Article pubs.acs.org/Langmuir
Characterization of Bupivacaine-Loaded Formulations Based on Liquid Crystalline phases and Microemulsions: The Effect of Lipid Composition Anan Yaghmur,*,‡ Michael Rappolt,† Jesper Østergaard,‡ Claus Larsen,‡ and Susan Weng Larsen‡ ‡
Department of Pharmaceutics and Analytical Chemistry, Faculty of Pharmaceutical Sciences, University of Copenhagen, Universitetsparken 2, DK-2100 Copenhagen, Denmark † Institute of Biophysics and Nanosystems Research (IBN), Austrian Academy of Sciences, Graz, Austria ABSTRACT: This report details the structural characterization and the in vitro drugrelease properties of different local anesthetic bupivacaine (BUP)-loaded invertedtype liquid crystalline phases and microemulsions. The effects of variations in the lipid composition and/or BUP concentration on the self-assembled nanostructures were investigated in the presence of the commercial distilled glycerol monooleate Myverol 18−99K (GMO) and medium-chain triglycerides (MCT). Synchrotron small-angle X-ray scattering (SAXS) and rotating dialysis cell model were used to characterize the BUP formulations and to investigate the in vitro BUP release profiles, respectively. The evaluation of SAXS data for the BUP-loaded GMO/MCT formulations indicates the structural transition of inverted-type bicontinuous cubic phase of the symmetry Pn3m → inverted-type hexagonal (H2) phase → inverted-type microemulsion (L2) with increasing MCT content (0−40 wt %). In the absence of MCT, the solubilization of BUP induces the transition of Pn3m → H2 at pH 7.4; whereas a transition of Pn3m → (Pn3m + H2) is detected as the hydration is achieved at pH 6.0. To mimic the drug release and transport from in situ formed self-assembled systems after subcutaneous administration, the release experiments were performed by injecting low viscous stimulus-responsive precursors to a buffer in the dialysis cell leaving the surface area between the selfassembled system and the release medium variable. Our results suggest that the pH-dependent variations in the lipidic partition coefficient, Kl/w, between the liquid crystalline nanostructures and the surrounding buffer solution are significantly affecting BUP release rates. Thus, a first step toward understanding of the drug-release mechanism of this drug-delivery class has been undertaken tackling the influence of drug ionization as well as the type of the self-assembled nanostructure and its release kinetics under pharmaceutically relevant conditions.
1. INTRODUCTION In recent years, there has been a surge of interest in the formulation of efficient and biocompatible injectable nanocarriers providing sustained drug release at the administration site.1−5 A promising injectable-delivery approach applicable to drugs with different physicochemical properties and controlling their release is the in situ formation of high-viscous drug-loaded lyotropic nonlamellar liquid crystalline phases.4,6−11 These nanoobjects are formed at the administration site via the selfassembly of biologically relevant lipids having the tendency to form nonlamellar structures in response to the biological environmental stimuli including hydration and temperature.4,10−14 Herein, the propensity to form inverted nonplanar structures originates from phase transitions induced in lowviscous stimulus-responsive preformulations to excess water.3,4,8,11,15 For instance in our recent study,11 we investigated the in situ behavior of the local anesthetic bupivacaine (BUP)-loaded inverted micellar (L2) and inverted hexagonal (H2) phases (preformulations of BUP) based on glycerol monooleate (GMO) upon exposure to phosphate buffer under physiological conditions of pH 7.4 and 37 °C. A key characteristic of this delivery technology, where lipids upon hydration induce the formation of well-defined © 2012 American Chemical Society
nanostructures, is that sustained drug release can be accomplished in the absence of chemical modifications of the drug substance or the drug formulation and without the inclusion of temperature-sensitive gellators such as polymeric hydrogels. The growing interest in the utilization of the inverted-type liquid crystalline phases based on generally safe lipids, is driven also by the biological relevance.16−19 The observations of these curved membrane phases in biological systems are plenty, and concern, for example the occurrence of cubic membranes in cytomembranes and cubic cuticular structures in butterfly wing scales.20 The interested reader will find further appealing reports in Hyde et al.,19 Deng and co-workers,17,21 Michielsen and Stavenga,20 Larsson,14 and Luzzati.16 The vast majority of experimental studies have focused on the formation of sustained-release drug-delivery systems involving inverted hexagonal (H2) or bicontinuous cubic (V2) phases in excess water under physiological conditions at pH 7.4 and 37 °C.4,7−9,22−24 In binary lipid−water systems, previous Received: September 13, 2011 Revised: January 16, 2012 Published: January 16, 2012 2881
dx.doi.org/10.1021/la203577v | Langmuir 2012, 28, 2881−2889
Langmuir
Article
at pH 7.4 (PBS 7.4) and 0.14 M at pH 6.0 (PBS 6.0). The ionic strength was kept constant for both buffers at a value of 0.18 M. Other reagents and solvents were of analytical or reagent grade. Deionized water was used throughout the study. Visking dialysis tubing size 27/32″, 21.5 mm, MW cutoff 12−14 kDa (VWR International, West Chester, Pennsylvania) was employed for the rotating dialysis cell. Preparation of BUP-Loaded GMO/MCT-Based SelfAssembled Systems. For SAXS investigations, BUP with appropriate concentrations was added to GMO or binary GMO/MCT mixtures and heated for 5−10 min at 40−50 °C. The obtained isotropic solutions were hydrated by adding the preheated PBS, heating the samples for 5−10 min at 50−60 °C, and then homogenizing by vigorous vortexing. In the binary GMO/MCT mixtures, the MCT concentration was in the range of 0−40 wt %. We investigated the effect of the weight ratio R (R = mass of solubilized BUP/mass of GMO (or GMO/MCT mixture)) on the nanostructures of the GMOand GMO/MCT-based phases. The fully hydrated samples were formed at a fixed total phosphate buffer concentration of 50 wt %. Both the low-buffer containing samples and the corresponding fully hydrated samples were incubated at 37 °C for 1−2 weeks before carrying out SAXS measurements. X-ray Measurements and Data Analysis. SAXS measurements52 were performed at the Austrian SAXS beamline at ELETTRA, Trieste, Italy, at an energy of 8 keV. The scattering patterns were recorded with a 2D image plate detector (Mar300, MarResearch, Norderstedt, Germany) using collection times of 30−120 s. The detector covered the q-range (q = 4π sin θ/λ, where λ is the wavelength and 2θ is the scattering angle) of interest from about 1/24 to 1/1.3 nm−1 at an X-ray energy of 8 keV. Silver behenate (CH3−(CH2)20− COOAg with a d spacing value of 5.84 nm provided by Eastman Kodak, NY, USA) was used as a standard to calibrate the angular scale of the measured intensity. The samples were measured in quartz glass capillaries (diameter 1.5 mm) and thermostatted in a custom-built sample holder block of brass, which was connected to a circulating water bath (Unistat CC, Huber, Offenburg, Germany). The temperature was set to 37 °C. Before analysis, the 2D-images were radially integrated using the software FIT2D.53 All lattice spacings of the L2, V2 of the symmetry Pn3m, and H2 phases were deduced from the reflections with strong intensity applying standard procedures (see for instance review54). Lipidic Partition Experiments. The lipidic partition coefficients of BUP between the GMO/MCT-based nanostructures formed and the phosphate buffer solutions (PBS 6.0 or PBS 7.4) were determined in triplicate at 37 ± 0.5 °C as previously described by Engström et al. 1999.55 Briefly, 0.50 mL (Vl) of the GMO/MCT mixtures was added to 5.0 or 15.0 mL (Vw) of a BUP solution containing approximately 1 mM BUP dissolved in PBS 6.0 or PBS 7.4, respectively. Equilibrium was attained after rotation in an incubator hood for 5 days and the aqueous phase was removed after 1 day of phase separation. We assume that all GMO and MCT molecules are taking part in the formation of the lipid nanostructures. The BUP concentration in the buffer solution before (Ci) and after equilibrium (Cw) was measured by HPLC. The lipidic partition coefficient55 was calculated according to eq 1 by assuming that the concentration of BUP in the aqueous channels of the nanostructure is equal to Cw.
reports have documented the self-assembly of these phases in excess water at a wide range of investigated temperatures.14,25−34 Used lipids include monoglycerides, phospholipids, urea-based lipids, endocannabinoids, synthetic lipid prodrugs, as well as glycolipids.14,18,25−31 The formation of the H2 and V2 phases can easily be modulated by varying the lipid composition, the inclusion of guest additives including hydrophobic agents, peptides, or salts.12,13,35−39 The hydrophilic nanochannels of these phases undergo significant change in their diameter in response to external stimuli such as changes in temperature or due to partial replacement of the investigated lipid or lipid mixtures by hydrophilic or hydrophobic agents.3,12,25,37,40 The formation of safe injectable depots of local anesthetics that prolong their analgesic effect and improve therefore the pain management is an important research area. In particular, there is a great interest in the solubilization of local anesthetics in efficient lipid-based carriers that provide prolonged analgesic effect with minimal systemic and local toxicity.2,41 It was reported that the delivery of large doses of local anesthetic agents without encapsulation in lipidic or polymeric carriers would be lethal.41 Different lipidic systems including oily vehicles2,42−44 and liposomes41,45 were suggested as sustainedrelease depot formulations of local anesthetics such as bupivacaine, lidocaine, and ropivacaine. For instance, Dyhre et al.44 described recently new MCT-based depot formulations of BUP that have promising pharmaceutical and pharmacologic characteristics. In the present study, we have investigated the effect of varying the lipid composition and the solubilized model drug (BUP) content in lipid-based systems, comprising the commercial distilled glycerol monooleate Myverol 18−99K (GMO) and medium chain triglycerides (MCT), on the formed self-assembled nanostructures and the in vitro drugrelease properties. The main attention was to evaluate in vitro the potential utility of inverted-type liquid crystalline phases and microemulsions as injectable depot types intended for subcutaneous administration. Three different self-assembled systems were formed by increasing the MCT content, which induced the structural transition of V2 → H2 → L2. Synchrotron SAXS was used to study the structural characteristics of these systems at physiological temperature in relation to the GMO/MCT weight ratio as well as the solubilized BUP content and the pH of the buffer. In addition, the release of BUP from these in situ formed self-assembled nanostructures was investigated in a membrane-based in vitro release method (the rotating dialysis cell model2,42,46−50), which allows the drug release to be followed immediately upon the exposure of the BUP preformulations to the aqueous buffer media.
2. MATERIALS AND METHODS Materials. Myverol 18−99K was a gift from Kerry BioSciences, Almere, Netherlands. This distillated food-grade emulsifier consists of 93% monoglycerides including 60.0% monoolein, and about 21% monolinolein. Medium chain triglycerides (MCT) consisting of approximately 60% caprylic acid and 40% capric acid was kindly donated from Brøste AS (Lyngby, Denmark). The free base form of the local anesthetic bupivacaine (BUP) was obtained from its corresponding hydrochloride (Unikem, Copenhagen, Denmark) as previously reported.43 BUP is a weak base51 and the corresponding acid has a pKa value of about 8.1 at 37 °C. Phosphate buffer solutions (PBS) were used at two different pH values: 0.067 M 2882
dx.doi.org/10.1021/la203577v | Langmuir 2012, 28, 2881−2889
Langmuir
Kl / w =
Article
Ci − C w V × w Cw Vl
(1)
where Vl and Vw are the volume of the GMO/MCT mixture and the aqueous buffer solution, respectively. In Vitro Release Experiments. The release experiments were carried out at 37 ± 0.5 °C using the rotating dialysis cell model as previously described.42 As illustrated in Figure 1, this release model consists of a small cylindrical donor compartment separated from a large acceptor compartment by a dialysis membrane. At time zero, the dialysis cell containing 1.0 mL of the lipid preformulation of BUP added to 5.0 mL buffer (PBS 6.0 or PBS 7.4) was placed inside a round-bottomed vessel containing 1000 mL of preheated release medium. After exposure of the injectable low-viscous stimulus-responsive preformulations to the buffer, the nanostructure (BUP-loaded lyotropic nonlamellar phase or microemulsion) was rapidly formed in the dialysis cell. Figure 1 schematically illustrates the applied rotating dialysis cell model and three representative examples on the formed self-assembled nanostructures. In this experimental setup, the net drug transport process include the following: i) the release of BUP from the in situ formed nanostructures into the phosphate buffer in the donor compartment, and then ii) the diffusion of BUP across the dialysis membrane from the donor to the acceptor compartment (Figure 1). To determine the permeability of BUP across the dialysis membrane, an additional experiment was performed where the dialysis cell contained 5.0 mL buffer solution (PBS 6.0) of BUP. The revolution speed of the dialysis cell was set to 50 rpm. At appropriate time intervals, samples were withdrawn from the acceptor phase and analyzed by HPLC. The release experiments (in triplicate) were run until equilibrium was attained in the system or until more than 60% of the drug (BUP) was released. The cumulated amount of BUP released, MA,t, was calculated according to the following equation: n
Figure 1. Schematic illustration of the applied rotating dialysis cell model for the in vitro drug-release experiments. (a) The dialysis cell (the donor compartment) with a membrane area of 22 cm2 in which 1.0 mL of the preheated low-viscous stimulus-responsive preformulation of BUP was added to 5.0 mL buffer with pH 7.4 (PBS 7.4) or with pH 6.0 (PBS 6.0). (b) The dialysis cell was placed inside a roundbottomed vessel containing 1000 mL of preheated release medium (the receptor compartment). The nanostructure (BUP-loaded lyotropic nonlamellar phase or microemulsion) was rapidly formed in the dialysis cell. Three photos of the dialysis cell containing the in situ formed nanostructure (without BUP) and the surrounding aqueous phase are shown. From bottom to top, they display a V2 phase of the symmetry Pn3m formed in the absence of MCT, a formed H2 phase with low MCT load, and a formed L2 phase in the presence of high amount of solubilized MCT. The photos clearly indicate a reduced interfacial area between the lipidic nanostructure of MCT-free (biphasic Pn3m/H2 system) and low MCT-loaded (H2 system with 5 wt % MCT) as compared to that for the microemulsion system (40 wt % MCT in the binary GMO/MCT mixture).
MA,t = VS ∑ Ci − 1 + VAC n i=1
(2)
where VS and VA are the volumes of the samples withdrawn from the acceptor phase (VS = 1.5 mL) and the volume of the acceptor phase at time t, respectively. Ci is the drug concentration in sample i. For the release experiments, an overall first-order release rate constant, k, for attainment of equilibrium was obtained by fitting MA,t vs time (t) using GraphPad Prism 5.02 for Windows (GraphPad Software, San Diego, CA, USA) applying the following equation:
MA,t = M∞(1 − e−kt )
(3)
where M∞ is the fitted amount of BUP released at equilibrium. HPLC Analysis. Samples from the release and partition experiments were analyzed by an HPLC system consisting of a LaChromElite system (VWR International, Tokyo, Japan) consisting of a L-2130 pump, a L-2300 column oven set at 30 °C, a L-2450 diode array detector and a L-2200 autosampler. Reversed phase chromatography was performed using a C18 Gemini RP column (150 × 4.6 mm, 5 μm particles) (Phenomenex, Torrance, CA, USA) equipped with a SecurityGuard precolumn (Phenomenex, Torrance, CA, USA). The flow rate was set at 1 mL/min and the column effluent was monitored at 205 nm. The mobile phase consisted
of 35% (v/v) methanol and 65% (v/v) of 0.1% (v/v) phosphoric acid. Quantification of BUP was done from peak area measurements in relation to those of standards chromatographed under the same conditions.
3. RESULTS AND DISCUSSION 3.1. Effect of Solubilized Drug Concentration on the Self-Assembled Nanostructures. In this section, SAXS 2883
dx.doi.org/10.1021/la203577v | Langmuir 2012, 28, 2881−2889
Langmuir
Article
solubilization of drugs such as ibuprofen, tetracaine, propranolol, and lidocaine induces the Pn3m → H2 structural transition in monoglyceride-based systems.56,57 The influence of solubilized hydrophobic agents on the spontaneous curvature of the monolayer leaflets in these lipid nanostructures is discussed in great detail in previous reports.12,13,35 In brief, the present results confirm that the structural transition from bicontinuous cubic phase to discontinuous phases, that is, concerning the H2, the L2, or the discontinuous micellar cubic Fd3m phases involves significant change in the amount of solubilized water and requires overcoming the packing limitations, which are the main energy barrier for the formation of these discontinuous phases.12,61,62 This means that the efficiency to enhance the Pn3m → H2 transition in the presence of poorly water-soluble drugs is mainly affected by the degree of incorporation of solubilized drug molecules into the membrane interfacial film and the concomitant variation of the spontaneous curvature of the monolayer leaflets. This also explains the full transition from the Pn3m phase to H2 phase at the same R value of 0.06, when the BUP-loaded sample is prepared with a higher pH value (using PBS 7.4). The increase in the pH value leads to a reduction of the fraction of ionized BUP molecules and hence to a significant decrease in the degree of partitioning of BUP into the hydrophilic channels of the V2 and H2 phases. In other words, increasing the pH enhances the up-take of the unionized BUP in the lipid bilayer, which triggers the formation of the highly curved H2 phase. In the presence of BUP (R = 0.06), a representative example of the effect of hydration of GMO with PBS 6.0 under equilibrium conditions is given in panel b of Figure 2. The investigated samples were prepared with different water concentrations in the range of 5−50 wt %. It is clear that the hydration enlarges the nanostructures owing to the waterswelling behavior and it induces transition in the monophasic region from the inverted micellar-type solution via the H2 phase to the fully hydrated Pn3m/H2 biphasic system. The detailed characterization of this sample during hydration under nonequilibrium conditions is discussed in our recent report.11 We demonstrated that a fast hydration-induced rearrangement of the lipid and the solubilized BUP molecules takes place in excess phosphate buffer; whereas the overall structural conversion of the preformulations to the corresponding equilibrated full hydrated systems approaches within about 1000 s. Panel a of Figure 3 presents an example on the effect of adding MCT to the GMO nanostructures with and without BUP. In the BUP-free sample, the addition of a small amount of MCT (binary GMO/MCT mixture consisting of 5 wt % MCT) induces a structural transition from the Pn3m cubic phase with a lattice parameter, a, of 8.40 nm to H2 phase with an a value of 5.85 nm (Table 1). In the present study, the recorded lattice parameter for the MCT-free Pn3m cubic phase under full hydration conditions is similar to the value of 8.37 nm that was previously reported for the binary Myverol 18−99K/water system63 at 40 °C. At a higher MCT content (binary GMO/ MCT mixture consisting of 20 or 40 wt % MCT), the SAXS patterns have a single broad peak indicating the formation of L2 phase (panel b of Figure 3). MCT especially helps to overcome packing limitations in the discontinuous mesophases.12,35 As a consequence, the addition of MCT causes the release of curvature frustration and is therefore also associated with a significant decrease in the amount of solubilized water in these inverted-type liquid crystalline nanostructures. This effect
experiments carried out on fully hydrated BUP-free and BUPloaded GMO- and GMO/MCT-based systems at 37 °C are presented. For the GMO-based systems, prepared with PBS 6.0, panel a of Figure 2 shows typical structural changes as the
Figure 2. BUP concentration-dependence behavior of GMO/water system at 37 °C. The samples were prepared with PBS 6.0. (a) SAXS scattering patterns were taken from fully hydrated samples prepared with different R values in the range of 0−0.06. (b) The effect of hydration on the BUP-loaded GMO nanostructure at R value of 0.06. The SAXS scattering patterns reveal the structural transitions from the L2 via H2 to a biphasic system of Pn3m coexisting with H2. The SAXS diffraction pattern of the drug-loaded fully hydrated sample (R = 0.06) is also presented in (b) to compare the three representative examples of the structures of the monophasic samples containing PBS 6.0 in the range of 5−16 wt % with the final structural form (the structure of the fully hydrated sample is independent of buffer content). The scattering curves are shifted for better visibility.
weight ratio of solubilized drug, R, increases in the range of 0− 0.06. In the absence of BUP (R value of 0), GMO selfassembles in excess water to form a bicontinuous cubic (V2) phase of the symmetry Pn3m (diamond type, QD). In this case, the lattice parameter could be identified by recording the first 8 reflections. As further seen, up to an R = 0.04 the Pn3m structure is retained, but the set of observed peaks shift to higher q values indicating a decrease in the lattice parameter (Table 1). At an R = 0.06, the system displays a coexistence of the Pn3m and a newly formed H 2 liquid crystalline nanostructure. This is consistent with previously reported results on the effect of solubilizing hydrophobic agents on fully hydrated lyotropic systems.12,35,56−60 It was reported that the 2884
dx.doi.org/10.1021/la203577v | Langmuir 2012, 28, 2881−2889
Langmuir
Article
Table 1. Structure Parameters (the Mean Lattice Parameter, a, for the V2 and the H2 Phases and the d Spacing That Is Called Characteristic Distance for the L2-Phase) at 37 °C as Derived from SAXS Investigations That Were Carried out on the GMOand GMO/MCT-Based Systems As Function of R Value investigated system no.
GMO content in the GMO/MCT mixture
R value
buffer
buffer content (wt %)
space group
mean a (nm)
1 2a 2b 2c 2d 3a 3b 3c 4a 4b 4c 4d 5a 5b 5c 5b 6a 6b 6c 6d
100 100 100 100 100 100 100 100 95/5 95/5 95/5 90/10 80/20 80/20 80/20 80/20 60/40 60/40 60/40 60/40
0 0.02 0.04 0.06 0.06 0.06 0.06 0.06 0 0.06 0.06 0.06 0 0.06 0.06 0.06 0 0.02 0.04 0.06
PBS 6.0a PBS 6.0 PBS 6.0 PBS 6.0 PBS 7.4 PBS 6.0 PBS 6.0 PBS 6.0 PBS 6.0a PBS 6.0 PBS 7.4 PBS 6.0 PBS 6.0a PBS 6.0 PBS 6.0 PBS 6.0 PBS 6.0a PBS 6.0 PBS 6.0 PBS 6.0
50b 50b 50b 50b 50b 5c 10c 16c 50b 50b 50b 50b 50b 50b 5c 10c 50b 50b 50b 50b
Pn3m Pn3m Pn3m Pn3m & H2 H2 L2 L2 H2 H2 H2 H2 H2 L2 L2 L2 L2 L2 L2 L2 L2
8.40 8.27 8.01 7.92 (Pn3m) & 5.75 (H2) 5.61
d (nm)
3.38 3.74 5.11 5.85 5.52 5.31 5.11 4.63 4.07 3.46 4.00 4.16 4.03 3.91 3.69
a
Identical nanostructures were detected for BUP-free samples prepared with PBS 7.4, bfully hydrated samples containing 50 wt % PBS 6.0 or PBS 7.4, and cthe effect of PBS 6.0 content was investigated on isotropic monophasic BUP-loaded GMO or GMO/MCT systems containing.
membrane in the applied rotating dialysis release cell is pH independent over pH 6.2−8.2. After exposing the BUP-loaded GMO/MCT preformulations to either PBS 6.0 or PBS 7.4, the in situ formation of the selfassembled nanostructures is supposed to take place very fast, within about 15 min as recently reported.11 The release experiments were adequately described by apparent first-order kinetics with correlation coefficients (R2 > 0.98) and the obtained rate constants are presented in Table 2. It should be noted that the release rates of solubilized drugs from invertedtype liquid crystalline systems have been analyzed by different mathematical models including Higuchi’s square root of time function,3,4 Ritger and Peppas empirical equation,64 and firstorder kinetics.65,66 By comparison of the BUP release rate from the in situ formed V2 and H2 nanostructures (in absence of MCT), it is evident that the change in pH (replacement of PBS 6.0 by PBS 7.4) results in a significant change in the obtained release rates: the release rate was significantly faster (t1/2 = 7.5 h) at pH 6.0 compared to that observed at pH 7.4 (t1/2 = 25 h) (panel a of Figure 4). The significant impact of pH on the release rates is most likely related to the higher affinity of BUP to the lipidic phase at pH 7.4 as also reflected in the determined lipidic partition coefficient, Kl/w, between the GMO-based liquid crystalline nanostructures and the phosphate buffer solution. As shown in Table 2, the determined Kl/w values were 212 and 21 at pH 7.4 and 6.0, respectively. This pH sensitivity is consistent with our SAXS data: the increased incorporation of BUP into the lipidic matrix induces a full structural transition from Pn3m cubic phase to H2 phase at pH 7.4; whereas a complete transition is prevented at the investigated drug loads (R values) at pH 6.0. In the latter case 99.2% of BUP is ionized; whereas the fraction of uncharged form of BUP increases to 16% at pH 7.4 leading to a relatively higher amount of BUP embedded in the lipid membrane (Table 2). This means that
increases the monolayer curvature modulus, which leads to the Pn3m → H2 → L2 transition with increasing the MCT content. It is worth noting that the BUP-free samples prepared with PBS 7.4 and PBS 6.0 display identical nanostructures. Both the H2 and the L2 phases exhibit the pH-insensitive lattice spacings in all preparations, that is the corresponding a values remain the same at pH 6.0 and 7.4. There is also no pH effect on the structure of the electrically neutral MCT-free Pn3m cubic phases. Figure 3 shows the significant impact of solubilizing BUP molecules on the three self-assembled systems (the cubic Pn3m, H2 and L2 phases): the lattice shrinking behavior of the nanostructures is indicated by the shift of the observed peaks in the diffraction patterns to higher q values. The structure parameters for these SAXS data are summarized in Table 1. 3.2. In Vitro Drug Release Studies. As mentioned above, the release experiments were designed to mimic the drugrelease profiles from in situ formed drug-loaded lyotropic nonlamellar phases or microemulsions after subcutaneous administration of injectable low viscous stimulus-responsive precursors. This was achieved by investigating BUP release after rapid exposure of the preformulations (stock solution of BUP dissolved in GMO or GMO/MCT mixtures) to excess of phosphate buffer in the donor compartment of the rotating dialysis cell model at 37 °C (Material and Methods and Figure 1). The rate of the diffusion process of BUP across the dialysis membrane from the donor to the acceptor compartment was determined by an additional experiment in which an aqueous buffer solution (PBS 6.0) of BUP was added to the donor compartment under the same experimental conditions. In this experiment, the appearance of BUP in the acceptor compartment adhered to first-order kinetics under sink conditions (panel a of Figure 4, green circles; k = 1.01 h−1, standard deviation (S.D.) = 0.04, n = 3) in line with previous findings.46 It has been found46 that the permeability of a solute across the 2885
dx.doi.org/10.1021/la203577v | Langmuir 2012, 28, 2881−2889
Langmuir
Article
Figure 3. Structural characterization of BUP-free (R = 0) and BUP-loaded (R = 0.06) fully hydrated systems in the presence of MCT at 37 °C. (a) pH-dependence of BUP-free and BUP-loaded fully hydrated H2 phase consisting of GMO/MCT at a weight ratio of 95/5. These investigated samples containing 50 wt % PBS 7.4 or PBS 6.0 were prepared under equilibrium conditions. (b) The effect of varying lipid composition on the BUP-free and BUP-loaded fully hydrated nanostructures of samples containing 50 wt % PBS 6.0. The MCT content in the GMO/MCT binary mixture was varied in the range of 0−40 wt %. The scattering curves are shifted for better visibility. Panels (c)−(e) show representative examples of SAXS 2D detector patterns for the following samples prepared with PBS 7.4: (c) MCT-loaded L2 phase contains GMO/MCT mixture at weight ratios of 95/5 (c), (d) MCT-loaded H2 phase contains GMO/MCT mixture at weight ratios of 80/20, (e) MCT-free Pn3m cubic phase.
aqueous buffer phase in the donor compartment since almost equal lipidic partition coefficients of BUP (Kl/w in the range of 18−21) were obtained with the GMO/MCT binary mixtures containing 0−40 wt % MCT (Table 2). Only for 20 and 40 wt % MCT in the GMO/MCT binary mixtures the slightly slower release of BUP from the microemulsion systems as compared to MCT alone can be ascribed, to some extent, to the small increase in lipidic partition coefficient (Table 2). As described in previous studies,42,43,68 the in vitro drug-release rates from oils such as MCT are dictated by the drug partition coefficient between the oil and the aqueous phase in the donor compartment of the rotating dialysis cell because the intensive stirring of the system leads to a large oil−water interface. In fact, partitioning phenomena are generally applied to describe release from various drug-delivery systems when membranebased in vitro release methods are used.2,42,46−48 3.3. Macro- and Nanostructure Couple to Release Rates. The observed decrease in the BUP release rate from the high-viscous inverted-type liquid crystalline phases compared to that obtained from the low-viscous inverted-type microemulsion systems can also be due to the significant reduction in the interfacial area between the lipidic phase and the
the increased level of BUP ionization with decreasing pH from 7.4 to 6.0 enhances the preferential localization of the drug molecules in the hydrophilic nanochannels of the selfassembled systems and in the surrounding excess buffer in the donor compartment, which subsequently leads to a faster drug release to the acceptor compartment. Likewise, a previous release study67 using a dialysis cell with a home−built cubic phase sample holder demonstrated that the release rate of tryptophan (a water-soluble model drug) from GMO-based cubic phase was decreased by alkylation because this chemical modification of tryptophan enhanced the affinity to the lipidic matrix (increased the lipidic partition coefficient). Panel b of Figure 4 illustrates the dependence of the BUP release rate at pH 6.0 on the lipid composition for five different systems. It is evident that the partial replacement of GMO by MCT is significantly affecting both the obtained nanostructures as shown in Figure 3 and the release rates as seen in Table 2. The obtained results indicate a faster release rate with increasing MCT content in the binary GMO/MCT system as MCT content varies in the range of 0−40 wt %. These observations cannot be explained by a partitioning process between the self-assembled nanostructure and the excess of 2886
dx.doi.org/10.1021/la203577v | Langmuir 2012, 28, 2881−2889
Langmuir
Article
enlarged. This means that the decrease in the overall release rate observed for the less elastic inverted-type liquid crystalline phases might be due to a slower establishment of the lipidbuffer partition equilibrium inside the donor compartment. To this end, a decrease in the drug-release rate from the in situ formed high-viscous cubic phases have been observed in a recent in vitro release study using dialysis bags.70 In line with the above, an interesting finding of the present study is that the formation of the H2 phase in the presence of MCT (GMO/ MCT mixture at a weight ratio of 95/5) involving a significant decrease in the hydrophilic channels (Table 1) does not lead to a slower release rate than that obtained from the GMO-based (MCT-free system) sample (part b of Figure 4). In contrast, recent studies3,71 on the release from inverted-type liquid crystalline phases found that the diffusion of solubilized model hydrophilic drugs from the H2 phase is slower than that from the Pn3m cubic phase potentially due to the significant decrease in the size of hydrophilic channels of the H2 phase. In theses investigations,3,71 the release of model hydrophilic drugs was followed from a constant and a well-defined surface area of the liquid crystalline phases prepared before initiating the experiments. However, to mimic the subcutaneous administration in a realistic manner our BUP release experiments were performed without fixing the self-assembled nanostructure to buffer surface area in the applied rotating dialysis method. Therefore, the inconsistency with the results of the present study should be seen in the light of the difference in the physicochemical properties of the used model drug and the employed in vitro release model. Furthermore, in contrast to the BUP release rate dependency on the lipid composition at pH 6.0, increasing the content of MCT in the preformulation mixtures did not lead to a significant difference in BUP release rate at pH 7.4 (Table 2). It is clear that further release studies are needed to fully investigate the effects of the structural changes and the variations in the surface area of the in situ formed selfassembled systems coexisting with excess buffer in the donor compartment on the drug-release rates. Figure 5 shows a schematic illustration of the formed nanostructures and emphasizes the influence of the investigated
Figure 4. BUP release profiles from different GMO- and GMO/MCTbased samples at 37 °C. (a) Release profiles of BUP from GMO-based formulations (R = 0.06) at pH 6.0 (red triangles) and 7.4 (blue diamonds), respectively, compared to an aqueous buffer solution (pH 6.0) (green circles). (b) Release profiles of BUP at pH 6.0 from the following GMO/MCT-based formulations: blue circles: 100 wt % MCT (R = 0.02), green diamonds: 40 wt % MCT (R = 0.06), magenta diamonds: 20 wt % MCT (R = 0.06), black triangles: 5 wt % MCT (R = 0.06), and red triangles: 0 wt % MCT (R = 0.06)). The full lines were obtained by fitting the data to first-order kinetics according to eq 3. Bars represent standard deviations for experiments in triplicate (n = 3).
Table 2. Lipidic Partition Coefficients for BUP at 37 °C between the GMO/MCT Mixtures and Phosphate Buffer Solutions of pH 7.4 (Kl/pH 7.4) or pH 6.0 (Kl/pH 6.0) and the Corresponding First-Order Release Rate Constant (kpH 7.4 and kpH 6.0)a wt% MCT in GMO/MCT mixture
Kl/pH 7.4 (±S.D.)
0 5 20 40 100
212 (3.6) 214 (4.0) n.d. n.d. 210 (7.8)
a
Kl/pH 6.0 ± S.D.) 21 19 21 18 13
(0.3) (0.7) (0.5) (0.1) (0.3)
kpH 7.4 (h−1) (±S.D.)
kpH 6.0 (h−1) (±S.D.)
0.028 (0.014) 0.022 (0.001) n.d. n.d. 0.027 (0.001)
0.093 (0.011) 0.13 (0.01) 0.17 (0.01) 0.19 (0.01) 0.27 (0.003)
Figure 5. Schematic illustration of the effect of lipid composition on the drug-release rates of different BUP-loaded lipidic formulations based on the binary GMO/MCT mixture.
n.d.: not determined, S.D.: standard deviation, n = 3−9.
lipid composition on the release profiles. It is worth mentioning that the in vitro drug release from different inverted-type liquid crystalline phases has been investigated in several studies;3,9,23,67 however, a full understanding of the release mechanism has not been provided. The release behavior is, besides being related to the drug properties (e.g., the degree of its lipophilicity and amphiphilicity) and the characteristics of the nanostructures of the formulations, also affected by the design and the experimental conditions of the employed in vitro drug-release model. Therefore in this research area, future
surrounding aqueous buffer in the donor compartment. Indeed the formation of high-viscous V2 and H2 phases69 (Figures 2 and 3) are accompanied with a significant reduction in the interfacial area between the lipidic BUP-loaded self-assembled system and the surrounding aqueous buffer in the donor compartment (panel b of Figure 1). In contrast, the formed low-viscous inverted-type microemulsion systems (at MCT content of 20 or 40 wt % in the GMO/MCT mixtures) spread out in the donor compartment and therefore the interfacial area between the lipid formulation and the buffer medium is 2887
dx.doi.org/10.1021/la203577v | Langmuir 2012, 28, 2881−2889
Langmuir
■
studies should focus on the comparison of the release profiles of hydrophilic, hydrophobic, and amphiphilic drugs from these phases by applying different in vitro release models, each chosen with relevance to the administration route of these formulations.
REFERENCES
(1) Pritchard, E. M.; Kaplan, D. L. Expert Opin. Drug Del. 2011, 8, 797−811. (2) Larsen, C.; Larsen, S. W.; Jensen, H.; Yaghmur, A.; Østergaard, J. Expert Opin. Drug Del. 2009, 6, 1283−1295. (3) Fong, W. K.; Hanley, T.; Boyd, B. J. J. Controlled Release 2009, 135, 218−226. (4) Rizwan, S. B.; Boyd, B. J.; Rades, T.; Hook, S. Expert Opin. Drug Del. 2010, 7, 1133−1144. (5) Dicko, A.; Mayer, L. D.; Tardi, P. G. Expert Opin. Drug Del. 2010, 7, 1329−1341. (6) Boyd, B. J.; Khoo, S. M.; Whittaker, D. V.; Davey, G.; Porter, C. J. Int. J. Pharm. 2007, 340, 52−60. (7) Boyd, B. J.; Whittaker, D. V.; Khoo, S. M.; Davey, G. Int. J. Pharm. 2006, 309, 218−226. (8) Chang, C. M.; Bodmeier, R. Int. J. Pharm. 1998, 173, 51−60. (9) Shah, J. C.; Sadhale, Y.; Chilukuri, D. M. Adv. Drug Delivery Rev. 2001, 47, 229−250. (10) Malmsten, M. J. Disper. Sci. Technol. 2007, 28, 63−72. (11) Yaghmur, A.; Larsen, S. W.; Schmitt, M.; Østergaard, J.; Larsen, C.; Jensen, H.; Urtti, A.; Rappolt, M. Soft Matter 2011, 7, 8291−8295. (12) Yaghmur, A.; Kriechbaum, M.; Amenitsch, H.; Steinhart, M.; Laggner, P.; Rappolt, M. Langmuir 2010, 26, 1177−1185. (13) Yaghmur, A.; Laggner, P.; Zhang, S.; Rappolt, M. PLoS ONE 2007, 2, e479. (14) Larsson, K. J. Phys. Chem. 1989, 93, 7304−7314. (15) Angelova, A.; Angelov, B.; Mutafchieva, R.; Lesieur, S.; Couvreur, P. Acc. Chem. Res. 2011, 44, 147−156. (16) Luzzati, V. Curr. Opin. Struct. Biol. 1997, 7, 661−668. (17) Almsherqi, Z.; Hyde, S.; Ramachandran, M.; Deng, Y. J. R. Soc. Interface 2008, 5, 1023−1029. (18) Yaghmur, A.; Glatter, O. Adv. Colloid Interface Sci. 2009, 147− 148, 333−342. (19) Hyde, S.; Blum, Z.; Landh, T.; Lidin, S.;, Ninham, B. W.; Andersson, S.; Larsson, K. The Language of Shape: The Role of Curvature in Condensed Matter: Physics, Chemistry, Biology; Elsevier: Amsterdam, 1997. (20) Michielsen, K.; Stavenga, D. G. J. R. Soc. Interface 2008, 5, 85− 94. (21) Almsherqi, Z. A.; Kohlwein, S. D.; Deng, Y. J. Cell Biol. 2006, 173, 839−844. (22) Peng, X. S.; Wen, X. G.; Pan, X.; Wang, R. C.; Chen, B.; Wu, C. B. AAPS Pharmscitech 2010, 11, 1405−1410. (23) Lee, K. W. Y.; Nguyen, T. H.; Hanley, T.; Boyd, B. J. Int. J. Pharm. 2009, 365, 190−199. (24) Han, K.; Pan, X.; Chen, M. W.; Wang, R. C.; Xu, Y. H.; Feng, M.; Li, G.; Huang, M.; Wu, C. B. Eur. J. Pharm. Sci. 2010, 41, 692− 699. (25) de Campo, L.; Yaghmur, A.; Sagalowicz, L.; Leser, M. E.; Watzke, H.; Glatter, O. Langmuir 2004, 20, 5254−5261. (26) Qiu, H.; Caffrey, M. Biomaterials 2000, 21, 223−234. (27) Kaasgaard, T.; Drummond, C. J. Phys. Chem. Chem. Phys. 2006, 8, 4957−4975. (28) Gong, X.; Sagnella, S.; Drummond, C. J. Int. J. Nanotechnol. 2008, 5, 370−392. (29) Sagnella, S. M.; Conn, C. E.; Krodkiewska, I.; Mulet, X.; Drummond, C. J. Soft Matter 2011, 7, 5319−5328. (30) Sagnella, S. S.; Gong, X. J.; Moghaddam, M. J.; Conn, C. E.; Kimpton, K.; Waddington, L. J.; Krodkiewska, I.; Drummond, C. J. Nanoscale 2011, 3, 919−924. (31) Gong, X. J.; Moghaddam, M. J.; Sagnella, S. M.; Conn, C. E.; Danon, S. J.; Waddington, L. J.; Drummond, C. J. ACS Appl. Mater. Interfaces 2011, 3, 1552−1561. (32) Kulkarni, C. V.; Wachter, W.; Iglesias-Salto, G.; Engelskirchen, S.; Ahualli, S. Phys. Chem. Chem. Phys. 2011, 13, 3004−3021. (33) Rappolt, M.; Hodzic, A.; Sartori, B.; Ollivon, M.; Laggner, P. Chem. Phys. Lipids 2008, 154, 46−55. (34) Rappolt, M.; Di Gregorio, G. M.; Almgren, M.; Amenitsch, H.; Pabst, G.; Laggner, P.; Mariani, P. Europhys. Let. 2006, 75, 267−273.
4. CONCLUSIONS The structural characteristics and the in vitro BUP release properties of a series of lipidic formulations based on GMO or binary GMO/MCT mixtures were investigated by synchrotron SAXS and rotating dialysis cell model. At 37 °C, the SAXS investigations of these formulations reveal the formation of different inverted-type nanostructures in excess buffer. These results show clearly how the structural parameters depend on pH, lipid composition, and solubilized BUP content. Among these factors, the obtained SAXS data emphasize the important role of pH and partial replacement of GMO by MCT on modulating both the self-assembled nanostructure and the in vitro drug-release profiles. The drug release from the inverted-type liquid crystalline phases and microemulsions was found to follow first-order kinetics similar to that previously reported on the release from oils when employing the rotating dialysis method.2,42,43 The studies reveal that both the self-assembled nanostructures and the BUP release rates are highly sensitive to pH changes. This pH sensitivity is most likely due to the change in the affinity of the solubilized BUP to the lipid-based nanostructures as reflected in the determined lipidic partition coefficient, Kl/w, between the GMO-based liquid crystalline nanostructures and the phosphate buffer solution. For instance, the obtained results indicate approximately 10-fold decrease in the Kl/w value of the MCT-free GMO-based systems upon decreasing pH from 7.4 to 6.0. In investigating the in vitro drug-release profiles, keeping the surface area between the investigated inverted-type liquid crystalline phase and the release medium constant is common practice.3,4,7,9 However, in order to mimic the subcutaneous administration in a realistic manner, the release experiments in this study were performed without defining a fixed surface area between the self-assembled system and the release medium. This explains the main difference between our results and most recent reports.3,23 Future studies are needed to fully understand the drug-release mechanism at the nanoscale from in situ formed self-assembled nanostructures by investigating various factors including the drug physicochemical properties (ionization and lipophilicity), the type of the self-assembled nanostructure and its release surface area, and the presence of external stimuli. This calls also for relevant in vivo investigations to verify these effects. These lipid-based formulations are attractive candidates for the solubilization of local anesthetic agents such as BUP and for extending the duration of their local anesthetic effect.
■
Article
AUTHOR INFORMATION
Corresponding Author
*Tel.: +45 35 33 65 41, Fax: +45 35336030, E-mail: aya@ farma.ku.dk. Notes
The authors declare no competing financial interest. 2888
dx.doi.org/10.1021/la203577v | Langmuir 2012, 28, 2881−2889
Langmuir
Article
(35) Yaghmur, A.; de Campo, L.; Sagalowicz, L.; Leser, M. E.; Glatter, O. Langmuir 2005, 21, 569−577. (36) Yaghmur, A.; Sartori, B.; Rappolt, M. Phys. Chem. Chem. Phys. 2011, 13, 3115−3125. (37) Yaghmur, A.; de Campo, L.; Sagalowicz, L.; Leser, M. E.; Glatter, O. Langmuir 2006, 22, 9919−9927. (38) Salentinig, S.; Sagalowicz, L.; Glatter, O. Langmuir 2010, 26, 11670−11679. (39) Fraser, S.; Separovic, F.; Polyzos, A. Eur. Biophys. J. Biophy. 2009, 39, 83−90. (40) Angelov, B.; Angelova, A.; Vainio, U.; Garamus, V. M.; Lesieur, S.; Willumeit, R.; Couvreur, P. Langmuir 2009, 25, 3734−3742. (41) Weiniger, C. F.; Golovanevski, M.; Sokolsky-Papkov, M.; Domb, A. J. Expert Opin. Drug Del. 2010, 7, 737−752. (42) Larsen, S. W.; Østergaard, J.; Friberg-Johansen, H.; Jessen, M. N. B.; Larsen, C. Eur. J. Pharm. Sci. 2006, 29, 348−354. (43) Larsen, S. W.; Frost, A. B.; Østergaard, J.; Marcher, H.; Larsen, C. Eur. J. Pharm. Sci. 2008, 34, 37−44. (44) Dyhre, H.; Soderberg, L.; Bjorkman, S.; Carlsson, C. Reg. Anesth. Pain Med. 2006, 31, 401−408. (45) Chorny, M.; Levy, R. J. Proc. Natl. Acad. Sci., U.S.A. 2009, 106, 6891−6892. (46) Pedersen, B. T.; Østergaard, J.; Larsen, S. W.; Larsen, C. Eur. J. Pharm. Sci. 2005, 25, 73−79. (47) Larsen, S. W.; Thomsen, A. E.; Rinvar, E.; Friis, G. J.; Larsen, C. Int. J. Pharm. 2001, 216, 83−93. (48) D’Souza, S. S.; Deluca, P. P. Pharm. Res. 2006, 23, 460−474. (49) Washington, C. Int. J. Pharm. 1990, 58, 1−12. (50) Washington, C. Int. J. Pharm. 1989, 56, 71−74. (51) Nielsen, A. B.; Buur, A.; Larsen, C. Eur. J. Pharm. Sci. 2005, 24, 433−440. (52) Amenitsch, H.; Rappolt, M.; Kriechbaum, M.; Mio, H. L. P.; et al. J. Synchrotron Rad. 1998, 5, 506−508. (53) Hammersley, A. P. ESRF Internal Report, ESRF97HA02T 1997. (54) Rappolt, M. In Planar Lipid Bilayers and Liposomes; Leitmannova, A., Ed.; Elsevier Book Series: Amsterdam, 2006; p 253. (55) Engström, S.; Norden, T. P.; Nyquist, H. Eur. J. Pharm. Sci. 1999, 8, 243−254. (56) Engström, S.; Engström, L. Int. J. Pharm. 1992, 79, 113−122. (57) Chang, C. M.; Bodmeier, R. Int. J. Pharm. 1997, 147, 135−142. (58) Dong, Y. D.; Larson, I.; Hanley, T.; Boyd, B. J. Langmuir 2006, 22, 9512−9518. (59) Yaghmur, A.; de Campo, L.; Salentinig, S.; Sagalowicz, L.; Leser, M. E.; Glatter, O. Langmuir 2006, 22, 517−521. (60) Engström, L. J. Disper. Sci. Technol. 1990, 11, 479−489. (61) Kozlov, M. M.; Leikin, S.; Rand, R. P. Biophys. J. 1994, 67, 1603−1611. (62) Mares, T.; Daniel, M.; Perutkova, S.; Perne, A.; Dolinar, G.; Iglic, A.; Rappolt, M.; Kralj-Iglic, V. J. Phys. Chem. B 2008, 112, 16575−16584. (63) Clogston, J.; Rathman, J.; Tomasko, D.; Walker, H.; Caffrey, M. Chem. Phys. Lipids 2000, 107, 191−220. (64) Shah, M. H.; Paradkar, A. Eur. J. Pharm. Biopharm. 2007, 67, 166−174. (65) Burrows, R.; Collett, J. H.; Attwood, D. Int. J. Pharm. 1994, 111, 283−293. (66) Phelps, J.; Bentley, M. V.; Lopes, L. B. Colloids Surface B 2011, 87, 391−398. (67) Clogston, J.; Craciun, G.; Hart, D. J.; Caffrey, M. J. Controlled Release 2005, 102, 441−461. (68) Larsen, S. W.; Jessen, M. N. B.; Østergaard, J.; Larsen, C. Drug Dev. Ind. Pharm. 2008, 34, 297−304. (69) Mezzenga, R.; Meyer, C.; Servais, C.; Romoscanu, A. I.; Sagalowicz, L.; Hayward, R. C. Langmuir 2005, 21, 3322−3333. (70) Ahmed, A. R.; Dashevsky, A.; Bodmeier, R. Eur. J. Pharm. Biopharm. 2010, 75, 375−380. (71) Negrini, R.; Mezzenga, R. Langmuir 2011, 27, 5296−5303.
2889
dx.doi.org/10.1021/la203577v | Langmuir 2012, 28, 2881−2889
