Rapid, inexpensive measurement of synthetic bacterial community composition by Sanger sequencing

Abstract

AbstractSimple synthetic bacterial communities are powerful tools for studying microbial ecology and evolution, as they enable rapid iteration between controlled laboratory experiments and theoretical modeling. However, their utility is hampered by the lack of fast, inexpensive, and accurate methods for quantifying bacterial community composition. For instance, while next-generation amplicon sequencing can be very accurate, high costs (>$30 per sample) and turnaround times (>1 month) limit the nature and pace of experiments. Here, we introduce a new approach for quantifying composition in synthetic bacterial communities based on Sanger sequencing. First, for a given community, we PCR-amplify a universal marker gene (here, the 16S rRNA gene), which yields a mixture of amplicons. Second, we sequence this amplicon mixture in a single Sanger sequencing reaction, which produces a “mixed” electropherogram with contributions from each community member. We also sequence each community member’s marker gene individually to generate “individual” electropherograms. Third, we fit the mixed electropherogram as a linear combination of time-warped individual electropherograms, thereby allowing us to estimate the fractional amplicon abundance of each strain within the community. Importantly, our approach accounts for retention-time variability in electrophoretic signals, which is crucial for accurate compositional estimates. Using synthetic communities of marine bacterial isolates, we show that this approach yields accurate and reproducible abundance estimates for two-, four-, and seven-strain bacterial communities. Furthermore, this approach can provide results within one day and costs ~$5 USD per sample. We envision this approach will enable new insights in microbial ecology by increasing the number of samples that can be analyzed and enabling faster iteration between experiments and theory. We have implemented our method in a free and open-source R package called CASEU (“Compositional Analysis by Sanger Electropherogram Unmixing”), available at https://bitbucket.org/DattaManoshi/caseu.