Optimisation of DNA extraction from nasal lining fluid to assess the nasal microbiome using third-generation sequencing

BackgroundSampling nasal lining fluid (NLF) via nasosorption is minimally invasive and well tolerated, but the feasibility of assessing the nasal microbiome using this technique is unknown. However, low biomass makes airway samples particularly susceptible to issues related to contaminant DNA. In this study, we evaluated the suitability of DNA isolated using methodologies for low-biomass respiratory samples and assessed how well lining fluid collected via nasosorption captures the nasal microbial diversity and composition compared to the traditional swab sampling approach. MethodsNasal swabs and NLF were collected from adult volunteers. DNA was extracted from a mock microbial community and NLF using a column-based kit (ZymoBIOMICS), a precipitation-based kit (Qiagen), or a previously published precipitation-based method. Quality and quantity of DNA was assessed and short-read 16S rRNA sequencing performed to assess feasibility and extraction bias. An optimised extraction methodology was then used to extract DNA from NLF and nasal swabs, and full-length 16S rRNA sequencing performed to compare microbial profiles between NLF and nasal swabs. Taxonomy was assigned using the nf-core/ampliseq pipeline, the PacificBiosciences/pb-16S-nf pipeline, or the software Emu, and downstream analyses were performed using R packages phyloseq and mixOmics. ResultsAll extraction methods recovered DNA from the mock community, but only precipitation-based methods yielded sufficient DNA from NLF. Extraction methodologies significantly affected microbial profiles, with mechanical lysis needed to minimize bias against specific genera. Profiles obtained from NLF and swabs were comparable with long-read sequencing. ConclusionsOur findings demonstrate the feasibility of profiling the nasal microbiome using NLF collected via nasosorption and validated two extraction methodologies as suitable for full-length 16S rRNA sequencing of low-biomass respiratory samples. Our data demonstrate the importance of unbiased DNA extraction methodologies in low-biomass respiratory samples, and the subsequent impact of DNA extraction on observed microbial profiles. Additionally, we demonstrated NLF may be an appropriate surrogate samples for nasal swabs to assess the nasal microbiome using 16S rRNA sequencing.