3.5 Toxicity of DSP toxins

3.5.1 Mechanism of toxicity

The discovery that OA acid caused long-lasting contraction of smooth muscle from human arteries was the first clue to elucidation of the mechanism of action of DSP toxins. Since smooth muscle contraction is activated by a sub-unit of myosin, it was supposed that the effect of OA was due to inhibition of myosin light chain phosphatase. Thereafter, OA was shown to be a potent inhibitor of the serine/threonine phosphatases PP1 and PP2A; PP2A is about 200 times more strongly inhibited than PP1. Protein phosphatases are a critical group of enzymes closely linked with many crucial metabolic processes within a cell. Phosphorylation and dephosphorylation of proteins is one of the major regulatory processes in eukaryotic cells. Processes as diverse as metabolism, membrane transport and secretion, contractility, cell division and others are regulated by these versatile processes. It is indicated that phosphatases, which are sensitive to OA, like PP1 and PP2A, are involved in entry into mitosis. It is suggested that diarrhoea in humans is caused by hyperphosphorylation of proteins that control sodium secretion by intestinal cells or by increased phosphorylation of cytoskeletal or junctional moieties that regulate solute permeability, resulting in passive loss of fluids (Van Egmond et al., 1993; Hallegraeff et al., 1995). Extensive structure-activity studies measuring the inhibition of protein phosphatase activity indicated that a free carboxyl group in the DSP molecule is essential for activity, since methyl and diol esters did not show phosphatase inhibition. However, the amide and reduced carboxyl (okadaol) derivatives are about half as active as OA, as are the naturally occurring DTX3 compounds (Hallegraeff et al., 1995).

3.5.2 Pharmacokinetics

studies in laboratory animals

studies in mice with okadaic acid (OA)

Adult Swiss mice received a single oral dose by gavage of 50 or 90 mg [³H]OA/kg bw dissolved in 0.2 ml sterile water and methanol (50:50 [v/v]). Urine and faeces were collected over a 24 hour period and thereafter the animals were killed. At 50 mg/kg bw no clinical signs of toxicity were seen, whereas at 90 mg/kg bw diarrhoea was observed from eight hours onwards. No mortality occurred. Radioactivity was determined in the brain, lung, spleen, heart, liver and gall bladder, kidney, stomach, intestine tissue, intestine content, skin, blood, muscle, urine and faeces, and OA was analysed with LC (fluorescence detection) after derivatization with 9-anthryldiazomethane (ADAM). Both methods gave similar results indicating that OA was not very much metabolised. OA was absorbed from gastrointestinal tract as it was found mostly in intestinal tissue and contents (49.2 percent of the dose) and urine (11.6 percent) after 24 hours. The high concentrations in intestinal tissue and contents after 24 hours demonstrated slow elimination of OA. OA was found in all tissues. The total amount of OA in organs at 50 mg/kg bw was low compared to the amount excreted in urine and faeces (11.6 and 6.6 percent of the dose, respectively) and by far lower than the amount in intestinal tissue plus contents. As the dose increased from 50 to 90 µg/kg bw, concentrations of OA in intestinal contents and faeces increased proportionally. The increase of OA in intestinal tissue at the higher dose correlated well with the diarrhoea observed. The fact that OA was present in liver and bile and all organs including skin and also fluids and the fact that concentrations in intestinal content were approximately two to seven fold higher than in faeces after 24 hours, confirmed that enterohepatic circulation occurred (Matias and Creppy, 1996a). This study also demonstrated that in acute OA intoxication, the concentration in intestinal tissue reaches cytotoxic concentrations in accordance with the diarrhoea seen (Matias et al., 1999a). In recent studies in mice using the anti-OA antibody, OA was detected in lung, liver, heart, kidney and small and large intestines just five minutes after oral administration. OA was detected in liver and blood vessels for two weeks after dosing and in the intestines for four weeks (EU/SANCO, 2001).

Male and female adult Swiss mice received a single intramuscular (i.m.) injection with 25 mg [³H]OA/kg bw dissolved in 0.1 ml sterile water and methanol (50:50 [v/v]). OA was detected in bile and intestinal contents one hour after injection. Its elimination pattern showed biliary excretion and enterohepatic circulation. Administration of cholestyramine, which prevents enterohepatic circulation, changed the cyclic elimination profile of OA (Matias and Creppy, 1996a).

observations in humans

No data

3.5.3 Toxicity to laboratory animals

acute toxicity

studies with mussel extracts

Toxicity of DSP toxins is usually measured by means of i.p. injection of extracts from contaminated mussels in mice. Although this is a crude comparison, it forms the basis of the most widely used screening and quality control methods.

When DSP toxins are given by the oral route, the lethal dose is 16 times higher than the i.p. dose but the symptoms are the same (Yasumoto et al., 1978).

Three to five mice (4 to 5 days old) receiving once orally by gavage 0, 0.05, 0.1, 0.2, 0.4 or 0.8 MU DSP toxins as 0.1 ml of a crude extract from contaminated scallops containing a drop of 1 percent Evans Blue solution per ml animals were kept for four hours at 25 °C and killed. The whole intestine was removed and fluid accumulation was determined as the ratio of intestinal weight to that of remaining body weight (FA ratio). FA ratios in control, 0.05, 0.1, 0.2, and 0.4 MU groups were 0.072, 0.073, 0.09, 0.108 and 0.112, respectively. At 0.8 MU mortality occurred. The diarrhoeagenicity (as FA ratio) of the components of the crude mixture (OA, DTX1, DTX3, PTX1) in the suckling mice was as follows: OA and DTX1 had the same potency; diarrhoea was seen at doses ³0.1 MU, with DTX3 diarrhoea was seen at doses ³0.05 MU and PTX1 did not show diarrhoeagenicity at the doses tested (0.025 to 0.4 MU) (Hamano et al., 1986).

oral studies in mice with OA

After oral administration of 75 µg OA/kg bw to adult mice, the weight of the small intestines increased slightly within one hour (by fluid accumulation) but that of the liver decreased slightly. The lowest observed adverse effect level (LOAEL) in mice by acute oral administration was deduced to be 75 µg/kg bw (EU/SANCO, 2001).

Within one hour after oral administration of OA to mice, severe mucosal injuries in the intestine were seen. The injuries could be divided into three consecutive stages (Matias et al., 1999a):

• extravasion of serum into the lamina propria of villi;

• degeneration of absorptive epithelium of iliac villi;

• desquamation of the degenerated epithelium from the lamina propria.

Rat small intestine was stated to be the most sensitive and reproducible organ for studies of the diarrhoeic effects of marine toxins. When OA was injected in ligated loops from the middle duodenum of male rats (200 g) the following changes were seen within 15 minutes. Enterocytes at the top of the villi became swollen and subsequently detached from the basal membrane. Globet cells were not affected at the doses applied (1-5 mg OA). After 60 to 90 minutes, most of the enterocytes of the villi were shed into the lumen and large parts of the flattened villi were covered by globet cells. The degree of the damage was dose-dependent: 3 mg OA affected only the top of the villi, while 5 mg led to collapse of the villous architecture. Intravenous injection induced similar but less extensive changes (Van Apeldoorn et al., 1998).

oral studies in mice with DTXs

At oral doses of 100, 200, 300 or 400 µg DTX1 to mice 1/5, 0/5, 2/4 and 3/4 animals died respectively (Ogino et al., 1997).

oral studies in mice with PTXs

When oral doses of 25, 100, 200, 300 or 400 µg PTX2 were given to mice 1/4, 0/4, 1/5, 2/5 and 1/4 animals died respectively. This study did not show a dose-response. The oral toxicity of PTX2 is comparable to its i.p. toxicity (i.p. lethal dose in mice of PTX2 is 260 µg/kg bw) (Ogino et al., 1997).

At oral doses of 1.0, 2.0 and 2.5 mg PTX2/kg bw to mice, diarrhoea was seen in 1/5, 2/5 and 2/5 animals respectively. At 0.25 mg/kg bw, no diarrhoea was observed but the small intestine was swollen and filled with fluid (EU/SANCO 2001). Oral doses of 0.25 to 2.0 mg PTX2/kg bw in mice caused histopathological changes in liver as well as stomach and whole intestines. The oral dose of 0.25 mg PTX2/kg bw in mice is considered to be a LOAEL (EU/SANCO, 2001).

oral studies in mice with YTXs

At a maximum oral dose of 1.0 mg yessotoxin (YTX)/kg bw no lethality in mice was observed by Ogino et al. (1997). At this dose-level the mice gained weight during three days observation time (Yasumoto and Satake, 1998). Aune et al. (2002) reported that oral doses up to and including 10.0 mg YTX/kg bw did not cause mortality of female mice (not fasted). Microscopy revealed only moderate changes in the heart (slight intercellular oedema) at 10 and 7.5 mg/kg bw. At 5, 2.5 and 1 mg/kg bw no changes in the heart were seen by light microscopy. Ultramicroscopy revealed swelling of heart muscle cells leading to separation of the organelles. Effects were more pronounced close to the capillaries. These effects were dose-dependent and were only very slight at 2.5 mg/kg bw, which was the lowest dose examined by ultramicroscopy.

When YTX was given orally by gavage to four-day old suckling ddY mice at dose levels of 0.1, 0.2 and 0.4 mg/mouse as a 1 percent suspension in Tween 60 solution, no intestinal fluid accumulation was seen after four hours, whereas this phenomenon was seen at all dose-levels of OA or DTX1 (Ogino et al., 1997).

intraperitoneal studies

Thirty minutes to several hours after i.p. injection of DSP toxins in mice, inactivation and general weakness were seen and at sufficiently high concentrations mice died between one and a half and 47 hours. Concerning the effects reported after oral administration, it is of interest to compare the intraperitoneal (i.p.) toxicity of the different toxins in the DSP complex (see Table 3.1).

Table 3.1 Acute toxicity (lethal dose) of DSP toxins after i.p. injection in mice

Source: Van Egmond et al., 1993 and Ritchie, 1993 (except as indicated)

* presumed from the toxicity of PTX1

** fasted suckling male mice; at 80 mg/kg bw 1/3 mice died, at 100 mg/kg bw all 3 mice died (Ogino et al., 1997)

*** fasted male mice (Aune et al., 2002)

^# data indicate damage to the heart

^## Ogino et al. (1997)

^@ data indicate damage to the liver

Mice receiving an i.p. injection with 160 mg DTX1/kg bw died within 24 hours while suffering from constant diarrhoea (Van Egmond et al., 1993).

After intraperitoneal injections of 50-500 mg DTX1/kg bw into suckling mice (7-10 g) duodenum and upper portion of small intestine became distended and contained mucoid, but not bloody, fluid. Villous and submucosal vessels were severely congested at the higher concentrations. No discernible changes in organs and tissues other than the intestines were seen. At ultrastructural level, three sequential stages of changes of intestinal villi were observed as was seen after oral administration (see preceding page: oral studies in mice with OA) (Van Apeldoorn et al., 1998).

Marked dilation or destruction of Golgi apparatus suggests that DTX1 may directly attack this organelle (Van Apeldoorn et al., 1998) OA and DTX1 induce also liver damage in mice and rats after oral as well as i.p. administration. The liver changes were expressed as degeneration of endothelial lining cells at the sinusoid. In addition, dissociation of ribosomes from the rough endoplasmic reticulum and autophagic vacuoles were seen in hepatocytes in midzone of hepatic lobuli. Haemorrhage in subcapsular region of the liver was observed. Furthermore OA, DTX1 and DTX3 induced damage to the epithelium in the small intestine after both oral and i.p. dosing (Van Apeldoorn et al., 1998).

PTX1 did not cause pathological findings in small and large intestine in suckling mice after i.p. injection but marked congestion of the liver and finely granulated surfaces of the liver were seen. Thirty to sixty minutes after i.p. injection of 1 000 mg/kg bw multiple vacuoles appeared around the periportal region of the hepatic lobules. Similar features were seen in livers from mice two hours after i.p. treatment with 500 or 700 mg/kg bw. Electron microscopy confirmed these light microscopic observations: Several portions of the microvilli of the hepatocytes became flat and the plasma membrane was invaginated into the cytoplasm. Within 30 minutes, vacuoles had increased in size and most of the cellular organelles had become compressed. Within 24 hours, almost all hepatocytes contained numerous vacuoles and granules and had become necrotic. Mice given i.p. 150-200 mg/kg bw showed only slight hepatic injuries after one hour (Van Apeldoorn et al., 1998).

YTX kills suckling mice at an i.p. dose of 100 mg/kg bw. Even at the lethal dose no intestinal fluid accumulation was seen in the suckling mice. Five week old male mice (bw 23-25 g) showed after i.p. doses above 300 mg YTX/kg bw normal behaviour for the first hours, but then suddenly dyspnea occurred and the mice died. No discernible changes in liver, pancreas, lungs, adrenal glands, kidneys, spleen or thymus were seen. Mice given i.p. 500 mg YTX/kg bw showed severe cardiac damage. Endothelial lining cells of the capillaries in the left ventricle were swollen and degenerated. Mice treated orally with 500 mg YTX/kg bw did not show changes (Van Apeldoorn et al., 1998).

Aune et al. (2002) gave i.p. injections of 0.1-1.0 mg YTX/kg bw in 1% Tween 60 to groups of three female white mice. At 1.0 mg/kg bw all three mice died and at 0.75 mg/kg bw two out of three mice died. Light microscopy revealed effects in the myocardium (slight intercellular oedema) at 0.75 and 1.0 mg/kg bw. Ultramicroscopy showed at 1.0 mg/kg bw swelling of myocardial muscle cells, separation of organelles, most pronounced near capillaries (no other dose-levels were examined by ultramicroscopy). The lethal effects seen at 0.75 mg/kg bw and higher, indicated a lower i.p. acute toxicity than reported in Table 3.1 above. Some of the reasons might be that in this study non-fasted female mice were used, whereas in the other studies fasted male mice of a different strain were used.

Male mice given 300 mg desulphated YTX (chemically prepared)/kg bw survived 48 hours. Desulphated YTX caused only slight deposition of fat droplets in the heart muscle. On the other hand, effects in liver and pancreas were seen. Within 12 hours after an i.p. dose of 300 mg/kg bw livers were pale and swollen. Fine fat droplets were found in all hepatocytes in the lobuli. Almost all mitochondria were slightly swollen and showed reduced electron density. Pancreatic acinar cells also showed degeneration. Disarrangement of the configuration of the rough endoplasmic reticulum was prominent within six hours. Mice treated orally with 500 mg desulphated YTX/kg bw developed fatty degeneration of the liver (Van Apeldoorn et al., 1998).

repeated administration

No data

reproduction/teratogenicity studies

Studies in pregnant mice demonstrated the transplacental passage of [³H]-OA by measuring the radio-labelled compound 24 hours after oral administration of 50 mg/kg bw (dissolved in sterile water and methanol 50:50) at day 11 of gestation. Foetal tissue contained more OA than maternal liver or kidney: 5.60 percent of the administered label compared to 1.90 and 2.55 percent respectively as measured by scintillation counting and LC with fluorescent detection after derivatization with ADAM (Matias and Creppy, 1996b).

mutagenic activity of okadaic acid

OA did not induce mutations in Salmonella typhimurium TA 98 or TA 100 in the absence as well as the presence of a metabolic activation system, but it was strongly mutagenic in Chinese hamster lung cells without metabolic activation (mutagenic activity was comparable to that of 2-amino-N⁶-hydroxyadenine, one of the strongest known mutagens). Diphtheria toxin resistance (DT^r) was used as marker of mutagenesis. Results indicated that OA increased the number of DT^r cells by induction of a mutation from the DT^r phenotype, and not by selection of spontaneously induced DT^r cells. The authors suggested that induction of DT^r mutation is not due to OA-DNA adduct formation, but probably operates via modification of the phosphorylation state of proteins involved in DNA replication or repair (Aune and Yndestad, 1993).

Using the ³²P-postlabelling method, DNA adduct formation was seen in two cell lines (BHK21 C13 fibroblasts and HESV keratinocytes) after treatment with OA for 24 hours (doses 0.01-5 nM). Low doses did not show adduct formation. Intermediate doses have given the most important number of adducts and with higher doses, the number of adducts decreased dose-dependently. Nineteen adducts were observed with BHK21 C13 cells and 15 with HESV cells. Ten adducts were similar in the two strains, while nine were specific of BHK21 C13 cell line, and five of HESV one (Fessard et al., 1996).

tumour promoting activity of okadaic acid and dinophysistoxin-1

OA and DTX1 are tumour promoters in two-stage experiments on mouse skin. OA and DTX1 do not activate protein kinase C as do the phorbol esters but inhibit the activity of protein phosphatase 1 and 2A, resulting in rapid accumulation of phosphorylated proteins. The effects of OA on protein phosphorylation in cellular systems emphasize the strong tumour-suppressing effect which PP1 and PP2A must have in normal cells. OA and DTX1 distinguish themselves from phorbol ester promoters by the fact that they do not bind to the same receptors. OA and DTX1 bind to a particulate fraction of the mouse skin. The binding sites of OA are also present in stomach, small intestine and colon, as well as in other tissues (Fujiki et al., 1988). OA and DTX1 and also PTX2 induce ornithine decarboxylase (ODC) in mouse skin (Fujiki et al., 1989). Furthermore OA induced ODC in rat stomach and enhanced the development of neoplastic changes (adenomatous hyperplasia and adenocarcinomas) in the rat glandular stomach after initiation with N-methyl-N’-nitro-N-nitrosoguanidine (Suganuma et al., 1992).

OA has been shown to promote morphological transformation of carcinogen (3-methyl-cholanthrene)-initiated BALB/3T3 cells. It was demonstrated that OA induced morphological transformation of BALB/3T3 cells also in the absence of an initiator (Sheu et al., 1995).

Induction of DNA adducts by okadaic acid was shown in Baby Hamster Kidney (BHK) cells, Human (HESV) keratinocytes and human bronchial epithelial cells. Also the induction of DNA adducts in zebra fish embryos was demonstrated. It was noted that the DNA adduct formation increased with the dose at lower and intermediate (non cytotoxic) concentrations whereas higher concentrations caused toxic stress (Huynh et al., 1998)

immunotoxicity of okadaic acid

The effect of OA on peripheral blood monocytes of humans in vitro by means of effects on the interleukin-1 (IL-1) synthesis was studied. OA induced a marked depression of IL-1 production in the monocytes at concentrations of 0.1-1.0 mg/ml. At higher concentrations OA killed the cells. The suppressive effect of OA on IL-1 is readily reversed by specific monoclonal anti-OA. The mode of action of this effect of OA is unknown (Aune and Yndestad, 1993).

in vitro toxicity

OA, DTX1, PTX1 and YTX were studied for their possible toxicity towards fresh rat hepatocytes by means of light and electron microscopy (Van Apeldoorn, et al., 1998). OA was the most toxic. At 1 mg/ml blebs on the cell surface were seen. At increasing concentration blebs increased in size and number. At high concentrations the cells lost their circular appearance and became irregular. DTX1 showed at 2.5 mg/ml effects similar to those of OA, although to a lower degree. PTX1 gave quite different results. Morphological changes manifested themselves as small grooves on the cell surface and vacuoles in the cytoplasm in a dose-dependent pattern, starting at 7.5 mg/ml. Electron microscopy revealed invagination of the cell membrane and development of vacuoles.YTX was far less toxic. Between 25 and 50 mg/ml very tiny blebs on the cell surface were observed without changing the general spheric appearance of the cells. None of the four purified DSP toxins studied caused enzyme (lactate dehydrogenase) leakage from the cells.

Protein and DNA synthesis in Vero cells (from monkey kidney) were both inhibited by OA in a concentration-dependent manner (IC₅₀ 3.3 x 10^-8 and 5.3 x 10^-8 M, respectively). RNA synthesis was inhibited with an IC₅₀ of 8.2 x 10^-8 M. The time lag before DNA and RNA synthesis inhibition occurred was longer (eight hours) than the time lag before protein synthesis occurred (four hours) indicating that protein synthesis is probably the main target and the first of OA’s cytotoxic effect (Matias et al., 1996).

In a later study (Matias et al., 1999b), the effect of OA on the production of oxygen reactive radicals as possible inducers of impairment of protein synthesis was studied in the presence and the absence of oxygen radical scavengers (SOD+catalase, vitamin E and/or vitamin C). Lipid peroxidation appeared to be a precocious marker of OA exposure. The radical scavengers (partially) prevented the lipid peroxidation, but the inhibition of protein synthesis induced by OA was not reduced to the same level. This indicates that a more specific mechanism might be responsible for inhibition of protein synthesis.

In the cell free rabbit reticulocyte lysate specific mRNA is translated into globin. This was used to ensure that protein synthesis is a direct target of okadaic acid. Indeed in this system protein synthesis was also inhibited by OA in a concentration-dependent manner (Matias et al., 1996).

Matias and Creppy (1998) studied the effect of OA on the five nucleosides (deoxycytosine, 5-methyldeoxycytosine, desoxythymidine, deoxyguanidine and deoxyadenine) in the DNA of Vero cells. At 7.5 ng OA/ml no significant inhibition of DNA synthesis was seen but hypermethylation of DNA was induced. The level of 5-methyldeoxycytosine increased from 3.8 to 7.8 percent, indicating possible interference with DNA regulation, replication and expression. Higher levels of OA inhibited DNA synthesis but failed to increase the rate of DNA methylation. Since OA is involved in tumour production, the most threatening effects are those possibly connected with DNA modification and/or regulation of gene expression, such as the rate of methylation. In other terms, the risks for humans and animals may be more related to repeated exposure to low OA concentrations in seafood that could assault the DNA several times within a life span.

Primary cultures of liver cells of 11 days old chick embryos were exposed to PTX1 and the effects were examined by fluorescence microscopy. PTX1 reduced the cell size. Microtubules were reduced in number and lost their radial arrangement. Stress fibres (actin filament bundles) disappeared and actin became accumulated at the cellular peripheries. At exposure to concentrations £0.5 mg/ml for less than four hours these effects were reversible within 24 hours (Zhou et al., 1994).

The effect of OA on cultured human intestinal epithelial T₈₄ cell monolayers was studied by measuring electrophysiological parameters, lactate dehydrogenase release, and ²²Na⁺ and [³H] mannitol flux rates. Protein phosphorylation studies were carried out to identify potentially involved proteins. OA did not directly stimulate Cl^- secretion but increased the paracellular permeability of intestinal epithelia. This alteration may contribute to the diarrhoea of DSP poisoning (Tripuraneni et al., 1997).

YTX appeared to have an effect on the cytosolic Ca²⁺ levels of freshly isolated human lymphocytes. YTX modulated intracellular Ca²⁺ of human lymphocytes by producing a slight non-capacitative calcium entry and inhibiting the Ca²⁺ entry produced by emptying of internal calcium stores. OA did not cause these effects. The authors suggested interaction of YTX with plasma membrane calcium channels (De la Rosa et al., 2001).

3.5.4 Toxicity to humans

Shellfish containing more than 2 mg OA/g hepatopancreas and/or more than 1.8 mg DTX1/g of hepatopancreas are considered unfit for human consumption (Hallegraeff, 1995). The predominant symptoms in humans include diarrhoea, nausea, vomiting and abdominal pain. The onset of symptoms, which are never lethal, ranged from 30 minutes to a few hours after ingestion of the toxic shellfish, with complete recovery within three days. The intensity of the symptoms in humans depends upon the amount of toxin ingested. Hospitalization is usually not needed. Among the DSP toxins, OA, DTX1 and DTX3 are the most important in causing diarrhoea in humans (Aune and Yndestad, 1993). DTX2 was reported to be the predominant diarrhoeic DSP toxin in Ireland during a prolonged DSP episode (Carmody et al., 1996). Epidemiological data from Japan (1976-1977) indicated that as little as 12 MU was enough to induce a mild form of poisoning in humans (EU/SANCO, 2001). MU was defined as the amount of toxin (later defined as DTX1 in the Japanese study) killing a mouse by i.p. injection within 24 hours and 12 MU corresponded to 43.2 µg, which can be considered as a LOAEL for DTX1 (EU/SANCO, 2001). However Yasumoto et al. (1985) reported that the minimum dose of DTX1 for the induction of toxic symptoms in human adults was 32 µg. Fernandez and Cembella (1995) reported that 1 MU corresponded to approximately 3.2 µg DTX1 and 4 µg OA which means that the minimum dose for toxic effects in humans is 38.4 and 48 µg for DTX1 and OA, respectively. The probable human health problems associated with tumour-promoting, mutagenic and immunosuppressive effects shown in animals and experimental systems by OA and DTX1 cannot yet be quantified.

Concerning two other chemical groups of the DSP complex, the PTXs and YTXs, the situation is unsatisfactory. PTXs have a low diarrhoeic potential and YTXs do not induce diarrhoea in rodents, but both groups of toxins are lethal to mice at i.p. injection and they exert toxicity to liver and heart, respectively, in rodents. It is unclear whether PTXs or YTXs pose a health threat to consumers of contaminated mussels (Aune and Yndestad, 1993). In a pipi DSP event (56 cases of hospitalisation) in New South Wales, Australia in December 1997 (ANZFA, 2001) PTX-2 seco acids may have contributed to the gastrointestinal symptoms, vomiting or diarrhoea in humans (Quilliam et al., 2000 in Aune, 2001). Burgess and Shaw (2001) reported that the patients consumed approximately 500 g of pipis containing 300 µg PTX-2SA/kg (~150 µg PTX-2SA/person~2.5 µg/kg bw for a 60 kg weighing person).

During a DSP episode in Norway in 1984, a few people were hospitalized with symptoms of severe exhaustion and cramps, in addition to the usual DSP symptoms. After intravenous injection of an electrolyte mixture, the patients recovered within a few days (Aune and Yndestad, 1993). In a recent incident in Norway, about 70 people were served blue mussels during the opening ceremony of a new mussel farm. Among the guests, 54 percent were intoxicated with typical DSP symptoms. DSP toxin levels in the left-over mussels were found to be around 55 to 56 µg OA eq/100 g mussel meat (Aune, 2001).

3.5.5 Toxicity to aquatic organisms

OA inhibited the growth of a variety of non-DSP producing microalgae at micromolar concentrations. The effects of DTX1 on microalgal growth were found to be equivalent to those of OA, and the effects of a mixture of both toxins were simply additive. The growth of the DSP toxin producing dinoflagellate P. lima was not affected (Windust et al., 1996).

Concentrations of (3 mM of both OA-diol ester and OA inhibited almost completely the growth of the diatom Thalassiosira weissflogii. (EC₅₀ 2.2 and 1.0 mM for OA-diol and OA, respectively). This result is in contradiction with the accepted idea that only the free acid toxins, such as OA and DTX1, are potent phosphatase inhibitors. Substantially higher concentrations of DTX4 were required to detect any effect on the growth. The OA-diol ester was shown to be partially hydrolysed to OA (7 percent hydrolysed, 20 percent unchanged OA-diol ester and 73 percent was unidentified). This phenomenon does suggest that cells exposed to inactive DSP toxin esters could metabolically activate them. In an additional experiment both DTX4 and OA-diol ester were hydrolysed (2.0 and 2.7 percent, respectively, within five days) to OA spontaneously and not by mediation of the presence of T. weissflogii.

The inactive DTX4 can apparently be hydrolysed through uncharged, lipophilic intermediates ultimately to yield the active, free acid toxin OA (Windust et al., 1997).

Ichthyotoxicity of yessotoxin (YTX) and bisdesulfated YTX (dsYTX) was studied using killifish (Oryzias latipes). YTX was diluted in 0.1 ml methanol and the solution was diluted with water to 50 ml to prepare a 1.0 or 0.5 mg/L test solution. A killifish was placed in a beaker with test solution and observed for 24 hours. The assay was run in triplicate. Similarly, dsYTX was tested at 0.5 mg/L only. None of the fish exposed to 1 or 0.5 mg/L YTX died within 24 hours. Three fish exposed to 0.5 mg/L dsYTX died after six hours (Ogino et al., 1997).