Characterization and Structural Elucidation of the Complex

Characterization and Structural Elucidation of the Complex Mixture Thearubigins in Black Tea using Advanced Analytical Mass Spectrometry

by

Ghada Yassin

A Thesis submitted in partial fulfillment of the requirements for the degree of

Doctor of Philosophy

in Chemistry

Approved Dissertation Committee

Prof. Dr. Nikolai Kuhnert (Jacobs University Bremen, Germany)

Prof. Dr. Gerd-Volker Roeschenthaler (Jacobs University Bremen, Germany)

Prof. Dr. Michael Clifford (University of Surrey, United Kingdom)

Date of Defense: 23 January 2014

School of Engineering and Science

“The beauty of Science elaborates when a scientist mixes her passion of

research with the scientific knowledge embracing novel findings in the

uncertain world”

Abstract

Thearubigins are the most abundant pigments found in black tea, contributing to a wide range of

heterogeneous fractions of polyphenolic oxidation products. The term thearubigins has been

introduced fifty years ago but up till now their chemical nature remained unresolved and a

challenge for the scientists despite many attempts and efforts made to understand their

composition. Studies based on the health promotion properties of the black tea extracts and

thearubigins were carried out in the last years. Due to the increase significance of their potential

beneficial role on human health, identification of thearubigins and their characterization is

necessary. Moreover, their detailed structural knowledge is of high interest to understand their

sensory properties in order to assist the manufacturing and quality of black tea as an

economically important beverage. By employing a strategy in combining a series of advanced

powerful mass spectrometry techniques, the chemical composition and structural elucidation of

these compounds were achieved. A series of analytical mass spectrometry experiments including

ESI-LC tandem mass and ESI-LC-TOF mass spectrometry, MALDI-TOF-TOF mass

spectrometry, ESI-UPLC ion mobility mass spectrometry, MALDI imaging mass spectrometry

were employed in combination with experimental potentiostatic studies and theoretical

computational analysis.

In this work, two classes of compounds were identified and a new structural hypothesis in

thearubigin formation was further substantiated. The data revealed the presence of

polyhydroxylated dimers of the theanaphthaquinone and theasinensin C structures, which were

consistent with the polyhydroxylation hypothesis previously formulated. Furthermore, new class

1

of peroxo-/ epoxi- compounds in the series of theasinensin A were identified, thus indicating the

presence of H2O2 and its important contribution in the tea fermentation process. A series of novel

trimeric and tetrameric catechins were identified. The results showed that the structures of these

compounds provide for the first time experimental evidence on oxidative oligomerisation of

flavan-3-ols through an oxidative coupling pathway, most notably between theaflavin/

theasinensin and oxidized flavan-3-ols. Furthermore, the regiochemistry of the new trimeric

structures were suggested based on computational studies. Isomeric structures were also resolved

based on the energy dimension in direct infusion MS experiments.

An electrochemical model oxidation system was established. Theasinensins, theaflavins and their

gallate esters were identified and shown as the main reaction products obtained from the

electrochemical oxidation process performed, thus acting as the main precursors for further

oxidation processes in thearubigin formation.

Since thearubigins are one of the most complex mixtures present in nature, being composed of

sets of multiple isomers, the differentiation between isomeric structure of theasinensins,

proanthocyanidins B-type, and rutin (Quercetin-3-O-rutinoside) were studied according to their

mobility drift times acquired and the assignment of the individual isomeric structures were based

on comparison between their experimental drift times and theoretical collisional cross sections..

Furthermore, polyphenolic components present in the tea leaves and their spatial distribution

were located allowing a molecular mapping of the green tea constituents to be achieved. A new

free matrix MS imaging method was developed for the first time, where the biochemical process

occurring in a young green tea leaf has been determined. Finally, the importance of thearubigins

on the health benefits by studying its inhibitory activity on nutrient cultures containing both gram

positive and gram negative bacteria was highlighted.

2

This study characterizes the thearubigin as a complex mixture, reveals the important reaction

mechanism occurring and identifies novel components, shows the main polyphenolic occurring

components as main precursors in the oxidation process, demonstrates the isomerism present in

thearubigin fractions, reveals their initial biochemical processing in the leaf matrix, and shows its

important biological activity. In a word, this study serves a step forward for unraveling the

mystery of black tea.

3

Acknowledgements

First and foremost, I would like to express my deep gratitude to my advisor Professor Dr. Nikolai

Kuhnert for his continuous support during my Ph D. study and research, for his patience,

motivation, enthusiasm and immense knowledge. His guidance and attitude for conveying the

spirit of adventure in regard to research helped me all the time during my Ph D.

In addition, I would like to express my valuable and deep gratitude to my committee members

Professor Gerd-Volker Roeschenthaler and Professor Michael Clifford for their encouragement

and interest in my research project.

My sincere thanks also go to Dr. Sujatha Jayaraman, Dr. Jean H. Koek, Dr. Christian Grun, and

Professor Dr. Hans Gerd Janseen for offering me spring internship in Unilever laboratories in

Vlaardingen. My special thanks to Dr. Christian Grun for his nice and friendly welcome in

Vlaardingen, for his valuable supervision in performing the ion mobility experiments. Besides

Dr. Grun, I would like to thank especially Dr. Jean H. Koek for providing me a nice atmosphere

during my stay in Vlaardingen laboratories as well, for his interest, generosity, collaboration in

my research project, and for the valuable discussions and his encouragement and motivation in

my research.

I would like to thank deeply our present dean Professor Dr. Werner Nau for his generous and

valuable collaboration providing theoretical calculations to be studied in his laboratories. My

special and valuable thanks go to his fellow Ph D. Khaleel Assaf for his interest and

collaboration in performing these theoretical calculations adding an interesting impact to our

findings. My special gratitude goes also to my colleague Dr. Nina Vankova and Professor Dr.

4

Thomas Heine for the valuable help in assisting me for performing the theoretical computational

studies based on density functional theory. My thanks go to Ahmed Deyab Rezk (Ph D. fellow by

Professor Dr. Ulrich Mathias) for preparing the antibacterial experimental assays in performing

our studies.

I would like to express my gratitude to Unilever for funding my research project. Without the

financial support this work would not have been accomplished. My deepest acknowledgements

go to my lovely home Jacobs University Bremen for funding me as well and keeping me all the

time in a very nice atmosphere. My sincere gratitude goes to our present provost Prof. Dr.

Bernard Kramer, who made my stay and continuity in Jacobs University in a nice and beautiful

atmosphere.

Also I would like to express my thanks to all my fellow labmates for the fun we have had in the

last years. Thanks to Dr. Rakesh Jaiswal, Mohamed Elsadig Karar, and Abhinandan Shrestha for

providing me some samples of interest for performing my analysis. Also I thank all my friends in

Jacobs University especially Amira Rizk, Tuhidul Islam Tuhin, Rajesh Gavara, Jose Antonio

Salgado, Florin Ionita, Rosalyn Harrison, Martin Kangwa, Nourdinne Zibouche. I would like to

thank my colleague Boudewijn Hollebrands in Unilever Vlaardingen for assisting me in MALDI

imaging instrument and his friendly welcome. My thanks also go to Mrs. Rosie Mills for her

kind assistance and help during all my stay in Vlaardingen.

Last but not least, I would like to thank deeply my parents for supporting me spiritually

throughout my life, encouraging me, believing in me, and who are always proud of me.

5

List of Publications

[1] Ghada H. Yassin., Jean H. Koek, Nikolai Kuhnert. Identification of trimeric flavan-3-ol

derivatives in the SII black tea thearubigin fraction using ESI-Tandem and MALDI-TOF Mass

Spectrometry (manuscript under preparation)

[2] Ghada H. Yassin, Christian Grun, Jean H. Koek., Khaleel Assaf, Mohamed G. Elsadig

Karar , Rakesh Jaiswal, Werner Nau, Sujatha Jayaraman, Nikolai Kuhnert. Investigation of

isomeric structures in complex mixtures using ultra performance liquid chromatography coupled

with hybrid quadropole / ion mobility / time of flight mass spectrometry (manuscript under

preparation)

[3] Ghada H. Yassin, Jean H. Koek, Sujatha Jayaraman, Nikolai Kuhnert. Identification of

Novel Series Containing Polyhydroxylated and Epoxy-/ Peroxo-Components in Thearubigin SΙΙ

fractions (manuscript under preparation)

[4] Vikas Pawar, Ghada Yassin, Nikolai Kuhnert, Sujatha Jayaraman. Molecular weight

distribution of thearubigins in black tea (manuscript under preparation)

[5] Christian Grun, Rakesh Jaiswal, Marius F. Matei, Ghada Yassin, Nikolai Kuhnert.

Differentiation of prototropic ions in regioisomeric caffeoyl quinic acids by electrospray ion

mobility mass spectrometry (manuscript under preparation)

[6] Nikolai Kuhnert, Farnoosh Diarpoosh, Ghada Yassin, Agneiska Golon, Rakesh Jaiswal.

What is under the hump? Mass spectrometry based analysis of complex mixtures in processed

6

food-lessons from the characterization of black tea thearubigins, coffee melanoidines and

caramel, Food Function (8):1130-47

[7] Amal H. Ismail; Bassem S. Bassil; Ghada H. Yassin.; Beneta Keita, and Ulrich Kortz.

Ring Opening: The Fe16 Containing, Ln4 Stabilized 49-Tungsto-8-Phosphate Open Wheel

[Fe16O2(OH)23(H2O)9(P8W49O189)Ln4(H2O)20]11-, Chem. Eur. J. (18), 6163-6166

[8] Tarek Farhat, Ghada H. Yassin, Stephan T. Dubas , and Joseph Schlenoff. Water and ion

pairing in polyelectrolyte multilayer, Langmuir (American Chemical Society).15 (20), 6621-6623

Communication Publications and Poster Presentations in Conferences

[1] Ghada Yassin and Nikolai Kuhnert, Structural Elucidation of Black Tea Thearubigins,

Polyphenols Communication 2012, Vol. (1), 171-172.

[2] Ghada Yassin and Nikolai Kuhnert, Identification of trimeric flavan-3-ol derivatives in

the SII black tea thearubigin fraction using ESI-Tandem and MALDI-TOF Mass Spectrometry.

2nd International Congress on Cocoa, Coffee, and Tea 2013 Napoli, Italy

[3] Ghada Yassin and Nikolai Kuhnert, Structural Elucidation of Thearubigins in Black Tea

for Human Health. XXVI International Conference on Polyphenols 2012 Florence, Italy

[4] Ghada Yassin and Nikolai Kuhnert, Structural Elucidation of Thearubigins in Black Tea

for Human Health. 5th International Conference for Polyphenols and Human Health 2011

Barcelona, Spain

7

[5] Ghada Yassin and Nikolai Kuhnert, Structural Elucidation of Thearubigins in Black Tea.

GDCH Wissenschaft 2011 Bremen, Germany

Conference Seminar Talks

[1] Talk Topic: “Advanced Analytical Mass Spectrometry for the Characterization and

Structural Elucidation for the Complex Mixture Thearubigins in Black Tea”. Mass Spectrometry

in Supramolecular Chemistry and Biochemistry Conference 2013, Freie Universitate Berlin,

Germany

[2] Talk topic: “Analysis and Structural Elucidation of Thearubigins in Black Tea”

Symposium in Jacobs University Bremen 2011 Bremen, Germany

8

Table of Contents

Chapter-1 Introduction ......................................................................................................................... 17

1.1 Background ................................................................................................................................. 17

1.2 Black tea Manufacturing ............................................................................................................. 18

1.3 Chemical Composition of black tea ............................................................................................ 19

1.3.1 Polyphenols-Flavonoids ...................................................................................................... 19

1.3.2 Main black tea constituents ................................................................................................. 22

1.3.3 Other constituents in black tea ............................................................................................ 28

1.4 Thearubigins: History, definition, and challenges ....................................................................... 31

1.5 Towards a definition of thearubigins........................................................................................... 36

1.6 Health Benefits of thearubigins in black tea ............................................................................... 38

1.7 Aim of the project ....................................................................................................................... 40

Chapter-2 Identification of Novel Series Containing Polyhydroxylated and Epoxy-/ Peroxo- Components in the Thearubigin SII Fractions ............................................................................................ 44

2.1 Experimental Part ........................................................................................................................ 47

2.1.1 Chemicals and reagents ....................................................................................................... 47

2.1.2 Preparation of thearubigins ................................................................................................. 47

2.1.3 LC-MSn Method .................................................................................................................. 48

2.2 Results and Discussions .............................................................................................................. 49

2.2.1 Theanaphthaquinone and its homologous series ................................................................. 50

2.2.2 Theasinensin C and its homologous series .......................................................................... 58

2.2.3 Theasinensin A and its homologous series .......................................................................... 68

2.3 Conclusions ................................................................................................................................. 77

Chapter-3 Identification of Trimeric and Tetrameric Flavan-3-ols using ESI-Tandem and MALDI –TOF Mass Spectrometry ............................................................................................................................. 78

3.1 Experimental part ........................................................................................................................ 80

3.1.1 Chemicals and reagents ....................................................................................................... 80

3.1.2 Preparation of thearubigins TRs .......................................................................................... 80

3.1.3 LC-MSn ............................................................................................................................... 81

3.1.4 High Resolution LC-MS ..................................................................................................... 82

3.1.5 HPLC analysis .................................................................................................................... 82

3.1.6 Direct Infusion Tandem MS ................................................................................................ 83

3.1.7 Direct Infusion TOF ............................................................................................................ 83

9

3.1.8 Data analysis ....................................................................................................................... 83

3.1.9 MALDI-TOF MS ................................................................................................................ 84

3.1.10 Matrix Preparation .............................................................................................................. 84

3.1.11 MALDI Target .................................................................................................................... 85

3.1.12 Sample preparation for MALDI .......................................................................................... 85

3.1.13 Amsterdam Density functional Software (ADF) ................................................................ 85

3.2 Results and Discussions .............................................................................................................. 85

3.2.1 Identified trimeric compounds ............................................................................................ 88

3.2.2 Identified tetrameric compounds....................................................................................... 101

3.2.3 Homologous series for the new trimeric compounds ........................................................ 105

3.2.4 Theoretical calculations and Computational Analysis ...................................................... 108

3.3 Conclusions ............................................................................................................................... 115

Chapter-4 Potentiostatic measurement, a New Approach for Identification of the Main Reaction Products, Precursors occurring in the Tea Oxidation Process for Thearubigin Formation ....................... 116

4.1 Experimental Part ...................................................................................................................... 118

4.1.1 Sample preparation ........................................................................................................... 118

4.1.2 Direct infusion tandem MS ............................................................................................... 118

4.1.3 LC-MSn Method ................................................................................................................ 119

4.1.4 Amsterdam Density functional Software (ADF) .............................................................. 119

4.2 Results and Discussions ............................................................................................................ 120

4.2.1 Direct Infusion and LC MS experiments .......................................................................... 123

4.2.2 LC-MSn experiments ......................................................................................................... 126

4.2.3 Identification of different reaction products by direct infusion experiments .................... 134

4.2.4 Theoretical calculations and Computational Analysis ...................................................... 137

4.3 Conclusion ................................................................................................................................ 144

Chapter-5 Investigation of isomeric structures in complex mixtures using Ultraperformance Liquid Chromatography Coupled with Hybrid Quadropole /Ion Mobility/ Time of Flight Mass Spectrometry . 145


5.1.1 Preparation of Proanthocyanidin extract ........................................................................... 148

5.1.2 Preparation of Thearubigins .............................................................................................. 149

5.1.3 Sample Preparation ........................................................................................................... 149

5.1.4 Acquity UPLC-IMS-MS Conditions................................................................................. 149

5.2 Results and Discussion ............................................................................................................. 151

5.3 Conclusions ............................................................................................................................... 169

Chapter-6 MALDI imaging of the polyphenolic components present in tea leaves .......................... 170

10


6.1.1 Consideration for MALDI-MSI imaging .......................................................................... 172

6.2 Experimental part ...................................................................................................................... 173

6.2.1 Leaf sample preparation .................................................................................................... 173

6.2.2 MALDI imaging method and software ............................................................................. 174

6.2.3 Data processing and visualisation ..................................................................................... 174




Chapter-7 Beyond black tea composition- Health benefits effect of black tea thearubigin fractions 185

7.1 Experimental part ...................................................................................................................... 186


7.3 Conclusion ................................................................................................................................ 188

Chapter-8 Conclusions ....................................................................................................................... 189

11

List of Abbreviations

AFM Atomic Force Microscopy

α -HCCA alpha Hydroxycinnamic acid

C Catechin

CCS Collisional Cross Section

CG Catechin gallate

CID Collision Induced Dissociation

CTC Crush Tear Curl

CV Cyclic Voltammetry

DFT Density Functional Theory

Dt Drift Time

EC Epicatechin

E. Coli Escherischia Coli

ECG Epicatechin gallate

EGC Epigallocatechin

EGCG Epigallocatechin gallate

EIC Extracted Ion Chromatogram

12

ESI MS Electrospray Ionisation Mass spectrometry

FAO Food Agency and Agricultural Organisation

FTICR Fourier Transform Ion Cyclotron

GC Gallocatechin

GCG Gallocatechin gallate

HOMO Highest Occupied Molecular Orbital

HPLC High Performance Liquid Chromatography

H.R.F. Heterolytic Ring Fission

IMS Ion Mobility mass spectrometry

IR Infrared

POD Peroxidase Enzyme

PP Polyphenol Product

PA Proanthocyanidin

TF Theaflavin

TS Theasinensin

TR Thearubigin

LC-MS Liquid Chromatography-Mass Spectrometry

LDL Low Density Lipoprotein

13

LUMO Lowest Unoccupied Molecular Orbital

MS Mass Spectrometry

MSn MS Tandem mass spectrometry

MALDI MS Matrix Assisted laser Desorption Ionization Mass Spectrometry

MALDI-MSI Matrix Assisted laser Desorption Ionization Mass Spectrometry Imaging

MeCN Acetonitrile

NMR Nuclear Magnetic Resonance

Q.M. Quinide Methide fission

R.D.A. Retro Diels Alder fragmentation

TIC Total Ion Chromatogram

TOF Time of Flight

TPPO Tea Polyphenol Oxidase enzyme

TFA Trifloroacetic acid

Rt Retention Time

UV/VIS Ultraviolet-Visible Spectroscopy

UPLC Ultra performance Liquid Chromatography

U.S. United States

VLDL Very Low Density Lipoprotein

14

XC Exchange Correlation Energy

amu atomic mass unit

Da Dalton

ev Electron volt

Hz Herz

j.cm-2s Joule/ centimeter square second

m/z mass –charge-ratio

ml milliliters

min. Minute

mM millimolar

mg milligram

mV millivolt

µm micrometer

µl microliters

Pa Anodic Peak

Pc Cathodic Peak

sc Second

ºC Degrees Celsius

15

ºK Degrees Kelvin

ºA Angstrom

V Volts

16

Introduction

Chapter-1 Introduction

1.1 Background

Black tea, an unusual drink, is the most widely used ancient beverage in the world and

economically considered as the most consumed beverage after water. Black tea is mainly

produced by a systematically manufacturing process from the fresh green leaves of Camellia

Sinensis or Camellia Assamica plant. Camellia Sinensis was grown initially in China dating back

to 350 A.D., and then widely distributed all over the world. The average tea consumption per

person per year worldwide stands at 0.5 kg, which accounts to a consumption of an estimated

150 liters per year or 500 ml per person per day (Poulter S. Daily mail online, dailymail, co. u.k.,

27th june 2008). In 2012, the united nation food agency and agricultural organization (FAO)

reported that the tea production reached 4.05 million tons, and this comprised mainly to 20 %

green tea, 2 % oolong tea, where as the remainder accounts to 75% black tea. China remained the

largest producing country globally with an output of 1.4 million tons followed by Kenya, India,

and Indonesia. All the three types of tea are consumed abundantly in different regions of the

world; however, these are processed differently to obtain various degree of fermentation in the

final product. When considering black tea, the UN food agency and agricultural organization

17

Introduction

(FAO) reported that over the past five years black tea accounted to 65 % of the total global tea

production, 67% of the global consumption and 80 % of the global trade volume. The

appreciation of its health benefits in the developed countries helped the exports and resulted in

increased consumption. Furthermore, FAO reported that in the next ten years black tea

production will grow annually by 1.9 % to reach alone 3.28 million tons in year 2021, coming to

an equilibrium with the demand at a price of 2.75 $ per Kg (www.fao.org), thus making black tea

as one of the most economically important agricultural product.

1.2 Black tea Manufacturing

Black tea, produced from “Camellia Sinensis” plant (rich in flavan-3-ols), has mainly different

composition than the fresh tea leaves as oxidation, polymerization, and other modification of

original components taking place during the manufacturing process. Its composition, taste and

aroma1 vary with soil, season, climate, variety, and manufacturing process. Therefore, there are

two major types of black tea manufacturing processes, the orthodox and the crush –tear –curl

(CTC) process. The orthodox process consists of withering, rolling, enzymatic oxidation

(fermentation), drying and sorting of the tea leaves. High quality teas are usually prepared by this

method. The CTC process consists of the withered tea leaf being cut into a uniform size and then

being directly fed into a machine, where they are crushed, torn and curled by metal rollers. The

enzymes that are released, in particular tea polyphenoloxidase (TPPO) are collected and added to

the disrupted leaves again, in which tea polyphenols and the tea polyphenoloxidase enzymes are

brought in contact. Within the enzymatic oxidation process, the tea leaves become darker and

aroma compounds, which are responsible for the characteristic sensory properties, in particular

18

http://www.fao.org/

Introduction

the flavor of the tea beverage, are obtained. After the required fermentation, the leaves are dried

to stop the enzymatic oxidation process and the final moisture content in black tea is brought to 3

%. Finally, black tea is sorted into grades on the basis of the origin and other criteria such as size

or presence of stems. In both processing methods, the principal aim is to bring tea leaf enzymes

and tea leaf constituents into contact, so that enzymatic oxidation of the constituents may occur.

The method of preparation in withering, length of oxidation stage and process of deactivating the

leaf enzymes (drying) vary in black tea production.

1.3 Chemical Composition of black tea

1.3.1 Polyphenols-Flavonoids

The major polyphenolic substances in green tea leaves are classified into a group of plant

secondary metabolites called flavonoids. Flavonoids are produced by Camellia Sinensis but as

well by many other dietary plants such as cocoa, apple, plum and many others.2,3,4 These are

formed by a series of enzymatic condensation reactions between hydroxycinnamic acid and

malonyl residue giving rise to C6-C3-C6 base structure (Scheme 1.1). The C3 carbon bridge

between the phenyl rings is cyclized to form a third ring.5 The subclasses of flavonoids formed in

plants are flavanones, flavones, flavonols, flavanols, and leukanthocyanidins.6 (Figure 1.1)

19

Introduction

HO

CoA

O

CoAS OH

O OO

OH

OOH

HO

Flavanone

OH

OH

OOH

HO

naringenin ChalconeMalonyl CoA Hydroxycinnamoyl CoA

3

Scheme1.1. Biosynthesis of flavonoids.

O

OH

OOH

HO

Flavanone

O

OH

OOH

HO

O

OH

OOH

HO

O

OH

OH

HO

O

OH

OHOH

HO

OH

Flavanolol

OH

Flavonol OH

Flavanol

OH

Leukoanthocyanidine

O

OH

OOH

HO

Flavone

1

2

3456

761 51

8 41

31

21

A

B

C

11

Figure 1.1 Basic structures of the main classes of flavonoids.

20

Introduction

Fresh green tea leaves contain the two main subclasses: flavonols and flavanols. However, the

main polyphenolic constituents in green tea leaves are primarily the flavanols. These are mainly

characterized by the meta-5, 7-dihydroxy substitution of the A-ring and di- or tri-hydroxy

substitution of the B-ring, where these are known as catechins or flavan-3-ols.7 Flavan-3-ols

account for 10-25 % of the dry weight of a tea leaf and are sequestered mostly in the vacuoles of

the leaf cells. Their concentration is higher in unfermented tea, and the principal of these flavan-

3-ols structures are: catechin (C), epicatechin (EC), catechin gallate (CG), epicatechin gallate

(ECG), gallocatechin (GC), epigallocatechin (EGC), gallocatechin gallate (GCG) and

epigallocatechin gallate (EGCG) where these are illustrated in figure 1.2. These main

polyphenolic components present in the green tea leaf are responsible for several biological

activities and also contribute to the tea sensory properties giving it the astringency and bitter

taste.8,9

O

OR2

OH

R1

OH

OH

HOA C

B

O

OR2

OH

R1

OH

OH

HO A C

B

a) Epicatechin (EC) b) Catechin (C)

Where R1= OH or H

R2 = H or Galloyl unit

O

OH

OH

OH

21

Introduction

EC or C EGC or GC ECG or CG EGCG or GCG R1 H OH H OH R2 H H Galloyl Galloyl MW(amu) 290 306 442 458

Figure 1.2 Structures of different catechins (flavan-3-ols) in green tea leaf (Camellia Sinensis)

During the fermentation process, the majority of flavan-3-ols in the green leaf are chemically

transformed, resulting in dramatically lowered concentration in black tea if compared to green

tea.10 This observation has been rationalized by the assumption that after enzymatic oxidation,

flavan-3-ols are chemically converted into more complex polyphenols.

1.3.2 Main black tea constituents

During the fermentation process most of the important chemical changes take place in black tea.

Fermentation, a term usually associated with the action of microorganisms, is a rather misleading

term, since tea fermentation constitutes an enzymatically driven oxidative process in the absence

of microorganisms. Both polyphenoloxidase (TPPO) and peroxidase enzymes (POD) were

found to be responsible for the oxidation of flavan-3-ols which are acting as the main substrates

for the various oxidase enzymes in the green tea leaf, that is, epigallocatechin gallate (5),

epicatechingallate (6), epigallocatechin (4) and epicatechin (2) 7 (figure 1.3, p. 27). Oxidation

occurs mainly on the catechol moiety of the B-ring, affording highly electrophilic o-quinones,

which subsequently react with any appropriate nucleophile present in the tea leaf. When

considering the structure of dimeric flavan-3-ols isolated from black tea as experimental

evidence for their mechanism of formation, a second catechin serves as nucleophile attacking the

22

Introduction

ortho-quinone subsequently forming a new C–C bond, yielding as a reaction intermediate a

carbocationic dimer in an aromatic electrophilic substitution reaction (scheme 1.2). The later

undergoes in turn cascades of further reactions and yields ultimately different dimeric species,

known as theaflavins (7), theasinensins (12), theanaphthaquinones (8), and theacitrins (13)

(figure 1.3, p. 27). These dimeric compounds may yet again be subjected to further oxidation or

may act as nucleophiles in reacting with oxidized flavan-3-ols to form higher polymeric and

oligomeric new products known as thearubigins.11

O

OH

OH

OH

OH

OH

HO O

OH

OH

O

OH

HO

O

OH

OH

OH

OH

OH

HO

O

OH

O-

OH

O

OH

HO

O

2 e oxidation

ortho-quinone

O

OH

OH

OH

OH

OH

HOTheasinensins (re-aromatisation)

Theacitrin (further 2 e oxidationfollowed by Aldol chemistry)

Theaflavin (extrusion of CO)

Scheme 1.2 Suggested mechanism for the oxidative formation of the complex black tea polyphenols 11

23

Introduction

a) Bisflavanols

Bisflavanols otherwise known as theasinensins (12) (figure 1.3, p. 27), are a class of flavan-3-ol

dimers, linked by the C-C bond at the B-ring, thus forming a biphenyl function. These structures

are obtained by oxidative dimerisation of two epigallocatechins in their free or ester forms

undergoing loss of hydrogen.7 Theasinensins have been reported in both black and green tea with

several stereoisomers identified. Their existence was firstly reported by Roberts. In his work

Roberts identified the existence of atropisomers termed as theasinensin A (R-configuration) and

theasinensin D (S-configuration), when isolating the products of a model enzymatic oxidation

from a mixture of epigallocatechin gallate and epigallocatechin.12,13 Later on Tanaka

biosynthesized, isolated, and characterized the dehydrotheasinensins when investigating the

mechanism of theasinensin production of black tea.14 In 1992, Graham reported that theasinensin

are assumed to undergo further changes as the tea fermentation proceeds, and therefore may take

part in the thearubigin formation.15

b) Theaflavins and thearubigins

Theaflavins (7-9) (figure 1.3, p.27) are produced by oxidative coupling of epicatechin EC (2)

and epigallocatechin EGC (4) (figure 1.3, p.27) and their gallate esters forming a

benzotropolone structure. Roberts et al. reported firstly that these are the compounds that

correspond to the orange red acidic pigment on paper chromatography.16,17,18 Takino elucidated

their structure and suggested that it is the benzotropolone chromophore that results in the

characteristic reddish-orange colour of theaflavins. Later on theaflavin formation was confirmed

by model enzymatic oxidation reactions of appropriate pairs of EC having ortho- dihydroxy

24

Introduction

group and EGC having vic- trihydroxy group.19 Theaflavins are present as four derivatives:

Theaflavin -3-monogallate (9) which is formed by oxidation of EGCG (5) and EC (2),

theaflavin 31-monogallate (10) formed by oxidation of EGC (4) and ECG (6), theaflavin 3-31

digallate formed from oxidative coupling of EGCG (5) and ECG (6) (figure 1.3, p.27). Later

on, theaflavic acids and epitheaflavic acids were also identified in black tea infusion, and these

are formed by the oxidative products from coupling between catechin C (1) or epicatechin EC

(2) with gallic acid.20,21 These ethyl acetate soluble compounds have been well characterized and

contribute importantly to the color and taste of the black tea beverage, although they are present

only at levels of approximately 2 weight % of the dry tea infusion.22 It was reported that

theaflavins represent an important markers for the black tea quality. Tanaka et al. reported that

theaflavins are further oxidized to give further oxidation dimeric products known as

theanaphthaquinones (8).23 Furthermore, a new theaflavin type trimer was identified and

characterized, known as theadibenzotropolone (11) (figure 1.3, p. 27), and this trimeric

compound was formed by oxidative coupling between the galloyl ester group of theaflavin

gallate and the B-ring vic-trihydroxy of EGC (4) or EGCG (5). Model enzymatic oxidation

reactions were performed by Sang et al., which showed that these trimeric structures were

formed in the presence of peroxidase enzymes and H2O2.24 Most of the research work previously

performed, was mainly based on in vitro oxidation experiments, which confirmed that

theaflavins may further be degraded undergoing oxidation and coupling reactions with the

quinones in the formation of new oxidized polyphenolic components, thus corresponding to

numerous heterogeneous oxidation products. These products are characterized by reddish-brown

pigment and by their solubility in water; such heterogeneous fraction is referred as thearubigins.

Thearubigins account for 60-70 % of the solids in black tea infusion and contribute to a

25

Introduction

considerable extent to its color, taste and quality. Thearubigins readily form insoluble complexes

with caffeine and are responsible for the creaming phenomenon in the black tea infusion.25,26

However, the structures of thearubigin components remained poorly characterized in spite of lots

of effort done by many research groups over the last 50 years and on the other hand the

mechanism for their formation was until recently completely unknown.

26

Introduction

O

OH

OH

OH

HO

Catechin (1)C15H14O6, m/z [M-H] 289

O

OH

OH

HO

Epicatechin (2)C15H14O6, m/z [M-H] 289

Epigallocatechin (4)C15H14O7, m/z [M-H] 305

Gallocatechin (3)C15H14O7, m/z [M-H] 305

O

OHO

OH OHOH

OH

OHOH

OHO

Epigallocatechin gallate (5)C22H18O11, m/z [M-H] 457

O

OHO

OH OH

OH

OHOH

OHO

Epicatechin gallate (6)C22H18O10, m/z [M-H] 441

O

OH

O

OH

O

OH

HO

OH

HO

HOOH

OH

Theaflavin (7)C29H24O12, m/z [M-H] 563

O

O

OO

HO

OH

HO

HOOH

OH

OH

O OH

OHOH

OH

Theaflavin 3-gallate (9)C36H28O16, m/z [M-H] 715

O

OH

OO

HO

OH

HO

HOOH

OH

O OHO

OH

OHHO

Theaflavin 3´-gallate (10)C36H28O16, m/z [M-H] 715

OO

O

OH

O

OH

HO

OHOH

HO

Theanaphthaquinone (8)C28H22O11, m/z [M-H] 533

O

O

O

HO

OH

OH

OH

HO

HO

O

HOOH

OH

OH

OHOH

OH

O

OOH

OH

OH

Theasinensin A (12)C44H34O22, m/z [M-H] 913

Theacitrin-3-gallate (13)C37H28O17, m/z [M-H] 743

O

O

OO

HO

OH

HO

HOOH

OH

O

O OH

OHOH

OHO

OH

OHHO

Theaflavin 3, 3´di-gallate (14)C43H32O20, m/z [M-H] 867

O

OH

HO

O

O

HO OH

O

O

HO

HO

OOHHO

OOH

OH

OH

O

OH

O

OH

HO

OH

OH

OH

O

O

OH

O

OH

HO

OH

HO

HOOH

OH

O

O

Theadibenzotropolone (11)C50H38O21, m/z [M-H] 973

O

OH

OHOH

OH

HOOH

O

OH

OHOH

OH

HOOH

OH

OH

OH OH

Figure 1.3 Structures of catechins and their formal dimers previously identified in black tea

27

Introduction

1.3.3 Other constituents in black tea

a) Flavonol glycosides

Black tea contains also a mixture of glycosides of three major flavanols: kaempferol,

quercetin, and myricetin (figure 1.4). The glycosides have the carbohydrate moiety attached

at 3- position. These moieties consisted specifically of glucose, rhamnose, galactose, and

arabinose, and were configured in mono-, di-, and tri-glyceride units. Their presence was

confirmed by Engelhardt et al.8

O

ORO

OH

OH

HO

R1

R2

R1 = R2= H Kaempferol; R1= OH, R2 = H Quercetin; R1 = R2= OH Myricetin

Figure 1.4 basic flavonol structures present in both green and black tea

In addition to flavone-O-glycoside, Engelhardt reported that both green and black tea contain C-

glycosides, mainly containing glucose and arabinose moieties linked to 6- and or 8- positions of

apigenin. Further isomers containing both 6-, 8- diglycoside C- apigenin were additionally

identified. The total amount of flavone-C-glycosides was determined to be 0.48-2.69 g/kg dry

weight of the tea leaf.7 Due to the stability of the C-C bond, these compounds were assumed not

to undergo further hydrolysis upon the enzymatic activity and may retain their structures.

However, the sensory contribution of the flavonol glycosides was estimated to be of less impulse

for taste and flavor if compared to the other polyphenolic components in black tea.

28

Introduction

b) The non-phenolic substances present in black tea:

Methyl Xanthines such as caffeine are present in black tea accounting for 3-4% of the dry weight

of the tea leaves. Other purines of small quantities have been reported and these include

theobromine and theophylline present in 0.02 % respectively.15 Caffeine is responsible for the

taste of briskness due to its association with the oxidized polyphenols forming complexes which,

additionally result in the formation of tea cream on the tea brew.26

N

N N

N

O

O

R

CH3

R1

Caffeine: R = R1= CH3, MW = 194; Theobromine: R = H, R1= CH3; Theophylline: R=CH3, R1=H, MW=180

Figure 1.5 Structures of methyl xanthines present in black tea

Furthermore, different amino acids were detected in black tea infusions at a level of 3 weight%,

while the main amino acid present was identified to be theanine, which accounts for half of the

total amino acids contents. In 1954, theanine was identified by Cartwright et al. as the N-

methylated derivative of glutamine accounting for 1-2% of the dry weight of the tea leaf.27

HOHN

O

NH2

O Theanine

Figure 1.6 Theanine, the most available amino acid in black tea

29

Introduction

In addition to that, tea leaves contain trace amounts of inorganic ionic materials containing

calcium, aluminum, magnesium, and higher amount of fluoride. Further secondary metabolites

identified in black tea include phytoestrogens, which are suggested to have anti-osteoporotic

activity.

A recent detailed investigation was performed on the average chemical composition of black tea

infusions from different origins by Unilever tea excellence center (unpublished work, Pawar et

al.). In this work the different percentage composition for the constituents present in black tea

infusions were established. Within the black tea infusion, the aqueous phase contained mostly the

polyphenols having a composition [~30 % w/w], alkaloids [~ 8 % w/w], proteins + amino acids

[~15 %w/w], total sugars [~ 9% w/w]. However, these figures need to be evaluated with some

care for several reasons. Total phenol composition (TPP) was experimentally determined by the

Folin Ciocalcau assay9 and percentage figures relate to gallic acid equivalents, which are not

representative for tea polyphenols. Secondly the protein and alkaloid fractions contain residual

amounts of phenolics due to their ability to form strong non-covalent interactions with these

constituents and figures obtained for these are probably overestimates. Only data for ash,

moisture and sugars can be considered as reliable.

The main components of black tea were found to be polyphenolics including catechins, TFs,

phenolic acids, flavonols and its glycosides, while major part was contributing to the term

thearubigins, and these are shown in figure 1.7.

30

Introduction

figure 1.7 Chemical composition of a) total black tea extract b) total polyphenols TPP in black tea

By further analyzing and quantifying selected components of the TPP fraction a contribution of 8

% of phenolic acids, 7 % flavanol-glycosides, 2 % theaflavins, 8 % catechins and 70 %

thearubigins was observed on average, with catechin levels varying dramatically depending on

the origin of tea and the manufacturing processes employed.

1.4 Thearubigins: History, definition, and challenges

Thearubigins (TR) were firstly observed by Roberts in 1959 but the term was only used till 1962

for the first time. As mentioned in the previous section TRs were isolated as a fraction of

reddish-brown phenolic pigment forming a broad band on a two dimensional paper

chromatography.17,18 In his work, Roberts suggested that the chief precursors for the production

of TRs are the epigallocatechin and its gallate ester due to their low redox potential.28,29,30 In

1960, Roberts achieved separating TRs into fractions classified as SI and SII fractions according

to their polarity and chromatographic behavior, in order to isolate a TR rich fraction.31 These TRs

were defined by their brown color, their slight acidity and high water solubility revealing their

ability to partition in the aqueous phase of an ethyl acetate / water mixture. However, the

heterogeneity of TRs then became apparent, which restricted progress towards understanding

Ash14% Moisture

3%

TPP51%

Alkaloids6%

Sugars9%

Proteins17%

Tea compositiona)

TRs75%

Catechins8%

TFs2%

FGs7%

Acids8%

Phenolic compositionb)

31

Introduction

their chemical composition in black tea. In 1989, Clifford reported “Before the role of the

individual TRs can be ascertained with a view to increasing and / or preserving the level of those

major importance, their separation and chemical characterization are prerequisite steps in

research “. In his work, Clifford employed gradient elution reversed phase HPLC method for

separating TRs in black tea liquor, which showed an unusual characteristic gaussian hump.32

Interestingly, Clifford performed in vitro model enzymatic oxidation based on tea polyphenol

oxidase enzyme and mixture of selected flavan-3-ols as substrates for synthesizing pigmented

compounds in black tea. In his work, he showed many of the TRs to be related to their catechin

precursor. However, the model system did not produce all components separated from black tea

liquor. In 1993, Clifford and Robertson performed in vitro enzymatic model fermentation in

combination with reversed phase HPLC.33,34 Clifford identified that theaflavins TF alone were

not the substrates of polyphenoloxidase as in catechins. However, a mixture of EC and TF were

converted rapidly to complex mixture of TR like characteristics. Clifford concluded that

theaflavins TF were degraded with coupled oxidation with (epi)catechin quinones. In a

subsequent publication in 1997 Clifford and Davis identified a new class of yellow polyphenols

present in TRs, which they termed theacitrin.35 In 2003, Haslam proposed his structural

mechanism for TR formation based on self-condensation of hydroxyquinones.22 Reviews on TRs

were also authored by Habowy and Balentine summarizing the current knowledge about the

nature of black tea including data on the molecular weight distribution of TRs, the classes and

the number of compounds present in the TRs fractions.36,37 However, their conclusions were

speculative and controversial. The reason behind the lack of knowledge on TRs was for around

five decades based on the small number of analytical techniques suitable for providing a broader

32

Introduction

picture of this material, mainly the unusual chromatographic behavior of TRs producing an

unresolved gaussian shaped hump.

Figure1.8 HPLC chromatogram of black tea showing the unresolved gaussian hump

On the other hand, data based on the molecular weight distribution of a typical thearubigin

fraction were contradictory and controversial. While some authors suggested that the

thearubigins are polymeric materials of molecular weight reaching to 40 000 Da.38 Based on size

exclusion chromatography, Clifford suggested that the molecular weight of the thearubigin

constituents were not exceeding 2000 Da with no larger compounds being present. Later on,

when performing studies based on conventional and modern analytical methods such as

elemental analysis, IR, diffusion NMR, atomic force microscopy (AFM), Kuhnert et al. provided

clear evidence that the upper limit for the molecular dimensions of thearubigins components are

not exceeding 2000 Da. (figure 1.9). The experimental evidence further suggested that the

individual components of the TRs strongly interact via non-covalent interactions, while

excluding metal cation chelation as the contributor for the gaussian hump in reversed phase

column chromatography. These poorly defined but important children of nature remained still

difficult to unravel.

33

Introduction

Until 2010, by employing advanced analytical techniques including electrospray transform

ion cyclotron resonance mass spectrometry (ESI-FTICR), Kuhnert et al. showed that

thearubigins comprise of several thousands of compounds, an order of magnitude higher than

previously expected. Around 10,000 molecular ions, without counting their isomers, could be

resolved in a single direct infusion MS experiment.39,40 From the most abundent signals 1500

molecular formulae have been assigned.

In our collaboration with Unilever center in Bangalore, Simple techniques like dialysis and

ultrafiltration were used in studying the molecular weight distribution of black tea. It was

found that the black tea extracts that were subjected to dialysis, showed 75 % or more of the

tea solids have a molecular weight less than 3.5 kDa. MALDI-TOF–MS analysis confirmed

the majority of black tea extracts are low molecular weight compounds.(unpublished results,

Pawar et al.).

Figure 1.9 molecular weight distribution in thearubigin fractions

Based on the experimental MS data, Kuhnert et al. introduced the “oxidative cascade

0-3 kDa85%

3-10 kDa12%

above 10 kD3%

Size fractionationultracentrifugation or dialysis

34

Introduction

theory” rationalising TR formation and structures.

The oxidative cascade hypothesis suggests three levels of oxidative transformations of tea

polyphenols.

In a first step flavan-3-ol monomers are oxidized by TPPO affording highly electrophilic o-

quinones, which subsequently react with any appropriate nucleophilic catechin present in the tea

leaf, thus forming a C–C bond carbocationic dimer in an electrophilic substitution reaction

(scheme 1.2). Catechin dimers can be formed by four main mechanistically distinct reactions: for

example, the loss of proton formally leads to a theasinensin (11), ring contraction and

rearrangement leads to theaflavins (9), theanaphthaquinones (7), and theacitrins (12) (figure 1.3,

p. 27). These compounds, corresponding formally to dimers of flavan-3-ols, may yet again act as

nucleophiles in reacting with TPPO oxidized flavan-3-ols to form trimers, tetramers up to

hexamers not exceeding 2000 Da molecular weight. This oligomerisation part of the oxidative

cascade hypothesis is based solely on the molecular formulae obtained from FT-ICR-MS data,

with no further tandem MS supporting this part of the hypothesis.

A second level of oxidation within the oxidative cascade hypothesis involves oxidation of

monomeric and oligomeric flavan-3-ol derivatives to ortho quinones followed by nucleophilic

addition of water, which is the most abundant nucleophile in green tea leaves during the

fermentation process, thus leading to a series of polyhydroxylated structures. The mechanism of

the second level is illustrated in scheme 1.3. In each step, formally oxygen is inserted into an

aromatic CH bond. Due to the increased electron density in the aromatic nuclei with every

oxygen insertion, the subsequent oxidation becomes more facile and hence this constitutes a

cascade.

35

Introduction

OHO

OHOH

O

OOHH

HOHO

OHOH

O

OHHO

Scheme 1.3 Proposed mechanism for thearubigins undergoing oxygen insertion sequence

Finally polyhydroxylated structures are in a third oxidation step converted into their quinone

redox counterparts. The latter two aspects of the oxidative cascade hypothesis have been

confirmed by employing targeted tandem LC-MS experiments, where structures of dimeric

thearubigin components including polyhydroxylated theaflavins and theacitrins and their quinone

derivatives were identified.40 The average number of isomers identified in these experiments

have suggested that the total number of unique chemical entities present in the TR fraction is

around 30 000 compounds. Thus, the unusual chromatographic hump can be explained by the

large number of compounds present, combined with peak broadening additionally arising from,

non-covalant interactions such as hydrogen bonding and aromatic aromatic interactions between

phenolic compounds and other tea constituents.39,41

1.5 Towards a definition of thearubigins

As mentioned above thearubigins, known as a mysterious fraction, are very important to define

due to their high contribution in black tea. Up till now a clear definition about thearubigin is still

missing. Therefore, a series of definitions should be considered and suggested.

36

Introduction

1. Firstly from the historical point of view TRs can be defined as those compounds

characterized by a chromatographic gaussian hump, which could not be characterized

before applying high resolution mass spectrometry to their characterization.

2. Alternatively, these could be defined by spectroscopic properties and parameters,

suggesting all compounds showing UV/VIS absorption between 400 and 500 nm.

Alternatively, these are the compounds contained within the thearubigin hump excluding

all the well resolved floating peaks. On the other hand, as a general explanation from

mass spectrometry they are these compounds characterized by certain boundaries in van

Kravelen elemental ratio plot.11,42

3. If we have to define them by an experimental procedure, there are some isolation

experimental procedures done. However, the most common procedure to my knowledge

is the classical Roberts protocol. So in my work, I can define TRs as the compounds

obtained by caffeine procedure method.31

4. Furthermore, thearubigins are defined by a class of compounds. If we want to think of the

most proposed class of compounds for explaining thearubigins, we can rationalize the

answer that according to the oxidative cascade hypothesis, TRs can be defined as the

compounds obtained by oxidative pathways from flavan-3-ols precursors. Accordingly,

TRs are the phenolic compounds having C6-C3-C6 backbone structures. Therefore, it is

importantly to rationalize further my definition for TRs and the explanation for these

definitions can be demonstrated in my research work.

37

Introduction

1.6 Health Benefits of thearubigins in black tea

Considerable interest on the health promotion properties of black tea in general and thearubigins

in particular has been developed in the last years due to their high contribution as the major class

of polyphenolic compounds.43,5,44,45 A large number of epidemiological studies demonstrated that

the daily intake of black tea is inversely associated with cardiovascular disease, hypertension and

cancer.46 In a cross sectional study done by Hakim et. al. on black tea consumption carried out on

women, it was found that taking 6 cups a day of black tea could reduce the level of triglycerides,

LDL, and VLDL in the blood plasma, thus revealing the long term intake of black tea to be

associated with the reduction of coronary heart disease.47,48 Furthermore, studies done by

Clifford and Ioannides revealed that black tea can reduce DNA damage and mutagenesis, where

theaflavins showed to inhibit mutagenicity on a number of food carcinogens.49,50

Due to the presence of a homologous series of hydroxylated components present in thearubigin

fraction,40 black tea possess high antioxidant activity51,52,53 thus scavenging free radicals and

exhibiting chelation of redox metals. On the other hand, they can exert prooxidant effects

catalyzing DNA degradation in the presence of transition metal ions such as iron and copper to

possess chemopreventive and therapeutic properties against cancer.54 Because of these

properties, black tea may provide protection against damage to red blood cells due to oxidative

stress. Osawa et al. investigated the effects of black tea infusions on oxidative stress induced by

H2O2 and tert-butyl hydroperoxide, which in turn prevented cellular oxidative DNA damage in

normal rat epithelial RL-34 cells.55 In his study, Osawa demonstrated that black tea polyphenols

theaflavin not only have a protective effect on oxidative stress and cellular DNA damage, but

they also inhibited activated carcinogen –related DNA damage by suppressing cytochrome P450

1A1 (CYP1A1) enzyme in cells. Furthermore, in vitro studies done by Hodgson et al., showed

38

Introduction

that black tea infusions were found to reduce Cu-induced oxidation of lipoproteins in human

serum, thus showing a property of preventing atherosclerosis.56 Interestingly, thearubigins in

black tea showed to possess a good anti-inflammatory activity. Maiety et al. performed a model

induced colitis study in mice for observing the effect of TRs on inflammatory bowel disease.

Thearubigins in black tea were found to prevent diarrhea and the damage of colonic architecture,

thus reducing the levels of NO and O2- and the up regulation of the mRNA expression of

proinflammatory cytokine response and inducible NO synthase.57 Furthermore, Chaudhuri et al.

demonstrated that the anti inflammatory activity of black tea extract is due to its suppression

activity of inducible forms of enzymes such as cyclooxygenase and lipooxygenase which can

generate lipid mediators having proinflammatory activity, thus preventing the generation of free

radicals.58 Recently in 2011, Kuhnert showed that theaflavin and thearubigin extracts are potent

inhibitors of DNA transferases in the µM range, which suggested the health benefits of

thearubigins for the improvement of brain performance and mental health.59 Mikutis et al. could

show that TF digallate and TRs are able to selectively stabilize telomeric DNA, giving a

conceptually novel approach for rationalizing the life-span promoting effects of tea beverages.60

Since black tea is a source of phytoestrogens which may have a role in maintaining the bone

mineral density, this indicates that phytoestrogenic compounds present in black tea can be

effective against osteoporotic damage caused by deficiency of estrogen. A U.S. study conducted

among menopausal women aged between 50-79 years found that the total body bone mineral

density increased with daily consumption of tea.61

On the other hand, black tea was found to lower iron absorption. The absorption of iron was not

affected by consumption of tea in place of water but decreased when iron was provided in tea.

However, in healthy people with no risk of iron deficiency, tea drinking is unlikely to have much

39

Introduction

effect on iron status. These properties revealing the highly beneficial role in tea indicate that one

can use black tea in drug applications in the near future and expect lots of potentials in terms of

health benefits.

At the end of this section a critical word concerning the general state of TR research in the

context of health benefits is appropriate. TRs in terms of their dietary intake in g/day/human, are

considered as the most important dietary polyphenols. However, data based on their absorption,

metabolism and excretion are not available until today. Only if those are known, then it is

possible to establish, which constituents of TRs are biologically active and which targets are

affected by TRs in the human body, thus leading ultimately to an understanding of their

mechanism of action and promoting their dietary advice. Up till now, it is unknown how TRs are

metabolized and what the chemical nature of their metabolites might be. Therefore, TR

bioavailability43 and metabolism constitutes an urgent issue in this area of research in the near

future.

1.7 Aim of the project

Thearubigins, the most abundant group of phenolic pigments in black tea, thereby accounting for

an estimated 75% of the solids in a typical black tea infusion, remained for a long time poorly

characterized. Roberts stated “ Despite the fact that the thearubigins are the most abundantly

occurring of the polyphenolic oxidation products in black tea they are the substances whose

chemistry is at present least understood.” Therefore, structure elucidation of the individual

constituents of thearubigins is important for the understanding those components contributing to

many physicochemical properties of black tea that stand behind the astringent and bitter or

desirable taste62, the dark color41 and other properties; thus facilitating the manufacturing of tea

40

Introduction

based products. On the other hand one can ascertain the identification of chemical compounds,

which are responsible for the biological activities supposedly associated with the consumption of

black tea.

During my research period, I performed different analytical studies based on powerful advanced

analytical techniques for investigating thearubigin fractions in order to understand the mystery of

black tea. These include the following:

1. Employing advanced analytical techniques: electrospray ESI-conjugated to HPLC mass

spectrometry, tandem mass spectrometry (ion trap) MSn, electrospray ionization time of

flight mass spectrometry (ESI-LC-TOF-MS)63,64 for structurally identifying the

components in the oxidative cascade hypothesis not exceeding 1000 Da, thus confirming

the presence of new homologous series of different thearubigins structures and their sites

of oxygen insertions occurring.

2. Employing Matrix assisted laser desorption ionization time-of -flight (MALDI-

TOF)65,66,67 and MALDI-TOF-TOF-MS for fragmentation for specifically identifying and

structurally elucidating the higher molecular weight components exceeding 1000 Da. In

this part, I was able to identify new series of higher molecular weight TRs components in

the range between 1000-1500 Da revealing the presence of a series of novel trimeric and

tetrameric compounds which have never been identified before.

3. Performing computational analysis work based on Density Functional Theory68 in order

to suggest and propose the novel structures obtained and the suggested mechanism

reaction occurring; thus confirming in my studies that TRs are mainly formed by

41

Introduction

oligomerisation through the oxidative cascade hypothesis previously postulated by

Kuhnert.

4. Performing Potentiostatic measurements69 LC-MSn and direct infusion experiments for

one of the main dominating constituent in tea flavan-3-ols, that is, epigallocatechin

gallate EGCG (5), which could serve a good model for thearubigin formations. In my

research work, the different oxidation products and the chemical reaction processes were

identified, where formation of the main products occurring in the oxidation process

acting as precursors for thearubigin formation were obtained.

5. When analyzing TRs in black tea, multiple sets of isomers were commonly observed

during ESI-LC tandem mass MSn and direct infusion experiments. However, isomeric

structures in a complex mixture are difficult to separate, hence it was worthy to follow up

my work by employing ultra performance liquid chromatography UPLC (the latest

advancement in liquid chromatography) (www.chromatographyonline.com) conjugated to

/ ion mobility IMS / orthogonal acceleration time of flight mass spectrometry,70,71,72

which makes possible the isomers separation based on their rotational cross section .

6. Performing MALDI imaging mass spectrometry73,74 for the first time on tea leaves, in

order to gain a knowledge on the main components present, their spatial distribution, and

understanding the biochemical processes occurring in the tea leaf, where in this part I

developed a new methodology for the experimentation in the MALDI imaging when used

for plant science and this could serve a new development for researchers in the future.

7. Since tea polyphenols showed to possess a broad spectrum of biological functions, I was

interested in investigating the inhibitory effect of SΙΙ thearubigin fraction in both gram

42

http://www.chromatographyonline.com/

Introduction

positive and gram negative bacteria, where they showed to possess good inhibitory

activity.

8. Since black tea is the “Ultimate master of chemical diversity, such a study is an attempt

to characterize the thearubigins, to provide interesting information for their structural

elucidation, and to understand the mystery behind black tea chemistry associated with the

health benefits.

Herein, I summarize my findings in unraveling the mystery of black tea thearubigins by

employing different advanced analytical techniques for identifying and structural elucidation

the components of the thearubigins and these are discussed in details in every chapter

followed.

43

Identification of Novel Series Containing Polyhydroxylated and Epoxy-/ Peroxo- Components in the Thearubigin SII Fractions

Chapter-2 Identification of Novel Series Containing Polyhydroxylated and

Epoxy-/ Peroxo- Components in the Thearubigin SII Fractions

When analyzing black tea, which is classified as a complex mixture, Kuhnert et al. adopted a

novel data interpretation strategy based on van Krevelen and Kendrick analysis as a powerful

graphical tool for understanding the complexity of the thearubigin constituents present.11 The van

Krevelen analysis, known as elemental ratio analysis, is calculated from molecular formulae

obtained from high resolution MS data. Most commonly the H/C and O/C ratios are plotted in a

two dimensional graph against each other, where in the third dimension it is possible to add the

intensity of the observed peaks corresponding to their molecular formulas.75 A set of elemental

ratios of a given peak is considered as characteristic for a certain class of compound. Since all

the classes of natural product are characterized by elemental ratio boundaries, tentative

assignments of compound classes become possible.76 On the other hand, the Kendrick analysis

displays information with respect to mass defects present in a class of compounds.77 From the

high resolution masses data obtained, the Kendrick mass defect and the Kendrick mass for one

particular mass increment of choice are plotted in two dimensional graph. The compounds

having the same mass increments lie on a series of parallel lines with identical slopes and can be

visually identified, thus obtaining the existence of main series occurring in the natural product.78

44


This petrolomic strategy was used previously for characterizing petroleum products, when

assuming a series of compounds having the same mass increments such as the existence of a

series of CH2 increments identified in these products.79,80 Kuhnert et al. adopted this type of

strategy and applied it on the black tea chemistry. In this work, Kuhnert showed the existence of

homologous series of a variety of mass increments when taking Kendrick plot to study the

thearubigins in black tea. Since nature uses repeatedly building blocks to produce more complex

structures, both in biosynthesis and food processing, the data generated, revealed the presence of

different mass increments of oxygen (O), hydrogen (H2), gallates (C7H5O4), and catechins in

various oxygenation degrees acting as main building blocks in the black tea chemistry.

According to the data generated, oxygen insertion was observed as mass increment in the largest

number.39 However, the type and nature of the compounds were still unknown. Later on, by

employing ESI-FTICR-MS81,82 and tandem LC MS experiments, the data revealed that a

significant fraction of thearubigins consists of polyhydroxylated derivatives of catechin dimers (a

stepwise increase of oxygen increments within the dimeric structures) so called ‘Homologous

series”.40 Based on the experimental data, the oxidative cascade hypothesis was proposed as

previously discussed (in chapter 1). This hypothesis proposes the formation of the ortho-

quinones of different dimeric and oligomeric structures in the oxidation process, where these

could be followed by nucleophilic addition of water in the fermentation process, leading to

polyhydroxylated structures and this continues until all aromatic hydrogens are replaced by OH

functionalities, thus increasing the number of the components within the thearubigin fractions.

Previous investigations carried out on the oxidative cascade hypothesis revealed the presence of

homologous series of theaflavins, their gallate esters theaflavin mono- and di-gallates, and

45


theacitrins structures. Kuhnert et al. additionally showed that a significant portion of these

compounds are in redox equilibrium with their quinone counterparts.40

The aim of this study is to further confirm the second level of the oxidative cascade hypothesis

and to structurally elucidate the novel thearubigin components. For this investigation

electrospray ionization tandem mass spectrometry (ESI-LC-ion trap-MS) was employed for

characterizing thearubigins in homologous series reaching up to m/z 1000. A particular emphasis

was put on derivatives of theasinensins and theanaphthaquinones, the two dimeric catechin

derivatives so far not investigated. The experimental strategy used, was based on generating

extracted ion chromatograms EIC at the mass of the parent ion and their corresponding

hydroxylated derivatives, where the signals are strong enough to obtain MS2 and MS3 spectra. In

this approach, I was able to obtain the different m/z signals of the parent ion of

theanaphthaquinone and theasinensin C dimeric structures and their hydroxylated components in

the homologous series, where several regioisomers were detected at distinct retention times

increasing the number of structural variety of the identified new TR compounds.

Furthermore, interestingly a new series of compounds could be also identified for the first time

revealing the presence of H2O2 as a contributor in the fermentation process.83 In previous studies

done by Sinkar et al., it was confirmed the presence of H2O2 and its important role in the

oxidation process leading to the formation of thearubigins in the presence of peroxidase

enzyme.84

In this investigation by focusing on the ESI-LC-tandem mass MSn measurements, the

homologous series of dimeric structures theasinensin C and theanaphthaquinone are obtained,

and a new series containing epoxide derivatives of theasinensin A that has not been previously

investigated are reported.

46


2.1 Experimental Part

2.1.1 Chemicals and reagents

All chemicals and reagents were purchased from sigma Aldrich. Black tea sample of Ceylon

Assam was purchased from a local supermarket.

2.1.2 Preparation of thearubigins

Freshly ground black tea leaves (8 g) were added to 150 ml freshly boiled water and kept for

10 minutes in a Thermos flask, which was inverted every 30 seconds. The flask contents

were filtered through a Whatman No 4 filter paper to remove the leaves, and the remaining

brew allowed to cool to room temperature. Caffeine sufficient to achieve 20 mM was added

to the brew, stirred to ensure dissolution, and allowed to stand at 4 ºC for two hours, and

centrifuged at 23,300 × g for 20 min. The resulting precipitate was recovered and suspended

in boiling water, and partitioned against aliquots of ethyl acetate (40 ml) until no further

color was extracted (usually ×5).

The ethyl acetate-supernatant was removed and evaporated to dryness under nitrogen below 35

ºC, and the residue (TF fraction) recovered in 10 ml distilled water. The aqueous phase was

partitioned at 80 ºC against two volumes of chloroform, the decaffeinated liquid stored overnight

at –80 ºC, and freeze-dried. The freeze-dried material (TR fraction) was stored at –20 ºC until

required and reconstituted as required for the analysis. The yield of thearubigins in this method

was mainly 10 % and obtained as orange to light brown fluffy powders.

47


2.1.3 LC-MSn Method

The extracted thearubigin solutions were reconstituted to a 500mg/l concentration in 1:1

methanol/water, filtered through a 0.45 µm HPLC filter, and analyzed by LC –MS method using

Aligent 1100 series LC equipment and a DAD detector with a light pipe flow cell (recording at

400 and 245 nm and scanning from 200 to 600 nm). This was coupled to an ion trap mass

spectrometer fitted with an ESI source (Bruker Daltonics HCT Ultra) operating in the negative

ion Auto MSn mode to obtain fragment ion m/z. Separations were achieved on a POLARIS 5-

C18-A column (length 250 mm, diameter 3 mm, particle size 5 μm) with a step gradient elution

employing acetonitrile (MeCN) and water containing 0.005% formic acid, as follows: 8% MeCN

from 0 to 50 minutes, then changing to 31% MeCN for 10 minutes then changing to 25% MeCN

for a further 5 minutes. The column eluent was first directed to the UV detector and then to the

ESI interface operating with a capillary voltage of 1 Volt, and fragmentation amplitude was set

starting at 30% and ending at 200%. The capillary temperature was also set at 300 o C, Nitrogen

gas was used here as nebulizing and drying gas at a flow rate of 10 L/min and pressure of 10 psi

respectively.

48


2.2 Results and Discussions

For further investigating the hydroxylation process of the dimeric components and confirming

their presence within the oxidative cascade hypothesis in TR fractions, ESI-LC-MSn experiments

were carried out on the homologous series of theasinensin C (15), its gallate ester theasinensin A

(12) (figure1.3, p. 27). These structures are formed via dimerisation of two catechin derivatives.

Additionally homologous series of theanaphthaquinones (8) (figure 1.3, p.27) were investigated,

which is formed via further oxidation of theaflavin. Both dimeric structures are likely implicated

in the TRs formation.

LC tandem MS experiments were performed on the selected m/z parent ions of these compounds,

which are potentially the first member of the anticipated homologous series. Extracted ion

chromatograms EIC were created from the m/z values of hypothetical members of the

homologous series and in case strong signals were observed at the corresponding m/z values

targeted fragmentation experiments were carried on corresponding to MS2 and MS3 spectra. The

presence of fragment ions with their neutral losses was located; therefore the fragmentation

pattern and its mechanism have been determined. Later on, oxygen mass increments were

introduced and different signals at m/z values corresponding to the majority of predicted

hydroxylated derivatives were established. The interpretation here was based on the fragment

ions and the neutral losses for the parent ions observed, where the mechanism of fragmentation

has been previously established either from investigation of authentic reference compounds or

from literature precedents.7 The assumption for identifying the homologous series of different

hydroxylated parent ions was based on their m/z values and their MS2 fragment spectra observed.

Identical fragmentation mechanism has to be obtained for the whole homologous series. For

example, theaflavins (7 shown in figure 1.3, p.27) undergo fragmentation with a neutral loss of

49


an enone of m/z 137 (C7H6O3), which can be rationalized in terms of retro Diels-Alder

fragmentation (R.D.A.) at one of the benzopyran moieties.85 It has been confirmed that the whole

homologous series of this compound underwent the same fragmentation.40 On the other hand, its

gallate esters e.g. theaflavin mono/di-gallates (9, 10, 14 in figure 1.3, p. 27) underwent a loss of

152 Da and 170 Da indicating the loss of galloyl moieties. The whole series underwent oxygen

insertions within the sites of the fragment ions obtained or within the available sites of the

galloyl neutral losses having the same fragmentation mechanism. Hence, it was suggested that it

is reasonable to expect that similar fragmentation mechanism will apply to all members of the

homologous series investigated. There is ample evidence in the literature for many classes of

compounds, where minor structural changes do not alter the basic fragmentation pathways.86,87

Therefore, the hydroxylated derivative will fragment by the same mechanism as the parent

compound. It was anticipated that generally retention time on reversed phase packings would

decrease relative to the parent compound as hydroxylation increased unless internal hydrogen

bonding significantly increased the hydrophobicity.40

2.2.1 Theanaphthaquinone and its homologous series

Theanaphthaquinone (8) showed an MS spectrum at m/z 533 characterized by its MS2 fragment

ions: at m/z 514 as a base peak revealing the loss of H2O, at m/z 464 having a neutral loss of 69

Da which is rationalized by the loss of C3HO2 from the quinone part of this compound, a

fragment ion of a very weak intensity at m/z 396 showing a neutral loss of 137 Da indicating a

retro Diels Alder R.D.A. fragmentation occurring on a part of this molecule and this is shown in

figure 2.1 . Theanaphthaquinones have different sites available to hydroxylation, and in order to

evaluate the assignment of these peaks as hydroxy-theanaphthaquinone, EICs were prepared to

50


their corresponding predicted masses and the data are exhibited in table 2.1. Furthermore, by

using constant neutral loss analysis of the tandem mass, the presence of particular fragment ions

and their neutral losses corresponding to their chromatographic peaks could be located and the

results were confirmed.

Figure 2.1 EIC for theanaphthaquinone (8) at m/z 533.0 showing its characteristic MS2 and MS3 fragment ions

0 10 20 30 40 50 60 70 Time [min]

0.2

0.4

0.6

0.8

1.0

5 x10

Intens.

THEANAPHT187111.D: EIC 533.0 -All MS

260.5 402.6 460.6 533.0

-MS, 47.8min

304.8 348.8 396.6 464.6 514.8

-MS2(533.0), 47.9min

348.7 376.8 424.8

452.8 304.7 -MS3(533.0->514.7), 47.9min

0.5

5 x10 Intens.

0 1 2

4 x10

0 500

1000 1500

200 400 600 800 1000 1200 m/z

Loss of H2O Loss of C3HO2

51


OO

O

OH

O

O

HO

OHOH

HO

Theanaphthaquinone (8)C28H21O11, m/z [M-H] 533

OH

OOH

O

O

HO

OHOH

OH

OH

HC

O

O

m/z 464 neutral loss: 69 Da

Scheme 2.1 Fragmentation mechanism of theanaphthaquinone (8) exhibiting a neutral loss of 69 Da which can be

rationalized as the loss of C3HO2 on the quinone moiety. (The sites of deprotonation in [M-H]- have been selected

randomly).

Extracted ion chromatograms were created for the predicted polyhydroxylated compounds of

theanaphthaquinone (8+Ox). These showed the presence of m/z values corresponding to the

components with two oxygen insertion at m/z 565.1(C28H21O13), three oxygen insertion at m/z

581.2 (C28H21O14), four oxygen insertion at m/z 597 (C28H21O15), and five oxygen insertion at

m/z 613.1 (C28H21O16). However, these were of low intensity when compared to other

homologous series. Previous FT-ICR-MS data confirmed the elemental composition of these

ions.39

Only a single chromatographic peak was detected for the m/z 565.4 corresponding to 8+O2 for

the predicted di-hydroxylated derivative at a retention time of 50.1 min. Tandem mass

experiment of this compound gave MS2 fragment ions: at m/z 544.9 revealing a loss of H2O

molecule, and a characteristic fragment ion at m/z 494.5 revealing a loss of C3HO2. The same

52


fragmentation was previously observed in the parent ion, thus showing that this compound

belongs to the homologous series of theanaphthaquinone with two oxygen insertions.

Four chromatographic peaks were detected in the EIC at the m/z 581.2 (8+O3) at different

retention times of 53.5 min., 43 min., 34.7 min., and 30.9 min., as expected for the tri-

hydroxylated theanaphthaquinone. All exhibited an MS2 fragment ion at m/z 512.5 as a base peak

revealing a loss of 69 Da, thus corresponding to the loss of C3HO2 and revealing three oxygen

insertions within the compound. Minor fragment ions with different intensities at m/z 444.5 were

also observed revealing R.D.A. fragmentation as previously identified in the parent ion.

One chromatographic peak was detected for the m/z 597 at a retention time of 48.4 min

corresponding to 8+O4. Tandem mass experiments showed a transition of this precursor ion to a

fragment MS2 ion at m/z 528.5 as a base peak characterized by a neutral loss of 69 Da of C3HO2

and revealing four oxygen insertions within this compound.

Two chromatographic peaks were detected for the m/z 613 at retention times of 39 min and 28.1

min. as expected for the penta-hydroxylated theanaphthaquinone. Tandem mass fragmentation

experiment showed a main fragment ion MS2 at m/z 544.5 having a neutral loss of 69 Da of

C3HO2, thus indicating all the oxygens are inserted in the fragment ion which corresponds to

8+O5.

The main neutral losses of all the chromatographic peaks identified revealing losses of H2O, loss

of C3HO2, and weak R.D.A. fragmentation are shown in figures 2.2& 2.3, thus confirming the

assignment made.

53


Figure 2.2 Constant neutral loss of H2O molecule for theanaphthaquinone (8) at m/z 533 and its dihydroxylated

component m/z 565.4

Figure 2.3 Constant neutral losses of the hydroxylated components of theanaphthaquinones: m/z 565.4, m/z

581.2, m/z 597.2, m/z 613.2 exhibiting losses of H2O, C3HO2, and R.DA. Fragmentation

240.7 -324.7

280.8 -284.6

-186.5

406.9 -158.5

494.5 -70.9

544.9 -20.5 378.9

-MS2(565.4), 50.1min

444.5 -136.7

-68.7 512.5

-MS2(581.2), 53.4min

460.5 -136.5

528.5 -68.5

526.6 -70.4

-MS2(597.0), 48.4min

476.5 -136.7

-68.7 544.5

-MS2(613.2), 39.0min

0.5 1.0 1.5 2.0

4 x10 Intens.

0.25 0.50 0.75 1.00 1.25

4 x10

0.5

1.0

1.5 4 x10

0.0 0.5 1.0 1.5 2.0

4 x10

200 400 600 800 1000 1200 m/z

8+O2

8+O3

8+O4

8+O5

Loss of C3HO2 Loss of C7H5O3

Loss of C3HO2

Loss of C3HO2

Loss of C7H5O3

Loss of C7H5O3

0

1

2

3

4 x10

Intens.

0 10 20 30 40 50 60 70 Time [min]

304.8 -228.3 348.8

-184.2 396.6 -136.5

464.6 -68.5

-18.2 514.8 -MS2(533.1), 47.9min

240.7 -324.7

280.8 -284.6

-186.5

406.9 -158.5

494.5 -70.9

544.9 -20.5 378.9

-MS2(565.4), 50.1min

0.5

1.0 1.5

4 x10 Intens.

0.0 0.5 1.0 1.5

4 x10

200 400 600 800 1000 1200 m/z

Loss of H2O Loss of C3HO2

54


Therefore, theanaphthaquinone undergoes nucleophilic addition of H2O in the fermentation

process for TRs formation, where successive oxygen insertion takes place giving its

polyhydroxylated derivatives. These are exhibited in scheme 2.2 showing the hydroxylation

process. Again it should be noted that in the absence of authentic reference compounds,

assignments of detailed structures are tentative and information on regiochemistry of oxygen

insertion is speculative with structures shown corresponding to positions of random oxygen

insertion. Therefore, a new homologous series of theanaphthaquinone was observed within the

oxidative cascade hypothesis extending the members of the TRs family.

55


O

OH

O

O

OH

O

OH

HO

OHOH

HO O

OH

O

O

OH

O

OH

HO

OHOH

HO O

OH

O

O

OH

O

OH

HO

OHOH

HO

O

OH

O

O

OH

O

OH

HO

OHOH

HOO

OH

O

O

OH

O

OH

HO

OHOH

HOO

OH

O

O

OH

O

OH

HO

OHOH

HO

+O +O

+O

+O+O

HO HO

OH

HO

OH

HO

OH

HO

OH

m/z 533 m/z 549m/z 565

m/z 581m/z 597m/z 613

C28H21O11(8) C28H21O12 (8 +O1)C28H21O13 (8 +O2)

C28H21O14 (8 +O3)C28H21O15 (8 +O4)C28H21O16 (8 +O5)

OH

OH OH

OHOH

HO

Scheme 2.2 Homologous series of theanaphthaquinone (8) with successive oxygen insertion (regioisomers

selected randomly)

56


Table 2.1 Homologous series for theanaphthaquinone confirming the presence of the hydroxylated

derivatives with their selected MS2 & MS3 fragment ions and neutral losses observed. (MS2 fragment ions

colored in red are subjected for further fragmentation for MS3 fragment ions. The fragment ions

highlighted in bold font indicate the identified fragmentation obtained having characteristic neutral losses)

Molecular formula

R.T.

(min)

LC-MS [M-H] (m/z)

Fragment ion

LC-MS2 (m/z)

Neutral loss δM(S1-S2)

[Da]

Fragment ion

LC-MS3 (m/z)

(8) C28H21O11 47.9 533 514.8 (100%), 464.6 (55.9%),

396.6 (31.6%) 18.2,69, 137 304.7 (100%), 376.8 (40%),

424.8 (39.8%)

C28H20O11 64.7 531.5 462.5 (100%) 70 402.5 (100%)

(8+O2) C28H21O13 50.1 565.1 544.9 (100%), 494.5 (40.5%), 406.9 (69%)

20.7, 70.1, 158.5

240.6 (100%)

378.9 (20%), 240.7 (46%) 186.7, 324.7

(8+O3) C28H21O14 53.5 581.2 512.5 (100%), 444.5 (33.1%), 376.5 (6%)

69, 138 444.5 (100%), 376.5 (18%)

43 581.3 512.5 (100%), 444.5 (19%) 69, 137 444.5 (100%), 376.5 (25%)

34.7 581.3 512.5 (100%), 444.5 (20%) 69, 137 444.5 (100%)

30.9 581.2 512.5 (100%), 444.5 (31.4%) 69, 137 444.5 (100%), 376.4 (21.3%)

(8+O4) C28H21O15 48.4 597 528.5 (100%) 69 402.5 (100%)

(8+O5) C28H21O16 39 613.2 544.5 (100%), 476.5 (5%) 69, 137 402.5 (100%)

28.1 613.1 544.5 (100%), 476.5 (5.2%) 69, 137 402.5 (100%)

57


2.2.2 Theasinensin C and its homologous series

Another systematic search for homologous series was carried out on theasinensin C (C30H25O14)

(15 as shown in figure 2.5a, p.60) as a parent compound. Extracted ion chromatograms EICs

were also generated for the mass of parent ions expected for this homologous series.

Theasinensin C (15) is characterized by it mass spectrum with [M-H] ion at m/z 609 revealing

the presence of two isomers at different retention times of 34.6 min, and 48.1 min., with the

major chromatographic peak of this compound detected at 34.6 min (figure 2.4). Performing

tandem MS experiments with the precursor ion at m/z 609 gave a MS2 fragment spectra

characterized by a base peak at m/z 538.6 exhibiting a neutral loss of 70 Da, thus indicating a

loss of C3H2O2. Further MS3 tandem mass experiments with a precursor ion at m/z 538.7 gave the

characteristic MS3 fragment ion at m/z 470.5 corresponding to a loss of 138 amu, and this

indicates that the molecule underwent firstly a loss of C3H2O2, with fragmentation on one of the

two benzopyran A-ring moieties, and then retro Diels-Alder fragmentation with the loss of an

enone fragment. The fragmentation mechanism is shown in scheme 2.3.

58


Figure 2.4 EIC of theasinensin C (15) at m/z 609 with their characteristic MS2 spectra and their fragment ions

O

HO OH

OHHO

HO OHOH

OHOHO

O

HO

HO

HO OHOH

OHOHO

O

HO

HO

HO

HO

-C7H6O3

MS2-C3H2O2

O

HO OH

OHO

HO OHOH

OHO

HO

MS2/MS3

m/z609Theasinensin C (15)

R.D.A.

m/z 470.5

m/z138

m/z 70

m/z 538.5

OH OH

OH

HO O

Scheme 2.3 Fragmentation mechanism of theasinensin C (15) at m/z 609 undergoing loss of C3H2O2 and Retro-

Diels Alder fragmentation (The sites of deprotonation in [M-H]- have been selected randomly)

0 10 20 30 40 50 60 70 Time [min] 0.0

0.5

1.0

1.5

6 x10 Intens.

THEASC300511001.D: EIC 609.0 -All MS

462.9 538.5

-MS2(608.9), 48.9min

470.5 -MS3(608.3->538.6), 48.9min

330.8 438.9 538.6 299.7

-MS2(609.0), 34.6min

0

100

Intens. [%]

0

100

[%]

0

100

[%]

100 200 300 400 500 600 700 800 900 1000 m/z

C30H26O14

C30H26O14

Loss of C3H2O2

Loss of C7H6O3

1

2

1

2

Loss of C3H2O2

59


It is worth noting that while theasinensins have a biphenyl connection on the catechol B-rings,

proanthocyanidins (another dimeric catechin structure) is connected from the resorcinol A-ring to

C- ring as indicated in figure 2.5. Both classes of compounds are constitutional isomers of one

another with Proanthocyanidins as well reported previously in tea chemistry. Both can be

identified and differentiated by their different MS2 fragment ions obtained in LC tandem MS.

Proanthocyanidins show different MS2 fragment spectra characterized by retro Diels Alder

fragmentation R.D.A. with a neutral loss of 168 Da giving m/z at 441, Quinide Methide fission

Q.M. giving m/z 305, and heterolytic ring fission H.R.F. giving an m/z 483 with a neutral loss of

126 Da. Further details for differentiating between these two isomeric structures are discussed

briefly in chapter 5.

O

HOOH

OHHO

HO

OH

OH

OH

O

OH

HO

HO

OH

Theasinensin C (15)

O

HOOH

OH

HO

HOOH

O

OH

HOOH

OH

OH

HO

Proanthocyanidin (16)

C

A

Figure 2.5 The connectivity difference in the isomeric structures of theasinensins (15) and proanthocyanidins (16)

a) b)

60


In order to evaluate the assignment of the hydroxylated components within the homologous

series of theasinensin C (15), extracted ion chromatograms EIC were prepared for their

corresponding expected m/z values. While theasinensin C (15) C30H25O14 was detected at m/z

609, the EIC of its hydroxylated components were detected at m/z 641 for C30H25O16, at m/z

673.1 for C30H25O18, and m/z 705.6 for C30H25O20. The identity of the molecular formulae was

previously established by FT-ICR-MS.39 A detailed study for the hydroxylated derivatives of this

compound and their MS2 and MS3 fragment ions are exhibited in table 2.2

One chromatographic peak was identified for m/z 641 C30H25O16, corresponding to theasinensin

C with two oxygen insertions (15+O2), eluting at a retention time of 33.2 min. This compound

gave a fragment ion at m/z 604.8 as a major peak exhibiting a neutral loss of 36 amu

corresponding to a loss of two H2O molecules from this ion, and another characteristic MS2 ion

at m/z 570.5 revealing the loss of C3H2O2. Further fragmentation gave MS3 as base peak at m/z

504.5 revealing the further R.D.A. fragmentation occurring in this compound. Hence the

fragmentation pattern observed match fragment ions expected for a di-hydroxylated theasinensin

derivative.

Three isomers in the EIC with m/z 673 were detected at different retention times of 20.6 min.,

40.1 min, and 40.9 min, corresponding to the predicted tetra-hydroxylated theasinensin 15+O4

(C30H25O18). One of them eluting at 20.1 min. showed a transition from m/z 673.1 to a fragment

ion at m/z 602.5 as a base peak, indicating the loss of C3H2O2 and the oxygen insertions within

the dimeric ring. The second isomer of the m/z 673 eluting at 40.1 min. showed a transition from

m/z 673 to m/z 604.5 and to m/z 535.5 exhibiting a major loss of C3H2O2 on the benzopyran ring

and a minor loss of 136.6 amu due to R.D.A. fragmentation occurring, where as all the hydroxyl

groups are inserted within the fragment ion of theasinensin C ring. The third isomer of the m/z

61


673 was eluting at 40.9 min. showing a transition from m/z 673 to m/z 655 exhibiting a neutral

loss of 18 amu which is the loss of water and indicating three hydroxyl insertions within this

compound. A further MS3 fragmentation for this compound gave a transition of the MS2

fragment ion at m/z 655 to MS3 at m/z 503, which indicates that two hydroxyl groups are inserted

within the theasinensin fragment ion and one hydroxyl insertion was occurring within the

eliminated benzopyran moiety, where as this molecule was undergoing a loss of water at the

same time. This is briefly illustrated in Scheme 2.4 &Table 2.2. Hence the fragmentation pattern

observed match the fragment ions expected for the tetra-hydroxylated theasinensin derivative. In

a part of fragmentation pattern of these compounds a loss of water was observed, which has not

been observed in the parent compound.

62


O

OH

OHO

HO OH

OH

OHO

OH

HO

OH

HO

O

HO OH

OH

OHO

OH

HO

OH

HOOH

HO

HOOH

O

HO

HO

-C7H6O3

HOOH

OH

OH

O

HO OH

OH

OHOHO

OH

HOHO

OH

HOOH

O

HO

HO

-H2O

m/z 154

m/z 535.5

m/z 673

m/z 503

m/z 138

m/z 655

R.T. 40 min

R.T. 41 mina)

b)

-C3H2O2

HO

OH

HO

Scheme 2.4 Mechanism of retro Diels Alder fragmentations R.D.A. of a selected member of theasinensin C of the

tetrahydroxylated derivative at m/z 673 showing the oxygen insertion process occurring (The sites of deprotonation

in [M-H]- and regioisomers are selected randomly)

63


Figure 2.6 EIC prepared for Retro-Diels-Alder .R.D.A fragment ions in the homologous series of theasinensin C;

a) EIC at m/z 471, b) EIC at m/z 503, c) EIC at m/z 535 (The sites of deprotonation in [M-H]- and regioisomers are

selected randomly)

An additional set of EICs were created in the all MSn mode searching for fragment ions that

would be characteristic for hydroxylated theasinensins. Representative data are shown in figure

2.6. Here fragment ions originating from R.D.A. fragmentations at m/z 471, 503 and 535 were

observed corresponding to bisflavanol-like compounds with two or four oxygens inserted

respectively. These observed fragment ions point towards the presence of hydroxylated

theasinensins.

It should as well be noted that the fragment spectra obtained showing frequently loss of two

water molecules from the precursor ion might as well correspond to hydroxylated theasinensin

THEASINENSCRDA1.D: EIC 471.0



0.0

0.2

0.4

0.6

0.8

7 x10 Intens.

0.0

0.2

0.4

0.6

0.8

1.0 6 x10

Intens.

0

1

2

3 6 x10 Intens.

0 10 20 30 40 50 60 Time [min]

HO

O

HO OHOH

OHO

OH

HO

OH

HO

O

HO OH

OH

OHO

OH

HO

OH

HO

OH

OH

HO

HOHO

O

HO OH

OH

OHOHO

OH

HOOH

HOHO

OH

64


derivatives, in which due to a change of the aromatic substitution pattern the fragmentation

pathway is altered.

Only one isomer of C30H25O20 corresponding to 15+O6 was observed at m/z 705.6 giving a

transition to m/z 636.4 as a major peak intensity which reveals a loss of C3H2O2 from the A-ring,

and indicating all the hydroxyl groups are inserted within the theasinensin fragment.

All the above mentioned results are exhibited in table 2.2 confirming the presence of a new

homologous series for theasinensin C present in thearubigin fraction.

65


O

HOOH

OHHO

HO

OH

OH

OHO

OH

HO

HO

OH

Theasinensin C (15)

O

HOOH

OHHO

HO

OH

OH

OHO

OH

HO

HO

OH

O

HOOH

OHHO

HO

OH

OH

OHO

OH

HO

HO

OH

O

HOOH

OHHO

HO

OH

OH

OHO

OH

HO

HO

OH

O

HOOH

OHHO

HO

OH

OH

OHO

OH

HO

HO

OH

O

HOOH

OHHO

HO

OH

OH

OHO

OH

HO

HO

OH

OH OH

HO

HO

OH

OH

HO HO

OHOH

HOHO

OH

C30H25O14, (15)m/z 609

C30H25O16, (15+O2)m/z 641C30H25O15, (15+O1)

m/z 625

C30H25O17, (15+O3)m/z 657

C30H25O18, (15+O4)m/z 673

C30H25O20, (15+O6)m/z 705

O O

O

O2O

HO

OHOH

Scheme 2.5 Homologous series of theasinensin C (15) with successive oxygen insertion (regioisomers selected

randomly)

66


Table 2.2 Homologous series for theasinensin C confirming the presence of the hydroxylated derivatives

with their selected MS2 & MS3 fragment ions and neutral losses observed. (MS2 fragment ions colored in

red are subjected for further fragmentation for MS3 fragment ions. The fragment ions highlighted in bold

font indicate the identified fragmentation obtained having their characteristic neutral losses)

Molecular Formula

LC-MS [M-H]

m/z

R.T.

(min.)

Fragment ions

LC-MS2

m/z

Neutral Loss δM (S1-S2)

[Da]

Fragment ions

LC-MS3

m/z

C30H25O14 609.0 34.6 538.6 (67.8%), 456.9 (22.56%),438.9 (41.1%)

70.4, 152.1, 170.1,

330.8 (6.8%), 299.7 (65.40%) 278.2, 309.3

609.0 34.1 538.5 (100%), 470.5 (15.56%), 300.7 (76.6%),

70., 138.5, 308

470.5 (100%), 402.5 (13.9%)

270.6 (7.8%), 228.6 (4.4%) 338.4, 380.6

15+O2 C30H25O16 641.0 33.2 604.8 (100%), 570.5 (92%) 36.2, 70, 504.5 (100%), 436.5 (26.2%)

504.5 (21.5%) 137

15+O4 C30H25O18 673 40.9 655 (100%), 626.6, 604.5 (22.3%) 18,46, 69 502.9 (100%)

673.1 40.1 604.5 (100%), 535.5 (20.4%) 69, 136.6 544.5 (100%),492.8(15.8%)

673.1 20.1 654.1 (7%), 602.5 (100%), 544.5 (10.3%)

19, 70, 128.6, 544.5(100%)

15+O6 C30H25O20 705.6 47.8 681.6 (6.3%), 658.5 (6.8%), 636.4 (100%)

24, 47.1, 69.2

578.4(100%), 520.5(28.3%)

570.5(41.5%), 502.5(13.5%) 135.1, 203.1 502.5(31.8%)

67


2.2.3 Theasinensin A and its homologous series

Theasinensin A (12) as indicated (in figure 1.3, p.27) is characterized by its mass spectrum with a

pseudomolecular ion at m/z 913.2 [M-H] in the negative ion mode, exhibiting different

characteristic fragment ions. In the EIC of m/z 913 three signals are observed at different

retention times of 1.7 min, 16.3 min., and 22.5 min. The first fragment ion was observed at m/z

761.2 with a neutral loss of 152 amu, and a second fragment ion also observed at m/z 743.2 with

a neutral loss of 170 amu, both revealing the loss of one of the galloyl moieties of this

compound. A third fragment ion was observed at m/z 609.1 exhibiting a loss of 304 amu, which

can be rationalized due to a loss of two galloyl moieties from both parts of theasinensin A. A

base peak fragment ion was exhibited in the MS2 spectra at m/z 591.2, indicating the loss of one

galloyl group from one part of the molecule and another gallate ester from the second part. The

fragmentation of this compound and its mechanism are exhibited in figure 2.7 & Scheme 2.6.

68


Figure 2.7 Extracted ion chromatogram EIC of theasinensin A (12) with its characteristic MS2 spectra showing

their main fragment ions obtained

0 10 20 30 40 50 60 Time [min] 0.0

0.2

0.4

0.6

0.8

1.0 6 x10

THEASINENAC1801.D: EIC 913.0 -All MS

743.2

591.2 -MS2(913.2), 22.5min

447.2 529.3 573.3 609.1 743.2

591.2 -MS2(913.2), 16.3min

447.2 529.3 761.1

591.2 -MS2(913.3), 1.7min

0 50

Intens. [%]

0 50

100 [%]

0

50 100 [%]

200 300 400 500 600 700 800 900 1000 m/z

1 2

3

1

2

3 [M-H-C7H6O5] [M-H-C15H11O9]

[M-H-C7H5O4]

69


O

O

HO

O

OH

OH

OH

OHO

OHHO

HO

OH

HO

HOHO

O

O

OH

OH

OH

O

OH

OH

OH

O

O

O

HO

O

OH

OH

OH

OHO

OHHO

HO

OH

HO

HOHO

OH

m/z 761

O

O

HO

HO

O

OH

OH

OH

OH

HO

HOHO

OH

m/z 609

O

OH

OH

OH

O

O

O

HO

O

OH

OH

OH

OHO

OHHO

HO

OH

HO

HO

HO

m/z 743

-C7H6O5

O

O

HO

O

OH

OH

OH

OH

OH

HO

HOHO

m/z 591

-C7H5O4

OH

O

OHHO

HO

-C7H5O4 -C7H5O4

O

-152 -152

O

OH

OH

OH

m/z 913Theasinensin A (12)

Scheme 2.6 Mechanism of fragmentation for theasinensin A (12) at m/z 913.1 exhibiting its characteristic fragment

ions (The sites of deprotonation in [M-H]- and regioisomers are selected randomly)

In order to evaluate the assignment of the hydroxylated components within the homologous

series of theasinensin A (12), extracted ion chromatograms EIC were prepared for their

corresponding expected m/z values. While theasinensin A (12) was detected at m/z 913.2

(C44H33O22), EIC for its predicted hydroxylated components were prepared at m/z 929

(C44H33O23), at m/z 945 (C44H33O24), at m/z 961 (C44H33O25), at m/z 977.1 (C44H33O26), at m/z

993.3 (C44H33O27). These results are shown in table 2.3. Targeted tandem MS experiments of the

hypothetical hydroxylated components of theasinensin A (12) were carried out for evaluating the

70


presence of oxygen insertions within these compounds, in order to confirm whether a

homologous series of theasinensin A is present.

As shown in table 2.3; the EIC at m/z 929 showed a single peak at a retention time of 32.6 min.

corresponding to a theasinensin A with one oxygen inserted (12+O1). The precursor ion at m/z

929.1 displayed MS2 fragment ion at m/z 913.3 as a base peak having a neutral loss of only 16

amu. However, such MS2 fragmentation did not show similar fragmentation as the parent ion of

theasinensin A, undergoing the neutral losses of the galloyl moieties, which is its characteristic

fragmentation mechanism. Hence, this could not be in agreement with the hypothesis of the

homologous series as previously described that the whole series undergoes similar if not

identical fragmentation as the parent ion described. Neutral loss of 16 amu is characteristic for

either peroxide functionalities or epoxide functionalities, so that the compound observed was

tentatively assigned to a peroxo or epoxy derivative of a bisflavanol.

Two chromatographic peaks corresponding to the predicted di-hydroxylated component of

theasinensin A (12+O2) were detected in the EIC at m/z 945.0.The first peak was observed at a

retention time of 52.3 min where its MS2 spectrum exhibited fragment ions at m/z 926.6 showing

a loss of H2O, a base peak at m/z 909.1 with a neutral loss of 36 amu, thus indicating a further

loss of H2O molecule, and a characteristic fragment ion with a very weak intensity at m/z 657.2

revealing a loss of two galloyl moieties from this compound undergoing a loss of oxygen (2

C7H5O4 –O) and three oxygen insertion within the fragment ion, which is in agreement as the

fragmentation pathway of the parent ion theasinensin A. A second isomer of m/z 945 was

detected at a retention time of 47.8 min., the MS2 spectrum of this compound gave fragment ion

with a medium intensity at m/z 909.2 revealing a loss of two H2O.

71


Only one chromatographic peak in the EIC at m/z 961 corresponding to three oxygen insertions

(12+O3) was observed at 64.0 min. Its MS2 spectrum showed a characteristic base peak at m/z

916.9 suffering a neutral loss of 44 Da (loss of CO2) and further fragment ions were observed,

where one of which was detected co-eluting with a low intensity at m/z 657 exhibiting a neutral

loss of 304 Da, which indicates the loss of the two galloyl group from this compound as in the

parent ion and three oxygen insertion within the fragment ion occurring. The latter fragment ion

is in agreement with a tri-hydroxylated theasinensin structure.

Two chromatographic signals in the EIC at m/z 977.1 for the predicted tetra-hydroxylated

derivatives of theasinensin A (12+O4) were detected. The first one was observed at a retention

time of 51.2 min., exhibiting MS2 fragment spectra characterized by a weak fragment ion at m/z

959.4 having a neutral loss of 17.5 which corresponds to a loss of one hydroxyl group, and

another characteristic MS2 ion with high intense peak was identified at m/z 920.6 having a

neutral loss of 56 Da (-C3H4O) characteristic for flavanol C-ring cleavage. The second isomer at

m/z 977.3 was observed eluting at 32.8 min., and the latter showed no characteristic

fragmentation of theasinensin A. Thus, these two mentioned isomeric compounds could not be

considered as a part of the homologous series of theasinensin A, so no tetra-hydroxylated

components of this series is present.

A single chromatographic peak was observed in the EIC at m/z 993.3, a value predicted for a

penta-hydroxylated derivative of theasinensin A (12+O5). MS2 spectrum of this component

exhibited a characteristic fragment ion at m/z 961.2 exhibiting neutral loss of 32 amu, which

indicates the loss of two oxygens while only three oxygens were inserted within the fragment

ions of this compound. As in the previously identified compounds, this new component did not

72


undergo the same fragmentation as the parent compound, theasinensin A, thus indicating that this

compound does not belong to the homologous series suggested.

Accordingly, this observation can be rationalized by suggesting that a new series of oxidation

products in thearubigin formation may exist. These new compounds undergo mainly oxygen

losses (-32 Da, -O2) during MS2 is a characteristic for peroxides or epoxides. It is proposed that

these compounds can be formed in the presence of H2O2 in black tea thearubigin fractions, with

hydrogen peroxide acting as a nucleophile adding to orthoquinone moieties. It should be noted

that due to the α-effect hydrogen peroxide is around 1000 times more nucleophilic if compared

to water, the conventional nucleophile involved in TR chemistry. The existence of H2O2 in black

tea has been previously discussed and confirmed by Sinkar et al. in an attempt to investigate the

biochemistry and chemistry of thearubigins using polyphenols oxidase and peroxidase enzyme.84

In his work, Sinkar demonstrated that TPPO while oxidizing catechins generates H2O2 in

concentrations of 290-390 µM, which may play a role in oxidation process in the thearubigin

formation. In this part of my research, I was able to identify a further oxygenation process taking

place but here by nucleophilic addition of H2O2 as the second suggested mechanism in the

oxidation process for the formation of new peroxide or epoxide structures. The new suggested

mechanism is shown in scheme 2.7.

73


OHO

OHOH

O

OOO H

OHO

OHOH

O

OOOH

OHO

OHOH

O

OO

Scheme 2.7 proposed mechanism undergoing oxygen insertion by nucleophilic addition of H2O2

74


Table 2.3 A new class of compounds for theasinensin A detected with their selected tandem mass data

and their neutral losses, showing most of them different fragmentation mechanism than the parent ion.

(The fragment ions highlighted in bold font indicate the identified fragmentation obtained having the

characteristic neutral losses)

Molecular Formula

LC-MS [M-H]

m/z

R.T. (min.)

Fragment ions

LC-MS2

m/z

Neutral loss δM(S1-S2)

[Da]

12 C44H33O22 913.2 22.1 761.2 (28%), 743.2 (40.59%), 609.1 (5.44%), 152, 170, 304

591.2 (100%), 573.2 (9.37%), 555.2 (3.02%) 322, 340, 358

16.3 761.2 (34.01%), 743.2 (23%), 609 (14.92%), 152, 170, 304

591.2 (100%), 573 (30.37%),555.3 (10.84%) 322, 340, 358

529.3 (6.89%), 511.3 (5.86%), 447.2 (7.56%) 384, 401.9

15.7 761.2 (39.67%), 743.2 (33.17), 609.2 (15.1%),

591.2 (100%) 152, 170, 304, 322

573.3 (33.8%),555.3 (12.9%),529.3 (6.6%), 340, 358, 384

1.7 893.2 (8.33%), 761.1 (32.29%), 743.2 (31.27%) 20.1, 152, 170

609.2 (13.98%), 591.2 (100%), 573.3 (17.85%) 304, 322, 340

529.3 (7.12%), 447.2 (5.87%) 383.9, 466.1

12+O1 C44H33O23 929.0 32.6 913.3 (76.06%), 870.7 (100%) 16 , 58.3

12+O2 C44H33O24 945.0 52.3 926.6 (11.78%), 909.1 (100%), 886.6 (46%) 18 , 36, 58.4

826.4 (7.84%), 657.2 (8%) (609+3O) 118.6, 287.7

945.3 47.8 909.2( 26%), 886.6 (100%), 846.6( 15.9%) 36, 58.4, 98.5

826.7 (12.5%) (761+4O) 118.3

12+O3 C44H33O25 960.9 64 916.9 (100%), 831.0 (61%) 44, 130

656.8 (12.4%) (609+3O) 304

618.8 (9.23%), 591.2 (8.7%) 342, 369.9

75


12+O4 C44H33O26 977.1 51.2 959.4 (19.1%), 920.6 (100%), 862.6 (28.6%) 17.5, 56, 114.5

977.3 32.8 918.5 (57.8%), 863.6 (22.68%), 721.0 (39.8%) 58.1, 113, 255.6

12+O5 C44H33O27 993.3 49 961.2 (19.2%), 934.6 (100%), 896.6 (31.35%), 572.8 (34.70%) 32, 58.4, 96.4, 419.9

76


2.3 Conclusions

In conclusion, I characterized new thearubigin components by employing tandem LC-MSn

experiments. The data generated in the homologous series of theanaphthaquinone and

theasinensin C were consistent with the hypothesis previously formulated, in which

polyhydroxylated dimers of catechins are formed by nucleophilic addition of water to the

quinone oxidized counterparts of catechins. On the other hand, new compounds containing

oxygen insertions in the series of theasinensin A were identified, where the latter were

rationalized according to the nucleophilic addition of hydrogen peroxide, thus forming TR

components of a different structural class. Data on the homologous series of theasinensin A were

inconclusive, with MS data not allowing tentative structure suggestions. In summary this chapter

has complemented previous knowledge on TR chemistry and added evidence for the presence of

homologous series of formal oxygen insertion for the two previously not studied dimers of

catechins, theasinensins and theanaphthaquinone.

77

Identification of Trimeric and Tetrameric Flavan-3-ols using ESI-Tandem and MALDI –TOF Mass Spectrometry

Chapter-3 Identification of Trimeric and Tetrameric Flavan-3-ols using ESI-

Tandem and MALDI –TOF Mass Spectrometry

Theaflavins, their gallate esters and Thearubigins are the major pigments in black tea, and it is

generally accepted that these are formed by the oxidation of flavan-3-ols acting as the main

building blocks in tea fermentation.12,13,88,89 Theaflavins, which are reddish-orange in color, were

clearly characterized and their structures were well studied. It is well known that theaflavins

contribute to the properties of tea such as color and taste.90,91 However, theaflavins content in

black tea is usually 0.8-2.8% of the dried leaf depending on the fermentation.92,93 On the other

hand, thearubigins constitute up to 70 % of the solids in black tea infusion and are therefore, very

important with respect to color, and taste, as well as biological activities.32,61 However, little is

known about thearubigins and their chemical structures remain unclear and speculative despite

substantial efforts by many research groups.94 And this is due to the difficulties in separating the

intermediates in a very complex mixture by conventional chromatography forming as discussed

before an unresolved gaussian hump in a reversed phase chromatography.32, 95 Substantial

progress was made by Sang et al. for identifying the trimeric structures such as

theadibenzotropolone A,24 by Tanaka et al. by characterizing theanaphthaquinones,23 and as well

as by Wan et al. when synthesizing theaflavic acid 96 during investigating the chemical oxidation

78


of epicatechin gallate. However, these materials were obtained in model reactions synthesized in

vitro in the lab., purified, isolated and determined their structure using spectroscopic methods.

By using ultra performance high resolution FTICR mass spectrometry, Kuhnert et al. showed

that thearubigins fraction constitute an enormous wealth of structural diversity contributing to 10

000 resolved signals without considering isomerism. Similarly these investigations showed that

in direct infusion experiments, even with 1 Da window set for the ion trap, numerous regio- and

stereoisomers were trapped. Based on their fragment spectra these were accompanied by

multiply charged ions, so that their spectra were not clean as in simple mixtures. However, the

resultant fragment spectra could depend strongly on the collision energy applied in the collision

induced dissociation (CID), where this contribution in high resolution mass spectrometry has not

yet been explored. Accordingly, the collision energy applied in the CID was varied since

different molecules require for bond breaking different levels of applied external energy

transferred into their normal modes of vibration. Hence, with varying the collision energy,

different structures trapped were fragmented at different applied collision energy. Here the

fragmentation energy was mainly used for the first time as an additional dimension for

compound separation in direct infusion tandem mass experiments. Additionally, MALDI-TOF

mass spectrometry under optimised conditions was employed as a complementary technique for

further structural elucidation of the compounds in black tea thearubigins.97,98 The main

advantages of MALDI-TOF-MS,99 is its high sensitivity and ability to produce exclusively singly

charged ions of higher molecular weight compounds with mass precision, and consequently by

using the LIFT method in MALDI-TOF-TOF-MS characteristic fragmentation spectra for each

identified parent ion were obtained.

79


The aim of this chapter is the analysis of selected TR components with molecular weights

ranging from 1000 to 1500 Da including possible identification of isomers, using a selection of

appropriate tandem mass spectrometric techniques. The SII fraction of the TRs were chosen from

four different commercial black teas (Assam, Ceylon, South Indian, Chinese, Assam Ceylon) by

a caffeine precipitation protocol originally developed by Roberts.11,31 From this black tea TR

samples, intense ions similarly and commonly appearing in the mass spectra within all the

fractions were selected, thereby ensuring that the analytes chosen are common components in

many black tea beverages.

Analyzing the fragmentation spectra at several collision energies and the aforementioned

techniques employed, provides structural information on many of the complex oligomeric

thearubigin components in a generic straightforward way without additional (laborious)

purification of the thearubigin fractions.

3.1 Experimental part

3.1.1 Chemicals and reagents

Matrix 2, 5 dihydroxyacetophenone and all other reagents such as methanol, water and acetone

were purchased from Sigma Aldrich Company. Black tea samples were provided by Unilever and

others were purchased from a local supermarket.

3.1.2 Preparation of thearubigins TRs

Black tea leaves (8 g) were added to 150 ml freshly boiled water and kept for 10 minutes in a

Thermos flask, which was inverted every 30 seconds. The flask contents were filtered through a

80


Whatman No. 4 filter paper to remove the leaves, and the remaining brew allowed for cooling to

room temperature. Caffeine sufficient to achieve 20 mM was added to the brew, stirred to ensure

dissolution, and allowed to stand at 4 ºC for two hours, and centrifuged at 23,300 ×g for 20 min.

The resulting precipitate was recovered and suspended in boiling water, and partitioned against

aliquots of ethyl acetate (40 ml) until no further color was extracted (usually ×5). The ethyl

acetate-supernatant was removed and evaporated to dryness under nitrogen below 35 ºC, and the

residue (TF fraction) recovered in 10 ml distilled water. The aqueous phase was partitioned at 80

ºC against two volumes of chloroform, the decaffeinated liquid stored overnight at –80 ºC, and

freeze-dried. The freeze-dried material (TR fraction) was stored at –20 ºC until required and

reconstituted as required for the analysis. The thearubigins were obtained as orange to light

brown fluffy powders.

3.1.3 LC-MSn

The LC equipment (Agilent 1100 series) comprised a binary pump, an auto sampler with a

100 µL loop, and a DAD detector with a light-pipe flow cell (recording at 400 and 254 nm

and scanning from 200 to 600 nm). This was interfaced with an Ion-trap mass spectrometer

fitted with an ESI source (Bruker Daltonics HCT Ultra) operating in the negative ion MSn

mode to obtain fragment ion m/z. Tandem mass spectra were acquired using a ramping of the

collision energy. Maximum fragmentation amplitude was set to 1 volt, starting at 30% and

ending at 200%. MS operating conditions (negative mode) had been optimised using

theaflavin-3-gallate (9) with a capillary temperature of 300 oC, a dry gas flow rate of 10

L/min, and a nebulizer pressure of 10 psi.

81


3.1.4 High Resolution LC-MS

High Resolution LC-MS in the negative ion mode was carried out using the same HPLC

equipped with a MicrOTOF Focus mass spectrometer (Bruker Daltonics) fitted with an ESI

source and internal calibration was achieved with 10 mL of 0.1 M sodium formate solution

injected through a six port valve prior to each chromatographic run. Calibration was carried

out using the enhanced quadratic calibration mode. It should be noted that in TOF calibration

the intensities of the measured peaks have a significant influence on the magnitude of the

mass error with high intensity peaks resulting in detector saturation displaying larger mass

errors. Where necessary this was avoided by using a more dilute sample. All MS

measurements were carried out in the negative ion mode.

3.1.5 HPLC analysis

The extracted thearubigins were analysed by HPLC using an Agilent 1200 HPLC pump with

a 5 μl loop, coupled to an Agilent 1100 autosampler, an Agilent 1100 DAD-UV-VIS

detector. Black tea SII thearubigin fractions were reconstituted in 0.5 mg/ml concentration in

water and filtered through a 0.45 μm HPLC filter prior to injection of a volume of 3 μl.

HPLC analysis used a POLARIS 5-C18-A column (length 250 mm, diameter 3 mm, particle

size 5 μm) with a step gradient elution employing acetonitrile (MeCN) and water containing

0.005 % formic acid, as follows: 8% MeCN from 0 to 50 minutes, then changing to 31 %

MeCN for 10 minutes then changing to 25 % MeCN for a further 5 minutes.

82


3.1.6 Direct Infusion Tandem MS

Samples were prepared by dissolving 5 mg from the crude isolated thearubigin SII fractions from

Assam, Ceylon, Assam Ceylon, South Indian tea in 10 ml methanol/water (1:1), then infused in

the ion trap mass spectrometer (Bruker HCT ultra) in the negative ion mode at a flow rate at 180

µl /min. MSn experiments were carried manually with an isolation width 1Da for targeted masses

in the mass range between 500-1500 Da.

3.1.7 Direct Infusion TOF

The thearubigin solutions were injected in ESI-TOF mass spectrometer in the negative ion mode

at a flow rate 180 µl / min.

3.1.8 Data analysis

Data were analysed using Bruker Data Analysis 4.0 software. Micro TOF data were analysed

after external enhanced quadratic calibration. Ion trap data were analysed by isolating and

trapping the parent target ion for a longer time, in which fragmentation was performed. The

pseudo-molecular ions having a particular fragmentation could be located., and further

analysis were performed in terms of TIC and EIC chromatograms using the implemented

software routines.

83


3.1.9 MALDI-TOF MS

MALDI-TOF spectra were acquired on an Autoflex ǁ MALDI-TOF-TOF mass spectrometer

(Bruker Daltonics) equipped with nitrogen laser (wavelength 337nm). The instrument was

operated in the reflector mode with ion Source 19.00 kv, lens 8.00 kv, and reflector 20.00 kv,

using an optimized extraction delay time of 50ns.The laser frequency was set at 25 Hz with 50

laser shots per acquisition. The laser-power was optimized to obtain a strong analyte signal, in

the range of 38 to 40%. Matrix suppression mode was set to deflection up to 500 amu.

Thearubigins were first co crystallized with the matrix, then exposed to the nitrogen laser inside

the ion source of the mass spectrometer. The crystals were desorbed and ionized, and then

accelerated by an electro-static field in the ion source with different velocities into the detector.

Spectra were acquired in the mass range of 500-2100 Da with a detector gain of 2.60x, an

optimized electronic gun of 100 mv, and were recorded with a detector voltage of 1400mv and

sample rate of 2 Gsignal/s.

The instrument was calibrated with different point external calibration using peptide calibration

standards; Angiotensin-ǁ (m/z 1046.5418 Da), Angiotensin-ǀ (m/z 1296.6847 Da), Bradykinin

fragment 1-7 (m/z 757.3992 Da), Renin substrate (m/z 1758.93 Da), and ACTH-(1-17) (m/z

2093.0861 Da).

3.1.10 Matrix Preparation

10 mg of 2, 5 dihydroxyacetophenone were dissolved in 1 ml (methanol: water) (7:3), vortex and

mixed for 1 min. and then centrifuged in order to use the clear liquid for the preparation.

84


3.1.11 MALDI Target

The matrix was applied on MTP Anchor Chip TM, target equipped with hydrophilic patches

(anchors) in hydrophobic surroundings. Anchor diameters used between 200-800µm.

3.1.12 Sample preparation for MALDI

1mg of the TRs was dissolved in 500 µl (methanol/ water).Sample mixed with the matrix

solution in a ratio (1:10).0.5-1µl of this mixture was mounted on an Anchor chip 800 µm.

3.1.13 Amsterdam Density functional Software (ADF)

ADF is quantum chemistry software based on density functional theory, which implies a picture

of many electron systems and yields in principle the exact electron density and the total energy

of the system. The exact energy correlation is unknown, but this software provides a large basis

set where the exchange correlation energy could be computed providing precise information

about the different structures. (www.scm.com)

3.2 Results and Discussions A selection of higher molecular weight ions in the range from m/z 1000 to 1500 from the four

commercial teas (Assam, Ceylon, South Indian, Chinese) were clearly identified. These ions,

present in the SII fractions, were investigated at m/z 1143, 1155, 1171, 1187, 1339, 1441, 1445

and 1461 in negative ion mode. Accurate mass values using ESI-FT-ICR measurements were

available through previous studies by Kuhnert et al.39

85

http://www.scm.com/


For these ions extracted ion chromatograms were generated from ESI-LC-MS runs (figure 3.1)

and targeted ESI-MSn experiments were carried out where their intensities were sufficient

(figure 3.2). Secondly, these ions were used as precursor ions in tandem MS direct infusion

experiments using an isolation width of 1 Da. The collision energies applied in the direct

infusion tandem MS experiments were varied yielding different fragmentation spectra. Further

MS3 spectra were acquired when possible. Thirdly under optimized matrix preparation, MALDI-

TOF MS and MALDI-TOF-TOF tandem MS experiments were carried out using the LIFT

method for the aforementioned precursor ions in the negative ion mode. The complete data set

are exhibited in table 3.1 and thus used for further structure assignments according to their

identified main fragment ions and their neutral losses. Accordingly, we assume that flavan-3-ols

are substrates for tea polyphenol oxidase and that oxidation occurs mainly at the B-ring of the

flavan-3-ol where these highly electrophilic substrates may react with any other polyphenol

acting as a nucleophile. According to the oxidative cascade hypothesis it was assumed that ions

in the mass range from 1000-1500 Da originate from three to four units of flavan-3-ols, where

these are formed by the reactions of a flavan-3-ol quinone with a dimeric or trimeric flavan-3-ol

respectively. The regiochemistry of C-C bond formation is always speculative and this stage not

amenable to MS characterization.

86


Figure 3.1 Extracted Ion Chromatograms: EIC of the trimeric compounds in the negative ion mode a) at m/z 1155:

C58H43O26, b) at m/z 1171 (with one oxygen inserted) C58H43O27, c) at m/z 1185 (the quinone counterpart with two

oxygen insertion) C58H41O28; The tetrameric compounds d) at m/z 1445 C72H52O33, e) at m/z 1461 (with one oxygen

insertion) C72H52O34

ASSFRAGM_051202.D: EIC 1155.0 -All MS





0.00 0.25 0.50 0.75 1.00 1.25 x10 Intens.

0 1

2 3 4 4 Intens.

0

2

4

4 x10 Intens.

0.0

0.5 1.0 1.5 2.0

4 x10 Intens.

0 1 2 3 4 4 x10 Intens.

0 10 20 30 40 50 60 70 Time [min]

a

b

c

d

e

x10

5

87


Figure 3.2 Tandem LC-MSn experiments for one of the EIC in the negative mode; for example: the m/z 1171 at

different retention times (Chromatograph shown in figure 3.1 b) revealing the structural variety of this compound

with different fragment MS2 spectra obtained

3.2.1 Identified trimeric compounds

Ion at m/z 1143

A pseudo molecular ion at [M-H]- 1143 Da was identified in all the tea samples and extracted

ion chromatograms (EICs) revealed the presence of four to six chromatographic peaks of

medium intensity of this mass indicating 4 to 6 isomeric structures. A direct infusion ESI-tandem

MS experiment of the m/z 1143 Da signal showed a fragment ion in MS2 at m/z 914.4 Da. An

MS3 experiment with a precursor ion at 914.4 yielded a fragment spectrum identical to that of

274.5 318.6

352.6 412.6

440.6 500.6

867.0 971.0

1049.0 1171.4

-MS, 51.2min

1017.0 1152.3

-MS2(1171.4), 51.2min

274.5 318.6 352.6 412.6 440.7 544.7 745.0 897.0 1171.1

-MS, 36.5min

712.8 973.0 1019.0

1073.9

1151.8 -MS2(1171.1), 36.5min

274.5 318.6 352.6 412.6 440.7 500.6 1155.1 1171.1

-MS, 28.7min

881.0 1001.0 1151.1

-MS2(1171.1), 28.7min

274.5 318.6 352.6 412.6

544.6 911.0 1171.3

-MS, 44.6min

936.9 1001.0 1086.4

1134.9 -MS2(1171.3), 44.7min

0 50

100 Intens. [%]

0 50

100 150

[%]

0 50

100 150 [%]

0 50

100 150

[%]

0 50

100 150

[%]

0 50

100 150 [%]

0 50

100 150 [%]

0 50

100 150 [%]

400 600 800 1000 1200 1400 m/z

88


theasinensin A C44H33O22. The observed neutral mass loss of 228 Da could not be further

rationalized, however the data clearly show the presence of a theasinensin A coupled to an

unknown nucleophile. By increasing the collision energy, almost no signals except for the parent

ion and a small signal at m/z 955 giving a loss of 188 also of unknown origin pointing that in this

case a different compound was observed. At still higher energy again a complete different pattern

was observed with the signal at m/z 715 indicating to a theaflavin gallate structural element. So

indeed the presence of at least three different compounds of m/z 1143 could be observed, in

which one had a theasinensin (12) type structure and another had a theaflavin gallates (9) type

structure as the main backbone. Tentative structures are exhibited in figure 3.3 showing the

different structures identified at different collision energies.

Figure 3.3 Direct infusion tandem mass experiment in the negative ion mode for [M-H]- 1143 Da with its different

fragment ion at m/z 914.4 Da (the isotope of theasinensin A), at m/z 955.1, and at m/z 715 exhibited at different

collision energies ranging from fragmentation amplitude of a) 0.2 V, b) 0.25V and c) 1V

495.3

914.4 1143.1

955.1

1143.1 -MS2(i1143.4), 6.9min

715.0 -MS2(1143.4), 10.4min

0

20

40

60

80

100 [%]

0

20

40

60

80

100 [%]

0

20

40

60

80

100 [%]

300 400 500 600 700 800 900 1000 1100 1200 m/z

Intens a

b

c

O

O

OO

HO

OH

HO

HOOH

OH

OH

OOH

OHOH

OH

O

O

O

HO

OH

OH

OH

HO

HO

O

HOOH

OH

OH

OHOH

OH

OO

OH

OH

OH

(12)

(9)

89


Ion at m/z 1155

A pseudo molecular ion at m/z [M-H]- 1155 was identified in all samples and extracted ion

chromatograms (EICs) revealed the presence of six chromatographic peaks of high to medium

intensities of this mass indicating the presence of at least six isomeric structures of this

compound. A direct infusion ESI experiment showed spectra with most intense ions identified at

m/z 913.1, and targeting fragmentation experiment with this precursor ion confirmed it to be the

theasinensin A C44H33O22 (12) structure as its fragment spectrum had characteristic fragment ions

at m/z 761.2 and 609.1. The direct infusion ESI experiment, also exhibited another intense peak

at m/z 867.1, corresponding to theaflavin digallate C43H32O20 (14), and this was confirmed by

applying further fragmentation, which gave its characteristic fragment ions at m/z 715.2 and

563.1 confirming this structure. Furthermore, fragmentation at higher collision energy of 1.0 V

of this compound gave again a fragment ion at m/z 867 of theaflavin digallate (14) corresponding

to a neutral loss of 288 Da the loss of the oxidized (epi)catechin (C15H14O6-2H), and also

exhibited other fragment ions with intense peaks observed at m/z 1003.1 and m/z 985.1 revealing

the loss of the galloyl moiety from this compound. On the other hand, by varying the collision

energy for fragmentation, another fragmentation of the precursor ion at m/z [M-H]- 1155 was

obtained and gave a fragment ion at m/z 1016.7,where according to its observed neutral loss of

138.3 Da, this fragment ion is rationalized in terms of a retro Diels-Alder fragmentation (RDA)

may occur at a flavan-3-ol moiety on one of its benzopyran part (C7H6O3) and these are

illustrated in figure 3.4.

MALDI-MS2 data confirmed the presence of this precursor ion at m/z 1155 having its

characteristic fragment ions at m/z 1018 and m/z 867. An additional fragment ion at m/z 986 is

90


observed corresponding to a neutral loss of 169 Da, which is consistent with loss of a galloyl

moiety (-C7H5O5) as shown in figure 3.5. So the signal at m/z 1155 could represent different six

isomeric structures having theasinensins and theaflavins as a backbone, where these structures

were resolved in the collision energy dimension. Both the ESI tandem mass and MALDI TOF-

MS data confirmed the presence of the isomeric structure, that is formed by oxidative coupling

between theaflavin di-gallate (14) and (epi)catechin (2).

609.0

715.0

745.0 761.1

813.0

897.0

913.1

1019.1 1155.1

713.0 815.0

833.0

867.0

985.1

1003.1

1137.2

7

0

20

40

60

80

100

0

20

40

60

80

100 [%]

300 400 500 600 700 800 900 1000 1100

-MS2(1155.6), 1.5min

-MS, 0.1min 867.1 [%] Intens

m/z

Loss of [C15H14O6 -2H]

Loss of C7H6O5

Loss of H2O

Loss of C7H5O4

a

b

O

O

OO

HO

OH

OH

OH

O

O OH

OHOH

OH

O

OHOH

HO

HO

HO

O

O

OO

HO

OH

OH

OH

O

O OH

OHOH

OH

O

OH

OH

HO

HO

HO

O

O

O

HO

OH

OH

OH

HO

HO

O

HOOH

OH

OH

OHOH

OH

OO

OH

OH

OH(14) (12)

91


Figure 3.4 Direct infusion experiment in the negative ion mode exhibiting the presence of the parent ion for [M-H]-

1155 and its MS2 spectra for the different fragment ions obtained at different collision energies; (b) At energy of 1.00

V giving fragment ion at m/z 867 Da of theaflavin digallate C43H32O20 revealing the loss of epicatechin, at m/z 985.1

corresponding to the loss of C7H6O5 and at m/z 1003.1 the loss of C7H5O4 revealing the loss of galloyl moiety; while

(c ) at a different collision energy of 0.3 V tandem mass experiments gave a fragment ion at m/z 1016.7 Da

indicating retro-Diels Alder fragmentation

1016.7

1154.9

-MS2(1155.0), 28.9min

0

20

40

60

80

100 [%]

300 400 500 600 700 800 900 1000 1100 1200 m/z

Intens.

Loss of C7H6O3

c

92


Figure3.5 MALDI-TOF Fragmentation of the pseudo-molecular ion in the negative mode m/z 1156.748 using the

LIFT method, this compound giving main fragment ions at m/z 986.588, and at m/z 868.471 (the regiochemistry for

the tentative structure exhibited is chosen randomly)

1156.746

580.046

986.588

868.466 716.236

437.877 738.283

287.655

0

2000

4000

6000

8000

Intens. [a.u.]

200 400 600 800 1000 1200 m/z

O

O

OO

OH

OH

OH

OH

OH

OH

O

O OH

OHOH

O

OH

OH

OH

OH

O

OH

OH

HO

OH

OH

93


Ion at m/z 1171

A pseudo molecular ion at m/z [M-H]- 1171 was identified in all samples and extracted ion

chromatograms (EICs) revealed the presence of seven chromatographic peaks of medium

intensity of this compound. A direct infusion ESI-tandem MS experiment showed a fragment ion

at m/z 867.9 as an intense peak. An MS3 experiment with this precursor ion yielded fragment

ions at m/z 715.2 and 563.1, corresponding to theaflavin digallate C43H32O20 (14). The observed

neutral loss between this new compound and the fragment ion obtained, was found to be 304 Da

and this corresponds to the oxidized (epi)gallocatechin (C15H14O7 -2H). By increasing the

collision energy in the tandem mass experiment, the fragmentation gave the isotope of the

molecular ion at m/z 1172.5 giving a main fragment ion at m/z 914.0. MS3 experiment of this

precursor ion yielded a fragment spectrum identical to that of theasinensin A C44H33O22 (12) with

its characteristic fragment ions. At another still higher energy, the fragmentation spectra revealed

the presence of fragment ions at m/z 1001.1 corresponding to the neutral loss of 169.9 which

indicates the loss of the galloyl moiety (-C7H6O5) from this compound, and a fragment ion at m/z

881.1 (C43H30O21) corresponding to the quinone of mono-hydroxylated theaflavin di-gallate

showing a neutral loss of (epi)catechin moiety (-C15H14O6).

MALDI-MS2 data of the 1171 molecular ion confirmed the presence of fragment ions at m/z

1001 and 867. An additional fragment ion at m/z 1155 is also observed corresponding to a neutral

loss of 16 Da consistent with loss of a single oxygen atom. These data clearly indicate the

presence of trimers of flavan-3-ols formed by oxidative coupling between theaflavin di-gallate

(14) and (epi)gallocatechin (4) or by oxidative coupling between theasinensin A (12) and an

unknown building block, and these data are exhibited in figures 3.6 & 3.7 and further illustrated

in table 3.1

94


Figure 3.6 Direct infusion experiment for the parent ion [M-H]- 1171Da in the negative ion mode and their MS2

spectra exhibiting different fragment spectra at different collision energies; (b) at 0.25 V giving the fragment ion at

m/z 867.9 Da of C43H32O20 with a loss of (epi)gallocatechin, (c) at 0.3 V giving the 1172.5 having the fragment ion

at m/z 914.0 of C44H33O22 (d) at 1.0 V giving the fragment ion of mono-quinone of the hydroxylated derivative of

theaflavin di-gallate C43H30O21 (regioisomerism chosen randomly)corresponding to the loss of (epi)catechin moiety

and at m/z 1001.1Da revealing the loss of galloyl moiety from this compound

558.9 578.9

729.0

744.9 829.0

881.0

957.0

983.1

1001.0

1017.0

1151.1

0

20

40

60

80

100

120

300 400 500 600 700 800 900 1000 1100 1200 m/z

-MS2(1171.0), 1.8min Loss of C7H6O5

Loss of C15H14O6

d

O

O

OO

HO

OH

HO

HOO

OH

O

OOH

OHOH

OH

O

OH

OHHO

O

396.6

440.9

456.9 609.0 715.0

761.0

913.0

965.1 1018.1 1171.0

867.9

1171.1

633.5

914.0

1172.5

0

20

40 60

80 100 [%]

0

20 40

60 80

100 [%]

0

20

40

60

80

100 [%]

400 600 800 1000 1200 1400 m/z

867.0

-MS2(i1171.0), 5.8min

-MS2(i1171.0), 5.9min

-MS, 24.3min

Loss of [C15H14O7 -2H]

a

b

c

O

O

OO

HO

OH

OH

OH

O

O OH

OHOH

OH

O

OHOH

HO

HO

HO

O

O

O

HO

OH

OH

OH

HO

HO

O

HOOH

OH

OH

OHOH

OH

OO

OH

OH

OH

(14)

(12)

95


Figure 3.7 MALDI-TOF fragmentation of the pseudo-molecular ion at m/z 1171 using the LIFT method in the

negative ion mode giving fragment ions at m/z 1001.751, 867.799, 715.655 ( the regiochemistry of this structure is

chosen randomly, for further discussion of the regiochemistry, see section d in this chapter).

1171.727

437.342 579.411

1001.751

715.512

1152.858

867.610

0

2000

4000

6000

Intens. [a.u.]

200 400 600 800 1000 1200 m/z

O

O

OO

OH

OH

OH

OH

OH

OH

O

OOH

OHOH

O

OH

OH

OH

OH

O

OH

OH

OH

OH

HO

OH

96


Ions at m/z 1187 and 1185

A pseudo molecular ion at m/z [M-H]- 1187 was identified, although EICs did not show the

presence of clear chromatographic peaks. A direct infusion ESI-tandem MS experiment showed a

fragment ion in MS2 at m/z 867.7 corresponding to the ion of theaflavin digallate C43H32O20 (14)

(figure 3.8). The observed neutral loss of 320 Da corresponds to a singly oxygenated (epi)

gallocatechin (C15H14O8). MALDI-MS2 data confirmed the presence of fragment ions at m/z 867.

In MALDI-MS an ion at m/z 1185 was observed with higher intensity if compared to the ion at

m/z 1187 yielding fragment ions at m/z 1169.8, 1018.2 and 867.6 the latter corresponding again

to theaflavin-di-gallate (14) resulting from a neutral loss of 318 Da (C15H12O8), indicating

formation of a quinone structure of this compound as expected within the oxidative cascade

framework. Fragment spectra and tentative structures are shown in figure 3.9. These data clearly

indicate the presence of a trimer of flavan-3-ols formed by oxidative coupling between theaflavin

di-gallate (14) and an oxygenated (epi)gallocatechin (C15H14O8).

97


Figure 3.8 Direct infusion experiment for the parent ion for [M-H]- 1187Da in the negative mode giving (a)

theaflavin digallate at m/z 867 and theasinensin A at m/z 913 as main peak ions, and (b) the tandem mass

experiments by setting collision energy of 0.15 V giving MS2 spectrum with most intense fragment ion at m/z 867.7

C43H32O20 corresponding to the loss of mono-oxygenated (epi)gallocatechin

867.7

1187.2

-MS2(1187.0), 17.3min

0

20

40

60

80

100

Intens. [%]

400 600 800 1000 1200 1400 m/z

Loss of [C15H14O7 +O -2H]

b

O

O

OO

HO

OH

HO

HO

OH

OH

O

OOH

OHOH

OH

O

OH

OHHO

(14)

440.9

456.9

483.0 519.5

609.0

635.0 665.0

685.7

715.0

732.0

761.0

813.1

867.1

885.1 897.1

913.0

927.1 965.0 1003.1 1051.0

1095.1

1155.2 1187.7

-MS, 0.1min

0

20

40

60

80

100

Intens. [%

]

300 400 500 600 700 800 900 1000 1100 1200 m/z

a

98


Figure 3.9 MALDI-TOF fragmentation of m/z 1185 component using the lift method in the negative ion mode

giving different fragment ions at m/z 1169.860, m/z 1018.152, m/z 867.56 (the regioisomerism of this structure is

selected randomly)

1185.982

1018.152

579.654 455.678

1169.860

715.925 867.561

0

500

1000

1500

Intens. [a.u.]

200 400 600 800 1000 1200 1400 m/z

O

O

OO

OH

OH

OH

OH

OH

OH

O

OOH

OHOH

O

OH

OH

OH

OH

O

OH

O

O

OH

HO

OH

OH

99


Ion at m/z 1339

A pseudo molecular ion at m/z [M-H]- 1339 was identified in three samples with the EICs not

showing clear chromatographic peaks. A direct infusion ESI-tandem MS experiment showed a

fragment ion in MS2 at m/z 1323.1 revealing the loss of oxygen and at m/z 1021.4, where the

latter can be tentatively assigned as a neutral loss of methylated (epi)gallocatechin (C16H17O8).

Figure 3.10 MS2 spectra for the parent ion for [M-H]- 1339 Da and its fragment ions at m/z1323.1 revealing the loss

of oxygen and at m/z 1021.4 Da the loss of methylated (epi)gallocatechin

1237.5

1323.1

1021.4 0

20

40

60

80

100

0

20

40

60

80

100 [%]

400 600 800 1000 1200 1400 m/z

-MS2(1339.0),17.6min

Intens [%]

-MS2(i1339.0), 15.9min

Loss of oxygen

Loss of (318Da) methylated epigallocatechin

100


3.2.2 Identified tetrameric compounds

m/z 1441, m/z 1445, and m/z 1461

A pseudo molecular ion at m/z [M-H]- 1441 was identified in the thearubigins fractions of the

four black tea samples The direct infusion ESI-tandem MS experiment showed a fragment ion in

MS2 at m/z 898.1, which can be tentatively assigned as a di-oxygenated theaflavin-digallate,

which presence has been previously confirmed in the SII black tea thearubigin fraction. Another

pseudo molecular ion at m/z [M-H]- 1445 was also identified in the four samples, where the EICs

showed the presence of five clear chromatographic peaks. A direct infusion ESI-tandem MS

experiment of this compound showed a main fragment ion in MS2 at m/z 879.5; this fragment ion

was previously identified as the di-quinone of mono-hydroxylated theaflavin digallate

(C43H28O21), revealing a neutral loss of 564.6 Da (C29H24O12), where this may indicate an

oxidative coupling could have occurred between the quinone of mono-oxygenated theaflavin

digallate and theaflavin (7) resulting in the tetrameric structure with m/z 1445. However, no

further MS3 data could be obtained. Although five signals were observed in the EIC no clear

additional fragments could be demonstrated in the direct infusion ESI-tandem MS experiment.

101


Figure 3.11: Direct infusion tandem mass experiment for the parent ion for [M-H]- 1445Da giving the quinone of

the mono-hydroxylated theaflavin digallate as its main fragment ion at m/z 879.5 Da

Furthermore, EICs showed the presence of one clear chromatographic peak for a new pseudo-

molecular ion at m/z 1461, where direct infusion ESI MS experiment for the latter compound

showed an intense peak ion at m/z 865.2. This ion can be tentatively assigned as a quinone

derived from theaflavin-digallate (C43H30O20), previously reported in the thearubigin fraction.

This ion arises from a neutral loss of 595.6 Da corresponding to a di-oxygenated theaflavin

derivative (C29H24O14), and this indicates that this compound could be formed from oxidative

coupling between theaflavin digallate and the di-hydroxylated theaflavin derivative again

suggesting a tetrameric flavan-3-ol structure now at m/z 1461. No further MS2 and MS3 data

could be obtained for this compound.

461.9

542.2

764.7

879.5

1444.1

-MS2(1445.0), 6.5min

0

20

40

60

80

100 [%]

400 600 800 1000 1200 1400 m/z

Intens.

[Theaflavin digallatate+O1-4H]: (C43H28O21)]

Loss of (C29H24O12)

102


Table 3.1 ESI-Tandem MSn data for the new trimeric and tetrameric components in thearubigin fractions

exhibiting their different characteristic fragmentation spectra by increasing the collision energy induced,

their molecular formulas confirmed by ESI-TOF, their main fragment ions MS2, their mass losses, their

attained MS3 fragment ions when possible; MALDI-TOF-TOF data, their main fragment ions using the

LIFT method, and their mass losses (energy values are given within the legends of the figures)

Mol.ion m/z

Molecular formula

Main Fragment ions MS2

Molecular Formula of

Fragment ions Mass loss

Main Fragment ions MS3

MALDI LIFT method main

Fragment ions

Mass losses in MALDI-

Fragmentation

1143

Energy-1 914.4 C44H33O22 228.6 894.1, 761.1, 591.1, 457.1

Energy-2 955.1 187.9 Energy-3 715 C36H28O16 428 563.1

1155 C58H43O26

868.471, 1018.681, 986.588 288,138, 170

Energy-1 867 C43H32O20 288 715.1, 563.1

1137.2 18

1003.1 152

985.1 170 925.4, 833.0, 797.1 Energy-2 1016.7 138.3

1171 C58H43O27

1155.385, 867.79, 1001.751 16, 304, 169

Energy-1 867.9 C43H32O20 304 715.1, 563.1 Energy-2 914 C44H33O22 258 Energy-3 881 C43H30O21 290 713.08, 563.01,

441.07

1001 C51H37O22 170

1187 C58H43O28

1169.860, 1018.152, 867.561 16, 168, 318

Energy-1 867.7 C43H32O20 320

1339

Energy-1 1323.1 16 Energy-2 1021.4 318

1441

103


Energy-1 898.1 C43H31O22 543

1444

Energy-1 C72H53O33 879.5 C43H28O21 564.6

1461 C72H53O34 865.4 C43H30O20 595.6

104


3.2.3 Homologous series for the new trimeric compounds

As the extracted ion chromatograms EIC confirmed the m/z parent ions in LC tandem MS

experiments, it was worthy to perform further identification for the presence of their

hydroxylated derivatives if they exist for investigating the second level of the oxidative cascade

hypothesis. EIC were generated for their predicted hydroxylated derivatives, and results

confirmed the presence of the homologous series of these compounds revealing the presence of

different stereo-and regioisomer at different retention times. In figure 3.12 is the homologous

series for trimeric compound at m/z 1143 C57H43O26, showing oxygen insertions at m/z 1159

C57H43O27, at m/z 1175 C57H43O28, at m/z 1207 C57H43O30, at m/z 1239 C57H43O32. Figure 3.13

also exhibits the homologous series of the trimeric structures at m/z 1155 of C58H43O26

undergoing oxygen insertions at m/z 1171 C58H43O27, at m/z 1187 C58H43O28, at m/z 1219

C58H43O30, at m/z 1235 C58H43O31, at m/z 1251 C58H43O32 In many cases we obtained a stepwise

insertion of oxygen. Several regioisomers were detected at different retention times and these are

exhibited in both figures. Tentative structures for the homologous series for the m/z 1155 are

shown in scheme 3.1.

105


Figure 3.12 EIC for the parent ion [M-H]- 1143 C57H43O26 and its hydroxylated series at m/z 1159 C57H43O27, at m/z

1175 C57H43O28, at m/z 1207 C57H43O30, at m/z 1239 C57H43O32 in the negative ion mode

Figure 3.13 .EIC of the parent ions in the homologous series in the negative mode for m/z 1155 of C58H43O26

undergoing oxygen insertions: at m/z 1171C58H43O27, at m/z 1187 C58H43O28, at m/z 1219 C58H43O30, at m/z 1235

C58H43O31, at m/z 1251 C58H43O32

EIC 1171NEW.D: EIC 1155.0 -All MS





0

1

4 x10

0

1000

2000

Intens.

0

500

1000

Intens.

0

1000

2000 Intens.

0

500

1000 Intens.

0 1000 2000 3000

Intens.

0 10 20 30 40 50 60 70 Time [min]


C58H43O26

C58H43O27

C58H43O28

C58H43O30

C58H43O31

C58H43O32

EIC 1143.D: EIC 1143.0 -All MS





0 1000

2000 3000

Intens.

0 500

1000 1500 2000

Intens.

0

2000

4000

Intens.

0 500

1000 1500 2000

Intens.

0 2000

4000

6000 Intens.

0 10 20 30 40 50 60 70 Time [min]

C57H43O26

C57H43O27

C57H43O28

C57H43O30

C57H43O32

Intens.

106


O

O

OO

OH

OH

OH

OH

OH

OH

O

O OH

OHOH

O

OH

OH

OH

OH

O

OH

OHHO

OH

OH

O

O

OO

OH

OH

OH

OH

OH

OH

O

O OH

OHOH

O

OH

OH

OH

OH

O

OH

OHHO

OH

OH

OH

O

O

OO

OH

OH

OH

OH

OH

OH

O

O OH

OHOH

O

OH

OH

OH

OH

O

OH

OHHO

OH

OH

O

O

OO

OH

OH

OH

OH

OH

OH

O

O OH

OHOH

O

OH

OH

OH

OH

O

OH

OHHO

OH

OH

HO

OHHO

O

O

OO

OH

OH

OH

OH

OH

OH

O

O OH

OHOH

O

OH

OH

OH

OH

O

OH

OHHO

OH

OH

HO

OHHO

OH

OH

O

O

OO

OH

OH

OH

OH

OH

OH

O

O OH

OHOH

O

OH

OH

OH

OH

O

OH

OHHO

OH

OH

HO

OHHO

OH

OH

OH

O O

2O

OO

C58H43O26 m/z 1155

C58H43O27 m/z 1171

C58H43O28 m/z 1187

C58H43O30m/z 1219

C58H43O31m/z 1235

C58H43O32m/z 1251

OH OH

OH

Scheme 3.1 Homologous series .of the m/z 1155 with successive oxygen insertion (regioisomerism selected

randomly)

107


3.2.4 Theoretical calculations and Computational Analysis

The experimental work based on both direct infusion and LC tandem mass spectrometry in the

analysis and characterization of TRs fractions, showed the presence of multiple isomers of the

components within a series, however reliable information of the regiochemistry of these

structures is unclear due to the absence of the authentic standards. In order to gain a rational

insight in the oxidative coupling reactions and the most likely regiochemistry of the products,

computational approach based on the density functional theory DFT (as defined in chapter 4)

offers a promising knowledge at which part of the substrates the oxidative coupling is more

likely taking place.

In both direct infusion MSn mass spectrometry and MALDI TOF experiments theaflavin digallate

was the main fragment ion obtained, where the experimental data revealed that the higher

molecular TRs components specifically the trimeric structures are mainly formed by oxidative

coupling between theaflavin digallate and TPPO oxidized catechins.

In order to further support the oxidative cascade hypothesis occurring for the trimeric structure

formation obtained, and in order to understand the mechanism behind the formation of these

structures, I performed computational analysis work based on the DFT, thus giving an insight at

which part of the substrates the oxidative coupling is mainly occurring. For this study, the

dimeric theaflavin digallate structure (14) acting as the main fragment ion in the targeting

fragmentation MS2 experiments and a monomeric flavan-3-ol epigallocatechin (4) activated by

TPPO to the orthoquinone was selected in the formation of the higher molecular weight

thearubigin trimeric components obtained such as C58H43O27.

108


The theoretical calculations were based on high exchange correlation energy functional Hybrid/

B3LYP, in order to attain maximum precision and accuracy. The results obtained, revealed the

electronic distribution showing the highest occupied molecular orbital (HOMO) and the lowest

unoccupied molecular orbital (LUMO) in the aforementioned substrates, thus reflecting the most

likely nucleophilic or electrophilic attacking sites. Furthermore, performing such theoretical

calculations, I was able to investigate the energy of the orbital coefficient EHOMO of these

substrates, which reveals the tendency of their electron donation.

The electronic distribution for the dimeric theaflavin digallate, acting as the main fragment ion in

the experiments, was observed. This study showed that the highest occupied molecular orbital

(HOMO) are distributed on the A-ring and the B-ring of the benzotropolone part of the dimeric

theaflavin digallate, while the lowest unoccupied molecular orbital (LUMO) lie on galloyl part of

this compound and at some sites in the benzotropolone ring. The TPPO oxidized

epigallocatechin was also observed and it was found to contain mostly the LUMO level on the

orthoquinone B-ring (oxidized part of the molecule), while the HOMO levels lies almost on the

A-ring and these are exhibited in figure 3.14.

Furthermore, theaflavin digallate acting as a nucleophile showed to possess a higher energy

EHOMO orbital coefficient than the oxidized epigallocatechin, revealing its tendency for donating

electrons to the oxidized part of epigallocatechin and this is shown in table 3.2. The data

obtained enabled to rationalize the routes of the reaction thus identifying the sites of the

oxidative coupling occurring for the formation of the new trimeric thearubigins in the

oligomerisation process. I was able to suggest that the coupling mainly occurs from the B-ring of

the benzotropolone of the dimeric theaflavin digallate acting as a nucleophile to the B-ring of

TPPO oxidized epigallocatechin, and also may occur from A-ring of theaflavin digallate dimer to

109


the B-ring of the oxidized epigallocatechin. Hence, I was able to suggest two alternative reaction

pathways for the formation of the higher molecular weight thearubigin components and these are

exhibited in figure 3.15 and figure 3.16 undergoing firstly intermolecular Aldol reaction then

followed by equilibrium to the more stable tautomer. Hence, these calculations support indeed

potential coupling routes as depicted and giving trimers connected through benzotropolone

substructures.

110


Figure 3.14 The electronic distribution for a) theaflavin digallate C43H32O20 as the main fragment ion and b) the

oxidized EGC (C15H12O7) (the dark blue and red orbitals present the HOMO levels where as the light blue and

brown are the LUMO levels)

a)

b)

111


Table 3.2 The energies of the HOMO levels within the theaflavin digallate dimer and the oxidized

substrate epigallocatechin, revealing the tendency of theaflavin digallate for donating electrons to the

highly electrophilic quinone acting itself as the nucleophile

EHOMO(e.V.)

Theaflavin digallate dimeric structure -5.727

Epigallocatechin oxidation state (quinone) -7.228

Theflavin digallate

Epigallocatechin quinone

HOMO

LUMO

HOMO

LUMO

Energy

O

O

OO

HO

OH

HO

HOOH

OH

O

O OH

OHOH

OHO

OH

OHHO

O

HO

OO

OH

OHHO

Scheme 3.2 Energy diagram for the orbital coefficient of theaflavin digallate and the epigallocatechin quinone

112


O

O

OO

O

OH

OH

OH

OH

OH

O

OOH

OH

OH

O

OH

OH

OH

OH

O

OH

O

O

OH

H+

O

O

OO

O

OH

OH

OH

OH

OH

O

OOH

OH

OH

O

OH

OH

OH

OH

O

O

OO

OH

OH

OH

OH

OH

OH

O

OOH

OH

OH

O

OH

OH

OH

OH

O

OH

OH

O

OHH

H+

H

O

OH

OH

OH

OH

H

H+

m/z 1171

Oxidized epigallocatechinm/z 304

Theaflavindigallate,m/z [M-H] 867

A

BHO

OH

HO

OH

HO

OH

Figure 3.15 Proposed mechanism reaction occurring from ring A of theaflavin digallate to ring B- of

epigallocatechin quinone for the proposed novel structure (1) in higher molecular weight components obtained from

the oxidative coupling between epigallocatechin and theaflavin digallates

113


O

O

OO

OH

OH

OH

OH

O

OH

O

OOH

OH

OH

O

OH

OH

OH

OH

Theaflavin digallate,m/z [M-H] 867

O

OH

O

O

OH

H+

O

O

OO

OH

OH

OH

OH

O

OH

O

OOH

OH

OH

O

OH

OH

OH

OH

O

OH

OH

O

OH

H

H

H+

O

O

OO

OH

OH

OH

OH

OH

OH

O

OOH

OH

OH

O

OH

OH

OH

OH

O

OH

OH

OH

OH

H

H+

Oxidized epigallocatechinm/z 304

m/z1171

B

BHO

OH

HO

OH

OH

HO

Figure 3.16 Proposed mechanism occurring from ring B of theaflavin digallate to ring B- of epigallocatechin

quinone monomer for the novel structure (2) in higher molecular weight components obtained from the oxidative

coupling between epigallocatechin and theaflavin digallates

114


3.3 Conclusions

I identified a series of novel constituents of the SII fraction of black tea thearubigins with

molecular weights in the range between 1000 and 1500 Da. A series of complementary mass

spectrometric techniques including high resolution ESI-tandem mass spectrometry, LC-MS and

MALDI-TOF tandem mass spectrometry clearly revealed, that the compounds observed are

trimers and tetramers of flavan-3-ols. Clear evidence was provided for the presence of

oligomerisation of flavan-3-ols taking place via oxidative coupling pathway. By performing

theoretical calculations based on the density functional theory, I was able to propose the reaction

pathway and suggest the likely regiochemistry of the new trimeric structures identified.

Furthermore, the presence of a series of hydroxylated derivatives of these compounds was

clearly observed.

In this work, I confirmed for the first time experimentally the oxidative cascade hypothesis

which suggested oligomerisation of catechins (flavan-3-ols) mediated by oxidative coupling

reactions. This study comprises a key step forward for unraveling the mystery of black tea

thearubigin structure.

115

Potentiostatic measurement, a New Approach for Identification of the Main Reaction Products, Precursors occurring in the Tea Oxidation Process for Thearubigin Formation

Chapter-4 Potentiostatic measurement, a New Approach for Identification of

the Main Reaction Products, Precursors occurring in the Tea Oxidation

Process for Thearubigin Formation

Thearubigins (TRs) have been shown to arise from a highly complex mixture of phenolic

coupling products comprising for at least 5000, may be up to 30 000 individual chemical

constituents. Any attempt for designing a model system in the tea fermentation process for TR

formation, that allows simplification of the system and subsequent meaningful mechanistic

investigations, would constitute a useful tool on TR research. Many previous attempts were

made in the Kuhnert group to establish such a model system, using chemical oxidants such as

KMnO4, FeCl3, Cu (ΙΙ), MnO2, Br2, I2, peroxidase, etc. In each case hump like features in the

chromatography were created, resembling the chromatographic features of TRs. However, a

close inspection of these “artificial TRs” by high resolution mass spectrometry demonstrated

them to be completely different in composition if compared to TRs. Indeed it appeared that any

chemical oxidant added additional complexity to the system, through metal chelation, over

oxidation of aromatic moieties or electrophilic aromatic substitution chemistry.40,41 Hence, to

produce a meaningful TR model system, two approaches were taken. Whereas the group of

Engelhardt investigated the use of TPPO rich flavan-3-ol-free tea leaves, in my research work I

116


adopted a second approach; an attempt to oxidize flavan-3-ols, reagent free, on an electrode

system, followed by LC-MS analysis for the obtained products.

Epigallocatechin gallate (EGCG) (5 in figure 1.3, p.27) is commonly known as the main

constituent in green tea leaves, characterized by its considerable contribution as a precursor in

the fermentation process for TRs formation. In this part of research a combination of

potentiostatic measurement100,101,102 LC-MSn, and direct infusion-MS experiments were

employed for EGCG (5) solutions subjected to electrochemical oxidation. The aim was to

investigate its primary oxidation processes leading to quinone formation, with quinones acting as

the main building block that could serve a good model for thearubigin formations. Furthermore

reaction products of EGCG derived quinones could be investigated in this approach. The

different elementary oxidation steps and overall oxidation process occurring in EGCG were

identified and the reaction pathways were mainly observed within these experiments. The

hydroxylation process was identified as the main reaction process occurring. Furthermore, I was

able to identify the formation of theaflavins and theasinensins as the main products obtained.

The results provided for the first time spectroscopic evidence for the structure of primary

oxidation products and led to the conclusion that oxidation is mainly taking place on the B-ring

and on the galloyl group of EGCG, where the oxidized components can undergo nucleophilic

addition of H2O leading to the formation of hydroxylated derivatives, but also undergoing

oxidative coupling as well for the formation of theaflavins (their hydroxylated components) and

theasinensins acting as the main products, precursors for thearubigin formation. All these

oxidative processes take place in the absence of TPPO and are solely based on the inherent

reactivity of EGCG.

117



4.1.1 Sample preparation

Epigallocatechin gallate EGCG (5) extract (1.0 × 10-3) solutions were prepared in different 5 %

(v/v). methanolic buffer solutions. Solution A: EGCG dissolved in neutral phosphate buffer of

(pH 7), Solution B: in acidic buffer of (pH 3) (phosphate buffer acidified with H3PO4). The stock

solutions were protected from light with aluminum foil, kept in the refrigerator and used within

one week.

For the supporting electrolytes KNO3 was used this was purchased from sigma Aldrich.

Cyclic voltammetry measurements were carried out on a Potentiostat Versastat 3 using three

electrode system with Au electrode as a working electrode, Ag/AgCl as reference electrode and

platinum wire as counter electrode. The cyclic voltammograms were recorded using a scan rate

of 1 V/sec using multiple cycles (10 cycles) and having 200 points in each cycle.

Solid phase extraction was performed using solvent CH3OH / H2O (1:1) as eluent.

All experiments were carried at room temperature

Direct infusion MSn and LC-MSn measurements were carried after the solid phase extraction.

4.1.2 Direct infusion tandem MS

The oxidized EGCG solution was infused into an ion trap mass spectrometer (Bruker HCT

Ultra) in the negative ion mode using the instrument settings, having nitrogen gas as nebulizing

and drying gas at a flow rate of 10 L / min and a pressure of 10 psi respectively. MS2 experiments

were carried out manually for targeted masses in the mass range between m/z 500 to 1000.

118


4.1.3 LC-MSn Method

The extracted oxidized EGCG solution was analyzed by LC–MS method using Aligent 1100

series LC equipment and a DAD detector with a light pipe flow cell (recording at 400 and 245

nm and scanning from 200 to 600 nm). This was coupled to an ion trap mass spectrometer fitted

with an ESI source (Bruker Daltonics HCT Ultra) operating in the negative ion mode; Auto MSn

mode to obtain fragment ion m/z. Separations were achieved on a POLARIS 5-C18-A column

(length 250 mm, diameter 3 mm, particle size 5 μm) with a step gradient elution employing

acetonitrile (MeCN) and water containing 0.005% formic acid, as follows: 8% MeCN from 0 to

50 minutes, then changing to 31% MeCN for 10 minutes then changing to 25% MeCN for a

further 5 minutes. The column eluent was first directed to the UV detector and then to the ESI

interface operating with a capillary voltage of 1 Volt, and fragmentation amplitude was set

starting at 30% and ending at 200%. MS Operating conditions (negative mode) had been

optimised using theaflavin-3-gallate as the standard calibrant. The capillary temperature was

also set at 300o C, Nitrogen gas was used here as nebulizing and drying gas at a flow rate of 10 L

/ min and pressure of 10 psi respectively.

4.1.4 Amsterdam Density functional Software (ADF)

ADF is quantum chemistry software based on density functional theory, which implies a picture

of many electron systems and yields in principle the exact electron density and the total energy

of the system. The exact energy correlation is unknown, but this software provides a large basis

set where the exchange correlation energy could be computed and provides information about

the binding energy of different structures.(www.scm.com)

119

http://www.scm.com/


4.2 Results and Discussions

Cyclic voltammetry was firstly applied for attaining an electrochemical oxidation process, which

provides a relevant model for studying the oxidation behaviour and the chemical processes

taking place on one of the main flavan-3-ol constituent in green tea leaf EGCG (5), acting as the

main building block in the tea fermentation. The cyclic voltammograms (CVs) are shown in

figures 4.1 & 4.2 of the oxidized solutions of EGCG (A &B) in both acidic (pH 3) and neutral

(pH 7) conditions.

Figure 4.1 Overlapped 10 cycles of 10-3M of EGCG extract at pH 3

Figure 4.1 shows the cyclic voltammogram of EGCG and hence its oxidation potentials

subjected to electrochemical oxidation at pH 3 in ten cycles. The forward response indicates that

Cycle 1

Cycle10 Pa1

Pa2

Pc1

Pc2

120


a two electron-oxidation of EGCG is taking place where as the backward response indicates that

the electro oxidized EGCG products of this process are reversibly reduced. As shown in the

above figure, two pairs of redox peaks appear at the anodic (Pa1, Pa2)and cathodic (Pc1, Pc2)

peak potentials at about 300mV, 700mV, 100mV, -200mV.

In first cycles counting from cycle no.1, it is clearly observed that the backward components of

the response show the presence of the reversible reduction states, indicating that reduction of the

oxidized derivative of EGCG is presumably taking place. When performing further electro-

oxidation cycles by increasing the number of cycles until ten, an increase in the current intensity

of the forward response is clearly observed, which indicates the formation of further oxidized

products at the expense of EGCG, where these are accumulated at the electrode surface causing

its deactivation for the reverse reduction reaction. The ratio of the peak currents of the forward

and backward components is quite high, which shows that chemical reactions following EGCG

oxidation are taking place during the electrochemical oxidation process.

121


Figure 4.2 overlapped 10 cycles of 10-3 EGCG extract at pH 7

Figure 4.2 shows the cyclic voltamogram for EGCG recorded at pH 7, where the potential is

shifted in the negative direction with increasing the pH and the peak current changed

significantly; thus revealing a lower oxidation current at a higher pH taking place, however, the

oxidation potential showed a higher reducing activity. This observation is in line with the

assumption that the deprotonated phenolate anion of EGCG is easier to oxidize if compared to

the protonated form at pH 3. Ten cycles were further performed revealing a stepwise increase in

the oxidation current due to a decrease of EGCG concentration according to Nernst law and

accumulation of the oxidized products. At pH 7 again a reversible reduction was observed. These

experiments demonstrate that EGCG can be repeatedly reversibly oxidized at the surface of an

electrode, generating presumably quinone structures as intermediates. Successive oxidation leads

to an appreciable reduction of EGCG concentration and the formation of TRs like products via

the quinone intermediates.

122


The standard oxidation mechanism for catechols is shown in Scheme 4.1 for obtaining the

conceivable quinone intermediates in the formation of the TRs products. The phenolic –OH

groups can be readily oxidized in a one electron loss followed by abstraction of a proton, which

leads to the formation of a phenoxy radical intermediate. The sp2 hybridised phenoxy radical is

stabilized by a hydrogen bond from one of the adjacent phenolic OH groups to the lone pair of

the oxygen. Further one electron oxidation followed by loss of a second proton yields the ortho-

quinone intermediates. These can react as electrophiles in reaction with other nucleophilic

reduced EGCG products in the elaboration of TR compounds.

OO

O

H

H

H

H-bond

Catechol or pyrogallol ring

-e, -HO

O

O

H

H

OO

OH

OO

OH

Orthoquinone intermediatesoxidized EGCG species

-e, -H

Scheme 4.1 The oxidation mechanism for the formation of the reactive quinone intermediates

4.2.1 Direct Infusion and LC MS experiments

Following the electrochemical oxidation of EGCG, direct infusion tandem mass MSn

experiments were performed for the oxidized EGCG solutions at pH 7 and at pH 3 in the

negative ion mode and the results are presented in figures 4.3 & 4.4.

Figure 4.3 shows direct infusion experiment for the EGCG solution at pH 7, representing the

oxidized components of this compound observed.

123


Figure 4.3. Direct infusion MS2 spectra for the different oxidized state of EGCG extract at pH 7

Spectrum a, shows tandem MS2 fragment spectrum of EGCG with the precursor ion at m/z 456.9

having a characteristic MS2 fragment ion at m/z 304.9 with a neutral loss of 152 which reveals

the loss of galloyl moiety from this compound. Another fragment ion was found at m/z 168.7

which corresponds to the gallate ester moiety of this compound.

Spectrum b in figure 4.3 shows MS2 fragment spectrum with a precursor ion at m/z 455.8

corresponding to the product of a two electron oxidation of EGCG. This precursor ion at m/z

455.8, gives a characteristic MS2 fragment ion at m/z 303.8, which corresponds to the two

electron oxidation product of epigallocatechin EGC ion exhibiting a neutral loss of 152

(C7H5O4), that is the loss of a galloyl moiety. Another MS2 fragment ion was observed at m/z

287.8 corresponding presumably to the EGC ion after loss of single oxygen. No MS3 data could

168.7

192.7 268.8 286.8

304.9

330.9

456.9

168.7

287.8 303.8 331.9

-MS2(455.8), 6.4min. 0.0

0.2

0.4

0.6

0.8

1.0 6 x10 Intens.

0

100

200

300

400

500

100 150 200 250 300 350 400 450 500 550

-MS2(456.9), 2.4min.

m/z

Loss of C7H5O4

2e oxidation of EGCG with a loss of C6H3O3

After the loss of C7H5O4

a

b

(5) (4)

O

OHO

OH OOH

O

OH

OH

OH

O-C7H5O4

-C6H3O3

O

OHO

OH OHOH

OH

OH

OH

OHO-C7H5O4

O

OHHO

OH OHOH

OH

O

OOH

O

OHOH

HO

124


be obtained on this ion to further probe its structure. Furthermore, a characteristic MS2 fragment

ion was detected at m/z 331.9, which corresponds to [M-H-C6H3O3] and this clearly indicates a 2

electron oxidation is taking place on the catechol B-ring, where this reveals interestingly that the

quinone intermediate is sufficiently stable to be observed in our tandem mass experiments and

gives for the first time direct experimental evidence for selective B-ring oxidation in EGCG.

Figure 4.4 Direct infusion MS2 spectra of EGCG extract at pH 3

168.7

192.7

268.8 286.8

304.8

318.9

330.9

456.9

-MS2(456.8), 0.8min

198.7

240.7

284.7 302.7

-MS2(454.9), 2.4min 0

20

40

60

80

100 Intens. [%]

0

20

40

60

80

100 [%]

100 150 200 250 300 350 400 450 500 m/z

a

b

Loss of C6H3O3 (4)

O

OHHO

OHO

OH

O

O

OHHO

OHOH

OH

OH

125


Furthermore, direct infusion experiment for the acidic solution of electrochemically oxidized

EGCG at pH 3 was also performed and this showed the presence of the oxidized product on the

EGC moiety as previously obtained, exhibiting their MS2 fragment ion at m/z 304.8 and m/z

302.7 as shown in figure 4.4, but giving also the m/z 331 as a characteristic MS2 fragment ion,

thus indicating as previously mentioned that oxidation mainly occurs on the B-ring of EGC

exhibiting a loss of C6H3O3 on the oxidized part of EGC moiety.

4.2.2 LC-MSn experiments

ESI-LC-MS experiments were carried out on the oxidized solutions of EGCG in the negative ion

mode. Extracted ion chromatograms (EIC) were created for identification of the oxidation

products. A number of oxidized derivatives of EGCG were clearly identified. Targeted MSn

experiments were carried out, where the chromatographic peak signals were strong enough to

obtain further MS2& MS3 fragment spectra. Such experiments were performed for elucidating

and confirming the structures obtained in the electrochemical oxidation.

Furthermore, the same LC-MSn experiments were employed for identifying the reaction process

mainly occurring for the major products obtained during oxidation. For the latter approach, the

majority of m/z values corresponding to the poly-hydroxylated derivatives of EGCG as

postulated by the oxidative cascade hypothesis were searched for and identified by creating EICs

of the appropriate m/z values and further fragmentation MS2 and MS3 experiments were

performed, thus confirming their structures, revealing the oxygen insertions predominantly

occurring, with several regioisomers were detected at distinct retention times.

Firstly, extracted ion chromatograms EIC of the compound EGCG and its oxidized product were

performed in the negative ion mode. The results confirmed the presence of the different

126


oxidation derivatives of epigallocatechin gallate, which were previously obtained and

characterized in direct infusion experiments.

Figure 4.5 represents the extracted ion chromatograms (EIC) of oxidized EGCG at m/z 456 and

m/z 455. In each EIC generated, the presence of more than one oxidized component for each

oxidized derivative was clearly observed, with two strong chromatographic peaks each

accompanied by one or two weak chromatographic peaks respectively. Accordingly, we can

argue to explain this observation that either two or more different quinones are present on the

oxidized gallocatechin, where one can assign that one of the peaks exhibited oxidation on the B-

ring as the major product obtained, whereas the second intense peak obtained could correspond

to a para-quinone-methide following B-ring oxidation. Additionally formation of a para-quinone

methide on the A-ring of EGC can be considered.

Another plausible suggestion for obtaining more than one peak on the EIC s involves assumption

of epimerization occurring at position-2 on the C- ring taking place during reversible oxidation /

reduction in the cyclic voltammetry cell. It is suggested that following one electron oxidation to

produce a phenoxy radical, this radical is in equilibrium with its tautomeric radical centered at

the A-ring. Such a tautomerisation leads to ring opening of the C-ring. Subsequent radical

addition on the para-quinone-methide can yield either the initial phenoxy radical or its

diastereoisomeric radical. The mechanism is shown in scheme 4.2 a. Alternatively, epimerisation

can be achieved thermally under the influence of high temperature during mass spectrometric

measurements. Here an ionic mechanism is proposed where ring opening of the C-ring via a

benzylic cation yields the diastereoisomer of the EGCG quinone derivative. The latter

alternative mechanism for this epimerisation phenomenon is illustrated in scheme 4.2 b.

127


O

OGHO

OH OOH

OH O

HO

OH OOH

OHO

HO

OH OOH

OH

O

OGHO

OH OO

O O

HO

OH OO

OO

HO

OH OO

O

a)

b)

OG OG

OG OG

Scheme 4.2 Suggested alternative mechanism for the epimerisation phenomenon a) via reversible oxidation and

tautomerisation; b) via ionic mechanism. Both leading to the presence of the diastereoisomer (G denotes the galloyl

moiety)

128


Figure 4.5 Extracted ion chromatograms in the negative ion mode for the oxidized epigallocatechin gallates at m/z

456 and m/z 455

Figure 4.6 exhibits one of the tandem MS3 of the extracted ion chromatograms, for example; the

oxidized product at m/z 456 showed the main MS3 fragment ion detected at m/z 303.9, thus

allowing assignment of this ion as the quinone of EGCG. At different retention times, the same

oxidized products were observed, however with different intensities and showing at m/z 303.9 as

the main MS3 fragment ion obtained, thus revealing that mainly oxidation is occurring mostly on

the EGC ring as previously obtained in direct infusion tandem mass experiments.

EGCGLCMSSPE.D: EIC 456.0 -All MS

EGCGLCMSSPE.D: EIC 455.0 -All MS 0.00

0.25

0.50

0.75

1.00

1.25

6 x10

Intens.

0

2

4

6

5 x10

Intens.

0 10 20 30 40 50 60 Time [min]

129


O

OOH

OH

OHOH

HO

Figure 4.6 Tandem mass experiments for the EIC at m/z 456 revealing the presence the oxidized part occurring in

EGCG

Secondly, further tandem LC MSn experiments were performed in order to investigate the

chemical reactions taking place for the formation of the products from the oxidized

epigallocatechin gallate derivatives, and EICs were generated for the hydroxylated components

of EGCG as expected from the oxidative cascade hypothesis. Figure 4.7 shows the presence of

the different polyhydroxylated derivatives of EGCG at m/z 473, m/z 489 and m/z 505. The

signals were strong enough to obtain MS2 and MS3 fragment ions.

544.5

913.0

455.9

-MS, 27.0min

168.7

303.9 379.9

-MS2(455.9), 27.0min

168.7

303.9 -MS3(456.1->378.7), 27.0min

743.0

913.0

455.9

-MS, 31.5min

168.7 303.9

438.6

379.9 -MS2(455.9), 31.5min

303.9 -MS3(456.0->379.9), 31.6min

0 50

100

Intens. [%]

0 50

100 [%]

0 50

100

[%]

0 50

100

[%]

0 50

100

[%]

0 50

100

[%]

100 200 300 400 500 600 700 800 900 1000

m/z

130


Figure 4.7.a) Total ion chromatogram, b) EIC for the mono-hydroxylated EGCG at m/z 473, c) EIC for the di-

hydroxylated EGCG at m/z 489, d) EIC for the tri-hydroxylated derivative at m/z 505

The mono-hydroxylated epigallocatechin gallates were detected at m/z 473 observed at retention

times of 31.1 min., 24.6 min. and 19.8 min., showing characteristic MS2 fragment ions with

different intensities at m/z 455 revealing the loss of H2O, at m/z 404.6 with a neutral loss of 69

which is characteristic of a loss of (C3HO2) , m/z 318 characterized by one oxygen insertion in

the oxidized EGC part (C15H12O8) of EGCG, and a fragment ion at m/z 168.7 corresponding to

the oxidized galloyl moiety. At the retention time of 31.1 min., a fragment ion of high intensity at

m/z 303.9 which is the quinone of epigallocatechin exhibiting a neutral loss of 169.9 is observed,

and this indicates the loss of the galloyl moiety and oxygen insertion within this part of the

molecule (C7H5O4+O) and this is shown in figure 4.8.

EGCG(OH)PH7(A).D: TIC -All MS

EGCG(OH)PH7(A).D: EIC 473.0 -All MS



0.5

1.0

1.5

8 x10

0 1

2

3

5 x10

Intens.

0.0 0.5 1.0 1.5 2.0 2.5 5

x10 Intens.

0.0 0.5 1.0 1.5 2.0 2.5 5

x10 Intens.

0 10 20 30 40 50 60 Time [min]

Intens.

a

b

c

d

EGCG+O1

EGCG+O2

EGCG+O3

131


Figure 4.8 Tandem mass experiments for EIC of the mono- hydroxylated EGCG at m/z 473.1

Extracted ion chromatogram for the tri-hydroxylated derivative at m/z 505 of EGCG are

exhibited in figure 4.9 and as shown this pseudomolecular ion gave MS2 fragment ions at m/z

436.6 as a main fragment ion revealing the loss of 69 which is (C3HO2), m/z 320.7 the

characteristic fragment ion of the mono-oxygenated epigallocatechin EGC, after oxygen

insertion within EGC, and exhibiting a neutral loss of 184 which is the loss of a dihydroxy

galloyl part, obtained after two oxygen insertions into the eliminated galloyl moiety. Further

fragmentation of the main MS2 fragment ion gave the quinone counterpart of the mono

oxygenated epigallocatechin MS3 at m/z 318 and this is illustrated in the following figure.

168.7

456.9

473.5

-MS, 31.0min

168.7 303.9 356.8

455.8 472.8

404.6 -MS2(473.5), 31.1min

262.7 376.7 402.6

-MS3(473.8->404.4), 31.1min

168.7

260.6 318.7

402.6 473.0

-MS, 19.8min

168.7 260.6

404.6 454.9

-MS2(473.0), 19.8min

262.6

318.6 402.6

-MS3(473.1->404.6), 19.8min

0 50

100

Intens. [%]

0 50

100 150 [%]

0 50

100 150 [%]

0 50

100 150 [%]

0 50

100 150 [%]

0 50 100 150 [%]

100 200 300 400 500 600 700 800 900 1000 m/z

C15H10O8

C15H12O7 with the loss of (-C7H5O5)

132


Figure 4.9 Tandem mass experiment for EIC of the tri- hydroxylated EGCG at m/z 505 showing oxygen insertion

within EGC and the loss of the galloyl part with 2 oxygen insertions in it. (The blue–OH corresponds to the tentative

regioisomers of the hydroxylated derivative).

From the above mentioned results, I was able to conclude that hydroxylation is the main reaction

process occurring in the formation of the model thearubigin-like products. Performing targeting

fragmentation experiments I was able to investigate that normally oxidation process is occurring

continuously within the epigallocatechin part and the galloyl part of epigallocatechin gallate,

where the H2O molecules acting the main nucleophiles for the formation of the aforementioned

hydroxylated reaction products. It should be noted that the hydroxylated flavan-3-ols have not

been detected on FT-ICR-MS in TR extracts, their derivatives are observed for the first time in

the model system, and this is due to the higher catechin concentration present in the tea leaves

159.6 192.8 320.7 368.7

417.0

459.0

485.0 505.0

436.6

-MS2(505.1), 64.0min

234.4

318.6

378.6

418.5

-MS3(505.2->437.0), 64.0min 0

50

100

150

Intens. [%]

0

50

100

150

[%]

100 200 300 400 500 600 700 800 900 1000 m/z

- (C7H5O4+ 2 O)

O

OHHO

HOOH

HO

OH

HO

O

OHHO

HOO

HO

O

HO

133


~200 mg/g acting as the main nucleophiles in the oxidation process. If electrochemical oxidation

is carried out at lower EGCG concentrations competitive nucleophilic addition by water

dominates over C-C coupling between two EGCG units.

4.2.3 Identification of different reaction products by direct infusion experiments

The main advantage of direct infusion tandem MSn experiments if compared to LC-tandem MS

runs, is its operating mode, providing higher signal intensities for much longer time, thus

permitting optimizing of trapping and fragmentation.40 It was worth to further apply direct

infusion MSn for characterizing other reaction products that can be formed from the oxidized

EGCG, acting as the main constituent in the tea leaves. I was able to identify the presence of

theasinensins and theaflavingallates as the main products.

Theasinensin A is characterized by its mass spectrum showing a pseudo-molecular ion at m/z 913

in the negative ion mode. This compound was identified from the electrochemical model

oxidation system, which gave its characteristic MS2 fragment ions. The first fragment ion was

observed at m/z 761 with a neutral loss of 152 amu, and a second fragment ion was also observed

at m/z 743 having a neutral loss of 170 amu revealing the loss of one of the galloyl moieties of

this compound. A third fragment ion was observed at m/z 609 and another fragment ion was

exhibited at m/z 591.2, indicating the loss of two galloyl groups from this molecule. These MS2

fragment ions were previously detected by us and confirmed in black tea thearubigin fraction to

be characteristic of theasinensin A. Interestingly, a characteristic MS2 fragment ion was detected

at m/z 457 corresponding to EGCG (5) and at m/z 455 to its oxidized form. Thus, I was able to

conclude that in black tea thearubigin, theasinensin (TS) (12) are the main products obtained by

134


the reaction of epigallocatechin gallate dimer, which means that oxidation is mainly taking place

on B-ring rather than A-ring and these are exhibited clearly in figure 4.10.

Figure 4.10 Direct infusion tandem mass experiment showing the presence of the compound theasinensin A (12) as

the main product at m/z 913 and its characteristic MS2 fragment ions at m/z 457 of EGCG (5) and its quinone

counterpart at m/z 455

Furthermore, Theaflavin (TF) mono- and di-gallates (9, 10 &14) were identified, where they

occurred in their hydroxylated and quinone products. For example, in figure 4.11 is shown the

presence of hydroxylated theaflavin-digallate with two oxygen insertion (C43H32O22) was

identified, having a neutral loss of 170 Da, which indicates a loss the galloyl part and one oxygen

from this molecule (C7H6O5) to give the quinone of mono-hydroxylated theaflavin-monogallate

457 590.9 742.9 760.9

-MS2(913.0), 10.5min

454.9 500.9 528.9

572.9

590.9

608.9 699.1 724.9

742.9

760.9

774.9

-MS2(913.0), 10.3min 0

20

40

60

80

100

120

Intens. [%]

0

20

40

60

80

100

120

[%]

300 400 500 600 700 800 900 1000 m/z

O

OHO

OH OOH

O

OH

OH

OH

O

O

OHO

OH OHOH

OH

OH

OH

OH

O

(12) (5)

O

O

O

HO

OH

OH

OH

HO

HO

O

HOOH

OH

OH

OHOH

OH

O

OOH

OH

OH

135


and these are denoted by arrows in the following figure. Theacitrin (13) and theanaphthoquinione

(8) derivatives could not be identified within EICs.

Figure 4.11 Direct infusion tandem mass experiment showing the presence of the hydroxylated theaflavin gallates

as main reaction product formed after oxidation of EGCG (regioisomerism for –OH insertions were selected

randomly)

By using potentiostatic experiments for oxidizing one of the main tea flavan3-ols constituents

present in green tea leaves, I was able to confirm that theaflavin mono- and di-gallates and

theasinensins are the main products formed, thus acting as the main precursors for thearubigin

728.9 898.9

-MS2(899.0), 26.5min

0

20

40

60

80

100

Intens. [%]

200 300 400 500 600 700 800 900 1000 m/z

C36H26O17

C43H32O22

(14) (9)

O

O

OO

HO

OH

HO

HOOH

OH

OH

OO

OHO

OH

HO

O

O

OO

HO

OH

HO

HOOH

OH

O

O OH

OHOH

OHO

OH

OHHO

HO

OH

136


formation, and these are normally formed by oxidative coupling reactions where oxidation is

mainly taking place on the B-ring.

In a conclusion, the oxidation products that mainly occur in a reagent free electrochemical

oxidation of EGCG (5) via their quinones are mainly the hydroxylated EGCG, TFs (9, 10, 14),

and TS (12). All compounds are formed following two electron oxidation in the absence of an

enzyme, thus reflecting the natural reactivity of EGCG in aqueous solution.

4.2.4 Theoretical calculations and Computational Analysis

In continuation for understanding the hypothesis behind the thearubigin TR formation, I

performed further computational work based on the density functional theory DFT using

Amsterdam density functional theory software. DFT is a quantum mechanical method which

allows studying the electronic structure of many body systems. DFT provides large basis sets,

which facilitates to study the exchange correlation energies of the whole electronic system, thus

providing accurate knowledge of the electronic distribution within the main reacting tea catechin

structures, acting as the main building blocks. Information on their relative stability is obtained,

which can be compared with the experimental values.

By performing (DFT) calculations based on high correlation energy Hybrid / B3LYP, I was able

to obtain orbital coefficients for HOMO and LUMO of the different catechin structures acting as

the main building blocks for thearubigin formation, and hence identify the most likely

nucleophilic and electrophilic sites. With this information on the electron distribution within the

main building blocks (epigallocatechin, gallocatechin, epicatechin, catechin), conclusions can be

drawn about the likely regiochemical outcome of C-C coupling reactions.

137


The electron distribution of the highest occupied molecular orbital (HOMO) and the lowest

unoccupied molecular orbital (LUMO) levels on the different parts of the mentioned flavan-3-ol

structures and their diastereoisomers were calculated. In figure 4.12 below, we can observe the

electronic distribution of the hydroxylated catechin, that is, the epigallocatechin and its isomer

revealing that the HOMO level mainly lies on the catechol (B-ring) of this compound, while the

LUMO level lies on the A- ring. Furthermore, performing the electronic distribution on

epicatechin structure and its isomer, the HOMO level was identified to occupy mainly on the

resorcinol (A-ring), while the LUMO level was lying on the catechol (B-ring) and this showed a

different behavior than its hydroxylated analogue (epigallocatechin). Furthermore, the isomer of

this compound, that is, catechin contained mainly the HOMO the LUMO level on the A- and C-

ring. This constitutes an interesting finding since it readily explains why tea chemistry of

catechins is B-ring coupling chemistry. In contrast in many other dietary plants oxidation of

catechin moieties leads to formation of Proanthocyanidins with A- ring chemistry.

138


a)

O

OHHO

OHOH

OH

OHA

B

C

b)

O

OHHO

OHOH

OH

OH

139


c)

OHO

OHOH

OH

OH

d)

OHO

OH

OH

OH

OH

Figure 4.12 The HOMO and LUMO levels of; a) epigallocatechin, b) gallocatechin, c) epicatechin and, d) catechin.

(the dark blue and red orbitals are the HOMO where as the light blue and brown orbitals are the LUMO level)

140


Furthermore, focusing on the epigallocatechin structural units following their different quinoid

forms resulting from their different oxidation regiochemistry, the energies of their orbital

coefficients were calculated. Based on the high energy correlation calculations of Hybrid/

B3LYP, the results obtained were very considerable and in agreement with the theoretical

mechanism for thearubigin formation. It was shown that the one electron oxidation state on the

catechol moiety has the highest energy in the HOMO level, which expresses its tendency for

donating a further second electron; while the highly electrophilic quinone oxidation state has the

tendency for accepting the electrons having the lowest energy in its orbital coefficient. These

calculations are illustrative and in agreement for understanding the mechanism for thearubigin

formation from their building blocks, which explains more favorably occurring 2 electron

oxidation on the catechol B-ring for the hydroxylated catechin structures, that is, the

epigallocatechins.

a)

O

O

OH

OH

OH

OH

HOB

141


b)

O

OH

OH

OH

OH

O

OH

A

c)

O

O

OH

OHOH

HO

O

Figure 4.13 epigallocatechin oxidation states: a) 1 electron oxidation on the catechol group, b) 1 electron oxidation

on the resorcinol group, c) 2 electron oxidation on the catechol moiety. (the dark blue and red orbitals are the

HOMO levels where as the light blue and brown orbitals are the LUMO levels)

142


Table 4.1 The energies of the HOMO levels in the different oxidation states of epigallocatechin

compounds.

EHOMO(e.V.)

Epigallocatechin oxidation state (catechol) -6.496

Epigallocatechin oxidation state (quinone) -7.228

Epigallocatechin oxidation state (resorcinol) -6.828

143


4.3 Conclusion

The aforementioned results explain one aspect behind the thearubigins formation in black tea,

where oxidation is more favored on the B-ring over the A-ring, which is opposite to chocolate or

red wine having preference of A-ring over B-ring and the following discussion illustrates our

hypothesis.

Since hydroxylation is the main reaction process occurring in the fermentation of the tea leaves,

where water acting as the main nucleophile, the more two electron oxidation is favorable to

occur normally on the trihydroxy- B-ring, that is, the epigallocatechins more over than the

epicatechins, where the former exhibits the oxidation on the catechol B-ring. And as a

consequence theasinensins, theaflavins and their gallate esters are the main reaction products

formed, acting as the main precursors for further oxidation processes in thearubigin formation

rather than proanthocyanidins. However, one cannot neglect the presence of proanthocyanidins

that may survive and may take part in the fermentation process, but the later are not observed

abundantly to be present as the main precursors for the thearubigin formation.

144

Investigation of isomeric structures in complex mixtures using Ultraperformance Liquid Chromatography Coupled with Hybrid Quadropole /Ion Mobility/ Time of Flight Mass

Spectrometry

Chapter-5 Investigation of isomeric structures in complex mixtures using

Ultraperformance Liquid Chromatography Coupled with Hybrid

Quadropole /Ion Mobility/ Time of Flight Mass Spectrometry

Thearubigins (TRs) are mainly characterized by the unresolved chromatographic behavior in

reversed phase column chromatography revealing the complexity of this material produced in

food processing. Kuhnert et al. showed by employing (ESI-FTICR-MS) in combination with LC-

tandem MSn mass spectrometry that TRs are comprised from around 30 000 distinct compounds,

when counting in their isomeric structures.103,104 Furthermore, when analyzing black tea by

electrospray ionization ESI–HPLC and tandem mass spectrometry, sets of multiple isomers were

observed, having identical masses. These observations were based on the difference of retention

times of the pseudo-molecular ions obtained in ESI-HPLC-MS experiments, and the different

fragment spectra obtained in direct infusion MSn mass spectrometry experiments, when varying

the collision energy applied in the collision induced dissociation (CID).

Isomerism can be studied in principle in three different ways using mass spectrometry. Firstly by

LC-MS method resolving isomeric structures by chromatography, secondly by energy resolved

mass spectrometry when varying collision energies and finally by ion mobility mass

spectrometry (IMS-MS) by separating isomeric ions in the gas phase. This chapter will show the

145


Spectrometry proof of the principle that IMS-MS provides a useful new approach towards investigating

isomers in TRs.

Previous studies demonstrated the importance of the ion mobility mass spectrometry IMS-MS as

a tool to investigate protein confirmation in the gas phase,105 however applications in small

molecule chemistry are still rare and not general. A series of reports on this method were able to

separate pairs of isomeric compounds have been reported in the literature starting with the

pioneering work by group of Rappaport on the ion mobility separation of E/Z pairs of isomers106

and further work including of separation of disaccharides by Gabryelski et al.,107 separation of

ephedrine/ pseudoephedrine pair of diastereoisomers by McCooeyey et al.,108 separation of

carbamazepine hydroxylated regioisomeric metabolits by Cuyckens et al.,109 separation of silver

ion adducts of regioisomeric pair of hesperidine and neohesperidine by Clovers et al.,110

separation of regioisomeric phthalic acids by Guevremont et al.,111 the separation of E/Z isomers

of caffeic acid derivatives by Xie et al.,112 separation of diastereosimeric terphenyl ruthenium

complexes by Williams et al.,113 separation of diastereoisomeric nickel complexes of all naturally

occurring amino acids by Mie at al.,114 and the separation of enantiomeric pairs of R/S atenolol

in the presence of (+)-2-butanol as a chiral collision gas.115 Nau et al. studied the inner phase

chemical reactions inside the molecular containers in the gas phase for understanding the

supramolecular reactivity which has implications in catalysis.116 In his investigation, Nau et al.

described the gas phase reactions of encapsulated guests such as azoalkanes in a macrocyclic

hosts such as cucurbiturils by using ion mobility experiments.

In my research work, for further improvement of isomer separation in TR analysis, ion mobility

mass spectrometry IMS-MS was employed as alternative tool for the distinction between

isomeric compounds having the same molecular formulas and nominal masses but differing in

146


Spectrometry their structures. IMS-MS separates ionic species as they drift through an inert gas under the

influence of an electric field. The drift time of the ion here depends not only on the mass of the

ion, but also on its collisional cross section. (the collisional cross section is the energy

transformed from the inert gas upon collision to the compound we are examining). More

compact ions with smaller collision cross section will drift more quickly than expanded ions.

Hence, ion mobility mass spectrometry constitutes a relevant option in differentiating between

isomeric small molecules in natural product chemistry based on the separation in the drift time

dimension. Therefore, the isomeric compounds are separated according to their collisional cross

section, which can be obtained experimentally and compared to the theoretical computed values,

thus allowing predictive structure elucidation.

It is worth noting that isomeric molecules present in complex mixture such as TRs are usually

very difficult to separate, therefore traditionally chromatographic separation prior to mass

spectrometric detection are crucial for isomer identification. Ultra performance liquid

chromatography (UPLC) is the latest advancement in chromatographic separation. When

compared to high performance liquid chromatography, UPLC has improved chromatographic

performance due to the small particle size of 1.7µm in the stationary phase providing high

resolution, enhanced sensitivity, and increased separation speed.117 UPLC when coupled to ion

mobility mass spectrometry offers the possibility for accurately distinguishing isomers whether

they are positional, structural, or stereo-isomers. As model study for the IMS-MS when

separating isomeric structures having identical masses, I used two relevant isomeric compounds

Theasinensin C (TS) and proanthocyanidin B (PA). TSs as investigated in my previous studies

as model compound due to their relevance to black tea chemistry. On the other hand, PAs are

present abundantly in many botanical and dietary sources including cocoa, apple and others.65, 118

147


Spectrometry Here extracts of Rhododendron and Ziziphus Spinae Christi species were employed as sources of

PAs as reported earlier 119 Both isomeric structures showed to possess a wide range of important

biological activities on the human health.120,121 In this contribution IMS/MS mass spectrometry

was employed to study the distinction between TS (15 in figure 2.5-chapter 2) present in SII

thearubigin fractions from Assam tea and Ceylon tea, and PA (16 in figure 2.5-chapter 2) present

in Zizizphus Spinae Christi and Rhododendron species.

The results obtained clearly indicated that IMS-MS has the potential to separate isomeric

structures in TR fractions and the Rhododendron extracts. These were separated according to

their mobility drift time and structurally confirmed by using the MSE collision mode.

Furthermore, the results indicated that under the influence of high temperature structural

variation may occur in the gas phase chemistry of black tea thearubigins fractions during the

IMS/MS mass spectrometry experiments.


5.1.1 Preparation of Proanthocyanidin extract

Green leaves (5g) from two different botanical source of Rhododendron and Ziziphus Spinae

Christi were freeze dried at -20 ºC overnight, extracted with aqueous methanol ( 100 ml, 70 %),

homogenized with a blender, and ultra sonicated for 10 min. These extracts were filtered through

a whatman no.1 filter paper. The solvents were removed by evaporation in vacuum and the

extracts were stored at -20ºC until required.

148


Spectrometry

5.1.2 Preparation of Thearubigins

Thearubigin fractions from two different types of tea: Assam and Ceylon Black tea were

prepared according to the caffeine precipitation method.11, 31

5.1.3 Sample Preparation

A) 5mg of the above mentioned samples of thearubigin fractions and Proanthocyanidine

extracts were dissolved in 1ml methanol: water (7:3).

B) 4 mM of a dimeric solution of epigallocatechin was synthesized in the lab. By dissolving

epigallocatechin in tap water and adding some FeCl3 as a trace metal to enhance

oxidation process. This solution is diluted to 100µM in a 50ml conical flask allowing

headspace of 5ml (containing sufficient amount of 1ml of O2 for oxidation).

5.1.4 Acquity UPLC-IMS-MS Conditions

Chromatographic separation was achieved using an ACQUITY UPLC system (Waters) on a C18

column (length 100mm, diameter 2.1 mm) having particle size of 1.7µm with a step gradient

elution employing water and acetonitrile with acetic acid 0.1% as follows: starting from 2%

MeCN and held for 2 min. followed by a 7 min linear gradient to 20 % MeCN, then stepped

to100 % at 30 min. and held for 5min. and was finally stepped to 2 % in order to equilibrate the

column. The column temperature was set at 35 ºC and a flow rate of 0.3 ml/min.

The MS detection was provided by Synapt HMDS (Waters). The negative ion electrospray mode

was used for data collection. The desolvation gas was nitrogen and the collision gas was Argon.

149


Spectrometry The carrier gas for ion mobility was nitrogen. The source temperature was 100ºC and the

desolvation temperature was 350ºC. The cone voltage was 40 volt.

The lock mass compound used was leucine enkephaline with a reference mass at m/z 554.2615

in the negative mode.

Data were acquired in the MSE mode in which two separate scan functions were programmed for

the MS/MS acquisition mode. One scan function was set at low collision energy of 4v in the trap

cell, and the other scan was set at high collision energy in the transfer cell (ramped from 25 to 40

V). The mass spectrometer switched between two functions during data acquisition, as a result

information about the intact precursor ions and fragment ions were obtained from one LC

injection.

For the IMS experiments, the traveling wave (T-wave) ion mobility cell was operated at a wave

velocity of 652 m/s and wave amplitude of 40 volts.

For the IMS-TOF experiments, the collision energies (CE) for both the trap and the transfer T-

wave cells were set at low values so that the intact [M-H]- ions were detected. For the IMS/MS-

TOF Time aligned parallel fragmentation experiments, the trap cell was set at a low collision

energy value to allow the intact [M-H]- ions to enter the IMS drift cell. Upon exiting the IMS

drift cell, these ions were drift time separated, and then entered the transfer cell (set at high

collision energy) which caused fragmentation within the transfer T-wave cell. As a result the

fragment ions were detected by the TOF detector and time aligned with their precursor ions.

150


Spectrometry

5.2 Results and Discussion

In this contribution we demonstrate the ability of UPLC-IMS/MS mass spectrometry to separate

and study different isomeric structures present in complex mixture such as TRs. Hence this

chapter serves as a proof of principle for the ability of IMS-MS to differentiate isomeric

structures in black tea polyphenol chemistry. In the negative ion electrospray mode, initially full

scan spectra at low collision energy were acquired of a black tea infusion. The total ion

chromatogram revealed the presence of isobaric ions in a complex mixture, where extracted ion

chromatograms were created restricted at m/z 609. These were exhibiting the specific retention

times of the different isomeric components of theasinensins (TS) in the thearubigin fractions and

the proanthocyanidins B (PA) in the Rhododendron and Ziziphus Spinae Christi extracts. The

attention was focused on the retention times, where these isomeric structures occurring.

For example: the isomeric structures at m/z 609 in TRs in Assam black tea were observed at

retention times of 1.28 min. and between 10.09 min. and 10.22 min. In contrast a synthetically

prepared solution containing oxidized epigallocatechin showed its signals in the EICs at m/z 609

at retention times of 2.91 min., 7.06 min., 7.74 min., 9.11 min., 9.48 min., which indicates that

the naturally biosynthesized or naturally oxidized components in black tea under the effect of tea

polyphenol oxidase enzymes (TPPO) vary significantly from the synthesized oxidation products

in the model system. On the other hand Proanthocyanidins PA present in e.g. Rhododendron

extract revealed the presence of different isomeric structures at m/z 609 ions having their

retention times at 2.35 min., 4.98 min., and 10.40 min. The mobility drift times of the

corresponding isomeric structures identified at m/z 609 including all the experiments performed

151


Spectrometry are summarized in table 5.1. Furthermore, under high collision energy (CE) signals

corresponding to the isomers of both theasinensins (TS) and Proanthocyanidins (PA) showed

characteristic fragment spectra allowing their assignment. TS isomers from the precursor ion at

m/z 609 gave characteristic fragment ions at m/z 471 corresponding to a retro Diels-Alder

fragmentation with a loss of 138 Da from one of the benzopyran moiety, at m/z 288 and m/z 287

having quinide methide fission with a loss of H2O. Even though all the isomers showed their

characteristic fragment ions, the relative abundances or their relative intensities were different.

On the other hand, PA isomers from the precursor ion at m/z 609 gave its characteristic fragment

ions at m/z 483 characterized by heterolytic ring fission (HRF) with a loss of 126 Da, at m/z 441

of retro Diels-Alder fragmentation (R.D.A.) with a loss of 168 Da, at m/z 423 R.D.A with a loss

of water characterized by a loss of 186 Da, at m/z 305 giving quinide methide fissin (Q.M.) with

a loss of 304 Da, where their signal intensities were clearly different. Figure 5.1 exhibits an

example for the difference in fragmentation spectra between theasinensins TS and

proanthocyanidins PA.

152


Spectrometry

Figure 5.1 an example of two different MS2 fragment spectra between the two isomeric structures a) theasinensin b)

Proanthocyanidin with a precursor ion at m/z 609 separated by UPLC/IMS in the negative ion mode

5 mg in 1ml meth:H2O(7:3)

m/z 50 100 150 200 250 300 350 400 450 500 550 600

%

0

100 Sudanese Plant_frag_01_rt_01 68 (2.351) 1.90e3 423.099

305.084

125.031 177.030

179.045 397.109

609.157 441.105

483.121

5 mg in 1ml meth:H2O(7:3)

m/z 50 100 150 200 250 300 350 400 450 500 550 600

%

0

100 Thearubigins_Theas_frag_01_rt_01 32 (1.289)

3.56e3 609.156

201.967

167.040

232.987

471.163

305.072

b)

a)

O

OH

HOOH

OH

OH

O

OH

OHHO

HO

HOOH

-168R.D.A.

Q.M.

H.R.F. -126

HO

O

HOOH

OHHO

HO OH

OH

OHO

OH

HO

OH

HO

-70

-138R.D.A

Q.M.

153


Spectrometry

The ion mobility drift time (Dt, ms or bins) of the identified peaks at their chromatographic

retention time (Rt; min.) were obtained with their accurate m/z- values. UPLC IMS/MS results

can be studied in the drift scope with three possible axis options (m/z/Dt, Dt/Rt, m/z/Rt).

Figure 5.2 shows two of these options , the first part exhibits the drift time (Dt) versus the m/z of

the isolated isomers for theasinensins in one of the SII thearubigin fractions (Assam tea) and the

second part exhibits its corresponding drift time (Dt) versus the retention time (Rt).

a)

Thearubigins_Theas_frag_01.raw:1

Thearubigins_Theas_frag_01.raw : 1

154


Spectrometry

b)

Figure 5. 2 A drift scope for m/z 609.155 of theasinensin C identified at retention time of 1.28 min

expressed in a) Dt versus m/z, b) Dt versus Rt; showing two distinct ion mobility signals in the IMS-MS

experiments which indicates the existence of an additional isomeric structure

In figure 5.2 chromatographic peak at retention time 1.28 min. was observed, exhibiting two

distinct mobility drift times at 5.56 ms and 5.18 ms. Extracting their mass spectra at both drift

times, showed different characteristic components, the former was giving an accurate mass at m/z

609.155 (RT 1.28 min), while the latter was a different component with an accurate mass at m/z

609.157. MS2 fragment spectra confirmed that both of these structures were isomeric of

theasinensins.

Thearubigins_Theas_frag_01.raw:1

Thearubigins_Theas_frag_01.raw : 1

155


Spectrometry

Furthermore, a new pseudomolecular ion at m/z 609.178 was observed at a retention time of

10.22 min. with high peak intensity, accompanied by a second peak at a very low intensity at

10.09 min. Both showed to have a specific mobility drift time at 5.67 ms. When investigating

their MS2 fragment spectra obtained in the MSE collision experiments, the data revealed the

presence of a flavonoid glycoside, Quercetin-3-O-rutinoside, which is characterized by its

fragment ion at m/z 300.047 and m/z 301.050 corresponding to the cleavage of the glycosidic

bond. Structure and fragmentation are illustrated in table 5.1.

By examining the second thearubigin fraction from Ceylon tea, the same behavior was obtained

if compared to the previous thearubigins SII fraction of Assam tea. A chromatographic single

peak of an analyte at m/z 609.159 showed the presence of two signals in their mobility drift

times. The former was found at a drift mobility time of 5.18 ms of the same compound at m/z

609.159, where as the latter was obtained at 5.51 ms showing the presence of a new isomeric

structure. Investigating their MS2 fragment spectra in the MSE collision experiments both were

identified as theasinensins TS. Furthermore, new compound was obtained at retention times of

10.25 min. having the highest chromatographic peak intensity and accompanied by a very small

peak intensity at retention time of 10.13 min., detected at m/z 609.183, exhibiting the same

specific mobility drift time of 5.62 ms, and showing fragmentation at m/z 300.047 and 301.050,

thus confirming the latter structures to be Quercitin-3-O-rutinoside.

Figure 5.3 shows a 3D representation of the IMS-MS experiment exhibited for the

chromatographic peak at 1.28 min, acquiring its mobility drift time at 5.18 ms and showing an

additional new distinct mobility drift signal for an additional isomeric component observed at

156


Spectrometry 5.56 ms. While figure 5.4 shows the IMS-MS experiment for Quercetin-3-O-rutinoside, which is

characterized by only one single peak in the mobility drift time acquired at 5.62 ms.

Figure5.3. 3D representation of the IMS-MS experiment obtained from the thearubigin fraction; where the

chromatographic peak obtained at 1.28 min showed two tentatively diastereoisomeric structures of TS in the

mobility experiment; the drift time range is exhibited between 5-6 ms (120-140 bins)

157


Spectrometry

Figure5.4. 3D representation of the IMS-MS experiment obtained from the thearubigin fraction corresponding to

the Quercitin 3-O-rutinoside dimeric structure at m/z 609.178 identified at a retention time of 10.22 min giving only

singlet in the mobility drift time at 5.67ms.

In this work, I was able to identify the presence of new structural components of theasinensin

(TS) obtained; to be present in both black tea thearubigin fractions and these were characterized

by exhibiting different mobility drift times in IMS/MS experiments. These were subsequently

identified as theasinensins by the MSE collision experiments. However, elucidating the detailed

structures of the two isomeric TS observed in the ion mobility experiment is a challenge due to

the unavailability of theasinensin authentic reference materials. One significant outcome of my

observations can be demonstrated as the following: Under the influence of high temperature,

which is normally employed during the spraying process in the ion mobility mass spectrometry

158


Spectrometry experiment structural interchange of theasinensins in thearubigins SII may take place. As

structures for the two isomeric TSs two types of isomerism might be considered; Firstly it is

feasible that two TS atropisomers are responsible for the two peaks observed in the mobility

dimension. Secondly epimerization may occur at position 2 of the C-ring of one of the two

catechin moieties in TS, which resulted in a pair of diastereoisomers, characterized by their

distinct collisional cross sections and therefore distinct ion mobility signals were obtained.

A further inspection on the MSE fragmentation experiments which are illustrated in table 5.1, we

can observe that theasinensins at the initial drift mobility time exhibited no loss of water H2O,

and if we compare the fragmentation spectra with the second isomeric structure obtained at 5.56

ms, a loss of H2O can be observed. This may indicate that a certain cis- trans- isomerism is

occurring at C-ring of catechin moieties in TS. Jaiswal et al. could show that in MS2 fragment

spectra of proanthocyanidins PAs, epicatechin moieties typically show a loss of H2O, where as

PA based on catechin do not loose H2O.119 The characteristic loss of H2O during fragmentation

therefore lead to suggest that isomerism observed here is based on change of stereochemistry at

the C-rings of TS rather than on atropisomerism. In order to substantiate this hypothesis

collisional cross section calculations of TS based on different isomeric structures including

epigallocatechin/epigallocatechin and gallocatechin/ epigallocatechin pairs were carried out. A

correlation between the calculated collisional cross sections of different isomeric structures and

their mobility drift times was established, so that I was able to tentatively assign the mobility

drift time for each isomer characterized by its calculated collisional cross section as shown in

figure 5.9. The results revealed that the isomeric component occurring at retention time of 1.28

159


Spectrometry min. and mobility drift time of 5.18 ms has a collisional cross section of 154 A2exhibiting a C-

ring trans-stereochemistry (catechin building block), where as the latter identified at 5.56 ms

showed a collisional cross section of 157 A2 having a C-ring cis-stereochemistry (epicatechin).

This phenomenon occurring can be attributed to epimerization occurring under the influence of

high temperature in the ion mobility experiment and the mechanism is illustrated in scheme 5.1.

O

OH

OH

OH

HO

H+

O

OH

OH

OH

HOH

O

OH

OH

OH

HOH

O

OH

OH

OH

HOH

O

OH

OH

OH

HO

OH

OH

OH

OH

OH

OH

OH

OH

OH

OH

Scheme 5.1 Mechanism of epimerization phenomenon occurring in thearubigin fractions

Furthermore, in my observation for trans-isomer in the 3D representation, these appeared to be

more compact than the cis-isomers, which confirmed experimentally that the former trans-

isomers are characterized by a shorter mobility drift time. Furthermore, Quercetin 3-O-rutinoside

was identified at retention time of 10.22 min. having its mobility drift time at 5.67 ms, where it

160


Spectrometry exhibited a calculated collisional cross section of 160 A2. The IMS-MS data corresponding to

these isomeric structures are summarized in figure 5.5 along with their specified calculated

collision cross section. The computed collisional cross section (CCS) values are in nice

agreement with the experimental results. Both TS diastereoisomers show distinct and different

CCS values, explaining their distinct experimental drift times. The Quercitin rutinoside show an

increased CCS value at 160 A2, in agreement with the increased drift time observed in

comparison to the two diastereoisomers.

O

HOOH

OHHO

HO

OH

OH

OH

O

OH

HO

HO

OH(154A2)

O

HOOH

OHHO

HO

OH

OH

OH

O

OH

HO

HO

OH(157A2)

O

OOH

HO

O

OH

OH

OHH

O

O

HOOH

O

H3CHO

HO OH

(160A2)

Figure 5.5 The isomeric structures of theasinensins and Quercetin-3-O-rutinoside with their characteristic

collisional cross sections (A2) correlated with their mobility drift times D.T. (ms) acquired: a) trans theasinensin :

mobility D.T. 5.18 ms having collisional cross section of 154 A2, b) cis- theasinensin: mobility D.T. 5.56 ms having

collisional cross section of 157A2, c) Quercetin-3-O-rutinoside: mobility D.T. 5.67 ms and having cross section of

160 A2

161


Spectrometry

On the other hand, the presence of Proanthocyanidins PAs was identified in Ziziphus Spinae

Christi species. Extracted ion chromatograms showed the presence of a compound at m/z

609.167 having retention time of 1.67 min. acquiring only one specific mobility drift time at 5.35

ms. Its characteristic fragment spectrum confirmed that this structure was a Proanthocyanidin

PA. Furthermore, Quercetin-3-O-glycoside was also observed in the Ziziphus Spinae Christi

extract showing a chromatographic peak at 10.22 min. and acquiring a single specific mobility

drift time of 5.62 ms and this is shown in figures 5.6 &5.7.

Figure 5.6 Quercitin-3-O-rutinoside in Ziziphus Spinae Christi showing only one clear signal in the drift scope

obtained from multiple chromatographic peaks

Sudanese Plant_frag_01.raw : 4

Sudanese Plant_frag_01.raw : 4

162


Spectrometry

Figure 5.7 3D representation of the IMS-MS experiment obtained from in the Ziziphus Spinae Christi; where the

chromatographic peaks for Quercitin-3-O-rutinoside at retention times of 10.22 and 10.36 min. showed only one

signal in the drift scope

Furthermore, table 5.1 shows proanthocyanidins PAs which were identified in the

Rhododendron species and these were confirmed by their characteristic fragment spectrum in the

MSE experiments. These two components were detected at different retention times of 2.35 min.

and 4.98 min. exhibiting in the IMS/MS drift scope only one signal for every chromatographic

peak. However, dramatically different discrete mobility drift times at 5.24 ms and at 8.37 ms

were observed for the two signals corresponding to two isomeric proanthocyanidins, as assigned

by their fragment spectra. This indicates that positional isomerism may exist, where these are

163


Spectrometry mainly present in two forms as epigallocatechin-(4,8)-epigallocatecin and epigallocatechin-

(4,6)-eigallocatechin in Proanthocyanidins B-type. The computational calculations for their

collisional cross section showed a discrete difference in their cross section areas; thus confirming

our experimental data obtained. Epigallocatechin-(4,8)-epigallocatechin showed a collisional

cross section of 154 A2, where as epigallocatechin-(4,6)-epigallocatechin showed a collisional

cross section of 178 A2. And this is representative and in agreement with the experimental data

obtained. The identified isomeric structures of proanthocyanidins are exhibited in figure 5.8

O

HOOH

OH

HO

HOOH

O

OH

HOOH

OH

OH

HO

epigallocatechin-(4,8)-epigallocatechin

Proanthocyanidine (154A2)

OHO

OH

HO

HO

OH

OH

OH

O

OH

HO

HO

OH

HO

Proanthocyanidineepigallocatechin-(4,6)-epigallocatechin

(178A2)

Figure 5.8 The isomeric structures of Proanthocyanidins corresponding to their collisional cross section areas (A2):

a) epigallocatechin-(4,8)-epigallocatechin with collisional cross section 154A2 having mobility drift time of 5.23 ms

b) epigallocatechin-(4,6) epigallocatechin having collisional cross section of 178 A2 and mobility drift time 8.37ms

164


Spectrometry

IMS-MS allows therefore the identification of linkage regioisomerism in the class of PAs. This is

a significant finding since MS2 spectra of 4,8- and 4,6- PAs are almost identical and do not allow

differentiation between the important classes of phenolic secondary metabolites.

All the above mentioned isomeric structures characterized by their collisional cross sections (A2)

and corresponding to their acquired mobility drift times are exhibited in table 5.2.

In order to compare CCS values from computational work and experimental drift times, both

parameters were plotted against one another in figure 5.9. It can be clearly seen from this graph

that a nice correlation exists between drift time and theoretical CCS, which according to the

literature follows the following mathematical relation Dt ~{CCS/ Q}, where Q is the charge

distributed and CCS is the collisional cross section of the compound.71

Figure 5.9 Correlation between the calculated collision cross section (ºA2) of the isomeric structures with their

acquired mobility drift times (ms)

165


Spectrometry

Table 5.1 The different isomeric components of theasinensin and proanthocyanidin identified from

thearubigin fractions in black tea and botanical sources from Rhododendron and Ziziphus spinae Christi;

the conformation of these structures was obtained by the MSE collision mode showing their characteristic

fragment ions and the neutral losses; each isomer attained at a chromatographic peak (min.) acquired its

specific mobility drift time (ms)

Parent ions

(m/z)

R.T. (min)

D.T. (ms.)

MSE Collision mode

(m/z)

Mass losses

[Da]

Assam Tea (TRs)

Theasinensins 609.156 1.28 5.18 471.163 138

609.157 5.56 591.129, 471.123, 303.067 18 , 138, 306

609.178 10.09 5.67 300.042, 301.050 309, 308

609.178 10.22 5.67 300.043, 301.050 309, 308.

Sri Lanka tea (TRs)

Theasinensins 609.159 1.27 5.18 471.118, 289.045 138, 320

609.162 5.51 591.154, 471.120 18, 138

609.183 10.13 5.62 300.047, 301.050 309, 308

609.183 10.25 5.62 300.047, 301.053 309, 308

Ziziphus Spinea Christi Proanthocyanidins B-

Type 609.167 1.67 5.35 483.121, 441.110, 423.096, 305.085 126 , 168, 186, 304

609.159 2.31 483.125, 441.106, 423.097, 305.084 126, 168, 186 , 304

609.157 2.35 441.105, 423.099, 305.084 168, 186, 304

609.185 10.22 5.62 300.043, 301.051 309, 308

Rhododendron

Proanthocyanidins B-Type

609.158 2.35 5.24 441.113, 423.097, 305.087,125.033 168 , 186, 304

609.169 4.98 8.37 441.110, 423.103, 305.088 168 , 186, 304

609.185 10.4 5.62 300.045, 301.051 309, 308

166


Spectrometry

Lab. synthesized

Dimeric structures .

609.165 2.91 5.61 484.612, 423.1, 305.089 124, 186, 304

609.169 7.74 591.158, 565.177, 441.111, 331.069,

305.089 18, 44, 168, 278, 304

609.163 7.06 535.407, 471.122, 305.088 74, 138 , 304

609.170 9.11 5.61 591.157, 573.141, 565.173, 349.079, 331.071, 305.088 18, 36, 44, 260, 278, 304

609.169 9.49 5.13 591.157, 565.175, 349.080, 331.070, 305.089 18, 36, 44, 260, 278,304

167


Spectrometry

Table 5.2 The different dimeric isomeric structures corresponding to Theasinensins, Rutin, and

Proanthocyanidin with their characteristic cross sections (A2) and the mobility drift times (ms) acquired

Drift time (ms.) Collisional cross section (A2)

A) Thearubigin fractions

Theasinensins (TS); trans 5.18 154

(TS); cis 5.56 157

Quercetin -3-O-rutinoside 5.67 160

B) Rhododendron species

Proanthocyanidins B-Type (PA)EGC-(4,8)-EGC 5.23 154

(PA) EGC-(4,6)-EGC 8.37 178

Quercetin -3O-rutinoside 5.62 160

168


Spectrometry

5.3 Conclusions

Our studies demonstrated that UPLC when coupled to IMS-MS mass spectrometry is a suitable

technique for complex mixture analysis such as the black tea thearubigins. With the aid of this

advanced instrumental analysis we were able to separate and identify different isomeric

components in the complex mixture which could previously not be differentiated by HPLC

tandem mass spectrometry only. UPLC-IMS/MS offered the detection of the different isomeric

structures in a very short period of time when compared with HPLC/MS mass spectrometry.

In this work, I studied the difference between isomeric structure Theasinensins,

Proanthocyanidins B-type, and Rutin (Quercetin-3O-rutinoside) and these are present abundantly

in plant sources. I was able to differentiate between these structures according to their acquired

mobility drift times, which differ from the traditional investigations in mass spectrometry.

Calculation of theoretical collisional cross sections allowed assignment of individual isomeric

structures, and these were found to be in good agreement with the experimental results.

The present work demonstrates UPLC-IMS-MS as an efficient technology for isolating and

separating isobaric and isomeric structures existing in complex mixtures discriminating between

them according to their characteristic fragment ions and mobility drift times. Different phenolic

secondary metabolites of varying structures could be successfully differentiated by IMS/MS,

where the calculated collisional cross sections can be correlated with experimental drift times.

Therefore, a rational assignment of isomeric structures based on the ion mobility data might be

useful in mass spectrometry based structure analysis in the future.

169

MALDI imaging of the polyphenolic components present in tea leaves

Chapter-6 MALDI imaging of the polyphenolic components present in tea

leaves

Tea is made from the dried fresh (green tea) or enzymatically oxidized (black tea) young buds

and leaves of varieties of Camellia Sinensis plant. Many chemical studies have been carried out

on green and fermented tea leaves and were focused on profiling the phenolic components

present.122,123 In the previous studies both conventional and advanced powerful mass

spectrometrical methods were employed, and these included mainly electrospray ionization

coupled to liquid chromatography (ESI/LC-TOF/MS), LC/MSn mass spectrometry, matrix

assisted laser desorption/ time of flight mass spectrometry (MALDI-TOF), Fourier Transform

Ion Cyclotron Resonance mass spectrometer (FTICR-MS), providing accurate elemental

composition and structural data.87,98,124,125 However, all these studies were performed on tea

extracts and fractions following different preparation steps for the sample prior to mass

spectrometric analysis and thus cannot provide further information on the precise localization of

the main and important secondary metabolites present in a green tea leaf.

Furthermore, locating the main polyphenolic components and studying their spatial distribution

on a green tea leaf may assist industrial processing for tea production. However, to the best of

my knowledge, there is no literature till now describing the detection and imaging of compounds

170


on the surface of a green tea leaf. MALDI imaging mass spectrometry (MSI) (www.maldi-

msi.org) is an interesting tool that provides a means of resolving in, intact tissue sections, both

spatial distribution and their relative abundance of many kinds of molecules allowing molecular

map of all species to be achieved on the micrometer scale.126 The versatility of this method has

been demonstrated more widely in clinical investigations; for example Caprioli’s group has

carried out MALDI imaging of proteins and peptides from thin sections of human and rat

tissues.127 It has been applied for cancer disease research by distinguishing the lipids and proteins

profiles of healthy tissues and cancerous ones,128 and for pharmaceutical research for studying

drugs and metabolites.129 Recently, this methodology has been used in plant research but still

limited e.g. profiling amino acids, sugars and phosphorylated compounds from wheat seeds130,

and fatty acids and flavonoids from fruits.131 However, in spite of the development of the matrix

deposition improvements, optimization of the experimental methods including improvements in

sample preparation and matrices of choice is still a challenge for scientists.

The aim of my work is to develop an optimized experimental method for imaging green tea leaf.

In this chapter two parts of work were performed. The first includes imaging a roasted green tea

leaf using an optimized matrix of choice for imaging. The second part included imaging on a

fresh green tea leaf without the use of a matrix for the first time, where optimizing of the

instrumental conditions were necessary. In the second part, the complexity of the biochemical

processes occurring in the green tea leaf plant tissue has been determined for the first time.

Furthermore, the main polyphenolic components were successfully located in both roasted tea

leaf and a very young fresh green tea leaf.

171



6.1.1 Consideration for MALDI-MSI imaging

MALDI-MSI imaging requires method development and optimization and this includes: dried

droplet sample preparation when selecting the most appropriate MALDI matrix, its

concentration, appropriate pH for crystallization and different instrumental settings.

Firstly during matrix application the incorporation of the analyte into the matrix crystals with

defined size is important for MALDI-MS imaging. The matrix of choice, its concentration, and

its deposition are crucial steps to be determined. Matrix crystals with small size can be obtained

if nearly saturated solution of matrix arrives homogeneously to the surface of the plant tissue.

The solvent normally extracts the compounds from the leaf tissue and during evaporation the

compounds are incorporated and co-crystallized with the matrix applied. The extraction of the

compound depends on the solvent composition and the time it remains on the tea leaf tissue to be

analyzed without causing delocalization of the compounds. Optimization of the matrix

deposition conditions to obtain small matrix crystals with small defined size and limited

delocalization of compounds is a crucial step for imaging experiment and this was overcome by

using an electrospray probe with a syringe pump under controlled gas pressure, where this

experimental part is more illustrated in Part I of this chapter. However, I developed a new

method without the use of a matrix and the optimization was mainly based on acquiring the right

instrumental conditions in MALDI-MS imaging experiments and this is described in Part ΙI of

this chapter.

The second aspect for MALDI-MS imaging is the desorption effect since it directly influences

the spatial resolution. If the desorbed area becomes smaller by using lower laser energy, the step

172


size of the sampled spots can be decreased to obtain a higher resolution, however the sensitivity

of the detected ions will decrease. A balance needs to be found between the sensitivity and the

spatial resolution within each MALDI-MSI imaging experiment and this can be achieved by

optimizing the laser fluence thus increasing the spatial resolution without decreasing the

sensitivity. Furthermore, several parameters need to be considered that affect the sensitivity of

the instrument such as the voltages in the ion source and the acquisition mode in the TOF

detector. (Such experimental conditions were optimized and described in Part Ι and Part II of this

chapter).

Part I


6.2.1 Leaf sample preparation

A roasted slightly fermented green tea leaf was coated with α- HCCA (6mg / 1ml dissolved in

methanol: H2O (7:3) / 0.1 % of trifloroacetic acid TFA). For homogeneous matrix deposition, the

tea leaf was sprayed with the methanolic matrix solution, and this was done by a homebuilt

pressure driven deposition device. This matrix spotter consists of a movable target plate holder,

an electrospray probe with sheath gas and a syringe pump. For optimal deposition, the flow rate

of the matrix solution was set at 6 ml / hour via a gas tight syringe. Nitrogen gas was used as

nebulizer gas and its pressure was set at 5 bar. The slide was mounted on a movable stage and

during matrix deposition the stage was manually moved for homogeneous deposition. The tea

leaf was left to dry under vaccum conditions, and then was attached on a double sided adhesive

173


tape on a MALDI target and then loaded into the MALDI source and the desired area was

imaged.

6.2.2 MALDI imaging method and software

A photo image was created, MALDI imaging pattern creator software (Waters) was used to select

the area to be sampled from the photo image. All locations were stored inside a pattern size and

mass spectra were acquired from each defined location. For acquisition a good resolved mass

spectra (HDMS) synapt G2 system (Waters) was used. The MALDI source was equipped with

nitrogen gas having a wavelength of 337 nm, operating at 200 Hz. The mixture was calibrated by

using sodium iodide as a Calibrant in the negative ion mode and choosing the mass range

between 100-2000Da. The MALDI source parameters were optimised using a sample plate,

extraction cone and hexapole voltages to be set at 0, 10, and 12 respectively. The laser energy

was set to be at 170 (j.cm-2s).

6.2.3 Data processing and visualisation

All mass spectra acquired during imaging experiments were combined into a image file of

several gegabytes in size. The mass spectra of each spatial resolution were binned with MALDI

imaging converter software (Waters). This was required for further image processing with the

Biomap Software. At the end an image of the intensity of each spatial location was created.

174



I investigated the presence and the distribution of the main polyphenolic components on one part

of a roasted slightly fermented tea leaf by using MALDI-TOF imaging when sprayed with

HCCA matrix providing a very uniform deposition. Figure 6.4 shows the comparison between

the normal optical microscopy and the MALDI imaging when deposited with the matrix layer.

Negative ion mode was selected and the ions corresponding to the selected phenolic secondary

metabolites were clearly located. The experimental protocol was very sensitive yielding mass

spectra with very clear signals [M-H]- pseudomolecular ions. The identity of the identified

compounds in MALDI-TOF mass spectra was confirmed by the accurate mass measurements of

their respective [M-H]- ions. The detectable peaks were: epicatechin C15H13O6 at m/z 289,

epigallocatechin C15H13O7 at m/z 305, theaflavin C29H23O12 at m/z 563, epigallocatechin gallate

C22H17O11at m/z 457, theasinensin and proanthocyanidins (epigallocatchin / epicatechin dimers )

C30H25O14 at m/z 609, C30H25O13 at m/z 593, C30H20O13 at m/z 588, dehydrotheaflavin C29H19O13

at m/z 575. These compounds were previously reported to be the main polyphenolic components

present in tea extracts.122 Their identified mass spectra are exhibited in figures 6.1, 6.2 & 6.3.

175


Figure 6. 1. Expanded region of the MALDI mass spectrum revealing the presence of thr epicatechin and

epigallocatechins in the negative ion mode

Figure 6.2. Expanded region for the MALDI mass spectrum showing the presence epigallocatechin gallate at m/z

457 in the negative ion mode

tea leaf sprayed with hcca 6 mg/ml 150 um spot

m/z280 285 290 295 300 305 310 315 320 325

%

0

10018042013-tea leaf_HCCA_Test 1 1349 (54.215) Sm (SG, 2x3.00); Cm (1274:1373) 1: TOF MS LD-

1.00e6287.0858

285.0451

284.0422

300.0355

289.1041

299.0292

295.0661293.0496

311.0695301.0434

305.0693

306.0780

316.0314

315.0840317.0348

323.0645317.0970

318.0475325.0477


m/z430 432 434 436 438 440 442 444 446 448 450 452 454 456 458 460 462 464 466

%

0

10018042013-tea leaf_HCCA_Test 1 1349 (54.215) Sm (SG, 2x3.00); Cm (1274:1373) 1: TOF MS LD-

1.00e6

456.1294431.1057

443.1059

432.1320

433.1140 437.1080

455.1123447.1087 454.1149

457.1289463.1022

459.1126465.0894

C15H13O6

C15H13O7

C22H17O11

176


Figure 6.3.Expanded region of the MALDI-MSI spectrum showing the presence of theaflavin abundantly at m/z 563

and revealing the presence of dehydrotheaflavin at m/z 575, proanthocyanidins at m/z 588, m/z 589 and at m/z 593,

& proanthocyanidins / theasinensins at m/z 609 in the negative ion mode

i) Optical microscopy for smaller crystals appearing where the matrix is totally embedded within the tea leaf

(resolution 1300 µm)


m/z545 550 555 560 565 570 575 580 585 590 595 600 605 610

%

0

10018042013-tea leaf_HCCA_Test 1 2095 (84.153) Cm (2055:2146) 1: TOF MS LD-

1.00e6563.1590

549.0257

544.1097

550.0300

560.0924551.0250

552.0311

588.1097

564.1554

575.0687565.1630

566.1091576.0651

581.1651

604.0874589.1068

593.1670

603.1744

594.1682

601.1684

609.1587

C29H23O12

C30H25O14 C30H25O13

C30H20O13

C29H19O13

177


Figure 6.4. ii) MALDI-MSI imaging for one of the coated tea leaf showing the distributions of the identified

compounds a) m/z 287, b) m/z 289 epicatechin, c) m/z 305 epigallocatechin, d) m/z 563 theaflvin, e) m/z 457

epigallocatechin gallate, f) m/z 593, g) m/z 609 of proanthocyanidins / theasinensins, h) m/z 575 of

dehydrotheaflavin. Spatial resolution (150 µm) in the negative ion mode

The MALDI-MSI imaging in figure 6.4 confirms the presence and the spatial distribution of the

important polyphenols in a roasted tea leaf. The results show that most phenolic compounds are

unevenly distributed within the tea leaf. However, (epi)catechin and (epi)gallocatechin were

present in high concentration and were found in the same location in the tea leaf. Interestingly,

a b c

d e f

g h

178


theaflavin was also observed to be located evenly with both epigallocatechin and epicatechin

acting as its main building blocks. Theasinensins and / or proanthocyanidins were also identified

in good concentration; however these were distributed unevenly in the roasted tea leaf.

Such studies have not been conducted before thereby adding a whole new layer of information to

the conventional analysis. An understanding of the spatial distribution of the main constituents in

the tea leaf and their metabolites will have a significant impact on the elucidation of metabolic

network in the tea plants. However, a particular disadvantage of the organic matrices is the large

number of the low molecular weight background peaks induced, these derived from the gas

phase protonation/ deprotonation, fragmentation and cluster formation which was observed

mostly in HCCA matrices.

Further improvement and modification in sample preparation is still required and this is

discussed in the following section of this chapter.

Part II


A very small young green tea leaf was picked freshly from Camellia Sinensis plant and then

directly mounted on an adhesive tape fixed on a target plate (no matrix is used). The target which

is moved within the source of MALDI TOF instrument is then irradiated by a fixed laser beam

for the whole small tea leaf. Since the irradiation was done without the use of a matrix, in this

experiment I decreased the laser energy at 100 Jcm-2s considering the direct ablation for the spots

irradiated, in order to have a good spatial resolution. And this generated a mass spectrum specific

to the region of the tea leaf irradiated by the laser beam having a spatial resolution of 100 µm.

Measurements were done in the positive ion mode using polyethylene glycol as a calibrant of

179


choice. The mass range acquired was in the region detected between 100-1000Da. The MALDI

source parameters were optimised using a sample plate, extraction cone and hexapole voltages to

be set at 0, 10, and 12 v respectively. The data set was displayed as an average spectrum where

the selected images of the individual m/z values were extracted, displaying their distribution

based on the main m/z values obtained in the fresh green tea leaf.


In this section a fresh tea leaf was imaged in the absence of a matrix. It was assumed that since

most common MALDI matrices are phenolic compounds, a high concentration of phenolic

secondary metabolites within the leaf might already serve as an internal matrix.

By performing matrix free imaging experiments, the results revealed the presence of the main

components in fresh green tea leaf having a mass range till 600 Da. It is worth to note, that here

the age of the green tea leaf was still very young. In this part, I was able to identify in the

positive ion mode the presence of (epi)gallocatechins (EGC) and its oxygenated component at

m/z 323 C15H15O8, undergoing hydrogenation process at m/z 325. Furthermore, the presence of

its methylated derivative was clearly identified at m/z 337.2112 C16H17O8 (EGC+O+CH3).

Furthermore, the presence of galloyl quinic acid +H+ was also detected at m/z 345 (C14H17O10),

undergoing hydrogenation at m/z 347.1980, methylation (+CH3) at m/z 359.1990 (C15H19O10),

and carboxylation (+ CO2) at m/z 391.2764 (C15H19O12) and these results are exhibited in figure

6.5.

180


Figure 6.5 Mass spectrum of a very small fresh tea leaf directly picked directly from the plant revealing the

presence of: a) m/z 323 C15H15O8, m/z 325.2083 C15H17O8, m/z 337.1980 (C16H17O8), b) m/z 345.1825 (C14H17O10),

m/z 347.1980 (C14H19O10), m/z 359.199(C15H19O10) in the positive ion mode.

325.2083

323.1949

337.2112

327.2279 331.1746

333.1967

347.1980

345.1825

341.1852

359.1990

349.2031 357.1820

391.2764

3052013 Tea leaf _6 610 (25.705) Cm (26:2286) 8.80e5

without matrix

m/z 320 325 330 335 340 345 350 355 360 365 370 375 380 385 390 395

%

0

100 TOF MS LD+

a

b c

181


Figure6.6 MALDI-MSI imaging for the young green tea leaf revealing the distribution of the different components:

a) m/z 212, b) m/z 323 (EGC+O), c) m/z 337 (EGC+O+CH3), d) m/z 345(galloyl quinic acid), e) m/z 391(galloyl

quinic acid+CO2), f) m/z 359 (galloyl quinic acid +CH3)

On the other hand, some low molecular weight components were appearing in the mass range

between m/z 170 and 290; specifically when examining the component at m/z 212, it was clearly

identified in the selected images, when the m/z values were individually extracted (figure 6.6 a);

this can be explained that new fragment ions were obtained due to the direct absorption of laser

energy that is exceeding the bond energies of the constituents present in the green tea leaf

causing new fragmentation.

In this section again a heterogeneous distribution of the phenolic constituents was observed in

the tea leaf. The chemical processes occurring that we could propose here are based on the m/z

ions identified in the positive mode and the mass losses occurring. These compounds were

previously identified in the literature122 and now clearly identified in the imaging part, but

however they are present in a low concentration when distributed in the whole tea leaf and this

can be attributed to the very small age of the tea leaf, where all the imaging are exhibited in

figure 6.6. It is worthy to note that I was able to identify that the middle part of the leaf was

d f e

182


found to contain the highest level of the identified components and these are exhibited in the

MSI imaging part and indicated by a dashed arrow in figure 6.7.

Figure 6.7 The most concentrated part of the leaf is observed in the above imaging (indicated by the dashed arrow)

where all the identified components are obtained in this part of the leaf at higher concentrations

183


In this part, I was able to prove that MALDI imaging can be performed without the use of matrix

in a plant tissue by modifying the laser energy, and this was done successfully for the first time,

which may save preparation time and costs for researchers in the future in plant science.

184

Beyond black tea composition- Health benefits effect of black tea thearubigin fractions

Chapter-7 Beyond black tea composition- Health benefits effect of black tea

thearubigin fractions

Tea polyphenols have recently showed much attention because they possess a broad spectrum of

biological functions.132,133,134,135,136 The main important properties of black tea are its high

antioxidant activity,137,138,139,140 which plays an important role in scavenging free

radicals,141,142,143,144 thus providing benefits in pathological situations associated with many

diseases such as hypertension and cardiovascular disease.145,146,147 On the other hand, they can

exert prooxidant effects by catalyzing DNA degradation in the presence of transition metal ions

through Haber-Weiss Fenton type reactions,148,149,150 possessing chemopreventive and therapeutic

properties against cancer.151,152,153 Such properties were confirmed in many in vitro studies,154,155

however; further in vivo studies156 based on the metabolism processes and excretion of these

polyphenolic compounds should be more addressed and studied in the future.

My investigations in structural elucidation of black tea revealed their chemical structures to

possess significant hydroxylated derivatives present in the homologous series of many

compounds, and these in turn allow black tea specifically to possess high antioxidant properties

for scavenging free radicals. On the other hand, theaflavin, an important precursor for

thearubigins formation, was identified to undergo further oxidative coupling reactions giving a

series of heterogeneous structures of new TRs components, which was shown and confirmed in

185


my research in identifying higher molecular weight components exceeding 1000 Da. Such

chemical structures and composition of the novel identified black tea components demonstrate

that these exhibit highly significant amphiphilic properties. Hence thearubigin fractions could be

used as antibacterial and antibiotic agent. Therefore, I was interested in investigating the health

benefits of the black tea SΙΙ thearubigins fractions for studying their inhibitory effect in both

gram positive and gram negative bacteria


A thearubigin fraction SΙΙ was prepared from Assam Ceylon tea according to the caffeine

precipitation method.31,11 2 mg of this fraction was dissolved in 500µl methanol/water at different

ratios (1:1) and at (7:3). Then these were injected in (LB media 2) nutrient plates with grown: A)

gram positive bacteria containing “Bacillus Subtilis” and B) gram negative containing

“Escherichia Coli” bacteria. These were prepared at 37ºC and later the experiments were

performed after 24 hours for measuring the inhibition zone.


The results showed that SΙΙ thearubigin fraction showed a good inhibitory activity in both gram

positive and gram negative bacteria and these are exhibited in both figures 7.1& 7.2.

186


Figure 7.1 Black tea thearubigins dissolved in CH3OH/ H2O solution at rations (1:1 and 7:3) and injected in a

nutrient containing gram positive bacteria (Bacillus Subtilis), showing an inhibitory activity of 0.6 cm

Figure 7.2 Black tea thearubigins dissolved in CH3OH/ H2O solution at rations (1:1 and 7:3) and injected in a

nutrient containing gram negative bacteria (E. Coli), and showing an inhibitory activity of 0.65 cm

187


7.3 Conclusion

I was able to investigate that black tea SΙΙ thearubigin fraction possesses good inhibitory activity

in both gram positive “Bacillus Subtilis” and gram negative “Escherichia Coli” bacteria. This can

be ascribed to the physicochemical properties of thearubigins (TRs) possessing both hydrophobic

and hydrophilic parts within their structures, where they can interact with both types of bacteria

(gram positive and gram negative) causing their inhibition activity. Furthermore, the presence of

the hydrogen peroxides during the enzymatic oxidation can play an important role in the

inhibitory effect of TRs as well. These properties revealing the highly beneficial role may

indicate that one can use black tea active conserves/isolates in many food as well as in drug

applications such as anti-inflammatory creams and expect lot of potentials in terms of health

benefits.

Additionally, this experiment indicates and suggests that a bacterial infection on a tea plant e.g.

when infection occurs in nature of a typical plant pathogen such as “Bacillus Subtilis” on the tea

plant (Camellia Sinensis), the cellular structure of the tea plant is normally damaged, thus

producing polyphenol oxidized products like TRs in tea fermentation, where these are

characterized with their antibacterial activity. Thus, thearubigins can serve in nature as defense

against microorganisms.

188

Conclusions

Chapter-8 Conclusions

A combination of advanced complementary mass spectrometric techniques have been employed

for an unprecedented high level of characterizing thearubigins in black tea, one of the most

complex mixtures which remained a challenge for the scientist until recently.

By employing high resolution mass spectrometry, ESI-LC-MS and tandem MSn spectrometry

both the identification and structural elucidation of previously described TR components were

allowed, where new members of TRs in the class of homologous series of theanaphthaquinone

and theasinensin C dimeric structures were identified. The structures of the series of these

dimeric compounds were consistent with the oxidative cascade hypothesis previously

formulated, in which polyhydroxylated dimers of catechins are formed by nucleophilic addition

of water to the quinone oxidized counterparts of catechins. On the other hand, this approach

allowed the identification of new class of TR components in the series of theasinensin A, which

were rationalized due to the presence of hydrogen peroxide and its contribution in the

nucleophilic addition reaction, thus forming new TR components of a different structural class,

adding novel reaction mechanisms and schemes to the chemistry of TRs. By employing both

MALDI-TOF and ESI tandem mass experiments, a series of novel constituents of the

thearubigins with molecular weights exceeding 1000 Da were identified, and these included a

series of trimeric and tetrameric structures of flavan-3-ols subunits. Furthermore, these

experiments revealed the heterogeneity of the molecular ions present in a complex mixture,

where the collision energy applied in the direct infusion experiments were used for the first time

for identification of different isomeric compounds, as a dimension for compound separation.

189

Conclusions

MALDI-TOF experiments confirmed the presence of oligomeric structures of these compounds,

where they showed to be formed by oxidative coupling pathway between most notably

theaflavins dimeric structures and the TPPO oxidized flavan-3-ols, thus confirming the

oligomerisation of catechins (flavan-3-ols) as previously postulated by Kuhnert rather than

polymerization. The reaction pathway and the regiochemistry was suggested for the first time by

performing computational calculations for the structural subunits acting as the main building

blocks for the higher molecular weight compounds formed, and these were based on the density

functional theory. The aspect behind the thearubigin formations were clearly hypothesized, when

performing potentiostatic measurements on one of the main polyphenolic constituents in green

tea leaf. The oxidation was confirmed both experimentally and theoretically that it is more

favored on B-ring in black tea formation rather than A- ring as in other dietary products such as

chocolate and red wine, where theaflavins and theasinensins were found the main primary

oxidation products, acting as the main precursors for further thearubigin formation products. The

oxidation theory was further explained and confirmed by the computational analysis work based

on density functional theory. The data showed clearly the presence of multiple isomers,

explaining the unusual chromatographic behavior of the thearubigin fractions. Highly advanced

mass spectrometry strategy based on UPLC-IMS-MS mass spectrometry was employed for the

separation and identification of isomeric structures in complex mixtures, and this was

successfully proved. This strategy allowed the differentiation of isomeric structures according to

their acquired mobility drift times, which differ from the traditional investigations in mass

spectrometry. Calculation of theoretical collisional cross sections allowed the assignment of

individual isomeric structures, and these were found to be in good agreement with the

experimental results. The latter strategy showed the UPLC-IMS-MS as an efficient technology

190

Conclusions

for isolating and separating isomeric structures existing in complex mixtures. Therefore,

structural analysis and characterization for many isobaric and isomeric components present in

different complex mixtures found in dietary products could be done in the future. MALDI

imaging mass spectrometry was employed for the first time on green tea leaf, thus allowing the

identification of the main polyphenolic components, their spatial distribution, and the

biochemical processing occurring on a fresh green tea leaf. A matrix free method was developed

for the first time, serving less costs and time for the researchers in the future. The benefits of

thearubigin fractions in black tea were also elaborated by investigating its antibacterial inhibitory

activity, where thearubigins and black tea extract can be used in medical applications in the near

future. These data have provided inspiration for the chemistry of black tea, which allowed the

development of new reaction mechanisms and structural hypothesis, and confirming previously

postulated hypothesis by Kuhnert. The data showed the ultimate diversity of thearubigins as a

complex mixture in unraveling the mystery of black tea.

191

Reference List

1. Sanderso.Gw; Graham, H. N., Black Tea Aroma and Its Formation. Abstr Pap Am Chem S 1972,

164 (Aug-S), 69-&.

2. Awad, M. A.; de Jager, A., Formation of flavonoids, especially anthocyanin and chlorogenic acid

in ‘Jonagold’ apple skin: influences of growth regulators and fruit maturity. Scientia Horticulturae 2002,

93 (3–4), 257-266.

3. Grassi, D.; Ferri, C., Chapter 78 - Cocoa, Flavonoids and Cardiovascular Protection. In

Polyphenols in Human Health and Disease, Watson, R. R.; Preedy, V. R.; Zibadi, S., Eds. Academic

Press: San Diego, 2014; pp 1009-1023.

4. Kim, D.-O.; Jeong, S. W.; Lee, C. Y., Antioxidant capacity of phenolic phytochemicals from

various cultivars of plums. Food Chemistry 2003, 81 (3), 321-326.

5. Crozier, A.; Jaganath, I. B.; Clifford, M. N., Dietary phenolics: chemistry, bioavailability and

effects on health. Nat Prod Rep 2009, 26 (8), 1001-1043.

6. Tomas-Barberen, F. A.; Clifford, M. N., Flavanones, chalcones and dihydrochalcones - nature,

occurrence and dietary burden. J Sci Food Agr 2000, 80 (7), 1073-1080.

7. Drynan, J. W.; Clifford, M. N.; Obuchowicz, J.; Kuhnert, N., The chemistry of low molecular

weight black tea polyphenols. Nat Prod Rep 2010, 27 (3), 417-462.

8. Biedrich, P.; Engelhardt, U. H.; Herzig, B., Determination of Quercetin-3-O-Rutinoside in Black

Tea by Hplc. Z Lebensm Unters For 1989, 189 (2), 149-150.

192

9. Lakenbrink, C.; Lapczynski, S.; Maiwald, B.; Engelhardt, U. H., Flavonoids and other

polyphenols in consumer brews of tea and other caffeinated beverages. J Agr Food Chem 2000, 48 (7),

2848-2852.

10. Shao, W. F.; Powell, C.; Clifford, M. N., The Analysis by Hplc of Green, Black and Puer Teas

Produced in Yunnan. J Sci Food Agr 1995, 69 (4), 535-540.

11. Kuhnert, N., Unraveling the structure of the black tea thearubigins. Arch Biochem Biophys 2010,

501 (1), 37-51.

12. Roberts, E. A. H., The Fermentation Process in Tea Manufacture .10. The Condensation of

Catechins and Its Relation to the Chemical Changes in Fermentation. Biochem J 1949, 45 (5), 538-542.

13. Roberts, E. A. H., The Chemistry of Tea Fermentation. J Sci Food Agr 1952, 3 (5), 193-198.

14. Tanaka, T.; Watarumi, S.; Matsuo, Y.; Kamei, M.; Kouno, I., Production of theasinensins A and D,

epigallocatechin gallate dimers of black tea, by oxidation-reduction dismutation of dehydrotheasinensin

A. Tetrahedron 2003, 59 (40), 7939-7947.

15. Graham, H. N., Green Tea Composition, Consumption, and Polyphenol Chemistry. Prev Med

1992, 21 (3), 334-350.

16. Roberts, E. A. H.; Wood, D. J., The Separation of Polyphenols in Tea Leaf by Paper

Chromatography with Water as a Mobile Solvent. Biochem J 1951, 49 (5), R33-R33.

17. Roberts, E. A. H.; Wood, D. J., Separation of Tea Polyphenols on Paper Chromatograms.

Biochem J 1953, 53 (2), 332-336.

18. Cartwright, R. A.; Roberts, E. A. H., Paper Chromatography of Phenolic Substances. Chem Ind-

London 1954, (45), 1389-1391.

193

19. Takino, Y.; Ferretti, A.; Flanagan, V.; Gianturc.M; Vogel, M., Structure of Theaflavin a

Polyphenol of Black Tea. Tetrahedron Lett 1965, (45), 4019-&.

20. Coxon, D. T.; Holmes, A.; Ollis, W. D., Theaflavic and Epitheaflavic Acids. Tetrahedron Lett

1970, (60), 5247-&.

21. Coxon, D. T.; Holmes, A.; Ollis, W. D.; Vora, V. C., Constitution and Configuration of Theaflavin

Pigments of Black Tea. Tetrahedron Lett 1970, (60), 5237-&.

22. Haslam, E., Thoughts on thearubigins. Phytochemistry 2003, 64 (1), 61-73.

23. Li, Y.; Shibahara, A.; Matsuo, Y.; Tanaka, T.; Kouno, I., Reaction of the Black Tea Pigment

Theaflavin during Enzymatic Oxidation of Tea Catechins. J Nat Prod 2010, 73 (1), 33-39.

24. Sang, S. M.; Tian, S. Y.; Meng, X. F.; Stark, R. E.; Rosen, R. T.; Yang, C. S.; Ho, C. T.,

Theadibenzotropolone A, a new type pigment from enzymatic oxidation of (-)-epicatechin and (-)-

epigallocatechin gallate and characterized from black tea using LC/MS/MS. Tetrahedron Lett 2002, 43

(40), 7129-7133.

25. Roberts, E. A. H., Phenolic Substances of Manufactured Tea .10. Creaming down of Tea Liquors.

J Sci Food Agr 1963, 14 (10), 700-&.

26. Powell, C.; Clifford, M. N.; Opie, S. C.; Ford, M. A.; Robertson, A.; Gibson, C. L., Tea Cream

Formation - the Contribution of Black Tea Phenolic Pigments Determined by Hplc. J Sci Food Agr 1993,

63 (1), 77-86.

27. Cartwright, R. A.; Roberts, E. A. H.; Wood, D. J., Theanine, an Amino-Acid N-Ethyl Amide

Present in Tea. J Sci Food Agr 1954, 5 (12), 597-599.

28. Roberts, E. A. H.; Wood, D. J., Oxidation of Catechol by Tea-Oxidase. Nature 1950, 165 (4184),

32-33.

194

29. Roberts, E. A. H., Oxidation-Reduction Potentials in Tea Fermentation. Chem Ind-London 1957,

(41), 1354-1355.

30. Roberts, E. A. H.; Wood, D. J., The Fermentation Process in Tea Manufacture .2. Oxidation of

Substrates by Tea Oxidase. Biochem J 1950, 47 (2), 175-186.

31. Roberts, E. A.; Smith, R. F., Spectrophotometric Measurements of Theaflavins and Thearubigins

in Black Tea Liquors in Assessments of Quality in Teas. Analyst 1961, 86 (101), 94-&.

32. Opie, S. C.; Robertson, A.; Clifford, M. N., Black Tea Thearubigins - Their Hplc Separation and

Preparation during Invitro Oxidation. J Sci Food Agr 1990, 50 (4), 547-561.

33. Opie, S. C.; Clifford, M. N.; Robertson, A., The Role of (-)-Epicatechin and Polyphenol Oxidase

in the Coupled Oxidative Breakdown of Theaflavins. J Sci Food Agr 1993, 63 (4), 435-438.

34. Opie, S. C.; Clifford, M. N.; Robertson, A., The Formation of Thearubigin-Like Substances by in-

Vitro Polyphenol Oxidase-Mediated Fermentation of Individual Flavan-3-Ols. J Sci Food Agr 1995, 67

(4), 501-505.

35. Davis, A. L.; Lewis, J. R.; Cai, Y.; Powell, C.; Davis, A. P.; Wilkins, J. P. G.; Pudney, P.; Clifford,

M. N., A polyphenolic pigment from black tea. Phytochemistry 1997, 46 (8), 1397-1402.

36. Harbowy, M. E.; Balentine, D. A., Tea chemistry. Crit Rev Plant Sci 1997, 16 (5), 415-480.

37. Balentine, D. A., The Chemistry of Tea and Its Manufacturing. Abstr Pap Am Chem S 1991, 202,

44-Agfd.

38. Sanderson, G. W., The Chemistry of Tea and Tea Manufacturing. In Recent Advances in

Phytochemistry, Runeckles, V. C.; Tso, T. C., Eds. Elsevier: 1972; Vol. Volume 5, pp 247-316.

195

39. Kuhnert, N.; Jaiswal, R.; Matei, M. F.; Sovdat, T.; Deshpande, S., How to distinguish between

feruloyl quinic acids and isoferuloyl quinic acids by liquid chromatography/tandem mass spectrometry.

Rapid Commun Mass Sp 2010, 24 (11), 1575-1582.

40. Kuhnert, N.; Drynan, J. W.; Obuchowicz, J.; Clifford, M. N.; Witt, M., Mass spectrometric

characterization of black tea thearubigins leading to an oxidative cascade hypothesis for thearubigin

formation. Rapid Commun Mass Sp 2010, 24 (23), 3387-3404.

41. Bruschi, L.; Copeland, E. L.; Clifford, M. N., Unexpected hyperchromic interactions during the

chromatography of theafulvins and simple flavonoids. Food Chem 1999, 67 (2), 143-146.

42. Kuhnert, N., What's under the Hill? Nachr Chem 2011, 59 (9), 866-871.

43. Crozier, A.; Del Rio, D.; Clifford, M. N., Bioavailability of dietary flavonoids and phenolic

compounds. Mol Aspects Med 2010, 31 (6), 446-467.

44. Clifford, M. N., Anthocyanins - nature, occurrence and dietary burden. J Sci Food Agr 2000, 80

(7), 1063-1072.

45. Catterall, F.; King, L. J.; Clifford, M. N.; Ioannides, C., Bioavailability of dietary doses of H-3-

labelled tea antioxidants (+)-catechin and (-)-epicatechin in rat. Xenobiotica 2003, 33 (7), 743-753.

46. Block, G.; Patterson, B.; Subar, A., Fruit, Vegetables, and Cancer Prevention - a Review of the

Epidemiologic Evidence. Nutr Cancer 1992, 18 (1), 1-29.

47. Hakim, I. A.; Alsaif, M. A.; Aloud, A.; Alduwaihy, M.; Al-Rubeaan, K.; Al-Nuaim, A. R.; Al-

Attas, O. S., Black tea consumption and serum lipid profiles in Saudi women: a cross-sectional study in

Saudi Arabia. Nutr Res 2003, 23 (11), 1515-1526.

48. Hakim, I. A.; Alsaif, M. A.; Alduwaihy, M.; Al-Rubeaan, E.; Al-Nuaim, A. R.; Al-Attas, O. S.,

Tea consumption and the prevalence of coronary heart disease in Saudi adults: Results from a Saudi

national study. Prev Med 2003, 36 (1), 64-70.

196

49. Catterall, F.; Copeland, E.; Clifford, M. N.; Ioannides, C., Contribution of theafulvins to the

antimutagenicity of black tea: their mechanism of action. Mutagenesis 1998, 13 (6), 631-636.

50. Catterall, F.; Copeland, E.; Clifford, M. N.; Ioannides, C., Effects of black tea theafulvins on

aflatoxin B-1 mutagenesis in the Ames test. Mutagenesis 2003, 18 (2), 145-150.

51. Sang, S. M.; Cheng, X. F.; Stark, R. E.; Rosen, R. T.; Yang, C. S.; Ho, C. T., Chemical studies on

antioxidant mechanism of tea catechins: Analysis of radical reaction products of catechin and epicatechin

with 2,2-diphenyl-1-picrylhydrazyl. Bioorgan Med Chem 2002, 10 (7), 2233-2237.

52. Bodewes, T. C. F.; Luttikhold, J.; van Stijn, M. F. M.; Visser, M.; van Norren, K.; Vermeulen, M.

A. R.; van Leeuwen, P. A. M., Antioxidative Properties of Flavonoids. Curr Org Chem 2011, 15 (15),

2616-2626.

53. Kuhnert, N., Chapter 29 - Chemistry and Biology of the Black Tea Thearubigins and of Tea

Fermentation. In Tea in Health and Disease Prevention, Preedy, V. R., Ed. Academic Press: 2013; pp 343-

360.

54. Quideau, S.; Deffieux, D.; Douat-Casassus, C.; Pouysegu, L., Plant polyphenols: chemical

properties, biological activities, and synthesis. Angewandte Chemie 2011, 50 (3), 586-621.

55. Feng, Q.; Torii, Y.; Uchida, K.; Nakamura, Y.; Hara, Y.; Osawa, T., Black tea polyphenols,

theaflavins, prevent cellular DNA damage by inhibiting oxidative stress and suppressing cytochrome

P450 1A1 in cell cultures. J Agr Food Chem 2002, 50 (1), 213-220.

56. Hodgson, J. M.; Proudfoot, J. M.; Croft, K. D.; Puddey, I. B.; Mori, T. A.; Beilin, L. J.,

Comparison of the effects of black and green tea on in vitro lipoprotein oxidation in human serum. J Sci

Food Agr 1999, 79 (4), 561-566.

57. Maity, S.; Ukil, A.; Karmakar, S.; Datta, N.; Chaudhuri, T.; Vedasiromoni, J. R.; Ganguly, D. K.;

Das, P. K., Thearubigin, the major polyphenol of black tea, ameliorates mucosal injury in trinitrobenzene

sulfonic acid-induced colitis. Eur J Pharmacol 2003, 470 (1-2), 103-112.

197

58. Nag Chaudhuri, A. K.; Karmakar, S.; Roy, D.; Pal, S.; Pal, M.; Sen, T., Anti-inflammatory activity

of Indian black tea (Sikkim variety). Pharmacological research : the official journal of the Italian

Pharmacological Society 2005, 51 (2), 169-75.

59. Rajavelu, A.; Tulyasheva, Z.; Jaiswal, R.; Jeltsch, A.; Kuhnert, N., The inhibition of the

mammalian DNA methyltransferase 3a (Dnmt3a) by dietary black tea and coffee polyphenols. Bmc

Biochem 2011, 12.

60. Mikutis, G.; Karakose, H.; Jaiswal, R.; LeGresley, A.; Islam, T.; Fernandez-Lahore, M.; Kuhnert,

N., Phenolic promiscuity in the cell nucleus - epigallocatechingallate (EGCG) and theaflavin-3,3 '-

digallate from green and black tea bind to model cell nuclear structures including histone proteins, double

stranded DNA and telomeric quadruplex DNA. Food Funct 2013, 4 (2), 328-337.

61. Sharma, V.; Rao, L. J. M., A Thought on the Biological Activities of Black Tea. Crit Rev Food Sci

2009, 49 (5), 379-404.

62. Kallithraka, S.; Bakker, J.; Clifford, M. N., Evaluation of bitterness and astringency of (+)-

catechin and (-)-epicatechin in red wine and in model solution. J Sens Stud 1997, 12 (1), 25-37.

63. Banerjee, S.; Mazumdar, S., Electrospray Ionization Mass Spectrometry: A Technique to Access

the Information beyond the Molecular Weight of the Analyte. Int J Anal Chem 2012.

64. Zhan, D. L.; Fenn, J. B., Electrospray mass spectrometry of fossil fuels. Int J Mass Spectrom

2000, 194 (2-3), 197-208.

65. Drynan, J. W.; Clifford, M. N.; Obuchowicz, J.; Kuhnert, N., MALDI-TOF Mass Spectrometry:

Avoidance of Artifacts and Analysis of Caffeine-Precipitated Sill Thearubigins from 15 Commercial

Black Teas. J Agr Food Chem 2012, 60 (18), 4514-4525.

66. Reed, J. D.; Krueger, C. G.; Vestling, M. M., MALDI-TOF mass spectrometry of oligomeric food

polyphenols. Phytochemistry 2005, 66 (18), 2248-2263.

198

67. Wang, J.; Sporns, P., MALDI-TOF MS analysis of isoflavones in soy products. J Agr Food Chem

2000, 48 (12), 5887-5892.

68. Zhang, H. Y., Structure-Activity Relationships and Rational Design Strategies for Radical-

Scavenging Antioxidants. Curr Comput-Aid Drug 2005, 1 (3), 257-273.

69. Roginsky, V.; Barsukova, T.; Hsu, C. F.; Kilmartin, P. A., Chain-breaking antioxidant activity and

cyclic voltammetry characterization of polyphenols in a range of green, oolong, and black teas. J Agr

Food Chem 2003, 51 (19), 5798-5802.

70. Kanu, A. B.; Dwivedi, P.; Tam, M.; Matz, L.; Hill, H. H., Ion mobility-mass spectrometry. J Mass

Spectrom 2008, 43 (1), 1-22.

71. Creaser, C. S.; Griffiths, J. R.; Bramwell, C. J.; Noreen, S.; Hill, C. A.; Thomas, C. L. P., Ion

mobility spectrometry: a review. Part 1. Structural analysis by mobility measurement. Analyst 2004, 129

(11), 984-994.

72. Asbury, G. R.; Hill, H. H., Evaluation of ultrahigh resolution ion mobility spectrometry as an

analytical separation device in chromatographic terms. J Microcolumn Sep 2000, 12 (3), 172-178.

73. Angel, P. M.; Caprioli, R. M., Matrix-Assisted Laser Desorption Ionization Imaging Mass

Spectrometry: In Situ Molecular Mapping. Biochemistry-Us 2013, 52 (22), 3818-3828.

74. Kaspar, S.; Peukert, M.; Svatos, A.; Matros, A.; Mock, H. P., MALDI-imaging mass spectrometry

- An emerging technique in plant biology. Proteomics 2011, 11 (9), 1840-1850.

75. Wu, Z. G.; Rodgers, R. P.; Marshall, A. G., Compositional determination of acidic species in

Illinois no. 6 coal extracts by electrospray ionization Fourier transform ion cyclotron resonance mass

spectrometry. Energ Fuel 2004, 18 (5), 1424-1428.

199

76. Kim, S.; Kramer, R. W.; Hatcher, P. G., Graphical method for analysis of ultrahigh-resolution

broadband mass spectra of natural organic matter, the van Krevelen diagram. Anal Chem 2003, 75 (20),

5336-5344.

77. Kendrick, E., A Mass Scale Based on Ch2=14.0000 for High Resolution Mass Spectrometry of

Organic Compounds. Anal Chem 1963, 35 (13), 2146-&.

78. Hughey, C. A.; Rodgers, R. P.; Marshall, A. G., Resolution of 11 000 compositionally distinct

components in a single Electrospray ionization Fourier transform ion cyclotron resonance mass spectrum

of crude oil. Anal Chem 2002, 74 (16), 4145-4149.

79. Hughey, C. A.; Galasso, S. A.; Zumberge, J. E., Detailed compositional comparison of acidic

NSO compounds in biodegraded reservoir and surface crude oils by negative ion electrospray Fourier

transform ion cyclotron resonance mass spectrometry. Fuel 2007, 86 (5-6), 758-768.

80. Koch, B. P.; Dittmar, T.; Witt, M.; Kattner, G., Fundamentals of molecular formula assignment to

ultrahigh resolution mass data of natural organic matter. Anal Chem 2007, 79 (4), 1758-1763.

81. Kim, S.; Rodgers, R. P.; Blakney, G. T.; Hendrickson, C. L.; Marshall, A. G., Automated

Electrospray Ionization FT-ICR Mass Spectrometry for Petroleum Analysis. J Am Soc Mass Spectr 2009,

20 (2), 263-268.

82. Schmid, D. G.; Grosche, P.; Jung, G., High-resolution analysis of a 144-membered pyrazole

library from combinatorial solid phase synthesis by using electrospray ionisation Fourier transform ion

cyclotron resonance mass spectrometry. Rapid Commun Mass Sp 2001, 15 (5), 341-347.

83. Nakayama, T.; Ichiba, M.; Kuwabara, M.; Kajiya, K.; Kumazawa, S., Mechanisms and structural

specificity of hydrogen peroxide formation during oxidation of catechins. Food Sci Technol Res 2002, 8

(3), 261-267.

84. Subramanian, N.; Venkatesh, P.; Ganguli, S.; Sinkar, V. P., Role of polyphenol oxidase and

peroxidase in the generation of black tea theaflavins. J Agr Food Chem 1999, 47 (7), 2571-2578.

200

85. Mulder, T. P. J.; van Platerink, C. J.; Schuyl, P. J. W.; van Amelsvoort, J. M. M., Analysis of

theaflavins in biological fluids using liquid chromatography-electrospray mass spectrometry. J

Chromatogr B 2001, 760 (2), 271-279.

86. Cuyckens, F.; Claeys, M., Mass spectrometry in the structural analysis of flavonoids. J Mass

Spectrom 2004, 39 (1), 1-15.

87. Cren-Olive, C.; Deprez, S.; Lebrun, S.; Coddeville, B.; Rolando, C., Characterization of

methylation site of monomethylflavan-3-ols by liquid chromatography/electrospray ionization tandem

mass spectrometry. Rapid Commun Mass Sp 2000, 14 (23), 2312-2319.

88. Roberts, E. A. H.; Oldschool, M., Enzymic Oxidation of Polyphenols to Benzotropolones. Chem

Ind-London 1958, (4), 99-100.

89. Myers, M.; Roberts, E. A. H.; Rustidge, D. W., Some Trace Constituents of Black Tea. Chem Ind-

London 1959, (30), 950-951.

90. Hashimoto, F.; Nonaka, G.; Nishioka, I., Tannins and Related-Compounds .114. Structures of

Novel Fermentation Products, Theogallinin, Theaflavonin and Desgalloyl Theaflavonin from Black Tea,

and Changes of Tea Leaf Polyphenols during Fermentation. Chem Pharm Bull 1992, 40 (6), 1383-1389.

91. Wang, D. M.; Kubota, K.; Kobayashi, A.; Juan, I. M., Analysis of glycosidically bound aroma

precursors in tea leaves. 3. Change in the glycoside content of tea leaves during the oolong tea

manufacturing process. J Agr Food Chem 2001, 49 (11), 5391-5396.

92. Hall, M. N.; Robertson, A.; Scotter, C. N. G., Near-Infrared Reflectance Prediction of Quality,

Theaflavin Content and Moisture-Content of Black Tea. Food Chem 1988, 27 (1), 61-75.

93. Temple, S. J.; Temple, C. M.; van Boxtel, A. J. B.; Clifford, M. N., The effect of drying on black

tea quality. J Sci Food Agr 2001, 81 (8), 764-772.

201

94. Ozawa, T.; Kataoka, M.; Morikawa, K.; Negishi, O., Elucidation of the partial structure of

polymeric thearubigins from black tea by chemical degradation. Biosci Biotech Bioch 1996, 60 (12),

2023-2027.

95. Bailey, R. G.; Nursten, H. E.; Mcdowell, I., Isolation and Analysis of a Polymeric Thearubigin

Fraction from Tea. J Sci Food Agr 1992, 59 (3), 365-375.

96. Wan, X. C.; Nursten, H. E.; Cai, Y.; Davis, A. L.; Wilkins, J. P. G.; Davies, A. P., A new type of

tea pigment - From the chemical oxidation of epicatechin gallate and isolated from tea. J Sci Food Agr

1997, 74 (3), 401-408.

97. Cai, Y. Z.; Xing, J.; Sun, M.; Zhan, Z. Q.; Corke, H., Phenolic antioxidants (hydrolyzable tannins,

flavonols, and anthocyanins) identified by LC-ESI-MS and MALDI-QIT-TOF MS from Rosa chinensis

flowers. J Agr Food Chem 2005, 53 (26), 9940-9948.

98. Menet, M. C.; Sang, S. M.; Yang, C. S.; Ho, C. T.; Rosen, R. T., Analysis of theaflavins and

thearubigins from black tea extract by MALDI-TOF mass spectrometry. J Agr Food Chem 2004, 52 (9),

2455-2461.

99. Drynan, J. W.; Clifford, M. N.; Obuchowicz, J.; Kuhnert, N., MALDI-TOF Mass Spectrometry:

Avoidance of Artifacts and Analysis of Caffeine-Precipitated SII Thearubigins from 15 Commercial Black

Teas (vol 60, pg 4514, 2012). J Agr Food Chem 2013, 61 (6), 1418-1418.

100. Chen, P. C.; Chang, F. S.; Chen, I. Z.; Lu, F. M.; Cheng, T. J.; Chen, R. L. C., Redox potential of

tea infusion as an index for the degree of fermentation. Anal Chim Acta 2007, 594 (1), 32-36.

101. Zheng, X. F.; Chen, A. Y.; Hoshi, T.; Anzai, J.; Li, G. X., Electrochemical studies of (-)-

epigallocatechin gallate and its interaction with DNA. Anal Bioanal Chem 2006, 386 (6), 1913-1919.

102. Novak, I.; Seruga, M.; Komorsky-Lovric, S., Square-wave and cyclic voltammetry of epicatechin

gallate on glassy carbon electrode. J Electroanal Chem 2009, 631 (1-2), 71-75.

202

103. Kuhnert, N., Assessing the Theatubigene of black Tea with ultra high-resolution MS Methods.

Deut Lebensm-Rundsch 2011, 107 (8), 388-391.

104. Kuhnert, N., Identification by Tandem Mass spectrometry. Deut Lebensm-Rundsch 2011, 107

(11), 521-524.

105. Wu, C.; Siems, W. F.; Klasmeier, J.; Hill, H. H., Separation of isomeric peptides using

electrospray ionization/high-resolution ion mobility spectrometry. Anal Chem 2000, 72 (2), 391-395.

106. Karpas, Z.; Stimac, R. M.; Rappoport, Z., Differentiating between Large Isomers and Derivation

of Structural Information by Ion Mobility Spectrometry Mass-Spectrometry Techniques. International

Journal of Mass Spectrometry and Ion Processes 1988, 83 (1-2), 163-175.

107. Gabryelski, W.; Froese, K. L., Rapid and sensitive differentiation of anomers, linkage, and

position isomers of disaccharides using High-Field Asymmetric Waveform Ion Mobility Spectrometry

(FAIMS). J Am Soc Mass Spectr 2003, 14 (3), 265-277.

108. McCooeye, M.; Ding, L.; Gardner, G. J.; Fraser, C. A.; Lam, J.; Sturgeon, R. E.; Mester, Z.,

Separation and quantitation of the stereoisomers of ephedra alkaloids in natural health products using

flow injection-electrospray ionization-high field asymmetric waveform ion mobility spectrometry-mass

spectrometry. Anal Chem 2003, 75 (11), 2538-2542.

109. Cuyckens, F.; Wassvik, C.; Mortishire-Smith, R. J.; Tresadern, G.; Campuzano, I.; Claereboudt,

J., Product ion mobility as a promising tool for assignment of positional isomers of drug metabolites.

Rapid Commun Mass Sp 2011, 25 (23), 3497-3503.

110. Clowers, B. H.; Hill, H. H., Influence of cation adduction on the separation characteristics of

flavonoid diglycoside isomers using dual gate-ion mobility-quadrupole ion trap mass spectrometry. J

Mass Spectrom 2006, 41 (3), 339-351.

111. Guevremont, R., High-field asymmetric waveform ion mobility spectrometry: A new tool for

mass spectrometry. J Chromatogr A 2004, 1058 (1-2), 3-19.

203

112. Xie, C.; Yu, K.; Zhong, D. F.; Yuan, T.; Ye, F.; Jarrell, J. A.; Millar, A.; Chen, X. Y., Investigation

of Isomeric Transformations of Chlorogenic Acid in Buffers and Biological Matrixes by

Ultraperformance Liquid Chromatography Coupled with Hybrid Quadrupole/Ion Mobility/Orthogonal

Acceleration Time-of-Flight Mass Spectrometry. J Agr Food Chem 2011, 59 (20), 11078-11087.

113. Williams, J. P.; Bugarcic, T.; Habtemariam, A.; Giles, K.; Campuzano, I.; Rodger, P. M.; Sadler, P.

J., Isomer Separation and Gas-Phase Configurations of Organoruthenium Anticancer Complexes: Ion

Mobility Mass Spectrometry and Modeling. J Am Soc Mass Spectr 2009, 20 (6), 1119-1122.

114. Mie, A.; Jornten-Karlsson, M.; Axelsson, B. O.; Ray, A.; Reimann, C. T., Enantiomer separation

of amino acids by complexation with chiral reference compounds and high-field asymmetric waveform

ion mobility spectrometry: Preliminary results and possible limitations. Anal Chem 2007, 79 (7), 2850-

2858.

115. Dwivedi, P.; Wu, C.; Matz, L. M.; Clowers, B. H.; Siems, W. F.; Hill, H. H., Gas-phase chiral

separations by ion mobility spectrometry. Anal Chem 2006, 78 (24), 8200-8206.

116. Lee, T. C.; Kalenius, E.; Lazar, A. I.; Assaf, K. I.; Kuhnert, N.; Grun, C. H.; Janis, J.; Scherman,

O. A.; Nau, W. M., Chemistry inside molecular containers in the gas phase. Nat Chem 2013, 5 (5), 376-

382.

117. Wren, S. A. C., Peak capacity in gradient ultra performance liquid chromatography (UPLC). J

Pharmaceut Biomed 2005, 38 (2), 337-343.

118. Shoji, T., Chapter 26 - Chemical Properties, Bioavailability, and Metabolomics of Fruit

Proanthocyanidins. In Polyphenols in Human Health and Disease, Watson, R. R.; Preedy, V. R.; Zibadi,

S., Eds. Academic Press: San Diego, 2014; pp 339-351.

119. Jaiswal, R.; Jayasinghe, L.; Kuhnert, N., Identification and characterization of proanthocyanidins

of 16 members of the Rhododendron genus (Ericaceae) by tandem LC-MS. J Mass Spectrom 2012, 47

(4), 502-515.

204

120. Hou, D. X.; Masuzaki, S.; Tanigawa, S.; Hashimoto, F.; Chen, J. H.; Sogo, T.; Fujii, M., Oolong

Tea Theasinensins Attenuate Cyclooxygenase-2 Expression in Lipopolysaccharide (LPS)-Activated

Mouse Macrophages: Structure-Activity Relationship and Molecular Mechanisms. J Agr Food Chem

2010, 58 (24), 12735-12743.

121. Gescher, K.; Hafezi, W.; Louis, A.; Kuhn, J.; Hensel, A., Anti-herpes simplex virus type 1 activity

of Rhododendron ferrugineum L. extracts. Planta Med 2009, 75 (9), 989-989.

122. Lin, L. Z.; Chen, P.; Harnly, J. M., New phenolic components and chromatographic profiles of

green and fermented teas. J Agr Food Chem 2008, 56 (17), 8130-8140.

123. Clifford, M. N.; Stoupi, S.; Kuhnert, N., Profiling and characterization by LC-MSn of the

galloylquinic acids of green tea, tara tannin, and tannic acid. J Agr Food Chem 2007, 55 (8), 2797-2807.

124. Temple, C. M.; Clifford, M. N., The stability of theaflavins during HPLC analysis of a

decaffeinated aqueous tea extract. J Sci Food Agr 1997, 74 (4), 536-540.

125. Stodt, U.; Engelhardt, U. H., Progress in the analysis of selected tea constituents over the past 20

years. Food Res Int 2013, 53 (2), 636-648.

126. Heeren, R. M., Innovation in Molecular Imaging with Mass Spectrometry: Running Towards

High Resolution. Osteoarthr Cartilage 2012, 20, S1-S2.

127. Schwartz, S. A.; Weil, R. J.; Johnson, M. D.; Toms, S. A.; Caprioli, R. M., Protein profiling in

brain tumors using mass spectrometry: Feasibility of a new technique for the analysis of protein

expression. Clin Cancer Res 2004, 10 (3), 981-987.

128. Schwamborn, K.; Caprioli, R. M., INNOVATION Molecular imaging by mass spectrometry -

looking beyond classical histology. Nat Rev Cancer 2010, 10 (9), 639-646.

205

129. Stoeckli, M.; Staab, D.; Schweitzer, A.; Gardiner, J.; Seebach, D., Imaging of a beta-peptide

distribution in whole-body mice sections by MALDI mass spectrometry. J Am Soc Mass Spectr 2007, 18

(11), 1921-1924.

130. Burrell, M. M.; Earnshaw, C. J.; Clench, M. R., Imaging Matrix Assisted Laser Desorption

Ionization Mass Spectrometry: a technique to map plant metabolites within tissues at high spatial

resolution. J Exp Bot 2007, 58 (4), 757-763.

131. Zhang, H.; Cha, S. W.; Yeung, E. S., Colloidal graphite-assisted laser desorption/ionization MS

and MSn of small molecules. 2. Direct profiling and MS imaging of small metabolites from fruits. Anal

Chem 2007, 79 (17), 6575-6584.

132. Kuhnert, N., In Garden, Industry, Medicine and Food Polyphenols: Versatile Plant ingredients.

Chem Unserer Zeit 2013, 47 (2), 80-91.

133. Wiseman, S.; Waterhouse, A.; Korver, O.; Clifford, M.; Engelhardt, U.; Wan, X. C.; Hoffman, P.

C. H.; Rice-Evans, C.; Terao, J.; Gross, M.; Beecher, G., Special issue: The health effects of tea and tea

components. Crit Rev Food Sci 2001, 41 (5), 387-412.

134. Tomas-Barberan, F. A.; Andres-Lacueva, C., Polyphenols and Health: Current State and Progress.

J Agr Food Chem 2012, 60 (36), 8773-8775.

135. Khan, N.; Mukhtar, H., Tea polyphenols for health promotion. Life Sci 2007, 81 (7), 519-533.

136. Riemersma, R. A.; Rice-Evans, C. A.; Tyrrell, R. M.; Clifford, M. N.; Lean, M. E. J., Tea

flavonoids and cardiovascular health. Qjm-Mon J Assoc Phys 2001, 94 (5), 277-282.

137. Wright, J. S.; Johnson, E. R.; DiLabio, G. A., Predicting the activity of phenolic antioxidants:

Theoretical method, analysis of substituent effects, and application to major families of antioxidants. J Am

Chem Soc 2001, 123 (6), 1173-1183.

206

138. de Heer, M. I.; Korth, H. G.; Mulder, P., Poly methoxy phenols in solution: O-H bond

dissociation enthalpies, structures, and hydrogen bonding. J Org Chem 1999, 64 (19), 6969-6975.

139. Okuda, T.; Yoshida, T.; Hatano, T., Chemistry and Antioxidant Effects of Phenolics from Licorice,

Tea and Composite and Labiate Herbs. Abstr Pap Am Chem S 1992, 204, 201-AGFD.

140. Pietta, P. G., Flavonoids as antioxidants. J Nat Prod 2000, 63 (7), 1035-1042.

141. Neudorffer, A.; Desvergne, J. P.; Bonnefont-Rousselot, D.; Legrand, A.; Fleury, M. B.; Largeron,

M., Protective effects of 4-hydroxycinnamic ethyl ester derivatives and related dehydrodimers against

oxidation of LDL: Radical scavengers or metal chelators? J Agr Food Chem 2006, 54 (5), 1898-1905.

142. Vaya, J.; Mahmood, S.; Goldblum, A.; Aviram, M.; Volkova, N.; Shaalan, A.; Musa, R.; Tamir, S.,

Inhibition of LDL oxidation by flavonoids in relation to their structure and calculated enthalpy.

Phytochemistry 2003, 62 (1), 89-99.

143. Terao, J.; Piskula, M.; Yao, Q., Protective Effect of Epicatechin, Epicatechin Gallate, and

Quercetin on Lipid-Peroxidation in Phospholipid-Bilayers. Arch Biochem Biophys 1994, 308 (1), 278-

284.

144. Masaki, H.; Atsumi, T.; Sakurai, H., Hamamelitannin as a New Potent Active Oxygen Scavenger.

Phytochemistry 1994, 37 (2), 337-343.

145. Grassi, D.; Aggio, A.; Onori, L.; Croce, G.; Tiberti, S.; Ferri, C.; Ferri, L.; Desideri, G., Tea,

flavonoids, and nitric oxide-mediated vascular reactivity. J Nutr 2008, 138 (8), 1554-1560.

146. Bahorun, T.; Luximon-Ramma, A.; Neergheen-Bhujun, V. S.; Gunness, T. K.; Googoolye, K.;

Auger, C.; Crozier, A.; Aruoma, O. I., The effect of black tea on risk factors of cardiovascular disease in a

normal population. Prev Med 2012, 54, S98-S102.

207

147. Bajerska, J.; Wozniewicz, M.; Jeszka, J.; Drzymada-Czyz, S.; Walkowiak, J., Green tea aqueous

extract reduces visceral fat and decreases protein availability in rats fed with a high-fat diet. Nutr Res

2011, 31 (2), 157-164.

148. Lopes, G. K. B.; Schulman, H. M.; Hermes-Lima, M., Polyphenol tannic acid inhibits hydroxyl

radical formation from Fenton reaction by complexing ferrous ions. Bba-Gen Subjects 1999, 1472 (1-2),

142-152.

149. Andjelkovic, M.; Van Camp, J.; De Meulenaer, B.; Depaemelaere, G.; Socaciu, C.; Verloo, M.;

Verhe, R., Iron-chelation properties of phenolic acids bearing catechol and galloyl groups. Food Chem

2006, 98 (1), 23-31.

150. Sugihara, N.; Ohnishi, M.; Imamura, M.; Furuno, K., Differences in antioxidative efficiency of

catechins in various metal-induced lipid peroxidations in cultured hepatocytes. J Health Sci 2001, 47 (2),

99-106.

151. Mira, L.; Fernandez, M. T.; Santos, M.; Rocha, R.; Florencio, M. H.; Jennings, K. R., Interactions

of flavonoids with iron and copper ions: A mechanism for their antioxidant activity. Free Radical Res

2002, 36 (11), 1199-1208.

152. Steele, V. E.; Kelloff, G. J.; Balentine, D.; Boone, C. W.; Mehta, R.; Bagheri, D.; Sigman, C. C.;

Zhu, S. Y.; Sharma, S., Comparative chemopreventive mechanisms of green tea, black tea and selected

polyphenol extracts measured by in vitro bioassays. Carcinogenesis 2000, 21 (1), 63-67.

153. Goupy, P.; Vulcain, E.; Carls-Veyrat, C.; Dangles, O., Dietary antioxidants as inhibitors of the

heme-induced peroxidation of linoleic acid: Mechanism of action and synergism. Free Radical Bio Med

2007, 43 (6), 933-946.

154. Nobre, H. V.; Cunha, G. M. A.; Maia, F. D.; Oliveira, R. A.; Moraes, M. O.; Rao, V. S. N.,

Catechin attenuates 6-hydroxydopamine (6-OHDA)-induced cell death in primary cultures of

mesencephalic cells. Comp Biochem Phys C 2003, 136 (2), 175-180.

208

155. Stoupi, S.; Williamson, G.; Drynan, J. W.; Barron, D.; Clifford, M. N., A comparison of the in

vitro biotransformation of (-)-epicatechin and procyanidin B2 by human faecal microbiota. Mol Nutr

Food Res 2010, 54 (6), 747-759.

156. Stoupi, S.; Williamson, G.; Viton, F.; Barron, D.; King, L. J.; Brown, J. E.; Clifford, M. N., In

Vivo Bioavailability, Absorption, Excretion, and Pharmacokinetics of [C-14]Procyanidin B2 in Male

Rats. Drug Metab Dispos 2010, 38 (2), 287-291.

Unpublished References

Pawar V., Yassin G., Kuhnet N., Jayaraman S. Molecular weight distribution of thearubigins in black

tea (Manuscript under preparation)

Reference websites

a) Poulter S. Daily mail online, dailymail, co. u.k., 27th june 2008.

b) www.fao.org

c) www.scm.com

d) www.chromatographyonline.com

e) www.maldi-msi.org

209

http://www.fao.org/

http://www.scm.com/

http://www.chromatographyonline.com/

http://www.maldi-msi.org/

Documents

Characterization and Structural Elucidation of the Complex