Article pubs.acs.org/Langmuir
Langmuir Balance Investigation of Superoxide Dismutase Interactions with Mixed-Lipid Monolayers Antonio P. Costa, Xiaoming Xu,† and Diane J. Burgess* Department of Pharmaceutical Sciences, University of Connecticut, 69 North Eagleville Road U3092, Storrs, Connecticut 06269, United States S Supporting Information *
ABSTRACT: Higher than theoretical encapsulation efficiencies in liposomes of the cytoplasmic protein, superoxide dismutase (SOD), were previously observed. The high encapsulation of SOD led to the consideration of lipid− protein interactions and the embedding of SOD in the lipid bilayer. Difficulty in other methods such as dynamic scanning calorimetry due to cholesterol obscuring the measurements brought about the interest for a modified Langmuir monolayer relaxation study. A novel method was devised to distinguish between different lipid compositions that formed either a favorable or an unfavorable environment for SOD. Normalized monolayer relaxations with SOD were compared between mixed-lipid compositions containing 1,2-dipalmitoyl-sn-glycero-3-phosphocholine (DPPC), 1,2-distearoyl-sn-glycero-3-phosphocholine (DSPC), and cholesterol (Chol). Lipid-monolayer relaxation with and without SOD in the subphase was plotted over 30 min to determine if the protein was altering the lipid-monolayer relaxation. The monolayer relaxation with SOD was normalized to the monolayer relaxation without SOD over the 30 min period. The results indicated that lipid length and mole percent of cholesterol were important parameters that must be adjusted in order to support a favorable environment for SOD interaction with the lipid. It was determined that hydrophobic interactions were dominant over electrostatic forces; thus, SOD was embedding into the lipid monolayer. Additionally, this study was correlated to a previous liposome study and proved that lipid− protein interactions were the reason for the higher encapsulation efficiencies. The significance of this method is that it (1) provides a connection between lipid−protein interactions observed in monolayers and bilayers and (2) establishes a simple and effective manner to test lipid compositions for lipid−protein interaction that will aid in optimization of liposome encapsulation efficiency.
1. INTRODUCTION In a previous study, it was shown that different lipid compositions had a significant impact on the encapsulation efficiency of the cytoplasmic protein, superoxide dismutase (SOD), in liposomes.1 SOD is a hydrophilic protein that is one of the most potent antioxidants in nature and has therapeutic use in inflammatory disorders.2 In some cases, the amount of SOD being encapsulated was much greater than the theoretically calculated encapsulation efficiency based solely on available volumetric space in the aqueous cores of the liposomes.3 This positive deviation indicated that SOD may also be embedded in the lipid bilayer. There are difficulties in assessing lipid−protein interactions, especially when they take place in the lipid bilayer of a liposome. A technique that has been used to evaluate liposome−protein interactions is dynamic scanning calorimetry (DSC).4 However, most liposomes include cholesterol as a bilayer stabilizing agent and the cholesterol−lipid interactions decrease the enthalpy changes associated with lipids and result in broader peaks.5 Accordingly, it is problematic to observe any changes in the DSC measurements that might be due to lipid− © 2012 American Chemical Society
protein interactions (unpublished results). In a previous study, an adapted light-scattering technique was developed to assess lipid−protein interactions in liposomes containing cholesterol in the bilayer.1 This method, suggested by Michel et al., involves measuring the optical properties of a material during temperature variation to determine its phase transition behavior.6 It was shown that for some lipid compositions the lipid phase transition was altered in the presence of SOD from a single to a dual transition, indicating that SOD was in fact in the lipid bilayer. To further determine lipid−protein interactions, a new method was developed using a Langmuir balance to investigate lipid-monolayer relaxation. Briefly, this method involves comparing the monolayer relaxation of a lipid mixture without protein to a lipid mixture with protein. Monolayer relaxation of the lipid mixture with protein was normalized to that without protein. Any positive or negative deviation of this normalized Received: April 20, 2012 Revised: June 1, 2012 Published: June 6, 2012 10050
dx.doi.org/10.1021/la301614t | Langmuir 2012, 28, 10050−10056
Langmuir
Article
2.3. SOD Adsorption/Desorption Isotherms. An isotherm for SOD was obtained to show the behavior of the protein at the surface. The protein was added to the subphase, and 60 min was allowed for the protein to diffuse to the surface. The barriers were moved at a rate of 1.5 mm/min until reaching the target surface pressure of 25 mN/m. Isotherms were plotted as surface pressure (mN/m) versus trough area (cm2). In order to further determine if SOD was desorbing from the interface over a range of surface pressures, 20 cycles of barrier compression/expansion over three units of surface pressure (i.e., 17− 19 mN/m) were recorded; only the last five cycles were used to allow for preconditioning. 2.4. Monolayer Compression Isotherms. Surface pressure versus mean molecular area isotherms of DPPC:Chol:SA (6:3:1) and DSPC:Chol:SA (7:2:1) were compared to the corresponding lipid compositions with SOD in the subphase up to a surface pressure of 40 mN/m. The isotherms were not recorded until the collapse pressure and used mainly to show protein adsorption/desorption at the interface. When the protein was added to the subphase, the mean molecular area data entered into the software was based only on the lipid and not on the SOD in the subphase. Accordingly, any positive change in the mean molecular area of the isotherm is due to the presence of SOD, but the exact amount cannot be quantified using this technique. 2.5. Monolayer Relaxation Study. Before initiating recording, the barriers were moved at a constant compression rate of 1.5 mm/ min until the desired target surface pressure (25 and 35 mN/m) was reached. The surface pressure was then held constant by allowing the barriers to move (at a rate of 1.5 mm/min) in order to compensate for any change in surface pressure due to lipid/protein reorganization and/or desorption. The trough area versus time profile was then recorded for 30 min after the target surface pressure was reached to assess the relaxation behavior of the monolayer. The trough area for each sample was normalized to the initial trough area at the target surface pressure. This normalized area was then plotted against time to determine the monolayer relaxation profile of each sample. To evaluate the effect of lipid−SOD interactions, the monolayer relaxation profile of the lipid mixture + SOD was divided by the relaxation of the corresponding lipid mixture. All samples were run in duplicate, and less than 1.5% error was observed.
relaxation would indicate that protein was favorably or unfavorably interacting with the lipid monolayer, respectively. Surface pressures for this study near 30−35 mN/m were chosen based on the bilayer equivalence pressure for a lipid monolayer.7 An equivalent surface pressure of approximately 35 mN/m was also established in a study where a lipid monolayer was formed at equilibrium with a suspension of unilamellar dioleoyl-PC liposomes in the subphase.8 However, it is difficult to directly relate lipid-monolayer studies to lipid-bilayer studies since these systems are inherently different. Specifically correlating a lateral surface pressure with the surface pressure found in the lipid bilayer is difficult since the lateral tension in a bilayer is inherently zero.7 However, a monolayer study can provide useful information on the phenomena that are occurring in a lipid bilayer. For example, the mean molecular area of lipid molecules in a monolayer has been found to be similar to those in lipid bilayers.9 At surface pressures around 20 mN/m, different molecular weight proteins on the surface of an air−water interface tend to desorb.10 Above this surface pressure the protein will continuously desorb unless hydrophobic interactions or electrostatic forces prevent the protein from desorbing. It is hypothesized that for different lipid compositions this Langmuir monolayer relaxation study will be able to differentiate between different lipid compositions (DPPC or DSPC in combination with cholesterol and stearylamine) that form favorable and unfavorable environments for SOD. Furthermore, it will be possible to relate these interactions to those in liposome formulations where high encapsulation efficiencies of SOD were observed.
2. MATERIALS AND METHODS 2.1. Lipid and Protein Preparation. 1,2-Dipalmitoyl-sn-glycero3-phosphocholine (DPPC), 1,2-distearoyl-sn-glycero-3-phosphocholine (DSPC), and cholesterol (Chol) (ovine wool, >98%) were purchased from Avanti Polar Lipids. Stearylamine (SA) was purchased from Sigma. The following lipid compositions were prepared: DPPC:Chol:SA (molar ratios 9:0:1, 6:3:1, and 7:2:1) and DSPC:Chol:SA (molar ratios 6:3:1 and 7:2:1). Lipid mixtures were dissolved in chloroform (Fischer Scientific, C297) to a final concentration of ca. 1 mg/mL. Bovine superoxide dismutase (SOD) was purchased from Sigma (S7571-300KU). SOD stock solution was prepared in 10 mM pH 7.4 HEPES buffer at a concentration of ca. 400 mM, and samples were stored at −25 °C in 330 μL aliquots prior to use. 2.2. Trough and Sample Preparation. The Langmuir−Blodgett Minitrough (KSV 1000) and barriers were extensively rinsed with distilled water (18.2 mOhms) and pure ethanol (Fischer Scientific, S93232) and subsequently dried with high-purity nitrogen (Airgas). A glass capillary tube, used to suction the surface prior to use, was cleaned in the same manner. The Wilhelmy plate was rinsed with distilled water and a water:ethanol (3:7) mixture and then heated using a Bunsen burner before each run. Ten millimolar pH 7.4 HEPES buffer was used as the subphase (0.22 μm filtered followed by 10 min sonication). For samples containing protein, a 0.29 mM SOD solution was prepared using SOD stock solution by adding it to HEPES buffer prior to filling the trough. The lipids (dissolved in chloroform) were spread over the subphase using a 50 μL Hamilton syringe (the initial surface pressure was less than 0.30 mN/m). Once the lipid was pipetted onto the surface, the chloroform was allowed to evaporate (10 min in the absence of SOD and 60 min in the presence of SOD) before initiating movement of the barrier using the KSV-Nima software. The lipid molecular weight, concentration, and volume were entered into the software to determine the mean molecular area of the lipid. The trough temperature was maintained at 25.0 ± 0.1 °C using a Haake B5 circulator.
3. RESULTS 3.1. SOD Behavior at the Air−Water Interface. An attempt was made to obtain a surface pressure vs mean molecular area isotherm of pure SOD to determine the amount of hydrophobic residues that support interfacial adsorption at the air−water interface. However, due to its hydrophilic nature, SOD that was added directly to the surface diffused into the subphase, thus rendering inadequate any mean molecular area measurement. Since the mean molecular area of SOD could not be determined, the surface pressure was plotted vs trough area (Figure 1). At low surface pressures (
