Lipid-Based Liquid Crystals As Carriers for Antimicrobial Peptides:
Phase Behavior and Antimicrobial E ff ect
Lukas Boge,*
,†,#Helena Bysell,
†Lovisa Ringstad,
†David Wennman,
‡Anita Umerska,
§Viviane Cassisa,
∥Jonny Eriksson,
⊥Marie-Laure Joly-Guillou,
∥Katarina Edwards,
⊥and Martin Andersson
#†SP Technical Research Institute of Sweden, Drottning Kristinas vag 45, Box 5607, Stockholm SE 11486, Sweden̈
‡SP Process Development, Forskargatan 18, Box 36, Södertälje SE 15121, Sweden
§Inserm U1066, University of Angers, 4 rue Larrey, Cedex 9 Angers FR 49933, France
∥Laboratoire de Bacteriologie-Hygié ne, CHU Angers, 4 rue Larrey, Angers FR 49000, Francè
⊥Department of Chemistry - BMC, Uppsala University, Husargatan 3, Box 579, Uppsala SE-75123, Sweden
#Department of Chemical and Chemical Engineering, Applied Chemistry, Chalmers University of Technology, Kemigården 4, Göteborg SE-41296, Sweden
*S Supporting Information
ABSTRACT: The number of antibiotic-resistant bacteria is increasing worldwide, and the demand for novel antimicrobials is constantly growing. Antimicrobial peptides (AMPs) could be an important part of future treatment strategies of various bacterial infection diseases. However, AMPs have relatively low stability, because of proteolytic and chemical degradation. As a consequence, carrier systems protecting the AMPs are greatly needed, to achieve efficient treatments. In addition, the carrier system also must administrate the peptide in a controlled manner to match the therapeutic dose window. In this work, lyotropic liquid crystalline (LC) structures consisting of cubic glycerol monooleate/water and hexagonal glycerol mono-
oleate/oleic acid/water have been examined as carriers for AMPs. These LC structures have the capability of solubilizing both hydrophilic and hydrophobic substances, as well as being biocompatible and biodegradable. Both bulk gels and discrete dispersed structures (i.e., cubosomes and hexosomes) have been studied. Three AMPs have been investigated with respect to phase stability of the LC structures and antimicrobial effect: AP114, DPK-060, and LL-37. Characterization of the LC structures was performed using small-angle X-ray scattering (SAXS), dynamic light scattering,ζ-potential, and cryogenic transmission electron microscopy (Cryo-TEM) and peptide loading efficacy by ultra performance liquid chromatography. The antimicrobial effect of the LCNPs was investigatedin vitrousing minimum inhibitory concentration (MIC) and time-kill assay. The most hydrophobic peptide (AP114) was shown to induce an increase in negative curvature of the cubic LC system. The most polar peptide (DPK-060) induced a decrease in negative curvature while LL-37 did not change the LC phase at all. The hexagonal LC phase was not affected by any of the AMPs. Moreover, cubosomes loaded with peptides AP114 and DPK-060 showed preserved antimicrobial activity, whereas particles loaded with peptide LL-37 displayed a loss in its broad-spectrum bactericidal properties.
AMP-loaded hexosomes showed a reduction in antimicrobial activity.
■
INTRODUCTIONThe number of antibiotic-resistant bacteria is constantly increasing and has become a vast problem in the global health sector. Antimicrobial peptides (AMPs) are present in all organisms as part of the innate defense system. These peptides are promising therapeutics to treat various infectious diseases, because of their fast and nonspecific mechanism of action and because bacteria are less prone to develop high-level resistance toward these antimicrobials.1,2AMPs are generally constituted of fewer than 50 amino acids, have a positive net-charge, and are often amphiphilic.2The main challenge for clinical translation of AMPs has previously been the lack of chemical and proteolytic stability, but today these issues can be overcome by using clever
modifications and formulation strategies, such as incorporating them into well-designed drug delivery vehicles.3
Different types of delivery systems have been investigated for peptide delivery, such as a variety of lipid-based formulations.4,5 Lyotropic liquid crystalline (LC) structures of polar lipids have potential applications as carriers and delivery systems in various pharmaceutical applications. This is due to their ability to solubilize and encapsulate both hydrophilic and hydrophobic substances. An important feature of these LC systems is that they
Received: January 29, 2016 Revised: March 15, 2016 Published: March 31, 2016
pubs.acs.org/Langmuir
can coexist with an excess of water, which enables fragmentation of bulk gels into liquid crystalline nanoparticles (LCNPs) in the presence of a suitable stabilizer.6,7Polar peptides and proteins can be incorporated in the lipid self-assembled structures and thus being less subjected to chemical and proteolytic degrada- tion.8−11LCNPs has been investigated as carriers for the delivery of several peptides and proteins; insulin,12 ovalbumin,13−15 somatostatin,16 and cyclosporine A17,18 are a few examples.
These studies report encouraging results, in terms of observed blood glucose values (insulin, in vivo, rat), sustained release properties (ovalbumin, in vivo, rat), increased half-life time (somatostatin, in vivo, rat), and increased peptide skin penetration (cyclosporine A,in vivo, mice).
One of the most extensively studied LC systems in drug delivery is the glycerol monooleate (GMO)/water system.19 Depending on the nature of introduced guest molecules in such LC systems (e.g., the polarity and molecular structure), the molecules will be located either in the water or lipophilic domains of the system.20It has been shown that the addition of lipophilic compounds to the GMO/water LC cubic system induces a transition to a reverse hexagonal phase (increased negative curvature), while molecules having a pronounced amphiphilic character can induce transition to a lamellar LC phase (decreased negative curvature).19,21,22AMPs comprising differences in amphiphilicity could be prone to induce phase transitions, as described above, when incorporated in such systems. LC phase transitions have previously been shown for a variety of protein−LCNP systems.23−29
Today, little is known about LCNPs as drug delivery vehicles for AMPs and whether the antimicrobial effect can be preserved or even improved. So far, only a few antimicrobial peptides have been formulated in LC gels or colloidal lipid formulations.
One example is the antimicrobial peptide melittin, which was successfully incorporated in a lipid-disk formulation.30The for- mulation showed a preserved antimicrobial effect, as compared to unformulated peptide. Also, the lipid-disk formulation protected the peptide for proteolytic degradation and demon- strated bactericidal properties upon repeatedly exposure to bacteria, indicating a sustained release of peptides from the lipid disks. Moreover, different liposomal formulations have been investigated for the delivery of several AMPs.31
In this study, the AMPs named AP114, DPK-060, and LL-37 were investigated, together with LC systems as carriers. The three water-soluble AMPs display differences in antimicrobial selectivity, origin, and physical properties. The peptide AP114 (also known as NZ2114) is an improved derivative of plectasin, which is a defensin peptide produced by the fungus Pseudoplectania nigrella.32,33In contrast to other AMPs, which act by disrupting bacterial membranes, AP114 inhibits the membrane biosynthesis by targeting the cellular precursor Lipid II.32 AP114 is active against Gram-positive bacteria,
including Staphylococcus aureus and its methicillin-resistant variety (MRSA), making it useful in the treatment of pneumonia.
The DPK-060 peptide is derived from the endogenous human protein kininogen and is intended for topical administration to treat a variety of infected skin conditions. Its effect has been confirmed by clinical phase II studies as treatment for infec- tions in atopic dermatitis and acute external otitis.3LL-37 is the only known human peptide in the cathelicidin family, found in different cells, tissues, and body fluids.34 LL-37 also has wound-healing properties, besides its broad-spectrum antibacte- rial activity, making it suitable for topical administration on infected wounds. The peptide is sensitive to various bacterial proteolytic enzymes, which so far has limited its therapeutic use.35
Briefly, the purpose of this present research was to investigate how three different AMPs influenced the LC structure of a cubic (GMO/water) and hexagonal (GMO/oleic acid (OA)/water) gel and to evaluate the antimicrobial effect of dispersions of those, that is cubosomes and hexosomes. Characterization of the LC gels was performed using small-angle X-ray scattering (SAXS). The peptide-loaded LCNPs were characterized in terms of particle size, ζ-potential, structure (SAXS and cryogenic transmission electron microscopy (Cryo-TEM)), and peptide loading efficacy (ultra performance liquid chromatography (UPLC)). The antimicrobial effect of the LCNPs was inves- tigatedin vitro, using minimum inhibitory concentration (MIC) tests and time-kill assays.
■
EXPERIMENTAL SECTIONMaterials. Glycerol monooleate RYLO MG 19 Pharma was purchased from Danisco A/S (Grindsted, Denmark). The composi- tion of the sample, according to the manufacturer, was min. 95%
monoglycerides, max. 10% diglycerides, and max. 2% triglycerides. The fatty acid content was min. 88% oleoyl (C18:1). Super-refined oleic acid NF-LQ-(MH) was purchased from Croda, Inc. (Snaith, U.K.). The triblock copolymeric stabilizer poly(ethylene oxide)−poly(propylene oxide)−poly(ethylene oxide) (PEO-PPO-PEO, tradename Lutrol F127) was obtained from BASF (Lundwigshafen, Germany) and had an approximate formula of PEO101PPO56PEO101 and an average molecular weight of 12 600 g/mol. AP114 (99.1% purity) was provided by Adenium Biotech ApS (Copenhagen, Denmark), DPK-060 (98.5% purity) was synthesized by Bachem AG (Bubendorf, Swizerland) and provided by Pergamum AB (Stockholm, Sweden) and LL-37 (94.7% purity) was synthesized and provided by PolyPeptide Laboratories (Limhamn, Sweden). The structure and properties of the peptides are summarized in Table 1. All substances were used as received.
Sample Preparation. LC gels were prepared by mixing melted GMO or GMO/OA mixtures (40°C) with a water solution contain- ing AMP. Samples were homogenized with a spatula, centrifuged at 3000 rpm for 1 h, and allowed to equilibrate at room temperature for at least 1 week prior to further use. The lipid:water ratio for the LC gels prepared without AMP were chosen to obtain the cubic or hexagonal Table 1. Structure and Properties of the AMPsa
aAll peptides were soluble in water at the used concentrations. Amino acid (AA) color code: black letters = neutral, red letters = lipophilic, blue letters = positively charged, yellow letters = negatively charged.
Langmuir Article
phase, so the peptide’s impact of the curvature could be investigated.
LCNP dispersions were made by fragmenting 0.5 g (5 wt %) of the LC gel in 9.5 g 5 mM acetic acid buffer pH 5.5 containing 1.0 % Lutrol F127, with a Ultra-Turrax high-shear mixer (Model IKA T25, Staufen, Germany) operated at 15 000 rpm for 1−2 min. The coarse dispersion were ultrasonicated using a Vibra-Cell VC 750 system (Sonics and Materials, Inc., Newtown, CT, USA) with a 6 mm probe operating at 40% of its maximum power in pulse mode (3-s pulses, followed by 7-s breaks) for a total time of 5 min (GMO gels) or 10 min (GMO/OA gels). The GMO/OA ratio (8:2) was chosen according to data presented by Salentinig et al.,22 and was shown to result in LCNPs with a hexagonal structure at pH 5.5.Table 2summarizes the prepared samples.
Particle Size Measurements.The particle size and its distributions was measured by dynamic light scattering (DLS), using a Zetasizer Zen3600 instrument (Malvern Instruments Ltd., Worcestershire, U.K.) using disposable cuvettes. The samples were diluted to a concentration of 2−2.5 mg/mL in Milli-Q water prior to analysis, and each sample was measured in triplicate. Refractive indices were set to 1.47 and 1.33, for lipids and water, respectively, and the temperature was held at 25°C during the measurements. The software calculated the hydrodynamic radius, assuming spherical particles, using the Stokes−Einstein relation to describe the Brownian motion of the particles.
Zeta (ζ) Potential.Theζ-potential of the particles was determined by measuring the electrophoretic mobility with a Zetasizer Zen3600 (Malvern Instruments, Ltd., Worcestershire, U.K.), using the Smoluchowski model. Samples were diluted in Milli-Q water to a particle concentration of 2−2.5 mg/mL and analyzed in disposable measuring cells at 25°C.
Each sample was measured in triplicate.
Small-Angle X-ray Scattering (SAXS). Synchrotron SAXS measurements were performed at Beamline I911-SAXS (MAX IV Laboratory, Lund, Sweden) equipped with a fast read-out pixel detector (PILATUS 1 M, Dectris, Ltd., Baden, Switzerland).36Liquid crystalline gels were mounted in a multiple sample holder made of steel andfitted between Kapton sheets, whereas liquid samples were loaded into quartz capillaries (ID 1.0 mm) mounted in a steel holder. Samples were sub- jected to aflux of 5×1010photons/s with a beam size of 0.3 mm×0.3 mm at a wavelength of 0.91 Å. The exposure time was 100 s for each sample.
Measurements were carried out in air at 22°C, and at 37°C for selected
samples. Collected raw-data images were analyzed using the Bli911-4 GUI software, Version 4.12, and the LC mesophases were identified according to the Bragg peak spacings.37
Cryogenic Transmission Electron Microscopy (Cryo-TEM).For Cryo-TEM, specimens were prepared in a controlled environment vitrification system that was maintained at 25°C with a humidity close to saturation, to prevent evaporation from the sample during preparation.
A droplet of the LCNP dispersion, diluted by a factor of 5 in Milli-Q water, was placed onto a carbon-coated holey polymerfilm supported by a copper grid, gently blotted with afilter paper to form a thin liquidfilm (10−500 nm), and immediately plunged into liquid ethane at−180°C for vitrification. The sample grid was then kept at liquid nitrogen temperature and transferred into a Zeiss LIBRA-120 transmission electron microscope (Carl Zeiss, Oberkochen, Germany) operated at 80 kV in zero-loss bright-field mode. Digital images were recorded under low-dose conditions with a TRS slow scan CCD camera system (TRS GmbH, Germany) and iTEM software (Olympus Soft Imaging Solutions GmbH, Germany). An underfocus of 1−3μm was used in order to enhance contrast. The cryo-TEM technique has been described in detail elsewhere.38
Ultra Performance Liquid Chromatography (UPLC). The peptide loading efficacy of AMPs in the LCNPs was quantified indirectly by separating the LCNPs from the surrounding liquid by centrifugation at 15 000gfor 20−40 min through Amicon Ultra-0.5filter devices (Ultracel-100 K, Merck Millipore, Ltd. Corc, Ireland) with a 100 kDa molecular weight cutoff. The peptide concentration in the filtrate was then quantified using a Waters Acquity UPLC system (Waters Corp., Milford, MA, USA) with an ultraviolet (UV) detector.
The system was equipped with a C18 column (1.7μm, 2.1 mm × 150 mm; Model Acquity BEH300, Waters Corp., Milford, MA, USA) operated at 40°C. The mobile phases used were acetonitrile/water 10/90 by volume plus 0.03% trifluoroacetic acid (phase A) and acetonitrile/water 90/10 by volume plus 0.03% trifluoroacetic acid (phase B) at aflow rate of 500μL/min. A gradient was used where the mobile phase composition was changed from 100% phase A to 39% over a period of 3.4 min. The injection volume was 5μL, and the absorbance was measured at 205 nm (LL-37) or 215 nm (AP114 and DPK-060).
The concentration of peptide in the filtrate was determined using a calibration curve and a correction factor compensating for adsorption Table 2. Summary of the Composition of Cubic and Hexagonal LC Gels and LCNP Dispersions That Were Analyzed in the Studya
sample code
lipid composition, GMO:OA (wt %)
lipid:aqueous phase for LC
gels (wt %) AMP
AMP concentration in LC
gels (wt %) final AMP concentration (μg/mL) in dispersions
Cub-ref 100:0 70:30 0 0
Cub-AP-0.25 100:0 70:30 AP114 0.25 125
Cub-AP-0.5 100:0 70:30 AP114 0.5 250
Cub-AP-1.0 100:0 70:30 AP114 1.0 500
Cub-AP-1.5 100:0 70:30 AP114 1.5 750
Cub-DP-0.5 100:0 70:30 DPK-060 0.5 250
Cub-DP-1.0 100:0 70:30 DPK-060 1.0 500
Cub-DP-1.5 100:0 70:30 DPK-060 1.5 750
Cub-LL-0.5 100:0 70:30 LL-37 0.5 250
Cub-LL-1.0 100:0 70:30 LL-37 1.0 500
Cub-LL-1.5 100:0 70:30 LL-37 1.5 750
Hex-ref 80:20 80:20 0 0
Hex-AP-0.5 80:20 80:20 AP114 0.5 250
Hex-AP-1.0 80:20 80:20 AP114 1.0 500
Hex-AP-1.5 80:20 80:20 AP114 1.5 750
Hex-DP-0.5 80:20 80:20 DPK-060 0.5 250
Hex-DP-1.0 80:20 80:20 DPK-060 1.0 500
Hex-DP-1.5 80:20 80:20 DPK-060 1.5 750
Hex-LL-0.5b 80:20 80:20 LL-37 0.5 250
Hex-LL-1.0 80:20 80:20 LL-37 1.0 500
Hex-LL-1.5b 80:20 80:20 LL-37 1.5 750
aThe LCNPs were made by dispersing 5 wt % LC gel in a 5 mM acetic acid buffer (pH 5.5) containing 1% Lutrol F127 stabilizer.bSamples were not analyzed by SAXS.
Langmuir
to thefilter. The amount peptide incorporated or adsorbed to the LCNPs could then be calculated, in terms of a percentage of total added peptide.
Minimum Inhibitory Concentration (MIC). The bacteriostatic behavior of pure peptides and peptide-loaded LCNPs was studied using the broth microdilution method. The bacteria used wereStaphylococcus Figure 1.(A, C, and E) SAXS diffractograms of cubic GMO:water (70:30 (w/w)) and (B, D, and F) hexagonal GMO:OA:water (64:16:20 (w/w/w)) LC gels containing AMPs AP114, DPK-060, and LL-37. Cubic gels with AP114 gradually changed its LC structure to hexagonal, while DPK-060 induced the formation of another cubic phase:Pn3m. LL-37 did not induce phase changes in the cubic system. The hexagonal GMO/OA gels did not change the LC phase upon the addition of any of the peptides. Representative peak indexing for the cubicIa3d(panel A), hexagonal (panel B), and cubicPn3m (panel C) symmetries are displayed as the corresponding Miller indices {hkl}. Measurements were carried out at 22°C.
Langmuir Article
aureus(reference strain ATCC 25923), methicillin-resistantS. aureus (MRSA, clinical strain No. 0702E0196), Pseudomonas aeruginosa (reference strain ATCC 27853 and clinical strain No. 0704C0134), Escherichia coli(reference strain ATCC 25922), ESBLE. coli(clinical strain ATCC 9007550201), andAcinetobacter baumannii(AYE refer- ence strain ATCC BAA-1710). Inoculum was prepared by taking 1−2 colonies directly from a Columbia agar plate supplemented with sheep blood (Oxoid, France) into 2 mL of 0.85% NaCl solution, and the density of the microorganism suspension was adjusted to∼3.3×108 colony forming units (CFU)/mL (S. aureus) or to∼1.5×108CFU/mL (P. aeruginosa,E. coli,A. baumannii). The former suspensions were further diluted by a factor of 100 (3.3×106CFU/mL), and the latter were diluted 10 times (1.5×107CFU/mL) with brain heart infusion broth (BHI) (bioMérieux, France) for test items that contain AP114 or 1% BHI in water for test items that contain DPK-060 and LL-37.
Serial 2-fold dilutions of the tested AMP-loaded LCNP samples in BHI or 1% BHI in water were prepared in order to obtain the desired concentration range. An aliquot of 100μL of the bacterial suspension was added into each well of a sterile 96-well plate containing 100μL of the tested sample. Positive controls, containing only the medium and the bacterial suspension (growth control) and negative control wells, containing the medium and the tested sample without the bacterial suspension (sterility control), were also prepared. The plates were incubated at 37 °C for 24 h. The MIC was defined as the lowest concentration of the sample that completely inhibited the growth of the bacteria as detected by the unaided eye. If the assessment was not possible due to turbidity of the sample, an amount of∼2 μL was withdrawn from each well, transferred onto a plate with Mueller Hinton agar using an multipoint inoculator (AQS Manufacturing, Ltd., U.K.), and incubated for 24 h at 37°C.
Time-Kill Assay. To evaluate the bactericidal activity of peptide- loaded LCNPs, time-kill assays were performed. An inoculum of∼3.3× 104CFU/mL was added to afinal volume of 2 mL in a polypropylene tube. The samples containing the tested LCNPs in BHI, as well as the controls were incubated at 37°C. At each sampling time (0, 3, 6, 18, and 24 h), an amount of 100μL was withdrawn from each tube. Serial 100-fold dilutions were prepared in distilled water when necessary.
A 100μL aliquot of the diluted and/or undiluted sample was delivered onto the surface of the agar and allowed to be absorbed into the agar.
Having incubated the agar plates for 24 h at 37°C, the colonies were counted.
■
Characterization of AMP-Loaded LC Gels.RESULTS First, the effect of incorporating AMPs into cubic and hexagonal LC gels was studied using SAXS, and diffractograms are presented inFigure 1.Characterizations of the dispersed LC gels are shown in the next section. As can be seen, when peptides were added to the cubic gel, the LC structure was retained (LL-37), changed to another cubic symmetry (DPK-060), or gradually turned into a hexagonal phase (AP114). The LC structure of the hexagonal gels was unaltered upon the incorporation of any of the AMP at the studied concentrations.
The unloaded reference cubic GMO LC gel showed eight distinct Bragg reflections with spacing ratios of 61/2:81/2:141/2:161/2:201/2:221/2:241/2:261/2. These reflections are characteristic for a not fully hydrated GMO phase, belonging to the body-centered cubic Ia3d space group. Calculated lattice parameters for the gels can be found in Table 3.
Figure 1A shows how the diffraction pattern changes upon increased AP114 concentrations. At low peptide concentra- tions (≤0.5 wt %), the negative curvature decreased, as the Pn3mcubic space group dominates, with peak spacing ratios of 21/2:31/2:41/2:61/2:81/2:91/2111/2:141/2. However, at higher AP114 concentrations, the structure was gradually transformed to a hexagonal phase with characteristic spacings of 1:31/2:41/2. This transition reflects an increase in negative curvature. Unit-cell
dimensions for thePn3mincreased slightly with increased AP114 content, whereas the dimension for the H2phase decreased, as displayed inTable 3. DPK-060 gradually changed the structure from cubic Ia3d to cubic Pn3m (Figure 1C), equivalent to a decrease in negative curvature. The lattice parameter for both the Ia3dandPn3mphases in the DPK-060 system decreased slightly with increasing peptide concentration. Moreover, the peptide LL-37 did not change the cubicIa3dLC phase at the studied concentrations (see Figure 1E). However, peaks were slightly shifted to higherq-values, reflecting a decrease in lattice parameter.
Unlike the cubic GMO gels, the hexagonal GMO/OA gels retained their LC structure upon peptide addition. The lattice parameter for the HIIphase without AMP (48.9 Å) was increased, irrespective of which AMP that was introduced. DPK-060 increased the hexagonal lattice parameter the most, up to 56.9 Å for the 1.0% sample. The lattice parameter for the hexagonal gel loaded with 1 wt % LL-37 was calculated to be 54.4 Å, which is between the lattice parameters for the AP114 and DPK-060 samples.
Characterization of AMP-Loaded LCNPs. Experiments using a high-pressure homogenizer (microfluidizer) to make LCNP dispersions were initially performed, but resulted in dispersions dominated by vesicles, as revealed by cryo-TEM.
Consistent with the cryo-TEM images, those samples did not display any LC structure (DPK-060 and LL-37) by SAXS, and low peptide loading efficacy was observed (AP114 and DPK-060).
Data are presented in the Supporting Information (Table S2, Figure S1, and Figure S2).
The internal LC structure of the dispersed LC gels was also characterized via SAXS at 22 °C, and at physiological tem- perature (37°C) for references and 1 wt % AMP-loaded LCNP samples. Diffractograms and calculated lattice parameters are dis- played inFigure 2andTable 4, respectively. For the GMO-based LCNPs (Figures 2A,2C, and2E), the reference sample showed Table 3. Calculated Lattice Parameter (a) Values for the Different Space Groups Present in the AMP-Loaded Cubic and Hexagonal LC Gelsa
Lattice Parameter,a(Å)
gel sample phase Ia3d Pn3m Im3m HII
Cub-ref Ia3d 141.3/134.8b
Cub-AP-0.25 Pn3m 84.3
Cub-AP-0.5 Pn3m, H2+ traces of Im3mandIa3d
139.5 86.0 113.3 60.3
Cub-AP-1.0 H2,Im3m 115.8 58.6
Cub-AP-1.5 H2 57.6
Cub-DP-0.5 Ia3d 133.7
Cub-DP-1.0 Pn3m,Ia3d 130.6 84.0
Cub-DP-1.5 Pn3m 82.9
Cub-LL-0.5 Ia3d 135.0
Cub-LL-1.0 Ia3d 130.4
Cub-LL-1.5 Ia3d 131.4
Hex-ref H2 48.9
Hex-AP-0.5 H2 51.6
Hex-AP-1.0 H2 51.9
Hex-AP-1.5 H2 51.0
Hex-DP-0.5 H2 56.1
Hex-DP-1.0 H2 56.9
Hex-DP-1.5 H2 55.1
Hex-LL-1.0 H2 54.4
aThe hydration level was 30 wt % for the cubic (GMO) LC gels and 20 wt % for hexagonal (GMO/OA) LC gels. Samples were measured at 22°C.bResults obtained from two different GMO batches.
Langmuir
peak spacing ratios of 21/2:41/2:61/2, indexed as (110), (200), and (211) reflections, distinctive for the Im3m cubic symmetry.
Independent of which symmetry the undispersed cubic LC gel
adopted, it contained the cubicIm3m space group when dis- persed. As for the AP114-containing GMO gels, a hexagonal phase was also present in the dispersions at the highest concentration Figure 2.SAXS diffractograms of GMO-based LCNPs loaded with (A) AP114, (C) DPK-060, and (E) LL-37, showing the presence of the cubicIm3m phase, and HIIin the case of AP114 loadings of 1.5 wt %, and at 1.0 wt %, kept at 37°C. The GMO/OA-based LCNPs with (B) AP114, (D) DPK-060, and (F) LL-37 all displayed a hexagonal LC structure for all peptide concentrations studied. All samples were measured at 22°C, except for samples loaded with 1 wt % AMP, which were also analyzed at a physiologically relevant temperature (37°C). The peptide loading in % reflects the initial peptide content in the undispersed LC gel. Representative peak indexing for the cubicIm3m(panel A) and hexagonal (panel B) symmetries are displayed as the corresponding Miller indices {hkl}. All samples contained 5 wt % gel dispersed in 5 mM acetic acid buffer (pH 5.5) containing 1% Lutrol F127 stabilizer.
Langmuir Article
studied. CoexistingIm3mand hexagonal phase were detected at 1.5% peptide loading or at 1% while being maintained at 37°C.
However, at 0.5 and 1.0% AP114 content at 22 °C, only the Im3mcubic phase was detected. The lattice parameter for the Im3mcubic phase decreased upon the addition of AP114, while the dimension of the observed hexagonal phase did not change.
Surprisingly, no diffraction peaks were detected for the GMO sample containing 0.5% DPK-060 (Figure 2B). Samples of 1.0%
(22 and 37°C) and 1.5% peptide content showed characteristic
peaks for theIm3mcubic symmetry, and the lattice parameter only changed slightly among the samples. LL-37-containing dispersions were all cubic Im3m and the lattice parameter analyzed at 22°C only differed slightly from the reference. For all samples analyzed at 37°C, peaks were slightly shifted to higher q-values, giving∼10 Å reductions in lattice parameter.
The GMO/OA-based LCNPs (Figures 2B, 2D, and 2F) displayed the presence of hexagonal structure, and the trends observed in the lattice parameter for the undispersed gels were also found for the LCNPs. The calculated hexagonal lattice parameters, inTable 4, varied a few Å from the gels. As for the cubic GMO-based dispersions, the lattice parameter of GMO/OA hexagonal LCNPs measured at 37°C was slightly lower, compared to those obtained at 22°C.
Furthermore, the cubic and hexagonal LCNP dispersions were characterized in terms of particle size,ζ-potential, and peptide loading efficacy, and the results are illustrated inFigure 3. For the cubic LCNPs, the mean particle size was always in the range of 100−130 nm with a narrow particle size distribution (PDI <
0.160). The mean particle size for the hexagonal LCNPs was in the range of 140−170 nm and showed slightly broader particle size distributions (PDI < 0.220). Theζ-potential for unloaded LCNPs was negative and increased upon peptide incorporation.
Generally, dispersed cubic GMO gels showed lower peptide loading efficacies (41%−86%), compared to dispersions of GMO/OA hexagonal LC gels (>94%). Different trends in loading efficacy were observed for the peptides loaded in the cubic LCNP, as observed in Figure 3A. The AP114 loading efficacy was higher as the initial loading was increased. The same trend was seen for peptide LL-37. However, DPK-060 adopted an opposite trend. No clear trends were observed for the hexagonal GMO/OA system (Figure 3B), taking the standard deviation into account (found inTable S1).
Representative cryo-TEM images of LCNPs loaded with 1.0 wt % AMP are shown inFigure 4. All GMO-based dispersions (Figures 4A−D) contained cubosome particles with ordered internal structure, in agreement with SAXS data. Unilamellar vesicles were also found, together with the cubosome particles.
The GMO/OA dispersions did also contain denser particles (hexosomes) with somewhat indefinite internal structure, together with vesicles.
In VitroAntibacterial Effect.Thein vitroantibacterial effect of the AMP-loaded LCNPs was characterized using MIC Table 4. Observed LC Phases and Their Lattice Parameter
(a) Values, for Dispersed GMO and GMO/OA LCNPs Loaded with AP114, DPK-060, and LL-37a
Lattice Parameter,a(Å) dispersed gel temperature,T(°C) phase Im3m H2
Cub-ref 22 Im3m 135.0
Cub-ref 37 Im3m 126.6
Cub-AP-0.5 22 Im3m 132.8
Cub-AP-1.0 22 Im3m 125.5
Cub-AP-1.0 37 Im3m, H2 114.2 59.5
Cub-AP-1.5 22 Im3m, H2 116.4 60.1
Cub-DP-0.5 22 no peaks
Cub-DP-1.0 22 Im3m 137.9
Cub-DP-1.0 37 Im3m 128.6
Cub-DP-1.5 22 Im3m 137.5
Cub-LL-0.5 22 Im3m 138.9
Cub-LL-1.0 22 Im3m 134.6
Cub-LL-1.0 37 Im3m 125.3
Cub-LL-1.5 22 Im3m 133.7
Hex-ref 22 H2 52.6
Hex -ref 37 H2 51.1
Hex -AP-0.5 22 H2 53.8
Hex -AP-1.0 22 H2 52.4
Hex -AP-1.0 37 H2 50.9
Hex -AP-1.5 22 H2 53.5
Hex -DP-0.5 22 H2 55.5
Hex -DP-1.0 22 H2 55.9
Hex -DP-1.0 37 H2 54.6
Hex -DP-1.5 22 H2 56.7
Hex -LL-1.0 22 H2 55.8
Hex -LL-1.0 37 H2 53.8
aSamples contained 5 wt % gel dispersed in 5 mM acetic acid buffer (pH 5.5) containing 1% Lutrol F127 stabilizer.
Figure 3.Peptide loading efficacy plotted as a function ofζ-potential, for dispersion of (A) cubic and (B) hexagonal LC gels. The diameter of the bubbles is proportional against its mean particle size. The cubic reference equals 127 nm and the hexagonal reference equals 159 nm (unfilled circles). Numbers inside circles represent the initial peptide loading (wt %) in dispersed gels. Detailed data can be found in the Supporting Information (Table S1).
Langmuir
determinations (Table 5) and time-kill assays (Figure 5). Peptide AP114, which is known to only be effective against Gram-positive bacteria, was not tested against P. aeruginosa, E. coli, and A. bumannii strains. LCNP dispersions with a total peptide concentration of 500μg/mL were tested. The majority of the peptide-loaded GMO-based LCNPs maintained the same anti- microbial activity as that for the unformulated peptide. DPK-060 formulated in GMO-based LCNPs significantly decreased the MIC (by 1−2 dilutions) on theE. colistrain, compared to the unformulated peptide. The LL-37 loaded GMO-sample showed a loss in its broad-spectrum effect and showed only preserved inhibition in growth ofP. aeruginosa(clinical strain) andE. coli.
However, an apparent decrease in antibacterial activity was noted for the AMP-loaded GMO/OA LCNPs. The samples did not show any signs of inhibiting bacterial growth up to a total peptide concentration of 16 μg/mL. None of the unloaded reference samples showed any antibacterial activity up to concentrations of 1200μg/mL.
Time-kill assays were performed in order to study the bactericidal properties of the formulations over a prolonged time period, on a selection of bacterial strains according to the MIC results. The hexagonal GMO/OA LCNPs were not tested,
because of their high MIC values. Data are presented inFigure 5.
AP114-loaded GMO-based LCNPs were tested on S. aureus (Figure 5A) and MRSA (Figure 5B) and showed a similar effect as that for the pure peptide (a difference of less than 2 logarithmic units). DPK-060-loaded LCNPs were tested at two different concentrations, 4 and 8μg/mL peptide, onS. aureus(Figure 5C) and MRSA (Figure 5D). The antimicrobial effect of the DPK-060 formulation did not differ significantly from the unformulated reference, except for the 4μg/mL sample tested on S. aureus, which showed a loss in bacterial killing after 6 h of incubation time (Figure 5C). LCNPs loaded with DPK-060 and LL-37 were also tested againstE. coliand, again, the effect was similar to unformulated peptide. Positive controls performed as anticipated and the LCNPs did not inhibit the growth of bacteria on their own (Control and Cub-ref curves).
■
DISCUSSIONIn this study, it was found that the cubic GMO/water gel was more sensitive to peptide additions, compared to the hexagonal GMO/OA systems, which always maintained their LC structure.
The nature of the peptides (hydrophobicity and net charge) strongly influenced the changes in curvature of the cubic gels.
Figure 4.Cryo-TEM images of dispersed (A−D) GMO and (E−H) GMO/OA LCNPs loaded with AP114, DPK-060, and LL-37. A quantity of 1 wt % peptide was initially loaded in the LC gel prior dispersion, resulting in a concentration of 500μg/mL in the formulations. Scale bar = 100 nm.
Table 5. MIC Values for Unformulated (UF) Peptide and Peptide Loaded in GMO and GMO/OA LCNPsa
MIC Values (μg/mL total AMP content)
AP114 DPK-060 LL-37
bacterial strain UF GMO GMO/OA UF GMO GMO/OA UF GMO GMO/OA
S. aureus 8 8 >16 4 2−4 >16 8−16 64 >16
MRSA 4 4 >16 4 2−4 >16 8−16 >16 >16
P. aeruginosa 8 16 >16 8−16 >16 >16
P. aeruginosaclinical strain 16 8−16 >16 8−16 8 >16
E. coli 8 2−4 >16 16 16 >16
ESBLE. coliclinical strain 4−8 2−4 >16 16 >16 >16
A. Baumannii 4−8 16 >16 16 >16 >16
aSamples with 1.0 wt % AMP in nondispersed LC gel, equal to a total AMP concentration of 500μg/mL in the LCNP dispersions, were tested.
AP114, which is only active against Gram-positive bacteria, were therefore only tested againstS. aureusand MRSA strains.
Langmuir Article
The antimicrobial effect of the cubic GMO LCNPs were, in most cases, comparable to unformulated peptide, while the effect for peptides loaded in GMO/OA LCNPs was drastically decreased.
As expected, the unloaded reference cubic LC gel belonged to theIa3dspace group, which is consistent with phase diagrams studies found in the literature at a hydration level of 30%.39,40 When the AP114 peptide was added to the system, having the
highest percentage of hydrophobic residues and lowest net charge among the tested AMPs, the structure gradually turned into the hexagonal phase. Hence, AP114 is suggested to interact strongly with the hydrophobic parts of the GMO membranes, thus increasing the critical packing parameter (CPP), and, in turn, increasing the negative curvature. Interestingly, the cubicPn3m phase was detected at lower peptide additions, representing a Figure 5.Time-kill assay curves for GMO-based LCNPs loaded with (A, B) AP114, (C−E) DPK-060, and (F) LL-37 on a selection of bacterial strains.
The bactericidal properties of the formulations (diluted to the chosen start concentration) were found to be very similar to the effect of pure peptide.
Each data point is represented by mean±standard deviation (n= 3).
Langmuir
decrease in negative curvature. Transitions fromIa3dtoPn3m, and also further to HII, have previously been observed for lipophilic molecules or amphiphilic peptides added to GMO-based systems.41 As seen for AP114 at low concentrations, DPK-060 changed the structure from cubicIa3dto cubicPn3mat peptide additions of >0.5%. This is probably a result of DPK-060 being the most hydrophilic peptide among the examined AMPs with the highest net charge. DPK-060 interacts strongly with the hydrophilic head groups of GMO, resulting in a decreased CPP and, in turn, the negative curvature. LL-37 did not change its cubicIa3dstructure, indicating that the peptide is mostly located in the water domains of the structure or is penetrating the bilayer interfaces without significantly changing the curvature. The lattice parameter of theIa3dphase decreased upon addition of all the investigated AMPs, at low concentrations. The change was∼10 Å for samples containing 1.0 and 1.5 wt % LL-37. The difference is not likely only due to the slight decrease in water content (some water is replaced by peptide in the aqueous phase). A 1% decrease in water content from 30 wt % gives a decrease of∼2 Å in theIa3dlattice parameter (extrapolation of data found in the literature39) and one can thus expect that LL-37 makes the packing of molecules more efficient by replacing part of the GMO molecules. An interesting observation for the dispersed cubic GMO gel loaded with 1% AP114 is the formation of a hexagonal phase upon heating at 37 °C (Table 4). An increase in temperature is known to result in dehydration of the polar head groups of nonionic amphiphiles, thus increasing the CPP and, in turn, the negative curvature toward hexagonal.42 The same reasoning could be adapted to explain the decrease in lattice parameter for LCNPs maintained at 37°C.
The observed increase in lattice parameter for the hexagonal GMO/OA peptide-loaded gels indicates a swelling of the structure. The most hydrophobic peptide, AP114, showed the smallest increase in the lattice parameter. The reason for this could be that the peptide is mostly located in the lipophilic parts of the structure, thus increasing the bulkiness and length of the hydrophobic tails, without changing the radius of the water channels. The greatest swelling was observed for DPK-060.
Because of its high positive charge and hydrophilicity, DPK-060 is most prone to interact with the polar head groups and water domains in the hexagonal structure, thus swelling the diameter of the cylinders, resulting in an increased lattice parameter. Also, electrostatic repulsion between the positive charges of the AMPs could explain the increase in lattice parameter.
The increase in ζ-potential upon peptide incorporation is probably due to localization of the cationic AMPs at the particle’s surface. This effect has previously been observed when the positively charged peptide hormone somatostatin was adsorbed onto negatively charged LCNPs.16The adsorbed AMPs are most likely in equilibrium with the free peptides present in the surrounding solution. The surface and/or structure of the GMO particles loaded with DPK-060 might already been saturated at 0.5% peptide loading, resulting in decreased loading efficacy and a smaller increase in ζ-potential at higher concentrations, compared to the trends for AP114- and LL-37-loaded LCNPs.
This indicates that the ζ-potential is not only reflecting the loading efficacy, but rather also the increase of nonencapsulated peptide. Our loading efficacies for the cubic GMO systems is comparable to the encapsulation of ovalbumin (47%−53%) and cyclosporine A (86%−92%) measured previously for similar carriers.13,18
Cryo-TEM images of both cubic and hexagonal dispersions displayed a fraction of dense particles with internal structure,
which is consistent with the diffraction peaks observed by SAXS.
Also, particle sizes were in agreement with DLS measurements.
FromFigure 4A, the repeating distance in the cubosome particles was measured to∼90−95 Å. This observation is in agreement with the calculated spacing from the (110) reflection in the SAXS data: 135/21/2= 95.5 Å. A large fraction of vesicles has previously been shown to be present in GMO/F127-based dispersions prepared by microfluidization or sonication.17,43,44The presence of bilayer vesicles has been proposed to be required during the cubosomes formation process, which appears to be slow at ambient temperatures.45One way of drastically increasing the fraction of cubosomes is to subject the sample to a heat treatment cycle in an autoclave.43However, this harsh treatment was tested for microfluidized samples and resulted in phase separation or dispersions that only contained large liposomes and degraded AMP (data not shown). Compared to the GMO-based dis- persions, the hexagonal GMO/OA particles appeared more spherically shaped, displayed a more fuzzy internal structure (Figures 4E−H) and did not contain any lamellar structures attached to the particle outer surfaces. These observations are consistent with cryo-TEM investigations of dispersed hexagonal phases that have been found elsewhere.17,46−48
DPK-060 formulated in GMO-based LCNPs appeared to be particularly effective in inhibiting growth ofE. coli, compared to unformulated peptide. This indicates a synergetic effect of encapsulating DPK-060 in cubic LCNPs, where not only the free (nonbounded) peptide could explain the observed antibacterial effect. The fact that the peptide-loaded GMO-based LCNPs performed as well as the unformulated reference indicates that there might be a burst release of peptide from the particles.
A burst release of lipophilic compounds from cubosomes has previously been shown,49 and could also be the case for the AMPs. The decrease in antimicrobial activity for the peptide- loaded GMO/OA LCNPs may be explained by the fact that peptides tend to prefer to remain in the hexagonal environment (no phase change observed by SAXS), resulting in high peptide loading efficacy (>94%) and a slow release from these particles.
Our MIC and time-kill assay data suggest that the release from the hexagonal particles is very low during thefirst 24 h.
The coexistence of vesicles and cubosome and hexosome particles inevitably bring about the question of which colloidal carrier is encapsulating and delivering the AMPs. The observed lower peptide loading in pure vesicles (microfluidized samples) and the fact that peptides did change the LC structure (which is clear in the case of AP114) indicates that the peptides are immobilized in the LC structure of the cubosomes. In case of the dispersed hexagonal systems, the antibacterial effect was determined to be poor, even though the peptide loading efficacy was high and the fraction of vesicles was comparable among the cubic and hexagonal systems. If the AMPs were mainly incorporated and delivered by vesicles, one would have expected that the hexagonal GMO/OA-based LCNPs should have given rise to an apparent bacterial killing effect in the MIC tests, which was not the case. Hence, it is believed that the cubosomes and hexosomes, and not the vesicles, are the major delivery vehicles for the AMPs.
■
CONCLUSIONIn this study, we have demonstrated that antimicrobial peptides (AMPs) could successfully be loaded in cubic and hexagonal liquid crystalline (LC) gels and in liquid crystalline nanoparticles (LCNPs) with high loading efficacies. Importantly, the cubic AMP-loaded LCNPs showed promising antimicrobial activity.
Langmuir Article
Moreover, it was demonstrated that the AMPs strongly influenced the LC structure of the cubic glycerol monooleate (GMO)/water system. Peptide net charge and degree of hydrophobicity are important factors affecting the curvature.
The hexagonal GMO/oleic acid (OA)/water system was found to be more robust upon the addition of any of the AMPs, and did not change the LC structure. Why peptides formulated in cubic LCNPs show sustained antimicrobial activity, but not if formulated in hexagonal LCNPs, is not fully understood, and this must be investigated further. Important aspects to investigate are peptide release profiles, possible release triggers from the LCNPs, and interaction with bacterial membranes. Such information would provide more details about the performance in the minimum inhibitory concentration (MIC) and time-kill assays. Moreover, the chemical and proteolytic stability of the AMPs formulated in LCNPs is of great interest to examine in the future.
■
ASSOCIATED CONTENT*S Supporting Information
The Supporting Information is available free of charge on theACS Publications websiteat DOI:10.1021/acs.langmuir.6b00338.
Additional tables andfigures (PDF)
■
AUTHOR INFORMATION Corresponding Author*Tel.: +46105166078. E-mail:lukas.boge@sp.se.
Notes
The authors declare no competingfinancial interest.
■
ACKNOWLEDGMENTSThis research performed in this study was funded by the European Union’s Seventh Framework Programme (FP7/2007- 2013), under Grant Agreement No. 604182 within the FORMAMP project. The MAX IV Laboratory is acknowledged for beam time at X-ray Synchrotron Beamline I911-SAXS. We would also like to thank Szymon Sollami Delekta (SP Technical Research Institute of Sweden, Stockholm, Sweden) for fine- tuning the preparation protocol for hexosomes and Anand Kumar Rajasekharan (Chalmers University of Technology, Gothenburg, Sweden) for running part of the SAXS samples.
■
(1) Hancock, R. E. W.; Sahl, H.-G. Antimicrobial and host-defenseREFERENCES peptides as new anti-infective therapeutic strategies. Nat. Biotechnol.2006,24(12), 1551−1557.
(2) Pasupuleti, M.; Schmidtchen, A.; Malmsten, M. Antimicrobial peptides: Key components of the innate immune system.Crit. Rev.
Biotechnol.2012,32(2), 143−171.
(3) Eckert, R. Road to clinical efficacy: Challenges and novel strategies for antimicrobial peptide development.Future Microbiol.2011,6(6), 635−651.
(4) Martins, S.; Sarmento, B.; Ferreira, D. C.; Souto, E. B. Lipid-based colloidal carriers for peptide and protein deliveryLiposomes versus lipid nanoparticles.Int. J. Nanomed.2007,2(4), 595−607.
(5) Matougui, N.; Boge, L.; Groo, A.-C.; Umerska, A.; Ringstad, L.;
Bysell, H.; Saulnier, P. Lipid-based nanoformulations for peptide delivery.Int. J. Pharm.2016,502, 80−97.
(6) Larsson, K. Cubic lipid−water phases: Structures and biomem- brane aspects.J. Phys. Chem.1989,93(21), 7304−7314.
(7) Gustafsson, J.; Ljusberg-Wahren, H.; Almgren, M.; Larsson, K.
Cubic Lipid−Water Phase Dispersed into Submicron Particles.
Langmuir1996,12(20), 4611−4613.
(8) Rizwan, S. B.; Boyd, B. J.; Rades, T.; Hook, S. Bicontinuous cubic liquid crystals as sustained delivery systems for peptides and proteins.
Expert Opin. Drug Delivery2010,7(10), 1133−1144.
(9) Shah, J. C.; Sadhale, Y.; Chilukuri, D. M. Cubic phase gels as drug delivery systems.Adv. Drug Delivery Rev.2001,47(2−3), 229−250.
(10) Angelova, A.; Angelov, B.; Mutafchieva, R.; Lesieur, S.; Couvreur, P. Self-Assembled Multicompartment Liquid Crystalline Lipid Carriers for Protein, Peptide, and Nucleic Acid Drug Delivery.Acc. Chem. Res.
2011,44(2), 147−156.
(11) Chen, Y.; Ma, P.; Gui, S. Cubic and Hexagonal Liquid Crystals as Drug Delivery Systems.BioMed Res. Int.2014,2014, 1.
(12) Chung, H.; Kim, J.; Um, J. Y.; Kwon, I. C.; Jeong, S. Y. Self- assembled “nanocubicle” as a carrier for peroral insulin delivery.
Diabetologia2002,45(3), 448−451.
(13) Rizwan, S. B.; Assmus, D.; Boehnke, A.; Hanley, T.; Boyd, B. J.;
Rades, T.; Hook, S. Preparation of phytantriol cubosomes by solvent precursor dilution for the delivery of protein vaccines.Eur. J. Pharm.
Biopharm.2011,79(1), 15−22.
(14) Gordon, S.; Young, K.; Wilson, R.; Rizwan, S.; Kemp, R.; Rades, T.; Hook, S. Chitosan hydrogels containing liposomes and cubosomes as particulate sustained release vaccine delivery systems.J. Liposome Res.
2012,22(3), 193−204.
(15) Rattanapak, T.; Young, K.; Rades, T.; Hook, S. Comparative study of liposomes, transfersomes, ethosomes and cubosomes for trans- cutaneous immunisation: Characterisation andin vitroskin penetration.
J. Pharm. Pharmacol.2012,64(11), 1560−1569.
(16) Cervin, C.; Vandoolaeghe, P.; Nistor, C.; Tiberg, F.; Johnsson, M.
A combinedin vitro andin vivo study on the interactions between somatostatin and lipid-based liquid crystalline drug carriers and bilayers.
Eur. J. Pharm. Sci.2009,36(4−5), 377−385.
(17) Lopes, L.; Ferreira, D.; de Paula, D.; Garcia, M. T.; Thomazini, J.;
Fantini, M. A.; Bentley, M. V. B. Reverse Hexagonal Phase Nano- dispersion of Monoolein and Oleic Acid for Topical Delivery of Peptides:In VitroandIn VivoSkin Penetration of Cyclosporin A.Pharm.
Res.2006,23(6), 1332−1342.
(18) Lai, J.; Lu, Y.; Yin, Z.; Hu, F.; Wu, W. Pharmacokinetics and enhanced oral bioavailability in beagle dogs of Cyclosporin A encapsulated in glyceryl monooleate/poloxamer 407 cubic nano- particles.Int. J. Nanomed.2009,5, 13−23.
(19) Milak, S.; Zimmer, A. Glycerol monooleate liquid crystalline phases used in drug delivery systems.Int. J. Pharm.2015,478(2), 569−
587.
(20) Caboi, F.; Amico, G. S.; Pitzalis, P.; Monduzzi, M.; Nylander, T.;
Larsson, K. Addition of hydrophilic and lipophilic compounds of biological relevance to the monoolein/water system. I. Phase behavior.
Chem. Phys. Lipids2001,109(1), 47−62.
(21) Borné, J.; Nylander, T.; Khan, A. Phase Behavior and Aggregate Formation for the Aqueous Monoolein System Mixed with Sodium Oleate and Oleic Acid.Langmuir2001,17(25), 7742−7751.
(22) Salentinig, S.; Sagalowicz, L.; Glatter, O. Self-Assembled Structures and pKa Value of Oleic Acid in Systems of Biological Relevance.Langmuir2010,26(14), 11670−11679.
(23) Angelova, A.; Ionov, R.; Koch, M. H. J.; Rapp, G. Interaction of the Peptide Antibiotic Alamethicin with Bilayer- and Non-bilayer- Forming Lipids: Influence of Increasing Alamethicin Concentration on the Lipids Supramolecular Structures.Arch. Biochem. Biophys.2000,378 (1), 93−106.
(24) Angelova, A.; Angelov, B.; Papahadjopoulos-Sternberg, B.;
Ollivon, M.; Bourgaux, C. Proteocubosomes: Nanoporous Vehicles with Tertiary Organized Fluid Interfaces.Langmuir2005,21(9), 4138−
4143.
(25) Angelova, A.; Angelov, B.; Papahadjopoulos-Sternberg, B.;
Ollivon, M.; Bourgaux, C. Structural organization of proteocubosome carriers involving medium- and large-size proteins.J. Drug Delivery Sci.
Technol.2005,15(1), 108−112.
(26) Angelova, A.; Angelov, B.; Lesieur, S.; Mutafchieva, R.; Ollivon, M.; Bourgaux, C.; Willumeit, R.; Couvreur, P. Dynamic control of nanofluidic channels in protein drug delivery vehicles.J. Drug Delivery Sci. Technol.2008,18(1), 41−45.
Langmuir
