Comparison of one-step and two-step PCR for 16S Illumina MiSeq

Despite general adoption of DNA high-throughput sequencing of PCR-amplified microbial taxonomic marker gene fragments for ecological analyses, distinct approaches for preparing PCR amplicon libraries exist. One approach utilises long fusion primers (ca. 80 nucleotides) and a single PCR (one-step) while another utilises normal length primers (35 nucleotides) in a first PCR, before transferring diluted amplicons to a second PCR for barcode index incorporation (two-step). We sought to investigate whether this transfer of diluted amplicons risked creating artificially simplified communities with low alpha- and beta-diversity. In a cropping and forest soil with inherently contrasting diversity, one-step detected 2 – 3 times more amplicon sequence variants. Quantitative comparisons of community distribution demonstrated that two-step covered 55.6% and 61.7% of cropping and forest communities, respectively. While beta-diversity between replicates was lowest for two-step forest, indicating high reproducibility, Bray-Curtis distance between cropping and forest centroids was five times greater under one-step. Driving the differences between approaches was confirmed as quantitative underestimation of relatively minor taxa under two-step. These taxa were low in abundance, yet play important roles in carbon and nitrogen fixation, recalcitrant carbon breakdown, organic phosphorus cycling, and plant-microbe interactions. We conclude that for studies focussed on assessing diversity or investigating relatively minor yet functionally important taxa, one-step amplicon library preparation is advisable.