Impact of DNA extraction methods on 16S rRNA-based profiling of bacterial communities in cheese

Numerous factors associated with sample preparation, DNA extraction, primer choice, sequencing platform and data analysis can affect the accuracy of 16S rRNA sequencing results. The DNA extraction method is considered critical for the success of sequencing as it can be the source of considerable variations in the analysis of the microbiome. In this study, the impact of various DNA extraction methods on the results of analysis of bacterial communities in cheese was evaluated. DNA was isolated from Mozzarella as a model cheese using optimized bead-based homogenization followed by different extraction procedures. Five commercial kits and two open-formula DNA extraction protocols were evaluated for amplicon sequencing of a 16S rRNA fragment of ~1460 bp. In addition, model cheese samples artificially contaminated by defined concentrations of Listeria monocytogenes and Escherichia coli, as representatives of Gram positive and Gram negative bacteria, were analysed. Six out of seven DNA extraction procedures were found to be able to provide amplifiable bacterial DNA suitable for 16S rRNA sequence analysis, but individual extraction procedures led to variable results. In particular, lysis supported with bead-beating led to a higher proportion of G+ bacteria in relative abundance profiles, probably because of the more efficient cell wall disruption. Artificially added bacterial species were reliably detected with a quantitative response. The results demonstrated a risk in comparing the data on bacterial communities in cheese when different DNA extraction protocols are used and highlighted the need to choose a standardized approach when comparison across multiple sequencing runs is required. •DNA extraction method is crucial for the success of 16S rRNA sequencing analysis.•Individual extraction procedures led to variable results of microbiome profiling.•Mechanically assisted lysis promoted higher representation of G+ bacteria.•Artificially added bacteria were detected with quantitative response.