Post on 09-Mar-2018
Analysis of Marine Toxins in Shellfish
Andrew Turner Food Safety Group
Cefas, Weymouth
Overview
• Background to shellfish biotoxins • Context & legislation • Current shellfish monitoring • Method development – PSP toxins • Ongoing work
Introduction
Cefas c525 Staff
Lowestoft (360)
RV Endeavour
Weymouth (135)
• >100 yr fishery research
• Executive agency since 1997
Groups at Weymouth
• Aquatic Animal Disease – Fish, shellfish, amphibians
• Environment and Animal Health
– Ecotox, welfare, endocrine disruption, pollution
• Food Safety – Shellfish contam: viruses, bacteria, biotoxins
The Problem: Algal Biotoxins ASP
(Domoic/epi-domoic acid)
• Nausea
• Diarrhoea
• Vomiting
• Confusion
• Memory loss (may be permanent)
• Can be fatal
DSP (Lipophilic toxins include:
OA, DTX. YTX, PTX and AZA groups)
• Nausea
• Abdominal pains
• Vomiting
• Diarrhoea
• May be tumourigenic
PSP (Saxitoxin, GTX and C toxins)
• Numbness/tingling
• Headaches
• Nausea, Vomiting
• Respiratory distress
• Paralysis
• Can be fatal
Pseudo-nitzschia spp.
Dinophysis spp. Prorocentrum lima
Alexandrium spp.
Azadinium spinosum
1995 1980s 1989 1920s
20 mg/kg 160 µg/kg
3.75 mg/kg 800 µg/kg
OVERALL ~20 algal species &>35 toxins all present in UK waters and requiring monitoring
The Problem: Algal Biotoxins ASP
(Domoic/epi-domoic acid)
• Nausea
• Diarrhoea
• Vomiting
• Confusion
• Memory loss (may be permanent)
• Can be fatal
DSP (Lipophilic toxins include:
OA, DTX. YTX, PTX and AZA groups)
• Nausea
• Abdominal pains
• Vomiting
• Diarrhoea
• May be tumourigenic
PSP (Saxitoxin, GTX and C toxins)
• Numbness/tingling
• Headaches
• Nausea, Vomiting
• Respiratory distress
• Paralysis
• Can be fatal
Pseudo-nitzschia spp.
Dinophysis spp. Prorocentrum lima
Alexandrium spp.
Azadinium spinosum
1995 1980s 1989 1920s
20 mg/kg 160 µg/kg
3.75 mg/kg 800 µg/kg
OVERALL ~20 algal species &~33 toxins all present in UK waters and requiring monitoring
Toxin levels known to reach extreme levels in shellfish under certain conditions
Toxins are heat stable: not destroyed by cooking or
processing
Distribution & Mitigation • 3 toxin groups regulated in Europe:
– PSP, ASP and Lipophilic toxins
• UK and EU affected by all 3 groups
• This is a global problem – worldwide distribution
• Serious Health Risks
• Shellfish must be tested
• Use approved methods
• Compliance vs Reg limits
Context & Current Monitoring
EU Regulations • Water and shellfish testing • All commercially active shellfish production areas • Min. testing frequency • Toxins to be quantified and methods • Actions to be taken • Validation of methods • Accreditation of OC labs • Responsibility of national competent authority • Requirement for reference laboratories
Programme Framework
ISO17025:2005 accreditation • Quality system • Training and management • Quality control & proficiency testing • Suppliers • Calibration and validation • Equipment • Improvements
Animal (Scientific Procedures) Act 1984
• Testing methods if involving animals • WHAT, HOW, WHEN, HOW MANY • Training and management • Facilities (min. standard, maintenance)
National Reference Laboratory • Standard methods • Recommendations • Audits • Proficiency testing
Community Reference Laboratory • Standard methods • Proficiency testing • Method development • Method validation
Food Standards Agency
• UK Competent Authority (CA) • Responsible for Official Controls on bivalve molluscs • Sub-contract testing to OC labs
Cefas
• Lead UK “Shellfish Partnership” for OC testing • Run biotoxin monitoring programmes • Conduct OC testing
Methods – use of animals • Official Reference Method for PSP is still a
Mouse Bioassay (MBA)a – Direct quantitative toxicity determination
• Until 2011 (with leeway until 2015), RM for DSP
also the MBAb
– Qualitative assay only – New EU RM for LT = LC-MS/MS
• No issues with ASP – RM is HPLC-UV
aAOAC 959.08; bbased on Yasumoto et al., 1978
Implementation of “new” methods Process is time-consuming:
• Method developed and single-lab validated: – Must follow full EC / IUPAC guidelines – Demonstrate “equivalence” with current ref methods
• Formal multi-lab collaborative study – Following specific guidelines (e.g. AOAC)
• Publication as Official Method (e.g. AOAC, CEN) • Method acceptable within EU legislation • Approval by Competent Authority and COT • Accreditation to ISO17025
Implementation now may be possible
Challenges – Chemical Methods Large numbers of
toxins Chemical diversity of toxins
Method performance variability between
shellfish from different sources
Requirement for standards
(calibration & matrix)
Method complexity
Species differences
Practical issues
Fast turnaround requirements +
contingency requirements
Using chemistry to estimate
toxicity
Challenges Large number of
toxins Chemical diversity Performance issues
Poor perception Incomplete validation to date Method complexity
Species differences Practical issues
Requirements for method validation
Current monitoring
Monitoring programmes for shellfish
Flesh monitoring programme Samples of shellfish are collected from pre-determined monitoring
points within commercially active, classified production and relaying
areas
•>200 sites
•>3,500 samples
•Covering all
of GB
Shellfish testing process
• Samples received at Cefas daily • Shellfish shucked, >100g tissue homogenised • Sub-samples for each of three testing methods • Extraction, clean-up, derivatisation • Analysis overnight • Results reported next day (customer requirement)
• Results >MPL = shellfish beds closed for harvest
• Two consecutive <MPL to re-open
Analyse for regulated toxins
PSP
ASP
Lipophilic toxins
Biotoxin Detection Methods
ASP • Domoic acid & epi-domoic acid – total content of
whole shellfish or edible part alone
• EU reference method: HPLC-UV
• Shellfish + 50% Methanol extraction
• With or without SPE clean-up • Very simple, reproducible – no
major issues
HPLC
HPLC Chromatogram
Matrix peak
Domoic acid Matrix peaks
LC-MS/MS for Lipophilic Toxins
• EU Reference Method • EU-RL SOP specifies:
– Aims and scope – Extraction and general conditions – Performance characteristics
• Can use any instrumentation as long as method is specific, robust, linear and acceptable performance verified in-house
From 1st July 2011
•MBAs may still be used until 31st December 2014 •Afterwards, the MBA shall be only used during the periodic monitoring of production
LT method overview • Determination of toxins in following
groups: – Okadaic acid (OA and DTXs) – Pectenotoxin (PTXs) – Azaspiracid (AZAs) – Yessotoxin (YTXs)
• Direct determination of toxins
available as reference standards – Indirect determination of other toxins
• LOQ & MU: varies depending on
toxin and shellfish species
Shellfish homogenate
100% methanol extraction
Alkaline hydrolysis for OA/DTX esters
Filtration
LC-MS/MS
LT LC-MS/MS • Proportion of OA/DTXs as acyl-esters • Either low or high pH mobile phase – both
present chromatographic challenges • +/- switching to encompass all groups • Known response drift issues with neg-
mode toxins (OA/DTX) • Highly variable matrix effects • Known TEFs enable calculation of sample
toxicities • Now implemented in >10 countries • PT results from Quasimeme still indicate
high variability in performance between different labs
Lipophilic toxins
Detection of PSP
PSP toxins • Saxitoxins (>55 analogues) • Sodium channel blocker • Highly potent neurotoxin*
– Mice LD50 ~ 10 µg/kg (I.P injection) for STX – Acute oral tox of STX in mice 209-588 µg/kg – Severe illness in human reported at 5.6 – 2,058 µg/kg STX eq
• Listed in Chemical Weapons Convention • Majority are hydrophilic – but analogues also exist with
hydrophobic substituents • Stable in weakly acidic solutions • MPL = 800 µg STX eq/kg shellfish tissue
*R. Munday (2014) Toxicology of Seafood Toxins: A critical review
PSP toxins
• N-hydroxyl – Carbamate NEO, GTX1&4 – Decarbamoyl dcNEO, dcGTX1&4 – N-sulfocarbamate GTX6, C3&4
• Non N-hydroxyl – STX, GTX2&3, dcSTX, dcGTX2&3, GTX5, C1&2
• Others – M toxins, GC toxins and more…
• All have different toxicities; TEF of some still unknown
Saxitoxin derivatives
PSP toxins
• N-hydroxyl – Carbamate NEO, GTX1&4 – Decarbamoyl dcNEO, dcGTX1&4 – N-sulfocarbamate GTX6, C3&4
• Non N-hydroxyl – STX, GTX2&3, dcSTX, dcGTX2&3, GTX5, C1&2
• Others – M toxins, GC toxins and more…
• All have different toxicities; TEF of some still unknown
Saxitoxin derivatives
Thankfully: PSTs commonly occurring in naturally contaminated shellfish in
UK/EU are available as standards and most have fairly well described TEFs
PSP Methods 1. AOAC 959.08 MBA (Ref method)
2. Alternative methods authorised in EU under reg.
2074/2005: – Includes: “Lawrence method” - HPLC (AOAC 2005.06
method)
3. Other methods authorised in US/Canada: – PCOX LC-FLD (AOAC 2011.02) – Receptor Binding Assay (AOAC 2011.27)
4. Until recently no validated LC-MS method – None authorised in EU or US legislation
1. PSP MBA • Until Oct 2006 – all PSP testing conducted by MBA • Shellfish homogenised & extracted (HCl) • Injected into multiple mice • Death time gives PSP toxicity • Assay calibrated using STX solution bimonthly • LOD: ~330 µg STX eq./kg flesh (<1/2 regulatory limit)
• Negative samples – no death – mice still euthanised
• Many thousands mice required for routine testing
2. PSP by LC-FLD – more complex
Taken from M. Boundy, 2015
PSP LC-FLD (AOAC 2005.06 OM)
Extraction
(1% Acetic acid)
C18 SPE clean-up / pH adj
Peroxide oxidation
Fraction1 (C toxins)
Fraction2 (GTX1/4, GTX2/3,
GTX5, GTX6*)
Fraction3 (STX, NEO, dcNEO,
dcSTX)
Perox. Period
STX, dcSTX, GTX2/3, GTX5, C1/2, dcGTX2,3
Perox. Period Perox. Period
C1/2 C3/4* GTX2/3,
GTX5 GTX1/4 GTX6*
STX, dcSTX NEO,
dcNEO * No certified standards currently available
Ion exchange SPE (fractionation)
Shellfish homogenate
Periodate ox (screen)
HPLC-FLD
HPLC-FLD
HPLC-FLD HPLC-FLD
Unoxidised Non-toxic co-extractives
HPLC-FLD
PSP LC-FLD
Extraction
(1% Acetic acid)
C18 SPE clean-up / pH adj
Peroxide oxidation
Fraction1 (C toxins)
Fraction2 (GTX1/4, GTX2/3,
GTX5, GTX6*)
Fraction3 (STX, NEO, dcNEO,
dcSTX)
Perox. Period
STX, dcSTX, GTX2/3, GTX5, C1/2, dcGTX2,3
Perox. Period Perox. Period
C1/2 C3/4* GTX2/3,
GTX5 GTX1/4 GTX6*
STX, dcSTX NEO,
dcNEO * No certified standards currently available
Ion exchange SPE (fractionation)
Shellfish homogenate
Periodate ox (screen)
HPLC-FLD
HPLC-FLD
HPLC-FLD HPLC-FLD
(AOAC 2005.06 OM) Today’s event summary? “This meeting aims to cover developments in
analytical instrumentation that make it possible to simultaneously analyse numerous pollutants
in complex matrices using minimal sample clean-up”
Other than the “complex matrices”, this method
misses this aim on many levels!
Even so – highly effective for monitoring Used for OC since 2008
Consequences: Use of mice
Total UK replacement of MBA for shellfish food safety
0
2000
4000
6000
8000
10000
12000
14000
16000
2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012
Year
Number of mice used
Projected mice used
0
2000
4000
6000
8000
10000
12000
14000
16000
2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012
Nu
mb
er
of
mic
e
Year
Projected mice used
3. Other PSP methods
• AOAC 2011.02 – Post-column oxidation LC-FLD: – Not in EU legislation – Requires at least 2 columns/systems to run
each sample – Use of ion-pairing chromatography – Very short column lifetime – Prone to matrix effects – false +/-
4. LC-MS/MS options?
• Considered the “Holy Grail” for PSP • HILIC-MS/MS approach first demonstrated
in 2007, but issues with: – Sensitivity and run-time – Huge matrix effects in shellfish extracts – In-source fragmentation implications
• All 3 issues needed resolving… • Cefas/Cawthron 2014 collaboration
a) Sensitivity & run-time • Waters Acquity &
Xevo TQ-S
• 10-50 times sensitivity of Xevo TQ
• Enabled very low LOD/Q for standards
• Fast UPLC – ideal for fast cycle time
b) In-source fragmentation mitigation
• Traditionally all PST MRM acquired in +ve mode – 100% source fragment – Shared parent ions – Shared MRMs – Specificity lost in source
i) –ve mode for some GTX transitions ii) Optimise HILIC separation
Two solutions:
HILIC separation • Waters BEH
Amide HILIC
• +/- switching
• Full separation of critical pairs, including epimers
• Total cycle time of 8.5 min for all PSTs
c) Matrix effects - salts • 4 most common metals in ocean are Na, Mg, K & Ca • Each observed as formate salt clusters in –ve ion
scans • High salt conc =
– Large PST suppression – Retention time changes – Poor chromatography
-ve ESI 2D plot MS scan showing elution of the four major salt cations K+, Na+, Mg2+, Ca2+
Both chromatography AND ionisation is
compromised
c) Removal of matrix effects • Will “dilute and
shoot” work for PSTs in shellfish?
Mean recoveries (n=8) and LOQ (µg STX.2HCl eq/kg) for mussel extracts diluted 1/20 in 80:20:0.25 acetonitrile/water/acetic acid
Even with x20 or higher dilution, matrix effects
remain
c) Removal of matrix effects • Supelclean ENVI-Carb 250mg/3mL GPC
SPE cartridges • Fast auto clean-up on Gilson SPE robot • Cartridges re-usable
c) Matrix effects - salts • Use of Carbon SPE • Levels of Na, Mg, K & Ca reduced by ~90-100% • Na formate remains:
– some interference with C1 MRM
2D Plot –ve ion mode MS scan: P. oyster before & after carbon SPE clean-up
c) Matrix effects - salts
• Na formate MRM
452.7 > 113
• Check elution with C1 routinely
c) Removal of matrix effects • Recovery of
carbamates hugely improved
• Peak shape acceptable
• Improved sensitivity
Comparison of mean recoveries (n=8) and LOQ (µg STX.2HCl eq/kg) for mussel extracts diluted 1/20 in 80:20:0.25 acetonitrile/water/acetic acid
and with carbon SPE cleanup.
Method sensitivity • Excellent
comparison with LC-FLD methods
• Up to x50 improvement in LOQ vs current LC-FLD method
HILIC-MS/MS vs LC-FLD Comparison of PST results (µg STX di-HCl eq/kg) generated by HILIC-MS/MS and LC-FLD
methods on shellfish tissues from the UK between 2009-2014 (n=1,141)
PST HILIC-MS/MS summary
• Good method performance – Key points are SPE, optimised HILIC and use
of sensitive TQ-S • Fully validated for 12 species of interest • Good equivalence vs. LC-FLD methods • Accepted for publication • Collaborative trial planned • Aiming for AOAC and/or CEN method
Other findings 1) Assessment of UK shellfish revealed first
detection of novel PSP toxins “M toxins”, not detected with LC-FLD
2) LCMSMS method incorporates Tetrodotoxin (TTX) • Pufferfish; Not expected in bivalves; never seen in European
bivalves • Potential bacterial source e.g. Vibrio
• Analysis of vibrio-positive oysters & mussels from south coast found contain TTX
• First evidence for TTX in UK waters (or N. Europe) • First proof of TTX in European bivalve molluscs
Ongoing work Method development •Further develop faster LC methods •Other sample clean-up methods •Further automation - where possible •Assessment of biomolecular methods in parallel
•RBA, rapid tests, SPR, “molecular” techniques
Expanding toxin discovery •Assessment of TTX impact within UK and EU •Cyanotoxin contamination with food products •Other new/emerging toxins for UK:
•Other AZA analogues •Pinnatoxins and other cyclic imines •Brevetoxins •Other fish toxins – potential concern re: import
Conclusions •Current chemical testing methods effective, but complex & time-consuming •UPLC-MS/MS has presented opportunities for method simplification and more accurate quantitation •HILIC-MS/MS is a great tool for small polar PSTs •Highly sensitive MS/MS systems are great – but do not solve all problems •Salt-removal is key for successful PST method – may be of benefit to other methods for small polar analytes?
Today’s event summary? “This meeting aims to cover developments in
analytical instrumentation that make it possible to simultaneously analyse numerous pollutants
in complex matrices using minimal sample clean-up”
•With UPLC-MS/MS – yes, BUT
•Cannot envisage completely losing need for sample clean-up •Highly sensitive mass specs help but not the solution to all problems
Acknowledgements •FSA England and Wales •FSA Scotland •Cawthron Institute, New Zealand •All in Biotoxin Team at Cefas, Weymouth
RSC for inviting me today