Optimization and Testing of a Commercial Viability PCR Protocol to Detect IEscherichia coli/I in Whole Blood

Bacteremia, specifically if progressed to sepsis, poses a time-sensitive threat to human and animal health. Escherichia coli is a main causative agent of sepsis in humans. The objective was to evaluate a propidium monoazide (PMA)-based viability PCR (vPCR) protocol to detect and quantify live E. coli from whole blood. We optimized the protocol by adding a eukaryotic-specific lysis step prior to PMA exposure, then used spiking experiments to determine the lower limit of detection (LOD) and linear range of quantification. We also compared the vPCR quantification method to standard colony count of spiked inoculum. Lastly, we calculated percent viability in spiked samples containing 50% live cells or 0% live cells. The LOD was 10[sup.2] CFU/mL for samples containing live cells only and samples with mixed live and heat-killed cells. The linear range of quantification was 10[sup.2] CFU/mL to 10[sup.8] CFU/mL (R[sup.2] of 0.997) in samples containing only live cells and 10[sup.3] CFU/mL to 10[sup.8] CFU/mL (R[sup.2] of 0.998) in samples containing live plus heat-killed cells. A Bland–Altman analysis showed that vPCR quantification overestimates compared to standard plate count of the spiked inoculum, with an average bias of 1.85 Log[sub.10] CFU/mL across the linear range when only live cells were present in the sample and 1.98 Log[sub.10] CFU/mL when live plus heat-killed cells were present. Lastly, percent viability calculations showed an average 89.5% viable cells for samples containing 50% live cells and an average 19.3% for samples containing 0% live cells. In summary, this optimized protocol can detect and quantify viable E. coli in blood in the presence of heat-killed cells. Additionally, the data presented here provide the groundwork for further development of vPCR to detect and quantify live bacteria in blood in clinical settings.