Quantification of the diversity sampling bias resulting from rice root bacterial isolation on popular and nitrogen-free culture media, using 16S amplicon barcoding

Abstract

Culturing bacteria from plant material is well known to introduce a strong bias compared to the real diversity present in the original samples. This bias is related to cultivability of bacteria, the chemical composition of media and culture conditions. The bias of recovery is often observed but was never quantified on different media using an amplicon barcoding approach comparing plant microbiota DNA extractions versus DNA extracted from serial dilutions of the same plant tissues grown on bacterial culture media. In this study, we i) quantified the culturing diversity bias using 16S amplicon barcode sequencing by comparing a culture-dependent approach (CDA) on rice roots on four popular bacterial media (Tryptone Soybean Agar-TSA-at two concentrations, 10% and 50%; a plant-based media with rice flour; Norris Glucose Nitrogen Free Medium-NGN; and Nitrogen Free -NFb) versus a culture-independent approach (CIA) assessed from DNA extracted directly on root and rhizosphere samples; ii) assessed enriched and missing taxa detected on the different media; iii) use biostatistics functional predictions to predict which metabolic profiles are enriched in the CDA and CIA. A comparative analysis of the two approaches revealed that among the 22 phyla present in the microbiota of the studied rice root samples, only five were present on the culture media approach (Proteobacteria, Firmicutes, Bacteroidetes, Actinobacteria, Verrucomicrobia). The Proteobacteria phylum was the most abundant in all cultured media samples, showing a high enrichment of gamma-Proteobacteria. The diversity of the combined culture media represented about 1/3 of the diversity of the total microbiota, and its genus diversity and frequency was documented. The functional prediction tool (PiCrust2) detected an enrichment of nitrogenase enzyme in bacterial taxa sampled from Nitrogen-free media, validating its predictive capacity. Further functional predictions also showed that the CDA missed mostly anaerobic, methylotrophic, methanotrophic and photosynthetic bacteria compared to the culture independent approach, delivering valuable insights to design ad-hoc culture media and conditions to increase cultivability of the rice-associated microbiota.