Estimating the bias related to DNA recovery from hemp stems for retting microbial community investigation

The industrial hemp plant Cannabis sativa is a source of vegetable fiber for both textiles and biocomposite applications. After harvesting, the plant stems are laid out on the ground and colonized by microorganisms (bacteria and fungi) naturally present in the soil and on the stems. By producing hydrolytic enzymes that degrade the plant wall polymers, the natural cement that binds the fiber bundles together is removed, thus facilitating their dissociation (retting process) which is required for producing high-performant fibers. To investigate temporal dynamics of retting microbial communities (density levels, diversity, and structure), a reliable protocol for extracting genomic DNA from stems is mandatory. However, very little attention has been paid to the methodological aspects of nucleic acid extraction, although they are crucial for the significance of the final result. Three protocols were selected and tested: a commercial kit (FastDNA™ Spin Kit for soil), the Gns-GII procedure, and a custom procedure from the Genosol platform. A comparative analysis was carried out on soil and two different varieties of hemp stem. The efficiency of each method was measured by evaluating both the quantity and quality of the extracted DNA and the abundance and taxonomy of bacterial and fungal populations. The Genosol protocol provides interesting yields in terms of quantity and quality of genomic DNA compared to the other two protocols. However, no major difference was observed in microbial diversity between the two extraction procedures (FastDNA™ SPIN Kit and Genosol protocol). Based on these results, the FastDNA™ SPIN kit or the Genosol procedure seems to be suitable for studying bacterial and fungal communities of the retting process. It should be noted that this work has demonstrated the importance of evaluating biases associated with DNA recovery from hemp stems. KEY POINTS: • Metagenomic DNA was successfully extracted from hemp stem samples using three different protocols. • Further evaluation was performed in terms of DNA yield and purity, abundance level, and microbial community structure. • This work exhibited the crucial importance of DNA recovery bias evaluation.