Normal Cellular Prion Protein Protects against Manganese-Induced Oxidative Stress and Apoptotic Cell Death

Abstract

The normal prion protein is abundantly expressed in the central nervous system, but its biological function remains unclear. The prion protein has octapeptide repeat regions that bind to several divalent metals, suggesting that the prion proteins may alter the toxic effect of environmental neurotoxic metals. In the present study, we systematically examined whether prion protein modifies the neurotoxicity of manganese (Mn) by comparing the effect of Mn on mouse neural cells expressing prion protein (PrP^C-cells) and prion-knockout (PrP^KO-cells). Exposure to Mn (10μM–10mM) for 24 h produced a dose-dependent cytotoxic response in both PrP^C-cells and PrP^KO-cells. Interestingly, PrP^C-cells (EC₅₀ 117.6μM) were more resistant to Mn-induced cytotoxicity, as compared to PrP^KO-cells (EC₅₀ 59.9μM), suggesting a protective role for PrP^C against Mn neurotoxicity. Analysis of intracellular Mn levels showed less Mn accumulation in PrP^C-cells as compared to PrP^KO-cells, but no significant changes in the expression of the metal transporter proteins transferrin and DMT-1. Furthermore, Mn-induced mitochondrial depolarization and reactive oxygen species (ROS) generation were significantly attenuated in PrP^C-cells as compared to PrP^KO-cells. Measurement of antioxidant status revealed similar basal levels of glutathione (GSH) in PrP^C-cells and PrP^KO-cells; however, Mn treatment caused greater depletion of GSH in PrP^KO-cells. Mn-induced mitochondrial depolarization and ROS production were followed by time- and dose-dependent activation of the apoptotic cell death cascade involving caspase-9 and -3. Notably, DNA fragmentation induced by both Mn treatment and the oxidative stress inducer hydrogen peroxide (100μM) was significantly suppressed in PrP^C-cells as compared to PrP^KO-cells. Together, these results demonstrate that prion protein interferes with divalent metal Mn uptake and protects against Mn-induced oxidative stress and apoptotic cell death.

manganese, neurotoxicity, prion disease, metals, ROS, caspases

Prion-related diseases are a family of neurodegenerative disorders that affect various mammalian species including animals and humans (Collinge, 2001; Collinge and Palmer, 1992; Prusiner, 1998). Prion-related encephalopathies include conditions such as scrapie in sheep, bovine spongiform encephalopathy (BSE) in cattle, and Creutzfeldt–Jacob disease (CJD) and Gerstmann–Straussler–Scheinker syndrome, among others, in humans (Bendheim et al., 1985; Prusiner, 1989; Prusiner et al., 1985). Prion diseases cause severe neuronal damage mainly in the brain regions that control motor function, but can affect other areas, including the basal ganglia, cerebral cortex, thalamus, and cerebellum (Chiesa and Harris, 2001; Collinge, 2001; Mastrangelo and Westaway, 2001; Porter et al., 2000; Schreuder, 1994). The characteristic feature of prion diseases is conversion of normal cellular prion (PrP^C) to the abnormal pathogenic scrapie isoform (PrP^Sc) through a still undefined mechanism, which eventually results in the extracellular accumulation of PrP^Sc aggregates (Brandner et al., 2000; Collinge and Palmer, 1993; Kretzschmar, 1999; Prusiner and Kingsbury, 1985). The infectious PrP^Sc is thought to act as a template and promote the conversion of more PrP^C to PrP^Sc, resulting in the amplification of the infectious and toxic species. Thus, the normal cellular PrP^C protein is required for the pathogenesis of spongiform encephalopathies (Cohen and Prusiner, 1998; Kretzschmar, 1999; Prusiner et al., 1988).

PrP^C is a cell surface glycoprotein expressed predominantly in neurons and glia (Bazan et al., 1987; Hafiz and Brown, 2000; Rudd et al., 2001). Studies have shown PrP^C possesses intrinsic superoxide dismutase (SOD)–like activity, conferring antioxidant properties which may aid in cellular resistance to oxidative stress (Brown et al., 1997c, 1999). Although the exact mechanism by which PrP^C acts as an antioxidant is unclear, there is evidence that PrP^C contributes to the cellular SOD activity and to the maintenance of glutathione (GSH) levels inside the cell (Brown and Besinger, 1998; Brown et al., 1999; Choi et al., 1998; White et al., 1999). PrP^C has been established as a copper (Cu²⁺)–binding protein, which is essential for its normal function (Thackray et al., 2002). The octapeptide repeat regions of PrP^C bind to Cu²⁺, Zn²⁺, and Mn²⁺ with varying affinities (Brown et al., 1997b; Choi et al., 2006; Hornshaw et al., 1995a). These findings are further supported by implications of altered metal binding and oxidative stress metabolism in the pathogenesis of prion disease (Wong et al., 2000). Prion protein–deficient mice exhibit decreased SOD activity, GSH reductase activity, and altered metal content compared to naive mice (Brown et al., 1997c; Klamt et al., 2001). Additionally, copper and manganese (Mn) content levels are significantly altered in prion protein-deficient mice, suggesting that the normal activity of PrP^C may regulate cellular metal homeostasis (Brown, 2003; Brown and Harris, 2003; Watt and Hooper, 2003). Recent studies have shown that overexpression of PrP^C results in a phenotype that is more resistant to oxidative stress, and the lack of expression results in a phenotype that is more sensitive to oxidative stress (Brown et al., 1997c; Sakudo et al., 2004). Alterations in metal-binding capacity of PrP^C have recently been implicated in propagation of prion diseases (Wong et al., 2001). Prolonged exposure to Mn has been suggested to lead to substitution of Cu in PrP^C, resulting in the loss of antioxidant activity; this may also aid in the conversion of PrP^C to PrP^Scin vitro (Brown et al., 2001).

Oxidative stress and programmed cell death have recently been implicated in a number of neurodegenerative disorders including prion disease (Choi et al., 2006; Kanthasamy et al., 2003; Tritschler et al., 1994). The neurotoxic peptide fragment of human PrP^C (PrP 106–126) has been shown to induce reactive oxygen species (ROS) production, mitochondrial dysfunction, mitogen-activated protein kinase activation, caspase-3 activation, and apoptosis in multiple cell types (Brown, 2000; Brown et al., 1997a; Burkle et al., 1999; Florio et al., 1996). We and others have recently shown that oxidative stress and caspases mediate Mn-induced apoptotic cell death in neuronal cells (Aschner, 1999; Kitazawa et al., 2005; Latchoumycandane et al., 2005; Roth and Garrick, 2003; Roth et al., 2002; Weber et al., 2002; Worley et al., 2002). In the current study, we characterized whether PrP^C alters Mn-induced oxidative stress apoptosis in neuronal cells and report that PrP^C protects against Mn-induced GSH depletion, ROS generation, caspase-3 activation, and DNA fragmentation, possibly by interfering with Mn uptake.

MATERIALS AND METHODS

Chemicals.

Hydrogen peroxide (H₂O₂), 3-(4,5-dimethylthiazol-3-yl)-2,5-diphenyl tetrazolium bromide (MTT), Trolox, and manganese chloride (MnCl₂) were purchased from Sigma (St Louis, MO); ApoAlert Glutathione Detection Kit was purchased from Clontech (Palo Alto, CA); Sytox Green nucleic dye, Dihydroethidium, DIOC₆, and Hoechst 33342 nucleic dye were purchased from Molecular Probes (Eugene, OR). Ac-DEVD-AFC (Acetyl-Asp-Glu-Val-Asp-7-amido-4-trifluoromethylcoumarin [AFC]), Ac-LEHD-AFC (Acetyl-Leu-Glu-His-Asp-AFC), Z-VAD-FMK (Z-Val-Ala-Asp-fluoromethyl ketone), and Z-DEVD-FMK (Z-Asp-Glu-Val-Asp-fluoromethyl ketone) were purchased from MP Biomedicals (Aurora, OH). Cell Death Detection ELISA^plus (enzyme-linked immunosorbent assay) Assay Kit was purchased from Roche Molecular Biochemicals (Indianapolis, IN). Bradford protein assay kit was purchased from Bio-Rad Laboratories (Hercules, CA). Dulbecco's modified Eagle's Medium (DMEM), fetal bovine serum, L-glutamine, penicillin, and streptomycin were purchased from Invitrogen (Gaithersburg, MD). Nitric acid was purchased from Fisher Scientific (Pittsburgh, PA).

Cell culture.

The mouse PrP cDNA and vector alone were engineered into a mouse neural cell line (CF10) derived from PrP knockout mice as described previously (Kocisko et al., 1994; Priola et al., 1994; Takemura et al., 2006). Complementary DNAs (cDNAs) encoding mouse PrP cDNA with a hamster epitope were first cloned into the retroviral vector pSFF in order to establish a stable cell line. The substitution of lysine and valine residues at positions 108 and 111 with methionine residues allows the mouse prion protein to be recognized by the 3F4 monoclonal antibody. Stable cell lines that constitutively expressed mouse wild-type PrP^C (PrP^C-cells) were selected with G418. CF10 cells stably expressing the pSFF vector alone were also established and designated as PrP^KO-cells. These cell lines were kindly provided by Dr Susan Priola, Rocky Mountain Lab, NIAIDS, MT. PrP^C- and PrP^KO-cells were grown in DMEM with 5% fetal bovine serum supplemented with 2mM L-glutamine, 50 units of penicillin, and 50 μg/ml streptomycin in a humidified atmosphere of 5% CO₂ at 37°C.

Immunocytochemistry.

Cells were plated onto coverslips coated with 0.1 mg/ml poly-L-lysine. After 24–48 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 in the prion protein (1:500; Signet Labs, Berkeley, CA) overnight at 4°C, followed by incubation with the appropriate Alexa 488 secondary antibody (1:1000; Molecular Probes, Eugene, OR) for 90 min at room temperature. The cells were then mounted on slides, viewed, and imaged under a Nikon 2000U microscope. For confocal microscopy, a Nikon C1 microscopy system was employed. All images were processed in MetaMorph 5.7 from Universal Imaging (Downingtown, PA).

Western blotting.

PrP^C- and PrP^KO-cells were lysed, homogenized, sonicated, and centrifuged as described previously (Kaul et al., 2003; Kitazawa et al., 2002). The supernatants were collected as cell lysates, and protein concentrations were determined and used for sodium dodecyl sulfate (SDS) gel electrophoresis. Cytoplasmic fractions containing equal amounts of protein were loaded in each lane and separated on a 10–12% SDS-polyacrylamide gel electrophoresis as described previously (Kaul et al., 2003; Kitazawa et al., 2002). Proteins were then transferred to nitrocellulose membrane, and nonspecific binding sites were blocked by treating with 5% nonfat dry milk powder. The membranes were then treated with primary antibody directed against the 3F4-epitope present in the prion protein (1:500 dilution), followed by treatment with secondary HRP-conjugated anti-mouse antibody. Secondary antibody-bound proteins were detected using Amersham's ECL (enhanced chemiluminescence) kit (Amersham, Piscataway, NJ). For determination of differences in divalent metal transporter-1 (DMT-1) and transferrin between PrP^C-cells and PrP^KO-cells, blots were probed with anti-mouse DMT-1 (1 μg/ml) and anti-mouse transferrin (1 μg/ml) (Alpha Diagnostic International, San Antonio, TX). Following incubation with primary antibodies, the blots were incubated with secondary HRP-conjugated anti-mouse antibody and processed for ECL detection. To confirm equal protein loading, blots were reprobed with a β-actin antibody (1:5000 dilution). All Western blot images were captured with a Kodak 2000 MM imaging system.

Cytotoxicity assays.

To examine cell death, cells were plated and grown on six-well plates at density of 5 × 10⁵ and treated with MnCl₂ in a range of concentrations from 10μM to 1mM for 24 h. Cell death was determined by the Trypan blue exclusion method with an Improved Nebauer-type Hemacytometer as described previously (Anantharam et al., 2002). The cell viability was normalized as percent of control. Cell death was also measured using the Sytox Green cell death assay as described previously (Latchoumycandane et al., 2005). Briefly, the cells were exposed to 100μM Mn and 1μM Sytox for 24 h. Sytox green is a cell-impermeable nucleic acid dye that enters dead cells and intercalates with DNA to produce green fluorescence. After treatment, the Sytox-positive cells were imaged using a Nikon TE2000 microscope and SPOT digital camera (Diagnostic Instruments, Sterling Heights, MI). Cells undergoing apoptosis exhibit a strong green fluorescence because the dye permeates through the cells and intercalates with DNA in the nuclei.

Determination of intracellular Mn concentration.

Inductively coupled plasma mass spectrometry (ICP-MS) was used to determine intracellular Mn concentration. The ICP-MS device was a high-resolution double focusing instrument operated in medium resolution (m/Δm = 4,000) in order to resolve ⁵⁵Mn⁺ from any potential interferences (Shum et al., 1992). The cells were treated with either 30 or 100μM Mn for 6 h. For measurements with antioxidants, cells were pretreated with 2mM Trolox (vitamin E analog), and then exposed to 100μM Mn for 6 h. After washing twice with phosphate-buffered solution (PBS), the number of cells was counted. Samples containing an equal number of cells were resuspended in 1% nitric acid and spiked with gallium. ⁶⁹Ga⁺ was chosen as an internal standard because its m/z ratio is similar to that of the element of interest, and it has no major spectroscopic interferences. A 10-μl aliquot of a 10 ppm Ga standard was added to each of the samples. A multielement standard was prepared using 10-μl aliquots of 10 ppm standard solutions of Mn, Cu, and Ga in 1 ml of 1% nitric acid. All solutions were prepared with Milli-Q 18.2 Mω water. The nitric acid blank, the three-element standard, and each of the samples were introduced into the ICP-MS via a 100-μl/min self-aspirating PFA nebulizer (Elemental Scientific, Inc., Omaha, NE). Between samples, the nitric acid blank was used to rinse the nebulizer. The results for each sample are calculated using the integrated average background-subtracted peak intensities from 20 consecutive scans. In order to correct for differences in elemental sensitivity in the ICP, the three-element standard was used to derive a normalization factor for each element. Using this normalization factor and the isotopic abundances of each element, the concentrations of Mn and copper were calculated for each sample.

Determination of mitochondrial depolarization by Mn.

Depolarization of mitochondrial membrane potential (ΔΨm) was assessed by flow cytometry using DIOC₆ as previously described (Kitazawa et al., 2001). Prior to analysis, cells were collected and incubated with 40nM of DIOC₆ for 15 min. At the end of the incubation period, cells were washed once, resuspended in PBS, and analyzed using the Becton Dickenson FACScan™ flow cytometer (Becton Dickenson, San Francisco, CA) with excitation at 484 nm and emission at 501 nm.

MTT assay.

This assay is widely used to assess cell viability by measuring the activity of mitochondrial dehydrogenase enzymes that cleave the tetrazolium ring to produce formazan (Brown et al., 1994; Kitazawa et al., 2001). After Mn treatment, cells were washed once and further incubated in serum-free DMEM containing 0.25mg/ml MTT for 1 h at 37°C. Supernatant was removed, and MTT crystals were solubilized in 200 μl of dimethyl sulfoxide. The mitochondrial activity was measured with the SpectraMax spectrophotometer (Molecular Devices Corporation, San Diego, CA) at 570 nm with the reference wavelength at 630 nm.

Determination of ROS production by Mn.

Flow cytometry analysis was performed with a Becton Dickenson FACScan flow cytometer (Becton Dickenson). Dihydroethidium, a sodium borohydride-reduced derivative of ethidium bromide, was used to detect ROS. When hydroethidium is loaded into the cells, it binds to cellular macromolecules. ROS are generated, which reduce hydroethidium to ethidium bromide, thus increasing the fluorescence signal (620 nm). This ROS assay was performed as per the method described previously (Kitazawa et al., 2002). Cells were resuspended in PBS with 2mM calcium at a density of approximately 1 × 10⁶ cells and incubated with 10μM dihydroethidium for 15 min at 37°C in the dark to allow the dye to load into the cells. Following addition of Mn, ROS generation was measured at 60 min. The flow cytometric data were analyzed with Cellquest data analysis software (Becton Dickenson).

Visualization of ROS production induced by Mn.

PrP^C- and PrP^KO-cells were plated at a density of 5 × 10⁵ cells in six-well plates and grown overnight. Cells were then incubated with 100μM Mn and 5μM dihydroethidium at 37°C for 5 min in the dark. After treatment, cells were washed once with PBS, and the fluorescence positive cells were viewed microscopically using 620 emission filter. Pictures were taken at 15-min intervals over 60 min using a Nikon TE2000 microscope and SPOT digital camera (Diagnostic Instruments).

GSH assay.

Methods adapted from ApoAlert Glutathione Detection Kit (Clontech) were used to measure intracellular GSH. After Mn treatment, the cell lysates were prepared according to the manufacturer's protocol. The cell lysates were centrifuged at 132,000 × g for 10 min. The resulting supernatant was collected and incubated with 1mM monochlorobimane (GSH-binding dye) for 20 min at 37°C. Binding of the monochlorobimane dye with GSH was measured using a fluorescence plate reader (excitation 395 nm, emission 480 nm, Molecular Devices Corporation, Gemini XS Plate Reader). Results were calculated by normalizing the fluorescence signals from the samples to the amount of protein. Protein concentration was measured using the Bradford protein assay as described previously.

SOD assay.

Methods were adapted from the Superoxide Dismutase Assay Kit (Cayman Chemical, Ann Arbor, MI). Both PrP^C-cells and PrP^KO-cells were grown to confluency, and collected for assay of both total SOD and MnSOD activity. Cells were homogenized in cell homogenization buffer (20mM 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid [HEPES], 1mM ethyleneglycol-bis(2-aminoethylether)-N,N,N′,N′-tetraacetic acid [EGTA], 210mM mannitol, and 70mM sucrose, pH 7.2). Cytosolic lysates were prepared by centrifugation at 10,000 × g for 15 min at 4°C, and the resulting supernatant was used for measurement of total SOD activity. The pellet was resuspended in the cell homogenization buffer with 3mM KCN to measure MnSOD activity. SOD standard was prepared from SOD standards provided by the kit. All the samples were incubated with diluted radical detector (hypoxanthine) and xanthine oxidase, and placed on the orbital shaker for 20 min at room temperature. After the incubation, absorbance was measured at 450 nm using a plate reader. Activity of SOD in the samples was calculated with the standard curve obtained from the SOD standard, and normalized to protein concentration. One unit of SOD was defined as the amount of enzyme required to exhibit 50% dismutation of the superoxide radical.

Enzymatic assay for caspases.

Caspase-3 and caspase-9 activities were measured as described in our previous publications (Anantharam et al., 2002; Kaul et al., 2003; Kitazawa et al., 2002, 2003). After exposure to Mn, the cells were washed with PBS, resuspended in lysis buffer containing 50mM Tris/HCl (pH 7.4), 1mM ethylenediaminetetraacetic acid, 10mM EGTA, and 10μM digitonin, and incubated at 37°C for 20 min. Lysates were centrifuged at 132,000 × g, and the cell-free supernatants were incubated with 50μM Ac-DEVD-AFC (fluorometric caspase-3 substrate) or 50μM Ac-LEHD-AFC (fluorometric caspase-9 substrate) at 37°C for 1 h. Formation of AFC, resulting from caspase cleavage, was measured using a fluorescence plate reader (excitation 400 nm, emission 505 nm). All the fluorescence signals from the samples were normalized to protein concentration, as determined with the Bradford protein assay.

In situ labeling of apoptosis.

Hoechst 33342 nucleic dye was used to visualize nuclear condensation, which was one of the distinct morphological changes observed (Kitazawa et al., 2003). The cells were grown on coverslips coated with poly-lysine (0.1 mg/ml) for 1 day at 37°C with 5% CO₂. Attached cells were treated with 100μM Mn for 24 h. After exposure, cells were washed once with PBS and treated with 10 μg/ml Hoechst 33342 for 10 min in the dark. Cells were washed once with PBS and fixed in 10% buffered formalin for 30 min at room temperature in the dark. After fixing, the cells were washed three times with PBS for 5 min each to remove the formalin, followed by one wash with H₂O. Then the coverslips were mounted on slides and observed with a Nikon TE2000 microscope and SPOT digital camera (Diagnostic Instruments).

DNA fragmentation assay.

DNA fragmentation assays were performed using a recently developed Cell Death Detection ELISA^plus Assay Kit. This is a fast, highly sensitive, and reliable assay for detection of early changes in cells undergoing apoptotic cell death (Anantharam et al., 2002; Kaul et al., 2003; Kitazawa et al., 2002). DNA fragmentation was measured in PrP^KO- and PrP^C-cells exposed to Mn at time points correlating to the maximum caspase-3 activation. For H₂O₂ study, both PrP^C- and PrP^KO-cells were incubated with 100μM H₂O₂ for 24 h. In inhibitor studies, Z-DEVD-FMK (50μM) or Z-VAD-FMK (100μM) was cotreated with 100μM Mn for 18 h. After treatment, 20 μl of cell lysate was prepared according to the manufacturer's protocol. The assay solution consisted of a mixture of anti-histone biotin and anti-DNA-horseradish peroxidase directed against various histones and antibodies to both single-stranded DNA and double-stranded DNA, which are the major constituents of nucleosomes. After 2 h of incubation, unbound components were removed by washing three times with the incubation buffer provided with the kit. The nucleosomes retained by anti-DNA-horseradish peroxidase in the immunocomplex were quantified spectrophotometrically with ABTS (2,2′-azino-di(3-ethoxybenzylthiazoline sulphonate) as horseradish peroxidase substrate (supplied by the kit). Measurements were made at 405 nm against an ABTS solution as a blank (reference wavelength 490 nm). All sample concentrations were normalized to protein concentration using the Bradford protein assay.

Statistical analysis.

Data were analyzed with Prism 3.0 software (GraphPad Software, San Diego, CA). Bonferroni's and Dunnett's multiple comparison testing were used to delineate significant differences between PrP^C-cells, Mn-treated PrP^C-cells, PrP^KO-cells, and Mn-treated PrP^KO-cells. For the cell viability assay, a nonlinear regression curve was fit onto the data, and EC₅₀ concentrations were extrapolated. For comparison between two samples, Student's t-test was performed to examine the differences. Differences with p < 0.05, p < 0.01, and p < 0.001 were considered to be significant and are indicated by asterisks. Nonsignificant differences were noted as ns. The sample size was calculated based on a statistical power analysis of our preliminary data. It was determined that n = 6 was large enough (95% confidence intervals) to distinguish significant differences among the treatment groups as well as the cell types.

RESULTS

Generation of Stable Prion Protein–Expressing Cell Lines

To understand the physiological role of prion protein, we used the mouse PrP^C and PrP^KO neural cells. Constitutive and stable expression of prion protein was verified using immunoblot and immunohistochemical analyses. Figure 1A shows stable expression of a di-glycosylated (32 kDa), mono-glyocosylated, and unglycosylated prion protein band in PrP^C-cell lysates, but not in PrP^KO-cell lysates, as determined by Western blot analysis using a monoclonal antibody directed against the 3F4 epitope in mouse prion protein. Stable expression of mouse prion protein was further confirmed by immunocytochemical analysis (Fig. 1B). The bright green immunofluorescence-staining pattern in the PrP^C-cells revealed that the prion protein was distributed on the surface of the cells, whereas no immunoreactivity was detected in the PrP^KO-cells. Blue staining represents nuclear staining with Hoechst 33342.

FIG. 1.

Prion protein expression. (A) Western blot was performed to verify prion protein expression in the PrP^C-cells, and the lack of prion protein expression in the PrP^KO-cells. Lane 1 represents PrP^C-cells and lane 2 represents PrP^KO-cells. β-Actin was used to confirm equal loading of proteins. (B) Immunocytochemistry was also performed to verify prion protein expression in the cells. Prion protein was stained with anti-3F4 antibody conjugated with Alexa 488 (green stain). Nuclear staining was performed with Hoechst 33342.

Prion Protein Expression Protects against Manganese-Induced Cytotoxic Cell Death

PrP^KO- and PrP^C-cells were exposed to Mn concentrations ranging from 0.01 to 10mM for 24 h, and the extent of cell death was determined by the Trypan blue exclusion method. Figure 2A shows a dose-dependent increase in cytotoxic cell death with Mn treatment. Data analysis yielded an EC₅₀ for Mn-induced cytotoxic cell death of approximately 59.9 and 117.6μM for PrP^KO- and PrP^C-cells, respectively. These results suggest that the PrP^C-cells were more resistant to cell death induced by Mn.

FIG. 2.

PrP^C-cells are more resistant to Mn toxicity. (A) Both PrP^KO- and PrP^C-cells were exposed to Mn (10–10,000μM), and cytotoxicity was measured using the Trypan blue exclusion assay. The EC₅₀ values were 59.9 and 117.6μM, respectively, for each cell line. Each of the points represents mean ± SEM Mn-induced cytotoxicity in PrP^KO-cells and PrP^C-cells. (B) Phase-contrast and Sytox® fluorescence staining images of PrP^C- and PrP^KO-cells with Mn treatment. Cells were exposed to 100μM Mn for 24 h and then stained with 1μM Sytox® green nucleic dye, and images were captured using a Nikon TE2000 microscope and SPOT digital camera (Diagnostic Instruments).

In addition, cytotoxicity was also monitored qualitatively with the Sytox fluorescence assay. Microscopic analysis clearly displayed the protective effect of prion protein, as evident by the reduced number of Sytox-positive green cells in Mn-treated PrP^C-cells when compared to Mn-treated PrP^KO-cells (Fig. 2B). Changes in cellular morphology due to the Mn treatment can be seen in the phase-contrast images.

Intracellular Manganese is Significantly Attenuated in Prion Protein–Expressing Cells

To verify whether cellular prion protein alters the intracellular Mn accumulation, the Mn concentration was measured by ICP-MS in PrP^C- and PrP^KO-cells. Table 1 shows the intracellular Mn levels in Mn-treated PrP^KO- and PrP^C-cells. ICP-MS revealed a significant difference in intracellular Mn levels in untreated PrP^C- and PrP^KO-cells. The basal intracellular Mn levels were determined to be 2.38 ± 0.6 ng/mg protein and 0.245 ± 0.034 ng/mg protein, respectively. PrP^C-cells had a ninefold higher basal level of Mn compared to PrP^KO-cells, suggesting that the PrP^C-cells were able to retain Mn more effectively than PrP^KO-cells. Exposure to 30 and 100μM Mn resulted in a dose-dependent increase in intracellular Mn levels in both PrP^C- and PrP^KO-cells (Table 1). Intracellular Mn accumulation was increased by 840- and 2200-fold in 30 and 100μM Mn-treated PrP^KO-cells, respectively, whereas intracellular Mn levels increased by only 48- and 180-fold, respectively, in PrP^C-cells. These results suggest significantly reduced Mn uptake in PrP^C-cells. Since PrP^C has been implicated in copper binding, copper levels for both PrP^C-cells and PrP^KO-cells were examined. Similar to Mn values, PrP^C-cells exhibited higher basal copper levels as compared to PrP^KO-cells (Table 1). Following Mn treatment, no significant alterations in copper levels were noted in PrP^C-cells, but a slight decrease was observed in PrP^KO-cells. This suggests that PrP^C may be involved in metal homeostasis, and the lack of the protein may result in lower than normal basal metal levels. Conversely, exposure to high concentrations of metals may result in reduced capability to prevent excessive metals from accumulating inside the cells. We also examined whether Mn exposure alters the levels of DMT-1 and transferrin in PrP^C-cells and PrP^KO-cells. As shown in Figure 3, Western blot analysis showed no significant differences in DMT-1 and transferrin between PrP^C-cells and PrP^KO-cells following 100μM Mn treatment, suggesting that the levels of these metal-binding proteins do not account for the difference in Mn uptake in PrP^C-cells and PrP^KO-cells. Also, basal levels of DMT-1 and transferrin were not altered in PrP^C-cells and PrP^KO-cells. Mouse liver lysates and/or mouse brain lysates were used for a positive control.

FIG. 3.

Manganese treatment does not alter DMT-1 or transferrin levels in PrP^C-cells and PrP^KO-cells. (A) Both PrP^C-cells and PrP^KO-cells were treated with 100μM Mn for 24 h and collected for analysis by Western blot for DMT-1 levels. The positive control was mouse liver lysate. β-Actin was used to confirm equal loading of proteins. (B) Western blot analysis of transferrin levels. Positive controls were mouse brain lysates and mouse liver lysates. β-Actin was used to confirm equal loading of proteins.

TABLE 1

Intracellular Mn and Copper Content in PrP^C and PrP^KO Cells

Treatment	PrPC-cells	Fold difference compared to control	PrPKO-cells	Fold difference compared to control
Mn content
Untreated	2.4 ± 0.6	1	0.25 ± 0.03***	1
30μM MnCl2	112.9 ± 4.8	47.5	205.7 ± 6.4***	839.4
100μM MnCl2	428.9 ± 13.2	180.2	537.2 ± 16.7***	2192.5
Copper content
Untreated	23.9 ± 5.2	1	7.3 ± 3.0***	1
30μM MnCl2	18.8 ± 8.0	0.8	4.3 ± 1.2***	0.6
100μM MnCl2	27.0 ± 5.5	1.1	4.5 ± 1.1***	0.6

Note. Values represent mean ng/mg protein ± SEM performed in n = 4 (***p < 0.001 for comparison between groups).

Prion Protein Attenuates Manganese-Induced Mitochondrial Dysfunction

One of the main cellular targets of Mn has been reported to be mitochondria, and as Mn accumulates in the mitochondria it inhibits the mitochondrial complex I activity (Gavin et al., 1992, 1999; Gunter et al., 2004, 2006). Accumulation of Mn in the mitochondria leads to mitochondrial impairment and increased ROS generation (Gunter et al., 2004). In the present study, we examined whether Mn treatment alters mitochondrial dysfunction in both PrP^C- and PrP^KO-cells. Cells were plated out onto 96-well plates and at 0-, 1-, 3-, 6-, 9-, 18-, and 24-h post–Mn-treatment; mitochondrial activity was measured by MTT assay at each time point. As shown in Figure 4A, PrP^KO-cells exhibited greater loss of mitochondrial activity as compared to PrP^C-cells following Mn treatment. The time course data of PrP^KO-cells showed a leftward shift, indicating increased sensitivity of PrP^KO-cells to Mn. Next, we measured the mitochondrial membrane protential (ΔΨm) in Mn-treated PrP^C- and PrP^KO-cells by flow cytometry with a DIOC₆ fluorescent probe to further determine the role of prion protein in Mn-induced mitochondrial dysfunction. After incubation with DIOC₆, cells were treated with Mn and at 6 h, ΔΨm was measured. PrP^C-cells did not exhibit any loss of ΔΨm at 6 h, whereas PrP^KO-cells showed a significant loss of ΔΨm (Fig. 4B).

FIG. 4.

Manganese-induced mitochondrial dysfunction in PrP^KO- and PrP^C-cells. (A) Time course of mitochondrial activity. Cells were treated with 100μM Mn and mitochondrial activity was measured for 0, 1, 3, 6, 9, 18, and 24 h using the MTT assay. (B) Mitochondrial membrane potential (ΔΨm) was measured using DIOC₆. Cells were collected and incubated with 40nM DIOC₆ for 15 min, and incubated with 100μM Mn for 6 h prior to analysis with flow cytometry. Data represent the mean ± SEM for two separate experiments performed in triplicate. **p < 0.01, ***p < 0.001, comparing PrP^C- and PrP^KO-cells.

Manganese Treatment Induces ROS Generation

Recent studies from our lab and others have shown that oxidative stress mediates Mn-induced apoptotic cell death in dopaminergic cells (Gunter et al., 2006; Kitazawa et al., 2005; Latchoumycandane et al., 2005; Stokes et al., 2000; Weber et al., 2002). Therefore, we attempted to determine whether Mn exposure induces ROS generation in PrP^C- and PrP^KO-cells. Flow cytometric analysis using the ROS-sensitive fluorescence probe hydroethidine revealed that treatment with 100μM Mn induces ROS generation. Figure 5A depicts a representative flow cytometric histogram of ROS generation in untreated and Mn-treated PrP^C- and PrP^KO-cells. Exposure to 100μM Mn for 60 min resulted in a 128% and 194% increase in ROS generation in PrP^C- and PrP^KO-cells compared to untreated control cells, respectively (Fig. 5B).

FIG. 5.

Manganese-induced ROS generation in PrP^KO- and PrP^C-cells. (A) Representative flow cytometric histogram of dihydroethidium fluorescence in PrP^C-cells and PrP^KO-cells treated with 100μM Mn. Dihydroethidium was added to the cells and incubated for 15 min at 37°C in the dark. Then Mn was added, and fluorescence intensity was measured at 60 min by flow cytometry. The shift of the curve to the right indicates increase in ROS generation with Mn treatment as indicated by arrows. The x-axis shows the log scale of fluorescence intensity and the y-axis represents the cell count. For both histograms, the solid curve represents the untreated cells, while the outline represents the treated cells. (B) Quantification of ROS generation with Mn treatment. Data represent the mean ± SEM for two separate experiments done in triplicate. ***p < 0.001, comparing treated cells to untreated cells, and treated PrP^C-cells to treated PrP^KO-cells.

For further confirmation of the flow cytometric data, we measured dihydroethidium fluorescence in Mn-treated cells. Increases in dihydroethidium fluorescence, an indicator of ROS levels, were monitored in PrP^KO- and PrP^C-cells every 15 min for up to an hour in 100μM Mn-treated cells. As shown in Figure 6, Mn treatment induced a time-dependent increase in the number of cells positive for dihydroethidium fluorescence in both PrP^C- and PrP^KO-cells. However, the number of dihydroethidium-positive cells was significantly less in Mn-treated PrP^C-cells as compared to PrP^KO-cells. Together, these data suggest that normal prion protein–expressing cells are more resistant to Mn-induced oxidative insult.

FIG. 6.

Visualization of ROS generation in Mn-treated PrP^KO- and PrP^C-cells over time. Following a 5-min incubation at 37°C with 1μM dihydroethidium, the cells were treated with 100μM Mn, and the fluorescence images were captured. The fluorescence of the treated cells increased over time. Mn treatment increased the number of fluorescent cells over time, whereas images of untreated cells showed relatively low fluorescence at the 60-min time point.

To verify that the increased uptake of Mn by PrP^KO-cells was not due to increased ROS generation, both PrP^C-cells and PrP^KO-cells were pretreated and cotreated with 2mM Trolox, which is a potent antioxidant that protects against ROS. As seen in Table 2, cotreatment with Trolox did not prevent Mn internalization into the cells, and no significant difference was observed between Mn treatment and Mn + Trolox cotreatment in both PrP^C-and PrP^KO-cells.

TABLE 2

Effect of Antioxidant Trolox on Manganese Uptake

Treatment	PrPC-cells	PrPKO-cells
Untreated	3.4 ± 0.2	0.6 ± 0.01
100μM MnCl2	447.2 ± 6.8	451.8 ± 6.2
100μM MnCl2 + 2mM Trolox	457.4 ± 7.6	446.6 ± 6.6

Note. Values represent mean ng/mg protein ± SEM performed in n = 4.

Prion Protein Expression Protects against Manganese-Induced Decreases in GSH Levels

To determine whether increases in ROS production are also accompanied by decreases in antioxidant GSH levels, we measured initial GSH levels in untreated PrP^KO- and PrP^C-cells. Figure 7A shows that there is no significant difference in baseline GSH levels between these two cell lines, suggesting that prion protein expression does not affect the basal intracellular GSH levels. We also investigated whether GSH metabolism is altered in prion protein–expressing cells. For this study, we exposed both PrP^C-cells and PrP^KO-cells to bulthionine sulfoximide (BSO), an inhibitor of GSH synthesis, and determined the intracellular GSH levels. Figure 7B shows that exposure to 100μM BSO for 6 h resulted in decreased intracellular GSH levels in both these cell types. There was no significant difference between PrP^C and PrP^KO-cells. We then investigated whether Mn treatment alters GSH levels in PrP^KO- and PrP^C-cells. Exposure to 100μM Mn for 24 h decreased intracellular GSH levels in both PrP^C- and PrP^KO-cells (Fig. 7C). However, the decrease in GSH levels was more pronounced in PrP^KO-cells than in PrP^C-cells. We also measured total cellular SOD as well as mitochondrial SOD (MnSOD) in PrP^C-cells and PrP^KO-cells to determine any difference in the basal activity of this important antioxidant enzyme. Our results show that neither total SOD nor MnSOD activity was altered in PrP^C-cells and PrP^KO-cells (Fig. 7D). Taken together with cellular GSH (Fig. 7A), these results indicate no significant difference in antioxidant status between PrP^C- and PrP^KO-cells.

FIG. 7.

Manganese-induced depletion of GSH levels in PrP^KO- and PrP^C-cells. (A) Baseline GSH levels of PrP^C-cells and PrP^KO-cells. (B) Effect of BSO on GSH levels in PrP^C-cells and PrP^KO-cells. The cells were treated with 100μM BSO for 6 h and then GSH levels were measured. (C) Effect of Mn on PrP^C-cells and PrP^KO-cells. The cells were treated with 100μM Mn for 24 h and then GSH levels were measured and compared. (D) Baseline levels of total SOD and MnSOD activity were compared between PrP^C-cells and PrP^KO-cells. Data represent the mean ± SEM for two experiments performed in triplicate. *p < 0.05, comparing Mn-treated PrP^C and PrP^KO-cells.

Manganese-Induced Caspase-9 and -3 Activation is Significantly Suppressed in PrP^C-Cells

ROS production in cells is known to activate many proapoptotic factors, including cytochrome c, which subsequently triggers activation of the caspase cascade (Cassarino et al., 1999; Tan et al., 1998). Sequential activation of the apoptotic factor caspase-9 followed by caspase-3 is important in mediating oxidative stress-induced cellular death.

We examined whether Mn-induced differences in cytotoxic cell death between PrP^C- and PrP^KO-cells are due to differences in the activation of the caspase cascade in these cell lines. Exposure to Mn resulted in time- and dose-dependent increases in caspase-9 enzyme activity. Figure 8 shows the activation of caspase-9 in PrP^C- and PrP^KO-cells treated with 30μM Mn (Fig. 8A) and 100μM Mn (Fig. 8C). The activation of caspase-9 started at 18 h, peaked at 24 h, and then started to decrease at 30 h in cells treated with 30μM Mn. Increasing the concentration of Mn to 100μM resulted in the earlier activation of caspase-9 activation, starting at 6 h, peaking at 18 h, and then decreasing at 24 h. Exposure to 30μM Mn resulted in an increase in caspase-9 enzyme activity by 610% in PrP^KO-cells and by 470% in PrP^C-cells at 24 h, respectively. Similarly, exposure to 100μM Mn resulted in an increase in caspase-9 enzyme activity by 690% by 18 h in PrP^KO-cells and by 549% in PrP^C-cells. Data analysis revealed that the Mn-induced increase in caspase-9 activation was significantly reduced in PrP^C-cells, suggesting that prion protein may protect cells against Mn-induced activation of early initiator apoptotic protease caspase-9.

FIG. 8.

Manganese-induced caspase-9 and -3 activation in PrP^KO- and PrP^C-cells. Cells were treated with 30μM or 100μM Mn for 24 or 30 h, and caspase-9 and -3 activities were measured with a fluorescence plate reader using specific caspase substrates (50μM Ac-LEDHD-AFC for caspase-9 activity and 50μM Ac-DEVD-AFC for caspase-3 activity). (A, B) A 30μM Mn treatment induced caspase-9 activation starting at 18 h (*p < 0.05); caspase-3 activation started at about the same time as caspase-9, and peaked at 24 h, with PrP^KO-cells having a higher activity at 24 h (p < 0.01). (C, D) A 100μM Mn treatment induced caspase-9 activation around 6 h, peaking around 18 h, with slight differences between the two cell lines (p < 0.05). Caspase-3 activity began around 6 h, and peaked around 18 h without any difference between the two cell lines. The value of each treatment group represents the mean ± SEM from two separate experiments performed in triplicate (*p < 0.05; **p < 0.01).

Manganese exposure also induced a time- and dose-dependent increase of effector caspase (caspase-3) activation in PrP^C- and PrP^KO-cells. Figure 8 also shows the activation of caspase-3 in PrP^C- and PrP^KO-cells treated with 30μM Mn (Fig. 8B) and 100μM Mn (Fig. 8D). The activation of caspase-3 started to appear after 12 h, peaked at 24 h, and then started to decrease at 30 h in 30μM Mn-treated cells. Increasing the Mn concentration to 100μM resulted in earlier activation of caspase-3, starting at 6 h, peaking at 18 h, and then decreasing at 24 h. Exposure to 30μM Mn resulted in an increase in caspase-3 enzyme activity by 760% at 24 h in PrP^KO-cells and by 400% in PrP^C-cells, respectively. Similarly, exposure to 100μM Mn increased caspase-3 enzyme activity by 646% at 18 min in PrP^KO-cells and by 579% in PrP^C-cells, respectively. Comparative analysis of caspase-3 data revealed that the Mn-induced increases in caspase-3 activation were significantly different between PrP^C-cells and PrP^KO-cells following 30μM Mn treatment. However, the extent of caspase-3 activation was similar, and not significantly different, between these two cell lines in 100μM Mn-treated cells. Together, these data suggest that cellular prion protein may protect against Mn-induced increase in caspase-3 activation.

In Situ Fluorometric Detection of Manganese-Induced Apoptosis

To understand the functional consequence of the activation of caspases, we examined whether Mn induces apoptosis. Chromosomal breakdown and DNA condensation are hallmarks of cells undergoing apoptosis. To identify whether Mn induces apoptotic cell death, we used in situ fluorometric analysis to detect chromatin condensation using Hoechst 33342 dye. After treatment with Mn, PrP^KO- and PrP^C-cells were fixed and stained with Hoechst 33342 dye. Exposure to 100μM Mn for 24 h induced chromatin condensation in both PrP^C- and PrP^KO-cells, as observed by the tightly bound dye to the highly condensed DNA (Fig. 9). Image analysis revealed that a significant number of Mn-treated PrP^KO-cells showed nuclear condensation, whereas only a small percentage of cells exhibited chromatin condensation in Mn-treated PrP^C-cells. The Hoechst 33342 staining of untreated control cells was weak and diffuse, indicating no significant nuclear condensation in these cells (Fig. 9B, far left). These results suggest that prion protein–expressing cells are less susceptible to Mn-induced apoptotic cell death.

FIG. 9.

Phase-contrast and Hoechst 33342 staining of Mn-treated PrP^KO- and PrP^C-cells. After Mn treatment, cells were incubated with Hoechst 33342 (10 μg/ml) for 30 min. Following incubation with Hoechst 33342, cells were visualized and images were captured using a Nikon TE2000 microscope and SPOT digital camera. (A) Phase-contrast images of untreated and treated cells for both PrP^C-cells and PrP^KO-cells. (B) Hoechst 33342 stained images showing chromatin condensation in cells, as indicated by arrows.

Prion Protein Expression Protects against Manganese-Induced DNA Fragmentation

To further confirm that Mn induces apoptotic cell death, we used DNA ELISA sandwich assay to quantify the extent of DNA fragmentation in PrP^KO- and PrP^C-cells. Mn treatment induced a dose-dependent increase in DNA fragmentation in both PrP^KO- and PrP^C-cells (Fig. 10). Exposure to 30μM Mn for 24 h resulted in 137% and 291% increases in DNA fragmentation in PrP^C- and PrP^KO-cells, respectively (Fig. 10A). Similarly, exposure to 100μM Mn for 18 h resulted in 540% and 809% increases in DNA fragmentation in PrP^C- and PrP^KO-cells, respectively (Fig. 10B). Together, these results further confirm that PrP^C-cells are less susceptible to Mn-induced apoptotic cell death.

FIG. 10.

DNA fragmentation in PrP^KO- and PrP^C-cells. PrP^C-cells and PrP^KO-cells were treated with different doses of Mn with or without 100μM Z-VAD-FMK (pan caspase inhibitor). (A) A 30μM Mn exposure increased DNA fragmentation, with PrP^KO-cells showing a higher percentage of DNA fragmentation at 24 h. (B) A 100μM Mn exposure increased DNA fragmentation to a greater extent; again, DNA fragmentation increased more significantly in PrP^KO-cells than in PrP^C-cells. (C) Inhibitor studies with 100μM Mn cotreated with 100μM Z-VAD-FMK significantly inhibited Mn-induced DNA fragmentation. (D) Treatment with 100μM H₂O₂ for 24 h resulted in significantly higher amounts of DNA fragmentation in PrP^KO-cells as compared to PrP^C-cells. Data represent mean ± SEM from two separate experiments done in triplicate. **p < 0.01, ***p < 0.001 comparing between cell lines and between treatments.

Next, we determined whether caspases mediate Mn-induced apoptotic cell death. Figure 10C shows that cotreatment with Z-VAD-FMK (a pan caspase inhibitor) significantly attenuated Mn-induced apoptotic cell death in PrP^KO- and PrP^C-cells. Cotreatment with 100μM Z-VAD-FMK blocked Mn-induced DNA fragmentation by 267% and 226% in PrP^C- and PrP^KO-cells, compared to Mn treatment alone. Treatment with Z-VAD-FMK alone did not have any effect on DNA fragmentation, i.e., the cells appeared similar to untreated cells. Since Z-VAD-FMK did not completely block DNA fragmentation in both cells, caspase-independent apoptotic cell death may also contribute to Mn-induced apoptotic cell death.

To determine whether cellular prion protein protects against apoptotic cell death during oxidative insult, we measured DNA fragmentation in H₂O₂ treated PrP^C- and PrP^KO-cells. DNA fragmentation was only slightly increased in PrP^C-cells following treatment with 100μM H₂O₂ for 24 h, whereas a twofold increase in DNA fragmentation was observed in PrP^KO-cells (Fig. 10D), indicating cellular prion can attenuate oxidative stress-induced apoptotic cell death.

DISCUSSION

In the present study, we show that expression of prion protein rescues PrP^KO-cells from Mn-induced apoptotic cell death by attenuating ROS production, caspase-3 activation, and DNA fragmentation. Our studies also demonstrate that expression of prion protein reduces intracellular accumulation of Mn during the metal exposure. To our knowledge, this is the first study demonstrating the protective role of prion protein in Mn-induced neurotoxicity. Furthermore, the excellent response of PrP^C-cells and PrP^KO-cells to the toxic effect of Mn is similar to the response observed in other neuronal cell models used in previous studies, indicating that PrP^C-cells and PrP^KO-cells are very useful in vitro models for studying the role of prion protein in metal neurotoxicity.

Prion protein is a cell surface glycosylated protein, predominantly expressed in the CNS of most mammalian species (Martins and Brentani, 2002; McKinley et al., 1987; Prusiner, 1984). It is a 254 amino acid peptide with several octapeptide repeat sequences (PHGGSWGQ) toward the N-terminus, which have variable binding affinity for divalent metals such as copper, zinc, and Mn, with preferential binding for copper (Hornshaw et al., 1995b; Viles et al., 1999). The number of octapeptide repeats differs from species to species, ranging from four to eight repeats (Goldfarb et al., 1991; Hornshaw et al., 1995b). Although the role of the octapeptide repeats is yet to be characterized, binding of the divalent cation to the octapeptide repeats may stabilize protein conformation by facilitating the folding of the largely unstructured N-terminus of the protein (Cereghetti et al., 2003). Previous studies have shown the prion protein to be a Cu-binding protein, and binding of Cu seems to be essential for its normal function (Garnett and Viles, 2003; Hornshaw et al., 1995b). Copper binding also enhances the antioxidant activity of the prion protein (Brown et al., 2001). Loss of Cu binding and decreased concentrations of Cu and other metals in the brain have been observed in prion diseases including CJD and BSE (Lehmann, 2002; Rossi et al., 2004; Thackray et al., 2002). A growing body of evidence indicates that Mn can also interact with the prion protein (Brown et al., 1997c; Giese et al., 2004; Tsenkova et al., 2004) to possibly alter its function. Recent studies have shown that PrP^C can be stimulated from the cells to be shed into the media (Parkin et al., 2004). This was first described by Parkin et al., who demonstrated that PrP^C can be cleaved from the surface of the cells through proteolytic cleavage at the site of the GPI-anchor. Similarly, Toni et al. (2005) also described shedding of PrP^C from cells induced through treatment with copper. Both these papers speculated that this mechanism could protect cells against scrapie infection, suggesting that this could be a normal pathway for PrP^C. We hypothesize that the shedding mechanism could be a way for cells to excrete excessive amounts of metals (i.e., copper, Mn, zinc) through a protein–metal complex. This could be a reason why the PrP^KO-cells treated with Mn exhibited higher Mn content, and increased cellular death. Our future studies will focus on this issue of metal-induced shedding of prion proteins to determine possible mechanisms underlying the protective effect of cellular PrP^C against metal neurotoxicity.

Manganese is an essential element in development and survival, but also has been classified as an environmental neurotoxin (HaMai and Bondy, 2004; Mergler et al., 1999; Pamphlett et al., 2001). Excess Mn accumulates in the mitochondria and inhibits both mitochondrial complexes I and II, which leads to disruption of the mitochondrial electron transport system (Galvani et al., 1995; Gunter et al., 2006). Disruption of mitochondrial respiration results in increased levels of oxygen in the cytosol, which can generate ROS. Oxidative stress resulting from Mn has been shown to induce apoptotic cell death in many different cell types (Castilho et al., 1999; Hirata, 2002; Kitazawa et al., 2005; Latchoumycandane et al., 2005; Stredrick et al., 2004). In the present study, Mn-induced cytotoxicity was significantly attenuated in a dose-dependent manner in cells that overexpress prion protein compared to prion protein knockouts. The EC₅₀ were determined to be 117.6μM and 59.9μM for PrP^C- and PrP^KO-cells, respectively. Our results agree with previous studies which show prion protein can protect against cytotoxic cell death in multiple cell types in response to various stimuli (Roucou et al., 2003), and loss of prion protein results in enhanced sensitivity to stress (Brown et al., 1997c, 1998; White et al., 1999). However, the molecular and cellular mechanisms underlying the neuroprotective role of prion protein are poorly understood. In this study we also show that PrP^C-expressing cells have higher levels of basal intracellular accumulated Mn compared to PrP^KO-cells. The elevated basal Mn in PrP^C-cells may be due to internalization of Mn-bound prion proteins. The intracellular Mn accumulation is dramatically suppressed in Mn-treated PrP^C-cells compared to PrP^KO-cells, and may explain the partial resistance of PrP^C-cells to Mn-induced cytotoxicity. Furthermore, our data showing lower copper and Mn levels in cells lacking PrP^C indicate that PrP^C can function to regulate cellular metal content.

We and others recently showed that Mn induces ROS production, mitochondrial dysfunction, cytochrome C release, activation of multiple caspases and apoptotic cell death in neuronal and nonneuronal cell types (Hirata, 2002; Kitazawa et al., 2005; Latchoumycandane et al., 2005; Malthankar et al., 2004; Roth et al., 2002). In this study we examined ROS production in PrP^C- and PrP^KO-cells following Mn treatment. Mn-induced ROS generation was significantly suppressed in a dose- and time-dependent manner in PrP^C-cells compared to PrP^KO-cells. The antioxidant activity of prion protein may in part contribute to the reduced ROS generation observed in PrP^C-cells. Recently, Rachidi et al. (2003) showed that expression of prion protein increases copper binding and antioxidant enzyme activities in the rat kidney RK13 cell line. In the present study, we found that Mn exposure reduced GSH levels to a greater extent in PrP^KO-cells than in PrP^C-cells. However, we found no significant differences in the baseline levels of GSH antioxidant levels between the PrP^C- and PrP^KO-cells. Furthermore, the ability of BSO (inhibitor of γ-glutamylcysteine synthase) to suppress baseline GSH production was identical in both PrP^C- and PrP^KO-cells, suggesting that the basal antioxidant system is not impaired or altered in PrP^C-cells. Our results are also in agreement with those of Senator et al. (2004) who showed that paraquat-induced decreases in GSH levels were significantly lower in prion protein overexpressing A74 kidney cells.

Recent studies from our lab and others have shown that Mn-induced mitochondrial dysfunction and oxidative stress cause a sequential activation of caspase-9 and caspase-3, and induce DNA fragmentation (Anantharam et al., 2002; Kaul et al., 2003; Kitazawa et al., 2002, 2003). In this study, PrP^C overexpression significantly reduced Mn-induced caspase-9 and caspase-3 activation compared to PrP^KO-cells, suggesting that prion protein expression suppresses activation of caspases via ROS or mitochondrial mechanisms. However, our results are not in agreement with those of Paitel et al. (2004) who showed that prion protein overexpression potentiated stuarosporine-stimulated increases in caspase-3 activation in HEK293 cells. This may be due to the type of stimulus and/or type of cell model used in their study.

The biochemical consequences of caspase-3 activation culminate in apoptotic cell death. DNA fragmentation and chromatin condensation are considered biochemical hallmarks of apoptosis and are terminal events in the apoptotic cell death process. Recent studies demonstrate that apoptosis plays a role in the pathology of prion diseases (Daniels et al., 2001; Massimino et al., 2002; Schatzl et al., 1997; Zanata et al., 2002). In particular, the small fragment of prion protein encompassing 106–126 residues has been shown to be neurotoxic to neurons through the apoptotic pathway (Brown, 1998; Brown et al., 1994; Chiarini et al., 2002; Thellung et al., 2000; Zanata et al., 2002). In the present study, in situ fluorometric experiments using Hoechst 33324 revealed that Mn exposure induces chromatin condensation in both PrP^KO- and PrP^C-cells; however, PrP^C-cells exhibited fewer apoptotic nuclei. Quantification of DNA fragmentation by the ELISA method further confirmed the effect of Mn on neuronal apoptosis. Manganese-induced DNA fragmentation was significantly suppressed in prion protein–expressing cells, suggesting that prion protein expression can rescue cells from Mn-induced apoptotic cell death. Pretreatment with the pan caspase inhibitor Z-VAD-FMK significantly blocked Mn-induced DNA fragmentation in both PrP^C- and PrP^KO-cells, suggesting that caspases mediate the apoptotic cell death process. Since the pan caspase inhibitor did not completely block cell death, caspase-independent cell death events may also contribute to Mn-induced apoptosis. Our studies with H₂O₂ suggest that cellular prion proteins not only protect against metal-induced apoptotic cell death but also protect against oxidative insult.

In conclusion, we demonstrate for the first time that prion protein expression can rescue neural cells from Mn-induced apoptotic cell death, possibly by affecting intracellular Mn accumulation, in addition to being neuroprotective against oxidative damage. Finally, this study emphasizes the importance of characterizing the effect of metal interactions with cellular prion protein and elucidating the possible role of metals in the pathogenesis of prion diseases.

The project described was supported by Grant Number W81XWH-05-1-0239 from National Institutes of Health (NIH), United States Department of Defense, United States Army Medical Research, and Materiel Command through Vanderbilt University and its contents are solely the responsibility of the authors and do not necessarily represent the official views of the NIH, United States Department of Defense, United States Army Medical Research and Materiel Command, and Vanderbilt University. This work was also in part supported by NIH grant ES38644. Mouse prion neural cell line was kindly provided by Sue Priola, Rocky Mountain Lab; SPEX-CertiPrep is acknowledged for donating the ICP standards used in this work. Use of the ICP-MS instrument was provided by the U.S. Department of Energy (USDOE), Office of Defense Nuclear Nonproliferation, and the Office of Basic Energy Sciences. Ames Laboratory is operated for the USDOE by Iowa State University under Contract No. W-7405-ENG-82. W. Eugene and Linda Lloyd Endowed Professorship to A.G.K. is also acknowledged. The authors acknowledge Ms Keri Henderson for her assistance in the preparation of this manuscript.