Human Cytotoxicity, Hemolytic Activity, Anti-Inflammatory Activity and Aqueous Solubility of Ibuprofen-Based Ionic Liquids

: Ionic liquids (ILs) are a potential solution to the general problem of low solubility, polymorphism and low bioavailability of active pharmaceutical ingredients (APIs). In this work, we report on the synthesis of three pharmaceutically active ILs (API-ILs) based on ibuprofen, one of the most commonly available over-the-counter nonsteroidal anti-inflammatory drugs (NSAIDs), with imidazolium cations ([C 2 C 1 Im][Ibu] and [C 2(OH) C1Im][Ibu]) and a cholinium cation ([N 1112(OH) ][Ibu]). An upgrade to the aqueous solubility (water and biological simulated fluids) for the ibuprofen-based ILs relative to the ibuprofen’s neutral and salt form (sodium ibuprofen) was verified. The cytotoxic profiles of the synthesized API-ILs were characterized using two human cells lines, Caco-2 colon carcinoma cells and HepG-2 hepatocellular carcinoma cells, up to ibuprofen’s maximum plasma concentration (C max ) without impairing their cytotoxicity response. Additionally, the EC 50 in the Caco-2 cell line revealed similar results for both parent APIs and API-ILs. The biocompatibility of the ibuprofen-based ILs was also evaluated through a hemolytic activity assay, and the results showed that all the ILs were hemocompatible at concentrations higher than the ibuprofen C max . Moreover, the anti-inflammatory properties of the API-ILs were assessed through the inhibition of bovine serum albumin (BSA) denaturation and inhibition of cyclooxygenases (COX-1 and COX-2). The results showed that [C 2 C 1 Im][Ibu], [C 2(OH) C1Im][Ibu] and [N 1112(OH) ][Ibu] maintained their anti-inflammatory response to ibuprofen, with improved selectivity towards COX-2, allowing the development of safer NSAIDs and the recognition of new avenues for selective COX-2 inhibitors in cancer chemotherapy and neurological diseases such as Alzheimer’s and Parkinson’s.


Introduction
The pharmaceutical industry has resorted to several strategies to improve drugs that present major physicochemical problems such as polymorphism, low solubility and low bioavailability [1]. Since most pharmaceutical compounds are obtained in a crystalline solid form, polymorphism often represents a crucial complication. Polymorphism can be defined as the ability of a solid material to exist in two or more crystalline forms; although chemically the same, the polymorphs present different lattice structures and, consequently, different physicochemical and biological properties. Thus, polymorphism can significantly influence the properties of an active pharmaceutical ingredient (API) such as the dissolution rate, solubility, stability and, therefore, its bioavailability [2]. Strategies, such as solid dispersions, cyclodextrin inclusions, nanoemulsions, micelles and nanoparticles, were widely adopted to deliver poorly water-soluble drugs [3]. However, one of the most used approaches by pharmaceutical companies to overcome the low bioavailability problem is to redesign drugs in their respective salt forms [4], for example, in the context of nonsteroidal anti-inflammatory drugs (NSAIDs), the commercially used sodium ibuprofen, sodium naproxen and sodium diclofenac.
Ionic liquids (ILs), in agreement with the usually accepted definition, are salts comprising cations and anions with a melting point below the conventional temperature of 100 °C [5]. The large number of possible cation/anion combinations allows for a great variety of tunable properties. The key physicochemical properties, such as melting point, density, viscosity, surface tension, thermal stability, solubility, hygroscopicity and even toxicity and biodegradability [6][7][8], can be tailored, making ILs task-specific designer materials suitable for the biotechnology and pharmaceutical areas [9,10]. The development of ILs based on active pharmaceutical ingredients (API-ILs) has arisen as a possible solution to some of the problems that crystalline solid APIs present [1,11], leading to APIs with lower melting temperatures and improved bioavailability [4].
For example, the API-IL approach has been applied to increase the solubility in aqueous media (water and simulated biological fluids) of the APIs nalidixic acid and niflumic acid by their conversion into cholinium-based ILs ([N1112(OH)][Nal] and [N1112(OH)][Nif]) successfully, allowing for increases of 3300-fold and 53,000-fold of the solubility in water at 25 °C, respectively. Furthermore, an in vitro study on two human cell lines, Caco-2 colon carcinoma cells and HepG2 hepatocellular carcinoma cells, revealed that the cytotoxicity of these APIs was preserved upon their conversion into ILs [12]. Fernández-Stefanuto and coworkers synthesized alverine-based ILs with a water solubility up to 39,937-fold higher than the parent API, a widely known smooth muscle relaxant. Additionally, they managed to synthesize API-ILs that were liquid at room temperature, avoiding any limitations related to polymorphism [13]. In addition, Florindo  , and an enhancement in the octanol-water partition coefficient was verified [14]. Although an API is converted into an IL, the API-ILs platform allows to maintain or even improve the pharmacological profile of the parent API. Zhao et al. developed a cholinium-based derivative of betulinic acid with higher solubility in water than betulinic acid (by 100-fold) and improved the biological activity for inhibition of HIV-1 protease [15]. Moreover, Demurtas and coworkers developed cholinium-based ILs from hydroxycinnamic acids with higher solubility and free radical scavenging activity than the parent hydroxycinnamic acids as well as negligible cytotoxicity activity [16].
NSAIDs are the most used pharmaceutical ingredients for inflammation relief [17]. Their main action is through the inhibition of cyclooxygenase (COX), minimizing the production of prostaglandins, which is responsible for pain and inflammation responses [18]. Most NSAIDs are poorly water-soluble, which is a major drawback when envisaging their incorporation into hydrophilic matrices for drug release or their bioavailability [19]. The oral administration of NSAIDs is currently the most used route [20]; however, this type of administration presents several drawbacks including possible side effects such as gastrointestinal [21] and renal [18] toxicity. Therefore, whenever possible, topical administration of anti-inflammatory drugs is an efficient way to reduce some of the side effects. The therapeutic anti-inflammatory action of NSAIDs is produced by the inhibition of COX-2, while the undesired side effects arise from inhibition of COX-1 activity. Thus, more selective COX-2 inhibitors have reduced side effects [22].
The API-ILs platform has already been implemented to tackle some of the NSAIDs' drawbacks. Wu et al. successfully improved ibuprofen aqueous solubility by its conversion into imizazolium-and phosphomium-based ILs [23], and Santos and coworkers [24] prepared a set of ibuprofen-based ILs with different organic cations (e.g., cholinium, imidazolium and acetylpyridine) with increased solubility in aqueous media and negligible cytotoxicity towards human dermal fibroblasts and ovarian carcinoma cells. Stocker and coworkers [25] successfully spray-dried an imidazolium-based ibuprofen IL into a polymer carrier in loadings of up to 75% w/w in order to transform it into a solid powder suitable for oral solid dosage formulation, and they demonstrated that aqueous solutions of this API-IL has the potential to offer thermodynamic stability upon release, avoiding in vivo recrystallization issues that can limit the bioavailability of amorphous solid dispersions and some high-energy crystalline forms. Chantereau et al. [26] increased the solubility in aqueous media (up to 100-fold) of different NSAIDs, such as ibuprofen, naproxen and ketoprofen, via the preparation of cholinium-based ILs, and incorporated these API-ILs into bacterial nanocellulose membranes envisaging their use in transdermal drug delivery systems. Ibuprofen-based ILs conjugated with pyrrolidonium and cholinium cations were prepared by Moshidur et al. [27] as effective biocompatible formulations for topical drug delivery. Abednejad and coworkers [28] exploited the dual nature of ILs by combining an analgesic licocainium cation with anti-inflammatory NSAID-based anions, namely, ibuprofen, naproxen and diclofenac, obtaining liquid API-ILs at 25 °C with higher solubility in aqueous media and higher permeation than the parent APIs, without impairing their anti-inflammatory activity. The dual nature of ILs was also explored by Panic' et al. [29], who modified the local anesthetic drug, procaine, into ionic liquids combined with distinct pharmaceutically active anions such as ibuprofenate, salicylate and docusate. Dual pharmaceutically active protic ILs containing NSAIDs (i.e., ibuprofen or naproxen) as anions and diphenhydramine, an H1-receptor antagonist, as the cation were developed by Wang and coworkers [30] and loaded into a mesoporous carrier for novel formulation development. Recently, ibuprofen-based ILs with diphenhydramine and ranitidine cations were prepared by Frizzo et al. [31], and antifungal activity that was not present in the precursor salts was observed in the ILs. Wust and coworkers [32] showed that the API-IL, diphenhydraminium ibuprofenate, supported onto mesoporous silicas, can be obtained through a simple and efficient process (i.e., adsorption from solution), and they determined the release profiles of the supported API-IL, which were dependent on the pore size of the silicas. Additionally, the API-IL had an antinociceptive effect greater than sodium ibuprofen; thus, the authors concluded that the diphenhydramine potentiates the antinociceptive effect of ibuprofen. Further, the development of ibuprofen-based API-ILs conjugated with amino acid derivatives were also attained in different works for improved solubility and permeability of the parent API [33,34].
Herein, we report on the synthesis, characterization and thermal properties of three API-ILs based on the imidazolium and cholinium cations with the ibuprofenate anion: , ibuprofen and sodium ibuprofen) at 25 °C were evaluated. The studied buffer solutions included simulated gastric fluid without enzymes (interchangeable with 0.1 N HCl pH 1.0), simulated intestinal fluid without enzymes (interchangeable with phosphate standard buffer pH 6.8) and 0.15 M NaCl (isotonic ionic strength solution). Additionally, to ascertain the biocompatibility of the ibuprofen-based ILs, we performed in vitro cytotoxicity assays with two different human cells lines, namely, Caco-2 colon carcinoma cells and HepG2 hepatocellular carcinoma cells, as well as hemocompatibility assays. Finally, the anti-inflammatory activity of the API-ILs was assessed by the inhibition of bovine serum albumin (BSA) denaturation and the inhibition of cyclooxygenases (COX-1 and COX-2) enzymes using a colorimetric COX (COX-2, human; COX-1, ovine) inhibitor screening assay kit, to evaluate the potential of the API-ILs platform to maintain/upgrade the pharmacological activity of the parent API and to improve selectivity towards COX-2.
Through a two-step anion exchange reaction methodology proposed in previous works [6,12], three ibuprofen-based ionic liquids were synthesized: 1-ethyl-3-methylimidazolium ibuprofenate ( ). Neat ibuprofen-based ionic liquids were obtained after eliminating the excess water and API by evaporation and washing, respectively. Their chemical structures and respective acronyms are shown in Table 1. More experimental details on the synthesis can be found in the Supplementary Materials. Additionally, the ionic liquids were characterized by 1 H and 13 C NMR and elemental analysis in order to examine their expected structures and final purities. The water content, determined by Karl Fischer titration, was less than 0.05 wt%. The prepared ibuprofen-based ionic liquids were further characterized by differential scanning calorimetry (DSC). In all experiments throughout the work, Milli-Q water (Milli-Q Integral Water Purification System, Merck, Darmstadt, Germany) was used. The simulated gastric fluid without enzymes (interchangeable with 0.1 N HCl at pH 1.0) was acquired from Carlo Erba Reagents, and the simulated intestinal fluid without enzymes (interchangeable with phosphate standard buffer at pH 6.8) from Honeywell. The sodium chloride physiological solution (0.15 M NaCl isotonic ionic strength) was purchased from Sigma-Aldrich. Tablets for the phosphate-buffered saline (PBS) solution preparation were purchased from Pan-Reac Applichem ITW Reagents Division (Chicago, IL, USA).

Nuclear Magnetic Resonance (NMR)
The prepared ibuprofen-based ILs were completely characterized by 1 H and 13 C NMR (Bruker Avance III 400) in order to determine their expected structures. The integration of the API-ILs characteristic 1 H NMR and 13 C NMR resonance peaks confirmed the expected cation/anion ratios. All characterizations are depicted in Supplementary Materials.

Differential Scanning Calorimetry (DSC)
The experiments were performed using a TA Instrument DSC Q200 Differential Scanning Calorimeter. Cooling was accomplished using a refrigerated cooling system capable of controlling the temperature down to −90 °C. The sample was continuously purged with 50 mL·min −1 nitrogen. Approximately 5 mg of sample was crimped in a standard aluminum hermetic sample pan. Indium (Tm= 157.61 °C) was used as the standard compound for the calibration of the DSC. The samples were cooled to −90 °C, tempered (30 min) and, finally, heated to 100 °C. The cooling-heating cycles were repeated three times at different rates (i.e., 10, 5 and 1 °C/min). The transition temperatures obtained from the second and subsequent cycles at the same rate were reproducible. The DSC curves of ibuprofen,

Solubility Studies
An excess of ibuprofen, ibuprofen sodium salt and ibuprofen-based ILs were added in 1.5 mL safe-lock microtubes with 1.5 mL of each solvent (Milli-Q water and buffer solutions suitable for dissolution testing, such as simulated intestinal fluid without enzymes (interchangeable with phosphate buffer at pH 6.8), simulated gastric fluid without enzymes (interchangeable with 0.1 N HCl at pH 1.0) and a 0.15 M NaCl isotonic ionic strength solution. The test samples were placed in an accuTherm (Labnet) microtube shaking incubator and kept under a controlled temperature of 25 °C (±1 °C) while stirring up to 1000 rpm. A study of the amount of ibuprofen solubilized in the solvents over time ensured that the equilibrium was achieved. Before sampling, each solution was centrifugated (VWR ® Mega Star 600 R) at isothermal conditions to enhance the physical separation of the two phases. Concentrations were determined by UV-Vis spectroscopy using a VWR ® spectrophotometer, model UV-6300PC, after appropriate dilution and interpolation from previously acquired calibration curves. All the solubilities experiments were repeated at least two times.

Cytotoxicity Assays
To screen the cytotoxicity of the ibuprofen-based ILs two different cell lines were chosen: human colon carcinoma cells, Caco-2, and the hepatocellular carcinoma cell line, HepG2. The Caco-2 cell line was grown in RPMI 1640 medium supplemented with 10% of inactivated fetal bovine serum (FBS) and 1% penicillin-streptomycin, and the HepG2 cells were cultured in MEM medium with 10% inactivated FBS, 2 mM glutamine, 1% MEM-NEAA and 1% sodium pyruvate. All media and supplements were supplied by Gibco from Thermo Fisher Scientific. Both cell lines were kept at 37 °C in a humidified incubator (Smart Biotherm S-Bt, bioSan) with 5% CO2 and routinely grown in 175 cm 2 culture flasks. Caco-2 cells were seeded at a density of 2 × 10 5 cells per well in 96-well plates, and the experiments were performed using cells after reaching a 90% confluence, 96 h after seeding. The HepG2 cells were seeded at a density of 6 × 10 5 cells per well, and the experiments were performed at a confluence of 80%, 24 h after seeding. Each ibuprofen-based IL was tested in a concentration range above/in the maximum plasma concentration (Cmax) of ibuprofen: 0.175 mM [26]. Additionally, the half maximal effective concentration (EC50) was determined for the Caco-2 cell line within concentrations ranging from 0.012 up to 4.8 mM for ibuprofen, 0.008 up to 7.5 mM for the API-ILs and 0.008 up to 320 mM for the ILs. All the stock solutions were homogenously prepared and diluted in 0.5% FBS culture media and added to a 96-well plate that was previously seeded. All cell lines were incubated for 24 h. Negative control cells were prepared containing culture medium with 0.25% (v/v) DMSO and the positive control cells only containing DMSO. To perform the cytotoxicity assay, an CyQUANT™ XTT Cell Viability test kit (Invitrogen, Waltham, MA, USA) was used. A preworking solution was prepared using an electron coupling reagent and the XTT reagent ((2,3-bis-(2-methoxy-4-nitro-5-sulfophenyl)-2H-tetrazolium-5-carboxanilide) according to the manufacturer's instructions. After the incubation period, all samples were removed and 100 µL of the working solution was added to each well and left to react for 4 h. The XTT reagent is sensitive to cellular redox potential; thus, actively respiring cells convert the water-soluble XTT compound into an orangecolored formazan product. The amount of formazan product that was soluble in the culture medium was proportional to the number of viable cells, and it was quantified spectrophotometrically at 450 nm in a Multiskan GO (ThermoFisher Scientific) microplate reader. Each sample was incubated in three different wells, and the obtained value was the average of three independent assays. Cell viability was determined by the ratio between the measured test compound-contacted cells and the measured absorbance of the control cells treated only with culture medium. Dose-independent viability curves were determined using the cell viability trends. Nonlinear regression analysis was used to determine the EC50 values for ibuprofen, ibuprofen-based ILs and IL/salt cation "suppliers" (i.e., [C2C1Im]Cl, [C2(OH)C1Im]Cl and [N1112(OH)]Cl).

Hemolytic Activity
The hemolytic activity was assessed according to the method optimized by Gaspar and coworkers [35]. Ethylene diamine tetraacetic acid (EDTA), preserved peripheral human blood obtained from voluntary donors, was used on the same day of all experiments. Briefly, the erythrocytes were centrifugated at 1000× g for 10 min to separate them from the serum and washed three times in a PBS solution. The ibuprofen-based ILs and sodium ibuprofen were prepared in PBS with a final concentration of 6 mM and ibuprofen, due to the fact of its low solubility, with a final concentration of 3 mM. The assay was performed in 96-well plates in which 100 µL/well of sample were diluted with 100 µL of the erythrocyte suspension. Moreover, the microplates were incubated at 37 °C for 1 h followed by a centrifugation at 800× g for 10 min. The supernatant absorbance was measured at 570 nm with a reference filter at 600 nm. The percentage of the hemolytic activity was calculated according to Equation (1): where AbsS is the average absorbance of the sample; AbsN is the average absorbance of the negative control; AbsP is the average absorbance of the positive control. All the compounds were tested in triplicate.

Protein Albumin Denaturation Assay
The inhibition of protein denaturation by the three ibuprofen-based ILs was evaluated according to the method described by Mizushima and Kobayashi [36] with slight modifications, and it was implemented to assess the anti-inflammatory activity of the API-ILs. Briefly, a 2% bovine serum albumin (BSA) solution was prepared in PBS at pH 7.4. Ibuprofen and sodium ibuprofen were used as positive control drugs. A solution of only PBS and 2% BSA was used as the negative control. Solutions of each API-IL and control drugs were also prepared in PBS. The reaction mixture containing 0.5 mL of 2% BSA and 1.5 mL of each test compound was incubated at 37 °C for 15 min and cooled at room temperature. Denaturation was induced by maintaining the mixtures at 80 °C ± 1 °C in a dry bath for 15 min. The turbidity of the solutions represented the level of protein precipitation. Thus, in the manual centrifuge, all samples were centrifuged and then measured at 660 nm. Each sample concentration consisted of at least two independent replicates.

Cyclooxygenases (COX-1 and COX-2) Imhibition Assays
The anti-inflammatory activity was directly determined by measuring the inhibition of the COX enzymes using a colorimetric COX (COX-1, ovine; COX-2, human) inhibitor screening assay kit (Cayman Chemical Co., No. 701050, Ann Arbor, MI, USA). The assay was conducted according to the manufacturer's instructions. Briefly, a reaction mixture containing 150 µL of assay buffer, 10 µL of heme, 10 µL of enzyme (either COX-1 or COX-2) and 10 µL of ibuprofen or ibuprofen-based ILs at 3 mM were added to the plate. To evaluate the 100% initial activity, a solution of 150 µL of assay buffer, 10 µL of heme, 10 µL of enzyme (either COX-1 or COX-2) and 10 µL of a solution 50/50 H2O and EtOH was added to the plate. The nonenzymatic solution designated as background contained 160 µL of assay buffer, while 10 µL of heme and 10 µL of a solution of 50/50 H2O and EtOH were also added to the plate. Cayman's COX colorimetric inhibitor screening assay measures the peroxidase component of COXs. The peroxidase activity was assayed colorimetrically by monitoring the appearance of oxidized N, N, N' and N'-tetramethyl-phenylenediamine (TMPD) at 590 nm [37]. Each plate was firstly incubated for 5 min at 25 °C and then 20 µL TMPD, and 20 µL of arachidonic acid was added to start the reaction. For a more accurate determination of the reaction rates, the samples were measured kinetically for precisely 2 min at 25 °C. All samples were assayed in triplicate. The COX-1 and COX-2 percent inhibition was calculated using Equation (2): where C is absorbance of the 100% initial activity sample; T is the subtraction of each inhibitor sample from the 100% initial activity sample. The absorbances are the average of all samples, and one must subtract the absorbance of the background wells from the absorbances of the 100% initial activity and the inhibitor wells.

Characterization of the Ibuprofen-based Ionic Liquids
The 1 H and 13 C NMR spectra were acquired for the API-ILs sodium ibuprofen and ibuprofen. The ibuprofen-based ILs presented the expected chemical shifts from the anion and the corresponding cation, meaning that the reactions were complete, and only one highly pure product was formed in each reaction. In agreement with the intended stoichiometry, a strict 1:1 proportion was always observed in the 1 H NMR spectra (see the Supplementary Materials). The hydroxyl group (-COOH) 1 H NMR chemical shift for the ibuprofen acid formed was detected at 12.15 ppm in DMSO and, as expected, no peak was observed in the ibuprofen-based ILs (the neutralization step in the two-step anion exchange reaction; see Section 2.1) and sodium ibuprofen 1 H NMR spectra in the same deuterated solvent. Additionally, there w a clear change in the chemical shifts upfield for protons 2 and 3 of ibuprofen for the APIs' salt forms in DMSO (see Table 2), corroborating the cations and API-based anion interactions. The same trend was observed in D2O, and it is depicted in Table S1 in the Supplementary Materials. The chemical shift deviations were more significant for the API-ILs, which is supported by the interionic hydrogenbonded network in the ILs, verified by experimental measurements and computer simulations [38,39].  [Ibu] was liquid, at room temperature. The addition of a hydroxyl group connected to the aromatic imidazolium ring seemed to strongly decrease the melting temperature. The same was observed by Ferreira et al. with fluorinated ionic liquids based on an imidazolium cation, where it was inferred that both structural features (i.e., hydroxyl group and anion) highly influence the thermal properties of ILs [40]. Although the API-IL [C2C1Im] [Ibu] was not liquid at room temperature or even at body temperature (a more relevant temperature for pharmaceutical applications; the Tm cut-off of API-ILs for use within the body should be approximately 37 °C), and the API-IL platform was not effective in reducing the melting point of the parent API in this IL (see Table 3), the DSC trace at a scan rate of 1 °C/min (see Figure S2 in the Supplementary Materials) indicated that [C2C1Im][Ibu] exhibited a cold crystallization temperature (26.98 °C) and, thus, depending on the cooling-heating cycles, this API-IL can easily be handled as a liquid at room temperature. This is advantageous for the development of new drug delivery systems such as the adsorption of API-ILs onto mesoporous silica-based materials for the controlled delivery of APIs [41] and incorporation of API-ILs into bacterial nanocellulose membranes for transdermal drug delivery systems [26].

Equilibrium Solubility in Water and Simulated Biological Fluids
A poor water solubility is a limiting factor in the efficacy and bioavailability of an API. The solubilities of the ibuprofen-based ILs, ibuprofen and ibuprofen sodium salt determined in the present work (see Section 2.4) are listed in Table 4. Ibuprofen, itself, by being an acid drug, has a low solubility in water (0.297 mM), a simulated gastric fluid solution at pH 1.0 (0.215 mM) and in the buffer, NaCl, a physiological solution mimicking the isotonic ionic strength in the bloodstream (0.353 mM). In the simulated intestinal fluid at pH 6.8, it had an increased solubility (20.21 mM) compared to the previous buffers and water.
The replacement of an acidic proton in ibuprofen by sodium significantly increases the Lewis-based properties of sodium ibuprofen [42]. Through the enhancement of basic properties, ibuprofen sodium salt presented a higher solubility and within the same order   Due to the anion-cation interactions, the solubilities of the API-ILs were much higher than those found for ibuprofen in the neutral and salt forms. The solubility studies attained in this work are illustrated in Figure 1, and Figure 1A depicts the solubility of ibuprofen, sodium ibuprofen and all three ibuprofen-based ILs in water, while Figure 1B [44]. Recently, Wu and coworkers also reported an increased water solubility of imidazoliumbased ibuprofen ILs ([C4C1Im][Ibu] and [C2(OH)C1Im][Ibu]) and correlated the higher solubility of the API-ILs to the higher permeability results and higher bioavailability [23].
Cholinium befits many physiological functions due to the fact of its incorporation into neurotransmitters, signaling molecules and membrane components [12]. In addition to imidazolium-based ibuprofen IL, cholinium-based ibuprofen IL was also prepared and investigated in this work for their potential use in pharmaceutical applications due to the prevalence of cholinium in the human body. Additionally, cholinium salts, such as chloride ([N1112(OH)]Cl-the cation "supplier" for [N1112(OH)][Ibu] in the implemented two-step anion exchange synthetic procedure, see Section 2.1), are currently cost-efficient chemicals, produced on a scale of millions per year. Accordingly, cholinium-based API-ILs propound biocompatibility and economic advantages. Thus, the solubility of [N1112(OH)] [Ibu] in simulated biological fluids at 25 °C were determined as well as for ibuprofen (parent API) and sodium ibuprofen (commercial salt form of the API). The results are depicted in Table 4 and Figure 1B [24], in agreement with the results presented herein.
The overall solubility results clearly demonstrate that the cholinium-and imidazolium-based API-ILs strategy was appropriate to twist the solubility in water and buffer solutions suitable for dissolution testing of the parent API and, consequently, enhanced the efficacy, bioavailability and potential membrane permeability. Solving the bioavailability problems of APIs is one of the critical challenges of the pharmaceutical industry, since nearly half of the new active substances being identified in high-throughput screening are either insoluble or poorly soluble in water [12].

Cytotoxicity Profile in Human Cell Lines
The cytotoxicity profiles of ibuprofen-based API-ILs as well as ibuprofen and sodium ibuprofen were accessed in the human colon carcinoma cell line, Caco-2, and the hepatocellular carcinoma cell line, HepG2. The studied concentrations were above/in the range of the pharmacokinetic parameter maximum plasma concentration of ibuprofen (Cmax, 0.175 mM; the highest level of ibuprofen that can be obtained in the blood usually following multiple doses) [45], which is above the possible intracellular concentrations. The dose-response cytotoxicity curves are depicted in Figure 2. Comparing the viability curves of ibuprofen, sodium ibuprofen and ibuprofen-based ILs, no significant changes were verified, indicating that the use of a modular IL strategy based on the cations [N1112(OH)] + , [C2C1Im] + and [C2(OH)C1Im] + produced no significant effect on the ibuprofen system cytotoxicity. NSAIDs are mainly associated with gastrointestinal problems [21]. Thus, to evaluate the relative toxicity in the in gastrointestinal tract, the EC50 (i.e., effective concentration reducing cell viability to 50%) in the Caco-2 cell line was acquired and is depicted in Table  5. In Figure S5, the nonlinear regression curves used to obtain the EC50 values are illustrated. In addition to API-ILs and parent API, the EC50 in the Caco-2 cell line was determined for IL/salt cation "suppliers" for the API-ILs (i.e.,  in Caco-2 (E) and HepG2 (F). Both Caco-2 and HepG2 cells were exposed for 24 h to the respective API-ILs and parent API at the concentrations shown. The EC50 values for ibuprofen and the three ibuprofen-based API-ILs were at the mM level (see Table 5). All four ibuprofen formulations (neutral and API-ILs) showed low toxicity against Caco-2 cells. Cl) also showed low toxicity, lower than the ibuprofen-based counterparts, which is in agreement with previous literature that compared the cytotoxicity of ibuprofen API-ILs and the corresponding precursor halide salts in different cell lines [23,24,48]. The EC50 for the precursor halide salts followed the trend: [N1112(OH)]Cl > [C2(OH)C1Im]Cl > [C2C1Im]Cl (see Table 5 and Figure S6), supporting the higher biocompatibility of the cholinium cation as discuss above (see Section 3.2) and indicating that the incorporation of a hydroxyl group in the alkyl substituent of the imidazolium ring significantly decreases the toxicity of imidazolium-based ILs ([C2C1Im] + versus [C2(OH)C1Im] + ) [49]. The overall analysis of the in vitro cytotoxicity assays indicates the suitability and biocompatibility of the proposed cholinium-and imidazolium-based API-ILs for pharmaceutical formulations.

Hemolytic Activity
Drug-induced immune hemolytic anemia is a serious condition that can be a rare side effect of commonly used over-the-counter medications such as NSAIDs [50,51]. Thus, an in vitro hemolysis study was performed prior to any pharmaceutical application to ensure that there was no serious potential pharmacologically mediated toxicity according to the guidelines of the European Medicines Agency (EMA) [52]. In this study, the red blood cell lysis through the hemoglobin release in the plasma was evaluated. Ibuprofen did not reveal hemolytic activity in the range of concentrations studied up to 1.5 mM. Additionally, ibuprofen sodium salt and ibuprofen-based ILs also displayed no hemolytic activity against human erythrocytes up to 3 mM. The studied concentration range for ibuprofen was limited to 1.5 mM due to the lower solubility in PBS (see Section 2.6). Still, the studied concentrations were above the Cmax of ibuprofen (0.175 mM) [45], which was above the possible intracellular concentrations. The structure-activity correlation suggests that an aromatic ring with a side chain containing a carbonyl group attached to a nitrogen atom is a requirement for stabilizing the erythrocyte membrane [53]. Additionally, our group reported the hemolytic activity of several ionic liquids and demonstrated that imidazolium and cholinium-based ILs are considered hemocompatible with ~0% hemolytic activity up to 3 mM [54]. Hemolysis also represents the most employed initial toxicity assessment in drug development that could be correlated to cytotoxicity assays, since the main reason for toxicity can be related to the disruption of cell membranes [55]. The hemocompatibility results agree with the cytotoxicity evaluation. Ibuprofen and ibuprofen API-ILs showed low toxicity against Caco-2 cells, their EC50 were at the mM level (see Table 5), higher than 4.0 mM for ibuprofen and 5.5 mM for the API-ILs.

Protein Albumin Denaturation Assay
Protein denaturation has been correlated with the formation of inflammatory disorders, such as rheumatoid arthritis, diabetes and cancer, since proteins lose their biological activity. Therefore, the ability of a substance to prevent protein denaturation may also help to prevent the inflammatory conditions [56]. Serum albumins are the most abundant proteins in plasma and are responsible for the movement of drugs through the blood stream. BSA is widely used as a model protein in many areas of research, because it holds similar properties to those of human serum albumin [57]. Herein, BSA was used to assess the ability of ibuprofen, ibuprofen sodium salt and ibuprofen-based ILs to prevent protein denaturation and, consequently, to evaluate their anti-inflammatory properties.
Ibuprofen  Figure 3 and Table S2). Cholinium cations can be considered as chaotropic agents [58,59] while imidazolium cations as kosmotropic agents [59,60]. A chaotropic agent can reduce the amount of order in the structure of a protein formed by water molecules, where as a kosmotropic agent can cause water molecules to favorably interact which, in effect, stabilizes the intramolecular interactions in macromolecules such as proteins [61]. Additionally, Pace et al. [62] investigated the contribution of polar groups and their hydrogen bonds to the conformational stability of proteins and disclosed that hydrogen bonds by side chain hydroxyl groups make a favorable contribution to protein stability. The hydrogen bonding and other interactions of -OH groups in folded proteins can be sometimes more favorable than interactions with water in the unfolded protein. Consequently, the hydroxyl group presented in the [C2(OH)C1Im][Ibu] API-ILs (see Table 1) contributed to stabilizing the BSA, preventing its denaturation in the tested conditions.

Cyclooxygenases (COX-1 and COX-2) Inhibition Assay
Cyclooxygenase (COX, also called prostaglandin H synthase (PGHS)) is a bifunctional enzyme exhibiting both COX and peroxidase activities. The COX component converts arachidonic acid to a hydroperoxyl endoperoxide (PGG2), and the peroxidase component reduced the endoperoxide to the corresponding alcohol (PGH2), the precursor of PGs, prostacyclins and thromboxanes [63,64]. Currently, it is well established that there are two distinct isoforms of COX. COX-1 is constitutively expressed in a variety of cell types and is involved in normal cellular homeostasis. The expression of a second isoform of COX, COX-2, is induced by a variety of mitogenic stimuli such as phorbol esters, lipopolysaccharides, and cytokines. COX-2 is responsible for the biosynthesis of PGs under acute inflammatory conditions [65]. All NSAIDs are essentially COX-2 inhibitors with differing degrees of COX-1 inhibition as a side effect. NSAIDs act by inhibiting these enzymes, which are involved in prostaglandin synthesis, resulting in their antipyretic, analgesic and anti-inflammatory effects. Drugs that inhibit COX-1 and COX-2 with comparable potency (non-selective NSAIDs, e.g., ibuprofen and ketoprofen) will not spare COX-1 activity after dosing, while drugs with intermediate COX-2 selectivity (e.g., diclofenac and nimesulide) or highly selective COX-2 inhibitors (e.g., rofecoxib and etoricoxib) have greater potential for sparing COX-1 activity [66]. In the gastrointestinal (GI) track, kidneys, and platelets, COX-1 is a commonly present enzyme responsible for the synthesis of prostaglandins, mainly in the GI mucosa and platelets. The produced prostaglandins help to maintain the GI mucosal integrity and renal blood flow as well as platelet activation. Symptomatic ulcers and ulcer complications associated with the use of nonselective NSAIDs is a very common side effect and may occur in approximately 1% of patients treated for three to six months and in 2-4% of patients treated for one year. Additionally, platelet function may also be impaired because of NSAIDs' inhibitory effect on thromboxane, a potent aggregating agent, resulting in bleeding [67]. Otherwise, COX-2 is primarily found at sites of inflammation. Additionally, the use of selective COX-2 inhibitors, as demonstrated by Simon and coworkers in long-term rheumatoid arthritis patients [68], allowed a significative reduction in pre-existing conditions, such as perforation and hemorrhages in in the GI, alongside anti-inflammatory and analgesic activity. Accordingly, it is advantageous to develop selective COX-2 NSAIDs with minimal COX-1 inhibition after dosing, allowing pain relief and inflammation reduction without no disruption of platelet and a lower risk for GI toxicity. Ibuprofen (nonselective NSAIDs) that inhibit COX-1 and COX-2 with comparable potency, which is observed in the relative COX-1/COX-2 selectivity [69], will not spare COX-1 activity after dosing, with all the unwanted side effects [70,71].
Ibuprofen-based ILs, as well as ibuprofen, were evaluated for their ability to inhibit cyclooxygenases (COX-1, ovine; COX-2, human) using a COX colorimetric inhibitor assay kit (see Section 2.8), providing direct insight into their anti-inflammatory properties. The results obtained are depicted in Figure 4 and Table 6.  These results clearly demonstrate the potential of the API-ILs platform in the development of safer NSAIDs with improved gastric and renal safety profiles and the recognition of new avenues for selective COX-2 inhibitors in cancer chemotherapy and neurological diseases such as Alzheimer's and Parkinson's. Table 6. Inhibition of COX-1 (ovine) and COX-2 (human) and COX-1/COX-2 selectivity for 3 mM ibuprofen and ibuprofen-based ILs.

Conclusions
In the present work, we reported on the synthesis of three NSAID-based ionic liquids, conjugating the ibuprofenate anion (derived from ibuprofen, one of the most commonly available over-the-counter NSAIDs) with imidazolium cations (i.e., ). An upgrade to the aqueous solubility (i.e., water and biological simulated fluids) for the ibuprofen-based ILs relative to the ibuprofen neutral and commercial ibuprofen salt form (i.e., sodium ibuprofen) was verified. Cytotoxicity tests on human cell lines and hemocompatibility assays substantiated the biocompatibility of the ibuprofen-based ILs, viz., ibuprofen-based ILs, sodium ibuprofen and ibuprofen, which displayed similar cytotoxic responses and all forms were hemocompatible, which boosts opportunities for creating advances in pharmaceutical challenges. The pharmacological action of the prepared API-ILs were tested to validate if a potential synergetic effect between the ion pairs existed, consequently conducting novel therapeutic advantages. Accordingly, the anti-inflammatory properties of the ibuprofen-based ILs were assessed through the inhibition of BSA denaturation and inhibition of cyclooxygenases (i.e., COX-1 and COX-2), showing that all prepared API-ILs,  Table S1: Chemical shifts for the hydrogens in positions 2 and 3 (as depicted in the chemical structure of ibuprofen on the right) for ibuprofen, sodium ibuprofen and API-ILs in D2O; Figure S5 Figure S6: EC50 values of ibuprofen, ibuprofen-based ILs and IL/salt cation "suppliers" for the API-ILs ([C2C1Im]Cl, [C2(OH)C1Im]Cl and [N1112(OH)]Cl) in the Caco-2 cell line exposed to the compounds for 24 h. The Rsquared was greater than 0.9540 with p-value < 0.0001 for all fitted curves; Table S2: Inhibition of BSA denaturation in PBS pH 7.4 at different concentrations for the neutral and salt forms of ibuprofen and API-ILs. The presented value is the mean of at least two independent measures ± standard deviation.