Molecular Vision 2013; 19:1096-1106
Received 23 January 2013 | Accepted 22 May 2013 | Published 24 May 2013
1Singapore Eye Research Institute, Singapore; 2Department of Ophthalmology, National University Health System & National University of Singapore, Singapore; 3Singapore National Eye Centre, Singapore; 4Neuroscience and Behavioral Disorders (NBD) Program, Duke-NUS Graduate Medical School, Singapore
Correspondence to: Eranga N. Vithana, Singapore Eye Research Institute, 11 Third Hospital Avenue, Singapore, 168751; Phone: +65 63224542; FAX: (+65) 6322 4599; email: email@example.com.
Purpose: To characterize the relative expression levels of all the solute carrier 4 (Slc4) transporter family members (Slc4a1–Slc4a11) in murine corneal endothelium using real-time quantitative (qPCR), to identify further important members besides Slc4a11 and Slc4a4, and to explore how close to the baseline levels the gene expressions remain after cells have been subjected to expansion and culture.
Methods: Descemet’s membrane-endothelial layers of 8–10-week-old C57BL6 mice were stripped from corneas and used for both primary cell culture and direct RNA extraction. Total RNA (from uncultured cells as well as cultured cells at passages 2 and 7) was reverse transcribed, and the cDNA was used for real time qPCR using specific primers for all the Slc4 family members. The geNorm method was applied to determine the most stable housekeeping genes and normalization factor, which was calculated from multiple housekeeping genes for more accurate and robust quantification.
Results: qPCR analyses revealed that all Slc4 bicarbonate transporter family members were expressed in mouse corneal endothelium. Slc4a11 showed the highest expression, which was approximately three times higher than that of Slc4a4 (3.4±0.3; p=0.004). All Slc4 genes were also expressed in cultured cells, and interestingly, the expression of Slc4a11 in cultured cells was significantly reduced by approximately 20-fold (0.05±0.001; p=0.000001) in early passage and by approximately sevenfold (0.14±0.002; p=0.000002) in late passage cells.
Conclusions: Given the known involvement of SLC4A4 and SLC4A11 in corneal dystrophies, we speculate that the other two highly expressed genes in the uncultured corneal endothelium, SLC4A2 and SLC4A7, are worthy of being considered as potential candidate genes for corneal endothelial diseases. Moreover, as cell culture can affect expression levels of Slc4 genes, caution and careful design of experiments are necessary when undertaking studies of Slc4-mediated ion transport in cultured cells.
Bicarbonate ions have been implicated to play a central role in human corneal endothelial ion pump to maintain corneal transparency . A substantial proportion of cellular HCO3– transport is mediated by proteins belonging to the solute carrier 4 (SLC4) family. The SLC4 family of transporters consists mainly of three functional groups: 1) electroneutral Na+-independent Cl–/HCO3-exchangers, 2) electrogenic or electroneutral Na+:HCO3– cotransporters, and 3) electroneutral Na+-driven Cl–/HCO3– exchangers [2,3]. SLC4A11, the most divergent member of this family, has been described to function as an electrogenic Na+-coupled borate co-transporter and as an electrogenic Na+/OH– co-transporter in the absence of borate .
Several members of the SLC4 gene family have been linked to ocular and corneal diseases in humans. Mutations in SLC4A4 cause proximal renal tubular acidosis as well as ocular anomalies, such as glaucoma , cataracts, and band keratopathy . Mice lacking sodium bicarbonate cotransporter (NBC3; Slc4a7) develop blindness and auditory impairment due to the degeneration of sensory receptors in the eye and inner ear . Moreover, SLC4A11 was identified to be responsible for three corneal endothelial dystrophies, recessive congenital hereditary endothelial dystrophy (CHED2) , late onset Fuchs endothelial dystrophy (FECD) , and corneal dystrophy with perceptive deafness (Harboyan syndrome) .
Despite the importance of HCO3– transport in the normal functioning of the cornea and the involvement of some of its members in corneal dystrophies, the entire SLC4 gene family has not been systematically characterized in the cornea. The expression of SLC4A4, which encodes the electrogenic sodium bicarbonate cotransporter NBC1, has been localized to the basolateral membrane of corneal endothelial cells [11-13]. Gene expression of SLC4A11 has been shown in the human corneal endothelium via SAGE and microarray  as well as reverse transcription PCR . Gene expression information is lacking for the other members of the SLC4 family.
Therefore, the purpose of this study was to characterize the expression levels of the entire Slc4 family of genes relative to those of Slc4a4 and Slc4a11 in the mouse corneal endothelium to identify further SLC4 members that can serve as candidate genes for analysis in corneal dystrophies. The expressional alterations that occur for Slc4 genes due to cell culturing procedures involving both early and late subcultures were also explored.
C57BL6 wild-type mice were ordered from the animal holding unit of the National University of Singapore and housed and bred in the Singhealth Experimental Medicine Center until a sufficient number for the study was obtained. Approval was sought from the SingHealth International Animal Care and Use Committee (IACUC), and all procedures were in accordance with the Association for Research in Vision and Ophthalmology resolution for the use of animals in research.
PCR primers were designed for all members of Slc4 gene family (Slc4a1 to Slc4a11) as well as housekeeping genes (HKGs; glyceraldehyde-3-phosphate dehydrogenase[Gapdh], 18s ribosomal RNA [18s rRNA], beta-actin [Actb], hypoxanthine phosphoribosyltransferase 1 [Hprt1]) and cell identifying marker genes (Aquaporin 1 [Aqp1], Zona occludens 1 [Zo 1], Collagen group VIII a2 [Col8a2], Collagen group I a1 [Col1a1]). The primers for the target genes were designed based on the mouse mRNA sequence using Primer 3 primers design software . Forward and reverse primers were designed to be located on separate exons (with a large intron in between) to ensure that the template used would be cDNA rather than genomic DNA. Each primer sequence was queried against the mouse DNA NCBI database, using BLAST to ensure that primer sequences were specific for the target mRNA transcript. The primers were also designed such that they were in a region that detects all known splice variants of the corresponding transcript. The primers were synthesized by AIT Biotech (Singapore). The optimized primer sequences used in this study are listed in the Table 1.
Adult 8–10-week-old C57BL6 wild-type mice (n=5 for the primary cell culture and n=10 for direct RNA extraction) were sacrificed with an overdose of sodium pentobarbital. Eyes were enucleated, and the globes were rinsed with sterile 1X PBS to remove traces of blood and other material. The corneas were then dissected from the globe and laid endothelial side up in a sterile Petri dish. Descemet’s membrane (DM)–endothelial layers were stripped under a dissecting microscope and incubated overnight to stabilize the cells in Opti-MEM I® (Invitrogen, Carlsbad, CA) medium. Endothelial cells were separated from the DM by 2 mg/ml collagenase A (Sigma, St. Louis, MO) treatment at 37 °C for 2–3 h in minimum essential media (MEM, Invitrogen) supplemented with 15% fetal bovine serum (FBS) and 20 μg/ml gentamicin. After detachment from DMs, the mouse corneal endothelial cells (MCECs) were subjected to direct RNA extraction or culture.
MCECs were washed with Dulbecco’s Modified Eagle’s Medium (DMEM, Invitrogen) supplemented with 0.1 mg/ml gentamicin and 1.25 μg/ml amphotericin B before culturing in MEM medium supplemented with 15% FBS and 20 μg/ml gentamicin in a humidified atmosphere with 5% CO2 at 37 °C. The cells were passaged on reaching 80% confluence. The MCECs at passage 2 and passage 7 were used for subsequent RNA extraction. The entire experiment was done three times independently to obtain three separate batches of cells at passages 2 and 7.
Total RNA from mouse uncultured corneal endothelial cells as well as cultured cells at passages 2 and 7 were isolated by the TRIZOL™ (Invitrogen) method following the manufacturer’s protocol with a few modifications. Briefly, the stripped Descemet’s membranes or the harvested cells were homogenized and lysed in the TRIZOL™ reagent (Invitrogen). Then 0.2 ml chloroform was added to each 1 ml of TRIZOL™ reagent, and centrifuged at 10,000 ×g for 20 min at 4˚C. The aqueous layer was mixed with 0.5 ml isopropanol and incubated overnight. The reaction was centrifuged at 10,000 ×g for 20 min at 4˚C, isopropanol was removed and mixed with 1 ml of cold 75% ethanol. The RNA wash with ethanol was done twice by centrifugation at 7500 ×g for 4 min at 4˚C and the resulting pellet was dissolved in RNase free water. Genomic DNA was removed by digestion with DNase I (AmpGrade; Invitrogen-Gibco) according to manufacturer’s protocol. Genomic DNA was removed by digestion with DNase I (AmpGrade; Invitrogen-Gibco). The dissolved RNA sample was measured on a spectrophotometer (Nanodrop 2000; Thermo Fischer Scientific, Waltham, MA) to determine the concentration and quality before proceeding to convert to cDNA. One microgram of total RNA was reverse transcribed with random hexamers by using SuperScript III™ first-strand synthesis system for reverse transcription (RT)-PCR (Invitrogen). A traditional PCR amplification was performed in a 50-μl reaction volume using GoTaq® DNA Polymerase (Promega, Madison, WI), an equal amount of first-strand cDNA template, and the optimized primers in a thermal cycler GeneAmp® PCR System 9700 (Applied Biosystems, Carlsbad, CA). We applied the following PCR parameters: 95 °C for 3 min, followed by 40 cycles of 95 °C for 30 s, 58 °C for 30 s, and 73 °C for 90 s. The resulting PCR products were separated by 2% agarose gel electrophoresis. The gel was then viewed under ultraviolet illumination, and the images were taken by a Hamamatsu image detection system (Hamamatsu, Japan).
Cells were seeded on coverslips until approximately 80% confluent. The cells were fixed in 4% paraformaldehyde for 30 min at 4 °C, incubated separately with primary anti-Na+K+ ATPase (200 µg/ml; Santa Cruz Biotechnology, Santa Cruz, CA) and anti-Zo-1antibodies (5 µg/ml; Life tech, Carlsbad, CA) at recommended dilutions (1:100 for Na+K+ ATPase and 1:50 for Zo-1) for 2 h at room temperature, washed three times with 1X PBS, and then incubated with Alexa Fluor® 488-conjugated secondary antibody (Invitrogen) for 1 h. Nuclear DNA was visualized by staining with DAPI. The cells were mounted on mounting medium (VECTORSHEILD, Vector Laboratories, Burlingame, CA) and then examined using a fluorescent microscope with an ApoTome attachment (Axio Imager Z1; Zeiss, Stuttgart, Germany).
The most stable HKG was selected by using the geNorm™ VBA applet for Microsoft Excel version 3.5 (Biogazelle, Gent, Belgium) according to the manufacturer’s protocol. Briefly, the expression data matrix with raw data of relative quantities of each HKG was loaded to calculate the M values, which measure gene expression stability, and the gene expression normalization factor for each sample based on the geometric mean. The detailed underlying principles and calculations were described by Vandesompele et al. .
qPCR with SYBR® Green I dye for detection was performed in a 10-μl mixture containing 5 μl of SYBR® green PCR master mix (Applied Bisosystems), 500 nM of each primer, and 1 μl of cDNA template, using the LightCycler® 480 (Roche, Basel, Switzerland). The threshold cycles (Ct) were calculated using the LightCycler® software version 1.5 (Roche). The cycling parameters used for amplifications of targets were as follows: initial denaturation for 10 min at 95 °C, followed by 45 cycles of 30 s denaturation at 95 °C, 30 s annealing at 58 °C, and 45 s extension at 72 °C. The expression levels of Slc4 bicarbonate transporters were normalized against that of the most stable HKG Hprt1, as shown by the geNorm™ analysis. Relative quantification using the comparative Ct method was used to analyze the data output. Values were expressed as the fold change over corresponding values for the control by the 2–ΔΔCt method .
All the data used for statistical analyses were obtained from three independent experiments. Differences between the groups was determined by the two-tailed Student t test using spreadsheet software (Excel 5.0; Microsoft, Redmond, WA), with significance determined at p<0.05.
In order for the ΔΔCt method to be valid, the amplification efficiencies for both the tested gene and the HKG must be optimally equal to 2 (100% efficiency). Therefore, we validated the amplification efficiencies of all primers of the Slc4 family as well as the HKGs by real-time qPCR, using the mouse kidney cDNA as the standard template. Mouse kidney cDNA was used for this purpose since all Slc4 genes were shown to be expressed in mouse kidney tissue  and sufficient quantities of RNA could be extracted from kidney tissue. Six serial dilutions of the template were subject to real-time qPCR to investigate the amplification efficiency of the primers. All the primers showed similar amplification efficiencies in the serially diluted samples of mouse kidney cDNA (Table 2). To ensure a single product was obtained without primer dimers, we also performed a melting curve analysis to verify that a single specific melting peak was observed for each primer pair.
The standard PCR with Taq DNA polymerase using the mouse corneal endothelial cDNA template and primers that were selected specifically to target the mRNA of the Slc4 genes revealed a single RT-PCR product of the predicted size for a given mRNA template for each primer pair. This indicated that there was no contamination by genomic DNA. The four HKGs (Gapdh, Actb, Hprt1, 18s rRNA) were used as amplification (positive) and normalizing controls. This analysis revealed that all Slc4 genes were expressed in murine corneal endothelium (Figure 1A). It also indicated Slc4a11 and Slc4a2 genes had the highest expression levels, while Slc4a1 and Slc4a9 appeared to have the lowest expression levels in uncultured corneal endothelium. However, RT-PCR is semiquantitative and not as sensitive (or always accurate) as real-time qPCR for the assessment of relative gene expression.
Further analysis by real-time qPCR using the most stable HKG, Hprt1 (Figure 2) revealed that among the Slc4 genes, Slc4a11 had the highest expression level while Slc4a1 showed the lowest (Table 3). The order of expression level was therefore Slc4a11 followed by Slc4a2, Slc4a4, Slc4a7, Slc4a3, Slc4a10, Slc4a5, Slc4a9, Slc4a8, and Slc4a1. The expression of Slc4a2 was approximately half that of Slc4a11 (0.55±0.03; p=0.007), while the expression of Slc4a1 was 300 times less than Slc4a11 (0.0026±0.0006; p=0.003). When comparing the expression levels of the two clinically important genes known for corneal diseases (i.e., Slc4a11 and Slc4a4), the expression of Slc4a11 was approximately three times greater than that of Slc4a4 (3.4±0.3; p=0.004).
The MCECs initially showed stellate morphology at low densities, and upon confluence they became polygonal in shape, characteristic of endothelial cells (Figure 3C,D). Cultures were split (i.e., passaged) upon reaching 80% confluence, and passaging was performed until the seventh passage for this study. In the latter passages, some cells were more elongated in appearance, while some appeared larger in size with more prominent nuclei (Figure 3E,F).
To characterize the MCECs, immunofluorescent analyses were performed on the cultured passage 2 MCECs with antibodies for Na+K+ATPase and Zo-1. Na+K+ATPase is involved in physiologic maintenance of corneal thickness by the corneal endothelium [20,21], while Zo-1 is a selective semipermeable tight junction-associated protein in the corneal endothelium [22,23]. The cultured MCECs expressed both Na+K+ ATPase and Zo-1 (Figure 4A,B). We further characterized the cells by examining the expressions of several genes normally present in corneal endothelial cells (Aqp1, Zo-1, Col8a2) as well as the fibroblastic marker gene Col1a1 by RT- PCR (Figure 4D,E). MCECs at both early and late passages expressed the endothelial markers Aqp1, Zo-1, and Col8a2. The fibroblast cell marker Col1a1 was not expressed by early passage 2 cells but was expressed by late passage 7 cells. RT-PCR also revealed the expression of all Slc4 genes in the cultured MCECs (Figure 1B,C).
Real-time qPCR using the normalization factor calculated from Hprt1, Actb, and 18s rRNA (Table 4) indicated that the expressions of all the Slc4 transporters, except Slc4a4 and Slc4a10, were downregulated in early passage 2 (Table 5). However, in late passage 7, Slc4a1, Slc4a2, Slc4a3, Slc4a4, and Slc4a7 were upregulated. The expression of Slc4a11 was significantly reduced by 20-fold (0.05±0.001; p=0.000001) in early passage but only by sevenfold (0.14±0.002; p=0.000002) in late passage. Interestingly, the expression of another clinically important gene, Slc4a4, was significantly upregulated by approximately 2.5-fold (2.52±0.07; p=0.0007) in passage 2 and 14-fold (14.57±0.16; p=0.00005) in late passage 7 (Figure 5).
Previous studies have investigated SLC4 family expression in other types of tissue  but not in ocular tissues. We found that all members of the Slc4 bicarbonate transporter family are expressed in MCECs, and some genes, such as Slc4a4, Slc4a2, and Slc4a7, are more highly expressed than others. Although SLC4A4 and Slc4a7 have been reported to be associated with ocular conditions in humans [5,6] and mice , respectively, a documented link between SLC4A2 and an eye disease is currently lacking. These genes can, therefore, serve as potential candidate genes to be analyzed for corneal endothelial diseases, such as FECD. We observed a similar result for humans, with these genes being among the highest expressed in corneal endothelial cells (unpublished data). Slc4a11 showed the highest expression in uncultured CECs, indicating that it plays a pivotal role in transporting solutes in the corneal endothelium, although we did not establish a functional correlation with its expression. The exact function of SLC4A11 in the corneal endothelium is unknown. However, the congenital corneal opacity seen in recessive CHED cases as well as the severe morphological alterations displayed by Slc4a11 mutant mice attests to the importance of SLC4A11 in maintaining normal corneal endothelial function .
The SLC4 family, except sodium-coupled borate cotransporter (SLC4A11), is functionally divided into three main groups [26,27]: anion exchangers (AEs), sodium bicarbonate cotransporters (NBCs), and sodium-driven Cl–/HCO3– exchangers (NDCBE). We found that the expression of genes encoding the two NBCs (Slc4a4 and Slc4a7) to be most highly expressed after Slc4a11, indicating that the NBCs are the main bicarbonate transporters in corneal endothelium. Our data indicated that for NBCs, NBCe1 (Slc4a4) is the primary member in the corneal endothelium and NBCn1 (Slc4a7) is secondary (Table 6). Similarly for AEs, AE2 (Slc4a2) appears to be primary and AE3 (Slc4a3) secondary. For the NDCBE family of proteins, AE1 (Slc4a1) is primary and NDCBE (Slc4a8) secondary. The redundancy seen with this family of genes is perhaps indicative of the important role that these genes play in ion transport within the corneal endothelium. It will also be interesting to explore the compensatory role played by the secondary gene in the event of a defect, i.e., mutation, involving the primary member of each class. Understanding the hierarchy of expression of bicarbonate transporters within the same functional group has also opened up the interesting possibility of compensatory therapeutics.
In this study, we found that Slc4 gene expression was significantly altered during cell culture. Interestingly, the expression of Slc4a11 gene was reduced by sevenfold in late passages while that of another clinically important gene, Slc4a4, was significantly up-regulated by approximately 2.5-fold in early passages and 14-fold in late passages. The morphology of endothelial cells was also less “endothelial like” in late passages. The environmental conditions of the cultured cells, for example the high salt content of the culture media, could be a key factor for the altered expression of solute transporters seen in cultured cells compared to native tissue. Another possible explanation for the observed alterations in gene expression is the epithelial/endothelial-to-mesenchymal transition (E/EnMT) during which endothelial cells lose endothelial markers and obtain mesenchymal markers, as suggested by the expression of Col1a1 in late passage cells in our study (see Figure 4E). In vitro studies have shown that endothelial cells can undergo EMT, which is speculated to depend on transforming growth factor ß1 [28,29]. The drastically altered expression levels of the Slc4 genes coincident with an altered cellular morphology indicate that further study should be undertaken to explore the possible link between SLC4 gene expression and EMT. More importantly, the effects of cell culture on the expression levels of Slc4 genes highlight the need for caution and the careful design of experiments with adequate controls when studying ion transport activity of these proteins in cultured cells.
The main limiting factor in this study is that Slc4 gene expression was only tested at the RNA level and not at the protein level. Ideally, the expression levels should have been confirmed by using specific antibodies to the various Slc4 genes through western analysis of protein extracts. The lack of commercially available antibodies for murine Slc4 members that are sensitive or specific for immunoblot analysis precluded further confirmation of our results at the protein level.
To the best of our knowledge, this is the first study to investigate transcript expression levels of the entire Slc4 bicarbonate transporter family in mouse corneal endothelial cells. We could establish expression profiles for each member in primary corneal endothelium as well as quantify the expressional alterations that occur for Slc4 genes due to our cell culturing procedure that involved both early and late subcultures. Identification and characterization of more genes causative of corneal endothelial diseases would further our understanding of pathologic mechanisms underlying this group of disorders and may lead to novel ways to treat these conditions.
This work was supported by grants from the following funding bodies of Singapore (NMRC grant NMRC/EDG/0018/2008, BMRC grant 07/1/35/19/520 and SHF/FG383P/2007). The authors would like to thank Dr Seet Li Fong for critical review of the manuscript.