In 1979, on Palm Island off the tropical coast of Queensland, Australia, a mystery illness with gastrointestinal symptoms caused the hospitalisation of more than 100 of the local residents, mostly children. The probable cause of this ilhess was determined retrospectively to be the potent toxin cylindrospermopsin from a bloom of the cyanobacterium Cylindrospermopsis raciborskii (C. raciborskii), in the reticulated water supply. This species was not previously known to be toxic. Since that time, numerous blooms have been reported from sub-tropical areas of Australia and overseas. In the summer of 1995-96, this organism was identified in bloom samples from a major water reservoir.
Lake Samsonvale (North Pine Dam), supplying the Queensland state capital, Brisbane. At the time little was known about the organism or its toxin and with preliminary mouse bioassay tests suggesting the material from the reservoir to be toxic, use of the reservoir was temporarily suspended. This resulted in an active research program being implemented at the National Research Centre for Environmental Toxicology in Brisbane.
To undertake fiiurther research, development and validation of reference methods for both cylindrospermopsin and GSH in liver tissue using HPLC-MS/MS has been completed as well as studies on the stability of GSH under various conditions relevant to these studies. Methods have also been developed to extract and purify cylindrospermopsin and deoxycylindrospermopsin from aqueous solution including culture media. Purified radiolabelled cylindrospermopsin has also been produced.
The structure of an analog of cylindrospermopsin, deoxycylindrospermopsin, has been determined and its lack of toxicity in Quackenbush mice, even when administered concurrently with cylindrospermopsin, has been demonstrated, implicating the hydoxyl group on the bridge to the uracil moiety as being essential to the toxicity of cylindrospermopsin. The LOAEL values were determined to be 0.005, 0.15, 0.005 and 0.05 mg/kg/day for pure cylindrospermopsin dosed by the IP route, oral route and cellular extract containing cylindrospermopsin dosed by the IP and oral routes respectively. The NOAEL values were determined as 0.001, 0.05, <0.005, and 0.05 mg/kg/day and the ED50 0.005, >0.05 <0.15, <0.005 and approx 0.05 mg/kg/day respectively for the same treatments as for the LOAEL. From the values for LOAEL and NOAEL, the oral bioavailability has been estimated as less than 5 %, which is consistent with other published studies. Radiolabelled cylindropsermopsin has been used to demonstrate the rapid urinary excretion of cylindrospermopsin and its retention in the liver in some animals. Cylindrospermopsin also reduces GSH, most likely by inhibition of synthesis, but this does not have sufficient impact to explain the toxicity. Inhibition of metabolism by administration of piperonyl butoxide does however protect Quackenbush mice from toxicity, suggesting metabolic activation is essential to the toxicity in vivo for this strain at least.
Previously published experiments using the rabbit reticulocyte model suggested inhibition of protein synthesis to be the mechanism of toxicity for cylindrospermopsin. Based on this present work, the probability exists that two mechanisms of toxicity exist, The first, occurring at lower concentrations, requires metabolic activation of cylindrospermopsin, resulting in a covalently bound product in the hepatocyte. The second, probably involving inhibition of protein synthesis occurs at higher concentrations and may not require metabolic activation of cylindrospermopsin. A novel focal haematological lesion described and investigated herein is most likely the result of a disturbance in protein synthesis in the liver. Given that the risks associated with cylindrospermopsin toxicity lie primarily with chronic dosing by low concentrations in drinking water, this second mechanism may be of lesser importance from a public health perspective in humans.
These present studies have also demonstrated the inter-individual variability in cyhndrospermopsin toxicity for Quackenbush mice. This is evidenced in particular in mice in control groups dosed at the ED50, where some mice succumb within 2-3 days and others appear totally unaffected. This may suggest a genetic component to the toxicity, perhaps in the expression of appropriate P450 isoforms to allow activation to occur. Further studies may however benefit from a concerted effort to find an inbred strain which produces more consistent responses.
Given that cylindrospermopsin is strongly zwitterionic, it is an unusual toxin, as such compounds are generally unable to cross the cellular membrane and are rapidly excreted from the body. The ability of cylindrospermopsin to cross the hepatocellular membrane is most likely due to the presence of aqueous pores in the cell. Uptake is then further assisted by the irreversible binding to intracellular components, creating a low unbound concentration of cylindrospermopsin intracellularly. This will increase the concentration difference across the cell wall, which is the driving force for passive transport.