Characterization of Titratable Amphiphiles in Lipid Membranes by

Oct 28, 2015 - PDF. la5b03258_si_001.pdf (798.94 kB). Citing Articles; Related Content. Citation data is made available by participants in Crossref's ...
2 downloads 17 Views 991KB Size
Article pubs.acs.org/Langmuir

Characterization of Titratable Amphiphiles in Lipid Membranes by Fluorescence Spectroscopy Philippe Pierrat and Luc Lebeau* Laboratoire de Conception et Application de Molécules Bioactives, Faculté de Pharmacie, UMR7199 CNRS - Université de Strasbourg, 74 route du Rhin - BP 60024, 67401 Illkirch, France S Supporting Information *

ABSTRACT: Understanding the ionization behavior of lipid membranes is a key parameter for successful development of lipidbased drug delivery systems. Accurate determination of the ionization state of a titratable species incorporated in a lipid bilayer however requires special care. Herein we investigated the behavior of titratable lipids in liposomes by fluorescence spectroscopy and determined which extrinsic parametersi.e., besides those directly related to their molecular structuredetermine their ionization state. Two fluorescent dyes, TNS and R18, have been used to investigate basic and acidic titratable lipids, respectively. Our results suggest that the titration behavior of the ionizable lipid in the membrane is more sensitive to the composition of the membrane and to its physical state than to the presence of solutes in the aqueous phase. Essentially overlooked in earlier studies on ionizable lipid assemblies, the concentration of the titratable lipid in the membrane was found to have a major effect on the ionization state of the lipid polar head. This may result in a shift in the apparent pKa value which may be as large as two pKa units and cannot be satisfactorily predicted.



INTRODUCTION Until the discovery of liposomes by Bangham more than 50 years ago,1 lipid nanoparticles have been extensively investigated and, in the past two decades, have been clinically translated as pharmaceutical drug delivery carriers for their ability to improve on drug tolerability, circulation half-life, and efficacy.2 More recent developments focus to the design of environmentally responsive lipid particles.3 These nanoparticles produce physicochemical changes when exposed to external stimuli that favor drug release at the target site. Among these environmental stimuli, pH gradients have been widely used to design pH-responsive liposome-based drug delivery systems. Indeed, most organs, tissues, and subcellular compartments, as well as their pathophysiological states, can be characterized by their pH levels and gradients. For example, in the gastrointestinal tract pH is ranging from 1−3, in the stomach lumen, to 6.6−7.5, in the duodenum and ileum.4 In tumor tissues, extracellular pH is decreased from 7.4 to ca. 6.5 due to the Warburg effect.5 At the subcellular level, the pH of endosomes drops from 7.2 to ca. 6.0−6.5 along the maturation process of these compartments.6 Thus, it is critical to the identification of structure−activity relationships which can be used to rationally design a liposome-based drug delivery systemi.e., with optimal biophysical propertiesto understand the ionization behavior of a titratable amphiphile in liposomes. The acid dissociation constant (pKa) of a titratable compound is an important value to determine the extent of ionization as a function of pH, especially to accurately consider © XXXX American Chemical Society

its interactions with the biological environment. However, considering titratable amphiphilic compounds, direct pKa measurement is quite difficult. First, the ionizable headgroup of a membrane-bound amphiphile is located at a hydrophobic/ hydrophilic interface where there is a nonisotropic distribution of the surrounding molecular species (neighbor amphiphilic molecules two-dimensionally distributed within the plane of the membrane and solutes in the three-dimensional aqueous phase facing the membrane). Consequently, standard methods used for pKa determination of water-soluble compounds, e.g. potentiometric titration, cannot be accurately applied. Indeed, this requires solubilization of the amphiphilic compound in micelles which may significantly alter its dissociation properties.7−10 Second, protonation/deprotonation of a titratable functional group is strongly influenced by the local dielectric constant. Fromherz and Fernandez estimated that the latter is in the range 10−30 at the surface of micelles, whereas it is 78 in pure water.11 Hence, the acid−base equilibrium is strongly shifted toward uncharged species and thus decreases the pKa of an amine and, oppositely, increases that of an acid titratable group. Shifts may represent one to two pKa units when compared with a “soluble” form of the analyte.7,11,12 The ionization state of titratable amphiphilic molecules can be however determined by specific means. For instance, the Received: August 31, 2015 Revised: October 23, 2015

A

DOI: 10.1021/acs.langmuir.5b03258 Langmuir XXXX, XXX, XXX−XXX

Article

Langmuir

Figure 1. Structure of the two series of titratable lipids investigated in this work. The compounds in the left series (1−9) will provide cationic species at low pH (ammonium form), whereas those in the right series (10−16) will provide anionic species at high pH (phenate form).

ionization state of fluorescent drugs associated with membranes has been determined using fluorometric titration.13 Other related nonfluorescent analytes were investigated by piezoelectric acoustic sensing.14 A mono-rhamnolipid has been evaluated by a conventional potentiometric method at low concentration (below its critical micelle concentration, CMC).9 While the usual sigmoidal shape was observed below the CMC of the compound, almost linear evolution of the pH was observed above the CMC, revealing distinct chemical environments due to the presence of heterogeneous aggregates. The authors proposed alternative analytical techniques for a more accurate pKa determination (1H NMR and ATR-FT-IR spectroscopy). Other authors described as well the use of NMR spectroscopy to determine the apparent pKa of cholic acid in a variety of physiologically relevant environments.15 The results revealed impressive shift of the related pKa that ranged from 4.2 for the monomer in water to 7.3 when associated with phospholipids. The reported 13C NMR experiments however required 90% 13C isotopic substitution which certainly is an obstacle to a wide use of this technique. In the same line, 2H NMR spectroscopy has been successfully used to determine the pKa of N,N-dimethylsphingosine, a protein kinase C (PKC) inhibitor, when inserted in deuterated phospholipid membranes.16 Monolayer isothermal compression at the air−water interface using a Langmuir trough also has been proposed as a reliable method to characterize the ionization state of insoluble materials.17 It has been applied to p-dioctadecanoylcalix[4]arene, and plotting the area per molecule as a function of pH at a fixed surface pressure (Π) allowed identification of two inflection points that were attributed to the stepwise dissociation of the phenol groups of the compound resulting in electrostatic repulsions between charged species at the interface. More recently, some authors have proposed zeta-

potential measurement as a surrogate to analyze the pHdependent surface charge of a membrane.18 As a part of an ongoing project focused to targeted drug delivery to cancer cells, we were interested in the development of titratable lipids that might help in the general drug delivery process by liposomes. The growth of solid tumors is characterized not only by the uncontrolled proliferation of cancer cells but also by increased glucose metabolism, the subsequent production of lactic acid leading to enhanced acidification of the extracellular milieu.19,20 Modification of the ionization state of titratable “helper” lipids in the acidic tumor microenvironment (pH-induced charge conversion) might be exploited to trigger specific interactions between the negatively charged membrane of the cancer cells and the drug-containing liposomes. For this strategy to be successful, titration of the helper lipids should occur in a narrow pH range centered around 6.5, corresponding to that of the acidic extracellular matrix of tumors.21 Two strategies may be developed in order to take advantage of the anomalous lactic acid accumulation in rapidly growing tumor cells. The first one involves a neutral lipid at physiological pH that will be protonated in the tumor microenvironment. The subsequent cationic lipid may thus interact with the negatively charged membrane of the tumor cells and trigger liposome fusion with the plasma membrane, or endocytosis, both mechanisms resulting in cell internalization of the antitumor drug loaded in the liposome. Though this approach was mainly investigated with polymeric drug delivery systems,22 there are some recent examples with lipidic carriers.23 The second strategy has not been investigated so far to the best of our knowledge and involves a negatively charged lipid at pH 7.4 that becomes neutralized at acidic pH, hence displaying a hydrophobic mobile portion that may promote fusion with cells upon membrane insertion.24 In order B

DOI: 10.1021/acs.langmuir.5b03258 Langmuir XXXX, XXX, XXX−XXX

Article

Langmuir

Figure 2. Structure of TNS (17) and of the cationic fluorescent dyes evaluated for the pKa determination of anionic lipids (18: ethidium bromide; 19: Cy5; 20: R18). apparent pKa was determined as the pH-giving rise to half-maximal fluorescence intensity. Determination of Apparent pKa of Amino Lipids in Lipoplexes. A dry lipid film containing DOPC/9 (10:90, mol %) was hydrated with salmon sperm DNA (1 mL, 1.8 mM [PO4] in buffered solution at pH 6) for 30 min at room temperature before vortex mixing for 2 min to achieve a 6 mM lipid final concentration. Lipoplexes obtained as a homogeneous preparation were diluted to 100 μM lipid in 1.5 mL of buffered solutions with pH ranging from 3.0 to 11.0 (vide supra) and were kept at room temperature for 1 h. Then TNS (30 μL, 33 μM in deionized water) was added to achieve a final TNS concentration of 0.65 μM. After vortex mixing, the solutions were allowed to equilibrate for 30 min at room temperature, and TNS fluorescence was measured as described above. Determination of Apparent pKa of Phenolic Lipids (PL) Using the Cationic Fluorescent Probe R18. The multilamellar vesicles (MLVs) used for apparent pKa determinations were prepared by hydrating a dry lipid film containing DOPC/PL/R18 (80:10:10, mol %) in PBS pH 7.0 (333 μL) for 15 min at room temperature before vortex mixing for 2 min to achieve a 6 mM lipid final concentration. A homogeneous clear preparation was obtained. Vesicles were stepwise diluted to 600 μM in PBS and to 12.0 μM in 1.5 mL of buffered solutions with pH ranging from 3.0 to 11.0 (vide supra). After 30 min at room temperature, rhodamine fluorescence was measured at 20 °C, using excitation and emission wavelengths of 568 and 594 nm, respectively. A sigmoidal best-fit analysis was applied as described above to determine the pKa of lipid. Large unilamellar vesicles (LUVs) were prepared by hydrating a dry lipid film containing DOPC/PL/R18 (80:10:10, mol %) in buffer at pH 7.0 (333 μL) for 15 min at room temperature before vortex mixing for 2 min to achieve a 6 mM lipid final concentration The homogeneous clear preparation was extruded 10 times through 100 nm polycarbonate filters to provide LUVs. Vesicles were diluted to 12.0 μM in 1.5 mL of buffered solutions with pH ranging from 3.0 to 11.0 (vide supra). After 30 min at room temperature, rhodamine fluorescence was measured as described above.

to develop these strategies in a safe and useful manner, it is of utmost importance that the pKa for these helper lipids embedded in a liposome bilayer can be determined accurately. Herein we describe the methods we have used to determine the ionization state of titratable lipids switching from neutral to positively charged (compounds 1−9) and negatively charged to neutral (compounds 10−16), upon acidification (Figure 1). They involve the use of fluorescent dyes which fluorescence changes as a function of the bulk pH could be correlated with the ionization state of the lipids of interest. We investigated the influence of the molar composition and physical state of the lipid membrane, and of the aqueous phase composition, on the titration behavior of the ionizable lipids. Especially the introduction of a PEGylated lipid (DSPE-mPEG2000) in the lipid membrane to produce stealth lipid particles was analyzed, and subsequent alteration of the titration behavior of the lipids was determined. Finally, nanoparticles resulting from the complexation of titratable liposomes with DNA were prepared, and their titratable behavior was examined and compared to that of original liposomes.



MATERIALS AND METHODS

Chemicals. Unless otherwise stated, all chemical reagents were purchased from Alfa Aesar (Bischeim, France) and were used as received. 1,2-Dioleoyl-sn-glycero-3-phosphocholine (DOPC) was from Lipoid AG (Cham, Switzerland). 1,2-Dipalmitoyl-sn-glycero-3-phosphocholine (DPPC) and 1,2-distearoyl-sn-glycero-3-phosphoethanolamine-N-[methoxy(poly(ethylene glycol))-2000] (ammonium salt) (DSPE-mPEG2000) were from Avanti Polar Lipids (Alabaster, AL). All buffer solutions were prepared with deionized water purified with an EMD Millipore Milli-Q integral system (resistivity ≤18.2 MΩ·cm) and filtered through a 0.2 μm carbonate membrane. The synthesis of compounds 1−16 will be reported elsewhere. Rhodamine B octadecyl ester (R18) was synthesized as previously reported for similar compounds.25 Determination of Apparent pKa of Amino Lipids (AL) Using the Anionic Fluorescent Probe TNS. The multilamellar vesicles (MLVs) used for apparent pKa determinations were prepared by hydrating a dry lipid film containing DOPC/AL (90:10, mol %) in PBS pH 7.0 (2 mL) for 15 min at room temperature before vortex mixing for 2 min to achieve a 6 mM lipid final concentration. A homogeneous clear preparation was obtained. The vesicles were diluted to 100 μM lipid in 1.5 mL of buffered solutions containing 10 mM HEPES, 10 mM MES, 10 mM NH4OAc, and 130 mM NaCl (unless otherwise stated), with pH ranging from 3.0 to 11.0. Then 2(p-toluidino)-6-naphthalenesulfonic acid (TNS, 30 μL, 33 μM in deionized water) was added to achieve a final TNS concentration of 0.65 μM. After vortex mixing, the solutions were allowed to equilibrate for 30 min at room temperature. The surface charge of the liposomes was monitored by determining the TNS fluorescence at each pH, which was measured at 20 °C with a Fluoromax-4 spectrofluorometer (Horiba Scientific) using excitation and emission wavelengths of 321 and 445 nm, respectively. A sigmoidal best-fit analysis (using the Origin software suite) was applied to the fluorescence data, and the



RESULTS AND DISCUSSION TNS Fluorescent Assay for Determination of the Apparent pKa of Amino Lipids in a Membrane as a Function of the Membrane and Aqueous Phase Compositions. Lipid vehicles bearing protonatable functional groups appeared in the literature for various purposes and mostly for improving drug delivery inside cells through progressive protonation of an amine group in the endosomal compartment.26−28 A convenient means for assessing the pKa value of amino lipids involves the fluorescent dye TNS (Figure 2, compound 17). The method was introduced by Eastman et al. to investigate transmembrane pH gradients in unilamellar vesicles.29 It has been further developed to determine the ionization behavior of amino lipids in LUVs or MLVs and calculate their pKa.10,30−32 That negatively charged fluorescent dye associates with cationic lipids in the outer monolayer of the membrane through electrostatic interactions. This results in partitioning of the negatively charged probe into the lipid bilayera milieu with a C

DOI: 10.1021/acs.langmuir.5b03258 Langmuir XXXX, XXX, XXX−XXX

Article

Langmuir low dielectric constantand correspondingly improved fluorescence emission. When the titratable lipid in the neutral form, there is no electrostatic interaction between the dye and the membrane, and hydrophobic interactions do not suffice to ensure stable anchoring of the dye to the membrane. Consequently, the fluorescence variation as a function of pH is a result of a delicate balance between hydrophobic and electrostatic interactions and can be used to investigate the surface charge of the lipid nanoparticles and thus the titration behavior of a compound of interest in the membrane. In this assay, it is assumed that fluorescence of the probe reaches a maximum when 100% of the titratable lipids are ionized, while it is minimum when they are in the un-ionized state. The pH value at which fluorescence is the mean between the highest and lowest fluorescence is proposed as the apparent pKa value of the titratable lipid in the membrane. It cannot be excluded that conformation and/or hydration of the probe within the membrane change(s) with the ionization of the surface, which may translate into some variation in the intensity of the fluorescence emission. The titration curve is thus likely to illustrate two concomitant processes: (1) immobilization of the water-soluble fluorescent probe at the membrane through ion pairing; (2) modification of the conformation and/or hydration of the membrane-bound probe (as a result of a modification of the ionization state of neighboring molecules). Insofar as TNS concentration in the assay is much less than that of the titratable lipid (e.g., less than 1/10), ionization of a small fraction of the titratable lipid (i.e., in stoichiometric amount with respect to the probe) is enough to lead to quantitative association of the probe with the membrane through electrostatic interaction. It thus may be assumed that the titration curve essentially translates the modification of the ionization state of the titratable lipid and thus can be used confidently to determine the apparent pKa value. We used this method to investigate the ionization behavior of the amino lipids 1−9. The latter (10 mol %) were incorporated in DOPC liposome formulations (MLVs). After dilution of the liposomes in various buffered solutions, the fluorescence of TNS was measured and plotted as a function of pH (Figure S1). Fitting the experimental data with a Boltzmann equation allowed calculation of an apparent pKa value, as exemplified in Figure 3

for compound 1. The results obtained for the full series of amino lipids 1−9 are presented in Table 1 and are compared to Table 1. Experimentally Determined and Calculated pKa Values for Amino Lipids 1−9 lipid

measured pKaa

predicted pKab

1 2 3 4 5

7.01 6.63 6.86 3.21 4.49

8.02 7.02 8.16 8.17 5.17

lipid

measured pKaa

predicted pKab

6 7 8 9

6.35 5.87 8.27 8.85

7.09 7.09 8.72 9.36

a

Experimental determinations were performed on liposomes DOPC/ lipid 90:10 (mol %) at 25 °C. bPredicted pKa values were obtained using the ACD/pKa v 6.00 software (ACD/Laboratories, Toronto, ON, Canada).

the predicted pK a values obtained using the ACD/pKa Predictor v 6.00 software suite (ACD/Laboratories, Toronto, ON, Canada).33 This algorithm calculates acid−base ionization constants under standard conditions (25 °C and zero ionic strength) and does account for anisotropic distribution of the titratable compound. As can be seen, the deviation of the pKa value measured according to the TNS fluorescence assay from the predicted value may exceed 1.3−1.4 pKa units (e.g., for compounds 3 and 7). In all cases, however, the experimental value was invariably less than the predicted one, which reflects that electrostatic repulsions between charged polar heads at the surface of the lipid membrane are not taken into account by the pKa calculation algorithm. Though the compounds in the series have similar molecular structures, the deviation of the measured value from the predicted one is highly fluctuating as was already reported for other amino lipids.10 This makes difficult, if not impossible, to determine some experimental correction factor that might help refine the calculated pKa value. It is interesting to note that Bailey and Cullis have previously determined a pKa value of 6.58 for compound 1 in a lipid matrix with a different composition (EggPC/cholesterol/1 55:45:10, mol %).11 This indicates that the lipid bilayer composition may have significant impact on the ionization of the lipid headgroup, as was already suggested by comparative measurements performed on liposomes using the TNS fluorescence method, and on micelles (by addition of neutral surfactants) by potentiometric titration, though the two techniques might have led to different results for other reasons.10 We thus investigated that point a step further and examined the influence of the lipid layer composition on the apparent pKa of amino lipid 1. We determined the apparent pKa of 1 in liposomes, varying the concentration of the titratable compound in the lipid layer from 10 to 90 mol % (Figure 4). A decrease of 2.12 pKa units was observed over the concentration range examined. Such a pKa drop is remarkable and reveals that pKa of the titratable amino lipid in the DOPC matrix is concentration dependent. That deserves to be raised given that for a couple of weak acid and base in dilute aqueous solution, the pKa value is not expected to be concentration-dependent. This unusual behavior is actually a consequence of the high local concentration of the titratable amine at the surface of the lipid layer, i.e., of the short distance separating adjacent titratable groups. Indeed, the decrease in pKa value is likely due to the increasing charge repulsion between protonated amine headgroups. These groups can thus no longer be considered to

Figure 3. Representative plot showing normalized TNS fluorescence intensity as a function of pH in the presence of liposomes of composition DOPC/1 90:10 (mol %). The apparent pKa value of the amino lipid is the pH at which TNS fluorescence is half of its maximum and was obtained after fitting the data with a sigmoid function. D

DOI: 10.1021/acs.langmuir.5b03258 Langmuir XXXX, XXX, XXX−XXX

Article

Langmuir

for the titratable lipids (10 mol %) in DOPC and DPPC matrices. Results support the previous conclusions (Figure 5).

Figure 4. Normalized fluorescence intensity of TNS as a function of pH in the presence of DOPC/1 liposomes with composition ranging from 90:10 to 10:90 (mol %) and apparent pKa (in parentheses) values determined for the titratable lipid.

Figure 5. Normalized fluorescence intensity of TNS as a function of pH in the presence of DOPC/3 90:10 and DPPC/3 liposomes (mol %) and apparent pKa values (in parentheses) determined for the titratable lipid in both cases.

be independent from each other as is the case in a dilute solution, and the titration properties of the lipid assembly are closer to those of a titratable polymer in solution.34 Though the pKa value of all the titratable amino lipids investigated herein showed concentration dependence, this tendency was attenuated when the amine group was connected to the hydrophobic part of the molecule through a long hydrophilic spacer (Figure S2). A reasonable explanation may be that the longer the spacer, the larger the volume the titratable group can distribute in and the closer to that in solution the titration behavior of the lipid in the membrane. Our observations however call for a more detailed examination. Indeed, determination of the apparent pKa value using the fluorescence assay is based on the implicit assumption that fluorescence intensity varies linearly with the amount of ionized species in the membrane. We tentatively checked this point by measuring the TNS fluorescence in the presence of DOPC liposomes incorporating increasing amount of the titratable lipid 1 (Figure S3). Experiments were conducted under acidic conditions (pH 3.0) to ensure full ionization of the titratable lipid. The results obtained indicate that the assumption about linear response is not strictly valid. Indeed, the relationship between TNS fluorescence intensity and amount of titratable lipid in the membrane follows a polynomial rather than linear distribution. However, considering liposomes with low content of titratable lipid in the membrane (up to ca. 30 mol %), a linear variation of the fluorescence emission can be comfortably considered (coefficient of determination R2 > 0.95). This is not the case at higher content of 1 in the membrane for which deviation from linearity rapidly increased. The same phenomenon was observed for all the amino lipids investigated herein, and in some cases fluorescence linearity could be approximated for higher content of titratable lipid in the membrane (Figure S4). This is e.g. the case with amino lipid 9 for which satisfactory linearity was observed with liposomes containing up to 50 mol % of the compound (R2 > 0.95). As a corollary of the nonlinearity of the fluorescence emission, the TNS fluorescence assay leads to some overestimation of the pKa value of amino lipids when at high concentration in a membrane. To further investigate the ionization state of the amino lipids in membranes, we compared the apparent pKa values measured

With a main phase transition temperature (TM) of −17 °C, DOPC forms bilayers in the fluid phase, with a surface per molecule of about 0.57 nm2.35 For mixed liposomes containing 10 mol % of a titratable lipid, the mean distance separating two neighboring titratable molecules can be estimated at about 2.4 nm. In the case of DPPC (TM = +41 °C, A = 0.41 nm2/ molecule), making the assumption that DPPC perfectly mixes with unsaturated lipids, this estimated distance decreases to ca. 2.0 nm. Consequently, the charge repulsion between protonated amine headgroups is increased which translates into a decrease of the pKa value. In the case of compound 3, the measured pKa downshift corresponds to 1.17 pKa unit. Again, the presence of a long hydrophilic spacer between the hydrophobic part of the titratable lipid and the polar head attenuates the pKa downshift as was observed with compound 8 (Figure S5). Most liposomal drug delivery systems incorporate cholesterol (chol) in their formulation. This sterol which is naturally found in cell membranes is used for fluidity regulation of the phospholipid membranes, decreasing their permeability to hydrophilic molecules, and increasing the liposome stability.36,37 We observed that neutral cholesterol also has significant impact on the ionization state of the membrane (Figure 6). Experiments conducted with lipid 3 (10 mol %) revealed that 30% cholesterol in the membrane provoked a pKa downshift from 6.86 to 5.39. This might be consistent with a partial dehydration of the titratable lipid headgroup due to the presence of the cholesterol polar head which is less hydrophilic than the phosphocholine moiety. The full validation of this hypothesis however would require a determination of the mean distance separating two neighboring titratable molecules in both DOPC/chol/3 90:0:10 and 70:30:10 liposome preparations. As previously observed in the absence of cholesterol (Figure 4), increasing the amount of the titratable lipid in the liposomes (5 to 25 mol %) further decreased the pKa value and had a cumulative effect with cholesterol incorporation. Liposome preparations used for in vivo applications must display prolonged circulation half-life and reduced immunogeE

DOI: 10.1021/acs.langmuir.5b03258 Langmuir XXXX, XXX, XXX−XXX

Article

Langmuir

90:10, mol %), the apparent pKa value shifted from 7.00 to 6.75. The pKa downshift observed at higher concentration of the ionizable lipid (DOPC/1 10:90, mol %) was even less important (4.88 to 4.79). From that it can be concluded that grafting PEG chains on the surface of the membrane has only little effect on the hydration of the polar head of the titratable lipid. Besides, experiments carried out with various amounts of NaCl (0−130 mM) in the buffered solution revealed as well a relatively weak dependence of the pKa value on medium salinity (Figure S6). Considering that the PEG moiety in the previous set of experiments preferentially distributed in the aqueous phase, our results may suggest that the ionization state of the titratable lipid headgroup is not much sensitive to the local presence of fully hydratedi.e. water-solublechemical species. Cationic lipids are powerful tools for delivering nucleic acid to cells. We thus attempted to get some insight into the ionization state of lipoplexes, these nanosized complexes that form through electrostatic interactions between negatively charged nucleic acid and cationic lipids. Lipoplexes were thus prepared using a 3-fold excess of cationic lipid 9 with respect to the DNA phosphate contents (N/P = 3) and DOPC (10 mol %), and TNS fluorescence measurements were performed for determining the titration behavior of the resulting nanoparticles (Figure 8). As can be observed, the lipid complexation with

Figure 6. Normalized fluorescence intensity of TNS as a function of pH in the presence of DOPC/chol/3 liposomes and apparent pKa values (in parentheses) determined for the titratable lipid.

nicity. PEGylation of liposomesi.e., covalent attachment of poly(ethylene glycol) (PEG) to the lipid layerhas been widely recognized one of the more promising methods for achieving this goal and produce what is usually called “stealth” liposomes.38 To date, several PEGylated liposomal medicines have already been approved by the U.S. Food and Drug Administration (FDA) and the European Medicines Agency (EMA). The hydrophilic and bioinert PEG chains act as “protein repellent” due to their flexibility, high number of possible conformations, and hydration degree. PEGylation of liposomes also might have consequences on the ionization state of titratable lipids in the membranes. To investigate this, we performed TNS fluorescence measurements with liposomes incorporating DSPE-mPEG2000 (Figure 7). The amount of PEG-lipid (5 mol %) introduced in the liposomes corresponds to the one which is most commonly used to prepare stealth liposomes. Rather unexpectedly, PEGylation only provoked minor alteration of the ionization state of the titratable lipid 1. At low concentration of the latter in the bilayer (DOPC/1

Figure 8. Titration curves determined for DOPC/9 10:90 liposomes and the corresponding lipoplexes prepared at a charge ratio N/P = 3. The apparent pKa values determined for the titratable lipid in both particles are indicated in parentheses.

DNA caused a left shift of the titration curve when compared to neat liposomes, which translated into a decrease of the apparent pKa value from 8.25 to 7.44. This was not a result of any direct interaction between TNS and DNA, as was checked over the pH range of interest (Figure S7). Consequently, the amine group of the titratable lipids in lipoplexes were less easily protonated than in liposomes. Considering that titratable headgroups revealed not much sensitive to local concentration of electrolytes (vide supra), this phenomenon most likely originated from a difference in the lipid arrangement within the particles. Indeed, it is well-known that lipid complexation with DNA usually results in formation of ordered phases, in most cases tight lamellar phases, although nonlamellar phases (inverted hexagonal phase and cubic phases) may also form depending on the lipid structure.39,40 These rearrangements are

Figure 7. Normalized fluorescence intensity of TNS as a function of pH in the presence of DOPC/1/DSPE-mPEG2000 liposomes and apparent pKa values (in parentheses) determined for the titratable lipid. F

DOI: 10.1021/acs.langmuir.5b03258 Langmuir XXXX, XXX, XXX−XXX

Article

Langmuir

at 568 nm (Figure 9). The collected data revealed that the fluorescence of the probe was stable over the whole pH range

driven by cooperative tethering of the DNA polyphosphate chain to the lipid membrane, imposing lateral compression of the lipids, i.e., reduction of the mean distance separating two neighboring titratable lipid headgroups. Similarly to what was observed when replacing DOPC for DPPC in liposomes (vide supra), this is expected to translate into a pKa decrease which indeed is what was observed experimentally. Fluorescent Approach To Determining the Apparent pKa of Titratable Acidic Lipids in a Membrane. While the titration behavior of protonatable amino lipids has received significant attention, the same is not true for lipids that may switch from a neutral to an anionic form under physiologically relevant conditions. Especially, we did not find any report in the literature on the use of a specific fluorescent probe for the determination of the pKa of acidic lipids in membranes. Hence, we had been looking for fluorescent dyes suitable for investigating the ionization state of such lipids. We selected a series of three cationic fluorescent dyes likely to establish pHsensitive interactions with liposomes incorporating anionic lipids (Figure 2, compounds 18−20). One prerequisite for the intended application is that fluorescence of the probe is not intrinsically pH-dependent. This means that the probe should not possess any functional group likely to react with hydroxyl or hydronium ions in the pH range investigated. As a first attempt, ethidium bromide (18) was examined as it is expected to have potential to interact with hydrophobic species and displayed satisfactory pH-independent fluorescence properties between pH 3 and pH 11 (data not shown). Furthermore, emission intensity and lifetime increase nearly 25 times when ethidium bromide is intercalated between the plates of hydrophobic base pairs in nucleic acid. However, no significant variation of the fluorescence regime could be observed when in the presence of titratable liposomes (DOPC/18 90/10, mol %) over the full pH range investigated (3−10). This probably is an indication that the hydrophobic interaction between ethidium bromide and the lipid membrane is not robust enough to ensure significant distribution of the dye at the surface of the liposomes. With the aim to improve the hydrophobicity of the probe while preserving its aqueous solubility, we then examined the polycationic compound 19 in the Cy5 dyes family. Unexpectedly, this compound displayed marked pH-dependent fluorescence that should be attributed to the titration of the indole group, though calculations predict a negative pKa value (−0.42 ± 0.40). Tuning the hydrophobicity of this dye through any structure modifications hence would have likely revealed inefficient in our approach and was not investigated further. Last we examined rhodamine B octadecyl ester (R18, 20). Since the seminal paper of Hoekstra et al.,41 this lipophilic dye has been extensively used to evaluate the kinetics of fusion in a wide range of systems such as liposomes, viruses, erythrocytes, etc. The method developed relies upon the relief of selfquenching of the probe incorporated into a bilayer of interest when the latter fuses with other membranes. Put another way, fusion of the labeled membrane with a nonlabeled one causes a decrease in the probe surface density (lipid dilution) which translates into fluorescence dequenching, i.e., increase in the intensity of the fluorescence emission. To evaluate the potential of R18 to deliver accurate information on the ionization state of lipids, we first examined the impact of pH on the fluorescence properties of this probe. To do so, a dry lipid film (DOPC/R18 90/10, mol %) was hydrated to fabricate liposomes. The latter were diluted in buffered solutions at various pH, and fluorescence emission was assessed at 594 nm under excitation

Figure 9. Intensity of fluorescence emission of the nonionizable DOPC/R18 90:10 vesicles as a function of bulk pH.

investigated, i.e., from 3 to 11. This result is not so trivial and has to be examined in detail. First, considering that charged molecules cannot permeate mechanically cohesive, nonleaky lipid bilayers for time periods from hours to days,42,43 it can be assumed that the aqueous core of liposomes was not subject to significant composition variation or pH equilibration with the bulk solution during the course of the experiment which only lasted a few minutes. Second, ΔpH-driven translocation of R18 from one membrane leaflet to the other can be ruled out as it was shown this cannot occur within such a short time frame for a nontitratable charged amphiphile.29,44,45 However, R18 was shown to massively move between the inner and outer leaflets of bilayers upon application of a transmembrane potential.46,47 The subsequent asymmetrical distribution of the fluorescent probe within the two leaflets has consequences on the selfquenching efficiency as it is directly proportional to the ratio of R18 to total lipid in the leaflet.41 Therefore, in the light of the data presented in Figure 9 and considering the stability of the rhodamine fluorescence, it can be concluded that in our experimental setup the transmembrane potential resulting from the difference in the composition of the intra- and extraliposomal milieux (cf. Materials and Methods section) was not high enough to provoke significant transport of the lipophilic probe from one leaflet to the other. In the same line, we can also conclude that pH variations outside the liposomes did not significantly change the local dielectric constant at the membrane outer leaflet; otherwise, fluorescence intensity modulation would also have been detected. It therefore appeared that R18 might indeed stand as a candidate for monitoring the ionization state of the titratable headgroup of acid lipids in membranes. Experiments thus were conducted on DOPC-based LUVs containing R18 (10 mol %) and compounds 10−16 (10 mol %), and the apparent pKa of these titratable compounds was determined by fluorescence measurements (Figure S8). The results obtained for the full series of phenol lipids are presented in Table 2 and are compared to the predicted pKa value (isotropic distribution, 25 °C and zero ionic strength) for each compound. Similar to what was previously observed with amino lipids, the G

DOI: 10.1021/acs.langmuir.5b03258 Langmuir XXXX, XXX, XXX−XXX

Article

Langmuir

membrane and found acceptable linearity up to 30 mol % titratable lipid (R2 > 0.98) (Figure S10). Beyond this ratio, deviation from linearity increased extremely rapidly. Thus, as previously observed with the amino lipids, at high concentration of the titratable lipid in the membrane, the assay leads to overestimation of the pKa value. To rule out the hypothesis of any artifactual interaction between the lipid and some divalent components of the buffered saline, a series of measurements were carried out replacing HEPES in the original buffer saline by the monovalent buffering agent N,N-bis(2-hydroxyethyl)-2aminoethanesulfonic acid (BES, 10 mM). Indeed, as a piperazine derivative, HEPES might potentially interact simultaneously with two anionic groups, thus facilitating phenol deprotonation through a cooperative process. Under the new buffer conditions, however, no significant variation of the pKa values was observed (Figure S11), indicating that the pKa decrease upon concentration was a result of intrinsic properties of lipid 13. Especially, high concentration of lipid 13 in the DOPC matrix favors π−π stacking interaction between phenol rings which might thus alter their titration behavior. There are some previous reports in the literature supporting this interpretation, and our results are consistent with those obtained e.g. by Olasz et al., who found that stacking interaction may reduce the pKa of phenol by up to two units.48

Table 2. Experimentally Determined and Predicted pKa Values for Phenol Lipids 10−16 lipid

measured pKaa

predicted pKab

10 11 12 13

9.13 8.74 9.38 6.64

8.12 8.46 8.22 6.68

lipid

measured pKaa

predicted pKab

14 15 16

8.19 5.14 5.61

7.19 5.22 5.21

a

Experimental determinations were performed on liposomes DOPC/ lipid 90:10 (mol %) at 25 °C. bPredicted pKa values were obtained using the ACD/pKa software.

experimentally determined pKa values may be quite distant from those predicted by the ACD/pKa calculation algorithm. In all cases, however, the experimental determination gave a pKa value equal to or higher than the predicted one. Again this is consistent with the failure by the computing algorithm to account for the anisotropic distribution of the ionizable headgroups. To establish the robustness of the method, we determined whether the liposome architecture might have some influence on the fluorescence measurements and pKa determination. Thus, multilamellar large vesicles (MLVs) were compared to large unilamellar vesicles (LUVs), and no significant difference between the apparent pKa value determined in these two liposome configurations was observed (Figure S9). Variation of the amount of fluorescent dye in the preparations (5 and 10 mol %) also had no effect on the apparent pKa value (data not shown). Finally, and intriguingly, increasing the proportion of the titratable lipid in the membrane led to lower apparent pKa value (Figure 10). In parallel to what was observed with amino



CONCLUSIONS Understanding the ionization behavior of lipid membranes is a key parameter for successful development of lipid-based drug delivery systems. Accurate determination of the ionization state of a titratable species incorporated in a lipid bilayer however requires special care. Herein we investigated the behavior of titratable lipids in liposomes by fluorescence spectroscopy and determined which extrinsic parametersi.e., besides those directly related to their molecular structuredetermine their ionization state. Our results suggest that the titration behavior of an ionizable lipid in a membrane is more sensitive to the composition of the membrane and to its physical state than to the presence of solutes in the aqueous phase. Essentially overlooked in previous studies on ionizable lipid assemblies, the concentration of the titratable lipid in the membrane was found to have a major effect on the ionization state of the lipid polar head. This may result in a shift in the apparent pKa value which may be as large as two pKa units and cannot be satisfactorily predicted. Besides, we brought some evidence that a linear relationship between the fluorescence emission of the probe and the ionization state of the titratable lipid in the membrane can only be invoked when the latter is present at moderate concentration in the membrane. At high concentration (>30− 50 mol %) deviation from linearity may be quite important, leading to overestimation of the pKa value.

Figure 10. Normalized fluorescence intensity of DOPC/R18 90:10 (mol %) liposomes containing variable amounts of titratable lipid 13 (10−70 mol %) as a function of pH. The apparent pKa value determined for the titratable lipid is indicated in parentheses.



ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.langmuir.5b03258.

lipid 1 (cf. Figure 4), the increasing charge repulsion between the phenate groups indeed was expected to result in an increase in the pKa value at the higher concentrations of titratable lipid 13 (the higher the density of anionic groups at the surface of the lipid layer, the harder to generate additional anionic sites). To tentatively interpret the observed phenomenon, we first determined the relationship between the rhodamine fluorescence intensity and the amount of titratable lipid in the



Figures S1−S11 (PDF)

AUTHOR INFORMATION

Corresponding Author

*(L.L.) E-mail: [email protected]. H

DOI: 10.1021/acs.langmuir.5b03258 Langmuir XXXX, XXX, XXX−XXX

Article

Langmuir Notes

(20) Neri, D.; Supuran, C. T. Interfering with pH regulation in tumours as a therapeutic strategy. Nat. Rev. Drug Discovery 2011, 10, 767−777. (21) Gao, W.; Chan, J. M.; Farokhzad, O. C. pH-Responsive nanoparticles for drug delivery. Mol. Pharmaceutics 2010, 7, 1913− 1920. (22) Drummond, D. C.; Zignani, M.; Leroux, J.-C. Current status of pH-sensitive liposomes in drug delivery. Prog. Lipid Res. 2000, 39, 409−460. (23) Mo, R.; Sun, Q.; Xue, J. W.; Li, N.; Li, W. Y.; Zhang, C.; Ping, Q. N. Multistage pH-responsive liposomes for mitochondrial-targeted anticancer drug delivery. Adv. Mater. 2012, 24, 3659−3665. (24) Pierrat, P.; Casset, A.; Didier, P.; Kereselidze, D.; Lux, M.; Pons, F.; Lebeau, L. Cationic DOPC-detergent conjugates for DNA and siRNA delivery: Structural requirements for transfection efficiency and safety. Submitted for publication. (25) Halterman, R. L.; Moore, J. L.; Mannel, L. M. Disrupting aggregation of tethered rhodamine B dyads through inclusion in cucurbit[7]uril. J. Org. Chem. 2008, 73, 3266−3269. (26) Ishida, T.; Okada, Y.; Kobayashi, T.; Kiwada, H. Development of pH-sensitive liposomes that efficiently retain encapsulated doxorubicin (DXR) in blood. Int. J. Pharm. 2006, 309, 94−100. (27) Akinc, A.; Zumbuehl, A.; Goldberg, M.; Leshchiner, E. S.; Busini, V.; Hossain, N.; Bacallado, S. A.; Nguyen, D. N.; Fuller, J.; Alvarez, R.; Borodovsky, A.; Borland, T.; Constien, R.; de Fougerolles, A.; Dorkin, J. R.; Narayanannair Jayaprakash, K.; Jayaraman, M.; John, M.; Koteliansky, V.; Manoharan, M.; Nechev, L.; Qin, J.; Racie, T.; Raitcheva, D.; Rajeev, K. G.; Sah, D. W. Y.; Soutschek, J.; Toudjarska, I.; Vornlocher, H.-P.; Zimmermann, T. S.; Langer, R.; Anderson, D. G. A combinatorial library of lipid-like materials for delivery of RNAi therapeutics. Nat. Biotechnol. 2008, 26, 561−569. (28) Semple, S. C.; Akinc, A.; Chen, J.; Sandhu, A. P.; Mui, B. L.; Cho, C. K.; Sah, D. W. Y.; Stebbing, D.; Crosley, E. J.; Yaworski, E.; Hafez, I. M.; Dorkin, J. R.; Qin, J.; Lam, K.; Rajeev, K. G.; Wong, K. F.; Jeffs, L. B.; Nechev, L.; Eisenhardt, M. L.; Jayaraman, M.; Kazem, M.; Maier, M. A.; Srinivasulu, M.; Weinstein, M. J.; Chen, Q.; Alvarez, R.; Barros, S. A.; De, S.; Klimuk, S. K.; Borland, T.; Kosovrasti, V.; Cantley, W. L.; Tam, Y. K.; Manoharan, M.; Ciufolini, M. A.; Tracy, M. A.; de Fougerolles, A.; MacLachlan, I.; Cullis, P. R.; Madden, T. D.; Hope, M. J. Rational design of cationic lipids for siRNA delivery. Nat. Biotechnol. 2010, 28, 172−176. (29) Eastman, S. J.; Hope, M. J.; Cullis, P. R. Transbilayer transport of phosphatidic acid in response to transmembrane pH gradients. Biochemistry 1991, 30, 1740−1745. (30) Bailey, A. L.; Cullis, P. R. Modulation of membrane fusion by asymmetric transbilayer distributions of amino lipids. Biochemistry 1994, 33, 12573−12580. (31) Heyes, J.; Palmer, L.; Bremner, K.; MacLachlan, I. Cationic lipid saturation influences intracellular delivery of encapsulated nucleic acids. J. Controlled Release 2005, 107, 276−287. (32) Jayaraman, M.; Ansell, S. M.; Mui, B. L.; Tam, Y. K.; Chen, J.; Du, X.; Butler, D.; Eltepu, L.; Matsuda, S.; Narayanannair, J. K.; Rajeev, K. G.; Hafez, I. M.; Akinc, A.; Maier, M. A.; Tracy, M. A.; Cullis, P. R.; Madden, T. D.; Manoharan, M.; Hope, M. J. Maximizing the potency of siRNA lipid nanoparticles for hepatic gene silencing in vivo. Angew. Chem., Int. Ed. 2012, 51, 8529−8533. (33) This software uses a fragment-based approach to prediction and Hammett equations derived from a reference database containing 31 000 experimental pK a values (www.acdlabs.com/products/ percepta/predictors/pka). (34) Blagbrough, I. S.; Metwally, A. A.; Geall, A. J.; Clifton, N. J. Measurement of polyamine pKa values. Methods Mol. Biol. 2011, 720, 493−503. (35) Lebeau, L.; Mioskowski, C.; Oudet, P. Systematic study of phospholipids linked to a steroid derivative, spread into a monolayer at the air/water interface. Biochim. Biophys. Acta, Biomembr. 1988, 939, 417−429.

The authors declare no competing financial interest.



ACKNOWLEDGMENTS The authors are grateful for the financial support received from la Ligue contre le Cancer (CCIR-GE 2013, ref 1FI11301OGGF).



REFERENCES

(1) Bangham, A. D.; Horne, R. W. Negative staining of phospholipids and their structural modification by surface-active agents as observed in the electron microscope. J. Mol. Biol. 1964, 8, 660−668. (2) Torchilin, V. P. Recent advances with liposomes as pharmaceutical carriers. Nat. Rev. Drug Discovery 2005, 4, 145−160. (3) Ganta, S.; Devalapally, H.; Shahiwala, A.; Amiji, M. A review of stimuli-responsive nanocarriers for drug and gene delivery. J. Controlled Release 2008, 126, 187−204. (4) Tomoyasu, Y.; Yasuda, T.; Maeda, S.; Higuchi, H.; Miyawaki, T. Liposome-encapsulated midazolam for oral administration. J. Liposome Res. 2011, 21, 166−172. (5) Karanth, H.; Murthy, R. S. R. pH-Sensitive liposomes-principle and application in cancer therapy. J. Pharm. Pharmacol. 2007, 59, 469− 483. (6) Guo, X.; Szoka, F. C. Chemical approaches to triggerable lipid vesicles for drug and gene delivery. Acc. Chem. Res. 2003, 36, 335−341. (7) Tsui, F. C.; Ojcius, D. M.; Hubbell, W. L. The intrinsic pKa values for phosphatidylserine and phosphatidylethanolamine in phosphatidylcholine host bilayers. Biophys. J. 1986, 49, 459−468. (8) Kanicky, J. R.; Shah, D. O. Effect of premicellar aggregation on the pKa of fatty acid soap solutions. Langmuir 2003, 19, 2034−2038. (9) Lebron-Paler, A.; Pemberton, J. E.; Becker, B. A.; Otto, W. H.; Larive, C. K.; Maier, R. M. Determination of the acid dissociation constant of the biosurfactant monorhamnolipid in aqueous solution by potentiometric and spectroscopic methods. Anal. Chem. 2006, 78, 7649−7658. (10) Zhang, J.; Fan, H.; Levorse, D. A.; Crocker, L. S. Ionization behavior of amino lipids for siRNA delivery: Determination of ionization constants, SAR, and the impact of lipid pK(a) on cationic lipid-biomembrane interactions. Langmuir 2011, 27, 1907−1914. (11) Fernandez, M. S.; Fromherz, P. Lipoid pH indicators as probes of electrical potential and polarity in micelles. J. Phys. Chem. 1977, 81, 1755−1761. (12) Egorova, E. M. Dissociation constants of lipid ionizable groups I. Corrected values for two anionic lipids. Colloids Surf., A 1998, 131, 7− 18. (13) Vazquez, J. L.; Berlanga, M.; Merino, S.; Domenech, O.; Vinas, M.; Montero, M. T.; Hernandez-Borrell, J. Determination by fluorometric titration of the ionization constants of ciprofloxacin in solution and in the presence of liposomes. Photochem. Photobiol. 2001, 73, 14−19. (14) Li, X.; Cooper, M. A. Measurement of drug lipophilicity and pK(a) using acoustics. Anal. Chem. 2012, 84, 2609−2613. (15) Cabral, D. J.; Hamilton, J. A.; Small, D. M. The ionization behavior of bile-acids in different aqueous environments. J. Lipid Res. 1986, 27, 334−343. (16) Lau, B.; Macdonald, P. M. Determination of the pK(a), of membrane-bound N,N-dimethylsphingosine using deuterium NMRspectroscopy. Biochim. Biophys. Acta, Biomembr. 1995, 1237, 37−42. (17) Esker, A. R.; Zhang, L.-H.; Olsen, C. E.; No, K.; Yu, H. Static and dynamic properties of calixarene monolayers at the air/water interface. 1. pH Effects with p-dioctadecanoylcalix[4]arene. Langmuir 1999, 15, 1716−1724. (18) Walsh, C. L.; Nguyen, J.; Szoka, F. C. Synthesis and characterization of novel zwitterionic lipids with pH-responsive biophysical properties. Chem. Commun. 2012, 48, 5575−5577. (19) Kroemer, G.; Pouyssegur, J. Tumor cell metabolism: cancer’s Achilles’ heel. Cancer Cell 2008, 13, 472−482. I

DOI: 10.1021/acs.langmuir.5b03258 Langmuir XXXX, XXX, XXX−XXX

Article

Langmuir (36) Chang, H.-I.; Yeh, M.-K. Clinical development of liposomebased drugs: Formulation, characterization, and therapeutic efficacy. Int. J. Nanomed. 2012, 7, 49−60. (37) Felice, B.; Prabhakaran, M. P.; Rodriguez, A. P.; Ramakrishna, S. Drug delivery vehicles on a nano-engineering perspective. Mater. Sci. Eng., C 2014, 41, 178−195. (38) Vllasaliu, D.; Fowler, R.; Stolnik, S. PEGylated nanomedicines: recent progress and remaining concerns. Expert Opin. Drug Delivery 2014, 11, 139−154. (39) Koynova, R.; MacDonald, R. C. Cationic O-ethylphosphatidylcholines and their lipoplexes: phase behavior aspects, structural organization and morphology. Biochim. Biophys. Acta, Biomembr. 2003, 1613, 39−48. (40) Tenchov, B. G.; Wang, L.; Koynova, R.; MacDonald, R. C. Modulation of a membrane lipid lamellar-nonlamellar phase transition by cationic lipids: A measure for transfection efficiency. Biochim. Biophys. Acta, Biomembr. 2008, 1778, 2405−2412. (41) Hoekstra, D.; de Boer, T.; Klappe, K.; Wilschut, J. Fluorescence method for measuring the kinetics of fusion between biological membranes. Biochemistry 1984, 23, 5675−5681. (42) Lasic, D. D.; Ceh, B.; Stuart, M. C. A.; Guo, L.; Frederik, P. M.; Barenholz, Y. Transmembrane gradient driven phase-transitions within vesicles - Lessons for drug-delivery. Biochim. Biophys. Acta, Biomembr. 1995, 1239, 145−156. (43) Ceh, B.; Lasic, D. D. A rigorous theory of remote loading of drugs into liposomes: Transmembrane potential and induced pHgradient loading and leakage of liposomes. J. Colloid Interface Sci. 1997, 185, 9−18. (44) Hope, M. J.; Cullis, P. R. Lipid asymmetry induced by transmembrane pH gradients in large unilamellar vesicles. J. Biol. Chem. 1987, 262, 4360−4366. (45) Hope, M. J.; Redelmeier, T. E.; Wong, K. F.; Rodrigueza, W.; Cullis, P. R. Phospholipid asymmetry in large unilamellar vesicles induced by transmembrane pH gradients. Biochemistry 1989, 28, 4181−4187. (46) Leenhouts, J. M.; De Kruijff, B. Membrane potential-driven translocation of a lipid-conjugated rhodamine. Biochim. Biophys. Acta, Biomembr. 1995, 1237, 121−6. (47) Melikyan, G. B.; Deriy, B. N.; Ok, D. C.; Cohen, F. S. Voltagedependent translocation of R18 and DiI across lipid bilayers leads to fluorescence changes. Biophys. J. 1996, 71, 2680−2691. (48) Olasz, A.; Mignon, P.; De Proft, F.; Veszpremi, T.; Geerlings, P. Effect of the p-p stacking interaction on the acidity of phenol. Chem. Phys. Lett. 2005, 407, 504−509.

J

DOI: 10.1021/acs.langmuir.5b03258 Langmuir XXXX, XXX, XXX−XXX