Overexpression of the inwardly rectifying potassium channel Kir4.1 or Kir4.1 Tyr⁹Asp in Müller cells exerts neuroprotective effects in an experimental glaucoma model

Abstract

Downregulation of the inwardly rectifying potassium channel Kir4.1 is a key step for inducing retinal Müller cell activation and interaction with other glial cells, which is involved in retinal ganglion cell apoptosis in glaucoma. Modulation of Kir4.1 expression in Müller cells may therefore be a potential strategy for attenuating retinal ganglion cell damage in glaucoma. In this study, we identified seven predicted phosphorylation sites in Kir4.1 and constructed lentiviral expression systems expressing Kir4.1 mutated at each site to prevent phosphorylation. Following this, we treated Müller glial cells in vitro and in vivo with the mGluR I agonist DHPG to induce Kir4.1 or Kir4.1 Tyr⁹Asp overexpression. We found that both Kir4.1 and Kir4.1 Tyr⁹Asp overexpression inhibited activation of Müller glial cells. Subsequently, we established a rat model of chronic ocular hypertension by injecting microbeads into the anterior chamber and overexpressed Kir4.1 or Kir4.1 Tyr⁹Asp in the eye, and observed similar results in Müller cells in vivo as those seen in vitro. Both Kir4.1 and Kir4.1 Tyr⁹Asp overexpression inhibited Müller cell activation, regulated the balance of Bax/Bcl-2, and reduced the mRNA and protein levels of pro-inflammatory factors, including interleukin-1β and tumor necrosis factor-α. Furthermore, we investigated the regulatory effects of Kir4.1 and Kir4.1 Tyr⁹Asp overexpression on the release of pro-inflammatory factors in a co-culture system of Müller glial cells and microglia. In this co-culture system, we observed elevated adenosine triphosphate concentrations in activated Müller cells, increased levels of translocator protein (a marker of microglial activation), and elevated interleukin-1β mRNA and protein levels in microglia induced by activated Müller cells. These changes could be reversed by Kir4.1 and Kir4.1 Tyr⁹Asp overexpression in Müller cells. Kir4.1 overexpression, but not Kir4.1 Tyr⁹Asp overexpression, reduced the number of proliferative and migratory microglia induced by activated Müller cells. Collectively, these results suggest that the tyrosine residue at position nine in Kir4.1 may serve as a functional modulation site in the retina in an experimental model of glaucoma. Kir4.1 and Kir4.1 Tyr⁹Asp overexpression attenuated Müller cell activation, reduced ATP/P2X receptor–mediated interactions between glial cells, inhibited microglial activation, and decreased the synthesis and release of pro-inflammatory factors, consequently ameliorating retinal ganglion cell apoptosis in glaucoma.

Introduction

Inwardly rectifying potassium (Kir) channels play a crucial role in establishing the hyperpolarized resting membrane potential of retinal Müller cells, which are essential for maintaining potassium and water homeostasis in the retina (Newman, 1993; Ishii et al., 1997; Kusaka and Puro, 1997; Bringmann et al., 1999). Accumulating evidence indicates that Kir channels, particularly the Kir4.1 subunit, are involved in activation of retinal Müller cells, a process that is known as gliosis. Gliosis is characterized by increased expression of glial fibrillary acidic protein (GFAP) in the context of experimental glaucoma (Francke et al., 1997; Xue et al., 2006; Ji et al., 2012; Gao et al., 2015; Wang and Yang, 2016). Glaucoma is a retinal neurodegenerative disease, with elevated intraocular pressure (IOP) being the most significant risk factor in its pathogenesis (Wang et al., 2024a, b; Adebayo and Adebayo, 2025). Animals with elevated IOP are commonly used as experimental models for studying glaucoma. In models of chronic ocular hypertension (COH) associated with experimental glaucoma, downregulation of Kir4.1 currents and a decrease in Kir4.1 protein expression have been observed in Müller cells. This downregulation leads to cell membrane depolarization and subsequent cell activation (Ji et al., 2012; Wu et al., 2018). Activated Müller cells lose their ability to functionally support retinal ganglion cells (RGCs) and release pro-inflammatory factors such as tumor necrosis factor-α (TNF-α) and interleukin-1 (IL-1) (Miao et al., 2023). In the COH retina, activated Müller cells can trigger microglial activation through the ATP/P2X7 receptor (P2X7R) pathway. The interaction between Müller cells and microglia exacerbates the release of pro-inflammatory factors, which contributes to RGC apoptosis (Hu et al., 2021; Miao et al., 2023).

Previous studies have demonstrated that downregulation of Kir channels, which induces Müller cell activation in COH-induced experimental models of glaucoma, is mediated by activation of group I metabotropic glutamate receptors (mGluR I) (Ji et al., 2012; Gao et al., 2015, 2017). The mGluR I agonist (S)-3,5-dihydroxyphenylglycine (DHPG), when injected intravitreally into rat eyes or applied to cultured primary Müller cells, induces downregulation of Kir4.1 expression. This occurs through the Ca²⁺-dependent phospholipase C/inositol triphosphate 3-ryanodine/protein kinase C signaling pathway, leading to an increase in GFAP expression (Ji et al., 2012; Gao et al., 2015, 2017; Wang and Yang, 2016). These findings suggest that Kir4.1 phosphorylation may contribute to reduced Kir4.1 protein expression, which is a key step in Müller cell activation in glaucomatous retinas. Conversely, preventing Kir4.1 phosphorylation, and thereby maintaining the stability of Kir4.1 expression, may serve as an effective strategy to mitigate Müller cell activation and, consequently, alleviate RGC injury in glaucoma. In this study, we first constructed expression systems to overexpress wild-type Kir4.1 and Kir4.1 mutated at various functional phosphorylation sites. We then investigated whether modulation of Kir4.1 expression and phosphorylation levels would disrupt Müller cell activation and reduce RGC damage in glaucomatous retinas and we explored the underlying mechanisms.

First, we constructed lentiviral expression systems overexpressing variations of Kir4.1 containing mutations at predicted phosphorylation sites and infected purified cultured Müller cells to screen the possible functional phosphorylated sites in Kir4.1. We then investigated whether Kir4.1 and Kir4.1 Tyr⁹Asp overexpression would impact Kir4.1 protein expression levels, Müller cell activation, and RGC apoptosis in COH retinas. Additionally, we explored the mechanisms underlying the neuroprotective effects of Kir4.1 and Kir4.1 Tyr⁹Asp overexpression. This involved examining the effects of Kir4.1 and Tyr⁹Asp overexpression on apoptosis-related proteins, glial cell activation, glial cell interactions, and the release of pro-inflammatory factors.

Methods

Animals

The in vivo experiments were conducted using 4-week-old male Sprague-Dawley rats (specific pathogen-free level), each weighing approximately 90 g. The rats were housed in a climate-controlled environment with a 12/12-hour light/dark cycle and were fed ad libitum in an independently ventilated cage system (GUX-56-25, Guxiu Experimental Equipment, Suzhou, China) maintained at 22–25°C with 40%–70% humidity. All animals were purchased from SLAC Laboratory Animal Co., Ltd. (Shanghai, China; license No. SCXK (Hu) 2017-0005). Animal procedures were carried out in accordance with the guidelines set by the National Institutes of Health and Fudan University for the Care and Ethical Use of Experimental Animals. Approval for these procedures was granted by the Bioethical Committee at Fudan University (approval No. IOBS 2018 JS-008, February 27, 2018). All experiments were designed and are reported in accordance with the Animal Research: Reporting of In Vivo Experiments (ARRIVE) guidelines (Percie du Sert et al., 2020). A total of 325 adult (4-week-old) rats were randomly assigned to groups (n = 6 per group), with 120 rats allocated for mRNA analysis, 148 rats for Western blot analysis, 48 rats for detection of RGC apoptosis, and 9 rats for immunohistochemistry. Additionally, 80 newborn (5-day-old) Sprague–Dawley rats were used for western blot analysis, and 408 rats for the Müller cells and microglia co-culture assay. Detailed numbers for each group are presented in the Results section.

Establishment of a rat model of glaucoma

The COH-induced rat model of glaucoma was established following procedures outlined in our previous studies (Gao et al., 2017; Zhou et al., 2023). The animals were anesthetized by intraperitoneal injection of 0.6% pentobarbital sodium (12 μL/g; Merck, Kenilworth, NJ, USA, Cat# P-010). Subsequently, 10–15 μL of iron oxide magnetic particles (Bangs Laboratories, Indianapolis, IN, USA) was injected into the anterior chamber of the rat eyes that had been infected with lentiviruses, using an ophthalmic surgical microscope (Carl Zeiss, Jena, Germany). IOP was measured using a handheld digital tonometer (TonoLab, Helsinki, Finland). The average value of five consecutive measurements, with a deviation of < 5%, was recorded. All measurements were conducted around midmorning to minimize potential circadian variations. IOP was assessed in both eyes prior to magnetic particle injection (day 0) and again at 1 week (G1w) and 2 weeks (G2w) following the injections.

Primary retinal Müller cells and microglia culture

Primary cultures of retinal Müller cells and microglia were prepared as previously described (Hu et al., 2021; Zhou et al., 2023). In brief, newborn Sprague–Dawley rats (5 days old; SLAC Laboratory Animal Co., Ltd.) were decapitated, and the retinas were isolated. The retinas were then digested with 0.25% trypsin (Gibco, Life Technologies, Rockville, MD, USA) at 37°C for 15 minutes to disperse the tissue clumps into a single-cell suspension. The cells were then cultured in cell culture flasks in Dulbecco’s Modified Eagle’s Medium/Nutrient Mixture F12 (Gibco), supplemented with 10% fetal bovine serum (Gibco) and penicillin (100 U/mL)–streptomycin (100 μg/mL) (Gibco) at 37°C in a 5% CO₂ incubator (Thermo Fisher Scientific, Waltham, MA, USA). For Müller cells, the medium was fully replaced with fresh medium every 2–3 days, and the cells were passaged every 7–8 days until the third generation. Purification of second-generation Müller cells was performed by tapping the flasks to remove non-adherent cells and microglia. For microglia, half of the medium was replaced every 4–5 days until days 14–16, at which point the cells were harvested for experiments. The purity of the Müller cells and microglia was confirmed by double immunofluorescence staining, achieving over 90% purity (Gao et al., 2017; Hu et al., 2021; Zhou et al., 2023).

Construction of lentiviral expression vectors

Lentiviral expression vectors containing the stable selection marker puromycin, along with the kcnj10 gene (GeneCopoeia^TM, Rockville, MD, USA), were constructed with seven different mutation sites in the kcnj10 gene using iGeneBio (Guangzhou, China). All lentiviral expression vectors were re-extracted through a de-endotoxin process using an Endofree Plasmid Mega Kit (Qiagen, Hilden, Germany). Sequencing primers (forward: 5′-GCG GTA GGC GTG TAC GGT-3′; reverse: 5′-CTG GAA TAG CTC AGA GGC-3′) were used to verify the mutated sites. All lentiviruses were packaged according to the instructions provided with the Lenti-Pac^TM HIV Expression Packaging Kit (GeneCopoeia^TM), and the concentrations were determined using Lenti-Pac^TM Lentivirus Concentration Solution (GeneCopoeia^TM). The titers of the harvested lentiviral stocks were assessed by counting fluorescent HEK293T cells infected with gradient-diluted lentiviruses. The mutated gene sites in each lentiviral expression vector and their corresponding amino acid changes are detailed in Additional Table 1.

Additional Table 1.

Gene mutated sites and corresponding amino acid mutated sites of Kir4.1

Name	Gene mutated site	Amino acid muated site
Kir4.1	-	-
Tyr9Asp	25 (T-G)	9 (Tyr-Asp)
Thr178Ala	532 (A-G)	178 (Thr-Ala)
Ser338Pro	1012 (T-C)	338 (Ser-Pro)
Thr346Ala	1036 (A-G)	346 (Thr-Ala)
Ser360Pro	1078(T-C)	360 (Ser-Pro)
Ser370Arg	1108 (A-C)	370 (Ser-Arg)
Ser377Gly	1129 (A-G)	377 (Ser-Gly)

Infection of Müller cells with lentiviruses

Purified Müller cells were added to a 96-well plate at a density of 1 × 10⁴ cells per well. Each well received different dilutions of lentiviral stock, ranging from 0.1 μL to 100 μL. The titers of lentiviruses were determined by counting the fluorescing cells and the multiplicity of infection (MOI) values were evaluated following the calculation: (MOI = (lentivirus titer × lentivirus volume)/total cell number). The titers and MOI values for each lentivirus used in this study are presented in Additional Table 2. Lentiviruses with calculated MOI values were used to infect the purified Müller cells. At 5 days post-infection, 2 μg/mL of puromycin (Gibco) was added to select for infected cells, ensuring consistent efficiency of lentiviral infection across all groups.

Additional Table 2.

Titers and MOI value of each lentivirus

Lentivirus	Titers of lentivirus (TU/mL)	MOI value
LV-NC	2×109	100
Kir4.1	4×108-8×108	120
Tyr9Asp	7×108-8×108	100
Thr178Ala	2×108	120
Ser338Pro	4×108	100
Thr346Ala	2×108	100
Ser360Pro	1×108	120
Ser370Arg	2×108-8×108	120
Ser377Gly	3×108	100

eGFP: Enhanced green fluorescent protein; LV-NC: eGFP control lentiviruses; MOI: multiplicity of infection.

Subretinal injection of the lentiviruses and intravitreal injection of drugs

The glial cell–specific promoter retinaldehyde binding protein 1 (RLBP1) was used to drive kcnj10 gene expression from the lentiviral vectors in the in vivo experiments. Lentiviral particles (≥ 10⁹ TU/mL, 2.5 μL; ≥ 10⁸ TU/mL, 5 μL) were injected subretinally using an ophthalmic surgical microscope (Carl Zeiss) and a microinjector (Hamilton, Reno, NV, USA). A 33-gauge needle was inserted at the limbus area and then advanced through the vitreous into the subretinal space on the opposite side of the retina (You et al., 2014; Peng et al., 2017). A lentivirus encoding enhanced green fluorescent protein (eGFP) was used as a negative control. At 3 weeks after the injections, 100 μM DHPG (Tocris, Bioscience, Ellisville, MO, USA) diluted in 3 μL of 0.9% normal saline was injected intravitreally (Hu et al., 2021; Zhou et al., 2023). Control groups were injected in the same manner with saline only (5 μL for subretinal injection and 3 μL for intravitreal injection).

Western blot analysis

Western blot analysis was performed following established protocols (Dong et al., 2015; Hu et al., 2021; Zhou et al., 2023). Total protein was extracted from retinas or cultured cells and quantified using a standard bicinchoninic acid assay kit (Thermo Fisher Scientific). The final protein samples, mixed with 1× sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE) loading buffer, were denatured at 95°C for 5 minutes. Membrane proteins were extracted with the ProteoExtract Transmembrane Protein Extraction Kit (Millipore, Billerica, MA, USA), which contained 2× SDS-PAGE loading buffer but no denaturant.

The proteins were separated by SDS-PAGE using a Mini Protean 3 electrophoresis system (Bio-Rad, Hercules, CA, USA) and then transferred to a polyvinylidene fluoride (PVDF) membrane (R&D, Minneapolis, MN, USA) using a Mini Transblot electrophoretic transfer system (Bio-Rad). To block non-specific adsorption, the PVDF membranes were treated with 5% skim milk at room temperature (23–25°C) for 1 hour. The primary antibodies used in this study included: mouse anti-GFAP (1:500, Sigma-Aldrich, St. Louis, MO, USA, Cat# G6171, RRID: AB_1840893), rabbit anti-Kir4.1 (1:400, Alomone, Jerusalem, Israel, Cat# APC-035, RRID: AB_2040120; APC-165 for membrane protein, RRID: AB_2041043), anti-B cell lymphoma/leukemia-2 (Bcl-2; 1:1000, Abclonal Biotechnology Co., Ltd, Wuhan, Hubei, China, Cat# A0208, RRID: AB_2757022), rabbit anti-Bcl-2-associated X protein (Bax; 1:1000, Abcam, Cambridge, MA, USA, Cat# ab32503, RRID: AB_725631), anti-myeloid differentiation primary response protein 88 (MYD88; 1:1000, Novus Biologicals, Littleton, CO, USA, Cat# NB100-56698SS, RRID: AB_838599), anti-nuclear factor kappa B p65 (NF-κB p65; 1:1000, Cell Signaling Technology, Danvers, MA, USA, Cat# 3033T, RRID: AB_331284), goat anti-translocator protein (TSPO; 1:1000, Novus Biologicals, Cat# NB100-41398, RRID: AB_788260), rabbit anti-cleaved caspase 3 (1:500, Zen-Bioscience, Chengdu, Sichuan, China, Cat# 341034, RRID: AB_2924432), rabbit anti-IL-1 receptor-associated kinase 1 (IRAK1; 1:1000, Abcam, Cat# ab180747, RRID: AB_2895218), rabbit anti-TNF receptor-associated factor 6 (TRAF6; 1:1000, Abcam, Cat# ab33915, RRID: AB_778572), mouse anti-β-actin (1:10,000, Sigma-Aldrich, Cat# A5441, RRID: AB_476744), rabbit anti-glyceraldehyde-3-phosphate dehydrogenase (GAPDH; 1:10,000, Sigma-Aldrich, Cat# G9545, RRID: AB_796208), and rabbit anti-Na-K ATPase (1:8000, Abcam, Cat# ab167390, RRID: AB_2890241). The secondary antibodies used were horseradish peroxidase (HRP)-conjugated donkey anti-mouse (1:10,000, Jackson ImmunoResearch Labs, West Grove, PA, USA, Cat# 715-035-150, RRID: AB_2340770), HRP-conjugated donkey anti-rabbit (1:10,000, Jackson ImmunoResearch Labs, Cat# 711-035-152, RRID: AB_10015282), and HRP-conjugated donkey anti-goat (1:10,000, Jackson ImmunoResearch Labs, Cat# 705-035-147, RRID: AB_2313587). The membrane was incubated with primary antibodies overnight at 4°C and with the secondary antibodies for 2 hours at room temperature. Enhanced chemiluminescent substrate (Thermo Fisher Scientific) was used for detecting the HRP signals on the immunoblots. Quantitative analysis of the protein blots was conducted using Alpha View software (version 3.3.0, Cell Biosciences, Inc., Santa Clara, CA, USA). Protein expression levels on each blot were normalized to the corresponding internal reference (β-actin, GAPDH, or Na-K ATPase) and then to the control samples.

Quantitative polymerase chain reaction

Total retinal RNA was extracted using Trizol, as described previously (Gao et al., 2015; Liu et al., 2018). For cultured microglia, total RNA was extracted using a TaKaRaMiniBEST Universal RNA Extraction Kit (Takara, Kusatsu, Japan). Reverse transcription was performed using a PrimeScript RT Reagent Kit with gDNA Eraser (Takara) to eliminate genomic DNA and synthesize complementary DNA (cDNA). The cDNAs were diluted (5-fold dilution for IL-1β, TNF-α, CNTF; 10-fold dilution for BDNF, NGF) and used as templates for quantitative PCR using a SYBR Green qPCR reaction system. The template cDNA and gene-specific primers were mixed with SYBR Premix Ex Taq^TM II (Tli RNaseH Plus) from Takara. The sequences of the primer used are shown in Additional Table 3. The reaction procedure included an initial step of 95°C for 30 seconds, followed by 40 cycles of 5 seconds at 95°C, 40 seconds at 60°C, and varying lengths of time at 72°C, depending on the target gene: 10 seconds for IL-1β, 15 seconds for nerve growth factor (NGF), and 20 seconds for TNF-α, brain-derived neurotrophic factor (BDNF), and ciliary neurotrophic factor (CNTF). All reactions were performed using a QuantStudio 3 Real-Time PCR System from Thermo Fisher Scientific (Waltham, MA, USA). Reactions showing a single peak in the melting curve were selected, and the Ct values were calculated using the 2^–ΔΔCt method.

Additional Table 3.

Primer sequences

Gene	Primer sequences	Length (bp)	Accession No.
IL-1β	5’-TCCAGTCAGGCTTCCTTGTG-3’ 5’-AGGTCATTCTCCTCACTGTCG-3’	100	NM_031512.2
TNFα	5’-AAGTTCCCAAATGGGCTC-3 ’ 5’-TCACAGAGCAATGACTCCAAAG-3’	522	NM_012675.3
Bdnf	5’-GTGACAGTATTAGCGAGTGGG-3’ 5’-GATTGGGTAGTTCGGCATT-3 ’	217	M61175.1
Cntf	5’-TTCGCAGAGCAAACACCTCT-3 ’ 5’-AGGCCCTGATGTTTTACATAAGATT-3’	125	NM_013166.2
Ngf	5’-GGACGCAGCTTTCTATCCTGG-3’ 5’-CCCTCTGGGACATTGCTATCTG-3’	129	NM_001277055.1
Actin	5’-CGCATCCTCTTCCTCCCTG-3’ 5’-CACAGGATTCCATACCCAGGA-3’	128	NM_031144.3

Bdnf: Brain-derived neurotrophic factor; Cntf: ciliary neurotrophic factor; IL-1β: interleukin-1β; Ngf: nerve growth factor; TNFα: tumor necrosis factor-α.

Immunofluorescence staining

Cultured Müller cells were fixed with 4% paraformaldehyde in phosphate-buffered saline (PBS, pH 7.4) for 20 minutes at room temperature. They were then incubated in 0.25% Triton X-100 for 10 minutes, followed by incubation in a solution containing 10% donkey serum and 1% bovine serum albumin in PBS with Tween for 1 hour at room temperature. The following primary antibodies were used: mouse anti-GFAP (1:400, Sigma-Aldrich, Cat# G6171, RRID: AB_1840893) and rabbit anti-GFP (1:500, Abcam, Cat# ab290, RRID: AB_303395). The secondary antibodies used were donkey anti-mouse Cy3 (1:500, Cat# 715-165-151, RRID: AB_2315777) and donkey anti-rabbit 488 (1:500, Cat# 711-545-152, RRID: AB_2313584, Jackson ImmunoResearch Labs). Immunofluorescence staining of retinal vertical slices was performed according to a previously described procedure (Dong et al., 2015). Briefly, the retinas were fixed with 4% paraformaldehyde in PBS (pH 7.4) for 2 hours at 4°C. After dehydration in a gradient of sucrose solutions, the retinal vertical slices were cut to a thickness of 14 μm using a Leica microtome (Nussloch, Germany) and mounted on Superfrost Plus slides (Thermo Fisher Scientific). The slices were then incubated with 0.4% Triton X-100, 5% donkey serum, and 1% bovine serum albumin for 1.5 hours. For the retinal slices, the primary antibodies used were anti-glutamine synthetase (1:300, Millipore, Cat# MAB302, RRID: AB_2110656) and rabbit anti-GFP (1:500, Abcam, Cat# ab290, RRID: AB_2110656). The same secondary antibodies mentioned above were used. The cultured cells and retinal slices were incubated with the primary antibodies overnight at 4°C and subsequently with the secondary antibodies for 2 hours at room temperature. Images were captured using an Olympus confocal laser-scanning microscope (FV1000, Monolith, Tokyo, Japan).

Terminal deoxynucleotidyl transferase-mediated dUTP nick end labeling assay

A terminal deoxynucleotidyl transferase-mediated dUTP nick end labeling (TUNEL) assay was performed to label apoptotic cells using a DeadEnd Fluorometric TUNEL System (Promega, Madison, WI, USA), following previously reported procedures (Chen et al., 2017; Cheng et al., 2021; Wu et al., 2024). Whole flat-mounted retinas were fixed and positioned with the ganglion cell layer (GCL) facing upwards. TUNEL-positive signals in the GCL of each retina were captured using a fast digital slide scanner (VS120, Olympus, Monolith). TUNEL-positive signals that overlapped with 4′,6-diamidino-2-phenylindole (Roche, Basel, Switzerland, Cat# 10236276001)-labeled nuclei were counted in each retina. Two independent individuals, who were blinded to the animal group assignments, conducted the experiments.

Müller cell and microglia co-culture assay

The Müller cell and microglia co-culture assay was conducted as previously described (Hu et al., 2021), with some modifications, to investigate the effects of lentivirus-infected Müller cells on microglia. Briefly, Müller cells were pre-activated with 100 µM DHPG for 12 hours and then co-cultured with microglia in the presence or absence of the P2X7R antagonist A-740003 (1 µM, MedChemExpress, Shanghai, China, Cat# HY-50697) and the P2X4R antagonist 5-BDBD (5 µM, MedChemExpress, Cat# HY101911) for 72 hours. mRNA levels of the inflammatory factors interleukin-1β (IL-1β) and tumor necrosis factor-α (TNF-α) in microglia were assessed by qPCR. MYD88, IRAK1, TRAF6, NF-κB p65, and TSPO protein levels were analyzed by Western blotting. Additionally, IL-1β and TNF-α concentrations in the co-culture medium were measured using enzyme-linked immunosorbent assay (ELISA) kits (R&D), while ATP concentrations were determined using an enhanced ATP detection kit (Beyotime Biotechnology, Shanghai, China).

Microglia proliferation assay

The effects of lentivirus-infected Müller cells on microglial proliferation were investigated using a Müller glial cell and microglia co-culture system (24-well, 0.4 μm pore size, Corning, Bedford, MA, USA). A Click-iT^TM 5-ethynyl-2′-deoxyuridine (EdU) Imaging Kit (Invitrogen, Thermo Fisher Scientific) was used, following previously reported procedures (Xu et al., 2022). Pre-activated Müller cells were co-cultured with EdU-labeled microglia for 24 hours. The microglia were then incubated overnight at 4°C with a primary antibody (rabbit anti-ionized calcium-binding adapter molecule 1 (Iba1), 1:800, Abcam, Cat# ab178846, RRID: AB_2636859) and subsequently for 2 hours at room temperature with a secondary antibody (donkey anti-rabbit 488, 1:500, Jackson ImmunoResearch Labs, Cat# 711-545-152, RRID: AB_2313584). Incorporated EdU was detected using reaction cocktails (Invitrogen, Thermo Fisher Scientific). EdU-positive microglia were detected using a fast digital slide scanner (VS120, Olympus). Experiments were conducted in triplicate, and in each sample, five random fields were selected to count the number of EdU and Iba1 double-labeled cells. The numbers of proliferating microglia in the non-lentivirus infection and lentivirus infection groups were normalized to their corresponding controls (no Müller cells/LV-NC + normal Müller cells).

Microglial migration assay

Cell migration was analyzed as previously described (Xu et al., 2022). Lentivirus-infected Müller cells were treated with 100 μM DHPG for 12 hours and then co-cultured with microglia for 48 hours using a 24-well insert with an 8 μm pore size (Corning). The number of migrated microglia in five randomly selected fields from each insert was determined using a Nikon Eclipse Ti microscope (NIKON Corporation, Tokyo, Japan). The experiments were conducted in triplicate. The numbers of migratory microglia in the non-lentivirus and lentivirus infection groups were normalized to their corresponding controls (no Müller cells/LV-NC + normal Müller cells) for comparison.

Prediction of phosphorylation sites in Kir4.1

Seven predicted phosphorylation sites in Kir4.1 were identified using various prediction websites, including NetPhos (https://services.healthtech.dtu.dk/service.php?NetPhos-3.1; Blom et al., 2004; Solayman et al., 2017), KinasePhos (http://KinasePhos.mbc.nctu.edu.tw/; Huang et al., 2005), Disphos (http://www.dabi.temple.edu/disphos/), Scansite (https://scansite4.mit.edu/; Obenauer et al., 2003), Musite (https://www.musite.net/; Gao et al., 2010; Wang et al., 2020), and PhosphoSitePlus (https://www.phosphosite.org/homeAction.action; Que et al., 2012).

Statistical analysis

All data are presented as mean ± standard error of the mean and were analyzed using GraphPad Prism version 10.0.0 for Windows (GraphPad Software, Boston, MA, USA, www.graphpad.com). Brown-Forsythe test and Bartlett’s test were used to assess the homogeneity of variances. When P > 0.05 (indicating equal variances), one-way analysis of variance (ANOVA) with Tukey-Kramer multiple comparisons test or unpaired two-tailed t-test was used. Conversely, when P < 0.05 (indicating unequal variances), the Kruskal-Wallis test (a nonparametric test) with Dunn’s multiple comparisons test was applied. P values less than 0.05 were considered statistically significant.

Results

Construction of an expression system for Kir4.1 with various mutation sites

Seven phosphorylation sites were predicted in Kir4.1, all of which are located within intracellular sequences. We mutated the codons encoding these putative sites in the kcnj10 gene (1140 bp, GenBank: NM_031602.2), cloned the mutant genes into lentiviral expression vectors (pEZ-Lv201), and confirmed correct mutation and insertion by sequencing (Figure 1A and B). All lentiviral expression vectors were successfully packaged into lentiviruses, which were subsequently used to infect purified cultured Müller cells (Figure 1C).

Figure 1.

Identification and verification of potential functional phosphorylation sites in Kir4.1.

(A) Schematic diagram illustrating the design of the Kir4.1 mutant lentiviral expression system. (B) Mutation sites in the kcnj10 gene. (C) Immunofluorescence staining showing primary cultured Müller cells infected with lentiviruses for overexpressing Kir4.1, the Kir4.1 mutants, and the eGFP control. All lentiviruses successfully infected purified cultured Müller cells. Scale bar: 20 μm. (D, E) Representative immunoblots (D) and quantification (E) demonstrating changes in total Kir4.1 protein expression in DHPG-treated cultured Müller cells infected with lentiviruses overexpressing Kir4.1 and seven different Kir4.1 mutants. All data are normalized to GAPDH levels and then to controls. One-way analysis of variance (Tukey–Kramer multiple comparisons test) was performed (*P < 0.05, ***P < 0.001, vs. Ctr group). (F, G) Representative immunoblots (F) and densitometric quantification (G) showing changes in membrane Kir4.1 protein expression in DHPG-treated cultured Müller cells infected with lentiviruses overexpressing Kir4.1, Kir4.1 Tyr⁹Asp, and Kir4.1 Ser³⁷⁰Arg. All data are normalized to Na-K-ATP levels and then to the Kir4.1 group. One-way analysis of variance (Tukey–Kramer multiple comparisons test) was used (**P < 0.01, vs. Kir4.1 group). CMV: Cytomegalovirus; Ctr: control; DHPG: (S)-3,5-dihydroxyphenylglycine; eGFP: enhanced green fluorescent protein; GAPDH: glyceraldehyde-3-phosphate dehydrogenase; SV40: simian vacuolating virus 40.

We previously showed that activation of mGluR I induces a downregulation of Kir currents and Kir4.1 protein expression in Müller cells, contributing to Müller cell activation in glaucomatous retinas in an animal model (Ji et al., 2012; Gao et al., 2015). Treatment with DHPG, an mGluR I agonist, significantly decreases Kir4.1 protein expression in cultured Müller cells (Ji et al., 2012; Gao et al., 2015). Thus, in this study, we first investigated the effects of DHPG treatment on Kir4.1 expression in cultured Müller cells overexpressing wild-type Kir4.1 or Kir4.1 harboring different phosphorylation site mutations. As shown in Figure 1D and E, total Kir4.1 expression levels in Müller cells overexpressing Kir4.1 Tyr⁹Asp or Kir4.1 overexpression were significantly higher than those in the non-infected lentivirus negative control group (385.6% ± 53.7% of control group, n = 3, P < 0.001 and 229.5% ± 35.1% of control group, n = 3, P < 0.05, respectively). Additionally, a trend toward increased Kir4.1 expression was observed in cells overexpressing Kir4.1 Ser³⁷⁰Arg (184% ± 39.4% of control group, n = 3, P > 0.05). We then assessed the levels of Kir4.1 proteins in the cell membrane components of Müller cells that exhibited Kir4.1 overexpression, the Kir4.1 Tyr⁹Asp mutation, and the Kir4.1 Ser³⁷⁰Arg mutation. The Western blot results indicated that Kir4.1 Tyr⁹Asp expression (168.9% ± 14.9% of control group, n = 3, P < 0.01) was significantly higher than Kir4.1 and Ser³⁷⁰Arg overexpression (Figure 1F and G) in Müller cells following DHPG treatment. These results suggest that the tyrosine residue at position nine (Tyr⁹) in Kir4.1 may be a key functional phosphorylation site for reversing the downregulation of Kir4.1 induced by mGluR I activation.

Kir4.1 and Kir4.1 Tyr⁹Asp overexpression attenuate Müller cell activation

Next, we investigated the effects of Kir4.1 and Kir4.1 Tyr⁹Asp overexpression on Müller cell activation by assessing changes in GFAP expression. As shown in Figure 2A and B, GFAP expression levels did not differ significantly among cultured Müller cells expressing the eGFP control lentivirus (LV-NC), those overexpressing Kir4.1, and those overexpressing Kir4.1 Tyr⁹Asp, compared with the control. However, after treatment with DHPG for 1 hour, GFAP protein levels in the Kir4.1 and Kir4.1 Tyr⁹Asp overexpression groups were significantly reduced to 56.6% ± 6.1% of the control (n = 3, P < 0.01) and 54.3% ± 5.7% of the control (n = 3, P < 0.01), respectively (Figure 2A and B). These findings suggest that both Kir4.1 overexpression and Kir4.1 Tyr⁹Asp overexpression attenuate Müller cell activation, which was supported by the following fluorescence immunostaining experiments. First, we confirmed expression of LV-NC (Figure 2C, c4–c6), Kir4.1 (Figure 2C, c7–c9), and Kir4.1 Tyr⁹Asp (Figure 2C, c10–c12) in cultured Müller cells. The GFAP fluorescence intensity in all three groups was comparable to that seen in the control group (Figure 2C, c1–c3). After 1 hour of DHPG treatment, the GFAP fluorescence intensity was significantly lower in Müller cells overexpressing Kir4.1 (Figure 2C, c19–c21) and Kir4.1 Tyr⁹Asp (Figure 2C, c22–c24) compared with the control group (Figure 2C, c13–c15).

Figure 2.

Effects of Kir4.1 overexpression and Kir4.1Tyr⁹Asp overexpression on GFAP expression in cultured Müller glial cells.

(A, B) Representative immunoblots (A) and densitometric quantification (B) showing changes in GFAP expression in non-treated and DHPG-treated cultured Müller cells infected with lentiviruses overexpressing the eGFP control (LV-NC), Kir4.1, and Kir4.1 Tyr⁹Asp. All data are normalized to GADPH and then to Ctr. One-way analysis of variance (Tukey–Kramer multiple comparisons test) was performed (**P < 0.01, vs. Ctr group). (C) Double immunofluorescence staining showing changes in GFAP (Cy3, red) expression in non-treated (c1–c12) and DHPG-treated (c13–24) cultured Müller cells infected with lentiviruses overexpressing the eGFP Ctr (LV-NC) (c4–c6 and c16–c18), Kir4.1 (c7–c9 and c19–c21), and Kir4.1 Tyr⁹Asp (c10–c12 and c22–c24) mutation. Kir4.1 overexpression and Kir4.1 Tyr⁹Asp overexpression reduced the DHPG-induced increase in GFAP expression in Müller cells. GFP (indicated by Alexa488) and DAPI are shown in green and blue, respectively. Scale bars: 20 µm. Ctr: Control; DHPG: (S)-3,5-dihydroxyphenylglycine; eGFP: enhanced green fluorescent protein; GAPDH: glyceraldehyde-3-phosphate dehydrogenase; GFAP: glial fibrillary acidic protein; GFP: green fluorescent protein; LV-NC: eGFP control lentiviruses.

For the in vivo experiments, we replaced the cytomegalovirus promoter in the lentiviral expression vectors with the glial cell–specific promoter RLBP1 (GenBank: NC_000073.6:79384575-79385286) (Bunt-Milam and Saari, 1983; Saari and Crabb, 2005; Vogel et al., 2007; Geller et al., 2008; Additional Figure 1 ^{(2.6MB, tif)} A). Double immunofluorescent co-immunostaining for glutamine synthetase, a Müller cell marker (Bringmann et al., 2006), and GFP demonstrated that, 4 weeks after lentiviral injection (Miyoshi et al., 1997; Cheng et al., 2005; Yu et al., 2016), retinal Müller cells were successfully infected by lentiviruses carrying the RLBP1 promoter (Additional Figure 1 ^{(2.6MB, tif)} B–D). In subsequent experiments, retinal Müller cells were infected with LV-NC, Kir4.1, and Kir4.1 Tyr⁹Asp lentiviruses. Three weeks after infection, DHPG was injected intravitreally, and changes in GFAP and Kir4.1 expression were examined by Western blot. When saline was injected intravitreally, GFAP and Kir4.1 protein levels in the LV-NC, Kir4.1, and Kir4.1 Tyr⁹Asp groups were comparable to those seen in the control group (Figure 3A–D). Although GFAP expression in the lentivirus-injected groups did not differ from that in the control group 1 week after DHPG injection, GFAP levels were significantly increased in the LV-NC group 2 weeks post-DHPG injection, reaching 182.9% ± 14.7% of the control (n = 6, P < 0.01). In comparison, GFAP levels reached 131.1% ± 11.7% of the control (n = 6, P > 0.05) in the Kir4.1 Tyr⁹Asp group and 117.9% ± 10.9% of the control (n = 6, P > 0.05) in the Kir4.1 group (Figure 3A and C). Additionally, Kir4.1 levels in the LV-NC group slightly decreased 1 week after DHPG injection, while they increased in both the Kir4.1 and Kir4.1 Tyr⁹Asp groups (Figure 3A and E). Two weeks post–DHPG injection, Kir4.1 protein levels in the Kir4.1 and Kir4.1 Tyr⁹Asp overexpression groups were slightly higher than those in the control group (Figure 3A and E). These results suggest that Kir4.1 and Kir4.1 Tyr⁹Asp overexpression in Müller cells can attenuate Müller cell activation.

Figure 3.

Kir4.1 and Kir4.1Tyr⁹Asp overexpression attenuate retinal Müller cell activation and inhibit DHPG-induced Kir4.1 downregulation.

(A) Representative immunoblots showing changes in GFAP and Kir4.1 expression in retinas from uninjected rats and rats subjected to DHPG intravitreal injection at 1 and 2 weeks post-injection. (B–E) Densitometric quantification of changes in GFAP (B, C) and Kir4.1 (D, E) expression under the conditions described in panel A. The retinas were infected with lentiviruses encoding eGFP (LV-NC), Kir4.1, and Kir4.1 Tyr⁹Asp. All data were normalized to GAPDH/β-actin levels and then to control values. One-way analysis of variance (Tukey–Kramer multiple comparisons test) was used for panel D, while the Kruskal–Wallis test (Dunn’s multiple comparisons test) was applied for panel E. *P < 0.05, **P < 0.01, vs. Ctr group; #P < 0.05, vs. DHPG 2w LV-NC group. (F) Representative immunoblots showing changes in GFAP expression in normal or COH retinas at G2w in Müller cells infected with LV-NC, Kir4.1, and Kir4.1 Tyr⁹Asp lentiviruses. (G, H) Bar charts summarizing the average densitometric quantification of the GFAP bands shown in panel F. All data were normalized to GAPDH levels and then to control values. Unpaired two-tailed t-test was performed for panel B, and one-way analysis of variance (Tukey–Kramer multiple comparisons test) was used for panel C. *P < 0.05, **P < 0.01, vs. Ctr group; #P < 0.05, vs. LV-NC group. Ctr: Control; DHPG: (S)-3,5-dihydroxyphenylglycine; eGFP: enhanced green fluorescent protein; G2w: chronic ocular hypertension at 2 weeks; GAPDH: glyceraldehyde-3-phosphate dehydrogenase; GFAP: glial fibrillary acidic protein; LV-NC: eGFP control lentiviruses.

Kir4.1 and Kir4.1 Tyr⁹Asp overexpression in Müller cells attenuate retinal ganglion cell apoptosis in rats with chronic ocular hypertension

In the rat COH model, Müller cells are activated, as indicated by increased GFAP expression (Ji et al., 2012). Activated Müller cells promote RGC apoptosis in glaucoma (Li et al., 2021; Miao et al., 2023). Here, we asked whether Kir4.1 and Kir4.1 Tyr⁹Asp overexpression in Müller cells can ameliorate Müller cell activation and RGC apoptosis in COH retinas. Consistent with our previous studies performed in rat models of COH (Zhao et al., 2018; Xu et al., 2022), the average IOP of the affected eyes increased from 10.7 ± 0.8 mmHg before the operation (0 days) to 23.4 ± 2.6 mmHg 1 week (G1w) and 28.4 ± 3.1 mmHg 2 weeks (G2w) after the operation (n = 24, P < 0.001). These values were significantly higher than those of the contralateral, unaffected eyes (11.5 ± 0.4 mmHg; n = 24, P < 0.001). As shown in Figure 3F and G, GFAP expression in the COH retinas was elevated to 305.6% ± 33.3% of the control (n = 6, P < 0.01) at G2w (Chen et al., 2017; Xu et al., 2022). In normal rats infected with LV-NC, Kir4.1, and Kir4.1 Tyr⁹Asp lentiviral vectors, retinal GFAP protein levels increased significantly to 202.4% ± 13.6% of the control (n = 6, P < 0.05), 273.6% ± 36.9% of the control (n = 6, P < 0.05), and 323.2% ± 46.7% of the control (n = 6, P < 0.01), respectively (Figure 3F and H). In COH retinas of rats injected with the LV-NC lentivirus, GFAP levels further increased to 370.8% ± 56.8% of the control (n = 6, P < 0.001 vs. control; P < 0.05 vs. LV-NC alone) at G2w (Figure 3F and H). However, no further increase in GFAP levels was observed in COH retinas injected with lentiviruses encoding Kir4.1 and Kir4.1 Tyr⁹Asp at G2w (Kir4.1 + G2w: 290% ± 31.1% of the control, n = 6, P < 0.05 vs. control; P > 0.05, vs. Kir4.1 group; Kir4.1 Tyr⁹Asp + G2w: 341.7% ± 36.2% of the control, n = 6, P < 0.001 vs. control, P > 0.05, vs. Kir4.1 Tyr⁹Asp group) (Figure 3F and H).

TUNEL staining revealed a significant increase in the total number of TUNEL-positive RGCs in whole flat-mounted COH retinas at G2w, rising from a control value of 10.0 ± 1.2 (n = 6) to 235.8 ± 43.9 (n = 6, P < 0.001 vs. control) (Figure 4A and B). The number of apoptotic RGCs also increased to 102.8 ± 10.0 (n = 6, P < 0.05), 99.8 ± 9.2 (n = 6, P < 0.05), and 147 ± 18.9 (n = 6, P < 0.05) in rats subjected to subretinal injection of lentiviruses encoding LV-NC, Kir4.1, and Kir4.1 Tyr⁹Asp, respectively, compared with control rats (Figure 4C and D). At G2w following LV-NC lentiviral injection, the number of apoptotic RGCs in the COH retinas had increased to 468.5 ± 66.2 (n = 6, P < 0.001 vs. control; P < 0.001 vs. LV-NC alone group) (Figure 4C and D). However, in the COH retinas at G2w from rats subjected to Kir4.1 and Kir4.1 Tyr⁹Asp lentiviral injections, there was no further increase in the number of apoptotic RGCs (Kir4.1 + G2w: 205.8 ± 37.9, n = 6, P < 0.01 vs. control, P > 0.05 vs. Kir4.1 alone group; Kir4.1 Tyr⁹Asp + G2w: 143.5 ± 18.8, n = 6, P < 0.05 vs. control, P > 0.05 vs. Kir4.1 Tyr⁹Asp alone group) (Figure 4C and D).

Figure 4.

Kir4.1 and Kir4.1 Tyr⁹Asp overexpression attenuate retinal Müller glial cell activation and ganglion cell apoptosis in a rat model of chronic ocular hypertension.

(A) Representative images of TUNEL staining detecting apoptotic RGCs in whole flat-mounted retinas from control (a1–a3) and COH (a4–a6) rats at G2w. The number of TUNEL-positive RGCs was significantly increased in COH retinas. Scale bar: 20 μm. (B) Bar charts summarizing the changes in the average number of TUNEL-positive signals. n = 6 for each group. ***P < 0.001, vs. Ctr. Unpaired two-tailed t-test. (C) Representative images of TUNEL staining of normal and COH retinas infected with LV-NC, Kir4.1, and Kir4.1 Tyr⁹Asp lentiviruses at G2w. Kir4.1 overexpression and Kir4.1 Tyr⁹Asp overexpression in Müller cells significantly reduced the number of TUNEL-positive signals in COH retinas at G2w. Scale bar: 20 μm. (D) Bar charts summarizing the changes in the average number of TUNEL-positive signals under the different conditions shown in panel C. n = 6 for each group. *P < 0.05, **P < 0.01, ***P < 0.001, vs. Ctr group; ###P < 0.001, vs. LV-NC group. One-way analysis of variance (Tukey–Kramer multiple comparisons test) was performed. COH: Chronic ocular hypertension; Ctr: control; DAPI: 4′,6-diamidino-2-phenylindole; eGFP: enhanced green fluorescent protein; G2w: chronic ocular hypertension at 2 weeks; LV-NC: eGFP control lentiviruses; TUNEL: terminal deoxynucleotidy transferase-mediated dUTP nick end labeling.

Next, we explored changes in the expression levels of apoptosis-related proteins. The levels of the pro-apoptotic protein Bax increased to 161.5% ± 21.1% of the control (n = 5, P < 0.05) in COH retinas at G2w (Figure 5A and B). In the COH retinas of rats subjected to LV-NC lentiviral injections, Bax expression remained elevated at 162.9% ± 10.4% of the control level (n = 5, P < 0.05 vs. LV-NC). However, Bax expression in the Kir4.1 + G2w (126.9% ± 5.7% of control, n = 5, P > 0.05 vs. LV-NC) and Kir4.1 Tyr⁹Asp + G2w (122.3% ± 5.1% of control, n = 5, P > 0.05 vs. LV-NC) groups was comparable to that in the LV-NC group (Figure 5A and C). At G2w, expression of the anti-apoptotic protein Bcl-2 in the COH retinas was not significantly different from controls (Figure 5A and D). However, Bcl-2 levels were significantly higher than those in the LV-NC group in COH retinas overexpressing Kir4.1 or Kir4.1 Tyr⁹Asp (Kir4.1 + G2w: 151.1% ± 7.8% of control, n = 5, P < 0.05 vs. LV-NC; Kir4.1 Tyr⁹Asp + G2w: 184.0% ± 13.1% of control, n = 5, P < 0.05 vs. LV-NC) (Figure 5A and E). The cleaved caspase-3 expression level showed a pattern similar to that of Bax. Specifically, cleaved caspase-3 levels significantly increased to 168.8% ± 13.7% of the control (n = 5, P < 0.05) at G2w in COH retinas. However, injection with the Kir4.1 or Kir4.1 Tyr⁹Asp lentivirus restored cleaved caspase-3 levels to those seen in the LV-NC group (Kir4.1 + G2w: 121.0% ± 21.4% of control, n = 5, P > 0.05 vs. LV-NC; Kir4.1 Tyr⁹Asp + G2w: 113.6% ± 11.7% of control, n = 5, P > 0.05 vs. LV-NC) (Figure 5A, F, and G). These results are consistent with the changes observed in GFAP expression, suggesting that Kir4.1 and Kir4.1 Tyr⁹Asp overexpression in Müller cells reduced RGC injury by suppressing Müller cell activation in glaucoma.

Figure 5.

Changes in Bax, Bcl-2, and cleaved-caspase 3 levels in COH retinas infected with lentiviruses overexpressing Kir4.1 or Kir4.1 Tyr⁹Asp.

(A) Representative immunoblots showing changes in Bax, Bcl-2, and cleaved-caspase 3 expression in normal or COH retinas infected with LV-NC, Kir4.1, and Kir4.1 Tyr⁹Asp lentiviruses at G2w. (B, D, F) Bar charts summarizing the average densitometric quantification of immunoreactive bands for Bax (B), Bcl-2 (D), and cleaved-caspase 3 (F) expression in the retinas of the control and COH groups at G2w. n = 5 for each group. *P < 0.05, vs. Ctr group. Unpaired two-tailed t-test. (C, E, G) Bar charts summarizing the average densitometric quantification of immunoreactive bands for Bax (C), Bcl-2 (E), and cleaved-caspase 3 (G) expression under different conditions, as shown in panel A. All data are normalized to β-actin levels and then to control values. n = 5 for each group. *P < 0.05, vs. LV-NC group. Repeated measures one-way analysis of variance (Tukey–Kramer multiple comparisons test) was performed. Bax: Bcl-2-associated X protein; Bcl-2: B cell lymphoma/leukemia-2; COH: chronic ocular hypertension; Ctr: control; eGFP: enhanced green fluorescent protein; G2w: chronic ocular hypertension at 2 weeks; LV-NC: eGFP control lentiviruses.

Kir4.1 and Kir4.1 Tyr⁹Asp overexpression attenuates retinal inflammatory cytokine release

Next we investigated how Kir4.1 and Kir4.1 Tyr⁹Asp overexpression in Müller cells attenuate RGC apoptosis in COH retinas. First, we examined changes in the expression of neuroprotective factors and growth factors, including BDNF, CNTF, and NGF (Chakrabarti et al., 1990; Cao et al., 1997; Taylor et al., 2003; Wilson et al., 2007). In the saline-injected retinas, BDNF, CNTF, and NGF mRNA levels did not show significant changes across the three lentiviral injection groups compared with the controls (Additional Figure 2 ^{(2.9MB, tif)} A, C, and E). In the DHPG-injected retinas, neither Kir4.1 overexpression nor Kir4.1 Tyr⁹Asp overexpression in Müller cells significantly affected BDNF mRNA expression (Additional Figure 2 ^{(2.9MB, tif)} B). An increase in CNTF mRNA levels was observed only in DHPG-injected retinas in rats injected with LV-NC 1 week post-injection (166.2% ± 12.0% of control, n = 6, P < 0.05; Additional Figure 2 ^{(2.9MB, tif)} D). Similarly, the NGF mRNA level in the LV-NC group increased to 191.0% ± 23.5% of the control (n = 6, P < 0.01) and 167.3% ± 20.9% of the control (n = 6, P < 0.05) at 1 and 2 weeks, respectively, after DHPG injection. In contrast, NGF levels did not increase in the retinas overexpressing Kir4.1 or Kir4.1 Tyr⁹Asp after DHPG injection (Additional Figure 2 ^{(2.9MB, tif)} F). These results suggest that the neuroprotective effects of Kir4.1 and Kir4.1 Tyr⁹Asp overexpression in Müller cells may not be mediated through an increase in the production of neurotrophic and growth factors.

Activated glial cells induce RGC apoptosis by releasing pro-inflammatory factors such as IL-1β and TNF-α in glaucomatous retinas (Cheng et al., 2021; Hu et al., 2021; Miao et al., 2023). Therefore, we next investigated changes in IL-1β and TNF-α mRNA levels in activated Müller cells overexpressing Kir4.1 or Kir4.1 Tyr⁹Asp. Lentiviral infection did not affect the basal mRNA levels of IL-1β and TNF-α in normal rats (Figure 6A and C). In the LV-NC retinas, DHPG injection induced a slight increase in IL-1β mRNA expression 1 and 2 weeks post-injection, and this effect was reversed by Kir4.1 or Kir4.1 Tyr⁹Asp overexpression (Figure 6B). Similarly, DHPG injection led to a trend in increased TNF-α mRNA expression in the LV-NC retinas. This upregulation was significantly reduced in the Kir4.1 overexpression group (83.2% ± 11.8% of control, n = 6, P < 0.05 vs. LV-NC) and in the Kir4.1 Tyr⁹Asp overexpression group (85.2% ± 9.0% of control, n = 6, P < 0.05 vs. LV-NC) (Figure 6D).

Figure 6.

Kir4.1 overexpression and Kir4.1 Tyr⁹Asp overexpression alleviate the Müller cell activation–induced increase in inflammatory factor mRNA expression levels.

Bar charts summarizing the relative mRNA levels of IL-1β in normal retinas (A) and in retinas from eyes injected intravitreally with DHPG at 1 and 2 weeks post-injection (B), following infection with the LV-NC, Kir4.1, and Kir4.1 Tyr⁹Asp lentiviruses. (C, D) Bar charts summarizing relative TNF-α mRNA levels in normal retinas (C) and in retinas from eyes injected intravitreally with DHPG at 1 and 2 weeks post-injection (D), infected with the same lentiviruses. #P < 0.05 vs. LV-NC group. Kruskal–Wallis test (Dunn’s multiple comparisons test) was performed. Ctr: Control; DHPG: (S)-3,5-dihydroxyphenylglycine; eGFP: enhanced green fluorescent protein; IL-1β: interleukin-1β; LV-NC: eGFP control lentiviruses; TNF-α: tumor necrosis factor-α.

Müller cell activation in COH-induced glaucoma model retinas induces microglial activation through the ATP/P2X7 receptor pathway, leading to increased production of pro-inflammatory factors (Xue et al., 2016; Hu et al., 2021; Xu et al., 2022; Zhu et al., 2023). Using a co-culture system, we investigated the effects of Kir4.1 and Kir4.1 Tyr⁹Asp overexpression in activated Müller cells on microglial activation and pro-inflammatory factor release. As shown in Figure 7, after co-culturing microglia with LV-NC lentivirus-infected and pre-activated Müller cells (LV-NC + activated Müller cells) for 72 hours, TSPO protein levels in microglia significantly increased to 158.9% ± 13.2% of levels observed in the LV-NC + normal Müller cells group (n = 5, P < 0.05). In contrast, TSPO protein levels in microglia co-cultured with Kir4.1 lentivirus-infected or Kir4.1 Tyr⁹Asp lentivirus-infected and pre-activated Müller cells (Kir4.1 + activated Müller cells and Kir4.1 Tyr⁹Asp + activated Müller cells) were comparable to those seen in the LV-NC + normal Müller cells group (122.7% ± 10.0%, n = 5, P > 0.05; 123.4% ± 9.9%, n = 5, P > 0.05, respectively) (Figure 7A and B). Additionally, we assessed changes in ATP concentrations in Müller cells. The ATP concentration in activated Müller cells increased to 3.63 ± 0.93 μM (n = 6, P < 0.05) compared with the LV-NC + normal Müller cells group (2.10 ± 0.57 μM; Figure 7C). In the Kir4.1 lentivirus-infected and Kir4.1 Tyr⁹Asp lentivirus-infected and pre-activated Müller cells (Kir4.1 + activated Müller cells and Kir4.1 Tyr⁹Asp + activated Müller cells), the ATP concentrations were similar to those in the LV-NC + normal Müller cells group (2.10 ± 0.56 μM, n = 6, P > 0.05 for the Kir4.1 + activated Müller cells group, 1.90 ± 0.44 μM, n = 6, P > 0.05 for the Kir4.1 Tyr⁹Asp + activated Müller cells group) (Figure 7C).

Figure 7.

Kir4.1 overexpression and Kir4.1 Tyr⁹Asp overexpression in Müller cells attenuate ATP release, microglia activation, and release of inflammatory factors.

(A) Representative immunoblots showing changes in TSPO protein levels in microglia co-cultured with normal or activated Müller cells infected with LV-NC, Kir4.1, and Kir4.1 Tyr9Asp lentiviruses. (B) Bar charts summarizing the average densitometric quantification of TSPO bands under different conditions. (C) Bar chart showing changes in ATP secretion by Müller cells under various conditions. n = 5–6. One-way analysis of variance (Tukey–Kramer multiple comparisons test). *P < 0.05, vs. LV-NC group. (D, E) Bar charts showing changes in TNF-α mRNA (D) and protein (E) levels in microglia co-cultured with normal or activated Müller cells infected with LV-NC, Kir4.1, and Kir4.1 Tyr9Asp lentiviruses. (F, G) Bar charts showing changes in IL-1β mRNA (F) and protein (G) levels in microglia co-cultured with normal or activated Müller cells infected with LV-NC, Kir4.1, and Kir4.1 Tyr9Asp lentiviruses. (H, I) Bar charts showing changes in IL-1β mRNA (H) and protein (I) levels in microglia co-cultured with Müller cells in which P2X7/P2X4 receptor signaling was inhibited. n = 5–9. *P < 0.05, vs. LV-NC group. Kruskal–Wallis test (Dunn’s multiple comparisons test) for panels F and H; one-way analysis of variance (Tukey–Kramer multiple comparisons test) for panels D and I; repeated measures one-way analysis of variance (Tukey–Kramer multiple comparisons test) for panels E and G. ATP: Adenosine triphosphate; eGFP: enhanced green fluorescent protein; IL-1β: interleukin-1β; LV-NC: eGFP control lentiviruses; TNF-α: tumor necrosis factor-α; TSPO: translocator protein.

When microglia were co-cultured with LV-NC lentivirus-infected and pre-activated Müller cells (LV-NC + activated Müller cells), there were no significant differences in TNF-α mRNA or protein levels among the different groups (Figure 7D and E). However, IL-1β mRNA and protein levels were significantly increased compared with the LV-NC + normal Müller cells group (n = 5 and 9, P < 0.05; Figure 7F and G). Microglia were then co-cultured with pre-activated Müller cells overexpressing Kir4.1 (Kir4.1 + activated Müller cells) or Kir4.1 Tyr⁹Asp (Kir4.1 Tyr⁹Asp + activated Müller cells). The IL-1β mRNA and protein levels in these groups were comparable to those seen in the LV-NC + normal Müller cells group (Figure 7F and G). Moreover, the changes in IL-1β mRNA and protein levels were reversed when microglia were pre-incubated with the P2X7R antagonist A740003 (1 μM) or the P2X4R antagonist 5-BDBD (5 μM) (Figure 7H and I).

In glaucomatous retinas, activated microglia are a primary source of pro-inflammatory factors. Previous studies have demonstrated that Toll-like receptors play a role in microglial activation and the release of pro-inflammatory factors in the COH retina, a process mediated by the MYD88/IRAK1/TRAF6/NF-κB p65 signaling pathway (Takeda and Akira, 2004; Luo et al., 2010; Weber et al., 2010; Miao et al., 2023). Thus, we investigated whether the MYD88/IRAK1/TRAF6/NF-κB p65 signaling pathway is involved in the effects of Kir4.1 and Kir4.1 Tyr⁹Asp overexpression in Müller cells on microglial release of pro-inflammatory factors. When microglia were co-cultured with pre-activated Müller cells (+ activated Müller cells), MYD88, IRAK1, TRAF6, and NF-κB p65 expression levels increased significantly to (176.2% ± 16.1%, n = 5, P < 0.01; 183.1% ± 26.1%, n = 5, P < 0.05; 143.7% ± 8.2%, n = 5, P < 0.05), and (150.9% ± 11.0%, n = 5, P < 0.05) respectively, of the levels seen in the normal Müller cells group (Figure 8A, B, D, F, and H). Similarly, in primary cultured microglia infected with the LV-NC lentivirus and co-cultured with pre-activated Müller cells (LV-NC + activated Müller cells), the protein levels of these factors were also significantly increased. Notably, this increase was reversed by either Kir4.1 or Kir4.1 Tyr⁹Asp overexpression in Müller cells (Figure 8A, C, E, G, and I). These results suggest that Kir4.1 and Kir4.1 Tyr⁹Asp overexpression attenuate Müller cell activation and reduce the release of ATP. This reduction subsequently suppresses microglial activation by blocking the ATP/P2X pathway and diminishing interactions between Müller cells and microglia. This inhibits MYD88/IRAK1/TRAF6/NF-κB p65 pathway signaling, leading to a decrease in the synthesis and release of inflammatory factors.

Figure 8.

Kir4.1 overexpression and Kir4.1 Tyr⁹Asp overexpression in Müller cells suppress the MYD88/IRAK1/TRAF6/NF-κB p65 inflammatory signaling pathway.

(A) Representative immunoblots showing changes in MYD88, IRAK1, TRAF6, and NF-κB p65 expression levels in microglia co-cultured with normal or activated Müller cells infected with LV-NC, Kir4.1, and Kir4.1 Tyr⁹Asp lentiviruses. (B, D, F, H) Bar charts summarizing the average densitometric quantification of immunoreactive bands for MYD88 (B), IRAK1 (D), TRAF6 (F), and NF-κB p65 (H) in microglia co-cultured with normal or activated Müller cells. n = 5 for each group. *P < 0.05, **P < 0.01, vs. the + normal Müller cells group. Unpaired two-tailed t-test. (C, E, G, I) Bar charts summarizing the average densitometric quantification of immunoreactive bands for MYD88 (C), IRAK1 (E), TRAF6 (G), and NF-κB p65 (I) in microglia co-cultured with normal or activated Müller cells. n = 5 for each group. *P < 0.05, **P < 0.01, vs. LV-NC + normal Müller cells group. One-way analysis of variance (Tukey–Kramer multiple comparisons test) was performed. eGFP: Enhanced green fluorescent protein; IRAK1: IL-1 receptor associated kinase 1; LV-NC: eGFP control lentiviruses; MYD88: myeloid differentiation primary response protein 88; NF-κB P65: nuclear factor kappa B P65; TRAF6: TNF receptor associated factor 6.

Kir4.1 overexpression in Müller cells attenuates microglial proliferation and migration

Activated Müller cells have been reported to possibly induce microglial proliferation and migration to the GCL, ultimately leading to RGC injury in COH retinas (Xu et al., 2022). Next, we investigated whether Kir4.1 and Kir4.1 Tyr⁹Asp overexpression in activated Müller cells affects microglial proliferation and migration. When microglia were co-cultured with pre-activated Müller glial cells (+ activated Müller cells), the number of EdU-positive microglia increased to 305.0% ± 59.0% of the control group (no Müller cells) (n = 5, P < 0.01; Figure 9A and B). In pre-activated Müller cells infected with LV-NC, the number of EdU-positive microglia was 215.3% ± 27.1% of the control group (LV-NC + normal Müller cells) (n = 5, P < 0.05; Figure 9C and D). Additionally, overexpression of Kir4.1 in pre-activated Müller cells reduced the number of EdU-positive microglia to 116.0% ± 19.6% of the control (n = 5, P > 0.05 vs. LV-NC + normal Müller cells, P < 0.05 vs. LV-NC + activated Müller cells; Figures 9C and D). However, infection of pre-activated Müller cells with the Kir4.1 Tyr⁹Asp lentivirus did not significantly impact the number of EdU-positive microglia (198.2% ± 30.9% of the control, n = 5, P < 0.05 vs. LV-NC + normal Müller cells; Figure 9C and D). Furthermore, when normal Müller cells were co-cultured with microglia, the number of migratory microglia increased. A significant increase in the number of migratory microglia was observed when microglia were co-cultured with pre-activated Müller cells (449.6% ± 58.0% of the control [no Müller cells], n = 5, P < 0.001; Figure 10A and B). In pre-activated Müller cells infected with LV-NC, the number of migratory microglia was 149.5% ± 6.1% of the control (LV-NC + normal Müller cells) (n = 5, P < 0.05) (Figure 10C and D). In contrast, Kir4.1 overexpression in pre-activated Müller cells reduced the number of migratory microglia to 80.5% ± 9.2% of the control (LV-NC + normal Müller cells) (n = 5, P > 0.05; Figure 10C and D). However, infection of pre-activated Müller cells with the Kir4.1 Tyr⁹Asp lentivirus did not significantly affect microglial migration (140.6% ± 8.4% of the control, n = 5, P < 0.05; Figure 10C and D). These results suggest that Kir4.1 overexpression in Müller cells may reduce both microglial proliferation and migration, whereas Kir4.1 Tyr⁹Asp overexpression in Müller cells appears to have no significant effect on microglial proliferation and migration.

Figure 9.

Effects of Kir4.1 overexpression and Kir4.1 Tyr⁹Asp overexpression in Müller cells on microglial proliferation.

(A) Representative images showing changes in the number of EdU-positive microglia (indicated by arrowheads) when microglia were co-cultured with an empty Transwell chamber (no Müller cells) (a1 and a2), normal Müller cells (a4 and a5), and activated Müller cells (a7 and a8). Panels a3, a6, and a9 are enlarged images of the areas outlined in white. The number of EdU-positive microglia increased significantly when microglia were co-cultured with activated Müller cells. (B) Bar chart illustrating the changes in the number of EdU-positive microglia under the different conditions shown in panel A. The data were normalized to the no Müller cells group. n = 5 for each group. **P < 0.01, vs. no Müller cells group. One-way analysis of variance followed by the Tukey–Kramer multiple comparisons test was performed. (C) Representative images showing changes in the number of EdU-positive microglia (indicated by arrowheads) when microglia were co-cultured with LV-NC + normal Müller cells (c1 and c2), LV-NC + activated Müller cells (c4 and c5), Kir4.1 + activated Müller cells (c7 and c8), and Kir4.1 Tyr⁹Asp + activated Müller cells (c10 and c11). Panels c3, c6, c9, and c12 are enlarged images of the areas outlined in white. The increased numbers of EdU-positive microglia induced by activated Müller cells were reduced by Kir4.1 overexpression, but not by Kir4.1 Tyr⁹Asp overexpression, in Müller cells. (D) Bar chart showing changes in the number of EdU-positive microglia under the different conditions depicted in panel C. The data were normalized to the LV-NC + normal Müller cells group. n = 5 for each group. *P < 0.05, vs. LV-NC + normal Müller cells group; #P < 0.05, vs. LV-NC + activated Müller cells group. One-way analysis of variance (Tukey–Kramer multiple comparisons test) was performed. Scale bars: 20 μm and 10 μm (for the enlarged panels). EdU: 5-Ethynyl-2′-deoxyuridine; eGFP: enhanced green fluorescent protein; Iba1: ionized calcium-binding adapter molecule 1; LV-NC: eGFP control lentiviruses.

Figure 10.

Effects of Kir4.1 overexpression and Kir4.1 Tyr⁹Asp overexpression in Müller cells on microglial migration.

(A) Representative images showing changes in microglial migration when microglia were co-cultured with an empty Transwell chamber (no Müller cells) (a1), normal Müller cells (a2), and activated Müller cells (a3). The number of migrated microglia increased significantly when co-cultured with activated Müller cells. (B) Bar chart illustrating the average number of migrated microglia under the different conditions depicted in panel A. The data were normalized to the no Müller cells group. n = 5 for each group. ***P < 0.001, vs. the no Müller cells group. One-way analysis of variance followed by the Tukey–Kramer multiple comparisons test was performed. (C) Representative images showing changes in microglial migration when microglia were co-cultured with LV-NC + normal Müller cells (c1), LV-NC + activated Müller cells (c2), Kir4.1 + activated Müller cells (c3), and Kir4.1 Tyr⁹Asp + activated Müller cells (c4). The increase in the number of migrated microglia induced by activated Müller cells was reduced by Kir4.1 overexpression, but not by the Kir4.1 Tyr⁹Asp overexpression, in Müller cells. Scale bars: 20 μm. (D) Bar chart showing the average number of migrated microglia under the different conditions depicted in panel C. The data were normalized to the LV-NC + normal Müller cells group. n = 5 for each group. *P < 0.05, vs. LV-NC + normal Müller cells group. ##P < 0.01, vs. LV-NC + activated Müller cells group. One-way analysis of variance (Turkey–Kramer multiple comparisons test) was performed. eGFP: Enhanced green fluorescent protein; LV-NC: eGFP control lentiviruses.

Discussion

The functional phosphorylation sites in regulating Kir4.1 channels

Decreased Kir4.1 expression in Müller cells is a common characteristic of retinal injuries and diseases, including glaucoma, diabetic retinopathy, macular edema, proliferative retinopathy, and ischemia–reperfusion injury (Bringmann et al., 2006). In glaucomatous retinas in experimental animal models, the reduction in Kir4.1 expression and Kir4.1-mediated currents contributes to membrane depolarization, leading to Müller cell activation. This activation, in turn, induces microglial activation, which exacerbates retinal inflammatory responses and promotes RGC apoptosis (Ji et al., 2012; Gao et al., 2015; Hu et al., 2021; Xu et al., 2022; Miao et al., 2023). Given that phosphorylation mediates downregulation of Kir4.1 expression and Kir current amplitudes (Ji et al., 2012), we identified seven potential phosphorylation sites in Kir4.1 based on its amino acid sequence using prediction tools. These phosphorylation sites included serine, threonine, and tyrosine residues. We mutated each of these putative phosphorylation sites and constructed lentiviruses overexpressing the resulting Kir4.1 mutants to explore their effects on Müller cell activation. In Müller cells overexpressing Kir4.1 or Kir4.1 Tyr⁹Asp, both in vivo and in vitro, mGluR I-induced downregulation of Kir4.1 expression and upregulation of GFAP expression were significantly attenuated. This result suggests that the Tyr⁹ residue of Kir4.1 is a crucial functional phosphorylation site. Overexpression of Kir4.1 or Kir4.1 Tyr⁹ reduced Müller cell activation. Notably, a previous study showed that the Tyr8 and Tyr⁹ residues of Kir4.1 can be phosphorylated in mouse initial distal convoluted tubules. Single-channel recordings demonstrated that the mutation of Tyr⁹ did not alter the channel conductance of the Kir4.1 homotetramer (Zhang et al., 2013).

Notably, in normal rats injected with LV-NC, Kir4.1 and Kir4.1 Tyr⁹Asp overexpression resulted in increased GFAP expression, indicating that subretinal injection may cause some degree of retinal injury and Müller cell activation. In COH retinas, GFAP expression was further elevated in Müller cells infected with the LV-NC lentivirus; however, it did not increase further in Müller cells infected with the Kir4.1 or Kir4.1 Tyr⁹Asp lentivirus. These findings demonstrate that both Kir4.1 and Kir4.1 Tyr⁹Asp overexpression effectively suppressed Müller cell activation. Additionally, in Müller cells overexpressing Kir4.1 Ser³⁷⁰Arg, Kir4.1 expression increased after DHPG treatment. Future studies should investigate the effect of mutating multiple Kir4.1 phosphorylation sites to gain a more detailed understanding of the transcriptional regulation mechanisms involved.

In our previous study, we identified K⁺-selective Kir currents in retinal Müller cells using patch-clamp techniques and found that these currents were primarily mediated by Kir4.1 (Ji et al., 2012). The changes in Kir currents observed in Müller cells from glaucomatous retinas correlated with alterations in Kir4.1 protein levels as assessed by Western blot, indicating that the decrease in Kir currents in Müller cells is due to reduced Kir4.1 expression. In this study, we found that overexpression of Kir4.1 or Kir4.1 Tyr⁹Asp significantly mitigated the reduction in total Kir4.1 expression levels in cell membranes in COH retinas. Furthermore, both Kir4.1 overexpression and Kir4.1 Tyr⁹Asp overexpression reduced Müller cell activation and decreased the release of retinal inflammatory factors. These results suggest that Kir4.1 and Kir4.1 Tyr⁹Asp overexpression indeed altered channel function. Future studies will include additional electrophysiological recordings to verify these changes in channel function.

Kir4.1 overexpression and Kir4.1 Tyr⁹Asp overexpression attenuate RGC apoptosis by suppressing the retinal inflammatory response

Activated glial cells may contribute to RGC injury in glaucoma (Xue et al., 2016; Xu et al., 2020; Hu et al., 2021; Miao et al., 2023). One of the main mechanisms involves retinal neuroinflammation–mediated RGC apoptosis (Miao et al., 2023; Sulak et al., 2024). In this study, we demonstrated that Kir4.1 overexpression and Kir4.1 Tyr⁹Asp overexpression in Müller cells inhibited the release of pro-inflammatory factors through two main pathways. First, these alterations inhibited Müller cell activation, thereby reducing the expression of pro-inflammatory factors. Indeed, an increase in the expression and release of pro-inflammatory factors has been observed in activated Müller cells in experimental glaucoma (Miao et al., 2023). Second, Kir4.1 and Kir4.1 Tyr⁹Asp overexpression attenuated microglial activation. Using a microglia–Müller cell co-culture system, we observed that these interventions reduced ATP production in activated Müller cells, leading to diminished microglial activation, as indicated by reduced TSPO expression. Our previous study showed that, in COH retinas, activated Müller cells release ATP through connexin 43 hemichannels, which activates microglia via P2X7/P2X4 receptors. Activated microglia can alter Müller cell functions, and the interactions between microglia and Müller cells exacerbate the retinal inflammatory response in a positive feedback loop in glaucoma (Hu et al., 2021; Xu et al., 2022). Activated microglia are a significant source of pro-inflammatory factors in glaucoma (Miao et al., 2023). Kir4.1 and Kir4.1 Tyr⁹Asp overexpression in Müller cells attenuated ATP/P2X receptor signaling and microglial activation, reducing the release of pro-inflammatory factors. Notably, when P2X7/P2X4 receptors in microglia were blocked, Kir4.1 overexpression and Kir4.1 Tyr⁹Asp overexpression no longer changed IL-1β mRNA expression levels in microglia. Additionally, the Toll-like receptor pathway may play a role in microglial activation and the release of pro-inflammatory factors in COH retinas (Luo et al., 2010; Miao et al., 2023). The expression of these molecules increased in microglia co-cultured with activated Müller cells, suggesting that Toll-like receptor signaling was modified, thereby affecting the synthesis and release of pro-inflammatory factors. Notably, Kir4.1 and Kir4.1 Tyr⁹Asp overexpression in Müller cells downregulated Toll-like receptor protein levels, indicating a direct relationship with the reduction in the synthesis and release of pro-inflammatory factors from microglia. A question that remains unanswered is whether the ATP/P2X receptor signaling pathway or other pathways mediate inhibition of the Toll-like receptor pathway, which we will explore in future studies. These findings suggest that Kir4.1 and Kir4.1 Tyr⁹Asp overexpression suppress Müller cell activation and attenuate interactions between Müller cells and microglia by blocking the ATP/P2X pathway. This ultimately reduces the synthesis and release of pro-inflammatory factors, thereby ameliorating RGC damage in the glaucomatous retina. Furthermore, while retinal glial cells can provide neuroprotective effects through the production of neurotrophic and growth factors, our results indicate that Kir4.1 and Kir4.1 Tyr⁹Asp overexpression in Müller cells did not significantly affect BDNF, CNTF, or NGF mRNA levels. This result suggests that the neuroprotective effects associated with Kir4.1 and Kir4.1 Tyr⁹Asp overexpression in Müller cells may not be mediated by an increase in the production of these neurotrophic factors and growth factors.

Microglia, the primary immune cells in the retina, are an essential component of the resident glial population, maintaining retinal homeostasis through constant immune surveillance, synaptic pruning, and the cleanup of cellular debris (Langmann, 2007; Guo et al., 2022; Murenu et al., 2022; Wang and Cepko, 2022). Under physiological conditions, microglia are primarily located in the plexiform layers of the retina. However, in response to retinal injuries and diseases, they can quickly transition to an activated state, proliferate, and migrate to the sites of injury. In a COH-induced experimental model of glaucoma, activated Müller cells in the retina induce microglial proliferation and migration to the GCL via the ATP/P2X receptor pathway. The accumulated microglia in the GCL release pro-inflammatory factors, exacerbating RGC impairment (Hu et al., 2021; Xu et al., 2022). In this study, we found that, when microglia were co-cultured with pre-activated Müller cells infected with a lentivirus overexpressing Kir4.1, the numbers of proliferating and migrating microglia were reduced. This reduction may be related to decreased ATP secretion from Müller cells. However, microglial proliferation and migration were not affected by co-culturing with activated Müller cells infected with a lentivirus overexpressing Kir4.1 Tyr⁹Asp, despite the reduction in ATP secretion. This discrepancy suggests that the functional states of activated Müller cells may differ between those overexpressing Kir4.1 and those overexpressing Kir4.1 Tyr⁹Asp. In addition to ATP, the release of chemokines and adhesion molecules may vary under these two conditions. The detailed and precise mechanisms underlying these effects warrant further exploration in future studies.

RGC apoptotic death is a fundamental cause of blindness in glaucoma. In this study, we provide evidence that Kir4.1 and Kir4.1 Tyr⁹Asp overexpression in Müller cells reduce RGC apoptosis in COH retinas. This reduction may result from decreased glial activation and the release of pro-inflammatory factors. Inflammatory factors such as TNF-α and IL-1β can induce RGC apoptosis by activating cell death signaling pathways, modulating ion channels, and increasing cell excitability (Cheng et al., 2021; Miao et al., 2023; Wang et al., 2023). The balance between Bax and Bcl-2 plays a crucial role in cell apoptosis. Activated Bax, a pro-apoptotic protein, increases the permeability of mitochondrial membranes and promotes the release of cytochrome c, which in turn activates caspase-3 by cleaving it into its active form, cleaved-caspase 3—an indicator of apoptosis (Mootha et al., 2001; Wei et al., 2001; Walters et al., 2009). Conversely, Bcl-2, an anti-apoptotic protein, is vital for maintaining mitochondrial function (Granville and Gottlieb, 2002; Wang et al., 2013). We demonstrated that IOP elevation increases Bax and cleaved-caspase 3 expression, and that this effect is reversed by Kir4.1 and Kir4.1 Tyr⁹Asp overexpression in Müller cells. Furthermore, although Bcl-2 levels in COH retinas did not show significant changes compared with control, both Kir4.1 overexpression and Kir4.1 Tyr⁹Asp overexpression in Müller cells increased Bcl-2 levels, thereby maintaining the Bax/Bcl-2 balance and exerting anti-apoptotic effects. These findings suggest that regulating Müller cells may be a promising strategy for ameliorating RGC injury in glaucoma retinas.

The present study had some limitations. Owing to the constraints of animal models and experimental methods, some experiments were conducted using primary cultured glial cells. The results obtained from cultured cells may differ from those observed in COH retinas to some extent, and will need to be validated in future studies. Furthermore, as mentioned above, the changes in Kir4.1 function in Müller cells resulting from Kir4.1 overexpression and Kir4.1 Tyr⁹Asp overexpression should be verified using electrophysiological recordings in future investigations.

In conclusion, our findings suggest that phosphorylation of the Tyr⁹ residue of Kir4.1 may induce downregulation of Kir4.1 expression and activation of Müller cells in an experimental model of glaucoma. Kir4.1 overexpression and Kir4.1 Tyr⁹Asp overexpression attenuated Müller cell activation, subsequently inhibiting microglial activation, proliferation, and migration through the ATP/P2X receptor pathway. This, in turn, reduced the release of pro-inflammatory factors, thereby decreasing RGC apoptosis in glaucoma. These findings suggest a novel approach to treating glaucoma by modulating Müller cells to inhibit retinal inflammation and reduce RGC injury.

Open peer reviewer: Min Ji, Nantong University, China.

Additional files:

Additional Table 1: Gene mutated sites and corresponding amino acid mutated sites of Kir4.1.

Additional Table 2: Titers and MOI value of each lentivirus.

Additional Table 3: Primer sequences.

Additional Figure 1 ^{(2.6MB, tif)} : Cellular localization of lentiviral infections in the retina.

Additional Figure 1

Cellular localization of lentiviral infections in the retina.

(A) Schematic diagram illustrating the construction of Kir4.1 overexpression and Kir4.1 Tyr⁹Asp mutation lentiviral expression system using the glial cell-specific promoter RLBP1. (B-D) Double fluorescent immunostaining showing co-localization of GFP and GS, a Müller cell marker, in retinas injected with LV-NC (B), Kir4.1 overexpression (C), and Kir4.1 Tyr⁹Asp mutation (D) lentiviruses. The retinal Müller cells were successfully infected by the lentiviruses carrying the RLBP1 promoter. Scale bar: 20 μm. DAPI: 4',6-Diamidino-2-phenylindole; eGFP: enhanced green fluorescent protein; GCL: ganglion cell layer; GS: glutamine synthetase; INL: inner nuclear layer; IPL: inner plexiform layer; LV-NC: eGFP control lentiviruses; ONL: outer nuclear layer; OPL: outer plexiform layer; RLBP1: retinaldehyde binding protein 1; SV40: simian vacuolating virus 40.

Additional Figure 2 ^{(2.9MB, tif)} : DHPG-induced changes in BDNF, CNTF, and NGF mRNA levels in the retina with Kir4.1 overexpression and Kir4.1 Tyr⁹Asp mutation in Müller glial cells.

Additional Figure 2

DHPG-induced changes in BDNF, CNTF, and NGF mRNA levels in the retina with Kir4.1 overexpression and Kir4.1 Tyr⁹Asp mutation in Müller glial cells.

(A, B) Bar charts showing changes in BDNF mRNA levels in normal retinas (A) and DHPG-injected retinas at 1 and 2 weeks post-injection (B) following infections with LV-NC, Kir4.1 overexpression, and Kir4.1 Tyr⁹Asp mutation lentiviruses. (C, D) Bar charts illustrating changes in CNTF mRNA levels in normal retinas (C) and DHPG-injected retinas at 1 and 2 weeks post-injection (D) following infections with LV-NC, Kir4.1 overexpression, and Kir4.1 Tyr⁹Asp mutation lentiviruses. (E, F) Bar charts displaying changes in NGF mRNA levels in normal retinas (E) and DHPG-injected retinas at 1 and 2 weeks post-injection (F) following infections with LV-NC, Kir4.1 overexpression, and Kir4.1 Tyr9Asp mutation lentiviruses. n = 6 for each group.^*P <0.05, ^**P < 0.01, vs. Ctr; #P < 0.05, vs. LV-NCgroup. Kruskal Wallis test (Dunn's multiple comparisons test) was performed. BDNF: Brain-derived neurotrophic factor; Ctr: control; CNTF: ciliary neurotrophic factor; DHPG: (S)-3,5-dihydroxyphenylglycine;eGFP: enhanced green fluorescent protein; LV-NC: eGFP control lentiviruses; NGF: nerve growth factor.

Additional file 1: Open peer review report 1 ^{(79.5KB, pdf)} .

Funding Statement

Funding: This study was supported by the National Natural Science Foundation of China, Nos. 32271043 (to ZW) and 82171047 (to YM); the both Science and Technology Major Project of Shanghai, No. 2018SHZDZX01 and ZJLab, and Shanghai Center for Brain Science and Brain-Inspired Technology (to ZW).

Data availability statement:

All data relevant to the study are included in the article or uploaded as Additional files.