Inhaled bosentan microparticles for the treatment of monocrotaline-induced pulmonary arterial hypertension in rats

Hyo-Jung Leea, Yong-Bin Kwona, Ji-Hyun Kanga, Dong-Won Oha, Eun-Seok Parkb,
Yun-Seok Rheec, Ju-Young Kimd, Dae-Hwan Shina, Dong-Wook Kime,⁎, Chun-Woong Parka,⁎⁎
a College of Pharmacy, Chungbuk National University, Cheongju 28160, Republic of Korea
b School of Pharmacy, Sungkyunkwan University, Suwon 16419, Republic of Korea
c College of Pharmacy and Research Institute of Pharmaceutical Sciences, Gyeongsang National University, Jinju 52828, Republic of Korea
d College of Pharmacy, Woosuk University, Wanju-gun 55338, Republic of Korea
e Department of Pharmaceutical Engineering, Cheongju University, Cheongju 28503, Republic of Korea

A R T I C L E I N F O

Keywords:
Pulmonary delivery Dry powder inhalers
Bosentan
Pulmonary arterial hypertension Spray-drying
Monocrotaline

A B S T R A C T

The conventional treatment of pulmonary arterial hypertension (PAH) with oral bosentan hydrate has limita- tions related to the lack of pulmonary selectivity. In this study, we verified the hypothesis of the feasibility of dry powder inhalation of bosentan as an alternative to oral bosentan hydrate for the treatment of PAH. Inhalable bosentan microparticles with the capability of delivery to the peripheral region of the lungs and enhanced bioavailability have been formulated for PAH. The bosentan microparticles were prepared by the co-spray- drying method with bosentan hydrate and mannitol at different weight ratios. The bosentan microparticles were then characterized for their physicochemical properties, in vitro dissolution behavior, and in vitro aerodynamic performance. The in vivo pharmacokinetics and pathological characteristics were evaluated in a monocrotaline- induced rat model of PAH after intratracheal powder administration of bosentan microparticles, in comparison to orally administered bosentan hydrate. The highest performance bosentan microparticles, named SDBM 1:1, had irregular and porous shape. These microparticles had not only the significantly highest aerosol performance (MMAD of 1.91 μm and FPF of 51.68%) in the formulations, but also significantly increased dissolution rate, compared with the raw bosentan hydrate. This treatment to the lungs was also safe, as evidenced by the cy- totoXicity assay. Intratracheally administered SDBM 1:1 elicited a significantly higher Cmax and AUC0-t that were over 10 times higher, compared with those of the raw bosentan hydrate administered orally in the same dose. It also exhibited ameliorative effects on monocrotaline-induced pulmonary arterial remodeling, and right ven- tricular hypertrophy. The survival rate of the group administrated SDBM1:1 intratracheally was 0.92 at the end of study (Positive control and orally administrated groups were 0.58 and 0.38, respectively). In conclusion, SDBM 1:1 showed promising in vitro and in vivo results with the dry powder inhalation. The inhaled bosentan microparticles can be considered as a potential alternative to oral bosentan hydrate for the treatment of PAH.

1. Introduction

Pulmonary arterial hypertension (PAH) is a severe and progressive disease of the pulmonary circulation that is characterized by constric- tion and remodeling of the pulmonary vasculature (especially the small pulmonary arterioles). Symptoms that suggest PAH are exertional dyspnea, fatigue or weakness, angina, syncope, peripheral edema, and abdominal distension. PAH results from increased progressive pulmonary vascular resistance, and ultimately leads to right heart failure and death [1,2]. The currently available drugs to treat the de- velopment and progression of PAH are endothelin receptor antagonists (ERAs; bosentan, macitentan, sitaxentan, and ambrisentan), inhibitors of phosphodiesterase-5 (PDE5; sildenafil, and tadalafil), and prosta- noids (epoprostenol, treprostinil, and ilioprost) [3,4]. To date, there is no convincing evidence that drugs targeting one specific pathway are superior to the others. Hence, in clinical practice, PAH can be controlled by the co-administration of multiple drugs. Although these drugs do not cure the disease, they can retard its progression [5,6].
Bosentan hydrate is a sulfonamide-based drug, competitive dual endothelin receptor antagonist that is approved by the US Food and Drug Administration as the first oral drug for PAH with the brand name ‘Tracleer®’ [7]. Even though it can alleviate the symptoms and improve the quality of life in patients with PAH, the drug can, after its hepatic metabolism and systemic exposure, produce adverse effect [2,8–10]. Oral administration of bosentan hydrate leads to exposure-dependent liver injury, as well as a reduction in hemoglobin level, inhibition of spermatogenesis, and headache [6,11,12]. In addition, bosentan-in- duced isoenzymes inhibit the effects of co-administrated drugs, such as warfarin, cyclosporine, oral estrogens, simvastatin, and sildenafil [13–16].
Alternatively, the pulmonary delivery system in PAH can release the drugs in the vicinity of the pulmonary circulation, and avoid hepatic first-pass metabolism. The alveoli in contact with the small pulmonary arteries are the target sites of bosentan hydrate. Hence, pulmonary delivery of bosentan hydrate is expected to demonstrate high local concentration and bioavailability, since it evades the hepatic first-pass metabolism. This results in a more effective pulmonary vasodilation, than that produced after oral administration of the drug. In addition, it can improve ventilation and perfusion matching by vessels supplying the ventilated regions, thus improving gas exchange [17,18]. The conventional treatment of PAH using pulmonary drug delivery has al- ready been approved in some countries with other drugs, such as iloprost inhalation solution (Ventavis®), and treprostinil inhalation solu- tion (Tyvaso®) [10]. We believe that pulmonary delivery of bosentan hydrate can overcome the limitations of oral administration by im- proving the bioavailability of the drug, resulting in a reduction in the dosage.
In this study, bosentan hydrate was formulated and evaluated as a dry powder inhaler (DPI). The DPI is a pulmonary delivery system that utilizes the breathing process to disperse and deliver micronized solid drug particles to the lungs. It is an attractive technique, as compared to liquid-based systems, such as nebulizer, and pressurized meter dose inhaler (pMDI). DPI is portable, propellant-free, easy to operate, has a flexible-dose capacity, and because of the dry state of the formulation, has high stability [19–23]. To the best of our knowledge, no previous studies have systemically investigated the feasibility of bosentan hy- drate to deliver the drug to the lung in PAH.
For bosentan hydrate DPI to be effective in the treatment of PAH, the drug should be delivered to the peripheral region of the lungs containing respiratory bronchioles and alveoli. The lung delivery effi- ciency of DPI is typically determined by the physicochemical properties of the drug particles, including the particle size distribution, density, shape, morphology, and interparticle force [8,22,24,25]. To tailor the physicochemical properties of the particles in DPI, various microniza- tion techniques have been used, such as jet-milling, spray-drying, spray freeze-drying, supercritical fluid technology, and micro crystallization [8,22,24,26–29].
In our previous study, bosentan microparticles, which were pre- pared by either jet-milling or spray-drying processes without excipients, and exploited as DPIs, exhibited significant differences in physico- chemical properties [30]. The jet-milled bosentan microparticles showed an irregular corrugated surface, and a crystalline solid state. In contrast, the spray-dried microparticles were spherical, with a smooth surface, and an amorphous solid state. Thus due to the differences in the physicochemical properties, the in vitro lung delivery efficiency and dissolution behavior of the DPIs were considerably different. The fine particle fraction (FPF) was significantly higher in jet-milled bosentan, while the dissolution rate was considerably higher in the spray-dried bosentan. To optimize the DPI formulation, bosentan microparticles with superior lung delivery and dissolution were required.
In this study, we proposed to investigate whether the pulmonary delivery of bosentan hydrate could produce advantages over oral bosentan hydrate, and whether the physicochemical properties of the drug particles in a DPI formulation affect drug delivery efficiency and therapeutic response in PAH-induced animals. Thus, this study was performed to i) prepare the inhaled microparticulate dry powder of bosentan hydrate by using the spray-drying method, ii) evaluate the physicochemical properties of the drug particles that can be delivered to the peripheral lungs, and iii) to elucidate the pharmacokinetics and pathological characteristics of bosentan DPI in PAH-induced animals.

2. Material and methods

2.1. Materials
Bosentan hydrate (C27H31N5O7S; molecular weight (MW), 569.63 g/mol) was obtained from Hanmi Pharm Co. Ltd. (Korea). D- mannitol (MNT, C6H14O6; MW, 182.17 g/mol) was obtained from Sigma-Aldrich (MO, USA). Ethanol, acetonitrile (HPLC grade, Honeywell Burdick & Jackson®, MI, USA), and all other reagents were of analytical grade.

2.2. Spray drying of bosentan hydrate and mannitol
To prepare the bosentan microparticles, the co-spray-drying process was performed using a laboratory scale spray dryer (EYELA SD-1000, Rikakikai Co., Ltd., Japan). The feeding solution was prepared by completely dissolving bosentan hydrate and D-mannitol in 70% ethanol (v/v) [each ratio of 3:1, 1:1, and 1:3 (w/w)], to obtain a total powder concentration of 1% (w/v). The following parameters were used during spray-drying: inlet temperature, 110 °C; outlet temperature, 65–75 °C; nozzle size, 0.4 mm; feed rate, 10 mL/min; atomization air pressure, 200 kPa; and drying air flow rate, 0.30 m3/min. The designations of the spray-dried bosentan microparticles (SDBMs) were SDBM 3:1, SDBM 1:1, and SDBM 1:3, depending on the ratio of bosentan hydrate to mannitol, respectively. The bosentan microparticles were kept in a glass vial containing silica gel at −20 °C, until used.

2.3. Physicochemical characterization
The SDBMs were characterized according to their morphology, size distribution, zeta potential, true density, surface area, and water con- tent. The morphology was examined by scanning electron microscopy (ZEISS-GEMINI LEO 1530, Zeiss, Germany). The particle size distribu- tion was carried out using laser diffraction particle sizing (Mastersizer 2000, Malvern Instruments, UK) by wet dispersion method after dis- persing the samples in the non-dissolving solvent, hexane. The zeta- potential was determined using a dynamic light scattering technique (Zetasizer Nano ZS, Malvern Instruments, UK). A pycnometer (AccuPyc 1330, Micromeritics Instrument Corp., GA, USA) was used to determine the true density of the particles. The surface areas were measured using an accelerated surface area and porosimetry analyzer (ASAP 2000, Micromeritics Instrument Corp., GA, USA). It was calculated according to the Brunauer–Emmett–Teller (BET) equation. The residual water content was quantified by Karl Fischer titration (736 GP Titrino, Metrohm, Switzerland). In addition, differential scanning calorimetry (DSC 2910, TA Instruments, DE, USA), X-ray Diffractometry (XDS 2000, SCINTAG, CA, USA), and Fourier-transform infrared spectrometry (IFS- 66/S, Bruker Optics, Germany) were used to investigate the solid-state of the particles.

2.4. In vitro aerodynamic performance evaluation by Andersen cascade impactor (ACI)
In accordance with the USP Chapter 601 specification on aerosols, the aerosol performance of the SDBMs as DPI was determined using the 8-stage non-viable Andersen cascade impactor (ACI) (TE-20-800, TISCH Environmental, Inc., OH, USA) and Handihaler® (Boehringer Ingelheim, Germany) DPI device. To prevent particle bounce and re-entrainment, the collection plates of ACI stage were pre-coated with silicone oil. Raw bosentan hydrate and SDBMs equivalent to 20 mg of a formulation were loaded into a hydroXypropyl methylcellulose hard capsule (HPMC, size 3). Two capsules were aerosolized for each experiment, with air drawn through it at a controlled flow rate of 60 L/min for 4 s. The quantity of the particles remaining in the capsule and deposited onto each collec- tion plate of the stage was measured using an HPLC method. The HPLC system (Ultimate 3000 series HPLC system, Thermo Scientific, MA, USA) was operated at 220 nm with a Luna L11 250 mm × 4.60 mm, 5 μm column (Phenomenex, CA, USA). The mobile phase consisting of acetonitrile and buffer (0.1% triethylamine solution, pH 2.5) at a 45:55 (v/v) ratio was eluted at a flow rate of 1.5 mL/min. The column tem- perature was maintained at 35 °C, and the volume of each injected sample was 10 μL. The emitted dose (ED) and fine particle fraction (FPF) were calculated based on the following equations:
Emitted dose (ED)%
= Initial mass in capsule final mass remaining in capsule × 100
Initial mass in capsule
Fine particle fraction (FPF)%
Mass of the particles in stages 0 through 5

2.7. In vitro cell transport study
The cell transport of SDBM was studied using a Calu-3 cell line derived from bronchial adenocarcinoma of the airway (ATCC®). The Calu-3 cells were grown using an air–interface culture, as previously described [31]. For the air–liquid interface model, the cells were pre- cultured in DMEM, and incubated in tissue culture flasks at 37 °C in 5%
CO2, until confluency was reached, according to ATCC® recommenda- tions. Then, cells were seeded onto Transwell polyester inserts (pore size 0.4 μm, Corning) at a density of 0.5 × 106 cells/cm2. After 24 h, the apical medium was removed, and the basolateral medium was re- placed with fresh medium. This was done over 12 days, to allow for air–liquid interface monolayer differentiation.
Raw bosentan hydrate and SDBM were deposited in a Transwell containing Calu-3 using Twin stage impinger (TSI) (Copley Scientific, UK) and Handihaler®. The TSI was modified as previously described [32], to allow the attachment of the Transwell insert to the connecting tube in the lower stage. Bosentan samples equivalent to 10 mg of the drug were loaded in a hydroXypropyl methylcellulose hard capsule (HPMC, size 3), and aerosolized at a controlled flow rate at 60 L/min for 4 s. In this condition, particles reaching the Transwell in the lower stage have an aerodynamic diameter of less than 6.4 μm. After aero- solization, the Transwell, in which bosentan particles less than 6.4 μm in aerodynamic diameter were deposited, was transferred to the Franz diffusion cell. The receptor compartment of the Franz diffusion cell was The mass median aerodynamic diameter (MMAD) and the geo- metric standard deviation (GSD) were calculated using a web-based tool (www.mmadcalculator.com/andersen-impactor-mmad.html).

2.5. In vitro dissolution study
The in vitro dissolution behavior of the SDBMs was evaluated using a Franz diffusion cell system (FCDS-900C, Labfine Instruments, Korea). The receptor compartment of the Franz diffusion cell was filled with 12 mL of phosphate-buffered saline solution, adjusted to a pH of 7.4 with Tween 80 (5% w/w), which was maintained in a sink condition for the experiment. A cellulose membrane filter (pore size, 0.45 μm; ADVANTEC®, Japan) was used as a barrier, and was placed on the re- ceptor. The membrane was in contact with the receptor medium, which allowed the air–liquid interface. The receptor medium was maintained at 37 ± 1 °C, and continuously stirred to ensure homogeneity. Raw bosentan hydrate and SDBMs equivalent to 10 mg of the drug was uniformly spread on the surface of the membrane at the air–liquid in- terface. At a defined time, a volume of 200 μL of the medium was taken, and the same volume of fresh buffer was added. The content of drug was quantified using HPLC, as described above.

2.6. In vitro cell viability assay
CytotoXicity of SDBM in A549 human alveolar epithelial cells (ATCC®) was evaluated using the 3-(4,5-dimethylthiazol-2-yl)-2,5-di- phenyltetrazolium bromide (MTT) assay. The A549 cells were cultured in Dulbecco’s Modified Eagle Medium (DMEM), containing fetal bovine serum (10% v/v), streptomycin (50 μg/mL), and penicillin (50 U/mL). Cultures were incubated in tissue culture flasks at 37 °C in 5% CO2, until confluency was reached, according to ATCC® recommendations. Confluent A549 cells were plated onto 96-well plates (Corning, NY, USA) at a density of 10,000 cells/well. Each well contained 100 μL DMEM, and the plate was incubated for 24 h at 37 °C and 5% CO2. After 24 h, the A549 cells were treated with sample dispersion in water (bosentan and mannitol, ~ 200 μg/mL), and incubated for 24 h at 37 °C and 5% CO2. The absorbance of each well was measured spectro-photometrically at 570 nm by microplate reader (SpectraMax® M3, Molecular Devices, CA, USA). filled with 10 mL of Hanks’ balanced salt solution (HBSS), maintained at (37 ± 1) °C, and continuously stirred. The Transwell was in contact with the receptor medium, which allows the air–liquid interface state of Calu-3 cells. At the defined time, a volume of 200 μL of medium was taken from the receptor, and replaced with 200 μL of fresh HBSS. The content of bosentan hydrate was quantified using the HPLC method described above.

2.8. Animal model
The in vivo animal study was carried out to determine the phar- macokinetics and pathological characteristics of SDBM 1:1 after the administration of DPI, as compared to those after oral administration of the drug. Male Sprague-Dawley (SD) rats weighing 220–240 g (Samtako, Korea) were used in the experiments. They were fed on a commercial pellet diet and freshwater, and housed at room temperature of 23 ± 1 °C, with a relative humidity of 50 ± 10%, and a 12 h light and dark cycle. The animals were randomly divided into five groups: NC (saline-treated, negative control, n = 9), PC (monocrotaline- treated, positive control, n = 12), OR (monocrotaline-treated and oral raw bosentan hydrate, n = 13), IR (monocrotaline-treated and inhaled raw bosentan hydrate, n = 14), and ISD (monocrotaline-treated and inhaled of SDBM 1:1, n = 12), as shown in Scheme 1. The animals from the PC, OR, IR, and ISD groups were induced with PAH through a single intraperitoneal injection of monocrotaline (50 mg/kg, Sigma, MO, USA). The NC rats received an equal volume of saline. Eleven days after monocrotaline injection, drug samples equivalent to 5 mg/kg of the bosentan were administered daily to the OR, IR, and ISD groups. The drug was administrated to the OR group by gavage (1.5 mg/mL in 1% sodium carboXymethyl cellulose), and the IR and ISD groups were treated using a dry powder insufflator™ (DP-4, Penn-Century, Inc., Philadelphia, PA, USA) for intratracheal administration of the drug as a DPI. All rats were weighed daily, until 31 days after monocrotaline injection. The Chungbuk National University Institutional Animal Care and Use Committee approved the experimental protocols and animal care methods used in the study.

2.9. Pharmacokinetic study
The pharmacokinetics of the drug in the OR, IR, and ISD groups were studied on the first day of drug administration (after 11 days of
Scheme 1. Scheme of the in vivo studies protocol.
monocrotaline injection). The blood samples were withdrawn after administration of the drug at pre-determined time periods. Plasma was immediately separated from the blood samples by centrifuging at 1500 rpm for 10 min. All samples were stored at −20 °C, until analysis. A volume of 10 μL of internal standard (2 μg/mL of sulfaphenazole) was added to 100 μL of rat plasma, and the samples were vortexed for 30 s. Precipitation was ensured by adding 300 μL of acetonitrile, followed by vortex for 30 s. The samples were then centrifuged at 3000 rpm for 5 min at 4 °C. The supernatant was injected into the LC-MS/MS system. Liquid chromatography was performed on an Acquity UPLC (Waters Co., MA, USA) with XBridge™ C18 50 mm × 2.1 mm, 5 μm column (Waters, MA, USA). A gradient program was used with the mobile phase combining solvent A (0.1% formic acid in water) and solvent B (0.1% formic acid in acetonitrile), as follows: 5% B (0–0.5 min), 5–95% B (0.5–1.5 min), 95% B (1.5–3.0 min), 95–5% B (3.0–3.1 min), and 5% B
(3.1–7.0 min). The flow rate was 0.4 mL/min, and the injection volume was 5 μL. The column temperature was maintained at 30 °C. MS/MS analysis was performed using Xevo TQ triple quadruple mass spectro- metry (Micromass Co., UK), operated in the positive electrospray io- nization (ESI) mode. The ESI ionization was optimized to an ion spray temperature of 550 °C, and a spray voltage of 5.5 kV. Nitrogen was used as a nebulizing gas. The mass transition ion pair, performed in the multiple reaction monitoring (MRM) mode of m/z 552.0 > 202.0, was followed for bosentan, and m/z of 315.2 > 158.1 was followed for IS.
The pharmacokinetic parameters of bosentan were estimated by non- compartmental analyses using WinNonlin® software version 5.2 (Pharsight Corporation, CA, USA). The Cmax and Tmax were attained directly from the concentration–time data. The AUC0–6 and AUCinf were estimated with the linear trapezoidal rule and extrapolation to infinity, as appropriate. The t1/2 was calculated by ln2/λz, where λz is the terminal elimination rate constant.

2.10. Plethysmography
After 30 days of monocrotaline injection, the respiratory functions of all animal groups were estimated by double-chamber plethysmo- graphy (Emka Technologies, France). The double-chamber plethysmo- graph consisted of two connected pneumotachographs that measured the time shift flow between the thoracic and nasal cavity, allowing the estimation of the functional outcomes of the lungs, such as respiratory rates, durations, volumes, and flows.

2.11. Echocardiography
Echocardiography allows non-invasive pathological measurement in PAH. After 30 days of monocrotaline injection, the rats were examined by transthoracic echocardiography using a GE vivid 7 ultrasound ma- chine with a 12 MHz transducer (GE Healthcare, NJ, USA), to assess the pulmonary artery flow and ventricular dimensions and functions. The animals were anesthetized with 2% isoflurane–oXygen miXture. Pulsed wave Doppler sonography was performed in the parasternal short-axis view at the base of the heart, to measure the pulmonary artery flow. Ventricular geometries were recorded at the parasternal short-axis view at the papillary muscle level. At this level, M-mode recordings were measured to investigate the left ventricular cardiac function to de- termine the cardiac dimensions.

2.12. Ventricle weight and lung histology
At the endpoint of the animal experiment, all the animals were sa- crificed with carbon dioXide after 31 days of monocrotaline injection. The hearts were dissected, and the right and left ventricles with the ventricular septa were weighted. The lungs were dissected, and fiXed with 4% paraformaldehyde for morphological analysis of the pul- monary vessels. Paraffin-embedded sections were processed for hema- toXylin–eosin (H&E) staining for optical microscopic studies. ImageJ software (NIH) was used to estimate the external diameter and the medial wall thickness of the pulmonary arteries in 10 muscular arter- ioles (< 100 μm external diameter) per lung section, and the ratio of the medial thickness was calculated.

2.13. Statistical analyses
Statistically significant differences were evaluated using a one-way analysis of variance (ANOVA) and Tukey’s post hoc test (SPSS, version 22.0, IL, USA). A p-value of < 0.05 was considered statistically sig- nificant.

3. Results and discussion

3.1. Particle morphology
Fig. 1 shows the scanning electron micrographs of raw bosentan hydrate and co-spray-dried particles of bosentan hydrate and mannitol (SDBMs), respectively. The raw bosentan hydrate had irregular and non-spherical shape, and polydispersed size range. When we observed the SDBMs, the macroscopic images (left columns) showed that all the SDBMs had a similar shape, and were micro-size. They were non- spherical, and had raisin-shaped morphology. However, the micro- morphologies of the SDBMs (middle and right columns) were different, depending on the formulation. SDBM 3:1 had a smooth surface mor- phology with tiny projections. In contrast, SDBM 1:1 had a rough sur- face with small pores, while SDBM 1:3 had a wrinkled smooth surface.
Fig. 1. SEM micrographs of SDBMs (Magnifications for samples were 15 K, 35 K, and 350 K).

3.2. Particle size, zeta potential, true density, surface area, and water content
Table 1 shows the particle size distribution, zeta potential, true density, surface area, and water content of bosentan particles. Since all the SDBMs were prepared under similar drying conditions, the final particle size between the samples were comparable, and exhibited a similar micronized size distribution. The SDBM 3:1 showed 1.96 μm (DV10), 5.20 μm (DV50), 9.18 μm (DV90) of size and 1.388 of span
Table 1
values. The SDBM 1:1 showed 2.62 μm (DV10), 6.17 μm (DV50),
11.5 μm (DV90) of size and 1.446 of span values. The SDBM 1:3 showed 1.94 μm (DV10), 4.96 μm (DV50), 13.5 μm (DV90) of size and 2.341 of span values. The particle size distribution of SDBMs decreased relative to the raw bosentan hydrate.
The zeta potential was −41.03 ± 2.80, − 30.1 ± 1.6, −34.4 ± 1.3, and − 24.2 ± 1.4 mV for raw bosentan hydrate, SDBM 3:1, SDBM 1:1, and SDBM 1:3, respectively. The absolute values of SDBMs were significantly lower than that of the raw bosentan hydrate
Particle size distribution, zeta potential, true density, surface area and water content of SDBMs (mean ± standard deviation, n = 3). (p < 0.05, ANOVA-Tukey). SDBM 1:1 exhibited a significantly more unipolar nature than SDBM 3:1 and SDBM 1:3 (p < 0.05, ANOVA- Tukey), indicating the higher stable surface energy of SDBM 1:1 [27,28].
The values of true density of raw bosentan hydrate, SDBM 3:1, SDBM 1:1, and SDBM 1:3 were 1.32, 1.28, 1.34, and 1.41 g/cm3, re- spectively. The true density slightly increased as the mannitol ratio increased, as the true density of mannitol (1.51 g/cm3) is larger than that of bosentan hydrate (1.32 g/cm3).
Significant differences could be seen in the surface area according to the formulation. The surface areas of the raw bosentan hydrate, SDBM 3:1, SDBM 1:1, and SDBM 1:3 were 2.93 ± 0.04, 3.36 ± 0.02, 3.19 ± 0.06, and 4.03 ± 0.06 m2/g. The surface areas of SDBMs were significantly higher than that of the raw bosentan hydrate (p < 0.05, ANOVA-Tukey), in which the data were seen to be consistent with the observations obtained from the particle images of SEM (Fig. 1).
The water content of bosentan particles was analytically quantified via KF. Raw bosentan hydrate, SDBM 3:1, SDBM 1:1, and SDBM 1:3 had water content of 3.51, 1.99, 2.22, and 2.33%, respectively. The water content of the SDBMs was lower than that of raw bosentan hydrate, indicating that during the spray-drying process, water was removed. In addition, as the mannitol ratio increased in the formulations, the water content of the SDBMs was slightly increased.

3.3. Differential scanning calorimetry
Fig. 2A shows the results of differential scanning calorimetry (DSC) that were used to characterize the thermal behavior, which can be di- vided into a region representing a thermogram of bosentan (range, 110–130 °C), and a region representing a thermogram of mannitol (range, 160–175 °C). The thermogram of raw bosentan hydrate ex- hibited two endotherm peaks at 113.6 and 125.9 °C. The first peak represents the water elimination of hydrate, and the second sharp peak is the melting point of the drug, indicating the highly crystalline state of raw bosentan hydrate. However, the drug in the solid-state was severely transformed by spray drying. A definite endotherm of bosentan hydrate around 125.9 °C was not observed in the spray-dried bosentan and SDBMs, indicating the change in the solid-state of bosentan hydrate from crystalline to amorphous form as a result of spray drying. In the thermogram of mannitol, the mannitol exhibited an endotherm peak at 166.5 °C that is the melting point, indicating the highly crystalline state of mannitol. The spray-dried mannitol and SDBMs also showed an en- dothermic melting peak around 165 °C, slightly shifted from the melting peak of mannitol. This means that mannitol maintains its crystalline form, even after the spray-drying process. The SDBMs em- bedded bosentan hydrate in the amorphous state, and mannitol in the crystalline state.

3.4. Powder X-ray diffraction
Fig. 2B shows the powder X-ray diffraction (PXRD) patterns. The diffractograms of the spray-dried bosentan hydrate had no specific diffraction peaks, due to complete phase transforming to an amorphous solid-state, whereas the raw bosentan hydrate had approXimately
Fig. 2. Physicochemical characteristics and in vitro dissolution behavior of SDBMs: (A) DSC thermograms, (B) PXRD patterns, (C) FT-IR spectra, and (D) dissolution profiles in Franz diffusion cell (mean ± standard deviation, n = 4). identical diffractograms with sharp and strong diffraction peaks at the main angles (2θ) of 9.28°, 15.55°, 16.69°, and 18.64°. This indicated that the raw bosentan had high crystallinity, and that spray-drying influenced the solid-state of the drug to the amorphous state. As ex- pected in the results of DSC, spray-dried mannitol was in the crystalline state, and showed differences in intensity, as compared to the raw mannitol at the main angles (2θ) of 13.46°, 17.06°, and 31.64°. SDBMs had diffraction peaks at angles (2θ) typically at 9.47°, 40.21° and 44.65°. The peak intensity was proportional to the amount of mannitol in the formulations.

3.5. Fourier-transform infrared spectroscopy
Fig. 2C presents the results of the Fourier-transform infrared spec- troscopy (FT-IR). When compared to the raw bosentan hydrate, the FT- IR spectra of the spray-dried bosentan showed a disappearance in the peaks at 3650–3600 cm−1, which corresponded to the OeH stretching. The FT-IR spectra distinguishing mannitol was present at 3400–3300 cm−1 corresponding to the OeH absorption, and 3000–2850 cm−1 corresponding to the CeH stretching. When mannitol and spray-dried mannitol were compared, no additional or absent peaks were observed, and the peak intensity was increased as a result of spray-drying. In the SDBMs, it was found that there were no major shifts in the positions of the peaks, as compared to those of the spray-dried bosentan and spray-dried mannitol. The main intensity of the peak increased with a higher mannitol amount.
In this study, we prepared the bosentan microparticles by the co- spray-drying method with bosentan hydrate and mannitol at different weight ratios. They have significantly different physicochemical prop- erties. The results of DSC, PXRD, and FT-IR of the bosentan particles (Fig. 2A–C) show that the SDBMs consisted of amorphous bosentan hydrate and crystalline mannitol, and the crystallinity of mannitol was proportional to the mannitol ratio in the formulations. The spray drying conditions that control the evaporation rate of droplets, such as drying temperatures, and composition; concentration; and solvent types of the feeding solutions offer the opportunity of changing the crystallinity and polymorphs of bosentan and mannitol [33,34], as shown in the results of DSC, PXRD, and FT-IR (Fig. 2A–C). The differences in the Peclet number (Pe) of bosentan and mannitol also have significant effects on the physicochemical properties of each spray-dried particle formula- tions, such as surface morphology and zeta potential (Table 1 and Fig. 1) [35–38].

3.6. In vitro aerosol performance
The aerosol performance characteristics of the bosentan particles as DPI were evaluated using an Andersen cascade impactor coupled with a Handihaler® DPI device. Table 2 tabulates the aerosol performance parameters. The ED % of the SDBMs ranges 83.54–90.22%, which was significantly higher that of the raw bosentan hydrate (58.91%; p < 0.05, ANOVA-Tukey). This corresponds to a fairly lower dose loss when the SDBMs were inhaled. The FPF % indicating the ability of the particles to reach the respirable region with aerodynamic size of about 51.68 ± 6.20, and 33.05 ± 2.40% with raw bosentan hydrate, spray- dried mannitol, SDBM 3:1, SDBM 1:1, and SDBM 1:3, respectively. The significantly highest value was noted with SDBM 1:1 (p < 0.01, ANOVA-Tukey). The MMAD values of raw bosentan hydrate, spray- dried mannitol, SDBM 3:1, SDBM 1:1, and SDBM 1:3 are 3.78 ± 0.33, 4.91 ± 0.22, 3.66 ± 0.27, 1.91 ± 0.07, and 5.01 ± 0.52 μm, re- spectively. SDBM 1:1 had a significantly smallest MMAD value (p < 0.01, ANOVA-Tukey), and it also correlated with the results of FPF %. The GSD value of SDBM 1:3 was (6.63 ± 2.14), which was significantly higher than that of the others (p < 0.01, ANOVA-Tukey). All SDBMs were micro-sized, irregular shaped, and contained crystal- line mannitol. Ultimately, the factors affecting the aerosol performance of SDBMs were the surface morphology (Fig. 1), and zeta potential (Table 1). SDBM 1:1 has a more rough and indented surface mor- phology that functions to reduce the true area of the contact between the particles, as compared with SDBM 3:1 and SDBM 1:3 [39]. In ad- dition, SDBM 1:1 had significantly higher absolute values of the zeta potentials, indicating more stable surface energy than the other for- mulations, and thus would be less aggregated [40,41]. The MMAD of SDBM 1:1 ranged 1–3 μm, referring to the respirable particle size de- livering optimally to the respiratory bronchioles and alveolar region (target sites of the inhaled bosentan hydrate for the treatment of PAH) by sedimentation and diffusion with small deviations [42].

3.7. In vitro dissolution rate
Fig. 2D shows the dissolution profiles of the bosentan particles ob- tained by a Franz diffusion cell. When comparing the dissolution be- havior of raw bosentan hydrate that showed a constant drug dissolution rate during the 60 min period, the SDBMs showed a significantly higher cumulative dissolution % during 60 min (p < 0.05, ANOVA). In ad- dition, the drug dissolution rate of the SDBMs significantly increased as the mannitol ratio increased in the formulations. SDBM 3:1 and SDBM 1:1 showed a rapidly reduced dissolution rate after 30 min, while SDBM 1:3 maintained the dissolution rate, and reached a cumulative dis- solution rate of about 100% at 60 min.
The kinetics of drug dissolution was calculated according to the zero-order, first-order, HiXson-Crowell, and Higuchi model. Table 3 shows each model’s correlation coefficient (r2) and dissolution constant (k). The release profiles of all samples fit well into each kinetic model, where the dissolution constants of SDBMs increased as the mannitol ratio increased in the formulations. The presence of mannitol as a hy- drophilic agent can increase the wettability, thus enhancing the dis- solution profile of bosentan [43].
There are many physicochemical factors that can influence the in vitro dissolution behavior of DPI, such as crystallinity, polymorphism, particle size, water solubility, and dose [44,45]. Raw bosentan hydrate is extremely insoluble in water (< 0.001 mg/mL), and is highly crys- talline in the solid-state. This property reduces the affinity between the solid particles and the dissolution medium, resulting in a low dissolu- tion rate [46,47]. In the case of SDBMs, the particles consisted of amorphous bosentan hydrate, as shown in DSC, PXRD, and FT-IR (Fig. 2A–C). The amorphous bosentan hydrate that was formed by spray-drying process and the mannitol that had high water solubility (of
Aerosol performance characteristics of SDBMs including emitted dose (ED), fine particle fraction (FPF), mass median aerodynamic diameter (MMAD), and geometric standard deviation (GSD) (mean ± standard deviation, n = 4).
HiXson-Crowell 216 g/L at 25 °C) led to an increased affinity for the dissolution medium, and eventually enabled rapid dissolution of the drug [46,47]. Collectively, the results of the in vitro aerosol characteristics (Table 2) and in vitro dissolution behavior (Table 3 and Fig. 2D) showed that SDBM 1:1 had not only the significantly highest aerosol perfor- mance in SDBMs, but also a significantly increased dissolution rate, compared with that of the raw bosentan hydrate. It is expected that after inhalation, SDBM 1:1 was delivered efficiently to the peripheral lungs, and was dissolved rapidly at that site, resulting in achieving higher bioavailability. Therefore, SDBM 1:1 was selected to evaluate the in vitro cell studies and in vivo animal studies with bosentan mi- croparticles as a DPI in PAH-induced animals.

3.8. In vitro cell viability
The in vitro cell viability was evaluated for the DPI using A549 human alveolar epithelium cell line, which is widely used in inhalation toXicology [48]. A549 cell line was selected in this study, as it is re- presentative of the peripheral lung epithelium at the intended delivery site of the bosentan particle for DPI. Fig. 3A shows that the raw bo- sentan hydrate, SDBM 1:1, and mannitol appeared to be well tolerated by the A549 cells, with a cell viability of 66.27 ± 0.93, 46.17 ± 4.73, and 50.06 ± 3.21%, respectively, at 200 μg/mL concentration after 24 h, indicating a good toXicity profile. This shows the safety and fea- sibility of bosentan particles for DPI.

3.9. In vitro cell transport
The in vitro cell deposition and transport were investigated for DPI using an air–interface Calu-3 epithelial cell line that was suitable for use in drug translocation experiments. Fig. 3B shows the drug transport (apical to basal) of SDBM 1:1 and raw bosentan hydrate through the air–interface Calu-3 epithelial cell line. The initial mass of drug de- posited on the cell lines varied 416 to 1107 μg for SDBM 1:1, and 185 to 495 μg for raw bosentan hydrate. The drug transport of both samples showed steady state during the 60 min period. The fluX values were 0.2381 and 0.1164 for SDBM 1:1 and raw bosentan hydrate, respec- tively. This correlated with the results of the in vitro aerosol perfor- mance (Table 2) and in vitro dissolution rate (Table 3, Fig. 2D), but there was no significant difference (p > 0.05, ANOVA-Tukey).

3.10. Survival rate
To investigate the pharmacokinetics and pathological responses of inhaled bosentan microparticles as a DPI, we conducted an in vivo study using PAH-induced rats by single intraperitoneal injection of mono- crotaline (Scheme 1). Fig. 4A shows the Kaplan–Meier plot of survival for 31 days after monocrotaline injection. There were no deaths in the negative group (NC) rats. The survival rates in the PC, OR, IR, and ISD groups were 0.58, 0.38, 0.77, and 0.92, respectively, at the end of the study. This graph demonstrates that the intratracheal powder admin- istration of raw bosentan hydrate and SDBM 1:1 improved survival in monocrotaline-induced rats. The IR and ISD groups showed no sig- nificant difference in survival at the endpoint of the study, as compared to the NC group (p > 0.05, log-rank test). In particular, only one an- imal died in the ISD group. Survivals of the PC and OR groups were significantly different, as compared to that of the NC group, at the end of the study (p < 0.05, log–rank test). The OR group did not show an improvement in survival as compared to PC, despite oral administration of 5 mg/kg/day of the drug.

3.11. Pharmacokinetics
The pharmacokinetic parameters were studied after administration of either oral or intratracheal powder administration of bosentan in monocrotaline-induced rats. Table 4 enumerates the pharmacokinetic parameters, while Fig. 4B depicts the plasma concentration–time curve. The pharmacokinetic experiments showed typically improved bioa- vailability when administrated with bosentan by pulmonary delivery. In OR, IR, and ISD groups, the Cmax were 96.1, 165.8, and 2316.1 ng/
Fig. 3. In vitro cell studies: (A) viability of A549 cells exposure to SDBM 1:1, raw bosentan hydrate and mannitol (mean ± standard deviation, n = 3), and (B) drug transport of SDBM 1:1 and raw bosentan hydrate through air-interface Calu-3 epithelial cell line (mean ± standard deviation, n = 3).
Fig. 4. In vivo pharmacokinetic results: (A) Kaplan–Meier survival curves; * represents a significant difference (p < 0.05, Paired comparison with NC group at 31 days after monocrotaline injection, Log–Rank test), and (B) mean plasma concentration versus time curves of bosentan in PAH induced rats (mean ± standard error, n = 6) mL, and the AUC0-t were 274.1, 570.3, and 3666.0, respectively. Compared to the OR group, the AUC0-t was 2.1 times higher than that in the IR group, and 13.4 times higher than that in the ISD group. The pharmacokinetic results of the OR and IR groups indicate that just by changing the route of administration from oral to intratracheal administration, the dry powder aerosol of raw bosentan hydrate has higher bioavailability. In addition, the ISD group that was administrated with SDBM 1:1 by intratracheal dry powder showed traerosol performance (Table 2) and dissolution rate (Table 3 and Fig. 2D) of SDBM 1:1, as described in Sections 3.6 and 3.7, respectively.
Fig. 5. Respiratory functions: (A) inspiratory time (Ti), (B) expiratory time (Te), (C) peak inspiratory flow (PIF), (D) peak expiratory flow (PEF), (E) tidal volume (TV), and (F) frequency of breathing (BF). */** represent significance, compared with NC (p < 0.05/0.01, ANOVA/Tukey), and † represents significant difference, compared with PC (p < 0.05, ANOVA/Tukey) (mean ± standard error, n > 7).
Fig. 6. Flow patterns of pulmonary artery by Pulsed Wave–Doppler echocardiography from the right ventricular outflow tract: (A) representative Doppler images, (B) pulmonary artery systolic acceleration time (PS AT), and (C) pulmonary artery systolic deceleration time (PS DcT). * represents significance, compared with NC (p < 0.05, ANOVA/Tukey), and † represents significant difference, compared with PC (p < 0.05, ANOVA/Tukey) (mean ± standard error, n > 7).

3.12. Respiratory functions
Respiratory disturbance is the main symptom of PAH, and in more advanced disease is experienced almost universally [49–51]. In this study, the respiratory parameters were assessed by plethysmography in conscious animals after 30 days of monocrotaline injection, and Fig. 5 presents the data. The inspiratory time (Ti) in all the groups did not significantly differ (p > 0.05, ANOVA-Tukey). The expiratory time (Te) value in the PC group was significantly shorter than that in the NC group (p < 0.05, ANOVA-Tukey); and the Te value in the ISD group was significantly longer than that in the PC group (p < 0.05, ANOVA- Tukey). The peak inspiratory flow (PIF) and peak expiratory flow (PEF) signify the maximum inspiratory and expiratory flow rates during a single breath, respectively. These data showed that after treatment with monocrotaline, the PIF and PEF in the PC, OR, IR, and ISD groups ex- hibited significantly lower values than those in the NC group, regardless of drug treatment (p < 0.05, ANOVA-Tukey). In addition, the tidal volume (TV) signifying the total volume breathed over 1 min based on the current breathing showed significantly lower values in the PC, OR, IR, and ISD groups, as compared to that in the NC group (p < 0.05, ANOVA-Tukey). The frequency of breathing was significantly higher in the PC group than that in the NC group (p < 0.05, ANOVA-Tukey). Initially, a pattern of hypopnea was observed in the PC group, evi- denced by a reduction in the Te, PIF, PEF, and TV values, and a com- pensatory increase in the BF as compared to the NC group, illustrating the severity of PAH. Treatment with drug (OR, IR, and ISD groups) was not normalized for most of these responses, maintaining the respiratory parameters in the PC group. However, it should be emphasized that, in spite of the compulsory intratracheal aerosolization of the dry powder, the IR and ISD groups exhibited no reduced respiratory function, as compared to that in the PC and OR groups.

3.13. Cardiac remodeling
The cardiac morphology and cardiac functions in the presence of PAH can be investigated by echocardiography [52–54] that was used in this study in conscious animals after 30 days of monocrotaline injec- tion. Chronic resistance of the pulmonary vasculature causes char- acteristic structural changes in the ventricles, and decreases the cardiac output [53,55]. Pulmonary blood flow from the right ventricle was assessed by Pulsed Wave–Doppler echocardiography that provides customary measures of the PAH severity. Fig. 6A–C show the re- presentative Doppler images, pulmonary artery systolic acceleration time (PS AT), and pulmonary artery systolic deceleration time (PS DcT), respectively. Under normal conditions, pulmonary artery flow velocity increases and decreases slowly, showing a round-shaped Doppler image. In severe PAH, the Doppler signal is triangular with mid-systolic notches, due to increased right ventricular stroke work and stiffness in the pulmonary artery [52,56]. The rats in the PC group had a sig- nificantly higher value of PS DcT than that in the NC group (p < 0.05, ANOVA-Tukey). The ISD group showed a significantly lower value of PS DcT than that in the PC group (p < 0.05, ANOVA-Tukey), and there was a normalized response in the NC group (p > 0.05, ANOVA-Tukey). Fig. 7A–F depict the representative and quantitative results of the
Fig. 7. Ventricular function by echocardiography: (A) representative parasternal short axis view of the ventricles, (B) right ventricular internal diameter (RVIDd), (C) right ventricular posterior wall end diastole (RVPWd), (D) left ventricular posterior wall end diastole (LVPWd), (E) interventricular septal end diastole (IVSd), and (F) left ventricular eccentricity index. Quantitation of vessel wall thickness: (G) ratio of right ventricle (RV) weight to left ventricular plus septum (LV + S) weight. * represents significance, compared with NC (p < 0.05, ANOVA/Tukey), and † represents significant difference, compared with PC (p < 0.05, ANOVA/Tukey) (mean ± standard error, n > 7). ventricular images of the experimental animals as observed in the echocardiograph. Rats with severe PAH had a flattened interventricular septum (IVS), expanded right ventricle (RV), and decreased sphericity of the left ventricle (LV). These echocardiographic characteristics of PAH were found in the PC, OR, and IR groups, whereas ISD showed palliative pathological findings. In the results of the right ventricular internal diameter end diastole (RVIDd), PC, OR, and IR groups had significantly higher values than those of the NC group (p < 0.05, ANOVA-Tukey). The ISD group showed a significantly lower value of RVIDd than that in the PC (p < 0.05, ANOVA-Tukey), and there was a normalized response in the NC group (p > 0.05, ANOVA-Tukey). The thickness of the right ventricular posterior wall (RVPWd), left ven- tricular posterior wall (LVPWd), and interventricular septal (IVSd) were also measured. The rats in the PC, OR, and IR groups had a significantly thick RVPWd, as compared to those in the NC group (p < 0.05, ANOVA-Tukey). The LVPWd and IVSd in all the groups were not sig- nificantly different (p > 0.05, ANOVA-Tukey). The morphological characteristics of the left ventricle appeared shrunken by right ventricle hypertrophy, resulting in an increased left ventricular eccentricity index. This demonstrated a ratio of the long axis to the short axis of the left ventricle during systole. Regardless of drug treatment, monocrota- line-induced rats in the PC, OR, IR, and ISD groups exhibited
Fig. 8. Histological analysis: (A) representative microphotographs of pulmonary vessels, and (B) ratio of vessel wall thickness. */** represent significance, compared with NC (p < 0.05/0.01, ANOVA/Tukey), and †/†† represent significant difference, com- pared with PC (p < 0.05/0.01, ANOVA/Tukey) (mean ± standard error, n = 30). significantly higher left ventricular eccentricity indices than that in the NC group (p < 0.05, ANOVA-Tukey).
Further, the monocrotaline-induced right ventricular hypertrophy was assessed by obtaining Fulton’s index after sacrifice at the endpoint of the animal experiment. This is the ratio of the right ventricular (RV) weight to the left ventricular plus septal (LV + S) weight. Fig. 7G shows that monocrotaline-induced rats in the PC, OR, IR, and ISD groups had a significantly increased ratio of the RV to (LV + S) weight, thus de- monstrating RV hypertrophy. Although it was not normalized, in the IR and ISD groups that were treated with bosentan DPI, this increment was significantly suppressed (p < 0.05, ANOVA-Tukey). This was con- sistent with the results of RVPWd in echocardiography.
The cardiac remodeling results demonstrated that induction of PAH by monocrotaline had a pathological influence on the right ventricle. This was due to compensation of the right ventricular pressure overload caused by chronic pulmonary resistance. The rats in the ISD group showed a tendency to suppress cardiac remodeling, such as RVIDd, which led to an increased survival rate of the monocrotaline-induced rats, as described in Section 3.10. However, despite the high bioavail- ability of SDBM 1:1 in the ISD group, it was limited to a sufficient re- duction in the thickened ventricular wall, and left ventricle shrinkage, due to gradual hypertrophy of the right ventricle, especially for 10 days after monocrotaline injection and before drug administration.

3.14. Pulmonary vascular remodeling
The muscularization of the pulmonary vessel was investigated at the endpoint of the animal experiment. Fig. 8 shows a representative image and the wall thickness ratio of the pulmonary arteries. It was confirmed that bosentan attenuated the obliterative remodeling of the small pul- monary arteries, as induced by monocrotaline. The PC group featured a medial hyperplastic response in the small pulmonary arteries. The ratio of the wall thickness was significantly increased in the PC group, as compared to that in the NC group (p < 0.05, ANOVA-Tukey). How- ever, the bosentan-treated group (OR, IR, and ISD) had a significantly reduced ratio, as compared to the PC group (p < 0.05, ANOVA- Tukey). In particular, the wall thickness ratio of the ISD group was not significantly different from that of the NC group, signifying normal- ization of the pulmonary vascular remodeling.
Collectively, the results demonstrate that the treatment of PAH in monocrotaline-induced rats with intratracheal powder administration of bosentan showed significant superiority over traditional oral bo- sentan in a small dose (5 mg/kg/day). The results of the survival rate and pharmacokinetics indicated that the intratracheal powder admin- istration of raw bosentan hydrate showed a higher pathological effect than oral administration by just changing the administration route from oral to inhalation. In addition, SDBM 1:1 had the highest bioavail- ability, pulmonary vessel vasodilation, ventricular function, and im- proved survival in PAH. This means that the physicochemical proper- ties, such as particle size distribution, morphology, surface charge, and crystalline state, significantly influenced the aerosol performance and dissolution rate, which resulted in advanced in vivo therapeutic effects.
In this study, we developed an efficient formulation of bosentan DPI of SDBM 1:1, which showed improvement in the treatment of PAH, as described above. Oral bosentan manifests drug interactions and adverse effects associated with hepatic metabolism, and hence an alternative drug delivery system is required. Thus, we postulated that inhalation therapy of optimized drug particle would provide a better treatment effect in PAH. In comparison, the oral administration of bosentan did not show pathological responses, whereas intratracheal powder ad- ministration showed significant palliative pathological effects, despite the small dose of bosentan (5 mg/kg/day). The important findings are not only the switching of oral administration to inhalation, but also the influence of the physicochemical properties of the drug particle on the delivery efficiency, eventually determining the therapeutic effects of PAH. Thus, if additional studies can be performed in various fields, inhaled bosentan could be used to treat PAH with low dose and side effects.

4. Conclusion

In this study, we evaluated the feasibility of inhaled bosentan mi- croparticles for the treatment of PAH. Co-spray-dried bosentan micro- particles produced respirable particles, and showed a higher dissolution rate, as compared to that of the raw bosentan hydrate. In vivo evalua- tion of intratracheally administered bosentan microparticles showed promising results in terms of bioavailability, pulmonary vessel vasodi- lation, ventricular function, and survival rate in monocrotaline-induced PAH, when compared to oral bosentan hydrate. Therefore, the for- mulated inhaled bosentan microparticles can be considered as a po- tential pulmonary drug delivery system as a DPI for the treatment of PAH.

Acknowledgments
This study was supported by National Research Foundation of Korea Grant, funded by the Korean government (NRF- 2018R1A1A1A05023012, NRF-2018R1D1A1B07050538, and 2017R1A5A2015541).

References

[1] M. Humbert, O. Sitbon, G. Simonneau, Treatment of pulmonary arterial hyperten- sion, N. Engl. J. Med. 351 (2004) 1425–1436.
[2] M. Iglarz, A. Bossu, D. Wanner, C. Bortolamiol, M. Rey, P. Hess, M. Clozel, Comparison of pharmacological activity of macitentan and bosentan in preclinical models of systemic and pulmonary hypertension, Life Sci. 118 (2014) 333–339.
[3] Y. Yokoyama, M. Tomatsuri, H. Hayashi, K. Hirai, Y. Ono, Y. Yamada, K. Todoroki, T. Toyo’oka, H. Yamada, K. Itoh, Simultaneous microdetermination of bosentan, ambrisentan, sildenafil, and tadalafil in plasma using liquid chromatography/tandem mass spectrometry for pediatric patients with pulmonary arterial hy- pertension, J. Pharm. Biomed. Anal. 89 (2014) 227–232.
[4] V.V. McLaughlin, S.L. Archer, D.B. Badesch, R.J. Barst, H.W. Farber, J.R. Lindner, M.A. Mathier, M.D. McGoon, M.H. Park, R.S. Rosenson, ACCF/AHA 2009 expert consensus document on pulmonary hypertension: a report of the American College of Cardiology Foundation Task Force on expert consensus documents and the American Heart Association developed in collaboration with the American College of Chest Physicians; American Thoracic Society, Inc.; and the Pulmonary Hypertension Association, J. Am. Coll. Cardiol. 53 (2009) 1573–1619.
[5] R. Ciracì, G. Tirone, F. Scaglione, The impact of drug–drug interactions on pulmonary arterial hypertension therapy, Pulm. Pharmacol. Ther. 28 (2014) 1–8.
[6] S.M. Markova, T. De Marco, N. Bendjilali, E.A. Kobashigawa, J. Mefford, J. Sodhi, H. Le, C. Zhang, J. Halladay, A.E. Rettie, Association of CYP2C9* 2 with Bosentan- induced liver injury, Clin. Pharmacol. Ther. 94 (2013) 678–686.
[7] A.P. Davenport, K.A. Hyndman, N. Dhaun, C. Southan, D.E. Kohan, J.S. Pollock, D.M. Pollock, D.J. Webb, J.J. Maguire, Endothelin, Pharmacol. Rev. 68 (2016) 357–418.
[8] H.-J. Lee, J.-H. Kang, H.-G. Lee, D.-W. Kim, Y.-S. Rhee, J.-Y. Kim, E.-S. Park, C.W. Park, Preparation and physicochemical characterization of spray-dried and jetmilled microparticles containing bosentan hydrate for dry powder inhalation aerosols, Drug Des. Dev. Ther. 10 (2016) 4017.
[9] V. McLaughlin, Looking to the future: a new decade of pulmonary arterial hypertension therapy, Eur. Respir. Rev. 20 (2011) 262–269.
[10] B. Vaidya, V. Gupta, Novel therapeutic approaches for pulmonary arterial hypertension: unique molecular targets to site-specific drug delivery, J. Control. Release 211 (2015) 118–133.
[11] L.J. Rubin, D.B. Badesch, R.J. Barst, N. Galiè, C.M. Black, A. Keogh, T. Pulido, A. Frost, S. RouX, I. Leconte, Bosentan therapy for pulmonary arterial hypertension, N. Engl. J. Med. 346 (2002) 896–903.
[12] C.P. Denton, J.E. Pope, H.-H. Peter, A. Gabrielli, A. Boonstra, F.H. van den Hoogen, G. Riemekasten, S. De Vita, A. Morganti, M. Dölberg, Long-term effects of bosentan on quality of life, survival, safety and tolerability in pulmonary arterial hyperten- sion related to connective tissue diseases, Ann. Rheum. Dis. 67 (2008) 1222–1228.
[13] S. Lepri, L. Goracci, A. Valeri, G. Cruciani, Metabolism study and biological evaluation of bosentan derivatives, Eur. J. Med. Chem. 121 (2016) 658–670.
[14] J.A. Morales-Molina, J.E. Martínez-de la Plata, O. Urquízar-Rodríguez, M.A. Molina-Arrebola, Bosentan and oral anticoagulants in HIV patients: what we can learn of cases reported so far, Hematol. Rep. 3 (2011).
[15] K. Fattinger, C. Funk, M. Pantze, C. Weber, J. Reichen, B. Stieger, P.J. Meier, The endothelin antagonist bosentan inhibits the canalicular bile salt export pump: a potential mechanism for hepatic adverse reactions, Clin. Pharmacol. Ther. 69 (2001) 223–231.
[16] N. Galiè, A. Torbicki, R. Barst, P. Dartevelle, S. Haworth, T. Higenbottam, H. Olschewski, A. Peacock, G. Pietra, L.J. Rubin, Guidelines on diagnosis and treatment of pulmonary arterial hypertension, Eur. Heart J. 25 (2004) 2243–2278.
[17] N.S. Hill, I.R. Preston, K.E. Roberts, Inhaled therapies for pulmonary hypertension, Respir. Care 60 (2015) 794–805.
[18] T. Gessler, Inhalation of repurposed drugs to treat pulmonary hypertension, Adv. Drug Deliv. Rev. 133 (2018) 34–44.
[19] M.J. Telko, A.J. Hickey, Dry powder inhaler formulation, Respir. Care 50 (2005) 1209–1227.
[20] B. Forbes, B. Asgharian, L.A. Dailey, D. Ferguson, P. Gerde, M. Gumbleton, L. Gustavsson, C. Hardy, D. Hassall, R. Jones, Challenges in inhaled product development and opportunities for open innovation, Adv. Drug Deliv. Rev. 63 (2011) 69–87.
[21] D.A. Groneberg, H. Paul, T. Welte, Novel strategies of aerosolic pharmacotherapy, EXp. ToXicol. Pathol. 57 (2006) 49–53.
[22] C.-W. Park, X. Li, F.G. Vogt, D. Hayes, J.B. Zwischenberger, E.-S. Park, H.M. Mansour, Advanced spray-dried design, physicochemical characterization, and aerosol dispersion performance of vancomycin and clarithromycin multi- functional controlled release particles for targeted respiratory delivery as dry powder inhalation aerosols, Int. J. Pharm. 455 (2013) 374–392.
[23] H.-G. Lee, D.-W. Kim, C.-W. Park, Dry powder inhaler for pulmonary drug delivery: human respiratory system, approved products and therapeutic equivalence guide-line, J. Pharm. Investig. (2017) 1–14.
[24] A.H. Chow, H.H. Tong, P. Chattopadhyay, B.Y. Shekunov, Particle engineering for pulmonary drug delivery, Pharm. Res. 24 (2007) 411–437.
[25] J. Weers, Inhaled antimicrobial therapy–barriers to effective treatment, Adv. Drug Deliv. Rev. 85 (2015) 24–43.
[26] S.A. Shoyele, S. Cawthorne, Particle engineering techniques for inhaled bio- pharmaceuticals, Adv. Drug Deliv. Rev. 58 (2006) 1009–1029.
[27] D.M. Kabary, M.W. Helmy, E.-Z.A. Abdelfattah, J.-Y. Fang, K.A. Elkhodairy, A.O. Elzoghby, Inhalable multi-compartmental phospholipid enveloped lipid core nanocomposites for localized mTOR inhibitor/herbal combined therapy of lung carcinoma, Eur. J. Pharm. Biopharm. 130 (2018) 152–164.
[28] M.M. Abd Elwakil, M.T. Mabrouk, M.W. Helmy, E.-Z.A. Abdelfattah, S.K. Khiste, K.A. Elkhodairy, A.O. Elzoghby, Inhalable lactoferrin–chondroitin nanocomposites for combined delivery of doXorubicin and ellagic acid to lung carcinoma, Nanomedicine 13 (2018) 2015–2035.
[29] H.M. Abdelaziz, M. Gaber, M.M. Abd-Elwakil, M.T. Mabrouk, M.M. Elgohary, N.M. Kamel, D.M. Kabary, M.S. Freag, M.W. Samaha, S.M. Mortada, Inhalable particulate drug delivery systems for lung cancer therapy: Nanoparticles, micro- particles, nanocomposites and nanoaggregates, J. Control. Release 269 (2018) 374–392.
[30] H.-J. Lee, J.-H. Kang, H.-G. Lee, D.-W. Kim, Y.-S. Rhee, J.-Y. Kim, E.-S. Park, C.- W. Park, Preparation and physicochemical characterization of spray-dried and jet-milled microparticles containing bosentan hydrate for dry powder inhalation aerosols, Drug Design, Develop. Therapy 10 (2016) 4017.
[31] M. Haghi, P.M. Young, D. Traini, R. Jaiswal, J. Gong, M. Bebawy, Time-and pas- sage-dependent characteristics of a Calu-3 respiratory epithelial cell model, Drug Dev. Ind. Pharm. 36 (2010) 1207–1214.
[32] C. Grainger, L. Greenwell, G. Martin, B. Forbes, The permeability of large molecular weight solutes following particle delivery to air-interfaced cells that model the re- spiratory mucosa, Eur. J. Pharm. Biopharm. 71 (2009) 318–324.
[33] S.G. Maas, G. Schaldach, E.M. Littringer, A. Mescher, U.J. Griesser, D.E. Braun, P.E. Walzel, N.A. Urbanetz, The impact of spray drying outlet temperature on the particle morphology of mannitol, Powder Technol. 213 (2011) 27–35.
[34] A. Burger, J.O. Henck, S. Hetz, J.M. Rollinger, A.A. Weissnicht, H. Stöttner, Energy/ temperature diagram and compression behavior of the polymorphs of D-mannitol, J. Pharm. Sci. 89 (2000) 457–468.
[35] R. Vehring, W.R. Foss, D. Lechuga-Ballesteros, Particle formation in spray drying, J. Aerosol Sci. 38 (2007) 728–746.
[36] R. Vehring, Pharmaceutical particle engineering via spray drying, Pharm. Res. 25 (2008) 999–1022.
[37] J.C. Sung, B.L. Pulliam, D.A. Edwards, Nanoparticles for drug delivery to the lungs, Trends Biotechnol. 25 (2007) 563–570.
[38] C. Weiler, M. Egen, M. Trunk, P. Langguth, Force control and powder dispersibility of spray dried particles for inhalation, J. Pharm. Sci. 99 (2010) 303–316.
[39] N.Y. Chew, H.-K. Chan, Use of solid corrugated particles to enhance powder aerosol performance, Pharm. Res. 18 (2001) 1570–1577.
[40] V.A. Philip, R.C. Mehta, M.K. Mazumder, P.P. DeLuca, Effect of surface treatment on the respirable fractions of PLGA microspheres formulated for dry powder in- halers, Int. J. Pharm. 151 (1997) 165–174.
[41] I. TASK, Task group on lung dynamics: deposition and retention models for internal dosimetry of the human respiratory tract, Health Phys. 12 (1966) 173–207.
[42] J. Heyder, J. Gebhart, G. Rudolf, C.F. Schiller, W. Stahlhofen, Deposition of parti- cles in the human respiratory tract in the size range 0.005–15 μm, J. Aerosol Sci. 17 (1986) 811–825.
[43] M.M. Elgohary, M.W. Helmy, E.-Z.A. Abdelfattah, D.M. Ragab, S.M. Mortada, J.- Y. Fang, A.O. Elzoghby, Targeting sialic acid residues on lung cancer cells by in- halable boronic acid-decorated albumin nanocomposites for combined chemo/herbal therapy, J. Control. Release 285 (2018) 230–243.
[44] S.K. Ei-Arini, H. Leuenberger, Modelling of drug release from polymer matrices: effect of drug loading, Int. J. Pharm. 121 (1995) 141–148.
[45] G. Gasparini, S. Kosvintsev, M.T. Stillwell, R. Holdich, Preparation and character- ization of PLGA particles for subcutaneous controlled drug release by membrane emulsification, Colloids Surf. B: Biointerfaces 61 (2008) 199–207.
[46] S. May, B. Jensen, M. Wolkenhauer, M. Schneider, C.M. Lehr, Dissolution techniques for in vitro testing of dry powders for inhalation, Pharm. Res. 29 (2012) 2157–2166.
[47] N.M. Davies, M.R. Feddah, A novel method for assessing dissolution of aerosol in-haler products, Int. J. Pharm. 255 (2003) 175–187.
[48] J. Zavala, B. O’Brien, K. Lichtveld, K.G. Sexton, I. Rusyn, I. Jaspers, W. Vizuete, Assessment of biological responses of EpiAirway 3-D cell constructs versus A549 cells for determining toXicity of ambient air pollution, Inhal. ToXicol. 28 (2016) 251–259.
[49] L.J. Rubin, Primary pulmonary hypertension, N. Engl. J. Med. 336 (1997) 111–117.
[50] A. Low, A. Medford, A. Millar, R. Tulloh, Lung function in pulmonary hypertension, Respir. Med. 109 (2015) 1244–1249.
[51] V.L. Wittmer, É.W. Junior, P.L. Gava, F.E.L. Pereira, M.C.C. Guimarães, S.G. de Figueiredo, H. Mauad, Effects of captopril on cardiovascular reflexes and re- spiratory mechanisms in rats submitted to monocrotaline-induced pulmonary ar- terial hypertension, Pulm. Pharmacol. Ther. 30 (2015) 57–65.
[52] J.W. Koskenvuo, R. Mirsky, Y. Zhang, F.S. Angeli, S. Jahn, T.-P. Alastalo, N.B. Schiller, A.J. Boyle, K. Chatterjee, T. De Marco, A comparison of echocardio- graphy to invasive measurement in the evaluation of pulmonary arterial hy- pertension in a rat model, Int. J. Card. Imaging 26 (2010) 509–518.
[53] A. Rathinasabapathy, E. Bruce, A. Espejo, A. Horowitz, D.R. Sudhan, A. Nair, D. Guzzo, J. Francis, M.K. Raizada, V. Shenoy, Therapeutic potential of adipose stem cell-derived conditioned medium against pulmonary hypertension and lung fibrosis, Br. J. Pharmacol. 173 (2016) 2859–2879.
[54] D. Urboniene, I. Haber, Y.-H. Fang, T. Thenappan, S.L. Archer, Validation of high- resolution echocardiography and magnetic resonance imaging vs. high-fidelity ca- theterization in experimental pulmonary hypertension, Am. J. Physiol. Lung Cell. Mol. 299 (2010) L401–L412.
[55] J.M. McLendon, S.R. Joshi, J. Sparks, M. Matar, J.G. Fewell, K. Abe, M. Oka, I.F. McMurtry, W.T. Gerthoffer, Lipid nanoparticle delivery of a microRNA-145
inhibitor improves experimental pulmonary hypertension, J. Control. Release 210 (2015) 67–75.
[56] D.S. Celermajer, T. Marwick, Echocardiographic and right heart catheterization techniques in patients with pulmonary arterial hypertension, Int. J. Cardiol. 125 (2008) 294–303.