Primary mouse Müller cells (MIC-iCELL-m018) were purchased from iCell Bioscience Inc., Shanghai) [25,26]. Their identity was validated by surface marker analysis with immunofluorescence staining (see Supplementary Materials for details). All cells tested negative for mycoplasma. Cells were maintained in DMEM/F12 medium (Boster Biological Technology, Wuhan, China; Catalog#DZPYG0074) containing 10% fetal bovine serum and 1% penicillin–streptomycin. This medium contains 17.17 mg/L (≈0.7 mmol/L) Mg²⁺ and 3151 mg/L (≈17.5 mmol/L) glucose. For the high glucose (HG) condition, glucose was added to DMEM/F12 medium to give a final concentration of 25 mmol/L [27]. For the HG and low Mg²⁺ (LM) condition, a customized Mg²⁺-free DMEM/F12 medium was used (Boster Biological Technology, Wuhan, China; Catalog#DZPYG0313). Glucose and MgSO₄ were added to this medium to achieve final glucose and Mg²⁺concentrations of 25 mmol/L and 0.5 mmol/L, respectively [28].

Oligonucleotides encoding murine TRPM7-specific shRNA (CCTTATCAAACG) or scrambled non-effective shRNA were cloned into the pAAV-U6-CMV-Kozak-EGFP vector (Cyagen, Guangzhou, China) to produce recombinant AAV vectors in which green fluorescent protein (GFP) functioned as the expression reporter. Cells were transduced with AAVs at concentrations between 1.8 and 3 × 10⁷ PFUs/mL. Following a 48-h transduction period, the cells were utilized in subsequent experiments.

Total RNA was extracted from Müller cells using the SteadyPure Quick RNA Extraction kit (Accurate Biotechnology, Changsha, China). This was subsequently reverse transcribed into complementary DNA using a Reverse Transcription kit as recommended by the manufacturer. Quantitative real-time PCR (qPCR) was carried out with the SYBR Green Premix Pro Taq HS qPCR kit (Accurate Biotechnology) and the QuantStudio™ 5 Real-Time PCR System (Thermo, MA, USA). The primer sequences for qPCR were: TRPM7, forward (F): 5′-CAGCTTTGTTACCGGATTGGTT-3′; reverse (R): 5′-GTGGAGGTACAGGAACGAAGG-3′; VEGF, F: 5′-CTGCTGTAACGATGAAGCCCTG-3′; R: 5′-GCTGTAGGAAGCTCATCTCTCC-3′; VDAC-1, F: 5′-TGTCGCCAAATGCAACTGTG-3′; R: 5′-AAGGTGAGCTTCAGTCCACG-3′; GAPDH, F: 5′-CAATGACCCCTTCATTGACC-3′; R: 5′-GACAAGCTTCCCGTTCTCAG-3′. The 2^-ΔΔCt method was used to calculate the relative expression level of various genes.

RIPA lysis buffer was used to extract total protein from Müller cells, as recommended by the manufacturer (Epizyme Biotech, Shanghai, China). A BCA protein assay kit was used to determine the protein concentration (Epizyme Biotech). Proteins were separated using SDS-PAGE and then transferred to PVDF membranes. QuickBlock™ Western blocking solution (Beyotime Institute of Biotechnology, Shanghai, China) was used to block the membranes for 15 min at 37 °C, after which they were incubated for 16 h at 4 °C with the following anti-mouse primary antibodies: TRPM7 (Cat#: ab245408, 1:5000, Abcam, Cambridge, UK), VEGF (Cat#: ab32152, 1:5000, Abcam, Cambridge, UK), voltage-dependent anion channel 1 (VDAC1) (Cat#: AF1027, 1:1000, Beyotime, China) and β-actin (Cat#: ab8227, 1:5000, Abcam, Cambridge, UK). The membranes were then incubated for 1 h at room temperature with corresponding secondary antibody, and the protein bands visualized with a chemiluminescence reagent (Beyotime, China).

The Cell Counting Kit-8 (CCK-8) assay was used to assess cell viability, as recommended by the manufacturer (Beyotime, China). Cells cultured in 96-well plates (1000 cells/well) were incubated at 37 °C for 2 h with CCK-8 solution (10 μL/well), followed by measurement of the absorbance at 450 nm using a microplate reader (Biotek Instruments Inc., USA).

For the analysis of apoptosis, cells were grown in 6-well plates and the culture medium containing floating cells was collected into centrifuge tubes. Adherent cells were washed once with phosphate-buffered saline (PBS), and the wash buffer was combined with the previously collected medium. Cells were detached by adding a 0.25% trypsin solution without EDTA to each well, causing them to round up. Trypsinization was stopped by adding complete culture medium. Cells were gently pipetted to ensure a single-cell suspension and then added to the same centrifuge tubes containing the original medium and washes. The cell suspension was centrifuged for 5 min at 4 °C at a speed of 300–400 ×g. After carefully discarding the supernatant, the cell pellet was resuspended in cold PBS and the cells were washed twice by repeating the centrifugation step.

The Annexin V-fluorescein isothiocyanate (FITC) apoptosis kit (Beyotime, China) was subsequently used to evaluate apoptosis in the Müller cells by flow cytometry. Briefly, cells were resuspended in Annexin V binding buffer (1×) at a density of 1 × 10⁶ cells/mL. One hundred µL of this cell suspension was then stained by incubating with FITC Annexin V (5 µL) and propidium iodide (5 µL) in the dark for 15 minutes at room temperature. After incubation, 400 µL of binding buffer (1×) was added and the samples analyzed with a BD Accuri™ C6 Plus Cytometer (BD Biosciences, USA).

Measurement of MMP was performed using an enhanced MMP assay kit with JC-1 (Beyotime, China) and flow cytometry. Following collection and resuspension in PBS, cells were incubated for 20 min with JC-1 solution at 37 °C in 5% CO₂, as recommended by the manufacturer. Subsequently, a microplate reader (Biotek Instruments Inc., USA) was used to measure fluorescence at 527 nm (JC-1 monomers) and 590 nm (JC-1 aggregates). The ratio of the fluorescence intensity of aggregates to that of monomers indicated changes in MMP, with a decreasing ratio indicating mitochondrial depolarization.

Opening of the mPTP was examined by cobalt quenching of Calcein-(acetoxymethyl ester) AM fluorescence with an mPTP Assay Kit (Beyotime, China). This green fluorescent dye accumulates in the mitochondria of living cells. When the mPTP opens, cytosolic Co²⁺ enters the mitochondria and forms a complex with Calcein-AM, causing the mitochondrial green fluorescent to diminish or disappear. In brief, cells were collected, resuspended in PBS, then incubated in the dark with Calcein-AM and CoCl₂ for 30 min at 37 °C, as recommended by the manufacturer. After washing 2–3 times with PBS, cell fluorescence was detected with a laser scanning confocal microscope (40×, Leica, Germany) and an Olympus FluoView FV1000 Confocal Microscope. The analysis of fluorescence was conducted with ImageJ software version 1.53i (https://imagej.nih.gov/ij/; Bethesda, USA).

Fura-2 AM is a membrane-permeable, Ca²⁺-sensitive fluorescent dye (Thermo Fisher Scientific, USA). It was used in this study to measure intracellular Ca²⁺ levels in Müller cells using a microplate reader (Tecan Infinite M200®, Switzerland), as previously reported [29]. Therefore, intracellular Ca²⁺ levels in individual wells are determined by the fluorescence emission intensity at 510 nm at alternating excitation wavelengths of 340 and 380 nm. In brief, cells were seeded into 96-well culture plates at a density of 3.0 × 10⁴ cells/well. After washing with PBS, they were loaded with dye solution and incubated in the dark for 1 h at room temperature. The fluorescence emission intensity at 510 nm at alternating excitation wavelengths of 340 and 380 nm was then determined for individual wells in a microplate reader with Tecan i-Control software (Männedorf, Switzerland). The ratio of the 510 nm emission in response to 340/380 nm excitation was calculated and normalized to the control.

The intracellular Adenosine triphosphate (ATP) level in Müller cells was measured with an enhanced ATP Assay Kit (Beyotime, China) as recommended by the manufacturer. In brief, cells were first lysed with lysis buffer and the supernatant collected. Standard working solutions of ATP were made by serial dilutions of the ATP stock solution. The samples and ATP standard working solutions were subsequently added to individual wells, and the bioluminescence measured utilizing a microplate reader (Biotek Instruments Inc., USA) in luminometer mode. The ATP concentration was then calculated based on an ATP standard curve.

The intracellular ROS level in Müller cells was measured with a Reactive Oxygen Species kit (Beyotime, China). Briefly, cells were incubated in the dark with 10 μM DCFH-DA for 20 min at 37 °C, then washed with PBS. The fluorescence intensity, representing the intracellular ROS level, was determined with an Olympus IX 73 microscope (Olympus, Japan).

The levels of cellular malondialdehyde (MDA), superoxide dismutase (SOD), and glutathione (GSH), as well as the nicotinamide adenine dinucleotide phosphate (NADP⁺)/nicotinamide adenine dinucleotide hydrogen (NADPH) ratio, were evaluated using corresponding assay kits for MDA, SOD, GSH, and NADP⁺/NADPH, as recommended by the manufacturer (Beyotime, China). Measurements were obtained with a microplate reader (Biotek Instruments Inc., USA), and corresponding standard curves were used to calculate the concentration.

Measurement data were analyzed by two-way ANOVA and SPSS 25 statistical analysis software (IBM Corp. Armonk, New York, USA). The least significant difference method (LSD) was used to compare homogeneity of variance, and a non-parametric test to determine non-uniformity of variance. Data are presented as the mean ± standard error of the mean (SEM). Results with a p-value of <0.05 were considered significantly different, and those with a p-value of <0.01 as highly significant.

We first evaluated the effects of low Mg²⁺ on cell viability and VEGF expression in Müller cells under HG conditions. LM/HG treatment for 48 h was found to markedly reduce cell viability and increase apoptosis compared with HG treatment alone (Fig. 1A–C). WB and PCR analyses revealed that both VEGF mRNA and protein levels were significantly increased by LM/HG treatment compared with HG treatment alone (Fig. 1D–F). These data demonstrate that low Mg²⁺ exacerbates injury to Müller cells under HG stress.

Fig. 1.

Low Mg²⁺ exacerbates the injury of Müller cells under HG stress. Müller cells were treated for 48 h with HG (25 mM), with or without LM (0.5 mM). (A) The CCK-8 assay was used to evaluate cell viability, expressed as the percentage of optical density relative to control cells (100%) (n = 5). (B) Flow cytometry: representative image of dot plot analysis. Flow cytometry was used to analyze cell apoptosis with the Annexin V-FITC/PI assay. (C) Histogram showing the percentage of apoptotic cells (n = 5 independent experiments). (D) VEGF gene expression level as determined by qPCR (n = 5). (E,F) VEGF protein expression level as determined by WB analysis (n = 3). Relative gene and protein expression levels were normalized to the level of β-actin. Con, cells cultured in basal media containing normal glucose and Mg²⁺; HG, cells cultured in media containing high glucose; HG/LM, cells cultured in media containing high glucose and low Mg²⁺; PI, propidium iodide; FITC, fluorescein isothiocyanate; VEGF, vascular endothelial growth factor; qPCR, quantitative real-time PCR; SEM, standard error of the mean. Data are presented as mean ± SEM. *p < 0.05, **p < 0.01.

Oxidative stress is a major pathogenic mechanism through which low Mg²⁺ disrupts cellular function, and also plays a central role in DR [30,31]. Therefore, we evaluated the impact of low Mg²⁺ on oxidative stress in Müller cells exposed to HG. LM/HG treatment for 48 h induced more severe oxidative damage, as manifested by increased ROS compared with HG treatment alone (Fig. 2A,B). Moreover, LM/HG reduced SOD activity (Fig. 2C) and GSH (Fig. 2D), increased MDA production (Fig. 2E), and increased the NADP⁺/NADPH ratio (Fig. 2F). These findings indicate that low Mg²⁺ amplifies oxidative stress, thereby compromising the function of Müller cells under HG conditions.

Fig. 2.

Low Mg²⁺ aggravates oxidative stress in Müller cells under HG stress. Müller cells were treated for 48 h with HG (25 mM), with or without LM (0.5 mM). (A) Representative fluorescent images of ROS. (B) Intracellular ROS levels were measured by immunofluorescence intensity and then normalized to control cells (n = 6). (C–F) SOD, GSH, and MDA levels, as well as the NADP⁺/NADPH ratio, were evaluated using corresponding assay kits (n = 4–5). Data are presented as the mean ± SEM; **p < 0.01, *p < 0.05. Scale bars: 200 μm. ROS, reactive oxygen species; SOD, superoxide dismutase; GSH, glutathione; MDA, malondialdehyde; NADP+/NADPH, nicotinamide adenine dinucleotide phosphate/nicotinamide adenine dinucleotide hydrogen.

TRPM7-mediated Ca²⁺ signaling is a key factor in oxidative stress-dependent cell death [32,33]. Ca²⁺-induced up-regulation of VDAC1 expression is also associated with the induction of cell death [34]. We next examined whether low Mg²⁺ worsens HG toxicity in Müller cells through TRPM7-mediated Ca²⁺ signaling, which subsequently induces VDAC1-mediated mitochondrial dysfunction. We therefore evaluated TRPM7 expression and the intracellular Ca²⁺ level, as well as the mitochondrial function markers of intracellular ATP, mPTP opening, MMP, and mitochondrial VDAC1. Compared with HG treatment alone, LM/HG treatment for 48 h markedly upregulated TRPM7 gene and protein expression (Fig. 3A–C), enhanced mPTP opening (Fig. 3D), and increased VDAC1 gene and protein expression (Fig. 3E–G). Moreover, LM/HG treatment resulted in more severe mitochondrial dysfunction, as evidenced by reduced MMP (Fig. 3H), decreased ATP production (Fig. 3I), and increased intracellular Ca²⁺ levels (Fig. 3J). To determine whether TRPM7 upregulation connects Ca²⁺ overload to mitochondrial injury, we transduced Müller cells with TRPM7-specific shRNA and validated the knockdown efficiency (Fig. 4A–C). Under LM/HG conditions, the silencing of TRPM7 largely preserved mitochondrial function (Fig. 4D–I) and significantly attenuated the increase in intracellular Ca²⁺ level (Fig. 4J). These data indicate that low Mg²⁺ activates the TRPM7/Ca²⁺/VDAC1 axis, leading to mitochondrial dysfunction in Müller cells during HG stress.

Fig. 3.

Low Mg²⁺ upregulates the TRPM7/Ca²⁺/VDAC1 axis, leading to mitochondrial dysfunction in Müller cells under HG stress. Müller cells were treated for 48 h with HG (25 mM), with or without LM (0.5 mM). (A) The TRPM7 gene expression level was quantified using qPCR (n = 6). (B,C) The TRPM7 protein expression level was determined using WB analysis (n = 4). (D) Opening of mPTP was determined by cobalt quenching of Calcein-AM fluorescence using an mPTP Assay Kit. (E) The VDAC1 gene expression level was determined by qPCR (n = 5). (F,G) The VDAC1 protein expression level was determined using WB analysis (n = 4). (H) Relative levels of MMP were quantified by measuring JC-1 fluorescence intensity and then normalizing to the control cells (n = 5). (I) The intracellular levels of ATP were measured with an ATP kit (n = 5). (J) Relative intracellular Ca2+ levels were quantified by measuring Fura-2 AM fluorescence intensity and then normalizing to the control cells (n = 5). The relative levels of gene and protein expression were normalized to β-actin. Data are presented as the mean ± SEM, **p < 0.01, *p < 0.05. Scale bars: 50 μm. TRPM7, Transient Receptor Potential Cation Channel, Subfamily M, Member 7; VDAC1, voltage-dependent anion channel 1; ATP, Adenosine triphosphate; mPTP, Mitochondrial Permeability Transition Pore; WB, Western Blot; MMP, Mitochondrial Membrane Potential; AM, acetoxymethyl ester.

Fig. 4.

The effects of TRPM7 silencing on intracellular Ca²⁺ levels and mitochondrial function in Müller cells under HG and LM stress. Control Müller cells, or Müller cells transfected with TRPM7-specific shRNA were treated for 48 h with HG (25 mM), with or without LM (0.5 mM). (A) The TRPM7 gene expression level was determined by qPCR (n = 5). (B,C) The TRPM7 protein expression level was determined by WB analysis (n = 4). (D) Opening of mPTP was determined by cobalt quenching of Calcein-AM fluorescence using the mPTP Assay Kit. (E–G) VDAC1 gene and protein expression levels were evaluated using qPCR and WB analysis, respectively (n = 4–5). (H) Relative levels of MMP were quantified by measuring JC-1 fluorescence intensity and then normalizing to the control cells (n = 5). (I) The intracellular levels of ATP were measured with an ATP kit (n = 5). (J) Relative intracellular Ca2+ levels were quantified by measuring Fura-2 AM fluorescence intensity and then normalizing to the control cells (n = 5). Relative levels of gene and protein expression were normalized to β-actin. Data are presented as mean ± SEM. **p < 0.01, *p < 0.05. Scale bars: 50 μm.

Finally, we examined whether TRPM7 knockdown protects Müller cells from oxidative stress and functional impairment under LM/HG conditions. After LM/HG treatment for 48 h, TRPM7-silenced cells exhibited significantly lower oxidative-stress indices (Fig. 5A–F), a pronounced rescue from apoptosis (Fig. 6A,B), and a marked reduction in VEGF expression (Fig. 6C–E) compared with the control. These results demonstrate that TRPM7 mediates both the oxidative stress and dysfunction observed in Müller cells under HG/LM stress.

Fig. 5.

The effects of TRPM7 silencing on oxidative stress in Müller cells under HG and LM Stress. Control Müller cells, or Müller cells transfected with TRPM7-specific shRNA were treated for 48 h with HG (25 mM), with or without LM (0.5 mM). (A) Representative fluorescent images of ROS. (B) Intracellular ROS levels were measured by immunofluorescence intensity and then normalized to control cells (n = 5). (C–F) SOD, GSH, and MDA levels, as well as the NADP⁺/NADPH ratio, were measured with the corresponding assay kit (n = 4–5). Data are presented as mean ± SEM, **p < 0.01, *p < 0.05. Scale bars: 200 μm.

Fig. 6.

The effects of TRPM7 silencing on function and apoptosis levels in Müller cells under HG and LM Stress. Control Müller cells or Müller cells transfected with TRPM7-specific shRNA were treated for 48 h with HG (25 mM), with or without LM (0.5 mM). (A) Flow cytometry: representative image of dot plot analysis. Flow cytometry was used to analyze cellular apoptosis with Annexin V-FITC/PI staining. (B) Histogram showing the percentage of cell apoptosis (n = 4). (C) The VEGF gene expression level was determined by qPCR (n = 5). (D,E) The VEGF protein expression level was determined by WB analysis (n = 4). Relative levels of gene and protein expression were normalized to β-actin. Data are presented as mean ± SEM. **p < 0.01.