In vitro toxicity and genotoxicity assessment of disinfection by-products, organic N-chloramines.

Disinfection by-products (DBPs) are of concern to both water industries and health authorities. Although several classes of DBPs have been studied, and there are regulated safe levels in disinfected water for some, a large portion of DBPs are not characterized, and need further investigation. Organic N-chloramines are a group of DBPs, which can be formed during common disinfection processes such as chlorination and chloramination, but little is known in terms of their toxicological significance if consumed in drinking water. Only a few in vitro studies using bacterial assays have reported some genotoxic potential of organic N-chloramines, largely in the context of inflammatory processes in the body rather than exposure through drinking water. In this study, we investigated 16 organic N-chloramines produced by chlorination of model amino acids and amines. It was found that within the drinking water-relevant micromolar concentration range, four compounds were both cytotoxic and genotoxic to mammalian cells. A small reduction of cellular GSH was also observed in the treatment with these four compounds, but not of a magnitude to account for the cytotoxicity and genotoxicity. The results presented in this study demonstrate that some organic N-chloramines, at low concentrations that might be present in disinfected water, can be harmful to mammalian cells.

Evaluation of the Abraxis Strip Test for Microcystins™ for use with wastewater effluent and reservoir water.

Rapid tests for the microcystin-type cyanobacterial toxins that are designed to be able to be used in the field have recently become available. The tests provide a semi-quantitative result over a relatively narrow concentration range (10-fold) and are available with detection limits relevant for drinking water and recreational water compliance testing (1 µg/L and 10 µg/L, respectively). The aim of this research was to assess the applicability of these tests for the determination of microcystin-related toxicity in treated effluent from the Western Treatment Plant and potable source water from Tarago Reservoir, both near Melbourne, Australia. Accuracy, precision, cross-reactivity, matrix effects and inter-operator variability were assessed. The claimed mLR concentration response range of the tests was confirmed within reasonable limits, although the false negative and false positive rates were significant for spike concentrations below 2.5 µg/L (Recreational Strip Test). Inter-operator variability was reasonably high (CV=23%) and this was exacerbated by the use of untrained scorers. Contributing to this was significant inter-assay variability in test band intensity (CV=28%). The strip tests responded to all 8 microcystin analogues tested and also to a mixture of another 7 analogues contained in a Certified Bloom Material. Cross-reactivity was always greater than 50%. Matrix effects due to the test waters or to cyanobacterial cell material were also relatively minor, being of the order of 2-fold at the maximum. Overall, these Strip Tests were found to be reliable for relatively rapid detection of microcystins around the upper limits of their response ranges, as recommended by the manufacturer. While the Recreational Water Strip test was less reliable in the lower ranges, it can be used in conjunction with the Drinking Water Strip test to reduce uncertainty around the 1 µg/L concentration. Despite limitations, both strip tests provide near real-time information which can assist with day to day operational decisions. When results indicate microcystin concentrations near compliance limits it is recommended that use of the test kits should be supported by accurate quantitative toxin testing together with traditional algal cell counts, and possibly emerging qPCR methods for species and toxin gene detection.

Morphological cell transformation of Syrian hamster embryo (SHE) cells by the cyanotoxin, cylindrospermopsin.

Cylindrospermopsin (CYN) is a cyanotoxin which has been implicated in human intoxication and animal mortality. Genotoxic activity of this hepatotoxin is known but its carcinogenic activity remains to be elucidated. In this work, CYN was assessed for its cell-transforming activity using the Syrian hamster embryo (SHE) cell transformation assay. This in vitro assay is used to evaluate the carcinogenic potential of chemical, physical and biological agents in SHE cells, which are primary, normal, diploid, genetically stable and capable of metabolic activation. We demonstrated that CYN induced a significant increase in morphological cell transformation in SHE cells following a 7-day continuous treatment in the range of non-cytotoxic concentrations 1 x 10(-7)-1 x 10(-2) ng/mL.

Comparison of analytical tools and biological assays for detection of paralytic shellfish poisoning toxins.

The paralytic shellfish poisoning toxins (PSTs) were, as their name suggests, discovered as a result of human poisoning after consumption of contaminated shellfish. More recently, however, the same toxins have been found to be produced by freshwater cyanobacteria. These organisms have worldwide distribution and are common in our sources of drinking water, thus presenting another route of potential human exposure. However, the regulatory limits for PSTs in drinking water are considerably lower than in shellfish. This has increased the need to find alternatives to the mouse bioassay, which, apart from being ethically questionable, does not have a limit of detection capable of detecting the PSTs in water at the regulated concentrations. Additionally, the number of naturally occurring PSTs has grown substantially since saxitoxin was first characterised, markedly increasing the analytical challenge of this group of compounds. This paper summarises the development of chromatographic, toxicity, and molecular sensor binding methodologies for detection of the PSTs in shellfish, cyanobacteria, and water contaminated by these toxins. It then summarises the advantages and disadvantages of their use for particular applications. Finally it recommends some future requirements that will contribute to their improvement for these applications.

Limited uptake of the cyanobacterial toxin cylindrospermopsin by Vero cells.

Cylindrospermopsin (CYN) is a cyanobacterial toxin increasingly found in drinking water sources worldwide. Toxicity studies have shown CYN can induce effects in a range of different cell types with primary hepatocytes consistently shown to be the most sensitive cellular model. How CYN enters the intracellular environment is not clear, although the size and hydrophilic nature of the toxin suggest it would not readily cross a lipid bilayer. In this study, a Vero cell line expressing green fluorescent protein (GFP) was used to monitor for CYN uptake based on the toxin's potent effects on protein synthesis. Effects on the GFP signal were compared with inhibitors cycloheximide (CHEX) and emetine. While CYN potency was demonstrated in a cell-free system (CYN>CHEX>emetine) it was considerably reduced in the Vero-GFP cell model (CHEX, emetine>>CYN). In contrast to other inhibitors, CYN effects on GFP signal increased 6 fold over 4-24 h incubation indicating slow, progressive uptake of the toxin. Confirming that the uptake process is not energy dependent CYN entry also occurred at 4 degrees C, while competition experiments excluded the uracil nucleobase transporter system as potential mechanism for CYN uptake. Dilution of media enhanced CYN uptake by Vero-GFP cells although mechanism by which this occurred is unknown.

Flow-cytometric analysis of in vitro micronucleus formation: comparative studies with WIL2-NS human lymphoblastoid and L5178Y mouse lymphoma cell lines.

The development of a flow cytometry-based micronucleus (FCMMN) assay for measuring the micronucleus (MN) frequency in cells following exposure to test chemicals has potential for improving reproducibility and reducing turn-around time when compared with the traditional microscopy-based micronucleus method. A major drawback of the FCMMN assay is that a false-positive interpretation could result from the presence of large numbers of apoptotic or necrotic bodies in the measured sample. Although several studies have reported ways in which the FCMMN assay could be improved using different staining techniques or electronic gating strategies, to date none of these protocols are suitable for use as a screening assay. To reduce the interference from apoptosis, performing the FCMMN assay with an apoptosis-resistant cell line may be an alternative approach. This study reports the use of p53-mutated cell lines to minimise the interference found in the FCMMN assay. Two commonly used cell lines (WIL2-NS and L5178Y) were investigated by comparison of (1) cytotoxicity and micronucleus induction in the FCMMN assay following treatment with model genotoxicants and (2) apoptotic responses after exposure to inducers of apoptosis. Both cell lines were responsive to all genotoxicants, producing concentration-dependent results with respect to genotoxicity. WIL2-NS cells were found to be more tolerant to apoptosis induction than L5178Y cells. This characteristic could be beneficial to minimise the interference from apoptotic nuclei in the FCMMN genotoxicity-screening assay.

Interaction of the cyanobacterial toxin cylindrospermopsin with the eukaryotic protein synthesis system.

The cyanobacterial toxin cylindrospermospin (CYN) was shown to inhibit the eukaryotic protein synthesis apparatus with similar potency in plant and mammalian cell extracts, IC(50) of 334 nM in wheat germ extract and 110 nM in reticulocyte lysate. [(14)C] CYN binding was assessed in reticulocyte lysates, following separation of ribosomes from free toxin on Sephadex G-50 size exclusion columns. While toxin binding was shown to be concentration-dependent (60-600 nM [(14)C] CYN), it did not correlate with ribosome content. The molar ratio of toxin to ribosomes was 0.02:1 at the binding plateau. Significant binding of [(14)C] CYN was identified in both the ribosome fraction (> 90% total ribosomes) and the first wash fraction eluting from the Sephadex G-50 column, despite the low ribosome content (<10% total ribosomes) in the latter. Bound [(14)C] CYN could be partially displaced by incubation with an excess of unlabelled CYN in both the ribosome and wash fractions, indicating a non-covalent linkage. Molecular weight cut-off filters identified that [(14)C] CYN was associated with high molecular weight material > 100 kDa. These results suggest that CYN's target may not be the ribosome itself, but rather one of the soluble proteins associated with the eukaryotic translation system.

The cyanobacterial toxin, cylindrospermopsin, induces fetal toxicity in the mouse after exposure late in gestation.

Cylindrospermopsin (cyn) is a cyanobacterial toxin implicated in human and wildlife poisonings. We have completed studies investigating the potential of purified cyn to induce developmental toxicity in mammals. The teratology study involved intraperitoneal injections (8.0-128 microg kg(-1)) on gestational days (GD) 8-12 with subsequent examination of term fetuses for viability, weight and morphological anomalies. Cyn was lethal to a significant portion of the dams receiving > or = 32 microg kg(-1). Surviving pregnant females were killed and fetuses removed for examination. Analysis indicates no adverse effects on litter size, fetal weight, or incidence of anomalies. Subsequently, 50 microg kg(-1) cyn was administered on GD 8-12 or 13-17. Animals were allowed to give birth and litters monitored for growth and viability. A reduction in litter size occurred in treated groups. Avg. pup wt. was only affected in the GD 13-17 group. GD 13-17 dams did not exhibit the toxicity noted in the GD 8-12 group but gave birth significantly earlier than controls. There was a significant number of dead GD 13-17 pups and incidences of blood in the gastrointestinal tract and hematomas in the tips of the tails in survivors. Pups were cross-fostered to control mothers in litters of 10. On postnatal days (PND) 5-6 there were no significant differences in weight gain or viability in GD 8-12 litters, while GD 13-17 litters had significantly reduced weight gain and viability. GD 13-17 exposed male pups still weighed significantly less than the controls after 15 months.

Cylindrospermopsin occurrence in two German lakes and preliminary assessment of toxicity and toxin production of Cylindrospermopsis raciborskii (Cyanobacteria) isolates.

Cylindrospermopsis raciborskii, a freshwater cyanobacterium of tropical origin, is not only increasingly found in (sub) tropical water bodies, but also in temperate regions. Since this species may produce potent toxins such as cylindrospermopsin (CYN) and paralytic shellfish poisons, its massive occurrence in water bodies used as drinking water sources or for recreation is of major concern. The proliferation of C. raciborskii in German water bodies has been documented for the past decade. We investigated the occurrence of CYN in field populations and isolates of C. raciborskii from two lakes, and assessed the toxicity of culture isolates using the mouse bioassay, primary rat hepatocytes and human derived cell lines. We show for the first time the occurrence of CYN in German water bodies. None of seven isolates of C. raciborskii contained CYN, however, all isolates were toxic to primary rat hepatocytes, human hepatoblastoma (HEP-G2) and human colon adenocarcinoma (CACO-2) cells. Methanolic extracts were more toxic than aqueous extracts. Three isolates tested in the mouse bioassay were toxic at a concentration of 800 mg kg(-1) showing liver and spleen damage and inflammation of the intestine. These results give strong evidence that the German isolates of C. raciborskii contain currently not identified or unknown toxins.

Oral toxicity of the cyanobacterial toxin cylindrospermopsin in male Swiss albino mice: determination of no observed adverse effect level for deriving a drinking water guideline value.

The cyanobacterial toxin cylindrospermopsin (CYN) is a frequent contaminant of freshwaters throughout the world, including those that are sources of drinking water. The first cases of human poisoning attributed to this toxin occurred from a treated drinking water supply in Queensland, Australia, in 1979. The toxin causes extensive damage to the liver, kidneys, spleen, heart, and other organs. It is known to be a potent protein synthesis inhibitor, but there is mounting evidence for genotoxicity and that it metabolizes to even more toxic forms. As part of a risk assessment process leading to a guideline for a safe drinking water level for this toxin, we performed a series of experiments to determine a no-observed-adverse-effect level (NOAEL) for this toxin. In the first trial male mice were exposed to CYN-containing cyanobacterial extract in their drinking water (0-657 microg CYN kg(-1) day(-1)) for 10 weeks. In the second trial mice received purified CYN by daily gavage (0-240 microg CYN kg(-1) day(-1)) for 11 weeks. Body and organ weights were recorded; urine, serum, and hematology analyses were performed; and histopathological examination of tissues was carried out. Body weights were significantly increased at low doses (30 and 60 microg kg(-1) day(-1)) and decreased at high doses (432 and 657 microg kg(-1) day(-1)). Liver and kidney weights were significantly increased at doses of 240 microg kg(-1) day(-1) and 60 microg kg(-1) day(-1), respectively. Serum bilirubin levels were significantly increased and bile acids significantly decreased at doses of 216 microg kg day(-1) and greater. Urine total protein was significantly decreased at doses above 60 microg kg(-1) day(-1). The kidney appeared to be the more sensitive organ to this toxin. If it is assumed that increased organ weights and changes in functional capacity are responses to an underlying toxic effect, then the NOAEL based on this data is 30 microg kg(-1) day(-1), which, with standard calculations and uncertainty factors, provides a proposed guideline safety value of 1 microg/L in drinking water.

Cell-free protein synthesis inhibition assay for the cyanobacterial toxin cylindrospermopsin.

The cyanobacterial toxin cylindrospermopsin (CYN) is known to be a potent inhibitor of protein synthesis. This paper describes the use of a rabbit reticulocyte lysate translation system as a protein synthesis inhibition assay for CYN. A dose response curve for protein synthesis inhibition by CYN was constructed and was modeled to a sigmoidal dose response curve with variable slope (R2 = 0.98). In this assay, CYN has an IC50 of 120 nM [95% confidence limits (Cl) = 111-130 nM] with a detection limit in the region of 50 nM in the assay solution. Application of the assay allows quantification of toxin samples within the range 0.5-3.0 microM (200-1200 micrograms/L) CYN. To assess the usefulness of this assay, a range of toxic and nontoxic Cylindrospermopsis raciborskii extracts, including both laboratory strains and environmental samples, were assayed by protein synthesis inhibition. These CYN quantifications were then compared to quantifications obtained by high performance liquid chromatography (HPLC) and HPLC-tandem mass spectrometry (HPLCMS-MS). The results demonstrate that the protein synthesis inhibition assay correlates well with both HPLCMS-MS (r2 = 0.99) and HPLC (r2 = 0.97) quantifications. We conclude that this is an accurate and rapid assay for the measurement of cylindrospermopsin in cyanobacterial extracts.

Preliminary evidence for in vivo tumour initiation by oral administration of extracts of the blue-green alga cylindrospermopsis raciborskii containing the toxin cylindrospermopsin.

New reports indicate that the toxic alkaloid cylindrospermopsin occurs in cyanobacteria in Israel, Florida, South America, and Australia in drinking water sources. This toxin is now recognised as a potential threat to human health. Furthermore, we have recently demonstrated the mutagenicity of cylindrospermopsin in vitro in a human lymphoblastoid cell-line. Therefore it is essential to determine whether cylindrospermopsin is also carcinogenic in vivo. In this preliminary study, 53 mice were treated up to three times orally with Cylindrospermopsis raciborskii extract containing cylindrospermopsin, while 27 control mice were treated with saline. A proportion of each group were then given O-tetradecanoylphorbol 13-acetate (10 microg/mouse, twice weekly in liquid food) for the duration of the experiment; the remainder were given a control diet. After 30 weeks, the mice were euthanased and the major organs were examined histologically. Five tumours were found in 53 cylindrospermopsin-treated mice while none were found in the 27 controls. Although the number of animals used was too low to provide statistical significance (p=0.16), the calculated relative risk (RR=6.2; 95% CI: 0.33-117) indicates a potential biological and public health significance requiring further investigation. Estimates are given of the size of experiment required to provide statistical proof of cylindrospermopsin carcinogenicity.

Preliminary evidence of toxicity associated with the benthic cyanobacterium Phormidium in South Australia.

In April 2000, the water supply for Yorke Peninsula in South Australia was deemed non-potable when extracts from a proliferation of the benthic cyanobacterium Phormidium aff. formosum in Upper Paskeville Reservoir were found to be lethally toxic by intraperitoneal injection into mice (400 mg kg-1). Routine water quality monitoring had failed to detect the development of the Phormidium until complaints of musty taste and odour, attributable to the production of 2-methyl-isoborneol (MIB), were received from the consumers. The 185 ML open-balancing storage, receiving filtered and chloraminated water from the River Murray, was isolated from the drinking water supply and a health alert was issued to approximately 15,000 consumers. The identity of the toxin(s) is thus far unknown, but clinical symptoms of toxicity in mice and chemical characteristics are distinct from the known major cyanotoxins. Preliminary characterisation of this toxin indicates that it has low solubility in water and organic solvents and is strongly associated with the particulate cellular material of the filaments. Toxicity of extracts was diminished by boiling and by treatment with chlorine, but not by chloramines. Further testing of floating cyanobacterial mats in the Torrens Lake in the city of Adelaide (Phormidium aff. formosum) and Myponga Reservoir (Phormidium aff. amoenum) in 2000/2001 was also found to be toxic by mouse bioassay. Toxicity is yet to be confirmed in monospecific cultured strains and further studies are required to identify the toxin and assess its health significance. Genetic characterisation of isolates has commenced in an attempt to classify their relatedness and to assist in the rapid identification of potentially toxic strains.

Micronucleus induction and chromosome loss in transformed human white cells indicate clastogenic and aneugenic action of the cyanobacterial toxin, cylindrospermopsin.

Cylindrospermopsin (CYN) is a potent inhibitor of protein synthesis produced by a number of cyanobacterial species, the most common being Cylindrospermopsis raciborskii. CYN contains a uracil moiety attached to a sulphated guanidino moiety, suggesting that it may have carcinogenic activity. This report describes the use of the WIL2-NS lymphoblastoid cell-line in the well-validated cytokinesis-block micronucleus (CBMN) assay to test this hypothesis. Centromeres (CENs) were identified in micronuclei (MNi) of binucleated cells (BNCs) by fluorescent in situ hybridisation of alpha centromeric DNA sequence repeats. The results indicate that CYN induced a significant increase in the frequency of MNi in BNCs exposed to 6 and 10microg/ml, and a significant increase in CEN-positive MNi at all concentrations of CYN tested (1, 3, 6, and 10microg/ml). However, despite this apparently greater sensitivity of WIL2-NS cells to induction of CEN-positive MNi at low CYN concentrations, at the higher concentrations the magnitude of the increase in CEN-positive MNi did not account for the greater increase in MNi in BNCs, indicating that both CEN-positive and CEN-negative MNi were induced. This suggests that CYN acts to induce cytogenetic damage via two mechanisms, one at the level of the DNA to induce strand breaks, the other at the level of kinetochore/spindle function to induce loss of whole chromosomes (aneuploidy). C. raciborskii occurs in a number of human drinking water sources worldwide and so these findings may have important public health implications.

Microcystins (cyanobacterial toxins) in drinking water enhance the growth of aberrant crypt foci in the mouse colon.

Microcystis aeruginosa produces toxic cyclic peptides called microcystins, potent hepatotoxins that have been implicated in tumor promotion in skin and liver. The model used in this investigation was the azoxymethane (AOM)-induced aberrant crypt focus (ACF) in the male C57Bl/6J mouse colon. Three intraperitoneal (i.p.) injections of 5 mg/kg AOM were administered at 7-d intervals to mice; 19 d after the last AOM injection, drinking water containing Microcystis extract was commenced and continued for a further 212 d. The content of microcystins in the drinking water was determined by mouse bioassay, high-performance liquid chromatography (HPLC), capillary eletrophoresis, and protein phosphatase inhibition. The doses employed were 0, 382, and 693 micrograms/kg bodyweight/d at the midpoint of the trial. Following postmortem examination blood cells, serum enzymes and organ pathology were investigated. A significant microcystin dose-dependent increase in the area of aberrant crypt foci was observed. There was no marked increase in the number of crypts/colon. Two overt colonic tumors (approximately 30 mm3) were seen in microcystin-treated mice, and one microscopic colonic tumor in an AOM-alone-treated mouse. This investigation provides the first evidence for the stimulation of preneoplastic colon tumor growth by microcystin.

Isolation and toxicity of Cylindrospermopsis raciborskii from an ornamental lake.

In Australia, the tropical/subtropical cyanobacterium Cylindrospermopsis raciborskii forms substantial blooms in some drinking water supply reservoirs, rivers and recreational water bodies during the warmer months of the year. This paper describes the isolation, culture and toxicity characterisation of Cylindrospermopsis from a water bloom in a small lake in NSW, Australia. The cyanobacterium grew as straight trichomes terminating with a characteristic heterocyst. The toxic alkaloid cylindrospermopsin was separated and identified by high-performance liquid chromatography at a concentration of 5.5 mg/g dry cells, 0.026 pg/cell. Intraperitoneal injection of sonicated cells caused liver, kidney, intestinal and lung damage, with an LD50 of 52 mg cells/kg mouse body weight at 24 hr, and 32 mg/kg at 7 days. The 24 hr mouse toxicity is not consistent with previous studies using pure cylindrospermopsin, and is suggestive of other toxic compounds in this isolate.