Inhibition of miR-205 Impairs the Wound-Healing Process in Human Corneal Epithelial Cells by Targeting KIR4.1 (KCNJ10)

Abstract

Purpose.: The aim of the study was to test the hypotheses that injury stimulates the expression of miR-205, which in turn inhibits KCNJ10 channels by targeting its 3′ UTR, thereby facilitating the wound-healing process in human corneal epithelial cells (HCECs).

Methods.: A stem-loop qRT-PCR was used to examine the miR-205 expression. BrdU cell proliferation assay and wound scratch assay were applied to measure the effect of miR-205 mimic or antagomer in HCECs. The patch-clamp technique, dual luciferase reporter assay, and Western blot analysis were employed to test whether miR-205 regulates KCNJ10, one of the target genes of miR-205. Both of the primary human and mouse corneal epithelial cells (pH/MCECs) were employed to further confirm the observations obtained in HCECs.

Results.: The scratch injury in pH/MCECs increased the expression of miR-205 and decreased the expression of KCNJ10 within 24 hours. The notion that miR-205 may target KCNJ10 was supported by dual luciferase reporter assay showing an inhibition effect of miR-205 on 3′ UTR of KCNJ10. Application of miR-205 antagomer significantly delayed the regrowth in wounded HCECs. However, inhibition of KCNJ10 partially abolished the effect from miR-205 antagomer and restored the healing process. Moreover, overexpression miR-205 antagomer enhanced the protein expression of KCNJ10 but not KCNJ16. In addition, patch-clamp demonstrated that inhibition of endogenous miR-205 expression increased Ba²⁺-sensitive inwardly rectifying K⁺ channels. In addition, an electrophysiological study of pHCECs showed the presence of KCNJ10-like 20 pS K⁺ channels and scratch injury significantly decreased the Ba²⁺-sensitive inwardly rectifying K⁺ currents.

Conclusions.: miR-205 stimulates wound healing by inhibiting its target gene KCNJ10.

Introduction

Human corneal epithelial cells (HCECs) compose the outermost five to seven layers of stratified epithelial cells covering the cornea. ¹ They function as the protective barrier for the corneal stroma, preventing microbial invasion, stromal desiccation, and permitting the proper refraction of light; they have a remarkable regrowth capability after injury. Wound repair occurs in distinct stages, which include migration, activation, proliferation, differentiation, and remodeling. ^2,3 Dysfunction of regeneration and repair would lead to vision loss due to scarring and ulceration. Multiple signaling pathways and diverse growth factors and cytokines have been shown to mediate the continuous wound-healing stages. ⁴ For instance, the epidermal growth factor (EGF) and EGF-related growth factors, TGFα, bFGF, heparin-binding EGF-like growth factor, among others, could stimulate the healing through the phospho-EGF receptor/mitogen activated protein kinase pathway ^5–10 in the early migration and the following proliferation stage in wounded HCECs. 

It has previously been demonstrated that the electric field (EF) substantively regulates corneal cell migration and proliferation, showing that the potassium (K⁺) current is inhibited to prevent loss of intracellular K⁺ during the regrowth after injury in HCECs. ^11–13 These studies reflected the crucial role of K⁺ channels in manipulating membrane potentials of HCECs, thereby leading to activation of downstream signaling by inducing Ca²⁺ influx. However, many questions still remain regarding the details of the mechanisms involved in the regulation of K⁺ channels in the wounded HCECs. Inwardly rectifying potassium channel KIR4.1 (KCNJ10), a member of the KIR families, plays an essential role in regulating cell membrane potential. ^14,15 The dysfunction or genetic mutation of this gene leads to the SeSAME syndrome, a central nervous system disorder characterized by deafness, ataxia, mental retardation, and renal loss of K⁺. ^16,17 The expression of KCNJ10 in retinal glia cells has been well documented. ¹⁸ However, the physiological function of KCNJ10 in the unexcitable corneal epithelial cells remains unknown. 

MicroRNAs are a class of potent small, noncoding RNA molecules that, mostly, negatively regulate gene expression in a posttranscriptional manner. ^19,20 They play important roles in regulating widespread physiological processes in mammalian cells. ^21,22 Comprehensive investigations have been conducted to define the correlation between the metastasis of tumor epithelial cells and the expression of miR-205. ^23–25 For instance, miR-205 has been shown to regulate the migration and proliferation of esophageal cancer cells. ²⁶ In wounded corneal cells, miR-205 has been reported to effect migration by targeting the phosphatase SHIP2. ²⁷ The aim of the present study was to explore the role of miR-205 in regulating the healing process in HCECs. Specifically, we hypothesize that miR-205 stimulates the healing process by inhibiting inwardly rectifying K⁺ channel member KIR4.1 (KCNJ10). 

Materials and Methods

Cell Line and Cell Culture

An immortalized human corneal epithelial cell line (HCEC) was a gift from Michal L. Schwartzman (Department of Pharmacology, New York Medical College, Valhalla, NY). ² The cells were subpassaged by TrypLE (Life Technologies, Grand Island, NY) digestion, and the disassociated cells were suspended and spun down at 500g for 2 minutes, the pelleted cells were resuspended with complete KGM-2 BulletKit (catalog number CC-3107; basal medium +SingleQuots Kit containing growth factors, cytokines, and supplements; LONZA, Allendale, NJ), and seeded in coating matrix FNC (AthenaES, Baltimore, MD) precoated T-75-cm flasks (BD Biosciences, San Jose, CA) and cultured in a humidified incubator at 37°C with 5% CO₂, the medium was replaced with fresh KGM-2 medium every 2 days. 

The primary mouse corneal epithelial cells were established by explant culture method according to many skilled experts reported before. ^28,29 Briefly, C57/BL6 mice (The Jackson Laboratory, Bar Harbor, ME), aged 4 to 5 weeks, were used in accordance with the guidelines in the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research. After mice were killed, the eyeballs were enucleated by the pathogen-free disposable punch (2.5 mm), and washed with Ca²⁺-free PBS (pH 7.2) containing 1X antibiotic-antimycotic (Life Technologies). Each corneal button was further delicately cut into half and free from the other attached tissue, for example, iris and sclera. Each half of the corneal button was placed in collagen-type I precoated 6-well plate (Life Technologies) with the epithelial side up. After approximately 10 to 15 minutes in the laminar air-flow culture hood, the half corneal button securely attached to the bottom of the plate, 1.5 mL of low-calcium, low-bovine pituitary extract (BPE), serum-free progenitor cell targeted medium CnT-50 (CELLnTEC, Bern, Switzerland) containing 1X antibiotic-antimycotic (Life Technologies) was added and cultured in a humidified incubator at 37°C with 5% CO₂; the fresh medium was replaced every 2 days. After 7 to 10 days, nearly 100% confluence of the cells formed, the half corneal button could be placed in a new cell culture plate to start a new explant culture as described above. The primary cells were digested in stem-pro accutase (Life Technologies) with gentle pipette up-down releasing, centrifuged at 500g for 5 minutes, and subcultured at CnT-50 containing 100 ng/mL cholera toxin (LIST Biological Laboratories, Inc., Campbell, CA), 1X antibiotic-antimycotic (Life Technologies). 

The primary human corneal epithelial cells (pHCECs) were purchased from Life Technologies, (epithelial cells isolated from dissected limbal regions of the eye), and subcultured in as the protocol for mouse primary corneal epithelial cells above. 

The primary human keratinocytes (pHKCs) were purchased from Zen-Bio, Inc. (Research Triangle Park, NC). pHKCs were subcultured as the same protocol as pHCECs. All the primary cell lines used for the experiments were within five passages. 

In Vitro Scratch Wound-Healing Migration Assay

HCECs (approximately 3 × 10⁶ cells) were seeded into an FNC (AthenaES) precoated 12-well plate 24 hours prior to treatments. When a monolayer of HCECs was formed (cell confluence reached to 90% or above), fresh medium containing 10 μg/mL Mitomycin-C was added. Two crossing linear scrape injuries (bared area) were generated with a sterile Fisher 10-μL fine tip (Fisher Scientific, Pittsburgh, PA), the width of each linear injury area was approximately 0.1 cm and the time immediately following scratching was considered as the start point for each experiment; thereafter, at certain time points, the images were acquired with an inverted microscope Axiovert25 (Carl Zeiss, Thornwood, NY) equipped with a camera. The healing progress was measured by the average decreasing percentage of these bared areas from each well at certain time points. 

BrdU Cell Proliferation Assay (BrdU Incorporation ELISA Method)

BrdU cell proliferation assay kit was purchase from Cell Signaling (Boston, MA). Briefly, the cells (5000 cells) transfected with vector control or miR-205 mimic or antagomer were plated in a 96-well plate and incubated with BrdU for 6 hours in an incubator. Medium was removed and replaced with 100 μL/well fixing/denaturing solution, after incubating at room temperature (RT) for 30 minutes, 100 μL/well-prepared detection antibody was added, the plate was placed at RT for 1 hour, washed with 1X wash buffer for three times, 100 μL/well-prepared 1X horseradish peroxidase (HRP)-conjugated secondary antibody solution was added, and kept at RT for 30 minutes. The buffer was completely removed and the plate was washed with 1X wash buffer for three times, and 100 μL/TMB substrate solution was added. The plate was incubated at RT for 30 minutes, 100 μL STOP solution was added, and absorbance (OD) was read by a microplate reader (Biotek Synergy, Winooski, VT) at 450 nm. 

DNA Constructs Used in the Experiments and Lentiviral Particles Production

The vectors pZip-miR-205 and pCDH were purchased from SBI Biosciences (Mountain View, CA) for the silencing or overexpression of miR-205. ³⁰ For the specific overexpression of miR-205, human pri-miR-205 was amplified from genomic DNA by PCR and inserted into the pCDH with EcoRI and BamHI sites; the primer sequence used, sense: ACGAATTCCGAGAAACAGTCAAGAGTC, antisense: ACTGGATCCCCAATCTGCCCATCACC, underlined nucleotides reflected the enzyme cutting sites used for cloning, and all the constructs were sequenced for confirmation in Genewiz (South Plainfield, NJ). For siRNA of KCNJ10, the annealed double strands of oligos were ordered from IDT (Coralville, IA), sense: 5′ GCUGAGACCAUUCGUUUCAdTdT; antisense: 5′ UGAAACGAAUGGUCUCAGCdTdT. 

The antagomers and mimics of miR-205 were purchased from Qiagen (Valencia, CA); these were used for further validation of the results from the above vector-based expression. For the production of lentiviral particles, 90% confluent cells of the package cell line 293T were plated in 75-cm cell culture T-flasks 24 hours prior to the day of transfection. For each T-75-cm flask, 10 mg of gene-expressing plasmid DNA, 6 μg of pSPAX2, and 3 μg of pMD2 (packaging plasmids) were mixed with 45 μL of Turbofect transfection reagent (Fisher Scientific) in 1 mL of opti-MEM (Life Technologies), incubated for 15 minutes at RT, and gently applied into the flask; incubated at 37°C with 5% CO_2, replaced with the fresh complete Dulbecco's modified Eagle's medium after 24 hours while collecting the virus particle–containing medium at 48 and 72 hours after the transfection. To purify the lentiviral particles, the collected medium from above was spun at 500g for 2 minutes, the pellet was discarded, and the supernatant was filtered with a sterile 0.45-μm filter unit. Poly-L-lysine was combined with the final concentration 0.005% (wt/vol) and incubated at 4°C for at least 3 hours to stabilize the virus particles. The lenti-particles were then pelleted at 10,000g for 2 hours; the pellet was formed and resuspended with KGM medium and applied to the HCECs. 

RNA Extraction and RT-qPCR Quantitation of MicroRNAs

Total RNA from HCECs was extracted with an miRneasy Mini kit (Qiagen) as described before. ³¹ Briefly, 700 μL QIAzol lysis buffer was added to the monolayer HCECs, the lysate was collected into 1-mL Rnase-free microcentrifuge tubes with a rubber cell scraper, and vortexed to mix. Then, 140 μL chloroform was also added and mixed completely. The sample was then placed on a rack to stand for 2 to 3 minutes at RT, centrifuged at 8000g for 15 minutes, the upper aqueous phase was carefully transferred to a new tube, and 1.5 volumes of ethanol was added and mixed thoroughly. All the samples were loaded to a Rneasy mini spin column and spun at 8000g for 15 seconds, then washed with 700 μL RWT, followed by twice with 500 μL RPE. The RNA sample was eluted with 30 μL Rnase-free water. The quality and quantity of RNA was confirmed with a spectrophotometer (Eppendorf, Enfield, CT). 

RT reaction for microRNA was performed with modified Maxima First Strand cDNA kits (Fisher Scientific) by adding a poly A tail in the routine RT process. ^32,33 Briefly, a reaction sample was formed consisting of 4 μL of 5X Reaction Mix, 2μL of Maxima enzyme mix, 2 μL of 10X poly(A) polymerase buffer, 0.1 mM ATP, 1 μL poly(A) polymerase, 1 μM RT- primer, up to 4 μg total RNA, then using Rnase-free water to bring the final volume to 20 μL. The sample was then incubated at 25°C for 10 minutes followed by 50°C for 30 minutes, the enzyme was inactivated by incubation at 85°C for 5 minutes. RT-primer for miR-205 sequence: 5′ GTTTTTTTTTTTTTTTCAGACTCC. qPCR of microRNA was employed as previously described; 12.5 μL of 2x Brilliant II QRT-PCR SYBR Green Low ROX master mix (Agilent Technologies, Santa Clara, CA) was combined with 250 nM forward and reverse primers and 50 ng cDNA from the above RT reaction in 25 μL total volume. The cycling parameters were 95°C 15 minutes followed by 40 cycles of 95°C 30 seconds, 60°C 1 minute. A melting curve analysis was performed after the program to ensure the specific amplification for all the PCR reactions. The relative expression of specific genes was determined by the 2^−ΔΔct method. The specific forward primer for miR-205 was: 5′ CTCTTGTCCTTCATTCCACC, the reverse primer for miR-205 was the same as the RT primer used above. The specific forward primer for miR-16 was 5′ TAAGTGTTGGACGGAGAACTG, and reverse primer was 5′ GCGAGCACAGAATTAATACGAC, the RT primer was 5′ GCGAGCACAGAATTAATACGACTCACTATAGGTTTTTTTTTTTTGC, for U6 primers: forward 5′ CTCGCTTCGGCAGCACATA, reverse 5′ ATATGGAACGCTTCACGAATT. 

Semi-RT-PCR Quantitation of Keratin-3, -5, -12, -14, and qRT-PCR of KCNJ10 Expression

Total RNA from pHCECs and pHKCs were extracted using the same protocol as above. RT-reaction was performed with Maxima First Strand cDNA kits (Fisher Scientific) according to the product sheet. Briefly, a reaction was formed by mixing 4 μL 5X Reaction Mix, 2 μL Maxima enzyme mix, up to 4 μg total RNA, then using Rnase-free water to bring the final volume to 20 μL. The sample was then incubated at 25°C for 10 minutes followed by 50°C for 30 minutes, the enzyme was inactivated by incubation at 85°C for 5 minutes, the resultant cDNA was used as the template for PCR cycling, the primers used were designed by complying with intron-spanning approaches, which could exclude the possible amplification from genomic DNA contamination, ³⁴ and are listed in Supplementary Table S1. The cycling parameters were 95°C 2 minutes followed by 30 cycles of 95°C for 10 seconds, 55°C for 5 seconds, and 68°C for 1 minute. After a 10-minute holding at 68°C, the samples were kept on ice and ready for 1.2% agarose gel examining. 

Luciferase Assay

To test the binding site of miR-205 on KCNJ10, the annealed oligo containing the 3′ UTR of KCNJ10 (NM_002241: nt 2888–94) was inserted into the XhoI and NotI sites of luciferase expression vector psi-Check2 (Promega, Madison, WI); for the wild-type (wt) sequence: sense: TCGACGGAAGTGAGATAGCTGATGAAGGAGC, antisense: GGCCGCTCCTTCATCAGCTATCTCACTTCCG; for the mutant (mut), which has disrupted the seed sequence by continuous 4 “Cs”: sense: TCGACGGAAGTGAGATAGCTGATGCCCC AGC, antisense: GGCCGCTGGGGCATCAGCTATCTCACTTCCG. Underlined nucleotides indicate the seed sequence of miR-205, the italicized nucleotides show the mutated seed sequence of miR-205. All the constructs were sequenced (Genewiz) for confirmation. The dual-luciferase reporter assay (Promega) was performed in HEK293 transfected with the constructs of pCDH-miR-205 and psiCheck2-Luc-site1, which is the wt 3′ UTR sequence of KCNJ10, or psiCheck2-Luc-muSite1, in which the wt sequence has been mutated to disrupt the complementary matched binding with the seed sequence of miR-205. The relative renilla luciferase activities are normalized with firefly luciferase for equal transfection efficiency. 

The Patch-Clamp Experiments

HCECs were transfected with pZip-control, or pZip-miR-205 or pCDH-miR-205 lentiviral particles as described above. Within 24 hours after transfection, the cells were treated with TrypLE Express (Life Technologies) for 10 minutes to detach the cells. The perforated whole-cell patch-clamp experiments were carried out at room temperature. The cells were incubated with a bath solution containing 140 mM KCl, 0.5 mM MgCl₂, 1.5 mM CaCl₂, and 10 mM HEPES (pH 7.4). Borosilicate glass (1.7-mm OD) was used to make the patch-clamp pipettes that were pulled with a Narishege electrode puller (Narishege, Inc., Long Island, NY). The pipette had a resistance of 2 to 4 MΩ when filled with 140 mM KCl. The tip of the pipette was filled with pipette solution containing 140 mM KCl, 2 mM MgCl₂, 1 mM EGTA, and 5 mM HEPES (pH 7.4). The pipette was then back-filled with amphotericin B (20 μg/0.1 mL) containing pipette solution. After forming a high-resistance seal (>2 GΩ), the membrane capacitance was monitored until the whole-cell patch configuration was formed. The cell membrane capacitance was measured and compensated. The K⁺ currents were measured by an Axon 200A patch-clamp amplifier (Molecular Devices, Sunnyvale, CA). The currents were low-pass filtered at 1 KHz and digitized by an Axon interface (Digidata 1200) and data were stored in an IBM-compatible computer and were analyzed using the pClamp software system 9 (Axon; Molecular Devices). K⁺ currents were presented as picoamperes/25 picofarads. The same patch-clamp equipment described above was used for the single channel recording and the pipette solution contains (in mM) 140 KCl, 1.8 MgCl₂, 10 HEPES (pH = 7.4). The currents were low-pass filtered at 1 KHz and digitized by an Axon interface. Data were analyzed using the pClamp software system 9 (Axon; Molecular Devices). Channel activity defined as NP_o (a product of channel number and open probability) was calculated from data samples of 60 seconds' duration in the steady state as follows:  

 

where t_i is the fractional open time spent at each of the observed current levels. The channel conductance was determined by measuring the current amplitudes over several voltages.  

Western Blot Analysis

The protein sample prepared for Western blot was harvested as previously described. ³¹ Briefly, the HCECs were washed with 1x PBS once and each well in the 12-well cell culture plate was lysed with 100 μL of ice cold radioimmune precipitation assay (RiPA) buffer (25 mM Tris-Cl, pH 7.4, 150 mM NaCl, 50 mM KCl, 1% SDS, 2 mM EDTA, 0.5% glycerol, 50 mM NaF) with 1:100 (vol/vol) dilution of the proteinase inhibitor and phosphotase I and II inhibitor mixture (Sigma-Aldrich, St. Louis, MO), then briefly vortexed and placed on ice for 15 minutes, spun down at 8000g for 15 minutes to pellet the debris and keep the supernatant for the following Western blot. Then, 20 μg of total protein lysate was boiled with 400 mM DTT in 2X sample buffer for 5 minutes and loaded into SDS-PAGE gel. After running at 180 volts, the protein was transferred to the pre-wet nitrocellulose membrane with cold tank buffer (48 mM Tris, 39 mM glysine, 20% methanol, and 0.0375% SDS) by 100 volts for 1 hour. After incubation with Licor's blocking buffer (LI-COR, Lincoln, NE) for 1 hour at RT, the membrane was probed with appropriate diluted first antibody for 1 hour at RT, and washed three times with 0.01% Tween 20 in PBS and 5 minutes for each. The membrane was then probed with the 1:10,000 (wt/wt) diluted second antibody IRDye 680RD goat antirabbit IgG (LI-COR) and 800 cw goat antimouse IgG (LI-COR) for another 1 hour at RT. After that completed, the membrane was washed three times with 0.01% Tween 20 in PBS for 5 minutes, then the membrane was ready for Odyssey imaginer (LI-COR) scanning. 

Statistical Analysis and Materials

The Student's t-test was used to determine the significance between the two comparing groups. Data are presented as mean ± SD. Rabbit polyclonal antibody of KCNJ10 was purchased from EMD Millipore (Billerica, MD) and monoclonal mouse antiproliferating cell nuclear antigen (anti-PCNA) and polyclonal rabbit anti-glyceraldehyde 3-phosphate dehydrogenase (anti-GAPDH) were obtained from Cell Signaling, monoclonal mouse anti-α-Tubulin was from Rockland (Gilbertsville, PA), and rabbit anti-KCNJ16, -Ki-67 was from Abcam (Cambridge, MA). All chemicals used in the present study were obtained from Sigma-Aldrich. 

Results

Scratch Injury Stimulated miR-205 Expression in Primary Mouse Corneal Epithelial Cells

In our current study, we established the primary corneal epithelial cell line from mouse (pMCECs) corneal button explant (Fig. 1A), used the scratch wound model, and measured the miR-205, miR-16 (as negative control) expression level at 12 hours after injury in pMCECs with qRT-PCR (shown in Fig. 1B). We found that miR-205 significantly increased in wounded cells, in contrast to miR-16, another stably expressed microRNA in cornea displayed unchanged expression at 12 hours after wounding. The same and consistent observation was also obtained in pHCECs (shown in Supplementary Fig. S1). The effect of wounding on the miR-205 expression was specific because wounding had no effect on miR-16, which is also highly expressed in cornea. ^35–37 To further confirm the differentiated epithelial feature of pHCECs, we used RT-PCR to examine the expression level of the stag-differentiation related keratin-3/-12 pairs (KRT-3/-12) in pHCECs, ³⁸ the results are shown in Supplementary Figure S2. Comparing with primary keratinocytes, which exclusively displayed the expression of typical keratin-5/-14 pairs (top), pHCECs expressed terminal differentiation-related keratin-3/-12 pairs besides the keratin-5/-14 pairs ^38,39 (bottom). GAPDH expression from pHCECs and keratinocytes reflected the equivalent cDNA template loading during PCR reaction. 

Figure 1. 

 

The expression of miR-205 increased in the injured pMCECs. (A) The representative phase-contrast images were acquired with an inverted microscope Axiovert25 equipped with camera after 4 days (top) and 10 days (bottom) from each explant (Et). The typical “cobblestone”-like corneal epithelial cells started to spread out from explant at day 2 to 4 and form nearly 100% confluence at day 7 to 10. (B) A crossing scratch damage was generated in the subcultured pMCECs and total RNA was extracted, the expression of miR-205/-16 was examined by qRT-PCR, it showed that the expression of miR-205 significantly increased (P ≤ 0.001; n > 3) at 12 hours after wounding (top); however (bottom), the expression of miR-16 did not change after wounding. Student's t-test: P = not significant, n > 3.

miR-205 Stimulated Migration in Wound-Healing Process in HCECs

We next examined the role of miR-205 in regulating the wound-healing process of HCECs by expression of exogenous miR-205 mimic, miR-205 antagomer, or the empty vector (as control) in HCECs. The efficiency of miR-205 mimic or miR-205 antagomer in altering miR-205 expression is shown in Figure 2. It was apparent that miR-205 antagomer decreased its expression to one-third of the control (in Fig. 2A), while expression of miR-205 mimic increased miR-205 expression by 3-fold after 48 hours of transfection (Fig. 2B). Figure 3 and Supplementary Figure S3 show the representative images of scratch injury, demonstrating that miR-205 mimic significantly stimulated, while antagomer inhibited, the cell migration in the injured HCECs in comparison with those transfected with empty vector (as control). The results in Figure 3 are summarized in Figure 4A, showing that overexpression of miR-205 increased the cell regrowth rate from 25.0% ± 4.3% (mean ± SD; n = 4) in control to 30.0% ± 3.2% (mean ± SD; n = 4), while inhibition of endogenous miR-205 decreased the healing process to 17.0% ± 0.9% (mean ± SD; n = 4). 

Figure 2. 

 

The efficiency of the lentiviral vector based enforced expression or silencing of miR-205 was examined by qRT-PCR in HCECs. (A) qRT-PCR showed that pZip-miR-205 could downregulate the expression of miR-205 to one-third of the control after 48 hours of transfection. n = 4. t-test: P ≤ 0.001. (B) Mammalian expression vector pCDH-miR-205 could forcedly express miR-205 3-fold over empty vector control after 48 hours of transfection. n = 4. Student's t-test: P ≤ 0.001.

Figure 3. 

 

MiR-205 increased wound closure in HCECs. A crossing scratch damage was generated in HCECs transfected with scramble control (shown in [A, B]) alone or miR-205 mimic (C, D) or antagomer (E, F), with the presence of mitomycin C (10 μg/mL). The representative wounding-scratch images were acquired at 0 hour (A, C, E) or 6 hours after injury (B, D, F). n = 4.

Figure 4. 

 

Data from wound-healing assays were statistically analyzed by Student's t-test. (A) Enforced miR-205 expression vigorously augmented, while inhibition of endogenous miR-205 decreased the migration 6 hours after injury. (B) The miR-205 antagomer significantly impaired wound closure after 24-hour and 48-hour injuries. Moreover, BaCl₂ incubation partially abolished this inhibition effect from miR-205 antagomer. (C) Silencing KCNJ10 expression by siRNA modestly increased regrowth after 12-hour injury in HCECs.

miR-205 Stimulated the Wound Healing by Suppressing KCNJ10

We further tested our hypothesis that miR-205 stimulated wound healing by suppressing KCNJ10. We found that the delayed effect induced by inhibiting miR-205 was abolished in the presence of 500 μM BaCl₂ in media, an inhibitor of inwardly rectifying K⁺ channels (Fig. 4B). Inhibition of K⁺ channels attenuated the delayed recovery of wounded HCECs transfected with miR-205 antagomer (Supplementary Fig. S4), suggesting the role of K⁺ channel in facilitating the wound-healing process. This notion was further supported by experiments in which downregulation of KCNJ10 channels in wounded HCECs also increased the cell regrowth after wounding (Fig. 4C, Supplementary Fig. S5). 

miR-205 Modulated KCNJ10 Expression by Targeting Its 3′ UTR

Since downregulation of KCNJ10 facilitated cell regrowth, we speculated that miR-205 may regulate KIR.4.1 channels in HCECs. Through the analysis and prediction from bioinformatics tools, we identified a potential targeting site for miR-205 on the 3′ UTR located on mRNA nucleotide sequence from 2878 to 2895 of KCNJ10 (Fig. 5A). This prediction was confirmed by dual luciferase reporter assay. Figure 5B summarizes the results, showing that overexpression of miR-205 significantly inhibited luciferase report gene activity (renilla luciferase) when containing wt sequence, but had no effect in the cells transfected with mutated 3′ UTR of KCNJ10, in which the matched seed sequence was disrupted. We next examined whether the scratch injury altered the expression of endogenous KCNJ10 in HCECs. Figure 6A showed the protein expression of KCNJ10 in the control and wounded HCECs, respectively. It was apparent that the KCNJ10 expression in the wounded cells was significantly lower than those of the control within 24 hours (Fig. 6B), the consistent observation was also obtained in pHCECs (Supplementary Figs. S6A, S6B). To test our hypothesis that KCNJ10 expression could be modulated by miR-205, we also examined the KCNJ10 expression in HCECs transfected with either miR-205 antagomer or mimic. Figure 6C is a Western blot demonstrating that expression of miR-205 antagomer increased KCNJ10 protein expression (Fig. 6D). In contrast, the expression of miR-205 mimic decreased KCNJ10 protein expression (Figs. 6E, 6F). 

Figure 5. 

 

KCNJ10 is the one of the targets of miR-205. (A) Graph showing the 3′ UTR of KCNJ10 contained a matched binding site of miR-205. (B) The dual luciferase reporter assay was employed in 293T transfected with the constructs of pCDH-miR-205 and psiCheck2-Luc-site1, which is the wt 3′ UTR sequence of KCNJ10, or psiCheck2-Luc-muSite1 in which the wt sequence had been mutated to disrupt the complementary matched binding within the seed sequence of miR-205. The rLuciferase activities were normalized with fLuciferase for equal transfection efficiency. The results showed that the enforced expression of miR-205 could inhibit the Luc expression in the wt sequence, which matched the seed sequence of miR-205 (psiCheck2-Luc-site1), and the inhibition was abolished when the matching site was mutated. n = 3, Student's t-test: P ≤ 0.001.

Figure 6. 

 

(A) Western blot showed that the expression of KCNJ10 was inhibited at 24 hours after injury in HCECs. (C) Silencing the endogenous miR-205 could enhance the expression of KCNJ10 in HCECs after 48-hour transfection. (E) Enforced expression of miR-205 decreased the expression of KCNJ10 but could not influence the expression of KCNJ16 in HCECs after 48-hour transfection. Data collected from (A), (C), and (E) were analyzed and are shown in (B) (P < 0.05), (D) (P < 0.01), and (F) (P < 0.01), n = 4.

To further examine the effect of miR-205 on KCNJ10, we used patch-clamp experiments to measure K⁺ currents in the HCECs transfected with miR-205 antagomer or mimic. After 24 hours of transfection, we used perforated whole-cell recording to measure the Ba²⁺-sensitive K⁺ currents. Figure 7 is a set of recordings showing K⁺ currents in the cells transfected with vector (control) (A), miR-205 antagomer (B), or miR-205 mimic (C). Data summarized in Figure 7D show that forced expression of miR-205 by pCDH-miR-205 decreased Ba²⁺ sensitive K⁺ currents from 60 ± 10 pA/pF to 25 ± 8 pA/pF (n = 4). In contrast, knocking down the endogenous miR-205 by pZip-miR-205 increased K⁺ currents to 135 ± 12 pA/pF (n = 4). 

Figure 7. 

 

Whole-cell recordings showing the Ba²⁺ -sensitive K⁺ currents in the cells transfected with control vector (A), miR-205 antagomer (pZip-miR-205 ) (B), and miR-205 mimic (pCDH-miR-205) (C). The bath solution and pipette solution contain the symmetrical 140-mM KCl and the K⁺ currents were measured from −100 mV to 60 mV. The protocol of the clamp is included. (D) A bar graph summarized the results measured at −100 mV.

The notion that scratch-injury inhibited inwardly rectifying K⁺ channels in corneal cells was also supported by the patch-clamp experiments performed in the pHCECs (Fig. 8). The single channel recording identified a 20 pS K⁺ channel in the pHCEC (Fig. 8A). It has been reported that the channel conductance of the homotetramer of KCNJ10 is approximately 20 pS. ⁴⁰ Thus, it is likely that the 20 pS K⁺ channel detected in the primary cultured HCEC might be a KCNJ10-like K⁺ channel. This view was further supported by the whole-cell patch clamp experiments in which the inwardly rectifying K⁺ currents were detected in pHCECs. Figure 8B shows whole-cell K⁺ currents measured in the control and wounded pHCECs. It was apparent that the whole-cell K⁺ currents in the primary cultured pHCECs were strongly inwardly rectifying. Moreover, scratch injury significantly decreased the inward K⁺ currents at −100 mV from 2000 ± 150 pA to 700 ± 100 pA (n = 4). 

Figure 8

 

(A) A typical K⁺ channel activity in primary culture HCEC measured at 0, −20, and −40 mV. The experiments were performed in cell-attached patch and the channel closed line is indicated by “C” and a dotted line. The cells were superfused with 140-mM NaCl-containing bath solution while the pipette solution has 140 mM KCl. A group of representative single-channel recordings showed the activity of the K⁺ channel at 0, −20, and −40 mV in the primary cultured HCECs. (B) A whole-cell recording showed the Ba²⁺ -sensitive K⁺ currents in the primary cultured HCECs under control condition or 3 hours after injuries. The K⁺ currents were measured by RAMP protocol, which clamps the cells from −100 to 100 mV. The bath solution and pipette solution contained the symmetrical 140-mM KCl solution.

miR-205 Increased the Proliferation of Wounded HCECs

BrdU incorporation cell proliferation assay (Fig. 9) revealed that miR-205 mimic stimulated, whereas antagomer significantly decreased proliferation. This observation was further supported by the immunoblot on cellular proliferation marker genes, PCNA, and Ki-67. In Figure 10A, Western blot shows that overexpression of miR-205 by pCDH-miR-205 stimulated, while inhibition of endogenous miR-205 by pZip-miR-205 decreased the expression of PCNA. Furthermore, inhibiting K⁺ channels by BaCl₂ incubation increased PCNA expression (Fig. 10C); statistical analysis of these data are shown in Figures 10B, 10D. The stimulation effect on PCNA expression by overexpression of miR-205 was observed after 48 hours but not 24-hour transfection. It is possible that 24 hours was not sufficient for pCDH-miR-205–induced expression. Supplementary Figure S7A shows that miR-205 mimic stimulated, while antagomer decreased, the expression of Ki-67, and such inhibition was abolished by KCNJ10-siRNA (Supplementary Fig. S7B). 

Figure 9. 

 

BrdU incorporation ELISA (cell proliferation assay) was performed on HCECs transfected with vector control (vec control) or pCDH-miR-205 (miR-205 mimic) or pZIP-miR-205 (miR-205 antagomer). Absorbance (OD) values showed that miR-205 mimic stimulated (P < 0.05, n = 4), antagomer decreased (P < 0.01, n = 4) proliferation in HCECs.

Figure 10. 

 

(A) Western blot showed that the forced expression miR-205 could induce the expression of PCNA, which is a cellular marker for proliferation after 48 and 72 hours of transfection. (C) The inhibition of miR-205 could decrease the expression of PCNA and this inhibition effect on PCNA expression by miR-205 antagomer could be abolished by the incubation of BaCl_2. The experiments were repeated at least four times and the representative data are analyzed in (B) and (D).

Discussion

The main finding of the present study was that the scratch injury significantly increases the expression of miR-205 in pHCECs, a highly expressed microRNA in corneal epithelial cells. ^35–37 The effect of injury on miR-205 expression is specific in HCECs because it is not observed in EA.hy926 cells, an endothelial derived immortalized cell line (data not shown here). This observation could be obtained from both primary human and mouse corneal epithelial cells. The primary human cell line from the commercial vendor has confirmed its epithelial differentiation feature by specific keratin-3/-12 examination, and the mouse cell line used in our experiment is established from explant culture from mouse corneal button. The observation that the injury failed to stimulate miR-16 expression suggests that miR-205 plays a specific and pivotal role in facilitating wound-healing progression. The finding that inhibition of miR-205 impaired wound healing in HCECs strongly reveals the indispensable role of miR-205 in wound repair in HCECs. Moreover, our finding that forced expression of the lentiviral vector carrying the miR-205 significantly stimulated wound closure further proves the key role of miR-205 in wounded HCECs. 

Three lines of evidence strongly indicate that miR-205 stimulates the healing process of wounded HCECs by inhibiting the inwardly rectifying K⁺ channel, KCNJ10. First, the inhibition of inwardly rectifying K⁺ channels by BaCl₂ rescued the compromised healing caused by an miR-205 antagomer. Second, silencing KCNJ10 expression by siRNA increased the regrowth percentage in injured HCECs. Third, injury not only significantly decreased the protein expression of endogenous KCNJ10, but also K⁺ channel activity, as evidenced by decreased inwardly rectifying K⁺ currents. Thus, we have demonstrated that KCNJ10 plays an important role in facilitating the wound-healing process in HCECs. The notion that an electric signal plays a role in facilitating the healing process in corneal cells has been suggested previously, and it was shown that the electric flux is extremely active and crucial in response to cornea injuries. ^12,41–43 The abnormal function of K⁺ channel activities could lead to the impaired wound healing in multiple organs and cell lines. ^44,45 Thus, our study shed some light on the mechanism by which modulation of K⁺ channel activities regulates the healing process following corneal epithelial injury. 

A previous study has demonstrated that miR-205 accelerated migration by targeting SHIP2 expression in wounded keratinocytes. ²⁷ We now show that KCNJ10 is also a target gene for miR-205. We speculate that miR-205 may play a profound role in facilitating the wound-healing process in HCECs at every stage of the wound-healing process, including migration and proliferation, by targeting a variety of genes directly or indirectly. This speculation is supported by the observation that overexpression of an miR-205 mimic enhanced BrdU incorporation and the expression of PCNA and Ki-67. The role of K⁺ channels in regulating PCNA or Ki-67 expression in response to miR-205 is indicated by the finding that BaCl₂ or KCNJ10-siRNA abolished the effect of inhibiting miR-205 on PCNA/Ki-67 expression and restored the PCNA/Ki-67 protein level. However, further experiments are required to investigate the mechanism by which miR-205 and KCNJ10 regulate PCNA/Ki-67 expression. 

The mechanism by which inhibition of KCNJ10 facilitates the wound-healing process is not completely understood so far. One possibility is that miR-205–induced inhibition of KCNJ10 would reduce K⁺-leak. ^14,46 Because K⁺ is the major essential intracellular cation for cell function, a decrease in the membrane K⁺ permeability would favor corneal epithelial cell proliferation, thereby repairing the wounded cells. ⁴³ Figure 11 illustrates another possibility. Because KCNJ10 has been shown to play an important role in regulation of membrane potential in a variety of cells, ^{14,15,17,18,40,47} downregulation of K⁺ channels in HCECs should lead to membrane depolarization, thereby activating the voltage-gated Ca²⁺ channels. The activation of the Ca²⁺ channels is expected to increase the intracellular Ca²⁺ level, which could be essential for stimulation of cytokines in the cornea. ⁴⁸ Thus, inhibiting KCNJ10 is expected to favor the release of growth factors or cytokines, thereby stimulating the healing process in HCECs. Further studies are required to test this hypothesis, including in eye-specific Kcnj10 knockout mice, since global Kcnj10 knockout is lethal. ^17,49  

Figure 11. 

 

Scheme showed our hypothesis that wounding injury in HCECs increased the expression of miR-205, which contributed to the downregulation of KCNJ10 by posttranscriptional regulation. The decreasing of KCNJ10 activities could lead to the depolarization by lowering the membrane potential, possible in the Ca²⁺ involved manner, therefore regulating the regeneration and healing process after wounding.

In summary, we demonstrate that injury increases the miR-205 expression in HCECs and that expression of KCNJ10 in injured HCECs is inhibited by the elevated miR-205 expression. The finding from the luciferase reporter assay reveals that miR-205 promotes healing through the posttranscriptional regulation of KCNJ10. Because prompt and uncomplicated closure of defects in the corneal epithelium is critical, and delayed healing leaves the cornea vulnerable to a loss of vision-threatening complications, ^9,10 miR-205 could be an important therapeutic target for stimulating healing in injured corneal cells. ⁵⁰  

Supplementary Materials

pdf S1 

Acknowledgments

The authors sincerely thank Michael W. Dunn, Wenhui Wang, and Michal L. Schwartzman for their critical and constructive comments in the preparation of the manuscript. 

Supported by American Heart Association 11SDG7360052 (DL). 

Disclosure: D. Lin, None; A. Halilovic, None; P. Yue, None; L. Bellner, None; K. Wang, None; L. Wang, None; C. Zhang, None