Comparative study of sample storage conditions on gut dysbiosis in peripheral artery disease

Background: Large-scale gut microbiome studies rely on fecal sample storage prior to batch sample preparation, sequencing, and analysis. Effects of storage methods have largely been studied using samples from healthy participants, where the microbial communities and the metabolic environment are in concordance. In diseased states, dysbiosis is more prone to environmental perturbation, which causes variable shifts in the communities. Cardiovascular diseases are associated with gut dysbiosis, but the effect of storage methods on the qualitative and quantitative aspects of dysbiosis is unknown. Thus, we examined the effects of 3 sample storage conditions on the fecal samples of patients with peripheral artery disease (PAD), a form of cardiovascular disease, and non-PAD controls. Methods and Results: This is a cross-sectional study of fecal samples collected from adults with PAD and non-PAD controls. All participants (12 non-PAD and 18 PAD) followed the home fecal sample collection protocol. Each sample was immediately frozen (IF), placed in modified Cary-Blair (CB), and stored in an OMNIgene•Gut vial. All samples were subjected to 16S rRNA gene amplicon sequencing of the hypervariable V4 region. A subset of glycerol stocks from IF and CB samples was thawed and cultured to compare revivification. We found significant differences in microbial composition and community structure between non-PAD and PAD groups based on storage conditions. Although we did not see the effect of an interaction term (disease group*storage condition) at the community level, we observed storage condition-specific differential abundance of genera in the PAD compared to the non-PAD group. The high number of differentially variable taxa in the PAD group samples further emphasize the need for standardizing storage conditions. The subset of samples stored in CB had less revivification potential than IF samples under both anaerobic and aerobic processing conditions. Conclusions: Sample storage conditions and room temperature storage time differentially affect the microbial communities of fecal samples and revivification of glycerol stocks from non-PAD and PAD groups. The effects of storage conditions can bias microbiome-related disease biomarker discovery. Careful consideration should be given to sample storage conditions when analyzing fecal samples from diseased populations and when combining data from cohorts with samples stored in different conditions.