Comparison of four DNA extraction methods for 16s rRNA microbiota profiling of human faecal samples

Growth in large population-based studies assessing contributions of the gut microbiota to health and disease requires high-throughput sample processing and analysis methods. This study assessed the impact that modifications to a commercially available magnetic bead based, semi-automated DNA extraction kit had on determination of microbial composition, relative to an established in-house method involving a combination of mechanical and chemical lysis. DNA was extracted from faecal samples from healthy adults (n = 12; 34-69 years), microbial composition was determined by V3-V4 16s rRNA sequencing and compared between extraction methods.

Diversity metrics did not differ between extraction methods. Differences in the relative abundance of key phyla, including a significantly lower abundance of the Firmicutes (p = 0.004) and higher relative abundance of the Bacteroidetes (p = 0.005) and Proteobacteria (p = 0.008) phyla were noted where the DNA extraction did not include additional chemical and mechanical lysis. Principal coordinate analysis of family and genera level data also suggested a potential for sample pre-processing to impact microbial composition. Observations of the potential for skewed microbial composition profiles from samples prepared using a semi-automated DNA extraction kit without additional sample pre-processing highlights a need for consideration of standardisation of methodological approaches to increase the comparability of microbial compositional data.