Rapid, multiplex-tandem PCR assay for automated detection and differentiation of toxigenic cyanobacterial blooms

Abstract

Cyanobacterial blooms are a major water quality issue and potential public health risk in freshwater, marine and estuarine ecosystems globally, because of their potential to produce cyanotoxins. To date, a significant challenge in the effective management of cyanobacterial has been an inability of classical microscopy-based approaches to consistently and reliably detect and differentiate toxic from non-toxic blooms. The potential of cyanobacteria to produce toxins has been linked to the presence of specific biosynthetic gene clusters. Here, we describe the application of a robotic PCR-based assay for the semi-automated and simultaneous detection of toxin biosynthesis genes of each of the toxin classes characterized to date for cyanobacteria [i.e., microcystins (MCYs), nodularins (NODs), cylindrospermopsins (CYNs) and paralytic shellfish toxins (PSTs)/saxitoxins (SXTs)]. We demonstrated high sensitivity and specificity for each assay using well-characterized, cultured isolates, and establish its utility as a quantitative PCR using DNA, clone and cell-based dilution series. In addition, we used 206 field-collected samples and 100 known negative controls to compare the performance of each assay with conventional PCR and direct toxin detection. We report a diagnostic specificity of 100% and a sensitivity of ≥97.7% for each assay.

Introduction

Cyanobacteria (‘blue–green algae’) are amongst the oldest, most abundant and widely distributed forms of life on the planet. These bacteria provide a basis for many of the world's aquatic ecosystems [1], and are one of the world's major carbon-sinks and oxygen producers [2]. However, in drinking or recreational waters, particularly during warmer months of the year, high-density cyanobacterial populations (i.e., blooms) may represent a significant threat to water quality and/or environmental health [3], can illicit allergic dermatitis [4], [5] and affect taste/palatability of the water [6]. Most significantly, many cyanobacteria species produce small molecules, including cyclic peptides [e.g., nodularins (NODs) and microcystins (MCYs)] and alkaloids [e.g., cylindrospermopsins (CYNs) and paralytic shellfish toxins (PSTs), such as, saxitoxin (SXTs)], that are potent neuro-, cyto- and/or hepatotoxins in vertebrates and represent a significant risk to animal and public health [3], [7], [8].

Currently, a range of methods exist for the detection of toxic cyanobacteria in water. These include immunological, microscopical and DNA-based approaches. Various immunoassays allow the direct detection of cyanotoxins in water samples [9]; however, although often easy to use, such methods have limitations, both with respect to sensitivity (false-negatives) and specificity (cross-reactivity and false-positives) [10]. Although more sensitive and specific chromatography and mass spectrometry approaches are available [9], they can be costly and time-consuming for routine use. The primary tool used to determine the level of risk for cyanotoxin production is routine water surveillance and the identification of potentially toxigenic cyanobacterial species by light microscopy [11]. Although this allows taxonomic identification, toxigenicity cannot be assessed based on morphological observation alone [12].

The ability to produce cyanotoxins has been demonstrated to relate to specific biosynthetic pathways encoded by complex gene operons, which have been characterized for the major toxins [13]. Various PCR assays have been developed to detect toxigenic cyanobacterial blooms [13], [14], [15]. The advantage of these approaches is that they are low-cost and highly sensitive, and when coupled to mutation scanning tools and/or direct DNA sequencing, can allow assessment of the toxigenic potential of a cyanobacterial bloom. A limitation is that numerous, highly distinct toxigenic species can occur in the same geographical region or bloom [16], [17], [18]. Therefore, samples must be assessed using multiple distinct PCR assays, often using different cycling conditions. Multiplexed PCR, wherein multiple primer pairs are used to target and amplify distinct loci simultaneously in a single reaction, has been pursued as a method to improve the efficiency of PCR-based detection of toxigenic cyanobacteria [18], [19], [20]. Although effective, competition among primer pairs in multiplexed reactions often presents a significant technical challenge and differentiation/quantification of each amplicon relies on the use of target-specific (e.g., Taqman) probes, each labelled with a distinct fluorochrome, and, thus, multi-channel thermocyclers increasing the costs of such methods. Furthermore, each of the multiplex-PCR assays currently available for toxigenic cyanobacteria targets a subset of taxa/toxin producers [19], [20], [21], and each has distinct running conditions. Their combination into a single assay has not yet been achieved.

Multiplexed-tandem PCR (MT-PCR) overcomes many of the limitations associated with standard PCR diagnostic methods [21]. This approach uses a primary ‘target enrichment’ phase, consisting of a 10–20 PCR cycles conducted in multiplex, followed by a parallelized analytical amplification phase, with nested primer pairs specific to each assay run in tandem. Because the multiplexed stage of the PCR is terminated prior to exponential amplification, competition among primer sets is minimized and quantitative capacity is maintained [22]. When coupled to a final melting-curve analysis, the entire process is conducted remotely using a liquid handling robot and real-time PCR thermocycler, allowing semi-automated processing and detection, identification and quantification in a single channel thermocycler using a standard fluorogenic dye (e.g., SYTO-9) [22]. In the present study, we assessed an MT-PCR assay for the rapid, automated detection of toxigenic cyanobacteria from blooms able to produce each of the major known classes of cyanotoxins (MCYs, NODs, CYNs and/or PSTs), and established its diagnostic sensitivity and specificity relative to conventional PCR and direct toxin detection.