Analysis of phytoplankton community structure using the metabarcoding method Korean Waters

Abstract

Phytoplankton are photosynthesizing microscopic organisms whose major function in aquatic food web is the primary production, the creation of organic compounds from carbon dioxide dissolved in the water. Their community structure is the result of the complicate interactions of various biological components, and its change means the change of a whole ecosystem where they involved. From the reason, it is inevitable to make various types of survey to monitor the change of phytoplankton community structure. Typical surveys have mainly depended on the microscopic observation, which require a significant time and efforts to obtain accurate data. In addition, it has been almost impossible to identify and quantify the nano- or pico-sized phytoplankton. For these reasons, a large scale survey of phytoplankton has been difficult and chlolophyl a are generally measured to quantify them. For the phytoplankton in the ocean, images from the satellite are alternatively often used, which represent the relative amount of phytoplankton on the surface water. However, those do not provide the species information and its community structure and there is a need for a faster and more reliable standard method of performing detailed analysis of phytoplankton community structures. Due to the recent advance in DNA sequencing and bioinformatics analysis, researchers can analyze the biodiversity simply by massive sequencing of DNA barcodes that are amplified from environmental samples, a process known as “metabarcoding”. Here, we first modified the plastid 23S rDNA primer set targeting photosynthetic microorganism, which was originally designed by Sherwood and Presting, 2007. The modified 23S primer set (Yoon’s 23S primer set) was designed to be optimized in the MiSeq platform, whose products size were between 404 ~ 411 bp. The next generation sequencing (NGS) data of the six water samples collected from the East/Japan Sea in May and July 2014 showed that the Yoon’s 23S primer set is reliable to analyze the phytoplankton community structure by the MiSeq planform. Total 345 operational taxonomic units (OTUs) were identified which covered most of the prokaryotic and eukaryotic algal phyla including Dinophyta, Rhodophyta, Ochrophyta, Chlorophyta, Streptophyta, Cryptophyta, Hapotphyta and Cyanophyta. Analysis of algal proportions in the sea samples showed that community structure differed depending on location, depth and season. Howver, Yoon’s 23S primer set was not good enough to quantify the phytoplankton species for their cross-reactivity for heterotrophic bacteria. From the reason, we further modified Yoon’s 23S primer set to amplify exclusively photosynthetic phytoplankton. Primer design was based on analysis of 1,473 phytoplankton and bacterial 23S ribosomal DNA sequences to increase taxon coverage and specificificity for the phytoplankton (Kang’s 23S primer set). Kang’s primer set showed the high specificity to phytoplankton sequences, which exclude heterotrophic bacterial sequences almost completely as well as increase the numbers of phytoplankton OTUs and its contigs. Using the Kang’s 23 primers, we analyzed phytoplankton community structure in the Nakdong River, which may be further used for the management of water quality of the river. As the result of two NGS results amplified by 16S and Kang’s 23S primers, 16S primer set predominantly amplified prokaryotic heterotroph sequences. By contrast, no heterotrophic bacteria was amplified by Kangs’s 23S primers, which proved that this primers effectively discriminate heterotrophs from autotrophs. In addition, Kang’s 23S primers produced 4 times higher numbers of eukaryotic algal OTUs than those by 16S primers. Although higher heterotrophic bacterial OTUs were identified, cyanobacteria occupied more than half of total biomass in all examined sites. According to the proportions of the contig numbers, 50 % of biomass of algae changes by the season and the other 50 % were detected throughout the year. Interestingly, two sample sites with similar environmental factors showed high similarity in phytoplankton community structure regardless of distance. Waters collected from the region (Geoje, Gyengsangnamdo) with the algal blooming were also analyzed by the Kang’s 23S primer set. First, we identified that biomass of heterotrophic bacteria was much higher than the harmful algal species. Second, we identified that community structure between the center of blooming region and water near the blooming region were extremely similar each other compared with those of distantly located region from the blooming. As the result, we were able to know that community structure changes before we detect the change in color in waters. In addition, several bacterial species change in correlation with the harmful algal blooming. This result would provide the useful information about the process of the algal blooming and its early forecasting to prepare the harmful algal invasion in ocean. In conclusion, we developed a pipeline to analyze the phytoplankton community structure using NGS techniques and bioinformatic analysis. Currently designed Kang’s 23S primer set would be a strong tool to analyze the phytoplankton structure when it is used with previously used 16S primer set. Athough relatively less amount of data is available in 23S rDNA sequence database compared with those of 16S rDNA, this pipeline exhibited the highest performance in analyzing the phytoplankton sequence in the mixture of many other microoranisms, which would be used for the further quantitative analysis. NGS-based phytoplankton analysis would allow laboratories to analyze large numbers of samples at the reasonable cost and extremely low examination time, which never been possible before. In addition, we expect that this strategy will generate a large amount of novel data that could potentially change the established methods and tools that are currently used in the realms of oceanography and marine ecology. And this strategy is a powerful tool for understanding interaction between autotrophic phytoplankton and heterotrophic bacteria in complicated aquatic microbial environments.