The Application of Nanopore Sequencing Technology to the Study of Dinoflagellates: A Proof of Concept Study for Rapid Sequence-Based Discrimination of Potentially Harmful Algae

Abstract

Harmful algal blooms (HABs) are a naturally occurring global phenomena that have the potential to impact fisheries, leisure and ecosystems, as well as posing a significant hazard to animal and human health. There is significant interest in the development and application of methodologies to study all aspects of the causative organisms and toxins associated with these events. This paper reports the first application of nanopore sequencing technology for the detection of eukaryotic harmful algal bloom organisms. The MinION sequencing platform from Oxford Nanopore technologies provides long read sequencing capabilities in a compact, low cost, and portable format. In this study we used the MinION to sequence long-range PCR amplicons from multiple dinoflagellate species with a focus on the genus Alexandrium. Primers applicable to a wide range of dinoflagellates were selected, meaning that although the study was primarily focused on Alexandrium the applicability to three additional genera of toxic algae, namely; Gonyaulax, Prorocentrum, and Lingulodinium was also demonstrated. The amplicon generated here spanned approximately 3 kb of the rDNA cassette, including most of the 18S, the complete ITS1, 5.8S, ITS2 and regions D1 and D2 of the 28S. The inclusion of barcode genes as well as highly conserved regions resulted in identification of organisms to the species level. The analysis of reference cultures resulted in over 99% of all sequences being attributed to the correct species with an average identity above 95% from a reference list of over 200 species (see Supplementary Material 1). The use of mock community analysis within environmental samples highlighted that complex matrices did not prevent the ability to distinguish between phylogenetically similar species. Successful identification of causative organisms in environmental samples during natural toxic events further highlighted the potential of the assay. This study proves the suitability of nanopore sequencing technology for taxonomic identification of harmful algal bloom organisms and acquisition of data relevant to the World Health Organisations "one health" approach to marine monitoring.

Keywords: MinION; alexandrium; dinoflagellate; eDNA; harmful algal bloom; nanopore sequencing; sequencing.

FIGURE 1

A simplified depiction of the rDNA cassette including the approximate location of both forward and reverse primers for the MinION sequencing and Sanger sequencing PCR reactions (note not to scale).

FIGURE 2

The results generated from analysis of pure cultures, showing the percentage of reads aligning to different spp. on the reference list for: (A) EPI2ME platform and (B) Custom data analysis pipeline.

FIGURE 3

Results from EPI2ME alignment after MinION sequencing for: (A) Spiked environmental samples (due to the larger abundance of A. tamarense sequences the X axis to be split so as to see of lower copy number species. (B) Environmental samples from each site as well as a negative control. (C) An HPLC-Fld chromatogram showing saxitoxin profile from site 1. (Note: sequence alignments shorter than 1000 bp removed for both A and B analyses).

FIGURE 4

(A) Proportional representation of the number of sequences aligning to the 10 most common dinoflagellate genera. (B) Depiction of number of sequences aligned to each species of Alexandrium as a percentage of total reads for the Site 1 environmental sample, with different column colors represents volume of data used for analysis with an approximate log2 between each dataset used. (C) Chi-squared distribution of the A. minutum data throughout the changes in data volume.

FIGURE 5

(A,B) Provide graphical representation of data generated by the EPI2ME custom alignment tool for environmental sites 4 and 5, cell count and toxin levels are also quoted. (C,D) Show respective chromatograms providing toxin profile for each site.

FIGURE 6

Frequency of nucleotides and position of candidate single nucleotide polymorphisms (SNPs) in each reference species using NanoPipe. A total of 5000 randomly extracted reads were used as the query and the consensus was used as the target.