Nanoincorporation of curcumin in polymer-glycerosomes and evaluation of their in vitro–in vivo suitability as pulmonary delivery systems

Abstract

The aim of this work was to deliver curcumin into the lungs by incorporating it into innovative vesicles obtained using phospholipids and high concentrations of glycerol (50%, v/v), so called glycerosomes, which were then combined with two polymers: sodium hyaluronate and trimethyl chitosan to form polymer-glycerosomes. These systems were prepared without the use of organic solvents or acidic solutions and their physico-chemical properties were fully characterized. Cryogenic transmission electron microscopy and small-angle X-ray scattering showed that both glycerosomes and polymer-glycerosomes were spherical, mainly unilamellar and of nanometric size (65–112 nm). The vesicles were readily nebulized with the largest amount of curcumin being found in the latest stages of the Next Generation Impactor™. In vitro results revealed the high biocompatibility of samples especially those containing the polymers. Curcumin loaded vesicles were also able to protect in vitro A549 cells stressed with hydrogen peroxide, restoring healthy conditions, not only by directly scavenging free radicals but also by indirectly inhibiting the production of cytokine IL6 and IL8. Moreover, in vivo results in rats showed the high capacity of these formulations to favour the curcumin accumulation in the lungs confirming their potential use as a target system for the treatment of pulmonary diseases.

---

1. Introduction

Local pulmonary disorders such as asthma, bronchitis, cystic fibrosis and chronic obstructive pulmonary disease are heterogeneous disorders generally characterized by common symptoms: inflammation of the airway, fibrosis, mucus plugging and emphysema. Infiltration of inflammatory cytokines, oxidative stress and chronic inflammation are the pathological conditions associated with the development and progression of these diseases. For this reason, treatment with drugs with antioxidant and anti-inflammatory properties may be beneficial in preventing or slowing the progression of these disorders. Several plant-derived secondary metabolites are capable of directly affecting inflammatory mediators, as well as the production and activity of second messengers, transcription factors and pro-inflammatory molecules expression, thereby, inhibiting consequent pathological conditions.^1,2 Polyphenols, such as flavonoids, are known to be potent antioxidants and anti-inflammatory agents and among all of these compounds curcumin has been shown to exhibit a wide range of pharmacological activities against many chronic diseases including type II diabetes, rheumatoid arthritis, multiple sclerosis, Alzheimer’s disease and atherosclerosis.^3–7 Further studies have demonstrated that curcumin can enhance wound healing⁸ and protect against liver and pulmonary injury or fibrosis.^9–11 Additionally, curcumin not only inhibits tumor cell proliferation and metastasis, but also induces their apoptosis by modulating several pro-inflammatory factors (e.g. interleukin IL1, IL1b, IL12), tumor necrosis factor α (TNF-α) and interferon γ (INF-γ).¹² In spite of its potential activity, curcumin bioavailability is limited by its very low aqueous solubility (0.6 mg ml⁻¹) and its susceptibility to degradation, especially under alkaline conditions.¹³ Indeed, its pulmonary delivery may represent a valid alternative; however, an appropriate vehicle is required because in water curcumin may aggregate in large particles affecting the atomization and homogeneous distribution of aerosol inhaled droplets. A possible strategy for improving the pulmonary delivery of curcumin is loading it into suitable nanocarriers which ameliorate the drug aerosol performance, local tissue deposition and distribution. Of such delivery systems, liposomes, are versatile systems able to load both lipophilic and hydrophilic compounds, and have been widely used as drug delivery systems as well as for the pulmonary delivery of active compounds because they are safe and biocompatible. Although encouraging results have been obtained when liposomes have been used as pulmonary delivery systems, their nebulization may be ameliorated by the addition of other components.^14–18 Indeed, their composition has been modified in by using polymers with the aim of improving the vesicle stability during atomization and their homogeneous distribution deep into the airway. To the best of our knowledge, in this work a new kind of phospholipid vesicle, polymer-glycerosomes, was used for the first time to prepare systems for the delivery of drugs into the lungs, in an attempt to improve the local deposition of the natural compound, curcumin, which has low bioavailabity . These systems are based on phospholipids and high concentrations of glycerol (50%, v/v), and their performances were further modified by the addition of two different hydrophilic polymers: sodium hyaluronate and trimethyl chitosan. The main components of the vesicles are phospholipids and glycerol, two harmless and fully accepted compounds for the administration of different types of pharmaceuticals. In previous works, glycerosomes have been prepared and tested as skin delivery systems of natural and synthetic compounds.^19,20 In the present work, the addition of the polymers to the vesicles made it possible for them to be prepared using an organic solvent-free and environmentally-friendly method previously proposed for the preparation of hyalurosomes.⁸

Hyaluronic acid is a natural, biocompatible and biodegradable polysaccharide that is distributed widely throughout the human body, mainly in the connective tissue, eyes, intestine and lungs where it is mainly metabolized by hyaluronidase.²¹ In addition hyaluronic acid plays an important role in the function of various inflammatory mediators and has beneficial effects on the mucociliary transport rate in airways, due to the mucoadhesivity of the polymer.^22–25 Thanks to its endogenous origin, easy degradation, elastic–viscous properties and stabilizing ability, it is often added to phospholipid vesicle formulations intended for lung delivery.^26,27 Due to its high water solubility, it is widely used in pharmaceutical formulations as a sodium salt, which was used in the present study, to prepare hyaluronan-glycerosomes (HY-glycerosomes).

Chitosan was selected because of its favourable and versatile properties that make it an attractive option for designing adequate and advanced lung delivery systems. In fact, it is a polymer which is highly biocompatible, biodegradable and mucoadhesive; moreover, it possesses antimicrobial and antioxidant properties which are potentially useful for the development of pulmonary delivery systems.²⁸ Despite its pharmaceutical suitability, chitosan has a limited aqueous solubility (only at pH < 5.6) and a lot of effort has been made to increase its water solubility especially at neutral and basic pH values, by modifying its structure.²⁹ Trimethyl chitosan is an example, and its use in pharmaceutical nanoformulations is promising.^30,31 In the present work, trimethyl chitosan chloride was synthesized and used to obtain trimethyl chitosan-glycerosomes (TMC-glycerosomes).

The purpose of the present work was to obtain new hyaluronan- or trimethyl chitosan-glycerol/phospholipid vesicles, combining the innovative composition of glycerosomes with the environmentally friendly technology previously used to prepare hyalurosomes and to test, for the first time, these polymer-glycerosomes as lung delivery systems for curcumin. In order to achieve this, (i) curcumin loaded glycerosomes, HY-glycerosomes and TMC-glycerosomes were formulated and characterized; (ii) the in vitro aerodynamic behaviour of the formulations were evaluated by means of the Next Generation Impactor™ (NGI); (iii) the in vitro biocompatibility, antioxidant and anti-inflammatory effects of curcumin loaded vesicles were assessed on pulmonary epithelial cells (A549); and (iv) the in vivo curcumin lung accumulation was studied after intratracheal administration of formulations to rats.

2. Results

2.1. Synthesis and characterization of trimethyl chitosan chloride

The water soluble trimethyl chitosan chloride was synthesised and the degree of quaternization was determined by means of ¹H-NMR spectroscopy using the procedure of water suppression. It was found to be ∼70%. The obtained polymer was soluble in water at the used concentration of 0.1% (w/v).

2.2. Preparation and characterization of vesicles

Taking into account the results obtained in a previous work,²⁰ which demonstrated that a high amount of glycerol is required to improve the characteristics of quercetin loaded glycerosomes, a preformulation study was carried out using the same phospholipid amount (180 mg ml⁻¹) and hydration by the direct addition of a water dispersion of curcumin (2 mg ml⁻¹; to obtain liposomes) or a glycerol/water mixture (30 or 50% v/v) to obtain glycerosomes. The liposome dispersion was very turbid like an emulsion, the vesicles were large (size ​not determined, ND) and highly polydispersed, and the curcumin precipitated immediately after preparation, the 30% glycerosome dispersion was turbid as well, the vesicles were smaller but the curcumin precipitated after the preparation. In contrast, curcumin (2 mg ml⁻¹) was stably incorporated in 50% glycerosomes resulting in a yellow-opalescent and highly viscous dispersion (Fig. 1A). Hence, 50% (v/v) of glycerol was selected as a suitable concentration to perform the study and this formulation was used as the phospholipid-base formulation (reference). Curcumin loaded polymer-glycerosomes were prepared by direct hydration of phospholipids with drug/polymer/glycerol/water mixtures avoiding the use of organic solvents. It is probable that a combination of glycerol and polymers facilitates the homogenous dispersion of curcumin and/or its partial solubilisation³² and the resulting polymer-glycerosomes give rise to a very fluid, yellow and transparent colloidal dispersion (Fig. 1B and C). Cryogenic transmission electron microscopy (Cryo-TEM) observation of the vesicles confirmed the presence of spherical and regular vesicles, which are mainly unilamellar, except for the TMC-glycerosomes, which partially formed oligolamellar structures (Fig. 1).

Fig. 1 Macroscopic appearance and cryo-TEM micrographs of the curcumin loaded glycerosomes (A), HY-glycerosomes (B), and TMC-glycerosomes (C).

Dynamic laser light scattering measurements (Table 1) showed that glycerosomes had the smallest mean diameter (∼62 nm, p < 0.05 versus all) in a fairly homogeneous system (P.I. = 0.28), while HY-glycerosomes and TMC-glycerosomes were larger (∼80 and 112 nm, p < 0.05 between all) with a higher polydispersity index (P.I. ≥ 0.33, p < 0.05 versus glycerosomes).

Table 1 Mean diameter (MD), polydispersity index (P.I.), zeta potential (ZP), entrapment efficiency (EE), and aggregation efficiency (AE) of curcumin (CUR) loaded glycerosomes, HY-glycerosomes and TMC-glycerosomes. Mean values ± standard deviations (SD) were obtained from at least 6 independent samples. The same symbols indicate values statistical equivalents (p > 0.05)

	Size ± SD (nm)	P.I.	ZP ± SD (mV)	EE ± SD (%)	AE ± SD (%)
Glycerosomes	°62 ± 4	*0.28	*−52 ± 2	*74 ± 4	*91 ± 3
HY-glycerosomes	*80 ± 2	°0.32	*−56 ± 5	*74 ± 2	*87 ± 4
TMC-glycerosomes	"112 ± 6	°0.34	°−43 ± 2	*73 ± 2	*93 ± 2

---

The zeta potential of the formulations was highly negative (∼−50 mV), predicting great stability of the dispersions. The surface charge of all the vesicles was negative mainly because P90G is a mixture of phosphatidylcholine, other phospholipids and fatty acids. As phosphatidylcholine is a zwitterionic molecule, the orientation and conformation of its headgroup is not constant and can be influenced by binding or adsorption of the negatively charged molecule as well sodium hyaluronate, which can interact with the choline groups on both of the bilayer surfaces to form a structured vesicle-polymer system.⁸ TMC-glycerosomes were found to have a lower negative zeta potential (−43 mV, p < 0.05 versus others) as a consequence of the partial neutralization of the phosphatidylcholine negative groups by the positive charges of trimethyl chitosan.

The curcumin concentration after preparation was checked by HPLC and the calculated amount corresponded to the amount initially used (∼2 mg ml⁻¹) in all the formulations. Both glycerosomes and polymer-glycerosomes were able to incorporate high amounts of curcumin (∼74%, p > 0.05 among groups), the great part of the used phospholipids were aggregated to form vesicles (AE% ∼ 90%, p > 0.05 among groups) and both parameters did not seem to be affected by the presence of polymers which did not hamper drug incorporation and vesicle formation. The vesicle stability was evaluated by monitoring their physico-chemical characteristics during a storage period of 90 days at room temperature (Fig. 2).

Fig. 2 Mean diameter (A), zeta potential (B) and entrapment efficiency (C) of glycerosomes, HY-glycerosomes and TMC-glycerosomes during 90 days of storage at room temperature (25 ± 1 °C). Bars represent standard deviation obtained from 3 samples.

The mean diameter of the glycerosomes (which were the smallest vesicles) increased up to ∼80 nm (p < 0.05 versus the initial value) after 30 days and then remained constant until 90 days of storage. HY-glycerosomes only increased in size (from ∼80 to ∼100 nm, p < 0.05 versus the initial value) after 60 days, whereas the size of the TMC-glycerosomes, which were the largest vesicles, remained constant (p > 0.05) during the storage period (Fig. 2A). The zeta potential of all the formulations tested remained unchanged while the entrapment efficiency decreased during the 90 days of storage, mainly for the glycerosomes and TMC-glycerosomes which diminished up to ∼55% (p < 0.05 versus the initial values and HY-glycerosome value at 90 days). The entrapment efficiency of HY-glycerosomes underwent only an initial reduction (∼10%, p < 0.05 versus start value) and then remained constant.

2.3. Small-angle X-ray scattering

The small-angle X-ray scattering (SAXS) pattern confirmed the cryo-TEM results. In fact the peak measured for the glycerosomes and HY-glycerosome was first order, broad and not sharp denoting well defined unilamellar vesicles, whereas, the TMC-glycerosome peak was asymmetrical and had a lateral hump probably due to the presence of low-lamellar vesicles. The parameters of the vesicle bilayer were calculated according to the Pabst model³³ and confirmed the effective formation of lamellar vesicles, which were liposome like and were not modified by the addition of polymer. The vesicle structure was very similar as well as the main parameters like bilayer thickness (d_B) ∼ 46 Å, polar head amplitude (σ_H) ∼ 2.8 Å, and the distance of the head group from the center of the bilayer (Z_H) ∼ 17.4 Å (Fig. 3). The SAXS patterns of the empty vesicles (data not shown) were very similar to those of the curcumin loaded vesicles.

Fig. 3 Representative SAXS diffraction profiles (A) of curcumin loaded glycerosome, HY-glycerosomes and TMC-glycerosomes and calculated bilayer values (B). d_B: bilayer thickness, Z_H: distance of head group from the centre of the bilayer, σ_H: polar head amplitude. Values represent mean ± standard deviation (n = 3).

2.4. Aerodynamic properties of samples

Formulations were sprayed using the PariSX® air jet nebulizer connected to the NGI, and the drug deposition and the aerodynamic diameter of each formulation was evaluated. Nebulization was performed for 15 min, which was appropriate to combine both patient compliance and nebulisation of a great amount of vesicular dispersion. For each formulation the nebulizer content was not completely converted into an aerosol as the Emitted Dose (ED%) is always lower than 100%: ∼43% for glycerosomes and ∼55% for polymer glycerosomes. Assessment of the aerodynamic parameters (Table 2) revealed that MMAD was always lower than 5 μm: ∼4.4 μm for glycerosomes and smaller for both polymer-glycerosomes (∼3.6 μm, p < 0.05 versus glycerosomes), proving that the droplets formed during the formation of the aerosol may reach the lungs. FPF was ∼60% for all the formulations (p > 0.05) confirming that the size of more than 59% of the aerosol droplets was below 5 μm and could potentially reach deep into the lungs.

Table 2 Fine particle dose (FPD), fine particle fraction (FPF) and mass median aerodynamic diameter (MMAD) of curcumin formulations using the NGI. All the values are shown as mean ± standard deviation of three experiments, while MMAD values are reported as mean ± geometric standard deviation. The same symbols indicate values statistically equivalent (p > 0.05)

	FPD (mg)	FPF (%)	MMAD
Glycerosomes	0.46 ± 0.07	*60 ± 2	4.5 ± 0.6
HY-glycerosomes	*0.64 ± 0.04	*61 ± 4	*3.6 ± 0.3
TMC-glycerosomes	*0.57 ± 0.09	*58 ± 5	*3.7 ± 0.5

---

The percentage of curcumin deposited in each stage of the impactor as a function of the droplet size was very similar for all the formulations (Fig. 4A) and no real impact of the polymers could be observed: ∼55% of curcumin was contained in the droplets ranging from 5.39 to 0.98 μm (stages 3–7). The mean diameter of the vesicles recovered in each stage of the impactor was measured to evaluate their stability during the process (Fig. 4B).

Fig. 4 Percentage of curcumin deposited in the different impactor stages (A) and mean diameter (B) of glycerosomes, HY-glycerosomes and TMC-glycerosomes measured before (BN) and after nebulization. Values represent mean ± standard deviation (n = 3).

The mean diameter of the glycerosomes deposited in the first three stages was ∼75 nm, which is only slightly higher than the initial size (62 nm, p < 0.05) and decreased up to ∼57 nm (p < 0.05 versus 75) in the next stages (4–8). Polymer-glycerosomes showed a similar behaviour, the size was larger, ∼89 nm in the initial stages (3) and subsequently decreased up to ∼70 nm (p < 0.05 versus 89) in the others stages. Results indicated that destabilizing forces involved during the nebulization process, caused only a small variation of vesicle size and P.I. but did not induce vesicle aggregation in large particles and the consequent drug leakage. Overall, in vitro nebulization results demonstrated a fairly good aptitude of the three formulations to be nebulized, which could be predictive of the improvement of curcumin deposition in the lung.

2.5. In vitro cytotoxicity, antioxidant and anti-inflammatory activity of curcumin-loaded vesicles, on A549 cells

Despite the well-known safety of all the used components (i.e. phospholipids, hyaluronic acid, trimetyl chitosan and curcumin), the cytotoxicity of curcumin loaded vesicles was checked, as a function of curcumin concentration, (Fig. 5A) and compared with that of curcumin dispersion. The viability of cells incubated with curcumin dispersion was the lowest (∼80%) without statistical differences (p > 0.05) among the different dilutions. When cells were treated with vesicles at the lowest dilution (curcumin 40 μg ml⁻¹) the same viability provided by dispersion (∼82%, p > 0.05) was obtained, regardless of the formulation used. Using glycerosomes at higher dilutions (curcumin 20, 10, 4 and 2 μg ml⁻¹) the viability slight increased (∼89%, p > 0.05) being higher, but not statistically different, than that obtained using the dispersion (p > 0.05). Incubation with HY-glycerosomes (except at the lowest dilution, 40 μg ml⁻¹ of curcumin) improved the cell viability with respect to curcumin dispersion (always ≥100%), in particular at the highest three dilutions (∼113%, p < 0.05). Incubation with TMC-glycerosomes also increased the cell viability with respect to curcumin dispersion but to a lesser extent (∼99%, p < 0.05) except at the lowest dilution. These results confirmed that incorporation of curcumin into vesicles, especially into polymer-glycerosomes, ameliorates drug biocompatibility especially when a concentration of 20 μg ml⁻¹ of curcumin was used. This concentration was used for further study.

Fig. 5 In vitro A549 cell bioavailability after 48 h of incubation with curcumin in dispersion or loaded in glycerosomes, HY-glycerosomes and TMC-glycerosomes at different dilutions (2, 4, 10, 20, 40 μg ml⁻¹ of curcumin) (A). Protective effect of curcumin dispersion and vesicles against H₂O₂-induced oxidative stress in A549 cells at 6 h (B). Inhibition percentage of IL-6 and IL-8 on A549 cells treated with curcumin loaded vesicles (C). Data are reported as mean values ± standard deviation. The same symbols indicate values statistical equivalents (p > 0.05). All symbols indicate statistical differences with respect to curcumin dispersion. In panel B, symbol ° indicates statistical differences with respect to curcumin dispersion and H₂O₂ and * indicates statistical differences with respect to H₂O₂.

The short-term (6 h) protective effect of curcumin formulations, against the oxidative stress induced in vitro on A549 cells by H₂O₂ was also evaluated (Fig. 5B). The H₂O₂ treatment reduced the cell viability up to 60% (positive control) and all the curcumin formulations exerted a restoring effect (p < 0.05 versus positive control) reducing the cell mortality. The protective effect of curcumin was slightly evident using its water dispersion and glycerosomes (∼80% viability, p < 0.05 versus positive control) and was especially reinforced by its incorporation into HY- and TMC-glycerosomes which totally counteracted the damaging effect of H₂O₂ and additionally allowed cell proliferation up to ∼120% of viability. HY- and TMC-glycerosomes seem to facilitate cellular uptake of curcumin which may exert its scavenging activity in the intracellular environment, thus, preventing cell damage and consequent death.

The ability of curcumin loaded vesicles to down regulate the production of two pro-inflammatory cytokines IL-6 and IL-8 was evaluated by means of an in vitro long-term study (4 days) using A549 cells. Cytokine expression was measured by FACS analysis (Fig. 5C). All of the formulations significantly inhibited the production of both IL6 and IL8 in comparison with untreated cells (control, p < 0.05). The formulation which was the most effective at inhibiting the secretion of both cytokines, comprised of curcumin loaded TMC-glycerosomes, the efficacy of this formulation reached ∼52%, while the efficacy of HY-glycerosomes was lower than ∼22%, p < 0.01.

2.6. In vivo biodistribution studies

In vivo biodistribution studies were performed by intratracheal administration of curcumin dispersion and curcumin loaded vesicles to rats, then measuring the plasma concentration of the drug and its retention in the trachea, lungs and liver. During the 24 h experiment, curcumin was not detected in the plasma or its concentration was below the limit of quantification, confirming its low systemic distribution and its reduced accumulation in non-targeted tissues. All of the formulations improved the total drug deposition in the whole lungs at 24 h with respect to the curcumin dispersion (used as control, p < 0.05) following the ranking: HY-glycerosomes > TMC-glycerosomes > glycerosomes > control (Fig. 6A). HY-glycerosomes gave the same drug deposition in the trachea and in each lobe of the lungs (∼0.75 μg g⁻¹, p > 0.05 among groups), except in the right inferior lobe in which the amount was slightly higher (∼0.85 μg g⁻¹, p < 0.05). TMC-glycerosomes provided an homogeneous distribution of curcumin (∼0.68 μg g⁻¹) in the trachea and the three lung lobes, the amount of curcumin was slightly lower only in the residual lung (∼0.45 μg g⁻¹, p < 0.05). Using glycerosomes, the amount of drug accumulated in each lobe of the lungs and trachea was homogeneous (∼0.37 μg g⁻¹) but the values were not statistically different to those found using the control (p > 0.05). For all of the tested formulations the amount of drug found in the liver was ≤0.004 μg g⁻¹ confirming the negligible accumulation in non-targeted tissues. As confirmation of previous findings fluorescent curcumin accumulation in the lung was visualized using the In Vivo FX PRO Imaging System (Fig. 6B).

Fig. 6 Curcumin deposited in the different parts of the respiratory tree (A) and curcumin fluorescence distribution in lungs and liver excised from rats 24 h after the intratracheal administration of the different formulations (B). Data are reported as mean values ± standard deviation. The same symbols indicate values of statistical equivalents (p > 0.05). All the symbols indicate statistical differences with respect to curcumin dispersion.

Following treatment with all the vesicles, a higher fluorescence than that of the control was detected in the lungs. In particular, lungs treated with HY-glycerosomes showed the highest fluorescence. The fluorescence intensity of lungs treated with TMC-glycerosomes was slightly lower while those treated with glycerosomes showed the weakest fluorescence. In the liver a slight fluorescence was detected using glycerosomes, showing their lower selectivity for lungs with respect to polymer-glycerosomes, which on the contrary, provided a high affinity for pulmonary tissues and any liver or systemic distribution was detected, confirming their potential use as a targeting system for the lungs.

3. Discussion

Considering the promising results obtained when glycerosomes have been used for drug delivery into skin, here their innovative composition was combined with the environmentally-friendly technology of hyalurosomes to obtain polymer-glycerosomes.^8,19,34,35 These new vesicles were tested for the first time as carriers for the pulmonary delivery of curcumin and their performances were compared with that of conventional glycerosomes which were used as the reference as it is not possible to obtain the corresponding conventional liposomes. In the preformulation study, a mixture of glycerol/water 50% v/v was selected as the most suitable to obtain curcumin loaded glycerosomes. The polymer was initially dispersed with curcumin in the glycerol/water mixture facilitating the curcumin stabilization in the polymeric network and probably its partial solubilisation as previously reported for Eudragit/curcumin dispersion.³² The obtained homogenous and transparent dispersions were used as hydration media for phospholipids, facilitating the assembly and stabilization of HY-glycerosomes.

The partial substitution of amino hydrogen with methyl groups converted chitosan to the water soluble trimethyl chitosan which enabled TMC-glycerosomes to be prepared using the same biocompatible procedure applied in the preparation of HY-glycerosomes, thereby, avoiding the use of the acidic solution usually needed to solubilise chitosan.^8,36 The safety of a drug delivery system is particular crucial for pulmonary formulations. For this reason, natural and well-known safe components were used to prepare the tested vesicles (i.e. phospholipids, sodium hyaluronate and trimethyl chitosan) and their biocompatibility with human lung epithelial cells was confirmed, especially that of HY-glycerosomes, which at concentrations of curcumin ranging from 20 to 2 μg ml⁻¹ improved cell viability up to ∼108% probably due to the proliferative effect of hyaluronan.³⁷ By virtue of its natural origin and polyfunctional properties, curcumin appears to be an ideal agent to treat inflammatory and oxidative stress conditions in the lungs.³⁸ Indeed, it can directly avoid the oxidative stress scavenging superoxide, hydroxyl radical and nitric oxide free radicals³⁹ and indirectly down regulate the expression of multifunctional proinflammatory cytokines (i.e. IL-6 and IL-8) actively involved in endogenous oxidative stress, inflammation and carcinogenesis.⁴⁰ In this work, these helpful curcumin properties were confirmed and its antioxidant activity in vitro was improved by its inclusion in polymer-glycerosomes. In particular, the endogenous production of oxidative and inflammatory mediators was greatly inhibited by TMC-glycerosomes. Additionally, the incorporation of the active into glycerosomes and polymer-glycerosomes provided a good curcumin nebulization in vitro overcoming its limited water solubility, which makes it difficult to achieve therapeutic concentrations in lungs. In fact, the jet nebulization method involves repeated cycles of aerosol droplet formation and their recapture into the nebulizer reservoir before the formulation leaves the device. Previous findings have disclosed that the considerable shearing forces applied during nebulization, facilitate solvent elimination and the formation of large drug aggregates which are unable to reach the last part of the respiratory tract or to be eliminated from the device. The combination of the nanotechnologies, the innovative preparation method and the employment of glycerosomes seems to overcome these drawbacks permitting curcumin to stay inside the vesicles, these can reach the last stages of the impactor intact probably due to their high resistance to the shearing forces.⁴¹ This vesicle resistance seemed to be independent from their composition probably because the structure and size of the vesicles were very similar, simply TMC-glycerosomes were oligolamellar and consequently slightly larger than glycerosomes and HY-glycerosomes which were unilamellar and smaller. The deposition into the lung epithelia of curcumin still loaded in vesicles and not as the free drug represents an important advantage due to the well-known ability of phospholipid vesicles to facilitate drug passage through biological membranes and overcoming the barrier formed by the lung epithelial cells that prevents the absorption of inhaled substances.⁴²

Despite the comparable physico-chemical properties of the vesicles and the similarity in the formation of their aerosols in vitro , the in vivo biodistribution study confirmed the superior ability of both HY- and TMC-glycerosomes compared to glycerosomes as lung delivery systems and demonstrated that the addition of both polymers to glycerosomes improves the in vivo accumulation of curcumin. A possible explanation for this could be that the polymer acted as a key component forming less viscous dispersions and favouring the vesicle adhesiveness and deposition in the lung tissue. Additionally, these carriers were able to maximize the in vitro curcumin protective effect against the oxidative stress on lung cells and provided an optimal inhibition of proinflammatory cytokine production indeed it is expected to exert the same activity in vivo after their deposition in the lungs.

4. Conclusions

During this study, new polymer-glycerosome formulations were tailored using the hyalurosome method for their preparation to deliver curcumin into lungs. Then, HY- and TMC-glycerosomes were obtained by means of a one-step method using only phospholipids, glycerol, polymers, curcumin and water. Overall, the in vitro and in vivo outcomes highlighted that the simultaneous addition of glycerol and hydrophilic polymer to phospholipids as the hydrating medium does not simply form a coating layer on the vesicle surface but a continuous polymer–glycerol network which avoids changes in the lamellar assembly and allows small but substantial modifications to the dispersion architecture, resulting in stable and in vivo performing phospholipid vesicles. Their safe composition, easy, biocompatible, environmentally-friendly preparation and the high delivery properties make them potential carriers to improve the local pulmonary activity of curcumin or other compounds that exhibit poor water solubility and low bioavailability. Actually, polymer-glycerosomes thanks to the ease with which they can be turned into aerosols, their selective deposition in lung tissue and their simultaneous ability to improve the antioxidant and anti-inflammatory activity of curcumin represent promising lung delivery systems which can improve patient compliance and prevent hepatic first-pass metabolism of the drug. Additionally, HY-glycerosomes will be tested as specific delivery system into tissues, such as tumors, which over express the hyaluronic acid cell surface receptor, CD44.

5. Materials and methods

5.1. Materials

Phospholipon®90G (P90G), a commercial mixture, containing the soy phosphatidylcholine, phosphatidylethanolamine, fatty acids and triglycerides, was kindly supplied by AVG S.r.l. (Garbagnate Milanese, Italy). Sodium hyaluronate low molecular weight (200–400 kDa), was purchased from DSM Nutritional Products AG Branch Pentapharm (Aesch/Switzerland). Curcumin, chitosan (low molecular weight 50–190 kDa and 75–85% deacetylated), sodium iodide, methyl iodide, N-methylpyrrolidinone, acetone, glycerol and all the other products, were of analytical grade and were purchased from Sigma-Aldrich (Milan, Italy). All the cell culture reagents were purchased from Life Technologies Europe (Monza, Italy).

5.2. Preparation and characterization of trimethyl chitosan chloride

The synthesis of trimethyl chitosan reported by Wang et al.⁴³ was slightly modified. Chitosan, sodium iodide and N-methyl pyrrolidinone were vigorously stirred at 60 °C for 20 minutes. NaOH solution (11 ml, 15% w/w) and methyl iodide (12 ml) were added, the mixture was maintained for 60 minutes at 60 °C with vigorous stirring, and then stored in a refrigerator to avoid the evaporation of CH₃I. To control the degree of quaternization of chitosan, methyl iodide (5 ml) and NaOH (10 ml, 15% w/w) were added again. The mixture was kept at 60 °C for 6 hours and then at room temperature overnight with stirring. The mixture was concentrated, purified by dialysis, treated with a NaCl solution (10% w/w) at 25 °C overnight, to exchange the I-counterions with Cl-counterions, and then freeze-dried. ¹H-NMR spectroscopy was performed using a Varian INOVA-500. The ¹H-NMR spectra were recorded at 27 °C, using deuterium oxide as the solvent. All the measurements were done with water suppression. The trimethyl chitosan degree of quaternization (DQ%) was calculated using eqn (1):⁴⁴where [N(CH₃)³] is the integral of the trimethyl amino group and H¹ is the integral of the proton on the C¹ of the glycoside ring.

5.3. Vesicle preparation

To obtain glycerosomes, curcumin (2 mg ml⁻¹) was dispersed in a mixture of glycerol/water (1/1 v/v) and to obtain HY-glycerosomes and TMC-glycerosomes, the same amount of drug was dispersed in a mixture of glycerol/water (1/1 v/v) containing 0.1% (w/v) of sodium hyaluronate or trimethyl chitosan. The dispersions were stirred for 5 h and used to hydrate phospholipids (180 mg ml⁻¹) overnight.^19,35 Finally, the dispersions were sonicated using a high intensity ultrasonic disintegrator (Soniprep 150, MSECrowley, London, United Kingdom), for 25, 10, 15 and 25 cycles (2 seconds on and 2 seconds off, 15 μm of probe amplitude) waiting 2 min between each cycle group to promote the cooling of the sample. A schematic illustration of glycerosomes and polymer-glycerosomes is presented in Fig. 7.

Fig. 7 Schematic illustration of the structure of glycerosomes and polymer-glycerosomes.

Samples (2 ml) were purified from the non-incorporated drug by dialysis against glycerol/water (2 l) using dialysis tubing (Spectra/Por® membranes: 12–14 kDa MW cut-off, 3 nm pore size; Spectrum Laboratories Inc., DG Breda, The Netherlands) at room temperature for 4 hours, by replacing the water every hour. The water (8 l in total) was enough to theoretically remove all the drug contained in 2 ml of dispersion (4 mg).

5.4. Vesicle characterization

The morphology of each formulation was checked by cryo-TEM (CCiT, University of Barcelona, Spain). A thin aqueous film was formed on a glow-discharged holey carbon grid. The films were vitrified by plunging the grid (kept at 100% humidity and room temperature) into ethane and maintained at their melting point, using a Vitrobot (FEI Company, Eindhoven, The Netherlands). The vitreous films were transferred to a Tecnai F20 TEM (FEI Company), and the samples were observed in a low dose mode. Images were acquired at 200 kV at a temperature between −170 and −175 °C, using low-dose imaging conditions with a CCD Eagle camera (FEI Company).²⁰

The size distribution (average diameter and polydispersity index, P.I.) and zeta potential of the samples were measured using a Zetasizer nano (Malvern Instrument, London, United Kingdom) after their dilution with the glycerol/water solution.

The entrapment efficiency (EE%) was expressed as the percentage of the amount of incorporated curcumin versus that initially used. The drug amount was determined by HPLC, after dilution of the sample with methanol (1/100) using a chromatograph Alliance 2690 (Waters, Milano, Italy) equipped with a photodiode array detector and a computer integrating apparatus (Empower™ 3). The column was a SunFire C18 (3.5 μm, 4.6 × 150 mm). The mobile phase was a mixture of acetonitrile, water and acetic acid (95 : 4.84 : 0.16, v/v), delivered at a flow rate of 0.7 ml min⁻¹. Curcumin content was quantified at 424 nm.

Quantitative determination of phospholipids was carried out using the Stewart assay.⁴⁵ Vesicle dispersions (10 μl) were added to the reagent and the obtained solutions were maintained in the dark at room temperature for 30 min and then analysed at 485 nm using a UV spectrophotometer (Spectrometer Lamba 25, Perkin Elmer, Milan, Italy). Aggregation efficiency (AE%) was calculated as the percentage of the P90G amount after dialysis (aggregated) against that initially used.

5.5. Small-angle X-ray scattering

The vesicle structure was studied by SAXS. Analysis was carried out at 25 °C using a S3-MICRO SAXS camera system (HECUS X-ray Systems, Graz, Austria). The working q-range was 0.003–0.6 Å⁻¹, where q = (4π sin θ)/λ is the modulus of the scattering wave vector, θ is the scattering angle and λ is the wavelength. All the scattering curves were reproduced three times and a representative curve was selected, plotting the scattering intensity (I) as a function of the scattering vector (q). SAXS patterns were analysed using the GAP (Global Analysis Program) software developed by Pabst.³³ GAP enables fitting of the SAXS pattern of the bilayer-based structures, i.e. vesicles and lamellar phases.

5.6. Nebulization and aerodynamic behaviour of vesicular formulations

A Pari SX® air jet nebulizer was used and connected to the Next Generation Impactor™ (NGI, Eur. Ph 7.2, Copley Scientific Ltd., Nottingham, United Kingdom). Samples (3 ml) were placed in the nebulizer and aerosols were prepared for 15 min at 15 l min⁻¹.⁴⁶ At the end of each experiment, the sample deposited in each single stage of the impactor and that undelivered was collected into glass vials using methanol. The amount of curcumin was determined by HPLC. The Emitted Dose (ED%) was calculated as the percentage of drug recovered in the NGI versus the amount of drug placed in the nebulizer. The Fine Particle Dose (FPD) was calculated as the amount of curcumin contained in droplets ≤5 μm in size and the Fine Particle Fraction (FPF), as the ratio of FPD to the total recovered dose. The total amount of curcumin recovered from the impactor was calculated as the sum of all of the recovered drug. The cumulative amount of curcumin-containing particles with a diameter lower than the stated size at each stage was plotted as a percentage of the recovered drug versus the cut-off diameter. The mass median aerodynamic diameter (MMAD) of the particles was extrapolated from the graph (Eur. Ph. 7.2). MMAD and geometric standard deviation (GSD) values were calculated without including the mass deposited in the induction port because of the unavailability of a precise upper size limit for particles deposited in this section.⁴⁷ The size and size distribution of the vesicles deposited in each stage of the impactor were measured.

5.7. In vitro cytotoxicity studies (MTT assay)

Human basal epithelial alveolar cells (A549) were grown in Dulbecco’s Modified Eagle Medium containing L-glutamine, supplemented with 10% foetal bovine serum, 1% penicillin/streptomycin and 1% fungizone, at 37 °C, 5% CO₂ and 100% humidity. Before the experiment, A549 cells were placed into 96-well plates at a density of 7.5 × 10³ cells per well. After 24 h, the cells were treated with curcumin (40, 20, 10, 4, 2 μg ml⁻¹) in dispersion or loaded into vesicles. The viability of the cells was evaluated after 48 h experiment by means of the MTT test. The reagent [3(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide] (0.5 mg ml⁻¹) was added to each well and after 2/3 h the formed crystals were dissolved in dimethyl sulfoxide. The sample intensity was measured spectrophotometrically at 570 nm with a microplate reader (Synergy 4, Reader BioTek Instruments, Bernareggio, Italy).¹⁸ All of the experiments were repeated at least three times and in triplicate. The results were expressed as the percentage of cell viability in comparison with non-treated control cells (100% viability).

5.8. Effect of curcumin formulations on cell oxidative stress damage

A549 cells were seeded in 96 well plates and incubated at 37 °C in 5% CO₂ until confluence was reached.⁴⁸ Cells were treated with H₂O₂ (1 : 30 000 dilution) and curcumin in a dispersion or loaded into vesicles (20 μg ml⁻¹), incubated for 6 hours and then washed with PBS. Untreated cells were used as the control (100% viability) while cells treated with H₂O₂ only, were the positive control. The MTT assay was used to assess the protective effect of curcumin on cell damage and death.²⁰

5.9. In vitro inhibition of cytokines by curcumin formulations

The ability of curcumin to inhibit the release of the proinflammatory cytokines IL-6 and IL-8 was measured. A549 cells (micoplasma free) were seeded in 96 well plates (10⁵ cells per well) and incubated at 37 °C in 5% CO₂. After 24 h, curcumin loaded vesicles (10 μg ml⁻¹) were added and the cells were incubated for 4 days. To inhibit the protein secretion, 18 h before the test, the protein transport inhibitor cocktail (eBiosciences, San Diego, CA, USA) was added. At the end of the experiment, cells were fixed with paraformaldehyde (3% in PBS) for 15 min at 25 °C, washed two times with foetal calf serum (FCS, 0.5% in PBS) at 4 °C, permeabilized with triton (0.3% in PBS) and stained with the anti-cytokine antibodies (anti human IL-8 and IL-6, Biolegend, London, UK) conjugated with fluorescein (FITC). Cell analysis was carried out by Facscalibur (Becton Dickinson, Milan, Italy) using a CellQuest software package (Becton Dickinson, Milan, Italy). For all the samples, 20 000 events were acquired in the R1 region gate, which was defined based on forward and side light scatter properties to exclude debris.

5.10. In vivo curcumin deposition in the lungs

Male Wistar rats, 2–3 months old and weighing 280–310 g, were obtained from the animal facilities of the Faculty of Pharmacy, University of Valencia. They were kept at a temperature of 23 ± 1 °C, humidity of 60%, with light–dark cycles of 12 hours, fed a standard diet and had ad libitum access to water. The protocol for the study was approved by the Animal Care Committee (protocol A1352991914316) of the Faculty of Pharmacy at the University of Valencia, Spain.

Rats were cannulated in the jugular vein to facilitate blood sample collection using a procedure reported previously.⁴⁹ Each curcumin formulation was diluted 1 : 1 in water and 100 μl of the obtained dispersion (0.10 μg) was intratracheally administered to each rat (6 animals for group) under isoflurane anaesthesia. Blood samples (200 μl) were taken with heparinized syringes connected to the jugular vein cannula at 0.5, 1, 2, 3, 4, 6, 8, and 24 h after dosing. Samples were centrifuged at 1500g for 5 min and the supernatant plasma was stored at −20 °C until analysis. At 24 h, the rats were sacrificed and their liver, lungs and trachea were excised. Curcumin fluorescence (excitation at 410 nm and emission at 535 nm) was observed using the In Vivo FX PRO Imaging System (Bruker BioSpin, Barcelona, Spain).

The curcumin concentration in the plasma and tissue samples was determined by HPLC. Prior to injection, the plasma samples were deproteinized with acetonitrile while the tissue samples were homogenized in a mixture of water and acetonitrile (30 : 70, v/v) using an Ultra-Turrax®T25 homogenizer (IKA®Werke GmbH & Co, Staufen, Germany) and centrifuged at 12 500g for 15 min. The left inferior lobe, right inferior lobe, right superior lobe and the rest of the whole lungs were separately treated and analysed at 425 nm, using a chromatograph Thermo Scientific (Madrid, Spain), a Novapak C18 column (Waters, Madrid, Spain) and a mobile phase of water/acetonitrile/acetic acid (24 : 75 : 1, v/v/v) delivered at a flow rate of 1 ml min⁻¹.

5.11. Statistical analysis of data

The results are expressed as the mean ± standard deviation and the significance was tested at the 0.05 level of probability (p). Analysis of variance (ANOVA) was used to substantiate the statistical differences between the groups while the Student’s t-test was used for comparison between the two samples using XLStatistic for Excel. The multiple comparisons of in vivo results were performed by the Scheffe or Dunnet tests using IBM SPSS statistics for Windows.

Acknowledgements

This work was partially supported by grants from MIUR (PRIN 2010–2011, Prot. 2010H834LS_004).

Notes and references

1. J. B. Calixto, M. F. Otuki and A. R. S. Santos, Planta Med., 2003, 69, 973.
2. J. B. Calixto, M. M. Campos, M. F. Otuki and A. R. Santos, Planta Med., 2004, 70, 93.
3. X. Hu, F.-F. Yang, L.-H. Quan, C.-Y. Liu, X.-M. Liu, C. Ehrhardt and Y.-H. Liao, Eur. J. Pharm. Biopharm., 2014, 88, 1064.
4. M. Masuda, N. Suzuki, S. Taniguchi, T. Oikawa, T. Nonaka, T. Iwatsubo, S. Hisanaga, M. Goedert and M. Hasegawa, Biochemistry, 2006, 45, 6085.
5. M. E. Egan, M. Pearson, S. A. Weiner, V. Rajendran, D. Rubin, J. Glöckner-Pagel, S. Canny, K. Du, G. L. Lukacs and M. J. Caplan, Science, 2004, 304, 600.
6. Herbal Drugs: Ethnomedicine to Modern Medicine, ed. K. G. Ramawat, Springer Berlin Heidelberg, Berlin, Heidelberg, 2009.
7. M. T. Huang, H. L. Newmark and K. Frenkel, J. Cell. Biochem., 1997, 27, 26.
8. M. L. Manca, I. Castangia, M. Zaru, A. Nácher, D. Valenti, X. Fernàndez-Busquets, A. M. Fadda and M. Manconi, Biomaterials, 2015, 71, 100.
9. K. K. Chereddy, R. Coco, P. B. Memvanga, B. Ucakar, A. des Rieux, G. Vandermeulen and V. Préat, J. Controlled Release, 2013, 171, 208.
10. M. Salahshoor, S. Mohamadian, S. Kakabaraei, S. Roshankhah and C. Jalili, Journal of Traditional and Complementary Medicine, 2015 DOI:10.1016/j.jtcme.2014.11.034.
11. Q. Yao, Y. Lin, X. Li, X. Shen, J. Wang and C. Tu, Toxicol. Lett., 2013, 222, 72.
12. C. Mohanty, M. Das and S. K. Sahoo, Mol. Pharm., 2012, 9, 2801.
13. O. Naksuriya, S. Okonogi, R. M. Schiffelers and W. E. Hennink, Biomaterials, 2014, 35, 3365.
14. M. Zaru, C. Sinico, A. De Logu, C. Caddeo, F. Lai, M. L. Manca and A. M. Fadda, J. Liposome Res., 2009, 19, 68.
15. M. Zaru, S. Mourtas, P. Klepetsanis, A. M. Fadda and S. G. Antimisiaris, Eur. J. Pharm. Biopharm., 2007, 67, 655.
16. Q. T. Zhou, S. S. Y. Leung, P. Tang, T. Parumasivam, Z. H. Loh and H.-K. Chan, Adv. Drug Delivery Rev., 2015, 85, 83.
17. Z. Liang, R. Ni, J. Zhou and S. Mao, Drug Discovery Today, 2015, 20(3), 380–389.
18. M. L. Manca, C. Sinico, A. M. Maccioni, O. Diez, A. M. Fadda and M. Manconi, Pharmaceutics, 2012, 4, 590.
19. M. L. Manca, M. Zaru, M. Manconi, F. Lai, D. Valenti, C. Sinico and A. M. Fadda, Int. J. Pharm., 2013, 455, 66.
20. M. L. Manca, I. Castangia, C. Caddeo, D. Pando, E. Escribano, D. Valenti, S. Lampis, M. Zaru, A. M. Fadda and M. Manconi, Colloids Surf., B, 2014, 123, 566.
21. A. J. Bollet, W. M. Bonner and J. L. Nance, J. Biol. Chem., 1963, 238, 3522.
22. B. T. Shannon and S. H. Love, Immunol. Commun., 1980, 9, 735.
23. Z. Darzynkiewicz and E. A. Balazs, Exp. Cell Res., 1971, 66, 113.
24. K. Pritchard, A. B. Lansley, G. P. Martin, M. Helliwell, C. Marriott and L. M. Benedetti, Int. J. Pharm., 1996, 129, 137.
25. S. T. Lim, G. P. Martin, D. J. Berry and M. B. Brown, J. Controlled Release, 2000, 66, 281.
26. S. Arpicco, C. Lerda, E. Dalla Pozza, C. Costanzo, N. Tsapis, B. Stella, M. Donadelli, I. Dando, E. Fattal, L. Cattel and M. Palmieri, Eur. J. Pharm. Biopharm., 2013, 85, 373.
27. S. Barbault-Foucher, R. Gref, P. Russo, J. Guechot and A. Bochot, J. Controlled Release, 2002, 83, 365.
28. F. Andrade, D. Rafael, M. Videira, D. Ferreira, A. Sosnik and B. Sarmento, Adv. Drug Delivery Rev., 2013, 65, 1816.
29. M. Rinaudo, Prog. Polym. Sci., 2006, 31, 603.
30. S. M. van der Merwe, J. C. Verhoef, J. H. M. Verheijden, A. F. Kotzé and H. E. Junginger, Eur. J. Pharm. Biopharm., 2004, 58, 225.
31. V. K. Mourya and N. N. Inamdar, J. Mater. Sci.: Mater. Med., 2009, 20, 1057.
32. J. Li, I. W. Lee, G. H. Shin, X. Chen and H. J. Park, Eur. J. Pharm. Biopharm., 2015, 94, 322.
33. G. Pabst, M. Rappolt, H. Amenitsch and P. Laggner, Phys. Rev. E: Stat. Phys., Plasmas, Fluids, Relat. Interdiscip. Top., 2000, 62, 4000.
34. M. Zaru, M. L. Manca, A. M. Fadda and G. Orsini, Glycerosomes and Use Thereof in Pharmaceutical and Cosmetic Preparations for Topical Applications, 2012.
35. M. L. Manca, M. Manconi, M. Zaru, I. Castangia, A. Cabras, N. Cappai and A. M. Fadda, Ialurosomi, Loro Uso in Composizioni Topiche Farmaceutiche O Cosmetiche E Relativo Procedimento Di Preparazione, 2014, RM2014A000687.
36. A. Jintapattanakit, S. Mao, T. Kissel and V. B. Junyaprasert, Eur. J. Pharm. Biopharm., 2008, 70, 563.
37. S.-N. Park, H. J. Lee, K. H. Lee and H. Suh, Biomaterials, 2003, 24, 1631.
38. S. Ganesan, A. N. Faris, A. T. Comstock, S. S. Chattoraj, A. Chattoraj, J. R. Burgess, J. L. Curtis, F. J. Martinez, S. Zick, M. B. Hershenson and U. S. Sajjan, Respir. Res., 2010, 11, 131.
39. R. Motterlini, R. Foresti, R. Bassi and C. J. Green, Free Radical Biol. Med., 2000, 28, 1303.
40. G. C. Jagetia and B. B. Aggarwal, J. Clin. Immunol., 2007, 27, 19.
41. R. Abu-Dahab, U. F. Schäfer and C.-M. Lehr, Eur. J. Pharm. Sci., 2001, 14, 37.
42. M. V Suresh, M. C. Wagner, G. R. Rosania, K. A. Stringer, K. A. Min, L. Risler, D. D. Shen, G. E. Georges, A. T. Reddy, J. Parkkinen and R. C. Reddy, Am. J. Respir. Cell Mol. Biol., 2012, 47, 280.
43. S. Wang, T. Jiang, M. Ma, Y. Hu and J. Zhang, Int. J. Pharm., 2010, 386, 249.
44. R. J. Verheul, M. Amidi, S. van der Wal, E. van Riet, W. Jiskoot and W. E. Hennink, Biomaterials, 2008, 29, 3642.
45. J. C. M. Stewart, Anal. Biochem., 1980, 104, 10.
46. V. A. Marple, B. A. Olson, K. Santhanakrishnan, D. L. Roberts, J. P. Mitchell and B. L. Hudson-Curtis, J. Aerosol Med., 2004, 17, 335.
47. M. Taki, C. Marriott, X.-M. Zeng and G. P. Martin, Int. J. Pharm., 2010, 388, 40.
48. M. L. Manca, M. Manconi, A. M. Falchi, I. Castangia, D. Valenti, S. Lampis and A. M. Fadda, Colloids Surf., B, 2013, 111, 609.
49. F. Torres-Molina, J. C. Aristorena, C. Garcia-Carbonell, L. Granero, J. Chesa-Jiménez, J. Pla-Delfina and J. E. Peris-Ribera, Pharm. Res., 1992, 9, 1587.

---