The Extent of Solubilization of Flurbiprofen and Ketoprofen by Cetyltrimethylammonium Micelles Using Semi-Equilibrium Dialysis

The partitioning of two non-steroidal 
anti-inflammatory drugs (NSAIDs), flurbiprofen and ketoprofen, into cationic 
cetyltrimethylammonium micelles was investigated using semi-equilibrium 
dialysis at 37℃ in phosphate buffered saline. The micellar-water 
solubilization equilibrium constants for both NSAIDs, in their deprotonated 
forms, were observed to decrease linearly with increasing mole fraction of drug 
in micelles. For flurbiprofen, the solubilization constant in the limit as mole 
fraction of drug in micelles approaches zero was found to be 11,200 (co = 1 M), while for ketoprofen the value was 1950 (co = 1 M). Using 1H-NMR and UV spectroscopic techniques, the locus of solubilization 
for ketoprofen was found to be towards the charged exterior of the micelles, in 
the Stern layer, whereas flurbiprofen was found to solubilize more in the 
micellar interior.


Introduction
Non-steroidal anti-inflammatory drugs (NSAIDs) are a classification of medication used primarily for their anti-inflammatory, antipyretic, and analgesic activities. Globally, NSAIDs are the most widely used pharmaceutical drug [1]. The principal mode of action for NSAIDs is inhibition of the cyclooxygenase enzymes, COX-1 and COX-2. Available evidence suggests that the analgesic and anti-inflammatory properties of NSAIDs arise from their inhibition of the COX-2 isoform, while common gastrointestinal side effects associated with their use are due to inhibition of COX-1. Recently, certain NSAIDs have also emerged as valuable drugs in the treatment or prevention of breast cancer [2], colorectal cancer [3], bone loss in postmenopausal women [4], and even certain neurodegenerative diseases [5]. Despite their prevalence, many NSAIDs have poor biopharmaceutical properties like low water solubility, rapid metabolism and excretion, and deleterious side effects [6]. Surfactant micelles have been shown to improve these biopharmaceutical properties when used as drug carriers [7], however few studies have investigated the physico-chemical interactions between NSAIDs and surfactants, especially as a function of micellar mole fraction of drug. Such studies are of interest to both the pharmaceutical sciences, because of micellar-based drug delivery systems and the use of surfactant systems as biological membrane mimics, and to the environmental sciences, because of the impact these drugs have on both terrestrial and marine ecosystems, and their potential removal using a surfactant-based remediation technology i.e. micellar-enhanced ultrafiltration (MEUF).
In aqueous solution, surfactant micelles are organized with hydrophobic tails aggregated in the core and hydrophilic heads comprising the exterior, mimicking structural features of biological membranes. This behavior makes micelles excellent tools for studying the complexity of membrane-drug interactions. A drug's therapeutic efficacy may be affected by the extent to which it binds to or solubilizes in pseudo-phase structures (i.e., micelles, membranes, or plasma proteins). The fraction of drug that is unbound and therefore able to diffuse into tissues and cross membranes is often very low when compared to the drug's total concentration. Differences observed in a drug's efficacy and toxicity, at similar total concentration levels, are often attributable to disparities in the drug's unbound fraction. The interactions of NSAIDs with cationic membrane mimics have been investigated using experimental methods for indomethacin with dihexadecyldimethylammonium bromide [8]; for piroxicam, meloxicam, and tenoxicam with cetyltrimethylammonium bromide (CTAB) [9]; and for diclofenac with small unilamellar vesicles [10]. The interactions of NSAIDs with lipid membranes have also been studied by molecular dynamics simulations [11].
In addition to serving as good mimics of biological membranes, surfactant micelles are also used as delivery vehicles for sparingly water-soluble drugs. Such drugs can partition into the hydrophobic interior of micelles, or may, in the case of oppositely charged surfactants, also electrostatically bind to the exterior of micelles. Advantages of surfactant micelles as delivery vehicles include their ability to: increase the drug's bioavailability; maintain a sustained release; and, protect drugs from degradation and metabolization [12]. The use of surfactant micelles as NSAID carries has been investigated using ibuprofen in anionic micelles [13]; naproxen and diclofenac in cationic micelles [14]; indomethacin in both anionic and cationic micellar assemblies [7]; and ibuprofen in amphiphilic polymers [2].
The partitioning of drugs into surfactant micelles can also be exploited through MEUF, a separation technique that could be employed to remove NSAIDs for contaminated waters. The global environmental impact of NSAIDs has only recently been assessed [15] [16]. Indications are that even low levels, μg·L −1 , of drugs like ketoprofen and diclofenac can have toxic effects on fish [17] and birds [18]. The MEUF process has been studied previously for the removal of various environmental contaminants, including haloacetates [19]; arsenic [20]; naphthenic acid [21]; chromium [22]; and nitrobenzene [23]. The efficiency of MEUF strategies is governed by the partitioning behavior of the target contaminants.
Micellar-water partition coefficients, which are also referenced as binding or solubilization constants, have been evaluated for many NSAIDs using a variety of techniques. Isothermal titration calorimetry (ITC) has been used to study the partitioning into surfactant micelles of diclofenac sodium [14]; of ibuprofen [13]; of indomethacin [7]; and of naproxen [24]. Other techniques that have been used to investigate partition coefficients for NSAIDs in surfactant-based systems include biopartitioning micellar chromatography [25], fluorescence anisotropy [26], and spectrophotometry [27]. A convenient expression for defining the partition coefficient, K, is provided in Equation (1), where X NSAID,mic is the mole fraction of drug in the micelles and c NSAID,bulk is the molar concentration of unsolubilized NSAID in bulk aqueous solution.
NSAID,mic NSAID,bulk The right-hand side of Equation (1) is easily evaluated using semi-equilibrium dialysis (SED) [28]. In the SED method, a membrane separates two compartments and retains micelles and solubilized/bound solute species on one side, while allowing monomeric surfactant and unsolubilized solutes to be transferred.
After a period of several hours a state of thermodynamic semi-equilibrium is achieved between unsolubilized solute on both sides of the membrane.
The SED technique has been used to investigate the partitioning of a variety of organic and inorganic solutes into micellar pseudophases [19]. Though not used extensively for the study of drug-micelle interactions, SED was employed to investigate the binding of sulfonamide antibiotics to CTAB micelles [31]. Advantages of SED compared to other methodologies include its relatively low cost and overall reliability. In addition, it is easy to design SED experiments to evaluate K as a function of X NSAID,mic , providing what are referred to as solubilization isotherms [32]. Such experiments often demonstrate that K is not a constant, but instead varies with the amount of solubilizate load incorporated in micelles. In this study, partition coefficients for the NSAIDs flurbiprofen and ketoprofen (see Figure 1) into CTAB micelles have been determined using SED. In addition, the interactions of these NSAIDs with CTAB micelles were examined by 1 H NMR spectrometry and UV spectrophotometry. Advances in Materials Physics and Chemistry

Materials
All materials utilized in this study were used as received from their respective manufacturers without further purification or modification.

SED Experiments
The

Analytical Methods
Concentrations of the NSAIDs were determined by UV absorption, using an Concentrations of CTAB in analyte solutions were determined using a methodology adapted from Adura et al. [33]. Briefly, to a 1 mL aliquot of analyte, 5 mL of CHCl 3 and 1 mL of bromophenol blue solution were added in a 16 × 150 mm glass test-tube. The resulting mixture was then vortexed for 30 minutes.
Next, the test-tubes were placed upright for 10 minutes. The organic layer was then removed using a Pasteur pipet and analyzed using an Agilent 8453 spectrophotometer, equipped with a 1 cm pathlength cuvette, at 606 nm.

Evaluation of K and X NSAID,mic
The critical micelle concentration (CMC) for CTAB in 1X PBS used in this study was 0.11 mM, the concentration of surfactant in the permeate after the equilibration period. Figure   mM NaCl at 25˚C [34] Finally, X NSAID,mic and K were calculated using Equations (2) and (3)

Solubilization Constants from SED
The partitioning results for flurbiprofen and ketoprofen are provided in Figure   4. For both NSAIDs, K is observed to decrease, almost linearly, with increasing X NSAID,mic . Across similar micellar mole fraction ranges, this trend has been observed for cresols with cetylpyridinium chloride (CPC) [35], for benzoate with CPC [36], and for phenol with CPC [37]. Dougherty and Berg [38] indicate that Advances in Materials Physics and Chemistry this linear relationship between K and X NSAID,mic suggests a Langmuir-type adsorption at cationic micellar surfaces.
In several studies utilizing the SED method to investigate the partitioning of polar solubilites in ionic micelles, an expression like that provided in Equation (4) has been shown to well represent the dependence of the solubilization constant on micellar composition for X NSAID,mic values less than 0.2 or 0.3. In Equation (4), K 0 is the value of the of the solubilization constant in the limit as X NSAID,mic approaches zero and B is obtained from the negative slope of the plot K vs X NSAID,mic divided by K 0 . The B parameter is an empirical constant that can be related to constants in the Langmuir adsorption equation [28] [39]. Values obtained for K 0 and B using the SED method, from this and previous studies are provided in Table 1.  The pKa values obtained at 0.1 M ionic strength in aqueous solution for flurbiprofen and ketoprofen are 4.50 [29] and 4.76 [30], respectively. While deprotonation constants are known to vary with surfactant concentration [40], in the case of cationic surfactants apparent pKa values decrease with increasing micelle concentration. Deprotonated forms dominate the speciation for both flurbiprofen and ketoprofen in 1× PBS and electrostatic attraction between negatively charged NSAIDs and cationic micelles certainly contributes to the extent of partitioning and the relatively large K and K 0 values. For comparison, K 0 (c˚ = 1 M) values of 833 and 786 were obtained for the neutral molecules p-bromophenol and p-chlorophenol in CPC micelles, respectively [41]. The larger K 0 for flurbiprofen compared to ketoprofen correlates with their respective octanol-water partition coefficients, provided in Figure 1, reflecting its greater lipophilicity.

Locus of Solubilization
Variations in the wavelength of maximum absorption, max λ , for flurbiprofen and ketoprofen as functions of CTAB concentration in aqueous solution at 37˚C are presented in Figure 5. For flurbiprofen, a bathochromic shift from 247 nm to 252 nm was observed. This red shift is attributable to hydrophobic interactions between π-electrons in the drug and alkyl tails of the surfactants, indicating that, for flurbiprofen, the locus of solubilization is more in the micellar interior.
Similar solubilization and corresponding electronic spectral behavior has been observed previously for crystal violet in CTAB micelles [43]. For ketoprofen, a slight hypsochromic shift from 260 nm to 257 nm was observed and indicates that electrostatic attraction between the drug's carboxylate group and the cationic micellar exterior plays a more significant role in partitioning. The locus of solubilization for ketoprofen is likely more towards the charged exterior of CTAB micelles, in the Stern layer. Interactions between methyl orange and dodecyltrimethylammonium bromide (DTAB) micelles display similar blue-shifted spectra and were attributed to incorporation of dye molecules into the shell of DTAB micelles [44]. Further insights into the locus of solubilization are depicted in Figure 6. A greater upfield shift of the proton on the α-carbon of flurbiprofen, about 0.29 ppm, was seen when compared to the proton on the α-carbon of ketoprofen, about 0.09 ppm. Large upfield shifts were observed by Vermathen et al. [45] for fluoro-substituted benzoates in tetradecyltrimethylammonium micelles and were attributed to proton location deeper inside the micellar palisade layer.
In addition to use in probing loci of solubilization, NMR techniques have also been employed to determine micellar solubilization constants [46]. In the future, both SED and NMR techniques will be used to investigate the partitioning of other NSAIDs into surfactant micelles, e.g., diclofenac, meclofenamic acid, diflunisal. Of particular interest will be the establishment of a methodology whereby NMR data can be used to assess solubilization constants as a function of micellar composition. Advances in Materials Physics and Chemistry

Conclusion
Semi-equilibrium dialysis was used to assess the partitioning of flurbiprofen and ketoprofen into CTAB micelles. The results were well correlated by the expression ( ) 0 NSAID,mic , with K 0 being larger for flurbiprofen than ketoprofen. A hypsochromic shift in the UV spectrum of ketoprofen indicates its lo-cus of solubilization is in the Stern layer of CTAB micelles, while the bathochromic shift in the spectrum of flurbiprofen suggests solubilization further into the micellar interior. These loci of solubilization were further supported by differences in chemical shifts in the 1 H-NMR spectra of both NSAIDs in the presence and absence of CTAB micelles. The extent of partitioning of both drugs into CTAB micelles suggests that MEUF separations are feasible for their removal from contaminated marine systems.