In Vivo Evaluation of the Chronic Oral Toxicity of the Marine Toxin Palytoxin

Palytoxin (PLTX) is one of the most poisonous substances known to date and considered as an emergent toxin in Europe. Palytoxin binds to the Na+-K+ ATPase, converting the enzyme in a permeant cation channel. This toxin is known for causing human fatal intoxications associated with the consumption of contaminated fish and crustaceans such as crabs, groupers, mackerel, and parrotfish. Human intoxications by PLTX after consumption of contaminated fishery products are a serious health issue and can be fatal. Different reports have previously explored the acute oral toxicity of PLTX in mice. Although the presence of palytoxin in marine products is currently not regulated in Europe, the European Food Safety Authority expressed its opinion on PLTX and demanded assessment for chronic toxicity studies of this potent marine toxin. In this study, the chronic toxicity of palytoxin was evaluated after oral administration to mice by gavage during a 28-day period. After chronic exposure of mice to the toxin, a lethal dose 50 (LD50) of 0.44 µg/kg of PLTX and a No-Observed-Adverse-Effect Level (NOAEL) of 0.03 µg/kg for repeated daily oral administration of PLTX were determined. These results indicate a much higher chronic toxicity of PLTX and a lower NOAEL than that previously described in shorter treatment periods, pointing out the need to further reevaluate the levels of this compound in marine products.


Introduction
The marine toxin palytoxin (PLTX) is one of the most toxic marine phycotoxins known to date. Since its first isolation, about 50 years ago, from marine corals of the genus Palythoa [1] and later on from algae of the genus Ostreopsis [2][3][4], the toxin was associated with high toxicity. Even though the toxic organisms producers of PLTX were initially located in the Pacific Sea, now the presence of the toxin and its congeners has expanded also to European waters. In fact, a recent report described a massive bloom of the dinoflagellate Ostreopsis cf. ovata producer of 5.6 pg of PLTX per cell in the Adriatic Sea [5]. This bloom was not an isolated case but rather massive Ostreopsis blooms have been reported in the Mediterranean Sea during the last 20 years as summarized in several reports [5] and increasingly reported in the literature [2,3,[5][6][7][8][9][10], and are in expansion mainly due to environmental factors [11][12][13].

Chronic Toxicity Elicited by Daily Oral PLTX
Based on the initial results for the acute toxicity of PLTX in mice [35], we further explored the chronic toxicity of this compound assessing its oral toxicity after a 28 daily exposure period. Thus, the first dose of palytoxin evaluated to test its chronic toxicity was 3.5 µg/kg/day over a 28 day treatment period, but with this dose a mortality of about 75% was observed, therefore, subsequent dilutions of the toxin were made in order to achieve the chronic NOAEL for the toxin. Table 1 represents the mortality rate at the different PLTX doses evaluated. As shown in this table, at 3.5 µg/kg/day, only two out of eight mice survived the 28-day treatment period. At the highest dose evaluated (10 µg/kg/day) animals survived 8, 10 and 18 days and none of the animals evaluated completed the entire treatment period. As shown in Figure 1A, non-linear fit of the data represented on Table 1 allowed to calculate a chronic lethal dose 50 (LD 50 ) for PLTX of 0.44 µg/kg (95% confidence interval (CI) from 0.22 to 0.9 µg/kg, R 2 = 0.9122, df (degrees of freedom) = 7), although the survival times differed considerably between the individual mice fed with each dose of PLTX as depicted in Figure 1B.
Toxins 2020, 12,489 3 of 18 evaluated (10 µg/kg/day) animals survived 8, 10 and 18 days and none of the animals evaluated completed the entire treatment period. As shown in Figure 1A, non-linear fit of the data represented on Table 1 allowed to calculate a chronic lethal dose 50 (LD50) for PLTX of 0.44 µg/kg (95% confidence interval (CI) from 0.22 to 0.9 µg/kg, R 2 = 0.9122, df (degrees of freedom) = 7), although the survival times differed considerably between the individual mice fed with each dose of PLTX as depicted in Figure 1B. Values are means ± SEM of the data obtained from three to ten animals.
All over the treatment period, mice weight was monitored weekly. As shown in Table 2, control mice and mice treated with PLTX at doses of 0.03, 0.1 and 0.3 µg/kg gained some weight during the 28 days treatment period while mice dosed with PLTX at 1 and 3.5 µg/kg lost weight during the treatment. Average weight of the animals during the treatment period and accumulative weekly weight change are shown in Table 2. Mice dosed with PLTX at 1 and 3.5 µg/kg exhibited significant weight lost at day 7 and day 14 of treatment and mice treated with 10 µg/kg lost weight at day 14, but this data was not significantly different due to the limited number of animals. Values are means ± SEM of the data obtained from three to ten animals.
All over the treatment period, mice weight was monitored weekly. As shown in Table 2, control mice and mice treated with PLTX at doses of 0.03, 0.1 and 0.3 µg/kg gained some weight during the 28 days treatment period while mice dosed with PLTX at 1 and 3.5 µg/kg lost weight during the treatment. Average weight of the animals during the treatment period and accumulative weekly weight change are shown in Table 2. Mice dosed with PLTX at 1 and 3.5 µg/kg exhibited significant weight lost at day 7 and day 14 of treatment and mice treated with 10 µg/kg lost weight at day 14, but this data was not significantly different due to the limited number of animals.
Even at PLTX doses as low as 0.1 µg/kg, animals fed with the toxin administered daily by gavage exhibited symptoms typical of the toxin that are summarized in Table 3, specifying the number of animals that exhibited the different symptoms. No symptoms were observed after dosing the mice with 0.03 µg/kg PLTX, while piloerection was the most frequent observation at the dose of 0.1 µg/kg PLTX. At the 0.3 µg/kg PLTX dose, the most common symptoms were lethargy, piloerection, and facial swelling. In addition, dyspnea appeared at the doses of 1, 3.5 and 10 µg/kg together with signs of abdominal pain and, as in our preliminary acute study [35], kyphosis and ataxia were also present after PLTX administration. Table 3.
Main symptoms after oral administration of PLTX by gavage (animals with symptomatology/total animals for each dose).

Chronic Oral PLTX Induced Macroscopic Changes in the Intestines and Stomachs of Treated Animals
After euthanasia, the animals were subjected to macroscopic observation on necropsy. As shown in Figure 2, no macroscopic alterations were observed neither in control animals nor in animals fed 28 days with 0.03 µg/kg PLTX. However, in one animal treated with 0.1 µg/kg PLTX that survived for 27 days, the stomach and small intestines were empty of solid contents and instead these structures contained dark-yellow mucous liquid. At the same dose, after daily feeding with 0.1 µg/kg PLTX, two mice showed swollen abdomen at death and survival times of 15 and 27 days. At the dose of 0.3 µg/kg, one mouse that survived 18 days of treatment showed gas accumulation in the stomach and intestines and absence of solid content. For this group of mice, two out of seven animals presented abdominal swelling and survival times of 18 and 24 days. Similarly, stomach and intestine alterations were observed in other mice fed daily with 1 µg/kg PLTX and that survived 25 days. At this dose, 4 out of 10 animals showed abdominal swelling and survival times of 2, 15, 25, and 28 days from the beginning of treatment. Similar alterations were observed in a mouse fed daily with 3.5 µg/kg PLTX that survived for 15 days. At this dose, five out of eight mice showed abdominal swelling at death (survival times ranged from 7 h for one animal, 9 days for another mouse, 15 days for 2 mice, and 28 days in the last animal with abdominal swelling). Notwithstanding, it is noteworthy to remark that abdominal swelling and the presence of gas and mucus in the stomach and intestines of PLTX-treated animals was a common feature. At the highest dose employed, 10 µg/kg, only one mouse, which survived 8 days, did not present changes at necropsy.

Chronic Oral PLTX Induced Macroscopic Changes in the Intestines and Stomachs of Treated Animals
After euthanasia, the animals were subjected to macroscopic observation on necropsy. As shown in Figure 2, no macroscopic alterations were observed neither in control animals nor in animals fed 28 days with 0.03 µg/kg PLTX. However, in one animal treated with 0.1 µg/kg PLTX that survived for 27 days, the stomach and small intestines were empty of solid contents and instead these structures contained dark-yellow mucous liquid. At the same dose, after daily feeding with 0.1 µg/kg PLTX, two mice showed swollen abdomen at death and survival times of 15 and 27 days. At the dose of 0.3 µg/kg, one mouse that survived 18 days of treatment showed gas accumulation in the stomach and intestines and absence of solid content. For this group of mice, two out of seven animals presented abdominal swelling and survival times of 18 and 24 days. Similarly, stomach and intestine alterations were observed in other mice fed daily with 1 µg/kg PLTX and that survived 25 days. At this dose, 4 out of 10 animals showed abdominal swelling and survival times of 2, 15, 25, and 28 days from the beginning of treatment. Similar alterations were observed in a mouse fed daily with 3.5 µg/kg PLTX that survived for 15 days. At this dose, five out of eight mice showed abdominal swelling at death (survival times ranged from 7 h for one animal, 9 days for another mouse, 15 days for 2 mice, and 28 days in the last animal with abdominal swelling). Notwithstanding, it is noteworthy to remark that abdominal swelling and the presence of gas and mucus in the stomach and intestines of PLTX-treated animals was a common feature. At the highest dose employed, 10 µg/kg, only one mouse, which survived 8 days, did not present changes at necropsy.

Chronic Oral PLTX at Doses of 1 and 3.5 µg/kg Decreased Urine Production After the 28 Days Treatment Period
On day 27 after the beginning of treatment, mice were housed individually in metabolic cages and the last dose of toxin was administered by gavage, before monitoring food consumption and urine and Toxins 2020, 12, 489 6 of 17 faeces production over a 24 h period. As shown in Table 4, at the highest doses evaluated after daily oral administration of PLTX, urine production decreased by about 88% at the doses of 1 and 3.5 µg/kg (Table 4). Urine samples were limited to the number of animals that survived the entire treatment. Table 4. Food intake, faeces and urine production over a 24 h period in control mice and in mice treated with PLTX. Urine production was decreased in mice treated with PLTX at doses of 1 and 3.5 µg/kg, but no significant changes were found mainly due to the limited number of animals.

Urine Parameters in Control Mice and in Mice Treated Daily with PLTX
Urine parameters were evaluated in samples collected during the 24 h period before sacrifice, after mouse placement in individual metabolic cages. The animals employed for this analysis were the mice that survived the 28 days treatment period. Urine parameters analyzed included colour, clarity, specific gravity, protein, glucose, ketones, blood/hemoglobin, bilirubin, and urobilinogen. Urine parameters in control and PLTX-treated mice are summarized in Table 5. No statistically significant differences were found between control mice and animals dosed daily with PLTX besides a bilirubin increase at PLTX doses of 0.1 µg/kg (t = 2.959; df = 9; p = 0.016) and 3.5 µg/kg (t = 2.579; df = 6; p = 0.0418). Table 5. Analytical results of urine in control animals and in animals treated daily with PLTX at doses ranging from 0.03 µg/kg to 3.5 µg/kg for 28 days. Values are expressed as mean ± SEM and the number of animals used to evaluate urine parameters is shown in parenthesis for each dose. * p < 0.05 versus control animals. For urine protein, glucose, ketones, blood haemoglobin, and bilirubin values represent mean ± SEM (in bold) and after the individual values for each mouse.

Effect of 28 Day Repeated Exposure of Mice to Low Doses of PLTX on Blood Biochemical Parameters
At the end of the 28 days of treatment, mice were euthanized, and biochemical analysis was performed in blood extracted by cardiac puncture immediately after euthanasia or, in some cases, after sudden death of animals. Blood samples were analyzed as in the previous study [35] and blood parameters analyzed included electrolyte levels (Cl − , Na + and K + ), alanine transaminase (ALT), aspartate transaminase (AST), lactate dehydrogenase (LDH), and creatine kinase (CK). As shown in Figure 3A, mean blood levels of ALT in control animals were 44 ± 4 U/L (n = 6) and 52 ± 5 U/L (n = 5) in animals dosed with 0.03 µg/kg. ALT blood levels were significantly increased to 163 ± 46 U/L in seven animals dosed with 0.1 µg/kg PLTX (t = 2.410; df = 11; p = 0.0346). In mice fed orally with PLTX at 0.3 µg/kg ALT blood levels were 103 ± 49 U/L, 116 ± 49 U/L in mice fed with 1 µg/kg PLTX, 128 ± 46 U/L in the four mice fed with 3.5 µg/kg PLTX, and 49 U/L in one mice that survived 10 days after daily PLTX dosage. Besides mean AST blood levels ( Figure 3B), were 157 ± 26 U/L (n = 6) in control mice, 210 ± 23 U/L in mice dosed daily with 0.03 µg/kg (n = 5), 535 ± 206 U/L (n = 7) in mice treated with 0.1 µg/kg PLTX, 869 ± 435 U/L (n = 5) in mice fed with 0.3 µg/kg PLTX, 343 ± 113 U/L (n = 6) in mice fed orally with PLTX at 1 µg/kg, 301 ± 47 U/L (n = 4) in mice fed with 3.5 µg/kg, and 194 U/L in one mice after 10 days administration of PLTX at 10 µg/kg. Statistical analysis of AST data did not yield significant differences between the control group and the PLTX-treated animals with the exception of mice fed with 3.5 µg/kg (t = 2.935; df = 8; p = 0.0189), but the levels of AST in 0.1, 0.3, 1 and 3.5 groups were above the normal reference range (59-247 U/L for the IDEXX). In contrast, creatine kinase Toxins 2020, 12, 489 8 of 17 blood levels ( Figure 3C) were 508 ± 147 U/L for the control group (n = 6), 708 ± 220 U/L for the group of mice dosed with 0.03 µg/kg (n = 5), 1781 ± 572 U/L (n = 6), 3040 ± 1319 U/L (n = 5), 512 ± 99 U/L (n = 5), 654 ± 208 U/L (n = 4), and 814 U/L (n = 1) for the groups of mice dosed with 0.1, 0.3, 1, 3.5 and 10 µg/kg PLTX, respectively. Statistical comparison of creatine kinase data did not show significant differences between control animals and animals treated daily with PLTX. Nevertheless, in this case, the blood level of CK in the animals dosed daily with 0.1 and 0.3 µg/kg of PLTX was well above the IDEXX reference range, which is between 68 and 1070 U/L, with some animals showing CK levels well above 1070 U/L. As shown in Figure 3D, mean LDH values were 1277 ± 229 U/L (n = 6) in the control group, 1904 ± 338 U/L in three animals dosed with 0.03 µg/kg, 6230 ± 2448 U/L (n = 6) in the 0.1 µg/kg group, 2716 ± 841 U/L (n = 3) for the 0.3 µg/kg group, 1626 ± 325 U/L (n = 4) in the 1 µg/kg group, 1275 ± 296 U/L (n = 4) in the 3.5 µg/kg group, and finally 1914 U/L in one mouse at the highest dose employed. In this case, no statistically significant differences were observed between control animals and those that received a daily PLTX dose. However, two out of the six mice fed with 0.1 µg/kg PLTX had LDH levels well above the IDEXX reference range for blood LDH, which is between 1105 and 3993 U/L. with 0.1 µg/kg PLTX, 869 ± 435 U/L (n = 5) in mice fed with 0.3 µg/kg PLTX, 343 ± 113 U/L (n = 6) in mice fed orally with PLTX at 1 µg/kg, 301 ± 47 U/L (n = 4) in mice fed with 3.5 µg/kg, and 194 U/L in one mice after 10 days administration of PLTX at 10 µg/kg. Statistical analysis of AST data did not yield significant differences between the control group and the PLTX-treated animals with the exception of mice fed with 3.5 µg/kg (t = 2.935; df = 8; p = 0.0189), but the levels of AST in 0.1, 0.3, 1 and 3.5 groups were above the normal reference range (59-247 U/L for the IDEXX). In contrast, creatine kinase blood levels ( Figure 3C) were 508 ± 147 U/L for the control group (n = 6), 708 ± 220 U/L for the group of mice dosed with 0.03 µg/kg (n = 5), 1781 ± 572 U/L (n = 6), 3040 ± 1319 U/L (n = 5), 512 ± 99 U/L (n = 5), 654 ± 208 U/L (n = 4), and 814 U/L (n = 1) for the groups of mice dosed with 0.1, 0.3, 1, 3.5 and 10 µg/kg PLTX, respectively. Statistical comparison of creatine kinase data did not show significant differences between control animals and animals treated daily with PLTX. Nevertheless, in this case, the blood level of CK in the animals dosed daily with 0.1 and 0.3 µg/kg of PLTX was well above the IDEXX reference range, which is between 68 and 1070 U/L, with some animals showing CK levels well above 1070 U/L. As shown in Figure 3D, mean LDH values were 1277 ± 229 U/L (n = 6) in the control group, 1904 ± 338 U/L in three animals dosed with 0.03 µg/kg, 6230 ± 2448 U/L (n = 6) in the 0.1 µg/kg group, 2716 ± 841 U/L (n = 3) for the 0.3 µg/kg group, 1626 ± 325 U/L (n = 4) in the 1 µg/kg group, 1275 ± 296 U/L (n = 4) in the 3.5 µg/kg group, and finally 1914 U/L in one mouse at the highest dose employed. In this case, no statistically significant differences were observed between control animals and those that received a daily PLTX dose. However, two out of the six mice fed with 0.1 µg/kg PLTX had LDH levels well above the IDEXX reference range for blood LDH, which is between 1105 and 3993 U/L.  Next, the effect of repeated exposure of mice to low doses of PLTX on blood electrolyte levels was evaluated (Figure 4). Statistically significant differences were found in sodium and potassium levels as well as in the Na + /K + ratio and on blood chloride levels, this dysregulation of electrolytes levels is in agreement with the previously reported alterations produced by a single oral dose of PLTX [35].

Ultrastructural Examination of Stomach Damage Induced by Repeated Oral Dosing of Mice with PLTX
Since, after necropsy, macroscopic changes in the digestive tract of PLTX-treated animals were observed, and some mice dosed with PLTX at 1 and 3.5 µg/kg exhibited significant weight loss, transmission electron microscopic studies (TEM) of the stomach were performed. Samples from control animals and from animals treated with PLTX were obtained immediately after euthanasia and evaluated. First, and in view of the macroscopic alterations of the stomach with dark-yellow liquid and dilatation with gas accumulation, mucous cells were evaluated. After TEM analysis, some evident changes in the structure of mucous cells, which were identified by their electron-lucent granules, were appreciated as Figure 5 shows. Thus, in control animals, a regular structure of the mucous cells was observed ( Figures 5A,B), while in animals treated daily with PLTX at doses of 1 µg/kg during 9 days (Figures 5C,D)   As shown in Figure 4A, mean Na + blood values in control animals were 155 ± 1 mmol/L (n = 4) while in PLTX-treated animals, sodium levels were 156 ± 2 mmol/L (n = 5) at 0.03 µg/kg, 152 ± 2 mmol/L in seven animals dosed with 0.1 µg/kg of PLTX. At the dose of 0.3 µg/kg of PLTX, five animals have mean blood Na + values of 149 ± 1 mmol/L and significant differences were observed versus control animals (t = 3.332, df = 7, p = 0.0126). Similarly, blood sodium levels were 159 ± 2 mmol/L (n = 6) and 158 ± 2 mmol/L (n = 4) in animals dosed with 1 and 3.5 µg/kg PLTX, respectively. Moreover, at the highest dose employed (10 µg/kg), the blood sodium level was 150 mmol/L, but only one mouse was evaluated. Interestingly, an increase in potassium levels was found in blood samples as indicated in Figure 4B. Animals fed daily with PLTX doses of 0.03, 0.1, and 0.3 µg/kg presented a statistically significant increase in blood potassium levels comparing with control animals. In control animals, K + levels were 6.8 ± 0.2 mmol/L (n = 4), 8.6 ± 0.1 mmol/L (n = 5) in mice fed daily with 0.03 µg/kg PLTX (t = 4.053; df = 7; p = 0.0048), 9.3 ± 0.3 mmol/L in seven mice fed with 0.1 µg/kg PLTX (t = 5.611; df = 9; p = 0.0003) and 8.6 ± 0.5 mmol/L (n = 5) in mice treated with 0.3 µg/kg PLTX (t = 3.255; df = 7; p = 0.0140). Despite the short survival time of the animals treated with the highest PLTX dose, blood K + levels were well above the normal range in one mouse that survived 10 days of treatment. Due to the decrease in blood sodium concentration and the increase in blood potassium levels, the ratio Na + /K + decreased ( Figure 4C), especially at low doses of PLTX, for which survival times were longer. In control mice, the ratio Na + /K + was 23 (n = 4) and it decreased to 19 ± 1 (t = 4.000, df = 6, p = 0.0071), 18 ± 1 (t = 5.547, df = 6, p = 0.0015) and 19 ± 1 (t = 3.313, df = 5, p = 0.0212) in animals treated with 0.03, 0.1 and 0.3 µg/kg PLTX, respectively. Next, the effects of oral PLTX on blood chloride levels were evaluated ( Figure 4D). Chloride blood levels increased compared with control, and statistically significant variations were found between control animals and PLTX-treated animals. Thus, Cl − levels were 112 ± 2 mmol/L in four control animals, 118 ± 1 mmol/L (t = 3.041, df = 7, p = 0.0188) in animals fed daily with 0.03 µg/kg PLTX, 120 ± 1 mmol/L (t = 5.066, df = 9, p = 0.0007) after the 0.1 µg/kg treatment, 117 ± 1 mmol/L (p = 0.0192) after the 0.3 µg/kg treatment, and 119 ± 1 mmol/L (p = 0.0047) after the 1 µg/kg treatment.

Ultrastructural Examination of Stomach Damage Induced by Repeated Oral Dosing of Mice with PLTX
Since, after necropsy, macroscopic changes in the digestive tract of PLTX-treated animals were observed, and some mice dosed with PLTX at 1 and 3.5 µg/kg exhibited significant weight loss, transmission electron microscopic studies (TEM) of the stomach were performed. Samples from control animals and from animals treated with PLTX were obtained immediately after euthanasia and evaluated. First, and in view of the macroscopic alterations of the stomach with dark-yellow liquid and dilatation with gas accumulation, mucous cells were evaluated. After TEM analysis, some evident changes in the structure of mucous cells, which were identified by their electron-lucent granules, were appreciated as Figure 5 shows. Thus, in control animals, a regular structure of the mucous cells was observed ( Figure 5A,B), while in animals treated daily with PLTX at doses of 1 µg/kg during 9 days ( Figure 5C,D) or 3.5 µg/kg for 22 days ( Figure 5E,F), the surface of mucous cells was damaged and most mucous cells showed an irregular nuclei with chromatin margination and vacuolated cytoplasm with numerous autophagosomes. Aside from mucous cells, the gastric epithelium secretes hydrochloric acid and digestive enzymes, and since palytoxin altered blood ionic homeostasis, the acid-producing cells of the stomach (parietal cells) were also analyzed by TEM. As shown in Figure 6, in control animals, parietal cells conserved their characteristic microvilli, membrane invaginations called canaliculi and clusters of round vesicles that transfer substances from the cytoplasm to the lumen of the canalicular system ( Figure 6A,B). In contrast, in animals fed daily with PLTX at the dose of 1 µg/kg ( Figure  6C,D), evident alterations of the parietal cells were observed, consisting of damage of the cytoplasm with dilatation of the canaliculi system, electron dense mitochondria and proliferation of autophagosomes (black arrows). Moreover, in PLTX-treated animals, nuclear membrane shrinkage, Aside from mucous cells, the gastric epithelium secretes hydrochloric acid and digestive enzymes, and since palytoxin altered blood ionic homeostasis, the acid-producing cells of the stomach (parietal cells) were also analyzed by TEM. As shown in Figure 6, in control animals, parietal cells conserved their characteristic microvilli, membrane invaginations called canaliculi and clusters of round vesicles that transfer substances from the cytoplasm to the lumen of the canalicular system ( Figure 6A,B). In contrast, in animals fed daily with PLTX at the dose of 1 µg/kg ( Figure 6C,D), evident alterations of the parietal cells were observed, consisting of damage of the cytoplasm with dilatation of the canaliculi system, electron dense mitochondria and proliferation of autophagosomes (black arrows). Moreover, in PLTX-treated animals, nuclear membrane shrinkage, chromatin margination, and development of membrane-bound remnants of the nuclear membrane also called apoptotic bodies were observed. These alterations are consistent with programmed cell death that leads to apoptosis of the parietal cells of PLTX-treated animals [44]. The same ultrastructural alterations were observed in mice dosed daily with 3.5 µg/kg of PLTX ( Figure 6E,F).
Toxins 2020, 12, 489 12 of 18 also called apoptotic bodies were observed. These alterations are consistent with programmed cell death that leads to apoptosis of the parietal cells of PLTX-treated animals [44]. The same ultrastructural alterations were observed in mice dosed daily with 3.5 µg/kg of PLTX ( Figures 6E,F).

Discussion
So far, the available in vivo toxicity studies with palytoxin have shown a high acute oral toxicity with LD50 ranging from 510 to 767 µg/kg. Nevertheless, in the latest opinion of the EFSA experts on PLTX-related risks [41], the lack of data on the subchronic and chronic toxicity of PLTX and its analogues was highlighted. In view of these requirements, in this work we pursued to determine the chronic LD50 and the chronic NOAEL for oral PLTX. In this context, the data presented in this report  A,B) and of animals treated daily with 1 µg/kg PLTX (C,D) or PLTX at 3.5 µg/kg (E,F). Scale bar is 5 µm, and 2 µm, respectively from the left to the right panels. N: nucleus, C: canalicular system, M: mitochondria. Animals fed with PLTX showed apoptotic nuclei with nucleus membrane shrinkage (white arrows), abundant autophagosomes (black arrows), electron dense mitochondria and dilatation of the oval vesicles of the canalicular system with evident wide clear centers.

Discussion
So far, the available in vivo toxicity studies with palytoxin have shown a high acute oral toxicity with LD 50 ranging from 510 to 767 µg/kg. Nevertheless, in the latest opinion of the EFSA experts on PLTX-related risks [41], the lack of data on the subchronic and chronic toxicity of PLTX and its analogues was highlighted. In view of these requirements, in this work we pursued to determine the chronic LD 50 and the chronic NOAEL for oral PLTX. In this context, the data presented in this report constitute the first initial approach for the chronic evaluation of the PLTX toxicity after daily ingestion. Moreover, as far as we know, this is the first work that evaluates the chronic toxicity of this potent marine toxin following the OECD recommendations to evaluate the chronic toxicity of chemicals [43]. The data presented here indicate that daily feeding of female Swiss mice by gavage with different PLTX doses for 28 days resulted in an oral LD 50 of 0.44 µg/kg and a NOAEL of 0.03 µg/kg. In view of the high increase of toxicity after repeated exposure to PLTX, it is crucial to consider a representative scenario in which successive exposures to low doses of PLTX might arise, especially for the population living near coastal areas affected by Ostreopsis blooms. Indeed, besides oral toxicity, many reports have described toxic corneal reactions [45,46], as well as frequent respiratory and dermal alterations in humans [18,20,47,48].
Until now, different in vivo toxicities for PLTX in mice have been reported, depending mainly on the time of exposure of the animals to the toxin. For instance, a NOAEL of 300 µg/kg and a LD 50 of 757 µg/kg have been reported after a single dose of PLTX and an observation period of 24 h [37]. In the same conditions, an LD 50 of 510 µg/kg was reported by other authors [49], while no death was observed in mice after oral administration of PLTX at doses up to 200 or 300 µg/kg [42] but stomach alterations were evident 2 h after toxin administration. However, the toxin lethality increased in mice after oral repeated administration of successive doses and this was associated to a decrease in the NOAEL up to 3 µg/kg after a repeated oral dose of PLTX for 7 days [28]. However, to date, no additional reports on the chronic toxicity of PLTX have been released. We demonstrated here that after daily feeding of mice with PLTX during 28 days, a daily oral dose of 3.5 µg/kg led to a mortality rate of about 75% with survival times between 7 h and 28 days. In this context, it is important to remark that the NOAEL was much lower than that previously reported, since even after oral administration of the toxin at doses of 1 µg/kg, animals exhibited stomach ultrastructural alterations after 9 days of treatment and characteristic symptoms, some of them initiating on the third day of treatment. The results presented here indicate that after the 28-day treatment period, mice dosed with 0.03 µg/kg of PLTX did not present changes in most of the blood parameters analyzed in this work. However, the only alteration observed in mice fed with 0.03 µg/kg PLTX was hyperkalemia that appeared in two out of five animals. However, the modifications in blood electrolyte levels were much lower at this dose than that observed at higher toxin concentrations. These results are in consonance with previous studies that reported electrolyte changes after PLTX administration [28,35,38].
As the molecular target of PLTX, the Na + -K + -ATPase pump is indispensable for different cellular functions, and its inhibition results in an excess of extracellular K + and accumulation of intracellular Na + because, in the presence of toxin the Na + -K + -ATPase can no longer pump K + into the cell or pump Na + out of the cell [50], the resulting blood electrolyte homeostasis dysregulation observed in animals fed daily with PLTX was expected. The pathophysiology expected after PLTX intoxication, includes cardiovascular alterations as a consequence of the Na + -K + inhibition [51] and could result in other clinical signs such as piloerection, dyspnea, and abdominal pain that were even present in some mice fed daily with 0.1 µg/kg of PLTX. Clinical signs including lethargy, dyspnea and consequently cyanosis in the tail of the animals and neuromuscular alterations, kyphosis, circling, and ataxia were also present after daily PLTX administration, in consonance with the symptoms reported in other in vivo studies using higher doses of PLTX and shorter exposure periods [28,29,35,38]. Additionally, macroscopic alterations of the digestive tract were also present in animals fed with PLTX at doses of 0.3 µg/kg and higher. This observation is in agreement with previous data indicating gastric lesions after acute oral administration of higher PLTX doses [35,38,42]. Regarding electron microscopy analysis of the stomach, the results presented here showed a strong ultrastructural damage of the mucous and parietal cells which could probably lead to erosion and detachment of the gastric epithelium. This fact could be related with earlier PLTX studies reporting that the Na + -K + -ATPase was transformed in an open cation channel in the presence of PLTX [23,25] which will lead to cell membrane depolarization and the consequent change in ion channel conductance. Thus, it was expected that the sustained overload of the Na + concentration would affect all the ion transport systems coupled to the sodium gradient like the Na + /H + exchanger, the Na + /Ca 2+ exchanger, the H + -K + -ATPase pump and hence, the osmotic balance [52,53]. In this regard, death of parietal cells has been observed after inhibition of the H + -K + -ATPase with omeprazol [54,55]. In addition to the gastric alterations observed by electron microscopy, it is noteworthy to point out that blood electrolytes imbalances together with gastric acid changes could lead to nutrient deficiency and consequently weight loss [56], as observed after daily administration of PLTX at doses of 1 and 3.5 µg/kg.
Altogether, the new findings described in the present study, together with the increase in Ostreopsis blooms in Europe, raise the need to perform additional studies on the oral in vivo toxicity of this potent marine toxin in order to prevent human illness and accurately regulate the food levels of this emerging marine toxin.

Conclusions
Daily feeding of female Swiss mice by gavage with different PLTX doses for 28 days resulted in an oral LD 50 of 0.44 µg/kg and a NOAEL of 0.03 µg/kg.

In Vivo Experimental Procedure
In vivo studies were performed with Swiss female mice weighing an average of 21 to 25 g in each group (4 weeks old). All in vivo studies employed in the manuscript followed the European directives The PLTX used in the experiments was purchased from Fujifilm Wako Pure Chemical Corporation (Osaka, Japan) extracted from Palythoa tuberculosa, (purity 88.9% and CAS Number: 77734-91-9) and dissolved in water. Before administration, the stock PLTX solution was diluted in 0.9% saline solution to achieve each dose. For the treatment, mice received a daily oral dose of PLTX by gavage (200 µL of physiological saline at 9:30 am) or the same volume of control saline. Animals were observed closely for the first 2 h and periodically thereafter, until sacrifice of surviving animals at the end of the treatments. During the course of the experiments, moribund animals or animals obviously in pain or showing signs of agony, as well as animals with weight loss higher than 10% and enduring distress, were euthanized and considered in the interpretation of the test results in the same way as animals that were euthanized at the end of the 28 days treatment period. During treatment, mice were housed in rooms with constant temperature, humidity, and a controlled photoperiod at the animal facilities of the School of Veterinary Medicine of the University of Santiago de Compostela (Code: AE-LU-002). During experiments, control and PLTX-treated mice were weighted weekly at day 0 (day of the fist treatment), day 7, day 14, day 21 and day 28, and food consumption was also registered at the same intervals. Euthanasia was performed on day 28. After the last PLTX-dosage (day 27), control and PLTX-treated mice were placed in metabolic cages for 24 h to monitor urine and faeces production and obtain samples. Finally, animals were euthanized in a CO 2 chamber, and immediately after death confirmation by cervical dislocation, blood and tissue samples were collected. All animals in the study were subjected to a full necropsy, including detailed visualization of heart, liver, lungs, kidneys, spleen, stomach, duodenum, rectum, and cerebrum.

Evaluation of Enzyme and Electrolyte Levels in Plasma of Control and PLTX-Treated Mice
Blood was obtained directly from the ventricle of euthanized mice as previously described [35,57] and conserved in microtubes containing lithium heparin (2.5 U.I./mL) until the end of the necropsy and tissue collection (30 min). Afterwards, each blood sample was mixed for 30 s and centrifugated for 90 s at 15,800 rpm/12,000× g (IDEXX Stat Spin, IDEXX Europe B.V., Hoofddorp, The Netherlands). The computer controlled biochemical analyzer IDEXX Catalyst Dx (IDEXX VetLab Station, IDEXX Europe B.V) was used for the analysis of plasma parameters. In most of the samples, 300 µL of plasma were used to evaluate electrolyte levels (Cl − , Na + and K + ), alanine transaminase, aspartate transaminase, lactate dehydrogenase, and creatine kinase. Initially, all the markers were evaluated in undiluted samples, but if the marker level in plasma was above the reference range of the analyzer, an automatic dilution with physiological saline was performed. Samples with altered visual properties that could lead to the misinterpretation of the biochemical parameters (high hemolysis, icterus or lipidemia) were not further processed.

Urine Analysis
Urine colour, clarity, protein, glucose, ketones, blood hemoglobin, bilirubin, and urobilinogen analysis were performed using a reflectance photometer (IDEXX VetLab UA), which reads and evaluates IDEXX UA strips. Urine specific gravity was measured with a refractometer. Because for some mice a small amount of blood or urine were obtained, it was not possible to measure all the parameters for each animal.

Ultraestructural Analysis
Stomach samples preparation for transmission electron microscopy was performed according to previously described procedures [35,57,58]. Concisely, organ samples (1 mm 3 ) were fixed by immersion (2.5% glutaraldehyde in phosphate buffered saline (PBS) containing in mM: 137 NaCl, 8.2 Na 2 HPO 4 , 1.5 KH 2 PO 4 and 3.2 KCl) for 2 h at 4 • C and rinsed three times with PBS. Postfixation by immersion in 1% OsO 4 in 0.1 M cacodylate trihydrate buffer was performed for 60 min. After, tissues were rinsed and dehydrated in graded ethanol solutions, including one bath with 70% ethanol and 0.5% uranyl acetate, rinsed in propylene oxide, and embedded in Epon 812 (Momentive Specialty Chemicals Inc., Houston, TX, USA). A Leica Ultracut UCT ultramicrotome (Leica Microsystems GmbH, Wetzlar, Germany) was used to obtain ultrathin sections of the tissue samples. Samples were counterstained with uranyl acetate and lead citrate, and ultrastructural analysis of 1 mm 2 thick samples was performed with a JEOL JEM-1011 Transmission Electron Microscope (Jeol Ltd., Tokyo, Japan).