Effect of terpene liposomes on the transdermal delivery of hydrophobic model drug , nimesulide : Characterization , stability and in vitro skin permeation

This study evaluates the ability of the terpenes incorporated in liposomes on the in vitro skin delivery of hydrophobic model drug, nimesulide (NE). To this purpose, so-called terpene liposomes (TPs), which are composed of phospholipid and three types of terpene, citral, limonene and cineole. The obtained formulations were characterized in terms of size distribution, zeta potential and morphology. The efficiency of TPs on skin delivery of NE was studied using in vitro Franz diffusion cells and abdominal rat skin in comparison with conventional liposomes and ethanolic solutions of NE. Results showed that all the used TPs had spherical structures with negative zeta potential, low polydispersity (PDI < 0.2), nanometric size range (z-average no more than 150 nm). TPs improved the entrapment efficiency (EE%) and gave good physical stability. In vitro skin permeation data showed that TPs were able to give a significant improvement of NE permeation through the rat skin in comparison with conventional liposomes and drug solution. Moreover, the TPs prepared with limonene were also able to deliver a higher amount of NE than the other formulation, thus suggesting that NE delivery to the skin was strictly correlated to type of terpenes incorporated liposomes.


INTRODUCTION
The liposomes have been studied as a system for dermal and transdermal delivery (Verma and Fahr, 2004;Biruss and Valenta, 2007;Knudsen et al., 2011).Many researchers had reported that the skin delivery of numerous drugs was enhanced following application of liposomes.For instance, tretinoin for the treatment of acne (Schafer- Korting et al., 1994), glucocorticoids for the treatment of atopic eczema (Korting et al., 1990), lignocaine and tetracaine as anesthetics (Gestztes and Mezei, 1988).The possible mechanisms by which conventional liposomes could increase skin delivery of drugs have been appraised.These include the vesicles intact with the skin surface and its components entering *Corresponding author.E-mail: mbadran75@gmail.com.Tel: + 966 1 4678533.
the intercellular lipid matrix of the stratum corneum (SC), modifying the lipid lamellae (Kirjavainen et al., 1996).Another mechanism reported that intact vesicles might penetrate the SC, due to the effect of transepidermal osmotic pressure (Cevc and Blume, 1992).Cevc et al. (1998) found that the conventional liposomes do not exhibited a penetration effect upon topical applications.Touitou et al., (2000) reported that the conventional liposomes could not deeply penetrate skin because they remain restricted to upper layers of the stratum corneum.Therefore, only localized or rarely transdermal effects of tradional liposomes have been observed.
Recently, several approaches have declared that elastic membrane of the liposomes could enhance the skin penetration.Therefore, the flexible liposomes could result in improved drug transport across the skin as compared to vesicles with rigid membrane (El Maghraby et al., 1999, 2001).As a result, a series of vesicles with flexible membranes were established in order to enhance the dermal or transdermal delivery of the drugs, for example, the liposomes with edge activators such as sodium cholate, Span 80 and Tween 80 (Cevc and Gebaure, 2003;Cevc and Blume, 2004).An edge activator destabilises lipid bilayers of the vesicles and increases flexibility of the membrane (Honeywell-Nguyen et al., 2003;Honeywell-Nguyen and Bouwstra, 2005).These vesicles had ability to penetrate intact skin underocclusive application efficiently.Moreover, the elasticity of the vesicle membranes could be improved by adding the ethanol to the lipids.Nevertheless, the incorporation of high concentration of ethanol to the vesicles causes interdigitation effect on lipid bilayers (Dubey et al., 2007).These defined carrier systems are not sufficient enough to transport the drugs across the SC.
Recently, Verma (2002) explored novel vesicles with high flexible membranes after incorporation ethanol and terpenes into the lipids.These vesicles were employed for topical application (Verma et al., 2004;Dragicevic-Curic et al., 2008;Chen et al., 2011).It is well known that the terpenes have widely been used as penetration enhancers and as constituents of vesicle dispersions due to their low toxicity and relative safety.There has also been proof of their low skin irritancy at low concentrations (Okabe et al., 1990).Cineole, citral and limonene are examples of terpenes which were extensively used for topical delivery of both hydrophilic and lipophilic drugs.When skin is treated with these terpenes, the existing network of hydrogen bonds between ceramides may get loose and break (Jain et al., 2002).Very recently, flexible vesicles prepared, using 1% of these terpenes, were described as transdermal carrier for systems (Dragicevic-Curic et al., 2009;Chen et al., 2011).
Therefore, the purpose of this study was to evaluate the liposomal systems containing terpenes, given a flexible bilayer as skin drug delivery of hydrophobic model drug, nimesulide (NE).To this purpose, this system is called terpene liposomes (TPs).NE is selected as drug model due to very low solubility in water (0.01 mg/mL), an octanol-water partition (log P) of 2.60 and a pKa value of 6.46 (Khan et al., 2011).For this reason, three different types of flexible liposomes containing cineol, citral and limonene were prepared.The obtained liposomes were characterized in terms of particles size, zeta potential, stability and morphology.Permeation experiments were conducted in order to investigate the permeation ability of these vesicles.

Preparation of different types of terpene liposomes (TPs)
The conventional liposomes (NE-CL) containing NE were prepared by a rotary evaporation method (Ita et al., 2007).The prepared liposomes and their composition are presented in Table 1.Briefly, the lipoid S75, NE and cholesterol were dissolved in organic solvent, chloroform/methanol (1:2) in a round bottom flask.The flask was connected to a rotor evaporator under vacuum (Rotavapor, Büchi, Germany) and immersed in a water bath preheated at temperature equal or more than the transition temperature of phospholipids, in the case of lipoid S75 is about 45°C.A thin film of the lipid was formed on the wall of the flask.The lipid film was then flashed with nitrogen gas for removal of possible traces of organic solvent.The liposomal dispersion was formed after film rehydration with PBS pH 7.4.The liposomal dispersion was sonicated by means of the probe sonicator to get liposomes of the smaller size.
In the case of terpene liposomes (TPs) containing cineole, limonene and citral were prepared by ethanol dissolution method (Verma et al., 2004).The composition of terpene liposomes is presented in Table 1.The liposomes were prepared by dissolving 1% w/v of terpenes and NE in the ethanolic solution of lipoid S75.This organic mixture was vortexed for 5 min and afterwards sonicated for 5 min until a clear solution was observed.The PBS pH 7.4 was added to the mixture by a syringe under constant vortexing.The vortexing was continued for additional 5 min.The liposome dispersions were then probe sonicated under ice to obtain liposomes of the smaller size.All liposomal dispersions were stored at 4°C until further investigation.The NE ethanolic solution was prepared for comparison with vesicles system.

Photon correlation spectroscopy (PCS)
Dynamic light scattering was measured at 25°C.The particle size, the polydispersity index (PDI), and zeta-potential (mV) were measured using a Zeta plus instrument (Brookhaven Instruments, Brookhaven, USA).The liposomes were diluted with the deionized water prior the measurement.The samples were analyzed 24 h after preparation.

Enterapment efficiency (EE%)
The EE% of NE was determined by using ultracentrifugation method (Heeremans et al., 1995).The free NE was separated from entrapped NE by using the ultracentrifugation (Optima TM Max-E, Ultra Centrifuge, Beckman Coulter, Pasadena, CA) at 50,000 rpm at 4°C, for 30 min.Purified sediment was then diluted to the initial volume using PBS (pH = 7.4) in order to keep a final PC concentration of 10% (w/v), and used directly for in vitro permeation study.Entrapment efficiency of NE was calculated indirectly from the amount of free drug, according to the following equation: Where NEf is the amount of free NE; NEt is the total amount of NE.

Morphology observation
Transmission electron microscopy (TEM) was used to visualize TPs vesicles.Samples were dried on a carbon-coated grid and negatively stained with a 1% w/v aqueous solution of phosphotungestic acid.The excess of this solution was removed and after drying, the samples were examined under microscope at 20 or 25 kV by TEM (JEM-200 CX, JEOL, Tokyo, Japan) (Manconi et al., 2003).

Storage stability studies
In order to determine the physical stability of liposomes, their particle size and PDI (by PCS) were measured.The vesicles were stored at 4°C for up to 2 months under light protection (Nasr et al., 2008).In predetermined time intervals, vesicles were subjected to macroscopical observation and characterized for their vesicle size and PDI.

In vitro permeation studies
It is difficult to get human skin, thus an abdominal rat skin was used.Previously, rat skin membranes model were well established for in vitro testing, as it is comparable to human skin in stratum corneum thickness, as well as water permeability (Walters and Roberts, 1993).The abdominal hair of male rat skin was removed with an electric clipper carefully.After the rats were scarified, the subcutaneous fat tissue was carefully removed from the skin by means of a scalpel and surgical scissors.Afterwards, the skin was wrapped into aluminum foil and stored at -20°C until use, less than 4 weeks (Kong et al., 2011).All studies were in accordance with the Guidelines of Animal Ethical Committee of King Saud University, and had its approval.Prior to the experiments, the skin samples were taken from the freezer and let thaw at room temperature for about 30 min.After thawing, the skin surface was carefully wiped with cotton wool balls wetted with PBS buffer.Skin samples were mounted onto Franz diffusion cells, with a nominal area for diffusion of 1.76 cm 2 and a receptor volume of about 12 ml.The epidermal side of the skin was exposed to ambient conditions while the dermal side was bathed with BPS buffer pH 7.4 containing 10% polyethylene glycol 400 to keep sink condition (Babu et al., 2003).The receptor fluid was kept at 37 ± 1°C throughout the experiments, to reach the physiological skin temperature (that is, 32 ± 1°C).The constant stirring was maintained by magnetic stirring at 500 rpm.Care was taken to remove all air bubbles between the underside of the skin (dermis) and the receptor solution throughout the experiment.After equilibration for 30 min, 300 µl of liposomal dispersion were applied to the skin surface, occlusively.At appropriate intervals, sample of 1 ml was withdrawn from the receptor solution and immediately replaced with fresh solution to maintain a constant volume.The incubation times with different drug formulations were 30 h.All samples were analyzed spectrophotometrically for NE content at 396 nm (Singh et al., 2005).No spectrophotometric interference by skin components with NE was observed at this λmax of 396 nm.All experiments were done in triplicate.

Particle size and zeta potential measurements
In the present study, three different terpene liposomes containing cineole, limonene and citral were prepared to overcome the SC barrier properties.Also, the ability of these terpene vesicles on skin delivery of hydrophobic model drug NE, through rat skin, was investigated and compared with an aqueous solution and conventional liposomes.The compositions of the different vesicular systems, their particle size distribution, PDI and zeta potential are presented in Table 1.
Particle size of the formulated vesicles after probe sonicated is presented in Table 1.The results showed that the average size of terpene liposomes containing citral (NE-TP1) was 194.1 with a PDI of 0.261 while the average size of the terpene liposomes containing limonene (NE-TP2) was 216.4 nm, with a PDI of 0.174.In case of vesicles containing cineol (NE-TP3), the vesicle size was 244.1 nm with a PDI of 0.105 was detected.The particle size of conventional liposomes was 164.4 nm.The increasing of particle size of liposomes containing terpenes was observed.These findings are in agreement with previous studies (Dragicevic-Curic et al., 2009;Badran et al., 2009).The PDI of the investigated formulations was below 0.3, which indicates the homogeneity of the prepared NE loaded liposomes (Verma et al., 2004;Elsayed et al., 2007).
Regarding the zeta potential measurements, all liposomal dispersions had a negative surface charge, indicating that the formulations are more stable and homogeneous in distribution (Table 1).Liposomes containing terpenes were more negative than conventional liposomes.These negative charge values of the obtained liposomes are attributed to the presence of ethanol.Ethanol was found to increase negativity of the liposomes (Touitou et al., 2000).

Entrapment efficiency (EE%)
The entrapment efficiency of NE in terpene liposomes NE-TP1, NE-TP2, and NE-TP3 reported, 83.7, 79.2, and 76.4%, respectively, compared with 55.3% reported for conventional liposomes NE-CL (Figure 1).It is obvious that the liposomes containing terpenes represented the largest EE%, which is accompanied with increasing in the particle size and zeta potential of the vesicles.The reason for that was attributed to the high value of zeta potential which frequently led to increase in the repulsion forces of the bilayer structure of the vesicles, which consequently increased the size of the liposomes.NE, being hydrophobic, is expected to be localized in the membrane compartment of lipid vesicles (El Maghraby et al., 2005).

Transmission electron microscope (TEM)
TEM revealed the spherical-shape vesicles (Figure 2).TEM showed slightly dark vesicular structure, as can be seen in Figure 2, where it is possible to notice the outermost bilayer.Dark vesicles are probably due to strong interactions between the terpenes and phosphotungestic acid, which is able to selectively deposit electrons in the sample and enhance structural details (Mura et al., 2009).It was obvious that the vesicles possess interior part that appear more dense towards the central part (Mats et al., 2000).In contrast, the oil droplets are often dark and the surroundings are bright (Gupta et al., 2010).

Stability studies
The physical stability of the conventional and terpene liposomes stored at 4°C for 60 days are presented in Figure 3.The stability results showed that a minimal effect on particle size and PDI of the investigated liposomes was noticed.The particle size and the PDI of different terpene liposomal dispersions have slightly increased after 60 days storage (Figure 3A and B).These results showed that the existence of the terpene in the vesicular formulations did not affect vesicle stability.

In vitro permeation studies
In order to study the influence of these terpene vesicles on the drug permeation through the skin, in vitro skin permeation of hydrophobic model drug, NE loaded terpene liposomes, were conducted using rat skin membrane.For comparison, the permeation of NE from its solution and conventional liposomes were also investigated (Table 2 and Figure 4).At the end of 30 h incubation of the rat skin with NE solution, only small amounts of NE in the receptor fluid were found.
Particularly, when the skin was incubated with the NEterpene liposomes, the higher amounts of NE in the receptor fluid were detected.That is, the terpene liposomes were much more efficient in enhancing the skin permeation of NE than conventional liposomes.Moreover, the relative amounts of NE from the conventional liposomes compared to the hydroethanolic solution were permeated (Table 2 and Figure 4).These influence revealed that the vesicles play a great role in the delivery of the drug into the deeper skin layers (Guo et al., 2000;Verma et al., 2004;El Maghraby et al., 2005;Elsayed et al., 2007).These results showed that the improved permeation of NE with liposomes containing 1% terpenes (and 3.3% ethanol, since all terpene liposomes contain ethanol) was not only a result of the amount of terpenes, but also a result of the potential effect of ethanol and terpenes.The result combination of ethanol and terpenes is already described in the literature (Ota et al., 2003).These data also showed the importance of the effect of liposomes, terpenes and ethanol and its higher skin penetration, compared to the ethanolic solution.Recently, terpene liposomes prepared using cineole, citral and limonene as penetration enhancers with both hydrophilic and lipophilic drugs were developed (Verma, 2002;Hiruta et al., 2006;Badran et al., 2009;Dragicevic-Curic et al., 2009;Chen et al., 2011).
After 30 h of in vitro skin permeation studies, the permea ted amounts of NE in case of terpene liposomes were significantly higher than that obtained in case of conventional liposome and NE solution.In details, the terpene liposomes containing limonene (NE-TP2) showed the highest drug amount in the receptor fluid 18.09, 9.43,    7.102.16and 2.16-fold compared to the hydroethanolic solution (NE-HE), the conventional liposomes (NE-CL), terpene liposomes containing cineol (NE-TP3) and citral (NE-TP1), respectively.Moreover, the amount of NE permeated in case of NE-TP1 were 3.29-fold, compared to NE-TP3 (Figure 5).Then, the permeation data were ranked in the following order NE-TP2 > NE-TP1 > NE-TP3 > NE-CL > NE-HE.
In addition, the flux of NE was enhanced to 20.510±1.298µg/cm 2 h with NE-TP2 (limonene), which is about 17.47, 8.60, 3.77 and 1.61 times compared NE-HE (control), NE-CL, NE-TP3 (cineol) and NE-TP1 (citral), respectively.There was no significant increase in the flux of the drug with NE-HE (1.174 ± 0.2 µg/cm 2 h) compared to that obtained for liposomal formulations.The permeability coefficient NE-NE (control) was 4.496 ± 0.938 cm/h.On presence of vesicles, there are significant increases in the permeability coefficient of the NE compared to the control (Table 2).The permeability coefficient of NE in terpene liposomes NE-TP1, 82.041 ± 5.191 and 21.761 ± 3.055 cm/h, respectively, compared with 9.544 ± 2.707 cm/h reported for conventional liposomes NE-CL (Table 2).These results showed the enhancing ability of three types of terpenes on the rat skin permeation of NE compared to the ethanolic solution and normal liposomes.Moreover, data also showed that the enhanced permeation of NE with liposomes containing 1% terpenes (and 3.3% ethanol, since all liposomes contain ethanol) was not only due to the presence of terpenes, but could be attributed to the combination of ethanol and terpenes.
The synergistic effect in the enhancement of skin permeation due to the combination of ethanol and terpenes is already described in the literature (Bhatia and Singh, 1999;Ota et al., 2003;Dragicevic-Curic et al., 2009).
These results are in agreement with results obtained by   other authors who proposed that ethanol and phospholipids, in combination with terpenes, have enhanced penetration of substances (Van den Bergh et al., 2001: Verma et al., 2004;Touitou et al., 2000;Paolino et la., 2005).The high transdermal enhancement by terpenes suggests that there are possible several mechanisms that could explain permeability of hydrophobic model drug, NE.They include an improvement of partition of NE in the SC lipids that is, higher drug solubility in the skin (Krishnaiah et al., 2002).Additionally, the terpenes may produce phase separation with the SC lipid lamellae (Moghimi et al., 1997;El-Kattan et al., 2000).Also, the terpene might modify the organization of the SC lipids (Verma et al., 2004;Dragicevic-Curic et al., 2009).However, several studies have reported that the addition of ethanol and terpenes makes the phospholipid bilayer highly flexible and deformed vesicles (Verma et al., 2004;Dragicevic-Curic et al., 2008;Curic et al., 2009;Chen et al., 2011).A number of studies also support the permeation enhancing effect for the flexible liposomes (Cevc et al., 1995;Trotta et al., 2002;Honeywell-Nguyen andBouwstra, 2003, Honeywell-Nguyen et al., 2006;Elsayed et al., 2007;Fang et al., 2008;Badran et al., 2009;Chen et al., 2011).According to literature, the improvement of skin permeation of NE in the present study should be interpreted as a prove to the flexibility of terpenes liposomes.
Comparison of results obtained from in vitro skin permeation with these three terpenes, limonene has the highest permeation of NE.This consequence could be attributed to the fact that limonene is hydrocarbon terpene, which has a high chemical enhancement and more efficient for NE permeation in comparison to oxygenated citral and cineole.The results might be due to limonene having highest log p (4.58) than citral (3.45) and cineol (2.82) (El-Kattan et al., 2000).The efficiency of hydrocarbon limonene had also been recognized as enhancer for transdermal delivery of ketoprofen, indomethacin and estradiol (Gupta and Garg, 2002;Williams and Barry, 2004).Conclusively, NE-TP2 (limonene) was considered to be a suitable system for the delivery of NE for 2 reasons.The first is its higher drug flux than that of other formulae, as shown in Table 2.The second reason is related to the cumulative percentage of the drug permeated through rat skin over 30 h, ~ 29% (≈ 869.708 μg).

Conclusion
The present study demonstrated that the terpene vesicular system which consisted of ethanol and terpenes, could enhance the skin permeation of hydrophobic model drug NE, compared to that of conventional liposomes and hydroethanolic solution (control).Moreover, the presence of the terpenes increased the EE% of NE compared to conventional liposomes.Consequently, the terpene liposomes containing limonene could be a good vesicular system for the transdermal delivery of the lipophilic model drug, NE.

Figure 3 .Figure 3 .
Figure 3. Stability study of NE loaded liposomal dispersions stored at 4 o C over 2 months duration.(A) Change of particle size (z-average), (B) change of the polydispersity index (PDI).

Figure 3 .
Figure 3. Stability study of NE loaded liposomal dispersions stored at 4°C over 2 months duration.(A) Change of particle size (zaverage), (B) change of the polydispersity index (PDI).

Figure 4 .
Figure 4.In vitro skin permeation of NE-loaded terpene liposomes, conventional liposomes and hydroethanolic solution after applying for 30 hr.

Figure 4 .
Figure 4.In vitro skin permeation of NE-loaded terpene liposomes, conventional liposomes and hydroethanolic solution after applying for 30 h.

Table 1 .
PBS pH 7.4 (ethanol 60%) composition of the different types of terpene liposomes and PBS control.

Table 2 .
Permeability parameters of the different formulations.