Opposing Role of Prion Protein in Oxidative Stress- and ER Stress-induced Apoptotic Signaling

Abstract

Although the prion protein is abundantly expressed in the CNS, its biological functions remain unclear. To determine the endogenous function of the cellular prion protein (PrP^c), we compared the effects of oxidative stress and endoplasmic reticulum (ER) stress inducers on apoptotic signaling in PrP^c-expressing and PrP^ko-knockout neural cells. H₂O₂, brefeldin-A (BFA) and tunicamycin (TUN) induced increases in caspase-9 and caspase-3, PKCδ proteolytic activation, and DNA fragmentation in PrP^c and PrP^ko cells. Interestingly, ER stress-induced activation of caspases, PKCδ, and apoptosis were significantly exacerbated in PrP^c cells, whereas H₂O₂-induced proapoptotic changes were suppressed in PrP^c compared to PrP^ko cells. Additionally, caspases-12 and -8 were activated only in BFA and TUN treatments. Inhibitors of caspase-9, caspase-3, and PKCδ significantly blocked H₂O₂-, BFA- and TUN-induced apoptosis, whereas the caspase-8 inhibitor attenuated only BFA- and TUN-induced cell death, and the antioxidant MnTBAP blocked only H₂O₂-induced apoptosis. Overexpression of the kinase inactive PKCδ^K376R or the cleavage site-resistant PKCδ^D327A mutants suppressed both ER- and oxidative stress-induced apoptosis. Thus, PrP^c plays a proapoptotic role during ER stress, and an anti-apoptotic role during oxidative stress-induced cell death. Together, these results suggest that cellular PrP^c enhances the susceptibility of neural cells to impairment of protein processing and trafficking, but decreases the vulnerability to oxidative insults, and that PKCδ is a key downstream mediator of cellular stress-induced neuronal apoptosis.

Introduction

Prion-related diseases are infectious diseases often called transmissible spongiform encephalopathies (TSEs), that affect many mammalian hosts [1–7]. Prion diseases include Creutzfeldt-Jakob disease (CJD) in humans, bovine spongiform encephalopathy (BSE) in cattle, scrapie in sheep, and chronic wasting disease (CWD) in deer and elk, transmissible mink encephalopathy (TME) in minks, and they all share considerable neuropathological and clinical similarities [1–7]. Prion diseases cause severe neuronal damage mainly in the brain regions that control motor function, including the basal ganglia, cerebral cortex, thalamus, and cerebellum. Neuronal damage is accompanied by the accumulation of the abnormal prion protein PrP^sc (scrapie isoform), derived from the post-translational conformational change of normal prion protein PrP^c (cellular isoform) through unknown pathogenic mechanisms. The abnormally folded isoform PrP^sc is partially resistant to proteinase K digestion and appears to be the infectious proteinaceous particle [8, 9]. Therefore, prion diseases are transmissible and threatening to both humans and animals.

Apoptosis is the major type of cell death observed in several neurodegenerative disorders, including prion diseases, and is characterized by a series of distinct morphological and biochemical changes caused by activation of an array of cell signaling molecules in the cell [10, 11]. Apoptotic cell death pathways can be triggered by either an extrinsic factors pathway triggered by cell death receptors such as Fas, and mediated by caspase-8 or caspase-10, or by an intrinsic pathway dependent on oxidative stress, mitochondrial release of cytochrome C, caspase-9, and caspase-3 [12–14]. More recently, apoptosis was also shown to be triggered by ER stress which proceeds via an alternative intrinsic pathway that is dependent on caspase-12 and caspase-8 [15–18].

Oxidative damage and apoptosis have been well documented in animal and human prion diseases [19–21]. Evidence indicates that the cellular antioxidant capacity is significantly compromised in cells lacking PrP^c, and that PrP^ko cells are more sensitive to oxidative stress than control cells [22–27]. Studies have also shown PrP^sc and the toxic fragment PrP106-126 induces apoptosis in a variety of neuronal cells via Bcl2/Bax dysregulation, mitochondrial dysfunction, ROS generation, cytochrome C release, caspase-3, and p38 MAP kinase activation [28–33]. Recently, Hetz et al. demonstrated that caspase-12 and ER stress mediate PrP^sc-induced apoptotic cell death in mouse Neuro2A cells [34]. Studies have also shown that normal host prion protein is necessary for scrapie-induced neurotoxicity, as evidenced by PrP^c-deficient mice unable to develop the disease when infected with scrapie isoform PrP^sc [35], and depletion of cellular PrP^c during ongoing prion infection prevents the neurodegeneration and pathology associated with the disease [36]. Thus, PrP^c seems to be fundamental to the neurotoxicity observed during PrP^sc infections [37–39]. Cross-linking PrP^c in vivo with specific monoclonal antibodies was recently shown to trigger neuronal apoptosis, suggesting that PrP^c functions in the control of neuronal survival [40]. However, the cellular mechanisms by which PrP^c is converted to PrP^sc to cause rapid and severe neuronal damage in prion diseases are poorly understood.

Therefore, characterization of neurobiological functions of PrP^c will assist in elucidating the pathogenic mechanisms underlying prion diseases. To more fully understand the biological role of PrP^c, a stable neural cell line derived from PrP knockout mice was compared to PrP knockout cells that had been engineered to express mouse PrP^c. In this study, we used these two cell lines to evaluate the contribution of cellular non-pathogenic PrP^c to oxidative and ER stress-induced apoptotic cell death mechanisms. Herein, we report that cellular PrP^c enhances the susceptibility of neural cells to ER stress-induced apoptotic cell death and protects against vulnerability to oxidative insults, and that PKCδ is a key downstream mediator of cellular stress-induced neuronal apoptosis.

Materials and Methods

Chemicals and reagents

Hydrogen peroxide (H₂O₂), Brefeldin A (BFA), Tunicamycin (TUN), cyclosporine A, β-actin (mouse monoclonal), histone H1, β-glycerophosphate, ATP, and protein-A-sepharose were obtained from Sigma-Aldrich (St. Louis, MO); rottlerin was purchased from Calbiochem (San Diego, CA). Z-VAD-FMK, (Z-Val-Ala-Asp-Fluoro Methyl Ketone), Z-DEVD-FMK (Z-Asp-Glu-Val-Asp-Fluoro Methyl Ketone), Z-IETD-FMK (Z-Ile-Glu-Thr-Asp-Fluoro Methyl Ketone), Z-LEHD-FMK (Z-Leu-Glu-His-Asp-Fluoro Methyl Ketone), Ac-DEVD-AFC (Acetyl Asp-Glu-Val-Asp-AFC), Ac-IETD-AFC (Acetyl Ile-Glu-Thr-Asp-AFC), and Ac-LEHD-AFC (Acetyl-Leu-Glu-His-Asp-AFC) were obtained from MP Biomedicals (Irvine, CA). Antibodies to PKCδ and PKCα were purchased from Santacruz labs (Santacruz, CA), and 3F4 monoclonal antibody was purchased from Signet Labs (Dedham, MA). Antibody for SAF32 was obtained from Cayman Chemical (Ann Arbor, MI). ECL chemiluminescence kit was purchased from Amersham Pharmacia Biotech (Piscataway, NJ). Hydroethidine was purchased from Molecular Probes Inc. (Eugene, OR). Cell Death Detection Elisa Plus Assay Kit was purchased from Roche Molecular Biochemicals (Indianapolis, IN). Mn(III)tetrakis(4-benzoic acid)porphyrin chloride (MnTBAP) was purchased from Oxis Health Products (Portland, Oregon). [γ-³²P]ATP was purchased from Perkin Elmer (Downers Grove, IL). Bradford protein assay kit was purchased from Bio-Rad Laboratories (Hercules, CA). Lipofectamine Plus reagent, RPMI, horse serum, fetal bovine serum, L-glutamine, penicillin, streptomycin, and PCEP4 plasmid were purchased from Invitrogen (Gaithersburg, MD). Plasmids for kinase-inactive dominant negative mutant PKCδ^K376-GFP fusion protein and pEGFP-N1 were kind gifts from Dr. Stuart Yuspa, National Cancer Institute, Bethesda, Maryland. Plasmids for caspase cleavage-resistant PKCδ mutant PKCδ^D327A-GFP fusion protein were kindly provided by Dr. Mary E. Reyland, University of Colorado (Boulder, CO).

Generation of the brain-derived PrP^0/0 cell line CF10

Immortalization of PrP^0/0 cells was done using the plasmid vector pSV3-neo and cells were derived from 129/Ola mice with an inactivated PrP gene accomplished via gene targeting. CF10 cell line lacking the cellular prion protein was generated from the brain of E15 mouse pups (Unpublished observations Vorberg and Priola).

PrP^c and PrP^ko cells

PrP^c cells express mouse prion protein with a hamster 3F4-epitope and PrP^ko cells were derived from prion knockout mice (Priola et al., 2001; Takemura et al., 2006). PrP^c mouse neural cell line was derived from CF10 mouse neural cell line lacking prion protein engineered to stably express the mouse PrP^c gene with 3F4 hamster epitope. As a non-PrP control, PrP-knockout cells expressing the empty vector PrPko were also established. Mouse PrP^c and PrP^ko neural cell lines (kindly provided by Dr. Suzette A Priola, National Institute of Allergy and Infectious Disease, NIH, Hamilton, Montana) were grown in DMEM medium supplemented with 10% fetal bovine serum, 2 mM L-glutamine, 50 units penicillin, and 50 μg/ml streptomycin. Cell cultures were maintained in a humidified atmosphere of 5% CO₂ at 37°C for 4–5 days and then used for experiments.

Treatment paradigm

After 2–4 days in culture, PrP^c and PrP^ko cells were exposed to 10–100 μM H₂O₂, 3–30 μM BFA, or 0.6–6 μM TUN over a 24 h time. In inhibitor studies, MnTBAP (ROS inhibitor, 2 μM), rottlerin (PKCδ inhibitor, 2 μM), cyclosporine A (mitochondrial permeability pore inhibitor, 1 μM), Z-VAD-FMK (pan-caspase inhibitor, 100 μM), Z-DEVD-FMK (caspase-3-specific inhibitor, 100 μM), Z-LEHD-FMK (caspase-9-specific inhibitor, 100 μM), or Z-IETD-FMK (caspase-8-specific inhibitor, 100 μM) were added at the same time as H₂O₂, BFA, and TUN. After treatment, cells were harvested at 1, 3, 6, 12, and 24 h and were assessed for caspase-3, caspase-8, caspase-9, caspase-12, and proteolytic activation of PKCδ and DNA fragmentation. DMSO (dimethyl sulfoxide, 0.1%) and ethanol (0.1%) were used as vehicles in control experiments.

Transient transfections

Transient transfection assays were performed as described previously and according to the manufacturer’s protocol supplied with the Cell Line Nucleofector™ Kit V [41]. Briefly, after 2–3 days in culture, PrP^c cells were harvested, washed with 1X PBS, and resuspended in transfection solution (supplied with the kit) at a cell density of 2x10⁶ cells/100 μl. Plasmids encoding pPKCδ^K376R-GFP, pPKCδ^D327A-GFP, and pEGFP-N1 DNA were added separately to 100 μl of cell suspension, transferred into a nucleofector cuvette, and electroporated using the Nucleofector device. Cells were immediately transferred from the cuvette to 24-well plates after dilution with growth media. After 24 h, GFP-expressing cells were observed under a TE2000 Nikon fluorescence microscope and the transfection efficiency was estimated to be between 70–75%. At this point, cells were harvested and used for apoptosis assays.

Immunocytochemistry

After PrP^c and PrP^ko cells were established, they were assessed for PrP^c expression using monoclonal antibody directed against 3F4 epitope. Cells were plated on poly-l-lysine (0.1 mg/ml) coated coverslips. After 12–24 h, the cells were fixed with 4% formaldehyde and permeabilized with 0.2% Triton X-100. The cells were then incubated with antibodies directed against the 3F4 epitope (3F4 mouse monoclonal 1:250) overnight at 4°C, followed by incubation with the mouse Alexa-488 conjugated secondary antibody (Jackson Laboratories, 1:1000) for 90 min at room temperature; cells were then washed with 1X PBS, mounted on slides, viewed, and images were captured using a Nikon C1 confocal microscope.

Cytotoxicity assays

Cytotoxicity assays were performed as described previously [41]. PrP^c and PrP^ko cells were exposed to 30 μM H₂O₂, 10 μM BFA, or 2 μM TUN for 24 h. Cell death was determined by the trypan blue exclusion method with an Improved Nebauer-type Hemacytometer. The cell viability was normalized as percent of control. Cell death was also determined using the Sytox green fluorescence-based nucleic acid probe after exposing PrP^c cells to 30 μM H₂O₂, 10 μM BFA, or 2 μM TUN for 24 h in the presence or absence of 100 μM Z-VAD-FMK. Sytox green is a cell-impermeable nucleic acid dye that enters dead cells and intercalates with DNA to produce green fluorescence, which was quantified using a fluorescence microplate reader (SpectraMax Gemini XS Model, Molecular Devices, Sunnyvale, CA) with excitation at 485 nm and emission at 538 nm [41].

Flow cytometric measurement of ROS generation

ROS generation was detected in the cells using hydroethidine with a Becton Dickenson FACScan™ flow cytometer (Becton Dickinson, San Francisco, CA) as described previously [42]. Flow cytometric data were analyzed by Cellquest™ data analysis software. PrP^c and PrP^ko cells were resuspended with Hanks balanced salt solution (HBSS) with 2 mM calcium at a density of 0.5 × 10⁶ cells/ml. Cells were then incubated with 10 μM hydroethidine for 15 min at 37°C in the dark to allow dye loading into the cells. After incubation with dye, excess dye was removed, and the cells were resuspended with HBSS. Following addition of 100 μM H₂O₂, 30 μM BFA, or 6 μM TUN, ROS generation was measured at 0, 5, 15, 30, 45, and 60 min post-exposure in the fluorescence-activated cell sorter in the flow cytometer.

Enzymatic assay for caspases

Caspase-3, caspase-8, and caspase-9 activity analyses were performed as previously described [42]. After treatment cells briefly were spun and the cell pellets were lysed with Tris buffer (pH 7.4, 50 mM Tris HCl, 1 mM EDTA, and 10 mM EGTA) containing 10 μM digitonin for 20 min at 37°C. Lysates were centrifuged at 900 × g for 3 min, and the resulting supernatants were incubated with specific fluorogenic caspase substrates at 37°C for 1 h. Ac-DEVD-AFC (50 μM), Ac-IETD-AFC (50 μM) and Ac-LEHD-AFC (50 μM) were used as substrates for determining caspase-3-, caspase-8-, and caspase-9-like protease activities, respectively. Levels of cleaved (active) caspase substrate were monitored at excitation λ 400 nm and emission λ 505 nm using a fluorescence microplate reader (SpectraMax Gemini XS Model, Molecular Devices, Sunnyvale, CA). Caspase activities were expressed as fluorescence units per mg protein per h. The protein concentrations were determined using the Bio-Rad protein assay kit.

Western blots

After exposure of PrP^c and PrP^ko cells (~1 × 10⁷ cells) to 100 μM H₂O₂, 10–30 μM BFA, or 2–6 μM TUN with or without pharmacological inhibitors over 12 h, cell lysates were prepared as described previously [42]. Briefly, PrP^c and PrP^ko cells were washed with 1X phosphate buffered saline (PBS), resuspended in homogenization buffer, and then sonicated for 10–15 sec. The homogenates from treated and control cells were centrifuged at 16,000 × g for 60 min at 4°C, and the supernatants were collected as cell lysates. Cell lysates containing equal amounts of protein were loaded in each lane, resolved on a 10–12% SDS-polyacrylamide gel, and blotted onto nitrocellulose membrane (Bio-Rad Laboratories). After blocking the non-specific binding sites with non-fat dry milk (Amersham Pharmacia Biotech), the membranes were then treated with either anti-3F4-PrP^c (1:500 mouse monoclonal), or SAF-32 monoclonal antibody (1:500) for detecting mouse prion, anti-procaspase-12 (1:1000, rabbit polyclonal), anti-PKCδ (1:2000, rabbit polyclonal), or anti-PKCα (1:1000 rabbit polyclonal) antibodies, followed by secondary HRP-conjugated anti-rabbit (1:2000) or anti-mouse (1:2000) antibody. Antibody-bound proteins were detected with an enhanced chemiluminescence (ECL) system using a Kodak Imager (Kodak Image Station 2000R, Eastman Kodak Company, New Haven, CT). To confirm equal protein in each lane, membranes were reprobed with β-actin (1:5000 mouse monoclonal).

Immunoprecipitation kinase assays

PKCδ enzymatic activity was determined using an in vitro immunoprecipitation kinase assay as previously described [42]. Briefly, PrP^c and PrP^ko cells were exposed to 100 μM H₂O₂, 10 μM BFA, or 2 μM TUN for 12 h, with or without a caspase-3 inhibitor (100 μM Z-DEVD-FMK), and cell lysates were collected. After immunoprecipitation with anti-PKCδ antibody, 25-μl samples containing PKCδ bound to sepharose A beads were incubated with 25 μl of reaction buffer containing 0.4 mg histone H1 and 5 μCi of [γ-³²P] ATP (4,500 Ci/mM) for 10 min at 30°C. The reaction was terminated by the addition of 2X SDS gel loading buffer and boiled for 5 min. The samples were separated on 12% SDS-PAGE and histone phosphorylated bands were detected using a Phospho Imager (Personal Molecular Imager FX, Bio-Rad Laboratories, Hercules, CA) and quantified using Quantity One 4.2.0 Software (Bio-Rad Laboratories).

DNA fragmentation assay

DNA fragmentation was measured using a Cell Death Detection ELISA Plus Assay Kit, as described previously [42]. Briefly, PrP^c and PrP^ko cells were exposed to 100 μM H₂O₂, 10 μM BFA, or 2 μM TUN in the presence or absence of pharmacological inhibitors for 24 h. After treatment, cells were spun down, washed with 1X PBS, and incubated with cell lysis buffer (supplied with the kit) for 30 min at RT, and then centrifuged. The supernatants were then dispensed onto streptavidin-coated 96-well microtiter plates containing 80 μl of HRP-conjugated antibody cocktail. After 2 h incubation at RT, the nucleosomes retained by the antibody cocktail in the immunocomplex were quantified spectrophotometrically with ABTS as an HRP substrate. Measurements were made at 405 nm and 490 nm using a SpectroMax 190 spectrophotometer (Molecular Devices, Sunnyvale, CA). The difference of absorbance between OD 405 and OD 490 nm was used to quantify the levels of DNA fragmentation.

Data analysis

Data analysis was performed using Prism 3.0 software (GraphPad Software, San Diego, CA). Data from caspase enzymatic activities and DNA fragmentation assays were first analyzed using one-way ANOVA. Neuman-Keuls or Dunnett’s post-tests were then performed to compare between untreated, H₂O₂-, BFA- and TUN-treated, and inhibitor-cotreated groups; differences with p<0.05 were considered significant.

Results

Characterization of PrP^c-expressing cell lines

First, we characterized the level of prion protein expression in PrP^c cells and PrP^ko cells. Stable expression of PrP^c was verified using 3F4 monoclonal antibody raised by immunocytochemical analysis. A bright immunofluorescence staining on the cell surface was observed in PrP^c cells, with a very weak staining in the cytoplasm or the nucleus (Fig 1A), indicating that PrP^c expression was predominantly localized in the plasma membrane. No staining was observed in PrP^ko cells. PrP^c expression was confirmed with Western blot analysis using two antibodies directed against two different epitopes in the mouse prion protein. 3F4-antibody detects the hamster epitope present in the exogenously expressed prion protein, whereas SAF32 antibody can detect both endogenous and expressed mouse prion protein. Since, PrP^ko cells were derived from prion knockout mice, we used SAF32- antibody to confirm that PrP^ko cells do not express the endogenous prion protein. Figure 1B shows stable expression of PrP^c protein in PrP^c cell lysates, but not in PrP^ko cell lysates with both 3F4- and SAF32-antibodies. Nitrocellulose membranes were reprobed with β-actin antibody and the thickness of the 43 kDa β-actin band was identical in all lanes, confirming equal protein loading. As shown in Fig. 1B, PrP^c cells were positive for mouse prion protein, while PrP^ko cells were negative for both SAF32 and 3F4-antibody staining. Western blots revealed similar prion protein expression, di-glycosylated (32–34 kDa), mono-glycosylated (24–28 kDa), and unglycosylated (17–20 kDa) with both SAF32 and 3F4 antibody staining (Fig. 1B), indicating that the PrP^c protein is synthesized and processed normally.

Fig. 1.

Effect of H₂O₂, BFA, and TUN on PrP^c protein expression. (A) Confocal images depicting stable expression of mouse PrP^c in PrP^ko neural cells using 3F4-antibody. A bright immunofluorescence with cell surface staining pattern was observed in PrP^c cells, with no or very weak staining in the cytoplasm or the nucleus, indicating that PrP^c was expressed predominantly in the cell surface. No staining was observed in PrP^ko cells. (B) Western blot showing stable expression of PrP^c protein in PrP^c and PrP^ko cell lysates. (C) Western blot showing the effect of 100 μM H₂O₂, 30 μM BFA, or 6 μM TUN on PrP^c expression following 24 h treatment. Membranes were reprobed with β-actin antibody to confirm equal protein loading in each lane.

Effect of H₂0₂, BFA, and TUN on PrP^c expression

We examined the effects of H₂O₂ (oxidative stress inducer), BFA (inhibitor of protein transport from the ER to Golgi), and TUN (N-glycosylation inhibitor) on PrP^c expression. Figure 1C shows the effect of 100 μM H₂O₂, 30 μM BFA, and 6 μM TUN on PrP^c expression. Antibody directed against the prion 3F4-epitope was used to determine the expression pattern. Untreated PrP^C cells expressed all three forms of prion protein: unglycosylated (17–20kDa), mono-N-glycosylated (24–28kDa) and di-N-glycosylated (32–34kDa). The pattern of expression in H₂O₂ was similar to that in untreated cells. As expected in the presence of N-glycosylation inhibitor (TUN), PrP^C cells expressed only the unglycosylated form. On the other hand, BFA-treated cells expressed mostly mono-N-glycosylated and unglycosylated forms and significantly reduced di-N-glycosylated form compared to untreated PrP^c cells. Nitrocellulose membranes were reprobed with β-actin antibody to confirm equal protein loading.

Caspases mediate oxidative stress- and ER stress-induced cytotoxic cell death

To determine the effect of oxidative stress- and ER stress-inducers on cytotoxic cell death, we compared the relative cytotoxic response of H₂O₂, BFA, and TUN treatment in PrP^c cells. Figure 2A shows that 24-h treatment with 30 μM H₂O₂, 10 μM BFA, or 2 μM TUN decreased the cell viability by 70%, 85%, and 77%, respectively, compared to the untreated control PrP^c cells, using the trypan blue cell death assay. There was no significant loss of cell viability in H₂O₂-, BFA-, or TUN-treated cells up to 6 h (data not shown). To determine if caspases mediate cytotoxic cell death, cells were co-treated with 100 μM Z-VAD-FMK (pan caspase inhibitor) for 24 h, and cell death was quantified by Sytox green fluorescence assay. Exposure to 30 μM H₂O₂, 10 μM BFA, or 2 μM TUN for 24 h increased cytotoxic cell death by 150%, 236%, and 200% respectively, compared to untreated control cells (control values are normalized to 100%). Co-treatment with 100 μM Z-VAD-FMK (pan-caspase inhibitor) almost completely attenuated H₂O₂-, BFA-, and TUN-induced cytotoxicity (Fig. 2B). These results suggest caspases mediate both oxidative and ER stress-induced cytotoxic cell death.

Fig. 2.

Effect of PrP^c on H₂O₂-, BFA-, and TUN-induced cytotoxic cell death and ROS generation. (A) Trypan blue assay: PrP^c-expressing neural cells were treated with 30 μM H₂O₂, 30 μM BFA, or 6 μM TUN, and cell viability was determined using the trypan blue exclusion technique by counting live (clear) and dead (stained) cells in three to five randomly selected fields. (B) Sytox assay: PrP^c cells were treated with 100 μM H₂O₂, 30 μM BFA, or 6 μM TUN ± 100 μM Z-VAD-FMK for 24 h, and the number of Sytox-positive cells was determined using the SYTOX® Green nucleic acid stain assay, as described in the Materials and Methods section. (C) ROS generation was measured in PrP^c and PrP^ko cells treated with 100 μM H₂O₂, 30 μM BFA, or 6 μM TUN, and hydroethidine fluorescence intensity was measured at 45 min by flow cytometry. Data represent mean ± SEM, N = 6, from two independent experiments performed in triplicate. Asterisks (*p<0.05 and **p<0.01) indicate significant differences between control and treatment groups. Pound sign (#p<0.05 and ##p<0.01) indicates significant differences between treated and Z-VAD-FMK-treated cells or H₂O₂-treated PrP^c and PrP^ko cells.

H₂O₂-induced ROS generation is diminished in PrP^c-expressing cells

Flow cytometric analysis using the ROS-sensitive fluorescence probe hydroethidine revealed that H₂O₂ treatment induced time-dependent increases in ROS generation, whereas exposure to BFA and TUN did not induce significant increases in ROS generation. As shown in Fig. 2C, exposure to 100 μM H₂O₂ for 45 min resulted in 188% and 232% increases in ROS production in PrP^c and PrP^ko cells, respectively, whereas 30 μM BFA- and 6 μM TUN-treatment only slightly (118% and 117% of control) increased ROS generation when compared to untreated PrP^c cells in the same time period. Since PrP^c cells were more sensitive to BFA and TUN, we measured ROS generation in PrP^c cells only. Overall, these results suggest that oxidative stress, but not ER stress, generates a significant amount of ROS in PrP^c and PrP^ko cells. Additionally, PrP^c expression partially protected against H₂O₂-induced increases in ROS generation.

PrP^c exacerbates ER stress, but protects against oxidative stress-induced increases in caspase-9, caspase-8, and caspase-3 activation

Previous studies have implicated multiple caspases, including caspase-9, caspase-3, and caspase-12, in mediating various stress-induced apoptotic cell death [12, 13, 15, 17]. Therefore, we examined the activation of various caspases, including caspase-9, caspase-3, caspase-8, and caspase-12, over a 12 h period with H₂O₂, BFA, or TUN. Initial time-course analysis revealed BFA and TUN induced maximal increases in caspase-8, caspase-9, and caspase-3 enzyme activities. Increases were observed at 3, 6, and 12 h in both PrP^c and PrP^ko cells, suggesting a sequential activation of caspase-8, caspase-9, and caspase-3. As shown in Fig. 3, BFA- and TUN-induced increases in caspase-9 (Fig. 3A), caspase-3 (Fig. 3B), and caspase-8 (Fig. 3C) were significantly augmented in PrP^c cells, whereas H₂O₂-induced caspase-9 and -3 were significantly attenuated in PrP^c cells compared to PrP^ko cells. H₂O₂-treatment did not induce any significant changes in caspase-8 activity in PrP^c and PrP^ko cells. Together, these results suggest that PrP^c partially protects against H₂O₂-induced increases, but exacerbates BFA- and TUN-induced increases in caspase activities.

Fig. 3.

Effect of PrP^c on H₂O₂-, BFA-, and TUN-induced caspase activations. (A) Caspase-9, (B) Caspase-3 and (C) Caspase-8. PrP^c and PrP^ko cells were treated with 100 μM H₂O₂, 30 μM BFA, or 6 μM TUN over a period of 12 h. Cytosolic extracts were collected and assayed for caspase-3 (12 h), caspase-9 (6 h), and caspase-8 (3 h) activities using the enzyme-specific fluorogenic substrate as described in the Materials and Methods section. Data represent the mean ± SEM, N = 6, from two independent experiments performed in triplicate. Asterisks (*p<0.05) and (**p<0.01) represent significant differences between treated and respective untreated cells.

Activation of caspase-12 in BFA- and TUN-treated, but not in H₂O₂-treated, PrP^c cells

Caspase-12 has been reported to be activated during ER stress [16, 17, 34, 43]. Activation of this caspase normally occurs in the ER and results from the proteolytic conversion of procaspase-12 (50 kDa) to activated caspase-12 (20 and 30 kDa), which is then translocated to the cytoplasm. In the cytoplasm, caspase-12 has been shown to directly activate caspase-8 and caspase-3 [18]. Therefore, we examined if caspase-12 is activated during treatment with BFA, TUN, or H₂O₂ in PrP^c cells. Since antibodies to detect cleaved caspase-12 are not available, we measured loss of native procaspase-12 as an indicator of increased caspase-12 activation Caspase-12 was activated as early as 1 h in BFA- and TUN-treated PrP^c and PrP^ko cells (data not shown). As shown in Fig. 4A, treatment with 30 μM BFA and 6 μM TUN for 3 h induced the activation of caspase-12, as determined by the loss of procaspase 12 (50 kDa) in PrP^ko and PrP^c cells. However, exposure to 100 μM H₂O₂ did not significantly decrease the 50 kDa caspase-12 band compared to untreated control cells (Fig. 4A). Further, there were no significant differences in caspase-12 activation between PrP^c and PrP^ko cells in the different treatments. Nitrocellulose membranes were reprobed with β-actin antibody to confirm equal protein loading. The results were deduced from densitometric analysis of the 50 kDa native caspase-12 band normalized to β-actin band (data not shown).

Fig. 4.

Effect of PrP^c on H₂O₂-, BFA-, and TUN-induced caspase-12 and PKCδ cleavage. (A) Caspase-12 activation in PrP^c and PrP^ko cells, (B) PKCδ cleavage in PrP^c cells, and (C) PKCδ cleavage in PrP^ko cells. Subconfluent undifferentiated PrP^c and PrP^ko cells were exposed to 30–100 μM H₂O₂, 10–30 μM BFA, or 2–6 μM TUN, and assayed for caspase-12 activation 3 h post-treatment (loss of procaspase-12), and PKCδ cleavage 12 h post-treatment. Cell lysates were obtained as described, and were separated by 10–12% SDS-polyacrylamide gel electrophoresis, and transferred to nitrocellulose membrane; PKCδ was detected using rabbit polyclonal antibodies directed against their catalytic subunits. To confirm equal protein loading in each lane, the membranes were reprobed with β-actin antibody. The immunoblots were visualized using the GE Healthcare’s ECL detection agents.

Oxidative and ER stress induce proteolytic cleavage of PKCδ

Recently, we demonstrated that PKCδ is one of the important endogenous substrates of caspase-3 in neuronal cells [44–46]. Activated caspase-3 cleaves PKCδ, yielding a 38 kDa regulatory fragment and a 41 kDa catalytic fragment, resulting in persistently active kinase. Since H₂O₂, BFA, and TUN exposure resulted in a profound activation of caspase-3, we decided to examine the proteolytic cleavage of PKCδ. After treatment of PrP^ko and PrP^c cells with 100 μM H₂O₂, 10 μM BFA, or 2 μM TUN for 12 h, we observed that a significant proportion of native PKCδ (72–74 kDa) protein was proteolytically cleaved to yield the 41 kDa catalytically active fragment when immunoblotted with an antibody raised against PKCδ. We observed a dose-dependent increase in the proteolytic cleavage of PKCδ in both PrP^c (Fig. 4B) and PrP^ko cells (Fig. 4C); however, proteolytic cleavage of PKCδ was more pronounced in PrP^c cells compared to treated PrP^ko cells. On the contrary, the 41 kDa PKCδ cleaved band was less pronounced in H₂O₂-treated PrP^c cells as compared to PrP^ko cells. Very little or no cleavage was observed in untreated cells. Further, oxidative and ER stress-induced proteolytic cleavage of PKCδ was also isoform-specific, as H₂O₂, BFA, and TUN treatment failed to induce proteolytic cleavage of PKCα (data not shown). In order to further, confirm that PKCδ cleavage is mediated by caspase-3, we used the caspase-3-specific inhibitor Z-DEVD-FMK to block cleavage. Cotreatment with 100 μM Z-DEVD-FMK for 12 h almost completely prevented the appearance of the 41 kDa catalytically active PKCδ fragment in 100 μM H₂O₂, 10 μM BFA, and 2μM TUN-treated PrP^c cells (data not shown). Nitrocellulose membranes were reprobed with β-actin antibody and the density of the 43 kDa β-actin band was identical in all lanes, confirming equal protein loading.

Caspase-3 mediates oxidative and ER stress-induced increases in PKCδ kinase activity

Immunoprecipitation kinase assays were performed to determine if the H₂O₂-, BFA-, and TUN-induced caspase-3-dependent accumulation of PKCδ cleaved product is attributed to an increase in PKCδ enzyme activity. PKCδ immunoprecipitates from cytosolic fractions were assayed for PKCδ kinase activity in the absence of DAG and phosphatidyl serine to solely determine the kinase activity of cleaved catalytic PKCδ fragment. The PKCδ enzymatic activity increased at 12 h exposure to 100 μM H₂O₂, 10 μM BFA, and 2 μM TUN in both PrP^c and PrP^ko cells (Fig. 5A). Densitometric analysis of phosphorylated histone H1 substrate bands in Fig. 5A revealed that increases in PKCδ kinase activity were more pronounced in BFA- and TUN-treated PrP^c cells compared to PrP^ko cells. On the other hand, increases in PKCδ kinase activity were less pronounced in H₂O₂-treated PrP^c cells compared to PrP^ko cells. Very little or no increase in kinase activity was observed in untreated cells. Cotreatment with 100 μM Z-DEVD-FMK significantly blocked H₂O₂-, BFA-, and TUN-induced increases in PKCδ kinase activity in PrP^c and PrP^ko cells (Fig. 5B). We attribute the increased kinase activity in H₂O₂-, BFA-, and TUN-treated cells to the persistently active proteolytically cleaved PKCδ catalytic fragment.

Fig. 5.

Effect of PrP^c on H₂O₂-, BFA-, and TUN-induced PKCδ kinase activity. (A) PKCδ kinase assay and (B) Effect of Z-DEVD-FMK. Subconfluent undifferentiated PrP^c and PrP^ko cells were harvested at 12 h post-treatment with 100 μM H₂O₂, 10–30 μM BFA, or 2–6 μM TUN with or without the inclusion of 100 μM Z-DEVD-FMK for 12 h. For in vitro PKCδ activity, substrate (histone H1) and [γ-³²P] ATP were added to the immunoprecipitated samples. The immunoprecipitation kinase assay was performed as described. The bands were quantified with a PhoshoImager after scanning the dried gel, and are expressed as percent of control (untreated cells). Data represent the mean ± SEM, N = 5, from two independent experiments. Asterisks (*p<0.05 and **p<0.01) indicate significant differences between treated and respective untreated control cells.

Effect of PrP^c on oxidative and ER stress-induced apoptotic cell death

To understand the functional consequence of the activation of many apoptotic factors including caspases and PKCδ, we examined oxidative and ER stress-induced DNA fragmentation. Chromosomal breakdown of DNA into 200-base pair nucleosomal fragments and DNA condensation are hallmarks of cells undergoing apoptosis [12, 13]. As shown in Fig. 6, exposure to 100 μM H₂O₂, 10 μM BFA, or 2 μM TUN for 24 h induced DNA fragmentation by 152, 355, and 379% in PrP^c and 244, 229, and 225% in PrP^ko cells, respectively, compared to untreated controls. BFA- and TUN-induced increases in DNA fragmentation were significantly higher in PrP^c cells, and H₂O₂-induced increases were significantly less in PrP^c cells compared to PrP^ko cells. Together, these results suggest that PrP^c protects against H₂O₂-induced increases, but exacerbates BFA- and TUN-induced increases in DNA fragmentation.

Fig. 6.

Effect of PrP^c on H₂O₂-, BFA-, and TUN-induced DNA fragmentation. PrP^c and PrP^ko cells were treated with 100 μM H₂O₂, 10 μM BFA, or 2 μM TUN; cells were lysed after 24 h treatment and assayed for DNA fragmentation as described in the Materials and Methods section. Data represent the mean ± SEM, N = 6, from two independent experiments performed in triplicate. Asterisks (*p<0.05 and **p<0.01) indicate significant differences between treated and respective untreated control cells.

Pharmacological dissection of ER and oxidative stress-induced cell death signaling mechanisms

Based on time-course of activation, sequential activation of caspase-12, caspase-8, caspase-9, and caspase-3, proteolytic activation of PKCδ, and apoptotic cell death occur in BFA- and TUN-treated cells, whereas in H₂O₂-treated cells, ROS is generated first and is followed by caspase-9, caspase-3, proteolytic activation of PKCδ and apoptotic cell death. To confirm this sequential activation, we used pharmacological inhibitors directed against ROS production (2 μM MnTBAP, SOD mimetic), mitochondrial permeability transition pore (1 μM cyclosporine A, CsA), caspase-8 (100 μM Z-IETD-FMK), caspase-9 (100 μM Z-LEHD-FMK), caspase-3 (100 μM Z-DEVD-FMK), and PKCδ (2 μM rottlerin). Co-treatment of PrP^c cells with these inhibitors did not prevent BFA- and TUN-induced loss of the 50 kDa procaspase-12 band in western blot experiments (data not shown). These results suggest that caspase-12 activation is not influenced by inhibitors directed against ROS, mitochondrial permeability transition pore, caspase-8, caspase-9, caspase-3 or PKCδ during in BFA- and TUN-treated PrP^c cells.

Next, we examined the effect of these inhibitors on H₂O₂- (Fig. 7A), BFA- (Fig. 7B), and TUN-induced (Fig. 7C) caspase-3 activation in PrP^c cells. As shown in Fig. 7, a 12 h cotreatment with either CsA or Z-LEHD-FMK significantly blocked H₂O₂-, BFA-, and TUN-induced increases in caspase-3 activation. Cotreatment with MnTBAP significantly blocked H₂O₂-, but not BFA- and TUN-induced increases in caspase-3 activation; however, cotreatment with Z-IETD-FMK significantly suppressed BFA- and TUN-, but not H₂O₂-induced increases in caspase-3 activation. These results suggest that MPT pore and caspase-9 mediate BFA-, TUN-, and H₂O₂-induced increases in caspase-3 activation, whereas ROS mediates only H₂O₂-induced increases, and caspase-8 mediates BFA- and TUN-induced increases in caspase-3 activity.

Fig. 7.

Effect of ROS-, MPT pore-, and caspase inhibitors on caspase-3 activation. Briefly, PrP^c cells were treated with (A) 100 μM H₂O_2, (B) 10 μM BFA, or (C) 2 μM TUN, in the presence or absence of 2 μM MnTBAP (ROS inhibitor), 2 μM CsA (MPT pore inhibitor), 100 μM Z-IETD-FMK (caspase-8 inhibitor), or 100 μM Z-LEHD-FMK (caspase-9 inhibitor). After a 12 h treatment, caspase-3 enzymatic activity was measured in cell samples as described. Data represent the mean ± SEM, N = 6, from two independent experiments performed in triplicate. Pound sign (## p<0.01) indicates significant differences between H₂O₂-, BFA-, and TUN-treated and inhibitor-treated groups.

Finally, we examined the effect of these inhibitors on H₂O₂- (Fig. 8A), BFA- (Fig. 8B), and TUN- (Fig. 8C) induced increases in DNA fragmentation. As shown in Fig. 8, cotreatment with inhibitors directed against mitochondrial permeability transition pore, caspase-9, caspase-3, and PKCδ significantly attenuated the H₂O₂-, BFA-, and TUN-induced increases in DNA fragmentation. Cotreatment with the caspase-8 inhibitor Z-IETD-FMK almost completely attenuated BFA- and TUN-, but not H₂O₂-induced increases in DNA fragmentation. Conversely, cotreatment with the ROS inhibitor MnTBAP significantly blocked H₂O₂-, but not BFA- and TUN-induced increases in DNA fragmentation over a 24 h treatment. To confirm the key results obtained with PrP^c cells, PrP^ko cells were exposed to 10 μM BFA and 2 μM TUN for 24 h in the presence of pharmacological inhibitors. As shown in Fig 9, rottlerin (2 μM), Z-IETD-FMK (100μM), and Z-DEVD-FMK (100μM) all significantly reduced BFA and TUN-induced increases in DNA fragmentation to that of untreated control PrP^ko levels. Together, these results suggest that oxidative stress initiates apoptotic cell signaling by generating ROS, and ER-stress initiates caspase-12 activation followed by caspase-8 activation, indicating both oxidative and ER stress share a common apoptotic cell signaling pathway that is mediated by mitochondria, caspase-9, caspase-3, and PKCδ.

Fig. 8.

Effect of ROS-, MPT pore-, caspase-, and PKCδ-inhibitors on apoptotic cell death in PrP^c expressing cells. PrP^c cells were treated with (A) 100 μM H₂O₂, (B) 10 μM BFA, or (C) 2 μM TUN, in the presence or absence of 2 μM MnTBAP (ROS inhibitor), 2 μM CsA (MPT pore inhibitor), 100 μM Z-IETD-FMK (caspase-8 inhibitor), 100 μM Z-LEHD-FMK (caspase-9 inhibitor), 100 μM Z-DEVD-FMK, or 2 μM rottlerin (PKCδ inhibitor). After 24 h treatment, DNA fragmentation was measured using an ELISA sandwich assay as described in the Materials and Methods section. Data represent the mean ± SEM, N = 6, from two independent experiments performed in triplicate. Pound sign (## p<0.01) indicates significant differences between H₂O₂-, BFA-, and TUN-treated and inhibitor-treated groups.

Fig. 9.

Effect of caspase and PKCδ-inhibitors on apoptotic cell death in PrP^ko cells. PrP^ko cells were treated with 10 μM BFA or 2 μM TUN, in the presence or absence 100 μM Z-IETD-FMK (caspase-8 inhibitor), 100 μM Z-DEVD-FMK, or 2 μM rottlerin (PKCδ inhibitor). After 24 h treatment, DNA fragmentation was measured using an ELISA sandwich assay as described in the Materials and Methods section. Data represent the mean ± SEM, N = 3, from one experiment performed in triplicate. Asterix sign (** p<0.01) indicates significant differences between BFA- and TUN-treated and untreated control cells. Pound sign (## p<0.01) indicates significant differences between H₂O₂-, BFA-, and TUN-treated and inhibitor-treated groups.

Overexpression of PKCδ^D327A-CRM and PKCδ^K376R-DNM rescue PrP^c cells from ER- and Oxidative stress-induced apoptotic cell death

We used loss-of-function approaches to demonstrate that caspase-3-dependent proteolytic activation of PKCδ mediates apoptotic cell death in PrP^c cells. We used two PKCδ mutants to evaluate the loss-of-function approach: the PKCδ^K376R-DNM mutant in which the caspase-3 cleavage site is intact but the catalytic site is mutated to render the kinase inactive, and the PKCδ^D327A-CRM mutant in which the cleavage site is mutated to render PKCδ resistant to cleavage by caspase-3 with the catalytic site intact (Fig. 10A). We examined whether H₂O₂-, BFA-, and TUN-induced apoptotic cell death is suppressed in PrP^c cells transiently expressing PKCδ^D327A-CRM and PKCδ^K376R-DNM mutants (Fig. 10B). The transfection efficiency was >75%. PrP^c cells expressing GFP alone were used as controls. Figure 10C shows that BFA- and TUN-induced DNA fragmentation was significantly reduced in PrP^c cells overexpressing PKCδ^K376R and PKCδ^D327A mutant proteins as compared to GFP vector-expressing cells. Figure 10C also shows similar results for H₂O₂-induced DNA fragmentation. Data were normalized to the % GFP expression in matched untreated Vector-GFP or PKCd^D327A-GFP, or PKCd^K376R-GFP expressing cells. Overall, these data suggest overexpression of PKCδ^K376R and PKCδ^D327A mutant proteins can rescue PrP^c cells from oxidative and ER stress-induced apoptotic cell death.

Fig. 10.

Overexpression of PKCδ^K376R-DNM and PKCδ^D327A-CRM mutant proteins rescue PrP^c cells from BFA- and TUN-induced apoptotic cell death. (A) Plasmid description, (B) transient expression, and (C) DNA fragmentation. Briefly, PrP^c cells transiently expressing caspase cleavage-resistant PKCδ^D327A-CRM, dominant negative PKCd^K376R-DNM, and GFP alone were exposed to 100 μM H₂O₂, 10 μM BFA, or 2 μM TFA for 24 h. DNA fragmentation was quantified using an ELISA assay as described in the Methods section. Data points represent mean ± SEM, N=6, from two independent experiments performed in triplicate. Asterisks (**p<0.01) indicate significant differences between treated and respective untreated GFP-vector expressing cells. Pound sign (## p<0.01) indicates significant differences between GFP vector-expressing cells and PKCδ^D327A-CRM or PKCδ^K376R-DNM-expressing PrP^c cells.

Discussion

In this study, we demonstrate the dual and opposing roles of PrP^c in mediating oxidative and ER stress-induced cell death. PrP^c plays a protective role during oxidative stress, and a proapoptotic role during ER stress-induced apoptotic cell death. This study also demonstrates that caspase-3-dependent proteolytic activation of PKCδ plays a role in both oxidative and ER stress-induced apoptotic cell death. To our knowledge, this is the first study showing proapoptotic function of PKCδ in cellular models of prion disease.

Previous studies have shown the antioxidant property of PrP^c during oxidative stress-induced cell death in cell culture and animal models of prion disease [22–27]. We found H₂O₂-induced increases in intracellular ROS levels were significantly attenuated in PrP^c expressing cells when compared to PrP^ko cells. ROS generation was observed as early as 15 min following H₂O₂ exposure in both cell types, indicating that ROS generation precedes the cytotoxic response. We attribute the partially protective effect of PrP^c on H₂O₂-induced ROS production to its antioxidant property. Further, exposure to BFA and TUN did not induce any increase in ROS generation up to 60 min in PrP^c cells, suggesting that ROS is a key mediator of H₂O₂-induced but not BFA- and TUN-induced cytotoxic response.

Previous reports have shown the involvement of ROS and caspases in mediating cell death induced by PrP^sc and the human PrP106–126 toxic fragment [16–18, 34, 47]. Perturbation of calcium homeostasis and accumulation of misfolded PrP^c proteins in the ER also elicit stress responses which are detrimental to the cells [48]. Molecular mechanisms linking ER stress and cell death in neuronal cell lines have recently been described in which activation of ER-resident caspase-12 leads to caspase-3 activation and cell death [15, 17, 43, 49]. However, the effect of PrP^c in oxidative and ER stress-induced activation of caspases and cell death is not well established. Here we show that H₂O₂-, BFA-, and TUN-induced cytotoxic cell death is mediated by caspases. Further, BFA- and TUN-induced activation of caspase-3 and caspase-9 was exacerbated in PrP^c cells compared to PrP^ko cells. Unlike BFA and TUN, H₂O₂-induced increases in caspase-9 and caspase-3 enzyme activities were significantly reduced in PrP^c cells. Additionally, BFA- and TUN-induced activation of caspase-8 and caspase-12 was also exacerbated in PrP^c cells. Caspase-8 and caspase-12 were not activated in either H₂O₂-treated PrP^c or PrP^ko cells. Time-course analysis revealed the activation sequence is caspase-12 (1 h), caspase-8 (3 h), caspase-9 (6 h), and caspase-3 (12 h) in BFA- and TUN-treated cells, while the activation sequence in H₂O₂-treated PrP^c cells was ROS generation followed by activation of caspase-9 and c aspase-3. Recently, Hetz et al. showed that caspase-12 and ER stress mediate neurotoxicity of PrP^sc [34]. In addition to caspase-12, other ER-resident caspases, such as caspase-4 have also been implicated in apoptotic cell death in infantile neuronal ceroid lipofuscinosis (INCL), a childhood neurodegenerative storage disorder [50]. On the contrary, studies have also shown caspase-12 and caspase-4 are not required for caspase-dependent ER stress-induced apoptotic cell death [51].

Recently, we and others established PKCδ as an emerging putative substrate for caspase-3 [44–46]. Here we show H₂O₂, BFA, and TUN exposures induced proteolytic activation of PKCδ in PrP^c and PrP^ko cells, which can be effectively blocked with the caspase-3 inhibitor Z-DEVD-FMK, indicating that the cleavage is mediated by caspase-3. However, the extent of PKCδ cleavage was significantly reduced in H₂O₂-treated PrP^c cells and significantly enhanced in BFA- and TUN-treated PrP^c cells compared to PrP^ko cells, suggesting that PrP^c is capable of attenuating oxidative stress-induced PKCδ cleavage and exacerbating ER stress-induced proteolytic activation of PKCδ. For the first time, we demonstrate PKCδ proteolytic activation during ER stress-induced apoptotic cell death.

DNA fragmentation and chromatin condensation are terminal events in the apoptotic process and are considered biochemical hallmarks of apoptosis [12, 13]. Previous reports have shown that PrP^sc and the human PrP106–126 toxic fragment induce DNA fragmentation [16–18, 34, 47]. Here we show the differential effect of PrP^c on oxidative and ER stress-induced apoptotic cell death. Quantitative DNA ELISA assays revealed H₂O₂, BFA, and TUN all induced DNA fragmentation in PrP^c and PrP^ko cells. Consistent with caspases and proteolytic cleavage of PKCδ, DNA fragmentation was significantly reduced in H₂O₂-treated cells, and significantly exacerbated in BFA- and TUN-treated PrP^c cells. Furthermore, pretreatment with pharmacological inhibitors directed against MPT pore, caspase-9, caspase-3, and PKCδ significantly suppressed H₂O₂-, BFA-, and TUN-induced DNA fragmentation in PrP^c cells, indicating that mitochondrial dysfunction, caspase-9, caspase-3, and PKCδ activities mediate the cell death events. Additionally, a ROS inhibitor selectively blocked H₂O₂-, but not BFA- and TUN-induced DNA fragmentation, whereas a caspase-8 inhibitor selectively blocked BFA- and TUN-, but not H₂O₂-induced DNA fragmentation. Further confirmation of the role of proteolytically activated PKCδ fragments in H₂O₂-, BFA-, and TUN-induced apoptosis is supported by loss-of-function dominant negative experiments performed in PrP^c cells overexpressing caspase cleavage resistant PKCδ^D327A and kinase dead PKCδ^K376R mutant proteins. H₂O₂, BFA, and TUN treatments all produced a significant increase in DNA fragmentation in GFP vector-expressing cells, and induced a moderate increase in DNA fragmentation in PKCδ^D327A or PKCδ^K376R overexpressing cells, thus confirming the contribution of proteolytically activated PKCδ in oxidative and ER stress-induced apoptotic cell death.

How PrP^c, a plasma membrane protein, helps to protect against oxidative stress and exacerbates ER stress-induced apoptotic cell death is unclear. Recently, a number of laboratories have reported altered subcellular distribution, misfolding, and trafficking of PrP^c protein during scrapie-infections and/or during stress [10, 52–60]. PrP^c misfolding at the plasma membrane results in endocytosis, intracellular retention, and degradation of PrP^c [61, 62]. Both proteasomes and ubiquitin are involved in the turnover of the wild-type prion protein [63]. Studies have also shown PrP^c positively regulates neural precursor proliferation during developmental and adult mammalian neurogenesis and is also expressed in long-term repopulating hematopoietic stem cells and is important for their self-renewal [64, 65]. Prion protein with a stop codon at Y145stop mutation was shown to be degraded very rapidly by the proteasome-mediated pathway, and blockage of proteasomal degradation resulted in the intracellular accumulation of PK-resistant PrP (Y145stop), suggesting that the mutant prion protein is degraded through the proteasomal pathway and acquires PK-resistance if degradation is impaired [66]. An amber mutation at codon 145 (Y145stop) of the PrP^c gene results in a variant of an inherited human prion disease known as Gerstmann-Straussler-Scheinker (GSS) syndrome. PrP^c with the Y145stop mutation also induces mitochondria-mediated apoptosis and PrP-containing deposits in vitro [67]. Conflicting studies have shown cytosolic accumulation of PrP^c is both not neurotoxic [68] and neurotoxic [39]. Further, PrP^c retrogradely transported out of the endoplasmic reticulum produced both amorphous aggregates and a PrP^sc-like conformation in the cytosol, suggesting that PrP^c has an inherent capacity to promote its own conformational conversion in mammalian cells [38]. However, other studies have shown PrP^c does not retrotranslocate from the ER into the cytoplasm prior to degradation by the proteasome [69]. In addition, PrP^c has been shown to contain cryptic nuclear localization signals, which may suggest abnormal PrP forms may accumulate in the nucleus and initiate neurotoxicity through hitherto unknown mechanisms [70]. Further, the apparent reduction in mono- and di-N-glycosylated PrP^c levels in BFA and TUN treated PrP^c cells observed in Western blot experiments could hurt cell survival in more than one way. Reduced PrP^c levels could lead to a loss of cell surface PrP^c resulting in reduced antioxidant capacity, and making the cells vulnerable to ER-stress, alternatively. Improperly processed PrP^c in the ER-could also elicit an unfolded protein response to facilitate its degradation. It is also possible that the retro-translocation of improperly processed PrP^c to the cytoplasm and nucleus may contribute to the exacerbated toxicity observed in BFA and TUN cells. Preliminary studies from our laboratory indicate that misfolded PrP^c is retrotranslocated to the cytosol and/or nucleus during ER stress, and during oxidative stress PrP^c with native confirmation is translocated to the cytosol from the plasma membrane (unpublished observations). We are currently investigating the hypothesis of whether accumulation of misfolded PrP^c in the cytosol or nucleus may be toxic, and whether accumulation of PrP^c with native conformation in the cytosol may be beneficial to the cell.

In conclusion, the present study emphasizes the opposing roles of PrP^c in mediating oxidative and ER stress-induced apoptotic cell death in spatial and temporal manner (see schematic Fig. 11). Pharmacological and genetic dissection revealed that ER-stressors BFA- and TUN 1.) activate caspase-12, caspase-8, caspase-9, caspase-3, and PKCδ in a sequential manner, and 2.) subsequently result in DNA fragmentation and cell death. Similarly, oxidative stressor H₂O₂ induces ROS production, cytochrome C release, activation of caspase-9, caspase-3, and PKCδ and subsequently DNA fragmentation and cell death. Further, there exists a cross-talk between ER and mitochondria during ER-stress but not during oxidative stress. For the first time, we demonstrate that PrP^c plays differential roles in cell death response i.e. a pro-apoptotic role during ER stress, and an anti-apoptotic role during oxidative stress-induced cell death. Our results also suggest that PKCδ may be a key downstream proapoptoic effector in cell death processes associated with misfolded prion protein-induced neurotoxicity.

Fig. 11.

A proposed model describing the apoptotic cell signaling events during oxidative and ER stress-induced cell death in PrP and PrP^ko cells. BFA and TUN induce activation of caspase-12 and caspase-8; H₂O₂ induces ROS production ; BFA, TUN, and H₂O₂ all activate caspase-9 which activates caspase-3, which then mediates proteolytic activation of PKCδ. Protealytically cleaved PKCδ induces DNA fragmentation and apoptotic cell death. The inhibitory action of pharmacological inhibitors: MnTBAP (ROS inhibitor); CsA, cyclosporine A (MPT, mitochondrial permeability transition pore inhibitor); Z-IETD-FMK (caspase-8 inhibitor); Z-LEHD-FMK (caspase-9 inhibitor); Z-DEVD-FMK (caspase-3 inhibitor); rottlerin (PKCδ inhibitor; and (loss-of-function dominant negative mutants (PKCδ^D327A-CRM and PKCδ ^K376R-DNM) are shown in the proposed pathway. In essence, our data suggest that PrP^c protects against oxidative stress and exacerbates ER stress-induced apoptotic cell death signaling events.

Abbreviations

BFA

brefeldin A

TUN

tunicamycin

H₂O₂

hydrogen peroxide

Z-VAD-FMK

Z-Val-Ala-Asp-Fluoro Methyl Ketone

Z-DEVD-FMK

Z-Asp-Glu-Val-Asp-Fluoro Methyl Ketone

Z-IETD-FMK

Z-Ile-Glu-Thr-Asp-Fluoro Methyl Ketone

Z-LEHD-FMK

Z-Leu-Glu-His-Asp-Fluoro Methyl Ketone

Ac-DEVD-AFC

Acetyl Asp-Glu-Val-Asp-AFC

Ac-IETD-AFC

Acetyl Ile-Glu-Thr-Asp-AFC

Ac-LEHD-AFC

Acetyl-Leu-Glu-His-Asp-AFC

MnTBAP

Mn(III)tetrakis(4-benzoic acid)porphyrin chloride

ROS

reactive oxygen species

ER

endoplasmic reticulum

CsA

cyclosporine A

MPT

mitochondrial permeability transition pore

GFP

green fluorescent protein

PrP^c

prion-expressing neural cells

PrP^ko

neural cells lacking the prion protein (knockout)

PKCδ

protein kinase Cδ