N,N-Bis(2-hydroxyethyl)−2-aminoethanesulfonic acid (BES), polysorbate 80 (Tween 80), sodium chloride (NaCl), calcium chloride (CaCl2), magnesium sulfate heptahydrate (MgSO4·7H2O), sodium acetate anhydrous (NaAc), cholesterol, mucin type II from porcine stomach, bovine serum albumin (BSA), sodium hydroxide (NaOH) pellets, 5 M NaOH solution, maleic acid, tris base, budesonide, febuxostat, fenofibrate, furosemide, hydrochlorothiazide, hydrocortisone, ketoprofen, prednisolone, sulfasalazine, theophylline, ticagrelor, acetonitrile (ACN), methanol (MeOH) and formic acid (FA) were obtained from Merck (Darmstadt, Germany). Natrosol™ 250 HHX PHARM (hydroxyethyl cellulose, HEC) was kindly provided by Ashland™ (Schaffhausen, Switzerland), phosphatidylcholine (PC) by Lipoid (Ludwigshafen, Germany), and felodipine from AstraZeneca (Gothenburg, Sweden). Oxprenolol (free form) and fexofenadine (free form) were purchased from Toronto Research Chemicals (Toronto, Canada), and apixaban from Thermo Scientific (Geel, Belgium). Fed-state simulated colonic fluid (FeSSCoF) powder was obtained from Biorelevant (London, UK).

Drug solubility was tested in 6 different media in this study (Table 1). BES buffer (10 mM) was prepared by dissolving 430 mg BES in 200 mL MilliQ water and the pH was adjusted to 7.0 with 5 M NaOH.

Mucus solution was prepared based on a recently reported porcine artificial colonic mucus formulation (17, 18), which was developed according to the characterized porcine proximal colonic mucus (14), but without the addition of the polymer. Briefly, mucin type II (0.91% w/v) and BSA (7.02% w/v) were dissolved in a non-isotonic buffer (10 mM BES with 1.3 mM CaCl2, 1.0 mM MgSO4, pH 7.0) under magnetic stirring. This concentration of BSA was selected based on the higher relative abundance observed in the porcine native mucus (14, 16). The lipid mixture containing cholesterol (2.1% w/v) and PC (1.2% w/v) was made by dissolving them in isotonic buffer (10 mM BES with 137 mM NaCl, 1.3 mM CaCl2 and 1.0 mM MgSO4, pH 7.0) in the presence of Tween 80 (1.1% w/v) and under sonication. Subsequently, the solution containing mucin type II and BSA was mixed with the lipid solution at a ratio of 9:1 v/v, and the pH was adjusted to 7.0 using 5 M NaOH solution.

A simplified system was employed to investigate the influence of the polymer on drug solubility. Instead of polyacrylic acid in the previously reported study (18), HEC, which has been found to better capture the surface charge properties of the native porcine colonic mucus (17), was added to the 10 mM BES buffer at the concentration of 0.25% w/v and 0.5% w/v under intense magnetic stirring. The obtained solutions were placed at room temperature for 30 min for equilibrium before usage.

FeSSCoF was selected as the representative colonic fluid. The preparation was according to the instructions from the manufacturer (Biorelevant). Briefly, FeSSCoF buffer (16.5 mM NaOH, 30 mM maleic acid, 30 mM tris base) was prepared by dissolving the required amount of NaOH pellets, maleic acid, and tris base into MilliQ water. Subsequently, the pH of the buffer was adjusted with 1 M NaOH to 6.0. To make FeSSCoF, 0.074% w/v of the FeSSCoF powder was added to the FeSSCoF buffer under magnetic stirring and the acquired solution was equilibrated for 1 h before usage.

Each drug was added in excess to 2 mL of media in a 4 mL glass vial. The vials were infilled with nitrogen to minimize risk of oxidation and sealed by parafilm. Afterwards, the vials were placed in an incubator (Termaks, Norway) and were shaken at 37℃ at 300 rpm. Samples were withdrawn after 24 h. For tests in BES buffer, mucus solution, FeSSCoF buffer and FeSSCoF, samples were centrifuged at 21,000 g, 37℃ for 15 min to separate the solid materials, while for tests in BES buffer with HEC, free drugs were collected by centrifuging at 21,000 g for 30 min. Appropriate dilution was conducted before quantifying drug concentrations in the supernatants. The pH of the testing media including BES buffer, mucus solution, FeSSCoF buffer and FeSSCoF were measured after solubility studies. Solubility was measured in triplicate for each drug.

The solid state of studied compounds was not further confirmed after the solubility measurements, as the scope was to understand the solubilization capacity of colonic mucus from a physiological point of view rather than a physicochemical point of view.

The concentration of sulfasalazine was detected using Spark® multimode microplate readers (Tecan Trading AG, Switzerland). Samples were measured in a transparent 96-well plate and UV absorbance was detected at 360 nm and drug concentration was determined based on a quality-controlled calibration curve.

Except for sulfasalazine, the concentrations of all other compounds were analyzed with HPLC (1290 Infinity, Agilent Technologies, Santa Clara, CA, USA) using a Zorbax Eclipse XDB-C18 column (4.6 × 100 mm kept at 40°C). The methods used are summarized in Table 2. The flow rate was 1 mL/min and the injection volume was 20 μL for all the compounds except for furosemide for which 10 μL was used. Calibration curves used for concentration determination and quality control samples were prepared to assure the accuracy of the analysis.

Molecular descriptors were generated from RDKit deriving from SMILES strings of each compound (19). Selected descriptors are listed in Table 3, while the other calculated descriptors can be found in the supplementary information.

Statistical analysis was done using GraphPad Prism 10 version 10.2.2 (397). The results are presented as mean ± standard deviation. One-way ANOVA test followed by Dunnett’s multiple comparison test was used to determine the statistical significance of different groups. P values less than 0.05 were considered statistically significant. Spearman’s correlation analysis was performed to interpret the relevance of molecular features towards the drug solubility in the mucus solution.