Introduction

Phytoplankton are microalgae and are the main contributor to the marine food web as well as oxygen on earth. The beneficial role of algae in the food chain arises from the fact that they are the only organisms that can readily make long chain polyunsaturated fatty acids (PUFAs). Shellfish feed on phytoplankton and the potential beneficial role of shellfish in the human diet has been attributed to the presence of oils that are rich in PUFAs. Shellfish are also a rich source of protein, essential minerals, especially iron, vitamins A and D. However, some marine microalgae are known to produce bioactive compounds that negatively impact on human health through their accumulation in shellfish [1,2]. Many of these toxins

Comprehensive Analytical Chemistry, Volume 51 © 2008 Elsevier B.V.

ISSN: 0166-526X, DOI 10.1016/S0166-526X(08)00013-5 All rights reserved.

are known to be highly potent [3]. The proliferation of toxin-bearing algae, mainly diatoms and dinoflagellates, called harmful algal blooms (HABs), has been a major cause of concern in recent years [4]. Bivalve molluscs filter large volumes of water when grazing on microalgae, and can concentrate both bacterial pathogens and phycotoxins [5]. These shellfish, especially mussels, scallops, oysters and clams, are major vectors for toxins that are implicated in several human toxic syndromes. These include amnesic shellfish poisoning (ASP) [6], diarrhetic shellfish poisoning (DSP) [7], paralytic shellfish poisoning (PSP) [8], neurotoxic shellfish poisoning (NSP) [9] and azaspiracid poisoning (AZP) [10].

ASP first came to attention in Canada in 1987 when human fatalities occurred from eating mussels (Mytilus edulis) [11]. Domoic acid (DA), which is produced by marine diatoms of the Pseudo-nitzschia spp., was identified as the causative toxic agent [12,13]. ASP causes gastric upset, headache and dizziness but the syndrome was named due to the persistent short-term memory impairment experienced by some patients [11].

Toxins produced mainly by marine dinoflagellates are some of the most potent poisons known and have a major impact on human health [1]. The accumulation of toxins in shellfish and fish has led to serious human toxicosis as well as animal and bird deaths. Table 1 shows the toxins produced by marine phytoplankton.

2. SEAFOOD-POISONING TOXINS

Six main classes that cause poisoning from the consumption of seafood are discussed, with an emphasis on the methods for their detection and determination. The phycotoxin poisonings are widely known by their acronyms and include: Paralytic Shellfish Poisoning (PSP), Diarrhetic Shellfish Poisoning (DSP), Azaspir-acid Poisoning (AZP), Amnesic Shellfish Poisoning (ASP), Yessotoxins (YTX) and Neurotoxic Shellfish Poisoning (NSP). Other less common seafood-poisoning toxins that are not discussed in this review include ciguatoxins [14], tetrodotoxins [15] and palytoxins [16].

3. PARALYTIC SHELLFISH POISONING (PSP) TOXINS

The PSP toxins are basic, water soluble compounds, which are extremely sensitive to alkaline pH and air-oxidation. PSP toxins are potent marine neurotoxins, which specifically block the excitation current in the nerve and muscle cells, and this finally results in signs of paralysis. To date, there are over 21 known saxitoxin congeners. The structures of saxitoxin (Figure 1) congeners vary by differing combinations of hydroxyl and sulphate substituents at four sites on the molecule (R:-R4). Based on the substitutions at R4, the saxitoxins can be subdivided into four sub-groups, the carbamate, sulfocarbamoyl, decarbamoyl and deoxydecar-bamoyl toxins. Substitutions at R4 result in substantial changes in toxicity, with the carbamate toxins being the most potent [17].

Table 1 Seafood toxic syndromes and toxin-producing phytoplankton

Toxic syndrome

Toxins

Affected seafood

Source of toxins

Paralytic shellfish

Saxitoxin (STX),

Shellfish,

Alexandrium spp.

poisoning (PSP)

Neosaxitoxin

crustacions

[136],

(NEO),

Gymnodinium spp.

Gonyautoxin

[137]

(GTX)

(21 analogues)

Diarrhetic shellfish

Okadaic acid (OA),

Shellfish

Dinophysis spp.

poisoning (DSP)

Dinophysistoxins

[35,38],

(DTXs),

Prorocentrum spp.

Pectenotoxins

[37,138]

(PTXs)

Yessotoxins (YTXs)a

Shellfish

Protoceratium

reticulatum [108],

Lingulodinium

polyedrum [139]

Neurotoxin shellfish

Brevetoxins (PbTx)

Shellfish, finfish

Karenia brevis

poisoning (NSP)

(Ptychodiscus

brevis) [140,9]

Amnesic shellfish

Domoic acid (DA)

Shellfish, finfish

Pseudo-nitzschia spp.

poisoning (ASP)

and analogues

[12,141]

Azaspiracid

Azaspiracids

Shellfish

Protoperidinium

poisoning (AZP)

(AZAs)

crassipes [76]

Ciguatera poisoning

Ciguatoxins (CTXs)

Finfish

Gambierdiscus toxicus

[142]

Palytoxin poisoning

Palytoxin

Seaweed, crabs,

Ostreopsis siamensis

finfish

[143]

aYTXs were originally included in the DSP category.

aYTXs were originally included in the DSP category.

Human symptoms of PSP vary depending on the dose and the individual. In the case of a mild intoxication, a sensation of tingling or numbness of the lips is experienced, gradually progressing to the face and neck; pins and needles in the extremities; headache, vertigo, nausea, vomiting and diarrhoea occur in tandem. Extreme intoxication results in muscular paralysis, choking and extreme respiratory difficulty often culminating in respiratory failure [18].

Although marine dinoflagellates have been identified as the progenitors of PSP toxins that contaminate bivalve shellfish, these toxins are also produced in freshwaters by cyanobacteria (Aphanizomenon flos-aquae) [19] and have been responsible for mass mortalities of cattle and sheep [20].

Several analytical techniques have been developed to determine the PSP toxins including: receptor binding assay [21], mouse bioassay [22], liquid chromato-graphy with fluorimetric detection (LC-FLD) [23], liquid chromatography with mass spectrometric detection (LC-MS) [24] and capillary electrophoresis [25].

Saxitoxins STX

L/OH"

R3 R2

R1 R2

R3 H

Figure 1

GTX II H OSO3

GTX III H H

NEO OH H

Structure of the most common PSP toxins; R4 = H; Carbamoyl.

OSO-

Hydrophilic interaction LC with MS detection has also been used for the analysis of PSP toxins [26].

The most frequently used methods for PSP determination involve LC-FLD and these are based on the oxidation of PSP toxins to fluorescent products. These sensitive determinations can be performed post-column, as first developed by Oshima [27], and these remain the definitive methods for the determination of a wide range of PSP analogues. This derivatization procedure has also been used for the determination of PSP toxins in body fluids in the forensic investigation of fatal intoxications following shellfish consumption [18]. However, the pre-column oxidation method, developed by Lawrence et al., is more convenient for the routine monitoring of PSP toxicity and has recently been adopted as an AOAC International reference method [28-30]. PSP toxins are basic and can be readily extracted from a shellfish homogenate using dilute hydrochloric acid. The oxidation reaction can be carried out using either periodate or peroxide and the fluorescent products can be separated using either reversed-phase or ion exchange columns. Figure 2 shows the chromatograms that were obtained from contaminated mussel tissue [29]. The main disadvantage of this method is that not all PSP analogues can be identified as the oxidation step can produce the same fluorescent product from more than one PSP toxin.

4. DIARRHETIC SHELLFISH POISONING (DSP) TOXINS

DSP is a widely distributed seafood contamination. Three classes of toxins were initially included in the DSP group (a) dinophysistoxins, (b) pectenotoxins (PTXs)

Figure 2 Sample chromatograms from the inter-laboratory study of determination of PSP toxins in various bivalve shellfish, using the Lawrence LC-FLD method that involves pre-column oxidation. (A) Analysis of mussels containing PSP toxins; (B) Analysis of clams containing PSP toxins. Adapted with permission from Ref. [29]. Copyright 2005 by AOAC International.

Figure 2 Sample chromatograms from the inter-laboratory study of determination of PSP toxins in various bivalve shellfish, using the Lawrence LC-FLD method that involves pre-column oxidation. (A) Analysis of mussels containing PSP toxins; (B) Analysis of clams containing PSP toxins. Adapted with permission from Ref. [29]. Copyright 2005 by AOAC International.

and (c) yessotoxins (YTXs) [7,31,32]. However, YTX is not diarrhetic and is no longer classified as a DSP toxin [33,34].

DSP was first reported in Japan in 1978 but the illness is now recognized as an important threat to public health throughout the world and outbreaks have resulted in prolonged closures of shellfish culturing industries. Dinophysistoxin-1 (DTX1) was identified in Japan as the causative toxin and was accumulated in bivalve shellfish through filter feeding on the dinoflagellate, Dinophysis fortii [35]. However, most incidents of DSP have involved the demethyl analogue, okadaic acid (OA), as the responsible toxin arising from a variety of Dinophysis sp. [36-38]. DTX2, an isomer of OA (Figure 3), was first isolated from Irish mussels and this was the predominant toxin during major DSP events in 1991 and 1994 [39,40]. DTX2 was subsequently identified in shellfish along the western coastline of Europe [25].

The PTXs were first isolated from toxic shellfish in Japan and frequently co-occur with OA and DTXs [7]. It is often observed that the toxin profiles in shellfish are more complex than in the marine phytoplankton where toxins originate. Thus, although PTX2 (Figure 4A, R = CH3) is the main PTX in Dinophysis spp., several PTX analogues are found in shellfish tissues that are formed from PTX2 by repeated oxidation; PTX1 (R = CH2OH), PTX3 (R = CHO) and PTX6 (R = COOH) as well as other PTX analogues that are formed by epimerization at C-7 [32]. Pectenotoxin-2 seco acids (PTX2SAs) have also been isolated and identified in dinoflagellates and shellfish [41,42]. PTX2SAs have an open chain structure and not a lactone ring like the rest of the PTXs (Figure 4B).

Figure 3 Structures of dinophysistoxins; Okadaic acid (OA); R1 = Me, R2 = H; dinophysistoxin-1 (DTX1); R1 = R2 = Me, dinophysistoxin-2 (DTX2); R1 = H, R2 = Me.
Figure 4 Structure of (A) Pectenotoxins; PTX2 (R = CH3), PTX1 (R = CH2OH), PTX3 (R = CHO), PTX6 (R = COOH) and (B) Pectenotoxin-2 seco acids (PTX2SAs).

4.1 Toxicity of DSP toxins

Serious diarrhetic effects have only been proven for DTXs. Hamano et al. first investigated the diarrhetic effects caused by OA [43]. The three major DTXs, OA, DTX2 and DTX1, are potent inhibitors of protein phosphatases (PP1 and PP2A) [44] and they are tumour promoters [45]. This phosphatase inhibition may be linked to degenerative changes in absorptive epithelium cells of the small intestine thus producing diarrhoea. The EU regulatory limit for total DTXs in shellfish is 0.16 mg/g edible tissues. Although PTX2 can induce diarrhetic symptoms, this only occurs at relatively high levels when compared with the DTXs [46]. The PTX2SAs exhibit substantially less toxicity than PTXs, indicating that the cyclic structure of PTX2 is important for toxicity [41].

4.2 Analysis of DSP toxins

Live animal bioassays were the first methods that were used for the detection of DSP toxins, but problems due to the lack of sensitivity, false positives, lack of method validation and the prohibition of such testing in many countries on ethical considerations, has led to an examination of a variety of alternative analytical methods [47,48].

4.2.1 Rapid screening methods for DSP toxins

Rapid screening methods, based on immunoassays provide convenient alternatives to live animal bioassays for screening shellfish tissues but these methods cannot be used for the precise quantitative determination of the DSP toxins [49-51]. A very sensitive method utilized LC linked with a protein phosphatase assay that used a 32P-labeled substrate but there are problems associated with using radioactive materials in regulatory laboratories [52]. More convenient colourimetric phosphatase inhibition assays have been developed in which the ability of PP2A to dephosphorylate a colourless substrate, p-nitrophenyl phosphate (p-NPP), to a coloured substrate, p-nitrophenol, was used to determine OA and DTXs [53,54]. A fluorescent inhibition assay for these toxins provides enhanced sensitivity for DTX detection [55].

4.2.2 Determination of DSP toxins using LC-FLD

LC-FLD using the carboxylic acid derivatizing reagent, 9-anthryldiazomethane (ADAM), has been the most widely used method for the determination of DTXs and PTX2SAs [38,56]. Other fluorimetric reagents have been successfully employed for the determination of DSP toxins and have been discussed in a previous review [48]. Unfortunately, PTX2 and most other PTXs cannot be determined using LC-FLD as they lack a carboxylic acid moiety. Although these LC-FLD methods are sensitive, sample clean-up, especially with shellfish samples, can be prolonged and relatively complicated [57,58]. The high sensitivity of LC-FLD was demonstrated when this method was used to show that the non-culturable phytoplank-ton, Dinophysis acuta, was the progenitor of DTX2 by the analysis of unialgal samples (22-100 cells), collected manually from microscope slides [38].

4.2.3 Determination of DSP toxins using LC-MS and LC-MSn methods

One of the earliest applications of LC-MS for the analysis of food contaminants was the determination of DSP toxins in shellfish [59]. LC-MS allows the determination of DTXs and PTXs without recourse to derivatization and multiple tandem MS methods do not require rigorous sample clean-up [60].

Table 2 summarizes the LC-MS methods that have been used for the analysis of DSP toxins (DTXs and PTXs). All of the MS methods used were carried out using an API source with an ionspray/electrospray interface, coupled to a single quadrupole or a triple quadrupole mass analyser. A major problem with LC-MS methods for the analysis of toxins in shellfish tissues is ion suppression and this is evident when using exhaustive extraction procedures, especially alcohols without sample clean-up. The consequence is that multi-toxin methods that encompass a wide range of analyte polarities are very difficult to validate for quantitative analysis but may play a role as screening methods for toxins. High performance size exclusion chromatography has been applied for the clean-up of raw extracts from mussel tissue. The proposed protocol can be performed automatically to enable the fast and sensitive analysis of a large number of samples. The recovery using this method was approximately 70% with good repeatability [61]. One method to determine the potential impact of ion suppression is to analyse an extract from a toxin-free shellfish sample with simultaneous infusion of a toxin standard. OA was used as a representative toxin and the signal for this toxin was acceptably consistent in the regions where toxins are expected to elute, thus showing that ion suppression did not significantly vary [62].

Different MS scans events have been applied for the identification and quantification of DSP toxins in shellfish and phytoplankton samples. Single scan MS experiments and selected ion monitoring (SIM) were used in most LC-MS applications. SIM is useful for the identification of toxins and screening samples but the lower selectivity when compared with multiple tandem MS does present difficulties, especially with shellfish samples whenever there is limited sample clean-up [63,64].

Both positive and negative modes have been used for the LC-MS determination of DSP toxins (Table 2). In positive ion mode, MS is notorious for generating multiple adduct ions for DSP toxins, with [M+NH4]+, [M+Na]+and [M+H]+ ions being observed together in the full scan spectra [64,65]. This has a detrimental effect on quantitation due to division of signal. It is prudent therefore to use an additive, for example, ammonium or sodium acetate, to assist in the formation of a predominant ion type. However, some adduct ions, especially the sodiated ions, are very stable and are difficult to fragment. The first LC-MS/MS method for polyether toxins was applied to the determination of four DTXs in shellfish using selected reaction monitoring (SRM). Three of these toxins are isomers with identical CID spectra and therefore chromatographic separation is essential [66].

Negative polarity does not exhibit the problems of multiple ion formation but in some instances the spectra may not be rich in product ions. Draisci et al. proposed a negative mode method for the different DTXs and PTXs but poor resolution was achieved for isomers [60]. Suzuki et al. developed another negative polarity method using LC-MS (SIM) for OA, DTX1 and PTX6 with reasonable detection limits [65].

Compound

Matrix

Stationary phase Mobile phase

OA, DTX1, DTX2

OA, DTX1, DTX2, DTX2B

OA, PTX2SAs

PTX2, PTX2SAs

OA, DTX1, PTX6, PTX2SAs PTX1, PTX2

Mussel (M. edulis)

Phytoplankton Mussels (M. edulis)

Mussel (M. edulis), Mussel

(M. galloprovincialis) Mussel (M. edulis), (M. galloprovincialis)

Scallop

(P. yessoensis) Mussels

(M. galloprovincialis) Scallop (P. yessoensis)

5 um Vydac 214TP (C4) column (250 x 2.1 mm) Vydac201TP C-18 (250 x 2.1 mm, 5 |j.m) Supelcosil LC18DB (300 x 1 mm, 5 |j.m) Supelcosil LC18DB (300 x 1 mm, 5 |j.m)

Symmetry CI 8 (150 x 2.1 mm, 3.5 |j.m) Develosil ODS-MG-5 (150 x 2 mm, 5 |j.m)

0.1% TFA (positive) and 2mM NH4OH (negative)

Ionization MS experiments Detection limit References source

injection

Compound

Matrix

Stationary phase

Mobile phase

Ionization source

MS experiments

Detection limit

References

PTX6, PTX2,

Greenshell Mussel

Luna-C18(2)

ACN:H20 4mM

ESI

SRM; [M+NH4]+

0.01 ng/g

[150] ([151])

PTX2SAs,

(P. canclicnhis)

(150 x 2 mm,

NH4OH and

PTXli,

5 |im)

50 mM formic

PTXISAs

acid.

OA, DTX1,

Mussel (M. ednlis)

Luna-C18(2)

ACN:H20 both

ESI

SRM; [M-H]-

0.5ng/g

[62] ([67])

DTX2, PTX2,

Phytoplankton cells

(150x2.2 mm,

containing 1 mM

PTX2SAs

(D. acuta)

5 |im)

NH4OAc

OA, DTX1,

Mussel (M. ednlis)

C8 (50 x 2.1 mm,

5mM ammonium

ESI

SIM; [M-H]";

nr

[152]

DTX2, PTX1,

Oysters (C. gigas)

3 |im)

acetate (pH 6.8)

[M+H]+;

PTX2

Scallop (P. maximns)

in water and ACN (95% v/v)

[M+Na]+

OA, DTX1,

Mussels (M. ednlis)

Hypersil-BDS-C8

A=water and

ESI

SRM or Product

nr

[153] ([154])

PTX2,

(50 x 2 mm,

B=ACN-water

Ion Scanning

PTX2SAs

3 |im)

(95:5), both with 50 mM formic acid and 2 mM ammonium formate.

ESI, electrospray; ISP, ionspray; SSI, sonic spray; TSI, turbospray; HP, hepatopancreas tissues; SRM, selected reaction monitoring; nr, not reported.

ESI, electrospray; ISP, ionspray; SSI, sonic spray; TSI, turbospray; HP, hepatopancreas tissues; SRM, selected reaction monitoring; nr, not reported.

x 17

x 136 5

Figure 5 LC-MS/MS of DSP toxins. (A) Irish mussels (M. edulis), (B) Norwegian mussels (M. edulis). (1) OA (2.8 min), (2) DTX2 (3.5 min), (3) PTX2SAi (4.9 min), (4) 7-epi-PTX2SA (6.5 min), (5) PTX2 (9.8 min), (6) DTX1 (6.65 min). (Signal amplification, if any, is shown above each peak.). LC conditions: a gradient of acetonitrile-water containing 1 mM ammonium acetate was used and the column was a Luna C-18(2) (150 x 2.1 mm, 5 mm, Phenomenex). Adapted with permission from Ref. [62]. Copyright 2004 by Elsevier.

x 23 rS

/VLA

x 233 rS

x 14

rS 3

x 28

Figure 5 LC-MS/MS of DSP toxins. (A) Irish mussels (M. edulis), (B) Norwegian mussels (M. edulis). (1) OA (2.8 min), (2) DTX2 (3.5 min), (3) PTX2SAi (4.9 min), (4) 7-epi-PTX2SA (6.5 min), (5) PTX2 (9.8 min), (6) DTX1 (6.65 min). (Signal amplification, if any, is shown above each peak.). LC conditions: a gradient of acetonitrile-water containing 1 mM ammonium acetate was used and the column was a Luna C-18(2) (150 x 2.1 mm, 5 mm, Phenomenex). Adapted with permission from Ref. [62]. Copyright 2004 by Elsevier.

The identification of DSP toxins in marine biological material is seriously hampered by the lack of commercial availability of many standard toxins and the isolation of toxins from shellfish tissues to use as standards is a very protracted process. Using LC-MS/MS, characteristic product ion spectra can be obtained for each component in a mixture without interference from the other components. However, for most applications where sensitivity of determination using LC-MS/MS is important, the monitoring of specific precursor/product ions with Q1 and Q3, respectively, whilst fragmentation occurs in Q2 (SRM/MRM) is necessary. Triple quadrupole MS offers excellent quantitative analysis with high selectivity and a wide calibration range. The polarity difference between neutral toxins, such as PTX2, and the acidic toxins has implications both for efficient LC separations and MS ionization efficiency. In Figure 5, the SRM negative mode (Q1/Q3 pairs) were: 803/255 (OA and DTX2), 876/137 (PTX2SAs), 817/255 (DTX1), 857/137 and (PTX2). The optimized LC gradient gave excellent resolution of the OA isomers, the PTX2SA isomers and PTX2 in relatively short run-times (10 min) [62,67]. The case with which toxins can be determine when present in widely different concentrations is demonstrated in Figure 5B. Thus, PTX2 (peak 5) has a signal that is 233 times less than DTXI (peak 6).

5. AZASPIRACID POISONING (AZP) TOXINS

AZP is the most recently discovered toxic syndrome from shellfish consumption [10]. The first confirmed event was in 1995 in the Netherlands and resulted from the consumption of mussels (M. edulis) from Killary Harbour in Ireland. Azaspiracids have also been found in other European countries, including the UK, Norway, France and Spain [68-70] and throughout the western coastline of Ireland [71-73]. Azaspiracids possess a polyether backbone, that include a trispiro ring assembly, an azaspiro ring fused with a 2,9-dioxabicyclo (3.3.1) nonane, and a terminal carboxylic group and co-occurs with an isomer, AZA6, and demethyl and methyl analogues, AZA1 and AZA3, respectively. Symptoms of AZP include nausea, vomiting, diarrhoea and abdominal cramps; similar to DSP. However, AZA1 is distinctly different from DSP toxins as it target organs include liver, spleen, the small intestine and it has also been shown to be carcinogenic [74,75].

The dinoflagellate, Protoperidinium crassipes, was identified as the responsible organism for the production of azaspiracids. This was achieved with LC-MS3 using only 200 cells that were manually collected from a microscope slide [76]. Protoperidinium spp. are difficult to culture in a laboratory environment and extensive sampling may need to obtain wild phytoplankton. P. crassipes is distributed in temperate to tropical waters and may even form blooms in warm water estuaries [77]. Following the total synthesis by Nicolaou et al., the structure of AZA1 (and consequently the structures of other AZAs) has been revised [78]. Figure 6 shows the revised structures of azaspiracids.

5.1 Determination of azaspiracids using LC-MS and LC-MSn methods

LC-MS quantitative analysis of a complex matrix, such as a shellfish tissue extract, can be problematic and interferences, including ion suppression, are to be expected. The first LC-MS method that was employed to determine AZA1-AZA3 involved a preliminary diol solid phase extraction (SPE) clean-up step [79]. A comprehensive study of the clean-up and toxin recoveries using various types of SPE showed that C-18 and diol phases were preferred for azaspiracids [80]. Sample clean-up using SPE should also be applicable for incorporation into less specific and less sensitive analytical methods for the determination of azaspiracids in seafood.

LC-MS(/MS) is a universal method for marine toxins, and has been successfully applied to the simultaneous determination of various groups of polyether toxins [60]. The first LC-MS/MS method was developed by Draisci et al. for the determination of the predominant azaspiracid, AZA1, in mussels using a triple quadrupole instrument [81]. This was later extended to the analysis of other azaspiracids using a monolithic column [82]. Although monolithic columns produce high chromatographic resolution, the high flow rates involved mean that these columns are rarely adopted for use with MS instruments.

Using an ion-trap detector (ITD) MS, LC-MS3 methods have been developed for a wide range of azaspiracids [83-86], that included the full chromatographic resolution of eleven azaspiracids using reversed-phase LC. The elucidation of the fragmentation pathways of azaspiracids was especially important for the development of analytical methods. To this end, the powerful complementary roles of hybrid quadrupole time-of-flight (QqTOF) MS that produces high mass accuracy data and ITD MS that yields MSn data (Figure 7, Table 3), was

Toxin

R1

R2

R3

R4

AZA1

H

ch3

H

H

AZA2

ch3

ch3

H

H

AZA3

H

H

H

H

AZA4

H

H

OH

H

AZA5

H

H

H

OH

AZA6

ch3

H

H

H

AZA7

H

CH3

OH

H

AZA8

H

ch3

H

OH

AZA9

ch3

H

OH

H

AZA10

ch3

H

H

OH

AZA11

ch3

ch3

OH

H

Figure 6 Structures of azaspiracids.

demonstrated in a study of AZA1-AZA3 fragmentations [87]. These toxins exhibit charge-remote fragmentation in positive mode following the initial water loss from the geminal diol moiety. A significant fragmentation is the loss of the A-ring in azaspiracids which was observed at the MS3 stage.

This results in the loss of the C1-C9 portion of azaspiracids that contains the R1 and R3 substituents, leaving a residual ion [M+H-H2O-C9H10O2R1R3]+, that retains R2 and R4 [87]. It was possible to select unique product ions for each azaspiracid isomer (Table 3), thus eliminating the requirement for complete chromatographic separation for the determination of isomers [86]. Figure 8 shows the chromatograms and spectra from the analysis of the isomers, AZA4 and AZA5, which have the same mass but were conveniently discriminated by the selection of characteristic

Figure 7 Multiple tandem mass spectrum (MS3) of AZA1, produced by positive electrospray ionization using a quadrupole ion-trap (ITD) instrument. The inset shows the main chargeremote fragmentation processes. (The G-ring fragmentation is not observed in ITD.) Adapted with permission from Ref. [86]. Copyright 2004 by Elsevier.

Figure 7 Multiple tandem mass spectrum (MS3) of AZA1, produced by positive electrospray ionization using a quadrupole ion-trap (ITD) instrument. The inset shows the main chargeremote fragmentation processes. (The G-ring fragmentation is not observed in ITD.) Adapted with permission from Ref. [86]. Copyright 2004 by Elsevier.

product ions formed by A-ring fragmentation [85]. Although LC-MSn using a ITD MS instrument is less sensitive than MRM methods using a triple quadrupole MS instrument, the ability to acquire spectra in MS2 and MS3 modes without a diminution in the sensitivity of detection is an important advantage of ITD MS.

6. AMNESIC SHELLFISH POISONING (ASP) TOXINS

DA (Figure 9) is an excitatory amino acid that has been implicated as responsible for a toxic outbreak in Canada in 1987 [13]. The intoxicated patients exhibited

Table 3 Azaspiracid structural assignments, fy-^ (see Figure 7), and the molecule-related and product ion masses using positive electrospray mass spectrometry

Name

R1

R2

R3

R4

[M+H]+

[M+H-H2O]+

[M+H-H2O-C9H10O2R1R3]+

AZA1

H

CH3

H

H

842.5

824.5

672.4

AZA2

CH3

CH3

H

H

856.5

838.5

672.4

AZA3

H

H

H

H

828.5

810.5

658.4

AZA4

H

H

OH

H

844.5

826.5

658.4

AZA5

H

H

H

OH

844.5

826.5

674.4

AZA6

CH3

H

H

H

842.5

824.5

658.4

AZA7

H

CH3

OH

H

858.5

840.5

672.4

AZA8

H

CH3

H

OH

858.5

840.5

688.4

AZA9

CH3

H

OH

H

858.5

840.5

658.4

AZA10

CH3

H

H

OH

858.5

840.5

674.4

AZA11

CH3

CH3

OH

H

872.5

854.5

672.4

gastrointestinal disorders, including nausea, vomiting, abdominal cramps and diarrhoea, manifesting about 24 h after the consumption of mussels. This was followed by the neurological symptoms of headache, confusion, disorientation, seizures and coma within 48-72 h. However, the permanent loss of short-term memory in some of the survivors led this toxic syndrome to be named ASP. The toxin in mussels causing the intoxication was identified as DA, which is a tricarboxylic amino acid (Figure 9). It was later shown that diatoms of the Pseudo-Nitzschia spp. were the causative organisms for producing DA [12]. Following this outbreak of ASP, there have been worldwide reports of DA contamination of seafood [88-91].

Several methods have been developed for the detection of DA in shellfish, including radioimmunoassay and enzyme immunoassay [92,93]. However, LC-UV is used by most regulatory agencies for the quantitative determination of DA in shellfish and a limit of 20 mg DA/g has been generally adopted [94,95]. Analysis is complicated somewhat by the presence of isomers of DA, as well as tryptophan, in naturally contaminated samples. The DA analogues differ from DA by isomerism involving diene side-chain and they are not always chromatographically resolved. A rapid and sensitive LC-UV method has been developed for analysis of DA in shellfish extracts without the need for SPE cleanup. Isocratic reversed-phase LC separation of DA and its isomers from shellfish matrix interferences and from the prevalent amino acid, tryptophan, was achieved by using an optimized mobile phase pH of 2.5 [96].

Fluorimetric derivatization of DA was used in the application of three LC-FLD methods and the improved sensitivity has allowed the determination of DA at much lower concentrations in seawater and marine phytoplankton. Derivatizing reagents used include fluorenylmethoxycarbonyl chloride (FMOC) [97], 6-aminoquinoloyl-N-hydroxysuccinimidyl carbamate [98] and 4-fluoro-7-nitro-2,1,3-benzoxadiazole (NBD-F) [99]. However, only the NBD-F method could be readily applied to the analysis of DA in shellfish without post-reaction

10095908580 -7570-

3.73

4.63

100-1 9080706050403020100

808.5

658.5

4826.6

851.8

808.5

674.4

362.5 438.

326.2

509.5587.1

736.8

826.4

1000

1000

600 m/z

Figure 8 (A) Chromatogram showing the separation of the isomers, AZA4 & AZA5, which was obtained using LC-MS3; AZA4 (3.73 min) and AZA5 (4.63 min). (B) and (C) are the mass spectra corresponding to AZA4 and AZA5, respectively. Reproduced with permission from Ref. [84]. Copyright 2002 by Elsevier.

'COOH

Figure 9 Structure of domoic acid.

COOH

COOH

'COOH

COOH

COOH

clean-up using SPE. Figure 10 shows the LC-FLD chromatograms for extracts from scallops (Pecten maximus), following derivatization with NBD-F, in which DA and two DA isomers (arrowed) are resolved as well as tryptophan (Try).

Time (min)

Figure 10 LC-FLD of domoic acid in an extract from scallops, contaminated with DA. DA isomers (arrowed) are resolved from DA and tryptophan (Try) does not interfere.

Time (min)

Figure 10 LC-FLD of domoic acid in an extract from scallops, contaminated with DA. DA isomers (arrowed) are resolved from DA and tryptophan (Try) does not interfere.

Other methods that have been employed for the determination of DA include capillary electrophoresis with UV detection [100] and derivatization followed by GC-MS [101]. Several LC-MS methods for DA have been developed [102,103], but the benefits of using multiple tandem MS, including improved selectivity and sensitivity, have been demonstrated [104,105].

0 0

Post a comment