- Email: [email protected]
Optimization of nanostructured lipid carriers loaded with methotrexate: A tool for inflammatory and cancer therapy
International Journal of Pharmaceutics 492 (2015) 65–72
Contents lists available at ScienceDirect
International Journal of Pharmaceutics journal homepage: www.elsevier.com/locate/ijpharm
Optimization of nanostructured lipid carriers loaded with methotrexate: A tool for inflammatory and cancer therapy Mara Ferreiraa,b , Luíse L. Chavesa , Sofia A. Costa Limaa,* , Salette Reisa a b
UCIBIO/REQUIMTE, Department of Chemistry, Faculty of Pharmacy, University of Porto, Portugal Faculty of Engineering of University of Porto, Portugal
A R T I C L E I N F O
A B S T R A C T
Article history: Received 13 May 2015 Accepted 5 July 2015 Available online 10 July 2015
The aim of this study was to optimize and assess the potential of nanostructured lipid carriers (NLC), prepared by the hot ultrasonication method, as carrier for methotrexate (MTX), highlighting the application of factorial design. Preliminary screening drug/lipid solubility, allowed us to select Witepsol1 E85 as the solid lipid and Mygliol1 812 as liquid lipid for the NLC loaded with MTX. Then, a 3-level, 3-factor Box-Behnken design and validated by ANOVA analysis; the correspondence between the predicted values and those measured experimentally confirmed the robustness of the design. Properties of optimized MTX-loaded NLCs such as morphology, size, zeta potential, entrapment efficiency, storage stability, in vitro drug release and cytotoxicity were investigated. NLCs loaded with MTX exhibited spherical shape with 252-nm, a polydispersity of 0.06 0.02, zeta potential of 14 mV and an entrapment efficiency of 87%. In vitro release studies revealed a fast initial release followed by a prolonged release of MTX from the NLC up to 24-h. The release kinetics of the optimized NLC best fitted the Peppas–Korsmeyer model for physiological and inflammatory environments and the Hixson–Crowell model skin simulation conditions. No toxicity was observed in fibroblasts. Thus, the optimized MTX-loaded NLC have the potential to be exploited as delivery system. ã 2015 Elsevier B.V. All rights reserved.
Keywords: Lipid colloidal carriers Methotrexate Witepsol1 E85 Mygliol1 812 Box–Behnken design Hot ultrasonication In vitro drug release Storage stability L929 fibroblasts
1. Introduction Lipid based colloidal carriers have been widely used for drug delivery because they offer the possibility of modulating drug release, by facilitating its transport to the different targets, by increasing local penetration, prolonging residence time and by a controlled release mechanism to provide an effective dose to the target site (Marianecci et al., 2014). Among these lipid colloidal delivery systems (e.g. lipid nanoparticles, liposomes, and nanoemulsions) solid lipid nanoparticles (SLN) emerged in the early nineties as an alternative and current trials applying SLN consider them very promising in drug delivery (Müller et al., 2000). In fact, it is widely accepted that SLN combine the advantages and avoid the disadvantages of other colloidal carriers (Müller et al., 2000). Low drug loading and drug expulsion during storage period are taken as the disadvantage of SLN (Huang et al., 2008). Nanostructure lipid carriers (NLC) have been developed to overcome the drug loading capacity and drug leaking phenomena
* Corresponding author at: Rua de Jorge Viterbo Ferreira, 228, 4050-313 Porto, Portugal. Fax: +351 226093483. E-mail address: [email protected] (S.A. C. Lima). http://dx.doi.org/10.1016/j.ijpharm.2015.07.013 0378-5173/ ã 2015 Elsevier B.V. All rights reserved.
that occurs after lipid polymorphic transition limitations of SLN. NLC are based on a mixture of solid and liquid lipids that results in an imperfect matrix thus, in an increase of the drug loading. As liquid lipids exhibit higher solubility for drugs, NLC have a higher loading capacity than SLN, as well as an improved drug controlled release (Uner, 2006). Many administration routes are being investigated for lipid nanoparticles (SLN and NLC), including topical, oral and parenteral ones. Also, these lipid colloidal carriers have been proposed for specific applications such as cancer treatment, gene therapy, diagnosis and medical devices production (Carbone et al., 2014). Methotrexate (MTX) has been used in the clinics since the fifteens for the treatment of different solid tumours (e.g. osteosarcoma, lung and breast cancer) (Abolmaali et al., 2013) and in the therapy of autoimmune and inflammatory diseases as rheumatoid arthritis, Crohn’s disease and psoriasis (Braun and Rau, 2009; Swierkot and ski, 2006). MTX is a folate antagonist that competitively Szechin bind to dihydrofolatereductase (DFHR) hampering cell growth and arresting cell cycle in G1/S phases (Genestier et al., 2000). It has been reported that MTX induced apoptosis in several cancer cell lines (Padmanabhan et al., 2009) but its low tumor accumulation results in an ineffective exposure. However, MTX causes toxic side effects to normal cells as well as several adverse effects (hepatotoxicity,
66
M. Ferreira et al. / International Journal of Pharmaceutics 492 (2015) 65–72
ulcerative colitis, nephrotoxicity) that hampers its therapeutic application (Visser and van der Heijde, 2009). To overcome these drawbacks colloidal delivery systems can be developed from biocompatible and biodegradable materials. A formulation design requires full knowledge of the relationship between the process parameters and the quality attributes. To reach an optimized formulation using traditional screening approach (one factor at a time) is difficult, inefficient and time consuming. A few studies (Cun et al., 2011; Liu et al., 2010; Pradhan et al., 2015) have optimized lipid nanoparticles using factorial design and is widely accepted that the ingredients have great influence on the physico-chemical properties (Hao et al., 2011). The statistical formulation design is a validated and useful approach to develop a formulation with less experimentation and providing enough information on the relationship between independent and dependent variables (Gohel and Amin, 1998; Liu et al., 2010). The present work reports the effect of the formulation composition in the optimization of MTX-loaded NLC by means of a Box–Behnken factorial design. Methotrexate was used as a model drug to be incorporated in the NLC due to its wide clinical application. The formulation was produced using the hot ultrasonication method and their physico-chemical properties (morphology, particle size, polydispersity, zeta potential, and entrapment efficiency), in vitro drug release studies and cytotoxicity were investigated. 2. Materials and methods
Table 1 Variables with respective coded levels of the Box–Behnken design. Factors
Coded levels
Independent variables Low Level Medium Level High Level (1) (0) (+1) X1 = liquid lipid (mg) X2 = surfactant (mg) X3 = drug (mg)
40 40 2
50 50 10
60 60 20
Dependent variables
Constrains
Y1 = particle size Y2 = polydispersity index Y3 = entrapment efficiency
Optimum (250 nm) Minimum Maximum
The polynomial equation generated from the experimental design is given below: Y ¼ b0 þ b1 X 1 þ b2 X 2 þ b3 X 3 þ b12 X 1 X 2 þ b13 X 1 X 3 þ b23 X 2 X 3 þ b11 X 12 þ b22 X 22 þ b33 X 32 where Y is the dependent variable, b0 is the intercept, X1, X2, X3 are the coded levels of independent variables, and b1 to b33 are the regression coefficients computed from the observed experimental values of Y; the terms X1X2 and Xi2 (i = 1, 2 or 3) represent the interaction and quadratic terms, respectively. The polynomial equation was statistically validated using ANOVA, by statistical significance of coefficients and r2 values. Statistical analysis was considered significant when the p values were 0.05.
2.1. Materials Methotrexate (MTX) was a kind gift from Excella (Feucht, Germany). The solid lipid, Witepsol1 E85 and the liquid lipid, Miglyol1 812, were acquired from Cremer Oleo (Hamburg, Germany). The surfactant polyvinyl alcohol (PVA) was purchased from Sigma–Aldrich (St Louis, MO, USA). All other reagents and solvents were of analytical reagent grade.
2.2.3. Optimization and validation The graphical and numerical analyses were done by STATISTICA 10 to obtain optimum values of the variables based on the criteria of desirability (Table 1). The optimum variables were used to prepare a checkpoint NLC formulation and were compared with the predicted values to calculate the predicted error, in order to validate the chosen experimental domain and polynomial equations.
2.2. Methods 2.2.1. Preparation of NLCs MTX-loaded NLCs and NLCs were prepared by hot ultrasonication method. Briefly, the lipid phase composed by Witepsol1 E85, Miglyol1 812 and MTX and the aqueous phase containing the surfactant (PVA) in 7 mL of double deionized water were heated to 60 C in a water bath, separately. The aqueous phase was poured into the lipid phase and homogenized using a probe-sonicator (VCX130, Sonics & Materials, 115 Newtown, CT, USA) with amplitude frequency of 70% during 10 min, in order to obtain a nanoemulsion. Blank NLCs were prepared in a similar way, without the drug. Then, formulations cool down at room temperature. 2.2.2. Experimental design The 3-level, 3-factor Box–Behnken design was applied to maximize the experimental efficiency, requiring a minimum of experiments to optimize NLCs produced by hot ultrasonication and study the effects of independent variables on dependent variables (Table 1). Independent variables were amount of liquid lipid (X1), amount of surfactant (X2) and amount of MTX (X3). Other parameters, i.e., amount of solid lipid, sonication time, sonication amplitude, final volume, were set at fixed levels. The established dependent variables were: Y1 = mean particles size; Y2 = polydispersity index and Y3 = encapsulation efficiency. For each factor, the lower (1), medium (0) and higher values (+1) were chosen on the basis of tested lower and upper values for each variable, according to pre-formulation studies and literature research. The data were analyzed using ANOVA by STATISTICA 10 (Statsoft1, Inc.) software.
2.2.4. Entrapment efficiency The entrapment efficiency of MTX within NLCs was determined as described previously (Pinto et al., 2014). Briefly, the nonentrapped drug was quantified at 303 nm, which is the wavelength of maximum absorption of MTX in aqueous solution (Lin et al., 2010). A standard curve of MTX was used to determine the concentration of MTX and the results are expressed as mean standard deviation (n = 3). 2.2.5. Particle size, polydispersity and zeta potential analysis The particle size, polydispersity index (PDI) and zeta potential of all NLCs dispersions were analyzed using a ZetaPALS, ZetaPotential Analyzer (Holtsville, NY, USA). All samples were diluted with double distilled water to reach a suitable concentration before measurement. All analyses were carried out with a fixed light incidence angle of 90 at 25 C. 2.2.6. Transmission electron microscopy (TEM) analysis Optimized NLCs morphology was observed by TEM (TEM Jeol JEM-1400). Images were obtained after one drop of nanoparticles suspension was placed over a grid followed by negative staining with uranyl acetate and placed at the accelerating voltage of 60 kV. 2.2.7. Fourier transform infrared (FT-IR) spectroscopy The freeze-dried optimized formulations of NLCs with and without MTX and pure MTX, as well as physical mixtures, were evaluated using an FT-IR spectrophotometer (FrontierTM, PerkinElmer; Santa Clara, CA, USA) equipped with a horizontal
M. Ferreira et al. / International Journal of Pharmaceutics 492 (2015) 65–72
attenuated total reflectance (ATR) sampling accessory with a diamond crystal. All samples were run in triplicate. Several controls were run in parallel, a background run (to remove the background noise of the instrument) was carried out as a negative control, and MTX as positive controls. The mid-infrared absorbance region was between 4000 and 600 cm1 and the spectra were measured at a spectral resolution of 4 cm1 with 200 scans coadded, to minimize differences between spectra due to baseline shifts. In order to perform the spectra comparison, spectra were truncated at 2000 and 750 cm1, based on the typical absorption bands for the analyzed compounds. 2.2.8. Assessment of storage stability of optimized NLCs NLCs stability during storage was assessed by the parameters: size, PDI and drug content in comparison to the day of production. NLCs dispersions were stored after production in closed glass vials at room temperature and examined on the day of production and after 1, 2, 3 and 4 weeks of storage. 2.2.9. In vitro release study The optimized formulation was subjected to in vitro release studies under three conditions defined to simulate physiological (pH 7.4, 37 0.5 C), inflammatory (pH 5, 37 0.5 C) and topical (pH 5, 32 0.5 C) environments. In vitro release studies were performed using the dialysis bag method, modified to maintain a sink condition and achieve satisfactory reproducibility. 1.5 mL of MTX-loaded NLC dispersion was first poured into the dialysis bag (molecular weight cut off 6000–8000 Da, CelluSep1 T2; Membrane Filtration Products Inc., Frilabo, Portugal) with the two ends fixed by thread and placed into the preheated dissolution media. The suspension was stirred at defined temperature, using a heating and magnetic stirring plate (IKAMAG1, Staufen, Germany) at 350 rpm. 1-mL of sample was withdrawn at fixed time intervals and the same volume of fresh buffer was added accordingly to the condition (phosphate buffer pH 7.4 or acetate buffer pH 5). The drug content determined spectrophotometrically (Jasco V-660 spectrophotometer, USA). The mathematical models for evaluation of drug release kinetics: zero order, first order, Higuchi, Peppas–Korsmeyer and Hixon–Crowell were fitted to the experimental data (Barzegar-Jalali et al., 2008). Regression coefficient (r2) was calculated to determine the best-fit model. 2.2.10. Cell culture and viability assessment Murine fibroblasts L929 from the American Type Culture Collection (Rockville, MD, USA) were cultured in DMEM supplemented with 10% (v/v) fetal bovine serum and 1% (v/v) penicillinstreptomycin. The cells were maintained in a humidified chamber at 37 C and 5% CO2, and the cells were detached using a scraper when reaching 80% confluence. For assessment of the formulations effect on the cell viability a MTT assay was performed. L929 cells were cultured in 96-well plates at 5 105 cells-mL1 density and after 4 h incubated with the NLCs and MTX-loaded NLCs at different concentrations (ranging from 0.1 to 100 mg mL1 in MTX). Empty NLC were added at equivalent lipid concentration to the MTX-loaded NLC (15–250 mg/mL). Upon 24 and 48 h of incubation, the culture medium was removed and replaced by 100 mL of MTT at 0.5 mg/mL in fresh culture medium. The plate was incubated for 3 h at 37 C and formazan crystals were solubilized using 100 mL of dimethyl sulfoxide. The absorbance (590 nm, 630 nm) was read using a SynergyTM HT Multi-mode microplate reader (BioTek Instruments Inc., Winooski, VT, USA). 2.3. Statistical analysis Data is expressed as the mean standard deviation, for a minimum of three independent experiments. Statistical
67
comparisons of the means were performed using one-way analysis of variance or Student’s t-test with Welch’s correction for the in vitro cytotoxicity data, with GraphPad Prism 6 software (La Jolla, California, USA). The differences were considered to be significant when the p-value was <0.05. 3. Results and discussion Initial steps for formulation of MTX-loaded NLCs assessed the drug solubility in the solid lipids. Six lipids with different physicochemical properties were studied (Cetyl Palmitate, Imwitor 308, LipocireTM CM, Witepsol1 E76, Witepsol1 E85, Witepsol1 H32, Witepsol1 S51). MTX solubilized in Cetyl Palmitate, Imwitor 308, LipocireTM CM and Witepsol1 E85 and no drug crystals were observed. So, the solid lipid Witepsol1 E85 with a melting range above body temperature and exhibiting the higher solubility was selected. Preliminary batches of NLCs were prepared to identify possible factors influencing the incorporation of MTX on NLCs, their size and polydispersity. The parameters studied were speed and time of sonication, type and amount of surfactant and optimum lipid loading in 10 mL of dispersion. Based on the preliminary formulation studies, three major variables affecting the particle size and drug EE of NLCs were identified: amount of liquid lipid, amount of surfactant and amount of drug. 3.1. Experimental design Fifteen experimental runs were conducted using Box–Behnken design with a triplicate of the central point for estimation of the experimental error. Three responses were determined as shown in Table 2. Statistical analysis and calculated p-values together with the fitting mathematical model involving the individual main effects and interaction factors are shown in Table 3, with 95% confident level. The positive sign before a factor revels that the response increases whereas the negative sign indicates that the response decreases with the factor. Interaction terms or quadratic relationships are represented by more than one factor or higher-order terms in regressions equations, respectively. It is also suggested non-linearity between factors and responses. A factor can produce a different degree of response when a factor is varied at different levels or more than one factor is varied simultaneously (Rahman et al., 2010). The intercept correspond to the mean of the responses. Analysis of variance for the relevance of the model
Table 2 Formulation composition and the effect on different formulation variables on particle size (Y1), polydispersity index (Y2) and entrapment efficiency (Y3). Sample
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
Factors
Responses
X1 (mg)
X2 (mg)
X3 (mg)
Y1 (nm)
Y2
Y3 (%)
40 60 40 60 40 60 40 60 50 50 50 50 50 50 50
40 40 60 60 50 50 50 50 40 60 40 60 50 50 50
10 10 10 10 2 2 20 20 2 2 20 20 10 10 10
275.9 291.7 273.9 257.4 261.2 243.7 263.4 256.5 257.2 242.3 266.6 255.5 265.5 250.0 269.9
0.079 0.058 0.055 0.084 0.073 0.108 0.089 0.092 0.090 0.062 0.062 0.085 0.088 0.069 0.087
92 90 89 58 65 60 95 94 65 61 95 90 89 88 88
68
M. Ferreira et al. / International Journal of Pharmaceutics 492 (2015) 65–72
Table 3 Summary of results of regression analysis for responses Y1,Y2 and Y3. Parameter
Intercept X1 X12 X2 X22 X3 X32 X1X2 X1X3 X2X3 R2
Size (Y1)
PDI (Y2)
EE (Y3)
Coefficient
p-value
Coefficient
p-value
Coefficient
p-value
262.282 3.009 3.431 7.727 3.031 4.700 6.492 8.075 2.307 1.086 0.863
0.000 0.334 0.157 0.040 0.202 0.155 0.026 0.097 0.585 0.795
0.078 0.005 0.001 0.000 0.007 0.001 0.004 0.013 0.008 0.013 0.876
0.000 0.150 0.726 0.922 0.028 0.850 0.164 0.037 0.148 0.036
80.069 4.799 1.333 5.494 1.708 15.375 4.438 7.250 1.367 0.110 0.940
0.000 0.072 0.427 0.048 0.319 0.001 0.035 0.058 0.664 0.972
are shown in Table 4, where the model is statistically significant when F is higher that Fcritic for all responses. Response surface analysis were plotted (Fig. 1) based on the model polynomial function in a three-dimensional model depicting the effect of significant independent factors on the observed responses of particle size, polydispersity index and entrapment efficiency. 3.1.1. Effect on particle size The particle size ranged from 242.3 nm (sample 10) to 291.7 nm (sample 2), with the selected levels of variables, while the mean was found to be 262.3 nm, which is the intercept of the model (Table 3). The most significant factors affecting particle size are the amount of surfactant (X2) and amount of drug (X32) (p < 0.05). Amount of surfactant had a negative effect on particle size. Higher amounts of surfactant may promote formation and stabilization of smaller particles due to the decrease in interfacial tension between the lipid and the external phase (das Neves and Sarmento, 2015). On the other hand, drug concentration has a positive effect, as expected, due to the rise molecular density in the inner phase. Interaction terms have non-statistically significant effects on Y1. Despite the significant effect of variables on particle size, and a correlation coefficient not as high as expected (R2 0.863), the relatively small ranges of the responses support that the hot ultrasonication method is relatively robust to factor changes. 3.1.2. Effect on polydispersity index The PDI ranged from 0.055 (sample 3) to 0.108 (sample 6) (Table 2), while the mean was 0.078 according to the model intercept (Table 3). The only independent factor that seem to influence on PDI is the amount of surfactant (X2) since all the effects involving this variables were statistically significant (Table 3). The PDI value increases with the increase of surfactant concentration, probably in a non-linear relationship, as the p-value of the linear coefficient (X22) was <0.05. In addition, the interaction
terms involving the amount of surfactant (X1X2 and X2X3) are also significant (p < 0.05). This phenomenon may be explained by the adsorption of PVA onto nanoparticle surface in a concentration dependent way, promoting aggregation due to its adhesive nature (das Neves and Sarmento, 2015). The same behavior may be achieved when two factor are changed simultaneously, i.e., increasing the amount of lipid (X1) or the amount or drug (X3). 3.1.3. Effect on entrapment efficiency The EE varied from 58% (sample 4) to 95% (sample 7) for the level combinations (Table 2), while the mean was 80.1%, according to the model intercept. The independent factors affecting EE were the amount of surfactant (X2) and the amount of drug (X32) (p < 0.05), as seen in Table 3. The value of the correlation coefficient (R2 0.940), indicating a good correlation between observed and predicted value. The amount of surfactant has a negative impact on the EE, due the negative value of the regression coefficient, on the contrary of the amount of drug, which both coefficients (linear and quadratic) presented positive signs. This might indicate a non-linearity correlation between amount of drug and EE. The increase of EE in the presence of higher concentrations of drug was expected, as more drug is available to be entrapped. In the case of the surfactant level, the opposite was found, i.e., which could be explained by the partition phenomena, in which higher concentration of surfactant in the external phase might increase drug partition from internal to external phase, leading to drug solubilization. Interaction terms have non-significant effect on EE. 3.2. Optimization and validation Desirability function of STATISTICA 10 was used to get the optimized formulation. Because the PDI was always below the desired value (<0.2), the formulation optimization were conducted regarding particle size (closer to 250 nm) and entrapment efficiency (maximum). Upon assessment of several responses and comprehensive search through desirability function, the composition of optimized formulation was 45 mg of liquid lipid, 47 mg of surfactant and 18 mg of drug, which fulfill the requirements of optimized formulations. Three replicates of the checkpoint were analyzed. All the responses were considered to be in good agreement with the predicted values (Fig. 2) which confirmed. The optimized formulation has an average size of 252 9 nm and an entrapment efficiency 87 1%, which were in good agreement with the predicted values. 3.3. Physico-chemical characterization of optimized NLCs Unloaded NLCs were produced based on the optimized MTX-loaded NLCs and then physico-chemically characterized for comparison. TEM studies revealed that MTX-loaded NLCs were
Table 4 ANOVA results from particle size, polydispersity index and entrapment efficiency. Response
Source
DF
Sum of squares
Mean of squares
F
Fcritic
Y1
Model Error Cumulative total
9 5 14
19390.700 3147.698 22538.398
1939.072 62.951
30.80
4.77
Y2
Model Error Cumulative total
9 5 14
0.003 0.001 0.003
0.003 0.000
34.19
4.77
Y3
Model Error Cumulative total
9 5 14
2890.592 1761.015 2950.933
2890.592 35.220
82.07
4.77
M. Ferreira et al. / International Journal of Pharmaceutics 492 (2015) 65–72
69
Fig. 1. Response surface and counter plots of size (a), polydispersity index (b) and drug entrapment efficiency (c).
300
100
280
80
260
60
240
40
220
20
200
Predicted Particle Size
Observed
Entrapment Efficiency (%)
Particle Size (nm)
spherical in shape with narrow size distribution (Fig. 3A). NLCs did not aggregated and no visible significant differences were observed between NLCs and MTX-loaded NLCs. The diameters of the particles observed by transmission microscopy (c.a. 250 nm) are in good agreement with the data obtained from dynamic light scattering. Production of nano-sized unloaded and MTX-loaded NLCs was confirmed by dynamic light scattering analyzes. The mean hydrodynamic diameters and particle size distribution (e.g.
0
Encapsultion efficiency
Fig. 2. Validation of the predicted optimal results with experimental values. (n = 3)
polydispersity index, PDI) of the formulations prepared by hot ultrasonication method are presented in Table 5. The results indicate that NLC size (246 2 nm) increase with the incorporation of MTX to 252 9 nm (p < 0.05) remaining homogenous, as suggested by the low values obtained for the PDI (<0.1). The optimized formulations exhibited a particle size below 300 nm, suitable for systemic and topical administrations (Albanese et al., 2012; Kohli and Alpar, 2004). It has been described that PDI values below 0.2 are indicative of homogeneity of the size distribution and with minimum tendency to aggregation (Mitri et al., 2011). In order to confirm the incorporation of MTX in the NLCs infrared spectra of free MTX, NLC and MTX-loaded NLC were obtained, and are presented on Fig. 3B. MTX spectrum shows, at 1638 cm1, a marked peak that indicates the presence of a C¼C stretching vibration, which is characteristic of the drug molecule (Khan et al., 2010; Kohler et al., 2005). This observation was present also in the MTX-loaded NLC, but not in the NLC, confirming the successful incorporation of MTX in the lipid nanoparticles. The characteristic peaks from NLC were not altered in the MTX-loaded NLC. 3.4. Storage stability assessment Physical stability was assessed by analyzing changes in particle size, polydispersity index and drug content of NLCs stored at room temperature. Changes in particle size are usually accepted as
70
M. Ferreira et al. / International Journal of Pharmaceutics 492 (2015) 65–72
Fig. 3. Characterization of the optimized NLCs. (A) Transmission electron microscopy (TEM) images of NLCs (top) and MTX-loaded NLCs (bottom). The scale indicated below the pictures is 200 nm. Amplification: 80,000. (B) FT-IR spectra of raw MTX powder and lyophilized nanoparticles NLCs and MTX-loaded NLCs.
Table 5 Particle size, PDI, z-potential, drug EE of optimized NLCs.
NLC MTX loaded-NLC
Size (nm)
PDI
z-potential (mV)
EE (%)
246 2 252 9
0.09 0.02 0.06 0.02
13 2 14 1
87 1
PDI, polydispersion index; EE, entrapment efficiency. Mean SD (n = 3).
indicator of formulation instability (Heurtault et al., 2003). For all samples no particle aggregation was found through visual observations, up to 4 weeks. The stability studies of MTX-loaded NLCs suggested that these nanoparticles were quite stable for a month with no significant change in the mean particle size and drug content (Fig. 4). As a broad conclusion, NLCs presented good stability after 4 weeks with average particle size between 212 and 264 nm, PDI below 0.2 and EE remains higher than 85%.
to simulate physiological, inflammatory and topical environments. In Fig. 5 shows the release profiles obtained that were further analyzed to determine the mechanism of release using the kinetic models first-order, Higuchi, Peppas–Korsmeyer and Hixson–Crowell. 3.5.1. Release simulation at physiological conditions In vitro MTX release studies from the NLCs were performed at physiological conditions (37 C, pH 7.4) to access the release profile upon systemic administration. The NLC formulations showed an initial fast release, the release reach 30% in 2 h, followed by a sustained release up to 50% within 24 h (Fig. 5). The observed profile was analyzed by several drug release kinetic models, and fitted best to the Peppas–Korsemeyer model (r2 = 0.9771), since the r2 value is much higher than any other kinetic model. In this model, the value of n characterizes the release mechanism of drug. For the present case n was 0.55, indicative of non Fickian diffusion, as a combination of both diffusion and erosion controlled rate release (Dash et al., 2010).
3.5. In vitro methotrexate release studies The in vitro of MTX release profile from the NLCs was investigated using a dialysis membrane in three conditions defined
3.5.2. Release simulation at inflammatory conditions MTX is considered a reference drug for rheumatoid arthritis. We performed an in vitro release study simulating the inflammatory
Fig. 4. Storage stability assessment. Optimized MTX-loaded NLC were stored up to 4 weeks at room temperature and evaluated for (A) size, (B) polydispersity and (C) drug entrapment efficiency alterations. Data expressed as mean SD (n = 3).
M. Ferreira et al. / International Journal of Pharmaceutics 492 (2015) 65–72
71
Fig. 5. MTX in vitro release from NLCs. On the left, drug cumulative release under physiologic (dark circles) and inflammatory environments (open squares). On the right, release simulation at skin environment. Data expressed as mean SD (n = 2).
Fig. 6. Fibroblasts viability. Unloaded and MTX-loaded NLCs were incubated with fibroblasts for (A) 24 h and (B) 48 h. Data expressed as mean SD (n = 3).
conditions (37 C, pH 5) in order to predict the MTX-loaded NLC in vivo kinetics (Müller et al., 2000). The drug release from the formulation under this conditions exhibited a fast initial release (40% in 2 h) followed by a sustained release up to 80% within 24 h (Fig. 5). Data fitted best to the Peppas–Korsemeyer model (r2 = 0.9865) with an n value of 0.31, suggesting Fickian diffusion that occurs by the usual molecular diffusion of the drug. 3.5.3. Release simulation at skin conditions Topical administration of MTX could represent an interesting approach for skin inflammatory diseases, as psoriasis. Release conditions were set to simulate the drug administration through the skin (32 C, pH 5). A slow release of the MTX from the NLC was observed reaching 30% after 24 h (Fig. 5). In this case, data fitted best to the Hixson–Crowell model (r2 = 0.9290), that describes the drug releases by dissolution mechanism that occurs upon a change in surface area and diameter of particles (Dash et al., 2010). A prolonged release is of interest for dermal formulations as it will contribute to a sustained effect (Pardeike et al., 2009). Gathering the release studies data it is possible to consider the NLCs formulations suitable for systemic and topical administration of MTX under physiological and inflammatory conditions, conferring protection and allowing drug controlled release. 3.6. Cytotoxicity of the optimized formulation To assess the effect of the formulations on cell viability, a MTT assay was performed on L929 fibroblasts. Cells were exposed to empty NLCs and MTX-loaded NLCs up to 250 mg mL1 of MTX equivalent to 6.5 mg mL1 in lipid for 24 and 48 h. The fibroblasts tolerate well empty formulation as no toxicity was observed (Fig. 6). While for MTX-loaded NLCs a slight effect on the viability of cells was observed in the presence of 250 mg.mL1 MTX-loaded NLCs after 48 h incubation. As for all studied conditions cell viability was above 80% it can be stated that the formulations are not toxic to the fibroblasts. The NLC system represents a promising option for delivery of MTX, since it did not affect cell viability, confirming the safety profile of the optimized delivery system.
4. Conclusion MTX loaded NLCs were successfully developed by hot ultrasonication method and optimized using 3- factor, 3-level Box–Behnken design. Based on all data achieved from the formulation design it is possible to conclude that, in general, the amount of surfactant and drug used were important factors for the production of MTX-loaded NLCs influencing their physical properties. This factorial design study has proven to be a useful tool in optimizing NLC for delivery of MTX. The optimized NLCs showed size of 252 9 nm with a polydispersity around 0.08 and an entrapment efficiency of 87%. FTIR analysis enabled confirmation of the efficient entrapment of MTX in the NLCs. The stability tests during 4 weeks at room temperature give a good indication of that at least in solution, the formulation maintain their initial properties. In vitro release profile of MTX from NLCs revealed to be biphasic under physiological and inflammatory environments while sustained at the skin environment. Thus, the optimized formulation revealed suitable for systemic and topical administration of MTX. No toxicity was observed in fibroblasts upon 24 and 48 h of contact with the optimized formulations, indicative of their safe application. This lipid colloidal carrier show interesting properties for delivering methotrexate. Future studies could focus on the evaluation of the therapeutic potential of the optimized formulation in clinically relevant models. Present research demonstrated that MTX loaded NLCs could be a promising tool for systemic and topical administration in the therapy of cancer and inflammatory diseases, such as psoriasis and rheumatoid arthritis. Acknowledgments This work received financial support from the European Union (FEDER funds through COMPETE) and National Funds (FCT, Fundação para a Ciência e Tecnologia) through project UID/Multi/04378/2013. This work was also funded by ON.2 QREN – Quadro de Referência Estratégico Nacional – QREN, by
72
M. Ferreira et al. / International Journal of Pharmaceutics 492 (2015) 65–72
FEDER funds through the Programa Operacional Factores de Competitividade – COMPETE and national funds through FCT (Portugal) through project NORTE-07-0124-FEDER-000067. Luíse Chaves thanks the CAPES Foundation, Ministry of Education of Brazil for the Doctoral fellowship 0831-12-3. The authors would like to acknowledge Excella for kindly provide the MTX. References Abolmaali, S.S., Tamaddon, A.M., Dinarvand, R., 2013. A review of therapeutic challenges and achievements of methotrexate delivery systems for treatment of cancer and rheumatoid arthritis. Cancer Chemother. Pharmacol. 71, 1115–1130. Albanese, A., Tang, P.S., Chan, W.C., 2012. The effect of nanoparticle size shape, and surface chemistry on biological systems. Annu. Rev. Biomed. Eng. 14, 1–16. Barzegar-Jalali, M., Adibkia, K., Valizadeh, H., Shadbad, M., Nokhodchi, A., Omidi, Y., Hasan, M., 2008. Kinetic analysis of drug release from nanoparticles. J. Pharm. Pharm. Sci. 11, 167–177. Braun, J., Rau, R., 2009. An update on methotrexate. Curr. Opin. Rheumatol. 21, 216–223. Carbone, C., Leonardi, A., Cupri, S., Puglisi, G., Pignatello, R., 2014. Pharmaceutical and biomedical applications of lipid-based nanocarriers. Pharm. Pat. Anal. 3, 199–215. Cun, D., Jensen, D.K., Maltesen, M.J., Bunker, M., Whiteside, P., Scurr, D., Nielsen, H. M., 2011. High loading efficiency and sustained release of siRNA encapsulated in PLGA nanoparticles: quality by design optimization and characterization. Eur. J. Pharm. Biopharm. 77, 26–35. das Neves, J., Sarmento, B., 2015. Precise engineering of dapivirine-loaded nanoparticles for the development of anti-HIV vaginal microbicides. Acta Biomater. doi:http://dx.doi.org/10.1016/j.actbio.2015.02.007. Dash, S., Murthy, P.N., Nath, L., Chowdhury, P., 2010. Kinetic modeling on drug release from controlled drug delivery systems. Acta Pol. Pharm. 67, 217–223. Genestier, L., Paillot, R., Quemeneur, L., Izeradjene, K., Revillard, J.P., 2000. Mechanism of action of methotrexate. Immunol. Pharmacol. 47, 247–257. Gohel, M.C., Amin, A.F., 1998. Formulation optimization of controlled release diclofenac sodium microspheres using factorial design. J. Controlled Release 51, 115–122. Hao, J., Fang, X., Zhou, Y., Wang, J., Guo, F., Li, F., Peng, X., 2011. Development and optimization of solid lipid nanoparticle formulation for ophthalmic delivery of chloramphenicol using a Box–Behnken design. Int. J. Nanomed. 6, 683–692. Heurtault, B., Saulnier, P., Pech, B., Proust, J.E., Benoit, J.P., 2003. Physico-chemical stability of colloidal lipid particles. Biomaterials 24, 4283–4300. Huang, Z., Hua, S., Yang, Y., Fang, J., 2008. Development and evaluation of lipid nanoparticles for camptothecin delivery: a comparison of solid lipid
nanoparticles, nanostructured lipid carriers, and lipid emulsion. Acta Pharmacol. Sin. 29, 1094–1102. Khan, M.S., Priyadarshini, M., Sumbul, S., Bano, B., 2010. Methotrexate binding causes structural and functional changes in lung cystatin. Acta Biochim. Pol. 57, 499–503. Kohler, N., Sun, C., Wang, J., Zhang, M., 2005. Methotrexate-modified superparamagnetic nanoparticles and their intracellular uptake into human cancer cells. Langmuir 21, 8858–8864. Kohli, A.K., Alpar, H.O., 2004. Potential use of nanoparticles for transcutaneous vaccine delivery: effect of particle size and charge. Int. J. Pharm. 275, 13–17. Lin, Y.-K., Huang, Z.-R., Zhuo, R.-Z., Fang, J.-Y., 2010. Combination of calcipotriol and methotrexate in nanostructured lipid carriers for topical delivery. Int. J. Nanomed. 5, 117–128. Liu, C.H., Wu, C.T., Fang, J.Y., 2010. Characterization and formulation optimization of solid lipid nanoparticles in vitamin K1 delivery. Drug Dev. Ind. Pharm. 36, 751–761. Marianecci, C., Di Marzio, L., Rinaldi, F., Celia, C., Paolino, D., Alhaique, F., Carafa, M., 2014. Niosomes from 80s to present: the state of the art. Adv. Colloid Interface Sci. 205, 187–206. Mitri, K., Shegokar, R., Gohla, S., Anselmi, C., Müller, R.H., 2011. Lipid nanocarriers for dermal delivery of lutein: preparation, characterization, stability and performance. Int. J. Pharm. 414, 267–275. Müller, R.H., Mäder, K., Gohla, S., 2000. Solid lipid nanoparticles (SLN) for controlled drug delivery – a review of the state of the art. Eur. J. Pharm. Biopharm. 50,161–177. Padmanabhan, S., Tripathi, D.N., Vikram, A., Ramarao, P., Jena, G.B., 2009. Methotrexate-induced cytotoxicity and genotoxicity in germ cells of mice: intervention of folic and folinic acid. Mutat. Res. 673, 43–52. Pardeike, J., Hommoss, A., Müller, R.H., 2009. Lipid nanoparticles (SLN, NLC) in cosmetic and pharmaceutical dermal products. Int. J. Pharm. 366, 170–184. Pinto, M.F., Moura, C.C., Nunes, C., Segundo, M.A., Costa Lima, S.A., Reis, S., 2014. A new topical formulation for psoriasis: development of methotrexate-loaded nanostructured lipid carriers. Int. J. Pharm. 477, 519–526. Pradhan, M., Singh, D., Singh, M.R., 2015. Development characterization and skin permeating potential of lipid based novel delivery system for topical treatment of psoriasis. Chem. Phys. Lipids 186, 9–16. Rahman, Z., Zidan, A.S., Khan, M.A., 2010. Non-destructive methods of characterization of risperidone solid lipid nanoparticles. Eur. J. Pharm. Biopharm. 76 (1), 127–137. doi:http://dx.doi.org/10.1016/j.ejpb.2010.05.003. ski, J., 2006. Methotrexate in rheumatoid arthritis. Pharmacol. Swierkot, J., Szechin Rep. 58 (4), 473–492. Uner, M., 2006. Preparation, characterization and physico-chemical properties of solid lipid nanoparticles (SLN) and nanostructured lipid carriers (NLC): their benefits as colloidal drug carrier systems. Pharmazie 61 (5), 375–386. Visser, K., van der Heijde, D.M., 2009. Risk and management of liver toxicity during methotrexate treatment in rheumatoid and psoriatic arthritis: a systematic review of the literature. Clin. Exp. Rheumatol. 27 (6), 1017–1025.
Contents lists available at ScienceDirect
International Journal of Pharmaceutics journal homepage: www.elsevier.com/locate/ijpharm
Optimization of nanostructured lipid carriers loaded with methotrexate: A tool for inflammatory and cancer therapy Mara Ferreiraa,b , Luíse L. Chavesa , Sofia A. Costa Limaa,* , Salette Reisa a b
UCIBIO/REQUIMTE, Department of Chemistry, Faculty of Pharmacy, University of Porto, Portugal Faculty of Engineering of University of Porto, Portugal
A R T I C L E I N F O
A B S T R A C T
Article history: Received 13 May 2015 Accepted 5 July 2015 Available online 10 July 2015
The aim of this study was to optimize and assess the potential of nanostructured lipid carriers (NLC), prepared by the hot ultrasonication method, as carrier for methotrexate (MTX), highlighting the application of factorial design. Preliminary screening drug/lipid solubility, allowed us to select Witepsol1 E85 as the solid lipid and Mygliol1 812 as liquid lipid for the NLC loaded with MTX. Then, a 3-level, 3-factor Box-Behnken design and validated by ANOVA analysis; the correspondence between the predicted values and those measured experimentally confirmed the robustness of the design. Properties of optimized MTX-loaded NLCs such as morphology, size, zeta potential, entrapment efficiency, storage stability, in vitro drug release and cytotoxicity were investigated. NLCs loaded with MTX exhibited spherical shape with 252-nm, a polydispersity of 0.06 0.02, zeta potential of 14 mV and an entrapment efficiency of 87%. In vitro release studies revealed a fast initial release followed by a prolonged release of MTX from the NLC up to 24-h. The release kinetics of the optimized NLC best fitted the Peppas–Korsmeyer model for physiological and inflammatory environments and the Hixson–Crowell model skin simulation conditions. No toxicity was observed in fibroblasts. Thus, the optimized MTX-loaded NLC have the potential to be exploited as delivery system. ã 2015 Elsevier B.V. All rights reserved.
Keywords: Lipid colloidal carriers Methotrexate Witepsol1 E85 Mygliol1 812 Box–Behnken design Hot ultrasonication In vitro drug release Storage stability L929 fibroblasts
1. Introduction Lipid based colloidal carriers have been widely used for drug delivery because they offer the possibility of modulating drug release, by facilitating its transport to the different targets, by increasing local penetration, prolonging residence time and by a controlled release mechanism to provide an effective dose to the target site (Marianecci et al., 2014). Among these lipid colloidal delivery systems (e.g. lipid nanoparticles, liposomes, and nanoemulsions) solid lipid nanoparticles (SLN) emerged in the early nineties as an alternative and current trials applying SLN consider them very promising in drug delivery (Müller et al., 2000). In fact, it is widely accepted that SLN combine the advantages and avoid the disadvantages of other colloidal carriers (Müller et al., 2000). Low drug loading and drug expulsion during storage period are taken as the disadvantage of SLN (Huang et al., 2008). Nanostructure lipid carriers (NLC) have been developed to overcome the drug loading capacity and drug leaking phenomena
* Corresponding author at: Rua de Jorge Viterbo Ferreira, 228, 4050-313 Porto, Portugal. Fax: +351 226093483. E-mail address: [email protected] (S.A. C. Lima). http://dx.doi.org/10.1016/j.ijpharm.2015.07.013 0378-5173/ ã 2015 Elsevier B.V. All rights reserved.
that occurs after lipid polymorphic transition limitations of SLN. NLC are based on a mixture of solid and liquid lipids that results in an imperfect matrix thus, in an increase of the drug loading. As liquid lipids exhibit higher solubility for drugs, NLC have a higher loading capacity than SLN, as well as an improved drug controlled release (Uner, 2006). Many administration routes are being investigated for lipid nanoparticles (SLN and NLC), including topical, oral and parenteral ones. Also, these lipid colloidal carriers have been proposed for specific applications such as cancer treatment, gene therapy, diagnosis and medical devices production (Carbone et al., 2014). Methotrexate (MTX) has been used in the clinics since the fifteens for the treatment of different solid tumours (e.g. osteosarcoma, lung and breast cancer) (Abolmaali et al., 2013) and in the therapy of autoimmune and inflammatory diseases as rheumatoid arthritis, Crohn’s disease and psoriasis (Braun and Rau, 2009; Swierkot and ski, 2006). MTX is a folate antagonist that competitively Szechin bind to dihydrofolatereductase (DFHR) hampering cell growth and arresting cell cycle in G1/S phases (Genestier et al., 2000). It has been reported that MTX induced apoptosis in several cancer cell lines (Padmanabhan et al., 2009) but its low tumor accumulation results in an ineffective exposure. However, MTX causes toxic side effects to normal cells as well as several adverse effects (hepatotoxicity,
66
M. Ferreira et al. / International Journal of Pharmaceutics 492 (2015) 65–72
ulcerative colitis, nephrotoxicity) that hampers its therapeutic application (Visser and van der Heijde, 2009). To overcome these drawbacks colloidal delivery systems can be developed from biocompatible and biodegradable materials. A formulation design requires full knowledge of the relationship between the process parameters and the quality attributes. To reach an optimized formulation using traditional screening approach (one factor at a time) is difficult, inefficient and time consuming. A few studies (Cun et al., 2011; Liu et al., 2010; Pradhan et al., 2015) have optimized lipid nanoparticles using factorial design and is widely accepted that the ingredients have great influence on the physico-chemical properties (Hao et al., 2011). The statistical formulation design is a validated and useful approach to develop a formulation with less experimentation and providing enough information on the relationship between independent and dependent variables (Gohel and Amin, 1998; Liu et al., 2010). The present work reports the effect of the formulation composition in the optimization of MTX-loaded NLC by means of a Box–Behnken factorial design. Methotrexate was used as a model drug to be incorporated in the NLC due to its wide clinical application. The formulation was produced using the hot ultrasonication method and their physico-chemical properties (morphology, particle size, polydispersity, zeta potential, and entrapment efficiency), in vitro drug release studies and cytotoxicity were investigated. 2. Materials and methods
Table 1 Variables with respective coded levels of the Box–Behnken design. Factors
Coded levels
Independent variables Low Level Medium Level High Level (1) (0) (+1) X1 = liquid lipid (mg) X2 = surfactant (mg) X3 = drug (mg)
40 40 2
50 50 10
60 60 20
Dependent variables
Constrains
Y1 = particle size Y2 = polydispersity index Y3 = entrapment efficiency
Optimum (250 nm) Minimum Maximum
The polynomial equation generated from the experimental design is given below: Y ¼ b0 þ b1 X 1 þ b2 X 2 þ b3 X 3 þ b12 X 1 X 2 þ b13 X 1 X 3 þ b23 X 2 X 3 þ b11 X 12 þ b22 X 22 þ b33 X 32 where Y is the dependent variable, b0 is the intercept, X1, X2, X3 are the coded levels of independent variables, and b1 to b33 are the regression coefficients computed from the observed experimental values of Y; the terms X1X2 and Xi2 (i = 1, 2 or 3) represent the interaction and quadratic terms, respectively. The polynomial equation was statistically validated using ANOVA, by statistical significance of coefficients and r2 values. Statistical analysis was considered significant when the p values were 0.05.
2.1. Materials Methotrexate (MTX) was a kind gift from Excella (Feucht, Germany). The solid lipid, Witepsol1 E85 and the liquid lipid, Miglyol1 812, were acquired from Cremer Oleo (Hamburg, Germany). The surfactant polyvinyl alcohol (PVA) was purchased from Sigma–Aldrich (St Louis, MO, USA). All other reagents and solvents were of analytical reagent grade.
2.2.3. Optimization and validation The graphical and numerical analyses were done by STATISTICA 10 to obtain optimum values of the variables based on the criteria of desirability (Table 1). The optimum variables were used to prepare a checkpoint NLC formulation and were compared with the predicted values to calculate the predicted error, in order to validate the chosen experimental domain and polynomial equations.
2.2. Methods 2.2.1. Preparation of NLCs MTX-loaded NLCs and NLCs were prepared by hot ultrasonication method. Briefly, the lipid phase composed by Witepsol1 E85, Miglyol1 812 and MTX and the aqueous phase containing the surfactant (PVA) in 7 mL of double deionized water were heated to 60 C in a water bath, separately. The aqueous phase was poured into the lipid phase and homogenized using a probe-sonicator (VCX130, Sonics & Materials, 115 Newtown, CT, USA) with amplitude frequency of 70% during 10 min, in order to obtain a nanoemulsion. Blank NLCs were prepared in a similar way, without the drug. Then, formulations cool down at room temperature. 2.2.2. Experimental design The 3-level, 3-factor Box–Behnken design was applied to maximize the experimental efficiency, requiring a minimum of experiments to optimize NLCs produced by hot ultrasonication and study the effects of independent variables on dependent variables (Table 1). Independent variables were amount of liquid lipid (X1), amount of surfactant (X2) and amount of MTX (X3). Other parameters, i.e., amount of solid lipid, sonication time, sonication amplitude, final volume, were set at fixed levels. The established dependent variables were: Y1 = mean particles size; Y2 = polydispersity index and Y3 = encapsulation efficiency. For each factor, the lower (1), medium (0) and higher values (+1) were chosen on the basis of tested lower and upper values for each variable, according to pre-formulation studies and literature research. The data were analyzed using ANOVA by STATISTICA 10 (Statsoft1, Inc.) software.
2.2.4. Entrapment efficiency The entrapment efficiency of MTX within NLCs was determined as described previously (Pinto et al., 2014). Briefly, the nonentrapped drug was quantified at 303 nm, which is the wavelength of maximum absorption of MTX in aqueous solution (Lin et al., 2010). A standard curve of MTX was used to determine the concentration of MTX and the results are expressed as mean standard deviation (n = 3). 2.2.5. Particle size, polydispersity and zeta potential analysis The particle size, polydispersity index (PDI) and zeta potential of all NLCs dispersions were analyzed using a ZetaPALS, ZetaPotential Analyzer (Holtsville, NY, USA). All samples were diluted with double distilled water to reach a suitable concentration before measurement. All analyses were carried out with a fixed light incidence angle of 90 at 25 C. 2.2.6. Transmission electron microscopy (TEM) analysis Optimized NLCs morphology was observed by TEM (TEM Jeol JEM-1400). Images were obtained after one drop of nanoparticles suspension was placed over a grid followed by negative staining with uranyl acetate and placed at the accelerating voltage of 60 kV. 2.2.7. Fourier transform infrared (FT-IR) spectroscopy The freeze-dried optimized formulations of NLCs with and without MTX and pure MTX, as well as physical mixtures, were evaluated using an FT-IR spectrophotometer (FrontierTM, PerkinElmer; Santa Clara, CA, USA) equipped with a horizontal
M. Ferreira et al. / International Journal of Pharmaceutics 492 (2015) 65–72
attenuated total reflectance (ATR) sampling accessory with a diamond crystal. All samples were run in triplicate. Several controls were run in parallel, a background run (to remove the background noise of the instrument) was carried out as a negative control, and MTX as positive controls. The mid-infrared absorbance region was between 4000 and 600 cm1 and the spectra were measured at a spectral resolution of 4 cm1 with 200 scans coadded, to minimize differences between spectra due to baseline shifts. In order to perform the spectra comparison, spectra were truncated at 2000 and 750 cm1, based on the typical absorption bands for the analyzed compounds. 2.2.8. Assessment of storage stability of optimized NLCs NLCs stability during storage was assessed by the parameters: size, PDI and drug content in comparison to the day of production. NLCs dispersions were stored after production in closed glass vials at room temperature and examined on the day of production and after 1, 2, 3 and 4 weeks of storage. 2.2.9. In vitro release study The optimized formulation was subjected to in vitro release studies under three conditions defined to simulate physiological (pH 7.4, 37 0.5 C), inflammatory (pH 5, 37 0.5 C) and topical (pH 5, 32 0.5 C) environments. In vitro release studies were performed using the dialysis bag method, modified to maintain a sink condition and achieve satisfactory reproducibility. 1.5 mL of MTX-loaded NLC dispersion was first poured into the dialysis bag (molecular weight cut off 6000–8000 Da, CelluSep1 T2; Membrane Filtration Products Inc., Frilabo, Portugal) with the two ends fixed by thread and placed into the preheated dissolution media. The suspension was stirred at defined temperature, using a heating and magnetic stirring plate (IKAMAG1, Staufen, Germany) at 350 rpm. 1-mL of sample was withdrawn at fixed time intervals and the same volume of fresh buffer was added accordingly to the condition (phosphate buffer pH 7.4 or acetate buffer pH 5). The drug content determined spectrophotometrically (Jasco V-660 spectrophotometer, USA). The mathematical models for evaluation of drug release kinetics: zero order, first order, Higuchi, Peppas–Korsmeyer and Hixon–Crowell were fitted to the experimental data (Barzegar-Jalali et al., 2008). Regression coefficient (r2) was calculated to determine the best-fit model. 2.2.10. Cell culture and viability assessment Murine fibroblasts L929 from the American Type Culture Collection (Rockville, MD, USA) were cultured in DMEM supplemented with 10% (v/v) fetal bovine serum and 1% (v/v) penicillinstreptomycin. The cells were maintained in a humidified chamber at 37 C and 5% CO2, and the cells were detached using a scraper when reaching 80% confluence. For assessment of the formulations effect on the cell viability a MTT assay was performed. L929 cells were cultured in 96-well plates at 5 105 cells-mL1 density and after 4 h incubated with the NLCs and MTX-loaded NLCs at different concentrations (ranging from 0.1 to 100 mg mL1 in MTX). Empty NLC were added at equivalent lipid concentration to the MTX-loaded NLC (15–250 mg/mL). Upon 24 and 48 h of incubation, the culture medium was removed and replaced by 100 mL of MTT at 0.5 mg/mL in fresh culture medium. The plate was incubated for 3 h at 37 C and formazan crystals were solubilized using 100 mL of dimethyl sulfoxide. The absorbance (590 nm, 630 nm) was read using a SynergyTM HT Multi-mode microplate reader (BioTek Instruments Inc., Winooski, VT, USA). 2.3. Statistical analysis Data is expressed as the mean standard deviation, for a minimum of three independent experiments. Statistical
67
comparisons of the means were performed using one-way analysis of variance or Student’s t-test with Welch’s correction for the in vitro cytotoxicity data, with GraphPad Prism 6 software (La Jolla, California, USA). The differences were considered to be significant when the p-value was <0.05. 3. Results and discussion Initial steps for formulation of MTX-loaded NLCs assessed the drug solubility in the solid lipids. Six lipids with different physicochemical properties were studied (Cetyl Palmitate, Imwitor 308, LipocireTM CM, Witepsol1 E76, Witepsol1 E85, Witepsol1 H32, Witepsol1 S51). MTX solubilized in Cetyl Palmitate, Imwitor 308, LipocireTM CM and Witepsol1 E85 and no drug crystals were observed. So, the solid lipid Witepsol1 E85 with a melting range above body temperature and exhibiting the higher solubility was selected. Preliminary batches of NLCs were prepared to identify possible factors influencing the incorporation of MTX on NLCs, their size and polydispersity. The parameters studied were speed and time of sonication, type and amount of surfactant and optimum lipid loading in 10 mL of dispersion. Based on the preliminary formulation studies, three major variables affecting the particle size and drug EE of NLCs were identified: amount of liquid lipid, amount of surfactant and amount of drug. 3.1. Experimental design Fifteen experimental runs were conducted using Box–Behnken design with a triplicate of the central point for estimation of the experimental error. Three responses were determined as shown in Table 2. Statistical analysis and calculated p-values together with the fitting mathematical model involving the individual main effects and interaction factors are shown in Table 3, with 95% confident level. The positive sign before a factor revels that the response increases whereas the negative sign indicates that the response decreases with the factor. Interaction terms or quadratic relationships are represented by more than one factor or higher-order terms in regressions equations, respectively. It is also suggested non-linearity between factors and responses. A factor can produce a different degree of response when a factor is varied at different levels or more than one factor is varied simultaneously (Rahman et al., 2010). The intercept correspond to the mean of the responses. Analysis of variance for the relevance of the model
Table 2 Formulation composition and the effect on different formulation variables on particle size (Y1), polydispersity index (Y2) and entrapment efficiency (Y3). Sample
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
Factors
Responses
X1 (mg)
X2 (mg)
X3 (mg)
Y1 (nm)
Y2
Y3 (%)
40 60 40 60 40 60 40 60 50 50 50 50 50 50 50
40 40 60 60 50 50 50 50 40 60 40 60 50 50 50
10 10 10 10 2 2 20 20 2 2 20 20 10 10 10
275.9 291.7 273.9 257.4 261.2 243.7 263.4 256.5 257.2 242.3 266.6 255.5 265.5 250.0 269.9
0.079 0.058 0.055 0.084 0.073 0.108 0.089 0.092 0.090 0.062 0.062 0.085 0.088 0.069 0.087
92 90 89 58 65 60 95 94 65 61 95 90 89 88 88
68
M. Ferreira et al. / International Journal of Pharmaceutics 492 (2015) 65–72
Table 3 Summary of results of regression analysis for responses Y1,Y2 and Y3. Parameter
Intercept X1 X12 X2 X22 X3 X32 X1X2 X1X3 X2X3 R2
Size (Y1)
PDI (Y2)
EE (Y3)
Coefficient
p-value
Coefficient
p-value
Coefficient
p-value
262.282 3.009 3.431 7.727 3.031 4.700 6.492 8.075 2.307 1.086 0.863
0.000 0.334 0.157 0.040 0.202 0.155 0.026 0.097 0.585 0.795
0.078 0.005 0.001 0.000 0.007 0.001 0.004 0.013 0.008 0.013 0.876
0.000 0.150 0.726 0.922 0.028 0.850 0.164 0.037 0.148 0.036
80.069 4.799 1.333 5.494 1.708 15.375 4.438 7.250 1.367 0.110 0.940
0.000 0.072 0.427 0.048 0.319 0.001 0.035 0.058 0.664 0.972
are shown in Table 4, where the model is statistically significant when F is higher that Fcritic for all responses. Response surface analysis were plotted (Fig. 1) based on the model polynomial function in a three-dimensional model depicting the effect of significant independent factors on the observed responses of particle size, polydispersity index and entrapment efficiency. 3.1.1. Effect on particle size The particle size ranged from 242.3 nm (sample 10) to 291.7 nm (sample 2), with the selected levels of variables, while the mean was found to be 262.3 nm, which is the intercept of the model (Table 3). The most significant factors affecting particle size are the amount of surfactant (X2) and amount of drug (X32) (p < 0.05). Amount of surfactant had a negative effect on particle size. Higher amounts of surfactant may promote formation and stabilization of smaller particles due to the decrease in interfacial tension between the lipid and the external phase (das Neves and Sarmento, 2015). On the other hand, drug concentration has a positive effect, as expected, due to the rise molecular density in the inner phase. Interaction terms have non-statistically significant effects on Y1. Despite the significant effect of variables on particle size, and a correlation coefficient not as high as expected (R2 0.863), the relatively small ranges of the responses support that the hot ultrasonication method is relatively robust to factor changes. 3.1.2. Effect on polydispersity index The PDI ranged from 0.055 (sample 3) to 0.108 (sample 6) (Table 2), while the mean was 0.078 according to the model intercept (Table 3). The only independent factor that seem to influence on PDI is the amount of surfactant (X2) since all the effects involving this variables were statistically significant (Table 3). The PDI value increases with the increase of surfactant concentration, probably in a non-linear relationship, as the p-value of the linear coefficient (X22) was <0.05. In addition, the interaction
terms involving the amount of surfactant (X1X2 and X2X3) are also significant (p < 0.05). This phenomenon may be explained by the adsorption of PVA onto nanoparticle surface in a concentration dependent way, promoting aggregation due to its adhesive nature (das Neves and Sarmento, 2015). The same behavior may be achieved when two factor are changed simultaneously, i.e., increasing the amount of lipid (X1) or the amount or drug (X3). 3.1.3. Effect on entrapment efficiency The EE varied from 58% (sample 4) to 95% (sample 7) for the level combinations (Table 2), while the mean was 80.1%, according to the model intercept. The independent factors affecting EE were the amount of surfactant (X2) and the amount of drug (X32) (p < 0.05), as seen in Table 3. The value of the correlation coefficient (R2 0.940), indicating a good correlation between observed and predicted value. The amount of surfactant has a negative impact on the EE, due the negative value of the regression coefficient, on the contrary of the amount of drug, which both coefficients (linear and quadratic) presented positive signs. This might indicate a non-linearity correlation between amount of drug and EE. The increase of EE in the presence of higher concentrations of drug was expected, as more drug is available to be entrapped. In the case of the surfactant level, the opposite was found, i.e., which could be explained by the partition phenomena, in which higher concentration of surfactant in the external phase might increase drug partition from internal to external phase, leading to drug solubilization. Interaction terms have non-significant effect on EE. 3.2. Optimization and validation Desirability function of STATISTICA 10 was used to get the optimized formulation. Because the PDI was always below the desired value (<0.2), the formulation optimization were conducted regarding particle size (closer to 250 nm) and entrapment efficiency (maximum). Upon assessment of several responses and comprehensive search through desirability function, the composition of optimized formulation was 45 mg of liquid lipid, 47 mg of surfactant and 18 mg of drug, which fulfill the requirements of optimized formulations. Three replicates of the checkpoint were analyzed. All the responses were considered to be in good agreement with the predicted values (Fig. 2) which confirmed. The optimized formulation has an average size of 252 9 nm and an entrapment efficiency 87 1%, which were in good agreement with the predicted values. 3.3. Physico-chemical characterization of optimized NLCs Unloaded NLCs were produced based on the optimized MTX-loaded NLCs and then physico-chemically characterized for comparison. TEM studies revealed that MTX-loaded NLCs were
Table 4 ANOVA results from particle size, polydispersity index and entrapment efficiency. Response
Source
DF
Sum of squares
Mean of squares
F
Fcritic
Y1
Model Error Cumulative total
9 5 14
19390.700 3147.698 22538.398
1939.072 62.951
30.80
4.77
Y2
Model Error Cumulative total
9 5 14
0.003 0.001 0.003
0.003 0.000
34.19
4.77
Y3
Model Error Cumulative total
9 5 14
2890.592 1761.015 2950.933
2890.592 35.220
82.07
4.77
M. Ferreira et al. / International Journal of Pharmaceutics 492 (2015) 65–72
69
Fig. 1. Response surface and counter plots of size (a), polydispersity index (b) and drug entrapment efficiency (c).
300
100
280
80
260
60
240
40
220
20
200
Predicted Particle Size
Observed
Entrapment Efficiency (%)
Particle Size (nm)
spherical in shape with narrow size distribution (Fig. 3A). NLCs did not aggregated and no visible significant differences were observed between NLCs and MTX-loaded NLCs. The diameters of the particles observed by transmission microscopy (c.a. 250 nm) are in good agreement with the data obtained from dynamic light scattering. Production of nano-sized unloaded and MTX-loaded NLCs was confirmed by dynamic light scattering analyzes. The mean hydrodynamic diameters and particle size distribution (e.g.
0
Encapsultion efficiency
Fig. 2. Validation of the predicted optimal results with experimental values. (n = 3)
polydispersity index, PDI) of the formulations prepared by hot ultrasonication method are presented in Table 5. The results indicate that NLC size (246 2 nm) increase with the incorporation of MTX to 252 9 nm (p < 0.05) remaining homogenous, as suggested by the low values obtained for the PDI (<0.1). The optimized formulations exhibited a particle size below 300 nm, suitable for systemic and topical administrations (Albanese et al., 2012; Kohli and Alpar, 2004). It has been described that PDI values below 0.2 are indicative of homogeneity of the size distribution and with minimum tendency to aggregation (Mitri et al., 2011). In order to confirm the incorporation of MTX in the NLCs infrared spectra of free MTX, NLC and MTX-loaded NLC were obtained, and are presented on Fig. 3B. MTX spectrum shows, at 1638 cm1, a marked peak that indicates the presence of a C¼C stretching vibration, which is characteristic of the drug molecule (Khan et al., 2010; Kohler et al., 2005). This observation was present also in the MTX-loaded NLC, but not in the NLC, confirming the successful incorporation of MTX in the lipid nanoparticles. The characteristic peaks from NLC were not altered in the MTX-loaded NLC. 3.4. Storage stability assessment Physical stability was assessed by analyzing changes in particle size, polydispersity index and drug content of NLCs stored at room temperature. Changes in particle size are usually accepted as
70
M. Ferreira et al. / International Journal of Pharmaceutics 492 (2015) 65–72
Fig. 3. Characterization of the optimized NLCs. (A) Transmission electron microscopy (TEM) images of NLCs (top) and MTX-loaded NLCs (bottom). The scale indicated below the pictures is 200 nm. Amplification: 80,000. (B) FT-IR spectra of raw MTX powder and lyophilized nanoparticles NLCs and MTX-loaded NLCs.
Table 5 Particle size, PDI, z-potential, drug EE of optimized NLCs.
NLC MTX loaded-NLC
Size (nm)
PDI
z-potential (mV)
EE (%)
246 2 252 9
0.09 0.02 0.06 0.02
13 2 14 1
87 1
PDI, polydispersion index; EE, entrapment efficiency. Mean SD (n = 3).
indicator of formulation instability (Heurtault et al., 2003). For all samples no particle aggregation was found through visual observations, up to 4 weeks. The stability studies of MTX-loaded NLCs suggested that these nanoparticles were quite stable for a month with no significant change in the mean particle size and drug content (Fig. 4). As a broad conclusion, NLCs presented good stability after 4 weeks with average particle size between 212 and 264 nm, PDI below 0.2 and EE remains higher than 85%.
to simulate physiological, inflammatory and topical environments. In Fig. 5 shows the release profiles obtained that were further analyzed to determine the mechanism of release using the kinetic models first-order, Higuchi, Peppas–Korsmeyer and Hixson–Crowell. 3.5.1. Release simulation at physiological conditions In vitro MTX release studies from the NLCs were performed at physiological conditions (37 C, pH 7.4) to access the release profile upon systemic administration. The NLC formulations showed an initial fast release, the release reach 30% in 2 h, followed by a sustained release up to 50% within 24 h (Fig. 5). The observed profile was analyzed by several drug release kinetic models, and fitted best to the Peppas–Korsemeyer model (r2 = 0.9771), since the r2 value is much higher than any other kinetic model. In this model, the value of n characterizes the release mechanism of drug. For the present case n was 0.55, indicative of non Fickian diffusion, as a combination of both diffusion and erosion controlled rate release (Dash et al., 2010).
3.5. In vitro methotrexate release studies The in vitro of MTX release profile from the NLCs was investigated using a dialysis membrane in three conditions defined
3.5.2. Release simulation at inflammatory conditions MTX is considered a reference drug for rheumatoid arthritis. We performed an in vitro release study simulating the inflammatory
Fig. 4. Storage stability assessment. Optimized MTX-loaded NLC were stored up to 4 weeks at room temperature and evaluated for (A) size, (B) polydispersity and (C) drug entrapment efficiency alterations. Data expressed as mean SD (n = 3).
M. Ferreira et al. / International Journal of Pharmaceutics 492 (2015) 65–72
71
Fig. 5. MTX in vitro release from NLCs. On the left, drug cumulative release under physiologic (dark circles) and inflammatory environments (open squares). On the right, release simulation at skin environment. Data expressed as mean SD (n = 2).
Fig. 6. Fibroblasts viability. Unloaded and MTX-loaded NLCs were incubated with fibroblasts for (A) 24 h and (B) 48 h. Data expressed as mean SD (n = 3).
conditions (37 C, pH 5) in order to predict the MTX-loaded NLC in vivo kinetics (Müller et al., 2000). The drug release from the formulation under this conditions exhibited a fast initial release (40% in 2 h) followed by a sustained release up to 80% within 24 h (Fig. 5). Data fitted best to the Peppas–Korsemeyer model (r2 = 0.9865) with an n value of 0.31, suggesting Fickian diffusion that occurs by the usual molecular diffusion of the drug. 3.5.3. Release simulation at skin conditions Topical administration of MTX could represent an interesting approach for skin inflammatory diseases, as psoriasis. Release conditions were set to simulate the drug administration through the skin (32 C, pH 5). A slow release of the MTX from the NLC was observed reaching 30% after 24 h (Fig. 5). In this case, data fitted best to the Hixson–Crowell model (r2 = 0.9290), that describes the drug releases by dissolution mechanism that occurs upon a change in surface area and diameter of particles (Dash et al., 2010). A prolonged release is of interest for dermal formulations as it will contribute to a sustained effect (Pardeike et al., 2009). Gathering the release studies data it is possible to consider the NLCs formulations suitable for systemic and topical administration of MTX under physiological and inflammatory conditions, conferring protection and allowing drug controlled release. 3.6. Cytotoxicity of the optimized formulation To assess the effect of the formulations on cell viability, a MTT assay was performed on L929 fibroblasts. Cells were exposed to empty NLCs and MTX-loaded NLCs up to 250 mg mL1 of MTX equivalent to 6.5 mg mL1 in lipid for 24 and 48 h. The fibroblasts tolerate well empty formulation as no toxicity was observed (Fig. 6). While for MTX-loaded NLCs a slight effect on the viability of cells was observed in the presence of 250 mg.mL1 MTX-loaded NLCs after 48 h incubation. As for all studied conditions cell viability was above 80% it can be stated that the formulations are not toxic to the fibroblasts. The NLC system represents a promising option for delivery of MTX, since it did not affect cell viability, confirming the safety profile of the optimized delivery system.
4. Conclusion MTX loaded NLCs were successfully developed by hot ultrasonication method and optimized using 3- factor, 3-level Box–Behnken design. Based on all data achieved from the formulation design it is possible to conclude that, in general, the amount of surfactant and drug used were important factors for the production of MTX-loaded NLCs influencing their physical properties. This factorial design study has proven to be a useful tool in optimizing NLC for delivery of MTX. The optimized NLCs showed size of 252 9 nm with a polydispersity around 0.08 and an entrapment efficiency of 87%. FTIR analysis enabled confirmation of the efficient entrapment of MTX in the NLCs. The stability tests during 4 weeks at room temperature give a good indication of that at least in solution, the formulation maintain their initial properties. In vitro release profile of MTX from NLCs revealed to be biphasic under physiological and inflammatory environments while sustained at the skin environment. Thus, the optimized formulation revealed suitable for systemic and topical administration of MTX. No toxicity was observed in fibroblasts upon 24 and 48 h of contact with the optimized formulations, indicative of their safe application. This lipid colloidal carrier show interesting properties for delivering methotrexate. Future studies could focus on the evaluation of the therapeutic potential of the optimized formulation in clinically relevant models. Present research demonstrated that MTX loaded NLCs could be a promising tool for systemic and topical administration in the therapy of cancer and inflammatory diseases, such as psoriasis and rheumatoid arthritis. Acknowledgments This work received financial support from the European Union (FEDER funds through COMPETE) and National Funds (FCT, Fundação para a Ciência e Tecnologia) through project UID/Multi/04378/2013. This work was also funded by ON.2 QREN – Quadro de Referência Estratégico Nacional – QREN, by
72
M. Ferreira et al. / International Journal of Pharmaceutics 492 (2015) 65–72
FEDER funds through the Programa Operacional Factores de Competitividade – COMPETE and national funds through FCT (Portugal) through project NORTE-07-0124-FEDER-000067. Luíse Chaves thanks the CAPES Foundation, Ministry of Education of Brazil for the Doctoral fellowship 0831-12-3. The authors would like to acknowledge Excella for kindly provide the MTX. References Abolmaali, S.S., Tamaddon, A.M., Dinarvand, R., 2013. A review of therapeutic challenges and achievements of methotrexate delivery systems for treatment of cancer and rheumatoid arthritis. Cancer Chemother. Pharmacol. 71, 1115–1130. Albanese, A., Tang, P.S., Chan, W.C., 2012. The effect of nanoparticle size shape, and surface chemistry on biological systems. Annu. Rev. Biomed. Eng. 14, 1–16. Barzegar-Jalali, M., Adibkia, K., Valizadeh, H., Shadbad, M., Nokhodchi, A., Omidi, Y., Hasan, M., 2008. Kinetic analysis of drug release from nanoparticles. J. Pharm. Pharm. Sci. 11, 167–177. Braun, J., Rau, R., 2009. An update on methotrexate. Curr. Opin. Rheumatol. 21, 216–223. Carbone, C., Leonardi, A., Cupri, S., Puglisi, G., Pignatello, R., 2014. Pharmaceutical and biomedical applications of lipid-based nanocarriers. Pharm. Pat. Anal. 3, 199–215. Cun, D., Jensen, D.K., Maltesen, M.J., Bunker, M., Whiteside, P., Scurr, D., Nielsen, H. M., 2011. High loading efficiency and sustained release of siRNA encapsulated in PLGA nanoparticles: quality by design optimization and characterization. Eur. J. Pharm. Biopharm. 77, 26–35. das Neves, J., Sarmento, B., 2015. Precise engineering of dapivirine-loaded nanoparticles for the development of anti-HIV vaginal microbicides. Acta Biomater. doi:http://dx.doi.org/10.1016/j.actbio.2015.02.007. Dash, S., Murthy, P.N., Nath, L., Chowdhury, P., 2010. Kinetic modeling on drug release from controlled drug delivery systems. Acta Pol. Pharm. 67, 217–223. Genestier, L., Paillot, R., Quemeneur, L., Izeradjene, K., Revillard, J.P., 2000. Mechanism of action of methotrexate. Immunol. Pharmacol. 47, 247–257. Gohel, M.C., Amin, A.F., 1998. Formulation optimization of controlled release diclofenac sodium microspheres using factorial design. J. Controlled Release 51, 115–122. Hao, J., Fang, X., Zhou, Y., Wang, J., Guo, F., Li, F., Peng, X., 2011. Development and optimization of solid lipid nanoparticle formulation for ophthalmic delivery of chloramphenicol using a Box–Behnken design. Int. J. Nanomed. 6, 683–692. Heurtault, B., Saulnier, P., Pech, B., Proust, J.E., Benoit, J.P., 2003. Physico-chemical stability of colloidal lipid particles. Biomaterials 24, 4283–4300. Huang, Z., Hua, S., Yang, Y., Fang, J., 2008. Development and evaluation of lipid nanoparticles for camptothecin delivery: a comparison of solid lipid
nanoparticles, nanostructured lipid carriers, and lipid emulsion. Acta Pharmacol. Sin. 29, 1094–1102. Khan, M.S., Priyadarshini, M., Sumbul, S., Bano, B., 2010. Methotrexate binding causes structural and functional changes in lung cystatin. Acta Biochim. Pol. 57, 499–503. Kohler, N., Sun, C., Wang, J., Zhang, M., 2005. Methotrexate-modified superparamagnetic nanoparticles and their intracellular uptake into human cancer cells. Langmuir 21, 8858–8864. Kohli, A.K., Alpar, H.O., 2004. Potential use of nanoparticles for transcutaneous vaccine delivery: effect of particle size and charge. Int. J. Pharm. 275, 13–17. Lin, Y.-K., Huang, Z.-R., Zhuo, R.-Z., Fang, J.-Y., 2010. Combination of calcipotriol and methotrexate in nanostructured lipid carriers for topical delivery. Int. J. Nanomed. 5, 117–128. Liu, C.H., Wu, C.T., Fang, J.Y., 2010. Characterization and formulation optimization of solid lipid nanoparticles in vitamin K1 delivery. Drug Dev. Ind. Pharm. 36, 751–761. Marianecci, C., Di Marzio, L., Rinaldi, F., Celia, C., Paolino, D., Alhaique, F., Carafa, M., 2014. Niosomes from 80s to present: the state of the art. Adv. Colloid Interface Sci. 205, 187–206. Mitri, K., Shegokar, R., Gohla, S., Anselmi, C., Müller, R.H., 2011. Lipid nanocarriers for dermal delivery of lutein: preparation, characterization, stability and performance. Int. J. Pharm. 414, 267–275. Müller, R.H., Mäder, K., Gohla, S., 2000. Solid lipid nanoparticles (SLN) for controlled drug delivery – a review of the state of the art. Eur. J. Pharm. Biopharm. 50,161–177. Padmanabhan, S., Tripathi, D.N., Vikram, A., Ramarao, P., Jena, G.B., 2009. Methotrexate-induced cytotoxicity and genotoxicity in germ cells of mice: intervention of folic and folinic acid. Mutat. Res. 673, 43–52. Pardeike, J., Hommoss, A., Müller, R.H., 2009. Lipid nanoparticles (SLN, NLC) in cosmetic and pharmaceutical dermal products. Int. J. Pharm. 366, 170–184. Pinto, M.F., Moura, C.C., Nunes, C., Segundo, M.A., Costa Lima, S.A., Reis, S., 2014. A new topical formulation for psoriasis: development of methotrexate-loaded nanostructured lipid carriers. Int. J. Pharm. 477, 519–526. Pradhan, M., Singh, D., Singh, M.R., 2015. Development characterization and skin permeating potential of lipid based novel delivery system for topical treatment of psoriasis. Chem. Phys. Lipids 186, 9–16. Rahman, Z., Zidan, A.S., Khan, M.A., 2010. Non-destructive methods of characterization of risperidone solid lipid nanoparticles. Eur. J. Pharm. Biopharm. 76 (1), 127–137. doi:http://dx.doi.org/10.1016/j.ejpb.2010.05.003. ski, J., 2006. Methotrexate in rheumatoid arthritis. Pharmacol. Swierkot, J., Szechin Rep. 58 (4), 473–492. Uner, M., 2006. Preparation, characterization and physico-chemical properties of solid lipid nanoparticles (SLN) and nanostructured lipid carriers (NLC): their benefits as colloidal drug carrier systems. Pharmazie 61 (5), 375–386. Visser, K., van der Heijde, D.M., 2009. Risk and management of liver toxicity during methotrexate treatment in rheumatoid and psoriatic arthritis: a systematic review of the literature. Clin. Exp. Rheumatol. 27 (6), 1017–1025.