A new method for the simultaneous determination of cyanotoxins (Microcystins and Cylindrospermopsin) in mussels using SPE-UPLC-MS/MS

1 A new method for the simultaneous determination of cyanotoxins (Microcystins and Cylindrospermopsin) in mussels using SPE-UPLC-MS/MS Leticia Díez-Quijada Jiméneza (ldiezqui[email protected]), Remedios Guzmán-Guilléna, *([email protected]), Giorgiana M. Cătunescub ([email protected]), Alexandre Camposc ([email protected]), Vitor Vasconcelosc,d ([email protected]), Ángeles Josa ([email protected]), Ana M. Cameána (ca[email protected]). aArea of Toxicology, Faculty of Pharmacy, University of Sevilla, Sevilla, Spain. bUniversity of Agricultural Sciences and Veterinary Medicine Cluj-Napoca, Calea Mănăștur 3-5, 400372, Cluj- Napoca, Romania cInterdisciplinary Centre of Marine and Environmental Research, University of Porto, Matosinhos, Portugal dDepartment of Biology, Faculty of Science, University of Porto, Portugal *Corresponding author: Remedios Guzmán-Guillén Area of Toxicology, Faculty of Pharmacy, University of Sevilla.C/ Profesor García González, nº 2. C.P. 41012, Seville. Spain. E-mail address: [email protected] 2 Abstract The aim of this study was to optimize the extraction conditions of Microcystin-LR (MC-LR), Microcystin-RR (MC-RR), Microcystin-YR (MC-YR) and Cylindrospermopsin (CYN) simultaneously from mussels by using response surface methodology (RSM) and to validate the method by a dual solid phase extraction (SPE) system combined with ultra-performance liquid chromatography-tandem mass spectrometry (UPLC-MS/MS). The optimal parameters were: 90% MeOH (% v/v) for the extraction, a solvent/sample ratio of 75 and 15% MeOH in the extract before loading onto SPE. Mussels were spiked at 10; 37.5 and 75 ng g-1 fresh weight (f.w) of the 4 toxins, showing linear ranges of 0.5-75 ng g-1 f.w; low values for the limits of detection (0.01-0.39 ng g-1 f.w.) and quantification (0.23-0.40 ng g-1 f.w.); acceptable recoveries (70.37-114.03%) and relative standard deviation (%RSDIP) values (2.61-13.73%). The method was successfully applied to edible mussels exposed to cyanobacterial extracts under laboratory conditions, and it could allow the monitoring of these cyanotoxins in environmental mussel samples. Keywords: Cyanotoxins, Response surface methodology, Box-Behnken Design, method validation, solid phase extraction, mussels 3 Funding sources -Spanish Ministry of Economy and Competitiveness (AGL2015-64558-R, MINECO/FEDER, UE) and the grant FPI (BES-2016-078773). -EFSA funding under the EU-FORA Programme -Projects MOREBIVALVES (PTDC/ASP-PES/31762/2017) and UID/Multi/04423/2013 cofinanced by NORTE 2020, Portugal 2020 and the European Union through the ERDF, and by Portuguese Science Foundation (Fundação para a Ciência e a Tecnologia, FCT) 4 1 Introduction Nowadays, there is an increase in the cyanobacteria blooms over the world as a consequence of anthropogenic activities and climate change (Sanseverino et al., 2017). These cyanobacteria have the ability to produce toxic secondary metabolites called cyanotoxins (Funari et al., 2017). Among the variables involved in the prediction of the cyanotoxins presence, water temperature, turbidity, total phosphorus and alkalinity have an important role (García Nieto et al., 2013). Cyanotoxins are different in their chemical structures and toxicities, and they are classified as hepatotoxins, cytotoxins, dermatotoxins, neurotoxins, and irritant toxins, according to their mechanism of action (Testai et al., 2016). Microcystins (MCs) and cylindrospermopsin (CYN) are usually the most investigated cyanotoxins because of their wide distribution and toxicity. MCs can be produced by different genera such as Microcystis, Planktothrix and Dolichospermum (Buratti et al., 2017). MCs are cyclic heptapeptides with the general structure cyclo(D-Ala1-X2 -D-MeAsp3-Z4- Adda5 -D-Glu6 -Mdha7), where Adda is (2S,3S,8S,9S)-3- amino-9-methoxy-2,6,8-trimethyl-10-phenyldeca-4,6-dienoic acid, D-MeAsp3 is D-erythro-ß- methyl aspartic acid, Mdha is N-methyldehydroalanine; and X, Z are variable L-amino acids (Sivonen and Jones, 1999). So far, at least 246 MCs congeners have been recognized to date (Spoof and Catherine, 2017), with MC-LR being the most toxic (Díez-Quijada et al., 2019a). Among MCs, MC-LR, MC-RR and MC-YR are the most common variants, as a result of the presence of leucine (L), arginine (R) and tyrosine (Y), respectively, in the positions 2 (X) and 4 (Z) (Harada et al., 1990; Hooser et al., 1991; Sivonen and Jones, 1999). The mechanism of action of MCs is related to the inhibition of protein serine/threonine phosphatases through their binding mainly to PP1 and PP2A (Buratti et al., 2017; Bagu et al., 1997). 5 Many cyanobacterial genera have been identified as CYN producers such as Cylindrospermopsis, Aphanizomenon, Umezakia, Chrysosporum and Anabaena, however, the main producers are C. raciborskii and Aph. ovalisporum (Buratti et al., 2017; Kokocinski et al., 2017). CYN is a stable tricyclic alkaloid with a guanidine moiety connected to a hydroxymethyluracil through a hydroxyl bridge (Ohtani et al., 1992). CYN is highly water soluble because of its zwitterionic properties (Chiswell et al., 1999). It is well known that CYN produces an inhibition of protein (Terao et al., 1994) and glutathione (Runnegar et al., 1995) synthesis, and it needs to be activated metabolically by cytochrome P450, suggesting its progenotoxic activity (Norris et al., 2002; Štraser et al., 2011; Díez-Quijada et al., 2019b). Often an 80% of human exposure to cyanotoxins is thought to result from ingestion of contaminated drinking water (Ibelings and Chorus, 2007; Greer et al., 2018), followed by consumption of contaminated vegetables (Machado et al., 2017; Cao et al., 2019; Llana-Ruíz- Cabello et al., 2019) and edible aquatic food (Gutiérrez-Praena et al., 2013). Several studies reported MCs accumulation in mussels (Vasconcelos et al., 1995; Vareli et al., 2012; Baralla et al., 2017; Kim et al., 2017), but data on CYN are scarcer (Anderson et al., 2003; Saker et al., 2004). This accumulation represents a potential hazard for human health (Gutiérrez-Praena et al., 2013; Testai et al., 2016; Díez-Quijada et al., 2019a, 2019c). Provisional Tolerable Daily Intakes (TDI) of 0.04 and 0.03 µg/kg of body weight (b.w.)/day were established for MCs and CYN, respectively (Chorus and Bartram, 1999; Humpage and Falconer, 2003). Estuaries and coast environments can be affected by contaminated freshwaters (Tatters et al., 2017). These situations are very particular but frequent in certain regions, and are documented in the freshwater and saltwater invertebrates (Miller et al., 2010; Gibble et al., 2016; Peacock et al., 2018). Bivalves are particularly useful for assessing accumulation of cyanotoxins due to their widespread, being 6 prey of marine mammals and consumed by humans; they are important vectors of these toxins in the marine trophic chain. Therefore, it is important to establish methods for the determination and quantification of these toxins in these organisms to monitor the state of these ecosystems and the degree of contamination. Mass spectrometry (MS) combined with liquid chromatography (LC-MS, LC-MS/MS) is used to determine MCs and CYN in various matrices, such as water, biomass of natural blooms, cyanobacterial strains, mollusks and fish (Cameán et al., 2004; Moreno et al., 2005; Ruiz et al., 2005; Guzmán-Guillén et al., 2012a, 2012b, 2015; Zervou et al., 2017). In particular, ultraperformance liquid chromatography tandem mass spectrometry (UPLC-MS/MS) is a quicker and more selective analytical method that is employed to determine cyanotoxins in water (Oehrle et al., 2010; Greer et al., 2016; Pekar et al., 2016), fish (Guzmán-Guillén et al., 2017; Prieto et al., 2017; Manubolu et al., 2018) and mussels (Turner et al., 2018). The European Food Safety Authority (EFSA) has set as a priority the development of analytical validated methods to determine cyanotoxins in complex matrices (Testai et al., 2016), and in this sense, methods for multitoxin determination are of particular interest. Studies aiming at a simultaneous determination of cyanotoxins in mussels are limited and only detect MCs, but not CYN (Gkelis et al., 2006; Mekebri et al., 2009; Papadimitriou et al., 2012; Preece et al., 2015a, 2015b; Baralla et al., 2017; Kim et al., 2017; Turner et al., 2018). Moreover, only two of them have been validated (Mekebri et al., 2009; Turner et al., 2018), and although Turner et al. (2018) tried to detect CYN on their analyses, it was not possible. Recently, a method for the simultaneous determination of MCs and CYN in vegetables was developed (Díez-Quijada et al., 2018). 7 Response surface methodology (RSM) is a useful approach of process optimization considering the experimental design. It is an effective mathematical and statistical tool designed to model, analyze and optimize complex processes. It is usually used when a response of interest (called a dependent variable), such as compound extraction yields in this study, is influenced simultaneously by several factors (independent variables) (Addinsoft, 2019). It has different variations, however the Box-Behnken Design (BBD) is of current interest because it contains a smaller number of experiments with reliable results (Pal and Jadeja, 2019; Rodsamran and Sothornvit, 2019). In the BBD the points of experiments are generated on a sphere around the center point, the number of different factor levels is minimized, and the center point is repeated in order to maximize the prediction precision around the supposed optimum similar to the Central Composite Design (CCD) (Addinsoft, 2019). Its efficiency has been showed to be greater than that of Central Composite Design and much greater than that of Three-Level Full Factorial (Ferreira et al., 2007; Pal and Jadeja, 2019). In addition, the BBD generates less extreme experimental combinations compared to CCD, which uses variables that have simultaneous extreme value (at the corners of the hypercube) (Otto, 2016). Thus, BBD can provide a model with a better fit. Many studies dealing with the optimization of extraction procedures used the BBD (Anuar et al., 2013; Yılmaz et al., 2015; Oliveira et al., 2016; Lim et al., 2019). The main objective of this study was for the first time the optimization of a sensitive method for the simultaneous extraction and determination of four cyanotoxins (MC-LR, MC-RR, MC-YR and CYN) from mussels, using a dual solid phase extraction (SPE) cartridge system combined with UPLC-MS/MS for their determination and RSM for optimization procedures. 8 2 Materials and methods 2.1 Chemicals and Reagents Commercial standards for MC-LR, MC-RR and MC-YR (99% purity) and CYN (95% purity) were supplied by Enzo Life Sciences (Lausen, Switzerland). Deionized water (18.2 MΩ cm resistivity) was obtained from an ultrapure water purification system (NANOpure DiamondTM, Barnstead, USA). HPLC-grade methanol (MeOH), dichloromethane (DCM), hexane, formic acid (FA) and sodium hydroxide (NaOH) were supplied by Merck (Darmstadt, Germany). Cartridges for the SPE were C18 Bakerbond® (500 mg, 6 mL, Dicsa, Andalucía, España), and graphitized carbon cartridges BOND ELUT® (500 mg, 6 mL, Agilent Technologies, Amstelveen, The Netherlands, Europe). Nanosep® Centrifugal Devices with Bio- Inert® Membrane (modified nylon) 0.45 µm and 25 mm Syringe Filters with 0.2 µm cellulose acetate membrane were supplied by PALL Corporation (New York, USA) and VWR (International Radnor, Pennsylvania, USA), respectively. LC–MS grade reagents were employed for UHPLC–MS/MS analyses: water and acetonitrile, purchased from VWR International (Fontenay-sous-Bois, France), and FA from Fluka (Steinheim, Germany). Standard multitoxin solutions containing the four cyanotoxins (MC-LR, MC-RR, MC-YR and CYN) were prepared when needed in 20% MeOH and diluted to three different concentrations (20, 75 and 150 ng mL- 1) to obtain 10, 37.5 and 75 ng g-1 f.w., respectively, by spiking 1 g of mussels f.w. with 500 µL of the standard solutions. Mussels were obtained from a local supermarket, ready for human consumption. 9 2.2 Optimization of the extraction method 2.2.1 Preliminary extraction tests Preliminary tests were performed on mussels spiked with 250 ng g-1 f.w. by applying the method of Díez-Quijada et al. (2018) for multitoxin analysis in vegetables with some modifications: the extraction procedure was repeated twice; extraction with or without hexane (10 mL) to eliminate the lipids; MeOH concentration was reduced to 25% before passing the sample through the cartridges in the SPE (Mekrebi et al., 2009; El Ghazali et al., 2010; Greer et al., 2016); and the extracts were cleaned by 2 options: using SPE with the dual cartridge system (C18 and porous graphitic carbon –PGC-) or by filtering (nanosep devices -0.45 µm- and syringe filters - 0.2 µm-) (Turner et al., 2018). The use of hexane in general yielded the poorest results (5-8%), while applying only SPE instead of filtering always resulted in higher recovery yields; indeed, the best recoveries were obtained when using SPE without hexane (46-66%) (data not shown). Taking this into account, we decided to continue with the optimization of the extraction method without hexane, using SPE, and reducing the percentage of MeOH to at least 25% before SPE. 2.2.2 Response surface methodology – Box-Behnken Design A BBD experimental design of 15 experimental runs was used to analyze the optimal extraction conditions for MC-LR (Y1), MC-RR (Y2), MC-YR (Y3) and CYN (Y4) from mussels spiked with a solution mixture. Three independent variables were considered at 3 levels: MeOH concentration (% v/v) (X1) of the solvent used for extraction (80%, 90%, 100%), solvent/sample ratio (X2) (25, 50, 75) and MeOH concentration of the extract for loading onto SPE (% v/v) (X3) (5%, 15%, 25%). These 3 factors were chosen because they were the ones with higher heterogeneity among the scientific literature regarding cyanotoxins extraction. In the case of the 16 Figure 1. Response linearity in mussel (Huber plot) for (a) MC-LR; (b) MC-RR; (c) MC-YR and (d) CYN. 3.1.2 Sensitivity LOD and LOQ are presented in Table 1, and both values were lower than the reported guideline values (GV) for waters of 1 µg L-1 proposed for MCs (WHO, 2003) and for CYN (Humpage and Falconer, 2003). Although these are different matrices, the values are compared to those GV in waters because this is the media where cyanotoxins reach the mussels. The lowest LOD was obtained for MC-RR, followed by MC-YR, MC-LR and finally CYN. The results obtained are lower than the LOD and LOQ values reported by other authors in mussels (Baralla et al., 2017; Turner et al., 2018), fish (Manubolu et al., 2018), fish and crayfish (Ríos et al., 2013), other aquatic organisms (Gkelis et al., 2006), and vegetables (Li et al., 2014; Díez- Quijada et al., 2018). Moreover, their study presents two limitations; one concerning quantification below these reported values, and the second related to the applicability of the 245 250 255 260 265 270 275 280 0 20 40 60 80 Signal/Concentration CYN concentration (µg L-1) Target Line Action Line 105% Action Line 95% 290 295 300 305 310 315 320 325 330 335 0 20 40 60 80 Signal/Concentration MC-LR concentration (µg L-1) Target Line Action Line 105% Action Line 95% a 440 460 480 500 520 540 560 580 600 0 20 40 60 80 Signal/Concentration MC-RR concentration (µg L-1) Target Line Action Line 105% Action Line 95% b d 190 195 200 205 210 215 220 225 0 20 40 60 80 Signal/Concentration MC-YR concentration (µg L-1) Target Line Action Line 105% Action Line 95% c 17 method for the extraction of other MC variants. Turner et al. (2018) reported higher or similar LOD and LOQ for some MCs congeners in mussels, but they do not include CYN on this analysis. Our study allows for the first time the analysis of MC congeners and CYN simultaneously, showing adequate and lower LOD and LOQ values for all of them compared to the existing values reported by other methods in mussels. 18 Table 1. Estimations of recoveries (experimental, predicted and error (%)); within-day repeatability (Sw); between-day repeatability (SB); intermediate precision (SIP); and its relative standard deviation (%RSDIP) for MC-LR, MC-RR, MC-YR and CYN. Limits of detection (LOD) and quantification (LOQ) for the mussel matrix. Validation parameters Toxin concentration level (ng g-1 f.w.a) Experimental Recovery (%) Predicted Recovery (%) % Error of prediction SW SB SIP RSDIP (%) LOD (ng g-1 f.w.a) LOQ (ng g-1 f.w.a) MC-LR 10.0 113.46 108.07 5.39 1.09 0.22 0.90 7.95 0.056 0.185 37.5 87.30 85.20 2.10 1.87 1.04 1.64 5.02 75.0 73.31 72.96 0.35 2.29 1.84 2.15 3.91 MC-RR 10.0 70.37 70.84 -0.47 0.39 0.28 0.35 5.04 0.048 0.161 37.5 74.93 74.93 0 2.77 4.68 3.52 12.54 75.0 88.41 87.05 1.36 4.91 11.26 7.64 11.52 MC-YR 10.0 114.03 109.01 5.02 0.28 0.34 0.30 2.61 0.063 0.209 37.5 83.93 83.03 0.90 2.82 2.18 2.62 8.34 75.0 80.76 80.29 0.47 5.03 9.38 6.80 11.22 CYN 10.0 111.78 110.22 1.56 1.47 1.65 1.53 13.73 0.037 0.123 37.5 85.18 85.56 -0.38 3.03 2.40 2.84 8.88 75.0 90.75 90.72 0.03 3.13 2.19 2.85 4.19 a f.w.: fresh weight. Acceptable ranges for RSDAOAC (%): 16% for 10 ng g-1 f.w. and 8-16% for 37.5 and 75 ng g-1 f.w. Acceptable ranges for RecoveryAOAC (%): 60- 115 % for all concentrations. 19 3.1.3. Matrix Effects The matrix effect could affect significantly the quantification, depending on different factors such as the composition and amount of matrix and the properties of the analyte (Li et al., 2014). The matrix effect was determined in mussels with the following expression: ME = 100 x [(slope matrix/slope solvent)-1], expressed in percentage. The values obtained were 28.28 for MC-LR, - 17.77 for MC-RR, 18.97 for MC-YR and -11.26 for CYN, meaning mild matrix effect (|10| < ME < |20|) for MC-RR, –YR and CYN, and medium (|20| < ME < |50|) in the case of MC-LR. No strong matrix effects were observed in the present study. However, mild decreases in the responses of the cyanotoxins were detected: a decrease in the signal of MC-RR, similarly to Moreno et al. (2005) in fish, and in the response of CYN in agreement with Díez-Quijada et al. (2018) in vegetables. Moreover, a slight increase was observed in the response of MC-YR, in concordance with Ríos et al. (2013) in fish and crayfish. These effects could affect the ionization efficiency and weaken the analyte response. Therefore, matrix-matched calibration curves should be employed for quantification in order to obtaining more accurate results. 3.2 Optimization procedures The yields obtained from the experimental runs were analyzed by RSM and the predicted values are shown in Table 2. The Mann-Whitney two-tailed test showed non-significant differences between the experimental and predicted values at a significance level of 0.001. The multiple regression analysis of the RSM generated a system of 4 quadratic polynomial equations that proved to be good statistical models for the recovery yield of the toxins from a combined mixture (Table 3). They all had high coefficient of determination values (R2) accounting for above 85% of the variability of the response (yields of the 4 toxins). 20 Table 2. Recovery yield of the cyanotoxins at different extraction conditions based on a Box-Behnken design for response surface methodology (RSM). Sort order Extraction conditions - Independent variables Response - dependent variables X1 methanol concentration (% v/v) X2 Samplesolvent ratio (w/v) X3 methanol concentration before SPE (% v/v) Y1 MC-LR recovery yield (%) Y2 MC-RR recovery yield (%) Y3 MC-YR recovery yield (%) Y4 CYN recovery yield (%) experimental predicted* experimental predicted* experimental predicted* experimental predicted* 1 80 25 15 60.66 58.98 70.30 70.886 74.03 72.099 100.49 95.22 2 90 50 15 78.49 74.39 91.91 83.637 82.53 80.623 102.37 96.35 3 100 50 25 69.32 69.40 65.61 66.734 74.98 75.610 95.78 88.06 4 80 50 25 52.85 48.13 56.13 49.114 63.94 61.038 65.11 66.52 5 100 25 15 91.29 84.82 85.00 77.446 92.99 87.526 96.84 100.69 6 80 50 5 45.94 45.86 40.83 39.706 57.94 57.310 45.10 52.82 7 90 25 25 71.21 77.61 77.29 83.720 77.27 82.104 96.51 100.38 8 90 50 15 68.59 74.39 81.74 83.637 76.15 80.623 93.75 96.35 9 90 25 5 69.97 71.72 80.77 81.308 75.12 77.681 89.33 86.89 10 90 75 25 78.81 77.06 96.02 95.483 85.91 83.349 92.06 94.50 11 90 50 15 76.10 74.39 77.26 83.637 83.19 80.623 92.94 96.35 12 100 75 15 83.57 85.25 98.89 98.304 88.46 90.391 119.82 125.10 13 90 75 5 86.73 80.34 90.54 84.110 86.93 82.096 98.88 95.01 14 80 75 15 60.15 66.62 57.04 64.594 69.43 74.894 76.91 73.06 15 100 50 5 64.34 69.06 55.34 62.356 70.76 73.663 90.19 88.78 p-value** 0.935 0.967 1.000 0.775 *the predicted value resulted from the model optimizing the simultaneous extraction of the 4 toxins. **Mann-Whitney two-tailed test (α = 0.001) of the experimental data versus the yields predicted by the model optimizing the simultaneous extraction of the 4 toxins. 21 The differences between root mean square error (RMSE) and predicted residual error sum of squares (Press's) RMSE were small, showing that the model was not sensitive to the presence or absence of specific observations in the model. Additionally, the Q2 (1- (Press's RMSE)/(sum of the squares of the errors) for the model) values were close to 1, suggesting that the model is good and robust. The statistical models were better for MC-RR and CYN recovery which had R2 values above 90% and both reached statistical significance. Table 3. Model parameters (coded coefficients) and goodness of fit statistics obtained by response surface methodology (RSM) for each of the four response variables (Yi). Y1 MC-LR recovery Y2 MC-RR recovery Y3 MC-YR recovery Y4 CYN recovery Coefficient intercept b0 74.39 83.64 80.62 96.35 linear b1 11.12*** 10.07** 7.73** 14.38*** b2 2.02 3.64 1.42 0.56 b3 0.65 3.45 1.42 3.25 quadratic b11 -9.52* -18.75** -6.90* -8.99* b22 9.05* 12.92** 7.50* 11.16** b33 -6.76* -10.41* -6.82* -13.31** interaction b12 -1.80 6.79 0.02 11.64** b13 -0.48 -1.26 -0.45 -3.61 b23 -2.29 2.24 -0.79 -3.50 R² 0.87 0.90 0.86 0.93 p-value 0.075* 0.047** 0.090* 0.022** RMSE 20.94 28.54 16.00 28.93 Press RMSE 27.09 31.74 21.83 28.73 Q² -0.67 -0.24 -0.86 0.01 * Significant at P < 0.15, **Significant at P < 0.05, *** Significant at P < 0.01. The MeOH concentration (% v/v) of the solvent for extraction had the greatest statistically significant influence upon the recovery yields of the 4 toxins. The positive linear coefficients showed that a high concentration of MeOH seems to generally increase recovery 22 yields, especially for the most hydrophobic MCs. However, the negative quadratic coefficients suggest that there is a maximum limit of MeOH concentration for optimal recovery, as seen in the plots obtained by RSM, and this could be the case for the more hydrophilic MC-RR and especially CYN (Figure 2). A significant positive interaction was found between the MeOH concentrations for extraction and the MeOH percentage of the extract for loading onto SPE only in the case of CYN. That means that the chosen concentration of MeOH for the extraction will influence the optimal %MeOH for loading onto SPE. Figure 2. Effect of the MeOH concentration (% v/v) of the solvent for the extraction on the recovery yield of cyanotoxins from complex mixtures.. y = -0.09x2+ 17.63x - 761.57 50 60 70 80 90 100 110 80 85 90 95 100 CYN recovery yield (%) methanol concentration (%) y = -0.10x2+ 18.25x - 796.90 50 60 70 80 90 100 110 80 85 90 95 100 MC-LR recovery yield (%) methanol concentration (%) y = -0.19x2+ 34.76x - 1,525.99 50 60 70 80 90 100 110 80 85 90 95 100 MC-RR recovery yield (%) methanol concentration (%) y = -0.07x2+ 13.19x - 547.79 50 60 70 80 90 100 110 80 85 90 95 100 MC-YR recovery yield (%) methanol concentration (%) 23 Thus, the present model shows that the maximum yields should be obtained at the following concentrations: 91.25% for MC-LR; 91.47% for MC-RR; 94.21% for MC-YR and 97.94% for CYN, when considering a MeOH concentration between 80% and 100%. This is in agreement with other authors who reported optimal recoveries when using 75-90% MeOH for MCs extraction from fish or mussel tissues (Karlsson et al., 2005; Moreno et al., 2005; Mekebri et al., 2009; Smith and Boyer, 2009; Manubolu et al., 2018; Turner et al., 2018). According to Turner et al. (2018), lower recoveries were observed by the use of concentrations below 70% MeOH. A higher solvent/sample ratio in the extraction seems to positively influence the recovery yields of the 4 toxins, but the linear regression coefficients did not reach statistical significance (Table 3). The positive influence of the quadratic coefficients was statistically confirmed, thus explaining the presence of the linear coefficients in the hierarchical model (Table 3). However, the quadratic regression showed the presence of 2 possible maximum optimal ratios depending on the position of the minimum with the tested 25-75 solvent to sample ratio interval (Figure 3). The tested solvent to sample ratio interval 25 to 75 was selected as following the preliminary tests and as indicated by the scientific literature in either fish or mollusks (Guzmán-Guillén et al., 2011, 2015; Freitas et al., 2014, 2016). In addition, a higher ratio would imply a higher consumption of extraction solvent, thus a wasteful approach as the solvent is evaporated almost entirely prior to analysis. Moreover, a lower ratio would make more difficult the manipulation and further extraction of the 4 cyanotoxins. 24 Figure 3. Effect of solvent/sample ratio in the extraction on the recovery yield of cyanotoxins from complex mixtures. Thus, when determining the maximum yields of the 4 toxins from this model the 75 emerges as the optimal solvent/sample ratios within the range of 25 to 75. The positive linear regression coefficients (Table 3) appeared to indicate that a higher MeOH concentration of the extract for loading onto SPE slightly increased the recovery yield of the 4 toxins. But, the higher quadratic coefficients, statistically significant, showed that there might be a maximum optimal concentration, as seen in the plots obtained by RSM (Figure 4). Although their significance was not confirmed statistically, the linear coefficients were y = 0.02x2- 1.76x + 139.85 50 60 70 80 90 100 110 25 35 45 55 65 75 CYN recovery yield (%) solvent to sample ratio y = 0.01x2- 1.37x + 106.54 50 60 70 80 90 100 110 25 35 45 55 65 75 MC-LR recovery yield (%) solvent to sample ratio y = 0.02x2- 1.92x + 128.05 50 60 70 80 90 100 110 25 35 45 55 65 75 MC-RR recovery yield (%) solvent to sample ratio y = 0.01x2- 1.14x + 107.81 50 60 70 80 90 100 110 25 35 45 55 65 75 MC-YR recovery yield (%) solvent to sample ratio 25 maintained within the model because of its hierarchical nature. This effect was consistent to that observed for the MeOH concentration used for loading onto SPE (Greer et al., 2016). Figure 4. Effect of MeOH concentration for loading onto the solid phase extraction (SPE) (% v/v) on the recovery yield of cyanotoxins from complex mixtures. In conclusion, our model suggests that the maximum yields might be obtained at the following concentrations of MeOH in the extracts for loading onto SPE: 14.92% for MC-LR; 17.35% for MC-RR; 15.64% for MC-YR and 16.61% for CYN, when considering a concentration interval of 5% to 25%. This is consistent with the desirability the plots obtained by RSM (Figure 5). y = -0.13x2+ 4.32x + 61.53 50 60 70 80 90 100 110 510 15 20 25 CYN recovery yield (%) methanol concentration (%) y = -0.07x2+ 2.09x + 58.21 50 60 70 80 90 100 110 510 15 20 25 MC-LR recovery yield (%) methanol concentration (%) y = -0.10x2+ 3.47x + 55.05 50 60 70 80 90 100 110 510 15 20 25 MC-RR recovery yield (%) methanol concentration (%) y = -0.07x2+ 2.19x + 63.15 50 60 70 80 90 100 110 510 15 20 25 MC-YR recovery yield (%) methanol concentration (%) 32 Diez-Quijada, L., Puerto, M., Gutiérrez-Praena, D., Llana-Ruiz-Cabello, M., Jos, A., Cameán, A.M., 2019a. Microcystin-RR: Occurrence, content in water and food and toxicological studies. A review. Environ. Res. 168, 467–489. https://doi.org/10.1016/j.envres.2018.07.019. Diez-Quijada, L., Llana-Ruiz-Cabello, M., Cătunescu, M.G., Puerto, M., Moyano, R., Jos, A., Cameán, A.N., 2019b. In vivo genotoxicity evaluation of cylindrospermopsin in rats using a combined micronucleus and comet assay. Food Chem. Toxicol. 132, 110664. https://doi.org/10.1016/j.fct.2019.110664. Diez-Quijada, L., Prieto, A.I., Guzmán-Guillén, R., Jos, A., Cameán, A.M., 2019c. Occurrence and toxicity of microcystin congeners other than MC-LR and MC-RR: A review. Food Chem. Toxicol. 125, 106–132. https://doi.org/10.1016/j.fct.2018.12.042. El Ghazali, I., Saqrane, S., Carvalho, A.P., Ouahid, Y., Del Campo, F.F., Oudra, B., Vasconcelos, V., 2010. Effect of different microcystin profiles on toxin bioaccumulation in common carp (Cyprinus carpio) larvaevia Artemia nauplii. Ecotoxicol. Environ. Saf. 73, 762- 770. https://doi.org/10.1016/j.ecoenv.2009.12.015. Ferreira, S.L.C., Bruns, R.E., Ferreira, H.S., Matos, G.D., David, J.M., Brandão, G. C., Da Silva, E.G.P., Portugal, L.A., Dos Reis, P.S., Souza, A.S., Dos Santos, W. N. L., 2007. Box-Behnken design: An alternative for the optimization of analytical methods. Anal. Chim. Acta 597, 179- 186. https://doi.org/10.1016/j.aca.2007.07.011. Freitas, M., Azevedo, J., Carvalho, A.P., Campos, A., Vasconcelos, V., 2014. Effects of storage, processing and proteolytic digestion on microcystin-LR concentration in edible clams. Food Chem. Toxicol. 66, 217-223. https://doi.org/10.1016/j.fct.2014.01.041. Freitas, M., Azevedo, J., Carvalho, A.P., Mendes, V.M., Manadas, B., Campos, A., Vasconcelos, V., 2016. Bioaccessibility and changes on cylindrospermopsin concentration in edible mussels with storage and processing time. Food Control 59, 567–574. https://doi.org/10.1016/j.foodcont.2015.06.025. 33 Funari, E., Manganelli, M., Buratti, F.M., Testai, E., 2017. Cyanobacteria blooms in water: Italian guidelines to assess and manage the risk associated to bathing and recreational activities. Sci. Total Environ. 598, 867–880. http://dx.doi.org/10.1016/j.scitotenv.2017.03.232. García, A.C., Bargu, S., Dash, P., Rabalais, N.N., Sutor, M., Morrison, W., Walker, N.D., 2010. Evaluating the potential risk of microcystins to blue crab (Callinectes sapidus) fisheries and human health in a eutrophic estuary. http://dx.doi.org/10.1016/j.fct.2013.03.025doi:10.1016/j.hal.2009.08.011. García Nieto, P.J., Alonso Fernández, J.R., de Cos Juez, F.J., Sánchez Lasheras, F., Díaz Muñiz, C., 2013. Hybrid modelling based on support vector regression with genetic algorithms in forecasting the cyanotoxins presence in the Trasona reservoir (Northern Spain). Environ. Res. 122, 1-10. http://dx.doi.org/10.1016/j.envres.2013.01.001. Gibble, C. M., Peacock, M. B., Kudela, R. M., 2016. Evidence of freshwater algal toxins in marine shellfish: Implications for human and aquatic health. Harmful Algae 59, 59–66. https://doi.org/10.1016/j.hal.2016.09.007. Gkelis, S., Lanaras, T., Sivonen, K., 2006. The presence of microcystins and other cyanobacterial bioactive peptides in aquatic fauna collected from Greek freshwaters. Aquat. Toxicol. 78, 32–41. https://doi.org/10.1016/j.aquatox.2006.02.001. González, A.G., Herrador, M.A., 2007. A practical guide to analytical method validation, including measurement uncertainty and accuracy profiles. Trends Anal. Chem. 26, 227–238. https://doi.org/10.1016/j.trac.2007.01.009. González, A.G., Herrador, M.A., Asuero, A.G., 2010. Intra-laboratory assessment of method accuracy (trueness and precision) by using validation standards. Talanta 82, 1995–1998. https://doi.org/10.1016/j.talanta.2010.07.071. 34 Greer, B., McNamee, S.E., Boots, B., Cimarelli, L., Guillebault, D., Helmi, K., Marcheggiani, S., Panaiotov, S., Breitenbach, U., Akçaalan, R., Medlin, L.K., Kittler, K., Elliott, C.T., Campbell, K., 2016. A validated UPLC-MS/MS method for the surveillance of ten aquatic biotoxins in European brackish and freshwater systems. Harmful Algae 55, 31–40. https://doi.org/10.1016/j.hal.2016.01.006. Greer, B., Meneely, J.P., Elliott, C.T., 2018. Uptake and accumulation of Microcystin-LR based on exposure through drinking water: An animal model assessing the human health risk. Sci. Rep. 8, 4913. DOI: 10.1038/s41598-018-23312-7. Gutiérrez-Praena, D., Jos, A., Pichardo, S., Moreno, I.M., Cameán, A.M., 2013. Presence and bioaccumulation of microcystins and cylindrospermopsin in food and the effectiveness of some cooking techniques at decreasing their concentrations: a review. Food Chem. Toxicol. 53, 139- 152. https://doi.org/10.1016/j.fct.2012.10.062. Guzmán-Guillén, R., Prieto, A. I., Moreno, I., Soria, M. E., Cameán, A. M., 2011. Effects of thermal treatments during cooking, microwave oven and boiling, on the unconjugated microcystin concentration in muscle of fish (Oreochromis niloticus). Food Chem. Toxicol. 49, 2060-2067. https://doi.org/10.1016/j.fct.2011.05.018. Guzmán-Guillén, R., Prieto, A.I., Gónzalez, A.G., Soria-Díaz, M.E., Cameán, A.M., 2012a. Cylindrospermopsin determination in water by LC-MS/MS: Optimization and validation of the method and application to real samples. Environ. Toxicol. Chem. 31, 2233–2238. https://doi.org/ 10.1002/etc.1954. Guzmán-Guillén, R., Prieto Ortega, A.I., Moreno, I., González, G., Soria-Díaz, M.E., Vasconcelos, V., Cameán, A.M., 2012b. Development and optimization of a method for the determination of Cylindrospermopsin from strains of Aphanizomenon cultures: Intra-laboratory assessment of its accuracy by using validation standards. Talanta 100, 356–363. https://doi.org/10.1016/j.talanta.2012.07.087. 35 Guzmán-Guillén, R., Moreno, I.M., Prieto, A.I., Soria-Díaz, M.E., Vasconcelos, V.M., Cameán, A.M., 2015. CYN determination in tissues from freshwater fish by LC–MS/MS: Validation and application in tissues from subchronically exposed tilapia (Oreochromis niloticus). Talanta 131, 452–459. https://doi.org/10.1016/j.talanta.2014.07.091. Guzmán-Guillén, R., Maisanaba, S., Prieto Ortega, A.I., Valderrama-Fernández, R., Jos, A., Cameán, A.M., 2017. Changes on cylindrospermopsin concentration and characterization of decomposition products in fish muscle (Oreochromis niloticus) by boiling and steaming. Food Control 77, 210–220. https://doi.org/10.1016/j.foodcont.2017.02.035. Harada, K., Matsuura, K., Suzuki, M., Watanabe, M.F., Oishi, S., Dahlem, A.M., Beasley, V.R,, Carmichael, W.W., 1990. Isolation and characterization of the minor components associated with microcystins LR and RR in the cyanobacterium (blue-green algae). Toxicon 28(1), 55–64. https://doi.org/10.1016/0041-0101(90)90006-S. Hooser, S. B., Kuhlenschmidt, M. S., Dahlem, A. M., Beasley, V. R., Carmichael, W. W., Haschek, W. M., 1991. Uptake and subcellular localization of tritiated dihydro-microcystin-LR in rat liver. Toxicon 29(6), 589-601. DOI: 10.1016/0041-0101(91)90053-t. Huber, L., 1998. Validation and Qualification in Analytical Laboratories; Interpharm: East Englewood, CO, USA, pp. 1–288. ISBN 9780849382673. Humpage, A.R., Falconer, I.R., 2003. Oral toxicity of the cyanobacterial toxin cylindrospermopsin in male Swiss albino mice: Determination of no observed adverse effect level for deriving a drinking water guideline value. Environ. Toxicol. 18, 94–103. https://doi.org/10.1002/tox.10104. Ibelings, B.W., Chorus, I., 2007. Accumulation of cyanobacterial toxins in freshwater ‘‘seafood’’ and its consequences for public health: a review. Environ. Pollut. 150, 177–192. https://doi.org/10.1016/j.envpol.2007.04.012. 36 ICH Harmonised Tripartite Guideline, 2005. Validation of analytical procedures: Text and methodology Q2(R1). ICH Working Group. Available online: http://www.ich.org/fileadmin/Public_Web_Site/ICH_Products/Guidelines/Quality/Q2_R1/Step4/ Q2_R1__Guideline.pdf. (accessed on 10 February 2020). Karlsson, K.M., Spoof, L.E.M., Meriluoto, J.A.O., 2005. Quantitative LC-ESI-MS analyses of microcystins and Nodularin-R in animal tissue – matrix effects and method validation. Environ. Toxicol. 20, 381–389. https://doi.org/10.1002/tox.20115. Kim, M.S., Lee, Y., Ha, S.Y., Kim, B.H., Hwang, S.J., Kwon, J.T., Choi, J.W., Shin, K.H., 2017. Accumulation of microcystin (LR, RR and YR) in three freshwater bivalves in Microcystis aeruginosa bloom using dual isotope tracer. Mar. Drugs 15, 226–240. https://doi.org/10.3390/md15070226. Kokocinski, M., Cameán, A.M., Carmeli, S., Guzmán-Guillén, R., Jos, A., Mankiewicz- Boczek, J., Metcalf, J.S., Moreno, I., Prieto, A.I., Sukenik, A., 2017. Chapter 12: Cylindrospermopsin and congeners. In: Meriluoto, J., Spoof, L., Codd, G.A. (Eds.), Handbook of Cyanobacterial Monitoring and Cyanotoxin Analysis, 1st ed. Wiley, NJ, USA, pp. 127–137. Li, Y.W., Zhan, X.J., Xiang, L., Deng, Z.S., Huang, B.H., Wen, H.F., Sun, T.F., Cai, Q.Y., Li, H. Mo, C.H., 2014. Analysis of trace microcystins in vegetables using solid-phase extraction followed by high performance liquid chromatography triple-quadrupole mass spectrometry detection. J. Agric. Food Chem. 62, 11831–11839. https://doi.org/10.1021/jf5033075. Lim, J.W., Shin, S.M., Jung, S.J., Lee, M.k., Kang, S.Y., 2019. Optimization of antibacterial extract from lacquer tree (Rhus verniciflua Stokes) using response surface methodology and its efficacy in controlling edwardsiellosis of olive flounder (Paralichthys olivaceus). Aquaculture 502, 40-47. https://doi.org/10.1016/j.aquaculture.2018.12.008. 37 Llana-Ruiz-Cabello, M., Jos, Á., Cameán, A.M., Oliveira, F., Barreiro, A., Machado, J., Azevedo, J., Pinto, E., Almeida, A., Campos, A., Vasconcelos, V., Freitas, M., 2019. Analysis of the Use of Cylindrospermopsin and/or Microcystin-Contaminated Water in the Growth, Mineral Content, and Contamination of Spinacia oleracea and Lactuca sativa. Toxins 11, 624-644. https://doi.org/10.3390/toxins11110624. Machado, J., Campos, A., Vasconcelos, V., Freitas, M., 2017. Effects of microcystin-LR and cylindrospermopsin on plant-soil systems: A review of their relevance for agricultural plant quality and public health. Environ. Res. 153, 191-204. https://doi.org/10.1016/j.envres.2016.09.015. Magnusson, B., Örnemark, U., (eds.) Eurachem Guide: The fitness for purpose of analytical methods – A laboratory guide to method validation and related topics, (2nd ed. 2014). ISBN 978- 91-87461-59-0. Available from www.eurachem.org. Manubolu, M., Lee, J., Riedl, K.M., Kua, Z.X., Collart, L.P., Collart, S.P., Ludsin, S.A., 2018. Optimization of extraction methods for quantification of microcystin-LR and microcystin-RR in fish, vegetables, and soil matrices using UPLC-MS/MS. Harmul Algae 2018, 76, 47–57. https://doi.org/10.1016/j.hal.2018.04.009. Mekebri, A., Blondina, G.J., Crane, D.B., 2009. Method validation of microcystins in water and tissue by enhanced liquid chromatography tandem mass spectrometry. J. Chromatogr. A 1216, 3147–3155. https://doi.org/10.1016/j.chroma.2009.01.095. Miller, M. A., Kudela, R.M., Mekebri, A., Crane, D., Oates, S.C., Tinker, M.T., Staedler, M., Miller, W.A., Toy-Choutka, S., Dominik, C., Hardin, D., Langlois, G., Murray, M., Ward, K., Jessup, D.A., 2010. Evidence for a novel marine harmful algal bloom: Cyanotoxin (microcystin) transfer from land to sea otters. PLOS One 5, 1–11. https://doi.org/10.1371/journal.pone.0012576. 38 Miller, N.J., Miller, J.C. (Eds.), 2000. Statistics and Chemometrics for Analytical Chemistry. Fourth Edition. Pearson Education Limited. ISBN: 0-13-022888-5. Moreno, I.M., Molina, R., Jos, Á., Picó, Y., Cameán, A.M., 2005. Determination of microcystins in fish by solvent extraction and liquid chromatography. J. Chromatogr. A 1080, 199–203. https://doi.org/10.1016/j.chroma.2005.05.029. Norris, R.L.G., Seawright, A.A., Shaw, G.R., Senogles, P., Eaglesham, G.K., Smith, M.J., Chiswell, R.K., Moore, M.R., 2002. Hepatic xenobiotic metabolism of cylindrospermopsin in vivo in the mouse. Toxicon 40, 471–476. https://doi.org/10.1016/S0041-0101(01)00243-4. Oehrle, S.A., Southwell, B., Westrick, J., 2010. Detection of various freshwater cyanobacterial toxins using ultra-performance liquid chromatography tandem mass spectrometry. Toxicon 55, 965–972. https://doi.org/10.1016/j.toxicon.2009.10.001. Oliveira, G. D. A., De Oliveira, A. E., Da Conceição, E. C., Leles, M. I., 2016. Multiresponse optimization of an extraction procedure of carnosol and rosmarinic and carnosic acids from rosemary. Food Chem. 211, 465-473. doi:10.1016/j.foodchem.2016.05.042. Ohtani, I., Moore, R.E., Runnegar, M.T.C., 1992. Cylindrospermopsin, a potent hepatotoxin from the blue-green alga Cylindrospermopsis raciborskii. J. Am. Chem. Soc. 114, 7941–7942. https://doi.org/10.1021/ja00046a067. Otto, M., (Ed.). 2016. Chemometrics: statistics and computer application in analytical chemistry. In: John Wiley & Sons. (3rd ed). ISBN: 978-3-527-34097-2 Pal, C. B. T., Jadeja, G. C., 2019. Microwave-assisted deep eutectic solvent extraction of phenolic antioxidants from onion (Allium cepa L.) peel: a Box–Behnken design approach for optimization. J.Food Sci. Technol. 56, 4211-4223. https://doi.org/10.1007/s13197-019-03891-7. 39 Papadimitriou, T., Kagalou, I., Stalikas, C., Pilidis, G., Leonardos, I.D., 2012. Assessment of microcystin distribution and biomagnification in tissues of aquatic food web compartments from a shallow lake and evaluation of potential risks to public health. Ecotoxicology 21, 1155–1166. http://doi.org/10.1007/s10646-012-0870-y. Peacock, M. B., Gibble, C. M., Senn, D. B., Cloern, J. E., Kudela, R. M., 2018. Blurred lines: Multiple freshwater and marine algal toxins at the land-sea interface of San Francisco Bay, California. Harmful Algae 73, 138–147. https://doi.org/10.1016/j.hal.2018.02.005. Pekar, H., Westerberg, E., Bruno, O., Lääne, A., Persson, K.M., Sundström, L.F., Thim, A.M., 2016. Fast, rugged and sensitive ultra high pressure liquid chromatography tandem mass spectrometry method for analysis of cyanotoxins in raw water and drinking water- First findings of anatoxins, cylindrospermopsins and microcystin variants in Swedish source waters and infiltration ponds. J. Chromatogr. A 1429, 265–276. https://doi.org/10.1016/j.chroma.2015.12.049. Preece, E.P., Moore, B.C., Hardy, F.J., Deobald, L.A., 2015a. First detection of microcystin in Puget Sound, Washington, mussels (Mytilus trossulus), Lake Reserv. Manage. 31, 50-54. https://doi.org/10.1080/10402381.2014.998398. Preece, E.P., Moore, B.C., Hardy, F.J., 2015b. Transfer of microcystin from freshwater lakes to Puget Sound, WA and toxin accumulation in marine mussels (Mytilus trossulus). Ecotoxicol. Environ. Saf. 122, 98–105. http://dx.doi.org/10.1016/j.ecoenv.2015.07.013. Prieto, A.I., Guzmán-Guillén, R., Valderrama-Fernández, R., Jos, A., Cameán, A.M., 2017. Influence of Cooking (Microwaving and Broiling) on Cylindrospermopsin Concentration in Muscle of Nile Tilapia (Oreochromis niloticus) and Characterization of Decomposition Products. Toxins 9, 177–189. https://doi.org/10.3390/toxins9060177. Puerto, M., Campos, A., Prieto, A.I., Cameán, A.M., Martinho de Almeida, A., Coelho, A.V., Vasconcelos, V., 2011. Differential protein expression in two bivalve species; Mytilus 40 Galloprovinciales and Corbicula Fluminea; exposed to Cylindrospermopsis Raciborskii Cells. Aquat. Toxicol. 101, 109-116. https://doi.org/10.1016/j.aquatox.2010.09.009. Ríos, V., Moreno, I., Prieto, A.I., Puerto, M., Gutiérrez-Praena, D., Soria-Díaz, M.E., Cameán, A.M., 2013. Analysis of MC-LR and MC-RR in tissue from freshwater fish (Tinca tinca) and crayfish (Procambarus clarkii) in tench ponds (Cáceres, Spain) by liquid chromatography–mass spectrometry (LC–MS). Food Chem. Toxicol. 57, 170-178. Doi: http://dx.doi.org/10.1016/j.fct.2013.03.025. Rodsamran, P., Sothornvit, R., 2019. Extraction of phenolic compounds from lime peel waste using ultrasonic-assisted and microwave-assisted extractions. Food Biosci. 28, 66-73. https://doi.org/10.1016/j.fbio.2019.01.017. Ruiz, M.J., Cameán, A.M., Moreno, I.M., Picó, Y., 2005. Determination of microcystins in biological samples by matrix solid-phase dispersion and liquid chromatography-mass spectrometry. J. Chromatogr. A 1073, 257–262. https://doi.org/10.1016/j.chroma.2004.08.128. Runnegar, M.T., Kong, S.M., Zhong, Y.Z., Lu, S.C., 1995. Inhibition of reduced glutathione synthesis by cyanobacterial alkaloid cylindrospermopsin in cultured rat hepatocytes. Biochem. Pharmacol. 49, 219–225. https://doi.org/10.1016/S0006-2952(94)00466-8. Saker, M.L., Metcalf, J.S., Codd, G.A., Vasconcelos, V.M., 2004. Accumulation and depuration of the cyanobacterial toxin cylindrospermopsin in the freshwater mussel Anodonta cygnea. Toxicon 43, 185–194. https://doi.org/10.1016/j.toxicon.2003.11.022. Sanseverino, I., Conduto, D., Loos, R., Lettieri, T., 2017. Cyanotoxins: methods and approaches for their analysis and detection. JRC Technical Reports. EUR-28624. https://doi.org/10.2760/36186. Sivonen, K., Jones, G., 1999. Cyanobacterial toxins. In: Chorus, I., Bartram, J. (Eds.), Toxic Cyanobacteria in Water: A Guide to Their Public Health Consequences, Monitoring, and 41 Management. E and FN Spon, London/New York, 41-111. http://apps.who.int/iris/bitstream/10665/42827/1/0419239308_eng.pdf. Smith, J.L., Boyer, G.L., 2009. Standardization of microcystin extraction from fish tissues: a novel internal standard as a surrogate for polar and non-polar variants. Toxicon 53, 238-245. https://doi.org/10.1016/j.toxicon.2008.11.007. Spoof, L., Catherine, A., 2017. Appendix 3. Tables of microcystins and nodularins. In: Meriluoto, J., Spoof, L., Codd, J. (Eds.), Handbook of cyanobacterial monitoring and cyanotoxin analysis. Wiley, Hoboken, 526–538. https://doi.org/10.1002/9781119068761. Štraser, A., Filipič, M., Žegura, B., 2011. Genotoxic effects of the cyanobacterial hepatotoxin cylindrospermopsin in the HepG2 cell line. Arch. Toxicol. 85, 1617–1626. https://doi.org/10.1007/s00204-011-0716-z. Tatters, A. O., Meredith, D.A.H., Nagoda, C., Busse, L., Gellene, A.G., Caron, D.A., 2017. Multiple stressors at the land-sea interface: Cyanotoxins at the land-sea interface in the Southern California Bight. Toxins 9, 95-111. doi:10.3390/toxins9030095. Terao, K., Ohmori, S., Igarashi, K., Ohtani, I., Watanabe, M.F., Harada, K.I., Ito, E., Watanabe, M., 1994. Electron microscopic studies on experimental poisoning in mice induced by cylindrospermopsin isolated from blue-green alga Umezakia natans. Toxicon 32, 833–843. https://doi.org/10.1016/0041-0101(94)90008-6. Testai, E., Buratti, F.M., Funari, E., Manganelli, M., Vichi, S., Arnich, N., Biré, R., Fessard, V., Sialehaamoa, A., 2016. Review and analysis of occurrence, exposure and toxicity of cyanobacteria toxins in food. EFSA supporting publication EN-998. 1-309. https://doi.org/10.2903/sp.efsa.2016.EN-998. Turner, A.D., Waack, J., Lewis, A., Edwards, C., Lawton, L., 2018. Development and singlelaboratory validation of a UHPLC-MS/MS method for quantitation of microcystins and