*Article* **A Comparative Analysis of Methods (LC-MS/MS, LC-MS and Rapid Test Kits) for the Determination of Diarrhetic Shellfish Toxins in Oysters, Mussels and Pipis**

**Penelope A. Ajani 1,\* , Chowdhury Sarowar <sup>2</sup> , Alison Turnbull <sup>3</sup> , Hazel Farrell <sup>4</sup> , Anthony Zammit <sup>4</sup> , Stuart Helleren <sup>5</sup> , Gustaaf Hallegraeff <sup>3</sup> and Shauna A. Murray <sup>1</sup>**


**Abstract:** Rapid methods for the detection of biotoxins in shellfish can assist the seafood industry and safeguard public health. Diarrhetic Shellfish Toxins (DSTs) are produced by species of the dinoflagellate genus *Dinophysis*, yet the comparative efficacy of their detection methods has not been systematically determined. Here, we examined DSTs in spiked and naturally contaminated shellfish– Sydney Rock Oysters (*Saccostrea glomerata*), Pacific Oysters (*Magallana gigas*/*Crassostrea gigas*), Blue Mussels (*Mytilus galloprovincialis*) and Pipis (*Plebidonax deltoides*/*Donax deltoides*), using LC-MS/MS and LC-MS in 4 laboratories, and 5 rapid test kits (quantitative Enzyme-Linked Immunosorbent Assay (ELISA) and Protein Phosphatase Inhibition Assay (PP2A), and qualitative Lateral Flow Assay (LFA)). We found all toxins in all species could be recovered by all laboratories using LC-MS/MS (Liquid Chromatography—tandem Mass Spectrometry) and LC-MS (Liquid Chromatography—Mass Spectrometry); however, DST recovery at low and mid-level concentrations (<0.1 mg/kg) was variable (0–150%), while recovery at high-level concentrations (>0.86 mg/kg) was higher (60–262%). While no clear differences were observed between shellfish, all kits delivered an unacceptably high level (25–100%) of falsely compliant results for spiked samples. The LFA and the PP2A kits performed satisfactorily for naturally contaminated pipis (0%, 5% falsely compliant, respectively). There were correlations between spiked DSTs and quantitative methods was highest for LC-MS (r<sup>2</sup> = 0.86) and the PP2A kit (r<sup>2</sup> = 0.72). Overall, our results do not support the use of any DST rapid test kit as a stand-alone quality assurance measure at this time.

**Keywords:** LC-MS; rapid test kit; biotoxins; shellfish; diarrhetic shellfish toxins; *Dinophysis*

**Key Contribution:** LC-MS continues to be a reliable DST detection method across labs and shellfish species; with low to mid-level toxin concentration recovery more variable than high-level. All rapid test kits delivered unacceptably high, falsely compliant results for spiked samples. The Neogen and the PP2A kits performed satisfactorily for naturally contaminated pipis. Overall, our results do not support the use of any DST rapid test kit as a stand-alone quality assurance measure at this time.

## **1. Introduction**

Marine biotoxins are toxic chemical compounds produced by certain microalgae, which can bioaccumulate in shellfish and other marine organisms, and cause poisoning to

**Citation:** Ajani, P.A.; Sarowar, C.; Turnbull, A.; Farrell, H.; Zammit, A.; Helleren, S.; Hallegraeff, G.; Murray, S.A. A Comparative Analysis of Methods (LC-MS/MS, LC-MS and Rapid Test Kits) for the Determination of Diarrhetic Shellfish Toxins in Oysters, Mussels and Pipis. *Toxins* **2021**, *13*, 563. https://doi.org/ 10.3390/toxins13080563

Received: 8 July 2021 Accepted: 9 August 2021 Published: 11 August 2021

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2021 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

seafood consumers. As well as seafood related illnesses, marine biotoxin contamination can lead to damaged public perceptions of seafood, direct economic losses and a restriction in the growth of the shellfish industry.

Diarrhetic Shellfish Toxins (DSTs) are produced by dinoflagellates of the planktonic genus *Dinophysis* and *Phalacroma*, and more rarely benthic *Prorocentrum*, and can bioaccumulate in shellfish and cause Diarrhetic Shellfish Poisoning (DSP). With approximately 11,000 human poisonings reported globally over the period 1985–2018 [1], DSP is a gastrointestinal disorder caused by the human consumption of seafood contaminated with DSTs. While symptoms are dose dependent and include diarrhea, nausea, vomiting and abdominal pain, DSTs are potent inhibitors of certain protein phosphatases and may promote tumor/cancer formation [2], although the impact of chronic exposure to DSTs is still not well known.

DSTs are a group of heat stable, polyether toxins consisting of okadaic acid (OA) and its isomer 19-epi-okadaic acid; the OA congeners dinophysistoxin-1 (DTX-1) and dinophysistoxin-2 (DTX-2); and the 7-acyl derivatives of OA, DTX-1 and DTX-2 that are collectively known as DTX-3. Together, they are referred to as the OA group toxins or the 'okadaates' (OAs). While OA, DTX-1 and DTX-2 only differ slightly in their molecular structure, the DTX-3 (group) includes a wide range of derivatives esterified with saturated and unsaturated fatty acids, products of metabolic transformations that occur in the shellfish [3]. Chemical compounds of this group are therefore generally described as either 'free' (unesterified) or 'esterified' [4].

DSP was first described after a large toxin event occurred in Japan in 1976 [5,6], whereby many people became sick after eating scallops (*Patinopecten yessoensis*). This contamination was linked to toxins produced by *Dinophysis fortii*. Following this event, further toxic episodes occurred in Japan, Spain and France, with several thousands of cases of human poisonings occurring over the 1970s and 1980s, and leading to the development of many regional monitoring programs. This monitoring has seen a gradual increase in reported DSP episodes in countries including Chile, Argentina, Mexico, the east coast of North America, Scandinavia, Ireland, Great Britain, Spain, Portugal, Italy, Greece, India, Thailand, Australia and New Zealand [5,7–9].

*Dinophysis* is common in Australian waters, with 36 species reported [10–12]. Toxic species include *D. acuminata* Claparede and Lachmann, *D. acuta* Ehrenberg, *D. caudata* Saville-Kent, *D. fortii* Pavillard, *D. norvegica* Claparede and Lachmann and *D. tripos* Gourret. There have been three serious DSP events in Australia. The first episode was caused by contamination of Pipis (*Plebidonax deltoides*) in New South Wales in 1997 (NSW) by *D. acuminata* [13]. One hundred and two people were affected and 56 cases of gastroenteritis were reported. A second episode occurred again in NSW in March 1998, this time with 20 cases of DSP poisoning reported [14]. The final event occurred in Queensland in March 2000, which was again linked to the consumption of Pipis [15]. While no human fatalities from DSP are known globally, DSTs continue to be a major food safety challenge for the shellfish industry.

Detection methods for DSTs using liquid chromatography with tandem mass spectrometry (LC-MS/MS and LC-MS) [4,16] and implemented as part of seafood safety programs, are considered the "gold standard" across the globe. These methods replaced the mouse bioassay (MBA), which was previously the most commonly used laboratory analysis tool (e.g., [17]). However, the development of more rapid, cost effective (on farm) testing methods for the presence of DSTs would potentially make harvest management simpler and faster and result in fewer closures. Three types of rapid test kits for the detection of DSTs are currently commercially available. These include an antibody-based enzyme-linked immunosorbent assay (ELISA) test; a functional protein phosphatase inhibition activity (PPIA) assay; and a lateral flow analysis (LFA) rapid test. ELISA assays involve an antigen immobilized on a (micro) plate, which are then complexed with an antibody that is linked to a reporter enzyme. These assays were first developed in the 1960s and 1970s for primarily medical diagnosis purposes [18]. Detection of OA, DTX-1 and DTX-2 (varying analogue

cross reactivity depending on kit) is accomplished by assessing the conjugated enzyme activity via incubation with a substrate to produce a quantifiable product. Functional PPIA assays quantify okadaic acid (OA) and DST analogues including DTX-1, DTX-2 and DTX3 by colorimetric phosphatase inhibition, based on the reversible inhibition of protein phosphatase type 2A (PP2A) by the toxin, and the resulting absorbance derived from enzymatic hydrolysis of the substrate. A lateral flow test involves the shellfish extract transported across a reagent zone in which OA, DTX-1, DTX-2 and DTX-3 specific antibodies are combined with colored particles. If a toxin is present, it is captured by the particle-antibody complex, and as its concentration increases, the intensity of the test "line" decreases [19].

In a comprehensive review by McLeod et al. [20] of the currently available field methods for detection of marine biotoxins in shellfish, it was concluded that the ELISAs and LFAs had poor reactivity to the DSP congener DTX-2 and can give false negative results when high levels of DTX-3 are present (and the hydrolysis step is not undertaken to release ester forms). LFAs were also found to give some false positive results when DSP was below the ML (Max Limit), but this was dependent on the toxin profile, geographic region and shellfish species involved. Pectenotoxins (PTXs) are not currently included in Codex Standard for Live and Raw Bivalve Molluscs [21], and therefore are not included in this study. Several other jurisdictions such as Canada, Chile and the European Union do regulate for PTX (but not PTX-2sa), but the European Food Safety Authority has issued an opinion to deregulate PTX [22]. Furthermore, DSP regulation in Australia is governed by Food Standards Australia New Zealand with a maximum regulatory limit of 0.2 mg OA eq/kg [23], while most international standards including the Codex Standard, state a ML of 0.16 mg OA eq/kg [21].

To date, these rapid detection kits have not been tested on various shellfish matrices in a systematic manner, nor a comparison made across multiple analytical laboratories to assess LC-MS/MS or LC-MS detection of DSTs in shellfish. With this in mind, the present study aimed to undertake a comparative study to detect DSTs in differing shellfish matrices using commonly implemented protocols for LC-MS/MS or LC-MS in several different laboratories, as well as compare five commercially available rapid test kits for the detection of DSTs in these same shellfish tissues. The rapid test kits included three quantitative ELISA kits by BeaconTM, Eurofins/AbraxisTM and EuroProximaTM; a quantitative PP2A kit by Eurofins/AbraxisTM, and a qualitative LFA kit by NeogenTM.

#### **2. Results**

#### *2.1. LC-MS/MS and LC-MS*

No toxins were detected in any of the four shellfish species matrices (Sydney Rock Oysters (*Saccostrea glomerata*) (SRO), Pacific Oysters (*Magallana gigas*/*Crassostrea gigas*) (PO), Blue Mussels (*Mytilus galloprovincialis*) (MUS) and Pipis (*Plebidonax deltoides*/*Donax deltoides*) (PIPI)) screened before spiking began (see Methods). Of the triplicate SROs spiked with OA at 0.02 mg/kg, Laboratory 1 detected OA in all three samples (x = 0.01, SD ± 0.00, min <0.01, max 0.02 mg/kg), Laboratory 2 and 4 reported concentrations below the detection limit for all samples (<0.01 mg/kg and <0.025 mg/kg respectively), and Laboratory 3 detected OA in all three samples (x = 0.013, SD ± 0.006, min 0.01, max 0.02 mg/kg). In summary, two out of the four laboratories detected OA at this low level, with recoveries between ~50–100% (Table 1).

Of the four shellfish species spiked with OA at 0.02 mg/kg, Laboratory 1 detected this toxin in all four matrices (x = 0.013, SD ± 0.005, min 0.01, max 0.02 mg/kg), Laboratory 2 did not detect OA in SRO or PO; however, it was detected in both MUS and PIPI (x = 0.015, SD ± 0.007; min <0.01, max 0.02 mg/kg), and Laboratory 3 did not detect OA in PO or MUS, but detected it in SRO and PIPI (x = 0.015, SD ± 0.007; min <0.01, max 0.02 mg/kg). Laboratory 4 did not detect OA at this concentration (less than detection limit <0.025 mg/kg). Laboratory 4, however, did detect OA in one PIPI sample at 0.03 mg/kg (>spike concentration). In summary, OA was detected in all matrices at this concentration, although not all laboratories detected toxins in all four matrices. Recovery across all laboratories ranged from ~50–150% (Table 2).

**Table 1.** Results of LC-MS/MS (Liquid Chromatography—tandem Mass Spectrometry) and LC-MS (Liquid Chroma-tography—Mass Spectrometry) for Sydney Rock Oysters (SRO) spiked with 0.02 mg/kg okadaic acid (no DTX-1 or DTX-2 added).


<LOR = below limit of reporting; Note: Spike below limit of reporting for Laboratory 4.

**Table 2.** Results of LC-MS/MS and LC-MS for Australian shellfish—Sydney Rock Oysters (SRO), Pacific Oysters (PO), Blue Mussels (MUS) and Pipis (PIPI) spiked with 0.02 mg/kg okadaic acid (no DTX-1 or DTX-2 added).


<LOR = below limit of reporting; Note: Spike below limit of reporting for Laboratory 4.

For the shellfish spiked with DTX-1 at 0.04 mg/kg, Laboratory 1 recovered this analogue in all matrices (x = 0.035, SD ± 0.006; min 0.03, max 0.05 mg/kg), with one PIPI sample returning a concentration of 0.01 OA mg/kg. Laboratory 2 detected DTX-1 in all matrices (x = 0.025, SD ± 0.006; min 0.02, max 0.03 mg/kg), also with a detection of OA in PIPI at 0.02 mg/kg. Laboratory 3 detected DTX-1 in all matrices (x = 0.025, SD ± 0.006; min 0.02, max 0.03 mg/kg), while Laboratory 4 did not detected this toxin in MUS (other matrices x = 0.026, min <0.025, max 0.04 mg/kg) (Table 3). In summary, DTX-1 was detected in all shellfish matrices at this concentration; however, one laboratory did not detect DTX-1 in MUS. The overall recovery of this analogue was ~50–100% across laboratories with two detections of OA in PIPIs.

For all shellfish spiked with DTX-2 at 0.01 mg/kg, Laboratory 1 did not recover this analogue in SRO or PIPI, and was only detected it in PO and MUS (both at 0.01 mg/kg). No toxin at this concentration was recovered from either Laboratory 2 nor Laboratory 3, while Laboratory 4 was unable to detect this toxin (below the limit of reporting <0.025 mg/kg) (Table 4). In summary DTX-2 was only detected in PO and MUS at this low concentration, and only at one laboratory. Overall recovery was ~50–100%.


**Table 3.** Results of LC-MS/MS and LC-MS for Australian shellfish—Sydney Rock Oysters (SRO), Pacific Oysters (PO), Blue Mussels (MUS) and Pipis (PIPI) spiked with 0.04 mg/kg DTX-1 (no OA or DTX-2 added).

<LOR = below limit of reporting.

**Table 4.** Results of LC-MS/MS and LC-MS for Australian shellfish—Sydney Rock Oysters (SRO), Pacific Oysters (PO), Blue Mussels (MUS) and Pipis (PIPI) spiked with 0.01 mg/kg DTX-2 (no OA or DTX-1 added).


<LOR = below limit of reporting; Note: Spike below limit of reporting for Laboratory 4.

When shellfish were spiked with all toxins (in varying concentrations between 2–10 × LOR depending on toxin analogue; see Methods), laboratory recovery of total toxin per sample for each laboratory was as follows: Laboratory 1: 53–75%; Laboratory 2: 35–88%; Laboratory 3: 13–41%; and Laboratory 4: 0–88% (Table 5). More specifically, all toxins were recovered in all matrices for Laboratory 1, with an individual toxin/sample recovery ranging from 40–200%, with the lowest matrix average recovery in SRO at 57% and the highest in PIPI at 103%. For Laboratory 2, DTX-2 was not detected in SRO or PO, while individual toxin/sample recovery ranged from 40–400%, with the lowest matrix average recovery in SRO at 43%, and the highest in PIPI at 170%. For Laboratory 3, OA was not detected in MUS or PIPI, and DTX-2 was not detected in PIPI. The individual toxin/sample recovery ranged from 20–50%, with the lowest matrix average in PIPI at 40% and the highest in MUS at 47%. Finally, for Laboratory 4, DTX-2 was not detected across all matrices and OA was not detected in MUS. Individual toxin/sample recovery ranged from 50–340% with the lowest matrix average in MUS at 50% and the highest in PIPI at 154%. Overall, most toxins were detected by all laboratories at these concentrations, individual recovery across all labs/matrices ranged from 0–88%, while the recovery across shellfish matrices varied.

In our final analysis to determine the recovery of CRM (OA/DTX-1/DTX-2), all laboratories detected all toxin analogues. Individual toxin recoveries ranged from 88 to 131% for Laboratory 1, 79–81% for Laboratory 2, 83–95% for Laboratory 3 and 101–262% for Laboratory 4 (Table 6). However, considering that these recoveries are the result of one sample per lab, they should be treated as indicative only.


**Table 5.** Results of LC-MS/MS and LC-MS for Australian shellfish—Sydney Rock Oysters (SRO), Pacific Oysters (PO), Blue Mussels (MUS) and Pipis (PIPI) spiked with a combination of DST analogues-OA 0.1 mg/kg; DTX-1 0.05 mg/kg; and DTX-2 0.02 mg/kg.

<LOR = below limit of reporting; Note: Spike of OA for MUS and PIPI below limit of reporting for Laboratory 4.

**Table 6.** Results of LC-MS/MS and LC-MS for Certified Reference Material CRM DSP-Mus-c.


\* CRM are certified for free toxin; they report higher total toxin concentration post hydrolysis but these are not certified.

#### *2.2. Rapid Test Kits*

#### 2.2.1. Wild Harvest Pipis

Prior to rapid test kit screening, OA, DTX-1 and DTX-2 analysis by LC-MS for wild harvest Pipis resulted in a OA toxin range of 0.1 to 0.3 mg/kg (Sample 4A—0.1 mg/kg, 4B—0.1 mg/kg, 4C—0.2 mg/kg, and 4D—0.3 mg/kg). After hydrolysis, no DTX-1 or DTX-2 was detected in any samples. Three batches comprising 10 replicates of each OA toxin concentration of 0.1, 0.2 and 0.3 mg/kg were subsequently screened using each rapid test kit.

#### 2.2.2. LC-MS

Using LC-MS (Laboratory 3), all control shellfish samples (no toxin added) returned a 'not detected' result (Table 7). For OA spiked samples, 43/46 (~93%) returned concentrations at, or slightly above, the spiked toxin concentrations 0.1 and 0.2 mg/kg (Tables 7 and 8). The three samples (7%) that returned concentrations lower that the spiked

concentration were all spiked Pipi samples: sample 22 reported 0.09 mg/kg when it was spiked with OA at 0.1 mg/kg; sample 23 reported 0.15 mg/kg when it was spiked with OA at 0.2 mg/kg; and finally, sample 24 reported 0.09 mg/kg when it was spiked with OA at 0.2 mg/kg (Tables 7 and 8). The latter two of these samples were falsely compliant at the regulatory limit (7%, 2/28). A Pearson's correlation analysis between LC-MS results and the concentration of spiked toxin revealed a very strong relationship (r<sup>2</sup> = 0.86) (Figure 1). Subsequently, this method returned a mean recovery of 106.5%, meeting the criteria set out in the AOAC Guidelines for Single Laboratory Validation of Chemical Methods for Dietary Supplements and Botanicals (AOAC 2002).

**Table 7.** Results of LC-MS and rapid test kits for Okadaic Acid spiked into Australian shellfish (Sydney Rock Oysters [SRO], Pacific Oyster [PO], Blue Mussel [MUS] and Pipis [PIPI]). Note: Neogen qualitative test (±) with Limit of Quantification = 0.08 mg/kg; Abraxis PP2A Working Range = 0.06 to 0.35 mg/kg; Beacon ELISA Limit of Quantification = 0.1 mg/kg; Abraxis ELISA Working Range = 0.1–5.0 mg/kg; Europroxima ELISA Limit of Quantification = 0.04 mg/kg.


ND = not detected (0.01 mg/kg detection limit).

2.2.3. Rapid Test Kits

Qualitative Test

Neogen

The Neogen kit returned negative readings for the eight negative control samples across all species-specific shellfish matrices. However, 23 out of 46 samples (50%) of spiked samples (across all shellfish matrices) returned a negative result when they contained okadaic acid (Tables 7 and 8). Within this group, 18% (5/28 samples again across all matrices) returned a false compliant result when they were spiked at, or above, the regulatory limit (=/> 0.2 mg OA eq/kg), while no naturally contaminated Pipis returned falsely compliant results with this kit.
