ROCK Inhibitor Enhances Resilience Against Metabolic Stress Through Increasing Bioenergetic Capacity in Corneal Endothelial Cells

Abstract

Purpose: To investigate the effect of Rho-associated protein kinase (ROCK) inhibitor Y27632 on bioenergetic capacity and resilience of corneal endothelial cells (CECs) under metabolic stress.

Methods: Bovine CECs (BCECs) were treated with Y27632 and subjected to bioenergetic profiling using the Seahorse XFp Analyzer. The effects on adenosine triphosphate (ATP) production through oxidative phosphorylation and glycolysis were measured. BCECs were also challenged with monensin to induce metabolic stress. Cell viability, apoptosis, intracellular sodium levels, and hexokinase localization were assessed using calcein AM assay, flow cytometry, fluorescence imaging, and immunostaining, respectively.

Results: Y27632 increased maximal ATP production rates via both oxidative phosphorylation and glycolysis, thereby expanding the overall bioenergetic capacity in BCECs. Under monensin-induced metabolic stress, ROCK inhibitor pretreatment significantly enhanced glycolytic ATP production and reduced apoptosis compared with untreated cells. Y27632 also facilitated sodium export by increasing Na/K-ATPase activity, as evidenced by lower intracellular sodium levels. Additionally, Y27632 promoted the translocation of hexokinase 2 to mitochondria under stress conditions, thereby enhancing glycolytic capacity. The effect of Y27632 on cell viability and sodium export was abrogated when cells were forced to rely on oxidative phosphorylation in galactose media, indicating that the protective effects of Y27632 are dependent on glycolytic ATP production under monensin stress.

Conclusions: ROCK inhibitor Y27632 enhances the bioenergetic capacity of BCECs, allowing the cells to better withstand metabolic stress by rapidly generating ATP to meet increased energy demands, maintaining ion homeostasis and reducing apoptosis.

The corneal endothelium is essential for maintaining corneal clarity through active ion pumping.¹ Owing to the limited mitotic ability of human corneal endothelial cells (CECs) in vivo, a decrease in cell density owing to factors such as endothelial dystrophies or surgical trauma can lead to corneal endothelial decompensation.² This process results in impaired fluid regulation, stromal edema, and subsequent visual impairment. 

To counteract the passive fluid influx in the cornea and to maintain corneal deturgescence, one of the mechanisms used by CECs is through active ion transport, especially Na/K-ATPase pump.³ During the process, substantial adenosine triphosphate (ATP) expenditure is required to maintain ionic gradients.¹ When the corneal endothelium is diseased or damaged by surgical trauma, there is contradictory increase of corneal endothelial pump site densities, suggesting that the remaining cells manage to compensate for the increased permeability.⁴ Correspondingly, CECs from patients with Fuchs endothelial corneal dystrophy (FECD) undergo initial compensation and resistance phases to generate more ATP by increasing mitochondrial mass and calcium, as well as membrane potential before burnout and apoptosis.⁵ Reduced CEC density owing to cell apoptosis further increase the energy burden of the remaining cells.⁵ These observations indicate that CECs compensate for the cell loss with increased pump densities, as well as the metabolic demand per cell to maintain corneal deturgescence. However, the metabolic stress may overwhelm the bioenergetic capacity, inflicting further stresses on the remaining cells and leading to a vicious cycle of accelerated cell death. 

Since its first application in the field of corneal endothelial regeneration, inhibitors of Rho-associated protein kinase (ROCK) have been shown to have multifunctional roles, including promoting CEC survival and proliferation, enhancing corneal endothelial wound healing, and facilitating CEC attachment onto the posterior cornea after cell injection.^6–8 Some of the benefits contribute to the development of CEC injection therapy to treat bullous keratopathy.^9,10 Furthermore, our previous study showed that part of the effect of ROCK inhibition in supporting CEC migration and wound healing is through the enhancement of oxidative phosphorylation (OXPHOS), indicating that ROCK inhibition is associated with changes in mitochondrial function.¹¹ However, the overall change in bioenergetic scope and the survival benefit for CECs under stress remain to be determined. 

In this study, we demonstrated that, in addition to OXPHOS, Y27632, a ROCK inhibitor, also increased ATP production through glycolysis and, thus, the overall bioenergetic capacity in CECs. We further used monensin, an ionophore that facilitates the transport of sodium ions into the cells, to increase the activity of the Na/K-ATPase and thus incur the metabolic stress in CECs.^12,13 Our results showed that, under monensin stress, Y27632 helped CECs to generate more ATP through glycolysis to facilitate sodium output and withstand apoptosis. This evidence suggests that ROCK inhibition enhances the adaptive response of CECs to increased energy demand, which may be used further to maintain the function of CECs under metabolic stress. 

Materials and Methods

Reagents

Dulbecco's modified Eagle's medium: Nutrient Mixture F-12, HEPES buffer, phosphate-buffered saline, gentamicin, amphotericin B, fetal bovine serum, trypsin, and selenium were purchased from Invitrogen Corp. (Carlsbad, CA, USA). Dimethyl sulfoxide, human epidermal growth factor, insulin, transferrin, cholera toxin, bovine serum albumin, 4′,6-diamidino-2-phenylindole, Triton X-100 and Pluronic F127 were purchased from Sigma-Aldrich (St. Louis, MO, USA). Antibodies against caspase 3 (#9662), cleaved caspase 3 (#9664), PARP (#9542), AMP-activated protein kinase (AMPK) (#2532) and phospho-AMPK (#2535) were purchased from Cell Signaling Technology (Danvers, MA, USA). Antibodies against hexokinase 1 (HK1) (#GTX105248), HK2 (#GTX111525), and phospho-acetyl-CoA carboxylase (#GTX133974) were purchased from GeneTex Inc. (Irvine, CA, USA). The antibodies against cleaved PARP (#ab32064) and acetyl-CoA carboxylase (#ab72046) and ION NaTRIUM Green-2 AM were purchased from Abcam (Waltham, MA, USA). The antibodies against TOM20 (#sc-17764) and GAPDH (#sc-25778) were purchased from Santa Cruz Biotechnology (Dallas, TX, USA). Y27632, ripasudil and Y39983 were purchased from Tocris Bioscience (Minneapolis, MN, USA). 

Primary Culture of Bovine CECs

Fresh bovine eyes were acquired from the local abattoir, disinfected by iodine solution for 3 minutes, and then washed with phosphate-buffered saline. The corneal buttons were harvested, and the Descemet's membranes were peeled under the dissecting microscope. After digestion with trypsin at 37°C for 30 minutes, the bovine CECs (BCECs) were collected by centrifugation, seeded into a dish, cultured in supplemented hormonal epithelial medium (SHEM) composed of equal volumes of HEPES-buffered Dulbecco's modified Eagle's medium: Nutrient Mixture F-12, supplemented with 5% fetal bovine serum, 0.5% dimethyl sulfoxide, 2 ng/mL human epidermal growth factor, 5 mg/mL insulin, 5 mg/mL transferrin, 5 ng/mL selenium, 1 nmol/L cholera toxin, 50 mg/mL gentamicin, and 1.25 mg/mL amphotericin B. The dish was incubated at 37°C in an atmosphere of 95% air/5% CO₂, and the culture medium was changed every 3 days. When the cells reached confluence, they were trypsinized and passaged at a ratio of 1:3. Cultivated BCECs at passage 1 were used for all experiments. 

Measurement of Bioenergetic Capacity and ATP Production

The bioenergetic capacity and ATP production in BCECs were measured by the Seahorse XFp Analyzer (Agilent Technologies, Santa Clara, CA, USA) according to the manufacturer's instructions. BCECs were seeded onto Seahorse XFp cell culture miniplates and cultured in SHEM with or without 1 µM Y27632 until reaching 90% confluence. The maximal ATP production from OXPHOS was elicited after serial injections of oligomycin (1 µM), carbonyl cyanide 4-(trifluoromethoxy) phenylhydrazone (2 µM), and a rotenone/antimycin A mixture (1 µM). In contrast, maximal ATP production from glycolysis was induced after serial injections of a rotenone/antimycin A mixture (1 µM) and monensin (20 µM). The oxygen consumption rate and the extracellular acidification rate were measured and were used to derive ATP production rate from OXPHOS and glycolysis.¹⁴ To measure ATP production under monensin stress in BCECs, cells were plated as described elsewhere in this article. Until reaching 90% confluence, cells were further incubated with 10 µM monensin for 24 hours. The oxygen consumption rate and extracellular acidification rate were measured after serial injections of oligomycin (1 µM) and a rotenone/antimycin A mixture (1 µM), and the value were used to derive ATP produced. After completing the measurement, the cells on the miniplates were fixed and stained with 4′,6-diamidino-2-phenylindole. The cell numbers were quantified with ImageXpress Nano automated imaging system (San Jose, CA, USA) for normalization of the measured parameters. 

Quantification of Viable Cells

For assessing the effect of ROCK inhibitor to rescue CECs from monensin-induced cell death, BCECs preincubated with or without 1 µM Y27632 until confluence were treated with solvent control or 10 µM monensin for 24 h. The relative levels of viable cells were determined using Calcein AM assay (Invitrogen Corp.) according to the manufacturer's instructions. 

Flow Cytometry Analysis of Cell Apoptosis

Apoptosis was assayed using the Annexin V-propidium iodide (PI) with a commercially available kit (BioLegend, San Diego, CA, USA) according to the manufacturer's instructions. BCEC cells were seeded in 12-well culture plates with or without 1 µM Y27632 in SHEM. Upon reaching confluence, cells were further treated with monensin 1 to 10 µM in SHEM for 24 h. Following treatment, cells were trypsinized, washed thoroughly, and resuspended in Annexin binding buffer according to the manufacturer's protocol. Cells were labeled by adding Annexin V-FITC and PI in each sample. After incubation for 15 minutes at room temperature in the dark, samples were analyzed on a flow cytometer (FACSLyric, BD Biosciences, Franklin Lakes, NJ, USA) for the detection of Annexin V- and PI-positive subpopulations. Nontreated cells were used as controls. Further analysis was performed with FlowJo FACSuite software. 

Western Blotting Analysis

BCECs incubated with SHEM with or without 1 µM Y27632 until confluence were further challenged with 10 µM monensin for 24 hours. Cells were lysed with RIPA lysis buffer (Pierce Biotechnology, Rockford, IL, USA) containing protease inhibitor cocktail (Roche, Mannheim, Germany) and 0.1% SDS. After SDS-PAGE, proteins were transferred to a polyvinylidene difluoride membrane (Bio-Rad, Hercules, CA, USA), blocked with a solution containing 5% bovine serum albumin in Tris-buffered saline with 0.1% TweenR 20 detergent (Sigma-Aldrich) overnight, and detected by the indicated primary antibodies. After washing and incubation with secondary antibodies conjugated with horseradish peroxidase, immunoreactive bands were observed by chemiluminescence and quantified by ImageJ software (Image Processing and Analysis in Java; National Institutes of Health, Bethesda, MD, USA; http://imagej.nih.gov/). 

Measurement of Intracellular Sodium

To measure the level of intracellular sodium, BCECs were plated into a u-slide (ibidi, Gräfelfing, Germany) and were cultured in the medium containing solvent control or 1 µM Y27632 until reaching confluence. The cells were incubated with ION NaTRIUM Green dye diluted in SHEM containing 0.02% Pluronic F127 (final concentration = 5 µM) for 1 hour. After washing, cells were further incubated with SHEM with or without 10 µM monensin, and were subjected to fluorescence time-lapse imaging every 30 minutes for 16 hours by Zeiss Axiovert 200 M system (Carl Zeiss, Jena, Germany). The fluorescence level indicating the concentration of intracellular sodium was determined by ImageJ software. 

Immunofluorescence Staining

To observe the translocation of hexokinases to the mitochondria under monensin stress, BCECs incubated with SHEM with or without 1 µM Y27632 until confluence were further challenged with 10 µM monensin for 24 hours. The cells were fixed in 4% paraformaldehyde (pH 7.4) for 30 minutes at room temperature, permeabilized with 0.5% Triton X-100 for 5 minutes, and blocked with 1% bovine serum albumin for 30 minutes. The cells were incubated with the primary antibody (1:200 dilution) against HK1, HK2, or TOM20 overnight at 4°C. After washing twice with phosphate-buffered saline for 15 minutes, samples were incubated with Alexa Fluor-conjugated secondary antibody (1:500 dilution; Thermo Fisher Scientific) at room temperature for 1 hour. Immunofluorescent images were obtained using laser scanning confocal microscope (LSM 880, Carl Zeiss). The Pearson's correlation coefficient between hexokinases and TOM20, a mitochondrial marker, was determined by Imaris software (Oxford Instruments PLC, Oxon, UK). 

Statistical Analysis

Data were representative of at least two independent experiments, and were analyzed by analysis of variance with the Tukey post hoc test. A P value of less than 0.05 was considered statistically significant. 

Results

ROCK Inhibitor Increases Bioenergetic Capacity in BCECs

The bioenergetic capacity of cells is essential for meeting energy demands, particularly for those under metabolic stress conditions.¹⁵ Previous studies also showed that the mitochondrial respiratory ability is associated with the phenotype and the function of CECs, and the administration of ROCK inhibitor upregulates OXPHOS inside the cells.^11,16 To delineate the bioenergetic capacity of CECs and the impact of ROCK inhibition, we evaluated the effect of Y27632, a potent and selective ROCK inhibitor, on the ATP production rates through OXPHOS and glycolysis. To induce maximal ATP production from OXPHOS (mitoATP) and glycolysis (glycoATP), BCECs were separately incubated with carbonyl cyanide 4-(trifluoromethoxy) phenylhydrazone and monensin, an ionophore that facilitate the transport of sodium across the cell membrane and activation of Na/K-ATPase, thereby further increasing the energy demand.¹⁷ There was no apparent effect of ROCK inhibitor Y27632 on the basal ATP production through OXPHOS in BCECs. However, BCECs exhibited increased maximal ATP production rate through OXPHOS at lower Y27632 concentrations until the rate was contradictorily reduced at the highest concentration of Y27632 we tested in this study (Fig. 1A). By contrast, Y27632 increased both basal and maximal ATP production rate through glycolysis (Fig. 1B). The overall bioenergetic capacity, illustrated by the bioenergetic capacity plot, showed a marked increase in the presence of Y27632 (Fig. 1C). In addition to Y27632, the bioenergetic capacity of BCECs treated with ripasudil and Y39983, another two specific ROCK inhibitors, was also increased (Supplementary Fig. S1). These results indicated that ROCK inhibition expands the bioenergetic scope of BCECs. 

Figure 1.

 

ROCK inhibition altered the bioenergetic capacity of BCECs. (A and B) Basal and maximal ATP production rate by mitochondrial OXPHOS (A, mitoATP) and glycolysis (B, glycoATP) in control cells or cells incubated with different concentrations of Y27632. (C) Bioenergetic capacity plot for BCECs generated as in (A and B). The shaded rectangle regions indicate maximum bioenergetic scope. Symbols indicate basal and maximal ATP production rates, separately. n = 5 in each group. Data are shown as mean ± SD. *P < 0.05, **P < 0.01, ***P < 0.001. ns, nonsignificant by analysis of variance with Tukey post hoc test. Data are representative of two independent experiments.

Increased Glycolytic ATP Production by ROCK Inhibitor Under Monensin Stress

CECs, particularly those from FECD, are susceptible to metabolic stress that may compromise the ability to maintain cellular homeostasis, leading to accelerated cell dysfunction and death.¹⁸ To test the effect of bioenergetic enhancement by ROCK inhibition under metabolic stress situation, we challenged the BCECs with monensin and measured the ATP production. Monensin is an ionophore that facilitate the transport of sodium across the cell membrane and activation of Na/K-ATPase, thereby further increasing the energy demand.¹⁷ Under monensin stress, mitoATP was reduced, while glycoATP was increased (Figs. 2A, 2B). Preincubation with Y27632 significantly increased glycoATP production rate at basal state as well as under monensin stress, whereas its effect on mitoATP production was inconspicuous (Figs. 2A, 2B). The total ATP production rate was markedly increased in Y27632 group with glycolysis as the major source of ATP under monensin stress (Figs. 2C). These data demonstrated the ability of ROCK inhibition in enhancing glycolytic ATP production under stress conditions. 

Figure 2.

 

Monensin shifted the bioenergetics of BCECs towards using glycolysis as the major ATP production resource and ROCK inhibition further increased this trend. (A and B) ATP production rates by mitochondrial OXPHOS (A) and glycolysis (B) in control cells or cells incubated with 1 µM Y27632 that were further challenged with vehicle control or 10 µM monensin for 24 hours. n = 5 in each condition. (C) Total ATP production rates generated from (A) and (B). Data are shown as mean ± SD. *P < 0.05, **P < 0.01, ***P < 0.001. ns, nonsignificant by analysis of variance with Tukey post hoc test. Data are representative of two independent experiments.

ROCK Inhibitor Reduces Apoptosis Under Monensin Stress

Next, we validated whether increased ATP production by ROCK inhibitor confers any survival advantage under monensin stress. Phase-contrast imaging revealed that Y27632-treated cells maintained typical hexagonal morphology and better viability under monensin stress compared with untreated cells (Figs. 3A, 3B). Flow cytometry analysis after Annexin V/PI staining revealed increasing apoptosis at higher concentration of monensin in control cells, while Y27632 reduced rate of apoptosis (Fig. 3C). Consistently, the levels of cleaved caspase 3 and cleaved PARP were significantly increased under monensin stress in control cells, whereas Y27632 pretreatment led to decreased caspase 3 and PARP cleavage (Figs. 3D–F). In addition to monensin, we tested whether ROCK inhibitor also confers a survival advantage under menadione or IFN-γ stress, because a previous study shows that elevated cytokines in aqueous humor owing to iris atrophy and disruption of the blood–aqueous barrier led to oxidative stress and mitochondrial dysfunction, which ultimately result in CEC loss.¹⁹ Interestingly, preincubation with Y27632 enhanced the survival of BCECs under menadione or IFN-γ treatment (Supplementary Fig. S2). These results implied that BCECs with greater bioenergetic capacity can potentially produce more ATP to match with these stress conditions. 

Figure 3.

 

ROCK inhibition mitigated monensin-induced BCEC apoptosis. (A) Representative phase-contrast images of control BCECs or cells incubated with 1 µM Y27632 that were further challenged with vehicle control or 10 µM monensin for 24 hours. Scale bar, 200 µm. (B) Cell viability relative to control group determined by calcein AM staining followed by measuring the fluorescence intensity. n = 4 in each condition. (C) The percentage of apoptotic cells determined by annexin V/PI staining followed by flow cytometry analysis. n = 5 in each condition. (D) Representative Western blotting results using cellular lysates prepared from the cells that were undergoing the same experiments as described in (A and B) by the indicated antibodies. (E and F) Relative levels of cleaved caspase 3 and PARP normalized by GAPDH. n = 5 independent experiments per condition. Data are shown as mean ± SD. *P < 0.05, **P < 0.01, ***P < 0.001. ns, nonsignificant by analysis of variance with Tukey post hoc test.

ROCK Inhibitor Facilitates Sodium Output Under Monensin Stress

One of the mechanisms of monensin-triggered apoptosis is through increased intracellular sodium, which further leads to energy depletion, oxidative stress, and subsequent apoptosis.^20,21 To confirm that alleviated apoptosis observed after ROCK inhibition under monensin stress is associated with the regulation of intracellular sodium levels, which largely depends on Na/K-ATPase,¹ we quantified cytosolic sodium ion changes by time-lapse fluorescence imaging. Compared with those treated with vehicle control, monensin stress significantly increased intracellular sodium levels over time, reflected by fluorescence intensity of ION NaTRIUM Green (Figs. 4A, 4B). Preincubation with Y27632 markedly decreased intracellular sodium levels over time and at the last time point (Figs. 4A–C). These data indicate that increased bioenergetic capacity by ROCK inhibitor is associated with reduced sodium levels inside BCECs. 

Figure 4.

 

ROCK inhibition counteracted the effect of monensin on the accumulation of intracellular sodium. (A) Representative time-series fluorescence images of control or Y27632 (1 µM) pretreated BCECs with or without additional monensin stress (10 µM) stained with ION NaTRIUM Green. Scale bar, 100 µm. (B and C) Quantification of fluorescence intensity over time (B) and at the last time point (C). n = 3 in each condition. Data are shown as mean ± SD. **P < 0.01, ***P < 0.001, ns, nonsignificant by analysis of variance with Tukey post hoc test. Data are representative of three independent experiments.

ROCK Inhibitor Increases Recruitment of HK2 to the Mitochondria

Our data showed that ROCK inhibitor treatment increased glycoATP production under monensin stress that was associated with reduced BCEC apoptosis and enhanced sodium output. Intriguingly, our previous study demonstrated that ROCK inhibitor treatment only enriches the expression of OXPHOS-related gene sets, but not the glycolysis-related ones.¹¹ Compatible with our previous results that ripasudil increased OXPHOS through AMPK pathway, Y27632 treatment also lead to increased phosphorylation of AMPK and its downstream target acetyl-CoA carboxylase (Supplementary Fig. S3). To further investigate the mechanism of increased glycolysis by ROCK inhibitor under monensin stress, we first checked the levels of HK1 and HK2, two crucial enzymes responsible for the initiation of glycolysis and bioenergetics capacity.²² However, the expression levels of HK1 and HK2 did not change significantly with Y27632 treatment or under monensin stress (Supplementary Fig. S4). Previous studies showed that not only the total expression levels, but also the subcellular localization of HKs, determines their role in metabolism.²³ We further investigated whether there were any changes in the subcellular localization of these two enzymes, and found that both enzymes were significantly relocated to mitochondria under monensin stress, with the recruitment of HK2 to mitochondria being even more evident after Y27632 pretreatment (Fig. 5). These results provide mechanistic insight about how ROCK inhibitor affects glycolysis in BCEC through regulating the localization, but not the total level, of HK. 

Figure 5.

 

ROCK inhibitor enhanced the presence of HKs at the mitochondria. (A and B) BCECs pretreated with or without 1 µM Y27632 that were further treated with vehicle control or 10 µM monensin were subjected to double immunostaining with the anti-TOM20 and anti-HK1 (A) or anti-HK2 (B) antibodies. Scale bars, 50 µm. (C and D) Quantification of the extent of colocalization of HK1 (C) and HK2 (D) with TOM20 by Pearson's correlation coefficient. n = 4 in each condition. Data are shown as mean ± SD. **P < 0.01, ***P < 0.001. ns, nonsignificant by analysis of variance with Tukey post hoc test. Data are representative of two independent experiments.

Importance of Glycolysis for ROCK Inhibitor-Facilitated Sodium Export and Cell Viability Maintenance Under Monensin Stress

To further substantiate the dependence on glycolysis for ROCK inhibitor to enhance sodium export and maintain cell viability, we tested the response of BCECs to monensin stress in either glucose or galactose-based culture medium. Compared with glucose medium, using galactose as the carbon source forced BCECs to rely on OXPHOS, whereas the glycolysis was suppressed (Supplementary Fig. S5). Consistent with previous results, Y27632 pretreated BCECs showed accelerated sodium export under monensin stress in glucose medium (Figs. 6A, 6C). However, the effect of Y27632 was abolished when glycolysis was inhibited (Figs. 6B, 6D). Consistently, BCEC viability was preserved by Y27632 pretreatment under monensin stress in glucose medium, but not in galactose medium (Figs. 6E, 6F). These data suggested that the enhanced sodium export and the preservation of cell viability by ROCK inhibitor under monensin stress depend on glycolytic ATP production, which may be used to fuel the Na/K-ATPase pump and maintain cellular homeostasis (Fig. 7). 

Figure 6.

 

Enhanced glycolysis is crucial for ROCK inhibitor to facilitate export of intracellular sodium and preserve BCEC viability under monensin stress. (A and B) BCECs pretreated with or without 1 µM Y27632 were further treated with vehicle control or 10 µM monensin in glucose (A) or galactose (B) media. Cells were stained with ION NaTRIUM Green and were subjected to time lapse imaging. The images taken at the last time point (16th hour) were shown. Scale bar, 100 µm. (C and D) Quantification of the fluorescent intensities of the images shown in (A) and (B). (E and F) Cell viability relative to control group as determined by calcein AM staining of BCECs under monensin stress in glucose (E) or galactose (F) media. n = 4 in each condition. Data are shown as mean ± SD. *P < 0.05, **P < 0.01, ***P < 0.001. ns, nonsignificant by analysis of variance with Tukey post hoc test. Data are representative of two independent experiments.

Figure 7.

 

Schematic presentation of increased bioenergetic capacity stimulated by ROCK inhibitor in BCECs. The enhanced ATP production from both OXPHOS and glycolysis in BCECs enables the cells to better withstand monensin-induced stress by facilitating sodium ion export, thereby reducing apoptosis and promoting cell survival. ADP, adenosine diphosphate.

Discussion

In this study, we found that ROCK inhibition by Y27632 increased bioenergetic capacity in BCECs. Furthermore, a similar phenomenon can be observed by using ripasudil and Y39983, the other two specific ROCK inhibitors. This result indicated that the ability to augment bioenergetic capacity in BCECs was a general effect across different ROCK inhibitors. The bioenergetic capacity of cells is crucial for maintaining energy homeostasis and supporting cellular functions, especially under stress conditions.^17,24 CECs commit a substantial amount of energy to sustain pump function.²⁵ As mentioned elsewhere in this article, decreased CEC density leads to an increase in pump sites to compensate for the elevated permeability,⁴ and the energy burden per remaining cells may overwhelm the bioenergetic capacity, resulting in metabolic stress and apoptosis. These evidences warrant further investigation into the mechanisms of enhancing bioenergetic capacity of CECs to improve their resilience against metabolic stress. 

In this study, we used monensin, an ionophore, to generate metabolic stress through artificially increasing the energy demand via the disruption of sodium ion gradients, thereby forcing the BCECs to expend more ATP via the Na/K-ATPase to maintain homeostasis.¹² The previous study showed that monensin inflicts various stresses on the CECs, including Golgi stress and endoplasmic reticulum stress, in addition to the sodium ionophore effect.²⁶ When facing endoplasmic reticulum stress, increased mitochondrial metabolism and enhanced cellular bioenergetics are critical for the cellular adaptation.²⁷ Our results showed that the increased bioenergetic capacity by ROCK inhibitor was associated with a lower intracellular sodium level, indicating that the energy, largely derived from glycolysis, supported Na/K-ATPase to pump out sodium ion. Therefore, from a bioenergetic perspective, ROCK inhibitor treatment enables BCECs to better cope with the metabolic demand under monensin stress that results in both endoplasmic reticulum stress and increased sodium ion inflow, thereby maintaining ion homeostasis and cellular function. 

Compatible with our previous study,¹¹ we found that Y27632, a specific ROCK inhibitor, increased phosphorylation of AMPK, a central mediator of cell energy homeostasis,²⁸ and the maximal ATP production through OXPHOS was correspondingly increased. Interestingly, Y27632 also increased maximal ATP production through glycolysis in a dose-dependent manner. Although the gene set enrichment analysis in the previous study showed no significant enrichment of glycolysis related genes, a ROCK inhibitor may regulate glycolysis through other mechanisms, such as the translocation of HK to the mitochondria.²⁹ Independent RNA sequencing assays had been performed in both Y-27632– and ripasudil–treated cell groups and further bioinformatics study consistently demonstrated that both ROCK inhibitors can elicit the expression of OXPHOS-related genes, presumably through the AMPK pathway, while have no effect on glycolysis related genes.¹¹ These findings are indeed congruous with the differential roles of OXPHOS and glycolysis in cellular bioenergetic; that is, OXPHOS generates ATP in a more efficient way, whereas glycolysis supplies ATP more rapidly.³⁰ Therefore, glycolysis is usually the cellular source of ATP under stressful situations. To enable glycolysis as a fast-reacting machinery, the regulatory signals should converge on a few molecules involved in glycolysis and better governed through a limited step such as post-translational modification of the putative key regulator instead of through more complicated signaling relays, which would generally require a readout of gene promoter activation.³¹ Previous studies showed that the subcellular localization of HK, the rate-limiting enzyme in glycolysis, dictates the fate of glucose metabolism, with mitochondria-bound hexokinases promote glycolysis, whereas cytosolic hexokinases increase glucose flux through anabolic pathway.^32,33 Indeed, we found that, although the overall levels of HK1 and HK2 remained unchanged, the recruitment of HK1 and HK2 to the mitochondria was enhanced under monensin stress. Furthermore, the extent of mitochondria-bound HK2 was even more obvious with the addition of Y27632, which may allow preferential access of HK2 to ATP generated by mitochondria, thereby facilitating the initiation of glycolysis and optimizing their function and efficiency in ATP production.³⁴ Whether ROCK or its downstream kinase such as AMPK can regulate the post-translational modification status of hexokinases and then their positional proximity to the mitochondria is an intriguing idea worth testing. 

Intriguingly, a previous study by Yamaguchi et al. demonstrated that CECs from bullous keratopathy showed upregulation of the complement activation pathway and downregulation of the glycolytic process and dysfunctional mitochondria. Furthermore, cytokine levels, including that of IFN-γ, and 8-OHdG level, a biomarker for oxidative stress, are elevated in the aqueous humor, indicating that microenvironmental changes in the aqueous humor of patients with bullous keratopathy predisposes to mitochondrial dysfunction and subsequent cell loss.¹⁹ In addition to monensin, we also found that ROCK inhibitor pretreatment maintained the BCEC viability under menadione and IFN-γ stresses, reflecting the findings from previous studies that cells, with greater bioenergetic capacity, can potentially produce more ATP if needed to cope with stress conditions, such as energetic and oxidative stresses.^24,35 In contrast with bullous keratopathy, Yamaguchi et al.¹⁹ showed that CECs from FECD exhibit upregulation of canonical glycolysis, highlighting differences in energy production impairments among corneal endothelial diseases. However, another study showed that CECs from FECD initially compensate for the cell loss and the sustained ATP requirement by increasing mitochondrial mass, calcium and membrane potential. After reaching maximum capacity, mitochondrial burnout ensues, leading to apoptosis.⁵ These studies showed that, although the initial presentation and energy production impairments vary across different clinical scenarios, mitochondrial dysfunction consistently plays a central role in disease progression. Because ROCK inhibitor boosted bioenergetic scope by increasing maximal ATP generation from both OXPHOS and glycolysis, thereby facilitating CECs to endure metabolic stress from monensin as well as stresses from menadione and IFN-γ, further studies are warranted to elucidate the protective effect of CECs from bullous keratopathy or FECD. 

There are some limitations to this study, including species differences in the cultured CECs and a lack of further validation of the pump function. Based on the observation that ROCK inhibitor-treated BCECs showed decreased intracellular sodium level under monensin stress, the pump function may be enhanced by ROCK inhibitor treatment by increasing bioenergetic capacity. This result not only sheds insight in how ROCK inhibitor boosts BCEC function and survival, but also warrants further validation in cultured human CECs as well as in vivo. In conclusion, our findings demonstrated that ROCK inhibitor enhances the bioenergetic capacity of BCECs by increasing ATP production through both OXPHOS and glycolysis. The enhancement from glycolysis enables BCECs to better endure metabolic stress by generating ATP rapidly to meet increased energy requirement and preserving ion balance, which is vital for the survival and function of CECs in scenarios such as corneal endothelial dystrophies or surgical trauma. 

Acknowledgments

Funded by the National Science and Technology Council, Taiwan (NSTC), Grant number: 112-2314-B-418-002, 112-2320-B-002 -021; and the Far Eastern Memorial Hospital, Taiwan (FEMH), Grant number: 111DN29, 112DN29, 113DN24, 109FTN14, 110FTN12, 112FTN0006. 

Disclosure: W.-T. Ho, None; J.-S. Chang, None; C.-J. Lei, None; T.-C. Chen, None; J.-K. Wang, None; S.-W. Chang, None; M.-H. Yang, None; T.-S. Jou, None; I.-J. Wang, None