Effects of variation in sample storage conditions and swab order on 16S vaginal microbiome analyses

Abstract

ABSTRACT Technical bias is a pressing issue in microbiome research, and variability can be introduced at any stage from sample collection to figure generation. In this study, we aim to reduce biases in studying the human vaginal microbiome by examining the impact of sample storage buffer and multiple swabbing events using 16S rRNA gene amplicon sequencing data generated from vaginal swabs. We show that AssayAssure Genelock, a clinically relevant preservative for urine samples, is effective in preserving vaginal samples for microbiome studies. When comparing Genelock to 95% (vol/vol) ethanol and no preservative (air only), host variability explained more variance in both weighted and unweighted UniFrac measurements than the preservation method. We further examined the impact of three successive self-swabbing events, as the relatively low biomass nature of vaginal samples can inherently introduce bias. It is important to know if taking multiple swabs can provide replicable results and thus allow for additional technical replicates and an increased sample size. We found that up to three swabbing events do not introduce bias when examining the presence or absence of taxa but can explain 3% of the variability in the amount of taxa calculated. A study with more participants is warranted to provide further validation of these findings, but in producing this pilot study, we aim to continue laying the groundwork so that universally standardized and accessible studies can be created. IMPORTANCE The composition of the human vaginal microbiome has been linked to a variety of medical conditions including yeast infection, bacterial vaginosis, and sexually transmitted infection. The vaginal microbiome is becoming increasingly acknowledged as a key factor in personal health, and it is essential to establish methods to collect and process accurate samples with self-collection techniques to allow large, population-based studies. In this study, we investigate if using AssayAssure Genelock, a nucleic acid preservative, introduces microbial biases in self-collected vaginal samples. To our knowledge, we also contribute some of the first evidence regarding the impacts of multiple swabs taken at one time point. Vaginal samples have relatively low biomass, so the ability to collect multiple swabs from a unique participant at a single time would greatly improve the replicability and data available for future studies. This will hopefully lay the groundwork to gain a more complete and accurate understanding of the vaginal microbiome.