2.5 Toxicity of PSP toxins

2.5.1 Mechanism of action

The pharmacological action of the PSP toxins strongly resembles that of TTX. Due to the almost identical action of STX and TTX, it was assumed that both molecules had the same interaction with the receptor. Much attention has been given to the elucidation of the mechanism via which the blockade of the voltage-gated sodium channel is achieved as STX and TTX are the only agents which block this channel in a selective manner and with high affinity.

The voltage-gated sodium channel is a protein of approximately 250 000 Da, which traverses the plasma membrane of many excitable cells and is characterized by uniform conduction, potential dependency and ion selectivity. Among these are all mammalian nerves, skeletal muscle fibres and most cardiac muscle fibres. Upon appropriate depolarization of the cell, a conformational change occurs in the sodium channel molecule such that an aqueous path opens to permit movement of Na⁺ from the extra-cellular phase into the cell under the existing electrochemical driving forces. The inward sodium current is responsible for the rising phase of the action potential. Voltage-gated potassium channels are also present in the membrane, and when open, they permit outward passage of intracellular K⁺ and consequent repolarization. STX and several other PSP toxins block the voltage-gated sodium channel with great potency, thus slowing or abolishing the propagation of the action potential. However, they leave the potassium channel unaffected.

The 7,8,9-guanidine function has been identified as being involved in the channel blockade. The C12-OH (as hydrated ketone) is important, whereas the carbamoyl side chain contributes but is not vital to channel blockade. Several hydrogen bonds between the toxin molecule and the binding side add to the binding energy. There is a general agreement among the investigators on the kinetic aspects of the toxin binding. The averaged blocking time of the channel is not dependent on the toxin concentration, but on the dissociation velocity. The lifetime of the open channel, however, is reversibly correlated with the toxin concentration and depends on the association constant (Mons et al., 1998).

2.5.2 Pharmacokinetics

studies in laboratory animals

rats

A single intravenous (i.v.) dose of radiolabeled [³H]-saxitoxinol (STXOL), an analogue of STX, given to male Wistar rats, was rapidly distributed to various tissues including the central nervous system (CNS). The rats had excreted 40 percent of the dose in urine within two hours and 80 percent after 48 hours. Half-lifetime (t_1/2) in plasma is 29.3 minutes. Radioactivity reached a maximum in most tissues, including brain, eight hours after dosing. In liver and gastro-intestinal tract (G.I.T.), radioactivity was low during early phase after dosing, and was highest 24 hours after dosing, suggesting an alternate route of elimination and excretion. STXOL was metabolized in various tissues. Ten minutes after dosing, 19 percent of activity in kidney extract was associated with unidentified metabolites, for lungs it was 28.5 percent and for heart it was 41.8 percent. After 48 hours, 75 percent of activity in these tissues was associated with unidentified metabolites. Minimum biotransformation was found in muscles (14.4 percent) 48 hours after dosing. In CNS 10 minutes after dosing, 31.8 percent of activity was associated with unidentified metabolites in brain and 37.4 percent in spinal cord. After 48 hours in the brain, 76 percent of activity was associated with unidentified metabolites. No STXOL metabolites were detected in urine (Naseem, 1996).

Rapid excretion in urine was observed in rats after i.v. administration of radioactively labelled STX at a sub-lethal dose (ca. 2 µg/kg). No radioactivity was detectable in faeces at any time. Four hours after injection, approximately 19 percent of the STX dose was excreted in urine. By 24 hours, approximately 58 percent of the administered dose was excreted. Average total urinary excretion of administered STX was approximately 68 percent for the full study period. No radioactivity was found in the faeces. The authors concluded that these results demonstrate that small quantities of non-metabolized STX can be detected in rat urine up to 144 hours after i.v. administration (Aune, 2001).

cats

Fourteen adult male cats (bw 2.5-5 kg) were anesthetized and permanently coupled to artificial ventilation. Then the cats received a single i.v. injection with 2.7 or 10 mg STX/kg bw. During four hours after injection, cardiovascular parameters such as blood pressure and electrocardiograms were recorded and urine and blood samples were collected. Then the cats were killed and STX levels in brain, liver, spleen, bile and medulla oblongata were measured. The low dose did not cause changes in hemodynamic parameters. However, the high dose drastically reduced blood pressure, caused myocardial failure and finally cardiac arrest. Administration of dobutamine (2.5 mg/kg per minute) restored hemodynamics and allowed the cats to overcome the shock. STX was excreted only by urine; within four hours, 25 percent of the administered dose at 2.7 mg/kg and 10 percent of the administered dose at 10 mg/kg. Renal clearance at the high dose was 0.81 ml/min/kg and at the low dose 3.99 ml/min/kg. These data suggest that STX excretion mainly involves glomerular filtration. No PSP toxins other than STX were detected in urine, blood or tissues analysed, indicating that no biotransformation had occurred. STX was detected in intensely irrigated organs such as liver and spleen but also in the central nervous system (brain [1.81 ng/g of wet tissue at the high dose] and medulla oblongata [2.5 ng/g of wet tissue at the high dose]) showing that STX was capable of crossing the blood-brain barrier (Andrinolo et al., 1999a).

observations in humans

Clinicians have observed that, if patients survive PSP for 24 hours either with or without mechanical ventilation, chances for a rapid and full recovery are excellent. Such observations suggest that toxin(s) responsible for PSP either undergo rapid excretion, metabolism or both. In spite of the fact that most PSP toxins are positively charged, they are readily absorbed through the gastrointestinal mucosa. Depending on the severity of poisoning, the symptoms vary somewhat. The determinants of the severity are the specific toxicity of the PSP toxin in the ingested food, the amount of food ingested, and the rate of elimination of the PSP toxin(s) from the body. If the amount of toxic food is high enough, the first symptoms occur within a few minutes (Mons et al., 1998).

In patients from four outbreaks of PSP in Alaska during May and June 1994, PSP toxin levels of 2.8-47 nM and 65-372 nM in serum and urine, respectively, were detected at acute illness and after acute symptom resolution. Severe hypertension was observed in the patients although only nanomolar serum levels were detected. The PSP toxin profile differed between mussels and human biological specimens, suggesting human metabolism had occurred. Clearance of PSP toxins from serum was evident within 24 hours and urine was identified as a major route of excretion (Gessner et al., 1997).

2.5.3 Toxicity to laboratory animals

acute toxicity

The toxicity of the PSP toxins is almost always expressed as STX or STX equivalents. The sulfocarbamoyl compounds are considerably less toxic than the other groups of PSP toxins. However, they might be converted to the more toxic carbamates under acidic conditions (Aune, 2001). The mouse is very sensitive to the PSP toxins when compared to species such as fish, amphibians, reptiles and animals of a low order. The LD₅₀ values for the different routes of administration are shown in Table 2.2. The oral LD₅₀ values for other species than the mouse are shown in Table 2.3.

Table 2.2 Acute toxicity of STX in mice (Mons et al., 1998)

Table 2.3 Oral LD50 values of STX in various species (Mons et al., 1998)

Aside from mouse lethality bioassays which are used to determine the relative potency of all analogues compared to that of STX (see Table 2.4), the full biological actions have been studied for only 50 percent of the natural analogues. However, from those that have been studied, the cellular mechanism of action seems to be basically the same. The N-sulfocarbamoyl compounds are appreciably less toxic than their counterparts of the carbamoyl series but they are readily converted to the corresponding carbamoyl compounds under acidic conditions with increases in toxicity of up to 40-fold. Such conversion has some potential clinical and public health significance because weakly toxic shellfish containing N-sulfocarbamoyl toxins could cause disproportional severe poisoning once ingested. Experimentally, however, it has been found that the conversion occurs in artificial gastric juice of the mouse and rat at a pH of 1.1, but not in genuine gastric juice remaining at a buffered pH of 2.2 (Mons et al., 1998).

Table 2.4 Relative toxicity of PSP toxins in the mouse bioassay

a = a/b epimeric mixture
Source: Usleber et al., 1997

type of toxic effects

In animal experiments effects of STX on the respiratory system, myocard, muscle and nervous tissue (both peripheral and central) have been studied (Mons et al., 1998).

effects on the respiratory system

If PSP intoxication occurs, the effects on the respiratory system are responsible for the fatal ending. The cause of death is asphyxiation due to progressive respiratory muscle paralysis. In animals (cat, rabbit) doses of 1-2 mg STX/kg bw administered intravenously caused a decreased respiratory activity reflected in both a decline in the amplitude and velocity. When the dose was raised to 4-5 mg STX/kg bw, a strong depression of the respiration occurred. By artificial respiration, death can be avoided. If the dose is not too excessive the respiration may return spontaneously. In animal experiments only peripheral paralysis has been noticed by direct effects on the muscle of the respiratory system. The respiratory centre of the nervous system is not inhibited, action potentials are sent off to the midriff and the middle rib muscles. Other investigators however suggest a central influence. The possibility of central effects on the respiratory neurones may therefore not be excluded. The occurrence of paresthesia and feeling of lightness are often connected with a central effect, however the peripheral effects on the nervous system may be the cause of these symptoms (Mons et al., 1998).

cardiovascular effects

In anesthetized animals doses above 1 mg STX/kg bw (i.v.) can already provoke hypotension (paralysis of muscles is already observed at lower dose levels). This cardiovascular effect is seldom observed in human cases of intoxications and is more likely the reflection of peripheral effects, although the central nervous system might be involved to a certain extent. About the peripheral action there are uncertainties. Apart from a direct effect on the muscle tissue the possibility of an axonal blockade of the sympathetic nervous system can not be excluded. Most investigators agree on the fact that there are no, or hardly any, direct cardial effects. As an exception a direct disturbance of the atrio-ventricular sinus conduction is mentioned (Mons et al., 1998).

neuromuscular effects

An intravenous dose of 1-2 mg STX causes a fast weakening of muscle contractions; both contractions by direct stimulation as contractions by indirect motoneuron stimulation are affected. The effects include all skeletal muscle tissues. This dose level induces also a decrease of the action potential-amplitude and a longer latency time in the peripheral nervous tissue. Both motor and sensory neurones are influenced but the sensory neurones are already inhibited at lower dose levels. By this influence on the sensory system, the numbness and the proprioceptive loss may be explained but not the paresthesia. No clarity has been achieved about the possible theory on the toxic mechanism and many scientific debates reflect this (Mons et al., 1998).

effects on the central nervous system

There are uncertainties about the existence of an effect of PSP toxins on the central nervous system. Most symptoms can be attributed to peripheral effects. However central effects can not be excluded. For example investigators reported the influence of STX on the hamstring reflex (Mons et al., 1998).

repeated administration

No data

mutagenicity

No data

reproduction/teratogenicity studies

No data

2.5.4 Toxicity to humans

acute toxicity

The level at which PSP intoxications occur in humans varies considerably. This variation is mainly due to individual difference in sensitivity and fluctuation in the method of determination. Oral intake causing mild symptoms varied from 144 to 1660 mg STX eq/person. Fatal intoxications were reported after a calculated consumption of 456-12400 mg STX eq/person. These values are only reconstructed from what remained of the toxic mussels and vary greatly. An oral consumption of 300 mg PSP toxin per person was in some cases reported as fatal, whereas others noted the absence of toxic symptoms after an oral dose of 320 ug toxin per person. In Alaska, PSP was fatal for one fisherman, while two others eventually recovered. The stomach contents of patients contained 370 ug PSP toxin (STX eq)/100 g (Mons et al., 1998). Other sources report mild poisoning at doses of PSP toxins between 304 and 4128 µg/person, while severe poisonings are caused by doses between 576 and 8272 µg/person (Aune, 2001).

In 1987, an outbreak of PSP with 187 cases and 26 deaths was reported after consumption of a clam (Amphichaena kindermani) soup. The fatalities were the highest among young children (50 percent) compared with 7 percent in adults. Some of the children who died ingested an estimated dose of 140-160 MU/kg bw (Rodrigue et al., 1990). Aune (2001) reported that the minimal lethal dose in this incident was estimated to be about 25 µg STX eq/kg bw for a child weighing 25 kg compared to 86-788 µg STX eq/kg bw in four adults who died.

The Australia New Zeleand Food Authority reported that 120 to 180 µg PSP toxins can produce moderate symptoms in humans, 400 to 1060 µg can cause death and 2 000 to 10 000 µg is more likely to constitute a fatal dose (ANZFA, 2001).

The mortality rate of PSP varies considerably. In recent outbreaks involving over 200 people in North America and Western Europe, no deaths occurred. However, in similar outbreaks in Southeast Asia and Latin America, death rates of 2 to 14 percent have been recorded. A large part of the difference is due to the fact that in the former cases intoxication often occurred in urban areas where victims already have access to hospital care, whereas in Southeast Asia and Latin America, intoxications often occurred in rural areas where such poisonings had never before been encountered by local people and health professionals (Mons et al., 1998).

toxic symptoms

Clinical symptoms of PSP in mild cases include a tingling sensation or numbness around lips, which appear mostly within 30 minutes. These effects are clearly due to local absorption of the PSP toxins through the buccal mucous membranes. These sensations then spread gradually to the face and neck. Prickly sensation in the fingertips and toes is frequent as are headaches, dizziness, nausea, vomiting and diarrhoea. Sometimes, temporary blindness is also observed. Most symptoms have a quick onset (hours) but may last for days, and are virtually invariant in all cases of paralytic shellfish poisoning. They precede distinct muscular weakness because sensory nerves, being thinner and having shorter internodes than motor nerves, are always affected first by any axonal blocking agents.

In moderately severe poisoning, paresthesia progresses to the arms and the legs, which also exhibit motor weakness. Giddiness and incoherent speech are apparent. Cerebellar manifestations such as ataxia, motor incoordination and dysmetria are frequent. Respiratory difficulties begin to appear as tightness around the throat. In severe poisoning, muscular paralysis spreads and becomes deeper. The pulse usually shows no alarming abnormality. Pronounced respiratory difficulty and death through respiratory paralysis may occur within 2 to 24 hours of ingestion (Mons et al., 1998).

Since STXs are charged, water-soluble molecules, it is probable that they do not penetrate the blood-brain barrier well and most of their effects are on peripheral nerves (Mons et al., 1998). In a study carried out in Alaska among patients suffering from PSP, hypertension also occurred corresponding with the ingested toxin dose (Gessner et al., 1997).

effects of alcohol consumption

The effect of alcohol consumption on PSP is still unclear. Some say that alcohol might be a protective factor against the adverse effects of PSP toxins but the mechanism through which alcohol might reduce the risk is unknown. Since the elimination of PSP toxins occurs at least in part through the urine, alcohol may influence illness by a diuretic effect. Alternatively, alcohol may cause hepatic enzyme induction. In a case-control study in Alaska, 47 outbreaks were studied for which the consumption histories of all persons were known. Alcohol consumption and eating cooked rather than raw shellfish were associated with a reduced risk of PSP. An association between illness and either the toxin level or dose ingested was not found (Mons et al., 1998).

treatment

The clinical management of poisoned victims is entirely supportive. If no vomiting had occurred spontaneously, induced emesis or gastric lavage should be used to remove sources of unabsorbed toxins. As the PSP toxins are strongly charged at the gastric pH, they would be effectively absorbed by activated charcoal. These steps are especially important in the management of child victims of poisoning, as the severity of the intoxication is directly dependent on the concentration of the toxins in the body. In the 1987 epidemic in Guatemala, the mortality rate in children up to six years of age was 50 percent while for adults it was 7 percent.

In moderately severe cases, maintenance of adequate ventilation is the primary concern. In uncomplicated PSP the airway is not obstructed by excessive excretion. As ventilatory failure is due to varying degrees of paralysis of the respiratory nerves and muscles, positive pressure assisted ventilation, when indicated, is desirable. Fluid therapy is essential to correct any possible acidosis. Additionally, it will facilitate the renal excretion of the toxins.

Time-honoured conservatively supportive management has proven effective. If the patient survives 18 hours, the prognosis is good, with complete and rapid recovery. Some say that even nine hours should be adequate for a physiological reduction of the toxins concentration to relatively harmless levels, except in those cases where the toxin concentration began at exceptionally high levels, or in victims with impaired renal function. Artificial ventilation and gastric lavage are still the only acceptable medical countermeasure against STX intoxication. However, in cases of severe intoxication, artificial ventilation may be inadequate (Mons et al., 1998).

Some studies in animals gave an indication that 4-aminopyridine may be useful as antidote for STX-intoxication. 4-Aminopyridine significantly reversed the respiratory rate, tidal volume and blood pressure to normal values in anesthesized STX-intoxicated rats. Furthermore, 4-aminopyridine, not only prolonged the survival time but also decreased the mortality of mice (71 to 43 percent) at a normally lethal dose (30 mg/kg i.p.) of STX (Chen et al., 1996). In guinea pigs, 4-aminopyridine was able to reverse the extent of cardiorespiratory infirmity and other sublethal effects of STX. At the point where cardiorespiratory performance was most seriously compromised (30 minutes after intramuscular STX injection) 4-aminopyridine was injected. Within minutes the STX induced diaphragmatic blockade, bradypnoea, bradycardia and depressed cortical activity were all restored to a level either comparable to, or surpassing, that of control. At the dose-level used to restore ventilatory function and cardiovascular performance, 4-aminopyridine produced no sign of seizures and convulsions. Although less secondary effects such as cortical excitant/arousal effect and transient periods of skeletal muscle fasciculation were seen, these events were of minor concern in view of the remarkable therapeutic effects (Chang et al., 1997b).

STX-induced lethality in guinea pigs can be reversed by 4-aminopyridine when it is administered at the time of respiratory arrest. 4-Aminopyridine showed to facilitate recovery and reduce the amount of time guinea pigs are dependent upon artificial ventilation (Benton et al., 1998).

2.5.5 Toxicity to aquatic organisms

The acute and chronic effects of A. tamarensis on the shrimp Neomysis awatchensis were studied. Mortality rate in the presence of 9 000 cells/litre was 55 percent for N. awatschensis. The 96 hour LC₅₀ for N. awatschensis was 7 000 cells/litre. In the presence of a cell-free filtrate 25 percent of N. awatschensis died after 96 hours. In a 62-day experiment, 37 percent of N. awatschensis died at a concentration of 900 cells/ml and the number of juveniles was 27 (only 16.4 percent of the number of juveniles in the control group). The first date of reproduction was prolonged for three days compared with controls. At 900 cells/ml length and weight of parent shrimps were 95.6 and 81.9 percent, respectively of control values. These last differences were not significant (Zhijun et al., 2001).

The hatching rate of fertilized eggs of the scallop Chlamys farreri was only 30, and 5 percent of controls at exposure for 36 hours to A. tamarensis cells or cellular fragments at concentrations of 100 and 500 cells/ml respectively. Exposure of the eggs to STX or cell contents (supernatant or re-suspended algal cells) did not elicit this inhibitory response. The alga also affected larvae at early D-shape stage of scallop. Survival rates began to decrease significantly at exposure for six days to concentrations ³3000 cells/ml; no larvae could survive after 14-day exposure to 10 000 cells/ml. This study indicated that developmental stages before blastula were the most sensitive periods to A. tamarensis toxins. And the alga at the early exponential stage had the strongest effect on egg hatching comparing with other growth phases (Yan et al., 2001).

Most of the bivalve shellfish filter-feeders are relatively insensitive to the PSP toxins because many of them have nerves and muscles operated mainly by voltage-gated calcium channels and STX and other PSP toxins block only the voltage-gated sodium channel with great potency (high affinity). This enables them to continue feeding and thereafter they become highly toxic. The high tolerance of for example blue mussels to STXs and their continued feeding on toxic algae can result in exceeding the 80 mg STX level in less than one hour (Mons et al., 1998).

There are even some species of bivalves known to avoid toxic dinoflagellates. One species of particular interest is the northern quahog or hard clam, Mercenaria mercenaria. In a laboratory study in the presence of A. tamarensis the quahog first retracts its siphons and then completely isolates itself from the external environment by means of shell valve closure. The animals did not re-open their valves until the addition of clean seawater (Mons et al., 1998).

Blooms of Alexandrium can cause fish kills. For an adult herring 10-20 mg STX eq is a lethal dose (Teegarden and Cembella, 1996). When milkfish (Chanos chanos) fingerlings were exposed (without any aeration) to toxic A. minutum (1.4 x 10⁴ cells/ml ~ 3.0 x 10⁴ cells/ml) or toxic algal extracts (5.13 x 10³ cells/ml ~ 2.05 x 10⁴ cells/ml) for one day, noticeable edema, hyperplasia and necrosis of secondary lamellae in the gill were observed by light microscopical examination. Similar symptoms were also seen in fish treated with pure STX (6.5 x 10^-2 mg/ml). At the same treatment, fingerlings showed a higher O₂ consumption rate and a higher demand of critical O₂ pressure. Treatment of milkfish fingerlings with non-toxic A. minutum cells or the algal cell extract did not cause any damage to the gills, nor a rise in O₂ consumption rate or critical O₂ demand. However, fingerlings died from suffocation at a cell concentration of 2.1 x 10⁴ cells/ml of nontoxic A. minutum without aeration support in 24 hours (Chou and Chen, 2001b).

In a study on the sublethal effects of Alexandrium, as might occur during pre- and post-bloom conditions, copepods grazing on Alexandrium, were fed to larval mummichogs (Fundulus heteroclitus). The exposed larvae consistently showed reduced swimming performance and prey capture. In some trials exposed copepods were captured more easily than unexposed ones. The results suggest that Alexandrium impacts both fish and their prey, therefore facilitating transfer of the toxin through the food web (Samson et al., 1999).

A variety of zooplankton responses towards toxic flagellates have been reported, ranging from avoidance to ingestion of the algae or even active selection. Bagøien et al. (1996) showed that all developmental stages of the zooplankton species Euterpina acutifrons were inactivated by A. minutum, but the effect was faster and more intense on the nauplii. However, nauplii and copepods not moving within a given time are often defined as inactive, but in many cases they are not dead. Still, inactivation of adults also reached high proportions (80 percent was inactivated) towards the end of a five-day large volume experiment. Many of the copepods were then clearly dead but a considerable proportion of adult females remained active and were able to produce viable eggs. Only traces of PSP toxins were detected in adult copepods after five days of exposure to A. minutum suggesting that E. acutifrons avoids feeding on the dinoflagellates after tasting a few cells. Extrapolating these results to natural conditions suggests that even if the toxic effects are not acute under moderate to dense bloom conditions, a significant proportion of zooplankton population can be killed, assuming that the zooplankton do not avoid the affected area.

Dutz (1998) studied the effects of the PSP toxins-producing dinoflagellate A. lusitanicum on the reproductive success of the common calonoid copepod Acartia clausi. Acartia clausi did not show mortality when fed on A. lusitanicum and were also able to produce eggs. However, the egg production was reduced. It is suggested that the ingested toxins probably interfere with digestive processes or cause enhanced energy expenditure due to detoxification. Consequently less energy is available probably resulting in reduced fecundity. In Egyptian coastal waters, mortality of wild fish and of fish kept in aquariums with filtered seawater has been associated with dense blooms of A. minutum (Bagøien et al., 1996).

2.5.6 Toxicity to water fowl

An acute poisoning incident in captive herring gull chicks (Larus argentatus) fed a batch of store-bought scallops was described. The chicks developed a characteristic acute syndrome that had not until then been reported in birds and the cause of which remained to be identified. The authors suggested that it was a variant of PSP insofar as it was paralytic and caused by shellfish. However, analyses were negative for STX, brevetoxin and domoic acid (Gochfeld and Burger, 1998).