Rapid, Culture-Free Detection of Staphylococcus aureus Bacteremia

Abstract

S. aureus bacteremia (SAB) is a common condition with high rates of morbidity and mortality. Current methods used to diagnose SAB take at least a day, and often longer. Patients with suspected bacteremia must therefore be empirically treated, often unnecessarily, while assay results are pending. In this proof-of-concept study, we describe an inexpensive assay that detects SAB via the detection of micrococcal nuclease (an enzyme secreted by S. aureus) in patient plasma samples in less than three hours. In total, 17 patient plasma samples from culture-confirmed S. aureus bacteremic individuals were tested. 16 of these yielded greater nuclease assay signals than samples from uninfected controls or individuals with non-S. aureus bacteremia. These results suggest that a nuclease-detecting assay may enable the rapid and inexpensive diagnosis of SAB, which is expected to substantially reduce the mortality and morbidity that result from this condition.

Fig 1. Inactivation of MN inhibitory activity with a heat treatment step.

Human serum and plasma that was spiked with calcium chloride and the indicated amounts of MN was either heated to 90°C or incubated at 4°C, for 20 minutes (A). Unheated samples and supernatants of heated samples were then incubated with the 11mer PolyT FAM probe and fluorescence was measured. Note the clear differences between the heated and unheated samples with MN concentrations below 10⁻² units/μl (panel in A). Error bars indicate standard deviations of triplicate measurements. The incubation time needed to inactivate inhibitory antibodies in plasma at 90 and 100°C was determined as follows. Human plasma, spiked with 10⁻⁵ units/μl MN and 10 mM calcium chloride, was heated to 90°C (upper panel in B) or 100°C (lower panel in B) for indicated amounts of time, centrifuged, and supernatants were incubated with the 11mer polyT FAM probe. Probe incubated with buffer only was included as control. Error bars indicate standard deviations of triplicate measurements.

Fig 2. Nuclease assay detection sensitivity in human plasma.

MN was immunoprecipitated from heat-treated human plasma that was spiked with MN (concentrations are indicated above), followed by incubation with a quenched fluorescent oligonucleotide probe (11mer PolyT FAM probe). Note the improvement in assay sensitivity when the immunoprecipitation step is included (gray) versus simply incubating heat-treated plasma directly with the probe (black). Fluorescence values are normalized to those of control samples in which no nuclease was included.

Fig 3. Detection of MN in SAB patient plasma.

First, the stability of MN in human plasma at 4°C was determined by spiking pooled human plasma with the indicated concentrations of MN and storing at 4°C for 3 days, followed by MN activity measurement (A). Equivalent amounts of MN were also added to pooled human plasma immediately before the nuclease assay for comparison. For the assay, calcium chloride was added to all samples, which were then heated to denature inhibitory antibodies. Supernatants of heated plasma samples were incubated with the 11mer PolyT FAM probe for 1 hour and fluorescence was measured with a plate-reader. All values are normalized to those produced by control samples in which nuclease was omitted. Error bars indicate standard deviations of triplicate measurements. In part B, MN activity assays were carried out with plasma specimens from S. aureus bacteremic (S. aureus infected) and individuals showing no signs of active infections (Presumed uninfected). 660 μl of plasma was used for each and the 11mer PolyT FAM probe was used for detection. Fluorescence values were normalized to those of control samples in which buffer was substituted for plasma; note normalization values included above each bar. Data shown are compiled from several independent experiments.

Fig 4. Sensitivity of MN detection with PolyT length and dye variants.

Each of the indicated probes was combined with the various indicated concentrations of MN (diluted in buffer) and incubated at 37°C for 1 hour and then fluorescence was measured. Fluorescence values are normalized to those of control samples in which no nuclease was included. Fluorescence values for the lowest MN concentrations are also plotted with a different scale for better differentiation of the lower values (inset). Error bars indicate standard deviations of triplicate measurements.

Fig 5. Detection of MN in plasma of patients with bacteremia.

Nuclease activity assays were carried out with plasma samples from uninfected patients (No growth (green bars)), patients with positive blood culture results for non-S. aureus pathogens (black and blue bars; pathogen is indicated) or for S. aureus (red bars). Blood used for blood cultures was drawn the same day (green bars without asterisks, blue and red bars) or one day prior to (green bars with asterisks and black bars) the blood drawn for the nuclease assays. 2 ml of plasma was used for each nuclease assay and the 4mer PolyT ATTO probe was used for detection. Fluorescence values are normalized to the average fluorescence value of control samples in which buffer was substituted for plasma. Data shown are compiled from several independent experiments. Some experiments are excluded from the compilation due to failed control reactions. Culture-confirmed SAB samples for which same day culture data are unavailable are also not included due to the difficulty in interpreting these data. The researcher carrying out the nuclease assays was blinded to the results of the culture-based assays until the nuclease assay results were obtained and reviewed.

Fig 6. Nuclease activity in culture supernatants of S. aureus, S. epidermidis and S. lugdunensis.

An overnight culture supernatant of each species was incubated with the 4mer PolyT ATTO probe for 1 hour at 37°C and fluorescence was measured. Fluorescence values are normalized to that of a control sample in which buffer was substituted for culture supernatant; note normalization values above each bar.