Distribution of Domoic Acid in the Digestive Gland of the King Scallop Pecten maximus

The king scallop Pecten maximus retains the amnesic shellfish poisoning toxin, domoic acid (DA), for a long time. Most of the toxin is accumulated in the digestive gland, but this organ contains several cell types whose contribution to the accumulation of the toxin is unknown. Determining the time-course of the depuration by analyzing whole organs is difficult because the inter-individual variability is high. A sampling method, using biopsies of the digestive gland, has been developed. This method allows for repetitive sampling of the same scallop, but the representativeness of the samples obtained in this way needs to be validated. In this work, we found that the distribution of DA in the digestive gland of the scallops is mostly homogeneous. Only the area closest to the gonad, and especially its outer portion, had a lower concentration than the other ones, probably due to a transfer of the toxin to the intestinal loop. Samples obtained by biopsies can therefore be considered to be representative. Most of the toxin was accumulated in large cells (mostly digestive cells), which could be due to differences during the toxin absorption or to the preferential depuration of the toxin from the small cells (mostly secretory).


Introduction
In 1987 in Prince Edward Island, Canada, four people died and nearly 100 became intoxicated by the consumption of blue mussels Mytilus edulis. The main symptoms of this intoxication were digestive problems and short-term memory loss, which led to the syndrome being named Amnesic Shellfish Poisoning [1][2][3]. Domoic acid (DA), a tricarboxylic amino acid previously isolated from the red alga Chondria armata [4], was identified as the responsible agent [5]. The diatom Nitzschia pungens f. multiseries (currently Pseudo-nitzschia multiseries) was later identified as the producing organism in the area [6]. Since then, this toxin has been recorded in many locations around the world including Europe [7][8][9][10][11], North America [12][13][14][15][16][17], South America [18], Oceania [19], Asia [20][21][22][23][24][25][26] and Africa [27][28][29]. DA, beside its effect on humans, also has severe effects on other marine mammals, the DA is distributed among both of the different cell types of the tissues involved. The digestive gland is constituted by several cell types whose functions are in some cases under discussion [62]. The two main types are absorptive (digestive) or secretory (basophile) cells. The first ones are in charge of the absorption of particles and their intracellular digestion and are larger than the secretory cells. Secretory cells produce enzymes and secrete them to the digestive lumen [62,63]. It could, therefore, be expected that the larger absorptive cells contain more DA than the smaller secretory ones, unless the transfer from absorptive to secretory cells is high.
In this study the distribution of the DA among different regions (Figure 1), sectors and cellular types of the digestive gland of Pecten maximus was determined to (1) evaluate the heterogeneity of the toxin concentration in the digestive gland, to contribute to the understanding and minimization of the variations observed in repeated samples of this organ obtained using biopsies, which could be a non-destructive way of monitoring the depuration process, and (2) gain some information about the actual depuration routes or mechanisms. discussion [62]. The two main types are absorptive (digestive) or secretory (basophile) cells. The first ones are in charge of the absorption of particles and their intracellular digestion and are larger than the secretory cells. Secretory cells produce enzymes and secrete them to the digestive lumen [62,63]. It could, therefore, be expected that the larger absorptive cells contain more DA than the smaller secretory ones, unless the transfer from absorptive to secretory cells is high. In this study the distribution of the DA among different regions (Figure 1), sectors and cellular types of the digestive gland of Pecten maximus was determined to (1) evaluate the heterogeneity of the toxin concentration in the digestive gland, to contribute to the understanding and minimization of the variations observed in repeated samples of this organ obtained using biopsies, which could be a non-destructive way of monitoring the depuration process, and (2) gain some information about the actual depuration routes or mechanisms.

Variability of Toxin Distribution in Different Parts of the Digestive Gland
The variation coefficients estimated from all sectors and regions of each scallop were lowmoderate, between 8.5% and 27.1%.  Table S1). Sector 1 had the highest relative concentration and sector 4 the lowest. On average, the difference between these two sectors was approximately 20% of the mean. When the effect of the distance to the gonad ( Figure 2B) and the adductor muscle ( Figure 2C) was examined, the toxin concentration was found to be lower as sector 4 was closer to the gonad (p = 0.0385, by a two way ANOVA, Supplementary Materials Table  S2), with the range also being approx. 20%. The toxin concentration was lower in the sectors closer to the adductor muscle (3 and 4), but the difference was not statistically significant (p = 0.2746 by a twoway ANOVA, Supplementary Materials Table S2).

Variability of Toxin Distribution in Different Parts of the Digestive Gland
The variation coefficients estimated from all sectors and regions of each scallop were low-moderate, between 8.5% and 27.1%.
Domoic acid was not homogeneously distributed among the four sectors of the digestive gland ( Figure 2A, p = 0.0159 by a two-way ANOVA, Supplementary Materials Table S1). Sector 1 had the highest relative concentration and sector 4 the lowest. On average, the difference between these two sectors was approximately 20% of the mean. When the effect of the distance to the gonad ( Figure 2B) and the adductor muscle ( Figure 2C) was examined, the toxin concentration was found to be lower as sector 4 was closer to the gonad (p = 0.0385, by a two way ANOVA, Supplementary Materials Table S2), with the range also being approx. 20%. The toxin concentration was lower in the sectors closer to the adductor muscle (3 and 4), but the difference was not statistically significant (p = 0.2746 by a two-way ANOVA, Supplementary Materials Table S2). The toxin was, in general, homogeneously distributed between the internal and external regions of the digestive gland ( Figure 3A, p = 0.8921 in a 2-way ANOVA on the relative concentration, Supplementary Materials Table S1.). The variability, nevertheless, was higher in the outer than in the inner parts ( Figure 3A) In two sectors (2 and 4) the concentration in these two parts was different (lower in the outer part in sector 4, and the opposite in sector 2) ( Figure 3B), which made the interaction between sector and part (inner-outer) near to the statistical significance (p = 0.08, a 2-way ANOVA on the relative concentration, Supplementary Materials Table S1).  The toxin was, in general, homogeneously distributed between the internal and external regions of the digestive gland ( Figure 3A, p = 0.8921 in a 2-way ANOVA on the relative concentration, Supplementary Materials Table S1.). The variability, nevertheless, was higher in the outer than in the inner parts ( Figure 3A) In two sectors (2 and 4) the concentration in these two parts was different (lower in the outer part in sector 4, and the opposite in sector 2) ( Figure 3B), which made the interaction between sector and part (inner-outer) near to the statistical significance (p = 0.08, a 2-way ANOVA on the relative concentration, Supplementary Materials Table S1). The toxin was, in general, homogeneously distributed between the internal and external regions of the digestive gland ( Figure 3A, p = 0.8921 in a 2-way ANOVA on the relative concentration, Supplementary Materials Table S1.). The variability, nevertheless, was higher in the outer than in the inner parts ( Figure 3A) In two sectors (2 and 4) the concentration in these two parts was different (lower in the outer part in sector 4, and the opposite in sector 2) ( Figure 3B), which made the interaction between sector and part (inner-outer) near to the statistical significance (p = 0.08, a 2-way ANOVA on the relative concentration, Supplementary Materials Table S1).

Distribution of Domoic Acid between Cellular Types
Domoic acid was preferentially accumulated into large (mostly digestive) cells ( Figure 4A, p = 0.0039 by Wilcoxon paired samples test). The measured concentration of this toxin in small cells (basophilic, mostly secretory) was less than 25% of that in large cells.

Distribution of Domoic Acid between Cellular Types
Domoic acid was preferentially accumulated into large (mostly digestive) cells ( Figure 4A, p = 0.0039 by Wilcoxon paired samples test). The measured concentration of this toxin in small cells (basophilic, mostly secretory) was less than 25% of that in large cells. The recorded differences in toxin burden ( Figure 4B) were higher than those in concentration because the large cells contributed more to the digestive gland weight than the small ones.

Discussion
The variability of the toxin concentration among areas of the digestive gland of the scallops is moderate-low. It is lower than the usual inter-individual variation in scallops [52,59] and other bivalves [47,64,65]. Most sectors of the digestive gland had very similar concentrations, with the only significant exception being the sector adjacent to the gonad. In addition, in this sector the outer region was shown to have a lower concentration than the inner one. The most likely cause for this difference is the migration of the toxin to the intestinal loop located in the gonad. In previous studies [66], it has been observed that gonadal growth during depuration produces an increase in DA burden in this organ. The origin of the toxin incorporated by the intestinal loop should be the digestive gland, which is where the DA in the scallops is mainly stored. The mechanism involved, nevertheless, is unknown. A possibility could be the reabsorption of the toxin excreted by the digestive gland by the intestinal cells. It has been shown that those cells can significantly absorb different substances [62], but the very low depuration rate of the scallops [53,54] suggests that the amount of DA that could be reabsorbed is very low. In any case, this mechanism would not be expected to affect some parts of the gonad differently. A second possibility could be the transfer of the toxin from the digestive gland cells to the newly generated intestinal ones. Transfers of DA between organs have been documented but are significant only in exceptional situations, such as post-spawning periods [53,66,67]. This mechanism could or could not have an effect on the distribution of DA in the different areas of the digestive gland. The third possibility is that the intestinal loop cells which contain the toxin derive from cells from the digestive gland. If the new cells of the intestinal loop are generated in the nearby area of the digestive gland, a reduction in the toxin concentration in that area would be expected. Intestinal cells are mostly epithelial and the presence of immature cells, as in the digestive gland, has not been observed [62]. It is therefore possible that the intestinal cells derive from immature cells in the The recorded differences in toxin burden ( Figure 4B) were higher than those in concentration because the large cells contributed more to the digestive gland weight than the small ones.

Discussion
The variability of the toxin concentration among areas of the digestive gland of the scallops is moderate-low. It is lower than the usual inter-individual variation in scallops [52,59] and other bivalves [47,64,65]. Most sectors of the digestive gland had very similar concentrations, with the only significant exception being the sector adjacent to the gonad. In addition, in this sector the outer region was shown to have a lower concentration than the inner one. The most likely cause for this difference is the migration of the toxin to the intestinal loop located in the gonad. In previous studies [66], it has been observed that gonadal growth during depuration produces an increase in DA burden in this organ. The origin of the toxin incorporated by the intestinal loop should be the digestive gland, which is where the DA in the scallops is mainly stored. The mechanism involved, nevertheless, is unknown. A possibility could be the reabsorption of the toxin excreted by the digestive gland by the intestinal cells. It has been shown that those cells can significantly absorb different substances [62], but the very low depuration rate of the scallops [53,54] suggests that the amount of DA that could be reabsorbed is very low. In any case, this mechanism would not be expected to affect some parts of the gonad differently. A second possibility could be the transfer of the toxin from the digestive gland cells to the newly generated intestinal ones. Transfers of DA between organs have been documented but are significant only in exceptional situations, such as post-spawning periods [53,66,67]. This mechanism could or could not have an effect on the distribution of DA in the different areas of the digestive gland. The third possibility is that the intestinal loop cells which contain the toxin derive from cells from the digestive gland. If the new cells of the intestinal loop are generated in the nearby area of the digestive gland, a reduction in the toxin concentration in that area would be expected. Intestinal cells are mostly epithelial and the presence of immature cells, as in the digestive gland, has not been observed [62].
It is therefore possible that the intestinal cells derive from immature cells in the digestive gland and consequently inherit the toxin they contain. This would explain the lower DA concentration in the sector closest to the gonad and also the greater reduction in the outer region relative to the inner one, because the cells in this region are the closest to the intestinal loop. The confirmation of this hypothesis requires additional studies.
Taking into account the involved mechanisms proposed and the observed variability, it seems clear that, in general, sampling any part of the digestive gland could be enough to trace the time-course of the toxin accumulation in that organ, which accumulates more than 90% of the DA in the scallops. Non-destructive sampling utilizing biopsies [61], for example, could be used without targeting a precise part of the organ. Nevertheless, avoiding sampling the sector adjacent to the gonad could reduce the variability, especially during periods of intense gonadal growth.
Biopsies could be especially useful for experimental studies of DA depuration in Pecten maximus because of its size, commercial value and high inter-individual variation. In most studies of bivalves, the problem of inter-individual variation is usually reduced by using a large number of organisms, which are analysed individually or in pooled samples. Large numbers of scallops are difficult to place in most experimental facilities and this, together with the economic cost, prevents its use in most cases. Sampling the same scallops during the depuration period would reduce the interindividual variability to the variability between biopsies and, from our study, it seems clear that the biopsies could be representative of the whole digestive gland (which contains more than 90% of the toxin in the scallop). For monitoring natural beds, using biopsies could also have some benefits because it would be not necessary to sacrifice the scallops to have a rough estimate of their toxin concentration.
Domoic acid was preferentially retained in large cells, mostly absorptive (digestive cells) [68]. The concentration measured in small cells, mostly secretory and immature cells, was substantially lower than in the digestive ones. In terms of toxin burden, the importance of large cells was even greater, because they are found in a larger amount in the digestive gland (more biomass). The difference in concentration opens two possibilities regarding the methods of depuration not taken into account to date: (a) the secretory cells depurate the DA they have accumulated during the intoxication phase (which could be the case if both cell types accumulate the toxin to the same extent); and (b) DA is transferred from absorptive to secretory cells, which excrete it. Unfortunately, no information about the distribution of DA during the intoxication phase exists and the transfer between cellular types has not been documented, so new studies would be required to shed some light on the subject. However, in our opinion, neither of the two ways is very likely. Firstly, it seems unlikely that absorptive and secretory cells would absorb and accumulate DA during the intoxication at the same rate, because of their main functions. Secondly, it is difficult to explain how the toxin could pass the cellular membrane of the absorptive cells to go to the secretory ones but not to be excreted (the depuration rate is very low [53,54]) or transferred to other tissues [53,54].
The same difference in toxin concentration between cellular types of the digestive gland was found in the mussel Mytilus galloprovincialis for okadaic acid. Okadaic acid is a lipophilic toxin [69], which is chemically very different to DA, suggesting that the mechanism underlying the observed differences could be independent of the chemical structure of the toxins.

Biological Material
The scallops used in this work were collected from the Ría de Arousa after a toxic Pseudo-nitzschia bloom and maintained for two to four days in running seawater.

Sector Distribution
The digestive glands of three scallops were dissected into four sectors, which were chosen based on their proximity to the adductor muscle (1 and 3) and to the gonad (2 and 4). The external and internal regions of each sector were also separated (Figure 1), taking into account that the first one is expected to be enriched with acini and the second one with primary and secondary ducts. Therefore, eight fractions of each gland were obtained. All inner regions of the dissected sectors contained some gastric tissue, which was not removed because it would have also been sampled in the biopsies.
Each fragment was weighed and extracted with methanol 50% (1:36 w/v) by homogenization with Ultraturrax (12,000 rpm, 3 min) and centrifuged at 9960× g for 25 min. Throughout the extraction process, the samples were kept in an ice bath. A 0.5-mL aliquot of each extract was collected for HPLC analysis.

Cell Type Separations
The dissociation of the digestive gland pieces to obtain the cells was carried out using the method of Giard et al. [70], which involves an enzymatic treatment. The digestive glands from nine scallops were divided into 1-2 mm pieces, in 20 mM, pH 7.4, HEPES buffer, containing 436 mM NaCl, 53 mM MgSO 4 , 10 mM KCl, 10 mM CaCl 2 and 11 mM glucose. The fragments were washed twice in this medium and then dissociated enzymatically through two 30-min incubations in the previous buffer containing 0.1% pronase under gentle agitation, at room temperature. The cell suspension obtained was filtered through a 100 µm Nylon mesh and the cell types were separated by centrifugation.
After several tests at different centrifugation speeds and checking the separation of cell types by microscope examination, 43.5 and 77× g were chosen to precipitate large and small cells, respectively. Thus, the cell suspension was centrifuged for five minutes at 43.5× g. The pellet containing the larger cell types (mostly digestive cells) was reserved. The smaller cell types (mainly secretory cells and non-differentiated cells) were then precipitated from the supernatant by centrifugation at 77× g for 5 min and the supernatant was discarded.
Each pellet was weighed after centrifugation and careful removal of the supernatant. The DA was extracted by sonication in 50% MeOH (1:10 w/v) for 10 min in a sonication bath and by centrifugation at 11,136× g for 5 min. A 0.5 mL aliquot of each extract was collected for HPLC analysis.

Chromatographic Analysis
The analyses were carried out by HPLC-UV using a Thermo Surveyor PDA and the method of Regueiro et al. [71] with small modifications and omitting the preliminary cleaning steps. A monolithic column Chromolith Performance RP-18 (100 × 3 mm, Merck) was used and the separation was carried out isocratically with ACN: HCOOH: H2O (7:0.2:92.8 v/v/v) as mobile phase, flow 1.2 mL/min and the column temperature was maintained at 30 • C. An injection volume of 20 µL was used. The absorbance was monitored at 242 and 265 nm, corresponding to the peak wavelengths at which DA and tryptophan (its main interfering compound) absorb, respectively. The limit of detection of the method is less than 0.1 µg g −1 .

Computations and Statistical Analysis
The ratios of the different parts or sectors to the mean were obtained by dividing the toxin concentration in that part or sector by the mean concentration in the digestive gland. The mean concentration in the digestive gland was calculated with the expression C p ·W p / W p , where C p is the concentration in each piece (part-sector) and W p are the corresponding weights.
All of the statistical analyses were carried out with R [72] and all of the figures were generated with the R package ggplot2 [73].
Supplementary Materials: The following are available online at http://www.mdpi.com/2072-6651/12/6/371/s1. Table S1: Two-way ANOVA of the DA concentration relative to the mean of the digestive gland as a function of the sector, the inner or outer area and their interaction, Table S2: Two-way ANOVA of the DA concentration relative to the mean of the digestive gland as a function of the distance to the gonad, the distance to the adductor muscle and their interaction.