- Email: [email protected]
Effect of beeswax modification on the lipid matrix and solid lipid nanoparticle crystallinity
Available online at www.sciencedirect.com
Colloids and Surfaces A: Physicochem. Eng. Aspects 315 (2008) 189–195
Effect of beeswax modification on the lipid matrix and solid lipid nanoparticle crystallinity Anthony A. Attama a,b,∗ , Christel C. M¨uller-Goymann a,1 a
Institut f¨ur Pharmazeutische Technologie, Technische Universit¨at Carolo-Wilhelmina zu Braunschweig, Mendelssohnstraße 1, D-38106 Braunschweig, Germany b Department of Pharmaceutics, University of Nigeria, Nsukka 410001, Enugu State, Nigeria Received 24 April 2007; received in revised form 10 July 2007; accepted 26 July 2007 Available online 8 August 2007
Abstract The influence of a heterolipid (Phospholipon 90G® , P90G), which directly modifies the surface of solid lipid nanoparticles (SLN), and goat fat on the crystallinity of beeswax matrix and the SLN prepared therefrom was studied. Lipid matrices composed of 30% (w/w) of P90G in beeswax and in 1:1 mixture of beeswax and goat fat were formulated and characterized. SLN containing polysorbate 80 with or without P90G were formulated by hot high-pressure homogenisation and characterized by particle size and zeta potential measurements. Differential scanning calorimetry (DSC) and wide-angle X-ray diffraction (WAXD) patterns of the SLN were compared with the lipid matrices. Results showed WAXD of the lipid matrices containing P90G contain amorphous portions. Most SLN formulated possessed low z-average diameters and polydispersity indices, which remained constant after 3 months, with high-negative potentials after 4 weeks and crystallized into stable modification within 48 h of preparation. The overall result indicated that P90G and goat fat influenced the crystallinity of beeswax matrix and its SLN, but the crystallinity of the mixed lipid matrix (1:1 beeswax/goat fat) and its SLN, beeswax containing P90G and its SLN was lower than that of beeswax alone and its SLN. There was no increase in crystallinity of SLN on storage. Modification of beeswax with P90G or goat fat offers a way of improving the SLN formulated with beeswax in terms of reduction of its crystallinity responsible for its low-drug incorporation efficiency. © 2007 Elsevier B.V. All rights reserved. Keywords: Beeswax; Goat fat; Mixed lipid; Phospholipon 90G® ; Modification; Solid lipid nanoparticles; Crystallinity; Drug delivery
1. Introduction
Abbreviations: SLN-1, SLN prepared with 30% (w/w) P90G in beeswax as lipid matrix and 0.3% (w/w) polysorbate 80; SLN-2, SLN prepared with 30% (w/w) P90G in beeswax as lipid matrix and 0.6% (w/w) polysorbate 80; SLN-3, SLN prepared with 30% (w/w) P90G in beeswax as lipid matrix and 1.0% (w/w) polysorbate 80; SLN-4, SLN prepared with beeswax as lipid matrix and 1.0% (w/w) polysorbate 80; SLN-A, SLN prepared with 1:1 mixture of beeswax and goat fat as lipid matrix and 1.0% (w/w) polysorbate 80; SLN-B, SLN prepared with 30% (w/w) P90G in 1:1 mixture of beeswax and goat fat as lipid matrix without polysorbate 80; SLN-C, SLN prepared with 30% (w/w) P90G in 1:1 mixture of beeswax and goat fat as lipid matrix and 0.3% (w/w) polysorbate 80; SLN-D, SLN prepared with 30% (w/w) P90G in 1:1 mixture of beeswax and goat fat as lipid matrix and 0.6% (w/w) polysorbate 80; SLN-E, SLN prepared with 30% (w/w) P90G in 1:1 mixture of beeswax and goat fat as lipid matrix 1.0% (w/w) polysorbate 80 ∗ Corresponding author at: Department of Pharmaceutics, University of Nigeria, Nsukka 410001, Nigeria. Tel.: +234 42 771911; fax: +234 42 771709. E-mail address: [email protected] (A.A. Attama). 1 Tel.: +49 531 3915654; fax: +49 531 3918108. 0927-7757/$ – see front matter © 2007 Elsevier B.V. All rights reserved. doi:10.1016/j.colsurfa.2007.07.035
Solid lipid nanoparticles (SLN) are an alternative carrier system to polymer nanoparticles or liposomes. They consist of physiological and biocompatible lipids, which are suitable for the incorporation of lipophilic, hydrophilic and poorly watersoluble active ingredients. Improved bioavailability, protection of sensitive drug molecules from the outer environment (water and light) and controlled release characteristics were claimed by incorporation of drugs in the solid lipid matrix [1]. Due to the rigidity of the nanoparticles, the mobility of an incorporated drug is reduced, preventing drug leakage. A major disadvantage of SLN is their inherent low-drug incorporation due to the crystalline structure of the solid lipid. The high crystalline order of nanoparticles prevents incorporation of high amounts of drug molecules, but the use of triglyceride mixtures results in particles with a less ordered matrix [2]. Moreover, addition of a second matrix component may specifically alter the crystallization behaviour of the lipid
190
A.A. Attama, C.C. M¨uller-Goymann / Colloids and Surfaces A: Physicochem. Eng. Aspects 315 (2008) 189–195
matrix. In objective of this study, beeswax was modified with goat fat and the nanoparticles formulated therefrom studied. The effect of the modification on crystallinity of the P90G or polysorbate 80 stabilized SLN formulated was studied. Properties of the SLN prepared with beeswax containing P90G were compared with SLN prepared with 1:1 mixture of beeswax and goat fat. Beeswax SLN containing polysorbate 80 as the only surfactant was compared with other triglycerides and it was found that the beeswax nanoparticles possessed excellent particle size stability but poor drug encapsulation efficiency despite the fact that polymorphism of waxes is very low compared with triglycerides [3,4]. Phospholipon 90G® has been used in parenteral emulsions and in formulation of liposomes, and was recently shown to be a good surface modifier for SLN [5,6] with resultant improvement in targeting and pharmacokinetics [7–9]. The phospholipid bilayer structure formed around the lipid core may increase the drug loading capacity, as biologically important molecules can be anchored on the colloidal particle surface, and surfacemodification also enables stabilization of colloidal particles especially when generation of the nanoparticles is carried out in an aqueous medium [10]. Beeswax is a natural product used in pharmaceutical, cosmetics, food and other industries [11]. It is highly crystalline [12], and this adversely affects its drug holding capacity. Addition of goat fat and phospholipid (heterolipid) is expected to greatly disturb its crystal order in addition to surface modification. Goat fat on the other hand, is derived from a domestic animal, Capra hircus and has been characterized and used in experimental drug delivery systems [13,14]. 2. Materials and experiments 2.1. Materials Beeswax (Cera alba) Ph. Eur., sorbitol (Caesar & Loretz, Hilden Germany), thimerosal (Synochem, Germany) and polysorbate 80 (Tween 80® , Across Organics, Germany), Phospholipon 90G® (Phospholipid GmbH K¨oln, Germany) were used as procured. Goat fat was obtained from a batch processed according to earlier procedure [13]. Double-distilled water was used for nanoparticle preparation. 2.2. Preparation of lipid matrix containing Phospholipon 90G® The lipid matrix composition was chosen based on previous experience with SLN prepared and evaluated in our laboratory [5,6], and contained 30% (w/w) of Phospholipon 90G® in beeswax or 1:1 mixture of beeswax and goat fat, and was prepared initially by fusion prior to nanoparticle preparation. 2.3. Preparation of nanoparticles Solid lipid nanoparticles were prepared to contain 5.0% (w/w) lipid matrix (30% (w/w) of P90G in beeswax and 30% (w/w) of P90G in 1:1 mixture of beeswax and goat fat), 0.3, 0.6
or 1.0% (w/w) polysorbate 80, 0.005% (w/w) thimerosal, 4.0% (w/w) sorbitol and enough double-distilled water to make 100% (w/w). SLN without P90G or polysorbate 80 were also prepared. The nomenclatures used for the different batches of SLN formulated are presented in the key below. Melt-homogenisation technique was adopted. In each case, the lipid matrix was melted at 75 ◦ C, which is more than 10 ◦ C above the melting temperature and avoids the lipid memory effect and makes new crystallization possible [15]. The double-distilled water containing the polysorbate 80, thimerosal and sorbitol at the same temperature, was added to the molten lipid matrix, mixed very well and dispersed at 75 ◦ C with Ultra-Turrax (T25 basic, Ika® Staufen Germany) at 24,000 rpm for 5 min and immediately passed through a heated high pressure homogeniser (EmulsiFlex-C5, Avestin Canada) at a pressure of 1000 bars for 20 cycles. 2.4. Photon correlation spectroscopy (PCS) Particle sizes were determined by PCS at 25 ◦ C using a Multi Angle Particle Size Analyser (Zetasizer 3 Model AZ6004, Malvern England) modified with a 35 mW He–Ne laser (Model 127-35, Spectra Physics USA). The detection was performed at a scattering angle of 90◦ in a cell AZ10 equilibrated at 293 K and at an accumulation time of 180 s. Samples were diluted with filtrated double-distilled water (0.2 m Sterifix® filter) and data were analysed by the cumulants method assuming spherical particles. Particle size studies of SLN were done 24 h, 1, 4 and 12 weeks after preparation. 2.5. Measurement of zeta potential The zeta potentials of the formulated SLN were determined after 1 month of preparation in a Zetasizer Nano Series (NanoZS, Malvern Instruments England). Each sample was diluted with bidistilled water and the electrophoretic mobility determined at 25 ◦ C and dispersant dielectric constant of 78.5 and pH of 7. The obtained electrophoretic mobility values were used to calculate the zeta potentials using the software DTS Version 4.1 (Malvern, England) and applying Henry equation [16]: UE =
2εZf (Ka ) 3η
(1)
where Z is the zeta potential, UE the electrophoretic mobility, ε the dielectric constant, η the viscosity of the medium and f(Ka ) is the Henry’s function. 2.6. Wide angle X-ray diffraction (WAXD) studies To study the crystalline characteristics, WAXD studies were done on the lipid matrices and all the SLN as earlier described [13,17]. Wide angle X-ray studies were done using an X-ray generator (PW3040/60 X’Pert PRO, Fabr. DY2171, PANalytical Netherlands) connected to a copper anode (PW3373/00 DK 147726 Cu LFF). WAXD diffractograms were obtained 3 months after lipid matrices preparation and 48 h, 1 week, 1 month and 3 months after SLN preparation.
A.A. Attama, C.C. M¨uller-Goymann / Colloids and Surfaces A: Physicochem. Eng. Aspects 315 (2008) 189–195
191
2.7. Differential scanning calorimetry (DSC) The degree of crystallinity of the nanoparticles was determined on a calorimeter (DSC 220C) connected to a disc station (5200H, Seiko, Tokyo, Japan). Approximately 5 mg of the SLN was weighed into an aluminium pan and sealed hermetically, and the thermal behaviour determined against an empty pan in the range of 5–125 ◦ C at a heating rate of 5 ◦ C min−1 . Transition temperatures were taken as the peak minima of the endothermic transitions, while transition enthalpies were determined by integration of the endothermic transitions using linear baselines. Measurement was done on the SLN after 1 month of preparation. 3. Results and discussion 3.1. Photon correlation spectroscopy (PCS) and zeta potential measurement The particle size distribution of the nanoparticles presented in Fig. 1a and b shows low-sized uniform nanoparticles were obtained, except for SLN prepared with mixed lipid and P90G only (SLN-B). The figure shows that increase in polysorbate 80 concentration reduced the particle size. However, the sizes of SLN-2 containing 0.6% (w/w) polysorbate 80 and SLN-3 containing 1.0% (w/w) polysorbate 80 were close, but the polydispersity indices (PI) were significantly different (p < 0.05) with PI of the former being consistently higher than those of the latter throughout the period of storage (Fig. 1a). Thus, SLN-3 had a narrower particle size distribution than SLN-2. Polydispersity index (PI) measures the width of the particle size distribution. Consistently high values of PI indicate either an aggregated or poorly prepared sample. If the PI is above 0.2, the PI loses its significance as an accurate measure of the width of the size of distribution, but can still be useful for comparative purposes. The lower PI of SLN-3 thus indicates a better control of particle size. In all the SLN, the z-average diameter and PI remained almost constant after 3 months, except in SLN-1 containing 0.3% (w/w) polysorbate 80, where there was an increase in PI after 12 weeks. Low standard deviations were obtained for all the SLN. Consistent with reported work [3], nanoparticles prepared with beeswax and polysorbate 80 alone (SLN-4) possessed good particle size stability with low PI compared with SLN-1 and SLN-2. But high crystallinity recorded in WAXD and DSC would preclude the use of SLN-4 because of difficulty in incorporation of drugs, although drugs could be coupled or adsorbed to the surface but may not achieve sustained release effect. The particle size stability coupled with low degree of crystallinty obtained for SLN-2 and SLN-3 could make them good candidate drug delivery systems with improved drug incorporation capacity. Highly crystalline particles and lipid matrices lead to drug expulsion [18]. All the SLN containing both P90G and polysorbate 80 may be suitable for intravascular delivery systems because of their low size, stability and in vivo tolerability of the component lipids. Beeswax has been used as a thickener and a humectant in the manufacture of ointments, creams, lipsticks and other cosmetics. However, recent studies have shown promising in vivo applications of beeswax and other long chain fatty alcohols and
Fig. 1. (a) Particle size distribution of SLN prepared with 30% (w/w) P90G in beeswax: SLN-1 to 3 contained 0.3, 0.6 and 1.0% (w/w) polysorbate 80, respectively, while SLN-4 contained 1.0% polysorbate 80 but no P90G. z-average and PI represent, respectively, z-average diameter and polydispersity index. (b) Particle size distribution of SLN prepared with 30% (w/w) P90G in 1:1 mixture of beeswax and goat fat: SLN-A contained 1.0% polysorbate 80 but no P90G; SLN-B contained P90G but no polysorbate 80; SLN-C to E contained 0.3, 0.6 and 1.0% (w/w) polysorbate 80, respectively. z-average and PI represent respectively, z-average diameter and polydispersity index.
acids [19–21], thus parenteral administration of this dosage form would not pose any metabolic problem. SLN prepared with mixed lipid matrix (SLN-A to E) produced lower z-average sized particles when P90G and polysorbate 80 were incorporated (SLN-D and SLN-E) compared with SLN prepared with beeswax alone at polysorbate concentrations of 0.6% (w/w) and 1.0% (w/w), but with higher PI (Fig. 1b). This was due to greater ease of particle disintegration and slower recrystallization in the mixed lipid as its melting point and crystallinity were lower than that of beeswax. SLN containing only beeswax and P90G could not be prepared. SLN-B
192
A.A. Attama, C.C. M¨uller-Goymann / Colloids and Surfaces A: Physicochem. Eng. Aspects 315 (2008) 189–195
prepared with mixed lipid matrix and P90G without polysorbate 80 showed high growth in particles up to micrometer scale within 12 weeks. Addition of a mobile surfactant is necessary for stabilization of the nanoparticles containing phospholipid. In terms of controlled and sustained or prolonged drug delivery, SLN prepared with beeswax containing both P90G and polysorbate 80 could be recommended because of the more crystalline nature compared with the mixed lipid matrix SLN. However, when high drug incorporation is needed, SLN formulated with mixed lipid core containing P90G and polysorbate 80 will be better as they were shown to possess low crystalline order which favours drug loading. All the same, both SLN proved better than SLN formulated with beeswax and polysorbate 80. They can be used in active or passive drug targeting depending on the intentions of the formulator. The nanocarrier formulated with 0.0% (w/w) of polysorbate 80 (SLN-B) exhibited gelation with storage. Although gelation was observed here, there was no increase in crystallinity detected by either DSC or WAXD providing evidence that the gelation was not a consequence of alkyl chain ordering (increase in crystallinity) and that a mobile surfactant is required to achieve optimum particle size stability in phospholipid-stabilized lipid nanoparticles. Insofar as there would be no drug expulsion in this SLN during storage, it is not suitable for use as drug delivery carrier because of high growth in particle size. Since there were neither increase in crystallinity nor gelation in the more stable SLN formulated with mixed lipid matrix (SLN-C to E), the novel mixed matrix lipid nanocarriers could overcome the low drug incorporation capacity of crystalline lipids such as beeswax, which is a drawback of lipid particulate drug delivery systems. Evaluation of the drug incorporation efficiency of these optimised SLN is planned for the next stage of this investigation. The zeta potential values obtained were −20.2 ± 5.3, −27.9 ± 8.3, −32.2 ± 6.3 and −34.3 ± 0.49 mV for SLN-1 to 4 respectively; and −20.0 ± 1.2, −19.2 ± 0.6, −15.9 ± 0.2, −18.4 ± 0.3 and −15.4 ± 0.6 mV for SLN-A to E, respectively, with SLN-E having the lowest absolute value and SLN-4 the highest. The magnitude of zeta potential gives an indication of the potential stability of a colloidal system. Absolute large negative or positive zeta potential is required for colloidal dispersion stability. The general dividing line between stable and unstable suspension is generally taken as either +30 or −30 mV [16]. Based on the zeta potentials obtained for the SLN in distilled water, the dispersion stability of the SLN could be stated in the following descending rank order: SLN-4 > SLN-3 > SLN2 > SLN-1 ≈ SLN-A > SLN-B ≈ SLN-D > SLN-C ≈ SLN-E. Although SLN-4 had highest absolute zeta potential value (highest stability), other properties associated with this SLN would preclude its use in formulation. The overall negative zeta potentials of all the SLN would also contribute to stability of the lipid nanodispersions, and provide surfaces for attachment of drugs and other biomacromolecules. The fact that SLN-C, SLN-D and SLN-E with low-absolute zeta potential values showed good particle size stability indicate P90G had a stabilizing effect at the interface. SLN formulated with mixed lipid possessed lower absolute zeta potential values compared with SLN prepared with beeswax alone with or without P90G.
Fig. 2. WAXD of lipid matrices: Beeswax containing 30% (w/w) P90G (BWP90G), 1:1 mixture of beeswax and goat fat (1:1 LM), lipid matrix containing 1:1 mixture of beeswax and goat fat and 30% (w/w) P90G (1:1 LM-P90G).
This effect may be related to the shape of the lipid particles but this has to be proved in a further investigation. 3.2. Wide angle X-ray diffraction (WAXD) studies WAXD was used to study the crystalline character of the lipid matrices and the SLN. Sharp high intensity reflections are produced by highly crystalline lipids which appear above the amorphous background of non-crystalline lipids. Fig. 2 shows the WAXD diffractograms of the lipid matrices. Pure beeswax reflections were sharp and conformed to the result of earlier study [22]. Beeswax containing 30% (w/w) P90G presented ˚ (high intensity), 2θ = 23.7◦ reflections at 2θ = 21.4◦ d = 4.15 A ˚ (low intensity) and 2θ = 5.6◦ d = 15.78 A ˚ and 2θ = 7.0◦ d = 3.75 A ˚ (very weak intensities). There is an amorphous pord = 12.63 A tion around 2θ = 19.0◦ . The slight amorphous portion present in the diffractogram of the beeswax containing phospholipid is not present in the diffractogram obtained for pure beeswax. This shows that the former matrix is less crystalline than the latter. It is thus expected that the amorphous portion would home any incorporated drug as there are spaces where drug could be incorporated, compared with pure beeswax which is more crystalline. Also the presence of higher intensity sharp reflections in the diffractogram of beeswax at the wide angle region indicates that most of the hydrocarbon chains of beeswax are highly ordered [23], in comparison with the reflections obtained for the lipid matrix containing phospholipid, attesting that beeswax has higher crystallinity. Phospholipon 90G® used in this study mainly contains linoleic, oleic, stearic and palmitic acids, which are fatty acids of different chain lengths and degree of saturation [24]. The interaction of these fatty acids with the diverse fatty acids present in beeswax [11,12] resulted in the partly amorphous nature of the lipid matrix containing the phospholipid. Reflections of medium and low intensities detected for the SLN prepared with beeswax containing P90G or beeswax ˚ and alone (SLN-1 to 4), respectively, at 2θ = 21.4◦ d = 4.15 A
A.A. Attama, C.C. M¨uller-Goymann / Colloids and Surfaces A: Physicochem. Eng. Aspects 315 (2008) 189–195
Fig. 3. (a) WAXD of SLN prepared with 30% (w/w) P90G in beeswax after 3 months: SLN-1 to 3 contained 0.3, 0.6 and 1.0% (w/w) polysorbate 80, respectively, while SLN-4 contained 1.0% polysorbate 80 but no P90G. (b) WAXD of SLN prepared with 30% (w/w) P90G in 1:1 mixture of beeswax and goat fat after 3 months: SLN-A contained 1.0% polysorbate 80 but no P90G; SLN-B contained P90G but no polysorbate 80; SLN-C to E contained 0.3, 0.6 and 1.0% (w/w) polysorbate 80, respectively.
˚ after 48 h, 1 week and 1 month (figures not 2θ = 23.8◦ d = 3.74 A shown) remained unchanged after 3 months (Fig. 3a). Absence of growth in intensities of these reflections on storage indicates absence of any dynamic polymorphic transformation, and that progression to increasingly ordered nanoparticles due to polymorphic transition would not occur on storage. In comparison, the intensities of the reflections of SLN-1 to 3 containing P90G were significantly lower (p < 0.05) than those of SLN-4 prepared with beeswax and polysorbate 80 only. In this study, the reflections detected for SLN-1 to 4 were in the same modification as the reflections detected for beeswax containing phospholipid attesting that no modification of the wax esters occurred in the nanoparticles. Thus, any included drug would not be expelled from the nanoparticles as a result of change in modification leading to a change in crystal order. Although the crystal order was
193
greatly disturbed in SLN-1 to 3, these drug carriers remained solid. Most published works are on glyceride SLN with little knowledge reported on wax carriers [25]. Glyceride SLN showed good-drug encapsulation, while physical stability was poor. In contrast, wax SLN possessed good physical stability but lacked sufficient drug encapsulation in the solidified state. These differences were attributed in part to different crystal packing. Less ordered crystal lattices favour successful drug inclusion while highly ordered crystal packing as in wax SLN leads to drug expulsion, but possess superior physical stability. The SLN formulated with this lipid matrix containing beeswax and phospholipid is ideal for achieving both good physical stability and possible high-drug encapsulation efficiency considering the obtained WAXD crystal characteristics. Diffractogram obtained for lipid matrix containing 1:1 beeswax/goat fat is also presented in Fig. 2. There was a slight ˚ increase in the intensity of the reflection at 2θ = 6.0◦ d = 14.73 A in the mixed lipid matrix containing phospholipid due to the goat fat content. The almost diffuse nature of the reflections presented by 1:1 lipid mixture containing P90G between 2θ = 15.0◦ and 22.8◦ compared with the diffractogram of beeswax or beeswax containing P90G shows reduced crystallinity, and confirms some degree of matrix modification on addition of goat fat. This lipid matrix containing P90G would offer more spaces for drug encapsulation compared with beeswax containing P90G. WAXD diffractograms obtained for SLN prepared with mixed lipid matrix with or without P90G (SLN-A to E) after 3 months are presented in Fig. 3b. All the SLN presented weak and very weak reflections 48 h, 1 week and 1 month after preparation (figure not shown). These reflections remained at the same positions and intensities after 3 months of preparation (Fig. 3b). The reflections occurred ˚ 2θ = 21.5◦ d = 4.13 A ˚ and 2θ = 23.9◦ at 2θ = 19.5◦ d = 4.55 A, ◦ ˚ for SLN-A; 2θ = 6.0 d = 14.73 A, ˚ 2θ = 19.4◦ d = 3.72 A ◦ ˚ and 2θ = 21.5 d = 4.13 A ˚ for SLN-B; 2θ = 19.4◦ d = 4.58 A ◦ ˚ 2θ = 21.5 d = 4.13 A ˚ and 2θ = 23.9◦ d = 3.72 A ˚ d = 4.58 A, ◦ ˚ 2θ = 21.4◦ d = 4.15 A ˚ and for SLN-C; 2θ = 19.3 d = 4.60 A, ˚ for SLN-D; and 2θ = 19.4◦ d = 4.58 A, ˚ 2θ = 23.8◦ d = 3.74 A ◦ ˚ and 2θ = 23.9◦ d = 3.72 A ˚ for SLN-E. The 2θ = 21.5 d = 4.13 A d spacings of these reflections are very close to the spacings of the reflections of the stable form of the mixed lipid matrix (Fig. 2), and could be concluded that the SLN recrystallized in the stable modification with less crystalline order compared with SLN prepared with beeswax or beeswax containing P90G. The fact that the SLN recrystallized in the stable modification within 48 h and remained in that modification for 3 months indicates that transition to highly ordered lipid particles would not occur on storage. Any incorporated drug would as a result, remain in the particle and the objective of increased drug incorporation capacity of SLN drug delivery system could be realized in this mixed lipid matrix containing phospholipid. 3.3. Differential scanning calorimetry (DSC) DSC was used to analyse the degree of crystallinity of the nanoparticles. The thermotropic phase behaviour of a lipid matrix system is highly affected by the presence of guest
194
A.A. Attama, C.C. M¨uller-Goymann / Colloids and Surfaces A: Physicochem. Eng. Aspects 315 (2008) 189–195
80 in comparison with SLN containing polysorbate 80 and no P90G where acceleration of the recrystallization in the stable modification occurred. For the SLN containing mixed lipids, very little thermal events could only be detected after 1 month of storage as shown in Fig. 4b. Their transition enthalpies could not be obtained because of the low heat of transition and unsteady baselines. This indicates overall crystallinity was low in accordance with the transition enthalpies too low to be determined. It shows that SLN-A to E prepared with mixed lipid matrix were less crystalline than SLN-1 to 3 prepared with beeswax containing P90G and very much less crystalline than SLN-4 prepared with beeswax without P90G. 4. Conclusions
Fig. 4. (a) DSC traces obtained for SLN prepared with 30% (w/w) P90G in beeswax: SLN-1 to 3 contained 0.3, 0.6 and 1.0% (w/w) polysorbate 80, respectively, while SLN-4 contained 1.0% polysorbate 80 but no P90G. (b) DSC traces obtained for SLN prepared with 30% (w/w) P90G in 1:1 mixture of beeswax and goat fat: SLN-A contained 1.0% polysorbate 80 but no P90G; SLN-B contained P90G but no polysorbate 80; SLN-C to E contained 0.3, 0.6 and 1.0% (w/w) polysorbate 80, respectively.
molecules, and the related thermodynamic variables (melting temperature and enthalpy changes) depend on the nature of the interaction between the constituents. Fig. 4a and b shows the DSC thermograms of the SLN. SLN prepared with beeswax or beeswax containing P90G presented peak melting transitions at 61.8 ± 0.8, 60.7 ± 0.7, 61.7 ± 0.6 and 63.2 ± 0.5 ◦ C (n = 3) for SLN-1, SLN-2, SLN-3 and SLN-4, respectively, which are close to the melting peaks of pure beeswax. However, SLN-4 without P90G had melting peak about that of beeswax showing P90G had a significant influence (p < 0.05) on the melting peaks and thus, the crystalline states of SLN-1 to 3. The enthalpies evaluated for these transitions were 3.84 ± 0.87, 3.62 ± 0.65, 3.76 ± 0.75 and 9.48 ± 0.23 mJ/mg, respectively, for SLN-1, SLN-2, SLN3 and SLN-4. Melting enthalpy of SLN-4 containing no P90G was significantly higher (p < 0.05) than SLN-1 to 3 with P90G indicating higher crystallinity. Considering the melting point and enthalpy of pure beeswax (ca. 63 ◦ C and 189 mJ/mg), the 3.5% (w/w) of beeswax within the nanosuspensions possessed melting enthalpies, which were less than the ideal value except the SLN prepared without P90G. There was thus suppression of recrystallization in SLN containing both P90G and polysorbate
This study has demonstrated that beeswax could be very useful in the preparation of stable lipid nanoparticles when combined with phospholipid with or without goat fat. In the presence of goat fat and/or P90G, there was reduction in crystallinity of the resulting lipid matrix and SLN formulated therefrom. This is favourable especially in lipid microparticulate or nanoparticulate drug delivery systems where drug incorporation is difficult because of high crystallinity of high-purity lipids available. The SLN produced had good particle size stability with low z-average and PI, and high negative potential at higher mobile surfactant content. SLN prepared with mixed lipid consisting of 1:1 beeswax and goat fat at 0.6% (w/w) and 1.0% (w/w) polysorbate 80 concentration possessed high-particle size stability comparable to SLN containing beeswax and P90G at equivalent polysorbate 80 concentration. These SLN could be further explored for use as drug delivery systems for intravascular and extravascular applications. Because of lower crystallinity of the mixed lipid nanoparticles, they will perform better as lipid nanoparticle drug delivery systems as they will hold more drugs compared with single lipid nanoparticles. Acknowledgements Dr. Attama wishes to acknowledge the support of Alexander von Humboldt Stiftung, Germany. We also thank Phospholipid GmbH, K¨oln for donation of Phospholipon 90G® . References [1] R.H. M¨uller, W. Mehnert, J.-S. Lucks, C. Schwarz, A. zur M¨uhlen, H. Weyhers, C. Freitas, D. R¨uhl, Eur. J. Pharm. Biopharm. 41 (1995) 62. [2] S.A. Wissing, O. Kayser, R.H. M¨uller, Adv. Drug Deliv. Rev. 56 (2004) 1257. [3] A. Jennings, S. Gohla, Int. J. Pharm. 196 (2000) 219. [4] A. Dingler, Feste Lipid Nanopartikel als Kolloidale Wirstofftr¨agersysteme zur dermalen Applikation, Ph.D. Thesis, Frei Universit¨at Berlin, 1998. [5] M.A. Schubert, B.C. Schicke, C.C. M¨uller-Goymann, Int. J. Pharm. 298 (2005) 242. [6] M.A. Schubert, C.C. M¨uller-Goymann, Eur. J. Pharm. Biopharm. 61 (2005) 77. [7] S.K. Huang, E. Mayhew, S. Gilani, D.D. Lasic, F.J. Martin, D. Papahadjopoulos, Cancer Res. 52 (1992) 6774.
A.A. Attama, C.C. M¨uller-Goymann / Colloids and Surfaces A: Physicochem. Eng. Aspects 315 (2008) 189–195 [8] A. Gabizon, R. Catane, B. Uziely, B. Kaufman, T. Safra, R. Cohen, F. Martin, A. Huang, Y. Barenholz, Cancer Res. 54 (1994) 987. [9] H. Heiati, N.C. Phillips, R. Tawashi, Pharm. Res. 13 (1996) 1406. [10] M. Sastry, Curr. Sci. 78 (2000) 1089. [11] J.L. Bernal, J.J. Jim´enez, M.J. del Noza, L. Toribo, M.T. Mart´ın, Eur. J. Lipid Sci. Technol. 107 (2005) 158. [12] T. Kameda, J. Insect Sci. 4 (2004) 1. [13] A.A. Attama, C.C. M¨uller-Goymann, Int. J. Pharm. 322 (2006) 67. [14] A.A. Attama, M.O. Nkemnele, Int. J. Pharm. 304 (2005) 4. [15] K. Jores, W. Mehnert, M. Drechsler, H. Bunjes, C. Johann, K. M¨ader, J. Control Release 95 (2004) 217. [16] Zetasizer Nano Series, Malvern Instruments England, User Manual Issue 2.2, 2005. [17] A.A. Attama, B.C. Schicke, T. Paepenm¨uller, C.C. M¨uller-Goymann, Eur. J. Pharm. Biopharm. 67 (2007) 48.
195
[18] M. Radtke, E.B. Souto, R.H. M¨uller, Pharm. Technol. Europe 17 (2005) 45. [19] J.L. Hargrove, P. Greenspan, D.K. Hartle, Exp. Biol. Med. 229 (2004) 215. [20] W.B. Rizzo, Mol. Genet. Metab. 65 (1998) 63. [21] W.B. Rizzo, D.A. Craft, A.L. Dammann, M.W. Phillips, J. Biol. Chem. 262 (1987) 17412. [22] A.A. Attama, B.C. Schicke, C.C. M¨uller-Goymann, Eur. J. Pharm. Biopharm. 64 (2006) 294. [23] V. Luzzati, X-ray diffraction of lipid-water systems, in: D. Chapman (Ed.), Biological Membranes Physical Facts and Function, Academic Press, London, 1968, pp. 71–123. ˇ ak, Biomed. Pap. 145 (2001) 17. [24] M. Stuchl´ık, S. Z´ [25] C.C. M¨uller-Goymann, Eur. J. Pharm. Biopharm. 58 (2004) 343.
Colloids and Surfaces A: Physicochem. Eng. Aspects 315 (2008) 189–195
Effect of beeswax modification on the lipid matrix and solid lipid nanoparticle crystallinity Anthony A. Attama a,b,∗ , Christel C. M¨uller-Goymann a,1 a
Institut f¨ur Pharmazeutische Technologie, Technische Universit¨at Carolo-Wilhelmina zu Braunschweig, Mendelssohnstraße 1, D-38106 Braunschweig, Germany b Department of Pharmaceutics, University of Nigeria, Nsukka 410001, Enugu State, Nigeria Received 24 April 2007; received in revised form 10 July 2007; accepted 26 July 2007 Available online 8 August 2007
Abstract The influence of a heterolipid (Phospholipon 90G® , P90G), which directly modifies the surface of solid lipid nanoparticles (SLN), and goat fat on the crystallinity of beeswax matrix and the SLN prepared therefrom was studied. Lipid matrices composed of 30% (w/w) of P90G in beeswax and in 1:1 mixture of beeswax and goat fat were formulated and characterized. SLN containing polysorbate 80 with or without P90G were formulated by hot high-pressure homogenisation and characterized by particle size and zeta potential measurements. Differential scanning calorimetry (DSC) and wide-angle X-ray diffraction (WAXD) patterns of the SLN were compared with the lipid matrices. Results showed WAXD of the lipid matrices containing P90G contain amorphous portions. Most SLN formulated possessed low z-average diameters and polydispersity indices, which remained constant after 3 months, with high-negative potentials after 4 weeks and crystallized into stable modification within 48 h of preparation. The overall result indicated that P90G and goat fat influenced the crystallinity of beeswax matrix and its SLN, but the crystallinity of the mixed lipid matrix (1:1 beeswax/goat fat) and its SLN, beeswax containing P90G and its SLN was lower than that of beeswax alone and its SLN. There was no increase in crystallinity of SLN on storage. Modification of beeswax with P90G or goat fat offers a way of improving the SLN formulated with beeswax in terms of reduction of its crystallinity responsible for its low-drug incorporation efficiency. © 2007 Elsevier B.V. All rights reserved. Keywords: Beeswax; Goat fat; Mixed lipid; Phospholipon 90G® ; Modification; Solid lipid nanoparticles; Crystallinity; Drug delivery
1. Introduction
Abbreviations: SLN-1, SLN prepared with 30% (w/w) P90G in beeswax as lipid matrix and 0.3% (w/w) polysorbate 80; SLN-2, SLN prepared with 30% (w/w) P90G in beeswax as lipid matrix and 0.6% (w/w) polysorbate 80; SLN-3, SLN prepared with 30% (w/w) P90G in beeswax as lipid matrix and 1.0% (w/w) polysorbate 80; SLN-4, SLN prepared with beeswax as lipid matrix and 1.0% (w/w) polysorbate 80; SLN-A, SLN prepared with 1:1 mixture of beeswax and goat fat as lipid matrix and 1.0% (w/w) polysorbate 80; SLN-B, SLN prepared with 30% (w/w) P90G in 1:1 mixture of beeswax and goat fat as lipid matrix without polysorbate 80; SLN-C, SLN prepared with 30% (w/w) P90G in 1:1 mixture of beeswax and goat fat as lipid matrix and 0.3% (w/w) polysorbate 80; SLN-D, SLN prepared with 30% (w/w) P90G in 1:1 mixture of beeswax and goat fat as lipid matrix and 0.6% (w/w) polysorbate 80; SLN-E, SLN prepared with 30% (w/w) P90G in 1:1 mixture of beeswax and goat fat as lipid matrix 1.0% (w/w) polysorbate 80 ∗ Corresponding author at: Department of Pharmaceutics, University of Nigeria, Nsukka 410001, Nigeria. Tel.: +234 42 771911; fax: +234 42 771709. E-mail address: [email protected] (A.A. Attama). 1 Tel.: +49 531 3915654; fax: +49 531 3918108. 0927-7757/$ – see front matter © 2007 Elsevier B.V. All rights reserved. doi:10.1016/j.colsurfa.2007.07.035
Solid lipid nanoparticles (SLN) are an alternative carrier system to polymer nanoparticles or liposomes. They consist of physiological and biocompatible lipids, which are suitable for the incorporation of lipophilic, hydrophilic and poorly watersoluble active ingredients. Improved bioavailability, protection of sensitive drug molecules from the outer environment (water and light) and controlled release characteristics were claimed by incorporation of drugs in the solid lipid matrix [1]. Due to the rigidity of the nanoparticles, the mobility of an incorporated drug is reduced, preventing drug leakage. A major disadvantage of SLN is their inherent low-drug incorporation due to the crystalline structure of the solid lipid. The high crystalline order of nanoparticles prevents incorporation of high amounts of drug molecules, but the use of triglyceride mixtures results in particles with a less ordered matrix [2]. Moreover, addition of a second matrix component may specifically alter the crystallization behaviour of the lipid
190
A.A. Attama, C.C. M¨uller-Goymann / Colloids and Surfaces A: Physicochem. Eng. Aspects 315 (2008) 189–195
matrix. In objective of this study, beeswax was modified with goat fat and the nanoparticles formulated therefrom studied. The effect of the modification on crystallinity of the P90G or polysorbate 80 stabilized SLN formulated was studied. Properties of the SLN prepared with beeswax containing P90G were compared with SLN prepared with 1:1 mixture of beeswax and goat fat. Beeswax SLN containing polysorbate 80 as the only surfactant was compared with other triglycerides and it was found that the beeswax nanoparticles possessed excellent particle size stability but poor drug encapsulation efficiency despite the fact that polymorphism of waxes is very low compared with triglycerides [3,4]. Phospholipon 90G® has been used in parenteral emulsions and in formulation of liposomes, and was recently shown to be a good surface modifier for SLN [5,6] with resultant improvement in targeting and pharmacokinetics [7–9]. The phospholipid bilayer structure formed around the lipid core may increase the drug loading capacity, as biologically important molecules can be anchored on the colloidal particle surface, and surfacemodification also enables stabilization of colloidal particles especially when generation of the nanoparticles is carried out in an aqueous medium [10]. Beeswax is a natural product used in pharmaceutical, cosmetics, food and other industries [11]. It is highly crystalline [12], and this adversely affects its drug holding capacity. Addition of goat fat and phospholipid (heterolipid) is expected to greatly disturb its crystal order in addition to surface modification. Goat fat on the other hand, is derived from a domestic animal, Capra hircus and has been characterized and used in experimental drug delivery systems [13,14]. 2. Materials and experiments 2.1. Materials Beeswax (Cera alba) Ph. Eur., sorbitol (Caesar & Loretz, Hilden Germany), thimerosal (Synochem, Germany) and polysorbate 80 (Tween 80® , Across Organics, Germany), Phospholipon 90G® (Phospholipid GmbH K¨oln, Germany) were used as procured. Goat fat was obtained from a batch processed according to earlier procedure [13]. Double-distilled water was used for nanoparticle preparation. 2.2. Preparation of lipid matrix containing Phospholipon 90G® The lipid matrix composition was chosen based on previous experience with SLN prepared and evaluated in our laboratory [5,6], and contained 30% (w/w) of Phospholipon 90G® in beeswax or 1:1 mixture of beeswax and goat fat, and was prepared initially by fusion prior to nanoparticle preparation. 2.3. Preparation of nanoparticles Solid lipid nanoparticles were prepared to contain 5.0% (w/w) lipid matrix (30% (w/w) of P90G in beeswax and 30% (w/w) of P90G in 1:1 mixture of beeswax and goat fat), 0.3, 0.6
or 1.0% (w/w) polysorbate 80, 0.005% (w/w) thimerosal, 4.0% (w/w) sorbitol and enough double-distilled water to make 100% (w/w). SLN without P90G or polysorbate 80 were also prepared. The nomenclatures used for the different batches of SLN formulated are presented in the key below. Melt-homogenisation technique was adopted. In each case, the lipid matrix was melted at 75 ◦ C, which is more than 10 ◦ C above the melting temperature and avoids the lipid memory effect and makes new crystallization possible [15]. The double-distilled water containing the polysorbate 80, thimerosal and sorbitol at the same temperature, was added to the molten lipid matrix, mixed very well and dispersed at 75 ◦ C with Ultra-Turrax (T25 basic, Ika® Staufen Germany) at 24,000 rpm for 5 min and immediately passed through a heated high pressure homogeniser (EmulsiFlex-C5, Avestin Canada) at a pressure of 1000 bars for 20 cycles. 2.4. Photon correlation spectroscopy (PCS) Particle sizes were determined by PCS at 25 ◦ C using a Multi Angle Particle Size Analyser (Zetasizer 3 Model AZ6004, Malvern England) modified with a 35 mW He–Ne laser (Model 127-35, Spectra Physics USA). The detection was performed at a scattering angle of 90◦ in a cell AZ10 equilibrated at 293 K and at an accumulation time of 180 s. Samples were diluted with filtrated double-distilled water (0.2 m Sterifix® filter) and data were analysed by the cumulants method assuming spherical particles. Particle size studies of SLN were done 24 h, 1, 4 and 12 weeks after preparation. 2.5. Measurement of zeta potential The zeta potentials of the formulated SLN were determined after 1 month of preparation in a Zetasizer Nano Series (NanoZS, Malvern Instruments England). Each sample was diluted with bidistilled water and the electrophoretic mobility determined at 25 ◦ C and dispersant dielectric constant of 78.5 and pH of 7. The obtained electrophoretic mobility values were used to calculate the zeta potentials using the software DTS Version 4.1 (Malvern, England) and applying Henry equation [16]: UE =
2εZf (Ka ) 3η
(1)
where Z is the zeta potential, UE the electrophoretic mobility, ε the dielectric constant, η the viscosity of the medium and f(Ka ) is the Henry’s function. 2.6. Wide angle X-ray diffraction (WAXD) studies To study the crystalline characteristics, WAXD studies were done on the lipid matrices and all the SLN as earlier described [13,17]. Wide angle X-ray studies were done using an X-ray generator (PW3040/60 X’Pert PRO, Fabr. DY2171, PANalytical Netherlands) connected to a copper anode (PW3373/00 DK 147726 Cu LFF). WAXD diffractograms were obtained 3 months after lipid matrices preparation and 48 h, 1 week, 1 month and 3 months after SLN preparation.
A.A. Attama, C.C. M¨uller-Goymann / Colloids and Surfaces A: Physicochem. Eng. Aspects 315 (2008) 189–195
191
2.7. Differential scanning calorimetry (DSC) The degree of crystallinity of the nanoparticles was determined on a calorimeter (DSC 220C) connected to a disc station (5200H, Seiko, Tokyo, Japan). Approximately 5 mg of the SLN was weighed into an aluminium pan and sealed hermetically, and the thermal behaviour determined against an empty pan in the range of 5–125 ◦ C at a heating rate of 5 ◦ C min−1 . Transition temperatures were taken as the peak minima of the endothermic transitions, while transition enthalpies were determined by integration of the endothermic transitions using linear baselines. Measurement was done on the SLN after 1 month of preparation. 3. Results and discussion 3.1. Photon correlation spectroscopy (PCS) and zeta potential measurement The particle size distribution of the nanoparticles presented in Fig. 1a and b shows low-sized uniform nanoparticles were obtained, except for SLN prepared with mixed lipid and P90G only (SLN-B). The figure shows that increase in polysorbate 80 concentration reduced the particle size. However, the sizes of SLN-2 containing 0.6% (w/w) polysorbate 80 and SLN-3 containing 1.0% (w/w) polysorbate 80 were close, but the polydispersity indices (PI) were significantly different (p < 0.05) with PI of the former being consistently higher than those of the latter throughout the period of storage (Fig. 1a). Thus, SLN-3 had a narrower particle size distribution than SLN-2. Polydispersity index (PI) measures the width of the particle size distribution. Consistently high values of PI indicate either an aggregated or poorly prepared sample. If the PI is above 0.2, the PI loses its significance as an accurate measure of the width of the size of distribution, but can still be useful for comparative purposes. The lower PI of SLN-3 thus indicates a better control of particle size. In all the SLN, the z-average diameter and PI remained almost constant after 3 months, except in SLN-1 containing 0.3% (w/w) polysorbate 80, where there was an increase in PI after 12 weeks. Low standard deviations were obtained for all the SLN. Consistent with reported work [3], nanoparticles prepared with beeswax and polysorbate 80 alone (SLN-4) possessed good particle size stability with low PI compared with SLN-1 and SLN-2. But high crystallinity recorded in WAXD and DSC would preclude the use of SLN-4 because of difficulty in incorporation of drugs, although drugs could be coupled or adsorbed to the surface but may not achieve sustained release effect. The particle size stability coupled with low degree of crystallinty obtained for SLN-2 and SLN-3 could make them good candidate drug delivery systems with improved drug incorporation capacity. Highly crystalline particles and lipid matrices lead to drug expulsion [18]. All the SLN containing both P90G and polysorbate 80 may be suitable for intravascular delivery systems because of their low size, stability and in vivo tolerability of the component lipids. Beeswax has been used as a thickener and a humectant in the manufacture of ointments, creams, lipsticks and other cosmetics. However, recent studies have shown promising in vivo applications of beeswax and other long chain fatty alcohols and
Fig. 1. (a) Particle size distribution of SLN prepared with 30% (w/w) P90G in beeswax: SLN-1 to 3 contained 0.3, 0.6 and 1.0% (w/w) polysorbate 80, respectively, while SLN-4 contained 1.0% polysorbate 80 but no P90G. z-average and PI represent, respectively, z-average diameter and polydispersity index. (b) Particle size distribution of SLN prepared with 30% (w/w) P90G in 1:1 mixture of beeswax and goat fat: SLN-A contained 1.0% polysorbate 80 but no P90G; SLN-B contained P90G but no polysorbate 80; SLN-C to E contained 0.3, 0.6 and 1.0% (w/w) polysorbate 80, respectively. z-average and PI represent respectively, z-average diameter and polydispersity index.
acids [19–21], thus parenteral administration of this dosage form would not pose any metabolic problem. SLN prepared with mixed lipid matrix (SLN-A to E) produced lower z-average sized particles when P90G and polysorbate 80 were incorporated (SLN-D and SLN-E) compared with SLN prepared with beeswax alone at polysorbate concentrations of 0.6% (w/w) and 1.0% (w/w), but with higher PI (Fig. 1b). This was due to greater ease of particle disintegration and slower recrystallization in the mixed lipid as its melting point and crystallinity were lower than that of beeswax. SLN containing only beeswax and P90G could not be prepared. SLN-B
192
A.A. Attama, C.C. M¨uller-Goymann / Colloids and Surfaces A: Physicochem. Eng. Aspects 315 (2008) 189–195
prepared with mixed lipid matrix and P90G without polysorbate 80 showed high growth in particles up to micrometer scale within 12 weeks. Addition of a mobile surfactant is necessary for stabilization of the nanoparticles containing phospholipid. In terms of controlled and sustained or prolonged drug delivery, SLN prepared with beeswax containing both P90G and polysorbate 80 could be recommended because of the more crystalline nature compared with the mixed lipid matrix SLN. However, when high drug incorporation is needed, SLN formulated with mixed lipid core containing P90G and polysorbate 80 will be better as they were shown to possess low crystalline order which favours drug loading. All the same, both SLN proved better than SLN formulated with beeswax and polysorbate 80. They can be used in active or passive drug targeting depending on the intentions of the formulator. The nanocarrier formulated with 0.0% (w/w) of polysorbate 80 (SLN-B) exhibited gelation with storage. Although gelation was observed here, there was no increase in crystallinity detected by either DSC or WAXD providing evidence that the gelation was not a consequence of alkyl chain ordering (increase in crystallinity) and that a mobile surfactant is required to achieve optimum particle size stability in phospholipid-stabilized lipid nanoparticles. Insofar as there would be no drug expulsion in this SLN during storage, it is not suitable for use as drug delivery carrier because of high growth in particle size. Since there were neither increase in crystallinity nor gelation in the more stable SLN formulated with mixed lipid matrix (SLN-C to E), the novel mixed matrix lipid nanocarriers could overcome the low drug incorporation capacity of crystalline lipids such as beeswax, which is a drawback of lipid particulate drug delivery systems. Evaluation of the drug incorporation efficiency of these optimised SLN is planned for the next stage of this investigation. The zeta potential values obtained were −20.2 ± 5.3, −27.9 ± 8.3, −32.2 ± 6.3 and −34.3 ± 0.49 mV for SLN-1 to 4 respectively; and −20.0 ± 1.2, −19.2 ± 0.6, −15.9 ± 0.2, −18.4 ± 0.3 and −15.4 ± 0.6 mV for SLN-A to E, respectively, with SLN-E having the lowest absolute value and SLN-4 the highest. The magnitude of zeta potential gives an indication of the potential stability of a colloidal system. Absolute large negative or positive zeta potential is required for colloidal dispersion stability. The general dividing line between stable and unstable suspension is generally taken as either +30 or −30 mV [16]. Based on the zeta potentials obtained for the SLN in distilled water, the dispersion stability of the SLN could be stated in the following descending rank order: SLN-4 > SLN-3 > SLN2 > SLN-1 ≈ SLN-A > SLN-B ≈ SLN-D > SLN-C ≈ SLN-E. Although SLN-4 had highest absolute zeta potential value (highest stability), other properties associated with this SLN would preclude its use in formulation. The overall negative zeta potentials of all the SLN would also contribute to stability of the lipid nanodispersions, and provide surfaces for attachment of drugs and other biomacromolecules. The fact that SLN-C, SLN-D and SLN-E with low-absolute zeta potential values showed good particle size stability indicate P90G had a stabilizing effect at the interface. SLN formulated with mixed lipid possessed lower absolute zeta potential values compared with SLN prepared with beeswax alone with or without P90G.
Fig. 2. WAXD of lipid matrices: Beeswax containing 30% (w/w) P90G (BWP90G), 1:1 mixture of beeswax and goat fat (1:1 LM), lipid matrix containing 1:1 mixture of beeswax and goat fat and 30% (w/w) P90G (1:1 LM-P90G).
This effect may be related to the shape of the lipid particles but this has to be proved in a further investigation. 3.2. Wide angle X-ray diffraction (WAXD) studies WAXD was used to study the crystalline character of the lipid matrices and the SLN. Sharp high intensity reflections are produced by highly crystalline lipids which appear above the amorphous background of non-crystalline lipids. Fig. 2 shows the WAXD diffractograms of the lipid matrices. Pure beeswax reflections were sharp and conformed to the result of earlier study [22]. Beeswax containing 30% (w/w) P90G presented ˚ (high intensity), 2θ = 23.7◦ reflections at 2θ = 21.4◦ d = 4.15 A ˚ (low intensity) and 2θ = 5.6◦ d = 15.78 A ˚ and 2θ = 7.0◦ d = 3.75 A ˚ (very weak intensities). There is an amorphous pord = 12.63 A tion around 2θ = 19.0◦ . The slight amorphous portion present in the diffractogram of the beeswax containing phospholipid is not present in the diffractogram obtained for pure beeswax. This shows that the former matrix is less crystalline than the latter. It is thus expected that the amorphous portion would home any incorporated drug as there are spaces where drug could be incorporated, compared with pure beeswax which is more crystalline. Also the presence of higher intensity sharp reflections in the diffractogram of beeswax at the wide angle region indicates that most of the hydrocarbon chains of beeswax are highly ordered [23], in comparison with the reflections obtained for the lipid matrix containing phospholipid, attesting that beeswax has higher crystallinity. Phospholipon 90G® used in this study mainly contains linoleic, oleic, stearic and palmitic acids, which are fatty acids of different chain lengths and degree of saturation [24]. The interaction of these fatty acids with the diverse fatty acids present in beeswax [11,12] resulted in the partly amorphous nature of the lipid matrix containing the phospholipid. Reflections of medium and low intensities detected for the SLN prepared with beeswax containing P90G or beeswax ˚ and alone (SLN-1 to 4), respectively, at 2θ = 21.4◦ d = 4.15 A
A.A. Attama, C.C. M¨uller-Goymann / Colloids and Surfaces A: Physicochem. Eng. Aspects 315 (2008) 189–195
Fig. 3. (a) WAXD of SLN prepared with 30% (w/w) P90G in beeswax after 3 months: SLN-1 to 3 contained 0.3, 0.6 and 1.0% (w/w) polysorbate 80, respectively, while SLN-4 contained 1.0% polysorbate 80 but no P90G. (b) WAXD of SLN prepared with 30% (w/w) P90G in 1:1 mixture of beeswax and goat fat after 3 months: SLN-A contained 1.0% polysorbate 80 but no P90G; SLN-B contained P90G but no polysorbate 80; SLN-C to E contained 0.3, 0.6 and 1.0% (w/w) polysorbate 80, respectively.
˚ after 48 h, 1 week and 1 month (figures not 2θ = 23.8◦ d = 3.74 A shown) remained unchanged after 3 months (Fig. 3a). Absence of growth in intensities of these reflections on storage indicates absence of any dynamic polymorphic transformation, and that progression to increasingly ordered nanoparticles due to polymorphic transition would not occur on storage. In comparison, the intensities of the reflections of SLN-1 to 3 containing P90G were significantly lower (p < 0.05) than those of SLN-4 prepared with beeswax and polysorbate 80 only. In this study, the reflections detected for SLN-1 to 4 were in the same modification as the reflections detected for beeswax containing phospholipid attesting that no modification of the wax esters occurred in the nanoparticles. Thus, any included drug would not be expelled from the nanoparticles as a result of change in modification leading to a change in crystal order. Although the crystal order was
193
greatly disturbed in SLN-1 to 3, these drug carriers remained solid. Most published works are on glyceride SLN with little knowledge reported on wax carriers [25]. Glyceride SLN showed good-drug encapsulation, while physical stability was poor. In contrast, wax SLN possessed good physical stability but lacked sufficient drug encapsulation in the solidified state. These differences were attributed in part to different crystal packing. Less ordered crystal lattices favour successful drug inclusion while highly ordered crystal packing as in wax SLN leads to drug expulsion, but possess superior physical stability. The SLN formulated with this lipid matrix containing beeswax and phospholipid is ideal for achieving both good physical stability and possible high-drug encapsulation efficiency considering the obtained WAXD crystal characteristics. Diffractogram obtained for lipid matrix containing 1:1 beeswax/goat fat is also presented in Fig. 2. There was a slight ˚ increase in the intensity of the reflection at 2θ = 6.0◦ d = 14.73 A in the mixed lipid matrix containing phospholipid due to the goat fat content. The almost diffuse nature of the reflections presented by 1:1 lipid mixture containing P90G between 2θ = 15.0◦ and 22.8◦ compared with the diffractogram of beeswax or beeswax containing P90G shows reduced crystallinity, and confirms some degree of matrix modification on addition of goat fat. This lipid matrix containing P90G would offer more spaces for drug encapsulation compared with beeswax containing P90G. WAXD diffractograms obtained for SLN prepared with mixed lipid matrix with or without P90G (SLN-A to E) after 3 months are presented in Fig. 3b. All the SLN presented weak and very weak reflections 48 h, 1 week and 1 month after preparation (figure not shown). These reflections remained at the same positions and intensities after 3 months of preparation (Fig. 3b). The reflections occurred ˚ 2θ = 21.5◦ d = 4.13 A ˚ and 2θ = 23.9◦ at 2θ = 19.5◦ d = 4.55 A, ◦ ˚ for SLN-A; 2θ = 6.0 d = 14.73 A, ˚ 2θ = 19.4◦ d = 3.72 A ◦ ˚ and 2θ = 21.5 d = 4.13 A ˚ for SLN-B; 2θ = 19.4◦ d = 4.58 A ◦ ˚ 2θ = 21.5 d = 4.13 A ˚ and 2θ = 23.9◦ d = 3.72 A ˚ d = 4.58 A, ◦ ˚ 2θ = 21.4◦ d = 4.15 A ˚ and for SLN-C; 2θ = 19.3 d = 4.60 A, ˚ for SLN-D; and 2θ = 19.4◦ d = 4.58 A, ˚ 2θ = 23.8◦ d = 3.74 A ◦ ˚ and 2θ = 23.9◦ d = 3.72 A ˚ for SLN-E. The 2θ = 21.5 d = 4.13 A d spacings of these reflections are very close to the spacings of the reflections of the stable form of the mixed lipid matrix (Fig. 2), and could be concluded that the SLN recrystallized in the stable modification with less crystalline order compared with SLN prepared with beeswax or beeswax containing P90G. The fact that the SLN recrystallized in the stable modification within 48 h and remained in that modification for 3 months indicates that transition to highly ordered lipid particles would not occur on storage. Any incorporated drug would as a result, remain in the particle and the objective of increased drug incorporation capacity of SLN drug delivery system could be realized in this mixed lipid matrix containing phospholipid. 3.3. Differential scanning calorimetry (DSC) DSC was used to analyse the degree of crystallinity of the nanoparticles. The thermotropic phase behaviour of a lipid matrix system is highly affected by the presence of guest
194
A.A. Attama, C.C. M¨uller-Goymann / Colloids and Surfaces A: Physicochem. Eng. Aspects 315 (2008) 189–195
80 in comparison with SLN containing polysorbate 80 and no P90G where acceleration of the recrystallization in the stable modification occurred. For the SLN containing mixed lipids, very little thermal events could only be detected after 1 month of storage as shown in Fig. 4b. Their transition enthalpies could not be obtained because of the low heat of transition and unsteady baselines. This indicates overall crystallinity was low in accordance with the transition enthalpies too low to be determined. It shows that SLN-A to E prepared with mixed lipid matrix were less crystalline than SLN-1 to 3 prepared with beeswax containing P90G and very much less crystalline than SLN-4 prepared with beeswax without P90G. 4. Conclusions
Fig. 4. (a) DSC traces obtained for SLN prepared with 30% (w/w) P90G in beeswax: SLN-1 to 3 contained 0.3, 0.6 and 1.0% (w/w) polysorbate 80, respectively, while SLN-4 contained 1.0% polysorbate 80 but no P90G. (b) DSC traces obtained for SLN prepared with 30% (w/w) P90G in 1:1 mixture of beeswax and goat fat: SLN-A contained 1.0% polysorbate 80 but no P90G; SLN-B contained P90G but no polysorbate 80; SLN-C to E contained 0.3, 0.6 and 1.0% (w/w) polysorbate 80, respectively.
molecules, and the related thermodynamic variables (melting temperature and enthalpy changes) depend on the nature of the interaction between the constituents. Fig. 4a and b shows the DSC thermograms of the SLN. SLN prepared with beeswax or beeswax containing P90G presented peak melting transitions at 61.8 ± 0.8, 60.7 ± 0.7, 61.7 ± 0.6 and 63.2 ± 0.5 ◦ C (n = 3) for SLN-1, SLN-2, SLN-3 and SLN-4, respectively, which are close to the melting peaks of pure beeswax. However, SLN-4 without P90G had melting peak about that of beeswax showing P90G had a significant influence (p < 0.05) on the melting peaks and thus, the crystalline states of SLN-1 to 3. The enthalpies evaluated for these transitions were 3.84 ± 0.87, 3.62 ± 0.65, 3.76 ± 0.75 and 9.48 ± 0.23 mJ/mg, respectively, for SLN-1, SLN-2, SLN3 and SLN-4. Melting enthalpy of SLN-4 containing no P90G was significantly higher (p < 0.05) than SLN-1 to 3 with P90G indicating higher crystallinity. Considering the melting point and enthalpy of pure beeswax (ca. 63 ◦ C and 189 mJ/mg), the 3.5% (w/w) of beeswax within the nanosuspensions possessed melting enthalpies, which were less than the ideal value except the SLN prepared without P90G. There was thus suppression of recrystallization in SLN containing both P90G and polysorbate
This study has demonstrated that beeswax could be very useful in the preparation of stable lipid nanoparticles when combined with phospholipid with or without goat fat. In the presence of goat fat and/or P90G, there was reduction in crystallinity of the resulting lipid matrix and SLN formulated therefrom. This is favourable especially in lipid microparticulate or nanoparticulate drug delivery systems where drug incorporation is difficult because of high crystallinity of high-purity lipids available. The SLN produced had good particle size stability with low z-average and PI, and high negative potential at higher mobile surfactant content. SLN prepared with mixed lipid consisting of 1:1 beeswax and goat fat at 0.6% (w/w) and 1.0% (w/w) polysorbate 80 concentration possessed high-particle size stability comparable to SLN containing beeswax and P90G at equivalent polysorbate 80 concentration. These SLN could be further explored for use as drug delivery systems for intravascular and extravascular applications. Because of lower crystallinity of the mixed lipid nanoparticles, they will perform better as lipid nanoparticle drug delivery systems as they will hold more drugs compared with single lipid nanoparticles. Acknowledgements Dr. Attama wishes to acknowledge the support of Alexander von Humboldt Stiftung, Germany. We also thank Phospholipid GmbH, K¨oln for donation of Phospholipon 90G® . References [1] R.H. M¨uller, W. Mehnert, J.-S. Lucks, C. Schwarz, A. zur M¨uhlen, H. Weyhers, C. Freitas, D. R¨uhl, Eur. J. Pharm. Biopharm. 41 (1995) 62. [2] S.A. Wissing, O. Kayser, R.H. M¨uller, Adv. Drug Deliv. Rev. 56 (2004) 1257. [3] A. Jennings, S. Gohla, Int. J. Pharm. 196 (2000) 219. [4] A. Dingler, Feste Lipid Nanopartikel als Kolloidale Wirstofftr¨agersysteme zur dermalen Applikation, Ph.D. Thesis, Frei Universit¨at Berlin, 1998. [5] M.A. Schubert, B.C. Schicke, C.C. M¨uller-Goymann, Int. J. Pharm. 298 (2005) 242. [6] M.A. Schubert, C.C. M¨uller-Goymann, Eur. J. Pharm. Biopharm. 61 (2005) 77. [7] S.K. Huang, E. Mayhew, S. Gilani, D.D. Lasic, F.J. Martin, D. Papahadjopoulos, Cancer Res. 52 (1992) 6774.
A.A. Attama, C.C. M¨uller-Goymann / Colloids and Surfaces A: Physicochem. Eng. Aspects 315 (2008) 189–195 [8] A. Gabizon, R. Catane, B. Uziely, B. Kaufman, T. Safra, R. Cohen, F. Martin, A. Huang, Y. Barenholz, Cancer Res. 54 (1994) 987. [9] H. Heiati, N.C. Phillips, R. Tawashi, Pharm. Res. 13 (1996) 1406. [10] M. Sastry, Curr. Sci. 78 (2000) 1089. [11] J.L. Bernal, J.J. Jim´enez, M.J. del Noza, L. Toribo, M.T. Mart´ın, Eur. J. Lipid Sci. Technol. 107 (2005) 158. [12] T. Kameda, J. Insect Sci. 4 (2004) 1. [13] A.A. Attama, C.C. M¨uller-Goymann, Int. J. Pharm. 322 (2006) 67. [14] A.A. Attama, M.O. Nkemnele, Int. J. Pharm. 304 (2005) 4. [15] K. Jores, W. Mehnert, M. Drechsler, H. Bunjes, C. Johann, K. M¨ader, J. Control Release 95 (2004) 217. [16] Zetasizer Nano Series, Malvern Instruments England, User Manual Issue 2.2, 2005. [17] A.A. Attama, B.C. Schicke, T. Paepenm¨uller, C.C. M¨uller-Goymann, Eur. J. Pharm. Biopharm. 67 (2007) 48.
195
[18] M. Radtke, E.B. Souto, R.H. M¨uller, Pharm. Technol. Europe 17 (2005) 45. [19] J.L. Hargrove, P. Greenspan, D.K. Hartle, Exp. Biol. Med. 229 (2004) 215. [20] W.B. Rizzo, Mol. Genet. Metab. 65 (1998) 63. [21] W.B. Rizzo, D.A. Craft, A.L. Dammann, M.W. Phillips, J. Biol. Chem. 262 (1987) 17412. [22] A.A. Attama, B.C. Schicke, C.C. M¨uller-Goymann, Eur. J. Pharm. Biopharm. 64 (2006) 294. [23] V. Luzzati, X-ray diffraction of lipid-water systems, in: D. Chapman (Ed.), Biological Membranes Physical Facts and Function, Academic Press, London, 1968, pp. 71–123. ˇ ak, Biomed. Pap. 145 (2001) 17. [24] M. Stuchl´ık, S. Z´ [25] C.C. M¨uller-Goymann, Eur. J. Pharm. Biopharm. 58 (2004) 343.