Oxidative stress impairs autophagic flux in prion protein-deficient hippocampal cells

Abstract

We previously reported that autophagy is upregulated in Prnp-deficient (Prnp^0/0) hippocampal neuronal cells in comparison to cellular prion protein (PrP^C)-expressing (Prnp^+/+) control cells under conditions of serum deprivation. In this study, we determined whether a protective mechanism of PrP^C is associated with autophagy using Prnp^0/0 hippocampal neuronal cells under hydrogen peroxide (H₂O₂)-induced oxidative stress. We found that Prnp^0/0 cells were more susceptible to oxidative stress than Prnp^+/+ cells in a dose- and time-dependent manner. In addition, we observed enhanced autophagy by immunoblotting, which detected the conversion of microtubule-associated protein 1 light chain 3 β (LC3B)-I to LC3B-II, and we observed increased punctate LC3B immunostaining in H₂O₂-treated Prnp^0/0 cells compared with H₂O₂-treated control cells. Interestingly, this enhanced autophagy was due to impaired autophagic flux in the H₂O₂-treated Prnp^0/0 cells, while the H₂O₂-treated Prnp^+/+ cells showed enhanced autophagic flux. Furthermore, caspase-dependent and independent apoptosis was observed when both cell lines were exposed to H₂O₂. Moreover, the inhibition of autophagosome formation by Atg7 siRNA revealed that increased autophagic flux in Prnp^+/+ cells contributes to the prosurvival effect of autophagy against H₂O₂ cytotoxicity. Taken together, our results provide the first experimental evidence that the deficiency of PrP^C may impair autophagic flux via H₂O₂-induced oxidative stress.

Keywords: :

• prion
• Prnp-deficient
• autophagy
• autophagic flux
• apoptosis
• oxidative stress
• neuron

Introduction

Prion diseases or transmissible spongiform encephalopathies (TSEs) are a group of fatal neurodegenerative disorders in animals and humans. These disorders are characterized by the accumulation of the pathological scrapie isoform (PrP^Sc) protein, which is post-translationally generated from PrP^C.¹ PrP^C is a glycosylphosphatidylinositol (GPI)-anchored, membrane-bound protein encoded by Prnp and is highly conserved in mammals.² The expression of PrP^C is most abundantly detected in neuronal cells, but it has also been observed in non-neuronal cells.^3,4 Therefore, PrP^C may play an important role in the central nervous system, although its exact physiological function remains unknown. Prnp^0/0 mice are comparatively healthy and do not develop a prion-like disease.⁵ These mice have also been reported to show abnormalities in synaptic physiology,⁶ circadian rhythms, and sleep.⁷ In the last several years, various biological functions for PrP^C have been suggested, including neurotransmitter metabolism,⁸ signal transduction,⁹ copper metabolism,¹⁰ cell adhesion,¹¹ neuritogenesis,¹² antioxidant activity,¹³ and immune cell activation.¹⁴ Moreover, many studies have suggested that PrP^C plays a neuroprotective role in response to serum deprivation and oxidative stress in neuronal and non-neuronal cell lines.^15-17 In response to oxidative stress, primary neurons cultured from the brains of Prnp^0/0 mice are more susceptible to cell death than those cultured from the brains of Prnp^+/+ mice.^18,19 However, the molecular mechanism related to the protective role of PrP^C against oxidative stress is unclear.

Autophagy is a major lysosomal degradation pathway for the elimination of damaged or long-lived proteins and organelles and plays important roles in physiology and pathophysiology in all cell types. Autophagy was originally described as a response to nutrient deprivation, thus aiding in cell survival through the turnover and recycling of cytoplasmic contents.²⁰ Autophagy is activated during the differentiation and development of multicellular organisms, such as flies, worms and slime molds,²¹ and is also involved in cancer, neurodegenerative diseases, pathogen infections, liver injury and myopathies.^22,23 Despite its role in cell survival, autophagy is also associated with cell death. Indeed, cell death occurs through three mechanisms: apoptosis (Type I), autophagic cell death (Type II) and necrosis (Type III).²⁴ Although autophagic cell death is caspase-independent and is clearly distinct from apoptotic cell death, these mechanisms can occur in mixed forms with autophagic and caspase-related apoptotic features.^25-28

Ultrastructural analyses of prion-infected cultured cells²⁹ and neurons from the brain tissues of TSE-affected patients and rodents³⁰ show the presence of autophagic vacuoles and multivesicular bodies. In addition, stimulator of chondrogenesis 1/scrapie responsive gene 1 (SCRG1) expression is increased in the autophagic vacuoles of neurons from scrapie-infected mice.³¹ Recently, it is reported that autophagy induction achieved by pharmacological autophagy inducers, such as trehalose or lithium, can reduce PrP^Sc expression in persistently prion-infected neuronal cells.^32,33 Interestingly, the ectopic overexpression of PrP protein 2 (dublet)/PrP-like protein doppel (PRND) in PrP^C-deficient central neurons of Ngsk prion protein-deficient (NP^0/0) mice showed that the expression levels of SCRG1 and the autophagic markers LC3B-II and sequestosome 1 (SQSTM1)/p62 were increased both before and during Purkinje cell death, whereas the mRNA expression levels of Lc3b and Sqstm1/p62 remained stable. These results may reflect an impairment of autophagic flux in the Purkinje cells of NP^0/0 mice.³⁴ A recent report revealed that downregulation of PrP^C expression using antisense oligonucleotides targeting the PRNP transcript led to autophagy-dependent cell death, with the absence of markers of apoptotic cell death in human malignant glioma cell lines and nonglial tumor cells.³⁵ This report suggested that PrP^C could directly modulate the autophagy-dependent cell death pathway.

In a previous report, we showed that autophagy is increased in Prnp^0/0 hippocampal neuronal cells compared with wild-type control cells under serum deprivation conditions. This increased autophagy is suppressed in Prnp^0/0 cells transfected with the Prnp gene.³⁶ However, it is unclear whether increased autophagy contributes to the activation of cell survival or cell death pathways. In transformed and cancer cell lines, oxidative stress-induced autophagic cell death persisted upon treatment with reactive oxygen species (ROS)-generating agents, such as hydrogen peroxide (H₂O₂) and 2-methoxyestradiol.³⁷ Low levels of such ROS as superoxide (O₂^•–), hydroxyl radical (HO^•) and H₂O₂ can serve as signaling molecules in various cellular pathways that lead to growth and survival. However, high levels of ROS can promote cellular damage and cell death through the oxidation of lipids, proteins and DNA.^38,39

To investigate further a mechanism for the protective role of PrP^C in oxidative stress, we examined the alteration and differentiation of autophagy under H₂O₂-induced oxidative stress in hippocampal neuronal cell lines expressing PrP^C (Prnp^+/+) and a control vector (Prnp^0/0).^36,40 In addition, we explored whether the inhibition or induction of autophagy could promote cell survival or cell death. In these experiments, we show that the expression of the autophagy marker LC3B-II was more increased in Prnp^0/0 neuronal cells than in Prnp^+/+ cells under oxidative stress conditions. Furthermore, we demonstrated that the enhanced expression of LC3B-II in Prnp^0/0 cells is due to impaired autophagic flux. Moreover, we show that increased autophagy may be associated with cell survival and cell death in Prnp^+/+ and Prnp^0/0 cells, respectively, under oxidative stress. Our findings suggest that PrP^C not only plays a novel protective role against oxidative stress-induced cell death but may also regulate autophagic flux during oxidative stress in neuronal cells.

Results

Prnp^0/0 neuronal cells exhibit increased susceptibility to H₂O₂-induced oxidative stress

In a previous report, it was shown that cultured primary neurons of Zürich I Prnp^0/0 mice are more susceptible to H₂O₂-induced oxidative stress than are Prnp^+/+ neurons.¹⁸ Recently, we established Zürich I Prnp^0/0 hippocampal neuronal cell lines, which are stably transfected with a vector encoding murine Prnp (Prnp^+/+ cells) or an empty vector (Prnp^0/0 cells).³⁶ The cells transfected with the empty vector were shown to be Prnp^0/0, while the cells transfected with Prnp were used as PrP^C controls. To confirm the expression levels of PrP^C in Prnp^+/+ and Prnp^0/0 cells, we performed western blot analysis using embryonic hippocampus lysates from ICR mice and previously established ICR hippocampal neuronal cell lines (ZW 13-1, ZW 13-2 and ZW 13-3). The results showed that PrP^C expression was slightly lower in the Prnp^+/+ cells than in the ICR hippocampal neuronal cell lines, whereas the expression level of PrP^C in the Prnp^+/+ cells was higher than in the ICR embryonic hippocampus (Fig. 1A). Using Prnp^+/+ and Prnp^0/0 cell lines, we investigated the neuroprotective role of PrP^C under H₂O₂-induced oxidative stress using light microscopy and a 4-[3-(4-iodophenyl)-2-(4-nitrophenyl)-2H-5-tetrazolio]-1,3-benzene disulfonate (WST-1)-based assay. As shown in Figure 1B and C, the morphological features of Prnp^+/+ and Prnp^0/0 cells exposed to H₂O₂ revealed irregular shrinkage and cell rounding compared with the untreated control cells. Although the number of cells attached to the surfaces of culture dishes gradually decreased in both cell lines, there were fewer attached Prnp^0/0 cells than Prnp^+/+ cells in a dose- and time-dependent manner following H₂O₂ treatment. Subsequently, we examined whether the cell lines differed in susceptibility to H₂O₂ toxicity using a WST-1-based assay. Although treatment with H₂O₂ significantly decreased cell viability in both Prnp^+/+ (doses of 500–1,000 µM) and Prnp^0/0 (doses of 250–1,000 µM) cells, the Prnp^0/0 cells were significantly more susceptible to the H₂O₂ toxicity (Fig. 1D). Moreover, the viability of Prnp^0/0 cells was significantly lower than that of Prnp^+/+ cells in a time-dependent manner (3–12 h) after treatment with 500 µM H₂O₂ (Fig. 1E). These findings clearly demonstrate that Zürich I Prnp^0/0 neuronal cells are more susceptible to H₂O₂-induced oxidative stress than Prnp^+/+ control cells, indicating that PrP^C plays a protective role against H₂O₂ cytotoxicity.

Figure 1. The protective effect of PrP^C against H₂O₂-induced cell death. (A) The expression of PrP^C in distinct stable cell lines and the embryo hippocampus of ICR mice. Western blot analysis was conducted using anti-PrP (3F10) in lysates from Zürich I Prnp^0/0 hippocampal neuronal cell lines transfected with murine Prnp (Prnp^+/+) or with an empty vector (Prnp^0/0), several hippocampi from 14-d-old embryos of ICR mice and 3 different WT hippocampal neuronal cell lines of ICR mice (ZW 13–2, ZW 13–2 and ZW 13–3). The expression of PrP^C was quantified using densitometry analysis (right panel). (B and C) Morphological appearance of Prnp^+/+ and Prnp^0/0 cells as observed under an inverted microscope. In (B and D), Prnp^+/+ and Prnp^0/0 cells were incubated with the indicated concentrations of H₂O₂ for 6 h. In (C and E), cells were treated with 500 μM H₂O₂ for the indicated time periods. (D and E) Cell viability was measured using the WST-1-based assay. The data are presented as the means ± standard deviation (S.D.) of 3 independent experiments performed in triplicate. *p < 0.02 and **p < 0.001, significant differences between treated and untreated cells; ^†p < 0.01 and ^‡p < 0.001, significant differences between Prnp^+/+ and Prnp^0/0 cells.

Prnp^0/0 neuronal cells are more susceptible to autophagic induction during H₂O₂-induced oxidative stress

To determine whether H₂O₂ cytotoxicity is associated with autophagy, we measured the expression of endogenous LC3B-II using western blot analysis. During the induction of autophagy, the cytoplasmic form of LC3B (LC3B-I, 18 kDa) is converted to the autophagosomal membrane-bound form (LC3B-II, 16 kDa). This conversion is a reliable marker for autophagy activation.⁴¹ As shown in Figure 2A and B, no differences in the basal expression levels of LC3B-II were detected between the untreated Prnp^+/+ and Prnp^0/0 cells. However, increased expression levels of LC3B-II were observed at doses of 250–750 µM H₂O₂ in both cell lines (Fig. 2A). In addition, enhanced expression of LC3B-II was detected at 3–24 h after treatment with 500 µM H₂O₂ in both cell lines (Fig. 2B). Interestingly, the expression of LC3B-II in the Prnp^+/+ cells was lower than that in the Prnp^0/0 cells after H₂O₂ treatment in a dose- and time-dependent manner (Fig. 2A and B).

Figure 2. The induction of autophagy in Prnp^+/+ and Prnp^0/0 cells exposed to H₂O₂. (A) Cells were incubated with the indicated concentrations of H₂O₂ for 6 h. (B) Cells were treated with 500 μM H₂O₂ for the indicated time periods. Cell lysates were subjected to western blot analysis using anti-LC3B or anti-β-actin antibodies (A and B), and the expression of LC3B-II was quantified using densitometry analysis (right panels). At least 3 independent experiments were used to assess the percentage of LC3B-II expression relative to that of β-actin. (C) Immunofluorescence microscopy analysis using anti-LC3B antibody and DAPI counterstaining in both Prnp^+/+ and Prnp^0/0 cells treated with 500 μM H₂O₂ for 3 h. The average number of LC3B puncta per cell is shown in the lower panel. At least 60 cells in several fields were counted in each experiment. The data represent the means ± SD of 3 independent experiments. *p < 0.05 and **p < 0.01, significant differences between H₂O₂-treated and untreated cells; ^†p < 0.05 and ‡p < 0.01, significant differences between Prnp^+/+ and Prnp^0/0 cells. Scale bar: 20 µm.

When autophagy was induced, LC3B was redistributed from a cytosolic diffuse pattern to a punctate staining pattern in the vacuolar membranes, which indicates the presence of autophagosomes.⁴¹ Therefore, we counted the number of endogenous LC3B-positive puncta per cell in both Prnp^+/+ and Prnp^0/0 cells exposed to 500 µM H₂O₂ for 3 h using immunofluorescence microscopy and an anti-LC3B antibody. Upon exposure to H₂O₂, the average number of LC3B puncta per cell was higher in the Prnp^0/0 cells than in the Prnp^+/+ cells (Fig. 2C). These results suggest that H₂O₂-induced cell death is associated with enhanced levels of autophagy in both neuronal cell lines.

H₂O₂-induced oxidative stress is impaired during the early stages of autophagic flux in Prnp^0/0 neuronal cells

To confirm further whether the H₂O₂-induced enhancement of autophagy arose from autophagy stimulation or impaired autophagic flux, we used a vacuolar H⁺-ATPase inhibitor, bafilomycin A₁ (Baf A₁), which prevents lysosome acidification and autophagosome-lysosome fusion. Prnp^+/+ and Prnp^0/0 cells were treated in the absence or presence of Baf A₁ combined with H₂O₂ for various time periods (3, 6 and 12 h). Treatment with Baf A₁ in combination with H₂O₂ increased the expression of LC3B-II at each time point (3–12 h) in the Prnp^+/+ cells compared with H₂O₂ treatment alone, indicating that H₂O₂ did not inhibit lysosomal activity or autophagic flux in the Prnp^+/+ cells (Fig. 3). In contrast, the expression of LC3B-II at early time points (3 and 6 h) in Prnp^0/0 cells treated with Baf A₁ and H₂O₂ was similar to that of Prnp^0/0 cells treated with H₂O₂ alone, suggesting that H₂O₂ treatment impairs autophagic flux in Prnp^0/0 cells. Interestingly, this impaired autophagic flux is rescued at a later time point (12 h) in Prnp^0/0 cells. To confirm whether PrP^C is associated with autophagic flux, we examined basal autophagic flux in the absence or presence of Baf A₁ without H₂O₂ in both cell lines. Interestingly, the expression of LC3B-II was higher in both cell lines treated with Baf A₁ for 3–12 h compared with untreated cells. This result indicates that the PrP^C deficiency did not affect the basal autophagy flux under normal culture conditions. Taken together, these data suggested that PrP^C is indirectly associated with autophagic flux under H₂O₂-induced oxidative stress.

Figure 3. Impaired autophagic flux in Prnp^0/0 cells exposed to H₂O₂. Cells were incubated with or without 500 μM H₂O₂ in the presence or absence of 20 nM bafilomycin A₁ (Baf A₁) for the indicated time periods. Cell lysates were subjected to western blot analysis using anti-LC3B or anti-β-actin antibody (upper panel), and the expression of LC3B-II was quantified using densitometry analysis (lower panel). At least 3 independent experiments were used to assess the percentage of LC3B-II expression relative to that of β-actin. The data represent the means ± SD of 3 independent experiments. *p < 0.05 and **p < 0.01, significant differences between Baf A₁-treated and untreated cells; ^†p < 0.05 and ^‡p < 0.01, significant differences between H₂O₂-treated and untreated cells.

H₂O₂-induced oxidative stress activates caspase-dependent and caspase-independent apoptosis at different time points in both Prnp^+/+ and Prnp^0/0 cells

Our previous study showed that Prnp^0/0 neuronal cells are more sensitive to autophagy and apoptosis than Prnp^+/+ neuronal cells, although expression levels of LC3-II and cleaved caspase 3 (CASP3) were elevated in both cell types under conditions of serum deprivation.³⁶ Therefore, we examined the generation of cleaved CASP3, a marker for apoptosis,⁴² in both cell types under exposure to H₂O₂. Interestingly, cleaved CASP3 was initially detected at a dose of 750 µM H₂O₂ or at 12 h after treatment with 500 µM H₂O₂ in both Prnp^+/+ and Prnp^0/0 cells (Fig. 4A and B). Unexpectedly, the expression of cleaved CASP3 in Prnp^+/+ cells was higher than that of Prnp^0/0 cells in a dose- and time-dependent manner. This result suggests that Prnp^+/+ cells are more susceptible to apoptotic cell death than Prnp^0/0 cells in the later stages of H₂O₂-induced oxidative stress.

Figure 4. Apoptosis activation in Prnp^+/+ and Prnp^0/0 cells exposed to H₂O₂. (A) Cells were incubated with the indicated concentrations of H₂O₂ for 6 h. (B) Cells were treated with 500 μM H₂O₂ for the indicated time periods. Cell lysates were subjected to western blot analysis with anti- CASP3 or anti-β-actin antibodies (A and B), and the expression of cleaved CASP3 was quantified using densitometry analysis (right panels). At least 3 independent experiments were used to assess the percentages of cleaved CASP3 expression relative to that of β-actin. ^†p < 0.05, significant differences between Prnp^+/+ and Prnp^0/0 cells. (C) Cells were preincubated with the indicated concentrations of zVAD-fmk for 3 h and treated with or without 500 μM H₂O₂ for 6 h. Cell viability was determined using the WST-1 reagent assay. The data are represented as the percent viability normalized to mock-treated control groups (100%). The data are presented as the means ± SD of 3 independent experiments performed in triplicate. (D) Cells were treated with 500 μM H₂O₂ for the indicated time periods and fractionated into cytoplasmic and nuclear fractions. Each fraction was analyzed by western blotting with anti-AIFM1, anti-ENDOG, anti-ENO1 (cytosolic fraction markers) and anti-HDAC2 antibody (nuclear fraction marker).

To confirm further the absence of CASP3-dependent apoptosis in H₂O₂ treatment at early time points (3–6 h), we investigated whether z-VAD-fmk, a pan-caspase inhibitor, could prevent H₂O₂-induced cell death. As expected, z-VAD-fmk treatment did not suppress H₂O₂-induced cell death in either cell type (Fig. 4C). These results are in agreement with the fact that cleaved CASP3 was not detected in either Prnp^+/+ or Prnp^0/0 cells at 3–6 h after treatment with 500 µM H₂O₂, which suggests that CASP3-dependent apoptosis does not contribute to H₂O₂-induced cell death at early stages.

Next, we investigated whether H₂O₂-induced cell death is associated with caspase-independent apoptosis using apoptosis-inducing factor, mitochondrion-associated, 1 (AIFM1) and endonuclease G (ENDOG). AIFM1 and ENDOG can directly activate nuclear DNA fragmentation when translocated from the mitochondrial intermembrane space to the nucleus through regulation of the pro-apoptotic BCL2 family.^43-46 Both AIFM1 and ENDOG are expressed as precursor proteins that are converted to mature forms upon mitochondrial import and cleavage of the N-terminal targeting sequence.^47,48 The cells were treated with H₂O₂ and fractionated into cytoplasmic and nuclear fractions. As shown in Figure 4D, the expression of mature AIFM1 was unchanged from 0 h to 12 h in both Prnp^+/+ and Prnp^0/0 cells, whereas mature ENDOG expression was markedly increased at 3 h. Enolase 1 (ENO1) and histone deacetylase 2 (HDAC2) were recovered exclusively in the cytoplasmic and nuclear fractions, respectively, confirming the adequate separation of cytoplasmic and nuclear fractions. This result indicates that H₂O₂-induced cell death is associated with ENDOG expression at an early stage (3 h) in both Prnp^+/+ and Prnp^0/0 cells. Taken together, these results suggest that H₂O₂ cytotoxicity is associated with caspase-independent apoptosis at an early stage (at 3 h after treatment with 500 µM H₂O₂) and caspase-dependent apoptosis at a later stage (at 12 h after treatment with 500 µM H₂O₂) in both cell types.

The administration of 3-MA reduced H₂O₂-induced autophagy, whereas treatment with rapamycin or trehalose promoted H₂O₂-induced autophagy in Prnp^0/0 cells

To evaluate the role of H₂O₂-induced autophagy, we investigated the effect of 3-methyladenine (3-MA), an inhibitor of phosphatidylinositol 3-kinase (PtdIns3K), which is a component of the autophagy pathway,⁴⁹ on H₂O₂-induced autophagy in both Prnp^+/+ and Prnp^0/0 cells. In this study, increased autophagy activation and reduced CASP3 cleavage were observed in both Prnp^+/+ and Prnp^0/0 cells at 6 h after treatment with 500 µM H₂O₂. We therefore examined the effect of 3-MA at 6 h after treatment with 500 µM H₂O₂ using western blot analysis and cell viability assays. Treatment with 3-MA slightly reduced the expression of LC3B-II in Prnp^0/0 cells treated with or without H₂O₂ and had no effect on LC3B-II expression in Prnp^+/+ cells (Fig. 5A). The WST-1 cell viability assay revealed that 3-MA significantly decreased H₂O₂-induced cell death in Prnp^0/0 cells but had no significant effects in Prnp^+/+ cells (Fig. 5B). Importantly, 3-MA is a widely used autophagy inhibitor that suppresses class III PtdIns3K activity.⁴⁹ However, it was recently reported that 3-MA can also activate autophagy under nutrient-rich conditions, such as normal culture conditions, via persistent inhibition of class I PtdIns3K, disruption of mechanistic target of rapamycin (MTOR) complex I activity and transient suppression of class III PtdIns3K.⁵⁰ Hence, we further explored the expression of phospho-AKT1, a major downstream target of class I PtdIns3K, to confirm the inhibition of class I PtdIns3K. As shown in Figure 5A, the treatment of 3-MA did not reduce the expression of phospho-AKT1, suggesting that the effect of 3-MA in Prnp^+/+ cells may occur through a class I PtdIns3K-independent pathway.

Figure 5. Effects of the autophagy inhibitor 3-MA and the autophagy inducers rapamycin (Rapa) and trehalose (Treha) on H₂O₂-induced cell death. (A–F) Prnp^+/+ and Prnp^0/0 cells were pretreated with or without 5 mM 3-MA, 10 nM Rapa or 100 mM Treha for 1 h (Rapa) or 3 h (3-MA and Treha) and with or without 500 μM H₂O₂ for 6 h. (A, C and E) Cell lysates were subjected to western blot analysis using an anti-LC3B, antibody in PathScan^® Multiplex Western Cocktail I to detect phosphorylated AKT1 or RPS6 or anti-β-actin antibody. (B, D and F) Cell viability was determined using the WST-1 reagent assay. The data are represented as the percent viability normalized to mock-treated control groups (100%). The data are presented as the means ± SD of 3 independent experiments performed in triplicate.

To study further the effect of H₂O₂-induced autophagy, we used two autophagy inducers, namely, rapamycin (MTOR-dependent)⁵¹ and trehalose (MTOR-independent).⁵² The MTOR-dependent and independent effects of these inducers were confirmed through the expression of phospho-ribosomal protein S6 (RPS6), a surrogate marker for MTOR activity, using western blot analysis (Fig. 5C and E). These two autophagy inducers markedly increased the expression levels of LC3B-II upon treatment with or without H₂O₂ in both Prnp^+/+ and Prnp^0/0 cells (Fig. 5C and E). Furthermore, these two autophagy inducers significantly enhanced H₂O₂-induced cell death in Prnp^0/0 cells. Likewise, trehalose significantly increased H₂O₂-induced cell death in Prnp^+/+ cells, although the effects of rapamycin in these cells were not significant (Fig. 5D and F). Taken together, these experiments suggest that the role of autophagy in H₂O₂ cytotoxicity might be associated with cell death in Prnp^0/0 cells. However, these results remain controversial, and further investigations are required to establish the effects of pharmacological inhibition or activation of autophagy in Prnp^+/+ and Prnp^0/0 cells.

The knockdown of Atg7 inhibits autophagy and promotes H₂O₂ cytotoxicity in Prnp^+/+ cells, but not Prnp^0/0 cells

The chemical inhibition of autophagy may be a consequence of the side effects from the drug used;⁵³ therefore, we assessed the role of autophagy in H₂O₂ cytotoxicity using an siRNA knockdown of Atg7, which is essential for the formation of autophagosomes. The expression of ATG7 protein was markedly decreased in Prnp^+/+ and Prnp^0/0 cells transfected with Atg7 siRNA. The knockdown of Atg7 also led to decreased expression levels of LC3B-II in both Prnp^+/+ and Prnp^0/0 cells treated with or without H₂O₂ (Fig. 6A). Interestingly, the knockdown of Atg7 resulted in enhanced H₂O₂-induced cell death in Prnp^+/+ cells but did not affect cell viability in the H₂O₂-treated Prnp^0/0 cells (Fig. 6B). Taken together, these results suggest that autophagy may be involved in cell survival in H₂O₂-treated Prnp^+/+ cells and that impaired autophagic flux in H₂O₂-treated Prnp^0/0 cells may contribute to cell death. Furthermore, the inhibition of autophagy may not affect cell death in H₂O₂-treated Prnp^0/0 cells because the autophagic flux is already impaired in Prnp^0/0 cells.

Figure 6. Augmentation of H₂O₂-induced cell death using Atg7 knockdown in Prnp^+/+ cells. (A and B) Prnp^+/+ and Prnp^0/0 cells were transfected with nontargeting (NT) or Atg7 siRNA, respectively. The siRNA-transfected cells were subsequently treated with or without 500 μM H₂O₂ for 6 h. (A) Cell lysates were subjected to western blot analysis using anti-ATG7, anti-LC3B or anti-β-actin antibody. (B) Cell viability was determined using the WST-1 reagent assay. The data are represented as the percent viability normalized to H₂O₂-untreated control groups (100%). The data are presented as the means ± SD of 3 independent experiments performed in triplicate.

Discussion

Building upon previous results,¹⁸ we showed that Prnp^0/0 cells are more susceptible to H₂O₂-induced oxidative stress than Prnp^+/+ cells. These results indicate that PrP^C has anti-oxidant activity. Indeed, many reports have shown that PrP^C can act as a superoxide dismutase 1 (SOD1)⁵⁴ and can modulate the activity of Cu/Zn SOD1 by binding to 5 Cu²⁺ ions on octapeptide repeat motifs in the N-terminal region.^10,55 Nevertheless, the molecular mechanism associated with the neuroprotective role of PrP^C against oxidative stress remains unclear. Therefore, we investigated whether neuronal autophagy and apoptosis are activated during H₂O₂-induced oxidative stress in Prnp^+/+ and Prnp^0/0 cells. In this study, we demonstrate that the induction of autophagy following H₂O₂ treatment was due to enhanced and impaired autophagic flux in Prnp^+/+ and Prnp^0/0 cells, respectively. Moreover, autophagy inhibition achieved by Atg7 siRNA promotes H₂O₂ cytotoxicity in Prnp^+/+ cells but does not affect Prnp^0/0 cells, indicating that the increased autophagic flux in Prnp^+/+ cells (at 3–12 h) and the impaired autophagic flux in Prnp^0/0 cells (at 3–6 h) contribute to neuronal cell protection and the acceleration of neuronal cell death, respectively, during H₂O₂-induced oxidative stress. Interestingly, the impaired autophagic flux is rescued at a later time (12 h), suggesting that reactivated or restored autophagic flux may play a protective role against H₂O₂-induced oxidative stress (Fig. 7).

Figure 7. Schematic representation of time-dependent H₂O₂-induced cell death in Prnp^+/+ and Prnp^0/0 cells. In 2 types of cells, caspase-independent and caspase-dependent apoptosis were observed through ENDOG and cleaved CASP3, respectively. Activated autophagic flux protects against oxidative stress in Prnp^+/+ cells, while impaired autophagic flux plays a role in neuronal cell death in Prnp^0/0 cells. This impaired autophagic flux is rescued at a later time point (12 h), suggesting that the reactivated or restored autophagic flux may play an additional protective role.

In addition, caspase-independent (at an early time point) and caspase-dependent apoptosis (at later time points) was observed through ENDOG and cleaved CASP3 in both Prnp^+/+ and Prnp^0/0 cells exposed to H₂O₂ (Fig. 7). This result is somewhat consistent with the previous finding that H₂O₂ treatment leads to ENDOG-dependent death but not CASP3 activation at an early time point in primary cortical neurons.⁵⁶ Unexpectedly, we found that Prnp^0/0 cells were less sensitive to apoptosis at later time points after treatment with H₂O₂ than Prnp^+/+ cells. This result suggests that the presence of PrP^C may promote CASP3-dependent apoptosis during H₂O₂-induced oxidative stress. However, this result remains controversial, and further investigation is required to establish the role of PrP^C in H₂O₂-induced CASP3-dependent apoptosis.

Although H₂O₂-induced oxidative stress increased autophagy in both cell types, it was found that autophagy induction resulted from enhanced and impaired autophagic flux in Prnp^+/+ and Prnp^0/0 cells, respectively. It was recently reported that H₂O₂ induces autophagic cell death with elevated autophagic activity in primary cortical neurons.⁵⁷ However, it is reasonable to propose that enhanced autophagy could eliminate oxidized or damaged intracellular components as a cellular defense mechanism against oxidative stress.

A number of studies have shown that autophagy dysfunction is directly associated with several neurodegenerative diseases, such as Alzheimer disease, Huntington disease, Parkinson disease, frontotemporal dementia and acute injuries.^58,59 This impairment of autophagy could result from alterations in the autophagy process, which is composed of several steps, resulting in neurodegeneration. However, concerning prion diseases, the evidence of impaired autophagic flux remains unclear. Oxidative stress is induced through ROS or free radicals and may play an important role in the pathogenesis of several neurodegenerative diseases.⁶⁰ Notably, increased markers of oxidative damage and decreased antioxidants are observed in the brains of prion diseases.^61,62 These changes imply that the conversion of PrP^C to the nonfunctional form PrP^Sc or the absence of PrP^C expression causes oxidative stress in prion diseases. Hence, the appearance of impaired autophagic flux through oxidative damage in prion diseases could accelerate PrP^Sc accumulation.

We observed inconsistent results in the viability assay using either pharmacological autophagy inhibition with 3-MA or induction using rapamycin and trehalose with H₂O₂ treatment, although the treatment of these drugs affected the Prnp^+/+ and Prnp^0/0 cells differently. However, the observed chemical inhibition and activation of autophagy may be a consequence of the side effects of the drug used. Hence, this result remains controversial, and further investigations are required to establish the effects of pharmacological autophagy inhibition and activation under H₂O₂-induced oxidative stress in neuronal cells. Interestingly, in H₂O₂-treated Prnp^+/+ cells, the viability was significantly decreased upon transfection with Atg7 siRNA compared with non-targeting (NT) siRNA. These data suggest that enhanced autophagy in H₂O₂-treated Prnp^+/+ cells is associated with a protective role. In contrast, in H₂O₂-treated Prnp^0/0 cells, Atg7 siRNA transfection did not affect cell viability. In fact, it is possible that reduced formation of autophagosomes through Atg7 siRNA could not rescue impaired autophagic flux because autophagic flux is already disrupted in H₂O₂-treated Prnp^0/0 cells. Hence, it is also possible that H₂O₂-induced oxidative stress may impair autophagosome/lysosome fusion or lysosomal degradation after successful formation of autophagosomes against H₂O₂ toxicity in Prnp^0/0 cells, as shown in the H₂O₂-treated Prnp^+/+ cells.

In addition, PrP^C may not be directly involved in autophagic flux because natural autophagic flux was shown in both Prnp^+/+ and Prnp^0/0 cells under normal culture conditions, suggesting that the deficiency of PrP^C may contribute to the impairment of autophagic flux under stress-induced conditions, such as H₂O₂ treatment or PRND overexpression. As previously described, it was shown in NP^0/0 mice that the impairment of autophagic flux resulted in the loss of Purkinje cells through the autophagic cell death. Interestingly, it has been reported that PRND alone is toxic to neurons, and this toxicity is associated with upregulated heme oxygenase 1 (HMOX1), an enzyme that mediates the formation of antioxidants, and nitric oxide synthases (NOSs), such as neuronal NOS (nNOS) and inducible NOS (iNOS), in cell culture models and mice.^63,64 Together, this evidence suggests that impaired autophagic flux in NP^0/0 mice could result from PRND toxicity-induced oxidative stress and the lack of PrP^C expression. However, we did not observe any relationship between PrP^C and autophagic flux in this study; therefore, further investigations are required to determine the relationship between PrP^C and autophagic flux.

Taken together, these results demonstrate that the deficiency of PrP^C contributes to H₂O₂-induced autophagic cell death in hippocampal neuronal cells via impaired autophagic flux. Furthermore, our present results provide insight into the protective mechanism of PrP^C in neuronal cells and the pathogenesis of prion diseases associated with autophagy.

Materials and Methods

Reagents

Reagents were obtained from commercial sources. The cell proliferation reagent WST-1 (11644807001) was purchased from Roche. H₂O₂ (1.07210) and methanol (1.13351) were obtained from Merck. Acetone (179124), 3-MA (M9281), trehalose dihydrate (T5251), rapamycin (R8781), bafilomycin A₁ (B1793) and dimethyl sulfoxide (DMSO; D2650) were purchased from Sigma-Aldrich. z-VAD-fmk (4800-520) was obtained from MBL. Dulbecco’s modified Eagle’s medium (DMEM; SH30243.01), fetal bovine serum (FBS, SH30396.03), Halt^TM protease inhibitor cocktail (87785) and Halt^TM phosphatase inhibitor cocktail (78420) were obtained from Thermo Scientific. Penicillin-streptomycin (15140) and puromycin (A11138-03) were purchased from Invitrogen. Hanks’ balanced salt solution (HBSS, LB203-04) was obtained from WelGENE. A bicinchoninic acid (BCA) protein assay kit (23225) and the Supersignal West Pico Chemiluminescent Substrate (34080) were purchased from Pierce.

Antibodies

Anti-PrP (3F10) antibody was employed as described in ref. 65. Anti-CASP3 (9662), anti-ATG7 (SAB1407006), anti-AIFM1 (4642S), anti-ENDOG (4969S) and PathScan Multiplex Western Cocktail I (5301), including anti-phospho-AKT1 and anti-phospho-RPS6, were purchased from Cell Signaling. Anti-LC3B (L7543) and anti-β-actin (A5441) antibodies were purchased from Sigma-Aldrich. Anti-LC3B antibody (L10382) for immunocytochemistry was purchased from Invitrogen. Anti-ENO1 antibody (LF-MA10098) was purchased from Ab Frontier. Anti-HDAC2 antibody (sc-9959) was purchased from Santa Cruz Biotechnology.

Cell culture

Prnp^+/+ and Prnp^0/0 cell lines were previously established as described in ref. 36. All cell lines were cultivated in DMEM supplemented with 10% FBS and 1% penicillin-streptomycin in a 5% CO₂ atmosphere at 37°C.

To analyze the effects of 3-MA, z-VAD-fmk, rapamycin and trehalose under H₂O₂ treatment, 8 × 10³ cells or 1 × 10⁶ cells were plated in 0.1 ml or 10 ml of culture media per well in a 96-well plate or 100-mm Petri dish, respectively. The next day, the various compounds were preincubated with cells for 1 h or 3 h before H₂O₂ treatment.

Cell viability assay

Cell viability was determined using a Cell Proliferation Reagent WST-1 (4-[3-(4-iodophenyl)-2-(4-nitrophenyl)-2H-5-tetrazolio]-1,3-benzene disulfonate)-based assay. Briefly, 8 × 10³ cells per well were plated in 96-well flat-bottomed plates with 0.1 ml of culture medium. The next day, the cells were incubated in fresh culture media without (control) or with designated concentrations of H₂O₂ for various durations. The Cell Proliferation Reagent WST-1 assay was performed according to the manufacturer’s instructions. The absorbance at 440 nm (A₄₄₀) was quantified using a plate reader. The A₄₄₀ value in the control cells was regarded as 100% viability.

Western blot analysis

To obtain lysates, cells (1 × 10⁶ cells per 100-mm Petri dish) were harvested through scraping in HBSS containing a protease inhibitor cocktail (Roche, 11697498001). The cell pellet was obtained after centrifugation at 2500 × g for 5 min at 4°C. The pellet was resuspended using a 1-ml syringe (26-gauge) in lysis buffer containing 10 mM TRIS-HCl (pH 7.4), 1 mM EDTA, 150 mM NaCl, 1% Nonidet P-40, 1% deoxycholic acid, 1 mM phenylmethyl sulfonyl fluoride, Halt^TM phosphatase inhibitor cocktail and protease inhibitor cocktail and was subsequently solubilized on ice for 30 min. To remove the insoluble materials, the cells were centrifuged at 5000 × g for 10 min at 4°C. The protein concentration was determined using a BCA protein assay, and equal amounts of protein were separated on 10% (ATG7, AIFM1, ENDOG, ENO1 and HDAC2), 12% (PathScan Multiplex Western Cocktail I) or 15% (PrP, LC3B, CASP3 and β-actin) SDS-PA gels and transferred to 0.45-μm nitrocellulose membranes. The blots were preincubated with 3% nonfat dried milk in Tris-buffered saline (TBS) containing 0.05% Tween 20 (TBS-T) for 1 h at room temperature and subsequently incubated with primary antibodies against ATG7 (1:1000), AIFM1 (1:1000), ENDOG (1:1000), ENO1 (1:5000), HDAC2 (1:200), PrP (1:3000), LC3B (1:1000), CASP3 (1:1000), PathScan Multiplex Western Cocktail I (1:200) or β-actin (1:10,000) for 16 h at 4°C or 1.5 h at room temperature (RT). The membranes were washed in TBS-T and subsequently incubated with species-specific secondary antibodies conjugated to horseradish peroxidase (goat anti-mouse IgG and goat anti-rabbit IgG; Pierce; dilution 1:3000) for 1 h at RT. The reactive protein bands were visualized using the Supersignal West Pico Chemiluminescent Substrate and quantified by densitometric analysis with ImageJ software.

Immunofluorescence microscopy

The cells were washed quickly with ice-cold HBSS and fixed using 4% paraformaldehyde in phosphate-buffered saline (PBS) for 15 min at RT. After 3 washes with PBS, the cells were permeabilized using ice-cold acetone:methanol (1:1) for 5 min at -20°C and then washed 3 more times with PBS. For endogenous LC3B staining, the cells were incubated with rabbit primary antibody against LC3B (dilution 1:1000 in 1% BSA blocking solution) for 2 h at RT. After washing 6 times with PBS, the cells were incubated with Alexa Fluor 568 goat anti-rabbit IgG antibody (Molecular Probes, A11011) for 2 h. After washing 6 times with PBS, the cells were subsequently mounted using Vectashield with DAPI (Vector Laboratories, H-1200). Imaging was performed using a Zeiss LSM-700 confocal laser-scanning microscope imaging system. The number of LC3B-positive puncta was determined using ImageJ software and expressed as the number of LC3B-positive puncta per cell.

siRNA knockdown

Nontargeting siRNA (sc-37007) and Atg7 siRNA (sc-41448) were purchased from Santa Cruz Biotechnology. Prnp^+/+ and Prnp^0/0 cells were seeded at 1.2 × 10⁶ cells per 100-mm Petri dish and cultured for 6 h in culture medium without antibiotics. Subsequently, the cells were transfected with 50 nM nontargeting siRNA or 50 nM Atg7 siRNA using Lipofectamine RNAiMAX (Invitrogen, 13778–150) for 18 h according to the manufacturer’s protocol. After transfection, the cells were seeded onto 96-well plates or a 100-mm Petri dish at a concentration of 8 × 10³ cells or 1 × 10⁶ cells, respectively. After 12 h, the cells were treated with H₂O₂ for 6 h.

Cell fractionation

Cell fractionation was performed using NE-PER^® Nuclear and Cytoplasmic Extraction Reagents (Thermo Scientific, 78833) according to the manufacturer’s instructions. ENO1 and HDAC2 served as cytosolic and nuclear markers, respectively.

Statistical analysis

All experiments were repeated at least 3 times. The data are expressed as the means ± standard deviation (S.D) All analyses were performed using the SIGMA STAT 3.5 software (Systat Software, Inc.). The statistical significance of the data was determined using one-way ANOVA followed by Tukey’s test.

Abbreviations:
PrPC	=	cellular prion protein
Prnp 0/0	=	Prnp-deficient
H2O2	=	hydrogen peroxide
LC3B	=	microtubule-associated protein 1 light chain 3 beta
TSE	=	transmissible spongiform encephalopathy
PrPSc	=	pathological scrapie isoform
GPI	=	glycosylphosphatidylinositol
3-MA	=	3-methyladenine
PtdIns3K	=	phosphatidylinositol 3-kinase
MTOR	=	mechanistic target of rapamycin
SOD1	=	superoxide dismutase 1
Baf A1	=	bafilomycin A1
AIFM1	=	apoptosis-inducing factor, mitochondrion-associated, 1
ENDOG	=	endonuclease G
NOS	=	nitric oxide synthase
RPS6	=	ribosomal protein S6
SQSTM1	=	sequestosome 1
HDAC2	=	histone deacetylase 2
CASP3	=	caspase 3
ENO1	=	enolase 1
SCRG1	=	stimulator of chondrogenesis 1/scrapie responsive gene 1
PRND	=	PrP protein 2 (dublet)/PrP-like protein doppel
HMOX1	=	heme oxygenase 1

Acknowledgments

This research was supported by Basic Science Research Program through the National Research Foundation of Korea (NRF) funded by the Ministry of Education, Science and Technology (2012-0000308) and by the National Research Foundation of Korea Grant funded by the Korean Government (NRF-2011-619-E0001).

Disclosure of Potential Conflicts of Interest

No potential conflicts of interest were disclosed.