phd thesis - jiang zhangjian - core · 2018. 1. 9. · chapter 2 determination of senkirkine and...
TRANSCRIPT
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
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
List of Abbreviations
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
List of Abbreviations
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
List of Abbreviations
XVII
UPLC Ultra performance liquid chromatography
Chapter 1 Introduction
1
Chapter 1 Introduction
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
Chapter 1 Introduction
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
Chapter 1 Introduction
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
Chapter 1 Introduction
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
Chapter 1 Introduction
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
Chapter 1 Introduction
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
Chapter 1 Introduction
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
Chapter 1 Introduction
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
Chapter 1 Introduction
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
Chapter 1 Introduction
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
Chapter 1 Introduction
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
Chapter 1 Introduction
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].
Chapter 1 Introduction
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
Chapter 1 Introduction
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
Chapter 1 Introduction
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,
Chapter 1 Introduction
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
Chapter 1 Introduction
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.
Chapter 1 Introduction
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
Chapter 1 Introduction
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.
Chapter 1 Introduction
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.
Chapter 1 Introduction
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
Chapter 1 Introduction
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
Chapter 1 Introduction
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
Chapter 1 Introduction
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.
Chapter 1 Introduction
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.
Chapter 1 Introduction
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.
Chapter 1 Introduction
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. Jim´enez-Carmona, V. Fern´andez-P´erez,
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.
Chapter 1 Introduction
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.
Chapter 1 Introduction
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.
Chapter 1 Introduction
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.
Chapter 1 Introduction
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)
Chapter 1 Introduction
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
Chapter 1 Introduction
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.
Chapter 1 Introduction
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, Anal. Chim. Acta 482 (2003) 81.
[142] E. S. Ong, S. M. Len, J. Sep. Sci. 26 (2003) 1533.
[144] E. S. Ong, S. M. Len, J. Chromatogr. Sci. 42 (2004) 211.
[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)
Chapter 1 Introduction
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.
Chapter 1 Introduction
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.
[170] E. S. Ong, S. M. Len, Anal. Chim. Acta 482 (2003) 81.
[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
Chapter 1 Introduction
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.
Chapter 1 Introduction
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.
Chapter 1 Introduction
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.
Chapter 1 Introduction
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.
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
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
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
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.
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
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
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
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
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
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.
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
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.
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
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
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
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.
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
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
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
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)
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
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
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
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].
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
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
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
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.
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
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
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
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
)
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
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
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
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
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
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
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
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
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
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.
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
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.
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
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.
[35] E. S. Ong, S. M. Len, Anal. Chim. Acta 482 (2003) 81.
[36] E. S. Ong, S. M. Len, J. Chromatogr. Sci. 42 (2004) 211.
[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.
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
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
Chapter 3 Preconcentration and separation of toxic alkaloids in herbal medicines by non-aqueous capillary electrophoresis (NACE)
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
Chapter 3 Preconcentration and separation of toxic alkaloids in herbal medicines by non-aqueous capillary electrophoresis (NACE)
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
Chapter 3 Preconcentration and separation of toxic alkaloids in herbal medicines by non-aqueous capillary electrophoresis (NACE)
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.
Chapter 3 Preconcentration and separation of toxic alkaloids in herbal medicines by non-aqueous capillary electrophoresis (NACE)
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
Chapter 3 Preconcentration and separation of toxic alkaloids in herbal medicines by non-aqueous capillary electrophoresis (NACE)
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
Chapter 3 Preconcentration and separation of toxic alkaloids in herbal medicines by non-aqueous capillary electrophoresis (NACE)
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.
Chapter 3 Preconcentration and separation of toxic alkaloids in herbal medicines by non-aqueous capillary electrophoresis (NACE)
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)
Chapter 3 Preconcentration and separation of toxic alkaloids in herbal medicines by non-aqueous capillary electrophoresis (NACE)
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.
Chapter 3 Preconcentration and separation of toxic alkaloids in herbal medicines by non-aqueous capillary electrophoresis (NACE)
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
Chapter 3 Preconcentration and separation of toxic alkaloids in herbal medicines by non-aqueous capillary electrophoresis (NACE)
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)
Chapter 3 Preconcentration and separation of toxic alkaloids in herbal medicines by non-aqueous capillary electrophoresis (NACE)
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,
Chapter 3 Preconcentration and separation of toxic alkaloids in herbal medicines by non-aqueous capillary electrophoresis (NACE)
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.
Chapter 3 Preconcentration and separation of toxic alkaloids in herbal medicines by non-aqueous capillary electrophoresis (NACE)
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.
Chapter 3 Preconcentration and separation of toxic alkaloids in herbal medicines by non-aqueous capillary electrophoresis (NACE)
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.
Chapter 3 Preconcentration and separation of toxic alkaloids in herbal medicines by non-aqueous capillary electrophoresis (NACE)
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.
Chapter 3 Preconcentration and separation of toxic alkaloids in herbal medicines by non-aqueous capillary electrophoresis (NACE)
83
(1)
(2)
(3)
Chapter 3 Preconcentration and separation of toxic alkaloids in herbal medicines by non-aqueous capillary electrophoresis (NACE)
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).
Chapter 3 Preconcentration and separation of toxic alkaloids in herbal medicines by non-aqueous capillary electrophoresis (NACE)
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.
Chapter 3 Preconcentration and separation of toxic alkaloids in herbal medicines by non-aqueous capillary electrophoresis (NACE)
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.
Chapter 3 Preconcentration and separation of toxic alkaloids in herbal medicines by non-aqueous capillary electrophoresis (NACE)
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.
Chapter 3 Preconcentration and separation of toxic alkaloids in herbal medicines by non-aqueous capillary electrophoresis (NACE)
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.
Chapter 3 Preconcentration and separation of toxic alkaloids in herbal medicines by non-aqueous capillary electrophoresis (NACE)
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)
senkirkinemonocrotalinesenecionineretrorsine
B
Chapter 3 Preconcentration and separation of toxic alkaloids in herbal medicines by non-aqueous capillary electrophoresis (NACE)
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)
senkirkinemonocrotalinesenecionineretrorsine
A
Chapter 3 Preconcentration and separation of toxic alkaloids in herbal medicines by non-aqueous capillary electrophoresis (NACE)
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)
senkirkinemonocrotalinesenecionineretrorsine
B
Chapter 3 Preconcentration and separation of toxic alkaloids in herbal medicines by non-aqueous capillary electrophoresis (NACE)
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).
Chapter 3 Preconcentration and separation of toxic alkaloids in herbal medicines by non-aqueous capillary electrophoresis (NACE)
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.
Chapter 3 Preconcentration and separation of toxic alkaloids in herbal medicines by non-aqueous capillary electrophoresis (NACE)
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.
Chapter 3 Preconcentration and separation of toxic alkaloids in herbal medicines by non-aqueous capillary electrophoresis (NACE)
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.
Chapter 3 Preconcentration and separation of toxic alkaloids in herbal medicines by non-aqueous capillary electrophoresis (NACE)
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.
[6] R. Lebada, A. Schreier, S. Scherz, C. Resch, L. Krenn, B. Kopp, Phytochem.
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.
Chapter 3 Preconcentration and separation of toxic alkaloids in herbal medicines by non-aqueous capillary electrophoresis (NACE)
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.
[25] R. Lebada, A. Schreier, S. Scherz, C. Resch, L. Krenn, B. Kopp, Phytochem.
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.
Chapter 3 Preconcentration and separation of toxic alkaloids in herbal medicines by non-aqueous capillary electrophoresis (NACE)
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.
Chapter 4 Metabolic Profiling of Berberine in Rats with GC/MS, LC/MS and 1H NMR Spectroscopy
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
Chapter 4 Metabolic Profiling of Berberine in Rats with GC/MS, LC/MS and 1H NMR Spectroscopy
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
Chapter 4 Metabolic Profiling of Berberine in Rats with GC/MS, LC/MS and 1H NMR Spectroscopy
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%
Chapter 4 Metabolic Profiling of Berberine in Rats with GC/MS, LC/MS and 1H NMR Spectroscopy
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.
Chapter 4 Metabolic Profiling of Berberine in Rats with GC/MS, LC/MS and 1H NMR Spectroscopy
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
Chapter 4 Metabolic Profiling of Berberine in Rats with GC/MS, LC/MS and 1H NMR Spectroscopy
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
Chapter 4 Metabolic Profiling of Berberine in Rats with GC/MS, LC/MS and 1H NMR Spectroscopy
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
Chapter 4 Metabolic Profiling of Berberine in Rats with GC/MS, LC/MS and 1H NMR Spectroscopy
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.
Chapter 4 Metabolic Profiling of Berberine in Rats with GC/MS, LC/MS and 1H NMR Spectroscopy
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
Chapter 4 Metabolic Profiling of Berberine in Rats with GC/MS, LC/MS and 1H NMR Spectroscopy
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
Chapter 4 Metabolic Profiling of Berberine in Rats with GC/MS, LC/MS and 1H NMR Spectroscopy
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)
Chapter 4 Metabolic Profiling of Berberine in Rats with GC/MS, LC/MS and 1H NMR Spectroscopy
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
Chapter 4 Metabolic Profiling of Berberine in Rats with GC/MS, LC/MS and 1H NMR Spectroscopy
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
Chapter 4 Metabolic Profiling of Berberine in Rats with GC/MS, LC/MS and 1H NMR Spectroscopy
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.
Chapter 4 Metabolic Profiling of Berberine in Rats with GC/MS, LC/MS and 1H NMR Spectroscopy
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
Chapter 4 Metabolic Profiling of Berberine in Rats with GC/MS, LC/MS and 1H NMR Spectroscopy
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.
Chapter 4 Metabolic Profiling of Berberine in Rats with GC/MS, LC/MS and 1H NMR Spectroscopy
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.
Chapter 4 Metabolic Profiling of Berberine in Rats with GC/MS, LC/MS and 1H NMR Spectroscopy
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]
Chapter 4 Metabolic Profiling of Berberine in Rats with GC/MS, LC/MS and 1H NMR Spectroscopy
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]
Chapter 4 Metabolic Profiling of Berberine in Rats with GC/MS, LC/MS and 1H NMR Spectroscopy
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]
Chapter 4 Metabolic Profiling of Berberine in Rats with GC/MS, LC/MS and 1H NMR Spectroscopy
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↑*
Chapter 4 Metabolic Profiling of Berberine in Rats with GC/MS, LC/MS and 1H NMR Spectroscopy
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 ↑*
Chapter 4 Metabolic Profiling of Berberine in Rats with GC/MS, LC/MS and 1H NMR Spectroscopy
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
Chapter 4 Metabolic Profiling of Berberine in Rats with GC/MS, LC/MS and 1H NMR Spectroscopy
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
Chapter 4 Metabolic Profiling of Berberine in Rats with GC/MS, LC/MS and 1H NMR Spectroscopy
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).
Chapter 4 Metabolic Profiling of Berberine in Rats with GC/MS, LC/MS and 1H NMR Spectroscopy
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
Chapter 4 Metabolic Profiling of Berberine in Rats with GC/MS, LC/MS and 1H NMR Spectroscopy
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.
Chapter 4 Metabolic Profiling of Berberine in Rats with GC/MS, LC/MS and 1H NMR Spectroscopy
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]
Chapter 4 Metabolic Profiling of Berberine in Rats with GC/MS, LC/MS and 1H NMR Spectroscopy
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,
Chapter 4 Metabolic Profiling of Berberine in Rats with GC/MS, LC/MS and 1H NMR Spectroscopy
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
Chapter 4 Metabolic Profiling of Berberine in Rats with GC/MS, LC/MS and 1H NMR Spectroscopy
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.
Chapter 4 Metabolic Profiling of Berberine in Rats with GC/MS, LC/MS and 1H NMR Spectroscopy
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,
Chapter 4 Metabolic Profiling of Berberine in Rats with GC/MS, LC/MS and 1H NMR Spectroscopy
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.
Chapter 4 Metabolic Profiling of Berberine in Rats with GC/MS, LC/MS and 1H NMR Spectroscopy
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.
Chapter 4 Metabolic Profiling of Berberine in Rats with GC/MS, LC/MS and 1H NMR Spectroscopy
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,
Metabolomics 5 (2009) 84.
[48] T. K. Karakach, E. C. Huenupi, E. C. Soo, J. A.Walter, L. O. B. Afonso,
Metabolomics 5 (2009) 123.
[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.
Chapter 4 Metabolic Profiling of Berberine in Rats with GC/MS, LC/MS and 1H NMR Spectroscopy
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)
Chapter 4 Metabolic Profiling of Berberine in Rats with GC/MS, LC/MS and 1H NMR Spectroscopy
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)
Normalized peak intensitya (%) NO. Metabolite Retention Time
(min) m/z Control Ave ±SD Treated
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
Chapter 4 Metabolic Profiling of Berberine in Rats with GC/MS, LC/MS and 1H NMR Spectroscopy
138
(n=8) and treatment group (n=8) is based on two tailed student t test (* p<0.05, ** p<0.01). (A)
Chapter 4 Metabolic Profiling of Berberine in Rats with GC/MS, LC/MS and 1H NMR Spectroscopy
139
Metabolites identified in rat urine samples for both the control group and treatment group as measured by LCMS, day 1 (positive mode)
Normalized peak intensitya (%) NO. Metabolite Retention Time
(min) m/z Control Ave ±SD Treated
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
Chapter 4 Metabolic Profiling of Berberine in Rats with GC/MS, LC/MS and 1H NMR Spectroscopy
140
(n=8) and treatment group (n=8) is based on two tailed student t test (* p<0.05, ** p<0.01). (B)
Chapter 4 Metabolic Profiling of Berberine in Rats with GC/MS, LC/MS and 1H NMR Spectroscopy
141
Metabolites identified in rat urine samples for both the control group and treatment group as measured by LCMS, day 2 (positive mode)
Normalized peak intensitya (%) NO. Metabolite Retention Time
(min) m/z Control Ave ±SD Treated
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
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
Chapter 4 Metabolic Profiling of Berberine in Rats with GC/MS, LC/MS and 1H NMR Spectroscopy
142
(n=8) and treatment group (n=8) is based on two tailed student t test (* p<0.05, ** p<0.01). (C)
Chapter 4 Metabolic Profiling of Berberine in Rats with GC/MS, LC/MS and 1H NMR Spectroscopy
143
Metabolites identified in rat urine samples for both the control group and treatment group as measured by LCMS, day 3 (positive mode)
Normalized peak intensitya (%) NO. Metabolite Retention Time
(min) m/z Control Ave ±SD Treated
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
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
Chapter 4 Metabolic Profiling of Berberine in Rats with GC/MS, LC/MS and 1H NMR Spectroscopy
144
(n=8) and treatment group (n=8) is based on two tailed student t test (* p<0.05, ** p<0.01). (D)
Chapter 4 Metabolic Profiling of Berberine in Rats with GC/MS, LC/MS and 1H NMR Spectroscopy
145
Metabolites identified in rat urine samples for both the control group and treatment group as measured by LCMS, day 4 (positive mode)
Normalized peak intensitya (%) NO. Metabolite Retention Time
(min) m/z Control Ave ±SD Treated
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
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
Chapter 4 Metabolic Profiling of Berberine in Rats with GC/MS, LC/MS and 1H NMR Spectroscopy
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.
Chapter 4 Metabolic Profiling of Berberine in Rats with GC/MS, LC/MS and 1H NMR Spectroscopy
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
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).
(A)
Chapter 4 Metabolic Profiling of Berberine in Rats with GC/MS, LC/MS and 1H NMR Spectroscopy
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
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).
(B)
Chapter 4 Metabolic Profiling of Berberine in Rats with GC/MS, LC/MS and 1H NMR Spectroscopy
149
Metabolites identified in rat urine samples for both the control group and treatment group as measured by LCMS, day 2 (negative mode) Normalized peak intensitya (%) NO. Metabolite Retention Time (min) m/z Control Ave ±SD Treated Ave ±SD
1 Fumaric Acid* 1.00 115 0.0504 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
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).
(C)
Chapter 4 Metabolic Profiling of Berberine in Rats with GC/MS, LC/MS and 1H NMR Spectroscopy
150
Metabolites identified in rat urine samples for both the control group and treatment group as measured by LCMS, day 3 (negative mode) Normalized peak intensitya (%) NO. Metabolite Retention Time (min) m/z Control Ave ±SD Treated Ave ±SD
1 Fumaric Acid** 1.00 115 0.0345 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
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).
(D)
Chapter 4 Metabolic Profiling of Berberine in Rats with GC/MS, LC/MS and 1H NMR Spectroscopy
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
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).
(E)
Chapter 4 Metabolic Profiling of Berberine in Rats with GC/MS, LC/MS and 1H NMR Spectroscopy
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.
Chapter 4 Metabolic Profiling of Berberine in Rats with GC/MS, LC/MS and 1H NMR Spectroscopy
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
Chapter 4 Metabolic Profiling of Berberine in Rats with GC/MS, LC/MS and 1H NMR Spectroscopy
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
Chapter 4 Metabolic Profiling of Berberine in Rats with GC/MS, LC/MS and 1H NMR Spectroscopy
155
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).
(A)
Chapter 4 Metabolic Profiling of Berberine in Rats with GC/MS, LC/MS and 1H NMR Spectroscopy
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
Chapter 4 Metabolic Profiling of Berberine in Rats with GC/MS, LC/MS and 1H NMR Spectroscopy
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
31 3-D-Hydroxybutyrate 1.20 d 0.27 0.11 0.21 0.15 32 Isobutyrate 1.11 d 0.09 0.05 0.07 0.05 33 Valine 1.04 d 0.10 0.04 0.09 0.04 Valine 0.98 d 0.11 0.03 0.10 0.04
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
Chapter 4 Metabolic Profiling of Berberine in Rats with GC/MS, LC/MS and 1H NMR Spectroscopy
158
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).
(B)
Chapter 4 Metabolic Profiling of Berberine in Rats with GC/MS, LC/MS and 1H NMR Spectroscopy
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
Chapter 4 Metabolic Profiling of Berberine in Rats with GC/MS, LC/MS and 1H NMR Spectroscopy
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
Chapter 4 Metabolic Profiling of Berberine in Rats with GC/MS, LC/MS and 1H NMR Spectroscopy
161
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).
(C)
Chapter 4 Metabolic Profiling of Berberine in Rats with GC/MS, LC/MS and 1H NMR Spectroscopy
162
Metabolites identified in rat urine samples for both control and treatment groups as measured by 1H NMR, day 3 Normalized peak intensitya NO. Metabolite Chemical Shift (ppm) Multiplicity Control Ave ±SD Treated Ave ±SD
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
Chapter 4 Metabolic Profiling of Berberine in Rats with GC/MS, LC/MS and 1H NMR Spectroscopy
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
31 3-D-Hydroxybutyrate 1.20 d 0.25 0.04 1.60 2.48 32 Isobutyrate 1.11 d 0.04 0.01 0.03 0.05 33 Valine 1.04 d 0.04 0.01 0.02 0.04 Valine 0.98 d 0.05 0.02 0.03 0.05
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
Chapter 4 Metabolic Profiling of Berberine in Rats with GC/MS, LC/MS and 1H NMR Spectroscopy
164
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).
(D)
Chapter 4 Metabolic Profiling of Berberine in Rats with GC/MS, LC/MS and 1H NMR Spectroscopy
165
Metabolites identified in rat urine samples for both control and treatment groups as measured by 1H NMR, day 4 Normalized peak intensitya NO. Metabolite Chemical Shift (ppm) Multiplicity Control Ave ±SD Treated Ave ±SD
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
Chapter 4 Metabolic Profiling of Berberine in Rats with GC/MS, LC/MS and 1H NMR Spectroscopy
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
Chapter 4 Metabolic Profiling of Berberine in Rats with GC/MS, LC/MS and 1H NMR Spectroscopy
167
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).
(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
Chapter 5 Conclusion and Future Work
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.
Chapter 5 Conclusion and Future Work
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
Chapter 5 Conclusion and Future Work
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
Chapter 5 Conclusion and Future Work
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.