2.1 Sampling
Every month from March 2019 to February 2020, shellfish samples were collected from the Buji seafood wholesale market in Luohu, which supplies 90% of shellfish consumed in Shenzhen. Other shellfish were collected from local retail stores. In total, 188 shellfish samples of 14 species were collected, including: Chlamys nobilis (scallop), Crassostrea rivularis (oyster), Perna viridis (mussel), Paphia undulata (clam), Atrina pectinate (bivalve), Babylonia areolate (sea snail), Scapharca broughtoni (clam), Scapharca subcrenata (bivalve), Meretrix meretrix (clam), Sinonovacula constricta (razor clam), Ruditapes philippinara (clam), Patinopecten yessoensis (scallop), Solen gouldi (bivalve), and Mytilus edulis (mussel).
Since the year-round availability of three shellfish species (Solen gouldi, Mytilus edulis and Scapharca subcrenata) was limited, their sample sizes were small relative that of other species.
2.2 Reagents
Methanol, acetonitrile, and formic acid were purchased from Merck (Darmstadt, Germany).
Methylene chloride was obtained from Damao Chemical Reagent Factory (Tianjin, China). Water was distilled and passed through a MilliQ water purification system (Millipore, Burlington, Mass. USA). Ammonium formate was purchased from Sigma (Bangalore, India). PST standard reagents (C1, C2, dcGTX2, dcGTX3, dcNEO, dcSTX, GTX1, GTX2, GTX3, GTX4, GTX5, STX, NEO) were purchased from Bedford Institute of Oceanography (Dartmouth, Nova Scotia, Canada).
2.3 Toxin extraction
Shellfish samples were cleaned with fresh water, emoved from their shells, and rinsed again with MiliQ water to remove silt from the flesh. Shellfish meat and viscera (200g of each batch) was removed and homogenized. An 5 g aliquot of homogenized sample was transferred to a centrifuge tube, 8 mL of 0.5% formic acid solution added, for vortex-mixed for 5 min, ultrasonically extracted for 10 min, followed by centrifugation at 9500 rpm/min for 10 min. The suspension was immersed in 0.5% formic acid solution to 10 mL. Two mL of this mixture were added to 5 mL dichloromethane, vortex-mixed for 3 min, and then centrifuged for 5 min at 9500 rpm. Then, 1 mL suspension was passed through a C-18 solid-phase extraction cartridge (Waters, Millford, Mass., USA) that had been activated by adding 3 mL acetonitrile and 3 mL 0.5% formic acid solution in that order, and then immersed in 1.5 mL 0.5% formic acid solution. The eluate was collected, adjusted to 6 mL acetonitrile solution, and vortexed. The resulting solution was allowed to rest for 30 min in a -4℃ refrigerator, then centrifuged at 9500 t/min for 5 min, passed through 0.22 μm nylon syringe filter and collected in a brown test tube.
2.4 Detection and quantification of PSTs
Firstly, a chromatographic analysis was performed with a Nexera X2 HPLC system (Shimadzu, Kyoto, Japan) comprising a degassing unit, an auto-sampler and a micro binary pump and equipped with a Waters Amide column (2.1*100 mm,3.5 μm, Waters, USA). The temperature of the column was set at 40℃. The injection volume was 5 μL, and the flow rate was set to 0.5 mL/min in all time. The mobile phase consisted of two eluates: A (1 mmol/L ammonium formate solution) and B (acetonitrile), both of which contained 1.3 mmol/L formic acid. The elution was performed starting with 5% A for 2 min, increasing to 15%-30% for 16 min, and finally 5% A for 2.5 min for column re-equilibration. PSP toxins were detected by mass spectrometry analysis using a Shimadzu 8050 series mass spectrometer equipped with an electrospray interface set in the positive ionization mode (ESI+). PSP toxins (and their corresponding PubChem CID) included: neo-dc saxitoxin (dcNEO, 100962170), gonyautoxin 5 (GTX5, 49789073), neosaxitoxin (NEO, 135562690 ), dc saxitoxin (dcSTX ,101936522), saxitoxin (STX, 56947150). A negative ionization mode (ESI-) was used to detect: N-sulfocarbamoyl-gonyautoxin-2 and -3(C1,2, 49789085), and other gonyautoxins, including dcGTX2 (101034662), GTX1 (135061918), GTX2 (101650338), GTX4 (440699577), GTX3 (46217347), and dcGTX3(101034664). The multiple reaction monitoring (MRM) mode was used, with specific transition parameters as shown in Table 2. The capillary voltage was 4000 V, the nebulizer gas flow rate was 3 L/min, the drying gas flow rate was 10 L/min, and the temperature of the ion source was 300℃.The limits of detection (LOD, μg/kg) were dcNEO (24), GTX5 (60), NEO (30), dcSTX (30), STX (30), C1(40.8), C2 (24), dcGTX2 (42), GTX1 (38.4), GTX2 (28.8), GTX4 (48), GTX3 (48, dcGTX3(48).
2.5 Dietary exposure assessment
Dietary exposure to shellfish toxins was performed by multiplying shellfish PST value and amount of shellfish consumed divided by the target population mean body weight (WHO, 2009; Wong, 2013).
2.5.1 Consumption data
The consumption data and mean body weight of the target population were obtained from the individual-based Shenzhen Food Consumption Survey 2008 (SZFCS 2008), which was conducted by the Shenzhen Center for Disease Control and Prevention (SZCDC). A three-stage cluster sampling was performed taking account of local population flow and density, together with geographic and economic characteristics (Yang et al. 2014). Referring to the sampling principles of China National Health and Nutrition Survey (2002), 244 households were selected throughout 4 urban regions and 2 rural regions, a total of 853 individuals was included. Food consumption data were assessed by a continuous 3-day door-to-door interview, which included weighing that was consumed at home and using a 24-h dietary-recall questionnaire to record food intake outside of the home.
2.5.2 PSTs contamination data
Sample PST contamination data were assessed using liquid chromatography-tandem mass spectrometry (LC-MS/MS). Since the data were left-censored, the protocol for assigning concentration values to ND results was applied in this study. In a lower bound (LB) scenario, 0 was used to substitute for ND results (FAO/WHO,2009; EFSA,2010); while, in an upper bound (UB) scenario, 1/2 LOD was used to substitute for ND results.