Human fatalities due to consumption of seafood suspected to be contaminated with PTX were reported in the Philippines, after consumption of the crab Demania reynaudii [40], and in Madagascar following consumption of the sardine Herklotsichthys quadrimaculatus [41]. In fact, PTX is now suggested to be the cause of clupeotoxism—a poorly understood toxic syndrome associated with the consumption of clupeoid fish, such as sardines, herrings and anchovies, and characterized by a high mortality rate [38,41]. Near fatal cases took place after consumption of a smoked fish (Decapterus macrosoma) in Hawaii [42] and after groupers (Epinephelus sp.) or blue humphead parrotfish (Scarus ovifrons) were eaten in Japan [43,44]. The most commonly reported complications of PTX poisoning appears to be rhabdomyolysis [42–44], a syndrome injuring skeletal muscle, causing muscle breakdown, and leakage of large quantities of intracellular (myocyte) contents into blood plasma [38]. Other symptoms associated with PTX poisoning in humans are characterized by a bitter/metallic taste, abdominal cramps, nausea, vomiting, diarrhea, paresthesia, bradycardia, renal failure, cyanosis and respiratory distress [40,42,43]. The latter two precede death in fatal cases [29].

Recent cases of human intoxication by PTX have shown different contours. In Germany [45] and in the USA [38], poisoning occurred through dermal absorption, after the victims touched zoanthid corals present in their home aquariums. Moreover, a case of human exposure to PTX via inhalation was also reported regarding a patient who attempted to kill a Palythoa coral in his aquarium [46]. Many other anecdotal evidences of intoxications related with aquarium zoanthids can be found in online marine aquarium forums [38]. Respiratory illness has also occurred when people were exposed to Ostreopsis ovata bloom aerosols during recreational or working activities, in Italy [47]. Symptoms caused by ovatoxin-a from aerosols include fever associated to serious respiratory disturbs, such as bronchoconstriction, mild dyspnea, and wheezes, while conjunctivitis was observed in some cases [11,47,48]. These reports highlight the existence of new routes of exposure to PTX and its analogs while it alerts for the risks experienced by aquarium hobbyists by keeping zoanthid in these artificial ecosystems. For a detailed review of human health risks, see [38].

PTX is hemorrhagic to mammals and has significant effects on cardiovascular, kidney, gastrointestinal and respiratory systems [49]. This toxin has been known to be extremely lethal to animals by intraperitoneal or intravenous administration, since its discovery [3,49]. It can be detected and quantitatively measured by the use of biological assays, although chemical analytical methods are necessary to confirm its presence in natural samples [50]. In a recent review about current assays for marine toxins, it was recognized that the mouse test is the most widely used for sample toxicity detection, and even the reference method for marine toxins in most countries [37]. Nevertheless, for palytoxin and analogs, this bioassay currently in-use is not yet fully developed [37,51]. For instance, data regarding mouse assays for PTX varies in the literature, since authors use different definitions to “mouse unit”, detection limits, LD₅₀ values (which ranges between 0.150 and 0.720 μg/kg), and observation time of mice (from 4 to 48 h) [51]. To reach to a consensus, Riobó et al. [51] deeply analyzed and described the proper employment of this bioassay for PTX, since it could constitute the reference for other methods, according to the authors.

The earliest study using animal models for PTX toxicity showed highly lethal effects of this substance in several species [49], although this work was done with semi-purified material [38]. In the first attempts to determine PTX toxicity, the crude ethanol extracts of Palythoa toxica proved to be so toxic that an accurate LD₅₀ was difficult to determine [52]. A summary of toxic effects and LD₅₀ of PTX for various exposure routes, as well as for several animal tested, can be found in [29] and [38]. In a similar way, a more complete toxicity assessment for PTX analogs from Ostreopsis spp. can be found in [52]. For comparative purpose, the LD₅₀ values of these toxins observed 24 hours after exposure in the two most common test animals are summarized in Table 1.

The intravenous LD₅₀ for PTX in rats and mice is 0.089 and 0.045 μg/kg, respectively (Table 1). Toxicity values obtained by this route in other mammals—rabbit, dog, monkey, and guinea pig—ranged between 0.025 and 0.45 μg/kg, being the former the most sensitive animal model tested so far [16,38]. Symptoms at high doses include ataxia, convulsions, dyspnea, and subsequently death within minutes [29]. In this case, the rapid death is attributed to heart failure. In contrast, at doses close to the LD₅₀, death occurs 8–10 h after administration, and cardiac activity still persists even after respiratory arrest [29].

Several studies [38,49] point to similar LD₅₀ values for PTX in mice either by intraperitoneal (0.4–0.72 μg/kg, variation due to different sources of PTX) or by intravenous (0.15–0.74 μg/kg) administration. Corroborating these findings, the 24 h LD₅₀ value for PTX in the mouse bioassay by intraperitoneal injection was recently [51] established to be 0.295 μg/kg (see Table 1). In rats, the acute toxicity by intraperitoneal administration is much lower than the intravenous LD₅₀ (0.63 μg/kg vs. 0.089 μg/kg, respectively). Regarding the PTX analogs, ostreocin-D injected by intraperitoneal route has a LD₅₀ of 0.75 μg/kg in mouse, a value similar to PTX, while mascarenotoxin-A from a crude extract of Ostreopsis mascarenensis has showed much lower toxicity, presenting a LD₅₀ value of 900 μg/kg [7,9,12]. Despite the reported [12] mouse lethality via intraperitoneal injection of ostreotoxin-1 and -3 produced by O. lenticularis, the classification of these compounds as PTX analogs is still unclear. This is because, so far, analytical methods for these molecules are still lacking [29,54]. Some studies [51,53,55] showed that the symptoms are quite consistent whether the intoxication is achieved by purified PTX or by analogs derived from crude extracts of Ostreopsis spp. or from toxic fish and crabs: uncoordinated movement and paralysis are early observations; dyspnea, cyanosis and exophthalmus precede death, while convulsions and diarrhea have been described in some cases. Riobó et al. [51] observed that at dose levels close to the LD₅₀, death may occur up to 48 h after toxin administration by intraperitoneal route. The characteristic initial symptoms of mice intoxication (within 15 min after administration) were stretching of hind limbs, lower backs and concave curvature of the spinal column, whether the mice died or stayed alive. The authors state that these early symptoms are sufficiently distinctive for allowing the identification of the presence of PTX regardless the presence of other toxins in the sample.

Although the risk associated with ingestion of contaminated seafood is well recognized [38], PTX is much less toxic orally than by parenteral administration aforementioned. Results from the few existing studies (Table 1) reports an oral LD₅₀ of >40 μg/kg in rats and 510 μg/kg in mice, by intragastric injection [29,49]. Recently it was shown that toxin administration by gavage in mice did not produce death at the maximum dosage of 200 μ/kg of PTX and 300 μg/kg of ostreocin-D [56]. Nevertheless, another recent study showed lethality by acute oral administration of PTX [53], despite these values were three orders of magnitude lower than that observed after intraperitoneal injection. The deaths started at 600 μg/kg and LD₅₀ was established at 767 μg/kg, a result somewhat comparable to that presented above. At lethal doses, the symptoms are similar to those observed after intraperitoneal administration (jumping, paralysis of the hind limbs and respiratory distress), suggesting that skeletal muscles, including the respiratory musculature, may be primary targets of PTX. These findings are supported by additional ultrastructural and hematoclinical data [53]. At sub-lethal doses of PTX and ostreocin-D administrations, weak erosion in the stomach caused by gastric juice secretion is observed, and also light injuries to the small intestine, lung and kidney [56]. Both toxins also induced organ injuries after 24 h when dosed by sublingual administration at about 200 μg/kg. The injuries became fatal when the PTX dose was given twice or three-times [56]. Controversially, evaluations for NOEL and LOEL values for acute oral administration of PTX in mice were estimated to be 300 μg/kg [53] and 200 μg/kg [56], respectively.

By intra-tracheal route, PTX caused mice mortality with doses above 2 μg/kg in 2 h and rats death at 5~7.5 μg/kg, with paralytic symptoms [56]. Values obtained earlier [49] in rats indicates a 24 h LD₅₀ of 0.36 μg/kg (Table 1). Ostreocin-D showed to be less potent than PTX, since mice died at 13 μg/kg after 1 h or 11 μg/kg after 6 h [56]. At sublethal dose (1 μg/kg), mice and rats were unable to walk for 1–2 h but recovered thereafter. After 24 h, surviving animals appeared to be in normal condition, but with multiple organ injuries (lung, gastro-intestines and kidney). The study [56] also indicates that the two toxins administered via the trachea move to the lung and further to other organs. Though slight differences could be seen, both toxins were very toxic by this route, which is a model for respiratory illness. This reinforces the concern to take into account the route of exposure whenever assessing the health risks (remember the aforementioned case in Italy [47], where inhalation of aerosols contaminated with dinoflagellate cells occurred).

Some other routes of exposure were investigated for the toxicity assessment of PTX [49]. It was demonstrated that this substance is highly toxic after intramuscular or subcutaneous injection; no toxicity was found after intrarectal administration. Furthermore, PTX caused significant, non-lethal effects when topically applied to skin or eyes [38].

Returning to human health risks, by extrapolation of the animal toxicity data presented above, the toxic dose in a human was estimated to be between 2.3 and 31.5 μg PTX [12,43,50]. Recently, an acute reference dose was suggested to be 64 μg for a 60 kg individual [57].

Not much is known about the effect of PTX and its analogs in invertebrates, and particularly in respect to developmental aspects. The few existing data are mainly related to ecological studies on the impact of Ostreopsis spp. in sea urchin communities and ecotoxicological effects in bivalves (see Section 3.3). The only documented effect of PTX in sea urchins reproduction is the inhibition of sperm motility [22]. Some other reported effects of PTX in invertebrates are retrieved from standard bioassays (Table 2). Toxicity caused by Ostreopsis siamensis isolates from New Zealand was studied for the brine shrimp Artemia salina and to larvae of the marine gastropod Haliotis virginea [59]. This dinoflagellate originating from New Zealand is known to produce PTX-like compounds [55]. O. siamensis killed brine shrimps, even at low cell numbers (250 cells per test well). The time until morbidity (tM₅₀, when test organisms showed minimal gill movement) was 4 h, while the time until 50% death (LT₅₀) was 24 h (Table 2). The dinoflagellate caused morbidity in gastropod larvae (tM₅₀–1 h), but death did not occur within the 24 h of the bioassay, even at 1000 cells per test well. In this case, morbidity was characterized by retracted viscera, settled and velum lost.

The motility of sperm from hamsters, guinea pigs, rabbits, cattle, humans and from the invertebrate sea urchins is inhibited by PTX exposure [22]. Its manifestation is characterized as a loss in flagellar-bend amplitude, which may be accompanied with an increase in beat frequency. The forward progression is lost gradually until the completely cessation of movement.

In animal developmental studies, the ability of PTX to block the Na⁺-K⁺ transporters and thus to depolarize the cell membrane is used. This capability was explored in such a study for Xenopus laevis [62], in which larvae were exposed to 2 nM PTX. It was shown that, at these concentrations, the larvae remain healthy and behave normally, although they do not regenerate their tails [62]. Some others potential physiological and developmental effects of PTX to vertebrates are also inferred from studies using the same model animal, e.g., by the frog embryo teratogenesis assay Xenopus (FETAX). By using this method, Franchini et al. [61] evaluated the toxicological effects of PTX in embryos at early gastrula stage of the anuran. It was found that at the highest toxin concentration tested (370 nM), the embryo population decreased by about 80% by the end of the assay (Table 2). An increase in the number of malformations and a delay in embryo growth were also observed. The modifications more frequently observed were folding along the antero-posterior body axis and swelling of the visceral mass. The histochemical analysis performed in the same study revealed the nature of the toxin-induced injuries. It was determined that the nervous system and the muscle tissue are sensitive target organs for PTX. Other relevant aspects include a general size reduction of main visceral organs, a severe damage to the heart structure and a negative inflammatory response.

These data highlight the putative morpho-functional changes that can be induced by exposure to PTX in vertebrates.

Toxin distribution, source organisms, and routes of exposure of PTX and its analogs are interrelated aspects that represent the main concern in respect to marine ecosystems and to human health issues. Since its discovery, PTX has been repeatedly documented in tropical Indo-Pacific seawaters, not only on the soft coral Palythoa spp. [3], but also in organisms associated to these zoanthid colonies [31]. Later, it was found that several species from the benthic dinoflagellates genus Ostreopsis also presents PTX-like compounds. Despite the uncertainty about the true origin for the production of these compounds discussed above (Section 1), Palythoa and Ostreopsis spp. are recognized as producers, and represent the major known sources for these toxins. In view of the fact that these two types of organisms were described from tropical and sub-tropical areas around the world, PTX and PTX-like compounds appeared to be confined to these biogeographic regions and thus largely overlooked. Nevertheless, possible new routes of exposure must be taken into account, such as the previously discussed cases of Palythoa corals present in aquariums. Moreover, Ostreopsis spp. have a broad geographical distribution than formerly thought [39], and are now a global concern since it can represent threats to human health whenever they develop and become prevalent [38]. Actually, while the presence of Ostreopsis spp. in tropical waters is well documented since the 1980s, the number of studies of these benthic dinoflagellates in temperate regions has increased substantially in the last few years [39]. These PTX-like producing microorganisms have been described for the China Sea [63], Pacific Ocean [55,63–65], Tasman Sea [65], Indic Ocean [9], Atlantic Ocean [50,66], Gulf of Mexico [67,68] and Mediterranean Sea [47,70–72,77]. The spread of toxin-producing Ostreopsis spp. to temperate regions may be due in part to ballast water of cargo ships and also to marginal changes in climate conditions, enough to induce bloom formation [39]. Dinoflagellate blooms are characterized by a brownish colored mucilaginous coverage of marine benthos.

Not surprisingly, due to the impact that these blooms may have, some studies [39,71] have been made in order to determine which are the optimal environmental conditions for the occurrence of this phenomena. Findings from these studies show that blooms are likely to be ephemeral and strongly related to seasonal patterns and to wave action [39,72]. The highest coverage occurs in hard substrata of shallow, sheltered areas such as harbors and estuaries [39,71], where the hydrodynamic conditions are gentle to moderate. In contrast to what occurs in other components of the microphytobenthos (e.g., diatoms, cyanobacteria), the mats formed are loosely attached to the substrata [71], and thus are easily resuspended in the water column by waves and mechanical action. This leads to cells dispersion, which may establish and develop in other areas whenever the conditions are favorable.

The high toxicity of PTX and its analogs has resulted several times in animal fatalities. Beside the few reports about human victims [38], animal deaths due to poisoning by PTX and analogs are also documented. These are the case of sea urchins poisoning after the occurrence of blooms of Ostreopsis spp., in Brazil and New Zealand [39,55,65], and the death of several pigs in Ryukyu Islands (Japan), after eating viscera of the filefish Alutera scripta [73]. In New Zealand, unexplained mouse deaths from regions where Ostreopsis spp. have been recorded raises the possibility that ostreocin-D or other PTX analog could have been the cause [59]. In Italy, massive mortalities of marine invertebrates and macroalgae [52], and visible impacts in sessile (cirripeds, bivalves, gastropods) and mobile (echinoderms, cephalopods, little fishes) epibenthos were also observed after summer blooms of Ostreopsis ovata [47,71].

Toxicological effects may arise directly from the toxin-producers itself (e.g., Palythoa spp. and Ostreopsis spp.) or indirectly via vectors that accumulate the toxin, which may be susceptible to the toxin or not. The entrance, diffusion and sequestration of PTX into the food chain have been recognized elsewhere [31]. For instance, in tropical and subtropical areas, PTX has been detected in several animals such as crabs [40], different species of fish [41,43] or organisms living in close association with Palythoa spp. The later include sponges, other soft corals, mussels, gorgonians and crustaceans or predators that feed on Palythoa spp. colonies such as the polychaete worm Hermodice carunculata, the starfish Acanthaster planci and the fish Chaetodon spp. [31]. These predators are examples of high tolerant organisms that store the toxin in its active form and may enable the distribution of PTX in other marine biota.

Filter-feeding invertebrates are also organisms in which PTX or analogs accumulation is a well known phenomenon, especially during harmful algal blooms [31,59,74,75]. Mussels, cockles, oysters, and scallops feed on toxic dinoflagellates, transferring them from the gills to digestive organs where the toxins accumulate [55,59]. PTX sequestration is also observed in sponges and mussels living near or among zoanthid colonies, which often exhibit higher PTX concentrations than Palythoa spp. [31]. It was showed that some shellfish are able to depurate the toxins at fairly rapid rates, while others can retain the toxins in its active form for months while continuing to bioaccumulate [31,76]. Moreover, the mode of accumulation can also differ in different bivalves. In an ecotoxicological study [55], the scallop Pecten novaezealandiae and the Pacific oyster Crassostrea gigas were fed with cells of a same culture of Ostreopsis siamensis (containing 0.3 pg palytoxin equivalents/cell), in numbers representative of a dense bloom. While the oysters contained detectable amounts of toxin in hepatopancreas, muscle, and roe, the scallops showed higher concentrations but only in the hepatopancreas. Unlikely these two shellfish, the green-lipped mussel Perna canaliculus, which was tested in the same study, did not present PTX-like substances in any of evaluated parts. Thus, it appears not to accumulate the toxin.

Notwithstanding the capability of shellfish to uptake PTX-like compounds after feeding on toxic Ostreopsis spp., no cases of human intoxications through consumption of contaminated shellfish have been noticed. The only suspected occurrence is a human shellfish poisoning that occurred with wild mussels collected in Tasmania [52]. As stated before, accumulated toxins may not be active. For instance, in the aforementioned Pacific oysters, no bioactivity was found by mouse bioassay [59], using extracts from individuals that fed on a sole diet of O. siamensis cells.

Concerning the effects that the toxin accumulation may have in such bivalves, little information is available. In one of the few exceptions [60,] it was found that PTX seriously increases phagocytosis (Table 2), a central process to the cell mediated immune response in invertebrates. The study was conducted on the mussel Mytilus galloprovincialis, a shellfish used for human consumption and widely used as a model. Nevertheless, it still lacks knowledge about the effects that this toxin can exert on mollusks in the long-term and in natural conditions.

Other marine organisms known to be affected by the toxins are sea urchins, an ecologically important herbivore. In southeastern Brazil, Equinometra lucunter individuals showed alterations on their exoskeleton, accompanied by high mortality. It did coincide with an outbreak of a benthic dinoflagellate, previously reported as belonging to Prorocentrum sp. but after confirmed to be Ostreopsis ovata [52]. Furthermore, algal extracts from this occurrence were used to carry out toxicity tests, namely the Artemia salina assay (Table 2). This marine invertebrate displayed 100% death after being exposed to an extract of no more than 125 O. ovata cells [66]. Chemical analyses confirmed the presence of a PTX-like compound in samples. A further toxicological test with O. ovata extracts from the Mediterranean Sea revealed a similar association between Artemia salina mortality after 24 h and the number of algal cells [52,77]. In this case, 542 to 906 cells/mL were sufficient to cause 65% to 100% death, respectively. Recently, it was also observed that summer blooms of O. siamensis killed sea urchins in New Zealand [57]. Another study from this austral country corroborates these findings [39]. A strong negative effect between the health of the sea urchin Evechinus chloroticus and the benthic cover of O. siamensis was demonstrated. The echinoderm densities declined by 56–60% at bloom sites over the study period (Table 2).