FORMULATION AND OPTIMIZATION OF TERBINAFINE HCl SOLID LIPID NANOPARTICLES FOR TOPICAL ANTIFUNGAL ACTIVITY

Methods: SLNs were prepared using the solvent injection technique. Glyceryl Monostearate (GMS) served as the lipid base. Three stabilizers; Tween 80, Cremophor RH40, and Poloxamer 188, were used. The effect of stabilizer type and concentration, as well as the lipid concentration, were studied, factorial design of 32*21was applied. The prepared SLNs were characterized regarding their particle size, zeta potential, polydispersity index (PDI), entrapment efficiency percent (EE %), and physicochemical stability. The selected formulae were subjected to further investigations such as morphological studies, in vitro release studies, and Infrared (IR) spectroscopy. They were compared with the marketed cream Lamifen® in term of their antifungal activity against Candida albicans.


INTRODUCTION
Nanoparticles (NPs) are colloidal drug carriers that range between 1 to 1000 nanometres (nm) in size with a surrounding interfacial layer. NPs have many advantages such as low cost of ingredients, ease of preparation as well as being suitable for the preparation of hydrophilic and lipophilic drugs. These particles also designed to enhance the solubility of poorly soluble drugs [1][2][3]. NPs may be classified according to the type of polymer used in their preparation, and, so, they are called polymeric NPs or according to the type of lipid used and, hence, called lipid-based NPs. Polymeric NPs are colloidal carriers composed of biodegradable macromolecular polymers. They increase the therapeutic performance of the poorly soluble drug in any route of administration, but unfortunately, they suffer from toxic degradation of toxic monomers aggregation [4]. Lipid-based colloidal particles have an advantage over polymeric systems. They avoid the toxic effect, which may be obtained by the polymers used in the polymeric system [5,6].
Lipid NPs can be classified based on the type of the lipid used. Many lipids can be used, such as fatty acids, glycerides, and waxes. Some important factors must be considered during the selection of the lipid phase, such as the solubility of the drug in it, the partitioning of the drug between the lipid and aqueous phase, and the lipid polymorphism. A solid lipid was selected in the study to produce solid lipid nanoparticles (SLNs). Using solid lipids in the formulation of NPs has an advantage of reducing drug mobility, decreasing drug leakage and also decreases its migration into the interfacial layer. SLNs also have the advantage of improving the bioavailability of the drug by increasing their dissolution and absorption [4].
The main components of SLNs are solid lipid, stabilizers, and active ingredients. Co-stabilizers may be used. Stabilizers are another name for the surface active agents or the emulsifier. They are the key of formation and stabilization of lipid NPs. They help in dispersing the lipid in the aqueous phase during the manufacturing process. The main action of stabilizer is lowering the surface free energy; thus, the surface tension between the two phases will decrease (first role). Once the lipid NPs were formed, the stabilizer performs its second function of stabilizing its structure (second role) [7].
Terbinafine HCl (TFH) is effective in the treatment of different skin diseases such as dermatophytosis, candidiasis, seborrhoea dermatitis, and onychomycosis. Topical therapy is usually the most preferred due to the ease of administration. SLNs possess several advantages for the topical route of administrations as they increase the penetration of the encapsulated drug into the skin along with the avoidance of the first-pass metabolism, the avoidance of side effects as allergic reactions, and also targeting the active ingredient to the affected areas over the skin [8,9].

Preparation of SLNs of TFH
In this study, the solvent injection technique was the method used to prepare TFH SLNs. The solid lipid was dissolved in a water-miscible solvent (e. g., ethanol, methanol, and acetone) or in a mixture of these solvents together. This lipid solvent mixture was injected through an injection needle into a stirred aqueous phase with or without a stabilizer. The presence of a stabilizer within the aqueous phase helped to produce lipid droplets at the site of injection and stabilize SLNs by reducing the surface tension between water and solvent [10,11]. TFH (1%) was mixed with glyceryl monostearate (GMS) and then dissolved into ethanol. They were heated together to above the melting point of the lipid (>55°C). The resulted organic solution was injected at once into the aqueous phase (distilled water and the nonionic stabilizer). The aqueous phase immediately turned milky, and this indicated the formation of lipid NPs suspension. The resulted suspension was stirred at 500 rpm for three hours on a magnetic stirrer. After that, SLNs suspension was sonicated for 15 min.

Factorial design
The factorial design was considered as an efficient statistical tool to study the influence of numerous factors concurrently. In addition, it was more practical to predict and estimate the interaction between selected variables and thus reduced the number of experimental runs. Stepwise multivariable regression analysis was also used to determine the significant factors.
Eighteen formulae were prepared and entered in JMP (version 14 software SAF instate inc., Cary NC), a factorial design program used to select the optimized formula. A 3 2 *2 1 full factorial design was used to define the optimum conditions regarding the selected factors. The design involved two factors, each with three levels and one factor with two levels. The two factors with the three levels reflect the concentration of GMS (2.5, 5 and 10%) and the stabilizer types (Tween 80, Cremophor RH40 and Poloxamer 188) and the one factor with the two levels reflect the stabilizer concentration (1 and 2%) as shown in table 1.

Selection of the optimized formula by factorial design
Applying the desirability function of the used software gave rise to a new formula F19 with the following composition: 6.25% GMS and 1.5% tween 80. This formula was the most optimized formula in the model (table 2). Fig. 1 illustrates the composition of the optimized formula F19 as well as its expected values of particle size and EE%.

Physicochemical characterization of TFH SLNs
Particle size analysis, polydispersity index, and zeta potential measurement Particle size analysis, polydispersity index (PDI), and zeta potential (ZP) were performed for all TFH SLNs formulae using Malvern Zetasizer. Prior to measurement, two drops of SLNs suspension were diluted in 10 ml of distilled water and sonicated for 10 min. After that, 0.1 ml was taken and further diluted in another 10 ml of distilled water and sonicated for 10 min. The second diluted sample was filled in a transparent sizing cuvette [12,13,6]. The PDI value gave an indication of the homogeneity of the preparations [14]. All measurements were performed in triplicate.

Determination of the drug entrapment efficiency percent (EE %)
The EE% gave an indication about the suitability of the drug to be formulated in a SLNs system. EE% was determined by the indirect method by measuring the concentration of the remaining free drug in the aqueous phase, which contained the SLNs dispersion by using the centrifugation method at 5000 rpm for 30 min at 25 °C [15]. The concentration of the free TFH was measured spectrophotometrically at λmax=283 nm [16]. All experiments were run in triplicates. EE % was determined using the following equation: Winitialdrug: was the amount of TFH initially used for the preparation of the SLNs, Wfree drug in suspension: was the amount of free drug determined in the aqueous phase after separation of the NPs [17].

Morphology of TFH SLNs
The morphology of the selected TFH SLNs suspension was observed using a transmission electron microscope (TEM). One drop of the diluted suspension sample was deposited on a film-coated 200-mesh copper grid, stained with one drop of 3% (w/v) aqueous solution of phosphotungstic acid and allowed to dry. Any excess fluid was removed with filter paper before the examination [18].

In vitro drug release study
One gram of the optimized formulae (equivalent to 10 mg of TFH) was placed in cellulose membrane, pore size (0.45 µm), at the dissolution vessel of the USP dissolution test apparatus II. The dissolution media was 250 ml of phosphate buffer (pH 5.5) containing 0.8% tween 80, which was added into the dissolution media to maintain the sink conditions. The release study was carried out at 37 °C±0.5 [9]. The stirring paddle was rotated at a speed of 50 rpm [19]. Two ml samples were withdrawn and replaced by 2 ml fresh phosphate buffer solution at 5, 10, 15, 30, 60, 90, 120, and 150 min time intervals. They were spectrophotometrically analyzed for TFH at its λmax 283 nm. The blank was prepared according to the same procedure by using a plain SLNs suspension. The cumulative amount of TFH released was plotted as the function of time, and the release rate was calculated from the slope of the straight-line portion. Experiments were repeated three times as the results were expressed as the mean values±SD.

Infrared (IR) spectroscopy
Infrared spectra of physical mixtures of pure TFH with each one of the used excipients, namely GMS, Tween 80 and Cremophor40, as well as those of the selected optimum formulae, were investigated using KBr discs. The IR spectra were obtained in the spectral region 4000-400 cm -1 [20].

Microbiological evaluation of the selected SLNs formulae
The cup-plate technique was used to evaluate the antifungal activity of the selected SLNs, as described by Barot et al., and Sahoo et al., [21,22]. The evaluation was carried out against Candida albicans ATCC 10231, which obtained from The National Research Centre, Cairo, Egypt. The selected formulae and the marketed lamifen ® cream were prepared at a concentration of 1% TFH. Sabouraud's dextrose agar was used for the evaluation of these formulae against Candida albicans. Dextrose agar was poured into a sterile petri dish followed with surface inoculation of Candida albicans (10 7 CFU/ml). Three wells (7 mm diameter hole cut) were made in the middle of the agar petri dish and filled with 0.1 ml of the selected SLNs formulae, their plain base and lamifen ® cream, respectively. The plate was then incubated at 37 °C for 48 h. The diameter of zone of inhibition was measured in mm after incubation, recorded and considered as an indication for antifungal activity. This method was done in triplicate and mean results were calculated [23][24][25].

Effect of the storage time on stability of the selected formulae
TFH loaded SLNs optimized formulae were investigated for their stability. The samples were stored at controlled ambient temperature (25°C) for three and six months. The samples were analyzed for their particle size, zeta potential, PDI and EE%.

Statistical analysis
All data were expressed as mean±standard deviation (n=3). They were subjected to one way ANOVA followed by Tukey Kramer's post-test using Instat 3 software. A difference between means was considered significant if the P-value was less than or equal to 0.05.

Physicochemical characterization
The effect of different formulae variables and their interaction was studied and statistically analyzed. The results are listed in table 3and illustrated in fig. 3.

Particle size analysis and polydispersity index (PDI)
Particle size was selected to be one of the dependent variables, and the PDI value gave an indication for the homogeneity of the preparations. PDI of all the prepared formulae was found to be (0.375-1).
The effect of stabilizer type on the particle size The particle size of the prepared SLNs was significantly affected by changing the stabilizer type as follows, using poloxamer as stabilizer significantly increased the particle size ( ** p<0.0001) while using any of Tween 80 or CremophorRH40 exerted a less significant negative effect and decreased the particle size of the SLNs ( * p<0.05). The SLNs formulae prepared using 1 and 2 % w/w poloxamer 188 (F13 to F18) showed larger particle size than those prepared using 1%and 2 % w/w tween 80 and cremophor RH40 (F1 to F12) (poloxamer188<tween80<cremophor RH40). This may be due to the higher molecular weight of poloxamer 188 compared to tween 80 and cremophor RH 40 [26]. Vitorino et al. [15] stated that the choice of stabilizer is important in the optimization of any NPs formulae. The authors added this component contributes not only to control the particle size and stability of the dispersions, but also to control the crystallization behavior, including polymorphs.

The effect of stabilizer concentration on the particle size
Changing the stabilizer concentration from 1 to 2% exerted a significant effect on particle size ( ** p<0.001). The average diameters of F7 (1% cremophor RH 40) were 334.80±11.21 nm compared to F8 (2% cremophor RH 40) 101±9.45 nm. This was attributed to higher stabilizer concentration, which prevented particles' aggregation [27]. These results were in agreement with Tiwari S et al., who observed that the stabilizer concentration affects the particles' diameter. They attributed this to the increase in stabilizer concentration, which led to the reduction of the interfacial tension between the lipid matrix and the aqueous phase resulting in a formation of SLNs with smaller particle size [28,29].
The effect of lipid concentration on the particle size Changing in lipid concentration exerted a significant main effect on the particle size of the prepared SLNs ( * p<0.05). The particle size of the SLNs formulae prepared using higher lipid concentrations of 10% (F5, F6, F11 [30,31]. The authors added that the lipid concentration was the most important factor affecting particles' diameter. They showed that increasing the lipid content of geleol, compritol 888 ATO and percirol ATO from 5 to 7.5, and 10% resulted in a subsequent increase in particle size [32].

Two-factor interaction
The factorial design analysis revealed that there was a significant two-factor interaction effect of lipid concentration and type of stabilizer on the particle size of the prepared SLNs ( * p<0.05).

Zeta potential
Zeta potential, which reflects the electric charge on the particles surface, was a useful parameter to predict the physical stability of colloidal systems [14,17,32]. The zeta potential indicated the degree of repulsion between close and similarly charged particles in the nanodispersion. The zeta potential values above |8-9 mV| were a prerequisite for the stability of the nanoparticles. Zeta potential less than or equal |21 mV| were considered as advisable for the stabilization of suspensions, since very little or no agglomeration takes place, while |30 mV| were indicated to full electrostatic stability [33].
The results of TFH SLNs zeta potential of different formulae are shown in table 3. The values of zeta potential were between-10.50 mV and-24.43 mV.

Entrapment efficiency
EE% was the second dependent variable selected to study the effect of different formulation variables and their interactions. The results are listed in table 3 and illustrated in fig. 3.

The effect of lipid concentration on EE %
The lipid concentration exerted a highly significant positive effect on EE% of the prepared SLNs ( ** p<0.0001). As shown in fig. 3, a direct relationship was investigated between lipid concentration and EE%. This was attributed either to the increase in lipids content, which may increase the space for the drug molecules to be entrapped or to the increase in the druglipid solubility, which may also lead to an increase the entrapment [12]. These results are in full agreement with (Soma et al., 2013), who studied the effect of increasing GMS on Irbesartan loaded SLNs. They proved that increasing concentration of GMS increased EE% [20].

The effect of stabilizer concentration on EE %
Changing stabilizer concentration from 1 to 2% exerted a significant effect on EE% of the prepared SLNs ( * p≤0.05). All the SLNs formulae prepared using 2% stabilizer concentrations showed higher EE% than their corresponding ones containing 1% stabilizer (table 3). This was because increasing the stabilizer concentration was found to enhance the solubility of the drug and so EE% was increased [26].

Effect of Stabilizer type on EE%
Studying the effect of stabilizer's type on the EE% showed that Tween 80 affected the EE% significantly ( * p<0.05), while both Cremophor RH40 and Poloxamer188 exerted a non-significant effect on it (p values>0.05). This may be due to the difference in chemistry of these stabilizers, which may affect the solubility of the drug in lipid either by increasing or decreasing it, and this will affect EE%.
These results of studying the effect of changing both the type and concentration of stabilizer are in agreement with Ekambaram and Abdul Hasan Sathali, who reported that increasing the concentration of stabilizers from 1% to 2% affected EE% irrespective of the type of stabilizer [26].

Selection of the optimized formulae
After a complete analysis of the factorial design with the previously set levels of the selected independent variables and their effect on the selected important dependant variable of particle size and EE%, the two optimum criteria set were suitable particle size and highest EE%. Suitable PDI was also taken into consideration. According to these criteria, the three optimum formulae (F6, F12, and F19) were selected for further evaluation.

Morphology of prepared SLN formulae
Optimized formulae were selected for studying their surface morphology as they show suitable particle size with the best EE%. The morphology of fresh TFH SLNs was observed using TEM. Morphology TEM was shown in fig. 4 for F6, F12, and F19. TEM photographs revealed that the particles possessed a nanometer-size with spherical uniform shape without aggregation. The particle size ranged from 480 to 490 nm for F6, from 455 to 470 nm for F12 and from 245 to 260 nm for F19.

In vitro drug release study
The drug was released from SLNs and diffused through the cellulose nitrate membrane into the release medium. Drug release was measured by determining the drug concentration in the release medium. Fig. 5 illustrates the release profile of TFH from optimized SLNs. The cumulative release profiles were constructed up to 150 min. For assessment and comparison, the release profiles evaluation was based on the cumulative drug % release and T50%. A nonsignificant difference was recorded when comparing the cumulative drug release from the formulae (P ˃ 0.05). The cumulative drug release was 74.58%, 70.56% and 66.91% for formulae F6, F12 and F19, respectively. In order to determine the release model, the in vitro drug release data were analyzed according to zero order, first order, and diffusion-controlled mechanism according to Higuchi model. The results revealed that the release of TFH from all the three optimized formulae obeyed the Higuchi model of matrix diffusion with highest R 2 values, as shown in table 4. This is in agreement with Gardouh et al. who stated that drug release by diffusion involves three steps which include penetration of water into system causing swelling of matrix followed by the conversion of solid lipid into rubbery matrix, and then the diffusion of drug from the swollen rubbery matrix takes place [34]. To confirm the release mechanism, the data were fitted to the Korsmeyer-Peppas equation to describe the drug release mechanism from matrix. It was revealed that the three formulae showed non-Fickian diffusion with n values 0.45<n<0.89. T50% recorded a value of 49.22±1.6, 59.46±0.7and 61.45±1.3 min for F6, F12 and F19, respectively. They showed a nonsignificant difference. (P ˃ 0.05).

Infrared (IR) spectroscopy
FT-IR spectroscopy was used to characterize possible interactions between the drug and the used excipients. The interactions between them often lead to identifiable changes in the IR profile of the formulae. FT-IR spectra of TFH, TFH/Tween 80 physical mixture (1:1), TFH/GMS physical mixture (1:1), TFH/Cremophor RH40 physical mixture (1:1), F6, F12 and F19were shown in fig. (6a, b, [35]. The IR spectrum of the developed physical mixtures, as well as the formulae (F6, F12, and F19), showed minor differences in the positions of the absorption bands of TFH. These indicated the incorporation of the drug in SLNs formulae without any interaction with neither the lipid nor the stabilizers.

In vitro microbiology assessment
This test was used to prove the efficacy of the selected SLNs formulae against fungal infection caused by Candida albicans. Fig. 7 show the diameters of zones of inhibition (mm) caused by the selected SLNs formulae, plain base for each formula, and the marketed lamifen ® cream. The larger the zone of inhibition of Candida albicans growth, the higher the release rate of the drug from the SLNs and the greater the efficacy of the released drug towards killing the Candida albicans. Formulae F6, F12 and F19, showed a mean zone of inhibition of 25.9±0.25 mm, 25±0.35 mm and 24.67±0.36 mm, respectively, which were significantly greater than the mean zone of inhibition exhibited by the commercial preparation lamifen ® cream (11.2±0.44 mm) ( * P<0.05). The optimized SLNs formulae were stored under ambient conditions for 3 and 6 mo at 25 °C. They were re-evaluated for their particle size and EE%. A slight increase in the particle size diameter ( fig. 8) and slight decrease of EE% ( fig. 9) were observed; both were statistically non-significant (P>0.05, paired t-test), proving the stability of TFH loaded SLNs.

CONCLUSION
In this design, it was noted that lipid concentration exerted a significant effect on particle size and EE%. The type of stabilizer chosen and its concentration also showed a significant effect on the studied parameters. The selected optimized formulae of the prepared TFH SLNs (F6, F12, and F19) obtained optimum particle size with the highest EE%. They showed satisfactory in vitro release profiles. In vitro microbiology testing depicted that the zone of inhibition of these three formulae was visibly more than that observed from the market cream (lamifen ® cream).