Approach to obtain live and dead bacterial cells
To obtain dead cells isopropanol was first tested. It has been reported that isopropanol increases permeability of the bacterial cell membrane and destroys protein function by denaturing them, thereby kills bacteria [18,19].
Bacterial cells from the same pre-culture were either treated with 70% isopropanol or with 0.9% NaCl solution. Isopropanol treated cells led to no colonies in the agar plating experiments, whereas expected number of colonies was obtained from the NaCl treated cells (data not shown). Therefore, cells treated with isopropanol were referred as dead cells in this study and NaCl solution treated cells as live cells. Furthermore, the differently treated cells exhibited similar values (with less than 10% difference) of optical density (OD). This result suggests that the cells, even if dead, kept structural integrity after the treatment with isopropanol. This suggestion was further supported by observation of similar numbers of green-colored cells (live) and red/yellow-colored cells (dead) with similar shape under the microscope (Additional file 1: Figure S1). Therefore, the accordingly treated cells were further used as live and dead cells for the staining tests.
Mixtures of different ratios of live and dead cells were stained with SYTO9 alone. The fluorescence intensity was measured with the microplate reader. As expected for staining with a membrane permeable dye like SYTO9, no difference in intensity was observed between live and dead cells of S. aureus (Figure 1a). However, for P. aeruginosa with the same total cell numbers 100% dead cells exhibited an 18-fold stronger signal than 100% live cells (Figure 1b). This finding is further supported by the intermediate signal intensity of the different mixtures, showing a linear increase with the increase of the fraction of dead cells. The effect of stronger SYTO9 staining of dead cells seems to be common for Gram-negative bacteria as we observed the same effect for Escherichia coli, but not for the Gram-positive Bacillus subtilis (data not shown).
SYTO9 signal after counterstaining with PI
To distinguish live cells from dead ones, PI was added to the mixtures having different live/dead ratios of SYTO9 stained cells. A clear reduction in SYTO9 staining was observed for the dead cells of both strains compared to control samples which were treated with NaCl solution (Figure 1). For 100% dead cells of S. aureus and P. aeruginosa the fluorescence intensity of SYTO9 was decreased 87% and 85%, respectively, compared to the control samples based on the measurement with the microplate reader. On the contrary, living cells were significantly less de-stained by the addition of PI, e.g. 5% reduction in SYTO9 signal for 100% S. aureus live cells and 20% for P. aeruginosa. Thus, the dead cells of S. aureus exhibited 9-fold weaker SYTO9 signal intensity than the living cells, whereas the dead cells of P. aeruginosa still displayed 2.7-fold higher SYTO9 intensity than the living ones after counterstaining with PI (Figure 1). These results demonstrate that the displacement of SYTO9 by PI takes place as expected. However, in P. aeruginosa even if the dead cells show strong reduction in SYTO9 fluorescence after PI counterstaining, they possess still stronger SYTO9 fluorescence than the living ones. Living cells showed no or only slight reduction in SYTO9 fluorescence after counterstaining, which is expected because PI should not enter intact cells to replace SYTO9.
During the experiments strong reduction of SYTO9 fluorescence with time was observed, which indicates that SYTO9 is prone for bleaching. Therefore, the possibility of SYTO9 bleaching was investigated by measuring green fluorescence of SYTO9 stained cells every 5 minutes. About 4-8% of the SYTO9 signal intensity was lost every 5 minutes, depending on the physiological state of the cell and cell number (Figure 2 relative values, Additional file 2: Figure S2 absolute values). Different trends can be observed. First, the reduction rate of SYTO9 signal decreases with higher cell numbers. Second, the reduction rate is higher for the same amount of dead compared to live cells. Interestingly, particular differences in bleaching were observed for live P. aeruginosa cells, which were also shown to be difficult to stain (Figure 1).
PI signal after counterstaining
Upon counterstaining, the PI signal in living S. aureus cells was almost zero after subtraction of background signals (cross-signal of SYTO9 and unbound PI signal), as expected for this membrane-impermeable dye. With increased proportion of dead to live cells the red PI fluorescence increased linearly (Figure 3). However, the absolute fluorescent intensity value was rather low. Unbound PI possessed strong background signal with a relative fluorescence intensity unit (RFU) of about 700. The dead cells exhibited a RFU of 1200 after the background subtraction (cross-signal of SYTO9 and unbound PI signal). The background signals of unbound PI could not be prevented in fluorescence readouts. Therefore, for reliable interpretation of the PI fluorescence data obtained from the microplate reader background controls and relatively high numbers of dead cells are needed. Precise quantitative determination of the amount of dead cells is hence rather difficult.
Microscopical examination of live/dead staining
Confocal laser scanning microscope (CLSM) was used to investigate individual cells stained with either SYTO9 alone or SYTO9/PI. The results gained from microscopy confirmed the data obtained with the microplate reader. Live and dead cells of S. aureus showed similar green fluorescence intensity when stained with SYTO9 alone, while live P. aeruginosa cells are stained clearly less than the dead ones (Figure 4, Additional file 3: Figure S3). Dead cells of both species, S. aureus and P. aeruginosa, exhibited red fluorescence after PI counterstaining (Figure 5). As expected, S. aureus cells that appear red after PI counterstaining show clearly weaker SYTO9 signal (Figure 5, Additional file 4: Figure S4). The mean integrated green fluorescence intensity was evaluated with CellProfiler software. It was found that dead S. aureus cells exhibited an almost 5-fold lower signal intensity compared to the living cells (Additional file 4: Figure S4d). Counterstaining of P. aeruginosa resulted in a strong reduction of SYTO9 fluorescence in dead cells. However, dead cells possessed much higher SYTO9 fluorescence than live cells before counterstaining. Therefore, the fluorescence reduction in dead cells after counterstaining only resulted in SYTO9 levels similar to that of living cells (Additional file 5: Figure S5).