Effects of Two Toxin-Producing Harmful Algae, Alexandrium catenella and Dinophysis acuminata (Dinophyceae), on Activity and Mortality of Larval Shellfish.

Harmful algal bloom (HAB) species Alexandrium catenella and Dinophysis acuminata are associated with paralytic shellfish poisoning (PSP) and diarrhetic shellfish poisoning (DSP) in humans, respectively. While PSP and DSP have been studied extensively, less is known about the effects of these HAB species or their associated toxins on shellfish. This study investigated A. catenella and D. acuminata toxicity in a larval oyster (Crassostrea virginica) bioassay. Larval activity and mortality were examined through 96-h laboratory exposures to live HAB cells (10-1000 cells/mL), cell lysates (1000 cells/mL equivalents), and purified toxins (10,000 cells/mL equivalents). Exposure to 1000 cells/mL live or lysed D. acuminata caused larval mortality (21.9 ± 7.0%, 10.2 ± 4.0%, respectively) while exposure to any tested cell concentration of live A. catenella, but not lysate, caused swimming arrest and/or mortality in >50% of larvae. Exposure to high concentrations of saxitoxin (STX) or okadaic acid (OA), toxins traditionally associated with PSP and DSP, respectively, had no effect on larval activity or mortality. In contrast, pectenotoxin-2 (PTX2) caused rapid larval mortality (49.6 ± 5.8% by 48 h) and completely immobilized larval oysters. The results indicate that the toxic effects of A. catenella and D. acuminata on shellfish are not linked to the primary toxins associated with PSP and DSP in humans, and that PTX2 is acutely toxic to larval oysters.

1. Introduction

Alexandrium catenella, associated with paralytic shellfish poisoning (PSP) and Dinophysis acuminata, associated with diarrhetic shellfish poisoning (DSP), are two common harmful algal bloom (HAB) species in the northeastern United States of America (USA). Regionally, cell concentrations of these two species are typically 1–100 cells/mL, but can exceed 1000 cells/mL, especially within subsurface thin layers and near bloom peaks [1,2,3]. Alexandrium catenella produces a suite of hydrophilic PSP toxins (PSTs), including saxitoxin (STX), while D. acuminata produces two classes of lipophilic toxins: the pectenotoxins (PTXs) and DSP toxins (DSTs)—consisting of okadaic acid (OA), dinophysistoxins (DTXs), and their derivatives.

These species and their toxins may also threaten aquatic life. In shellfish, the most common vector of PSP and DSP to humans, these HAB species have wide-ranging effects that are sometimes, but not exclusively, related to human syndrome toxins. For instance, PSTs impair burrowing activity in some clam populations (Mya arenaria [4]) and interfere with the immunocompetence of oyster hemocytes (Crassostrea gigas [5]). PST-producing Alexandrium cells or their exudates reduce swimming activity and survival in many larval shellfish species [6,7,8]. Similarly, cultures of A. catenella cause adductor-muscle paralysis in adult oysters (Crassostrea virginica [9]), larval inactivity and a decrease in egg viability in pearl oysters (Pinctada fucata martensii [10,11]), and aberrant or arrested development and reduced growth, settling rates, and survival in larval oysters (C. gigas [12]).

These cell and exudate effects are not always related to PST exposure. Non-PST bioactive compounds produced by Alexandrium species have been shown to reduce disease resistance in clams (Ruditapes philippinarum and M. arenaria [13]). Additionally, exposure of larval pearl oysters (P. fucata martensii) to non-PST-producing Alexandrium affine leads to larval inactivity [10]. Similarly, larval oysters (C. gigas) exhibit reduced swimming activity, feeding, growth, development, settlement yield, and malformations of the mantle on exposure to non-PST-producing Alexandrium minutum [14,15].

The effects on shellfish from Dinophysis species and/or associated toxins are comparable, though reports to date are more limited. Mussels (Mytilus edulis) exposed to OA exhibit reduced protein phosphatase activity and reduced larval viability [16,17], while PTX2 exposure to oysters (C. gigas) causes reduced fertilization success and larval mortality [18,19]. Likewise, exposure to D. acuminata causes oyster (C. gigas) oocyte mortality and reduced fertilization success, while gametes exposed to Dinophysis sacculus exhibit abnormal larval development [18].

Furthermore, co-occurrences and successive occurrences of A. catenella and D. acuminata blooms, as well as the co-occurrence of PSTs and DSTs, have been reported in Northport Bay, NY, USA [3]. These co-occurrences occurred at temperatures at which C. virginica in the region have been observed to spawn [20]. The potential combined effects of these HAB species or their toxins on bivalves have not been studied; however, a recent study that explored the individual and combined effects of A. catenella and D. acuminata on two larval, estuarine finfishes found reduced growth and swimming activity on exposure to either HAB species [21]. Lethality in the latter study was associated with exposure to A. catenella alone, with no additive effects through co-exposure with D. acuminata.

A goal of this study was to explore the effects of these HAB species and toxins—traditionally associated with human health issues—on shellfish health, and to assess whether effects were due to PSTs and/or DSTs, or other modes of toxicity. We assessed the individual and combined effects of A. catenella and D. acuminata and their toxins on the survival and swimming activity of oyster larvae (C. virginica). These animals were too small to consume A. catenella or D. acuminata directly, so all assays evaluated non-feeding interactions with live HAB cells, suspended particles, extracellular toxins and bioactive compounds, depending on the bioassay. Bioassays included 96-h exposures to: (1) A. catenella and D. acuminata live cells, administered individually; (2) A. catenella and D. acuminata lysates, administered individually and combined; and (3) pure toxins from A. catenella (STX) and D. acuminata (OA and PTX2), administered individually and combined (Table 1). Co-exposures were limited to toxin and lysate experiments to avoid complications that might have arisen from interactions between the two live HAB species.

Table 1

Treatments within the Alexandrium catenella and Dinophysis acuminata bioassays.

Bioassay	Treatments a	Cells/mL or Cells/mL Equivalent	Species
Live-Cell	Fed (Pav) b	25,000	Pavlova pinguis
	Unfed	0	None
	Acat 10	10	Alexandrium catenella
	Acat 100	100	Alexandrium catenella
	Acat 500	500	Alexandrium catenella
	Acat 1000	1000	Alexandrium catenella
	Dacum 10	10	Dinophysis acuminata
	Dacum 100	100	Dinophysis acuminata
	Dacum 500	500	Dinophysis acuminata
	Dacum 1000	1000	Dinophysis acuminata
Lysate	Fed (Pav) b	25,000	Pavlova pinguis
	Unfed	0	None
	Acat 100	100	Alexandrium catenella
	Acat 1000	1000	Alexandrium catenella
	Dacum 1000	1000	Dinophysis acuminata
	Acat 1000 × Dacum 1000	1000 *	Alexandrium catenella and Dinophysis acuminata
Pure Toxin	Carrier b	0	None
	OA	10,000	None
	PTX2	10,000	None
	STX	10,000	None
	OA × PTX2	10,000	None
	OA × STX	10,000 *	None
	PTX2 × STX	10,000 *	None
	OA × PTX2 × STX	10,000 *	None

a Pav = Pavlova pinguis, Acat = Alexandrium catenella, Dacum = Dinophysis acuminata, OA = okadaic acid, PTX2 = pectenotoxin-2, and STX = saxitoxin. b Control treatments within each bioassay. * Cells/mL equivalent of each algal species represented in the treatment, independently.

2. Results

2.1. Larval Inactivity and Mortality

2.1.1. Live-Cell Bioassay

Depression of larval swimming activity, i.e., inactivity including death, was observed in the higher cell concentration treatments of live-cell A. catenella and D. acuminata. By 96 h, all of the A. catenella treatments had significantly higher larval inactivity than the Fed and Unfed treatments (Fed: 11.6 ± 4.0%, Unfed: 5.7 ± 2.5%, Acat10: 51.9 ± 12.7%, Acat100: 90.1 ± 3.0%, Acat500: 92.8 ± 4.0%, Acat1000: 89.8 ± 3.4% standard error [SE], n = 10 wells per treatment, Tukey–Bonferroni: all p < 0.0125; Figure 1, Table S1). Additionally, larvae exposed to the Acat500 treatment exhibited moderate mortality (22.8 ± 8.7% SE; Figure 1), significantly higher than the Fed and Unfed treatments (Fed: 0.9 ± 0.9% SE, Unfed: 0%, Tukey–Bonferroni: all p < 0.0125), while the Acat10, Acat100, and Acat1000, treatments exhibited low mortality (Acat10: 5.7 ± 4.6%; Acat100: 9.2 ± 5.0% SE; Acat1000: 0%; Figure 1, Table S2). With D. acuminata, larval inactivity and mortality were only significant in the Dacum1000 treatment, with inactivity peaking at 24 h (49.2 ± 10.6% SE; Figure 1, Table S1). By 48 h, the Dacum1000 treatment had significantly higher mortality than the Fed and Unfed treatments, peaking at 96 h (21.9 ± 7.0% SE, Tukey–Bonferroni: all p < 0.0125; Figure 1, Table S2). Significant predictors in the models of larval inactivity and mortality included time, treatment, and their interaction (full model outputs in Tables S3 and S4, respectively). Cell concentrations of A. catenella and D. acuminata stayed the same or increased from 0 to 96 h across all treatments (Figure S1).

Figure 1

Percentage of inactive and dead larvae over time (h) in the live-cell bioassay, when larval oysters were exposed to Alexandrium catenella (Acat) or Dinophysis acuminata (Dacum), at four different initial cell concentrations. Error bars show the standard error (n = 10 wells per treatment). Data points that were significantly different from the Fed and Unfed treatments are denoted by an asterisk.

2.1.2. Lysate Bioassay

Effects on larval oysters from exposure to lysates were generally lower than from live-cell exposure (Figure 1 and Figure 2). Lysates of A. catenella had no significant effects while D. acuminata lysate exposure caused significant larval inactivity compared to Fed and Unfed treatments at all timepoints, peaking at 96 h (Dacum1000: 31.8 ± 4.8%, Fed: 2.0 ± 1.4%, Unfed: 2.9 ± 1.5% SE, Tukey–Bonferroni: all p < 0.0125; Figure 2, Table S5). Mixture of D. acuminata and A. catenella lysates moderated the effect of D. acuminata lysate exposure (Figure 2), but both the Acat1000 × Dacum1000 and Dacum1000 lysate treatments caused low to moderate mortality by 96 h, significantly higher than Fed or Unfed treatments (Acat1000 × Dacum1000: 6.2 ± 2.3%, Dacum1000: 10.2 ± 4.0%, Fed: 0%, Unfed: 2.0 ± 1.3% SE, Tukey–Bonferroni: all p < 0.0125). No other lysate treatments produced significant mortality in this set of experiments (Acat100: 0%, Acat1000: 0.9 ± 0.9% SE, Tukey–Bonferroni: all p > 0.0125; Figure 2, Table S6). Significant predictors in the model of larval inactivity included treatment and the interaction between treatment and time (full model output in Table S7). Significant predictors in the model of larval mortality included time, treatment, and their interaction (full model output in Table S8).

Figure 2

Percentage of inactive and dead larvae over time (h) in the lysate bioassay, when larval oysters were exposed to Alexandrium catenella (Acat) and Dinophysis acuminata (Dacum) lysate treatments. Numbers represent cell concentration equivalents (cells/mL equiv.). Error bars show the standard error (n = 10 wells per treatment). Data points that were significantly different from the Fed and Unfed treatments are denoted by an asterisk.

2.1.3. Pure Toxin Bioassay

Among the three pure toxins tested (STX, OA, and PTX2), only PTX2 caused severe inactivity and mortality. All larval oysters exposed to pure PTX2 were inactive within 24 h and experiments were therefore terminated after 48 h (i.e., 2 days earlier than the other single toxin treatments evaluated; Figure 3). All treatments that included PTX2 exhibited significantly higher mortality than the Carrier treatment at both 24 h (Kruskal–Wallis: χ2 = 47, df = 7, p < 0.0001, Dunn tests: all p < 0.05) and at 48 h (Kruskal–Wallis: χ2 = 71, df = 7, p < 0.0001, Dunn tests: all p < 0.05). Highest mortality was observed on exposure to all three toxins (OA × PTX2 × STX: 61.1 ± 4.9% SE) but PTX2 alone and other pairwise combinations produced similar levels of larval death (PTX2: 49.6 ± 5.8%, OA × PTX2: 50.0 ± 4.7%, PTX2 × STX: 36.5 ± 2.8% SE; Figure 3).

Figure 3

Percentage of inactive and dead larvae over time (h) in the pure toxin bioassay, when larval oysters were exposed to okadaic acid (OA), pectenotoxin-2 (PTX2), and saxitoxin (STX), alone or in combination at toxin concentrations equivalent to 10,000 cells/mL. Treatments PTX2, OA × PTX2, PTX2 × STX, and OA × PTX2 × STX were terminated at 48 h to collect toxin samples. Error bars show the standard error (n = 10 wells per treatment). Data points that were significantly different from the Carrier treatment are denoted by an asterisk.

In contrast, larval inactivity over the 96 h in the OA, STX, and OA × STX treatments was low (all < 6%; Figure 3), and not significantly different from the Carrier treatment (Carrier: 1.0 ± 1.0% SE, full model output in Table S9). Similarly, larvae from the OA, STX, and OA × STX treatments all exhibited very low mortality by 96 h (OA: 1.0 ± 1.0%, STX: 1.0 ± 1.0%, OA × STX: 2.0 ± 1.3%, Carrier: 1.0 ± 1.0% SE; Figure 3; full model output in Table S10).

2.2. Toxins in HAB Cultures, Water, and Oysters

Initial hatchery water used in the bioassays did not contain quantifiable toxin concentrations. Trace levels of PTX2, the DST dinophysistoxin-1 (DTX1), and the PSTs gonyautoxin 2 (GTX2) and decarbamoylneosaxitoxin (dcNEO) were detected but not quantifiable (

The toxin profile for the A. catenella culture was dominated by the STX congener C2, with 3.9 pg C2/cell (>90%); STX and GTX3 were also detected but were

The toxin profile for the D. acuminata culture was dominated by PTX2. Intracellular toxins included 8.0 pg PTX2/cell (91%), 0.4 pg OA/cell (5%), and 0.2 pg DTX1/cell (2%), while only PTX2 was quantifiable (0.2 pg PTX2/cell, 2%) in the extracellular fraction, with trace (

Larval oysters in all of the bioassays bioaccumulated PTX2 (Table 2). In the live-cell bioassay, oysters bioaccumulated increasing amounts of PTX2 with exposure to increasing D. acuminata concentration. Oysters exposed to the highest concentration of live or lysed D. acuminata, 1000 cells/mL or equivalent, accumulated similar PTX2 concentrations (5.2 and 3.7 pg PTX2/oyster, respectively). Oysters exposed to concentrations of pure PTX2, equivalent to 10,000 cells/mL, accumulated an order of magnitude more PTX2 (40.2–50.0 pg PTX2/oyster) than oysters exposed to 1000 D. acuminata cell equivalents/mL of live cells or lysate. PTX2 was below the detection limit (

Table 2

Oyster bioaccumulation of toxins associated with Dinophysis acuminata.

Bioassay	Treatments a	Time Collected (h)	Oysters per Sample b	Toxin (pg/oyster) c
				OA	DTX1	PTX2
N/A	Control	0	70 *	<DL	<LOQ	<DL
Live-Cell	Dacum 10	96	80	<DL	<LOQ	<LOQ
	Dacum 100	96	69	<DL	<DL	<LOQ
	Dacum 500	96	72	<LOQ	<DL	1.8
	Dacum 1000	96	70	<DL	<DL	5.2
Lysate	Dacum 1000	96	68	<LOQ	<LOQ	3.7
	Acat 1000 × Dacum 1000	96	65	<DL	<LOQ	2.5
Pure Toxin	PTX2	48	74	<DL	<DL	40.2
	OA × PTX2	48	70	<DL	<DL	45.3
	PTX2 × STX	48	71	<DL	<DL	50.0
	OA × PTX2 × STX	48	70	<DL	<DL	47.9

N/A = not applicable. Oysters were pooled into one sample per treatment and results were normalized to pg toxin/oyster. a Acat = Alexandrium catenella, Dacum = Dinophysis acuminata, DTX1 = dinophysistoxin-1, OA = okadaic acid, PTX2 = pectenotoxin-2, and STX = saxitoxin; numbers represent cell concentrations or cell concentration equivalents (cells/mL or cells/mL equiv.). b Direct counts of oysters were performed by microscope, unless otherwise noted. c

In the pure toxin bioassay, PTX2 in well water samples declined over 48 h, with or without oysters present, from 70.7 ± 14.5 standard deviation (SD) ng PTX2/mL at the start of treatment to 28.1 ± 3.7 SD ng PTX2/mL without oysters, and to 34.7 ± 5.9 SD ng PTX2/mL with oysters. Trace amounts of OA and DTX1 (

N/A = not applicable. Oysters were pooled into one sample per treatment and results were normalized to pg toxin/oyster. a Acat = Alexandrium catenella, Dacum = Dinophysis acuminata, DTX1 = dinophysistoxin-1, OA = okadaic acid, PTX2 = pectenotoxin-2, and STX = saxitoxin; numbers represent cell concentrations or cell concentration equivalents (cells/mL or cells/mL equiv.). b Direct counts of oysters were performed by microscope, unless otherwise noted. c