Detection of biologically relevant anions by fluorescence and NIR molecular probes

.. .DETECTION OF BIOLOGICALLY RELEVANT ANIONS BY FLUORESCENCE AND NIR MOLECULAR PROBES QUEK YI LING (B Sc.(Hons.), National University of Singapore) A THESIS SUBMITTED FOR THE DEGREE OF DOCTOR OF. .. Results and discussion .140 6.3.1 Synthesis of complexes and 140 6.3.2 Spectroscopic characterization of complexes and .142 6.3.3 Comparison of NIR band shift for [Fe2]4+, 1, and ... convenient way for the detection of the HS− generation rate of a H2S donor of medical importance The addition of π-acceptor ligands such as cyanide and isocyanide ligands to the NIR active isovalent

DETECTION OF BIOLOGICALLY RELEVANT ANIONS BY FLUORESCENCE AND NIR MOLECULAR PROBES QUEK YI LING NATIONAL UNIVERSITY OF SINGAPORE 2011 DETECTION OF BIOLOGICALLY RELEVANT ANIONS BY FLUORESCENCE AND NIR MOLECULAR PROBES QUEK YI LING (B. Sc.(Hons.), National University of Singapore) A THESIS SUBMITTED FOR THE DEGREE OF DOCTOR OF PHILOSOPHY DEPARTMENT OF CHEMISTRY NATIONAL UNIVERSITY OF SINGAPORE 2011 Acknowledgements First and foremost, I wish to express my deepest gratitude to my supervisor Prof Huang Dejian for his valuable supervision, patient guidance and encouragement given throughout the project. I feel honoured to have him as my supervisor and have learnt from him his invaluable ideas, profound knowledge, and rich research experience. Without his encouragement, I would not have embarked on my PhD studies. I am also deeply grateful to his kindness throughout the four years and his efforts of guiding me to complete my research project and thesis. I also wish to extend my sincere gratitude to NUS for research scholarship and Singapore Ministry of Education (Grant No. R-143-000-299-112) and Science and Engineering Research Council of the Agency for Science, Technology and Research (A*Star) of Singapore (Grant No. 072-101-0015) for financial support in the project. Next, I would like to express my sincere gratitude to Ms Tan Ying Ying and Ms Wenie Chin, my UROPs and honours year students respectively, for their contribution in determining the pro-oxidant activity of tea leaves, tea catechins and doing the DNA cleavage experiments. I also wish to thank Prof Li Tianhu and Dr Wang Yifan for their knowledge and expertise in the DNA cleavage assay setup. I would also like to express my heartfelt appreciation to Dr Wang Suhua, Dr Yao Wei, Dr Feng Shengbao, Dr Viduranga Yashasvi Waisundara, Dr Koh Lee Wah, Dr Fu Caili, and Ms Chen Wei for their advice and technical assistance rendered throughout the project. In addition, I would like to express i my appreciation to Ms Lee Chooi Lan, Ms Lew Huey Lee, Ms Jiang Xiaohui, and Mr Abdul Rahman bin Mohd Noor for their excellent technical support. I am also grateful to NUS Chemical, Molecular and Materials Analysis Centre (CMMAC) for its technical assistance. Last but not least, I would like to thank my family members for their love and unwavering support. I wish to express my sincere appreciation to my boyfriend for his constant encouragement during the process of my thesis writing. I wish to sincerely thank my mentor in life Dr Daisaku Ikeda, my cousin, my friends, and my comrades in Singapore Soka Association for their encouragement and continuous support during the four years of my PhD studies. From the bottom of my heart, I thank all my friends whose name may not have been mentioned here, but have never hesitated to give me their helping hands whenever I need help. Quek Yi Ling August 2011 ii Table of Contents page Summary...........................................................................................................x List of Figures...........................................................................................xii List of Tables.................................................................................................xvii List of Abbreviations...................................................................................xviii Part I: Hydroethidine as a Fluorescent Probe for Quantifying Pro-oxidant Activity of Polyphenolic Compounds Chapter 1 Introduction on Pro-oxidants 1.1. Reactive oxygen species and oxidative stress……………....................1 1.2. Structure and antioxidant activity of flavonoids.....................................2 1.2.1 Chemical structure of flavonoids.....................................................3 1.2.2 Antioxidant activity of flavonoids...................................................4 1.3 Pro-oxidant activity of flavonoids……………………………………...8 1.3.1 Pro-oxidant activity in the absence of transition metals..................9 1.3.2 Pro-oxidant activity in the presence of transition metals or peroxidases……………………………………………………………10 1.3.3 Pro-oxidant activity in terms of DNA damage and lipid peroxidation……………………………………………………………13 1.3.4 Pro-oxidant activity in terms of enzyme and topoisomerase inhibitors.................................................................................................14 1.3.5 Pro-oxidant activity in terms of cancer therapy.............................16 1.4. Pro-oxidant activity assays...................................................................17 1.4.1 Deoxyribose assay.........................................................................17 iii 1.4.2 Other pro-oxidant assays…………………………………………18 1.5 Detection methods of superoxide..........................................................21 1.5.1 Spectrophotometric probes............................................................22 1.5.2 Fluorescent probes ………………………………………………23 1.5.3 Luminescence probes.....................................................................25 1.5.4 Electron spin resonance and spin trapping ………………………27 1.6 Aim of this research...............................................................................28 Chapter 2 Pro-oxidant Activity of Flavonols Quantified by a Fluorescent Probe Hydroethidine 2.1 Introduction............................................................................................30 2.2 Materials and methods..........................................................................31 2.2.1 Materials........................................................................................31 2.2.2 Instruments....................................................................................32 2.2.3 Preparation of stock solutions........................................................32 2.2.4 HPLC analysis of oxidation product of HE with flavonols..........33 2.2.5 Reaction between HE, flavonol and potassium superoxide...........34 2.2.6 Pro-oxidant assay procedure..........................................................34 2.2.7 UV-vis kinetics measurement of HE oxidation by myricetin.......35 2.2.8 pH dependency of HE oxidation by myricetin..............................36 2.2.9 Myricetin decomposition in the presence and absence of HE.......36 2.2.10 Determination of acid dissociation constants (pKa) of flavonols.37 2.2.11 Measurement of oxidation potentials of flavonols.......................37 2.2.12 Determination of DNA cleavage.................................................38 2.3 Results and discussion...........................................................................39 2.3.1 HE oxidation by myricetin.............................................................39 iv 2.3.2 Quantification of pro-oxidant activity of flavonols.......................45 2.3.3 DNA cleavage activity of flavonols......................................47 2.4 Conclusion.............................................................................................51 Chapter 3 Evaluation of Pro-oxidant Activity of Different Tea Leaves 3.1 Introduction...........................................................................................53 3.1.1 Tea processing and its polyphenol composition............................53 3.1.2 Health effects of tea.......................................................................55 3.1.3 Pro-oxidant activity of tea..............................................................56 3.1.4 Aims & objectives..........................................................................57 3.2 Materials and methods.........................................................................59 3.2.1 Materials........................................................................................59 3.2.2 Instruments....................................................................................60 3.2.3 Preparation of stock solutions........................................................60 3.2.4 Extraction of tea samples...............................................................60 3.2.5 Pro-oxidant assay procedure…………………………….............61 3.2.6 Quantification of polyphenols in tea samples................................61 3.2.7 Total phenolic assay procedure......................................................62 3.2.8 Oxidation product of HE with tea catechins and theaflavins.........63 3.2.9 Determination of acid dissociation constants (pK a ) of tea catechins……………………………………………………………….63 3.2.10 Measurement of oxidation potentials of tea catechins.................63 3.2.11 Determination of DNA cleavage.................................................63 3.3 Results and discussion...........................................................................64 3.3.1 Oxidation product of HE with tea catechins and theaflavins….....64 v 3.3.2 Quantification of pro-oxidant activity of tea catechins, theaflavins, gallic acid, methyl gallate, pyrogallol, and tea samples….....................64 3.3.3 Quantification of major polyphenols in tea extracts......................68 3.3.4 Effect of pH on pro-oxidant activity of EGCG, theaflavins and tea extracts…………………………………………………........................77 3.3.5 Redox potential and acid dissociation constants of tea catechins..79 3.3.6 DNA damage induced by tea catechins……………….................83 3.4 Conclusion.............................................................................................87 Part II: Bimetallic Complexes of Ruthenium and Iron as Near-IR Probes for Detection of Redox-Active Molecules Chapter 4 Literature Review on NIR Active Bimetallic Complexes of Ru and Fe 4.1 Introduction on NIR active metal complexes........................................89 4.2 NIR absorption by metal complexes containing radical ligands...........90 4.2.1 Iron(II)-2,2’-bipyridine complexes…............................................91 4.2.2 Ruthenium(II) dioxolene complexes………..................................92 4.3 NIR absorption by mixed-valence dinuclear complexes……...............96 4.3.1 Classification of mixed-valence dinuclear complexes...................96 4.3.2 Physical properties of mixed-valence complexes..........................98 4.3.3 Mixed-valence complexes with bis-monodentate bridging ligands…………………………………………………………….........99 4.3.4 Mixed-valence complexes with bis-bidentate bridging ligands...101 4.3.5 Mixed-valence complexes with bis-tridentate bridging ligands..102 4.3.6 Mixed-valence complexes with bis-tetradentate bridging ligands…………………………………………………………….......103 vi 4.4 NIR absorption from mixed-valency of coordinated radical ligands..108 4.4.1 Bis(-iminopyridine)iron(II) complexes.....................................108 4.4.2 Bis(dithiolene)iron(III) complex………………………………109 4.5 Applications of NIR active dinuclear complexes…………………....110 4.5.1 Application in electro-optic switching…………………….........110 4.6 Aim of this research.............................................................................112 Chapter 5 Air Oxidation of HS- Catalyzed by a Mixed-Valence Diruthenium Complex, a Near-IR Probe for HS- Detection 5.1. Introduction.........................................................................................114 5.2 Materials and methods.........................................................................115 5.2.1 Materials......................................................................................115 5.2.2 Instruments...................................................................................116 5.2.3 Preparation of stock solutions......................................................117 5.2.4 Construction of standard curve of NaHS with [Ru2]+…............117 5.2.5 Reaction of NaHS with [Ru2]+ under a nitrogen atmosphere…..118 5.2.6 Reusability of [Ru2]+ for HS− quantification….........................118 5.2.7 Oxidation of HS− catalyzed by 5% [Ru2]+.................................119 5.2.8 Measurement of H2O2 formed from HS− oxidation...................119 5.2.9 Extraction of HS2− from HS− oxidation with 5% [Ru2]+......…120 5.2.10 Selectivity of [Ru2]+ towards anions and reductants................120 5.2.11 Replacement of axial Cl− ligands of [Ru2]+ with F−................121 5.2.12 Method validation......................................................................121 5.2.13 Measurement of H2S release from GYY4137 using [Ru2]+…..122 5.3 Results and discussion.........................................................................122 5.3.1 Sensitivity of [Ru2]+ towards HS−.............................................122 vii 5.3.2 Reversibility of the reaction of [Ru2]+ with HS− in the presence of oxygen...................................................................................................123 5.3.3 Reusability of [Ru2]+ for HS− quantification.............................124 5.3.4 Oxidation of HS− catalyzed by 5% [Ru2]+ ………….…............125 5.3.5 Selectivity of [Ru2]+ towards anions and reductants………....130 5.3.6 Method validation…………………………….........................131 5.3.7 Measurement of H2S release from GYY4137 using [Ru2]+…...132 5.4 Conclusion...........................................................................................134 Chapter 6 Synthesis and Characterization of NIR Active Diiron Complexes and Their Reactivity with Redox-Active Molecules 6.1 Introduction..........................................................................................135 6.2 Materials and methods.........................................................................135 6.2.1 Materials......................................................................................135 6.2.2 Instruments...................................................................................136 6.2.3 Preparation of stock solutions......................................................136 6.2.4 Synthesis of Fe 2 TIED(CN) 4 and [Fe 2 TIED(RNC) 4 ] 4 + complexes…………………………………………………………….137 6.2.5 Procedures for sensing redox-active molecules……...................139 6.3 Results and discussion.........................................................................140 6.3.1 Synthesis of complexes 1 and 2...................................................140 6.3.2 Spectroscopic characterization of complexes 1 and 2.................142 6.3.3 Comparison of NIR band shift for [Fe2]4+, 1, and 2...................153 6.3.4 Reactivity of complexes 1 and 2 with redox-active molecules…154 6.4 Conclusion...........................................................................................157 6.5 Future work..........................................................................................158 viii List of Publications and Patent...................................................................159 References.....................................................................................................160 Appendices (CD attached) ix Summary The first part of the thesis documented the efforts to establish a convenient assay making use of a fluorescent probe hydroethidine (HE) to quantify the superoxide radical forming pro-oxidant activity of flavonols under physiologically relevant conditions. In the presence of the flavonols myricetin and quercetin, oxidation of HE by superoxide yielded ethidium instead of 2hydroxyethidium. The reaction is inhibited by added superoxide dismutase, suggesting that superoxide is involved in the rate limiting step of the oxidation. The superoxide formation rates were quantified from the oxidation kinetics and myricetin was found to have the highest pro-oxidant activity. This assay was then applied in quantifying the pro-oxidant activity of different tea leaves. The pro-oxidant activity decreases in the order of black tea > oolong tea > green tea. Hence there is evidence that the pro-oxidant activity of the tea leaves at pH 7.40 increases with the degree of fermentation. The major tea catechins, theaflavins, and gallic acid present in the tea samples were quantified using reverse phase-high performance liquid chromatography. Theaflavins are found to be better pro-oxidants than tea catechins. However, the total pro-oxidant activities of each tea sample correlate poorly with the sum of the weighted pro-oxidant activities of the phenolic compounds quantified in the tea extract. In the presence of Cu(II), the DNA damaging pro-oxidant activity is high for myricetin under a wide range of concentrations, whereas for quercetin and tea catechins, they cause DNA cleavage at low concentration but suppress DNA cleavage at high concentration. Our results illustrated the dual roles of polyphenolic compounds as pro-oxidants and antioxidants. x The second part of the thesis examined the versatile chemical reactivity of near infrared (NIR) active bimetallic complexes of ruthenium and iron, which includes redox reaction and ligand substitution, for sensing application. The NIR active mixed-valence diruthenium complex, [Ru2TIEDCl4]Cl, where TIED = tetraiminoethylenedimacrocycle, was found to be a highly active catalyst for air oxidation of HS−, forming hydrogen peroxide, disulfane, and elemental sulfur. The NIR probe was selective towards HS− and did not react with other common biological anions. It provides a convenient way for the detection of the HS− generation rate of a H2S donor of medical importance. The addition of π-acceptor ligands such as cyanide and isocyanide ligands to the NIR active isovalent diiron complex, [Fe2TIED(CH3CN)4]4+, was able to replace the axial acetonitrile ligands to form two novel, water-soluble complexes, neutral [Fe2TIED(CN)4] (1) and cationic [Fe2TIED(C5H11NC)4]4+ (2), which were characterized by UV-vis, IR, ESI, and NMR spectroscopy. Both complexes contain a strong π-acidic ligand, but the NIR absorption band differs by 129 nm. We believe it is due to the charge on 2. 1 did not show any reactivity with ROS and HS−, while 2 exhibited good selectivity for HS− and Angeli’s Salt (NO− donor). However, because of its poor sensitivity, 2 was not suitable for use as a molecular probe for sensing HS− and Angeli’s Salt. The combination of the NIR spectra of the two complexes covers a board range from 803 nm to 932 nm. This may be of use as NIR materials in future applications. xi List of Figures Figure 1.1 Basic structure of flavonoid, 2-phenylbenzopyran. Figure 1.2 Subclasses of flavonoids. Figure 1.3 Chemical structures of major polyphenolic catechins present in green tea. Figure 1.4 Metal chelating sites in flavonoids. Figure 1.5 Pro-oxidant mechanism of catechol-type flavonoids, (A) luteolin and (B) quercetin, with GSH. Figure 1.6 CL mechanism of the reaction of luminol with O2●−. Figure 1.7 CL mechanism of the reaction of lucigenin with O2●−. Figure 2.1 Chemical structure of the flavonols used. Figure 2.2 HPLC chromatograms of E+, potassium superoxide induced oxidation of HE and the product formed from oxidation of HE by myricetin. Figure 2.3 Oxidation products of HE by myricetin, KO2, and KO2 in the presence of flavonol (myricetin, quercetin). Figure 2.4 (A) Kinetic traces with [HE] = 25.6 µM and various concentrations of myricetin. (B) Standard calibration curve of fluorescence intensity vs E+ concentration. Figure 2.5 Kinetic traces of E+ formed from the reaction of [HE] = 25.6 µM with various concentrations of myricetin. (B) Oxidation rate of HE as a function of myricetin concentration in the absence and presence of SOD. Figure 2.6 (A) Normalized absorbance of myricetin at 392 nm in the absence and presence of HE (30 µM) in phosphate buffer (pH 7.4); initial [myricetin] = 7.5 µM. (B) UV-vis kinetic measurements at 479 nm during the reaction of HE (30 xii µM) with variable myricetin over 2 hours. Figure 2.7 (A) pH dependency of HE oxidation by myricetin in buffered solution at 37 °C. (B) pKa curve showing the variation of absorption maximum of myricetin at 324 nm with pH. Figure 2.8 Proposed reaction mechanism of HE oxidation in the absence and presence of myricetin. Figure 2.9 (Top) Agarose gel electrophoretic analysis of pBR322 DNA damage induced by myricetin in the presence of Cu(II). (Bottom) Flavonol concentration dependent DNA damage in the presence of Cu(II). Figure 3.1 Chemical structures of major theaflavins and other oxidative products present in oolong and black tea. Figure 3.2 Calibration curves of GA, ECG, EGCG, EC, and EGC. Inset shows the calibration curves of TF, TFMG-a, TFMG-b, and TFDG. Figure 3.3 HPLC chromatograms of the tea extracts (A) Tea 2: Gold Kili Green Tea, (B) Tea 7: China Fujian Tie Guan Yin and (C) Tea 12: Roma English Breakfast Tea. Figure 3.4 Relation of the total radical generation rate of the tea extracts with the radical generation rate due to the quantified polyphenolic compounds by HPLC. Figure 3.5 Relation between the pro-oxidant activity and TPC of the tea extracts. Figure 3.6 pH dependency of HE oxidation by EGCG in buffered solution at 37 °C. Figure 3.7 Cyclic voltammograms for (A) EGCG, (B) EGC, (C) ECG, and (D) EC at pH 5.50 and pH 7.40. xiii Figure 3.8 Agarose gel electrophoretic analysis of pBR322 DNA damage induced by catechins. Figure 3.9 DNA damage induced by catechins in the presence of Cu(II). Figure 3.10 Catechin concentration dependent DNA damage in the presence of Cu(II). Figure 3.11 DNA damage induced by (+)-catechin and EC in the absence and presence of Cu(II) in pH 7.4 phosphate buffer (10 mM) at 37 °C for 2 hours under aerobic condition. Figure 4.1 NIR absorption of radical ions. Figure 4.2 Redox reactions of catecholate, semiquinone and quinone. Figure 4.3 Four reversible redox interconversions of 1n+ (n = 0-4). Figure 4.4 Electronic spectra of [Ru(bpy)2(sq)]+ and [1]2+. Figure 4.5 Electronic spectra of [2]n+ in acetonitrile, the numbers 1, 2, 3, 4 refer to the charge n+. Figure 4.6 IVCT and IC (interconfigurational) transitions in d6-d5 mixed-valence systems such as RuII-L-RuIII complexes. Figure 4.7 Structure of bimetallic tetraiminoethylenedimacrocycle, (M2TIEDL4)n+, formula M2N8C20H36L4, where L = Cl−, CH3CN or DMF and M = Ru or Fe. Figure 4.8 Absorption spectral changes of the binuclear mixedvalence ruthenium complexes 14-17 as they undergo redox reactions in aqueous solution. Figure 5.1 Absorption spectra of [Ru2]+ (16 µM) in a 50 mM TrisHCl buffer (pH 7.40) recorded 1 minute after reactions with various concentrations of HS−. xiv Figure 5.2 Absorption spectra of [Ru2]+ (16 µM) in a 50 mM Tris-HCl buffer (pH 7.40) before and after the addition of HS− (16 µM) under a N2 atmosphere with time and upon exposure to air after 42 minutes. Figure 5.3 Catalytic circles of [Ru2]+ in the HS−/O2 system observed by the absorbance ratio changes recorded in real time of [Ru2]+ (16 µM) in a 50 mM Tris-HCl buffer (pH 7.40) after the first addition of HS− (8 µM) and the second, third, and fourth additions of HS− (8, 16, and 32 µM). Figure 5.4 Absorption spectra of HS− with 5% [Ru2]+ in 50 mM TrisHCl buffer pH 7.40 with time. Figure 5.5 Standard curve of H2O2 with FOX reagent. Figure 5.6 Concentration of H2O2 generated from HS− oxidation catalyzed by 5% [Ru2]+ with time. Figure 5.7 Absorption spectra of HS− with 5% [Ru2]+ (30 minutes), NaHS2 and Na2SO3 in 50 mM Tris-HCl buffer pH 7.40. Figure 5.8 Absorption spectra of NaHS2 and HS2− extract in DCM. Figure 5.9 Absorbance ratio changes of [Ru2]+ (16 µM) in a 50 mM Tris-HCl buffer (pH 7.40) recorded 1 minute after the addition of various anions (1600 µM) and reductants (16 µM). Figure 5.10 (A) Standard curve of HS− with DTNB (247 µM). (B) Standard curve of HS− with [Ru2]+ (16 µM). Figure 5.11 Release of H2S from GYY4137 (200 µM) in 50 mM Tris-HCl buffer (pH 7.40) incubated at 37 °C under a N2 atmosphere as determined spectrophotometrically with the use of [Ru2]+ (16 µM) recorded after 1 minute. Figure 6.1 Absorption spectra of [Fe2]4+ (21 µM) in H2O recorded 5 minutes after reactions with various concentrations of with KCN (84, 210, and 420 µM). xv Figure 6.2 Absorption spectra of [Fe2]4+ (21 µM) in CH3CN and [Fe2TIED(CN)4] (21 µM) in MeOH, EtOH, H2O. Figure 6.3 Absorption spectra of [Fe2]4+ (21 µM) in CH3CN and [Fe2TIED(C5H11NC)4]4+ (21 µM) in MeOH, and H2O. Figure 6.4 IR spectrum of 1 recorded as a KBr disc. Figure 6.5 IR spectrum of 2 recorded as a KBr disc. Figure 6.6 ESI-MS spectrum of 1 in deionized H2O recorded in the cationic mode. Figure 6.7 ESI-MS spectrum of 2 in MeOH recorded in the cationic mode. Figure 6.8 1 H NMR spectrum of 1 in CD3OD. Figure 6.9 1 H NMR spectrum of 2 in CD3OD. Figure 6.10 Absorption spectra of [Fe2]4+ (21 µM) in CH3CN, [Fe2TIED(CN)4] (21 µM) in H2O, and [Fe2TIED(C5H11)4]4+ (21 µM) in H2O. Figure 6.11 Absorbance changes of 2 (21 µM) in a 50 mM Tris-HCl buffer (pH 7.40) recorded 30 minutes after the addition of various ROS, CN−, HS− (420 µM), and NO−, NO (84 µM). Figure 6.12 Absorption spectra of 2 (21 µM) with HS− (420 µM) in a 50 mM Tris-HCl buffer (pH 7.40) with time. xvi List of Tables Table 2.1 Measured pKa1, oxidation potentials and pseudo-first order rate constants of oxidation of flavonols by oxygen at 37 °C, pH 7.40. Table 3.1 Name and origin of the tea samples. Table 3.2 Pro-oxidant activity of tea samples at pH 7.40 in terms of pseudo-first order rate constant k′. Table 3.3 Pro-oxidant activity of tea catechins, theaflavins, gallic acid, methyl gallate and pyrogallol at pH 7.40 in terms of pseudo-first order rate constant k′. Table 3.4 Composition of the tea extracts expressed as percentage of the dry weight of the tea leaves. Table 3.5 Concentration of phenolic compounds (g/L) in the tea extracts and k′ of the tea extracts. Table 3.6 Radical generation rate due to each phenolic compound, the total radical generation rate of the tea extract and the percentage of unaccounted radical generation rate calculated based on 16 g/L of tea extract. Table 3.7 Comparison of the Gallic Acid Equivalents (GAE) to give an estimation of the percentage of polyphenols unquantified in the HPLC quantification compared to the total phenolic assay. Table 3.8 Comparison of pro-oxidant activities of theaflavins at pH 7.40 and pH 5.60. Table 3.9 Measured pKa1 and oxidation potentials of tea catechins. Table 5.1 Accuracy of DTNB assay. Table 5.2 Accuracy of [Ru2]+ assay. xvii List of Abbreviations Abbreviation Description ADP adenosine-diphosphate ATP adenosine-5'-triphosphate BHT butylated hydroxytoluene CHD coronary heart disease CL chemiluminescence CYP cytochrome p 450 DCF dichlorofluorescein DCM dichloromethane DEA/NO diethylamine nonoate diethylammonium salt DEPMPO 5-diethoxyphosphoryl-5-methyl-1-pyrroline-n-oxide DMF dimethylformamide DMPO 5,5-dimethylpyrroline-n-oxide DMSO dimethyl sulfoxide DNA deoxyribonucleic acid DTNB 5,5’-dithiobis-(2-nitrobenzoic acid) DTPA diethylenetriaminepentaacetic acid xviii E+ ethidium EC (-)-epicatechin ECG (-)-epicatechin gallate EGC (-)-epigallocatechin EGCG (-)-epigallocatchin gallate EI-MS electron ionization mass spectrometry EPR electron paramagnetic resonance ESI-MS electrospray ionization mass spectrometry ESR electron spin resonance EtOH ethanol FOX ferrous ion oxidation-xylenol orange GA gallic acid GAE gallic acid equivalents GSH glutathione GSHPx glutathione peroxidase GYY4137 morpholin-4-ium-4methoxyphenyl(morpholino)phosphinodithioate HE hydroethidine HOMO highest occupied molecular orbital xix HPLC high performance liquid chromatography HRP horseradish peroxidase HVA homovanillic acid IC interconfigurational IC50 the half maximal inhibitory concentration IL intra-ligand ITO indium-doped tin oxide IVCT intervalence charge-transfer LDL low-density lipoproteins LLIVCT ligand-to-ligand intervalence charge transfer LMCT ligand-to-metal charge-transfer LUMO lowest unoccupied molecular orbital MDA malondialdehyde MeOH methanol MF+ monoformazan MLCT metal-to-ligand charge-transfer MPO myeloperoxidase MSA methanesulfinic acid xx NBT nitroblue tetrazolium NIR near-infrared 2-OH-E+ 2-hydroxyethidium ORAC oxygen radical absorbance capacity PBS phosphate buffer saline PPO polyphenol oxidase RNS reactive nitrogen species ROS reactive oxygen species RP-HPLC reverse phase-high performance liquid chromatography SOD superoxide dismutase SOMO singly occupied molecular orbital TAE tris-acetate-edta TBA thiobarbituric acid TF theaflavin TFDG theaflavin-3,3’-digallate TFMG-a theaflavin-3-gallate TFMG-b theaflavin-3’-gallate TIED tetraiminoethylenedimacrocycle xxi TON turn over number TPC total phenolic content UV ultraviolet xxii Part I: Hydroethidine as a Fluorescent Probe for Quantifying Pro-oxidant Activity of Polyphenolic Compounds Chapter 1 Introduction on Pro-oxidants Chapter 1 1.1 Reactive oxygen species and oxidative stress A free radical is defined as any species capable of independent existence, which contains one or more unpaired electron [1]. Reactive oxygen species (ROS) refer to a group of oxygen radicals such as superoxide anion radical (O2●−), hydroxyl radical (●OH) and peroxyl radical (ROO●), and non radicals such as hydrogen peroxide (H2O2), hypochlorous acid (HOCl) and singlet oxygen (1O2) [ 2 ]. ROS are formed in large amounts as unavoidable byproducts of many biochemical processes or as the result of exogeneous factors such as smoking and air pollution [ 3 ]. They can cause severe oxidative damage to biological molecules, especially to DNA, lipids, and proteins [4]. Oxidative stress refers to the imbalance between the production of ROS and the activity of the antioxidant defense system [5,6]. Increased production of ROS and lack of antioxidant defense will result in oxidative stress. Steadily accumulating scientific evidence supports that severe oxidative stress is the causative factor in aging and several degenerative diseases, such as cataract [7], Parkinson’s disease [8], atherosclerosis [9], and cancer [10,11]. To protect against oxidative stress-related diseases, living organisms have developed an antioxidant-defense system, which are made up of endogeneous ROS scavengers. They include superoxide dismutase (SOD), catalase and glutathione peroxidase (GSHPx). SOD enzymes are transition metalcontaining complexes, which catalyze the dismutation of superoxide into molecular oxygen and hydrogen peroxide according to eq. 1.1 [ 12 ]. The catalase enzymes then convert the hydrogen peroxide formed into molecular oxygen and water (eq. 1.2) [ 13 ]. Glutathione peroxidases make use of 1 Chapter 1 hydrogen peroxide as an oxidant to convert reduced glutathione (GSH) to oxidized glutathione (GSSG) (eq. 1.3) [14]. SOD + 2H  H2O2 + O2 (1.1) catalase  2 H2O + O2 2 H2O2 (1.2) 2 O2 ●− + GSHPx 2 GSH + H2O2  GSSG + 2 H2O (1.3) The non-enzymatic antioxidants are made up of endogeneous antioxidants such as glutathione (GSH), and dietary antioxidants such as ascorbic acid (Vitamin C), α-tocopherol (Vitamin E) and flavonoids [15 ]. Studies have shown that a high dietary intake of vegetables and fruits can reduce the risk of cardiovascular diseases [16] and cancer [17,18] due to the ROS scavenging antioxidant role played by the polyphenols in such diets. 1.2 Structure and antioxidant activity of flavonoids Polyphenols are naturally occurring plant secondary metabolites. Over 4000 different types of flavonoids have been found and the number is still increasing. They are widely consumed as fruits and vegetables in the human diet for their health benefits as antioxidants. Flavonoids belong to the big family of polyphenols and can be said to be one of the most nutritionally important classes of dietary compounds. They can be found in red wine, tea, coffee, cocoa, nuts, fruits, and vegetables at a high concentration [19,20]. They have multiple biological activities including vasodilatory [21], antitumor [22], anti-inflammatory [ 23 ], antibacterial, immune-stimulating, antiallergic, and antiviral effects [24]. 2 Chapter 1 1.2.1 Chemical structure of flavonoids Flavonoids have a basic structure of 2-phenyl benzopyran with three components, A-, C- and B-ring (Figure 1.1), and can be classified into subclasses according to the functional groups they contain. These subclasses include flavones, flavonols, flavanols, flavanonols, flavanones, and isoflavones (Figure 1.2) [25]. Flavonol and flavanonol differ from flavone and flavanol respectively by the presence of a hydroxy (OH) group attached at the 3-position. Individual flavonoids are distinguished mainly by the number and position of OH groups substituted in the framework.   Figure 1.1. Basic structure of flavonoid, 2-phenylbenzopyran. Figure 1.2. Subclasses of flavonoids. Flavones are mainly found in parsley, rosemary and thyme [26], while flavonols are predominantly found in onions, broccoli, apples, berries, cherries, and in drinks such as tea and red wine [27]. Flavanones are mainly found in 3 Chapter 1 citrus fruits [28], and flavanols are found in tea, grapes, cocoa, pomegranate, apricots, apples, and cherries [22]. The predominant source of flavanols can be found in fresh tea leaves. Fresh tea leaves contain four major catechins, namely (-)-epicatechin (EC), (-)epicatechin gallate (ECG), (-)-epigallocatechin (EGC), and (-)-epigallocatchin gallate (EGCG) (Figure 1.3). EGCG, the most abundant catechin in green tea, accounts for 60–65% of the total catechin content. These colorless, watersoluble compounds contribute bitterness and astringency to green tea. Flavonols such as quercetin, kaempferol, myricetin, and their glycosides, which are characterized by a 4-oxo-3-hydroxy C-ring, can also be found in green tea [29,30]. Figure 1.3. Chemical structures of major polyphenolic catechins present in green tea. 1.2.2 Antioxidant activity of flavonoids 4 Chapter 1 In the past decade, the antioxidant activity of flavonoids has been given much attention and chemically, there are three features that confer on flavonoids their remarkable antioxidant properties [31]: • The hydrogen donating substituents (OH groups), attached to aromatic ring structures of flavonoids, which enable the flavonoids to undergo a redox reaction that helps them to scavenge free radicals more easily; • A stable delocalization system, consisting of aromatic and heterocyclic rings as well as multiple unsaturated bonds, which helps to delocalize the resulting free radicals, and • The presence of certain structural groups, which are capable of forming transition metal-chelating complexes that can regulate the production of reactive oxygen species such as OH and O2●−. In addition, three criteria need to be fulfilled to achieve good antioxidant activity for the flavonoids, [20]: (1) presence of 3’,4’-dihydroxy (catechol) moiety to stabilize the radical formed, (2) a C(2)=C(3) double bond providing conjugation among the B-, C- and A-ring, and (3) 3- and 5-hydroxy group with a C(4)=O oxo group. Examples of flavonols that fulfill these criteria are quercetin and myricetin. Flavonoids are reported to function as antioxidants by four modes of action, namely as free radical scavengers, metal ion chelators, lipid peroxidation inhibitors, and enzyme inactivators. Free radical scavenging activity Flavonoids are reported to be good free radical scavengers due to their excellent hydrogen- or electron-donating ability, and the flavonoid radical was quite stable due to electron delocalization and intramolecular hydrogen 5 Chapter 1 bonding [32]. They are capable of scavenging reactive oxygen species (ROS) and reactive nitrogen species (RNS) such as hydroxyl (HO●) [33], superoxide (O2●−) [34], peroxyl (ROO●) radicals [35,36], and peroxynitrite (ONOO−) [37,38]. The free radical scavenging ability of flavonoids is mainly due to the number and arrangement of their phenolic OH groups attached to ring structures. A catechol moiety (3’,4’-dihydroxy) in the B-ring is a main feature of the most potent scavengers of peroxyl [37], superoxide [ 39 ], and peroxynitrite radicals [38]. The peroxyl radical scavenging ability of quercetin substantially exceeds that of kaempferol, as the latter lacks the catechol moiety in the B-ring [ 40 ]. In addition, the free radical scavenging activity of flavonoids is highly dependent on the presence of a free 3-hydroxy group in the C-ring [41]. Flavonoids with a 3-hydroxy and 3’,4’-dihydroxy structure are reported to be 10 fold more potent than ebselen, a known RNS scavenger, against peroxynitrite [38]. Since flavonoids can stop free radical chain reactions by scavenging the reactive species, they are said to suppress aging, carcinogenesis, and the development of cardiovascular diseases, cancer, immune deficiency, and atherosclerosis [42]. Chelation of transition metals Flavonoids act as antioxidants by chelating redox-active metals such as copper and iron, so as to prevent metal-catalyzed free radical formation [43,44]. When traces of transition metals are present in their free states in biological systems, they will accelerate the auto-oxidation reactions and the decomposition of lipid hydroperoxide (LOOH) to LOO●, LO● radicals and cytotoxic aldehydes [45]. Flavonoids which chelate these transition metals will 6 Chapter 1 help to prevent such reactions from occurring and decrease their biological effects dramatically. There are three possible metal chelating sites in flavonoids, namely at the catechol moiety (3’,4’-dihydroxy groups) in the B-ring, 5-hydroxy-4-oxo group and 3-hydroxy-4-oxo group between A- and C-ring as illustrated in Figure 1.4. O M O HO O O O O M M Figure 1.4. Metal chelating sites in flavonoids. The metal chelating ability of flavonoids is attributed to the presence of aromatic OH groups, their positions in the three rings, the oxidation state of the C-ring, and the overall number of OH groups present [46,47]. It has been shown that the chelating ability of the catechol moiety on the B-ring increases as pH increases, which make it easier for metal chelation. The effect of flavonoids on iron-induced lipid peroxidation showed an antioxidant activity, suggesting the formation of an inert complex between iron and flavonoid [48,49,50]. In addition, the prevention of the catalytic effect of copper(II) through chelation with flavonoid has been reported as a major antioxidant mechanism [51]. Inhibition of enzymes The third mode of antioxidant action is by inhibiting the ability of myeloperoxidase (MPO) to oxidize low-density lipoproteins (LDL) [52] and 7 Chapter 1 also by inhibiting an array of enzymes such as xanthine oxidase [ 53 ], phospholipase A2 [ 54 ], lipoxygenase [ 55 ], cyclooxygenase [55], monooxygenase [56], protein kinase, and ATPase [57]. Xanthine oxidase will catalyze the oxidation of both xanthine to uric acid, while reducing molecular oxygen to superoxide and hydrogen peroxide. Green tea catechins have been found to inhibit the activity of xanthine oxidase in vitro, with EGCG showing the highest inhibition, thereby preventing the formation of ROS [ 58 ]. Isoflavones [59] and tea polyphenols [60] are reported to inhibit lipoxygenases and prevent the development of inflammatory diseases such as atherosclerosis, inflammatory bowel disease, atopic dermatitis, and psoriasis. Green and black tea polyphenols have also shown the inhibition of cyclooxygenase-2 and 5-, 12-, and 15-lipoxygenase activities in human colon mucosa cells and human colon cancer cells [60]. In addition, studies have shown that flavonoids inhibit –amylase and –glucosidase [61,62], impair starch digestion and slow down the increase in glucose concentration in the blood, which greatly benefits diabetic patients. Owing to their antioxidative properties, flavonoids show a wide spectrum of action on the mammalian cell, including anti-atherosclerotic, antitumoral, antiplatelet, anti-ischemic, anti-allergic, antiviral, and anti-inflammatory activities [63]. 1.3 Pro-oxidant activity of flavonoids The potential health benefits of flavonoids are well-known to be attributed to their antioxidative and free radical scavenging activities demonstrated in vitro. However, the opposite side of these dietary polyphenolic compounds as 8 Chapter 1 pro-oxidants has not been thoroughly studied. Studies have shown that flavonoids act as pro-oxidants under certain conditions [64]. 1.3.1 Pro-oxidant activity in the absence of transition metals It is known for a long time that in the absence of transition metal, certain flavonoids, like pyrogallol and myricetin, undergo pH dependent autooxidation to generate semiquinone radical and superoxide, which will disproportionate to persistent hydrogen peroxide (eq. 1.4 - 1.6) [64,65]. ArOH  ArO− + H+ (1.4) ArO− + O2  ArO● + O2●− (1.5) 2 O2●− + 2H+  H2O2 + O2 (1.6) Hodnick et al. has demonstrated that auto-oxidation of myricetin involved O2●−, since the addition of SOD inhibited the auto-oxidation [66]. In addition, myricetin and quercetin had lower oxidation potentials than those that did not undergo auto-oxidation, indicating that the auto-oxidation was thermodynamically feasible for myricetin and quercetin. In consistency with the results from Hodnick et al., Canada et al. also found that the rate of autooxidation for myricetin was much faster than that for quercetin, but no autooxidation was observed for rutin due to glycosylation of the 3-OH group [67]. This indicated that the pyrogallol (3’,4’,5’-trihydroxy) moiety in the B-ring and the 3-OH group in the C-ring were critical for the high pro-oxidant activity in terms of auto-oxidation rate. Other than auto-oxidation, the flavonoids can be oxidized by peroxidase to generate semiquinone radicals in the absence of metals. The flavonoid semiquinone radical can co-oxidize glutathione (GSH) to regenerate the 9 Chapter 1 flavonoid and also generate the thiyl radical of glutathione (GS●) in eq. 1.7. The thiyl radical can then react with another GSH to form a disulfide radical anion (GSSG●−) in eq. 1.8, which will reduce molecular oxygen rapidly to generate superoxide anion radical [68] in eq. 1.9. This pro-oxidant effect has resulted in semiquinone radical mediated hemolysis and thyroid peroxidase inactivation by flavonoids. ArO● + GSH  GS● + ArOH (1.7) GS● + GS− (1.8) GSSG●− + O2 GSSG●− GSSG + O2●− (1.9) Green tea polyphenols are also unstable in the presence of air and can undergo auto-oxidation to generate ROS under typical cell culture conditions [69]. It was found that EGCG incubated in the absence of cells in cell culture medium at 37 °C resulted in the time-dependent formation of H2O2, followed with a decrease in the concentration of EGCG (half-life < 30 min), and an increase in the concentration of theasinensin A (oxidative dimer of EGCG) formed [69]. Halliwell and coworkers found that tea and coffee accumulate hydrogen peroxide upon ageing, presumably also due to aerial oxidation of polyphenolic compounds [70,71]. 1.3.2 Pro-oxidant activity in the presence of transition metals or peroxidases It has been reported that flavonoids will exhibit pro-oxidant behaviour in systems containing redox-active metals. The hydrogen peroxide formed when combined with redox-active transition metal ions such as Fe(II) can lead to generation of highly reactive hydroxyl radical (eq. 1.10) [72]. Flavonoids 10 Chapter 1 acted as pro-oxidants in these cases by regenerating the Fe(II) through single electron reduction: H2O2 + Fe(II)  Fe(III) + HO● + HO− (1.10) ArOH + Fe(III)  Fe(II) + ArO● + H+ (1.11) The consequences of the metal-catalyzed oxidation of flavonoids are not just only the generation of ROS, but also the generation of semiquinone radicals and eventually quinone or quinone methide intermediates for those with a 3’,4’-catechol moiety such as luteolin and quercetin. The highly electrophilic quinone species can then further react with free thiol compounds such as glutathione to form stable flavonoid glutathionyl adducts [73,74] as shown in Figure 1.5. However, it is unlikely that such metal-catalyzed autooxidation occurred in vivo, since transition metals are bound to proteins [75]. In the absence of transition metals in vivo, the oxidative action in converting the flavonoids to quinone or quinone methide intermediates will be carried out by tyrosinase, or hydrogen peroxide and horseradish peroxidase (HRP) or other peroxidases instead [74,76,77]. Catechol estrogens can also be bioactivated in the same manner to their quinone GSH conjugates, which plays a role in tumor formation due to excessive exposure to estrogens. This is one of the mechanisms proposed to be responsible for the connection between estrogen exposure and the risk of developing cancer [78,79,80]. EGCG was oxidized by peroxidase/hydrogen peroxide to form o-quinone, which reacted with glutathione to form thiol conjugates [73]. In addition, the reduced Fe(II) formed from eq. 1.11 can catalyze the decomposition of lipid hydroperoxide and hydrogen peroxide to generate lipid alkoxyl and hydroxyl radicals respectively. 11 Chapter 1 tyrosinase/O2 B O OH OH HO Metal-catayzed HO oxidation O OH OH O Quercetin O O OH HO O O or HRP/H2O2 OH OH OH O Semiquinone OH Disproportionation HO Ortho-quinone OH O O Tautomerization O O OH Quinone methide O OH Tautomerization O O O O OH OH Quinone methide GSH OH H O GS O O OH OH Meisenheimer complex OH OH O GS HO O OH OH OH Isomerization OH OH GS HO O OH OH O Quercetin-glutathione conjugate Figure 1.5. Pro-oxidant mechanism of catechol-type flavonoids, (A) luteolin and (B) quercetin, with GSH [74]. Moreover, the production of ROS from the auto-oxidation of flavonoids in the presence of transition metals can accelerate the LDL oxidation during the propagation phase. However, in vivo, most transition metal ions are unable to 12 Chapter 1 catalyze free radical reactions due to their sequestered forms [81]. Low levels of free copper ions may be released by tissue injury and result in oxidative damage to cells and proteins [82]. These results suggest the role that the transition metal played in some of the pro-oxidant behaviour of flavonoids. 1.3.3 Pro-oxidant activity in terms of DNA damage and lipid peroxidation The ROS and semiquinone radicals formed from the redox cycling of flavonoids in the presence of transition metals can lead to oxidative damage of deoxyribonucleic acid (DNA), lipids and other biological molecules [83,84,85]. Unrepaired oxidative DNA damage can result in DNA strand breaks and mutations [86,87], which may lead to cancer induction [88,89]. Yen et al. reported that the auto-oxidation of some flavonoids cause DNA damage to human lymphocytes [90]. Sahu et al. indicated that myricetin, kaempferol, morin, and naringenin induced lipid peroxidation and DNA strand breaks in isolated rat liver nuclei [91,92,93]. The generation of hydroxyl radicals from the auto-oxidation of flavonoids in the presence of transition metals such as Fe(III) and Cu(II) may initiate lipid peroxidation by a metal-catalyzed HaberWeiss mechanism (eq. 1.12), in which the ROS formed will oxidize DNA, resulting in DNA damage [91,92]. In addition, the covalent binding of the flavonoid semiquinone radical (ArO●) to DNA has been suggested to induce DNA cleavage [94]. Moreover, the semiquinone radical can be oxidized by molecular oxygen to form quinone and generate superoxide anion (eq. 1.13). The oxidized quinone may also react with DNA directly to form DNAflavonoid or DNA-copper-flavonoid adducts which can result in genotoxicity [91,92,95]. 13 Chapter 1 O2●− + H2O2  O2 + HO● + OH− (1.12) ArO● + O2  oxidized quinone + O2●− (1.13) The green tea catechin, EGCG, can also induce H2O2 generation and oxidative damage to isolated and cellular DNA in the presence of transition metal ions [96]. Moreover, studies have shown that administration of high doses (up to 50 µM) of purified flavonoids for six hours can result in chromosome translocation in human cell line studies [97]. Several studies have shown that catechin-related oxidation is attributed to the presence of metal ions. It was demonstrated that tea extracts, especially the green, pouchong, and oolong tea extracts, markedly stimulated the oxidation of deoxyribose in the presence of Fe(III) and H2O2 [98]. An experiment using 32 P-labeled DNA fragments obtained from human cancer-related genes showed that caffeic acid induced DNA damage in the presence of metals such as Cu(II) complexes [99]. Another in vitro study also demonstrated that tea catechins with Cu(II) ion caused extensive DNA cleavage and fatty acid peroxidation under aerobic conditions [100]. It was reported that catechins-Cu(II)-induced DNA cleavage was significantly inhibited by catalase but not by superoxide dismutase, indicating that H2O2 may be involved in the DNA cleavage [100]. These results suggested that the pro-oxidant properties of tea catechins are attributed to reactive oxygen species which are generated by reduction of O2 through a combination action of catechins and Cu(II) ion. 1.3.4 Pro-oxidant activity in terms of enzyme and topoisomerase inhibitors 14 Chapter 1 The activity of flavonoids as topoisomerase II inhibitors has resulted in their cytotoxicity. Genistein, myricetin and quercetin have been identified to be potent topoisomerase II inhibitors at low concentrations [101,102,103]. Topoisomerase II inhibitors will cause the topoisomerase II-DNA covalent intermediates called cleavable complexes to accumulate, which may then lead to double-stranded DNA cleavage at specific topoisomerase binding sites and cell death. It has been found that increasing maternal consumption of DNA topoisomerase II inhibitor-containing food elevated the risk of acute myeloid leukemia in infants by about 10-fold [104]. Some studies also demonstrated that flavonoids can inhibit glutathione reductase, a key antioxidant enzyme, thereby promoting extensive oxidative stress in vivo [105,106,107]. Several flavonoids such as quercetin, kaempferol and naringenin can inhibit thyroxine synthesis by acting as alternative substrates for tyrosine iodination, yielding mono-, di-, and tri-iodoisoflavones [108]. It was also found that these flavonoids can inhibit the enzyme thyroid peroxidase, thus interfering with thyroid hormone biosynthesis, and eventually lead to thyroid gland growth and thyroid dysfunction [108]. Flavonoids with a free resorcinol (meta-hydroxylphenol) moiety have the ability to inhibit thyroid peroxidase and lactoperoxidase due to persistent and irreversible covalent binding [ 109 ]. The proposed mechanism for enzyme inhibition involves the conversion of thyroid peroxidase to a free radical, which reacts with a flavonoid with a resorcinol moiety to yield a flavonoid radical. The flavonoid radical can then covalently bind to the catalytic amino acid residues on the enzyme, thus inactivating the enzyme. Therefore, care must be taken when consuming flavonoids with a resorcinol moiety at high concentrations as 15 Chapter 1 they may be potential thyroid carcinogens. It has also been suggested that increased consumption of soy isoflavones can reduce fertility and retard sexual maturation [110]. Flavonoids such as naringenin in grapefruit can inhibit drugmetabolizing enzymes like CYP 3A4, a member of cytochrome P 450 (CYP) enzyme system, involved in xenobiotic metabolism, and thus interact with medicinal drugs increasing the risk of overdose and harm [111]. Therefore, it is important to consider the pro-oxidant effects of flavonoids on biological molecules. 1.3.5 Pro-oxidant activity in terms of cancer therapy The pro-oxidant action of flavonoids may be important in terms of cancer therapy due to their anticancer and apoptosis-inducing properties since ROS can mediate apoptotic DNA fragmentation [95, 112 , 113 ]. Several known anticancer drugs make use of such a mechanism [113], which is similar to certain pro-oxidant properties of flavonoids in terms of DNA cleavage and generation of ROS in the presence of transition metal ions [95]. The production of ROS has also been reported to play an important role in the proapoptotic effects of EGCG and other tea polyphenols against cancer cell lines [114,115,116]. Addition of EGCG, EGC and theaflavin-3,3’-digallate (TFDG) to Ha-ras gene transformed (21BES) human bronchial epithelial cells showed a dose-dependent growth inhibition with IC50 values of 22 to 24 µM [114]. EGCG, EGC and TFDG induced apoptosis in 21BES cells. Similarly, when HL-60 cells were treated with 50 µM EGCG, there was formation of ROS as detected by dichlorofluorescein (DCF) and also an increase in the number of 16 Chapter 1 apoptotic cells [117]. It was found that addition of catalase or SOD reduced the levels of ROS and the number of apoptotic cells. 1.4 Pro-oxidant activity assays Various methods have been used to measure the pro-oxidant activities of flavonoids, and the pro-oxidant activity of the flavonoids in these methods is generally defined as its ability to generate H2O2 and in turn ●OH radicals through the Fenton reaction and further in terms of oxidative DNA damage. 1.4.1 Deoxyribose assay The deoxyribose assay is the most commonly used assay for determining the pro-oxidant activity of test compounds [118,119]. In this assay, the prooxidant is tested for its ability to reduce Fe(III) to Fe(II) (eq. 1.11) and the Fe(II) in turn reacts with H2O2 to produce ●OH through Fenton reaction. The radical then attacks deoxyribose to produce malondialdehyde (MDA). Upon heating under acidic conditions, MDA reacts with thiobarbituric acid (TBA) to form a pink chromogen, which can be quantified spectrophotometrically at 532 nm (eq. 1.14). Laughton et al. and Puppo has used the deoxyribose assay to investigate the pro-oxidant activity of flavonoids [64,120]. Laughton et al. found that the prooxidant activity for myricetin was higher than that for quercetin and interestingly, addition of SOD inhibited their pro-oxidant effect [64]. Their 17 Chapter 1 group suggested that Fe(III)-ethylenediaminetetraacetic acid (EDTA) oxidized the flavonols to form O2●− which caused the reduction of Fe(III)-EDTA to Fe(II)-EDTA as no pro-oxidant activity was observed for Fe(III) alone or physiologically relevant Fe(III)-adenosine-diphosphate (ADP) or Fe(III)citrate complexes. Puppo’s results showed that the ●OH generating activity of the flavonoids decreased in the following order: myricetin > quercetin > morin > kaempferol, while flavones had no effect [120]. These results were in agreement with the pro-oxidant hypothesis of Laughton et al. [64]. However, when biologically relevant chelators adenosine-5'-triphosphate (ATP) and citrate are present, the flavonoids did not act as pro-oxidants [120]. This in turn suggests that the presence of iron chelators will influence the pro-oxidant activity of flavonoids. However, one disadvantage is that it is only applicable to water-soluble compounds, as organic solvents such as ethanol (EtOH) and dimethyl sulfoxide (DMSO) can scavenge ●OH. 1.4.2 Other pro-oxidant assays The pro-oxidant effects of myricetin and quercetin on DNA have been investigated using the bleomycin assay [64]. Bleomycin, being an antitumor antibiotic, will bind both DNA and Fe(III). The bound Fe(III)-bleomycin complex will only damage DNA in the presence of molecular oxygen and a reducing agent such as ascorbic acid or flavonoid [118]. It has been found that myricetin and quercetin cause greater DNA damage, most likely by reducing the Fe(III)-complex to Fe(II)-complex, leading to the generation of ROS, which can result in DNA damage [64]. 18 Chapter 1 The pro-oxidant activity of flavonoids has also been investigated by Hanasaki et al. [121] using a method whereby the ●OH generated from Fenton reaction will oxidize DMSO to form the stable compound methanesulfinic acid (MSA) as shown in eq. 1.15 [122]. MSA is then derivatized with a diazonium salt to form diazosulfone (eq. 1.16), which can be quantified by high performance liquid chromatography (HPLC) [123]. CF3SOOH + Ar-N=N+  H+ + Ar-NH-SO2-CH3 (1.16) MSA diazonium salt diazosulfone (coloured, hydrophobic) The results showed that baicalein (5,6,7-trihydroxyflavone), morin, quercetin, and myricetin increased ●OH production, demonstrating their prooxidant behaviour [121]. It was suggested that these pro-oxidant flavonoids generated H2O2 during auto-oxidation, which then started the Fenton reaction in the presence of Fe(II) [121]. One drawback is that it may give rise to artifactual DMSO oxidation during the assay procedure itself due to prolonged incubation or storage time, as opposed to biological DMSO oxidation by ●OH, giving rise to false positive results [122]. The other drawback is that sulfinic acids may be degraded in biological samples before measurement, giving rise to false-negative results [122]. The pro-oxidant activity of flavonoids has also been investigated by Cao et al. using the oxygen radical absorbance capacity (ORAC) assay [36]. In a Cu(II)-H2O2 system, the flavonoids will act as antioxidants by scavenging ● OH. However in the absence of H2O2, the flavonoids acted as pro-oxidants instead. Once the reaction started, they would form H2O2. It was found that the 19 Chapter 1 pro-oxidant activity decreased in the following order: myricetin > quercetin > kaempferol > taxifolin, while flavone and 6-hydroxyflavone did not show any pro-oxidant activity [36]. The results demonstrated the presence of C(2)=C(3) double bond and more OH groups are critical for initiating the pro-oxidant behaviour of flavonoids [36]. Hayakawa et al. [124] measured the generation of hydrogen peroxide and hydroxyl radical from tea catechins in the presence of copper ions through a fluorometric method described by Guilbault et al. [125]. The probe used was homovanillic acid (HVA), which is non-fluorescent, but upon oxidation, it is converted to a strongly fluorescent compound (λex = 315 nm, λem = 425 nm). It was discovered that in the presence of Cu(II), the ●OH radical is generated from EGC and EC, but not from EGCG and ECG [124]. The results explained the reason why EGCG showed lesser pro-oxidant activity than EGC could be due to the chelating ability of catechin gallates to metal ions under the experimental conditions [124]. Inui et al. observed that simple polyphenols induced the aerobic oxidation of ethanol to acetaldehyde in water in the presence of Fe(II)diethylenetriaminepentaacetic acid (DTPA) [ 126 ]. Therefore, this method served as an indicator of the pro-oxidant activity of the polyphenols from the amount of acetaldehyde formed due to oxidation by hydroxyl radical generated from the Fenton reaction induced by polyphenolic compound. The detection of acetaldehyde was carried out by reacting it with sodium hydrogen sulfite to form acetaldehyde-hydrogensulfite adduct, which was then passed through a C18 reversed-phase column at pH 3.2 (eluting solution: 10 mM ammonium acetate and 20 mM acetic acid) to convert it into 1-hydroxyethanesulfonic acid. 20 Chapter 1 The 1-hydroxyethanesulfonic acid formed was then passed through a postcolumn at pH > 7.0 to convert it into hydrogen sulfite, which reacts with phthalaldehyde to form 2H-isoindole-1-sulfonic acid, a fluorescent compound (λex = 320 nm, λem = 390 nm). The reaction product was then analyzed by HPLC equipped with fluorescence detector. The disadvantages of this method are that all the manipulations were done by HPLC and can be quite tedious and time-consuming to ensure that the reaction is complete or with consistent conversion. There is a similarity in all the reported pro-oxidant assays, in which all requires the presence of transition metals such as Fe(III) or Cu(II) to induce the Fenton reaction for ●OH production. Up to now, there is no convenient assay reported in quantifying the pro-oxidant activity in terms of superoxide radical formation activity of polyphenolic compounds in the absence of transition metals. There is an urgent need to come up with a simple and convenient assay for the quantification of pro-oxidant activity of flavonoids in terms of superoxide radical formation by looking at the various probes which have been developed for the detection of superoxide. 1.5 Detection methods of superoxide Superoxide has an absorption maximum in the ultraviolet (UV) region at 245 nm, with an extinction coefficient of 2350 M-1cm-1 [127], which is not easy to be detected since most compounds also absorb in UV region. Therefore, a number of detection methods have been developed for superoxide detection, which generally includes using UV-vis spectrophotometry, fluorescence, chemiluminescence, and electron spin resonance technique with 21 Chapter 1 spin traps [ 128 , 129 , 130 ]. However, the detection and quantification of superoxide has met with a couple of technical difficulties in the detection methods mentioned above. In addition, the half-life of O2●− is short (~ 0.05 seconds) [ 131 ], which greatly hinders its detection. The commonly used methods for detection of superoxide and their advantages and limitations will be discussed in detail. 1.5.1 Spectrophotometric probes The two most commonly used spectrophotometric detection methods of superoxide are the cytochrome c reduction assay and nitroblue tetrazolium reduction assay. Cytochrome c In the cytochrome c reduction assay, ferricytochrome c is reduced by O2●− to ferrocytochrome c, which has an intense absorption at 550 nm as shown in eq. 1.17 [132,133]. The reaction is then monitored spectrophotometrically at 550 nm. Fe(III) cyt c + O2●−  Fe(II) cyt c + O2 (1.17) There are several drawbacks of this method. Firstly, this method is nonspecific for O2●−, as ferricytochrome c can also react with other oneelectron reductants such as ascorbate and glutathione present in cell tissues. Antioxidants such as polyphenolic compounds can also reduce ferricytochrome c, hence interfering with the assay. Secondly, the ferrocytochrome c formed can be easily oxidized by other oxidants such as H2O2 and ONOO− [ 134 ]. This will in turn lead to a lower absorbance measured at 550 nm and as a result underestimate the amount of superoxide 22 Chapter 1 detected. Hence this method can only be used when competing extracellular antioxidants are not present. Nitroblue tetrazolium In the nitroblue tetrazolium (NBT) reduction assay, NBT is reduced by O2●− through one-electron transfer to monoformazan (MF+, eq. 1.18-1.19), which has an intense absorption at 560 nm and can be monitored spectrophotometrically [135]. NBT2+ + O2●−  NBT+ + O2 (1.18) 2 NBT+ + H+  NBT2+ + MF+ (1.19) NBT+ + O2  NBT2+ + O2●− (1.20) Similar to the cytochrome c reduction assay, other reductants present can also react with NBT, so this method is not exclusive to O2●−. In addition, under aerobic conditions, the tetrazoinyl radical (NBT+) can also react with atmospheric oxygen to form O2●− as shown in eq. 1.20 [136]. The artifact O2●− formed can then react with NBT to form monoformazan. This will in turn overestimate the amount of O2●− present. Hence the cytochrome c and NBT reduction assay are not suitable methods for detecting superoxide especially in the presence of other interfering reductants. Being spectrophotometric methods, the major disadvantage is their low sensitivity in comparison with other probes. 1.5.2 Fluorescent probes Hydroethidine The classical fluorescent probe for detection of superoxide is hydroethidine (HE), It has been believed all along that HE is oxidized by superoxide to yield 23 Chapter 1 ethidium (E+), a fluorescent compound (λex = 520 nm and λem = 610 nm) [137]. In contrast to the NBT probe, there is little artifactual formation of O2●−. However, there are two limitations of the assay. Firstly, HE can also be oxidized by cytochrome c, which may serve as interference in situations when the main source of O2●− generation is the mitochondria [138]. In such cases, it is difficult to assume that the oxidation of HE is due to O2●− only. Secondly, HE can increase the O2●− dismutation rate to H2O2, which means that O2●− quantification using HE as probe may not be accurate [139]. Thirdly, HE can also be oxidized by other oxidants such as hypochlorous acid and H2O2 via the non-specific peroxidases such as horseradish peroxidase and myeloperoxidase [ 140 ]. The fluorescent oxidation products can cause interference as their emission bands (580–600 nm) were found to be near to the ethidium and recently clarified 2-hydroxyethidium (2-OH-E+) emission bands [141,142]. Zhao et al. have shown that HE was oxidized by superoxide to give selectively 2-OH-E+ instead of E+ as it was used to be believed as shown in eq. 1.21 [141,142]. The major advantage is that the fluorescent product, 2-OH-E+ (λem = 567 nm), is highly specific for O2●−. It is a unique marker product that is not generated from HE reaction with other oxidants such as ONOO−, H2O2, HO●, and ROO● [141,142]. This assay seems unique because HE is oxidized by 24 Chapter 1 O2●− instead of being reduced by O2●− [143]. Hence it may be a suitable probe to be used for measuring the pro-oxidant activity of flavonoids, as the flavonoids being reductants will not interfere with the assay. Since there is an extensive overlap in the emission bands of 2-OH-E+ and E+ [142], the two species cannot be detected and quantified reliably by using fluorescence technique. HPLC-fluorescence technique can be used for the separation and detection of 2-OH-E+ and E+ [144]. 1.5.3 Luminescence probes The two widely used luminescence probes for the detection of O2●− include luminol and lucigenin. Luminol The chemiluminescence (CL) mechanism of the reaction of luminol with O2●− is shown in Figure 1.6. First, luminol undergoes a ketoenol tautomerism and a simple ionization reaction to form luminol anion. One-electron oxidation of the luminol anion by strong oxidants forms the luminol radical, which could reduce O2 to O2●−. The luminol radical will then react with O2●− to form an unstable endoperoxide, whose decomposition produces N2 and an electronically excited aminophthalate. Luminescence occurs when the aminophthalate emits a photon upon relaxation to the ground state [145]. One drawback is that luminol is not specific for O2●−, as other ROS such as H2O2, ●OH and ONOO− can also result in luminol CL [146]. In addition, same as the NBT assay, luminol in the presence of any univalent oxidant can form O2●− and therefore is an unreliable detector of O2●−. 25 Chapter 1 Figure 1.6. CL mechanism of the reaction of luminol with O2●− [145]. Lucigenin Lucigenin-amplified CL is very useful for the detection of low concentration of O2●−. The CL mechanism of the reaction of lucigenin with O2●− is shown in Figure 1.7 [147]. First, lucigenin (Luc2+) undergoes a oneelectron reduction to its cation radical (Luc●+), which can also reduce O2 to form O2●−. Luc●+ will then react with O2●− to form an unstable dioxetane. Decomposition of the dioxetane molecule produces two molecules of Nmethylacridone, whereby one is in an excited state. Luminescence occurs when the excited acridone emits a photon upon relaxation to the ground state [145]. The drawback is that lucigenin, like luminol, can also produce artifactual O2●−, but the difference is that it does so through catalytic redox cycling. This in turn leads to an overestimation of the O2●− production. Therefore, 26 Chapter 1 chemiluminescent techniques using luminol and lucigenin are not desirable for the detection of O2●−. CH3 CH3 CH3 N N N +e O2 - O -e O N N N CH3 CH3 CH3 Lucigenin Lucigenin radical Lucigenin dioxetane + O2 O2 - CH3 O * N N O CH3 O hv N CH3 N-methylacridone Figure 1.7. CL mechanism of the reaction of lucigenin with O2●− [147]. 1.5.4 Electron spin resonance and spin trapping Electron spin resonance (ESR) is the only analytical method that allows for the direct detection of O2●−. Direct detection of O2●− by ESR is difficult due to its short half-life and its low concentration. This can be resolved by adding spin traps to form more stable secondary radical species so that they can be measured by ESR, since the sensitivity limit of ESR is 1.0 nM [148]. Spin trap is generally an organic nitrone compound to produce relatively stable and detectable nitroxide radical adducts. O2●− specific spin traps include 5,527 Chapter 1 dimethylpyrroline-N-oxide (DMPO) [ 149 ] and 5-diethoxyphosphoryl-5methyl-1-pyrroline-N-oxide (DEPMPO) [150,151]. The problem with DMPO is that the superoxide adduct has a short half-life of about 50 seconds at pH 7 [152]. In addition, decomposition of this adduct readily occurs, forming a species which may be identical to the hydroxyl radical adduct. A more stable spin trap like DEPMPO has been developed. The half-life of the superoxide adduct is about 14 minutes at room temperature [152]. Moreover, DEPMPOOOH does not decompose to DEMPO-OH [151,153]. Therefore, DEPMPO is very useful in measuring cell and tissue O2●− production [154,155]. In addition, spin trapping with DMPO or DEPMPO is selective and more sensitive for detection of superoxide than the cytochrome c method. The drawbacks with ESR are the high cost and limited availability of ESR spectrometers and the necessity to perform experiments in close proximity to an ESR spectrometer. 1.6 Aim of this research The aim of this research is to come up with a simple and convenient assay to quantify the pro-oxidant activity of flavonoids such as flavonols and tea polyphenols in terms of superoxide radical formation and understanding the structure and activity relationship and reaction mechanisms. Specifically, this project will include the following research tasks: (1) A convenient assay making use of hydroethidine as a fluorescent probe will be established to quantify the superoxide radical forming pro-oxidant activity of flavonols under physiologically relevant conditions. 28 Chapter 1 (2) The assay will then be used to quantify the pro-oxidant activity of green tea, oolong tea and black tea to correlate the different degrees of fermentation of tea with their pro-oxidant activity. The pro-oxidant activity of the main tea polyphenols namely, (-)-epicatechin (EC), (-)-epicatechin gallate (ECG), (-)epigallocatechin (EGC), (-)-epigallocatechin gallate (EGCG), theaflavin (TF), theaflavin-3-gallate (TFMG-a), theaflavin-3’-gallate (TFMG-b), and theaflavin-3,3’-digallate (TFDG), will also be measured to assess whether the difference in pro-oxidant activity of the teas was due to the different prooxidant activity of the tea components. (3) The tea catechins, theaflavins and gallic acid present in the tea samples will be quantified using reverse phase-high performance liquid chromatography (RP-HPLC) to determine if there is a relation between the composition of polyphenols in the tea extract and the pro-oxidant activity of the tea extracts. (4) The DNA damaging pro-oxidant activity of flavonols and tea catechins will be examined using gel electrophoresis to correlate the DNA damaging ability with the superoxide radical forming pro-oxidant activity of flavonols and tea catechins. 29 Chapter 2 Pro-oxidant Activity of Flavonols Quantified by a Fluorescent Probe Hydroethidine Chapter 2 2.1 Introduction Dietary polyphenolic compounds are well-known for their radical scavenging antioxidant property, which is believed to be beneficial to human health by reducing the risk factors of chronic diseases [ 156 , 157 ]. Comparatively, much less thoroughly studied is the opposite side of the polyphenolic compounds as pro-oxidants. Shahidi and coworkers discovered that aqueous green tea extract acted as pro-oxidant in fish oils, likely due to the catalytic effects of residual chlorophylls [158]. When the pigment was removed from the extract, it behaved as an antioxidant [158]. At low concentrations, ascorbic acid and gallic acid accelerate the oxidation of deoxyribose induced by Fe(III)-EDTA-H2O2, acting as pro-oxidants [159]. It was suggested that the hydrogen peroxide formed when combined with redox active transition metal ions such as Fe(II) can lead to generation of highly reactive hydroxyl radical (HO•) (eq. 1.10) [72,160]. Polyphenolic compounds acted as pro-oxidants in these cases by regenerating the Fe(II) through single electron reduction (eq. 1.10 - 1.11). In the absence of transition metal, polyphenols like pyrogallol and myricetin undergo pH dependent auto-oxidation to generate semiquinone radical and superoxide, which will disproportionate to persistent hydrogen peroxide (eq. 1.4 - 1.6) [64, 65]. Tea and coffee accumulated hydrogen peroxide upon ageing, likely due to such mechanism [70, 71]. The bioactivity of the phenolic compounds measured in cell line was suggested to be artifacts due to auto-oxidation of these compounds added to cell culture media [161]. To understand the structure-activity relationship of the polyphenolic compound structures and their radical generating pro-oxidant activity, it is of 30 Chapter 2 great importance to establish a convenient method for quantifying free radical generating pro-oxidant activity of polyphenolic compounds under physiologically relevant conditions. While the deoxyribose assay was commonly applied in measuring the pro-oxidant activity of polyphenolic compounds through catalyzing the formation of hydroxyl radical in a Fe(III)H2O2 mixture [119], there is no convenient assay in quantifying the superoxide radical formation activity of polyphenolic compounds in the absence of transition metals. This study aims to take advantage of hydroethidine (HE) as the probe of choice for the quantification of pro-oxidant activity of flavonols. The flavonols used in this study are myricetin, quercetin, kaempferol, and galangin as shown in Figure 2.1. In addition, the DNA damaging pro-oxidant activity of these flavonols will be examined using gel electrophoresis to correlate their DNA damaging ability with their superoxide radical forming pro-oxidant activity. Figure 2.1. Chemical structure of the flavonols used. 2.2 Materials and methods 2.2.1 Materials Myricetin and galangin were bought from Indofine Chemical Company, Inc (Hillsborough, NJ). Bovine erythrocyte copper-zinc superoxide dismutase (SOD), potassium superoxide (KO2), kaempferol, quercetin, and gel loading solution type 1 (6X) were purchased from Sigma-Aldrich, Inc (St Louis, MO, USA). Hydroethidine was obtained from Polysciences, Inc (Warrington, PA). 31 Chapter 2 Ethidium bromide (E+Br−) solution (10 mg/mL) was purchased from Research Organics, Inc (Cleveland, OH). Dipotassium hydrogen phosphate (K2HPO4) and potassium dihydrogen phosphate (KH2PO4) were from Merck (Darmstadt, Germany). pBR322 DNA was purchased from New England Biolabs. Agarose (electrophoresis grade) was from Life Technologies, Inc (USA), EDTA was from BDH Chemicals Ltd (England) and copper(II) sulfate pentahydrate was from J. T. Baker, Inc (USA). Phosphate buffer saline (PBS) stock solution of ultrapure grade and tris-acetate-EDTA (TAE) buffer (10X) were purchased from National University Medical Institute (Singapore). All other chemicals used were of analytical reagent grade. 2.2.2 Instruments Fluorescence measurements were obtained with a Synergy HT microplate fluorescence reader from Bio-Tek Instruments, Inc (Winooski, Vermont), installed with the software KC4. Fluorescence spectra were acquired using Perkin Elmer LS 55 luminescence spectrometer (UK). HPLC analysis was done on a Waters 2695 HPLC system equipped with Waters 2996 photodiode array detector (Milford, MA, USA) and installed with Empower program. A C18 reversed-phase column (Waters Atlantis T3, 6 µm, 4.6 x 250 mm) was used throughout this study. The PDA acquisition wavelength range was set at 200–400 nm. 2.2.3 Preparation of stock solutions All flavonol stock solutions (400 µM) were prepared in methanol (MeOH) and stored at 4 °C. Myricetin and quercetin stock solutions were then diluted 32 Chapter 2 to 100 µM with MeOH. HE (3.17 mM) was prepared in CH3CN, aliquotted into 200 µL portions in brown vials and stored at –20 °C until use. Immediately prior to use, HE stock solution was diluted with phosphate buffer working solution to 31.7 µM. pBR322 DNA was split into 10 µL aliquots of 1.0 mg/mL and stored at –40 °C. PBS stock solution was diluted 10 times with deionized water to 1.0 L and a few drops of 1.0 M sodium hydroxide (NaOH) were added to yield a pH 7.40 phosphate buffer working solution. Both dipotassium hydrogen phosphate (1.0 M) and potassium dihydrogen phosphate (1.0 M) were prepared in deionized water and diluted to 100 mM with deionized water before use. SOD enzymes (30,000 units) were diluted with pH 7.40 phosphate buffer working solution to make 30 units/mL per vial and stored in a –20 °C freezer until use. 10X TAE was diluted with deionized water to 1X TAE solution. 2.2.4 HPLC analysis of oxidation product of HE with flavonols The reaction product from the oxidation of HE (50 µM) with flavonols (50 µM) was analyzed by HPLC. The mobile phase consists of 0.1% formic acid in HPLC water (v/v) (Solvent A) and CH3CN (Solvent B). Sample (20 µL) was injected into HPLC system with the column equilibrated with 10% solvent B. Gradient elution was employed with a linear increase in solvent B concentration from 10% to 40% in 5 minutes at a flow rate of 0.5 mL/min, followed by isocratic elution at 40% solvent B for another 30 minutes and then back to 10% solvent B over an additional 5 minutes and kept constant at 10% solvent B for 3 minutes. 33 Chapter 2 2.2.5 Reaction between HE, flavonol and potassium superoxide Equimolar (50 µM) flavonol (myricetin, quercetin) was each added separately to HE (50 µM) in pH 7.40 phosphate buffer in two vials, and shaken for 1 minute, after which potassium superoxide was added in excess to the reaction mixture in one of the vials, shaken well and left to react for 2 hours. Fluorescence spectra of the two reaction mixtures were obtained with an excitation wavelength of 485 nm and scanned over an emission wavelength from 540 nm to 700 nm. 2.2.6 Pro-oxidant assay procedure A serial dilution of 6Xs, 12Xs, 24Xs, 48Xs, and 96Xs was performed for the test samples with deionised water using Precision XS microplate sample processor from Bio-Tek Instruments, Inc (Winooski, Vermont, USA). Diluted samples (20 µL) were automatically pipetted into individual wells of a 96-well flat-bottom microplate, followed by dispensing 20 µL phosphate buffer working solution into all the wells. E+Br− solution was diluted with pH 7.40 phosphate buffer to 0.25, 0.5, 1, 2, and 4 µM. The microplate was incubated at 37 °C for 10 minutes, before HE working solution (31.7 µM, 160 µL) was dispensed into all the wells. The total liquid volume per well was 200 µL. Phosphate buffer control (20 µL) was also run together in the same microplate. SOD enzymes (30 units/mL, 20 µL) were also added in place of 20 µL phosphate buffer working solution to see if SOD enzymes will inhibit the reaction. After dispensing HE working solution, the microplate was shaken for 10 seconds with low intensity. Fluorescence intensity was then recorded every 3 minutes for 18 minutes at an excitation wavelength of 485/20 nm and an 34 Chapter 2 emission wavelength of 645/40 nm. The kinetic experiments were conducted by following the rate of oxidation of HE to E+. The rate of fluorescence produced in the presence of sample was denoted as V and this was related to the generation rate of radicals. The plot of the rate versus the concentration of sample gave a linear curve, where slope k′ is the apparent rate constant for the pseudo first order reaction with respect to the limiting reagent sample and indicates the pro-oxidant activity of the sample. The molar concentration of the E+ formed was obtained from a standard curve which plotted the authentic E+ concentration versus the fluorescence intensity under the same conditions. k′ values were determined in triplicates and reported as mean values ± SD. The E+ standard calibration curve was performed in duplicates. Ten phosphate buffer blanks were performed for each assay. Significant difference between the means was assessed by using the Student’s t-test (p < 0.05 as significant). Statistical analysis was performed using Microsoft Excel 2007. 2.2.7 UV-vis kinetics measurement of HE oxidation by myricetin Automatic pipetting was carried out to dilute the myricetin sample with deionized water to five different concentrations (300 M, 150 M, 75 M, 37.5 M, and 18.75 M) using Precision XS microplate sample processor from Bio-Tek Instruments, Inc (Winooski, Vermont, USA). Diluted samples (30 L) were automatically pipetted into the first half of a 96-well flat-bottom microplate in triplicates, followed by dispensing 30 L pH 7.4 phosphate buffer working solution and 240 L HE working solution (37.5 µM) into these 18 wells. The total liquid volume per well was 300 L. In the second half of the plate, E+ standards (120 M, 60 M, 30 M, 15 M, and 7.5 M) were 35 Chapter 2 prepared and 300 L were pipetted into the wells. Phosphate buffer control (300 L) acting as blank were also run together in the same plate. The microplate was then shaken for 20 seconds at an intensity of one. The kinetic experiments were monitored over a period of 2 hours at 37 C by following the rate of oxidation of HE to E+, whereby absorbance reading of E+ at 479 nm was recorded at every 3 minutes. The absorbance readings taken were first corrected from the blank readings and then converted to concentration of E+ based on E+ standard calibration curve. The turn over number (TON) of the reaction was calculated by dividing the concentration of E+ formed by the concentration of myricetin, at the time of the completion of the reaction (t = 45 min). 2.2.8 pH dependency of HE oxidation by myricetin Phosphate buffers at various pHs (4.60 to 8.51) were prepared by mixing 100 mM dipotassium hydrogen phosphate and 100 mM potassium dihydrogen phosphate in appropriate volumes according to the Henderson-Hasselbach equation using a pKa value of 6.865. Effects of pH on the reaction of HE with myricetin was then examined by mixing myricetin (40 µM) and HE (25.6 µM) in the phosphate buffers at various pHs and the reaction was monitored using the method described in the pro-oxidant assay procedure. 2.2.9 Myricetin decomposition in the presence and absence of HE The rate of decomposition of myricetin in pH 7.4 phosphate buffer working solution in the absence and presence of HE was monitored using a Shimadzu UV-1601 UV-visible spectrophotometer. In the absence of HE, myricetin (75 36 Chapter 2 µM, 150 µL) was added to 1350 µL phosphate buffer. In the presence of HE, myricetin (75 µM, 150 µL) and HE (37.5 µM, 1200 µL) was added to 150 µL phosphate buffer. The reaction mixtures were scanned from 200 nm to 700 nm at 0, 3, 6, and 9 minutes. Triplicates were performed. The absorbance of myricetin at 392 nm, where there was no interference from HE and E+, was obtained and plotted against time. 2.2.10 Determination of acid dissociation constants (pKa) of flavonols The pKa values of flavonols were determined experimentally by spectrophotometric method using Synergy HT UV-vis microplate reader from Bio-Tek Instruments (Winooski, Vermont). Phosphate buffers at various pHs (4.60 to 12.49) were prepared by mixing 100 mM dipotassium hydrogen phosphate and 100 mM potassium dihydrogen phosphate in appropriate volumes according to the Henderson-Hasselbach equation using a pKa value of 6.865. Samples (400 µM, 50 µL) were pipetted into individual wells of a 96well flat-bottom microplate, followed by dispensing phosphate buffer at various pHs (150 µL) into all the wells. The microplate was then shaken for 15 seconds with low intensity and the samples were scanned from 200 nm to 500 nm. From the absorption spectra of the flavonols at various pH values, the wavelength at which the conjugate base absorbs was determined and the absorbance at that particular wavelength was then plotted against pH to get a sigmoid curve. pKa occurs at the midpoint of the spectrophotometric titration curve and was determined mathematically. 2.2.11 Measurement of oxidation potentials of flavonols 37 Chapter 2 Flavonols (400 µM) were diluted to 100 µM in pH 7.40 phosphate buffer working solution in a three-necked round bottom flask. The solution was purged with nitrogen gas for 15 minutes. Cyclic voltammetry was performed by a three-electrode system, which comprised of a glassy carbon working electrode, a platinum counter electrode and a silver/silver chloride reference electrode. Voltammograms were obtained by scanning each solution once from –1.0 V to +0.6 V with a scanning speed of 50 mV/s, using a µ Autolab Type II potentiostat (Eco Chemie BV). Results were obtained by the program General Purpose Electrochemical System Version 4.7. The oxidation potentials were determined as anodic peak potentials in the region from –0.1 V to +0.6 V vs Ag/AgCl electrode. The oxidation potentials of the flavonols in pH 5.50 sodium acetate buffer were also determined. 2.2.12 Determination of DNA cleavage Phosphate buffer (100 mM) containing EDTA (10 mM) was prepared by mixing K2HPO4 (100 mM, 19.354 mL), KH2PO4 (100 mM, 5.646 mL), and EDTA (0.093 g). This stock solution was then diluted 10 times to give phosphate buffer (10 mM) containing EDTA (1.0 mM) and a few drops of NaOH (1.0 M) were added to adjust the pH to 7.40. pBR322 DNA (1.0 mg/mL) was then diluted to 0.10 mg/mL with pH 7.40 phosphate buffer (10 mM) containing EDTA (1.0 mM). Agarose gel (0.10 g) was completely dissolved in 1X TAE solution (100 mL) and was heated until boiling. The gel solution was cooled down to 60 °C before loading. Ethidium bromide (10 mg/mL) was diluted to 0.5 µg/mL by deionized water. The reaction mixture of 20 µL contained plasmid DNA (0.4 µg), CuSO4 (200 µM), sample (0 – 600 µM), pH 38 Chapter 2 7.40 phosphate buffer (10 mM) without EDTA and deionized water. DNA alone was run together as a control. The reaction mixture was then incubated at 37 °C for 2 hours. After incubation, 5 µL of a gel loading buffer containing 0.25% (w/v) bromophenol blue, 0.25% (w/v) xylene cyanole FF and 40% (w/v) sucrose in water was added, and the reaction mixtures were loaded into agarose gel. The gel was then loaded into the SVG-SYS Vari-gel MINI electrophoresis system from Scie-Plas which is filled up with 1X TAE solution. Gel electrophoresis was then run for 2 hours using Consort EV243 power supply from Scie-Plas with voltage set at 100 V, current at 100 mA and power at 10 W, followed by staining overnight with ethidium bromide. The gel was then viewed and photographed on a G:Box Chemi HR16 from Synoptics Ltd, using software GeneSnap from SynGene (version 7.00.01) and GeneTools from SynGene (v3.07.01) to quantitate the DNA cleavage in terms of the percentage of supercoiled (Form I), open circular (Form II), and linear (Form III) conformation. The percentage of DNA damage is expressed as (Form II + Form III) / total DNA x 100%. 2.3 Results and discussion 2.3.1 HE oxidation by myricetin Mixing equimolar amounts of myricetin (50 µM) and HE in pH 7.40 phosphate buffer for 2 hours at 37 °C, led to clean formation of E+ with no detectable amount of 2-hydroxyethidium (2-OH-E+), as revealed by HPLC analysis (Figure 2.2). Addition of KO2 to the HE solution under the same conditions yielded mainly 2-OH-E+ with E+ as the minor product [141]. We suspected that myricetin oxidation of HE may not involve superoxide anion. 39 Chapter 2 Yet, when equal molar amounts of the flavonol (myricetin, quercetin) was mixed with HE immediately before addition of excess KO2, E+ was the only observed product (Figure 2.3). The oxidation product of HE by superoxide was definitely altered in the presence of flavonol.     E+ A 2-OH-E+ C E+ B D Figure 2.2. HPLC chromatograms of E+, a potassium superoxide induced oxidation of HE and the product formed from oxidation of HE by myricetin. (A) Mixtures of HE (50 µM) and myricetin (50 µM) in phosphate buffer (pH 7.4). (B) HPLC trace of authentic E+ (50 µM) in phosphate buffer (pH 7.4). (C) Solution containing HE (50 µM) and KO2 (excess) in phosphate buffer (pH 7.4). (D) Same as in (C) but spiked with authentic E+ (50 µM). The detector was set at 290 nm.   Figure 2.3. Reaction products of HE with superoxide anion in the presence or absence of flavonol (myricetin, quercetin). 40 Chapter 2 To further establish the stoichiometric relationship of E+ formation with the concentration of myricetin, the progress of the reaction was monitored by fluorescence signal of E+ by converting the fluorescence intensity measured with time (Figure 2.4A) into molar concentration of E+ formed using the E+ standard curve (Figure 2.4B) to yield the kinetic plot of E+ formed from the reaction of HE with various concentrations of myricetin in Figure 2.5A. In the presence of a large excess of HE, the reaction was first order with respect to myricetin concentrations (Figure 2.5). The plot of the rate of E+ formation versus the concentration of myricetin gave a straight line and the slope k′ was the apparent rate constant of E+ formation, which indicates the pro-oxidant activity of myricetin to be 0.1938 min-1. To test if O2●− was involved in the oxidation, SOD was first mixed with HE before myricetin was added. In the presence of SOD, the oxidation rate decreased dramatically (Figure 2.5B). 4.2 3.6 3.0 2.4 1.8 1.2 0.6 A 12000 B y = 3279.5x - 257.81 R² = 0.9997 FL (a.u.) FL (k.a.u.) This result clearly demonstrated the involvement of O2●−. 8000 4000 0 0 4 8 12 16 Time (min) 0.0 1.0 2.0 3.0 E+ (µM) 4.0 Figure 2.4. (A) Kinetic traces with [HE] = 25.6 µM and variable myricetin concentration (■: 0.833 µM; ♦: 0.417 µM; ▲: 0.208 µM; ×: 0.104 µM) in phosphate buffer (pH 7.4) monitored by fluorescence at ex = 485 nm, em = 645 nm. (B) Standard calibration curve of fluorescence intensity (FL) vs E+ concentration. Results are mean±SD; n = 2. We noticed from Figure 2.5 that the formation of E+ was much greater with respect to the amount of myricetin added, indicating that myricetin might act as a catalyst for the reaction. This was verified independently by monitoring the decomposition kinetics of myricetin using UV-vis absorbance at 392 nm 41 Chapter 2 (Figure 2.6A). Indeed, myricetin decomposition was slowed down in the presence of HE, indicating that myricetin was regenerated during HE oxidation. By variation of the molar ratio between myricetin and HE, the best turn over number (TON) we measured was about 9 when 6.25% (mol) of 4.4 3.6 2.8 2.0 1.2 0.4 A V (µM min -1) E+ (µM) myricetin was used to catalyze the oxidation (Figure 2.6B). B y = 0.1938x + 0.0061 0.16 R² = 0.9946 0.12 0.08 + SOD 0.04 0.00 0 4 8 12 Time (min) 0.0 16 0.2 0.4 0.6 0.8 [Myricetin] (µM) 40 1.0 A B 30 0.9 E+ (µM) Normalised A392 Figure 2.5. Kinetics of the HE/myricetin reaction in phosphate buffer (pH 7.4) monitored by fluorescence detection at ex = 485 nm, em = 645 nm. (A) Kinetic traces with [HE] = 25.6 µM and variable myricetin concentration (■: 0.833 µM; ♦: 0.417 µM; ▲: 0.208 µM; ×: 0.104 µM). (B) Oxidation rate of HE as a function of myricetin concentration in the absence (♦) and presence (■) of SOD (3 U/mL). Results are mean±SD; n = 3. 0.8 0.7 TON = 9 20 10 0 0 2 4 6 Time (min) 8 0 20 40 60 80 100 120 Time (min) Figure 2.6. (A) Normalized absorbance of myricetin at 392 nm in the absence (▲) and presence (■) of HE (30 µM) in phosphate buffer (pH 7.4); initial [myricetin] = 7.5 µM. (B) UV-vis kinetic measurements at 479 nm during the reaction of HE (30 µM) with myricetin (■: 30 µM; ♦: 15 µM; ▲: 7.5 µM; ×: 3.75 µM; ●: 1.875 µM) over 2 hours. It is known that myricetin can generate free radicals under basic conditions due to deprotonation of myricetin to its phenolate, which is more sensitive to oxidation [162]. We also observed that the HE oxidation rate was highly pH 42 Chapter 2 dependent (Figure 2.7A). It is apparent that at pH below 6.5 the reaction rate is rather small but it increases dramatically, and reaches a plateau at pH higher than 8.0. This midpoint of this curve was found at pH 7.53, which is close to the reported pKa of myricetin (7.73) obtained by UV-vis spectrophotometric method as shown in Figure 2.7B [163]. The pKa values of other flavonols were also measured (Table 2.1). Consistently, we observed similar trends of pH dependency on HE oxidation. The increasing pH value leads to increasing concentrations of the phenolates, and hence increasing rates of radical formation through single electron reduction of dissolved molecular oxygen 6.0 0.60 A B 0.50 4.0 A324 V (x 103 AU min-1) [67]. 2.0 0.40 0.30 0.20 0.0 4.5 5.5 6.5 pH 7.5 8.5 4.5 5.5 6.5 7.5 8.5 9.5 10.5 pH Figure 2.7. (A) pH dependency of HE oxidation by myricetin expressed as the increase in rate of the fluorescence signal at 37 °C, buffered solution. Results are mean±SD; n = 3. (B) pKa curve showing the variation of absorption maximum of myricetin at 324 nm with pH. Auto-oxidation of flavonols was known to generate superoxide anion. Under physiological conditions, pyrogallol decomposes to a colored compound and the reaction is inhibited by SOD [164]. Taking advantage of this reaction, SOD-like activity assay was developed and applied to a limited extent in studying superoxide scavenging activity of natural products. Canada et al. measured the auto-oxidation rates of flavonols using oxygen consumption measurement at a concentration range of 150–450 µM and found 43 Chapter 2 that myricetin is the most reactive followed by quercetin, while kaempferol shows little activity [67]. The authors reported that the rate of polyphenol auto-oxidation is highly sensitive to the pH of the medium studied. In addition, added superoxide dismutase slows down the oxidation of myricetin, indicating that superoxide anion promotes the auto-oxidation of myricetin. We also observed the similar phenomenon as shown in Figure 2.5. Using HE as the probe, the sensitivity is increased by one hundred times. Intuitively, the electron transfer reaction from deprotonated myricetin to oxygen should not be affected by added SOD. However, it is possible that superoxide may accelerate the oxidation of myricetin by shifting the reaction towards myricetin quinone through hydrogen atom abstraction from the semiquinone. When SOD is added, the superoxide is removed from the system and slows down the oxidation (Figure 2.5B). Taken together, we propose a reaction mechanism of myricetin-catalyzed oxidation of HE as follows (Figure 2.8). The reaction of HE with the superoxide generated from the auto-oxidation of a myricetin anion leads to HE●+ as suggested by Zielonka et al. [165]. HE●+ may react with another superoxide to form 2-OH-E+ [166]. Alternatively, a myricetin semiquinone radical may abstract a hydrogen atom from HE●+ and regenerate myricetin and form E+. The reaction cycle stops when myricetin is irreversibly oxidized to myricetin quinone, which does not react with HE, as we found independently. The reaction products of HE with oxidants were nicely reviewed with illustration of the complexity of the products formed from typical molecular biological laboratory practices [167]. In a chemical system like ours, the reaction mixture is much less complex than that of a biological system and the 44 Chapter 2 presence of flavonols is enough to channel the reaction product to E+ for convenience of quantification purpose by using microplate reader instead of HPLC separation which can be quite time-consuming. OH OH H2N NH2 HO O N Et OH OH pH dependent OH O OH O- E+ HO H atom transfer O OH OH + -H OH H2N NH2 H O2 H2N N Et O NH2 H HO N Et O OH OH O myricetin O2 O2 OH OH O semiquinone HE HE O2 catalyst deactivation O OH H2N NH2 N Et O HO O OH OH OH O 2-OH-E+ Figure 2.8. Proposed reaction mechanism for the oxidation of HE in the presence of myricetin. 2.3.2 Quantification of pro-oxidant activity of flavonols In the presence of a large excess of HE as it is the case in our experiments, the oxidation of HE is pseudo first order with respect to myricetin and other flavonols as demonstrated in Figure 2.5B for myricetin. The slope of the kinetic curve is the pseudo rate constant for the flavonols, which also indicates their pro-oxidant activity and the results are listed in Table 2.1. To understand the structure-activity relationship, we measured the oxidation potentials (Epa) of flavonols at pH 7.40 and pH 5.50 using a three-electrode cyclic voltammetry system made up of a glassy carbon working electrode, a platinum counter electrode and a silver/silver chloride reference electrode. Myricetin has the lowest Epa values +0.030 V and +0.179 V (vs Ag/AgCl) at pH 7.40 and 45 Chapter 2 pH 5.50 respectively (Table 2.1). Therefore, deprotonation of myricetin enables it to react with dissolved molecular oxygen. The formation of E+ as the sole product of HE air oxidation promoted by flavonols makes HE ideal for quantifying radical generating pro-oxidant activity of flavonols. The dose response relationship between HE oxidation and the flavonol concentration provides a highly sensitive method to quantify their pro-oxidant activity. Under the assay conditions, the HE oxidation rate is pseudo first order with respect to the flavonol concentration and the apparent rate constant, k′, of the reaction can be defined as the radical generating prooxidant activity. Table 2.1. pKa1, oxidation potentials, and pseudo first order rate constants of the oxidation of flavonols by oxygen at 37 °C, pH 7.40. Flavonol -1 k′ (min ) pKa1 Oxidation Potential (V, vs Ag/AgCl) pH 5.50 pH 7.40 Myricetin 0.1938 ± 0.0152 7.95 +0.179 +0.030 Quercetin 0.0308 ± 0.0015 6.05 +0.256 +0.089 Kaempferol 0.0020 ± 0.0001 6.89 +0.324 +0.158 Galangin 0.0006 ± 0.0001 7.78 +0.500 +0.388 As shown in Table 2.1, the pro-oxidant activity increases with the increasing number of hydroxy groups on the B-ring, with myricetin being the most active. In contrast, galangin and kaempferol have no detectable prooxidant activity. This shows that the ortho-trihydroxy moiety on the B-ring is critical for the high radical forming pro-oxidant activity of myricetin. Thus radical generation is related to the position and number of hydroxy groups on the B-ring with ortho-trihydroxy moiety having higher activity than ortho- 46 Chapter 2 dihydroxy moiety. Our results are consistent with the earlier study by Canada et al. [67] and this enhances the validity of HE as a probe for rapid and convenient quantification of radical forming pro-oxidant activity. The prooxidant activity correlates with the measured pKa values of flavonols, which fall in the range of 7.95–6.05 (Table 2.1). At pH 7.40, there are significant amounts of deprotonated flavonols for all four compounds, but their reactivity with oxygen is apparently different, as revealed by their oxidation potentials (Epa), which increase from myricetin to galangin. The oxidation potentials at pH 5.50 (due to phenols) are all much higher than that at pH 7.40 (for phenolates), accounting for the lack of pro-oxidant activity under weakly acidic conditions, which is typically the case in plant tissues where these polyphenolic compounds are normally found. The pH switch, after these compounds enter the small intestine of animal and in the bloodstream, will activate the radical generating pro-oxidant activity of these compounds. The toxicology effects of such reactivity switch are worthy of further study. We found that, in test tube, myricetin is able to catalyze the oxidation of cysteine to cystine under physiological conditions. Whether it will catalytically lead to oxidative damage of ubiquitous glutathione in a biological system remains to be answered. 2.3.3 DNA cleavage activity of flavonols The native pBR322 DNA in its undamaged state exists in a supercoiled conformation (Form I), so it has a high electrophoretic mobility. Upon damage, the supercoiled conformation was converted to an open circular conformation (Form II), which has a lower electrophoretic mobility. At times, 47 Chapter 2 linear form (Form III) may also be formed, and this has an intermediate mobility between that of supercoiled and open circular conformation [168]. Hence analysis by gel electrophoresis will give information on the damage to DNA based on the different eletrophoretic mobility of the three conformations of plasmid DNA. DNA cleavage caused by flavonols in the presence of Cu(II) ion was examined by gel electrophoresis and the results are shown in Figure 2.9. Neither the flavonol nor Cu(II) ion alone caused apparent DNA damage under the experimental conditions (Figure 2.9 top Lane 9 and 10). In order to clarify the role of superoxide in DNA cleavage, DNA was exposed to 25, 50 and 100 µM KO2 in DMSO with 18-Crown-6. Results showed that DNA nicking caused by superoxide anion radical was comparable to the control, indicating that superoxide anion radical generated by flavonols does not induce direct DNA damage. Therefore, it can also be concluded that the semiquinone radical generated by flavonol oxidation does not cause direct damage to DNA. However, addition of Cu(II) ion to the flavonols enhanced the percentage of DNA cleavage as illustrated in Figure 2.9. At 25 and 50 µM of myricetin, supercoiled DNA was completely cleaved into Form II and III (Lane 1 and 2). For concentrations above 100 µM, supercoiled DNA was cleaved by about 83 – 95% into Form II DNA only. Band density analysis gives rise to the dose response of degree of DNA cleavage as shown in Figure 2.9 (bottom). Among the flavonols, myricetin showed the greatest percentage of DNA damage (100%) from 25 to 100 µM and decreased to about 83% at the highest concentration (600 µM). Next was quercetin, which showed a high extent of DNA damage at concentrations 50 µM and 200 µM. It was generally observed 48 Chapter 2 that the flavonols showed an extensive DNA cleavage at very low concentration (50 µM) except for galangin, and the ability of the flavonols on DNA cleavage in combination with Cu(II) ion increased at lower concentration and then decreased at higher concentration. Therefore, although radicals generated in the auto-oxidation of flavonols are not potent enough to cleave DNA, the presence of transition metal greatly enhances the DNA damage potency, which is positively correlated to the radical generation rate. In addition, our results show that the flavonols present at higher concentrations may act as antioxidants in preventing the DNA damage. 1 2 3 4 5 6 7 8 9 10 11 Form II Form III Form I DNA Damage (%) 100 80 60 40 20 0 0 100 200 300 400 Flavonol (µM) 500 600 Figure 2.9. (Top) Agarose gel electrophoretic analysis of pBR322 DNA damage induced by myricetin in the presence of Cu(II). DNA was incubated with myricetin (25. 50, 100, 200, 300, 400, 500, and 600 µM, lane 1 to 8) in the presence of CuSO4 (200 µM) in pH 7.4 phosphate buffer (10 mM) at 37 °C for 2 hours under aerobic condition; lane 9, 600 µM myricetin without Cu(II); lane 10, Cu(II); lane 11, pBR322 alone (control). (Bottom) Flavonol concentration dependent DNA damage in the presence of Cu(II) (■: myricetin, ♦: quercetin, ▲: kaempferol, Δ: galangin). 49 Chapter 2 It is remarkable but not surprising that the superoxide generating activity of the flavonols does not translate to their DNA damaging activity as all of them do not show any DNA cleavage property in the absence of transition metals. The superoxide and subsequently hydrogen peroxide formed from autooxidation of flavonols are not known to be DNA cleavage agents. However, in the presence of Cu(II) ion, the ability of DNA damage is ranked from myricetin > quercetin > kaempferol > galangin at low concentrations (< 300 µM). It was observed that the rate of DNA cleavage increases with increasing number of hydroxy groups on the B-ring of flavonols, which is consistent with their trend of pro-oxidant activity determined earlier, with myricetin being the most potent pro-oxidant and galangin being the least active pro-oxidant. The presence of ortho-trihydroxy and ortho-dihydroxy moiety on the B-ring was important in enhancing the DNA cleavage activity of the flavonol, which might be related to the ease of Cu(II) chelation at these positions. Ahmad et al. has also reported the activities of flavonoids on the cleavage of DNA in the presence of Cu(II), and their findings show that the rate of DNA degradation by various flavonoids is as follows: myricetin > quercetin > epicatechin > rutin ≈ apigenin > galangin [169], which is consistent with our findings except that they did not study on kaempferol to show a clear trend in correlating the number of hydroxy groups on the B-ring of flavonols with their DNA damaging ability. The role of Cu(II) ion in the DNA cleavage reaction by polyphenols has been extensively studied. Studies have shown that the DNA cleavage reaction by flavonols in the presence of Cu(II) is linked to the reduction of Cu(II) to Cu(I) due to auto-oxidation of flavonols, resulting in the generation of reactive 50 Chapter 2 oxygen species (eq. 2.1 - 2.2) [169,170,171]. Cu(I) then reduces superoxide anion radical to hydrogen peroxide with formation of Cu(II) (eq. 2.3). It was proposed that the main reactive oxygen species responsible for DNA cleavage is the hydroxyl radical produced in the Fenton reaction through the reduction of hydrogen peroxide by Cu(I) (eq. 2.4) [169,170]. It was shown that flavonols such as quercetin was able to bind to DNA and form a charge-transfer complex with Cu(II) which would decay under oxygen conditions and its decay was accelerated by DNA [95]. It was suggested that the binding of the Cu(II)-quercetin complex to DNA generated the hydroxyl radical in close proximity to the DNA backbone, that led to the DNA scission [95]. The overall chemical reactions involved in DNA damage are summarized in eq. 2.1 - 2.4 [168]. ArOH + Cu(II)  ArO• + Cu(I) + H+ ArO• + O2  oxidized quinone + O2●− Cu(I) + O2●− + 2H+  Cu(II) + H2O2 Cu(I) + H2O2  Cu(II) + HO• + HO− (2.1) (2.2) (2.3) (2.4) Above the concentration of 300 µM, the flavonols protect DNA from cleavage, with the exception of myricetin. We propose at high concentration, the flavonol effectively traps the hydroxyl radicals generated from Cu(I) and hydrogen peroxide to prevent the DNA from being cleaved. Under the assay conditions, myricetin would have been completely oxidized by the dissolved oxygen and thus cannot prevent DNA damage. Our results demonstrate the double edge nature of flavonol as pro-oxidant and antioxidant. 2.4 Conclusion 51 Chapter 2 In summary, we have demonstrated that HE is oxidized to E+ under the prooxidative effect of flavonols under physiological conditions and is a sensitive fluorescent probe for rapid quantification of the radical forming pro-oxidant activity of polyphenolic compounds. The pro-oxidant activity of flavonols was found to be related to the position and number of hydroxy groups on the Bring of flavonols with ortho-trihydroxy moiety having higher activity than ortho-dihydroxy moiety. The flavonols themselves do not cleave DNA even though they can generate superoxide radicals. In the presence of Cu(II) ions, the flavonols act as pro-oxidants at low concentrations but antioxidants at high concentrations. We are positioned to explore the application of HE as an ultrasensitive probe in quantifying the pro-oxidant activity of common beverages that are known to generate hydrogen peroxide as observed [70]. 52 Chapter 3 Evaluation of Pro-oxidant Activity of Different Tea Leaves Chapter 3 3.1 Introduction 3.1.1 Tea processing and its polyphenol composition Tea is an infusion of the leaves of the Camellia sinensis plant and it is presently the second most widely consumed beverage in the world, next to water. Tea is classified by the degree of fermentation and oxidation of the polyphenols in the tea leaves. Green tea is non-fermented, oolong tea is partially fermented, while black tea is fully fermented [172]. In green tea processing, the enzymatic activities are halted within 1 to 3 hours after harvest to prevent fermentation. Either pan-fixing (>180 °C) or steam fixing (100 °C) is used to fix the tea leaves because above 65 °C, polyphenol oxidase (PPO) activity becomes deactivated [172]. Further processing of the tea leaves during the rolling and drying stages results in the oxidation, hydrolysis, polymerization, and transformation of some tea polyphenols, therefore the polyphenol content of processed green tea is about 15% less than that of fresh tea leaves [172]. The composition and amount of tea polyphenols differs from cultivar to cultivar, and can be affected by many factors, including the harvesting condition. The polyphenols in green tea consist mainly of six groups of compounds, of which about 60-80% are flavanols (mainly catechins). 90% of the catechins are constituted by (-)epigallocatechin gallate (EGCG), (-)-epigallocatechin (EGC), (-)-epicatechin gallate (ECG), and (-)-epicatechin (EC) (Figure 1.3) in order of decreasing concentration in green tea infusions [173]. Green tea is typically consumed in East Asian countries and its extracts are widely used in dietary supplements due to its well-known beneficial health properties. 53 Chapter 3 In black tea processing, tea leaves are first left to wither after harvest. The following rolling step serves to release and activate the PPO for the fermentation step. In black tea, PPO is the key enzyme in the oxidation and condensation of catechins into theaflavins and thearubigins (Figure 3.1). This process is termed as tea fermentation and has been well studied. During fermentation, the B-ring of the catechins is oxidized by PPO to form oquinone. The o-quinones of two green tea catechins then rapidly react to form theaflavins and other oxidative products like thearubigins, which provide distinctive flavor and color to black tea [174,175,176]. Eleven theaflavins, which have a characteristic benzotropolone unit, have been reported. The relative proportions are theaflavin (TF)-18%, theaflavin-3-gallate (TFMG-a)18%, theaflavin-3’-gallate (TFMG-b)-20%, theaflavin-3,3’-digallate (TFDG)40%, isotheaflavin and theaflavic acid-4% as shown in Figure 3.1. The precursors of TF are EC and EGC, that of TFMG-a are EGCG and EC, that of TFMG-b are ECG and EGC, while EGCG and ECG react to form TFDG [175]. The theaflavic acids are formed when gallic acid (GA) reacts with catechin quinones and are present only in small quantities in black tea [177]. The thearubigins, which are more extensively oxidized and polymerized than theaflavins, have not been well characterized, but their dry weight is about 1020 times higher than that of theaflavins in black tea infusions [172]. After fermentation, the tea leaves are dried to cease the enzymatic activities [172]. Oolong tea processing combines green and black tea processing. The withering process is first carried out. Rotating, which is a special operation in oolong tea processing, causes damage to the leaf edges, and takes place at around 20–25 °C at 75–85% humidity. Fermentation hence begins from the 54 Chapter 3 leaf edges and spreads inwards. Before the whole leaf is fermented, the leaves are fixed at high temperature to deactivate the PPO. The processing ends with rolling and drying. The amount of theaflavin in heavily fermented oolong tea is about 10% of that in black tea [172]. Catechins and other homobisflavan compounds like oolonghomobisflavan, theasinensin and oolongtheanin have also been identified [172]. OH OH OH OR2 OH OH HO HO O O OH OH HO HO O O O O OH OH OR1 OH OH OH Theaflavin Theaflavin-3-gallate Theaflavin-3'-gallate Theaflavin-3,3'-digallate TF TFMG-a TFMG-b TFDG R1 H Gallate H Gallate Isotheaflavin R2 H H Gallate Gallate OH OH OH HO OH O OH OH HO HOOC O O OH OH OR HO O HO Epitheaflavic acid OH OH O OR O n Thearubigins Figure 3.1. Chemical structures of major theaflavins and other oxidative products present in oolong and black tea. 3.1.2 Health effects of tea The health effects of green tea, oolong tea and black tea have been well reported in literature but with conflicting results. Many epidemiological 55 Chapter 3 studies have been conducted to investigate the effects of tea consumption on the incidence of cardiovascular diseases and cancer in humans. Wiseman et al. reported on the protective effects of tea against cardiovascular diseases like mycocardial infarctions and stroke [178]. The protective effects of the green tea and black tea studied were attributed to the antioxidant properties and beneficial effects on vascular function. On the other hand, Higdon and Frei reviewed nine cohort studies that examined relationships between tea consumption and coronary heart disease (CHD), and concluded that two found inverse relation, one found positive relation, while the rest failed to find a relationship [ 179 ]. Similarly for the preventive effect of tea on cancer, Wiseman et al. reported that on the whole, the epidemiologic studies on tea consumption and cancer to date do not provide support that increased tea consumption has an anti-carcinogenic effect in human based on the results of the studies reviewed [178]. However, the anti-carcinogenic activity of tea extracts on rats has been demonstrated [173,180,181]. As reviewed by Tijburg et al., epidemiologic studies, in overall, do not provide conclusive evidence for a protective effect of tea consumption on the risk of cardiovascular diseases in humans, although several studies have demonstrated significant risk reduction of cardiovascular diseases in consumers of black and green tea [182]. 3.1.3 Pro-oxidant activity of tea The possible health benefits of tea are often attributed to the antioxidative and free radical scavenging activities of the polyphenols demonstrated in vitro. However, the correlation of tea consumption and disease prevention is less obvious in vivo, especially in human studies. Furthermore, there is also 56 Chapter 3 evidence that at pharmacological concentrations, in certain vulnerable populations and in certain diseases, polyphenolic compounds could have deleterious health effects [160]. It was found that tea catechins are unstable under cell culture conditions and undergo oxidative polymerization with the cogeneration of hydrogen peroxide [160]. Aruoma et al. also reported that some phenolic antioxidants accelerate the oxidative damage of DNA, carbohydrate and protein in vitro [ 183 ]. These pro-oxidant effects demonstrated in vitro suggests that tea catechins may also induce oxidative stress in vivo. Therefore, there is strong evidence that the tea polyphenolic compounds can act as both antioxidants and pro-oxidants. In addition, the stability of tea catechins and theaflavins was studied by Su et al. using HPLC [184]. It was found that the stability of both tea catechins and theaflavins is greater at lower pH of the phosphate buffer. However, at the same pH, the theaflavins are less stable than the catechins. This gives some indication that theaflavins auto-oxidize to a greater extent than catechins and may be more potent pro-oxidants. The observed pro-oxidant activity of tea extract is often attributed to the pro-oxidant effect of tea polyphenols present due to the auto-oxidation of tea polyphenols in neutral and alkaline pH. Roginsky et al. also noted that although the auto-oxidation of tea polyphenols has been reported in literature, the quantitative information on the autooxidation of tea extracts and the individual tea polyphenols is rather limited [185]. 3.1.4 Aims & objectives In this study, the pro-oxidant activity of tea polyphenols, green tea, oolong tea, and black tea extracts was assayed by a hydroethidine (HE) oxidation 57 Chapter 3 method which has been described in detail in Chapter 2. It was demonstated in Chapter 2 that flavonol reacts with HE to form only E+ as the oxidative product. A confirmation that HE when reacted with tea catechins and theaflavins formed only E+ as the oxidative product was also performed in this study. Therefore, in this study, the pro-oxidant activity of the tea polyphenols was measured by their ability to generate E+. This study aims to examine the effect of fermentation on the pro-oxidant activity of tea leaves. The pro-oxidant activities of the main tea polyphenols namely, EGC, EGCG, EC, ECG, TF, TFMG, and TFDG, were also measured to assess if the difference in pro-oxidant activities of the teas was due to the different pro-oxidant activities of the tea components. The tea catechins, theaflavins and gallic acid present in the tea samples were quantified using RP-HPLC to determine if there is a relation between the composition of polyphenols in the tea extract and the pro-oxidant activity of the tea extract. In order to better understand the mechanism of the observed pro-oxidant activity of the tea extracts and individual tea polyphenols, the pro-oxidant assay was carried out at the physiological pH of 7.40, as well as in an acidic medium close to the tea extract’s natural pH. In addition, DNA damaging prooxidant activitiy of tea catechins was examined using gel electrophoresis to correlate the DNA damaging ability with the pro-oxidant activity of tea catechins. (The following section on determining the pro-oxidant activity of tea catechins and DNA cleavage by tea catechins was done by Ms Wenie Chin as part of her honours year project and the section on determining the pro-oxidant activity of tea leaves was done by Ms Tan Ying Ying as part of her UROPs project.) 58 Chapter 3 3.2 Materials and methods 3.2.1 Materials (-)-Epicatechin (≥ 90%, HPLC, sum of enantiomers), (-)-epicatechin gallate (≥ 98%, HPLC from green tea), (-)-epigallocatechin (≥ 95%, HPLC from green tea) and (-)-epigallocatechin gallate (≥ 95%, HPLC from green tea) were purchased from Sigma-Aldrich, Inc (St Louis, MO, USA). Theaflavin (87.9%, HPLC), theaflavin-3,3’-gallate [mixture of theaflavin-3-gallate (TFMG-a, 65.4%, HPLC) and theaflavin-3’-gallate (TFMG-b, 19.5%, HPLC)], theaflavin-3,3’-digallate (88.0%, HPLC) were purchased from Chromadex, Inc (St Santa Ana, CA, USA). Gallic acid monohydrate was purchased from Acros Organics (New Jersey, USA). Folin-Ciocalteu reagent was from Merck (Damstadt, Germany). All other chemicals used were of analytical reagent grade. The tea samples were purchased from the local supermarket and were listed as follows in Table 3.1. Table 3.1. Name and origin of the tea samples. Type Tea Name 1 2 Green Tea 3 4 5 6 7 Oolong Tea 8 9 10 11 12 Black Tea 13 14 15 Chin Guan Hin Fujian Green Tea Gold Kili Green Tea Lipton Yellow Label Green Tea UJiIno TSUYU Japanese Green Tea ShiFeng China Lungching Sea Dyke Brand, China Fujian Oolong Tea Sea Dyke Brand, China Fujian Tie Guan Yin Rickshaw Oolong Tea Lipton Yellow Label Oolong Tea Dilmah Oolong Tea Lipton Yellow Label Tea Roma English Breakfast Tea Twining English Breakfast Tea Adam’s Peak English Breakfast Tea Dilmah English Breakfast Tea Origin Fujian, China Taiwan Indonesia Kyoto, Japan Zhejian, China Xiamen, China Xiamen, China China China Ceylon, Sri Lanka Indonesia Ceylon, Sri Lanka Kenya and India Ceylon, Sri Lanka Ceylon, Sri Lanka 59 Chapter 3 3.2.2 Instruments Fluorescence measurements were obtained with a Synergy HT microplate fluorescence reader from Bio-Tek Instruments, Inc (Winooski, Vermont), installed with the software KC4. HPLC analysis was done on a Waters 2695 HPLC system equipped with Waters 2996 photodiode array detector (Milford, MA, USA) and installed with Empower program. A C18 reversed-phase column (Waters Atlantis T3, 6 µm, 4.6 x 250 mm) was used throughout this study. The PDA acquisition wavelength range was set at 200–400 nm. 3.2.3 Preparation of stock solutions (-)-Epicatechin, (-)-epicatechin gallate, (-)-epigallocatechin, (-)- epigallocatechin, theaflavin, theaflavin-3,3’-gallate, theaflavin-3,3’-digallate and gallic acid (400µM) were prepared in MeOH and stored at –20 °C. Gallic acid (400 mg/L) was prepared in deionised water. Folin Ciocalteu solution was diluted 10X from the stock solution with deionised water. Sodium carbonate solution (30 g/L) was prepared in deionised water. All reagents were stored at 4 oC. Hydroethidine, pH 7.40 phosphate buffer working solution, pBR322 DNA, and 1X TAE solution were prepared as described in Section 2.2.3. Phosphate buffer stock solution of pH 5.60, 5.47 and 4.98 were also prepared by adding HCl to the pH 7.40 stock phosphate buffer solution. 3.2.4 Extraction of tea samples Each tea leaves sample (0.5 g) was packed into a filter bag. The filter bag was immersed in 50 mL of boiling tap water for 10 minutes. During the immersion period, the solution was removed from heat but subjected to 60 Chapter 3 continuous stirring by magnetic stirrer. The filter bag was then rinsed with tap water before immersing into another 50 mL of boiling water. Three extractions were carried out for each tea sample, which were then combined into a 250 mL volumetric flask after filtration with Whatman No. 4 filter paper. After cooling to room temperature, tap water was used to top up the solution to prepare 2 g/L of tea extract. Each tea sample (15 mL) was kept at –20 °C for a maximum period of 4 weeks. For the HPLC analysis, 8 g/L of green tea extract and 16 g/L of oolong and black tea extract were prepared similarly. 3.2.5 Pro-oxidant assay procedure The pro-oxidant activity of the tea catechins, theaflavins and tea extracts were assayed by the HE oxidation method using the procedure described in Section 2.2.6. To test for the pro-oxidant activity at the tea’s original pH, the phosphate buffer pH 7.40 used for the HE oxidation assay was replaced with phosphate buffer pH 5.60, 5.47 and 4.98 for green tea, oolong tea and black tea respectively. The pro-oxidant activity of theaflavins was also assayed at pH 5.60. The pH dependency of HE oxidation with EGCG was also examined using the procedure described in Section 2.2.8. 3.2.6 Quantification of polyphenols in tea samples Green tea samples (8 g/L), oolong tea and black tea samples (16 g/L) were analysed by HPLC. The injection volume was 20 μL and the gradient elution program as described in Su et al. was used [184]. The mobile phase consists of 0.1% formic acid in HPLC water (v/v) (Solvent A) and CH3CN (Solvent B). After the injection of the sample, solvent B was increased from 8% to 15% 61 Chapter 3 over 28 minutes, to 31% over an additional 52 minutes, and then back to the starting ratio over an additional 5 minutes. The flow rate was maintained at 1.0 mL/min. Standard calibration curves with at least five points were established for the tea catechins, theaflavins and gallic acid. Identification of polyphenols present in the tea samples were based on comparison of retention time of the unknown peaks to those of authentic standards of EGCG, EGC, ECG, EC, TF, TFMG, TFDG, and GA run under identical conditions. 3.2.7 Total phenolic assay procedure The total reducing power of the tea samples was assayed by the FolinCiocalteu reagent method. The same plate reader used for the pro-oxidant assay was used to measure the endpoint absorbance at 765 nm. The preheating temperature was set at 37 oC. A serial dilution of 6Xs, 12Xs, 24Xs, 48Xs, and 96Xs was performed for the test samples with deionised water using Precision XS microplate sample processor from Bio-Tek Instruments, Inc (Winooski, Vermont, USA). Gallic acid stock solution was diluted with deionised water to 200, 100, 50, 25, 12.5, and 6.25 mg/L to establish a 5-point calibration curve. Samples and gallic acid standards (20 µL) were pipetted into individual wells of a 96-well flat-bottom microplate, followed by dispensing 100 µL Folin reagent and 80 µL sodium carbonate stock solution. The microplate was shaken for 10 seconds with low intensity. Absorbance was then recorded every 5 minutes for 30 minutes to obtain 7 readings, of which only the last reading was used for analysis. The samples and gallic acid calibration curve were performed in duplicates. Twelve deionised water blanks were assayed in each plate. Data were 62 Chapter 3 expressed as milligram gallic acid equivalents per gram extract (mg GAE/g extract). 3.2.8 Oxidation product of HE with tea catechins and theaflavins The mobile phase, injection volume, elution gradient, and flow rate are identical to those used for the quantification of tea samples in Section 3.2.6. HE (50 μM) was reacted with EC, ECG, EGC, EGCG (50 μM), TF, TFMG, and TFDG (400 μM) in pH 7.40 phosphate buffer at room temperature for 6 hours and injected into the HPLC column. The detector wavelength was set at 290 nm. The chromatograms were compared with the chromatograms of E+ in buffer after 6 hours of incubation under similar conditions. 3.2.9 Determination of acid dissociation constants (pKa) of tea catechins The pKa values of tea catechins were determined experimentally by spectrophotometric method using the procedure described in Section 2.2.10. 3.2.10 Measurement of oxidation potentials of tea catechins The oxidation potentials of tea catechins were determined at pH 5.50 and 7.40 by using cyclic voltammetry according to the procedure described in Section 2.2.11. 3.2.11 Determination of DNA cleavage The DNA cleavage by the tea catechins were determined by using gel electrophoresis according to the procedure described in Section 2.2.12. 63 Chapter 3 3.3 Results and discussion 3.3.1 Oxidation product of HE with tea catechins and theaflavins From the RP-HPLC analysis, E+ was the only oxidative product observed when HE was incubated with TF, TFMG, TFDG, EC, ECG, EGC, and EGCG. This was similar to the results of the reaction of HE with myricetin to form only E+ as the oxidative product. The reaction mechanism in Figure 2.8 has been proposed to explain the results that in the presence of polyphenolic compounds, only E+ is generated in the oxidation of HE, in contrast to the 2OH-E+ reported as the only oxidation product of HE with superoxide radical [142]. The tea catechins with pro-oxidant activities attributed to reactions mainly on the B-ring also reacted similarly like myricetin to form E+ as only the oxidative product. 3.3.2 Quantification of pro-oxidant activity of tea catechins, theaflavins, gallic acid, methyl gallate, pyrogallol, and tea samples Since the only oxidation product of HE and tea polyphenols was E+, the fluorescence intensity measured was converted into E+ equivalence using the E+ standard curve as shown in Figure 2.4B. The pro-oxidant activity of the tea samples was found and shown in Table 3.2. The order of pro-oxidant activity of the tea samples is as follows: black tea > oolong tea > green tea. Since the pro-oxidant activity of tea extracts is generally attributed to the tea phenolic compounds, the pro-oxidant activity of tea catechins, theaflavins and gallic acid were also assayed and the results are shown in Table 3.3. For comparison purpose, pyrogallol and methyl gallate were also included in Table 3.3. 64 Chapter 3 Table 3.2. Pro-oxidant activity of tea samples at pH 7.40 in terms of pseudo first order rate constant k′. Type Tea 1 2 Green Tea 3 pH 5.60 4 5 6 7 Oolong Tea 8 pH 5.47 9 10 11 12 Black Tea 13 pH 4.98 14 15 pH 7.40 Name 3 -1 k' (x 10 min ) Average Chin Guan Hin Fujian Green Tea 1.2 ± 0.0 Gold Kili Green Tea 1.5 ± 0.2 1.3 ± 0.3 Lipton Yellow Label Green Tea 0.9 ± 0.0 UJiIno TSUYU Japanese Green Tea 1.4 ± 0.1 ShiFeng China Lungching 1.6 ± 0.1 Sea Dyke Brand, China Fujian Oolong Tea 1.6 ± 0.2 Sea Dyke Brand, China Fujian Tie Guan Yin 1.8 ± 0.0 2.2 ± 0.6 Rickshaw Oolong Tea 3.1 ± 0.2 Lipton Yellow Label Oolong Tea 2.3 ± 0.2 Dilmah Oolong Tea 2.3 ± 0.3 Lipton Yellow Label Tea 2.5 ± 0.2 Roma English Breakfast Tea 3.6 ± 0.3 3.3 ± 0.7 Twining English Breakfast Tea 4.1 ± 0.2 Adam’s Peak English Breakfast Tea 2.6 ± 0.2 Dilmah English Breakfast Tea 3.7 ± 0.2 Table 3.3. Pro-oxidant activity of tea catechins, theaflavins, gallic acid, methyl gallate, and pyrogallol at pH 7.40 in terms of pseudo first order rate constant k′. Sample k′ (x 103 min-1) EGCG 8.3 ± 0.6 EGC 3.6 ± 0.3 ECG 1.2 ± 0.3 EC 0.4 ± 0.1 Pyrogallol 12.3 ± 0.6 Methyl Gallate 1.8 ± 0.2 Gallic Acid 8.2 ± 0.3 TF 41.9 ± 7.3 TFMG 29.8 ± 1.0 TFDG 30.9 ± 1.6 65 Chapter 3 As shown in Table 3.3, at pH 7.40, the order of pro-oxidant activity is as follows: TF ≈ TFDG ≈ TFMG >> pyrogallol > gallic acid ≈ EGCG > EGC > methyl gallate > ECG > EC. EC exhibited the lowest ability to generate superoxide via auto-oxidation among the four tea catechins (k′ = 0.0004 min-1). An additional hydroxy group at C-5’ position in the B-ring in EGC caused a substantial increase of k′ value (0.0036 min-1) due to the pyrogallol moiety. Comparing the k′ values of epicatechin gallate esters, it can be seen that EGCG, with a pyrogallol moiety in the B-ring, also showed a greater ability to generate superoxide (k′ = 0.0083 min-1) than ECG (k′ = 0.0012 min-1). Therefore, the results implied that the additional hydroxy group at C-5’ position in the B-ring, forming the pyrogallol moiety for EGCG and EGC, increases the generating superoxide ability as compared to ECG and EC with only catechol group on the B-ring. This shows a good agreement with the results obtained in Chapter 2, which examined the ability of flavonols in generating superoxide anion. Myricetin, which has a pyrogallol moiety on the B-ring, exhibited a higher k′ value (0.1938 min-1) than quercetin, which has a catechol moiety on the B-ring (0.0308 min-1). In addition, it was found that pyrogallol alone shows a great ability to generate superoxide (k′ value = 0.0123 min-1). Hence, as compared with catechol group, pyrogallol group is more prone to undergo auto-oxidation and generate superoxide, resulting in a greater pro-oxidant activity. On the other hand, k′ value of EC was found to be much lower than its gallate derivative (ECG), and also EGC was found to have a much lower pro-oxidant activity than its gallate derivative (EGCG). This indicates that the addition of a gallate group at C-3 position in the C-ring of catechins also increases the pro-oxidant activity though the effect is less than 66 Chapter 3 the pyrogallol moiety on the B-ring. Results also showed that methyl gallate is inefficient in generating superoxide anion (k′ = 0.0018 min-1). Therefore, it can be concluded that the pyrogallol moiety on the D-ring in the gallate group is less susceptible to auto-oxidation in comparison to the pyrogallol moiety on the B-ring. Our results showed that theaflavins have higher pro-oxidant activity than tea catechins, which is consistent with the study done by Jovanovic et al. [186]. Through a pulse radiolysis study, it was found that the reduction potential of theaflavin radicals in neutral and alkaline medium were higher than that of gallocatechin radicals, explaining the greater pro-oxidant activity of theaflavins [186]. Jovanovic et al. also proposed that in the auto-oxidation of theaflavins, the benzotropolene moiety might be responsible for electron donation due to the existence of resonance forms [186]. This was supported by the observation that the influence of the dihydroxybenzopyran and gallate substituents on the theaflavin radical reduction potential was minimal. Consequently, the pro-oxidant activity of TF, TFMG and TFDG would not be expected to differ significantly as observed in our results. The greater tendency of the benzotropolene moiety to undergo auto-oxidation compared to the pyrogallol and gallate moiety would provide an explanation of the greater prooxidant activity observed by the theaflavins, even though this experiment was not carried out in this study. We can also attribute the higher pro-oxidant activity of the more heavily fermented teas to be due to the theaflavins present in the fermented teas, which contribute to the high pro-oxidant activity. However, before this claim could be made, the quantification of the tea catechins, theaflavins and gallic acid in the tea samples was performed. 67 Chapter 3 3.3.3 Quantification of major polyphenols in tea extracts The major polyphenols in the tea extracts were quantified by RP-HPLC. Calibration curves of the standards were established with identical RP-HPLC conditions over the concentration range from 10 to 640 µM for EC, ECG, EGC, and EGCG, while that of TF, TFMG-a, TFMG-b, and TFDG were established over 10 to 160 µM, while for GA, the calibration curve was established over 25 to 400 µM as shown in Figure 3.2. All calibration curves showed linearity with correlation coefficient > 0.99 over the concentration range. 1.0E+07 Peak Area 2.0E+06 8.0E+06 ECG TFDG 3.0E+06 TFMG-a TF 1.0E+06 GA TFMG-b EGCG Peak Area 0.0E+00 0 6.0E+06 40 80 120 160 Concentration (µM) 4.0E+06 EC 2.0E+06 EGC 0.0E+00 0 100 200 300 400 500 600 Concentration (µM) Figure 3.2. Calibration curves of GA, ECG, EGCG, EC, and EGC. Inset shows the calibration curves of TF, TFMG-a, TFMG-b, and TFDG. HPLC chromatograms of one representative green tea, oolong tea and black tea sample are shown in Figure 3.3. 68 Chapter 3 A EGCG ECG EGC EC   1.00 GA 5.355 6.545 0.75 AU 10.679 B 0.50 0.25 0.00 3.00 6.00 9.00 12.00 15.00 18.00 21.00 24.00 Minutes EGCG ECG EC 1.20 GA 0.60 8.510 5.422 AU 0.90 0.30 13.919   C 6.577 EGC 0.00 3.20 EGC EGCG 4.80 6.40 ECG 8.00 9.60 11.20 12.80 14.40 16.00 Minutes TFMG-a TF TFDG Figure 3.3. HPLC chromatograms of the tea extracts (A) Tea 2: Gold Kili Green Tea, (B) Tea 7: China Fujian Tie Guan Yin and (C) Tea 12: Roma English Breakfast Tea. Detector wavelength was set at 280 nm for the polyphenol detection and 217 nm for the gallic acid detection. Retention time of the assigned polyphenol has been confirmed with the spiking of the standard. 69 Chapter 3 Table 3.4. Composition of the tea extracts expressed as percentage of the dry weight of the tea leaves. Tea 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 [Polyphenol] (%) EGC EC EGCG ECG 1.022 3.469 1.998 4.519 0.619 0.425 0.414 0.183 0.505 1.992 0.231 0.650 0.444 0.029 0.592 0.419 0.762 0.648 0.878 0.224 1.956 4.459 3.108 2.462 3.892 1.400 1.475 0.224 0.580 0.348 0.245 0.632 0.182 0.205 0.060 0.743 0.850 2.354 5.027 0.275 0.162 0.073 0.154 0.134 0.304 0.609 0.181 0.132 0.052 0.110 0.135 1.000 0.946 1.815 1.272 0.808 1.515 1.008 0.046 0.144 0.300 TF TFMG-a TFMG-b TFDG Gallic acid 0.099 0.024 0.052 0.036 0.022 0.041 0.024 0.050 0.030 0.019 0.048 0.027 0.068 0.043 0.017 0.078 Total (%) 3.720 9.269 6.101 8.104 5.427 2.749 2.990 0.183 4.307 8.874 2.577 2.386 1.487 1.865 2.037 All the tea extracts, even those with the same degree of fermentation had different phenolic compound composition. This was expected as the exact composition depends on many factors, for example the exact tea cultivar, other than the degree of fermentation. However, some general trends could be observed from Table 3.4, which showed the composition of the tea extracts determined by HPLC in this study. The four tea catechins, EGC, EC, EGCG, ECG (in increasing order of retention time), were detected in all the green tea samples. Gallic acid is the main phenolic acid detected in the tea samples. However, the quantified amount for Tea 1 was much less than the other green teas. As expected, the amount of green tea catechins quantified in oolong tea and black tea was generally lower than that of green tea, as the tea catechins undergo oxidative polymerization during the fermentation process to form other compounds, for example theaflavins. Since the amount of EC present in green tea was already quite low, the even lower EC concentration could not be detected in some oolong teas and all the black teas. In particular, for Tea 8 70 Chapter 3 Rickshaw Oolong Tea, only EGC was detected, therefore Tea 8 was left out in the subsequent discussion. Theaflavins were also not detected in the oolong teas, probably due to the low concentrations. The concentration of theaflavins in even the most heavily fermented oolong tea has been reported to be about 10% that of black teas [172]. As expected, theaflavins were detected in black teas. However, the concentration of theaflavin (g/L) was only about 10–20% of the concentration of tea catechins as shown in Table 3.5. The retention time for TFDG and TFMG-b was rather close, therefore in some cases the UV-vis spectrum of the peaks were used to identify the two compounds. The concentration of gallic acid increases as the degree of fermentation increases, similar to that observed by Zuo et al. [187]. Table 3.5. Concentration of phenolic compounds (g/L) in the tea extracts and k′ of the tea extracts. Tea 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 [Polyphenol] (g/L) EGC EC EGCG ECG 0.1635 0.5550 0.3197 0.7230 0.0990 0.0680 0.0662 0.0293 0.0808 0.3187 0.0370 0.1040 0.0710 0.0046 0.0947 0.0670 0.1219 0.1037 0.1405 0.0358 0.3130 0.7134 0.4973 0.3939 0.6227 0.2240 0.2360 0.0358 0.0928 0.0557 0.0392 0.1011 0.0291 0.0328 0.3766 0.8043 0.0440 0.0259 0.0117 0.0246 0.0214 0.0486 0.0974 0.0290 0.0211 0.0083 0.0176 0.0216 0.0074 0.0230 0.0480 TF 0.0038 0.0083 0.0058 0.0035 0.0066 TFMG-a TFMG-b TFDG 0.0038 0.0080 0.0048 0.0030 0.0077 0.0043 0.0109 0.0069 0.0027 0.0125 Sum of [polyphenols] Gallic acid (g/L) 0.0158 0.5952 1.4830 0.9762 1.2966 0.0096 0.8683 0.1189 0.4398 0.1360 0.4784 0.0293 0.1600 0.6891 0.1514 1.4198 0.2904 0.4123 0.2035 0.3818 0.1293 0.2379 0.2424 0.2984 0.1613 0.3259 -1 k'(min ) 0.0012 0.0015 0.0009 0.0014 0.0016 0.0016 0.0018 0.0031 0.0023 0.0023 0.0025 0.0036 0.0041 0.0026 0.0037 To determine if there is a relation between the composition of polyphenols in the tea extract and the pro-oxidant activity of the tea extract, the radical generation rate due to each phenolic compound present in the tea extract is calculated by multiplying the pro-oxidant activity of the phenolic compound, 71 Chapter 3 k′ (min-1), by the concentration of the phenolic compound in 16 g/L of tea extract. The total radical generation rate was obtained by multiplying the prooxidant activity of the tea sample by the concentration of the tea sample (16 g/L). The difference between the sum of the radical generation rate due to each phenolic compound and the total radical generation rate was divided by the total radical generation rate, and expressed as the percentage of total radical generation rate unaccounted for. The results were shown in Table 3.6. The high percentage of unaccounted radical generation rate leads to the performance of the total phenolic assay to estimate the amount of phenolic compounds unquantified by HPLC. Table 3.6. Radical generation rate due to each phenolic compound, the total radical generation rate of the tea extract and the percentage of unaccounted radical generation rate calculated based on 16 g/L of tea extract. Radical Total radical generation rate generation rate unaccounted -1 -1 EC EGCG ECG TF TFMG-a TFMG-b TFDG Gallic acid Sum (mg L min ) (%) 0.03 2.38 0.09 0.13 3.07 19.20 84.03 0.05 5.42 0.23 7.20 24.00 69.99 0.04 3.78 0.14 4.82 14.40 66.51 0.06 2.99 0.10 5.10 22.40 77.23 0.01 4.73 0.25 0.08 5.35 25.60 79.12 1.70 0.07 0.97 2.93 25.60 88.54 1.79 0.08 1.12 3.17 28.80 88.99 0.00 0.00 0.00 0.08 49.60 99.84 0.01 2.86 0.12 1.31 4.52 36.80 87.71 0.02 6.11 0.24 1.24 8.48 36.80 76.96 0.33 0.07 0.12 0.16 0.17 2.38 3.33 40.00 91.67 0.20 0.05 0.25 0.33 0.43 1.67 3.21 57.60 94.43 0.09 0.02 0.17 0.20 0.28 1.06 2.01 65.60 96.93 0.19 0.04 0.11 0.12 0.11 1.99 2.57 41.60 93.82 0.16 0.05 0.20 0.31 0.51 1.32 2.82 59.20 95.24 -1 -1 Radical generation rate due to each phenolic compound (mg L min ) Tea EGC 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 0.44 1.50 0.86 1.95 0.27 0.18 0.18 0.08 0.22 0.86 0.10 0.28 0.19 0.01 0.26 When the radical generation rate due to the individual tea polyphenolic compounds was calculated as shown in Table 3.6, the percentage of radical generation rate unaccounted for was significantly high (ranging from 66% to 99%), with black and oolong tea having more than 80% of their radical 72 Chapter 3 generation rate unaccounted for. Probably substantial amount of compounds in oolong and black teas were not detected and quantified by HPLC. As shown in Figure 3.4, considering all the tea samples as a group, there is no strong relation between the sum of radical generation rate contributed by the polyphenols quantified and the total radical generation rate of the tea samples. However, if we consider the teas as three groups, within each group, when the radical generation rate due to the accounted polyphenolic compounds is higher, there is generally a higher total radical generation rate observed. However, the rather low correlation coefficient (< 0.7) indicates that while there is a positive trend, the relation is not strongly linear. Radical generation rate due to accounted phenolic compounds (mg L-1 min-1) 10.00 Green Tea 8.00 Oolong Tea R² = 0.2489 R² = 0.5840 Black Tea 6.00 4.00 R² = 0.0004 2.00 0.00 0 20 40 60 Total radical generation rate (mg L-1 80 min-1) Figure 3.4. Relation of the total radical generation rate of the tea extracts with the radical generation rate due to the quantified polyphenolic compounds by HPLC. The radical generation rates are calculated based on 16 g/L of tea extract. Tea 1, Tea 8 and Tea 13, which fall out of the group, were left out in the plot. The total phenolic assay was performed on all the tea samples to determine the total phenolic content (TPC) of the tea samples so as to give an estimation of the percentage of polyphenols unaccounted for in the HPLC quantification. 73 Chapter 3 The concentration of the various quantified phenolic compound in each tea extract was expressed in terms of milligram gallic acid equivalents per gram extract (mg GAE/g extract) for comparison with the results obtained from the total phenolic assay which was also expressed in the same units. The results from the total phenolic assay were taken to be 100% of the phenolic compounds in the tea samples. The difference between the results from HPLC and total phenolic assay was divided by the total phenolic content and expressed as the percentage of unaccounted phenolic compounds. These results were shown in Table 3.7. Table 3.7. Comparison of the Gallic Acid Equivalents (GAE) to give an estimation of the percentage of polyphenols unquantified in the HPLC quantification compared to the total phenolic assay. Quantified TPC Phenolic compounds Tea polyphenols by HPLC (mg GAE/g unaccounted Type (mg GAE/g extract) extract) (%) 1 16.41 72.65 77.42 2 42.60 129.12 67.00 Green Tea 3 58.01 99.10 41.47 4 40.95 100.43 59.23 5 84.17 111.41 24.45 6 8.21 74.92 89.04 7 8.95 75.84 88.20 Oolong Tea 8 1.21 81.17 98.51 9 13.45 110.34 87.81 10 33.54 139.61 75.98 11 3.45 103.07 96.65 12 5.80 106.56 94.56 Black Tea 13 4.32 69.89 93.82 14 1.48 94.74 98.43 15 6.53 103.48 93.69 The comparison in Table 3.7 is only an estimation because despite its name, the total phenolic assay actually measures the total reducing capacity of a sample and is not only specific to phenolic compounds [188]. Furthermore, 74 Chapter 3 some phenolic compounds may have a reducing capacity greater than that of gallic acid, but this was not considered in the conversion of the phenolic content quantified by RP-HPLC into gallic acid equivalents. From Table 3.7, the percentage of unquantified phenolics was found to be very high, ranging from 59.2–98.5%. However, the high percentage of phenolic compounds unaccounted for does indicate that there might be significant phenolic compounds undetected in the RP-HPLC analysis. Since the quantification of the tea polyphenolic compounds in Table 3.5 is in good agreement with those reported in literature [187,189], it is unlikely that the low percentage of polyphenol accounted for is due to poor extraction. In green tea, the four quantified catechins accounts for 20–40% of the phenolic compounds. However, in the case of oolong tea, there are actually many unquantified catechin oxidative products, for example homobisflavan compounds oolonghomobisflavan A, B, oolongtheanin and theasinensin D, E, F, and G [172]. Theaflavin, which should be present in the partially fermented tea, was also not detected in our RP-HPLC analysis. In black tea, a major group of the polyphenol compounds, the thearubigins along with the other minor groups like theaflavic acid, were not quantified. In all three kinds of teas, the flavonols, flavones and phenolic acids other than gallic acids were also not quantified by HPLC. These compounds contribute to the reducing capacity of the tea, hence they would be detected by the total phenolic assay. Similarly, these compounds might also contribute to the pro-oxidant activity of the tea extracts. There is a weak association between the reducing capacity of the tea extracts as measured by the total phenolic assay, and the pro-oxidant activity 75 Chapter 3 of the tea extracts as shown in Figure 3.5. This suggests that the tea components do not contribute proportionally towards the reducing capacity and the pro-oxidant activity towards HE oxidation. A better relation may be obtained if the total phenols of the tea extracts were assayed using the method described by Stevanato et al. [190]. The enzymatic assay method using PPO is more specific to polyphenols. However, the reproducibility might not be as good as the total phenolic assay by Folin-Ciocalteu reagent. TPC (mg GAE/g extract) 140 R² = 0.8118 120 R² = 0.3487 R² = 0.4041 100 80 60 Green Tea 40 Oolong Tea 20 Black Tea 0 0 0.001 0.002 0.003 0.004 k' (min-1) Figure 3.5. Relation between the pro-oxidant activity and TPC of the tea extracts. Therefore, the unaccounted radical generation rate observed in the tea sample can be attributed to two factors: (1) the failure to quantify some polyphenolic compounds which might contribute significantly in terms of prooxidant activity, (2) the possible synergistic effect of the tea polyphenolic compounds in pro-oxidant activity. To test for the synergistic effect of the polyphenol, the polyphenols could be mixed together in proportions similar to their proportions in a tea extract and assayed for the pro-oxidant activity. 76 Chapter 3 3.3.4 Effect of pH on pro-oxidant activity of EGCG, theaflavins and tea extracts Reports from Jovanovic et al. have shown that the pro-oxidant activity of theaflavins and catechins is pH-dependent [186]. Since the oxidation of molecular oxygen to superoxide radical is dependent on the deprotonated polyphenol, it is of interest to investigate if the deprotonation of polyphenol and hence the generation of radicals (superoxide radical as well as polyphenol radical) would occur in the various tea extracts at their natural pH. The pH of all the tea samples was measured, and an average pH value was calculated for the different degree of fermentation. The pro-oxidant activity of green tea, oolong tea and black tea was assayed at pH 5.60, 5.47 and 4.98 respectively, and the results showed that the pro-oxidant activity of the teas at their respective pH was too small to be determined accurately. The pro-oxidant activity of theaflavins was also measured at pH 5.60 and compared with their pro-oxidant activity at pH 7.40 as shown in Table 3.8. Table 3.8. Comparison of pro-oxidant activities of theaflavins at pH 7.40 and pH 5.60. Sample k′ (x 103 min-1) pH 7.40 pH 5.60 TF 41.9 ± 7.3 0.6 ± 0.1 TFMG 29.8 ± 1.0 0.7 ± 0.1 TFDG 30.9 ± 1.6 1.4 ± 0.1 When the pH of the assay conditions was lowered, the pro-oxidant activity decreases for both the tea samples and the theaflavin standards. This provides 77 Chapter 3 evidence that it is mainly the deprotonated form of the phenolic compounds which reacts with molecular oxygen to result in the final E+ product. This result was also observed by Roginsky et al. [185]. It was hypothesized that the protonated polyphenols (o-quinones) also react with molecular oxygen to form superoxide at high pH. However, calculations of the enthalpy of this reaction show that this reaction is highly endothermic, and with rate constants of below 1.0 M-1s-1 at pH 7.40. The reaction of the deprotonated polyphenol with molecular oxygen is expected to have lower activation energy, and thus proceed at a higher rate [185]. In order to investigate the influence of pH on the pro-oxidant behavior of tea catechins, EGCG was used as a model and its flux rate of superoxide was monitored over a pH range from 4.5 to 10. As shown in Figure 3.6, it follows that flux rate of superoxide of EGCG increased with increasing pH of the medium. Under acidic condition (pH 5.5–6.5), the ability of catechins in generating superoxide was shown to be negligible. As the pH increased from 6.5 to 8, flux rate of superoxide increased markedly. After pH 8, flux rate of superoxide reached a plateau until pH 9, where a significant increase in flux rate of superoxide was again observed. More specifically, it follows that the generating superoxide ability of EGCG increased substantially at pH 6.5–8 and pH 9–10. Comparing this to acid dissociation constants of EGCG, it was found that these two pH ranges are close to the first and third dissociation constants of EGCG (7.55 and 9.43 respectively) determined by Inoue et al. [ 191 ]. This indicated that dissociated catechins show a greater ability to generate superoxide than undissociated ones. In addition, pH of brewed green tea measured at 33.3 °C was found to be 6.15. Therefore, major form of tea 78 Chapter 3 catechins present in green tea is in its undissociated form, which does not have pro-oxidative effect. V (x 103 AU min-1) 10.0 8.0 6.0 4.0 2.0 0.0 4.0 6.0 8.0 10.0 pH Figure 3.6. pH dependency of HE oxidation by EGCG in buffered solution at 37 °C, expressed by fluorescence increase rate. Results are mean±SD; n = 3. 3.3.5 Redox potential and acid dissociation constants of tea catechins Table 3.9 represents the acid dissociation constants and oxidation potentials of the tea catechins. The acid dissociation constants were found to be close to the literature reported values. Slight difference could be attributed to the differences in the source of the material and in the experimental method. As observed in Table 3.9, EGC and EC were found to have a higher value (8.50 and 8.38 respectively) than EGCG and ECG. A second dissociation constant was observed for EGC and EC (10.22 and 10.11 respectively) but it was not the case for EGCG and ECG. As compared to the literature reported values, the second dissociation constant of EGC and EC was found to be closer to the third dissociation constant instead of the second value. This is probably because the difference between first and second dissociation constant was not 79 Chapter 3 very large (less than 1) and the unit of pH increase was insufficiently small. As a result, second dissociation constants of catechins were not detected using UV-vis method. It was found that first proton dissociation constants of EGCG and ECG were close to that of methyl gallate (measured pKa value: 7.81) and were lower than that of EGC and EC. Their second proton dissociation occurred in almost the same pH range as the first proton dissociation of EGC and EC, which do not have gallate moiety in the structure. Therefore, the results suggested that proton dissociation of EGCG and ECG are likely to occur at the gallate group, and subsequently at the pyrogallol or catechol group. This observation is in consistency with Muzolf et al., who calculated the relative deprotonation energies (DE) of the hydroxy groups of the catechins [192]. The calculated DE values for deprotonation indicated that hydroxy groups on C-3’, C-4’ in B-ring, and/or C-4’’ position in gallate moiety (D-ring) are the ones that preferably deprotonate [192]. In general, it was observed that acid dissociation constants of the catechins fall within the physiological pH range, indicating that deprotonation of these compounds at physiological pH values needs to be taken into account when measuring their redox properties. It is well-known that cyclic voltammetry can be used to characterize the reducing ability and electrochemical behavior of polyphenols [193,194]. It has been used for the evaluation of antioxidant capacity of several polyphenols and their mixtures. However, no research on the relationship between the oxidation potential and pro-oxidant activity has been conducted so far. Hence the oxidation potentials of tea catechins were determined by cyclic voltammetry since the electrochemical studies reveal general trends in the electron-donating abilities of polyphenols. Table 3.9 shows that at pH 7.40 and 80 Chapter 3 scan rate of 50 mV/s, oxidation potential of EC has the highest value of +0.222 V among the four catechins. On the other hand, EGCG was found to have two oxidation peaks, the first at +0.065 V and a second at +0.162 V. The first oxidation peak of EGCG was found to correspond to that of EGC since its oxidation potential (+0.065 V) was close to EGC’s oxidation potential (+0.081 V), while the second oxidation peak of EGCG corresponds to that of ECG since its oxidation potential (+0.162 V) was close to ECG’s oxidation potential (+0.171 V). The results indicated that the first peak is attributed to the oxidation of pyrogallol moiety on the B-ring, whereas the second peak comprises oxidation reactions involving the gallate moiety at C-3 position on the C-ring. Since lower oxidation potential implies easier electrochemical oxidation, pyrogallol moiety is more susceptible to oxidation as compared to gallate moiety. These results support the observation that introduction of pyrogallol group to the catechin structure conferrred more potent pro-oxidant activity. Table 3.9. Measured pKa1 and oxidation potentials of tea catechins. Oxidation Potential (V, vs Ag/AgCl) Sample EGCG EGC ECG EC pKa1 pKa2 pKa3 pH 5.50 1 2nd st pH 7.40 1 2nd st 7.92 7.55b 8.74b 8.50 10.22 8.51b 9.38b 9.43b +0.220 10.90b +0.289 +0.081 +0.347 +0.171 +0.353 +0.222 +0.306 +0.065 +0.162 7.86 7.60b 8.80b 8.38 10.11 8.76b b Ref 191. 9.46b 11.10b 81 Chapter 3 The oxidation potentials of tea catechins were then examined at a lower pH 5.50. A similar voltammetric behavior was observed at both pHs. However, decrease in pH shifted the oxidation potential toward more positive value for all catechins as shown in Figure 3.7. A B D C Figure 3.7. Cyclic voltammograms for (A) EGCG, (B) EGC, (C) ECG, and (D) – – EC at pH 5.50 ( ) and pH 7.40 ( ). These results showed that the oxidation of tea catechins was dependent on pH. It was observed that, for the catechins studied, oxidation potential shifted toward more positive value at lower pH, indicating that they are oxidized more easily at higher pH. As pH increases to be greater than their acid dissociation constants, deprotonation of tea catechins occur readily, making electron transfer easier, and leading to a greater ability to generate superoxide anion. In other words, pro-oxidant activity is closely related to the acid dissociation constant of catechins. This shows a great consistency with the observation 82 Chapter 3 obtained by Mochizuki et al., who employed a peroxidase-based bioelectrochemical sensor of hydrogen peroxide and a Clark-type oxygen electrode to conduct continuous monitoring and kinetic analysis of the oxidation of catechins [195]. However, in their study, it was suggested that pH dependence of oxidation of the catechin is not simply interpreted from the acid dissociation constants alone. Stabilization of semiquinone and superoxide anion at increased pH might also be responsible for the pH dependence of oxidation [185]. 3.3.6 DNA damage induced by tea catechins Tea catechins, especially EGCG and EGC, were demonstrated to have a strong ability to generate superoxide. The relationship between their ability to generate superoxide and their pro-oxidative effect on DNA was examined. It was observed that all the tea catechins studied did not cause apparent DNA cleavage (Figure 3.8). 1 2 3 4 5 Form II Form I Figure 3.8. Agarose gel electrophoretic analysis of pBR322 DNA damage induced by catechins. DNA was incubated with catechins (600 µM) in pH 7.4 phosphate buffer (10 mM) at 37 °C for 2 hours under aerobic condition. Lane 1, pBR322 alone (control); lane 2, EGCG; lane 3, EGC; lane 4, ECG; lane 5, EC. Next, DNA cleavage caused by tea catechins (25–600 µM) was examined in the presence of Cu(II) ion (200 µM) under the same conditions. The electrophoretic analysis of DNA damage induced by tea catechins (300 µM) in the presence of Cu(II) ion (200 µM) was presented in Figure 3.9. Neither 83 Chapter 3 oxygen nor Cu(II) ion alone caused apparent DNA damage under the experimental conditions. However, presence of Cu(II) ion enhanced the DNA cleavage induced by catechins. As shown in Figure 3.9, supercoiled DNA was cleaved into progressively smaller heterogeneous sized fragments by both EGC and EC, leading to a faint tail at the lower part of the gel. EGCG also showed a complete DNA cleavage, with detectable formation of Form III DNA, while the ability of ECG to cleave DNA in the presence of Cu(II) ion was found to be the weakest (56.67%) among the tea catechins. 1 2 3 4 5 6 Form II Form III Form I Figure 3.9. DNA damage induced by catechins in the presence of Cu(II). DNA was incubated with catechins (300 µM) and CuSO4 (200 µM) in pH 7.4 phosphate buffer (10 mM) at 37 °C for 2 hours under aerobic condition. Lane 1, pBR322 alone (control); lane 2, Cu(II); lane 3, EGCG; lane 4, EGC; lane 5, ECG; lane 6, EC. The extent of DNA damage shown in Figure 3.10 was determined by fluorometric analysis with ethidium bromide. The extent of DNA damage varied at different concentrations of tea catechins. Moreover, it was found that all catechins showed an extensive DNA cleavage even at very low concentration (25 µM). At this concentration of 25 µM, the extent of DNA damage induced by catechins increased significantly to at least 50%. For 84 Chapter 3 EGCG, EGC and EC, as the concentration increased from 100 µM to 300–400 µM, the extent of DNA damage plateaued. DNA damaging activities of these catechins dropped as the concentration increased to above 300–400 µM. In the case of ECG, its ability to cleave DNA was found to decrease progressively as the concentration increased from 100 to 600 µM. In general, for EGCG, EGC and ECG, DNA damage increased at lower concentration and decreased at higher concentration. DNA Damage (%) 100 80 60 40 EGCG EGC ECG EC 20 0 0 100 200 300 400 Catechin (µM) 500 600 Figure 3.10. Catechin concentration dependent DNA damage in the presence of Cu(II). DNA was incubated with catechins (25–600 µM) and CuSO4 (200 µM) in pH 7.4 phosphate buffer (10 mM) at 37 °C for 2 hours under aerobic condition. In addition, DNA damaging activity of (+)-catechin was also investigated. Similar to other catechins, (+)-catechin alone did not induce apparent DNA cleavage. Upon addition of Cu(II) ion, DNA damaging activity of (+)-catechin was enhanced to a great extent as shown in Figure 3.11. From the comparison of (+)-catechin to EC, DNA damaging activity induced by EC was greater than (+)-catechin. 85 Chapter 3 1 2 3 4 5 Form II Form III Form I Figure 3.11. DNA damage induced by (+)-catechin and EC in the absence and presence of Cu(II) in pH 7.4 phosphate buffer (10 mM) at 37 °C for 2 hours under aerobic condition. Lane 1, pBR322 alone (control); lane 2, 600 µM EC; lane 3, 300 µM EC + 200 µM Cu(II); lane 4, 600 µM (+)-catechin; lane 5, 300 µM (+)-catechin + 200 µM Cu(II). The role of Cu(II) ion in the DNA cleavage reaction by tea catechins has been extensively studied [96,100,196]. Many of the studies have suggested that hydroxyl radical is the main reactive oxygen species that cause DNA cleavage [100]. It is proposed that the tea catechins may function to reduce Cu(II) to Cu(I), and the generated Cu(I) then produces reactive oxygen species such as hydroxyl radicals, which attack the DNA. It is known that Cu(II) can chelate to the ortho-dihydroxy moiety in the B-ring of catechins, from which molecular oxygen can react with the catechin through Cu(II), and generate superoxide and hydrogen peroxide, leading to DNA damage [100, 197 ]. Furukawa et al. suggested that, in addition to hydroxyl radical, Cu(I) may react with hydrogen peroxide to generate Cu(I)-hydroperoxo complex, which plays an important role in Cu(II)-mediated DNA damage induced by EGCG [96,198] as shown in eq. 3.1. Cu(I) + H2O2  Cu(I)OOH + H+  Cu(II) + HO● + HO− (3.1) As shown in Figure 3.10, DNA damaging activity of tea catechins in the presence of Cu(II) decreased in the following order: EGC > EC > EGCG > 86 Chapter 3 ECG. Our results is consistent with the study conducted by Hayakawa et al. on the DNA cleavage activities of EGC, EC, EGCG, and (+)-catechin in the presence of Cu(II), with EGC > EC > > EGCG > (+)-catechin [199]. It was also observed that DNA damage induced by catechins in the presence of Cu(II) increased at lower concentration and decreased at higher concentration, with the exception of EC. The observation is probably attributed to a combination of antioxidant and pro-oxidant action of catechins on DNA. At lower concentration, catechins undergo oxidation upon deprotonation, and ultimately, result in the generation of hydroxyl radical, which causes DNA damage in the presence of Cu(II) ion. Since the concentration of Cu(II) ion is kept constant, the amount of catechins becomes excessive as its concentration increases. Therefore, excessive catechins act as antioxidants by chelating metal ion or scavenging hydroxyl radical. The ability of catechins to scavenge hydroxyl radical was found to decrease in the order of ECG > EC > EGCG >> EGC [200]. On the other hand, the order of their ability to generate superoxide anion was found to be EGCG > EGC > ECG > EC. As a result, EGC being a potent pro-oxidant and weakest antioxidant causes a great extent of DNA damage, thereby degrading DNA into small sized heterogeneous fragments even at low concentration (100 µM). However, high DNA damaging activity of EC cannot be simply interpreted by this. In addition, it was found that EC was more active than (+)-catechin in inducing DNA cleavage although they are stereoisomers. It is possible that the EC or catechin radicals may be involved in DNA cleavage. 3.4 Conclusion 87 Chapter 3 In summary, the pro-oxidant activity assay with HE as the fluorescent probe has its advantage in its sensitivity. The pro-oxidant activity of the tea extracts was found to increase with the degree of fermentation of the tea leaves in the order: black tea > oolong tea > green tea. Similar to flavonols, we have demonstrated that the presence of pyrogallol group on the B-ring is important for the high pro-oxidant activity of EGCG and EGC. The pro-oxidant activity of tea catechins, theaflavins and tea samples are pH-dependent. Their prooxidant activity increases with increasing pH of surrounding medium. The tea catechins themselves do not cleave DNA, even though they can generate superoxide radicals. DNA damaging activity of tea catechins is enhanced in the presence of Cu(II) ion. This pro-oxidant activity of catechin is believed to be related to the generation of hydroxyl radical through the reaction of hydrogen peroxide and copper ion, or through the formation of Cu(II)-catechin complex. At higher concentrations, catechins behave as antioxidants presumably by scavenging the formed hydroxyl radicals or acting as metal chelators. In conclusion, tea catechins, which are commonly known as antioxidants, can also exert pro-oxidative action, indicating that they could be more of an oxidative risk than benefit under certain conditions and in certain tissues. Therefore, the bioavailability of tea catechins, which should also be investigated in further studies, becomes a critical issue. Prudent consumption of large amounts of tea catechins in the form of supplements or in tea samples is encouraged until the bioactivity of these compounds is better understood. 88 Part II: Bimetallic Complexes of Ruthenium and Iron as Near-IR Probes for Detection of Redox-Active Molecules Chapter 4 Literature Review on NIR Active Bimetallic Complexes of Ru and Fe Chapter 4 4.1 Introduction on NIR active metal complexes Redox active transition-metal complexes, which absorb strongly in the near-infrared (NIR) region (750–2500 nm) due to a change of oxidation state, are known as NIR active metal complexes [201]. A lot of attention has been turned to these NIR active metal complexes in recent years due to several reasons [201]. Firstly, transition-metal complexes usually display reversible electrochemical behaviour. Secondly, most of these complexes show very strong electrochromism in the NIR region due to charge-transfer transitions, whereby a simple one-electron redox change will result in a huge change in the intensity of the NIR absorption band. Thirdly, it is possible to fine-tune the metal complexes by modifying the ligand substituents so as to have effective control of the redox potentials and the absorption maxima. Finally, the complexes with substituents can be functionalised for their incorporation into thin films by either polymerization or adsorption onto a metal oxide surface. The NIR region has attracted particular interest due to the ‘telecommunication window’, whereby the transmission of the optical signals through the silica fibre-optic cables lie in the 1300–1500 nm region, where silica is transparent and will not weaken the signal. Besides that, the NIR active metal complexes are also receiving increasing attention for their applications in solar cells [ 202 ], optical signal processing [201, 203 ] and medical applications [204,205]. Because of these useful applications, it is of significant importance to understand the chemistry behind their strong NIR absorption. There are generally three main approaches behind their strong NIR absorption, which are due to (i) coordination by radical ligands [206], (ii) metal-metal mixed valency [ 207 , 208 , 209 ], and (iii) mixed-valency of 89 Chapter 4 coordinated radical ligands, where selected examples will be discussed in detail in the next section. The spectroscopic methods used for the detection and characterization of the NIR active complexes generally include electron paramagnetic resonance (EPR), vibrational spectroscopy (IR, Raman), and electronic spectroscopy (UV-vis-NIR). Most research studies carried out on NIR active metal complexes are mainly focused on FeIIFeIII and RuIIRuIII systems in the d6/d5 configuration, due to the importance of iron and ruthenium in biochemistry and catalysis [210]. In addition, the two oxidation states (II and III) of iron and ruthenium are thermodynamically stable and the redox potentials for II/III conversion can be determined conveniently [210]. Hence the NIR active metal complexes discussed in this literature review will focus on dinuclear iron and ruthenium complexes for a clear understanding of their chemistry, so as to avoid the complications of coordination compounds of higher nuclearity such as tri-, tetra- or hexanuclear complexes. 4.2 NIR absorption by metal complexes containing radical ligands Metal complexes containing radical ligands generally have strong NIR absorption arising from low-energy shifted metal-to-ligand charge-transfer (MLCT), ligand-to-metal charge-transfer (LMCT) or intra-ligand (IL) π-π* transitions. Singly occupied molecular orbitals (SOMOs) of the radical ions are formed when they undergo one-electron oxidation to form radical cations or one-electron reduction to form radical anions. Low-energy transitions occur when the fully occupied MOs in radical cations transit to the half-occupied 90 Chapter 4 SOMO (former HOMO) or when the half-occupied SOMO (former LUMO) in radical anions transit from the SOMO to empty molecular orbitals lying near the SOMO as shown in Figure 4.1 [210]. Figure 4.1. NIR absorption of radical ions [210]. An example of a small molecule which acts as a potential ligand for metal coordination is the radical anion of 2,2’-bipyridine (bpy) [211, 212], which will display NIR absorption due to π-π* transitions. Other examples include the ruthenium(II) dioxolene complexes in which the ligand ions undergo oxidation and give rise to NIR absorption due to MLCT transition [213]. 4.2.1 Iron(II)-2,2’-bipyridine complexes The Fe(II) complex of 2,2’-bipyridine, [Fe(bpy)3]2+ undergoes a single oneelectron reversible oxidation which is metal-centered, and three reversible oneelectron reductions which are ligand-based [211]. The UV-vis-NIR spectra of the dicationic parent complex show two main absorption bands at ~33 000 cm1 (due to π-π* transition of the coordinated diimine ligand) and at ~20 000 cm-1 (due to MLCT transition) [214]. There is no NIR absorption observed for the 91 Chapter 4 parent Fe(II) complex. However, the reduced Fe(II) complexes of 2,2’bipyridine, namely [Fe(bpy)3]+ and [Fe(bpy)3], display three intra-ligand π-π* transitions in the UV-vis-NIR regions at ~27 000 cm-1 [π(6)→π(7)], 18 000 cm-1 [π(7)→π(10)], and 10 500 cm-1 [π(7)→π(8,9)], with the added electrons being localized on separate ligands [211]. These features are characteristics of the radical anion of 2,2’-bipyridine and complexation with metal does not effect much changes to the bands. The intensity of the band in the NIR region at 10 500 cm-1 is typically weak. 4.2.2 Ruthenium(II) dioxolene complexes Lever et al. found that the mononuclear Ru(II) complex [Ru(bpy)2(cat)] (cat = catecholate) has no absorption in the NIR region [213]. However, when the cat ligand ions in the Ru(II) complex undergo two reversible oxidations to form 1,2-benzosemiquinone anion (sq) first and then 1,2-benzoquinone (q) as shown in Figure 4.2 [201], a ‘hole’ created in the dioxolene ligand orbital leads to the RuII[d(π)] → sq(SOMO) and RuII[d(π)] → q(π*) MLCT transitions, with absorption bands at 890 nm and 640 nm respectively, both being quite intense with molar absorption coefficient, ε ≈ 10 000 M-1cm-1. These two oxidations are mainly ligand-centered in nature, with the Ru metal centre maintaining an oxidation state of +2 throughout the redox reaction. The interconversion between [Ru(bpy)2(cat)] and [Ru(bpy)2(sq)]+ will result in the reversible appearance or disappearance of the NIR absorption band at 890 nm, leading to good NIR electrochromic behaviour for these complexes [213]. 92 Chapter 4 Figure 4.2. Redox reactions of catecholate (cat), semiquinone (sq) and quinone (q) [201]. Figure 4.3. Four reversible redox interconversions of 1n+ (n = 0 – 4). [201]. To make the NIR transitions more intensive, Ward et al. showed that when two of the dioxolene ligands present in [Ru(bpy)2(cat)] are linked ‘back-toback’ by a conjugated bridging ligand to form complex [1]n+ (n = 0 – 4), the dinuclear complexes displayed very rich redox and spectroscopic behaviour [215]. Complex 1 showed four ligand-centred redox reactions, covering from the fully reduced bis-catecholate form to the fully oxidized bis-quinone form (Figure 4.3). Among the various forms of complex 1, the most interesting one is the bis-semiquinone (sq-sq) form, which is diamagnetic due to the bridging double bond between the two phenyl rings and the unpaired electron at each end of the semiquinone being shared equally between the two oxygen atoms. The highly conjugated structure for this bis-semiquinone (sq-sq) form allows 93 Chapter 4 the two unpaired electrons at each end to pair up, giving rise to a planar and extensively delocalized bridging ligand. This in turn results in a RuII → sq MLCT transition at 1080 nm, with ε = 37 000 M-1cm-1, which is more intense than that in the [Ru(bpy)2(cat)] as shown in Figure 4.4. Figure 4.4. Electronic spectra of [Ru(bpy)2(sq)]+ and [1]2+. [215]. However, this strong NIR absorption for the fully reduced (cat-cat) form shifts to 750 nm in the fully oxidized form. In the mixed-valence form [1]3+, the NIR absorption is red-shifted to 1225 nm and is equally intense. Hence this redox interconversion of complex 1 offers a lot more possibilities for NIR electrochromism than the mononuclear [Ru(bpy)2(cat)]. Another dinuclear complex [2]n+ (n = 1-4) is also made up of two chelating dioxolene-like binding sites, except that the bridging ligand is now 9-phenyl2,3,7-trihydroxy-6-fluorone [216]. Complex 2 shows three reversible, one- 94 Chapter 4 electron, ligand centered redox processes. Similarly, this ligand-centred oxidation creates ‘hole’ in the π-system, leading to a highly intense RuII → bridging ligand MLCT transition at 1237 nm (ε = 41 000 M-1cm-1) for [2]3+ as shown in Figure 4.5. The fully reduced form [2]+ has no absorptions at wavelength longer than 900 nm, which is consistent with the behaviour of fully reduced 1. Figure 4.5. Electronic spectra of [2]n+ in acetonitrile, the numbers 1, 2, 3, 4 refer to the charge n+. [216]. As illustrated in the above examples, it can be concluded that ruthenium(II) dioxolene complexes are polyelectrochromic, with a large number of stable oxidation states possible through the oxidation of the ligands, thereby resulting in the NIR absorption of the MLCT transitions to change in a redox-dependent manner. 95 Chapter 4 4.3 NIR absorption by mixed-valence dinuclear complexes Mixed-valence dinuclear complexes are made up of two metal centers, which exist in different oxidation states. They play an important role in the study of electron transfer processes. These mixed-valence complexes typically exhibit intense intervalence charge-transfer (IVCT) transitions in the NIR region from the electron-rich (donor) to the electron-deficient (acceptor) metal centers. These very intense IVCT transitions [217], which do not exist in the isovalent states, make the mixed-valence dinuclear complexes ideal for the preparation of NIR electrochromic materials and devices. 4.3.1 Classification of mixed-valence dinuclear complexes The structure of mixed-valence dinuclear complexes consists of two metal centers, which are either linked directly to each other or through a bridging ligand. The optical and redox properties of these complexes are dependent on the strength of the electronic interactions between the two metals. The degree of interaction between the two metal centers is dependent on the metal-tometal distances, the extent of conjugation in the bridging ligand, the nature of the ancillary ligands, the structure of the bridging ligand, and the solvent [218]. Robin and Day classified the mixed-valence compounds reported in literature into three broad categories according to the degree of electronic interaction between the two metal centers, based on their electronic spectra [207]. In class I compounds, the electronic interaction between the metal centers is negligible or weak so IVCT transition is not possible. Class I complexes show the optical and electronic properties of the individual metal site M1n and M2n+1 or M1n+1 and M2n. Class II compounds exhibit moderate 96 Chapter 4 electronic coupling between the metal centers and generally have broad, solvent-dependent, Gaussian-shaped IVCT bands. They show new optical and electronic properties in addition to that of the individual metal sites due to the possibility of electron transfer between the metal centers. However, due to the weak interaction, the complexes are either valence trapped or charge localized. In class III compounds, there is strong coupling between the metal centers, such that the odd electron is fully delocalized over the two metals, resulting in an averaging of the oxidation states of the two metal centers. As a result, class III complexes have new electronic and optical properties instead of the usual properties of the individual metal centers. Their IVCT bands are narrow, solvent-independent and highly intense [219]. Besides Robin and Day’s classification, another piece of work was contributed by Hush [220], on correlating the optical IVCT transition with the energy barrier to thermally activated electron transfer. Simple Hush theory was useful in determining the two most important electron-transfer parameters from the IVCT band shape, namely the reorganization energy λ and the electronic matrix coupling Hab between the metal centers. The relation between λ and Hab will determine the class to which a mixed-valence complex belongs. For a class I complex, the charge is completely localized (Hab = 0), so there is no intramolecular electron transfer between the two metal centers. When Hab is very small and there is no IVCT band observed, the complex can be assigned to be class I. When Hab < λ/2, the complex is weakly coupled and belongs to class II. When Hab ≥ λ/2, the complex is completely delocalized and belongs to class III. 97 Chapter 4 The Hush theory only applies to weakly interacting systems like class I and II, so care must be taken when applying this theory to class III systems [221,222]. In class III system, where the electron is fully delocalized, the IVCT band can no longer be considered to be the electron transfer from one metal center to the other, since the two metals share the electron equally. Creutz, Newton, and Sutin, have introduced a new optical method (CNS model) to determine Hab by analyzing the higher energy, and easier to recognize MLCT band in UV-vis spectroscopy. It has been applied in class II and III mixed-valence diruthenium complexes [223,224]. 4.3.2 Physical properties of mixed-valence complexes A physical property relating to the stability of mixed-valence intermediates is the comproportionation constant, Kc, as shown in eq. 4.1 - 4.3. c  Int 2 Red Ox 10∆ /59mV at 298 K; ∆ 2 1    4.2            RT ln Kc = F(∆E), F (Faraday constant) = 96485.33 C/mol (4.3) In mixed-valence dinuclear complexes, due to strong electronic interaction between the two metal centers, there will be reduction and oxidation occurring at the same time. The potential difference (∆E) between the two reversible waves (E1 and E2) reflects the thermodynamic stability of the mixed-valence state relative to the other redox states [225,226]. ∆E or Kc is an indicator of the 98 Chapter 4 electronic interaction between the two metal centers. If ∆E is close to zero, this indicates that the metals are not interacting and the complex belongs to class I. A small ∆E indicates that there is a weak interaction between the metals with a small value for the comproportionation constant, Kc, as shown in eq. 4.2. It is usually assigned to a class II valence-trapped complex. A large ∆E represents a completely delocalized system with a very large Kc, as shown in eq. 4.2. In this situation, the complex is stabilized and is assigned to class III [227]. There is strong coupling between the two metal centers that the electron becomes delocalized and the two metal centers have a single average valence state. 4.3.3 Mixed-valence complexes with bis-monodentate bridging ligands The famous mixed-valence ion, Creutz-Taube ion [(NH3)5Ru(µ- pz)Ru(NH3)5]5+ (3, py = pyrazine), was the first to be studied for its interesting properties [228]. The low-energy absorption in the NIR region is due to the result of the weak-to-moderate interactions between the two metal centers, while the intensity of the IVCT transition is determined by the metal orbital overlap [229,230]. One or more transitions may occur in the NIR or even midinfrared region for the classical d6/d5 system using the example of 3 as shown in Figure 4.6 [231,232]. Figure 4.6. IVCT and IC (interconfigurational) transitions in d6-d5 mixedvalence systems such as RuII-L-RuIII complexes [231]. 99 Chapter 4 3 is quite stable with a Kc value of ~ 107 and it exhibits an intense, lowenergy absorption band in the NIR region at 1570 nm (ε = 5000 M-1cm-1) [208,228,229]. This band is narrow, asymmetrical and solvent independent [229,230]. The insensitivity of the IVCT band towards the solvent variation has led to 3 being commonly assigned as a class III (delocalized) complex, but in fact, it seems to have the properties of both class II and class III, and is better assigned to be a class II-III mixed valence complex as the exchanging electron is localized [233]. Studies have shown that when the neutral NH3 donors are replaced by the anionic CN− π-acceptors in [(NC)5Ru(µ-pz)Ru(CN)5]5− (4), the comproportionation constant, Kc, decreases to ~ 104.7 in CH3CN and the NIR absorption band due to the IVCT transition is broadened [ 234 ]. Due to competition from both the cyanide ligands and the pyrazine bridge for the electrons of ruthenium, the Ru-Ru metal interaction via the pyrazine bridge is weakened, giving rise to a more localized valence state. 4 has a very broad (∆v1/2 = 4200 cm-1) and symmetrical IVCT band at 1760 nm (ε = 2600 M-1cm1 ), and is classified as a class II complex [234]. On the other hand, when the metal is changed from ruthenium to iron in [(NC)5Fe(µ-pz)Fe(CN)5]5− (5) in aqueous solution, the metal-metal interaction 100 Chapter 4 is weakened, since there is less efficient dπ-π*(pz) overlap for the 3d orbitals compared to the 4d orbitals [235,236]. This in turn leads to a lower energy shift of the IVCT band at about 1300 nm, which is of lower intensity (ε = 2200 M-1cm-1), but still broad (∆v1/2 = 4800 cm-1) and symmetrical [236]. Thus, 5 is classified as a class II complex. 4.3.4 Mixed-valence complexes with bis-bidentate bridging ligands Besides looking at complexes 3, 4, and 5, which have bis-monodentate bridging ligands, it is of interest to look at complexes with symmetrical bisbidentate bridging ligands since the chelating effect can make the complexes more inert towards ligand substitution. An example of a bis-bidentate bridging ligand, which also enables efficient metal-metal interaction, is the 1,2,4,5tetrazine derivatives. For instance, the bis(chelating) 3,6-bis(2-pyridyl)1,2,4,5-tetrazine (bptz) can react with two Ru metals to form a RuIIRuIII mixed-valence complex, 6, which has a very high Kc value of 1015 [237]. Even though 6 has a large Kc value, the intensity of the IVCT band is about 10 times weaker than that of 3. This may be due to the rigid chelate conformation in 6, whereby the dπ/ligand π*/dπ orbital overlap is not as favourable as in 3, hence resulting in a weaker IVCT band. 101 Chapter 4 (NH3)4 Ru (CN)4 Fe 5+ N N N N N N N N Ru (NH3)4 Fe (CN)4 7 8 (CN)4 Fe N 3- N N N N (CN)4 Fe 3- N Fe (CN)4 3- N N N N N N N N Fe (CN)4 9 10 Other complexes such as ammineruthenium {(µ-bpym)[Ru(NH3)4]2}5+ (7, bpym = 2,2’-bipyrimidine) [238] and cyanoiron {(µ-bpym)[Fe(CN)4]2}3− (8), {(µ-bptz)[Fe(CN)4]2}3− (9), {(µ-bmtz)[Fe(CN)4]2}3− (10, bmtz = 3,6-bis(2pyrimidyl)-1,2,4,5-tetrazine) [239,240,241] all have low intensity IVCT bands, which is a characteristic feature for d5/d6 mixed-valence systems bridged by bis-bidentate acceptor ligands. 4.3.5 Mixed-valence complexes with bis-tridentate bridging ligands Bis-tridentate bridging ligands such as π-acceptor 2,3,5,6-tetrakis(2pyridyl)pyrazine (tppz), 11 [242,243,244] or conjugated π system such as 12, which have two “terpy-type” binding sites, have been found in diruthenium mixed-valence complexes [245]. These complexes, which are bridged by πacceptor, have small Kc values of < 106 and moderately intense IVCT bands in the NIR region due to their fairly rigid conformation [243, 244, 245]. In addition, the tppz ligand cannot adopt a coplanar conformation because of CH/CH bond repulsion between the neighbouring pyridyl rings [243, 244]. On 102 Chapter 4 the other hand, cyclometalated analogues of 12 with π-donor bridges have very intense IVCT bands (ε > 20 000 M-1cm-1), but small Kc values of < 103 [246]. This is because the bis-meridional configuration at the metal centers in such complexes tends to favour the overlap of the orbitals for IVCT transition even though the Kc value is smaller due to large Ru-Ru distances. 4.3.6 Mixed-valence complexes with bis-tetradentate bridging ligands The more interesting mixed-valence complexes with a conjugated bistetradentate bridging ligand are a series of bimetallic tetraiminoethylenedimacrocycles, M2TIEDL4, where L is Cl−, CH3CN, dimethylformamide (DMF), M is Ru or Fe, and TIED = tetraiminoethylenedimacrocycle, synthesized by Spreer and co-workers as shown in Figure 4.7 [247,248,249,250,251]. The properties of M2TIEDL4 complexes are demonstrated by spectroscopic and electrochemical studies to be as follows [252]: (i) Three distinct oxidation states (II,II), (II,III) and (III,III) are accessible. (ii) Mixed-valence form (II,III) gives rise to a narrow and intense absorption band in the NIR region from 800–900 nm. This intense absorption band is assigned as an IVCT band (ε ranges from ~24 600–68 000 M-1cm-1). (iii) They are assigned as Robin and Day class III valence-delocalized complexes, with large comproportionation constants Kc of 1011–1015. 103 Chapter 4 Figure 4.7. Structure of bimetallic tetraiminoethylenedimacrocycle, (M2TIEDL4)n+, formula M2N8C20H36L4, where L = Cl−, CH3CN or DMF, and M = Ru or Fe [252]. In addition, the macrocyclic metal binding sites promote stability by inhibiting metal dissociation and the cross-conjugated π system facilitates electronic interaction between the metals [251]. High degree of mixing of metal and ligand orbitals contributes to coupling between the metals and the unusual optical properties [251]. As a result of that, the M-N(imine) bond length is very short, with the iron(II)-nitrogen bond length being 1.89 A, which is the shortest among the other iron(II)-nitrogen distances [247]. The properties of some selected examples of M2TIEDL4 complexes will be looked at in detail. 4.3.6.1 [Ru2(TIED)Cl4]Cl [Ru2(TIED)Cl4]Cl (denoted as [Ru2]+) was prepared by aerobic oxidative dehydrogenation of trans-[Ru(cyclam)Cl2]Cl (cyclam = 1,4,8,11- tetraazacyclotetradecane) [250]. This mixed-valence complex has a very narrow and intense absorption band at 805 nm in CH3CN, with a very high molar absorption coefficient of 68 000 M-1cm-1. The unpaired electron in the complex is completely delocalized (class III complex) as suggested from the low dipole moment changes upon excitation and the NIR band is the result of 104 Chapter 4 the delocalized intervalence transition (IT) [253]. Since this band does not shift when the solvent is changed, this indicates that it is solvent-independent. This again corresponds well to a Robin and Day class III valence-averaged complex [207], where the IVCT bands of class III complexes are generally narrow, highly intense and solvent-independent. [Ru2]+ displays two reversible waves at +0.27 V and +1.19 V relative to a Ag/Ag+ reference electrode [250]. E at +0.27 V is assigned to [Ru2] oxidation to [Ru2]+, while E at +1.19 V is due to oxidation of [Ru2]+ to [Ru2]2+ [250]. The large difference of 920 mV in the potentials gives a very large comproportionation constant Kc of 3.5 x 1015 [250]. This indicates that the mixed-valence [Ru2]+ is resonance stabilized and has a delocalized ground state [208]. The valence-averaged [Ru2]+ could be reduced electrochemically at 0.0 V to obtain neutral [Ru2], which could be easily oxidized by oxygen to regenerate [Ru2]+ (eq. 4.4) [250]. The reduced species [Ru2] also has an intense NIR band at 910 nm, with a molar absorption coefficient of 19 000 M-1cm-1. This band has been assigned to be due to a low energy MLCT transition. 4.3.6.2 [Fe2(TIED)(CH3CN)4]5+ Fully reduced, isovalent [Fe2(TIED)(CH3CN)4](ClO4)4●2CH3CN (denoted as [Fe2TIED(CH3CN)4]4+) was prepared by aerobic oxidative dehydrogenation reaction of cyclam with Fe(OH2)6(ClO4)2 in oxygenated CH3CN [248]. This dark green complex has three prominent absorption bands 105 Chapter 4 at 240, 340 and 874 nm in CH3CN. The intense absorption band at 874 nm in the NIR region has a high molar absorption coefficient at 24 600 M-1cm-1 and has been assigned as a low energy MLCT transition from the d orbitals of Fe to a low-lying π* orbital of the tetraiminoethylenedimacrocycle ligand [248]. This NIR band is highly solvent-dependent as it shifts to 1050 nm in DMF and 1005 nm in water [247]. The shifts are due to axial CH3CN ligand exchange by coordinating solvent molecules. [Fe2TIED(CH3CN)4]4+ complex shows two oxidation waves at +1.18 V and +1.68 V in dried CH3CN (vs NHE) [248]. The E1/2 value for the oxidation wave is also solvent-dependent as it shifts from +1.18 V in CH3CN to +0.70 V in DMF. This is consistent with the NIR band being assigned as a MLCT transition, as MLCT will show large solvent shifts. The [Fe2TIED(CH3CN)4]4+ complex can be oxidized at +1.25 V to yield a yellow mixed-valence [Fe2TIED(CH3CN)4]5+ complex, which has been characterized to be the first low-spin Robin and Day class III valencedelocalized complex [249]. The UV-vis absorption spectrum of the mixedvalence [Fe2TIED(CH3CN)4]5+ complex shows the shift of the 340 nm band, the disappearance of the NIR band at 874 nm and the appearance of new bands at 500 nm and 940 nm, with the latter NIR band having a very high intensity (ε = 27 000 M-1cm-1) [249]. Hush’s equation [220] was applied to the 940 nm band by assuming it as an intervalence transition band and the result indicates that the [Fe2TIED(CH3CN)4]5+ complex was delocalized [249]. In addition, the difference in the two oxidation potentials of the [Fe2TIED(CH3CN)4]4+ complex in dried CH3CN gives ∆E1/2 = 500 mV, which corresponds to a large comproportionation constant Kc of 1011, indicating that the complex is a valence-averaged class III complex [249]. Besides using electrochemical 106 Chapter 4 oxidation, the mixed-valence [Fe2TIED(CH3CN)4]5+ complex can also be generated from chemical oxidants such as Ce4+ or Cu2+ in CH3CN [251]. 4.3.6.3 [Fe2(TIED)(DMF)4]5+ [Fe2(TIED)(DMF)4]4+ was formed from [Fe2TIED(CH3CN)4]4+ in DMF due to axial ligand exchange [251]. The NIR band shifts from 874 nm in CH3CN to 1075 nm in DMF as the CH3CN ligands were replaced with DMF ligands [248]. The mixed-valence [Fe2(TIED)(DMF)4]5+ can be generated electrochemically from the isovalent complex and is stable in the presence of dioxygen. There is a blue shift in the λmax from 1075 nm to 883 nm for the mixed-valence species in DMF, but the shift is in the opposite direction for [Fe2TIED(CH3CN)4]4+ complex. [Fe2TIED(CH3CN)4]4+ is rather stable in aerated CH3CN solution. Unlike [Fe2TIED(CH3CN)4]4+, the DMF complex is unstable in aerated DMF solution and decomposes into two monomeric keto macrocycles as shown in eq. 4.5. The first oxidation wave of [Fe2(TIED)(DMF)4]4+ is at +0.24 V in DMF vs the ferrocene/ferrocinium electrode. Dioxygen will be able to remove the electron from the DMF species at the linking carbon-carbon double bond, resulting in two monomeric keto macrocycles. 4.3.6.4 [Fe2(TIED)Cl4]+ 107 Chapter 4 Yellow, isovalent tetraethylammonium [Fe2(TIED)Cl4] chloride was to formed by [Fe2TIED(CH3CN)4]4+ adding or [Fe2(TIED)(DMF)4]4+ and has an NIR band at 985 nm [251]. When dioxygen is injected into the yellow [Fe2(TIED)Cl4] in CH3CN or DMF, a pink, mixedvalence [Fe2(TIED)Cl4]+ is formed immediately in good yield [251]. [Fe2(TIED)Cl4]+ has a Kc value of 1.7 x 1013, which is larger than the Kc value of 2.8 x 108, observed for [Fe2TIED(CH3CN)4]5+ [251]. 4.4 NIR absorption from mixed-valency of coordinated radical ligands Another way to achieve NIR absorption is to have coordination compounds with non-innocent ligands, whereby the ligands have different oxidation states. In this way, electron transfer between the two mixed-valence ligands can result in low energy ligand-to-ligand intervalence charge transfer (LLIVCT) transitions. These LLIVCT transitions can give rise to high intensity absorption bands when the π systems are planar and when the ligand orbitals can interact with the d orbitals of the bridging metal. Examples of such a system include the neutral Fe(II) complex containing -diimines as the noninnocent ligands and the bis(dithiolene)iron(III) complex [254,255]. 4.4.1 Bis(-iminopyridine)iron(II) complexes The neutral bis(-iminopyridine)iron(II)complex, [FeII(L●)2]0 undergoes two ligand-centered, one-electron oxidations to form [FeII(L●)(L)]+ and [FeII(L)2]2+ respectively as shown in eq 4.6 and 4.7, where M = Fe [254]. The UV-vis spectrum of the oxidized species [FeII(L●)(L)]+ shows an intense broad 108 Chapter 4 band in the NIR region (~ 900–1400 nm), with ε = 3000 M-1cm-1, which is the result of the intramolecular LLIVCT transition. The cationic complex [FeII(L●)(L)]+ exhibits Robin and Day class III behaviour with its fully delocalized ligand radical. 4.4.2 Bis(dithiolene)iron(III) complex, [Fe2(pdt-p-tBu)4]z, z = 1+, 0, 1-, 2The dianionic bis(dithiolene)iron(III)complex, [Fe2(pdt-p-tBu)4]2−, where pdt = p-anisyl-1,2-dithiolate, undergoes three ligand-centered, one-electron oxidations to form [Fe2(pdt-p-tBu)4]−, [Fe2(pdt-p-tBu)4] and [Fe2(pdt-p-tBu)4]+ respectively as shown in eq. 4.8, where L = (pdt-p-tBu). [255,256]. The UV-vis spectrum of the oxidized species [Fe2(pdt-p-tBu)4]+ shows that the intensity of the charge transfer band at ~ 650 nm is increased when the dianionic complex is oxidized to the monocationic complex [257], which is the result of the intramolecular LLIVCT transition. Besides, the oxidation state of the iron remains +III throughout, indicating that the electron transfer processes shown in eq. 4.8 occur through reversible ligand-centered redox reactions. In addition, the infrared spectrum of the monoanion and neutral complexes show a new band at ~ 1164 cm-1, which is being assigned to the C=S● stretching 109 Chapter 4 frequency [257,258], which further confirms that the oxidation of the dianionic complex happens at the ligand site. 4.5 Applications of NIR active dinuclear complexes Redox-active transition-metal complexes, which display strong absorption in the NIR region for some oxidation states but not in others, are attracting interest for their commercial applications as NIR switchable electrochromic dyes for electro-optic switching, as sensitizers for solar cells, and for their use in smart windows which can filter out NIR radiation, while transmitting visible light. The application in electro-optic switching will be illustrated in detail with selected examples to understand the usefulness of these NIR active complexes in daily life. 4.5.1 Application in electro-optic switching NIR active metal complexes can be used as electro-optic switching materials, whereby an electrical input is applied to modulate light transmission in the NIR region. The first example is based on ruthenium(II) dioxolene complexes described in Section 4.2.2, for instance, bis(2,2’-bipyridine-4,4’dicarboxylic acid) (tetrachlorocatecholato)ruthenium(II) complex 13 [259,260]. It was designed with four carboxylic acid anchoring groups at the periphery of the complex and the dioxolene unit uses tetrachlorocatechol instead of catechol so that the former has a more positive redox potential within the conducting potential window of the Sb-doped SnO2 surface. The complex is adsorbed onto the surface via its carboxylic acid anchoring groups. The SnO2:Sb film containing the adsorbed complex 13 was then deposited on an 110 Chapter 4 indium-doped tin oxide (ITO) surface. When the applied potential is sweeped from negative to positive through the redox potential of the reversible cat/sq couple of the complex, a quick colour change from blue-gray to pink is observed, together with a decrease in transmittance at 940 nm of about 30% due to the RuII → sq MLCT transition, when the dioxolene ligand is oxidized. When the potential sweep is reversed, reversible changes are observed. This system is effective as a visible and an NIR electrochromic window based on a single cat/sq redox interconversion at a modest potential. The second example is based on mixed-valence dinuclear complexes described in Section 4.2.3. Asymmetric dinuclear complexes 14-17 have a carboxylate group attached to the terminal pyridine ligand, which is linked to the Ru-ammine unit, which has a RuIII/RuII redox couple potential close to 0.0 V [261]. 111 Chapter 4 Figure 4.8. Absorption spectral changes of the binuclear mixed-valence ruthenium complexes 14-17 as they undergo redox reactions in aqueous solution [261]. The MII-RuII complexes (reduced form) did not have any NIR absorptions, but when the Ru-ammine unit is oxidized to RuIII, a MII → RuIII IVCT transition is observed at between 763 nm (for 14) and 1149 nm (for 17) as shown in Figure 4.8. This behaviour is observed in both solution and when the complexes are adsorbed onto transparent SnO2:Sb film or ITO electrode via the carboxylate anchoring group. The switching is reversible with a switching time on the millisecond timescale. Since these complexes are able to display distinct colour changes in the different oxidation states, they can be used as electro-optic switching materials in both the visible and NIR region. 4.6 Aim of this research The aim of this research is to examine the versatile chemical reactivity of the mixed-valence bimetallic complexes, which includes redox reaction and 112 Chapter 4 ligand substitution. This research will focus on the synthesis of the bimetallic tetraiminoethylenedimacrocycles, M2TIEDL4, where L is Cl−, CH3CN, M is Ru or Fe, and TIED = tetraiminoethylenedimacrocycle, which have NIR active electronic absorption bands, and to examine the utility of these complexes for the sensing of redox-active molecules based on monitoring the strong NIR absorbance signal changes since oxidation or reduction can occur at the complexes influencing their NIR absorption band frequency. The frequency and the absorption coefficient of the NIR band are highly dependent on the nature of the metal, bridging ligand, and axial ligands, hence we expect that replacement of the axial ligands, L, in M2TIEDL4 with π-acceptor ligands through simple substitution reactions will alter the NIR band frequency, leading to new complexes. Their properties and reaction with redox-active molecules will be examined. The structural characterization of the new complexes will be carried out via routine tools including UV-vis-NIR, IR, and NMR spectroscopy. 113 Chapter 5 Air Oxidation of HS- Catalyzed by a Mixed-Valence Diruthenium Complex, a Near-IR Probe for HS- Detection Chapter 5 5.1 Introduction Compounds with strong near-infrared (NIR, 750–2500 nm) absorbance and/or luminescent property are highly sought after optical materials for a number of applications. These include photovoltaic devices [262], spectrophotometric analysis [263], and biomedical application by taking advantage of the deep-tissue penetration of NIR waves with low background interference (either absorbance or autofluorescence from the biological matrix) [264,265]. Mixed-valence transition-metal complexes may give rise to a strong NIR absorption band from intervalence charge-transfer (IVCT) transition [207]. Classical examples of these NIR dyes include Prussian Blue (Fe) and Creutz–Taube ion (Ru) [228]. Fundamental chemistry questions such as the synthesis, electronic spectra, and redox chemistry have been extensively studied on mixed-valence complexes [266,267], yet the application of these intensively NIR-active complexes is less studied and has great potential as molecular probes, NIR light harvest devices, and/or medical applications [210,263]. We were inspired by the mixed-valence diruthenium complex [Ru2TIEDCl4]Cl (TIED = tetraiminoethylenedimacrocycle, denoted as [Ru2]+, eq. 4.4), first reported by Spreer and co-workers because of its narrow and intense band at ~800 nm with a very high molar absorption coefficient of 68 000 M-1cm-1 [250]. The unpaired electron in the complex is completely delocalized (class III complex), as suggested from the low dipole moment changes upon excitation, and the NIR band is the result of the delocalized intervalence transition [253]. 114 Chapter 5 The valence-averaged [Ru2]+ could be reduced electrochemically at 0.0 V to obtain neutral [Ru2], which could be easily oxidized by oxygen to regenerate [Ru2]+ (eq. 4.4) [250]. One can take advantage of the reversible redox reaction accompanied with strong NIR absorbance signal changes for the reversible sensing of redox-active molecules. Indeed, documented herein are our results on using [Ru2]+ as a selective and sensitive NIR probe for the reversible sensing and quantification of hydrogen sulfide (H2S). H2S is a colorless and toxic gas well-known for its pungent odor of rotten eggs yet lesser known for its important biological role as a gaseous mediator for cardiovascular and neuronal health [268,269]. In the human body, H2S is formed endogenously from cysteine and homocysteine by two pyridoxal-5’phosphate-dependent enzymes, cystathionine-γ-lyase and cystathionine-βsynthetase [270,271]. HS− (the pKa of H2S is 6.76 [272]) has been detected at micromolar concentration in the lung, heart, brain and blood [273,274]. 5.2 Materials and methods 5.2.1 Materials All experiments were performed with analytical grade reagents. All chemical solvents and compounds were used without further purification. Cyclam, potassium aquapentachlororuthenate(III), sodium hydrosulfide, tris(hydroxymethyl)aminomethane, butylated hydroxytoluene (BHT), xylenol orange sodium salt, ammonium ferrous sulfate, sodium nitrite, sodium bicarbonate, potassium superoxide, cysteine, ascorbic acid, uric acid, and 5,5’dithiobis-(2-nitrobenzoic acid) (DTNB) were purchased from Sigma-Aldrich, Inc (St Louis, MO, USA). Hydrogen peroxide solution (30%), hydrochloric 115 Chapter 5 acid (37%), sodium acetate, potassium nitrate, potassium sulfate, and dipotassium hydrogen phosphate were from Merck (Damstadt, Germany). Potassium chloride was from Alfa Aesar (Heysham, Lancs, UK), potassium iodide was from GCE Lab Chemicals, potassium bromide was from Sino Chemicals (Singapore), Trolox, and cesium fluoride were from Acros Organics (New Jersey, USA), and Sephadex LH-20 was purchased from GE Healthcare BioSciences AB (Sweden). [Ru2TIEDCl4]+Cl- (denoted as [Ru2]+) was synthesized according to a literature method [250]. The purity of commercial sodium hydrosulfide (NaHS) was determined by iodometric titration with sodium thiosulfate to be 57.2%. NaHS2 was prepared according to a literature method with slight modifications as follows [275]. NaHS (21 mg, 0.183 mmol) was dissolved in DMF (1 mL) and S (6.3 mg, 0.196 mmol) was then added. The reaction mixture (20 µL) was withdrawn after 1 h and diluted with MeOH. The diluted NaHS2 in MeOH was further diluted in 50 mM Tris-HCl buffer (pH 7.40) and DCM and the UV-vis absorption spectra were obtained. Morpholin-4-ium-4methoxyphenyl(morpholino)phosphinodithioate, GYY4137 was synthesized according to a literature method [276]. 5.2.2 Instruments UV-vis absorption spectra were recorded with a Shimadzu UV-1601 UVvisible spectrophotometer. Spectra were acquired at room temperature in 1-cm path length 1.5 mL quartz cuvette. Electron ionization mass spectrometry (EIMS) spectrum was obtained with an Agilent 5975 DIPMS mass spectrometer from m/z 10 to 300. Electrospray ionization mass spectrometry (ESI-MS) 116 Chapter 5 spectra were obtained with a Finnigan/MAT LCQ ion trap mass spectrometer from m/z 50 to 1000. 5.2.3 Preparation of stock solutions [Ru2]+ was dissolved in deionized water to give a 0.651 mM stock solution and stored in the dark at –80 °C. Tris buffer (50 mM) was prepared in deionized water to 1.0 L and a few drops of 1.5 M hydrochloric acid were added to yield a pH 7.40 Tris-HCl buffer (50 mM). NaHS was dissolved in 50 mM Tris-HCl buffer (pH 7.40) to give a 12 mM stock solution and pipetted into vials of 500 µL per vial. The vials were then stored in the dark at –80 °C. The concentration of NaHS stock solution was determined by using its molar extinction coefficient of 7200 M-1cm-1 at 230 nm [277]. Stock solutions (0.12 M) of potassium salts of NO3−, Cl−, Br-, I−, SO42−, HPO42−, and sodium salts of NO2−, HCO3−, CH3COO−, and stock solutions (1.2 mM) of cysteine, ascorbic acid, uric acid, and Trolox were prepared in 50 mM Tris-HCl buffer (pH 7.40). Stock solution (0.12 M) of potassium superoxide was freshly prepared in DMSO before use. DTNB was dissolved in 100 mM phosphate buffer (pH 7.00) to give a 10 mM stock solution. GYY4137 was dissolved in deionized water to give a 15 mM stock solution and pipetted into vials of 120 µL per vial and stored in the dark in freezer at –80 °C. 5.2.4 Construction of standard curve of NaHS with [Ru2]+ NaHS (150, 300, 600, 900, 1200, 2400, and 4800 µM) were obtained by dilution from the 12 mM stock solution in 50 mM Tris-HCl buffer (pH 7.40). The different concentrations of NaHS prepared (20 µL) was then added to 16 117 Chapter 5 µM [Ru2]+ in 50 mM Tris-HCl buffer (pH 7.40) to give final concentrations of 2, 4, 8, 12, 32, 64, and 160 µM NaHS (V = 1.5 mL). The reaction mixtures were quickly mixed and the UV-vis absorption spectra were recorded after 1 minute. The absorbance ratio A895/A789 was then obtained and plotted against the concentration of HS−. Triplicates were performed and the results are expressed as mean ± SD. 5.2.5 Reaction of NaHS with [Ru2]+ under a nitrogen atmosphere This operation was carried out under a nitrogen atmosphere. NaHS (1200 µM) was obtained by dilution from the 12 mM stock solution in 50 mM TrisHCl buffer (pH 7.40). NaHS (40 µL, 1200 µM) was then added via syringe to [Ru2]+ (16 µM) in 50 mM Tris-HCl buffer (pH 7.40) in a specially designed air free glass cuvette (V = 3 mL) to give a final concentration of 16 µM NaHS. The reaction solution was then deoxygenated by freeze-pump-thaw for 3 cycles and filled with N2. UV-vis absorption spectrum of the reaction solution was then recorded at 22, 32 and 42 minutes. After 42 minutes, the reaction solution was then exposed to air for 20 minutes and the UV-vis absorption spectrum was recorded again. 5.2.6 Reusability of [Ru2]+ for HS− quantification [Ru2]+ (16 µM) was reacted with 8 µM NaHS in 50 mM Tris-HCl buffer (pH 7.40) at room temperature over time (V = 1.5 mL). Kinetic reading of the absorbance ratio A895/A789 of the reaction solution was then taken at every 8.5 seconds to obtain 55 readings. To test its reversibility, 8 µM NaHS was then added again at 10, 20 and 30 minutes to achieve a total of 4 additions. In the 118 Chapter 5 second experiment, the first addition of NaHS was kept at 8 µM but the second to fourth addition were 16 µM (final concentration). In the third experiment, the first addition of HS- is again 8 M and the subsequent additions (2nd to 4th) were 32 µM (final concentration). The absorbance ratio A895/A789 was plotted against time. 5.2.7 Oxidation of HS− catalyzed by 5% [Ru2]+ [Ru2]+ (16 µM) was added to NaHS (320 µM) in 50 mM Tris-HCl buffer (pH 7.40) and was left to react at room temperature over time (V = 1.5 mL). The progress of the reaction was monitored by UV-vis absorption spectra of the reaction mixture at 0, 10, 20, 30, 40, 50, 60, and 70 minutes. The absorption spectrum of the reaction mixture at 30 minutes was then compared with the absorption spectrum of Na2SO3 and NaHS2 prepared independently in Tris-HCl buffer. In a separate experiment, [Ru2]+ (16 µM) was reacted with NaHS (320 µM) in H2O (pH 7.40) for 1 day and EI-MS spectrum of the reaction mixture was obtained. 5.2.8 Measurement of H2O2 formed from HS− oxidation The ferrous ion oxidation-xylenol orange (FOX) reagent was first prepared by mixing 9 volumes of FOX-1 reagent (4.4 mM BHT in HPLC grade MeOH) with 1 volume of FOX-2 reagent (1 mM xylenol orange sodium salt and 2.56 mM ammonium ferrous sulfate in 250 mM sulfuric acid). The concentration of H2O2 stock solution was determined by using its molar extinction coefficient of 43 M-1cm-1 at 240 nm [278] and standard solutions of H2O2 (2, 4, 8, 16, and 32 µM) were then prepared from the stock solution and used to calibrate the 119 Chapter 5 FOX assay. The FOX assay was then used to measure the concentration of H2O2 generated from the HS− oxidation with 5% [Ru2]+ catalyst as follows: A mixture of [Ru2]+ (16 µM) and NaHS (320 µM) in 50 mM Tris-HCl buffer (pH 7.40) was left to react at room temperature over time (V = 1.5 mL). At 10, 20, 30, 40, 50, 60, and 70 minutes, the reaction mixture (150 µL) was taken out and added to FOX reagent (1350 µL), which was vortexed and incubated at room temperature for 30 minutes. The absorbance of the reaction mixtures and standards were measured at 560 nm. The absorbance values of the reaction mixtures were converted to concentration of H2O2 by using the standard curve of H2O2 with FOX reagent and were plotted against time. 5.2.9 Extraction of HS2− from HS− oxidation with 5% [Ru2]+ [Ru2]+ (97.5 µM) was added to NaHS (1.95 mM) in 50 mM Tris-HCl buffer (pH 7.40) (V = 0.7 mL) and was left to react at room temperature over time. At 30 minutes (where the concentration of H2O2 formed was maximum), the reaction mixture (350 µL) was taken out and acidified with HCl to pH 3, and extracted with DCM (350 µL). The faint yellow organic layer was then diluted 2 times with MeOH and adjusted to pH 9 with 6% NH3 (aq). The diluted DCM extract was further diluted in DCM to obtain its UV-vis absorption spectrum, which was compared with the UV-vis absorption spectrum of prepared NaHS2 in DCM. 5.2.10 Selectivity of [Ru2]+ towards anions and reductants In each separate experiment, 20 µL of the various solutions of 0.12 M of anion (NO2−, NO3−, Cl−, Br−, I−, SO42−, HPO42−, HCO3−, CH3COO−, and O2•−) 120 Chapter 5 or 1.2 mM of reductant (HS−, cysteine, ascorbic acid, uric acid, and Trolox) was added to [Ru2]+ (37 µL, 0.651 mM) in 50 mM Tris-HCl buffer (pH 7.40) to give a final concentration of 1600 µM anion or 16 µM reductant and 16 µM [Ru2]+ (V = 1.5 mL). The UV-vis absorption spectra of the reaction mixtures were measured after 1 minute. The absorbance ratio A895/A789 were then obtained and normalized to [Ru2]+ before any addition. 5.2.11 Replacement of axial Cl- ligands of [Ru2]+ with F− [Ru2]+ (0.99 mg, 0.00129 mmol) and cesium fluoride (9 mg, 0.0592 mmol) were left to stir at room temperature in anhydrous CH3CN or THF (1 mL) for 1 day under a nitrogen atmosphere. The reaction mixture was then analyzed by UV-vis spectrophotometry and ESI-MS. 5.2.12 Method validation The standard curve for NaHS with DTNB was obtained as follows: 10 mM DTNB stock solution (37 µL) was diluted in 50 mM Tris-HCl buffer to 247 µM. A series of concentrations of NaHS (2, 4, 8, 16, and 32 µM) were obtained by dilution and mixed with DTNB solution to a final volume of 1.5 mL. The absorbance at 412 nm was recorded after 1 minute and plotted against the concentration of HS−. Triplicates were performed and the results are expressed as mean±SD. Accuracy validation were performed on the DTNB and [Ru2]+ assay as follows: NaHS solutions (5, 10 and 15 µM) were each added separately in triplicates to DTNB (247 µM) and [Ru2]+ (16 µM) in 50 mM Tris-HCl buffer (pH 7.40). The amount of NaHS added was then measured using the DTNB and [Ru2]+ assay based on their standard curves. 121 Chapter 5 The percentage accuracy was expressed as a ratio: measured value/standard value x 100%. 5.2.13 Measurement of H2S release from GYY4137 using [Ru2]+ GYY4137 stock solution (240 µL) was diluted to 1.2 mM in 50 mM TrisHCl buffer (pH 7.40) (V= 3 mL) and the resulting solution was then deoxygenated by freeze-pump-thaw for 3 cycles under a nitrogen atmosphere and incubated at 37 °C. GYY4137 (250 µL, 1.2 mM) was then removed via syringe at 12, 26, 43, 58, 80, and 103 minutes and added to [Ru2]+ (1.25 mL, 19.2 µM) in 50 mM Tris-HCl buffer (pH 7.40) to give a final concentration of 200 µM GYY4137 and 16 µM [Ru2]+. The UV-vis absorption spectrum of each reaction solution was recorded after 1 minute. The concentration of H2S released from GYY4137 was calculated from the standard curve of NaHS obtained using the [Ru2]+ assay and was plotted against time. Triplicates were performed and the results are expressed as mean ± SD. 5.3 Results and discussion 5.3.1 Sensitivity of [Ru2]+ towards HS− The addition of different concentrations of NaHS in a Tris-HCl buffer solution (50 mM, pH 7.40) to a [Ru2]+ (16 µM) solution led to a quick (~1 minute) decay of the IVCT band at 789 nm and the concurrent appearance of a new peak at 895 nm due to [Ru2], as shown in Figure 5.1. The spectra also feature an isosbestic point at 840 nm, indicating a quantitative reaction. The plot of the ratio of absorbance at 895 nm and that at 789 nm versus the concentration of HS− gives a good linear response up to 16 M (inset of 122 Chapter 5 Figure 5.1), but it takes over 32 M HS− to completely reduce [Ru2]+. With this linear relationship, the quantification of HS− is realized using [Ru2]+ with a good limit of detection at 1.35 µM and the limit of quantification at 2.32 µM. 0 2.0 4.0 8.0 12.0 16.0 32.0 64.0 160.0 0.4 Absorbance 2.5 [HS-] (µM) 0.5 A895 / A789   0.3 0.2 2.0 1.5 1.0 0.5 0.0 0 40 80 120 160 [HS- ] (µM) 0.1 0 550 650 750 850 950 1050 Wavelength (nm) Figure 5.1. Absorption spectra of [Ru2]+ (16 µM) in a 50 mM Tris-HCl buffer (pH 7.40) recorded 1 minute after reactions with various concentrations of HS− (2, 4, 8, 12, 16, 32, 64, and 160 µM). Inset: Plot of the absorbance ratio A895/A789 vs [HS-]. There is a linear response in the range below 16 µM HS− with regression line, A895/A789 = 0.0729[HS−] + 0.1353, R2 = 0.9996. 5.3.2 Reversibility of the reaction of [Ru2]+ with HS− in the presence of oxygen When the reaction solution was kept at room temperature for an extended period (30 minutes), we observed clean re-formation of [Ru2]+. To further verify that the reversibility of the reaction is indeed caused by oxygen dissolved in the solution, we conducted the equimolar reaction of [Ru2]+ (16 µM) with NaHS under a N2 atmosphere. The NIR spectra show that [Ru2]+ is reduced to [Ru2] and remains reduced (Figure 5.2, dotted lines) for an extended period (42 min), indicating that the reaction between [Ru2]+ and HS- 123 Chapter 5 is not reversible per se. Upon exposure to air, [Ru2] is oxidized back to [Ru2]+ quantitatively (Figure 5.2, red and black solid lines). Therefore, [Ru2]+ may be utilized as a reusable molecular probe for HS− with the help of oxygen in the system. 0.5 [Ru2] + [Ru2] + + HS- (22min) Absorbance 0.4 [Ru2] + + HS- (32min) [Ru2] + + HS- (42min) 0.3 [Ru2] + + HS- (42min, then expose to air) 0.2 0.1 0 300 400 500 600 700 800 900 1000 1100 Wavelength (nm) Figure 5.2. Absorption spectra of [Ru2]+ (16 µM) in a 50 mM Tris-HCl buffer (pH 7.40) before and after the addition of HS− (16 µM) under a N2 atmosphere with time and upon exposure to air after 42 minutes. 5.3.3 Reusability of [Ru2]+ for HS− quantification The reusability of [Ru2]+ for HS− quantification was examined. When a limiting amount of HS- (8 µM) was added to a buffered solution of [Ru2]+ (16 µM), about 50% of [Ru2]+ was quickly reduced by HS− within 1 minute, and slowly oxidized, as demonstrated by a A895/A789 plot versus time plot (Figure 5.3). To our surprise, a second addition of 8 M HS− only reduces about oneeighth of the total [Ru2]+ revealed from the A895/A789 ratio. The third and fourth addition of 0.5 equiv of HS− led to results similar to those of the second addition (Figure 5.3, black line). When 16 M HS− were added for the second, 124 Chapter 5 third, and fourth round, 50% of the signal (A895/A789) was regained (Figure 5.3, red line). Full recovery of A895/A789 is achieved after the addition of 32 M HS- each time (Figure 5.3, blue line).   0.65 1st 0.5 0.5 0.5 0.55 A895 / A789 2nd 3rd 2 1 0.5 4th addition 2 1 0.5 equiv HS- 2 1 0.5 0.45 0.35 0.25 0.15 0 10 20 Time (min) 30 40 Figure 5.3. Catalytic circles of [Ru2]+ in the HS−/O2 system observed by the absorbance ratio changes recorded in real time of [Ru2]+ (16 µM) in a 50 mM Tris-HCl buffer (pH 7.40) after the first addition of HS− (8 µM) and the second, third, and fourth additions of HS− (8, 16, and 32 µM). 5.3.4 Oxidation of HS− catalyzed by 5% [Ru2]+ The progress of the reaction of HS− with 5% [Ru2]+ was monitored by UVvis absorption spectra of the reaction mixture at 0, 10, 20, 30, 40, 50, 60, and 70 minutes as shown in Figure 5.4. A new absorption band at 328 nm was observed. To gain more insight into the reaction, we analyzed the reaction products. Under 5% [Ru2]+ loading, HS− oxidation generates about 0.5 equiv of H2O2 per HS− oxidized, as quantified using a FOX assay, where the standard curve is shown in Figure 5.5 [ 279 ]. The concentration of H2O2 generated was observed to increase from 0–30 min and start to decrease after that (Figure 5.6). The maximum concentration of H2O2 observed at 30 minutes 125 Chapter 5 in Figure 5.6 is about 12.66 µM, which is about 2 times less than the concentration of HS−, indicating that 0.5 equiv of H2O2 is generated from 1 equiv of HS−. 1 Time (min) 0 10.0 0.8 Absorbance 20.0 30.0 328 nm 40.0 0.6 50.0 60.0 0.4 70.0 0.2 0 200 300 400 500 600 700 800 900 1000 1100 Wavelength (nm) Figure 5.4. Absorption spectra of HS− with 5% [Ru2]+ in 50 mM Tris-HCl buffer pH 7.40 with time.   1.6 1.4 y = 0.0461x + 0.095 R² = 0.9989 1.2 A560 1.0 0.8 0.6 0.4 0.2 0.0 0 4 8 12 16 20 24 28 32 [H2O2] µM Figure 5.5. Standard curve of H2O2 with FOX reagent. Results are expressed as mean±SD (n = 3). 126 Chapter 5   14.0 12.0 [H2O2] (µM) 10.0 8.0 6.0 4.0 2.0 0.0 -2.0 0 10 20 30 40 Time (min) 50 60 70 Figure 5.6. Concentration of H2O2 generated from HS− oxidation catalyzed by 5% [Ru2]+ with time. HS− was converted to H2S2 as indicated by the similarity of the UV-vis spectra of the reaction products and that of authentic NaHS2 prepared independently by following a literature method (Figure 5.7) [275]. In addition, the diluted DCM extract from the HS− oxidation with 5% [Ru2]+ also confirmed the formation of H2S2 as its UV-vis spectrum corresponds well to that of authentic NaHS2 in DCM (Figure 5.8). On the basis of these data, we proposed the following catalytic reaction mechanism: HS− + [Ru2]+  [Ru2] + S●− + H+ (5.1) [Ru2] + O2  [Ru2]+ + O2●− (5.2) S●− + HS−  [HSS●2−] (5.3) HSS●2− + O2  O2●− + HSS− (5.4) 2O2●− + 2H+  H2O2 + O2 (5.5) Net reaction: 2HS− + H+ + O2  HSS− + H2O2 (5.6) 127 Chapter 5 Firstly, a rapid single-electron-transfer reaction resulted in the formation of HS●, which has a low pKa value and will dissociate to give S●− under the reaction conditions (eq. 5.1) [280]. In the presence of oxygen, [Ru2] transfers an electron to molecular oxygen in the solution at a slower rate and generates superoxide and [Ru2]+ to complete the catalytic cycle (eq. 5.2). The formed S●− would react with HS− to give HSS●2− (eq. 5.3), which may lose an electron to oxygen and form HSS− (eq. 5.4). Finally, the formed superoxide anion dismutates to oxygen and H2O2 (eq. 5.5). The net reaction is a catalyzed oxidation of HS− to HS2− and H2O2 (eq. 5.6). Upon an extended reaction time (1 day), elemental sulfur was formed, as detected by electron ionization mass spectrometry of the reaction mixture, which shows peaks at m/z 64, 96, 128, 160, 192, and 256. It takes a little over 2 equiv of HS− to fully reduce [Ru2]+, as shown in Figure 5.1, indicating that HS− is consumed by other pathway (eq. 5.3) in addition to reacting with [Ru2]+. It was proposed that S●− could react with O2 to give a sulfur peroxyl anion radical, SOO●−, which rearranges to a sulfur dioxide anion radical, SO2●−. SO2●− could reduce the oxygen to give a superoxide anion and SO2 [281]. Under pH 7.4 conditions, we would expect the formation of SO32− because the pKa of HSO3− is 7.20 [282]. However, the UV-vis spectra of the reaction products bear little similarity to that of SO32− (Figure 5.7). In addition, if this mechanism were operating, we would expect one HS− to generate one H2O2, which is not supported by our experimental data. The partial reduction of [Ru2]+ in the subsequent addition of HS− shown in Figure 5.3 is likely due to the reaction of formed H2O2 with added HS− [283,284]. Consequently, more than 1 equiv of HS− has to be added in order to reach the same degree of reduction as that in the first addition. It is foreseeable 128 Chapter 5 that removal of the formed H2O2 from the system would lead to full response in the subsequent addition of HS−. This may be achieved by grafting [Ru2]+ on a solid matrix that can be washed with clean buffer before further usage.   1 30.0 NaHS + 5% [Ru2]+ (30min) Absorbance 0.8 NaHS NaHS2 2 Na2SO3 SO32 - 0.6 0.4 0.2 0 200 300 400 500 600 700 800 900 1000 Wavelength (nm) Figure 5.7. Absorption spectra of HS− with 5% [Ru2]+ (30 minutes), NaHS2 and Na2SO3 in 50 mM Tris-HCl buffer pH 7.40.   1.4 1.2 NaHS2 in DCM Seri es1 HS2 extract in DCM Seri es6 Absorbance 1 0.8 0.6 0.4 0.2 0 225 275 325 375 425 475 Wavelength (nm) Figure 5.8. Absorption spectra of NaHS2 and HS2− extract in DCM. 129 Chapter 5 5.3.5 Selectivity of [Ru2]+ towards anions and reductants For a good molecular probe, high selectivity is an essential feature and [Ru2]+ exhibits good selectivity for HS− (1 equiv) compared to other anions of biological relevance. These include NO2−, NO3−, Cl−, Br−, I−, SO42−, HPO42−, HCO3−, CH3COO−, and O2•− even though they are in large excess (100 equiv; Figure 5.9). Not all anions tested herein possess reducing activity, but ligand substitution (replacement of the Cl−) may occur to cause a shift of the IVCT band. This is not the case here because we did not observe much change of the NIR spectra caused by these anions under the assay conditions. Although the presence of large excessive amount of Cl− in the buffer may suppress ligand substitution reaction, our attempts failed to purposefully replace the chloride with other anionic ligand (e.g. F−, large excess, in CH3CN or THF) in the hope to alter the redox property of [Ru2]+, indicating that the Cl− is rather tightly bound to Ru. This is in sharp contrast to its much labile [Fe2TIED(CH3CN)4]4+ counterpart, which readily undergoes axial ligand exchange [251]. The reactivity of [Ru2]+ was examined with common biological reductants such as cysteine, ascorbic acid (vitamin C), uric acid, and vitamin E (in the form of Trolox, a water-soluble -tocopherol analogue). Eqimolar amounts of cysteine and ascorbic acid were able to reduce [Ru2]+ to about half its amount compared to that by HS− (Figure 5.9). The negative charge on HS− may help its interaction with the cationic [Ru2]+ and facilitate the reaction. Uric acid and Trolox failed to react. Nonetheless, for biological fluids containing a mixture of thiols, vitamin C, and HS−, quantification of HS− is only accurate after these 130 Chapter 5 interfering compounds are selectively removed either chemically or enzymatically.   Normalized A895 / A789 8 16 µM 7 6 5 4 3 2 1600 µM 1 Figure 5.9. Absorbance ratio changes of [Ru2]+ (16 µM) in a 50 mM Tris-HCl buffer (pH 7.40) recorded 1 minute after the addition of various anions (1600 µM) and reductants (16 µM). Cys, AA, UA, and TL denote cysteine, ascorbic acid, uric acid, and Trolox, the water-soluble vitamin E analogue, respectively. All the data are normalized with respect to the absorbance ratio of [Ru2]+ before the addition. 5.3.6 Method validation To first confirm the accuracy of our method, we validated our developed [Ru2]+ method with the DTNB colorimetric method for determining sulfhydryl groups. The standard curves of the test reagent (DTNB or [Ru2]+) with HS− are shown in Figure 5.10, where both curves showed a very good linearity of R2 > 0.9995. As shown in Table 5.1 and Table 5.2, the accuracy of the two assays lies within the usual accepted limits of 98–105%. The accuracy of our [Ru2]+ method is determined to be excellent at 98 ± 4% recovery of control sample of HS−. This result compares nicely with the established DTNB method [285]. 131 Chapter 5 0.7 1.4 A 0.6 A895 / A789 y = 0.0202x + 0.0023 R² = 0.9998 0.5 A412 B 1.2 0.4 0.3 1.0 0.8 0.6 0.2 0.4 0.1 0.2 0.0 y = 0.0729x + 0.1353 R² = 0.9996 0.0 0 4 8 12 16 20 [HS-] µM 24 28 32 0 2 4 6 8 10 [HS-] µM 12 14 16 Figure 5.10. (A) Standard curve of HS− with DTNB (247 µM). (B) Standard curve of HS− with [Ru2]+ (16 µM). Results are expressed as mean±SD (n = 3). Table 5.1. Accuracy of DTNB assay [HS-] (µM) 5.0 a Mean SD %RSDb Recoverya (µM) %R 5.3 107 5.1 103 5.3 106 5.2 105 ±0.1 ±2 ±2.0 ±2 [HS-] (µM) 10.0 Recoverya (µM) %R 10.3 103 10.8 108 10.8 108 10.6 106 ±0.1 ±3 ±2.0 ±3 [HS-] (µM) 15.0 Recoverya (µM) %R 15.8 105 15.3 102 14.8 99 15.3 102 ±0.5 ±3 ±3.1 ±3 Recovery or accuracy (%R) is expressed as a ratio: recovery (µM)/standard [HS−] x 100%. b Precision (%RSD) is the ‘relative percent difference’ between triplicate test results, %RSD = (SD/mean) x 100%. Table 5.2. Accuracy of [Ru2]+ assay [HS-] (µM) 5.0 a Mean SD %RSDb Recoverya (µM) %R 5.2 104 4.9 98 5.2 103 5.1 102 ±0.2 ±3 ±3.0 ±3 [HS-] (µM) 10.0 Recoverya (µM) %R 10.4 104 9.8 98 9.5 95 9.9 99 ±0.4 ±4 ±4.4 ±4 [HS-] (µM) 15.0 Recoverya (µM) %R 15.3 102 13.7 91 13.1 87 14.0 93 ±1.1 ±8 ±8.2 ±8 Recovery or accuracy (%R) is expressed as a ratio: recovery (µM)/standard [HS−] x 100%. b Precision (%RSD) is the ‘relative percent difference’ between triplicate test results, %RSD = (SD/mean) x 100%. 5.3.7 Measurement of H2S release from GYY4137 using [Ru2]+ A potential application of [Ru2]+ is in quantifying H2S by NIR absorbance. To illustrate such an application, we measured the H2S releasing rate from a 132 Chapter 5 H2S donor agent, morpholin-4-ium-4- methoxyphenyl(morpholino)phosphinodithioate, GYY4137 (eq. 5.7) using our validated [Ru2]+ assay [276,286]. GYY4137 hydrolyzes slowly in buffer to release H2S and has been investigated for its use as a therapeutic agent for cardiovascular diseases as a H2S donor. Using [Ru2]+ as the probe, the GYY4137 hydrolysis rate constant, k, is determined to be (3.33 ± 0.25) x 10-7 s-1 (Figure 5.11). We noticed that, at the time zero, the concentration of H2S was 2.40 µM. This is attributed to the residual H2S already present in GYY4137 crystals, which are highly hydroscopic and may be hydrolyzed during storage. Compared to the concentration of GYY4137, the degree of hydrolysis of 1.2% is low.    2.90 [HS-] (µM) 2.80 GYY4137 2.70 y = 0.0040x + 2.4053 R² = 0.9978 2.60 2.50 2.40 0 20 40 60 80 Time (min) 100 120 Figure 5.11. Release of H2S from GYY4137 (200 µM) in 50 mM Tris-HCl buffer (pH 7.40) incubated at 37 °C under N2 atmosphere as determined spectrophotometrically with the use of [Ru2]+ (16 µM) recorded after 1 minute. Results are expressed as mean ± SD (n = 3). 133 Chapter 5 5.4 Conclusion In conclusion, we have demonstrated that NIR active [Ru2]+ can catalyze the air oxidation of HS− through a single-electron-transfer reaction to generate hydrogen peroxide, disulfane, and elemental sulfur. The NIR absorbance ratio of the two resting states of the catalyst, [Ru2]+ and [Ru2], allows for the convenient determination of HS−. The reaction also provides a convenient method for the detection of HS− generation rate of a H2S donor of medical importance. 134 Chapter 6 Synthesis and Characterization of NIR Active Diiron Complexes and Their Reactivity with Redox-Active Molecules Chapter 6 6.1 Introduction Besides the [Ru2TIEDCl4]Cl complex (TIED = tetraiminoethylenedimacrocycle, denoted as [Ru2]+) reported by Spreer and co-workers, we are also interested in the isovalent [Fe2(TIED)(CH3CN)4](ClO4)4●2CH3CN (denoted as [Fe2]4+) because of its intense band at 874 nm in CH3CN with a high molar absorption coefficient of 24 600 M-1cm-1 [248]. One interesting aspect of the [Fe2]4+ complex is its NIR band, which is highly solvent-dependent because it shifts to 1050 nm in DMF and 1005 nm in water [247]. The large optical shifts are due to the exchange of the axial CH3CN ligands with coordinating solvent molecules. This makes the [Fe2]4+ complex interesting to study because the identity of the axial ligands can influence the optical, redox, and reaction properties of the bimetallic iron complex. In this study, the addition of π-acceptor ligands such as cyanide and 1-pentyl isocyanide to [Fe2]4+ were able to replace the axial acetonitrile ligands to form two new stable and water-soluble complexes Fe2TIED(CN)4 (1) and [Fe2TIED(RNC)4]4+ (2, where RNC = 1-pentyl isocyanide). Complexes 1 and 2 have been structurally characterized. Their reactivity towards reactive oxygen species (ROS) such as ClO−, H2O2, HO●, 1O2, O2●−, ONOO−, NO, NO− and reductant/π-acceptor ligand such as HS−, CN− were examined. 6.2 Materials and methods 6.2.1 Materials All experiments were performed with analytical grade reagents. All chemical solvents and compounds were used without further purification. Cyclam, iron(II) tetrafluoroborate hexahydrate, sodium hydrosulfide (NaHS), 135 Chapter 6 tris(hydroxymethyl)aminomethane, iron(II) sulfate heptahydrate, sodium nitrite (NaNO2), potassium superoxide (KO2), 1-pentyl isocyanide, rose bengal, and diethylamine NONOate diethylammonium salt (DEA/NO, NO donor) were purchased from Sigma-Aldrich, Inc (St Louis, MO, USA) and used without further purification. Hydrogen peroxide solution (30%), hydrochloric acid (37%), and potassium cyanide (KCN) were from Merck (Darmstadt, Germany). Sodium hypochlorite (NaOCl, 12–15%) was from Alfa Aesar (Heysham, Lancs, UK), sodium hydroxide was from Sino Chemicals (Singapore), and Angeli’s Salt (Na2N2O3, NO− donor) was from Cayman Chemical Company (Ann Arbor, MI, USA). Fe2TIED(CH3CN)4]4+ (denoted as [Fe2]4+) was synthesized according to a literature method [248]. 6.2.2 Instruments UV-vis absorption spectra were recorded with a Shimadzu UV-1601 UVvisible spectrophotometer at room temperature. ESI-MS spectra were obtained with a Finnigan/MAT LCQ ion trap mass spectrometer from m/z 50 to 1000. 1 H NMR spectra were recorded in CD3OD with a Bruker AC300 spectrometer operating at 300 MHz. Infrared spectra were recorded on a BIO-RAD FTS 165 FTIR spectrometer. 6.2.3 Preparation of stock solutions NaHS was dissolved in deionized H2O to give a 6 mM stock solution. The concentration of NaHS stock solution was determined by using its molar extinction coefficient of 7200 M-1cm-1 at 230 nm [277]. Tris buffer (50 mM) was prepared in deionized H2O to 1.0 L and a few drops of 1.5 M hydrochloric 136 Chapter 6 acid were added to yield a pH 7.40 Tris-HCl buffer (50 mM). DEA/NO was dissolved in deionized H2O to give a 4.847 mM stock solution. Angeli’s Salt was dissolved in 0.01 M NaOH to give a 25.2 mM stock solution. KCN and iron(II) sulfate heptahydrate were dissolved in deionized H2O to give a 6 mM stock solution. Stock solutions of NaOCl, and H2O2 were freshly prepared in deionized H2O with a stock solution concentration of 6 mM. The concentrations of NaOCl and H2O2 stock solutions were determined by measuring their UV absorbance immediately before use, at 292 nm in basic conditions (pH = 12.0, ε292 = 350 M-1cm-1 for NaOCl) [287], and at 240 nm (ε240 = 43 M-1cm-1 for H2O2) [278] respectively. Stock solution of KO2 was freshly prepared in DMSO with a stock solution concentration of 6 mM. ONOO− was prepared according to a literature method [ 288 ] and was neutralized with diluted HCl before use to give a 1.4 mM stock solution. The concentration of ONOO− stock solution was determined by using its molar extinction coefficient of 1670 M-1cm-1 at 302 nm [289]. HO● was generated in situ by adding one equiv iron(II) sulfate heptahydrate in the presence of one equiv of H2O2 in deionized H2O. Rose bengal was dissolved in deionized H2O to give a 1.0 mM stock solution. 6.2.4 Synthesis of Fe2TIED(CN)4 and [Fe2TIED(RNC)4]4+ complexes (i) Fe2TIED(CN)4 (1). [Fe2]4+ (16.54 mg, 0.0151 mmol) was dissolved in CH3CN (0.3 mL). Potassium cyanide (47.56 mg, 0.730 mmol) was dissolved in deionized H2O (2.85 mL), which was then added to [Fe2]4+. The reaction mixture was left to stir at room temperature for 1 hour. The green reaction mixture turned yellowish brown over the period of 1 hour. The reaction 137 Chapter 6 mixture (5 µL) was then taken out and diluted with 995 µL deionized H2O for UV-vis spectrophotometric analysis, which showed a shift of the NIR band to 932 nm, indicating that the complex was formed. The reaction mixture was evaporated to dryness to yield Fe2TIED(CN)4 as a yellowish brown complex (8.8 mg, 96.3%). 1 H 3 4 5 CN N N 2 6 H N Fe Fe N H CN N N N N CN CN H Fe2TIED(CN)4 (1) Fe2TIED(CN)4 (1). 1H NMR (300 MHz, CD3OD): δ 9.17 (m, 1H, H6), 3.33 (t, J = 3 Hz, 2H, H5), 2.74 (t, J = 3 Hz, 2H, H4), 2.65 (s, 2H, H3), 1.73 (m, 1H, H2). MS (ESI, +c): 640.9 [M+2H2O+H]+. IR (KBr): 3430, 2170, 2083, 2042, 1654, 1388, 1084, 885, 630, 533, 521 cm-1. (ii) [Fe2TIED(RNC)4]4+ (2, where RNC = 1-pentyl isocyanide). [Fe2]4+ (39.6 mg, 0.0362 mmol) was dissolved in CH3CN (0.3 mL). 1-Pentyl isocyanide (176 µL, 7.94 M) was dissolved in MeOH (5.524 mL), which was then added to [Fe2]4+. The reaction mixture was left to stir at room temperature for 1 day. The green reaction mixture turned blue-green over the period of 1 day. After 1 day, 6 µL of the reaction mixture was then taken out and diluted with 994 µL deionized H2O for UV-vis spectrophotometric analysis, which showed a shift of the NIR band to 803 nm, indicating that the complex was formed. The reaction mixture was then evaporated to dryness to yield [Fe2TIED(RNC)4]4+ as a blue-green complex (15.4 mg, 47.9%). 138 Chapter 6 [Fe2TIED(RNC)4]4+ (2, where RNC = 1-pentyl isocyanide). 1H NMR (300 MHz, CD3OD): δ 9.49 (m, 1H, H11), 3.84 (m, 2H, H10), 2.92 (m, 4H, H9, H8), 1.68 (s, 4H, H7, H5), 1.31 (m, 6H, H4, H3, H2), 0.92 (m, 3H, H1). MS (ESI, +c): 222.2 [M]4+. IR (KBr): 3433, 2958, 2208, 1638, 1446, 1055, 847, 805, 758, 533, 521 cm-1. 6.2.5 Procedures for sensing redox-active molecules In each separate experiment, 70 µL of the various solutions of 6.0 mM of ROS (ClO−, H2O2, HO●, and O2●−) or reductant/π-acceptor ligand (HS−, CN−) was added to complex 1 or 2 (21 µL, 1.0 mM) in 50 mM Tris-HCl buffer (pH 7.40) to give a final concentration of 420 µM test reagents and 21 µM complex 1 or 2 (V = 1 mL), while for ONOO−, 300 µL of 1.4 mM ONOO− was added instead. DEA/NO (20 µL, 4.2 mM) was added to complex 1 or 2 (21 µL, 1.0 mM) in 50 mM Tris-HCl buffer (pH 7.40) to give a final concentration of 84 µM NO and 21 µM complex 1 or 2 (V = 1 mL). Angeli’s Salt (20 µL, 8.4 mM) was added to complex 1 or 2 (21 µL, 1.0 mM) in 50 mM Tris-HCl buffer (pH 7.40) to give a final concentration of 168 µM Angeli’s 139 Chapter 6 Salt and 21 µM complex 1 or 2 (V = 1 mL). Rose bengal (7 µL, 1 mM) was added to complex 1 or 2 (21 µL, 1.0 mM) in 50 mM Tris-HCl buffer (pH 7.40) to give a final concentration of 7 µM rose bengal and 21 µM complex 1 or 2 (V = 1 mL). 1O2 was generated by illuminating the solution of 7 µM rose bengal and 21 µM complex 1 or 2 using Thorlabs OSL1-EC fibre illuminator at a high intensity for 5 and 10 minutes. The UV-vis absorption spectra of the reaction mixtures were recorded at a 5 minute interval for 30 minutes. The absorbance maxima of the complex 1 or 2 were obtained and normalized to that of complex 1 or 2 before any addition. All operations were performed under a nitrogen atmosphere. 6.3 Results and discussion 6.3.1 Synthesis of complexes 1 and 2 The synthesis of 1 was quite straightforward and was carried out by mixing [Fe2]4+ with excess (~ 48 equiv) potassium cyanide in deionized H2O. After continuous stirring for 1 hour at room temperature, a yellowish brown solution was obtained and was evaporated to dryness to yield 1 as a yellowish brown solid in good yield (96.3%). 1 is soluble in polar solvents like MeOH, EtOH, CH3CN, and H2O. In addition, unlike [Fe2]4+ which decomposes readily in H2O and 50 mM Tris-HCl buffer, 1 is very stable in H2O and 50 mM Tris-HCl buffer (pH 7.40). Its 1H NMR results confirm the low spin, dinuclear Fe(II) configuration. It was found that the addition of 84, 210, and 420 µM KCN to 21 µM [Fe2]4+ in H2O after mixing for 5 minutes showed the red shift of the NIR band from 884 nm to 932 nm with higher intensity, and the red shift of the visible band from 345 nm to 388 nm with higher intensity as shown in 140 Chapter 6 Figure 6.1. From the red shift in the absorption bands as compared to the absorption bands of [Fe2]4+ at 240, 340 and 874 nm in acetonitrile reported by Spreer et al. [247], it is believed that the axial acetonitrile ligands are gradually being replaced with the cyanide ligands when higher amounts of KCN were added to [Fe2]4+. This is because the ligand field strength of CN− is relatively stronger than CH3CN in the spectrochemical series, so the strongly π-acidic CN− ligand is able to displace the weaker CH3CN axial ligands. Hence in the synthesis of 1, excess (~ 48 equiv) potassium cyanide was used to ensure that all the axial acetonitrile ligands were fully replaced with cyanide ligands with a shorter reaction time of 1 hour. 0.7 84 µM KCN Absorbance 0.6 210 µM KCN 420 µM KCN 0.5 0.4 0.3 0.2 0.1 0 200 300 400 500 600 700 800 900 1000 1100 Wavelength (nm) Figure 6.1. Absorption spectra of [Fe2]4+ (21 µM) in H2O recorded 5 minutes after reactions with various concentrations of with KCN (84, 210, and 420 µM). The synthesis of 2 was done by mixing [Fe2]4+ with excess (~ 64 equiv) 1pentyl isocyanide in MeOH. After continuous stirring for 24 hours at room temperature, a blue-green solution was obtained and was evaporated to dryness to yield 2 as a blue-green solid. Complex 2 is slightly soluble in DCM, 141 Chapter 6 but has a good solubility in polar solvents like MeOH and H2O. In addition, similar to 1, complex 2 is very stable in H2O and 50 mM Tris-HCl buffer (pH 7.40). In the synthesis of 2, excess (~ 64 equiv) 1-pentyl isocyanide was used to ensure that all the axial acetonitrile ligands were fully replaced with 1pentyl isocyanide ligands. However, unlike 1, a longer reaction time of 24 hours is needed for the synthesis of 2. It was observed that there was no colour change in the reaction mixture after 1 hour, so a small amount of reaction mixture was withdrawn after 1 hour and scanned by UV-vis spectrophotometry to determine if a reaction has occurred to result in a shift of the NIR band. Surprisingly, there was only a slight blue shift of the NIR band from 874 nm to 866 nm and the visible band remained unshifted at 340 nm. Thus we decided to leave the reaction mixture to stir overnight for 24 hours as we thought a longer reaction time might be able to displace the axial acetonitrile ligands. Indeed, after stirring overnight, the reaction mixture turned blue-green and UV-vis spectrum of the reaction mixture showed a blue shift of the NIR band to 775 nm. 6.3.2 Spectroscopic characterization of complexes 1 and 2 UV-vis spectroscopy. Complex 1 in deionized H2O exhibits an intense band in the NIR region at 932 nm (ε = 26 695 M-1cm-1) and another intense band at 388 nm (ε = 21 017 M-1cm-1) in Figure 6.2. Since Spreer et al. has assigned the NIR band at 874 nm for [Fe2]4+ to be due to a MLCT transition [247], it seems reasonable to assign the intense NIR band at 932 nm for 1 as the expected MLCT transition from the filled d orbitals of Fe(II) to the empty low-lying π* orbital of the conjugated tetraiminoethylene dimacrocyclic ligand 142 Chapter 6 and the intense band at 388 nm can be assigned as the LMCT transition from the non-bonding electrons of the anionic cyanide ligands to the lowest unoccupied d-orbital of Fe(II). Since there is a red shift in the NIR band from 874 nm in CH3CN for [Fe2]4+ to 932 nm in H2O for complex 1, and that this NIR band is very stable and does not decay with time, we know that the tetraiminoethylene dimacrocyclic ligand is intact and propose that the axial CH3CN ligands on [Fe2]4+ may have been replaced with CN− ligands to form 1. There is also a red shift observed for the visible band from 340 nm for [Fe2]4+ to 388 nm for 1. This is because the anionic cyanide ligands being more electron rich can give away electrons more easily to the metal center, giving rise to a lower LMCT energy and hence red shift in the visible absorption band maxima. In addition, the NIR band of 1 is further red shifted to 966 nm in MeOH and EtOH, compared to 932 nm in H2O (Figure 6.2). The visible band of 1 is only slightly red shifted to 396 nm in MeOH and EtOH, compared to 388 nm in H2O as shown in Figure 6.2. It can be observed that the NIR absorption band for 1 is solvent dependent, while that of the visible absorption band is almost solvent independent. The MLCT absorption band’s shift to longer wavelength in MeOH and EtOH may be because of the higher Gutmann donor number of MeOH (19.0 kcal/mol) and EtOH (31.5 kcal/mol) when compared to H2O (18.0 kcal/mol) [290]. The donor number is defined as the molar reaction enthalpy for the formation of a 1:1 adduct between SbCl5 and the added solvent in 1,2-dichloroethane. It provides a measure of the ability of individual solvent molecules to donate an electron pair to SbCl5 [291]. The higher the donor number, the stronger is the Lewis basicity of the solvent. As the Gutmann donor number increases, the electron pair donation from the 143 Chapter 6 solvent to the metal increases the energy level of the d orbitals of Fe(II), resulting in a decrease in the MLCT band energy and thus a longer wavelength absorption [292]. 0.7 [Fe2]4+ in CH 3CN [Fe2TIED(CN)4] in MeOH, EtOH [Fe2TIED(CN)4] in H2O Absorbance 0.6 0.5 0.4 0.3 0.2 0.1 0 200 300 400 500 600 700 800 900 1000 1100 Wavelength (nm) Figure 6.2. Absorption spectra of [Fe2]4+ (21 µM) in CH3CN (blue line) and [Fe2TIED(CN)4] (21 µM) in MeOH, EtOH (pink line), H2O (green line). The UV-vis absorption spectrum of 2 in deionized H2O shows an intense band in the NIR region at 803 nm (ε = 24 625 M-1cm-1), a band in the visible region at 340 nm (ε = 25 170 M-1cm-1), and another band in the UV region at 247 nm (ε = 40 272 M-1cm-1) as shown in Figure 6.3. The intense NIR band at 803 nm can be assigned as the expected MLCT transition from the filled dπ orbitals of Fe(II) to the empty low-lying π* orbital of the tetraiminoethylene dimacrocyclic ligand, by analogy to spectral assignment for similar NIR absorption in the analogous [Fe2]4+ [247]. The intense band in the visible region at 340 nm is assigned as the LMCT transition from the non-bonding electrons of the neutral 1-pentyl isocyanide ligands to the lowest unoccupied d orbital of Fe(II). The strong absorption band in the UV region at 247 nm can 144 Chapter 6 be assigned as intra-ligand (IL) π-π* transitions of the tetraiminoethylene dimacrocyclic ligand. Since the NIR band at 803 nm is very stable, which is similar to the case of complex 1, we propose that the axial CH3CN ligands on [Fe2]4+ have been replaced with 1-pentyl isocyanide ligands to form 2. However, unlike complex 1 in which the NIR band is red-shifted to 932 nm in H2O, for complex 2, the NIR band is blue shifted to 803 nm in H2O, when compared to 874 nm in CH3CN for [Fe2]4+. On the other hand, no shift is observed for the visible band of 2 at 340 nm when compared to [Fe2]4+. This is because both the acetonitrile and 1-pentyl isocyanide ligands are neutral and not as electron rich as the anionic cyanide ligands, hence the LMCT energy remained unchanged. In addition, the NIR band of 2 is further blue shifted to 775 nm in MeOH, compared to 803 nm in H2O (Figure 6.3). This observation for complex 2 is totally opposite to the case of complex 1 whereby the NIR band is red-shifted in MeOH instead of blue-shifted. The visible band of 2 is only slightly blue shifted to 335 nm in MeOH, compared to 340 nm in H2O as shown in Figure 6.3. Similar to complex 1, the NIR absorption band for 2 is found to be solvent dependent, while that of the visible absorption band is almost solvent independent. The MLCT absorption band of 2 is shifted to shorter wavelength in MeOH. This is due to the increased solvation of the lone pairs of electrons in non-bonding orbitals of MeOH, which lowers the energy of the metal orbitals, resulting in a larger energy gap and hence shorter wavelength is observed. 145 Chapter 6 0.9 [Fe 2]4+ in CH3CN [Fe 2TIED(C 5H11NC)4]4+ in MeOH [Fe 2TIED(C 5H11NC)4]4+ in H 2O 0.8 Absorbance 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 200 300 400 500 600 700 800 900 1000 1100 Wavelength (nm) Figure 6.3. Absorption spectra of [Fe2]4+ (21 µM) in CH3CN (blue line) and [Fe2TIED(C5H11NC)4]4+ (21 µM) in MeOH (pink line), and H2O (green line). IR vibrational spectroscopy. Complex 1 has a D2h symmetry and should show one IR-detectable vibrational band for the CN stretching. Indeed, in the IR spectrum of 1, a strong absorption peak at 2170 cm-1 was observed, together with two weak absorption peaks at 2083 and 2042 cm-1 (Figure 6.4). These three absorption peaks correspond to the characteristic CN groups, which provides evidence of the coordination of the CN− ligands to the Fe centers, since a free CN− ligand’s stretching frequency at 2100 cm-1 is not observed [293]. It is well-known that CN− acts both as an σ-donor to the metal and as a π-acceptor from the metal. Studies have shown that σ-donation generally increases the value of CN, while π-back bonding decreases the value of CN [294]. The lower stretching frequency observed for π-back bonding is because of a weakened CN bond due to the donation of electron from the metal orbital to the π* orbital of CN−. Since complex 1 is a neutral complex, the metal is more electron rich to engage in metal to ligand π-back 146 Chapter 6 bonding, giving rise to two additional weaker absorption peaks at 2083 and 2042 cm-1. In the case of complex 1, CN− plays a dual role as an σ-donor observed at 2170 cm-1 and as a π-acceptor at 2083 and 2042 cm-1. In addition, the IR spectrum of 1 also shows NH at 3430 cm-1, C=N at 1654 cm-1, and C=C at 1388 cm-1, which corresponds well to that obtained for the [Fe2]4+ complex [295], indicating that the tetraiminoethylene dimacrocyclic ligand is present.   32.0 Complex 1 30 28 26 885.15 24 22 630.08 20 533.58 521.64 18 %T 16 14 1084.04 12 2042.51 10 2083.39 8 6 4 1654.10 3430.84 2170.20 1388.40 2 0.0 4000.0 3000 2000 -1 cm cm-1 1500 1000 450.0 Figure 6.4. IR spectrum of 1 recorded as a KBr disc. Isocyanides generally coordinate to transition metals with almost linear MCN-R bonds. Because of this, complex 2 has a D2h symmetry so it has one IR-active CN stretching mode. Indeed, the IR spectrum of 2 shows a strong absorption peak at 2208 cm-1, which corresponds to the characteristic RNC groups (Figure 6.5). This shows that the isocyanide ligands are coordinated to the Fe centers since uncoordinated isocyanide ligand generally has a stretching frequency of ~ 2130 cm-1 [296]. The observed increase in stretching frequency upon coordination to metal indicates that the 1-pentyl isocyanide ligand acts a strong σ-donor [297]. The IR spectrum of 2 also displays NH at 3433 cm-1, 147 Chapter 6 C=N at 1638 cm-1, and C=C at 1446 cm-1, which corresponds well to that obtained for the [Fe2]4+ complex [295]. Besides, CN of coordinated acetonitrile at 2285 cm-1 is not observed, indicating that the acetonitrile ligands have been replaced with the 1-pentyl isocyanide ligands.   30.0 Complex 2 28 26 24 22 20 18 847.95 758.54 16 533.87 521.95 805.90 %T 14 12 10 1446.03 8 1638.34 6 4 2208.37 2 0.0 4000.0 3433.24 2958.82 3000 1055.07 2000 -1 cm cm-1 1500 1000 450.0 Figure 6.5. IR spectrum of 2 recorded as a KBr disc. ESI-MS spectrometry. To confirm the number of CN− ligands coordinated to the Fe center, ESI-MS was conducted on the isolated complex 1. The ESIMS spectrum of 1 in deionized H2O recorded in the cationic ionization mode shows a molecular ion peak at m/z 640.9, corresponding to [1+2H2O+H]+ (Figure 6.6). This result demonstrated that all the four axial CH3CN ligands in [Fe2]4+ have been replaced by CN− ligands to form complex 1, which is a neutral complex. As for complex 2, the ESI-MS spectrum of 2 diluted in MeOH in the cationic ionization mode shows a molecular ion peak at m/z 222.2, corresponding to [2]4+ (Figure 6.7). This result confirmed that all the four axial acetonitrile ligands in [Fe2]4+ have been replaced by 1-pentyl isocyanide 148 Chapter 6 ligands to form complex 2. Since 1-pentyl isocyanide ligand is a neutral ligand, complex 2 carries a charge of +4. 100511Fe2TIEDCN4inH2O-1mM #60-135 RT: 1.16-2.42 AV: 76 SB: 17 0.83-1.11 NL: 2.04E7   T: + c ESI Full ms [50.00-1000.00] 102.4 100 95 90 85 80 75 70 Relative Abundance 65 60 55 152.7 50 45 40 35 116.8 30 25 266.6 640.9 20 212.7 88.5 15 164.8 208.6 10 5 380.5 638.8 87.5 816.1 642.9 278.6 236.7 494.4 316.9 364.5 392.6 404.5 464.4 592.5 509.8 608.5 692.3 722.3 800.2 817.1 854.0 914.3 0 100 200 300 400 500 600 700 800 900 m/z Figure 6.6. ESI-MS spectrum of 1 in deionized H2O recorded in the cationic mode. 050411Fe2TIEDPeNC4inMeOH #68-73 RT: 1.27-1.35 AV: 6 SB: 6 1.14-1.24 NL: 1.90E8   T: + c ESI Full ms [100.00-1000.00] 576.0 100 95 90 85 571.8 80 75 70 Relative Abundance 65 60 55 50 45 40 659.6 35 578.0 286.4 30 488.8 25 600.1 20 15 612.6 353.6 201.2 222.2 119.6 658.7 10 377.7 5 0 100 291.0 309.1 165.6 200 300 383.0 400 660.7 569.9 489.9 559.8 474.3 547.1 500 627.6 600 668.3 670.2 700 699.5 769.4 834.1 800 850.1 921.0 900 m/z Figure 6.7. ESI-MS spectrum of 2 in MeOH recorded in the cationic mode. NMR spectroscopy. The proposed structure of 1 has a plane of symmetry cutting through the middle of C=C bond in the TIED ligand, which is parallel to the Fe-CN bond. It has another plane of symmetry through the C=C bond, two Fe atoms, the four CN ligands and two of the opposite methylene carbon 149 Chapter 6 atoms. Because of the two symmetry planes, complex 1 has six types of chemically non-equivalent protons labeled as shown in Figure 6.8. 1 H 3 4 5 CN N N 2 N H N Fe Fe N N N H CN 6 N CN CN H Fe2TIED(CN)4 (1) water 3 5 4 2 6 Figure 6.8. 1H NMR spectrum of 1 in CD3OD. The numbering scheme is depicted in the figure. The 1H NMR spectrum of 1 in CD3OD confirms its D2h symmetry by showing six signals due to the six types of chemically non-equivalent protons (Figure 6.8) and a broad water peak at 4.87 ppm. The peak at 9.17 ppm is assigned to the proton of the azomethine carbons at position 6 and is a multiplet due to coupling with the two protons at position 5, which has a triplet at 3.33 ppm. Since the proton of the azomethine carbons at position 6 is on the same plane as the conjugated C=C double bond, it is highly deshielded by the anisotropic effect of the conjugated double bond and shifts to low field as 150 Chapter 6 observed by the high chemical shift. The peak at 2.74 ppm is assigned to the two protons at position 4 and is a triplet due to coupling with the two protons at position 5 (J = 3 Hz). The singlet at 2.65 ppm is assigned to the two protons at position 3, while the multiplet at 1.73 ppm is assigned to the two protons at position 2. No peak is assigned to the amine proton (N-H) at position 1 since it rapidly undergoes deuterium exchange with the deuterated NMR solvent to form N-D. The coordination of the 1-pentyl isocyanide ligands to the Fe centers also forms a low spin, diamagnetic, dinuclear Fe(II) complex since 1-pentyl isocyanide is a strong π-acceptor ligand. The structure of 2 is confirmed by its 1 H NMR spectrum in CD3OD. Similar to complex 1, complex 2 has two planes of symmetry, one cutting through the middle of C=C bond in the TIED ligand, parallel to the Fe-CN bond, while the other cuts through the C=C bond, two Fe atoms, the four RNC ligands and two of the opposite methylene carbon atoms. Because of these two symmetry planes, complex 2 has eleven types of chemically non-equivalent protons labeled as shown in Figure 6.9. 151 Chapter 6 water 1 4,3,2 solvent 9,8 10 7,5 11 Figure 6.9. 1H NMR spectrum of 2 in CD3OD. The numbering scheme is depicted in the figure. The 1H NMR spectrum of 2 in CD3OD shows eleven signals due to the eleven types of chemically non-equivalent protons (Figure 6.9) and a solvent peak at 3.31 ppm and a water peak at 4.81 ppm. The deshielded peak at 9.49 ppm is assigned to the proton of the azomethine carbons at position 11 and is a multiplet due to coupling with the two protons at position 10, which has a multiplet at 3.84 ppm. Another multiplet at 2.92 ppm is assigned to the protons at position 9 and 8. The closeness of the chemical shifts of the four protons at position 8 and 9 makes them appear as overlapping clusters. The slightly split singlet at 1.68 ppm is assigned to the two protons at position 7 and 5. The multiplet at 1.31 ppm is assigned to the pentyl isocyanide protons at position 4, 3 and 2, while the multiplet at 0.92 ppm is assigned to the three terminal pentyl isocyanide protons at position 1. No peak is assigned to the amine proton (N-H) at position 6 since it rapidly undergoes deuterium exchange with the deuterated NMR solvent to form N-D. 152 Chapter 6 Hence based on our UV-vis, IR, ESI-MS, and 1H NMR results, the structures of complexes 1 and 2 are confidently assigned. In addition to their solubility and stability in H2O, both complexes 1 and 2 are also very soluble and stable in 50 mM Tris-HCl buffer (pH 7.40) to give an NIR band at 932 nm and 803 nm respectively, which makes them ideal for sensing ROS under physiological conditions. 6.3.3 Comparison of NIR band shift for [Fe2]4+, 1, and 2 In comparison to the NIR band of [Fe2]4+ at 874 nm, it is interesting to note that the NIR band for complex 1 in deionized H2O is red-shifted to 932 nm by 58 nm, while that of the NIR band for complex 2 in deionized H2O is blueshifted to 803 nm by 71 nm as shown in Figure 6.10. This huge shift in the NIR absorption band maximum indicates that the axial acetonitrile ligands have replaced with the cyanide and 1-pentyl isocyanide ligands to form a neutral complex 1 and a cationic complex 2 respectively. The red shift observed for complex 1 is due to its neutral charge so the metal center is more electron rich and it will be easier to remove an electron from the metal center. This in turn results in a smaller MLCT energy gap, which causes the MLCT absorption band of 1 to shift to longer wavelength (red shift). It is observed that the NIR absorption band maxima of the two complexes differ by 129 nm, which is considered to be a huge difference even though the axial ligands in 1 and 2 are both strong π-acidic ligands. We believe it is due to the high positive charge (+4) of complex 2 which resulted in the large blue shift of the NIR band. The high positive charge on the complex makes it more difficult for the metal center to give away the electron and this lowers the energy level of the d 153 Chapter 6 orbitals of Fe(II), resulting in the MLCT band to shift to higher energy. Hence a shorter wavelength (blue shift) is observed for complex 2. 0.7 [Fe 2]4+ in CH3CN [Fe 2TIED(CN)4] in H 2O [Fe 2TIED(C 5H11NC)4]4+ in H 2O Absorbance 0.6 0.5 0.4 0.3 0.2 0.1 0 200 300 400 500 600 700 800 900 1000 1100 Wavelength (nm) Figure 6.10. Absorption spectra of [Fe2]4+ (21 µM) in CH3CN (blue line), [Fe2TIED(CN)4] (21 µM) in H2O (pink line), and [Fe2TIED(C5H11)4]4+ (21 µM) in H2O (green line). 6.3.4 Reactivity of complexes 1 and 2 with redox-active molecules The reactivity of complex 1 and 2 with the common biologically relevant ROS (ClO−, H2O2, HO●, O2●−, 1O2, ONOO−, NO, and NO−) or reductant/πacceptor ligand (HS−, CN−) in a 50 mM Tris-HCl buffer (pH 7.40) was first examined to see if there is any good selectivity for sensing application. It is found that complex 1 is very stable and shows no reactivity with all the ROS and HS− tested even though they are added in large excess (20 equiv). Hence complex 1 cannot be utilized as a molecular probe for sensing application. On the other hand, complex 2 is comparatively more reactive than complex 1 and shows good reactivity with HS− (20 equiv) and NO− (4 equiv) in Figure 6.11, while the rest of the ROS, and CN− do not show much reactivity though they are in large excess (20 equiv). The non-reactivity of CN− towards complex 2 154 Chapter 6 indicates that the four 1-pentyl isocyanide ligands on complex 2 are quite tightly bound to the Fe centers and cannot be replaced with other strong πacceptor ligands like CN−. 1.8 420 µM Normalized A803 1.6 84 µM 1.4 1.2 1.0 Figure 6.11. Absorbance changes of 2 (21 µM) in a 50 mM Tris-HCl buffer (pH 7.40) recorded 30 minutes after the addition of various ROS, CN−, HS− (420 µM), and NO−, NO (84 µM). All the data are normalized with respect to the absorbance of 2 before the addition. The spectroscopic changes of complex 2 upon addition of HS− (20 equiv) in 50 mM Tris-HCl buffer (pH 7.40) with time were shown in Figure 6.12. It can be seen that the NIR absorption band of complex 2 at 803 nm is quenched by about 1.64 times at the end of 30 minutes. However, when compared to [Ru2]+, the sensitivity of complex 2 towards HS− is not good since the addition of a higher concentration of HS− (420 µM) is still not able to fully quench the NIR absorption band of complex 2. Hence the linear calibration curve for the detection of HS− using complex 2 is not obtained. 155 Chapter 6 0.6 0 min 1 min 5 min 10 min 15 min 20 min 25 min 30 min Absorbance 0.5 0.4 0.3 0.2 0.1 0 200 300 400 500 600 700 800 900 1000 1100 Wavelength (nm) Figure 6.12. Absorption spectra of 2 (21 µM) with HS− (420 µM) in a 50 mM Tris-HCl buffer (pH 7.40) with time. Next, the reactivity of complex 2 with Angeli’s Salt (8 equiv), which is a strong oxidant, in a 50 mM Tris-HCl buffer (pH 7.40) was investigated as shown in Figure 6.13. At physiological pH of 7.40, Angeli’s Salt exists mainly in the form of monoanion HN2O3− (pKa1 = 2.5 [298]), which decomposes to HNO and NO2− (eq. 6.1) [299]. As HNO is a weak acid (pKa = 4.7), it will exist as NO− under physiological pH conditions (eq. 6.2) [299]. HN2O3−  HNO + NO2− HNO  NO− + H+ (6.1) (6.2) It can be observed that addition of Angeli’s Salt (168 µM) is able to quench the NIR absorption band of complex 2 by about 1.37 times after 55 minutes, which also demonstrated that complex 2 did not exhibit good sensitivity for the detection of Angeli’s Salt. Hence complex 2 cannot be usefully utilized as a good molecular probe for sensing application since its sensitivity towards ROS is not high enough. 156 Chapter 6 0.5 0 min 25 min 30 min 35 min 40 min 45 min 50 min 55 min Absorbance 0.4 0.3 0.2 0.1 0 200 300 400 500 600 700 800 900 1000 1100 Wavelength (nm) Figure 6.13. Absorption spectra of 2 (21 µM) with Angeli’s Salt (168 µM) in a 50 mM Tris-HCl buffer (pH 7.40) under a N2 atmosphere with time. 6.4 Conclusion In conclusion, we have demonstrated that the addition of π-acceptor ligands such as cyanide and isocyanide ligands to [Fe2]4+ was able to replace the axial acetonitrile ligands to form two new stable complexes, neutral [Fe2TIED(CN)4] (1) and cationic [Fe2TIED(C5H11NC)4]4+ (2), which are soluble in water. The structures of complexes 1 and 2 have been confirmed by UV-vis, IR, ESI, and NMR spectroscopy. Both complexes contain a strong π-acidic ligand, but the NIR absorption band is different by 129 nm. We believe it is due to the positive charge on 2. Complex 1 was very stable and did not show any reactivity with ROS and HS-, while complex 2 exhibited good selectivity for HS− and Angeli’s Salt. However, because of its poor sensitivity, complex 2 was not suitable for use as a molecular probe for sensing HS− and Angeli’s Salt. The combination of the NIR spectra of the two complexes covers a very 157 Chapter 6 board range from 803 nm to 932 nm. This may be of use as NIR materials for future applications. 6.5 Future work One potential application that we thought of is utilizing complexes 1 and 2 as ligand donors through NIR light activation to release the axial ligands, which could be beneficial for biomedical applications. When [Fe2TIEDL4]n+, where L = CN−, n = 0 for 1, and L = C5H11NC, n = 4 for 2, is excited by NIR light, the metal's π-donation capacity will be weakened and as a result, the πacceptor ligand is more likely to dissociate from the Fe centers (eq. 6.3). However, if the ligand dissociation is reversible, it may be very hard to detect changes unless a reagent (M) is used to trap the dissociated ligand (eq. 6.4). NIR light [Fe2TIEDL4]n+  ([Fe2TIEDL4]n+)*  [Fe2TIEDL3]n+ + L L + M  LM (6.3) (6.4) For example, CN− may be trapped by transition metal such as Fe and the reaction can then be followed by monitoring the NIR spectra for a long period of time. The rate of the ligand release with the on-off response to NIR light would be different if NIR light triggers the release of the ligand. The πacceptor ligand to be released will be CN− and RNC. This only works when there is no ligand dissociation in the dark. Any NIR induced releasing of CN− and RNC would have interesting biological and medical applications. 158 List of Publications and Patent 1. Yi Ling Quek, Choon Hong Tan, Jinsong Bian, and Dejian Huang. Air oxidation of HS– catalyzed by an mixed-valence diruthenium complex, an Near-IR Probe for HS– Detection. Inorganic Chemistry. 2011, 50, 7379– 7381. 2. Yi Ling Quek, and Dejian Huang. Hydroethidine as a probe for measuring superoxide formation rates during aerial oxidation of myricetin and quercetin. Tetrahedron Letters. 2011, 52, 5384–5387. 3. Caili Fu, Wei Chen, Yi Ling Quek, Runyan Ni, Amylia Bte Abdul Ghani, Wendy Wen Yi Leong, Huaquang Zeng, and Dejian Huang. Sustainability from agricultural waste: chiral ligands from oligomeric proanthocyanidins via acid-mediated depolymerization. Tetrahedron Letters. 2010, 51, 6322– 6324. 4. Yi Ling Quek, Ying Ying Tan, Wenie Chin, and Dejian Huang. Radicalforming pro-oxidant activities of different tea leaves quantified by a fluorescent probe hydroethidine. Manuscript in preparation. 5. Yi Ling Quek, and Dejian Huang. Synthesis and characterization of NIR active bis-macrocyclic diiron complexes. Manuscript in preparation. 6. Dejian Huang, Wei Chen, Runyan Ni, Caili Fu, Yi Ling Quek. Method of preparing derivatives/oligomers of epicatechin and applications thereof. PCT Int. Appl. (2011), 47pp, WO 2011073950. 159 References 1. Halliwell B, Gutteridge JMC. Free Radicals in Biology and Medicine, Clarendon Press: Oxford, 1989. 2. Halliwell B. Free Radical Res. 1996, 25, 57–74. 3. Droge W. Physiol. Rev. 2002, 82, 47–95. 4. Betteridge DJ. Metabolism 2000, 49, Suppl. 1, 3–8. 5. Gutteridge JMC, Halliwell B. Antioxidants in Nutrition, Health, and Disease, Oxford University Press: New York, 1994. 6. Sies H. Oxidative Stress, Academic Press: London, 1985. 7. Taylor A. Ann. N.Y. Acad. Sci. 1992, 669, 111–123. 8. Jenner P. Lancet. 1994, 344, 796–798. 9. Witztum JL. Lancet. 1994, 344, 783–795. 10. Cerutti PA. Lancet. 1994, 344, 862–863. 11. Wiseman H, Halliwell B. Biochem. J. 1996, 313, 17–29. 12. McCord JM, Fridovich I. J. Biol. Chem. 1969, 244, 6049–6055. 13. Deisseroth A, Dounce AL. Physiol. Rev. 1970, 50, 319–375. 14. Mills GC. J. Biol. Chem. 1957, 229, 189–197. 15. Halliwell B. Annu. Rev. Nutr. 1996, 16, 33–50. 16. Gaziano JM. Nutrition 1996, 12, 583–588. 17. Willett WC. Science 1994, 264, 532–537. 18. Byers T, Perry G. Annu. Rev. Nutr. 1992, 12, 139–159. 19. Hertog G, Hollman P, Katan M. J. Agric. Food Chem. 1992, 40, 2379– 2383. 20. Rice-Evans CA, Miller NJ, Paganga G. Free Radical Biol. Med. 1996, 20, 933–956. 160 21. Duarte J, Perez-Vizcaino F, Zarzuelo A, Jimenez J, Tamargo J. Eur J Pharmacol. 1993, 239, 1–7. 22. Cargenas M, Marder M, Blank VC, Roguin LP. Bioorg. Med. Chem. 2006, 14, 2966–2971. 23. Middleton E, Kandaswami C. Biochem. Pharmacol. 1992, 43, 1167– 1179. 24. Brown JP. Mutat. Res. 1980, 75, 243–277. 25. Crozier A, Jaganath IB, Clifford MN. Phenols, polyphenols and tannins: an overview. In Plant Secondary Metabolites (Crozier A, Clifford MN, Ashihara H, eds.) 2006, pp. 1–24, Blackwell Publishing Ltd, Oxford. 26. Pierpoint WS. Prog. Clin. Biol. Res. 1986, 213, 125–140. 27. Miean KH, Mohamed S. J. Agric. Food Chem. 2001, 49, 3106–3112. 28. Kuhnai J. Wld. Rev. Nutr. Diet. 1976, 24, 117–191. 29. Del Rio D, Stewart AJ, Mullen W, Burns J, Lean MEJ, Brighenti F, Crozier A. J. Agric. Food Chem. 2004, 52, 2807–2815. 30. Balentine DA, Wiseman AS, Bouwens LCM. Crit. Rev. Food Sci. Nutr. 1997, 37, 693–704. 31. Rice-Evans CA, Miller NJ, Paganga G. Trends Plant Sci. 1997, 2, 152– 159. 32. Shahidi F, Janitha PK, Wanasundara PD. Crit. Rev. Food Sci. Nutr. 1992, 32, 67–103. 33. Husain SR, Cillard J, Cillard P. Phytochemistry 1987, 26, 2489–2491. 34. Nakagawa T, Yokozawa T. Food Chem. Toxicol. 2002, 40, 1745–1750. 35. Torel J, Cillard J, Cillard P. Phytochemistry 1986, 25, 383–385. 36. Cao G, Sofic E, Prior RL. Free Radical Biol. Med. 1997, 22, 749–760. 161 37. Haenen GRMM, Paquay JBG, Korthouwer REM, Bast A. Biochem. Biophys. Res. Commun. 1997, 236, 591–593. 38. Heijnen CGM, Haenen GRMM, van Acker FAA, van der Vijgh WJF, Bast A. Toxicol. In Vitro 2001, 15, 3–6. 39. Hu JP, Calomme M, Lasure A, DeBruyne T, Pieters A, Vlietinck A, Vanden Berghe DA. Biol. Trace. Elem. Res. 1995, 47, 327–333. 40. van Acker SABE, de Groot MJ, van den Berg DJ, Tromp MNJL, den Kelder GDO, van der Vijgh WJF, Bast A. Chem. Res. Toxicol. 1996, 9, 1305–1312. 41. Burda S, Oleszek W. J. Agric. Food Chem. 2001, 49, 2774–2779. 42. Yagi K, Ed. Active Oxygens, Lipid Peroxides, and Antioxidants, CRC Press: Boca Raton, FL, 1993. 43. Mira L, Fernandez MT, Santos M, Rocha R, Florencio MH, Jennings KR. Free Radical Res. 2002, 36, 1199–1208. 44. Hajji HE, Nkhili E, Tomao V, Dangles O. Free Radical Res. 2006, 40, 303–320. 45. Halliwell B, Gutteridge JMC. Methods in Enzymol. 1990, 186, 1–85. 46. Kostyuk VA, Potapovich AI, Vladykovskaya EN, Korkina LG, Afanas’ev IB. Arch. Biochem. Biophys. 2001, 385, 129–137. 47. Moridani MY, Pourahmad J, Bui H, Siraki A, O’Brien PJ. Free Radical Biol. Med. 2003, 34, 243–253. 48. Morel I, Lescoat G, Cogrel P, Sergent O, Pasdeloup N, Brissot P, Cillard P, Cillard J. Biochem. Pharmacol. 1993, 45, 13–19. 49. Morel I, Lescoat G, Cillard P, Cillard J. Methods in Enzymol. 1994, 234, 437–443. 162 50. Afanas’ev BI, Dorozhko IA, Brodskii VA, Kostyuk AV, Popapovitch IA. Biochem. Pharmacol. 1989, 38, 1763–1769. 51. Thompson M, Williams CR, Elliot GEP. Anal. Chim. Acta 1976, 85, 375–381. 52. Steffen Y, Schewe T, Sies H. Free Radical Res. 2006, 40, 1076–1085. 53. Chang WS, Lee YJ, Lu FJ, Chiang HC. Anticancer Res. 1993, 13, 2165– 2170. 54. Lindahl M, Tagesson C. Inflammation 1993, 17, 573–582. 55. Hoult JRS, Moroney MA, Paya M. Methods in Enzymol. 1994, 234, 443– 455. 56. Siess MH, Leclerc J, Canivenc-Lavier MC, Rat P, Suschetet M. Toxicol. Appl. Pharmacol. 1995, 130, 73–78. 57. Middleton E, Kandaswami C, Theoharides TC. Pharmacol. Rev. 2000, 52, 673–751. 58. Aucamp J, Gaspar A, Hara Y, Apostolides Z. Anticancer Res. 1997, 17, 4381–4385. 59. Mahesha HG, Singh SA, Rao AGA. Arch. Biochem. Biophys. 2007, 461, 176–185. 60. Hong J, Smith TJ, Ho CT, August DA, Yang CS. Biochem. Pharmacol. 2001, 62, 1175–1183. 61. Havsteen B. Biochem. Pharmacol. 1983, 32, 1141–1148. 62. Kim JS, Kwon CS, Son KH. Biosci. Biotechnol. Biochem. 2000, 64, 2458–2461. 63. Nijveldt RJ, van Nood E, van Hoorn DEC, Boelens PG, van Norren K, van Leeuwen PAM. Am. J. Clin. Nutr. 2001, 74, 418–425. 163 64. Laughton MJ, Halliwell B, Evans PJ, Hoult JRS. Biochem. Pharmacol. 1989, 38, 2859–2865. 65. Ong KC, Khoo HE. General Pharmacol. 1997, 29, 121–126. 66. Hodnick WF, Kung FS, Roettger WJ, Bohmont CW, Pardini RS. Biochem. Pharmacol. 1986, 35, 2345–2357. 67. Canada AT, Giannella E, Nguyen TD, Mason RP. Free Radical Biol. Med. 1990, 9, 441–449. 68. Galati G, Chan T, Wu B, O’Brien PJ. Chem. Res. Toxicol. 1999, 12, 521– 525. 69. Sang S, Lee MJ, Hou Z, Ho CT, Yang CS. J. Agric. Food Chem. 2005, 53, 9478–9484. 70. Long LH, Halliwell B. Free Radical Res. 2000, 32, 463–467. 71. Lambert, JD, Kwon SJ, Hong J, Yang CS. Free Radical Res. 2007, 41, 850–853. 72. Stohs SJ, Bagchi D. Free Radical Biol. Med. 1995, 18, 321–336. 73. Sang S, Lambert JD, Hong J, Tian S, Lee MJ, Stark RE, Ho CT, Yang CS. Chem. Res. Toxicol. 2005, 18, 1762–1769. 74. Galati G, Moridani MY, Chan TS, O’Brien PJ. Free Radical Biol. Med. 2001, 30, 370–382. 75. Galati G, Sabzevari O, Wilson JX, O’Brien PJ. Toxicology 2002, 177, 91–104. 76. Awad HM, Boersma MG, Boernen S, van Bladeren PJ, Vervoort J, Rietjens IMCM. Arch. Biochem. Biophys. 2000, 378, 224–233. 77. Awad HM, Boersma MG, Boeren S, van der Woude H, van Zanden J, van Bladeren PJ, Vervoort J, Rietjens IMCM. FEBS Lett. 2002, 520, 30–34. 164 78. Inverson SL, Shen L, Anlar N, Bolton JL. Chem. Res. Toxicol. 1996, 9, 492–499. 79. Bolton JL, Pisha E, Zhang F, Qiu S. Chem. Res. Toxicol. 1998, 11, 1113– 1127. 80. Bolton JL, Trush MA, Penning TM, Dryhurst G, Monks TJ. Chem. Res. Toxicol. 2000, 13, 135–160. 81. Halliwell B, Gutterridge JMC. Arch. Biochem. Biophys. 1990, 280, 1–8. 82. Smith C, Mitchinson MJ, Arudma OI, Halliwell B. Biochem. J. 1992, 286, 901–905. 83. Decker EA. Nutr. Rev. 1997, 55, 396–407. 84. Yamanaka A, Oda O, Nagao S. FEBS Lett. 1997, 401, 230–234. 85. Li Y, Trush MA. Cancer Res. 1994, 54, 1895s–1898s. 86. Breimer LH. Molec. Carcinogen. 1990, 3, 188–197. 87. Vuillaume M. Mutat. Res. 1987, 186, 43–72. 88. Ames BN. Science 1983, 221, 1256–1264. 89. Cerutti PA. Science 1985, 227, 375–380. 90. Yen GC, Duh PD, Tsai HL, Huang SL. Biosci. Biotech. Biochem. 2003, 67, 1215–1222. 91. Sahu SC, Gray GC. Cancer Lett. 1993, 70, 73–79. 92. Sahu SC, Gray GC. Cancer Lett. 1994, 85, 159–164. 93. Sahu SC, Gray GC. Food Chem. Toxicol. 1997, 35, 443–447. 94. Naito S, Ono Y, Somiya I, Inoue S, Ito K, Yamamoto K, Kawanishi S. Mut. Res. 1994, 310, 79–88. 95. Rahman A, Shahabuddin, Hadi SM, Parish JH. Carcinogenesis 1990, 11, 2001–2003. 165 96. Furukawa A, Oikawa S, Murata M, Hiraku Y, Kawanishi S. Biochem. Pharmacol. 2003, 66, 1769–1778. 97. Strick R, Strissel PL, Borgers S, Smith SL, Rowley JD. Proc. Natl. Acad. Sci. USA 2000, 97, 4790–4795. 98. Yen GC, Chen HY, Peng HH. J. Agric. Food Chem. 1997, 45, 30–34. 99. Inoue S, Ito K, Yamamoto K, Kawanishi S. Carcinogenesis 1992, 13, 1497–1502. 100. Hayakawa F, Kimura T, Maeda T, Fujita M, Sohmiya H, Fujii M, Ando T. Biochim. Biophy. Acta 1997, 1336, 123–131. 101. Austin CA, Patel S, Ono K, Nakane H, Fisher LM. Biochem. J. 1992, 282, 883–889. 102. Chang YC, Nair MG, Nitiss JL. J. Natural Prod. 1995, 58, 1901–1905. 103. Azuma Y, Onishi Y, Sato Y, Kizaki H. J. Biochem. 1995, 118, 312– 318. 104. Ross JA. Int. J. Cancer 1998, Suppl. 11, 26–28. 105. Iio M, Kawaguchi H, Sakota Y, Otonari J, Nitahara H. Biosci. Biotech. Biochem. 1993, 57, 1678–1680. 106. Elliott AJ, Scheiber SA, Thomas C, Pardini RS. Biochem. Pharmacol. 1992, 44, 1603–1608. 107. Zhang K, Yang EB, Tang WY, Wong KP, Mack P. Biochem. Pharmacol. 1997, 54, 1047–1053. 108. Divi RL, Doerge DR. Chem. Res. Toxicol. 1996, 9, 16–23. 109. Doerge DR, Divi RL. Xenobiotica 1995, 25, 761–767. 110. Martin KR, Appel CL. Nutr. Diet. Suppl. 2010, 2, 1–12. 111. Kiani J, Imam S. Nutr. J. 2007, 6, 33–41. 166 112. Hadi SM, Asad SF, Singh S, Ahmad A. IUBMB Life 2000, 50, 167–171. 113. Kaufmann SH. Cancer Res. 1989, 49, 5870–5878. 114. Yang GY, Liao J, Li C, Chung J, Yurkow EJ, Ho CT, Yang CS. Carcinogenesis 2000, 21, 2035–2039. 115. Nakagawa H, Hasumi K, Woo JT, Nagai K, Wachi M. Carcinogenesis 2004, 25, 1567–1574. 116. Vittal R, Selvanayagam ZE, Sun Y, Hong J, Liu F, Chin KV, Yang CS. Mol. Cancer Ther. 2004, 3, 1091–1099. 117. Elbling L, Weiss RM, Teufelhofer O, Uhl M, Knasmueller S, SchulteHermann R, Berger W, Micksche M. FASEB J. 2005, 19, 807–809. 118. Arunos OI. JAOCS 1996, 73, 1617–1625. 119. Smith C, Halliwell B, Aruoma OI. Food Chem. Toxicol. 1992, 30, 483– 489. 120. Puppo A. Phytochemistry 1992, 31, 85–88. 121. Hanasaki Y, Ogawa S, Fukui S. Free Radical Biol. Med. 1994, 16, 845– 850. 122. Babbs CF, Steiner MG. Methods Enzymol. 1990, 186, 137–147. 123. Fukui S, Hanasaki Y, Ogawa S. J. Chromatogr. 1993, 630, 187–193. 124. Hayakawa F, Ishuzu Y, Hosino N, Yamaji A. Biosci. Biotechnol. Biochem. 2004, 68, 1825–1830. 125. Guilbault G, Brignac PJ, Juneau M. Anal. Chem. 1968, 40, 1256–1263. 126. Inui T, Nakahara K, Uchida M, Miki W, Unoura, Kokeguchi Y, Hosokawa T. Bull. Chem. Soc. Jpn. 2004, 77, 1201–1207. 127. Bielski BHJ. Photochem. Photobiol. 1978, 28, 645–649. 128. Tarpey MM, Fridovich I. Circ. Res. 2001, 89, 224–236. 167 129. Gomes A, Fernandes E, Lima JLFC. J. Biochem. Biophys. Methods 2005, 65, 45–80. 130. Bartosz G. Clin. Chim. Acta 2006, 368, 53–76. 131. Czapski G, Bielski BHJ. J. Phys. Chem. 1963, 67, 2180–2184. 132. Kirby TW, Fridovich I. Anal. Biochem. 1982, 127, 435–440. 133. Margoliash E, Frohwirth N. Biochem. J. 1959, 71, 570–572. 134. Thomson L, Trujillo M, Telleri R, Radi R. Arch. Biochem. Biophys. 1995, 319, 491–497. 135. Beauchamp C, Fridovich I. Anal. Biochem. 1971, 44, 276–287. 136. Liochev SI, Fridovich I. Arch. Biochem. Biophys. 1995, 318, 408–410. 137. Munzel T, Afanas’ev IB, Kleschyov AL, Harrison DG. Arterioscler. Thromb. Vasc. Biol. 2002, 22, 1761–1768. 138. Green DR, Reed JC. Science 1998, 281, 1309–1312. 139. Benov L, Sztejnberg L, Fridovich I. Free Radical Biol. Med. 1998, 25, 826–831. 140. Patsoukis N, Papapotolou I, Georgiou DC. Anal. Bioanal. Chem. 2005, 381, 1065–1072. 141. Zhao H, Kalivendi S, Zhang H, Joseph J, Nithipatikom K, VasquezVivar J, Kalyanaraman B. Free Radical Biol. Med. 2003, 34, 1359–1368. 142. Zhao H, Joseph J, Fales HM, Sokoloski EA, Levine RL, Vasquez-Vivar J, Kalyanaraman B. Proc. Natl. Acad. Sci. USA 2005, 102, 5727–5732. 143. Zhang L, Huang D, Kondo M, Fan E, Ji H, Kou Y, Ou B. J. Agric. Food Chem. 2009, 57, 2661–2667. 144. Fink B, Laude K, McCann L, Doughan A, Harrison DG, Dikalov S. Am. J. Physiol. Cell Physiol. 2004, 287, C895–902. 168 145. Faulkner K, Fridovich I. Free Radical Biol. Med. 1993, 15, 447–451. 146. Kobayashi H, Gil-Guzman E, Mahran AM, Sharma RK, Nelson DR, Thomas AJ, Agarwal A. J. Androl. 2001, 22, 568–574. 147. Spasojevic I, Liochev SI, Fridovich I. Arch. Biochem. Biophys. 2000, 373, 447–450. 148. Laurindo FR, Pedro Md, Barbeiro HV, Pileggi F, Carvalho MH, Augusto O, da Luz PL. Circ. Res. 1994, 74, 700–709. 149. Finkelstein E, Rosen GM, Rauckmann EJ. J. Am. Chem. Soc. 1980, 102, 4994–4999. 150. Frejaville C, Karoui H, Tuccio B, Le Moigne F, Culcasi M, Pietri S, Lauricella R, Tordo P. J. Med. Chem. 1995, 38, 258–265. 151. Roubaud V, Sankarapandi S, Kuppusamy P, Tordo P, Zweier JL. Anal. Biochem. 1997, 247, 404–411. 152. Tuccio B, Lauricella R, Frejaville C, Bouteiller J-C, Tordo P. J. Chem. Soc. Perkin Trans. 2 1995, 2, 295–298. 153. Roubaud V, Sankarapandi S, Kuppusamy P, Tordo P, Zweier JL. Anal. Biochem. 1998, 257, 210–217 154. Vasquez-Vivar J, Kalyanaraman B, Kennedy MC. J. Biol. Chem. 2000, 275, 14064–14069. 155. Dambrova M, Baumane L, Kalvinsh I, Wikberg JE. Biochem. Biophys. Res. Commun. 2000, 275, 895–898. 156. Vijg J, Campisi J. Nature 2008, 454, 1065–1071. 157. Wang Y, Ho CT. J. Agric. Food Chem. 2009, 57, 8109–8114. 158. Wanasundara UN, Shahidi F. Food Chem. 1998, 63, 335–342. 159. Yen GC, Duh PD, Tsai H L. Food Chem. 2002, 79, 307–313. 169 160. Lambert JD, Sang S, Yang CS. Chem. Res. Toxicol. 2007, 20, 583–585. 161. Halliwell, B. Free Radical Biol. Med. 2009, 46, 531–542. 162. Green JR, Branch GEK. J. Am. Chem. Soc. 1941, 63, 3441–3444. 163. Stöggl WM, Huck CW, Stecher G, Bonn GK. Electrophoresis 2006, 27, 787–792. 164. DiSilvestro RA, David C, David EA. Free Radical Biol. Med. 1990, 9, 507–510. 165. Zielonka J, Sarna T, Roberts JE, Wishart JF, Kalyanaraman B. Arch. Biochem. Biophys. 2006, 456, 39–47. 166. Zielonka J, Zhao H, Xu Y, Kalyanaraman B. Free Radical Biol. Med. 2005, 39, 853–863. 167. Zielonka J, Kalyanaraman B. Free Radical Biol. Med. 2010, 48, 983– 1001. 168. Yoshino M, Haneda M, Naruse M, Murakami K. Mol. Genet. Metab. 1999, 68, 468–472. 169. Ahmad MS, Fazal F, Rahman A, Hadi SM, Parish JH. Carcinogenesis 1992, 13, 605–605. 170. Rahman A, Shahabuddin, Hadi SM, Parish JH, Ainley K. Carcinogenesis 1989, 10, 1833–1839. 171. Fazal F, Rahman A, Greensill J, Ainley K, Hadi SM, Parish JH. Carcinogenesis 1990, 11, 2005–2008. 172. Xu N, Chen ZM. Tea: bioactivity and therapeutic potential, Taylor & Francis, London, 2002. 173. Lin YL, Juan IM, Chen YL, Liang YC, Lin JK. J. Agric. Food Chem. 1996, 44, 1387–1394. 170 174. Jhoo JW, Lo CY, Li S, Sang S, Ang CYW, Heinze TM, Ho CT. J. Agric. Food Chem. 2005, 53, 6146–6150. 175. Tanaka T, Mine C, Inoue K, Matsuda M, Kouno I. J. Agric. Food Chem. 2002, 50, 2142–2148. 176. Wang H, Provan GJ, Helliwell K. Trends Food Sci. Technol. 2000, 11, 152–160. 177. Roberton A. The chemistry and biochemistry of black tea production: the non-volatiles. In Tea Cultivation to Consumption. Eds. Wilson KC, Clifford MN. 1992, pp. 553–601, Chapman and Hall, London. 178. Wiseman S, Zijp I, Weggemans R, Rietveld Anton. Protective Effects of Tea against Cardiocascular Diseases. In Protective Effects of Tea on Human Health. Eds. Jain NK, Siddiqi M, Weisburger J. 2006, pp. 65–75, London: CABI. 179. Higdon JV, Frei B. Crit. Rev. Food Sci. Nutr. 2003, 43, 89–143. 180. Conney AH, Lu YP, Lou YR, Xie JG, Huang MT. Proc. Soc. Exp. Biol. Med. 1999, 220, 229–233. 181. Landau JM, Wang ZY, Yang GY, Ding W, Yang CS. Carcinogenesis 1998, 19, 501–507. 182. Tijburg LBM, Mattern T, Folts JD, Weisgerber UM, Katan MB. Crit. Rev. Food Sci. Nutr. 1997, 37, 771–785. 183. Aruoma OI, Murcia A, Butler J, Halliwell B. J. Agric. Food Chem. 1993, 41, 1880–1885. 184. Su YL, Leung LK, Huang Y, Chen ZY. Food Chem. 2003, 83, 189–195. 185. Roginsky V, Alegria AE. J. Agric. Food Chem. 2005, 53, 4529–4535. 171 186. Jovanovic VS, Hara Y, Steenken S, Simic MG. J. Am. Chem. Soc. 1997, 119, 5337–5343. 187. Zuo Y, Chen H, Deng Y. Talanta 2002, 57, 307–316. 188. Huang D, Ou B, Prior RL. J. Agric. Food Chem. 2005, 53, 1841–1856. 189. Lin JK, Lin CL, Liang YC, Lin-Shiau SY, Juan IM. J. Agric. Food Chem. 1998, 46, 3635–3642. 190. Stevanato R, Fabris S, Momo F. J. Agric. Food Chem. 2004, 52, 6287– 6293. 191. Inoue MB, Inoue M. Fernando Q, Valcic S, Timmermann BN. J. Inorg. Biochem. 2002, 88, 7–13. 192. Muzolf M, Szymusiak H, Gliszczynska-Swiglo A, Rietjens IMCM, Tyrakowska, B. J. Agric. Food Chem. 2008, 56, 816–823. 193. Simic A, Manojlovic D, Segan D, Todorovic, M. Molecules 2007, 12, 2327–2340. 194. Janeiro P, Brett AMO. Anal. Chim. Acta 2004, 518, 109–115. 195. Mochizuki M, Yamazaki S, Kano K, Ikeda, T. Biochim. Biophys. Acta 2002, 1569, 35–44. 196. Azam S, Hadi N, Khan NU, Hadi SM. Toxicol. in Vitro 2004, 18, 555– 561. 197. Weber G. Chromatographia 1988, 26, 133–138. 198. Schweigert N, Acero JL, von Gunten U, Canonica S, Zehnder AJB, Eggen RIL. Environ. Mol. Mutagen. 2000, 36, 5–12. 199. Hayakawa F, Kimura T, Hoshino N, Ando T. Biosci. Biotech. Biochem. 1999, 63, 1654–1656. 172 200. Guo Q, Zhao B, Li M, Shen S, Xin W. Biochim. Biophys. Acta, Lipids and Lipid Metabolism 1996, 1304, 210–222. 201. Ward MD. J. Solid State Electrochem. 2005, 9, 778–787. 202. Grätzel M. J. Photochem. Photobiol C: Photochem. Rev. 2003, 4, 145– 153. 203. Ward MD, McCleverty JA. J. Chem. Soc., Dalton Trans. 2002, 275– 288. 204. Delpy DT, Cope M. Phil. Trans. R. Soc. Lond. B 1997, 352, 649–659. 205. Detty MR, Gibson SL, Wagner SJ. J. Med. Chem. 2004, 47, 3897–3915. 206. Kaim W. Coord. Chem. Rev. 1987, 76, 187–235. 207. Robin MB, Day P. Adv. Inorg. Chem. Radiochem. 1967, 10, 247–422. 208. Creutz C. Prog. Inorg. Chem. 1983, 30, 1–73. 209. Crutchley RJ. Adv. Inorg. Chem. 1994, 41, 273–325. 210. Kaim W. Coord. Chem. Rev. 2011, 255, 2503–2513. 211. Braterman PS, Song JI, Peacock RD. Inorg. Chem. 1992, 31, 555–559. 212. Braterman PS, Song JI, Peacock RD. Spectrochim. Acta A Chem. 1992, 48A, 899–903. 213. Haga M, Dodsworth ES, Lever ABP. Inorg. Chem. 1986, 25, 447–453. 214. Palmer RA, Piper TS. Inorg. Chem. 1966, 5, 864–878. 215. Joulie LF, Schatz E, Ward MD, Weber F, Yellowlees LJ. J. Chem. Soc., Dalton Trans. 1994, 799–804. 216. Barthram AM, Ward MD. New. J. Chem. 2000, 24, 501–504. 217. Ward MD. Chem. Soc. Rev. 1995, 24, 121–134. 218. Aguirre-Etcheverry P, O’Hare D. Chem. Rev. 2010, 110, 4839–4864. 173 219. Gao LB, Liu SH, Zhang LY, Shi LX, Chen ZN. Organometallics 2006, 25, 506–512. 220. Hush NS. Prog. Inorg. Chem. 1967, 8, 391–444. 221. LeVanda C, Bechgaard K, Cowan DO, Mueller-Westerhoff UT, Eilbracht P, Candela GA, Collins RL. J. Am. Chem. Soc. 1976, 98, 3181– 3187. 222. Westmoreland TD, Wilcox DE, Baldwin MJ, Mims WB, Solomon EI. J. Am. Chem. Soc. 1989, 111, 6106–6123. 223. Evans CEB, Naklicki ML, Rezvani AR, White CA, Kondratiev VV, Crutchley RJ. J. Am. Chem. Soc. 1998, 120, 13096–13103. 224. Nelsen SF. Chem. Eur. J. 2000, 6, 581–588. 225. Paul F, Lapinte C. Coord. Chem. Rev. 1998, 178-180, 431–509. 226. Kowalski K, Linseis M, Winter RF, Zabel M, Zalis S, Kelm H, Kruger HJ, Sarkar B, Kaim W. Organometallics 2009, 28, 4196–4209. 227. Ceccon A, Santi S, Orian L, Bisello A. Coord. Chem. Rev. 2004, 248, 683–724. 228. Creutz C, Taube H. J. Am. Chem. Soc. 1969, 91, 3988–3989. 229. Creutz C, Taube H. J. Am. Chem. Soc. 1973, 95, 1086–1094. 230. Creutz C, Chou MH. Inorg. Chem. 1987, 26, 2995–3000. 231. Rocha RC, Rein FN, Jude H, Shreve AP, Concepcion JJ, Meyer TJ. Angew. Chem. Int. Ed. 2008, 47, 503–506. 232. Concepcion JJ, Dattelbaum DM, Meyer TJ, Rocha RC. Phil. Trans. R. Soc. A 2008, 366, 163–175. 233. Demadis KD, Hartshorn CM, Meyer TJ. Chem. Rev. 2001, 101, 2655– 2685. 174 234. Scheiring T, Kaim W, Olabe JA, Parise AR, Fiedler J. Inorg. Chim. Acta 2000, 300-302, 125–130 235. Felix F, Hauser U, Siegenthaler H, Wenk F, Ludi A. Inorg. Chim. Acta 1975, 15, L7–L8. 236. Felix F, Ludi A. Inorg. Chem. 1978, 17, 1782–1784. 237. Poppe J, Moscherosch M, Kaim W. Inorg. Chem. 1993, 32, 2640–2643. 238. Baumann F, Kaim W, Garcia Posse M, Katz NE. Inorg. Chem. 1998, 37, 658–660. 239. Brewer KJ, Jr Murphy WR, Petersen JD. Inorg. Chem. 1987, 26, 3376– 3379. 240. Glockle M, Kaim W, Katz NE, Garcia Posse M, Cutin E, Fiedler J. Inorg. Chem. 1999, 38, 3270–3274. 241. Ketterle M, Fiedler J, Kaim W. Chem. Commun. 1998, 1701–1702. 242. Ruminski R, Kiplinger J, Cockroft T, Chase C. Inorg. Chem. 1989, 28, 370–373. 243. Chandra N, Sakar B, Fiedler J, Kaim W, Lahiri GK. Dalton Trans. 2003, 3550–3555. 244. Ghumaan S, Sarkar B, Chanda N, Sieger M, Fiedler J, Kaim W, Lahiri GK. Inorg. Chem. 2006, 45, 7955–7961 245. Collin JP, Laine P, Launay JP, Sauvage JP, Sour A. J. Chem. Soc., Chem. Commun. 1993, 434–435. 246. Patoux C, Launay JP, Beley M, Chodorowki-Kimmes S, Collin JP, Stuart J, Sauvage JP. J. Am. Chem. Soc. 1998, 120, 3717–3725. 247. Mountford HS, Spreer LO, Otvos JW, Calvin M, Brewer KJ, Richter M, Scott B. Inorg. Chem. 1992, 31, 718–720. 175 248. Mountford HS, MacQueen DB, Li A, Otvos JW, Calvin M, Frankel RB, Spreer LO. Inorg. Chem. 1994, 33, 1748–1752. 249. Spreer LO, Li A, MacQueen DB, Allan CB, Otvos JW, Calvin M, Frankel RB, Papaefthymiou GC. Inorg. Chem. 1994, 33, 1753–1755. 250. Spreer LO, Allan CB, MacQueen DB, Otvos JW, Calvin M. J. Am. Chem. Soc. 1994, 116, 2187–2188. 251. Zhou J, Li A, Lange C, Allan CB, Spreer LO, Otvos JW, Calvin M. Inorg. Chim. Acta 1996, 246, 241–248. 252. Marin TW, Homoelle BJ, Spears KG, Hupp JT. J. Phys. Chem. A 2002, 106, 1131–1143. 253. Karki L, Williams RD, Hupp JT, Allan CB, Spreer LO. Inorg. Chem. 1998, 37, 2837–2840. 254. Lu CC, Bill E, Weyhermuller T, Bothe E, Wieghardt K. J. Am. Chem. Soc. 2008, 130, 3181–3197. 255. Patra AK, Bill E, Weyhermüller T, Stobie K, Bell Z, Ward MD, McCleverty JA, Wieghardt K. Inorg. Chem. 2006, 45, 6541–6548. 256. Ray K, Bill E, Weyhermüller T, Wieghardt K. J. Am. Chem. Soc. 2005, 127, 5641–5654. 257. Sproules S, Wieghardt K. Coord. Chem. Rev. 2010, 254, 1358–1382. 258. Petrenko T, Ray K, Wieghardt K, Neese F. J. Am. Chem. Soc. 2006, 128, 4422–4436. 259. Garcia-Canadas J, Meacham AP, Peter LM, Ward MD. Angew. Chem. Int. Ed. 2003, 42, 3011–3014. 260. Garcia-Canadas J, Meacham AP, Peter LM, Ward MD. Electrochem. Comm. 2003, 5, 416–420. 176 261. Biancardo M, Schwab PFH, Argazzi R, Bignozzi CA. Inorg. Chem. 2003, 42, 3966–3968. 262. Mayerhöffer U, Deing K, Gruß K, Braunschweig H, Meerholz K, Würthner F. Angew. Chem. Int. Ed. 2009, 48, 8776–8779. 263. Xun S, LeClair G, Zhang J, Chen X, Gao JP, Wang ZY. Org. Lett. 2006, 8, 1697–1700. 264. Kiyose K, Kojima H, Nagano T. Chem. Asian J. 2008, 3, 506–515. 265. Escobedo JO, Rusin O, Lim S, Strongin RM. Curr. Opin. Chem. Biol. 2010, 14, 64–70. 266. Kaim W, Sarkar B. Coord. Chem. Rev. 2007, 251, 584–594. 267. Sando GM, Ray A, Spreer LO, Spears KG, Hupp JT. Inorg. Chem. 2000, 39, 3911–3914. 268. Yang G, Wu L, Jiang B, Yang W, Qi J, Cao K, Meng Q, Mustafa AK, Mu W, Zhang S, Snyder SH, Wang R. Science 2008, 322, 587–590. 269. Lee SW, Hu YS, Hu LF, Lu Q, Dawe GS, Moore PK, Wong PT, Bian JS. Glia 2006, 54, 116–124. 270. Erickson PF, Maxwell IH, Su LJ, Baumann M, Glode LM. Biochem. J. 1990, 269, 335–340. 271. Russo CD, Tringali G, Ragazzoni E, Maggiano N, Menini E, Vairano M, Preziosi P, Navarra P. J. Neuroendocrinol. 2000, 12, 225–233. 272. Szabo C. Nat. Rev. Drug Discovery 2007, 6, 917–935. 273. Zhao W, Zhang J, Lu Y, Wang R. EMBO J. 2001, 20, 6008–6016. 274. Goubern M, Andriamihaja M, Nübel T, Blachier F, Bouillaud F. FASEB J. 2007, 21, 1699–1706. 177 275. de Graaf W, Damsté JSS, de Leeuw JW. J. Chem. Soc. Perkin Trans. I 1995, 635–640. 276. Li L, Whiteman M, Guan YY, Neo KL, Cheng Y, Lee SW, Zhao Y, Baskar R, Tan CH, Moore PK. Circulation 2008, 117, 2351–2360. 277. Hughes MN, Centelles MN, Moore KP. Free Radical Biol. Med. 2009, 47, 1346–1353. 278. Beers RF, Sizer IW. J. Biol. Chem. 1952, 195, 133–140. 279. Lee HL, Ng ABL, Teng FYH, Halliwell B. Free Radical Res. 1999, 31, 67–71. 280. Mills G, Schmidt KH, Matheson MS, Meisel D. J. Phys. Chem. 1987, 91, 1590–1596. 281. Zhu J, Petit K, Colson AO, DeBolt S, Sevilla MD. J. Phys. Chem. 1991, 95, 3676–3681. 282. Hayon E, Treinin A, Wilf J. J. Am. Chem. Soc. 1972, 94, 47–57. 283. Orbán M, Epstein IR. J. Am. Chem. Soc. 1985, 107, 2302–2305. 284. Rábai G, Orbán M, Epstein IR. J. Phys. Chem. 1992, 96, 5414–5419. 285. Nashef AS, Osuga D T, Feeney RE. Anal. Biochem. 1977, 79, 394–405. 286. Li L, Salto-Tellez M, Tan CH, Whiteman M, Moore PK. Free Radical Biol. Med. 2009, 47, 103–113. 287. Morris JC. J. Phys. Chem. 1966, 70, 3798–3805. 288. Beckman JS, Beckman TW, Chen J, Marshall PA, Freeman BA. Proc. Natl. Acad. Sci. USA 1990, 87, 1620–1624, 289. Hughes MN, Nicklin HG. J. Chem. Soc. A 1968, 450–452. 290. Gutmann V. Coord. Chem. Rev. 1976, 18, 225–255. 178 291. Gutmann, V. The Donor-Acceptor Approach to Molecular Interactions, Plenum Press: New York, 1978. 292. Creutz C, Chou MH. Inorg. Chem. 1987, 26, 2995–3000. 293. McCoy S, Caughey WS. Biochemistry 1970, 9, 2387–2393. 294. Nakomoto K. Infrared and Raman Spectra of Inorganic and Coordination Compounds, John Wiley & Sons, Inc., New York, 1997. 295. Suh MP, Kong GY, Kim IS. Bull. Korean Chem. Soc. 1993, 14, 439– 444. 296. Cotton FA, Zingales F. J. Am. Chem. Soc. 1961, 83, 351–355. 297. Taylor RC, Horrocks WD. Inorg. Chem. 1964, 3, 584–589. 298. Sturrock PE, Ray JD, McDowell J, Hunt HR. Inorg. Chem. 1963, 2, 649–650. 299. Bazylinski DA, Hollocher TC. Inorg. Chem. 1985, 24, 4285–4288. 179 [...]... structure of the flavonols used Figure 2.2 HPLC chromatograms of E+, potassium superoxide induced oxidation of HE and the product formed from oxidation of HE by myricetin Figure 2.3 Oxidation products of HE by myricetin, KO2, and KO2 in the presence of flavonol (myricetin, quercetin) Figure 2.4 (A) Kinetic traces with [HE] = 25.6 µM and various concentrations of myricetin (B) Standard calibration curve of fluorescence. .. catalyst for air oxidation of HS−, forming hydrogen peroxide, disulfane, and elemental sulfur The NIR probe was selective towards HS− and did not react with other common biological anions It provides a convenient way for the detection of the HS− generation rate of a H2S donor of medical importance The addition of π-acceptor ligands such as cyanide and isocyanide ligands to the NIR active isovalent diiron... addition of HS− (8 µM) and the second, third, and fourth additions of HS− (8, 16, and 32 µM) Figure 5.4 Absorption spectra of HS− with 5% [Ru2]+ in 50 mM TrisHCl buffer pH 7.40 with time Figure 5.5 Standard curve of H2O2 with FOX reagent Figure 5.6 Concentration of H2O2 generated from HS− oxidation catalyzed by 5% [Ru2]+ with time Figure 5.7 Absorption spectra of HS− with 5% [Ru2]+ (30 minutes), NaHS2 and. .. induced by myricetin in the presence of Cu(II) (Bottom) Flavonol concentration dependent DNA damage in the presence of Cu(II) Figure 3.1 Chemical structures of major theaflavins and other oxidative products present in oolong and black tea Figure 3.2 Calibration curves of GA, ECG, EGCG, EC, and EGC Inset shows the calibration curves of TF, TFMG-a, TFMG-b, and TFDG Figure 3.3 HPLC chromatograms of the... EGCG, (B) EGC, (C) ECG, and (D) EC at pH 5.50 and pH 7.40 xiii Figure 3.8 Agarose gel electrophoretic analysis of pBR322 DNA damage induced by catechins Figure 3.9 DNA damage induced by catechins in the presence of Cu(II) Figure 3.10 Catechin concentration dependent DNA damage in the presence of Cu(II) Figure 3.11 DNA damage induced by (+)-catechin and EC in the absence and presence of Cu(II) in pH 7.4... good selectivity for HS− and Angeli’s Salt (NO− donor) However, because of its poor sensitivity, 2 was not suitable for use as a molecular probe for sensing HS− and Angeli’s Salt The combination of the NIR spectra of the two complexes covers a board range from 803 nm to 932 nm This may be of use as NIR materials in future applications xi List of Figures Figure 1.1 Basic structure of flavonoid, 2-phenylbenzopyran... 7.40 Figure 5.8 Absorption spectra of NaHS2 and HS2− extract in DCM Figure 5.9 Absorbance ratio changes of [Ru2]+ (16 µM) in a 50 mM Tris-HCl buffer (pH 7.40) recorded 1 minute after the addition of various anions (1600 µM) and reductants (16 µM) Figure 5.10 (A) Standard curve of HS− with DTNB (247 µM) (B) Standard curve of HS− with [Ru2]+ (16 µM) Figure 5.11 Release of H2S from GYY4137 (200 µM) in 50... constants of oxidation of flavonols by oxygen at 37 °C, pH 7.40 Table 3.1 Name and origin of the tea samples Table 3.2 Pro-oxidant activity of tea samples at pH 7.40 in terms of pseudo-first order rate constant k′ Table 3.3 Pro-oxidant activity of tea catechins, theaflavins, gallic acid, methyl gallate and pyrogallol at pH 7.40 in terms of pseudo-first order rate constant k′ Table 3.4 Composition of the... unavoidable byproducts of many biochemical processes or as the result of exogeneous factors such as smoking and air pollution [ 3 ] They can cause severe oxidative damage to biological molecules, especially to DNA, lipids, and proteins [4] Oxidative stress refers to the imbalance between the production of ROS and the activity of the antioxidant defense system [5,6] Increased production of ROS and lack of antioxidant... high concentration Our results illustrated the dual roles of polyphenolic compounds as pro-oxidants and antioxidants x The second part of the thesis examined the versatile chemical reactivity of near infrared (NIR) active bimetallic complexes of ruthenium and iron, which includes redox reaction and ligand substitution, for sensing application The NIR active mixed-valence diruthenium complex, [Ru2TIEDCl4]Cl,