Octadecyl ferulate behavior in 1,2-Dioleoylphosphocholine liposomes

Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy 153 (2016) 333–343

Contents lists available at ScienceDirect

Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy journal homepage: www.elsevier.com/locate/saa

Octadecyl ferulate behavior in 1,2-Dioleoylphosphocholine liposomes☆ Kervin O. Evans a,⁎, David L. Compton a, Nathan A. Whitman a,1, Joseph A. Laszlo a, Michael Appell b, Karl E. Vermillion c, Sanghoon Kim d a Renewable Products Technology Research Unit, National Center for Agricultural Utilization Research, Agricultural Research Service, United States Department of Agriculture, 1815 N. University Street, Peoria, IL 61604, USA b Bacterial Foodborne Pathogens and Mycology, National Center for Agricultural Utilization Research, Agricultural Research Service, United States Department of Agriculture, 1815 N. University Street, Peoria, IL 61604, USA c Functional Foods Research, National Center for Agricultural Utilization Research, Agricultural Research Service, United States Department of Agriculture, 1815 N. University Street, Peoria, IL 61604, USA d Plant Polymer Research, National Center for Agricultural Utilization Research, Agricultural Research Service, United States Department of Agriculture, 1815 N. University Street, Peoria, IL 61604, USA

a r t i c l e

i n f o

Article history: Received 27 February 2015 Received in revised form 17 July 2015 Accepted 4 August 2015 Available online 11 August 2015 Keywords: Fluorescence Antioxidation Octadecyl ferulate Partition coefficient Liposomes

a b s t r a c t Octadecyl ferulate was prepared using solid acid catalyst, monitored using Supercritical Fluid Chromatography and purified to a 42% yield. Differential scanning calorimetry measurements determined octadecyl ferulate to have melting/solidification phase transitions at 67 and 39 °C, respectively. AFM imaging shows that 5-mol% present in a lipid bilayer induced domains to form. Phase behavior measurements confirmed that octadecyl ferulate increased transition temperature of phospholipids. Fluorescence measurements demonstrated that octadecyl ferulate stabilized liposomes against leakage, maintained antioxidant capacity within liposomes, and oriented such that the feruloyl moiety remained in the hydrophilic region of the bilayer. Molecular modeling calculation indicated that antioxidant activity was mostly influenced by interactions within the bilayer. Published by Elsevier B.V.

1. Introduction Plants and fungi containing bioactive compounds have been used for centuries for their medicinal benefits. In recent years, identification and exploration of these bioactive compounds have been of particular interest to researchers. One such compound is ferulic acid (FA), which itself and its derivatives, are readily found in foods such as tomatoes, grain foods, citrus fruits, and coffee [1]. FA has been shown to inhibit oxidative stress in human lymphocytes [2] and its derivatives have been shown to have anti-cancer activity [3]. Octadecyl ferulate (ODF) is one of many FA derivatives that are found throughout the plant and fungi kingdoms. For instance, ODF can be found naturally in fungi such as Annulohypoxylon squamulosum [4], in plants such as potato tubers [5], and in herbs [6]. Despite the natural sources, ODF is not particularly abundant in nature and thus the need for de novo synthesis of ODF and other phenolic derivatives in order ☆ Mention of trade names or commercial products in this publication is solely for the purpose of providing specific information and does not imply recommendation or endorsement by the U.S. Department of Agriculture. USDA is an equal opportunity provider and employer. ⁎ Corresponding author. E-mail address: [email protected] (K.O. Evans). 1 Current address: Department of Chemistry, University of North Carolina — Chapel Hill, Chapel Hill, NC 27599, USA.

http://dx.doi.org/10.1016/j.saa.2015.08.009 1386-1425/Published by Elsevier B.V.

to assess their potential as cosmeceutical, nutraceutical, and pharmaceutical ingredients [3,7–15]. One of the features sought in these cosmeceutical, nutraceutical and pharmaceutical ingredients is their ability to retard lipid oxidation. Lipid peroxidation is damaging to cells because the radical products can propagate to polyunsaturated lipids creating a chain reaction within a cell. Nature counters lipid peroxidation through use of both enzymatic and non-enzymatic antioxidants [16]. Non-enzymatic antioxidants include a large variety of antioxidants that are either aqueous-based or lipophilic like ODF, which is readily produced by plants. We recently demonstrated the antioxidant capacity of the lipophilic feruloyl glyceride derivative, feruloyl dioleoylglycerol [17]. We further explore feruloyl derivatives for antioxidant capacity by demonstrating the antioxidant capacity of ODF. Specifically, we used a Trolox-based assay because Trolox can readily incorporate into liposomes [18]. Peroxyl radicals play a substantial role in peroxidation of polyunsaturated fatty acids of cells which leads to membrane structural loss [19]. The use of liposomes is important because they can readily serve as model cellular systems and can be used as delivery systems of active agents like antioxidants in formulations [20,21]. Thus, it is important to characterize the behavior of ODF within liposomes. In this work, we report the chemical synthesis of ODF from ferulic acid and octadecanol using a solid acid catalyst as monitored by Supercritical Fluid Chromatography (SFC), the phase transition temperature

334

K.O. Evans et al. / Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy 153 (2016) 333–343

of ODF and its behavior in a lipid matrix of phospholipid liposomes. We compare it to that of slightly less lipophilic ethyl ferulate (EF) in liposomes. 2. Materials and methods 2.1. Reagents and materials Ferulic acid (4-hydroxy-3-methoxy cinnamic acid), 1-octadecanol, p-toluenesulfonic acid monohydrate (p-TSA), anhydrous toluene, magnesium sulfate, cobalt (II) chloride hexahydrate, ethylenediaminetetraacetic acid disodium dehydrate (EDTA), 2,2′-azobis(2-amidinopropane) dihydrochloride (AAPH), 5,5′-dithiobis(2-nitrobenzoic acid) (DTNB), and all common solvents were purchased from Sigma-Aldrich. Ferulic acid ethyl ester (ethyl ferulate) was obtained from Sinova Corporation (Ningbo, China). Ethanol (200 proof) was purchased from Decon Laboratories, Inc. (King of Prussia, PA). Supercritical Fluid Chromatography (SFC)/supercritical fluid extraction (SFE) grade CO2 was purchased from Airgas Products Co. (Radnor Township, PA). 1,2-Dioleoyl-snglycero-3-phosphocholine (DOPC) was bought from Avanti Polar Lipids (Alabaster, AL). High purity calcein, 4,4-difluoro-5-(4phenyl-1,3-butadienyl)-4-bora-3a,4a-diaza-s-indacene-3-undecanoic acid (C11 -Bodipy ® 581/591), 1,6-diphenylhexatriene (DPH), and 1-(4-trimethylammoniumphenyl)-6-phenyl-1,3,5-hexatriene ptoluenesulfonate (TMA-DPH) were purchased from Invitrogen (Carlsbad, CA). Tris(hydroxymethyl)aminomethane hydrogen chloride (Tris–HCl), potassium phosphate monobasic, potassium phosphate dibasic, Sephadex G-75 column beads, columns and sodium chloride were obtained from Fisher Scientific. Octaethylene glycol monododecyl ether (C12E8) was purchased from Sigma-Aldrich. 2.2. Octadecyl ferulate (ODF) synthesis The procedure was adapted from Taniguchi et al. [22]. Ferulic acid (10.07 g, 520 mmol), octadecanol (21.10 g, 780 mmol), and p-TSA (1.05 g, 5.5 mmol) were combined in a 500-mL Schlenk flask under a dry, nitrogen atmosphere using standard Schlenk line techniques. Anhydrous toluene (250 mL) was added via syringe, and the Schlenk flask was fitted with a reflux condenser. The slurry was heated to 85 °C with stirring. Octadecanol and p-TSA dissolved after 24 h, but not the ferulic acid. The entire slurry became a clear, yellow solution after stirring and heating for an additional 30 h (54 h total). The solution was then cooled to ambient temperature for 18 h, resulting in the formation of an off-white precipitate (confirmed by SFC analysis to be unreacted ferulic acid). MgSO4 (1.00 g) was added and the solution stirred for 15 min. The solid was removed by filtration through a fine frit, and the solvent removed from the filtrate resulting in ~ 30.0 g of off-white solid. The crude product was purified in 10-g batches by flash chromatography. A 10-g fraction of the crude product was placed in a 1-L round bottom flask and dissolved in 500 mL of acetone. Silica gel (20 g, Silica Gel 60, 70–230 mesh ASTM) was added and the slurry thoroughly mixed. The acetone was then removed by rotor-evaporation. The product-loaded silica was slurried in hexane and poured into a Pyrex, 600-mL, glass frit (10–15 ASTM) fitted on top of a 1-L Erlenmeyer filtration flask. A vacuum was applied and the resultant 5-cm bed of productloaded silica was developed with 3.5 L of 2% acetone/hexane, 2.5 L of 5% acetone/hexane, 0.5 L of 10% acetone/hexane, and finally 0.5 L of 100% acetone, respectively. Fractions containing ODF (as monitored by SFC) were placed on ice and a white precipitate formed. The precipitates were collected by filtration (Whatman #2, 4.25 cm filter paper) and combined. The supernatants were combined and placed on ice, and a second precipitant formed. The second precipitant was collected by filtration and combined with the first, and the white precipitate was dried at 35 °C in vacuo.

Yield: 9.77 g, 42.2% (based on ferulic acid). Analytical calculated for C28H46O4: C, 75.29 and H, 10.38; found: C, 73.36 and H, 10.86. ESI-MS (m/z) calculated for C28H47O4 [M + H]+: 447.34; found: 447.34749. 1 H NMR (500 MHz, d6-acetone): δ 8.12 (bs, 0.60 H, \\OH), 7.60 (d, J = 16.0 Hz, 1.00 H, α), 7.35 (d, J = 1.90, 1.00 H, A2), 7.15 (dd, J = 8.12, 2.03, 1.99 Hz, 1.00 H, A6), 6.88 (d, J = 8.16 Hz, 0.98 H, A5), 6.40 (d, J = 15.90 Hz, 1.01 H, β), 4.16 (t, 2.04 H, 1), 3.93 (s, 2.99 H, A7), 1.63 (p, 2.07 H, 2), 1.37 (bm, 30.97 H, 3–17), and 0.89 (t, 3.15 H, 18) ppm. 13C NMR (125.77 MHz, d6-acetone): δ 166.6 (γ), 149.2 (A3), 147.9 (A4), 144.6 (α), 126.6 (A1), 123.0 (A6), 115.2 (A5), 115.1 (β), 110.4 (A2), 63.8 (1), 55.4 (A7), 31.7 (16), 29.4 (11 C, 5–15), 29.2 (4), 28.7 (2), 25.8 (3), 22.4 (17), and 13.4 (18) ppm (Fig. 1). 2.3. Supercritical fluid chromatography (SFC) The synthesis of ODF was monitored by SFC using a Selerity Technologies Series 4000 chromatograph. Samples (100 μL) taken from the reduction reactions were diluted in 1.0 mL of ethanol and 10-μL aliquots were used for injection. The samples were developed with CO2 on a Selerity Technologies SB-methyl-100 SFC column (PN AE002, 50 μm × 10 m × .25 μm film) starting at 100 atm and 100 °C. The starting pressure was held for 5 min and ramped at 15 atm/min to a final pressure of 310 atm, which was held for 11 min. The column temperature was isothermal for the duration of the analysis. The FID was set at 350 °C. Under these analytical SFC conditions octadecanol, ferulic acid, and ODF resulted in Rt of 11.96, 12.52, and 15.00 min, respectively. 2.4. NMR 1 H and 13C NMR spectra were obtained on a Bruker Avance 500 spectrometer (500 MHz 1H/125.77 MHz 13C) using a 5 mm BBI probe. All samples were dissolved in d6-acetone, and all spectra were acquired at 27 °C. Chemical shifts are reported as ppm from tetramethylsilane calculated from the lock signal (ΞD = 15.350609%). Noteworthy: The reaction as monitored by SFC showed nearly 85% conversion of ferulic acid to ODF, which belies the reported yield of 40%. Much of the product was lost in optimizing the flash chromatography and in the precipitations. The ODF is N 95% pure by SFC and 1H NMR.

2.5. Differential scanning calorimetry (DSC) The thermal properties of ODF were characterized using a Q2000 MDSC™ (TA Instruments; New Castle, DE). ODF (2–3 mg) was weighed into aluminum pans and hermetically sealed. The pans were heated from 10 °C to 100 °C at a heating rate of 1.0 °C/min, followed by cooling from 100° to 10 °C at the same rate (1.0 °C/min). The heating/cooling cycle was repeated twice per sample to determine if ODF exhibited any property changes. All measurements were conducted under a nitrogen atmosphere and run in quadruplicate. Universal Analysis 2000 software (TA Instruments) was used to conduct analysis of the thermographs. 2.6. Fluorescence measurements Fluorescence emission measurements were conducted using a Jobin-Yvon Horiba Fluorolog 3-21 spectrofluorometer (Edison, NJ) equipped with a 450-W xenon lamp and four-position cuvette holder. Constant temperature was maintained using a Thermo Neslab RTE-7 refrigerated water circulator. Fluorescence measurements were conducted in a 10 mm × 10 mm quartz cuvette (Starna Cells, Inc., Atascadero, CA) with constant stirring. All spectra were corrected and background subtracted using a solution containing only buffer or buffer/liposomes with the corresponding amount of ODF only. Excitation and emission slits were 5 and 10 nm, respectively (unless stated otherwise).

K.O. Evans et al. / Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy 153 (2016) 333–343

335

Fig. 1. Octadecyl ferulate structure.

2.7. Liposome preparation Liposomes were generated as previously described [14,17,21]. Briefly, appropriate amounts of DOPC and ODF in chloroform were added to 4-mL amber vials to give the following DOPC/ODF mole ratios: 100/0, 99/1 and 95/5. The vials were gently swirled to give a homogenous mix of lipids which were subsequently dried to a thin film under a slight stream of argon and placed under vacuum overnight to remove residual solvent. The lipid film was resuspended in the appropriate buffer and periodically mixed over 60 min. Afterwards, the newly formed multilamellar liposomes were subjected to freeze/thaw cycles using dry ice in 2-propanol and a 60 °C water bath, respectively. Multilamellar liposomes were then extruded 11-times through two stacked polycarbonate filters of the appropriate pore size using a LiposoFast extruder (AVESTIN, Inc., Ottawa, Canada) at ambient temperature. 2.8. Partitioning into DOPC liposomes measuring via fluorescence spectroscopy DOPC liposomes without either EF or ODF were extruded through 100-nm pore filters in 10 mM HEPES, 100 mM NaCl, and pH 7.4 buffer as described above. Resulting liposomes were diluted to the appropriated concentrations and incubated for an hour with 40 μM of either EF or ODF. Fluorescence measurement was repeated in triplicate; excitation was at 330 nm. Fitting Eq. (1) below at a fixed wavelength for the partitioning of EF or ODF into DOPC liposomes resulted in calculating the apparent partition coefficient KP, F ðLÞ ¼ F 0 þ F max

K p ½L ; ½W  þ K p ½L

ð1Þ

where F0 is the initial fluorescence signal of EF or ODF in water with no liposomes present [23]; Fmax is the maximal fluorescence signal when full partition has occurred; [L] and [W] are respectively, the molar concentration of lipid and the molar concentration of water (taken as 55.3 M); and Kp is the molar fraction partition coefficient. Alternatively, partitioning was determined using the n-octanol/water method based on the description given by Turina et al. [24]. Briefly, an aliquot of EF or ODF in chloroform was dried under argon in a vial and equal amounts of nanopure water and n-octanol (6 mL each) were added to give a final concentration of 40 μM of EF or ODF in a 1/1 n-octanol/water n-octanol. The mix was stirred and equilibrated overnight in the dark. The solution was centrifuged at 1000 ×g for 10 min to ensure phase separation. EF (or ODF) absorbance was measured for the water (Aw) and n-octanol (Ao) phases. Partitioning was determined by taking the logarithm of the ratio Ao/Aw at 327 nm (the peak of absorbance in n-octanol). 2.9. Liposome leakage (Calcein–Cobalt) Liposomes containing 0–5-mol% of either EF or ODF were monitored for membrane integrity via a leakage assay [17,25]. Briefly, the assay is designed such that liposomes entrapped calcein–cobalt complexes which exhibit a quenched fluorescent signal. Leakage in the membrane results in the complex passing through the bilayer where EDTA chelates the cobalt, resulting in increased fluorescence. The leakage buffer

(0.6 mM CoCl2, 0.25 mM calcein, 2.3 mM potassium phosphate, pH 7.4) was added to resuspend the dried lipid film. Liposomes were created as described above. Liposomes were passed down a G-75 Sephadex column equilibrated with 5 mM NaCl, 2.3 mM potassium phosphate 10 mM EDTA, and pH 7.4 (column buffer used to osmotically balance liposomes) to remove unencapsulated cobalt–calcein complex. Liposomes were maintained on ice in the dark until just prior to the measurement; this reduces any significant leakage. Immediately before making the measurement, liposomes were diluted in column buffer which had been preheated to 37 °C in a cuvette inside the fluorometer. Fluorescent intensity was measured at 520 nm (excitation at 490 nm) for 30 min in 5 s intervals. Leakage percentage (L), was calculated Ft − Fo Þ as L ¼ 100ð ð F max − F o Þ with F t being the fluorescent signal over time, F o

the extrapolated initial signal at time zero, and Fmax the fluorescent signal at the end of the experiment after rupture by addition of C12E8 solution. Regression analysis within Sigma Plot software was used to fit the leakage data to the general equation L ¼ ∑Ln ð1−e−kn t Þ where n was 1 or 2 for single or double-exponential, respectively. 2.10. Atomic force microscope (AFM) AFM images were taken using a Nanoscope IV controller (Bruker Nano Surface, Santa Barbara, CA) in tapping mode via a fluid cell. The fluid cell was cleaned via sonication in 1/1 (v/v) mix of methanol and water for 5 min, rinsed thoroughly and dried under a stream of argon. Image scan rate was 1 Hz. Fresh cleaved mica was exposed for 15 min to liposomes that were extruded at 50-nm (it has been our experience that larger liposomes can adsorb unfused and lead to misinterpretation of the image). The fluid cell was then rinsed with buffer. The buffer contained 2 mM CaCl2, 10 mM HEPES, 100 mM NaCl, and pH 7.4. 2.11. Phase behavior monitored with DPH and TMA–DPH Liposomes composed of 1,2-dielaidoylphosphocholine (DEPC) were made as described earlier (DEPC was chosen because it has the same diacyl chain lipid as DOPC but the double bond is 18:1, Δ9 trans versus 18:1, Δ9 cis, allowing us to measure the phase transition temperature above 0 °C). DPH or TMA–DPH at 1 μM was added to liposomes at 100 μM to maintain a 100:1 lipid:probe ratio. Liposomes were incubated for an hour to allow maximum incorporation of probe. Excitation and emission were at 350 nm and 428 nm, respectively. Liposomes were equilibrated for 10 min at each temperature prior to measurement. The steady-state anisotropy, r, was monitored as a function of temperature to determine phase behavior of DEPC liposomes containing ODF at various concentrations. Anisotropy (r) was determined as follows:

r¼

ðIVV −G  IVH Þ IHV ; where G ¼ ðIVV þ 2G  IVH Þ IHH

ð2Þ

and VV, VH, HV, and HH denote the excitation/emission polarizers in the vertical/vertical, vertical/horizontal, horizontal/vertical and horizontal/ horizontal positions, respectively; G is the correction factor for polarization sensitivity of the detector.

336

K.O. Evans et al. / Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy 153 (2016) 333–343

2.12. Lipid peroxidation EF and ODF abilities to inhibit lipid peroxidation were measured using an assay incorporating the aqueous free-radical initiator AAPH and the oxidation-sensitive reporter lipid C11-Bodipy [26]. C11-Bodipy, which resides well within the bilayer because of lipophilicity, shifts its fluorescence from 595 nm to nearly 520 nm when it is oxidized [27]. It is assumed that the affinity for generated radicals is the same for C11-Bodipy and phospholipids in the bilayer; thus, it is believed that C11-Bodipy reports the oxidation of all lipids in the bilayer. Sample conditions used were similar as detailed previously [17]. Briefly, DOPC liposomes containing 36 μM C11-Bodipy and either EF, ODF (each at 0-, 1- or 5-mol% of total lipids), or Trolox were generated as previously described above [17,25,28]. Liposomes without C11-Bodipy were used as background subtraction. Samples were diluted to a final concentration of 2.5 mM DOPC (2.4 μM C11-Bodipy) and equilibrated to 37 °C for 15 min prior to the introduction of AAPH. Fluorescence intensity (excitation at 540 nm, emission at 595 nm) was measured every minute for 180 min from the moment of AAPH addition at a final concentration of 40 mM. Experiments were conducted in triplicate at least. The relative fluorescence intensity was calculated by dividing the reading at each minute by the initial fluorescence at the beginning of the experiment (Ft/Fo). Background subtraction was accomplished using the appropriate liposomes without C11-Bodipy. The lipid peroxidation inhibition ability of ODF and EF were compared to Trolox. Trolox, being lipophilic and somewhat water-soluble [18], was incorporated along with C11-Bodipy within DOPC liposomes in the same manner as EF or ODF, but in the final range of 10 to 125 μM. The peroxyl radical scavenging capacity (i.e., the relative antioxidation activity) was calculated accordingly: Relative‐Antioxidant‐CapacityðRACÞ ¼

ΔAUCsample ½sample = ; ΔAUCTrolox ½Trolox

ð3Þ

where ΔAUC is the difference in the area under the fluorescence curve (AUC) of liposomes containing antioxidant or standard curve minus empty liposomes [29]. The AUC was calculated as described by Cao and et al. [30]. 2.13. Bilayer location using nitroxide-labeled lipids The location of the feruloyl group of EF or ODF within the phospholipid bilayer of liposomes was determined using the parallax analysis [31]. This is based on the expectation that ODF fluorescence intensity is reduced by collisional quenching with nitroxide-labeled lipids as described by Kachel et al. [32]. The nitroxide-labeled lipids used had the nitroxide quencher positioned either in the headgroup region (TEMPO-PC), just below the headgroup region (5DOXYL-PC), or farther down the acyl chain (12DOXYL-PC). It is expected that the most quenching would occur from the nitroxide group closest to the feruloyl group. Nitroxide-labeled lipids were incorporated at 15 mol% into the membrane in conjunction with ODF at various concentrations. Measurements were conducted in potassium phosphate buffer (PPB) at pH 7.4. Fluorescence emission spectra were recorded using an excitation wavelength of 330 nm and band pass slits of 5 nm by 5 nm with the appropriate background subtracted. The location of the feruloyl group was determined (Chattopadhyay and London, 1987; Kondo et al., 2008) as follows: h i zc f ¼ Lcl þ − ln ð F 1 =F 2 Þ=πC−L221 =2L21 :

ð4Þ

Fig. 2. Representative DSC scan of ODF from 10 to 100 °C at a 10 °C/min scan rate. Sample replicates were scanned through two cycles of heating and cooling, monitoring for property changes. Positive heat flow values represent exothermic transitions.

(12DOXYL-PC), Lcl is the distance between the “shallow” quencher and the center of the bilayer, C is the concentration of quenchers per area (mole fraction of nitroxide-labeled lipids/area of lipid), and L21 is the difference in depth between quenchers. The area of the phosphatidylcholine lipids was taken to be 70.1 Å2 per molecule [33] and the distances from the bilayer center for the nitroxide group were taken from Kachel et al. where TEMPO-PC was located at 19.5 Å, 5DOXYL-PC at 12.2 Å, and 12DOXYL-PC at 5.9 Å [32]. 2.14. Molecular modeling Molecular models were constructed using Hyperchem v8.0.10 (Hypercube Inc. Gainesville, Florida). Models of EF, ODF, and DOPC were initially optimized in vacuo using Charmm27 molecular mechanics force field and the PM3 semi-empirical method with Polak–Rebiere conjugate gradient method and the convergence criteria of 0.01 kcal mol−1. The conformational preferences of ODF were studied by geometry optimization using the B3LYP three parameter density functional at the 6–31 + G* level as implemented in Spartan'10. Areas and volumes were calculated using the QSAR properties module in Hyperchem on B3LYP/6–31 + G* optimized geometries. 3. Results and discussion 3.1. Thermal properties of ODF The thermal properties of ODF were evaluated over a 10 to 100 °C temperature range employing DSC. Fig. 2 exhibits a representative heating and cooling thermograph scan of ODF. DSC measurements determined that ODF has a melting phase transition at 67 °C. This is approximately 9–18° above the range determined for other synthetic octadecyl esters [34–37] and lyso-stearoylphosphatidylcholine [38], indicating that the feruloyl moiety contributes significantly to the thermotropic characteristics. Table 1 Thermal properties of ODF measured by DSC at 10 °C/min. Melting

where zcf is the distance the fluorophores are from the center of the membrane, F1 is the normalized emission fluorescence intensity (normalization done with respect to liposomes with EF or ODF only) in the presence of the “shallow” quencher (TEMPO-PC), F2 is the normalized fluorescence intensity in the presence of the “deep” quencher

Solidification

Onset temp (°C)

Peak temp (°C)

Latent heat (J/g)

Onset temp (°C)

Peak temp (°C)

Latent heat (J/g)

62.9 ± 0.5

66.7 ± 0.5

127.5 ± 1.7

38.9 ± 0.3

38.5 ± 0.3

100.2 ± 8.7

Reported results are the average of 4 measurements of 2 heating/cooling cycles per measurement.

K.O. Evans et al. / Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy 153 (2016) 333–343

337

The solidification transition was determined to occur at 38.5 °C. The onset temperatures for both melting and solidification phase transitions were roughly 63 and 39 °C, respectively (Table 1). The latent heat was determined to be 128 and 100 J/g for the melting and solidification phase transitions, respectively. Both phase transitions exhibited sharp peaks which indicate pure material. A pre-transition phase occurred around 39 °C during the second cycle (data not shown), which may indicate formation of an ordered, tilted gel state [39]. However, the pre-transition phase was not reproducible and thus may not be thermodynamically stable for ODF. It should be noted that a tilted gel-state was readily found in the thermotropic behavior of C16 and C18 saturated diacylphospholipids like dipalmitoylphosphocholine (C16) [40] and distearoylphosphocholine (C18) [41], but not found for C18 monoacyl ester lipids [34,37] or a C18-lysophospholipid [38]. The possible presence of a pre-transition state further indicates the influence of the feruloyl moiety on the packing behavior of ODF. 3.2. AFM The data from the thermal properties of ODF indicates that ODF would form a gel state within a phospholipid bilayer at temperatures below 63 °C. AFM imaging was undertaken to explore the possible presence of gel state of 5-mol% ODF within a DOPC supported bilayer. Fig. 3a confirms that a supported bilayer formed on mica with two areas of two different materials (demonstrated by phase behavior imaging; not shown). One area was dark and the other one was light. The dark area was consistent with a uniform bilayer formed on the mica (not shown) and thus classified. The light regions, or domains (Fig. 3a) extended at least 2 nm above the uniform bilayer (based on bearing analysis, not shown) and as much as 4 nm (demonstrated by typical height profile in Fig. 3b). Fig. 3c shows that the domains are mobile in that several of them coalesced into larger units or became more spherical (see symbols in Fig. 3c). Thus, the light domains were interpreted as the result of gel state formations with the bilayer that were likely ODF-enriched. 3.3. Phase behavior Further investigation of the effect that ODF has on liposomes was done by monitoring the phase behavior of liposomes composed of DEPC which has a phase transition temperature of 12 °C [42]. The presence of domains as revealed by AFM is expected to raise the phase transition (Tm) of DEPC. Fig. 4 depicts the phase transition of the acyl chain region (as determined by DPH anisotropy) and the phase transition of the headgroup regions (as determined by TMA– DPH anisotropy) of DEPC liposomes as a function of ODF concentration. It is clear that the presence of ODF steadily increased the phase transition temperature of DEPC liposomes in both acyl chains and headgroup regions, further suggesting that ODF caused a more “gel” state within the liposomes. 3.4. Log P — Partition coefficient ODF has potential use in cosmeceutical, nutraceutical, and pharmaceutical ingredients. This means that ODF may have to be formulated within a lipid matrix as lipid matrices are regularly used to deliver water insoluble ingredients [43,44]. Then the question becomes, how well does ODF associate with a lipid matrix? The answer to this question was pursued using DOPC liposomes as the chosen lipid matrix (because of the abundance of DOPC and the simplicity of its liposomes) to explore the partitioning of ODF. It was previously reported that molecular Fig. 3. Topographical AFM images of a supported bilayer of DOPC containing 5-mol% ODF showing (a) that domains formed and (b) a height profile across two of the regions (solid black line); (c) domains shown 3 h, 10 min, and 20 s later where some nearest neighbors have migrated together (star). Dimensions for both images are 5 μm × 5 μm.

338

K.O. Evans et al. / Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy 153 (2016) 333–343

Fig. 4. Phase behavior temperature (Tm) as a function of ODF mole percent within DEPC liposomes. The circles represent the Tm of the acyl chains as measured by the anisotropy of DPH; the inverted triangles represent the Tm of the headgroups as measured by TMA–DPH.

partitioning into the membrane can be determined by monitoring the fluorescence signal of a molecule as it is titrated with liposomes [45]. The determination of a partitioning coefficient for ODF and EF was attempted using this approach, but reliable measurements were possible only for EF (Supplemental Data). As an alternative approach, the affinity of EF and ODF for a 1/1 n-octanol/water mixture was determined and compared. Table 2 shows the values of n-octanol/water partitioning coefficient indicating that EF and ODF partitioned into n-octanol similarly with ODF partitioning slightly more. This correlated well with the difference in hydrophobicity expressed by the difference in longer acyl chain length [39].

3.5. Liposome stability It was shown earlier in this work that the thermotropic properties of ODF have a phase transition temperature at approximately 67 °C, indicating that ODF would exist in a gel state at 37 °C. AFM measurements and phase behavior demonstrate a more “gel-like” state with ODF present within liposomes. It is thus important to understand how ODF affects liposomal stability at pharmaceutical/nutraceutical/cosmeceutical concentrations. EF and ODF effects on liposome stability were studied using the calcein–cobalt content leakage assay described previously [17,25]. Fig. 5a and b displays the results of the content leakage studies EF and ODF, respectively. It is clear that EF and ODF had distinctively different effects on the leakage of DOPC liposomes. EF (Fig. 5a) at both concentrations explored displayed a two-step leakage process where there a rapid release of entrapped contents (fast component of leakage) followed by a slow release (slow component), whereas liposomes empty of EF displayed a single-step leakage process. More specifically, the fast component of leakage induced by EF (not shown) was comparable to the single-exponential leakage observed for FDOG [17] and lipoyl dioleoylglycerol [28] lipids in DOPC liposomes. The slow

Table 2 Log P partitioning determine via experiments. Log Kp

EF ODF a

a

b

2.45 ± 0.12 2.71 ± 0.21

n-octanol/water

Determined using 1/1 n-octanol/water mixture. Calculated from least-squares fit of experimental data for partitioning between liposomes and aqueous buffer. b

component, on the other hand, dominated the leakage process and increased in percentage as a function of EF concentration (Table 3). This suggests that lipids with EF present retard the leakage rate as the extent of leakage approaches that of the control liposomes. A more straight-forward analysis was possible using the initial rates or “velocities” of leakage at the start (t = 0 s) of the experiment (Table 3). Leakage initial rates showed that overall leakage induced by 1- and 5-mol% EF was twice and 1.5, respectively, as fast as control liposomes (0-mol% EF). The indication was that despite any lipid movement

Table 3 Leakage properties of DOPC liposomes with ethyl and octadecyl ferulates incorporated. Ferulate

Amount (mol%)

Initial ratesa, b, c (%*s−1 × 10−2)

Slow component percentd

EF

0 1 5 1 5

2.96 (±0.33) 6.22 (±0.94) 4.33 (±0.52) 1.72 (±0.72) 2.18 (±0.83)

– 87.7 (±1.8) 94.3 (±0.6) – –

ODF

Liposomes 4.85 ± 0.13 –

Fig. 5. (a) Content leakage as a function of time for EF incorporated at 0-, 1-, and 5-mol% in DOPC liposomes. (b) ODF effect on DOPC liposome content leakage.

a Leakage initial rates reported as the mean and standard deviation of determinations conducted in 3–12 replicates. b Statistical analysis was done using one-way ANOVA where each was statistically different (P b 0.05) from the control sample (zero percent antioxidant). c L was obtained by fitting data to the equation: L ¼ ∑Ln ð1−e−kn t Þ; where n ¼ 1 or 2. d 2 Fraction was determined by L1L¼L  100. 2

K.O. Evans et al. / Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy 153 (2016) 333–343 Table 4 Relative antioxidant capacity of ethyl and octadecyl ferulates in DOPC liposomes. Antioxidant

Concentration (μM)

Relative antioxidant capacitya

EF

25 62.5 125 25 62.5 125

1.70 ± 0.07 1.43 ± 0.02 1.40 ± 0.02 0.89 ± 0.04 0.65 ± 0.02 0.53 ± 0.01

ODF

a Data is reported as the mean and standard deviation of values calculated. Experiments were conducted in triplicate at least. Values are relative to 50 μM Trolox. One-way ANOVA was applied to the data. All are considered to be statistically different (P b 0.05).

within the bilayer, the leakage profile in each case was quite similar (as borne out by the traces overlapping and reaching similar %leakage values). Thus, it was interpreted that EF did not induce any bilayer changes that resulted in significant bilayer destabilization.

339

ODF-loaded liposomes displayed a single-exponential leakage profile. Statistical analysis indicated that the initial rates, though similar in magnitude, were significantly different and less that either the control sample or the EF-loaded liposomes. The indication here was that leakage is slower in ODF-loaded liposomes than non-loaded (control) or EF-loaded liposomes. Additionally, each leakage profile for ODF-loaded liposomes resulted in nearly one-third of the percent leakage exhibited by unloaded liposomes (control). This suggests that ODF stabilized the liposomes against leakage. Based on the facts that the liposomes were in a fluid state under the current experimental conditions at 37 °C (DOPC has a phase transition temperature at −20 °C), ODF lipids have phase transition temperature of 67 °C, and the neutrally charged lipids had a minimum electrostatic interaction with negatively charged calcein [46], it is interpreted that content leakage correlated with the fluidity of the membrane. Therefore, it stands to reason to conclude that ODF reduced membrane fluidity by altering the membrane packing density. This is consistent with an earlier report that a similarly structured lyso-lipid altered the fluidity of phosphatidylcholine membranes

Fig. 6. Antioxidation behavior of (a) ethyl ferulate — EF and (b) octadecyl ferulate — ODF in DOPC liposomes.

340

K.O. Evans et al. / Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy 153 (2016) 333–343

Fig. 6 (continued).

[47] and consistent with the thermotropic behavior described earlier in this work. The ability of ODF to reduce membrane fluidity is, however, in contrast to the effect exhibited by FDOG on the same lipid system [17] where FDOG had no effect on liposome stability (leakage remained unchanged). The liposomal leakage exhibited by ODF can be explained by the structural of ODF. ODF has a feruloyl moiety esterified to a single octadecyl group, which resulted in ODF being a solid at room temperature and up to 67 °C. This suggests that ODF would form a more rigid region within the bilayer, presenting a barrier to local content leakage unlike FDOG which was a “fluid state” lipid (the result transesterification of ethyl ferulate to triolein in the sn−1 or sn−3 position) and did not affect leakage [17]. Thus, ODF can limit permeability due to increased local packing stress in the membrane, unlike local structural changes that resulted in greater permeability of 1-palmitoyl-3-oleoyl phosphocholine liposomes under shear flow [48].

3.6. Lipid peroxidation inhibition Ferulic acid has well-known antioxidant characteristics [49]. It is unclear, though, about the extent of antioxidant properties for ODF, especially within phosphatidylcholine liposomes. Therefore, the antioxidant properties of ODF in DOPC liposomes were tested and compared to EF, a smaller feruloyl derivative (Table 4). The method utilized AAPH, a free-radical generator and C11-Bodipy as the radical identifier. AAPH thermally decomposed into two peroxyl radicals which, upon interaction with C11-Bodipy molecules, caused a spectral shift from 595 nm. This spectral shift was seen as a fluorescence signal reduction when detected at 595 nm. Experiments were conducted at roughly a 1042 to 1 lipid-to-probe molar ratio. DOPC liposomes incorporated with EF demonstrated significant antioxidant behavior, as displayed by the lag plateau present in all concentrations explored (Fig. 6a, top). Clearly, there was linearity between

K.O. Evans et al. / Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy 153 (2016) 333–343

341

Fig. 7. B3LYP/6–31 + G* conformations of ODF, EF, and FDOG rendered in space filling overlapping spheres.

ΔAUC and EF concentration (Fig. 6a, bottom). Linearity between ΔAUC and concentration was also present for ODF (Fig. 6b), demonstrating that ODF maintain antioxidant capability. It is noteworthy to mention that ODF did not exhibit any lag plateau and the ΔAUC values for ODF were 2–3 times less than values for EF at corresponding concentration. This suggests that ODF was less effective as an antioxidant in liposomes than EF. However, the relationship between ΔAUC and antioxidant concentration may rely on multiple factors and thus make interpretation complex. It therefore may be simpler to normalize the ΔAUC with respect to Trolox, the standard used. Normalization of the ΔAUC values based on one concentration of the standard gave an antioxidant capacity number that was relative to Trolox. Table 4 summarizes that although there was a slight decrease in relative antioxidant capacity with concentration for EF and ODF, ODF maintained antioxidant capacity but was less than Trolox and EF. It should be noted that ODF proved to have nearly the same antioxidant capacity as FDOG. It is believed that this similarity in relative antioxidation capacity was primarily due to the fact that both have 18-carbon acyl chains and less due to the presence of multiple acyl chains and chain saturation/unsaturation. 3.7. Modeling Molecular modeling was undertaken to better understand the structural nature of ODF and how its intrinsic curvature affects partitioning into a DOPC membrane. Intrinsic curvature properties calculated on single molecules have been shown to be related to partitioning and micelle/bilayer formation [49–52]. ODF molecular model was optimized using SPARTAN'10 and the B3LYP density functional and the 6–31 + G* basis set. Molecular modeling shows that energetically preferred ODF confirmation has the feruloyl moiety angled with respect to the acyl chain tail (Fig. 7). This resulted in a positive headgroup-to-acyl chain area ratio (Table 5), suggesting that ODF has a conical shape or positive intrinsic curvature [50]. This positive intrinsic curvature indicates that Table 5 Parameters from B3LYP/6–31 + G* calculations on three conformations of ODF.

ΔE (kcal mol−1) Dipole (Debye) εHOMO (eV) εLUMO (eV) Band gap (eV) Area (Å2) Volume (Å3) Areaheadgroup/Areaacyl chain

ODF 1

ODF 2

ODF 3

EF

FDOG

0 1.55 −5.93 −1.85 −4.08 980.9 1602.8 1.42

1.29 5.55 −5.96 −1.86 −4.10 932.3 1581.6 2.16

8.28 3.76 −6.05 −2.03 −4.02 968.9 1578.3 1.99

0 1.76 −5.94 −1.87 −4.07 453.7 710.8 1.80

0 1.87 −6.23 −2.06 −4.17 1389.9 2587.5 0.33

ODF has tendency to form a spherical micelle and partition into membranes [51]. In contrast, the EF has a much shorter acyl chain, smaller molecular volume and smaller headgroup-to-acyl chain area ratio than ODF. Despite this, the similar intrinsic curvature of EF still suggests the formation of possible micelles. However, structurally EF may be better described as a hydrotrope [52] since EF as an aromatic ring attached to a short chain. Thus, described as a hydrotrope EF likely will form sphere-like aggregates due to a conical molecular shape. This similarity in molecular shape explains the resemblance in portioning coefficients of ODF and EF. In fact, the similarities between aggregates formed by short-chained hydrotropes, such as EF, and more traditional long tail surfactants similar to ODF have been extensively investigated [53]. FDOG, on the other hand, has a headgroup-to-acyl chain area ratio less than one which gives FDOG a cylindrical shape with negative intrinsic curvature and thus likely forming bilayers [51] and expected to have a much higher (presently immeasurable) partition coefficient. Surprisingly, ODF, EF and FDOG have similar frontier molecular orbitals (εHOMO and εLUMO) and band gap energies despite their significant differences in acyl chain sizes and volumes, suggesting that the feruloyl moiety was the main contributor to orbitals. Furthermore, modeling shows ODF, EF and FDOG to possess similar chemical reactivity and electron accepting/donating properties to free ferulic acid [54], leading to the conclusion that interaction within the lipid bilayer exerts the greatest influence on antioxidant activity. 3.8. ODF Position in DOPC bilayer ODF at 5-mol% was incorporated into DOPC liposomes to determine membrane depth of the feruloyl group using nitroxide labeled phospholipids as fluorescent quenchers. ODF presented a strong fluorescent signal within DOPC liposomes which was quenched the most by 12DOXYL-PC, the deepest quencher employed (data not shown). Table 6 summarizes the results of EF and ODF locations within the DOPC bilayer. The finding was the feruloyl group of EF and ODF were found on average approximately 12 Å from the bilayer center, which is right at or above the carbonyl groups of the phospholipids that are at the lower limit of bilayer hydrophilic region, just slightly deeper than the feruloyl moiety for FDOG [17]. Table 6 Approximate location of feruloyl moiety in phosphocholine bilayer. Sample

Approximate distance from the bilayer center, zcF (Å)

EF ODF

12.38 12.15

342

K.O. Evans et al. / Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy 153 (2016) 333–343

4. Conclusions In summary, ODF was synthesized and investigated for its thermotropic properties and behavior within DOPC liposomes using differential scanning calorimetry, fluorescence methods and molecular modeling. It was ascertained that ODF has melting and solidification temperatures of 67 and 39, respectively. It was demonstrated that DOPC supported bilayer containing 5-mol% ODF formed domains on mica and the presence of ODF increased liposomal phase transition temperature. Studies of ODF behavior show readily partitioning into n-octanol and the ability ODF to induce an increase in lipid packing density. Further studies revealed that ODF retained antioxidation behavior and resided within a DOPC similar to FDOG. We expect this work to deepen the basic understanding of increasing the lipophilicity of phenolic compounds. Supplementary data to this article can be found online at http://dx. doi.org/10.1016/j.saa.2015.08.009.

Acknowledgments We graciously thank Leslie Smith, Judy Blackburn, Ray Holloway and Jason Adkins for their technical assistance.

References [1] F. Di Domenico, C. Foppoli, R. Coccia, M. Perluigi, Antioxidants in cervical cancer: chemopreventive and chemotherapeutic effects of polyphenols, Biochim. Biophys. Acta (BBA) — Mol. Basis Dis. 1822 (2012) 737–747. [2] N.R. Prasad, S. Ramachandran, K.V. Pugalendi, V.P. Menon, Ferulic acid inhibits UVB-induced oxidative stress in human lymphocytes, Nutr. Res. 27 (2007) 559–564. [3] W. Li, N. Li, Y. Tang, B. Li, L. Liu, X. Zhang, H. Fu, J.-A. Duan, Biological activity evaluation and structure–activity relationships analysis of ferulic acid and caffeic acid derivatives for anticancer, Bioorg. Med. Chem. Lett. 22 (2012) 6085–6088. [4] M.-J. Cheng, M.-D. Wu, I.-S. Chen, S.-Y. Hsieh, G.-F. Yuan, Chemical constituents from the endophytic fungus Annulohypoxylon squamulosum, Chem. Nat. Compd. 48 (2012) 218–220. [5] M.A. Bernards, N.G. Lewis, Alkyl ferulates in wound healing potato tubers, Phytochemistry 31 (1992) 3409–3412. [6] B.T. Ngadjui, E. Dongo, E.N. Happi, M.-T. Bezabih, B.M. Abegaz, Prenylated flavones and phenylpropanoid derivatives from roots of Dorstenia psilurus, Phytochemistry 48 (1998) 733–737. [7] D.L. Compton, J.A. Laszlo, Sunscreens from vegetable oil and plant phenols, The United States of America as Represented by the Secretary of Agriculture (Washington, DC) USA, 2002. [8] L.J. DeFilippi, S.G. Grall, J.F. Kinney, J.A. Laszlo, D.L. Compton, Formulations With Feruloyl Glycerides and Methods of Preparation, Biotech Research and Development Corporation, (Peoria, IL) USA, 2010. [9] J.A. Laszlo, D.L. Compton, L.J. DeFilippi, S. Grall, Methods of Making Compositions Comprising a UV-absorbing Chromophore Biotechnology Research & Development Corporation (Peoria, IL), The United States of America as represented by the Department of Agriculture (Washington, DC), USA, 2009. [10] J.A. Laszlo, D.L. Compton, L.J. DeFilippi, S. Grall, Compositions Comprising a UVabsorbing Chromophore Biotechnology Research & Development Corporation (Peoria, IL), United States of America Department of Agriculture, (Washington, DC) USA, 2010. [11] J.A. Laszlo, D.L. Compton, L.J. DeFilippi, S. Grall, Compositions Comprising a UVabsorbing Chromophore. Biotechnology Research & Development Corporation (Peoria, IL), United States Department of Agriculture, Peoria, IL, 2013. [12] L. Panzella, L. Verotta, L. Goya, S. Ramos, M.A. Martín, L. Bravo, A. Napolitano, M. d'Ischia, Synthesis and bioactivity profile of 5-s-lipoylhydroxytyrosol-based multidefense antioxidants with a sizeable (poly)sulfide chain, J. Agric. Food Chem. 2012 (1710–1717) 61. [13] M. Kalinowska, J. Piekut, A. Bruss, C. Follet, J. Sienkiewicz-Gromiuk, R. Świsłocka, Z. Rzączyńska, W. Lewandowski, Spectroscopic (FT-IR, FT-Raman, 1H, 13C NMR, UV/VIS), thermogravimetric and antimicrobial studies of Ca(II), Mn(II), Cu(II), Zn(II) and Cd(II) complexes of ferulic acid, Spectrochim. Acta A Mol. Biomol. Spectrosc. 122 (2014) 631–638. [14] S. Zhang, X. Sun, R. Kong, M. Xu, Studies on the interaction of apigenin with calf thymus DNA by spectroscopic methods, Spectrochim. Acta A Mol. Biomol. Spectrosc. 136 (Part C) (2015) 1666–1670. [15] S. Zhang, X. Sun, F. Qu, R. Kong, Molecular spectroscopic studies on the interaction of ferulic acid with calf thymus DNA, Spectrochim. Acta A Mol. Biomol. Spectrosc. 112 (2013) 78–83. [16] J.-M. Lü, P.H. Lin, Q. Yao, C. Chen, Chemical and molecular mechanisms of antioxidants: experimental approaches and model systems, J. Cell. Mol. Med. 14 (2010) 840–860. [17] J.A. Laszlo, K.O. Evans, K.E. Vermillion, M. Appell, Feruloyl dioleoylglycerol antioxidant capacity in phospholipid vesicles, J. Agric. Food Chem. 58 (2010) 5842–5850.

[18] J. Zhang, R.A. Stanley, L.D. Melton, Lipid peroxidation inhibition capacity assay for antioxidants based on liposomal membranes, Mol. Nutr. Food Res. 50 (2006). [19] M. Comporti, Three models of free radical-induced cell injury, Chem. Biol. Interact. 72 (1989) 1–56. [20] M. Chorilli, G. Calixto, T.C. Rimério, M.V. Scarpa, Caffeine encapsulated in small unilamellar liposomes: characterization and in vitro release profile, J. Dispers. Sci. Technol. 34 (2012) 1465–1470. [21] D. Krilov, M. Kosović, K. Serec, Spectroscopic studies of alpha tocopherol interaction with a model liposome and its influence on oxidation dynamics, Spectrochim. Acta A Mol. Biomol. Spectrosc. 129 (2014) 588–593. [22] H. Taniguchi, E. Nomura, T. Tsuno, S. Minami, Ferulic Acid Ester Antioxidant/UV Absorbent, 1999. [23] M. Mohapatra, A.K. Mishra, Photophysical behavior of fisetin in dimyristoylphosphatidylcholine liposome membrane, J. Phys. Chem. B 115 (2011) 9962–9970. [24] A.D.V. Turina, M.V. Nolan, J.A. Zygadlo, M.A. Perillo, Natural terpenes: self-assembly and membrane partitioning, Biophys. Chem. 122 (2006) 101–113. [25] K.O. Evans, Room-temperature ionic liquid cations act as short-chain surfactants and disintegrate a phospholipid bilayer, Colloids Surf. A Physicochem. Eng. Asp. 274 (2006) 11–17. [26] J. Zhang, R.A. Stanley, L.D. Melton, Lipid peroxidation inhibition capacity assay for antioxidants based on liposomal membranes, Mol. Nutr. Food Res. 50 (2006) 714–724. [27] G.P.C. Drummen, L.C.M. van Liebergen, J.A.F. Op den Kamp, J.A. Post, C11BODIPY581/591, an oxidation-sensitive fluorescent lipid peroxidation probe: (micro)spectroscopic characterization and validation of methodology, Free Radic. Biol. Med. 33 (2002) 473–490. [28] J.A. Laszlo, K.O. Evans, D.L. Compton, M. Appell, Dihydrolipoyl dioleoylglycerol antioxidant capacity in phospholipid vesicles, Chem. Phys. Lipids 165 (2012) 160–168. [29] Y.M.A. Naguib, A fluorometric method for measurement of peroxyl radical scavenging activities of lipophilic antioxidants, Anal. Biochem. 265 (1998) 290–298. [30] G. Cao, H.M. Alessio, R.G. Cutler, Oxygen-radical absorbance capacity assay for antioxidants, Free Radic. Biol. Med. 14 (1993) 303–311. [31] F.S. Abrams, A. Chattopadhyay, E. London, Determination of the location of fluorescent probes attached to fatty acids using parallax analysis of fluorescence quenching: effect of carboxyl ionization state and environment on depth, Biochemistry 31 (1992) 5322–5327. [32] K. Kachel, E. Asuncion-Punzalan, E. London, The location of fluorescence probes with charged groups in model membranes, Biochim. Biophys. Acta 1374 (1998) 63–76. [33] B.A. Lewis, D.M. Engelman, Lipid bilayer thickness varies linearly with acyl chain length in fluid phosphatidylcholine vesicles, J. Mol. Biol. 166 (1983) 211–217. [34] J.A. Kenar, Thermal characteristics of oleochemical carbonate binary mixtures for potential latent heat storage, Eur. J. Lipid Sci. Technol. 114 (2012) 63–70. [35] S.T. Reddy, P.K. Tarafdar, R.K. Kamlekar, M.J. Swamy, Structure and thermotropic phase behavior of a homologous series of bioactive n-acyldopamines, J. Phys. Chem. B 117 (2013) 8747–8757. [36] P.K. Tarafdar, S.T. Reddy, M.J. Swamy, Effect of Hofmeister series anions on the thermotropic phase behavior of bioactive o-acylcholines, J. Phys. Chem. B 117 (2013) 9900–9909. [37] J. Watanabe, B.R. Harkness, M. Sone, H. Ichimura, Rigid-rod polyesters with flexible side chains. 4. Thermotropic behavior and phase structures in polyesters based on 1,4-dialkyl esters of pyromellitic acid and 4,4′-biphenol, Macromolecules 27 (1994) 507–512. [38] J.L. Slater, C.H. Huang, R.G. Adams, I.W. Levin, Polymorphic phase behavior of lysophosphatidylethanolamine dispersions. A thermodynamic and spectroscopic characterization, Biophys. J. 56 (1989) 243–252. [39] D.A. Small, Handbook of lipid research, The Physical Chemistry of Lipids, Plenum Press, New York 1986, pp. 43–88. [40] T. Schaffran, J. Li, G. Karlsson, K. Edwards, M. Winterhalter, D. Gabel, Interaction of N, N, N-trialkylammonioundecahydro-closo-dodecaborates with dipalmitoyl phosphatidylcholine liposomes, Chem. Phys. Lipids 163 (2010) 64–73. [41] K. Takeda, H. Okuno, T. Hata, M. Nishimoto, H. Matsuki, S. Kaneshina, Interdigitation and vesicle-to-micelle transformation induced by a local anesthetic tetracaine in phospholipids bilayers, Colloids Surf. B: Biointerfaces 72 (2009) 135–140. [42] E.A. Smith, C. Smith, B. Tanksley, P.K. Dea, Effects of cis- and trans-unsaturated lipids on an interdigitated membrane, Biophys. Chem. 190–191 (2014) 1–7. [43] W.T. Al-Jamal, K. Kostarelos, Liposomes: from a clinically established drug delivery system to a nanoparticle platform for theranostic nanomedicine, Acc. Chem. Res. 44 (2011) 1094–1104. [44] G.M.E. Maghraby, B.W. Barry, A.C. Williams, Liposomes and skin: from drug delivery to model membranes, Eur. J. Pharm. Sci. 34 (2008) 203–222. [45] A. Kyrychenko, F. Wu, R.P. Thummel, J. Waluk, A.S. Ladokhin, Partitioning and localization of environment-sensitive 2-(2′-pyridyl)- and 2-(2′-pyrimidyl)-indoles in lipid membranes: a joint refinement using fluorescence measurements and molecular dynamics simulations, J. Phys. Chem. B 114 (2010) 13574–13584. [46] T. Shimanouchi, H. Ishii, N. Yoshimoto, H. Umakoshi, R. Kuboi, Calcein permeation across phosphatidylcholine bilayer membrane: effects of membrane fluidity, liposome size, and immobilization, Colloids Surf. B: Biointerfaces 73 (2009) 156–160. [47] M. Pantusa, L. Sportelli, R. Bartucci, Phase behaviour of DPPC/Lyso-PPC mixtures by spin-label ESR and spectrophotometry, J. Spectro. 22 (2008) 153–163. [48] T. Natsume, M. Yoshimoto, Membrane permeability and stability of liposomes suspended in shear flow, J. Dispers. Sci. Technol. 34 (2013) 1557–1562. [49] M. Heinonen, D. Rein, M.T. Satué-Gracia, S.-W. Huang, J.B. German, E.N. Frankel, Effect of protein on the antioxidant activity of phenolic compounds in a lecithin–liposome oxidation system, J. Agric. Food Chem. 46 (1998) 917–922. [50] M.E. Haque, B.R. Lentz, Roles of curvature and hydrophobic interstice energy in fusion: studies of lipid perturbant effects, Biochemistry 43 (2004) 3507–3517.

K.O. Evans et al. / Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy 153 (2016) 333–343 [51] Y.C. Lee, T.F. Taraschi, N. Janes, Support for the shape concept of lipid structure based on a headgroup volume approach, Biophys. J. 65 (1993) 1429–1432. [52] J. Eastoe, M.H. Hatzopoulos, P.J. Dowding, Action of hydrotropes and alkylhydrotropes, Soft Matter 7 (2011) 5917–5925.

343

[53] M. Hopkins Hatzopoulos, J. Eastoe, P.J. Dowding, S.E. Rogers, R. Heenan, R. Dyer, Are hydrotropes distinct from surfactants? Langmuir 27 (2011) 12346–12353. [54] T.R. Heera, L. Cindrella, Molecular orbital evaluation of charge flow dynamics in natural pigments based photosensitizers, J. Mol. Model. 16 (2010) 523–533.