2015 Volume 62 Issue 1 Pages 65-72
New sequencing technologies such as the Roche 454 pyrosequencing platform have made it possible to obtain millions of sequence reads in a single experiment, and massively parallel sequencing (MPS) is currently revolutionizing the survey of eukaryotic diversity, as this technology enables the detection of thousands of operational taxonomic units (OTUs) of eukaryotes from various ecosystems and facilitates the detection of low-abundance populations in complex eukaryote communities. However, this new technology brings with it different sources of sequencing error to traditional dideoxy capillary sequencing, such as ambiguity of homopolymer length, particularly for long homopolymers. MPS of PCR amplicon libraries can lead to overestimates of microbial diversity through the generation of low-frequency, error-prone reads. For this reason, empirical studies to evaluate per-base accuracy of MPS and to identify factors that can be used to eliminate low-quality reads have been employed. The nuclear 18S-rRNA gene has been used as a metabarcoding marker in MPS-based environmental surveys for plankton biodiversity research. However, different hypervariable regions have been used in different studies, and their utility has been debated among researchers. In this study, detailed investigations of 18S-rRNA were carried out, namely, the number of sequences deposited in international nucleotide sequence databases (INSDs), and the amplicon sequence variability among the three variable regions, V1–3, V4–5, and V7–9, using in silico PCRs based on INSDs. We also examined the taxonomic identification power, using MPS-based environmental surveys of a planktonic eukaryote community, to determine which region is more useful for MPS-based monitoring. We succeeded in concluding that the V1‒3 region will be the best one to apply to MPS-based monitoring of natural eukaryote communities in future, as the registration number of sequences in INSDs increases.
In this study, we compared the eukaryote biodiversity between Hiroshima Bay and Ishigaki Island in Japanese coastal waters by using the MPS-based technique to collect preliminary data. The relative abundance of Alveolata was highest in both localities, and the second highest groups were Stramenopiles, Opisthokonta, or Hacrobia, which varied depending on the samples considered. For microalgal phyla, the relative abundance of operational taxonomic units (OTUs) and MPSs was highest for Dinophyceae in both localities, followed by Bacillariophyceae in Hiroshima Bay, and by Bacillariophyceae or Chlorophyceae in Ishigaki Island. The number of detected OTUs in Hiroshima Bay and Ishigaki Island was 290 and 236, respectively, and ca. 40％ of the OTUs were common between the two localities. In the non-metric multidimensional scaling analysis, the samples from the two localities were plotted in different positions, reflecting geographic differences in biodiversity. Thus, we succeeded in demonstrating biodiversity differences between the two localities, although the read numbers of the MPSs were generally low. The MPS-based technique shows a great advantage of high performance by detecting several hundreds to thousands of OTUs simultaneously.
