"TOTAL FLAVONOID" CONTENT ASSESSMENT VIA ALUMINUM [AL(III)] COMPLEXATION REACTIONS. WHAT WE REALLY MEASURE?

Abstract

Frequently applied protocols for "total flavonoid" determination are examined with respect to limitations of Al(III)-flavonoid complexation reactions employed. Flavonoids and other phenolics tested were selected on the basis of their structure. The validity of observations and suggestions was then examined for various olive leaf extracts. Al-complexation reaction in the presence of NaNO^sub 2^ seems to be nonselective for flavonoid assessment as codetermination of o-diphenols strongly affects the results. The latter were found to correlate well with the respective Folin-Ciocalteau data (r^sup 2^ = 0.76). Under acidic conditions complexation with catecholic groups is not favored so that such protocols are considered more appropriate for flavonoid content estimation. The expression "total flavonoid" content is not always fully representative of what is really measured. Researchers should acknowledge analytical limitations of each protocol when they interpret their results. [PUBLICATION ABSTRACT]

Headnote

ABSTRACT

Frequently applied protocols for "total flavonoid" determination are examined with respect to limitations of Al(III)-flavonoid complexation reactions employed. Flavonoids and other phenolics tested were selected on the basis of their structure. The validity of observations and suggestions was then examined for various olive leaf extracts. Al-complexation reaction in the presence of NaNO^sub 2^ seems to be nonselective for flavonoid assessment as codetermination of o-diphenols strongly affects the results. The latter were found to correlate well with the respective Folin-Ciocalteau data (r^sup 2^ = 0.76). Under acidic conditions complexation with catecholic groups is not favored so that such protocols are considered more appropriate for flavonoid content estimation. The expression "total flavonoid" content is not always fully representative of what is really measured. Researchers should acknowledge analytical limitations of each protocol when they interpret their results.

- Key words: aluminium-flavonoid complexation, flavonoids, Oleaceae, olive (Olea europaea L.) leaves, plant extracts -

INTRODUCTION

Flavonoids, compounds found in all plant tissues (seeds, fruit skin or peel, bark, flowers, leaves, etc.) have been appraised for numerous biological and pharmacological effects since the antiquity (HAVSTEEN, 2002). Currently, many of the health-related effects of flavonoids are attributed to their antioxidant properties, whereas a wide variety of products containing them are commercially available (BENAVANTE-GARCÍA et al, 1999; HAVSTEEN, 2002; MALESEV and KUNTIC, 2007; KIOKAS etat, 2008).

Flavonoid analysis in a particular plant extract, as well as analysis of other phenolic compounds, includes not only identification of individual or groups of compounds using either simple or sophisticated separation techniques coupled to selective detection means but also estimation of their total content. Thus, except for the well-known Folin-Ciocalteu assay for the estimation of the "total phenol" content colorimetrie assays can also be found in literature for the assessment of "total flavonoid" content.

Flavonoids due to their structure (Fig. 1), easily chelate metal ions and form complex compounds. It seems that metal-flavonoid complexation reactions are particularly appropriate for analytical objectives as formed complexes bear exceptional spectrophotometric characteristics. The respective colorimetrie protocols, mainly those involving Al(III), are simple, rapid, and inexpensive and have met wide application even as official methods (MALESEV and KUNTIC, 2007). Nevertheless, current analytical demands that require validation of all analytical steps so that a method can be accepted for use led us to reconsider applicability of some of these protocols for the "total flavonoid" assessment.

Al(III) is likely to bind with 1, 2 or 3 molecules of bidentate ligands to form 1:1, 1:2, and less often 1:3, complex compounds, respectively, In the case of flavonoids, those that can act as ligands bear 5 or 3-hydroxy-4-keto and/ or o-dihydroxy groups (Fig. 1). The 3', 4'-dixydroxy group of B ring (cateholic moiety) seems to be less important though in a few studies it appears to be the basic complexation site (BODINI et al, 1999; CORNARD étal, 2001; CORNARD and MERLIN, 2003). Besides the above factors, reaction environment (pH, solvent) also influences analytical data (PORTER and MARKHAM, 1970; CORNARD et al, 2001; MALESEV and KUNTIC, 2007). Methanol is reported to be preferable than ethanol when used in mixture with water (PORTER and MARKHAM, 1970). Last but not least, aluminum forms acid stable complexes with the 5 or 3-hydroxy-4-keto groups of flavonoids and acid labile ones with any o-dihydroxy group present (MARKHAM and MARBY 1968; VOIRIN, 1983).

In the present study the Al(III) -flavonoid complexation as an analytical procedure was examined with respect to reaction environment. Flavonoids tested have been (or can be) used as reference compounds. Other phenolics were also examined to show method limitations. The validity of observations and suggestions was then tested for selected plant extracts.

MATERIALS AND METHODS

Chemicals

Oleuropein (98%) was purchased from Extrasynthèse (Genay, France); tyrosol (98%) was from Sigma-Aldrich (Steinheim, Germany); morin, quercetin, catechin hydrate minimum (98%), caffeic acid, ferulic acid and pcoumaric acid were from Sigma Chemical Co. (St. Louis, MO). DPPH- radical (1,1-diphenyl2-picrylhydrazyl, 90%) was also from Sigma Chemical Co. (St. Louis, MO, USA). Folin-Ciocalteu (F-C) reagent and sodium carbonate anhydrous were purchased from Panreac Química (Barcelona, Spain). Aluminum chloride anhydrous was from Fluka Chemie (Buchs, Switzerland). MeOH (99.7%) was from J. T. Baker (Baker Analyzed, Holland). Sodium hydroxide, sodium nitrite and acetic acid were from Merck (Darmstadt, Germany). All other common reagents and solvents were of the appropriate purity from various suppliers.

Leaf sampling

A representative olive leaf sample was used. The latter resulted from pooling together new and mature leaves collected on different dates (June 2006-February 2008) from trees belonging to various cultivars of an olive orchard (Agricultural Research Station, Agios Mamas, Chalkidiki, Greece). Detailed information on sampling parameters and postharvest treatment of the plant material are described in PAPOTI and TSIMIDOU (2009).

Leaf extracts preparation

Decoction: A suitable aliquot of distilled water was boiled in an Erlenmeyer flask. Then, an appropriate amount of the lyophilized plant material (1 g dry sample/ 100 mL H2O) was added and the mixture was further boiled for 10 min.

Infusion A: A suitable aliquot of boiled distilled water was immediately transferred to an Erlenmeyer flask, which contained an appropriate amount of the lyophilized plant material (1 g dry sample/ 100 mL H2O). The mixture was left to stand at 75°C for 10 min.

Infusion B: A suitable aliquot of boiled distilled water was immediately transferred to an Erlenmeyer flask, which contained an appropriate amount of the lyophilized plant material (1 g dry sample/ 100 mL H2O). The mixture was left to stand at room temperature for 5 min.

Tincture: A suitable amount of the lyophilized plant material (1 g dry sample/ 50 mL solvent) was extracted either in ethanol or in aqueous ethanol (50%, v/v) for 24 h at 25°-30°C under shaking (150 rpm).

Extract by sonication: Lyophilized plant material (1/40 w/v) was treated in an ultrasonic bath at room temperature for 5 min. Extraction solvent was either methanol or aqueous methanol (50%, v/v).

Determination of total polar phenol (TPP) content

TPP content of extracts was estimated using the F-C assay. Analytical details for the applied protocol are given in (PAPOTI and TSIMIDOU, 2009). Oleuropein was used as the external standard.

Determination of flavonoid content (FL)

The flavonoid content was determined according to two widely applied aluminium chloride colorimetrie methods based on the formation of aluminium-flavonoid complexes. Protocols applied were based on previously described methodologies (MARKHAM and MARBY, 1968; ZHISHEN and JIANMING, 1999; CVEK et al, 2007) that were modified in consistence to experimental requirements. The flow diagram of the employed protocols (protocol A and B) is given in Fig. 2 while details are presented below.

Protocol A is subdivided in protocols A1 and A^sub 2^ as follows:

Protocol A^sub 1^: Flavonoid content was estimated via a modified protocol based on that employed by CVEK et al (2007). Briefly, an aliquot (0.1 mL) of an aluminium chloride solution (2% w/v aluminium chloride in methanol) was added to 1 mL of the test solution (standard or extract 1,500 ppm as dry extract) and subsequently 1.4 mL of methanol was added. The mixture was left for 30 min in the darkness at room temperature and thereafter the absorbance was measured at 415 nm against a control. Absorbance measurements were corrected by subtracting initial sample absorbance at 415 nm.

Protocol A^sub 2^: Flavonoid content was estimated via the validated protocol of CVEK et al (2007) applying the same procedure as in protocol A1 in the presence of acid. Therefore, the aluminium chloride solutions added to the test solution, as well as the subsequently added methanol were acidified (5% v/v acetic acid).

The repeatability of measurement calculated for a morin standard solution and an extract was found satisfactory (CV% = 1 for both, n = 5). All determinations were carried out in triplicate.

The complexation activity of the tested standards with Al(III) was expressed via y = bx + a equations with y to represent the absorbance and x the final concentration (mM) tested. The respective concentration ranges tested for protocols A^sub 1^ and A^sub 2^ were: quercetin (0.01-0.09 mM and 0.006-0.059 mM); catechin (0-14 mM for both); morin (0.01-0.10 mM for both); oleuropein (0.6-4.5 mM for both); caffeic (0.22-0.78 mM for both); tyrosol (0-29 mM for both); p-coumaric acid (0-7 mM for both); ferulic acid (0-15 mM for both).

Results for the extracts were expressed as µg FL/g dry leaf through a morin calibration curve.

Protocol B: Aliquots of the test solution (standard or extract 1,500 ppm as dry extract) were appropriately diluted with doubly distilled water to achieve final volume of 1.775 mL. Then, at zero time (t=0) 75 µL NaNO^sub 2^ (5%) were added to the mixture. Subsequently, at 6 and 11 min 150 µL AlCl^sub 3^ (10%) and 0.5 mL NaOH 1M were added respectively to the mixture. Reactions took place in the darkness at room temperature. Absorbance was then measured at 5 10 nm against a control. The repeatability of measurement calculated for a catechin standard solution and an extract was found satisfactory (CV% < 5 for both, n = 9). Absorbance measurements were corrected by subtracting initial sample absorbance at 510 nm. All determinations were carried out in triplicate. The complexation activity of the tested standards with Al(III) was expressed via y = bx + a equations with y to represent the absorbance and x the final concentration (mM) tested. Concentration ranges for quercetin, catechin, morin, oleuropein, caffeic, tyrosol, p-coumaric and ferulic acids were 0.05-0.57 mM, 0.03-0.14 mM, 0.1-1.6 mM, 0.02-0. 1 1 mM, 0.02-0. 13 mM, up to 29 mM, up to 7 mM, up to 15 mM, respectively.

Results for the extracts were expressed as µg FL/g dry leaf through a catechin calibration curve.

RESULTS AND DISCUSSION

Careful search in literature indicated variation in experimental conditions upon application of Al(III)-flavonoid complexation reaction. It seems that critical issues such as the co-presence of compounds structurally related to flavonoids (e.g. o-diphenolic compounds) and chemistry of the protocols applied are not always taken into consideration or justified in the experimental design or results discussion, In the majority of relevant published papers (Table 1) the protocols applied are those examined in our study.

Protocol A involves addition of methanolic solution of AlCl3 to a sample tested and can be applied in the presence or absence of acid. The presence of acid does not allow complexation with catecholic groups (VOIRIN, 1983). Literature search for "total" flavonoid determination in plant material indicated that version A1 is almost exclusively applied (Table 1). Version A2 is met mainly in publications aiming at structure elucidation and detection of o-dihydroxyl in combination with A1 (MARKHAM and MARBY, 1968; VOIRIN, 1983). Similar protocols have been used for account fluorescent chelates formed among aluminum and certain flavonoids (flavonols that contain a free 3-hydroxyl and 4-keto oxygen binding site) (HOLLMAN et al, 1996).

Solubility limitations of aluminum salts in various solvents have been reported in the past (HOLLMAN et al, 1996). By trial and error 99.7% v/v aqueous methanol was found preferable in the case of protocol A, Additionally, as reaction time interval varied in literature, trials on the kinetics of reaction (0-30 min) were performed for an indicative morin solution and findings showed the importance of this parameter (data not shown). For the rest of the analysis measurements were obtained after 30 min of the addition OfAlCl3.

Complexation reaction in the presence of NaNO^sub 2^ (protocol B) is met equally often for "total flavonoid" determination as that OfA1. However, reaction conditions of protocol B are close to those applied in the past for the determination of o-diphenols (BARNUM, 1977). In that case, according to BARNUM (1977), the addition of sodium nitrite results in nitration of any aromatic ring bearing a catechol group with its 3 or 4 positions unsubstituted or sterically unblocked. In order to better understand chemistry of reaction, reagents were gradually added and spectra were recorded. The latter showed that in cases of reaction a complex was immediately formed after AlCl^sub 3^ addition, which resulted in a yellow color formation that was then turned to red immediately after NaOH addition. Kinetics showed that reactions are almost immediately completed after NaNO^sub 2^ and AlCl^sub 3^ addition; however time intervals of 5 min were included in accordance to literature practices,

In the majority of studies employing protocols A and B for "total flavonoid" determination in plant extracts quercetin and catechin (see Table 1) are the standards of choice for result expression. The flavonol quercetin (Fig. 3a) is rather used upon application of protocol A and the flavanonol catechin (Fig. 3b) seems to be preferred in protocol B. To our knowledge no documentation appears for these preferences in the respective papers. Nevertheless, it is questionable whether all flavonoids present in the samples examined can be determined via these experimental conditions or the selected standards are adopted through validation procedure.

In our study apart from these two widely employed reference flavonoids (quercetin and catechin) some other phenolics were selectively tested on the basis of structural characteristics. The chosen compounds were the flavonoid morin and the phenolic compounds oleuropein, caffeic acid, tyrosol, p-coumaric and ferulic acids (Fig. 3c-h). Morin, a T hydroxy - flavonol, was selected because it possesses the 5 and 3-hydroxy-4-keto groups of interest and lacks the catecholic moiety in B ring (Fig. 3c). Other characteristic flavonoids e.g. luteolin (flavone), apigenin (flavone) and respective glucosides as rutin (flavonol glucoside) were not included in the study. This was not only due to their high cost but also because they bear the 5 -hydroxy-4-keto group that is expected to be less potent for complexation in comparison to the 3 -hydroxy-4-keto one under neutral, acidic or alkaline environment (CORNARD et al, 2001). Besides, HOLLMAN et al, (1996) that studied the required structural properties of flavonols employing an aluminum complexating protocol, similar to that of protocol A, have stressed the importance of the presence of a free 3-hydroxyl group. Oleuropein (Fig. 3d), present in a great number of Oleaceae plants (NENADIS and TSIMIDOU, 2009), along with caffeic acid (Fig. 3e) were tested to examine whether, as catechol derivatives, could interfere with the total flavonoid complexation process. Furthermore, tyrosol (Fig. 3f), p-coumaric acid (Fig. 3g) and ferulic acid (Fig. 3h) were tested to complement discussion about possible sources of interference. Results for the above compounds under different reaction conditions are presented in Table 2.

Obviously, the presence of monophenols (tyrosol and p-coumaric acid; Table 2) does not influence quantitative data in both protocols. The presence of a guaicol moiety (ferulic acid; Table 2) was of some importance only under the conditions of protocol B whereas that of a catecholic group became crucial for complexation, irrespectively of the class of the molecule (flavonoid or not). Thus, slope values of oleuropein and caffeic acid indicated a high complexation activity of these phenolics, comparable to that of catechin. However, in the presence of 5 and/ or 3-hydroxy-4-keto groups (quercetin; Table 2) complexation via a catecholic moiety seems to diminish under reaction conditions of protocol B. The fact that complexation through 5 and 3-hydroxy-4-keto sites (characteristic moieties of many flavones, flavonols, flavanonones, isoflavones) is not favored under conditions of protocol B, initially noted for quercetin, was also verified for morin.

Protocol A was found incompatible with catechin in line with sited references (Table 1) that avoid it as a reference compound. It seems that its catecholic moiety (ring B) does not act as a site of complexation under acidic conditions. Protocol A was found appropriate for the determination of quercetin and morin as 5 and 3-hydroxy-4-keto moieties seem to be the main sites of complexation under these experimental conditions. To our view morin has to replace quercetin as an external standard in relevant studies because it lacks catecholic groups that may interfere; quercetin is more appropriate than rutin because the latter lacks a free 3-hydroxyl group and escapes fluorescence (HOLLMAN et al, 1996). Our data drawn for morin (Table 2) were as expected since the absence or presence of acid in the reaction medium was not crucial.

The influence of potent interfering compounds indicates that protocol B is not that selective for flavonoid determination in the presence of compounds bearing catecholic moieties. However, most plant extracts contain different classes of phenolic compounds, each one of importance for their overall functional properties. In the case of such extracts as most of non flavonoids bear catecholic moieties, determination of "total flavonoids" in the presence of sodium nitrite is nonspecific. This limitation is not always considered (references for protocol B in Table 1). In addition co-determination of other groups of phenolics (e.g. secoiridoids, simple phenols) via protocol A1 is negligible in comparison to findings upon application of protocol B. Using protocol A2 complex formation via o-dihydroxy moieties is totally eliminated. This was also verified for catechin, as the slight absorbance observed under protocol A2 conditions disappeared in excess of acid (10% v/v instead of 5% v/v). It can be argued that the three protocols examined give biased results and the expression "total flavonoids" is rather illusive. It should not be forgotten that such methods are usually applied before chemical characterization of an extract that possibly contains other types of phenolic compounds except for flavonoids. This is next exemplified for the case of olive leaf extracts.

Case study: Assessment of "total flavonoid" content of various olive leaf extracts

The applicability of the above mentioned aluminum-flavonoid complexation protocols was examined for various extracts of a representative olive leaf sample (Table 3). Olive leaf is a source rich in simple phenols (tyrosol, hydroxytyrosol and derivatives, phenolic acids: i.e. caffeic, chlorogenic, gallic, protocatechic, homovanillic, ferulic), secoiridoids (oleuropein, oleuroside, oleuropein aglycone, 3,4-DHPEA-EDA, demethyloleuropein), flavonoids (luteolin, apigenin, and their derivatives, quercetin, quercitrin, rutin) and other relevant compounds (e.g. verbascoside) (RYAN et al, 2002; TSIMIDOU and PAPOTI, 2010). For the same extracts assessment of total polar phenol content was also carried out via the F-C assay.

As expected (PAPOTI and TSIMIDOU, 2009; GOULAS et al, 2010) all of the extracts were identical qualitatively with regard to RP- HPLC phenol and flavonoid profile but varied in terms of quantitative data. Nevertheless, an interesting correlation was found between F-C data and protocol B ones (r^sup 2^= 0.76) whereas the respective correlations for F-C and A protocol versions were not significant [r^sup 2^= 0.33 (A1), r^sup 2^= 0.27 (A^sub 2^)], The above finding is strong evidence on the unsuitability of protocol B for "total flavonoid" content assessment. Taking into consideration the structural characteristics of flavonoids as well as of other phenolics expected in the studied extracts, protocol B can not be considered specific and selective for their determination. Many of olive leaf non flavonoid components (hydroxytyrosol and derivatives, oleuropein and derivatives, verbascoside, caffeic, chlorogenic, gallic, protocatechic acids) bearing catecholic moieties must influence quantitative data upon application of this protocol (HRNCIRIC and FRITSCHE, 2004).

The relatively low correlations for F-C and A1 and A^sub 2^ protocols indicated that groups other than the catecholic moieties contribute to the final quantitative data drawn upon application of protocol A versions. BCtWeCnA1 and A2 protocols the second one seems more specific for flavonoid determination. In this case "total flavonoid" content should mainly correspond to quercetin levels of an olive leaf extract. Other flavonoids, which lack a free 3 hydroxyl group, are expected to contribute less to complex formation reaction. The magnitude of respective results between A^sub 1^ and A^sub 2^ varied for the same extract and also among extracts. Alcoholic ones contained higher amounts of flavonoids in comparison to aqueous or aqueous alcoholic extracts as expected (Table 3). Moreover, in relevance to the protocol used even the order of samples in terms of their flavonoid content alters.

As an epilogue it can be said that protocol A^sub 2^ is preferable for the assessment of the presence of flavonoids that can form complexes with Al(III). The expression "total flavonoid" content is not fully representative of what is really measured in a plant extract as the method is dependent on the structure of the individual flavonoids present. Researchers should acknowledge this analytical limitation when they interpret their results.

References

REFERENCES

Al M. L., Daniel D., Moise A., Bobis O., Laslo L. and Bogdanov S. 2009. Physico-chemical and bioactive properties of different floral origin honeys from Romania. Food Chem. 112: 863.

Barnum D.W. 1977. Spectropho tome trie determination of catechol, epinephrine, dopa, dopamine and other aromatic vic-diols. Anal. Chim. Acta 89: 157.

Barreira J. C. M., Ferreira I. C. F. R, Oliveira M. B. P. P. and Pereira J. A. 2008. Antioxidant activities of the extracts from chestnut flower, leaf, skins and fruit. Food Chem. 107: 1106.

Benavente -Garcia O., Castillo J., Marin F. R, Ortunö A. and Del Rio J.A. 1997. Uses and properties of Citrus flavonoids. J. Agric. Food Chem. 45: 4505.

Bodini M.E., Copia G., Tapia R, Leighton F and Herrera L. 1999. Iron complexes of quercetin in apro tic medium. Redox chemistry and interaction with superoxide anion radical. Polyhedron 18: 2233.

Chen H.-Y., Lin Y.-C. and Hsieh C-L. 2007. Evaluation of antioxidant activity of aqueous extract of some selected nutraceutical herbs. Food Chem. 104:1418.

Conforti F., Sosa S., Marrelli M., Menichini F., Statti G. A., Uzunov D., Tubaro A., Menichini F. and Della Loggia R. 2008. In vivo anti-inflammatory and in vitro antioxidant activities of Mediterranean dietary plants. J. Ethnopharmacol. 116: 144.

Cornard J. P. and Merlin J.C. 2003. Comparison of the chelating power of hydroxyflavones. J. of MoI. Struct. 651-653:381.

Cornard J. P., Boudet A.C. and Merlin J.C. 2001. Complexes of Al(III) with 3%4%-dihydroxy-flavone: characterization, theoretical and spectroscopic study. Spectrochim. Acta A 57: 591.

Cvek J., Medic-Saric M., Jasprica L, Zubcic S., Vitali D., Mornar A., Verdina-Dragojevic I. andTomic S. 2007. Optimisation of an extraction procedure and chemical characterisation of Croatian propolis tinctures. Phyto chem. Anal. 18: 451.

Ebrahimzadeh M. A, Pourmorad F. and Bekhradnia A. R. 2008. Iron chelating activity, phenol and flavonoid content of some medicinal plants from Iran. Afr. J. Bio technol. 7: 3188.

Galvez M., Martín-Cordero C, Houghton P.J. and Ayuso M.J. 2005. Antioxidant activity of methanol extracts obtained from plantago species. J. Agrie. Food Chem. 53: 1927.

Goulas V., Papoti T. V., Exarchou V., Tsimidou Z. M. and Gerothanasis I. 2010. Contribution of flavonoids to the overall radical scavenging activity of olive Olea europaea L.) leaf polar extracts. J. Agrie. Food Chem. 58: 3303.

Havsteen B. H. 2002. The biochemistry and medical significance of the flavonoids. Pharmacology & Therapeutics 96: 67.

Hollman P.C. H., van Trijp J.M.P. and Buysman M.N.C.P. 1996. Fluorescence detection of flavonols in HPLC by postcolumn chelation with aluminium. Anal. Chem. 68: 3511.

Hrnciric K. and Fritsche S. 2004. Comparability and reliability of different techniques for the determination of phenolic compounds in VOO. Eur. J. Lipid Sci. Technol. 10: 540.

KimD.-O., ChunO.K., KimY.J., MoonH.-Y. andLeeC.Y. 2003. Quantification of polyphenolics and their antioxidant capacity in fresh plums. J. Agrie. Food Chem. 51: 6509.

Kiokias S., Varzakas T. and Oreopoulou V. 2008. In vitro activity of vitamins, flavonoids, and natural phenolic antioxidants against the oxidative deterioration of oil-based systems. Crit. Rev. Food Sci. 48: 78.

Lin J. -Y. and Tang C-Y. 2007. Determination of total phenolic and flavonoid contents in selected fruits and vegetables, as well as their stimulatory effects on mouse splenocyte proliferation. Food Chem. 101: 140.

Makris D. P., Boskou G. and Andrikopoulos N.K. 2007. Polyphenols content and in vitro antioxidant characteristics of wine industry and other agri-food solid waste extracts. J. Food Comp. Anal. 20: 125.

Malesev D. and Kuntic V. 2007. Investigation of metal-flavonoid chelates and the determination of flavonoids via metal-flavonoid complexing reactions. J. Serb. Chem. Soc. 72: 921.

Marinova D., Ribarova F and Atanassova M. 2005. Total phenolics and total flavonoids in bulgarian fruits and vegetables. Journal of the University of Chemical Technology and Metallurgy 40: 255.

Markham K. R. and Mabry T.J. 1968. A procedure for the ultraviolet spectral detection of ortho-dihydroxyl groups in flavonoids. Phyto chemistry 7: 1197.

Meda A., Lamien CE., Romito M., Millogo J. and Nacoulma O.G. 2005. Determination of the total phenolic, flavonoid and proline contents in Burkina Fasan honey, as well as their radical scavenging activity. Food Chem. 91: 571.

Mersch-Sundermann V., BahorunT., Stahl T., Neergheen V.S., Soobrattee M.A., Wohlfarth R., Sobel R., Brunn H. E., Schmeiser T., Lamy E. and Aruoma O.I. 2006. Assessment of the DNA damaging potency and chemopreventive effects towards BaP-induced geno toxicity in human derived cells by Monimiastrumglobosum, an endemic mauritian plant. Toxicol, in Vitro, 20: 1427.

Nenadis N. and Tsimidou M. Z. 2009. Oleuropein and related secoiridoids. Antioxidant activity and sources other than Olea europaea L. (olive tree). Recent Prog. Med. Plants 25: 53.

Ordonez A.A.L., Gomez J.D. Vattuone M.A. and Isla M.I. 2006. Antioxidant activities of Sechium edule (Jacq.) Swartz extracts. Food Chem. 97: 452.

Papoti V.T. and Tsimidou M. Z. 2009. Impact of sampling parameters to the radical scavenging potential of olive Olea europaea L.) leaves. J. Agrie. Food Chem. 57: 3470.

Porter L.J. and Markham K.R. 1970. The unsuitability of ethanol as a solvent for the spectroscopic detection of functional groups in hydroxyflavones with aluminum chloride. Phytochemistry 9: 1363.

Rathee J. S., Hassarajani S.A. and Chattopadhyay S. 2006. Antioxidant activity of Mammea longifolia bud extracts. Food Chem. 99: 436.

Ryan D., Antolovich M., Prenzler P., Robards K. and Lavee S. 2002. Biotransformations of phenolic compounds in Olea europaea L. Sci. Hort. 92: 147.

Simirgiotis M.J., Adachi S., To S., Yang H., Reynertson K.A., Basile M.J., Gil R.R., Weinstein LB. and KennelIy E. J. 2008. Cytotoxic chalcones and antioxidants from the fruits of Syzygium samarangense (Wax Jambu). Food Chem. 107: 813.

Tsai P.-J., Tsai T.-H. and Ho S. -C 2007. in vitro inhibitory effects of rosemary extracts on growth and glucosyltransferase activity of Streptococcus sobrinus. Food Chem. 105: 311.

Tsai T-H., Tsai T-H., Chien Y. -C, Lee C-W. and Tsai P-J. 2008. In vitro antimicrobial activities against cariogenic streptococci and their antioxidant capacities: A comparative study of green tea versus different herbs. Food Chem. 110: 859.

Tsimidou M.Z. and Papoti V.T. 2010. Bioactive ingrethents in olive leaves, p. 35 1-358. In: "Olives and olive oil in health and disease prevention". V. P. Preedy and R. R. Watson (Ed.), Elsevier Inc.

Viña S.Z., Mugridge A., García M.A, Ferreyra R. M., Martino M. N., Chaves A.R. and Zaritzky N.E. 2007. Effects of Polyvinylchloride films and edible starch coatings on quality aspects of refrigerated Brussels sprouts. Food Chem. 103: 701.

Voirin B. 1983. UV spectral differentiation of 5-hydroxyand 5-hydroxy-3- methoxyflavones with mono-(4';), di(3', 4') or tri- (3', 4', 5') - substituted B rings. Phyrochemistry 22: 2107.

Wang Y. -C, Chuang Y. -C and Hsu H.-W. 2008. The flavonoid, carotenoid and pectin content in peels of citrus cultivated in Taiwan. Food Chem. 106: 277.

Zhishen J.M.T and Jianming W 1999. The determination of flavonoid contents in mulberry and their scavenging effects on superoxide radicals. Food Chem. 64: 555.