1. Introduction
With the increasing use of fertilizers in agriculture, discharge of industrial and domestic wastewater, coastal eutrophication, climatic change, the world trade market, overseas transportation and harmful algal blooms have become widespread marine environmental problems, particularly in large river estuaries and semi-enclosed bays [
1,
2]. Marine algal toxins are secondary metabolites produced by toxic marine algae that can accumulate in bivalve mollusks through filter-feeding behavior [
3]. Paralytic shellfish toxins (PSTs) are a group of neurotoxic algal toxins that are produced mainly by marine dinoflagellates belonging to the genera
Alexandrium,
Gymnodinium and
Pyrodinium, which are distributed throughout the world [
4,
5]. Depending on their chemical structure (different R
4 groups), PSTs can be reorganized into several subgroups (
Figure 1) [
6]. For example carbamate (saxitoxin (STX), neoSTX, gonyautoxins (GTX1, GTX4, GTX2, GTX3)), N-sulfocarbamoyl (GTX5) and decarbamoyl toxins (dcSTX, dcneoSTX) [
7,
8]. As PSTs readily accumulate in filter-feeding bivalve mollusks, the potential risk of shellfish to consumers is considered [
9,
10].
PST poisoning events in Europe and Japan have been reported since the late 1970s and early 1980s and have mainly occurred in countries near the coastlines of the Atlantic and Pacific Oceans [
11,
12,
13]. PST poisonings in some coastal cities have also been recorded in China in the past two decades [
13,
14]; for example, in 2017, 164 human intoxications were recorded in Zhangzhou City, Fujian Province, due to the presence of PSTs in shellfish (oysters and mussels) [
13]. The main clinical manifestations of poisoning were muscle paralysis, difficulty breathing and weakness, and the average latency was 3 h; 8 out of the 10 shellfish samples collected had PSTs [
13].
Additionally, some studies have shown that the PST levels in shellfish differ among different sea areas in China and that the PST levels are easily affected by red tides [
15,
16]. Because the Bohai Sea is a semi-enclosed inland sea, red tides occur more frequently. Du et al. collected 21 scallop samples from Dalian, Liaoning Province, and PSTs were detected in 19 samples with the highest concentration being 23.84 MU/g, which exceeded the limit of China’s national food safety standard (GB2733-2015) [
15,
17]. This study suggested that the presence of high PST levels in shellfish may be related to toxic algae [
15]. However, national monitoring of PSTs in shellfish has been carried out in China since 2016, but all shellfish samples were collected from the local market, and data on PST contents during farming and fishing are lacking.
In 2009, 2004 and 2016, the European Food Safety Authority (EFSA) and the Joint FAO/WHO Expert Committee on Food Additives (JECFA) estimated the acute dietary exposure to PSTs in shellfish, and the results showed that consumers with high levels of shellfish consumption were concerned [
18,
19,
20]. Most studies on acute dietary exposure to PSTs have adopted the method of point assessment, but the high percentile values of PST content in shellfish and shellfish consumption have been found in point assessments; for example, P99 of the PST content and P99 of shellfish consumption significantly overestimate these results [
9,
21]. Zhou et al. used the point assessment method to estimate the acute dietary exposure to PSTs in scallop consumers; this value exceeded the acute reference dose (ARfD) of 0.5 μg STX.2HCl eq./kg bw established by the EFSA [
9]. However, studies on the accurate estimation of acute dietary exposure to PSTs among Chinese consumers, especially those who prefer to eat shellfish in coastal areas, are lacking.
Boundy M. and Turner A.D. et al. developed the HILIC liquid chromatography coupled to mass spectrometry in tandem (HILIC–LC–MS/MS) method for the analysis of PSTs [
6,
7]. This method was also used to analyze the PST content in shellfish in this study. The purpose of this study was to understand the contamination status and temporal variation in the contaminant levels of PSTs in shellfish in the Dalian area from the Yellow-Bohai Sea from 2019 to 2020. Moreover, for a more accurate assessment, a probabilistic approach was used to assess the health risks of acute dietary exposure to PSTs in shellfish for general consumers and consumers in coastal areas in China.
2. Materials and Methods
2.1. Chemicals and Reagents
The certified reference materials of GTX1&4, GTX2&3, dcGTX2&3, neoSTX, STX, dcSTX and GTX5 were purchased from the National Research Council-Institute for Marine Biosciences (Halifax, NS, Canada). HPLC-grade acetonitrile was obtained from TEDIA (Fairfield, OH, USA). HPLC-grade formic acid, ammonium formate and analytical grade ammonium hydroxide (>25%, w/w) were obtained from Thermo Fisher Scientific (Shanghai, China). The water used in this study was 18.2 MΩ cm−1 ultrapure water which was prepared by Millipore (Bedford, MA, USA).
2.2. Standard Solution Preparation
Mixed standard intermediate solutions of STX, neoSTX, dcSTX, GTX5, GTX1, GTX4, GTX2, GTX3, dcGTX2 and dcGTX3 were prepared with different concentrations of matrix solution (
Table 1). All the standard solutions were stored at −20 °C and kept away from light.
2.3. Sample Collection
A total of 199 shellfish samples were collected, including 34 Pacific oysters (Crassostrea gigas), 25 Mediterranean blue mussels (Mytilus galloprovincialis), 34 Manila clams (Ruditapes philippinarum), 36 bay scallops (Argopecten irradians), 34 veined rapa whelks (Rapana venosa) and 36 blood clams (Scapharca subcrenata). All these samples were collected from four representative aquaculture zones around the Yellow-Bohai Sea in the Dalian area from October 2019 to October 2020. The sampling sites were Qidingshan Street (Manila clams), Dalijia Street (bay scallops and Pacific oysters), Daweijia Street (Mediterranean blue mussels) and Xingshu Street (veined rapa whelks and blood clams), Jinzhou District, Dalian. In this study, according to the occurrence time of toxic microalgae in the Yellow-Bohai Sea, shellfish samples were collected once every two weeks from October 2019 to March 2020 and once a week from April 2020 to October 2020 (no sampling occurred in February and August due to COVID-19), for a total number of sampling times of 35.
After each sample was collected, the sediment on the shell of the shellfish samples was washed with clean seawater, placed separately in a clean plastic bag, and sealed to avoid external contamination. All the samples were stored in ice bags at low temperatures and transported to the laboratory at −4 °C for analysis as soon as possible. All the samples were collected in duplicate.
2.4. Sample Preparation
Sample preparation was carried out according to the methods of Tuner et al., and the steps are summarized as follows [
7]. Each shellfish sample was approximately 2.5 kg in weight. The sample was washed and shaken. After the silt inside the shellfish was washed away, the edible flesh tissues (including digestive glands) were removed and subsequently homogenized. A total of 2.0 g of homogenized tissue was weighed into a 10 mL centrifuge tube, and 8 mL of 0.5% (
v/
v) acetic acid aqueous solution was added. The mixture was vortexed for 90 s and then extracted by ultrasonication for 5 min. The centrifuge tube was then sealed in boiling water for 5 min and cooled to room temperature quickly. The extract was centrifuged at 10,000 rpm for 5 min and prepared for purification.
An aliquot of 1.0 mL of the supernatant was decanted into a 2.0 mL centrifuge tube, 5 μL of ammonium hydroxide was added, and the mixture was mixed well. A solid-phase extraction (SPE) procedure using Supelco ENVI-Carb cartridges (250 mg/3 mL, Supelco) was used for shellfish extract purification before the chromatographic analysis. The stationary phase was conditioned with 3.0 mL of aqueous acetonitrile (20%, v/v), which contained acetic acid (0.8%, v/v) and 3.0 mL of ammonium hydroxide (0.1%, v/v). An aliquot of 250 μL of extract was loaded onto the SPE column. The stationary phase was then rinsed with 700 μL of ultrapure water and drained for 5 s. The toxins were subsequently eluted with 2 mL of aqueous acetonitrile (20%, v/v) containing 0.8% of acetic acid (v/v) and drained under a vacuum for 5 s. The eluent was collected in a 5 mL centrifuge tube, mixed, and filtered through a 0.22 μm syringe membrane filter prior to chromatographic analysis.
2.5. LC–MS Analysis
The purified sample extracts and standards were analyzed using a HILIC–LC–MS/MS [
7]. An 8060 LC–MS triple quadrupole mass spectrometer (Shimadzu, Kyoto, Japan) coupled with an LC-30AD ultrahigh-performance liquid chromatography system (Shimadzu, Kyoto, Japan) was used. An XBridge Amide column (Lot No.0163370685, 3.0 mm × 150 mm, 1.7 µm; Waters, Milford, MA, USA) was used for analysis. Mobile phase A was 10 mmol/L formic acid and 5 mmol/L ammonium formate, and mobile phase B was acetonitrile. The linear gradient program was performed as follows: 0–0.01 min, 85% B; 0.01–5 min, 85% to 45% B; 5–10 min, 45% B; 10–10.5 min, 45% to 85% B; and 10.5–14 min, 85% B. The injection volume was 5 μL at a flow rate of 0.4 mL/min for the mobile phase, and the column temperature was set at 40 °C.
The tandem mass spectrometer equipped with an electrospray source acquired masses in fast polarity switching (negative and positive). Mass acquisition was carried out in multiple reaction monitoring (MRM) mode. The electrospray ion source parameters were as follows: interface voltage 4500 V; desolvent line; MS interface and heat block temperatures 250 °C, 300 °C and 400 °C, respectively; heating gas (air, 10 L/min); nebulizing gas (nitrogen, 3 L/min); drying gas (nitrogen, 10 L/min); and collision gas (argon, 270 kPa). The MS parameters are listed in
Table 2.
2.6. Analytical Validation
The LC–MS/MS method was validated in the laboratory, and the validation parameters of the method are presented in
Table 3. The limits of detection (LODs) were 5–20 µg/kg according to this method (
Table 2). The chromatogram of PST is shown in
Figure 2. The precision and accuracy of the method satisfied the quality-control requirements of this study.
2.7. Acute Dietary Exposure Risk Assessment for PSTs from Shellfish for Chinese General Consumers
2.7.1. Shellfish Consumption Data of Chinese General Consumers
Shellfish consumption data were obtained from the China National Food Consumption Survey, which was conducted in 2018–2020 [
22]. This survey used multistage random cluster sampling and was conducted in 18 provinces (12 coastal and 6 inland provinces) in China. Food consumption was recorded using three discontinuous, face-to-face interviews of 24 h diet recalls that were performed on one weekend day (Saturday or Sunday) and two weekdays, and two adjacent surveys of the three interviews were conducted at least five days apart. In this survey, a total of 55,700 participants, including 3200 shellfish consumers aged 3 to 92 years were surveyed. The shellfish commonly consumed included scallops, mussels, oysters and blood clams (
Table 4).
2.7.2. The Equivalent Concentrations of STX in Shellfish Samples
The toxicity equivalence factors (TEFs) of STX and its analogs, published by JECFA in 2016, were used to define the toxicity ratio of a compound from a chemical group that shares the same mode of action as a reference compound in the same group [
20]. The toxicity of the analog was expressed as a fraction of the toxicity of the reference compound [
20].
Table 5 presents the TEFs published by JECFA. In this study, the total toxin group of STX and its analogs in the shellfish samples were expressed in μg STX.2HCL equivalents (eq.) per kg shellfish tissue, using the TEFs published by JECFA.
2.7.3. Probabilistic Model of Acute Dietary Exposure to PSTs from Shellfish among Chinese General Consumers
Monte Carlo (MC) simulations were used for the probabilistic modeling of acute dietary exposure [
23]. Daily shellfish consumption was simulated by sampling from the food consumption database combined with a random sample from an empirical STX equivalent concentration distribution for each shellfish. Due to the limited sample size of shellfish in this study, a
lognormal distribution was used to fit the STX equivalent concentrations in the shellfish samples [
24]. The daily acute dietary exposure to PSTs (μg STX.2HCl eq./kg bw) from shellfish was calculated by summing the intake of shellfish consumed in a single day, adjusted by the measured individual body weight of each consumer. The shellfish consumption and STX equivalent concentration data were harmonized by coding food according to the Codex Classification of Foods and Animal Feeds promoted by the Codex Alimentarius Commission (CAC) [
25,
26]. Based on the WHO-recommended approach for undetected values, the worst case was used to conservatively estimate acute dietary exposure to PSTs by shellfish consumers; i.e., the undetected values were assumed to be LODs [
27].
The number of MC iterations was 100,000 for each group of the population to be assessed. The acute dietary exposures were specified at the percentiles P50, P75, P90, P99 and P99.9 and at the mean value of the intake distribution. Bootstrap resampling was used to evaluate the uncertainty of the MC variation in consumers and compound concentrations quantitatively, and relevant statistics (such as P50, P95 and P99) and approximate confidence intervals (95% CIs) of the percentiles were calculated. The bootstrap sample in this study was B = 100.
The results of daily acute dietary exposures were subsequently compared with the ARfD of STX (0.5 μg STX.2HCl eq./kg bw) established by the EFSA, which estimated the probability of consumers with acute health risks [
18].
2.8. Estimation of Acute Dietary Exposure to PSTs from Shellfish by Consumers in Coastal Areas of China
Given that PST poisoning mainly occurs in coastal areas, a specific survey of shellfish consumption was carried out in coastal areas in China to assess the acute health risk of PSTs to typical consumers.
2.8.1. Shellfish Consumption Data of Special Consumers in the Bohai Rim Region
A specific survey of shellfish consumption in the Bohai Rim region in China was conducted in 2019 [
28]. This survey involved random sampling in local main aquatic product markets, communities, fishing villages and large factories. The consumption frequency, consumption and cooking habits of shellfish were recorded via face-to-face interviews with diet recalls of the month with the highest shellfish consumption. In this survey, a total of 875 participants aged 13 to 91 years were shellfish consumers, including 481 males and 394 females, representing the consumers who frequently consumed shellfish in the Bohai Rim area
. The detailed consumption data for consumers are shown in
Table 6.
2.8.2. Probabilistic Model of Acute Dietary Exposure to PSTs from Shellfish Consumed by Typical Consumers
The acute dietary exposure of shellfish consumers to PSTs from shellfish in coastal areas of China was calculated probabilistically using @Risk software (version 8.2, Palisade). The probabilistic modeling used in this study was implemented by Monte Carlo simulations, which simulate the daily consumption by sampling from the shellfish consumption distribution and combining these data with a random sample from the distribution of equivalent concentrations of PSTs. Random sampling from the concentration distribution was performed according to the percentage of the sample with detectable concentrations and the percentage of the sample with undetectable concentrations. Each consumer’s acute dietary exposure to PSTs from shellfish (μg STX.2HCl eq./kg bw) was calculated by shellfish consumption combined with the equivalent concentrations of PSTs in shellfish on a single day, adjusted by the measured individual body weight of each consumer.
@Risk software was used to fit the equivalent concentration data, shellfish consumption data and body weight data to obtain the appropriate distribution: the
lognorm distribution for the equivalent concentrations of STX in shellfish samples above the LOD. Considering the uncertainty and variability in the concentrations of the undetected samples, the equivalent concentrations of PSTs in the undetected samples were assumed to have a distribution ranging from 0 to the maximum LOD for PSTs, also expressed as the equivalent concentrations of PSTs. A
Uniform distribution was assumed for the equivalent concentrations of PSTs of the LOD in the case of undetectable samples. The
Pert distribution was applied to consumption data. The
Normal distribution was used for shellfish consumers’ body weights. The details of the variables and models used for this acute exposure assessment are shown in
Table 7. The number of Monte Carlo iterations was 100,00, and the number of simulations was 100. The exposures were specified at percentiles P50, P90, P95, P97.5 and P99 and the mean from the intake distribution and compared with the ARfD of the STX.
2.9. Statistical Analysis
The data were analyzed using SPSS (version 25.0 for Windows, Armonk, NY: IBM Corp., USA). The PST levels in the samples are shown as the minimum to maximum values of the detected PSTs. The acute dietary exposure of Chinese general consumers to PSTs from shellfish was statistically analyzed via SAS software (version 9.4). @Risk software was used to carry out acute dietary exposure assessments of PSTs for typical consumers in coastal areas of China.
4. Conclusions
The current study determined the PST levels in shellfish from 2019 to 2020 in the Dalian area from the Yellow-Bohai Sea using the HILIC–LC–MS/MS method and estimated their potential acute health risks to general Chinese consumers and typical consumers in coastal areas of China. Among the samples, scallops and blood clams were the shellfish species with the highest detection rate of PSTs (94.4%), and the highest total content of PSTs was detected in scallops (3953.5 μg STX.2HCl eq./kg), followed by blood clams (993.4 μg STX.2HCl eq./kg). The highest detection rate of PSTs was in the scallop viscera (94.4%), and the maximum level of total PSTs was in the scallop viscera (2605.5 μg STX.2HCl eq./kg). Consumption of scallop viscera was the most significant risk factor for PST poisoning. In addition, autumn was the season of concern for PST contamination and seafood poisoning in the Yellow-Bohai Sea in Dalian. For the Chinese general consumers, the probability of acute health risks to shellfish consumers from dietary exposure to PSTs was 13.14%; for typical consumers in the coastal areas of China, especially those with higher shellfish intake, there was an acute health risk associated with exposure to PSTs through bivalve consumption during the occurrence of harmful algal blooms. Therefore, it is suggested that the government continue to strengthen the monitoring of the source of PSTs and the monitoring of harmful algal blooms and reduce the production of industrial and domestic wastewaters, as well as give reasonable advice on shellfish consumption for the consumers in coastal areas, such as not eating scallop viscera.