new capillary electrophoresis methods for the … keyon... · 2014. 12. 17. · bolch, c.j. and...
TRANSCRIPT
NEW CAPILLARY ELECTROPHORESIS
METHODS FOR THE ANALYSIS OF
PARALYTIC SHELLFISH POISONING
TOXINS
By
Aemi Syazwani Abdul Keyon
M.Sc. (Chemistry)
School of Physical Science
Submitted in fulfilment of the requirements for the Degree of
Doctor of Philosophy
University of Tasmania (November, 2014)
i
Declaration
This thesis contains no material which has been accepted for a degree or diploma by
the University or any other institution, except by way of background information and
duly acknowledged in the thesis, and to the best of my knowledge and belief no
material previously published or written by another person except where due
acknowledgement is made in the text of the thesis, nor does the thesis contain any
material that infringes copyright.
The publishers of the papers in this thesis (comprising Chapter 2 to 4) hold the
copyright for that content, and access to the material should be sought from the
respective journals. The remaining non published content of the thesis may be made
available for loan and limited copying and communication in accordance with the
Copyright Act 1968
Aemi Syazwani Abdul Keyon
November 2014
ii
Acknowledgement
In the name of Allah, the Most Gracious and the Most Merciful, all praises to Allah for
the strengths and His blessing in completing this PhD thesis. Sincere appreciation goes
to my PhD supervisors, Prof. Michael Breadmore, Dr. Rosanne Guijt and Dr.
Christopher Bolch for their valuable supervisions, positive encouragement and
constructive advice. They were the ‘motivators’ for guiding me in completing my PhD
study and the ‘teachers’ of the most precious possession in the world, knowledge. I
would like to extend my heartfelt gratitude to all members of Australian Centre for
Research in Separation Science (ACROSS) for all the time they spent in helping me with
my research as well as their kindness and friendship. My special appreciation goes to
these ACROSS friends; Aliaa, Chiing and Petr (the microfluidic experts), Leila and
Hong Heng (the home built electrophoresis system experts), Ryan and Sui Ching (the
Beckman CE experts), Soo Hyun, Tom, Jason, Mohammad and the list goes on. They
have restlessly helped me, as well as giving technical and moral support whenever I need
them. My acknowledgement also goes to these UTAS Central Science Laboratory staff:
Mr. John Davis, Mr. Paul Waller and Mr Chris Young, who enabled my research
experiments through construction of electrical and mechanical hardware, as well as Dr.
Noel Davies and Dr. David Nichols for sharing their MS knowledge. To my husband
Burhanuddin Mohamad, thank you for your utmost moral support, sacrifices and
sometimes technical assistance. It has been a lot of joys, tears, sweat, and very late night
in the lab for the past three years pursuing this dream. I am really grateful to all my
family members (Mama, Papa, Ibu, Adik) for always being there whenever I need their
encouragement. Last but not least, my deepest gratitude goes to Ministry of Education
Malaysia (MOE) and Universiti Teknologi Malaysia (UTM) for scholarship and study
leave opportunities. I am bringing back valuable knowledge to Malaysia so that I can
spread it out to others. Thank you so much.
iii
Statement of co-authorship
The following people and institutions contributed to the publication of work undertaken
as part of this thesis:
Author details and their roles:
Paper 1, < Keyon, A.S.A., Guijt, R.M., Gaspar, A., Kazarian, A.A., Nesterenko, P.N.,
Bolch, C.J. and Breadmore, M.C. Capillary Electrophoresis for the Analysis of
Paralytic Shellfish Poisoning Toxins in Shellfish: Comparison of Detection
Methods, Electrophoresis, 35, (2014) 1496-1503>:
This paper comprises the majority of Chapter 2
Aemi Abdul Keyon was the primary author (70%) and conducted all the experiments,
analysed data and wrote the manuscript. The co-authors contributed a total of 30% to
the published work. Michael Breadmore, Rosanne Guijt and Chris Bolch contributed to
idea, its formalisation and development. Andras Gaspar, Artaches Kazarian and Pavel
Nesterenko offered experimental assistance. All co-authors assisted with refinement and
presentation.
Paper 2, < Keyon, A.S.A., Guijt, R.M., Bolch, C.J. and Breadmore, M.C. Transient
Isotachophoresis-Capillary Zone Electrophoresis with Contactless Conductivity
and Ultraviolet Detection for the Analysis of Paralytic Shellfish Toxins in Mussel
Sample, Journal of Chromatography A, (2014) 1364, 295-302 >:
This paper comprises the majority of Chapter 3
Aemi Abdul Keyon was the primary author (75%) and conducted all the experiments,
analysed data and wrote the manuscript. The co-authors contributed a total of 25% to
the published work. Michael Breadmore, Rosanne Guijt and Chris Bolch contributed to
the concept, development, refinement and presentation.
iv
Paper 3, < Keyon, A.S.A., Guijt, R.M., Bolch, C.J. and Breadmore, M.C., Droplet
Microfluidics for Post-Column Reactions In Capillary Electrophoresis, Analytical
Chemistry (2014), Article in press, DOI: 10.1021/ac5033963>:
This paper comprises the majority of Chapter 4
Aemi Abdul Keyon was the primary author (75%) and conducted all the experiments,
interpreted the results and wrote the manuscript. The co-authors contributed a total of
25% to work. Michael Breadmore and Rosanne Guijt contributed to the idea, its
formalisation, development and presentation. Chris Bolch assisted with refinement and
presentation.
We the undersigned agree with the above stated “proportion of work undertaken” for
each of the above published (or submitted) peer-reviewed manuscripts contributing to
this thesis:
Signed: __________________ ______________________
Prof. Michael C. Breadmore Prof. John Dickey
Supervisor Head of School
School of Physical Science School of Physical Science
University of Tasmania University of Tasmania
Date: _____________________
v
List of publications and presentations
Parts of research works described in this thesis have been or will be reported in the
following publications and presentations:
1. Keyon, A.S.A., Guijt, R.M., Gaspar, A., Kazarian, A.A., Nesterenko, P.N.,
Bolch, C.J. and Breadmore, M.C. Capillary Electrophoresis for the
Analysis of Paralytic Shellfish Poisoning Toxins in Shellfish: Comparison
of Detection Methods, Electrophoresis, 35, (2014) 1496-1503. DOI:
10.1002/elps.201300353. (Chapter 2)
2. Keyon, A.S.A., Guijt, R.M., Bolch, C.J. and Breadmore, M.C., Transient
Isotachophoresis-Capillary Zone Electrophoresis with Contactless
Conductivity and Ultraviolet Detection for the Analysis of Paralytic
Shellfish Toxins in Mussel Sample. J. Chromatogr. A, (2014) 1364, 295-302.
DOI: 10.1016/j.chroma.2014.08.074. (Chapter 3)
3. Keyon, A.S.A., Guijt, R.M., Bolch, C.J. and Breadmore, M.C., Droplet
Microfluidics for Post-Column Reactions in Capillary Electrophoresis,
Analytical Chemistry (2014), Article in press, DOI: 10.1021/ac5033963.
(Chapter 4)
4. Keyon, A.S.A., Guijt, R.M., Bolch, C.J. and Breadmore, M.C., Liquid
Chromatography and Capillary Electrophoresis for The Analysis of
Major Phycotoxins: A Review (manuscript in preparation)
vi
5. Keyon, A.S.A., Guijt, R.M., Bolch, C.J. and Breadmore, M.C., Capillary
Electrophoresis for the Analysis of Paralytic Shellfish Poisoning Toxins
in Shellfish: A Preliminary Study, UTAS Postgraduate Conference,
University of Tasmania, Australia, 6-7 September 2012.
6. Keyon, A.S.A., Guijt, R.M., Bolch, C.J. and Breadmore, M.C., Capillary
Electrophoresis for the Analysis of Paralytic Shellfish Poisoning Toxins
in Shellfish: Comparison of Detection Methods, The 28th International
Symposium On Microscale Bioseparations, Shanghai, China, 21-24 October
2012.
7. Keyon, A.S.A., Guijt, R.M., Bolch, C.J. and Breadmore, M.C., Capillary
Electrophoresis for the Analysis of Paralytic Shellfish Poisoning Toxins
in Shellfish: Comparison of Detection Methods, 9th International
Conference on Molluscan Shellfish Safety, Sydney, Australia, 17-22 March
2013.
8. Keyon, A.S.A., Guijt, R.M., Bolch, C.J. and Breadmore, M.C., Transient
Isotachophoresis-Capillary Zone Electrophoresis for The Analysis of
Paralytic Shellfish Toxins in Mussel Sample, 40th International Symposium
on High-Performance Liquid-Phase Separations and Related Techniques
(HPLC 2013), Hobart, Australia, 18-21 November 2013.
9. Keyon, A.S.A., Guijt, R.M., Bolch, C.J. and Breadmore, M.C., Droplet-
based Method for Post-Column Fluorescence Analysis in Capillary
Electrophoresis, 5th Australia and New Zealand Micro/Nanofluidics
Symposium (ANZMNF), Hobart, Australia, 14-16 April 2014.
vii
List of Abbreviations
ACN Acetonitrile
AOAC Association of Official Analytical Chemists
AQC 6-aminoquinolyl-N-hydroxysuccinimidyl carbamate
ASP Amnesic shellfish poisoning
AZA Azaspiracid
BGE Background electrolyte
CFP Ciguatera fish poisoning
CZE Capillary zone electrophoresis
C4D Capacitively coupled contactless conductivity detection
dcGTX2 decarbamoylgonyautoxin2
dcGTX3 decarbamoylgonyautoxin3
dcNEO decarbamoylneosaxitoxin
dcSTX decarbamoylsaxitoxin
DA Domoic acid
DSP Diarrhetic shellfish poisoning
DTX Dinophysistoxin
EIA Extracted ion electropherogram
ECD Electrochemical detection
ELISA Enzyme-linked immunosorbent assay
EOF Electroosmotic flow
ex Excitation wavelength
λem Emission wavelength
viii
µep Electrophoretic mobility
µeff Effective mobility
FLD Fluorescence detection
GTX1 gonyautoxin1
GTX2 gonyautoxin2
GTX3 gonyautoxin3
GTX4 gonyautoxin4
GTX5 gonyautoxin5
GTX6 gonyautoxin6
He-Cd Helium-Cadmium
HILIC Hydrophilic interaction liquid chromatography
HPLC High performance liquid chromatography
H2O2 Hydrogen peroxide
ITP Isotachophoresis
LE Leading electrolyte
LED Light emitting diode
LIF Laser induced fluorescence
LLE Liquid-liquid extraction
LOD Limit of detection
MEKC Micellar electrokinetic chromatography
MRM Multiple reactions monitoring
MS Mass spectrometry
NSP Neurotic shellfish poisoning
OA Okadaic acid
OPA o-phthaldialdehyde
ix
Pt Platinum
PTX Pectenotoxin
PSP Paralytic shellfish poisoning
PSTs Paralytic shellfish toxins
RBA Receptor binding assay
RP Reversed phase
RSD Relative standard deviation
SDS Sodium dodecyl sulfate
SEF Sensitivity enhancement factor
S/N Signal-to-noise ratio
SPE Solid phase extraction
TEFs Toxicity equivalent factors
TE Terminating electrolyte
tITP Transient isotachophoresis
tm Migration time
UV Ultraviolet
YTX Yessotoxin
x
Abstract
Paralytic shellfish poisoning (PSP) toxins, or usually termed as paralytic shellfish
toxins (PSTs), produced by marine and freshwater microalgae during algal blooms
can accumulate in filter-feeding bivalve shellfish. Early detection of PSTs in
shellfish is therefore important for food and public health safety. High performance
liquid chromatography (HPLC) methods with pre- or post-column oxidation for
fluorescence detection (FLD) and HPLC-mass spectrometry (MS) are the most
widely used instrumental analytical methods for PSTs, but are not easily miniaturised
for field-deployable portable analyser. Capillary electrophoresis (CE) can be
developed as an alternative method as it is compatible with miniaturisation, making it
an attractive method for a portable analyser for early on-site detection.
In order to develop appropriate portable instrumentation for CE of PSTs, it is
necessary to develop appropriate methods. This was first done by developing CE
methods with different detection techniques namely ultraviolet (UV), capacitively
coupled contactless conductivity detector (C4D), MS, and FLD - making this the first
report of the use of C4D and an improved FLD detection for various PSTs with CE.
Due to the fact that most oxidised PSTs were neutral, micellar electrokinetic
chromatography (MEKC) was used in combination with FLD. The capillary zone
electrophoresis-UV (CZE-UV) and CZE-C4D methods provided better resolution,
selectivity and separation efficiency compared to CZE-MS and MEKC-FLD
methods. However, CZE-UV and CZE-MS methods did not provide sufficient
sensitivity to detect PSTs at the regulatory concentration limit, while CZE-C4D and
MEKC-FLD did show sensitivity below or close to the regulatory limit. The latter
most portable methods were evaluated for the screening of PSTs in a naturally
xi
contaminated mussel sample. MEKC-FLD was successfully used for PSTs screening
in the periodate-oxidised sample, whilst CZE-C4D method suffered from significant
interferences from sample matrix; a result that motivated further investigation of an
on-line preconcentration method to deal with the high conductivity sample matrix
and improve the sensitivity.
Therefore, CZE with C4D was examined with counter-flow transient
isotachophoresis (tITP). The tITP system exploited the naturally high sodium
concentration in mussel sample to act as a leading ion, in combination with one
electrolyte acting as terminating electrolyte (TE) and background electrolyte (BGE).
Optimisation of the BGE concentration, duration of counter-flow and injected sample
volume suitable for tITP resulted in sensitivity enhancement of at least two-fold over
CZE-C4D method developed in the first body of work. In particular, the modest gain
in sensitivity was achieved in the existence of a high concentration of sodium, a
sample matrix property that was problematic in previous method. This allowed the
analysis of PSTs in mussel sample at below or close to the regulatory concentration
limit.
The pre-column periodate oxidation MEKC-FLD method described in the first body
of work enabled direct screening of PSTs in shellfish sample; however some toxins
produced multiple and/or identical oxidation products, affecting selectivity and
specificity of the method. The findings initiated investigation of CE with droplet
microfluidic post-column reaction system for the separation and FLD of PSTs. The
concept was that PSTs were separated using CZE and electrophoretically transferred
into droplets segmented by oil. Formation of droplets and electrical connection in the
CE-droplet microfluidic system were first evaluated. Depending on the total flow
rate of both aqueous and oil phases, nL-sized droplets could be formed having
xii
frequencies between 0.7-3.7 Hz. The use of an off-the-shelf micro cross for
positioning a salt bridge across the droplet flow from the separation capillary outlet
enabled the compartmentalisation of the analytes while maintaining the electrical
connection. Further, the potential of the system was investigated for post-column
oxidation of PSTs. Compartmentalised in the droplets, PSTs reacted with periodate
oxidant already present in the droplets, in which only a single peak for each PST was
detected by FLD.
Given that the general objective of this research study is to develop suitable CE
methods that can be implemented for on-site PSTs detection, the potentials of the
developed methods compatible with miniaturisation and portability have been
demonstrated. The CE methods with different detection techniques, combined with
an on-line preconcentration and ability to be coupled with post-column reaction
indicates the versatility of CE as alternative analytical method for PSTs.
xiii
Table of content
Declaration .......................................................................................................... i
Acknowledgement .............................................................................................. ii
Statement of co-authorship ............................................................................... iii
List of publications and presentations ............................................................... v
List of Abbreviations........................................................................................ vii
Abstract .............................................................................................................. x
Table of content ............................................................................................... xiii
Preface ................................................................................................................ 1
The importance of analysing paralytic shellfish poisoning toxins...................... 1
Capillary electrophoresis and its potential as portable analyser ......................... 3
Project aims and scopes of study ...................................................................... 5
References ....................................................................................................... 6
Chapter 1 ................................................................................................................ 9
Literatures review: Instrumental analytical methods for the analysis of paralytic
shellfish poisoning toxins and other major phycotoxins ........................................ 9
1.1 Introduction ............................................................................................ 9
1.2 Overview of toxins ................................................................................10
1.3 Instrumental analytical methods ............................................................15
1.4 Analysis of PSTs ...................................................................................18
1.4.1 HPLC methods ...............................................................................18
1.4.2 CE methods ....................................................................................28
xiv
1.5 Analysis of ASP toxins ..........................................................................32
1.5.1 HPLC methods ...............................................................................32
1.5.2 CE methods ....................................................................................39
1.6 Analysis of DSP toxins ..........................................................................42
1.6.1 HPLC methods ...............................................................................42
1.6.2 CE methods ....................................................................................50
1.7 The analysis of NSP toxins and CFP toxins ...........................................50
1.7.1 HPLC methods ...............................................................................51
1.8 Conclusions ...........................................................................................56
1.9 References.............................................................................................58
Chapter 2 ...............................................................................................................70
Capillary electrophoresis for the analysis of paralytic shellfish poisoning toxins in
shellfish: Comparison of detection methods ........................................................70
2.1 Introduction ...........................................................................................70
2.2 Experimental .........................................................................................73
2.2.1 Chemicals and methods ..................................................................73
2.2.2 Preparation of mussel sample .........................................................75
2.2.3 CE ..................................................................................................76
2.2.4 HPLC-FLD ....................................................................................77
2.3 Results and discussion ...........................................................................78
2.3.1 CE of PSTs and comparison of detection methods ..........................78
2.3.2 Application of CZE-C4D and MEKC-FLD to mussel sample..........93
xv
2.4 Conclusions ...........................................................................................97
2.5 References.............................................................................................98
Chapter 3 ............................................................................................................. 101
Transient isotachophoresis-capillary zone electrophoresis with capacitively
coupled contactless conductivity and ultraviolet detections for the analysis of
paralytic shellfish poisoning toxins in mussel sample ........................................ 101
3.1 Introduction ......................................................................................... 101
3.2 Experimental ....................................................................................... 103
3.2.1 Chemicals and methods ................................................................ 103
3.2.2 CE ................................................................................................ 103
3.2.3 tITP-CZE ..................................................................................... 105
3.2.4 Preparation of mussel sample ....................................................... 105
3.3 Results and discussion ......................................................................... 105
3.3.1 Development of tITP-CZE method ............................................... 105
3.3.2 Effect of BGE/TE concentration on the stacking and separation
performance .............................................................................................. 108
3.3.3 Effect of counter-flow and optimisation of sample volume injected
111
3.4 Application of tITP-CZE method to mussel sample ............................. 121
3.5 Conclusions ......................................................................................... 125
3.6 References........................................................................................... 126
Chapter 4 ............................................................................................................. 128
xvi
Droplet microfluidics for post-column reactions system in capillary
electrophoresis: Application to paralytic shellfish poisoning toxins ................... 128
4.1 Introduction ......................................................................................... 128
4.2 Experimental ....................................................................................... 131
4.2.1 Chemicals and methods ................................................................ 131
4.2.2 Droplet formation ......................................................................... 131
4.2.3 Interfacing of CE with droplets .................................................... 134
4.2.4 Droplet microfluidics post-column oxidation ................................ 135
4.3 Results and discussion ......................................................................... 136
4.3.1 Characterisations of droplets and CE-droplet microfluidic interface
136
4.3.2 CE-droplet compartmentalisation ................................................. 141
4.3.3 CE with droplet-based microfluidics post-column reaction for PSTs
149
4.4 Conclusions ......................................................................................... 156
4.5 References........................................................................................... 157
Chapter 5 ............................................................................................................. 159
General conclusions and future directions ......................................................... 159
Appendix .......................................................................................................... 164
1
Preface
The importance of analysing paralytic shellfish poisoning toxins
Microalgae produce a structurally diverse array of phycotoxins that can accumulate
through the food chain and affect human health. Such toxins are saxitoxin (STX) and
its analogues, collectively known as paralytic shellfish toxins (PSTs) due to their
neurotoxicity in mammals, including humans. At least 30 PSTs analogues have been
confirmed to have toxicological effects [1, 2], with their toxicity relative to STX
indexed through the Toxicity Equivalent Factors (TEFs) [3, 4]. Blooms of
dinoflagellates producing PSTs frequently occur in both Northern and Southern
Hemispheres, most commonly along the cold water marine coasts.
PSTs are of particular concern in shellfish-producing areas because they can
accumulate to high concentrations in shellfish including mussels, oysters and
scallops, and can cause human illness known as paralytic shellfish poisoning (PSP)
upon consumption. The first poisoning associated with human severe intoxification
event was reported in 1927 in California, which resulted in 102 being ill and six
deaths. Outbreaks in Europe, North America, Japan, South Africa, Australia and
New Zealand soon followed [5]. The impact of PSTs is also significant to seafood
industry. As the shellfish trade increases globally with increased exports and imports
to and from regions of the world where these toxins are widespread, effective
monitoring of toxins in shellfish products have to be in place. Relevant to the
Australian shellfish industry, a recent incident (in 2012) that has highlighted the
seriousness of this issue was the discovery of intolerable level of PSTs in a shipment
of blue mussels to Japan originating from the east coast of Tasmania [6]. This
2
resulted in a global recall of products from this source and extended closures (up to
100 days) of important shellfish production areas. The total economic impact was
estimated to be AUD25.6 million. The incident was attributed to a breakdown in the
biotoxin management plan procedures, which resulted in inadequate toxin
monitoring been carried out in the affected region before and while the shellfish were
being harvested. This resulted in failure to detect toxic bloom and the PSTs in the
shellfish, which led to the export of contaminated products. As such, there is a strong
need to detect and analyse PSTs in shellfish samples prior to toxins accumulating to
dangerous levels, in which early detection and prevention can further protect
economy activities and human health.
Mouse bioassay (MBA) is the common method used for the detection of PSTs in
regular monitoring; however this method is currently being phased out due to ethical
issues. Various alternative biological and analytical methods have been developed
and employed, including enzyme-linked immunosorbent assay (ELISA) [7-10],
receptor binding assay (RBA) [11, 12], saxiphilin-based assay [13, 14] and high
performance liquid chromatography (HPLC) [15, 16]. Among these methods, ELISA
is the only to have been commercially developed in a test kit for PSTs detection in
contaminated shellfish or water samples (i.e. www.r-biopharm.com,
www.biosense.com, www.abraxiskits.com, www.anconbiotech.com and
www.beaconkits.com). While these commercial kits are considered portable and can
be used to detect PSTs on-site, Campbell et al. [17] pointed out they have the
potential to produce subjective interpretation of results.
The Association of Official Analytical Chemists (AOAC) approves HPLC with pre-
or post-column oxidation and fluorescence detection (FLD) as official methods [15,
16]. On top of that, HPLC-mass spectrometry (MS) is also widely used, and is
3
expected to be accepted in the near future [18, 19]. Often only a few centralised
laboratories have the expertise and capacity to analyse PSTs by HPLC methods,
leading to significant analytical turnaround time. Moreover, HPLC is not prone for
miniaturisation to be used in portable instrumentation. Relevant to the purpose of this
research study, a technique that can be miniaturised and function as rapid and
selective detection method is much desired. Rapid and quantitative on-site PSTs
analysis at shellfish farms using portable instrumentation would improve toxin
management by providing timely toxin information to shellfish producers to manage
stock harvests around toxin closure periods, and reduce losses due to
harvested/packed seafood product loss or recall, and the risk of the sale of seafood
exceeding limits for safe human consumption.
Capillary electrophoresis and its potential as portable analyser
Electrophoresis has the potential as an alternative analytical method as it provides
rapid and efficient separation with minimal consumption of sample and reagents.
Capillary or microchip electrophoresis is well suited to portability because only a
separation capillary (or channel for microchip), a high voltage power supply and a
detector are needed to perform the separation. Significantly, it is inherently more
miniaturisable than HPLC since no highly sophisticated pressure pumping systems
are required. In general, a number of portable capillary and microchip electrophoresis
instrumentations/devices have been developed. This includes the Mars Organic
Analyser [20-22], hand-held portable isotachophoresis system [23] and the Medimate
Multireader® [24, 25]. The challenging part in CE is the detection system, which has
always been a significant since its inception three decades ago. Miniaturisation of the
4
detector is therefore a key step in the pursuit of making CE portable – it should
provide good sensitivity and be universal for a range of analytes that can be
separated in CE [26]. Capacitively coupled contactless conductivity detector (C4D)
and FLD are the most miniaturisable detection methods that have been routinely used
to improve portability [27]. In general, a C4D can be constructed from two axially
positioned electrodes through which the separation capillary (or channel for
microchip) is fitted [28]. Alignment of the electrodes is unnecessary and compared to
contact conductivity, electrode fouling is impossible as these are not in direct contact
with the solution. Lasers are currently the most important light source for FLD. The
wavelength range available covers almost the whole spectrum from near infrared
down to near UV. The applicability of incorporation into portable analyser increases
when employing laser diodes or light emitting diodes (LED) as excitation light
source [27]. While miniaturised MS has been described [29, 30], they are highly
sophisticated and require significant infrastructure (e.g. vacuum) to be operational.
Ultraviolet (UV) photometric detector based on LEDs has the potential to be portable
[31], however LEDs currently do not exist with a wavelength lower than 240 nm.
Other miscellaneous detection systems such as nuclear magnetic resonance and
inductively coupled plasma-MS are not suited to coupling to CE and miniaturisation
in the near future, despite their increasing use for inorganic speciation [27].
5
Project aims and scopes of study
From the above discussion, the analysis of PSTs is of high importance and early on-
site detection in shellfish would be significantly changing the way that shellfish
quality is managed. The advantage of CE being miniaturisable and portable makes it
a perfect choice to reach this goal. Therefore, the general aim of this research study is
to explore the potential of CE as new alternative instrumental analytical method for
the analysis of PSTs. The findings will serve as a fundamental understanding towards
the realisation of field-deployable portable analyser. It is believed that once method
is developed properly, then it will considerably increase simplicity of use on-site to
produce laboratory-quality results.
The specific aims of the project are to:
1. Develop new CE methods for the analysis of PSTs.
2. Develop and compare several detection methods for CE, namely C4D, FLD,
UV and MS as the first step towards the development of portable analyser.
3. Maximise the sensitivity and applicability of the most portable CE method
(CE-C4D) by combining an on-line preconcentration method.
4. Develop an improved method to the CE-FLD method based on an innovative
technique in order to improve separation selectivity and specificity.
5. Optimise, validate and implement the developed CE methods to PSTs-
contaminated shellfish sample.
6
References
[1] Rossini, G. P., Hess, P., EXS 2010, 100, 65-122.
[2] Vale, P., Phytochem. Rev. 2010, 9, 525-535.
[3] Oshima, Y., J. AOAC Int. 1995, 78, 528-532.
[4] Panel, E., EFSA J. 2009, 1019, 1-76.
[5] Otero, J. J. G., in: Botana, L. M. (Ed.), Seafood and Freshwater Toxins:
Pharmacology, Physiology and Detection, CRC Press Taylor and Francis Group,
Florida 2014, pp. 124-167.
[6] Campbell, A. M., C. Pointon, A. Hudson, D. Nicholls, C., Review of the 2012
paralytic shellfish toxin event in Tasmania associated with the dinoflagellate alga,
Alexandrium tamarense. A SafeFish FRDC Project 2012/060., Fisheries Research
and Development Corporation and South Australian Research and Development
Institute, Adelaide 2013.
[7] Chu, F. S., Fan, T. S., J. AOAC Int. 1985, 68, 13-16.
[8] Usleber, E., Schneider, E., Terplan, G., Lett. Appl. Microbiol. 1991, 13, 275-277.
[9] Chu, F. S., Hsu, K. H., Huang, X., Barrett, R., Allison, C., J. Agric. Food Chem.
1996, 44, 4043-4047.
[10] Usleber, E., Donald, M., Straka, M., Märtlbauer, E., Food Addit. Contam. 1997,
14, 193-198.
[11] Van Dolah, F. M., Field, T. A. L., Doucette, G. J., Bean, L., Niedzwiadek, B.,
Rawn, D. F. K., J. AOAC Int. 2009, 92, 1705-1713.
[12] AOAC, Official Method 2011.27 First Action 2011: Paralytic shellfish
poisoning in shellfish by receptor binding assay, AOAC International, Gaithersburg,
MD 2012.
[13] Llewellyn, L. E., Doyle, J., Negri, A. P., Anal. Biochem. 1998, 261, 51-56.
7
[14] Llewellyn, L. E., Doyle, J., Toxicon 2000, 39, 217-224.
[15] AOAC, Official Method 2005.06 Paralytic shellfish poisoning toxins in
shellfish.Pre-chromatographic oxidation and liquid chromatography with
fluorescence detection, AOAC International, Gaithersburg, MD 2005.
[16] AOAC, Official Method 2011.02 First Action 2011: Paralytic shellfish toxins in
mussels, clams, scallops and oysters by liquid chromatography post-column
oxidation, AOAC International, Gaithersburg,MD 2012.
[17] Campbell, K., Rawn, D. F. K., Niedzwiadek, B., Elliott, C. T., Food Addit
Contam Part A Chem Anal Control Expo Risk Assess 2011, 28, 711-725.
[18] Dell'Aversano, C., Hess, P., Quilliam, M. A., J. Chromatogr. A 2005, 1081,
190-201.
[19] Sayfritz, S. J., Aasen, J. A. B., Aune, T., Toxicon 2008, 52, 330-340.
[20] Skelley, A. M., Scherer, J. R., Aubrey, A. D., Grover, W. H., Ivester, R. H. C.,
Ehrenfreund, P., Grunthaner, F. J., Bada, J. L., Mathies, R. A., Proc. Natl. Acad. Sci.
U. S. A. 2005, 102, 1041-1046.
[21] Benhabib, M., Chiesl, T. N., Stockton, A. M., Seherer, J. R., Mathies, R. A.,
Anal. Chem. 2010, 82, 2372-2379.
[22] Kim, J., Jensen, E. C., Stockton, A. M., Mathies, R. A., Anal. Chem. 2013, 85,
7682-7688.
[23] Kaigala, G. V., Bercovici, M., Behnam, M., Elliott, D., Santiago, J. G.,
Backhouse, C. J., Lab Chip 2010, 10, 2242-2250.
[24] Floris, A., Staal, S., Lenk, S., Staijen, E., Kohlheyer, D., Eijkel, J., Van Den
Berg, A., Lab Chip 2010, 10, 1799-1806.
[25] Ávila, M., Floris, A., Staal, S., Ríos, A., Eijkel, J., van den Berg, A.,
Electrophoresis 2013, 34, 2956-2961.
8
[26] Kubáň, P., Nguyen, H. T. A., Macka, M., Haddad, P. R., Hauser, P. C.,
Electroanalysis 2007, 19, 2059-2065.
[27] Ryvolová, M., Macka, M., Brabazon, D., Preisler, J., TrAC-Trends Anal Chem
2010, 29, 339-353.
[28] Zemann, A. J., TrAC-Trends Anal Chem 2001, 20, 346-354.
[29] Gao, L., Sugiarto, A., Harper, J. D., Cooks, R. G., Ouyang, Z., Anal. Chem.
2008, 80, 7198-7205.
[30] Ouyang, Z., Noll, R. J., Cooks, R. G., Anal. Chem. 2009, 81, 2421-2425.
[31] Johns, C., Macka, M., Haddad, P. R., Electrophoresis 2004, 25, 3145-3152.
9
Chapter 1
Literatures review: Instrumental analytical methods for the
analysis of paralytic shellfish poisoning toxins and other major
phycotoxins
1.1 Introduction
In general, phycotoxins are small to medium-sized natural metabolites (molecular
mass ranges 300 to over 3000 Da) that belong to many different classes of chemical
compounds including tetrahydro purines, amino acids, alkaloids or polyethers [1, 2].
Toxins produced primarily by dinoflagellates and diatom species affect human health
through seafood consumption while cyanobacteria toxins are through domestic water
usage, although some cyanobacteria toxins can enter benthic food chain. Ingestion of
toxic seafood leads to several major human poisoning syndromes namely paralytic
shellfish poisoning (PSP), amnesic shellfish poisoning (ASP), diarrhetic shellfish
poisoning (DSP), neurotic shellfish poisoning (NSP) and ciguatera fish poisoning
(CFP).
Mouse bioassay (MBA), being the common detection method is expensive, requires
specialised animal facilities and cannot be automated [3]. The MBA’s lack of
selectivity, sensitivity and the difficulties of quantitation have driven the
development of instrumental analytical methods such as the use of separation
techniques like chromatography (i.e. HPLC) or electrophoresis (i.e. CE). Other than
rapid method development, instrumental analytical methods are capable of
10
identification and quantitation of individual compounds in most toxin classes, due to
the high sensitivity and selectivity these methods can offer.
In this chapter, developments in chromatographic and electrophoretic methods for
major phycotoxins are discussed, with attention given on the strategies used to
achieve efficient, sensitive and rapid separation methods, as well as to the endeavour
on introducing multi-toxin analysis methods. In addition, discussions are provided on
criteria needed and challenges faced on the development of portability of these
analytical separation methods towards early on-site detection/screening.
1.2 Overview of toxins
The chemical class, main poisoning effects and the source of major phycotoxins are
summarised in Table 1.1. More details on the mechanism of poisoning action and
toxicological evaluations can be found in several reviews on the subject [2, 4, 5]. The
most toxic phycotoxins are the paralytic shellfish toxins (PSTs), which include
saxitoxin (STX) and its analogues ((decarbamoylsaxitoxin (dcSTX), neosaxitoxin
(NEO), decarbamoylneosaxitoxin (dcNEO), gonyautoxin1 to gonyautoxin6 (GTX1–
GTX6), decarbamoylgonyautoxin1 to decarbamoylgonyautoxin4 (dcGTX1–
dcGTX4), and C1 to C4), with their toxicity relative to STX indexed through TEFs
[3, 6]. These originate from Gonyaulax catenella [7], Alexandrium minitum,
Alexandrium tamarense, Gymnodium catenatum, Pyrodinium bahamensevar,
Compressum dinoflagellates and Anabaena circinalis cyanobacteria [8, 9].
Consumption of PSTs-contaminated shellfish often results in severe vomiting,
nausea and even death due to respiratory arrest and cardiovascular shock due to
blockage of voltage-gated sodium channels [2, 10].
11
Table 1.1 Characteristics of major phycotoxins
Phycotoxins Toxin source Chemical class and structure Lipophilicity Main poisoning effect
PSTs
(STX, dcSTX, NEO,
dcNEO, GTX1-GTX6,
dcGTX2-dcGTX3, C1-C4
Dinoflagellates,
cyanobacteria (for
STX only)
Tetrahydro purine, two guanidium
groups
Hydrophilic Severe vomiting, nausea,
respiratory arrest, cardiovascular
shock leading to death
ASP toxin
(DA)
Diatom algae Cyclic amino acid, three carboxy
groups
Hydrophilic Severe gastrointestinal and
neurological disorders
DSP toxins
(OA, DTX1-DTX5)
Dinoflagellates Polyether, spiro-keto assembly Lipophilic Gastrointestinal disorder, diarrhoea,
abdominal cramps, nausea,
vomiting, tumour promoter
PTXs
(PTX1-PTX7, PTX2sa,
7-epi PTX2sa)
Dinoflagellates Polyether ester macrocycle
Lipophilic Gastrointestinal disturbance
YTXs Dinoflagellates Ladder-shaped polyether Lipophilic Gastrointestinal disturbance,
12
(homoYTX,
carboxyYTX,
hydroxyYTX)
abdominal pain
AZAs
(AZA2, AZA3),
gymnodimines and
spirolides
Dinoflagellates AZAs: Polyether, second amine,
three spiro ring
Gymnodimines and spirolides:
Cyclic imine, macrocycle
Lipophilic Nausea, vomiting, abdominal
cramps, diarrhoea
NSP toxins
(brevetoxins and
analogues)
Dinoflagellates Ladder-shaped polyether Lipophilic Respiratory irritation, mild
gastroenteritis, neurologic
symptoms
CFP toxins
Dinoflagellates Ladder-shaped polyether Lipophilic Gastrointestinal and neurological
disorders
13
Domoic acid (DA) and its isomers are ASP toxins produced by Pseudo-nitzschia, a
diatom widely distributed in coastal waters around the world [2, 11]. The
intoxication is characterised by severe gastrointestinal and neurological disorders,
hence the ASP category. The mechanism of toxicity is explained by its structural
similarity with the excitatory neurotransmitter glutamic acid but with a much
stronger receptor affinity. It binds predominantly to N-methyl-D-aspartate receptors
in the central nervous system, resulting in depolarisation of neurons and subsequent
cell dysfunction or death [12].
DSP toxins such as okadaic acid (OA) and its analogues (dinophysistoxin1 to
dinophysistoxin5 (DTX1-DTX5)) are lipid-soluble polyether compounds produced
by Dinophysis dinoflagellate [13]. These toxins are potent inhibitors of Type 1 and
2A protein phosphatases and also a powerful tumour promoter [14, 15]. Ingestion of
DSPs-contaminated shellfish results in gastrointestinal disorder, diarrhoea,
abdominal cramps, nausea and vomiting. Pectenotoxin (PTX) and its analogues
(pectenotoxin1 to pectenotoxin7 (PTX1-PTX7), pectenotoxin-2 seco acid (PTX2sa)
and 7-epi pectenotoxin-2 seco acid (7-epi PTX2sa)) are also produced by
Dinophysis, which are initially associated with DSP. These toxins do not induce
diarrhoea [16] and show histopathological changes in the liver and stomach when
administered into mice [17], indicating a toxicity mechanism that differs from OA
and DTX [2]. Yessotoxin (YTX) and its analogues (homoYTX, carboxyYTX, and
hydroxyYTX) are initially misclassified as DSP toxins because of their frequent co-
occurrence with DSP toxins. Despite evidence that YTXs do not cause diarrhetic
effects when administered orally to mice [18-20], acute toxicity relevance is still
taken into account for maintaining safety of the public health. YTXs are produced by
Protoceratium reticulatum [21], Lingulodinum polyedrum [22] and Gonyaulax
14
spinifera dinoflagellates [23]. Other lipophilic biotoxins usually analysed together
with DSP toxins include azaspiracid (AZA) and its analogues (AZA2 and AZA3),
spirolides and gymnodimines. The ingestion of AZAs-contaminated shellfish can
lead to health effects similar to DSP [24]. The gastrointestinal effects may be
explained by the alterations they induce in cytoskeletal structures and the E-cadherin
system, with disruption of cell-matrix interactions, as well as perturbation of the
intestinal barrier function [25]. There is no clear evidence of the toxicity of
gymnodimines and spirolides to human, but as these are highly toxic to mice, they
constitute a source of false positives by the MBA [26].
Brevetoxins and analogues, causative agents for NSP, are produced by Karenia
brevis dinoflagellate [27]. Respiratory irritation and mild gastroenteritis with
neurologic symptoms may occur after inhaling aerosol containing these or after
ingesting contaminated seafood. Finally, ciguatoxins are another group of lipophilic
marine toxins associated with CFP. Unlike other shellfish poisonings, these toxins
are produced in fish in tropical and subtropical areas due to biotransformation and
acid-catalysed spiroisomerisation of gambiertoxins by the Gambierdiscus toxicus
microalgae [28, 29].
15
1.3 Instrumental analytical methods
Detection and quantitation of major phycotoxins has been accomplished by
biological assays and/or instrumental analytical techniques. MBA involves injecting
a contaminated sample intra-peritoneally into a mouse and measuring the time to
death. Despite this still being the method of choice for some parts of the world,
ethical arguments against the continued use of live animals has seen the MBA
banned from many jurisdictions [30]. Several non-animal biological methods have
been developed for the detection of phycotoxins such as enzyme-linked
immunosorbent assay (ELISA), receptor binding assay (RBA), cell based assay and
biosensing. RBA method has been accepted by the AOAC for the detection of PSTs
[31], which like the MBA provides a single integrated toxic potency value to reflect
the health risk to humans, but unable to identify the toxins present. Analytical
methods for detection, quantitation and possibly for routine monitoring of
phycotoxins are thus required and must be able to handle the increasing numbers of
toxin variants discovered around the world. This includes the application of
separation techniques such as chromatographic and electrophoretic methods that
mostly can separate and identify each of toxins classes and analogues. The
advantages of analytical method are the superior sensitivity for individual toxins as
compared to MBA, and data can be collected faster. With regards to toxin screening,
this helps in verifying the stage of contamination which aids in subsequent action
strategies (i.e. manage stock harvest, recall product or close farm). Whilst a single
and rapid method would be preferred, this is not straight forward with the chemical
diversity of phycotoxins. Accurate detection and quantitation of the PSTs in shellfish
is a substantial challenge because of this diversity as well as the difference in
toxicities. It is imperative that these issues are factored in during method
16
development, in addition to the type of detection system, sensitivity, extraction and
clean-up method and sample matrix interferences.
Analytical separation methods are particularly attractive given the global efforts
towards on-site detection through miniaturisation. The key to functionality and
simplicity of portable instrumentation lies in the miniaturisable separation method,
detector system and easy-to-perform sample preparation. While electrophoresis is a
method naturally predetermined for miniaturisation, in case of LC, it has been a
much more challenging task. Considering the technical and instruments requirement
of LC, it is not easy to imagine that LC can be miniaturised and be made portable
without extensive efforts; although it might be technically possible (i.e. portable LC
for mobile laboratories [32] and portable immunoextraction-LC for herbicide
analysis [33]). The drawback inherently comes from the requirement of eluent and
pumping system. In regards to the analysis of phycotoxins, the requirement of
multiple columns, mobile phases and sample injections for some of the toxins (i.e.
PSTs and DSP toxins) complicates this further. On the other hand, electrophoresis is
a promising technique for this purpose. Generally speaking, the effectiveness and
significance of CE in toxin analyses is evidenced by the numerous publications of
compounds ranging from fungal metabolic products (i.e. mycotoxins) [34-37],
bacterial toxins [38-40] and animal venoms [41-44]. Extending from that, portable
capillary and microchip electrophoresis methods have been successfully developed
for various toxin analyses [45-48].
17
The following section discusses the HPLC and CE methods for the analysis of major
phycotoxins. Potential and challenges of most of the methods towards portability are
pointed out. Table 1.2-1.5 provides information about the separation protocols,
detection and method sensitivity (in terms of limit of detection (LOD) in sample
extract) for the analysis of these toxins in various shellfish, microalgae or water
samples.
18
1.4 Analysis of PSTs
In most countries, the current regulatory limit set for PSTs in seafood to ensure
public health safety is 800 μg/kg [49, 50], which corresponds with 800 ng/mL of
PSTs in sample extract following the extraction procedure described in the AOAC
methods [49, 51]. An overview of selected HPLC and CE protocols for the analysis
of PSTs are tabulated in Table 1.2, with the most important works discussed below in
separate sections.
1.4.1 HPLC methods
HPLC is one of the first instrumental analytical separation methods applied to PSTs
analysis seeking alternatives to the MBA. The breakthrough of PSTs oxidation to
enable FLD was first achieved by Bates and Rapoport using hydrogen peroxide
(H2O2) [52]. When oxidised, a conjugated imino purine was formed from the
guanidium on the tetrahydro purine backbone which had an excitation maximum at
330 nm [52]. Unfortunately, not all PSTs yielded fluorescent product (i.e. N-
hydroxylated toxins, NEO and GTX1,4), a problem solved by Oshima et al. [53] who
used tert-butyl hydroperoxide as a post-column oxidant after PSTs separation with
HPLC. The fluorescence intensity increased seven times relative to the H2O2
oxidation method, with LODs in sample extract in the range of 0.32-14 µg/mL.
Using this reagent, all PSTs including the N-hydroxylated toxins were oxidised, but
the reaction had to be conducted > 100 °C, an experimental complication compared
to H2O2 oxidation carried out at room temperature.
19
Table 1.2: Selected HPLC and CE protocols for the analysis of PSTs.
Protocol Sample Conditions LOD Ref
Post-column oxidation
RP-ion pair-HPLC-FLD
Shellfish Column: Develosil C8 or Inertsil C8
3 separate isocratic elutions:
1) 2 mM sodium heptanesulfonate in 30 mM ammonium phosphate
(pH 7.1) and ACN (STX and NEO)
2) 2 mM sodium heptanesulfonate in 10 mM ammonium phosphate
(pH 7.1) (gonyautoxins)
3) 1 mM tetrabutylammonium phosphate (pH 5.8) (C toxins)
0.8-4 ng/mL [6]
Post-column oxidation
RP-ion pair-HPLC-FLD
Shellfish Column: Agilent Zorbax Bonus RP and Thermo BetaBasic 8
2 separate gradient elutions:
1) 11 mM heptanesulfonate, 5.5 mM phosphoric acid (pH 7.1)/11
mM heptanesulfonate, 16.5 mM phosphoric acid, 11.5% ACN (pH
7.1) (STX, NEO and gonyautoxins)
4-70 ng/mL [54]
20
2) 2 mM tetrabutyl ammonium phosphate (pH 5.8)/2 mM tetrabutyl
ammonium phosphate (pH 5.8) in 4% ACN (C toxins)
Pre-column oxidation
RP-HPLC-FLD
Shellfish Column: Supelcosil LC-18
Gradient elution: 100 mM ammonium formate (pH 6.0)/100 mM
ammonium formate (pH 6.0) with 5% ACN
3-10 ng/mL
[55, 56]
Pre-column oxidation
RP-HILIC-FLD
Microalgae Column: µBondapak NH2
Gradient elution: 50 mM sodium acetate buffer (pH 6.5)/water
0.04-1 ng/mL [57]
Pre-column oxidation
RP-HPLC-FLD
Shellfish Column: Ascentis Express C18
Gradient elution:100 mM ammonium formate (pH 6.5)/100 mM
ammonium formate with 5% ACN
NAa [58]
Pre-column oxidation
RP-HPLC-FLD
Shellfish Column: Kinetex C18 and Poroshell C18
Gradient elution: 100 mM ammonium formate (pH 6.5)/100 mM
ammonium formate with 5% ACN
5-80 ng/mL [59]
HILIC-MS/MS Microalgae, Column: TSK-gel Amide-80 2-10 ng/mL [60]
21
shellfish Isocratic elution: (35/65, %v/v) 2 mM ammonium formate/95%
ACN, both contain 3.6 mM formic acid (pH 3.5)
HILIC-MS/MS Shellfish Column: TSK-gel Amide-80
Gradient elution: 2 mM ammonium formate/99% ACN, both contain
20 mM acetic acid (pH 3.5)
2-70 ng/mL
[61]
HILIC-MS/MS Shellfish Column: X-bridge Amide
Gradient elution: 3.6 mM formic acid/95% ACN and 3.6 mM formic
acid
0.08-3 ng/mL [62]
HILIC-MS/MS Shellfish,
fish
Column: TSK-gel Amide-80
Gradient elution: 2 mM ammonium formate /ACN, both contain
0.1% formic acid
0.4-0.6 ng/mL [63]
CZE-UV Shellfish BGE: 20 mM sodium citrate (pH 2.0) 1500-1800 ng/mL [64]
tITP-CZE-UV Shellfish BGE: 35 mM morpholine adjusted to pH 5.0 with formic acid (also
as LE), 10 mM formic acid as TE
3-8 ng/mL [65]
22
Stacking-CZE-UV-MS Shellfish BGE: 100 mM morpholine adjusted to pH 4.0 with formic acid NAa [66]
tITP-CZE-UV Shellfish BGE: 50 mM morpholine adjusted to pH 5.0 with formic acid (LE),
10 mM formic acid as TE
50-60 ng/mL [67, 68]
tITP-CZE-UV Microalgae BGE: 50 mM morpholine adjusted to pH 5.0 with formic acid (LE),
10 mM formic acid pH 2.7 as TE
32-80 ng/mL [69]
CZE-MS/MS Shellfish BGE: Trizma buffer pH 7.2 560-5600 ng/mL [70]
Pre-column derivatisation
CZE-FLD
NAa BGE: 50% 1.9 mM phosphate buffer and 50% ACN 0.3 fg/mL [71]
a) NA Not available
23
The same group further refined their post-column method using periodic acid [6]
with FLD at ex = 330 nm and λem = 390 nm. Oxidation could be done at relatively
lower temperature (85°C) and the bubble generation that was known to be
problematic when using peroxide could be avoided. In order to analyse the PSTs
prior periodate oxidation, three separate isocratic chromatographic conditions using a
reversed phase (RP) column and three separate injections of STX and NEO (group
1), gonyautoxins (group 2) and C toxins (group 3) had to be used (conditions as
outlined in Table 1.2). Three different mobile phases containing ion pair reagents
were employed. To further simplify instruments set-up, two separate binary gradients
using two separate RP columns (sequentially changed) and two separate injections of
STX, NEO, gonyautoxins (group 1) and C toxins (group 2) were used instead [54].
The AOAC adapted the latter approach as AOAC HPLC-FLD Official Method
2011.02 [72].
As an alternative to post-column HPLC-FLD method, Lawrence et al. developed a
pre-column oxidation HPLC-FLD which was instrumentally much simpler [55, 56].
Periodate and/or peroxide oxidation solution at basic pH was used, with ammonium
formate used to improve the sensitivity of the N-hydroxylated toxins. However, this
approach had two major disadvantages. First, that for some PSTs, such as isomeric
toxins (GTX2 and GTX3), the same oxidation product was formed (Figure 1.1) and
thus they could not be distinguished from each other. Second, some of the PSTs
produced multiple oxidation products, thus careful selection of primary peak for
quantitation was crucial. For example, NEO, and GTX1,4 produced three oxidation
peaks, but only the second peaks were used for quantitation. The oxidised PSTs were
separated on a RP column using single gradient system with LODs in sample extract
of 3-10 ng/mL. In order to be used as regular monitoring method, the method was
24
refined in which initial screening of overall PST content in samples using periodate
oxidation was conducted. If positive, this screen was followed by chromatographic
fractionations (C18 or ion exchange solid phase extraction), and periodate and/or
peroxide oxidations of the fractions to enable quantitation of the individual PSTs.
The method was approved as AOAC HPLC-FLD Official Method 2005.06 [49].
Further details on both pre- and post-column oxidation HPLC methods are further
described by Rodriguez et al. [73] and De Grasse et al. [74] in their reviews.
25
Figure 1.1 Chromatographic patterns showing oxidation products formed after
periodate and peroxide oxidations of PSTs. The same quantity of each toxin was
used for each pre-column oxidation reaction. Arrows indicate peaks used for
quantitation. Chromatographic conditions: RP HPLC, gradient elution with 100 mM
ammonium formate (pH 6.0)/100 mM ammonium formate (pH 6.0) with 5% ACN.
Reproduced from [56] with permission.
26
There have been a number of publications that have focused on improving the PSTs
separation after pre-column oxidation through the use of new column types. He et al.
[57] used a μBondapak NH2 column in hydrophilic interaction mode for the
separation of PSTs in microalgae. All PSTs were eluted with water/acetate buffer
solution (pH 6.5) without the use of organic solvent usually needed during separation
using RP column. LODs of 0.04-1 ng/mL were achieved, as such the method was
very sensitive to analyse trace level of PSTs especially in water samples. De Grasse
et al. [58] evaluated the state-of-the-art solid core particle RP column (Ascentis
Express) with the advantage of increased resolution due to shorter diffusion paths
which reduced peak broadening. Compared to elution times using a conventional RP
column, the new column was faster which reduced the separation time from 15 min
to 5 min. Hatfield et al. examined superficially porous sub 2-µm silica beads
columns (Kinetex and Poroshell), whereby a separation time of 6 min was achieved,
half the time using conventional RP column (Gemini C18). LODs ranging from 5-80
ng/mL were achieved, improved at least two-fold compared to a RP column (LODs
of 10-140 ng/mL) [59]. Similar performance to the Ascentis Express column was
shown [58] but the sample throughput doubled and analysis turnaround time (~24 h)
improved.
The most common issue when using a universal, non-selective detector is
interference by toxin variants as well as sample matrix components, which can lead
to false positives or negatives. Thus, HPLC combined with MS allows unambiguous
identification, confirmation and quantitation of the toxins and/or new analogues.
Dell’ Aversano et al. [60] used hydrophilic interaction LC-MS/MS (HILIC-MS/MS)
with a TSK-gel Amide-80® column. By using triple quadrupole MS, protonated
27
and/or fragment ions were selected for monitoring, which enabled the detection and
quantitation of isomeric toxins (GTX1-6, dcGTX1-4 and C1-4). Simultaneous
determination of all major PSTs was achieved in a single 30 min analysis, with high
degree of selectivity and LODs of 2-10 ng/mL.
To obtain cleaner chromatograms in HILIC-MS/MS, improvements on the
preparation methods have been published [61-63]. After extraction of shellfish
samples in acidified aqueous ACN, Sayfritz et al. [61] further cleaned-up the
extracts using hydrophilic SPE columns (i.e. Oasis HLB and Carbograph activated
carbon), in which sample matrix interference were removed. The method provided
LODs in sample extract from 2-70 ng/mL. More recently, Watanabe et al.[62] used a
basic alumina column to clean-up microalgae extracts. The alumina column removed
interfering peaks, as well as reducing peak shifting during separation and controlling
variable ionisation efficiency during MS detection. Almost all the PSTs could be
detected at < 0.8 ng/mL while GTX2 and dcGTX2 were slightly higher (3 ng/mL),
due to the lower ionisation efficiency. Their method gave a lower LOD than either
HPLC-FLD [6, 54, 55] or HPLC-MS [60, 61] developed prior to this work.
In brief, HPLC-FLD and HPLC-MS methods give adequate sensitivity to be used for
PSTs analysis in shellfish. This is a promising criterion to be used in portable
instrumentation, however much effort is required to miniaturise the separation
technique itself. In addition, the multiple chromatographic conditions required to
separate PSTs prior to post-column oxidation makes realising miniaturisation more
difficult compared to pre-column HPLC-FLD, however this also has an issue with
regards to selectivity. Detection wise, FLD is easily miniaturised, which is in contrast
to MS which has not been miniaturised at all for any liquid-phase technique.
28
1.4.2 CE methods
PSTs contain two basic guanidium moieties in their structure with pKa values > 8.5
and hence are positively charged in acidic medium (except C toxins which are
neutral as a result of protonated sulfonic acid group). The successful separation of
STX by cellulose acetate electrophoresis [75] prompted Thibault et al. [64] to
investigate capillary zone electrophoresis (CZE) for the analysis of STX and NEO.
The molar absorptivity of PSTs above 220 nm was very low (< 0.001 mol/cm);
however they showed some response at 200 nm [76] allowing UV detection. The
toxins were successfully separated using 20 mM sodium citrate at pH 2.1 as the
BGE; however, the LODs were only 1500-1800 ng/mL, which were inadequate for
trace analysis of PSTs. Transient isotachophoresis-CZE (tITP-CZE) was used as an
on-line preconcentration approach to improve the LOD [65]. An electrolyte
containing 35 mM morpholine adjusted to pH 5.0 with formic acid acted as the
BGE/leading electrolyte (LE) while 10 mM formic acid as the terminating electrolyte
(TE). The separation efficiency and resolution of all PSTs were impressive with
sensitivity improvement up to 500-fold over Thibault’s method [64], achieving
detection limits at least 100-times lower than the regulatory limit.
Gago-Martinez and co-workers [67, 68] reported optimisation of tITP-CZE-UV
method for the analysis of PSTs in shellfish samples. The use of lower separation
voltage in combination with higher concentration of BGE/LE (50 mM morpholine
added with formic acid) improved resolution and separation efficiency of all PSTs
including the isomeric toxins (i.e. GTX1 and GTX4, GTX2 and GTX3, and GTX5)
(Figure 1.2). LODs between 50-60 ng/mL were obtained, which enabled detection of
PSTs in shellfish extracts at concentrations lower than the regulatory limit.
29
For the analysis of PSTs in microalgae, Wu et al. [69] compared tITP with two other
on-column preconcentration (field amplification and stacking) used for CZE-UV.
Due to the high concentrations of acetic acid and other compounds in microalgae
extract, tITP was concluded as the better choice to enrich PSTs and to partly remove
matrix ions. The LODs obtained were 32-80 ng/mL, which was comparable to report
by Gago-Martinez et al. [68].
30
Figure 1.2 CE-UV analysis of PSTs standards. Conditions: 50 mM morpholine
adjusted to pH 5.0 with formic acid as BGE/LE, 10 mM formic acid as TE; sample
injected at 50 mbar for 90 s and separation was made at 20 kV. Note that different y-
axis was used in upper and lower figures. Reproduced from [68] with permission.
31
MS provides unambiguous identification of toxins; however, combination of MS
with CE is relatively harder than with HPLC. Volatile buffer usually possessing low
resolving power characteristics for CE must be used, limiting the freedom for BGEs
selection. By using 100 mM acetic acid (pH 2.9) as BGE for CE and ionspray
ionisation MS, Thibault et al. [64] identified peaks obtained from the separation
using CZE-UV for shellfish and microalgae samples, which were confirmed to be
STX and NEO. Pleasance et al. [70] developed a method based on co-axial
arrangement as ionspray ionisation interface. Sheath flow (formic acid) could be
delivered independently of the CE column effluent, thus providing greater flexibility
in the selection of BGE for MS; trizma buffer (pH 7.2) was used for separation.
However, dilution of the analyte zone by the sheath flow required as part of the
interface was expected. This method provided LODs between 560-5600 ng/mL
which were higher than the regulatory limit. Nevertheless, it allowed identifying and
distinguishing isomeric toxins (e.g. GTX2 and GTX3) in shellfish samples.
Given the popularity of HPLC-FLD and the fact that FLD can be miniaturised, it is
surprising that only a report of CE-FLD for the analysis of PSTs was published.
Wright et al. [71] reported initial attempt using CE and FLD, where several reagents
including dansyl chloride, fluorescamine or o-phthaldialdehyde (OPA) were tested
for saxitoxin (STX) labelling. The LOD of single STX-OPA derivative was 0.3
fg/mL, but the method was not robust for quantitative study due to the fact that the
OPA derivative degraded within an hour producing at least five additional neutral
derivatives.
32
1.5 Analysis of ASP toxins
To ensure safety of shellfish for human consumption, European Union specifies a
regulatory limit of 20 mg/kg for ASP toxins (domoic acid, (DA)) [4, 77], which
corresponds with 20 µg/mL in sample extract following the extraction procedure
described in the official method [77]. DA exhibits good UV absorbency at ~240 nm
due to its conjugated diene structure [2, 78, 79], thus allowing UV detection.
Selected HPLC and CE protocols for the analysis of ASP toxins are tabulated in
Table 1.3 with the most significant methods discussed below.
1.5.1 HPLC methods
Lawrence et al. [79, 80] used a RP column to separate DA in shellfish samples
within 6-8 min, with LODs in sample extract of 500-1000 ng/mL using UV
absorbance at 242 nm. The method was later approved as AOAC HPLC-UV Official
Method 991.26 [77]. One of the issues of using a generic detector is the susceptibility
to interferences, a problem overcome by Lopez-Rivera et al. [81] by careful control
of mobile phase pH. Careful selection of acidic conditions (pH 1.5-4.5) was used to
improve the resolution of DA and interference peak (tryptophan) during separation
by changing the extent of ionisation. This approach was susceptible to small changes
in pH, thus a well buffered mobile phase was required. This method allowed the
injection of shellfish extract without the need for SPE clean-up with a LOD better
than 25 ng/mL; approximately 20-times lower than official method [77]. Regueiro et
al. [82] used a monolithic RP column and acidic mobile phase (pH 2.8), allowing
rapid separation under 3 min and resolution between DA and tryptophan (> 10). The
separation time was reduced by two as compared to previous method [81] and a LOD
better than 40 ng/mL was achieved.
33
Table 1.3: Selected HPLC and CE protocols for the analysis of ASP toxins
Protocol Sample Conditions LOD Ref
RP-HPLC-UV Shellfish Column: Supelcosil LC-18
Isocratic elution:(12/88, %v/v) ACN and 2% phosphoric acid (pH 2.5)/
2% phosphoric acid (pH 2.5)
500-1000 ng/mL [79, 80]
RP-HPLC-UV Shellfish Column: Luna-C18
Isocratic elution: (12/87.8/0.2/0.01, %v/v) ACN/water/phosphoric
acid/triethylamine
< 25 ng/mL [81]
RP-HPLC-UV Shellfish Column: Chromolith Performance RP-18e
Isocratic elution: 7.5% ACN in water containing formic acid (pH 2.8)
< 40 ng/mL [82]
Pre-column derivatisation
RP-HPLC-FLD
Seawater,
microalgae
Column: Vydac 201TP
Isocratic elution: ACN containing 0.1% trifluoroacetic acid
0.02 ng/mL [83]
34
Pre-column derivatisation
RP-HPLC-FLD
Microalgae Column: Nova-Pak C18
Gradient elution: mixture of sodium acetate, phosphoric acid,
triethylamine, sodium azide/aqueous ACN and 0.3% acetone
0.1 ng/mL [84]
Pre-column derivatisation
RP-HPLC-FLD
Microalgae,
shellfish
Column: Luna-C18
Isocratic elution: (40/60, %v/v) ACN and 0.1% trifluoroacetic
acid/0.1% trifluoroacetic acid
< 15 ng/mL [85]
Pre-column derivatisation
RP-HPLC-FLD
Shellfish Column: Beckman C18
Isocratic elution:(38/62, %v/v) ACN/ 0.05% trifluoroacetic acid
0.1 ng/mL [86]
Post-column
derivatisation
RP-HPLC-FLD
Shellfish Column: Nucleosil C18
Isocratic elution: (13/87, %v/v) ACN/0.1% trifluoroacetic acid
25 ng/mL [87]
RP-HPLC-MS/MS Sea lion
faeces
Column: C18 column
Gradient elution: 1-95% methanol in 0.1% trifluoroacetic acid
NAa [88]
RP-HPLC-MS/MS Shellfish Column: Luna-2 C18 < 20 ng/mL [89]
35
Gradient elution: ACN/water, both contain 0.05% trifluoroacetic acid
HILIC-MS/MS Shellfish Column: TSK-gel Amide-80
Gradient elution: 2 mM ammonium formate /ACN, both contain 3.6
mM formic acid
3 ng/mL [90]
RP-HPLC-MS/MS Shellfish Column: Kinetex C18
Gradient elution: 0.2% formic acid/methanol and 0.2% formic acid
< 0.08 ng/mL [91]
RP-HPLC-MS/MS Seawater Column: Zorbax Eclipse C18
Gradient elution: ACN and 0.1% formic acid/0.1% formic acid
0.02 ng/mL [92]
CZE-UV Shellfish BGE: N-cyclohexyl-3-aminopropanesulfonic acid pH 10.2 ~µg/mL range [93]
CZE-UV Shellfish BGE: β-cyclodextrine-borate pH 9.0 40 ng/mL [94]
cITP-CZE-UV Shellfish,
microalgae
BGE: 20 mM caproic acid+ 20 mM β-alanine+5% methanol+0.1%
hydroxypropylmethylcellulose
LE: 10 mM hydrochloric acid+20 mM β-alanine+0.05%
2 ng/mL [95]
36
hydroxyethylcellulose (pH 3.5)
TE: 5 mM caproic acid+5% methanol
CE-EIA-EC Shellfish BGE: 1 mM H2O2 + 10 mM Briton-Robinson buffer+1%
polyvinylpyrolidone (pH 5.0)
0.02 ng/mL [96]
a) NA Not available
37
In contrast to the analysis in shellfish samples, determining trace level of DA in
microalgae or seawater requires greater sensitivity and the use of FLD is therefore
popular, although as DA is not natively fluorescent it requires derivatisation.
Pocklington et al. [83] used 9- fluorenylmethyl chloroformate for pre-column
derivatisation and FLD (ex = 264 nm, λem= 313 nm). Derivatised DA was eluted
within 15 min, with a LOD of 0.02 ng/mL, five orders of magnitude lower than the
official method [77]. However, prior to analysis, excess reagent had to be extracted
which further lengthened the analysis procedure. Sun and Wong [84] used 6-
aminoquinolyl-N- hydroxysuccinimidyl carbamate (AQC) with FLD (ex = 290 nm,
λem = 395 nm). The derivatised DA eluted 20 min after excess AQC, which made
extraction of excess reagent unnecessary. The LOD was 0.1 ng/mL, however the
separation time was long (31 min). James et al. [85] and Chan et al. [86] both used 4-
fluoro-7-nitrobenzofurazan (NBD-F) to derivatise DA for FLD (ex = 470 nm, λem =
530 nm). Derivatised DA eluted after the reagent’s hydrolysed products and the total
separation was < 15 min. The reported LODs were 15 ng/mL and 0.1 ng/mL
respectively, significant difference between these due to the combination of an
additional preconcentration step (SPE based on amorphous titania material) in the
latter method. Maroulis et al. [87] used a similar reagent, 4-chloro-7-nitrobenzo-2-
oxa-1, 3-diazole (NBD-Cl), for post-column derivatisation of DA. A two-step
derivatisation was used in which DA was reacted with NBD-Cl while the second
reaction degraded the interfering reagent hydrolysis product to allow detection of the
toxin. A LOD of 25 ng/mL was achieved, which was comparable to pre-column
derivatisation using NBD-F [85], however the post-column method was more
reliable due to less interference of hydrolysed products.
38
For the reasons stated for the detection of PSTs, MS has become a dominant
detection method to accurately determine DA in various sample matrices [88, 89,
97]. A single method that allows hydrophilic multi-toxin analysis is also desirable
because DA is usually co-extracted with PSTs. This type of method was reported by
Ciminiello et al. [90] who used HILIC-MS/MS for simultaneous separation of both
toxin class. Using multiple reactions monitoring (MRM) acquisition at positive
mode, DA was eluted at the chromatography front, which was at least 8 min before
the first PST peak with a better LOD for DA in sample extract (3 ng/mL) than the
FLD method of both James et al. [85] and Maroulis et al. [87].
Despite the power of using MS for detection, the occurrence of matrix effects is an
issue and always demands efficient sample purification. This was usually done using
laborious off-line strong anion-exchange SPE procedure. Therefore, Regueiro et al.
[91] developed an automated on-line SPE-HPLC-MS/MS using column switching
for rapid analysis of DA in shellfish samples. The total analysis time including
sample injection, on-line clean-up and chromatographic separation was less than 11
min, which was relatively shorter compared to using off-line SPE. An excellent LOD
in the sample extract better than 0.08 ng/mL was achieved. Ion exchange SPE was
not advised for highly saline samples due to the high ionic strength. Therefore, de la
Iglesia et al. [92] used C18 SPE to desalt and preconcentrate DA from acidified
seawater. Complete resolution between DA and its isomers was obtained in less than
3 min with LOD for DA of 0.02 ng/mL.
In short, hydrophilic multi-toxin analysis can be realised through the usage of
HILIC-MS/MS. This substantially reduces the total analysis time and is seen to
improve turnaround time when applied as regular monitoring method. Similar to
39
HPLC methods for PSTs, coupling of HPLC and MS is not suited to portability for
the foreseeable future.
1.5.2 CE methods
DA can be separated by electrophoresis due to possessing three carboxy groups and
an amino group, in which the charge states (cation or anion) were determined by pKa
values (1.85, 4.47, 4.75 and 10.6) and solution pH. CE was initially demonstrated by
Nguyen et al. [93] as an alternative method to LC. A basic buffer (N-cyclohexyl-3-
aminopropanesulfonic acid, pH 10.2) was used, in which DA was separated < 10 min
and detected by UV at 242 nm; LOD in the µg/mL level was achieved. Following
this, Zhao et al. [94] used β-cyclodextrin added to the BGE (borate, pH 9.0) to
separate DA and its isomers in shellfish samples. Two step SPE using anion and
cation exchange columns were used for clean-up and preconcentration, in which the
methods combined to provide a LOD in sample extract of 40 ng/mL. Kvasnika et al.
[95] developed a method combining capillary isotachophoresis (cITP) with CZE-UV
for on-column sample clean-up and to improve the sensitivity. The electrolyte
compositions were complicated, involving caproic acid in β-alanine with methanol
and hydroxypropyl methylcellulose as BGE, hydrochloric acid, β-alanine and
hydroxyethylcellulose as LE, and caproic acid with methanol as TE. This method
eliminated SPE steps because cITP enabled sample clean-up through migration of
matrix interference from the separation capillary; thus the sample was cleaned before
subsequent injection into analytical capillary (zone electrophoresis step). This
method afforded LOD of 2 ng/mL which was 20 times lower than previous method
[94].
40
Zhang and Zhang [96] developed a novel CE-based enzyme immunoassay (CE-EIA)
in which it was combined with electrochemical detection (ECD), another highly
miniaturisable detection method. The method was based on non-competitive
immunoreaction between free DA antigen (Ag) and excess horseradish peroxidase
(HRP)-labelled anti-DA antibody tracer (Ab*) in liquid phase. The bound enzyme-
labelled complex (Ab*-Ag) and unbound Ab* were separated by CE and the system
of HRP catalysing hydrogen peroxide/o-aminophenol reaction was adopted for ECD.
As shown in Figure 1.3, DA was separated within 4 min with LOD as low as 0.02
ng/mL, which was at least 100 times lower than CE-UV methods [94, 95]. The
combination of efficiency from CE, specificity of immunoassay, and sensitivity and
miniaturisable property of ECD is deemed to be perfect for future development of
portable DA analyser.
While FLD and MS are used widely with HPLC for DA, they have not been reported
with CE.
41
Figure 1.3 Electropherograms of Ab* and the Ab*–Ag immunocomplex. The
concentration of HRP-labelled anti-DA (Ab*) were constant. Parts a, b, c, and d
correspond to 0, 0.2, 0.8, and 3.0 ng/mL DA, respectively. The immune sample was
diluted 10-fold with 10 mM pH 5.0 Briton-Robinson buffer containing 0.2% SDS.
Peak 1: free Ab*; peak 2: Ab*–Ag immunocomplex. Reproduced from [96] with
permission.
42
1.6 Analysis of DSP toxins
Diarrhetic shellfish poisoning (DSP) was first reported in The Netherlands in 1960s
and is now known to occur worldwide [98-107]. Analyses of DSP toxins (i.e.
okadaic acid (OA) and dinophysistoxins (DTXs)) often include other lipophilic
biotoxins such as pectenotoxins (PTXs), yessotoxins (YTXs), azaspiracids (AZAs),
gymnodimines and spirolides; thus, instrumental analytical methods are usually
preferred for routine monitoring [108]. The permitted level for DSP toxins in
shellfish samples is 160 µg/kg [109, 110], which corresponds to 80 ng/mL of DSP
toxins in sample extract following the extraction procedure described in the official
method [109]. The level set for YTXs is 1000 µg/kg (500 ng/mL), 160 µg/kg (80
ng/mL) for AZAs [109, 110], while spirolides and gymnodimines are not yet
regulated. Selected HPLC and CE procedures for the analysis of DSP toxins are
given in Table 1.4 and the most significant papers discussed below.
1.6.1 HPLC methods
Due to the lipophilic properties of DSP toxins, RP HPLC methods can be used,
particularly after derivatisation for FLD. For instance, pre-column derivatisation of
DSP toxins with 4-bromomethyl-7 methoxycoumarin (detected as coumarine-esters
at λex = 325 nm, λem = 390 nm) [111] or 3-bromomethyl-6,7-dimethoxy-1-methyl-
2(1H)-quinoxalinone (λex = 370 nm, λem = 450 nm) [112] gave sensitivities in the low
ng/mL level (4-50 ng/mL). While the HPLC-FLD methods provide excellent
sensitivity, nonetheless, finding the most suitable derivatisation conditions that work
simultaneously for multiple lipophilic biotoxins is always challenging; thus, making
LC-MS a method of choice.
43
The separation of multiple lipophilic biotoxins using HPLC with MS/MS can be
performed in a number of different conditions: acidic, less extreme pH (near neutral
or slightly alkaline) or alkaline, in which each condition can change the extent of
toxin ionisation to achieve good separation and detection for all toxins. Following
initial work by Quilliam et al. [113], Goto et al. [114] developed comprehensive
multi-toxin separations by sequentially changing four different RP columns
isocratically eluted with different acidic mobile phases (Table 1.4).
44
Table 1.4 Selected HPLC and CE protocols for the analysis of DSP toxins
Protocol Sample Conditions LOD Ref
Pre-column derivatisation
RP-HPLC-FLD
Shellfish Column: Supelcosil-C18 DB
Isocratic elution: (58/4.5/37.5, %v/v) ACN/methanol/water
4 ng/mL [111]
Pre-column derivatisation
RP-HPLC-FLD
Microalgae Column: Discovery C18
Isocratic elution: (47/53, %v/v) ACN/water
50 ng/mL [112]
RP-HPLC-MS Shellfish Column: Symmetry C18, Capcellpak C18, Inertsil ODS-3 and Develosil
ODS
Isocratic elution:
1) (7/3, %v/v) ACN/0.05% acetic acid (OA, PTX1, PTX2, PTX6,
PTX2sa)
2) (98/2, %v/v) methanol/2.5 mM acetic acid (DSP toxin analogues)
3) (8/2, %v/v) methanol/0.2 M ammonium acetate (YTX, hydroxyYTX)
0.6-9 ng/mL [114]
RP-HPLC-MS/MS Shellfish Column: Luna C18 0.1-1.0 ng/mL [115]
45
Gradient elution: ACN/33 mM ammonium hydroxide and 500 mM formic
acid (pH 2.0)
RP-HPLC-MS Shellfish Column: Hypersil C8
Gradient elution: 5 mM ammonium acetate (pH 6.8)/95% ACN
0.50-12 ng/mL [116]
RP-HPLC-MS/MS Shellfish Column: Gemini NX
Gradient elution: 5 mM ammonium bicarbonate (pH 7.9)/5 mM
ammonium bicarbonate in 95% ACN
0.2-0.3 ng/mL [117]
RP-HPLC-MS/MS Shellfish Column: X-Bridge C18
Gradient elution: 6.7 mM ammonium hydroxide (pH 11)/ 6.7 mM
ammonium hydroxide (pH 11) in 90% ACN
0.08-0.9 ng/mL [118]
RP-HPLC-MS/MS Shellfish X-Bridge C8
Gradient elution of mobile phases as in method by McNabb et al.[115] or
Stobo et al. [116] or These et al. [117] or Gerssen et al.[118]
0.15-38 ng/mL [119]
MEKC-UV Shellfish BGE: 12.5 mM borate + 20 mM SDS (pH 9.2) 5 ng/mL [120]
46
MEKC-UV Shellfish,
microalgae
BGE: 10 mM disodium phosphate + 35 mM SDS +20% methanol
(pH 8.5)
200 ng/mL [121]
CZE-MS Shellfish BGE: 10 mM ammonium acetate (pH 8.5) 20 ng/mL [121]
47
The total separation time was 60 min. DSP toxins, YTXs and PTXs were monitored
with [M-H]- ion, [M-Na]- ion and [M+NH4]+ ion, respectively. In combination with
liquid-liquid extraction (LLE) and several SPE clean-ups, the LODs in sample
extract range from 0.6-9 ng/mL. Nonetheless, this method was not favourable for
routine monitoring due to the requirement of multiple chromatographic conditions.
McNabb et al. [115] performed HPLC-MS/MS under acidic condition using a single
RP column with gradient elution mixture of ACN and ammonium formate.
Separation of all lipophilic toxins was achieved within 30 min. While the peak
shapes for some of the toxins was poor, the method’s simplicity and sensitivity (LOD
in sample extract of 0.1-1.0 ng/mL) were significant attributes that encouraged use
by others for research and routine monitoring [122-125].
Two methods using less extreme pH chromatographic conditions were also
developed which could help to extend the lifetime of RP column compared to
methods with extreme acidic or alkaline conditions. Stobo et al. [116] used
ammonium acetate/ACN (pH 6.8) as mobile phase, and gradient eluted lipophilic
toxins from RP column within 13 min. This method showed good performance with
LODs in sample extract ranging from 0.50-12 ng/mL and offered high sample
throughput capacity. These et al. [117] used a slightly alkaline (pH 7.9) mobile
phase comprised of ammonium carbonate and ACN. Applied together with SPE and
LLE, the method could be used to quantitate toxins at LODs of 0.2-0.3 ng/mL.
However, these two methods did not demonstrate the ability to detect two emerging
lipophilic toxins classes, the gymnodimines and spirolides.
Gerssen et al. [118, 126] developed a method based on extreme alkaline
chromatographic condition for all lipophilic toxins including gymnodimines and
spirolides. The separation was conducted using mobile phase comprising aqueous
48
ACN/ammonium hydroxide at pH 11. The high pH required a column that can
withstand extreme alkaline condition and could maintain reasonable toxins
separation, thus the X-Bridge C18 column was used. In contrast with acidic
chromatographic conditions, OA, DTXs and YTXs and AZAs are negatively charged
while PTXs and gymnodimines remain neutral at alkaline pH; thus the elution order
was different. A major advantage was that the toxins could be clustered in three
MRM time windows allowing MS/MS detection without continuous electrospray
ionisation polarity switching. All 28 toxins were separated within 16 min with
improved LODs in sample extracts (0.08-0.9 ng/mL) compared to the acidic
condition. This method has been adapted by several research or monitoring
laboratories [107, 127-130]. Recently, Garcia-Altares et al. [119] implemented and
compared all four chromatographic conditions for HPLC-MS/MS of multi lipophilic
toxins. The chromatograms of standard toxin mixtures are shown in Figure 1.4. It
was concluded that alkaline condition provided higher sample throughput, lower
limit of quantitation and the best sensitivity and method accuracy.
The development of analytical separation methods for DSP toxins and other
important lipophilic toxins have been specifically focused to multi-toxin analysis, in
which HPLC-MS, possessing the utmost merit to resolve and identify each toxin was
used. High throughput analyses will be a highly valuable tool considering the
increase of the number of samples to be processed during routine monitoring.
49
Figure 1.4 Example of chromatograms of DSP toxins and lipophilic toxins under four
chromatographic conditions. Chromatographic conditions; (a) Acidic (pH 2.0):
Gradient elution with 2 mM ammonium formate/2 mM ammonium formate in 95%
ACN, both acidified with 50 mM formic acid; (b) Slight alkaline (pH 7.9): 5 mM
ammonium bicarbonate/ 5 mM ammonium bicarbonate in 95% ACN; (c) Near
neutral (pH 6.8): Gradient elution with 5 mM ammonium acetate/5 mM ammonium
acetate in 95% ACN; (d) Alkaline (pH 11): 6.7 mM ammonia/ 6.7 mM ammonia in
95% ACN. Reproduced from [119] with permission.
50
1.6.2 CE methods
DSP toxins (i.e. OA) and other lipophilic toxins (i.e. YTXs) are usually analysed
under separate CE conditions due to their different charge state in acidic or alkaline
medium. OA possess carboxyl group in its structure that shows molar absorptivity at
200 nm. This enabled Bouaïcha et al. [120] to use CE with UV detection for the
determination of OA; a detection method never utilised in HPLC for this particular
toxin. A borate buffer (pH 9.5) with 20 mM SDS could separate the negatively
charged OA from other sample matrix components in under 8 min. The LOD was 5
ng/mL, which was comparable to sensitivity obtained using HPLC-FLD method
[111]. de la Iglesia et al. [121] evaluated and compared micellar electrokinetic
chromatography-UV (MEKC-UV) and CZE-MS for the analysis of YTXs. In
MEKC-UV, a BGE comprising disodium phosphate (pH 8.5) and 35 mM sodium
dodecyl sulfate (SDS) was able to separate YTX and its analogue within 15 min. The
identity of the analogue was further confirmed by CZE-MS in which poor selectivity
inadvertently caused by CZE was encountered by the ability of MS to monitor
individual toxins ion. With the aid of sample stacking, the LOD of YTX was 200
ng/mL for MEKC-UV and 20 ng/mL for CZE-MS.
1.7 The analysis of NSP toxins and CFP toxins
Neurotic shellfish poisoning (NSP) cases have been primarily reported in the United
States (Gulf of Mexico and Florida) [131, 132] and New Zealand [133]. Since these
events, the Food and Drug Administration of United States set a regulatory limit of
800 µg/kg NSP toxins (brevetoxins) in shellfish [134], which corresponds to 800
ng/mL in sample extract following the extraction procedure described in the official
method [134, 135]. On the other hand, ciguatera fish poisoning (CFP) was initially
51
confined to tropical and subtropical areas [2], but is now ranked as one of the most
common food-borne illnesses that represents a global threat to human health [136].
Currently, there are no regulatory limits for ciguatoxin and its analogues in fish, but
it is required that contaminated fish are not placed in the market [136]. Compilation
of different HPLC conditions used for the analysis of NSP toxins and CFP toxins are
summarised in Table 1.5.
1.7.1 HPLC methods
HPLC with UV detection was initially used for NSP toxins (brevetoxins) [137, 138];
however the preference was to use HPLC-MS to obtain more structural information
[131, 132, 139, 140]. Initial HPLC-MS method developed by Hua et al. [139] used
methanol/water as mobile phase for elution of brevetoxins in RP column. Several
analogues of brevetoxins in microalgae extract were successfully identified within 35
min; however the LODs were only 40-900 ng/mL. To enable detection of trace level
brevetoxins in seawater and microalgae samples, preconcentration procedure has to
be incorporated. Twiner et al. [140] used SPE with C18 material whereby extensive
optimisation was conducted to find suitable elution solvent and drying steps.
Seawater matrices were removed, 20 mL methanol used for elution and extracts were
dried by centrifugation before been resuspended in 1 mL methanol. The toxins were
separated by RP HPLC-MS using acidified ACN/water as mobile phase. The method
afforded a LOD down to 1 ng/mL, and was shown to efficiently quantitate
brevetoxins well below the regulatory limit.
Additional brevetoxins analogues are expected to exist and MS/MS is essential to
enable structural characterisation and detection of these toxins. Following their
52
previous work on HPLC-MS of brevetoxins [141, 142], McNabb et al. [143] recently
developed an improved HPLC-MS/MS for shellfish samples. This includes gradient
elution using four different mobile phases (Table 1.5) to enable simultaneous
separation of all toxins. Six brevetoxins including new analogues were successfully
separated and detected within 15 min, with LODs in sample extract ranging from 4-
10 ng/mL.
53
Table 1.5 Selected HPLC protocols for the analysis of NSP toxins and CFP toxins
Toxin Sample Protocol Conditions LOD Ref
NSP toxins Microalgae RP-HPLC-MS/MS Column: Spherisorb C18
Isocratic elution: (85/15, %v/v) methanol/water
40-900 ng/mL [139]
NSP toxins Microalgae RP-HPLC-MS/MS Column: Luna C8
Gradient elution: water/ACN with 0.1% acetic acid
< 1 ng/mL [140]
NSP toxins Shellfish RP-HPLC-MS/MS Column: BDS Hypersil C8
Gradient elution:
1) 50% methanol/2.5% isopropanol
2) 97.5% methanol/2.5% isopropanol
3)30 mM ammonium formate/470 mM formic acid
4) 90% ACN
4-10 ng/mL [143]
CFP toxins Fish RP-HPLC-MS Column: PRP-1 column
Gradient elution: ACN/0.1% trifluoroacetic acid (pH 2.0)
100 ng/mL [144]
54
CFP toxins Fish RP-HPLC-MS/MS Column: Vydac TP52
Gradient elution: 0.05% trifluoroacetic acid /90% ACN and
0.05% trifluoroacetic acid
10-70 ng/mL [145]
CFP toxins Fish RP-HPLC-MS/MS Column: Luna C18
Gradient elution: 2 mM ammonium formate /95% ACN and
2 mM ammonium formate
< 25 ng/mL [146]
CFP toxins Fish RP-HPLC-MS/MS Column: Luna C18
Gradient elution: 0.1% formic acid /ACN and 0.1% formic
acid
< 30 ng/mL [147]
CFP toxins Fish RP-UPLC-MS/MS Column: NAa
Gradient elution: 2 mM ammonium formate /95% ACN and
2 mM ammonium formate
250-2500 ng/mL [148]
a) NA Not available
55
Since the introduction of MS method to characterise ciguatoxins [144, 149, 150], this
method has been studied in depth for confirmatory separation method [145] and is
becoming more popular than pre- or post-column oxidation RP-HPLC-FLD methods
[151-153]. However, laborious samples extraction and clean-up steps must be
conducted to obtain cleaner chromatogram in HPLC-MS. Lewis et al. [145] used
acetone extraction for sample (50-100 g) and subsequent multiple drying step. The
extracts were subjected to two steps LLE (methanol/hexane) to remove low polarity
lipid, additional three steps LLE (ethanol/ether) to extract toxins and finally clean-up
with a silica column. Acidified ACN as mobile phase successfully eluted the
polyether toxins on a RP column. The toxins were later detected by MS/MS by
selecting dominant fragment ions [M+NH4]+, in which the LOD in sample extract
was within ng/mL level (10-70 ng/mL). Lack of rapid extraction and clean-up steps
usually limits the usefulness of this method; improvements to the preparative steps
gained priority since then. The same group [146] developed a rapid extraction
method to simplify detection and quantitation. The method streamlined the extraction
of toxin by (i) replacing initial multiple acetone extraction-drying step with one-step
extraction (small sample portion, < 2 g) and LLE (methanol/hexane), (ii) reducing
number of transfers and drying steps, (iii) using centrifugation LLE and (iv)
incorporation of two orthogonal SPE clean-up steps (RP and normal phase SPE). The
turnaround time between analyses was significantly reduced to 12 min, and
ciguatoxins < 25 ng/mL could be detected. While not an official standard method,
Lewis’s method has been widely utilised by other research groups [147, 148, 154]
and public health analytical service laboratories [155].
There are currently no published reports of CE of NSP toxins and CFP toxins;
nevertheless seeing CE development works on these emerging toxins would be
56
interesting in the near future, particularly when implemented in a highly portable
analyser.
1.8 Conclusions
This chapter highlights the developments of LC and CE methods for the analysis of
major phycotoxins as alternative methods to move away from animal assay. Despite
the chemical diversity of each toxin class that contributes to the complexity of
analysis, accurate identification of individual toxin in almost all toxin classes can be
achieved due to inherent specificity of separation and detection methods. In
summary, almost all methods developed for major phycotoxins are able to detect
toxins at the regulatory limit. Due to frequent occurrence of poisonings, there is a
possibility that the regulatory limit of toxins in seafood products are to be reduced;
therefore methods with capability for trace detection (lower LOD) are sought. Many
efforts are seen on enhancing the sensitivity and efficiency of the methods (i.e. the
use of efficient chromatographic/electrophoretic conditions and highly sensitive
FLD).
Due to seafood samples often containing several toxins from multiple classes of
toxins, considerable effort is being devoted to develop multi-toxin analysis of
hydrophilic or lipophilic toxin classes. In this case, sophisticated MS method will
soon be the new ‘gold standard’, in line with its high resolving power that can detect
structurally different existing and emerging toxins. This has catalysed the widespread
combination of HPLC with MS through HILIC-MS for hydrophilic toxins (PSTs and
ASP toxins) and RP HPLC-MS for lipophilic toxins (DSP toxins, YTXs, PTXs,
AZA, gymnodimines and spirolides). Successful multi-toxin analysis provides high
57
throughput analysis for multiple toxin classes. High throughput analyses will be a
highly valuable tool considering the increase of the number of samples to be
processed during routine monitoring. With the increasing demands of rapid
verification of contamination in seafood samples, trends are seen to develop portable
methods so that on-site detection can be realised. While CE has the potential to be
used in portable instrumentation, portability of HPLC with either FLD or MS
detection systems is not likely to happen in the near future. Finally, from the result of
the literatures, it is expected that analytical instrumental methods will be able to
eliminate the dependence on animal assay in the near future.
58
1.9 References
[1] Quilliam, M. A., J. Chromatogr. A 2003, 1000, 527-548.
[2] Rossini, G. P., Hess, P., EXS 2010, 100, 65-122.
[3] Panel, E., EFSA J. 2009, 1019, 1-76.
[4] Quilliam, M. A., J. AOAC Int. 1999, 82, 773-781.
[5] Bhakuni, D. S., Rawat, D. S., Bioactive marine natural products, Springer, New
York 2005.
[6] Oshima, Y., J. AOAC Int. 1995, 78, 528-532.
[7] Schantz, E. J., Lynch, J. M., Vayvada, G., Matsumoto, K., Rapoport, H.,
Biochemistry (Mosc). 1966, 5, 1191-1195.
[8] Negri, A. P., Jones, G. J., Toxicon 1995, 33, 667-678.
[9] Hallegraeff, G. M., McCausland, M. A., Brown, R. K., J. Plankton Res. 1995, 17,
1163-1176.
[10] Narahashi, T., Haas, H. G., J. Gen. Physiol. 1968, 51, 177-198.
[11] Bates, S. S., Can. J. Fish. Aquat. Sci. 1989, 46, 1203-1215.
[12] Berman, F. W., Murray, T. F., J. Neurochem. 1997, 69, 693-703.
[13] MacKenzie, L., Beuzenberg, V., Holland, P., McNabb, P., Suzuki, T., Selwood,
A., Harmful Algae 2005, 4, 75-85.
[14] Bialojan, C., Takai, A., Biochem. J 1988, 256, 283-290.
[15] Haystead, T. A. J., Sim, A. T. R., Carling, D., Honnor, R. C., Tsukitani, Y.,
Cohen, P., Hardie, D. G., Nature 1989, 337, 78-81.
[16] Dominguez, H. J., Paz, B., Daranas, A. H., Norte, M., Franco, J. M., Fernández,
J. J., Toxicon 2010, 56, 191-217.
59
[17] Miles, C. O., Wilkins, A. L., Munday, R., Dines, M. H., Hawkes, A. D., Briggs,
L. R., Sandvik, M., Jensen, D. J., Cooney, J. M., Holland, P. T., Quilliam, M. A.,
MacKenzie, A. L., Beuzenberg, V., Towers, N. R., Toxicon 2004, 43, 1-9.
[18] Aune, T., Sorby, R., Yasumoto, T., Ramstad, H., Landsverk, T., Toxicon 2002,
40, 77-82.
[19] Tubaro, A., Sosa, S., Carbonatto, M., Altinier, G., Vita, F., Melato, M., Satake,
M., Yasumoto, T., Toxicon 2003, 41, 783-792.
[20] Tubaro, A., Sosa, S., Altinier, G., Soranzo, M. R., Satake, M., Della Loggia, R.,
Yasumoto, T., Toxicon 2004, 43, 439-445.
[21] Satake, M., MacKenzie, L., Yasumoto, T., Nat. Toxins 1997, 5, 164-167.
[22] Draisci, R., Ferretti, E., Palleschi, L., Marchiafava, C., Poletti, R., Milandri, A.,
Ceredi, A., Pompei, M., Toxicon 1999, 37, 1187-1193.
[23] Rhodes, L., McNabb, P., De Salas, M., Briggs, L., Beuzenberg, V., Gladstone,
M., Harmful Algae 2006, 5, 148-155.
[24] Furey, A., O'Doherty, S., O'Callaghan, K., Lehane, M., James, K. J., Toxicon
2010, 56, 173-190.
[25] Ito, E., Satake, M., Ofuji, K., Kurita, N., McMahon, T., James, K., Yasumoto,
T., Toxicon 2000, 38, 917-930.
[26] Vilariño, N., Fonfría, E. S., Molgó, J., Aráoz, R., Botana, L. M., Anal. Chem.
2009, 81, 2708-2714.
[27] Poli, M. A., Mende, T. J., Baden, D. G., Mol. Pharmacol. 1986, 30, 129-135.
[28] Yasumoto, T., Chemical Records 2001, 1, 228-242.
[29] Litaker, R. W., Vandersea, M. W., Faust, M. A., Kibler, S. R., Nau, A. W.,
Holland, W. C., Chinain, M., Holmes, M. J., Tester, P. A., Toxicon 2010, 56, 711-
730.
60
[30] Humpage, A. R., Magalhaes, V. F., Froscio, S. M., Anal. Bioanal. Chem. 2010,
397, 1655-1671.
[31] AOAC, Official Method 2011.27 First Action 2011: Paralytic shellfish
poisoning in shellfish by receptor binding assay, AOAC International, Gaithersburg,
MD 2012.
[32] Baram, G. I., J. Chromatogr. A 1996, 728, 387-399.
[33] Nelson, M. A., Gates, A., Dodlinger, M., Hage, D. S., Anal. Chem. 2004, 76,
805-813.
[34] Peña, R., Alcaraz, M. C., Arce, L., Ríos, A., Valcárcel, M., J. Chromatogr. A
2002, 967, 303-314.
[35] Maragos, C. M., Appell, M., J. Chromatogr. A 2007, 1143, 252-257.
[36] Arribas, A. S., Bermejo, E., Zapardiel, A., Téllez, H., Rodríguez-Flores, J.,
Zougagh, M., Ríos, Á., Chicharro, M., Electrophoresis 2009, 30, 499-506.
[37] Güray, T., Tuncel, M., Uysal, U. D., Oncu-Kaya, E. M., J. Liq. Chromatogr.
Related Technol. 2013, 36, 1366-1378.
[38] Shintani, T., Yamada, K., Torimura, M., FEMS Microbiol. Lett. 2002, 210, 245-
249.
[39] Alarcón, B., García-Cañas, V., Cifuentes, A., González, R., Aznar, R., J. Agric.
Food. Chem. 2004, 52, 7180-7186.
[40] Hu, A., Lo, A. A. L., Chen, C. T., Lin, K. C., Ho, Y. P., Electrophoresis 2007,
28, 1387-1392.
[41] Corzo, G., Villegas, E., Lee, H. S., Nakajima, T., Protein Peptide Lett. 2001, 8,
375-383.
[42] Graham, R. L. J., McClean, S., O'Kane, E. J., Theakston, D., Shaw, C.,
Biochem. Biophys. Res. Commun. 2005, 333, 88-94.
61
[43] Kokot, Z. J., Matysiak, J., Urbaniak, B., Dereziński, P., Anal. Bioanal. Chem.
2011, 399, 2487-2494.
[44] Leitão, A. A., de Lacerda Miranda, P. C. M., Simionato, A. V. C.,
Electrophoresis 2011, 32, 1074-1079.
[45] Meagher, R. J., Hatch, A. V., Renzi, R. F., Singh, A. K., Lab Chip 2008, 8,
2046-2053.
[46] Borowsky, J., Collins, G. E., Analyst 2007, 132, 958-962.
[47] Beroovici, M., Kaigala, G. V., Baokhouse, C. J., Santiago, J. G., Anal. Chem.
2010, 82, 1858-1866.
[48] Fruetel, J., Horn, B., West, J. A., Stamps, J., Vandernoot, V., Stoddard, M.,
Renzi, R., Padgen, D., in: Baba, Y., Shoji, S., van den Berg, A. (Eds.), Micro Total
Analysis Systems 2002, Springer Netherlands 2002, pp. 524-526.
[49] AOAC, Official Method 2005.06 Paralytic shellfish poisoning toxins in
shellfish.Pre-chromatographic oxidation and liquid chromatography with
fluorescence detection, AOAC International, Gaithersburg, MD 2005.
[50] EFSA, EFSA J. 2009, 1019, 1-76.
[51] AOAC, Official Method 959.08. Paralytic shellfish poison. Biological method.
Final action, AOAC International, Gaithersburg, MD 2005.
[52] Bates, H. A., Rapoport, H., J. Agric. Food. Chem. 1975, 23, 237-239.
[53] Oshima, Y., Machida, M., Sasaki, K., Agric. Biol. Chem. 1984, 48, 1707-1711.
[54] Van De Riet, J., Gibbs, R. S., Muggah, P. M., Rourke, W. A., MacNeil, J. D.,
Quilliam, M. A., J. AOAC Int. 2011, 94, 1154-1176.
[55] Lawrence, J. F., Menard, C., J. Assoc. Off. Anal. Chem. 1991, 74, 1006-1012.
[56] Lawrence, J. F., Niedzwiadek, B., Menard, C., De Rojas Astudillo, L., Biré, R.,
Burdaspal, P. A., Ceredi, A., Davis, B., Dias, E., Eaglesham, G., Franca, S.,
62
Gallacher, S., Graham, D., Hald, B., Heinze, L., Hellwig, E., Jonker, K. M., Kapp,
K., Krys, S., Kurz, K., Lacaze, J. P., Gago Martinez, A., McNabb, P., Ménard, C.,
Milandri, A., Nsengiyumva, C., Pereira, P., Pineiro, N., Poletti, R., Riddell, G.,
Selwood, A., Stern, A., Tiebach, R., Van Den Top, H., Wezenbeek, P., Yen, I. C., J.
AOAC Int. 2005, 88, 1714-1732.
[57] He, H. Z., Li, H. B., Jiang, Y., Chen, F., Anal. Bioanal. Chem. 2005, 383, 1014-
1017.
[58] DeGrasse, S. L., DeGrasse, J. A., Reuter, K., Toxicon 2011, 57, 179-182.
[59] Hatfield, R. G., Turner, A. D., J. AOAC Int. 2012, 95, 1089-1096.
[60] Dell'Aversano, C., Hess, P., Quilliam, M. A., J. Chromatogr. A 2005, 1081,
190-201.
[61] Sayfritz, S. J., Aasen, J. A. B., Aune, T., Toxicon 2008, 52, 330-340.
[62] Watanabe, R., Matsushima, R., Harada, T., Oikawa, H., Murata, M., Suzuki, T.,
Food Addit. Contam. 2013, 30, 1351-1357.
[63] Zhuo, L., Yin, Y., Fu, W., Qiu, B., Lin, Z., Yang, Y., Zheng, L., Li, J., Chen, G.,
Food Chem. 2013, 137, 115-121.
[64] Thibault, P., Pleasance, S., Laycock, M. V., J. Chromatogr. 1991, 542, 483-501.
[65] Locke, S. J., Thibault, P., Anal. Chem. 1994, 66, 3436-3446.
[66] Buzy, A., Thibault, P., Laycock, M. V., J. Chromatogr. A 1994, 688, 301-316.
[67] Piñeiro, N., Leão, J. M., Gago Martínez, A., Rodríguez Vázquez, J. A., J.
Chromatogr. A 1999, 847, 223-232.
[68] Gago-Martínez, A., Leão, J. M., Piñeiro, N., Carballal, E., Vaquero, E.,
Nogueiras, M., Rodríguez-Vázquez, J. A., Int. J. Environ. Anal. Chem. 2003, 83,
443-456.
63
[69] Wu, Y., Yam Tat Ho, A., Qian, P. Y., Leung, K. S. Y., Cai, Z., Lin, J. M., J.
Sep. Sci. 2006, 29, 399-404.
[70] Pleasance, S., Thibault, P., Kelly, J., J. Chromatogr. 1992, 591, 325-339.
[71] Wright, B. W., Ross, G. A., Smith, R. D., J. Microcolumn Sep. 1989, 1, 85-89.
[72] AOAC, Official Method 2011.02 First Action 2011: Paralytic shellfish toxins in
mussels, clams, scallops and oysters by liquid chromatography post-column
oxidation, AOAC International, Gaithersburg,MD 2012.
[73] Rodríguez, P., Alfonso, A., Botana, A. M., Vieytes, M. R., Botana, L. M.,
Toxicon 2010, 56, 448-457.
[74] DeGrasse, S. L., van de Riet, J., Hatfield, R., Turner, A., Toxicon 2011, 57, 619-
624.
[75] Fallon, W. E., Shimizu, Y., J. Environ. Sci. Health., A 1977, A12, 455-464.
[76] Schantz, E. J., Mold, J. D., Howard, W. L., Bowden, J. P., Stanger, D. W.,
Lynch, J. M., Wintersteiner, O. P., Dutcher, J. D., Walters, D. R., Riegel, B., Can. J.
Chem. 1961, 39, 2117-2123.
[77] AOAC, Official Method 991.26: Domoic acid in mussels. Liquid
chromatographic method, AOAC International, Gaithersburg, MD 1997.
[78] Quilliam, M., Sim, P. G., McCulloch, A. W., McInnes, A. G., Int. J. Environ.
Anal. Chem. 1989, 36, 139-154.
[79] Lawrence, J. F., Charbonneau, C. F., Menard, C., Quilliam, M. A., Sim, P. G., J.
Chromatogr. 1989, 462, 349-356.
[80] Lawrence, J. F., Ménard, C., Fresenius J. Anal. Chem. 1991, 339, 494-498.
[81] López-Rivera, A., Suárez-Isla, B. A., Eilers, P. P., Beaudry, C. G., Hall, S.,
Fernández Amandi, M., Furey, A., James, K. J., Anal. Bioanal. Chem. 2005, 381,
1540-1545.
64
[82] Regueiro, J., Álvarez, G., Mauriz, A., Blanco, J., Food Chem. 2011, 127, 1884-
1891.
[83] Pocklington, R., Milley, J. E., Bates, S. S., Bird, C. J., De Freitas, A. S. W.,
Quilliam, M. A., Int. J. Environ. Anal. Chem. 1990, 38, 351-368.
[84] Sun, T., Wong, W. H., J. Agric. Food. Chem. 1999, 47, 4678-4681.
[85] James, K. J., Gillman, M., Lehane, M., Gago-Martinez, A., J. Chromatogr. A
2000, 871, 1-6.
[86] Chan, I. O. M., Tsang, V. W. H., Chu, K. K., Leung, S. K., Lam, M. H. W., Lau,
T. C., Lam, P. K. S., Wu, R. S. S., Anal. Chim. Acta 2007, 583, 111-117.
[87] Maroulis, M., Monemvasios, I., Vardaka, E., Rigas, P., J. Chromatogr. B 2008,
876, 245-251.
[88] Hirooka, E. Y., Pinotti, M. H. P., Tsutsumi, T., Yoshida, F., Ueno, Y., Nat.
Toxins 1999, 7, 85-92.
[89] Furey, A., Lehane, M., Gillman, M., Fernandez-Puente, P., James, K. J., J.
Chromatogr. A 2001, 938, 167-174.
[90] Ciminiello, P., Dell'Aversano, C., Fattorusso, E., Forino, M., Magno, G. S.,
Tartaglione, L., Quilliam, M. A., Tubaro, A., Poletti, R., Rapid Commun. Mass
Spectrom. 2005, 19, 2030-2038.
[91] Regueiro, J., Martín-Morales, E., Álvarez, G., Blanco, J., Food Chem. 2011,
129, 672-678.
[92] de la Iglesia, P., Giménez, G., Diogène, J., J. Chromatogr. A 2008, 1215, 116-
124.
[93] Nguyen, A. L., Luong, J. H. T., Masson, C., Anal. Lett. 1990, 23, 1621-1634.
[94] Zhao, J. Y., Thibault, P., Quilliam, M. A., Electrophoresis 1997, 18, 268-276.
[95] Kvasnička, F., Ševčík, R., Voldřich, M., J. Chromatogr. A 2006, 1113, 255-258.
65
[96] Zhang, X. W., Zhang, Z. X., Toxicon 2012, 59, 626-632.
[97] Quilliam, M. A., Xie, M., Hardstaff, W., J. AOAC Int. 1995, 78, 543-554.
[98] Yasumoto, T., Oshima, Y., Yamaguchi, M., Bull. Jpn. Soc. Sci. Fish 1978, 44,
1249-1255.
[99] Scoging, A., Bahl, M., Lancet 1998, 352, 117.
[100] García, C., Schonstedt, V., Santelices, J. P., Lagos, N., J. Toxicol. Sci. 2006,
31, 305-314.
[101] Elgarch, A., Vale, P., Rifai, S., Fassouane, A., Mar. Drugs 2008, 6, 587-594.
[102] Igboeli, O. O., Asuzu, I. U., Bull. Environ. Contam. Toxicol. 2010, 84, 15-18.
[103] Li, A., Ma, J., Cao, J., McCarron, P., Toxicon 2012, 60, 420-425.
[104] Armi, Z., Turki, S., Trabelsi, E., Ceredi, A., Riccardi, E., Milandri, A.,
Environ. Monit. Assess. 2012, 184, 5085-5095.
[105] Naustvoll, L. J., Gustad, E., Dahl, E., Food Addit. Contam. 2012, 29, 1605-
1615.
[106] Lee, K. J., Mok, J. S., Song, K. C., Yu, H., Lee, D. S., Jung, J. H., Kim, J. H.,
J. Food Prot. 2012, 75, 2000-2006.
[107] Hattenrath-Lehmann, T. K., Marcoval, M. A., Berry, D. L., Fire, S., Wang, Z.,
Morton, S. L., Gobler, C. J., Harmful Algae 2013, 26, 33-44.
[108] Regulation, C., Off. J. Eur. Commun. 2011, 3-4.
[109] EFSA, EFSA J. 2008, 589, 1-62.
[110] Commission, E., Off. J. Eur. Union L 2004, 226, 22-82.
[111] Shen, J. L., Hummert, C., Luckas, B., Fresenius J. Anal. Chem. 1997, 357,
101-104.
[112] Nogueiras, M. J., Gago-Martínez, A., Paniello, A. I., Twohig, M., James, K. J.,
Lawrence, J. F., Anal. Bioanal. Chem. 2003, 377, 1202-1206.
66
[113] Quilliam, M. A., Hess, P., Dell’Aversano, C., in: de Koe, W. J., Samson, R. A.,
van Egmond, H. P., Gilbert, J., Sabino, M. (Eds.), Mycotoxins and Phycotoxins in
Perspective at the Turn of the Century, Wageningen, The Netherlands 2001, pp. 383-
391.
[114] Goto, H., Igarashi, T., Yamamoto, M., Yasuda, M., Sekiguchi, R., Watai, M.,
Tanno, K., Yasumoto, T., J. Chromatogr. A 2001, 907, 181-189.
[115] McNabb, P., Selwood, A. I., Holland, P. T., J. AOAC Int. 2005, 88, 761-772.
[116] Stobo, L. A., Lacaze, J. P. C. L., Scott, A. C., Gallacher, S., Smith, E. A.,
Quilliam, M. A., J. AOAC Int. 2005, 88, 1371-1382.
[117] These, A., Scholz, J., Preiss-Weigert, A., J. Chromatogr. A 2009, 1216, 4529-
4538.
[118] Gerssen, A., Mulder, P. P. J., McElhinney, M. A., de Boer, J., J. Chromatogr.
A 2009, 1216, 1421-1430.
[119] García-Altares, M., Diogène, J., de la Iglesia, P., J. Chromatogr. A 2013, 1275,
48-60.
[120] Bouaïcha, N., Hennion, M. C., Sandra, P., Toxicon 1997, 35, 273-281.
[121] de la Iglesia, P., Gago-Martinez, A., Yasumoto, T., J. Chromatogr. A 2007,
1156, 160-166.
[122] Hess, P., Nguyen, L., Aasen, J., Keogh, M., Kilcoyne, J., McCarron, P., Aune,
T., Toxicon 2005, 46, 62-71.
[123] Fux, E., McMillan, D., Bire, R., Hess, P., J. Chromatogr. A 2007, 1157, 273-
280.
[124] Villar-González, A., Rodríguez-Velasco, M. L., Ben-Gigirey, B., Botana, L.
M., Toxicon 2007, 49, 1129-1134.
67
[125] Chapela, M. J., Reboreda, A., Vieites, J. M., Cabado, A. G., J. Agric. Food.
Chem. 2008, 56, 8979-8986.
[126] Gerssen, A., Van Olst, E. H. W., Mulder, P. P. J., De Boer, J., Anal. Bioanal.
Chem. 2010, 397, 3079-3088.
[127] Kilcoyne, J., Fux, E., J. Chromatogr. A 2010, 1217, 7123-7130.
[128] Regueiro, J., Rossignoli, A. E., Álvarez, G., Blanco, J., Food Chem. 2011, 129,
533-540.
[129] Rossignoli, A. E., Fernández, D., Regueiro, J., Mariño, C., Blanco, J., Toxicon
2011, 57, 712-720.
[130] van der Fels-Klerx, H. J., Adamse, P., Goedhart, P. W., Poelman, M., Pol-
Hofstad, I. E., van Egmond, H., Gerssen, A., Food Addit. Contam. 2012, 29, 1616-
1629.
[131] Poli, M. A., Musser, S. M., Dickey, R. W., Eilers, P. P., Hall, S., Toxicon 2000,
38, 981-993.
[132] Pierce, R., Henry, M., Blum, P., Cont. Shelf Res. 2008, 28, 45-58.
[133] Ishida, H., Muramatsu, N., Nukaya, H., Kosuge, T., Tsuji, K., Toxicon 1996,
34, 1050-1053.
[134] FDA, Fish and Fishery Products Hazards and Controls Guidance, Food and
Drug Administration, Washington 2011.
[135] APHA, Recommended Procedures for the Examination of Seawater and
Shellfish, American Public Health Association, Washington 1970.
[136] EFSA, EFSA J. 2010, 8, 1627-1665.
[137] Alam, M., Trieff, N. M., Ray, S. M., Hudson, J. E., J. Pharm. Sci. 1975, 64,
865-867.
[138] Pierce, R. H., Toxicon 1986, 24, 955-965.
68
[139] Hua, Y., Lu, W., Henry, M. S., Pierce, R. H., Cole, R. B., Anal. Chem. 1995,
67, 1815-1823.
[140] Twiner, M. J., Bottein Dechraoui, M. Y., Wang, Z., Mikulski, C. M., Henry,
M. S., Pierce, R. H., Doucette, G. J., Anal. Biochem. 2007, 369, 128-135.
[141] McNabb, P., Rhodes, L., Adamson, J., Holland, P., Afr. J. Mar. Sci. 2006, 28,
375-377.
[142] Shi, F., McNabb, P., Rhodes, L., Holland, P., Webb, S., Adamson, J., Immers,
A., Gooneratne, R., Holland, J., N. Z. J. Mar. Freshwat. Res. 2012, 46, 149-165.
[143] McNabb, P. S., Selwood, A. I., Van Ginkel, R., Boundy, M., Holland, P. T., J.
AOAC Int. 2012, 95, 1097-1105.
[144] Lewis, R. J., Jones, A., Toxicon 1997, 35, 159-168.
[145] Lewis, R. J., Jones, A., Vernoux, J. P., Anal. Chem. 1999, 71, 247-250.
[146] Lewis, R. J., Yang, A., Jones, A., Toxicon 2009, 54, 62-66.
[147] Wu, J. J., Mak, Y. L., Murphy, M. B., Lam, J. C. W., Chan, W. H., Wang, M.,
Chan, L. L., Lam, P. K. S., Anal. Bioanal. Chem. 2011, 400, 3165-3175.
[148] Mak, Y. L., Wu, J. J., Chan, W. H., Murphy, M. B., Lam, J. C. W., Chan, L.
L., Lam, P. K. S., Anal. Bioanal. Chem. 2013, 405, 3331-3340.
[149] Lewis, R. J., Sellin, M., Poli, M. A., Norton, R. S., MacLeod, J. K., Sheil, M.
M., Toxicon 1991, 29, 1115-1127.
[150] Lewis, R. J., Holmes, M. J., Alewood, P. F., Jones, A., Nat. Toxins 1994, 2, 56-
63.
[151] Sick, L., Hansen, D., Babinchak, J., Higerd, B., Mar. Fish. Rev 1986, 48, 29-
34.
[152] Goto, J., Goto, N., Shamsa, F., Saito, M., Komatsu, S., Suzaki, K., Nambara,
T., Anal. Chim. Acta 1983, 147, 397-400.
69
[153] Yasumoto, T., Satake, M., Fukui, M., Nagai, H., Murata, M., Legrand, A.-M.,
in: Smayda, T. J., Shimizu, Y. (Eds.), Toxic Phytoplankton Blooms in the Sea,
Elsevier Science Publishers, New York 1993, pp. 455–460.
[154] Mak, Y. L., Wai, T. C., Murphy, M. B., Chan, W. H., Wu, J. J., Lam, J. C. W.,
Chan, L. L., Lam, P. K. S., Environ. Sci. Technol. 2013, 47, 14070-14079.
[155] Stewart, I., Eaglesham, G. K., Poole, S., Graham, G., Paulo, C.,
Wickramasinghe, W., Sadler, R., Shaw, G. R., Toxicon 2010, 56, 804-812.
This chapter has been removed for
copyright or proprietary reasons.
Chapter 2
Capillary electrophoresis for the analysis of paralytic shellfish poisoning toxins in shellfish: Comparison of detection methods
Published in:
http://www.ncbi.nlm.nih.gov/pubmed/24591173
Keyon, A.S.A., Guijt, R.M., Gaspar, A., Kazarian, A.A., Nesterenko, P.N., Bolch, C.J. and Breadmore, M.C. Capillary Electrophoresis for the Analysis of
Paralytic Shellfish Poisoning Toxins in Shellfish: Comparison of Detection Methods, Electrophoresis, 35, (2014) 1496-1503.
DOI: 10.1002/elps.201300353
101
Chapter 3
Transient isotachophoresis-capillary zone electrophoresis with
capacitively coupled contactless conductivity and ultraviolet
detections for the analysis of paralytic shellfish poisoning toxins
in mussel sample
3.1 Introduction
In Chapter 2, CZE-UV and CZE-MS were compared with another two most portable
methods namely CZE-C4D and MEKC-FLD for the separation and detection of
PSTs. Whilst MEKC-FLD had good sensitivity (60.9-1040 ng/mL), selectivity was
compromised by the AOAC pre-column oxidation HPLC-FLD method required to
render the PSTs fluorescence. CZE-C4D achieved similar LODs to MEKC-FLD
without the need for this reaction, but was unable to detect PSTs when applied to a
mussel sample due to sample matrix interferences. C4D is prone to these
interferences, one of them being the salt (i.e. sodium chloride) in shellfish that might
also reduced the sensitivity. On-line preconcentration in CE can be used to enhance
sensitivity, in which normal stacking is usually utilised. This method requires BGE
having very high conductivity than that of the sample, which is not practical with
C4D. In this situation, transient isotachophoresis (tITP) becomes an alternative
approach. Uniquely, the ITP stacking criterion can be met by taking the major matrix
component (sodium) to serve as leading ion. When a suitable terminating ion is used,
the analyte ions with mobilities intermediate between the supporting ions will fall in
the ITP range and experience concentrating effect. The combination of tITP and CE
102
enables injection of larger sample volume and therefore enhancement in sensitivity
and separation efficiency.
Fukushi et al. [1] shows that tITP can be applied for the determination of inorganic
ions in seawater, in which sulfate/bromate were introduced into the BGE to bracket
mobilities of analytes at the low mobility side. Sensitivity enhancement of 20 was
achieved. Riaz et al. [2] combined tITP with metal chelation to improve sensitivity
for quantitation of transition metals in another highly saline sample, urine. The urine
matrix chloride and phenylfluorone 4-(2-pyridylazo) resorcinol added to the sample
acted as the leading and terminating electrolytes, respectively. In a sample containing
250 mM sodium chloride, more than 400-fold enhancement was realised compared
to the normal CE injection mode. While these methods were successfully developed
with absorbance detector, combining tITP-CE with C4D is relatively challenging due
to the requirement of low conductive BGE to maximise analytes response - a
limitation that is not a major issue using absorbance detector.
In this chapter, CZE with tITP is developed for the preconcentration of the PSTs and
reduction of matrix interferences by using sodium ion from the saline mussel sample
as the leading ion. This is the first report of tITP being used in combination with CE
and C4D. Parameters such as BGE concentration, duration of counter-flow and
sample volume injected are investigated to achieve optimum tITP-CZE conditions.
The method is used to determine concentration of PSTs in a contaminated mussel
sample and the outcomes are validated against data obtained from the AOAC
HPLC-FLD method [3].
103
3.2 Experimental
3.2.1 Chemicals and methods
A PST mixture for method development consisted of 2.5 µM of STX (750 ng/mL),
dcSTX (640 ng/mL), NEO (790 ng/mL), dcNEO (680 ng/mL), GTX3 (990 ng/mL),
GTX4 (1050 ng/mL) and dcGTX3 (880 ng/mL), and 7.5 µM of GTX1 (3080
ng/mL), GTX2 (2960 ng/mL) and dcGTX2 (2640 ng/mL). The mixture was prepared
at low concentration due to the limited concentration of the stock standard for each
PST in order to combine all PSTs. The BGE, which also acted as the terminating
electrolyte (TE), contained 500 mM L-alanine (EGA-Chemie, Steinheim, Germany)
and 500 mM acetic acid (pH 3.5), prepared by mixing 1.1 g of L-alanine and 1.4 mL
of glacial acetic acid with 25 mL Milli-Q water. To induce tITP during method
development, 25 mM sodium chloride solution (Sigma Aldrich, St. Louis, Missouri,
United States) was added to the PSTs mixture to allow the sodium ion to function as
leading ion for the cationic toxins.
3.2.2 CE
All experiments were performed with an Agilent HP3DCE system. A TraceDec® C4D
was located inside the cassette; operated at optimum conditions of -12 Db and a gain
of 150% (optimum after comparison with 100% and 200%); the signal filter function
was turned on to reduce baseline drift and noise. Separation was carried out in 100
cm long 50 μm I.D polyimide coated fused-silica capillary with the UV detector
positioned 8.5 cm from the outlet (monitored at 200 nm, bandwidth 16 nm), with the
C4D 13.5 cm from the outlet. New capillaries were rinsed for 20 min each with 1 M
NaOH, Milli-Q water and BGE before analysis. Between runs, the same conditioning
104
procedure was performed (5 min each) with additional voltage conditioning with
BGE for 5 min. All separations were performed at 25°C and +30 kV applied to the
inlet vial. Samples were injected at 50 mbar for 103 s (86 nL, 5% of the total
capillary volume), unless otherwise specified. Signals were collected using an
Agilent 35900E analogue-to-digital convertor. Peaks were integrated and processed
using 3D-CE ChemStation software available in Agilent CE. Normal integration
option was applied, however for asymmetric peaks, tangent skim option was used for
integration.
105
3.2.3 tITP-CZE
The tITP-CZE procedures are schematically presented in Figure 3.1a-b. First, the
capillary was filled with the BGE/TE and then the sample containing sodium
(leading ion) was introduced into the capillary by hydrodynamic injection at 50
mbar. A short BGE/TE plug was then introduced into the capillary by electrokinetic
injection (25 kV) with application of a negative hydrodynamic pressure of 50 mbar
(for Figure 3.1b). A voltage of +30 kV was applied for separation of the PSTs.
3.2.4 Preparation of mussel sample
The procedures for preparation of mussel sample were as indicated in Chapter 2.
Briefly, the mussel extraction was performed according to the AOAC HPLC-FLD
method. The extract solution (0.5 mL) was further filtered using a SPE C18 cartridge.
The effluent containing the toxins was collected. No sodium chloride was added to
the extract.
3.3 Results and discussion
3.3.1 Development of tITP-CZE method
In ITP, target analytes are focused into discrete zones between leading electrolyte
(co-ion with electrophoretic mobility, ep higher than the target analytes) and TE (co-
ion with ep lower than target analytes). For the PSTs, sodium ion is an attractive
leading ion because its ep (51.9 x 10-9 m2/Vs [4]) is greater than many organic
cations. Conveniently, there is a high abundance of sodium in seawater and marine or
freshwater organism extracts. As marine species preserve a constant internal salt
106
concentration within the same sea water body having Practical Salinity 35 (common
salinity in sea water) [5], sodium concentration in different mussels is expected to
vary insignificantly, allowing direct usage of sodium as leading ion. Subsequently,
this allowed establishment of an ITP system solely by the introduction of TE after
injection in the sodium-containing sample. This can be done through injection of a
discreet TE zone followed by BGE or by simply using the same electrolyte as TE and
BGE. In this work, the latter approach was selected because the low conductive TE
was more appropriate for use with C4D than using the higher mobility, higher
conductivity sodium acetate reported in CZE-C4D developed in Chapter 2. To aid in
BGE selection, PeakMaster software version 5.3 was used to select a suitable
electrolyte composition and for separation simulation.
107
Figure 3.1 Schematic representation of the procedures used in tITP-CZE; (a) without
counter-flow and (b) with counter-flow (1) Filling the capillary with BGE which also
acted as terminating electrolyte (T). (2) Hydrodynamic injection of sample (S)
containing sodium ion as leading ion (L). (3) Electrokinetic injection of T (BGE)
(while applying hydrodynamic counter-flow (CF) in Figure 3.1b). (4) Starting tITP-
CZE mode.
108
Based on the pKa values of PSTs (>8.5) [6] and the effective mobilities ranging from
13 to 45 x 10-9 m2/Vs, BGE/TEs based on glutamic acid (pH 2.6), -alanine (pH 4.1),
and L-alanine (pH 3.5) were identified as potential candidates. Encouraging results
were obtained with L-alanine (based on PeakMaster simulation and experiment),
although some of the toxin peaks were only partially resolved from the Na+ peak;
this BGE/TE was selected for optimisation of the tITP conditions.
3.3.2 Effect of BGE/TE concentration on the stacking and separation
performance
From a practical point of view, it was important to first establish the BGE
concentration required for good separation selectivity using C4D. Figure 3.2a-b
shows the effect of BGE concentrations on tITP-CZE of the toxins with C4D and UV
detection, respectively, when 86 nL (5% of the capillary volume) of the PSTs
mixture was injected. From Figure 3.2a, using 100 mM L-alanine, a very large Na+
peak was observed (nearly 15 min wide) and six small peaks corresponding to PSTs
migrated just after the sharp edge of the Na+ peak. These peaks corresponded to the
monovalent PSTs (dcGTX2, dcGTX3, GTX2, GTX3, GTX1 and GTX4) indicating
that the four faster divalent PSTs (dcSTX, STX, dcNEO and NEO) were not
separated from the Na+ peak, most likely due to them still been concentrated as ITP
zones at the rear boundary of the Na+ peak. As the concentration of BGE was
increased, the Na+ peak became narrower, and the initial ITP stage established from
injection of the saline sample becomes shorter and thus the divalent PSTs peaks
emerged at 500 mM L-alanine. All of the PSTs could be fully resolved from Na+
when the concentration of L-alanine was increased to 850 mM. These trends were
confirmed using UV detection (Figure 3.2b) where both NEO and STX/dcNEO
109
started to migrate out from Na+ at a BGE concentration of 500 mM. Note that the UV
detection window was located 5 cm after C4D; providing a longer effective
separation length for UV detection than for C4D.
110
Figure 3.2 Electropherograms of PSTs during tITP-CZE detected with (a) C4D and
(b) UV showing effect of various BGE concentrations on stacking and separation. L-
alanine (pH 3.5) as BGE/TE, 86 nL sample was injected. The concentration of PSTs
mixture (mixed with 25 mM sodium chloride) injected and other conditions were
described in experimental part. Separation conducted at +30 kV and 25°C. Peak
identification: (1) dcSTX, (2) STX, (3) dcNEO, (4) NEO, (5) dcGTX2, (6) dcGTX3,
(7) GTX2, (8) GTX3, (9) GTX1, (10) GTX4.
111
While the higher concentration of BGE (i.e. 850 mM) allowed for better resolution,
the detection signals decreased significantly for C4D (approximately a 10-fold
reduction for the monovalent PSTs) as a result of the increased background
conductivity of the BGE. The signal decreased only by a factor of two for UV
detection. For C4D in particular, this indicated a trade-off between the higher
concentrations of BGE required to rapidly dissipate the ITP zone and low
concentrations of BGE required for good sensitivity. The divalent PST just emerged
from the Na+ peak at 500 mM L-alanine; hence this concentration was selected for
further optimisation.
3.3.3 Effect of counter-flow and optimisation of sample volume injected
Resolving the divalent PSTs from the Na+ zone using 500 mM L-alanine required
more effective separation length. This could be realised by minimising the movement
of the ITP zone towards the detector during the ITP stage thereby maximising the
capillary length remaining for separation of the PSTs. This was most easily done by
applying a counter-flow during ITP [7-9], which is schematically shown in Figure
3.1b. To investigate the effect of the counter-flow, its duration was varied from 1-5
min using maximum pressure available in the instrument (-50 mbar) and at an
optimum voltage of 25 kV. The Na+ peak dissipated faster as the counter-flow
duration was increased (Figure 3.3). As shown in Figure 3.4a, the migration time of
divalent PSTs decreased steeply from 23-24 min to 18-19 min when the ITP duration
was increased from 1 to 5 min. This could be explained by the fact that the ITP
occurred close to the capillary entrance during application of counter-flow, allowing
all PSTs to migrate away from the ITP zone. The monovalent PSTs showed a
decrease in migration time at duration of 3 min, after which they slightly increased
112
again, indicating the maximum counter-flow duration that could be applied.
Increased counter-flow duration allowed faster separation, but as illustrated in Figure
3.4b, sensitivity decreased with counter-flow duration due to the diffusion between
zones when applying the counter-flow. As a compromise between sensitivity and
selectivity, 3 min was selected as the optimum duration for the counter-flow.
113
Figure 3.3 Electropherograms showing Na+ peak dissipating faster as the counter
flow duration increased from 1-5 min. Counter-flow was applied during TE (BGE)
electrokinetic injection at 25 kV. 172 nL sample was injected. BGE/TE: 500 mM L-
alanine (pH 3.5). Other conditions were described in experimental part.
114
Figure 3.4 Effect of the duration of counter-flow on (a) migration time and (b) peak
area of PSTs standard mixture. 172 nL sample was injected. BGE/TE: 500 mM L-
alanine (pH 3.5). Other conditions were described in experimental part.
115
Having established the optimal counter-flow and duration for ITP, the injected
sample volume was optimised between 86 nL (5% of total capillary volume) and 344
nL (20% of total capillary volume) to enhance the sensitivity of the method.
However, the separation time was compromised with the increased injected sample
volume (as shown in Figure 3.5a for volume of 172 and 344 nL). Consequently, there
was a decrease in resolution (Figure 3.5b), indicating shorter effective capillary
length available for zone electrophoresis. An injection plug of 172 nL of sample
provided the highest PSTs sensitivity while still allowing for baseline resolution.
116
Figure 3.5 Effect of the sample volume on the migration time and resolution; (a)
electropherograms of PSTs mixture when 172 and 344 nL sample was injected. (b)
plot of resolution vs. sample volume injected. 3 min counter-flow was applied during
electrokinetic injection of TE (BGE) at 25 kV. BGE/TE: 500 mM L-alanine (pH
3.5). The concentration of PSTs mixture injected was described in experimental part.
Peak identification as indicated in Figure 3.2.
117
The optimum conditions were: 500 mM L-alanine (pH 3.5) as BGE/TE, injection of
172 nL sample (10% of total capillary volume), electrokinetic injection of BGE/TE
at 25 kV with a counter-flow of 50 mbar for duration of 3 min. Using these
conditions and as showed in Figure 3.6, the PSTs were successfully concentrated and
separated, with the exception of STX and dcNEO which could not be fully resolved.
The optimised tITP-CZE-C4D and tITP-CZE-UV were evaluated in terms of
linearity, LOD, intra-day and inter-day variations, and sensitivity enhancement
factors (SEFs), with the results being presented in Table 3.1. The linear ranges for
the divalent and monovalent PSTs analysed by tITP-CZE-C4D were between 70-
6200 ng/mL (regression coefficient (R2) > 0.980) and between 210-12,000 ng/mL
(R2 > 0.985), respectively. The LODs ranged from 74.2-1020 ng/mL for tITP-CZE-
C4D and 175-461 ng/mL for tITP-CZE-UV. The LODs for STX and NEO were
superior when compared to the LODs using capillary ITP-CZE-UV method
demonstrated by Thibault et al. [6]. Importantly, the reported LOD values for all
PSTs were below or close to the regulatory limit for shellfish sample except for
GTX4 (1020 ng/mL). The tITP-CZE methods afforded SEFs between 8-97 folds
with C4D and 19-84 with UV (Table 3.1). The presented method was at least two
times more sensitive than CZE-UV or CZE-C4D methods developed in Chapter 2. It
was important to note that this modest gain in sensitivity was achieved in the
presence of a high concentration of salt, a sample matrix property that was
problematic in CZE-C4D method developed in Chapter 2 limiting the applicability of
CE to shellfish samples.
118
Repeatability in terms of %RSD for intra-day variations in migration time and peak
height for PSTs analysed by tITP-CZE-C4D were between 0.82-1.5 and 2.2-9.2,
respectively. Repeatability for inter-day variations (%RSD) of the same performance
criteria were 0.88-1.8 and 4.8-11, respectively. The %RSD were slightly higher than
expected, probably due to slight changes in viscosity between electrolytes in the
counter-flow during the ITP stage.
119
Figure 3.6 Electropherograms of PSTs standards mixture under optimum tITP-CZE
conditions detected with; (a) C4D and (b) UV. Optimum conditions: L-alanine (pH
3.5) as BGE/TE, sample injection of 172 nL. Counter-flow was applied at 25 kV for
3 min during electrokinetic injection of TE (BGE). Separation was conducted at +30
kV and 25°C. Peak identification: (1) dcSTX, (2) STX, (3) dcNEO, (4) NEO, (5)
dcGTX2, (6) dcGTX3, (7) GTX2, (8) GTX3, (9) GTX1, (10) GTX4.
120
Table 3.1 Performance of tITP-CZE detected by UV and C4D
Toxin
tITP-CZE-UV tITP-CZE-C4D
LOD
(ng/mL)
SEFa Intraday
(%RSD, n = 3)
Inter-day
(%RSD, n = 3)
LOD
(ng/mL)
SEFa Intra-day
(%RSD, n = 3)
Inter-day
(%RSD, n = 3)
tm b Height tm
b Height tm b Height tm
b Height
dcSTX 280 64 0.81 5.9 1.1 8.1 74.2 97 1.2 7.0 1.5 7.2
STX 178c 71 1.1 5.1 1.2 5.0 91.0c 80 0.93 9.2 1.7 9.4
dcNEO 280 c 82 0.78 5.3 1.3 6.3 84.6c 92 0.82 6.9 1.7 7.2
NEO 175 79 1.0 4.6 0.76 4.2 92.8 87 1.1 7.2 1.1 7.4
dcGTX2 461 24 1.2 1.2 1.9 1.1 471 26 1.1 5.0 1.4 6.6
dcGTX3 219 29 1.4 8.2 1.9 8.9 350 22 0.95 6.9 1.8 11
GTX2 302 39 1.2 1.3 0.89 1.6 428 34 1.1 2.2 1.5 4.8
GTX3 274 34 1.3 9.1 1.1 10 566 27 1.1 3.8 1.8 5.6
GTX1 460 23 1.3 1.5 1.2 1.2 568 31 1.2 5.4 1.4 5.6
GTX4 308 19 1.3 4.3 1.3 4.7 1020 8 1.5 5.0 0.88 6.0
a) The SEF values were calculated by comparing the LODs between tITP-CZE method with conventional CZE of PSTs standard mixture.
b) tm was migration time
c) The LODs of STX and dcNEO were determined by conducting separation using individual standard solution due to co-migration.
121
3.4 Application of tITP-CZE method to mussel sample
The tITP-CZE method was applied for the determination of PSTs in a contaminated
mussel sample (replicates n = 3). As illustrated in the electropherograms in Figure
3.7a-b, STX and/or dcNEO, dcGTX2, GTX2, GTX3 and GTX1 were detected in the
mussel sample (S/N = 3) with both UV and C4D. Although some sample matrix
compounds were also concentrated and migrated near to PSTs, the identity of PSTs
peaks (S/N = 3) were confirmed by standard addition. C4D and UV are universal
detection techniques [10, 11], thus more susceptible to background interferences
when compared to more specific detector such as FLD. The toxin concentration
found in mussel sample was determined with a 5-point calibration and was reported
as STX-equivalent (µg STX-eq/kg) (Table 3.2). STX-equivalent is an internationally
accepted reporting method for PSTs found in contaminated shellfish [3, 12-14],
which is calculated by incorporating TEFs [15, 16]. To further confirm the validity
and potential applicability of the tITP-CZE-C4D method, the results obtained from
the analysis of the same contaminated mussel sample were compared to the AOAC
HPLC-FLD method. In contrast to the current method, the pre-column oxidation
method unavoidably produced multiple oxidation products for some toxins which
result in multiple overlapping peaks, thus the concentration of epimeric toxin pairs
was reported together for easier comparison [3]; i.e. dcGTX2,3, GTX2,3 and
GTX1,4. Analysis of the contaminated mussel sample using the HPLC-FLD method
(replicates n = 3) confirmed the presence of STX, dcGTX2,3, GTX2,3 and GTX1,4
in the mussel sample (Table 3.2). Only STX was detected in HPLC-FLD method
while dcNEO was absent, thus, other than confirmation by STX standard addition,
peak STX and/or dcNEO detected by tITP-CZE-C4D method (Figure 3.7) was
determined to be STX only. The concentration of all toxins determined by HPLC-
122
FLD agreed with the level determined from the current method (variation ≤ 17.5%),
confirming the potential applicability of the tITP-CZE-C4D method. The HPLC-FLD
method indicated the presence of dcSTX (TEFs = 1), where the level was lower than
the regulatory limit and could not be detected by the current tITP-CZE-C4D method.
However, the current tITP-CZE-C4D method had better selectivity for the separation
of low to highly potent epimeric toxin pairs (i.e. GTXs species having TEFs values
from 0.1 to 1) compared to the HPLC-FLD method. It was expected that the current
method using Na+ as leading ion to induce tITP can be used to analyse PSTs in
multiple mussel samples. Insignificant variation in Na+ concentration is expected in
multiple mussel samples [5]. Moreover, given that the aim of this work is to develop
a CE method that could be implemented easily in a microchip using detection
approaches already widely reduced to practice, the potential of this method for on-
site PSTs screening in shellfish had been demonstrated. More work on
implementation in portable instrumentation and analysis of shellfish samples are
required in the future to validate the real potential of this approach for improved
shellfish management.
123
Figure 3.7 Electropherograms for the analysis of PSTs in a contaminated mussel
sample by tITP-CZE using (a) C4D and (b) UV detection. Optimum conditions as
indicated in Figure 3.6. Peak identification: (1) dcSTX, (2) STX, (3) dcNEO, (4)
NEO, (5) dcGTX2, (6) dcGTX3, (7) GTX2, (8) GTX3, (9) GTX1, (10) GTX4.
124
Table 3.2: The concentration of PSTs found in a contaminated mussel sample using
tITP-CZE-C4D and HPLC-FLD methods.
Toxin Concentration of PSTs found
(µg STX-eq/kg)
tITP-CZE-C4D HPLC-FLD
dcSTX ND a 50
STX 918 b 1040
dcNEO ND a ND a
dcGTX2,3 383 300
GTX2,3 1960 2100
GTX1,4 696 560
a) ND Not detected due to reduced sensitivity of tITP-CZE-C4D for the particular toxin
b) Peak was determined to be STX, as confirmed by STX standard addition and the absence of dcNEO when analysed using HPLC-FLD method
125
3.5 Conclusions
This chapter presents a new method for the analysis of PSTs in mussel sample using
CZE-C4D in combination with tITP for analyte concentration and reduction of matrix
effects. CZE-C4D was selected because of its potential for miniaturisation and
portability, enabling its deployment for on-site screening to inform the shellfish
producers of potential PSTs contamination so they could manage harvest-scheduling
around harvest closures. The tITP system developed in this study used the naturally
high sodium concentration in mussel samples to act as leading ion in the tITP stage,
in combination with one electrolyte acting as BGE and TE (BGE/TE). L-alanine was
chosen as BGE/TE and its concentration was further optimised to improve selectivity
in C4D. Higher BGE concentration required to rapidly dissipate ITP zone resulted in
better peak resolution, however at the expense of reduced sensitivity. This elegant
separation chemistry was combined with a hydrodynamic counter-flow during tITP
step to allow for separation of all PSTs from the ITP zone. The obtained LODs of
PSTs ranged between 74.2-1020 ng/mL for tITP-CZE-C4D and from 175-461 ng/mL
for tITP-CZE-UV with enhancement in detection sensitivity ranging from 8-97 folds
over conventional CZE and at least two-fold over CZE-C4D method developed in
Chapter 2. Importantly, this enhanced sensitivity was achieved while reducing
sample matrix effects in the high salinity mussel sample and allowed the
determination of PSTs in mussel sample below or close to the regulatory limit. The
amount of PSTs found in a naturally contaminated mussel sample by using tITP-
CZE-C4D method was similar to those obtained by using AOAC HPLC-FLD
method; thus showing validity of the developed method.
126
3.6 References
[1] Fukushi, K., Miyado, T., Ishio, N., Nishio, H., Saito, K., Takeda, S., Wakida, S.
I., Electrophoresis 2002, 23, 1928-1934.
[2] Riaz, A., Kim, B., Chung, D. S., Electrophoresis 2003, 24, 2788-2795.
[3] AOAC, Official Method 2005.06 Paralytic shellfish poisoning toxins in
shellfish.Pre-chromatographic oxidation and liquid chromatography with
fluorescence detection, AOAC International, Gaithersburg, MD 2005.
[4] Pospíchal, J., Gebauer, P., Boček, P., Chem. Rev. 1989, 89, 419-430.
[5] Sumich, J. L., Morrissey, J. F., Introduction to the biology of marine life, Jones &
Bartlett Learning, Massaschusetts 2004.
[6] Thibault, P., Pleasance, S., Laycock, M. V., J. Chromatogr. 1991, 542, 483-501.
[7] Everaerts, F. M., Vacík, J., Verhegen, T. P. E. M., Zuska, J., J. Chromatogr. A
1970, 49, 262-268.
[8] Enlund, A. M., Westerlund, D., Chromatographia 1997, 46, 315-321.
[9] Reinhoud, N. J., Tjaden, U. R., Van der Greef, J., J. Chromatogr. 1993, 641, 155-
162.
[10] Gaš, B., Demjaněnko, M., Vacík, J., J. Chromatogr. A 1980, 192, 253-257.
[11] Brito-Neto, J. G. A., Fracassi da Silva, J. A., Blanes, L., do Lago, C. L.,
Electroanalysis 2005, 17, 1198-1206.
[12] AOAC, Official Method 959.08: Paralytic shellfish poison Biological method:
Final action, AOAC International, Gaithersburg, MD 2005.
[13] AOAC, Official Method 2011.27 First Action 2011: Paralytic shellfish
poisoning in shellfish by receptor binding assay, AOAC International, Gaithersburg,
MD 2012.
127
[14] AOAC, Official Method 2011.02 First Action 2011: Paralytic shellfish toxins in
mussels, clams, scallops and oysters by liquid chromatography post-column
oxidation, AOAC International, Gaithersburg,MD 2012.
[15] Oshima, Y., J. AOAC Int. 1995, 78, 528-532.
[16] Panel, E., EFSA J. 2009, 1019, 1-76.
This chapter has been removed for
copyright or proprietary reasons.
Chapter 4
Droplet microfluidics for post-column reactions system in capillary electrophoresis: Application to paralytic shellfish
poisoning toxins
Published in:
http://www.ncbi.nlm.nih.gov/pubmed/25310381
Keyon, A.S.A., Guijt, R.M., Bolch, C.J. and Breadmore, M.C., Droplet Microfluidics for Post-Column Reactions In Capillary Electrophoresis,
Analytical Chemistry (2014), Article in press.
DOI: 10.1021/ac5033963
159
Chapter 5
General conclusions and future directions
The impact of paralytic shellfish toxins (PSTs) on human health and economy
activities (through contamination of commercialised seafood products) continues to
increase worldwide, stimulating a renewed interest and research on these natural
toxins. There is an urgent need to detect and analyse PSTs, in which early detection
at shellfish farms can helps to implement prevention programs. There are several
general conclusions can be made from this research study regarding the successful
development and application of capillary electrophoresis (CE) aims to be new
alternative instrumental analytical method for PSTs.
CE has the advantage of being miniaturisable and portable, in which with correct
fundamental approaches taken, increases the simplicity of use on-site. The key step is
to develop CE with detection methods compatible with miniaturisation. This was
presented in Chapter 2, whereby CE with different detection methods having
sufficient sensitivity to detect PSTs at the regulatory limit was investigated. CZE-
C4D, MEKC-FLD, CZE-UV and CZE-MS were successfully developed and
compared; making this the first report of the use of C4D and an improved FLD
detection for various PSTs with CE. Good PSTs separation was achieved with CZE-
UV method; however this method as well as CZE-MS was unable to meet the
required LODs. On contrary, the two most portable options such as CZE-C4D and
MEKC-FLD showed LOD values below or close to the regulatory limit. CZE-C4D
method allowed for separation and identification of the charged PSTs within a few
minutes, but could not be used for the analysis of a contaminated mussel sample due
160
to background interferences. The MEKC-FLD method was suitable for conducting a
periodate screen of the mussel sample and included the neutral toxins to the analyte
set. When compared with HPLC-FLD method, the MEKC-FLD method was less
sensitive but allowed for direct quantitation of three of the toxins, in comparison with
one for the HPLC-FLD screen. The results confirmed the suitability of MEKC-FLD
for the periodate screen of PSTs, making it an attractive option for future direction in
the development of field-deployable portable analyser for on-site PSTs screening.
Whilst MEKC-FLD had good sensitivity, selectivity was inadvertently compromised
by the pre-column oxidation required to render the PSTs fluorescence. C4D achieved
similar sensitivity to FLD without the need for this reaction, but did not succeed
when applied to mussel sample. To deal with the high conductivity sample matrix
and for sensitivity improvement, tITP was successfully developed as an on-line
preconcentration method for CZE-C4D (Chapter 3). The tITP method utilised the
naturally high sodium concentration in mussel to act as leading ion, in combination
with one electrolyte acting as BGE and TE, in which L-alanine was used. As marine
species preserve a constant internal salt concentration within the same sea water
body, sodium was a suitable candidate as leading ion due to insignificant salt
variation between mussels. Uniquely, a hydrodynamic counter-flow was applied
during tITP step for improved stacking and separation performance. The obtained
LODs of PSTs ranged between 74.2-1020 ng/mL for tITP-CZE-C4D with
enhancement in detection sensitivity (8-97 folds) over conventional CZE and at least
two-fold over CZE-C4D method developed in Chapter 2. The most attractive part
was that the method achieved sensitivity enhancement while reducing sample matrix
effects in the high salinity mussel sample and allowed the determination of PSTs in
mussel sample below or close to the regulatory limit. It is believed that tITP-CZE-
161
C4D holds bright future to be deployed for screening of PSTs on-site, given that C4D
itself is inherently miniaturisable.
The limitation of pre-column oxidation MEKC-FLD method developed in Chapter 2
in fact opened up the prospect to investigate the potential of post-column oxidation.
An entirely new approach based on droplet microfluidics had been shown in Chapter
4 to be a useful tool for post-column reaction system. Uniquely, the system was
interfaced with a commercial CE instrument using only off-the-shelf components.
The use of a micro cross for positioning a salt bridge across the droplet flow from the
separation capillary outlet enabled the compartmentalisation and was the key element
for maintaining the electrical connection. As the first of its kind, positioning the
droplets in the electric field eliminated the need for EOF or hydrodynamic flow and
allowed electrophoresing effluents into the droplets. Droplets of water-in-oil could be
formed having frequencies between 0.7-3.7 Hz at different total flow rate,
corresponding to approximately 14 nL per droplet. By using fluorescent dyes as
model analytes, the relationship between total flow rate of droplets and signal was
successfully established. Low droplet frequencies resulted in higher proportion of the
electrophoresed components being in each droplet and significant increased in signal.
Higher droplet frequencies can maintain resolution better, even at the expense of
dilution. At optimum total flow rate (4 uL/min, 1 Hz), fluorescent peak was
compartmentalised into approximately 50 droplets and was found to be diluted 10-
times. Upon successful characterisations, the potential of this system was
investigated for post-column periodate oxidation of PSTs. PSTs were successfully
separated by CZE and post-column oxidised with periodic acid already in the
droplets. Due to the enhanced mixing in droplets, oxidation occurred within 14 s to
produce fluorescent product. The LODs for NEO and dcGTX2 were 670 ng /mL and
162
430 ng /mL, respectively. These values were approximately 11-times higher than
previously reported with pre-column oxidation and on-capillary detection in Chapter
2, and were consistent with the comparison of fluorescent dyes suggesting that
comparable oxidation between the two approaches was achieved. It is envisaged that
this new approach for post-column droplet collection in CE can be very valuable
when miniaturised further to decrease the droplet size which would decrease the
dilution factor.
The general aim of this research study is to explore the potential of CE as new
alternative instrumental analytical method for the analysis of PSTs. Suitability of CE
with different detection methods, combination with on-line preconcentration and
ability to be coupled with post-column reaction are indicative of CE versatility. The
results will serve as primary understanding towards the realisation of portable
analyser. Finally, it should be noted that further research for future direction is
essential in the following areas:
1. To confirm the potential of tITP-CZE-C4D method to be adapted as portable
analyser, the method should be extended to microchip format in which
miniaturised C4D is embedded within chip. Subsequent to that, the
application to various shellfish samples is also necessary to validate its ability
for on-site analysis.
2. In order to enable droplet microfluidics post-column reaction to be useful for
the analysis of PSTs content in shellfish samples, it is necessary for the
method to be able to detect toxins lower than the regulatory limit of 800
ng/mL. Therefore, the sensitivity needs to be improved further. This could be
163
by various stacking approaches, such as the tITP method previously
developed for CZE-C4D in Chapter 3. Another approach is through reduction
of the droplet size to minimise the dilution factor upon compartmentalisation.
Dilution can be countered in the future by using connector with smaller thru-
holes or by simply engineering the system with microfabricated structures
having small orifice.
3. Upon pre- and post-column oxidation, PSTs were excited at wavelength of
330 nm and emitted fluorescent at wavelength of 390 nm; therefore cheap
readily available He-Cd metal-vapour laser was utilised. While this laser was
useful during method development in laboratory, its bulky feature may
restrict its usefulness to be used in portable analyser for on-site PSTs
screening. He-Cd laser can be replaced with a laser diode module (i.e. 325
nm diode laser), which is more compact and tuneable over conventional
metal-vapour laser.
4. Integration of sample clean-up and preconcentration steps in portable
capillary or microchip electrophoresis is another aspect that can enhance the
sensitivity of existing methods developed in this research study. In retrospect,
this will reduce one off-line step (i.e. off-line SPE clean-up) prior separation,
which enhances the automation. For example, SPE concentrator with suitable
solid particles (such as molecular imprinted polymers, porous graphitised
carbon or perhaps immobilised aptamers) can be incorporated in-line with
capillary or microchip electrophoresis, in which large volume sample loading
can also be done at fast flow rate.
164
Appendix
Supplementary information for Figure 3.2: Electropherograms of tITP-CZE using six different BGE concentrations
Electropherograms of PSTs during tITP-CZE detected with (a) C4D and (b) UV showing
effect of six BGE concentrations on stacking and separation. L-alanine (pH 3.5) as BGE/TE,
86 nL sample was injected. Separation conducted at +30 kV and 25°C. Peak identification:
(1) dcSTX, (2) STX, (3) dcNEO, (4) NEO, (5) dcGTX2, (6) dcGTX3, (7) GTX2, (8) GTX3,
(9) GTX1, (10) GTX4.