Paralytic Shellfish Toxins of Pyrodinium bahamense (Dinophyceae) in the Southeastern Gulf of Mexico

In September and November 2016, eight marine sampling sites along the coast of the southeastern Gulf of Mexico were monitored for the presence of lipophilic and hydrophilic toxins. Water temperature, salinity, hydrogen potential, dissolved oxygen saturation, inorganic nutrients and phytoplankton abundance were also determined. Two samples filtered through glass fiber filters were used for the extraction and analysis of paralytic shellfish toxins (PSTs) by lateral flow immunochromatography (IFL), HPLC with post-column oxidation and fluorescent detection (FLD) and UHPLC coupled to tandem mass spectrometry (UHPLC-MS/MS). Elevated nutrient contents were associated with the sites of rainwater discharge or those near anthropogenic activities. A predominance of the dinoflagellate Pyrodinium bahamense was found with abundances of up to 104 cells L−1. Identification of the dinoflagellate was corroborated by light and scanning electron microscopy. Samples for toxins were positive by IFL, and the analogs NeoSTX and STX were identified and quantified by HPLC-FLD and UHPLC-MS/MS, with a total PST concentration of 6.5 pg cell−1. This study is the first report that confirms the presence of PSTs in P. bahamense in Mexican waters of the Gulf of Mexico.


Introduction
The dinoflagellate Pyrodinium bahamense Plate 1906 (Dinoflagellata: Dinophyceae: Goniodomaceae) is a tropical-subtropical species that was originally described from the Atlantic Ocean at New Providence Island (Bahamas) [1]. The first record of paralytic shellfish poisoning (PSP) caused by P. bahamense was recorded in 1972 in Southeast Asia at Port Moresby, Papua New Guinea [2,3]. It is one of the main dinoflagellate species producing paralytic shellfish toxins (PSTs) that have caused PSP from the consumption of bivalves, planktivorous fish and gastropods in the world [4][5][6][7][8], mainly in tropical and subtropical areas [7,[9][10][11]. Their proliferations have also been associated with epizootics evaluate physicochemical factors that may be associated with its proliferation and confirm the presence of PSTs in this dinoflagellate in the southeastern Gulf of Mexico.

Physicochemical Variables
The variation of the water temperature (Table 1) was low in both months; in November, at station 8, it reached a maximum of 38.63 • C. Salinity showed high variation both between stations and between months, with a minimum value of 17 at station 2 in November and a maximum value of 41.57 at station 4 in September. The pH showed a maximum value of 9.63 at station 7 in September and a minimum value of 6.5 at station 6.3 in November. Dissolved oxygen (DO) showed a minimum value of 4.1 mg L −1 at station 7 in November. In September, nitrites at all stations were above the maximum limit established for marine waters (MLEMW) [56]. As in November, a maximum value of 0.045 µmol L −1 was recorded at station 2. In September and November, nitrates were above the MLEMW. At stations 1, 2, 4 and 8, the ammonium concentrations were above the MLEMW. At all stations, orthophosphates were well below the MLEMW (Table 1).

Identification and Abundance of Pyrodinium bahamense
The cell width was measured above or below the cingulum, not taking into account the cingular lists: length without spines, including the apical horn, 45  1 -4 , the apical plates; 1 -6 , the precingular plates; 1 -5 , the postcingular plates; 1 and 2 , the antapical plates; aa-the anterior attachment pore, aas-the antapical spine, ah-the apical horn, as-the apical spine, cl-the cingular lists, cp-the canopy (also known as the cover plate or the closing plate), pa-the posterior attachment pore, Sp-the posterior sulcal plate, tp-the trichocyst pores, vp-the ventral pore. The plates are named mainly according to Balech [33]. Scale bars: 10 µm in (A-C,E,F), 1 µm in (D).  In September, P. bahamense showed a maximum abundance of 3.13 × 10 4 cells L −1 at station 7, followed by 1.90 × 10 4 cells L −1 at station 2. At the other stations, the abundances ranged from undetected to 10 3 cells L −1 in November ( Figure 3).

Lateral Flow Immunochromatographic Test (Rapid PSP Test)
Positive identification of PSTs by qualitative analysis (saxitoxin and analogs) for lateral flow immunochromatography (LFIC) was found in phytoplankton extracts. The intensity of the marked line was similar to a positive control (STX, FDA Reference Standard STD, was obtained from the US National Institute of Standards and Technology (NIST, RM 8642).

HPLC-FLD Analysis
The toxin profile determined by post-column oxidation HPLC-FLD of P. bahamense in comparison to certified reference standards showed the presence of two analogs from the group of carbamoyl toxins (STX and NeoSTX; Figure 4). No sulfocarbamoyl and decarbamoyl toxins were detected in this study. The toxin content was quantified at 6.5 pg cell −1 (against external calibration standards. The molar percentages recorded for P. bahamense showed high molar percentages of carbamoyl toxins with 86.7% and 13.3% of STX and NeoSTX, respectively. Figure 5 illustrates the MRM chromatograms obtained following HILIC-UHPLC-MS/MS analysis of STX and NeoSTX standards and the sample of P. bahamense. Peaks were observed for both STX and NeoSTX in the algal sample, with retention times identical to those determined in the calibration standard. Ion ratios determined between standard and sample were found to be similar (±10% deviation).

Discussion
The physicochemical variables (temperature, salinity, pH and DO) measured at the end of the rainy season and at the beginning of the windy season coincided with those reported by Poot-Delgado et al. [62], who concluded that their variation was influenced by the shallow depth in the study area (~1 m), recording higher values compared to the area more than 2.5 km offshore [63][64][65][66]. Similarly, the registered values of nitrogenous compounds are above the MLEMW, which could be influenced by the proximity of the study area to point sources of wastewater discharges resulting from various anthropogenic activities [67,68].
Phlips et al. [69] described, as physiological limits for P. bahamense in Florida, a temperature of 20 • C as the lower temperature limit and a tolerance to salinity between 10 and 45. The bloom potential of this dinoflagellate was most closely associated with shallow ecosystems with long water residence times, and peak biomass levels were correlated to nutrient concentrations in regions of high abundance.
In the present study, we followed [33], who found no significant morphological differences between the two varieties, var. bahamense and var. compressum [37]. Mertens [11] suggested that instead, they are probably different life stages and that P. bahamense is a species complex represented by two clearly separated ribotypes, Indo-Pacific and Atlantic-Caribbean ones.
Cusick and Duran [75] reported the existence of genotypes in natural P. bahamense subpopulations in Florida and Puerto Rico, with relevant insights on underlying genetic factors influencing the potential for toxin variability among natural subpopulations of HAB species.
Studies on bivalves contaminated by P. bahamense in Maylasia and the Phiippines showed that there were few differences in toxin profiles of dinoflagellate and shellfish [7,82,83].
Chromatographic profiles have been determined by HPLC-FLD and UHPLC-MS/MS from shellfish sampled during the main bloom of P. bahamense, which had the greatest impact on public health in Mexico [14,55]. A concentration of 7396 µg eq STX kg −1 was determined in the rock oyster Crassostrea iridescens (Hanley) from the Bay of Acapulco, Guerrero, during the 1995 HAB event; between 8889.3 and 32,272.5 µg eq STX kg −1 occurred in the clam Donax gracilis Hanley from Zapotal and 14,445 µg eq STX kg −1 in the mangrove mussels Mytella strigata (Hanley) from Puerto Madero, both in Chiapas, during the HAB of 2001 (regulatory limit for PSTs in flesh is 800 µg saxitoxin equivalents (STX eq.) per kg of shellfish flesh). The profiles by HPLC-FLD were STX, GTX2, dcSTX, dcGTX2, dcGTX3 and B1 in C. iridescens, STX, NeoSTX, GTX2, GTX3 and B1 in D. gracilis and STX, GTX2, GTX3 and B1 in M. strigata. In Salina Cruz, Oaxaca, in 2002, 306.3 µg eq STX kg −1 was still detected in the spiny oyster Spondylus calcifer Carpenter, with only STX present. Additionally, UHPLC-MS/MS analysis was used to detect GTX1, GTX5, GTX6, doSTX, dcNEO and C1 [14,55]. The two PST detection methods in this study are both thoroughly validated and have been found to generally compare well when applied to shellfish samples. A recent study assessing the use of six detection methods for analysis of shellfish from Latin America, including Mexico, showed good correlations between all instrumental methods applied.
The HPLC-FLD method has been previously validated for a wide range of PST analogs, although the LC-MS/MS has the further advantage of incorporating a larger number overall. Both methods are, however, well suited to the accurate quantitation of individual PST analogs and the summation of PSTs to estimate total toxin levels. Puffer fish (Tetraodontidae) from southeastern Mexico can also be vectors of PSTs associated with HABs of P. bahamense and could be the cause of SPFP in human cases [44,55] reported in the Yucatan Peninsula [85]. The profiles of sodium channel blocker neurotoxins (TTX and analogs + STX and analogs) in five species of puffer fish of the genus Sphoeoroides and Lagocephalus were recently evaluated along the central coast of Campeche, are attributed to P. bahamense, and were composed of PSTs (STX, NeoSTX, GTX2, GTX3, dcGTX2 and dcGTX3 [42] and TTX analogs. The presence of TTX and STX analogs was also confirmed by UHPLC-MS/MS, detecting TTX, 4-epiTTX, MonodeoxyTTX, 4,9-anhydroTTX and 11-norTTX-6(R)-ol as well as STX and dcSTX [54,55].
This was observed by Núñez-Vázquez et al. [42,54] indirectly in puffer fish, probably due to the toxins transmitted via bivalves and other invertebrates that regularly make up part of the diet of these fish in Campeche coastal waters [44]. Previous studies by [22,86] demonstrated the presence of high concentrations of PSTs in puffer fish from Florida, resulting from the P bahamense HAB event.
PSP has been the main public health problem due to bivalves in Mexico [14,28,55]. The HABs of P. bahamense have caused 92.78% of PSP cases, with 476 human intoxications, all occurring along the southeastern Mexican Pacific coast [55].
Studies that confirm the presence of PSTs in natural samples are scarce, and few reports are available that have confirmed toxin-producing microalgae, either from environmental samples or laboratory cultures (Table 3) worldwide with this species. This study, in addition to generating data on the microalgal toxin profile, has allowed for comparison with those toxin profiles from other regions of the world and from other species of dinoflagellates (e.g., Alexandrium spp. and Gymnodinium catenatum) that produce PSTs. Pyrodinium bahamense is one of the harmful microalga that has caused the highest number of human poisonings and deaths from PSP at global and regional levels (8). To the best of our knowledge, this is only the second time that the production of PSTs by P. bahamense has been confirmed in the Atlantic of the American continent, and worldwide, it is one of the few cases of toxins confirmation in this species. As such, this study contributes to the evaluation of possible risks to public and animal health in the area, as well as socioeconomic impacts related to fisheries and aquaculture industries. Additionally, the knowledge of the toxin profile of P. bahamense has allowed for confirmation of the possible etiology of these toxins that can be transferred to other trophic levels, including for example, puffer fish that are consumed and have caused human poisonings in this region.

Conclusions
The occurrence of the gonyaulacoid dinoflagellate P. bahamense in the southeastern Gulf of Mexico using light microscopy and SEM was related to elevated nutrient content at the sites of rainwater discharge or those near anthropogenic activities. STX and NeoSTX were detected by chromatographic techniques (IFL, HPLC-FLD and UHPLC-MS/MS). This study is the first report that confirms the presence of PST in P. bahamense in Mexican waters of the Gulf of Mexico.

Seawater Samples
In September and November 2016, seawater samples were taken at eight stations (site depth ca. 1 m) located along the central coast of the state of Campeche. All sampling sites were influenced by rainwater discharge or were near the sites with anthropogenic activities (Figure 6).

Phytoplankton Sampling and Nutrient Determination
At each site, surface seawater samples were collected with a 1 L plastic bottle; a 100 mL aliquot was used to determine cell abundances of phytoplankton taxa. Immediately after collection, samples were fixed with a neutral Lugol solution and subsequently preserved by adding 37% neutralized formalin to a final concentration of 4% [87]. Cells were counted after sedimentation in a 10 mL cylinder following Reguera et al. [88]. Additionally, horizontal tows were taken with a 20 µm mesh conical hand net. The material collected for the P. bahamense analysis was placed in glass vials and fixed using the same procedure as for the cell counting. Net samples were also examined under a light microscope for accurate identification of P. bahamense. In situ water temperature ( • C), salinity, pH and DO were measured using a using a Hanna multiparameter probe, model HI9828, equipped with a sensor, model HI769828, and a Hach multiparameter probe, model HQ40d (Hanna Instruments Inc., Woonsocket, RI, USA). Orthophosphate, ammonium, nitrite, nitrate and silicate analyses were performed in the laboratory following [89].

Identification of Pyrodinium bahamense
Cell measurements (n = 50) were made using the 40×/0.65 objective and an Olympus BX51 compound microscope. Four samples were examined in a JEOL JSM-7600F scanning electron microscope (SEM) at a working distance of 15 to 21 mm and a voltage of 1.2 to 5.0 kV after a preliminary wash in distilled water, followed by dehydration in a series of ethanol solutions of increasing concentration (30,50,70,90 and 100%), air drying on 0.5 aluminum mounts and sputter coating with gold-palladium using a Polaron SC7640 High Resolution Sputter Coater (Quorum Technologies, Newhaven, UK).

Lateral Flow Immunochromatography Test (Rapid PSP Test)
Analysis for paralytic shellfish toxins (saxitoxin and analogs) were conducted by Lateral flow immunochromatography (LFIC) using rapid test kits provided by Scotia Rapid Testing (Scotia Rapid Testing Ltd., Nova Scotia, Canada), a qualitative lateral flow test for detection of PST in shellfish and phytoplankton for PSP (SRT PSP). The development and validation of the SRT PSP kit using the standard protocol is described by Scotia test [92].

HPLC-FLD Analysis
Phytoplankton samples from station 7 were harvested by filtration through glass fibers (GF/F Whatman) and immediately frozen at −20 • C. These filters were used for toxin analysis. PSTs were analyzed by high-performance liquid chromatography and fluorescence detection (HPLC-FLD) [77] modified by [93]. PSTs were extracted by adding 2 mL of acetic acid (0.03 N) to each P. bahamense sample containing the filter and sonicated three times for 5 min each. The extract was centrifuged for 15 min at 14,000 rpm (HERMLE Z 216 microcentrifuge, Labortechnik GmbH, Wasserburg, Germany), and the supernatant was filtered through 0.45 µm filters using syringe filters of glass fiber (PVDF Millex membrane, 25 mm diameter). For the derivatization, 150 µL of each extract was mixed with 37 µL of hydrochloric acid 1 M and heated for 15 min at 90 • C. Cool to room temperature, after adding 75 µL of 1 N sodium acetate. The extracts were injected in the liquid chromatograph in separate runs for the identification and quantification of the PSTs (standards of saxitoxin (STX), neosaxitoxin (NeoSTX), goniautoxin-1,4 (GTX 1,4), decarbamoylsaxytoxin (dcSTX), decarbamoylgoniautoxin-2,3 (dc GTX 2, 3), and N-sulfocarbamoyl-11-hydrosulfate (C1,2).
The chromatography system used was an HP 1100 (Agilent Technologies, Santa Clara, CA, USA) consisting of an autosampler, degasser, quaternary pump, two binary pumps used for post-column reactions, a fluorescence detector, a C-18 column, and a post-column reactor. The PSTs were detected using an excitation wavelength of 333 nm and an emission wavelength of 390 nm. Two criteria were used for the identification of the PSTs: the retention times and co-elution with commercial standards of (STX), (NeoSTX) (National Research Council Canada, Institute for Marine Biosciences, Certified Reference Material Program, Halifax, NS, Canada).

UPLC-MS/MS
A Waters (Manchester, UK) Acquity UPLC I-Class coupled to a Waters Xevo TQ-S tandem quadrupole mass spectrometer (MS/MS) was used for UHPLC-MS/MS analysis. A Waters 1.7 µm, 2.1 × 150 mm Waters Acquity BEH Amide UPLC column held at +60 • C was used for chromatographic separation of toxin analogs in conjunction with a Waters VanGuard BEH Amide guard cartridge. Cleaned-up sample extracts were injected onto the UHPLC with an injection volume of 2 µL. Mobile phases and UHPLC gradient conditions were exactly as described by [52]. The MS/MS tune parameters were as follows: 150 • C source temperature, 600 • C desolvation temperature, 1000 L h −1 desolvation gas flow, 7.0 Bar nebulizer gas flow, 150 L h −1 cone gas flow and 0.15 mL min −1 collision gas flow. Capillary voltage was held at 0.5 and 2.5 kV for positive and negative ionization modes, respectively. Multiple Reaction Monitoring (MRM) transitions were exactly those described by [52,94]. Quantitation of toxins was performed against toxin standards available as certified reference standards. Five additional analogs (C3, C4, dcGTX1, dcGTX4 and GTX6) were incorporated into the method, with quantitation performed using the calibrations generated from their nearest structural analog, using experimentally determined relative response factors [51]). Toxicity equivalence factors (TEFs) were taken from EFSA recommendations where feasible [94], with others used as described by [94]. The method has been validated in shellfish internationally through collaborative study with well defined method performance characteristics [53]. Assessment of the method when applied to water or algae samples has previously demonstrated the absence of any significant matrix effects.