Preparation, optimization, and evaluation of midazolam nanosuspension: enhanced bioavailability for buccal administration

Abstract

Midazolam is considered as one of the best first-line drugs in managing status epilepticus in children who require emergency drug treatment. Due to poor water solubility, oral bioavailability of midazolam is relatively low. To improve its dissolution and absorption, midazolam nano-suspensions were formulated with different stabilizers using the ultrasonic technique. A combination of Tween 80 and Poloxamer (TP) was considered as one stabilizer and 3-methyl chitosan (TMC) as another stabilizer. The ratio of the stabilizers was selected as an independent variable, and their effects on the particle size and the zeta potential were evaluated by the simplex lattice mixture method. The freeze-dried optimized midazolam nano-suspension powder was characterized by particle-size analysis, SEM, the stability test, and the dissolution test. The optimized midazolam nano-suspension (containing 76% TMC and 24% TP) had a mean particle size of 197 ± 7 nm and a zeta potential of 31 ± 4 (mV). The stability test showed that the midazolam nano-suspension is stable for 12 months. In the in vitro dissolution test, the midazolam nano-suspension showed a marked increase in the drug dissolution percentage versus coarse midazolam. In the in vivo evaluation, the midazolam nano-suspension exhibited a significant increase in the C_max and the AUC_0-5, and a major decrease in T_max. The overall results indicate the nano-suspension of midazolam is a promising candidate for managing status epilepticus in children in emergency situation.

Supplementary Information

The online version contains supplementary material available at 10.1007/s40204-020-00148-x.

Introduction

Midazolam is considered as one of the best first-line drugs for managing status epilepticus in children who require emergency drug treatment. Due to high lipophilicity at the physiologic pH, it can be absorbed from the rectal, nasal, and buccal mucosa (Garnock-Jones 2012; McIntyre et al. 2012). It has a rapid effect on the central nervous system (Anderson and Saneto 2012). It is administered by different ways. The intravenous route is the most effective one; however, this method is not preferred for children (Kupietzky and Houpt 1993). Intranasal administration may result in coughing, sneezing, and the expulsion of a part of the drug (Latson et al. 1991). Rectal midazolam is not preferred by the patients and oral administration needs a higher drug dose considering its low bioavailability through this route. Buccolam was the first product of oromucosal midazolam that was approved for the long-term treatment of seizures and epilepsy in pediatric patients. C_max in Buccolamoro mucosal administration in pediatric patients aged 3 months to 18 varies between 87 and 104 (ng/mL) (Garnock-Jones 2012). Wermeling reviewed different formulations of midazolam for intranasal delivery, it has been reported that intranasal formulations have C_max between 50 and 80 (ng/mL) and absolute bioavailability compared to an intravenous dose between (71 and 83%) (Wermeling 2009).

One way to overcome the problem of poorly soluble drugs with low bioavailability is the use of nano-suspensions (Millána et al. 2017; Mishra et al. 2016). According to the specific application of midazolam by pediatrics, the midazolam nanosuspension for buccal administration can be a good option. Nanosuspensions are sub-micron colloidal dispersions of drug particles that are stabilized by surfactants and polymers, or mixtures of both (Wang et al. 2017; Knieke et al. 2015). Nanosuspensions can be obtained by two basic methods: particle size reduction of larger crystals to nano-size (the top–down approach) and by the precipitation of dissolved molecules into solid particles (the bottom–up approach). The top–down process involves techniques, such as high-pressure homogenization (Severino et al. 2012), media milling (Li et al. 2017), microfluidization, ultrasonic, etc. (Gao et al. 2008). Nanosuspension formulation has high drug loading, low occurrence of side effects by excipients, and low cost.

Zhang et al. prepared midazolam nanocrystals through wet milling with different stabilizers to obtain qualified, stable, and low-muscle-irritation nanosuspensions. They used combination of different types of stabilizers to prepare midazolam nanocrystal for muscle injection. The formulation had lower muscle irritation with 2.5% of hydroxypropyl methyl cellulose (HPMC) E5 and 1.0% sodium dodecyl sulphate (SDS). The dissolution velocity of midazolam was accelerated by the nanocrystals. The pharmacokinetics study showed that the AUC0–t of the midazolam nanocrystals was 2.72-fold higher than that of a midazolam solution (Zhang et al. 2020).

Bilgili et al. assessed the role of polymers/surfactants on the particle size of griseofulvin suspensions (Bilgili and Afolabi 2012). The narrowest particle size distribution and the smallest particle size of nanosuspensions were found within the combination of hydroxypropyl cellulose (HPC) and SDS used as stabilizers in comparison to the isolated use of HPC SDS as stabilizers.

Ryde et al. investigated the combination of one polymeric stabilizer and one anionic surfactant in the production of the nanosuspension. The anionic surfactant or the polymeric stabilizer alone cannot produce highly re-dispersible solid dose nanoparticulate compositions. In combination, however, the two compounds exhibit a synergistic effect of stabilizing the active agent and a superior re-dispersibility after administration in mammals (Ryde et al. 2002).

For nanogrinding miconazole, Cerdeira et al. found that the combination of 0.025–0.05% SDS and 5% HPC was the most suitable in providing a synergistic effect for the particle size and nanosuspension stabilization (Cerdeira et al. 2010).

Tween 80 is a small surfactant molecule that effectively decreases the interfacial tension of the particles; in addition, it forms a thin protective outer layer. This thin layer is insufficient for particle stability (Sinha et al. 2013). Combinations of polymer-surfactant are considered for the synergistic effects of nanosuspension preparation and stability. Poloxamers, comprising hydrophobic and hydrophilic chains, are more promising stabilizers in comparison with homopolymers. The most common poloxamers are Poloxamer407 (F127) and Poloxamer 188 (F68). It has been reported that a combination of poloxamers and Tween 80 showed the highest adsorption immediately after particle formation (Sinswat et al. 2005). Ionic polymers like three methyl chitosan (TMC), adsorbed on particle surfaces, lead to higher energy barriers by electrostatic repulsive forces and result in good stability.

Preparation of a new dosage form of midazolam with high bioavailability and easy administration in emergency situation is beneficial and necessary for patients. The purpose of the present study is to: (1) formulate a midazolam nanosuspension for buccal administration to enhance its bioavailability, (2) investigate the effect of the surfactant and stabilizer on physicochemical properties of midazolam nanosuspension, and (3) using optimization study to select the optimal formulation. The in vivo studies were also performed to evaluate the buccal delivery of the midazolam nanosuspension in comparison to coarse midazolam.

Materials and methods

Materials

The midazolam hydrochloride was provided by Combrex (Italy). Pluronic F68 (poloxamer 188) was purchased from BASF (USA). Tween 80 was purchased from the Fisher Chemical Company (USA). Chitosan (low M_w, DD > 75, MDL number MFCD00161512) was obtained from Sigma-Aldrich (USA). All the other analytical-grade chemicals were supplied by Merck (Germany).

Synthesis of N-trimethyl chitosan (TMC)

TMC prepared by the reductive methylation of low Mw chitosan by the procedure described by Jafary et al. (2014). The obtained product was characterized through ¹H NMR and FTIR.

The degree of quaternization is calculated by following formula. This degree shows the percentage of substitution of CH₃ by H.

$$DQ=\left[\frac{[(CH3)3]}{[H]},\times ,1,/,9\right]\times 100$$ 1

where DQ (%) = degree of quaternization as a percentage, (CH₃) 3 integral of the trimethylamino group (quaternary amino group) peak at 3.3 ppm, and [H] integral of the ¹H peaks at 4.7–5.7 ppm.

Experimental design and optimization

For evaluation of the surfactant and polymer concentration effects on particle size and zeta potential, the simplex lattice method of mixture design was chosen. For this study, Design Expert software (version 7.0.0, State Ease Inc, Minneapolis.MN) was used. Good fit of the model was tested using analysis of variance (ANOVA), the prediction multiple correlation coefficient (pre R²), adjusted multiple correlation coefficient (adjusted R²) and the lack of fit, P value and F value provided by Design Expert software were used as factors for selection of adequate models for responses. The quadratic model was selected as a good fit model. The effect of TP (56% of Tween 80 and 44% of Poloxamer 188, the percentage of Tween 80 and poloxamer 188 were obtained from previous work under publication) and TMC on particle size and zeta potential was evaluated. Various combinations of stabilizers in different formulations are shown in Table 1. The percentages of all stabilizers (Tween 80, Poloxamer 188 and TMC) were calculated based on 100 mg of midazolam. The particle size and the zeta potential of the nano-suspensions were optimized by minimizing the particle size and maximizing the zeta potential. The responses were selected in the same order to obtain an optimized formulation. The optimized formulation predicted by the statistical model was prepared experimentally and the responses (i.e., particle size and zeta potential) were determined.

Table 1.

Nano suspension formulation

Formulation Number	T.P (%)	TMC (%)	Particle size (nm)	zeta patential (mV)	PDI	Solubility (μg/ml)
1	25.000	75.000	204	34.0	0.20	492
2	75.000	25.000	355	17.0	0.23	600
3	00.000	100.00	289	26.5	0.26	385
4	00.000	100.00	297	24.4	0.28	392
5	100.00	00.000	308	23.2	0.25	627
6	50.000	50.000	290	23.3	0.22	523
7	50.000	50.000	283	24.6	0.20	517
8	100.00	00.000	319	21.0	0.24	633

Nano-suspension preparation

For preparation of the midazolam nano-suspension, the physical mixtures of the stabilizers with different combinations (according to Table 1) were prepared by mixing the 100 mg coarse midazolam powder and the stabilizers in a mortar until a homogenous mixture was obtained. After that, 10 mL water was added to mixture. A Hielscher sonicator (UP400S) with tip diameter of 3 mm and a length of 100 mm was applied to produce sonication waves for 30 min and 50% amplitude (200 W) at 37 °C. The prepared nano-suspension was thereafter freeze-dried.

Particle size and zeta potential analyses

The particle size (Z-AVE) and the polydispersity index (PDI) of the midazolam nano-suspension were determined by photon correlation spectroscopy (PCS) using a Zetasizer (3000SH, Malvern Instruments Ltd., UK). Before measurement, the midazolam nano-suspensions were diluted with distilled water to achieve the required dilution for analysis. Each sample was measured at a fixed angle of 90° at 25 °C in triplicates. The zeta potential values were assessed by determining the particle electrophoretic velocity using the same instrument.

Dissolution test

Dissolution test (TYPE DT 820, ERWEKA) was conducted using a freeze-dried midazolam and coarse powders using the United States Pharmacopoeia (USP) apparatus II (paddles) at 100 rpm in sink situation. The dry powder containing 10 mg of midazolam was weighed out and added into dialysis bag in 900 mL of the phosphate buffer (pH 7.4), as a dissolution medium. At pre-determined time intervals (1, 2.5, 5, 10, 20, 30, 45, and 60 min), 1 mL of each sample was withdrawn and replaced with fresh buffer. The amount of dissolved midazolam in the sample solution was determined by HPLC (Guo et al. 2013).

HPLC methods

The concentration of midazolam in the dissolution media and plasma was measured by the same validated HPLC analysis. HPLC was assayed (Agilent, SDV30) with a DAD detector at 245 nm. The HPLC separation system consisted of a PerfectSil Target (column: 250 × 4.6 mm L × I.D; 5 µm) equipped with a guard column. For the dissolution media, the mobile phase consisted of methanol and K₂HPO₄ (10 mM; in a ratio of 70:30), which was adjusted to a pH value of 4.7 ± 0.1 with the addition of H₃PO₄. The flow rate of the mobile phase was 1 mL/min. For plasma media, the mobile phase consisted of 10 mM of sodium acetate buffer (pH 4.7 ± 0.1) and acetonitrile (55:45, v/v); the flow rate was 1 mL/min at an ambient temperature (Jurica et al. 2007).

Morphology study

The morphology of coarse midazolam and optimized freeze-dried midazolam nano-suspension were evaluated using scanning electron microscopy (SEM, KYKY-EM3200, China). Prior to examination, the samples were mounted onto a gold stub using a double-sided adhesive tape. The scanning electron microscope was operated at a voltage of 25 kV.

X-ray diffraction studies

The crystalline state of midazolam in the coarse midazolam and freeze-dried optimized midazolam nano-suspensions was clarified with an X-ray diffractometer (Philips, Netherland). The X-ray diffraction (XRD) was carried out in symmetrical reflection mode using the CuK_α line as the source of radiation with a wavelength of 1.54 Å. The standard was generated at 40 kV and 30 mA with a scanning rate of 0.02^°/min and scanning range of 2θ from 0° to 70°.

Differential scanning calorimetry

The coarse midazolam powder, optimized freeze-dried midazolam nano-suspension powder and physical mixtures of the nanosuspension excipients were analyzed by differential scanning calorimetry (DSC-60, Shimadzu Co., Japan) at a heating rate of 10 °C/min from 40 to 300 °C to study the phase transition. This study was done under dry nitrogen atmosphere using Al₂O₃ as a reference.

Particle size stability during storage

Nano-suspensions were stored at 5 °C and 25 °C, at 60% RH for a period of three months. The stability was assessed in terms of particle size distribution, drug solubility, and zeta-potential at one, two and three months.

In vivo pharmacokinetic studies

Rabbits (weighing between 2–2.5 kg) were used as the experiment animals. Ten rabbits were randomly divided into two groups with five animals in each group. Prior to the experiment, the rabbits were fasted for 12 h with free access to water. In the next day, the midazolam nanosuspension and the coarse midazolam suspension, 1 mg/kg body weight expressed as midazolam equivalents were administered to the buccal of rabbits under light anesthesia by ether in two groups. About 1 mL of blood samples was collected via the ear veins at 5–300 min after administration. Extraction of midazolam from blood samples was done in the following method and related recovery was calculated. The collected blood samples were placed in a heparinized tube, and then, separated immediately by centrifugation at 4000 rpm for 5 min and stored at − 40 °C until analysis. All the experiments were carried out in accordance with the NIH Guidelines for the Care and Use of Laboratory Animals. The preparation of the plasma samples was as follows: 200 μL NaOH (0.1 M), 50 μL diazepam (as an internal standard), and 4 mL of diethyl ether were added to each 450 μL rabbit plasma sample. The mixtures were vortexed (MS3, IKA^®, Germany) for 10 min and centrifuged (H 2050R, Xiang Yi Centrifuge Instrument Co., Ltd., Hunan, China) at 3000 rpm for 5 min. The screw-capped tubes, containing the samples, were frozen at − 40 °C for about 35 min. The organic layer was decanted into conical glass centrifuge tubes and concentrated (at 40 °C) under a gentle stream of nitrogen. After evaporation, the residues were re-dissolved in 150 μL of the mobile phase; 50 μL was injected in the HPLC system (Guo et al. 2013; Jurica et al. 2007).

Results and discussion

Midazolam nano-suspension characteristics

Midazolam nano-suspension prepared with different combinations of stabilizers (TP and TMC), at first, it was necessary to synthesise and characterize TMC. The results of ¹H NMR and FTIR test showed that the quaternization degree of TMC was 44% and had CH₃ peak on 1475 cm⁻¹ that confirmed that product is TMC. The particle size and the zeta potential of the eight formulations containing 10 mg/mL midazolam hydrochloride were studied to determine the effect of TP and TMC as stabilizers (Fig. 1 and Table 1). The quadratic model fitted the results very well and there was no lack of fit. The adjusted R² and prediction R² for the particle size and the zeta potential were summarized in Table 2. The parameters of TP and TMC were statistically significant. The final models for particle size (Y₁) and the zeta potential (Y₂) are presented by Eqs. (2) and (3), where A and B are the TP and the TMC percentage, respectively.

$${\text{Y}}_1=389.50\times \text{A}+193.00\times \text{B}-356.00\times \text{AB}(\text{A}-\text{B})$$ 2

$${\text{Y}}_2=19.69\times \text{A}+26.98\times \text{B}+41\text{.36}\times \text{AB}$$ 3

Fig. 1.

a Particle size and b zeta patential responses

Table 2.

Analysis of variance of particle size and zeta potential of TP and TMC nanosuspensions

	F value	P value	Adjusted R2	prediction R2
(a) Particle size
Model	95.01	0.0004	0.91	0.87
Linear mixture	271.91	< 0.0001
AB	7.39	0.0531
AB (A–B)	5.72	0.0750
Lack of fit	4.30	0.1298
(a) zeta potential
Model	22.07	0.0033	0.94	0.90
Linear Mixture	11.46	0.0196
AB	32.68	0.0023
Lack of fit	1.51	0.3510

Considering Fig. 1 and Table 1, at low percentages of TMC (less than 25%), the particle size increased and the zeta potential decreased. The increase of particle size can be explained by the formation of bridging and the flocculation of the TMC molecules that tend to occur at these relatively low polymer concentrations (Mühle 1985). The particle size thereafter decreased and the zeta potential increased with an increase in TMC percentage (from 25 to 75%). This implies that the electrostatic stabilization contributes to the decrease in particle size. Moreover, the particle size started to increase when the percentage of TMC increased from 75%. It seems that at this point, the TMC concentration is sufficient to cover the surface of the nanoparticles. By increasing the TMC percentage (from 75 to 100%) and decreasing the TP percentage (from 25 to 0%), the drug particle size increased. TP can facilitate the adsorption of TMC on midazolam, thus promoting the formation of an entropic barrier which prevents the aggregation of drug particles (Choi et al. 2005). TP percentages less than 25% are not sufficient for adsorbing TMC on the drug particle’s surface. On the other hand, a positive charge of TMC might have repulsive interactions with the amine groups of midazolam (Choi et al. 2005). TMC, therefore, needs a sufficient amount of TP as surfactant to exhibit better adsorbing properties on drug particles.

The solubility of drug particles has been also studied. Based on the solubility test results, it was found that by increasing TP, the solubility of drug particles increased (Table 1). Increasing the solubility of drugs in solvents probably depends on the emulsification property of the surfactant (Sinswat et al. 2005).

Optimization and characterization of the optimum formulation

In this study, optimization was performed to minimize particle size and maximize zeta potential simultaneously. The optimized formulation as well as the upper and the lower levels of confidence of response were evaluated by the software as shown in Table 3. The optimized formulation was predicted to be at a composition of 24% TP and 76% TMC. The predicted particle size and zeta potential for the optimum formulation were 202 ± 5 nm and 32 ± 2 mV, respectively. The experiments were carried out at optimum formulation to verify the statistical model. The particle size and zeta potential of the optimum formulation were as 197 ± 7 nm and 31 ± 4 mV, respectively. This result indicated that the experimental and predicted responses were closely related; therefore, the model was successful in predicting the experimental results.

Table 3.

Experimentally and the predicted responses of optimized nanosuspension

predicted results by software							Experimentally results
Particle size (nm)			Zetapotential (mV)			Desirability	Particle size (nm)	Zeta potential (mV)
Predicted	Low level	High level	Predicted	Low level	High level	0.94	197.0 ± 7	31.0 ± 4
210.0	190.9	229.0	33.6	30.9	36.4

The optimum formulation was selected for morphological study. As shown in Fig. 2, the midazolam nanoparticles were more uniform in shape in comparison to the pure drug powder. Micronized midazolam powder showed irregular shapes, broad particle size distribution, and generally larger particle size than midazolam nanoparticles.

Fig. 2.

Scanning electron microscopy of (a) freeze-dried midazolam nanosuspension and b midazolam coarse powder

The optimized formulation was also selected for dissolution study. Figure 3 shows that the drug dissolution of the midazolam nano-suspension is 2.3 times higher than the midazolam coarse suspension. Since drug nanoparticles have much more surfaces available under the same condition, they can dissolve easily in buffer medium and have much drug uptake (Jung et al. 1999).

Fig. 3.

Dissolution profiles of the (■) midazolam coarse powder and the (♦) freeze-dried midazolam nanosuspension powder, (n = 3)

Differential scanning calorimetry and XRD analysis

For determining whether amorphization occurred at the particle surfaces, the DSC analysis was performed on the freeze-dried nanosuspension, drug-stabilizers physical mixture and pure midazolam. Midazolam demonstrated only a single sharp endothermic peak at 292.2 °C corresponding to its melting point. The DSC of the drug-stabilizers physical mixture and nanosuspensions was also performed (Fig. 4), both showing an endothermic peak at 289.7 °C and 288.4 °C, respectively, ascribing to the melting of the drug. However, the melting point of the nanosuspension was lower than that of the raw crystals, which might be due to the particle size reduction. This result implied that there was no interaction between the drug and excipients. In the powder X-RD patterns of the pure midazolam and midazolam nanosuspension, all the midazolam peaks appeared which was a proof of crystalline state of midazolam. The results are evident in Fig. 5

Fig. 4.

The DSC pattern of (a) pure midazolam, b drug-stabilizer physical mixture and c midazolam nanosuspension

Fig. 5.

The XRD analysis of (a) midazolam and b midazolam nanosuspension

Stability

The particle size and the zeta potential of the optimized nanosuspension were monitored for three months (Table 4). The appearance of the optimized nanosuspension did not change after 12 months’ storage at 4 °C and 25 °C. A loose and thin layer of sediment was observed when the nanosuspension was stored at 25 °C for 12 months. However, this layer disappeared by slight shaking. According to Table 4, the optimized nanosuspension had a zeta potential of about 30 mV. An absolute zeta potential value higher than 60 mV was considered to be extremely stable; 30 mV meant good stability; 20 mV showed acceptable short-term stability; and less than 5 mV induced fast particle aggregation (Wang et al. 2017; Jurica et al. 2007; Wu et al. 2011).

Table 4.

Stability results of optimized nanosuspension

	First month	Second month	Third month	Sixt month	Twelfth month
T = 25 °C
Particle size (nm), PDI	207 ± 10, 0.22 ± 0.05	210 ± 5, 0.24 ± 0.05	213 ± 7, 0.27 ± 0.06	215 ± 4, 0.27 ± 0.05	215 ± 5, 0.27 ± 0.06
Zetapotential (mV)	33 ± 3	30 ± 5	31 ± 2	29 ± 3	28 ± 5
Solubility (μg/ml)	492 ± 8	489 ± 7	494 ± 5	485 ± 6	483 ± 8
T = 4 °C
Particle size (nm)	203 ± 3, 0.20 ± 0.03	200.24 ± 0.05 0.21 ± 0.02	208 ± 3 0.22 ± 0.04	211 ± 4, 0.24 ± 0.06	213 ± 2, 0.25 ± 0.05
Zetapotential (mV)	33 ± 2	33 ± 3	31 ± 2	31 ± 3	32 ± 1
Solubility (μg/ml)	488 ± 6	488 ± 6	490 ± 8	482 ± 9	479 ± 8

In vivo bioavailability

To confirm the advantage of the application of midazolam nanosuspension in improving the buccal bioavailability and its fast effects, an in vivo test was carried out in rabbits and pharmacokinetic parameters were obtained. The recovery of midazolam extraction method was 97% ± 2. The plasma concentration–time curves of the midazolam coarse suspension and the midazolam nanosuspension are shown in Fig. 6. The pharmacokinetic parameters were given in Table 5. As shown, the midazolam coarse suspension and nanosuspension differ from each other in the corresponding parameters. The midazolam nanosuspension had higher C_max (111.90% higher), higher AUC0–t (275.08%), and shorter T_max (15 min), which indicated that midazolam nanosuspension was more easily absorbed in vivo. The particle size has been recognized as a key parameter for absorption from the buccal and the gastrointestinal tissue (Lee et al. 2008).

Fig. 6.

Themean plasma concentration–time curve of the midazolam coarse suspension (♦) and the midazolam nanosuspension (■), (n = 3)

Table 5.

Main pharmacokinetic parameters of midazolam coarse suspension and nanosuspension after buccal administration in rabbits (n = 5)

Parameters	Midazolam coarse suspension	Midazolam nanosuspension
Cmax (ng/mL)	420 ± 12	890 ± 14
Tmax (minute)	30 ± 3	15 ± 2
AUC (ng/mLh)	28.9 ± 2.6	108.4 ± 8.2

Conclusion

In this study, a nanosuspension of midazolam was prepared and the effects of the formulation compositions on the particle size and the zeta potential of the midazolam nanosuspension were determined. It was found that TMC, with positive charge effect was the most effective parameter on the particle size and the zeta potential of the midazolam nanosuspension. The optimized nanosuspension was stable for three months. Drug dissolution studies showed that the optimized midazolam nanosuspension was more soluble than the coarse drug suspension. Buccal bioavailability can be greatly enhanced by reducing the drug particle size to fall within the nanometer range. The overall results indicate the nanosuspension of midazolam is promising candidate for managing status epilepticus in children in emergency situation.

Supplementary Information

Below is the link to the electronic supplementary material.

Compliance with ethical standards

Conflict of interest

The authors would like to thank the Iran National Science Foundation (INSF) for the financial support of this work. Authors declare that have no conflict of interest. All in vivo experiments were performed in accordance with the international guide for the care and use of laboratory animals. The experimental procedures were evaluated and approved by the Committee for Ethics in Animal Research at Tarbiat Modares University (Approval No: IR.TMU.REC.1396.706).