Advanced Spray Dried Proliposomes of Amphotericin B Lung Surfactant-Mimic Phospholipid Microparticles/Nanoparticles as Dry Powder Inhalers for Targeted Pulmonary Drug Delivery

Abstract

The purpose of this study was to design, develop and characterize inhalable proliposomal microparticles/nanoparticles of Amphotericin B (AmB) with synthetic phospholipids, dipalmitoylphosphatidylcholine (DPPC) and dipalmitoylphosphatidylglycerol (DPPG) which are lung surfactant-mimic phospholipids. Organic solutions of AmB and phospholipids, were co-spray dried using an advanced closed-mode system and a high performance cyclone. Scanning electron microscopy (SEM) was employed to visualize the surface structure, morphology, and particles size. The residual water content of the proliposomes was quantified by Karl Fisher coulometric titration (KFT). Degree of crystallinity/non-crystallinity was measured by X-ray powder diffraction (XRPD). Phase behavior was measured by differential scanning calorimetry. The chemical composition by molecular fingerprinting was established using attenuated total reflectance (ATR)-Fourier-transform infrared (FTIR) spectroscopy. The amount of AmB loaded into the proliposomes was quantified using UV-VIS spectroscopy. The in vitro aerosol dispersion performance was conducted using the Next Generation Impactor (NGI) and the human dry powder inhaler (DPI) (Handihaler^®) that is FDA-approved. Different human lung cell lines were employed to demonstrate in vitro safety as a function of dose and formulation. Smooth, spherical microparticles/nanoparticles were formed at medium and high spray drying pump rates and had low residual water content. A characteristic peak in the XRPD diffraction pattern as well as an endotherm in DSC confirmed the presence of the lipid bilayer structure characteristic in the DPPC/DPPG proliposomal systems. Superior in vitro aerosol performance was achieved with engineered microparticles/nanoparticles demonstrating suitability for targeted pulmonary drug delivery as inhalable dry powders. The in vitro cellular studies demonstrated that the formulated proliposomes are safe. These AmB proliposomes can be a better option for targeted treatment of severe pulmonary fungal infections.

INTRODUCTION

Pulmonary fungal infections have emerged as an important cause of morbidity and mortality in immunocompromised patients such as patients with Cystic Fibrosis (CF)^1–4. It is primarily acquired by inhalation of Aspergillus fumigatus spores that can easily deposit and colonize in the peripheral and lower airways^{3, 5}. Amphotericin B (AmB) is one of the antifungal treatment options identified for invasive fungal infections such as aspergillosis ¹. However, when administered systemically AmB is associated with numerous side effects such as nausea, vomiting, rigors, fever, hypertension/hypotension, hypoxia and acute toxicity that sometimes require discontinuation of the treatment⁶. In contrast, pulmonary drug delivery has shown to offer many advantages to treat local diseases in a targeted manner^{7, 8} including pulmonary fungal infections⁹. Pulmonary delivery offers better therapeutic control, high local drug concentration, and reduced side effects ¹⁰ due to the rapid onset of therapeutic action⁷, low enzymatic activity ^{11, 12}, extensive blood supply ¹², and large surface area for high drug absorption^11–13.

The therapeutic activity and toxicity of AmB are related to its structure. As it can be seen in Figure 1a, the molecule owns a macrolide ring, containing an internal lactone, a heptane chromophore, and a hydroxyl hydrophilic region¹⁴. Due to the nature of the molecule, having both polar and non-polar region, it tends to associate with itself and with other lipids. The therapeutic potential of this antifungal is related to its affinity for the sterols that form part of the cell membranes¹⁴. AmB in its monomeric form has a higher affinity for ergosterol, the principal sterol in fungal cell membranes. In contrast, AmB in the self-aggregated state has a higher affinity for cholesterol which is the main sterol in mammalian cell membranes, causing the toxicity to the host cells and severe side effects¹⁴. In order to decrease the toxicity and improve the therapeutic effect of AmB, strategies to decrease the tendency of the antifungal to self-associate have been utilized. Lipid-based formulations such as liposomes lead to the release of the drug in both, its monomeric form and controlled manner¹⁴. In this way, the dose of AmB can be increased without causing several side effects to the host, therefore the elimination of the fungal infection can be successfully achieved.

Figure 1.

Chemical Structure of: (a) Amphotericin B (AmB); (b) dipalmitoylphosphatidylcholine (DPPC); and (c) dipalmitoylphosphatidylglycerol (DPPG) sodium salt.

Spray drying is a particle engineering design technique used for advanced tailored formulation of drug particles in a controllable manner such as particle size, particle distribution, surface morphology, which are critical for pulmonary drug delivery¹⁵. Spray drying using organic solvent has shown to successfully reduce residual water content that is present in solid-state powders using components such as phospholipids¹⁶. Dipalmitoylphosphatidylcholine (DPPC) and dipalmitoylphosphatidylglycerol (DPPG) are the primary phospholipid components in the lung surfactant, and they can offer many advantages in drug delivery due to their transition temperature being above body temperature, resulting in an exceptional controlled drug release kinetics^{16, 17}. In addition, these phospholipids are biocompatible and biodegradable excipients that can improve inhaled particle migration to the peripheral regions of the lung¹⁸. It is worth mentioning that these endogenous phospholipids are available in the market as lung surfactant replacement products administered locally to the lungs¹⁹.

Phospholipid encapsulation of drug molecules can decrease drug toxicity and provide controlled drug release²⁰. Considering the toxicity of unencapsulated AmB, it is rationale to encapsulate this drug using phospholipids and deliver it directly to the airways which can be achieved by inhalation aerosols. Using the spray drying technique, co-spray drying (co-SD) the drug with the phospholipids will render it into thermodynamically stable multilamellar liposomes which in solid state have been called proliposomes¹⁹.

Dry powder inhaler (DPI), which delivers solid state powders to the airways can be used to administer this AmB proliposomes. Our group first reported the advanced formulation of spray dried particles containing DPPC and DPPG with paclitaxel (a highly effective anticancer drug) having exceptional properties such as improved aerosol performance, high drug loading, controlled drug release¹⁶, and thermodynamic stability²¹. In addition, these delivery systems may require smaller doses for efficacy, exhibit reduced toxicity, fewer side effects, and increased formulation stability¹⁹.

The objective of the current study was to design and characterize solid state proliposomal microparticles/nanoparticles of DPPC/DPPG containing AmB for the treatment of pulmonary fungal infections. To the author’s knowledge, this is the first time to report a systematic and comprehensive study on these innovative antifungal lung surfactant-mimic inhalable formulations with in vitro aerosol dispersion performance with an FDA-approved human DPI device, in vitro cellular studies, and comprehensive physicochemical characterization. It is also the first study to engineer AmB lung surfactant-mimic microparticles/nanoparticles from advanced organic solution co-spray drying in a closed-mode.

EXPERIMENTAL: MATERIALS AND METHODS

MATERIALS

AmB [United Stated Pharmacopeia (USP) grade, Spectrum Chemical Manufacturing Corp., USA] [C₄₇ H₇₃ NO₁₇; molecular weight (MW): 924.08 g/mol], synthetic DPPC (MW 734.039 g/mol, >99% purity) and synthetic DPPG (MW 744.952 g/mol, >99% purity) were obtained from Avanti Polar Lipids (Alabaster, Alabama). The chemical structures are shown in Figure 1 (ChemDraw Ultra Ver. 15.0.; CambridgeSoft, Cambridge, MA). Methanol (HPLC grade, ACS-certified grade, purity 99.9%) was obtained from Fisher Scientific (Fair Lawn, NJ). Hydranal^®-Coulomat AD was from Sigma-Aldrich (Milwaukee, WI). Resazurin sodium salt was from Acros Organics (Fair Lawn, NJ). DMSO LC-MS grade was from Thermo Scientific (Milwaukee, WI). Raw AmB, DPPC, and DPPG were stored in sealed glass desiccators over indicating Drierite/Drierite™ desiccant at −20°C under ambient pressure. Other chemicals were stored under room conditions. The nitrogen gas used was ultra-high purity (UHP) nitrogen gas (Cryogenics and Gas facility, The University of Arizona, Tucson, AZ).

Human pulmonary cell lines A549 (ATCC® CCL-185™) and H358 (ATCC® CRL-5807™) were purchased from the American Type Culture Collection ATCC^® (Manassas, VA). Dulbecco’s modified Eagle’s medium (DMEM), Advanced 1X, Fetal Bovine Serum (FBS), Pen-Strep, Fungizone^®, and L-Glutamine were obtained from Gibco^® by Life Technologies (Thermo Fisher Scientific Inc, Milwaukee, WI)

METHODS

Advanced Spray Drying and Co-Spray Drying in Closed-Mode using Organic Solvent

Co-spray dried (Co-SD) of AmB-loaded lung surfactant-mimic particles were designed using a Büchi advanced mini spray dryer B-290 coupled with a Büchi inert loop B-295 and a high performance cyclone (Büchi Labortechik AG, Switzerland). This was performed in closed mode using ultra-high purity nitrogen dry gas as the atomizing gas. A stainless-steel nozzle of 0.7mm diameter was utilized for atomization. The multicomponent feed solution was prepared by co-dissolving DPPC and DPPG in the molar ratio of DPPC:DPPG 75:25 with 10% of amphotericin B on a molar basis to the total DPPC and DPPG and mixed with methanol to form a dilute concentration of 0.1% w/v. This feed concentration successfully produces small particles in the solid-state^{16, 17, 22} and was also the final concentration of the single component feed suspension containing AmB dissolved in methanol. Based on previous studies^{16, 17, 22, 23}, the following conditions were used for spray drying: atomization gas flow rate of 600 L/hr (50 mm), aspirator rate of 35 m³/hr (90%), inlet temperature of 150°C and different pump rates. The different spray drying pump rates were rationally chosen to cover the entire range of 0%−100% as “very low P”, “low P”, “med P”, and “high P” pump rates which correspond to 7.5 mL/min (25%), 15 mL/min (50%), 22.5 mL/min (75%), 30 mL/min (100%) mL/min, respectively. All the SD and co-SD powders were collected in glass vials and sealed with parafilm in desiccant for storage at −20°C under ambient pressure. The spray drying conditions are summarized in Table 1.

Table 1.

Advanced Spray Drying Parameters for Spray Dried (SD) and Co-Spray-Dried (Co-SD) Powders from Organic Solution in Closed-Mode

Powder Composition	Molar Ratio (AmB:DPPC/DPPG)	Feed Concentratio n (% w/v)	Pump Rate (%)	Inlet T (°C)	Outlet T (°C)
SD AmB	100:0	0.1	High (100%)	150	46
SD AmB	100:0	0.1	Med (75%)	150	59
Co-SD AmB:DPPC/DPPG	10:90	0.1	High (100%)	150	46
Co-SD AmB:DPPC/DPPG	10:90	0.1	Med (75%)	150	58–66
Co-SD AmB:DPPC/DPPG	10:90	0.1	Low (50%)	150	71–73

Scanning Electron Microscopy (SEM)

Using conditions similar to previously reported^{15, 16, 21–24}, visual imaging and analysis of particle size, morphology, and surface morphology were achieved by scanning electron microscopy (SEM). Samples were gold sputter-coated for 3 minutes using 15 AC milliAmperes with 7kV voltage. 30kV electron voltage was used for imaging at a working distance of 9–12.5 mm.

Particle Sizing and Size Distributions by Image Analysis

The mean size, standard deviation, and size range of all powders were determined using SigmaScan^™ Pro 5.0.0 (Systat, Inc., San Jose, CA) based on the micrographs obtained by scanning electron microscopy, using a similar procedure that we have previously reported ^{16, 24}. Representative micrographs of each particle sample at 10000x magnification were analyzed by measuring the diameter of at least 100 particles per sample.

X-Ray Powder Diffraction (XRPD)

The degree of long-range molecular order (crystallinity) for all powders was measured by X-ray powder diffraction (XRPD), using conditions comparable to previously reported^{15, 16, 21–24}. The X-ray radiation used was Ni-filtered Cu Kα (45 kv, 40 mA and λ = 1.5418 Å). Measurements were made between 5.0° and 40.0° (2θ) with a scan rate of 2.00°/min at ambient temperature.

Differential Scanning Calorimetry (DSC)

Using conditions similar to previously reported^{15, 16, 21–24}, thermal analysis and phase transition measurements were performed. Approximately 1–10 mg sample was heated from 0.00°C to 250.00°C at a scanning rate of 5.00°C/min for the detection of first order transitions and at a scan rate of 20.00°C/min and 40.00°C/min for the detection of second order transitions.

Hot-Stage Microscopy (HSM) under Cross-Polarizers

Using conditions similar to previously reported^{15, 16, 21–24}, hot-stage microscopy (HSM) was performed. Powder samples were fixed on a glass slide and heated from at least 25.0°C to 250.0°C at a heating rate of 5.00°C/min. The images were digitally captured using a Nikon coolpix 8800 digital camera (Nikon, Tokyo, Japan) under 10x optical objective and 10x digital zoom.

Karl Fisher Titration (KFT)

Using conditions similar to previously reported ^{15, 16, 21–24}, the residual water content of all SD and co-SD powders were quantified analytically by Karl Fischer titration (KFT) coulometrically. Approximately 2–10 mg of powder was added to the titration cell containing Hydranal^® Coulomat AD reagent.

Attenuated Total Reflectance (ATR)-Fourier-Transform Infrared (FTIR) Spectroscopy

Spectrum was collected for 32 scans at a resolution of 2 cm⁻¹ for the wavenumber range of 4000–400 cm⁻¹. A background spectrum was carried out under the same experimental conditions and was subtracted from each sample spectrum. These conditions are similar to our previous reports^{15, 16, 21–24}.

Amphotericin B Loading Analysis by UV-VIS Spectroscopy

UV-Vis spectroscopy was used to quantify the amount of AmB content in the co-SD powders. Raw AmB was dissolved in methanol to make a stock solution of 10 μg/mL concentration, which was subsequently diluted to get concentrations of 7 μg/mL, 5 μg/mL, 2 μg/mL, 1 μg/mL, and 0.5 μg/mL to make a calibration curve. Co-SD powders were also dissolved in methanol at a concentration of 10 μg/mL. A volume of 250 μL of each sample were transferred to a a UV-Vis transparent 96-well plate. The absorbance of the samples was then read at a wavelength (λ) of 405 nm using a PowerWave HT Microplate Spectrophotometer (BioTek Instruments, Inc., Winooski, Vermont ) and analyzed using the Gen5 Microplate Reader and Imager Software (BioTek Instruments, Inc., Winooski, Vermont). The AmB encapsulation efficiency (EE) and drug content (DC) were calculated by Equations 1 and 2:

$$\text{Encapsulation}\text{Efficiency}(\%)=\frac{\text{Actual}\text{mass}\text{of}\text{AmB}}{\text{Initial}\text{mass}\text{of}\text{AmB}}\times 100\%$$ (1)

$$\text{Drug}\text{content}(\text{mg}/\text{mg})=\frac{\text{Actual}\text{mass}\text{of}\text{AmB}}{\text{Mass}\text{of}\text{particles}}$$ (2)

In Vitro Aerosol Dispersion Performance

The in vitro aerosol dispersion performance experiment followed the specifications reported in previous studies^{15, 16, 21, 22, 24} and in accordance with US Pharmacopeia (USP) Chapter <601> on aerosols²⁵. Briefly, three hydroxypropyl methyl cellulose (HPMC) inhalation grade size 3 capsules (Quali-V, Qualicaps, North Carolina) were each loaded with 10mg of powder. One capsule at a time was loaded into the FDA-approved unit-dose capsule-based human DPI device, the Handihaler^® (Boehringer Ingelheim, Ingelheim, Germany). The Next Generation Impactor (NGI)^™ was operated at a flow rate (Q) of 60 L/min (Copley DFM 2000 digital flow meter, Copley Scientific, Nottingham, United Kingdom) with a delay time of 10 seconds before actuation of the inhaler followed by 10 seconds of airflow (Copley TPK 2000 critical flow controller, Copley Scientific, Nottingham, United Kingdom) through the inhaler which is similar to the 10-second breathhold time that patients do when using DPIs. The aerosol deposition was gravimetrically using type A/E glass fiber filters with diameter 55mm (PALL Corporation, Port Washington, New York) and 75mm (Advantec, Japan) which were measured before and after actuation on each stage to determine the particle stage deposition. The fine particle dose (FPD), fine particle fraction (FPF), respirable fraction (RF), and emitted dose (ED) were calculated using the 3–6 equations:

$$\mathrm{FPD}(\%)=\text{Mass}\text{of}\text{particles}\text{deposited}\text{on}\text{stage}2\text{through}7$$ (3)

$$\mathrm{FPF}(\%)=\frac{\text{Fine}\text{particle}\text{dose}}{\text{Initial}\text{particle}\text{mass}\text{loaded}\text{into}\text{capsules}}\times 100\%$$ (4)

$$\mathrm{RF}(\%)=\frac{\text{Fine}\text{particle}\text{dose}}{\text{Total}\text{particle}\text{mass}\text{on}\text{all}\text{stages}}\times 100$$ (5)

$$\text{ED}(\%)=\frac{\text{Initial}\text{mass}\text{in}\text{capsules-Final}\text{mass}\text{remaining}\text{in}\text{capsules}}{\text{Initial}\text{mass}\text{in}\text{capsules}}\times 100\%$$ (6)

The mass mean aerodynamic diameter (MMAD) and the geometric standard deviation (GSD) were determined by using Wolfram Mathematica (Wolfram Research Inc., Champaign, Illinois) written by Dr. Warren Finlay.

In Vitro Human Pulmonary Cell Viability

The effects of AmB formulations on cell proliferation were analyzed by measuring the response of human lung adenocarcinoma and bronchoalveolar carcinoma cells (A549 and H358, respectively) to different concentrations. The A549 pulmonary cell line is a human alveolar epithelial lung adenocarcinoma cell line and is also used as a model of the alveolar type II pneumocyte cell for in vitro pulmonary drug delivery and metabolism studies^{16, 24, 26}. The H358 pulmonary cell line is a human bronchoalveolar epithelial cell line similar to alveolar type II cells and expresses lung surfactant associated protein A (SP-A)^{24, 26}. Cell lines were grown in a growth medium including Dulbecco’s modified Eagle’s medium (DMEM), Advanced 1x, 10% (v/v) fetal bovine serum (FBS), Pen-Strep (100 U/mL penicillin, 100 μg/mL), Fungizone (0.5 μg/mL amphotericin B, 0.41 μg/mL sodium deoxycholate), and 2 mM L-Glutamine in a humidified incubator at 37°C and 5% CO₂, as previously reported ^{16, 24}.

As formerly reported¹⁶, A549 and H358 cells were seeded in 96-well plates at 5,000 cells/well and 100 μL/well and were allowed 48 hours to attach. The cells were then exposed to different concentrations of AmB formulations. The drug solution was prepared by dissolving the drug particles in 3% DMSO and 97% DMEM media. One hundred microliters (μL) of this drug solution or control solution (3% DMSO and 97% DMEM media) were added to each well. Seventy two (72) hours after exposure, 20 μL of 20 μM resazurin sodium salt were added to each well and incubated for 4 hours. At this point, the fluorescence intensity of the resorufin (metabolite) produced by viable cells was detected at 544 nm (excitation) and 590 nm (emission) using the Synergy H1 Multi-Mode Reader (BioTek Instruments, Inc., Winooski, VT). The relative viability of cell line was calculated by equation 7:

$$\text{Relative}\text{viability}(\%)=\frac{\text{Sample}\text{fluorescence}\text{intensity}}{\text{Control}\text{fluorescence}\text{intensity}}\times 100\%$$ (7)

Statistical Analysis

All experiments were performed in at least triplicate (n = 3). The results were analyzed statistically using Microsoft Office Excel 2007 (Microsoft Corporation, Redmond, Washington). The results are expressed as mean ± standard deviation. All graphs were plotted using Sigma Plot 13.0 (Systat Software, Inc, San Jose, CA).

RESULTS

Advanced Spray Drying and Co-Spray Drying in Closed-Mode using Organic Solvent

Particles were only obtained at medium and high pump rate in single component SD systems, whereas particles were obtained at low, medium and high pump rate in co-SD systems. The corresponding outlet temperatures of each system are shown in table 1.

Scanning Electron Microscopy

The particle size and morphology were visualized via SEM. Figures 2 and 3 include the micrographs at different magnifications (10,000x and 20,000x, respectively) of raw AmB, SD AmB, and co-SD AmB:DPPC/DDPG powders at different pump rates. All SD powders showed smooth, uniform, and spherical morphology, while raw AmB showed needle-like morphology. For Co-SD AmB:DPPC/DPPG low P and high P powders, as shown in Figures 2d and 2f, respectively, moderate agglomeration can be observed. At moderate pump rate (Figure 2e), some particles possessed slightly crinkled feature (Figure 3b, c, and e).

Figure 2.

SEM Micrographs of Raw AmB, SD Particles, and co-SD particles: (a) Raw AmB; (b) SD AmB (med P); (c) SD AmB (high P), (d) co-SD AmB:DPPC/DPPG (low P); (e) co-SD AmB:DPPC/DPPG med P); and (f) co-SD AmB:DPPC/DPPG (high P). Magnifications for all micrographs were 10,000x.

Figure 3.

SEM Micrographs of Raw AmB, SD Particles, and co-SD particles: (a) Raw AmB; (b) SD AmB (med P); (c) SD AmB (high P), (d) co-SD AmB:DPPC/DPPG (low P); (e) co-SD AmB:DPPC/DPPG (med P); and (f) co-SD AmB:DPPC/DPPG (high P). Magnifications for all micrographs were 20,000x.

Particle Sizing and Size Distribution by Image Analysis

As shown in Table 2, all the co-SD systems had a geometric mean diameter in the range of 1.10 μm −1.31 μm, while single-component SD samples had a geometric mean diameter of 1.136 μm and 1.152 μm. The majority of the particles were nanopartices/microparticles. All the SD and co-SD powders had particle size range ≤ 5 μm, which is preferred for inhalation therapy. Particles were sized at 10,000x magnification.

Table 2.

Particle Sizing Using Image Analysis on SEM Micrographs (n ≥100 Particles)

System Composition	Mean (µm)	Range (µm)
Raw AmB	1.314±0.605	0.015–2.521
SD AmB (75% PR)	1.136±0.330	0.087–1.979
SD AmB (100% PR)	1.152±0.296	0.663–1.903
Co-SD AmB:DPPC/DPPG (50% PR)	1.105±0.461	0.010–3.893
Co-SD AmB:DPPC/DPPG (75% PR)	1.110±0.317	0.642–2.322
Co-SD AmB:DPPC/DPPG (100% PR)	1.311±0.581	0.611–4.580

X-Ray Powder Diffraction (XRPD)

The XRPD pattern of raw AmB showed (Figure 4) sharp and intense peaks (i.e long-range molecular order) suggesting crystallinity of the drug. The following diffraction peaks 14.17°, 15.41°, 17.39°, 21.39°, and 21.90° 2θ were seen, which are similar to what has been previously reported ²⁷. In contrast, SD AmB at both pump rates of 75% PR and 100% PR did not show characteristic peaks indicating the absence of long-range molecular order. XRPD diffractograms of Co-SD AmB:DPPC:DPPC at 50% P, 75% PR, and 100% PR (Figure 3) showed the presence of a strong peak at 21° 2θ, which corresponds to the presence of the phospholipid bilayer structure. This is in good agreement with what has been previously reported¹⁶.

Figure 4.

X-ray Powder Diffractograms of Raw AmB, SD AmB, and co-SD AmB:DPPC/DPPG.

Differential Scanning Calorimetry (DSC)

DSC thermograms of raw AmB, SD single-component, and co-SD particles can be seen in Figure 5. Raw AmB, SD AmB (med P), and SD AmB (high P) exhibited a single endothermic transition at 148°C, 160°C, 151°C, respectively, suggesting a transition from an ordered to a disordered phase. At faster heating scan rates of 20 °C/min and 40 °C/min, glass transition temperatures (T_g) were not identified (data not shown). As described in previous studies^{16, 17}, the miscibility of DPPC with DPPG and the presence of the phospholipid bilayer is confirmed in all formulated co-SD AmB:DPPC/DPPG particles which, exhibited the characteristic bilayer main transition phase (T_m) at ~64°C with the exception of the high P particles which exhibited transition at ~72°C. As can be seen in Table 3, the decrease in enthalpy on the bilayer T_m was observed for the co-SD AmB:DPPC/DPPG particles with increasing pump rates. The two endothermic peaks were observed for the low, med and high pump rates particles suggesting decreased molecular miscibility between AmB and DPPC/DPPG components.

Figure 5.

Representative DSC Thermograms for: (a) raw AmB; (b) SD AmB (med P); (c) SD AmB (high P); (d) co-SD AmB:DPPC/DPPG (low P); (e) co-SD AmB:DPPC/DPPG (med P); and (f) co-SD AmB:DPPC/DPPG (high P).

Table 3.

DSC Thermal Analysis (n=3, mean±SD)

Powder Composition	Spray Drying Pump Rate (%)	Tpeak1 (°C)	ΔH1 (J/g)	Tpeak2 (°C)	ΔH2 (J/g)
Raw AmB	N/A	148.74±2.20	97.44±37.57	N/A	N/A
SD AmB	Med (75%)	160.14±3.88	73.33±45.59	N/A	N/A
SD AmB	High (100%)	151.20±8.00	87.35±27.29	N/A	N/A
Co-SD AmB:DPPC/DPPG	Low (50%)	64.00±0.27	24.45±6.87	157.94±13.88	42.61±27.44
Co-SD AmB:DPPC/DPPG	Med (75%)	65.23±1.51	20.22±0.72	155.02±2.96	60.83±6.98
Co-SD AmB:DPPC/DPPG	High (100%)	61.59±0.87	24.24±3.50	168.27±7.73	27.57±34.12

Hot-Stage Microscopy (HSM) Under Cross-Polarizer Lens

Co-SD AmB:DPPC/DPPG powders were visualized under cross-polarized light microscopy as a function of increasing temperature for the presence or absence of birefringence. HSM micrographs are presented in figure 6. All the samples showed no birefringence in all temperatures, suggesting the lack of crystallinity. Furthermore, there was no difference in birefringence observed at room temperature (25°C) and at body temperature (37°C), suggesting the stability of these powders at those temperatures. For co-SD AmB low P and med P, an evident phase transition was observed between 95–98°C, while for high P it was observed between 95–111°C. Melting of the particles produced at low P, med P, and high P was observed at 150°C, 167°C, and 142°C, respectively.

Figure 6.

Representative HSM Images for: (a) co-SD AmB:DPPC/DPPG (low P); (b) co-SD AmB:DPPC/DPPG (med P); and (c) co-SD AmB:DPPC/DPPG (high P) (scale bar: 3mm).

Karl Fisher titration (KFT)

The residual water content of raw and spray dried powders is shown in Table 4. The residual water content of raw AmB was 9.42% (w/w), whereas the amount of water content significantly decreased for all the powders after spray drying ranging from 2.25% (w/w) to 6.50% (w/w) with the exception of SD AmB (high P) powder which slightly decreased to 8.27% (w/w) Statistical analysis showed no significant difference compared to raw AmB (P >0.05). It can be noted from Table 4 that as the pump rate increased for the co-SD AmB:DPPC/DPPG powders, the residual water content decreased.

Table 4.

Residual Water Content (n=3, mean ± SD)

Powder Composition	Spray Drying Pump Rate (%)	Residual Water Content (% w/w)
Raw AmB	N/A	9.42±1.23
SD AmB	Med (75%)	5.12±1.14
SD AmB	High (100%)	8.27±0.22
Co-SD AmB:DPPC/DPPG	Low (50%)	6.77±0.40
Co-SD AmB:DPPC/DPPG	Med (75%)	6.50±0.16
Co-SD AmB:DPPC/DPPG	High (100%)	2.25±0.35

Attenuated Total Reflectance (ATR)-Fourier Transform Infrared (FTIR) Spectroscopy

Raw AmB, SD AmB, and co-SD AmB:DPPC/DPPG powders were analyzed by ATR-FTIR spectroscopy, as shown in Figure 7. The spectra of raw AmB and SD AmB powders (med P and high P) obtained are similar to the spectra reported in Scifinder^® and in good agreement with the literature²⁷. As can be seen in Figure 7, a broad peak was observed around ~3,500 cm⁻¹, which represents O-H stretching present in the molecule. Furthermore, -CH₂ antisymmetrical stretching around ~2,917 cm⁻¹, -CH₃ symmetrical stretching around ~2,849 cm⁻¹, C=O stretching of lactone around ~1,734 cm⁻¹, all these characteristics peaks representing of AmB were seen. Co-SD DPPC:DPPG spectra were consistent with the previously reported spectra ¹⁶.

Figure 7.

ATR-FTIR Spectra for: (a) raw AmB; (b) SD AmB (med P); (c) SD AmB (high P); (d) co-SD AmB:DPPC/DPPG (low P); (e) co-SD AmB:DPPC/DPPG (med P); and (f) co-SD AmB:DPPC/DPPG (high P).

Amphotericin B Loading Analysis via UV-Vis Spectroscopy

The encapsulation efficiency (EE) and drug loading of AmB in the co-SD systems were analyzed via UV-Vis spectroscopy. As shown in Table 5, all of the samples had an EE ranging between 1.20% w/w - 1.67% w/w. The AmB loading in co-SD systems at low P, med P, and high P were 0.202 mg, 0.146 mg, and 0.155 mg of AmB per mg particle, respectively.

Table 5.

List of Co-SD Formulation Compositions, AmB Encapsulation Efficiency (EE) and AmB content (mg/mg) (n=3, mean ± SD)

Powder Composition	Spray Drying Pump Rate (%)	AmB EE (%)	AmB loading (mg/mg)
Raw AmB	N/A	N/A	N/A
SD AmB	Med (75%)	N/A	N/A
SD AmB	High (100%)	N/A	N/A
Co-SD AmB:DPPC/DPPG	Low (50%)	1.67±0.98	0.202±0.118
Co-SD AmB:DPPC/DPPG	Med (75%)	1.20±0.16	0.146±0.0.019
Co-SD AmB:DPPC/DPPG	High (100%)	1.27±0.50	0.155±0.060

In Vitro Aerosol Dispersion Performance

The in vitro aerosol performance was analyzed for all the powders using the NGI^™ coupled with a Handihaler^® human DPI device that is FDA-approved. As listed in Table 6, the ED values ranged from 66.6% −78.0% for SD AmB powders and ranged from 74.5% −82.2% for co-SD AmB:DPPC/DPPG powders. The FPF ranged from 47% - 50% for SD AmB, while it ranged 13% - 47% for co-SD powders. The calculated MMAD for SD AmB was 2 μm - 2.2 μm, for co-SD it was 2.2 μm - 12.1 μm. Figure 8 demonstrates the in vitro aerosol dispersion performance of SD and co-SD powders by showing the percentage deposition of the particles on each of the NGI stages. Particle deposition on all stages was observed for all the powders.

Table 6.

In Vitro Aerosol Performance Parameters Using the Next Generation Impactor for SD and Co-SD Aerosol Systems Including Mass Median Aerodynamic Diameter (MMAD), Geometric Standard Deviation (GSD), Fine Particle Fraction (FPF), Respirable Fraction (RF), and Emitted Dose (ED) (n=3, mean ± SD)

System Composition (Spray Drying Pump Rate)	ED (%)	FPF (%)	RF (%)	MMAD (µm)	GSD
Raw AmB	89.7±3.6	30.0±3.6	57.2±4.0	4.9±0.5	2.9±0.5
SD AmB (med P)	78.0±6.4	49.9±9.7	86.1±7.8	2.2±0.5	2.0±0.2
SD AmB (high P)	66.6±21.3	47.3±20.0	93.2±1.7	2.0±0.0	1.9±0.0
Co-SD AmB:DPPC/DPPG (low P)	75.9±16.9	13.0±1.6	32.4±9.6	12.1±5.2	3.5±0.5
Co-SD AmB:DPPC/DPPG (med P)	82.2±11.9	22.4±4.1	46.0±4.3	5.3±1.0	2.7±0.2
Co-SD AmB:DPPC/DPPG (high P)	74.5±9.9	46.8±5.4	93.6±0.7	2.2±0.1	1.8±0.1

Figure 8.

In Vitro Aerosol Dispersion Performance Using the NGI^® Under an Airflow Rate (Q) of 60 L/min with the HandiHaler^® Human DPI Device for SD AmB Powders and Co-SD AmB:DPPC/DPPG (n=3, mean ±SD).

In vitro Human Pulmonary Cell Viability

In vitro cell viability response, as seen in Figure 9, was analyzed by exposing H358 and A549 to different concentrations of AmB using SD and co-SD powders. For A549 there was not significant change in relative viability when they were exposed to spray dried powders (including co-SD formulations) in concentrations ranging from 0.001 μM to 100 μM with the exception of co-SD AmB:DPPC/DPPG med P which relative viability decreased to ~80% at 0.01 μM and to ~75% at 100 μM. For H358 cells, there was a significant decrease in relative viability for all the powders at 100 μM with the exception of co-SD AmB:DPPC/DPPG high P.

Figure 9.

In vitro Drug Dose-Response Curves for: (a) A549 Cell Dose-Response to Raw AmB, SD AmB, and Co-SD AmB:DPPC/DPPG Powders and (b) H358 Cell Dose-Response to Raw AmB, SD AmB, and Co-SD AmB:DPPC/DPPG Powders; (n=6, mean ± SD).

DISCUSSION

This study showed the physicochemical and in vitro aerosol performance properties of rationally formulated co-SD AmB-loaded DPPC/DPPG and SD AmB as dry powders for pulmonary therapy. This is the first time to the author’s knowledge that these surfactant-based dry powders have been formulated encapsulating AmB. Using these microparticles/nanoparticles biomimetic surfactant-based dry powders offers many advantages including appropriate size range for targeted deposition in the respiratory tract, enhanced stability, and better solubility¹⁵.

This systematic experimental design resulted in the development of multi-component co-SD AmB:DPPC/DPPG powders (low P, med P, and high P) and single-component SD AmB powders (med P and high P). Furthermore, this study compares the effect of spray drying pump rate for the optimization of the SD powder with appropriate aerosol properties. As our group has previously described ^{16, 17}, organic solution spray drying results in smaller particle sizes due to the decreased surface tension of the solvent in comparison to water. In addition, this method also reduces the residual water content in the final powders resulting in enhanced stability and in vitro aerosol dispersion performance ²¹.

SEM analysis showed that all SD powders including single component and multiple component powders had ideal size ranges < 2 μm, which is comparable to previously developed polymer based amphotericin-B loaded in PEG-g-PLA nanoparticles²⁸. Alternate lipid based drug delivery systems like lipid nanoemulsion had a volume median diameter of ~5 μm, however, this is liquid aerosol delivered through nebulization²⁹. Instead, the size range of these particles are ensuring the potential to effectively deliver a high payload to lower regions of the lung, as well³⁰.

XRPD data indicates that single component SD powders are non-crystalline due to the lack of long range molecular order compared to raw AmB. This is in good agreement with HSM images since birefringence was not observed in any of these powders. However, amorphous nature was not verified in DSC analysis since no glass transition temperature (T_g) was observed at faster scan rates (data not shown). Spray drying can produce amorphous particles which are thermodynamically less stable than crystalline particles³¹. However, in the above formed AmB particles, the lower residual water content can attribute to increased stability of the powder for prolonged storage time³². In addition, decreased residual water content can lead to enhanced aerosol dispersion performance²⁴.

XRPD and DSC of the formulated co-SD particles demonstrated the presence of the lipid bilayer structure, evident from the signature peaks of XRPD and characteristic endothermic peaks (T_m) of DSC thermograms. HSM confirmed the presence of the phase transitions of the co-SD AmB:DPPC:DPPG powders as observed in DSC also. It also confirmed the stability of these powders at room (25°C) and physiological temperatures (37°C). ATR-FTIR analysis of particles in solid state confirmed the presence of DPPC and DPPG in all the co-SD powders together with the presence of AmB similar to previous reports^{16, 27}. Furthermore, ATR-FTIR analysis also showed that there was no chemical change in AmB after spray drying.

UV-Vis spectroscopy was used to analyze encapsulation efficiency and drug loading resulting with effective encapsulation and confirming the presence of AmB within the phospholipid nanocarriers. Moreover, excellent in vitro aerosol dispersion performance was achieved for all the SD powders at Q = 60 L/min (average adult airflow rate and standard Q for DPI testing)³³ using the NGI coupled with Handihaler^® device. Pump rate effect from spray drying was observed for all SD powders due to that a decreased in MMAD and GSD was observed as pump rate was increased, as seen in Table 5. The FPF and RF values also increased upon the increasing pump rate only for the co-SD AmB:DPPC/DPPG powders while there was not a noticeable trend for single component SD AmB powders. For MMAD, there was no noticeable trend due to pump rate effect. The amount of particle deposited on the lower stages increased with increasing pump rate.

The observed physicochemical properties and in vitro aerosol deposition profiles indicate that these particles are suitable for targeted pulmonary delivery as inhalation aerosols providing high local drug deposition particularly in the small airways and alveolar region. The powders developed at high pump rate (100%) showed higher particle deposition on the lower NGI stages 4 to 7 (Figure 8) which suggests that these powders might be less aggregated than the particles spray dried at the lower pump rates. This means that the formulated particles are capable of reaching the smaller airways and the bronchioalveolar region which are crucial for an effective antifungal treatment ⁹. Increasing spray drying pump rate resulted in particles that produced higher RF values.

In vitro human pulmonary cellular studies revealed important information about SD and co-SD AmB powders safety. The dose response (Figure 9) indicates that all the formulated powders are completely safe for pulmonary therapy 72 h after exposure. The addition of phospholipids DPPC and DPPG may allow the possibility of increasing dose concentration without causing cytotoxic effects that are normally related to AmB. This is advantageous since high doses are normally required in order to treat fungal infections⁹. These findings indicate the significant potential of these powders based on phospholipids as drug carriers to be utilized as an effective delivery mechanism for the treatment of pulmonary fungal infections as inhaled dry powder aerosols.

CONCLUSIONS

For the first time, this comprehensive and systematic study reports on the successful design of lung surfactant-mimic DPPC/DPPG dry powder microparticles and nanoparticles which are inhalable containing AmB are suitable for non-invasive targeted lung delivery as inhalation aerosols designed and created by advanced particle engineering technology for targeted pulmonary delivery as advanced co-spray dried DPIs. The in vitro aerosol performance study demonstrated these formulations have the proper aerodynamic properties for targeting the deep lung region. The in vitro pulmonary cellular studies also confirm the ability to use phospholipids to conceal AmB from self-associating, which is the prime reason for toxicity that can eventually lead to targeting fungal cell membranes instead of host cell membranes.