- Email: [email protected]
Reaction of iron(III) with theaflavin: complexation and oxidative products
JOURNAL OF
Inorganic Biochemistry Journal of Inorganic Biochemistry 98 (2004) 657–663 www.elsevier.com/locate/jinorgbio
Reaction of iron(III) with theaflavin: complexation and oxidative products Mairtin O’Coinceanainn a
a,*
, Samuel Bonnely a, Beate Baderschneider a, Michael J. Hynes b
Unilever Research and Development, Colworth Laboratory, Sharnbrook, Bedford MK44 1LQ, UK b Department of Chemistry, National University of Ireland, Galway, Ireland
Received 25 July 2003; received in revised form 20 November 2003; accepted 29 December 2003
Abstract Theaflavins are a family of compounds, whose chemistry has been sparsely investigated. They can comprise up to 40% the dry weight of black tea. They are known to chelate metals, however very little knowledge exists on the mechanisms involved. There is some correlation between both of these areas in that following degradation of the iron theaflavin complex, subsequent redox reactions may lead to the formation of similar products on both occasions. The interaction of iron(III) with theaflavin at pH <3.0 is investigated by means of liquid chromatography mass spectroscopy (LC-MS), stopped flow spectroscopy and multivariate data analysis. Iron theaflavin complexes are formed which subsequently decay to form a number of oxidative species. The difficulties involved in the elucidation of the structure of polymeric phenolic compounds from black tea has been highlighted by numerous authors. The intermediates and major low molecular weight oxidised theaflavin products from the reaction of excess iron with theaflavin have been detected and identified using multivariate data analysis of diode array spectroscopic data. It is not possible to characterise the extremely polar high molecular weight oxidation products obtained from polyphenol oxidation. High performance liquid chromatography (HPLC) and electrospray mass spectroscopy (ES-MS) detected the low molecular weight oxidised theaflavin species present in the system. Enzymatic oxidation of theaflavin using peroxidase (POD) resulted in the formation of one major low molecular weight species oxidative product, which was fully characterised using nuclear magnetic resonance spectroscopy (NMR), high performance liquid chromatography (HPLC), electrospray mass spectroscopy (ES-MS), UV–visible (UV–Vis) and Fourier transform infra-red spectroscopy (FT-IR). The major objective of this work is to investigate the reaction of iron(III) with theaflavin and to add some insight into the mechanistic interaction of iron(III) with this family of compounds. Ó 2004 Elsevier Inc. All rights reserved. Keywords: Theaflavin; Iron(III); Complex; Oxidation
1. Introduction The theaflavins are a family of compounds, which have been sparsely investigated. Theaflavins can become involved in the generation of high molecular weight polyphenolic species during black tea fermentation [1,2]. Numerous health benefits have been attributed to this family of compounds. Their method of antioxidative * Corresponding author. Tel.: +44-1234-222504; fax: +44-1234222844. E-mail address: [email protected] (M. O’Coinceanainn).
0162-0134/$ - see front matter Ó 2004 Elsevier Inc. All rights reserved. doi:10.1016/j.jinorgbio.2003.12.018
behaviour in the body has been investigated by a number of authors [3,4]. Theaflavins interact with certain metal ions and may substantially alter their bioavailability [3]. The antioxidative properties of theaflavin and the gallated theaflavins has been partially ascribed to their ability to scavenge free metals [3], although blood plasma concentrations of theaflavin are extremely low when compared to those of green catechins [4]. It has been more recently suggested that their beneficial effect may be more due to their strong enzymic inhibition capabilities than their antioxidative effects [5]. The levels of theaflavin detected in blood plasma are quite low [4],
658
M. O’Coinceanainn et al. / Journal of Inorganic Biochemistry 98 (2004) 657–663
hence the possible antioxidative effect of theaflavin through the scavenging of metals in the lower pH region of the alimentary canal is of interest. The formation of a variety of metal theaflavin species strongly influences mobility and transport processes and also the bioavailability of metals and ligands [3]. In addition to the beneficial effects, there are possible adverse effects of polyphenolic compounds associated with tea drinking. There is evidence that some tea constituents can have a negative association with measures of iron status in a number of different populations [6–9]. A more recent study has suggested that in the well balanced diet of western populations this effect is negligible [10]. Polyphenols, together with other metal chelators have also been reported to have a positive influence on the treatment of iron overloading in humans [11,12]. There is currently little data available in the literature on the chemistry of the reaction of iron with this family of compounds. This paper reports on the reaction of iron(III) with theaflavin, the parent member of this family and the determination of the resulting products. The work is mainly carried out at low pH due to the inherent problems arising from the hydrolysis of iron and the redox reactions between iron(III) and theaflavin. The interaction of iron with theaflavin under physiological conditions has been previously investigated [3]. The postulated low molecular weight oxidation products are compared to those obtained by enzymatic oxidation carried out on theaflavin under a more controlled environment. The biological importance of this family of compounds is without question, while very little is known of their chemistry, in particular through which they react with metals, iron in particular. It is the aim of this work to contribute to the understanding of the chemistry of this family of compounds.
2. Materials and methods 2.1. High performance liquid chromatography and electrospray ionisation The HPLC system was a Waters Alliance system (model 2690) coupled with a Waters ZMD single quadrupole electrospray ionisation mass spectrometer (ESI) (Micromass, Manchester, UK). The UV detection was done at 280 nm. The separation was performed on a Synergi Hydro column (150 mm 4.6 mm i.d.; particle size, 5 lm) (Phenomenex, Macclesfield, Cheshire, UK) at 30 °C. The flow rate was 0.2 ml/min. The gradient elution was as follows: solvent A, 0.1% formic acid in water; solvent B, 0.1% formic acid in acetonitrile; from 0 to 30 min, 10–100% B; from 30 to 35 min 100% B. The instrument was operated in the positive
and negative electrospray mode with a capillary voltage at 3.2 and 2.8 kV, respectively. Total ion chromatograms were recorded in the full scan mode from m/z 200 to 1000 at cone a voltage of 30 and 25 V in the positive and negative modes respectively. Data were collected between 0 and 35 min. Source temperature was set at 120 °C, desolvation temperature at 400 °C. Nitrogen was used as nebulising gas and drying gas with flow rates of 25 and 500 l/h, respectively. Iron– theaflavin samples were investigated for the presence of metal–polyphenol complexes by flow-injection-MS analysis. The micromass ZMD mass spectrometer was operated using the electrospray ionisation source. The sample solutions were introduced into the electrospray source at a flow rate of 10 ll/min using a syringe pump. The electrospray source conditions were optimised for the generation of the molecular related ion and data were acquired by accumulating several continuum spectra over a mass range of m/z 10–2000. The following instrumental settings were applied: source temperature 100 °C, desolvation temperature 200 °C. Nitrogen was used as nebulising gas and drying gas with flow rates of 25 and 200 l/h, respectively. The capillary voltage was 3.2 kV for positive electrospray and 2.8 kV for negative electrospray ionisation, respectively. The cone voltage was set at 65 and 50 V, respectively. 2.2. Spectroscopy Spectra and kinetic runs for the slowest reactions were obtained as a function of time by use of a Hewlett– Packard 8453A diode array spectrophotometer equipped with a HI-TECH SFA-20 rapid kinetics stopped-flow accessory. All spectra and kinetic runs were recorded at 25 °C. Kinetic data were obtained on a HI-TECh SF20 stopped-flow apparatus, which was interfaced to a PC via a Datalab DL901 transient recorder. The reactant reservoirs were thermostatted at 25 °C. All reagents were of the highest grade available, as detailed elsewhere [13–15]. The FT-IR spectra were recorded from dried powder using a Bio-Rad FTS-6000 spectrometer (Bio-Rad Laboratories, Cambridge, MA, USA) and a Golden-Gate single reflection diamond ATR unit (Specac Inc., Smyrna GA, USA). A spectral resolution of 2 cm1 was used in all measurements and 1000 interferograms were co-added before Fourier transformation. 2.3. NMR spectroscopy 1
H NMR spectra were measured at 400 MHz on a Bruker AMX400 spectrometer using a multinuclear 5 mm inverse probe at 300 K. The solvent used was 2 H2 O and internal reference TSP. HMQC spectra were run at 300 K using the multinuclear 5-mm inverse probe.
M. O’Coinceanainn et al. / Journal of Inorganic Biochemistry 98 (2004) 657–663
2.4. Purification of theaflavin Theaflavin was isolated and purified according to Unilever Bestfoods in-house proceedures as detailed elsewhere [14,15]. 2.5. Global kinetic analysis Spectral-kinetic data were processed as previously described [13] by use of the program SPECFIT/32TM (Spectrum Software Associates, Malborough, MA), which is based on the published works of Zuberb€ uhler et al. [16–20]. 2.6. Oxidation with a tea leaf homogenate [21] The method employed is the subject of a separate paper, but a short summary is as follows. Tea leaves (25 g) were homogenised in a domestic blender (3 10 s) with water (250 ml). The mixture was transferred in a 10 l container and water (4.8 l) was added. The mixture was stirred (using a Heidolph ST1 stirrer, speed 3) and continuously sparged with helium for 2 h. The mixture was filtered and the leaf residue washed a further three times. Washed leaves (6 g) were added to purified theaflavin (100 mg dissolved in 100 ml of water) and the mixture stirred (using a Heidolph ST1 stirrer, speed 2). Peroxidase catalysed oxidation was carried out under helium (to prevent polyphenoloxidase oxidation) and hydrogen peroxide (1.6 ml, 1.5% v/v) was added to the medium. After 1 h oxidation, the mixture was filtered and lyophillised.
3. Results and discussion Theaflavin and its derivatives are known to form complexes with iron and hence have good metal scavenging ability in vitro under physiological conditions [3]. There is however very little theaflavin detected in blood plasma, hence the current investigation is carried out under more acidic conditions in order to investigate the effect of oxidative environment and metals on this biologically significant family of phenols. The reaction of metal ions, Fe3þ in particular, with phenolic-based compounds has been widely investigated, however theaflavin is a much more complicated phenolic compound than previously investigated [22–25]. The general reaction mechanism involves the formation of a metal complex, followed by subsequent decomposition due to differing modes of electron transfer [22–25]. SVD analysis of the spectral data obtained from the reaction of an excess of iron(III) for a period of 10 min with theaflavin indicated the presence of only four eigenvectors in the time and spectral domain corresponding to three different reaction steps as indicated in
659
Fig. 1. The other eigenvectors obtained contained no significant colorometric information related to the reaction. The program SPECFIT/32TM enables the isolation of the spectral data of the major species detected and the elucidation of the major pathways involved. Theaflavin contains two possible binding sites. With the large excess of metal employed the first two reactions could be interpreted as the formation of 1:1 and 2:1 metal to theaflavin complexes on the basis of similar studies in this area [13]. The formation of the iron theaflavin complexes was too fast to be accurately further analysed by conventional diode array spectroscopy, but was detected by fast kinetic analysis using stopped flow spectroscopy with k obs of the order of 3 s1 at 700 nm. Both iron(III) theaflavin complexes subsequently decomposed, resulting in the build up of benzoquinone (kmax ¼ 378 nm). The nature of the redox reaction leading to the decomposition was not further investigated. The spectral data for the second step of the reaction can be partially isolated on the basis of the sharp isosbestic point at 390 nm as indicated by the spectral data in Fig. 2. The formation of the benzoquinone moiety is quite consistent with the enzymatic oxidative work carried out by Tanaka [5]. The benzoquinone is unstable in the oxidative environment and rearranges to a further product as indicated in Fig. 3 with a kmax ¼ 440 nm. Tanaka has dealt in detail with the mechanism involved during this rearrangement [5]. This complete reaction of iron(III) with theaflavin in solution was further investigated using HPLC and ES-MS Fig. 4. The theaflavin peak (RT 14.48 min) and the signal of the oxidation product (RT 14.16 min) were not resolved in the total ion chromatogram due to the slow scan speed of the mass spectrometer, but this is possible through the use of UV detection at 280 nm. Two clearly separated signals for the two compounds can be obtained by selective ion detection Fig. 5. The main oxidative species detected in the final solution after 20 min had a molecular weight (M + Hþ ) of 535 with a retention time of 14.15 min. A species with the same molecular weight and spectral characteristics (kmax ¼ 440 nm) (theanapthoquinone) was obtained by Tanaka following the rearrangement of benzoquinone. Further support for the assignment of this species arises during the enzymatic work. Evolving factor analysis of the spectral data obtained from the reaction of iron(III) with theaflavin indicates the evolution of only four significant species over time domains of 10 min or less (see Fig. 6). These correspond to the original theaflavin, metal–theaflavin complex, theabenzoquinone, and theanaphthoquinone. No other significant species appear after the first 30 s of the reaction from spectroscopic data. Two further minor species both of molecular weight 578, with retention times of 9.14 and 11.59 min,
660
M. O’Coinceanainn et al. / Journal of Inorganic Biochemistry 98 (2004) 657–663
Fig. 1. (a) First four temporal eigenvectors from U S (x-axis is s). (b) Four spectral eigenvectors V (x-axis is wavelength). Iron(III) ¼ 2 mM, theaflavin ¼ 0.2 mM, [Hþ ] ¼ 0.01 M. This represents three separate spectral data transitions or three individual reactions.
respectively, were detected by MS. These two species were not clearly detected as ‘real species’ from UV spectroscopy, however the chromatogram ionisation peaks for these species are quite strong, thus enabling their detection through this technique. The structures of these compounds can only be postulated on the basis of mass spectral data, however this together with strong literature support enables the confident assignment of
structures. The first of these can be tentatively assigned as dehydrotheaflavin on the basis of literature comparison with the results of Tanaka [26] (product C in Fig. 3). The second minor component could also be assigned tentatively as another structural analogue of dehydrotheaflavin (product B). Tanaka has previously detected dehydrotheaflavin as an oxidation product of theaflavins during the black tea fermentation process
M. O’Coinceanainn et al. / Journal of Inorganic Biochemistry 98 (2004) 657–663
661
Fig. 4. Total ion chromatogram of the reaction of iron(III) with theaflavin. Iron(III) ¼ 2 mM, theaflavin ¼ 0.2 mM, [Hþ ] ¼ 0.01 M. Fig. 2. Spectral data obtained for the formation of theabenzoquinone following decomposition of the iron theaflavin complex. Iron(III) ¼ 2 mM, theaflavin ¼ 0.2 mM, [Hþ ] ¼ 0.01 M.
under controlled conditions [24], however the structural analogue was not detected. It is not possible to obtain a complete mass balance of this system due to the longstanding inability to characterise the high molecular weight phenolic component of black tea. We can however ascertain from HPLC analysis that following completion of the reaction there is only 5% of the original theaflavin starting material remaining. The other species are also only present in low levels, which are difficult to quantify. It has not been possible to characterise the polymeric phenolic component, which is the major oxidative product of the reaction [2]. Potentiometric analysis was carried out on the interaction of iron(III) with an excess of theaflavin. The oxidative reactions and the formation of theanaphthoquinone rendered it quite difficult to obtain reliable stability constant values, a problem which is prevalent with iron(III) polyphenol systems. Samples were withdrawn from the potentiometric analysis at pH’s 3.50, R
OH
R
O
OH
2 Fe +
O
R
1
OH
O
O
O
R
4.50 and 5.50. Analysis of samples which were 1 mM theaflavin to 0.1–0.3 mM iron(III) using the ES-MS method above, indicated the presence of equivalent levels of only theanapthoquinone over the three pH regions. The levels present were however linearly dependent on the levels of iron(III) initially employed.
R
Fe
3+
Fig. 5. Chromatogram of the reaction of iron(III) with theaflavin with selective ion detection.
Fe
O
Theabenzoquinone High molecular weight oxidation products
O 2
3
O
R OH
Theaflavin
O H
R
O HO
HO R OH
O
O
HO
OH
HO O
O O
O
HO
C
OH
O
O OH
OH
HO R=
O
OH OH
O A Theanaphthoquinone
OH
OH
O HO
OH
O
OH
OH
B
Fig. 3. Mechanism of reaction of iron(III) with theaflavin and the subsequent products.
662
M. O’Coinceanainn et al. / Journal of Inorganic Biochemistry 98 (2004) 657–663
Fig. 6. Evolving factor analysis of the reaction of iron with theaflavin. Only four significant eigenvectors corresponding to theaflavin, iron complex, theabenzoquinone and theanaphthoquinone. Iron(III) ¼ 2 mM, theaflavin ¼ 0.2 mM, [Hþ ] ¼ 0.01 M.
It is well known that enzymatic oxidation of theaflavin is difficult to achieve with polyphenol oxidase while peroxidase is described to be an efficient oxidative enzyme for this substrate. Enzymatic oxidation of theaflavin under more controlled conditions than the harsh chemical treatment, is quite useful in the elucidation of the mechanistic interactions occurring in the harsher environment. Enzymatic oxidation was carried out on pure theaflavin as indicated in the materials and method section, and the major low molecular weight product was characterised. FT-IR spectra of the theaflavin before and after 1 h peroxidase oxidation indicates the formation of an extra carbonyl peak at 1670 cm1 corresponding to an additional carbonyl group. HPLC analysis indicated the formation of one major oxidative product with a retention time of 14.22 min and kmax ¼ 440 nm. ES-MS confirmed a molecular ion peak of (535 M + Hþ ). 1 H NMR and 13 C analysis confirmed the product as theanaphthoquinone [27]. The characterisation under these conditions is unequivocal, hence arising from the retention time and mass spectrum of this product being the same as that obtained for the major product of the interaction of iron(III) with theaflavin, it can be employed as a standard to confirm the assignment of the product obtained in the iron(II) system on the basis of retention time and spectra obtained. It is noteworthy that the main oxidation product obtained under the different systems of oxidation, theanaphthoquinone, unlike most other phenolic products is quite stable under extremes of pH and oxidative environment employed under the two different environments.
4. Conclusion Polyphenols are generally quite unstable under oxidative conditions, degrading to form products of
varying molecular weights. Spectroscopic data suggest that earlier part of the reaction of the reaction of iron(III) with theaflavin is not complicated as illustrated by stopped flow spectroscopic data and multivariate analysis of spectroscopic data under the conditions investigated suggested only three separate steps. However all of the evidence suggests that all of the species formed during these steps are quite unstable due to oxidative polymerisation, resulting in the formation of polymerised species. The formation of these polymeric species is not accompanied by any significant spectral changes, which suggests that the benzotroplone ring of theaflavin has been fully degraded. HPLC analysis confirms the disappearance of theaflavin. The enzymatic oxidation was carried out in order to confirm the formation of theanapthoquinone as one of the earlier species formed following degradation of the iron(III) theaflavin complexes. The current work has focussed on the issue of the stability of theaflavin and the nature of the degradation products, under varied conditions. The interaction of iron(III) with theaflavin under acidic conditions results in the formation of a number of products, none of which are stable. It is clearly evident that after short reaction times very little theaflavin remains in the presence of excess iron. All the oxidative species generated chemically are unstable under the extremes of pH and oxidative conditions employed during this work. There has been a recent focus on the investigation of natural alternative metal chelators for chelation therapy, due to the possible harmful side effects of those currently employed [10,11]. Theaflavins fulfil most of the required criteria, although their susceptibility to free iron(III) induced oxidative damage would seemingly render them unsuitable. The pathways leading to the oxidation products through both chemical and enzymatic oxidation are quite similar. The presence of high levels of iron(III) at the beginning and during the oxidation has apparent little interference on the direction of the oxidation pathway.
Acknowledgements M.O’C. and S.B. gratefully acknowledge the financial support of the EC through the Marie-Curie-Fellowship programme (Contract No. HPMI-CT-1999-00040).
References [1] E. Haslam, Phytochemistry 64 (1) (2003) 61–73. [2] T. Ozawa, M. Kataoka, K. Morikawa, O. Negishi, Biosci. Biotech. Biochem. 60 (12) (1996) 2023–2027. [3] N.J. Miller, C. Castelluccio, L. Tijburg, C. RiceEvans, Febs Lett. 392 (1996) 40–44. [4] T.P.J. Mulder, C.J. van Platerink, P.J.W. Schuyl, J.M.M. van Amelsvoort, J. Chromatogr. B 760 (2001) 271–279.
M. O’Coinceanainn et al. / Journal of Inorganic Biochemistry 98 (2004) 657–663 [5] T. Tanaka, I. Kouno, Food Sci. Technol. Res. 9 (2003) 128–133. [6] I.B. Razagui, P.J. Barlow, M.G.A. Izmeth, K.D.A. Taylor, Eur. J. Clin. Nutr. 45 (1991) 331–340. [7] P. Galan, S. Hercberg, Y. Soustre, M.C. Dop, H. Dupin, Hum. Nutr.-Clin. Nutr. 39C (1985) 279–287. [8] W. Doyle, H. Crawley, H. Robert, C.J. Bates, Eur. J. Clin. Nutr. 53 (1999) 552–559. [9] B.V. Roebothan, R.K. Chandra, Int. J. Vit. Nutr. Res. 66 (1996) 66–70. [10] E.H.M. Temme, P.G.A. Van Hoydonck, Eur. J. Clin. Nutr. 56 (2002) 379–386. [11] C. Ghisalberti, European Patent No. EP1074254 A2, 2001. [12] C. Ghisalberti, European Patent No. EP10722265 A1, 2001. [13] M.J. Hynes, M. O’Coinceanainn, J. Inorg. Biochem. 85 (2001) 131–142. [14] A.L. Davis, Y. Cai, A.P. Davies, Magn. Reson. Chem. 33 (1995) 549–552. [15] M. O’Coinceanainn, C. Astill, S. Schumm, Dalton Trans. (2003) 801–806. [16] H. Gampp, M. Maeder, C.J. Meyer, A.D. Z€ uberbuehler, Comment Inorg. Chem. 6 (1987) 41–60.
663
[17] H. Gampp, M. Maeder, C.J. Meyer, A.D. Z€ uberbuhler, Talanta 32 (1985) 1133–1139. [18] H. Gampp, M. Maeder, C.J. Meyer, A.D. Z€ uberbuhler, Talanta 33 (1986) 943–951. [19] H. Gampp, M. Maeder, A.D. Z€ uberbuhler, Trend Anal. Chem. 7 (1988) 147–150. [20] H. Gampp, M. Maeder, A.D. Z€ uberbuhler, Trend Anal. Chem. 7 (1988). [21] S. Bonnely, A.L. Davis, J. Lewis, C. Astill, Food Chem. 83 (2003) 485–492. [22] W. Linert, R.F. Jameson, E. Herlinger, Inorg. Chim. Acta 187 (1991) 239–247. [23] E. Pelizzet, E. Mentasti, E. Pramauro, G. Saini, J. Chem. Soc., Dalton Trans. (1974) 1940–1944. [24] E. Mentasti, E. Pelizzeti, J. Chem. Soc., Dalton Trans. (1973) 2605–2608. [25] E. Mentasti, E. Pelizzeti, G. Saini, J. Chem. Soc., Dalton Trans. (1973) 2609–2614. [26] T. Tanaka, C. Mine, I. Kouno, Tetrahedron 58 (2002) 8851–8856. [27] T. Tanaka, Y. Betsumiya, C. Mine, I. Kouno, Chem. Commun. (2000) 1365–1366.
Inorganic Biochemistry Journal of Inorganic Biochemistry 98 (2004) 657–663 www.elsevier.com/locate/jinorgbio
Reaction of iron(III) with theaflavin: complexation and oxidative products Mairtin O’Coinceanainn a
a,*
, Samuel Bonnely a, Beate Baderschneider a, Michael J. Hynes b
Unilever Research and Development, Colworth Laboratory, Sharnbrook, Bedford MK44 1LQ, UK b Department of Chemistry, National University of Ireland, Galway, Ireland
Received 25 July 2003; received in revised form 20 November 2003; accepted 29 December 2003
Abstract Theaflavins are a family of compounds, whose chemistry has been sparsely investigated. They can comprise up to 40% the dry weight of black tea. They are known to chelate metals, however very little knowledge exists on the mechanisms involved. There is some correlation between both of these areas in that following degradation of the iron theaflavin complex, subsequent redox reactions may lead to the formation of similar products on both occasions. The interaction of iron(III) with theaflavin at pH <3.0 is investigated by means of liquid chromatography mass spectroscopy (LC-MS), stopped flow spectroscopy and multivariate data analysis. Iron theaflavin complexes are formed which subsequently decay to form a number of oxidative species. The difficulties involved in the elucidation of the structure of polymeric phenolic compounds from black tea has been highlighted by numerous authors. The intermediates and major low molecular weight oxidised theaflavin products from the reaction of excess iron with theaflavin have been detected and identified using multivariate data analysis of diode array spectroscopic data. It is not possible to characterise the extremely polar high molecular weight oxidation products obtained from polyphenol oxidation. High performance liquid chromatography (HPLC) and electrospray mass spectroscopy (ES-MS) detected the low molecular weight oxidised theaflavin species present in the system. Enzymatic oxidation of theaflavin using peroxidase (POD) resulted in the formation of one major low molecular weight species oxidative product, which was fully characterised using nuclear magnetic resonance spectroscopy (NMR), high performance liquid chromatography (HPLC), electrospray mass spectroscopy (ES-MS), UV–visible (UV–Vis) and Fourier transform infra-red spectroscopy (FT-IR). The major objective of this work is to investigate the reaction of iron(III) with theaflavin and to add some insight into the mechanistic interaction of iron(III) with this family of compounds. Ó 2004 Elsevier Inc. All rights reserved. Keywords: Theaflavin; Iron(III); Complex; Oxidation
1. Introduction The theaflavins are a family of compounds, which have been sparsely investigated. Theaflavins can become involved in the generation of high molecular weight polyphenolic species during black tea fermentation [1,2]. Numerous health benefits have been attributed to this family of compounds. Their method of antioxidative * Corresponding author. Tel.: +44-1234-222504; fax: +44-1234222844. E-mail address: [email protected] (M. O’Coinceanainn).
0162-0134/$ - see front matter Ó 2004 Elsevier Inc. All rights reserved. doi:10.1016/j.jinorgbio.2003.12.018
behaviour in the body has been investigated by a number of authors [3,4]. Theaflavins interact with certain metal ions and may substantially alter their bioavailability [3]. The antioxidative properties of theaflavin and the gallated theaflavins has been partially ascribed to their ability to scavenge free metals [3], although blood plasma concentrations of theaflavin are extremely low when compared to those of green catechins [4]. It has been more recently suggested that their beneficial effect may be more due to their strong enzymic inhibition capabilities than their antioxidative effects [5]. The levels of theaflavin detected in blood plasma are quite low [4],
658
M. O’Coinceanainn et al. / Journal of Inorganic Biochemistry 98 (2004) 657–663
hence the possible antioxidative effect of theaflavin through the scavenging of metals in the lower pH region of the alimentary canal is of interest. The formation of a variety of metal theaflavin species strongly influences mobility and transport processes and also the bioavailability of metals and ligands [3]. In addition to the beneficial effects, there are possible adverse effects of polyphenolic compounds associated with tea drinking. There is evidence that some tea constituents can have a negative association with measures of iron status in a number of different populations [6–9]. A more recent study has suggested that in the well balanced diet of western populations this effect is negligible [10]. Polyphenols, together with other metal chelators have also been reported to have a positive influence on the treatment of iron overloading in humans [11,12]. There is currently little data available in the literature on the chemistry of the reaction of iron with this family of compounds. This paper reports on the reaction of iron(III) with theaflavin, the parent member of this family and the determination of the resulting products. The work is mainly carried out at low pH due to the inherent problems arising from the hydrolysis of iron and the redox reactions between iron(III) and theaflavin. The interaction of iron with theaflavin under physiological conditions has been previously investigated [3]. The postulated low molecular weight oxidation products are compared to those obtained by enzymatic oxidation carried out on theaflavin under a more controlled environment. The biological importance of this family of compounds is without question, while very little is known of their chemistry, in particular through which they react with metals, iron in particular. It is the aim of this work to contribute to the understanding of the chemistry of this family of compounds.
2. Materials and methods 2.1. High performance liquid chromatography and electrospray ionisation The HPLC system was a Waters Alliance system (model 2690) coupled with a Waters ZMD single quadrupole electrospray ionisation mass spectrometer (ESI) (Micromass, Manchester, UK). The UV detection was done at 280 nm. The separation was performed on a Synergi Hydro column (150 mm 4.6 mm i.d.; particle size, 5 lm) (Phenomenex, Macclesfield, Cheshire, UK) at 30 °C. The flow rate was 0.2 ml/min. The gradient elution was as follows: solvent A, 0.1% formic acid in water; solvent B, 0.1% formic acid in acetonitrile; from 0 to 30 min, 10–100% B; from 30 to 35 min 100% B. The instrument was operated in the positive
and negative electrospray mode with a capillary voltage at 3.2 and 2.8 kV, respectively. Total ion chromatograms were recorded in the full scan mode from m/z 200 to 1000 at cone a voltage of 30 and 25 V in the positive and negative modes respectively. Data were collected between 0 and 35 min. Source temperature was set at 120 °C, desolvation temperature at 400 °C. Nitrogen was used as nebulising gas and drying gas with flow rates of 25 and 500 l/h, respectively. Iron– theaflavin samples were investigated for the presence of metal–polyphenol complexes by flow-injection-MS analysis. The micromass ZMD mass spectrometer was operated using the electrospray ionisation source. The sample solutions were introduced into the electrospray source at a flow rate of 10 ll/min using a syringe pump. The electrospray source conditions were optimised for the generation of the molecular related ion and data were acquired by accumulating several continuum spectra over a mass range of m/z 10–2000. The following instrumental settings were applied: source temperature 100 °C, desolvation temperature 200 °C. Nitrogen was used as nebulising gas and drying gas with flow rates of 25 and 200 l/h, respectively. The capillary voltage was 3.2 kV for positive electrospray and 2.8 kV for negative electrospray ionisation, respectively. The cone voltage was set at 65 and 50 V, respectively. 2.2. Spectroscopy Spectra and kinetic runs for the slowest reactions were obtained as a function of time by use of a Hewlett– Packard 8453A diode array spectrophotometer equipped with a HI-TECH SFA-20 rapid kinetics stopped-flow accessory. All spectra and kinetic runs were recorded at 25 °C. Kinetic data were obtained on a HI-TECh SF20 stopped-flow apparatus, which was interfaced to a PC via a Datalab DL901 transient recorder. The reactant reservoirs were thermostatted at 25 °C. All reagents were of the highest grade available, as detailed elsewhere [13–15]. The FT-IR spectra were recorded from dried powder using a Bio-Rad FTS-6000 spectrometer (Bio-Rad Laboratories, Cambridge, MA, USA) and a Golden-Gate single reflection diamond ATR unit (Specac Inc., Smyrna GA, USA). A spectral resolution of 2 cm1 was used in all measurements and 1000 interferograms were co-added before Fourier transformation. 2.3. NMR spectroscopy 1
H NMR spectra were measured at 400 MHz on a Bruker AMX400 spectrometer using a multinuclear 5 mm inverse probe at 300 K. The solvent used was 2 H2 O and internal reference TSP. HMQC spectra were run at 300 K using the multinuclear 5-mm inverse probe.
M. O’Coinceanainn et al. / Journal of Inorganic Biochemistry 98 (2004) 657–663
2.4. Purification of theaflavin Theaflavin was isolated and purified according to Unilever Bestfoods in-house proceedures as detailed elsewhere [14,15]. 2.5. Global kinetic analysis Spectral-kinetic data were processed as previously described [13] by use of the program SPECFIT/32TM (Spectrum Software Associates, Malborough, MA), which is based on the published works of Zuberb€ uhler et al. [16–20]. 2.6. Oxidation with a tea leaf homogenate [21] The method employed is the subject of a separate paper, but a short summary is as follows. Tea leaves (25 g) were homogenised in a domestic blender (3 10 s) with water (250 ml). The mixture was transferred in a 10 l container and water (4.8 l) was added. The mixture was stirred (using a Heidolph ST1 stirrer, speed 3) and continuously sparged with helium for 2 h. The mixture was filtered and the leaf residue washed a further three times. Washed leaves (6 g) were added to purified theaflavin (100 mg dissolved in 100 ml of water) and the mixture stirred (using a Heidolph ST1 stirrer, speed 2). Peroxidase catalysed oxidation was carried out under helium (to prevent polyphenoloxidase oxidation) and hydrogen peroxide (1.6 ml, 1.5% v/v) was added to the medium. After 1 h oxidation, the mixture was filtered and lyophillised.
3. Results and discussion Theaflavin and its derivatives are known to form complexes with iron and hence have good metal scavenging ability in vitro under physiological conditions [3]. There is however very little theaflavin detected in blood plasma, hence the current investigation is carried out under more acidic conditions in order to investigate the effect of oxidative environment and metals on this biologically significant family of phenols. The reaction of metal ions, Fe3þ in particular, with phenolic-based compounds has been widely investigated, however theaflavin is a much more complicated phenolic compound than previously investigated [22–25]. The general reaction mechanism involves the formation of a metal complex, followed by subsequent decomposition due to differing modes of electron transfer [22–25]. SVD analysis of the spectral data obtained from the reaction of an excess of iron(III) for a period of 10 min with theaflavin indicated the presence of only four eigenvectors in the time and spectral domain corresponding to three different reaction steps as indicated in
659
Fig. 1. The other eigenvectors obtained contained no significant colorometric information related to the reaction. The program SPECFIT/32TM enables the isolation of the spectral data of the major species detected and the elucidation of the major pathways involved. Theaflavin contains two possible binding sites. With the large excess of metal employed the first two reactions could be interpreted as the formation of 1:1 and 2:1 metal to theaflavin complexes on the basis of similar studies in this area [13]. The formation of the iron theaflavin complexes was too fast to be accurately further analysed by conventional diode array spectroscopy, but was detected by fast kinetic analysis using stopped flow spectroscopy with k obs of the order of 3 s1 at 700 nm. Both iron(III) theaflavin complexes subsequently decomposed, resulting in the build up of benzoquinone (kmax ¼ 378 nm). The nature of the redox reaction leading to the decomposition was not further investigated. The spectral data for the second step of the reaction can be partially isolated on the basis of the sharp isosbestic point at 390 nm as indicated by the spectral data in Fig. 2. The formation of the benzoquinone moiety is quite consistent with the enzymatic oxidative work carried out by Tanaka [5]. The benzoquinone is unstable in the oxidative environment and rearranges to a further product as indicated in Fig. 3 with a kmax ¼ 440 nm. Tanaka has dealt in detail with the mechanism involved during this rearrangement [5]. This complete reaction of iron(III) with theaflavin in solution was further investigated using HPLC and ES-MS Fig. 4. The theaflavin peak (RT 14.48 min) and the signal of the oxidation product (RT 14.16 min) were not resolved in the total ion chromatogram due to the slow scan speed of the mass spectrometer, but this is possible through the use of UV detection at 280 nm. Two clearly separated signals for the two compounds can be obtained by selective ion detection Fig. 5. The main oxidative species detected in the final solution after 20 min had a molecular weight (M + Hþ ) of 535 with a retention time of 14.15 min. A species with the same molecular weight and spectral characteristics (kmax ¼ 440 nm) (theanapthoquinone) was obtained by Tanaka following the rearrangement of benzoquinone. Further support for the assignment of this species arises during the enzymatic work. Evolving factor analysis of the spectral data obtained from the reaction of iron(III) with theaflavin indicates the evolution of only four significant species over time domains of 10 min or less (see Fig. 6). These correspond to the original theaflavin, metal–theaflavin complex, theabenzoquinone, and theanaphthoquinone. No other significant species appear after the first 30 s of the reaction from spectroscopic data. Two further minor species both of molecular weight 578, with retention times of 9.14 and 11.59 min,
660
M. O’Coinceanainn et al. / Journal of Inorganic Biochemistry 98 (2004) 657–663
Fig. 1. (a) First four temporal eigenvectors from U S (x-axis is s). (b) Four spectral eigenvectors V (x-axis is wavelength). Iron(III) ¼ 2 mM, theaflavin ¼ 0.2 mM, [Hþ ] ¼ 0.01 M. This represents three separate spectral data transitions or three individual reactions.
respectively, were detected by MS. These two species were not clearly detected as ‘real species’ from UV spectroscopy, however the chromatogram ionisation peaks for these species are quite strong, thus enabling their detection through this technique. The structures of these compounds can only be postulated on the basis of mass spectral data, however this together with strong literature support enables the confident assignment of
structures. The first of these can be tentatively assigned as dehydrotheaflavin on the basis of literature comparison with the results of Tanaka [26] (product C in Fig. 3). The second minor component could also be assigned tentatively as another structural analogue of dehydrotheaflavin (product B). Tanaka has previously detected dehydrotheaflavin as an oxidation product of theaflavins during the black tea fermentation process
M. O’Coinceanainn et al. / Journal of Inorganic Biochemistry 98 (2004) 657–663
661
Fig. 4. Total ion chromatogram of the reaction of iron(III) with theaflavin. Iron(III) ¼ 2 mM, theaflavin ¼ 0.2 mM, [Hþ ] ¼ 0.01 M. Fig. 2. Spectral data obtained for the formation of theabenzoquinone following decomposition of the iron theaflavin complex. Iron(III) ¼ 2 mM, theaflavin ¼ 0.2 mM, [Hþ ] ¼ 0.01 M.
under controlled conditions [24], however the structural analogue was not detected. It is not possible to obtain a complete mass balance of this system due to the longstanding inability to characterise the high molecular weight phenolic component of black tea. We can however ascertain from HPLC analysis that following completion of the reaction there is only 5% of the original theaflavin starting material remaining. The other species are also only present in low levels, which are difficult to quantify. It has not been possible to characterise the polymeric phenolic component, which is the major oxidative product of the reaction [2]. Potentiometric analysis was carried out on the interaction of iron(III) with an excess of theaflavin. The oxidative reactions and the formation of theanaphthoquinone rendered it quite difficult to obtain reliable stability constant values, a problem which is prevalent with iron(III) polyphenol systems. Samples were withdrawn from the potentiometric analysis at pH’s 3.50, R
OH
R
O
OH
2 Fe +
O
R
1
OH
O
O
O
R
4.50 and 5.50. Analysis of samples which were 1 mM theaflavin to 0.1–0.3 mM iron(III) using the ES-MS method above, indicated the presence of equivalent levels of only theanapthoquinone over the three pH regions. The levels present were however linearly dependent on the levels of iron(III) initially employed.
R
Fe
3+
Fig. 5. Chromatogram of the reaction of iron(III) with theaflavin with selective ion detection.
Fe
O
Theabenzoquinone High molecular weight oxidation products
O 2
3
O
R OH
Theaflavin
O H
R
O HO
HO R OH
O
O
HO
OH
HO O
O O
O
HO
C
OH
O
O OH
OH
HO R=
O
OH OH
O A Theanaphthoquinone
OH
OH
O HO
OH
O
OH
OH
B
Fig. 3. Mechanism of reaction of iron(III) with theaflavin and the subsequent products.
662
M. O’Coinceanainn et al. / Journal of Inorganic Biochemistry 98 (2004) 657–663
Fig. 6. Evolving factor analysis of the reaction of iron with theaflavin. Only four significant eigenvectors corresponding to theaflavin, iron complex, theabenzoquinone and theanaphthoquinone. Iron(III) ¼ 2 mM, theaflavin ¼ 0.2 mM, [Hþ ] ¼ 0.01 M.
It is well known that enzymatic oxidation of theaflavin is difficult to achieve with polyphenol oxidase while peroxidase is described to be an efficient oxidative enzyme for this substrate. Enzymatic oxidation of theaflavin under more controlled conditions than the harsh chemical treatment, is quite useful in the elucidation of the mechanistic interactions occurring in the harsher environment. Enzymatic oxidation was carried out on pure theaflavin as indicated in the materials and method section, and the major low molecular weight product was characterised. FT-IR spectra of the theaflavin before and after 1 h peroxidase oxidation indicates the formation of an extra carbonyl peak at 1670 cm1 corresponding to an additional carbonyl group. HPLC analysis indicated the formation of one major oxidative product with a retention time of 14.22 min and kmax ¼ 440 nm. ES-MS confirmed a molecular ion peak of (535 M + Hþ ). 1 H NMR and 13 C analysis confirmed the product as theanaphthoquinone [27]. The characterisation under these conditions is unequivocal, hence arising from the retention time and mass spectrum of this product being the same as that obtained for the major product of the interaction of iron(III) with theaflavin, it can be employed as a standard to confirm the assignment of the product obtained in the iron(II) system on the basis of retention time and spectra obtained. It is noteworthy that the main oxidation product obtained under the different systems of oxidation, theanaphthoquinone, unlike most other phenolic products is quite stable under extremes of pH and oxidative environment employed under the two different environments.
4. Conclusion Polyphenols are generally quite unstable under oxidative conditions, degrading to form products of
varying molecular weights. Spectroscopic data suggest that earlier part of the reaction of the reaction of iron(III) with theaflavin is not complicated as illustrated by stopped flow spectroscopic data and multivariate analysis of spectroscopic data under the conditions investigated suggested only three separate steps. However all of the evidence suggests that all of the species formed during these steps are quite unstable due to oxidative polymerisation, resulting in the formation of polymerised species. The formation of these polymeric species is not accompanied by any significant spectral changes, which suggests that the benzotroplone ring of theaflavin has been fully degraded. HPLC analysis confirms the disappearance of theaflavin. The enzymatic oxidation was carried out in order to confirm the formation of theanapthoquinone as one of the earlier species formed following degradation of the iron(III) theaflavin complexes. The current work has focussed on the issue of the stability of theaflavin and the nature of the degradation products, under varied conditions. The interaction of iron(III) with theaflavin under acidic conditions results in the formation of a number of products, none of which are stable. It is clearly evident that after short reaction times very little theaflavin remains in the presence of excess iron. All the oxidative species generated chemically are unstable under the extremes of pH and oxidative conditions employed during this work. There has been a recent focus on the investigation of natural alternative metal chelators for chelation therapy, due to the possible harmful side effects of those currently employed [10,11]. Theaflavins fulfil most of the required criteria, although their susceptibility to free iron(III) induced oxidative damage would seemingly render them unsuitable. The pathways leading to the oxidation products through both chemical and enzymatic oxidation are quite similar. The presence of high levels of iron(III) at the beginning and during the oxidation has apparent little interference on the direction of the oxidation pathway.
Acknowledgements M.O’C. and S.B. gratefully acknowledge the financial support of the EC through the Marie-Curie-Fellowship programme (Contract No. HPMI-CT-1999-00040).
References [1] E. Haslam, Phytochemistry 64 (1) (2003) 61–73. [2] T. Ozawa, M. Kataoka, K. Morikawa, O. Negishi, Biosci. Biotech. Biochem. 60 (12) (1996) 2023–2027. [3] N.J. Miller, C. Castelluccio, L. Tijburg, C. RiceEvans, Febs Lett. 392 (1996) 40–44. [4] T.P.J. Mulder, C.J. van Platerink, P.J.W. Schuyl, J.M.M. van Amelsvoort, J. Chromatogr. B 760 (2001) 271–279.
M. O’Coinceanainn et al. / Journal of Inorganic Biochemistry 98 (2004) 657–663 [5] T. Tanaka, I. Kouno, Food Sci. Technol. Res. 9 (2003) 128–133. [6] I.B. Razagui, P.J. Barlow, M.G.A. Izmeth, K.D.A. Taylor, Eur. J. Clin. Nutr. 45 (1991) 331–340. [7] P. Galan, S. Hercberg, Y. Soustre, M.C. Dop, H. Dupin, Hum. Nutr.-Clin. Nutr. 39C (1985) 279–287. [8] W. Doyle, H. Crawley, H. Robert, C.J. Bates, Eur. J. Clin. Nutr. 53 (1999) 552–559. [9] B.V. Roebothan, R.K. Chandra, Int. J. Vit. Nutr. Res. 66 (1996) 66–70. [10] E.H.M. Temme, P.G.A. Van Hoydonck, Eur. J. Clin. Nutr. 56 (2002) 379–386. [11] C. Ghisalberti, European Patent No. EP1074254 A2, 2001. [12] C. Ghisalberti, European Patent No. EP10722265 A1, 2001. [13] M.J. Hynes, M. O’Coinceanainn, J. Inorg. Biochem. 85 (2001) 131–142. [14] A.L. Davis, Y. Cai, A.P. Davies, Magn. Reson. Chem. 33 (1995) 549–552. [15] M. O’Coinceanainn, C. Astill, S. Schumm, Dalton Trans. (2003) 801–806. [16] H. Gampp, M. Maeder, C.J. Meyer, A.D. Z€ uberbuehler, Comment Inorg. Chem. 6 (1987) 41–60.
663
[17] H. Gampp, M. Maeder, C.J. Meyer, A.D. Z€ uberbuhler, Talanta 32 (1985) 1133–1139. [18] H. Gampp, M. Maeder, C.J. Meyer, A.D. Z€ uberbuhler, Talanta 33 (1986) 943–951. [19] H. Gampp, M. Maeder, A.D. Z€ uberbuhler, Trend Anal. Chem. 7 (1988) 147–150. [20] H. Gampp, M. Maeder, A.D. Z€ uberbuhler, Trend Anal. Chem. 7 (1988). [21] S. Bonnely, A.L. Davis, J. Lewis, C. Astill, Food Chem. 83 (2003) 485–492. [22] W. Linert, R.F. Jameson, E. Herlinger, Inorg. Chim. Acta 187 (1991) 239–247. [23] E. Pelizzet, E. Mentasti, E. Pramauro, G. Saini, J. Chem. Soc., Dalton Trans. (1974) 1940–1944. [24] E. Mentasti, E. Pelizzeti, J. Chem. Soc., Dalton Trans. (1973) 2605–2608. [25] E. Mentasti, E. Pelizzeti, G. Saini, J. Chem. Soc., Dalton Trans. (1973) 2609–2614. [26] T. Tanaka, C. Mine, I. Kouno, Tetrahedron 58 (2002) 8851–8856. [27] T. Tanaka, Y. Betsumiya, C. Mine, I. Kouno, Chem. Commun. (2000) 1365–1366.