Comparison of Two 16S rRNA Primers (V3-V4 and V4-V5) for Studies of Arctic Microbial Communities

Abstract

Microbial communities of the Arctic Ocean are poorly characterized in comparison to other aquatic environments as to their horizontal, vertical, and temporal turnover. Yet, recent studies showed that the Arctic marine ecosystem harbors unique microbial community members that are adapted to harsh environmental conditions, such as near-freezing temperatures and extreme seasonality. The gene for the small ribosomal subunit (16S rRNA) is commonly used to study the taxonomic composition of microbial communities in their natural environment. Several primer sets for this marker gene have been extensively tested across various sample sets, but these typically originated from low-latitude environments. An explicit evaluation of primer-set performances in representing the microbial communities of the Arctic Ocean is currently lacking. To select a suitable primer set for studying microbiomes of various Arctic marine habitats (sea ice, surface water, marine snow, deep ocean basin, and deep-sea sediment), we have conducted a performance comparison between two widely used primer sets, targeting different hypervariable regions of the 16S rRNA gene (V3-V4 and V4-V5). We observed that both primer sets were highly similar in representing the total microbial community composition down to genus rank, which was also confirmed independently by subgroup-specific catalyzed reporter deposition-fluorescence in situ hybridization (CARD-FISH) counts. Each primer set revealed higher internal diversity within certain bacterial taxonomic groups (e.g., the class Bacteroidia by V3-V4, and the phylum Planctomycetes by V4-V5). However, the V4-V5 primer set provides concurrent coverage of the archaeal domain, a relevant component comprising 10-20% of the community in Arctic deep waters and the sediment. Although both primer sets perform similarly, we suggest the use of the V4-V5 primer set for the integration of both bacterial and archaeal community dynamics in the Arctic marine environment.

Keywords: Arctic Ocean; amplicon sequencing; method comparison; microbial communities; molecular observatory; universal primers.

FIGURE 1

Chao1 richness estimates in the different sample types. Different primer sets represented by colors and shapes. Please note the differences of y-axis between the panels.

FIGURE 2

Major taxonomic families in V3–V4 and V4–V5 datasets. The x-axis represents the total sequence proportion of each family in V3–V4 (left panel) and V4–V5 (right panel) datasets. The numbers at each column represent the number of observed ASVs affiliated with each taxonomic family. Different taxonomic classes are represented by color code. Only families that comprised at least 1% of sequences in at least one of the datasets were included in the visualization.

FIGURE 3

Taxonomic compositions of the microbial communities. Different taxonomic classes are represented by color code. Classes with sequence proportion below 2% were classified as “Other taxa”.

FIGURE 4

Non-metric Multi-dimensional Scaling (NMDS) of the microbial communities in merged V3–V4 and V4–V5 datasets, based on Jensen-Shannon Divergence. The different types of samples are represented by colors, and the primer set are represented by shapes. Ellipses encompass clustering of each microbiome type with normal confidence of 0.95.