Synergistic activity of phage PEV20-ciprofloxacin combination powder formulation—A proof-of-principle study in a P. aeruginosa lung infection model

Abstract

Combination treatment using bacteriophage and antibiotics is potentially an advanced approach to combatting antimicrobial-resistant bacterial infections. We have recently developed an inhalable powder by co-spray drying Pseudomonas phage PEV20 with ciprofloxacin. The purpose of this study was to assess the in vivo effect of the powder using a neutropenic mouse model of acute lung infection. The synergistic activity of PEV20 and ciprofloxacin was investigated by infecting mice with P. aeruginosa, then administering freshly spray-dried single PEV20 (10⁶ PFU/mg), single ciprofloxacin (0.33 mg/mg) or combined PEV20-ciprofloxacin treatment using a dry powder insufflator. Lung tissues were then harvested for colony counting and flow cytometry analysis at 24 h post-treatment. PEV20 and ciprofloxacin combination powder significantly reduced the bacterial load of clinical P. aeruginosa strain in mouse lungs by 5.9 log₁₀ (p < 0.005). No obvious reduction in the bacterial load was observed when the animals were treated only with PEV20 or ciprofloxacin. Assessment of immunological responses in the lungs showed reduced inflammation associating with the bactericidal effect of the PEV20-ciprofloxacin powder. In conclusion, this study has demonstrated the synergistic potential of using the combination PEV20-ciprofloxacin powder for P. aeruginosa respiratory infections.

Graphical Abstract

INTRODUCTION

Pseudomonas aeruginosa is a common pathogen that exacerbates disease and mortality among patients with cystic fibrosis (CF), bronchiectasis and chronic obstructive pulmonary disease [1–4]. The pathogenic prominence of P. aeruginosa is largely due to an intrinsic resistance to antibiotics [5]. Bacteriophages (or phages) are considered a potential alternative or supplementary treatment over antibiotics when antibiotic resistance has become increasingly life-threatening [6]. Phages are bactericidal virus that attach to the surface receptor of their host and then eject genetic materials into the cell. While lytic phages hijack the host protein machinery, self-replicate and then cause bacteriolysis, lysogenic phages can integrate their genome within the host chromosome to allow phage DNA, known as prophage, to be copied and passed on along with the host cells [7]. At present, phage therapy (PT) utilizes virulent lytic phages to treat bacterial infections [8]. Lytic phages can not only kill the bacteria, but also enable synergistic antibacterial effect with antibiotics by exerting selective pressure [8, 9].

Antibiotic-phage synergy has been shown repeatedly in vitro and in vivo pre-clinical studies [10]. Synergistic antibacterial effect against planktonic cells and biofilms of P. aeruginosa was achieved in vitro using anti-Pseudomonas phages combined with streptomycin, tobramycin, amikacin, meropenem, ceftazidime or ciprofloxacin [1, 11–14]. Furthermore, improved animal survival rate, increased phage activity, reduced emergence of resistance, and synergistic effects on bacterial load have been reported in different animal models after combination therapy [15–22]. Phage have also been administered as an adjunct to existing antibiotic therapy in clinical settings [9, 23, 24]. Recent clinical reports of simultaneous or sequential administration of phages and antibiotics have shown promising therapeutic outcomes for treatment of multi-drug resistant (MDR) respiratory infections [25–29]. In one notable case, PT in combination with antibiotics was associated with resolution of infection and apparent eradication of bacteria colonization in a 77 year-old woman with ventilator-associated pneumonia and empyema [27].

In another study, improved antibiotic susceptibility was observed after PT on P. aeruginosa strains isolated from a 57 year-old non-CF patient [26]. PT is also associated with clinical efficacy when used as an adjunct to antibiotics in lung transplant patients [28]. Importantly, no PT-related adverse effects were observed upon administration of phages via nebulization.

Inhaled delivery of antimicrobials can benefit patients due to high local bioavailability (i.e. the site of infection) and low systemic exposure [30]. Dry powder inhalers (DPIs) are designed to deliver aerosol particles in the range of 1–5 μm to the lungs, thereby exerting local therapeutic effects and are increasingly used for treating of pulmonary diseases [31]. The use of DPIs can improve patient handling characteristics and compliance due to rapid delivery [32]. In our previous study, we produced inhalable powders containing phage PEV20 and ciprofloxacin [33]. By co-spray drying them together, both antimicrobials can be administered in a single inhalation dose for treating P. aeruginosa respiratory infections. The resulting dry powders were inhalable and in vitro bactericidal synergy was maintained against clinical strains. To the best of our knowledge, the efficacy of phage and antibiotic dry powder combination in the lung has never been reported. Hence, further in vivo investigation would be necessary to warrant potential clinical applications of the combination powder.

In our previous study, a neutropenic mouse lung infection model was established using the P. aeruginosa strain FADD1-PA001. Single PEV20 powder treatment was administered intratracheally, which enabled precise dose given in the mice lung [34]. The same experimental approach was adopted in the present proof-of-concept study to investigate the synergistic effects of PEV20-ciprofloxacin combination powder treatment by assessing bacterial load and immune responses in the lungs.

MATERIALS AND METHODS

Bacterial strain

Clinical P. aeruginosa FADD1-PA001 was freshly subcultured from −80°C stock prior to the experiment. This strain is resistant to ciprofloxacin, aztreonam and amikacin, with minimum inhibitory concentrations at 32, 256 and 64 mg/L, respectively. A single bacterial colony was inoculated in 20 mL of nutrient broth (NB) for 18 h at 37°C with continuous shaking (150 revolutions per minute, rpm). Overnight culture (10 mL) was mixed with 20 mL of fresh NB and further incubated for 2 h until early-log phase was reached. The early log suspension was centrifuged for 10 minutes and then the pellet was collected. Pellet was washed and diluted by saline for mice lung inoculation.

Phage PEV20 and ciprofloxacin

Phage PEV20 was isolated from the sewage treatment plant in Olympia (WA, USA) by Kutter lab (Evergreen phage lab). Ciprofloxacin hydrochloride was purchased from Sigma-Aldrich Inc.

Spray drying

A Buchi spray dryer (B-290, Buchi Labortechnik AG, Flawil, Switzerland) was employed to prepare ciprofloxacin, PEV20 and combination powder. Liquid feed of ciprofloxacin powder contained 18 mL water solution of 8 mg/mL ciprofloxacin, 8 mg/mL lactose and 8 mg/mL L-leucine and 2 mL phosphate-buffered saline (PBS; 0.01 M phosphate buffer, 0.0027 M KCl and 0.137 M NaCl; pH 7.5). The liquid feed of PEV20 powder was composed of 18 mL water solution of lactose (16 mg/mL) and L-leucine (8 mg/mL) and 2 mL phage suspension (10⁹ PFU/mL) in PBS. For the ciprofloxacin and PEV20 combination powder, the liquid feed was composed of 18 mL water solution of ciprofloxacin (8 mg/mL), lactose (8 mg/mL) and L-leucine (8 mg/mL) and 2 mL of phage suspension (10⁹ PFU/mL) in PBS (Table 1). A conventional two fluid nozzle with diameter 0.7 mm was used for atomization. The suspension was fed at a constant feed rate of 1.8 mL/min and an atomizing airflow of 742 L/h with an aspiration rate of 35 m³/h. The drying inlet air was heated to 60°C and the outlet temperature range was between 41 and 42°C. Dried powders were collected in a vial after passing through the cyclone. Powders were prepared and stored in a desiccator with silica beads at room temperature before administration.

Table 1.

Concentration of each excipient in liquid feed for spray drying of single Ciprofloxacin, single PEV20 and the combination powder

Powders	Excipients (mg/mL) in 18 mL water			Phage titre (PFU/mL) in 2 mL PBS
	Ciprofloxacin	Lactose	L-leucine
Ciprofloxacin	8	8	8	-
PEV20	-	16	8	109
Combination	8	8	8	109

Animal experiments

All animal experiments were conducted with approval of the University of Sydney Animal Ethics Committee (Sydney, NSW, Australia) [AEC approval number: 2016/017C]. Female BALB/c mice (8–10 weeks, 18–21 g) were obtained from Australian BioResources (Moss Vale, NSW, Australia) and were kept under specific pathogen-free conditions in the animal facility at the Centenary Institute (Camperdown, NSW, Australia). Neutropenia was induced by intraperitoneal (IP) injection of cyclophosphamide (Baxter Healthcare Pty Ltd., New South Wales, Australia) on one (dose 150 mg/kg) and four (dose 100 mg/kg) days before inoculation. On the day of inoculation, mice were anesthetized by IP injection of ketamine/ xylazine solution (80/8 mg/kg), and then fixed in the supine position against the restraining platform angled at 60 to 70° from the horizontal. A total of 25 μL FADD1-PA001 suspension (approximately 10⁶ cells) was sprayed directly into trachea using a MicroSprayer (model IA-1C; Penn-Centuray, Philadelphia, PA, USA). Mice were maintained in the upright position for 1 minute and then placed on a warm pad for recovery. After 2 hours, powders (1 mg) of single ciprofloxacin (0.33 mg), single PEV20 (10⁶ PFU/mg) and the combination were aerosolized into the trachea of anesthetized mice from corresponding groups (n=4) using a Dry powder insufflator (model DP-4M; Penn-Century). Two additional groups of (i) untreated control (healthy without any treatment) and (ii) neutropenic control (infected and untreated) were set up as controls. Mice were culled via CO₂ overdose at 24 h post-treatment. The left lung was harvested aseptically and homogenized in 2 mL saline for colony counting. Superior, middle and inferior lobes were harvested for flow cytometry analysis.

Colony counting

Series dilution (1:10) was made for homogenized left lung sample. The Nutrient broth (NB) and agar plate (NB broth with 1.4% agar) was inoculated with 50 μL diluted suspension. Bacterial density was counted and calculated after an overnight incubation at 37°C.

Flow cytometry analysis

Two million lung cells were incubated with 1.25 μg ml–1 anti-CD32/CD16 (eBioscience, San Diego, CA) in FACS wash buffer (PBS/2% FCS/0.1%) for 30 min to block Fc receptors, then washed and incubated for 30 min with anti-CD4-phycoerythrin (PE)-Cy7 (clone RM4–5, BD), anti-CD8a-Pacific Blue (clone 53–6.7, BD), anti-SiglecF-PE (clone E50–244, BD), anti-B220-allophycocyanin (APC) (clone RA3–6B2, BD), anti-CD11b-APCCy7 (clone M1/70, BD), anti-CD11c-BV785 (clone HL3, BD), anti-Ly6G-BV510 (clone 1A8, BD), anti-Ly6C-PerCPCy5.5 (clone AL-21, BD) and anti-NK1.1-fluorescein isothiocyanate (FITC) (clone PK136, BD). Fixable Blue Dead Cell Stain (Life Technologies) was added to allow dead cell discrimination. Cells were then fixed with formalin. All samples were acquired on a BD LSR-Fortessa flow cytometer (BD), and analysed using FlowJoTM analysis software (Treestar, Macintosh Version 9.8, Ashland, OR). The leukocyte gating strategy is described in [35].

Statistical analysis

The statistical significance of differences between experimental groups was evaluated by one- or two-way analysis of variance (ANOVA), with pairwise comparison of multi-grouped data sets achieved using Tukey or Dunnet post hoc test. Bacterial survival rate at 24 h was calculated by dividing bacterial load of treatment group with that of negative control group. The additive survival rate was calculated by multiplying bacterial survival rates under the treatment of single phage with that of single antibiotics to represent sum of efficacy. Phage-antibiotic synergy was defined if the observed bacterial survival rate of phage-antibiotic combinations was statistically lower than the calculated additive survival rate [1].

RESULTS AND DISCUSSION

In this study, MDR P. aeruginosa load in non-treated control group increased by 4 log₁₀ to 10 log₁₀ CFU/lung after 24 h inoculation (Figure 1). Treatment with intratracheally delivered PEV20-ciprofloxacin combination powder significantly reduced the bacterial load in lungs by 5.9 log₁₀ (p < 0.005), whereas single treatments failed to reduce the burden. The efficacy was synergistic as the observed killing effect for the combination powder was statistically higher than the additive effect of single treatments, both showing nil effect at 24 h. This is the first proof-of-concept study demonstrating synergistic efficacy of combined phage-antibiotic powder treatment in a mouse lung infection model. Other studies showed positive interactions between phage and ciprofloxacin against P. aeruginosa in vitro and in vivo using liquid lysates [17, 36–38]. In an endocarditis rat model, combination therapy of ciprofloxacin and phage cocktail was able to lower bacterial counts in catheter-induced aortic vegetation by 6 log₁₀ CFU/g, while phage cocktail monotherapy led to a reduction of only 2 log₁₀ CFU/g [17]. Although the mechanisms underlying synergy is yet to be elucidated, the observed synergistic bacterial killing in our current study was likely due to a selective pressure under which the bacteria mutate in one trait to improve fitness while suffers decrease in another trait [36]. A recent study demonstrated an evolutionary trade-off effect when phage treatment imposed a selective pressure on MDR P. aeruginosa strain. By acquiring phage-resistance via loss of phage-binding receptor, the bacteria regained sensitivity to several classes of antibiotics including ciprofloxacin [37]. Another study showed morphologic changes to bacterial cells when exposed to sub-inhibitory concentrations of β-lactam antibiotics. The host cells became elongated as a result of antibiotic exposure but did not divide, which might facilitate phage assembly and maturation [39].

Figure 1.

Bactericidal activity of PEV20 only, ciprofloxacin only and PEV20-ciprofloxacin combination powders (PEV20: 10⁶ PFU; Ciprofloxacin: 0.33 mg) in mouse lung infection model (initially 10⁶ bacterial cells) 24 h after treatment (n=4). Statistical significance between groups was determined by ANOVA (*p < 0.005).

In the current study, single PEV20 administration at an MOI of 1 showed no significant bacterial killing at 24 h (Figure 1). Treatment with higher phage dose can potentially improve the observed therapeutic effect at 24 h post-treatment. In our previous study, inhaled delivery of PEV20 powder resulted in a significant bacterial killing at 24 h post-treatment in a similar mouse lung infection model using P. aeruginosa FADD1-PA001 with an MOI of 100 [34]. Such dose-dependent effect was also observed in an Acinetobacter baumannii-induced pneumonia mouse model, which resulted in a rate of survival of 100%, 50% or 16% when single phage were given intranasally at MOI of 10, 1 or 0.1, respectively [40]. Theoretically, the amount of phage should reach beyond an inundation threshold at a certain point in time to observe therapeutic efficacy [41]. This can be done by giving a high dose of phage, or by relying on phage replication in vivo, where the latter is complexed by concurrent systemic clearance. Thus, administration of different phage doses would result in different therapeutical outcomes at a fixed time point. Lower phage dose may have less probability of reaching the inundation threshold at which phages can efficiently infect the host bacteria and result in a therapeutic outcome in vivo. However, in the presence of ciprofloxacin, even a low phage dose of 10⁶ PFU could significantly reduce the bacterial burden in the mouse lungs (Figure 1). The more effective in vivo bacterial killing was perhaps due to a selective pressure [36]. It is much more difficult for bacteria to develop resistance to both antimicrobials due to adaptation trade-offs to the bacteria when facing two concurrently acting selective pressures (i.e. phage and antibiotic), which constrain the resistance mechanisms to antibiotics or phages [37]. In addition, ciprofloxacin inhibits DNA gyrase inside the bacteria while PEV20 could change the activity of the drug efflux pump and thereby increases ciprofloxacin concentration in bacterial cells [1].

However, the ex vivo killing effect of phage and ciprofloxacin might also contribute to the observed in vivo synergy. Harvested samples were processed immediately at cold temperature to minimize the risk of phage-bacterium interactions in vitro [42]. Further experimental technical improvement is necessary to sufficiently remove ciprofloxacin and PEV20 immediately after lung collection.

Powder formulations

The neutropenic mouse lung infection model has been frequently used to minimize immunological impacts on the interactions between phages and bacterial host cells [18, 42–47]. The neutrophil-depleted animals become highly susceptible to bacterial infection, which resembles respiratory infection in immune-deficient patients [10]. In addition, phage may persist longer in a neutropenic model as they can be removed or degraded by both the innate and adaptive immune system [48]. On the other hand, it has also been reported that phage treatment was less effective in neutropenic mice than in normal mice aided by neutrophils [49]. Thus, the treatment outcome depends on both the biological fate of phage and phage-immune synergistic effect [50]. The present study showed that cyclophosphamide injection was able to effectively induce neutropenia in treated mice, reducing lung neutrophil proportions by half (Figure 2a). PEV20, ciprofloxacin or PEV20-ciprofloxacin combination treatment did not result in altered neutrophils levels, indicating no increase in inflammation upon administration of treatment. There are significantly lower proportions of B cells in the ciprofloxacin and PEV20-ciprofloxacin combination treatment groups compared to phage treatment (Figure 2b). This was likely due to off-target effects of ciprofloxacin. P. aeruginosa infection of neutropenic mice is associated with a strong Mo/Mφ and CD8⁺ T cell response in the lungs. However, the addition of PEV20 to ciprofloxacin treatment significantly lowered proportions of lung monocyte/macrophage (Mo/Mφ) and CD8⁺ T cells in comparison to ciprofloxacin treatment alone (Figure 2c–d). There is also a trend towards reduced proportions of Mo/Mφ and CD8⁺ T cells in the PEV20-ciprofloxacin combination treatment group in comparison to the neutropenic control and phage treatment group. However, the reduction in CD8⁺ T cells, B cells and Mo/Mφ cells correlates with reduction in bacterial CFU and this may indicate effective clearance of bacteria and thus more rapid resolution of immune response in lung.

Figure 2.

Leukocyte recruitment to lungs post-infection. BALB/c mice (n = 4) were culled via CO₂ overdose 24 hours post treatment. Lungs were harvested 24h later and cell composition determined by flow cytometry. (A) Neutrophil (Ly6G⁺ CD11b⁺), (B) B cell (B220⁺), (C) CD8+ T cell (CD8⁺), and (D) monocyte/macrophage (Mo/Mϕ, Ly6G⁻ CD11b⁺) proportions were calculated. Data (average ± SEM) is representative of three independent experiments. Statistical significance between groups was determined by ANOVA (*p < 0.05; **p < 0.01).

Inhaled antimicrobial therapy is a promising strategy to tackle bacterial infections in the lung [30]. Our study indicated that the efficacy of ciprofloxacin was enhanced by phage PEV20 and the co-spray dried powder can readily release both antimicrobial agents to have synergistic killing effect on the infection site. Phage can be used simultaneously or sequentially with antibiotics in clinical practice and usually these two agents are delivered through different routes, i.e., IV or oral, leading to variable concentration ratios. The less controlled concentrations may result in negative interactions for certain phage-antibiotic pair on specific bacterial strain. This animal study provided a proof of concept to use co-spray dried phage and antibiotic combination powder to concomitantly treat MDR respiratory infections. However, to implement this particular PEV20-ciprofloxacin combination on human respiratory infections necessitates individual in vitro synergy test. In our previously study, both nebulized and co-spray dried ciprofloxacin-PEV20 combinations showed equivalent synergistic bacterial killing effects in vitro and completely suppressed bacterial regrowth [1,33]. Powder formulations provide easy storage, transport and potentially better patient compliance over liquid formulations, but require dissolution to exert clinical efficacy. In addition, aerosolized powder and solution would have different delivered dose and particle deposition in deep lung. The fine particle fraction (FPF) were reported as 69% and 53% for the air-jet and vibrating mesh nebulized aerosols, respectively. The median mass aerodynamic diameter (MMAD) were 3.7–3.6 μm and 5.1–5.3 μm, respectively. For powder formulation used in the current animal study, the FPF was around 60% and the MMAD was 2.5 μm. The impact of these factors on bacterial killing effects needs further clinical investigation. Chronic safety and pharmacokinetics are also necessary to support clinical dosing regimen design.

CONCLUSION

Co-spray dried phage PEV20 and ciprofloxacin combination powder showed synergistic bacterial killing effect compared with single PEV20 or ciprofloxacin powders in an acute mouse lung infection model caused by P. aeruginosa. The bactericidal effect of combination PEV20-ciprofloxacin was associated with reduced inflammation in the lung. This proof-of-principle study showed the combination powder can potentially be a promising approach to treating MDR P. aeruginosa respiratory infections.