Optimizing rpoB metabarcoding for low-concentration bacterial DNA in eukaryotic matrices using nested PCR: application to insect microbiota

In recent years, microbial ecosystems have been mainly studied using 16S rRNA gene metabarcoding. However, this strategy does not provide sufficient taxonomic resolution and often fails to determine the bacterial species. The aim of this work was to develop a novel method to improve the amplification of samples with low bacterial DNA concentrations embedded in eukaryotic matrices, while allowing to keep species-level taxonomic resolution. To this end, we optimized a metabarcoding procedure by combining nested PCR and targeting of a region of the rpoB housekeeping gene. First, we validated our method using in silico approaches. Then, we tested it on commercial and custom-made mock samples. Finally, we applied this procedure to whole larvae of the lepidopteran insect Spodoptera frugiperda collected in the field in Senegal , and to their oral secretions (OS). We demonstrated that rpoB nested PCR outperforms rpoB direct PCR for the detection of bacterial amplicon sequence variants ASVs in low-concentration bacterial DNA samples (OS samples) or bacterial DNA embedded in complex insect matrices (larvae samples), while allowing species-level resolution, unlike the short amplicon from 16S rRNA genes, which is limited to genus-level identification. Finally, we propose a new strategy to facilitate the description of insect-associated microbiota, which could be extended to other host-associated microbiota.