Full Text No Signature
Full Text No Signature
Full Text No Signature
Doctor of Philosophy
Front matter II
0.6. Acknowledgements XX
1.3. Risks of HM 4
III
Table of contents
Endoherb™ 12
selected analytes 24
1.6.5.3. Precision 27
1.6.5.4. Linearity 27
1.6.5.6. Stability 28
arvense L. 29
1.7.3. Chemometrics 30
IV
Table of contents
1.8. Aims 34
2.2. Equipment 38
2.3. Reagents 39
solution 40
solutions 41
solution 41
V
Table of contents
umbelliferone by GC-MS 45
3.1. Samples 49
3.2.1. TLC 50
VI
Table of contents
4.3.3.2. Accuracy 70
4.3.3.3. Precision 72
4.3.3.5. Linearity 73
4.3.3.6. Stability 74
VII
Table of contents
6.1. Summary 99
6.2.1. A systematic method for the selection of analytes for the quantitative
solvents. 101
6.2.3. Evaluation of the extraction efficiencies of the three solvents tested. 101
VIII
Table of contents
References 106
IX
List of figures
complementary medicines.24 7
Figure 1.2: A flowchart for determining the need for analytical method
validation.27 25
Figure 4.1: The fractionation scheme used for the analysis of the analytes. 58
standard (red) and the 95 % ethanolic extract (black). λdet = 254 nm. 59
Figure 4.9: Comparison between the standard (red) and sample (black) UV
X
List of figures
Figure 4.10: Comparison between the standard (red) and sample (black) UV
spectra of paeoniflorin. 65
Figure 4.11: Comparison between the standard (red) and sample (black) UV-Vis
spectra of rhein. 65
different extracts. 76
chicoric acid). 80
Figure 5.4: The number of peaks detected in the TLC, LC-PDA and LC-MS
chromatograms. 85
XI
List of figures
Figure 5.9: matK DNA barcodes of the original plant material used to produce
Figure 5.10: rbcL DNA barcodes of the original plant material used to produce
Figure 5.11: The antioxidant activity of the various E. arvense extracts using the
Figure 5.12: The antioxidant activity of the various E. arvense extracts as profiled
Figure 5.13: A representative chromatogram of the China 8 sample using the on-
with the DPPH absorbance at 515 nm. Analytes that have DPPH
XII
List of figures
using (A) LC-MS and (B) LC-PDA and (C) the proposed MS
fragmentation pattern. 96
acetylglucoside using (A) LC-MS and (B) LC-PDA and (C) the
XIII
List of tables
China.46) 14
Table 1.2: Chemicals in Endoherb™ that may be used for listing on the ARTG.
Table 1.3: The analytes recommended for analysis in the respective herb based
Table 1.4: The ranking system used to select analytes for the compound-based
analysis of Endoherb™. 20
Table 1.5: The analytes chosen in the Endoherb™ formulation, based on British
standard. 42
XIV
List of tables
Table 4.5: Summary of the calibration curve linearity for the analytes. 74
Table 4.6: Summary of the results obtained for each of the Endoherb™ extracts
tested 75
Table 5.1: The parameters used for chromatographic processing and a brief
Suresh Govindaraghavan. 95
XV
List of publications
genome and metabolome: Towards comprehensive quality standards for medicinal herb
Sucher, N. J., Hennell, J. R. & Carles, M. C. DNA fingerprinting, DNA barcoding, and
next generation sequencing technology in plants. Methods Mol Biol 862, 13-22 (2012).
Hennell, J. R., D'Agostino, P. M., Lee, S., Khoo, C. S. & Sucher, N. J. Using
GenBank(R) for Genomic Authentication: A Tutorial. Methods Mol Biol 862, 181-200
(2012).
Diaz, P. L., Hennell, J. R. & Sucher, N. J. Genomic DNA extraction and barcoding of
Lee, S., Khoo, C. S., Pearson, J. L., Hennell, J. R. & Bensoussan, A. Liquid
L.) in the form of the raw herb and of the dried aqueous extract. J AOAC Int 92, 789-
796 (2009).
XVI
List of publications
Lee, S. Khoo, C. S., Hennell, J. R., Pearson, J. L., Jarouche, M., Halstead, C. W. &
lactiflora) as a raw herb and dried aqueous extract. J AOAC Int 92, 1027-1034 (2009).
Hennell, J. R., Lee, S., Khoo, C. S., Gray, M. J. & Bensoussan, A. The determination
of glycyrrhizic acid in Glycyrrhiza uralensis Fisch. ex DC. (Zhi Gan Cao) root and the
dried aqueous extract by LC-DAD. J Pharm Biomed Anal 47, 494-500 (2008).
Halstead, C. W., Lee, S., Khoo, C. S., Hennell, J. R. & Bensoussan, A. Validation of a
method for the simultaneous determination of four schisandra lignans in the raw herb
and commercial dried aqueous extracts of Schisandra chinensis (Wu Wei Zi) by RP-LC
XVII
Statement of authentication
The work presented in this thesis is, to the best of my knowledge and belief, original
except as acknowledged in the text. I hereby declare that I have not submitted this
material, either in whole or in part, for a degree at this or any other institution.
Candidate’s signature:
B.Sc. (Honours)
July 2012
XVIII
Dedication
0.5. Dedication
I would like to dedicate this work to my parents, who have given their fullest love,
XIX
Acknowledgements
0.6. Acknowledgements
One) for all the gems of knowledge and pearls of wisdom that he has provided over the
studies. Your passion for science, research and knowledge is admirable. I also wish to
Many thanks go to the Centre for Complementary Medicine Research, PT Soho Industri
Pharmasi and the University of Western Sydney for supporting this research.
Specifically I wish to thank Professor Gerald Muench, Dr Sam Lee (The Lab Manager),
Dr Ray Helliwell, Jarryd Pearson, Mitchell Low, John Truong, Patricia Diaz, Professor
Alan Bensoussan, Ros Priest, Micki Macdonald and the rest of the CompleMED team
for their generous support. I also wish to thank the chemistry technical officers Dr
Shane Griffin, Tuan Nguyen and Zoran Polic for all the help they have given me.
enjoyable and memorable. Especially to Ben Harper, Ben Singh, Dr Elise Wright, Paul
D’Agostino, Justin Sissng, Madhura Manohar, Dr Nikita Orkey, Anwen Krause- Heuer,
Greg Czaban and Caitlin Short. I would also like to convey my appreciation to my
friends at the Macarthur Symphonic Wind and Concert Bands, namely Ian Birt,
XX
Acknowledgements
Stephanie Parkin, Brooke Anderson and Kim Jackson for being a constant source of joy
Finally my deepest gratitude to my mother Joan, father Stephen, sister Ariane and the
rest of my family for their love and accommodation of my eccentric needs during these
academic years.
XXI
List of abbreviations
Abbreviation Definition
(-)-ESI Negative mode ESI
-
[M-H] Molecular ion
AAPH 2,2’-Azobis(2-amidinopropane) hydrochloride
ARGCM Australian Guidelines for Complementary Medicines
ARTG Australian Register of Therapeutic Goods
BLAST Basic Local Alignment Search Tool
BP British Pharmacopoeia
CBoL Consortium for the Barcode of Life
cGMP Code of Good Manufacturing Practice
CM Complementary Medicine
DPPH 2,2-di(4-tert-octylphenyl)-1-picrylhydrazyl
EI Electron Impact
ESI Electrospray Ionization
ET Electron Transfer
FID Flame Ionization Detector
GC Gas Chromatography
HAT Hydrogen Atom Transfer
HCA Hierarchical Cluster Analysis
HM Herbal medicine
KNN k-Nearest Neighbor Analysis
LC High-Performance Liquid Chromatography
matK Maturase K
MS Mass Spectrometry
NPPEG Natural Products (diphenylboric acid 2-aminoethyl ester) and Polyethylene
Glycol 4000 (PEG) reagent
OCM Office of Complementary Medicines
ORAC Oxygen Radical Absorbance Capacity
P-PRC Pharmacopoeia of the People's Republic of China
PCA Principal Component Analysis
PCR Polymerase Chain Reaction
PVDF Polyvinylidene Difluoride
Q1 / Q3 Quadrupole 1 / Quadrupole 3
QA Quality Assurance
QC Quality Control
rbcL Ribulose-1,5-bisphosphate carboxylase/oxygenase large subunit
RSD Relative Standard Deviation
SD Standard Deviation
SIM Selective Ion Monitoring
TCM Traditional Chinese Medicine
TGA Therapeutic Goods Administration
TGO Therapeutic Goods Orders
XXII
List of abbreviations
Abbreviation Definition
TLC High-Performance Thin Layer Chromatography
US$ US Dollar
UV Ultraviolet
WHO World Health Organization
λdet Wavelength of detection
λmax Wavelength of maximum absorption
XXIII
Abstract
0.8. Abstract
The past decade has seen an unprecedented growth in the popularity of complementary
continues to grow, serious concerns have been raised about their quality and safety.
Herbal medicine quality control (QC) and assurance (QA) poses a great challenge, as
the large assortment of complex chemicals and the compositional variation found within
herbal mixtures makes analysis especially difficult. Here, the two commonly used
paradigms for herbal quality assessment; the compound- and pattern-based approaches
product consistency. The complex nature of the 13-herb Endoherb™ formulation made
it an exemplar model for assessing this approach. The methodology used in this study
carried out using a logical and systematic method for the selection of the analytes, and
then developing a method for their sensitive, specific and accurate analysis. Several
photodiode array (PDA) and mass spectrometric (MS) detection, as well as gas
chromatography (GC) with MS and flame ionisation detection (FID) were employed for
the specific analysis of the selected analytes based on factors such as their polarity,
utilising sonication as the extraction method. The results show that increasing the
XXIV
Abstract
organic modifier in the extraction solvent, even for predominantly polar analytes can
The pattern-based method compares the chromatographic profile of the extract of a new
batch with that of a target or reference batch. Combined with new genetic and
to the quality of the raw materials. E. arvense was used as a model to observe the effect
software ‘R’ was necessary for subsequent statistical analysis as the instrumental
contribution to the profiles made the detection of peaks cumbersome and the statistical
inferences inaccurate. The novel use of k-nearest neighbour analysis combined with
equivalence. Importantly, the results indicate that the implied biological effect of the
extracts is not reflected in their chemical profile. Finally, DNA barcoding using the loci
rbcL and matK was successfully used to authenticate the raw materials.
XXV
CHAPTER 1.
GENERAL INTRODUCTION
General introduction
medicine (HM).2 The herbs prescribed are believed to work in concert together to treat
the diagnosed ailment. In HM, the herbs are mixed, decocted and generally consumed as
a tea.
The past decade has seen the HM industry grow from a modest base to a global industry
grossing over US$15.6 billion annually.3 In parts of Asia and Africa, over 80 % of the
population depend on traditional medicine for health care. It is currently estimated that
countries with organised primary health care systems, HM has been integrated into
preventative care.5
The World Health Organisation (WHO) has identified that education, training and
research to support the safe use of complementary medicines is currently lacking, due in
secondary metabolites.7
2
General introduction
depending on their role in the plant.8 Primary metabolites play a critical role in the
reproduction, and includes chemicals such as amino acids and lipids.9 The remaining
chemicals are known as secondary metabolites and are thought to perform roles in
signalling and plant defence, such as protecting plants from herbivores, microbial
infection, other plants and UV damage.9 Three basic chemical families exist, which are
Secondary metabolites are extensively analysed as part of herbal regulation and in the
genus of plant, and even individual species of plant can produce unique secondary
Plant secondary metabolites have been used throughout history, including in Western
culture, for a diverse range of tasks such dyes, glues, oils, waxes, flavours and
perfumes.9 In the past 50 years research has increased significantly into the uses of
novel drugs. Secondary metabolites have been reported in the literature to have a
3
General introduction
1.3. Risks of HM
Serious concerns have been raised about the quality and safety of HM, particularly in
Western countries. In the past, HM have been contaminated with organic and inorganic
matter and some have been adulterated with prescription drugs.14,15 Furthermore, the
components is generally not known. In some cases the concentrations of putative active
ingredients are lower than expected but occasionally a component could be too high, so
the user may consume more than the maximum recommended dosage.3
reactions, which are directly related to the active medicine itself, can be either type A,
where they have predictable toxicity such as an overdose; or type B where they have an
in nature, and are as a result of some failure of good handling or good manufacturing
4
General introduction
Herbal nomenclature and a general lack of certified herbal reference materials are
ways: the common name, the translated or pinyin name, the pharmaceutical name, and
the binomial botanical name.17 Often the common and pinyin names can represent more
than one species or even different plant parts. The use of authenticated herbs can also
lead to ambiguity, considering the term species represents ‘more or less arbitrary and
subjective man-made units’ and ‘there are no objective infallible criteria for rank-
‘QC-by-input’ whereby the manufacturer only needs to show that the quantity of each
herbal extract in the herbal formulation complies with the label claim when the bulk
Goods Administration (TGA) that putative actives be quantified, though the regulatory
suppliers now provide standardised extracts, which are extracts standardised to achieve
a target concentration of one marker and occasionally two markers. The selected marker
concentration in the herb. The assumption is made that the other non-standardised
raw material. Factors such as genetic drift; environmental conditions such as soil,
5
General introduction
climate, rainfall; the age and plant part used; time of harvest, post-harvest treatment,
storage and processing all contribute to this variation.19,20 These variations can
contribute significantly to batch-to-batch variation of the product and hence alter its
effectiveness.
the common and pinyin names for herbs can sometimes cover more than one botanical
species or for more economic reasons such as a shortage or increased cost of the
original ingredient.15 Regardless of the reason, the consequence may be a poor quality
tetranda with Aristolochia fangchi in a medicine designed to help with weight loss.
A. fangchi contains the nephrotoxin aristolochic acid that resulted in rapidly progressive
interstitial nephritis with terminal renal failure for those taking it.15
HM are sometimes adulterated with Western prescription drugs in order to increase their
efficacy.15 For example, an epileptic patient fell into a coma after taking an HM in
China laced with phenytoin, an anti-epileptic drug. No mention was made of any
The TGA is the Australian Government body responsible for the regulation of
Act 1989i.23
6
General introduction
As presented in Figure 1.1, CM is legally a general term for many kinds of medicines,
Figure 1.1: Different types of medicines classified by the TGA as complementary medicines.24
The Therapeutics Goods Act 1989i requires therapeutic goods that are imported or
HM can be recorded with the TGA as either a listed product or as a registered product.20
Formulations can be listed and assigned an AUST L number if they contain ingredients
that the TGA considers to be of low public health concern. Listed products must be
they contain herbs that are either restricted by the Standards for the Uniform Scheduling
of Drugs and Poisons, if the substance is a designated active ingredient that has an
established identity and tradition of use, or has been identified by the TGA as being of
7
General introduction
The British Pharmacopoeia (BP) and Therapeutic Goods Orders (TGOs) are the official
standards for regulatory purposes in Australia.27 The BP and TGO monographs provide
the minimum requirements that must be met for a medicine to comply with the cGMP.
The requirements of the monographs must be met except where a justification for not
doing so is authorised by the TGA. The TGA can consider the suitability of other
Besides ensuring safety, the aim of a QC program is to ensure a consistent product and
hence some predictable outcome for the consumer. The aim of ensuring product safety
1. Chemical testing
2. Biochemical testing
8
General introduction
Ideally chemical, biochemical and biological testing should all be performed but this
carry out, though the analytical method development process itself may be quite time
consuming.
Methods of analysis must therefore focus on being able to separate these analytes and
chromatography (LC) and gas chromatography (GC) are extensively used for the quality
assessment of HM.
TLC is commonly used for the rapid identification of herbal extracts. TLC has been
Unfortunately a number of key steps rely heavily on operator skill, giving rise to poor
LC is by far the most popular chromatographic technique for analysing herbal products,
9
General introduction
plants.29 One of the major advantages of LC is the ability to hyphenate with various
detectors, the main ones being the photodiode array (PDA) and mass spectrometer (MS)
detectors.
formulation are used to quantify quality and hence product consistency.30 This approach
assumes that those constituents not quantified do not make a significant contribution to
the therapeutic effect of the medication. Consequently, this method does not give a
complete picture of an herbal product, as many components may be responsible for the
purported therapeutic effect and may work synergistically or antagonistically with each
other.7 The extent to which these assumptions are valid will likely depend on how
competently the type and number of analytes have been selected for monitoring.
Obviously the more analytes the better, but this has to be balanced against the cost,
HM have been referred to as a ‘black box’ due to the plethora of unknown chemicals
profile of an extract of a new batch with that of a target or reference batch, without
however desirable if they have been shown to have some pharmacological effect.7
10
General introduction
Traditionally, comparison between batches is performed using TLC and LC, however
the LC profile is more reproducible as the result is less dependent on operator skill.
Statistics are often utilised to determine whether the variation between the observed
Characterisation of plants at the genome level has been touted as the method by which
where polymerase chain reaction (PCR) is used for the amplification of a small locus of
genomic DNA, usually chloroplast DNA of the plant.35 DNA sequences are then
the raw herbal starting material and not to the prepared extracts. This can be
problematic since most commercial herbal formulations are prepared by mixing single
herb extracts rather than producing an extract from the raw herb mixture. Thus the
several regulatory agencies such as the European Medicines Agency (EMA) to serve as
testing may include cell line testing, animal testing and genomic response testing.
11
General introduction
One of the simplest ways to characterise the biological effects of an herb is by assessing
some overall pharmacological property like DPPH free radical scavenging activity,
phenols, nitric oxide production by the Griess assay and Fe3+ reducing antioxidant
power. This is particularly applicable to plant extracts that containing phenols, which
The most commonly used approach in HM QC is to identify, quantitate and monitor the
bioactive analytes are identified, then the quantitation of these will provide a reliable
measure of product quality and hence enable the manufacturer to produce a consistent
product.
Section 4.4.2.2. of the ARGCM guidelines recommends the analysis of at least one
unique analyte from each constituent herb part in an herbal formulation.20 This is to
prevent adulteration of the herbal formulation with different herb parts or with a
different herb altogether. The analytes selected should ideally contribute to the
therapeutic effect of the formulation. This ideal is often not achieved because relatively
few plants have chemicals that are absolutely unique to them, but it may be sufficient
for the chemical just to be unique in the context of the formulation. There is currently
no regulatory criteria for the selection of analytes in complex herbal medicines for QC
and QA purposes, except that the analyte analysed should contribute to the effect of the
formulation.20
12
General introduction
outside of the uterus and is commonly associated with chronic inflammation, pelvic
40 % of infertile women around the world.41,43 The aetiology and pathogenesis of this
disease is poorly understood, and while medical and surgical treatment are effective for
common.44
clinical trial, Endoherb™ was shown to have significant benefits for relieving
marked pain relief and improved quality of life without unpleasant side effects.45 The
herbs used and their percentage composition in the Endoherb™ formulation is presented
in Table 1.1. For consistency, the herbs used in this formulation are referred to by their
pinyin name.
In HM, herbs are usually consumed as a water decoction, though aqueous alcoholic
extracts are also used using wines and spirits. In this work, three different extraction
solvents were used for the preparation of the Endoherb™ extract, these are: pure water,
35 % aqueous ethanol and 95 % aqueous ethanol. A water extraction was used for the
13
Pinyin name
Chinese name Percent in formulation (%) Latin pharmaceutical name Latin botanical name
14
General introduction
Glycyrrhiza glabra L.
General introduction
The first step in the QC of an HM is to decide on the set of analytes to monitor. If the
QC results are to reasonably reflect the quality of the medication, then this selection
activity and concentration of the analyte, factors such as regulatory compliance and the
analyte uniqueness in the context of the formulation are also considered. Finally, the
practicality of the task in terms of the number of analytes, availability of the pure
The ‘Substances that may be used in Listed Medicines’ publication by the TGA lists the
herbs and chemicals eligible for listing on the ARTG.47 The list includes the approved
role of the substance as an active (A), excipient (E), and/or component (C) ingredient.
Based on the herbs listed Table 1.1, the Endoherb™ formulation is not eligible to be a
listed medicine, principally because it contains taoren, which contains amygdalin.47 The
accepted therapeutic use of the herbs and major constituents are presented in Table 1.2.
Substances marked as C are not approved as substances for use in their own right and
can only be used in conjunction with an approved source. This table was used as a
15
Pinyin name
Latin botanical name Use Notes Active chemicals Accepted Notes
in herb use
Aiye
Artemisia argyi Levl. et Vant. A, E Oil derived from this species is a
Customs Prohibited Import.
Baishao
Paeonia lactiflora Pall. A, E
Taoren
Prunus persica (L.) Batsch Only Prunus dulcis var. dulcis Amygdalin C Listed
seed is permitted. Amygdalin and medicines must
Hydrocyanic acid are mandatory not contain any
components. amygdalin.
16
General introduction
General introduction
Only 6 out of the 13 herbs in the Endoherb™ formulation have a BP monograph for
their analysis.48 These herbs are gancao, dahuang, guizhi, danggui, baishao and chishao
(as P. lactiflora). The Pharmacopoeia of the People’s Republic of China (P-PRC) has a
wider breadth of coverage for HM than the BP, and was therefore used in this project as
the reference for the remaining herbs.46 As each pharmacopoeia reference is for the
individual herb and not the formulation, the references were only used as a guide for
analyte selection. Based on the BP and P-PRC monographs, Table 1.3 presents the
17
General introduction
Table 1.3: The analytes recommended for analysis in the respective herb based on the British
recommendations.
aiye, baizhu, chuanxiong, fuling and guizhi, analytes were chosen based on published
literature. Criteria were developed within CompleMED to examine all the reported
analytes of each constituent herb in a formulation for their importance in the action
towards the targeted disease.49,50 This is done to reduce the number of analytes required
The analytes are ranked in importance and selected according to the following criteria:30
18
General introduction
• Toxicity of the analyte and the safety limits applied to its use
• The part of the plant used in the formulation, such as the seed, root or rhizome
These criteria were used to create a numerical ranking system from 0 – 6 as presented in
Table 1.4. The higher the number, the more important the analyte is for monitoring
purposes.
19
General introduction
Table 1.4: The ranking system used to select analytes for the compound-based analysis of
Endoherb™.
Ranking Details
6 Analyte has highest bioactivity related to major symptoms of disease for which there
is sufficient evidence to support its activity (≥ 2 studies)
AND
Analyte is present in high yield in the herb
AND
Traditional/other use of the herb and its therapeutic action supports the activity of the
analyte
AND
Bioavailability of analyte and/or its metabolites are known to be bioavailable
OR
Analyte may be toxic and needs to be screened to comply to safety limits
5 Analyte has high bioactivity related to major symptom/s of the disease for which
there is sufficient evidence to support its activity (≥ 2 studies)
AND
Analyte is present in high yield in the herb
AND
Traditional/other use of the herb and its therapeutic action supports the activity of the
analyte
4 Analyte has bioactivity related to a symptom (major/minor) of the disease, however
there is only 1 study to support its action
AND
Analyte is present in high yield in the herb
AND
Traditional/other use of the herb and its therapeutic action supports the activity of the
analyte
3 Analyte has bioactivity related to a symptom (major/minor) of the disease, however
there is only 1 study to support this action
AND
Analyte is present in low yield in the herb
AND
Traditional/other use of the plant maybe be related to action of the analyte
2 Analyte’s activity is indirectly related to symptom of disease
OR
Analyte is present in too low a yield in the herb for it to be screened
OR
Analyte’s activity is not supported by traditional and therapeutic use of herb
1 Analyte’s activity is not related to indication of the disease of interest or no
commercial source of standard is available.
0 There are no bioactivity studies currently available for this analyte
20
General introduction
This ranking system was also used to rank the previously identified analytes for
to 1-2 per herb since this is a 13-herb formulation and the analysis would be
impractically large otherwise. In total, 12 analytes were chosen for analysis. This
number of analytes is probably the close to the maximum number a industry QC lab can
The herbs baishao, chishao and mudanpi are of the same genus, Paeonia, and are
consequently too similar for a unique analyte to be found to differentiate them. Instead,
paeoniflorin and paeonol, which are common to the Paeonia genus are analysed.
danggui.
21
General introduction
Table 1.5: The analytes chosen in the Endoherb™ formulation, based on British Pharmacopoeia
Administration (TGA), and Hong Kong Department of Health (DOH) recommendation and the
Paeonol 5 BP48
Bensky12
Tang13
Ligustilide 5 Bensky12
Tang13
Wagner53
DOH54
22
General introduction
23
General introduction
1.6.4. Development and validation of analytical methods of analysis for the selected
analytes
analytical procedure is suitable for its intended purpose, for example, identification,
flowchart presented in Figure 1.2, the first step is to develop the analytical method for
Endoherb™ and to test and characterise its performance through the validation process.
The validation parameters outlined in Section 1.6.5 are adapted from the European
96/23/EC.64 Once the analytical method is developed, it can be used to analyse different
sources of the herb or herbal formulation. A full method validation involves inter-
Analytical Chemists.
24
General introduction
Figure 1.2: A flowchart for determining the need for analytical method validation.27
standard peaks. If the PDA is used, identity confirmation is achieved by comparing the
UV-Vis spectra of the standard and sample peaks to determine their match. Comparison
of the UV-Vis spectra is informative if the spectrum has multiple peaks for several
points of evaluation. For the MS detector, spectral comparison is more reliable. In the
25
General introduction
enables the identification of the molecular ion, which can then be fragmented by
collision-induced dissociation to form distinct daughter ions. In the case of GC-MS, the
molecular ion is usually fragmented by electron impact (EI), generally forming many
ion fragments. The MS spectra of sample and standard peaks can then be compared for
the presence of these ions as well as their relative intensities. Additionally, if the
Accuracy is assessed to determine whether the analytical method gives results that are
close to the true value. In this study accuracy is determined from fortification (spiking)
recoveries carried out at the 50, 100 and 200 % levels. Fortification is carried out using
a mixed standard fortification solution, where the ratio of the concentration of the
shown in Figure 1.3. A volume of fortification solution is added to the sample such that
the analyte peak height (or area) will increase by approximately 50, 100 and 200 %, to
give fortification levels of 50, 100 and 200 % respectively. After adding the fortification
solution the solvent is allowed to evaporate before starting the analysis. Each
fortification level and the un-fortified sample are determined with n = 7 replicates to
26
General introduction
1.6.5.3. Precision
set of replicate results. Precision is assessed for both the sample and standards. This
agreement is expressed in terms of the standard deviation (SD) and relative standard
1.6.5.4. Linearity
Instrumental linearity refers to the ability of the detector to produce results that are
concentration of the chemical standard. Method linearity, that is the linearity starting
from extraction to instrumental analysis, is tested over a range in which the analyte
of an expected lowest value to 100 % above an expected highest value. Method linearity
tests can reveal, for example, if the extraction solvent has reached saturation for the
analyte.
The limit of detection (LOD) is defined as the lowest amount of analyte that can be
method LOD is typically defined as three times the SD of a set of replicate (typically
27
General introduction
The limit of quantitation (LOQ) is the lowest measured amount of analyte in a sample
that can be quantified within a specified degree of accuracy and reproducibility. The
method LOQ is generally defined as ten times the SD of a set of replicate (typically
LOD and LOQ are calculated statistically in this way as both the sample preparation and
instrumental uncertainties are accounted for. Determination of the detection limits using
limit. If the LOD is determined from the SD of replicates of multiple injections of the
same solution, the uncertainty also incorporates the error contribution from the
autosampler.
1.6.5.6. Stability
standard and sample under specified storage conditions as a function of time. This is
determined by analysing the samples at specific time intervals using the validated
analytical method. The samples and standards are discarded when the analyte peak area
standards at room temperature (25 °C), in a fridge (4 °C) and in a freezer (-20 °C).
28
General introduction
arvense L.
tool in assessing herbal medicine quality.65-67 The pattern-based approach considers all
constituents. The pattern-based approach may be used to determine the similarity of the
rather than the mixture itself because it is difficult to identify the component herb(s) that
from a target profile. It is a reasonable expectation that by using individual herbs that
have a comparable profile to a reference, it is more likely that a mixture will meet its
target profile.
It was originally intended that the pattern-based approach be applied to one of the herbs
samples from different manufacturers of any of the individual herbs used in order to
29
General introduction
sources of the extract of the herb E. arvense, so this herb was used for this pattern-based
In this study E. arvense was used as a model to observe the effect of worldwide
E. arvense was selected because it is distributed worldwide and the Equisetum species
plus its hybrids are widely reported to possess extensive morphological, morphometric
1.7.3. Chemometrics
The pattern-based approach to the assessment of herbal raw material quality using
the starting material make the comparison somewhat ambiguous.65,76 The term
chemometrics has been used since the 1970s to describe the application of mathematical
as principal component analysis (PCA) and hierarchical cluster analysis (HCA) are
order to minimise the ‘garbage in, garbage out’ principle, thereby reducing the
result.79
30
General introduction
The R language for statistical computing (known simply as ‘R’) is a powerful tool for
chemometric analysis.80 One of R’s major advantages is that it is free, with many users
around the world donating add-on packages that can perform a wide array of functions.
chromatographic data. msProcess can remove instrumental noise, baseline drift, peak
retention time variations, identify peaks, and quantify peak height.76 The ‘Stats’
package included in R contains many of the commonly used statistical techniques such
PCA works by explaining a large number of highly correlated variables such as various
principal components (PCs).67 Plotting the samples on the PCs can yield a 1-, 2- or 3-
dimensional plot depending on the number of PCs chosen. Similar samples often group
together in the same area of the scores plot. Plotting the variables on the loadings plot
can identify the variables responsible for the grouping observed in the scores plot.67
HCA-based techniques can be used in conjunction with PCA to aid in the classification
of sample groups.67 K-nearest neighbour clustering (KNN), for example, can be used to
that uses the linear correlation coefficient (r) to compare two chromatographic
profiles based on the ratio of each peak intensity (n) with each of the other (n-1) peak in
31
General introduction
This process was used as a starting point for subsequent PCA analysis. The rationale for
using the ratio information instead of simply the peak intensity values is that the ratio
provides measurement relative to each of the other peaks in the chromatogram. This is
specific variations in the DNA sequences of various chloroplast and nuclear regions.33
PCR-based methods only require minute amounts of DNA and can be applied to fresh
DNA barcoding is a tool that has the potential to allow rapid and unequivocal
sequence of DNA often between 400 and 800 base pairs long. While DNA barcoding
has proven to be considerably more demanding in plants than in animals, where a single
locus has proved sufficient, a concerted effort of scientists worldwide has led to the
identification of four regions that can be used for DNA barcoding of land plants with
reasonable success. The three plastid markers rbcL, matK and intergenic spacer trnH–
psbA have proved to be effective in a wide range of land plants. In addition, the nuclear
ribosomal marker ITS, which has been used for the genomic authentication of medicinal
plants, has also proved suitable for species discrimination in a large number of plants.33
Due to the highly variable nature of the flavonoid and phenyl carboxylic acid content of
32
General introduction
capacity, the radical scavenging capacity can be used as a basic and preliminary
the chemical sense and should not be used in lieu of a cell line-based or in vivo assay.39
detectable by the chemical analysis used. Similarly, extracts with different chemical
profiles may exhibit the same biological activity if the phytochemicals responsible for
The two main methods by which a compound can function as an antioxidant are
hydrogen atom transfer (HAT) and electron transfer (ET).39 This study assesses the
radical scavenging capacity of the E. arvense extracts using both HAT and ET
mechanisms.
HAT reactions such as ORAC are kinetic based methods, whereby fluorescein and the
antioxidant being measured compete for peroxyl radicals generated by the thermal
stronger the antioxidant is, the longer the substrate takes to be degraded.
involve a redox reaction between the DPPH and the antioxidant compound being
measured. DPPH is a rapid and simple antioxidant assay that is commercially available.
In its oxidised form DPPH has an intense purple colour (λmax 515 nm) and when it is
reduced it becomes yellow (λmax 320 nm), the colour change being proportional to the
antioxidant concentration.
33
General introduction
Both the ORAC and DPPH methods use gallic acid as a reference for antioxidant
capacity. That is, these assays measure how much better (or worse) the E. arvense
extracts are at being antioxidants than gallic acid. An on-line LC-PDA DPPH assay was
also used to detect the chemicals that may contribute the bulk antioxidant capacity of
the extract.
1.8. Aims
1. Use a systematic method for the selection of analytes for the quantitative analysis of
Endoherb™.
2. Develop and validate an analytical method for the quantitation of the selected
analytes in the dried Endoherb™ extract prepared from the pure aqueous, 35 %
3. Use the validated analytical method to evaluate the extraction efficiencies of the
6. Use chemical antioxidant assays as a rapid and simple method for assessing the
34
General introduction
materials.
35
CHAPTER 2.
Three powdered extracts of approximately 200 g were produced by The Centre for
extracts were prepared using 95 % aqueous ethanol, 35 % aqueous ethanol, and water as
the extraction solvent to investigate the effect of extraction solvent polarity on the
quantity of the analytes in the final extract. All extracts were stored at room
Each dried plant material was ground to ≤ 500 µm and then mixed in the in the ratio
given in Table 1.1 to produce 2 kg of the mixture. Dried plant material was supplied by
Beijing Tong Ren Tang (Sydney, Australia). The extracts were then prepared as
follows.
For the 95 % aqueous ethanol extract, 2 kg of the herbal mixture was placed in a 50 L
ethanol was added and the mixture allowed to steep with a gentle rocking motion for 24
h. The 20 L of extract was then filtered through 3 × Whatman No. 1 filter papers. The
extract was then concentrated using a 20 L rotary evaporator. The yield for this extract
was lower than the desired 10 : 1 extraction ratio, so another 20 L of 95 % ethanol was
added to the herb residue and steeped for a further 24 h, filtered, concentrated and
combined with the first extract. The extract was then placed on a freeze drier as a final
37
Experimental for Endoherb™
For the 35 % aqueous ethanol extract, the same procedure was followed except that
35 % ethanol was used instead of 95 % ethanol and only one aliquot of extraction
2 kg of the herbal mixture was placed in a 50 L stainless steel rocking extraction vessel.
20 L of water was added and allowed to steep with a heating element inserted into the
extraction vessel. The extract was heated at 65 °C for 2 h with regular stirring. A filter
head was then fitted to the extraction vessel and allowed to rock for 3 h to improve
extraction. The extract was then filtered and dried as for the aqueous ethanol extracts to
2.2. Equipment
Sartorius SE-2 micro analytical balance (Sartorius Australia, Australia) were used to
weigh the samples and standards. A Powersonic 420 ultrasonic bath (Thermoline
The LC-PDA analysis of glycyrrhizic acid, paeoniflorin, and rhein was performed on a
Varian Prostar system comprising of 2 × 210 single pumps, a column valve module 500,
a 430 auto-sampler, and a 335 PDA detector with a ‘9×0 mm’ analytic flow cell (Varian
Inc., Australia). Solvents were degassed using a model AF DG2 in-line degasser
(Waters, USA). The system was controlled using Varian Star Workstation version 6.20.
38
Experimental for Endoherb™
The column was a Luna C18 (150 × 4.6 mm, 5 µm) equipped with a C18 SecurityGuard
The LC-MS analysis of amygdalin and pachymic acid was performed on the same
Varian LC-PDA system except that the column eluent was passed through a splitter
which diverted 20 % of the flow to a Varian 1200L triple quadrupole mass spectrometer
XL EI/CI mass selective detector (MS) and CombiPal autosampler (Agilent, Australia).
The column was a HP-5MS (30 m × 0.25 mm ID, 0.25 µm; J&W scientific, USA). The
GC-FID system was the same as used for the GC-MS, though the column was
2.3. Reagents
Acetonitrile was of LC grade (Mallinckrodt Chemical Ltd., UK). Ethanol (95 %),
methanol and formic acid (90 %) were reagent grade (Biolab, Australia). Air, argon,
helium, hydrogen and nitrogen were of ultra-high purity grade (Coregas, Australia).
Purified water (> 18 MΩ cm) was obtained from an Elga Purelab Prima and Purelab
isomers, 99.4 %), trans-cinnamaldehyde (99.4 %), glycyrrhizic acid (93.4 %) and
39
Experimental for Endoherb™
Atractylenolide III (99.5 %), ligustilide (98.8 %), paeonol (100 %), paeoniflorin
(98.7 %), pachymic acid (97.9 %), rhein (99.7 %) and umbelliferone (100 %) were of
secondary grade (Phytomarker Ltd., Tianjin, China). The primary grade standards have
purity and spectroscopic characterisation; the secondary grade standards have purity by
LC only. The results presented in the results section have been corrected for standard
purity.
Mobile phase A (0.1 % formic acid in water) was prepared by adding 900 mL water to a
1000 mL volumetric flask followed by 1.1 mL formic acid before making up to volume
with water. Mobile phase B was acetonitrile. Mobile phases were degassed by
sonication for 5 min and filtered through a 0.45 µm polyvinylidene difluoride (PVDF)
glycyrrhizic acid and 1000 µg/mL paeoniflorin were prepared by weighing 10.0, 20.0
and 5.0 mg of the respective standard into a 5 mL volumetric flask and adding
solid has dissolved. The solution was allowed to cool before making up to volume with
50 % aqueous methanol.
40
Experimental for Endoherb™
the mixed fortification solution to 1000 µL with 50 % aqueous methanol to give a 20-
fold dilution of the mixed fortification solution. This intermediate mixed standard was
diluted 1, 1/5, 1/10, 1/50 and 1/100 to give the mixed working standard solutions.
A mixed fortification solution containing 3000 µg/mL pachymic acid and 1700 µg/mL
rhein was prepared by weighing 30.0 and 17.0 mg of the respective standards into a 10
for 5 min or until the standards have dissolved. The solution was allowed to cool before
the mixed fortification solution to 1000 µL with 95 % aqueous methanol to give a 20-
fold dilution of the mixed fortification solution. This intermediate mixed standard was
diluted 1, 1/5, 1/10, 1/50 and 1/100 to give the mixed working standard solutions.
A mixed fortification solution containing 4000 µg/mL atractylenolide III, 5500 µg/mL
ligustilide and 18,200 µg/mL paeonol was prepared by adding 20.0 mg, 27.5 mg and
approximately 3 mL ethanol. Ethanol was first placed in the flask because ligustilide is
41
Experimental for Endoherb™
volatile. The flask was loosely stoppered and sonicated until the solid just dissolved
the mixed fortification solution to 1000 µL with 95 % aqueous methanol to give a 40-
fold dilution of the mixed fortification solution. This intermediate mixed standard was
diluted as shown in Table 2.1 to give the mixed working standard solutions.
Table 2.1: Preparation of working standard solutions from the intermediate standard.
fortification solution
prepared by adding 3 µL (or 2.87 mg), 2.5 µL (or 278 mg), 1.0 mg and 4.0 mg of the
95 % ethanol. The flask was loosely stoppered and sonicated for approximately 5 min or
till the solid is just dissolved. After cooling the solution was made up to volume with 95
42
Experimental for Endoherb™
% ethanol. The 95 % ethanol was first added to the flask, as carveol and
the mixed fortification solution to 1000 µL with 95 % aqueous methanol to give a 40-
fold dilution of the mixed fortification solution. This intermediate mixed standard was
diluted as shown in Table 2.1 to give the mixed working standard solutions.
Two different extractions were used, utilising different solvents to selectively extract
particular analytes. For the analysis of amygdalin, glycyrrhizic acid and paeoniflorin,
50 % aqueous methanol was used as the extraction solvent, while 95 % aqueous ethanol
200 mg of the dried extract was weighed into a 25 mL volumetric flask, approximately
20 mL of the extraction solvent added and the mixture sonicated for 30 min. The
solution was allowed to cool before making up to volume with the extraction solvent.
The sample was passed through a 0.45 µm PVDF syringe filter into a 1.5 mL
Separate LC methods were used for the determination of amygdalin, glycyrrhizic acid,
paeoniflorin, pachymic acid and rhein. GC-MS was used for the simultaneous
43
Experimental for Endoherb™
FID was used for the simultaneous determination of atractylenolide III, ligustilide and
paeonol.
maintained for 10 min before changing to 95 % solvent B for 5 min to wash the column
before returning to the initial composition for 5 min to equilibrate the column for the
next analysis. The injection volume was 10 µL. The MS was operated using (-)-ESI
with a drying gas temperature of 350 ºC, capillary voltage of -60 V, needle voltage of -
4850 V, shield voltage of -400 V and a detector voltage of 1500 V. Q1 and Q3 both
operated with a peak width of 3 AMU and a scan time of 2 s. The [M-H]- ion (m/z 456)
was monitored for quantitation and two transitions are monitored for qualification (m/z
The initial mobile phase was 58 % solvent A and 42 % solvent B, maintained for 10
min before changing to 95 % solvent B for 5 min to wash the column before returning
to the initial composition for 5 min to equilibrate the column for the next analysis. The
injection volume was 10 µL. The glycyrrhizic acid peak was monitored at 254 nm.
The initial mobile phase was 85 % solvent A and 15 % solvent B, maintained for 10
min before changing to 95 % solvent B for 5 min to wash the column before returning
to the initial composition for 5 min to equilibrate the column for the next analysis. The
injection volume was 10 µL. The paeoniflorin peak was observed at 230 nm.
44
Experimental for Endoherb™
The mobile phase was 20 % solvent A and 80 % solvent B. The injection volume was
10 µL. The MS was operated using (-)-ESI with a drying gas temperature of 300 ºC,
capillary voltage of -100 V, needle voltage of -5000 V, shield voltage of -275 V and a
detector voltage of 1500 V. Q1 and Q3 both operated with a peak width of 3 AMU and
a scan time of 2 s. The [M-H]- ion (m/z 528) was monitored for quantitation and two
daughter ions were monitored for qualification (m/z 528→465, 46 V and m/z 528→467,
45 V).
The initial mobile phase was 55 % solvent A and 45 % solvent B, maintained for 10
min before changing to 95 % solvent B for 5 min to wash the column before returning
to the initial composition for 5 min to equilibrate the column for the next analysis. The
GC-MS
The autosampler was programmed for a 1 µL injection, pre-cleaning the syringe with
ethanol once, then the sample five times before each injection. The syringe was rinsed
with ethanol five times after each injection. The injector was set at 200 ºC using a split
1.5 mL/min. The initial oven temperature is held at 100 ºC for 1 min and then increased
to 220 ºC at a rate of 10 ºC/min and held at 220 ºC for 2 min. The oven temperature was
then increased to 250 ºC for 2 min after the run to clean the column. The MS transfer
line was maintained at 250 ºC, the EI source at 230 ºC and the quadrupole at 150 ºC.
45
Experimental for Endoherb™
Quantitation was performed in selected ion monitoring (SIM) mode, monitoring the
The operating conditions for the GC, autosampler and column are the same as those
used in Section 45. The FID was set at 300 ºC with a H2 flow of 30 mL/min, air flow of
(approximately 100 µg/mL in methanol) into the ESI interface at 20.0 µL/min. The ESI
parameters optimised include needle voltage, shield voltage, capillary voltage, drying
46
Experimental for Endoherb™
Once the optimum conditions for maximum molecular ion signal were determined,
20 µg mL-1 solution of chemical reference and determining the most intense and highest
47
CHAPTER 3.
3.1. Samples
Pharmaceuticals Ltd (NSW, Australia). The authenticity of the extracts was established
Where the raw material used to produce the extracts were available, these were used for
genomic authentication.
The excipient was removed from the commercial extracts to minimise sample
variability due to different excipients used and their different extract-to-excipient ratios.
Approximately 4 g of each commercial extract was weighed into a 250 mL conical flask
and 250 mL of 80 % aqueous methanol was added. The mixture was sonicated for 1 h
with occasional stirring and then centrifuged at 4000 g for 5 min to pellet out the
insoluble excipient. The supernatant was filtered though a 0.45 µm PVDF syringe filter
and the filtrate evaporated to dryness under vacuum at 60 ºC to remove the methanol
and the residue freeze-dried for 12 h to remove the remaining water. The resultant solid
49
Experimental for E. arvense
3.2.1. TLC
A CAMAG (Muttenz, Switzerland) TLC system equipped with a sample applicator and
visualisation chamber was used with Merck (Darmstadt, Germany) silica gel 60 F254
TLC plates (20 cm × 10 cm) for TLC profiling. The TLC profiling method was from
Wagner et al, using a mobile phase of ethyl acetate : formic acid : glacial acetic acid :
extract and then filtered through a 0.45 µm PVDF syringe filter. 2 µL bands were
To visualise the flavonoid and phenyl carboxylic acid profile, the plate was developed
radical, the plate was developed in DPPH reagent (200 µg/mL in ethanol) and visualised
in white light. Chemicals that scavenge the DPPH radical appear yellow.
A Varian (California, USA) ProStar system equipped with a 430 autosampler, 335
photodiode array detector (PDA) and 1200L quadrupole MS/MS detector was used for
50
Experimental for E. arvense
(150 mm × 4.6 mm, 5 µm) with a Phenomenex (California, USA) Security C18 guard
Working solutions of each extract were prepared by dissolving 50.0 mg of the purified
The LC analysis was carried out using a 10 µL injection volume and a mobile phase
flow rate of 1 mL/min consisting of 0.1 % aqueous formic acid (mobile phase A) and
acetonitrile (mobile phase B). The mobile phase program was 10 % B for 10 min with a
linear increase to 50 % B between 10 - 63 min. The column was then washed with 100
% B for 10 min and equilibrated with starting mobile phase for 10 min between each
analysis.
The eluate was split to send 80 % to the PDA and 20 % to the MS. PDA chromatograms
are acquired at 280 nm. The MS was set to acquire in the (-)-ESI mode, scanning
between 70-700 AMU using a nebulisation gas temperature of 400 ºC at 19 psi, needle
voltage -3900 V at 15 µA, shield voltage -400 V, capillary voltage -100 V, and MS
detector at -1700 V.
Software supplied by the ‘R Project for Statistical Computing’ was used for the data
processing and statistical analysis. Specific packages used with R are detailed as
follows.87
instrumental noise, baseline drift, identifying peaks, removing peak retention time
51
Experimental for E. arvense
variations between samples and to quantify peak height.88 This was done to minimise
PCA was used together with KNN to differentiate samples and highlight the chemical
components potentially responsible these differences using the ‘stats’ package included
with R.81 PCA was first conducted on the corrected chromatograms and the results are
plotted using the first two PCs. KNN was then applied to the first two PCs in order to
highlight samples that cluster together. Three groups were specified for the KNN, based
on the country of origin of the sample: 1) USA, 2) China/Europe and 3) India. The
group-specific peaks and their corresponding UV and MS spectra were compiled and
Using the chromatogram correction technique outlined, the average number of peaks
detected was determined using the usual techniques of TLC, LC-PDA and LC-MS to
estimate their detection power. The determination of the statistical significance (p <
0.05) between the analytical techniques was measured by one-way ANOVA with a
Genomic DNA was extracted and purified from the dried stem using a Qiagen DNeasy
mini plant mini kit (Victoria, Australia) following the manufacturer’s instructions
except that water was used instead of buffer AE. The loci used for genomic
authentication are the chloroplast genes matK (Maturase K) and rbcL (ribulose-1,5-
the Barcode of Life (CBoL).35 For the PCR amplification of matK, the primers
52
Experimental for E. arvense
the Royal Botanic Gardens, Kew.90 For the PCR amplification of rbcL the primers
by CBoL.91 The iProof high-fidelity DNA polymerase PCR kit from Bio-Rad (NSW,
Australia) was used for PCR amplification as per the manufacturer instructions for a
extension 72 ºC, 5 min. The PCR products were purified using the Qiagen QIAquick
PCR Purification Kit according to the manufacturer’s instructions except that water was
used instead of buffer AE. PCR products were sequenced at the Australian Genome
Research Facility Ltd (NSW, Australia). Data processing was performed using the
program Geneious™.92
eluent using a third pump (0.6 mL/min) and reacting the solution in a coil
(5.0 m × 0.5 mm) based on the work by Bandoniene et al, as presented in Figure 3.1.93
The PDA detector was set to monitor at 280 nm for the chromatogram and 515 nm for
53
Experimental for E. arvense
A method adapted from Blois et al and Molyneux et al was used to estimate the DPPH
standard.94,95 All reagents are prepared in 80 % aqueous methanol. The gallic acid
standard curve was made by diluting a gallic acid stock (3 mM) to give 0.3, 0.6, 0.9 and
1.5 mM working standards. Samples are prepared by dissolving 1.0 mg of the extract in
10 mL 80 % aqueous methanol which was also the solvent used as the reagent blank.
180 µL of the DPPH reagent (250 µM) was pipetted into each microtitre plate well
used. In triplicate, 20 µL of each working standard, sample or blank was pipetted into
the DPPH reagent to make a total volume of 200 µL in each well. To correct for sample
absorbance (i.e. absorbance not due to the DPPH), sample blanks were also prepared in
sample. The plate was vortexed at 700 rpm for 30 min in the dark prior to measuring the
absorbance at 515 nm. The sample antioxidant scavenging capacity is reported as the
54
Experimental for E. arvense
The oxygen radical absorbance capacity (ORAC) assay measures the ability of E.
arvense extracts to protect fluorescein from degradation by peroxyl radicals using the
method described in the BMG LABTECH (Ortenberg, Germany) application note 148
using gallic acid as the reference standard.96 All reagents were prepared in pH 7.4
phosphate buffer (10 mM). The gallic acid standard curve was prepared by diluting the
gallic acid stock (200 µM) to give 12.5, 25, 50 and 100 µM working standards. Samples
which was also the solvent used as the reagent blank. 150 µL fluorescein (10 nM) and
25 µL of either gallic acid standard, sample or blank was pipetted into each microtitre
plate well used for analysis. This solution was vortexed for 30 min at 37 ºC before
plate every 90 s (excitation 485 nm, emission 520 nm). The areas under the signal
degradation curves of the samples were compared to the gallic acid standard and the
being active using the LC-DPPH-PDA assay were compared to the literature for
55
CHAPTER 4.
The three Endoherb™ extracts were deliquescent, making accurate weighing and
sampling difficult. The 95 % ethanol extract was the most difficult to work with as it did
not dry to a solid, presumably because it contained oils. Both the 35 % and water
extracts could be dried and powdered for homogeneous sampling when desiccated.
The extraction method used involved soaking the powdered herb in water or aqueous
ethanol and is therefore a relatively mild method of extraction. The extraction method
approximates home extraction where the sliced or ground-up herbs are soaked in spirits.
Home extraction by water usually involves boiling the herb to produce the decoction.
Many solvents were trialled in order to maximise the analyte extraction efficiency, and
methanol or ethanol caused a white cloudy solution to form in the ethanolic extracts of
Endoherb™ instead of the clear coloured liquid obtained through extraction with pure
ethanol. The cloudiness is probably due to non-polar components that are dispersed in
the solvent but are not soluble in it. Consequently two extraction solvents: 50 %
aqueous methanol and 95 % aqueous ethanol were needed to extract the high and low
57
Results and discussion for Endoherb™
The variety of chemical groups and hence chemical properties of the analytes present in
the mixture made their analysis in a single analytical run difficult. Thus several methods
had to be used for their analysis. The techniques of LC-PDA, LC-MS, GC-FID and GC-
chemicals analysed removed the need for long and involved sample preparation
A summary of the methods of analysis and the rationale for their use is presented in
Figure 4.1.
Figure 4.1: The fractionation scheme used for the analysis of the analytes.
Chromatographic methods were developed for all analytes except catalpol. All the
methods showed good resolution of the analyte peak and satisfactory signal-to-noise for
chemical standards and herbal extracts are presented in Figure 4.2 - Figure 4.8.
58
Results and discussion for Endoherb™
Figure 4.2: Representative LC-PDA chromatograms of the glycyrrhizic acid standard (red) and the
Figure 4.3: Representative LC-PDA chromatograms of the paeoniflorin standard (red) and the
59
Results and discussion for Endoherb™
Figure 4.4: Representative LC-PDA chromatograms of the rhein standard (red) and the 95 %
Figure 4.5: Representative LC-MS chromatograms of the amygdalin standard (red) and the 95 %
60
Results and discussion for Endoherb™
Figure 4.6: Representative LC-MS chromatograms of the pachymic acid standard (red) and the
Figure 4.7: Representative GC-FID chromatograms of the mixed standard (red) and the 95 %
61
Results and discussion for Endoherb™
Figure 4.8: Representative GC-MS chromatograms of the mixed standard (red) and the 95 %
MS detection was required for most of the chemicals markers as they were present at
low concentrations in the extract (< 1 mg/g). LC-PDA detection could not detect many
of the analytes even if they had chromophore as it was not specific enough to separate
potentially hundreds of other analytes unless an impractically long run (up to 2 h per
sample) was used. MS can provide a means of specific detection of these analytes even
without complete resolution from other peaks. GC-MS was used for the more volatile
analytes as the significantly decreased peak width compared to LC made their analysis
Catalpol was unable to be measured with the available instrumentation, as it was too
chromophore for sensitive and specific detection using a PDA detector. Catalpol was
also not detected in any of the extracts using the MS detector during method
62
Results and discussion for Endoherb™
development. The result of this is that the herb Shudihuang is not measured in the final
formulation. A possible reason for why catalpol has not been detected is that
Shudihuang is honey baked; therefore the cells may not be being ruptured during the
ethanol extraction solvent. It was also not possible to grind finely due to its stickiness.
Tetramethylpyrazine is below the limit of detection in the extracts tested. Even though
tetramethylpyrazine was not detected, method validation was still conducted on this
It was necessary to use LC-PDA for the analysis glycyrrhizic acid, paeoniflorin and
rhein as matrix effects prevented their accurate quantitation using the ESI source. The
presence of strong chromophores, high concentration, and the robustness of the LC-
PDA system negated the necessity to use LC-MS to analyse these chemicals.
All the methods showed good resolution of the analyte peak with a satisfactory signal-
to-noise and good precision. Method validation was completed on all analytes except
Reasonable identity confirmation was achieved for those analytes analysed by LC-PDA
by comparison of the UV-Vis spectra of the peaks obtained from the sample extract and
standard solutions. Each analyte displayed a characteristic spectrum with a λmax together
with at least one other peak. The sample and standard spectrum show good overlap as
63
Results and discussion for Endoherb™
can be observed in Figure 4.9 - Figure 4.11. Identity confirmation was supported by LC-
Figure 4.9: Comparison between the standard (red) and sample (black) UV spectra of glycyrrhizic
acid.
64
Results and discussion for Endoherb™
Figure 4.10: Comparison between the standard (red) and sample (black) UV spectra of
paeoniflorin.
Figure 4.11: Comparison between the standard (red) and sample (black) UV-Vis spectra of rhein.
Identity confirmation for the chemicals markers analysed by LC-MS and GC-MS was
achieved with greater confidence than by comparison of the UV-Vis spectrum, namely
65
Results and discussion for Endoherb™
by comparing the m/z of the ions obtained, as well as their relative abundances with
those obtained for the standard. Additional confidence is gained if one is able to
rationalise the observed ions to an assumed chemical structure. The relative intensities
for the analytes in the standard and samples fall within acceptable tolerances.64 The MS
66
Analyte Standard Sample Percentage Permitted Pass / Fail Proposed fragmentation pattern
difference (%) tolerance (%)
67
Results and discussion for Endoherb™
Analyte Standard Sample Percentage Permitted Pass / Fail Proposed fragmentation pattern
difference (%) tolerance (%)
68
Results and discussion for Endoherb™
Analyte Standard Sample Percentage Permitted tolerance Pass / Fail Proposed fragmentation pattern
difference (%) (%)
69
Results and discussion for Endoherb™
Results and discussion for Endoherb™
4.3.3.2. Accuracy
The 50, 100 and 200 % spike recovery results for the 35 % ethanol extract of the
Endoherb™ mixture are presented in Table 4.2. Considering all the analytes, the
average recovery was 104.2 % (range 79.0 to 151.9 %) with an average RSD of 10.7 %
Where the percentage recovery decreases with increasing fortification, for example for
umbelliferone, this may be due to the chemical being more difficult to dissolve
rhein, this may be due to a constant loss of chemical, such as by adsorption onto the
surface of glass flasks or the filters used. The loss of a constant amount of analyte is
concentration. This may be due to uncertainties regarding the purity of the standard
purity characterisation of the standard using elemental analysis and thermal analysis of
(for example water) are not detected. The more common explanation for > 100 %
recovery is a coeluting peak, but this is less likely in our case as a MS detector was used
and the ion ratio comparison between sample and standard peaks are a good match.
70
a
Accuracy a
Recovery (%) RSD (%) Recovery (%) RSD (%) Recovery (%) RSD (%) Recovery (%) RSD (%)
Atractylenolide III GC-FID 109.7 7.7 110.7 3.5 113.5 3.3 111.3 5.2
Cinnamaldehyde GC-MS 90.9 11.2 77.4 13.1 75.1 7.5 81.1 13.6
Glycyrrhizic acid LC-PDA 87.9 4.7 96.6 2.8 97.2 1 93.9 5.5
Pachymic acid LC-MS 114.6 1.9 107.1 3.3 101.6 1.2 107.8 5.5
Paeoniflorin LC-PDA 107 4.7 109.3 2.4 109.5 1.3 108.6 3.2
Paeonol GC-FID 101.5 2.6 99.9 2.1 97.6 2.8 99.6 2.9
Rhein LC-PDA 93.6 1.7 94.6 1.5 95.4 1.4 94.5 1.7
Umbelliferone GC-MS 173.3 4.3 141.9 4.3 140.7 3.5 151.9 10.9
71
Results and discussion for Endoherb™
Results and discussion for Endoherb™
4.3.3.3. Precision
Good instrumental and method precision was obtained using all methods of analysis as
Precision a
Analyte Technique
Amount ± SD (mg/g) tR ± SD (min)
Amygdalin LC-MS 0.85 ± 0.03 7.433 ± 0.008
Atractylenolide III GC-FID 1.08 ± 0.05 15.081 ± 0.001
(+)-Carveol GC-MS 0.078 ± 0.005 4.462 ± 0.000
(-)-Carveol GC-MS 0.007 ± 0.001 4.592 ± 0.004
Cinnamaldehyde GC-MS 0.146 ± 0.007 5.074 ± 0.000
Glycyrrhizic acid LC-PDA 1.80 ± 0.03 6.43 ± 0.02
Ligustilide GC-FID 11.2 ± 0.3 11.1622 ± 0.0005
Pachymic acid LC-MS 0.46 ± 0.01 8.33 ± 0.02
Paeoniflorin LC-PDA 14.5 ± 0.1 8.84 ± 0.03
Paeonol GC-FID 8.1 ± 0.1 7.7908 ± 0.0005
Rhein LC-PDA 0.119 ± 0.005 8.23 ± 0.03
b
Tetramethylpyrazine GC-MS 0.003 ± 0.001 3.08 ± 0.01 b
Umbelliferone GC-MS 0.013 ± 0.001 11.561 ± 0.000
a
Average and RSD calculated from n = 7 replicates.
b
Not determined in the unfortified sample, value determined from the 50 % fortification.
The LOQs are sufficiently low to quantify the concentrations of these chemicals that
Table 4.4.
72
Results and discussion for Endoherb™
Table 4.4: Summary of the analytical method detection and quantification limits for the analytes.
Detection limits a
Analyte Technique
LOD (mg/g) LOQ (mg/g)
Amygdalin LC-MS 0.1 0.3
Atractylenolide III GC-FID 0.1 0.5
(+)-Carveol GC-MS 0.02 0.06
(-)-Carveol GC-MS 0.003 0.01
Cinnamaldehyde GC-MS 0.02 0.07
Glycyrrhizic acid LC-PDA 0.09 0.3
Ligustilide GC-FID 1 3.2
Pachymic acid LC-MS 0.04 0.1
Paeoniflorin LC-PDA 0.4 1.5
Paeonol GC-FID 0.3 1
Rhein LC-PDA 0.01 0.05
b
Tetramethylpyrazine GC-MS 0.003 0.009 a
Umbelliferone GC-MS 0.003 0.012
a
Average and RSD calculated from n = 7 replicates.
b
Not determined in the unfortified sample, value determined from the 50 % fortification.
4.3.3.5. Linearity
The calibration curves show good linearity with correlation coefficients (r2) > 0.995 as
73
Results and discussion for Endoherb™
Table 4.5: Summary of the calibration curve linearity for the analytes.
Linearity
Analyte Technique
Range (µg/mL) r2
Amygdalin LC-MS 1 - 100 0.996
Atractylenolide III GC-FID 10.5 - 105 0.9998
(+)-Carveol GC-MS 0.4 - 4 0.9995
(-)-Carveol GC-MS 0.4 – 4 0.9994
Cinnamaldehyde GC-MS 0.12 – 12 0.996
Glycyrrhizic acid LC-PDA 2 - 211 0.9992
Ligustilide GC-FID 12 - 120 0.9998
Pachymic acid LC-MS 1.5 - 150 0.998
Paeoniflorin LC-PDA 10 - 1000 0.9999
Paeonol GC-FID 45 - 450 0.9998
Rhein LC-PDA 0.8 - 42 0.9994
Tetramethylpyrazine GC-MS 0.25 – 2.5 0.9995
Umbelliferone GC-MS 1 – 10 0.998
4.3.3.6. Stability
For accurate quantitation, standards and samples need to be prepared freshly on the day
4.6 and Figure 4.12. The 95 % ethanolic extract contained the highest concentrations of
the analytes, followed by the 35 % ethanolic extract, and lastly the aqueous extract as
74
Results and discussion for Endoherb™
Table 4.6: Summary of the results obtained for each of the Endoherb™ extracts tested
Amount ± SD (mg/g)
Analyte Technique Aqueous 35 % ethanol 95 % ethanol
extract extract extract
Amygdalin LC-MS 0.298 ± 0.002 0.85 ± 0.03 2.078 ± 0.1
Atractylenolide III GC-FID < LOD 1.08 ± 0.05 2.1 ± 0.3
(+)-Carveol GC-MS < LOD 0.078 ± 0.005 0.2021 ± 0.0002
(-)-Carveol GC-MS < LOD 0.007 ± 0.001 0.03 ± 0.03
Cinnamaldehyde GC-MS < LOD 0.146 ± 0.007 0.512 ± 0.07
Glycyrrhizic acid LC-PDA 0.6 ± 0.1 1.80 ± 0.03 4.0 ± 0.2
Ligustilide GC-FID < LOD 11.2 ± 0.3 8.4 ± 1.5
Pachymic acid LC-MS < LOD 0.46 ± 0.01 0.50 ± 0.02
Paeoniflorin LC-PDA 11.6 ± 0.1 14.5 ± 0.1 36 ± 1
Paeonol GC-FID < LOD 8.1 ± 0.1 11.4 ± 0.6
Rhein LC-PDA < LOQ 0.119 ± 0.005 0.14 ± 0.02
Tetramethylpyrazine GC-MS < LOD < LOD < LOD
Umbelliferone GC-MS 0.05 ± 0.05 0.013 ± 0.001 0.0242 ± 0.0001
75
Results and discussion for Endoherb™
Figure 4.12: A comparison of the amount of chemical determined in each of the different extracts.
76
CHAPTER 5.
obtained using TLC, LC-PDA and LC-MS, especially in relation to phenyl carboxylic
The flavonoid and phenyl carboxylic acid TLC profile, while being useful as a rapid
profiling technique, was insufficient at revealing the complex nature of the chemical
variation between samples, as presented in Figure 5.1. The TLC profile was on average
able to resolve 9 ± 3 peaks, however only a single peak was detected in the India 13
sample. The TLC profile was however sufficient to indicate a general quantitative
three general groups: 1) the USA extracts 2) the European and Chinese extracts 3) the
Indian extract. Due to the poor sensitivity of the TLC technique, it is difficult to
78
Results and discussion for E. arvense
Figure 5.1: Chromatographic characterisation of the E. arvense extracts using TLC stained with
Natural Products and Polyethylene Glycol reagent view under 366 nm UV light.
number of peaks contained in the TLC profile, as presented in Figure 5.2. The same
general trend in phenyl carboxylic acid and flavonoid concentration in the extracts from
different was similar to that in the TLC. The LC-PDA profile has the added benefit of
acquiring the UV-Vis spectrum of each peak, which can be useful in downstream
process such as structural elucidation. Clear qualitative differences also exist between
the samples, especially in regards to the Indian sample, detectable due to the increased
79
Results and discussion for E. arvense
Figure 5.2: Chromatographic characterisation of the E. arvense extracts using LC-PDA viewed at
The LC-MS technique detected on average 43 ± 8 peaks, displaying both qualitative and
quantitative differences between the extracts as shown in Figure 5.3. The same general
trend in phenyl carboxylic acid and flavonoid concentration between countries of origin
80
Results and discussion for E. arvense
Figure 5.3: Chromatographic characterisation of the E. arvense extracts using LC-MS. (Note: The
81
Results and discussion for E. arvense
Figure 5.4: The number of peaks detected in the TLC, LC-PDA and LC-MS chromatograms using
the msProcess peak detection software. (Note: * represents a statistical significance of p < 0.05).
The detection of peaks in a chromatogram is crucial for both qualitative and quantitative
analyses, because the amount of information increases as more peaks are detected.97 The
where TLC is the profiling method of choice for herbal authentication. TLC is
straightforward to perform but the information it provides is somewhat limited. The LC-
MS is much more informative but the cost of the instrumentation is more than double
that of the TLC and involves much more maintenance and is more costly to operate.
The package msProcess was successfully used to remove instrumental noise, remove
baseline drift, identify peaks, and remove peak retention time variations to accurately
82
Results and discussion for E. arvense
quantify peak height. Although it is desirable to use peak area for accurate quantitation
using LC, peak height was used due to a limitation of the software. Satisfactory
identification of peaks was obtained using a trial and error method, resulting in the
Table 5.1: The parameters used for chromatographic processing and a brief outline of their
function.
msDetrend span 0.1 The fraction of the observations in the span of the
running lines smoother.
Figure 5.5.
83
Results and discussion for E. arvense
Figure 5.5: A representative chromatogram of the China 8 sample indicating various corrections
A) The original chromatogram. B) The black line represents the noise detected in the
chromatogram. The red line represents the average. C) The black line represents the de-noised
chromatogram and the red line represents the detected baseline drift. D) The black line represents
the chromatogram with the baseline drift removed; the red circles represent the detected peaks. E)
Once each individual chromatogram is processed, they are compared to each other to
84
Results and discussion for E. arvense
presented in Figure 5.6. From this, a table of intensity values was produced that was
used for the subsequent statistical analysis. Ultimately, 107 intensity values were
(Note: Red circles represent where a peak is detected in each chromatogram. Where that peak is
identified in most chromatograms, the peak height is measured in all chromatograms at the
The multivariate statistical techniques of PCA and KNN were successfully used to
identify differences and similarities between the different E. arvense extracts based on
peak intensity. KNN was used to identify grouping obtained using PCA, by highlighting
samples that were classified into the 3 groups. PCA was also used to determine which
85
Results and discussion for E. arvense
chromatographic peaks and therefore phytochemicals are responsible for the observed
differentiation.
As presented in Figure 5.7, grouping in the scores plot (left) is observed, which is
consistent with the country of origin, that is: USA (red), China / Europe (blue), and
India (green). KNN clustering represented by coloured circles proved very effective in
loadings plot (right) to the sample groups in the scores plot, the peaks responsible for
sample differentiation were identified. Three representative peaks have been highlighted
in the same colours as the sample groups to indicate the groups they affect. Chicoric
example of how this peak is detected only in the India 13 sample by visual examination,
and confirmed as a differentiating variable in the PCA. Other chemicals that have been
tentatively identified in the E. arvense extracts based on comparison of LC, UV-Vis and
86
Results and discussion for E. arvense
Figure 5.7: Principal component analysis (PCA) of LC-MS chromatographic peaks identified using
msProcess.
The colour and ellipses on the scores plots denote grouping obtained from KNN clustering using 3
specified clusters. The proportion of variance encompassed by each PC is given in brackets. The
scores plot (left) is based on the absolute amplitude of all 107 detected peaks, showing that the
geographical origin of the extracts is primarily associated with the 3 specified groups (red = USA,
blue = Europe and China, green = India). The loadings plot (right) highlights peaks that are
representative of the grouping observed in the scores plot (3-hydroxyflavone for USA, methyl
caffeoylquinic acid for Europe and China, chicoric acid for India).
As presented in Figure 5.8, the PCA based on the intensity ratios (PSI) showed better
delineation between the USA and European extracts. The downside to this method is
many more data points are produced (left) which make peak identification difficult.
87
Results and discussion for E. arvense
Figure 5.8: Principal component analysis (PCA) of LC-MS chromatographic peaks identified using
msProcess.
The colour and ellipses on the scores plots denote grouping obtained from KNN using 3 specified
clusters. The proportion of variance encompassed by each PC is given in brackets. The scores plot
(left) is of the 5,671 peaks intensity ratios obtained from the 107 detected peaks using the rational of
Tilton et al showing the geographical origin of the extracts is primarily associated with the 3
specified groups (red = USA, blue = Europe and China, green = India). The loadings plot (right)
Both the matK and rbcL loci were successfully used to authenticate the representative
China, Europe and India samples with reasonable certainly. The matK locus was better
at differentiating E. arvense from the other Equisetum species than rbcL, with a BLAST
search of GenBank® yielding between 97.3 – 99.9 % (India and Europe respectively)
identical sites to the E. arvense database entries using the matK products compared to
98.9 – 100 % (Europe and India respectively) for rbcL. Although the percentage match
using rbcL is higher, the percentages are equally shared with other Equisetum species,
88
Results and discussion for E. arvense
for example India 13 shared the 100 % match with both E. fluviatile and E. diffusum.
Numerous single nucleotide polymorphisms (SNPs) are present in the matK sequence
for the India 13 sample, including an insertion between 465-472 bp not present in any
other GenBank® entries, which may indicate it is a different species yet to be published
in GenBank®. Nucleotide alignments of the China 8, Europe 11 and India 13 matK and
rbcL sequences against other species in the GenBank® database are presented in Figure
89
Results and discussion for E. arvense
Figure 5.9: matK DNA barcodes of the original plant material used to produce the China 8, Europe
11 and India 13 extracts compared to other Equisetum species entries in the GenBank® database.
(Note: Differences between the sequences are marked with a coloured box (red = A, green = T, blue
= C, yellow = G).
90
Results and discussion for E. arvense
Figure 5.10: rbcL DNA barcodes of the original plant material used to produce the Europe 11 and
India 13 extracts compared to other Equisetum species entries in the GenBank® database. (Note:
(Note: Differences between the sequences are marked with a coloured box (red = A, green = T, blue
= C, yellow = G).
91
Results and discussion for E. arvense
Assessment of the total radical scavenging capacity of the extracts was measured using
two different antioxidant techniques to account for the different kinetic models involved
the different E. arvense extracts. Generally speaking, the ORAC and DPPH results are
comparable, indicating that the flavonoids and phenyl carboxylic acids can function in
both the HAT and ET mechanisms. The China 8 and USA 7 samples showed the
highest antioxidant capacity of the extracts. This is contrary to what is implied by the
phytochemical profiling, which indicates that the China and European extracts are
Figure 5.11: The antioxidant activity of the various E. arvense extracts using the DPPH and ORAC
assays.
92
Results and discussion for E. arvense
The DPPH-based antioxidant assay was very useful for the assessment of antioxidant
activity due primarily to the rapidity of the reaction, as well as being performed in a
solvent compatible with the herbal extracts. The major downside of this assay is that the
Due to the rapidity of the reaction, DPPH is useful as a TLC stain. When in the presence
presented in Figure 5.12, the Chinese and European extracts contain approximately 5
Figure 5.12: The antioxidant activity of the various E. arvense extracts as profiled using TLC
developed in DPPH reagent, viewed under white light. (Note: Pink/purple regions are unreacted
DPPH, lighter regions are where the DPPH radical has been scavenged by an antioxidant).
The online LC-DPPH-PDA method was adapted in order to determine the analytes
potentially responsible for the antioxidant activity of the extracts. The preliminary
93
Results and discussion for E. arvense
results in tandem with the tentative structural elucidation provide a useful method to
presented in Figure 5.13, peaks at 280 nm (similar to that in Figure 5.2) that have a
Figure 5.13: A representative chromatogram of the China 8 sample using the on-line DPPH assay.
(Note: The chromatogram at 280 nm is overlayed with the DPPH absorbance at 515 nm. Analytes
that have DPPH antioxidant activity are observed as a negative peak at 515 nm).
The retention time, as well as the UV-Vis and MS spectra of the peaks observed in the
LC-PDA and LC-MS chromatograms respectively was compared to the literature for
94
Results and discussion for E. arvense
Table 5.2: The tentative structural elucidation of several chemical constituents contained in the E.
95
Results and discussion for E. arvense
Figure 5.14: Tentative spectral structural elucidation of dicaffeoyltartaric acid using (A) LC-MS
96
Results and discussion for E. arvense
Figure 5.15: Tentative spectral structural elucidation of genkwanin acetylglucoside using (A) LC-
97
CHAPTER 6.
GENERAL CONCLUSIONS
General conclusions
6.1. Summary
The past decade has seen an unprecedented growth in the popularity of complementary
continues to grow, serious concerns have been raised about their quality and safety. HM
quality control is a significant challenge due to the large assortment of chemicals found
applying quality control to multi-herb medicines increases with each additional herb.
This study shows how quality control of a complex herbal mixture may be carried out
and how the quality and identity of the single herb may be established.
analysis for the separation of the complex assortment of chemicals contained in herbal
and hence product consistency. In contrast, the pattern-based method compares the
chromatographic profile of the extract of a new batch with that of a target or reference
characterisation of plants.
99
General conclusions
The complex nature of Endoherb™ made it an exemplar model for assessing the
compound-based approach to herbal medicine QA. The methodology used in this study
has demonstrated how the systematic QC of a complex herbal mixture can be carried
out. The process firstly involves applying a logical and systematic method for the
selection of the analytes, and then developing a method for their sensitive, specific and
6.2.1. A systematic method for the selection of analytes for the quantitative analysis of
Endoherb™
A major contribution of this work is the logical and systematic process that has been
applied for the careful selection of analytes relevant to the safety and efficacy of the
Endoherb™ formulation. The need for such a process is self evident, as a typical multi-
herb formula, generally containing between 3 to 20 herbs, could contain many hundreds
of analytes, therefore making the total chemical analysis impractical. With no currently
established and accepted criteria for the selection of marker chemicals in the QA
foundation on which some form of informed decision can be made. Accurate chemical
100
General conclusions
6.2.2. A validated analytical method for the quantitation of the selected analytes in the
dried Endoherb™ extract prepared from the pure aqueous, 35 % aqueous ethanol
analytical method used. Considerable time was invested in ensuring that the
performance of the developed method was well characterised with regards to specificity,
The diverse chemical properties of the selected analytes made it difficult to analyse
instead used for the specific analysis of these chemical components based on factors
such as their polarity, volatility and presence of a chromophore. LC-PDA, LC-MS, GC-
The sample preparation is rapid and simple, utilising sonication as the extraction
differences in the chemistry of the analytes made it necessary to use two extraction
The results of the extraction solvent comparison showed that increasing the organic
modifier in the extraction solvent, even for predominantly polar analytes, could increase
of preparation used in the TCM industry, which predominantly use aqueous extraction
solvents. It is therefore likely that most TCM preparations in the market today have only
101
General conclusions
a small amount the active components extracted, with the rest of the herb mass left over
going to waste. As discussed in the ‘future work’ section, these results indicate that an
investigation into the extraction efficiency of the individual raw herb should be carried
out, as it is possible that the starting material does not contain the target analyte. This
would require that each individual herb in the formulation be analysed prior to
extraction.
Characterisation of herbal authenticity and quality is one of the major challenges facing
the herbal medicine industry, a task made difficult due to fundamental challenges in
species delineation. The pattern-based approach to quality evaluation, namely the use of
of the fingerprinting technique is that it does not have the resolving power to
instrument for profiling than TLC, the significant increase in information content will
make this a worthwhile addition to herbal chemical profiling. Limitations of the LC-
102
General conclusions
MS, for example the ionisation source, still limit the chemicals that can be detected with
good sensitivity, for example large complex molecules. Other spectral characterisation
LC-MS, can make the detection of peaks cumbersome and the statistical inferences
inaccurate. The package msProcess was successfully used to remove instrumental noise,
remove baseline drift, identify peaks, and remove peak retention time variations to
PCA is a well-established statistical method for the analysis of complex data. Similar
samples tend to form clusters on the scores plot whereas dissimilar samples will be
found at greater distances, with the loadings plot indicating the original variables
responsible for the similarity or dissimilarity. The novel use of KNN combined with
PCA allows for a more objective classification of sample grouping, classifying extracts
with minimal subjectivity. As with all statistical analysis, the overall reliability of the
result is dependent on the sample size. For a large QC laboratory, the potentially
hundreds of batches of product will make this statistical analysis more appropriate.
Similarly, more complex statistical analyses such as soft independent modelling of class
103
General conclusions
6.3.3. Chemical antioxidant assays as a rapid and simple method for assessing the
biologically useful way, albeit fairly basic and preliminary, to establish quality
that the implied biological effects of the extracts are not reflected in their chemical
profile. This is useful knowledge for industry, where extracts are commonly
relates to this case, it appears that pharmacologically active chemicals may not be
detectable by the chemical analyses used or that many of the analytes detected in the
materials
genomic methods to achieve species delineation. Although there are a large number of
used the proposed standardised DNA loci of rbcL and matK to confirm the identity of
the E. arvense raw materials. As DNA barcoding becomes more standardised and
The work reported in this thesis has considerable scope for expansion.
104
General conclusions
With respect to the analysis of the Endoherb™ formulation, the proposed method
should be subject to inter-laboratory testing and used to analyse more extracts from
and catalpol were unable to be measured in final formulation, the starting material
should be analysed prior to being used in the formulation to assure that analyte is
present. The accuracy of the umbelliferone result requires significant improvement. The
excessive peak tailing due to its polarity. A trimethysilyl derivative using for example
With respect to the analysis of E. arvense, many of the results presented here are
chemistry QC lab.
Finally, the techniques used here need to be communicated to industry and the scientific
105
REFERENCES
References
1 Li, G., Duke, C. & B, R. The quality and safety of traditional Chinese
medicines. Aust Prescr 26, 128-130 (2003).
2 Foster, S., Yue, C.-X. & Chongxi, Y. Herbal Emissaries - Bringing Chinese
Herbs to the West: A Guide to Gardening, Herbal Wisdom and Well-being.
(Inner Traditions Bear and Company, 1992).
5 Economic and Social Council. Economic and Social Council Panel Discussion:
Potential of Traditional Medicine should be Fostered, 2009).
7 Liang, Y.-Z., Xie, P. & Chan, K. Quality control of herbal medicines. Journal of
Chromatography, B: Analytical Technologies in the Biomedical and Life
Sciences 812, 53-70, doi:10.1016/j.jchromb.2004.08.041 (2004).
107
References
12 Bensky, D., Clavey, S., Stöger, E. & Gamble, A. Chinese Herbal Medicine
Materia Medica. 3 edn, (Eastland Press, Inc., 2004).
17 Wu, K. M., Farrelly, J. G., Upton, R. & Chen, J. Complexities of the herbal
nomenclature system in traditional Chinese medicine (TCM): lessons learned
from the misuse of Aristolochia-related species and the importance of the
pharmaceutical name during botanical drug product development.
Phytomedicine 14, 273-279, doi:10.1016/j.phymed.2006.05.009 (2007).
108
References
19 Tilton, R., Paiva, A. A., Guan, J. Q., Marathe, R., Jiang, Z., van Eyndhoven, W.,
Bjoraker, J., Prusoff, Z., Wang, H., Liu, S. H. & Cheng, Y. C. A comprehensive
platform for quality control of botanical drugs (PhytomicsQC): a case study of
Huangqin Tang (HQT) and PHY906. Chin Med 5, 30, doi:10.1186/1749-8546-
5-30 (2010).
21 Zhao, Z. Z., Yuen, J. P. S., Wu, J. L., Yu, T. & Huang, W. H. A systematic
study on confused species of chinese materia medica in the Hong Kong market.
Ann Acad Med Singap 35, 764-769 (2006).
109
References
30 Li, S., Han, Q., Qiao, C., Song, J., Lung Cheng, C. & Xu, H. Chemical markers
for the quality control of herbal medicines: an overview. Chin Med 3, 7,
doi:10.1186/1749-8546-3-7 (2008).
31 Liang, X.-m., Jin, Y., Feng, J.-t. & Ke, Y.-x. in Traditional Herbal Medicine
Research Methods 139-223 (John Wiley & Sons, Inc., 2011).
32 Yao, H., Shi, P., Shao, Q. & Fan, X. Chemical fingerprinting and quantitative
analysis of a Panax notoginseng preparation using HPLC-UV and HPLC-MS.
Chin Med 6, 9, doi:10.1186/1749-8546-6-9 (2011).
110
References
39 Huang, D., Ou, B. & Prior, R. L. The chemistry behind antioxidant capacity
assays. J Agric Food Chem 53, 1841-1856, doi:10.1021/jf030723c (2005).
40 Ames, B. N., Gold, L. S. & Willett, W. C. The causes and prevention of cancer.
Proc Natl Acad Sci U S A 92, 5258-5265 (1995).
111
References
49 Murray, C., Bensoussan, A., Lee, S., Khoo, C. & Hennell, J. Quality control of
herbal medicines: Criteria for the selection of marker compounds. In progress
(2011).
112
References
51 Wagner, H., Bauer, R., Peigen, X., Jianming, C. & Resch, M. Rhizoma
Atractylodis macrocephalae. Vol. 2 (Verlag für Ganzheitliche Medizin Dr. Erich
Wühr GmbH, 1999).
52 Wagner, H., Bauer, R., Peigen, X., Jianming, C. & Bächer, S. Radix Paeoniae
albae / rubrae. Vol. 5 (Verlag für Ganzheitliche Medizin Dr. Erich Wühr
GmbH, 2004).
53 Wagner, H., Bauer, R., Peigen, X. & Jianming, C. Radix Ligustici chuanxiong.
Vol. 3 (Verlag für Ganzheitliche Medizin Dr. Erich Wühr GmbH, 2001).
56 Wagner, H., Bauer, R., Peigen, X. & Jianming, C. Radix Angelicae sinensis.
Vol. 3 (Verlag für Ganzheitliche Medizin Dr. Erich Wühr GmbH, 2001).
58 Gapter, L., Wang, Z., Glinski, J. & Ng, K.-y. Induction of apoptosis in prostate
cancer cells by pachymic acid from Poria cocos. Biochemical and Biophysical
Research Communications 332, 1153-1161 (2005).
113
References
62 Wagner, H., Bauer, R., Peigen, X., Jianming, C. & Nenninger, A. Radix
Rehmanniae. Vol. 1 (Verlag für Ganzheitliche Medizin Dr. Erich Wühr GmbH,
1996).
66 Zeng, Z., Chau, F. T., Chan, H. Y., Cheung, C. Y., Lau, T. Y., Wei, S., Mok, D.
K., Chan, C. O. & Liang, Y. Recent advances in the compound-oriented and
pattern-oriented approaches to the quality control of herbal medicines. Chin Med
3, 9, doi:10.1186/1749-8546-3-9 (2008).
67 Kong, W. J., Zhao, Y. L., Xiao, X. H., Jin, C. & Li, Z. L. Quantitative and
chemical fingerprint analysis for quality control of rhizoma Coptidischinensis
based on UPLC-PAD combined with chemometrics methods. Phytomedicine 16,
950-959, doi:10.1016/j.phymed.2009.03.016 (2009).
114
References
68 Veit, M., Beckert, C., Hohne, C., Bauer, K. & Geiger, H. Interspecific and
Intraspecific Variation of Phenolics in the Genus Equisetum Subgenus
Equisetum. Phytochemistry 38, 881-891 (1995).
71 Lemus, I., Garcia, R., Erazo, S., Pena, R., Parada, M. & Fuenzalida, M. Diuretic
activity of an Equisetum bogotense tea (Platero herb): evaluation in healthy
volunteers. J Ethnopharmacol 54, 55-58 (1996).
73 Do Monte, F. H., dos Santos, J. G., Jr., Russi, M., Lanziotti, V. M., Leal, L. K.
& Cunha, G. M. Antinociceptive and anti-inflammatory properties of the
hydroalcoholic extract of stems from Equisetum arvense L. in mice. Pharmacol
Res 49, 239-243 (2004).
74 Oh, H., Kim, D. H., Cho, J. H. & Kim, Y. C. Hepatoprotective and free radical
scavenging activities of phenolic petrosins and flavonoids isolated from
Equisetum arvense. J Ethnopharmacol 95, 421-424,
doi:10.1016/j.jep.2004.08.015 (2004).
75 Beckert, C., Horn, C., Schnitzler, J. P., Lehning, A., Heller, W. & Veit, M.
Styrylpyrone biosynthesis in Equisetum arvense. Phytochemistry 44, 275-283
(1997).
115
References
76 Guo, F. Q., Liang, Y. Z., Xu, C. J., Huang, L. F. & Li, X. N. Comparison of the
volatile constituents of Artemisia capillaris from different locations by gas
chromatography-mass spectrometry and projection method. J Chromatogr A
1054, 73-79, doi:Doi 10.1016/J.Chroma.2004.08.122 (2004).
85 Apak, R., Guclu, K., Demirata, B., Ozyurek, M., Celik, S. E., Bektasoglu, B.,
Berker, K. I. & Ozyurt, D. Comparative evaluation of various total antioxidant
116
References
95 Molyneux, P. The use of the stable free radical diphenylpicryl- hydrazyl (DPPH)
for estimating antioxidant activity. Songklanakarin J. Sci. Technol. 26, 211-219
(2004).
117
References
99 Amaral, J. S., Ferreres, F., Andrade, P. B., Valentao, P., Pinheiro, C., Santos, A.
& Seabra, R. Phenolic profile of hazelnut (Corylus avellana L.) leaves [of]
cultivars grown in Portugal. Nat. Prod. Res. 19, 157-163,
doi:10.1080/14786410410001704778 (2005).
101 Veit, M., Geiger, H., Czygan, F.-C. & Markham, K. R. Malonylated flavone 5-
O-glucosides in the barren sprouts of Equisetum arvense. Phytochemistry 29,
2555-2560 (1990).
102 Plazonic, A., Bucar, F., Males, Z., Mornar, A., Nigovic, B. & Kujundzic, N.
Identification and quantification of flavonoids and phenolic acids in burr parsley
(Caucalis platycarpos L.), using high-performance liquid chromatography with
diode array detection and electrospray ionization mass spectrometry. Molecules
14, 2466-2490 (2009).
118
References
104 Veit, M., Strack, D., Czygan, F. C., Wray, V. & Witte, L. Di-E-caffeoyl-meso-
tartaric acid in the barren sprouts of Equisetum arvense. Phytochemistry 30, 527-
529 (1991).
119