LC QQQ MARINE BIOTOXINS

Dr. Jerry Zweigenbaum Agilent Technologies Agilent Americas Food Group Breaking News......

Recently, legislation in the EU introduced liquid chromatography coupled to tandem mass spectrometry (LC-MS-MS) as the soon to be, reference method for the detection of lipophilic marine biotoxins in live bivalve molluscs, tunicates and echinoderms.

Annex 1, SCoFCAH, 18-11-2010, SANCO/6831/2009 rev 7 amending EU Regulation 2074/2005.

This change in legislation allows for significant developments in the area of mass spectrometric detection for phycotoxins. Marine Biotoxins: Why analyse for them?

Marine biotoxin related illness can range from headaches, vomiting and diarrhoea to neurological problems, and in extreme cases can lead to death.

Is it really an issue? Agilent FC 24 Website

RASFF Alert notifications are sent when a food or feed presenting a serious health risk is on the market and when rapid action is required.

The RASFF member that identifies the problem and takes the relevant actions (e.g. withdrawal of the product) triggers the alert and allows other countries to search there own outlets. Responsibilities for Producers of shell fish

They must ensure their product must not contain marine biotoxins in quantities that exceed.... •800 micrograms per kilogram for paralytic shellfish poison (PSP), • 20 milligrams of domoic acid per kilogram for amnesic shellfish poison (ASP) •160 micrograms of okadaic acid equivalents per kilogram for okadaic acid, dinophysistoxins and pectenotoxins in combination, •1 milligram of yessotoxin equivalents per kilogram for yessotoxins, •160 micrograms of azaspiracid equivalents per kilogram for azaspiracids. The picture beforeSite Closuresharvesting 2005 – Ireland (locality) Assessment of Irish beds in 2005

Yellow: ASP Red: DSP Blue: AZP Green: DSP+AZP

Courtesy of Dr. Philipp Hess, Ifremer Examples of shell fish poisoning incidents Examples of shellfish poisoning incidents. Note that no new toxin groups have been reported since the discovery of azaspiracids in 1995 (Hess et al., 2008). Large-scale poisoning events for okadaic acid group toxins have still occurred during the last decade despite the toxic algae and toxins involved being known for over 20 years.

NO. OF SHELLFISH LOCATION POISONING REFERENCE CASES SPECIES OF ILLNESS MUSSELS AND DSP 164 JAPAN YASUMOTO ET AL., 1978 SCALLOPS BLUE MUSSELS (M. DSP > 300 NORWAY, SWEDEN UNDERDAHL ET AL., 1985 EDULIS) BLUE MUSSELS (M. ASP 107 CANADA PERL ET AL., 1990 EDULIS) CLAMS (A. PSP 187 GUATEMALA RODRIQUE ET AL., 1990 KINDERMANII) EASTERN OYSTER NSP 48 UNITED STATES MORRIS ET AL., 1991 (C. VIRGINICA) BLUE MUSSELS (M. AZP 24 IRELAND MCMAHON AND SILKE, 1998 EDULIS) BLUE MUSSELS (M. DSP > 300 BELGIUM DE SCHRIJVER ET AL., 2002 EDULIS) BROWN CRAB (C. DSP 200 NORWAY AUNE ET AL., 2006 PAGURUS) BLUE MUSSELS (M. DSP 159 UNITED KINGDOM UK COT, 2006 EDULIS) Provided courtesy of Dr. Philipp Hess, Ifremer During the last 2 years

•EFSA publish scientific opinion that bioassay has shortcomings and is not considered an appropriate tool for control of some toxins in Shellfish because of the high variability in results, the insufficient detection capability and the limited specificity.....A key, dirty dozen is the key focus. •Ring trials take place across Europe. As a result an LCMSMS approach becomes effectively validated by the Union Reference Laboratory for marine biotoxins and is declared to be an approach which maintains and ensures a full protection of consumer health without the shortcomings of the biological test, such as the high variability in results, the insufficient detection capability and the limited specificity. •In November Member States endorsed the relevant European Commission proposal during meeting of the Standing Committee of the Food Chain and Animal Health (SCoFCAH). In Annex III to Regulation (EC) No 2074/2005, Chapter III to be replaced by ....

LIPOPHILIC TOXIN DETECTION METHODS: Chemical methodology

The EU-RL LC-MS/MS method shall be the reference method for the detection of marine toxins

as referred to in Chapter V(2)(c), (d) and (e) of Section VII of Annex III, to Regulation (EC) No 853/2004.

This method shall determine at least the following compounds:

- okadaic acid group toxins : OA, DTX1, DTX2, including their esters

- pectenotoxins group toxins : PTX1 and PTX2,

- yessotoxins gp toxins: YTX, 45 OH YTX, homo YTX, and 45 OH homo YTX,

-azaspiracids group toxins: AZA1, AZA2 and AZA3.

Total toxicity equivalence shall be calculated using toxicity equivalent factors (TEFs) as recommended by EFSA.

If new analogues of public health significance are discovered, they should be included in the analysis. Total toxicity equivalence shall be calculated using toxicity equivalent factors (TEFs) as recommended by EFSA. In Annex III to Regulation (EC) No 2074/2005, Chapter III to be replaced by ....

LIPOPHILIC TOXIN DETECTION METHODS: Biological methods

A series of mouse bioassay procedures, may be still used until 31 December 2014 for detecting marine toxins.

After that period, the mouse bioassay shall be used only during the periodic monitoring of production areas and relaying areas for detecting new or unknown marine toxins on the basis of the national control programmes elaborated by the Member States. Dirty Dozen lipophilic toxins:

Compound Compound Toxicity Compound Compound Toxicity class factor class factor OA DSP 1 PTX1 DSP 1

DTX1 DSP 1 PTX2 DSP 1

DTX2 DSP 0.6 YTX DSP 1

AZA1 AZP 1 homo-YTX DSP 1

AZA2 AZP 1.8 OH-YTX DSP 1

AZA3 AZP 1.4 OH-homo-YTX DSP 0.5 Determination of lipophilic marine biotoxins in shellfish using Triple Quadrupole LC/MS/MS

Experimental Work and Data provided by:

Oliver Keuth Chemical and Veterinary Analytical Institute Münsterland-Emscher-Lippe, Münster, Germany

Dr. Begoña Ben Gigirey, Prof. Ana Gago-Martinez European Union Reference Labatory for Marine Biotoxins (EU-RL-MB) and Dept. Analytical and Food Chemistry, Faculty of Chemistry, University of Vigo Review of lipophilic marine biotoxin analysis:

• There are 2 LC-MS/MS methods proposed for the analysis of lipophilic marine biotoxins and both methods are currently evaluated in ring trials: • Acidic method (based on McNabb et al., 2005) – Reverse phase conditions – Acidic mobile phases – Electrospray ionization with polarity switching – Multiple reaction monitoring

• Alkaline method (based on Gerssen et al., 2009) – Reverse phase conditions – Basic mobile phases (pH 11) – Electrospray ionization with negative and positive time segment – Multiple reaction monitoring Influence of pH value of mobile phase:

DTX-1:

O O CH3 CH CH 3 O 3 CH3 O O O H H O + H H HO +H O HO O H H O O H H O H C - 3 OH H3C O H3C OH O H3C HO OH O H HO H CH2 H C O CH O 3 H 2 H3C OH H OH

Gerssen et al., J. Chromatogr. A, 1216 (2009) 1421-1430 Reaction Monitoring

Quad Mass Filter (Q1) Quad Mass Filter (Q3)

Collision Cell Spectrum with Q1 lets only Collision cell Q3 monitors only background target ion 210 breaks ion 210 characteristic ions (from ESI) pass through apart fragments 158 210 210 from ion 210 for 222 quant 158 268 280 158 165 191 210 170 210 250 290 190 210 150 170 190 210 160

Chromatogram High background

Low background Ionization of lipophilic toxins in electrospray AZA1 Compound class Formula ESI mode Precursor ion Okadaic acid group - OA C44H68O13 negative [M-H] = 803.5 positive [M+Na]+ = 827.5 - DTX1 C45H70O13 negative [M-H] = 817.6 positive [M+Na]+ = 841.5 - DTX2 C44H68O13 negative [M-H] = 803.5 positive [M+Na]+ = 827.5 Azaspiracids + PTX2 AZA1 C47H71NO12 positive [M+H] = 842.6 + AZA2 C48H73NO12 positive [M+H] = 856.5 + AZA3 C46H69NO12 positive [M+H] = 828.5 Pectenotoxins + PTX1 C47H70O15 positive [M+NH4] = 892.5 positive [M+Na]+ = 897.5 + PTX2 C47H70O14 positive [M+NH4] = 876.6 positive [M+Na]+ = 881.5 Yessotoxin group - YTX YTX C55H82O21S2 negative [M-H] = 1141.5 negative [M-2H]2- = 570.3 - homo-YTX C56H84O21S2 negative [M-H] = 1155.4 negative [M-2H]2- = 577.4 - OH-YTX C55H82O22S2 negative [M-H] = 1157.4 negative [M-2H]2- = 578.4 - OH-homo-YTX C56H84O22S2 negative [M-H] = 1171.4 negative [M-2H]2- = 585.4 Collisionally induced dissociation of OA positive

CE 45 V

negative

CE 60 V Application Number 1: Acidic Method Single polarity Methodology – Sample preparation:

2 g cooked and grinded mussel tissue

Extraction with aqueous solvent (80 % methanol)

Shaking or blending mixture and centrifugation

Repeat extraction, supernatants decanted into volumetric flask, filled up to 50 ml

Filter extract, using 0.45 µm membrane filter

Inject 10 µl LC-MS/MS Methodology – HPLC conditions:

Column: C-18, 150 x 2 mm, 5 µm (positive mode) Agilent ZORBAX Eclipse Plus C-8, 75 x 4.6 mm, 3.5 µm (negative mode)

Flow rate: 0.2 mL/min

Column temp: 30°C Injection volume: 10 μL (needle wash in flushport 5 seconds)

Solvent Channel A: 0.1% formic acid (positive mode) 2 mM ammonium acetate (negative mode) Solvent Channel B: methanol

Gradient: Time (min) A (%) B (%) 0.0 90 10 10.0 10 90 22.0 10 90 23.0 90 10 30.0 90 10

Overall Cycle Time: 30.0 min Methodology – Agilent Jet Stream conditions:

Drying gas temp: 300°C Gas flow rate: 5 L/min LC flow Super-heated N2 Nebulizer pressure: 45 psi sheath gas

Nebulizer N2 gas (near sonic Sheath gas temp: 250°C velocity) Sheath gas flow: 11 L/min Nozzle Voltage: +/- 500 V

Thermal focusing delta EMV: 400 V Electrospray Capillary voltage: +/- 3500 V Methodology – MRM transition parameters:

Analyte Polarity Prec Ion Prod Ion Frag CE Quantifier m/z m/z [V] [eV] OA and DTX2 pos 827.5 723.4 220 55 X pos 827.5 809.2 220 45 DTX1 pos 841.5 737.2 220 55 X pos 841.5 823.2 220 45 PTX1 pos 897.5 555.3 230 70 X pos 897.5 853.5 230 60 PTX2 pos 881.5 539.3 230 70 X pos 881.5 837.5 230 60 PTX2sa * pos 899.5 855.5 230 60 X pos 899.5 557.3 230 70 YTX neg 1141.5 1061.3 135 35 X neg 1141.5 925.5 135 60 Homo-YTX * neg 1155.4 1075.5 135 35 X OH-YTX neg 1157.4 1077.5 135 35 X OH-Homo-YTX * neg 1171.4 1091.5 135 35 X AZA1 pos 842.5 824.5 200 40 X pos 842.5 806.5 200 55 AZA2 pos 856.5 838.5 200 40 X pos 856.5 820.5 200 55 AZA3 pos 828.5 810.5 200 40 X pos 828.5 792.5 200 55 * Transitions based on literature information Chromatograms for blue mussel extract and QC

Blue mussel extract AZA1 to AZA3: < 20 µg/kg OA: 37 μg/kg DTX2: 120 μg/kg DTX1: 69 μg/kg

QC-sample AZA1: 15 µg/kg OA: 15 μg/kg PTX2: 15 μg/kg Lipophilic marine toxins calibration curves

Calibration curve (matrix matched) AZA1, OA, PTX2, and YTX reference standards spiked into blue mussel extract

Calibration range: AZA1: 2.5 to 25 ng/mL OA: 1 to 38 ng/mL Quantitation of marine toxins

• There are currently just 4 reference compounds (AZA1, OA, PTX2, YTX) available from the National Research Council, Institute for Marine Biosciences (NRC-CNRC), Halifax, Canada • All other compounds need to be quantified based on the calibration response of related compounds.

OA reference standard (10 ng/ml)

Blue mussel extract containing OA and DTX-2 (isomeric compounds) Quantitation of DTX2 based on OA calibration

• Area of OA for each calibration point is copied to DTX2 and allows automated quantitation of DTX2 based on calibration of OA • Other to application of average response factors, concept is amenable to standard addition, matrix spikes, QCs etc. • Baseline separation of signals is mandatory if compounds are isomeric; too good separation can have effect on quantitation result Method validation

• The method has been validated within an international collaborative study. • The collaborative study was conducted in the framework of the working group §64 LFGB “Phycotoxins”, which is hosted by the federal Office of Consumer Protection and Food Safety (BVL). • LODs and LOQs in cooked, grinded mussels:

Compound LOD LOQ 1OA 6 μg/kg 20 μg/kg 1DTX1 & 2 6 μg/kg 20 μg/kg 2AZA1 to 3 6 μg/kg 20 μg/kg 1PTX1 & 2 6 μg/kg 20 μg/kg 3YTX 10 μg/kg 35 μg/kg

1MRL in raw mussel material for sum of OA, DTX1 & 2, PTX1 & 2: 160 μg/kg OA equivalents 2MRL in raw mussel material for sum of azaspiracids: 160 μg/kg AZA1 equivalents 3MRL in raw mussel material for sum of yessotoxins: 1000 μg/kg YTX equivalents Application Number 2: Alkaline Method, both polarities • Based on a method published by Gerssen et al., 2009 • Alkaline mobile phase @ pH 11 • Chromatographic separation of compounds ionized in positive and negative mode

Negative ionization Positive ionization

AZA1

YTX

-

YTX

-

YTX/ homo

- -

AZA2

45 OH 45 OH 45

YTX/Homo

AZA3

DTX2

PTX2

OA DTX1

• Peak shapes of azaspirazids negatively effected Data provided by courtesy of Prof. Dr. Ana Gago-Martinez, EURLMB, Vigo Methodology – HPLC conditions:

Column: Waters X-Bridge C18, 150 × 3 mm, 3.5 μm

Flow rate: 0.4 mL/min

Column temp: 40°C Injection volume: 10 μL (needle wash in flushport 5 seconds)

Solvent Channel A: 0.05% ammonia (v/v) in water (pH 11) Solvent Channel B: 0.05% ammonia (v/v) in 90% acetonitrile

Gradient: Time (min) A (%) B (%) 0.0 90 10 1.0 90 10 10.0 10 90 13.0 10 90 15.0 90 10

Overall Cycle Time: 20.0 min Methodology – MRM transition parameters: Analyte Polarity Prec Ion Prod Ion Frag CE Quantifier m/z m/z [V] [eV] OA and DTX2 neg 803.5 255.2 330 52 X neg 803.5 113.1 330 66 DTX1 neg 817.6 255.2 330 52 X neg 817.6 113.1 330 66 PTX1 pos 892.5 839.5 170 22 X pos 892.5 213.1 170 42 PTX2 pos 876.6 823.7 170 22 X pos 876.6 213.1 170 42 YTX neg 570.3 467.1 185 30 X neg 570.3 396.2 185 34 Homo-YTX neg 577.4 474.4 185 30 X 577.4 403.3 185 34 OH-YTX neg 578.4 467.3 185 30 X 578.4 396.3 185 34 OH-Homo-YTX neg 585.4 474.4 185 30 X 585.4 403.3 185 34 AZA1 pos 842.6 824.6 220 29 X pos 842.6 672.4 220 53 AZA2 pos 856.5 838.5 220 29 X pos 856.5 672.4 220 53 AZA3 pos 828.5 810.5 220 29 X pos 828.5 658.4 220 53 Chromatograms for mussel extract and QC

Mussel extract

45-OH-YTX: 214 µg/kg YTX

OA: 133 μg/kg AZA1 YTX DTX2: 125 μg/kg - YTX: 380 µg/kg DTX1: 31 μg/kg OH 45 AZA3 25 µg/kg

AZA1 81 µg/kg AZA2

AZA3 DTX2

AZA2: 18 µg/kg OA PTX2 PTX2: 5 µg/kg DTX1

QC-sample OA: 80 μg/kg (100.3%) YTX: 500 µg/kg (102.7%) YTX AZA1: 80 µg/kg (102.6%) AZA1

PTX2: 80 μg/kg (99.5 %) PTX2 Data provided by courtesy of Prof. Dr. Ana Gago-Martinez, OA EURLMB, Vigo Lipophilic marine toxins calibration curves

Calibration curve (matrix matched) AZA1, OA, PTX2, and YTX reference standards spiked into mussel extract

Calibration range: AZA1, OA and PTX2: 2 to 24 ng/mL YTX: 12.5 to 150 ng/mL AZA1

YTX Data provided by courtesy of Prof. Dr. Ana Gago-Martinez, EURLMB, Vigo Quality control criteria • Chromatographic resolution OA/DTX ≥ 1.0  • Sensitivity based on S/N of qualifier ion: S/N ≥ 3 for lowest standard (matrix matched)  • Linear calibration curves with a correlation coefficient r2 ≥ 0.98  for at least five calibration points • Response drift over sample series ≤ 25% measured via the  slope variation of the two sets of calibration curves • Carry-over less than 10% of lowest calibration point  • Retention time drift less than 3% 

** EURLMB results from pre-trial Where to go from here?

• Add more compounds where applicable • Improve chromatographic separation and reduce runtimes by using UHPLC (requires intelligent management of duty cycles) • Make use of fast polarity switching capabilities of the G6460A QQQ system for extra speed

• Reduce injection volumes for enhanced robustness

YTX/ homoYTX

- -

DA

GYM

OA

SPX1

AZA2

AZA3 AZA1

DTX1

PTX2

YTX/ homoYTX

45 OH 45 OH 45

2

- DTX

* Real mussel extract, responses normalized to 100% Methodology – HPLC conditions:

Column: ZORBAX RRHD SB C-8, 50 x 2.1 mm, 1.8 µm

Flow rate: 0.4 mL/min

Column temp: 40°C Injection volume: 5 μL (needle wash in flushport 5 seconds)

Solvent Channel A: 2mM ammonium formate + 25 mM formic acid in water Solvent Channel B: 2mM ammonium formate in 95% acetonitrile

Gradient: Time (min) A (%) B (%) 0.0 88 12 0.5 88 12 3.0 50 50 6.5 10 90 7.0 10 90 7.1 88 12

Overall Cycle Time: 8.0 min Chromatograms for mussel extract and QC Mussel extract DA: 1490 μg/kg AZA1 GYM: 7 μg/kg

SPX1: 3 μg/kg homoYTX OA: 720 μg/kg - DTX2: 60 μg/kg

YTX: 230 µg/kg

AZA2

YTX/45 OH YTX/45 -

DTX1: 100 μg/kg 2

AZA3 -

AZA3 25 µg/kg DA

OA

GYM

SPX1

PTX2

DTX 45 OH 45 AZA1 150 µg/kg DTX1 AZA2: 4 µg/kg YTX/homoYTX PTX2: 7 µg/kg

QC-sample

DA: 787 μg/kg (109.9%) SPX1

GYM: 27 μg/kg (111.1%) GYM SPX1: 30 μg/kg (109.8%) YTX: 38 µg/kg (134.6%) AZA1

OA: 31 μg/kg (107.7%) DA

PTX2: 25 μg/kg (105.3%) PTX2

DTX1: 4 μg/kg (110.6%) OA DTX1

AZA1: 10 µg/kg (105.5%) YTX Lipophilic marine toxins calibration curves

Calibration curve (matrix matched) DA, GYM, SPX, OA, DTX1, AZA1, OA, PTX2, and YTX reference standards spiked into mussel extract

Calibration range: AZA1: 0.1 to 46 ng/mL OA: 0.2 to 200 ng/mL AZA1 PTX2: 0.2 to 120 ng/mL YTX: 0.3 to 180 ng/mL

Solid points: Calibration before sample series YTX Blue triangles: Calibration at end of sample series New triggered MRM functionality for confirmation

• Improved identification by triggering additional confirmatory MRMs if primary MRMs are detected above threshold • Triggered MRM advantageous over data dependant product ion scan because: • more sensitive • Peak shapes of primary transitions are not compromised (constant cycle time) • Stable in-spectra fragment ratios • User defined library with search capabilities in MassHunter software.

Page 39 Identification of compounds by library search

Library search functionality in MassHunter Qual

Excellent peak shapes for primary transitions!

Page 40 Identification of compounds in batch review

Page 41 Summary and Outlook

• There are two methods for lipophilic shellfish toxins currently evaluated in international ring trials (alkaline method vs. acidic method) • We have methods for both approaches available and our system has proved its sensitivity and robustness in both ring trials • There is a special script for the MassHunter Quant software available to deal with the lack of reference compounds • An application note together with CVUA Münster is now available for the analysis of lipophilic toxins under acidic conditions (5990-6377EN) • A similar note will be published for the alkaline method in January together with the EURL for marine toxins in Vigo, Spain, followed by a note on an UHPLC method with pos/neg switching for extra speed • Future work will focus on emerging toxins like Palytoxin, Brevetoxins, Ciguatoxins by LC-QQQ and the use of accurate mass instruments for the identification of toxins and toxin metabolites. Thank you for you attention

Questions about the application note, please contact:

Dr. Thomas Glauner EMEA LC/MS Food Segment Scientist Chemical Analysis Group Agilent Technologies Sales & Services GmbH & Co.KG Hewlett-Packard-Str. 8 76337 Waldbronn Phone: +49 (0)7243 602-2719 [email protected]