Induction of necrotic cell death by oxidative stress in retinal pigment epithelial cells

Abstract

Age-related macular degeneration (AMD) is a degenerative disease of the retina and the leading cause of blindness in the elderly. Retinal pigment epithelial (RPE) cell death and the resultant photoreceptor apoptosis are characteristic of late-stage dry AMD, especially geographic atrophy (GA). Although oxidative stress and inflammation have been associated with GA, the nature and underlying mechanism for RPE cell death remains controversial, which hinders the development of targeted therapy for dry AMD. The purpose of this study is to systematically dissect the mechanism of RPE cell death induced by oxidative stress. Our results show that characteristic features of apoptosis, including DNA fragmentation, caspase 3 activation, chromatin condensation and apoptotic body formation, were not observed during RPE cell death induced by either hydrogen peroxide or tert-Butyl hydroperoxide. Instead, this kind of cell death can be prevented by RIP kinase inhibitors necrostatins but not caspase inhibitor z-VAD, suggesting necrotic feature of RPE cell death. Moreover, ATP depletion, receptor interacting protein kinase 3 (RIPK3) aggregation, nuclear and plasma membrane leakage and breakdown, which are the cardinal features of necrosis, were observed in RPE cells upon oxidative stress. Silencing of RIPK3, a key protein in necrosis, largely prevented oxidative stress-induced RPE death. The necrotic nature of RPE death is consistent with the release of nuclear protein high mobility group protein B1 into the cytoplasm and cell medium, which induces the expression of inflammatory gene TNFα in healthy RPE and THP-1 cells. Interestingly, features of pyroptosis or autophagy were not observed in oxidative stress-treated RPE cells. Our results unequivocally show that necrosis, but not apoptosis, is a major type of cell death in RPE cells in response to oxidative stress. This suggests that preventing oxidative stress-induced necrotic RPE death may be a viable approach for late-stage dry AMD.

Figure 1

Lack of apoptotic hallmarks in ARPE-19 cells subjected to oxidative stress. (a) Light microscopy pictures of MTT crystals in ARPE-19 cells showing decrease in cell number and viability of ARPE-19 cells treated with 300 or 500 μM of H₂O₂, or 150 μM of tBHP. Test was performed at 24 h after inducing oxidative stress. (b) Analyses of DNA fragmentation in ARPE-19 cells treated with 300 or 500 μM of H₂O₂, or 150 μM of tBHP. DNA from UV-irradiated Hela cells was used as a positive control for apoptosis. DNA was purified from both dead and live cells 24 h after inducing cell death and analyzed by gel electrophoresis. Partial DNA fragmentation in ARPE-19 cells is visible as a smear in the agarose gel. (c) Western blotting detection of cleaved caspase-3 and PARP products as a result of cell death. 25 μg of total cell extract was prepared from both dead and live cells 24 h after subjecting cells to oxidative stress (ARPE-19) or UV irradiation (Hela). GAPDH served as loading control. (d) Western blot measurement of DFF45 level in control ARPE-19 cells and cells subjected to oxidative stress. Of note, the DFF antibody recognizes both DFF45 and DFF35. Hela cells were used as reference. GAPDH served as loading control. (e) Analyses of ATP levels in ARPE cells subjected to oxidative stress. ATP level was measured by recording luminescence in indicated time points after treating cells with 300 or 500 μM of H₂O₂, or 150 μM of tBHP. UV-irradiated Hela cells were used as a control for apoptosis. *P<0.05; ***P<0.001

Figure 2

Necrostatins but not z-VAD can rescue ARPE-19 from cell death induced by oxidative stress. (a) ARPE-19 cells were treated with 33 μM of z-VAD for 24 h before inducing oxidative stress with different concentrations of H₂O₂ or tBHP. Cell viability was measured by MTT assay at 24 h after induction of oxidative stress. (b) MTT assay after ARPE-19 cells were treated with 33 μM necrostains 1, 5 or 7 for 24 h before inducing oxidative stress as described above. *P<0.05; **P<0.01; ***P<0.001

Figure 3

Membrane changes during RPE cell death by oxidative stress. (A) Cell membrane changes in non-fixed ARPE-19 cells in the indicated conditions shown by staining with PI (a–d) and CellMask Orange Plasma Stain (e–h). Cell nucleus was stained with DAPI. Arrows marked the membrane blebs formed in the treated cells. Scale bar equals 25 μm. (B) Mitochondrial network and nuclear envelope permeability were tracked by YFP (green) and RFP (red) signal after ARPE-19 cells were transfected with HMGB1-YFP and ANT1-RFP expression plasmids and treated H₂O₂, or tBHP for indicated time. a and d are untreated control cells. b and c, e and f are cells treated with H₂O₂ and tBHP respectively. Scale bar equals 25 μm

Figure 4

RIPK3 activation and its requirement in ARPE-19 cell death in response to oxidative stress. (A) Sequential images showing the distribution and aggregation of RIPK3 in ARPE-19 cells transfected with RIPK3-GFP expression plasmid and treated with 300 μM H₂O₂ (a–e) or 150 μM tBHP (f–j) for indicated times. Scale bar equals 25 μm. The arrows denote the RIPK3 aggregates in the treated cells. (B) Knockdown of RIPK3 by transfection with RIPK3 siRNAs shown by real-time qPCR. ***P<0.001. (C) Rescue of RPE cell death in RIPK3 siRNA-transfected cells at 24 h after H₂O₂ (300 μM) or tBHP (150 μM) treatment. *P<0.05. Results from 500 μM H₂O₂ are not statistically significant

Figure 5

Dying ARPE-19 cells from oxidative stress induce the expression of pro-inflammatory genes in healthy cells. (a) HMHB1 released to the cell medium as measured by YFP fluorescence in ARPE-19 cells transfected with HMGB1-YFP expression plasmid and treated with 300 or 500 μM H₂O₂, or 150 μM tBHP for 24 h. (b) Inflammatory gene TNFα expression measured by real-time RT-PCR in differentiated THP-1 cells after 24 h treatment with conditioned medium collected from dying ARPE-19 cells subjected to oxidative stress. Gene expression in THP-1 cells treated with cell medium from healthy ARPE-19 cells was used as control. **P<0.01; ***P<0.001. (c) Inflammatory gene TNFα expression in healthy ARPE-19 cells after 24 h treatment with conditioned medium collected from dying ARPE-19 cells subjected to oxidative stress. Gene expression in ARPE-19 cells treated with cell medium from healthy ARPE-19 cells was used as control. *P<0.05; **P<0.01; NS, nonsignificant

Figure 6

No evidence of pyroptosis or autophagy in ARPE-19 cells in response to oxidative stress. (a) No evidence of caspase-1 activation as measured by western blot analysis in ARPE-19 cells at 24 h after treatment with 300 or 500 μM of H₂O₂. Caspase 1 activation in ARPE-19 transfected with Alu RNA was used as positive control. (b) Lack of autophagosomes by LC3B antibody in ARPE-19 subjected to oxidative stress for indicated times. Choloquine (100 μM)-treated cells were used as positive controls. Scale bar equals 25 μm. Arrows point to the LC3 aggregates in the positive controls. (c) Lack of autophagosomes in LC3-GFP-transfected ARPE-19 cells treated by 300 or 500 μM of H₂O₂, or 150 μM of tBHP. LC3 distribution in the cells was visualized by GFP fluorescent signal. Choloquine (100 μM)-treated cells were used as positive controls. Scale bar equals 25 μm. Arrows point to the LC3 aggregates in the positive controls

Figure 7

Detection of necrosis in ARPE-19 cells in response to prolonged low oxidative stress. (A) Lack of apoptotic DNA degradation in APRE-19 cells subjected to low oxidative stress (2 h/day for 4 days). (B) Increase in nuclear membrane permeability in HMGB1-YFP-transfected ARPE-19 cells subjected to prolonged low oxidative stress. HMGB1-YFP was localized in nucleus in non-treated cells (a), but released to the cytoplasm under the indicated treatments (b–e). A small percentage of cells seem to undergo apoptosis as shown by the apoptotic bodies in f. (C) RIPK3 aggregation detected in RIPK3-GFP-transfected ARPE-19 cells upon prolonged low oxidative stress. No obvious RIPK3 aggregation was observed in control cells or cells treated with 100 μM H₂O₂ for 2 days (a and e). Pattern of RIPK3 aggregation (b–d) and normal RIPK3 distribution (f–h) were observed in cells treated with 200 μM H₂O₂, 75 μM tBHP or 100 μM tBHP for 2 days