Caspase-8–Mediated Apoptosis in Human RPE Cells

Abstract

purpose. Tumor necrosis factor (TNF)-α is an important cytokine associated with age-related macular degeneration (AMD) and proliferative vitreoretinopathy (PVR). TNF-α activates the extrinsic apoptotic pathway. In many cells, nuclear transcription factor (NF)-κB upregulates antiapoptotic proteins and prevents TNF-α–mediated apoptosis. However, retinal pigment epithelial (RPE) cells are resistant to TNF-α–induced apoptosis, even after specific NF-κB blockade. Herein, the authors investigated the role of caspase-8 in RPE cell resistance to TNF-α–mediated cell death.

methods. Caspase-8 mRNA and protein expression were measured in human RPE cells, human lens epithelial cells, human trabecular meshwork (TM) cells, human choroidal endothelial cells, human uveal melanoma cells (OCM-1, 92.1 and MKT-BR), T-98G, OVCAR-3, HCT116, and Jurkat cancer cells by real-time reverse transcription-polymerase chain reaction and Western blot, respectively. RPE cells were coinfected with adenovirus encoding caspase-8 and Cre. RPE and T-98G cells were infected with adenovirus encoding mutant inhibitory (I)-κB and then were treated with media alone or with TNF-α. Cell viability was determined by WST-1 assay, and apoptosis was evaluated with DNA fragmentation assay and M30 assay. Caspase-3, -7, -9 expression and Bid protein expression after caspase-8 overexpression were examined by Western blot.

results. Human RPE cell caspase-8 mRNA and protein levels were low compared with levels in nonneoplastic ocular cells and cancer cells. Overexpression of caspase-8 significantly decreased cell number, caused caspase-8 and caspase-3 activation, decreased full-length Bid, caspase-9, and caspase-7, and significantly increased DNA fragmentation and M30-positive RPE cells. Without TNF-α treatment, NF-κB blockade had no effect on caspase-8–mediated RPE cell death. In the presence of TNF-α, NF-κB blockade slightly but significantly enhanced caspase-8–mediated RPE cell death.

conclusions. RPE cell caspase-8 protein levels are low compared with levels for other cell types and may be regulated posttranscriptionally. Low caspase-8 levels may protect RPE cells from apoptosis normally and in diseases such as AMD and may promote the survival of abnormal cells in PVR. Introduction of caspase-8 into RPE cells may be a potential strategy to treat PVR.

The retinal pigment epithelium is a polarized monolayer of specialized cells located adjacent to the neural retina and is critically important in maintaining neural retinal function. Most retinal pigment epithelial (RPE) cells survive for a person’s lifetime, despite continued sublethal injury incurred by oxidants and inflammatory stimuli. 

Many diseases of the retina originate in or affect the retinal pigment epithelium. For example, proliferative vitreoretinopathy (PVR), the principal cause of retinal reattachment surgical failure, is characterized by RPE cell migration, proliferation, and collagen secretion that contributes to membrane formation. ^{1 2} Age-related macular degeneration (AMD) is the leading cause of irreversible vision loss among persons older than 65 years in the Western world. ^{3 4} RPE cell death is an important feature of the advanced forms of this disease. ^{5 6} It is important to understand how RPE cells survive or die in diseases such as PVR and AMD so that effective therapy can be designed. However, the mechanism(s) that maintain RPE cell survival in PVR or that contribute to their demise in advanced forms of AMD are not well understood. 

Macrophage-released tumor necrosis factor (TNF)-α is an important cytokine associated with PVR and AMD. ^{7 8 9} TNF-α activates the extrinsic apoptotic pathway and the nuclear transcription factor (NF)-κB survival pathway through TNF receptors. ¹⁰ In many cells, NF-κB upregulates antiapoptosis proteins and thereby prevents TNF-α–mediated apoptosis. ^{10 11} However, we have shown that RPE cells are resistant to TNF-α–induced apoptosis, even after specific NF-κB blockade. ¹²  

Caspases are a family of cysteine proteases that play important roles in regulating apoptosis. Caspase-8 lies at the apex of an apoptotic cascade and initiates proteolytic activation of downstream caspase family members, resulting in apoptosis. ^{13 14} Previously, we showed that the endogenous caspase-8 inhibitor, cellular Fas-associated death domain (FADD)-like interleukin-1β–converting enzyme-like inhibitory protein (c-FLIP), is expressed by RPE cells and that protein levels are increased in an NF-κB–dependent manner after TNF-α stimulation. ^{12 15} It is thought that the relative abundance of c-FLIP to caspase-8 determines whether caspase-8 is activated. ¹⁶ Our data showing that RPE cells do not die, even when NF-κB is inhibited, suggested that an extrinsic apoptotic signaling component maybe functionally reduced. In this report, we hypothesized that low RPE cell caspase-8 levels might explain RPE cell resistance to TNF-α–mediated cell death. 

Materials and Methods

Cell Culture

Human donor eyes were obtained from the North Carolina Organ Donor and Eye Bank, Inc. (Winston-Salem, NC) in accordance with the provisions of the Declaration of Helsinki for research involving human tissue. RPE cells for culture studies were harvested from eyes as previously described. ¹⁷ In addition, the RPE cell line ARPE-19, a generous gift from Leonard Hjelmeland, University of California at Davis, was also used in experiments. Cells were grown in Eagle minimal essential medium (MEM; Invitrogen, Carlsbad, CA) with 10% fetal bovine serum (FBS; Hyclone Laboratories, Logan, UT) and 1× antibiotic–antimycotic (Invitrogen) at 37°C in a humidified environment containing 5% CO₂. T-98G glioma cells, HCT116 colon carcinoma cells, and Jurkat cells (American Type Culture Collection [ATCC], Rockville, MD) were grown in MEM, McCoy 5a medium, and RPMI 1640 medium containing 10% FBS and 1× antibiotic–antimycotic (Invitrogen), respectively. OVCAR-3 ovarian carcinoma cells (ATCC) were maintained in RPMI 1640 medium supplemented with 2 mM l-glutamine, 1.5 g/L sodium bicarbonate, 4.5 g/L glucose, 10 mM HEPES, 1.0 mM sodium pyruvate, 10% FBS, and 1× antibiotic–antimycotic (Invitrogen). Human lens epithelial cells (cell line SRA 01/04) ^{18 19} were cultured in Dulbecco modified Eagle medium (DMEM) containing 20% FBS and 20 μg/mL gentamicin (Invitrogen). Human trabecular meshwork (TM) cells (donor age, 25 years), a generous gift from Pedro Gonzalez and Paloma Liton of the Duke University Eye Center, ²⁰ were maintained in low-glucose DMEM with l-glutamine and 110 mg/L sodium pyruvate, supplemented with 10% FBS, 100 μM nonessential amino acids, 100 U/mL penicillin, 100 μg/mL streptomycin sulfate, and 0.25 μg/mL amphotericin B (Invitrogen). Human choroidal endothelial cells (donor age, 20 years), a generous gift from Mary Hartnett of the University of North Carolina, ²¹ were maintained in endothelial growth media with growth factors (EGM-2; Cambrex, East Rutherford, NJ) with 10% FBS and 100 U/mL penicillin/100 μg/mL streptomycin sulfate. OCM-1, 92.1, and MKT-BR uveal melanoma cells ²² were cultured in RPMI 1640 medium supplemented with 10% FBS and 100 U/mL penicillin/100 μg/mL streptomycin sulfate. T-98G, HCT116, and OVCAR-3 cells were chosen as controls for the apoptosis-inducing effect of TNF-α after NF-κB blockade. ^{12 23} Jurkat cells were chosen as a positive control for caspase-8 antibody. 

Adenoviral Infection and Stimulation

RPE cells (2 × 10⁵) and T-98G cells (1.5 × 10⁵) were seeded in six-well plates (Corning-Costar Incorporated, Corning, NY). RPE cells (7 × 10⁴) were seeded in eight-well chamber slides (Nalge Nunc International, Naperville, IL), and RPE cells (6 × 10³) and T-98G cells (5 × 10³) were seeded in 96-well plates (Corning-Costar Inc.). Twenty-four hours later, cells were incubated with fresh medium for an additional 24 hours. RPE cells were infected with adenovirus encoding either β-galactosidase (LacZ) or caspase-8 (Riken, Tsukuba, Japan; prepared by Gene Therapy Center Virus Vector Core Facility, University of North Carolina [UNC], Chapel Hill, NC) at various multiplicities of infection (MOI) in MEM containing 1% FBS for 1 or 2 hours, and then the virus was removed. Adenovirus encoding caspase-8 was coinfected with adenovirus encoding Cre (used as an on-off switching unit of caspase-8 [Riken, prepared by Gene Therapy Center Virus Vector Core Facility at UNC]) at a 2:1 ratio of MOI. Cre is a site-specific recombinase that can recognize loxP sites and excise the neo gene between the loxP sites so that the promoter and the caspase-8 genes are joined. ^{24 25} The total MOI adenovirus used to infect each cell was kept the same in all experiments by supplementing with the LacZ construct. RPE cells and T-98G cells were then infected overnight with adenovirus encoding either LacZ or mutant inhibitory (I)-κB ²⁶ (Gene Therapy Center Virus Vector Core Facility, UNC) in MEM containing 1% FBS and then stimulated with TNF-α (1.1 × 10³ U/mL [22 ng/mL]; R&D Systems Inc., Minneapolis, MN) in MEM containing 1% FBS for various times. 

Real-Time RT-PCR Analysis

Total RNA was isolated, and real-time quantitative reverse transcription-polymerase chain reaction (RT-PCR) was performed, as we have previously described. ¹⁵ Briefly, duplicate reactions were prepared with 20 μL PCR master mix consisting of 10 μL master mix (iQ SYBR Green Supermix; Bio-Rad, Hercules, CA), 1 μL cDNA template, 1 μL each of caspase-8 primer pair (20 nM; forward, CTGCTGGGGATGGCCACTGTG; reverse, TCGCCTCGAGGACATCGCTCTC), and 7 μL RNase-free water. Reactions were denatured at 95°C for 2 minutes and amplified for 50 cycles at 95°C for 15 seconds, 60°C for 15 seconds, and 72°C for 15 seconds. Real-time quantification of caspase-8 gene was normalized to the threshold cycle (C _T) value of glyceraldehyde-3-phosphate dehydrogenase (GAPDH) in the corresponding cell types, where C _T equals the PCR cycle number at which the amount of amplified sample product reached 100 relative fluorescence units (RFUs). Fold difference of caspase-8 mRNA expression, relative to T-98G expression, was calculated by comparing C _T (2^-ΔΔCT). A melting curve for all products was obtained immediately after amplification by increasing temperature in 0.4°C increments from 65° for 85 cycles of 10 seconds each. The experiments were separately repeated three times with similar results. 

Cell Extracts and Western Blot

Cell extracts were prepared, and Western blot analysis was performed as previously described. ¹⁵ For Western blot, membranes were incubated overnight at 4°C with the following antibodies (Cell Signaling Technology, Beverly, MA) diluted in 3% milk: mouse monoclonal antibody directed against caspase-8 (9746, 1:1000), rabbit monoclonal antibody directed against caspase-3 (9665, 1:1000), rabbit polyclonal antibody directed against Bid (2002, 1:1000), caspase-7 (9492, 1:1000), and caspase-9 (9502, 1:1000). The following antibodies diluted in 5% milk were used for stripped membranes: mouse monoclonal antibody directed against cytokeratin 18 (1:1500; Sigma, St. Louis, MO) and GAPDH (1:5000; Chemicon, Temecula, CA). Blots were then washed three times (20 minutes per wash) in Tris-buffered saline containing 0.1% Tween-20 (TBST) and incubated with anti–mouse or anti–rabbit IgG conjugated with horseradish peroxidase (1:5000 in 3% milk; Jackson ImmunoResearch Laboratories, Inc., West Grove, PA) for 60 minutes at room temperature (RT). Immunoreactive bands were visualized using an enhanced chemiluminescence light (ECL) detection kit (Amersham, Piscataway, NJ). 

Protein Extraction from Native Human RPE Cells

Eyes were obtained 26 hours and 9 hours after death, respectively, from donors without any known ocular diseases (donor 1, 65-year-old woman; donor 2, 57-year-old man). After the neural retina was removed, 0.15 mL mammalian protein extraction reagent (Pierce, Rockford, IL) containing protease inhibitor cocktail (Roche, Indianapolis, IN) was added to the eyecup, and the RPE was gently scraped from the Bruch membrane with a rubber policeman. ²⁷ Another 0.05 mL extraction reagent was added to rinse the eyecup and then collected. Combined lysates were sonicated, and Western blot analysis was performed as described for cultured RPE cells. 

Cell Viability Assay

Colorimetric assay was performed based on the cleavage of the tetrazolium salt WST-1 (4-[3-(4-lodophenyl)-2-(4-nitrophenyl)-2H-5-tetrazolio]-1.3-benzene disulfonate) by mitochondrial dehydrogenases in viable cells (Roche). After 12 hours of stimulation with TNF-α, the cells were incubated with WST-1 solution (10 μL/well) for 1 hour at 37°C. The plate was read on a spectrophotometer at 440 nm with a reference wavelength at 690 nm. 

Immunofluorescence Detection of M30

RPE cell medium was removed, and cells were rinsed twice with phosphate-buffered saline (PBS) fixed with ice-cold 100% methanol for 30 minutes at −20°C, washed with washing buffer (PBS containing 0.1% Tween-20), and incubated with M30 CytoDEATH mouse monoclonal antibody (1:50 in washing buffer containing 1% bovine serum albumin (BSA; Roche) for 1 hour at RT. Cells were washed with washing buffer three times and then incubated with fluorescein isothiocyanate (FITC)–conjugated goat anti–mouse IgG antibody (1:100 in washing buffer; Jackson ImmunoResearch Laboratories, Inc.) for 30 minutes at RT. Cells were washed with washing buffer and then incubated with 4′,6-diamidino-2-phenylindole dihydrochloride (DAPI; Sigma) for 5 minutes. Fluorescence stain was observed with a light microscope and epifluorescence attachment. A masked observer determined the percentage of cells with M30-positive stain out of the total number stained with DAPI. 

Detection of DNA Fragmentation

After the cell number was counted by hemocytometer, cells (3 × 10⁵) were harvested by lysis in buffers provided in an enzyme-linked immunosorbent assay (ELISA) kit (Cell Death Detection ELISA^plus; Roche). Lysates were cleared by centrifugation, and DNA fragmentation, a late marker of apoptosis, was quantified by ELISA according to the manufacturer’s instructions. 

Statistical Analysis

Data are expressed as the mean ± SD. Student’s t-test was used to determine whether there were statistically significant differences between treatment groups determined by cell viability assay, M30 assay, and DNA fragmentation assay. P < 0.05 was considered statistically significant. Western blot analysis performed in duplicate, cell viability assays performed in quadruplicate, and M30 assay performed in triplicate were separately repeated three times in three individual experiments with similar results. DNA fragmentation assay in triplicate was separately repeated two times in two individual experiments with similar results. Data shown in Figures 1to 9are from representative experiments conducted on RPE cells from a 61-year-old donor. Cell viability assay, M30 assay, DNA fragmentation assay, and Western blot to detect cleaved caspase-8, cleaved caspase-3, and decreased full-length Bid were conducted in quadruplicate, triplicate, and duplicate experiments, respectively. These assays were also performed at least twice in a 7-year-old donor with similar results (not shown). 

Results

RPE Cell Caspase-8 mRNA Levels

We postulated that low caspase-8 levels could help explain RPE cell resistance to TNF-α–mediated RPE cell death. To test this hypothesis, we first determined RPE caspase-8 mRNA expression in human RPE cells from 11 different donors and in the ARPE-19 cell line by real-time RT-PCR. Human RPE cells and ARPE-19 cells expressed low steady state caspase-8 mRNA levels compared with HCT116, OVCAR-3, and Jurkat cells (Fig. 1A) . To determine whether low RPE cell caspase-8 levels were unique among ocular cell types, we evaluated caspase-8 mRNA expression in human lens, TM, and uveal melanoma cells (OCM-1, 92.1, and MKT-BR). Interestingly, RPE cell caspase-8 mRNA levels were also generally low compared with these ocular cells (Fig. 1B) , though the disparity was not as great as it was compared with non-ocular cells. 

RPE Cell Caspase-8 Protein Levels

We next determined whether low steady state mRNA levels correlated with low RPE cell caspase-8 protein. In initial experiments, we found that human RPE cells from 3 different donors and ARPE-19 cells had very low caspase-8 protein levels compared with T-98G cells, HCT116 cells, OVCAR-3 cells, and Jurkat cells (Fig. 2A) . To confirm this result, we tested cultured human RPE cells from eight other donors and found that all the RPE cells tested expressed very low levels of caspase-8 protein (Fig. 2B) . 

Additional Western blots were performed to determine whether RPE cell caspase-8 protein levels were also lower than those observed in human lens epithelial cells, TM cells, and uveal melanoma cells. Caspase-8 protein levels were low in RPE cells compared with nonneoplastic ocular cells, includingchoroidal endothelial cells, and very low compared with cancer cells (Fig. 3A) . In contrast, caspase-3 protein levels in RPE cells were higher than those observed in the ocular cells (Fig. 3B) . 

To determine whether the low caspase-8 levels observed in vitro reflected RPE cell caspase-8 levels in situ, we performed Western blot analysis on freshly isolated native human RPE cells. The results mirrored those found in vitro; caspase-8 levels were undetectable in freshly isolated RPE cells (Fig. 4) . 

Effect of Caspase-8 Overexpression on RPE Cell Death

In the preceding experiments, we established that RPE cells express low caspase-8 levels compared with other cell types. To determine the functional consequence of these low caspase-8 levels, we next determined whether the introduction of caspase-8 into RPE cells induced RPE cell death. RPE cells were coinfected with adenovirus encoding caspase-8 and Cre (Ad8), and cell viability was examined by WST-1 assay. We first confirmed caspase-8 transduction by Western blot. After Ad8 infection, a dose-dependent increase was observed in full-length and cleaved caspase-8 protein. Caspase-8 protein was not detected in caspase-8–infected RPE cells without Cre coinfection (data not shown). Cell number was significantly lower in Ad8-infected RPE cells than in LacZ-infected RPE cells (Fig. 5) . Without TNF-α treatment, NF-κB blockade had no effect on caspase-8–mediated RPE cell death. In the presence of TNF-α, NF-κB inhibition slightly but significantly enhanced caspase-8–mediated RPE cell death and markedly enhanced mutant IκB-infected T-98G cell death (Fig. 5) . 

Caspase-8–Mediated RPE Cell Apoptosis

To further explore the mechanism of caspase-8–mediated RPE cell death, the downstream effectors of caspase-8 were evaluated by Western blot. We identified cleaved caspase-8 and caspase-3 protein in untreated Ad8-infected RPE cells and in TNF-α–treated mutant IκB (TNF-AdmIκB)-infected T-98G cells (Fig. 6) . Interestingly, the caspase-3 cleavage pattern in Ad8-infected RPE cells (p17/15) was different from that observed in TNF-AdmIκB-infected T-98G cells (p19/17), and endogenous caspase-3 levels in RPE cells were higher than in T-98G cells (Fig. 6) . 

Caspase-8–mediated Bid cleavage generates truncated Bid (t-Bid) that forms a signaling link between the extrinsic TNF receptor family apoptotic pathway and the intrinsic mitochondrial-based apoptotic pathway. ^{13 14 28} Endogenous full-length Bid and caspase-7 protein levels were lower in RPE cells than in T-98G cells (Fig. 7) . After caspase-8 transduction, RPE cell full-length Bid levels were lower than those observed in noninfected and LacZ-infected controls, and caspase-9 and -7 levels were lower at an MOI of 3 than in a LacZ control (Fig. 7) . We were unable to directly detect cleaved Bid, caspase-9, or capase-7 in RPE cells or T-98G cells. 

Cytokeratin cleavage is a specific early apoptotic event in epithelial cells. M30 is an antibody that specifically recognizes a neoepitope exposed by cytokeratin cleavage when cells undergo apoptosis, and it has been widely used as a marker for early caspase-dependent apoptotic cell death. ^{29 30} We next further confirmed caspase-8–mediated RPE cell apoptosis. Few M30-positive cells were observed in noninfected or LacZ-infected RPE cells. However, when cells were coinfected with caspase-8 and Cre, the number of M30-positive cells was significantly increased (Fig. 8) . 

DNA fragmentation is a hallmark of apoptosis. As shown in Figure 9 , DNA fragmentation was significantly increased in Ad8-infected RPE cells compared with that in noninfected and LacZ-infected RPE cells. 

Discussion

In this study, we have demonstrated for the first time that human RPE cells express low caspase-8 steady state mRNA and very low caspase-8 protein levels but relatively higher levels of downstream executioner caspase proteins compared with other cell types. The low caspase-8 protein levels functionally prevented executioner caspase activation and prevented apoptotic RPE cell death. Transduction of caspase-8 into RPE cells triggered downstream executioner caspase activation and apoptotic cell death in the absence of TNF-α; cell death was further increased following TNF-α exposure after NF-κB was blocked. 

Caspase-8 protein levels are highly cell type specific. RPE cell caspase-8 protein levels were lower than those we observed in nonneoplastic ocular cells such as lens epithelial cells, TM cells, choroidal endothelial cells, and a variety of neoplastic cell lines, including human uveal melanoma cells, T-98G cells, and OVCAR-3 cells. In addition, several different cultured nonneoplastic epithelial cells, including renal tubular epithelial cells, ³¹ human gingival epithelial cells, ³² and keratinocytes, ³³ contain significant amounts of endogenous caspase-8 protein that can be processed to an activated form when TNF receptor family member receptors are ligated. The reason for low endogenous caspase-8 levels in human RPE cells, which precludes caspase-8 activation by TNF-α, and higher caspase-8 levels in other nonneoplastic epithelial cells, which permits caspase-8 activation, is unknown. However, we speculate that the need for long-term survival in normal RPE cells, compared with other epithelial cell types that turn over more rapidly, may help to explain these differences. We further hypothesize that low human RPE cell caspase-8 protein levels in persons of all ages, as observed in the present report, may help RPE cells to survive for the duration of a person’s entire life. 

Low RPE caspase-8 protein levels observed in the present report could reflect gene silencing or posttranscriptional regulation. Some cells use the former mechanism almost exclusively. For example, the caspase-8 gene CASP8 is frequently inactivated in a variety of cancers, including neuroblastoma, Ewing sarcoma, and malignant brain tumors, and in certain cancer cell lines including small cell lung carcinoma, neuroblastoma, and retinoblastoma cells. ^{34 35 36 37 38 39 40} In these tissues andcells, the gene is silenced through a combination of DNA methylation and allelic deletion. ^{34 35 36 37 38 39} Reexpression of caspase-8 through demethylation of its promoter using 5-aza-2′-deoxycytidine or overexpression with a caspase-8 expression vector increased sensitivity of cancer cells to TNF-related apoptosis-inducing ligand (TRAIL)-induced apoptosis. ^{34 41 42} Although we cannot exclude DNA silencing as a possible mechanism for decreased caspase-8 protein, low levels of caspase-8 transcripts were identified by real-time PCR, which suggests that in RPE cells caspase-8 is, at least in part, posttranscriptionally regulated. Studies to address these possible mechanisms are under way in our laboratory. 

RPE cell procaspase-8 was transduced by adenoviral vector infection. However, after transduction, both procaspase-8 and cleaved caspase-8 were observed. It is likely that in human RPE cells, as has been demonstrated in other cell types, ^{24 43 44 45} caspase-8 overexpression leads to caspase-8 self-aggregation and cleavage to active forms without any external apoptotic signal. In contrast, overexpression of caspase-3, a key effector caspase, does not cause caspase-3 self-activation. ^{46 47 48} Resistance to downstream caspase-3 autoactivation, in addition to low caspase-8 levels, may provide additional control to prevent unwanted RPE cell death. 

Low levels of endogenous RPE cell caspase-8 protein precluded an analysis of endogenous caspase-8 activity. In most cells with functional caspase-8, TNF-α induces cell death, especially when NF-κB, a transcription factor that upregulates antiapoptotic factors, is blocked. In contrast, TNF-α does not induce RPE cell death, even when NF-κB is inhibited. ¹² Furthermore, when we introduced exogenous caspase-8 into RPE cells by an adenoviral vector, cell number was decreased dramatically, caspase-8 was self-cleaved, and caspase-3, a downstream effector, was cleaved, suggesting that the introduction of exogenous caspase-8 restores the missing caspase-8 activity in RPE cells. Together, these results strongly support a lack of significant endogenous caspase-8 activity. 

The activation of caspase-8 by TNF receptor family members leads to apoptosis by two pathways. In the so-called type 1 cells, the level of activation of procaspase-8 initiated at the death-inducing signaling complex (DISC) is sufficient to cleave procaspase-3 directly. ^{13 14} However, in type 2 cells, less procaspase-8 is activated, and the mitochondrial pathway is required to amplify the weak death signal. ^{13 14} When caspase-8 triggers apoptosis through the mitochondria, small amounts of activated caspase-8 can efficiently cleave t-Bid, which then translocates to mitochondria and induces the release of cytochrome c, leading to the activation of caspase-3 through Apaf1. ^{13 14 49} Endogenous cleaved t-Bid is difficult to identify in some cultured cells. Accordingly, reduced full-length Bid levels have been taken to indicate t-Bid cleavage after TNF receptor engagement. ^{49 50 51 52 53} We observed reduced Bid and pro-caspase-9 after relatively low-dose procaspase-8 gene transduction. These data suggest that the mitochondria amplification loop could be used in caspase-8–mediated RPE cell apoptosis. 

We have previously reported that the antiapoptotic protein c-FLIP is produced by cultured human RPE cells. It is thought that the relative ratio of c-FLIP level to caspase-8 level determine whether caspase-8 can be activated by extrinsic death-inducing signals. ¹⁶ Together, our results suggest that low RPE cell caspase-8 levels compared with c-FLIP levels would inhibit TNF-α—induced death signaling. However, it remains to be determined whether treatment strategies to increase the relative ratio of caspase-8 to c-FLIP could prove beneficial to induce the apoptosis of actively proliferating RPE cells in PVR, a condition characterized by unwanted RPE cell proliferation, or whether strategies to increase c-FLIP/caspase-8 levels would be advantageous to promote RPE cell survival in advanced AMD. Studies to address these questions are under way in our laboratory. 

 Supported by National Eye Institute Grants RO1 EY9106 (GJJ) and 930EY05722 (core grant).

 Submitted for publication November 7, 2006; revised February 7 and 28, 2007; accepted April 25, 2007.

 Disclosure: P. Yang, None; J.J. Peairs, None; R. Tano, None; N. Zhang, None; J. Tyrell, None; G.J. Jaffe, None

 The publication costs of this article were defrayed in part by page charge payment. This article must therefore be marked “advertisement” in accordance with 18 U.S.C. §1734 solely to indicate this fact.

 

Figure 1.

 

Caspase-8 mRNA expression. RNA from cultured human RPE cells, ocular cells, and cancer cell lines was extracted, reverse transcribed to cDNA, amplified with caspase-8 gene-specific primer pairs, and quantified by real-time RT-PCR simultaneously in one plate. Expression of caspase-8 genes was calculated relative to expression of T-98G cells using 2^-ΔΔCT. (A) RPE cell caspase-8 mRNA expression compared with non-ocular caspase-8 mRNA expression. Ages of donors 1 to 11 are 7, 31, 17, 50, 61, 58, 51, 61, 48, 33, and 48 years, respectively. (B). A separate experiment to show RPE cell caspase-8 mRNA expression levels compared with neoplastic and nonneoplastic ocular cells. RPE1 and RPE5 denote cells from donor 1 and donor 5 (A), respectively.

Figure 2.

 

Caspase-8 protein expression. Proteins (60 μg) from cultured human RPE cells and cancer cell lines were separated by SDS-PAGE and transferred to a nitrocellulose membrane for Western blot analysis. Jurkat cells were pretreated with cycloheximide (CHX; 1 μg/mL) for 30 minutes and were stimulated with TNF-α (200 ng/mL) for 6 hours as a positive control for antibody. (A) Cells of T-98G, RPE from three donors, ARPE-19, HCT116, OVCAR-3, and Jurkat. (B) Cells of RPE from eight donors and Jurkat. (A, B) Donor numbers are matched with the order shown in Figure 1 .

Figure 3.

 

Caspase-8 protein levels in ocular cells. Proteins (60 μg) from cultured human RPE cells and ocular cells were separated by SDS-PAGE and transferred to a nitrocellulose membrane for Western blot analysis. RPE1 and RPE5 are cells from donors 1 and 5, respectively, as shown in Figure 1in a separate experiment. Cultured RPE cells coinfected with adenovirus encoding caspase-8 and Cre were used as a positive control. (A) Western blot probed with antibody to caspase-8. (B) Blot in (A) was stripped and reprobed with antibody to caspase-3. (C) Blot in (B) was stripped and reprobed with antibody to GAPDH. The relative quantity of each protein is shown separately below each lane.

Figure 4.

 

Caspase-8 protein expression in native human RPE cells. (A) Proteins (60 μg) from freshly isolated human RPE cells were separated by SDS-PAGE and transferred to a nitrocellulose membrane for Western blot analysis. (A) Western blot probed with antibody to caspase-8. (B) Blot in (A) was stripped and reprobed with antibody to cytokeratin 18. (C) Blot in (B) was restripped and reprobed with antibody to GAPDH. Lane 1: donor 1. Lane 2: donor 2. Lane 3: cultured RPE cells coinfected with adenovirus encoding caspase-8 and Cre at MOI of 3. Lane 4: T-98G cells used as a nonepithelial caspase-8 positive control.

Figure 5.

 

Caspase-8–induced RPE cell death. RPE cells were infected or coinfected with adenovirus encoding either LacZ or caspase-8/Cre at various MOIs for 2 hours, and the virus was removed. RPE cells and T-98G cells were then infected with adenovirus encoding LacZ or mutant IκB (used as an NF-κB blocker) overnight and stimulated with TNF-α (22 ng/mL) for 12 hours. Absorbance of samples against a background control was measured on a spectrophotometer at 440 nm with a reference wavelength at 690 nm. Results are expressed as mean ± SD (n = 3). *P < 0.0001 versus LacZ-infected cells. **P ≤ 0.001 versus LacZ-infected cells. ^P ≤ 0.0001 versus LacZ-infected cells. #P < 0.00001 versus LacZ-infected cells and P ≤ 0.01 versus untreated cells in mutant IκB-infected group. ##P = 0.005 versus untreated cells.

Figure 6.

 

Caspase-8 overexpression–induced caspase-8 and caspase-3 cleavage. RPE cells in duplicate wells were infected or coinfected with adenovirus encoding LacZ or caspase-8/Cre at MOIs of 2 and 3 for 1 hour and were replaced with fresh 1% FBS-MEM for 16 hours. T-98G cells were infected with adenovirus containing LacZ or mutant IκB overnight and stimulated with TNF-α (22 ng/mL) for 6 hours. Proteins (60 μg) were separated by SDS-PAGE and transferred to a nitrocellulose membrane for Western blot analysis.

Figure 7.

 

Effect of caspase-8 overexpression on downstream effectors of caspase-8. RPE cells in duplicate wells were infected or coinfected with adenovirus encoding LacZ or caspase-8/Cre at MOIs of 2 and 3 for 1 hour and were replaced with fresh 1% FBS-MEM for 16 hours. T-98G cells were infected overnight with adenovirus containing LacZ or mutant IκB and were stimulated with TNF-α (22 ng/mL) for 6 hours. Proteins (60 μg) were separated by SDS-PAGE and transferred to a nitrocellulose membrane for Western blot analysis. (A) Proteins were probed with antibody against Bid and GAPDH. (B) Proteins were probed with antibody against caspase-7, caspase-9, and GAPDH. Cytochrome c-treated Jurkat cell extracts (9663; Cell Signaling Technology) used as a control for antibodies.

Figure 8.

 

Caspase-8–mediated RPE cell apoptosis. RPE cells in triplicate wells were infected or coinfected with adenovirus encoding either LacZ or caspase-8/Cre at an MOI of 2 for 1 hour and were replaced with fresh 1% FBS-MEM for 16 hours. First horizontal panel: morphology before cells were assayed (original magnification, ×100). Cells were immunostained to detect apoptotic cells. Green cells stained with M30 antibody (second horizontal panel). Nuclei were stained with DAPI (third horizontal panel). Bar, 10 μM. The percentage of M30-positive cells was expressed as mean ± SD (n = 3). *P = 0.001 versus noninfected cells, and P = 0.006 versus LacZ-infected cells.

Figure 9.

 

Caspase-8 overexpression-induced DNA fragmentation in RPE cells. RPE cells in triplicate wells were infected or coinfected with adenovirus encoding LacZ or caspase-8/Cre at an MOI of 2 for 1 hour and were replaced with fresh 1% FBS-MEM for 22 hours. DNA fragmentation was measured by ELISA. Results were expressed as mean ± SD (n = 3). *P = 0.0001 versus noninfected cells and LacZ-infected cells.

The authors thank Izumu Saito, University of Tokyo, and Hirofumi Hamada, Sapporo Medical University, for the purchase of Ad8 and LacZ virus from Riken. They also thank Goldis Malek for assistance with protein extraction from native human RPE cells.