Identification Of Human Gut Microbial Markers In Patients With Peripheral Artery Disease Using Different Sample Preservation Methods

Introduction: The gut microbiome is associated with cardiovascular diseases, but there is a paucity of microbial markers associated specifically with peripheral artery disease (PAD). We sought to identify gut microbial markers associated with PAD and to evaluate different sample preservation methods for future biomarker discovery. Methods: This is a single-center cross-sectional study of participants with and without documented PAD who performed home fecal sample collection using 3 preservation methods: immediately frozen (FR), Cary-Blair media (CB), and OMNIgene-GUT (GT). Samples were collected and stored at home for up to 4 days before being returned to the lab and stored at -80°C. In order to study the direct effects of room temperature (RT) storage on CB and GT samples, 5 samples from each group were stored for 0, 24, and 48 hours at RT before sample processing. Microbial communities were profiled using 16S rRNA (V4) gene amplicon sequencing; sequence data were analyzed using QIIME2 and Phyloseq in R. Taxonomic annotation was done using SILVA database and microbiome functional prediction was conducted using PICRUSt2 software package. Results: Thirty-three participants (14 PAD and 19 non-PAD) returned samples using all preservation methods. There were no significant differences in alpha diversity between PAD and non-PAD samples. FR samples had overall lower ASV richness compared to CB and GT. Microbial community structures were significantly different between PAD and non-PAD samples (PERMANOVA p<0.05), but CB and GT were less sensitive than FR in detecting the differences between PAD and non-PAD based on rare lineages. Paired differences in community structure were significantly different (p<0.05) between PAD and non-PAD after 24 and 48 hours at RT. Differentially abundant features were not consistently identified across all preservation methods, and these methods were also associated with differences in differential expression patterns of predicted functional gene content between groups. Conclusions: This study identifies gut microbial features associated with PAD and demonstrates the impact of sample preservation methods on these features. Ongoing work will integrate clinical variables in a larger cohort to validate these findings.