC Oleuropein COM
nm) nm)
nm) nm)
OMWW samples of different varieties Lianolia
Asprolia
Koroneiki
Coratina
Cellina
√ (1065) √ (1386) – – √ (1681) √ √ √ √ (15174) √ (9137) √ √ √ √ √ √ √ (507) √ (564) √ √ – √ √
√ (1471) √ (744) – – √ (207) √ √ √ √ (3148) √ (6853) √ – √ √ – √ √ (a) √ (a) √ √ – – √
√ (1684) √ (605) – – √ (558) √ √ √ √ (1089) √ (2590) √ – √ √ – √ √ (37) √ (39) √ √ – – √
√ (626) √ (2143) √ √ √ (255) √ √ √ √ (8719) √ (5654) √ √ √ √ √ √ √ (983) √ (1072) – √ √ √ √
√ (1066) √ (872) – – √ (2206) √ √ √ √ (9055) √ (16405) √ – √ √ √ √ √ (a) √ (a) √ √ √ – √
√ (3805) √ √ √ √
√ (5860) √ √ √ √
√ (568) √ √ √ √
√ (4038) √ √ √ √
√ (4045) √ √ √ √
nm) nm)
These compounds are detected in traces in the OMWW. The integral was taken for both compounds 18 and 19.
I. D'Antuono et al. / Food Research International 65 (2014) 301–310
Rodríguez, Fernández-Bolaños, Guillén, and Jiménez (2007) demonstrated that the antioxidant efficiency of 3,4-dihydroxyphenyl glycol in water is 2–3 times higher than that of ascorbic acid or hydroxytyrosol, whereas in lipidic medium it is comparable to that of vitamin E despite its high polarity. Moreover, De Roos et al. (2011) demonstrated that an alperujo extract may protect against platelet activation, platelet adhesion and possibly have anti-inflammatory properties, referring that a combination of hydroxytyrosol and 3,4dihydroxyphenylglycol were, at least partly, responsible for this effect. So it would be interesting to study in depth the in vitro antioxidant properties of these fractions from the OMWWs, as they might be of use in protecting food and pharmaceutical products by retarding the process of oxidation and deterioration during processing and storage. Also further studies of other bioactivities associated with the nonphenolic secoiridoids (elenolic acid and its derivatives) such as antimicrobial activity could be investigated. 4. Conclusion In this study OMWWs coming from Italian and Greek cultivars were processed using membrane technologies. With this system three fractions were recovered MF, UF and NF, containing a different percentage of polyphenols. The fractions were characterized by HPLC analysis, to quantify the main polyphenols present: HT, Tyr, CAA, CUA, VB, IsoVB, SEC and COM. The most abundant compound was HT that represent about 70–80% of the total phenolic concentration, followed by Tyr. On the contrary, in Coratina MF, the main phenolic compound was VB, that instead, was lower and in some cases absent, in all the other cultivars. The UF fractions that represented the best balance between the purification degree and the polyphenol enrichment were characterized by LC/DAD/ESI–MSn analysis. Overall twenty three compounds, most of them detected in trace amounts, were identified in almost all the UF fractions of the different cultivars. Results showed no significant qualitative differences, however, quantitative differences were observed in some of the main phenolic compounds. In detail, differences were observed among the different cultivars regarding the presence of elenolic acid derivatives, hydroxytyrosol glucoside, and β-hydroxyverbascoside diastereoisomers. This deeper characterization could help to understand how the olive cultivars can impact on the phenolic compositions of OMWW. Supplementary data to this article can be found online at http://dx. doi.org/10.1016/j.foodres.2014.09.033. Acknowledgments The authors want to thank the INTERREG IV European Territorial Cooperation Programme “Greece–Italy 2007–2013”: “Utilization of biophenols from Olea europaea products — Olives, virgin olive oil and olive mill wastewater-BIO-OLEA” project. Special thanks are given to the Mass Spectrometry Unit of the University of Ioannina for providing access to LC–MS/MS facilities. References Bianco, A., Buiarelli, F., Cartoni, G. P., Coccioli, F., Jasionowska, R., & Margherita, P. (2003). Analysis by liquid chromatography–tandem mass spectrometry of biophenolic compounds in olives and vegetation waters, part I. Journal of Separation Science, 26(5), 409–416. Bouaziz, M., Jemai, H., Khabou, W., & Sayadi, S. (2010). Oil content, phenolic profiling and antioxidant potential of Tunisian olive drupes. Journal of the Science of Food and Agriculture, 90(10), 1750–1758. Bouzid, O., Navarro, D., Roche, M., Asther, M., Haon, M., Delattre, M., et al. (2005). Fungal enzymes as a powerful tool to release simple phenolic compounds from olive oil by-product. Process Biochemistry, 40(5), 1855–1862. Capasso, R., Cristinzio, G., Evidente, A., & Scognamiglio, F. (1992). Isolation, spectroscopy and selective phytotoxic effects of polyphenols from vegetable waste waters. Phytochemistry, 31, 4125–4128.
309
Cardinali, A., Cicco, N., Linsalata, V., Minervini, F., Pati, S., Pieralice, M., et al. (2010). Biological activity of high molecular weight phenolics from olive mill wastewater. Journal of Agricultural and Food Chemistry, 58, 8585–8590. Cardinali, A., Pati, S., Minervini, F., D'Antuono, I., Linsalata, V., & Lattanzio, V. (2012). Verbascoside, isoverbascoside, and their derivatives recovered from olive mill wastewater as possible food antioxidants. Journal of Agricultural and Food Chemistry, 60, 1822–1829. Cardoso, S. M., Falcão, S. I., Peres, A. M., & Domingues, M. R. M. (2011). Oleuropein/ ligstroside isomers and their derivatives in Portuguese olive mill wastewaters. Food Chemistry, 129(2), 291–296. Cardoso, S. M., Guyot, S., Marnet, N., Lopes da Silva, J. A., Renard, C. M. G. C., & Coimbra, M. A. (2005). Characterization of phenolic extracts from olive pulp and olive pomace by electrospray mass spectrometry. Journal of the Science of Food and Agriculture, 85, 21–32. Cicco, N., Lanorte, M. T., Paraggio, M., Viggiano, M., & Lattanzio, V. (2009). A reproducible, rapid and inexpensive Folin–Ciocalteu micro method in determining phenolics of plant methanol extracts. Microchemical Journal, 91, 107–110. De La Torre-Carbot, K., Jauregui, O., Gimeno, E., Castellote, A. I., Lamuela-Raventós, R. M., & López-Sabater, M. C. (2005). Characterization and quantification of phenolic compounds in olive oils by solid-phase extraction, HPLC-DAD, and HPLC–MS/MS. Journal of Agricultural and Food Chemistry, 53(11), 4331–4340. De Marco, E., Savarese, M., Paduano, A., & Sacchi, R. (2007). Characterization and fractionation of phenolic compounds extracted from olive oil mill wastewaters. Food Chemistry, 104, 858–867. De Roos, B., Zhang, X., Rodriguez Gutierrez, G., Wood, S., Rucklidge, G. J., Reid, M. D., et al. (2011). Anti-platelet effects of olive oil extract: In vitro functional and proteomic studies. European Journal of Nutrition, 50(7), 553–562. Di Maio, I., Esposto, S., Taticchi, A., Selvaggini, R., Veneziani, G., Urbani, S., et al. (2013). Characterization of 3,4-DHPEA-EDA oxidation products in virgin olive oil by high performance liquid chromatography coupled with mass spectrometry. Food Chemistry, 138(2–3), 1381–1391. Esposito, E., Mazzon, E., Paterniti, I., Dal Toso, R., Pressi, G., Caminiti, R., et al. (2010). PPAR-α contributes to the anti-inflammatory activity of verbascoside in a model of inflammatory bowel disease in mice. PPAR Research, http://dx.doi.org/10.1155/ 2010/917312 (Article ID 917312, 10 pp.). European Food Safety Authority (EFSA) (2011). Polyphenols in olive related health claims. EFSA Journal, 9, 2033–2057. Garcia-Castello, E., Cassano, A., Criscuoli, A., Conidi, C., & Orioli, E. (2010). Recovery and concentration of polyphenols from olive mill wastewaters by integrated membrane system. Water Resources, 44, 3883–3892. Ginos, A., Manios, T., & Mantzavinos, D. (2006). Treatment of olive mill effluents by coagulation–flocculation–hydrogen peroxide oxidation and effect on phytotoxicity. Journal of Hazardous Materials, 133(1–3), 135–142. Gouveia, S. C., & Castilho, P. C. (2010). Characterization of phenolic compounds in Helichrysum melaleucum by high-performance liquid chromatography with on-line ultraviolet and mass spectrometry detection. Rapid Communications in Mass Spectrometry, 24(13), 1851–1868. Innocenti, M., La Marca, G., Malvagia, S., Giaccherini, C., Vincieri, F. F., & Mulinacci, N. (2006). Electrospray ionisation tandem mass spectrometric investigation of phenylpropanoids and secoiridoids from solid olive residue. Rapid Communications in Mass Spectrometry, 20(13), 2013–2022. Kanakis, P., Termentzi, A., Michel, T., Gikas, E., Halabalaki, M., & Skaltsounis, A. L. (2013). From olive drupes to olive oil. An HPLC–orbitrap-based qualitative and quantitative exploration of olive key metabolites. Planta Medica, 79(16), 1576–1587. Korkina, L. G. (2007). Phenylpropanoids as naturally occurring antioxidants: From plant defense to human health. Cellular and Molecular Biology, 53, 15–25. Lattanzio, V. (1982). High performance reversed phase liquid chromatography of free and bound phenolic acids in eggplant (Solanum melongena L.). Journal of Chromatography, 250, 143–148. Limiroli, R., Consonni, R., Ranalli, A., Bianchi, G., & Zetta, L. (1996). 1H NMR study of phenolics in the vegetation water of three cultivars of Olea europaea: Similarities and differences. Journal of Agricultural and Food Chemistry, 44(8), 2040–2048. Lozano-Sanchez, J., Giambanelli, E., Quirantes-Piné, R., Cerretani, L., Bendini, A., SeguraCarretero, A., et al. (2011). Wastes generated during the storage of extra virgin olive oil as a natural source of phenolic compounds. Journal of Agricultural and Food Chemistry, 59, 11491–11500. Medina, E., Brenes, M., Romero, C., García, A., & De Castro, A. (2007). Main antimicrobial compounds in table olives. Journal of Agricultural and Food Chemistry, 55(24), 9817–9823. Mulinacci, N., Romani, A., Galardi, C., Pinelli, P., Giaccherini, C., & Vincieri, F. F. (2001). Polyphenolic content in olive oil waste waters and related olive samples. Journal of Agriculture and Food Chemistry, 49, 3509–3514. Nissiotis, M., & Tasioula-Margari, M. (2002). Changes in antioxidant concentration of virgin olive oil during thermal oxidation. Food Chemistry, 77, 371–376. Obied, H. K., Allen, M. S., Bedgood, D. R., Prenzler, P. D., Robards, K., & Stockmann, R. (2005). Bioactivity and analysis of biophenols recovered from olive mill waste. Journal of Agricultural and Food Chemistry, 53(4), 823–837. Obied, H. K., Allen, M. S., Bedgood, D. R., Jr., Prenzler, P. D., & Robards, K. (2005). Investigation of Australian olive mill waste for recovery of biophenols. Journal of Agricultural and Food Chemistry, 53(26), 9911–9920. Obied, H. K., Bedgood, D., Mailer, R., Prenzler, P. D., & Robards, K. (2008). Impact of cultivar, harvesting time, and seasonal variation on the content of biophenols in olive mill waste. Journal of Agricultural and Food Chemistry, 56(19), 8851–8858. Obied, H. K., Bedgood, D. R., Jr., Prenzler, P. D., & Robards, K. (2007). Chemical screening of olive biophenol extracts by hyphenated liquid chromatography. Analytica Chimica Acta, 603(2), 176–189.
310
I. D'Antuono et al. / Food Research International 65 (2014) 301–310
Obied, H. K., Bedgood, D. R., Jr., Prenzler, P. D., & Robards, K. (2008). Effect of processing conditions, prestorage treatment, and storage conditions on the phenol content and antioxidant activity of olive mill waste. Journal of Agricultural and Food Chemistry, 56(11), 3925–3932. Obied, H. K., Karuso, P., Prenzler, P. D., & Robards, K. (2007). Novel secoiridoids with antioxidant activity from Australian olive mill waste. Journal of Agricultural and Food Chemistry, 55(8), 2848–2853. Obied, H. K., Prenzler, P. D., Konczak, I., Rehman, A. U., & Robards, K. (2009). Chemistry and bioactivity of olive biophenols in some antioxidant and antiproliferative in vitro bioassays. Chemical Research in Toxicology, 22(1), 227–234. Obied, H. K., Prenzler, P. D., & Robards, K. (2008). Potent antioxidant biophenols from olive mill waste. Food Chemistry, 111, 171–178. Paraskeva, C. A., Papadakis, V. G., Tsarouchi, E., Kanellopoulou, D. G., & Koutsoukos, P. G. (2007). Membrane processing for olive mill wastewater fractionation. Desalination, 213, 218e229. Ramos, P., Santos, S. A. O., Guerra, Â. R., Guerreiro, O., Felício, L., Jerónimo, E., et al. (2013). Valorization of olive mill residues: Antioxidant and breast cancer antiproliferative activities of hydroxytyrosol-rich extracts derived from olive oil by-products. Industrial Crops and Products, 46, 359–368.
Rodríguez, G., Rodríguez, R., Fernández-Bolaños, J., Guillén, R., & Jiménez, A. (2007). Antioxidant activity of effluents during the purification of hydroxytyrosol and 3,4-dihydroxyphenyl glycol from olive oil waste. European Food Research and Technology, 224(6), 733–741. Romero, C., Brenes, M., Garcia, P., & Garrido, A. (2002). Hydroxytyrosol 4-β-D-glucoside, an important phenolic compound in olive fruits and derived products. Journal of Agricultural and Food Chemistry, 50, 3835–3839. Russo, C. (2007). A new membrane process for the selective fractionation and total recovery of polyphenols, water and organic substances from vegetation waters (VW). Journal of Membrane Science, 288, 239–246. Savarese, M., De Marco, E., & Sacchi, R. (2007). Characterization of phenolic extracts from olives (Olea europaea cv. Pisciottana) by electrospray ionization mass spectrometry. Food Chemistry, 105(2), 761–770. Speranza, L., Franceschelli, S., Pesce, M., Reale, M., Menghini, L., Vinciguerra, I., et al. (2010). Antiinflammatory effects in THP-1 cells treated with verbascoside. Phytotherapy Research, 24, 1398–1404. Visioli, F., Romani, A., Mulinacci, N., Zarini, S., Conte, D., Vincieri, F. F., et al. (1999). Antioxidant and other biological activities of olive mill waste waters. Journal of Agriculture and Food Chemistry, 47, 3397–3401.