Topical drug delivery is a convenient mode of drug administration for ocular diseases. Yet, bioavailability through traditional ocular dosage forms, such as eye drops, is very poor. Several factors—such as pH, lachrymal secretion, blinking, tear flow, and epithelial barriers—influence ocular drug bioavailability [1
]. Drainage of most of the drug into the nasolacrimal duct within a few minutes and systemic absorption via conjunctional circulation decrease ocular drug concentration. To provide effective ocular drug concentration, repeated dosing may be required, which may lead to patient noncompliance. Prolongation of pre-corneal residence time is needed to improve the drug bioavailability of topically administered ocular drugs [2
]. Additionally, eye drops are not a suitable dosage form for all purposes––for instance, delivery to the posterior segment––and alternative drug delivery systems are needed to achieve effective concentration in target sites [3
]. A perfect ocular dosage form should be safe and provide selective targeting to the ocular tissue and prolonged delivery with minimal systemic effect [5
Nanoparticle-based products demonstrate three capabilities, including enhanced drug permeation, sustained and controlled drug delivery, and targeted drug delivery. Encapsulation of drugs in these colloidal carriers improves therapeutic effectiveness in comparison with traditional ocular dosage forms (such as eye drops), as has been shown by studies of different nanostructured carriers [6
]. The effectiveness of nanoparticles as ocular delivery systems depends on many factors, including increased pre-corneal retention time, the method of preparation, pre-corneal biodegradation, optimization of the lipophilic–hydrophilic properties of the carrier drug system, and the effects of nanoparticles on corneal structure [10
Nanostructured lipid carriers (NLC) are prepared using solid and liquid lipids, surfactant, and water and range in size from 50 to 1000 nm [11
]. Previous studies have reported the use of solid lipid nanoparticles as ocular delivery systems [12
]. NLC interaction with the corneal mucosa based on biocompatibility and mucoadhesive properties increases the drug’s corneal contact time and improves the ocular bioavailability [13
]. The effect of solid lipid nanoparticles (SLN) on tobramycin ocular delivery was reported previously [14
]. Furthermore, SLN significantly increases drug bioavailability in the aqueous humor. In another study, poorly water-soluble drugs (such as hydrocortisone, estradiol, and pilocarpine) were incorporated into SLN [15
] and corneal permeability was evaluated. The study demonstrated prolonged drug release in all formulations. Ex vivo studies of trans-corneal permeation in animal models have been used to characterize passive cornea permeation. Although permeation studies in such models neglect the complications of tear flow, tear drainage, and blinking, they have provided information about targeting similar molecules from the same pharmacological class [16
]. The aim of the present study was the preparation of the NLC and evaluation of rabbit corneal permeation of propranolol hydrochloride as a beta-blocker agent.
3.1. Formulation Components, Entrapment Efficiency (EE%) and Loading Capacity (LC%)
Different formulations’ properties are illustrated in Table 1
. The experimental design was performed on the basis of full-factorial design, with three variables on two levels. The independent variables were surfactant/lipid ratio (S/L), liquid lipid percentage (%L) and Transcutol percentage (%T). These independent variables were selected based on preformulation study and previously reported research.
Factorial and variance analysis were performed in order to evaluate the impact of independent variables on EE%. The results illustrate that all variables had a significant impact on EE%; however, an increase in variables led to elevated drug loading. Since propranolol hydrochloride is a naturally hydrophilic compound, while the nanoparticles are lipophilic, the maximum loading capacity of 61% (Formulation (1)) is acceptable. On the other hand, the values of LC% were between 1.1% and 2.31%, with maximum and minimum values provided by Formulations (2) and (4), respectively.
3.2. Nanostructure Lipid Carrier(NLC) Particle Size Distribution
summarizes the results regarding particle size and polydispersity index (PDI) for different formulations. The only independent variable that had a significant impact on particle size was %L: increasing the particle size showed a decreasing trend.
As mentioned above, surfactant concentration impact on particle size was not significant, a finding that shows consistency with some previous experiments, though it differs from others. For example, for nitrofurazone, the particle size increased with elevation of co-surfactant concentration [20
], while the opposite results were obtained regarding chitosan-coated SLNs and repaglinide-loaded SLNs [21
]. In addition, evaluation of PDI parameters demonstrates that particle size distribution follows a mono model in most formulations.
3.3. Drug Release from Lipid Nanoparticle Nanostructure Lipid Carrier (NLC)
The experiment was carried out in phosphate buffer with a pH of 7 and drug release was followed for48 h. In order to determine the effect of independent variables on drug release, the percentages of drug released after 4 h (R4) and 48 h (R48) were measured. R4 values measure rapid release of the component, whereas R48 quantifies the slow release rate (Table 3
and Figure 1
). Aqueous solution of propranolol hydrochloride with the same concentration was used as control, and results showed that more than 98% of drug passed through membrane during 3 h. This means that permeation through acetate cellulose membrane was not the limiting step.
As can be observed, the S/L ratio and %L percentage had a significant impact on drug release after 48 h. The results illustrate that an increase in the mentioned variables leads to a decreased drug release rate. These two variables were the reasons for increasing drug loading. Therefore, an increase in oleic acid and surfactant contents results in increasing propranolol hydrochloride loading and simultaneously causes a decrease in the drug’s release rate. Minimum amounts of R48 correspond to Formulation (1), where all three variables were at the highest level. On the other hand, Formulation (4), in which two variables (S/L and %L) were at minimum levels, had the highest value of R48. It seems that increasing the surfactant content promotes drug solubility in lipid matrix; consequently, drug loading in nanoparticles increases, whereas release rate decreases.
The drug release profile from NLC follows a two-step process: an initial rapid release with higher slope followed by a slow release with lower slope in the release profiles. Our results demonstrate that S/L ratio alone had a significant effect on R4 (p < 0.05); an increase in S/L ratio promotes drug loading after 4 h. However, R48 showed similar behavior. In addition, although liquid lipid percentage significantly affected R48, an increase in liquid lipid content resulted in a meaningful decrease at R48.This impact was not significant for R4 (p = 0.22). A comparison between the impact of %L on R4 and R48 indicates that oleic acid (OA) had no short-term impact on drug release, while the longer time period induced elevated drug loading in nanoparticles followed by a decrease in drug release rate. Therefore, S/L ratio and surfactant content play an important role in adjustment of optimum drug loading and release.
In a recent study, we used Compritol ATO 888 as the main solid lipid. It is composed of 64–72% mono- and diglycerides with a melting point of 71.1 °C. HLB 2 and its sustained release properties were previously reported [23
]. In order to evaluate the drug release mechanism of NLC, we studied the release profile in three kinetic models: zero, first-order, and Higuchi models (Table 4
Correlation coefficients and velocity constants in three situations were determined for all formulations. The results indicate that the Higuchi model was more consistent with the release profile. Accordingly, the main mechanism for control release of drug is diffusion, which strongly depends on the concentration gradient between the inside and outside environment of nanoparticles.
3.4. Nanoparticle Morphology
shows Scanning Electron Microscope SEM imaging for propranolol-hydrochloride-loaded NLCs. The figure above shows that the particles are mainly spherical and homogenized, on the basis of the solid solution pattern, while the impact layer surrounding the nanoparticles may account for the drug-enriched shell pattern, though it could be due to the topography of nanoparticles.
3.5. Nanostructure Lipid Carrier (NLC) Permeation through Rabbit Cornea
In order to evaluate the effect of different formulations on propranolol-hydrochloride-NLC permeation, static diffusion cells and isolated rabbit cornea were used. The amount of permeated drug was measured hourly for 5 h. The results are shown in Figure 3
Different permeation parameters were determined, including amount of drug permeating the surface area after 5 h (Q5
) and the permeation rate, which can be obtained from the slope of drug amount against the time curve (Jss
was also determined by using the cumulative amount of permeated drug against time in a steady state. These parameters were measured on the basis of an infinite dose, considering sink condition. The results demonstrate that less than 10% of drug amount in the donor phase permeates through cornea, while the maximum concentration in the receiver phase was not more than 10% of drug saturation concentration in the receiver phase. Thus, sink condition and steady state were maintained. Table 5
summarizes different permeation parameters.
The maximum Q5 was 1.705 for Formulation (4), whereas the minimum content was 0.625 for Formulation (6). It should be mentioned that all formulations had significantly higher values of Q5 than the control (p = 0.001). It seems that only the S/L ratio had a significant but opposite effect on Q5. In other words, increased surfactant content led to an increase in loading and a significant decrease in Q5. We conclude that the effect of surfactant on Q5 was mainly due to drug loading. In addition, Transcutol impact on Q5 was not significant, perhaps due to its percentage in formulations. The effect of different concentrations of Transcutol as a permeation enhancer should be evaluated in another study. NLCs with a lipophilic nature and tendency toward cornea increased drug partitioning into cornea, causing promotion in Q5. The effect of independent variables on Jss was similar to Q5, with the lowest and highest amounts of Jss at 0.155 and 0.4 for Formulations (6) and (4), respectively. In addition, NLC formulations caused a significant p value increase in comparison to the control group (p = 0.001). According to our results, S/L ratio impact upon Jss was significant; S/L increase leads to Jss decrease.
T% and S/L ratio integration with Jss is similar to Q5. Since Q5 and Jss parameters were strongly influenced by drug concentration in the donor phase, and because of different loading efficiency in formulations, a negligible variation in Jss and Q5 was observed due to different drug concentrations. The Papp in cornea was determined in order to normalize Jss regarding drug concentration in the donor phase. The highest permeation coefficient was 0.068 for Formulation (4), while the lowest was 0.012 for Formulation (1). Comparison between permeation coefficients illustrates that although drug amounts in the NLC formulations’ donor phase were much higher than the control’s, all formulations had significant impact on P parameter.
3.6. Differential Scanning Calorimetric (DSC) of Rabbit Cornea
In order to evaluate cornea heat behavior, thermograms were prepared by heating in temperatures ranging from −20 to 120 °C. Different thermograms were adjusted for cornea exposed to buffer, cornea in contact with NLC formulation, and NLC formulation, respectively (Figure 4
Several parameters, such as phase transition enthalpy ∆H and phase transition temperature, were calculated.
The results illustrate that cornea has three transition phases at temperatures of 2.7, 74, and 85 °C, respectively. The transition phase was thoroughly eliminated for cornea in contact with NLC formulations at 2.7 °C. In addition, in other two-phase transitions showed negative shifts; phase transition temperatures were 12.5 and 19.5 °C, respectively. Phase transition enthalpy significantly decreased. Therefore, it seems that formulations absolutely affect cornea structure and alter phase transition. It should be mentioned that phase transition at 2.7 °C occurs due to melting free water existing in cornea [18
As a conclusion, NLC formulations absorb free water in cornea, and due to film forming and mucoadhesion properties they result in extended existence in the ocular system. For human cornea, one transition due to collagen denaturation in 56–74 °C was reported [24
]. The impact of different factors, such as surfactant, on cornea water content was also evaluated. For example, cetylpyridinium chloride as a cationic surfactant caused a decrease in cornea water content, while benzalconium chloride as a cationic surfactant had the opposite effect. It has been proven that cetylpyridinium has no impact on water binding [18
]. Therefore, NLC formulation due to the presence of surfactant or other factors can decrease free water in the cornea. This effect didn’t cause any irritation in rabbit cornea, but more study is need for judgment about safety of NLC.
NLC is the second generation of lipid nanoparticles with advantages of SLNs, while overcoming limitations such as low EE%, poor long-term stability, and possibility of drug expulsion. NLC shows great ability for ocular drug delivery due to better compatibility and modified drug release kinetics [25
]. In the present study, propranolol-hydrochloride-loaded NLCs were prepared and characterized, and propranolol hydrochloride permeability through rabbit cornea by NLCs was evaluated. The percentage of liquid lipid in NLCs has an impact on formulation properties. An increase in liquid lipid promotes drug loading and simultaneously decreases particle size and drug release from lipid nanoparticles. Different results have been reported about impact of liquid lipid on NLC properties. Sangsen et al. indicated that increasing the amount of liquid oil increased the particle size and decreased size distribution, while curcumin entrapment efficiency and release profile were not affected by amount of liquid lipid [27
]. Curcumin is lipophilic with high affinity to loading into NLCs, so an increase in liquid oil did not show any impact on EE%. However, propranolol hydrochloride is a hydrophilic compound with improvement in loading by increasing in liquid oil. In NLCs, liquid and solid lipids produce imperfection in crystal order, which causes higher drug loading by leaving enough space to accommodate drug molecules [28
]. The degree of imperfection depends on the liquid oil. Oleic acid, which is a monounsaturated fatty acid form of stearic acid, produces low imperfection in crystal order [29
]. It seems that the reason for the low EE% of propranolol hydrochloride is low imperfection in NLCs. On the other hand, diethyl glycol monoethylether (Transcutol) is a new enhancer which has solubilization ability apart from integration with polar or non-polar solvents. The concentration of Transcutol was 1% while used as an absorbance enhancer [21
]. Transcutol increased propranolol hydrochloride solubility in NLCs, thus increasing EE%. A similar result was previously reported for didanoside-loaded in NLC as hydrophilic compound [30
]. In addition, surfactant/lipid ratio has an effect on nanolipid properties; it was studied in nitrofurazone permeation through rat skin in a previous experiment [31
]. In the present study, higher surfactant/lipid ratios produced smaller particle size and higher drug solubility in NLCs. This effect was reported for genistein and fluocinolone acetonide-loaded in NLCs [32
Drug loading and release profile of NLCs depend on several factors such as production parameters (method, temperature, etc.), lipid and drug properties, and surfactant concentration. Propranolol hydrochloride is a water-soluble compound which has no tendency toward the lipid phase; on the other hand, the preparation method for nanoparticles is cold homogenization. The drug release profile suggested the loading model followed by solid solution and drug-enriched shell patterns. Mostly solid solution patterns occurred during cold homogenization, where the drug substance was homogenized in solid lipid matrix. Shell patterning predominated during warm homogenization, especially for hydrophilic drug substances. However, during cooling, redistribution from the aqueous phase to lipid core occurs and drug remains in the outer matrix by film formation. As cold homogenization was applied in this study and the burst effect was negligible, the solid solution pattern dominated. Drug migration towards the shell occurred during cooling, as it has high hydrophilicity, so that drug content was much higher in the outer matrix than in the core, also resulting in a drug-enriched shell pattern. We conclude that the drug release profile follows both solid solution and drug-enriched patterns.
The effect of interaction between drug and Compritol ATO 888 on the loading model and drug release through SLN has been reported for tetracaine and etomidate with low melting points and prednisolone with a high melting point [27
]. The results showed approximately 90% loading of the tetracaine and etomidate as lipophilic compounds in SLNs. On the other hand, 20–80% of loaded tetracaine was released in6 h and particle size was introduced as a major factor influencing the release. In the latter article, particle surface area and diffusion coefficient between oily and aqueous phases were mentioned as major factors affecting burst release. Prednisolone loading and release depended on lipid type, as loading was 83.8% without any burst release for 5 weeks for the SLNs prepared with cholesterol. However, loading in SLN-containing Compritol was 80% without burst effect and 37.2% drug release during a 5-week period. The authors concluded that SLN properties strongly depended on lipid melting point; they believed that lipids with lower melting points can produce superior practical sustained release properties.
Additionally, in another study the impact of different factors (such as surfactant, preparation method, and lipid type) was studied for nitrofurazone-loaded SLN properties [31
]. Nitrofurazone loading was reported at 64% and 54.1%, using sodium lauryl sulphate (SLS) and Tween 80 as surfactant, respectively. In comparison with the present study, it was observed that although propranolol hydrochloride is more hydrophilic than nitrofurazone, loading percentage was similar to that for nitrofurazone due to liquid lipid impact on nanostructures. The results demonstrated that S/L ratio has a significant impact on P parameter, while this effect was not observed by other independent variables. Therefore, although Jss
was normalized by concentration adjustment in the donor phase, permeability through cornea was affected by S/L ratio.
In terms of permeability, the results show that independent variables did not significantly relate to Tlag; thus, practically prepared formulations in the present study did not show any obvious effect on shortening the time needed for approaching cornea equilibrium. In general, drug permeability through cornea includes two steps: first, partitioning or rapid distribution of drug from carrier to cornea, and second, permeation in different cornea layers. In order to determine the effects of different formulations on each step, the relation between nano-carrier properties and permeation parameters through cornea was recognized initially. The result indicates that there are no significant integrations between particle size, Jss and P (p = 0.244). Therefore, particle size was not an important factor for Jss and P, practically speaking. Since NLC formulations are not capable of Tlag alteration, we conclude that formulations did not change cornea structure and consequently had no impact on diffusion through different cornea layers.
On the other hand, since the formulations influenced P coefficient, they mainly affect drug partitioning into cornea. In other words, NLC formulations, due to their lipophilic nature, reach the inner parts of the eye. This property was also reported regarding tobramycin ocular delivery in rabbit eye, which showed that drug concentration was significantly higher in all ocular tissues after ocular and intravenous administration of tobramycin SLN formulation, with respect to reference formulations, and only tobramycin SLN allowed drug penetration into retina [14
]. In addition, NLC formulations were also used for enhancing ibuprofen ocular absorption [19
]. The results demonstrate that Gelucire (Gattefosse, Saint-Priest, France) as lipid and Transcutol promote drug permeation, as Papp
coefficient was about 1.5 times higher than control. However, in a recent study Transcutol didn’t significantly impact ocular absorption, while NLCs increased P
coefficient up to 20 for propranolol hydrochloride.
In conclusion, due to the hydrophilic nature of propranolol hydrochloride, its impact on Papp
parameter was much higher than that of lipophilic compounds such as ibuprofen. Mucoadhesion, increased corneal retention time, and enhanced permeation due to cellular uptake by corneal epithelial cells were reported as the main reasons for ocular delivery of topical lipid nanoparticles [34
]. Based on the hydrophilic property of propranolol hydrochloride and obtained results in this study, it seems that NLCs increased drug partitioning into cornea without any alteration in cornea structure.