Effects of Polyvinyl Alcohol on Drug Release from Nanocomposite Particles Using Poly (L-lactide- co -glycolide)

properties nanoparticles drug delivery; size of nanoparticles inhalation. Nanocomposite particles with an aerodynamic diameter suitable for inhalation problem. these particles use diluents the alveoli lining fluid, they decomposed into nanoparticles after reaching the alveoli 10 ） . Recently, it was reported that nanocomposite particles for inhalation improved the concentration of drug in the lungs. These Abstract: The effects of polyvinyl alcohol (PVA) on the release behavior of polymer nanoparticles from nanocomposite particles using amino acids were investigated. Rifaximin (RFX) was used as a hydrophobic drug model. RFX-loaded poly(L-lactide- co -glycolide) (PLLGA) nanoparticles were prepared using an antisolvent diffusion method. They were then spray-dried with equal amounts of amino acids to prepare the nanocomposite particles. The mean diameters of nanocomposite particles were 2.86-5.42 μm. The particle size increased as the concentration of PVA aqueous solution increased. The mean diameters of RFX-loaded PLLGA nanoparticles were 150-160 nm; however, the particle size distributions of those prepared using 0.25% (w/v) PVA aqueous solution differed significantly immediately after preparation and after redispersion from nanocomposite particles. The release test results of nanocomposite particles revealed that those prepared using 0.25% and 0.50% (w/v) aqueous PVA solutions rapidly released RFX. In contrast, particles prepared using 2.00 and 4.00% (w/v) PVA aqueous solution showed sustained drug release. The results of drug release tests of nanoparticles redispersed from nanocomposite particles showed that the nanoparticles prepared using 0.50% and 2.00% (w/v) PVA aqueous solution suppressed the initial burst. Therefore, we considered that the results of the drug release behavior of the nanoparticles in these particles reflectsreflect the release behavior of the nanoparticles from the nanocomposite particles. These results indicate that the rate of redispersion from nanocomposite particles to nanoparticles can be controlled by changing the concentration of PVA aqueous solution.


Introduction
Parenteral drug administration has attracted attention in drug delivery systems. Bioavailability can be improved by avoiding the hepatic first-pass effect, which can prematurely metabolize drugs 1,2 . Inhalation systems are one of the non-invasive and self-administering methods. Inhalation formulations are a promising system for systemic administration because of the large surface area of the lungs and thin absorption barrier 3 . In topical administration for the treatment of pulmonary diseases, they can deliver drugs directly and rapidly to the diseased lung tissue site, which is advantageous over oral formulations in which only a small proportion is delivered to the diseased site 4 . To efficiently deliver the drug to the target site, it is necessary to control the size and shape of the inhalation formulation. It has been reported that particles with a mean aerody-namic diameter of 1-3 μm are deposited minimally in the mouth and throat and maximally in the lung parenchymal regions, such as the alveolar or deep-lung region 5 . Thus, inhalable nanocomposite particles are a promising inhalation system. Furthermore, nanoparticles can overcome mucus clearance 6 . In addition to macrophages, they are also taken up by other cells such as cancer cells and epithelial cells 7 10 . These properties of nanoparticles are useful for drug delivery; however, the size of nanoparticles is not suitable for inhalation. Nanocomposite particles with an aerodynamic diameter suitable for inhalation are used to solve this problem. Since these particles use diluents soluble in the alveoli lining fluid, they are decomposed into nanoparticles after reaching the alveoli 10 . Recently, it was reported that nanocomposite particles for inhalation improved the concentration of drug in the lungs. These studies with nanocomposite particles containing sildenafil 11 and rifampicin 12 -loaded nanoparticles demonstrate the usefulness of transpulmonary administration using polymer nanoparticles. Polymer nanoparticles are useful carriers because of their biocompatibility, biodegradability, ease of surface modification, localization, and reduced systemic toxicity 13 . In addition, they are capable of controlling drug release behavior. In previous studies, we investigated the physicochemical properties that affect the drug release behavior of nanoparticles prepared using poly lactide-coglycolide 14,15 . In studies using curcumin 4 , tadalafil 16 , and simvastatin 17 , nanocomposite particles containing drug nanoparticles were prepared using the spray drying method. It was reported that nanocomposite particle formation increased solubility, mainly due to the amorphization of the drug. Studies have been conducted on the behavior of drug release from nanocomposite particles, whereas there are few studies on the behavior of the release of nanoparticles from them. While sugars have conventionally been used as diluents for nanocomposite particles, amino acids have recently attracted attention for improving pulmonary delivery in high humidity 18 20 . In general, amino acids have a lower solubility in water than sugars. Hence, research focusing on their dissolution behavior is necessary.
In this study, we controlled the release behavior of poly L-lactide-co-glycolide PLLGA nanoparticles from nanocomposite particles using polyvinyl alcohol PVA . PVA is used in various pharmaceutical applications, such as stabilizing agents for emulsions 21 , artificial tears 22 , and sustained-release formulations for oral administration 23 . An evaluation of oral toxicity showed that this additive was safe even at high levels of consumption 24 . The nanoparticle surfaces are easily coated with PVA using an aqueous PVA solution 25 . We considered that the PVA-derived hydrated layer on the nanoparticles surface caused an efficient influx of water into the nanocomposite particles and aided in the redispersion of the nanoparticles. PLLGA was suitable for this study because it can prepare nanoparticles with slower drug release than those using PLGA 14,15 . Rifaximin RFX , a hydrophobic antibacterial agent, was used as a model drug 26 . Release tests were performed on the nanocomposite particles to investigate nanoparticles release behavior after reaching the lungs. Similarly, release tests were performed on nanoparticles redispersed from nanocomposite particles to investigate drug release behavior after reaching the lungs. The relationship between nanoparticles, nanocomposite particles, and nanoparticles redispersed from the nanocomposite particles is shown in Fig. 1. Fig. 1 Schematic diagram of the relationship between nanoparticles, nanocomposite particles, and nanoparticles redispersed from nanocomposite particles.

Preparation of RFX-loaded PLLGA nanoparticles
RFX-loaded PLLGA nanoparticles were prepared using an antisolvent diffusion method 25,27 . Briefly, 190 mg of PLLGA and 10 mg of RFX were dissolved in 4 mL of a mixed solution of acetone and ethanol acetone:ethanol 5:3 . The solution was injected into 40 mL of purified water, 0.25, 0.50, 2.00, and 4.00 w/v PVA aqueous solution, and then the RFX-loaded PLLGA nanoparticles were immediately precipitated. The suspension was dialyzed for 24 h in a dialysis tube C36-32-100, molecular weight cutoff: 14,000, EIDIA Co., Ltd., Tokyo, Japan to remove unloaded-RFX. Purified water was used as the external liquid and was replaced after 2, 4, 6, and 8 h. After dialysis, RFXloaded PLLGA nanoparticle suspensions were obtained. The mean volume diameters and size distributions of the nanoparticles were determined using a dynamic light scattering system ELSZ-2, Otsuka Electronics Co., Ltd., Hirakata, Japan , which measures the scattered light that is generated when the laser light is irradiated onto particles that are in Brownian motion at 25 28 .

Preparation of nanocomposite particles
The preparation of nanocomposite particles was performed similar to our previous study 19 . The nanoparticles were redispersed in purified water, and a physical mixture of arginine hydrochloride and leucine arginine hydrochloride:leucine 1:6 was added to an equal weight of precipitated nanoparticles. The suspensions were then spraydried to prepare nanocomposite particles using a spray dryer Mini Spray Dryer B-290, BÜCHI Corp., Flawil, Switzerland . Spray drying was performed under the following conditions: an outlet temperature of 37-40 , an air volume of 22.5 m 3 /h, and a pump flow rate of 1.4 mL/min. After freezing at 30 , the nanocomposite particles were lyophilized using a freeze dryer FD-1000, Tokyo Rikakikai Co., Ltd., Tokyo, Japan for 12 h. The size of the nanocomposite particles in the air was measured using a sizer LD-SA-3500A, Nikkiso Co., Ltd., Tokyo, Japan . The mean volume diameters and the size distributions of PLLGA nanoparticles redispersed from the nanocomposite particles were determined using a dynamic light scattering system ELSZ-2 at 25 . RFX content in the particles was measured using high-performance liquid chromatography HPLC, SIL-20A prominence, SPD-20A prominence, LC-20AD prominence, CTO-10ASvp, DGU-20A3 prominence, Shimadzu Co., Kyoto, Japan at 290 nm with an ODS column STR ODS-M, size: 4.6 mm 150 mm, Shinwa Chemical Industries Ltd., Kyoto, Japan . The mobile phase consisted of 52 water, 0.1 acetic acid, and 48 ethanol. Samples were filtered through a 0.45 μm pore membrane Tokyo Roshi Kaisha, Ltd., Tokyo, Japan . The samples were dissolved in 10 mL of solution. HPLC measurements were performed at ambient temperature at a flow rate of 0.9 mL/min, and 20 μL of sample solution was applied. All HPLC measurements were performed under the same conditions 26 .

2.4
In vitro release study of RFX from nanocomposite particles To evaluate the effect of PVA on nanocomposite particles drug release behavior, the release ratios of RFX from nanocomposite particles were determined. The cumulative release ratio was expressed as the ratio of RFX released from the particles divided by the amount of RFX initially contained in the nanocomposite particles 29 . The particles were added to 500 mL of phosphate-buffered saline PBS, pH 7.4 in a vessel to an RFX concentration of 1.0 μg/mL. Release tests were performed using a dissolution tester NTR-6100A, Toyama Sangyo Co., Ltd., Osaka, Japan at 100 rpm at 37 . After 1, 2, 4, 6, 8, 10, 20, 30, 40, 50, and 60 min from the test initiation, 1.0 mL of each sample was collected. The cumulative release ratios of RFX from nanocomposite particles at each time point were calculated by measuring the amount of RFX using HPLC.

2.5
In vitro release study of RFX from redispersed RFXloaded PLLGA nanoparticles To evaluate the effect of PVA on drug release behavior, the release ratios of RFX from redispersed RFX-loaded PLLGA nanoparticles were determined. The cumulative release ratio was expressed as the ratio of RFX released from nanoparticles divided by the amount of RFX initially contained in the nanoparticles 29 . Ten milligrams of nanocomposite particles were redispersed in 50 mL of PBS pH 7.4 using a vortex mixer Vortex-Genie 2, Electro Scientific Industries, Inc., Portland, OR . The suspension was diluted with PBS to an RFX concentration of 0.5 μg/mL. In total, 3 mL of the suspension was placed in a dialysis tube 30 and added to 97.0 mL of PBS pH 7.4 . The sample suspensions were shaken at 100 rpm and incubated at 37 for 0.5, 1, 2, 4, 8, 12, or 24 h. The amount of RFX remaining in the dialysis membrane was quantified using HPLC. The amount of RFX released from the nanoparticles at each time was cal-culated by subtracting the measured residual amount from the initial amount 29 .

Characterization of nanocomposite particles and
RFX-loaded PLLGA nanoparticles The mean diameters of nanocomposite particles prepared using purified water, 0.25, 0.50, 2.00, or 4.00 w/v PVA aqueous solution were 2.86 0.18, 2.89 0.10, 4.33 0.27, 4.86 0.27, and 5.42 1.12 μm, respectively. Figure 2 shows the particle size distributions of the nanocomposite particles. The particle size increased as the concentration of PVA aqueous solution increased. The aerodynamic diameter, d aer , of the particles is calculated using the following equation: where d mass , ρ, and F are the geometrical particle diameter, density of the particle, and shape function, respectively.  Figure 3 shows the particle size distributions of RFX-loaded PLLGA nanoparticles. Their mean diameters and coefficients of variation are shown in Table 1. In the nanoparticles prepared using 0.25 w/v PVA aqueous solution, the redispersed nanoparticles had a smaller mean diameter and a higher coefficient of variation than the nanoparticles immediately after preparation. PLGA nanoparticles have negative charges at neutral pH because of the ionization of the terminal carboxyl groups of PLGA 33 . This also applies to PLLGA nanoparticles, which contribute to improved dispersion stability of the particles in the solvent. In PLGA nanoparticles using PVA, it was reported that the negative charge derived from PLGA was shielded by the PVA covering the surface of the nanoparticles 25 . Therefore, we considered that the PVA-derived hydrated layer adsorbed on the surface contributes to the nanoparticles dispersion stability. The instability of the nanoparticles prepared using 0.25 w/v PVA aqueous solution suggests that both the negative charge and the hydrated layer on their surfaces were insufficient. In the other nanoparticles, no significant changes were observed between those immediately after preparation and those after redispersion from nanocomposite particles. From these findings, we confirmed that various nanocomposite particles and PLLGA nanoparticles were successfully prepared.   Figure 4 shows the cumulative release ratios of RFX from nanocomposite particles prepared using purified water, 0.25, 0.50, 2.00, and 4.00 w/v PVA aqueous solution at 37 . The release ratios of the particles prepared using purified water, 0.25, 0.50, 2.00, and 4.00 w/v PVA aqueous solution after 10 min from the initiation of the tests were 36.2 8.4 , 85.9 4.9 , 93.1 4.3, 37.1 3.1, and 40.0 1.9 , respectively. In the particles prepared using 0.25 and 0.50 w/v PVA aqueous solution, the cumulative release ratios of RFX increased rapidly. This result suggests that a large amount of RFX was released from the nanoparticles, or the RFX-loaded nanoparticles were dispersed in the solvent due to efficient amino acid dissolution. In addition, a visual observation confirmed that these nanocomposite particles collapsed faster than the other particles. In the particles prepared using 2.00 and 4.00 w/v PVA aqueous solution, 40 min after the tests initiation, the cumulative release ratios of RFX reached 85.0 9.5 and 82.8 5.8 , respectively. It is challenging to obtain molecularly dispersed PVA solutions in water 34 PVA, which could not be adsorbed on the particle surface, physically mixed with the amino acids, and prevented their dissolution. The cumulative release ratio of RFX from the particles prepared using purified water did not reach 50 even after 1 h from the tests initiation. We visually confirmed the presence of unbroken nanocomposite particles.

In vitro release of RFX from redispersed RFX-loaded
PLLGA nanoparticles The cumulative release ratios of RFX-loaded PLLGA nanoparticles prepared using purified water, 0.25, 0.50, 2.00, or 4.00 w/v PVA aqueous solution at 37 are shown in Fig. 5. For all types of particles, the release ratios reached 97 8 h after the initiation of the tests Fig. 5a . Focusing on the early stages of drug release, we confirmed that after 0.5 and 1 h from the initiation of the tests, the release ratios of the nanoparticles prepared using 0.50 and 2.00 w/v PVA aqueous solution were significantly lower than those of the others Fig. 5b . RFX is a hydrophobic drug and is almost insoluble in water 35 . These results indicate that the PVA-derived hydrated layer on their surfaces suppressed RFX release from the nanoparticles. However, the nanoparticles prepared using a 4.00 w/v PVA aqueous solution showed high release ratios similar to those prepared using purified water and a 0.25 w/v PVA aqueous solution. PVA has been reported to effectively increase the apparent solubility of poorly water-soluble drugs by functioning as a concentration-enhancing polymer 24 . Therefore, we considered that the presence of a large amount of PVA in the poor solvent increased the apparent solubility of RFX in water, and some RFX was transferred to water without being contained in the particles. The RFX transferred into the water was contained in the nanocomposite particles amino acid portion and caused high drug release ratios in the early stages of drug release.

Conclusions
In this study, we controlled the drug release behavior of nanocomposite particles using PVA. The nanocomposite particles prepared using a 0.50 w/v PVA aqueous solution succeeded in the rapid release of nanoparticles from nanocomposite particles and suppressing the initial burst. These particles can be expected to redisperse into nanoparticles after reaching the lungs rapidly. This property is similar to that of nanocomposite particles prepared using conventional sugars. Therefore, these particles could be used to treat lung diseases and the stable storage of nanoparticles. Because the lungs have less water than the release test conditions used in this study, it may be necessary to develop nanocomposite particles that more easily release nanoparticles. Nanocomposite particles prepared using 0.25-2.00 w/v PVA aqueous solution should be studied in detail. The nanocomposite particles prepared using a 2.00 w/v PVA aqueous solution succeeded in the sustained release of nanoparticles from nanocomposite particles and suppressing the initial burst. In addition to the release of the drug from the nanoparticles, these particles can be expected to be sustained-release in releasing the nanoparticles from the nanocomposite particles. The particles may reduce the frequency of drug administration in diseases that require long-term treatment, such as tuberculosis and lung cancer. This study conducted release tests on nanoparticles redispersed from nanocomposite particles to investigate the particles and drug behavior after reaching the lungs. In the future, to examine the effect of PVA on drug release behavior from nanoparticles in detail, it may be necessary to focus on nanoparticles immediately after preparation and nanoparticles from which surrounding drugs have been removed after redispersion.