04-Gomez:gomez 9-06-2011 16:39 Pagina 22

L AVORO ORIGINALE

A.M. G ÓMEZ -C ARAVACA 1, 2 , Qualitative phenolic profile (HPLC- L. C ERRETANI 1, DAD-MS) from mill waste A. S EGURA -C ARRETERO 2, waters at different states of storage and A. F ERNÁNDEZ -G UTIÉRREZ 2, G. L ERCKER 1, evaluation of hydrolysis process as a T. G ALLINA TOSCHI 1 pretreatment to recover their antioxidants

PROGRESS IN NUTRITION Summary VOL . 13 , N. 1, 22-30, 2011 Olive oil mill waste water (OOMWW) represents a waste that it is gene - rated during the extraction process of olive oil from olives and it is an im - TITOLO portant environmental problem because of the huge volume generated in Profilo fenolico qualitativo the mills. Nevertheless, it is highlighted in literature the high concentra - (HPLC-DAD-MS) di acque tion of antioxidants, especially phenolic compounds, in the OOMWW. reflue di vegetazione Principally, the qualitative phenolic profiles of OOMWW differ depen - dell’industria olearia a diversi ding on the technological process of production of olive oil and the time stadi di conservazione e of storage. The aim of this work was to compare qualitatively the pheno - valutazione dell’idrolisi quale lic profile of OOMWW by HPLC-DAD-MS. Glucosilated secoiridoids pretrattamento per il recupero as and ligstroside, their aglycon derivatives and simple phenols degli antiossidanti were the principal compounds found in these matrixes. Furthermore, dif - ferent hydrolysis reactions have been studied for the phenolic fraction in the oleuropein standard as well as in the OOMWW samples. These steps KEY WORDS Olive, olive oil mill waste waters, have been evaluated as pretreatment for the selection and a possible sub - antioxidants, phenols sequent recovery of most interesting phenolic compounds as hydroxytyro - sol that has been proved to possess an important antioxidant activity

PAROLE CHIAVE Oliva, acque reflue, antiossidanti, Riassunto fenoli Le acque reflue di vegetazione (AV) dell’industria olearia rappresentano un rifiuto generato dal processo di estrazione dell’olio dalle olive e pertan - to sono un problema ambientale per l’elevato volume generato presso gli impianti di produzione. Tuttavia, è nota dalla letteratura, l’elevata concen - 1 Department of Analytical trazione nelle AV di composti antiossidanti a struttura fenolica. I profili Chemistry, Faculty of Sciences, fenolici qualitativi delle AV differiscono principalmente in funzione delle University of Granada, condizioni di produzione e delle modalità e tempi di conservazione. Que - C/Fuentenueva s/n, E-18071 sto lavoro ha l’obiettivo di confrontare qualitativamente i profili fenolici Granada, Spain delle AV mediante analisi HPLC-DAD-MS. Nelle acque reflue di vege - 2 Department of Food Science, tazione appena prodotte sono state ritrovate principalmente secoiridoidi University of Bologna, Piazza glicosilati (oleuropeina e ligstroside) nonché le loro forme derivate aglico - Goidanich 60, I-47521 Cesena (FC), Italy niche e basse concentrazioni di fenoli semplici. Allo stesso tempo sullo standard puro di oleuropeina e sugli stessi campioni di AV sono state in - Indirizzo per la corrispondenza: dagate diverse modalità di idrolisi della frazione fenolica. Tale operazione Dr. A.M. Gómez-Caravaca è stata valutata come pretrattamento per la selezione e l’eventuale succes - Tel. +390547338117 E-mail: [email protected] or sivo recupero dei composti fenolici di maggiore interesse come [email protected] l’idrossitirosolo al quale è riconosciuta una elevata attività antiossidante.

22 04-Gomez:gomez 9-06-2011 16:39 Pagina 23

PROGRESS IN NUTRITION 1/2011

Introduction tioxidant activity makes them in - hydrolysis reactions were studied teresting for the food, pharmaceu - using oleuropein as standard and Olive oil mill waste water tical and cosmetic industry. Be - better results were apply to the (OOMWW) represents the main cause of that, to reutilize them, OOMWW phenolic samples. environmental problem in the oli - their extraction from OOMWW The aim of this treatment was to ve oil production process. 100 kg is a very important objective that evaluate the hydrolysis and to use of olives are necessary to produce also helps to diminish the volume it as a pretreatment for the selec - just 10-20 kg of olive oil, but it re - of olive oil industry by-products. tion and a possible subsequent re - sults in a huge amount of Many phenolic compounds have covery of most interesting pheno - OOMWW (40-100 kg been identified in OOMWW, al - lic compounds as OOMWW) (1) that depends on though the phenolic fraction is very (HYTY) that has been proved to the olive cultivar, ripening index, different from that of olive fruit. possess an important antioxidant time of storage and quantity of OOMWW presents a high con - activity. water added during the olive oil centration of secoiridoids derivati - extraction process. ves, such as hydroxytyrosol, and the The treatment of OOMWW is dialdehydic form of decarboxy - Material and methods extremely difficult due to its large methyl oleuropein aglycone (1). volume and the high concentra - The profile of OOMWW is cha - Samples tion of organic matter (BOD va - racterized by a great complexity (9). ries between 15000 and 50000 A wide variety of methods have Three different OOMWWs that mg/L and COD can reach values been suggested for the treatment vary on the state of storage and the of around 250 g/L). Its principal of OOMWW, such as compo - ripeness of the olives were studied: components are polysaccharides, sting (10), anaerobic digestion OOMWW just produced from sugars, polyphenols, polyalcohols, (11), aerobic treatment (12), mi - fresh olives (AV1), OOMWW proteins,organic acids, and oil (2). xing with municipal wastewater just produced from very ripe olives Moreover, OOMWW contains (13), direct land application (irri - (AV2) and OOMWW from very considerable amounts of suspen - gation) (14), chemical oxidation in ripe olives and storage during three ded solids that may reach up to combination with biological treat - months (AV3). 190 g/L. Despite that, one of the ment (15) or adsorption (2), and major factors of the environmental even the utilization of fungi spe - Extraction of polar phenolic fraction problems caused by the cies (16). However, these metho - OOMWW is the high concentra - dologies had given rise to decom - Phenolic compounds were extrac - tion of polyphenols, because they position or destruction of the phe - ted from OOMWW by a liquid- present phytotoxicity (3, 4), toxi - nolic compounds contained on it liquid extraction method. 10 mL city against aquatic organisms (5), and not to their exploitation. of sample were centrifuged at or suppression of soil microorga - In this work the phenolic profile of 3500 rpm during 10 min. After nisms (6) and are difficult to de - three different OOMWWs concer - that, they were washed twice with compose (7,8). Although polyphe - ning their state of storage were stu - 15 mL of hexane in order to re - nols constitute an important envi - died by HPLC-DAD-MS (TOF, move lipids. The extraction was ronmental problem, their high an - time of flight). After that, different carried out with 20 mL of ethyl

23 04-Gomez:gomez 9-06-2011 16:39 Pagina 24

VOLUME 13

acetate for three times. Finally, the HPLC-DAD-MS(TOF) equip - USA). Parameters for analysis we - ethyl acetate fraction was evapora - ped with a reverse phase C18 Lu - re set using negative ion mode ted under vacuum and then the naTM column according to Ro - with spectra acquired over a mass dry residue was redissolved in 1 tondi et al. (17). Phenolic com - range from m/z 50–1000. The op -

mL MeOH/H 2O (50/50). pounds were tentatively identified timum values of the ESI-MS based on their UV-vis and mass parameters were: capillary voltage, Acid and basic hydrolysis spectra obtained by HPLC- +4.5 kV; drying gas temperature, DAD-MS(TOF). 190°C; drying gas flow, 9 L/min; Four different kind of hydrolyses The HPLC system was coupled to and nebulizing gas pressure, 2 bar. were studied: strong and weak acid a Bruker Daltonik microTOF External calibration was perfor - hydrolyses, basic hydrolysis and mass spectrometer (Bruker Dalto - med using a sodium formiate so - enzymatic hydrolysis. nik, Bremen, Germany) using an lution injected at the beginning of Three modalities of strong acid orthogonal electrospray interface the run and all the spectra were hydrolyses were performed: 0.5 M (ESI) (model G1607A from Agi - calibrated prior to the polyphenol

HCl/H 2SO4 2 h stirring at 80ºC; lent Technologies, Palo Alto, CA, identification.

1 M HCl/H 2SO4 2 h stirring and 1 M HCl/H 2SO4 22 min ultra - Figura 1 - Decrease of oleuropein after hydrolysis Vs time of hydrolysis. sounds. a) hydrolysis with HCl, b) hydrolysis with citric acid. The weak acid hydrolysis was car - ried out using a saturated solution

of citric acid in MeOH/H 2O du - ring 6 h at 80ºC. The basic hydrolysis was carried out in the following way: 2 M NaOH/KOH during 4h stirring, after that it was acidified with HCl until pH=1. Finally, the sam - ple was extracted with diethyl ether/ethyl acetate. To perform the enzymatic hydroly - sis 1 mg oleuropein (Ol) in sodium acetate buffer 0.1 M, pH=5.5 and 5 mg β-glucosidase (from almonds) ≥ 6 units/mg were heated at 37ºC in a water bath during 1 h.

Chromatographic procedure

Determination of the phenolic fraction was performed using an

24 04-Gomez:gomez 9-06-2011 16:39 Pagina 25

PROGRESS IN NUTRITION 1/2011

The accurate mass data for the Figura 2 - Changes in the concentration of oleuropein and hydroxytyrosol du - molecular ions were processed ring the HCl hydrolysis at different times. using the software Data Analysis 3.4 (Bruker Daltonik), which pro - vided a list of possible elemental formulas by using the Generate Molecular Formula TM editor.

Results and discussion

Hydrolisis

After carrying out the different hydrolyses, optimum results were obtained with 1M HCl and satu - rated citric acid solution both in a reflux condenser at 80 ºC. Both hydrolyses were proved at different times from 30 min until 8h. Optimum times of hydrolysis were individualized for each of them at 4h for HCl and 6h for ci - tric acid (Figure 1). Phenolic compounds in OOMWW have been characterized by Figure 2 shows how HCl hydroly - before and after hydrolysis HPLC-DAD-MS in order to see sis affects the extract. It can be ob - the differences among the pheno - served that oleuropein (Ol) de - OOMWWs have a high content lic profiles depending on the time creases with time, while the con - of phenolic compounds that come of storage of the OOMWW and centration of HYTY increases. from olives after olive oil produc - the freshness of the olives. Fur - This fact is due to the fragmenta - tion. These phenolics are princi - thermore, the hydrolysis of the tion of Ol after the hydrolysis. pally secoiridoids as oleuropein, samples have been carried out as a On the other hand, the citric acid ligstroside and its derivatives. Ho - pretreatment of the extracts to ob - hydrolysis also leads to the decrea - wever, the enzymes that exist in tain simple phenols as hydroxyty - se of Ol. However, this kind of olives, the milling of olives and rosol, compound that has been hydrolysis is softer than the one the storage lead to a hydrolysis of proved to possess a high antioxi - with HCl and the fragmentation phenolic compounds changing the dant activity. This process would gives caused to an intermediate initial composition of fresh be a previous step to isolate these fragment as oleuropein aglycone OOMWWs. simple compounds to further use (Figure 3). AV1, AV2 and AV3 phenolic ex - them as antioxidants in the food tracts from different OOMWWs or pharmaceutical industry.

25 04-Gomez:gomez 9-06-2011 16:39 Pagina 26

VOLUME 13

Figura 3 - Changes in the concentration of oleuropein and oleuropein aglyco - of the olives. Olives are not fresh ne during the citric acid hydrolysis at different times. and the storage of them before milling causes the beginning of natural hydrolysis in the olives. Because of that, after hydrolysis the phenolic profile of AV2 is very similar to the one before hydroly - sis (Table 2). Finally, OOMWWs from very ripe olives and storage during a long ti - me (AV3) as AV2 do not possess Ol; but they are characterised by the presence of simple phenols as HYTY and (TY), and oxi - dised forms of secoiridoids as oxidi - sed form of decarboxymethyl oleu - ropein aglycone (OxDOA), oxidi - sed form of elenolic acid (OxEA) and decarboxymethyl oleuropein aglycone/oxidised form of decar - boxymethyl ligstroside aglycone (DOA/OxDLA). As well as AV2, the acid hydrolysis do not affect particularly to the phenolic profile of AV3. This kind of OOMWW also comes from very ripe olives and besides it has been storage for a long time, which is the reason be - cause it does not contain Ol but Tables 1, 2 and 3 show the pheno - seen the formation of HYTY and just oxidised form of secoiridoids lic profile in the different TY (Table 1). and some simple phenols. OOMWW studied and the chan - The OOMWWs just produced ges that the hydrolyses produce on from very ripe olives (AV2) do not Conclusions them. have Ol, not even after produc - AV1 (OOMWW just produced tion; while it can be noticed the Different hydrolyses have been from fresh olives) is characterised presence of simple phenols as carried out and, after optimiza - for the presence of Ol and the ab - HYTY and also elenolic acid tion, acid hydrolyses with HCl sence of simple phenols. Never - (EA) that come from the rupture and citric acid show a good effi - theless, after the hydrolysis Ol of secoiridoids. This fact can be ciency (>95%) after 4 and 6 h re - disappears and mainly, it can be explained because of the ripeness spectively.

26 04-Gomez:gomez 9-06-2011 16:39 Pagina 27

PROGRESS IN NUTRITION 1/2011

Tabella 1 - Phenolic compounds found in OOMWW AV1 before and after hydrolysis.

Possible compound Molecular formula m/z calculated AV1 AV1HCl AV1CIT Vanillin C8H7O3 151.0401 X- - Vanillic acid C8H8O4 167.0328 XX X Hydroxy DEA C9H14O5 201.0788 XX X HYTY C8H10O3 153.0557 -X X OxDEA C9H12O5 199.0589 XX X TY C8H10O2 137.0608 -X - Caffeic acid C9H8O4 179.0331 XX X DEA C9H12O4 183.0663 XX X Syringic acid C9H10O5 197.045 -X X p-coumaric acid C9H8O3 163.0376 -- - Syringaresinol C22H26O8 417.1556 XX - OxDOA C17H20O7 335.1136 -- - EA C11H14O6 241.0722 XX X Luteolin glucoside C21H20O11 447.0941 -- - p-coumaroyl-6’-secologanoside C25H28O13 535.142 -- - OxEA C12H18O6 257.1019 -- - 10-H-Ol Agl C19H22O9 393.1178 -- - Oleuropein C25H32O13 539.178 X- - Ligstroside C25H32O12 523.1833 -- - Luteolin C15H10O6 285.0407 X- - DOA/OxDLA C17H20O6 319.1187 -- - Ac Pin C22H24O8 415.139 X- - Apigenin C15H10O5 269.0432 X- - DLA C19H20O5 303.1238 -- - Hydroxy DEA, hydroxy decarboxymethyl elenolic acid; HYTY, hydroxytyrosol; OxDEA, oxidised decarboxymethyl elenolic acid; TY, tyrosol; DEA, decarboxymethyl elenolic acid; OxDOA, oxidised decarboxymethyl oleuropein aglycone; EA, elenolic acid; OxEA, oxidised elenolic acid, DOA/OxDLA, decarboxymethyl oleuropein aglycone/oxidised decarboxymethyl ligstroside aglycone. AV1, OOMWW just produced from fresh olives; AV1HCl, OOMWW hydrolysed with HCl; AV1CIT, OOMWW hydrolysed with citric acid.

27 04-Gomez:gomez 9-06-2011 16:39 Pagina 28

VOLUME 13

Tabella 2 - Phenolic compounds found in OOMWW AV2 before and after hydrolysis

Possible compound Molecular formula m/z calculated AV2 AV2HCl AV2CIT Vanillin C8H7O3 151.0401 X- - Vanillic acid C8H8O4 167.0328 XX X Hydroxy DEA C9H14O5 201.0788 XX X HYTY C8H10O3 153.0557 XX X OxDEA C9H12O5 199.0589 XX X TY C8H10O2 137.0608 -X - Caffeic acid C9H8O4 179.0331 XX X DEA C9H12O4 183.0663 XX X Syringic acid C9H10O5 197.045 -X - p-coumaric acid C9H8O3 163.0376 XX X Syringaresinol C22H26O8 417.1556 -- - OxDOA C17H20O7 335.1136 -- - EA C11H14O6 241.0722 XX X OxEA C12H18O6 257.1019 -- - Luteolin glucoside C21H20O11 447.0941 X- X p-coumaroyl-6’-secologanoside C25H28O13 535.142 XX X 10-H-Ol Agl C19H22O9 393.1178 X- - Oleuropein C25H32O13 539.178 -- - Ligstroside C25H32O12 523.1833 X- - Luteolin C15H10O6 285.0407 XX X DOA/OxDLA C17H20O6 319.1187 -- - Ac Pin C22H24O8 415.139 X- X Apigenin C15H10O5 269.0432 X- X DLA C19H20O5 303.1238 -- - Abbreviations see Table 1. AV2, OOMWW just produced from very ripe olives; AV2HCl, OOMWW hydrolysed with HCl; AV2CIT, OOMWW hydrolysed with ci - tric acid.

28 04-Gomez:gomez 9-06-2011 16:39 Pagina 29

PROGRESS IN NUTRITION 1/2011

Tabella 2 - Phenolic compounds found in OOMWW AV3 before and after hydrolysis

Possible compound Molecular formula m/z calculated AV3 AV3HCl AV3CIT Vanillin C8H7O3 151.0401 X- - Vanillic acid C8H8O4 167.0328 XX - Hydroxy DEA C9H14O5 201.0788 XX - HYTY C8H10O3 153.0557 XX X OxDEA C9H12O5 199.0589 -- - TY C8H10O2 137.0608 XX X Caffeic acid C9H8O4 179.0331 -X - DEA C9H12O4 183.0663 XX X Syringic acid C9H10O5 197.045 XX X p-coumaric acid C9H8O3 163.0376 XX X Syringaresinol C22H26O8 417.1556 -- - OxDOA C17H20O7 335.1136 X- - EA C11H14O6 241.0722 -- - Luteolin glucoside C21H20O11 447.0941 -- - p-coumaroyl-6’-secologanoside C25H28O13 535.142 -- - OxEA C12H18O6 257.1019 XX - 10-H-Ol Agl C19H22O9 393.1178 -- - Oleuropein C25H32O13 539.178 -- - Ligstroside C25H32O12 523.1833 -- - Luteolin C15H10O6 285.0407 -X - DOA/OxDLA C17H20O6 319.1187 XX - Ac Pin C22H24O8 415.139 -- - Apigenin C15H10O5 269.0432 -- - DLA C19H20O5 303.1238 XX - Abbreviations see Table 1. AV3, OOMWW from very ripe olives and storage during three months; AV3HCl, OOMWW hydrolysed with HCl; AV3CIT, OOMWW hydrolysed with citric acid.

29 04-Gomez:gomez 9-06-2011 16:39 Pagina 30

VOLUME 13

The state of ripeness of the olives References dem mass spectrometry of biopheno - lic compounds in olives and vegeta - and the time of storage of 1. De Marco E, Savarese M, Paduano A, tion waters, Part I. J Sep Sci 2003; 26: OOMWW influence the phenolic Sacchi A. Characterization and fractio - 409-16. profile, specially affects to secoiri - nation of phenolic compounds entracte 10. Papadimitriou EK, Chatjipavlidis I, doids. Ol disappears if from olive oil mill wastewaters. Food Balis C. Application of composting to olive mill wastewater treatment. En - OOMWW are not produced Chem 2007; 104: 858-67. 2. Zouari N. Decolorization of olive oil Viron Technol 1997; 18: 101-7. from fresh olives. mill effluent by physical and chemical 11. Sabbah I, Yazbak A, Haj J, Saliba A, Furthermore, the storage for seve - treatment prior to anaerobic digestion. J Basheer S. Biomass selection for opti - ral months of OOMWW causes Chem Technol Biotechnol 1998; 73: mal anaerobic treatment of olive mill wastewater. EnViron Technol 2005; the formation of oxidised forms of 297-303. 3. Aliotta G, Fiorentino A, Oliva A, Te - 26: 47-54. secoiridoids. mussi F. Olive oil mill wastewater: Iso - 12. Hoyos SEG, Nieto LM, Rubio FC, Because of that, it will be very im - lation of polyphenols and their phytoto - Cormenzana AR. Kinetics of aerobic portant to assess the state of the xicity in vitro. Allelopathy J 2002; 9: 9- treatment of olive-mill wastewater (OMW) with Aspergillus terreus. olives and the time of storage of 17. 4. Capasso R, Cristinzio G, Evidente A, Process Biochem 2002; 37: 1169-76. OOMWW in order to use them Scognamiglio F. Isolation, spectroscopy 13. Croce F, Poulsom S, Hendricks D. to optimize a methodology to se - and selective phytotoxic effects of poly - Combined treatment of olive mill ef - parate the phenolic compounds phenols from vegetable waste waters. fluent and municipal waste-water in a small tourist community. Water Sci present in these olives by-pro - Phytochem 1992; 31: 4125-28. 5. Fiorentino A, Gentili A, Isidori M, et Technol 1994; 29: 105-10. ducts. al. Environmental effects caused by oli - 14. Lopez R, Martinez-Bordiu A, de Lo - Finally, the performances of the ve mill wastewaters: Toxicity compari - me ED, Cabrera F, Sanchez MC. Soil phenolic fraction isolated will have son of low-molecularweight phenol properties after application of olive oil mill wastewater. Fresenius EnViron to be studied in order to use them components. J Agric Food Chem 2003; 51: 1005-9. Bull 1996; 5: 49-54. as food antioxidants. 6. Kotsou M, Mari I, Lasaridi K, Chatzi - 15. Fiorentino A, Gentili A, Isidori M, pavlidis I, Balis C, Kyriacou A. The ef - Lavorgna M, Parrella A, Temussi F. Acknowledgment fect of olive oil mill wastewater (OMW) Olive oil mill wastewater treatment on soil microbial communities and sup - using a chemical and biological ap - Ana Mª Gómez-Caravaca is grateful to pressiveness against Rhizoctonia solani. proach. J Agric Food Chem 2004; 52: “Fundación Alfonso Martín Escudero” Appl Soil Ecol 2004; 26: 113-21. 5151-54. and Spanish Ministry of Education (Pro - 7. Davies LC, Vilhena AM, Novais JM, 16. Aggelis G, Iconomou D, Christou M, et al. Phenolic removal in a model oli - grama de estancias de moulidad postdoc - Martins-Dias S. Olive mill wastewater ve oil mill wastewater using Pleurotus torad EX-2009-0343). characteristics: Modelling and statistical ostreatus in bioreactor cultures and The research was also supported by Mi - analysis. Grasas Aceites 2004; 55: 233- biological evaluation of the process. nisterio de Ciencia e Innovación (Accio - 41. 8. Obied HK, Allen MS, Bedgood DR, Water Res 2003; 37: 3897-904. ne Integradas España - Italia IT2009- Prenzler PD, Robards K, Stockmann R. 17. Rotondi A, Bendini A, Cerretani L, 0061) and MIUR (Italian Ministry of In - Bioactivity and analysis of biophenols Mari M, Lercker G, Gallina Toschi T. struction, University and Scientific Re - recovered from olive mill waste. J Agric Effect of olive ripening degree on the search) CAzioni Integrate Italia - Spagna Food Chem 2005; 53: 823-37. oxidative stability and organoleptic prot. IT10GL2347). 9. Bianco A, Buiarelli F, Cartoni G, Coc - properties of cv. Nostrana di Brisighel - cioli F, Jasionowska R, Margherita P. la extra virgin olive oil. J Agric Food Analysis by liquid chromatography tan- Chem 2004, 52: 3649-54.

30