Molecular detection of harmful cyanobacteria and expression of their toxin genes in Dutch lakes using multi-probe RNA chips

Abstract Harmful cyanobacterial blooms are a major threat to water quality and human health. Adequate risk assessment is thus required, which relies strongly on comprehensive monitoring. Here, we tested novel multi-probe RNA chips developed in the European project, μAqua, to determine the abundance of harmful cyanobacterial species and expression of selected toxin genes in six Dutch lakes. All of the targeted cyanobacterial genera, except for Planktothrix , were detected using the microarray, with predominance of Dolichospermum and Microcystis signals, of which the former was found across all sites and detected by the probes for Anabaena where it was formerly placed. These were confirmed by microscopic cell counts at three sites, whereas at the other sites, microscopic cell counts were lower. Probe signals of Microcystis showed larger variation across sites but also matched microscopic counts for three sites. At the other sites, microscopic counts were distinctly higher. We detected anatoxin-a in the water at all sites, but unfortunately no genes for this toxin were on this generation of the toxin array. For microcystins, we found none or low concentrations in the water, despite high population densities of putative microcystin producers (i.e. Microcystis , Dolichospermum ). The described method requires further testing with a wider range of cyanobacterial communities and toxin concentrations before implementation into routine cyanobacterial risk assessment. Yet, our results demonstrate a great potential for applying multi-probe RNA chips for species as well as toxins to eutrophic waters with high cyanobacterial densities as a routine monitoring tool and as a predictive tool for toxin potential.