Comparing saliva collection and DNA extraction methods for saliva-based microbiome profiling

Aims: Mounting preclinical evidence demonstrates the importance of the human microbiome in health and disease. Saliva presents a particularly appealing medium for microbiome research due to its non-invasive collection and the availability of extensive biobanked samples across various conditions. However, methodological challenges remain- particularly regarding sample storage and the variability introduced by different nucleic acid extraction kits, which can exhibit selective affinities for certain bacterial taxa. In this study, we systematically compared multiple saliva collection and DNA extraction methods to optimize protocols for 16S rRNA-based microbiome profiling. Our approach incorporated rigorous quality control measures, including the analysis of water controls, differential abundance testing, and correlation analyses across groups, to identify the most reliable and reproducible methods for salivary microbiome characterization. Methods and results: We compared four commercially available kits for at-home saliva collection to determine their effectiveness at preserving the salivary microbiome following 1 week of storage at room temperature (RT). We also compared three commercially available DNA extraction kits marketed for salivary microbiome characterization. We discovered that the DNA extraction kit used significantly impacted the microbiome composition. One week of incubation in the preservative solution shifted the bacterial composition of the saliva. Additionally, we demonstrated that contaminants in the environment and kits reagents may be increased during the incubation period, posing a significant challenge. Conclusion: Our work demonstrates the feasibility of, and provides a framework for, microbiome characterization in saliva that applies to clinical and population-based studies. Our findings indicate that saliva microbiome studies using different extraction kits may introduce systematic biases, which should be accounted for when comparing results across studies. Using the same nucleic acid collection and extraction kits in various experiments is essential for reproducibility due to their different affinity to specific bacteria and contamination rates.