A Base-Triggerable Catanionic Mixed Lipid System: Isothermal

Oct 11, 2010 - A Base-Triggerable Catanionic Mixed Lipid System: Isothermal Titration Calorimetric and Single-Crystal X-ray Diffraction Studies ... Se...
2 downloads 7 Views 2MB Size
Download PDF
13710

J. Phys. Chem. B 2010, 114, 13710–13717

A Base-Triggerable Catanionic Mixed Lipid System: Isothermal Titration Calorimetric and Single-Crystal X-ray Diffraction Studies Pradip K. Tarafdar, S. Thirupathi Reddy, and Musti J. Swamy* School of Chemistry, UniVersity of Hyderabad, Hyderabad-500 046, India ReceiVed: May 27, 2010; ReVised Manuscript ReceiVed: August 16, 2010

Lipid-based, base-triggerable systems will be useful for colon specific targeted delivery of drugs and pharmaceuticals. In light of this, a catanionic surfactant system, composed of O-lauroylethanolamine hydrochloride (OLEA · HCl) and sodium dodecyl sulfate (SDS), has been designed. The aggregates formed by near equimolar mixtures of OLEA · HCl-SDS have shown lability at basic pH, indicating that the system may be useful for developing colon specific drug delivery system(s). Turbidimetric and isothermal titration calorimetric studies revealed that OLEA · HCl forms a 1:1 (mol/mol) complex with SDS. The three-dimensional structure of the equimolar OLEA-SDS complex has been solved by single-crystal X-ray diffraction. Analysis of the molecular packing and intermolecular interactions in the crystal lattice revealed a hydrogen bonding belt in the headgroup region of the complex and dispersion interactions among the acyl chains as the main factors stabilizing the complex. These observations will be useful in understanding specific interactions between lipids in more complex systems, e.g., biomembranes. Introduction Liposomal drug delivery systems that self-destruct upon application of a stimulus have drawn considerable interest among the biomedical community because of the temporal and spatial specificity afforded by them in the delivery of therapeutics. Liposomes which unload the encapsulated drugs upon the application of different types of stimuli such as acid, radiation, heat, and redox-potential have been designed earlier.1-4 In such systems, release of the drugs occurs either by significantly increasing the permeability of the liposomal membrane or by completely disrupting the supramolecular structure of the vesicular assembly under the condition of applied stimuli. In this context, base-triggerable systems attracted the attention of researchers due to their potential application in colon-specific drug delivery. The pH-responsive colon-specific systems are based on the reported gradual increase of the luminal pH along the small intestine (from 6.63 ( 0.53 in the jejunum) reaching a peak at the ileocecal junction (7.49 ( 0.46).5 The most commonly used system for colon-specific delivery is Eudragit S, a polymer which was designed to dissolve above pH 7. This was the first colon-specific pH-responsive delivery agent to be developed and is still being used for site-specific delivery of the anti-inflammatory drug mesalazine to the large intestine for the treatment of ulcerative colitis.6,7 However, so far there have been no reports on lipid-based, base-triggerable systems for colon-specific delivery. Mixtures of cationic and anionic single-chain lipids can spontaneously form a variety of phases, including vesicles.8,9 Such catanionic mixed lipid systems, also known as catanionic surfactants, have been subjected to extensive experimental studies, since they are easy to prepare and possess superior stability and versatile physicochemical properties.10 These features put them in contention as an alternative to conventional phospholipid vesicles for drug delivery applications. In view * To whom correspondence should be addressed. Tel.: +91-40-23134807. Fax: +91-40-2301-2460/0145. E-mail: [email protected], [email protected]. Website: http://202.41.85.161/∼mjs/.

of this, we have designed a base-triggerable catanionic mixed lipid system from a mixture of O-lauroylethanolamine hydrochloride (OLEA · HCl) and sodium dodecyl sulfate (SDS). The interactions between the lipids were investigated by turbidimetry and isothermal titration calorimetry (ITC). Dissociation of the aggregates formed by near equimolar mixtures of OLEA · HCl and SDS at high pH revealed that the system may be a promising candidate for developing base-triggerable drug delivery systems. Finally, the molecular structure and intermolecular interactions of the catanionic mixed lipid system and OLEA · HCl were determined by single-crystal X-ray diffraction. To the best of our knowledge, the present report is the first description on the determination of crystal structure of a mixed lipid system, although single-crystal X-ray diffraction has been employed for the structural characterization of a large number of lipids and surfactants over the last several decades. Details of the molecular packing and intermolecular interactions in the mixed lipid system, derived from the studies reported here, are expected to be useful in understanding specific interaction between different types of lipids in membrane microdomains. Experimental Methods Materials. Lauric acid, sodium dodecyl sulfate, N,N’dicyclohexylcarbodiimide (DCC), and 4-dimethylaminopyridine (DMAP) were purchased from Sigma-Aldrich (U.S.A.). Ammonium chloride (NH4Cl), sulfuric acid, and dioxane were obtained from Merck (Germany). Ethanolamine, BOC-anhydride (di-tert-butyl dicarbonate), solvents, and other chemicals used were of analytical grade and purchased locally. Milli-Q water was used in all experiments. Synthesis of O-Lauroylethanolamine Hydrochloride. OLauroylethanolamine hydrochloride (OLEA) was synthesized by a minor modification of a reported procedure,11 adopted earlier for the synthesis of O-stearoylethanolamine.12 In this procedure, the amino group of ethanolamine was protected using BOC-anhydride, and then the hydroxyl group was esterified by condensation with lauric acid using DCC as a coupling reagent. A catalytic amount of DMAP was used in condensation. The

10.1021/jp104841k  2010 American Chemical Society Published on Web 10/11/2010

Base-Triggerable Catanionic Mixed Lipid System BOC-protected ester was then deblocked with 4 M HCl in dioxane (prepared by dissolving dry HCl gas into dioxane) to get the hydrochloride salt of O-lauroylethanolamine. Turbidimetry. Turbidimetric measurements were performed at 25 °C using a Cary 100 UV-visible spectrophotometer (VARIAN). Aqueous solutions of OLEA and SDS (0.2 mM concentration each, which is well below the critical micellar concentration (cmc) of either surfactant) were mixed at different ratios, vortexed, and kept for 30 min. To investigate the effect of pH on the turbidity of OLEA/SDS (1:1) mixture, the aggregated catanionic lipid mixtures at near equimolar ratio of OLEA · HCl and SDS were hydrated with buffers of different pH and incubated for 30 min, and then the turbidity was measured. The following buffers were used: 20 mM KCl-HCl (pH 2.0), 20 mM citrate-phosphate (pH 3.0-6.0), 20 mM TrisHCl (pH 7.0-9.0), and 20 mM glycine-NaOH (pH 10.0). Turbidity was measured by recording the optical density from 350 to 450 nm, and turbidity at 400 mm was considered for further analysis. Isothermal Titration Calorimetry. An isothermal titration calorimeter (VP-ITC from MicroCal LLC, Northampton, MA) was used to measure the cmc of OLEA · HCl and to investigate thermodynamics of the interaction between OLEA · HCl and SDS at 25 °C. The cmc of OLEA · HCl was determined by measuring the enthalpy change resulting from injection of aliquots of 100 mM OLEA · HCl solution into water.13 To study the interaction between OLEA · HCl and SDS, 10 µL aliquots of 0.5 mM SDS solution in water were added via a rotating stirrer syringe to 1.445 mL of a 50 µM OLEA · HCl solution in water in the sample cell. It may be noted that these concentrations are 1 and 2 orders of magnitude lower, respectively, when compared to the cmc of SDS14 and OLEA (see below). Each injection lasted 20 s, and an interval of 240 s was given between successive injections. The solution in the reaction cell was stirred at a speed of 300 rev min-1 throughout the experiment. Usually the first injection was found to be inaccurate; therefore, a 1 or 2 µL injection was added first, and the resultant point was deleted before the remaining data were analyzed as described below. For a system of one set of interaction sites, the total heat evolved (or absorbed) during the binding process at the end of the ith injection, Q(i), is given by eq 115

Q(i) ) nPt∆HbV{1 + Xt/nPt + 1/nKbPt - [(1 + Xt/nPt + 1/nKbPt)2 - 4Xt/nPt]1/2}/2 (1) where n is the stoichiometry of binding, Pt is the total concentration of lipid in the cell, Xt is the concentration of lipid in the syringe, V is the cell volume, Kb is the binding constant, and ∆Hb is the binding enthalpy. The heat corresponding to the ith injection, ∆Q(i), is equal to the difference between Q(i) and Q(i - 1) and is given by eq 2, which involves the necessary correction factor for the displaced volume (the injection volume dVi):

∆Q(i) ) Q(i) + dVi /2V[Q(i) + Q(i - 1)] - Q(i - 1) (2) The ITC unit measures the ∆Q(i) value for every injection. These values are then fitted to eqs 1 and 2 by a nonlinear leastsquares method using the data analysis program Origin from MicroCal.16 The fit process involves initial guess of n, Kb, and ∆Hb, which allows calculation of ∆Q(i) values as mentioned

J. Phys. Chem. B, Vol. 114, No. 43, 2010 13711 above for all injections and comparison of them with the corresponding experimentally determined values. Based on this comparison, the initial guess of n, Kb, and ∆Hb is improved, and the process is repeated until no further significant improvement in the fit can be obtained. The thermodynamic parameters ∆Gb and ∆Sb are calculated according to the basic thermodynamic eqs 3 and 4:

∆Gb ) -RT ln Kb

(3)

∆Gb ) ∆Hb - T∆Sb

(4)

Crystallization, X-ray Diffraction, and Structure Solution. Equimolar O-lauroylethanolamine-dodecylsulfate (OLEA-DS) complex and OLEA · HCl alone were independently crystallized from dichloromethane containing a trace of methanol. While OLEA-DS complex yielded colorless, plate-type crystals, OLEA alone gave colorless, needle-shaped crystals. X-ray diffraction measurements were carried out at room temperature (ca. 25 °C) with a Bruker SMART APEX CCD area detector system using a graphite monochromator and Mo-KR (λ ) 0.710 73 Å) radiation obtained from a fine-focus sealed tube. Data reduction was done using Bruker SAINTPLUS program. Structure solution was carried out in the orthorhombic space group. Absorption correction was applied using the SADABS program. The structure of OLEA-DS complex was solved successfully by direct methods in the space group P - 1, and refinement was done by full matrix least-squares procedure using the SHELXL-97 program.17 The refinement converged into a final R1 ) 0.089, wR2 ) 0.187, and goodness of fit ) 1.04. The structure of OLEA · HCl was solved in the P212121 space group. The refinement converged into a final R1 ) 0.11, wR2 ) 0.286, and goodness of fit ) 1.03. Crystal Parameters of OLEA-DS. Molecular formula: C26H55NO6S. Molecular formula weight: 509.8. Crystals were plate type and colorless. Crystal system, triclinic; space group, Sg ) P - 1; ambient temperature, T ) 298(2) K; radiation wavelength (λ) ) 0.710 73 Å; minimum resolution ) 0.84 Å; radiation type, Mo-KR; radiation source, fine-focus sealed tube; radiation monochromator, graphite; number of reflections collected, 9110; unique reflections, 5199; reflection with I > 2σ(I), 3077; number of parameters, 310. Unit cell dimensions (with standard deviation in parentheses): a ) 5.478(1), b ) 7.272(1), c ) 37.771(5) Å; R ) 87.989(3), β ) 89.759(2), γ ) 88.604(2)°; volume of the cell, V ) 1503.3(4) Å3; number of molecules in the unit cell, Z ) 2; F(000) ) 564; absorption coefficient, µ ) 0.144 mm-1; T ) 298(2) K. Crystal Parameters of OLEA · HCl. Molecular formula: C14H30ClNO2. Molecular formula weight: 279.8. Crystals were needle type and colorless. Crystal system, orthorhombic; space group, Sg ) P212121; ambient temperature, T ) 298(2) K; radiation wavelength (λ) ) 0.710 73 Å; minimum resolution ) 0.84 Å; radiation type, Mo-KR; radiation source, fine-focus sealed tube; radiation monochromator, graphite; number of reflections collected, 7914; unique reflections, 2776; reflection with I > 2σ(I), 1507; number of parameters, 165. Unit cell dimensions (with standard deviation in parentheses): a ) 5.397(2), b ) 5.400 (2), c ) 60.66(2) Å; volume of the cell, V ) 1768.0(10) Å3; number of molecules in the unit cell, Z ) 4; F(000) ) 616; absorption coefficient, µ ) 0.213 mm-1; T ) 298(2) K. 1 H NMR and FTIR Spectroscopy. 1H NMR spectra were recorded on a Bruker Avance 400 MHz NMR spectrometer with

13712

J. Phys. Chem. B, Vol. 114, No. 43, 2010

Tarafdar et al.

a mixture of CDCl3 and CD3OD as the solvent. FTIR spectra were recorded in KBr pellets on a Jasco Model 5300 FTIR spectrometer. An equimolar mixture of OLEA and SDS was prepared by dissolving appropriate quantities of the two compounds in dichloromethane containing a trace of methanol, followed by evaporating the solvent under a stream of dry nitrogen gas and vacuum desiccation for 3 h. Small-Angle X-ray Scattering. SAXS measurements with OLEA · HCl and OLEA/SDS equimolar mixture were performed at 25 °C using a Hecus S3-Micro System (Graz, Austria) equipped with a one-dimensional position sensitive detector. OLEA · HCl and the equimolar OLEA/SDS mixture at 25 mM concentration dispersed in water were filled in sealed 1 mm diameter quartz capillaries and used for SAXS measurements. Data was collected for ca. 3-4 h, and the diffraction patterns were calibrated using silver behenate as the standard. Lattice Energy Calculations. Lattice energy calculations of OLEA · HCl and OLEA-DS complex were performed using the Dreiding force field in Cerius2 suite of programs (Accelrys Inc., San Diego, CA; web site: http://www.accelrys.com). Results and Discussion Synthesis and Characterization of O-Lauroylethanolamine. The synthetic protocol described under Experimental Methods gave OLEA · HCl in 57% yield. The product was characterized by 1H NMR and FTIR spectroscopy. The 1H NMR spectrum (Figure S1) gave the following resonances: 0.88 δ (3H, t), 1.25δ (18H, s), 1.61δ (2H, bs), 2.43δ (2H, t), 3.30δ (2H, m), 4.42δ (2H, t), 8.5δ (3H, bs). The FTIR spectrum (Figure S2) of OLEA · HCl contained strong absorption bands at 1740 cm-1 (ester carbonyl stretch), 2920 and 2851 cm-1 (C-H stretching), and 1167 cm-1 (C-O stretching), a broad band at ∼3000 cm-1 (N-H stretching, ammonium group), and two sharp bands of medium intensity at 1470 cm-1 (C-H bending) and 721 cm-1 (C-H rocking). These data are consistent with the structure of OLEA · HCl. Determination of the Cmc of O-Lauroylethanolamine. The critical micellar concentration of OLEA · HCl was determined using isothermal titration calorimetry by measuring the enthalpy changes resulting from titration of lipid solution into water at 25 °C. Calorimetric titrations were performed by the sequential injection of 10 µL aliquots from a 100 mM OLEA · HCl solution in water into a 1445 µL reaction cell containing water. The dependence of change in enthalpy per mole of OLEA · HCl (∆Hi) injected into the reaction cell on the lipid concentration in the reaction cell was calculated by integration of the heat flow versus time profiles (Figure 1). Initially, a series of relatively large endothermic peaks was observed when the lipid was injected into the reaction cell. These enthalpy changes are the result of micelle dissociation because the lipid concentration in the reaction cell was initially below the cmc.18 The endothermic nature of these peaks (∆H > 0) indicates that demicellization must lead to an increase in the overall entropy of the system, since micelle dissociation is thermodynamically favorable below the cmc (∆G < 0); therefore, T∆S > ∆H. This entropy increase can be attributed to the disruption of water structure when micelles break down to monomers.18 After a certain number of injections, there was an appreciable decrease in peak height because by then the surfactant concentration in the reaction cell exceeded the cmc and so the micelles titrated into the reaction cell no longer dissociated. The enthalpy change above the cmc is therefore solely due to micelle dilution effects. The cmc of OLEA · HCl was determined from the inflection point in the ∆Hi versus lipid concentration curves as 7.3 ( 0.2 mM.

Figure 1. Determination of cmc of OLEA · HCl by ITC. The upper panel shows raw data for the dilution of 100 mM OLEA · HCl into water. The lower panel shows the integrated data obtained from the raw data shown in the upper panel. The difference in enthalpy between the two sections of the curve at the cmc corresponds to the enthalpy of micellization. From the curve the enthalpy of micellization was obtained as 0.4 kcal · mol-1.

Figure 2. Turbidity of various mixtures of OLEA · HCl and SDS.

Turbidimetric Study of the Interaction of OLEA with SDS. The optically clear solution of OLEA · HCl becomes turbid upon addition of negatively charged sodium dodecyl sulfate (SDS). This suggests that the two lipids interact, resulting in the formation of larger aggregates. Since OLEA · HCl is cationic and SDS is anionic, we expect the formation of catanionic mixed lipid system upon addition of SDS. The formation of catanionic lipid system by various admixtures of the two lipids was monitored by turbidimetry. The turbidity of OLEA · HCl-SDS mixtures at constant overall concentration of 0.2 mM is shown in Figure 2. Initially the turbidity increases with increase in SDS concentration and reaches a maximum level near the XSDS value of 0.4-0.5. Above that turbidity again decreases with further increase in SDS concentration. The higher turbidity near the XSDS value of 0.5 suggests the formation of 1:1 (mol/mol) complex between two lipids. Since the aggregates are large enough to give significant turbidity to the sample, they are

Base-Triggerable Catanionic Mixed Lipid System

J. Phys. Chem. B, Vol. 114, No. 43, 2010 13713

Figure 4. Effect of pH on the turbidity of the aggregates formed by near equimolar mixtures of OLEA · HCl and SDS.

Figure 3. ITC profile of interaction between OLEA · HCl and SDS at 25 °C. The upper panel shows raw data for titration of 50 µM OLEA · HCl with 0.5 mM SDS. The lower panel shows the integrated data obtained from the raw data shown in the upper panel, after correcting for dilution effects. The solid line in the bottom panel represents the best curve fit to the experimental data, using the one set of sites model from MicroCal Origin.

unlikely to be micelles; instead they will most likely correspond to larger structures, e.g., multilamellar liposomes. Thermodynamics of OLEA-SDS Interaction. The interaction of OLEA with SDS was further investigated by ITC. Experiments were carried out at low concentration (below cmc) of both the lipids to avoid their self-aggregation, which sometimes may obstruct the data analysis. A representative calorimetric titration for the interaction of OLEA and SDS is given in Figure 3A. From this figure it is seen that the exothermic heat of binding decreases monotonically with successive injections until saturation is achieved. A plot of incremental heat released as a function of the ratio of SDS to OLEA · HCl is shown in Figure 3B, together with a nonlinear least-squares fit of the data to eq 1. The experimental data could be fitted satisfactorily to the one set of sites model available in the Origin software provided by the instrument manufacturer. The fit obtained for the data is shown as a solid line in Figure 3B. The fit yielded values of the following parameters: stoichiometry of binding, n ) 0.94 ( 0.12; binding constant Kb ) 1.37 ((0.09) × 106 M-1; enthalpy of binding, ∆Hb ) -21.8 ( 1.0 kcal · mol-1, and entropy of binding, ∆Sb ) -45.2 ( 3.75 cal · mol-1 · K-1. The Gibbs free energy, ∆Gb, was calculated using eq 4 as -8.33 kcal · mol-1. The stoichiometry of binding (n) is close to 1.0 within the range of experimental error and indicates that OLEA and SDS form a 1:1 complex. Since the equimolar complex is formed at concentrations that are well below the cmc of either of the two lipids, it appears that the first step of the formation of a catanionic mixed lipid system is the formation of a 1:1 complex, which then self-assembles to form an aggregated supramolecular structure above a critical aggregation concentration. Disruption of Aggregates at High pH. OLEA · HCl belongs to a class of lipids named O-acylethanolamines (OAEs), which are ester derivatives of ethanolamine. OAEs with palmitoyl and arachidonyl acyl chains have been reported to be unstable at

high pH and transform to the corresponding N-acylethanolamines via acyl chain migration.11 It is expected that like other OAEs, OLEA · HCl would also exhibit low stability at high pH. Therefore, catanionic lipid aggregates containing OLEA · HCl as the cationic lipid would be expected to decompose at high pH due to its conversion to N-lauroylethanolamine which may disturb the ionic as well as H-bonding interaction between the two oppositely charged lipids. Therefore, the system has considerable potential for use in drug delivery applications for the release of cargo at high pH, e.g., for delivery to colon. Figure 4 shows pH dependence of the turbidity of near equimolar OLEA · HCl-SDS mixtures, wherein concentrations of the two lipids were maintained constant in all the samples. It is seen from this figure that turbidity of the sample decreases at high pH, suggesting disruption of the aggregates with increase in pH. The aggregates exhibited a tendency to precipitate upon keeping for a long time and altering the lipid ratio from equimolar to near equimolar with slightly higher percentage of cationic lipid or anionic lipid did not significantly decrease the precipitation of the mixed lipid system. Tuning the structure of the two lipids may help in developing a more stable basetriggerable catanionic liposomal system. Description of the Structure. Molecular structure of the 1:1 (mol/mol) complex of O-lauroylethanolamine and dodecyl sulfate (OLEA-DS) is shown in the ORTEP given in Figure 5A, along with the atom numbering for all the non-hydrogen atoms. Bond distances, bond angles, and torsion angles involving all the non-hydrogen atoms are given in Tables S1 and S2. It is clearly seen from Figure 5A that the hydrocarbon portion (C1-C12) of the acyl chains of dodecyl sulfate (DS) and (C16-C26) of OLEA are in all-trans conformation. The torsion angles observed for the acyl chain region are all close to 180° (Table S2). The gauche conformation of the C1-O4 bond of DS results in a minor bend at the headgroup region, although the overall shape of the molecule is close to linearity. The alltrans conformation in the acyl chain portion of OLEA in the complex leads to a linear geometry for this molecule also. In the complex both the linear molecules are oriented appropriately, facilitating hydrogen bond interactions between their hydrophilic groups besides appreciable dispersion interaction between the hydrocarbon chains of the two molecules. The molecular structure of OLEA · HCl is shown in the ORTEP given in Figure 5B. Bond distances, bond angles, and torsion angles involving all the non-hydrogen atoms are given in Table S3. The gauche conformation at the terminal N1-C1 bond bends the amino group toward the chloride ion, although due to all-trans conformation of the hydrocarbon portion (C4-C14) of the acyl chains the overall geometry of the molecule is essentially linear. Comparison between the molec-

13714

J. Phys. Chem. B, Vol. 114, No. 43, 2010

Tarafdar et al.

Figure 5. (A) ORTEP showing the structure of OLEA-DS equimolar complex. (B) ORTEP showing the molecular structure of OLEA · HCl.

ular structure of OLEA · HCl alone and in complex with an anionic lipid suggests that the acyl chain conformation in both the crystal lattices is very similar. Molecular Packing. Packing diagrams of OLEA-DS complex along the a-axis and the b-axis (Figure 6A and 6B, respectively) show that the entire crystal lattice is made up of repeats of the OLEA-DS (1:1) complex. The molecules are packed in layers that are stacked in such a way that the methyl groups of the O-lauroyl chains of OLEA from one layer face the methyl groups of the O-lauroyl chains of the DS of the next layer; i.e., the chain packing is of the mixed type. The length of the DS molecule is ∼17.2 Å, which is slightly less than the length of OLEA (∼18.3 Å). This difference in size most likely favors the mixed type chain packing since symmetric packing would lead to voids in the crystal lattice, resulting in poor chain packing. The methyl ends of the stacked bilayers are in van der Waals contacts, with the closest methyl-methyl (C12-C26) distance between opposite layers and the same layer being 3.93 and 4.26 Å, respectively. The bilayer thickness in the crystal structure of OLEA-DS is 31.8 Å and the all-trans O-lauroyl chains of OLEA and DS are tilted by 23.8° and 26.6°, respectively, with respect to the normal to the respective methyl end planes. Packing diagrams of OLEA · HCl along the a-axis and the b-axis are given in Figures S3A and S3B, respectively. The OLEA · HCl molecules are packed in a head-to-head (and tailto-tail) manner in stacked bilayers. The methyl ends of the stacked bilayers are in van der Waals’ contacts, with the closest methyl-methyl contact distance (C14-C14) between the opposing layers and within the same layer being 4.08 and 5.39 Å, respectively. The bilayer thickness (N1-N1 distance) of OLEA · HCl is 34.3 Å, and the all-trans acyl chains are tilted by 42.2° with respect to bilayer normal. Subcell Packing. The different lateral packing modes, adopted by hydrocarbon chains in lipid crystals, are generally described by subcells that specify the relations between equivalent positions within the chain and its neighbors.19,20 Examination of the hydrocarbon chain packing in the acyl chains of OLEA-DS complex revealed that the hydrocarbon chains pack according to the orthorhombic type (O⊥), whereas the chain packing of OLEA · HCl was according to triclinic subcell (T//).

Figure 6. Packing diagrams of OLEA-DS: (A) view along the a-axis; (B) view along the b-axis. Color code: gray, carbon; red, oxygen; blue, nitrogen; yellow, sulfur.

This suggests that the chain packing of OLEA is different in the absence and in the presence of SDS, i.e., when complexed to SDS. Unit cell dimensions of the O⊥ subcell are a ) 4.96 Å, b ) 7.27 Å, and c ) 2.54 Å, whereas the corresponding values for the T// subcell are a ) 4.32 Å, b ) 5.40 Å, and c ) 2.52 Å. Hydrogen Bonding and Intermolecular Interactions. To understand the intermolecular interactions in the OLEA-DS mixed lipid system, the molecular packing in the crystal lattice was carefully examined from different angles. The observed hydrogen bonding pattern in the crystal lattice is shown in Figure 7. From Figure 7A it can be seen that each hydrogen atom of the ammonium group of OLEA is involved in a hydrogen bond with one of the oxygen atoms of the sulfate moiety of the dodecylsulfate anion. Thus, each OLEA molecule forms three such N-H · · · O type hydrogen bonds: two with the headgroups of adjacent DS molecules in the same layer and one with a DS molecule in the opposing layer. The hydrogen bond distances (H · · · O) and angles of the three types of N-H · · · O interactions

Base-Triggerable Catanionic Mixed Lipid System

J. Phys. Chem. B, Vol. 114, No. 43, 2010 13715

Figure 8. SAXS profiles of OLEA · HCl and OLEA · HCl-SDS equimolar mixture in water.

Figure 7. Hydrogen bonding pattern in the crystal lattice of OLEA-DS. (A) Close-up view displaying N-H · · · O type hydrogen bonding. (B) Hydrogen- bonded motif involving N-H · · · O interactions in the polar region. Color code: gray, carbon; red, oxygen; blue, nitrogen; yellow, sulfur.

are 1.929, 2.02, 2.03 Å and 168.8°, 172.9°, 160.9°, respectively. The hydrogen-bonded network in the headgroup region displays a pleated-sheet-like form as shown in Figure 7B. Each pleatlike region contains two ammonium groups and two sulfate moieties in alternate positions, with a total of four hydrogen bonds formed between them. The observed hydrogen bonding pattern in the crystal lattice of OLEA · HCl shows that each chloride ion is hydrogen bonded to three N-H hydrogens (Figure S4A); while two of these H-bonds are formed with adjacent molecules in the same layer, the third H-bond is formed with the N-H group of a OLEA molecule from the opposite layer. Thus the chloride ion effectively bridges the two opposing layers of the bilayer structure. The three hydrogen bonds formed by each chloride ion are distinctly different, with the hydrogen bond (H · · · Cl) distances and angles corresponding to the three N-H · · · Cl interactions being 2.29, 2.30, 2.28 Å and 168.0°, 158.1°, 172.4°, respectively. The three types of N-H · · · Cl interactions give rise to a hydrogen bonding motif in the headgroup region, which is topologically similar to the arrangement observed in super black phosphorus (Figure S4B).21 FTIR Spectroscopy. In order to investigate the effect of complexation between OLEA · HCl and SDS on the vibrational bands corresponding to the functional groups of these two molecules, we obtained FTIR spectra of OLEA, SDS, and their equimolar mixture (Figures S2, S5, and S6). Shifting of the stretching bands corresponding to N-H and CdO groups in OLEA · HCl (seen at ca. 3000 and 1740 cm-1) to 3090 and 1738 cm-1, respectively, is consistent with the change in the hydrogen bonding pattern due to complexation. Similarly, two strong absorption bands at 1250 and 1219 cm-1 in the spectrum of SDS, corresponding to the stretching modes of the sulfate moiety, coalesce in the complex and shift to 1209.5 cm-1, which is also consistent with the complex formation. The sharp bands of medium intensity seen at 1470 and 721 cm-1 in the FTIR spectrum of OLEA · HCl, which correspond to CH2 bending and rocking vibrations, respectively, are split

into doublets in the spectrum of OLEA-DS complex (seen at 1474and 729 cm-1, respectively). This is consistent with the orthorhombic (O⊥) chain packing in the complex, as inferred from the crystal structure of the complex. Similar observations have been made with respect to the CH2 bending and rocking vibrations in several other lipid molecules which also have O⊥ chain packing.22-26 Lattice Energy Calculation. Lattice energies of OLEA-DS complex and OLEA · HCl were computed using the Dreiding force fields in the Cerius2 program package in order to estimate the lattice energies. The calculations yielded an overall energy of -124.7 kcal · mol-1 for OLEA-DS complex and -99.4 kcal/ mol for OLEA · HCl. Therefore the equimolar complex is thermodynamically more stable than the single tail lipid. The higher lattice energy of the equimolar complex may be responsible for the high Krafft temperature of the system (see below under Structure and Phase Behavior). Small-Angle X-ray Scattering. Figure 8 gives SAXS profiles of OLEA · HCl and an equimolar mixture of OLEA · HCl and SDS dispersed in water. While OLEA · HCl gave a broad scatter that is consistent with micellar structure, the equimolar mixture gave a single sharp reflection over the scattering range studied. Although the single reflection cannot be assigned unambiguously as due to lamellar structure, it is likely that the phase is lamellar in view of the lamellar structure of the OLEA-DS complex in the crystalline state, as discussed above. The SAXS data presented in Figure 8 gives a repeat spacing of 39.2 Å for the OLEA/SDS equimolar mixture at 25 °C, which is higher than the value of 31.8 Å obtained for the bilayer thickness from the crystal structure determination. The increment in the thickness of the hydrated sample can be assigned to the thickness of the water layer separating adjacent bilayers and indicates substantial hydration of the headgroup region. This suggests that the hydrated complex most likely adopts a core-shell structure, with the hydrated charged groups forming the shell and hydrophobic acyl chains forming the core. It may be noted here that SAXS measurements on another O-acylethanolamine, O-stearoylethanolamine as well as an N-acylethanolamine, N-palmotoylethanolamine also yielded a single sharp reflection in SAXS studies, which was interpreted as arising from the lamellar phase.12,27 Similarly, SAXS studies on several other surfactant systems also yielded single sharp reflections, which were assigned to the lamellar structure.28,29 The higher order peaks may be absent due to low electron

13716

J. Phys. Chem. B, Vol. 114, No. 43, 2010

density of the headgroup of OAEs and NAEs as compared to phospholipids where electron-rich phosphorus is present. Also, the second-order reflection could be missing due to the phasing problem.30 Structure and Phase Behavior. Biological membranes are composed of a variety of different lipids, and an understanding of the intermolecular interaction among them will be of immense value to rationalize their structure-activity relationship. Among the different techniques available for investigating the intermolecular interactions, single-crystal X-ray diffraction is the most powerful since it provides precise information on the structure and packing of molecules at atomic resolution. Although the crystal structures of various natural and synthetic lipids have been solved successfully, to the best of our knowledge there are no reports where the structure and packing of a mixed lipid system have been analyzed. This could be due to the difficulty in growing high-quality single crystals of mixed lipid systems, required for X-ray diffraction studies. In addition, the diffraction patterns of lipids are dominated by strong subcell reflections arising from the regular hydrocarbon chain matrix, which complicate or sometimes hinder the structure solution.22 Here, we have solved the structure of a mixed lipid system, and the intermolecular interactions in the crystal have been analyzed. Identification of a hydrogen bonding belt between the headgroups and dispersion interactions between the acyl chains in the crystal lattice of the mixed lipid system are the most significant observations derived from an analysis of the crystal structure. It is possible that similar intermolecular interactions will be found in other mixed lipid systems, e.g., lysophosphatidylcholine and fatty acid, lysophosphatidylcholine and cholesterol, N-myristoylethanolamine and cholesterol, and sphingomyelin and cholesterol, which have been reported to form stoichiometric complexes.31-34 X-ray diffraction studies on the crystals of the complexes formed by these lipid mixtures are necessary to obtain details of specific interactions between the different molecules involved in the complexes. The catanionic mixed lipid aggregates of OLEA · HCl and SDS have shown lability under basic conditions, and hence this system was expected to be a potential candidate for developing drug delivery vehicles targeting colon. Since the mixed lipid aggregates tend to precipitate near equimolar mixture, which could be due to the high Krafft temperature of the mixture, the usefulness of this system for drug delivery application may be limited. The Krafft temperature is recognized as the temperature above which the solubility of a surfactant/lipid increases dramatically in aqueous systems, and interpreted as the melting temperature of a hydrated solid surfactant.35,36 The insolubilization of the lipids or surfactants below Krafft temperature is usually accompanied by the crystallization of alkyl chains of the surfactant molecules inducing the formation of hydrated solids in water. Therefore change in Krafft temperature can be regarded as a change in the stability of hydrated solid, and it is expected that crystallized lipids with higher lattice energy may possess higher Krafft temperature. The Krafft temperature of OLEA · HCl is well below room temperature, and the corresponding temperature for the equimolar mixture is much higher. The calculated lattice energies show that the complex possesses more lattice energy than OLEA · HCl alone. Therefore in the complex the acyl chains are more tightly packed than in OLEA · HCl. Further support for this comes from the packing coefficient, which was 65.6% for the complex and 60.7% for OLEA · HCl, suggesting that the acyl chains in the complex are packed more tightly than in OLEA · HCl. It is expected that structural modifications in the acyl chain region such as

Tarafdar et al. branching or introduction of unsaturation, or complex formation between surfactants with mismatched chains, will weaken the acyl chain packing, resulting in a lowering of the Krafft temperature of the system. Further studies using OAEs with different acyl chains and other anionic surfactants may yield better catanionic lipid compositions for the development of new base-triggerable lipid systems for targeting colon. In summary, in the present study, the interaction between OLEA · HCl and SDS has been investigated by isothermal titration calorimetry and turbidimetry, which provided strong evidence for the formation of an equimolar complex between the two lipids. Disruption of aggregates formed by near equimolar mixture of OLEA · HCl and SDS at high pH suggests that the present study can serve as a springboard for the design of novel base-triggerable systems for practical applications, e.g., in delivery of drugs and pharmaceuticals to organs/organelles that are at basic pH such as colon. Analysis of the crystal structure of the mixed lipid system shows that the complex between OLEA and SDS is stabilized by a hydrogen-bonded belt between the headgroups and strong dispersive interactions between the acyl chains. The packing and intermolecular interaction between two lipids can serve as a model to understand the specific interaction among the lipids in more complex systems, e.g., in biomembranes. Acknowledgment. This work was supported by the Centre for Nanotechnology established by the Department of Science and Technology (India) at the University of Hyderabad in which M.J.S. is a coinvestigator. P.K.T. and S.T.R. were supported by Senior and Junior Research Fellowships, respectively, from the Council of Scientific and Industrial Research (India). Use of the National Single Crystal Diffractometer Facility (SMART APEX CCD single-crystal X-ray diffractometer) at the School of Chemistry, University of Hyderabad, funded by Department of Science and Technology (India), is gratefully acknowledged. The Hecus S3-Micro SWAXS equipment used in this study was funded by the Department of Science and Technology (India) under its FIST program. We thank the University Grants Commission (India) for their support through the UPE (to University of Hyderabad), and CAS (to School of Chemistry) programs. The authors are grateful to Prof. Samudranil Pal of this school for advice in X-ray data analysis. Supporting Information Available: Bond distances, bond angles, and torsion angles observed in the crystal structures of OLEA-SDS complex and OLEA · HCl (Tables S1-S3); 1H NMR and FTIR spectra of OLEA · HCl (Figures S1 and S2); molecular packing and hydrogen bonding pattern observed in OLEA · HCl (Figures S3 and S4); FTIR spectra of SDS and OLEA-DS complex (Figures S5 and S6); detailed data for the crystal structures submitted in the electronic Crystallographic Information File (CIF) format. This material is available free of charge via the Internet at http://pubs.acs.org. References and Notes (1) Huang, Z.; Guo, X.; Li, W.; MacKay, J. A.; Szoka, F. C., Jr. J. Am. Chem. Soc. 2006, 128, 60–61. (2) Shum, P.; Kim, J.-M.; Thompson, D. H. AdV. Drug DeliVery ReV. 2001, 53, 273–284. (3) Needham, D.; Dewhirst, M. W. AdV. Drug DeliVery ReV. 2001, 53, 285. (4) Ong, W.; Yang, Y.; Cruciano, A. C.; McCarley, R. L. J. Am. Chem. Soc. 2008, 130, 14739–14744. (5) Evans, D. F.; Pye, G.; Bramley, R.; Clark, A. G.; Dyson, T. J.; Hardcastle, J. D. Gut 1988, 29, 1035–1041. (6) Dew, M. J.; Hughes, P. J.; Lee, M. G.; Evans, B. K.; Rhodes, J. Br. J. Clin. Pharmacol. 1982, 14, 405–408.

Base-Triggerable Catanionic Mixed Lipid System (7) McConnell, E. L.; Short, M. D.; Basit, A. W. J. Controlled Release 2008, 130, 154–160. (8) Kaler, E. W.; Murthy, A. K.; Rodriguez, B. E.; Zasadzinski, J. A. Science 1989, 245, 1371–1374. (9) Dubois, M.; Deme´, B.; Gulik-Krzywicki, T.; Dedieu, J.-C.; Vautrin, C.; De´sert, S.; Perez, E.; Zemb, T. Nature 2001, 411, 672–675. (10) Lioi, S. B.; Wang, X.; Islam, M. R.; Danoff, E. J.; English, D. S. Phys. Chem. Chem. Phys. 2009, 11, 9315–9325. (11) Markey, S. P.; Dudding, T.; Wang, T. L. J. Lipid Res. 2000, 41, 657–662. (12) Tarafdar, P. K.; Swamy, M. J. Biochim. Biophys. Acta 2010, 1798, 872–881. (13) Heerklotz, H.; Seelig, J. Biochim. Biophys. Acta 2000, 1508, 69– 85. (14) Umlong, I. M.; Ismail, K. Colloids Surf., A 2007, 299, 8–14. (15) Wiseman, T.; Williston, S.; Brandts, J. F.; Lin, L. N. Anal. Biochem. 1989, 179, 131–137. (16) ITC data analysis in Origin, Tutorial guide 7.0; Microcal: Northampton, MA, 2002; p 99. (17) Sheldrick, G. M. SHELXL97 program for the refinement of crystal structures; University of Go¨ttingen: Go¨ttingen, Germany, 1997. (18) Bijma, K.; Engberts, J. B. F. N.; Blandamer, M. J.; Cullis, P. M.; Last, P. M.; Irlam, K. D.; Soldi, L. G. J. Chem. Soc., Faraday Trans. 1997, 93, 1579–1584. (19) Abrahamsson, S.; Dahle´n, B.; Lo¨fgren, H.; Pascher, I. Prog. Chem. Fats Lipids 1978, 16, 125–143. (20) Maulik, P. R.; Ruocco, M. J.; Shipley, G. G. Chem. Phys. Lipids 1990, 56, 123–133. (21) Ermer, O.; Eling, A. J. Chem. Soc., Perkin Trans. 2 1994, 925– 944.

J. Phys. Chem. B, Vol. 114, No. 43, 2010 13717 (22) Pascher, I.; Lundmark, M.; Nyholm, P.-G.; Sundell, S. Biochim. Biophys. Acta 1992, 1113, 339–373. (23) Casal, H. L.; Mantsch, H. H.; Cameron, D. G.; Snyder, R. G. J. Chem. Phys. 1982, 77, 2825–2830. (24) Kodali, D. R.; Atkinson, D.; Small, D. M. J. Lipid Res. 1990, 31, 1853–1864. (25) Akita, C.; Kawaguchi, T.; Kaneko, F. J. Phys. Chem. B. 2006, 110, 4346–4353. (26) Kamlekar, R. K.; Tarafdar, P. K.; Swamy, M. J. J. Lipid Res. 2010, 51, 42–52. (27) Swamy, M. J.; Ramakrishnan, M.; Wu¨rz, U.; Marsh, D. Biochim. Biophys. Acta 2003, 1616, 174–183. (28) Yamashita, I.; Kawabata, Y.; Kato, T.; Hato, M.; Minamikawa, H. Colloids Surf., A 2004, 250, 485–490. (29) Aswal, V. K.; Goyal, P. S.; De, S.; Bhattacharya, S.; Amenitsch, H.; Bernstorff, S. Chem. Phys. Lett. 2000, 329, 336–340. (30) Harper, P. E.; Mannock, D. A.; Lewis, R. N. A. H.; McElhaney, R. N.; Gruner, S. M. Biophys. J. 2001, 81, 2693–2706. (31) Jain, M. K.; Van Echteld, C. J. A.; Ramirez, F.; De Gier, J.; De Haas, G. H.; Van Deenen, L. L. M. Nature 1980, 284, 486–487. (32) Ramsammy, L. S.; Brockerhoff, H. J. Biol. Chem. 1982, 257, 3570– 3574. (33) Ramakrishnan, M.; Kenoth, R.; Ravikanth, R. K.; Chandra, M. S.; Radhakrishnan, T. P.; Swamy, M. J. FEBS Lett. 2002, 531, 343–347. (34) Slotte, J. P. Chem. Phys. Lipids 1999, 102, 13–27. (35) Shinoda, K. Colloidal Surfactants; Academic Press: New York, 1963; p 7. (36) Tsujii, K.; Mino, J. J. Phys. Chem. 1978, 82, 1610–1614.

JP104841K