Pulmonary delivery of antitubercular drugs using spray-dried lipid–polymer hybrid nanoparticles

Abstract

The present study aimed to develop lipid–polymer hybrid nanoparticles (LPNs) for the combined pulmonary delivery of isoniazid (INH) and ciprofloxacin hydrochloride (CIP HCl). Drug-loaded LPNs were prepared by the double-emulsification solvent evaporation method using the three-factor three-level Box–Behnken design. The optimized formulation had a size of 111.81 ± 1.2 nm, PDI of 0.189 ± 1.4, and PDE of 63.64 ± 2.12% for INH-loaded LPN, and a size of 172.23 ± 2.31 nm, PDI of 0.169 ± 1.23, and PDE of 68.49 ± 2.54% for CIP HCl-loaded LPN. Drug release was found to be sustained and controlled at lower pH and followed the Peppas model. The in vitro uptake study in alveolar macrophage (AM) showed that uptake of the drugs was increased significantly if administered in the form of LPN. The stability study proved the applications of adding PLGA in LPN as the polymeric core, which leads to a much more stable product as compared to other novel drug delivery systems. Spray drying was done to produce an inhalable, dry, powdered form of drug-loaded LPN. The spray-dried (SD) powder was equally capable of producing nano-aggregates having morphology, density, flowability and reconstitutibility in the range ideal for inhaled drug delivery. The nano aggregates produced by spray drying manifested their aerosolization efficiency in terms of the higher emitted dose and fine particle fraction with lower mass median aerodynamic diameter. The in vivo study using pharmacokinetic and pharmacodynamic approaches revealed that maximum internalization efficiency was achieved by delivering LPN in SD powdered forms by pulmonary route.

Keywords::

• dry powder inhaler
• lipid
• nanoparticles
• spray drying
• tuberculosis
• uptake

Introduction

Tuberculosis (TB) is an extensively spreading pandemic infecting over 1.8 billon people worldwide, causing around 1.5 million deaths annually. Despite many treatment approaches, the failure rate is still very high. Possible reasons for this are the longer duration of treatment, and other problems like patient non-compliance (Ahmad et al. 2006). Current delivery systems are not that efficient from the perspective of their intracellular targeting ability, leading to a serious lack in the therapeutic efficiency for a disease like tuberculosis, which has its infection locus inside the macrophage cells (Bhardwaj et al. 2013). Therefore, for tuberculosis treatment, research should be emphasized in such a way that maximum intracellular targeting can be achieved (Pandey and Ahmad 2011). It is highly desirable that by using available optional strategies, different formulations are developed in order to achieve higher drug concentration in the cellular tropics of infection compared to that in the blood plasma pool.

It is well known that nano-sized particles (size range 0.5–2 μm) are slowly cleared from the lungs and escape both phagocytic and mucociliary clearance. These nanoparticles also have the capacity to achieve controlled and optimal release of drugs. Several approaches have been developed to formulate nanoparticles with the desired size range and release profile, but nanoparticle formulations still suffer from some limitations like lack of control of particle size, instability of the formulation, and poor loading efficiency (Chan et al. 2009). In the recent years, a new class of nanolipid carriers—popularly known as lipid–polymer hybrid nanoparticles (LPNs)—have been introduced, which show higher potential than nanoparticles and liposomes. LPNs are polymeric nanoparticles enveloped by lipid layers, which combine the main beneficial characteristics of liposomes, like cell affinity and targeting ability, with those of polymeric nanoparticles like controlled and sustained release of drug, structural integrity, and high serum stability (Chan et al. 2009; Salvador-Morales et al. 2009). All these features have transformed LPNs into preferred carriers for the delivery of both lipophilic as well as hydrophilic drugs (Cheow and Hadinoto 2011). In contrast to oral administration, pulmonary administration of therapeutic nanoparticles leads to enhanced bioavailability of nanoparticles, resulting in prolonged time of drug residence in the lung (Yee et al. 2003; Rogueda and Traini 2007; Geller 2009; Patil and Sarasija 2012). Pulmonary administration leads to better therapeutic effect at a lower fraction of dose. For example, an inhaled dose of 100–200 μg of salbutamol is therapeutically equivalent to an oral administration of 2–4 mg (Labiris and Dolovich 2003, Chono et al. 2008; Wang et al. 2012).

Instead of using a single drug, a newer drug combination was introduced, comprising Isoniazid (INH) and Ciprofloxacin Hydrochloride (CIP HCl). INH is a well-known first-line antitubercular drug, which acts by inhibiting mycolic acid synthesis. It acts on extracellular as well as on intracellular M. tuberculosis. On the other hand, ciprofloxacin hydrochloride (CIP HCl) is a second-line antitubercular and second-generation fluoroquinolone antibacterial drug. It kills mycobacteria by interfering with the enzymes that cause DNA to rewind after being copied, which stops synthesis of DNA and of protein (Kahana and Spino 1991; Masood et al. 2010). Many studies evidence the usefulness of CIP in tuberculosis treatment, especially in drug-resistant tuberculosis. Thus, the combination of both these drugs was introduced for the first time in tuberculosis treatment.

Thus, an attempt was made to prepare a nanosystem with well-controlled size distribution, high loading efficiency, and no leakage of encapsulated material, as well as a product with better stability from the market's point of view. In order to obtain good internalization of vesicles into lungs, LPNs were developed, converted into dry powder inhaler (DPI), and compared for efficacy with plain drug in term of drug delivery into the target site.

Material and methods

Instruments

Spectral and absorbance measurements were made on a UV–visible spectrophotometer (Shimadzu, Japan, UV 1700 Pharma Spec). A digital balance (Mettler Toledo AG, Laboratory & Weighing Technologies, Switzerland, AB265-S/FACT) was used for weighing the samples. A shaking incubator (LSB- 1005RE, Daihan Labtech, Korea) was used for the drug release study. A particle size analyzer (Beckman Coulter Pvt. Ltd., Delsa NanoC) was used for particle size and PDI measurement. A probe sonicator (Vibra-Cell, Sonics, USA) was used to reduce the size of particles in nano range. Centrifugation was done using a refrigerated centrifuge (R-24C, Remi Instrument Division Maharashtra, India). Fourier transform infrared (FT-IR) spectra were measured using Nicolet-380, Thermo Nicolet, USA. Spray drying was done using a Labultima spray dryer (LU222 ADVANCED, Labultima, Mumbai, India).

Chemicals

The drugs Ciprofloxacin Hydrochloride (CIP HCl) and Isoniazid were kind gifts from Lupin Pharmaceutical Pvt. Ltd., Aurangabad, India. Soy Lecithin (LC) and 1, 2-distearoyl-sn-glycero-3-phosphoethanolamine-N-methoxy (Polyethylene-glycol)-2000] (DSPE-PEG₂₀₀₀) were generous gift from Lipoid, Germany. Poly(lactide-co-glycolide) (PLGA), Dichloromethane (DCM), Dialysis tubing, and Sephadex G-50 were purchased from Sigma Aldrich, USA. Dimethyl sulfoxide (DMSO), disodium hydrogen phosphate, D-mannitol, methanol, and potassium dihydrogen phosphate were purchased from Rankem Laboratory Reagent, New Delhi, India. Triton X-100 (TX-100) was purchased from HiMedia Laboratories Pvt. Ltd., Mumbai, India.

Method

Experimental design

In the present study, a three-factor, three-level Box–Behnken experimental design (BBD) was used to optimize formulation and process parameters for the preparation of LPN. In order to optimize the formulation components, the lipid/polymer (L/P) ratio (% w/w) (X₁), concentration of DSPE-PEG₂₀₀₀ (mM) (X₂), and sonication time (min) (X₃) were selected as independent variables. They were optimized using critical response parameters such as Y₁: particle size (PS); Y₂: polydispersity index (PDI); and Y₃: percent drug entrapment (PDE) (Solanki et al. 2007). Different variables of INH- and CIP HCl-loaded LPN and their levels in the BBD are shown in Table I.

Table I. Different variables for INH- and CIP HCl-loaded LPNs and their levels in Box Behnken Design (BBD).

Level	Low	Medium	High
Independent variables
X1 = Lipid/Polymer (L/P) (%w/w)	10%	20%	30%
X2 = DSPE-PEG (mM)	0.5	1.0	1.5
X3 = Sonication time (min)	1	2	3
Transformed values	(− 1)	(0)	(+ 1)
Dependent variables
Y1 = Particle size (PS)	Optimum
Y2 = Polydispersity Index (PDI)	Optimum
Y3 = Percent drug entrapment (PDE)	Maximum

The polynomial equation generated by the BBD is described in the equation (1):

$${\text{Y}}_{\text{i}}={\text{b}}_0+{\text{b}}_1{\text{X}}_{\text{1}}+{\text{b}}_2{\text{X}}_{\text{2}}+{\text{b}}_3{\text{X}}_{\text{3}}+{\text{b}}_{12}{\text{X}}_{\text{1}}{\text{X}}_{\text{2}}+{\text{b}}_{13}{\text{X}}_{\text{1}}{\text{X}}_{\text{3}}+{\text{b}}_{23}{\text{X}}_{\text{2}}{\text{X}}_{\text{3}}+{\text{b}}_{11}{\text{X}}_{\text{12}}+{\text{b}}_{22}{\text{X}}_{\text{22}}+{\text{b}}_{33}{\text{X}}_{\text{32}}$$

Where, Y is the measured response associated with each factor level combination (dependent variable); b₀ is an intercept; b₁–to b₃₃ are regression coefficients computed from the observed experimental values of Y from experimental runs; and A, B, and C are the coded levels of independent variables. The “Point Optimization” feature of the software was used to determine the most optimum concentration of selected variables (Rane and Prabhakar 2013; Singh et al. 2005). The resultant experimental values of the responses were quantitatively compared with those of the predicted values to validate experimental design for validation of the proposed model (Rambali et al. 2001).

Preparation of LPNs by double-emulsification solvent evaporation method

Drug-loaded LPNs of INH and CIP HCl were prepared according to the method reported by Wang and group, with suitable modification. Briefly, the oil phase consisted of 20 mg of LC, 80 mg of PLGA, and 4.0 mg of DSPE-PEG₂₀₀₀ in 5 mL of DCM, while 10 mg of the drug was dissolved in 300 μL of deionized water to form the internal aqueous phase. Next, the aqueous drug solution was emulsified in the PLGA organic solution by sonication for 1 min. The resultant nanoemulsion was poured into 12 mL of deionized water and sonicated again for 1 min at room temperature. Afterwards, the nanoemulsion was stirred overnight at room temperature to evaporate the organic phase, and the resultant nanoparticle suspension was centrifuged twice at 11,000 rpm to remove the non-encapsulated drug and excess lipid (Wang et al. 2012).

Spray drying of optimized LPNs

Instrument condition and preparation method. The LPN dispersions (INH and CIP HCl) were spray dried (SD) using a spray dryer (Labultima spray dryer, Mumbai, India) (Pourshahab et al. 2011). Mannitol was used as an excipient along with the LPN dispersion because of its good flow properties and excellent yield. The total solid concentration (i.e., LPN + excipient) in the feed was optimized at 1.0% (w/v). The flow behavior of the SD powder was evaluated using different essential parameters. The homogeneous solution was atomized using a two-fluid pressure nozzle. Liquid feed was pumped to the nozzle by a syringe pump (model M362, Orion Sage). The inlet temperature of 100°C, the feed flow rate of 1 ml/min, and the cool temperature of 65°C were optimized. The SD powder was then filled into the Eppendorf tubes and stored in desiccators until further use.

Characterization of LPNs

Optimized formulations of the LPNs were characterized and evaluated using various parameters like morphology, PS, size distribution, entrapment efficiency (EE), in vitro drug release study, in vitro cellular uptake study, and stability study.

Particle size (PS), Polydispersity Index (PDI), and Percent Drug Entrapment (PDE). The PS and PDI were analyzed using a particle size analyzer. For the determination of PDE, LPN dispersions were subjected to a high-speed centrifugation at 10,000 rpm for 30 min at 4°C. Initially, the supernatant containing free drug was collected, and then the pellets obtained containing drug-loaded LPNs were further dispersed in 5 ml of phosphate-buffered saline (PBS, pH 7.4). The supernatant and pellets were analyzed for the drug content with the UV–visible spectrophotometer using PBS (pH 7.4) as blank solution. Two methods were used to determine the EE, namely, the direct method and indirect method. Both the methods [direct (Eq. 2) and indirect (Eq. 3)] were found to be useful to determine the EE (Kumbhar and Pokharkar 2013a).

$$\text{\% EE (Direct method)}=\frac{\text{ Amount of drug in pellets }}{\text{Amount of drug in precentrifuge sample}}\times 100$$$$\text{\% EE (Indirect method) }=\frac{\text{Amount of drug in precentrifuge sample}-\text{Amount of drug in supernatant}}{\text{Amount of drug in precentrifuge sample}}\times 100$$

In vitro drug release study

The in vitro drug release study of the optimized LPN formulation was carried out by the dialysis bag technique. LPN formulation equivalent to 25 mg of drug was filled in the dialysis bag (MWCO 12–14 kDa, pore size 2.4 nm) and immersed in a receptor compartment containing 150 mL of PBS, at three different pH values: 6.8 (simulating intestinal fluid pH), 5.2 (endosomal pH of macrophages), and 7.4 (physiological pH). The system was stirred at 100 rpm and maintained at a temperature of 37 ± 0.5°C. At predetermined time intervals, 5 mL of sample was withdrawn and diluted appropriately, and absorbance was measured by UV–visible spectrophotometer (UV 1700, Shimadzu, Japan) at 269 nm for INH and 276 nm for CIP HCl (Rao and Murthy 2000). Various models (Zero-order, First-order, Higuchi model, and Korsmeyer–Peppas model) were applied to determine the mechanism of drug release from the LPN (Gao et al. 2013; Arifin et al. 2006; Zhang et al. 2010).

In vitro uptake study. Uptake characteristics of free drugs (INH and CIP HCl) and drug-loaded LPNs were studied using J774A.1 AM cell lines [National Centre for Cell Science (NCCS), Pune, India]. The cells were suspended at a concentration of 10⁶ cells/mL in RPMI 1640 medium (Sigma Chemical Co., St. Louis, MO. USA) containing 10% fetal bovine serum (Sigma Chemical Co.). Then, 1 mL aliquots of the cell suspension were transferred to 24-well culture plates (Becton Dickinson, Lincoln Park, NJ, USA), and the plates were incubated for 90 min at 37°C with 5% CO₂ (to develop J774A.1 cell monolayers). After incubation, non-adherent cells were removed and RPMI 1640 medium containing 1% fetal bovine serum was added to the monolayers. These cells were digested with 1 ml NaOH solution and analyzed for drug content using HPLC. Free drug and drug-loaded LPNs were added to the J774A.1 cell monolayers, and cells were incubated at 37°C with 5% CO₂. Free solutions of INH and CIP HCl were used as control, for comparison, and their concentration in the medium was 100 nmol for INH/ml and 100 nmol for CIP HCl/mL respectively. After incubation, the medium was removed and the cells were washed thrice with RPMI 1640 medium (Chono et al. 2008). Thus, in vitro cellular uptake of the free drug and drug-loaded LPNs were examined to demonstrate the penetration of the nanoparticles into the alveolar macrophage (AM).

Stability study. Stability studies of the optimized LPN formulations were carried out in a stability chamber maintained at 30°C/65% RH, 40°C/75% RH, and 5°C for a period of 6 months, as per ICH guidelines. Three parameters, namely, particle size, PDI, and PDE were evaluated for the stability of LPN formulations (Kumbhar and Pokharkar 2013b).

Characterization of DPI

Particle flow properties. The Angle of Repose was determined by the fixed funnel method (Muttil et al. 2007; Kaur et al. 2014a). Further, SD formulations were taken, and the bulk density was noted. The formulations were then tapped 100 times and the tapped density was noted. Further, Carr's index was calculated by using the formula C = 100[(V_b-V_t)/V_b], where C = Carr‘s index, V_b = Bulk volume, V_t = Tapped volume.

Characterization of aerodynamic properties of dry powder. An eight-stage Andersen Cascade Impactor was used to determine the fine particle fraction (FPF) of the INH- and CIP HCl-loaded LPN formulations. A dose equivalent to 10 mg of free drug was loaded directly into the chamber with the help of the respirable device, Cyclohaler (Cipla Ltd), and dispersed at an aspiration rate of 28.3 L/min for an inhalation time of 10 s. The powder deposited on each plate and throat piece was extracted and analyzed by UV spectroscopy at the wavelength of 269 nm for INH and 276 nm for Ciprofloxacin HCl against PBS (pH 7.4). The Emitted Dose (ED), Fine Particle Fraction (FPF), and Mass Median Aerodynamic Diameter (MMAD) were determined (Sinsuebpol et al. 2013).

Differential scanning calorimetry (DSC). Thermal behavior of drugs (INH, CIP HCl, and their combination), polymer (PLGA), lipid (DSPE-PEG₂₀₀₀) and SD LPNs were analyzed by DSC using Pyris software manager. The sample (approximately 2 mg) was placed in 20 μL aluminum pans and heated at the scanning rate of 20°C/min from 50 to 350°C in a nitrogen atmosphere (Kumbhar and Pokharkar 2013b).

Fourier transform infrared spectroscopy. FT-IR spectra of the pure drugs (INH and CIP HCl), polymer (PLGA), and SD LPNs were scanned between 400 and 4000 cm^{− 1} using an FT-IR spectrophotometer (Nicolet- 380, Thermo Nicolet, USA) to check the effect of interaction between different excipients, drugs, and formulation.

Scanning electron microscopy (SEM). SEM (JSM 6100 JEOL, Tokyo, Japan) was employed to study the morphology of the prepared LPNs, and SEM photographs were taken at suitable magnification (Kho and Hadinoto 2011).

Moisture content. Moisture content of the dry powders was checked by the Karl Fisher volumetric titration method. The sample (approximately 15 mg) was weighed and analyzed using Karl Fischer titrimetry for the moisture content (Oliveira et al. 2005).

Recovery of nanoparticles from spray-dried powders in aqueous medium. Approximately 50 mg of the SD powders were incubated in 3 mL of PBS (pH 7.4) for 90 min, under mild magnetic stirring at room temperature. The particle size, PDI, and PDE of re-dispersed LPNs after spray drying were compared with those of the freshly prepared LPN (Sinsuebpol et al. 2013).

In vivo studies

In vivo studies were conducted on male mice (6–8 weeks old, weighing 18–30 grams) according to protocol No. 202 (ISFCP/IAEC/CPCSEA/Meeting No. 11/2014/Protocol No. 202) approved by the Institutional Animals Ethical Committee. Groups of 3 mice per time point were used in this study. Mice were grouped into Group 1, in which oral free drug powder [(INH and CIP HCl combination) (INH and CIP HCl = 10 mg each)] was administered to them, and Group 2, in which oral drug (INH and CIP HCl combination)-loaded LPNs (powder form) were administered by the pulmonary route with the help of dry powder insufflator (DPI). The animals were bled at several time points by the retro-orbital plexus. The plasma was obtained from each mouse and deproteinized with 100 ml of acetonitrile (ACN), vortexed for 5 min, and centrifuged at 5000 g for 20 min at 4–8°C. The supernatant was used for the analysis of INH and CIP HCl. The drugs were analyzed by HPLC and compared with calibration graphs obtained by analyzing pooled blank mice plasma spiked with known amounts of drug, to obtain the profile of plasma drug concentration vs time. Peak plasma concentration (C_max) and time taken to reach C_max (T_max) were determined. Elimination half-life (t_½), Mean Residence Time (MRT), and area under curve of plasma drug concentration over time (AUC_0-∞) were calculated using a Thermo Kinetica Version 50 (Kingsley et al. 2006; Li and Huang 2008). For the tissue homogenate study, animals were sacrificed at different time points, that is, 30 min, 2 h, 8 h, 12 h, and 24 h. Drug levels were determined in 20% w/v of different tissue homogenates (lungs, liver, kidney, and spleen) by following the same analytical procedure as that described for plasma (Ahmad et al. 2006).

Results

LPNs were successfully prepared using the double-emulsion solvent evaporation method and further characterized using various parameters to achieve the best result.

Particle size (PS), poly dispersity index (PDI), and Percent Drug Entrapment (PDE)

The optimized PS and PDI of optimized INH-loaded LPNs were found to be 111.81 ± 1.2 nm and 0.189 ± 1.4 respectively, while PS of optimized CIP HCl-loaded LPN was 172.23 ± 2.31 nm and PDI was 0.169 ± 1.23. The PDE of the optimized INH-loaded LPN and CIP HCl-loaded LPN were found to be 63.64 ± 2.12% and 68.49 ± 2.54% respectively, which showed good drug-loading capacity of the developed LPN (optimization of LPN for PS, PDI, and PDE using BBD is shown as supplementary data to be found online at http://informahealthcare.com/doi/abs/10.3109/21691401.2015.1062389).

In vitro release study and release kinetics

Release studies were carried out in three different release media in order to simulate the physiological condition (pH 7.4), intestinal condition (6.8), and the macrophage environment (pH 5.2), respectively, as shown in Figures 1 and 2. At pH 7.4, about 30–38% of the drug was released immediately in 1 h from INH and CIP HCl-loaded LPN. Similarly, at pH 6.8, about 26–31% of the drug was released immediately in 1 h and at pH 5.2, about 19–25% of the drug release occurred in 1 h from both the LPN formulations. This finding indicated that some of the drug was localized on the surface of the LPN due to the partition of the drug into the surface-active agent layer adsorbed at the surface of the emulsion droplets. After this initial burst, drug release was almost constant, and around 60–63% of the drug release was observed from both the LPN formulations at pH 5.2, approximately 80–82% drug release occurred at pH 7.4, and 70–72% of drug release occurred at pH 6.8, over the duration of 12–24 h. The initial burst release obtained from all the LPN formulations could be the combined effect of release of unentrapped drugs deposited at the outer shell of polymer, and may be due to the hydrophilic nature of the lipid matrix to some extent. However, diffusion and desorption processes might play a significant role in drug release from LPN (Xie et al. 2008). It was concluded that the release of the drug from the PLGA LPNs was pH-dependent. Drug release was found to be sustained at a lower pH than at around neutral pH (pH 5.2 ˂ pH 6.8 ˂ pH 7.4). This is the consequence of the higher solubility of PLGA at neutral pH, whereas lower solubility at acidic pH. Moreover, drugs INH and CIP HCl also have good solubility at physiological pH. To investigate the mechanism of drug release from LPN, release data were modeled using BCP software. The parameters such as correlation coefficient (r²) and rate constant (k) were determined for various models, and summarized as follows: For zero order (r² = 0.6295; k₀ = 0.8725), first order (r² = 0.8855; k₁ = 0.025), Higuchi matrix (r² = 0.729), and for Korsmeyer–Peppas (r² = 0.9775; k₀ = 0.239). Based on the high correlation coefficient, the Peppas model (r² = 0.977) was selected as a best-fit model for INH-loaded LPN. Similarly, CIP HCl-loaded LPN release data were also modeled using BCP software, and parameters such as correlation coefficients (r²) and rate constants (k) were determined for various models, and summarized as follows: For zero order (r² = 0.7702; k₀ = 0.9790), first order (r² = 0.9563; k₁ = 0.028), Higuchi matrix (r² = 0.8277), and for Korsmeyer–Peppas (r² = 0.9938; k₀ = 0.2943). Based on the high correlation coefficient, the Peppas model (r² = 0.998) was selected as the best-fit model for CIP HCl-loaded LPN (Kumbhar and Pokharkar 2013a).

Figure 1. In vitro release profile of INH-loaded LPN.

Figure 2. In vitro release profile of CIP HCl-loaded LPN.

Uptake of formulations by J774A.1 cells in vitro

The uptake of INH- and CIP HCl-loaded LPN was higher as compared to the free INH and CIP HCl solution at each point. These results showed that LPN formulations were useful for enhancing the uptake efficiency of drug. From the uptake study, it was observed that maximum cellular uptake was achieved by delivering the drug in the form of LPN, as compared to the free solution form. From the Figure 3, it can be observed that the uptake efficiency of INH-loaded LPN was significantly higher as compared to that of the free INH at different time points (i.e., 1, 2, 3, and 4 h). Similar results were also observed for free CIP HCl and CIP HCl-loaded LPN.

Figure 3. Uptake characteristics of free drug and LPN formulations (conventional and ligand-anchored LPN) by J774A.1 cell lines in vitro. Values are mean ± SD, n = 3, p ˂ 0.001 with respect to free drug, according to two-way ANOVA.

Stability study

Stability is of prime importance to ensure the final performance of any colloidal system. Stability studies were performed on the INH-loaded LPN and CIP HCl-loaded LPN dispersions, for a period of 6 months, at an accelerated temperature and percent RH as per ICH guidelines. At the end of 6 months, INH-loaded LPN had sizes of 120.21 ± 1.4 nm, 135.21 ± 1.2 nm, and 119.22 ± 2.2 nm at 30°C/65% RH, 40°C/75% RH, and 5°C, respectively. INH-loaded LPN had a PDI of 0.201 ± 2.3, 0.234 ± 1.4, and 0.198 ± 1.4 at 30°C/65% RH, 40°C/75% RH, and 5°C, respectively, at the end of 6 months. On the other hand, INH-loaded LPN had PDE of 60.22 ± 2.3%, 57.13 ± 2.2%, and 62.12 ± 1.6% at 30°C/65% RH, 40°C/75% RH, and 5°C, respectively, at the end of 6 months. Similarly, at the end of 6 months, CIP HCl-loaded LPN had sizes of 184.18 ± 2.3 nm, 198.19 ± 2.1 nm, and 180.16 ± 3.1 nm at 30°C/65% RH, 40°C/75% RH, and 5°C, respectively. CIP HCl-loaded LPN had a PDI of 0.178 ± 1.5, 0.198 ± 2.4, and 0.174 ± 2.4 at 30°C/65% RH, 40°C/75% RH, and 5°C, respectively, at the end of 6 months. On the other hand, CIP HCl-loaded LPN had a PDE of 62.13 ± 2.7%, 59.12 ± 1.9%, and 65.21 ± 3.1% at 30°C/65% RH, 40°C/75% RH, and 5°C, respectively, at the end of 6 months. The nominal changes in PS, PDI, and PDE suggested the stability of INH and CIP HCl-loaded LPN formulations at all the three temperature conditions. The insignificant changes in PDE assured the potential of the designed carrier, LPN, to hold the hydrophilic drugs, INH and CIP HCl, for a sufficiently long time. At the same time, the DSPE-PEG₂₀₀₀ formed a complex with the polymer that further intensified its steric stabilization (Florence et al. 1985).

Characterization of the DPI of LPN

Flow properties of DPI. The optimized LPN formulations were spray dried to convert them into SD powders. The data observed from the Table II indicate good flow property of the powders. Solid-state characteristics like angle of repose, Carr's index, bulk and tapped density are essential parameters that could determine the delivery efficiency of DPI formulations. Angle of repose, an index of flow characteristics of the DPI, were well below 30^o, ensuring good flow. Bulk and tapped density values were found to be very low, predicting the possibilities of good aerodynamic flow behavior.

Table II. Flow properties of dry powder formulation of INH- and CIP HCl-loaded LPNs.

Formulation	Tapped density (g/ml) (mean ± SD*)	Angle of repose (mean ± SD*)	Carr's compressibility index (mean ± SD*)	Bulk density (g/cm^3)
INH-loaded LPN	0.13 ± 0.011	24.2 ± 2.13	10.1 ± 2.14	0.991 ± 1.32
CIP HCL-loaded LPN	0.14 ± 1.301	25.4 ± 1.02	7.6 ± 1.03	0.924 ± 1.31

Characterization of aerodynamic properties of the formulations

For aerodynamic properties, powders of INH-loaded LPNs and powders of CIP HCl-loaded LPNs were taken. No significant difference was observed in both the formulations. These results demonstrated that 10–12% of drug was retained in the device (Table III), indicating good dispersibility of the prepared powder. The nature of mannitol might also influence the surface characteristics of dry powder formed in spray drying, resulting in a good dispersibility of dry powder. The values for FPF% obtained in this range were considered to be satisfactory (Corrigan et al. 2006). A MMAD of less than 5 μm was prerequisite to achieve an inhalation of powders into the lower region of the lung. The MMAD values are indicative of particles behaving as individual particles upon aspiration. Thus, the improved aerodynamic properties of LPN formulations confirmed the previous reports that mannitol could improve the yield and FPF of the aerosol powder (Rabbani and Seville 2005).

Table III. Summary of aerosol deposition data of spray-dried formulations (n = 3).

DPI	Emitted Dose (%)	FPF (%)	MMAD
INH-loaded LPN	88.34 ± 1.42	64.1 ± 1.2	2.49 ± 0.12
CIP HCl-loaded LPN	87.41 ± 1.32	62.2 ± 2.3	2.41 ± 1.42

DSC

The recorded thermograms of pure drugs (INH and CIP HCl), PLGA, DSPE-PEG₂₀₀₀, INH-loaded LPN & CIP HCl-loaded LPN were observed using DSC. INH showed a melting endothermic peak at 174.46°C, whereas CIP HCl showed two peaks, that is, at 150.30°C and 318.49°C respectively. Also, the DSC thermogram of combination of INH and CIP HCl showed two melting endothermic peaks, that is, one at 135.58°C and the other at 171.30°C. Their thermograms recommended that both the drugs would not interfere with each other, suggesting their potential for combined use. The DSC thermogram of INH-loaded LPN showed a melting endothermic peak at 164.95°C, while the DSC thermogram of CIP HCl-loaded LPN had endothermic peaks at 165.33°C. However, the characteristic drug-melting endotherm of INH and CIP HCl were not seen in the DSC thermogram of their LPN formulations. This clearly indicated that the drug was molecularly dispersed within the matrix of the LPNs (figures are shown as supplementary data to be found online at http://informahealthcare.com/doi/abs/10.3109/21691401.2015.1062389).

FT-IR

The interaction between the drug and components of LPN were investigated using FT-IR spectroscopy. The IR spectra of the pure drug INH displayed the absorption peaks at 3303.46 cm^{− 1} (N–H amide), 1667.38 cm^{− 1} (C = O amide), 1556.34 cm^{− 1} (C = C aromatic), and a peak at 3112.13 cm^{− 1} (N–H stretch). The IR peaks that were noted confirmed the structure of INH. The IR spectra of PLGA presented absorption bands in the region of 3000–3600 cm^{− 1}. The IR spectra of the pure drug CIP HCl revealed the absorption peaks at 1708.97 cm^{− 1} (C = O), 1192.42 cm^{− 1} (C–N–C), 3084.71 cm^{− 1} (C–H stretch), and a peak at 1624.49 cm^{− 1} (C = C stretch), which confirmed the structure of CIP HCl. However, IR spectra of the SDLPN formulation retained most of the peaks in the fingerprint region, suggesting amorphous form of the drug. Also, in the spectra of the SD CIP HCl-loaded LPN, most of the peaks of CIP HCl, like 1708.97 cm^{− 1} (C = O), were hidden. Similarly, in the spectra of SD INH-loaded LPNs, the peaks due to N–H stretch (3112.13 cm^{− 1}) in the drug were hidden in the spectra. Also, hydrogen bonding interaction was suspected between the drug, polymer, and LPN. The FT-IR overlay of INH, PLGA, INH-loaded LPN, and CIP HCl, PLGA, and CIP HCl-loaded LPN are depicted in Figure 4a and b.

Figure 4. (a) FT-IR overlay of INH-loaded LPN. (b) FT-IR overlay of CIP HCl-loaded LPN.

SEM

The optimized formulations were visualized using scanning electron microscopy (SEM) for surface morphology. The SEM photographs revealed that the surface of LPN were smooth and spherical in shape. SEM images of INH and CIP HCl-loaded LPN are shown in Figure 5 respectively.

Figure 5. SEM images of INH-loaded LPN and CIP HCl-loaded LPN.

Moisture content

Moisture content is an important criterion for the determination of drug stability upon storage and de-aggregation upon inhalation. The moisture content (% w/w) of the delivery systems was determined by Karl Fischer Titration, shown in Table IV. The moisture content of the optimized INH-loaded LPN (1.4 ± 2.4%) and CIP HCl-loaded LPN (1.8 ± 2.7%) formulations were found to be significantly less. This suggested that the conditions optimized for the development of SD powder (inlet temperature, outlet temperature, and atomization pressure) were best suited for the removal of moisture content from the prepared powder.

Table IV. Percentage moisture content of the INH-and CIP HCl-loaded LPN.

Formulation	Karl Fischer Titrimetry (% Moisture Content)
INH-loaded LPN	1.4 ± 2.4
CIP HCl-loaded LPN	1.8 ± 2.7

Recovery of nanoparticles from spray-dried powders in aqueous medium

The SD powders of LNPs were prepared using an inlet temperature of 100°C in the aqueous medium under low stirring rate, and mannitol was dissolved, resulting in a clear solution. The effect on PS, PDI, and PDE of the LPNs recovered from SD powders and freshly prepared LPN was compared, and the comparison is shown in Table V. The size of freshly prepared LPN and LPN recovered after spray drying were apparently unchanged. Both samples showed almost similar PDI values, and indicated similar size distribution. Also, the PDE of the LPN was not altered even after processing in the form of powders. Therefore, it was hypothesized that when the SD powders reached the deep lung, the soluble component (mannitol) in the LPN was dissolved in the lung fluid and released the drug.

Table V. PS, PDI, and PDE of freshly prepared LPNs and LPN recovered from spray-dried powders (Mean ± SD, n = 3).

	Isoniazid (INH)		Ciprofloxacin Hydrochloride (CIP HCl)
Responses	Freshly prepared LPN	Spray dried powder	Freshly prepared LPN	Spray dried powder
Particle size (nm)	111.81 ± 1.2	115.43 ± 2.1	172.23 ± 2.31	176.23 ± 2.54
PDI	0.189 ± 1.4	0.191 ± 0.11	0.169 ± 1.23	0.173 ± 1.11
PDE (%)	63.64 ± 2.12	62.12 ± 1.1	68.49 ± 2.52	66.32 ± 0.12

In vivo study

In vivo studies were performed on male mice by administering the SD formulations through the pulmonary route, and drug concentration was analyzed in different organs (lung, liver, spleen, and kidney). The HPLC bioanalytical method was used for estimation of INH & CIP HCl in the LPN. A comparative study of in vivo bioavailability was done to determine a clear distinction between drug-loaded LPN versus free drug plasma concentration at the same dose, using mice (n = 3). The plasma concentration after administration of plain drug powders (combination of free INH and free CIP HCI) and the combination of SD powders of INH loaded-LPN and CIP HCI-loaded LPN using dry powder insufflators through the pulmonary route was estimated by the HPLC method. SD INH-loaded LPN (7.192 ± 0.378) and SD CIP HCl-loaded LPN (8.931 ± 0.321) were detectable even after 24 h in the plasma, as compared to free drugs (INH and CIP HCl) which were not detectable for upto 24 h (Figure 6). Thus, it provided a clear difference between the plasma concentrations of SD LPN-encapsulated drug and free drug. This study indicated that drug level could be maintained for a longer duration if given in the form of SD LPN by the pulmonary route. Various pharmacokinetic (PK) parameters were evaluated to determine any differences between free and nano-encapsulated antitubercular drugs. PK parameters obtained after non-compartmental analysis of the plasma concentration vs. time data are summarized in Table VI. Free INH and free CIP HCl reached C_max at approximately 5 h and 9 h respectively, INH- and CIP HCl-loaded LPN reached C_max at approximately 3.2 h and 3.3 h respectively. The values of AUC_0-∞ for free INH and free CIP HCl were 1399891 ± 0.34 μg/ml*min and 1123272 ± 0.23 μg/ml*min respectively. Similarly, values for AUC_0-∞ for INH- and CIP HCl-loaded LPN were 523187 ± 0.14 μg/ml*min and 458483 ± 0.72 μg/ml*min respectively. Thus, it was observed that the area under the curve (AUC) estimated after pulmonary administration of the developed formulation (LPN) was significantly lower than from free drug administration, suggesting sustained release of drug. In addition, half-life (t_1/2) of these drug-loaded formulations was significantly longer (p ≤ 0.0001) compared to that of free drug, at 1.324 ± 0.23 h (free INH) and 0.181 ± 0.32 h (free CIP HCl), respectively. SD powders of INH-loaded and CIP HCl-loaded LPN maintained controlled and sustained drug release even after 24 h, and free drugs only for 12 h, when administered using DPI through pulmonary route, which was also observed from the tissue distribution study. Thus, it provided a clear distinction between LPN drug release versus free drug.

Figure 6. Profile of plasma concentration versus time for free drug (INH and CIP HCl) and INH- and CIP HCl-loaded LPNs. Data is shown as ± SD.

Table VI. Pharmacokinetics of antitubercular drugs encapsulated in LPN as compared to free drugs, in mice (n = 3).

	Ciprofloxacin Hydrochloride		Isoniazid
Parameters	CIP-HCl Free	CIP-HCl encapsulated	INH Free	INH encapsulated
AUC0-t (ug/ml*min)	1156 ± 0.12	436 ± 0.27	151 ± 0.12	451 ± 0.54
AUC0-∞ (ng/ml*min)	1123272 ± 0.23	458483 ± 0.72	1399891 ± 0.34	523187 ± 0.14
t1/2 (hr)	0.181 ± 0.32	3.54 ± 0.27	1.324 ± 0.23	6.013 ± 0.21
Cmax (ng/ml)	56342 ± 0.82	51200 ± 0.18	38713 ± 1.21	30741 ± 0.13
Tmax (min)	540 ± 0.29	240 ± 0.26	3000 ± 1.04	180 ± 0.24
AUMC0-t (ng/ml*min^2)	16422123 ± 0.82	5368267 ± 0.25	738220 ± 0.64	6552213 ± 0.82
MRT (min)	14.2 ± 0.24	12.3 ± 0.34	4.8 ± 0.14	14.5 ± 1.01

Values are mean ± SD, n = 3. p < 0.005 with respect to free drugs, according to two-way ANOVA.

AUC_0-t: area under the concentration–time curve for time 0 to time t; AUC_0-∞: area under the concentration–time curve for time 0 to infinity (∞); t_1/2 = elimination half-life; C_max = peak plasma concentration; T_max = time taken to reach C_max; AUMC_0-t: area under the first moment curve; MRT = mean residence time.

Further, SD forms of LPNs were observed to localize in organs of interest, namely, lungs, liver, spleen, and kidney. SD powders of INH-loaded LPN and CIP HCl-loaded LPN were compared in organ homogenates with the free drug powders. The percent drug content in lungs, liver, spleen, and kidney of plain drug combinations (powder form) and their DPI form are shown in Figure 7, the graphs plotted by Bonferroni post-test showed significant p values (p ˂ 0.0001). These significant p values in the case of lungs showed that the drugs reached a high concentration in the lung and remained in the lung for a longer duration when given in the form of DPI, thus accomplishing our aim to achieve maximum internalization of the vesicles in the target site. In the case of free-drug inhalation, only 36.17 ± 3.15% of INH and 36.37 ± 3.74% of CIP HCl were found in the lungs 30 min post-administration, and were not detectable in the lungs after 24 h. The drugs were readily distributed to systemic circulation, from where they were rapidly taken up by the liver and cleared (digested) by the tissue macrophages there. However, In the case of DPI formulation, 52.23 ± 3.59% of the administered dose was found in the lungs 30 min post-administration for SD INH-loaded LPN, and 52.34 ± 3.25% of the administered dose was found in the lungs 30 min post-administration for SD CIP HCl-loaded LPN. SD forms of INH-loaded LPN and CIP HCl-loaded LPN were observed even after 24 h in the lungs, which suggested higher drug accumulation in the lungs (Table VII). Thus, an improved pharmacokinetic as well as tissue distribution profile was achieved through the use of SD LPNs. This system led to reduction in dose as well as dosing frequency, and reduced toxicity could be achieved. Also, SD INH-loaded LPN and CIP HCl-loaded LPNs maintained controlled and sustained drug release over several hours, as compared to free drugs. All these factors cumulatively led to increase in patient compliance. These observations suggested that DPI was not only effective in rapid attainment of high drug concentrations in alveolar macrophages (lungs) but could also maintain the drug concentration over a prolonged period of time when compared to the free drug. These results showed that maximum amount of drug reached in lungs when given in the form of DPI by the pulmonary route.

Figure 7. Comparative percent drug recovered from different organs and lungs after administration of various formulations. Values are mean ± SD, n = 3. p < 0.001 with respect to free drugs, according to two-way ANOVA.

Table VII. Organ Distribution of Free Drugs (INH and CIP HCl) and LPN in different organs.

		% Dose recovered after
Formulations	Organ	30 min	2 h	8 h	12 h	24 h
Free INH (PF)	Lung	36.17 ± 3.15	22.01 ± 2.72	15.26 ± 2.49	9.34 ± 1.72	0.00
	Liver	6.51 ± 2.17	13.58 ± 3.41	10.78 ± 2.33	5.31 ± 1.01	0.12 ± 0.03
	Spleen	1.58 ± 0.42	4.41 ± 0.52	1.12 ± 0.33	0.00	0.00
	Kidney	0.00	1.47 ± 0.36	2.82 ± 0.48	2.53 ± 0.17	1.31 ± 0.25
INH-Loaded LPN (SD)	Lung	52.23 ± 3.59	43.41 ± 3.67	33.72 ± 3.41	25.59 ± 3.18	7.82 ± 1.29
	Liver	4.17 ± 0.98	15.38 ± 3.72	8.59 ± 2.37	3.38 ± 1.08	0.39 ± 0.04
	Spleen	1.37 ± 0.32	2.38 ± 0.31	3.91 ± 0.93	0.00	0.00
	Kidney	0.00	1.29 ± 0.52	2.37 ± 0.79	1.61 ± 0.38	0.37 ± 0.06
Free CIP HCl (PF)	Lung	36.37 ± 3.74	21.48 ± 2.56	13.24 ± 2.73	7.93 ± 1.42	0.00
	Liver	9.25 ± 1.39	17.39 ± 3.41	11.35 ± 2.17	5.31 ± 1.23	1.29 ± 0.21
	Spleen	1.52 ± 0.52	3.94 ± 0.46	1.32 ± 0.34	0.00	0.00
	Kidney	0.00	1.45 ± 0.52	2.41 ± 0.52	1.46 ± 0.34	0.19 ± 0.11
CIP HCl loaded-LPN (SD)	Lung	52.34 ± 3.25	43.57 ± 4.38	31.56 ± 3.62	16.45 ± 3.16	6.63 ± 1.52
	Liver	3.73 ± 3.29	11.21 ± 2.58	6.38 ± 1.42	2.04 ± 0.23	0.00
	Spleen	1.41 ± 0.47	2.56 ± 0.52	3.45 ± 0.37	0.00	0.00
	Kidney	0.00	1.28 ± 0.41	3.41 ± 0.28	1.03 ± 0.46	0.24 ± 0.10

SD, Spray Dried; PF, Powder Form.

Discussion

Initially, preliminary studies were performed in order to determine the interaction between the excipients and drug, purity of the drugs (INH and CIP HCl), and selection of an appropriate lipid and polymer for the development of LPNs. The three distinct functional components used for LPN preparation were (i) lipid, soy lecithin (LC), used to construct the shell of lipoparticles to promote electrostatic interaction with oppositely charged polymers; (ii) lipid, 1,2-distearoyl-sn-glycero-3-phosphoethanolamine-N-methoxy (Polyethylene-glycol)-2000] (DSPE-PEG₂₀₀₀) PEGylated DSPE used to facilitate stealth NP formulation, to escape recognition by the reticuloendothelial system (RES) (Thevenot et al. 2007); and (iii) polymer, poly (lactide-co-glycolide) PLGA, a FDA-approved non-cytotoxic, biocompatible, and biodegradable polymer, used to form the polymer core matrix, which was wrapped by the mixed-lipid monolayer shell and provided targeting versatility with a quantitative control on the targeting effect by controlling the lipid component. Further, LPNs prepared by the double-emulsification solvent evaporation technique were successfully explored for the targeted delivery of INH and CIP HCl in lung. The Box–Behnken design (BBD) showed that concentrations of L/P, DSPE-PEG, and sonication time significantly affect the PS, PDI, and PDE. LPNs were found to be spherical in shape with two distinct boundaries representing the drug–lipid core as observed from SEM. The presence of polymer (PLGA) and lipids played significant role in modulating drug release from LPN. In vitro drug release showed that the entire drug was released in a sustained and controlled manner from the LPN at pH 5.2 (macrophage pH). It was observed that diffusion and desorption played important roles in drug release from the LPN. The in vitro uptake study in AM showed that the uptake efficiency of INH- and CIP HCl-loaded LPN was significantly higher when compared to that of the free INH and CIP HCl at different time points, that is, at 1, 2, 3, and 4 h. Stability study showed that only nominal changes occurred in the developed LPN after 6 months, ensuring the purpose of adding PLGA as a polymeric core which kept the developed LPN stable for prolonged durations and overcoming the stability-related problems associated with other novel drug delivery systems. DPI of the INH-loaded LPN and CIP HCl-loaded LPN were successfully developed and characterized in terms of their efficiency to deliver drug into the lung. Parameters like angle of repose, Carr's index, and tapped density were found to be satisfactory, suggesting the possibility of good aerodynamic properties of the SD LPN. The MMAD of less than 5 μm showed that powders remained in the lower region of lung for longer duration. Thus, SD LPN formulations encapsulating antitubercular drugs demonstrated promising properties suitable for pulmonary delivery. The pharmacokinetic (PK) study showed that LPNs maintained controlled and sustained drug release over 24 h, and free drugs only for 12 h, which was also confirmed from the biodistribution study which showed that the drug remained in lung for more than 24 h when given in the form of SD INH-loaded LPNs and SD CIP HCl-loaded LPNs, as compared to free drug powders. Further study would be conducted to compare the currently developed formulations with the existing formulations in term of their drug delivery efficacy to the lung. In addition this formulation would be tested in models of diseased animals for better results.

Conclusion

Lipid–polymer hybrid nanoparticles (LPNs) for inhalation of antitubercular drugs were successfully developed using the spray drying technique. They were found to be spherical in shape, with size less than 200 nm, as observed with the particle size analyzer. Maximum cellular uptake was observed with drug-loaded LPN, as compared to that of free drugs. The 6-month stability study showed that the optimized formulations were stable, and thus suitable for delivery as per ICH guidelines. Spray drying leads to improved aerosolization of the developed systems. The PK and biodistribution studies revealed that not only was high drug concentration achieved in the lungs using DPI of LPN, but it was maintained over a prolonged period of time, as compared to free drug. Thus, we can claim that delivering drug through the pulmonary route could be advantageous for local action in the lungs, as maximum amount of drug concentration was achieved in lung. Based on these findings, it can be concluded that drug-loaded LPNs developed in the present study may be considered as a potential drug delivery system for the delivery of antitubercular drugs for tuberculosis treatment.

Supplementary material available online

Supplementary Data to be found online at http://informahealthcare.com/doi/abs/10.3109/21691401.2015.1062389.

Acknowledgement

I want to express my sincere thanks to Punjab Technical University, Kapurthala, for allowing me to proceed with the research proposal. The authors acknowledge DBT, New Delhi for financial assistance [Project No. BT/PR5237/MED/29/641/2012]. I also express my sincere thanks to ISF College of Pharmacy, Moga (Punjab) for providing the necessary facilities.

Declaration of interest

The authors report no conflicts of interest. The authors alone are responsible for the content and writing of the paper.