|Home | About | Journals | Submit | Contact Us | Français|
Amyloid β (Aβ) is thought to promote neuronal cell loss in Alzheimer’s disease (AD), in part through the generation of reactive oxygen species (ROS) and subsequent activation of mitogen-activated protein kinase (MAPK) pathways. Protein phosphatase 5 (PP5) is a ubiquitously expressed serine/threonine phosphatase which has been implicated in several cell stress response pathways and shown to inactivate MAPK pathways through key dephosphorylation events. Therefore we examined whether PP5 protects dissociated embryonic rat cortical neurons in vitro from cell death evoked by Aβ. As predicted, neurons in which PP5 expression was decreased by siRNA treatment were more susceptible to Aβ toxicity. In contrast, overexpression of PP5, but not the inactive PP5 mutant, H304Q, prevented MAPK phosphorylation and neurotoxicity induced by Aβ. PP5 also prevented cell death caused by direct treatment with H2O2, but did not prevent Aβ-induced production of ROS. Thus, the neuroprotective effect of PP5 requires its phosphatase activity and lies downstream of Aβ-induced generation of ROS. In summary, our data indicate that PP5 plays a pivotal neuroprotective role against cell death induced by Aβ and oxidative stress. Consequently, PP5 might be an effective therapeutic target in AD and other neurodegenerative disorders in which oxidative stress is implicated.
Alzheimer’s disease (AD) is a progressive neurodegenerative disorder characterized by senile plaques, neurofibrillary tangles (NFTs), and selective loss of neurons in the brain (Morishima et al. 2001; Goedert & Spillantini 2006). The accumulation of senile plaques within the brains of AD patients is considered a central hallmark of the disease. Senile plaques are primarily composed of fibrillar deposits of Aβ, a peptide ranging 39–43 amino acids derived from proteolytic processing of the amyloid precursor protein (APP; Selkoe 2001). In cases of familial AD, mutations in APP and the presenilins, which proteolize APP, lead to increased accumulation of fibrillar Aβ (Hardy & Selkoe 2002). Results from transgenic mouse models expressing mutations associated with familial AD have suggested a possible causal relationship between Aβ deposition, neuritic degeneration and neuronal loss (Urbanc et al. 2002; Meyer-Luehmann et al. 2008; Yankner & Lu 2009). These findings highlight the importance of Aβ deposition in the pathology of AD and provide the impetus for investigating the signal transduction mechanisms by which Aβ neurotoxicity can be prevented.
Oxidative stress plays a key role in AD as well as other neurodegenerative disorders (Lin & Beal 2006). Aβ impairs mitochondrial redox activity and increases the levels of reactive oxygen species (ROS), and antioxidants protect against Aβ-induced neurotoxicity (Behl et al. 1994; Kadowaki et al. 2005; Lin & Beal 2006; Sayre et al. 2008). Although little is known about how Aβ peptides induce ROS generation, in vitro studies revealed that Aβ1-42 binds to copper and iron to trigger the production of ROS (Cuajungco et al. 2000; Dai et al. 2007). This finding, together with the high levels of redox active metals found in AD brains, has reinforced the notion that oxidative stress plays a major role in AD (Lovell et al. 1998; Sayre et al. 2008).
In cultured cortical neurons fibrillar Aβ causes apoptotic or necrotic cell death (Yankner et al. 1990; Morishima et al. 2001; Perini et al. 2002; Geci et al. 2007). The type of cell death may depend on the Aβ concentration used for treatment (Geci et al. 2007). Nonetheless, Aβ-induced toxicity appears to involve one or more of the three major mitogen activated protein kinase (MAPK) pathways, c-jun N-terminal kinase (JNK), p38, and extracellular signal-regulated kinase (ERK), which are known to mediate oxidative stress-induced neuronal death (Crossthwaite et al. 2002; Tamagno et al. 2003; Kadowaki et al. 2005; Zhu et al. 2005; Frasca et al. 2008).
Protein phosphatase 5 (PP5) is a ubiquitously expressed serine/threonine protein phosphatase related to PP1, PP2A and PP2B (Hinds & Sanchez 2008). Brain contains high levels of PP5 mRNA, and PP5 protein is widely expressed in central neurons (Becker et al. 1996; Bahl et al. 2001; Rossie et al. 2006). PP5 dephosphorylates tau, the major component of neurofibrillary tangles, at sites that are hyperphosphorylated in AD (Gong et al. 2004) and PP5 activity is decreased in the neocortex of AD patients (Liu et al. 2005a; Liu et al. 2005b). These findings suggest a potential role for PP5 in protecting cells from AD. Several lines of evidence also suggest that PP5 plays a key role inhibiting MAP kinase pathways through dephosphorylation of Raf-1, a MAPK kinase kinase initiating the ERK MAPK pathway (von Kriegsheim et al. 2006), and apoptosis signal-regulating kinase 1 (ASK1), which activates the JNK and p38 MAPK pathways (Ichijo et al. 1997; Morita et al. 2001; Shinoda et al. 2003; Zhou et al. 2004). ASK1 has been shown to mediate Aβ-induced ROS-dependent death in cultured neurons and PC12 cells (Kadowaki et al. 2005). Together, these observations suggest that PP5 may help prevent neuronal cell death evoked by Aβ-induced oxidative toxicity.
In the present study, we examined whether PP5 protects neurons from death induced by fibrillar Aβ or by the direct application of a ROS, hydrogen peroxide (H2O2). We monitored cell death in embryonic rat cortical neurons in the presence or absence of overexpressed PP5 or a catalytically inactive PP5 mutant. PP5 overexpression reduced neuronal death induced by these ROS-generating agents. In addition, neurons expressing reduced levels of PP5 were more susceptible to cell death induced by Aβ and H2O2. These findings suggest that PP5 may play a protective role against neuronal cell death induced by Aβ and oxidative stress and raise the possibility that PP5 may be an important regulator in AD.
Monoclonal antibodies for microtubule associated protein-2B (α-MAP-2B) and PP5 (α-PP5) were obtained from BD Biosciences (San Jose, CA). Polyclonal antibodies against phospho ERK (α-p-ERK), total ERK (α-ERK), phospho JNK (α-p-JNK), total JNK (α-JNK), phospho p38 (α-p-p38), and total p38 (α-p38) were purchased from Cell Signaling (Danvers, MA). Polyclonal antibody against PP2A catalytic subunit (α-PP2Ac) was purchased from Santa Cruz Biotechnologies (Santa Cruz, CA). Monoclonal antibody for glyceraldehyde-3-phosphate dehydrogenase (α-GAPDH) was acquired from Ambion (Austin, TX). The α-glial fibrillar acidic protein (GFAP) polyclonal antibody, α-FLAG monoclonal antibody, and α-Aβ 22–35 polyclonal antibody were purchased from Sigma (Saint Louis, MO). All reagents were purchased from Sigma unless otherwise indicated.
Primary cortical neurons were prepared from embryonic day 17 rat fetuses as described by Brewer and colleagues (Brewer et al. 1993) with minor modifications. All procedures involving animal handling were approved by the Animal Care and Use Committee of Purdue University in compliance with NIH guidelines. Cells dissociated from embryonic rat cerebral cortices were plated at a density of 7.5 × 104 cells/cm2 in 8-well chamber slides coated with 25 µg/mL PDL for immunocytochemistry, or 1.2 × 105 cells/cm2 in PDL-coated 24-well plates for western blot analysis experiments. Neurons were grown in Neurobasal medium (Invitrogen, Carlsbad, CA) supplemented with 2% B-27 (Invitrogen) and 300 µM glutamine (Invitrogen). Cultures were treated with 1 µM cytosine arabinoside for 24 h during 2–3 days in vitro (DIV). Experiments were performed at 6 DIV.
Peptide treatments were performed as described (Morishima et al. 2001). The Aβ25-35 and the control reversed sequence peptide Aβ35-25 (American Peptide Company, Sunnyvale, CA) were dissolved in sterile water at 1 mM. To prepare aggregated Aβ peptides, the stock peptide solution was diluted with an equal volume of phosphate-buffered saline (PBS) and incubated at 37°C for 5 d. The peptides Aβ1-40 and Aβ40-1 (Bachem, Torrance, CA) were prepared in sterile water at 500 µM, then incubated at 37°C for 7 d to allow aggregation (Grace & Busciglio 2003). The extent of Aβ aggregation and soluble oligomers in these preparations was verified with the thioflavin-T (Th-T) assay and western blot analysis (described below and in Fig. S3, respectively). Peptides were added directly to cultures at a final concentration of 30 µM for Aβ25-35 or Aβ35-25, and 25 µM for Aβ1-40 or Aβ40-1, and cells incubated for 48 h. Hydrogen peroxide (Mallinckrodt Baker, Phillipsburg, NJ) was added at a final concentration of 50 µM for 24 h and STS was added at 200 nM for 24 h. All treatments were performed in Neurobasal medium supplemented with B27 without antioxidants.
Aβ fibril formation was measured by the fluorometric Th-T assay as described (LeVine 1999). Fluorescence was read in a Spectramax Gemini microplate reader (Molecular Devices, Sunnyvale, CA) with excitation at 440 nm and emission at 485 nm.
A catalytically inactive mutant of rat PP5, PP5(H304Q), was generated and characterized as described in Figure S1 and Figure S2. Adenoviruses expressing rat FLAG epitope-tagged PP5 [Ad-PP5(WT) or Ad-PP5(H304Q)] were generated in the gene transfer vector core facility at the University of Iowa (Iowa City, IA) as described (Messner et al. 2006). Cells were incubated with adenovirus at 4 DIV for 48 h. After this period, virus was removed and cells were subjected to toxic treatments. Cells infected with virus lacking a PP5 insert or incubated without adenovirus served as controls.
The siRNA against PP5 corresponded to coding region 277–297 (AAGAAGTACATCAAAGGTTAC) of rat PP5 (Messner et al. 2006) and the control siRNA constituted (AACACTATTCGACGAACTGGA). Control and PP5 siRNA were synthesized using an siRNA construction kit (Ambion). At DIV 4, cortical neurons were transfected in a 96-well plate with 9.5 pmol siRNA using the GeneSilencer siRNA transfection reagent (San Diego, CA). Briefly, siRNA and GeneSilencer were each diluted with culture media, then mixed together for 15 min. Cells were incubated with the transfection mixture for 4 h, then one volume of media was added to cells. Two days later, half the volume of culture media was replaced with fresh media. Transfected cells were used for Aβ and H2O2 toxicity experiments or harvested for western blot analysis at DIV 11.
Neurons were fixed with 4% (w/v) formaldehyde in PBS for 12 min at room temperature (RT), followed by methanol (−20°C) for 10 min, and then blocked for 1 h with PBS containing 0.3% (w/v) Tween-20 and 3% (w/v) BSA. Cells were then incubated with primary antibody for 1 h at RT, followed by fluorophore-conjugated secondary antibody (Invitrogen) together with the nuclear stain DAPI (1 µg/mL) for 1 h at RT. All antibodies were prepared in blocking solution supplemented with 1% (w/v) goat serum. Cells were analyzed with a Leica DMRX epifluorescence microscope. Images were acquired with OpenLab 5.0.1 software and formatted using Adobe Photoshop.
Cell viability experiments were performed by measuring the ability of neurons to metabolize MTT into a formazan salt by viable mitochondrial enzymes (Stanciu & DeFranco 2002). In brief, neurons were plated into 96-well plates (43,000 cells/well). At 6 DIV neurons were subjected to the indicated toxic treatments, followed by incubation with 10 µL of 3-(4,5-Dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) solution (5 mg/mL) for 2.5 h at 37°C. Cells were then solubilized (50% dimethylformamide and 20% SDS, pH 4.8) overnight at room temperature and the absorbance read at 570 nm in a Spectramax 340 microplate reader (Molecular Devices). The percentage of cell survival was expressed relative to non-treated control neurons.
This assay was performed as described previously (Levinthal & Defranco 2005) with minor modifications. Cell death was assessed based on the principle that only the nuclei of cells with compromised plasma membranes will be stained by propidium iodide (PI). Following various toxic treatments, neurons were incubated with 10 µg/mL PI for 10 min at 37°C, washed with PBS, and immunocytochemistry was performed using α-FLAG antibody to identify cells overexpressing PP5. The nuclear stain DAPI was used to label all cells. The percent cell death was calculated as the number of PI-positive cells among the infected population (PP5-overexpressing cells). Approximately 200 cells were counted among 20 – 25 randomly chosen fields per experiment.
Detection of ROS generation in neurons was performed using hydroethidine (HEt) as described by Martin and colleagues (Martin et al. 2001) with minor modifications. Neurons were treated with Aβ peptides for 8 h, then incubated with 3.0 µg/mL HEt (Invitrogen) for 30 min at 37°C, and processed for immunocytochemistry analysis using α-FLAG antibody to label cells overexpressing PP5 as described above. The percent of ROS-positive neurons was determined as the number of cells with fluorescent oxidized HEt among those overexpressing PP5.
Neurons were treated with 50 µM H2O2 or 30 µM Aβ25-35 for 8 h, washed with PBS, and solubilized with PBS containing 2% (w/v) sodium dodecyl sulfate (SDS), protease inhibitor cocktail (Roche, Indianapolis, IN), 5 mM sodium fluoride (NaF), 1 mM sodium orthovanadate (NaVO4), and 10 mM β-glycerophosphate. The lysates were then sonicated and centrifuged at 14,000 g for 10 min. Protein concentration was determined with the bicinchoninic acid assay (Thermo Fisher Scientific, Waltham, MA). Equal amounts of protein (15 – 20 µg) were subjected to SDS-PAGE using a 10% (w/v) separation gel and then transferred to nitrocellulose membranes. Membranes were blocked with Tris-buffered saline containing 0.1% (w/v) Tween-20 and 5% (w/v) bovine serum albumin (BSA) and then incubated for 1 h with primary antibodies. Immunoreactivity was detected by incubation with the corresponding horseradish-conjugated secondary antibodies (Jackson Immunoresearch, West Grove, PA) and ECL reagents (GE Healthcare, Piscataway, NJ). Immunoblots were then stripped, and re-probed as indicated in figure legends, using the Re-Blot Plus stripping solution (Millipore, Billerica, MA), following the manufacturer’s instructions. Several exposure times were used for each blot to ensure linearity of band intensities. Densitometry was performed using ImageJ software (National Institutes of Health, USA).
All experiments were performed at least three times using independent cultured neuron preparations. Data are expressed as the mean ± SEM. Comparisons of multiple mean values were performed by one-way analysis of variance followed by the Tukey’s-Kramer post hoc test for significance. Comparisons of two means were performed using a two-tailed paired t test. p values of <0.05 were considered statistically significant. Data were analyzed using GraphPad Prism, version 3.01 (GraphPad Software, San Diego, CA).
Figure 1a shows an example of the cultured neurons used in this study. To determine whether PP5 overexpression can prevent cell death induced by Aβ, we employed Aβ1-40, a major physiological form of Aβ required to develop fully mature senile plaques in AD brains (Shin et al. 1997; Goedert & Spillantini 2006). To test the effect of PP5 on Aβ toxicity, cultured cortical neurons were infected with adenovirus expressing PP5(WT), the catalytically inactive mutant PP5(H304Q), or control adenovirus lacking the PP5 insert. After exposing cells to Aβ1-40 or control reversed sequence peptide (Aβ40-1) for 48 h, cell death was evaluated with the PI exclusion assay. Treatment of neurons with 25 µM Aβ1-40, but not Aβ40-1, induced approximately 50% cell death in neurons infected with control virus (Figs. 1b and d). This level of toxicity achieved with micromolar concentrations of Aβ1-40 is consistent with previous reports in cultured neurons (Morishima et al. 2001; Rymer & Good 2001; Liu et al. 2004; Matharu et al. 2009). Aβ1-40–induced toxicity was reduced in neurons overexpressing PP5(WT). However, neurons overexpressing inactive PP5(H304Q) or subjected to adenoviral infection alone showed a similar toxic response as uninfected cells (Figs. 1c and d). As shown in Figure 1e, Aβ1-40, but not Aβ40-1, exhibited enhanced Th-T fluorescence, confirming the formation of β-sheet-amyloid structured species by Aβ1-40 (LeVine 1999). Western blot analysis using a polyclonal antibody to residues 22–35 of Aβ showed the absence of soluble oligomers in the preparations of fibrillar Aβ1-40 used in this study (Fig. S3).
Similar results were obtained using Aβ25-35, a fragment of Aβ that comprises the biologically active domain of physiological Aβ peptides and causes cell death (Yankner et al. 1990; Morishima et al. 2001) (Fig. 2). Taken together, these observations show that WT-PP5, but not inactive PP5, protects cultured cortical neurons against the toxicity of physiological Aβ aggregates.
Aβ-induced toxicity of cultured neurons is thought to occur through oxidative stress (Behl et al. 1994; Martin et al. 2001; Kadowaki et al. 2005; Sayre et al. 2008). In non-neuronal cells, PP5 overexpression was shown to block H2O2-induced ASK1 activation and cell death (Morita et al. 2001) and also prevented ROS-induced toxicity in a human breast cancer cell line, MCF-7 (Golden et al. 2008). This suggests that the protective effect of PP5 acts downstream of ROS generation. To examine this, we assessed whether PP5 overexpression decreased Aβ-induced ROS generation. For this purpose, neurons were incubated with HEt, a lipophilic dye that serves as an indicator of ROS production (Bindokas et al. 1996). This dye enters the cell and is oxidized to the fluorescent derivative ethidine if ROS are generated, particularly superoxide, and accumulates intercalated into DNA. This probe has been used previously to demonstrate increased superoxide levels in neurons treated with Aβ (Martin et al. 2001) and was compatible with the permeabilization step required for immunocytochemistry in our analysis (Swift & Sarvazyan 2000; Ungvari et al. 2003). Neurons infected with control virus or virus coding for either PP5(WT) or PP5(H304Q) were treated with Aβ for 8 h and analyzed for ROS generation with HEt. Little fluorescence was detected in neurons infected with control virus and treated with control peptide, Aβ35-25, showing that these cells contained only modest amounts of ROS. By contrast, control neurons treated with Aβ25-35 exhibited strong fluorescence, indicative of HEt oxidation to ethidine by ROS (Figs. 3a and c). Neurons infected with PP5(WT) or PP5(H304Q) also accumulated fluorescent ethidine in response to treatment with Aβ25-35 (Figs. 3b and c). Increased ethidine fluorescence was also observed when neurons were exposed to Aβ1-40, compared to the reversed sequence Aβ40-1 peptide, regardless of whether they were infected with control virus or virus encoding PP5(WT) or PP5(H304Q) (Fig. 3d). These results are consistent with previous reports showing that Aβ generates ROS, including superoxide (Martin et al. 2001; Sompol et al. 2008). The observation that PP5 does not prevent the generation of ROS by Aβ (i.e., PP5 does not act as an antioxidant) is consistent with the hypothesis that the neuroprotective effect of PP5 lies downstream of ROS production.
Because oxidative stress is associated with the pathogenesis of AD as well as other neurodegenerative diseases (Lin & Beal 2006; Sayre et al. 2008) and Aβ neurotoxicity is partly mediated through free radical generation (Behl et al. 1994; Kadowaki et al. 2005; Lin & Beal 2006; Sayre et al. 2008), we also examined whether overexpression of PP5 can block neuronal death induced by direct application of ROS. We first determined an H2O2 concentration that induced levels of neuronal cell death similar to Aβ. Figure 4 compares toxicity induced by fibrillar Aβ25-35 or H2O2. We had established that treatment of neurons with fibrillar Aβ25-35 for a period of 48 h induced a concentration-dependent toxicity; approximately 50% cell viability occurred with 30 µM Aβ25-35, comparable to previous reports (Morishima et al. 2001; Abad et al. 2006, Fig. 4a). Treatment with H2O2 for 24 h also caused concentration-dependent cell death. Under these conditions 50 µM H2O2 reduced neuronal viability to approximately 50%, whereas 100 µM H2O2 resulted in less than 25% cell viability (Fig. 4b). These neurotoxic H2O2 concentrations are similar to those previously reported in the literature (Olivieri et al. 2002; Nakamichi et al. 2005; Fuenzalida et al. 2007). We therefore used 50 µM H2O2 for a period of 24 h in our experiments. Neurons infected with control adenovirus or adenovirus encoding PP5(H304Q) were equally susceptible to death induced by this treatment. In contrast, overexpression of PP5(WT) caused a marked reduction in H2O2-induced cell death (Fig. 4c). In contrast, PP5 overexpression did not prevent neurotoxicity induced by staurosporine, a non-selective inhibition of cellular protein kinases and potent inducer of apoptosis (Krohn et al. 1998; Pong et al. 2001) (Fig. S4). These observations are consistent with the hypothesis that the neuroprotective effect by PP5 occurs via blockade of oxidative toxicity and is dependent on the catalytic activity of PP5. However, PP5-mediated neuroprotection against Aβ and oxidative stress is not a universal response to all treatments inducing neuronal toxicity.
To further investigate the role of PP5 in neuroprotection against H2O2 or Aβ-induced toxicity, we used siRNA to suppress PP5 expression in cortical neurons (Fig. 5a). PP5-depleted neurons were more susceptible to toxicity induced by Aβ25-35 than neurons expressing endogenous PP5 levels (Fig. 5b). Similar results were observed with H2O2O toxicity (Fig. 5c). Importantly, this indicates that endogenous PP5 protects neurons from oxidative stress and amyloid β toxicity.
The major stress-activated MAPK pathways, including ERK, JNK, and p38, are implicated as key regulators of toxicity initiated by oxidative stress and Aβ in primary cultures of cortical neurons (Morishima et al. 2001; Crossthwaite et al. 2002; Frasca et al. 2008; Chen et al. 2009), although the role of each pathway in cell death may depend on treatment conditions. Because PP5 can negatively regulate MAPK signaling cascades (Morita et al. 2001; von Kriegsheim et al. 2006), we asked if oxidative stress activated MAPK pathways in our experiments and if PP5 prevented this response. In neurons infected with either control adenovirus or with adenovirus expressing inactive PP5, both H2O2 and Aβ activated ERK (Figs. 6a–c) and JNK (Figs. 6d–f). In contrast, overexpression of PP5(WT) prevented both Aβ- and H2O2-induced activation of ERK or JNK. These results indicate that PP5 overexpression blocks activation of the ERK and JNK MAPK pathways during oxidative stress and this effect requires the catalytic activity of PP5. Chen and colleagues recently reported that treatment of PC12 cells and primary mouse neurons with a toxic dose of H2O2 decreased PP5 protein levels concomitant with activation of MAPK pathways (Chen et al. 2009). Because PP5 can block activation of all three MAPK pathways, they proposed that reduced levels of PP5 contributed to MAPK pathway activation and cell death. Under the conditions used in our study, PP5 protein levels did not change in the time frame of ERK, JNK, or p38 MAPK activation (Fig. 7). Our results indicate that downregulation of PP5 protein levels is not required for oxidative stress- or Aβ-induced activation of the ERK, JNK, or p38 MAPK pathways.
In this study we have shown that PP5 can protect neurons from toxicity induced by Aβ. The neuroprotective effect of PP5 lies downstream of ROS generation by Aβ since PP5 did not suppress the generation of ROS by Aβ, and because PP5 also prevented cell death caused by direct application of the ROS H2O2. The deposition of Aβ is thought to be central in the process of neuronal loss in patients affected with AD. Our results suggest that understanding the function and regulation of PP5 in brain can provide insight into neuronal responses to Aβ, as well as potential therapeutic strategies for the prevention of Aβ-induced neurodegeneration. We also observed that PP5 overexpression blocked activation of the ERK and JNK kinase pathways by Aβ or H2O2 treatment. The catalytic activity of PP5 is essential for its neuroprotective role against Aβ and oxidative stress, as well as for blockade of MAPK pathways. Although others have recently reported that H2O2 treatment of neurons decreases PP5 protein levels (Chen et al. 2009), we did not observe a change in PP5 protein levels during MAPK pathway activation (Fig. 7). This discrepancy may derive from differences in H2O2 concentrations [0.25 – 2 mM (Chen et al. 2009) compared to 50 µM in the present study], duration of treatments, species differences, or cell culture conditions. Regardless of any differences in experimental conditions, our findings indicate that decreased PP5 protein levels are not required for H2O2-induced MAPK pathways activation and neuronal cell death. It has been reported that activity, though not protein levels, of PP5 decreases in the neocortex of brains from patients affected with AD (Liu et al. 2005b). Although we observed no change in protein levels, decreases in PP5 activity could potentially contribute to the activation of MAPK pathways. It will be important in future studies to investigate whether the activity of PP5 is changed in response to Aβ and oxidative stress and to determine if PP5 inhibition of one or more MAPK pathways is responsible for its neuroprotective effect.
In PC12 cells, Chen and colleagues observed that overexpression of PP2A or PP5 partially prevented H2O2-induced cell death (Chen et al. 2009). In our hands, PP5 also partly prevented death caused by H2O2 treatment (2-fold reduction in cell death, Fig. 5), but completely blocked Aβ-induced neuronal cell death (Fig. 1 and Fig 2), despite the fact that the extent of cell death by these two agents was similar in our study. This suggests that direct treatment with H2O2 activates additional cell stress responses that contribute to cell death. Recent evidence indicates that non-fibrillar, oligomeric species of Aβ also contribute to neuronal loss in AD, though a causal relationship to cognitive decline has yet to be established (Yankner & Lu 2009). Levels of Aβ oligomers are elevated in the cerebrospinal fluid and cortex in AD (Gong et al. 2003) and studies in cultured neurons have revealed that Aβ oligomers are more toxic than fibrils, inducing cell death at nanomolar concentrations (Dahlgren et al. 2002; Resende et al. 2008). Oligomeric Aβ has been suggested to mediate neuronal death through inhibition of fast axonal transport to induce synaptic dysfunction (Pigino et al. 2009). To better understand the neuroprotective role of PP5 against Aβ toxicity, it will be important in future studies to address the role of PP5 in oligomeric Aβ-induced neuronal death, as well as in mouse models of familial AD where both oligomeric and fibrillar Aβ have been shown to contribute to neurodegeneration (Goate et al. 1991; Hsiao et al. 1996; Price et al. 1998; Gotz et al. 2001; Meyer-Luehmann et al. 2008).
The deposition of neurofibrillary tangles (NFTs) induced by the accumulation of hyperphosporylated tau protein into paired helical filaments is another important hallmark of AD pathology (Mattson & Magnus 2006; Liu et al. 2005b). Tau hyperphosphorylation pathology correlates with symptom presentation in AD patients (Sayre et al. 2008). Although PP2A is the main phosphatase for the majority of sites hyperphosphorylated in tau found in AD brains, PP5 also dephosphorylates AD tau sites in vitro (Gong et al. 2004) and in neuronal cell lines (Liu et al. 2005b). As mentioned above, PP5 activity is decreased in postmortem neocortex brain samples of AD patients (Liu et al. 2005b). A direct connection between the amyloid and tau hypotheses has been established since Aβ has been shown to increase the levels of NFTs in an AD mouse model with human tau (Gotz et al. 2001). Whether PP5 reduces Aβ-induced tau hyperphosphorylation is another important issue to be addressed. Thyroid hormone has been shown to signal through PP5 (Gentile et al. 2006). In addition to its well-established role in brain development, thyroid hormone plays an important role in protection against neurodegeneration (Tan & Vasan 2009). Therefore it will be of interest to investigate neuroendocrine regulation of PP5 by thyroid hormone and determine whether this pathway protects neurons from toxic insults such as Aβ or ROS.
In addition to AD, oxidative stress plays a major neurotoxic role in other neurodegenerative disorders such as Parkinson’s disease (PD), Huntington’s disease (HD) and Amyotrophic lateral sclerosis (ALS) (Lin & Beal 2006; Mattson & Magnus 2006; Sayre et al. 2008). The observations in this study, together with those of Chen and colleagues (Chen et al. 2008; Chen et al. 2009) demonstrate that PP5 protects neurons against the direct application of ROS, and ROS generated by other toxins [i.e., Aβ in our study and cadmium (Chen et al. 2008)]. This suggests that PP5 may be important in protecting neurons from oxidative stress induced by environmental toxins, aging, or protein aggregation processes in other neurodegenerative diseases as well. Besides neuronal generation of ROS, fibrillar Aβ activates microglia to generate and secrete ROS and pro-inflammatory cytokines, a process thought to further aggravate AD (Kitazawa et al. 2004). In the CNS PP5 is primarily present in neurons (Bahl et al. 2001); thus PP5 may also help combat the toxic effects of ROS generated by microglia, as well as neurons. In summary, the findings of this study indicate that PP5 is an important regulator of neuronal stress response in Aβ and oxidative toxicity and suggest that PP5 may be a useful a target for therapeutic intervention in AD as well as other neurodegenerative diseases.
This study was supported by NIH grant NS031221 (S.R.) and the NIH Intramural Research Program at National Institute of Environmental Health Sciences through Grant Z01-ES080043 (D.L.A.). PP5 adenovirus constructs were prepared at the Gene Transfer Vector Core at the University of Iowa, supported in part by the NIH and the Roy J. Carver Foundation.