phd thesis - jiang zhangjian - core · 2018. 1. 9. · chapter 2 determination of senkirkine and...

EXTRACTION, DETERMINATION AND METABOLIC

PROFILING OF ALKALOIDS IN TRADITIONAL

CHINESE MEDICINES BY MODERN ANALYTICAL

TECHNIQUES

JIANG ZHANGJIAN

(B. Sc., Soochow University)

A THESIS SUBMITTED FOR THE DEGREE OF

DOCTOR OF PHILOSOPHY

DEPARTMENT OF CHEMISTRY

NATIONAL UNIVERSITY OF SINGAPORE

2009

brought to you by COREView metadata, citation and similar papers at core.ac.uk

provided by ScholarBank@NUS

https://core.ac.uk/display/48636305?utm_source=pdf&utm_medium=banner&utm_campaign=pdf-decoration-v1

Acknowledgements

I

Acknowledgements

Foremost, I express my most sincere gratitude to my supervisor, Professor Sam

Fong Yau Li, for his long-term guidance, support and patience during my PhD

study.

I wish to extend my thanks to all the kind staffs for their patient support for my

projects, in particular to Dr. Eng Shi Ong in Department of Department of

Epidemiology and Public Health, Ms Frances Lim in Department of Chemistry.

I would like to thank all of my colleagues in Prof. Li’s group, who have helped me

in various ways: Dr. Hua Tao Feng, Dr. Lin Lin Yuan, Dr. Li Jun Yu, Dr. Yan Xu,

Dr. Xin Bing Zuo, Dr. Hua Nan Wu, Dr. Wai Siang Law, Dr. Ma He Liu, Miss Hiu

Fung Lau, Ms Junie Tok, Miss Elaine Teng Teng Tay, Miss Gui Hua Fang, Ms

Grace Birungi, Miss Feng Liu, Miss Ai Ping Chew, Mr. Jon Ashely, Miss Pei Pei

Gan and Mr. Jun Yu Lin.

I sincerely appreciate the National University of Singapore for providing me the

financial support during my research.

Finally, a million thanks to my parents for their selfless love and unfailing support.

Table of Contents

II

Table of Contents

Acknowledgements I

Table of Contents II

Summary VIII

List of Tables XI

List of Figures XII

List of Abbreviations XV

Chapter 1 Introduction 1

1.1 Traditional Chinese Medicine (TCM) 1

1.2 Separation and Analysis of TCM 3

1.2.1 Separation Technology 3

1.2.1.1 Headspace Extraction Techniques 3

1.2.1.1.1 Headspace Solid-phase Microextraction (HS-SPME) 4

1.2.1.1.2 Headspace Liquid-phase Microextraction (HS-LPME)5

1.2.1.2 Supercritical-fluid Extraction (SFE) 5

1.2.1.3 Ultrasonic Extraction (UE) 6

1.2.1.4 Microwave-assisted Extraction (MAE) 7

1.2.1.5 Pressurized-Liquid Extraction (PLE) 8

1.2.1.6 Microwave Distillation (MD) 9

1.2.2 Analysis of TCMs by Chromatographic Techniques 10

1.2.2.1 Analysis of TCMs by GC 11

1.2.2.2 Analysis of TCMs by LC 12

Table of Contents

III

1.2.2.3 Analysis of TCMs by Capillary Electrophoresis (CE) 15

1.3 Metabonomics of TCMs 18

1.3.1 Metabonomics 18

1.3.2 Metabonomics Samples 19

1.3.3 Metabonomics Analysis Technologies 19

1.3.3.1 NMR Spectroscopy 20

1.3.3.2 Mass Spectrometry 21

1.3.4 Metabonomics Data Analysis 22

1.4 Research Scope 23

References 26

Part I: Isolation and Determination of Pyrrolizidine Alkaloids in Traditional

Chinese Medicine with Modern Analytical Techniques 41

Chapter 2 Determination of Senkirkine and Senecionine in Tussilago Farfara

using Microwave Assisted Extraction and Pressurized Hot Water Extraction

with Liquid Chromatography Tandem Mass Spectrometry 42

2.1 Introduction 42

2.2 Experimental 45

2.2.1 Chemicals and reagents 45

2.2.2 Preparation of reference standards 45

2.2.3 Extraction 46

2.2.3.1 Microwave-assisted Extraction 46

2.2.3.2 Pressurized Hot Water Extraction 46

Table of Contents

IV

2.2.3.3 Heating under reflux 47

2.2.4 LC and LC/ESI-MS analyses of MAE Extracts 47

2.2.5 LC/ESI-MS analyses of PHWE Extracts 49

2.3 Results and Discussion 49

2.3.1 HPLC separation of MAE extract 49

2.3.1.1 HPLC separation with UV detector 49

2.3.1.2 Optimization of MAE 51

2.3.1.3 LC/ESI-MS analysis for MAE 53

2.3.2 HPLC analysis for PHWE 57

2.3.2.1 LC/ESI-MS analysis for PHWE 57

2.3.2.2 Optimization of PHWE 57

2.3.3 Matrix-induced interference 61

2.4 Conclusion 62

References 63

Chapter 3 Preconcentration and Separation of Toxic Pyrrolizidine Alkaloids

in Herbal Medicines by Non-aqueous Capillary Electrophoresis (NACE) 67

3.1 Introduction 67

3.2 Experimental Section 70

3.2.1 Materials 70

3.2.2 Instruments and methods 71

3.2.3 Standard sample and running buffer preparation 72

3.2.4 Extraction of PAs in herbal medicines 72

Table of Contents

V

3.2.4.1 Heating under reflux 72

3.2.4.2 Microwave-assisted extraction 72

3.3 Results and discussion 73

3.3.1 Optimization of NACE conditions 73

3.3.1.1 Effect of concentration of acetic acid 73

3.3.1.2 Effect of concentration of ammonium acetate 75

3.3.1.3 Effect of concentration of ACN 76

3.3.1.4 Effect of applied voltage 77

3.3.1.5 Linearity, precision, LODs and LOQs 78

3.3.1.6 Application 81

3.3.2 Optimization of online preconcentration conditions 82

3.3.2.1 Large volume sample stacking (LVSS) 82

3.3.2.2 Field-amplified sample stacking (FASS) 86

3.3.2.2.1 Choice of organic solvent plug 86

3.3.2.2.2 Optimization of sample matrix 87

3.3.2.2.3 Optimization of organic solvent plug injection time 88

3.3.2.2.4 Optimization of sample injection time and injection voltage89

3.3.2.2.5 Linearity, precision, LODs and LOQs 91

3.3.3 Application to real sample 94

3.4 Conclusion 95

References 96

Part II: Metabonomic Study of Natural Alkaloid in Rats 99

Table of Contents

VI

Chapter 4 Metabolic Profil ing of Berberine in Rats with gas

chromatography/mass spectrometry, liquid chromatography/mass

spectrometry and 1H NMR spectroscopy 100

4.1 Introduction 100

4.2 Experimental 103

4.2.1 Chemicals 103

4.2.2 Animal Studies 103

4.2.3 Histopathological Examination 104

4.2.4 Sample preparation for metabonomics profile of rat urine samples 104

4.2.5 Reversed-phased LC/MSMS 104

4.2.6 Analysis of urine samples by 1H NMR 105

4.2.7 Analysis of liver samples by GC/MS 105

4.2.8 Chemometric analysis 106

4.2.9 Statistical analysis 107

4.3 Results and Discussion 107

4.3.1 Results 107

4.3.1.1 Body weight and histopathology 107

4.3.1.2 Determination of metabolites in rat livers samples by GC/MS109

4.3.1.3 Determination of metabolites in rat urine samples by 1H NMR

113

4.3.1.4 Determination of metabolites in rat urine samples by LC/MS 122

4.3.2 Discussion 129

Table of Contents

VII

4.4 Conclusion 131

References 132

Appendix for Chapter 4 137

Chapter 5 Conclusion and Future Work 168

5.1 Summary of Results 168

5.2 Limitation and Future Work 170

List of Publications 172

Conference Papers 173

Summary

VIII

Summary

In this dissertation, efforts were dedicated to the study of alkaloids in traditional

Chinese medicines (TCMs). Various techniques of separation and determination

of natural alkaloids in Chinese herbs have been developed and the metabolic

profiling of alkaloid in rat model has been investigated.

This dissertation is broadly divided into two parts: the first part (chapter 2 and

chapter 3) focused on the separation and determination of toxic pyrrolizidine

alkaloids (PAs) in Chinese herbal medicine, called Tussilago farfara (Kuan

Donghua), and the second part (chapter 4) focused on the metabonomics studies

of a commonly used TCM, named berberine, in rat model.

In chapter 1, TCM was briefly reviewed from various aspects, including the

background, separation and analysis and metabonomics of TCMs.

Two new extraction techniques, microwave-assisted extraction (MAE) and

pressurized hot water extraction (PHWE) were applied to the separation of toxic

PAs from Tussilago farfara (Kuan Donghua) (Chapter 2). Conditions for MAE

and PHWE were optimized. It was found that a binary mixture of MeOH:H2O (1:1)

acidified using HCl to pH 2–3 was the optimal solvent for the extraction of the

PAs in the plant materials. The results obtained from MAE and PHWE were

compared against heating under reflux. LC with UV detection and electrospray

ionization mass spectrometry (ESI-MS) in the positive mode were used for the

determination and quantitation of PAs in the botanical extract. The proposed

extraction methods with LC/MS allow for a rapid detection of both the major and

Summary

IX

the minor PAs in T. farfara in the presence of co-eluting peaks. With LC/MS, the

quantitative analysis of PAs in the extract was done using internal standard

calibration and the precision was found to vary from 0.6% to 5.4% (n=6) on

different days. The limits of detection (LODs) and limits of quantitation (LOQs)

for MAE and PHWE were found to be 0.26 to 1.04 µg/g and1.32 to 5.29 µg/g,

respectively. The method precision of MAE and PHWE were found to vary from

3.7% to 10.4% on different days. The results showed that extraction efficiencies

for major and minor PAs extracted using MAE and PHWE were comparable to

that by heating under reflux. Our results also showed that significant ion

suppression was not observed in the LC/MS analysis.

In chapter 3, a simple and efficient non-aqueous capillary electrophoresis (NACE)

method was established for the determination of toxic PAs in Tussilago farfara

(Kuan Donghua) firstly. Influences from the background electrolyte (BGE) and

separation voltage were investigated. Then two online preconcentration methods

for NACE, named large volume sample stacking (LVSS) and field-amplified

sample stacking (FASS), were investigated. The stacking conditions, such as the

length of sample zone in LVSS, choice of organic solvent plug, organic solvent

plug length, sample injection voltage and injection time in FASS, were optimized.

Under the optimized conditions, the FASS could provide 18 to 89–fold sensitivity

enhancements with satisfactory reproducibility, while the LVSS could only

provide 5 to 7-fold.

In chapter 4, methods using gas chromatography/mass spectrometry (GC/MS),

Summary

X

liquid chromatography/mass spectrometry (LC/MS) and 1H NMR with pattern

recognition tools such as principle components analysis (PCA) were used to study

the metabolic profiles of rats after the administration of berberine. From the

normalized peak areas obtained from GC/MS analysis of liver extracts and

LC/MS analysis of urine samples and peak heights from 1H NMR analysis of

urine samples, statistical analyses were used in the identification of potential

biomarkers. The results from non-targeted 1H NMR data processing had proved

the reliability and accuracy of targeted LC/MS data processing. The proposed

approach provided a more comprehensive picture of the metabolic changes after

administration of berberine in rat model. The multiple parametric approach

together with pattern recognition tools is a useful platform to study metabolic

profiles after ingestion of botanicals and medicinal plants.

List of Tables

XI

List of Tables

Table 2.1 Comparison of MAE with heating under reflux for the analysis of senkirkine in Tussilago farfara by LC with UV detection. 52 Table 2.2 Comparison of MAE between heating under reflux and PHWE between heating under reflux. 59 Table 3.1 Resolution at different applied voltage. 77 Table 3.2 Linearity, precision, LODs and LOQs of NACE. 81 Table 3.3 Precision, linearity, LODs and LOQs of FASS. 94 Table 3.4 Comparison of limits of detection (LOD) and sensitivity enhancement (SE). 92 Table 4.1 Selected metabolites identified in rat urine samples for both control and treatment group as measured by 1H NMR. 114 Table 4.2 Selected metabolites identified in rat urine samples for both the control and treatment group as measured by LC/MS (positive and negative mode). 122 Table 4.3 Metabolites identified in rat urine samples for both the control and treatment group as measured by LC/MS (positive mode). 135 Table 4.4 Metabolites identified in rat urine samples for both the control and treatment group as measured by LC/MS (negative mode). 145 Table 4.5 Metabolites identified in rat urine samples for both the control and treatment group as measured by 1H NMR. 151

List of Figures

XII

List of Figures

Figure 2.1 Chemical structures of (A) senkirkine and (B) senecionine. 43 Figure 2.2 Chromatogram of Tussilago farfara extract A) before clean-up step, and B) after clean-up step. 51 Figure 2.3 Effect of different acids and solvents on microwave-assisted extraction for senkirkine using: A) different acids with MeOH: H2O (1:1) as solvent, and B) MeOH: H2O (1:1) and EtOH: H2O (1:1) as solvents with HCl. 52 Figure 2.4 MS2 Fragmentation pattern of A) Senkirkine, B) Senecionine, and C) MS3 fragmentation pattern of Senecionine. 55 Figure 2.5 A) Total ion chromatogram (TIC) of Tussilago farfara extract before clean-up step. B) Extracted ion chromatogram (EIC) of senecionine observed at m/z 336 and C) EIC of senkirkine observed at m/z 366. 57 Figure 2.6 Effect of different extraction temperatures on the recoveries of (A) senecionine and (B) senkirkine by PHWE (n=3) with flow rate: 1.5 ml/min for 40 min. 58 Figure 2.7 Effect of extraction time on the recoveries of senecionine and senkirkine by PHWE. 59 Figure 3.1 Structures of PAs of senkirkine, monocrotaline, sesecionine and retrorsine. 71 Figure 3.2A Effect of acetic acid concentration on the separation of PAs. 74 Figure 3.2B Effect of acetic acid concentration on apparent mobilities of PAs. 74 Figure 3.3A Effect of ammonium acetate concentration on separation of PAs. 75 Figure 3.3B Effect of ammonium acetate concentration on apparent mobilities of PAs. 76 Figure 3.4A Effect of acetonitrile concentration on the separation of PAs. 77 Figure 3.4B Effect of acetonitrile (ACN) concentration on apparent mobilities of PAs. 77 Figure 3.5 Effect of applied voltage on the separation of PAs. 78

List of Figures

XIII

Figure 3.6 The electropherograms of the standards mixture solution and the real sample. 81 Figure 3.7 CE electropherograms of the 4 PAs obtained by NACE and LVSS, respectively. 83 Figure 3.8 LODs of the 4 PAs with LVSS, respectively. 85 Figure 3.9 CE electropherograms of PAs obtained by NACE and LVSS. 86 Figure 3.10 Effect of organic solvent plug on FASS. 87 Figure 3.11 Effect of sample matrix on FASS. 88 Figure 3.12 Effect of organic solvent plug injection time on FASS. 89 Figure 3.13 Effect of sample injection time and injection voltage on peak areas.90 Figure 3.14 Electropherograms of the standard mixture (A) and the real sample (B) under the optimum conditions. 94 Figure 4.1 Chemical structure of berberine. 100 Figure 4.2 The trends of body weights of rats. 108 Figure 4.3 Histopathological photomicrographs of rat livers and kidneys from control and treatment group with dose of 50 mg/kg of berberine. 109 Figure 4.4 Typical GC/MS chromatograms of the lipid fraction of rat liver extracts. 110 Figure 4.5 (A) PCA scores plot based on GC/MS analysis of rat liver samples from all the control (n=8) and treatment group (n=8). 112 Figure 4.5 (B) Simplified pathway illustrating perturbed metabolites in the rat liver samples between the control and treatment group. 112 Figure 4.6 Typical 1H NMR spectrum (0-5 ppm) of rat urine samples obtained from pre-dose and treatment group (day 1, 2, 3 and 4). 113 Figure 4.7 Changes of selected metabolites as measured by 1H NMR on different days. 117

Figure 4.8 Data analysis of 1H NMR rat urine metabolic profiles. 118

List of Figures

XIV

Figure 4.9 Simplified pathway illustrating perturbed metabolites in the rat urine samples on day 4 between the control and treatment group. 121 Figure 4.10 Simplified pathway illustrating metabolites involved in purine metabolism in the rat urine samples on day 4 between the control and treatment group. 122 Figure 4.11 Typical total ion chromatograms (TIC) of rat urine samples obtained from control group and treatment group using LC/MS (positive mode). 123 Figure 4.12 Changes of selected metabolites as measured by LC/MS (positive mode) on different days. 126 Figure 4.13 Changes of selected metabolites as measured by LC/MS (negative mode) on different days. 127 Figure 4.14 Multivariate data analysis of LC/MS data (positive mode) metabolic profiles. 128

List of Abbreviations

XV


ACN Acetonitrile

BGE Background electrolyte

CE Capillary electrophoresis

CEC Capillary electrochromatography

CGE Capillary gel electrophoresis

CID Collision induced dissociation

CIEF Capillary isoelectric focusing

CITP Capillary isotachophoresis

CZE Capillary zone electrophoresis

DA Discriminant analysis

EIC Extracted ion chromatogram

EOF Electroosmotic flow

ESI-MS Electrospray ionization mass spectrometry

FASS Field-amplified sample stacking

FTIR Fourier transform infrared

GC/MS Gas chromatography/mass spectrometry

HS-SPME Headspace solid-phase microextraction

HS-LPME Headspace liquid-phase microextraction

LC/MS Liquid chromatography/mass spectrometry

LOD Limit of detection

LOQ Limit of quantitation


XVI

LVSS Large volume sample stacking

LXD Longdan Xiegan Decoction

MAE Microwace-assisted extraction

MD Microwave distillation

MEKC Micellar electrokinetic chromatography

NACE Non-aqueous capillary electrophoresis

NMR Nuclear magnetic resonance

PA Pyrrolizidine alkaloid

PC Principal components

PCA Principle components analysis

PHWE Pressurized hot water extraction

PLE Pressurized-liquid extraction

PLS Partial least squares/projection to latent structures

R.S.D. Relative standard deviation

RT Retention time

SE Sensitivity enhancement

SFE Supercritical-fluid extraction

SIM Selected ion monitoring

TCM Traditional Chinese medicine

TIC Total ion chromatogram

TMAO Trimethylamine N-oxide

UE Ultrasonic extraction


XVII

UPLC Ultra performance liquid chromatography

Chapter 1 Introduction

1


1.1 Traditional Chinese Medicine (TCM)

Traditional Chinese medicines (TCMs) which have been used to treat as well as

prevent various diseases for more than 2000 years in China are getting more and

more popular nowadays in the whole world. They have been modified to some

extent in other Asian countries, such as Korea and Japan and have attracted

significant attention in European, Australia and North American countries during

the last two decades [1]. Since Chinese medicine is generally extracted from

natural products without artificial additives which creates mild healing effects and

incurs fewer side effects, it has been considered as an important complementary

and alternative medicine in Western countries. At the same time Chinese herbs

have always been the most important resources for screening lead compounds.

Chinese Pharmacopoeia has recorded more than 500 examples of crude drugs

from plants and 400 TCMs that are widely used all over the world [2, 3]. These

drugs are multi-component systems, containing usually hundreds of chemically

different constituents which can act in a synergistic manner within the human

body, and can provide unique therapeutic properties with minimal or no

undesirable side-effects [4]. However, only a few, if not one, compounds are

responsible for the beneficial and/or hazardous effects [5]. For example, as the

most famous and commonly used Chinese herb in Chinese history, Ginseng, the

root and rhizome of Panax spp., has multifold bioactivities including

antimitogenic effect, improving impaired memory and inhibition of tumor cell

growth. The pharmacological properties of Ginseng are generally attributed to its

triterpene glycosides, called ginsenosides [5]. The dried root of Salvia miltiorrhiza

Bunge (Chinese name ‘Danshen’) is another Chinese herb of the most well-known


2

traditional Chinese medicines. It is widely used to treat coronary heart diseases,

cerebrovascular diseases, bone loss, hepatitis, hepatocirrhosis and chronic renal

failure, dysmenorrheal and neurasthenic insomnia. Phenolic acids are the water

soluble active constituents of Danshen [5].

However, due to the complicate constitutions of herbal medicines, besides the

therapeutic effects, TCMs show toxic effects also. For example, Tussilago farfara

(Kuan Donghua) is commonly used for the relief of coughs and as an expectorant,

blood pressure raiser, platelet activating factor and anti-inflammatory agent [6]. It

also can be used for the treatment of asthma, silicosis, pulmonary tuberculosis,

obesity, type 2 diabetes, and hepatitis [7–12]. However, the toxic pyrrolizidine

alkaloids included in Tussilago farfara are hepatotoxic, lung carcinogenesis,

neurotoxic and cytotoxic. Lilu, the roots and rhizomes of several Veratrum species,

has been used to treat aphasia arising from apoplexy, wind-type dysentery,

jaundice, scabies and chronic malaria for centuries in China. Nevertheless,

Veratrum nigrum L. is a very poisonous plant. The steroidal alkaloids isolated

from this plant were reported to exert teratogenic effects in several laboratory

animals. Perharic et al. [13] reported the toxicological problems resulting from

exposure to TCMs in 1994.

Until recently, there are still a lot of components unknown in TCMs, just like a

“black-box” system. What's more, the constituents in them are influenced by three

principal factors: heredity (genetic composition), ontogeny (stage of development)

and environment (e. g. climate, associated flora, soil and method of cultivation).

Therefore, the action mechanisms of TCMs are often difficult to be clearly

understood. As a result, developing effective techniques for isolation or separation

and analysis of effective or toxic components in TCMs is a very important subject


3

in order to ensure their reliability and repeatability of pharmacological and clinical

research.

1.2 Separation and Analysis of TCM

1.2.1 Separation Technology

Sample preparation is the crucial first step in the chromatographic analysis of

TCMs, because it is necessary to extract the desired chemical components from

the herbal material for further separation and characterization. Thus, the

development of novel sample-preparation techniques with significant advantages

over conventional methods (e.g. reduction in organic solvent consumption and in

sample degradation, elimination of additional sample clean-up and concentration

steps before chromatographic analysis, improvement in extraction efficiency,

selectivity, and/or kinetics, ease of automation, etc.) for the extraction and

analysis of medicinal plants plays an important role in the overall effort of

ensuring and providing high quality herbal products to consumers worldwide.

Herein, the recent sample-preparation techniques including headspace solid-phase

microextraction (HS-SPME), headspace liquid-phase microextraction (HS-LPME),

supercritical-fluid extraction (SFE), ultrasonic extraction (UE),

microwave-assisted extraction (MAE), pressurized-liquid extraction (PLE) and

microwave distillation (MD) will be introduced.

1.2.1.1 Headspace Extraction Techniques

The medicinal properties of TCMs can be partly related to the presence of volatile

constituents (e.g. essential oils) in the plant matrix, and GC–MS and GC–FID are

frequently used for determination of these volatile components. Because the

sample to be injected should be free from non-volatile components, a fractionation

step is necessary before GC analysis. The disadvantages of commonly used


4

sample preparation techniques, such as distillation and liquid solvent extraction,

are that they usually require large amounts of organic solvents and manpower.

These methods also tend to be destructive in nature and significant artifact

formation can occur due to the sample decomposition at high temperatures [14].

Recently, the two techniques of HS-SPME and HS-LPME have been developed

for the extraction of volatile constituents from TCMs.

1.2.1.1.1 Headspace Solid-phase Microextraction (HS-SPME)

In 1990, Arthur and Pawliszyn [15] introduced a completely solvent-less method,

named solid-phase microextraction (SPME), in which a fused silica fiber coated

with a stationary phase is exposed to the sample or its headspace and the target

analytes partition from the sample matrix to the fiber coating [16]. After

extracting for a set period of time, the fiber is transferred to the heated injection

port for GC or GC–MS analysis. The method has been applied widely in recent

years to the determination of the volatile chemical components of plants and

flowers [17–22]. Subsequently, HS-SPME was successfully applied to the

analyses of volatile components in TCMs, such as Schisandra chinensis Bail,

Chinese arborvitae, Angelico pubescens, and Angelico sinensis [23-26]. It was

proved that the reproducible and rapid determination of volatile compounds in

TCMs could be achieved when HS-SPME was used coupling with GC–MS, with

the advantages of eliminating the extraction or fractionation step and reducing

artifact formation. HS-SPME was also developed as a quality assessment tool for

Flos Chrysanthemi Indici from different growing areas [27]. In 2004, HS-SPME

was developed for the analysis of 35 volatile constituents in Rhioxma Curcumae

Aeruginosae [28], and 27 compounds in the TCM prescription of

Xiao-Cheng-Qi-Tang [29]. In 2006, Guo and Huang [30] analyzed Atractylodes


5

macrocephala (baizhu) and Atractylodes lancea (cangzhu) with HS-SPME and

found 23 common components. Qi et al. analyzed the volatile compounds from

Curcuma wenyujin and Houttuynia cotdata by using HS-SPME–GC–MS [31, 32].

Compared to distillation, HS-SPME was a simple, rapid, and solvent-free sample

extraction and concentration technique which has a strong potential for monitoring

the quality of TCMs.

1.2.1.1.2 Headspace Liquid-phase Microextraction (HS-LPME)

LPME was firstly introduced by Jeannot and Cantwell [33, 34] and He and Lee

[35]. It is performed by suspending 1µL drop of organic solvent on the tip of

either a Teflon rod or the needle tip of a microsyringe immersed in the stirred

aqueous sample. Then the microdrop was injected to GC-MS. The LPME

technique has been successfully applied to environmental analysis and drug

analysis [36–40]. HS-LPME was firstly introduced by Theis et al. for volatile

organic compounds in an aqueous matrix in 2001[41]. Cao and Qi found similar

results by HS-LPME and HS-SPME for the analysis of 66 volatile compounds

from a common TCM, C. wenyujin [42].

HS-SPME and HS-LPME have similar capabilities in terms of precision and

speed of analysis and are very suitable for quality assessment of TCMs. However,

the latter is prior to the former considering the choice of solvents is wider than the

limited number of stationary phases for SPME and the cost of the few microliters

of solvent is negligible compared to the cost of commercially available SPME

fibers.

1.2.1.2 Supercritical-fluid Extraction (SFE)

Supercritical-fluid extraction has been used for many years for the extraction of

volatile components, e.g. essential oils and aroma compounds, from plant


6

materials, on a laboratory and industrial scale [43, 44]. As far as SFE is concerned,

the extraction time is short, the use of hazardous solvents can be reduced. It is also

very convenient to couple with GC, LC and supercritical-fluid chromatography.

The application of SFE to the extraction of active compounds from medicinal

plants has attracted a lot of attention due to the avoiding of degradation caused by

lengthy exposure to elevated temperatures and atmospheric oxygen.

Supercritical carbon dioxide extraction has been successfully applied to extract the

essential oil for GC–MS from Aloe vera [45], Polygonum cuspidatum [46], radix

Angelicae dahuricae [47], ginger [48], and Cinnamomum cassia presl [49]. SFE

with methanol-modified supercritical carbon dioxide, followed by LC has been

reported for the analysis of sinomenine from Sinomenium acutum [50], berberine

from rhizome of Coptis chinensis Franch [51], triterpenoids in fruiting bodies of

Ganoderma lucidum [52], and saponins from Ginseng [53]. The application of

SFE coupling with LC and LC×LC to the analysis of G. lucidum has also been

reported [54, 55].

Recently, the coupling of SFE with high-speed counter-current chromatography

(HSCCC) has been used for the extraction and isolation of flavonoids [56, 57],

aurentiamide acetate from Patrinia villosa Juss [58] and psoralen and isopsoralen

from Fructus Psoraleae [59]. Today, many active compounds are obtained by

using SFE–HSCCC technique.

1.2.1.3 Ultrasonic Extraction (UE)

The first application of ultrasonic energy to the extraction of medicinal

compounds from plant materials can be tracked back to 1950s. The mechanisms

of the useful ultrasonically assisted extraction could be described into two

directions [60]. Firstly, some plant cells occur in the form of glands (external or


7

internal) filled with essential oil. A characteristic of external glands is that their

skin is very thin and can be easily destroyed by sonication, thus facilitating release

of essential oil contents into the extraction solvent. The other is ultrasound can

also facilitate the swelling and hydration of plant materials to cause enlargement

of the pores of the cell wall. Better swelling will improve the rate of mass transfer

and, sometimes, break the cell walls, thus resulting in increased extraction

efficiency and/or reduced extraction time.

As a novel approach to extraction and sample preparation for medicinal herbs,

Huie et al. reported the application of ultrasound to assist the surfactant-mediated

extraction of ginsenosides from American ginseng [61]. In ultrasonically assisted

extraction the aqueous surfactant solution containing 10% Triton X-100 as the

extraction solvent can be used to fasten the extraction kinetics and obtain higher

recovery compared to methanol and water.

1.2.1.4 Microwave-assisted Extraction (MAE)

The use of microwave energy for heating the solution (microwave-assisted

extraction, MAE) results in significant reductions in both the extraction time and

the consumption of organic solvents compared with conventional liquid–solid

extraction methods, such as soxhlet extraction [62]. This is because the

microwaves heat the solvent or solvent mixture directly, thus improving the

efficiency of heating. Solvent composition, solvent volume, extraction

temperature, and matrix characteristics are the most important parameters that can

affect the MAE efficiency. Until recently, MAE has been widely applied for the

extraction of flavonoids from Acanthopanax senticosus Harms, Mahonia bealei

(Foft.) leaves, and Chrysanthemum morifolium (Ramat.) petals, for rutin and

quercetin from Flos Sophorae, for six ginsenosides from ginseng root and for the


8

water-soluble bioactive constituents from the traditional Chinese medicinal

preparation Tongmaichongji [63–68]. In 2005, Liu et al. reported the use of

high-pressure MAE for the extraction of flavonoids and saponins from A.

senticosus leaves [69]. Most recently, MAE has been combined with HS-SPME or

HS-LPME for the quantitative analysis of volatile active components in TCMs

[70–72]. This technique provided a simple, rapid and solvent-free tool for the

quantitative analysis of active compounds in TCMs.

MAE has also been successfully applied to the analysis of heavy metals in herbal

products which play an important role in the therapeutic effects, despite their

reported toxicity. The determination of metals, such mercury, arsenic and lead in

traditional Chinese medicines have been reported [73-76].

1.2.1.5 Pressurized-liquid Extraction (PLE)

Pressurized-liquid extraction (PLE) emerged in the mid-1990s. However, the first

comprehensive study on the feasibility/usefulness of applying PLE in medicinal

herb analysis was carried out until 1999 by Benthin et al. [77]. The sample was

extracted with water or organic solvent or the mixture of both in a stainless-steel

cell under elevated temperature and pressure. Normally, the temperature ranges up

to 350oC and the pressure is kept as high as enough to keep the extraction solvents

in a liquid state. Under these conditions, the solubility of the analytes and the mass

transfer are increased, and the viscosity and the surface tension of the solvents are

decreased. These can improve the contact of the analytes with the solvent and thus

enhance the extraction [78]. Recently, Li’s group has coupled PLE with capillary

electrophoresis (CE), GC and HPLC for the determination of 5 anthraquinones

from Rhubarb [79], 11 sesquiterpenes from Ezhu, which is derived from three

species of Curcuma [80, 81], 11 major triterpene saponins from Panax


9

notoginseng [82–85], 43 nucleosides, bases and their analogues in natural and

cultured Cordyceps [86], Z-ligustilide, Z-butylidenephthalide, and ferulic acid

from Angelica sinensis [87], saponins and fatty acids from Suanzaoren [88], and

alkaloids and limonoids from Cortex Dictamni [89]. Ong and co-workers [90]

found that PLE is superior to conventional extraction methods such as ultrasonic

and Soxhlet extraction for the extraction of berberine and aristolochic acids in

medicinal plants. Later he reported the use of PLE coupled with CE for the

determination of glycyrrhizin in Radix glycyrrhizae or liquorice [91].

Since the polarity of water decreases markedly when liquid water is under

elevated temperature (ranging from 100 to 374 C) and elevated pressure. It can be

used for the pressurized hot water extraction (PHWE) of a wide range of analytes.

Fernandez- Perez et al. reported the application of PHWE to the extraction of

essential oils in plant materials [92]. Zhang et al. reported the use of PHWE for

the analysis of α-asarone in the dry rhizome of the common TCM Acorus

Tatarinowii Schott [93], Z-ligustilide and E-ligustilide, from Ligusticum

chuanxiong and A. sinensis [94, 95], and volatile active compounds such as

Fructus amomi [96,97]. Ong et al. investigated the application of PHWE for

extraction of berberine in coptidis rhizoma, glycyrrhizin in radix

glycyrrhizae/liquorice and baicalein in scutellariae radix [98] and compared PLE

with PHWE for the extraction of thermally labile components such as tanshinone I

and IIA in Salvia miltiorrhiza [99]. The application of PHWE for bioactive or

marker compounds in botanicals and medicinal plant materials was reviewed

[100].

1.2.1.6 Microwave Distillation (MD)

Microwave distillation (MD), which is a combination of microwave heating and


10

dry distillation at atmospheric pressure was invented by Chemat et al. in 2003

[101]. MD involves placing plant material in a microwave reactor, without any

added solvent or water. The internal heating of the water within the plant material

distends the plant cells and leads to rupture of the glands and oleiferous

receptacles. This frees essential oils which are evaporated by the water of the plant

material. A cooling system outside the microwave oven condenses the distillate.

The excess of water was refluxed to the extraction vessel in order to restore the

water to the plant material [102]. Lucchesi et al. reported the application of MD to

extract essential oils from aerial parts of three aromatic herbs: basil (Ocimum

basilicum L.), garden mint (Mentha crispa L.) and thyme (Thymus vulgaris L.)

[103]. Recently, Wang et al. applied the MD technique for the extraction of

essential oils from Cuminum cyminum L. and Zanthoxylum bungeanum Maxim

[104]. MD combined with headspace techniques, such as HS-SPME and

HS-LPME, had been successfully applied for the extraction and concentration of

volatile compounds from TCMs [105–107] using a short analysis time and no

solvent.

1.2.2 Analysis of TCMs by Chromatographic Techniques

As mentioned above, TCMs are complex mixtures containing up to hundreds or

even thousands of different components with significant difference in the content

and physical and chemical properties. However, only a few compounds are

responsible for the pharmaceutical and/or toxic effects. The large numbers of

other components in the TCM make the screening and analysis of the bioactive

components extremely difficult. Consequently, many methods, such as GC/MS,

HPLC/MS, CE, etc., have been proposed in recent years. In addition, hyphenated

instruments, such as GC/MS, HPLC/MS, etc., combining a chromatographic


11

separation system on-line with a spectroscopic detector in order to obtain

structural information on the analytes, have great potential in analyzing herbal

medicines. These instruments generate huge amounts of data. The

chromatographic profile of a complex mixture such as TCM extracts almost

always contained overlapped peaks. These overlapped peaks hinder the

identification of chemical components, as pure spectra of the corresponding

components cannot be obtained. The chemometric techniques are required to

retrieve the information these data contained.

1.2.2.1 Analysis of TCMs by GC

As we all know that a lot of therapeutic components in TCMs are volatile. Hence,

GC is a very important and useful technique for the analysis of these volatile

compounds in the past decades [108-114] due to its advantages: (1) both the

characteristic compounds of the particular plants and impurities can be detected,

(2) with the development of the extraction of volatile oil techniques, the

pharmacologically active components can be possibly identified using GC/MS

analysis, (3) the high sensitivity of GC and/or GC/MS ensures the detection for

almost all the volatile chemical compounds, (4) many volatile compounds can be

separated simultaneously within comparatively short times due to the high

selectivity of capillary columns. Until recently, MS is the most sensitive and

selective method for molecular analysis which can yield information on the

molecular weight as well as the structure of the molecule. The coupling of GC

with MS provides the advantages of both chromatography as a separation method

and MS as an identification method. For GC/MS, electron impact (EI) is primarily

configured to select positive ions from the analytes, while electron capture

ionization is usually configured for negative ions. GC/MS has been the first


12

successful online combination of chromatography with MS, and has been widely

applied for the analysis of essential oil in herbal medicines [115– 127]. With the

help of GC/MS, not only a separated chromatographic profile of the essential oil

of the herbal medicine could be obtained but also the information related to its

most qualitative and relative quantitative composition could be produced

[128–134]. For example, the combination of PHWE, HS-SPME and GC/MS was

successfully applied for the determination of three volatile compounds of

eucalyptol, camphor, and borneol in Chrysanthemum flowers [129] and

Z-ligustilide and E-ligustilide [130]. HS-SPME followed by GC/MS was also

applied to determine asarone in rabbit plasma at different time points after oral

adminstration of the essential oil from A. tatarinowii [131].

1.2.2.2 Analysis of TCMs by LC

With the advantages of high reproducibility, good linear range, ease of automation

and ability to analyze the number of constituents in botanicals and herbal

preparation, liquid chromatography (LC) with an isocratic/gradient elution

remains to be the primary method of choice in the analysis of TCMs. For LC, the

reversed octadecyl silica (C18) is one of the most commonly used columns. Ong

reported that columns with smaller inner diameter, such as 1.0 or 2.1 mm i.d. were

well suited to the analysis of components present in botanicals. Most important of

all, methods using columns with smaller inner diameter and the right mobile phase

can be readily adopted to mass spectrometry [135]. Applications of LC method to

medicinal plants and Chinese traditional medicines are outlined [136]. Methods

using gradient elution HPLC coupling with reversed phase columns had been

applied for the analysis of multiple constituents present in medicinal plants and

herbal preparations [137–139].


13

Until recently, the most common mode of detection remains to be ultraviolet (UV)

detection. Gradient elution HPLC with UV detection, using a C18 reversed phase

column had been successfully applied to profile components present in C. rhizoma,

Radix aristolochiae, ginseng, R. glycyrrhizae (liquorice), S. radix, R. codonopsis

pilosula and S. miltiorrhiza [140-144]. And due to the complexity of the matrix,

co-eluting peaks were often observed in the chromatograms obtained from the

analysis of marker compounds in herbal preparations with two or more medicinal

plants [145]. The co-eluting peaks might be reduced by additional sample

preparation steps, such as liquid-liquid partitioning, solid phase extraction,

preparative LC and TLC fractionation.

In recent years, there is a dramatical increase in the study of LC/MS for analysis

of TCMs due to the low sensitivity and specificity of UV detection. For

HPLC/MS, the most common mode of sample ionization includes ESI and

chemical API.

The LC-ESI/MS technique has been used for the analysis of 17 compounds and

their plant derivations of Xue-Fu-Zhu-Yu decoction, consisting of six crude drugs

[146], for the identification and quantification of paeonol [147], paeoniflorin [148],

oxysophocarpine and its active metabolite sophocarpine [149] in rat plasma, for

the simultaneous determination of tanshinone I, dihydrotanshinone I, tanshinone

IIA and cryptotanshinone, the active components of Salvia miltiorrhiza in rat

plasma [150], and for determination of oxymatrine and matrine in beagle dog

plasma [151]. It has also been successfully applied for the analysis of chemical

and metabolic components in TCM combined prescription containing Radix

Salvia miltiorrhiza, for quantification of pinane monoterpene glycosides in Cortex

Moutan [152] and Radix Panax notoginseng [153] and the determination of


14

adonifoline, a retronecine-type hepatotoxic pyrrolizidine alkaloid in Senecio

scandens Buch.–Ham. ex D. Don [154].

The chemical API source has been employed to tackle the matrix interference in

analyzing Chinese medicinal materials and to minimize the associated matrix

effects that were commonly encountered with other ionization modes. Moreover,

the method allowed direct interface to conventional HPLC systems [155]. The

HPLC/MS method using a chemical API source has been applied for separation

and identification of four major bioactive sesquiterpene alkaloids (Wilfortrine,

wilfordine, wilforgine and wilforine) in Tripterygium wilfordii Hook. F. [156], for

the biological fingerprinting analysis of bioactive components in a TCM

prescription, Longdan Xiegan Decoction (LXD) [157], and for the differentiation

of three ginseng species: Panax quinquefolium (American ginseng), P. ginseng

(Chinese ginseng) and P. notoginseng (sanqi) species [158].

Recently, the technique of Ultra Performance-LC (UPLC) has been developed.

With UPLC, the time and solvent consumption can be decreased. This new

technique combined with MS has been used for chemical profiling of Epimedium

brevicornum Maxim., as well as endogenous metabolite profiles of rats pre- and

post-hydrocortisone interfered and treated with this herbal medicine [159] and for

analysis of many other TCMs [160-168].

One of the challenges for the analytical methods developed was to study the

effects of batch-to-batch variations in the medicinal plants [135, 141, 169-170].

Ong et al. investigated the batch-to-batch variations of the plant materials. The

PLE-LC-UV method was used to assay the content of baicalein in S. radix from

four different sources [170]. They concluded that the assayed of markers or active

compounds together with chemical fingerprinting, using HPLC, would be able to


15

provide further information about the quality of the botanicals and herbal

preparations.

1.2.2.3 Analysis of TCMs by Capillary Electrophoresis (CE)

Capillary Electrophoresis (CE) was introduced in the early 1980s as a powerful

analytical and separation technique [230]. Until recently, several modes of CE are

available: (i) capillary zone electrophoresis (CZE), (ii) micellar electrokinetic

chromatography (MEKC), (iii) capillary gel electrophoresis (CGE), (iv) capillary

isoelectric focusing (CIEF), (v) capillary isotachophoresis (CITP), (vi) capillary

electrochromatography (CEC) and (vii) non-aqueous capillary electrophoresis

(NACE). The simplest and most versatile CE mode is CZE, in which the

separation is based on the differences in the charge-to-mass ratio and analytes

migrate into discrete zones at different velocities in the electrophoretic buffer in

the capillary. For MEKC, another most commonly used CE method, the main

separation mechanism is based on solute partitioning between the micellar phase

and the solution phase. The pH of running buffer, ionic strength, applied voltage,

concentration and type of micelle added and additives such as organic, ionic

liquids and nanostructures, and so on, are important parameters that can affect the

separation in CZE and MEKC. Compared with LC, CE shows several advantages:

(1) high separation efficiency, (2) specific selectively, (3) reduction of organic

solvent consumption, (4) small sample volume, (5) short analysis time, and (6)

low cost of accessories, such as the use of capillaries instead of more expensive

LC columns. As a result, CE has been proved to be a powerful alternative to LC in

the analysis of natural medicines or natural products in complex matrix.

Until recently, CE has been widely used for the analysis of TCMs. It has been

applied to determine the amount of catechin and others in tea composition,


16

phenolic acids in coffee samples and flavonoids and alkaloids in plant materials

[171-172], to differentiate between medicinal plant, such as S. radix from

Astragali radix [173], to determine aristolochic acids in R. aristolochiae,

strychnine in Strychnos nux-vomica, berberine in Rhizoma coptidis and

glycyrrhizin in R. glycyrrhizae [174-176], to analyse quaternary alkaloids of

Coptis chinensis [177], to estimate Synephrine level in Evodia fruit and eight

samples of TCM [178], to quantitate apigenin in Chamomilla recutita L.

Rauschert [179] and to determine jasminoidin, paeoniflorin and paeonol in jiawei

xiaoyao pills [180], bufadienolides in toad venom and their Chinese medicinal

preparations [181] as well as synephrine, hesperidin, naringenin and naringin in

Fructus anrantii Immaturus and Fructus aurantii [182]. CZE has also been widely

used to separate and determine four phenylpropanoid glycosides from T.

chamaedrys [183], to identify and determine ordycepin (3’-deoxyadenosine) in

Cordyceps kyushuensis Kob [184], to determine five active components in the

TCM preparation, Huangdan Yinchen Keli [185], rhein, baicalin and berberine in

TCM preparations, Sanhuangpian, Niuhuangjiedu-pian and

Huanglianshangqing-pian [186], quercetin, luteolin, kaempferol and isoquercitrin

in stamen nelumbinis [187] and genistein, rutin, baicalin and gallic acid in

Huaijiao pills [188].

Among the different modes of CE, MEKC is another commonly used method for

the analysis of TCMs. Until recently, it has been widely applied to determine

icariin, rhein, chrysophanol, physcion, glycyrrhetic acid and glycyrrhizic acid in

traditional Chinese herbal preparations [189], to determine three bioactive

compounds (andrographolide, deoxyandrographolide and neoandrographolide) in

Andrographis paniculata [190], to analyze two TCM preparations, Chuanxinlian


17

and Xiaoyan Lidan tablets [191], to determine hesperidin, naringin, puerarin and

daidzein in medicinal preparations [192-193], to analyze G. rhodantha, G. kitag,

G. scabra, G. rigescens, and G. macrophylla in Gentiana samples from Tibetian

medicines [194],to determine Gastrodin and tetramethylpyrazine in three TCM

preparations, Zhennaoning jiaonang, Yangxue shengfa and Xiaoshuan zaizaowan

[195], to identify and determine diterpenoid triepoxides in Tripterygium wilfordii

Hook F. and its preparations [196] and determine syringin and chlorogenic acid in

Acanthopanax senticosus from different parts [197]. In our group, we have

developed a new MEKC system with organic modifier for the analysis of

mutagenic pyrrolizidine alkaloids in TCMs [198].

NACE is a newly developed method as an alternative for the analysis of TCMs

due to its specific characterics. In our group, the determination of five toxic

alkaloids in aconitine root (Radix aconitini praeparata), seeds of Strychnos

pierrian and TCM preparation Shen Jin Huo Luo Wan by NACE was reported for

the first time [199]. Three aconitine alkaloids (hypoconitine, aconitine and

mesaconitine) and other unknown compounds coexisting in the five TCM,

Chuanwu, Caowu, Fuzi, Aconitum tanguticum Maxim. and Aconitum

gymnandrum were completely separated by non-aqueous CE within 13 min [200].

Li et al. reported the use of NACE for the determination of fangchinoline and

tetrandrine in Stephania tetrandra and Fengtongan capsule [201]. NACE has also

been successfully applied for the separation and determination of palmatine and

jatrorrhizine in Tinospora capillipes Gagn from different original areas [202].

It is clearly that CE has been proved to be a valuable tool for the analysis of

TCMs.


18

1.3 Metabonomics of TCMs

As far as TCMs is concerned, the toxicity is another important research area. All

the information mentioned above can only tell us the existence of certain

components, for the toxicity study, metabonomics will be introduced.

1.3.1 Metabonomics

Metabonomics is defined as “the quantitative measurement of the dynamic

multiparametric metabolic response of living systems to pathophysiological

stimuli or genetic modification” and aims at “the augmentation and

complementation of the information provided by measuring the genetic and

proteomic responses to xenobiotic exposure” [203].

Metabolite or metabolic profiling, the compositional analysis of low

molecular-weight species in biological samples, has existed for at least 35 years

already. Mass spectrometry (MS) coupled with some separation techniques such

as gas chromatography (GC) has been used for resolution and detection of

metabolites [204].

In other words, metabonomics encompasses “the comprehensive and simultaneous

systematic profiling of metabolite levels and their systematic and temporal

changes through such effects on diet, lifestyle, environment, genetics, and

pharmaceuticals, both beneficial and adverse, in whole organisms. This is

achieved by the study of biofluids and tissues with the data being interpreted using

chemometrics techniques” [205].

Recently, with the success of genomics and proteomics, a novel field name

metabolomics has been established. It is defined as the comprehensive

identification and quantification of the set of metabolites synthesized by an

organism, or metabolome [206-208]. Comparing metabolomics and


19

metabonomics, the former is concerned with the comprehensive metabolic

profiling at some scale, while the latter primarily focused on the history of time

dependent metabolic profiles in an integrated system.

For toxicity studies, the metabolic profiling of biofluids in addition to tissue

analysis can reveal the integrated physiological and not just tissue-specific

behavior of an organism. And multiple or continuous sampling through time from

the same individual is possible due to its minimally invasive character.

In short, information on in vivo multiorgan functional integrity in real time can be

obtained through metabonomics. Hence, it might provide the key to achieving a

“systems biology” approach to toxicology: the combination of genomic,

proteomic, and metabolic data from toxicological studies.

1.3.2 Metabonomics Samples

Biofluids or cell or tissue extracts are generally used for the metabonomics studies.

These samples are easy to collect. The mammalian biofluids can provide an

integrated view of the whole systems biology. Among the different kinds of

biofluids, urine and plasma are the most commonly used samples. In addition,

tissue biopsy samples and their lipid and aqueous extracts have been used for a lot

of metabonomics studies.

1.3.3 Metabonomics Analysis Technologies

There are many analytical techniques that can be used for metabonomics studies,

such as Nuclear Magnetic Resonance (NMR) spectroscopy , GC-MS, LC-MS,

CE-MS, Fourier transform infrared (FTIR), and so on. Metabonomics make use of

these instruments to detect the minute quantities of metabolites in the organisms.

Here, we will focus on the main methods: NMR, GC-MS and LC-MS.


20

1.3.3.1 NMR Spectroscopy

Historically, the definition of metabonomics has been NMR based. It arose from

the application of NMR to study the metabolic composition of biofluids, cells, and

tissues. As a nondestructive technique, NMR has been widely used in chemistry to

provide detailed information on molecular structure, both for pure compounds and

in complex mixtures [209]. It can also report on hundreds of compounds in a

single measurement with minimal sample preparation. In NMR spectra, individual

signals are dispersed according to the chemical environment of the source nuclei

and are directly proportional to the amount of material present. So NMR spectra

can provide abundant structural information and quantification basis. Therefore,

NMR spectroscopic methods are useful tools for probing metabolite molecular

dynamics and mobility as well as substance concentrations through the

interpretation of NMR spin relaxation times and by the determination of

molecular diffusion coefficients [210].

A typical 1H NMR spectrum of urine sample contains thousands of sharp peaks

from predominantly low molecular weight metabolites. The large interfering water

signal in NMR spectra of biofluids can be eliminated by use of appropriate

standard NMR water suppression method. The most commonly used reference

compound in aqueous media is the sodium salt of 3-trimethylsilylpropionic acid

(TSP) with methylene groups deuterated to avoid giving rise to peaks in the 1H

NMR spectrum.

Until recently, NMR has been successfully used for metabonomics research

[211-216]. For example, it has been used to compare metabonomics of differential

hydrazine toxicity in the rat and mouse [211], study the differentiation of gender,

diurnal variation and age in human urinary metabolic profiles [214], and so on.


21

1.3.3.2 Mass Spectrometry

Although NMR has its effectiveness, it suffers from two major drawbacks: poor

sensitivity and resolution. While mass spectrometry coupling with

chromatographic techniques such as GC or LC, can counteract these problems. A

recently introduced ultra-performance liquid chromatography (UPLC)-MS has

resulted in a highly sensitive and high resolution instrument that can also be used

for metabonomics studies. Due to the much improved chromatographic resolution

of UPLC, the typical problem of ion suppression for MS, which can impair the

sensitivity to any particular chemical species, could be greatly reduced.

Until recently, MS has been widely applied for metabonomics studies on plant

extracts, model cell system extracts and mammalian cells. For plant metabolic

studies by GC/MS, chemical derivatization has been used to ensure volatility and

analytical reproducibility for most investigations. For metabonomics applications

on biofluids such as urine, LC/MS with electrospray ionization is commonly used.

Both the positive and negative ion chromatograms would be measured. In a full

mass spectrum, each sampling point is three dimensional in nature, i.e., retention

time, mass, and intensity. We can select any peaks we want, or cut out any mass

peaks from interfering substances without affecting the integrity of the data set.

Mass Spectrometry has been successfully applied to many metabonomics studies

so far [217-222]. For instance, it has been used to study profiling of serum fatty

acids from Type 2 diabetic patients [217], liver diseases [219], within-day

reproducibility of human urine samples [221], and so on.

In summary, NMR and MS are complementary methods. Normally, it is necessary

to use both of them for full molecular characterization. MS can provide high

sensitivity given the analytes can be ionized, while NMR spectroscopy can


22

distinguish isomers, obtain molecular conformation information and study

molecular dynamics and compartmentation. There are also many studies reporting

the use of a combination of both MS and NMR or more analytical techniques

[223-227]. In our group, metabonomics of human urine after ingestion of green

tea was investigated through the combination of GC/MS, LC/MS and NMR [225]

and combination of NMR and LC/MS was used to evaluate metabolic profiles of

patients with albuminuria [226].

1.3.4 Metabonomics Data Analysis

An NMR or MS spectrum of a biofluid sample can be considered as an object

with a multidimensional set of metabolic coordinates, whose values are the

spectral intensities at each data point. As a result, the spectrum is a point in a

multidimensional metabolic hyperspace. The initial objectives of metabonomics

are: (1) to classify a spectrum based on identification of its inherent patterns of

peaks and (2) to identify those spectral features responsible for the classification.

Therefore, reducing the dimensionality of complex data sets is important in

metabonomics to enable easy visualization of any clustering or similarity of the

various samples.

Principle component analysis (PCA) may be the most extensively used

multivariate statistical technique in metabonomics. It can express most of the

variance within a data set through a smaller number of factors or principal

components (PC). Each PC is orthogonal and independent of other PCs. Since the

important PCs can describe the noise variation in the spectra simply, the variation

in the spectral set is usually described by fewer PCs. Each PC consists of “scores”

and “loadings”, where “scores” means a set of values defining the position of each

sample in the new coordinate space and “loadings” defines a set of values giving


23

the relative contributions of each variable in calculating the scores. As a result, the

multidimensional metabolic coordinates can be approximated by totaling the

products of each set of scores by the corresponding set of loadings.

Since PCA provides a reduced dimensional model that summarizes the major

variation in the data into a few axes, it can be used to monitor the systematic

variation, which can visualize the similar or dissimilar samples in a data set and

the spectral regions, corresponding to biomarkers, that causes the

treatment-related separation. Due to this important value, PCA has been exploited

extensively in metabonomic toxicology [228, 229]. PCA can also be used for

classification through soft independent modeling of class analogy [230, 231].

Firstly, an individual PCA model is established for each class of samples,

followed by fitting the new samples to each model and classifying to the

model/class that produces the lowest residual.

Partial least squares or projection to latent structures (PLS), a widely used

supervised method, derives a model that describes the correlation between the data

matrix containing independent variables from samples, such as spectral intensity

values (an X matrix) and the matrix containing dependent variables (e.g.,

measurements of response, such as toxicity scores) for those samples (a Y matrix).

It can be used to predict the dependent variables when presented with new data

and inform us about importance of each variable in prediction. PLS combined

with discriminant analysis (DA) can also be applied to classification problems.

1.4 Research Scope

The primary objective of this dissertation is to expand the analytical applicability

of sample preparation methods and capillary electrophoresis to Chinese herbal

medicines and develop a targeted data processing approach for the metabolic


24

profiling of alkaloids in rat model with multiple analytical approaches.

Since there are so many kinds of TCMs around the world and various research

directions on TCMs, it is impossible to cover every aspect in this dissertation.

Instead, two kinds of alkaloids existing in TCMs were selected for this study.

Pyrrolizidine alkaloids (PAs), widely distributed in TCMs around the world, was

investigated first. Different sample preparation methods such as

microwave-assisted extraction (MAE), pressurized hot water extraction (PHWE)

and heating under reflux were combined with LC/MS for isolation and

determination of PAs. All the key factors affecting the extraction efficiencies were

optimized. The results of this study provide an alternative tool for sample

preparation of natural products in pharmaceutical industries or natural products

research.

Then a simple, economical and efficient NACE system was developed for the

separation and determination of PAs in Chinese herbal medicine. The NACE

conditions including buffer consistence and separation voltage were optimized.

Subsequently, different on-line preconcentration methods of NACE were explored

to enhance the sensitivity. All the parameters that could affect the preconcentration

efficiency were optimized. The results of this study provide an alternative method

for quality control in pharmaceutical industries to monitor those toxic compounds

in natural products which would be harmful to human health.

Berberine, a commonly used TCM in China, was chosen for the toxicity study in

rat model. The metabolic profiling of berberine was investigated by NMR and

LC/MS of urine samples, and GC/MS of tissue extracts. A new data processing

method was used for PCA. The results of this study provided further insights on

the potential therapeutic application of berberine.


25

The entire dissertation will be divided into two parts: Part I will focus on the

separation and analysis of pyrrolizidine alkaloids in TCM, and Part II will focus

on metabonomic studies of berberine in the rat model.


26

References

[1] S. Bent, R. Ko, Am. J. Med. 116 (2004) 478.

[2] Committee of National Pharmacopoeia, Pharmacopoeia of PR China, Press of

Chemical Industry, Beijing, 1995

[3] Committee of Jiangsu New Medical College, Encyclopedia of Traditional

Chinese Medicine, Shanghai Science and Technology Press, Shanghai, 1995.

[4] P. B. Kaufman, L. J. Csake, S. Warber, J. A. Duke, H. L. Brielmann, (eds)

Natural products from plants. CRC Press, Boca Raton (1999)

[5] H. Wang, H. Zou, L. Kong, J. Ni, J. Basic Clic. Physiol. Pharmacol.11 (2000)

155.

[6] K. Y. Liu, T. J. Zhang,W.Y. Gao, H. X. Chen, Y. N. Zheng, Zhongguo

Zhongyao Zazhi 31 (2006) 1837.

[7] C. L. Fan, Faming Zhuanli Shenqing Gongkai Shuomingshu, (2007) CN

101036491.

[8] Y. M. Shi, Faming Zhuanli Shenqing Gongkai Shuomingshu, (2002) CN

1381252.

[9] S. X. Wang, Faming Zhuanli Shenqing Gongkai Shuomingshu, (2002) CN

1931022.

[10] H. S. Lee, M. C. Rho, Y. K. Kim, M. Y. Chung, S. W. Lee, H. R. Park, J. H.

Choi, J. Y. Kang, D. H. Kim, Repub. Korean Kongkae Taeho Kongbo (2007) KR

2007008228.

[11] Awasthy, Narendra Kumar, Indian Patent Application (2006) IN

2004DE01564.

[12] H. R. Park, M. Y. Yoo, J. H. Seo, I. S. Kim, N. Y. Kim, J. Y. Kang, L. Cui, C.

S. Lee, C. H. Lee, H. S. Lee, J. Agric. Food Chem. 56 (2008) 10493.


27

[13] L. Perharic, D. Shaw, M. Colbridge, I. House, C. Leon, V. Murray, Drug

Safety, 11 (1994) 284.

[14] M. D. Luque de Castro, M. M. Jiménez-Carmona, V. Fernández-Pérez,

Trends Anal. Chem. 18 (1999) 708.

[15] C. L. Arthur, J. Pawliszyn, Anal. Chem. 62 (1990) 2145.

[16] H. Lord, J. Pawliszyn, J. Chromatogr. A 885 (2000) 153.

[17] J. A. Field, G. Nickerson, D. D. James, C. Heider, J. Agric. Food Chem. 44

(1996) 1768.

[18] M. Kovacevic, J. Kac, J. Chromatogr. A 918 (2001) 159.

[19] J. Rohloff, Phytochemistry 59 (2002) 665.

[20] G. Flamini, P. L. Cioni, I. Morelli, J. Agric. Food Chem. 50 (2002) 4647.

[21] L. N. Fernando, I. U. Grun, Flav. Frag. J. 16 (2001) 289.

[22] L. Jirovetz, G. Buchbauer, M. B. Ngassoum, M. Geissler, J. Chromatogr. A

976 (2002) 265.

[23] C. H. Deng, G. X. Song, Y. M. Hu, X. M. Zhang, Chromatographia 57 (2003)

35.

[24] C. H. Deng, G. X. Song, Y. M. Hu, X. M. Zhang, Chromatographia 58 (2003)

289.

[25] G. X. Song, C. H. Deng, D. Wu, Y. M. Hu, Chromatographia 58 (2003) 769.

[26] G. X. Song, C. H. Deng, Y. M. Hu, Chromatographia 59 (2004) 343.

[27] S. Shen, Y. F. Sha, C. H. Deng, X. M. Zhang, D. X. Fu, J. K. Chen, J.

Chromatogr. A 1047 (2004) 281.

[28] Y. F. Sha, S. Shen, G. L. Duan, Z. Naturf. C 59 (2004) 533.

[29] Y. F. Sha, S. Shen, G. L. Duan, J. Pharm. Biomed. Anal. 36 (2004) 381.

[30] F. Q. Guo, L. F. Huang, Anal. Chim. Acta 570 (2006) 73.


28

[31] J. Cao, M. L. Qi, J. Pharm. Biomed. Anal. 40 (2006) 552.

[32] M. M. Liang, M. L. Qi, C.B. Zhang, Anal. Chim. Acta 531 (2005) 97.

[33] M. A. Jeannot, F. F. Cantwell, Anal. Chem. 68 (1996) 2236.

[34] M. A. Jeannot, F. F. Cantwell, Anal. Chem. 69 (1997) 235.

[35] Y. He, H. K. Lee, Anal. Chem. 69 (1997) 4634.

[36] L. S. DeJager, A. R. J. Andrews, Chromatographia 50 (1999) 733.

[37] L. S. DeJager, A. R. J. Andrews, Analyst 125 (2000) 1943.

[38] E. Psillakis, N. Kalogerakis, J. Chromatogr. A 907 (2001) 211.

[39] H. Y. Zhang, A. R. J. Andrews, J. Environ. Monit. 2 (2000) 656.

[40] L. S. DeJager, A. R. J. Andrews, J. Chromatogr. A 911 (2001) 97.

[41] A. L. Theis, A. J. Waldack, S. M. Hansen, M. A. Jeannot, Anal. Chem. 73

(2001) 6561.

[42] J. Cao, M. L. Qi, Anal. Chim. Acta 561 (2005) 88.

[43] R. M. Smith, M. D. Burford, J. Chromatogr. 600 (1992) 175.

[44] J. L. Hedrick, L. J. Mulcahey, L. T. Taylor, Mikrochim. Acta 108 (1992) 115.

[45] Q. H. Hu, Y. Hu, J. Xu, Food Chem. 91 (2005) 85.

[46] W. L. Yu, B. Shu, Y. P. Zhao, J. Sci. Food Agric. 85 (2005) 489.

[47] H. Mi, L. L. Qu, Y. L. Ren, Chin. J. Anal. Chem. 33 (2005) 366.

[48] X. W. Liu, Z. Y. Li, S. H. Chen, Chin. J. Chem. Eng. 13 (2005) 572.

[49] S. Zhao, H. Liang, Pol. J. Chem. 80 (2006) 99.

[50] B. Liu, H. L. Jiang, B. Shen, Y. L. Chang, J. Chromatogr. A 1075 (2005)

213.

[51] B. Liu, W. J. Li, Y. L. Chang, W. H. Dong, L. Ni, J. Pharm. Biomed. Anal.

41 (2006) 1056.

[52] J. Zhang, J. C. Duan, Z. Liang, W. B. Zhang, L. H. Zhang, Y. S. Huo, Y. K.


29

Zhang, Chin. J. Anal. Chem. 34 (2006) 447.

[53] C. M. Liu, X. L. Zhao, H. M. Li, Z. Q. Liu, J. P. Xing, Chem. Res. Chin.

Univ. 21 (2005) 757.

[54] J. Zhang, J. C. Duan, Z. Liang, L. H. Zhang, W. B. Zhang, Y. S. Huo, Y. K.

Zhang, Chem. J. Chin. Univ. 26 (2005) 1832.

[55] J. Zhang, L. H. Zhang, J. C. Duan, Z. Liang, W. B. Zhang, Y. S. Hu, Y. K.

Zhang, J. Sep. Sci. 29 (2006) 2514.

[56] J. Y. Peng, G. R. Fan, Y. T. Wu, J. Chromatogr. A 1115 (2006) 103.

[57] J.Y. Peng, G. R. Fan, Y. T. Wu, J. Chromatogr. A 1102 (2006) 44.

[58] J. Y. Peng, G. R. Fan, Y. T. Wu, J. Chromatogr. A 1083 (2005) 52.

[59] X. Wang, Y. Q. Wang, J. P. Yuan, Q. L. Sun, J. H. Liu, C. C. Zhen, J.

Chromatogr. A 1055 (2004) 135.

[60] M. Vinatoru, Ultrason Sonochem 8 (2001) 303

[61] Q. Fang, H. W. Yeung, H. W. Leung, C. W. Huie, J. Chromatogr. A 904

(2000) 47.

[62] E. Martino, E. Ramaiola, M. Urbano, F. Bracco, S. Collina, J. Chromatogr. A

1125 (2006) 147.

[63] W. Shi, Y. T. Wang, X. J. Quan, L. F. Bai, H. R. Zhang, X. D. Chen, L. Ding,

H. Q. Zhang, Chin. J. Anal. Chem. 34 (2006) 243.

[64] F. Zhang, B. Chen, S. Xiao, S. Z. Yao, Sep. Purif. Technol. 42 (2005) 283.

[65] M. J. Li, J. Y. You, S. Yao, L. Ding, Z. Y. Liu, H. Q. Zhang, Chem. Res.

Chin. Univ. 20 (2005) 703.

[66] S. Gao, W. Han, X. Deng, Flav. Frag. J. 19 (2004) 244.

[67] Z. Y. Liu, G. Q. Yan, F. Q. Bu, J. Sun, M. J. Li, L. Ding, H. Q. Zhang, Chin.

J. Anal. Chem. 32 (2004) 360.


30

[68] J. T. He, Z. H. Shi, W. B. Chang, J. Liq. Chromatogr. Rel. Technol. 27 (2004)

1769.

[69] Z. Y. Liu, G. Q. Yan, F. Q. Bu, J. Sun, X. L. Hu, H. Q. Zhang, Z. Q. Liu,

Chem. Analityczna 50 (2005) 851.

[70] C. H. Deng, Y. Mao, N. Yao, X. M. Zhang, Anal. Chim. Acta 575 (2006)

120.

[71] C. H. Deng, J. Ji, N. Li, Y. J. Yu, G. L. Duan, X. M. Zhang, J. Chromatogr. A

1117 (2006) 115.

[72] C. H. Deng, N. Yao, B. Wang, X. M. Zhang, J. Chromatogr. A 1103 (2006)

15.

[73] X. Wang, Z. Zhuang, D. Sun, J. Hong, X. Wu, F. S. C. Lee, M. S. Yang, H.

W. Leung (1999) At Spectrosc 20:86.

[74] E. S. Ong, Y. L. Yong, S. O. Woo, L. K. Kee , Anal. Sci. 16 (2000) 391.

[75] E. S. Ong, Y. L. Yong, S. O. Woo, J AOAC 83 (2000) 382.

[76] E. S. Ong, Y. L. Yong, S. O. Woo, J AOAC 82 (1999) 963.

[77] B. Benthin, H. Danz, M. Hamburger, J. Chromatogr. A 837 (1999) 211.

[78] L. Ramos, E. M. Kristenson, U. A. Th. Brinkman, J. Chromatogr. A 975

(2002) 3.

[79] Y. X. Gong, S. P. Li, Y. T. Wang, P. Li, Electrophoresis 26 (2005) 1778.

[80] F. Q. Yang, S. P. Li, Y. Chen, S. C. Lao, Y. T. Wang, T. T. X. Dong, K. W.

K. Tsim, J. Pharm. Biomed. Anal. 39 (2005) 552.

[81] F. Q. Yang, Y. T. Wang, S. P. Li, J. Chromatogr. A 1134 (2006) 226.

[82] J. X. Wei, Y. C. Du, Modern Science Research and Application of Panax

Notoginseng, Yunnan Science and Technology Press, Kunming, China, 1996.

[83] J. B. Wang, P. Li, S. P. Li, Y. T. Wang, T. T. X. Dong, K. W. K. Tsim, J. Sep.


31

Sci. 29 (2006) 2190.

[84] J. B. Wan, F. Q. Yang, S. P. Li, Y. T. Wang, X. M. Cui, J. Pharm. Biomed.

Anal. 41 (2006) 1596.

[85] J. B. Wan, C. M. Lai, S. P. Li, J. Pharm. Biomed. Anal. 41 (2006) 274.

[86] H. Fan, S. P. Li, J. J. Xiang, C. M. Lai, F. Q. Yang, J. L. Gao, Y. T.Wang,

Anal. Chim. Acta 567 (2006) 218.

[87] P. Li, S. P. Li, S. C. Lao, C. M. Fu, J. Pharm. Biomed. Anal. 40 (2006) 1073.

[88] J. Zhao, S. P. Li, F. Q. Yang, P. Li, Y. T.Wang, J. Chromatogr. A 1108 (2006)

188.

[89] Y. Jiang, S. P. Li, J. Chromatogr. A 1108 (2006) 268.

[90] E. S. Ong, Y. L. Yong, S. O. Woo, J Chromatogr A 313 (2000) 57

[91] E. S. Ong, J. Sep. Sci. 25 (2002) 825.

[92] V. Fernandez-Perez, M. M. Jimenez-Carmona, M. D. Luque de Castro,

Analyst 125 (2000) 481.

[93] C. H. Deng, N. Li, X. M. Zhang, J. Chromatogr. A 1059 (2004) 149.

[94] C. H. Deng, J. Ji, X. C. Wang, X. M. Zhang, J. Sep. Sci. 28 (2005) 1237.

[95] C. H. Deng, A. Q. Wang, S. Shen, J. Pharm. Bioanal. Anal. 38 (2005) 326.

[96] C. H. Deng, N. Yao, X. M. Zhang, Anal. Chim. Acta 536 (2005) 237.

[97] C. H. Deng, X. H. Yang, X. M. Zhang, Talanta 68 (2005) 6.

[98] E. S. Ong, S. M. Len, Anal. Chim. Acta 482 (2003) 81.

[99] E. S. Ong, S. M. Len, J. Chromatogr. Sci. 42 (2004) 211.

[100] E. S. Ong, J. S .H. Cheong, D. Goh, J. Chromatogr. A 1112 (2006) 92

[101] F. Chemat, J. Smadja, M. E. Lucchesi, European Patent Demand,

03001183.7, 2003.

[102] C. H. Deng, N. Liu, M. X. Gao, X. M. Zhang, J. Chromatogr. A, 1153 (2007)


32

90.

[103] M. E. Lucchesi, F. Chemat, J. Smadja, J. Chromatogr. A 1043 (2004) 323.

[104] Z. M. Wang, L. Ding, T. C. Li, X. Zhou, L. Wang, H. Q. Zhang, L. Liu, Y.

Li, Z. H. Liu, H. J. Wang, H. Zeng, H. He, J. Chromatogr. A 1102 (2006) 11.

[105] C. H. Deng, X. Q. Xu, N. Yao, N. Li, X. M. Zhang, Anal. Chim. Acta 556

(2006) 289.

[106] N. Li, C. H. Deng, Y. Li, H. Ye, X. M. Zhang, J. Chromatogr. A 1133 (2006)

29.

[107] C. H. Deng, Y. Mao, F. L. Hu, X. M. Zhang, J. Chromatogr. A 1131 (2006)

45

[108] P. Majlat, J. Chromatogr. 241 (1982) 399.

[109] C. J. Briggs, K. J. Simons, J. Chromatogr. 257 (1983) 132.

[110] G. J. Soleas, E. P. Diamandis, A. Karumanchiri, D. M. Goldberg, Anal.

Chem. 69 (1997), 4405.

[111] R. J. Molyneux, N. Mahoney, P. Bayman, R. Y. Wong, et al., J. Agric. Food

Chem. 50 (2002) 1393.

[112] F. Angerosa, N. Alessandro, F. Corana, G. Mellerio, J. Chromatogr. A 736

(1996) 195.

[113] M. Bunzel, J. Ralph, J. M. Marita, R. D. Hatfield, H. Steinhart, J. Sci. Food

Agric. 81 (2001) 653.

[114] M. Ylinen, T. Naaranlahti, S. Lapinjoki, A. Huhtikangas, et al., Planta Med.

52 (1986) 85.

[115] G. Guetens, G. De Boeck, M. Wood, R. A. A. Maes, et al., J. Chromatogr.

A 976 (2002) 229.

[116] F. Gong, Y. Q. Song, Y. G. Peng, Y. Z. Liang, et al., Acta Pharm. Sin. 35


33

(2000) 394.

[117] F. Gong, Y. Liang, A. K. M. Leung, F. T. Chau, Chin. Anal. Chem. 28

(2000) 860.

[118] F. Gong, Y. Z. Liang, C. Hui, F. T. Chau, B. T. P. Chan, J. Chromatogr. A

905 (2001) 193.

[119] F. Gong, Y. Z. Liang, Q. S. Xu, F. T. Chau, J. Chromatogr. A 909 (2001)

237.

[120] X. N. Li, H. Cui, Y. Q. Song, Y. Z. Liang, Acta Pharm. Sin. 36 (2001) 215.

[121] F. Gong, Y. Z. Liang, Y. Q. Song, Y. G. Peng, et al., Chem. J. Chin. Univ.

22, (2001)1481.

[122] X. N. Li, Y. Z. Liang, Phytochem. Anal. 14 (2003) 23.

[123] F. Gong, Y. Z. Liang, J. Sep. Sci. 26 (2003) 112.

[124] B. Y. Li, Y. Z. Liang, Chromatogrphia 57 (2003) 235.

[125] P. P. Fu, Y. C.Yang, Q. S. Xia, M. W. Chou, et al., J. Food Drug Anal. 10

(2002) 198.

[126] M. C. Tseng, M. J. Tsai, J. H. Lin, K. C. H. Wen, J. Food Drug Anal. 8

(2000) 315.

[127] C. Gherman, M. Culea, O. Cozar, Talanta 53 (2000) 253.

[128] K. K. To, Y. P. Ho, S. C. Au-Yeung, J. Chromatogr. A 947 (2002) 319.

[129] L. Dong, J. Y. Wang, C. H. Deng, X. Z. Shen, J. Sep. Sci. 30 (2007) 86.

[130] C. H. Deng, L. Ji, X. C. Wang, X. M. Zhang, J. Sep. Sci. 28 (2005) 1237.

[131] C. H. Deng, S. Lin, T. M. Huang, G. G. Duan, X. M. Zhang, Rapid

Commun. Mass Spectrom. 20 (2006) 2120.

[132] A. Velasco-Negueruela, M. J. Perez-Alonso, P. L. Perez de Paz, J. Pala-Paul,

J. Sanz, J. Chromatogr. A 984 (2003) 159.


34

[133] A. Velasco-Negueruela, M. J. Perez-Alonso, P. L. Perez de Paz, J. Pala-Paul,

J. Sanz, J. Chromatogr. A 1011(2003) 241.

[134] Y. P. Li, Z. Hu, L.C. He, J. Pharm. Biomed. Anal. 43 (2007) 1667.

[135] E. S. Ong, J. Chromatogr. B 812 (2004) 23.

[136] F. Li, S. Sun, J. Wang, D. Wang, Biomed. Chromatogr. 12 (1998) 78.

[137] S. Dheu, K. Li, J. High Resol. Chromatogr. 21 (1998) 569.

[138] S. Li, S. S.K. Chan, G. Lin, L. Ling, R. Yan, H. S. Chung, Y. K. Tam,

Planta Medica 69 (2003) 445.

[139] X. Zhang, M. Yu, J. Chen, J. Chromatogr. Sci. 41 (2003) 241.

[140] E. S. Ong, S. O. Woo, Y. L. Yong, J. Chromatogr. A 904 (2000) 57.

[141] H. K. Lee, H. L. Koh, E. S. Ong, S. O. Woo, J. Sep. Sci. 25 (2002) 160.


[142] E. S. Ong, S. M. Len, J. Sep. Sci. 26 (2003) 1533.


[145] T. Saeki, K. Koike, T. Nikaido, Planta Medica 65 (1999) 428.

[146] H. J. Zhang, Y. Y. Cheng, Biomed. Chromatogr. 20 (2006) 821.

[147] Y. Xiao, Y. H. Zhang, Y. X. Sheng, J. L. Zhang, Biomed. Chromatogr. 25

(2008) 527.

[148] Q. Wang, H. Y. Yang, W. N. Liu, X. L. Feng, Biomed. Chromatogr. 20

(2006) 173.

[149] S. J. Wang, M. Yang, X. T. Li, G. J. Wang, Biomed. Chromatogr.21 (2007)

768.

[150] E. J. Park, H. Y. Ji, N. J. Kim, W. Y. Song, Biomed. Chromatogr. 22 (2008)

548.

[151] Y. Q. Wang, Y. Y. Ma, X. H. Li, F. Qin, Biomed. Chromatogr. 21 (2007)


35

876.

[152] Y. Q. Song, H. P. Li, Y. Y. Cheng, J. Sep. Sci. 31 (2008) 64.

[153] Y. J.Wei, P. Li, B. Shu, H. J. Li, Biomed. Chromatogr.21 (2007) 797.

[154] F. Zhang, C. H. Wang, W. Wang, L. X. Chen, Phytochem. Anal. 19 (2008)

25.

[155] S. Liu, L. Z. Yi, Y. Z. Liang, J. Sep. Sci. 31 (2008) 2113.

[156] X. K. Ouyang, M. C. Jin, C. H. He, Phytochem. Anal. 18 (2007) 320.

[157] Y. Wang, L. Kong, L. H. Hu, X. Y. Lei, J. Chromatogr. B 860 (2007) 185.

[158] K. S. Y. Leung, K. Chan, A. Bensoussan, M. J. Munroe, Phytochem. Anal.

18 (2007) 146.

[159] F. M. Li, X. M. Lu, H. P. Liu, M. Liu, Z. L. Xiong, Biomed. Chromatogr.

21 (2007) 397.

[160] J. L. Lu, D. M. Wang, A. Q. Miao, D. P. Yang, Chin. Pharm. J. 42, (2007)

1456.

[161] X. Zhang, F. R. Song, L. S. Wang, Z. Q. Liu, S. Y. Liu, Acta Chim. Sin. 65

(2007) 829.

[162] L. Wang, Y. L. Lu, D. Yang, Z. M. Xue, Y. P. Wang, J. Xi ,an Jiao tong

Univ. (Med. Sci.) 28 (2007) 395.

[163] W. J. Liao, H. Zhang, Y. P. Ren, Chin. J. Anal. Chem. 38 (2006) 1175.

[164] X. L. Wang, M. Cai, Y. Zhou, Nat. Prod. Res. Dev. 18 (2006) 624.

[165] D. M. Wang, L. L. Du, J. L. Lu, X. G. Shi, X. H. Song, Nat. Prod. Res. Dev.

18 (2006) 978.

[166] X. Zhang, F. R. Song, Z. Q. Liu, S. Y. Liu, Chem. J. Chin. Univ. 27 (2006)

1640.

[167] J. Q. Lu, Y. G. Liu, D. Liu, Y. J. Qiao, J. Beijing Uni. TCM 29 (2006) 270.


36

[168] S. P. Liang, F. Liu, F. F. Du, W. Niu, Y. Sun, Chin. J. Clin. Pharmacol. Ther.

11 (2006) 491.

[169] P. Molgaard, S. Johnsen, P. Christensen, C. Cornett, J. Agric. Food Chem.

51 (2003) 6922.


[171] H. J. Issaq, Electrophoresis 18 (1997) 2438.

[172] H. J. Issaq, Electrophoresis 20 (2000) 3190.

[173] G. Chen, X. Ying, J. Ye, Analyst 125 (2000) 815.

[174] E. S. Ong, S. N. binte Apandi, Electrophoresis 22 (2001) 2723.

[175] E. S. Ong, J. Sep. Sci. 25 (2002) 825.

[176] E. S. Ong, S. O. Woo, Electrophoresis 22 (2001) 2236.

[177] K. J. Dai, Q. Z. Luo, J. B. Luo, A. D. Ma, Diyi Junyi Daxue Xuebao 23

(2003) 683.

[178] S. Ihara, H. Shimoda, Y. Akiho, F. Yoshizaki, Nat. Med. 57 (2003) 110.

[179] F. N. Fonseca, M. F. Tavares, Phytochem. Anal. 15 (2004) 65.

[180] X. Zhou, T. Lu, Z. G. Chen, J. Zhang, Chin. J. Anal. Lab. 26 (2007) 5.

[181] Q. Su, Y. B. Wang, X. P. Luo, J. Beijing Univ. TCM 30 (2007) 203.

[182] Y. Y. Peng, J. N. Ye, J. Instr. Anal. 26 (2007) 694.

[183] B. Avula, R. B. Manyam, E. Bedir, I. A. Khan, Chromatographia 58 (2003)

751.

[184] Y. J. Sun, J. Y. Ling, P. Lü, H. X. Zhang, C. K. Zhang, Chin. Chem. Lett.

14 (2003) 724.

[185] A. Wang, Y. Zhou, F. Wu, P. He, Y. Fang, J. Pharm. Biomed. Anal. 35

(2004) 959.

[186] H. T. Liu, K. T. Wang, X. G. Chen, Z. D. Hu, Biomed. Chromatogr. 18


37

(2004) 288.

[187] R. J. Jing, X. Y. Jiang, S. R. Hou, X. J. Li, Z. B. Yuan, Chin. J. Anal. Chem.

35 (2007) 1187.

[188] S. Y. Gao, G. C. Dang, H. Li, Chin. J. Pharm. 38 (2007) 290.

[189] H. T. Liu, K. T. Wang, H. Y. Zhang, X. G. Chen, Z. D. Hu, Analyst 125

(2000) 1083.

[190] J. Zhao, G. Yang, H. Liu, D. Wang, X. Song, Y. Chen, Phytochem. Anal. 13

(2002) 222.

[191] S. Guo, Z. Wu, Yaowu Fenxi Zazhi 22 (2002) 149.

[192] Y. Cao, C. Lou, X. Zhang, Q. Chu, Y. Fang, J. Ye, Anal. Chim. Acta 452

(2002) 123.

[193] S. G. Zhao, X. G. Chen, Z. D. Hu, Chromatographia 57 (2003) 593.

[194] S. Zhao, Q. Liu, X. Chen, Z. Hu, Biomed. Chromatogr. 18 (2003) 10.

[195] R. Y. Wang, W. Y. Ma, D. Y. Chen, J. Sep. Sci. 26 (2003) 543.

[196] J. X. Song, G. Yang, J. Zhao, Y. Chen, H. Sun, H. Zhang, Phytochem. Anal.

14 (2003) 132.

[197] C. Y. Zheng, S. M. Liu, H. T. Li, W. Y. Niu, W. X. Ping, Chin. Pharm. J. 42

(2007) 1248.

[198] L. J. Yu, Y. Xu, H. T. Feng, S. F. Y. Li, Electrophoresis, 26 (2005) 3397.

[199] H. T. Feng, S. F. Y. Li, J. Chromatogr. A 973 (2002) 243.

[200] S. Zhao, Z. Pan, X. Chen, Z. Hu, Biomed. Chromatogr. 18 (2004) 381.

[201] Y. Q. Li, X. G. Chen, Y. X. Qi, K. Li, B. X. Jia, Chin. J. Chin. Maeria Med.

35 (2007) 1992.

[202] Y. W. Wang, K. Yu, Y. Y. Cheng, Chin. Pharm. J. 142 (2007) 935.

[203] J. K. Nicholson, J. C. Lindon, E. Holmes, Xenobiotica 29, (1999) 1181.


38

[204] E. C. Horning, Clin Chem 17 (1971) 802.

[205] J. C. Lindon, E. Holmes, J. K. Nicholson1, Pharmaceutical Research, 23

(2006), 1075

[206] S. G. Oliver, M. K. Winson, D. B. Kell, F. Baganz, Trends Biotechnol. 16

(1998), 373.

[207] O. Fiehn, Plant Mol. Biol. 48 (2002) 155.

[208] L. W. Sumner, P. Mendes, R. A. Dixon, Phytochemistry 62 (2003) 817.

[209] T. D. W. Claridge, Elsevier, Oxford, UK. ISBN 0080427995, 1999.

[210] M. Liu, J. K. Nicholson, J. C. Lindon, Anal. Chem. 68 (1996) 3370.

[211] M. E. Bollard, H. C. Keun, O. Beckonert, T.M.D. Ebbels, H. Antti, A. W.

Nicholls, J. P. Shockcor, G. H. Cantor, G. Stevens, J. C. Lindon, E. Holmes, J. K.

Nicholson, Toxicol. Appl. Pharmacol. 204 (2005) 135.

[212] V. Mäkinen, P. Soininen, C. Forsblom, M. Parkkonen, P. Ingman, K. Kaski,

P. Groop, M. Ala-Korpela, Mol. Sys. Biol. 4 (2007) 1.

[213] R. M. Salek, M. L. Maguire, E. Bentley, D. V. Rubtosov, T. Hough, M.

Cheeseman, D. Nunez, B. C. Sweatman, J. N. Haselden, R. D. Cox, S. C.Connor,

J. L. Griffin, Physiol. Genomics 29 (2007) 99.

[214] C. M. Slupsky, K. N. Rankin, J. Wagner, H. Fu, D. Chang, A. M. Weljie, E. J.

Saude, B. Lix, D. J. Adamko, S. Shah, R. Greiner, B. D. Skyes, T. J. Marrie, Anal.

Chem. 79 (2007) 6995.

[215] S. C. Zhang, G. A. N. Gowda, V. Asiago, N. Shanaiah, C. Barbas, D. Raftery,

Anal. Biochem. 383 (2008) 76.

[216] A. D. Maher, D. Crockford, H. Toft, D. Malmodin, J. H. Faber, M. I.

McCarthy, A. Barrett, M. Allen, M. Walker, E. Holmes, J. C. Lindon, J. K.

Nicholson, Anal.. Chem. 80 (2008) 7354.


39

[217] J. Yang, G. W. Xu, Q. F. Hong, H. M. Liebich, K. Lutz, R. M. Schmülling, H.

G. Wahl, J. Chromatogr. B 813 (2004) 53.

[218] C. Wang, H. W. Kong, Y. F. Guan, J. Yang, J. R. Gu, S. L. Yang, G. W. Xu,

Anal. Chem. 77 (2005) 4108.

[219] J. Yang, G. W. Xu, Y. F. Zheng, H. W. Kong, C. Wang, X. J. Zhao, T. Pang, J.

Chromatogr. A 1084 (2005) 214.

[220] P. Y. Yin, X. J. Zhao, Q. R. Li, J. S. Wang, J. S. Li, G. W. Xu, J. Proteome

Res. 5 (2006) 2135.

[221] H. G. Gika, G. A. Theodoridis, J. E. Wingate, I. D. Wilson, J. Proteome Res.

6 (2007) 3291.

[222] L. W. Jia, J. Chen, P. Y. Yin, X. Lu, G. W. Xu, Metabolomics 4 (2008) 183.

[223] E. M. Lenz, J. Bright, R. Knight, F. R. Westwood, D. Davies, H. Major, I. D.

Wilson, Biomarkers 10 (2005) 173.

[224] F. P. J. Martin, M. Dumas, Y. Wang, C. Legido-Quigley, I. K. S. Yap, H.

Tang, S. Zirah, G. M. Murphy, O. Cloarec, J. C. Lindon, N. Sprenger, L. B. Fay, S.

Kochhar, P. van Bladeren, E. Holmes, J. K. Nicholson, Mol. Sys. Biol. 3 (2007) 1.

[225] W. S. Law, P. Y. Huang, E. S. Ong, C. N. Ong, S. F. Y. Li, K. K. Psikanti, E.

C. Y. Chan, Rapid Commun. Mass Spectrom. 22 (2008) 2436.

[226] W. S. Law, P. Y. Huang, E. S. Ong, S. K. Sethi, S. Saw, C. N. Ong, S. F. Y.

Li, J. Proteome Res. 8 (2009) 1828.

[227] E. S. Ong, C. F. Chor, L. Zou, C. N. Ong, Mol. Biosyst. 5 (2009) 288.

[228] B. M. Beckwith-Hall, J. K. Nicholson, A. W. Nicholls, P. J. Foxall, J. C.

Lindon, S. C. Connor, Chem. Res. Toxicol. 11 (1998) 260.

[229] D. G. Robertson, M. D. Reily, R. E. Sigler, D. F. Wells, D. A. Paterson, T.

K. Braden, Toxicol. Sci. 57 (2000) 326.


40

[230] E. Holmes, A. W. Nicholls, J. C. Lindon, S. Ramos, M. Spraul, P. Neidig,

NMR Biomed. 11 (1998) 235.

[231] E. Holmes, A. W. Nicholls, J. C. Lindon, S. C. Connor, J. C. Connelly, J. N.

Haselden, Chem. Res. Toxicol. 13 (2000) 471.

41

Part I: Isolation and Determination of

Pyrrolizidine Alkaloids in

Traditional Chinese Medicine with

Modern Analytical Techniques

Chapter 2 Determination of Senkirkine and Senecionine in Tussilago Farfara using Microwave-assisted Extraction and Pressurized Hot Water Extraction with Liquid Chromatography Tandem Mass Spectrometry

42

Chapter 2 Determination of Senkirkine and Senecionine in Tussilago Farfara

using Microwave Assisted Extraction and Pressurized Hot Water Extraction

with Liquid Chromatography Tandem Mass Spectrometry

2.1 Introduction

Tussilago farfara (Kuan Donghua), the dried flower bud of T. farfara L., is an

important Chinese herbal medicine which has been commonly used for the relief

of coughs and as an expectorant, blood pressure raiser, platelet activating factor

and anti-inflammatory agent [1].It is also used for the treatment of asthma,

silicosis, pulmonary tuberculosis, obesity, type 2 diabetes, and hepatitis [2–7].

However, besides therapeutic bioactive compounds present in the herb, it has been

found to contain toxic pyrrolizidine alkaloids (PAs), mainly senkirkine and traces

of senecionine (Figure 2.1) [8]. The PAs can be hepatoxic, causing damage to the

liver and may even cause liver cancer when minute quantities are consumed

[9–11]. As a result, in order to minimize the amount of toxic PAs ingested, the

German health authorities have limited the daily intake of toxic PAs to 1µg. And

in Austria only drugs and preparations free of 1,2-unsaturated PAs have been

authorised since 1994 [8]. Therefore, in order to reduce the risk presented to

consumers, it is very important to develop a simple method for the isolation and

quantification of the toxic PAs in this herb.


43

N

O

O

O

O

H CH3

O

H

H3C

HO CH3

CH3

A

N

O

O

O

H CH3

O

H

H3C

HO CH3

H

B

Figure 2.1 Chemical structures of (A) senkirkine and (B) senecionine.

Usually, the bioactive or marker compounds of T. farfara are extracted by soxhlet

extraction [8], or by heating under reflux [8, 12–14]. However, these extraction

methods are time-consuming and involve the use of large volumes of solvent.

Recently, simpler and more environmental friendly extraction methods have been

developed. These include supercritical fluid extraction (SFE), microwave-assisted

extraction (MAE), pressurized fluid extraction (PFE) and so on. Their

applicability in extraction of solid matrices and natural products has been

discussed [15,16]. Among these new methods, MAE is applied on the extraction

of a variety of sample matrices due to its faster extraction time and better

extraction yields. This is because the heating by microwaves is instantaneous and

occurs in the heart of the sample, leading to rapid extractions. Moreover, as the

radiation can be focused directly onto the sample, microwave heating is more

efficient and thus homogeneity and reproducibility improve significantly [16].

Until recently, MAE has been widely applied on the extraction of bioactive

compounds from natural products, such as isolation of alkaloids from Rhizoma

coptidis [17], Nelumbo nucifera leaves [18], Lycoris radiata [19] and Solanum


44

nigrum L. [20].

Pressurized hot water extraction (PHWE) is another extraction technique which

can be applied for the isolation of bioactive or marker compounds from botanicals

and medicinal plants while reducing or completely eliminating the use of organic

solvents [21]. In PHWE, water is used as the extraction solvent at temperatures of

100–350 C and at a pressure high enough to keep it as a liquid. At elevated

temperatures and pressures, the dielectric constant of water will be similar to those

of ethanol or acetone. As a result, it can be used with other organic solvents to

extract medium- or low-polarity compounds [22]. Until recently, PHWE has

already been successfully applied to the extraction of less polar organic

compounds from the environmental samples and traditional herbal medicines [21,

23–26].

For the analysis and separation of PAs in T. farfara, analytical techniques such as

gas chromatography [8], high performance liquid chromatography [12, 27], and

capillary electrophoresis [12, 13] have been used. Sometimes, only the major

alkaloid senkirkine, but not the minor alkaloid senecionine in T. farfara could be

determined [8, 12, 14]. LC/MS is able to detect low levels of target compounds in

a complex matrix as MS provides a second dimension (mass to charge) which

separates co-eluting components. With tandem mass spectrometry, characteristic

fragmentation pattern required for the rapid identification of unknown compounds

can be obtained. To the best of our knowledge, the isolation, quantitative analysis

and effects of ion suppression on senkirkine and senecionine in medicinal plant

extracts using LC/MS have not been reported. Furthermore, reports on the rapid

detection of major and minor components in the presence of co-eluting peaks in

botanical extracts have been limited.


45

For the determination of PAs present in medicinal plants, proper sample

preparation is of utmost importance. Thus, the aim of the current work is to

develop a method for the rapid extraction and analysis of alkaloids such as

senkirkine and senecionine using MAE and PHWE for determination by LC and

LC/MS. The extraction efficiency of MAE and PHWE will be compared with that

of heating under reflux. Concurrently, the effect of ion suppression in the

botanical extracts will be investigated.

2.2 Experimental

2.2.1 Chemicals and reagents

Methanol (HPLC Grade) and dichloromethane (HPLC Grade) were the products

of Merck (Nordic European Center, Singapore). Absolute ethanol (HPLC Grade)

was purchased from Fisher scientific (Loughborough, Leicestershire LE11 5RG,

UK). Anhydrous ether (HPLC Grade) was obtained from Hayman (Witham, Essex,

UK). Acetonitrile (HPLC Grade) was the product of Tedia (Fairfield, OH 45014,

USA). Ultra-pure water was obtained by a NANOpure ultrapure water system

(Barnstead Int., Dubuque, IA, USA). Formic acid was obtained from Fluka

(Sigma–Aldrich pte Ltd., Science Park II, Singapore). Ammonium formate, sand

(white quartz, 50–70 mesh, suitable for chromatography) and chlorpheniramine

maleate were purchased fromSigma (St. Louis, MO, USA). Ammonium acetate

was the product of Merck (Darmstadt, Germany). Powdered and dried flowers of

T. farfara, aswell as crystallized standards of pyrrolizidine alkaloids were kindly

provided by Health Sciences Authority, Singapore.

2.2.2 Preparation of reference stantards

Stock solution of senkirkine at 3000 mg/L was prepared in methanol. Senecionine

was prepared at 2500 mg/L in methanol. For the LC analysis of MAE, the


46

working solutions of PAs were prepared in the calibration range of 5–200 mg/L in

methanol. For the LC analysis of PHWE, the working solutions of PAs were

prepared in the calibration range of 10–200 µg/L in methanol.

2.2.3 Extraction

2.2.3.1 Microwave-assisted Extraction

A closed vessel system technique was employed using Marsx from CEM

Corporation (Matthews, NC, USA). Powdered air-dried T. farfara flowers (1 g)

were placed into a 100 mL Teflon extraction vessel with 40 mL of extraction

solvent (Acidified to pH 2–3 with acid). Multiple extractions (2–3 samples

simultaneously) were performed under different conditions. Samples were

automatically stirred during the extraction process. The pressure in the vessel was

set to be that of its own vapor pressure, and the temperature was set to be at

boiling point of the binary mixture. After extraction, the vessel was allowed to

cool to room temperature, extracts filtered, reduced to half volume, and extracted

twice using dichloromethane, and twice using ether. The aqueous layer was then

basified using 1M NaOH (pH 8–9) before it was extracted thrice using

dichloromethane. The combined extracts were then evaporated to dryness using

rotary evaporator and residues were dissolved in 1mL of methanol for analysis

using HPLC and 5mL for analysis using LC–MS. The mixture was centrifuged at

13,000 rpm for 15 min before it was analysed.

2.2.3.2 Pressurized Hot Water Extraction

The instrumentation of PHWE was the same as the one used by Ong et al. [21].

Briefly, the dimensions of the stainless steel columns obtained from Phenomenex

were 10.0 mm ID×1/2 in OD×25 cm. The high pressure was generated by an

isocratic Shimadzu LC 10 series pump (Kyoto, Japan). The pump flow was set at


47

1.5 mL/min and the pressure in the system indicated by the HPLC pump was

between 15 and 23 bars. The elevated temperature for PHWE was maintained by a

HP5890 GC oven (Hewlett-Packard, CA, USA). And a CLIC miniature back

pressure regulator (Switzerland) was used to control the outlet flow. Generally

0.25 g of the powdered and dried T. farfara was weighed directly in to a 50 mL

plastic tube and mixed thoroughly with around 15 mL sand. The mixture was then

transferred to the extraction cell, whose ends were filled with cotton. The cell was

pre-filled with extraction solvent to check the possible leakage before setting the

temperature to a certain value.

The extraction efficiency was investigated with extractions at various

temperatures of 60 C, 80 C, 100 C and 120 C for 40 min. The extract collected

was rota-evaporated or heated to less than 50mL, followed by being transferred to

a 50 mL volumetric flask. The kinetic study of PHWE was performed by

collecting extracts at an interval of 10 min over a period of 80 min. In between

runs, the system was washed with methanol in water (1:1) for 5 min.

2.2.3.3 Heating under Reflux

Powdered air-dried T. farfara flowers (1 g) were placed in a 150mL round bottom

flask with 60mL of solvent. 1M HCl was added dropwise such that the pH of the

resulting solution fell to around 2–3. A condenser was fitted to prevent solvent

loss from occurring and the mixture was refluxed for 60min. The mixture was

cleaned up in the same way as described in MAE section.

2.2.4 LC and LC/ESI-MS analyses of MAE Extracts

HPLC analyses were performed on an Agilent 1100 HPLC instrument

(Waldbronn, Germany) coupled to binary gradient pump, autosampler, column

oven, and diode array detector. Detection wavelength was set at 220 nm.


48

For the analysis of MAE extracts by HPLC, the samples were separated on a

Waters Xterra C18 column (5 µm, 3.9 mm 150 mm, USA). The gradient elution

involved a mobile phase consisting of (A) 0.1% formic acid in 20 mM ammonium

acetate and (B) 0.1% formic acid in acetonitrile. The initial condition was set at

10% B, gradient up to 40% B in 20 min before returning to initial conditions for

10 min. The oven temperature was set at 40 C and flow rate was set at 0.7

mL/min. For all experiments, volumes of 10 µL of the diluted standards and

samples were injected.

For LC/ESI-MS analysis, the system comprised of an Agilent 1100 series

(Waldbronn, Germany) binary gradient pump, autosampler, column oven, diode

array detector and a LCQ-Duo ion trap mass spectrometer (ThermoFinnigan, San

Jose, CA, USA).

ESI-MS was performed in the positive mode. The LCQ mass spectrometer was

operated with the capillary temperature at 270 C, sheath gas at 80 (arbitrary units)

and auxillary gas at 20 (arbitrary units). The target was fixed at 2×107 ions and the

automatic gain control was turned on. The electrospray voltage was set at 4.5 kV,

the capillary voltage at 10 V and lens tube offset at 0 V. Mass spectra were

recorded from m/z 100–800.

The ESI-MS spectrum was acquired in the positive mode. The instrument was

operated in the selected ion monitoring (SIM) mode, where m/z of 336, m/z of 366

and m/z of 230 were isolated. The product ions from 100 to 800 were collected.

The heated capillary temperature was maintained at 350 C, the flow rate of drying

gas was set at 10 L/min and the pressure of nebulizer nitrogen gas was 50 psi,

respectively. The target was set at 30,000, maximum accumulation time: 300 ms,

the number of average scans was 5spectra/s and Smart-SelectTM was used.


49

Linearities for PAs were established between 5 mg/L and 200 mg/L (correlation

coefficient≥0.99). S/N was at least 3 for peak identification. To quantify the two

compounds in the medicinal herb, a three-point calibration based on the linearity

established was used. The R.S.D.s (n = 6) for senkirkine and senecionine were

found to be less than 5% on different days.

2.2.5 LC/ESI-MS analyses of PHWE Extracts

For LC/ESI-MS analyses of PHWE extracts, the gradient elution consisted of

mobile phase of (A) 30 mM ammonium formate with 0.1% formic acid and (B)

acetonitrile with 0.1% formic acid. The initial condition was set at 5% B, gradient

up to 100% B in 15 min before returning to initial condition for 10 min. Oven

temperature was set at 50 C and flow rate was set at 0.2 mL/min. For all

experiments, 5 µL of standards and sample extracts were injected. The column

used for separation was Luna 3u C18 (2) 100A, 100 mm × 2.0 mm (Phenomenex,

Torrance, CA, USA). All the conditions for ESI-MS were the same as mentioned

above. Linearitis of PAs were established between 10 and 200 µg/L for LC/MS

with correlation coefficients of r2 ≥0.99. And the three-point calibration based on

the linearity established was used to quantify the two compounds in the medicinal

herb. The R.S.D.s (n = 6) for senkirkine and senecionine was found to vary from

0.6% to 5.4% on different days.

2.3 Results and Discussion

2.3.1 HPLC separation of MAE extract

2.3.1.1 HPLC separation with UV detector

The optimized separation conditions were applied to the MAE extracts of T.

farfara. The content of senkirkine in samples of T. farfara was found to be present

from 19.5 ppm to 46.6 ppm. For senecionine, it was either not detected or present


50

in less than 1 ppm [8]. From the chromatograms obtained, co-eluting peaks were

observed with the analyte, senkirkine, before any clean-up step was performed

(Figure 2.2A). In order to determine the amounts of PAs present in the plant

extract, a clean-up step to remove co-eluting substances was performed to

quantitate senkirkine accurately (Fig. 2.2B). Based on other reports, a clean-up

step with solid phase extraction with diol or cation exchange cartridges was used

[8, 28–29]. For the current work we used a liquid–liquid extraction clean-up step

as reported earlier [12]. To check if the co-eluting peaks had been removed

efficiently, UV spectra (not shown) were obtained for the senkirkine peak in the

botanical extract and compared against that obtained using a standard. From the

UV spectra obtained, it can be seen that the clean-up step has effectively removed

the co-eluting components as the UV spectrum of the peak from the sample

closely resembles that of the UV spectrum of the peak from the standard.


51

0 2 4 6 8 10 12 14 16 18-600

-400

-200

0

200

400

600

800

UV

Abs

orba

nce

(mA

U)

Time (min)

Co-eluting peaks4.

478

10.9

2911

.200

11.8

12

A

0 2 4 6 8 10 12 14 16 18-600

-400

-200

0

200

400

600

800

UV

Abs

orba

nce

(mA

U)

Time (min)

Co-eluting peaks4.

478

10.9

2911

.200

11.8

12

A

0 2 4 6 8 10 12 14 16 18

0

10

20

30

40

UV

Abs

orba

nce

(mA

U)

Time (min)

Senkirkine peak

10.7

54B

0 2 4 6 8 10 12 14 16 18

0

10

20

30

40

UV

Abs

orba

nce

(mA

U)

Time (min)

Senkirkine peak

10.7

54B

Figure 2.2 Chromatogram of Tussilago farfara extract A) before clean-up step, and B) after clean-up step. Running conditions are as stated above. 2.3.1.2 Optimization of MAE

As the pyrrolizidine alkaloids exist mainly in basic form, the alkaloids are usually

extracted in acidified solution with weak acids such as acetic acid or other strong

acids [8, 12, 28, and 29]. For the current work, two different acids, acetic acid and

HCl in MeOH: H2O (1:1) were tested with MAE under the identical experimental

conditions. From the results obtained in Figure 2.3A, it can be seen that extraction

using HCl gives better extraction efficiency. Hence, all further extractions were


52

done using solution acidified with HCl.

The amount of microwave radiation absorbed is dependent on the dielectric

constants of the solvents used. For the extraction of marker compounds in T.

farfara, it was reported that a binary mixture of MeOH: H2O (1:1) gave better

extraction efficiency compared to using purely just water or methanol itself [8].

Two different solvent mixture MeOH: H2O (1:1) and EtOH: H2O (1:1) were

evaluated under the same experimental conditions. From the result in Figure 2.3B,

MeOH: H2O (1:1) was able to give better extraction efficiency. Hence, MeOH:

H2O (1:1) acidified with HCl was selected for all further works. Although the use

of a closed vessel allowed the use of temperatures above the boiling point of the

solvent, the boiling point of the binary mixture was chosen for the current work.

This is to prevent any buildup of excessive vapor pressure in the vessel. An

extraction time of 15min was selected as it was reported that extraction using

MAE could be completed in less than 15 min [16, 30–33].

Extraction using different solvent types

0

20

40

60

80

100

120

140

MeOH:H2O EtOH:H2O

Con

cent

ratio

n (p

pm)

B Extraction using different solvent types

0

20

40

60

80

100

120

140

MeOH:H2O EtOH:H2O

Con

cent

ratio

n (p

pm)

B

Figure 2.3 Effect of different acids and solvents on microwave-assisted extraction for senkirkine using: A) different acids with MeOH: H2O (1:1) as solvent, and B) MeOH: H2O (1:1) and EtOH: H2O (1:1) as solvents with HCl. Values expressed as means ± S.D (n=2). To determine the extraction efficiency of MAE, the amount of senkirkine present

in the same medicinal plant was compared with that obtained by heating under

reflux. From Table 2.1, the amount of senkirkine in the medicinal plant extracted

A Extraction usingaceticic acid and HCl

0

50

100

150

200

Acetic acid HCl

Concentration ( ppm )

A Extraction using acetic acid and HCl

0

50

100

150

200

Acetic acid HCl

Con

cent

ratio

n (p

pm)


53

by MAE was comparable to that obtained by heating under reflux. This shows that

the conditions such as the extraction temperature and time of extraction for MAE

were optimal. The recovery of the clean-up step using liquid–liquid extraction was

found to be 103.1±3.7% (n = 3).

Comparison between

heating under reflux

and MAE on 2 days

Senkirkine (µg/g)

Heating under reflux

(n=2)

MAE (n=3)

Mean ± S.D. RSD (%) Mean ± S.D. RSD (%)

Senkirkine 94.0 ± 4.2 4.50 111.7 ± 11.3 9.66

Senkirkine 118.5 ± 0.04 0.04 123.6 ± 3.57 3.05

Table 2.1 Comparison of MAE with heating under reflux for the analysis of senkirkine in Tussilago farfara by LC with UV detection. Taking into account the number of steps in the proposed procedures, sample size

used and the non-homogeneity of the medicinal plant samples, the method

precision for the current work was comparable with other reports for the

determination of marker or bioactive compounds in medicinal plants by LC with

UV detection [34–37].

The main reasons for the enhanced performance when using MAE over reflux

extraction are the higher solubility of analytes in solvent and higher diffusion rates

as a result of a more homogenous heating due to the principle of microwave

heating. A more homogenous heating resulted in the strong solute-matrix

interaction caused by van der Waals forces, hydrogen bonding and dipole

attractions between solute molecules and active sites on the matrix to be disrupted.

2.3.1.3 LC/ESI-MS analysis for MAE

For the positive ion ESI experiments, senkirkine and senecionine showed a very


54

high tendency to form [M+H]+ at 366.3 and 336.2 respectively. The two

compounds were injected into the ESI/MS system and then fragmented in the ion

trap up to MS2 or MS3 using the collision induced dissociation (CID). The

advantage of multistage MS is that further fragmentation can be induced when

significant fragmentation pattern cannot be observed from the ESI/MS2 spectra.

This can be seen for baicalin in Scutellariae radix and Salvinorin B in Salvia

divinorum where significant fragmentation pattern was not observed with the MS2

spectra and MS3 was required to generate characteristic fragmentation pattern [38,

39]. As seen from each fragmentation pathways in Figure 2.4, the main losses

were water and carbon monoxide. The losses of H2O and CO have been observed

in the MSn spectra of natural occurring substances such as salvinorin A and others

[38, 40]. The MS/MS experiments of senkirkine and senecionine (Figure 2.4A–C)

provided characteristic fragmentation ions that were consistent with that of

retrosine-type and otonecinetype alkaloids [28, 41]. The otonecine-type PA,

senkirkine, showed characteristic fragment ions at m/z 150, and 168. The

retronecine-type PA, senecionine, showed characteristic fragment ions at m/z 138

and 120. The fragmentation pathways for both senkirkine and senecionine were

consistent with other reports [28, 41].


55

106.9

121.9

149.8

167.9

198.8 249.8 267.9 338.0347.9

+MS2(366.3), 0.8min #70

0

1

2

3

7x10Intens.

100 150 200 250 300 350 m/z

N

CH2+

CH3

m/z = 121.9

N

O

CH3

CH2+

m/z = 149.8

N

O CH2+OH

CH3

m/z = 167.9

[MH-CO]+ [MH-H2O]+106.9

121.9

149.8

167.9

198.8 249.8 267.9 338.0347.9

+MS2(366.3), 0.8min #70

0

1

2

3

7x10Intens.

100 150 200 250 300 350 m/z

N

CH2+

CH3

m/z = 121.9

N

O

CH3

CH2+

m/z = 149.8

N

O CH2+OH

CH3

m/z = 167.9

[MH-CO]+ [MH-H2O]+

A

108.1

119.9

137.8

152.8 186.7219.9 273.9

290.0

308.0

318.2

+MS2(336.2), 0.7min #61

0.0

0.5

1.0

1.5

7x10Intens.

100 125 150 175 200 225 250 275 300 325 m/z

N

CH2+

m/z = 119.9

N

O CH2+H

m/z = 137.8

[MH-CO]+

[MH-H2O]+[MH-H2O-CO]+

108.1

119.9

137.8

152.8 186.7219.9 273.9

290.0

308.0

318.2

+MS2(336.2), 0.7min #61

0.0

0.5

1.0

1.5

7x10Intens.

100 125 150 175 200 225 250 275 300 325 m/z

N

CH2+

m/z = 119.9

N

O CH2+H

m/z = 137.8

[MH-CO]+

[MH-H2O]+[MH-H2O-CO]+

B

108.0

119.9

137.9

152.9

290.1

+MS3(336.0->308.0), 0.7min #23

0.0

0.5

1.0

1.5

2.0

2.5

6x10Intens.

100 125 150 175 200 225 250 275 300 325 m/z

N

CH2+

m/z = 119.9

N

O CH2+H

m/z = 137.8

[MH-H2O]+108.0

119.9

137.9

152.9

290.1

+MS3(336.0->308.0), 0.7min #23

0.0

0.5

1.0

1.5

2.0

2.5

6x10Intens.

100 125 150 175 200 225 250 275 300 325 m/z

N

CH2+

m/z = 119.9

N

O CH2+H

m/z = 137.8

[MH-H2O]+

C

Figure 2.4 MS2 Fragmentation pattern of A) Senkirkine, B) Senecionine, and C) MS3 fragmentation pattern of Senecionine. For senkirkine, only MS2 spectra was collected as the abundance of daughter ions

at m/z 347.9 [MH−H2O]+ and 338.0 [MH−CO]+ were low. The selection of lower

m/z for MS3 did not give useful information. Thus, further fragmentation was not

done for senkirkine. As for senecionine, the abundance of daughter ion at m/z

308.0 [MH−CO]+ was high. From Figure 2.4B and C, it can be seen that the

daughter ions at m/z 119.9 decreased while other ions at m/z 137.9 and m/z 152.9

increased. Although both senkirkine and senecionine have similar chemical


56

structures, the fragmentation patterns can differ significantly. Finally, the

characteristic fragmentation patterns for both senkirkine and senecionine are

required for the rapid identification of unknown compounds in botanical extracts

with reference to pure standards.

For the determination of aristolochic acids in multi-components herbal remedies, a

method without any clean-up or concentration steps with LC/MS2 had been

developed [42]. To the authors’ best knowledge, reports on the use of LC/MS for

identification of target compounds in the presence of co-eluting peaks in botanical

extracts are rather limited. As seen in Figure 2.2A and B, when HPLC was used,

long and tedious clean-up step had to be employed to remove co-eluting peaks

before senkirkine could be detected. However, by using LC/MS in the SRM mode

where ions of m/z of 336 and 366 were isolated and detected, trace component

senecionine as well as senkirkine could be detected successfully without the use of

any clean-up step (Figure 2.5). Despite the presence of co-eluting peaks and

possible ion suppression, the presence of low level of senecionine in the extract

can be detected confidently. This shows that the proposed method using LC/MS

was sensitive and selective. From Table 2.2, we can see that the comparable

extraction efficiency was obtained between MAE and heating under reflux

extraction. The limits of detection (LODs) and limits of quantitation (LOQs) of

the MAE method developed were found to be 0.26µg/g and 1.32µg/g, 0.47µg/g

and 1.80µg/g for senkirkine and senecionine respectively.


57

RT: 0.00 - 29.99

0 5 10 15 20 25

Time (min)

0

20

40

60

80

1000

20

40

60

80

1000

20

40

60

80

100 23.062.51

2.063.09 10.02

8.70

9.13

16.1616.02

22.94

8.862.47 22.22

10.06

22.83

9.38

:A: MAE extract without clean-up

B:Senecionine

C:Senkirkine

Rel

ativ

e A

bund

ance

RT: 0.00 - 29.99

0 5 10 15 20 25

Time (min)

0

20

40

60

80

1000

20

40

60

80

1000

20

40

60

80

100 23.062.51

2.063.09 10.02

8.70

9.13

16.1616.02

22.94

8.862.47 22.22

10.06

22.83

9.38

RT: 0.00 - 29.99

0 5 10 15 20 25

Time (min)

0

20

40

60

80

1000

20

40

60

80

1000

20

40

60

80

100 23.062.51

2.063.09 10.02

8.70

9.13

16.1616.02

22.94

8.862.47 22.22

10.06

22.83

9.38

:A: MAE extract without clean-up

B:Senecionine

C:Senkirkine

Rel

ativ

e A

bund

ance

Figure 2.5 A) Total ion chromatogram (TIC) of Tussilago farfara extract before clean-up step. B) Extracted ion chromatogram (EIC) of senecionine observed at m/z 336 and C) EIC of senkirkine observed at m/z 366. 2.3.2 HPLC analysis for PHWE

2.3.2.1 LC/ESI-MS analysis for PHWE

A LC/ESI-MS method without the use of splitting was used for the analysis of

senkirkine and senecionine in the herbal medicine. Linearities for senkirkine and

senecionine were established between 10µg/L and 200µg/L (correlation

coefficient≥0.99). The relative standard deviation (R.S.D.) values of the peak

heights for senkirkine and senecionine were observed to vary from 0.6% to 5.4%

and 3.0% to 4.2% (n = 6) respectively on different days.

2.3.2.2 Optimization of PHWE

The main parameters that can affect the extraction efficiency of PHWE are


58

pressure, temperature, extraction time and addition of organic modifiers [21,

43-44]. Generally, the pressure has little effect on the extraction efficiency since it

is only for maintaining the extraction solvent in liquid state at elevated

temperature [45–47]. For the current work, the optimal extraction solvent in MAE

was also used in PHWE. As a result, temperature and extraction time were the

main parameters considered. Since the physicochemical properties of water such

as viscosity, surface tension, diffusibility, and solvent strength could be changed

at elevated temperature, they could affect the solubility of the target compound in

water. Since significant changes in the extraction efficiency from 60 C to 120 C

were not observed in Figure 2.6, a temperature of 60 C was chosen as the

optimized temperature. Figure 2.7 shows that most of the main PA, senkirkine,

was extracted out after 50min, while most of senecionine was extracted out after

20 min. As a result, 50 min was selected as the optimal extraction time.

A

0

0.5

1

1.5

2

2.5

3

60 80 100 120

Temperature (oC)

Con

cent

ratio

n of

sene

cion

ine

extra

cted

(

µg/g

)


59

Figure 2.6 Effect of different extraction temperatures on the recoveries of (A) senecionine and (B) senkirkine by PHWE (n=3) with flow rate: 1.5 ml/min for 40 min.

Figure 2.7 Effect of extraction time on the recoveries of senecionine and senkirkine by PHWE.

B

10

15

20

25

30

35

40

45

50

55

60 80 100 120

Temperature (oC)

Con

cent

ratio

n of

senk

irkin

e ex

tract

ed

(µg/

g)

-10

123456789

10111213

0-10 10-20 20-30 30-40 40-50 50-60 60-70 70-80

Time (min)

Peak

Hei

ght (

×106 )

Senecionine

Senkirkine


60

Table 2.2 Comparison of MAE between heating under reflux and PHWE between heating under reflux.

Conc. of alkaloids on 2 days (µg/g)

Comparison between MAE and heating under reflux Comparison between PHWE and heating under reflux MAE (n=3) Heating under reflux (n=2) PHWE (n=6) Heating under reflux (n=2)

Mean ± S.D. RSD (%) Mean ± S.D. RSD (%) Mean ± S.D. RSD (%) Mean ± S.D. RSD (%) Senkirkine (Day 1) 102.2 ± 10.6 10.4 109.6 ± 7.1 6.5 85.28 ± 5.00 5.9 84.78 ± 3.14 3.7 Senecionine (Day 1) 2.16 ± 0.08 3.7 2.65 ± 0.04 1.5 3.20 ± 0.19 5.9 3.05 ± 0.04 1.3 Senkirkine (Day 2) 112.8 ± 11.5 10.2 118.6 ± 8.3 7.0 80.95 ± 2.45 4.0 84.70 ± 3.73 5.8 Senecionine (Day 2) 2.57 ±0.12 4.7 2.76 ± 0.15 5.4 3.18 ± 0.16 5.0 3.30 ± 0.18 5.5


61

The extraction efficiencies of pyrrolizidine alkaloids in T. farfara by PHWE were

found to be comparable with heating under reflux extraction (Table 2.2). The

method precision (R.S.D.) was found to vary from 4.0% to 5.9%. The LODs and

LOQs of the PHWE method established were found to be 1.04 µg/g and 5.29 µg/g,

0.72 µg/g and 2.86 µg/g for senkirkine and senecionine respectively. From Table

2.2, we can also see that there are some differences between the two sets of results

for heating under reflux extractions. This might be due to the batch to batch

variation.

2.3.3 Matrix-induced interference

Ion suppression is one form of matrix effect that is a major concern in LC/MS

techniques. It is typically observed in sample extracts from biological samples and

is most possibly caused by the high concentration of nonvolatile materials present

in the spray with the analyte. Any co-eluting nonvolatile solute such as salts may

cause the suppression of ionization, giving rise to irreproducible results. Ion

pairing agents, surface activities of the analyte and interfering compounds may

also play an important role in ionization suppression [48]. Modifying instrumental

components and parameters, chromatographic separation and sample preparation

are strategies to reduce or eliminate ion suppression. Various calibration methods

can be applied to compensate this effect. Standard addition is an effective method

and is able to give good results even with variable matrices [49]. For the analysis

of naturally occurring substances in botanical extracts, the effect of ion

suppression was often not investigated [50–54]. As for the analysis of

Toosendanin in the botanical extracts carried out using external standard

calibration, the effect of matrix induced interference was investigated and

significant ion suppression was not observed [55]. For plant analysis, standard


62

addition experiments remained to be the most effective method for the evaluation

of ion suppression in LC/MS. From the chromatograms obtained in Fig. 2B, the

presence of senkirkine was confirmed by the retention time and the UV spectra

with the standard compounds. The effect of ion suppression or matrix induced

interference in the botanical extracts was investigated using the standard addition

method. For extracts obtained from MAE, the internal standard calibration plot (y

= 0.0024x−0.035, y = 0.0031x + 0.075) and the standard addition (with the use of

internal standard) plot (y = 0.0025x+0.1831, y = 0.0031x + 0.4857) were

essentially parallel for both senecionine and senkirkine respectively. For extracts

obtained from PHWE, the external standard calibration plot (y = 4321.1x−7089.6,

y = 2774.9x−4978.7) and the standard addition (without the use of internal

standard) plot (y=4112x + 97678, y = 2986.7x + 122142) were also essentially

parallel for both senecionine and senkirkine. Hence, it can be deduced that without

the use of any sample clean-up steps, significant ion suppression was not present

even for the analysis of low level of senecionine in the plant extract.

2.4 Conclusion

In this work, both the combination of MAE and PHWE with LC/MS allowed for

the simple, rapid isolation and quantification of the major and the minor

pyrrolizidine alkaloids in T. farfara. Using LC/MSn, characteristic fragmentation

pattern for both senkirkine and senecionine were obtained which could be used for

the rapid identification of unknown compounds in botanical extracts with

reference to a pure standard. Accompanied with the use of internal and external

standard calibration, significant matrix-induced interferences or ion suppression in

the presence of co-eluting peaks were not observed. As a result, the MAE and

PHWE methods proposed in this work could be alternatives for the extraction of


63

bioactive or marker compounds in medical plants.

References

[1] K. Y. Liu, T. J. Zhang,W. Y. Gao, H. X. Chen, Y. N. Zheng, Zhongguo

Zhongyao Zazhi 31 (2006) 1837.

[2] C. L. Fan, Faming Zhuanli Shenqing Gongkai Shuomingshu (2007) CN

101036491.

[3] Y. M. Shi, Faming Zhuanli Shenqing Gongkai Shuomingshu (2002) CN

1381252.

[4] S. X.Wang, Faming Zhuanli Shenqing Gongkai Shuomingshu (2002) CN

1931022.

[5] H. S. Lee, M.C. Rho, Y. K. Kim, M. Y. Chung, S. W. Lee, H. R. Park, J. H.

Choi, J. Y. Kang, D. H. Kim, Repub. Korean Kongkae Taeho Kongbo (2007) KR

2007008228.

[6] Awasthy, Narendra Kumar, Indian Patent Application (2006) IN

2004DE01564.

[7] H. R. Park, M. Y. Yoo, J. H. Seo, I. S. Kim, N. Y. Kim, J. Y. Kang, L. Cui, C.

S. Lee, C. H. Lee, H. S. Lee, J. Agric. Food Chem. 56 (2008) 10493.

[8] R. Lebada, A. Schreier, S. Scherz, C. Resch, L. Krenn, B. Kopp, Phytochem.

Anal. 11 (2000) 366.

[9] Environmental Health Criteria 80, Pyrrolizidine Alkaloids,World Health

Organization, Geneva, 1988.

[10] P. P. Fu, Y. C. Yang, Q. S. Xia, M. W. Chou, Y. Y. Cui, G. Lin, J. Food Drug

Anal. 10 (2002) 198.

[11] P. P. Fu, Q. Xia, G. Lin, M. W. Chou, Drug Metab. Rev. 36 (2004) 1.


64

[12] L. J. Yu, S.F.Y. Li, Electrophoresis 26 (2005) 4360.

[13] H. T. Feng, S. F. Y. Li, J. Chromatogr. A 973 (2002) 243.

[14] T. Mroczek, K. Glowniak, A.Wlaszczyk, J. Chromatogr. A 949 (2002) 249.

[15] V. Camel, Analyst 126 (2001) 1182.

[16] B. Kaufmann, P. Christen, Phytochem. Anal. 13 (2002) 105.

[17] C. Sun, H. Z. Liu, Anal. Chim. Acta 612 (2008) 160.

[18] G. Q. Xiao, Y. Tian, X. Y. Lu, J. Fang, Anjisuan He Shengwu Ziyuan 29

(2007) 76.

[19] H. J. Fan,W. Luan, G. K. Li, Fenxi Ceshi Xuebao 25 (2006) 27.

[20] Q. Liu, X. Q. Chen, X. Y. Jiang, S. D. Wu, Tianran Chanwu Yanjiu Yu Kaifa

17 (2005) 65.

[21] E. S. Ong, J. S. H. Cheong, D. Goh, J. Chromatogr. A 1112 (2006) 92.

[22] R. Carabias-Martinez, E. Rodriguez-Gonzalo, P. Revilla-Ruiz, J. Hernandez-

Mendez, J. Chromatogr. A 1089 (2005) 1.

[23] V. Fernandez-Gonzalez, E. Concha-Grana, S. Muniategui-Lorenzo, P. Lopez-

Mahia, D. Prada-Rodriguez, J. Chromatogr. A 1196–1197 (2008) 65.

[24] C. C. Teo, S. N. Tan, J. W. H. Yong, C. S. Hew, E. S. Ong, J. Chromatogr. A

1182 (2008) 34.

[25] K. Hartonen, J. Parshintsev, K. Sandberg, E. Bergelin, L. Nisula, M.

Riekkola, Talanta 74 (2007) 32.

[26] L. Dong, J.Y. Wang, C. H. Deng, X. Z. Shen, J. Sep. Sci. 30 (2007) 86.

[27] G. Hosch, H. Wiedenfeld, T. H. Dingermann, E. Roder, Phytochem. Anal. 7

(1996) 284.

[28] T. Mroczek, K. Ndjoko, K. Glowniak, K. Hostettmann, J. Chromatogr. A

1056 (2004) 91.


65

[29] T. Mroczek, K. Ndjoko-Ioset, K. Glowniak, A. Mietkiewicz-Capala, K.

Hostettmann, Anal. Chim. Acta 566 (2006) 157.

[30] A. Brachet, P. Christen, J. L. Veuthey, Phytochem. Anal. 13 (2002) 162.

[31] X. Pan, G. Niu, H. Liu, J. Chromatogr. A 922 (2001) 371.

[32] V. Camel, Trends Anal. Chem. 19 (2000) 229.

[33] F. Zhang, B. Chen, S. Xiao, S. Yao, Sep. Purif. Technol. 42 (2005) 283.

[34] E. S. Ong, S. O. Woo, Y. L. Yong, J. Chromatogr. A 904 (2000) 57.



[37] A. T. W. Eng, M.Y. Heng, E. S. Ong, Anal. Chim. Acta 583 (2007) 289.

[38] C. Medana, C. Massolino, M. Pazzi, C. Baiocchi, Rapid Commun. Mass

Spectrom. 20 (2006) 131.

[39] E. S. Ong, J. Chromatogr. B 812 (2004) 23.

[40] Y. Zhang, Q. Shi, P. Shi, W. Zhang, Y. Cheng, Rapid Commun. Mass

Spectrom. 20 (2006) 2328.

[41] S. Pu, D. Xu, M. Zhang, H. Zhou, Z. Wang, G. Yu, Chin. J. Nat. Med. 2

(2004) 293.

[42] G. C. Kite, M. A. Yule, C. Leon, M. S. J. Simmonds, Rapid Commun. Mass

Spectrom. 16 (2002) 585.

[43] J. Kronholm, K. Hartonen, M. L. Riekkola, Trends Anal. Chem. 26 (2007)

396.

[44] R. M. Smith, J. Chromatogr. A 975 (2002) 31.

[45] S. B. Hawthorne, Y. Yang, D. J. Miller, Anal. Chem. 66 (1994) 2912.

[46] M. S. Krieger, J. L. Wynn, R. N. Yoder, J. Chromatogr. A 897 (2000) 405.

[47] E. S. Ong, J. Sep. Sci. 25 (2002) 825.


66

[48] R. King, R. Bonfiglio, C. Fernandez-Metzler, C. Miller-stein, T. Olah, J. Am.

Soc. Mass Spectrom. 11 (2000) 942.

[49] L. E. Jessome, D.A. Volmer, LC GC 24 (2006) 498.

[50] Y. Bao, C. Li, H. Shen, F. Nan, Anal. Chem. 76 (2004) 4208.

[51] R. Kaneko, S. Hattori, S. Furuta, M. Hamajima, Y. Hirata, K. Watanabe, H.

Seno, A. Ishii, J. Mass Spectrom. 41 (2006) 810.

[52] D. Pigini, A.M. Cialdella, P. Faranda, G. Tranfo, Rapid Commun. Mass

Spectrom. 20 (2006) 1013.

[53] T. Ligor, A. Ludwiczuk, T.Wolski, B. Buszewski, Anal. Bioanal. Chem. 383

(2005) 1098.

[54] P. Zou, H. L. Koh, Rapid Commun. Mass Spectrom. 21 (2007) 1239.

[55] E. S. Ong, C. N. Ong, Rapid Commun. Mass Spectrom. 21 (2007) 589.

Chapter 3 Preconcentration and separation of toxic alkaloids in herbal medicines by non-aqueous capillary electrophoresis (NACE)

67

Chapter 3 Preconcentration and Separation of Toxic Pyrrolizidine Alkaloids

in Herbal Medicines by Non-aqueous Capillary Electrophoresis (NACE)

3.1 Introduction

Pyrrolizidine alkaloids (PAs) are a class of natural products including about 350

structures isolated from plants (especially Senecio, Eupatorium-Asteraceae,

Crotalaria-Fabaceae, and Heliotropium-Boraginaceae) [1 – 3]. About 3% of all

angiosperms contain these alkaloids. Until recently, more than 660 PAs and PA

N-oxides have been identified in over 6,000 plants of these three families, and

about half of them exhibit toxic activities [4]. Plants known or suspected to

contain PAs are widely used for medicinal purpose as remedies all over the world

[5]. For example, leaves of Tussilago farfara L. (Asteraceae), coltsfoot, are

traditionally used to treat bronchial infections. However, this herb contains toxic

PAs which can cause liver damage in minute quantities. As far as Senecio is

concerned, it contains toxic PAs which have a cumulative poisonous effect when

the whole herb is consumed [5]. So the risk to human health posed by exposure to

these compounds has been a concern due to the wide distribution of toxic

pyrrolizidine alkaloid-containing plants in the world. To avoid any risk to

consumers, the German public health authorities limited the daily intake of toxic

PAs to 1 µg. For use as herbal tea, the amount of these compounds in coltsfoot is

limited to 10 µg /day [6]. As a result, quantitative analysis of Pyrrolizidine

Alkaloids in these herbs plays an important role in avoiding the risks to

consumers.

Traditional Chinese medicine (TCM) has been used in China for treating illness

for more than 2000 years. Many people in other Asian countries have also taken

TCM for clinical practice or improving health for a long time. During the last two


68

decades, the use of Chinese herbal medicines in Europe, Australia, and North

America for health care has been increasing. There are several different usage of

the Chinese herbal plants; for new drug development, as natural remedies,

functional foods (mixing functional herbs in conventional food), and dietary

supplements. So the qualitative and quantitative analysis of active components in

TCM have become very important. Tussilago farfara L. (Kuan Donghua) is a

plant species which has been reported to be officially used as the plant source for

TCM herbs among the 49 PA-containing Chinese herbal plants produced in China.

It has been used to treat chronic bronchitis, asthma, and influenza [4]. Besides

acidic polysaccharides, which are considered to be responsible for the cough

relieving effect, the plant contains toxic pyrrolizidine alkaloids (PAs), mainly

senkirkine and traces of senecionine [6]. Therefore, in order to utilize this plant

more rationally and safely, it is necessary for us to develop a simple, economical

and efficient method for the simultaneous separation and determination of these

bioactive compounds in the plant.

Until recently, several methods such as GC [7-8], HPLC [9] and photometric

determination [10] have been employed for the analysis and quantification of PAs

in coltsfoot. However, GC lacks resolving power and quantitative precision, while

HPLC presents lower efficiency and is time-consuming. Therefore, it is necessary

to establish a rapid and effective method for quantitative analysis of these

alkaloids.

Capillary electrophoresis (CE) has been widely used for the analysis of the

components in herbal medicines because of its high resolution, minimum sample

volume, short analysis time, and high separation efficiencies. Various alkaloids

contained in herbal plants have been separated and determined using different CE


69

modes, such as capillary zone electrophoresis (CZE), nonaqueous CE (NACE),

and micellar electrokinetic chromatography (MEKC) [11–26]. However, due to

the narrow optical path length and the small sample volumes that can be injected

into the capillary, the sensitivity in CE is poor especially when using UV detection.

In order to overcome this problem, the application of on-line preconcentration

techniques is a much simpler and more economical method compared with

improving the sensitivity of the detector, such as laser-induced fluorescence

detection and mass detection. Recently, non-aqueous capillary electrophoresis

(NACE) has been found to be a good alternative for the analysis of Chinese herbs,

which are difficult to separate in aqueous media. Indeed, compared with aqueous

systems, the different chemical and physical properties of organic solvents

(viscosity, dielectric constant, polarity, auto-protolysis constant, electric constant

conductivity, etc.) have shown several advantages in terms of selectivity,

efficiency, rapidity, mass spectrum compatibility and analyte solubility and

stability. And some on-line preconcentration techniques in NACE have been

reported [27-35]. Among these stacking methods, large volume sample stacking

(LVSS) and field-amplified sample stacking (FASS) are the simplest and most

commonly used techniques. In LVSS, the sample is dissolved in a low

conductivity buffer and the sensitivity is improved by enhancing the injection

volume of the sample. In this condition, a typical peak broadening will happen

with the increasing of the sample volume. Normally, the improvement of the

stacking efficiency is not so high. While in FASS, a short zone of low

conductivity is presented at the capillary inlet, across which an electric field up to

several hundred times higher than that employed in normal CE is established. This

will permit the charged analytes to be injected at high velocity. With the


70

electrokinetic injection, analytes are stacked at the boundary between the

low-conductivity zone and the running buffer.

To the best of our knowledge, the applications of NACE on the analysis of PAs in

Tussilago farfara L. have not been reported. In this study, firstly, a highly

selective NACE system will be developed for the separation and determination of

the alkaloids in the Chinese herb, Tussilago farfara L. The effects of the

concentrations of electrolyte, acetic acid, organic solvent and applied voltage will

be investigated. Then the two on-line preconcentration techniques, LVSS and

FASS will be established for the analysis of PAs in Tussilago farfara L.

3.2 Experimental Section

3.2.1 Materials

Ammonium acetate and acetic acid (99.8%) were the products of Merck

(Darmstadt, Germany). Methanol and ethanol (HPLC grade) were obtained from

Fisher Scientific (Loughborough, Leicestershire LE11 5RG UK). Acetonitrile

(ACN) ( HPLC grade) was purchased from Tedia (Fairfield, OH45014, USA).

Deionized water (≥18 MΩ) used throughout the experiments was generated by a

NANOpure ultrapure water system (Barnstead Int., Dubuque, IA, USA). PAs of

senecionine, monocrotaline, senkirkine, and retrorsine (Chromadex, CA, USA)

were provided by the Health Sciences Authority (HSA, Singapore). Chemical

structures of these four PAs are shown in Figure 3.1. Tussilago farfara L (Kuan

Donghua) (Eu Yan Sang, Singapore) were also provided by HSA.


71

Figure 3.1 Structures of PAs of senkirkine, monocrotaline, sesecionine and retrorsine. 3.2.2 Instruments and methods

A 65 cm with 50 cm effective length long fused-silica capillary of 50 µm ID and

363 µm OD (Polymicro Technologies, Phoenix, AZ, USA) was used to carry out

the experiments. A portable CE instrument, CE-P2 Auto sampler (CE Resources,

Singapore) equipped with a UV-VIS detector (SPD-10AVP, Shimadzu, Japan)

was utilized for the NACE-UV experiments. The pressure for hydrodynamic

injection was set as 0.3psi. Direct UV detection was employed at a wavelength of

200 nm. Data was collected by CSW software (DataApex, Czech Republic).

When a new capillary was first used, it was rinsed with 1 M NaOH for 30 min and

deionized water for 10 min. After each run the capillary was rinsed for 3 min with

1 M NaOH and 3 min with deionized water, followed by 2 min with running

buffer, then conditioned under the applied voltage with running buffer for 2 min in

order to obtain highly reproducible electroosmotic flow (EOF). The buffer was


72

renewed after every two runs for good reproducibility. All operations were carried

out at a temperature of 22oC.

3.2.3 Standard sample and running buffer preparation

The stock standard sample solutions were prepared by dissolving 5 mg of each PA

into 1.5 mL methanol to obtain the final concentration of 3333.3 µg /mL.

Methanol in water (50:50) was used to dilute the stock sample solutions to a

desired concentration level. Stock solutions of ammonium acetate was prepared

and kept in the refrigerator (4oC). Running buffer was prepared daily.

3.2.4 Extraction of PAs in herbal medicines

3.2.4.1 Heating under reflux

Each of the pulverized herbal medicines (1 g) was refluxed for 30 min with 60 mL

of 50% methanol acidified with citric acid to pH 2–3. After filtration (0.2 µm)

the solution was reduced to half the volume and partitioned twice with 30 mL

dichloromethane and twice with 30 mL ether in order to remove lipophilic

accompanying substances. The aqueous solution was alkalized with 25%

ammonium (pH 9–10) and the PAs, as free bases, were extracted by threefold

partition each against 30 mL dichloromethane. The combined organic layers were

evaporated to almost dryness and then the residue was redissolved in 1 mL of

methanol [25].

3.2.4.2 Microwave-assisted extraction

A closed vessel system technique was employed using Marsx from CEM

Corporation (Matthews, NC, USA). Powdered air-dried Tussilago farfara flowers

(1g) were placed into a 100mL Teflon extraction vessel with 40mL of

CH3OH:H2O=50:50 (Acidified to PH 2-3 with acid). Samples were automatically

stirred during the extraction process. The pressure in the vessel was set to be that


73

of its own vapor pressure, and the temperature was set to be at boiling point of the

binary mixture. After 15 min extraction, the vessel was allowed to cool to room

temperature. The clean-up steps were the same as mentioned in section 3.2.4.1.

3.3 Results and discussion

3.3.1 Optimization of NACE conditions

3.3.1.1 Effect of concentration of acetic acid

The acidity of the buffer has an effect on the separation because it determined the

extent of ionization of each analytes. In this experiment the acidity of the buffer

was modified by acetic acid. Hereby, the concentration of acetic acid ranging from

0-4% (v/v) was examined to investigate the effect of concentration of acetic acid

on the migration behavior. From Figure 3.2A and Figure 3.2B we can see that

with the increase of the concentration of acetic acid, the migration time of PAs

decreased and the apparent mobilities of PAs increased. This might be due to the

excess H+ protonized the PAs, which move in the direction of the electroosmotic

flow (EOF). Considering both sensitivity and resolution, 1% (v/v) of acetic acid

was chosen as the optimal concentration.


74

Figure 3.2A Effect of acetic acid concentration on the separation of PAs. Buffer: 50 mM Ammonium Acetate and 20% (v/v) ACN in methanol medium, (1) 0% acetic acid; (2) 1% acetic acid; (3) 2% acetic acid; (4) 3% acetic acid; (5) 4% acetic acid. Capillary: 65 cm (50 cm to detector) x 50 µm i.d.; Applied voltage: 20 kv; Detection: 200nm. A: Senkirkine; B: Monocratoline; C: Senecionine; D: Retrorsine.

Figure 3.2B Effect of acetic acid concentration on apparent mobilities of PAs. The experimental conditions are the same as Figure 3.2A.

7.000

8.000

9.000

10.000

11.000

12.000

13.000

14.000

15.000

16.000

0 1 2 3 4Acetic acid concentration (%, v/v)

SenekirkineMonocrotalineSenecionineRetrorsine

App

aren

t mob

ilitie

s（cm

2.m

in-1

. kv-1）


75

3.3.1.2 Effect of concentration of ammonium acetate

Buffer concentration has a significant effect on separation performance because it

could influence the viscosity of electrolyte. In order to obtain the best resolution

of PAs, the ammonium acetate concentration ranging from 25 mM to 100 mM

was examined. Figure 3.3A and Figure 3.3B show that the migration time of PAs

increased with ammonium acetate concentration increasing. Also the noise

became higher with the increasing ammonium acetate concentration. This might

be due to the higher Joule heating caused by the increasing conductivity of the

buffer. It was found that the best sensitivity and resolution was obtained when the

concentration of ammonium acetate was 50 mM. So the optimal concentration of

ammonium acetate was chosen as 50 mM.

Figure 3.3A Effect of ammonium acetate concentration on the separation of PAs. Buffer: 1% (v/v) acetic acid and 20% (v/v) ACN in methanol medium, (1) 25 mM ammonium acetate, (2) 50 mM ammonium acetate, (3) 75 mM ammonium acetate, (4) 100 mM ammonium acetate.


76

Figure 3.3B Effect of ammonium acetate concentration on apparent mobilities of PAs. The experimental conditions are the same as Figure 3.3A. 3.3.1.3 Effect of concentration of ACN

Organic solvents used as buffer additives can influence the viscosity and dielectric

constant of electrolyte, so they could cause the change of velocity of EOF and

migration time of solute. The concentration of acetonitrile (ACN) ranging from

10% (v/v) to 35% (v/v) was tested to see its effect on the separation. From Figure

3.4A and 3.4B we can see that the apparent mobilities of PAs increased with

increasing concentration of ACN. As the best resolution was obtained at the

concentration of 20% (v/v) ACN, 20% (v/v) ACN was chosen as the optimal

condition.

10.000

12.000

14.000

16.000

18.000

20.000

22.000

25 50 75 100Ammonium acetate concentration (mM)

App

aren

t mob

ilitie

s（cm

2.m

in-1

. kv-1）

SenekirkineMonocratolineSenecionineRetrorsine


77

Figure 3.4A Effect of acetonitrile (ACN) concentration on the separation of PAs. Buffer: 50 mM ammonium acetate and 1% (v/v) acetic acid in methanol medium, (1) 10% (v/v) ACN, (2) 15% (v/v) ACN, (3) 20% (v/v) ACN, (4) 25% (v/v) ACN, (5) 30% (v/v) ACN, (6) 35% (v/v) ACN.

Figure 3.4B Effect of acetonitrile (ACN) concentration on apparent mobilities of PAs. The experimental conditions are the same as in Figure 3.4A. 3.3.1.4 Effect of applied voltage

The effect of applied voltage on the separation was examined in the range from 10

kv to 25 kv. The applied voltage can influence the migration behavior and

8.000

10.000

12.000

14.000

16.000

18.000

20.000

10 15 20 25 30 35ACN concentration (%, v/v)

SenekirkineMonocratolineSenecionineRetrorsine

App

aren

t mob

ilitie

s（cm

2.m

in-1

. kv-1）


78

resolution directly. From Figure 3.5 we can see that the migration time decreased

with the increase of applied voltage. And Table 3.1 shows that when the voltage is

20 kv, the best resolution was reached. Hence, 20 kv was selected as the optimal

applied voltage.

Figure 3.5 Effect of applied voltage on the separation of PAs. Buffer: 50 mM ammonium acetate, 1% (v/v) acetic acid and 20% (v/v) ACN in methanol medium, (1) 10 kv; (2) 15 kv; (3) 20 kv; (4) 25 kv. Capillary: 65 cm (50 cm to detector) x 50 µm i.d. Detection: 200 nm.

Conditions Resolution between a and b

Resolution between b and c

Resolution between c and d

V=10 kv 9.109 2.035 1.692 V=15 kv 9.295 2.166 1.688 V=20 kv 10.264 2.355 1.766 V=25 kv 8.074 2.192 1.615

Table 3.1 Resolution at different applied voltages.

According to the work mentioned above, optimal separation was achieved with a

fused-silica capillary column of 65 cm (50 cm to detector) × 50 µm i.d. and a

running buffer containing 50 mM ammonium acetate, 1.0% (v/v) acetic acid and

20% (v/v) acetonitrile in methanol medium. The applied voltage was 20.0 kV. The

analytes were detected by UV at 200 nm.

3.3.1.5 Linearity, precision, LODs and LOQs

In order to evaluate the practical applicability of the proposed NACE method,


79

linearity, precision, LODs and LOQs were investigated using the optimal

separation conditions (Table 3.2). The results show good linearity between the

peak area (y) and the concentration (x) with a good regression coefficient (R2 ≥

0.9940) in the range of 8.3-166.7 mg/mL. The relative standard deviations

(R.S.D.s) (n = 4) of the migration time and peak area for the four PAs varied from

0.37-0.50% and 0.98-2.67%, respectively. LODs and LOQs for the four PAs

varied from 1.30-5.36 and 4.36-17.99 µg/mL at signal-to-noise ratios of 3 and 10,

respectively.


80

Pyrrolizidine Alkaloids

Regression equation

(y=kx+b)a

Correlation coefficient

(R2)

RSD in migration time

(n=4) (%)

RSD in peak area (n=4) (%)

Linear range

(µg/mL)

Limit of detection (LOD, µg/mL)b

Limit of quantification

(LOQ, µg/mL)c

Senkirkine y=0.077x+0.16 0.9960 0.50 0.98 8.3-166.7 5.36 17.99 Monocrotaline y=0.064x+0.09 0.9991 0.38 1.34 8.3-166.7 4.34 14.24 Senecionine y=0.129x+0.35 0.9991 0.37 2.67 8.3-166.7 2.19 7.15 Retrorsine y=0.292x+0.81 0.9966 0.39 1.72 8.3-166.7 1.30 4.36

Table 3.2 Linearity, precision, LODs and LOQs of NACE a y and x stand for the peak area and the concentration (mg/l) of the analytes, respectively. b the detection limit was defined as the concentration where the signal-to-noise ratio is 3. c the limit of quantification was define as the concentration where the signal-to-noise ratio is 10.


81

3.3.1.6 Application

The NACE method proposed in this study was applied for the determination of

PAs in Kuan Donghua. The sample solution was diluted one time with methanol

in water (50%) for CE analysis. The typical electropherograms of Kuan donghua

after extraction are illustrated in Figure 3.6. Only senkirkine was detected in Kuan

donghua. Peaks were identified by comparison of the migration times and by

spiking the standards into the sample solution. The concentration of detected

senkirkine in Kuan donghua was calculated to be 79.1µg/g.

Figure 3.6 The electropherograms of the standards mixture solution and the real sample. Buffer: 50 mM ammonium acetate, 1% (v/v) acetic acid and 20% (v/v) ACN in methanol medium; high voltage, 20 kv.; injection time, 10 s; injection pressure, 0.3 psi. Capillary: 65 cm (50 cm to detector) × 50 µm i.d. Detection: 200 nm. Real sample solution was diluted one time with methanol in water (50%) before injection.


82

3.3.2 Optimization of online preconcentration conditions

3.3.2.1 Large volume sample stacking (LVSS)

In LVSS, the optimized buffer mentioned above was used as the background

solution (BGS). The samples were prepared in a buffer with 100-fold lower

concentration of ammonium acetate resulting in a low conductivity sample zone.

After hydrodynamic injection was completed, +20 kV was applied to power the

CE separation. This procedure permits the PAs to move rapidly in the sample zone,

and then slowdown at the junction between the sample zone and the background

solution. As a result, the samples become concentrated at the boundary.

Firstly, the LVSS was applied to the 4 different PAs, respectively. From Figure

3.7 we can see that with the increase of the length of sample zone, the peaks

became broader and the peak height increased initially. However, when the length

of sample zone reached a certain level, the peak height decreased. These would

affect the efficiency of the stacking. Therefore, the LODs of the LVSS method

decreased at first, and then increased, which can be seen in Figure 3.8. For

senkirkine, senecionine and monocrotaline, the length of the sample zone could be

up to 15 cm, while the length was 10 cm for retroesine. In short, the improvement

in LOD for a 15 cm injection compared to a typical injection of 0.6 cm is 10, 20

and 12 times greater for senkirkine, senecionine and monocrotaline. For retrorsine,

the improvement is 160 times greater under the optimized sample zone length.


83

(1)

(2)

(3)


84

(4)

Figure 3.7 CE electropherograms of the 4 PAs obtained by NACE and LVSS, respectively. (1) Senkirkine A: Normal NACE of 8.3 ppm senkirkine with the optimal conditions. LVSS: BGS: 50 mM ammomnium acetate, 1% (v/v) acetic acid, 20% (v/v) ACN in methanol; sample zone: 0.5 mM ammonium acetate, 1% (v/v) acetic acid, 20% (v/v) ACN in methanol. Sample injection length: (B) 5 cm, (C) 10 cm, (D) 15 cm and (E) 20 cm. (2) Senecionie A: Normal NACE of 8.3 ppm senecionine with the optimal conditions. LVSS: BGS: 50 mM ammomnium acetate, 1% (v/v) acetic acid, 20% (v/v) ACN in methanol; sample zone: 0.5 mM ammonium acetate, 1% (v/v) acetic acid, 20% (v/v) ACN in methanol. Sample injection length: (B) 5 cm, (C) 10 cm, (D) 15 cm, (E) 20 cm and (F) 25 cm. (3) Monocrotaline, experimental conditions were the same as (1). (4) Retrorsine, experimental conditions were the same as (1).


85

0

0.2

0.4

0.6

0.8

1

1.2

1.4

1.6

1.8

0.6 5 10 15 20

Length of sample zone (cm)

LOD

(ppm

, S/N

=3)

SenkirkineSenecionineMonocrotalineRetrorsine

Figure 3.8 LODs of the 4 PAs with LVSS, respectively.

The peak broadening became even worse when the LVSS was applied to the

mixture of 4 PAs. It became obvious that the sample could not be “focused”

completely. Figure 3.9 shows typical CE electropherograms of PAs mixture when

the normal NACE and LVSS modes were used. As a result, 2cm was chosen as

the optimized sample zone length to obtain good resolutions. From Table 3.4, we

can see that around 5.23-7.65 fold improvement in detection sensitivity was

obtained under the optimal sample length.


86

Figure 3.9 CE electropherograms of PAs obtained by NACE and LVSS. (1) Normal NACE of 4PAs (16.7ppm each) with the optimal conditions. LVSS: BGS: 50 mM ammonium acetate, 1% (v/v) acetic acid, 20% (v/v) ACN in methanol; sample zone: 0.5 mM ammonium acetate, 1% (v/v) acetic acid, 20% (v/v) ACN in methanol. Sample injection length: (2) 1 cm, (3) 2 cm, (4) 5 cm. 3.3.2.2 Field-amplified sample stacking (FASS)

3.3.2.2.1 Choice of organic solvent plug

For FASS of NACE, a short plug of organic solvent was pre-introduced into the

capillary at the anodic end before the electrokinetic injection of the sample. The

mixture of 4 PAs with 1% (v/v) acetic acid in methanol was injected at a potential

of 10 kV for 20 s after a 10 s organic solvent plug was injected at a pressure of 0.5

psi. ACN, MeOH, and EtOH were used as the pre-injected plug. From Figure 3.10

we can see that when ACN was used as the organic solvent plug, the largest peak

areas were obtained for all the four PAs. Even though the peak areas obtained

from no plug were sometimes a little higher or compared with those obtained from

ACN plug, ACN was still chosen as the organic solvent plug. This is because the

repeatability of FASS with no plug is not ideal.


87

0

5

10

15

20

25

30

35

40

45

50

No plug ACN Methanol Ethanol

Types of organic solvent plug

Peak

Are

a (m

v.s)

senkirkinemonocrotalinesenecionineretrorsine

Figure 3.10 Effect of organic solvent plug on FASS. Sample: 7.5 ppm senkirkine, 6.25 ppm senecionine, 7.5 ppm monocrotaline and 7.5 pm retrorsine, 1% (v/v) acetic acid in methanol. Buffer: 50 mM ammonium acetate, 1% (v/v) acetic acid, 20% (v/v) ACN in methanol. The organic solvent plug was pre-injected at 0.5 psi for 10 s. Sample was injected at +10 kv for 20 s. The separation voltage was +20 kv. 3.3.2.2.2 Optimization of sample matrix

As far as FASS is concerned, the difference between the conductivity of the

sample matrix and the buffer is another parameter that could affect the stacking

efficiency. The lower the conductivity of sample matrix, the higher the stacking

efficiency. Here, 3 different kinds of sample matrixes were investigated. It can be

seen from Figure 3.11 that samples dissolved in methanol give the highest

stacking efficiency. As a result, methanol was chosen as the optimized sample

matrix.


88

Figure 3.11 Effect of sample matrix on FASS. Sample matrix: (A) 4 PAs, 1% (v/v) acetic acid in CH3OH:H2O=50:50, (B) 4 PAs , 1% (v/v) acetic acid in H2O, (C) 4 PAs, 1% (v/v) acetic acid in CH3OH. All the other conditions were the same as Figure 3.10. 3.3.2.2.3 Optimization of organic solvent plug injection time

The effect of the ACN plug injection time on FASS stacking efficiency was

investigated from 0 to 20 s at the pressure of 0.5 psi. Figure 3.12A shows that the

peak heights of the 4 PAs increased when the ACN plug was injected from 0 to 10

s at 0.4 psi. This is because the ACN plug provides a higher electric field at the tip

of the capillary, which could improve the sample stacking procedure. Further

prolongation of the injection time reduced the peak heights. The change of peak

areas was similar to that of the peak heights as shown in Figure 3.12B. The

resolutions also became poorer with the further longer injection time. Considering

both of the stacking efficiency and resolution, 10 s was selected as the optimized

injection time for ACN plug.


89

Figure 3.12 Effect of organic solvent plug injection time on FASS. A: Overlaid CE electropherograms of FASS with different ACN plug injection time, (a) 0 s, (b) 5 s, (c) 10 s, (d) 15 s and (e) 20 s. B: Effect of ACN plug injection time on peak areas of 4 PAs. 3.3.2.2.4 Optimization of sample injection time and injection voltage

Herein, the effect of sample injection times (from 15 to 50s) and injection voltage

(from 6 to 14 kV) on FASS were studied. Figure 3.13A and Figure 3.13B show

that the peak areas of the 4 PAs increased with increasing injection time or

0

5

10

15

20

25

30

35

40

0 5 10 15 20

Injection time of ACN (s)

Peak

Are

a (m

v.s)


B


90

injection voltage. However, the resolution became poorer with the increasing of

sample injection time or injection voltage. This might be caused by the PAs

passing through the boundary between the ACN plug and the running buffer and

dispersing into the running buffer. As a result, the stacking of the analytes became

incomplete and the resolution decreased. High resolution and stacking efficiency

were achieved at an injection voltage of 10 kv and injection time of 35 s.

0

10

20

30

40

50

60

70

80

15 20 25 30 35 40 45 50

Sample injection time (s)

Peak

Are

a (m

v.s)


A


91

Figure 3.13 Effect of sample injection time and injection voltage on peak areas. According to the experiments mentioned above, the optimal separation and

stacking conditions were obtained when the sample was injected at 10 kv for 35 s

after preliminary pressure injection of ACN (0.5 psi) for 10 s and separated with

the buffer containing 50 mM ammonium acetate, 1% (v/v) acetic acid, 20% (v/v)

ACN in methanol at 20 kv.

3.3.2.2.5 Linearity, precision, LODs and LOQs

Since the stacking efficiency is not so significant for LVSS, only the linearity,

precision, limits of detection and limits of quantification were investigated under

the optimized conditions to evaluate the practical applicability of the proposed

FASS method (Table 3.3). The results showed good linearity between the peak

area (y) and concentration (x) with good regression coefficient (R2>0.9957) in the

range of 0.47-7.5 ppm for senkirkine, monocrotaline, retrorsine and 0.39-6.25

0

10

20

30

40

50

60

70

80

90

6 8 10 12 14

Sample injection voltage (KV)

Peak

Are

a (m

v.s)


B


92

ppm for senecionine. R.S.D.s (n=4) in migration time and peak area of the four

PAs varied from 0.12 - 0.37% and 0.53 - 4.87%, respectively. Limits of detection

(LODs) and limits of quantification (LOQs) for the four PAs varied from

60.5-135.9 and 199.92-237.01 ppb at signal-to-noise ratios of 3 and 10,

respectively. Table 3.4 shows the limits of detection (S/N=3) of different CE

methods for the analysis of PAs. Compared with normal NACE as well as LVSS,

sensitivity improvement ranging from 18.56 to 89.33 folds was realized for the

studied PAs by the proposed FASS method and it permits the direct analysis of

trace compound in Tussilago farfara L. (Kuan Donghua).


93

Pyrrolizidine Alkaloids

Regression equation

(y=kx+b)a

Correlation coefficient

(R2)

RSD in migration time

(n=4) (%)

RSD in peak area (n=4) (%)

Linear range

(µg/mL)

Limit of detection

(LOD, µg/L)b

Limit of quantificatrion (LOQ, µg/L)c

Senkirkine y=0.113x-0.33 0.9957 0.12 0.53 0.47-7.5 60.5 203.11 Monocrotaline y=0.292x+0.16 0.9976 0.23 3.90 0.47-7.5 135.9 445.75 Senecionine y=0.150x+0.19 0.9964 0.27 4.87 0.39-6.25 61.2 199.92 Retrorsine y=0.125x+0.22 0.9971 0.37 4.45 0.47-7.5 70.6 237.01

Table 3.3 Precision, linearity, LODs and LOQs of FASS. a y and x stand for the peak area and the concentration (mg/l) of the analytes, respectively. b the detection limit was defined as the concentration where the signal-to-noise ratio is 3. c the limit of quantification was define as the concentration where the signal-to-noise ratio is 10.

Pyrrolizidine alkaloids

Limits of detection (LOD, µg/mL)a and sensitivity enhancement (SE) NACE LVSS FASS LOD LOD SE LOD SE

Senkirkine 5.36 0.87 6.16 0.06 89.33 Monocrotaline 4.34 0.83 5.23 0.14 31.00 Senecionine 2.19 0.39 5.62 0.06 36.50 Retrorsine 1.30 0.17 7.65 0.07 18.56

Table 3.4 Comparison of limits of detection (LOD) and sensitivity enhancement (SE). a the detection limit was defined as the concentration where the signal-to-noise ratio is 3.


94

3.3.3 Application to Real Sample

The FASS method proposed was applied for the determination of PAs in Kuan

Donghua. One gram pulverized powder of Kuan Donghua was extracted by using

the MAE method described in section 3.2.4.2. The sample solutions were diluted

10 times with methanol. The typical electropherogram of the standard mixture and

Kuan Donghua is illustrated in Figure 3.14. Both the major senkirkine and minor

senecionine were detected. Peaks were identified by comparison of the migration

times and by spiking the standards into the sample solution. The concentration of

detected senkirkine and senecionine in Kuan donghua were calculated to be 77.36

and 2.86µg/g, respectively. The results were consistent with those obtained from

the dynamic PH junction-sweeping strategy [36] and LC/MS method discussed in

chapter 2.

Figure 3.14 Electropherograms of the standard mixture (A) and the real sample (B) under the optimum conditions. (a): senkirkine; (b): monocrotaline; (c): senecionine; (d): retrorsine.


95

3.4 Conclusion

A simple, economical and efficient NACE method was developed for the

separation and determination of PAs in the Chinese herbal medicine, Kuan

Donghua. Two stacking methods, LVSS and FASS were established for the

analysis of PAs in Kuan Donghua also. Compared with LVSS, FASS was found

to be superior in this study in terms of sensitivity enhancement, which could

provide 20-100 fold enhancements. Generally, this method provides a rapid and

sensitive alternative approach for the analysis of PAs in plants and for the quality

control of natural products.


96

References [1] A. R. Mattocks, Chemistry and Toxicology of Pyrrolizidine Alkaloids,

Academic Press, New York, London, 1986.

[2] A. F. M. Rizk, In Naturally Occurring Pyrrolizidine Alkaloids, A. M. Rizk, Ed.,

CRC Press, Boca Raton, FL, 1991, p. 1.

[3] T. Hartmann and L. Witte, Alkaloids: Chemical and Biological Perspectives, S.

W. Pelletier, Ed., Vol. 9, Pergamon Press, Oxford, 1995, p. 155.

[4] P. P. Fu, Y. C.Yang, Q. S. Xia, M. W. Chou, Y. Y. Cui, G. Lin, J. Food Drug

Anal. 10 (2002), 198.

[5] Pyrrolizidine Alkaloids, World Health Organization, Geneva (1988), EHC80.


Anal. 11 (2000), 366.

[7] W. Helmut, L. Roxana, K. Brigitte, Deutsche Apotheker Zeitung 135 (1995),

17.

[8] R. A. Steinbach, H. Miething, Pharmazeutische Zeitung 134 (1989), 25-6,

28-9.

[9] T. Mroczek; K. Glowniak, A. Wlaszczyk, J. Chromatogr. A 949 (2002), 249.

[10] J. P. B. Bartkowski, H. Wiedenfeld, E. Roeder, Phytochemical Analysis 8

(1997), 1.

[11] H. T. Feng, S. F.Y. Li, J. Chromatogr. A 973 (2002), 243.

[12] Y. Q. Li, S. D. Qi, X. G. Chen, Z. D. Hu, Electrophoresis 25 (2004), 3003.

[13] A. Marsh, B. J. Clark, K. D.Altria, Electrophoresis 25 (2004), 1270.

[14] Y. Q. Li, S. Y. Cui, Y. Q. Cheng, X. G. Chen, Z. D. Hu, Anal. Chim. Acta

508 (2004), 17.

[15] Y. Q. Yu, P. L. Ding, D. F. Chen, Anal. Chim. Acta 523 (2004), 15.


97

[16] S. Kreft, J. Zel, M. Pukl, A. Umek, B. Strukelj, Phytochem. Anal. 11 (2000),

3.

[17] S. X. Gong, X. D. Su, T. Bo, X. Zhang, H. W. Liu, K. A. Li, J. Sep. Sci. 26

(2003), 549.

[18] M. M. Reddy, V. Suresh, G. Jayashanker, B. S. Rao, R. K. Sarin,

Electrophoresis 24 (2003), 1437.

[19] L. Mateus, S. Cherkaoui, P. Christen, K. M.Oksman-Caldentey,

Phytochemistry 54 (2000), 517.

[20] J. Koyama, I. Toyokuni, A.Kino, K. Tagahara, Phytochem. Anal. 10 (1999),

175.

[21] C. W. Chiu, H. H. Liang, H. Y. Huang, Electrophoresis 28 (2007), 4220.

[22] S. F. Wang, H. B. Qu, Y. Y. Cheng, Electrophoresis 28 (2007), 1399.

[23] Q. Liu, Y. J. Liu, M. L. Guo, X. B. Luo, S. Z. Yao, Talanta 70 (2006), 202.

[24] Y. Q. Li, S. D. Qi, X. G. Chen, Z. D. Hu, Electrophoresis 25 (2004), 3003.


Anal. 11 (2000), 366.

[26] Y. Q. Li, S. Y. Cui, Y. Q. Cheng, X. G. Chen, Z. D. Hu, Anal. Chim. Acta

508 (2004), 17.

[27] S. Morales, R. Cela, J. Chromatogr. A 846 (1999), 401.

[28] B. Kim, D. S. Chung, Electrophoresis 23 (2002), 49.

[29] Z. K. Shihabi, Electrophoresis 23 (2002), 1628.

[30] C. H. Tsai, C. Fang, J. T. Liu, C. H. Lin, Electrophoresis 25 (2004), 1601.

[31] F. Y. He, R. M. Liu, Fenxi Shiyanshi 22 (2003), 38.

[32] S. Morales, R. Cela, Electrophoresis 23 (2002), 408.

[33] W. H. Gao, S. Y. Lin, Y. W. Chen, A. J. Chen, Y. Q. Li, X. G. Chen, Z. D.


98

Hu, J. Sep. Sci. 28 (2005), 639.

[34] F. Ca˜nada-Ca˜nadaa, A. Kasselouria, P. Prognona, P. Maillardb, D. S.

Griersonb, S. Descroixa, M. Tavernaa, J. Chromatogr. A 1068 (2005), 123.

[35] C. H. Tsai, C. C. Tsai, J. T. Liu, C. H. Lin, J. Chromatogr. A 1068 (2005),

115.

[36] L. J. Yu, S. F. Y. Li, Electrophoresis 26 (2005) 4360.

99

Part II: Metabonomic Study of Natural Alkaloid in Rats

Chapter 4 Metabolic Profiling of Berberine in Rats with GC/MS, LC/MS and 1H NMR Spectroscopy

100

Chapter 4 Metabolic Profiling of Berberine in Rats with gas

chromatography/mass spectrometry, liquid chromatography/mass

spectrometry and 1H NMR spectroscopy

4.1 Introduction

Berberine (Figure 4.1), an isoquinoline alkaloid of the protoberberine type, is

derived from the root, rhizome, and stem bark of many plant species such as

Coptis chinensis Franch., Coptis japonica Makino., Berberis thunbergii DC.,

Hydrastis canadensis L., and Thalictrum lucidum Ait. With the quaternary base in

the chemical structure, the commercial product of berberine is generally prepared

as a chloride or sulfate for clinical applications. Berberine has been commonly

used for the treatment of gastroenteritis and secretory diarrhea in traditional

Chinese medicine for a long history. To date, many pharmacological studies have

been done on berberine. It was found that berberine also shows antimicrobial,

anti-HIV, anti-inflammatory, antiarrhythmic, cerebroprotective, anticancer,

vasorelaxing, antimutagenic, analgesic, antifungal, cardioprotective,

immunoregulative, antimalarial, antioxidative and anxiolytic effects [1-17] and

liver might be the metabolic site for berberine [18, 19]. However, the

metabonomic studies on berberine have been limited.

Figure 4.1 Chemical structure of berberine.


101

Metabonomics is defined as the quantitative measurement of the dynamic

multiparametric metabolic response of a living system to pathophysiological

stimuli or genetic modifications [20]. It is based on the analysis of endogenous

metabolites of different kinds of biofluids and tissues extractions, and aims to

explore a potential relationship between the changed metabolic profiles and the

physiological status of the biosystem. As a result, it can augment and complement

the information provided by measuring the genetic and proteomic responses to

xenobiotic exposure. Metabonomics techniques have been found to be extremely

useful tools for probing the mechanisms and evaluating the safety of traditional

Chinese medicines [21-24].

As far as the analytical techniques used in metabonomics studies are concerned,

high-resolution nuclear magnetic resonance (NMR) spectroscopy is the most

commonly used technique. A lot of the original metabonomics studies were

performed using NMR spectroscopy. The ubiquitous used 1H NMR has the

advantages of being non-destructive, applicable to intact biomaterials and rich in

structural information. It is an inherently robust technique which can report a

number of compounds in a single measurement with minimal sample preparation.

Today NMR spectroscopy has been extensively applied for metabonomics studies

[25-36].

Another commonly used analytical technique is mass spectroscopy (MS) normally

coupled with LC or GC. This technique provides much greater sensitivity and

higher resolution than NMR spectroscopy. In addition, MS can provide useful

information for molecular identification, especially using tandem MS (MS-MS)

methods. However, the quantitation of analytes might be impaired by variable

ionization and ion suppression effects. Recently, ultra performance liquid


102

chromatography (UPLC) has been introduced to improve the chromatographic

peak resolution, separation speed and sensitivity. The application of

chromatography-mass spectrometry to metabonomics has been reviewed [37]. In

recent years, MS has been successfully used for metabonomics research [38-45].

At the same time, the combination of 1H NMR and MS has been used for the

investigation of metabolic profiles associated with the effects of drugs and

chemical substances [46-51]. They are highly complementary approaches for

metabonomics studies, and the use of both techniques is often necessary for full

molecular characterization. However, the manual processing and exporting of

LC/MS data can be rather tedious [52, 53]. It was demonstrated that a targeted

profiling with 1H NMR produced robust models, generated accurate metabolite

concentration data and provided information that can be used to help understand

metabolic differences in a healthy population [54-56]. However, the used of

targeted profiling with LC/MS has not been reported.

A successful metabonomics study can’t be achieved without the multivariate

analysis. Principal components analysis (PCA) is one of the simplest techniques

that have been used extensively in metabonomics. And most of the variance

within a data set can be expressed by a smaller number of factors or principal

components with PCA. For visualization, each sample would occupy a position

based on its metabolite composition. As a result, it is much easier for us to

investigate the evolution of metabolites.

As the manual processing of LC/MS can be rather tedious, we use an approach

based on targeted profiling with LC/MS. In this study, we report a comparison of

urinary metabolites profiles from normal rats and rats treated with berberine (50

mg/kg) using targeted profiling with LC/MS and non-targeted profiling with 1H


103

NMR. The current approach is used to study the urinary metabolic profiles

associated with the glucose reduction, lipid and cholesterol lowering effects of

berberine in the livers of Sprague Dawley rats.

4.2 Experimental

4.2.1 Chemicals

Methanol and acetonitrile of HPLC grade were purchased from APS (NSW,

Australia). Pure water was obtained from Millipore Alpha-Q water system

(Bedford, MA, USA). Formic acid was purchased from Merck (Darmstadt,

Germany). Valeric acid, L-leucine, D-serine, homocysteine, lysine, creatinine,

phenylalanine and hippuric acid were purchased from Sigma (St. Louis, MO,

USA).

4.2.2 Animal Studies

Rats (Sprague Dawley) were obtained from Laboratory Animals Centre, National

University of Singapore. Animals were acclimatized in standard rodent cages with

individual ventilation. The animal room was maintained at 25 + 2 oC with natural

day/night cycle. Following a 7 day acclimatization period, animals were randomly

allocated into 2 groups comprising 8 animals each. Food and water were provided

ad libitum. The treatment group was administered berberine (Sigma, Singapore) at

a dose of 50 mg/kg by intraperitoneal injection at 0 and 48 hours. All animals

were housed individually in metabolism cages for the ease of urine collection. The

control group received saline by intraperitoneal injection. Body weights of

individual rats in each group were measured at the beginning of the experiment,

weekly and at the time of sacrifice. All animals were sacrificed at the end of the

experiment. Post mortem examination was done on all animals. The livers and

kidneys of each rat from the control and treatment groups were preserved in 10%


104

buffered formalin solution for histopathological examination.

4.2.3 Histopathological Examination

The livers and kidneys of control and treated animals were processed and stained

with hematoxylin and eosin. Sections were examined under the light microscope.

4.2.4 Sample preparation for metabonomics profile of rat urine samples

Samples of rat urine were collected during the same (9 to 12 noon) time interval

daily for a total of 6 days after treatment. The rat urine collected from the control

and treatment groups were stored frozen at –20oC prior to analysis. For HPLC

assay, a 20 μl aliquot of rat urine was diluted to 100 μl with distilled water.

4.2.5 Reversed-phased LC/MSMS

For LC/MS assay, a 30 μL aliquot of urine was diluted to 70 μL with distilled

water. An Agilent 1200 RRLC system (Waldbronn, Germany) equipped with a

binary gradient pump, auto-sampler, column oven and diode array detector was

coupled with an Agilent 6410 triple quadrupole mass spectrometer. The gradient

elution involved a mobile phase consisting of (A) 0.1% formic acid in water and

(B) 0.1% formic acid in acetonitrile. The initial condition was set at 5% of (B),

gradient up to 100% in 10 minutes and returning to initial condition for 5 minutes.

Oven temperature was set at 50°C and flow rate was set at 200 μL/min. For all

experiments, 5 μL of samples were injected. The column used for the separation

was a reversed-phase Zorbax SB18 (50 x 2.0 mm, 1.8 μm, Agilent Technologies,

USA). The ESI-MS was acquired in the positive ion mode. The product ions of

m/z range from 100 to 800 were collected. The heated capillary temperature was

maintained at 350oC, the drying gas and nebulizer nitrogen gas flow rates were

10L/min and 50 psi respectively.


105

4.2.6 Analysis of urine samples by 1H NMR

300 µL of urine samples were buffered with 300 µL of 0.2 M phosphate

buffer/D2O (pH7.4) prior to analysis by 1H NMR. The solutions were pipetted into

NMR tubes (5mm O.D, 7 inch length, Sigma-Aldrich) and one dimensional 1H

NMR spectra were obtained on a Bruker DRX500 operating at 500.15 MHz

observation frequency. Water suppression was achieved by applying the standard

Noesypresat pulse sequence (Bruker Biospin Ltd) with secondary irradiation of

the dominant water signal during the mixing time of 150 ms and the relaxation

delay of 2 s. Spectra were referenced to the internal reference standard TSP

dissolve in D2O to provide a field-frequency lock.

4.2.7 Analysis of liver samples by GC/MS

Individual liver was dried in freeze dryer and extracted with 0.5 mL of

CHCl3/CH3OH (3:1). After centrifugation at 161,000 rcf for 5 minutes, the

supernatant was lyophilized. The dried sample (lipid fraction) was kept at -30 °C

prior to analysis using GC/MS. For the liver tissue extract, the dried sample was

reconstituted in 50 μL EA. 50μL of BSTFA, pyridine and EA (3:1:1, v/v/v)

mixture was added to the tissue extract together with the standard solutions. The

resulting solution was vortexed for 1 minute at room temperature and transferred

to an amber glass vial for analysis using GC/MS.

1.0 μL aliquot of the derivatized sample with standard was injected using the

splitless mode with an Agilent 7683 Series autosampler (Agilent Technologies)

into an Agilent 6890 GC system equipped with an Agilent HP-1MS capillary

column (15 m × 0.25 mm; ID × 0.25 μm). The inlet temperature was set at 300°C.

Helium was used as the carrier gas with a constant flow rate 1.40 mL/min through

the column. The initial temperature was set at 100 °C. 1 minute after injection the


106

GC temperature was increased at a rate of 10 °C/min to 300 oC and held for 3

minutes at 300 oC. The transfer line temperature was set at 300 oC. Detection was

achieved using MS in electron impact mode and full scan monitoring (m/z 50 to

800). The temperature of the ion source was set at 200 °C, and the quadrupole was

set at 150 °C.

4.2.8 Chemometric analysis

Fourier transformed 1H NMR spectra were manually phased and baseline

corrected using XWINNMR 3.5 (Bruker Biospin, Rheinstetten, Germany). Each

spectrum was integrated between 0.5 to 4.5 and 5.1 to 10 ppm. The spectral region

containing the water resonance was removed from all data sets prior to

normalization and multivariate data analysis in order to eliminate variation due to

water suppression efficiency. The resulting two-dimensional data, 1H chemical

shift, and peak heights were generated.

For GC/MS, each sample was represented by a GC/MS total ion chromatogram

(TIC). The data was exported to Genespring 1.1.1 (Agilent Technologies, USA)

for the determination of perturbed metabolites. Among the detected peaks, a

multi-dimensional vector was constructed to characterize the biochemical pattern.

Each vector was normalized to the total sum of vector intensity, thereby partially

accounting for concentration due to different sample sizes used. Peaks due to

column bleed and derivatization reagent were removed. The identification of

peaks was based on the use of reference standards and NIST98 library. The mass

spectra obtained were inspected manually and only those molecules with

probability matching higher than 90% were considered.

The resulting LC-MS data served as raw data for PCA analysis. The LC-MS data

were peak-detected and noise-reduced such that only true analytical peaks were


107

further processed by the PCA software. For targeted profiling, a list of the peak

areas of the peaks detected from Table 4.3A and Table 4.5A and m/z 100-110,

200-210, 300-310 and 400-410 were then generated manually. The data was

tabulated into Microsoft Excel for each sample run, using the retention tine (RT)

and m/z data pairs as the identifier for each peak. The peak areas for each peak

detected were then normalized within each sample, to the sum of the peak area in

that sample. Normalization was required to remove concentration differences

between dilute and concentrated urine samples. To account for any difference in

concentration between samples, all 1H NMR and LC/MS data were normalized to

a total value of 100. The resulting three-dimensional data for 1H NMR and LC/MS

were analyzed by PCA. The resulting data were then exported to Simca-P+

Software package (Umetrics, Umea, Sweden) for subsequent processing by

unsupervised and supervised methods. For PCA, the data were reduced to 2

latent variables (or principal components, PCs) that would describe maximum

variations within the data. The PCs which were obtained from the scores would

highlight clustering, trends and outliers in the observation direction in the data set.

4.2.9 Statistical analysis

From the normalized data obtained from LC/MS and 1H NMR, indication of

significance was based on a two-tailed student t-test performed with SPSS 14.0

for Windows (SPSS, Chicago, IL). For the identification of potential biomarkers,

two-tailed student t-test (P < 0.05 and P<0.01) were used.

4.3 Results and Discussion

4.3.1 Results

4.3.1.1 Body weight and histopathology

The body weights of all the rats kept on increasing in the acclimatization period


108

(results not shown here). Figure 4.2 shows the trends of body weights of rats after

the injection of saline and berberine. The body weights of the control group kept

on increasing over the 5 days period. However, the body weights of the treatment

group decreased after the administration of berberine. It was found that they had

recovered from the berberine effect with slight increasing in body weights on day

5. This phenomenon suggested the cholesterol lowering effect caused by berberine

in the treated animals.

150

160

170

180

190

200

210

220

230

240

250

1 3 4 5

Days

Wei

ght o

f Rat

s (g)

Figure 4.2 The trends of body weights of rats. The results shown are averages with standard deviations (n=8). (Blue: control group, red: treatment group) Histopathological findings after the administration of berberine are summarized in

Figure 4.3. Significant changes on liver and kidney pathology were not observed

in the treatment group.


109

(A) (B)

(C) (D)

Figure 4.3 Histopathological photomicrographs of rat livers and kidneys from control group and treatment group with dose of 50 mg/kg of berberine. H&E staining, Magnification 400×. (A) control rat liver, (B) treated rat liver, (C) control rat kidney, (D) treated rat kidney. 4.3.1.2 Determination of metabolites in rat livers samples by GC/MS

A capillary GC/MS using EI was carried out after the derivatization of the liver

extracts. Figure 4.4 shows the representative GC/MS chromatograms for the lipid

fraction of liver extracts. Significant information can be obtained by visual

examination of the TICs from the liver samples of the control and treatment

groups. The two groups exhibited unique metabolic profiles.

For the GC/MS data, the peak areas were normalized to a constant sum and the

PCA score plot (Figure 4.5A) shows two distinct clusters for data obtained from

the lipid fraction of liver extracts of both control and treatment groups. From


110

Figure 4.5A we can also see that the control group was clustered in a small region.

This might be due to the biological variation.

The perturbed metabolites detected using GC/MS on different liver extracts are

shown in Figure 4.5B. Metabolites that were observed to be significantly different

(based on a two-tailed Student’s t-test) in the livers of the control and treated rats

included cholesterol, glucose, maltose and fatty acids. The compounds that were

found to be lower in the treated group included cholesterol, glucose, maltose,

butanoic acids, propanoic acids, tetradecanoic acid, palmitoleic acid, arachidonic

acid and 5,8,11,14,17 eicosapentaenoic acid methy ester.

(A)

23.50 24.00 24.50 25.00 25.50 26.00 26.50

Control group

Treatment group

23.50 24.00 24.50 25.00 25.50 26.00 26.5023.50 24.00 24.50 25.00 25.50 26.00 26.50

Control group


111

(B)

(C)

Figure 4.4 Typical GC/MS chromatograms of the lipid fraction of rat liver extracts. (A) cholesterol, (B) glucose and (C) total ion chromatogram (TIC) of the lipid fraction of rat liver extracts. (Black: control group, Blue: treatment group)

4.00 6.00 8.00 10.00 12.00 14.00 16.00 18.00 20.00 22.00 24.00 26.00

14.00 14.20 14.40 14.60 14.80 15.00 15.20 15.40 15.60 15.80

Control group

Treatment group

14.00 14.20 14.40 14.60 14.80 15.00 15.20 15.40 15.60 15.8014.00 14.20 14.40 14.60 14.80 15.00 15.20 15.40 15.60 15.80

Control group

Counts vs. Acquisition Time (min)


112

(A)

(B)

Figure 4.5 (A) PCA scores plot based on GC/MS analysis of rat liver samples

Citrate

Isocitrate

2-Oxoglutarate

Succinyl-CoASuccinate

Fumarate

Malate

Oxaloacetate

TCA cycle

Fatty acids 1) butanoic acids↓** 2) propanoic acids↓* 3) tetradecanoic acid ↓** 4) palmitoleic acid↓** 5) arachidonic acid↓** 6)5,8,11,14,17 eicosapentaenoic acid methy ester ↓*

Glucose phosphates

Maltose ↓*

Lactate

Cholesterol ↓**

Triose phosphates

Acetyl-CoA

Pyruvate

Glucose ↓**

Treatment group

Control group


113

from all the control (n=8) and treatment group (n=8) (PCA component 1: 34.7 % variance, PCA component 2: 17.1 % variance), (B) Simplified pathway illustrating perturbed metabolites in the rat liver samples between the control and treatment group. The statistics are as follows: significant difference between the control group (n=8) and treatment group (n=8) is based on two tailed student t test (* p<0.05, ** p<0.01). 4.3.1.3 Determination of metabolites in rat urine samples by 1H NMR

A typical time course of 1H NMR spectra of the pre-dose and treatment groups’

urine samples acquired using the standard 1D pulse sequence for water

suppression is shown in Figure 4.6. It can be seen from Figure 4.6 that there exist

differences in some of the peak intensities, suggesting that the two groups might

have distinct metabolic profiles.

0.00.40.81.21.62.02.42.83.23.64.04.44.8

(ppm)

Figure 4.6 Typical 1H NMR spectrum (0-5 ppm) of rat urine samples obtained from pre-dose and treatment group (day 1, 2, 3 and 4).

Leucine/valine

Pyruvate

Pre-dose

Day 1

Day 2

Day 3

Day 4

acetoactetae creatinine

acetate

glutamate

lactate

citrate

creatine

TSP

3-D-Hydroxybutyrate

alaninesuccinate

TMAO

glucose


114

The metabolites being identified had their resonance determined previously and

the compounds that can be detected included the amino acids, carboxylic acids,

ketone bodies and methylamines, and others. The 1H chemical shift assignments

of the metabolites in the rat urine samples were based on other reports and online

databases [57-61]. The completed data were summarized in Table 4.5 (Appendix).

After the administration of berberine, an elevation in creatine, glucose and glycine

was observed. Decreases in tryptamine, tyrosine, α-ketoglutarate, dimethylglycine,

trimethylamine (TMA), aspartic Acid, glutamine, succinate and leucine were also

observed. Table 4.1 shows the selected metabolites identified in rat urine samples

for both control and treatment group on different days. The changes of these

selected components were more significant on the next day of berberine

administration. Figure 4.7 shows the changes of other selected metabolites as

measured by 1H NMR on different days. From Figure 4.7 we can see that pyruvate

could recover from the berberine effect while glucose and lactate couldn’t. The

concentrations of glucose and lactate were higher in the treatment group than in

the control one.


115

Normalized peak intensitya NO. Metabolite Chemical Shift

(ppm) Multiplicity Control

Ave ±SD Treated

Ave ±SD

Pre-dose 1 Fumurate 6.53 s 0.18 0.04 0.17 0.04 2 Creatinine 3.08 s 0.15 0.03 0.15 0.07 3 Lysine 3.74 t 0.41 0.21 0.38 0.26 4 Citrate 2.60 d 0.32 0.05 0.34 0.06 Citrate 2.50 d 0.37 0.07 0.32 0.08 Day 1

1 Fumurate 6.53 s 0.26 0.03 0.23 0.09 2 Creatinine 3.08 s 0.21 0.03 0.17 0.06 3 Lysine 3.74 t 0.65 0.27 1.79 1.48 4 Citrate 2.60 d 0.34 0.04 0.28 0.08 Citrate 2.50 d 0.41 0.09 0.49 0.38 Day 2 1 Fumurate** 6.53 s 0.27 0.02 0.13 0.06 2 Creatinine** 3.08 s 0.19 0.03 0.33 0.04 3 Lysine** 3.74 t 0.71 0.10 1.08 0.19 4 Citrate 2.60 d 0.28 0.12 0.32 0.08 Citrate 2.50 d 0.36 0.09 0.41 0.03 Day 3 1 Fumurate 6.53 s 0.20 0.03 0.31 0.33 2 Creatinine* 3.08 s 0.25 0.03 0.13 0.12 3 Lysine* 3.74 t 0.91 0.11 2.58 1.94


116

4 Citrate 2.60 d 0.32 0.13 0.15 0.13 Citrate 2.50 d 0.30 0.05 0.17 0.16 Day 4

1 Fumurate 6.53 s 0.15 0.04 0.14 0.07 2 Creatinine** 3.08 s 0.18 0.03 0.27 0.06 3 Lysine** 3.74 t 0.48 0.06 0.77 0.17 4 Citrate 2.60 d 0.30 0.03 0.44 0.21 Citrate 2.50 d 0.36 0.14 0.39 0.07

aThe relative intensity of metabolites is expressed with their normalized peak area. Values are represented as mean ±SD (standard deviation). The statistics are as follows: significance difference between the control group (n=8) and treatment group (n=8) is based on two tailed student t test (* p<0.05, ** p<0.01).

Table 4.1 Selected metabolites identified in rat urine samples for both control and treatment group as measured by 1H NMR.


117

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0 1 2 3 4 5

day

inte

nsity

0

0.5

1

1.5

2

2.5

0 1 2 3 4 5

day

inte

nsity

(A) (B)

0

0.2

0.4

0.6

0.8

1

1.2

0 1 2 3 4 5

day

inte

nsity

(C)

Figure 4.7 Changes of selected metabolites as measured by 1H NMR on different days (blue colour: control group, n=8, pink colour: treatment group, n=8). (A) Glucose, (B) Pyruvate and (C) Lactate.

For the 1H NMR data, the peak heights were normalized to a constant sum to

compensate for the differences in overall concentration between individual urine

samples. The PCA scores plots of the control and treatment groups are illustrated

in Figure 4.8. The plot indicates that there exists a separation between the control

and treatment groups. It suggests that there might be a difference in the metabolic

urinary excretion of metabolites from the two groups.


118

(A)

(B)

-20

-10

0

10

-20 -10 0 10 20

t[1]

t[2]

-12 -10 -8

-6

-4

-2

0 2 4 6 8

10 12

-20 -10 0 10 20

t[1]

t[2]


119

(C)

(D)

-10

0

10

-20 -10 0 10 20

t[1]

t[2]

-10

0

10

-30 -20 -10 0 10 20 30

t[1]

t[2]


120

(E)

Figure 4.8 Data analysis of 1H NMR rat urine metabolic profiles. (A) PCA scores plot for data from all rat urine samples collected on day 1; (B) PCA scores plot for data from all rat urine samples collected on day 2; (C) PCA scores plot for data from all rat urine samples collected on day 3; (D) PCA scores plot for data from all rat urine samples collected on day 4; (Black: control group, red: treatment group). (E) PCA scores plot for data from all rat urine samples collected (black triangle: control group from day 0, black open square: control group at day 4 and red diamond: treatment group at day 4) (R2X[1] = 0.346, R2X[2]= 0.114)

-10

0

10

-20 -10 0 10 20

Control group Treated group

t[1]

t[2]


121

Figure 4.9 and Figure 4.10 demonstrate the integrated metabolic profile on day 4.

Figure 4.9 Simplified pathway illustrating perturbed metabolites in the rat urine samples on day 4 between the control and treatment group. The statistics are as follows: significance difference between the control group (n=8) and treatment group (n=8) is based on two tailed student t test (* p<0.05, ** p<0.01).

Acetyl-CoA

Citrate

Isocitrate

α-Ketoglutarate↓**

Succinyl-CoASuccinate

Fumarate

Malate

Oxaloacetate

TCA cycle

Fatty acids and lipids 1) adipic acid ↑* 2) mystric acids ↑* 3) palmitoleic acid↑* 5) stearamide ↑* 6) oleamide ↑**

Glucose phosphates

Arginine↑** Glutamate↑** Glutamine↑**

Leucine↑** Methionine↑** Valine↑**

Phenylalanine↑**

Alanine↑** Glycine↑** Serine↑**

Aceto-acetyl-CoA

Leucine ↑** Phenylalanine↑** lysine↑**

Glucose ↑**

Triose phosphates

Pyruvate Lactate↑*


122

Figure 4.10 Simplified pathway illustrating metabolites involved in purine metabolism in the rat urine samples on day 4 between the control and treatment group. The statistics are as follows: significance difference between the control group (n=8) and treatment group (n=8) is based on two tailed student t test (* p<0.05, ** p<0.01). 4.3.1.4 Determination of metabolites in rat urine samples by LC/MS

Urine is of interest in metabonomics studies because its collection is non-invasive

and it actually amplifies the circulating levels of metabolites by renal

concentration, which consequently provides a distinct representation of metabolic

response. The composition of rat urine is very complicated. It contains amino

acids, organic acids, amines, lipids, sugars, peptides and metabolic end products

such as sulfate and glucuronide conjugates. In this study, both the positive and

negative modes were carried out for the LC/MS analyis of rat urine samples.

Figure 4.11 demonstrates the quantitative and qualitative differences between the

control and treatment groups by comparing the total ion chromatogram (TIC) for

AMP

Adenosine

Inosine

Hypoxantine ↑*

GMP

Guanosine ↑

Xanthine ↑

Uric acid ↑** (urine)

Guanine ↑**

N1-methyladenosine ↑*


123

the positive ESI/MS of urine samples. Visual inspection of the data suggested that

the two groups may have different metabolic profiles.

Figure 4.11 Typical total ion chromatograms (TIC) of rat urine samples obtained from control group (black) and treatment group (red) using LC/MS (positive mode). As seen in Table 4.3 (Appendix), there were several peaks in the urine samples

from the treatment group that showed significant changes when compared to urine

samples from the control group when the positive mode was used. And there were

also significant changes when the negative mode was applied (Table 4.4,

Appendix). Table 4.2 shows the selected metabolites identified in rat urine

samples for both the control group and treatment group as measured by LC/MS

with positive and negative modes.

1

2

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16

Counts vs. Acquisition Time (min)

×107


124

Normalized peak intensitya (%) NO. Metabolite Retention Time

(min) m/z Control Ave ±SD Treated

Ave ±SD

Predose 1 Fumaric Acid 1.00 115 0.0227 0.0118 0.0263 0.0205 2 Creatinine 0.80 114 6.09 1.08 5.55 1.10 3 Lysine 0.62 147 0.038 0.018 0.050 0.017 4 Citric Acid 0.94 191 6.10 1.51 5.97 1.71 Day 1 1 Fumaric Acid 1.00 115 0.0247 0.0071 0.0247 0.0170 2 Creatinine* 0.80 114 3.88 0.56 6.49 2.51 3 Lysine 0.62 147 0.052 0.018 0.056 0.018 4 Citric Acid** 0.94 191 6.24 1.66 10.33 2.96 Day 2 1 Fumaric Acid* 1.00 115 0.0504 0.0316 0.0167 0.0165 2 Creatinine** 0.80 114 4.24 0.38 10.42 1.89 3 Lysine** 0.62 147 0.051 0.027 0.182 0.087 4 Citric Acid 0.94 191 6.68 1.04 5.39 3.09 Day 3 1 Fumaric Acid** 1.00 115 0.0345 0.0303 0.1690 0.0912 2 Creatinine** 0.80 114 4.02 1.16 10.10 3.02 3 Lysine 0.62 147 0.112 0.104 0.113 0.092 4 Citric Acid** 0.94 191 4.61 1.92 26.42 12.61 Day 4


125

1 Fumaric Acid 1.00 115 0.0392 0.0245 0.0413 0.0247 2 Creatinine** 0.80 114 4.68 1.24 9.55 2.00 3 Lysine* 0.62 147 0.074 0.056 0.158 0.075 4 Citric Acid 0.94 191 6.74 1.81 8.08 10.17

The relative intensity of metabolites is expressed with their normalized peak area. Values are represented as mean ± SD (standard deviation). The statistics are as follows: significance difference between the control group (n=8) and treatment group (n=8) is based on two tailed student t test (* p<0.05, ** p<0.01).

Table 4.2 Selected Metabolites identified in rat urine samples for both the control group and treatment group as measured by LC/MS (positive and negative mode).


126

Figure 4.12 and Figure 4.13 also show changes of selected metabolites as

measured by LC/MS with positive and negative modes on different days. It can be

seen from Figure 4.12 that the concentration of hypoxanthine, xanthine, uric acid,

guanine and N1-methyladenosine increased for the treatment group. Figure 4.13

illustrates that the fatty acids of the treatment group including adipic acid, lauric

acid, palmitoleic acid, linoleic acid, furmaric acid and citrate increased. In

addition, adipic acid, lauric acid, furmaric acid and citrate could recover from the

cholesterol lowering effect of berberine in rat liver.

0

0.1

0.2

0.3

0.4

0.5

0.6

0 1 2 3 4 5

day

inte

nsity

0

0.2

0.4

0.6

0.8

1

1.2

0 1 2 3 4 5

day

inte

nsity

(A) (B)

0

0.2

0.4

0.6

0.8

1

1.2

1.4

1.6

0 1 2 3 4 5

day

inte

nsity

0

0.05

0.1

0.15

0.2

0.25

0.3

0.35

0.4

0.45

0.5

0 1 2 3 4 5

day

inte

nsity

(C) (D)

0

0.5

1

1.5

2

2.5

3

0 1 2 3 4 5

day

Inte

nsity

(E)

Figure 4.12 Changes of selected metabolites as measured by LC/MS (positive mode) on different days (Pink colour: control group, n=8, blue colour: treatment


127

group, n=8). (A) hypoxanthine, (B) xanthine, (C) uric acid, (D) guanine and (E)

N1-methyladenosine.

0

5

10

15

20

25

30

0 1 2 3 4 5

day

Inte

nsity

0

0.01

0.02

0.03

0.04

0.05

0.06

0.07

0.08

0.09

0.1

0 1 2 3 4 5

day

Inte

nsity

(A) (B)

0

0.01

0.02

0.03

0.04

0.05

0.06

0.07

0.08

0.09

0 1 2 3 4 5

day

Inte

nsity

0

0.02

0.04

0.06

0.08

0.1

0.12

0.14

0.16

0 1 2 3 4 5

day

Inte

nsity

(C) (D)

0

0.05

0.1

0.15

0.2

0.25

0.3

0 1 2 3 4 5

day

inte

nsity

0

5

10

15

20

25

30

35

40

45

0 1 2 3 4 5

day

inte

nsity

(E) (F)

Figure 4.13 Changes of selected metabolites as measured by LC/MS (negative mode) on different days (Pink colour: control group, n=8, blue colour: treatment group, n=8). (A) adipic acid, (B) lauric acid, (C) palmitoleic acid, (D) linoleic acid, (E) furmaric acid and (F) citrate.


128

After data normalization, the multivariate statistical method PCA was carried out.

The PCA scores plot showed that the two groups were scattered into two different

regions (Figure 4.14) and all control samples were clearly separated from the

treated samples.

(A)

(B)

Figure 4.14 Multivariate data analysis of LC/MS data (positive mode) metabolic profiles. (A) PCA scores plot for data from all rat urine samples collected (Black triangle:

-10

0

10

-10 0 10t[1]

t[2]

-10

0

10

-10 0 10t[1]

t[2]


129

control at day 0, black circle control at day 1, Black diamond: control at day 2, Red triangle: treated at day 1, Blue triangle: treated at day 2) (R2X[1] = 0.177, R2X[2]= 0.128). (B) PCA scores plot for data from all rat urine samples collected (black triangle: control at day 3, black circle control at day 4, Red triangle: treated at day 3, Blue triangle: treated at day 4) (R2X[1] = 0.209, R2X[2]= 0.140). 4.3.2 Discussion

All the analyses performed (GC/MS of liver extracts, LC/MS of urine sample and

1H NMR of urine samples) revealed substantial berberine-induced changes and a

high degree of recovery by the end of the study.

One important aim of this study is to investigate the metabolic response revealed

by biomarkers after berberine adiministration by metabonomic approach. It has

been reported that liver might be the metabolic site for berberine. GC/MS data of

the lipid fraction of the liver extracts revealed that cholesterol, glucose, maltose

and fatty acids decreased after the berberine exposure (Figure 4.5B).

In this study, both LC/MS and 1H NMR were used to investigate the rat urinary

metabolic profile associated with administration of berberine. The results obtained

from both 1H NMR and LC/MS revealed the variations of certain metabolites in

urinary composition. Therefore, it allowed us to obtain information about the

mechanisms involved in metabolite excretion. For the current work, one of the

challenges with GC/MS, LC/MS and 1HNMR was the analysis of the target

metabolites in the presence of other peaks. Hence, peaks that had not been found

in the control group were not included in the treatment group for data

normalization purposes.

From the results mentioned above, it can be seen that the changes in metabolite

profiles observed by 1H NMR and LC/MS with pattern recognition tools were

similar. However, the markers which indicated the separations of treatment group

from control group observed were different. For example, betaine, taurine,


130

α-ketoglutarate, and glutamate were noticeable markers for 1H NMR, but due to

either poor ionization abilities in ESI or poor elution/retention by LC, these

compounds were not identified by LC/MS. On the other hand, markers such as

pantothetic acids, riboflavin, and xanthurenic acid that were prominent

compounds in LC/MS were not obvious for 1H NMR. Figure 4.11, Figure 4.12

and Figure 4.13 show the changes of selected metabolites.

In our earlier studies [52-53, 59], it has been found that using a multiple analytical

approach not only provides additional information, but also enhances the

confidence of the data obtained. In the current work, we used a targeted approach

for the urine LC/MS data process. A non-targeted approach was applied to the

processing of 1H NMR data. The results demonstrate that analysis of metabolites

in rat urine samples from two different groups using with LC/MS (targeted)

produced consistent and reliable results. Based on the targeted ions selected with

LC/MS, the PCA score plots from different days were consistent with that

obtained from using the non-targeted approach with 1H NMR where distinctive

clusters are observed. Since berberine was administered on day1 and day3, the

results on these two days could be interfered with the metabolites from berberine.

As a result, the data from pre-dose, day2 and day4 were chosen for comparison.

Both the data obtained from targeted and non-targeted processing on the pre-dose

samples showed no significant changes. The results of day2 and day4 obtained by

non-targeted processing of 1H NMR data showed the same changes with the

results obtained by targeted processing of LC/MS data. The close agreement of

results from 1H NMR and LC/MS on the change trends of selected metablotes in

Table 4.1 and Table 4.2 proved the reliability and accuracy of the targeted data

processing method. Finally, from Figure 4.9 and Figure 4.10 we can see that the


131

current approach allows us to study a number of metabolites present in

carbohydrate metabolism, amino acid metabolism, TCA cycle, fatty acid

metabolism, purine metabolism, metabolism and synthesis of major bile salts.

4.4 Conclusion

In this work, we develop a metabonomic approach based on GC/MS, LC/MS and

1H NMR to elucidate the metabolic patterns induced by berberine in rat. Results

obtained from GC/MS show that metabolites such as cholesterol, glucose, maltose

and fatty acids were affected. The use of normalization to a constant sum for the

LC/MS and 1H NMR with statistical analysis allowed for the identification of a

network of potential biomarkers corresponding to the exposure of berberine. The

close agreement of the results from targeted data processing of LC/MS and the

results from non-targeted data processing of 1H NMR has proved the reliability

and accuracy of the targeted data processing method. The proposed approach

combined with pattern recognition tools such as PCA provided a more

comprehensive picture of the metabolic changes induced by berberine in rat

compared to any single analytical technique alone. The result obtained was

information rich and provides a firm platform for future analysis.


132

References

[1] Y. Kaneda, M. Torii, T. Tanaka, M. Aikawa, Ann Trop Med Parasitol 85

(1991) 417.

[2] S. O. Gudima, L. V. Memelova, V. B. Borodulin, D. K. Pokholok, B. M.

Mednikov, O. N. Tolkachev, S. N. Kochetkov, Mol Biol (Mosk) 28 (1994) 1308.

[3] K. Ckless, J. L. Schlottfeldt, M. Pasqual, P. Moyna, J. A. Henriques, M.

Wajner, J Pharm Pharmacol 47 (1995) 1029.

[4] E. Kupeli, M. Kosar, E. Yesilada, K. Husnu, C. Baser, Life Sci. 72 (2002) 645.

[5] J. Sanchez-Chapula, Br J Pharmacol 117 (1996)1427.

[6] L. Ma, P. Xiao, B. Guo, J. Wu, F. Liang, S. Dong, Zhongguo Zhong Yao Za

Zhi 24 (1999) 238.

[7] N. Iizuka, K. Miyamoto, K. Okita, A. Tangoku, H. Hayashi, S. Yosino, T. Abe,

T. Morioka, S. Hazama, M. Oka, Cancer Lett. 148 (2000) 19.

[8] V. Kettmann, D. Kosfalova, S. Jantova, M. Cernakova, J. Drimal, Pharmazie

59 (2004) 548.

[9] W. H. Ko, X. Q. Yao, C. W. Lau, W. I. Law, Z. Y. Chen, W. Kwok, K. Ho, Y.

Huang, Eur J Pharmacol 399 (2000) 187.

[10] M. Cernakova, D. Kost, V. Kettmann, M. Plodova, J. Toth, J. Drimal, BMC

Complement Altern Med 2 (2002) 2.

[11] E. Yesilada, E. Kupeli, J Ethnopharmacol 79 (2002) 237.

[12] A. Vollekova, D. Kost, V. Kettmann, J Toth, Phytother Res 17 (2003) 834.

[13] L. Zheng, Z. Zhou, D. Tao, T. Lan, Sichuan Da Xue Xue Bao Yi Xue Ban 34

(2003) 452.

[14] T. S. Kim, B. Y. Kang, D. Cho, S. H. Kim, Immunology 109 (2003) 407.

[15] Q. L. Tran, Y. Tezuka, J. Y. Ueda, N. T. Nguyen, Y. Maruyama, K. Begum,


133

H. S. Kim, Y. Wataya, Q. K. Tran, S. Kadota, J Ethnopharmacol 86 (2003) 249.

[16] L. Rockova, M. Majekova, D. Kost, M .Stefek, Bioorg. Med. Chem. 12 (2004)

4709.

[17] W. H. Peng, C. R. Wu, C. S. Chen, C. F. Chen, Z. C. Leu, M. T. Hsieh, Life

Sci 75 (2004) 2451.

[18] P. L. Tsai, T. H. Tsai, Drug Metabolism and Disposition 32 (2004) 405.

[19] F. Zuo, N. Nakamura, T. Akao, M. Hattori, Drug Metabolism and Disposition

34 (2006) 2064

[20] J. K. Nicholson, J. C.Lindon, E. Holmes, Xenobiotica 1999, 29, 1181–1189.

[21] J. K. Nicholson, I. D. Wilson, Nat. Rev. Drug Discov. 2 (2003) 668.

[22] H. Kitano, Nat. Rev. Drug Discov.6 (2007) 202

[23] L. Wei, P. Liao, H. Wu, X. Li, F. Pei, W. Li, Y. Wu, Toxicol. Appl. Pharmacol.

227 (2008) 417.

[24] X. Zhang, H. Wu, P. Liao, X. Li, J. Ni, F. Pei, Food Chem. Toxicol. 44 (2006)

1006.

[25] T. M.Tsang, J. N. Haselden, E. Holmes, Neurochem. Res. 34 (2009) 1261

[26] S. Rezzi, F. P. J. Martin, D. Shanmuganayagam, R. J. Colman, J. K.

Nicholson, R. Weindruch, Experimental Gerontology 44 (2009) 356.

[27] P. Q. Liao, L. Wei, H. F. Wu, W. S. Li, Y. J. Wu, X. J. Li, J. Z. Ni, F. K. Pei,

Journal of rare earths 27 (2009) 280.

[28] A. Waldram, E. Holmes, Y. L. Wang, M. Rantalainen, I. D. Wilson, K. M.

Tuohy, A. L. McCartney, G. R. Gibson, J. K. Nicholson, J. Proteome Res. 8 (2009)

2361.

[29] F. P. J. Martin, S. Rezzi, I. Montoliu, D. Philippe, L. Tornier, A. Messlik, G.

Hoelzlwimmer, P. Baur, L. Quintanilla-Fend, G. Loh, M. Blaut, S. Blum, S.


134

Kochhar, D. Haller, J. Proteome Res. 8 (2009) 2376.

[30] I. Montoliu, F. P. J. Martin, S. Collino, S. Rezzi, S. Kochhar, J. Proteome Res.

8 (2009) 2397

[31] B. J. Blaise, J. Giacomotto, M. N. Triba, P. Toulhoat, M. Piotto, L. Emsley, L.

Segalat, M. E. Dumas, B. Elena, J. Proteome Res. 8 (2009) 2542.

[32] K. E. Price, C. K. Larive, C. E. Lunte, Metabolomics 5 (2009) 219.

[33] H. C. Gao, Q. Lu, X. Liu, H. Cong, L. C. Zhao, H. M. Wang, D. H. Lin,

Cancer Sci. 100 (2009) 782.

[34]Q. H. He, X. F. Kong, G. Y. Wu, P. P. Ren, H. R. Tang, F. H. Hao, R. L. Huang,

T. J. Li, B. Tan, P. Li, Z. R. Tang, Y. L. Yin, Y. N. Wu, Amino Acids 37 (2009)

199.

[35] C. N. Xiao, F. H. Hao, X. R. Qin, Y. L. Wang, H. R. Tang, Analyst 134 (2009)

916.

[36] Q. Lu, Y. H. Huang, H. Cong, X. Liu, H. C. Gao, D. H. Lin, H. M. Wang,

Chin. J. Anal. Chem. 37 (2009) 194.

[37] Z. M. Wu, Z. Q. Huang, R. Lehmann, C. X. Zhao, G. W. Xu,

Chromatographia 69 (2009) S23.

[38] T. Uehara, A. Yokoi, K. Aoshima, S. Tanaka, T. Kadowaki, M Tanaka, Y. Oda,

Anal. Chem. 81 (2009) 3836.

[39] K. Shimbo, T. Oonuki, A. Yahashi, K. Hirayama, H. Miyano, Rapid Commun.

Mass Spectrom. 23 (2009) 1483.

[40] R. S. Plumb, P. D. Rainville, W. B. Potts, K. A. Johnson, E. Gika, I. D.

Wilson, J. Proteome Res. 8 (2009) 2495.

[41] H. H. Yoo, H. J. Yoon, H. J. Shin, S. H. Lee, H. R. Yoon, Biomolecules &

Therapeutics 17 (2009) 181.


135

[42] M. P. Hodson, G. J. Dear, J. L. Griffin, J. N. Haselden, Metabolomics

5 (2009) 166.

[43] M. Z. Ding, J. S. Cheng, W. H. Xiao, B. Qiao, Y. J. Yuan,

Metabolomics 5 (2009) 229.

[44] N. R. Kitteringham, R. E. Jenkins, C. S. Lane, V. L. Elliott, B. K. Park, J.

Chromatogr., B: Anal. Technol. Biomed. Life Sci. 877 (2009) 1229.

[45] G. X. Xie, M. Ye, Y. G. Wang, Y. Ni, M. M. Su, H. Huang, M. F. Qiu, A. H.

Zhao, X. J. Zheng, T. L. Chen, W. Jia, J. Agric. Food Chem. 57 (2009) 3046.

[46] M. Sieber, D. Hoffmann, M. Adler, V. S. Vaidya, M. Clement, J. V. Bonventre,

N. Zidek, E. Rached, A. Amberg, J. J. Callanan, W. Dekant, A. Mally, Toxicol.

Sci. 109 (2009) 336.

[47] J. R. McKelvie, J. Yuk, Y. P. Xu, A. J. Simpson, M. J. Simpson,


[48] T. K. Karakach, E. C. Huenupi, E. C. Soo, J. A.Walter, L. O. B. Afonso,


[49] J. C. Sun, L. K. Schnackenberg, R. D. Holland, T. C. Schmitt, G. H. Cantor, Y.

R. Dragan, R. D. Beger, J. Chromatogr., B: Anal. Technol. Biomed. Life Sci. 877

(2009) 105.

[50] E. C. Y. Chan, P. K. Koh, M. Mal, P. Y. Cheah, K. W. Eu, A. Backshall, R.

Cavill, J. K. Nicholson, H. C. Keun, J. Proteome Res. 8 (2009) 352.

[51] J. C. Sun, L. K. Schnackenberg, R. D. Holland, T. C. Schmitt, G. H. Cantor, Y.

R. Dragan, R. D. Beger, J. Chromatogr., B: Anal. Technol. Biomed. Life Sci. 871

(2008) 328.

[52] W. S. Law, P. Y. Huang, E. S. Ong, C. N. Ong, S. F. Y. Li, K. K. Psikanti, E. C.

Y. Chan, Rapid Commun. Mass Spectrom. 22 (2008) 2436.


136

[53] W. S. Law, P. Y. Huang, E. S. Ong, S. K. Sethi, S. Saw, C. N. Ong, S. F. Y. Li,

J. Proteome Res. 8 (2009) 1828.

[54] C. M. Slupsky, K.N. Rankin, J. Wagner, H. Fu, D. Chang, A.M. Weljie, E.J.

Saude, B. Lix, D.J. Adamko, S. Shah, R. Greiner, B.D. Sykes, T.J. Marrie, Anal.

Chem. 79 (2007) 6995.

[55] S. Y. Um, M. W. Chung, K. B. Kim, S. H. Kim, J. S. Oh, H. Y. Oh, H. J. Lee,

K. H. Choi, Anal. Chem. 81 (2009) 4734.

[56] C. M. Slupsky, A. Cheypesh, D. V. Chao, H. Fu, K. N. Rankin, T. J. Marrie, P.

Lacy, J. Proteome Res. 8 (2009) 3029.

[57] M. E. Bollard, H. C. Keun, O. Beckonert, T. M. D. Ebbels, H. Antti, A. W.

Nicholls, J. P. Shockcor, G. H. Cantor, G. Stevens, J. C. Lindon, E. Holmes, J. K.

Nicholson, Toxicol. Appl. Pharmacol. 204 (2005) 135.

[58] F. P. J. Martin, M. Dumas, Y. Wang, C. Legido-Quigley, I. K. S. Yap, H. Tang,

S. Zirah, G. M. Murphy, O. Cloarec, J. C. Lindon, N. Sprenger, L. B. Fay, S.

Kochhar, P. V. Bladeren, E. Holmes, J. K. Nicholson, Mol. Sys. Biol. 3 (2007) 1.

[59] E. S.Ong, C. F. Chor, L. Zou, C. N. Ong, Mol. Biosyst. 5 (2009) 288.

[60] Human Metalome Database. http://www.hmdb.ca/ (last accessed 15 Mar

2009)

[61] Biological Magnetic Resonance Data Bank.

http://www.bmrb.wisc.edu/metabolomics/ (last accessed 15 Mar 2009)


137

Appendix for Chapter 4

Metabolites identified in rat urine samples for both the control group and treatment group as measured by LCMS, pre-dose (positive mode)



Ave ±SD

1 Stearamide 11.44 256 0.022 0.007 0.015 0.009 2 Oleamide* 11.64 282 0.038 0.015 0.021 0.006 3 Serine 0.79 118 0.825 0.436 0.949 0.289 4 Homocysteine 0.85 132 0.057 0.035 0.125 0.077 5 Lysine 0.62 147 0.038 0.018 0.059 0.029 6 Phenylalanine 0.94 166 5.957 0.913 5.154 0.445 7 Hypoxanthine 1.00 137 0.020 0.030 0.029 0.039 8 Xanthine 0.76 151 0.493 0.310 0.235 0.061 9 Uric Acid* 0.88 169 0.098 0.048 0.186 0.060 10 Creatinine 0.80 114 6.091 1.082 5.559 1.102 11 Kynurenic Acid 3.41 190 8.260 0.656 6.962 1.792 12 Xanthurenic Acid 4.64 206 0.028 0.008 0.029 0.011 13 Riboflavin 4.33 377 11.562 2.003 9.539 2.273 14 Guanine 0.78 152 0.157 0.084 0.104 0.082 15 Adenosine 1.01 268 2.180 0.533 1.896 0.232 16 N1-Methyladenosine 0.89 282 2.240 0.278 1.969 0.219 17 Guanosine 0.99 284 0.007 0.003 0.009 0.002 18 N4-Acetylcytidine 1.06 286 0.134 0.144 0.199 0.235 19 N2-Methylguanosine* 1.25 298 0.094 0.013 0.076 0.010

The relative intensity of metabolites is expressed with their normalized peak area. Values are represented as mean ± SD (standard deviation). The statistics are as follows: significance difference between the control group


138

(n=8) and treatment group (n=8) is based on two tailed student t test (* p<0.05, ** p<0.01). (A)


139

Metabolites identified in rat urine samples for both the control group and treatment group as measured by LCMS, day 1 (positive mode)



Ave ±SD

1 Stearamide 11.44 256 0.028 0.008 0.046 0.023 2 Oleamide 11.64 282 0.030 0.009 0.074 0.074 3 Serine 0.79 118 0.995 0.197 1.093 0.335 4 Homocysteine** 0.85 132 0.155 0.051 0.289 0.064 5 Lysine 0.62 147 0.052 0.018 0.056 0.018 6 Phenylalanine** 0.94 166 4.310 0.680 5.753 1.099 7 Hypoxanthine 1.00 137 0.013 0.012 0.028 0.023 8 Xanthine 0.76 151 0.181 0.096 0.173 0.037 9 Uric Acid** 0.88 169 0.203 0.079 0.327 0.073 10 Creatinine* 0.80 114 3.889 0.567 6.492 2.511 11 Kynurenic Acid 3.41 190 10.408 1.414 10.904 2.797 12 Xanthurenic Acid** 4.64 206 0.019 0.006 0.006 0.004 13 Riboflavin** 4.33 377 6.118 1.538 2.159 1.073 14 Guanine 0.78 152 0.095 0.082 0.106 0.048 15 Adenosine* 1.01 268 1.717 0.328 1.369 0.171 16 N1-Methyladenosine* 0.89 282 1.557 0.216 1.823 0.269 17 Guanosine** 0.99 284 0.009 0.004 0.028 0.012 18 N4-Acetylcytidine 1.06 286 0.171 0.140 0.200 0.133 19 N2-Methylguanosine* 1.25 298 0.074 0.021 0.053 0.006

The relative intensity of metabolites is expressed with their normalized peak area.Values are represented as mean ± SD (standard deviation). The statistics are as follows: significance difference between the control group


140

(n=8) and treatment group (n=8) is based on two tailed student t test (* p<0.05, ** p<0.01). (B)


141




Ave ±SD

1 Stearamide 11.44 256 0.035 0.009 0.084 0.085 2 Oleamide 11.64 282 0.086 0.126 0.499 1.226 3 Serine 0.79 118 1.015 0.341 2.803 2.393 4 Homocysteine 0.85 132 0.143 0.080 0.743 1.205 5 Lysine** 0.62 147 0.051 0.027 0.182 0.087 6 Phenylalanine** 0.94 166 4.139 0.418 5.815 0.933 7 Hypoxanthine 1.00 137 0.434 1.203 0.099 0.175 8 Xanthine* 0.76 151 0.149 0.045 0.311 0.154 9 Uric Acid 0.88 169 0.137 0.037 0.647 0.740 10 Creatinine** 0.80 114 4.240 0.386 10.426 1.892 11 Kynurenic Acid 3.41 190 10.826 1.401 9.627 3.712 12 Xanthurenic Acid 4.64 206 0.019 0.008 0.012 0.009 13 Riboflavin 4.33 377 4.448 1.871 2.985 1.279 14 Guanine 0.78 152 0.125 0.069 0.145 0.063 15 Adenosine 1.01 268 1.945 0.102 1.845 0.429 16 N1-Methyladenosine** 0.89 282 1.390 0.082 1.944 0.326 17 Guanosine 0.99 284 0.017 0.008 0.011 0.007 18 N4-Acetylcytidine** 1.06 286 0.291 0.063 0.132 0.041 19 N2-Methylguanosine 1.25 298 0.067 0.025 0.091 0.032



142

(n=8) and treatment group (n=8) is based on two tailed student t test (* p<0.05, ** p<0.01). (C)


143




Ave ±SD

1 Stearamide** 11.44 256 0.046 0.023 0.090 0.032 2 Oleamide 11.64 282 0.069 0.055 0.090 0.023 3 Serine 0.79 118 1.199 0.739 3.702 3.552 4 Homocysteine 0.85 132 0.127 0.081 29.233 42.377 5 Lysine 0.62 147 0.112 0.104 0.113 0.092 6 Phenylalanine 0.94 166 4.072 1.012 5.837 2.287 7 Hypoxanthine** 1.00 137 0.011 0.014 0.331 0.240 8 Xanthine 0.76 151 0.130 0.029 0.329 0.480 9 Uric Acid** 0.88 169 0.173 0.073 0.528 0.318 10 Creatinine** 0.80 114 4.025 1.167 10.101 3.025 11 Kynurenic Acid 3.41 190 10.496 1.859 11.865 4.267 12 Xanthurenic Acid 4.64 206 0.019 0.010 0.013 0.011 13 Riboflavin** 4.33 377 4.885 1.557 1.920 1.099 14 Guanine* 0.78 152 0.102 0.071 0.246 0.128 15 Adenosine 1.01 268 1.677 0.465 1.890 1.014 16 N1-Methyladenosine 0.89 282 1.466 0.293 1.642 0.558 17 Guanosine* 0.99 284 0.014 0.010 0.029 0.011 18 N4-Acetylcytidine 1.06 286 0.290 0.084 0.248 0.194 19 N2-Methylguanosine 1.25 298 0.062 0.015 0.044 0.020



144

(n=8) and treatment group (n=8) is based on two tailed student t test (* p<0.05, ** p<0.01). (D)


145




Ave ±SD

1 Stearamide* 11.44 256 0.026 0.010 0.049 0.028 2 Oleamide** 11.64 282 0.027 0.005 0.061 0.025 3 Serine** 0.79 118 1.155 0.596 3.282 1.306 4 Homocysteine* 0.85 132 0.172 0.176 10.492 12.370 5 Lysine* 0.62 147 0.074 0.056 0.158 0.075 6 Phenylalanine** 0.94 166 3.831 0.982 5.523 1.247 7 Hypoxanthine* 1.00 137 0.010 0.008 0.130 0.135 8 Xanthine 0.76 151 0.130 0.020 0.430 0.617 9 Uric Acid** 0.88 169 0.182 0.065 0.553 0.222 10 Creatinine** 0.80 114 4.687 1.242 9.556 2.006 11 Kynurenic Acid 3.41 190 7.665 1.432 7.960 2.975 12 Xanthurenic Acid 4.64 206 0.019 0.012 0.014 0.013 13 Riboflavin 4.33 377 4.109 0.888 2.564 2.549 14 Guanine** 0.78 152 0.109 0.069 0.352 0.120 15 Adenosine 1.01 268 1.934 0.519 2.258 0.730 16 N1-Methyladenosine* 0.89 282 1.327 0.183 1.761 0.427 17 Guanosine 0.99 284 0.011 0.007 0.019 0.013 18 N4-Acetylcytidine 1.06 286 0.236 0.184 0.117 0.045 19 N2-Methylguanosine 1.25 298 0.076 0.023 0.081 0.033



146

(n=8) and treatment group (n=8) is based on two tailed student t test (* p<0.05, ** p<0.01). (E)

Table 4.3 Metabolites identified in rat urine samples for both the control group and treatment group as measured by LC-MS (positive mode). (A) Pre-dose, (B) Day 1, (C) Day 2 (D) Day 3 and (E) Day 4.


147

Metabolites identified in rat urine samples for both the control group and treatment group as measured by LCMS, pre-dose (negative mode) Normalized peak intensitya (%) NO. Metabolite Retention Time (min) m/z Control Ave ±SD Treated Ave ±SD

1 Fumaric Acid 1.00 115 0.0227 0.0118 0.0312 0.0228 2 Adipic Acid 2.34 145 0.4166 0.1802 0.4438 0.4703 3 Lauric Acid 9.92 199 0.0065 0.0043 0.0083 0.0058 4 Mystric Acid 9.95 227 0.0020 0.0030 0.0049 0.0024 5 Palmitoleic Acid 11.05 253 0.0035 0.0016 0.0109 0.0104 6 γ-Linoleic Acid 10.79 277 0.0011 0.0007 0.0025 0.0034 7 Linoleic Acid 11.20 279 0.0132 0.0044 0.0155 0.0090 8 Oleic Acid* 6.31 281 0.0236 0.0116 0.0735 0.0408 9 Decosahexaenoic Acid (DHA) 11.01 327 0.0024 0.0017 0.0024 0.0021 10 Deoxycholic Acid (DCA) 8.79 391 0.0045 0.0071 0.0042 0.0029 11 Cholic Acid (CA) 7.51 407 0.3680 0.3719 0.2860 0.3231 12 Glycodeoxycholic Acid(GDCA) 7.80 448 0.0014 0.0011 0.0022 0.0020 13 Glycocholic Acid (GCA) 6.82 464 0.0281 0.0435 0.0116 0.0079 14 Tauroursdeoxycholic Acid (TUDCA) 6.16 498 0.0027 0.0021 0.0040 0.0033 15 Taurochenodeoxycholic Acid (TCDCA)* 6.62 498 0.0026 0.0014 0.0009 0.0006 16 Taurodeoxycholic Acid (TDCA) 6.72 498 0.0027 0.0021 0.0021 0.0019 17 Taurocholic Acid (TCA) 6.21 514 0.0407 0.0208 0.0402 0.0283 18 Phenol Sulphate 2.09 173 63.0481 10.4543 62.2522 10.6504 19 Resorcinol Sulphate 1.81 189 25.5235 7.9718 26.0519 7.9606 20 Citric Acid 0.94 191 6.1097 1.5170 5.9763 1.7126 21 Sebacic Acid 5.92 201 4.3768 1.1372 4.7756 1.0962


(A)


148

Metabolites identified in rat urine samples for both the control group and treatment group as measured by LCMS, day 1 (negative mode) Normalized peak intensitya (%) NO. Metabolite Retention Time (min) m/z Control Ave ±SD Treated Ave ±SD

1 Fumaric Acid 1.00 115 0.0247 0.0071 0.0247 0.0170 2 Adipic Acid** 2.34 145 0.9176 0.2861 2.8841 1.6902 3 Lauric Acid 9.92 199 0.0178 0.0161 0.0147 0.0080 4 Mystric Acid 9.95 227 0.0042 0.0032 0.0102 0.0102 5 Palmitoleic Acid 11.05 253 0.0065 0.0034 0.0114 0.0070 6 γ-Linoleic Acid 10.79 277 0.0025 0.0014 0.0041 0.0045 7 Linoleic Acid 11.20 279 0.0163 0.0090 0.0339 0.0291 8 Oleic Acid 6.31 281 0.1290 0.0692 0.1034 0.0930 9 Decosahexaenoic Acid (DHA)** 11.01 327 0.0024 0.0017 0.0083 0.0043 10 Deoxycholic Acid (DCA)** 8.79 391 0.0034 0.0028 0.0092 0.0034 11 Cholic Acid (CA)** 7.51 407 0.6888 0.2848 0.0987 0.0636 12 Glycodeoxycholic Acid(GDCA) 7.80 448 0.0023 0.0021 0.0023 0.0030 13 Glycocholic Acid (GCA)* 6.82 464 0.0289 0.0237 0.0075 0.0073 14 Tauroursdeoxycholic Acid (TUDCA) 6.16 498 0.0088 0.0068 0.0109 0.0089 15 Taurochenodeoxycholic Acid (TCDCA) 6.62 498 0.0029 0.0026 0.0044 0.0036 16 Taurodeoxycholic Acid (TDCA) 6.72 498 0.0021 0.0013 0.0045 0.0068 17 Taurocholic Acid (TCA) 6.21 514 0.0114 0.0116 0.0096 0.0062 18 Phenol Sulphate 2.09 173 56.0683 7.3503 46.6392 11.4806 19 Resorcinol Sulphate 1.81 189 29.1860 6.7124 35.4629 8.3315 20 Citric Acid** 0.94 191 6.2478 1.6675 10.3336 2.9629 21 Sebacic Acid 5.92 201 6.6282 2.9841 4.3224 2.0137


(B)


149


1 Fumaric Acid* 1.00 115 0.0504 0.0316 0.0167 0.0165 2 Adipic Acid* 2.34 145 0.9271 0.1963 3.3375 3.0076 3 Lauric Acid 9.92 199 0.0169 0.0189 0.0393 0.0286 4 Mystric Acid 9.95 227 0.0061 0.0025 0.0096 0.0082 5 Palmitoleic Acid 11.05 253 0.0205 0.0367 0.0361 0.0483 6 γ-Linoleic Acid 10.79 277 0.0030 0.0018 0.0023 0.0018 7 Linoleic Acid* 11.20 279 0.0121 0.0079 0.0360 0.0269 8 Oleic Acid 6.31 281 0.1524 0.0745 0.1445 0.1306 9 Decosahexaenoic Acid (DHA)* 11.01 327 0.0032 0.0033 0.0096 0.0075 10 Deoxycholic Acid (DCA)* 8.79 391 0.0034 0.0035 0.0151 0.0133 11 Cholic Acid (CA) 7.51 407 0.6240 0.3219 0.6345 0.7700 12 Glycodeoxycholic Acid(GDCA) 7.80 448 0.0013 0.0015 0.0021 0.0025 13 Glycocholic Acid (GCA) 6.82 464 0.0157 0.0086 0.0181 0.0189 14 Tauroursdeoxycholic Acid (TUDCA) 6.16 498 0.0060 0.0034 0.0096 0.0088 15 Taurochenodeoxycholic Acid (TCDCA) 6.62 498 0.0041 0.0035 0.0089 0.0103 16 Taurodeoxycholic Acid (TDCA) 6.72 498 0.0036 0.0022 0.0080 0.0070 17 Taurocholic Acid (TCA) 6.21 514 0.0098 0.0068 0.0230 0.0178 18 Phenol Sulphate 2.09 173 49.0436 9.2321 46.0537 13.4921 19 Resorcinol Sulphate* 1.81 189 35.3092 8.2934 25.8835 8.2143 20 Citric Acid 0.94 191 6.9364 1.2124 4.7452 3.3889 21 Sebacic Acid 5.92 201 6.8511 2.1696 18.9666 16.9646


(C)


150


1 Fumaric Acid** 1.00 115 0.0345 0.0303 0.1690 0.0912 2 Adipic Acid** 2.34 145 1.1230 0.3561 13.5351 10.9277 3 Lauric Acid** 9.92 199 0.0215 0.0106 0.0605 0.0280 4 Mystric Acid 9.95 227 0.0059 0.0041 0.0160 0.0137 5 Palmitoleic Acid* 11.05 253 0.0065 0.0040 0.0349 0.0293 6 γ-Linoleic Acid 10.79 277 0.0018 0.0008 0.0039 0.0036 7 Linoleic Acid 11.20 279 0.0219 0.0177 0.0463 0.0453 8 Oleic Acid** 6.31 281 0.0178 0.0132 0.2163 0.1773 9 Decosahexaenoic Acid (DHA) 11.01 327 0.0052 0.0041 0.0265 0.0338 10 Deoxycholic Acid (DCA) 8.79 391 0.0052 0.0050 0.0607 0.0903 11 Cholic Acid (CA) 7.51 407 0.3782 0.1937 2.8123 5.3944 12 Glycodeoxycholic Acid(GDCA) 7.80 448 0.0034 0.0050 0.0070 0.0136 13 Glycocholic Acid (GCA) 6.82 464 0.0090 0.0076 0.0127 0.0114 14 Tauroursdeoxycholic Acid (TUDCA) 6.16 498 0.0069 0.0063 0.0086 0.0040 15 Taurochenodeoxycholic Acid (TCDCA) 6.62 498 0.0046 0.0043 0.0122 0.0125 16 Taurodeoxycholic Acid (TDCA)* 6.72 498 0.0023 0.0019 0.0071 0.0048 17 Taurocholic Acid (TCA) 6.21 514 0.0087 0.0055 0.1294 0.2408 18 Phenol Sulphate** 2.09 173 51.8815 9.2533 28.6927 12.3776 19 Resorcinol Sulphate** 1.81 189 33.9433 8.9576 19.1343 7.4234 20 Citric Acid** 0.94 191 4.6160 1.9202 26.4212 12.6128 21 Sebacic Acid 5.92 201 7.9030 2.9628 8.5933 9.4349


(D)


151

Metabolites identified in rat urine samples for both the control group and treatment group as measured by LCMS, day 4 (negative mode) Normalized peak intensitya (%)

NO. Metabolite Retention Time (min) m/z Control Ave ±SD Treated Ave ±SD 1 Fumaric Acid 1.00 115 0.0392 0.0245 0.0413 0.0247 2 Adipic Acid* 2.34 145 0.9119 0.3779 4.1145 3.6067 3 Lauric Acid 9.92 199 0.0189 0.0150 0.0269 0.0209 4 Mystric Acid* 9.95 227 0.0043 0.0019 0.0141 0.0114 5 Palmitoleic Acid* 11.05 253 0.0039 0.0013 0.0287 0.0282 6 γ-Linoleic Acid 10.79 277 0.0032 0.0030 0.0060 0.0055 7 Linoleic Acid 11.20 279 0.0319 0.0246 0.0770 0.0641 8 Oleic Acid 6.31 281 0.1229 0.1030 0.4016 0.7211 9 Decosahexaenoic Acid (DHA) 11.01 327 0.0043 0.0035 0.0067 0.0045 10 Deoxycholic Acid (DCA)* 8.79 391 0.0035 0.0017 0.0099 0.0075 11 Cholic Acid (CA) 7.51 407 0.3641 0.3192 0.8373 0.6912 12 Glycodeoxycholic Acid(GDCA) 7.80 448 0.0025 0.0032 0.0005 0.0006 13 Glycocholic Acid (GCA) 6.82 464 0.0111 0.0103 0.0238 0.0304 14 Tauroursdeoxycholic Acid (TUDCA) 6.16 498 0.0056 0.0045 0.0061 0.0068 15 Taurochenodeoxycholic Acid (TCDCA) 6.62 498 0.0042 0.0021 0.0172 0.0351 16 Taurodeoxycholic Acid (TDCA) 6.72 498 0.0029 0.0013 0.0189 0.0361 17 Taurocholic Acid (TCA)* 6.21 514 0.0058 0.0026 0.0364 0.0342 18 Phenol Sulphate 2.09 173 48.9471 5.3559 41.2996 12.3225 19 Resorcinol Sulphate 1.81 189 34.7741 3.1317 29.4535 11.3337 20 Citric Acid 0.94 191 6.7468 1.8129 8.0855 10.1793 21 Sebacic Acid* 5.92 201 7.9918 4.8422 15.4946 8.3890


(E)


152

Table 4.4 Metabolites identified in rat urine samples for both the control group and treatment group as measured by LC-MS (negative mode). (A) Pre-dose, (B) Day 1, (C) Day 2, (D) Day 3 and (E) Day 4.


153

Metabolites identified in rat urine samples for both control and treatment group as measured by 1H NMR, pre-dose Normalized peak intensitya NO. Metabolite Chemical Shift (ppm) Multiplicity Control Ave ±SD Treated Ave ±SD

1 Tryptamine 7.69 d 0.26 0.03 0.31 0.20 Tryptamine 7.32 d 0.31 0.03 0.32 0.08 Tryptamine * 7.24 d 0.17 0.02 0.21 0.04 2 Hippurate 7.62 t 0.50 0.15 0.39 0.27 Hippurate 7.55 t 1.73 0.55 1.63 0.74 3 Phenylacetylglycine 7.40 m 0.83 0.31 0.52 0.30 Phenylacetylglycine 7.36 m 1.36 0.47 1.08 0.24 4 Tyrosine 7.16 m 0.28 0.10 0.23 0.07 Tyrosine 6.87 m 0.34 0.10 0.26 0.04 5 Fumurate 6.53 s 0.18 0.04 0.17 0.04 6 Creatine 3.91 s 0.24 0.03 0.25 0.03 7 Creatinine 3.08 s 0.15 0.03 0.15 0.07 8 Betaine 3.88 s 0.42 0.09 0.45 0.07 9 Glucose 3.83 m 0.33 0.03 0.36 0.06 Glucose 3.66 m 0.56 0.04 0.51 0.35 Glucose 3.51 m 0.39 0.05 0.39 0.03

10 Methionine 3.76 m 0.54 0.27 0.73 0.26 Methionine 2.14 m 0.51 0.04 0.57 0.07 Methionine 2.13 s 0.42 0.04 0.46 0.06

11 Lysine 3.74 t 0.41 0.21 0.38 0.26 12 Ethanol 3.64 q 0.73 0.12 0.66 0.30 13 Glycine 3.52 s 0.37 0.07 0.37 0.06 14 Taurine 3.41 t 0.33 0.03 1.02 1.40 Taurine 3.18 t 0.25 0.04 0.24 0.05

15 Hypotaurine 3.35 t 0.19 0.06 0.22 0.08


154

16 Trimethylamine –N-oxide (TMAO) 3.25 s 0.52 0.10 1.00 1.21 17 Creatinine/Creatine 3.01 s 4.58 0.93 4.22 1.44 18 α-Ketoglutarate 3.00 t 2.47 0.39 2.22 1.05 α-Ketoglutarate 2.43 t 2.22 0.63 2.55 0.96

19 Dimethylglycine 2.93 s 1.55 0.39 1.24 0.65 20 Trimethylamine (TMA) 2.85 s 0.25 0.11 0.20 0.09 21 Aspartic Acid 2.84 m 0.40 0.22 0.29 0.05 22 Citrate 2.62 d 0.28 0.04 0.32 0.04 Citrate 2.60 d 0.24 0.06 0.23 0.14 Citrate 2.60 d 0.32 0.05 0.34 0.06 Citrate 2.50 d 0.37 0.07 0.32 0.08

23 Glutamine 2.44 m 4.17 1.80 3.74 2.62 24 Succinate 2.42 s 0.77 0.17 1.76 2.50 25 Pyruvate 2.39 s 0.52 0.12 0.52 0.08 26 Glutamate 2.36 m 0.44 0.06 0.43 0.07 Glutamate 2.33 m 0.60 0.07 0.64 0.15

27 Acetoacetate 2.27 s 0.37 0.06 0.37 0.08 28 Arginine 1.63 m 0.30 0.03 0.32 0.05 29 Alanine 1.49 d 0.42 0.07 0.45 0.08 Alanine 1.47 d 0.44 0.06 0.42 0.11

30 Lactate* 1.34 d 0.42 0.08 0.62 0.15 Lactate 1.32 d 0.38 0.05 0.48 0.14

31 3-D-Hydroxybutyrate 1.20 d 0.48 0.08 0.48 0.08 32 Isobutyrate 1.11 d 0.15 0.04 0.15 0.05 33 Valine 1.04 d 0.14 0.03 0.15 0.04 Valine 0.98 d 0.14 0.02 0.14 0.03

34 Leucine 1.00 m 0.14 0.02 0.15 0.03 Leucine 0.94 m 0.43 0.22 0.70 0.49


155


(A)


156

Metabolites identified in rat urine samples for both control and treatment group as measured by 1H NMR, day 1 Normalized peak intensitya NO. Metabolite Chemical Shift (ppm) Multiplicity Control Ave ±SD Treated Ave ±SD

1 Tryptamine* 7.69 d 0.15 0.07 0.04 0.09 Tryptamine* 7.32 d 0.30 0.07 0.20 0.10 Tryptamine 7.24 d 0.23 0.06 0.13 0.15 2 Hippurate 7.62 t 0.43 0.10 0.37 0.08 Hippurate 7.55 t 1.51 0.36 1.23 0.27 3 Phenylacetylglycine 7.40 m 0.46 0.11 0.39 0.17 Phenylacetylglycine * 7.36 m 0.74 0.21 0.47 0.25 4 Tyrosine 7.16 m 0.30 0.08 0.18 0.12 Tyrosine ** 6.87 m 0.41 0.08 0.18 0.10 5 Fumurate 6.53 s 0.26 0.03 0.23 0.09 6 Creatine* 3.91 s 0.47 0.10 1.37 0.80 7 Creatinine 3.08 s 0.21 0.03 0.17 0.06 8 Betaine** 3.88 s 0.58 0.24 1.94 0.90 9 Glucose** 3.83 m 0.58 0.09 2.14 1.11 Glucose 3.66 m 1.02 0.09 0.91 0.31 Glucose ** 3.51 m 0.63 0.09 2.17 1.07

10 Methionine** 3.76 m 0.93 0.10 1.52 0.26 Methionine 2.14 m 0.46 0.04 0.45 0.15 Methionine 2.13 s 0.38 0.03 0.32 0.12

11 Lysine 3.74 t 0.65 0.27 1.79 1.48 12 Ethanol 3.64 q 1.26 0.12 1.12 0.40 13 Glycine** 3.52 s 0.61 0.07 1.59 0.71 14 Taurine** 3.41 t 2.43 1.47 6.05 2.01 Taurine * 3.18 t 0.35 0.04 0.23 0.10

15 Hypotaurine 3.35 t 0.43 0.12 0.37 0.07


157

16 Trimethylamine –N-oxide (TMAO)** 3.25 s 2.55 0.84 5.79 1.79 17 Creatinine/Creatine** 3.01 s 4.13 1.30 1.23 0.76 18 α-Ketoglutarate** 3.00 t 2.41 0.73 0.83 0.42 α-Ketoglutarate ** 2.43 t 1.95 0.58 0.70 0.38

19 Dimethylglycine** 2.93 s 1.54 0.39 0.30 0.23 20 Trimethylamine (TMA)** 2.85 s 0.29 0.04 0.14 0.07 21 Aspartic Acid** 2.84 m 0.49 0.09 0.22 0.11 22 Citrate 2.62 d 0.28 0.02 0.28 0.11 Citrate 2.60 d 0.20 0.13 0.15 0.15 Citrate 2.60 d 0.34 0.04 0.28 0.08 Citrate 2.50 d 0.41 0.09 0.49 0.38

23 Glutamine** 2.44 m 3.61 1.14 1.17 0.71 24 Succinate** 2.42 s 0.89 0.07 0.46 0.31 25 Pyruvate 2.39 s 0.51 0.10 0.43 0.16 26 Glutamate 2.36 m 0.23 0.20 0.28 0.14 Glutamate ** 2.33 m 0.56 0.04 0.36 0.14

27 Acetoacetate 2.27 s 0.30 0.04 0.24 0.10 28 Arginine 1.63 m 0.25 0.04 0.20 0.08 29 Alanine 1.49 d 0.27 0.04 0.23 0.07 Alanine 1.47 d 0.28 0.04 0.24 0.08

30 Lactate 1.34 d 0.37 0.03 0.40 0.10 Lactate 1.32 d 0.33 0.03 0.38 0.10


34 Leucine 1.00 m 0.10 0.03 0.09 0.04 Leucine** 0.94 m 0.32 0.03 0.23 0.08


158


(B)


159

Metabolites identified in rat urine samples for both control and treatment groups as measured by 1H NMR, day 2 Normalized peak intensitya NO. Metabolite Chemical Shift (ppm) Multiplicity Control Ave ±SD Treated Ave ±SD

1 Tryptamine 7.69 d 0.12 0.06 0.16 0.07 Tryptamine 7.32 d 0.23 0.07 0.17 0.08 Tryptamine 7.24 d 0.21 0.06 0.15 0.05 2 Hippurate* 7.62 t 0.48 0.09 0.29 0.19 Hippurate * 7.55 t 1.67 0.33 1.02 0.55 3 Phenylacetylglycine* 7.40 m 0.34 0.08 0.49 0.12 Phenylacetylglycine 7.36 m 0.52 0.12 0.66 0.22 4 Tyrosine 7.16 m 0.24 0.06 0.20 0.04 Tyrosine ** 6.87 m 0.36 0.12 0.21 0.06 5 Fumurate** 6.53 s 0.27 0.02 0.13 0.06 6 Creatine 3.91 s 0.37 0.07 0.41 0.06 7 Creatinine** 3.08 s 0.19 0.03 0.33 0.04 8 Betaine 3.88 s 0.56 0.07 0.49 0.06 9 Glucose** 3.83 m 0.50 0.07 0.64 0.05 Glucose 3.66 m 0.83 0.09 0.99 0.24 Glucose * 3.51 m 0.54 0.09 0.67 0.05

10 Methionine 3.76 m 0.75 0.10 0.82 0.35 Methionine ** 2.14 m 0.38 0.04 0.63 0.09 Methionine ** 2.13 s 0.33 0.04 0.57 0.06

11 Lysine** 3.74 t 0.71 0.10 1.08 0.19 12 Ethanol** 3.64 q 1.04 0.08 1.22 0.06 13 Glycine* 3.52 s 0.51 0.07 0.59 0.04 14 Taurine 3.41 t 3.89 1.16 3.06 1.49 Taurine * 3.18 t 0.35 0.05 0.42 0.06

15 Hypotaurine 3.35 t 0.34 0.11 0.37 0.10


160

16 Trimethylamine –N-oxide (TMAO) 3.25 s 3.19 0.94 3.54 4.14 17 Creatinine/Creatine** 3.01 s 4.28 1.27 1.58 0.73 18 α-Ketoglutarate* 3.00 t 2.40 0.67 1.03 0.95

α-Ketoglutarate * 2.43 t 1.99 0.53 1.09 0.87 19 Dimethylglycine 2.93 s 1.78 0.56 1.61 0.98 20 Trimethylamine (TMA) 2.85 s 0.30 0.16 0.21 0.03 21 Aspartic Acid 2.84 m 0.52 0.30 0.28 0.07 22 Citrate 2.62 d 0.30 0.05 0.19 0.12 Citrate 2.60 d 0.27 0.08 0.24 0.03 Citrate 2.60 d 0.28 0.12 0.32 0.08 Citrate 2.50 d 0.36 0.09 0.41 0.03

23 Glutamine** 2.44 m 3.86 1.12 1.34 0.84 24 Succinate 2.42 s 0.84 0.08 0.81 0.16 25 Pyruvate 2.39 s 0.44 0.08 1.01 1.18 26 Glutamate* 2.36 m 0.30 0.13 0.48 0.10 Glutamate ** 2.33 m 0.49 0.06 0.71 0.09

27 Acetoacetate* 2.27 s 0.31 0.05 0.36 0.02 28 Arginine** 1.63 m 0.22 0.04 0.33 0.03 29 Alanine** 1.49 d 0.23 0.04 0.35 0.08 Alanine ** 1.47 d 0.25 0.04 0.39 0.05

30 Lactate** 1.34 d 0.34 0.06 0.55 0.17 Lactate ** 1.32 d 0.29 0.05 0.61 0.10

31 3-D-Hydroxybutyrate** 1.20 d 0.24 0.11 0.36 0.09 32 Isobutyrate 1.11 d 0.09 0.05 0.14 0.05 33 Valine 1.04 d 0.09 0.04 0.13 0.03 Valine * 0.98 d 0.09 0.03 0.14 0.03

34 Leucine 1.00 m 0.08 0.04 0.13 0.03 Leucine ** 0.94 m 0.31 0.04 0.47 0.14


161


(C)


162


1 Tryptamine 7.69 d 0.08 0.03 0.04 0.03 Tryptamine ** 7.32 d 0.22 0.03 0.07 0.06 Tryptamine ** 7.24 d 0.15 0.02 0.04 0.04 2 Hippurate** 7.62 t 0.52 0.17 0.04 0.04 Hippurate ** 7.55 t 1.79 0.57 0.16 0.14 3 Phenylacetylglycine 7.40 m 0.30 0.09 0.26 0.23 Phenylacetylglycine 7.36 m 0.49 0.14 0.30 0.29 4 Tyrosine 7.16 m 0.17 0.03 0.09 0.09 Tyrosine * 6.87 m 0.24 0.04 0.12 0.11 5 Fumurate 6.53 s 0.20 0.03 0.31 0.33 6 Creatine* 3.91 s 0.43 0.06 1.43 1.15 7 Creatinine* 3.08 s 0.25 0.03 0.13 0.12 8 Betaine 3.88 s 0.65 0.14 1.79 1.48 9 Glucose* 3.83 m 0.65 0.08 2.04 1.66 Glucose * 3.66 m 0.99 0.13 0.53 0.35 Glucose * 3.51 m 0.63 0.10 2.40 1.97

10 Methionine 3.76 m 0.96 0.13 1.34 0.77 Methionine * 2.14 m 0.45 0.06 0.27 0.20 Methionine * 2.13 s 0.35 0.05 0.19 0.14

11 Lysine* 3.74 t 0.91 0.11 2.58 1.94 12 Ethanol* 3.64 q 1.10 0.17 0.67 0.45 13 Glycine* 3.52 s 0.60 0.09 1.80 1.41 14 Taurine 3.41 t 4.81 1.25 6.11 4.22 Taurine ** 3.18 t 0.37 0.06 0.15 0.14

15 Hypotaurine 3.35 t 0.39 0.15 0.23 0.12


163

16 Trimethylamine –N-oxide (TMAO) 3.25 s 3.48 1.66 4.75 2.91 17 Creatinine/Creatine* 3.01 s 3.53 1.62 1.37 1.53 18 α-Ketoglutarate* 3.00 t 1.97 0.88 0.72 0.84 α-Ketoglutarate 2.43 t 1.58 0.66 0.82 0.71

19 Dimethylglycine** 2.93 s 1.95 0.57 0.39 0.36 20 Trimethylamine (TMA)** 2.85 s 0.22 0.03 0.06 0.09 21 Aspartic Acid** 2.84 m 0.34 0.05 0.09 0.12 22 Citrate** 2.62 d 0.26 0.07 0.10 0.08 Citrate ** 2.60 d 0.27 0.07 0.12 0.10 Citrate 2.60 d 0.32 0.13 0.15 0.13 Citrate 2.50 d 0.30 0.05 0.17 0.16

23 Glutamine* 2.44 m 3.04 1.37 1.26 1.37 24 Succinate* 2.42 s 0.78 0.11 0.49 0.34 25 Pyruvate 2.39 s 0.41 0.06 0.56 0.58 26 Glutamate 2.36 m 0.32 0.04 0.56 0.34 Glutamate 2.33 m 0.50 0.07 0.59 0.60

27 Acetoacetate* 2.27 s 0.25 0.04 0.12 0.12 28 Arginine** 1.63 m 0.18 0.02 0.08 0.07 29 Alanine 1.49 d 0.21 0.04 0.20 0.14 Alanine 1.47 d 0.22 0.04 0.21 0.14

30 Lactate 1.34 d 0.35 0.04 7.09 10.11 Lactate 1.32 d 0.29 0.06 6.77 9.67


34 Leucine 1.00 m 0.05 0.01 0.03 0.05 Leucine ** 0.94 m 0.35 0.03 0.16 0.11


164


(D)


165


1 Tryptamine** 7.69 d 0.16 0.04 0.68 0.31 Tryptamine 7.32 d 0.23 0.07 0.21 0.07 Tryptamine 7.24 d 0.18 0.04 0.18 0.09 2 Hippurate** 7.62 t 0.61 0.26 0.20 0.10 Hippurate ** 7.55 t 2.53 0.87 0.62 0.29 3 Phenylacetylglycine* 7.40 m 0.30 0.07 0.77 0.52 Phenylacetylglycine 7.36 m 0.48 0.10 1.12 0.91 4 Tyrosine 7.16 m 0.15 0.03 0.20 0.06 Tyrosine 6.87 m 0.19 0.07 0.26 0.04 5 Fumurate 6.53 s 0.15 0.04 0.14 0.07 6 Creatine** 3.91 s 0.21 0.03 0.34 0.08 7 Creatinine** 3.08 s 0.18 0.03 0.27 0.06 8 Betaine* 3.88 s 0.33 0.03 0.42 0.08 9 Glucose** 3.83 m 0.32 0.03 0.45 0.09 Glucose * 3.66 m 0.55 0.04 0.73 0.18 Glucose ** 3.51 m 0.35 0.04 0.50 0.07

10 Methionine** 3.76 m 0.50 0.05 0.81 0.21 Methionine** 2.14 m 0.49 0.07 0.84 0.17 Methionine** 2.13 s 0.43 0.06 0.75 0.16

11 Lysine** 3.74 t 0.48 0.06 0.77 0.17 12 Ethanol** 3.64 q 0.66 0.08 0.98 0.11 13 Glycine** 3.52 s 0.31 0.04 0.44 0.08 14 Taurine 3.41 t 3.11 1.41 2.28 1.96 Taurine * 3.18 t 0.25 0.03 0.32 0.07

15 Hypotaurine 3.35 t 0.21 0.09 0.26 0.09


166

16 Trimethylamine –N-oxide (TMAO) 3.25 s 2.48 1.06 1.53 1.37 17 Creatinine/Creatine** 3.01 s 4.28 1.40 0.97 0.45 18 α-Ketoglutarate** 3.00 t 2.19 0.69 0.54 0.28

α-Ketoglutarate ** 2.43 t 2.52 0.75 0.85 0.36 19 Dimethylglycine 2.93 s 1.57 0.48 1.76 1.06 20 Trimethylamine (TMA) 2.85 s 0.21 0.13 0.23 0.06 21 Aspartic Acid 2.84 m 0.34 0.23 0.41 0.25 22 Citrate** 2.62 d 0.31 0.04 0.22 0.04 Citrate * 2.60 d 0.27 0.05 0.22 0.03 Citrate 2.60 d 0.30 0.03 0.44 0.21 Citrate 2.50 d 0.36 0.14 0.39 0.07

23 Glutamine** 2.44 m 5.04 1.63 1.15 0.61 24 Succinate 2.42 s 0.90 0.05 0.84 0.23 25 Pyruvate 2.39 s 0.49 0.08 0.64 0.28 26 Glutamate** 2.36 m 0.39 0.06 0.51 0.10 Glutamate ** 2.33 m 0.57 0.09 0.81 0.10

27 Acetoacetate** 2.27 s 0.33 0.06 0.47 0.11 28 Arginine** 1.63 m 0.31 0.06 0.45 0.06 29 Alanine** 1.49 d 0.35 0.06 0.59 0.12 Alanine * 1.47 d 0.37 0.07 0.58 0.15

30 Lactate** 1.34 d 0.54 0.06 0.93 0.13 Lactate * 1.32 d 0.50 0.08 0.76 0.27

31 3-D-Hydroxybutyrate* 1.20 d 0.43 0.06 0.72 0.30 32 Isobutyrate 1.11 d 0.15 0.04 0.18 0.05 33 Valine** 1.04 d 0.14 0.03 0.19 0.04 Valine ** 0.98 d 0.14 0.03 0.20 0.04

34 Leucine** 1.00 m 0.13 0.02 0.19 0.03 Leucine ** 0.94 m 0.50 0.05 0.66 0.03


167


(E)

Table 4.5 Metabolites identified in rat urine samples for both the control group and treatment group as measured by 1H NMR. (A) Pre-dose, (B) Day 1, (C) Day 2, (D) Day 3 and (E) Day 4

Chapter 5 Conclusion and Future Work

168


This dissertation mainly focused on the study of traditional Chinese medicines

(TCMs). The results indicated that modern analytical techniques, such as CE,

LC/MS, GC/MS, NMR, MAE and PHWE, are powerful tools for the investigation

of natural products.

5.1 Summary of Results

In part I, the isolation and determination of PAs from Tussilago farfara (Kuan

Donghua) was achieved using the proposed MAE and PHWE methods coupled

with LC/MS. The binary mixture of CH3OH:H2O=1:1 acidified using HCl to PH

2-3 was found to be the optimal solvent for the extraction. This might be because

the polarity of the solvent is the most similar to those of the target compounds. For

PHWE, the effect of extraction temperature was investigated and 60oC was

selected as the optimal temperature. Since the physicochemical properties of

extraction solvent could be changed at elevated temperature, the solubility of the

target compounds could be affected. Under the optimized conditions, the MAE

and PHWE could be completed within 15 min and 50 min, respectively. The

coupling of the optimized extraction methods with LC/MSn allowed a rapid

detection of both major alkaloid, senkirkine and the minor alkaloid, senecionine.

Significant matrix-induced interferences or ion suppression in the presence of

co-eluting peaks was not observed. The extraction efficiencies of MAE and

PHWE were comparable to heating under reflux. All the results obtained indicated

that the MAE and PHWE methods proposed in this work could be alternatives for

the extraction of bioactive or marker compounds in medical plants. The methods

coupled with LC/MSn could be used for a rapid identification of unknown

compounds in botanical extracts with reference to a pure standard.


169

Furthermore, separation of four pyrrolizidine alkaloids (PAs), named senkirkine,

senecionine, retrorsine and monocrotaline, was achieved using the proposed

NACE system, and PAs contained in certain TCM were successfully detected

using this method. With the optimized separation conditions, the four PAs were

baseline separated in 15 min with very good repeatability. In order to improve the

sensitivity of NACE, two simple online preconcentration methods including large

volume sample stacking (LVSS) and field-amplified sample stacking (FASS) were

examined. In LVSS, with the increasing of the sample injection time, the peak area

increased. However, peak broadening always occurred when the injection time

was too long. This can be attributed to the fact that the analytes could not be

focused completely when the sample plug is too long. It was also found that the

limits of detection (LODs) were increased about 5~7-fold with the optimized

sample plug length. Since the increase of LODs was not so high in LVSS, FASS

was investigated for further increasing of LODs. In FASS, an organic solvent plug

was injected before sample injection to increase the sensitivity and reproducibility.

It was found that when ACN was injected as an organic plug, the peak areas of the

alkaloids were the highest. After the optimization of ACN plug length, sample

injection time and sample injection voltage, the LODs were increased about

18~89-fold for the four alkaloids. This might be because the ACN plug provided

an empty region for sample stacking and the sample migrated very quickly in

ACN plug and slowed down at the boundary between the organic plug and buffer

solution. As a result, the sample was focused in the boundary. All the results

indicate that the NACE method proposed, which is superior to routine HPLC and

aqueous CE methods with regard to selectivity, rapidity and separation efficiency,

is potentially an effective alternative tool for quality control and quantitative


170

analysis of herbal medicine in pharmaceutical industries.

In part II, metabolic profiling of berberine in rat model was achieved with

combination of GC/MS, LC/MS and NMR analysis of biofluids, such as the liver

extracts and urine samples. The study on the body weights and histopathology

suggested that rats could recover from the cholesterol lowering effect induced by

berberine. Results obtained from the GC/MS analysis of liver extracts has

revealed that the level of fatty acids, cholesterol, glucose and maltose decreased

after the administration of berberine. The level of fatty acids, glucose and so on

has been found increased in the urine samples with LC/MS and NMR analysis.

The close agreement of the results from targeted data processing of LC/MS and

the results from non-targeted data processing of 1H NMR has proved the

reliability and accuracy of the targeted data processing method. The proposed

approach combined with PCA provided a further insight on the metabolic changes

caused by berberine in rats. All the results obtained in this study have proved the

flexibility and usefulness of the approach proposed for metabonomic study.

5.2 Limitation and Future Work

With the work presented in this dissertation, modern analytical techniques, such as

CE, LC/MS, GC/MS, NMR, MAE and PHWE are proved to be powerful tools for

the natural product research. However, there still exist several aspects which

would need to be improved in further research:

1. MAE and PHWE were successfully applied to the extraction of bioactive or

marker compounds in medical plants. They were friendlier to the environment

and much simpler. However, the extraction efficiences of these methods had

not been improved significantly for the alkaloids investigated in this

dissertation. Further studies on how to improve extraction efficiences will be


171

valuable. For example, certain additives could be mixed with the sand together

for PHWE.

2. NACE was a novel CE mode developed for the analysis of natural products. It

could show some advantages over HPLC and normal aqueous CE. However,

the sensitivity enhancements were not so significant with the proposed LVSS

and FASS methods for NACE. In further studies, new preconcentration

techniques should be established. One possible method is to add some

modifiers in the BGE for FASS, such as the ionic liquid, which could affect

the EOF of the BGE zone.

3. In this dissertation, the metabolic profiling of berberine in rat model was

achieved using a combination of GC/MS, LC/MS and NMR for the analysis of

biofluids. However, the blood biochemistry had not been investigated. It was

found that berberine could cause the decrease of cholesterol. Further studies

on the mechanism of the cholesterol reduction should be valuable. The rats

could be fed with high cholesterol food first, and then berberine will be

injected. The metabonomic studies should be able to reveal how the berberine

affects the cholesterol levels in rats.

List of Publications

172

List of Publications

1. Zhangjian Jiang, Feng Liu, Jennifer Jia Lei Goh, Lijun Yu, Sam Fong Yau Li,

Eng Shi Ong, Choon Nam Ong

Determination of senkirkine and senecionine in Tussilago farfara using

microwave-assisted extraction and pressurized hot water extraction with liquid

chromatography tandem mass spectrometry, Talanta, 2009, 79, 539-546.

2. Feng Liu, Sow Yin Wan, Zhangjian Jiang, Eng Shi Ong, Sam Fong Yau Li,

Jhon Carlos Castaño Osorio

Determination of pyrrolizidine alkaloids in comfrey by liquid

chromatography-electropray ionization mass spectrometry, Talanta, 2009, 80,

916-923.

3. Zhangjian Jiang, Lijun Yu, Sam Fong Yau Li

On-column preconcentration and separation of toxic pyrrolizidine alkaloids in

herbal medicines by nonaqueous capillary electrophoresis (NACE), to be

submitted soon.

4. Zhangjian Jiang, Feng Liu, Pei Pei Gan, Eng Shi Ong, Sam Fong Yau Li

Metabolic Profiling of Berberine in Rats with gas chromatography/mass

spectrometry, liquid chromatography/mass spectrometry and 1H NMR

spectroscopy, submitted to Molecular Biosystems.

Conference Papers

173

Conference Papers

1. Zhangjian Jiang, Sam Fong Yau Li

Analysis of toxic alkaloids in herbal medicines by non-aqueous capillary

electrophoresis (NACE) with on-column preconcentration. 5th Singapore

International Chemical Conference, Singapore, Dec 2007.

2. Yan Xu, Zhangjian Jiang, Sam Fong Yau Li

Sensitive and Fast Analysis of Nitroaromatics and Nitramines in Enviromental

Soil by Micellar Electrokinetic Chromatography. 5th Singapore International

Chemical Conference, Singapore, Dec 2007.

phd thesis - jiang zhangjian - core · 2018. 1. 9. · chapter 2 determination of senkirkine and...

Documents