CYANOBACTERIAL TOXINS - OCCURRENCE AND LEVELS IN RAW WATER SOURCES AND REMOVAL IN WATERWORKS

Project leader: Kirsti Lahti, Finnish Environment Institute, P.O.Box 140,
FIN-00251 Helsinki, Finland, tel. +358-9-40300 854, e-mail: Kirsti.Lahti@vhvsy.fi
 
 

PUBLICATIONS

TIIVISTELMÄ SUOMEKSI

Researchers:
Jarkko Rapala, Finnish Environment Institute, P.O.Box 140, 00251 Helsinki, Finland, tel. +358-9-40300 861, email: Jarkko.Rapala@vyh.fi [project leader from May to
December 2001]
Jaana Kukkonen, Finnish Environment Institute, P.O.Box 140, 00251 Helsinki
Mariliina Lifländer, Finnish Environment Institute, P.O.Box 140, 00251 Helsinki
Kaarina Sivonen, Department of Applied Chemistry and Microbiology, P.O.Box 56, FIN-00014 Helsinki University, tel. +358-9-1915 9270, e-mail: Kaarina.Sivonen@helsinki.fi
Kirsti Erkomaa, Finnish Environment Institute, P.O.Box 140, 00251 Helsinki, tel. +358-9-40300 825
Liisa Lepistö, Finnish Environment Institute, P.O.Box 140, 00251 Helsinki, tel. +358-9- 40300 312, e-mail: Liisa.Lepisto@vyh.fi

Consortium: Microbial risks of drinking water contaminated with protozoans, viruses or cyanobacterial toxins
Financing SYTTY organisation: Tekes
Funding from SYTTY / Total funding of project (€): 182492 /  485022
Person-months of work funded by SYTTY / Total person-months of work: 35 / 88

KEY WORDS: cyanobacteria, toxins, water works, microcystins, drinking water
 

EXTENDED ABSTRACT

1 Introduction

Microcystins (MCYST) are a group of hepatotoxins produced by freshwater cyanobacteria. They are inhibitors of serine/threonine protein phosphatase enzymes and among the most potent known tumour-promoting compounds. In addition that toxic cyanobacteria have caused animal poisonings, they may cause a health hazard for humans through the use of water for drinking or recreation. WHO has proposed a guide value of 1 µg l-1 for the most common variant, MCYST-LR, in drinking water. Conventional water treatment processes do not  remove all microcystins from drinking water. The lack of  methods to detect low microcystin concentra-tions has hampered the assessment of water treatment efficiency and the monitoring of microcystins in drinking water. Only recently have highly sensitive methods, such as protein phosphatase inhibition assays (PPI) and enzyme-linked immunosorbent assays (ELISA) been developed.

For cyanobacterial neurotoxins (anatoxin-a and its derivatives, anatoxin-a(S) and saxitoxins) there exist no guide values in drinking water because they are not as common as microcystins and they are not known to have chronic health effects. There is a lack of experimental and human data for derivation of tolerable daily intake and due to the lack of suitable sensitive methods for monitoring.

The aim of the study was to evaluate the suitability of PPI and ELISA assays for monitoring  microcystins and to study occurrence of toxins in raw water and treated drinking water of Finnish water treatment plants. Variation of toxin concentrations in lakes was monitored, and sensitive detection methods for cyanobacterial neurotoxins were taken into use. The hygienic role of endotoxins associated with cyanobacteria and their removal during drinking water treatment was also assessed.

2 Methods

Water samples from raw water sources and after different treatment processes were collected from 10 municipal waterworks. One waterworks was monitored at least fortnightly during the open water season of the whole four-year study. Samples from the other ones were taken at least three times during two summers. In addition, microcystins were analysed from small-scale water treatment plants. Diurnal, horizontal and vertical variation of microcystin concentrations were monitored in lakes. Microscopical examination was used to identify and calculate the number of cyanobacteria. Microcystins were monitored  using EnviroGardTM  microcystins ELISA plate kits,  a colorimetric PPI assay and HPLC. Artemia salina bioassay was used for preliminary detection of neurotoxic cyanobacteria. Anatoxin-a was detected with HPLC, anatoxin-a(S) with acetylcholinesterase inhibition assay, and saxitoxins with MISTTM Alert rapid test for PSP-toxins. Endotoxins were determined with Limulus ameobocyte lysate assay.

3 Results and Discussion

Method development
Application of the colorimetric PPI assay for the detection of microcystins was developed. Quantification of microcystins using pure compounds, toxin-producing cyanobacterial strains and natural water samples gave comparable results with ELISA, PPI and HPLC, except with certain demethylated microcystins. These demethylated variants were common in cyanobacteria that often occurred in lakes and raw water sources of waterworks.

The sensitive HPLC method using fluorometric detection was taken into use and applied for the detection of anatoxin-a and its derivatives. Other methods taken into use for the detection of cyanobacterial neurotoxins included acetylcholinesterase inhibition assay for anatoxin-a(S) and a rapid MISTTM-test based on immunoaffinity chromatography for saxitoxins.

The use of mussels for indication of occurrence and levels of microcystins in water bodies was shown to be possible but the methodology for toxin detection from tissue samples needs further development.

Toxins in raw water sources
Microcystin concentrations varied significantly during different years. In 1999 they were detected at nine raw water supplies of the ten waterworks studied, and great numbers of cyanobacteria occurred in three of the raw water sources. The highest measured concentrations in the raw waters entering the waterworks exceeded 10 µg l-1 (Fig. 1) which is ten times higher than the guide value proposed by WHO for drinking water. In 2000 low numbers of cyanobacteria were detected in raw waters of the nine waterworks studied. Microcystins were detected at only three waterworks, and their concentration in the raw waters remained below 1 µg l-1.

Fig. 1. Concentration of microcystins (  ) and total length of Planktothrix agardhii filaments (?) in the raw water of one waterworks studied in 1999.

Mass occurrences of neurotoxic cyanobacteria were observed at two waterworks. Although we were capable of detecting all known groups of cyanobacterial neurotoxins, the compound responsible for the toxicity remained unknown. The toxicity was associated with appearance of a straight-filamentous Anabaena, and with HPLC a compound possibly related to anatoxin-a was always detected. At one waterworks such an Anabaena appeared regularly each year in late June and early July.

Removal of cyanobacteria and toxins during water treatment
Water treatment processes effectively removed cyanobacterial and other algal cells as well as microcystins. Traces of microcystins were occasionally detected in treated drinking water but the concentrations were less than 0.1 µg l-1 which is below the guide value proposed by WHO. The most significant removal occurred in the early stages of water treatment i.e. during coagulation, clarification and sand filtration thereafter simultaneously with the removal of cyanobacterial cells. Rapid filtration without coagulation/clarification was insufficient for microcystins. During slow sand filtration and artificial recharge of groundwater biological degradation enhances the removal. Nanofiltration removed microcystins effectively.

Endotoxins
Endotoxin concentrations of pure cyanobacterial strains were low, but in the samples from cyanobacterial mass occurrences the concentrations were high, due to Gram negative bacteria associated with cyanobacteria. In raw waters of waterworks the concentrations varied from 18 to 356 EU ml-1 (EU, endotoxin unit). Different treatment processes removed 59-97% of them. Risk assessment on the basis of the results is not possible since there exist no guide values for endotoxins in water and only a limited number of studies on their concentrations has been published.

Variation of cyanobacteria and toxins in lakes
Variation of the occurrence of microcystins was studied in several lakes. Demethylated microcystins were common in water samples and in cyanobacterial strains isolated from them. Of cyanobacterial neurotoxins only anatoxin-a and its derivatives were detected. The highest measured microcystin concentrations during mass occurrences of cyanobacteria were approximately 20 mg l-1, which exceeded 1000-fold the concentrations in raw waters of the waterworks. Horizontal, vertical and diurnal variation of cyanobacteria and microcystins were also studied. Horizontally microcystin concentrations of the water sample varied more than 10-fold. Planktothrix agardhii that was often associated with high microcystin concentrations occasionally developed a biomass maximum at deeper water layers (Fig. 2), which is especially inconvenient for waterworks since such a mass occurrence is invisible. Toxic cyanobacteria were most common in summer and early autumn, but occasionally they were detected in spring and late autumn until ice cover.

Fig. 2. Example of vertical distribution of chlorophyll-a (o, µg l-1), microcystins ( , µg l-1)  and Planktothrix agardhii (?, mg l-1).

4 Conclusions

Microcystins were quite common in raw water sources of Finnish waterworks during summer and autumn. Yearly their concentration varied highly, from nanograms to milligrams per litre. Neurotoxins of cyanobacterial origin were detected less frequently. ELISA and PPI methods were suitable methods for monitoring of total microcystin concentrations in water. Compared to HPLC the methods are more sensitive and less time consuming. During the study analytical methodology was established which can be applied for sensitive determination of cyanobacterial hepatotoxin and neurotoxins for monitoring purposes. However, attention should also be paid to the identification of different microcystin variants and unknown neurotoxins that occur in our waters. For these studies more complicated chemical analysis methods are needed. Although microcystin concentrations detected in treated drinking water were low, small-scale waterworks with conventional purification methods, adsorption and desorption of toxins in activated carbon filtration and toxin release during washing of sand filters may pose a temporal risk to water quality. In conclusion, the yield of microcystins in drinking water distributed by Finnish waterworks is clearly below the tolerable daily intake presented by WHO. On the contrary, recreational waters may pose a health hazard, since high toxin concentrations were relatively common.
 

[ Projects | Main Page ]