High Throughput Assay Identifies Glafenine as a Corrector for the Folding Defect in Corneal Dystrophy–Causing Mutants of SLC4A11

Abstract

Purpose: Protein misfolding, causing retention of nascent protein in the endoplasmic reticulum (ER), is the most common molecular phenotype for disease alleles of membrane proteins. Strategies are needed to identify therapeutics able to correct such folding/trafficking defects. Mutations of SLC4A11, a plasma membrane transport protein of the human corneal endothelial cell layer, cause cases of congenital hereditary endothelial dystrophy, Harboyan syndrome, and Fuchs' endothelial corneal dystrophy. Most SLC4A11 mutations induce SLC4A11 misfolding and retention in the ER.

Methods: An assay amenable to high-throughput screening was developed to quantify SLC4A11 at the plasma membrane, enabling a search for potential traffic-correcting small molecules. The assay was validated by comparing cell surface abundance of SLC4A11 mutants measured in the assay to observations from confocal immunofluorescence and values from cell surface biotinylation. Functionality of mutant proteins was assessed, using a confocal microscopic green fluorescent protein (GFP) water flux assay where relative rates of cell swelling are compared.

Results: A small-scale screen revealed that the nonsteroidal anti-inflammatory drugs (NSAIDs), glafenine, ibuprofen, and acetylsalicylic acid dissolved in 0.2% dimethyl sulfoxide (DMSO), partially rescued the trafficking defect in some SLC4A11 mutants, expressed in HEK293 cells. These SLC4A11 mutants retained functional activity when rescued to the plasma membrane by glafenine treatment. Glafenine was effective with an EC₅₀ of 1.5 ± 0.7 μM.

Conclusions: These data suggest that glafenine, and perhaps other NSAIDs, hold potential as therapeutics for misfolded membrane proteins, like SLC4A11. The high throughput approach described here can be modified to identify correctors of other misfolded plasma membrane proteins that cause eye disease.

Among genetic diseases of membrane proteins, the most common phenotype is protein ER-retention, secondary to protein misfolding. A classic example is the most common cystic fibrosis allele, CFTR F508del, which impairs the folding and trafficking of CFTR protein, resulting in endoplasmic reticulum (ER)-retention.^1,2 Potential small molecules correcting F508del folding have been identified that rescue F508del CFTR to the plasma membrane,^3–5 including one that has entered recent clinical use.^6,7 

Small molecule-folding correctors are thus proven to be effective therapeutics for ER-retained membrane proteins. Development of assays amenable to high throughput drug screening is the essential first step toward identification of folding corrector drugs. Here we have targeted for folding correction an integral membrane protein, SLC4A11, whose mutations cause ER-retention.^8,9 Approximately 60 disease-causing point mutations of human SLC4A11 have been identified.^8–25 Some ER-retained mutants of SLC4A11 could be rescued to the cell surface in cells cultured at 30°C, suggesting a temperature-sensitive folding defect.²⁶ Moreover, the rescued protein displayed functional activity upon rescue,²⁷ suggesting that rescue from the ER is a promising therapeutic approach. SLC4A11 is an attractive therapeutic target as its mutations cause corneal defects leading to blindness. The relative accessibility of the cornea opens the possibility to apply small molecule-folding correctors as eye drops, meaning that the treatment approach could be relatively facile and localized to the site of disease. 

In human cornea, high solute concentration of the stromal layer creates an osmotic gradient that draws water from the aqueous humor.²⁸ This osmotic gradient is opposed by water reabsorption into the aqueous humor by the endothelial monolayer. Posterior endothelial corneal dystrophies develop when this reabsorption is interrupted,²⁸ giving rise to an accumulation of fluid in the stroma. The edematous corneas develop a ground-glass appearance, leading to vision loss and eventual blindness. 

Congenital hereditary endothelial dystrophy (CHED; Mendelian Inheritance in Man [MIM] 217700),²⁹ Harboyan syndrome (HS; MIM 217400),³⁰ and Fuchs' endothelial corneal dystrophy (FECD; MIM 613268)³¹ are three forms of genetic corneal blindness that arise in some cases from mutations in the integral membrane protein, SLC4A11.^8,15,20 Harboyan syndrome patients present with sensorineuronal hearing loss in addition to progressive blindness.²⁸ Congenital hereditary endothelial dystrophy and HS are recessive, whereas FECD is dominant with a 4% prevalence in North America.²⁸ Fuchs' endothelial corneal dystrophy has also been linked to mutations of the COL8A2,³² LOXHD1,³³ ZEB1,³⁴ and the transcriptional regulator, TCF4.^35,36 

SLC4A11 is a member of the SLC4 family of bicarbonate transporters, but does not transport bicarbonate.³⁷ Plant SLC4A11 orthologs are established borate transporters.³⁸ Human SLC4A11 was originally reported to be a Na⁺-coupled borate transporter,³⁷ but other groups have not been able to replicate this finding.^39,40 Instead, SLC4A11 has been found to facilitate Na⁺/OH⁻ transport,⁴¹ NH₃/H⁺ cotransport,⁴² and electrogenic H⁺ (OH⁻) permeation.⁴³ Human SLC4A11 mediates water movement when expressed in Xenopus laevis oocytes and HEK293 cells,⁴⁴ which makes it the first identified water transporter that is not a member of the major intrinsic protein family. 

Depletion of SLC4A11 leads to degeneration and apoptosis of corneal endothelial cells.⁴¹ Mutant SLC4A11 does not induce apoptosis in HEK293 cells on its own.²⁷ One report found that HEK293 cells transfected with mutant SLC4A11 decrease expression of antioxidant proteins, rendering the cells more susceptible to oxidative stress.⁴⁵ Finally, three independent slc4a11^−/− mouse lines manifest varying degrees of corneal abnormalities,^46–48 indicating that loss of SLC4A11 function causes corneal dysfunction. 

Congenital hereditary endothelial dystrophy, HS, and FECD differ in their genetics of inheritance and age of onset, which in part is explained by SLC4A11 dimerization.²⁶ Congenital hereditary endothelial dystrophy and HS show autosomal recessive inheritance with symptom onset in the first decade of life.^20,28 SLC4A11 CHED mutant/wildtype (WT) heterodimers traffic to the cell surface, enabling sufficient traffic of WT protein to the cell surface to prevent disease symptoms in carriers.²⁶ The recessive ER-retention phenotype of CHED SLC4A11 explains this disease's inheritance. Individuals with FECD do not experience symptoms until the fourth or fifth decade of life.²⁸ Fuchs' endothelial corneal dystrophy SLC4A11/WT heterodimers do not traffic to the cell surface, indicating that the FECD phenotype is dominant over WT.²⁶ 

In this study, we developed and tested an assay amenable to high throughput screening to identify small molecule correctors of the mutant SLC4A11 cell surface trafficking defect. The assay approach was modified from one used to identify correctors of F508del CFTR folding.^49,50 To further assess the ability of the technique to identify SLC4A11 therapeutics, we performed a small scale screen of potential correctors. One candidate compound, glafenine, was characterized and found to show a promising ability to rescue SLC4A11 cell surface functional activity. 

Materials and Methods

Materials

Oligonucleotides were from Integrated DNA Technologies (Coralville, IA, USA). Q5 Site Directed Mutagenesis Kit was from New England Biolabs (Ipswich, MA, USA). Dulbecco's modified Eagle's medium (DMEM), fetal bovine serum (FBS), calf serum (CS), penicillin-streptomycin-glutamine, Geneticin, and Amplex UltraRed Reagent were from Life Technologies (Carlsbad, CA, USA). Cell culture dishes were from Sarstedt (Montreal, QC, Canada). Complete protease inhibitor tablets were from Roche Applied Science (Indianapolis, IN, USA). Immobilized Streptavidin Sepharose resin, sulfo-NHS-SS-biotin, glass coverslips and 10% formalin in phosphate buffer were from Thermo Fisher Scientific (Ottawa, ON, Canada). Poly-L-lysine was from Sigma-Aldrich (Oakville, ON, Canada). Hydrogen peroxide was from Ricca Chemical Company (Arlington, TX, USA). Immobilon-P PVDF was from Millipore (Billerica, MA, USA). Monoclonal antibodies against HA epitope (clone 16B12) and Glyceraldehyde 3-phosphate dehydrogenase (GAPDH) were from Covance (Princeton, NJ, USA) and Santa Cruz Biotechnology (Santa Cruz, CA, USA), respectively. Horseradish peroxidase-conjugated sheep anti-mouse immunoglobulin was from GE Healthcare Bio-Sciences Corp. (Piscataway, NJ, USA). Luminata TM Crescendo Western HRP Substrate chemiluminescence reagent was from Millipore. All compounds for screening were from Sigma-Aldrich, Thermo Fisher Scientific, or Cayman Chemical (Ann Arbor, MI, USA). 

DNA Constructs

The eukaryotic expression construct (pSKL1) for splicing variant 2 of human SLC4A11, encoding an 891 amino acid protein (NCBI reference: NG_017072.1) with an N-terminal Hemagglutinin tag (HA-tagged) was described previously.²⁷ Double HA-epitope tags were inserted at cDNA positions encoding amino acid 530 or 564, using untagged pSKL1 as template. HA530-SLC4A11 was constructed, using the forward primer 5′-ggtaaagtccacctgctgtc-3′ and the reverse primer 5′-gctgacaagggatgaagtagcgtaatctggaacatcgtatgggtaccctccgccagcgtaatctggaacatcgtatgggtacctttttgtgtgatagtcgtcc-3′, which contains the double HA-epitope tag (underlined). The resulting PCR product was used as a mega-primer in conjunction with the reverse primer 5′-agcagcaacagggacagg-3′ and subcloned with pSKL1 using KpnI and BsrFI restriction sites. HA564-SLC4A11 was constructed using the same strategy but replaced the initial reverse primer with 5′-cctgagtgtgtggccgaagcgtaatctggaacatcgtatgggtaccctccgccagcgtaatctggaacatcgtatgggtagggcagctccgtggggct-3′ containing the double HA-epitope tag (underlined). 

A shortened version of HA564-SLC4A11 removing the first 35 amino acids of the protein, resulting in an expression construct (pAMC1) encoding an 856 amino acid protein, was created using the Q5 site directed mutagenesis kit. All subsequent modifications to pAMC1 were carried out with the Q5 Kit. For ease of reference, numbering conventions for point mutations and HA tags are the same as in the 891 amino acid protein. Expression constructs for the point mutations E143K (c.427G>A [p.Glu143Lys]), A269V (c.806>T [p.Ala269Val]) and G709E (c.2126G>A [p.Gly709Glu]) were created, using HA tagged pAMC1 as template. Primers are listed in Supplementary Table S1. Expression constructs were confirmed by DNA sequencing (Institute of Biomolecular Design, Department of Biochemistry, University of Alberta, Edmonton, AB, Canada). 

Tissue Culture

HEK293 cells were grown in complete DMEM (cDMEM, DMEM supplemented with 5% [vol/vol] FBS, 5% [vol/vol] CS, and 1% [vol/vol] penicillin-streptomycin-glutamine) at 37°C in an air/5% CO₂ environment. HEK293 cells were transiently transfected, using the calcium phosphate method.⁵¹ All experiments using transiently transfected cells were carried out 48 hours post transfection. For experiments requiring treatment of cells, compounds made up in cDMEM and dimethyl sulfoxide (DMSO) solution were added 16 to 24 hours before cell harvesting. To make stably transfected cell lines, HEK293 cells were transfected and grown in cDMEM/G (cDMEM, containing 0.75 mg/mL geneticin). Cell lines were monoclonally selected. 

Cell Lysis

HEK293 cells were solubilized in immunoprecipitation buffer (IPB; 1% [vol/vol] IGEPAL CA-630, 5 mM EDTA, 150 mM NaCl, 0.5% sodium deoxycholate [wt/vol], 10 mM Tris-HCl, pH 7.5; Roche Applied Science, Laval, QC, Canada), containing complete, EDTA-free, protease inhibitors (Roche Applied Science) and incubated for 20 minutes on ice. Samples were centrifuged at 13,400g for 10 minutes and the resulting supernatant was stored at −20°C. 

Poly-L-Lysine Coating of Culture Dishes and Coverslips

Ninety-six well plates or 25-mm round glass coverslips (Thermo Fisher Scientific) in a 100-mm dish were coated with poly-L-lysine, using sterile solutions in a cell-culture hood. NaOH (150 μL per well of 5 M) was added to 96-well plates, which were incubated for 15 minutes. NaOH was then removed and wells were washed with H₂O. Ethanol (150 μL per well of 95%) was added for 5 minutes. Ethanol was removed and wells were rinsed with H₂O. Wells were washed twice with PBS (140 mM NaCl, 3 mM KCl, 6.5 mM Na₂HPO₄, 1.5 mM KH₂PO₄, pH 7.4) and 100 μL per well of 1 mg/mL poly-L-lysine in PBS was added for 15 minutes. Plates with poly-L-lysine solution remaining in wells were left overnight under ultraviolet light. Remaining solution was removed the next day and the plates were stored at 22°C. Before use, plates were rinsed with PBS to remove dried salt. 

Amplex Red High Throughput Assay

This assay is a modified version of one previously described.^49,50 Stably transfected cells were grown in T75 flasks to a confluence of approximately 90%. Growth medium was removed and the cells were rinsed with PBS. Cells were removed from the flask using 5 mL of 0.5 mM EDTA in PBS per flask and transferred to a 15-mL conical tube. Cells were centrifuged at 1000g for 5 minutes. Supernatant was removed and the cell pellet resuspended in 5 mL cDMEM/G. Cells were counted with a hemocytometer and plated onto a poly-L-lysine coated 96 well plate at a cell density of 2.2 × 10⁵ cells per well in 150 μL with cDMEM/G/well and the cells were placed in a 37°C incubator for 3 hours to allow the cells to settle on the plate. After 3 hours, cells were treated with compounds for screening and returned to the 37°C incubator for 18 to 24 hours. Concentrated stocks of compounds for testing were added to each well of the plate in 50 μL cDMEM/G, containing 0.4 or 0.8% (vol/vol) DMSO, to yield appropriate final compound concentrations in 0.1 or 0.2% DMSO. 

Using a multichannel pipetter, cells were rinsed with 200 μL/well of PBSCM (PBS, containing 0.1 mM CaCl₂ and 1 mM MgCl₂). Cells were then fixed with 10% formalin in phosphate buffer for 10 minutes at 22°C and then quenched with 150 μL/well of 50 mM NH₄Cl in PBS for 10 minutes at 22°C. Cells were blocked with 150 μL/well of PBS-B (1% [wt/vol] BSA in PBS) for 5 minutes at 22°C and then incubated with monoclonal mouse anti-HA antibody (16B12) at 1:1000 dilution in PBS-B for 1 hour at 22°C. Cells were rinsed three times with 200 μL/well PBS and incubated with sheep anti-mouse IgG conjugated to horseradish peroxidase (NXA931) at 1:1500 dilution in PBS-B for 1 hour at 22°C. Cells were rinsed three times with 200 μL/well PBS. Amplex UltraRed stock solution was made to 10 mM in DMSO and frozen at −20°C. Amplex Red working solution (200 μL/well 50 μM Amplex UltraRed, 0.0068% H₂O₂ in PBS) was added. Plates were placed on ice and covered with aluminum foil for 10 minutes after which, 150 μL of the resulting solution was collected and dispensed into black opaque 96 well plates (Greiner CELLSTAR; Sigma-Aldrich). After a 5-second shaking step, fluorescence of the solution was read in a SynergyMX Plate Reader (BioTek, Winooski, VT, USA) with excitation wavelength 530 nm and emission wavelength 590 nm. 

Immunoblotting

Cell lysates were prepared in ×2 SDS-PAGE sample buffer (10% [vol/vol] glycerol, 2% [wt/vol] SDS, 0.5% [wt/vol] bromophenol blue, 75 mM Tris, pH 6.8). Before electrophoresis, lysates were made to 1% (vol/vol) 2-mercaptoethanol and heated for 5 minutes at 65°C. Samples were resolved by SDS-PAGE on 7.5% (wt/vol) acrylamide gels.⁵² Proteins were electrotransferred onto Immobilon-P PVDF membranes (Millipore). Mouse anti-HA, or mouse anti-GAPDH were used at 1:2000 or 1:4000 dilution, respectively, in TBS-TM (5% skim milk powder in TBS-T: 0.1% [vol/vol] Tween-20, 0.15 M NaCl, 50 mM Tris, pH 7.5). After incubation for one hour at 22°C with sheep anti-mouse HRP-conjugated secondary antibody at 1:4000 dilution in TBSTM, immunoblots were developed, using Luminata TM Crescendo Western HRP Substrate chemiluminescence reagent and visualized, using an ImageQuant LAS 4000 (GE Healthcare Life Sciences, purchased from Cedarlane Corporation, Burlington, ON, Canada). Densitometry was performed, using ImageQuant TL 1D software, v8.1 (GE Healthcare Life Sciences). 

Cell Surface Biotinylation Assay

Cells were rinsed with PBS and washed with 4°C borate buffer (154 mM NaCl, 7.2 mM KCl, 1.8 mM CaCl₂, 10 mM boric acid, pH 9.0). Cells were then incubated with Sulpho-NHS-SS biotin (0.5 mg/mL in borate buffer) for 30 minutes on ice. Cells were incubated with quenching buffer (192 mM glycine, 25 mM Tris, pH 8.3) three times for 5 minutes. Cells were solubilized in 500 μL IPB buffer, containing protease inhibitors for 20 minutes on ice. Samples were centrifuged at 13,400g for 10 minutes at 22°C. Supernatant was recovered and split into two equal fractions. One fraction was reserved for later SDS-PAGE analysis (total protein, T). Immobilized streptavidin-Sepharose (100 μL of 50% suspension) was added to the other fraction of lysate and placed on a rotator and left to incubate at 4°C overnight. After incubation, the sample was centrifuged at 9800g for 2 minutes and the supernatant collected (unbound protein, U). U and T fractions were subjected to SDS-PAGE, immunoblotting, and densitometry as described. Most membranes proteins, including SLC4A11, are very vulnerable to aggregation when heated above 65°C. As a consequence, we avoid the process of elution from streptavidin Sepharose by measuring the difference between total and unbound. The formula (U-T)/T × 100% was used to determine percentage of protein biotinylated. 

Confocal Immunofluorescence

HEK293 cells, grown on poly-L-lysine coated 18-mm circular glass cover slips placed at the bottom of a 100-mm tissue culture dish, were transiently transfected with cDNA encoding SLC4A11 WT (N-terminally HA- tagged or HA 564-SLC4A11), indicated mutants (in HA 564-SLC4A11 background), or pcDNA 3.1(−). Glafenine treatment was carried out 24 hours post transfection, as described earlier. Forty-eight hours later, cells were washed in PBS, fixed with 4% paraformaldehyde in PBS for 10 minutes, followed by a quenching step with 100 mM glycine in PBS for 5 minutes at 20°C. Where indicated, cells were permeabilized with 0.2% Triton X-100 in PBS (15 minutes at 20°C). After three 5 minutes washes with PBS, nonspecific sites were blocked with 1% BSA in PBS for 30 minutes. Cells were incubated with polyclonal rabbit anti-HA (1:500; sc-805; Santa Cruz Biotechnology) or monoclonal mouse anti-GAPDH (1:500; sc-47724; Santa Cruz Biotechnology) for 1 hour at 20°C, washed with PBS (three times, 5 minutes) and, thereafter, incubated with donkey anti-rabbit IgG conjugated with Alexa Fluor 488 (1:500; 711-545-152; Jackson Immuno Research Laboratories, Inc., West Grove, PA, USA) or chicken anti-mouse IgG conjugated with Alexa 594 (1:500; A21201; Invitrogen) for 1 hour at 20°C. Final staining with DAPI was carried in the dark, followed by a washing step with PBS. Coverslips were mounted on a glass microscope slide, using Dako fluorescence mounting medium (Mississauga, ON, Canada), containing antifading agent and then visualized on Zeiss LSM 510 laser scanning confocal microscope, using 60× objective lens (Toronto, ON, Canada). 

Osmotically-Driven Water Flux Assay

HEK293 cells were grown on poly-L-lysine-coated 25-mm round glass coverslips and cotransfected with cDNA encoding enhanced green fluorescent protein (GFP; peGFP-C1 vector; Clontech, purchased from Cedarlane Corporation) and pcDNA 3.1 (empty vector) or the indicated SLC4A11 plasmid constructs in a 1:8 molar ratio. Twenty-four hours post transfection, cells were treated with 800 μL 67.5 μM glafenine in 2.7% DMSO (vol/vol), resulting in a final concentration of 5 μM glafenine in 0.2% DMSO (vol/vol). Untreated cells received DMSO to a final concentration of 0.2%. Forty-eight hours later, coverslips were mounted in a 35-mm diameter Attofluor Cell Chamber (Molecular Probes, Ottawa, ON, Canada) and washed with PBS. During experiments, the chamber was perfused with isotonic MBSS buffer (90 mM NaCl, 5.4 mM KCl, 0.4 mM MgCl₂, 0.4 mM MgSO₄, 3.3 mM NaHCO₃, 2 mM CaCl₂, 5.5 mM glucose, 100 mM D-mannitol, and 10 mM HEPES, pH7.4, 300 mOsm/kg) and then with hypotonic (200 mOsm/kg) MOPS buffered saline solution (MBSS) buffer, pH 7.4 (same composition as previous but lacking D-mannitol). Solution osmolarity was measured using an osmometer (Advance Instruments, Inc., Norwood, MA, USA). The chamber was mounted on the stage of a Wave FX spinning disc confocal microscope (Quorum Technologies, Guelph, ON, Canada), with a Yokogawa CSU10 scanning head (Tokyo, Japan). The microscope has a motorized XY stage with Piezo Focus Drive (MS-4000 XYZ Automated Stage; ASI, Eugene, OR, USA) and a live cell environment chamber (Chamlide, Seoul, Korea), set to 24°C during the duration of the experiment. Acquisition was performed with a Hamamatsu C9100–13 Digital Camera (EM-CCD; Chamlide) and a ×20 objective during excitation with a laser (Spectral Applied Research, Richmond Hill, ON, Canada) at 491 nm. Green fluorescent protein fluorescence, collected though a dichroic cube (Quorum Technologies) at wavelengths 520 to 540 nm, was acquired at 1 point s⁻¹ for 4 minutes. Quantitative image analysis was performed by selecting a region of interest for each HEK293 cell with Volocity 6.0 software (PerkinElmer, Waltham, MA, USA). Following the switch to hypotonic MBSS buffer, the rate of fluorescence change was determined from the initial 15 seconds of linear fluorescence change. 

Statistical Analysis

Values are expressed as mean ± standard error of measurement. Statistical analyses were performed, using Prism software (Graphpad v5; San Diego, CA, USA). Groups were compared with 1-way ANOVA and unpaired t-test with P < 0.05 considered significant. 

Results

Amplex Red High Throughput Assay of Cell Surface SLC4A11 Abundance

A chemical therapy to treat corneal endothelial dystrophies associated with missense mutations of SLC4A11 requires rescue of the ER-retained, mutant SLC4A11 protein to the plasma membrane. Accurate, high throughput measurement of the level of SLC4A11 at the cell surface is needed to identify suitable chemical correctors. Existing techniques such as cell surface biotinylation⁵³ can measure the level of protein at the plasma membrane, but are time and material intensive. We developed a high throughput method to identify HA-tagged SLC4A11 at the plasma membrane, which can be applied to any epitope-tagged plasma membrane protein. This assay was inspired by an assay to measure movement of CFTR-F508del to the cell surface.⁵⁰ 

The first step was to insert an epitope tag into an extracellular position, allowing detection of protein at the cell surface in intact cells. Positions for epitope insertion were selected on the basis of proteolytic accessibility of the extracellular loops.⁵⁴ Double hemagglutinin epitope (HA) tags were inserted into extracellular loop 3, following amino acid R530 or P564 (Fig. 1A). 

Figure 1

 

SLC4A11 topology model and characterization of epitope-tagged SLC4A11. (A) SLC4A11 topology model established by homology to AE1 (SLC4A1), in situ limited proteolysis and immunolocalization of the C-terminus.⁵⁸ Mutations associated with FECD (blue), CHED2 (red), and HS (orange) are highlighted. Positions of extracellular double HA epitope tag and cytosolic N-terminal HA tags are indicated by arrows at P564 and the N-terminus, respectively. Structures at amino acids 545 and 553 mark glycosylation sites. (B) Osmotically driven water influx activity of HEK293 cells transiently cotransfected with cDNA encoding GFP and WT SLC4A11 and SLC4A11 with double HA tag insertion following H530, or H564, or vector. Green fluorescent protein fluorescence was monitored in a region of interest in cells subjected to a shift to hypo-osmotic medium. The initial rate of fluorescence decrease following shift to hypo-osmotic medium (a surrogate for cell swelling) was measured as the rate of cell swelling. Data were corrected for the rate in vector transfected cells (41% ± 3% relative to noncorrected WT) and normalized to WT SLC4A11. Data represent the mean ± SEM water flux of three to six independent experiments of 10 to 20 cells per coverslip. N.S., not significant. Unpaired t-test results: WT versus HA 530, P = 0.31; WT versus HA 564, P = 0.24; HA 530 versus HA 564, P = 0.80.

To assess functional activity of the epitope insertion mutants, HEK293 cells were cotransfected with cDNA encoding GFP and cDNA encoding SLC4A11 constructs. Cells were perfused with iso-osmotic medium and then hypo-osmotic medium. Green fluorescent protein fluorescence in a small region of interest in each cell was monitored as a measure of cytosolic GFP concentration.⁴⁴ The rate of fluorescence change, which is a surrogate for cell volume change, was measured and corrected for the rate found in vector-transfected cells (Fig. 1B).⁴⁴ Water fluxes of cells expressing WT and the two extracellular loop 3 HA-tagged SLC4A11 variants were not significantly different (Fig. 1B, Supplementary Fig. S1). These data reveal that the insertion of a double HA-tag in the extracellular loop does not alter functional activity of the two mutants, indicating that their folding is not compromised by epitope tag insertion (WT versus HA530; P = 0.31, WT versus HA564; P = 0.24). Because functional activity was not significantly different in the HA-epitope insertion mutants, we arbitrarily chose to perform subsequent experiments with HA564-SLC4A11. 

To assess the cellular localization and HA-epitope accessibility of HA-tagged SLC4A11, we performed confocal immunofluorescence microscopy of transfected HEK293 cells. Cells were either permeabilized with Triton X-100, or left intact. N-terminally HA-tagged SLC4A11 (N-HA-WT) showed a strong signal in permeabilized, but not in nonpermeabilized cells (Fig. 2). This indicates that the SLC4A11 N-terminus and its epitope tag are cytosolic and inaccessible to antibodies unless cells are permeabilized. In contrast, cells expressing HA 564-SLC4A11, a protein with an HA epitope insertion in extracellular loop 3 (Fig. 1A), showed strong signal with and without cell permeabilization. This indicates that the HA-epitope is readily accessible to extracellular antibodies in intact cells. Under permeabilized conditions the level of signal is not appreciably higher than without permeabilization, indicating that HA564-SLC4A11 predominantly targets to the cell surface. Finally, as a control for specificity of the immunolocalization data, untransfected HEK293 cells were probed with anti-HA and anti-GAPDH antibodies (Supplementary Fig. S2). No signal for anti-HA antibody was observed, whereas GAPDH signal was seen only following cell permeabilization. Thus, the anti-HA antibody specifically detects HA-tagged SLC4A11 and the permeabilization protocol is effective in revealing cytosolic protein. Together these data indicate that HA564-SLC4A11 provides an appropriate background in which to study SLC4A11 trafficking, using an anti-HA antibody. 

Figure 2

 

Localization of SLC4A11 variants by confocal immunofluorescence microscopy. Confocal microscopy images of HEK293 cells expressing SLC4A11 tagged with HA at the cytosolic N-terminus (N-HA), or externally double HA tagged (HA 564). SLC4A11 was either WT, or indicated mutants. Nonpermeabilized (left) and permeabilized (right) cells were probed for SLC4A11, using polyclonal rabbit anti-HA antibody. Donkey anti-rabbit IgG conjugated with Alexa Fluor 488 (green) was used as secondary antibody, while nuclei were counterstained with DAPI (blue). Scale bars: 17 μm. Exposure time and gain were constant for all images.

HEK293 cell lines, stably expressing HA-tagged SLC4A11 variants, were established for the amplex red high throughput assay. The first had an N-terminal HA tag (on the cytoplasmic domain), allowing it to be used as a control for detection of accessible cytosolic epitopes (Fig. 1A). The second had the double HA-tag following H564, to detect cell surface protein. The remaining three cell lines express the SLC4A11 disease alleles A269V (CHED), E143K (CHED), and G709E (FECD),^9,12,14 along with a double HA tag after H564. A269V was chosen because culture of this mutant in cells at 30°C induced partial rescue to the cell surface.⁵⁴ E143K was chosen because dimerization with monomers that traffic to the plasma membrane resulted in a heterodimer at the plasma membrane.²⁷ These strategies suggest that these proteins have the potential to be cell-surface rescued by small molecules. G709E was chosen as a representative mutation for FECD. 

In this high throughput assay (HTA) cells stably expressing SLC4A11 variants are cultured in 96-well dishes. Incubation with anti-HA antibody allows binding of cell surface HA epitopes. The primary antibody is detected by a secondary antibody, conjugated to horse radish peroxidase (HRP). Amplex red reagent in the presence of H₂O₂ is cleaved by HRP to produce resorufin, a highly fluorescent red compound (Supplementary Fig. S2).⁵⁵ Red fluorescence is then quantified on a 96-well plate fluorescence reader. 

We first assessed whether the HTA can differentiate between cell surface and ER-retained SLC4A11. The maximum resorufin fluorescence was found for cells expressing WT SLC4A11 with extracellular HA tag (HA 564), a value that was used to normalize the fluorescence level for the other variants (Fig. 3A). Cells expressing N-terminally HA-tagged WT SLC4A11 had a fluorescence signal of 19% ± 3% of the externally HA-tagged WT (Fig. 3A). Because the HA tag of the N-terminally tagged WT is not present at the cell surface, this represents nonspecific background of the assay. Cells expressing A269V SLC4A11 had fluorescence 62% ± 13% relative to WT. A269V SLC4A11 partially traffics to the plasma membrane at levels approximately 50% of WT.⁵⁴ Cell expressing E143K and G709E SLC4A11 both had fluorescence 24% ± 3% of WT level, which is indistinguishable from N-terminal HA tag WT, consistent with earlier reports that these mutants are nearly completely ER-retained.^9,26,54 

Figure 3

 

Comparison of cell surface SLC4A11 level as measured by high throughput assay (HTA) and cell surface biotinylation. (A) In the HTA protocol, HEK293 cells stably expressing the indicated SLC4A11 variants (with an N-terminal cytosolic HA tag (N-HA), or with an external double HA tag [HA 564]) and processed through the HTA. Cells were incubated with amplex red reagent, which in the presence of HRP/H₂O₂, is converted to the fluorescent red product, resorufin. Red fluorescence in each well was quantified on a 96-well plate fluorescence reader. Data were normalized to HA 564 WT. *Significant difference (P < 0.05). Data represent the mean ± SEM of fluorescence relative to HA 564 WT (n = 8 wells). Dashed line indicates the fluorescence level of cells expressing N-terminally tagged SLC4A11, representing the background level of the assay. (B) Cell surface biotinylation assay of cells stably expressing the indicated SLC4A11 variants. HEK293 cells were incubated with membrane-impermeant biotinylating reagent. Cell lysates were incubated with streptavidin-Sepharose to remove the biotinylated proteins. The fraction of biotinylated protein, representing cell surface protein, was quantified. The level of biotinylation of GAPDH (6% ± 11%), representing the background of the assay, was subtracted from all values. Data represent the mean ± SEM of % total protein biotinylated (n = 3).

To assess the validity of the cell surface processing measurements determined by the HTA, we also quantified the level of cell surface protein, using cell surface biotinylation. In these assays, cells were treated with membrane-impermeant biotinylating reagent, and the fraction of biotinylated protein (extracellular-accessible) was determined (Supplementary Fig. S4, Fig. 3B). Cytosolic GAPDH served as a negative control for the assay and its level of biotinylation was subtracted from the values for SLC4A11 variants. Relative to externally tagged WT SLC4A11, 190% ± 2%, 54% ± 3%, 0% ± 8%, and 0% ± 7% protein was biotinylated for N-terminal tagged WT, A269V, E143K, and G709E SLC4A11, respectively. 

To compare the measurements of surface SLC4A11 determined by the HTA and cell surface biotinylation, background values were subtracted from the data in each assay. That is, in the HTA (Fig. 3A) fluorescence observed for cells expressing N-terminally HA-tagged WT SLC4A11 was subtracted and in cell surface biotinylation (Fig. 2B), the level of biotinylation of cytosolic GAPDH was subtracted. Relative cell surface processing values (Figs. 3A, 3B) were compared for the two techniques. In each case (each of the extracellular tagged WT, A269V, E143K, and G709E SLC4A11) the relative cell surface level obtained for each variant was not significantly different when data from the two methods were compared (not shown). SLC4A11 in the N-terminal HA-tagged WT stable cell line had significantly higher biotinylation compared with HA 564 WT SLC4A11, suggesting that the extracellular tag affected processing efficiency (Fig. 2B). 

As an additional way to assess the validity of the amplex red high throughput assay data, we performed confocal immunofluorescence of SLC4A11 mutants, tagged with HA564. Confocal immunofluorescence detected G709E-SLC4A11 only after cell permeabilization (Fig. 2). This indicates that the protein does not process to the cell surface, but rather is retained intracellularly. In contrast, pericellular (plasma membrane) A269V-SLC4A11 was detected in nonpermeabilized cells, with additional intracellular signal upon cell permeabilization (Fig. 2). This indicates that A269V partially processes to the cell surface. Interestingly, the confocal immunofluorescence data align well with results from the amplex red high throughput assay and cell surface biotinylation (Figs. 3A, 3B). 

Together these results indicate that the amplex red HTA reliably quantifies the fraction of SLC4A11 at the cell surface. 

Small-Scale Screen of Folding Corrector Drugs

We next examined the ability of the HTA to identify chemical correctors of mutant SLC4A11 misfolding. We thus carried out a small screen, focused on compounds effective in rescue of ER-retained CFTR mutants.^56–64 Concentrations of each of the compounds screened were selected on the basis of the effective concentrations for CFTR rescue. We chose to test acetylsalicylic acid because it is a nonsteroidal anti-inflammatory drug (NSAID) and other NSAIDs rescue some ER-retained proteins to the plasma membrane.^56,60,64 To emulate a large scale screen, compounds were initially screened with four replicates on a single plate (Fig. 4). Because this was a small-scale screen, compounds were added manually, using a multichannel pipetter, whereas a large screen would make use of robotic plating. A269V, E143K, and G709E mutant stable cells were treated with each of 11 different small molecules, and with solvent carrier (0.1% or 0.2% DMSO; Fig. 3). Dimethyl sulfoxide was used at two concentrations because high concentrations of DMSO act to assist folding of some misfolded proteins.⁶⁵ We wanted to determine whether there might be a synergistic effect of low concentration DMSO and other small molecules. DMSO at 0.1% and 0.2% did not significantly affect cell surface levels of SLC4A11 mutants, as assessed by the HTA (Figs. 4A, 4B). Similarly, DMSO had no effect on surface abundance of WT-SLC4A11 at either 0.1 or 0.2% (Supplementary Fig. S5). 

Figure 4

 

A small-scale screen of potential SLC4A11 folding corrector compounds. In the HTA assay protocol, cells stably expressing the indicated SLC4A11 mutants were treated with compounds at indicated concentrations for a period of 18 to 24 hours. Assays were carried out in the presence of (A) 0.1% DMSO (except TMAO, 4-PBA, and glycerol, which were screened in the absence of DMSO) or (B) 0.2% DMSO. Red fluorescence arising from amplex red conversion to resorufin, was measured on a 96-well plate reader. Data represent the mean ± SEM of fluorescence relative to the level found for each untreated variant, (n = 4). *Significant difference (P < 0.05) by 2-tailed unpaired t-tests against the relevant control. ASA, acetylsalicylic acid; Carb, carbamazepine; Geld, geldanamycin.

In this initial screen, results for most compounds clustered around 100% of untreated cell surface expression (Fig. 3). Two compounds, 4-phenylbutyric acid and glycerol, reduced fluorescent signal developed in the HTA, possibly because of cytotoxic effects. These compounds were not subjected to further analysis. Three compounds, glafenine, acetylsalicylic acid, and ibuprofen, stood out as leading to mutant SLC4A11 cell surface abundance approximately 1.5-fold higher than in untreated cells. Reflecting the limitations of this HTA, these compounds did not show significant difference from untreated controls for each mutant when subjected to screening on a single plate. These three compounds were subjected to additional intensive screening to determine whether their effect on cell surface processing was significant (Fig. 5). 

Figure 5

 

High-repetition screening of compounds identified in preliminary screen. High throughput assay was used to assess the level of SLC4A11 cell surface abundance in the presence of (A) 0.1% or (B) 0.2% DMSO. Data represent the mean ± SEM of fluorescence relative to the level found for each untreated variant, (n = 12). *Significant difference (P < 0.05) by t-tests against the relevant control.

As measured by the HTA, glafenine (10 μM) in the presence of 0.2% DMSO significantly increased cell surface abundance of all three mutants tested, compared with 0.2% DMSO treatment (Fig. 5B). Ibuprofen (10 μM) in 0.2% DMSO had a significant effect on cell surface abundance only of A269V and E143K SLC4A11. Acetylsalicylic acid (10 μM) in 0.2% DMSO only had a significant effect on A269V SLC4A11 cell surface level. Most compounds screened had no significant effect on HTA-measured cell surface abundance for any of the mutants (Figs. 4, 5). P values of compound treatments leading to significantly different mean fluorescence are listed in Supplementary Tables S2 and S3. These results show that the HTA discriminates between compounds that affect the cell surface level of SLC4A11 mutants from those that do not. Because glafenine treatment gives rise to an increase in relative fluorescence for all three of the mutants, we subjected glafenine to additional tests. 

Dose-Response to Glafenine of G709E-SLC4A11 Trafficking

To further probe the effect of glafenine on mutant SLC4A11, we established the relationship for the compound's effects. We tested G709E SLC4A11 as this mutant is profoundly affected, showing nearly complete ER-retention (Fig. 3). Stable cells expressing the G709E mutant were treated with 0-20 μM glafenine, and the amount of fluorescence produced in the HTA was measured. The dose-response curve displayed classical saturating kinetics, revealing that G709E-SLC4A11 was rescued with an EC₅₀ of 1.5 ± 0.7 μM (Fig. 6). 

Figure 6

 

Correction of G709E-SLC4A11 trafficking by glafenine. Cells stably expressing G709E-SLC411 were subjected to the HTA protocol on 96-well plates in the presence of indicated concentrations of glafenine and 0.2% DMSO. Red fluorescence arising from amplex red conversion to resorufin, was measured in each well. Fluorescence values of cells treated with 0 μM glafenine were subtracted from each value and data were normalized to the maximum red fluorescence observed. Data represent the mean ± SEM fluorescence (n = 8). EC₅₀ was calculated to be 1.5 ± 0.7 μM glafenine.

Effect of Glafenine on SLC4A11 Cell Surface Trafficking

To assess further the effect of glafenine on mutant SLC4A11, we measured cell surface processing, using the cell surface biotinylation assay (Fig. 7, Supplementary Tables S6). For this experiment, we used transiently transfected cells rather than stable cell lines, to provide an assessment in a different cell context. Because the effect of glafenine reaches a maximum at approximately 5 μM (Fig. 6), we used this as a fixed concentration in subsequent investigations. Transfected cells were treated with or without 5 μM glafenine with 0.2% DMSO 24 hours prior to biotinylation. Data were corrected for the background level of biotinylation, represented by the level of cytosolic GAPDH biotinylation. For all three mutants, a significant difference of biotinylation was found between the untreated and treated cohorts (Fig. 7, Supplementary Tables S6). Interestingly, 5 μM glafenine did not significantly alter WT SLC4A11 cell-surface processing. Additionally, E143K and G709E SLC4A11 had respective cell-surface processing of 58% ± 11% and 46% ± 14% relative to WT SLC4A11. This is similar to the cell-surface processing values determined in the HTA in the presence of 10 μM glafenine (39% ± 3% and 34% ± 2%, respectively relative to WT; Fig. 5). These data confirm that glafenine in the presence of 0.2% DMSO increased the level of cell surface processing of three ER-retained SLC4A11 mutants. 

Figure 7

 

Effect of glafenine on mutant SLC4A11 cell surface processing efficiency. Cells were transiently transfected with vector or cDNA encoding the indicated SLC4A11 type. Twenty-four hours post transfection, cells were treated with 0.2% DMSO (−) or 5 μM glafenine and 0.2% DMSO (+). Cells were subjected to cell surface biotinylation assays 48 hours post transfection. Samples have been corrected for GAPDH biotinylation which represents the background of the assay. *Significant difference (P < 0.05, n = 3–4).

To test further the effect of glafenine on mutant SLC4A11 cell-surface trafficking, we performed confocal immunofluorescence microscopy (Fig. 8). SLC4A11 mutants E143K, A269V and G709E were all tagged with the HA564 extracellular epitope tag and expressed in HEK293 cells. Without glafenine treatment, SLC4A11 G709E was not found at the cell surface, whereas some A269V and E143K were at the cell surface. Without glafenine, all three mutants revealed additional intracellular-retained SLC4A11 upon permeabilization. The effect of glafenine was striking in comparing nonpermeabilized cells with or without glafenine. Glafenine treatment led to a marked increase in cell surface SLC4A11 for cells expressing E143K and G709E. The effect was less pronounced for A269V, which traffics significantly to the cell surface without glafenine. These confocal immunofluorescence data indicate that glafenine increases cell surface trafficking of intracellular-retained SLC4A11 mutants. 

Figure 8

 

Confocal immunofluorescence microscopic assessment of the effect of glafenine on cell surface trafficking of SLC4A11 mutants. HEK293 cells were transiently transfected with cDNA encoding externally double HA-tagged (HA 564) SLC4A11 mutants (indicated) in a 100-mm dish, containing 18-mm round coverslips. Cells were treated with 5 μM glafenine in 0.2% DMSO 24 hours post transfection, or were untreated. Cells were either permeabilized with Triton X-100, or not, as indicated. Cells were probed for SLC4A11, using polyclonal rabbit anti-HA. Donkey anti-rabbit IgG conjugated with Alexa Fluor 488 (green) was secondary antibody, while nuclei were counterstained with DAPI (blue). Scale bars: 17 μm. Exposure time and gain were constant for all images.

Glafenine-Rescued Mutant SLC4A11 Retains Transport Function

Glafenine rescued SLC4A11 to the plasma membrane, but the results have little therapeutic significance unless the rescued protein is functional. We thus assessed the effect of glafenine on osmotically-driven water flux in HEK293 cells expressing the three mutants (Fig. 9, Supplementary Tables S7). A269V, E143K, and G709E SLC4A11 mutants without glafenine treatment showed a low water flux activity, 16% ± 1%, 13% ± 1%, and 12% ± 1%, relative to WT SLC4A11 respectively. Glafenine treatment induced a significant increase in water flux activity for all three mutants (67% ± 8%, 87% ± 6%, and 114% ± 11%, relative to WT SLC4A11, respectively). Glafenine did not affect water flux in cells expressing WT SLC4A11, consistent with the findings of the cell surface biotinylation assay (Fig. 7), where glafenine treatment did not significantly affect WT SLC4A11. Furthermore, cells expressing the mutants had their water flux increase to a level not significantly different to WT SLC4A11. These data indicate that SLC4A11 mutants retain functional activity and, if rescued to the plasma membrane, are as effective as WT SLC4A11 in terms of functional activity. Glafenine treatment may represent a feasible pharmacologic therapeutic for some mutations of SLC4A11. 

Figure 9

 

Effect of glafenine on osmotically-driven water flux by ER-retained mutant SLC4A11. HEK293 cells were transiently cotransfected with cDNA encoding external HA 564 epitope-tagged SLC4A11, either WT, A269V, E143K, G709E, or vector, along with GFP cDNA. The level of green florescence was quantified in regions of interest in cells as medium was changed from iso-osmotic to hypo-osmotic. The rate of fluorescence change upon switching to hypo-osmotic medium was measured as a surrogate for the rate of cell swelling. Data were corrected for rates observed in vector transfected cells and normalized to WT SLC4A11. Data represent the mean ± SEM of three to five independent experiments of 10 to 20 cells per coverslip. *Significant difference in water flux (P < 0.05). No significant difference compared with WT SLC4A11 without glafenine treatment.

Discussion

The goal of this study was to develop a high throughput assay (HTA) able to screen for compounds that increase the processing of ER-retained SLC4A11 mutants to the cell surface. The assay, working in a 96-well format, provided reliable measurements of surface abundance of SLC4A11. An initial small screen with potential correcting molecules identified glafenine as increasing the level of mutant SLC4A11 at the plasma membrane. Independent assessment by cell surface biotinylation revealed that glafenine does increase mutant processing to the cell surface. Glafenine increased the water flux activity of cells expressing SLC4A11 mutants to levels found in cells expressing WT-SLC4A11. The efficacy of glafenine suggests that corrector drugs with potential efficacy in treating corneal dystrophies caused by SLC4A11 mutants can be identified, using the HTA described here. 

The data indicate that the HTA provides reliable data, reflecting the relative cell surface abundance of SLC4A11 variants. Cell surface biotinylation data correlated well with HTA data, indicating that the HTA represents the relative cell surface abundance of SLC4A11 variants. We found a similar protein biotinylation pattern across the mutant types compared with the relative fluorescence from the HTA. Cells that express an extracellularly inaccessible, cytosolic HA tag (N-Terminal HA Tag WT), or that express extracellularly tagged HA-SLC4A11 that is ER-retained (E143K and G709E) produce approximately 20% fluorescence relative to the external HA Tag WT. This background fluorescence may arise from non-HRP catalyzed amplex red conversion, or reflect some degree of cell lysis, allowing antibody access to intracellular HA-epitopes. Finally, confocal immunofluorescence data provided additional support for the validity of the HTA: (1) intracellularly-HA-tagged SLC4A11 was inaccessible to anti-HA antibody, (2) extracellularly-tagged SLC4A11 (HA 564) was present at the plasma membrane, and (3) the level of immunofluorescence at the surface of cells correlated well with the degree of surface trafficking found in the HTA. 

We were concerned that the epitope insertion required for the assay might affect SLC4A11 folding. Transport function was unaffected by the epitope insertion, as measured in water flux assays in transiently-transfected cells indicating no effect on cell-surface processing and folding. In contrast, cell surface biotinylation of stably-transfected cells suggested a reduced cell-surface processing of the H564 HA insertion mutant, relative to the N-terminally tagged protein. While this suggests that the epitope insertion had an effect, we were reassured that cell surface biotinylation of H564 HA tagged proteins (Fig. 2B) was not different from that reported for untagged versions.^26,54 

Other high throughput assays to detect protein at the cell surface have been developed.^66–69 Although there are other methods to measure protein abundance at the plasma membrane, including cell surface biotinylation, flow cytometry, or confocal microscopy based immunofluorescence, these are relatively time and material intensive. We developed a simpler assay, including elements of an assay developed to assess F508del CFTR cell-surface processing.^49,50 We modified that assay, moving from a 24 well format to 96 wells, enabling higher throughput. To promote cell adhesion during the assay we introduced three modifications: (1) culture to almost 100% confluency, (2) coating plates with poly-L-lysine to promote cell adhesion, and (3) cell fixation, using formalin. We also found that deplating the stable cell lines with trypsin adversely affected our results because of the proteolytic sites located on the extracellular loop where the HA epitope is located.⁵⁴ Instead, we deplated cells, using 0.5 mM EDTA in PBS. Finally, we used a 1% BSA in PBS instead of DMEM-HEPES in immune steps of the protocol. Together these modifications provide an assay amenable to high throughput screening of SLC4A11 folding correctors. 

Assay capacity is a significant issue as we look ahead to screens of larger compound libraries. We found that at least four replicates in a single dish were needed to produce a reliable result. In a 96-well format, this could translate to 22 unique conditions per plate, with the positive and negative controls also performed in four replicates. Although we could detect a difference in fluorescence for cells treated with compounds in four replicates, further replicates were needed to lessen standard error to produce statistically significant results. The assay could be further refined by use of larger format (e.g., 128 or 256 well) plates. One major limitation of the HTA is the skill and reproducibility of the person performing the assay. This limitation could be overcome with robotic automation, especially if larger format plates are used. 

Currently, the only permanent treatment strategy for corneal endothelial dystrophies is corneal transplant. Corneal transplant is complicated by tissue availability, graft rejection, and the high amount of skill and resources required to perform the transplant. Results from the HTA, cell surface biotinylation and confocal immunofluorescence were consistent in revealing that glafenine with 0.2% DMSO is a pharmacologic corrector for the trafficking defect of SLC4A11 A269V, E143K, and G709E. Ibuprofen and acetylsalicylic acid also corrected surface targeting of some SLC4A11 mutants. It remains to be determined if other mutations of SLC4A11 will respond in a similar fashion to these drugs. 

We decided to further characterize glafenine as it was the only compound effective in rescuing all three SLC4A11 mutants. Glafenine is NSAID, used as an analgesic mainly in Europe in the 1970s. Its use was discontinued because it was implicated in anaphylaxis, gastrointestinal injury and kidney failure.^70,71 Glafenine rescued the F508del mutation of CFTR and the G601S mutation of the hERG K⁺ channel to the plasma membrane.^56,60 While the mechanism of rescue by glafenine is unclear, most NSAIDs act by inhibition of one or both of the cyclooxygenase (COX) isoforms.^72,73 Both COX-1 and COX-2 are integral membrane proteins,⁷⁴ but the role of COX inhibition in the rescue of mutant membrane proteins is unclear. We determined that the EC₅₀ of glafenine in SLC4A11 rescue was 1.5 μM (Fig. 5), but could not find literature identifying the EC₅₀ of glafenine as a COX inhibitor. Clinically, glafenine can achieve plasma concentrations up to 10 μM,^60,75 suggesting that the effective concentration we found is in the range that targets COX. 

The initial HTA screen of compounds included the NSAIDs acetylsalicylic acid and ibuprofen. When used with 0.2% DMSO, acetylsalicylic acid had a significant effect on A269V SLC4A11 (Fig. 4). Ibuprofen with 0.2% DMSO had a significant effect on A269V and E143K mutants (Fig. 4). Ibuprofen is effective in rescuing F508del CFTR and is proposed to do so through inhibition of COX-1.⁶⁴ Knockdown of COX-1 in HEK293 cells expressing F508del CFTR rescued the mutant protein, but COX-2 knockdown did not.⁶⁴ Rescue of mutant CFTR, and now SLC4A11, with the use of NSAIDs suggests that prostaglandin synthesis may play a role in ER-retention of mutant membrane proteins. Further probing of this mechanism could lead to a novel treatment strategy for diseases involving ER-retained membrane proteins. 

Despite rescue of mutant SLC4A11 to the plasma membrane, these results are only clinically relevant if the mutant proteins remain functional once at the plasma membrane. Earlier investigations found that to delay onset of symptoms associated with corneal dystrophies, 27% of WT SLC4A11 function is required and to completely avoid symptoms, 50% of WT SLC4A11 function is required.²⁷ Our in vitro investigations revealed that the 50% benchmark can be exceeded for all three SLC4A11 mutants treated with glafenine (Fig. 7). Treatment with glafenine may, thus, represent an effective pharmacologic therapy. 

This study leads to three major conclusions. First, the HTA we developed is an effective and reliable method to determine the abundance of SLC4A11 at the plasma membrane. The approach could be modified for application to identify correctors of other misfolded membrane proteins. Second, glafenine in 0.2% (vol/vol) DMSO corrects the trafficking defect for A269V, E143K, and G709E mutations of SLC4A11. Third, mutant SLC4A11 remains functional once localized to the plasma membrane and this functionality may suffice to prevent symptoms of corneal dystrophies. Here, we presented a paradigm for the high throughput identification and in vitro testing of potential SLC4A11 folding correctors. These data suggest that glafenine is a potential pharmacologic therapeutic for some corneal dystrophy-causing mutants of SLC4A11. 

Acknowledgments

The authors thank Anita Quon for preliminary work on the HTA. They also thank Darpan Malhotra for helpful comments on the manuscript. 

Supported by an operating grant from the Canadian Institutes of Health Research (Ottawa, ON, Canada). Supported by summer studentship awards from Alberta Innovates Health Solutions and Women's and Children's Health Research Institute (JM; Edmonton, AB, Canada). Supported by a graduate studentship and postdoctoral fellowship, respectively, from the International Research Training Group in Membrane Biology funded by the National Sciences and Engineering Research Council (SL and KA; Ottawa, ON, Canada). 

Disclosure: A.M. Chiu, None; J.J. Mandziuk, None; S.K. Loganathan, None; K. Alka, None; J.R. Casey, None