Bio Toxin 1

Accepted Manuscript Title: A Convenient and Cost-Effective Method for Monitoring Marine Algal Toxins with Passive Samplers Authors: Thomas Rundberget, Eli Gustad, Ingunn A. Samdal, Morten Sandvik, Christopher O. Miles PII: DOI: Reference:

S0041-0101(09)00046-4 10.1016/j.toxicon.2009.01.010 TOXCON 3404

To appear in:

Toxicon

Received Date: 3 July 2008 Revised Date: 6 January 2009 Accepted Date: 16 January 2009 Please cite this article as: Rundberget, T., Gustad, E., Samdal, I.A., Sandvik, M., Miles, C.O. A Convenient and Cost-Effective Method for Monitoring Marine Algal Toxins with Passive Samplers, Toxicon (2009), doi: 10.1016/j.toxicon.2009.01.010

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

ARTICLE IN PRESS

A Convenient and Cost-Effective Method for Monitoring Marine Algal

IP T

Toxins with Passive Samplers

Christopher O. Miles1,3

SC R

Thomas Rundberget1, Eli Gustad2, Ingunn A. Samdal1, Morten Sandvik1,

National Veterinary Institute, PB 8156 Dep., NO-0033 Oslo, Norway

2

Institute of Marine Research, Flødevigen Research Station, Flødevigen, N-4817 His, Norway

3

AgResearch Ltd., Ruakura Research Centre, Private Bag 3123, Hamilton, New Zealand

CE PT ED

M AN U

1

*Corresponding author: National Veterinary Institute Tel: +47 2321-6231; Fax: +47 2321-6201

AC

E-mail address: [email protected]

1

ARTICLE IN PRESS

Abstract Passive sampling disks were developed based on the method of MacKenzie et al. (2004) and

IP T

protocols were formulated for recovering toxins from the adsorbent resin via elution from small columns. The disks were used in field studies to monitor in situ toxin dynamics during

SC R

mixed algal blooms at Flødevigen in Norway. Examples are given from time-integrated sampling using the disks followed by extraction and high performance liquid chromatography-mass spectrometry (HPLC-MS) analysis for azaspiracids, okadaic acid

M AN U

analogues, pectenotoxins, yessotoxins and spirolides. Profiles of accumulated toxins in the disks and toxin profiles in blue mussels (Mytilus edulis) were compared with the relative abundance of toxin-producing algal species. Results obtained showed that passive sampling disks correlate with the toxin profiles in shellfish. The passive sampling disks were cheap to produce and convenient to use and, when combined with HPLC-MS or enzyme-linked

CE PT ED

immunosorbent assay (ELISA) analysis, provides detailed time-averaged information on the profile of lipophilic toxin analogues in the water. Passive sampling is therefore a useful tool

AC

for monitoring the exposure of shellfish to the toxigenic algae of concern in northern Europe.

2

ARTICLE IN PRESS

Keywords: Dinophysis, okadaic acid, dinophysistoxin, azaspiracid, passive sampling,

AC

CE PT ED

M AN U

SC R

IP T

shellfish toxin, algal toxin

3

ARTICLE IN PRESS

1. Introduction Over the last decade there has been an increase in the commercial cultivation and exploitation of shellfish along the Norwegian coast. Contamination of shellfish with biotoxins from micro-

IP T

algae can be a problem for public health not only in Norway, but world wide (Hallegraeff,

1993; Toyofuku, 2006; Camacho et al., 2007), and many countries regulate the biotoxins in

SC R

shellfish (FAO/WHO/IOC, 2005). The Norwegian marine biotoxin monitoring programme involves phytoplankton identification and enumeration, together with analysis of shellfish

flesh. The Norwegian Food Safety Authorities have a public surveillance program for algal

M AN U

toxins in mussels. During the 2007/2008 season the algal monitoring was performed weekly while chemical analysis of shellfish was performed monthly, from February to December and only at selected places (35–40 locations), and the programme is not able to cover all of the vast Norwegian coastline.

CE PT ED

Analysis of biotoxins in the shellfish flesh is required to determine the safety of the product for consumption. However, analysis of shellfish is time consuming, technically demanding and expensive, so it is not ideal as a tool for monitoring the progress of toxigenic blooms. In addition, many of the toxins are metabolised in shellfish during digestion and assimilation, and the increased variety and complexity of the metabolite profile makes toxin quantification even more challenging. Phytoplankton monitoring involves collecting a concentrated sample

AC

of the algae, shipping the sample to a suitable laboratory, and then enumerating the identifiable toxigenic species (Lund et al., 1958). Phytoplankton monitoring has the ability to provide early warning of toxin problems, but is subject to some serious limitations: it is relatively labour intensive, requires specialised laboratories and expertise, provides only a snapshot of the algal population at the time of sampling, and is limited to species which have definitively been linked to particular toxins. Nevertheless, the combination of phytoplankton

4

ARTICLE IN PRESS

monitoring and shellfish analysis has historically provided a reasonable degree of protection to shellfish consumers in Norway and elsewhere (Hallegraeff, 1993; van Egmond et al., 1993;

IP T

Batoreu et al., 2005).

Recently, alternatives have been sought to improve marine biotoxin monitoring. Of these,

SC R

passive sampling methods have shown much promise as tools for measuring aqueous

concentrations of a wide range of priority pollutants. The first passive sampling methods were aimed at monitoring the concentrations of dissolved inorganic compounds in surface water

M AN U

(Benes and Steinnes, 1974). Since then, there has been a rapid development in the use of passive sampling devices (Huckins et al., 1990; Sodergren, 1990; Alvarez et al., 2004). Some of the general features of different passive sampling devices have previously been reviewed (Vrana et al., 2005; Stuer-Lauridsen, 2005). In comparison to traditional water sampling, passive samplers offer the ability to integratively sample a range of environmental

CE PT ED

contaminants over an exposure period, mimicking biological uptake while potentially avoiding the heterogeneity and clean-up problems implicit with biological matrices (Verhaar et al., 1995; Kot-Wasik et al., 2007). Recently, MacKenzie et al. (2004) introduced the idea of monitoring algal toxins by passively adsorbing them directly from seawater using solid-phase adsorbents. These so-called solid-phase adsorption toxin tracking (SPATT) devices, consisting of bags sewn from polyester mesh containing activated polystyrydivinylbenzene

AC

resin, adsorb lipophilic algal toxins dissolved in seawater. The SPATT bags provide a more convenient means to perform time-averaged sampling prior to, or during, algal blooms than by shellfish or phytoplankton analyses.

Chemical analysis of the algal toxins from the SPATT devices was much easier than for shellfish. Sample preparation was rapid and simple, and few interfering components were

5

ARTICLE IN PRESS

present in the sample extracts. Also, since the devices adsorb toxins released directly from the algae into the water, the toxin profile is much simpler than the metabolite profile usually found in shellfish. This results in easier assays, fewer toxins to quantify, and lower detection

IP T

limits for the targeted toxins. The resin used in the SPATT bags was tested and validated by MacKenzie et al. (2004) for a range of algal toxins found in New Zealand (pectenotoxin-2

SC R

(PTX-2), PTX-2 seco acid (PTX-2 SA), yessotoxin (YTX), ocadaic acid (OA) and dinophysistoxin-1 (DTX-1)).

M AN U

The suitability of this approach for monitoring algal toxins in Norwegian waters was investigated. As part of the study, the practicality of the device was improved by introducing a frame in which the HP-20 resin is restrained to form a “passive sampling disk”. This design is simple, cheap, more easily assembled and disassembled than the sewn SPATT bags, and is well suited to high throughput processing of samples. In the trials, results obtained from

CE PT ED

analysis of the passive sampling disks were compared to those from shellfish analyses and phytoplankton monitoring at Flødevigen in Norway. Because passive sampling devices containing the HP-20 resin have been validated for analysis of PTXs, YTX, OA/DTXs and azaspiracids (AZAs) (MacKenzie et al., 2004; Fux et al., 2008), no attempt was made to perform validation in the present investigation. Parts of the work have been reported in a

AC

preliminary form in an earlier communication (Rundberget et al., 2006).

2. Materials and Methods 2.1 Reagents

All reagents and solvents (methanol, acetonitrile, formic acid and ammonium formate) were of HPLC grade or analytical-reagent grade and obtained from Rathburn Chemical (Walkerburn, UK). Water was deionized (MilliQ). Standards of purified YTX (Miles et al.,

6

ARTICLE IN PRESS

2004a), PTX-2 SA (Miles et al., 2004b), PTX-2 (Miles et al., 2004c), OA, DTX-1 and DTX-2 (Larsen et al., 2007) and a semi purified mixture of AZA-1, -2 and -3 (unpublished) were

IP T

available in our laboratory from previous work.

2.2 Passive sampling disks.

SC R

Passive samplers were constructed from 100-µm nylon mesh (Sefar AG, Heiden, Switzerland) folded in half, a 75 mm diameter plastic embroidery frame (Permin, Copenhagen, Denmark) and HP-20 resin (DIAION HP-20, Mitsubishi Chemical Corporation, Tokyo, Japan). The

M AN U

resin (3.0 g) was placed between the two layers of nylon mash, and clamped tightly in the embroidery frame so as to form a thin layer of resin between the layers of mesh. A No. 2 fishing swivel (Mustad, Gjøvik, Norway) was attached to the outer ring of the embroidery frame to provide a point of attachment during deployment (Figure 1). The resin was activated by soaking the packed disk in methanol for 15 min and washing in deionised water, as

CE PT ED

described in the resin-manufacturer’s instructions. The activated passive sampling disks were placed in an air-tight plastic bags and stored cold (but not below 0 °C) prior to and after deployment in the sea.

2.3 Extraction of toxins from disks.

The embroidery ring was opened, and the used resin was quantitatively transferred to a 25 mL

AC

Varian Bond-elute reservoir fitted with a 20 µm nylon frit (Varian, Palo Alto, CA) and washed free of salts with 30–50 mL deionized water. Excess water was drawn from the column by application of a vacuum. MeOH was added to the column and the resin was stirred gently then left to stand for 15 min. The column was then eluted slowly (0.5–1 drop/sec) and when finished, the process was repeated with another 10 mL MeOH. Finally, an additional 3 mL MeOH was pushed through to flush the column, and the combined eluate evaporated to

7

ARTICLE IN PRESS

dryness in vacuo. The residue was dissolved in 1.0 mL 80% MeOH, centrifuged, and the supernatant analyzed by HPLC-MS. Alkaline hydrolysis was performed by mixing 200 μL of 5 M NaOH with 0.8 mL methanolic HP-20 extract. The mixture was left to react at 37 °C for

IP T

45 min, followed by addition of 210 μL of 5 M HCl. Samples were filtered through 0.2 μm

SC R

Spin-X filters prior to chromatographic analysis.

2.4 Field trials.

Trials were performed at the Marine Research Institute, Flødevigen, on the south-west coast

M AN U

of Norway. Passive sampling disks were taken from their packaging and deployed by attaching them to a fixed point at 1 m depth, and leaving them for the required time. The disks were then rinsed briefly with fresh tap water, sealed in an air-tight plastic bag, and shipped to the laboratory for analysis. Simultaneously, shellfish were harvested weekly and kept at −20

CE PT ED

°C prior to analysis, while algal cell counting was performed 3 times weekly.

2.5 Shellfish samples.

Frozen samples of blue mussels (Mytilus edulis) were thawed, and the flesh was removed from the shells and homogenized using an Ultra Turrax® homogenizer (IKA®, Werke GmbH & Co. KG, Staufen, Germany). The homogenates were stored at −20 °C until extracted. Homogenized shellfish (2 g) was extracted three times with 6 mL methanol by vortex mixing

AC

for 2 min, and centrifuged at 2500 g for 5 min between extractions. The three extracts were combined in a 20 mL volumetric flask, and the volume was adjusted to 20 mL with methanol. Alkaline hydrolysis of the esters was performed by mixing 200 μL of 5 M NaOH with 0.8 mL methanolic shellfish-extract. The mixture was left to react at 37 °C for 45 min, followed by addition of 210 μL of 5 M HCl. Samples were filtered through 0.2 μm Spin-X® filters prior to chromatographic analysis.

8

ARTICLE IN PRESS

2.6 Algal cell counts. Phytoplankton is routinely monitored in Flødevigen Bay three times per week. Every

IP T

Monday, Wednesday, and Friday, samples for enumeration and identification of phytoplankton were taken as an integrated sample using a flexible hose, from 0–3 m depth,

SC R

from the same location as mussels and passive samplers. The water sample was preserved using neutral Lugol’s solution. Smaller flagellates and algae in high concentration were

counted under the light microscope using a Palmer–Maloney chamber (200× magnification),

M AN U

with a detection limit of 104 cells/L (Palmer and Maloney, 1954). Larger dinoflagellates were counted on semitransparent filters according to the description of Fournier (1978). Examination in light microscope (100 × magnification) was performed on 50 mL of the sample that was gently filtered onto the filter for cell-counting, giving a detection limit of 20

CE PT ED

cells/L.

2.7 HPLC-MS analysis.

Liquid chromatography was performed on a Symmetry C18 column (3 µm, 50 × 2.1 mm) (Waters, Milford, MA) using a Waters 2670 HPLC module. Separation was achieved by linear gradient elution, starting from acetonitrile–water (35:65 v/v, both containing 5 mM ammonium formate and 0.01% formic acid) rising to 100% acetonitrile over 10 min, held for

AC

5 min, then switched back to the start-eluent. The HPLC system was coupled to a Quattro Ultima Pt triple quadrupole mass spectrometer operating with an electrospray ionization (ESI) interface (Waters Micromass, Manchester, UK). Typical ESI parameters were a spray voltage of 3–3.5 kV, desolvation temperature at 250 °C, source temperature at 100 °C and cone gas and desolvation gas at 40 and 600 L N2/h, respectively. The mass spectrometer was operated in MS/MS mode with argon as collision cell gas at 10 × 10−3 Torr. Ionization and MS/MS

9

ARTICLE IN PRESS

collision energy settings were optimized while continuously infusing (syringe pump) 20 ng/mL of the toxin standards at 3 µL/min. Detection of the analytes was performed by multiple reaction monitoring (MRM) in either positive (AZA-1 842.5>672.5, AZA-2

IP T

856.5>672.5, AZA-3 828.5>658.5, PTX-2 876.5>823.5, PTX-2 SA 894.5>823.5, PTX-12 874.5>821.5, 20-methylSPX-G/SPX-C 706.5>164.2) or negative (OA/DTX-2 803.5>255.1,

SC R

DTX-1 817.5>255.1, YTX 1141.5>1061.5) ionization mode. Except for PTX-12 and 20-

methylSPX-G/SPX-C, and DTX-2, which were quantified from calibration curves of PTX-2 and OA, respectively, all toxins were quantified using external calibration curves of standard

3. Results and discussion

M AN U

specimens dissolved in 80% MeOH.

CE PT ED

3.1 Practical aspects of the improved disks

The HP-20 resin used in the disks has been tested and validated for a range of lipophilic biotoxins by others (MacKenzie et al., 2004; Fux et al., 2008), and no attempts were made to perform validation in this study. The main improvement over the SPATT bags of MacKenzie et al. (2004) lies in the design of the frame in which the HP-20 resin is retained. This design simplifies the preparation of the activated disks, their deployments and the subsequent toxin

AC

extraction compared to the sewn SPATT bags. The new design was quick and easy to use, the frames and algal mesh could be washed and reused, and the frames hold the resin in a reproducible thin layer (Figure 1E) that increases the exposure of the resin to the toxins in the water. The frames are fitted with an attachment so that they can easily be tied or clipped to moorings, buoys and ropes. The adsorbent resin (HP-20) in the disks adsorbs and retains lipophilic toxins, and thus provides a time-averaged indicator of toxin concentration in the

10

ARTICLE IN PRESS

water in much the same way as mussels might. However, the disks have the advantage that there is no toxin metabolism, they are more easily stored and cheaper to transport, and provide

IP T

a much cheaper and cleaner extract for the analytical laboratory.

The adsorption rate of lipophilic toxins from sea water by HP-20 resin is fast. MacKenzie et

SC R

al. (2004) found that after only 3.5 h exposure, significant amounts of toxins were adsorbed on the resin even though the Dinophysis cell numbers were low (100 cells/L). In the

Norwegian trial there was a mixed bloom containing high amounts of different flagellates and

M AN U

microalgae, typically ca 1–3 × 106 cells/L and the D. acuta and D. acuminata numbers ranged from 100–360 cells/L. In this period the HP-20 material also became dark green, indicating a high concentration of algal pigments in the water. It can not be ruled out that the HP-20 material can become saturated or that the 100 µm nylon mesh can clog during the exposure time and consequently the toxin levels can be underestimated. This needs further

CE PT ED

investigation.

3.2 Sample preparation.

Recovery of the lipophilic algal toxins from the HP-20 resin was straight forward. A fresh water rinse was necessary, prior to elution with MeOH, to remove salts which may disturb ionization in the HPLC-MS. The ESI interface on the mass spectrometer is susceptible to salt

AC

effects (Gustavsson et al., 2001), and a high salt content can influence the relative intensities of the H+, NH4+, and Na+ adducts ions used for MRM quantitation of the toxins. The resulting methanolic extract is clean enough to be analysed on HPLC-MS without further preparation (Figure 3).

11

ARTICLE IN PRESS

It was found necessary to elute the 3 g of HP-20 resin with at least 2 × 10 mL of methanol to fully recover the adsorbed toxins and this is in accordance with the findings of Fux et al. (2008). It was also important to use a low flow rate through the column, typically 1–2 bed-

IP T

volumes per hour (Manufacturer’s recommendation). Elution with 23 mL of solvent gave a diluted sample, but a concentration step can be included depending on the required detection

SC R

limits and the sensitivity of the HPLC-MS system. The detection limits obtained with the instrument used in this work were typically 0.1–0.3 ng/disk, depending on the toxin. By

omitting the concentration step and adjusting the extract volume to 25 mL, detection limits of

M AN U

about 2–5 ng/disk were obtained.

3.3 Toxin profile of disks versus cell counts and blue mussels.

The OA/DTX concentrations in the disks and blue mussels, and Dinophysis spp. (D. acuta and D. acuminata) cell concentrations in the water, are shown in Figure 2. The amount of

CE PT ED

OA/DTXs in the disks fluctuated from 120 to 660 ng/g disk, with maxima in weeks 30, 34 and 38. The cell numbers of D. acuta and D. acuminata also fluctuated during this period, with numbers ranging from 0 to 360 cells/L and one major peak around week 29. In shellfish, the sum of both free OA and DTX and their esters was about 65 ng/g at the beginning of the trial and about 220 ng/g when the trial ended, with peaks at weeks 34 and 40. During the monitoring period (weeks 28–41), three events can be described. The first was the increase of

AC

Dinophysis cell densities and OA/DTX levels in the disks around week 30, but the toxin levels in the shellfish did not show a corresponding increase (Figure 2). The reason for this might be that the overall amounts of other algae present during this period were high (typically ca 3 × 106 cells/L) while the percentage of Dinophysis was low, so that the amount of OA/DTXs ingested by the shellfish was low.

12

ARTICLE IN PRESS

In the second incident, in weeks 33–34, levels of OA/DTXs in the disks increased to ca 650 ng/g, and the Dinohpysis counts also rose to a moderate level (ca 100 cells/L) in weeks 33 and 34 (Figure 2). The level of OA/DTXs in the shellfish reached a peak in week 34, as did the

IP T

levels of OA/DTXs in the disks. OA/DTXs decreased in the shellfish in the following weeks (35–36), when the Dinophysis cell numbers and toxin level in the disks also declined.

SC R

Depuration of algal toxins from shellfish is poorly understood (Duinker et al., 2007). Passive samplers could be a useful tool in studies of the depuration of toxins in shellfish through

M AN U

improved monitoring of the toxin-exposure of the investigated shellfish.

The third event occurred when OA/DTXs in the disks and levels of Dinophysis in the water reached a maximum in week 38 and 39, respectively (Figure 2). Levels of OA/DTXs in the shellfish increased in week 39 and reached a maximum in week 40. During this period the Dinophysis numbers were moderate (ca 140 cells/L) but the amount of other algae was lower

CE PT ED

(typically ca 1 × 106 cells/L) than earlier in the period when the algae population was typically 2–3 × 106 cells/L.

Based on these three events, it is difficult to recommend the passive samplers as an early warning tool. The first incident had increased toxin levels in the disks, with no corresponding increase in the shellfish. However, in the second and third events there was a marked increase

AC

in OA/DTXs in the disks (weeks 33 and 38) some time before the levels in the shellfish were observed to increase (weeks 34 and 39).

Dinophysis concentrations were below the regulatory limits set by the Norwegian Food Safety Authorities of 1000 D. acuminata or 200 D. acuta cells/L, throughout the trial. In week 30, Dinophysis reached 360 cells/L (220 D. acuminata and 140 D. acuta), while the other peaks at

13

ARTICLE IN PRESS

week 34 and 38 had 120 (120 D. acuminata and 0 D. acuta) and 140 (20 D. acuminata and 120 D. acuta) cells/L, respectively. Algal cell-counting precision is usually good at high cell densities if over 200 cells of the target species are counted, but poor at low cell densities when

IP T

fewer cells are counted (Lund et al., 1958). Except for in weeks 29 and 30, levels of Dinophysis were below 200 cells/L (corresponding to only 10 cells counted), and this may

SC R

account for the lack of a precise correlation between the Dinophysis cell counts and the toxin levels in the disks and shellfish. Lindal, et al 2008 reported substantial variations in toxin content of both D. acuta and D. acuminata due to population density and environmental

to the numbers of algae counted.

M AN U

variations and this may also affect the toxin levels found in the disks and shellfish compared

One difficulty with algal counting as a monitoring tool is that algal blooms can be short-lived and mobile, and thus occur between algal samplings. This is especially so at locations prone

CE PT ED

to tidal flows and/or exposed to wind and wave motion such as in Flødevigen where this trial was performed. Passive sampling disks should be a valuable tool at such locations, where the algal counts can change quickly from no Dinophysis, up to 360 cells/L and back to a few cells/L again during one week (Figure 2B). With passive samplers, the water column is continuously being sampled and hence provides an integrated measurement of toxin levels

AC

throughout the exposure period.

Another problem with algal counting is that it can only provide effective monitoring for toxins when the identity of the toxigenic species is known, whereas this information is not necessary when using passive sampling devices.

3.4 Detection of AZAs in the disks.

14

ARTICLE IN PRESS

HP-20 resin has been shown to adsorb OA/DTXs, PTXs, and YTX in New Zealand waters (MacKenzie et al., 2004) and recently Fux et al. (2008) detected AZAs in Irish waters using HP-20. In northern European waters, the AZA group is commonly detected in shellfish and is

IP T

often present at levels that make shellfish unsuitable for consumption (Hess et al., 2005; Aasen et al., 2006). During the summer of 2005, blue mussels at Flødevigen contained low

SC R

levels of AZA-1, AZA-2, AZA-3 and AZA-6, in a ratio of approximately 3:1:1:0.3

respectively, with concentrations of 20–50 µg/kg (Figure 4). In the disks, however, only AZA-1 and AZA-2 (in a ratio of ca 5:1), and no AZA-3 or AZA-6, were detected (Figure 4).

M AN U

Similarly, Fux et al. (2008) found AZA-1 and AZA-2 in a ratio of ca 4:1 together with traces of AZA-3, and recently Krock et al. (2008) isolated and cultured an alga producing AZA-1, AZA-2 and an isomer of AZA-2 but not AZA-3 or AZA-6. This suggests that AZA-3 and AZA-6 may be produced by metabolism of ingested AZA-1 and AZA-2. Little is known about the formation and metabolic transformation of the AZAs, but in shellfish a whole range

CE PT ED

of AZA analogues has been identified (Satake et al., 1998; Ofuji et al., 1999; Ofuji et al., 2001; James et al., 2003; Rehmann et al., 2008).

3.5 Detection of SPXs in the disks

A spirolide, most likely 20-methylSPX-G (Aasen et al., 2005), was also detected in the disks throughout the trial, but only at low levels (ca 5–40 ng/disk, relative to PTX-2). The MRM

AC

transition of 706.5>164.2, which corresponds to 20-methylSPX-G and SPX-C, was chosen based on the findings of (Aasen et al., 2005), where 20-methylSPX-G was found to be the predominant spirolide in algae and shellfish in Norwegian waters. However, analysis of SPXs in shellfish was not performed in this work. Nevertheless, our results show the potential of the passive samplers for monitoring spirolides.

15

ARTICLE IN PRESS

3.6 PTX-2 and PTX-2 SA in disks. The detection of both PTX-2 and PTX-2 SA in the disks shows that formation of seco acids can take place outside of the shellfish, before the algal cells and their PTX-2 are ingested.

IP T

PTX-2 SA has previously been observed as a constituent of Dinophysis (Daiguji et al., 1998; James et al., 1999; Suzuki et al., 2001; MacKenzie et al., 2002) and it appears that conversion

SC R

of PTX-2 into PTX-2 SA can be mediated by enzymes present in the algae (MacKenzie et al., 2002). Esterases responsible for the seco acid formation may leak from damaged algal cells together with PTX-2, resulting in hydrolysis before adsorption to the HP-20 resin in the disk.

M AN U

The ratio of PTX-2 to PTX-2 SA in the disks was typically 1:1, compared to 10:1 PTX-2 in the trial of MacKenzie et al. (2004) in New Zealand, showing that the degree of PTX-2 conversion can vary greatly.

3.7 OA/DTXs and their esters in the disks.

CE PT ED

Dinophysis spp. can contain OA diol esters (Suzuki et al., 2004; Miles et al., 2004b; Miles et al., 2006). However, basic hydrolysis of extracts from passive samplers indicated that little or no OA or DTX esters were present in the disks. MacKenzie et al. (2004) performed basic hydrolysis on some of their samples to determine the levels of esterified DTXs, and found only low amounts of esterified forms (0–30%) in their extracts. Miles et al. (2004b) isolated a substantial amount of OA C8-diol ester from harvested D. acuta cells, and this diol ester was

AC

converted rapidly to OA by a homogenate from the hepatopancreas of the green-lipped mussel (Perna canaliculus). Also it is known that the complex OA-ester DTX-4 is very short lived outside the alga due to enzymatic hydrolysis (Quilliam and Ross, 1996). It is therefore possible that diol esters of OA/DTXs can be hydrolysed by esterases from the Dinophysis in the same manner as DTX-4, prior to adsorption to the HP-20 resin in the disks.

16

ARTICLE IN PRESS

Concluding remarks. The passive sampling disk system is cheap to produce and convenient to use and, when combined with HPLC-MS or ELISA analysis, provides detailed time-averaged information on

IP T

the profile of lipophilic toxin analogues in the water. The passive sampling disks have now been shown to accumulate azaspiracids, okadaic acid analogues, pectenotoxins, yessotoxins

SC R

and spirolides. The HP-20 resin in the samplers should also be able to accumulate other lipophilic algal toxins such as brevetoxin and ciguatoxins. Passive sampling disks have the potential to be a convenient tool for monitoring the exposure of shellfish and other bivalves to

M AN U

toxigenic algae containing lipophilic toxins, and may also be useful for monitoring exposure of aquatic ecosystems to these compounds as well as to a range of lipophilic pollutants.

Acknowledgement

This study was supported by the Norwegian Research Council grant 139593/140, by the

CE PT ED

BIOTOX project (partly funded by the European Commission, through the 6th Framework Programme contract no. 514074, priority Food Quality and Safety, and by the New Zealand Foundation for Research, Science and Technology (FRST) International Investment

AC

Opportunities Fund (IIOF contract number C10X0406).

17

ARTICLE IN PRESS

Captions for Figures

Figure 1. Fully assembled passive sampling disk (E), and its component parts: (A) 100 µm

IP T

nylon mesh; (B) HP-20 resin; (C) inner and (D) outer rings of a 75 mm diameter embroidery

SC R

ring with (F) a No. 2 fishing swivel attached.

Figure 2. A) Concentrations of OA/DTXs in passive sampling disks and shellfish (ng/g), and

M AN U

B) D. acuta + D. acuminata concentration (cells/L) in water for weeks 28–41 of 2005.

Figure 3. Typical MRM HPLC-MS chromatogram of toxins in an extract from a passive sampling disk (week 30) containing 20-methyl-SPX-G, AZA-1, AZA-2, OA, DTX-1, DTX-2, PTX-2, PTX-12 and YTX.

CE PT ED

Figure 4. Chromatogram of AZA profile in extracts of: (A) a passive sampling disk and: (B)

AC

blue mussels (M. edulis).

18

ARTICLE IN PRESS

References

Aasen, J., MacKinnon, S. L., LeBlanc, P., Walter, J. A., Hovgaard, P., Aune, T., Quilliam, M.

IP T

A., 2005. Detection and identification of spirolides in Norwegian shellfish and plankton. Chem. Res. Toxicol. 18, 509-515.

SC R

Aasen, J., Torgersen, T., Dahl, E., Naustvoll, L. J., Aune, T., 2006. Confirmation of

azaspirazids in mussels in Norwegian coastal areas, and full profile at one location. Proceedings of the 5th International Conference on Molluscan Shellfish Safety, 13-18

M AN U

June 2004, Galway, Ireland.

Alvarez, D. A., Petty, J. D., Huckins, J. N., Jones-Lepp, T. L., Getting, D. T., Goddard, J. P., Manahan, S. E., 2004. Development of a passive, in situ, integrative sampler for hydrophilic organic contaminants in aquatic environments. Environ. Toxicol. Chem.

CE PT ED

23, 1640-1648.

Batoreu, M. C. C., Dias, E., Pereira, P., Franca, S., 2005. Risk of human exposure to paralytic toxins of algal origin. Environ. Toxicol. Pharmacol. 19, 401-406. Benes, P., Steinnes, E., 1974. In situ dialysis for the determination of the state of trace elements in natural waters. Water Res. 8, 947-953.

AC

Camacho, F. G., Rodriguez, J. G., Miron, A. S., Garcia, M. C. C., Belarbi, E. H., Chisti, Y., Grima, E. M., 2007. Biotechnological significance of toxic marine dinoflagellates. Biotechnol. Adv. 25, 176-194.

19

ARTICLE IN PRESS

Daiguji, M., Satake, M., James, K. J., Bishop, A., MacKenzie, L., Naoki, H., Yasumoto, T., 1998. Structures of new pectenotoxin analogs, pectenotoxin-2 seco acid and 7-epipectenotoxin-2 seco acid, isolated from a dinoflagellate and greenshell mussels. Chem.

IP T

Lett. 653-654. Duinker, A., Bergslien, M., Strand, O., Olseng, C. D., Svardal, A., 2007. The effect of size

SC R

and age on depuration rates of diarrhetic shellfish toxins (DST) in mussels (Mytilus edulis L.). Harmful Algae 6, 288-300.

FAO/WHO/IOC, 2005. Report of the Joint FAO/IOC/WHO ad hoc Expert Consultation on

M AN U

Biotoxins in Bivalve Molluscs. ftp://ftp. fao. org/es/esn/food/biotoxin_report_en. pdf

Fournier, R. O., 1978. Membran filtering, in: Sournia, A. (Ed), Phytoplankton Manual.

CE PT ED

Monographs on Oceanographic Methodology 6, 108-112.

Fux, E., Marcaillou, C., Mondeguer, F., Bire, R., Hess, P., 2008. Field and mesocosm trials on passive sampling for the study of adsorption and desorption behaviour of lipophilic toxins with a focus on OA and DTX1. Harmful Algae 7, 574-583.

Gustavsson, S. A., Samskog, J., Markides, K. E., Langstrom, B., 12-7-2001. Studies of signal

AC

suppression in liquid chromatography-electrospray ionization mass spectrometry using volatile ion-pairing reagents. J. Chromatogr. A 937, 41-47.

Hallegraeff, G. M., 1993. A review of harmful algal blooms and their apparent global increase. Phycologia 32, 79-99.

20

ARTICLE IN PRESS

Hess, P., Nguyen, L., Aasen, J., Keogh, M., Kilcoyne, J., McCarron, P., Aune, T., 2005. Tissue distribution, effects of cooking and parameters affecting the extraction of azaspiracids from mussels, Mytilus edulis, prior to analysis by liquid chromatography

IP T

coupled to mass spectrometry. Toxicon 46, 62-71.

SC R

Huckins, J. N., Tubergen, M. W., Manuweera, G. K., 1990. Semipermeable membrane

devices containing model lipid: A new approach to monitoring the bioavaiiability of lipophilic contaminants and estimating their bioconcentration potential. Chemosphere

M AN U

20, 533-552.

James, K. J., Bishop, A. G., Draisci, R., Palleschi, L., Marchiafava, C., Ferretti, E., Satake, M., Yasumoto, T., 1999. Liquid chromatographic methods for the isolation and identification of new pectenotoxin-2 analogues from marine phytoplankton and

CE PT ED

shellfish. J. Chromatogr. A 844, 53-65.

James, K. J., Sierra, M. D., Lehane, M., Magdalena, A. B., Furey, A., 2003. Detection of five new hydroxyl analogues of azaspiracids in shellfish using multiple tandem mass spectrometry. Toxicon 41, 277-283.

AC

Kot-Wasik, A., Zabiegala, B., Urbanowicz, M., Dominiak, E., Wasik, A., Namiesnik, J., 2007. Advances in passive sampling in environmental studies. Anal. Chim. Acta 602, 141-163.

Krock, B., Tillmann, U., John, U., Cembella, A. D., 2008. Characterization of azaspiracids in plankton size-fractions and isolation of an azaspiracid-producing dinoflagellate from the North Sea. Harmful Algae, (in press, doi:10.1016/j.hal.2008.06.003).

21

ARTICLE IN PRESS

Larsen, K., Petersen, D., Wilkins, A. L., Samdal, I. A., Sandvik, M., Rundberget, T., Goldstone, D., Arcus, V., Hovgaard, P., Rise, F., Rehmann, N., Hess, P., Miles, C. O.,

IP T

2007. Clarification of the C-35 stereochemistries of dinophysistoxin-1 and dinophysistoxin-2 and its consequences for binding to protein phosphatase. Chem.

SC R

Res. Toxicol. 20, 868-875.

Lindal, O., Lundve, B., Johansen, M., 2007. Toxicity of Dinophysis spp. in relation to

Algae 6, 218-231.

M AN U

population density and environmental conditions on the Swedish west coast. Harmful

Lund, J. W. G., Kipling, C., Cren, E. D., 1958. The inverted microscope method of estimating algal numbers and the statistical basis of estimations by counting. Hydrobiologia 11,

CE PT ED

143-170.

MacKenzie, A. L., Holland, P., McNabb, P., Beuzenberg, V., Selwood, A., Suzuki, T., 2002. Complex toxin profiles in phytoplankton and Greenshell mussels (Perna canaliculus), revealed by LC-MS/MS analysis. Toxicon 40, 1321-1330. .

AC

MacKenzie, L., Beuzenberg, V., Holland, P., McNabb, P., Selwood, A., 2004. Solid phase adsorption toxin tracking (SPATT): a new monitoring tool that simulates the biotoxin contamination of filter feeding bivalves. Toxicon 44, 901-918.

. Miles, C. O., Wilkins, A. L., Hawkes, A. D., Jensen, D. J., Cooney, J. M., Larsen, K., Petersen, D., Rise, F., Beuzenberg, V., MacKenzie, A. L., 2006. Isolation and

22

ARTICLE IN PRESS

identification of a cis-C-8-diol-ester of okadaic acid from Dinophysis acuta in New Zealand. Toxicon 48, 195-203.

IP T

Miles, C. O., Wilkins, A. L., Hawkes, A. D., Selwood, A., Jensen, D. J., Aasen, J., Munday, R., Samdal, I. A., Briggs, L. R., Beuzenberg, V., MacKenzie, A. L., Holland, P. T.,

SC R

2004a. Isolation of a 1,3-enone isomer of heptanor-41-oxoyessotoxin from Protoceratium reticulatum cultures. Toxicon 44, 325-336. .

M AN U

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., 2004b. Isolation of pectenotoxin-2 from Dinophysis acuta and its conversion to pectenotoxin-2 seco acid, and preliminary

CE PT ED

assessment of their acute toxicities. Toxicon 43, 1-9.

Miles, C. O., Wilkins, A. L., Samdal, I. A., Sandvik, M., Petersen, D., Quilliam, M. A., Naustvoll, L. J., Rundberget, T., Torgersen, T., Hovgaard, P., Jensen, D. J., Cooney, J. M., 2004c. A novel pectenotoxin, PTX-12, in Dinophysis spp. and shellfish from Norway. Chem. Res. Toxicol. 17, 1423-1433. .

AC

Ofuji, K., Satake, M., McMahon, T., James, K. J., Naoki, H., Oshima, Y., Yasumoto, T., 2001. Structures of azaspiracid analogs, azaspiracid-4 and azaspiracid-5, causative toxins of azaspiracid poisoning in Europe. Biosci. Biotechnol. Biochem. 65, 740-742.

.

23

ARTICLE IN PRESS

Ofuji, K., Satake, M., McMahon, T., Silke, J., James, K. J., Naoki, H., Oshima, Y., Yasumoto, T., 1999. Two analogs of azaspiracid isolated from mussels, Mytilus edulis, involved

IP T

in human intoxication in Ireland. Nat. Toxins. 7, 99-102.

Palmer, C. M., Maloney, T. E., 1954. A new counting slide for nannoplankton. Limnology

SC R

and Oceanography, Special Publication 21, 1-7.

Quilliam, M. A., Ross, N. W., 1996. Analysis of diarrhetic shellfish poisoning toxins and

M AN U

metabolites in plankton and shellfish by ion-spray liquid chromatography mass spectrometry. Biochem. Biotechnol. Applic. ESI. Mass Spec. 619, 351-364.

Rehmann, N., Hess, P., Quilliam, M. A., 2008. Discovery of new analogs of the marine biotoxin azaspiracid in blue mussels (Mytilus edulis) by ultra-performance liquid

CE PT ED

chromatography/tandem mass spectrometry. Rap. Commun. Mass Spec. 22, 549-558.

Rundberget, T., Sandvik, M., Hovgaard, P., Nguyen, L., Aasen, J. A. B., Castberg, T., Gustad, E., Miles, C., 2006. Use of SPATT disks in Norway: detection of AZA's & DTX's and comparison with algal cell counts and toxin profiles in shellfish. Marine Biotoxin

AC

Science Workshop No. 23. NZFSA, Wellington, New Zealand 37-39.

Satake, M., Ofuji, K., Naoki, H., James, K. J., Furey, A., McMahon, T., Silke, J., Yasumoto, T., 1998. Azaspiracid, a new marine toxin having unique spiro ring assemblies, isolated from Irish mussels, Mytilus edulis. J. Am. Chem. Soc. 120, 9967-9968.

24

ARTICLE IN PRESS

Sodergren, A., 1990. Monitoring of persistent, lipophilic pollutants in water and sediment by solvent-filled dialysis membranes. Ecotoxicol. Environ. Saf. 19, 143-149.

IP T

Stuer-Lauridsen, F., 2005. Review of passive accumulation devices for monitoring organic

SC R

micropollutants in the aquatic environment. Environ. Pollut. 136, 503-524.

Suzuki, T., Beuzenberg, V., MacKenzie, L., Quilliam, M. A., 2004. Discovery of okadaic acid esters in the toxic dinoflagellate Dinophysis acuta from New Zealand using liquid

1138.

M AN U

chromatography tandem mass spectrometry. Rap. Commun. Mass Spec. 18, 1131-

Suzuki, T., MacKenzie, A. L., Stirling, D., Adamson, J., 2001. Pectenotoxin-2 seco acid: a toxin converted from pectenotoxin-2 by the New Zealand Greenshell mussel, Perna

CE PT ED

canaliculus. Toxicon 39, 507-514.

Toyofuku, H., 2006. Joint FAO/WHO/IOC activities to provide scientific advice on marine biotoxins (research report). Mar. Pollut. Bull. 52, 1735-1745.

van Egmond, H. P., Aune, T., Lassus, P., Speijers, G. J. A., Waldock, M., 1993. Paralytic and

AC

diarrhoeic shellfish poisons: Occurence in Europe, toxicity, analysis and regulation. J. Nat. Toxins. 2, 41-83.

Verhaar, H. J. M., Busser, F. J. M., Hermens, J. L. M., 1995. Surrogate parameter for the base-line toxicity content of contaminated water - simulating the bioconcentration of mixtures of pollutants and counting molecules. Environ. Sci. Technol 29, 726-734.

25

ARTICLE IN PRESS

Vrana, B., Mills, G. A., Allan, I. J., Dominiak, E., Svensson, K., Knutsson, J., Morrison, G., Greenwood, R., 2005. Passive sampling techniques for monitoring pollutants in water.

AC

CE PT ED

M AN U

SC R

IP T

Trends Anal. Chem. 24, 845-868.

26

D

C

AC

B

CE PT ED

A

M AN U

SC R

IP T

ARTICLE IN PRESS

F

E

ARTICLE IN PRESS

Sum DTX ng/g in shellfish

500

SC R

400 300 200 100 0 400

B 300 200 100 0

CE PT ED

Cell counts (cell/L)

IP T

600

Sum DTX ng/g disks

A

M AN U

Concentration (ng/g)

700

D. acuta + D. acuminata

AC

28 29 30 31 32 33 34 35 36 37 38 39 40 41 week

5

AC

0

CE PT ED

M AN U

SC R

IP T

ARTICLE IN PRESS

10

Time (min)

15

YTX DTX-1 PTX-2 AZA-2

PTX-12a,b OA, DTX-2 AZA-1 PTX-2SA 20-methyl-SPX-G 20

AC CE PT ED

M AN U

SC R

IP T

ARTICLE IN PRESS