Combining Nanopore Sequencing with Recombinase Polymerase Amplification Enables Identification of Dinoflagellates from the Alexandrium Genus, Providing a Rapid, Field Deployable Tool

Abstract

The armoured dinoflagellate Alexandrium can be found throughout many of the world's temperate and tropical marine environments. The genus has been studied extensively since approximately half of its members produce a family of potent neurotoxins, collectively called saxitoxin. These compounds represent a significant threat to animal and environmental health. Moreover, the consumption of bivalve molluscs contaminated with saxitoxin poses a threat to human health. The identification of Alexandrium cells collected from sea water samples using light microscopy can provide early warnings of a toxic event, giving harvesters and competent authorities time to implement measures that safeguard consumers. However, this method cannot reliably resolve Alexandrium to a species level and, therefore, is unable to differentiate between toxic and non-toxic variants. The assay outlined in this study uses a quick recombinase polymerase amplification and nanopore sequencing method to first target and amplify a 500 bp fragment of the ribosomal RNA large subunit and then sequence the amplicon so that individual species from the Alexandrium genus can be resolved. The analytical sensitivity and specificity of the assay was assessed using seawater samples spiked with different Alexandrium species. When using a 0.22 µm membrane to capture and resuspend cells, the assay was consistently able to identify a single cell of A. minutum in 50 mL of seawater. Phylogenetic analysis showed the assay could identify the A. catenella, A. minutum, A. tamutum, A. tamarense, A. pacificum, and A. ostenfeldii species from environmental samples, with just the alignment of the reads being sufficient to provide accurate, real-time species identification. By using sequencing data to qualify when the toxic A. catenella species was present, it was possible to improve the correlation between cell counts and shellfish toxicity from r = 0.386 to r = 0.769 (p ≤ 0.05). Furthermore, a McNemar's paired test performed on qualitative data highlighted no statistical differences between samples confirmed positive or negative for toxic species of Alexandrium by both phylogenetic analysis and real time alignment with the presence or absence of toxins in shellfish. The assay was designed to be deployed in the field for the purposes of in situ testing, which required the development of custom tools and state-of-the-art automation. The assay is rapid and resilient to matrix inhibition, making it suitable as a potential alternative detection method or a complementary one, especially when applying regulatory controls.

Keywords: Alexandrium; HABs; RPA; VolTRAX; aquaculture; food safety; harmful algal bloom; in-field sequencing; nanopore sequencing; paralytic shellfish poisoning; saxitoxin.

Figure 1

Graphical representation of ribosomal DNA cassette, location targeted by primers and example alignment of primer regions. Positions in the alignment matching the primers are represented by a dot and variants are shown as letters.

Figure 2

Primer performance matrix. Numbers refer to reaction number with gel electrophoresis bands at ~500 bp. (original Gel available in Supplementary File S1).

Figure 3

Environmental samples analysed by capillary electrophoresis showing a (i) gel and (ii) chromatogram for: (A) a negative sample with an interference peak from non-specific amplification, (B) a positive sample with no interference peak, and (C) a sample experiencing very low levels of amplification and a significant level of interference. Note: the peaks at 25 and 1500 bp are reference peaks. Concentrations for the amplicon and total concentration are shown.

Figure 4

Sequences from the mock community sample spiked with Alexandrium minutum cells were aligned against references for Alexandrium spp. using MinKNOW 5.4.3.

Figure 5

Phylogenetic tree using a maximum likelihood approach and 10,000 bootstraps with branch lengths not shown (a tree including branch lengths is included in the Supplementary File S2). Sequences in bold were generated in this study.

Figure 6

Chronological workflow of sample preparation for VolTRAX barcoding and library preparation. * The time required for sequencing will be variable, primarily dependent on the pores available on the flow cell.

Figure 7

Environmental sample data from four sites experiencing sub action limit toxin events (<800 µgSTXeq/kg). * Shellfish flesh saxitoxin concentration result obtained by semi-quantitative HPLC-Fld. Red cells identify where toxin data were not generated that day; thus, data from closest sampling data were used.

Figure 8

Field deployable sample preparation tools: (A) component parts of Osci-lyser, (B) the Osci-lyser assembled and ready to use, (C) USB adaptor for battery, providing 1.0- and 2.1-amp outputs, (D) Battery powering Qubit 2.0.

Figure 9

Field deployable sequencing device: (A) XavION sequencing computer, (B) External Power supply, (C) 200-watt solar panel. (i) deployed for use, (ii) packed for transport.