Slc4a11 ^−/− (KO) mice were originally provided by Dr. Eranga Vithana (Singapore Eye Research Institute) (Vithana et al., 2006). Slc4a11 ^+/+ (Normal, Wild Type, WT) and KO mice were housed and maintained in pathogen-free conditions. Animals were used in the experiments in accordance with institutional guidelines and the current regulations of the National Institutes of Health, the United States Department of Health and Human Services, the United States Department of Agriculture and Association for Research in Vision and Ophthalmology (ARVO) Statement for the Use of Animals in Ophthalmic and Vision Research.

Conditionally immortalized mouse corneal endothelial cells (MCEC) Slc4a11 ^+/+ and Slc4a11 ^−/− were generated and maintained in our lab (Zhang et al., 2017b). Cells were cultured in Complete Media, which contains OptiMEM-I medium (#51985; Thermo Fisher Scientific, Canoga Park, CA, United States), 14 mM Glucose and 4 mM L-Alanyl Glutamine supplemented with 8% heat-inactivated fetal bovine serum (FBS) (#10082139; Thermo Fisher Scientific), EGF 5 ng/ml (#01–107 Millipore, Darmstadt, Germany), pituitary extract 100 μg/ml (Hyclone 15 Laboratories, Logan, UT, United States), calcium chloride 200 mg/L, 0.08% chondroitin sulfate (#G6737; SigmaAldrich Corp., St. Louis, MO, United States), gentamicin 50 μg/ml (#15710072; Thermo Fisher Scientific), antibiotic/antimycotic solution diluted 1:100 (#15240062; Thermo Fisher Scientific), and 44 units/mL IFN-γ (#485-MI; R&D Systems, Minneapolis, MN, United States).

For experiments, cells were incubated in Assay Media, which contained Dulbecco’s Modified Eagle Medium (no glutamine, no sodium pyruvate, no phenol red, contains 5.5 mM glucose) (#11054001; Thermo Fisher Scientific), supplemented with 0.5 mM glutamine (#250030-081, Thermo Fisher Scientific) and 0.5% dialyzed FBS (#26400-036; Thermo Fisher Scientific) at 33°C for 16 h. Corneal endothelium is exposed to 0.5 mM glutamine in vivo (Langford et al., 2007), therefore this concentration was used in our experiments. Drug treatments, 2 μM MitoQ (#317102, Medkoo Biosciences, Morrisville, NC, United States), or 0.1 μM Thapsigargin (SML 1845, Sigma Aldrich) were added into the assay media for 16 h. All drug treatments were carried out in assay media. Significant elevation in mitochondrial ROS was observed after 16 h treatment of cells in assay media (Shyam et al., 2021).

Total RNA was isolated using RNA mini kit (#74104, Qiagen, Germantown, Maryland, United States). 1μg of RNA was used to prepare cDNA using a high capacity RNA to DNA kit (#4388950, Thermo Fisher Scientific). Previously published primer designs were used to amplify XBP1, us-XBP1, s-XBP1, and β-actin (Yoon et al., 2019). NCBI primer designing tool was used to design all other primers used for this study. List of primers used can be found in Table 1. Real time PCR was conducted using SYBR green dye using a BioRad CFX96 system. Relative quantitation was performed using 2^−ΔΔCt method against housekeeping gene. Fold Change (FC) is calculated as 2^−∆∆CT. Data is plotted on Log10 scale.

Immortalized Slc4a11 ^+/+ (WT) and Slc4a11 ^−/− (KO) corneal endothelial cells were cultured on fibronectin pre-coated 25-mm diameter glass coverslips (GG-25-pdl; Neuvitro Corporation, Vancouver, WA, United States) for 24 h in Assay media with or without glutamine. 1 mM stock of Fura 2-AM (#F1221, ThermoFisher Scientific) was prepared in DMSO. 5 μL of Fura stock solution along with 5 μL of 20% Pluronic F-127 solution in DMSO (#P3000MP, ThermoFisher Scientific) were mixed in 1 ml Hanks Balanced Salt Solution (HBSS) to obtain a final Fura 2-AM concentration of 5 μM. Cells were loaded with Fura-2AM for 30 min at 37°C. Cover slips were washed in HBSS for 30 min at room temperature, before they were mounted into a perfusion chamber, connected to a stage warmer (37°C) of an inverted microscope (Eclipse TE200; Nikon, Tokyo, Japan). Cells were perfused with HBSS containing 500 μM non-cell permeant BAPTA for 50 s to establish baseline. After establishing the fluorescence baseline, cells were perfused with HBSS containing 50 μM Ionomycin (#2092, Tocris, Minneapolis, MN, United States). Perfusing solutions were adjusted to pH 7.5 using NaOH, and kept at 37°C in a warming box. The flow of the perfusate (∼0.5 ml/min) was by gravity. Osmolarity of all solutions was adjusted to 295 mOsm with sucrose. Cells were imaged with a 40x oil-immersion objective (Nikon). Cells loaded with Fura-2AM were excited at 340 and 380 nm and the emission was collected at 505 nm to measure changes in cytosolic calcium. Ca²⁺ bound Fura-2 has an excitation maximum of 340 nm, while Ca²⁺ free Fura-2 has its excitation maximum of 380 nm. In both states, the emission maximum is 510 nm. The 340/380 nm excitation ratio for Fura-2 is a measure of intracellular [Ca²⁺].

Endothelium-Descemet’s tissue samples from WT and KO mice were fixed with 2.5% glutaraldehyde (#16020, Electron Microscopy Sciences, Hatfield, PA, United States), 4% paraformaldehyde (#15710, Electron Microscopy Sciences), in 0.1 M sodium cacodylate buffer, pH 7.2 at 4°C and post-fixed with 1% osmium tetroxide (# 19,150, Electron Microscopy Sciences) in 0.1 M sodium cacodylate buffer (#12300, Electron Microscopy Sciences), pH 7.2 at 4°C. Samples were dehydrated in a graded ethanol series to 100% ethanol, transitioned to propylene oxide (#20401, Electron Microscopy Sciences), and infiltrated with Embed 812 resin (#14120, Electron Microscopy Sciences). Infiltrated samples were placed in flat embedding molds and polymerized at 65°C for 18 h. Resin blocks were cut with a diamond knife using a Leica Ultracut UCT ultramicrotome (Leica Systems, Buffalo Grove, IL). Sections were placed on 300 mesh copper TEM grids (#0300-CU, Electron Microscopy Sciences) and stained with saturated uranyl acetate in aqueous solution, and lead citrate. Stained sections were viewed with a JEM-1010 TEM (JEOL, Peabody, MA, United States) at 80 kV and photographed with a Gatan MegaScan 794 CCD camera or JEM-1400plus TEM (JEOL) with a Gatan OneView CMOS digital camera.

Corneal endothelial cell layer was removed from dissected corneas. Protein lysates were prepared by pooling the tissues from two animals, using radioimmunoprecipitation (RIPA) lysis buffer containing protease and phosphatase inhibitors. Equal amounts of protein (1.5 µg) were loaded into 12–230 kDa separation module kit, and analyzed using the Protein Simple Wes System (Protein Simple, San Jose, CA, United States) following the manufacturer’s instructions as previously described (Shyam et al., 2021). Antibodies used are BIP (#3177, Cell Signaling technologies, Danvers, MA, United States), GADD153 (#NB600-1,335, Novus, Centennial, CO, United States), and α-tubulin (#NB100-690, Novus).

Wes immunoassay was carried out since it is challenging to obtain sufficient total protein from corneal endothelial peelings to conduct traditional western blots. Our lab has used this approach in several recent studies to quantify changes in protein expression (Ogando et al., 2021; Shyam et al., 2021; Shyam et al., 2022; Ogando and Bonanno, 2022).

Slc4a11 ^+/+ corneal endothelial cells were treated with 0.25 μM Antimycin A (#A8674, Sigma-Aldrich), 0.5 μM Rotenone (#R8875, Sigma-Aldrich), or 0.25 μM Antimycin A+ Rotenone in assay media for 18 h. Control cells were cultured in assay media for the same duration.

Slc4a11 ^+/+ corneal endothelial cells were trypsinized and stained in triplicate with Pacific Blue Annexin V kit with PI (#640928, BioLegend, San Diego, CA, United States) or MitoSOX (#M36008, Thermo Fisher Scientific) following manufacturer’s instructions. Cells were collected in 2 ml micro centrifuge tubes following filtration using CellTrics Fliters (#04-004-2,327, Sysmex, Gorlitz, Germany). Flow cytometry analysis was conducted on MACSQuant VYB (Miltenyi Biotech, Germany). 10,000 cells were collected for each acquisition. Data were analyzed using FCS Express (De Novo software, Pasadena, CA, United States).

MCEC cultures in 12-well format were trypsinized, washed and fixed using 4% Paraformaldehyde. Following permeabilization, cells were stained for aggresomes using the Aggresome detection kit (#ENZ51035, Enzo Life Sciences, NY) following manufacturer’s instructions. The aggresome detection reagent emits strong fluorescence when bound by misfolded proteins but not in solution. Therefore, the increase in fluorescence intensity is indicative of the levels of protein aggregates present in the cells. After filtration using 50 μm sterile CellTrics Filters flow cytometry analysis was conducted on MACSQuant VYB. 10,000 cells were collected per acquisition. Cells treated with ER stress inducer, Thapsigargin (#1138, Tocris, Minneapolis, MN, United States), served as a positive control, unstained cells were used as negative control. Data were analyzed with FCS Express.

MitoQ, 1 mM (#317102; Medkoo Bisociences) was prepared in equal volume of ethanol: distilled sterile water mixture. 100 μL of this solution containing 68 μg of MitoQ was injected intraperitoneally into Slc4a11 ^−/− mice on alternate days. Littermate control mice were injected with equal volume of vehicle for the same duration at the same time/day. To prevent stress induced weight loss due to increased handling, animals were provided with a vet-approved supplement. One tablespoon of commercially available peanut butter was provided to both control animals and experimental animals, once a day immediately after handling.

All experiments were performed at least three times on different days. Error bars represent mean ± SD. Statistical significance was calculated using unpaired t-tests when two groups were involved. ANOVA with Tukey’s multiple comparisons test was used to determine statistical significance if more than two groups were analyzed. Statistical analyses were conducted using Graph Pad Prism software (La Jolla, CA, United States).

The model figures were created with the aid of BioRender.com.

ER is the major site for protein folding in eukaryotes. When the protein folding machinery is overwhelmed by the influx of nascent polypeptides, ER stress ensues. The Unfolded Protein Response (UPR) pathway is activated during ER stress to promote cell survival through three distinct pathways (Figure 1A). However, chronic UPR activation leads to ER-stress-induced apoptosis (Adams et al., 2019; Bartoszewska and Collawn, 2020).

Misfolded proteins trigger the unfolded protein response (UPR) through three primary sensors, Inositol-Requiring Enzyme 1 α (IRE1-α), Protein kinase R-like Endoplasmic Reticulum Kinase (PERK), and Activating transcription factor 6 (ATF6). Activation of UPR leads to increased expression of ER chaperone protein BIP (Binding Immunoglobulin Protein) and chronic ER stress induces apoptosis through ER-related apoptosis protein GADD153 (Growth Arrest DNA Damage protein153, also known as CHOP) (Figure 1A) (Qi et al., 2017). In the mouse corneal endothelium, we observed increased BIP and GADD153 expression in 10 week old Slc4a11 ^−/− animals but not in the age-matched Slc4a11 ^+/+ animals (Figures 1B,C). Another established characteristic of ER stress is the presence of dilated ER lumen (Bernales et al., 2006; Despa, 2009; Hartley et al., 2010; Schönthal, 2012). Therefore, we conducted electron microscopy to determine whether there are any structural changes in the ER of the corneal endothelia of Slc4a11 ^−/− mice. Ten-week old Slc4a11 ^−/− mice corneal endothelia reveal the presence of dilated ER lumen that was not observed in WT samples of age-matched animals (Figure 1D).

Glutamine-induced mitochondrial ROS is the primary cellular stress in Slc4a11 ^−/− corneal endothelium (Ogando et al., 2019). To determine its contribution toward ER stress, we treated Slc4a11 ^+/+ and Slc4a11 ^−/− immortalized corneal endothelial cells with or without glutamine. Flow cytometry analysis showed a significant increase in misfolded protein aggregates (aggresomes) in Slc4a11 ^−/− compared to WT cells in glutamine media (Figures 2A,B). Unstained cells were used as negative control, and cells treated with the ER stress inducer Thapsigargin, was used as positive control for aggresome formation (Supplementary Figure S1).

Glutamine incubation also elevated expression of UPR associated proteins, BIP and GADD153 in Slc4a11 ^−/− cells consistent with the accumulation of misfolded proteins as an outcome of ER stress (Figures 3A,B). Splicing of X-Box Binding Protein-1 (XBP1) mRNA occurs in response to the IRE-1α pathway (Figure 1A). Therefore, if there is ER stress and IRE-1α is activated we expect an increase in the levels of spliced XBP1, but not the unspliced version. Using primers (Table 1) that can bind to unspliced XBP1, and spliced XBP1 (Yoon et al., 2019), we conducted real-time PCR. Slc4a11 ^−/− cells in the presence of glutamine had increased levels of the spliced XBP1 transcripts (Figure 3C), whereas the unspliced XBP1 levels were decreased. This result is consistent with ER stress induced activation of IRE-1α pathway.

ER is the largest reservoir of Ca²⁺ in eukaryotic cells. We used the calcium indicator dye, Fura-2, to measure cytosol [Ca²⁺]. A baseline fluorescence ratio was established using Hanks Balanced Saline Solution (HBSS) containing calcium chelator BAPTA so that extracellular calcium was minimal. Following this, the cells were perfused with the calcium ionophore ionomycin, which under these conditions releases Ca²⁺ from ER intracellular stores. Live-cell microscopy revealed similar baseline and ionomycin stimulated change in cytosolic [Ca²⁺] for both Slc4a11 ^+/+ and Slc4a11 ^−/− cells when perfused without glutamine (Figure 4A). However in the presence of glutamine, ionomycin released significantly less calcium into the cytosol of Slc4a11 ^−/− cells indicating that the ER calcium store is deficient (Figure 4B).

To determine if oxidative stress was sufficient to induce ER stress, WT corneal endothelial cells were treated with mitochondrial ROS inducers - rotenone (mitochondrial electron transport chain complex 1 inhibitor) or antimycin A (mitochondrial electron transport chain complex III inhibitor) or a combination of both drugs. Flow cytometry analysis of MitoSOX indicated that all three treatment conditions increased mitochondrial oxidative stress (Figure 5A), while significant apoptotic cell death was evident only with 0.5 μM rotenone treatment (Figure 5B). Significant increase in aggresome levels (Figure 5C) as well as elevation in ER stress marker, BIP (Figures 5D,E), were observed with 0.5 μM Rotenone treatment indicating that mitochondrial oxidative stress alone is sufficient to cause ER stress and that it is not simply due to the lack of Slc4a11.

In the absence of the ammonia sensitive mitochondrial uncoupling by Slc4a11, glutamine catabolism leads to mitochondrial membrane hyperpolarization that elevates mitochondrial ROS in corneal endothelial cells (Ogando et al., 2019). Since we now show that glutamine is causing ER stress, we asked whether quenching mitochondrial ROS can decrease ER stress in Slc4a11 ^−/− cells. Slc4a11 ^−/− corneal endothelial cells were incubated with 2 μM MitoQ for 24 h in assay media, which was previously shown to significantly reduce mitochondrial ROS (Shyam et al., 2021). Figures 6A,B show that this treatment significantly reduced BIP and GADD153 expression, and aggresome levels in KO cells (Figures 6C,D). Interestingly, Figure 6E shows that MitoQ in the presence of glutamine lowered baseline calcium in both WT and KO corneal endothelial cells and restored the release of calcium from internal stores by ionomycin in KO cells to a level similar to that observed in glucose alone (see Figure 4A).

To determine whether MitoQ was sufficient to improve ER stress in vivo, 8-week old Slc4a11 ^−/− animals were injected with this drug on alternate days over the course of 4 weeks (Shyam et al., 2021). Following MitoQ treatment, BIP expression in dissected corneal endothelium was determined using Wes immunoassay. Significant reduction in BIP levels were observed in Slc4a11 ^−/− animals compared to littermate controls when treated with MitoQ (Figures 6F,G). These observations are consistent with our previous findings in which MitoQ treatment alleviated autophagy and lysosomal dysfunction in Slc4a11 ^−/− animals (Shyam et al., 2021).