Production of triacetylhydroxytyrosol from olive mill waste waters for use as stabilized bioantioxidant

Article abstract of DOI:10.1021/jf061290r

A hydroxytyrosol triacetyl derivative was very efficiently produced as a highly pure stabilized antioxidant compound by a short treatment of olive mill waste water (OMWW) organic extracts, rich in hydroxytyrosol, with an acetylating mixture composed of HC

Full text of DOI:10.1021/jf061290r

J. Agric. Food Chem. 2006, 54, 9063−9070 9063
Production of Triacetylhydroxytyrosol from Olive Mill Waste
Waters for Use as Stabilized Bioantioxidant
RENATO CAPASSO,*,† FILOMENA SANNINO,† ANTONIO DE MARTINO,† AND
‡
CATERINA MANNA
Dipartimento di Scienze del Suolo, della Pianta e dell’Ambiente, Facolta` di Agraria, Universita` degli
Studi di Napoli Federico II, Via Universita` 100, 80055 Portici (NA), Italy, and Dipartimento di
Biochimica e Biofisica “F. Cedrangolo”, Facolta` di Medicina e Chirurgia, Seconda Universita` degli
Studi di Napoli, Via de Crecchio 7, 80138 Napoli, Italy
A hydroxytyrosol triacetyl derivative was very efficiently produced as a highly pure stabilized antioxidant
compound by a short treatment of olive mill waste water (OMWW) organic extracts, rich in
hydroxytyrosol, with an acetylating mixture composed of HClO4-SiO2 and Ac2O (Chakborti and Gulhane
reaction), in mild and safe conditions. A successive single step of middle pressure liquid chroma-
tography (MPLC) purification of the reaction product was performed, with an overall yield of 35.6%.
(This process, including both the Chakborti and Gulhane reaction and the MPLC purification, is
protected by an international patent under PCT/IT2005/000781.) The o-diphenol triacetyl derivative
was also produced by direct reaction of hydroxytyrosol, previously purified by MPLC, with HClO4-
SiO2 and Ac2O, with an overall yield of 29.5%. A further procedure for the production of the
hydroxytyrosol triacetyl derivative was consistent with the direct treatment of raw OMWW with the
acetylating agent and a single step of MPLC purification, with an overall yield of 27.6%. The purified
natural triacetylhydroxytyrosol confirmed the same strong protective effects against the oxidative stress
in human cells as the corresponding synthetic compound, likely because of the biochemical activation
of the acetyl derivative into the active parent hydroxytyrosol by esterases. We therefore propose the
utilization of OMWW for recovering hydroxytyrosol as a natural antioxidant in a chemically stabilized
form, with a good yield, which can be potentially used as a nontoxic functional component in nutritional,
pharmaceutical, and cosmetic preparations.
KEYWORDS: Olive mill waste waters; continuous liquid-liquid extraction; chromatographic purification;
hydroxytyrosol; triacetylhydroxytyrosol; bioantioxidants
INTRODUCTION
Hydroxytyrosol [4-(2-hydroxyethyl)-1,2-benzenediol, 1a] (Fig-
ure 1) is the main polyphenol naturally occurring in olive oil
(1). This o-diphenol is also the most abundant polyphenol
present in either olive mill waste waters (OMWW) or in solid-
liquid waste (OMW), which are byproducts of three-phase (2,
3) and two-phase olive oil processing (4) plants, respectively.
Compound 1a also occurs naturally in virgin olive oil in the
form of a monoacetyl derivative (5) and has an antioxidant effect
very close to the parent compound (6).
Compound 1a is endowed with a free radical scavenging
activity (7) and has proven to be more active than the usual
industrial synthetic antioxidants such as 2,6-di-tert-butyl-p-
hydroxytoluene (BHT) and 3-tert-butyl-6-hydroxyanisole (BHA)
(8). This o-diphenol confers cell protection against oxidative
Figure 1. Structure of hydroxytyrosol (1a) and triacetylhydroxytyrosol (1b).
damages both in cellular (9, 10) and in vivo experimental models
(11-13); moreover, recent studies have pointed out that 1a is
endowed with significant antithrombotic, antiatherogenic, and
anti-inflammatory activities (14-16).
Compound 1a permeates the cell membrane via a passive
diffusion mechanism (17). Moreover, it is characterized by high
bioavailability in humans when orally administered (18), and it
* Corresponding author. Tel: +39 081 2539173. Fax: +39 081 2539186.
E-mail: renato.capasso@unina.it.
† Universita` degli Studi di Napoli Federico II.
‡ Seconda Universita` degli Studi di Napoli.
10.1021/jf061290r CCC: $33.50
© 2006 American Chemical Society
Published on Web 11/09/2006

9064 J. Agric. Food Chem., Vol. 54, No. 24, 2006
Capasso et al.
distributes in all organs and tissues when intravenously injected
in rats (19). Compound 1a is rapidly metabolized in vivo,
yielding metabolites modified either in the aromatic moiety or
in the lateral chain, some of them sharing the same antioxidant
activity of the parent compound (19). Finally, the treatment of
rats with up 2 g kg-1 induced no significant organ damage,
indicating that 1a can be regarded as a nontoxic compound (19).
These properties suggest that 1a could be utilized as a potential
antioxidant for nutritional, pharmaceutical, and cosmetic prepa-
rations (20).
Since 1a is not commercially available, many scientists of
multidisciplinary areas have been prompted to develop several
methods for its production: by direct recovery from OMWW
(2, 3) and OMW (4) or by chemical (2, 21), biochemical (22),
or biotechnological (23) syntheses starting from a synthetic
precursor.
The recent recovery methods have been consistent with the
extraction of OMW using a water stream under a pressure of
42 kg cm-2 at 160-240 °C in a pilot reactor (4), or of three-
phase OMWW, using ethyl acetate at room temperature in a
polyethylene mixer settler apparatus (3). However, because 1a
is unstable, particularly in solution, it has to be preserved dried
and in darkness in the absence of air.
Therefore, the recovery of this compound from OMWW as
a chemically stable and lipophilic derivative proves to be more
advantageous in this form, offering three practical points: (i)
higher yield with respect to the recovery from OMWW in the
native form; (ii) increased efficiency when added to alimentary,
pharmaceutical, or cosmetic matrices as a protective agent
against reactive oxygen species (ROS) in human cells; and (iii)
possible exploitation as a nontoxic additive to lipophilic
matrices. Clearly, a stable derivative of 1a should be able to be
biologically converted in vivo into the native form, which is
the effective protective form against ROS in human cells.
In our recent paper, we demonstrated 1a synthetic triacetyl
derivative [4-(acetoxyethyl)-1,2-diacetoxybenzene, 1b] (Figure
1) to be the most suitable compound able to satisfy these
objectives since it is transformed into 1a and nontoxic acetate
by esterases in human cells. This investigation, indeed (20),
showed 1b to be as effective as the parent 1a in protecting
human cells from oxidative stress-induced toxicity, after the
metabolization by esterases, although it is devoid of chemical
antioxidant activity. Therefore, the aim of this paper is to select
the most efficient method for the production of pure 1b from
OMWW.
In this perspective, both crude samples of OMWW and their
organic extracts were comparatively treated with three different
acetylating agents, consisting of Ac2O mixed with H2SO4,
pyridine, or HClO4 fixed on SiO2 (Chakborti-Ghulane reagent),
and the acetylated mixtures were submitted to middle pressure
chromatography (MPLC) purification to obtain pure 1b. Pure
1b was also produced by direct acetylation of 1a, obtained by
a previous chromatographic purification of an aliquot of the
same organic extracts stated previously. Finally, 1a and 1b
purified from OMWW were comparatively tested to the cor-
responding synthetic compounds, to confirm their chemical and
biological antioxidant behavior.
mental aspects, and cost of production of natural and synthetic
1b was also performed.
MATERIALS AND METHODS
Materials. Samples of OMWW (pH 4.5) were supplied by traditional
mills located in Marrakech (Morocco). They were lyophilized in a
Cryodos-Totelstar lyophilisator apparatus equipped with varian vacuum
technologies in the laboratory of Prof. Ismail El Hadrami (Faculty of
Semlalia, Marrakech, Morocco) and then redissolved in our laboratory
in high purity water (hpw) to the initial volume. Solvents were of
analytical and HPLC grade and were purchased from Carlo Erba (Milan,
Italy). High purity water was obtained through a double filtration
system, consisting of a deionized column and a MilliQ (Millipore)
apparatus. Authentic synthetic samples of 1a and 1b, which were used
as standards for chemical and biological antioxidant activity tests, were
synthesized according to our previously reported method (2). All of
the other materials used for the chemical and biological antioxidant
activity tests were the same described in the previous work (20).
Analytical TLC. TLC analysis was performed on silica gel plates
(Merck, Kieselgel 60 F254 0.25 mm) eluted with acetone/petroleum ether
50:50 (v/v) for monitoring 1a and acetone/petroleum ether 40:60 (v/v)
for monitoring 1b in the OMWW organic extracts and chromatographic
fractions. The spots were visualized by exposure to UV radiation and/
or by spraying first with 10% sulfuric acid in methanol and then with
5% phosphomolybdic acid in methanol, followed by heating at 110 °C
for 10 min. The qualitative monitoring of natural 1a and 1b was
performed by comparison with the corresponding synthetic standards.
HPLC Analysis. The quantitative and qualitative analysis of 1a and
1b was performed using an Agilent 1100 series liquid chromatograph
equipped with a DAD array. Detection was accomplished at 264 and
280 nm for the compounds 1b and 1a, respectively. A Nucleosil 100-5
C-18 column (stainless steel 250 × 4, Macherey-Nagel, 5 µm) was
utilized for both the compounds. An isocratic elution at a flow rate of
1.0 mL min-1 with acetonitrile/water (45:55, v/v) as mobile phase was
used for 1b (tR ) 7.62 min), and a flow rate of 1.2 mL min-1 with a
mixture of acetonitrile (15%) and water (85%, containing 0.35% acetic
acid) was used as mobile phase for 1a (tr ) 3.15 min). A sample volume
of 20 µL was used for the injection. HPLC analysis was conducted
after every extraction and purification step. The quantitative analysis
of 1b and 1a from OMWW samples was performed elaborating their
corresponding calibration curves (100-2000 and 50-1000 µg/mL,
respectively) using the synthetic standards.
Liquid-Liquid Extraction of OMWW. OMWW samples of 50
mL at native pH (4.5) were extracted with 100 mL of ethyl acetate by
refluxing for 8 h in a continuous liquid-liquid Friedrich extractor,
obtaining a total organic extract of 1102.0 mg. The amount of 1a was
quantitatively monitored by HPLC analysis during the extraction process
until a total recovery of 252.2 mg was obtained.
Middle Pressure Liquid Chromatography (MPLC). The purifica-
tion procedure was performed using a column of 36 mm (i.d.) × 460
mm (A) packaged with silica gel (Merck, Kieselgel 0.040-0.063 µm)
and eluted under middle pressure (20 bar) using a Buchi 681 pump.
The homogeneous fractions with the same Rf on the TLC analysis were
pooled and evaporated under reduced pressure at 30 °C.
1H NMR, EI-MS, and FAB-MS Analysis. The 1a and 1b
1
compounds obtained pure from OMWW were identified by H NMR
(400 MHz on a Bruker AC 400 spectrometer, using tetradeuterated
methanol as solvent for 1a and deutochloroform for 1b), EI-MS (70
eV on a Fisons Trio 2000 spectrometer), and FAB-MS (Kratos MS
50, xenon as the bombarding atoms at 8 kV acceleration potential)
analyses.
Production of Triacetylhydroxytyrosol [4-(Acetoxyethyl)-1,2-
diacetoxybenzene, 1b]. Preparation of HClO4-SiO2. The preparation
of the HClO4-SiO2 catalyst to perform the acetylation reaction with
Ac2O was carried out according to our modified procedure of Chakborti
and Gulhane (24). HClO4 (1.25 mmol, as a 65% aqueous solution) was
added to the suspension of silica gel (2.37 g, 230-400 mesh) in 7.5
mL of diethyl ether. The mixture was kept for 3 h at room temperature
under magnetic stirring and successively was heated at 70 °C for 28 h
under vacuum to afford HClO4-SiO2 as a free flowing powder.
The exploitation of natural 1b from OMWW for use as
stabilized lipophilic antioxidant additives in pharmaceutical,
nutritional, and cosmetic preparations should be worth while
because this waste material represents an economic source of
1a, a potential high added value substance. Its recovery could
also contribute toward resolving the disposal problem of this
waste material. A brief comparative study on safety, environ-

Stabilized Bioantioxidant Triacetyl Hydroxytyrosol from Olive Mill
J. Agric. Food Chem., Vol. 54, No. 24, 2006 9065
Figure 2. Production scheme of triacetylhydroxytyrosol (1b) according to the protocols a, b, and c.
Production of Pure 1b by Treatment of OMWW Organic Extracts
with a Mixture of HClO4-SiO2 and Ac2O and a SuccessiVe MPLC
Purification Step (Figure 2, Protocol a). An aliquot of 167 mg
(containing 38.3 mg of 1a) of extract obtained by liquid-liquid
extraction of OMWW was acetylated using as a catalyst HClO4-SiO2
(22.5 mg) and Ac2O (1 mL) in the presence of diethyl ether (1 mL)
under magnetic stirring and at room temperature. The reaction,
monitored by TLC, showed 1b to be quantitatively obtained after 5
min. The reaction mixture was then washed with ethyl acetate, filtered
under vacuum, and evaporated under reduced pressure at 30 °C, leaving
a residue of 262 mg. An aliquot of the raw acetylated mixture (260
mg) was purified on column A using as eluent hexane/ethyl acetate
1:1 (v/v) and a flow rate of 2.5 mL/30 s (fraction volume collected
was of 2.5 mL). The pooled fractions monitored by TLC and
quantitative HPLC showed 24.8 mg of pure 1b, which was also
identified by 1H NMR and EI-MS analyses. These analytical procedures
were also carried out on the same compound obtained following the
other protocols described next.
Production of Pure 1b by Treatment of Purified 1a with a Mixture
of HClO4-SiO2 and Ac2O (Figure 2, Protocol b). An aliquot of 155
mg (containing 35.5 mg of 1a) of the organic extracts obtained by
liquid-liquid extraction of OMWW was purified by column A using
as eluent hexane/ethyl acetate 15:85 (v/v) and a flow rate of 2.5 mL/
30 s. The homogeneous fractions with the same Rf of the synthetic
standard 1a were pooled, leaving a residue of 14.6 mg of pure natural
1a. This compound was analyzed as reported previously. An aliquot
of 13 mg of purified 1a was completely acetylated into 1b after 5 min
(the reaction was monitored by TLC) using the mixture of HClO4-
SiO2 (17.5 mg) and Ac2O (0.70 mL) in the presence of diethyl ether,
under magnetic stirring. The reaction mixture was worked up as
described previously, leaving a residue of 16.9 mg of pure 1b.
Production of Pure 1b by Direct Treatment of Lyophilized OMWW
with a Mixture of HClO4-SiO2 and Ac2O and a SuccessiVe MPLC
Purification Step (Figure 2, Protocol c). A sample of 1 g of lyophilized
OMWW (containing 30.24 mg of 1a) was acetylated under constant
stirring using as catalyst HClO4-SiO2 (106 mg) and Ac2O (4 mL) in
the presence of diethyl ether (6 mL). After 24 h, the reaction mixture
was washed with ethyl acetate, filtered under vacuum, and evaporated
under reduced pressure at 30 °C. The residue was purified on column
A using as eluent hexane/acetone 70:30 (v/v) and a flow rate of 10
mL/50 s (the fraction volume collected was 10 mL). The homogeneous
fractions with the same Rf of the synthetic standard 1b were pooled,
leaving a residue of 15.0 mg.
Appropriate amounts of antioxidant were added to 1 mL of working
solution, prepared by mixing 25 mL of 300 mM acetate buffer [pH
3.6], 2.5 mL of TPTZ solution [10 mM TPTZ in 40 mM HCl], and 2.5
mL of 20 mM FeCl3‚6H2O; after incubation for 6 min at room
temperature (23-25 °C), the absorbance was read at 593 nm.
Biological Tests for the Antioxidant Activity of 1a and 1b in
Caco-2 Cells and Human Erythrocytes. The antioxidant protective
effects on human Caco-2 cells and erythrocytes against oxidative injury
exerted by natural 1a and 1b isolated from OMWW or by the synthetic
ones were measured using the tests of cellular viability and hemolysis,
respectively.
Human colon carcinoma (Caco-2) cells were grown, and oxidative
stress was induced by t-BHP-treatment as previously reported (20). At
the end of 6 h incubation, t-BHP-induced cytotoxicity was measured
evaluating cell viability using the 3-(4,5-dimethylthiazol-2-yl)-2,5-
diphenyltetrazolium bromide (MTT) assay (26) , as previously described
(20).
Human erythrocytes were selected as a second model system to test
the antioxidant activity of the tested compounds. Cell suspensions (2%
hematocrit) were treated with t-BHP (500 µM final concentration). At
the end of 2 h of incubation, the extent of hemolysis was measured
spectrophotometrically, as previously described (20).
Evaluation of Lipoperoxidation. After the removal of the medium,
Caco-2 cells were solubilized, and lysates were centrifuged at 12000g
for 5 min; the obtained supernatants were assayed for TBARS
(thiobarbituric acid reactive substances) formation (9). Aliquots of
supernatants were added to thiobarbituric acid (TBA) in 0.05 N NaOH
(0.2% w/v final concentration) and heated in a boiling water bath for
10 min. The absorbance of the developed pink chromophore was
determined at 532 nm. For the measure of lipoperoxidation extent in
human erythrocytes, samples were mixed with trichloroacetic acid (10%
w/v final concentration) and centrifuged at 5000g for 15 min;
supernatants were finally assayed for lipoperoxidation products as
reported previously (20).
Statistical Analysis. The results of the biological tests are reported
as means ( SD, n ) 4; Student’s paired t-test (p < 0.05) (Stat-Works
software running on Apple Macintosh LC) was routinely utilized.
RESULTS AND DISCUSSION
The production of pure 1b by treatment with HClO4-SiO2
and Ac2O and MPLC purification either directly by the organic
extracts of OMWW or by the previously purified 1a was
performed according to the respective protocols a and b,
described in the scheme of Figure 2. Both methods have a
common first step consistent with the continuous liquid-liquid
Ferric Reducing Antioxidant Power (FRAP) Assay. The ferric
reducing ability of the tested compounds was measured using the FRAP
assay, a colorimetric method based on the reduction of a ferric-
tripyridyltriazine (ferric-TPTZ) complex to its ferrous form (25).

9066 J. Agric. Food Chem., Vol. 54, No. 24, 2006
Capasso et al.
confirmed to be coincident with those of the corresponding
synthetic and natural compound 1b previously described (2, 28,
29).
The acetylating agent composed of HClO4-SiO2 and Ac2O
appeared to be the most effective for the production of 1b (yield
of 35.6%), as compared with the mixtures of Ac2O and H2SO4
or pyridine, which gave a yield of 26.1 and 29.8% (experimental
data not shown), respectively.
The reaction of Chakborti and Gulhane provides many
advantages with respect to those shown by the reaction
performed with the other two reagents. In particular, the first
reaction is performed in 5 min, at room temperature; in addition,
its work-up is simpler, safer, and of lower cost because the
reagents of Chakborti and Gulhane are fixed on SiO2.
Both the mixtures composed of Ac2O and H2SO4 or pyridine
represent the usual acetylating agents, described in our previous
papers for the acetylation of purified polyphenols, either those
naturally occurring in OMWW (28) or synthetic ones (2).
However, the acetyl derivative 1b was prepared in our previous
studies as a main goal, to chemically characterize 1a and to
investigate the mass spectrometry behavior and the structure-
bioactivity relationship of the phenol compounds from agricul-
tural waste (2, 28, 29). The efficiency of the acetylation method
and its possible industrial applications were not investigated.
The increased efficiency of the catalyst HClO4-SiO2 with respect
to the traditional H2SO4 or pyridine can be attributed to the
fact that the former probably limits the secondary reactions that
occur in the latter reagents, consequently producing a higher
yield of the acetyl compounds. The higher yield of 1b obtained
using the first mixture is further confirmation of the efficiency
of this reagent for the preparation of acetyl phenols with high
yield, according to the results shown by Chakborti and Gulhane
(24). In addition, these authors demonstrated that the catalyst
could also be recycled, with obvious economical advantages in
potential industrial applications.
With regard to the production of 1b, obtained with the
protocol b (Figure 2), the organic extracts first underwent
MPLC purification obtaining pure 1a (yield 41%). The HPLC
analysis of this compound showed a corresponding single and
sharp peak at tR ) 3.15 min (data not shown). Moreover, its
Figure 3. HPLC chromatogram of the OMWW extract (a), acetylated
OMWW extract (b), and triacetylhydroxytyrosol (1b) (c).
extraction of raw OMWW with ethyl acetate. The total organic
extracts showed the maximal amount of 1102.0 mg containing
252.2 mg of 1a, quantitatively detected by HPLC analysis
(Figure 3a).
This revealed itself to be the most efficient extraction
procedure as compared with those described previously for the
extraction of phenols from OMWW (27). As shown in the
previous work, this method provided higher yield, and the
simpler work-up required smaller volumes of ethyl acetate,
therefore reducing its overall cost as compared to the continuous
liquid-liquid counter-current extraction method or the discon-
tinuous one, in batch mode.
Protocol a (Figure 2) produced a raw acetylated mixture
containing the acetyl derivative 1b as revealed by TLC (see
Materials and Methods) and HPLC controls (Figure 3b).
Compound 1b was quantitatively obtained after 5 min and at
room temperature. The reacted mixture was chromatographed
by a single MPLC step giving pure 1b (24.8 mg), with a yield
of 35.6%, related to the amount of 1a (38.3 mg) naturally
occurring in the organic extracts (167 mg). The high purity grade
of 1b was indicated by the corresponding single and sharp peak
at tR ) 7.62 min, as shown in the relative HPLC -chromato-
gram (Figure 3c). In addition, the 1H NMR, EI-MS, and FAB-
MS data reported in Table 1 proved to be fully consistent with
the structure of this compound (Figure 1, 1b) and were
1
respective H NMR and EI-MS data (Table 1) proved to be
fully consistent with the corresponding structure (Figure 1).
The purified 1a was then acetylated using the same procedure
followed protocol a (Figure 2), with a yield of 72%. Therefore,
the overall yield for the production of 1b using this latter
protocol proved to be 29.5%.
It is worth noting that the preparation of 1a from the same
OMWW source for the production of 1b was also motivated
by the necessity to homogeneously compare the data on the
chemical and biological antioxidant effects of the o-diphenol
with those of its corresponding acetylated form. The findings
of this comparative investigation are illustrated later.
In summary, protocol a appeared to be more efficient than
protocol b (Figure 2), 1b being produced by the first method
with a yield of 35.6% and by the second with an yield of 29.5%.
Very probably, compound 1a is adsorbed more strongly than
1b on the silica gel stationary phase and undergoes oxidative
and photochemical processes during the preparation by protocol
b, whereas it is protected by the acetyl groups in 1b during the
preparation by protocol a.
The only work that describes the production of ester deriva-
tives of 1a with the perspectives of industrial applications
regards a recent patent (30). This work reports a general
production of 1a esters that proved to be more resistant than

Stabilized Bioantioxidant Triacetyl Hydroxytyrosol from Olive Mill
J. Agric. Food Chem., Vol. 54, No. 24, 2006 9067
1a to air oxidation in an edibile matrix at 120 °C. Moreover,
this work reports on formulations in juice, nutritional beverages,
and butter preparations and on beneficial effects against oxida-
tive stress and inflammatory diseases. In particular, it describes
the production of a tristearyl derivative of 1a by the reaction of
1a with stearic acid, with a yield of 32%. This production
obviously requires a preliminary preparation of synthetic or
natural 1a, which is usually obtained with a yield ranging
between 41%sthe amount of the natural compound obtained
in the present worksand 79%, which is the amount reported in
the literature (2). Therefore, the production of a tristearyl
derivative of 1a can be obtained with an overall maximum yield
of 25.6%. In our work, the production of the acetyl derivative
1b according to the protocol a (Figure 2) is more convenient
(yield of 35.6%) because it is obtained by the direct treatment
of the OMWW organic extracts with the acetylating mixture,
without any previous isolation and purification of 1a.
mz/
Another method for the production of 1b, described in Figure
2, protocol c, was consistent with the direct treatment of
lyophilized OMWW (see Materials and Methods) with HClO4-
SiO2 and Ac2O and a successive MPLC step. But, the yield of
this procedure was lower (27.6%) than that obtained by treating
the organic extracts because of the chemical complex composi-
tion of the raw lyophilized OMWW mixture. In addition, this
procedure requires proportionally larger amounts of a chro-
matographic stationary phase and solvents for the purification
of the acetylated product considered, and consequently, it is
more laborious and expensive than the other methods considered
previously. The comparative yield for the production of 1b,
obtained by the different methods, is reported in Table 2.
z
m
J
J
In conclusion, the production of 1b appears to be much more
efficient and advantageous using the method that employs the
treatment of the phenol extracts with HClO4-SiO2 and Ac2O
(protocol a, Figure 2), as compared with the other methods
summarized in Table 2.
A brief evaluation of the production of 1b from OMWW as
compared to that of the synthetic form, using as a common step
the Chakborti and Gulhane reagent, with regard to safety,
environmental aspects, and cost, led us to assess the production
of natural 1b to be more effective than the corresponding
synthetic compound. In particular, assuming the production of
natural 1b to be obtained according to protocol a of Figure 2,
it is performed in three steps, with an overall yield of 35.6%.
Instead, the production of synthetic 1b would require four
steps: (i) the synthesis of 1a from 3,4-hydroxyphenylacetic acid;
(ii) the work-up of this reaction; (iii) the chromatographic
purification of the impure 1a [these first three steps and the
corresponding yield % are reported in detail by Capasso et al.
(2)]; and (iv) the Chakborti and Gulhane reaction. The yield of
this production is calculated to be 56%. The production of
synthetic 1b involves additional costs with respect to that of
the natural compound: (i) the reactor for the synthesis of 1a;
(ii) 3,4-hydroxyphenylacectic acid and reagents and solvents
related to the transformation of the substrate into 1a; (iii) the
disposal or recycle of the considered reagents and solvents; and
(iv) the expenses for one additional step (cost efficiency). Even
if the yield % of the production of synthetic 1b is higher than
that of the corresponding natural compound by almost 20%,
this increased production is not likely to justify the additional
costs considered previously. Moreover, the first step relative to
the synthesis of 1a also implies additional environmental and
safety problems with respect to the production of 1b from
OMWW. Finally, the utilization of this agricultural waste
material for the recovery of the natural stabilized antioxidant
z
m
J
J
J
J
J
J

9068 J. Agric. Food Chem., Vol. 54, No. 24, 2006
Capasso et al.
Table 2. Comparative Yield of Overall Production of 1b and Tristearyl-1a from OMWW
overall yield%
production of 1b by treatment of OMWW organic extracts with HClO4-SiO2 and Ac2O
production of 1b by treatment of OMWW organic extracts with H2SO4 and Ac2O
production of 1b by treatment of OMWW organic extracts with pyridine and Ac2O
production of 1b by treatment of preliminarily purified 1a from OMWW organic extracts with HClO4-SiO2 and Ac2O
production of tristearyl-1a by reaction of purified 1a with stearic acida
35.6
26.1
29.8
29.5
25.6
27.6
production of 1b by treatment of lyophilized OMWW with HClO4-SiO2 and Ac2O
a This method has been published in a recent patent (30). It is reported with a yield of 32% because it is related to the reaction with a preliminary purified synthetic or
natural 1a. Since the production of pure 1a is reported in literature with a maximum yield of 80% (2, 3), we calculated the maximal yield of 25.6% considering the overall
process.
extent of the native compound 1a (Figure 4a). Similarly, a
complete protection against oxidative stress-induced hemolysis
was observable in the antioxidant-treated erythrocytes, the
antioxidant effect of 1b being of some order of magnitude of
1a (Figure 4b). Finally, 1a and its acetyl analogue equally
prevented at the same concentration the oxidative alterations
of membrane phospholipids, as indicated by TBARS evaluation
as a lipoperoxidation marker (data not shown).
It is worth noting that natural 1a and 1b, obtained by
treatment of OMWW extracts as described in the present work,
show protective effects to be very close to those of the
corresponding synthetic compounds (Figure 4) prepared and
assayed as described in our previous work (20), indicating the
efficiency as biological antioxidants also of the compounds
obtained from the OMWW source; the data also indicate that
both the synthetic and the natural compounds have been obtained
with the same high purity grade.
All together, the findings of both the chemical and the
biological tests performed using natural 1a and 1b confirm our
previous data (20), indicating that synthetic 1b is endowed with
a strong biological antioxidant activity similar to that of the
parent compound, although devoid of ferric reducing ability,
likely after its activation into 1a at a cellular level by esterases.
It should be stressed, in this respect, that stable 1b that is
preserved against possible chemical oxidation and is activated
into 1a only at the cellular level could ensure the highest
antioxidant activity and therefore the highest protecting effect
in vivo.
On the basis of these considerations, we propose a possible
industrial production of natural 1b for its use as a nontoxic
functional component in a wide range of biological matrices
for human purposes, not only in cosmetic and pharmaceutical
preparations, but mainly as an active ingredient in food
preparation.
Figure 4. Comparative antioxidant effects of natural and synthetic
hydroxytyrosol (1a) and triacetylhydroxytyrosol (1b) on t-BHP-induced
oxidative stress in Caco-2 cells (a) and erythrocytes (b). These effects
were measured using the tests of cellular viability (MTT assay) and
hemolysis, respectively.
The food industry, indeed, is deeply involved in the identi-
can contribute to an integrative solution for disposal problems
fication of new functional ingredients with health-promoting
since the exhausted material from the phenol organic extraction
properties for designing high added value so-called functional
is recyclable potentially as an amendment because of its high
organic content (31, 32).
foods, innovative foodstuffs claiming to improve health and/or
to reduce the risk of diseases, by beneficially affecting one or
We then comparatively tested the chemical antioxidant
activity of natural 1a and its acetyl derivative 1b. As expected,
differently from its precursor 1a endowed with a high antioxi-
dant capacity, the acetylated compound does not shown any
ferric reducing activity, as measured using the FRAP assay,
confirming that 1b, lacking the o-diphenol moiety, is totally
inactive as a hydrogen donor and therefore does not exert any
antioxidant effect (data not shown).
The biological antioxidant activity of natural 1b was evaluated
using human cells, including colon carcinoma cells (Caco-2
cells) and erythrocytes (Figure 4a,b). Preincubation in the
presence of 50 µM 1b significantly prevents the decrease in
cell viability induced by t-BHP in Caco-2 cells to the same
more target functions in the body (33, 35).
Antioxidants have being amply used by the food industry as
additives in food production to prevent any alteration in their
taste, appearance, and safety. They play such an important role
in delaying the spoiling of foodstuff, including oxidation, a
chemical process due to exposure to air or to the effects of heat
and light, responsible for a decrease in the shelf life of products.
It is important to underline that protection against oxidation
exerted by antioxidants results in a loss of their concentration
in food, and a further decrease can occur during cooking;
therefore, only a small part of the antioxidant originally present
in the raw material is ingested. Therefore, specific compounds,
such as 1b, which do not exert antioxidant properties in food

Stabilized Bioantioxidant Triacetyl Hydroxytyrosol from Olive Mill
J. Agric. Food Chem., Vol. 54, No. 24, 2006 9069
(12) Scaccini, C.; Nardini, M.; D’Aquino, M.; Gentili, V.; Di Felice,
M.; Tomassi, G. Effect of dietary oils on lipid peroxidation and
on antioxidant parameters of rat plasma and lipoprotein fractions.
J. Lipid Res. 1992, 33, 627-633.
(13) Wiseman, S. A.; Mathot, J. N.; de Fouw, N. J.; Tijburg, L. B.
Dietary non-tocopherol antioxidants present in extra virgin olive
oil increase the resistance of low density lipoproteins to oxidation
in rabbits. Atherosclerosis 1996, 120, 15-22.
(14) Owen, R. W.; Giacosa, A.; Hull, W. E.; Haubner, R.; Wurtele,
G.; Spiegelhalder, B.; Bartsch, H. Olive oil consumption and
health: the possible role of antioxidants. Lancet Oncol. 2000,
1, 107-112.
(15) Visioli, F.; Galli, C. Biological properties of olive oil phy-
tochemicals. Crit. ReV. Food Sci. Nutr. 2002, 42, 209-221.
(16) Tuck, K. L.; Hayball, P. J. Major phenolic compounds in olive
oil: metabolism and health effects. J. Nutr. Biochem. 2002, 13,
636-644.
(17) Manna, C.; Galletti, P.; Maisto, G.; Cucciolla, V.; D’Angelo,
S.; Zappia, V. Transport mechanism and metabolism of olive
oil hydroxytyrosol in Caco-2 cells. FEBS Lett. 2000, 470, 341-
344.
(18) Visioli, F.; Galli, C.; Bornet, F.; Mattei, A.; Patelli, R.; Galli,
G.; Caruso, D. Olive oil phenolics are dose-dependently absorbed
in humans. FEBS Lett. 2000, 468, 159-160.
(19) D’Angelo, S.; Manna, C.; Migliardi, V.; Mazzoni, O.; Morsica,
P.; Capasso, G.; Pontoni, G.; Galletti, P.; Zappia, V. Pharma-
cokinetics and metabolism of hydroxytrosol, a natural antioxidant
from olive oil. Drug Metab. Dispos. 2001, 29, 1492-1498.
(20) Manna, C.; Migliardi, V.; Sannino, F.; De Martino, A.; Capasso,
R. Protective effects of the synthetic hydroxytyrosol acetyl
derivatives against oxidative stress in human cells. J. Agric. Food
Chem. 2005, 53, 9602-9607.
(21) Chikamatsu, Y; Ando, H.; Yamamoto, A; Kyo, S.; Yamashita,
K.; Dojo, K. Hydroxytyrosol as melanin and lipid peroxide
formation inhibitor and its application to topical and bath
preparations. Japanese Patent 8119825, 1996; pp 1-10.
(22) Espin, J. C.; Soler-Rivas, E.; Cantos, F. A.; Tomas-Barberan,
H. J.; Wichers, X. Synthesis of the antioxidant hydroxytyrosol
using tyrosinase as biocatalyst. J. Agric. Food Chem. 2001, 49,
1187-1193.
(23) Allouche, N.; Damak, M.; Ellouz, R.; Sayadi, S. Use of whole
cells of Pseudomonas aeruginosa for synthesis of the antioxidant
hydroxytyrosol via conversion of tyrosol. Appl. EnViron. Mi-
crobiol. 2004, 70, 2105-2109.
(24) Chakborti, A. K.; Gulhane R. Perchloric acid adsorbed on silica
gel as a new, highly efficient, and versatile catalyst for acetylation
of phenols, thiols, alcohols, and amines. Chem. Commun. 2003,
1896-1897.
(25) Benzie, I. F. F.; Strain, J. J. Ferric reducing/antioxidant power
assay: Direct measure of total antioxidant activity of biological
fluids and modified version for simultaneous measurement of
total antioxidant power and ascorbic acid concentration. Methods
Enzymol. 1999, 299, 15-27.
(26) Denizot, F.; Lang, R. Rapid colorimetric assay for cell growth
and survival. Modifications to the tetrazolium dye procedure
giving improved sensitivity and reliability. J. Immunol. Methods
1986, 89, 271-277.
but are converted in vivo in the active antioxidant compound,
could be regarded as a tool designed to specifically increase
the antioxidant potential in humans for delaying the progression
of several ROS-related pathologies, including CHD. However,
there are two important aspects that must always be taken into
account, including bioavailability and safety of the potential
additive. As pointed out in the Introduction, several observations
on both humans (11, 14) and animal (12, 13) models indicate
the high bioavailability and safety of 1a.
Referring to the possibility to use 1b in cosmetic preparations,
several authors have proposed that topically applied phenolic
antioxidants could represent a tool for protecting the skin against
UV-mediated oxidative damage (36, 37). In particular, recent
data indicate that caffeic acid and, at a higher degree, ferulic
acid (because of its higher lipophilicity) are able to permeate
through the stratum corneum and to exert significant protection
against UVB-induced skin erythema in healthy human volun-
teers.
Moreover, based on the important biological and pharmaco-
logical activities exerted by 1a, we consider that its acetylated
form could be proposed as a therapeutic agent in pharmacologi-
cal preparation. In this case, however, further studies should
provide unequivocal evidence of the therapeutic efficacy of 1b
and, more important, its safety at pharmacological doses.
LITERATURE CITED
(1) Manna, C.; Della Ragione, F.; Cucciola, V.; Borriello, A.;
D’Angelo, S.; Galletti, P.; Zappia, V. Biological effects of
hydroxytyrosol, a polyphenol from olive oil endowed with
antioxidant activity. AdV. Exp. Med. Biol. 1999, 472, 115-130.
(2) Capasso, R.; Evidente, A.; Avolio, S.; Solla, F. A highly
convenient synthesis of hydroxytyrosol and its recovery from
agricultural waste waters. J. Agric. Food Chem. 1999, 47, 1745-
1748.
(3) Allouche, N.; Fki, I.; Sayadi, S. Toward a high yield recovery
of antioxidants and purified hydroxytyrosol from olive mill
wastewaters. J. Agric. Food Chem. 2004, 52, 267-273.
(4) Ferdinando-Bolanos, J.; Rodriguez, G.; Rodriguez, R.; Heredia,
A.; Guillen, R.; Jimenez, A. Production in large quantities of
highly purified hydroxytyrosol from liquid-solid waste of two-
phase olive oil processing or Alperujo. J. Agric. Food Chem.
2002, 50, 6804-6811.
(5) Brenes, M.; Garcia, A.; Garcia, P.; Rios, J.; Garrido, A. Phenolic
compounds in Spanish olive oils. J. Agric. Food Chem. 1999,
47, 3535-3540.
(6) Gordon, M. H.; Martins, F. P.; Almenda, M. Antioxidant activity
of hydroxytyrosol acetate as compared with that of other olive
oil polyphenols. J. Agric. Food Chem. 2001, 49, 2480-2495.
(7) Visioli, F.; Bellomo, G.; Galli, C. Free radical scavenging
properties of olive oil polyphenols. Biochem. Biophys. Res.
Commun. 1998, 247, 60-64.
(8) Ranalli, A.; Lucera, L.; Contento, S. Antioxidizing potency of
phenol compounds in olive oil mill wastewater. J. Agric. Food
Chem. 2003, 51, 7636-7641.
(27) Sannino, F.; De Martino, A.; Capasso, R.; Allouche, N.; Sayadi,
S. Recovery of hydroxytyrosol from olive mill wastewater by
two different continuous liquid-liquid extraction procedures.
Polyphenol. Commun. 2004, 803-804. Ibid, XXII International
Conference on Polyphenols: August 25-28, 2004, Helsinki,
Finland.
(28) Capasso, R.; Evidente, A.; Scognamiglio, F. Isolation, spectros-
copy, and selective phytotoxic effects of polyphenols from
vegetable wastewaters. Phytochemistry 1992, 31, 4125-4128.
(29) Capasso, R. A review on the electron ionization and fast atom
bombardment mass spectrometry of polyphenols naturally oc-
curring in olive wastes and of their synthetic derivatives.
Phytochem. Anal. 1999, 10, 299-306.
(9) Manna, C.; Galletti, P.; Cucciola, V.; Moltedo, O.; Leone, A.;
Zappia, V. The protective effect of olive oil polyphenols (3,4-
dihydroxyphenyl)-ethanol counteracts reactive oxygen metabolite-
induced cytotoxicity in Caco-2-cells. J. Nutr. 1997, 127, 286-
292.
(10) Manna, C.; Migliardi, V.; Golino, P.; Scognamiglio, A.; Galletti,
P.; Chiariello, M.; Zappia, V. Oleuropein prevents oxidative
myocardial injury induced by ischemia and reperfusion. J. Nutr.
Biochem. 2004, 15, 461-466.
(11) Visioli, F.; Caruso, D; Galli, C.; Viappiani, S.; Galli, G.; Sala,
A. Olive oils rich in natural catecholic phenols decrease
isoprostane excretion in humans. Biochem. Biophys. Res. Com-
mun. 2000, 278, 797-799.

9070 J. Agric. Food Chem., Vol. 54, No. 24, 2006
Capasso et al.
(30) Geerlings, A.; Lopez-Huertas, L. E.; Morales, S. J. C.; Boza, P.
J.; Jimenez Lopez, J. Natural compounds and derivatives thereof,
including tyrosol and hydroxytyrosol esters with fatty acids, for
the prevention and treatment of cardiovascular, hepatic, and renal
diseases and for cosmetic applications. International Patent
2003082798, 2003; pp 1-39.
(31) Capasso, R.; De Martino, A.; Arienzo, M. Recovery and
characterization of the metal polymeric organic fraction (poly-
merin) from olive oil mill waste waters. J. Agric. Food Chem.
2002, 50, 2846-2855.
(32) Capasso, R.; De Martino, A.; Cristinzio, G. Production, char-
acterization, and effects on tomato of humic acid-like polymerin
metal derivatives from olive oil mill waste waters. J. Agric. Food
Chem. 2002, 50, 4018-4024.
(33) Katan, M. B.; De Roos, N. M. Promises and problems of
functional foods. Crit. ReV. Food Sci. Nutr. 2004, 44, 369-377.
(34) Kwon, T. W.; Hong, J. H.; Moon, G. S.; Song, Y. S.; Kim, J. I.;
Kim, J. C.; Kim, M. J. Food technology: challenge for health
promotion. Biofactors 2004, 22, 279-287.
(35) Hirata, T. Key factors for the success of functional foods.
Biofactors 2004, 22, 289-294.
(36) Mantena, S. K.; Kativar, S. K. Grape seed proanthocyanidins
inhibit UV radiation-induced oxidative stress and activation of
MAPK and NF-kappaB signaling in human epidermal kerati-
nocytes. Free Radical Biol. Med. 2006, 40, 1603-1614.
(37) Sajia, A.; Tomaino, A.; Trombetta, D.; De Pasquale, A.; Uccella,
N.; Barbuzzi, T.; Paolino, D.; Bonina, F. In vitro and in vivo
evaluation of caffeic acids as topical photoprotective agents. Int.
J. Pharm. 2000, 199, 39-47.
Received for review May 8, 2006. Revised manuscript received
September 21, 2006. Accepted September 25, 2006. This work was
supported by EU grants within an INCO-IAC3-CT2002-10033 project.
DiSSPA number 123.
JF061290R