Solubilized Ubiquinol for Preserving Corneal Function

Abstract

Defective cellular metabolism, impaired mitochondrial function, and increased cell death are major problems that adversely affect donor tissues during hypothermic preservation prior to transplantation. These problems are thought to arise from accumulated reactive oxygen species (ROS) inside cells. Oxidative stress acting on the cells of organs and tissues preserved in hypothermic conditions before surgery, as is the case for cornea transplantation, is thought to be a major reason behind cell death prior to surgery and decreased graft survival after transplantation. We have recently discovered that ubiquinol – the reduced and active form of coenzyme Q10 and a powerful antioxidant – significantly enhances mitochondrial function and reduces apoptosis in human donor corneal endothelial cells. However, ubiquinol is highly lipophilic, underscoring the need for an aqueous-based formulation of this molecule. Herein, we report a highly dispersible and stable formulation comprising a complex of ubiquinol and gamma cyclodextrin (γ-CD) for use in aqueous-phase ophthalmic products. Docking studies showed that γ-CD has the strongest binding affinity with ubiquinol compared to α- or β-CD. Complexed ubiquinol showed significantly higher stability compared to free ubiquinol in different aqueous ophthalmic products including Optisol-GS® corneal storage medium, balanced salt solution for intraocular irrigation, and topical Refresh® artificial tear eye drops. Greater ROS scavenging activity was noted in a cell model with high basal metabolism and ROS generation (A549) and in HCEC-B4G12 human corneal endothelial cells after treatment with ubiquinol/γ-CD compared to free ubiquinol. Furthermore, complexed ubiquinol was more effective at lowering ROS, and at far lower concentrations, compared to free ubiquinol. Complexed ubiquinol inhibited lipid peroxidation and protected HCEC-B4G12 cells against erastin-induced ferroptosis. No evidence of cellular toxicity was detected in HCEC-B4G12 cells after treatment with complexed ubiquinol. Using a vertical diffusion system, a topically applied inclusion complex of γ-CD and a lipophilic dye (coumarin-6) demonstrated transcomeal penetrance in porcine corneas and the capacity for the γ-CD vehicle to deliver drug to the corneal endothelium. Using the same model, topically applied ubiquinol/γ-CD complex penetrated the entire thickness of human donor corneas with significantly greater ubiquinol retention in the endothelium compared to free ubiquinol. Lastly, the penetrance of ubiquinol/γ-CD complex was assayed using human donor corneas preserved for 7 days in Optisol-GS® per standard industry practices, and demonstrated higher amounts of ubiquinol retained in the corneal endothelium compared to free ubiquinol. In summary, ubiquinol complexed with γ-CD is a highly stable composition that can be incorporated into a variety of aqueous-phase products for ophthalmic use including donor corneal storage media and topical eye drops to scavenge ROS and protect comeal endothelial cells against oxidative damage.

INTRODUCTION

Tissue and organ preservation is essential for successful transplant surgery. Preservation in cold storage conditions after procurement is a common practice to minimize energy consumption and slow metabolic activity. However, cold temperature paradoxically contributes to the formation of reactive oxygen species (ROS) and cytokines that lead to graft dysfunction and failure [1–8]. In the case of corneal transplantation (51,294 performed in 2018 in the U.S.), the survival of the avascular graft is inversely proportional to the amount of time that the tissue spends preserved in media at 4° C before surgery [9]. Mitochondria-rich corneal endothelial cells that line the inner cornea and pump ions to maintain corneal transparency undergo a specific form of non-apoptotic cell death – ferroptosis – that can be induced by hypothermia and is mediated by iron and lipid peroxidation [10, 11]. Additionally, oxygen levels throughout preservation remain at least 3 times higher than physiologic concentrations for corneal endothelial cells stored in the most commonly used preservation media, increasing oxidative stress on stored tissue [12]. Despite the presence of constituent antioxidants in corneal storage media such as β-mercaptoethanol and glutathione added for defense, the development of specific molecular strategies to mitigate ROS accumulation in cold storage has been limited to date.

Ferroptosis can be inhibited by ubiquinol, the reduced and active form of coenzyme Q10 (CoQ10) [13, 14]. CoQ10 has at least three major actions, serving as i) an essential carrier of electrons in mitochondrial oxidative phosphorylation for ATP synthesis, ii) an antioxidant in the cytosol by trapping peroxyl radicals that mediate lipid peroxidation [13, 15–19], and iii) a direct inhibitor of mitochondrial permeability transition pore (PTP) opening and apoptosis, independent of its ROS scavenging activity [20]. As an antioxidant, ubiquinol is an essential participant in the FSP1-CoQ10-NAD(P)H pathway, an independent system that works in parallel with GPX4 and glutathione to suppress phospholipid peroxidation and ferroptosis [13, 21], wherein ferroptosis suppressor protein 1 (FSP1) catalyzes NADPH reduction of ubiquinone (the oxidized form of CoQ10) and endogenous ubiquinol is recycled. Ferroptosis inhibition has been proposed as a treatment strategy specifically for corneal endothelial cell pathologies related to oxidative stress [22, 23]. As a potent antioxidant and inhibitor of cell death, ubiquinol is a promising therapeutic molecule for protecting cells with high exposure to stressful conditions (oxygen, ultraviolet) that generate ROS such as corneal endothelial cells [24–26]. However, ubiquinol is unstable, owing to its reactivity to oxygen and light, and its extreme lipophilicity limits its solubility, bioavailability, and practical use. As we found recently [12], ubiquinol (10 μM) boosts mitochondrial spare respiratory capacity and reduces apoptosis in transplant suitable donor corneal endothelial cells; however, absolute ethanol maintained at 37° C was required to solubilize and prevent immediate drug precipitation when used as a supplement to corneal storage media. While ubiquinol is considered the superior form of Coenzyme Q10 as it is readily active [27, 28], poor aqueous solubility and instability in water, oxygen and light are major hurdles preventing rigorous investigation and therapeutic translation in organ storage.

In order to solubilize the reduced form of CoQ10 in a single-drug formulation, our team has developed a supramolecular inclusion complex of ubiquinol and cyclodextrin, a well-characterized excipient used previously in aqueous ophthalmic formulations [29]. Cyclodextrins (CDs) are cyclic oligosaccharides formed of 6, 7, or 8 (α-1,4-linked) α-D-glycopyranose (dextrose) units (α-, β-, or γ-CDs, respectively) via enzymatic degradation of starch [30, 31]. CDs have hydrophobic inner cavities allowing them to form inclusion complexes that host many “guest” lipophilic drug molecules, while their hydrophilic surfaces provide better aqueous solubility and dissolution rates of poorly soluble drugs [32–35]. CDs have been widely employed to improve oral bioavailability [36, 37], mask undesirable taste [38, 39], enhance stability [31, 39], improve pharmacological activity [34, 40], and reduce the toxicity of many drugs [41]. Herein, we report the preparation and characterization of a highly water dispersible ubiquinol/γ-CD complex, and stability of the drug complex in aqueous media (Optisol-GS® corneal storage media, balanced salt solution, and Refresh® artificial tears). We demonstrate the ubiquinol/γ-CD complex increases ROS scavenging activity using A549 human adenocarinomic alveolar epithelial cells, which generate elevated levels of ROS, and that this ubiquinol complex scavenges ROS and inhibits lipid peroxidation in HCEC-B4G12 human corneal endothelial cells (B4G12), a genetically homogeneous model with high fidelity to native cells. Importantly, we also report the anti-ferroptotic activity of the complex in B4G12 cells. Finally, we tested transcorneal penetrance using a topical application model and donor tissue penetrance using a cornea storage model in porcine and human corneas, and demonstrate increased drug bioavailability to corneal endothelial cells using the ubiquinol/γ-CD complex compared to free ubiquinol in both topical and media applications. Creating this highly dispersible form of ubiquinol readily available in the aqueous phase will permit its addition to corneal storage media and ophthalmic formulations such as eye drops to provide a powerful, stable, and easy to handle therapeutic drug that protects corneal endothelial cell function in conditions characterized by ROS accumulation and oxidative damage.

METHODS

All experimental procedures conformed to the tenets of the Declaration of Helsinki. The Institutional Review Board at the University of Iowa determined that approval was not required for this study.

Materials

Ubiquinol was purchased from Sigma Aldrich (USP analytical standard, St. Louis, MO). Coenzyme Q10 and alpha-, beta-, and gamma cyclodextrin (α-, β-, and γ-CD, respectively) were purchased from TCI America (Portland, OR). Dihydroethidium (DHE)-based assay kits were purchased from Abcam (Cambridge, MA), and used according to the manufacturer’s protocol. Coenzyme Q9 was purchased from Cayman Chemicals (Ann Arbor, MI). C11-Bodipy 581/591 lipid peroxidation assay kits were purchased from Thermo Fisher (Waltham, MA). All other reagents and solvents were at least of analytical grade and used as received.

Molecular Docking Study

The 3D structures of α, β and γ-CDs (PDB ID: 4FEM, 1Z0N, and 5E70, respectively) were obtained from the RCSB Protein Data Bank (PDB) database (http://www.rcsb.org). BIOVIA Discovery Studio Visualizer v20.1.0.19295 (Accelrys, San Diego, CA) software was used for the 3D structure preparation. The 2D structure of ubiquinol [PubChem: 9962735] and 3D conformation of coumarin-6 [PubChem: 100334] were retrieved from NCBI-PubChem Compound database [42]. All required file conversions and 3D structure generations of ubiquinol were performed using the open source chemical toolbox Open Babel version 2.3.1 (www.openbabel.org) [43]. Geometry optimization and conformation searches were carried out by Avogadro 1.2.0 (http://avogadro.cc/) [44]. The Merck molecular force field 94 (MMFF94) was used to perform energy minimization of ligands, and the charge calculation method used was Gasteiger. In the present investigation, α, β and γ-cyclodextrin were considered as targets (hosts), and ubiquinol and coumarin-6 were considered as ligands (guests). All rotatable bonds of the ligands were considered as non-rotatable to allow rigid docking and minimize standard errors [45]. Missing residues and hydrogen atoms were supplemented to the protein structure using AutoDock Tools (ADT, Scripps Research Institute, La Jolla, CA), a free graphic user interface of MGL software packages v1.5.7rc1 [46]. Grid box parameters were set by AutoDock Tools (ADT) to accommodate both target and ligand. Docking was operated according to the previously published literature [47]. The molecular docking program AutoDock Vina (version 1.1.2) was used to perform the docking simulation [45]. Molecular Docking simulations were carried out utilizing the Lamarckian genetic algorithm (LGA) to explore the best conformational space for the ligand with a population size of 150 individuals. In the current study, maximum numbers of generation and evaluation were set at 27,000 and 2,500,000, respectively. The inclusion complex conformation was selected according to the binding energy and optimum scoring pose. Docking results were analyzed in BIOVIA Discovery Studio Visualizer and PyMOL molecular graphic system, version 2.3 (https://pymol.org/). Grid box parameters selected for targets α-, β- and γ-cyclodextrins are displayed in Table S1.

All subsequent assays were executed using the most appropriate host molecule for the target molecule, ubiquinol (see Results).

Phase Solubility Study

Different concentrations of γ-CD solutions (0, 10, 20, 30, 40, and 50 mM) were prepared in Nanopure water (Bamstead Thermolyne Nanopure water purification system, Thermo Fisher). Excess amounts of ubiquinol were dispersed in these solutions in Eppendorf tubes; the tubes were shaken (3000 rpm) at room temperature for 3 days, then they were left without being shaken for 1 day. All of these steps were carried out protected from light. Later, these dispersions were centrifuged (21,000 xg, 10 min) and 50 μl aliquots of each solution were collected, diluted, and measured using high performance liquid chromatography (HPLC) as described below.

Preparation of the Ubiquinol/γ-CD Complex

In order to overcome ubiquinol’s high reactivity to oxygen and light, we utilized the kneading method to complex ubiquinol with the cyclodextrin excipient. Ubiquinol was mixed with γ-CD in a drug:CD molar ratio of 1:10 using geometric dilution in a porcelain mortar. A hydro-alcoholic solution (1:1) was added portion-wise to the mixture, then mixed using a pestle until a semi-liquid paste was formed. Mixing was continued for about 1 hour in dark conditions, and finally the mixture was vacuum dried to produce a white to off-white powdered complex.

Characterization of the Complex

Morphology.

The morphology of the formed complex was examined using scanning electron microscopy (SEM) and compared to the morphologies of ubiquinol, γ-CD, and their physical mixture (ubiquinol:γ-CD molar ratio of 1:10). The powdered samples were loaded onto an aluminum SEM stub, then sputter-coated with gold and palladium using an Emitech K550 sputter-coater. Images were captured using a Hitachi S-4800 scanning electron microscope operated at 5 kV accelerating voltage (Hitachi High Technologies America Inc., Schaumburg, IL).

Differential scanning calorimetry (DSC).

Accurately weighed amounts of ubiquinol, γ-CD, their physical mixture (ubiquinol:γ-CD molar ratio of 1:10), and the complex (ubiquinol:γ-CD molar ratio of 1:10) were loaded in crimped pans and DSC thermograms were obtained using a TA Instruments model Q20 DSC (New Castle, DE). A temperature ramp rate of 5° C/min, within a range of 0 to 75° C, was used.

X-Ray diffraction (XRD).

An X-ray diffractometer (Bruker D8 Advance X-ray diffractometer, Bruker AXS, Inc., Madison, WI) was used to analyze the crystal structures of ubiquinol, γ-CD, ubiquinol/γ-CD physical mixture (1:10 molar ratio), and ubiquinol/γ-CD complex (1:10 molar ratio).

High Performance Liquid Chromatography (HPLC)

The Agilent 1100 series HPLC station with a Waters RP-C18 4.6 x 150 mm column and pore size of 5 μm, set at room temperature, was used to measure the concentration of ubiquinol and ubiquinone. The mobile phase consisted of acetonitrile:THF:water (at a ratio of 60:35:5) with a flow rate of 1 ml/min. Detection was carried out using an Agilent diode array detector set at a wavelength of 290 nm for ubiquinol, and 280 nm for ubiquinone and coenzyme Q9 (internal standard, IS). The injection volume was set to 50 μl.

Stability of the Complex in Different Media

Equal amounts of free ubiquinol and complexed ubiquinol were dispersed in tubes each containing 0.5 ml of either Optisol-GS® (Bausch+Lomb, Rochester, NY), BSS (Balanced salt solution, Alcon, Fort Worth, TX), or Refresh® (Allergan US, Madison, NJ) by means of vortexing and sonication for 5 min, then stored at 4° C. At predetermined time points (0 day, 4 days, and 1 week), each sample was spiked with 10 μl of 1 mg/ml solution of coenzyme Q9 (Internal standard, IS) dissolved in THF:acetonitrile at a ratio of 38:62, then two ml of ethyl acetate were added to each tube. Liquid/liquid extraction was performed twice on each tube to extract ubiquinol by vortexing for 5 min, to collect a total of 4 ml of ethyl acetate, which were then evaporated under nitrogen (under light protection conditions). The residues were then reconstituted in THF:acetonitrile:water at a ratio of 87.5:150:12.5, respectively, then injected in the HPLC to quantify ubiquinol, ubiquinone, and total coenzyme Q10 in each sample as described above. The percentage remaining was calculated as follows: % remaining = amount remaining (mg) x 100/initial amount (mg).

ROS Assay in A549 Epithelial Cells

A549 human adenocarcinoma alveolar epithelial cells (gifted from Dr. Meng Wu, University of Iowa) were seeded in RPMI medium supplemented with 10% v/v FBS and 1% penicillin/streptomycin (Life Sciences, Waltham, MA) in 6-well plates at 200,000 cells/well for approximately 48 hours (37° C, 5 % CO2). The medium was then removed, and the treatments were added in 4 ml/well each (n = 3). Treatments added were (1) ubiquinol dispersed in RPMI medium at three different concentrations (100, 50, and 10 μM), (2) ubiquinol/γ-CD complex (molar ratio of 1:10) dispersed in RPMI medium at three different equivalent concentrations (100, 50, and 10 γM), and (3) γ-CD dissolved in RPMI at concentrations equivalent to those added with 100 γM of ubiquinol in the previous groups. Six wells were left untreated. After 24 h, media was removed, and wells were washed with 5 mM sodium pyruvate in PBS, then cells were detached with 0.5 ml of 0.25% v/v trypsin/EDTA in PBS. Trypsin action was quenched with 3 ml of media (containing 10% v/v FBS), then cells were collected, washed with 3 ml of 5 mM sodium pyruvate in PBS once, then re-suspended in 1 ml 5 mM sodium pyruvate in PBS. Two μl each of antimycin A (5 mM in DMSO) and dihydroethidium (DHE, 5 mM in DMSO) were added to each well, then the plates were incubated for 40 min at 37° C. The cells were then resuspended and transferred to round-bottom tubes, and analyzed by flow cytometry (FACScan, Becton Dickinson, Franklin Lakes, NJ). Data analysis was carried out by FlowJo (Becton Dickinson).

Cellular Uptake in A549 Cells

A549 cells were seeded in 6-well plates at 150,000 cells/well. After 48 hours (37° C, 5 % CO₂) of incubation, the cells were treated with 100 μM of ubiquinol either as a complex or as free drug. After 1 or 3 hours of treatment, 1 ml of cell lysis solution (1:1 mixture of 2% w/v SDS and 1% w/v Triton X) was added to each well and the plates were incubated for 15 minutes at 37° C. Later, 0.5 ml of cell lysate was spiked with 10 μl of 1 mg/ml solution of coenzyme Q9 (IS) and mixed. Two ml of ethyl acetate were added to each sample, and then the sample tube was vortexed for 5 minutes to extract the drug and IS, then centrifuged (3000 xg, 5 min). After the extraction was repeated, 4 ml of ethyl acetate were then evaporated under nitrogen stream, then the residue was reconstituted in 87.5 μl of tetrahydrofuran (THF). It was centrifuged, then the supernatant was diluted with acetonitrile and water at a ratio of THF:acetonitrile:water of 35:60:5. The samples were injected into the HPLC instrument as described above.

ROS Assay in Human Corneal Endothelial Cells

Human corneal endothelial cells (HCEC-B4G12, DSMZ-German Collection of Microorganisms and Cell Culture, GmbH) were seeded in endothelial cell growth medium (Genlantis, San Diego, CA) supplemented with 10 μg/L of Recombinant Human Basic Fibroblast Growth Factor (bFGF, Fisher Scientific) in 6-well plates at 150,000 cells/well for 48 hours (37° C, 5 % CO₂). The medium was then removed, and the treatments were added in 4 ml/well each (n = 3). Treatments added were (1) ubiquinol dispersed in the same medium at either 1 or 100 μM, (2) ubiquinol/γ-CD complex (molar ratio of 1:10) dispersed in the same medium at either 1 or 100 γM of ubiquinol equivalent concentration, and (3) γ-CD dissolved in the same medium at concentrations equivalent to those added with 100 μM of ubiquinol in the previous groups. Six wells were also left untreated. After 24 h, the media were removed and wells were washed with 5 mM sodium pyruvate in PBS; then, cells were detached with 0.5 ml of a 1:1 mixture of trypsin/EDTA 0.25%:accutase solution (Sigma Aldrich, St. Louis, MO) per well. After, 3 ml of media were added to each well, and then cells were collected, washed with 3 ml of 5 mM sodium pyruvate in PBS once, and resuspended in 1 ml of the same solution. Two γl each of antimycin A (5 mM in DMSO) and DHE (5 mM in DMSO) were added to each well and plates were incubated at 40 min. The cells were resuspended and transferred to round-bottom tubes, then analyzed by flow cytometry (FACScan). Data analysis was carried out by FlowJo.

Cytotoxicity Assay in Human Corneal Endothelial Cells

HCEC-B4G12 cells were seeded in endothelial cell growth medium supplemented with 10 μg/L of bFGF in 96-well plates at 10,000 cells/well for 24 h (37° C, 5 % CO₂). Different concentrations (0, 1, 10, 50, and 100 μM) of ubiquinol, equivalent concentrations of the complex, or relevant concentrations of γ-CD that coexist with complex amounts that are equivalent to the previously mentioned concentrations of ubiquinol were added to the cells. After 24 and 72 h, the media were removed, and 20 μl of MTS reagent (CellTiter-Glo® 2.0 Cell Viability Assay, Promega, Madison, WI) and 80 μl of media were added to each well. The plates were incubated at 37° C for 1 hour, then the absorbance was read at 490 nm according to the manufacturer’s protocol using a Spectra Max plus 384 Microplate Spectrophotometer (Molecular Devices). The absorbance obtained from untreated cells was considered to represent 100% viability, and the relative cell viability of other treatments were expressed as a percentage based on untreated cell absorbance (n = 3).

Lipid Peroxidation Assay in Human Corneal Endothelial Cells

B4G12 cells were seeded at 100,000 cells/well in a 6-well plate in endothelial cell growth media supplemented with 10 μg/L of bFGF. After incubation at 37° C and 5% CO₂ for 18 h, medium was removed from wells and ubiquinol, ubiquinol/γ-CD complex, or γ-CD, all in culture media, were added at 1, 10, 50, and 100 μM concentrations (n = 3). After incubation for 24 h at 37° C and 5% CO₂, 10 μM of erastin in DMSO was added to all wells after washing cells twice with 1X DPBS, except untreated and untreated-unstained groups, and the plates were incubated at 37° C and 5% CO₂ for 24 h. After incubation, 2 μl of C11-Bodipy 581/591 stock (in DMSO) were added to each well (except the untreated-unstained group), mixed vertically, and incubated for 20 min at 37° C and 5% CO₂. Cells were then washed in 1X DPBS, detached by trypsin, transferred to 50 ml Falcon tubes, and washed again in 1X DPBS by centrifugation at 230 xg for 5 min. Finally, cells were resuspended in Live Cell Imaging Solution (Thermo Fisher), and the fluorescence was analyzed using flow cytometry (FACScan). Data analyzed was performed using FlowJo.

Ferroptosis Assays in Human Corneal Endothelial Cells

MTS assay.

B4G12 cells were seeded at 2,500 cells/well in 96-well plates and incubated at 37° C and 5% CO₂ in endothelial cell growth media supplemented with 10 μg/L of bFGF. After 18 hours, the medium was removed, and treatments were added as follows: (1) ubiquinol dispersed in cell culture medium at 1, 10, 50, and 100 μM, (2) ubiquinol/γ-CD complex (molar ratio of 1:10) dispersed in the same medium at the same equivalent concentrations, and (3) γ-CD dissolved in the same medium at concentrations equivalent to those added with the complex (n = 3). We also used 5 μM of ferrostatin-1 (Sigma-Aldrich) in DMSO as a positive control to confirm ferroptosis. Two groups were left untreated (n = 3) with only medium added instead of treatments. After 24 h of treatment, cells were washed twice with 1X DPBS. In all treatment groups, and one of the untreated groups (n = 3), 10 μM of erastin in DMSO was added to induce ferroptosis. In another set of 96-well plates treated similarly, 2 μM of RSL3 in DMSO was added instead of erastin. After 24 h, bright field microscopic images were captured (Evos Cell Imaging System, Thermo Fisher). Cells were washed once with 1X DPBS buffer and 100 μl of medium and 20 μl of MTS reagent (Cell Titer-96 Aqueous One Solution Reagent, Promega Corporation) and 80 μl of medium were added to each well, then the plates were incubated at 37° C and 5% CO₂. After 3 h, the absorbance of the wells was read at 490 nm according to the manufacturer’s protocol using the Spectra Max plus 384 Microplate Spectrophotometer (Molecular Devices, Sunnyvale, CA).

LDH assay.

Lactate dehydrogenase (LDH) assay was conducted. In brief, 2,500 B4G12 cells/well were seeded in 96-well plates and incubated at 37° C and 5% CO₂ for 18 h. Then free ubiquinol, γ-CD, and ubiquinol/γ-CD complex at 1, 10, 50 and 100 μM concentrations were added to each respective well. Ferrostatin-1 (1 μM) was used as a positive control. Two control groups were used: one for measuring spontaneous LDH activity, and the other for measuring maximum LDH activity. Both control groups were treated with DMSO at a similar volume that would be used later in the erastin-treated group. After 24 h, cells were washed twice with 1X DPBS buffer and treated with erastin (10 μM) in DMSO and incubated at 37° C and 5% CO₂ for 24 h. Then, 10 μl of 10X lysis buffer was added to the control group designated for maximum LDH activity measurement and incubated for 45 min at 37° C and 5% CO₂. Following incubation, 50 μl of each sample medium were transferred to a 96-well flat-bottom plate. To perform the positive control assay, 50 μl of 1X LDH positive control was transferred into triplicate wells. Reaction mixture of 50 μl was added to each well and incubated at room temperature for 30 min in dark conditions. After 30 min, the reaction was stopped by adding 50 μl of stop solution. Absorbance was measured at 490 nm as mentioned above according to the manufacturer’s protocol. Percent cell viability was calculated by the following formula:

$$\%\text{Cell viability =100-}(\frac{CompoundtreatedLDHactivity-SpontaneousLDHactivity}{MaximumLDHactivity-SpontaneousLDHactivity})\times 100$$

Transcorneal Penetrance Assays

Porcine corneas.

Coumarin-6 γ-CD complex was prepared using the kneading method described previously at a coumarin-6: γ-CD ratio of 1:10. An amount of coumarin-6 or an equivalent amount of coumarin-6/γ-CD complex was suspended in 1X DPBS to make a concentration of 300 μM. The suspensions were vortexed for 5 min, then sonicated for 10 minutes to break any aggregates. Porcine corneas (freshly excised porcine eyes were obtained from a local abattoir, and the corneas were collected in our lab) were attached to the receiving side of a NaviCyte vertical diffusion chamber system (Harvard Apparatus, Holliston, MA, USA) with the epithelial side facing the donor side of the diffusion cell. One ml of the complex suspension in DPBS or free coumarin-6 suspension in DPBS (300 μM each) was added to the donor side. One ml of DPBS was added to the receiver side. A mixture of oxygen and CO² at an O²:CO² ratio of 95:5 (carbogen) was bubbled through the cells. The cells were kept at 37° C. After 2 hours, the receiver solution was collected; then, the corneas were removed from the diffusion cells, rinsed thoroughly in PBS, attached on a glass coverslip with an anti-fade mounting medium (ProLong Gold Antifade reagent, Thermo Fisher), and imaged using a confocal microscope (Leica SP8 STED Super Resolution confocal microscope, Leica Microsystems, Buffalo Grove, IL).

Human corneas.

Corneoscleral tissues were obtained, inspected, and stored in Optisol-GS® (Bausch+Lomb) at 4° C according to the Eye Bank Association of America and Iowa Lions Eye Bank (ILEB) standard protocols. All human tissues used in this research were deemed suitable for corneal transplantation according to standard ILEB protocols, and all experimental tests were performed within 14 days of procurement. For each donor pair, one cornea was treated with complexed ubiquinol and the other was treated with free ubiquinol.

Eight human donor corneas were attached to the Navicyte vertical diffusion system as described above. Either free ubiquinol or ubiquinol/γ-CD complex (n = 3/group) was suspended in 1X DPBS at a concentration of 300 μM and sonicated for 5 min to break aggregates, then 1 ml of each was added to the donor side of the diffusion cell, and 1 ml of 1X DPBS was added to the receiving side. The diffusion system was kept at 37° C and carbogen was bubbled on both sides. At predetermined time intervals (0, 30, 60, and 120 min), the entire volume on the receiving side was collected from each cell, and replaced with fresh 1X DPBS.

Each sample of 1 ml was transferred to a 15 ml Falcon tube and the IS in methanol (100 ng) was added to each tube. Three ml of ethyl acetate were added to each tube and the tube was vortexed for 5 minutes and centrifuged at 3000 xg for 5 minutes, then the ethyl acetate portion was transferred to a glass tube. Extraction was repeated for another time and ethyl acetate part was transferred to the same glass tube, then evaporated under nitrogen flow. Following evaporation, each sample was reconstituted in 1 ml methanol and centrifuged at 16,000 xg for 10 min, and the supernatant was collected for liquid chromatography/tandem mass spectrometry (LC/MS/MS) analysis.

Corneas were collected immediately after the diffusion experiment, washed thoroughly in 1X DPBS, and the endothelial cell-Descemet membrane complex (EDM) was peeled (using the full corneal diameter). EDM tissues were collected in 2 ml screw cap tubes, and 20 zirconia beads (2.5 mm) along with 500 μl of methanol were added to each tube. Tubes were shaken horizontally for 2 h at room temperature (500 rpm) and were kept at 4° C. After 20 h at 4° C, IS in methanol (100 ng) was added to each tube, then 1 ml of ethyl acetate was added to each tube and the tissues were homogenized in a bead homogenizer (Fisher Brand Bead Mill 4 Homogenizer, Hampton, NH) for 2 min. Tubes were shaken for an additional hour (500 rpm) at room temperature. Then, the entire contents of these tubes were transferred to new 15 ml tubes; 3 ml of ethyl acetate were added to these tubes, and they were again shaken for 1 h and finally centrifuged (3000 xg for 5 minutes). The supernatants (4.5 ml) were transferred to glass tubes and evaporated completely under nitrogen flow. Residues were reconstituted in 0.5 ml methanol and centrifuged at 16,000 xg for 10 min, and ubiquinol and ubiquinone were analyzed in the supernatants by LC/MS/MS as described below.

Corneal Storage Uptake Assay

Human donor corneas were stored in Optisol-GS® supplemented with either 10 μM of ubiquinol or ubiquinol/γ-CD complex equivalent to 10 μM of ubiquinol (n = 4/group), under standard ILEB protocols for 7 days at 4° C. For each donor pair, the right cornea was treated with complexed ubiquinol and the left cornea was treated with free ubiquinol. After 7 days of incubation, corneas were washed with 1X DPBS. EDMs were peeled and after IS was added, ubiquinol and ubiquinone were extracted as described above, then analyzed by LC/MS/MS as described below.

Liquid Chromatography with tandem mass spectrometry (LC/MS/MS)

The LC/MS/MS system used in the penetrance and uptake assays consisted of a Waters Acquity TQD (Milliford, MA), which includes a triple quadruple mass spectrometer and Acquity H-Class Ultra Performance Liquid Chromatographer (UPLC). The mobile phase consisted of 100% methanol with 5 mM ammonium formate at 0.8 ml/min (isocratic) and the column used was a Phenomenex Gemini RP-C18 (50 x 4.6 mm, 5 μm). Quantitative analysis of ubiquinol, ubiquinone, and coenzyme Q9 (IS) was carried out using positive electrospray ionization in multiple reaction monitoring (MRM) mode. Ubiquinol, ubiquinone, and coenzyme Q9 were each detected at m/z 882.795→197.19, m/z 880.78→197.107, and m/z 812.696→197.124 transition channels, respectively (Fig. S1). The standard curves were linear over concentration ranges of 10-100 ng/ml and 10-90 ng/ml for ubiquinol and ubiquinone, respectively.

Statistical Analysis

All data values were reported as mean ± SD. All experiments depicted in figures 2, 3, 4, 5 (B–E), S4, and S6 are based on the measurements of at least 3 biological replicates, with each run in technical triplicate. Statistical significance was calculated first with 1-way ANOVA, followed by Tukey’s post-hoc test. Two tailed Student’s T-test was used when comparing two experimental groups. P values of less than or equal to 0.05 were considered significant.

Figure 2:

Stability of ubiquinol alone versus ubiquinol/γ-CD complex in (A) Optisol-GS® and (B) BSS. Stability is measured with regard to ubiquinol (the reduced form), ubiquinone (the oxidized form), and total coenzyme Q10. Values represent mean ± SD, * p < 0.05, ** p < 0.01, n = 3.

Figure 3:

(A) Flow cytometric histograms of A549 cells. (B) and (C), bar graph figures representing the values obtained from the statistical analysis (geometric means) of the DHE fluorescence signals from histograms (values are means ± SD, **** p < 0.0001, *** p < 0.001, ** p < 0.01, and * p < 0.05). All groups represented by B and C were processed simultaneously, and they were split into two Figures to emphasize the specific message that each set conveys, and to improve clarity. (D) Amount of total CoQ10, ubiquinol and ubiquinone taken up into A549 cells after incubation of either free ubiquinol or ubiquinol/γ-CD complex with the cells for 1 or 3 hours at 37° C. Values represent mean ± SD, **p < 0.01, n = 4-6).

Figure 4:

(A) Flow cytometric analysis following DHE-based ROS assay in HCEC-B4G12 cells. (B) DHE fluorescence signal values (means ± SD, **** p < 0.0001, *** p < 0.001, and ** p < 0.01) following the ROS assay in HCEC-B4G12 cells measured by flow cytometry (geometric means). Unless otherwise mentioned, all group were treated with antimycin A (AM; Fig. A and B). (C) MTS-based cytotoxicity assay of HCEC-B4G12 cells following 24 h and 72 h of incubation with different concentrations of ubiquinol, ubiquinol/γ-CD complex, and γ-CD (values are means ± SD). (D) Schematic showing the proposed role of ubiquinol to inhibit lipid peroxidation and the subsequent ferroptosis (Cys: cysteine, Glu: glutamate, GSH: reduced glutathione, GSSG: oxidized glutathione). (E) Flow cytometric analysis following C11-Bodipy 581/591-based lipid peroxidation assay in HCEC-B4G12 cells. (F) Bodipy 581/591 fluorescence signal values (means ± SD, **** p < 0.0001, *** p < 0.001, and ** p < 0.01) following lipid peroxidation assay in HCEC-B4G12 cells measured by flow cytometry (median). Unless otherwise mentioned, all group were treated with 10μM erastin (Fig. E and F). (G) MTS cytotoxicity assay of HCEC-B4G12 cells following erastin treatment (values are means ± SD, **** p < 0.0001). (H) Light microscopic images of cells treated with different treatment groups. (I) LDH-based cytotoxicity assay of HCEC-B4G12 cells (values are means ± SD, **** p < 0.0001).

Figure 5:

(A) Coumarin-6/γ-CD complex (molar ratio of 1:10) shows much higher corneal penetrance compared to free coumarin-6 when applied to the corneal epithelium. Complexed coumarin-6 was able to diffuse uniformly across and through the anterior cornea from the epithelial side (top panel), and it was able to penetrate the entire cornea and reach the endothelial side (bottom panel) while free coumarin-6 could not. (B) Cumulative amount (ng) versus time of total coenzyme Q10 (ubiquinol and ubiquinone) detected in fluid on the endothelial side after permeation across human donor corneas fixed in a Navicyte vertical diffusion system after either ubiquinol/γ-CD complex or free ubiquinol (300 μM, n = 3/group) was added to the epithelial side of the diffusion apparatus. (C) Amount of total coenzyme Q10 (ng) retained in the endothelial cell Descemet membrane complex (EDM) of donor corneas isolated after permeation of either complexed or free ubiquinol (n = 4/group). (D) Amount of total coenzyme Q10 (ng) retained in the EDM of human donor corneas stored for 1 week (4° C) in Optisol-GS® supplemented with either complexed or free ubiquinol (10 μM, n = 4/group) in standard corneal storage containers (shown in the inset). Data represent average ± SD, p < 0.05 following two tailed Student’s T-test). (E) Pairwise comparison of complexed versus free ubiquinol uptake in the EDM of human donor corneas stored in Optisol-GS® supplemented storage media depicted in (D). For each donor pair, the right cornea was treated with complexed ubiquinol and the left cornea was treated with free ubiquinol.

RESULTS

Molecular Docking

Molecular docking studies were performed in order to model the most suitable excipient host molecule for ubiquinol dispersion in the aqueous phase. Ubiquinol showed stronger binding affinity to γ-CD in comparison to α-CD and β-CD. This is due to the relatively larger cavity size of γ-CD, which can accommodate a large portion of the ubiquinol hydrophobic tail and head group (Fig. 1A, 1B, and 1C). The head group of ubiquinol displayed multiple hydrogen bonding (H-bonding) formations with γ-CD (Fig. 1A), which is important for the stability of the complex as well as maintenance of reduced hydroxyl groups in the benzoquinol unit. The hydrophobic interactions between the lipophilic tail and the hydrophobic cavity of the host also provide strong non-covalent bonding that increase the stability of the complex (Fig. 1A). The intermolecular interaction energy (kcal/mol) had the highest value using γ-CD as the host molecule compared to other cyclodextrins (Fig. 1C), which indicates stronger binding affinity and likely results from the coexistence of both H-bonding and hydrophobic interactions. Docking studies also showed that γ-CD possibly forms inclusion complexes with other locations in the trans-isoprene side chain of ubiquinol, with both H-bonding and hydrophobic interactions involved (Fig. S2), indicating that one molecule of ubiquinol may interact with more than one molecule of γ-CD. Therefore, we selected γ-CD as our host for the supramolecular inclusion complex formation, with ubiquinol as the guest molecule.

Figure 1:

Molecular interactions between ubiquinol and α-, β- and γ-cyclodextrins. (A) 3D structures of α-, β-, and γ-cyclodextrins as soft surfaces, docked pose of ubiquinol in the cavity of α-, β-, and γ-cyclodextrins, interactions between ubiquinol and α-, β-, and γ-cyclodextrins with the H-bond distances, and representative hydrophobic interactions between ubiquinol and α-, β-, and γ-cyclodextrins with distances. Dashed green lines represent conventional H-bonds, dashed blue lines represent carbon-hydrogen bonds, and dashed magenta lines represent hydrophobic interactions. BIOVIA Discovery Studio Visualizer was used to generate H-bond distances, and PyMOL was used to generate the distances of hydrophobic interactions. (B) 3D structure of ubiquinol. (C) Predicted binding energy of ubiquinol to each of the three types of cyclodextrins used. (D) Phase solubility diagram of ubiquinol in different concentrations of γ-CD. (E) Photo: (on left): dispersion of 50 mg of ubiquinol/γ-CD complex (equivalent to 3.125 mg ubiquinol) added to 10 ml of water and shaken for 2 hours; (on right): dispersion of 5 mg ubiquinol alone added to 10 ml of water and shaken for 2 hours. (F) DSC thermograms and (G) XRD patterns of ubiquinol, γ-CD, ubiquinol/γ-CD physical mixture, and ubiquinol/γ-CD complex. (H) SEM images of ubiquinol, γ-CD, and ubiquinol/γ-CD complex. Top panel: X150, middle panel: X600 and bottom panel: X1200.

Phase Solubility

The solubility of ubiquinol in different molar concentrations of γ-CD increased slowly with increasing γ-CD concentrations. While only less than 15 μM of ubiquinol dissolved in 10 mM of γ-CD, the concentration of ubiquinol increased to 150-200 μM in 20 mM of γ-CD. After ubiquinol concentration showed a brief plateau at 30 and 40 mM γ-CD, it increased sharply in 50 mM to 1500 μM (Fig. 1D).

Preparation and Characterization of Complexed Ubiquinol

The behavior of the prepared ubiquinol/γ-CD complex after dispersion in water confirms the formation of a new composition (complex) that possesses selective characteristics of both ingredients. While the free γ-CD is completely soluble in water at concentrations higher than 200 mg/ml, and ubiquinol is completely insoluble in water, the formed complex appeared highly dispersed in water at a concentration of about 3.125 mg/ml (Fig. 1E and Fig. S3). DSC thermograms showed a sharp ubiquinol endothermic peak at approximately 48° C (Fig. 1F). This endothermic peak was markedly decreased in value and slightly shifted to approximately 46° C in the complex sample only, and not the physical mixture. An incomplete interaction between ubiquinol and γ-CD may be the reason why, unexpectedly, the melting peak did not disappear completely (Fig. 1F).

Similarly, XRD patterns showed that the major crystallinity peak of ubiquinol (2 theta value of 19) is retained slightly in the physical mixture only (Fig. 1G). SEM images show signs of interaction, as characteristics of both ubiquinol and γ-CD crystals can be seen in the same particles of the complex. The complex (Fig. 1H) shows that the smooth crystals of γ-CD appear to be coated with a rough surface characteristic of ubiquinol particles (X1200).

Complexed Ubiquinol is Stable in Aqueous Media

The stability of complexed ubiquinol in various aqueous phase products used in donor cornea tissue preservation, cornea transplant surgery, and clinical ophthalmic care was characterized. Because ubiquinol is prone to oxidize to ubiquinone in aqueous solutions due to dissolved oxygen (manifest as a yellow discoloration of the powder or solution/suspension), the analytical method employed quantified the remaining ubiquinol, ubiquinone, and total coenzyme Q10. Ubiquinol/γ-CD complex showed significantly higher stability in suspension at 4° C when added to Optisol-GS® (the most common corneal storage medium) and BSS (the most common intraocular irrigating solution) when compared to free ubiquinol (Fig. 2A and 2B). When free ubiquinol was added to these media, Optisol-GS® ubiquinone levels increased over time due to oxidation and then declined. Ubiquinol, ubiquinone, and total coenzyme Q10 all declined markedly after 4 and 7 days in both media in the free ubiquinone assays. After complexed ubiquinol was added to these media, the stability of all three entities was significantly higher after 4 and 7 days in both media, compared to free ubiquinol (with the exception of ubiquinone after 4 days in BSS). Additionally, the ubiquinol/ γ-CD complex showed almost 100% stability after one week in Refresh® eye drops at 4° C (Fig. S4) with regard to levels of ubiquinol, ubiquinone, and total coenzyme Q10.

Complexed Ubiquinol Lowers ROS in High-ROS Adenocarinomic Epithelial Cells (A549)

The ability of complexed ubiquinol to scavenge intracellular ROS was tested, starting with a carcinomatous cell culture model with abnormally high metabolism expected to generate high levels of intracellular ROS [48]. Complexed ubiquinol was able to completely mitigate ROS generation induced by antimycin-A (AM) in human adenocarinomic alveolar epithelial cells (A549), while free ubiquinol was unable to inhibit ROS generated by the same concentration of AM (Fig. 3A and 3B). Free ubiquinol was able to lower cellular ROS levels in these cells below basal levels, as did the complex, without ROS induction by AM. Even though the free drug was able to decrease ROS levels after AM treatment (compared to the untreated group), it was still significantly higher than the ROS levels in the ubiquinol-treated cells without induction by AM. ROS levels in complex-treated cells (with AM) were significantly less than those in untreated cells even before AM induction. In addition, an increase in the concentration of ubiquinol in complex resulted in a significant increase of ROS inhibition (Fig 3C), while a similar increase in the free ubiquinol concentration did not result in any significant change in ROS levels. We also found that the complex, but not free ubiquinol, at a concentration of 50 μM significantly lowered ROS levels compared to the untreated group.

In addition, we determined that ubiquinol cellular uptake was significantly higher after 3 hours of incubation with cells when the complex was used compared to free ubiquinol (Fig.3D).

The chromatogram of coenzyme Q9, ubiquinone, and ubiquinol (at 290 nm) in the cell lysate, and the chromatographic conditions are displayed in Fig. S5.

Complexed Ubiquinol Lowers ROS in Human Corneal Endothelial Cells (HCEC-B4G12)

The ability of complexed ubiquinol to scavenge intracellular ROS was also tested using a high-fidelity immortalized human endothelial cell culture model [49, 50]. ROS levels were significantly lower in B4G12 cells when complexed ubiquinol at low and high concentrations (1 μM and 100 μM) were used, compared to free ubiquinol in suspension (Fig. 4A, and 4B, p < 0.0001). γ-CD alone was not active. The complex at low concentration (1 μM) was more efficient than ubiquinol at high concentration (100 μM) in ROS mitigation in B4G12 cells (p < 0.01). The complex at an equivalent ubiquinol concentration of 100 μM showed complete peak shift by an order of magnitude, and the fluorescence (geometric mean) of cells in this group was about one tenth of that of cells treated with 100 μM of free ubiquinol (p < 0.0001), indicating total abolition of ROS levels in corneal endothelial cells at this concentration.

Complexed Ubiquinol Is Not Toxic To Human Corneal Endothelial Cells

Cytotoxicity evaluations were performed to assess for toxicity of complexed ubiquinol in corneal endothelial cells. MTS assay results using HCEC-B4G12 cells tested after 24 and 72 hours of exposure showed no inhibition of proliferation or signs of cell death for complexed ubiquinol, free ubiquinol, or γ-CD at all concentrations tested (Fig. 4C). No microscopic signs indicating signs of cell stress or death after 72 h of treatment groups were detected for all groups.

Complexed Ubiquinol Inhibits Lipid Peroxidation and Ferroptosis

C11-Bodipy 581/591, a fluorescent lipid analog used as a reporter for lipid peroxidation that exhibits fluorescence shift from red to green upon exposure to ROS [51, 52], was used to investigate the potential of γ-CD complexation to improve ubiquinol scavenging of lipid peroxides (Fig. 4). Erastin – a lipid peroxidation inducer that, along with RSL3 (known collectively as RAS-selective lethal, or RSL, compounds) leads to ferroptosis [11] – was used to induce lipid peroxidation. Erastin inhibits cystine cellular uptake in cystine-glutamate antiporter (also known as system X_c⁻ or xCT), which depletes reduced glutathione (GSH) and subsequently inhibits GPX4, while RSL3 directly inhibits GPX4 by attacking its nucleophilic moiety [11, 53, 54]. GPX4 inhibition results in accumulation of lipid peroxides and cell death by ferroptosis [11, 19, 53, 54]. In addition to GPX4, another recently found anti-ferroptotic cellular defense mechanism involves the ferroptosis suppressor protein 1 (FSP1) driven reduction of endogenous coenzyme Q10 to produce ubiquinol. The highly dispersible ubiquinol/γ-CD complex provides sufficient ubiquinol to the cell to inhibit lipid peroxidation completely and protect cells against ferroptosis. Fig. 4D summarizes this pathway.

Erastin treatment significantly increased the levels of lipid peroxides as shown in Fig. 4E and 4F. Free ubiquinol was ineffective at inhibiting lipid peroxidation at low concentrations (1 μM, Fig. 4E and 4F), and its efficacy started to become manifest at 100 μM. In contrast, complexed ubiquinol significantly inhibited lipid peroxidation at 1 μM. The complex at an equivalent ubiquinol concentration of 10 μM significantly inhibited lipid peroxidation compared to cells treated with erastin (p < 0.0001), γ-CD (p < 0.0001), and 100 μM ubiquinol (p < 0.001).

We then assessed cell survival following treatment with erastin using MTS and LDH assays, with both assays giving similar results. Ferrostatin-1 was used as a positive control to confirm that cell death was caused by ferroptosis. Only complexed ubiquinol and ferrostatin-1 were able to prevent erastin-induced ferroptosis, while free ubiquinol was ineffective, even at high concentrations (Fig. 4G, 4H, and 4I). Similar results were obtained when RSL3 was used to induce ferroptosis (Fig. S6).

Complexation Increases Transcorneal Ubiquinol Delivery to Donor Corneal Endothelial Cells

Coumarin-6, a lipophilic dye with readily detectable fluorescent signal detection, was used to assay transcorneal penetrance of the γ-CD host using porcine corneas. Coumarin-6 formed a stable complex with γ-CD (Fig S7A–C), with a predicted binding energy of −6.6 kcal/mol (Fig. S7D), and it was chosen as a model fluorescent probe to enable monitoring of γ-CD-based complexes owing to its lipophilicity. This complex was highly dispersed in DPBS compared to coumarin-6 (Fig. S7E). Confocal microscopy images show that the complex was able to cover a larger area of the corneal epithelial side than free coumarin-6 (Fig. 5A), mainly due to the increased dispersibility that γ-CD imparted on the highly lipophilic dye. Using a Navicyte vertical diffusion system over 2 hours, the coumarin-6/ γ-CD complex was able to penetrate deeper into the posterior cornea and it reached the endothelial side of the cornea in large amounts. In contrast, no coumarin-6 signal could be detected on the endothelial side when the free dye was used (Fig. 5A). This finding highlights the ability of γ-CD complexation to penetrate the corneal epithelium and stroma, and reach the corneal endothelium to deliver the drug load.

Next, the transcorneal penetrance of ubiquinol/γ-CD complex compared to free ubiquinol was assayed in human donor corneas using a Navicyte vertical diffusion system over 2 hours. We found that the average cumulative amount of ubiquinol that crossed the entire corneal thickness (after application on the epithelial side, in fluid sampled from the endothelial side) was higher with complexed ubiquinol treatment compared to free ubiquinol (Fig. 5B). The amount of ubiquinol retained in the EDM was measured by LC/MS/MS. Higher amounts of ubiquinol were detected in the corneal endothelium of corneas treated with complexed ubiquinol compared to free ubiquinol (Fig. 5C). This finding confirms the ubiquinol/γ-CD complex’s ability to penetrate the corneal epithelium and stroma, and deliver ubiquinol to corneal endothelial cells.

Complexation Increases Ubiquinol Delivery to Donor Corneal Endothelial Cells in Storage Media

Lastly, the penetrance of ubiquinol/γ-CD complex compared to free ubiquinol was assayed using human donor corneas preserved in storage media per standard industry practices. When corneas stored in Optisol-GS® were supplemented with complexed ubiquinol (10 μM) for 1 week, the amount of ubiquinol taken up into corneal endothelial cells was significantly higher (p < 0.05) compared to corneas stored in Optisol-GS® supplemented with the same concentration of free ubiquinol (Fig. 5D). In pairwise comparison, in all cases complexed ubiquinol delivered the higher amount of ubiquinol to donor corneal endothelial cells (Fig. 5E).

DISCUSSION

This research reports the development of a highly dispersible and stable formulation of ubiquinol in a single-drug formulation for the supplementation of aqueous-phase ophthalmic products including media used for the preservation of human donor corneal tissue, intraocular irrigation solutions, and topical eye drops. To our knowledge, no such formulation has been developed to incorporate the reduced and active form of coenzyme Q10, ubiquinol, as a single agent in an aqueous, non-lipid based formulation for therapeutic biomedical purposes.

Ubiquinol’s classical mechanism of action as a ROS scavenger has been well studied; however, we employed our formulation in light of newly described functions of ubiquinol as a ferroptosis suppressor to improve cell health. Doll et al. and Bersuker et al. recently reported that FSP1 counters this form of non-apoptotic cell death by catalyzing the recycling of coenzyme Q10 by NADPH to generate ubiquinol [13, 21]. Ubiquinol then goes on to protect cells by mitigating peroxyl radicals, inhibiting the lipid peroxidation that results in ferroptosis [13, 21]. Previously, we demonstrated native ubiquinol to be an effective inhibitor of corneal endothelial cell damage and proposed its use, after optimization, as a supplement added to cold storage prior to transplant surgery. In this study, we showed that while free ubiquinol had limited activity against erastin-induced lipid peroxidation and ferroptosis in human corneal endothelial cells, complexed ubiquinol abolished lipid peroxidation and provided complete protection against ferroptosis, with no cell death reported, even at the lowest concentration tested. We also demonstrated that complexed ubiquinol is more efficient at mitigating ROS levels compared to free ubiquinol using the same cell model. No evidence of toxicity was noted (Fig. 4), thereby paving the way for future studies using this formulation. Inhibition of ferroptosis in the corneal endothelium has been suggested as a proposed treatment in patients with corneal dysfunction secondary to elevated oxidative stress including Fuchs endothelial corneal dystrophy (FECD), the most common indication for cornea transplant surgery [15, 17, 22, 23, 26]. Ubiquinol complexation now makes it possible to tune drug concentration to the cell of interest (e.g. human donor corneal endothelial cells), perform mechanistic studies of ferroptosis mediated cell death across this and other hypothermic organ and tissue presurgical preservation platforms, and translate the use of this macromolecular antioxidant formulation to other ophthalmic applications.

We have designed and tested this highly water dispersible ubiquinol complex for compatibility with other clinical applications, particularly as a topically applied therapy for ophthalmic conditions characterized by oxidative damage. Regarding stability in aqueous media, after complexation with γ-CD, we found little to no decline in ubiquinol levels in all three aqueous applications tested (Optisol-GS® corneal storage media, BSS intraocular irrigating solution, and Refresh® artificial tear lubricating eye drops) after 1 week, while native free crystalline ubiquinol levels declined significantly in Optisol-GS® and BSS after 4 days. Cyclodextrins are well known to improve stability of labile molecules [55, 56], and hydrogen bonding between the benzoquinol group of ubiquinol and the cavity of γ-CD may be the reason behind the improved stability. Regarding corneal penetrance, lipophilic molecules like ubiquinol encounter difficulties passing through cell membranes, due to the permeability barrier of the cellular membrane itself as well as the barrier exerted by the unstirred water layer (UWL) formed as a result of the concentration gradient across the membrane [57, 58]. Inclusion complexes comprising lipophilic drug molecules and cyclodextrins render such drug molecules capable of penetrating the UWL to increase the number of drug molecules adjacent to the cell membrane, which consequently improves their overall permeation through membranes [59]. In addition to good cell uptake when applied directly to donor corneal endothelial cells, our data show significant corneal penetration through the corneal epithelium and stroma to the corneal endothelium. We used a topically applied inclusion complex of γ-CD with a lipophilic dye (coumarin-6) to demonstrate the capacity for the γ-CD vehicle to deliver drug load. Then, using the ubiquinol/γ-CD complex in the same topical application model, we demonstrated transcorneal penetrance across the entire thickness of human donor corneas and markedly higher endothelial retention of packaged ubiquinol compared to free drug, indicating the potential for clinically significant bioavailability using this strategy. Rapid clearance of eye drops from the surface of the eye through tearing is another barrier in ophthalmic drug delivery. However, formulations comprised using cyclodextrin can also overcome this problem [60, 61]. More importantly, we have shown that CD complexation not only maximizes the ROS mitigation/anti-ferroptotic capacities of ubiquinol, but also it enables significantly higher corneal uptake of the powerful antioxidant during storage in hypothermic storage media (e.g. Optisol-GS®) as shown in Fig. 5D and 5E. With solubility, stability, safety, function, and penetrance now demonstrated in topically applied aqueous vehicles, the spectrum of applicability of this soluble ubiquinol composition easily may be widened to include eye drops for the treatment of endothelial cell diseases characterized by oxidative damage such as FECD [15, 17] and can be investigated as a nonsteroidal adjuvant in the maintenance of corneal endothelial cell health towards minimizing oxidative stress following keratoplasty [18].

Complexing with γ-CD also achieved a more stable form of ubiquinol with respect to storage and handling. As a vacuum-dried powder, ubiquinol/γ-CD did not undergo observable yellowing – an indication of oxidative degradation to ubiquinol – unlike native ubiquinol, which can degrade quickly. The final powdered product, stored at −20° C, was not observed to undergo functional degeneration across experiments whereas native ubiquinol has been noted to degrade at −20° C [62]. The improved handleability of complexed ubiquinol makes this formulation immediately translatable into clinical usage, for example in eye banking where it can be packaged for addition to corneal storage media at the time of tissue procurement in the field. We have also considered this formulation in the context of future developments in eye banking and presurgical tissue handling. Fungal infections transmitted from donor corneal tissue to transplant recipient – primarily caused by Candida species – can be devastating and are becoming more common after cornea transplantation [63–65]. Our group and others have investigated the supplementation of Optisol-GS® with antifungal medication such as amphotericin B to reduce fungal transmission; this approach has been tested in preclinical studies [66] but not yet in clinical trials. We found that supplementation of Optisol-GS® with both amphotericin B and ubiquinol/γ-CD did not have adverse effects on fungicidal activity of amphotericin B against Candida species (Table S2; experimental details can be found in the supplementary section). Such experimentation provides early evidence that complexed ubiquinol does not interfere with the efficacy of other drug additives to media, though additional studies are needed to ensure optimum drug efficacy in primary corneal endothelial cells.

CONCLUSION

In summary, the supramolecular inclusion complex formed between ubiquinol and γ-CD is a highly stable composition that can be incorporated into a variety of aqueous phase products for ophthalmic use, including corneal storage media and topical eye drops, to protect corneal endothelial cells against oxidative stress. This formulation achieved high levels of ROS scavenging activity in cell culture models, complete inhibition of lipid peroxidation, higher corneal endothelium uptake during hypothermic storage, and full protection against ferroptosis in human endothelial corneal cells, and remarkable penetrance through the entire cornea and into the corneal endothelium when applied to the front of the eye. After establishing dose and efficacy of ubiquinol/γ-CD for stored donor corneal endothelial cells, prospective randomized controlled trials of keratoplasty using tissue with or without ubiquinol/γ-CD supplementation can occur. Extending the preservation time of donor corneas will directly increase tissue availability domestically and internationally as only 1 cornea is available for every 70 cases in need [67], and improving the health of corneal endothelial cells by reducing oxidative damage during preservation has a high probability of increasing transplant graft survival. Additionally, preclinical and clinical studies can be conducted to investigate complexed ubiquinol as a topical medical therapy to prevent vision loss and reduce the need for surgery to treat FECD – a condition that affects 4% of the population [68] – and other conditions in which ROS accumulation damages the corneal endothelium.

Data availability

The raw/processed data required to reproduce these findings cannot be shared at this time due to technical or time limitations. Most of the processed data has been shown in figures and tables with detail experimental parameters which are sufficient to reproduce data. However, more details raw data can be provided later as needed.