Rohan M. Shah*a,
Gary Bryantb,
Matthew Taylorb,
Daniel S. Eldridgea,
Enzo A. Palomboa and
Ian H. Harding*a
aDepartment of Chemistry and Biotechnology, Faculty of Science, Engineering and Technology, Swinburne University of Technology, Hawthorn, Melbourne, Australia. E-mail: iharding@swin.edu.au; rshah@swin.edu.au; Fax: +61 3 9819 0834; Tel: +61 3 9214 8715
bCentre for Molecular and Nanoscale Physics (NanoPHYS), School of Applied Sciences, RMIT University, Melbourne, Australia
First published on 8th April 2016
We have recently reported a novel microwave-assisted microemulsion technique for the production of solid lipid nanoparticles (SLNs). SLNs are colloidal carriers made from physiologically well-tolerated lipids that are normally solid at room and body temperature. These microwave-produced SLNs have small size, moderate zeta potential, high encapsulation efficiency and low crystallinity. The drug release studies conducted on drug-loaded SLNs are consistent with a core–shell structure for the microwave-produced SLNs, but with significantly different release profiles depending on the drug used. We further employed multi-angle static and dynamic light scattering (SLS/DLS) and small angle X-ray scattering (SAXS) techniques to help elucidate the structure of microwave-produced SLNs. The SLS/DLS data for the SLNs prepared in this study are consistent with a core–shell structure with a shell thickness of ∼13 nm. SAXS data suggest that the SLNs have a lipid lamellar structure with a repeat spacing of 41.0 ± 0.1 Å.
Suspensions of solid lipid nanoparticles (SLNs) stabilized with emulsifiers have been thoroughly investigated since their inception in the 1990s.2,3 These are emulsions at high temperature when first prepared that, upon cooling, crystallize predominantly into SLNs. Our group has recently reported a novel microwave-assisted microemulsion-based technique to prepare such SLNs.4
The findings from our previous experiments indicate the suitability of these SLNs as potential carriers for lipophilic cargos.4 A variety of techniques have been used to characterize the SLNs including dynamic light scattering, zeta potential analysis, differential scanning calorimetry and X-ray diffraction. However, these techniques have either provided information about the whole assembled (macroscopic) structure or about atomic level structure, rather than the nano-sized structure of interest. It is believed that the structure of SLNs significantly changes in the presence of different drugs and in the presence of various physical stimuli (e.g. temperature, ionic condition of solvent). The structure of SLNs not only influences drug release properties and stability but also influences some of the important biological processes including biodistribution and cellular uptake.5,6 Drug association with SLNs has been well studied,7–10 however, the location of the drugs (within the nanoparticle core, within the shell, particle associated, or dispersed homogenously) often remains unclear. To facilitate further development of SLNs as drug vehicles, it is necessary to obtain a more thorough understanding of their structure at all scales.
The aim of this study is to investigate the nanostructure of the SLNs, and to study how structural morphology changes upon drug loading. Indomethacin and miconazole nitrate were selected as lipophilic “model” drugs based on their physicochemical properties (Table 1) and previous successes in encapsulation studies.11–14 To the best of our knowledge, this is the first study that investigates the structure of the microwave-produced SLNs encapsulated with and without drugs of interest, using light scattering and X-ray scattering techniques.
SLNs | PCS radius (nm) | Zeta potential (mV) | Multi-angle DLS average radius (nm) |
---|---|---|---|
a PCS: photon correlation spectroscopy, DLS: dynamic light scattering. The results presented in this table are mean values ± SD. | |||
Drug-free SLNs | 118 ± 2 | −25.4 ± 1.6 | 136 ± 2 |
Indomethacin-loaded SLNs | 137 ± 4 | −21.4 ± 1.4 | 157 ± 3 |
Miconazole nitrate-loaded SLNs | 143 ± 2 | +19.8 ± 1.0 | 164 ± 4 |
The encapsulation efficiency (EE) of the SLNs was determined by centrifugal ultrafiltration described previously by Shah et al.4 The amount of drug was determined by high performance liquid chromatography (HPLC) (see Section 2.4). The EE of SLNs was calculated as,
A preliminary investigation was also undertaken to study the crystallinity of the SLNs by thermal and X-ray diffraction studies. Thermal studies were undertaken by differential scanning calorimetry (DSC) using a 2920 Modulated DSC (TA Instruments, USA). Samples (∼5 mg) sealed in aluminum pans were kept under isothermal conditions at 25 °C for 10 min and further scanned between 25 °C and 80 °C at a heating rate of 10 °C min−1. The melting enthalpy (ΔH) was obtained by integration of the area under the transition thermogram peak and recrystallization index (RI) was calculated as,
The X-ray diffraction (XRD) patterns were obtained using a D8 Advance diffractometer (Bruker, Germany). The crystalline characteristics of bulk lipid and SLNs were analyzed using primary monochromatic radiation (Cu Kα1, λ = 1.5406 Å). Samples (∼5 mg) were placed on glass sample holders and scanned between 5° and 40° with an angular scan speed of 0.6° min−1. The operating voltage was 40 kV and the current was 40 mA.
A comparison of drug release profiles is often performed to evaluate changes in the release profiles of a newly developed product when compared to the release profiles of a previously approved product. Several approaches (including statistical, model-dependent and model-independent approaches) have been proposed for comparison of drug release profiles.17 In this study, drug release profile comparison was carried out by model-independent approaches (determination of similarity and difference factors, determination of Rescigno indices and the bootstrap f2 method).18–20
The drug release profiles were compared to achieve a preliminary indication of the SLN structure (with respect to drug localization). The structure of nanomaterials often dictates the drug release behavior. It must, however, be acknowledged that these methods are generally used for comparison of two products – an approved product and a “changed” product. Therefore, the results here can only be used as an indicator of similar drug release profile, and hence similar SLN structure, but not proof of that structure.
For SLS, scattered intensities were measured as a function of angle from 20–150° in 5° increments. The average intensity was plotted as a function of scattering angle, and fitted using Rayleigh-Gans-Debye scattering theory.21
The scattering intensity, P(q) for a homogeneous sphere of radius R can be written as,
The scattering factor, P(q), for core–shell spheres with a shell of thickness t can be expressed:
To take into account polydispersity, fits are carried out for Gaussian, Schultz and log-normal distributions of particle sizes, characterized by a mean radius R, and a polydispersity σ. The analysis was carried out using an in-house program written in MATLAB as described previously.22,23
![]() | ||
Fig. 1 SEM images of (a) drug-free SLNs, (b) indomethacin-loaded SLNs and (c) miconazole nitrate-loaded SLNs viewed at 60![]() |
There was a significant (p < 0.05) increase in the particle size of drug-loaded SLNs. The negative zeta potential of the SLNs and indomethacin-loaded SLNs has been attributed to the dissociation of stearic acid and/or functional groups on the surface of the SLNs. Similar attribution of charge has been previously reported.24 Ionisation of indomethacin (which is an acidic drug) on the particle surface may also contribute to this charge. The positive zeta potential of miconazole nitrate-loaded SLNs is probably due to the exposure of cationic nitrogen atoms present in the structure of miconazole nitrate (which is an ionic drug) and must negate any intrinsic negative charge from the stearic acid. The magnitude of the zeta potential in all cases is sufficiently high to provide good physical stability of systems stabilized by non-ionic surfactants such as the one in this study.25
The crystallinity of the SLNs was investigated by DSC analysis. The DSC data of SLNs are summarized in Table 2 whilst the XRD patterns of the SLNs and stearic acid are shown in Fig. 2.
SLNs | Tonset (°C) | Tmax (°C) | ΔH (J g−1) | RI (%) | EE (%) |
---|---|---|---|---|---|
a RI: recrystallization index, EE: encapsulation efficiency, NA: not applicable. | |||||
Stearic acid | 66.9 | 71.5 | 173.0 | 100 | NA |
Drug-free SLNs | 54.5 | 64.4 | 59.6 | 91 | NA |
Indomethacin-loaded SLNs | 55.0 | 64.1 | 53.4 | 83 | 82.6 ± 3.8 |
Miconazole nitrate-loaded SLNs | 62.1 | 63.3 | 52.2 | 80 | 86.6 ± 0.7 |
The presence of a melting endotherm (i.e. Tmax) indicates the solid nature of SLNs.26 Although the onset of melting (i.e. Tonset) remains above body temperature (i.e. >37 °C), the melting transition was reduced compared to bulk stearic acid. This may be due the presence of surfactant molecules on the surface or due to the Gibbs–Thomson “small-size” effect.27 The high surface energy associated with the small size of LNs increases the energy cost of the crystalline state, and thus decreases the melting point of stearic acid in SLNs. The reduction of melting enthalpy (ΔH) and RI is attributed to the reduced crystallinity of stearic acid in SLNs compared to their bulk counterparts.28 The crystallinity was further reduced in drug-loaded SLNs which can be attributed to the entrapment of drug molecules within the lipid imperfections.
The reduction in crystallinity of stearic acid was also evaluated by XRD analysis. The results in Fig. 2 indicate that the characteristic peaks present in the XRD pattern of bulk stearic acid (i.e. one sharp peak at 2θ = 21.7°, three medium-intensity peaks at 2θ = 6.7°, 20.6° and 24.3° and two low intensity peaks at 11.1° and 36.4°) were either reduced or absent in the XRD of SLNs, which is indicative of reduced crystallinity in SLNs29 and aligns well with the DSC data (Table 2).
• An initial burst release (∼75% in first 4 h) followed by a sustained release of drugs (∼15% in next 20 h) – in the case of indomethacin-loaded SLNs.
• A slow, sustained (∼15% in first 4 h) but incomplete release of drugs (∼50% in total after 24 h) – in case of miconazole nitrate-loaded SLNs.
The difference in the release pattern is evident in Fig. 3 which is an overlay of release profiles of indomethacin and miconazole nitrate. This may be due to the difference in physicochemical properties of the drugs: indomethacin is weakly acidic (due to the presence of carboxylic acid functional groups) and has a low partition coefficient in octanol–water systems, with logP = 3.4;30 miconazole nitrate is basic (due to the presence of imidazole ring) and has a high partition coefficient in octanol–water system, with log
P = 6.3.31 These values suggest that miconazole nitrate has higher lipophilicity (and therefore, solubility in the lipid phase) than indomethacin. Thus, one interpretation of the release data is that indomethacin (low partition coefficient) is present predominantly in the shell (i.e. drug-enriched shell model) and released quickly. On the other hand, miconazole nitrate (high partition coefficient) should be solubilized in the molten lipid and crystallize within the lipid core (i.e. drug-enriched core model), thus providing a slower release. These findings support the hypothesis that indomethacin is primarily present in the shell and its release is diffusion-controlled.32 In contrast, the release of miconazole nitrate, which is predominantly present in the lipid core, is controlled by dissolution and diffusion.32
The drug release profiles from both the drug-loaded SLNs were compared by data modelling. Table 3 summarizes the results obtained after profile comparison by model-independent approaches. The indomethacin release data was considered to be the “reference” and miconazole nitrate data was considered to be the “test” formulations. The comparison was done by determination of similarity and difference factors, Rescigno indices and the Bootstrap f2 method. Such methods of profile comparison are often used to compare a “changed” form to an already “approved” product and rarely used to compare release profiles of different drugs from similar drug carriers, as in this study. Hence, the difference should be seen as a proof-of-concept and not as an absolute difference.
Parameter | Value | Accept or reject |
---|---|---|
Difference factor (f1) | 68.94 | Reject |
Similarity factor (f2) | 16.71 | |
Rescigno index (ξ1) | 0.4451 | Reject |
Rescigno index (ξ2) | 0.4314 | |
Bootstrap f2 | 16.706 (16.156–17.308) | Reject |
The difference factor f1 is a measure of the relative error between two release profiles, while the similarity factor f2 is a measure of the similarity in the percent of drug release between two release profiles. For the profiles to be considered “similar”, f1 should be less than 15 (i.e. f1 ∈ [0, 15]) and f2 should be greater than 50 (i.e. f2 ∈ [50, 100]) according to current FDA guidelines for comparison of “changed” and “approved” products.33 The results in Table 3 show that none of these criteria are met which is an indication that the release profiles of indomethacin and miconazole nitrate from SLNs are dissimilar.
The results from difference and similarity factors were consistent with the results obtained by another pairwise procedure of profile comparison: determination of Rescigno indices. The Rescigno index takes on values from zero (indicating no difference between the reference and test formulations) to one (indicating complete release from one formulation before the other begins to release the drug). The results in Table 3 indicate that the profiles were not similar (since the Rescigno indices were ∼0.4 > 0).
The bootstrap f2 method was also used to compare the release profiles. The release profiles are similar, at a 0.05 significance level, if the 90% lower confidence limit of f2 is greater than 50. The results in Table 3 indicate that the 90% lower confidence limit (i.e. 16.156) is less than 50 and, therefore, the release profiles are dissimilar.
All these findings suggest that the release profiles of indomethacin and miconazole nitrate are dissimilar, and therefore the drug encapsulated SLNs may have distinctively different structures. As discussed earlier, however, it is not proof of different nanostructures. In the case of these two drugs, it is quite possible that the different release profiles are due to the different charges on the drugs, with the positively charged miconazole nitrate drug finding it more difficult to leave the SLN than the negatively charged indomethacin drug.
The results in Table 1 are in broad agreement with the PCS measurements, but are slightly larger in all cases compared to the single angle measurements (PCS measurements are essentially single-angle DLS measurements). This is presumably due to slight polydispersity in particle size, which gives a different intensity weighting for each particle, meaning that measurements at a single angle can be different to an average over multiple angles. In other words, the measured particle size depends on the measuring technique. Both sets of results, however, show that the drug-free SLNs are considerably smaller than the indomethacin- and miconazole nitrate-loaded SLNs.
![]() | ||
Fig. 4 SLS data fitted to log normal PSD. (a) Drug-free SLNs, (b) indomethacin-loaded SLNs (I SLNs) and (c) miconazole nitrate-loaded SLNs (M SLNs). |
In all cases the best fit points to particles with a polydispersity of 60% (meaning, for example, that for drug-free SLNs the size of individual particles can vary from as little as 55 nm to as high as 215 nm). However, despite this, the shell thickness has a well-defined value of 13 ± 1 nm across the full range of particle sizes, i.e. it is the same for each individual particle. These results are in agreement with the drug release studies which indicate a core–shell type of structure of the SLNs. It was, however, very interesting that the shell thickness was independent of the drug, given that the release profiles of the two drugs were significantly different. This suggests that the different release profiles were not associated with different nano-structures but more likely due to intrinsic differences in kinetics of dissolution (possibly a result of the different charges).
Trying to fit the data using uniform spheres did not yield a good fit, regardless of polydispersity, for any of the samples. To illustrate this point, the dashed line in Fig. 4a shows the best fit for a homogeneous sphere of the same radius. These findings suggest that the SLNs present in this study are likely to be core–shell type structures.
![]() | ||
Fig. 5 SAXS patterns of drug-free SLNs, indomethacin-loaded SLNs and miconazole nitrate-loaded SLNs. |
Crystallization from the melt and vapor phases produces the C-form of stearic acid, whereas A-, B- and C-forms all crystallize from solutions.40 Aquilano et al. reported the C-form of stearic acid after crystallization of molten stearic acid into lipospheres.41 The microwave-assisted technique of SLN production used in this study is based on similar principles of using microemulsion templates to produce lipospheres. The SAXS analysis indicated that stearic acid in the SLNs crystallized in the C-form. These results align well with the studies of Sato and Boistelle and Aquilano et al.40,41 Being the most stable polymorphic form, further transitions to other forms are less likely and, therefore, drug leakage from the SLNs during storage is less likely to occur.
This journal is © The Royal Society of Chemistry 2016 |