Evaluating detection limits of next-generation sequencing for the surveillance and monitoring of international marine pests

Abstract

Most surveillance programmes for marine invasive species (MIS) require considerable taxonomic expertise, are laborious, and are unable to identify species at larval or juvenile stages. Therefore, marine pests may go undetected at the initial stages of incursions when population densities are low. In this study, we evaluated the ability of the benchtop GS Junior™ 454 pyrosequencing system to detect the presence of MIS in complex sample matrices. An initial in-silico evaluation of the mitochondrial cytochrome c oxidase subunit I (COI) and the nuclear small subunit ribosomal DNA (SSU) genes, found that multiple primer sets (targeting a ca. 400 base pair region) would be required to obtain species level identification within the COI gene. In contrast a single universal primer set was designed to target the V1-V3 region of SSU, allowing simultaneous PCR amplification of a wide taxonomic range of MIS. To evaluate the limits of detection of this method, artificial contrived communities (10 species from 5 taxonomic groups) were created using varying concentrations of known DNA samples and PCR products. Environmental samples (water and sediment) spiked with one or five 160 hr old Asterias amurensis larvae were also examined. Pyrosequencing was able to recover DNA/PCR products of individual species present at greater than 0.64% abundance from all tested contrived communities. Additionally, single A. amurensis larvae were detected from both water and sediment samples despite the co-occurrence of a large array of environmental eukaryotes, indicating an equivalent sensitivity to quantitative PCR. NGS technology has tremendous potential for the early detection of marine invasive species worldwide.

Figure 1. Detailed experimental design.

A) DNA samples from ten species were pooled together at equimolar (T1), decreasing (T2), and increasing concentrations (T3); each treatment was then PCR-amplified using specific fusion primers. B) Three distinct PCR-amplifications were run for each species individually, using distinct fusion primers. PCR products with identical primer tags were then pooled together at varying concentrations (T4, T5, T6). C) One water and one sediment samples, were collected; each sample was divided into three sub-samples and spiked with either 1 larva (T7, T9), 5 larvae (T8, T10), or no larva (controls T11, T12) of the Northern-Pacific seastar Asterias amurensis (genus Asterias does not occur in New Zealand), and PCR-amplified. Treatments 1 to 6 and 7 to 12 were pooled together and analysed in multiplex on the 454 GS Junior™ pyrosequencer.

Figure 2. Contrived community experiment.

Histograms show the number of recovered sequences per investigated species (circled numbers; see Figure 1) and for pooled DNA (T1, T2, T3) and pooled PCR (T4, T5, T6) treatments at A) equimolar, B) decreasing, and C) increasing concentration of starting material (shown in relative abundance of DNA/PCR products; see doted line and vertical scale on right of graphs). Two species, Perna perna and Perna canaliculus, were pooled together due to lack of SSU marker differentiation.

Figure 3. Environmental DNA/Spiking experiment.

Pie charts depict the proportion of marine phyla identified via BLASTn searches from environmental samples spiked with 1 larva (T7, T9), 5 larvae (T8, T10) or no larva (controls T11, T12) of Asterias amurensis (i.e., phylum Echinodermata; shown in bold) from A) one water and B) one sediment sample. A detailed taxonomic list of marine taxa, sequences lengths, minimum e-values, mean similarity, and number of best sequence hits to known NCBI sequences are shown for each treatment in Table S2.