Per-6-Thiolated Cyclodextrins: A Novel Type of Permeation Enhancing Excipients for BCS Class IV Drugs.

The purpose of the study was to develop a per-6-thiolated α-cyclodextrin (α-CD) by substituting all primary hydroxyl groups of α-CD with thiol groups and to assess its solubility-improving and permeation-enhancing properties for a BCS Class IV drug in vitro as well as in vivo. The primary hydroxyl groups of α-CD were replaced by iodine, followed by substitution with -SH groups. The structure of per-6-thiolated α-CD was approved by FT-IR and 1H NMR spectroscopy. The per-6-thiolated was characterized for thiol content, -SH stability, cytotoxicity, and solubility-improving properties by using the model BCS Class IV drug furosemide (FUR). The mucoadhesive properties of the thiolated oligomer were investigated via viscoelastic measurements with porcine mucus, whereas permeation-enhancing features were evaluated on the Caco-2 cell monolayer and rat gut mucosa. Furthermore, oral bioavailability studies were performed in rats. The per-6-thiolated α-CD oligomer displayed 4244 ± 402 μmol/g thiol groups. These -SH groups were stable at pH ≤ 4, exhibiting a pKa value of 8.1, but subject to oxidation at higher pH. Per-6-thiolated α-CD was not cytotoxic to Caco-2 cells in 0.5% (m/v) concentration within 24 h. It improved the solubility of FUR in the same manner as unmodified α-CD. The addition of per-6-thiolated α-CD (0.5% m/v) increased the mucus viscosity up to 5.8-fold at 37 °C within 4 h. Because of the incorporation in per-6-thiolated α-CD, the apparent permeability coefficient (Papp) of FUR was 6.87-fold improved on the Caco-2 cell monolayer and 6.55-fold on the intestinal mucosa. Moreover, in vivo studies showed a 4.9-fold improved oral bioavailability of FUR due to the incorporation in per-6-thiolated α-CD. These results indicate that per-6-thiolated α-CD would be a promising auxiliary agent for the mucosal delivery of, in particular, BCS Class IV drugs.

Introduction

For drug delivery, oral drug administration is the most appropriate route with a high patient compliance. Numerous drugs, however, cannot be delivered orally because of poor solubility and low drug absorption in the intestine leading to an insufficient bioavailability.[1] In fact, more than 90% of drugs approved since 1995 show either poor solubility or poor permeability or both.[2] Strategies to address these formulation challenges include the co-administration of auxiliary agents, improving drug solubility, and enhancing membrane permeability. Among those auxiliary agents that exhibit solubility-improving properties are cyclodextrins (CDs) being in focus of research for many decades. CDs can enhance the solubility of poorly soluble drugs via the formation of inclusion complexes,[3,4] raising the available concentration of these drugs at the surface of the absorption membrane.[5] In addition, even some inhibitory properties over intestinal P-glycoprotein were shown.[6] More recently, thiolated CDs moved into the limelight of pharmaceutical research because of their mucoadhesive behavior by the formation of disulfide bonds with mucus glycoproteins.[7−10] Although thiolated oligo- and polymers in general are well-known for their permeation-enhancing properties,[11] thiolated CDs have not been so far tested for these properties.

Therefore, the purpose of our study was to evaluate the permeation-enhancing properties of thiolated CDs. For this purpose, a highly thiolated CD, namely, a per-6-thiolated α-cyclodextrin (α-CD) was synthesized. A model BCS Class IV drug, furosemide (FUR), was chosen as this commonly used diuretic and antihypertensive drug is poorly soluble and exhibits a low permeability of the phospholipid bilayer of epithelial cells.[12] In addition, FUR is a substrate of P-glycoprotein that contributes also to its low oral bioavailability.[13] Drug permeation studies with and without per-6-thiolated α-CD were performed in vitro and in vivo.

Materials and Methods

Materials

α-CD (molecular mass: 972.84 Da), FUR, triphenylphosphine (Ph3P), iodine, thiourea, anhydrous N,N-dimethylformamide (DMF), dimethyl sulfoxide-d6 (DMSO-d6), anhydrous methanol, ethylenediaminetetraacetic acid (EDTA), sodium methoxide, sodium hydroxide (NaOH), potassium bisulfate (KHSO4), potassium hydroxide (KOH), 5,5′-dithiobis(2-nitrobenzoic acid) (Ellman’s reagent), Triton X-100, Spectra/Por dialysis membrane (molecular mass cut-off: 0.1–0.5 KDa), and minimum essential eagle medium (MEM) were obtained from Merck, Germany.

MEM powder (9.66 g/L) altered with Earle’s salts containing 19 amino acids, such as l-ala, l-asp, l-pro, l-glu, l-asn l-gly, and l-ser, 2 mM l-glutamine, sodium bicarbonate (NaHCO3) (2.2 g/L), 10% (v/v) fetal bovine serum (FBS), and phenol red was used as the cell culture medium. Phosphate buffered saline (PBS) and FBS were obtained from Invitrogen (Lofer, Austria).

Methods

Preparation of Iodinated α-CD

The primary hydroxyl groups of α-CD were replaced by a halogen, as described by Gadelle and Defaye.[14] In brief, 2.7 g of α-CD was added to 50 mM iodine and 50 mM triphenylphosphine solution while stirring in 50 mL of anhydrous DMF in a three-neck round-bottom flask (500 mL). The resulting solution was stirred at 80 °C under a nitrogen atmosphere for 18 h. The solution was then concentrated using a Hei-VAP Value Digital Rotary Evaporator (Heidolph Instruments GmbH & Co. Schwabach, Germany) to half the volume by the removal of DMF, and pH was adjusted between 9 and 10 by the addition of 10 mL of methanolic 3 M sodium methoxide on ice cool water. The reaction mixture was kept at 25 °C for half an hour to eliminate formate esters. It was then stirred for 5 min and poured into 500 mL of anhydrous methanol. The resulting yellow precipitate was filtered, washed with methanol after drying, and Soxhlet-extracted with anhydrous methanol for 24 h until the solvent became colorless. After rigorous drying, iodinated α-CD was obtained in 87% yield.

Synthesis of Per-6-thiolated α-CD

Iodinated α-CD was further treated with thiourea to substitute iodine with −SH groups.[15] Briefly, 950 mg of iodinated α-CD was dissolved in 10 mL of anhydrous DMF. Then, 300 mg of thiourea was added, and the resulting mixture was stirred at 70 °C under an inert environment for 15 h followed by the removal of DMF under reduced pressure by using a Hei-VAP Value Digital Rotary Evaporator (Heidolph Instruments, Schwabach, Germany). The remaining yellow oil was dissolved in 50 mL of water. Afterward, 250 mg of NaOH was added with heating under reflux and inert atmosphere for 1 h. The formed target compound was precipitated by acidifying with aqueous potassium bisulfate, filtered, thoroughly washed with demineralized water, and dried. The target compound was resuspended in 50 mL of demineralized water to remove the remaining traces of DMF. KOH was added until the target compound was completely dissolved, which was precipitated in the following again by acidifying with aqueous potassium bisulfate. This precipitate was filtered, redissolved in water, and dialyzed (MWCO of 0.1–0.5 KDa; cellulose membrane) for 3 days against demineralized water. Demineralized water was replaced every 8 h. The purified product was lyophilized to yield almost 80% per-6-thiolated α-CD.

Characterization of Per-6-thiolated α-CD

IR spectra of per-6-thiolated α-CD were analyzed by a Bruker ALPHA FT-IR apparatus (Billerica, USA). FT-IR spectra of test compounds were collected by placing these compounds on the tip of a platinum ART module. Results were obtained by 32 scans at a scanning speed of 4 cm–1 between 4000 and 500 cm–1. FT-IR spectra were presented as the average of these 32 scans.

1H NMR spectra were recorded using a Bruker-400 spectrometer (1H: 400 MHz) at 30 °C in 5 mm tubes containing the compounds dissolved in DMSO-d6. As the internal standard served the center of the DMSO-d6 multiplet that was correlated to tetramethylsilane with δ 2.49 ppm (1H).

Quantification of the −SH Content

Per-6-thiolated α-CD was assessed for the amount of thiol by using Ellman’s reagent.[16] The calibration curve was established using l-cysteine.

Cytotoxicity Studies

The impact of per-6-thiolated α-CD on viability of human colorectal carcinoma (Caco-2) cells was evaluated by resazurin assay to confirm the safety of the newly synthesized oligomer. For this purpose, Caco-2 cells were obtained from the European Collection of Cell Cultures (Salisbury, UK). Caco-2 cells at a density of 2.5 × 104 cells/well were suspended in 500 μL of MEM and seeded to a 24-well plate. The plate was incubated at an atmosphere of CO2 (5%) and relative humidity (95%) at 37 °C for 2 weeks. During this culture period, the medium was replaced on alternative days.

Toxicity assay was performed to investigate the cytotoxic potential of per-6-thiolated α-CD by quantifying resorufin formed by the reduction of resazurin. On the day of the experiment, cells of around 80% confluency were washed two times with preheated 10 mM PBS pH 7.4. The oligomer solutions (500 μL) of per-6-thiolated and unmodified α-CD (0.5% m/v) in white MEM, 4% (v/v) Triton X-100 acting as a negative control and white MEM acting as a positive control were transferred to the cell culture wells.

After 3 and 24 h, test samples were withdrawn, and the cells were washed two times with preheated PBS. In the following, 500 μL of resazurin (2.2 μM) solution was added, and cells were further incubated for 3 h. Aliquots (100 μL) were inserted to a 96-well plate, and the fluorescence was analyzed at emission and excitation wavelength of 590 and 540 nm utilizing a microplate reader (M-200 spectrometer; Tecan infinite, Austria). Cell viability of cells was measured using eq

Measurement of Thiol Stability

As the stability of sulfhydryl groups depends on thiolate anions concentration, the pKa value of −SH groups on per-6-thiolated α-CD was determined via an established photometric method with some modifications.[17] Briefly, ionic strength was stabilized by dissolving per-6-thiolated α-CD (0.1% m/v) in 0.001 M HCl containing 0.1 M NaCl. Increasing volumes of 0.01% NaOH were added to per-6-thiolated CD solutions that were continuously measured regarding pH and absorbance at 242 nm. The pKa was determined by plotting pH versus −log[(Amax – A)/A].

Furthermore, the stability of the thiolated CD was measured under different pH. In short, per-6-thiolated α-CD (0.25% m/v) was hydrated in water, and pH was adjusted to 4 and 5 with acetate buffer (50 mM), to 6 with phosphate buffer (50 mM), and to 7.2 with Tris buffer (50 mM). Test samples were kept under continuous stirring at 37 °C. After specific time intervals, aliquots of 500 μL were taken. Any further reaction was quenched by adding 50 μL of 1 M HCl to each aliquot.[18] Sulfhydryl groups were quantified using Ellman’s reagent.

Phase-Solubility Studies

Solubilizing properties of unmodified α-CD and per-6-thiolated α-CD for the model BCS Class IV drug FUR in aqueous solution were studied at room temperature according to an already established method.[19] In detail, FUR was added (CD and FUR in molar ratio of 1:2) to an aqueous solution containing increasing amounts of α-CD and per-6-thiolated α-CD ranging from 0.5 to 5 mM. The suspension was homogenized at room temperature using a thermomixer (Thermomixer C, Eppendorf, Austria) at 750 rpm until equilibrium in solubility was reached (3–4 days). After filtration (polyvinylidene fluoride filter with pore size: 0.45 μm), the obtained clear solutions were diluted 10 times, and FUR was quantified photometrically at 276 nm. The apparent stability constant (KC) was determined from the slope of the straight line using eq where S0 represents the solubility of FUR in water.

High-Performance Liquid Chromatography Analyses

The concentration of FUR was analyzed using high-performance liquid chromatography (HPLC) following a previous method[20] using a Hitachi LaChrom Elite HPLC-System equipped with a L-2200 autosampler, L-2130 pump, and L-2450 diode array detector. In brief, FUR-containing samples were analyzed on a reversed phase C-18 column being used as a stationary phase. Acetonitrile and 10 mM sodium dihydrogen phosphate pH 3.5 (70:30, v/v) at a flow rate of 1.0 mL/min was used as the mobile phase. The sample injection volume was 10 μL, and FUR was monitored at a wavelength of 229 nm.

Serial concentrations of the FUR in the range of 0.75–50 μg/mL (0.75, 1.5, 2.5, 5, 10, 20, 30, 40, 50 μg/mL) were prepared by spiking blank plasma samples with FUR stock solution in methanol. The calibration curve showed high linearity with a regression coefficient of 0.9795. The lower limit of quantification was 0.75 μg/mL determined after replicate injections of serial dilutions.

Inclusion Complex Formation of α-CD and Per-6-thiolated α-CD with FUR

The FUR−α-CD complex was formed according to the freeze-drying method.[21] In brief, equimolar amounts FUR and α-CD were added to demineralized water, and the resulting aqueous mixture was stirred for 4 days at 25 °C. Subsequently, the suspensions were filtered (polyvinylidene fluoride filter with pore size: 0.45 μm) and lyophilized to yield complexes of FUR with α-CD and per-6-thiolated α-CD, respectively. The amount of drug-loaded to the complexes was quantified by HPLC.

Rheological Investigation

The viscoelastic features of unmodified and per-6-thiolated α-CD in the presence of isolated intestinal mucus were evaluated. Intestinal mucus was purified as described previously.[22] It was freshly collected from the pig intestine, provided by a local slaughterhouse. The intestine was opened lengthwise and cut into smaller units to collect the mucus with a scraper. It was stirred on ice for 1 h after suspending in 0.1 M NaCl. Afterward, the suspension was centrifuged at 13,000g and 4 °C for 2 h. The purified mucus was separated by removing the supernatant.

Immediately after this mucus purification process, dynamic oscillatory measurements were executed on a plate–plate viscometer (HAAKE MARS Rheometer, Karlsruhe, Germany) within the viscoelastic region at a frequency of 1 Hz. Unmodified α-CD and per-6-thiolated α-CD in 0.2 M phosphate buffer pH 7 in 0.5% (m/v) concentration were homogenized with mucus in a ratio of 1:4. For disulfide bonds formation between the mucus glycoproteins and thiolated CDs, CD–mucus mixtures were incubated without stirring till 4 h at 37 °C. Thereafter, 0.5 mL of each mixture was placed on the viscometer plate for rheological analyses.

Permeation-Enhancing Properties

In Vitro Studies

In Vitro Permeation Studies on the Caco-2 Cell Monolayer

Permeation studies on the Caco-2 cell monolayer were executed as described previously.[23] Caco-2 cells at a density of 0.6 × 105 cells/cm2 were seeded onto 24-well plates with ThinCert inserts (pore-size: 400 μm, Greiner Bio-One, Kremsmünster, Austria). Cells were cultured in MEM containing 20% FCS in an atmosphere of 5% CO2 and 95% humidity at 37 °C. The medium was renewed every other day for 21 days. For permeation studies, only cell monolayers showing a transepithelial electrical resistance (TEER) in the range of 300–500 Ω cm2 were applied, which were determined with an EVOM instrument. TEER was also determined at time point 3 h representing the end of the permeation experiment and after 24 and 48 h to confirm monolayer integrity. Before the start of the experiment, cells were washed with PBS. Thereafter, 500 μL of MEM without FCS was transferred to the donor chamber and 1000 μL of the same medium was transferred to an accepter chamber. Upon equilibration for half an hour, the transport medium was replaced with 500 μL solution of unmodified α-CD–FUR and per-6-thiolated α-CD–FUR complexes corresponding to 0.05% (m/v) FUR in a final concentration. FUR solution (500 μL, 0.05% m/v) in white MEM was used as a control.

After every 30 min, aliquots of 100 μL were withdrawn from the acceptor chamber and replaced by the same volume of fresh preheated transport medium. At the end of the permeation experiment after 3 h, test solutions in the donor and acceptor compartments were replaced by fresh a prewarmed transparent medium. The amount of drug transported across the Caco-2 cell monolayer was analyzed through HPLC.

Apparent permeability coefficients (Papp) for FUR were measured as described previously[24] according to eq where “Q” is the total amount of FUR (μg) permeated the monolayer, “A” is the diffusion area (1.13 cm2), “c” is the FUR initial concentration (μg/cm3) in the donor chamber, and “t” is the time (s) of the permeation study.

The permeation enhancement ratio (R) was measured according to eq

In Vitro Permeation Studies on Intestinal Mucosa

For in vitro permeation studies, nonfasting Sprague-Dawley rats (250–300 g) were used. After sacrificing the rats, the gut was removed and kept in normal saline. The distal ileum and jejunum segments were cut into strips of about 1.5 cm, opened longitudinally, and rinsed with normal saline to free the luminal contents. Subsequently, the tissue was mounted in Ussing chambers with a permeation surface area of 0.64 cm2 without removing the underlying muscle layer. The apical side of the chamber was filled with 1 mL of medium containing 5 mM KCl, 138 mM NaCl, and 10 mM glucose buffered with 10 mM HEPES pH 6.8. The basolateral side of the chamber was also filled with 1 mL of the same medium but additionally containing 1 mM MgCl2 and 2 mM CaCl2. These salts were omitted in the apical chamber to prevent the complex formation of thiolated oligomers with calcium and magnesium. Gas consisting of 5% CO2 and 95% O2 was bubbled through each compartment to ensure oxygenation and agitation. Ussing chambers were equilibrated at 37 °C for 30 min to simulate physiological intestinal conditions. The buffer medium of the apical side of the chamber was then replaced with a solution of unmodified α-CD and per-6-thiolated α-CD complexed with FUR in 0.05% (m/v) concentration for permeation studies. FUR in concentration of 0.05% (m/v) served as a control. At predetermined time intervals, 100 μL aliquots were taken from the basolateral chamber and replaced by the same volume of fresh medium prewarmed at 37 °C. FUR having permeated the intestinal membrane was quantified by HPLC. Overall corrections were made for withdrawn aliquots. The apparent permeability coefficients (Papp) and permeation enhancement ratio (R) were calculated as described above.

In Vivo Studies

All animal experiments were performed according to the Principles of Laboratory Animal Care as approved by the Animal Ethical Committee of Austria (BMBWF-66.008/0035-V/3b/2019). For the in vivo studies to access oral permeation enhancement of the formulations, Sprague-Dawley rats (average weight 250–300 g) were used. Rats were randomly divided into four groups: first group (n = 6) for an unmodified α-CD–FUR complex, second group (n = 6) for per-6-thiolated α-CD–FUR, and third group (n = 6) for positive control having FUR. All three groups were dosed via oral gavage. The fourth group (n = 3) received FUR in sterile 10 mM PBS pH 7.4 via iv injection. No food was supplied 2 h prior to dosing, but water was provided ad libitum. The control and test groups were treated with a dose of 30 mg of FUR per kg body weight. Blood samples (200 μL) were collected from the prominent lateral tail veins into EDTA-treated microcontainers at predetermined time points of 0, 1, 2, 3, 5, 8, and 24 h. To separate plasma, samples were centrifuged at 3,000 rpm for 10 min and kept at −20 °C until further analysis.

Drug Extraction from Plasma

FUR was extracted from plasma with a solid-phase extraction (SPE) method, as reported previously in detail[25] by using Oasis HLB cartridges (Waters Corporation, USA). As FUR is extensively bound to serum proteins (albumin), samples were acidified with 5 M HCl prior to extraction in order to disrupt drug–protein interactions. First, the Oasis HLB cartridges were conditioned with methanol and then equilibrated with demineralized water. The 100 μL of the mobile phase (NaH2PO4/acetonitrile: 30/70, v/v) was added to the prepared sample (200 μL). After vortexing for 5–10 min, the sample was added to the cartridge and then washed with 1000 μL of methanol, and the eluate was evaporated. The obtained residues were then reconstituted in 0.2 mL of the mobile phase for further analysis with HPLC as described above. FUR recovery from plasma was found 89%.

Pharmacokinetic Analyses

Pharmacokinetic analyses were applied to the plasma concentration–time data in a noncompartmental manner to obtain PK parameters of FUR after intravenous and oral application of samples. The software PKSolver was utilized for this purpose. Relative bioavailability was evaluated according to eq

Statistical Analyses

Statistical analyses of all data were performed by using the student’s t-test. A confidence interval of p < 0.05 was used for the analysis of two groups. One-way ANOVA was applied to compare different group difference with 95% CI (confident interval). (GraphPad Prism, GraphPad Software, Inc.). All results were shown as the mean (±SD), n = 3.

Results and Discussion

Synthesis of Per-6-thiolated α-CD

α-CD was chosen for per-6-modification of its primary −OH groups applying a selective reaction.[26] In detail, α-CD was directly halogenated in the C-6 positions by substitution of primary −OH groups with iodine using triphenylphosphine (Ph3P) in anhydrous DMF, as illustrated in Figure . To activate primary −OH groups, a phosphonium salt was formed via this reaction.[27] The method was chosen as comparatively less reactive reagents attack more selectively the C-6-position −OH groups.[28] Moreover, halogenation of CDs is temperature-dependent as temperatures above 90 °C degrade CDs, resulting in a lower degree of substitution (DS).[7] To obtain on the one hand a high DS value and to avoid on the other hand the degradation of CD, the reaction was performed at 80 °C in step one and continued at 70 °C in step two. As per-6-iodinated α-CD was shown to be more stable and soluble in DMF,[28] it was used as a solvent. In the second step, per-6-iodinated α-CD was deiodinated with thiourea to generate per-6-thiolated α-CD.

Figure 1

Synthesis pathway of per-6-thiolated α-CD. The first step is the formation of iodinated α-CD in the presence of Ph3P followed by the formation of per-6-thiolated α-CD in the presence of thiourea.

Per-6-thiolated α-CD was prepared as odorless powder. The covalent attachment of sulfhydryl groups was verified by FTIR. The overlaid FT-IR spectra of per-6-iodinated, per-6-thiolated, and unmodified α-CDs derivatives are given in Figure S1. The IR spectrum of the per-6-thiolated α-CD derivative significantly differed from unmodified α-CD. The −OH group stretching vibrations at 3300–3400 cm–1 is a characteristic peak of unmodified α-CD. Because of C–H stretching, an intense peak between 2924 and 2931 cm–1 was identified for all modified α-CD derivatives. Additionally, strong bands at 1136 cm–1 and moderate bands at 610 cm–1 can be explained by the C–S stretching.[8]

Moreover, per-6-thiolated α-CD was also analyzed via 1H NMR. The 1H NMR spectrum of per-6-thiolated α-CD showed signals at 8.6, 8.2, and 7.3 ppm, likely due to the protons of the sulfhydryl groups, as shown in Figure S2.[8]

The amount of −SH and −S–S– groups on the CD backbone was determined to be 4244 ± 402 and 786 ± 139 μmol/g, respectively.

Cytotoxicity of Per-6-thiolated α-CDs

The effect of per-6-thiolated α-CDs on Caco-2 cells was analyzed by the resazurin assay to assess cell viability. This assay works on the principle that resazurin is reduced into resorufin by viable cells. The color of the dye changes from its oxidized blue color to a reduced pink color.[29] As dead cells cannot reduce resazurin, the dye serves as an indicator for cell viability. Cells were treated with 0.5% (m/v) per-6-thiolated α-CD. Cells showed >88% viability at the tested concentration after 3 and 24 h without any signs of significant cytotoxicity, as depicted in Figure . Per-6-thiolated α-CDs were synthesized without ring-opening as it was the case for oligomers that were found toxic at higher concentrations.[8]

Figure 2

Cell viability of Caco-2 cells with per-6-thiolated α-CD (0.5% m/v) and unmodified α-CD after 3 and 24 h. White MEM (negative control) and Triton X-100 (4% m/v) (positive control) was used. All the results are shown as mean ± SD, n = 3.

Stability of Thiol Groups

As the conversion of sulfhydryl ligands to thiolate ions (−S–) is the important step for the formation of disulfide bonds, the pKa of the sulfhydryl ligand was calculated. Thiol groups (−SH) exhibit a covalent bond (nonpolar) between S and H due to a minor difference in electronegativity that makes it easy for sulfur to lose its proton (H+) to form a thiolate anion (−S–). This pH-dependent deprotonation increases the reactivity of thiol groups toward electrophiles. An increase in pH leads to higher absorbance, which reflects the increase in −S– anions, as reflected in Figure a. The pKa of per-6-thiolated α-CD was measured by plotting pH versus −log value of [(Amax – A)/A], as shown in Figure b. The pKa of per-6-thiolated α-CD was found to be 8.1 that is almost the same as that of free thiols of cysteine.[30]

Figure 3

Absorbance of the per-6-thiolated α-CD at different pH [A] and pKa determination by plotting pH vs −log value of (Amax – A)/A [B].

Furthermore, the stability of per-6-thiolated α-CD was measured at different pH. At pH 4, the −SH groups of per-6-thiolated α-CD were found stable, whereas with an increase in pH over 5, a decrease in the sulfhydryl groups was noted as illustrated in Figure . After 3 h at pH 5, already 15% of −SH groups were oxidized due to decrease in proton concentration that results in an increased concentration of −S– anions that represent the active form for oxidation. The reactivity of sulfhydryl ligands is highly pH-dependent and can be controlled by pH adjustment.[31] At pH > 5, the thiolated oligomer formed inter- and intramolecular disulfide bonds. On contrary, at pH < 5 disulfide bond formation can almost be eliminated due to less thiol-anion concentration.[18]

Figure 4

Stability of −SH groups on per-6-thiolated α-CD (0.25% m/v) toward oxidation at 37 °C at pH 7.2 (▼), pH 6 (▲), pH 5 (■), and pH 4 (●). All the results are shown as mean ± SD, n = 3.

Solubilizing Properties of Per-6-thiolated α-CD

As CDs were identified as solubility enhancers for poorly soluble drugs,[3,4] the effect of unmodified and per-6-thiolated α-CD on the solubility of model BCS Class IV drug FUR in aqueous solution was investigated. Interactions of the drug with CDs displayed typical AL type solubility curves, indicating a soluble binary complex formation at a stoichiometric ratio of 1:1. Both unmodified and per-6-thiolated α-CD displayed almost the same solubility-improving effect for FUR, as depicted in Figure . The apparent stability constant (KC) values of α-CD and per-6-thiolated α-CD with FUR were determined to be 1050.7 and 1090.9 M–1, respectively. The successful oral delivery of BCS Class IV drugs depends upon their dissolution properties and membrane permeability. Per-6-thiolated CD increases the solubility of FUR by entrapping FUR in its lipophilic cavity.

Figure 5

Phase solubility studies: effect of unmodified and per-6-thiolated α-CD on the solubility of FUR. All the results are shown as mean ± SD, n = 3.

Interactions of Per-6-thiolated α-CD with Mucus

The dynamic viscosity of thiolated oligomer/mucin mixtures strongly depends on interactions that are taking place between the two components like covalent and noncovalent interactions.[32] The impact of unmodified and modified oligomers on the dynamic viscosity of intestinal mucus used as a model was therefore evaluated. In the presence of per-6-thiolated α-CD, dynamic viscosity of mucus was increased 5.8-fold, whereas with unmodified α-CD just an increase of 1.7- fold was noted. Results as depicted in Figure provide evidence for the disulfide bond formation between mucus and thiolated oligomer.

Figure 6

Dynamic viscosity (Pas) of per-6-thiolated and unmodified α-CD (0.5% m/v) in the presence of mucus on a plate–plate viscometer at 37 °C at a frequency of 1 Hz in 0.2 M PBS at pH 7. All the results are shown as the mean ± SD (n = 3), (**P < 0.01).

In a previous study,[7] thiolated CDs exhibited on average just one −SH group per CD molecule. As these thiolated CDs could not crosslink with the mucus, they had much more the character of mucolytic agents such as N-acetylcysteine cleaving disulfide bonds within the mucus. Because of the much higher degree of thiolation of per-6-thiolated α-CD, however, a pronounced increase in viscosity with mucus was observed. As highly thiolated CD forms obviously more than just one new disulfide bond with mucus glycoprotein, it caused a crosslinking of mucins. Furthermore, it is confirmed that the increase in rheological behavior of thiomers–mucus mixtures indicates high mucoadhesivity,[33] and thiolated CDs showed enhanced mucosal residence due to these high mucoadhesive properties.[7,8]

The results of transport studies on Caco-2 cells are shown in Figure . The absorptive transport of FUR was performed in the presence of 0.05% (m/v) of FUR complexed with the unmodified and thiolated CD. The unmodified α-CD–FUR complex had only a minor effect on the permeation properties of FUR, as shown in Table , whereas the Papp value strongly increased in the presence of a per-6-thiolated α-CD–FUR complex. The transport enhancement of FUR due to per-6-thiolated α-CD was 6.87-fold in comparison to control. The mechanism of enhanced permeation is due to the inhibitory effect of the protein tyrosine phosphatase (PTP) that is involved in tight junction (TJ) opening.[34] This enzyme dephosphorylates the transmembrane protein occludin that builds up TJs. Phosphorylation of occludin via protein tyrosine kinase leads to an opening of TJs, and PTP dephosphorylates these groups resulting in a closing of TJs.[35] As a consequence, PTP inhibition leads to phosphorylation and opening of TJs, favoring an improved paracellular transport.[36] PTP bears cysteine for its activity that is inhibited by reduced glutathione (GSH) via disulfide bond formation. Thiomers convert GSSG (oxidized glutathione) to GSH that in turn inhibits the enzyme.[34] Furthermore, the prolonged residence time of thiomers on the mucosa[7] can raise the GSH concentration that results in TJ opening.

Figure 7

Permeation enhancement of per-6-thiolated oligomer on diffusion of FUR on Caco-2 cells. Results show the permeation of FUR on Caco-2 cells indicated as the %age of the total concentration of FUR applied as control 0.05% (m/v) FUR (▲), 0.05% (m/v) unmodified α-CD–FUR complex (□), and 0.05% (m/v) per-6-thiolated α-CD–FUR complex (●). All the results are shown as mean ± SD, n = 3, (***P < 0.001).

Table 1

Comparison of Unmodified and Per-6-thiolated α-CD Complexed with FUR 0.05% (m/v) on Apparent Permeability Coefficients (Papp) of FUR and Transport Enhancement Ratio (R) on Caco-2 Cell and Rat Gut Mucosa (Mean ± SD, n = 3)

	apparent permeability coefficient [Papp × 10–6 (cm/s)]		transport enhancement ratio R = [Papp(sample)/Papp(control)]
sample	Caco-2 monolayer	rat intestinal mucosa	Caco-2 monolayer	rat intestinal mucosa
FUR	0.81 ± 0.2	0.65 ± 0.2
unmodified α-CD–FUR complex	1.63 ± 1.0	1.14 ± 0.5	2.01	1.76
per-6-thiolated α-CD–FUR complex	5.57 ± 2.1	4.26 ± 1.1	6.87	6.55

Recently, an additional process of TJ opening was discovered for thiolated polymers.[37] When thiomers get into contact with the epithelium, they tend to interact with receptors like epidermal growth factor receptor due to the interaction between their −SH groups and cysteine of these receptors. By phosphorylation, this results in activation of protein tyrosine kinases Src. Phosphorylated Src regulates in turn claudin-4 proteins and results in TJ opening. Moreover, chemical substructures on thiolated polymers like −NH2 groups additionally control TJ function by receptor interaction.

As TEER measurements report TJ integrity,[38] they were performed in connection with transport studies on the Caco-2 cells. As shown in Figure , TEER of the control was constant during the whole experiment, whereas the electrical resistance of the monolayer treated with unmodified α-CD decreased almost 15%. On the contrary, in the presence of per-6-thiolated α-CD, a continuous and pronounced decrease in TEER up to 28% was observed, confirming the opening of TJs. After cell washing and incubating with MEM, TEER regained values of 82 and 90% in comparison to initial values after 24 and 48 h, respectively. TJ opening was therefore reversible, and the Caco-2 cell monolayer remained intact. A relation of the minor cytotoxic effect of the thiolated CDs, as shown in Figure , with this TJ opening effect can therefore be excluded. In case of other TJ opening agents like latrunculin A or N-ethylmaleimide, for instance, such a reversible effect could not be shown.[39,40]

Figure 8

TEER measurements with FUR, unmodified α-CD, and per-6-thiolated α-CD for the permeation study before the start of the experiment (white bars), at time point 3 h representing the end of the permeation experiment when test solutions were replaced (dark gray bars), after 24 h (gray bars) and after 48 h (black bars). Results are presented as mean (±SD), n = 3.

Permeation studies on rat gut were performed in Ussing chambers.[41] As CDs are also known as permeation enhancers,[5] the effect of both unmodified and thiolated α-CD was tested. Permeation enhancement of per-6-thiolated α-CD was much more pronounced than that of unmodified α-CD. In the presence of per-6-thiolated α-CD, permeation of FUR was 6.55-fold increased as compared to control, as expressed in Figure . These results are similar to the results of transport studies across the Caco-2 cell monolayer. The Papp value of per-6-thiolated α-CD was found to be 4.26 × 10–6 cm/s, which was significantly higher as compared to control. The permeation enhancement ratio (R) of per-6-thiolated α-CD is shown in Table . The permeation enhancement of per-6-thiolated α-CD was found to be even more pronounced than that of other thiomers, for instance, thiolated poly acrylic acid[1] and polysulfonate thiomers.[11]

Figure 9

Transport of FUR on gut mucosa of nonfasting Sprague-Dawley rats. Permeation data are shown as %age of the total dose of FUR applied as control 0.05% (m/v) FUR (▲), 0.05% (m/v) unmodified α-CD–FUR complex (□), and 0.05% (m/v) per-6-thiolated α-CD–FUR complex (●). All the results are presented as the mean ± SD, n = 3, (***P < 0.001).

In Vivo Proof-of-Concept

Plasma concentration–time profiles of all samples administered intravenously and orally with an equal concentration of FUR (30 mg/kg)[42] are displayed in Figure . The PK-parameters are summarized in Table .

Figure 10

Comparative rat plasma concentration–time profile of FUR: equivalent concentrations of FUR (30 mg/kg) complexed with α-CD and per-6-thiolated α-CD were administered to three groups (n = 6) orally and fourth group (n = 3) was dosed with FUR intravenously. Blood was collected from all groups at indicated time points. FUR was extracted from plasma via the SPE method, and the concentration of FUR was analyzed by HPLC. All indicated values are means ± SD, (***P < 0.001).

Table 2

Pharmacokinetic Parameters of Per-6-thiolated α-CD–FUR Complex, α-CD–FUR Complex, and FUR Obtained after Noncompartmental Analysis by PKSolver

	formulations
parameters	FUR	α-CD–FUR complex	per-6-thiolated α-CD–FUR complex
dose (mg/kg)	30.0	30.0	30.0
t1/2 (h)	8.50	5.70	6.36
Tmax (h)	2.00	3.00	3.00
Cmax (μg/mL)	3.87	12.9	36.6
AUC0–24 (μg/mL h)	32.1	86.6	159
Vd/F (mg/kg)/(μg/mL)	9.11	2.72	1.62
Cl/F (mg/kg)/(μg/mL)/h	0.75	0.33	0.18
Fa (%)	4.00	11.0	19.8

For relative oral bioavailability (F) calculation, AUCiv after administration of 30 mg of FUR per kg body weight was 801 ± 74 μg/mL h.

FUR was poorly absorbed after 2 h of oral administration, as exhibited by a peak plasma concentration (Cmax) of 3.87 μg/mL showing a low AUC and bioavailability. On the contrary, α-CD–FUR and per-6-thiolated α-CD–FUR complexes showed enhanced Cmax, AUC, and bioavailability following oral administration. The per-6-thiolated α-CD–FUR complex increased FUR Cmax and AUC by 8.06-fold and 4.95-fold, respectively. Furthermore, 19.8% relative bioavailability (F) was obtained for the per-6-thiolated α-CD–FUR complex as compared to 4% in case of the FUR control formulation.

This improvement in oral bioavailability of FUR by per-6-thiolated α-CD is likely caused by various factors. First, the thiolated oligomer is highly mucoadhesive due to the high degree of thiolation providing an intimate contact with the intestinal mucosa to enhance the gastrointestinal residence time. As absorption window of FUR is in the proximal segment of the duodenum,[43] 2.5-fold higher drug concentration in plasma 8 h after administration can likely only be explained by the improved mucoadhesive properties of the thiolated CD. Moreover, an enhanced drug concentration at the absorption site facilitates drug transport. Second, solubility improvement by per-6-thiolated α-CD also contributes to a raised bioavailability of FUR. Third, per-6-thiolated α-CD exhibits permeation-enhancing properties by the opening of TJs. All these factors make per-6-thiolated α-CD an ideal excipient for enhancing FUR bioavailability. Furthermore, in vivo results were also in good correlation with ex vivo results.

Conclusions

In the present study, all primary hydroxyl groups of α-CD were replaced with sulfhydryl groups to produce nontoxic per-6-thiolated α-CD. The dynamic viscosity of mucus was increased 5.8-fold with thiolated CD. The thiomer showed a marked enhancement in permeation of FUR on Caco-2 cells and rat gut mucosa. Moreover, per-6-thiolated α-CD showed increased solubility of FUR as compared to FUR as a control. Furthermore, per-6-thiolated α-CDs improved the bioavailability of FUR 4.95-fold in vivo. Per-6-thiolated α-CD is the first functionalized CD exhibiting solubility-improving, mucoadhesive, and permeation-enhancing properties. This multifunctionality makes it likely to a powerful tool for the oral application of BCS Class IV drugs.