2026 Volume 51 Issue 3 Pages 227-237
OE2imC3C, a zwitterionic liquid, has attracted considerable attention as an emerging solvent with low toxicity. However, the safety data in research reports are limited, and its applicability to toxicity tests remains unclear. Therefore, we investigated the detailed safety profile of OE2imC3C and found that it was negative in both the bacterial reverse mutation and chromosome aberration tests, indicating no concern for genotoxicity. The hepatocellular toxicity of OE2imC3C stock solution (OS) was lower than that of dimethyl sulfoxide (DMSO). OS was widely applied and gave correct hepatocellular toxicity of ellagic acid (a functional food molecule with anticancer, antioxidant, and other properties), based on the high solubility. In contrast, DMSO revealed false toxicity in the test because the undissolved ellagic acid crystals damaged the cells. The general applicability of OS to food-related compounds was confirmed using Hansen solubility parameters. OS is a unique solvent capable to prevent artifacts in toxicity assays involving poorly soluble compounds such as food and drug molecules.
There is a growing international movement to reduce and replace animal testing in line with increasing concerns for animal welfare. The EU promoted the reduction of animal testing by introducing the Registration, Evaluation, Authorization, and Restriction of Chemicals regulation in 2007 and by banning animal testing for cosmetics in 2013 (European Commission, 2013; Sharma et al., 2022). In response to this trend, non-animal testing methods are being actively conducted to evaluate the safety and efficacy of new compounds. The importance of such methods has been recognized in the cosmetics, pharmaceutical, agrochemical, and food industries. In particular, in silico predictions and in vitro tests have attracted considerable attention (Pallocca et al., 2022; Zushin et al., 2023). However, in silico prediction is difficult to apply to mixtures and compounds having limited training data. As a result, most of newly approved test methods are based on in vitro testing (Westmoreland et al., 2022), in addition to classical in vitro toxicity tests such as bacterial reverse mutation test, chromosome aberration test, and hepatocellular toxicity test used in drug discovery (Chalasani et al., 2025).
Despite these advances, a critical challenge in vitro testing is compound insolubility, although such insoluble solid compounds can be administered orally in animal tests. During in vitro experiments, these solids float or precipitate in the culture medium and have little to no biological activity; in the worst-case scenario, hard and sharp crystals can physically damage the cells. In the field of drug discovery, it has been reported that 75% of drug development candidates exhibit low solubility and belong to Biopharmaceutical Classification System classes II and IV (Di et al., 2009). The efficacy of poorly soluble compounds is often underestimated during the early stages of drug development (Di and Kerns, 2005). A similar situation occurs in the food industries, where many compounds used in functional foods are insoluble, even as the development of new functional foods has been accelerating. For example, ellagic acid, a dimeric derivative of gallic acid, has attracted increasing attention due to its anticancer, antioxidant, cardiovascular, antidiabetic, anti-inflammatory, and anti-aging effects (Ríos et al., 2018). However, ellagic acid is barely soluble in alcohol and water (Li et al., 2021), limiting its accessibility to toxicity tests. Flavonoids are another example: although they possess various functional properties, most are insoluble in water (Neeraja et al., 2017).
The solvent most used for poorly soluble compounds in in vitro tests is dimethyl sulfoxide (DMSO). DMSO is classified by the International Council for Harmonization of Technical Requirements for Pharmaceuticals for Human Use (ICH) as a class 3 (non-toxic) solvent (ICH, 2011). However, DMSO cannot dissolve all hydrophobic compounds, and dissolving compounds insoluble in both water and DMSO remains challenging. Additionally, DMSO often causes precipitation of the dissolved compounds upon dilution with culture media, which is also a problematic issue. Ethanol and dimethylformamide are usable as alternatives; however, they are not preferable options due to their high toxicity. Therefore, when DMSO is ineffective in vitro tests, no suitable alternatives are currently available.
Recently, zwitterionic liquids have attracted attention as new candidate solvents when DMSO is ineffective. OE2imC3C, a type of zwitterionic liquid, has demonstrated low toxicity in certain fibroblasts, iPS cells, zebrafish embryos, and Escherichia coli and is capable of dissolving hydrophobic drugs and cellulose (Kuroda et al., 2017; Kuroda et al., 2020). However, safety data for OE2imC3C remain insufficient for its application in in vitro non-animal testing. In this study, the basic toxicity of OE2imC3C was evaluated, and then OE2imC3C was applied to an in vitro hepatocellular toxicity test of poorly soluble ellagic acid.
Zwitterionic liquids are a family of salts with melting points below 100°C, consisting of organic cations and anions covalently linked within the same molecule. OE2imC3C (Fig. 1a) was synthesized as previously reported (Kuroda et al., 2017; Sharma et al., 2022). A stock solution of OE2imC3C (OS) was prepared by dissolving in distilled water at a concentration of 70 wt.% without sterilization such as autoclave and filtration.
Cytotoxicity of water, DMSO, and OS (70 wt.% OE2imC3C aqueous solution) in HepG2 cells. a Structures of the solvents used in this study. b Relative number of living HepG2 cells cultured for 24 hr with water, DMSO, or OS at the indicated concentrations (n = 4, experimental quadruplicate). Bars marked with different lowercase letters (a, b, c) indicate statistically significant differences among solvent groups at the same concentration (p<0.05). All error bars indicate standard error.
The bacterial reverse mutation test was conducted at Mediford Corporation, Japan (formerly LSIM Safety Institute Corporation). The study followed OECD Guideline 471 (Economic and (OECD), 1997a), using pre-incubation method in the presence or absence of the S9 metabolic activation mix.
Salmonella typhimurium strains (TA98, TA100, TA1535, and TA1537) and E. coli strain (WP2 uvrA) were purchased from the Japan Bioassay Research Center, Japan. OE2imC3C was prepared as the test sample, and distilled water was used as the negative control. For assays without metabolic activation, 2-(2-furyl)-3-(5-nitro-2-furyl)acrylamide (AF-2, Wako Pure Chemical Industries, Japan) was used as the positive control for TA98 and TA100, sodium azide (NaN3, Fujifilm Wako Pure Chemical Corporation, Japan) was used for TA1535, and 9-aminoacridine hydrochloride (9AA, Sigma-Aldrich, USA) was used for TA1537. Under metabolic activation conditions, 2-aminoanthracene (2AA, Fujifilm Wako Pure Chemical Corporation, Japan) was used as the positive control for all tester strains. Although 2AA was used as the sole positive control under metabolic activation conditions, compliance with OECD Test Guideline 471 was ensured by separately confirming the metabolic competence of the S9 mix. Specifically, the lot of S9 mix used in this study was qualified using benzo[a]pyrene in a batch certificate, demonstrating sufficient enzyme activity for promutagen activation. Each sample was assayed in duplicate.
The S9 mix (Ieda Trading Corporation, Japan) was prepared from the livers of Sprague-Dawley rats pretreated with phenobarbital and 5,6-benzoflavone. For the assay, 0.1 mL of bacterial culture was added to a test tube. In the absence of metabolic activation, 0.5 mL of 0.1 M sodium phosphate buffer (pH 7.4) was added, whereas in the presence of metabolic activation, 0.5 mL of the S9 mix (containing 10% S9 and salt-cofactor-I solution) was added instead. After adding 0.1 mL of the test chemical suspension, the mixture was incubated for 20 min at 37°C with shaking (120 rpm). After incubation, 2.0 mL of top agar was added and poured onto a minimal glucose agar plate. The top agar consisted of 0.5% NaCl and 0.6% Bacto™ Agar (Becton, Dickinson and Company, USA) in distilled water and was supplemented with 0.5 mM histidine and biotin for S. typhimurium tests, or 0.5 mM tryptophan for E. coli tests. Each culture plate was incubated in a CO2 incubator at 37°C for 48 hr. After incubation, the number of revertant colonies was counted using a colony analyzer, and the background bacterial growth was also visually examined under a stereomicroscope.
A response was considered positive if the test compound yielded more than twice the number of revertants compared to the negative control. In addition, dose-dependence in the number of revertants was considered when determining test results.
Chromosomal aberration assay in cell culturesThe chromosomal aberration assay was performed according to OECD guideline TG473 (Economic and (OECD) 1997b) in the presence or absence of the S9 mix.
Chinese hamster lung cell line (CHL/IU) cells were obtained from the Food and Drug Safety Center, Hatano Research Institute, Japan. The cells were grown in Eagle's minimum essential medium (MEM; Nissui Pharmaceutical, Japan) containing kanamycin sulfate at 60 mg (potency)/L, supplemented with 10% heat-inactivated calf serum (Global Life Sciences Technologies Japan K.K., Japan). The cells were cultured in plates at 37°C in a humidified atmosphere containing 5% CO2. OS was assayed up to a concentration that caused more than 50% reduction in relative population doubling (RPD). The RPD was calculated as follows:
If the RPD decreased rapidly and the dose interval was less than the common ratio of 2, the assay was continued until the dose fell below the negative population doubling level. Mitomycin C (MMC, Kyowa Hakko Kogyo Co., Ltd., Japan), dissolved in distilled water was used as a positive control in the absence of S9 at concentrations of 0.1 and 0.05 μg/mL for the 6- and 24-hr treatments, respectively. Cyclophosphamide monohydrate (CP, Sigma–Aldrich, USA) dissolved in DMSO, was used as a positive control in the presence of S9 at 10 μg/mL.
Cells were seeded at a density of 2.0 × 104 cells/60 mm plate and cultured in a 5.0% CO2 incubator (BNA-111; Tabai Espec, Japan) at 37°C. After 3 days of incubation, short-term treatments were conducted under both non-metabolic (−S9) and metabolic activation (+S9) conditions. For the −S9 treatment, 3.0 mL of medium containing OS was added and cells were incubated for 6 hr. For the +S9 treatment, 0.5 mL of S9 mix and 2.5 mL of medium containing OS were added for 6 hr. After exposure under both conditions, the medium was removed, cells were rinsed with MEM, and further cultured in 5.0 mL of fresh medium for 18 hr. For the continuous treatment (−S9 only), cells were cultured in 5.0 mL of medium containing OS for 24 hr. Two hours before the end of the culture period, 25 μL of colcemid solution (20 μg/mL, dissolved in distilled water) was added to arrest mitosis. The cells were trypsinized and suspended in Dulbecco's phosphate-buffered saline [PBS (-), Nissui Pharmaceutical, Japan]. The suspension was centrifuged, and the pellet was resuspended in 6 mL of 0.075 M hypotonic potassium chloride. Cells were fixed in 3:1 (v/v) methanol:acetic acid solution. After centrifugation, the supernatant was removed and the cells were treated with 10 mL of fresh fixative. The fixation procedure was repeated twice. The cells were suspended in 1 mL of fixative, and a drop was placed on a glass slide, air-dried, and stained with 3% Giemsa solution (Nacalai Tesque, Japan).
Specimens were coded for randomization. Using a microscope (Eclipse 55i, Nikon, Japan), 150 metaphase chromosomes were analyzed per specimen, and a total of 300 cells/concentration (for two specimens) were evaluated. Microscopic observations were conducted at 100 times magnification to screen for the presence or absence of chromosomal aberrations. Subsequently, observations at 400 times magnification were performed to differentiate specific aberrant findings, such as distinguishing chromatid gaps from chromatid breaks, or chromatid exchanges from simple overlapping of chromosomes, as well as other structurally ambiguous features. Chromosomal aberrations were classified as structural or numerical. Statistical analyses were conducted to determine whether experimental groups were negative or positive.
Solubility assay of ellagic acidEllagic acid (0.1%, 0.2%, 0.5%, or 1.0%) was added to the solvent (water, DMSO, or OS). The samples were vortexed for approximately 10 sec at room temperature and inspected visually. Samples were considered insoluble if undissolved or dispersed matter was observed, and soluble if a clear solution was observed. Ellagic acid (≥98% purity) was purchased from Nacalai Tesque, Inc.
In vitro hepatocellular toxicity testHepG2 cells were obtained from the American Type Culture Collection (Manassas, VA, USA). The cells were maintained in RPMI 1640 medium (Thermo Fisher Scientific, USA) containing 10% fetal bovine serum (Serana Europe GmbH, Germany), 2 mM sodium pyruvate (Wako Pure Chemical Industries, Japan), 4 mM L-Alanyl-L-Glutamine (Fujifilm Wako Pure Chemical Corporation, Japan), and 100 U/mL penicillin-streptomycin (Thermo Fisher Scientific, USA) in a 5% CO2 atmosphere at 37°C.
For the solvent toxicity assay, water, DMSO, and OS were added to serum-free media containing NADPH regeneration system solutions A and B (solution A: 26 mM NADP+, 66 mM Glucose-6-phosphate, and 66 mM MgCl2 in H2O; solution B: 40 U/mL Glucose-6-phosphate dehydrogenase in 5 mM sodium citrate) (Corning, NY, USA) at 0, 1, 2, 5, and 10 vol%, respectively.
For the ellagic acid toxicity test, ellagic acid was dissolved in DMSO or OS at concentrations of 3, 10, 30, or 60 mM. These stock solutions were diluted 100-fold in serum-free medium containing NADPH, resulting in final ellagic acid concentrations of 30, 100, 300, and 600 μM, with 1 vol% DMSO or OS present. Different final OS concentrations (2 vol%) were also tested. In this case, ellagic acid was dissolved in OS at 1.5, 5, 15, and 30 mM. These solutions were diluted 50-fold, yielding the same final ellagic acid concentrations of 30, 100, 300, 600 µM.
HepG2 cells were seeded at a density of 2.0 × 104 cells/well into biocoat collagen I 96-well plates (IWAKI, Japan) and incubated for 24 hr. Serum-free medium containing the respective test solutions was added to the cells, and the plates were incubated for an additional 24 hr. At the end of the treatment, the medium was removed, and the cells were rinsed with phosphate-buffered saline (PBS) before being cultured in fresh medium for another 24 hr. Cell morphology was observed using a microscope (Eclipse Ti-U; Nikon, Japan) at 40 times magnification. The number of living cells was counted to quantify cytotoxicity by staining the nuclei. The staining medium was prepared by adding 10 μL of Hoechst 33342 stock solution (H3570; Thermo Fisher Scientific, USA) to 10 mL of culture medium. After morphological observation, HepG2 cells were rinsed with Hanks' balanced salt solution (HBSS), and 100 μL of staining medium was added to each well. The cells were stained for 45 min under light-protected conditions. Subsequently, the staining medium was removed, and the cells were washed by adding fresh HBSS and shaking for 10 sec. This rinsing step was repeated three times. Automated image acquisition and analysis were performed using the CellInsight NXT High-Content Screning Platform (Thermo Fisher Scientific) based on the protocol used in Suntory Holdings Limited.
Precipitation of ellagic acid after dilutionThe ellagic acid solutions (300 µM) were prepared in 96-well plates (AS ONE Corporation, Japan) in the absence of cells as in the protocol of hepatocellular toxicity assay. The bottom of the well was observed at 4 times magnification under a microscope (ECLIPSE Ts2, Nikon, Japan) at the end of the 24-hr incubation with ellagic acid, as well as immediately before and 24 hr after rinsing with PBS, to check for precipitation of ellagic acid.
HSPiP MethodHansen Solubility Parameter software (HSPiP, ver. 5.4.08) was used to calculate the Hansen solubility parameters for each solvent.
Statistical analysisAll statistical analyses were performed using JMP software (SAS Institute, Cary, NC, USA), with a significance level set at p< 0.05.
For the chromosomal aberration test, the frequency of metaphase cells with structural chromosomal aberrations was compared between treatment and control groups using Fisher’s exact test. A Cochran-Armitage trend test was applied to evaluate dose-dependent responses.
For the hepatocellular toxicity tests, the relative number of living cells was analyzed using Dunnett’s test to compare each treatment group against the control (for hepatocellular toxicity tests of OS), and Tukey-Kramer multiple comparison test was used for pairwise comparisons among all groups (for concentrations of 300 μM and above in hepatocellular toxicity tests of ellagic acid).
OE2imC3C must exhibit no genotoxicity to handle it. It also exhibits low cytotoxicity in hepatocytes. Hepatotoxicity is a major cause of drug development discontinuation or market withdrawal (Chalasani et al., 2025) and hepatocellular toxicity is therefore recognized as a mandatory in vitro assay (Kim and Park 2025).
Preliminary study: Genotoxicity TestNegative genotoxicity is important for operator safety, and we evaluated it as a preliminary study. OS can be handled by pipetting and readily work as a general solvent, although the pure OE2imC3C solution has a high viscosity and cannot be pipetted.
The bacterial reverse mutation test was performed to assess overall survival. In both the -S9 and +S9 treatments, no increase in the number of revertant colonies exceeding twice that of the negative control group was observed (Table 1). No growth inhibition occurred in any strain treated with OE2imC3C and the positive control exhibited a clear mutagenic effect on each test strain. Therefore, under these test conditions, the mutagenic activity of OE2imC3C against bacteria was determined to be negative. During the preliminary assessments, including the Ames test, we encountered considerable difficulties in handling OE2imC3C due to its high viscosity. To improve the experimental handling of OE2imC3C as a solvent, we evaluated OS, which exhibits low viscosity.
;%0A%09%09%09newWindow.document.open();%0A%09%09%09newWindow.document.write('<img src=%22./Graphics/51_227-t1.jpg%22>');%0A%09%09%09newWindow.document.close();%0A%09%09)
The chromosome aberration test was performed with up to 5 vol% OS because 7.5 vol% OS significantly damaged the cells and decreased the RPD values (Tables 2 and S1). There is no statistically significant increase in the frequency of structurally and numerically abnormal cells, compared to the negative control group (Tables 2 and S1). A statistically significant increase was observed in the positive control group (Table 2). Based on these results, OS does not have chromosomal aberration-inducing potential under the conditions used in this study.
;%0A%09%09%09newWindow.document.open();%0A%09%09%09newWindow.document.write('<img src=%22./Graphics/51_227-t2.jpg%22>');%0A%09%09%09newWindow.document.close();%0A%09%09)
Since both the bacterial reverse mutation and the chromosome aberration tests were negative, we concluded that OS presents no genotoxicity concern. The results also suggest the applicability of OE2imC3C as a solvent in genotoxicity tests. This is an important insight given that genotoxicity tests are commonly performed in the early stages of compound development.
Hepatocellular toxicity testHepG2 cells were incubated in medium containing various concentrations of OS for 24 hr, followed by an additional 24-hr culture in fresh medium. The number of living cells was counted, and relative number of living cells was calculated by comparing to the untreated control group. The relative number of living cells was 95% and 91% at 1 and 2 vol% OS, respectively (Fig. 1b), whereas it decreased significantly to 75% and 4% at 5 and 10 vol%, respectively. These results suggest that OS should be diluted at least fifty-fold when used in hepatocellular toxicity assays for new compounds. A significant difference was observed between the relative viability at 5 vol% OS and DMSO but the relative number of living cells were 75 and 1%, indicating that OS is less toxic to hepatocytes than DMSO, which is considered a safe organic solvent.
Toxicity test of an DMSO-insoluble compounds by dissolving in OS Selection of a suitable test compoundTo demonstrate the potential of OS, we focused on DMSO-insoluble compounds that may give false-negative results in the hepatocellular toxicity test. We here selected ellagic acid, a hydrophobic functional food molecule found in pomegranate (Fig. 2a).
Solubility of ellagic acid in water, DMSO, and OS. a Chemical structure of ellagic acid. b Solubility of ellagic acid in OS, water, and DMSO. c Appearance of the mixtures after addition of ellagic acid.
The solubility of ellagic acid in OS was up to 1 wt% under the tested conditions (Figs. 2b and c) whereas that in DMSO is approximately 0.2 wt% (Bala et al., 2006). The OE2imC3C contained in OS has a highly polar imidazolium cation and carboxylate anion, and a less polar oligoether moiety (see Fig. 1a). OE2imC3C presumably well interacts to ellagic acid because ellagic acid also possesses both polar and less-polar components (hydroxyl groups and a hydrocarbon moiety) (Fig. 2a).
Hepatocellular toxicity test of ellagic acidCells were incubated for 24 hr with ellagic acid (30, 100, 300, and 600 µM) and the relative number of living cells was measured after PBS washing and a subsequent 24-hr incubation in fresh medium. Ellagic acid is utilized as a dietary supplement, and its safety is demonstrated in a 90-day repeated-dose toxicity study in rats, in which no liver toxicity is observed even at the highest dose of approximately 3000 mg/kg/day (Tasaki et al., 2008). Despite the absence of hepatic hazard in vivo, ellagic acid at concentrations of 300 µM or higher clearly decreased the living cell number when DMSO was used as the solvent (Fig. 3a). In the DMSO-treated group, precipitated ellagic acid was observed at the bottom of the wells after the test (Figs. 3b and S1), which can cause physical damage to cells. In contrast, OS treatment resulted in significantly higher living cell numbers than DMSO at the concentrations ≥300 µM (Fig. 3a). The amount of ellagic acid precipitated was significantly lower than that observed with DMSO. Despite a slight amount of precipitate at 1 vol% OS, no cellular damage was detected when comparing the effects of 1% and 2 vol% OS. These results suggest that hepatocellular toxicity of ellagic acid dissolved in DMSO was overestimated at concentrations above 300 µM: in other words, a false positive in hepatocellular toxicity test. In contrast, we think that accurate results were obtained using OS, owing to its higher solubility.
Effect of solvents on ellagic acid cytotoxicity in HepG2 cells. a Relative number of living HepG2 cells after 24 hr culture with ellagic acid dissolved in OS or DMSO (n = 4, experimental quadruplicate). The final concentration of DMSO was 1 vol%, and those of OS were 1 and 2 vol%, respectively. * Statistically significant compared with 1 vol% DMSO, p<0.05. All error bars indicate standard error. b Microscopic images of the well bottom after the test described in “a.”
We discuss here the importance of conducting in vitro studies of ellagic acid at concentrations above 300 µM. A no observed adverse effect level (NOAEL) of 3254 mg/kg/day for ellagic acid was reported in 90-day repeated mixed-feed administration using F344 rats, with no hepatotoxicity observed even at the dose (Tasaki et al., 2008). Although hepatic and plasma concentrations of ellagic acid were not examined in that study, another investigation reported a plasma Cmax of 0.203 mg/mL after administration of 800 mg/kg pomegranate extract (containing 85.3 mg/kg ellagic acid) to male Wistar rats (Lei et al., 2003). Given the molecular weights of ellagic acid (302.2 g/mol), this plasma concentration corresponds to 0.672 μM. The Cmax of ellagic acid in plasma at NOAEL (3254 mg/kg) is therefore briefly translated as 25.6 μM, if the relationship between the administered amount and concentration in plasma is proportional. Although hepatic concentrations of ellagic acid were not measured in this or other previous reports, they are expected to exceed plasma concentration. In fact, in silico studies estimating plasma and intrahepatic concentrations of 246 chemicals have demonstrated that hepatic concentrations are generally several times higher than plasma concentrations (Kamiya et al., 2021). When the coefficient is unknown, the hepatic concentration is generally estimated as ten times higher, for safety. Given the risk of overdose, particularly for food-derived ingredients compared to pharmaceuticals, the maximum dose should be several to ten times higher than the anticipated exposure level in early development stages. In summary, hepatocellular toxicity test of ellagic acid up to the concentration range of 100–300 µM (0.672 µM × several × ten × ten) is important to confirm the safety of ellagic acid only by in vitro tests. This is a possible case: some companies which prioritize safety hesitate to market ellagic acid if hepatocellular toxicity is observed in vitro using DMSO and the in vivo safety data lack. We believe that OS will enable the accurate interpretation of toxicity data in the early stage screening of compound development.
On the other hand, there is still a limitation: we could not evaluate interactions between OS and ellagic acid at the present study because the study of interaction is complicated. We need to clarify that OS–solute interactions do not alter the cytotoxicity of solutes in future, to conclusively claim OS as a universal solvent for toxicity testing.
Prediction of the versatility of OE2imC3CWe demonstrated the potential of OS using ellagic acid in this study; however its versatility was not fully explored. Therefore, we predicted the solubility of various food components in OS. One common method for such predictions is using Hansen solubility parameters (HSPs) composed of polar forces (𝛿P), hydrogen bond forces (𝛿H), and dispersion forces (𝛿D), calculated from the molecular structures (C.M. and Hansen 2007). These three parameters are calculated for both the solvent (a) and solute (b), and the distance R is calculated using the following equation:
A smaller R value indicates a higher affinity between the solvent and solute, indicating better solubility.
The solubilities of ellagic acid in DMSO, OE2imC3C (not OS because mixed solutions are complex), and water were calculated to evaluate predictability. The R values of ellagic acid in DMSO, OE2imC3C, and water were 7.2, 7.5, and 34.3, respectively (Table 3a). These results predict ellagic acid to be soluble in DMSO and OE2imC3C but insoluble in water, which is not much different from its actual solubility. Therefore, we examined an additional 40 food-derived compounds. For all compounds, the R values in DMSO and OE2imC3C were lower than those in water (Table 3b), suggesting that OS can serve as a solvent. This indicates that OE2imC3C can dissolve hydrophobic food components as well as DMSO.
;%0A%09%09%09newWindow.document.open();%0A%09%09%09newWindow.document.write('<img src=%22./Graphics/51_227-t3.jpg%22>');%0A%09%09%09newWindow.document.close();%0A%09%09)
Notably, the R values of DMSO and OE2imC3C for ellagic acid were similar, despite OS exhibiting higher solubility. From these results, we assume that HSPs are challenging to use for quantitatively predicting solubility but are useful for qualitatively predicting soluble/insoluble. Nevertheless, several reports have demonstrated the efficacy of HSPs in predicting solubility in ionic liquids (Zhu et al., 2019). HSPs were calculated based on the evaporation energies of the target molecules. The vapor pressure of ionic liquids is extremely low due to strong intermolecular (inter-ion) interactions (Ravula and Larm, 2019; Semavin et al., 2024; Chen et al., 2022). Consequently, the prediction software did not accurately reproduce the ionic structures. For instance, when OE2imC3C was input into the software, the carboxylate was output as a carboxylic acid. This suggests that the actual intermolecular interactions are not adequately captured by typical HSPs. This limitation is likely responsible for the lower accuracy in quantitative solubility predictions.
In summary, OS, a practically applicable 70 wt.% OE2imC3C aqueous solution, exhibits no genotoxicity concerns and lower hepatocellular toxicity than DMSO, indicating its suitability as a solvent for hepatocellular toxicity tests. Furthermore, OS exhibits a higher solubility of ellagic acid than DMSO and water (1 vs 0.2 vs <0.1 wt.%, respectively). High solubility produced accurate results in the hepatocellular toxicity test, whereas DMSO overestimated cell damage due to insoluble precipitates. Such overestimation can lead to false-positive results. However, the possibility cannot be excluded that interactions between OS and ellagic acid suppressed the cytotoxicity of ellagic acid. Therefore, further investigation is needed. These findings suggest that OS is a potential solvent for toxicity tests, including genotoxicity and hepatocellular toxicity tests. Although this study focused on in vitro tests, OS also has a potential for in vivo toxicity tests.
FundingSuntory Holdings Co., Ltd. provided collaborating research grant. This study was partly supported by KAKENHI (23H03824, 24K21250 from the Japan Society for the Promotion of Science), Super Highway (for K.K., from Japan Science and Technology Agency), Kanazawa University SAKIGAKE project 2020/2022/2024.
Conflict of interestY. K., E. H., and K. K. declare no competing financial interests. Y.K. is employee of Suntory Holdings Co., Ltd. This study received funding from Suntory Holdings Co., Ltd. The funder was not involved in the study design, collection, analysis, interpretation of data, the writing of this article or the decision to submit it for publication.
Data availabilityThe data in this study are included in the article/supplementary materials. Contact the corresponding author directly to request the underlying data.
Author contributionsConceptualization: Yusuke Kubota, Kosuke Kuroda
Funding acquisition: Yusuke Kubota, Kosuke Kuroda
Investigation: Yusuke Kubota, Ayako Hohsaka
Supervision: Kosuke Kuroda
Visualization: Yusuke Kubota, Ayako Hohsaka
Writing – original draft: Yusuke Kubota, Kosuke Kuroda
Writing – review & editing: Yusuke Kubota, Eishu Hirata, Kosuke Kuroda
Ethical approval and consent to participateNot applicable.
Patient consent for publicationNot applicable.
