Abstract
The treatment of skin infections caused by Staphylococcus aureus is limited by acquired antibiotic resistance and poor drug delivery into pathogen and host cells. Here, we investigated the antibacterial activities of six topically used antimicrobials and a cationic polymer, polyhexamethylene biguanide (PHMB), against intracellular MSSA strain RN4420 and MRSA strains EMRSA-15 and USA 300.
The MICs of antimicrobials were determined for MSSA and MRSA strains, and the bactericidal activities of nadifloxacin and PHMB against intracellular MRSA were determined using infected keratinocytes. Fluorescein-tagged PHMB (PHMB–FITC) was used to study PHMB uptake, co-localization with intracellular EMRSA-15 and retention in keratinocytes. The mechanism(s) of PHMB uptake into keratinocytes were studied using a dynamin inhibitor, dynasore.
Gentamicin, nadifloxacin and PHMB showed the lowest MICs for MRSA. Nadifloxacin at 10 mg/L killed 80% of intracellular EMRSA-15, but was not effective against USA 300. PHMB at 4 mg/L killed almost 100% of intracellular EMRSA-15 and USA 300. PHMB entered keratinocytes, co-localized with intracellular EMRSA-15 and was retained by the cells for over 5 h. PHMB uptake and its intracellular antibacterial activities were inhibited by the dynamin inhibitor, dynasore.
PHMB kills intracellular MRSA via direct interaction with pathogens inside keratinocytes and host cell entry is dynamin dependent.
Introduction
Skin infections caused by invasive pathogens start when outer skin structures are damaged due to injury, such as a burn,1 or skin diseases, such as atopic dermatitis.2 The majority of invasive skin infections are caused by Staphylococcus aureus, including MRSA.3 Invasive strains of S. aureus are difficult to manage therapeutically due to their thick cell barriers and acquired drug resistance. Additionally, S. aureus is able to gain entry and replicate within various host cell types, including keratinocytes, macrophages, endothelial cells, epithelial cells, fibroblasts and osteoblasts, where antibiotic concentrations in situ are often sub-therapeutic.4 Recent work by Lehar et al.5 demonstrated that the three major antibiotics that are currently administered for invasive MRSA infections (linezolid, vancomycin and daptomycin) fail to kill intracellular S. aureus USA 300 in infected mice. Consequently, intracellular invasion of mammalian cells by S. aureus can induce pyroptosis (cell death), followed by accumulation of bacteria, which can then escape to infect nearby cells.6 Therefore, drug-resistant and intracellular S. aureus can result in incomplete eradication during therapy.
The control of S. aureus, particularly MRSA, is further complicated by the emergence of epidemic strains that display increased virulence. These strains are reported to be highly invasive and able to cause bacteraemia in patients.7 In the UK, epidemic MRSA 15 (EMRSA-15) and EMRSA-16 strains constitute more than 60% and 35% of MRSA associated infections, respectively.7 These two strains are considered to be the most successful strains at surviving, colonizing and spreading in the hospital environment.8–10 Similarly, in the USA, strains USA 300 and USA 400 often cause community-acquired MRSA infections, with USA 300 being the predominant cause of skin and soft tissue infections.11 The USA 300 strain is widely disseminated across the USA and capable of causing clinical illness ranging from uncomplicated bacteraemia to endocarditis, necrotizing pneumonia and osteomyelitis.12
The recommended treatments for skin infection caused by S. aureus include administration of topical and systemic antimicrobials.13 Topical administration is often needed to ensure drugs reach sites of infection. For example, in the case of burn injuries, damage to the local vascular system could limit the distribution of oral antimicrobials to infected sites. Therefore, topical antimicrobials such as mupirocin, fusidic acid and bacitracin are frequently used.14 Unfortunately, increasing resistance to these antimicrobials has reduced effectiveness.15 Among the recently introduced topical antimicrobials, nadifloxacin, a fluoroquinolone, is effective against aerobic and anaerobic bacteria isolated from skin infections.16 Importantly, nadifloxacin is effective against ciprofloxacin-resistant strains.17 Nadifloxacin is currently approved for acne treatment and skin infections.18
Synthetic cationic polymers are becoming widely used in clinics as antimicrobials due to their broad-spectrum bactericidal activities and high therapeutic index. For example, polyhexamethylene biguanide (PHMB) (Figure 1) is a cationic polymer that provides potent topical antimicrobial effects against Gram-positive and Gram-negative bacteria,19,20 fungi21,22 and certain viruses.23 PHMB is a relatively safe compound with no reported adverse reactions when used on the skin, eyes, epithelium of the nose and wounds.24,25 It is available both as a disinfecting solution and impregnated into biocellulose dressings.26 Although PHMB has been used in the clinic for >40 years, there are no reports of acquired resistance to this compound.24 Due to its excellent antibacterial properties and safety profile, PHMB is often administered topically to prevent wound infection.27
Structure of PHMB. PHMB is a cationic polymer of repeating hexamethylene biguanide groups, with n average=10–12 (n is the number of structural unit repeats).
Recently we observed that PHMB enters bacteria (S. aureus, Escherichia coli and Mycobacterium smegmatis) and several types of mammalian cells grown in cell culture (including epithelial cells, macrophages and fibroblasts).28 On the basis of these findings and the established antibacterial properties of PHMB, we hypothesized that PHMB may also possess previously unrecognized antibacterial activity against intracellular bacteria. Here, we investigated the potential antibacterial activities of six topical antimicrobials and PHMB against intracellular MRSA. We further characterized the route(s) of PHMB uptake into keratinocytes and the mechanism(s) of intracellular killing by PHMB.
Materials and methods
Bacterial strains and growth conditions
S. aureus strain RN4420 was obtained from Dr Staffan Arvidson, Karolinska Institutet, Stockholm, Sweden,29 strain EMRSA-15 was obtained from Dr Sean Nair, University College London,30 and strain USA 300 was obtained from Dr Jonathan Otter, Imperial College Healthcare NHS Trust, UK.31 Bacteria were grown in Mueller–Hinton broth (MHB; Sigma–Aldrich, UK) followed by incubation at 250 rpm (for liquid cultures) at 37°C for 18 h.
Eukaryotic cell lines and growth conditions
HaCaT cells were obtained from Dr Amir Sharili, Queen Mary University of London, and maintained in DMEM with 10% FBS (Sigma–Aldrich, UK), supplemented with 5% penicillin/streptomycin (Sigma–Aldrich, UK). Cells were maintained at 37°C in 5% carbon dioxide.
Determination of MICs
All antimicrobials were purchased from Sigma–Aldrich, UK, except nadifloxacin, which was from Santa Cruz Biotechnology, UK. PHMB and fluorescein-tagged PHMB (PHMB–FITC) were from Tecrea Ltd, UK. All antimicrobials were prepared in stock solution at 10 000 mg/L. Fusidic acid, gentamicin, bacitracin and PHMB were dissolved in sterile distilled water, mupirocin was dissolved in methanol, erythromycin was dissolved in ethanol and nadifloxacin was dissolved in 0.1 M sodium hydroxide solution.
MICs were determined using the broth microdilution method.32 Briefly, a range of concentrations of antimicrobials were prepared in a 96-well microplate, followed by inoculation of bacterial culture to yield ∼5 × 105 cfu/mL in a 250 μL final volume. The plate was then incubated at 37°C for 18 h. The lowest concentration of antimicrobial that inhibited growth of bacteria was scored as the MIC.
Intracellular infection of keratinocytes by S. aureus
Intracellular infections of keratinocytes by S. aureus RN4420, EMRSA-15 and USA 300 were established using a gentamicin protection procedure.33 Keratinocytes were seeded at 1.2 × 105 cells/well in a 12-well plate and cultured overnight in DMEM with 10% FBS, without antibiotic. In parallel, all S. aureus strains were cultured overnight in MHB at 37°C in an incubator shaker. One millilitre of overnight bacterial culture was centrifuged at 8000 rpm for 3 min and the pellet was resuspended in PBS (Sigma–Aldrich, UK). These steps were performed three times to remove bacterial toxin residues. Bacteria were diluted to a final concentration of ∼107 cfu/mL in DMEM with 10% FBS, without antibiotic. Aliquots of bacteria (1 mL, 107 cfu/mL) were added to keratinocyte cultures after the original medium was removed. Bacteria were co-incubated with keratinocytes for 3 h. Gentamicin diluted in medium at 200 mg/L was added and incubated for 3 h. Medium containing bacteria and gentamicin was removed and cells were rinsed with PBS. The rinse step was performed because this medium contained very high concentrations of gentamicin and therefore did not result in cfu of bacteria when plated directly. Therefore, aliquots of PBS from the rinsing process were plated on nutrient agar to determine the remaining number of extracellular bacteria. Next, 1 mL of 0.5% Triton X-100 prepared in PBS was added to each well to lyse cells. Lysed cells were serially diluted in PBS and plated on nutrient agar (Sigma–Aldrich, UK) for enumeration of intracellular bacteria. Uninfected cells were also subjected to the lysis procedure to confirm sterility.
Visualization of intracellular S. aureus within keratinocytes
Keratinocytes were grown on glass cover slips in 12-well plates followed by EMRSA-15 infection as described above. Following gentamicin exposure to kill extracellular bacteria, cells were rinsed with PBS and fixed with 4% paraformaldehyde (Santa Cruz Biotechnology, UK). Cells were stained with 5 mg/L DAPI (Life Technologies, UK) for nuclear staining and 5 mg/L wheat germ agglutinin (WGA)-conjugated Alexa Fluor 555 (Life Technologies, UK) for membrane staining. Cover slips were mounted on glass slides with FluorSave™ (Calbiochem, UK). Images were visualized using a Leica SP5 confocal microscope using Advanced Fluorescence software (Leica Microsystems, Milton Keynes, UK). Sequential scan z-stacks (139 slices 1024 × 1024) were compiled at a line average of 96, using Volocity® software. Three-dimensional (3D) image analysis software was used for analysis and to produce 3D images.
Intracellular killing of EMRSA-15 and USA 300 by nadifloxacin and PHMB
Keratinocytes were infected with EMRSA-15 and USA 300 followed by incubation with gentamicin to kill extracellular bacteria. Following gentamicin exposure, keratinocytes were rinsed with PBS, and antimicrobials (nadifloxacin and PHMB in medium) were added to the wells containing infected cells. Plates were incubated for another 3 h to kill intracellular bacteria. Following this period, antibiotic solutions were removed and cells were rinsed and lysed. Lysed cells were serially diluted and plated on nutrient agar. For each experiment, gentamicin-treated cells (without nadifloxacin or PHMB treatments) were used as controls.
Resazurin assay
Keratinocytes (4 × 104 cells/well) were added to a 96-well plate and cultured with increasing concentrations of nadifloxacin or PHMB at 37°C for 24 h. Non-treated cells and medium only were used as controls. Resazurin sodium salt (Sigma–Aldrich, UK) was prepared as a stock solution at 440 μM in PBS and added to each well at 44 μM final concentration and plates were incubated for an additional 48 h. OD was then measured using a Tecan Infinite plate reader (Tecan Group Ltd, Switzerland) at 550 and 630 nm. The OD value change (or percentage dye reduction) is proportional to the viable cell number and was used to calculate IC50 values based on the intercept theorem.
Visualization and quantification of PHMB uptake into keratinocytes
For both experiments, keratinocytes were grown on glass cover slips in a 12-well plate. PHMB–FITC was added to cultures at 4 mg/L for 3 h. Cells were rinsed with PBS and stained as described above before confocal microscopy. To quantify PHMB uptake into keratinocytes, PHMB–FITC-treated cells were incubated with 0.04% trypan blue (Invitrogen, UK) in PBS for 15 min to quench membrane-bound PHMB–FITC, followed by flow cytometry (FACS BD machine and BD FACSDiva™ software, BD Biosciences) using the FITC filter. For each sample, 10 000 gated cells were analysed.
Exocytosis of PHMB from keratinocytes
Keratinocytes were seeded in 35 mm tissue culture treated dishes (Greiner Bio-One, Austria). Keratinocytes were incubated with 4 mg/L of PHMB–FITC in medium for 3 h. Following exposure, PHMB–FITC-containing medium was removed and new medium containing 5 mg/L WGA was added. Fluorescence quenching rates were determined and used to select an appropriate light intensity. Cellular fluorescence was monitored with images captured every 15 min for 5 h using confocal microscopy. Mean fluorescence intensity (MFI) was calculated at the beginning and after 5 h of experiment to assess retention of PHMB–FITC in the cells.
Statistical analysis
All experiments were performed in triplicate. Statistical analysis was performed using one-way analysis of variance (ANOVA) followed by Tukey tests using the statistical package Prism 6.0 (GraphPad, San Diego, CA, USA). Data are presented as mean ± SD. Differences were considered to be statistically significant where P ≤ 0.05. For histogram and graphs, error bars represent standard deviations. *P ≤ 0.05; ***P ≤ 0.001; ****P ≤ 0.0001; and NS, not significant.
Results
Susceptibility of S. aureus strains to topically used antimicrobials and PHMB
Six topically used antimicrobials were tested, and the MICs of all six were higher for the clinical MRSA strains (EMRSA-15 and USA 300) than for the MSSA strain RN4420. In contrast, for PHMB, MICs were similar for all strains. Overall, gentamicin, nadifloxacin and PHMB displayed the lowest MICs for MRSA strains (Table 1).
MICs of topical antimicrobials for S. aureus RN4420, EMRSA-15 and USA 300
| Antimicrobial | Strain | ||
|---|---|---|---|
| RN4420 | EMRSA-15 | USA 300 | |
| Mupirocin | 0.125 | 32 | 2 |
| Fusidic acid | 0.25 | 32 | 0.5 |
| Gentamicin | 0.25 | 4 | 0.25 |
| Nadifloxacin | 0.125 | 2 | 2 |
| Erythromycin | 0.125 | 32 | 16 |
| Bacitracin | 16 | >32 | 32 |
| PHMB | 1 | 1 | 1 |
| Antimicrobial | Strain | ||
|---|---|---|---|
| RN4420 | EMRSA-15 | USA 300 | |
| Mupirocin | 0.125 | 32 | 2 |
| Fusidic acid | 0.25 | 32 | 0.5 |
| Gentamicin | 0.25 | 4 | 0.25 |
| Nadifloxacin | 0.125 | 2 | 2 |
| Erythromycin | 0.125 | 32 | 16 |
| Bacitracin | 16 | >32 | 32 |
| PHMB | 1 | 1 | 1 |
MICs of topical antimicrobials for S. aureus RN4420, EMRSA-15 and USA 300
| Antimicrobial | Strain | ||
|---|---|---|---|
| RN4420 | EMRSA-15 | USA 300 | |
| Mupirocin | 0.125 | 32 | 2 |
| Fusidic acid | 0.25 | 32 | 0.5 |
| Gentamicin | 0.25 | 4 | 0.25 |
| Nadifloxacin | 0.125 | 2 | 2 |
| Erythromycin | 0.125 | 32 | 16 |
| Bacitracin | 16 | >32 | 32 |
| PHMB | 1 | 1 | 1 |
| Antimicrobial | Strain | ||
|---|---|---|---|
| RN4420 | EMRSA-15 | USA 300 | |
| Mupirocin | 0.125 | 32 | 2 |
| Fusidic acid | 0.25 | 32 | 0.5 |
| Gentamicin | 0.25 | 4 | 0.25 |
| Nadifloxacin | 0.125 | 2 | 2 |
| Erythromycin | 0.125 | 32 | 16 |
| Bacitracin | 16 | >32 | 32 |
| PHMB | 1 | 1 | 1 |
Given that nadifloxacin and PHMB displayed low MICs relative to the other antimicrobials tested, we chose to focus on these two compounds in further studies. There are no reports for nadifloxacin and PHMB activities against intracellular bacteria. Gentamicin is an antibiotic that is active against extracellular bacteria, but not against intracellular bacteria, due to its poor penetration into mammalian cells.34 Therefore, we chose to further examine nadifloxacin and PHMB activities against intracellular MRSA, and gentamicin was not examined further.
Intracellular invasion of keratinocytes by S. aureus
To determine the activities of nadifloxacin and PHMB against intracellular bacteria, we first established an in vitro intracellular infection assay using MRSA and HaCaT cells.33 HaCaT is a well-established keratinocyte cell line that is often used experimentally as a host for intracellular S. aureus.6,35 To establish conditions for S. aureus RN4420, EMRSA-15 and USA 300 invasion in keratinocytes, bacteria were incubated with keratinocytes, followed by co-incubation with 200 mg/L gentamicin to kill extracellular bacteria. All strains were able to invade keratinocytes, as indicated by their ability to survive gentamicin treatment. Three hours of infection followed by 3 h of gentamicin treatment resulted in reproducible infections (Figure 2a, b and c). Killing of mainly extracellular bacteria was evident for all three strains, where lysis of keratinocytes following gentamicin exposure released ∼103 cfu/mL S. aureus RN4420, 105 cfu/mL EMRSA-15 and 104 cfu/mL USA 300. These colony counts presumably reflected S. aureus that were inside keratinocytes and thus protected from gentamicin's extracellular antibacterial activities. As anticipated, the two clinical isolate strains, EMRSA-15 and USA 300, showed higher invasion levels than S. aureus RN4420.
S. aureus invasion in keratinocytes. (a–c) Each panel represents survival of three different S. aureus strains following gentamicin exposure. After gentamicin exposure, lysis of keratinocytes released ∼103 cfu/mL S. aureus RN4420, 105 cfu/mL EMRSA-15 and 104 cfu/mL USA 300.
S. aureus invasion of keratinocytes was visualized by confocal microscopy. Keratinocytes were infected with EMRSA-15 and stained with DAPI and WGA. Using 3D analysis, we observed localization of EMRSA-15 (blue dots) inside keratinocyte membranes (red) (Figure 3a and b), confirming invasion of EMRSA-15.
Localization of S. aureus in keratinocytes. Invasion of EMRSA-15 in keratinocytes was visualized using confocal microscopy. Prior to imaging, keratinocytes were stained with DAPI (blue) for keratinocyte and EMRSA-15 nuclear staining and WGA (red) for keratinocyte membrane staining. Confocal microscopy z-stack projection moved through 139 slices across keratinocytes. (a) Horizontal cross-section of keratinocytes. (b) Ventricle cross-section of keratinocytes. The small blue dots are EMRSA-15 cells inside keratinocytes, indicating invasion. White scale bar = 7.5 μm. This figure appears in colour in the online version of JAC and in black and white in the print version of JAC.
Susceptibility of intracellular EMRSA-15 and USA 300 to nadifloxacin and PHMB
To investigate the potential of nadifloxacin and PHMB to kill intracellular MRSA, keratinocytes were infected with EMRSA-15 and USA 300, exposed to gentamicin to kill extracellular bacteria and then treated with nadifloxacin or PHMB at 1, 2, 4, 6, 8 or 10 mg/L for 3 h, followed by enumeration of cfu of surviving intracellular bacteria. Figure 4 shows the percentages of surviving intracellular MRSA after treatment with nadifloxacin and PHMB, relative to untreated infected cells. Nadifloxacin at 10 mg/L killed 80% of intracellular EMRSA-15, but failed to kill USA 300 after 3 h of treatment. PHMB at 4 mg/L killed almost 100% of both strains as compared with 100% survival of the untreated infected cells. Therefore, while nadifloxacin displayed similar MICs for both EMRSA-15 and USA 300, its activities were inconsistent against intracellular S. aureus, whereas PHMB displayed consistent killing of both EMRSA-15 and USA 300.
Bactericidal activities of nadifloxacin and PHMB against intracellular EMRSA-15 and USA 300. Keratinocytes infected with EMRSA-15 and USA 300 were either untreated or treated with increasing concentrations of nadifloxacin or PHMB. Untreated cultures were used to establish cfu values corresponding to 100% survival.
Nadifloxacin and PHMB toxicity in keratinocytes
Resazurin assays were performed to determine the in vitro toxicity levels of nadifloxacin and PHMB to keratinocytes. The IC50 values for nadifloxacin and PHMB were 1428 ± 247 and 19.7 ± 10 mg/L, respectively. Therefore, keratinocytes tolerated high concentrations of both compounds, which were far higher than the concentrations required to kill intracellular S. aureus: 10 mg/L nadifloxacin and 4 mg/L PHMB. These results indicate that both compounds can be used to kill intracellular S. aureus with a high therapeutic index.
PHMB uptake into keratinocytes
Killing of intracellular S. aureus by antimicrobials normally requires drug entry into host cells and direct contact with bacteria. To test whether or not PHMB can enter keratinocytes, we added PHMB–FITC to cultures and tracked its intracellular localization using confocal microscopy. As can be seen in Figure 5(a), a mixture of predominantly diffuse and punctate spots of PHMB–FITC were observed. The diffuse signals suggest localization in the cytosol, and the punctate spots suggest localization in vesicles. To quantify PHMB entry into keratinocytes, we analysed treated cells using flow cytometry. Consistent with microscopy results, flow cytometry data indicate that PHMB–FITC entered 99% of cells (Figure 5b).
PHMB–FITC uptake and retention in keratinocytes. (a) PHMB–FITC localization visualized by confocal microscopy. Free FITC was used as a control. (b) PHMB–FITC uptake was quantified by flow cytometry. The overlay histogram image illustrates the uptake of PHMB–FITC into keratinocytes. The red profile shows the population of keratinocytes treated with FITC and serves as a negative control. The green profile shows the population of keratinocytes treated with PHMB–FITC. Uptake of PHMB–FITC was observed in >99% of keratinocytes. (c) Keratinocytes following treatment with PHMB–FITC and 5 h after rinsing to remove extracellular PHMB–FITC. (d) MFI of PHMB–FITC before and after 5 h of experiments. This figure appears in colour in the online version of JAC and in black and white in the print version of JAC.
Exocytosis of PHMB–FITC from keratinocytes
Antimicrobial-mediated killing of intracellular pathogens may benefit from host cell retention. To determine PHMB's host cell retention, we exposed keratinocytes to PHMB–FITC, rinsed the cells, added new medium and monitored cell-associated fluorescence over 5 h. We observed only a minor reduction in fluorescence, with 90% of the fluorescence retained (Figure 5c and d). Therefore, the majority of PHMB was not excreted from cells, which may help to explain its potent killing of intracellular EMRSA-15 and USA 300.
Co-localization of PHMB and EMRSA-15 in keratinocytes
The results described above are most easily explained by a mechanism of direct killing of EMRSA-15 and USA 300, where PHMB enters keratinocytes and then kills bacteria. However, other indirect killing mechanisms are formally possible. To further characterize the PHMB-mediated killing mechanism(s), localization of EMRSA-15 and PHMB–FITC inside keratinocytes was assessed using the gentamicin protection method together with microscopy. After gentamicin treatment, PHMB–FITC was added to cells and after 3 h of incubation the cells were rinsed and stained with DAPI and WGA. Cells were then fixed and imaged using confocal microscopy.
Co-localization of blue-stained bacteria and green PHMB–FITC was apparent inside host cells, as shown in Figure 6. Indeed, co-localization of PHMB–FITC and EMRSA-15 in keratinocytes was apparent for the majority of intracellular bacteria in all cells examined. Therefore, PHMB–FITC appears to directly access intracellular EMRSA-15 inside keratinocytes, suggesting direct killing of intracellular bacteria.
Co-localization of PHMB–FITC with EMRSA-15 in keratinocytes. Keratinocytes were infected with EMRSA-15 followed by treatment with PHMB–FITC (green). Prior to microscopy, keratinocytes were labelled with DAPI (blue) and WGA (red). Top panels are images of infected cells visualized using filters for DAPI, FITC and WGA, and merged images. Bottom panels are enlarged images that clearly show co-localization between PHMB–FITC (green) and EMRSA-15 (blue). White scale bar = 25 μm. This figure appears in colour in the online version of JAC and in black and white in the print version of JAC.
Effect of dynamin inhibition on PHMB uptake and antibacterial activity in keratinocytes
Cell entry pathways that transport PHMB and S. aureus to a common final destination in the cell could explain their co-localization in keratinocytes. Recent work in our laboratory demonstrates that PHMB enters macrophages predominantly via a dynamin-dependent endocytic pathway.28 This pathway requires dynamin (a GTPase) to excise newly formed vesicles from the membrane.36 Without dynamin-mediated vesicle excision, vesicles do not form and enter cells.
If PHMB kills intracellular S. aureus via direct contact with bacteria in keratinocytes, as suggested by the co-localization results above, then inhibitors that block cell entry should reduce both PHMB uptake and intracellular killing. To test the effect of dynamin inhibition on PHMB uptake, we pretreated keratinocytes with a dynamin inhibitor, dynasore, before addition of PHMB–FITC. Using confocal microscopy, we observed that dynasore reduced PHMB–FITC uptake (Figure 7a). Quantification using flow cytometry confirmed that dynasore reduced PHMB–FITC uptake into keratinocytes by 85% (Figure 7b). Similarly, dynasore reduced PHMB-mediated killing of intracellular EMRSA-15 by ∼50% (Figure 7c). These results suggest that PHMB entry and intracellular antibacterial activity in keratinocytes involves dynamin-dependent endocytic pathways.
Effects of dynasore on PHMB uptake into keratinocytes and killing of intracellular EMRSA-15. Uptake of PHMB and intracellular antibacterial activities in dynasore-pretreated keratinocytes. Cells were pretreated with dynasore (in 0.5% DMSO) for 30 min. Then cells were incubated with PHMB–FITC for uptake studies or PHMB for antibacterial activity studies. Control samples were keratinocytes treated with PHMB–FITC alone. Vehicle samples were keratinocytes treated with PHMB–FITC in medium containing 0.5% DMSO. Following treatment, keratinocytes were (a) visualized by confocal microscopy and (b) analysed by flow cytometry. (c) Effects of dynasore on PHMB-mediated killing of intracellular bacteria. This figure appears in colour in the online version of JAC and in black and white in the print version of JAC.
Discussion
Intracellular bacteria are typically difficult to kill because they are shielded from host immune mechanisms and protected from antimicrobials. Here, we found that nadifloxacin and PHMB kill intracellular MRSA within keratinocytes, with PHMB being more potent. Nadifloxacin-mediated killing of intracellular S. aureus is perhaps expected, because it is a small molecule belonging to the fluoroquinolone group, a class of antibiotic that can enter and accumulate within mammalian cells.37 However, nadifloxacin displayed inconsistent activity, being effective against EMRSA-15 but not against USA 300. In contrast, the intracellular activity of PHMB was unexpected, because it is a relatively large molecule (average molecular weight around 3025 g/mol)38 and has only recently been reported to enter mammalian cells.28 Nevertheless, our results clearly show that PHMB enters keratinocytes and is retained for several hours.
Intracellular pathogens such as S. aureus enter mammalian cells through phagocytosis (one of the endocytic pathways) via a zipper-uptake mechanism.39 Inside host cells, the fate of S. aureus is to be killed by innate immune mechanisms within a phagolysosome, replicate inside an endosome or escape into the cytosol.39 Inside keratinocytes, the polymer PHMB–FITC localizes mainly inside vesicles and in the cytosol. In this study, we observed co-localization between PHMB–FITC and almost all intracellular EMRSA-15, suggesting direct interactions that could occur in both vesicles and the cytosol. These interactions may explain the potent intracellular bactericidal activities of PHMB. Furthermore, inhibition of dynamin, a GTPase that pinches off newly formed endocytic vesicles from the membrane,36 impeded both PHMB–FITC uptake and its intracellular killing activity. Therefore, PHMB uptake into keratinocytes appears to be dynamin dependent.
There are three main findings from this study. First, both nadifloxacin and PHMB are effective against intracellular S. aureus, with PHMB being more potent against intracellular MRSA. Second, PHMB co-localizes with intracellular MRSA in keratinocytes, indicating that killing occurs by direct interactions inside host cells. Third, PHMB's uptake and antibacterial activities inside keratinocytes appear to involve a dynamin-dependent cell uptake mechanism. Together, these findings suggest that nadifloxacin and particularly PHMB have previously unrecognized potential for treating skin infections caused by intracellular MRSA and possibly other intracellular bacteria.
Funding
This project is funded by the Ministry of Education Malaysia.
Transparency declarations
None to declare.
Acknowledgements
We thank Dr Staffan Arvidson, Dr Sean Nair and Dr Jonathan Otter for providing the strains and the following individuals for their opinion and technical assistance: Dr Kantaraja Chindera, Dr Adebowale Oluwawemimo, Dr Sharon Kendall, Dr Shan Goh, Dr Andrew Hibbert, Dr Charlotte Lawson, Rebecca Bristow, Klaudia Kloc-Muniak and Harry Horsley. All microscopy work was conducted at the Imaging Suite Facility, Royal Veterinary College, London.
