Abstract
The wide distribution of cyanobacteria in aquatic environments leads to the risk of water contamination by cyanotoxins, which generate environmental and public health issues. Measurements of cell densities or pigment contents allow both the early detection of cellular growth and bloom monitoring, but these methods are not sufficiently accurate to predict actual cyanobacterial risk. To quantify cyanotoxins, analytical methods are considered the gold standards, but they are laborious, expensive, time-consuming and available in a limited number of laboratories. In cyanobacterial species with toxic potential, cyanotoxin production is restricted to some strains, and blooms can contain varying proportions of both toxic and non-toxic cells, which are morphologically indistinguishable. The sequencing of cyanobacterial genomes led to the description of gene clusters responsible for cyanotoxin production, which paved the way for the use of these genes as targets for PCR and then quantitative PCR (qPCR). Thus, the quantification of cyanotoxin genes appeared as a new method for estimating the potential toxicity of blooms. This raises a question concerning whether qPCR-based methods would be a reliable indicator of toxin concentration in the environment. Here, we review studies that report the parallel detection of microcystin genes and microcystin concentrations in natural populations and also a smaller number of studies dedicated to cylindrospermopsin and saxitoxin. We discuss the possible issues associated with the contradictory findings reported to date, present methodological limitations and consider the use of qPCR as an indicator of cyanotoxin risk.