Challenges and Solutions in Low-Biomass Respiratory Microbiome Profiling: A Workflow for Bronchoalveolar Lavage Fluid Sequencing in Guinea Pigs.

Abstract

Accurate profiling of the respiratory microbiome in low-biomass samples remains technically challenging due to host DNA contamination and limited microbial yield. This study aimed to optimize a methodological workflow for 16S rRNA sequencing of bronchoalveolar lavage fluid (BALF) obtained from guinea pig models with differing microbial statuses-specific pathogen-free (SPF) and conventionally bred (CON) animals. Using a comparative approach, we evaluated six commercial DNA extraction kits and tested different input DNA concentrations (1 ng vs. 0.5 ng) to enhance microbial detection while minimizing host DNA interference. Among the tested kits, only the ZymoBIOMICS DNA Microprep Kit yielded sufficient microbial DNA for downstream analysis. Real-time PCR and droplet digital PCR confirmed the microbial origin of the extracted DNA. Sequencing libraries were prepared from the V1-V3 regions of 16S rRNA genes and sequenced using the Illumina iSeq 100 platform. Taxonomic assignment and diversity metrics were analyzed using the MicrobAT pipeline. Our findings revealed significant differences in microbial composition between SPF and CON animals, notably in Mycoplasma abundance, which dominated the microbiota in CON but was nearly absent in SPF animals. Alpha and beta diversity metrics showed consistent stratification by animal group and input DNA concentration. However, a high proportion of unclassified reads-particularly in SPF samples-correlated strongly with sequences mapping to the Cavia porcellus genome, indicating substantial host DNA contamination. This study demonstrates the feasibility of microbiome profiling from low-biomass BALF samples in guinea pigs while highlighting the limitations of current sequencing and bioinformatic tools in distinguishing microbial from host-derived DNA. Our optimized workflow supports future respiratory microbiome studies in animal models and provides a foundation for improving host DNA depletion and reference databases tailored to non-human species.