Chemico-Biological Interactions 180 (2009) 421–432

Contents lists available at ScienceDirect

Chemico-Biological Interactions

journa l homepage: www.e lsev ier .com/ locate /chembio int

Hypoglycemic and antioxidant effects of phenolic extracts and purifiedhydroxytyrosol from olive mill waste in vitro and in rats

Khaled Hamdena,!,1, Noureddine Alloucheb,1, Mohamed Damakb, Abdelfattah Elfekia

a Animal Ecophysiology, Faculty of Sciences, Sfax, Tunisiab Laboratory of Chemistry of Natural Products, Faculty of Sciences of Sfax, B.P. 1171, 3000 Sfax, Tunisia

a r t i c l e i n f o

Article history:Received 17 March 2009Received in revised form 7 April 2009Accepted 8 April 2009Available online 23 April 2009

Keywords:DiabetesHydroxytyrosolIn vitroMetabolic disordersLiver and kidney

a b s t r a c t

This study aimed to evaluate the effect of phenolic extract and purified hydroxytyrosol (HT) from olive millwaste (OMW) on oxidative stress and hyperglycemia in alloxan-induced diabetic rats. The OMW biophe-nols were extracted using ethyl acetate. The obtained extract was fractionated by solid phase extraction(SPE) experimentation to generate two fractions: (F1) and (F2). HPLC-UV and HPLC-MS analysis showedthat (F1) was made of known OMW monomeric phenolics mainly hydroxytyrosol (HT) while (F2) con-tained oligomeric and polymeric phenols such as verbascosid and ligstrosid. (HT) was purified from (F1)using silica gel-column chromatography and silica gel-TLC techniques. In incubated pancreas, supple-mentation of OMW fractions enhanced insulin secretion. The administration of OMW extract fractions(F1) and (F2) as well as purified (HT) in diabetic rats caused a decrease in glucose level in plasma andan increase in renal superoxide dismutase (SOD), catalase (CAT) and glutathione peroxidase (GPX) activ-ities in liver and kidney. Furthermore, a protective action against hepatic and renal toxicity in diabeticrats was clearly observed. Furthermore, a significant decrease in hepatic and renal indices toxicity wasobserved, i.e. alkalines phosphatases (ALP), aspartate and lactate transaminases (AST and ALT) activitiesand the thiobarbituric acid-reactive substances (TBARs), total and direct bilirubin, creatinine and urealevels. In addition, (F1), (F2) and especially (HT) decreased triglycerides (TG), total-cholesterol (T-Ch) andhigher HDL-cholesterol (HDL-Ch) in serum. These beneficial effects of OMW biophenols were confirmedby histological findings in hepatic, renal and pancreatic tissues of diabetic rats. This study demonstratesfor the first time that OMW polyphenols and especially (HT) are efficient in inhibiting hyperglycemiaand oxidative stress induced by diabetes and suggests that administration of HT may be helpful in theprevention of diabetic complications associated with oxidative stress.

© 2009 Elsevier Ireland Ltd. All rights reserved.

1. Introduction

Type 1 diabetes is a metabolic disorder caused by complete orrelative insufficiency of insulin secretion [1] and is the main causeof numerous complications related to many diseases. Chronic ele-vation of blood glucose will eventually lead to tissue damage, withconsequent often serious disease. While evidence of tissue damagecan be found in many organ and systems [1,2]. Hyperglycemia leadsto long-term complications of diabetes, which are the major causesof morbidity and mortality in human populations [3]. Increased freeradical generation and oxidative stress are hypothesized to play animportant role in the pathogenesis of diabetes and its late compli-cations [3]. In the past few years, polyphenolics substances have

! Corresponding author at: Lab. of Animal Ecophysiology, Department of Life Sci-ence, Faculty of Science, University of Sfax, PB 802, 3018 Sfax, Tunisia.Tel.: +216 97 469 111; fax: +216 74 274 437.

E-mail address: khaled.hamden@yahoo.fr (K. Hamden).1 This work was performed equally.

received widespread attention because of their potential for pre-venting some highly prevalent chronic diseases. In fact it has beenreported that polyphenols are endowed with interesting biologi-cal activities such as anti-inflammatory, antioxidant, antiallergic,hepatoprotective, antithrombotic, antiviral, and anticarcinogenicactivities [4–6]. In addition, olive mill waste (OMW) biophenolsare efficient components for the low incidence of cardiovasculardiseases in the Mediterranean area [7]. A hen, olive mill waste ispotentially a rich source of a diverse range of biophenols with awide array of biological activities [7]. The beneficial health effects ofOMW have been mainly attributed to its elevated phenols content[8]. Among them, hydroxytyrosol (3,4-dihydroxyphenylethanol)stands out as a compound of high added-value, due to its interest-ing antioxidant and potential beneficial human health properties[9]. Hydroxytyrosol has various biological activities, such as down-regulation of the immunological response [10], preventing humanerythrocytes from oxidative damage induced by hydrogen peroxide[11], anti-inflammatory, antithrombotic, and hypocholesterolemiceffects in rats [12–14]. Animal study has shown that administra-tion of 100 mg/kg hydroxytyrosol for 12 days enhanced resistance

0009-2797/$ – see front matter © 2009 Elsevier Ireland Ltd. All rights reserved.doi:10.1016/j.cbi.2009.04.002

422 K. Hamden et al. / Chemico-Biological Interactions 180 (2009) 421–432

of dissociated brain cells to oxidative stress, as shown by reducedbasal and stress-induced lipid peroxidation [13]. Furthermore, aprevious report has indicated that hydroxytyrosol is able to crossthe blood–brain barrier [14]. The present study describes the bio-phenols extraction, fractionation and identification from OMW. Inaddition, it deals with the evaluation of hypoglycemic and antiox-idant activities of the OMW monomeric and polyphenolic phenolsas well as of purified hydroxytyrosol.

2. Materials and methods

2.1. Materials

Fresh olive mill waste (OMW) was supplied by a discon-tinuous three phase olive processing mill from a cooperativein Sfax (Tunisia). This sample was taken at the middle ofthe olive harvest season (January 2008) and conserved at 4 "C.Tyrosol, caffeic acid, para-coumaric acid, protocatechuic acid, para-hydroxyphenylacetic acid (PHPA), luteolin, apigenin and rutin werepurchased from Sigma–Aldrich. Hydroxytyrosol was purified in ourlaboratory from OMW as described previously [15]. Oleuropein andligstrosid were obtained from Extrasynthese (France). All solventswere analytical grade and used without further purification.

2.2. Extraction of phenolic compounds from OMW

Liquid–liquid extraction with ethyl acetate was carried out onOMW (4 l) in a separatory funnel. The mixture (solvent-OMW) wasvigorously shaken for 10 min to achieve equilibrium and was thenallowed to settle for 30 min. The phases were separated and theextraction was repeated successively four times. All the runs wereperformed at ambient temperature (20–25 "C). The ethyl acetatecombined fractions were dried over anhydrous sodium sulphateand the ethyl acetate was evaporated in a rotary evaporator at 40 "Cto yield 47.31 g. The obtained dry residue was used for characteri-sation, quantification and fractionation of phenolic compounds aswell as for hydroxytyrosol purification.

2.3. OMW extracts fractionation

Fractionation of OMW phenolic extract was carried out bysolid phase extraction (SPE) experimentation. OMW extract (45 g)was subjected to a flash chromatography on reversed-phase(VersaPakTM, C18 Cartridge 40 mm # 75 mm) under medium pres-sure using a gradient mobile phase consisting of H2O–MeCN from80:20 to 0:100. The flow rate was adjusted to 10 ml/min and 8 mlfractions were collected. These fractions were measured by opticaldensity at 280 nm and the chromatogram (optical density versusfraction number) was represented (data not shown). Two separatedpeaks were obtained. The corresponding fractions were collectedinto two sub-fractions: (F1) and (F2) related to the first and thesecond separated peaks, respectively.

2.4. Hydroxytyrosol purification

For hydroxytyrosol purification, an aliquot (4 g) of the dryfirst fraction (F1) was chromatographed on a silica gel-column(4 cm # 80 cm) eluted with a solvent gradient: hexane–ethylacetate–methanol. The chromatographic behavior was followed byTLC analysis using silica gel plates (Merck, 60 F-254). Nine groups ofhomogeneous fractions were collected. The third group was puri-fied by prep.TLC, eluted with a mixture of chloroform:methanol(8:2, v/v), which afforded one pure compound with Rf = 0.6. Thiscompound was identified as hydroxytyrosol (designed F3) bymeans of HPLC-UV and HPLC-MS analyses.

2.5. HPLC separation and identification of phenolic compounds

The presence and amount of phenolic compounds in the OMWextract were studied by reversed-phase HPLC analysis using abinary gradient elution. The analysis was performed by reversed-phase HPLC on a Waters autopurification system equipped witha binary pump (Waters 2525), a UV–vis diode array detector(190–600 nm, Waters 2996) and a PL–ELS 1000 ELS detector Poly-mer Laboratory. The chromatographic separation was achieved ona Kromasil C18 column (250 mm # 4.6 mm, I.D., 5 m, Thermo). Itstemperature was maintained at 40 "C. The mobile phase was 0.1%formic acid in water (A) versus 0.1% formic acid in acetonitrile (B)for a total running time of 60 min. The elution conditions were:0–30 min, 20–50% B; 30–35 min, 50% B; 35–45 min, 50–100% B;45–55 min, 100% B; 55–60 min, 100–20% B. Finally, the column wassubjected to washing and reconditioning steps for 10 min with 20%B. The flow rate was 0.6 ml/min and the injection volume was 50 !l.The main phenolic compounds were identified by comparison withrelative retention times and UV spectra of pure standards, whenavailable, or by comparing the relative elution order and UV spectrawith those reported in the literature [16,17].

The identity of each peak was confirmed by LC–MS, performedon an Agilent 1100 LC system consisting of degasser, binary pump,auto sampler, and column heater. The column outlet was coupledto an Agilent MSD Ion Trap XCT mass spectrometer equipped withan ESI ion source. Data acquisition and mass spectrometric evalu-ation was carried out on a personal computer with Data Analysissoftware (Chemstations). For the chromatographic separation, thesame conditions (coloumn, solvent gradient and flow rate), givenpreviously, were used. The following parameters were employedthroughout all MS experiments: for electrospray ionisation withpositive ion polarity the capillary voltage was set to 3.5 kV, the dry-ing temperature to 350 "C, the nebulizer pressure to 40 psi, and thedrying gas flow to 10 l/min. The maximum accumulation time was50 ms, the scan speed was 26,000 mz$1 s$1 (ultra scan mode) andthe fragmentation time was 30 ms. Identification of compounds byLC–MS analysis was carried out by comparing retention times andmass spectra of the unknown peaks to those of standards, whenavailable.

2.6. Animals

Male Wistar rats, with body weights of 180–200 g and bred inthe Central Animal House and obtained from the Central Pharmacy,Tunisia, were used in this study. The animals were fed on a pelletdiet (Socco, Sfax, Tunisia) and water ad libitum. The animals weremaintained in a controlled environment under standard conditionsof temperature and humidity with an alternating light-and-darkcycle. The handling of the animals was approved by the TunisianEthical Committee for the care and use of laboratory animals.

2.7. Preparation of pancreas incubation

Male Wistar rats (180–250 g) were anesthetized with sodiumpentobarbital solution (50 mg/kg, i.p.) and the pancreas was iso-lated and perfused at 37 "C according to the method describedelsewhere [18,19]. F1, F2 and purified hydroxytyrosol were dis-solved in KRB buffer (pH 7.4). The pancreas were removed,transferred and then incubated in KRB buffer (pH 7.4). The perfusatewas incubated for 40 min in 5 ml KRB medium at 37 "C equilibratedwith 95% O2 + 5% CO2 gas. Pancreas were divided in to three sets:set 1: control, tissues incubated in KRB medium a final concen-tration of glucose 1 g/l; set 2: tissues incubated in KRB medium afinal concentration of glucose 4 g/l; sets 3–5: slices incubated inKRB medium a final concentration of glucose 4 g/l + 50 !g/ml F1,F2 and F3, respectively. Triplicate cultures were set up for each

K. Hamden et al. / Chemico-Biological Interactions 180 (2009) 421–432 423

concentration to minimize the errors. Effluent was fractionally col-lected every 10 min, and its insulin concentration was measured bya radioimmunoassay kit for Bi-insulin RIA Diagnostic, Pasteur, Paris,France.

2.8. Experimental induction of diabetes

The methods of diabetes induction in rats by alloxan as describedelsewhere [20]. Briefly rats were single injected intraperitoneallywith a freshly prepared solution of alloxan monohydrate in normalsaline at a dose (150 mg/kg BW) freshly dissolved in NaCl 0.9% buffer(pH 7). Because alloxan is capable of producing fatal hypoglycemiaas a result of massive pancreatic insulin release, rats were treated

with 20% glucose solution orally after 6 h. The rats were then keptfor the next 24 h on 5% glucose solution bottles in their cages toprevent hypoglycemia. After 2 weeks, the rats a fatal hyperglycemiawere chosen for the experiment (i.e. with blood glucose levels of2 g/l) [20].

2.9. Experimental procedure

The rats were randomly divided into five experimental groups(n = 8). Group 1: normal rats. Group 2: diabetic rats (glycemiawas superior to 2 g/l). Groups 3: diabetic rats treated with insulin0.5 IU/rat/day (Novo Nordisk AIS, Danemak). Groups 4–6: diabeticrats received daily, respectively; OMW monomeric phenols (F1),

Fig. 1. HPLC chromatograms of OMW ethyl acetate extract (A) and the two fractions obtained from the OMW extract by means of reversed-phase SPE: F1 (B) and F2 (C). 1,hydroxytyrosol; 2, 3, unknown compounds; 4, tyrosol; 5, cafeic acid; 6, protocatechuic acid; 7, p-coumaric acid; 8, luteolin 7-O-glucosid; 9, ligstrosid; 10, luteolin; 11, elenolicacid; 12, apigenin; 13, verbascosid; 14, oleuropein; 15, rutin; P, polymeric substances.

424 K. Hamden et al. / Chemico-Biological Interactions 180 (2009) 421–432

Fig. 2. Mass spectrum obtained in full scan mode acquisition for hydroxytyrosol.

Table 1Identification and concentrations of the principal compounds in the OMW extract fractions: F1 and F2.

Peak Identity Retention times (min) !max (nm) Major ESI$ peaks Concentration (mg/l)a

1 Hydroxytyrosol 16.59 220, 277 153 1453 ± 532 Unknown 18.12 226, 301 181 n.d.b

3 Unknown 18.43 230, 286 193 n.d.b

4 Tyrosol 19.34 219, 275 137 536 ± 435 Cafeic acid 21.42 220, 245, 322 179 423 ± 676 Protocatechuic acid 26.86 235, 289 153 326 ± 357 p-Coumaric acid 29.91 234, 308 163 72 ± 168 PHPAc 31.49 235, 280 151 125 ± 349 Verbascosid 33.42 216, 291, 328 623 n.d.b

10 Luteolin 34.32 240, 265, 345 285 234 ± 5611 Elenolic acid 36.46 240, 280 241 n.d.b

12 Apigenin 37.86 235, 270, 335 269 86 ± 2613 Ligstrosid 39.12 217, 286, 315 523 86 ± 1514 Oleuropein 40.18 228, 278 539 342 ± 2215 Rutin 43.09 218, 254, 352 609 36 ± 12

a Concentrations in crude OMW.b Not determined.c Para-hydroxyphenylacetic acid.

OMW polymeric phenols (F2) and purified hydroxytyrosol (F3)which were administrated by intraperitonial injection i.p. at dose20 mg/kg BW. After 2 months of treatment, the animals were sac-rificed by decapitation, and the trunk blood collected. The serumwas prepared by centrifugation (1500 # g, 15 min, 4 "C) and thekidneys and liver were removed, cleaned of fat and weighed; allthese samples were stored at $80 "C until used. Pieces of pancreas,liver and kidney were fixed in a Bouin solution for histologicalstudies.

2.10. Analytical methods

The lipid peroxidation in the liver and kidneys of control and alltreated groups of animals was measured by the quantification ofthiobarbituric acid-reactive substances (TBARS) determined by themethod of Buege and Aust [21]. The activity of superoxide dismu-tase was assayed by the spectrophotometric method of Marklundand Marklund [22]. The GPx activity was measured by the methoddescribed by Pagila and Valentine [23]. CAT was assayed colorimet-

Table 2Effects of OMW phenolic extracts on hepato-toxicity indices (AST, ALT, ALP, total and direct bilirubin) in plasma of diabetic rats. Values are given as mean ± S.D. for 8 rats ineach group. Values differ significantly at p < 0.05.

Groups AST ALT ALP T-bilirubin d-Bilirubin

Control 32.7 ± 4.5 37.9 ± 07 157 ± 13 1.21 ± 0.2 0.16 ± 0.05Diab 109 ± 21a 119 ± 20a 243 ± 29a 2.86 ± 0.3a 0.66 ± 0.11a

Diab + Ins 83.5 ± 15ab 73.9 ± 19ab 221 ± 23ab 2.19 ± 0.8ab 0.42 ± 0.13ab

Diab + F1 73.1 ± 9ab 70.5 ± 06ab 195 ± 24ab 1.76 ± 0.1abc 0.32 ± 0.11ab

Diab + F2 77.4 ± 18ab 59.5 ± 10abcd 170 ± 09abc 1.55 ± 0.2abcd 0.28 ± 0.03abcd

Diab + F3 63.3 ± 4abcde 60.6 ± 07abcd 167 ± 10abcde 1.38 ± 0.1abcde 0.26 ± 0.03abcde

ap < 0.05 as control rats.bp < 0.05 as diabetic rats.cp < 0.05 as diabetic rats treated with insulin.dp < 0.05 as diabetic rats treated with F1.ep < 0.05 as diabetic rats treated with F2.

K. Hamden et al. / Chemico-Biological Interactions 180 (2009) 421–432 425

Fig. 3. Effects of monomeric and polymeric substances (F1 and F2) and purifiedhydroxytyrosol (F3) on plasma glucose levels and liver glycogen levels in normal andalloxan-induced diabetic male wistar rats for 60 days. Values are given as mean± S.D.for 8 rats in each group. Values differ significantly at p < 0.05. Statistical analysis asin legend of Table 2.

rically at 240 nm and expressed as moles of H2O2 consumed perminute per milligram of protein, as described by Aebi [24]. The levelof total protein was determined by the method of Lowry et al. usingbovine serum albumin as the standard at 660 nm [25]. The activityof alkalines phosphatases (ALP), aspartate and alanine transami-nases (AST & ALT) and the level of blood and hepatic glucose, totaland direct bilirubin, creatinine, urea, cholesterol, triglyceride, HDL-cholesterol in serum were measured using commercial kits fromBiomagreb, Tunis, Tunisia and Biomerieux, Lyon, France.

For histological studies, pieces of pancreas were fixed in a Bouinsolution for 24 h, and then embedded in paraffin. Sections of 5 !mthickness were stained with hematoxylin–eosin and examinedunder the Olympus CX41 light microscope.

Table 3Effects of monomeric and polymeric substances (F1 and F2) and purified hydrox-ytyrosol (F3) on rat total-cholesterol (T-Ch), triglycerides (TG), and high-densitylipoprotein cholesterol (HDL-C) in diabetic rats. Values are given as mean± S.D. for8 rats in each group. Values differ significantly at p < 0.05. Statistical analysis as inlegend of Table 2.

Groups TG T-Ch HDL-Ch

Control 0.81 ± 0.02 0.63 ± 0.12 0.53 ± 0.08Diab 2.01 ± 0.34a 1.52 ± 0.11a 0.41 ±0.03a

Diab + Ins 1.68 ± 0.43ab 1.19 ± 0.14ab 0.43a

Diab + F1 1.51 ± 0.05ab 0.91 ± 0.18abc 0.51 ± 0.03bc

Diab + F2 1.09 ± 0.1abc 0.9 ± 0.17abc 0.58 ± 0.05bc

Diab + F3 0.93 ± 0.160abcd 0.72 ± 0.10abcde 0.69 ± 0.12abcde

2.11. Statistics

Data are presented as means ± S.D. The determinations wereperformed from 8 animals per group and the differences wereexamined by the one-way analysis of variance (ANOVA) followedby the Fisher test (Stat View) and the significance was accepted atp < 0.05.

3. Results

3.1. Polyphenols extraction and fractionation from OMW

OMW represents a complex medium containing mainlypolyphenols of different molecular mass. Several previous inves-tigations have demonstrated that the OMW monomeric fraction isendowed with interesting biological activities. For its recovery fromOMW, we have used a liquid–liquid solvent extraction procedure. Inprevious investigations, it was reported that among all proceduresthat are employed for natural antioxidants removal, liquid–liquidsolvent extraction represents a simple and convenient alternative,and it is widely used in pilot-scale production and in ultimate com-mercial recovery [4]. As solvent for the extraction, ethyl acetate waschosen which is frequently used to extract biophenols from aque-ous matrices such as OMW. In our previous investigation, we havedemonstrated that ethyl acetate exhibits a higher extraction powerrespect to other solvents, such as methyl isobutyl ketone, methylethyl ketone, diethyl ether, even though it is somewhat selectivetowards low (about 180 Da) and medium (about 13 kDa) molecularmass phenolic compounds [15].

The complexity of the OMW ethyl acetate extract inducesresearchers to make any effort in order to quantify and isolatethe extracted phenolic compounds prior to the determination oftheir biological activities. For these purposes, fractionation of theobtained OMW phenolic extract was performed by means of amedium-pressure reversed-phase chromatographic system (seeSection 2). The obtained chromatogram (data not shown) showedtwo separated peaks and the corresponding fractions have beencollected into two different sub-fractions (F1 and F2).

3.2. Identification and quantification of phenolic compounds inthe OMW fractions

The two fractions F1 and F2 resulted from the OMW extract frac-tionation have been characterized by HPLC analysis (Fig. 1). Theobtained chromatograms (A, B, and C) showed several simple peakscorresponding to different phenolic compounds. Identification ofphenolic compounds in the OMW extract fractions (F1 and F2) waspreliminary performed by HPLC-UV, comparing the relative reten-tion times and UV spectra with those of standard solutions. Thesestandards were selected on account of their structural diversityand polarity differences and their documented presence in OMW[15,17]. The presence of formic acid in the mobile phase suppressedthe dissociation of the phenolic compounds and enhanced the sep-aration of the eluates.

Table 4Effects of monomeric and polymeric substances (F1 and F2) and purified hydrox-ytyrosol (F3) on creatinine and urea levels in diabetic rats. Values are given asmean ± S.D. for 8 rats in each group. Values differ significantly at p < 0.05. Statisticalanalysis as in legend of Table 2.

Groups Creatinine Urea

Control 19.8 ± 2.1 0.65 ± 0.06Diab 54.6 ± 8.1a 1.81 ± 0.24a

Diab + Ins 34.2 ± 6.7ab 1.09 ± 0.19ab

Diab + F1 33.1 ± 3.3ab 0.75 ± 0.07bc

Diab + F2 30.6 ± 2.3ab 1.34 ± 0.12abcd

Diab + F3 26.5 ± 0.4.1abcde 0.88 ± 0.07abce

426 K. Hamden et al. / Chemico-Biological Interactions 180 (2009) 421–432

Fig. 4. Effects of monomeric and polymeric substances (F1 and F2) and purified hydroxytyrosol (F3) on SOD, CAT, GPX activities and TBARS levels in liver of diabetic maleWistar rats for 60 days. Values are given as mean± S.D. for 8 rats in each group. Values differ significantly at p < 0.05. Statistical analysis as in legend of Table 2.

The first OMW fraction (F1) was made mainly of somemonomeric biophenols: hydroxytyrosol, tyrosol, caffeic acid, pro-tocatechuic acid and para-coumaric acid. The second OMW fraction(F2) contained principally oligomeric compounds: luteolin 7-O-glucosid, ligstrosid, luteolin, elenolic acid, apigenin, verbascosid,oleuropein, rutin and polymeric substances. Minor unidentifiedcomponents were also present in F1 and F2. To our knowledge, thisis the first time that verbascosid and ligstrosid were identified in theTunisian OMW. This result shows the efficiency of the used SPE sys-tem in the separation of monomers and oligomers from the OMWextract. When there was no available standard, as with verbascosidand elenolic acid, identification was based on the comparison oftheir UV spectra with literature data [16,17].

Identification of phenolic compounds in the OMW extract frac-tions was confirmed by LC–MS, in particular for compounds whichlacked of pure standards. Deprotonated molecular ions representedthe base peak in the negative ion spectra of all the identified com-pounds as shown in Table 1. Protonated molecular ion and sodiumadduct peaks were observed in the positive ion spectra of only fewidentified compounds, namely oleuropein and ligstrosid (data notshown). The structure assignment was based on a systematic searchfor molecular ions using extracted ion mass chromatograms andcomparing with data in the literature [7]. For example, the elutionof hydroxytyrosol at 16.59 min was confirmed by the presence ofvery clean and distinct peaks in both the positive and negative ionmass chromatograms at m/z 155 and 153, respectively, with sodiumadduct at m/z 177 in the positive ion mass chromatogram. Further-

more, the mass spectrum of the peak at 16.59 min (Fig. 2) in thenegative mode was consistent with the assignment of this peak ashydroxytyrosol by comparison with the standard mass spectrum.The presence of ion fragments at m/z 123 and 109 is due to theloss of the (CH2OH) and (CH2–CH2OH) groups, respectively. On theother hand, scanning at m/z 625 and 647 (positive ion mode) andm/z 623 (negative ion mode) produced a single peak in each of themass chromatograms at 33.42 min. These m/z values correspondto the protonated molecular ion, sodium adduct and deprotonatedmolecular ion of verbascosid, respectively.

The identified compounds in the OMW extract were quanti-fied based on standard curves, in the concentration range 50 ppmto 2000 ppm, according to the method reported by Akasbi et al.[26]. Table 1 shows the determined concentrations of biophenols.Hydroxytyrosol is the major phenolic compound in the studiedOMW (1453 mg/l in crude OMW). In order to evaluate the hypo-glycaemic activity of hydroxytyrosol, we were interested in thepurification of this compound from the monomeric fraction (F1)using silica gel-column chromatography and prep. TLC as describedin Section 2. The obtained pure hydroxytyrosol was designed as (F3).

3.3. Blood glucose and hepatic glycogen and OMW derivingpreparations (Fig. 3)

In diabetic rats, the blood glucose level increased by 137%(p < 0.001) compared to the control animals. In F1, F2 and F3 OMWderiving preparations treated rats, a decrease in blood glucose

K. Hamden et al. / Chemico-Biological Interactions 180 (2009) 421–432 427

Fig. 5. Effects of monomeric and polymeric substances (F1 and F2) and purified hydroxytyrosol (F3) on renal SOD, CAT, GPX activities and TBARS of diabetic rats for 60 days.Values are given as mean ± S.D. for 8 rats in each group. Values differ significantly at p < 0.05. Statistical analysis as in legend of Table 2.

by 51, 12 and 55%, respectively after F1, F2 and F3 administra-tion was observed. The hepatic glycogen level is lower in diabeticrats liver compared with that in normal rats, however, after F1,F2 and F3 administration, a significant increase by 67, 36 and222%, respectively was detected compared to untreated diabeticrats.

3.4. Hepatic function, diabetes and OMW preparations

This study showed a decrease in the SOD, CAT and GPX activitiesin hepatic tissues of diabetic rats. In addition, diabetes reduced theHDL-Ch levels in plasma. At the same time an increase in TBARS theindices of hepatic dysfunction parameters (AST, ALT and ALP activi-ties and direct and total bilirubin content) were obtained in plasma.The hepatic toxicity leads to problem in metabolism observed bythe increase in total-cholesterol and triglycerides in plasma andliver (Tables 2 and 3).

In diabetic rats treated with F1, F2 and F3 fractions, a clear pro-tective effect was observed in hepatic function and metabolism. Infact administration of OMW phenolic extract fractions inhibited allchanges caused by alloxan-induced diabetes. This positive effect ofthese fractions was confirmed by histological finding (Figs. 4 and 5).As shown in Fig. 6B, fatty cysts appeared in hepatic tissues ofdiabetic rats indicated by arrow (Fig. 6B). However, F1, F2 andhydroxytyrosol administration to diabetic rats reduced the appear-ance of fat cells in liver (Fig. 6C and D). This ameliorative action ismore pronounced with purified HT (Fig. 6E).

3.5. Renal function, diabetes and OMW preparations

In diabetic rats, a decrease of the SOD, CAT and GPX activities by82, 65 and 78% in kidney at the end of the experiment was showed.Moreover, a significant increase in TBARS in kidney by 115% andcreatinine and urea contents in plasma by 175 and 178%, respec-tively was obtained in alloxan-treated rats (Fig. 5). However, afterOMW phenolic extract fractions administration, a positive actionfrom nephropathy was clearly observed. In fact the renal SOD, CATand GPx antioxidant activities increased after administration ofOMW extract fractions. Furthermore, the OMW biophenol fractionsdecreased lipid peroxidation content and renal dysfunction indicessuch as urea and creatinine levels in plasma (Table 4). Diabeticcontrol rat’s kidney showed fatty infiltration indicated by arrow(Fig. 7B). These changes were reduced in F1, F2 and F3 fractionstreated rats (Fig. 7C, D, and E).

3.6. Pancreas, diabetes and OMW preparations

The pancreas of control rats exhibited normal islets (Fig. 8A). Inalloxan-treated rats pancreas, a "cells degeneration was observed(Fig. 8B). However, in diabetic rats treated with the monomericbiphenols (F1), the polymeric biphenols (F2) and the purifiedhydroxytyrosol (F3) from OMW; a patent protective action of "cellswas recorded mainly in F3 (Fig. 8E). In vitro, Fig. 9 shows that higherglucose concentration causes a decline in insulin secretion after40 min of pancreas incubation. In with HG concentration and F2

428 K. Hamden et al. / Chemico-Biological Interactions 180 (2009) 421–432

Fig. 6. (A) Normal rat liver. (B) Diabetic rat liver: sinusoidal congestion and fatty degeneration in the form of fat lake. (C–E) Diabetic treated with insulin, F1 or F2. (F) Diabeticrats treated with purified hydroxytyrosol: a positive effect was observed (H&E 100#).

fraction not positive action was observed. However, in pancreasincubated at same time with HG concentration and F1 or puri-fied hydroxytyrosol, we investigated in the first time a significantincrease in the insulin secretion until the 40 Min and this positiveaction was more pronounced in purified HT.

4. Discussion

An accumulating body of evidence has been showing that type1 diabetic patients manifest oxidative stress resulting from hyper-glycemia, hyperinsulinemia and insulin resistance. Overwhelmingfree radicals generated due to oxidative stress may develop sev-eral adverse effects commonly seen in diabetes mellitus such ashepathology and nephropathy disorders [2,20]. As a strategy to

counteract the negative effect of oxidative stress, antioxidant-basedtherapy is promising to minimize the complications associatedwith oxidative stress in diabetes mellitus. Recent observationshave shown that many of these complications are diminishedupon supplementation with certain dietary antioxidants such asflavonids and polyphenols [27,28]. The present work is the firstinvestigation dealing with the protective effects of monomericand oligomeric biophenols as well as purified hydroxytyrosol fromOMW on diabetes and their complications in liver, pancreas andkidney functions. Hydroxytyrosol, the main phenolic compound inolives and their by-products, is a phenolic compound which hasbeen shown to possess diverse healing properties for its antioxidantand anti-inflammatory activities [29,30]. This study showed thatpurified hydroxytyrosol (F3) is efficient to prevent alloxan-induced

K. Hamden et al. / Chemico-Biological Interactions 180 (2009) 421–432 429

Fig. 7. (A) Normal rat kidney. (B) Diabetic treated rat kidney: tubular epithelial damage messangial capillary proliferation and fatty infiltration. (C) Diabetic tread with insulin.(D, E and F) Diabetic rats treated with F1 or F2 or F3: (H&E 100#).

hyperglycemia. It was established that administration of hydrox-ytyrosol decreases the glucose level in plasma by 55% comparedto untreated diabetic rats. This hypoglycaemic effect of hydroxyty-rosol in diabetes could be explained by three reasons: (i) protectionof pancreatic "cells from progressive damage enhanced by alloxanand/or the enhancement of the regeneration of theses cells similarto other substances such as oxovanadium [31]; (ii) like tungstate,HT can enhance insulin secretion by its insulinotropic effects: HTinhibits KATP channels and increases the voltage-dependent cal-cium channel which plays a key role in insulin secretion [32]; (iii)hydroxytyrosol increases peripheral uptake of glucose as was estab-lished in a previous study for oleuropein [33]; (vi) hydroxytyrosolwas reported to be effective scavenger of the 1,1-diphenyl-2-picrylhydrazyl (DPPH) radical as was established by Allouche et

al. [29]. This antioxidant activity protects pancreatic "cells fromdamage and death [34] resulting in the increase of insulin secre-tion which decreases glucose level in plasma (iii) HT can increasesome enzymes which catalyze the phosphorylation of glucose suchas hexokinase, pyruvate kinase and decrease enzymes which catal-yse the dephosphorylation of glucose-6-phosphate to free glucosesuch as glucose-6-phosphatase and fructose-1,6-bisphosphatase[3,5,6].

The hypoglycaemic effect of hydroxytyrosol prevents glucoseauto-oxidation reaction and inhibits the formation of advanced gly-cation endproducts (AGEs). Consequently, low rate in free radicalsand more activity in antioxidant enzymes such as SOD, CAT and GPXin liver and kidney tissues were observed. The hypoglycaemic andantioxidant activities of hydroxytyrosol prevent oxidative stress

430 K. Hamden et al. / Chemico-Biological Interactions 180 (2009) 421–432

Fig. 8. Histopathological studies of pancreas in control and experimental groups of rats. Section of pancreas from (A) control rat showing normal architecture; (B) diabeticrat showing vascular degenerative changes in the islets and decrease in the number of "cells; (C) diabetic rats treated with insulin; (D, E and F) diabetic rat treated with F1,F2 or F3.

[35–37] and preserve liver function as was observed by low ratesin AST, ALT, ALP, total and direct bilirubin. In addition, administra-tion of hydroxytyrosol to diabetic rats prevented kidney toxicityobserved by low rates in creatinine and urea in plasma whichincreased in diabetic rats [38–42]. Furthermore, the protection ofliver and kidney function by HT inhibited metabolic disorders indiabetic rat’s evidences by low level in TCh and TG and more HDL-Ch which linked to a lower risk of cardiovascular diseases [43]. Thisresult is in full agreement with the fundings of Gorinstein et al.[43] who reported that olive oil polyphenols decrease plasma LDLlevels and prevent their oxidation in vivo. The mechanism of thishypocholesterolemic action may be due to the inhibition of dietarycholesterol absorption in the intestine or its production by liver [44]or stimulation of the biliary secretion of cholesterol and cholesterol

excretion in the faeces [45]. These results agree with few humanstudies showing that nutritional doses of olive oil decrease TG ina group of mildly dyslipidemic patients [46,47]. In diabetic treatedwith the OMW monomeric fraction (F1), our study showed a goodant-diabetic and antioxidant actions. This would be related to thefree radicals scavenging, anti-inflammatory and antioxidant activ-ities of hydroxytyrosol [48–50]. These activities protect pancreatic"cells from damage and death induced by oxidative stress as shownin Fig. 9.

The hypoglycaemic and antioxidant actions of F1 prevented thedifferent types of oxidative damage and metabolic disorders asso-ciated with diabetes in liver and kidney.

On the other hand, our study showed that the polymeric frac-tion of OMW (F2) prevent diabetes and their toxicity in liver, kidney

K. Hamden et al. / Chemico-Biological Interactions 180 (2009) 421–432 431

Fig. 9. Effect of glucose, F1, F2 and purified hydroxytyrosol on the time course ofglucose-induced insulin decline.

and pancreas. This effect principally resulted from the presenceof hydroxytyrosol derivatives (oligomeric polyphenols) which arecomposed of hydroxytyrosol units attached to other compoundsvia ester or/and glucosidic linkages such as verbascosid, ligstrosid,oleuropein and salidroside [4]. In rat as in human these oligomerswould be hydrolyzed by enzymes such as "-glucosidase, lipase,lipoxygenase and phenoloxidase to release hydroxytyrosol andother metabolites such as elenolic acid. This statement can be sup-ported by the findings of Feki et al. who reported that during OMWstorage the hydroxytyrosol derivatives are hydrolyzed by enzymesnaturally occurring in OMW [51].

5. Conclusion

This study demonstrated the beneficial effect of hydroxytyrosolas an effective hypoglycemic and antioxidant agent in alleviatingoxidative stress and free radicals as well as in enhancing enzymaticdefenses in diabetic rats. The use of hydroxytyrosol may be of pro-phylactic value in reducing the complications usually resulting fromoxidative stress in diabetes mellitus.

Conflict of interest

No conflict in this work.

References

[1] C. Henquin, A. Debuyser, G. Drews, T.D. Plant, Regulation of K+ permeabilityand membrane potential in insulin-secretory cells, in: P.R. Flatt (Ed.), NutrientRegulation of Insulin Secretion, Porland, London, 1992, pp. 173–192.

[2] K. Hamden, S. Carreau, S. Lajmi, D. Aloulou, D. kchaou, A Elfeki, Protective effectof 17"-estradiol on hyperglycemia, stress oxidant, liver dysfunction and histo-logical changes induced by alloxan in male rat pancreas and liver, Steroids 94(2008) 495–501.

[3] P. Palsamy, S. Subramanian, Modulatory effects of resveratrol on attenuatingthe key enzymes activities of carbohydrate metabolism in streptozotocin-nicotinamide induced diabetic, Chem. Biol. Interact. 179 (2009) 356–362.

[4] H.K. Obied, M.S. Allen, D.R. Bedgood, P.D. Prenzler, K. Robards, R. Stockmann:,Bioactivity and analysis of biophenols recovered from olive mill waste, J. Agric.Food Chem. 53 (2005) 823–837.

[5] T. Ferre, A. Pujol, E. Riu, F. Bosch, A. Valera, Correction of diabetic alterations byglucokinase, Proc. Natl. Acad. Sci. U.S.A. 93 (1996) 7225–7230.

[6] J.Y. Chou, D. Matern, B.C. Mansfield, Y.T. Chen, Type I glycogen storage diseases:disorders of the glucose-6-phosphatase complex, Curr. Mol. Med. 2 (2002)121–143.

[7] H.K. Obeid, M.S. Allen, D.R. Bedgood, P.D. Prenzied, K. Robards, R. Stockmann,Bioactivity and analysis of biophenols recovered from olive mill waste, J. Agric.Food Chem. 53 (2005) 823–837.

[8] E. Casalino, G. Calzaretti, C. Sblano, V. Landriscina, M.F. Tecce, C. Landriscina,Antioxidant effect of hydroxytyrosol (DPE) and Mn2+ in liver of cadmium-intoxicated rats, Comp. Biochem. Phys. C 133 (2002) 625–632.

[9] J.I. Dudley, I. Lekli, S. Mukherjee, M. Das, A.A. Bertelli, D.K. Das, Does white winequalify for French paradox? Comparison of the cardioprotective effects of red

and white wines and their constituents: resveratrol, tyrosol, and hydroxyty-rosol, J. Agric. Food Chem. 56 (2008) 9362–9373.

[10] S. D’Angelo, D. Ingrosso, V. Migliardi, A. Sorrentino, G. Donnarumma, A. Baroni,L. Masella, M.A. Tufano, M. Zappia, P. Galletti, Hydroxytyrosol, a natural antiox-idant from olive oil, prevents protein damage induced by long-wave ultravioletradiation in melanoma cells, Free Radical Biol. Med. 38 (2005) 908–919.

[11] X. Zhang, L. Jiang, C. Geng, H. Yoshimura, L. Zhong, Inhibition of acrylamide geno-toxicity in human liver-derived HepG2 cells by the antioxidant hydroxytyrosol,Chem. Biol. Interact. 176 (2008) 173–178.

[12] M. Deiana, A. Incani, A. Rosa, G. Corona, A. Atzeri, D. Loru, M.P. Melis, M.A. Dessì,Protective effect of hydroxytyrosol and its metabolite homovanillic alcohol onH2O2 induced lipid peroxidation in renal tubular epithelial cells, Food Chem.Toxicol. 46 (2008) 2984–2990.

[13] F. Visioli, G. Bellomo, C. Galli, Free radical-scavenging properties of olive oilpolyphenols, Biochem. Biophys. Res. Commun. 247 (1998) 60–64.

[14] M.I. Covas, K. De la Torre, M. Farre-Albaladejo, et al., Postprandial LDL phenoliccontent and LDL oxidation are modulated by olive oil phenolic compounds inhumans, Free Radical Biol. Med. 40 (2006) 608–616.

[15] N. Allouche, I. Fki, S. Sayadi, Toward a high yield recovery of antioxidants andpurified hydroxytyrosol from olive mill wastewaters, J. Agric. Food Chem. 52(2004) 267–273.

[16] P. Rovellini, N. Cortesi, Liquid chromatography–mass spectrometry in the studyof oleuropein and ligstroside aglycons in virgin olive oil: aldehydic, dialdehy-dic forms and their oxidized products, Rivista Italiana Sostanze Grasse, LXXIL(2004) 1–14.

[17] H.K. Obied, M.S. Allen, D.R. Bedgood, P.D. Prenzler, K. Robards, Investigation ofAustralian olive mill waste for recovery of biophenols, J. Agric. Food Chem. 53(2005) 9911–9920.

[18] M.H. Giroux, B. Portha, M. Kergoat, D. Bailbe, L. Picon, Glucose insensitivityand amino-acid hypersensitivity of insulin release in rats with non-insulin-dependent diabetes: a study with the perfused pancreas, Diabetes 32 (1983)445–451.

[19] J.M.A. Hannan, L. Marenah, L. Ali, B. Rokeya, P.R. Flatt, Y.H. Abdel-Wahab, Insulinsecretory actions of extracts of Asparagus racemosus root in perfused pancreas,isolated islets and clonal pancreatic "-cells, J. Endocrinol. 192 (2007) 159–168.

[20] R. Ananthan, M. Latha, K.M. Ramkumar, L. Pari, C. Baskar, M.V. Narmatha Bai.,Modulatory effects of gymnema montanum leaf extract on alloxan-inducedoxidative stress in Wistar rats, Nutrition 20 (2004) 280–285.

[21] J.A. Buege, S.D. Aust, Microsomal lipid peroxidation, Methods Enzymol. 105(1984) 302–310.

[22] S. Marklund, G. Marklund, Involvement of the superoxide anion radical in theautoxidation of pyrogallol and convenient assay for superoxide dismutase, Eur.J. Biochem. 47 (1975) 469–474.

[23] D.E. Pagila, W.N. Valentine, Studies on the quantitative and qualitative char-acterization of erythrocyte glutathione peroxidase, J. Lab Clin. Med. 70 (1967)158–169.

[24] H. Aebi, Catalase in vitro, Methods Enzymol. 105 (1984) 121–126.[25] O.H. Lowry, N.J. Rosebrough, A.L. Farr, R.J. Randall, Protein measurement with

Folin phenol reagent, J. Biol. Chem. 193 (1951) 265–275.[26] M. Akasbi, D.W. Shoeman, A.S. Csallany, High-performance liquid chromatog-

raphy of selected phenolic compounds in olive oils, J. Am. Oil Chem. Soc. 70(1993) 367–370.

[27] A. Kishore, G.K. Nampurath, S.P. Mathew, R.T. Zachariah, B.K. Potu, M.S. Rao, M.Valiathan, M.R. Chamallamudi, Antidiabetic effect through islet cell protectionin streptozotocin diabetes: a preliminary assessment of two thiazolidin-4-onesin Swiss albino mice, Chem. Biol. Interact. 177 (2009) 242–246.

[28] K. Hamden, S. Carreau, K. Jamoussi, F. Ayadi, S. Miladi, S. Lajmi, D. Aloulou,A. Elfeki, Title:1#,25dihydroxyvitaminD3: therapeutic and preventive effectsagainst oxidative stress and hepatic, pancreatic and renal injury in diabetic rats,J. Nutr. Sci. Vitam., in press.

[29] N. Allouche, I. Fki, S. Sayadi, The use of olive mill wastewaters as a cheap sourceof natural antioxidants: biological activities of a continuous solvent extract andpurified hydroxytyrosol, Polyphénols Actualités 23 (2003) 16–19.

[30] J. Zhang, L. Cao, Zhong, Hydroxytyrosol inhibits pro-inflammatory cytokines,iNOS, and COX-2 expression in human monocytic cells, Naunyn SchmiedebergsArch. Pharmacol. 379 (2009) 581–586.

[31] B. Ramachandran, K. Ravi, V. Narayanan, M. Kandaswamy, S. Subramanian,Protective effect of macrocyclic binuclear oxovanadium complex on oxidativestress in pancreas of streptozotocin induced diabetic rats, Chem. Biol. Interact.149 (2004) 9–21.

[32] R.A. Silvestre, E.M. Egido, R. Hernandez, J. Marco, Tungstate stimulates insulinrelease and inhibits somatostatin output in the perfused rat pancreas, Eur. J.Pharmacol. 519 (2005) 127–134.

[33] M. Gonzalez, A. Zarzuelo, M.J. Gamez, M.P. Utrilla, J. Jimenez, I. Osuna, Hypo-glycemic activity of olive leaf, Planta Med. 58 (1992) 513–515.

[34] S.N. Nivitabishekam, M. Asad, V.S. Prasad, Pharmacodynamic interaction ofMomordica charantia with rosiglitazone in rats, Chem. Biol. Interact. 177 (2009)247–253.

[35] A. Siritantikorn, K. Johansson, K. Ahlen, R. Rinaldi, T. Suthiphongchai, P. Wilairat,R. Morgenstern, Protection of cells from oxidative stress by microsomal glu-tathione transferase 1, Biochem. Biophys. Res. Commun. 355 (2007) 592–596.

[36] E. Mosialou, A.E.P. Adang, R. Morgenstem, Evidence that rat liver microsomalglutathione transferase is responsible for glutathione-dependent protectionagainst lipid peroxidation, Free Radical Biol. Med. 9 (1990) 133.

[37] E. Mosialou, G. Ekström, A.E.P. Adang, R. Morgenstern, Evidence that rat livermicrosomal glutathione transferase is responsible for glutathione-dependent

432 K. Hamden et al. / Chemico-Biological Interactions 180 (2009) 421–432

protection against lipid peroxidation, Biochem. Pharmacol. 45 (8) (1993)1645–1651.

[38] J. Eliza, P. Daisy, S. Ignacimuthu, V. Duraipandiyan, Normo-glycemic andhypolipidemic effect of costunolide isolated from Costus speciosus (Koen ex.Retz.)Sm. in streptozotocin-induced diabetic rats. Chem. Biol. Interact. 179(2009) 329–334.

[39] H.S. Moon, H.G. Lee, J.H. Seo, C.S. Chung, T.G. Kim, Y.J. Choi, C.S. Cho, Antiobe-sity effect of PEGylated conjugated linoleic acid on high-fat diet-induced obeseC57BL/6J (ob/ob) mice: attenuation of insulin resistance and enhancement ofantioxidant defenses, J. Nutr. Biochem. 20 (2009) 187–194.

[40] T.J. Lyons, Oxidized low density lipoproteins: a role in the pathogenesis ofatherosclerosis in diabetes? Diabet. Med. 8 (1991) 411–419.

[41] A. Kiersztan, A. Baranska, M. Hapka, M. Lebiedzinsk, K. Winiarska, M. Dudziak, J.Bryl, Differential action of methylselenocysteine in control and alloxan-diabeticrabbits, Chem. Biol. Interact. 177 (2009) 161–171.

[42] Z. Liu, J. Li, Z. Zeng, M. Liu, M. Wang, The antidiabetic effects of cysteinylmetformin, a newly synthesized agent, in alloxan- and streptozocin-induceddiabetic rats, Chem. Biol. Interact. 173 (2008) 68–75.

[43] S. Gorinstein, H. Leontowicz, A. Lojek, M. Leontowicz, M. Ciz, R. Krzeminski,M. Gralak, J. Czerwinski, Z. Jastrzebski, S. Trakhtenberg, N. Grigelmo-Miguel, R.Soliva-Fortuny, Olive oils improve lipid metabolism and increase antioxidantpotential in rats fed diets containing cholesterol, J. Agric. Food Chem. 50 (2002)6102–6108.

[44] R. Krzeminski, S. Gorinstein, H. Leontowicz, M. Leontowicz, M. Gralak, J. Czer-winski, A. Lojek, M. Ciz, O. Martin-Belloso, N. Gligelmo-Miguel, S. Trakhtenberg,

Effect of different olive oils on bile excretion in rats fed cholesterol-containingand cholesterol-free diets, J. Agric. Food Chem. 51 (2003) 5774–5779.

[45] K. Prasad, J. Kalra, Oxygen free radicals and hypercholesterolemic atherosclero-sis, Am. Heart J. 125 (1993) 958–971.

[46] F. Visioli, D. Caruso, S. Grande, et al., Virgin Olive Oil Study (VOLOS): vasopro-tective potential of extra virgin olive oil in mildly dyslipidemic patients, Eur. J.Nutr. 6 (2004) 1–7.

[47] T. Weinbrenner, M. Fito, R. de la Torre, et al., Olive oils high in phenolic com-pounds modulate oxidative/antioxidative status in men, J. Nutr. 134 (2004)2314–2321.

[48] R. Masella, R. Vari, M. D’Archivio, R. Di Benedetto, P. Matarrese, W. Mal-orni, B. Scazzocchio, C. Giovannini, Extra virgin olive oil biophenols inhibitcell-mediated oxidation of LDL by increasing the mRNA transcription ofglutathione-related enzymes, J. Nutr. 134 (2004) 785–791.

[49] C. Giovannini, E. Straface, D. Modesti, E. Coni, A. Cantafora, M. De Vin-cenzi, W. Malorni, R. Masella, Tyrosol, the major olive oil biophenol, protectsagainst oxidized-LDL-induced injury in Caco-2 Cells, J. Nutr. 129 (1999)1269–1277.

[50] J.L. Quiles, A.J. Farquharson, D.K. Simpson, I. Grant, K.W.J. Wahle, Olive oil phe-nolics: effects on DNA oxidation and redox enzyme mRNA in prostate cells, Br.J. Nutr. 88 (2002) 225–234.

[51] M. Feki, N. Allouche, M. Bouaziz, A. Guargoubi, S. Sayadi, Effect of storage of olivemill waste waters on hydroxytyrosol concentration, Eur. J. Lipid Sci. Technol. 108(2006) 1021–1027.