Characterizing microbial community viability using propidium monoazide

Modern humans spend most of their life time in built environments (BE). Characterization of BE microbiomes is of great importance given the associations between microbial exposure and occupant health in indoor settings. Although many studies have explored the taxonomic composition of BE microbiomes using DNA sequencing, this method on its own suffers from an inability to discern viability. Non-viable microorganisms are likely of much greater prevalence and abundance in BEs than in natural settings due to the specific chemical characteristics of those environments. Previous studies have used propidium monoazide (PMA) in the pre-treatment of samples, coupled with downstream molecular profiling (e.g. qPCR or DNA sequencing), to characterize the viable fraction of a microbial community. While these studies have met with some success, most have focused entirely on interpretation of the resulting “viable” microbial community profiles without a systematic evaluation of the accuracy of this measurement.

Here, we present our work to rigorously benchmark “PMA-seq” (PMA treatment followed by 16S rRNA amplicon sequencing) as a screen for microbial viability in both synthetic and environmental microbial communities, ranging from simple in vitro co-culture to full-complexity BE communities. Our first validation focused on synthetic mixtures of living and heat-killed Escherichia coli and Streptococcus sanguinis in known proportions. PMA-seq was able to successfully reconstruct the communities of viable/heat-killed E. coli and S. sanguinis. We next evaluated the effects of biomass and sample diversity on PMA-seq by testing communities of variable biomass (computer screens and mice, soil, and saliva) and spiked them with known concentrations of viable and heat-killed E. coli. Against a background of realistically complex communities, viability was no longer accurately assessed by PMA-seq, with its performance largely affected by initial biomass and compositional diversity. Finally, we applied this technique to environmental swabs from the Boston subway system to identify and quantify living bacteria in that BE. Several taxa were significantly different after PMA treatment including Haemophilus genera, but not all samples respond consistently to PMA treatment. Overall, we revealed that PMA treatment was effective for qPCR or 16S readouts of pure cultures or very simple synthetic communities, but such usage may be premature and poorly quantify viable microbes in realistically complex community samples.

This study provides a comprehensive evaluation of the performance of PMA-treatment to test viability of microbes in BE communities. In the future, we plan to expand our work to the validation of other methodologies to distinguish between viable and non-viable microbes, such as RNA-based amplicon sequencing or metatranscriptomic sequencing. We expect characterization of viable microbes to further elucidate the interplay between microbes and their ecosystems/hosts, potentially unveiling new interactions between BE microbes and human health.