All ESKAPE strains used in this study are clinical isolates and are listed in Table 1. These isolates were collected over the past 20 years directly from the clinical microbiology laboratory at the University at Buffalo, which performed strain identification by approved laboratory methodology. Isolates from de-identified plates were frozen and stored at −80°C. Enterococcus faecium and Staphylococcus aureus were grown in Trypticase soy broth (Sigma-Aldrich; TSB) or TSB agar. All other strains were cultured in Luria-Bertani (LB; BD Biosciences) broth or on LB agar. The incubation temperature for all strains was 37°C. Histatin 5 (Hst 5), FITC-labeled Hst 5 (F-Hst 5), and spermidine conjugated Histatin 5 (Hst 5-Spd) were synthesized by Genemed Synthesis Inc. (San Antonio, Texas). Spermidine was purchased from Sigma-Aldrich.

Bactericidal assays were performed by microdilution plate method as described for candidacidal assays (Jang et al., 2010), with some modifications. Briefly, a single colony of each strain was inoculated into 10 mL of media and grown for 16 h at 37°C, cultures were diluted to an OD₆₀₀ = 0.1 in fresh media and incubated with shaking at 37°C to mid-log phase (OD₆₀₀ ≈ 1.0). Cultures were spun down at 1800 × g for 3 min and washed three times with 10 mM sodium phosphate buffer (pH 7.4; NaPB). Cells were re-suspended (10⁷ cells/mL) in 200 μL NaPB (control) or in NaPB containing peptides (30 μM); then incubated at 37°C (except for S. aureus that was incubated at room temperature for optimal growth). Aliquots were removed after 1 min, 1 h and 5 h incubation, then were serially diluted with phosphate buffered saline (PBS; 137 mM NaCl, 10 mM sodium phosphate, 2.7 mM KCl, pH of 7.4); plated and incubated for 24 h at 37°C to visualize surviving CFUs. Assays were performed in triplicate. Percentage killing was calculated as [1 − number of colonies from peptide-treated cells/mean number of colonies from control cells] × 100%. Bactericidal assays were also performed in 100 mM NaPB and PBS.

Confocal microscopy was performed using live cells of six ESKAPE strains attached to chambered borosilicate coverglass (Lab-Tek #155411, 8 chamber). Cells in mid-log phase were collected and washed three times and re-suspended in 10 mM NaPB. Cells (2 × 10⁷ in 200 μL NaPB) were added to each well, allowed to settle for 15 min, then propidium iodide (PI; 2 μg/mL; Sigma) and F-Hst 5 (30 μM) were added. Time-lapse confocal images were recorded immediately after addition of peptide and PI. Confocal images were acquired with a Zeiss LSM510 Meta Confocal Microscope (Carl Zeiss, Germany) using a Plan Apochromat 63X/1.4 Oil Immersion objective. For E. faecium, S. aureus, K. pneumoniae, and E. cloacae, images were collected every 10 min; and for A. baumannii and P. aeruginosa every 2 min. In order to detect F-Hst 5 and PI simultaneously, the 488 nm line of the argon ion laser and a 561 nm DPSS laser were directed over an HFT UV/488/561 beam splitter, and fluorescence was detected using a Mirror or NFT 565 beam splitter in combination with a BP 500–550 filter for F-Hst 5 and an LP 575 or BP 650–710 filter for PI detection. ImageJ software was used for image acquisition and analysis.

To evaluate killing of bacterial biofilms by Hst 5, cells were inoculated at 1 × 10⁷ cells/ml into 96 well plates and grown overnight at 37°C and 90 RPM for 12 h to form biofilms. After 12 h, spent media was replaced with fresh media, and biofilms were grown for another 4 h in the same conditions. Media was removed and biofilms were incubated with 30 μM Hst 5 suspended in 10 μM NaPB with 10 μM propidium iodide (PI) for 1 h at 37°C, except for S. aureus which was incubated at RT. After 1 h, biofilms were gently scraped to remove them from each well, diluted, and placed on microscope slides for image acquisition using a Zeiss Axioobserver.Z1 inverted fluorescent microscope, and an Axiocam 503M camera. Percent killing was calculated as the number of PI positive cells divided by the total number of cells from at least 25 fields from two independent wells.

Overnight cultures of E. faecium, S. aureus, A. baumannii, and E. cloacae were diluted and pre-incubated with 10% of media with or without 10 mM NaN₃ at 37°C. P. aeruginosa cells were pre-incubated in 10 mM NaPB with 10 mM NaN₃ at 37°C due to altered cell viability in the presence of 10% media with NaN₃. After 3 h incubation, cells were collected by centrifugation and washed two times with NaPB; then cells were used for Hst 5 bactericidal assay in NaPB as described above.

Minimum inhibitory concentration (MIC) was determined by broth microdilution method based on the guide of Clinical and Laboratory Standards Institute (Clinical Laboratory Standards Institute, 2012) with some modifications. Briefly, all bacterial strains except K. pneumoniae were cultured in respective media and grown overnight at 37°C. Cells were washed with NaPB buffer, diluted in 10% of Mueller-Hinton (MH) broth (Sigma) to obtain a concentration of 5 × 10⁶ CFU/mL that were used as the inoculum. Hst 5 peptide was serially diluted in 10% MH broth (since it is inactive in undiluted broth) in 96-well flat-bottomed plates (Falcon). After adding equal volume of target bacterial suspension to the peptide solution, the 96-well plates were incubated at 37°C for 24 h. The MIC values were determined by visibly recording the lowest concentration of Hst 5 that inhibited growth.

Statistical analyses were performed using GraphPad Prism software version 5.0 (GraphPad Software, San Diego, CA, USA) using unpaired Student's t-tests. Differences of P < 0.05 were considered significant. All experiments were performed at least thrice.

In order to determine the bactericidal activity of Hst 5 against ESKAPE pathogens, six clinical isolates (Table 1) were tested using bactericidal assays (performed in 10 and 100 mM NaPB buffer) to determine percent bacterial killing following incubation for 1 min, 1, and 5 h with 30 μM Hst 5. E. faecium showed time-dependent killing since only 18.1 ± 8.9 percent of cells were killed by Hst 5 at 1 h, while killing was enhanced to 63.5 ± 3.5 percent after 5 h incubation (Figure 1A). S. aureus cells incubated for only 1 min with 30 μM Hst 5 showed 28.3 ± 2.3 percent killing which increased to 60.3 ± 5.7 and 69.7 ± 0.2 percent after 1 and 5 h, respectively (Figure 1B). A MRSA strain of S. aureus had lower but still significant killing by Hst 5 after 1 h (33.8 ± 3.5%). Interestingly, Hst 5 did not show any killing activity against K. pneumoniae for up to 5 h of incubation (Figure 1C). A. baumannii AB307-0294 had substantial sensitivity to Hst 5 with 31.4 ± 3.7, 90.0 ± 1.1, and 95. ± 2.3 percent killing at 1 min, 1, and 5 h, respectively (Figure 1D), while the A. baumannii HUMC1 strain had even higher sensitivity with 100% killing at 1 h. Similarly, Hst 5 also showed strong killing activity against P. aeruginosa 94-323-0635, with a faster rate of killing. At just 1 min of incubation, 83.2 ± 1.3 percent killing was observed; while over 99.9 percent of cells were killed after 1 h treatment (Figure 1E). P. aeruginosa PAO1 had identical sensitivity to Hst 5 with 100% of cells killed at 1 h. E. cloacae sensitivity was similar to that of A. baumannii showing 20.3 ± 1.5, 66.1 ± 0.6, and 78.1 ± 1.7 percent of cells killed by Hst 5 at 1 min, 1 h, and 5 h, respectively (Figure 1F).

Since the activity of Hst 5 against C. albicans is abolished at higher phosphate buffer concentrations (Helmerhorst et al., 1999), Hst 5 killing activity against ESKAPE strains was also tested in a buffer with higher ionic strength. Killing activities of Hst 5 against E. faecium, P. aeruginosa, and E. cloacae were greatly reduced (Figures 1A,E,F) when carried out in 100 mM NaPB. However, bactericidal activity against S. aureus at 100 mM NaPB was similar to that observed in 10 mM NaPB (Figure 1B), while killing efficiency for A. baumannii were decreased by about 30 percent in 100 mM NaPB after 1 or 5 h incubation (Figure 1D). Effects of PBS on killing were similar to those seen in 100 mM NaPB (data not shown).

Based on these bactericidal results, we next tested growth inhibition by Hst 5 for these ESKAPE pathogens by determining MIC values in 10% MH broth (since Hst 5 is inactive in undiluted broth). Hst 5 MIC values of 38, 47, and 90 μM were observed for A. baumannii, P. aeruginosa, and E. cloacae; while we could not determine MIC values for E. faecium and S. aureus due to media components that inactivate Hst 5 (Table 2).

The activity of Hst 5 against ESKAPE pathogens grown as biofilms was determined after 16 h growth of each bacteria in 96-well plates (Table 2). Biofilms were treated with 30 μM Hst 5 for 1 h, dead cells stained with PI, and percent killing calculated. Neither E. faecium or E. cloacae formed robust biofilms in our hands so that Hst 5 killing could not be reliably determined for these two pathogens. As for planktonic cells, both strains of P. aeruginosa in biofilms were effectively killed by Hst 5 after 1 h (59.5%); however Hst 5 killing of A. baumannii biofilm cells was reduced to 15.1%; while killing of S. aureus biofilm cells (both strains) was negligible (5.1%). Surprisingly, we observed that Hst 5 killed 19.5% of K. pneumoniae when in biofilms, although it had no activity against planktonic cells.

Overall, our results show that Hst 5 has a strong bactericidal activity against P. aeruginosa and E. cloacae (and weaker killing with E. faecium) at lower ionic strength environments; while exerting killing against A. baumannii and S. aureus over a range of ionic strengths. However, biofilm cells of A. baumannii and S. aureus were more resistant to Hst 5 killing. To confirm these findings in real time and to assess whether ionic strength may influence metabolic or membrane lytic mechanisms of killing, we next measured Hst 5 binding, uptake, and time of killing for these ESKAPE pathogens in 10 mM NaPB.

Time-lapse confocal microscopy showed that Hst 5 (30 μM) rapidly (< 1 min) bound with the surface of E. faecium cells, although no internalization was evident until after 10 min of incubation (Figure 2A). This slow internalization resulted in only 15% of cells taking up F-Hst 5 by 120 min; but in each instance, Hst 5 internalization was accompanied by PI uptake (indicative of cell death; Figure 2B). This close correlation between intracellular Hst 5 and PI uptake suggested that targets of Hst 5 are intracellular rather than its effects being membrane lytic. However, the percentage of killing from the bactericidal assay was higher than apparent PI staining, suggesting that some portion of killing was delayed. To evaluate this, we pretreated cells with an inhibitor of energy metabolism (NaN₃) as a means to determine whether the killing activity of Hst 5-treated E. faecium might be dependent upon cell metabolism. As expected, pretreatment of cells with NaN₃ significantly reduced the killing efficiency of Hst 5 at 1 and 5 h (Figure 2C), showing that some portion of Hst 5 killing of E. faecium requires target cell energy, likely for Hst 5 internalization or intracellular organelle localization.

E. cloacae cells had a very similar response when treated with F-Hst 5, in that the peptide associated with the surface of all cells almost immediately, however entry of F-Hst 5 (and accompanying PI staining) was slow and only occurred in 20% of cells within 30 min (Figure 3A). As for E. faecium, there was a close correlation between intracellular Hst 5 and PI uptake at all time points (Figure 3B), also pointing toward intracellular targets for bactericidal activity. E. cloacae cells pretreated with 10 mM NaN₃ prior to exposure with 30μM Hst 5 were more resistant to the killing action of F-Hst 5 (Figure 3C), similar to the energy dependent mechanism found with E. faecium.

In agreement with its high bactericidal activity, F-Hst 5 completely covered the cell surface of P. aeruginosa within 1 min, and PI positive cells could be visualized within 2 min (Figure 4A). Quantitative analysis showed that 75% of cells contained F-Hst 5 and were PI positive within 5 min of addition of F-Hst 5 (Figure 4B). Both intracellular localization of Hst 5 and PI staining were rapid and simultaneous, pointing toward membrane lytic activity of Hst 5 mediated killing of P. aeruginosa. Pretreatment of P. aeruginosa cells with 10 mM NaN₃ did not reduce the killing activity of Hst 5 (Figure 4C), also suggesting energy-independent membrane disruption as the main pathway for Hst 5 killing activity against P. aeruginosa.

A. baumannii showed a very similar profile to P. aeruginosa after exposure to F-Hst 5 in that F-Hst 5 was associated with the surface of all cells within 2 min; and PI positive cells were visualized with 2 min (Figure 5A). Although intracellular uptake of F-Hst 5 was very rapid with 70% of cells containing F-Hst 5 in just 2 min, PI staining did not occur simultaneously so that 70% PI positive cells was seen only after 25 min (Figure 5B). Interestingly, the portion of A. baumannii cells that showed slower PI staining already contained F-Hst 5; so we questioned whether bactericidal activity of these cells was energy-dependent. However, NaN₃-pretreated A. baumanni cells had no difference in F-Hst 5 killing (Figure 5C), indicating that Hst 5-mediated killing of A. baumanni is partially due to membrane lysis and is energy independent, although some portion of killing might also be a result of its effect on intracellular targets.

F-Hst 5 was visualized binding to the cell surface of all S. aureus cells within 1 min after addition, however intracellular localization and PI uptake occurred slowly (Figure 6A), so that by 120 min only 7% of cells contained F-Hst 5 and were PI positive (Figure 6B). This was surprising since bactericidal activity of Hst 5 after 1 h incubation was found to have 60% killing (Figure 1). Furthermore, bactericidal assay with S. aureus pretreated with 10 mM NaN₃ did not decrease the killing efficiency (Figure 6C). These results, combined with the salt-insensitive killing, suggest that there may be multiple targets for Hst 5-mediated killing of S. aureus that are non-lytic and energy independent.

Since Hst 5 showed negligible killing against K. pneumoniae (Figure 1C), we examined the reason for this using confocal microscopy. F-Hst bound to most K. pneumoniae cells within 2 min, however by 10 min most binding was lost suggesting detachment of Hst 5 from the surface capsule (Figure 7A). As expected, F-Hst 5 and PI uptake were extremely limited (only 2% of cells), even up to 150 min (Figure 7B). These results suggest that F-Hst 5 is released from the K. pneumoniae after initial binding (perhaps due to lower binding efficacy in the presence of its capsule) and thus Hst 5 is unable to gain entry or lyse the cells and therefore is ineffective against K. pneumoniae.

We have previously reported that spermidine-conjugated Hst 5 (Hst 5-Spd) has greater activity against C. albicans because the spermidine conjugate translocates more efficiently into the yeast cells (Tati et al., 2014). Therefore, to determine if Hst 5-Spd has improved bactericidal activity against ESKAPE pathogens as well, we tested the killing activity of Hst 5-Spd [30 μM] against all the six strains (Figure 8). Interestingly, Hst 5-Spd had significantly higher killing activity compared with Hst 5 against E. faecium, E. cloacae, and A. baumannii (Figure 8). Hst 5-Spd bactericidal effects were most improved against E. faecium; increased by three-fold after 1 h incubation and by 1.5-fold after 5 h, while Spermidine (Spd) alone had no killing activity even after 5 h treatment (Figure 8A). Hst 5-Spd bactericidal activity was also increased significantly in E. cloacae compared with Hst 5, although by only 15% at 1 h and at 5 h. Hst 5-Spd also showed a small but significant increase in killing activity against A. baumannii compared with Hst 5 (23% at 1 min and 10% at 1 h). However, after 5 h there was no significant difference perhaps due to the high killing that occurred as a result of Spd (30 μM) itself (Figure 8D). These three species of ESKAPE pathogens were also ones that we found to involve Hst 5 uptake and intracellular targets (Figures 2, 3, 5), suggesting that improved Hst 5-Spd activity in these strains is a result of improved uptake. This is supported by our data showing that the killing ability of Hst 5-Spd against P. aeruginosa was not improved, in agreement with a membrane lytic mechanism for this organism. Furthermore, the Hst 5 resistant pathogen K. pneumoniae was not affected by Hst 5-Spd showing that this conjugate does not improve binding to this encapsulated ESKAPE pathogen.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.