Optimization of inhalable liposomal powder formulations and evaluation of their in vitro drug delivery behavior in Calu-3 human lung epithelial cells

Abstract

Inhalation therapy has advantages for the treatment of multidrug resistant bacterial lung infections with high drug concentrations at the infection sites in the airways and reduced systemic exposure. We have developed liposomal formulations for pulmonary delivery of synergistic ciprofloxacin (Cipro) and colistin (Col) as the potential candidate for treatment of lung infections caused by multidrug resistant Gram-negative bacteria. This study aims to: (1) further optimize the powder formulation by adding drying stabilizers (polyvinyl pyrrolidone or poloxamer) to protect the liposomes during spray-freeze-drying; (2) evaluate the transport and cellular uptake of drugs in a human lung epithelial Calu-3 cell model. The liposomal powder formulations were produced using the ultrasonic spray-freeze-drying technique. The optimal formulation (F5) used mannitol (8% w/v) and sucrose (2% w/v) as the internal lyoprotectants. Adding external lyoprotectants/aerosolization enhancers (i.e. 8% w/v mannitol, 2% w/v sucrose and 1%, w/w PVP 10) produced the superior rehydrated EE values of ciprofloxacin and colistin (50.2 ± 0.9 % for Cipro and 37.8 ± 1.2 % for Col) as well as satisfactory aerosol performance (FPF: 34.2 ± 0.8 % for Cipro and 33.6 ± 0.9 % for Col). The cytotoxicity study indicated that F5 with the colistin concentration at 50 μg/mL and ciprofloxacin at 200 μg/mL was not cytotoxic to human lung epithelial Calu-3 cells. The intracellular uptake of ciprofloxacin was concentration-dependent in Calu-3 cells and the uptake of A-B was more than that of B-A for all samples (p < 0.05). This study demonstrates that co-delivery of ciprofloxacin and colistin in a single liposome can lower the transport capability of both drugs across the Calu-3 cell monolayer and their accumulation in the cells. These findings indicate that co-loaded liposomal powder of ciprofloxacin and colistin is a promising potential treatment for respiratory infections caused by multidrug resistant Gram-negative bacteria.

1. Introduction

Multidrug-resistant (MDR) Pseudomonas aeruginosa poses a serious global risk of difficult-to-treat respiratory infections with recurrence, mortality, and morbidity frequently occurring in patients (Wang et al., 2018; Yeung et al., 2020; Yu et al., 2020). Deaths most often occur from extreme airway inflammation and lung damage. Within recent years, pulmonary delivery of antibiotics has emerged as a promising treatment method because it delivers drugs directly to the infection sites in the lungs, achieving high local drug concentration and less systemic exposure (Mangal et al., 2019a; Mehta et al., 2019; Xu et al., 2019a; Zhou et al., 2015). Dry powder inhalers (DPIs) generally have greater portability, improved chemical and physical stability as well as higher delivery efficiency in comparison to their liquid counterparts (Mangal et al., 2019b; Shetty et al., 2020; Shetty et al., 2018b).

Of the many manufacturing techniques for producing inhalable drug particles, spray-freeze-drying (SFD) has considerable utility because this method creates low-density porous particles that optimize aerosolization performance for pulmonary delivery (D’Addio et al., 2013; Isleroglu and Turker, 2019; Ye et al., 2017). The ultrasonic spray-freeze-drying technique uses an ultrasonic nozzle to atomize a solution into droplets, which are sprayed into liquid nitrogen. The frozen substances are retrieved from the liquid nitrogen and ice crystals are sublimed by lyophilization. Low-density, porous particles can be obtained with relatively smaller aerodynamic diameters after spray-freeze-drying (D’Addio et al., 2012; Edwards et al., 1997; Itatani et al., 2000).

Ciprofloxacin is a fluoroquinolone antibiotic that is often used to treat respiratory tract infections with P. aeruginosa (Darweesh and Sakagami, 2018; Millar et al., 2020; Shetty et al., 2018c). Unfortunately, extensive use of ciprofloxacin has caused an alarming increase of resistance in P. aeruginosa (Mahapatra et al., 2020; Rehman et al., 2019).

Colistin is a cationic lipopeptide antibiotic that has been increasingly used as the last-resort therapy for lung infections (MacNair et al., 2018; Yang et al., 2020). However, intravenous colistin at high doses has led to neurotoxicity and nephrotoxicity (Chien et al., 2020; Giacobbe et al., 2019). Combination antibiotic therapy has emerged as a desirable strategy to counteract antibiotic resistance. Ciprofloxacin and colistin in combination are efficacious in lung infections with Gram-negative pathogens (Buyck et al., 2015; Hoiby et al., 2005; Shetty et al., 2018a).

A controlled release inhalable formulation would reduce the burden of treatment by prolonging and sustaining antibiotic exposure in the lungs (Cipolla et al., 2013; Melis et al., 2016). Previous studies showed that liposomes can extend the drug exposure in the lung (Li et al., 2018; Mehta et al., 2020; Wang et al., 2018; Xu et al., 2019b). Liposomal ciprofloxacin and colistin (Cipro-Col-Lips) for inhalation has been formulated with lipids and cholesterol constructed for direct administration to the infection site and minimized systemic side effects (Ambrus et al., 2019; Schreier et al., 1993; Wang et al., 2017).

Our previous studies have developed the liposomes that were co-loaded with ciprofloxacin and colistin (Yu et al., 2020). The combination liposomes showed greater in vitro antimicrobial activity against multidrug resistant P. aeruginosa than each pure drug (Wang et al., 2018). The formulations also reduced transport capability of both drugs across the human lung epithelial cells (Wang et al., 2018; Yu et al., 2020), which is beneficial for local antimicrobial action at the infection sites on the airway epithelium.

In this study, we optimized our liposomal dry powder inhaler formulations consisting of ciprofloxacin and colistin by adding additional stabilizers of polyvinyl pyrrolidone or poloxamer into the liposomal suspensions before spray-freeze-drying. The physicochemical characteristics, cytotoxicity and the transport of the ciprofloxacincolistin-loaded liposomal (Cipro-Col-Lips) suspensions were examined. The protective effect of lyoprotectant combinations and stabilizers on liposomal suspensions was studied. Furthermore, we applied a human lung epithelial Calu-3 cell model to evaluate drug delivery performance (Grainger et al., 2006; Haghi et al., 2010; Nurbaeti et al., 2018; Zhang et al., 2019). Information gained from these studies will aid further development of such formulations.

2. Materials and Method

2.1. Materials

Ciprofloxacin hydrochloride monohydrate and colistin sulfate were purchased from BetaPharma Co., Ltd. (Wujiang, Jiangsu, China). Hydrogenated soybean phosphatidylcholine (HSPC), 1,2-distearoyl-sn-glycero-3-phosphoglycerol, sodium salt (DSPG-Na, referred to as DSPG in the text), N-(methylpolyyoxyethylene oxycarbonyl)-1,2-distearoyl-sn-glycero-3-phosphoethanolamine and sodium salt (DSPE-PEG-OMe, referred to as PEG in the text) were all obtained from the NOF America Corporation (White Plains, NY, USA). Cholesterol, poloxamer 188 (F68), poloxamer 407 (F127), PVP 10, PVP K30 and PVP K90 were obtained from Sigma-Aldrich (St. Louis, MO, USA).

Dowerx® 50WX4 ion-exchange resin (200~400 mesh) was purchased from Acros Organics (Geel, Belgium), sucrose from Fisher scientific (Fair Lawn, NJ, USA) and DMannitol (mannitol) from DOT Scientific Inc. (Burton, MI, USA). Ammonium sulfate (AS) was obtained from Alfa Aesar (Ward Hill, MA, USA), and 1,2-dipalmitoyl-snglycero-3-phoaphoethanolamine-N-(lissamine rhodamine B sulfonyl) (ammonium salt) (i.e. 16:0 Liss Rhod PE) from Avanti Polar Lipids, Inc. (Alabaster, AL, USA).

The Calu-3 human bronchial epithelial cell line was supplied by American Type Culture Collection (ATCC, Manassas, VA, USA). Penicillin streptomycin (10,000 U/mL penicillin, 10,000 μg/mL streptomycin), PBS pH 7.4 (10X), Hanks’ Balanced Salt Solution (HBSS), MEM non-essential amino acids, fetal bovine serum (heat inactivated, US origin), 0.5 % trypsin-EDTA, goat serum and Dulbecco’s Modified Eagle Medium Nutrient Mixture F-12 (Ham) (DMEM/F-12, 1:1) were all acquired from Gibco (Life Technologies Corporation, Grand Island, NY, USA).

Paraformaldehyde 4 % in PBS was supplied by Biotium, Inc. (Fremont, CA, USA); FITC Annexin V Apoptosis Detection Kit and Hoechst 33342 by BD Biosciences (San Diego, CA, USA). Triton X-100, BCA protein assay kit, radioimmunoprecipitation assay buffer (RIPA buffer) and monensin were purchased from Thermo Scientific (Rockford, IL, USA). Cell Counting Kit-8 was purchased from Dojindo molecular technologies (Tokyo, Japan). Goat anti-mouse IgM antibody conjugated with Alexa Fluor 647, mouse anti-polymyxin B IgM mAb, prolong™ gold antifade reagent and 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) were all acquired from Invitrogen (Life Technologies Corporation, Eugene, OR, USA).

Transwell permeable supports (6.5 mm insert, 0.4 μm polyester membrane) and cell culture flasks were from Corning Incorporated (Kennebunk, ME, USA); Glass bottom dishes (35 mm) from MatTek Corporation (Ashland, MA, USA).

2.2. Cell Culture

The Calu-3 cell line (passage 25–40) was cultured in DMEM/F-12, containing 10 % (v/v) fetal bovine serum, 1 % (v/v) nonessential amino acid solution, penicillin (100 U/mL) and streptomycin (100 μg/mL). Cells were incubated in 5 % CO₂ at 37 °C and RH of 95 % until 80~90 % confluence was achieved. Medium was changed every 2 days, and the cells were passaged weekly in accordance with ATCC guidelines.

To construct the monolayers, Calu-3 cells were seeded at 5×10⁵ cells/cm² in 100 μL medium onto the apical side of the 24-well Transwell® polyester insert membrane and 500 μL medium was added into the basolateral chamber. After 24 hours of incubation, the medium was aspirated from both the apical and basolateral chambers with 500 μL medium being added into the basolateral chamber. The cell monolayer could differentiate for 12–14 days under air interfaced culture (AIC) conditions as previously reported (Haghi et al., 2018; Panduga et al., 2017). After 12–14 days of culture, the monolayer integrity was evaluated by determining the transepithelial electrical resistance (TEER) using a Voltohmmeter (EVOM with STX-3 chopstick electrode, Word Precision Instruments, Sarasota, FL, USA). TEER measurements were taken prior to and post transport experiments.

2.3. Preparation of ciprofloxacin and colistin Co-loaded Liposomes

Cipro-Col-Lips were generated using the established thin film evaporation method (Wang et al., 2018). HSPC, DSPG, PEG and Chol (3:2:0.5:1.7, w/w) were weighed and dissolved in chloroform. A rotary evaporator extracted the organic solvent under lowered pressure to create a thin film at 37 °C. The film was hydrated with 500 mM (NH₄)₂SO₄ solution containing 20 mg/mL sucrose and 80 mg/mL mannitol in a 65 °C water bath for 1 hour. To decrease the size of liposomes, the multilamellar liposomal suspension was then sonicated for 15 min (400 W) using an ultrasonic probe (Sonifier 250, Branson Ultrasonics Corporation, Danbury, CT, USA). Then the blank liposomes were extruded through the polycarbonate membranes of 0.88, 0.45 and 0.22 μm five times in sequence to produce the unilamellar liposomes. The cation-exchange resin spin column removed the extra liposomal NH ⁺₄ of the blank liposomal suspension to establish a transmembrane pH gradient. Finally, the blank liposomal suspension was combined with the solution of ciprofloxacin (16 mg/mL) and colistin (4 mg/mL) containing 20 mg/mL sucrose and 80 mg/mL mannitol, and then incubated 10 min in a bath of 60 °C water for drug loading. The rhodamine-labelled drugs-loaded liposomes were prepared with an addition of rhodamine B with the drugs during the production of liposomes.

2.4. Drug Quantification

We used high-performance liquid chromatography (HPLC) to analyze ciprofloxacin hydrochloride and colistin sulfate (Shetty et al., 2018a; Shetty et al., 2018c; Wang et al., 2018), which includes an Agilent HPLC system with an Agilent Eclipse Plus C18 column (150 mm × 4.6 mm, 5 μm, Agilent, Waldbronn, Germany). The mobile phase consisted of A: 30 mM solution of sodium sulfate (adjusted to pH 2.5 with H₃PO₄, 76 % v/v) and B: acetonitrile (24 % v/v). To measure both drugs, the wavelength was set to 215 nm with a flow rate of 1.0 mL/min.

An established LC-MS/MS method was applied to measure the drugs in the cell studies (Chai et al., 2019). A Kinetex C18 column (2.6 μm, 100 Å, 50 × 3 mm; Phenomenex, Torrance, CA, USA) and a mobile phase of acetonitrile and water with 0.1 % v/v formic acid were used for the analysis with an injection volume of 15 μL and a flow rate of 0.6 mL/min. The chromatographic gradient was set as: 0–0.5 min, 10 % acetonitrile; 1.5 min, 70 % acetonitrile; 2.5 min, 90 % acetonitrile; 3.0 min, 90 % acetonitrile; 3.5 min, 10 % acetonitrile; 5.5 min.

Multiple reaction monitoring (MRM) was used to measure each drug (Table S1) at the condition of: gas temperature at 350 °C, gas flow at 9 L/min, nebulizer pressure at 35 psi, sheath gas temperature at 300 °C, sheath gas flow at 9 L/min, capillary voltage at 4000 V and nozzle voltage at 1000 V.

To determine ciprofloxacin and colistin concentration in the cell study, cells were lysed, and a protein precipitation method was employed. An aliquot of 200 μL cell monolayer lysate supernatant was transferred into a 1.5 mL low protein binding tube. An aliquot of 400 μL acetonitrile with 0.1 % FA was then added into the tube, vortexed for 10 seconds, and centrifuged at 15, 000 rpm for 15 minutes at 4 °C. An aliquot of 300 μL supernatant was placed into an HPLC polypropylene vial for LC-MS/MS analysis.

2.5. Transmission Electron Microscopy of the Liposomal Formulation

The morphology of the liposome suspension was obtained using Cryogenic Transmission Electron Microscopy (cryo-TEM) (Li et al., 2019; Wang et al., 2018). We pipetted an aliquot (3 μL) of the diluted liposomal suspension onto a carbon-coated copper grid and then dabbed with filter paper to acquire a translucent film. The grids were placed in liquid ethane chilled by liquid nitrogen. The translucent films were loaded to a Talos TEM (FEI Company, Hillsboro, OR, USA), and imaged at 300 kV and approximately −173 °C (FEI Company, Hillsboro, OR, USA).

2.6. Liposome Characterization

2.6.1. Zeta Potential and Particle Size Measurement

A Malvern Zetasizer Nano ZS90 (Malvern Instruments Inc., Malvern, UK) was used to determine sizes and zeta potentials of liposomes. Liposomal suspension diluted by distilled water was placed in the sample cuvette, which was then set inside the Zetasizer. Samples were analyzed in triplicate at 25 °C.

2.6.2. Drug Encapsulation Efficiency

Drug encapsulation efficiency was measured as reported previously (Yu et al., 2020). Briefly, the liposome’s suspension (0.5 mL) was transferred into the Amicon® Ultra centrifugation filter (MWCO: 50 K Da, Merck Millipore Ltd., Tullagreen, Ireland) and centrifuged for 30 min at 3000 × g using a MX-200 centrifuge (Tomy Kogyo Co. LTD, Tokyo, Japan). After centrifugation, an aliquot of 100 μL liquid collected in the sample recovery chamber was combined with 100 μL 0.2 M HCl solution, 100 μL distilled water and 700 μL methanol. Then the samples were centrifuged at 15,000 rpm for 15 min, with the supernatant drawn and analyzed by HPLC to determine the free drug content (C₁). Liposomal suspension (100 μL) was combined with 100 μL 0.2 M HCl solution, 100 μL distilled water and 700 μL methanol, and then the entire quantity of the drug (C_o) was evaluated by HPLC. The encapsulation efficiency was established as

$$\text{EE}(\%)=\left[\left({\text{C}}_{\text{o}}-{\text{C}}_1\right)/{\text{C}}_{\text{o}}\right]\times 100\%$$ (Eq. 1).

2.7. Preparation of Liposomes Dry Powder

USFD was used to prepare the liposomal dry powder as described previously (Yu et al., 2020). Previous research has shown that steric stabilizers play a vital role in protecting drug nanoparticles and proteins from freeze-drying (Abdelwahed et al., 2006; Beirowski et al., 2012; Gombotz et al., 1994; Shamblin et al., 1996; Wang, 2000). In our study poloxamer (F68 and F127) and polyvinylpyrrolidone (PVP 10 and PVP K30) were added as the stabilizer in combination with the lyoprotectants of sucrose and mannitol (Beck-Broichsitter et al., 2012; Beirowski et al., 2011; Zeng et al., 2001).

Fig. S1 shows the USFD process. Excipient concentrations are listed in the Table 1. Prior to USFD, poloxamer (F68 and F127) and polyvinylpyrrolidone (PVP 10 and PVP K30) were added into the liposomal suspension. The liposomal suspension containing stabilizers and lyoprotectants were pumped (1B.1003-R/65, Petro Gas ausrüstungen Berlin GmbH, Berlin, Germany) into the ultrasonic nozzle at 1.0 mL/min by an ultrasonic controller at 3.5 watt (BÜCHI Labortechnik AG, Flawil, Switzerland). The liquid nitrogen was evaporated under ambient conditions after atomization and the frozen droplets were lyophilized (Labconco FreeZone, Kansas, MO, USA). The freeze-drying process was conducted at −52 °C and the vacuum level of 0.014 mbar for 48 hours. The Cipro-Col-Lips dry powder was reconstituted with distilled water where applicable.

Table 1.

Compositions of excipients in liposomal suspensions.

Formula	Hydrated solution	Added stabilizers before USFD
F1	8 % Man+2 % Suc+0.5 mol/L AS	2 % F68
F2		5 % F68
F3		2 % F127
F4		5 % F127
F5		1 % PVP 10
F6		5 % PVP 10
F7		1 % PVP K30
F8		5 % PVP K30
F9		-

^*Suc, sucrose; Man, mannitol; AS, ammonium sulfate.

2.8. Particle Morphology

Cipro-Col-Lips particles were characterized by scanning electron microscopy (SEM, NOVA nanoSEM, FEI Company, Hillsboro, OR, USA) at 5.0 kV. Each sample was sprinkled on double-sided tape and coated with a gold film at 40 mA for 90 s (Ling et al., 2019).

2.9. Powder X-ray Diffraction (PXRD)

Powder diffraction patterns were evaluated using a diffractometer (Rigaku Americas, Texas, USA) at 40 kV and 44 mA. The scanning was performed at 5–60 ° 2θ at 4 °/min and a step size of 0.02 ° (Bhujbal et al., 2018).

2.10. In Vitro Aerosol Performance

A Multi-Stage Liquid Impinger (MSLI) (Copley Scientific Limited, Nottingham, UK) was used to measure the aerosolization properties of each powder formulation as previously described (Yu et al., 2020). Each formulation powder (approximately 10 mg) was briefly placed into a size 3 hydroxypropyl methylcellulose capsule (Qualicaps, Whitsett, NC, USA). For each measurement, 10 capsules were fired through a low-resistant RS01 inhaler device (Plastiape S.p.A., Osnago, Italy) at an air flow of 100 L/min for 2.4 s (Zhou et al., 2013b). Each formulation was measured in triplicate. The emitted dose (ED) was defined as proportion of drug collected from capsules and inhaler over the recovered dose. The fine particle fraction (FPF) was calculated as drug particles in Stage 3, 4 and filter (with aerodynamic diameter smaller than 4.9 μm) relative to the total drug recovered (Zhou et al., 2013a).

2.11. Cell Viability

A CCK-8 kit was used to evaluate the cytotoxicity of colistin-ciprofloxacin solution, colistin-ciprofloxacin co-loaded liposomal suspension and blank liposomal suspension in Calu-3 cells (Shao et al., 2018; Zhang et al., 2018). The 96-well tissue culture plates were seeded with Calu-3 cells at a density of 1 × 10⁴ cells/well and incubated at 37 °C for 4 days. The free drugs (ciprofloxacin/colistin, 8 mg/ 2 mg in 1 mL) were dissolved in 8 % mannitol and 2 % sucrose solution and diluted with the culture medium in order to acquire a series of drug solutions, e.g. 200/50, 100/25, 80/20, 40/10, 20/5 and 10/2.5 μg/mL (ciprofloxacin/colistin). Culture medium was used to dilute the ciprofloxacin colistin-loaded liposomes, producing the same drug concentrations as the free drug solutions. Subsequently, the culture medium in the cells was replaced by the drug solutions or liposome suspensions and incubated for 24 hours. After that, the CCK-8 reagents were added and incubated for another 2 hours. The absorbency of each well was measured at 450 nm with a microplate absorbance reader (BioTek, Winooski, VT, USA).

2.12. In vitro Drug Transport Using an Air Interface Pulmonary Epithelial Model

The in vitro drug transport behavior of Cipro-Col-Lips was evaluated in the human lung epithelial Calu-3 cell monolayer model.

2.12.1. Transepithelial Transport Studies

Ciprofloxacin (50, 100, 200 and 400 μg/mL) and colistin (12.5, 25, 50 and 100 μg/mL) transport across the Calu-3 cell monolayer was measured in both apical to basolateral (A-B) and basolateral to apical (B-A) directions. The liposomal suspensions and drug solutions were diluted using HBSS to the pre-determined concentrations. The TEER value of the cell monolayer was measured and both compartments of the cell inserts were rinsed twice with HBSS buffer prior to the experiment.

For the A-B transport experiment, an aliquot of 0.2 mL prewarmed HBSS containing the antibiotics (i.e. ciprofloxacin/colistin with concentrations at 50/12.5, 100/25, 200/50, 400/100 μg/mL) was placed in the apical chamber. An aliquot of 0.6 mL blank HBSS was added into the basolateral chamber. The 0.6 mL sample was collected every hour over a 4-hour period from the basolateral compartments and then replaced with 0.6 mL fresh prewarmed HBSS.

For B-A experiments, prewarmed HBSS (0.6 mL) with the antibiotics (same concentrations as those from A-B) was added into the basolateral compartment. An aliquot of 0.2 mL blank HBSS was added into the apical side of the monolayers. An aliquot of 0.2 mL sample was drawn from the apical side at chosen time points (1, 2, 3 and 4 hours) and subsequently replaced with 0.2 mL fresh prewarmed HBSS. At both the beginning and completion of the study, 0.01 mL was drawn from the donor compartments to determine the initial and terminal concentrations. After the transport experiment, the cells were washed twice with HBSS and then incubated with fresh HBSS in both the apical and basolateral sides for 30 min. At the end of the study, TEER values of the cell monolayer were determined.

Every experiment was performed in quadruplicate and all samples were analyzed via the established LC-MS/MS method. The apparent permeability coefficient (P_app, cm·s⁻¹) was calculated with the equation below:

$$P_{app}=(\text{dQ}/\text{dt})/\left(\text{A}\cdot {\text{C}}_0\right)$$ (Eq. 2)

Where dQ/dt, represented in microgram per second, is the linear appearance rate. A is the surface area of the cell monolayer (0.33 cm²) and C₀ is the starting drug concentration in the donor chamber.

2.12.2. Drug Uptake Studies

In order to analyze the drug content inside the cells, the cell monolayers along with the insert membrane were removed from the Transwell® and transferred into 1.5 mL tubes (Chai et al., 2019; Panduga et al., 2017). RIPA buffer was used to lyse the cells. An aliquot of 200 μL cell lysate was mixed with 400 μL acetonitrile, vortexed for 10 seconds, and centrifuged at 4 °C for 15 min at 15,000 rpm. Drug content in the supernatant was measured using the LC-MS/MS method described above. The Calu-3 cell uptake of drugs was normalized with the cellular protein content. The protein content in the cell lysates was measured using the BCA protein assay kit.

2.13. Drug Distributions in Calu-3 Monolayers Viewed by CLSM

A Nikon-A1R confocal microscope (Nikon America Inc., Melville, NY, USA) was used to image the distribution of ciprofloxacin, colistin and liposomes in the Calu-3 cell monolayer (Golet et al., 2001; Lee et al., 2007). The deposition of ciprofloxacin in the Calu-3 cell monolayer is visible under the confocal microscope with nuclei staining as a position reference (Chai et al., 2019). Colistin is also visible under the confocal microscope after immunofluorescent staining (Ahmed et al., 2019; Velkov et al., 2016). For three-dimensional imaging throughout the monolayers, images were taken every 0.5 μm along the Z-axis.

2.13.1. Ciprofloxacin Distribution

The cells were grown for 14 days at the air interface to enable monolayer differentiation. Unlabeled-liposomal powder (1.0 mg) was sprayed on cell monolayers and incubated for 1 to 4 hours. After incubation, cells were rinsed with PBS, and fixed with paraformaldehyde (4 % in PBS) at room temperature for 1 hour. After three rinses, cells were incubated with propidium iodide solution (1 μg/mL) for 30 min to stain nuclei. Once stained, culture inserts were examined under the microscope.

2.13.2. Colistin Distribution

The cells were grown for 14 days at the air interface to enable monolayer differentiation. Liposome-labeled powder (1.0 mg) was sprayed on cell monolayers and incubated for 1 to 4 hours. After incubation, cells were washed with PBS and treated with 4 % paraformaldehyde for 1 hour and then 0.1 % Triton X-100 for 1 hour. Subsequently, cells were treated with 1 % goat serum for 3 hours and incubated overnight with a mouse polymyxin B IgM mAb (1:500 in blocking buffer, in a darkened chamber at 4 °C). The cells were incubated with the primary antibody overnight and then incubated with the goat anti-mouse IgM conjugated with Alexa Fluor 647 (1:500 in blocking buffer) for 3 hours. The nuclei were stained with Hoechst 33342 (2 μg/mL) for 30 min.

2.14. Statistical Analysis

Statistical analysis was conducted using One-way ANOVA and Dunnett’s multiple comparison test. It was deemed as significant difference if p < 0.05.

3. Results and Discussion

3.1. Characteristics of Ciprofloxacin-Colistin-Liposomal Suspension

The average diameters of the liposomes were around 100 nm and PDI values were under 0.2 (Table 2). The mean EE values of ciprofloxacin and colistin were all above 90 % and 50 %, respectively. All the parameters remained unchanged at 4 °C throughout two months of storage.

Table 2.

Characteristics of liposomes containing ciprofloxacin and colistin over the storage (mean ± SD, n = 3).

Parameter		Days after preparing
		1	30	60
Size (nm)		97.1 ± 0.8	100.3 ± 2.7	102.8 ± 0.9
PDI		0.147 ± 0.051	0.164 ± 0.008	0.173± 0.012
Zeta potential (mV)		−10.1 ± −0.5	−10.1 ± −0.6	−9.8 ± −0.7
EE (%)	Cipro	93.8 ± 0.4	93.5 ± 0.3	95.7 ± 0.3
	Col	55.2 ± 2.4	56.9 ± 0.6	58.5 ± 7.5

Cryo-TEM was used to visualize the morphology of liposomes (Fig. S2). Liposomes are spherical and the sizes are consistent with those determined by DLS (Table 2).

3.2. Optimization of the Formulations

The liposomal suspensions are often physically and chemically unstable. Liposomal dry powder inhaler formulations present potential advantages: greater stability, controlled delivery and minimized airway toxicity. During freeze-drying, addition of the protecting excipients (cryo-/lyo-protectants) are necessary to adequately maintain the properties of liposomes (Beck-Broichsitter et al., 2012; Umerska et al., 2018). Combination of lyoprotectant and steric stabilizer excipient may also further improve the stability when lyophilizing nanosuspensions (Beirowski et al., 2012; Fonte et al., 2016). In previous studies, sucrose (2 %, w/v) and mannitol (8 %, w/v) were used as lyoprotectants in combination with the F68, F127, PVP 10 or PVP K15. These systems were evaluated here, and the concentrations of the steric stabilizers are shown in Table 1. Milli-Q water was used to reconstitute the Cipro-Col-Lips dry powders into suspensions. The steric stabilizers are optimized regarding particle size and EE of suspensions as well as regarding aerosol performance of the powder.

3.2.1. Effect of Steric Stabilizers on the EE of Reconstituted Cipro-Col-Lips Suspensions

As shown in Table 2 and 3, the mean vesicle sizes of reconstituted liposomal suspensions became larger for all the dry powder formulations than those prior to spray-freeze-drying. F6 showed the smallest size of 153.8 ± 2.3 nm after reconstitution. Ultrasonic spray-freeze-drying had a smaller effect on the EE of colistin than that of ciprofloxacin, possibly due to the different positions of two drugs in the liposome (Khatib et al., 2019; Wang et al., 2018). Ciprofloxacin was mostly loaded through the (NH₄)₂SO₄ gradient into the inner aqueous phase of liposomes. However, most of the colistin was adsorbed on the liposome surface via weak hydrophobic interaction. Therefore, atomization could disturb the membrane integrity of liposomes and lead to severe leakage of ciprofloxacin. The formulations with poloxamer as the stabilizer generally had relatively lower EE values (<25 %) of ciprofloxacin and colistin (F1–F4). This could be caused by an increase of poloxamer solubility in the bulk solution during freezing, leading to liposome aggregation (Fonte et al., 2016). When replacing poloxamer with PVP (F5–F8), the EE of drugs significantly increased. In general, the leakage tendency of ciprofloxacin was less apparent with higher steric stabilizer concentrations. Compared to F9, EE of ciprofloxacin significantly increased from 33.6 ± 1.3 % (F9) to 50.2 ± 0.9 % (F5), 55.4 ± 3.7 % (F6) and 53.4 ± 1.0 % (F8) with adding 1 % PVP 10, 5 % PVP 10 and 5 % PVP K15 into the formulation, respectively (p < 0.05). When adding 1 % PVP K15 into the formulation, EE of ciprofloxacin had a slight increase to 37.6 ± 2.0 % (F7). Stabilization of liposomes by PVP could be due to the effects of reducing freezing-induced stresses, surface activity, steric hindrance of nanoparticle interactions (particle isolation hypothesis), and/or increased solution viscosity restricting the movement of liposomes (Beirowski et al., 2012; Wang, 2000).

Table 3.

Properties of reconstituted Cipro-Col-Lips suspension (mean ± SD, n = 3).

Formula	Reconstituted Cipro-Col-Lips suspension
	EE (%)		Size (nm)	PDI	Zeta potential(mV)
	Cipro	Col
F1	14.6 ± 4.6	14.5 ± 6.5	246.8 ± 5.1	0.243 ± 0.017	−18.6 ± −0.4
F2	13.0 ± 2.0	16.9 ± 4.2	228.4 ± 7.3	0.222 ± 0.042	−19.4 ± −0.8
F3	11.4 ± 1.8	20.3 ± 3.7	255.9 ± 1.0	0.275 ± 0.068	−20.4 ± −0.7
F4	12.7 ± 4.5	22.6 ± 3.2	215.1 ± 8.3	0.238 ± 0.018	−21.0 ± −0.6
F5	50.2 ± 0.9	37.8 ± 1.2	183.3 ± 4.2	0.254 ± 0.055	−20.9 ± −0.7
F6	55.4 ± 3.7	41.5 ± 0.2	153.8 ± 2.3	0.185 ± 0.022	−11.4 ± −0.4
F7	37.6 ± 2.0	36.5 ± 2.1	223.7 ± 6.1	0.216 ± 0.004	−21.3 ± −1.0
F8	53.4 ± 1.0	45.8 ± 11.8	191.6 ± 2.3	0.239 ± 0.037	−9.2 ± −0.4
F9	33.6 ± 1.3	46.8 ± 11.9	378.9 ± 8.5	0.256 ± 0.077	−12.4 ± −0.7

3.2.2. Particle Morphology

Fig. 1 shows representative SEM images of various Cipro-Col-Lips dry powder formulations. All F1 – F9 samples have rough surfaces, porous structures and mixture shapes of irregular and near-spherical.

Figure 1.

Representative SEM images of dry powder formulations. Scale bars indicate 20 μm.

Previous studies indicated that an increase in hollow and surface coarseness of the inhalable dry powders could reduce the particle density and area of contact, and produce better aerosol performance (Chew and Chan, 2001; Shetty et al., 2018c). In this study, such porous particles with rough surfaces were produced and we expected they would have satisfactory aerosol performance, which was tested below.

3.2.3. Powder X-ray Analysis

P-XRD patterns of the USFD liposomal powders, raw drugs and USFD materials are shown in Fig. 2. The colistin pattern had no peaks, suggesting the amorphous form; while the patterns of ciprofloxacin had sharp crystalline peaks. All the USFD materials, except PVP, had crystalline peaks. The Cipro-Col-Lips powder formulations exhibited crystalline peaks, corresponding to mannitol and sucrose.

Figure 2.

Powder X-ray diffractograms (PXRD) of the liposomal powders, raw drugs and USFD excipients.

3.2.4. Effect of Steric Stabilizers on the Aerosol Performance of Dry Powder Formulations

Fig. S3 shows the aerosol deposition behavior of the USFD liposomal powders. As shown in Fig. 3, FPF values were significantly decreased when adding 2 % F68, 5 % F68, 5 % F127 or 5 % PVP 10 into the formulations (F1, F2, F4 and F6, p < 0.05), compared to F9 that contains only sucrose and mannitol. When adding 2 % F127 (F3) or 5 % PVP K30 (F8) into the formulation, FPF values were similar to those of F9. However, when 1 % PVP 10 or 1 % PVP K30 was added into the formulations, FPF values were significantly increased to 34.2 ± 0.8 % and 33.6 ± 0.9 % (for F5) or 33.0 ± 0.4 % and 33.4 ± 0.5 % (for F7), ciprofloxacin and colistin, respectively (p < 0.05). Based on the results of EE and FPF, we chose F5 for further studies. The EEs of ciprofloxacin and colistin for F5 were 50.2 ± 0.9 % and 37.8 ± 1.2 %, respectively; and the FPFs of ciprofloxacin and colistin for F5 were 34.2 ± 0.8 % and 33.6 ± 0.9 %, respectively.

Figure 3.

FPF and ED profiles of the powder formulations (mean ± SD, n = 3).

3.3. Cytotoxicity Study

The viability of Calu-3 cells incubated with free drugs, blank liposome suspension or liposome formulation (F5) are shown in Fig. 4. The drug-loaded liposome suspension, rehydrated drug-loaded liposome suspension and blank liposome suspension did not show significant cytotoxic activity on Calu-3 cells for concentrations from 10/2.5 to 200/50 μg/mL (Cipro/Col). Only free drugs showed significant cytotoxicity at the highest concentration (Cipro/Col: 200/50 μg/mL, P < 0.05). These results proved that the liposome formulations caused less cytotoxicity than pure drug solutions to Calu-3 human lung epithelial cells.

Figure 4.

Viability profiles in Calu-3 cells (mean ± SD, n = 6). Note: *: significant difference as compared to other samples at that drug concentration, p < 0.05.

3.5. In Vitro Drug Transport in the Calu-3 Cell

3.5.1. Ciprofloxacin Transport

For both absorptive (A-B) and secretory (B-A) transport, the apparent permeability coefficient (P_app) of ciprofloxacin across the Calu-3 cell monolayers decreased with an increase in drug concentrations from 50 to 400 μg/mL (Fig. 5). For ciprofloxacin liposomes, there was a difference in P_app values between A-B and B-A directions (p < 0.05), whereas there was no distinction in P_app values between the two directions for the rehydrated liposomes suspension or for the free drug. These findings could be due to a high amount of free drug in the rehydrated liposomes suspension. The process of spray-freeze-drying decreased the EE of ciprofloxacin (lowered from 93.8 ± 0.4 % to 50.2 ± 0.9 % from Tables 2 and 3) and the free drug could have been rapidly absorbed and cleared from the lung (Lamy et al., 2018). Therefore, P_app values of the rehydrated liposomal suspension and free drug were higher than those of liposomal suspensions. Cavet-Megan E. et al., have also reported that the flux of ciprofloxacin solution in the absorption and secretory directions was similar in magnitude and dependent on the drug concentration (Cavet et al., 1997). As shown in Fig. 5B, the intracellular uptake of ciprofloxacin was a concentration-dependent process and the uptake of A-B was more than that of B-A for all samples (p < 0.05).

Figure 5.

Ciprofloxacin transport across the Calu-3 cell (mean ± SD, n = 4). A: P_app; B: drugs transported into the cell; C: drugs remaining in the donor chamber. Note: *: significant difference, p < 0.05.

3.5.2. Colistin Transport in the Calu-3 Cell

For both (A-B) and (B-A) transport, the P_app values of colistin across the Calu-3 cell were decreased with increasing drug concentration (from 12.5 to 100 μg/ml) (Fig. 6A). However, the P_app values of A-B were greater than those of B-A for all three groups, indicating that colistin transport across the cell monolayer is greater from A to B than from B to A (p < 0.05). In comparison to ciprofloxacin, the colistin transported into the Calu-3 epithelial cell monolayer was concentration-independent, and the uptake of BA was more than that of A-B over 4 hours (p < 0.05) (Fig. 6B). Fig. 6C shows that more than 90 % of the drug was retained in the donor compartment.

Figure 6.

Colistin transport across the Calu-3 cell (mean ± SD, n = 4). A: P_app; B: drugs transported into the cell; C: drugs remaining in the donor chamber. Note: *: significant difference, p < 0.05.

Transepithelial electrical resistance (TEER) is commonly used to evaluate membrane integrity and tight junction dynamics in epithelial cells (Zhang et al., 2019). For both ciprofloxacin and colistin transport, the TEER results (showed in Fig. S4) were all above 500 Ω·cm², which suggested that the transport, at the drug concentration of 50/12.5 μg/mL to 400/100 μg/mL, had no adverse impact on the integrity of the epithelial cell monolayer (Chai et al., 2019; Haghi et al., 2018).

Rapid antibiotic absorbance and clearance from the lung is one of the largest obstacles for localized treatment of lung infections (Brillault et al., 2017). To overcome this obstacle, our study developed a Ciprofloxacin/Colistin-loaded liposomal formulation with controlled release capability, which could provide the prolonged residence time of ciprofloxacin and colistin at the airway surface.

3.6. Drug Distribution throughout the monolayers viewed by CLSM

As demonstrated in Fig. 7 and Fig. 8, the intensity of ciprofloxacin and colistin in the cell monolayers increased over time (from 1 to 4 hours). A greater amount of ciprofloxacin was accumulated in the monolayers, which was consistent with our previous studies (Chai et al., 2019). In vivo efficacy study of our formulation is warranted in the established mouse lung infection model (Lin et al., 2017a; Lin et al., 2017b).

Figure 7.

Ciprofloxacin distributions in the Calu-3 cell at different time points. The red color denotes propidium iodide labeled nucleus; the blue color denotes autofluorescence of ciprofloxacin. Scale bar 100 μm.

Figure 8.

Colistin distributions in the Calu-3 cell at different time points. The blue color denotes Hoechst 33342 labeled nucleus; the red color denotes rhodamine B labeled liposomes; the green color denotes immunofluorescence staining colistin. Scale bar 100 μm.

4. Conclusion

The co-loaded ciprofloxacin and colistin liposomal dry powder has the potential to become a viable therapy for Gram-negative lung infections. The optimized formulation (F5) had the rehydrated EE values of 50.2 ± 0.9 % for Cipro and 37.8 ± 1.2 % for Col, with fair aerosol performance (FPF: 34.2 ± 0.8 % for Cipro and 33.6 ± 0.9 % for Col). The cell viability results indicate that the optimized formulation (F5) was safe at least at the colistin concentration up to 50 μg/mL and ciprofloxacin up to 200 μg/mL.

Ciprofloxacin transported more rapidly into the Calu-3 cells than colistin. Meanwhile, co-delivery of ciprofloxacin and colistin in even a single liposome can lower the transport capability of both drugs across the Calu-3 cell monolayer and accumulation in the cells. With prolonged retention and sustained release of both drugs at the airway surface, the liposomal formulation has a great potential to become an antibiotic drug delivery system that can better treat respiratory tract infections.