This study asked whether ratiometric pre-rRNA analysis, which was originally developed for analysis of drinking water, could also assess the viability of bacterial cells in human specimens. Although human fluids such as serum are rich in certain nutrients, they are usually limited in some required nutrients, and bacterial growth rates are sub-maximal in such environments. Therefore, we hypothesized that the provision of limiting nutrients would stimulate pre-rRNA synthesis in viable bacteria derived from such samples. The results showed this to be true for all organisms tested. When bacteria were not viable, as seen in P. aeruginosa after incubation in the first lot of serum, nutritional stimulation failed to elicit the production of pre-rRNA.
The five species used in this study were phylogenetically diverse, with the bacterial phyla Proteobacteria, Firmicutes, and Actinobacteria represented. They were also diverse physiologically, with laboratory generation times ranging from <1 hour to ≥24 hours and varying numbers of rRNA biosynthetic genes. All of them responded to nutritional stimulation in similar fashion, producing significant amounts of pre-rRNA in less than 1–2 generation times, even after extended incubation in growth-limiting conditions. S. aureus sometimes exhibited a delayed response to changing nutritional environments. Relative to the other two species, it required longer periods to drain pre-rRNA pools in serum, and its response to nutritional stimulation peaked at 2 hours rather than 1 hour. These observations might be explained by its relatively slow growth rate. Despite these variations, we have found that a 90 minute nutritional stimulation consistently induces a strong pre-rRNA signal in A. baumannii, P. aeruginosa, and S. aureus. Pre-rRNA synthesis is among the earliest steps in growth initiation, significantly outpacing DNA replication and cell division. The phylogenetic conservation of this phenomenon enhances the utility of molecular viability testing by ratiometric pre-rRNA analysis.
In practical terms, pathogen viability testing can be conducted in a clinical sample by dividing the sample into two aliquots, one of which is nutritionally stimulated while the other is held as a non-stimulated control. After brief stimulation, nucleic acid is extracted from both aliquots and subjected to RT-qPCR to quantify species-specific pre-rRNA. When the pre-rRNA increases in the stimulated sample relative to the control sample, the presence of viable cells of the targeted species is indicated. As shown previously 
, the magnitude of pre-rRNA upshift in viable cells is sufficient to enable their detection even when they are greatly outnumbered by inactivated cells. A potential confounding factor is that cells in stimulated and non-stimulated samples may differ with regard to pre-rRNA release upon lytic treatment. For example, stimulated cells might have weaker cell envelopes that release nucleic acid more readily than non-stimulated cells. Additionally, inhibitors present in serum might reduce the efficiency of pre-rRNA amplification in non-stimulated samples relative to stimulated samples. Either factor could create the appearance of increased pre-rRNA in the stimulated aliquot, even if no new pre-rRNA was produced. In the present study, these factors were controlled by normalizing pre-rRNA to genomic DNA, as in , , , S1
, and S2
. However, this control adds cost and complexity to the procedure, and in at least some cases it may not be needed, as seen in and S3
The nutritional stimulation step in pre-rRNA analysis (1–4 hours depending on the targeted species), and the requirement for two measurements to obtain ratiometric results, add complexity to the overall procedure. In samples for which the mere presence of pathogen nucleic acid is often a satisfactory indication of disease (for example, M. tuberculosis DNA in sputum), this added complexity would be disadvantageous. But for diagnostic indications that require differentiation of viable and dead cells (e.g. antibacterial treatment monitoring), then the speed, sensitivity, and specificity of pre-rRNA analysis may offer advantages. With additional development and automation, it may be possible to generate results faster and with less hands-on work.
The results in , although generated with simulated samples, illustrate the potential clinical value of pre-rRNA analysis. Viewed in isolation, the genomic DNA signals in would have suggested dense infections with P. aeruginosa and A. baumannii, and somewhat lower-grade infection with S. aureus. However, the P. aeruginosa cells were inactivated while the S. aureus cells were partially viable. Therefore, the latter might present a more serious threat to a patient if seen in a real sample. Ratiometric pre-rRNA analysis was able to make this distinction.