|
Advertisement | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
J. Biol. Chem., Vol. 281, Issue 19, 13559-13565, May 12, 2006
The Interaction between Cytoplasmic Prion Protein and the Hydrophobic Lipid Core of Membrane Correlates with Neurotoxicity*![]() ![]() ![]() ![]() 1
From the
Received for publication, November 16, 2005 , and in revised form, March 10, 2006.
Prion protein (PrP), normally a cell surface protein, has been detected in the cytosol of a subset of neurons. The appearance of PrP in the cytosol could result from either retro-translocation of misfolded PrP from the endoplasmic reticulum (ER) or impaired import of PrP into the ER. Transgenic mice expressing cytoplasmic PrP (cyPrP) developed neurodegeneration in cerebellar granular neurons, although no detectable pathology was observed in other brain regions. In order to understand why granular neurons in the cerebellum were most susceptible to cyPrP-induced degeneration, we investigated the subcellular localization of cyPrP. Interestingly, we found that cyPrP is membrane-bound. In transfected cells, it binds to the ER and plasma/endocytic vesicular membranes. In transgenic mice, it is associated with synaptic and microsomal membranes. Furthermore, the cerebellar neurodegeneration in transgenic mice correlates with the interaction between cyPrP and the hydrophobic lipid core of the membrane but not with either the aggregation status or the dosage of cyPrP. These results suggest that lipid membrane perturbation could be a cellular mechanism for cyPrP-induced neurotoxicity and explain the seemingly conflicting results concerning cyPrP.
Prion diseases are a group of fatal neurodegenerative disorders that are sporadic, inherited, or transmissible (1). The resistance of PrP2 knock-out mice to prion disease revealed the essential role of endogenous PrP in developing the two characteristics of prion disease, infectivity and neurodegeneration (2). In prion disease, a portion of PrP converts from its normal conformation, PrPC, to an insoluble and protease-resistant pathogenic conformation, PrPSc (1). Compelling evidence supports that PrPSc is the transmissible element in prion disease (3-5).
Newly synthesized PrP contains an N-terminal signal sequence, which directs PrP to be co-translationally imported into the lumen of the ER. The signal sequence is cleaved within the ER, along with a C-terminal glycosylphosphatidylinositol signal sequence that is responsible for glycosylphosphatidylinositol anchor addition (6). Once folded correctly, PrP passes through the ER quality control machinery, maturing along the secretory pathway en route to the cell surface. In addition to cell surface and luminal localizations, cytosolic localized PrP has also been detected (7, 8). Recent studies revealed that the cytosolic appearance of PrP could result from two independent but not mutually exclusive pathways. In a variety of cell lines, including primary neurons expressing endogenous PrP, it has been found that some PrP molecules misfold in the ER and are retro-translocated to the cytosol for proteasome degradation via the ER-associated degradation pathway (9-13). Disease-associated PrP mutants have an increased tendency to misfold and are subjected to ER-associated degradation (9, 10, 12). An alternative pathway, the unsuccessful import of PrP because of the insufficiency of the PrP signal sequence, was revealed by an elegant study in mammalian cells (14) and by heterologous expression of PrP in yeast (15). In addition, mutations at the C-terminal globular domain of PrP have also been found to interfere with PrP import into the ER (16). We have investigated the biological consequences of PrP in the cytosol by using a variety of techniques (12, 17-19), and we found that the cytosolic accumulation of PrP correlates with toxicity in some neuronal cells (19). In transgenic mice expressing cyPrP (mouse PrP23-230 without signal sequences) under the control of the PrP promoter, rapid cerebellar granular neuron degeneration was observed in one transgenic line, 2D1, although neurodegeneration progressed slower in a separate line, 1D4 (19). The two transgenic lines developed identical neuropathology in a transgene dosage-dependent manner (19), demonstrating the correlation between cyPrP transgene expression and neurodegeneration. Notably, cyPrP was present at a much lower level than endogenous PrP and sensitive to proteinase K digestion, indicating that the neurotoxic species is not the PrPSc form. Interestingly, the subset of neurons in the hippocampus, neocortex, and thalamus of mouse brain appeared healthy even though cytosolic localized PrP was detected in these neurons (7, 8). Similarly, conflicting results on cyPrP-induced toxicity in cultured cells have been obtained by different groups (11, 14, 19-22). These discrepancies suggest that the cellular mechanism of cyPrP-induced toxicity is likely to be influenced by other cellular factors, which might include differences in cell types or growth environments. In this study, we investigated cyPrP in transgenic mice and in cultured cells in order to understand the cellular mechanism of its toxicity. A specific correlation between cyPrP-induced neurotoxicity and its interaction with the hydrophobic lipid core of membrane was found, suggesting membrane perturbation as the cellular basis for cyPrP-induced neurotoxicity.
Cell Culture and TransfectionN2A cells were purchased from the American Type Culture Collection (ATCC) and cultured according to ATCC-recommended conditions, with 10% fetal bovine serum (ATCC) in Eagle's minimum essential medium (ATCC) media. Penicillin/streptomycin (Invitrogen) was included in the culture media. ExGen 500 reagent (Fermentas) was used to transfect N2A cells. Preparation of Post-nuclear Supernatant (PNS)From cultured cells, one well of transfected cells on a 6-well cell culture plate was scraped into 350 µl of membrane preparation buffer (8% sucrose, 20 mM Tris-HCl, pH 8.0, 1 mM EDTA), homogenized, and centrifuged at 1,000 x g for 5 min at 4 °C to remove nuclei and unbroken cells. For mouse brains, half of the cerebellum or the rest of the brain was homogenized in the membrane preparation buffer and centrifuged at 1,000 x g for 5 min at 4 °C to remove nuclei and unbroken cells. Protein concentration of the PNS was determined with the Bio-Rad protein assay. Pools of eight transgenic mice brains were used for the membrane extraction experiments. Separating Membrane and Cytosolic Fractions Using Iodixanol GradientPNS containing 100-300 µg of protein from mouse brain or 280 µl of PNS from cultured cells was mixed with stock iodixanol solution (OptiPrep, Axis-Shield PoC, AS) to generate 1 ml of 36% iodixanol solution, and the mixture was loaded at the bottom of an ultracentrifuge tube. 1 ml of 31% iodixanol solution and 0.4 ml of 5% iodixanol solution were loaded on top of the sample sequentially. The gradient was centrifuged at 199,000 x g for 3 h at 4°C in a Sorvall micro-ultracentrifuge. Two hundred-microliter fractions were collected from the top to the bottom, and proteins in all fractions were precipitated by adding 200 µl of 40% trichloroacetic acid. Precipitated proteins were sonicated in SDS-PAGE sample buffer and subjected to electrophoresis on a 14% SDS-polyacrylamide gel (NOVEX Tris-glycine gel; Invitrogen) followed by immunoblot analysis. For membrane extractions, PNS prepared from cerebella or the rest of the brains was centrifuged at 346,000 x g for 30 min at 4 °C. The pellet was resuspended in a solution containing 0.5 M NaHCO3, pH 11, 1 M KCl, 1.5 M KCl, or 1.5 M KCl plus 10 mM NaOH and loaded on the iodixanol gradient. Fractions were collected from the top to the bottom, and the presence of cyPrP was determined by immunoblot analysis.
Subcellular Fractionation Using Sucrose GradientSubcellular fractionation was performed essentially according to the method reported previously (23). Briefly, cell pellets from two 100-mm cell culture dishes were homogenized to prepare PNS. The discontinuous sucrose gradient consisted of PNS in 0.25 M sucrose ( Determining the Aggregation Status of cyPrP from Transgenic Mouse BrainsThe cerebellum and the rest of the brain were separated and homogenized in phosphate-buffered saline (1:10, w/v). Protein concentration was determined using Bio-Rad protein assay. Brain lysates used for determining cyPrP aggregation status are as follows: 10 µg of cerebellum lysate from a wild-type mouse, 5 µg of the rest of the brain lysate from a wild-type mouse, 50 µg of cerebellum lysate from a transgenic mouse on the PrP knock-out background, 50 µg of the rest of the brain lysate from a transgenic mouse on the PrP knock-out background. These brain lysates were resuspended in 100 µl (final volume) of buffer containing 150 mM NaCl, 5 mM Tris-HCl, pH 8.0, 2 mM EDTA, 0.5% Triton X-100, and 0.5% sodium deoxycholate (final concentrations). After incubation on ice for 10 min, lysates were centrifuged at 16,300 x g for 30 min at 4 °C. The supernatant was carefully transferred to a new tube, precipitated with 4 volumes of cold methanol (-20 °C), and incubated at -20 °C for more than 30 min. Pellet fraction and precipitated supernatant proteins were sonicated in SDS-PAGE sample buffer containing 5% (w/v) SDS. After boiling, samples were separated in 14% SDS-PAGE followed by immunoblot analysis. Subcellular Fractionation of Neuronal StructuresFractionation of mouse brain homogenate was performed according to published protocols (24). Immunoblot Analysis and AntibodiesSamples separated by SDS-PAGE were transferred to poly(vinylidene difluoride) membrane for immunoblot analyses. Blots were developed using peroxidase-conjugated secondary antibodies (Bio-Rad Laboratories) and ECL-plus reagent (Amersham Biosciences). Quantification was performed on a Storm PhosphorImager System (Amersham Biosciences). Antibodies used were as follows: 3F4 monoclonal anti-PrP antibody (Signet Laboratories) at 1:2500; POM1 anti-PrP antibody (25, 26) at 1:2500; anti-calnexin antibody (StressGen) at 1:2000; anti-Hsp70 antibody (Stress-Gen) at 1:1000; anti-syntaxin 6 antibody (Santa Cruz Biotechnology) at 1:200, anti-porin antibody (Sigma) at 1:1000; anti-synaptophysin antibody (Sigma) at 1:1000; anti-synapsin I (Chemicon) at 1:1000; and anti-cPLA2 antibody (Santa Cruz Biotechnology) at 1:200. Immunohistochemical StainingThe tissue processing for immunohistochemical staining was performed as described previously (19). The anti-cPLA2 antibody (Santa Cruz Biotechnology) was used at 1:50. A rabbit polyclonal anti-GFAP antibody (DAKO) was used at 1:400 for GFAP staining in the mouse brains. ABC, AKP, and DAB kits were purchased from the Vector Laboratories. Preparation of LiposomeOne gram of mouse brain was homogenized in 10 ml of methanol for 1 min, followed by addition of 20 ml of chloroform to the mixture and homogenization for an additional 2 min. The homogenate was filtered, and the solid residues were resuspended in chloroform/methanol (2:1, v/v, 30 ml) and homogenized for 3 min. After repeating this step once more, the combined filtrates were washed with 0.25 volume of 0.88% KCl in water followed by a wash with 0.25 volume of water/methanol (1:1). The purified lipids in the bottom layer were collected and dried under nitrogen. To prepare liposomes, 20 mM Tris buffer, pH 7.5, was added to the dried lipids, and the solution was sonicated until it became clear.
Higher Level of Cytosolic Phospholipase A2 (cPLA2) in the Degenerating CerebellumThe cPLA2 is a calcium-responsive phospholipase, belonging to a large PLA2 family. It has been implicated as a contributor to the pathogenesis of a variety of neurodegenerative disorders, including prion diseases (27-30). We studied the protein level of cPLA2 in transgenic mice expressing cyPrP under the control of the PrP promoter (19). The 2D1 line of transgenic mice developed neurodegeneration with a synchronized onset. Ultrastructural changes were detected by electron microscopy at 3 weeks of age, ataxia at 4 weeks, severe loss of granular neurons at 5 weeks, and overt gliosis at about 7 weeks (19). Taking advantage of the synchronized onset of neurodegeneration in this transgenic line, we asked whether there is a correlation between cPLA2 induction and cyPrP-induced neurotoxicity. An increase of cPLA2 was detected in the cerebellum of a 3-week-old transgenic mouse but not in the rest of the brain, nor in the cerebellum or the rest of the brain of a wild-type littermate (Fig. 1A). Because the cPLA2 induction correlated with the onset of neurodegeneration (19), we tested whether the cPLA2 level continued to increase once neurodegeneration was in progress. Indeed, a higher level of cPLA2 was found in the cerebellum of a 4-week-old transgenic mouse compared with its wild-type littermate (Fig. 1B). These results indicated that higher level of cPLA2 was associated with cyPrP-caused neurodegeneration. Generally, cPLA2 is involved in the production of eicosanoids and is associated with reactive astrocytes (27, 28). To determine the relationship between the increase in cPLA2 and the reactive astrocytes, we performed immunohistochemical staining with antibodies against both cPLA2 and GFAP, a marker for reactive astrocytes. Compared with their wild-type littermates, a clear increase of cPLA2 immunoreactivity was detected in the cerebellum of transgenic mice at 4 weeks of age. However, a significant increase of GFAP immunoreactivity was not detected until mice reached 7 weeks of age (Fig. 1C). This result indicates that the increase of cPLA2 immunoreactivity might not be a direct result of astrogliosis.
CyPrP Is Associated with Membranes in N2A CellsAn alternative explanation of increased cPLA2 levels could be that cyPrP disrupts lipid membranes, causing increased cytosolic calcium concentrations and subsequently an increase of the calcium-responsive cPLA2. If this were the case, cyPrP should be associated with cell membranes. To test this possibility, post-nuclear supernatants were prepared from N2A cells transfected with cyPrP. Iodixanol reagent was mixed with PNS to reach a final concentration of 36% and loaded at the bottom of a discontinuous iodixanol gradient (13). In this gradient, membrane fractions float to the top, which avoids the influence of the aggregation status of cyPrP in separating membrane and cytosolic fractions. CyPrP from N2A cells was detected in the top fractions of the gradient, along with the transmembrane protein calnexin (Fig. 2A), while the cytosolic chaperone Hsp70 remained in the bottom fractions (Fig. 2A). When recombinant PrP23-230 alone was tested in the same gradient, it remained in the bottom fractions as well. These results indicate that a large part of cyPrP in N2A cells is membrane-associated. CyPrP Is Associated with ER and Plasma/Endocytic Vesicular Membranes in N2A CellsTo understand the consequence of cyPrP membrane association, we determined the identity of the subcellular organelles with which it associated. PNS from N2A cells transfected with cyPrP was loaded onto a discontinuous sucrose gradient (23). The migration of subcellular organelles in the gradient was determined by immunoblot analysis with antibodies raised against specific subcellular organelle antigens. The transfected cyPrP contains a 3F4 epitope (sequences derived from human and hamster PrP) that is not present in murine endogenous PrP from N2A cells. Therefore, cyPrP can be specifically detected without interference from endogenous PrP. We found that cyPrP migrated into two peaks in the sucrose gradient (Fig. 3, cyPrP). One of the peaks co-localized with endogenous glycosylated PrP (Fig. 3, endogenous PrP), indicating that cyPrP was associated with plasma/endocytic vesicular membranes (6). The second peak co-migrated with the peak of the ER marker calnexin (Fig. 3, calnexin). There was some overlap between the first cyPrP peak and the Golgi marker syntaxin 6, although the highest peak of each protein appeared in different fractions. No association of cyPrP with the mitochondrial marker, porin, was detected. Therefore, we concluded that in N2A cells, cyPrP is mainly associated with either ER or plasma/endocytic vesicular membranes.
CyPrP Is Associated with Synaptic and Microsomal Membranes in Transgenic MiceTransiently transfected cells generally express high levels of cyPrP. To rule out the possibility that membrane association of cyPrP is an artifact because of high levels of cyPrP expression, we tested whether cyPrP is membrane-associated in 1D4 transgenic mice. In this line of transgenic mice, a very low protein level of cyPrP was detected, and the progression of neurodegeneration was much slower compared with 2D1 mice (19). Because these mice were generated on the wild-type mouse background and the cyPrP transgene contains no easily recognizable epitope, we crossed cyPrP transgenic mice on to the PrP knock-out background. The newly generated transgenic mice expressing cyPrP on the PrP knock-out background developed a phenotype similar to that of 1D4 transgenic mice, specifically, degeneration of the cerebellar granular neurons but no detectable degeneration in other brain regions.3 PNS was prepared from the cerebella of these mice and subjected to iodixanol gradient separation. PrP knock-out mouse brain lysate was included in the immunoblot analysis as a negative control (Fig. 4A, lane C). Although it is expressed at a much lower level compared with transfected N2A cells, cyPrP in the transgenic mice was also membrane-associated, migrating to the top of the gradient (Fig. 4A).
To determine which subcellular membrane cyPrP is associated with in the transgenic mice, we separated different neuronal membrane structures using differential centrifugation (24). Synaptophysin and synapsin I were used as markers for the separation (Fig. 4B). Synaptophysin is enriched in the LP2 fraction, representing a synaptic vesicle-enriched fraction (31). In addition to LP2, synapsin I was also detected in the LP1 fraction, a fraction enriched with synaptosomal membranes (32). Notably, cyPrP was detected in both LP1 and LP2 fractions, indicating that it is associated with synaptic membranes (Fig. 4B). In addition, cyPrP was also detected in the P3 fraction, a fraction enriched with light membranes and microsomes (24). Therefore, the localization of cyPrP in transgenic mice is similar to that in transfected N2A cells, namely the plasma and ER/microsomal membranes. Cerebellar Neurodegeneration Caused by cyPrP Does Not Correlate with Its Expression Level, Aggregation Status, or Membrane AssociationNext, we investigated the relationship between the biochemical properties of cyPrP and neurotoxicity. The PrP promoter was used to drive expression of the cyPrP transgene in mice, and most intriguingly, despite broad pan-neuronal expression of this promoter (33), the only detectable pathology occurred in the granular neurons of the cerebellum (19). We reasoned that a comparison between cyPrP in the cerebellum and the rest of the brain might reveal the cellular basis for cyPrP-induced neurotoxicity.
We first compared the aggregation status of cyPrP to determine whether it correlates with the development of neurotoxicity. Using mild detergent extraction followed by sedimentation, we found that almost all the endogenous PrP from a wild-type mouse was soluble (Fig. 5A). In contrast, cyPrP from both the cerebellum and the rest of the brain was aggregated, appearing in the pellet fraction. This result rules out the possibility that a specific neuronal environment induces cyPrP aggregation and leads to neurotoxicity. An alternative possibility could be that a higher level of cyPrP expression in the cerebellum is responsible for the rapid neurodegeneration. By using serial dilutions, we compared the protein level of cyPrP in the cerebellum to that in the rest of the brain (Fig. 5B). As we reported previously (19), cyPrP is expressed at a very low level in the brain of these transgenic mice. Quantification using a homozygous cyPrP transgenic mouse revealed that the levels of cyPrP expression in the cerebellum were about 3% of endogenous PrP, whereas cyPrP in the rest of the brain was about 6% of endogenous PrP. Thus, the expression level of cyPrP cannot account for the rapid cerebellar granular neuron degeneration, as cyPrP expression level is notably lower in the cerebellum.
Because cyPrP is membrane-bound, it is possible that cyPrP binds to the membranes in the cerebellar granular neurons but not in other neurons, which could explain the cerebellar neurodegeneration in these mice. However, when PNS prepared from rest of the brain was tested, cyPrP was again found to be associated with membranes (Fig. 5C). A small amount of cyPrP remained in the bottom fractions (compare Fig. 5C with Fig. 4A), indicating some cyPrP molecules might not be membrane-bound. But this small difference cannot explain the severe degeneration in the cerebellum granular neurons, compared with the almost absence of degeneration in other brain regions (19). The Interaction between cyPrP and the Hydrophobic Lipid Core of Membrane Correlates with cyPrP-caused Cerebellar NeurodegenerationMembrane interactions are divided into interaction with the hydrophobic lipid core or peripheral membrane association. The latter is achieved through interaction with membrane proteins or the charged head group of phospholipids. Compared with a peripherally membrane-associated protein, a protein interacting with the hydrophobic lipid core is much more likely to cause membrane disruption. Therefore, we tested the hypothesis that cyPrP interacts with the hydrophobic lipid core of membrane in cerebellar granular neurons, whereas it is only peripherally associated with membrane surfaces in other brain regions. Membranes prepared from either cerebella or rest of the brains were extracted with 0.5 M NaHCO3 buffer, pH 11, to remove peripheral proteins from membranes. Notably, all the cyPrP from the rest of the brain appeared in the bottom fractions after the alkaline extraction (Fig. 6, rest of the brain). In contrast, part of the cyPrP from the cerebellum resisted the extraction and remained in the top membrane fractions of the gradient (Fig. 6, cerebellum). Coomassie Blue staining of the blots revealed that there was no significant difference in the protein distribution pattern between the two, supporting the specificity of the experiment. Because cyPrP is not an integral membrane protein, extraction with the 0.5 M NaHCO3 buffer, pH 11, might be too harsh and lead to underestimating the amount of cyPrP interacting with the hydrophobic lipid core of the membrane. To test this possibility, we extracted membranes with a series of buffer conditions that were known to release peripheral proteins. As shown in Fig. 7A, the majority of cyPrP from the cerebellum remained in the membrane fractions after all these treatments. In contrast, the majority of cyPrP from the rest of the brain was removed from membranes after 1.5 M KCl extraction and migrated to the bottom of the gradient. When 10 mM NaOH was included to eliminate charge-based interactions, almost all the cyPrP in the rest of the brain was released from membranes (Fig. 7A). The PNS used in these experiments contained pools of eight different transgenic mice to ensure that the observation was not associated with a specific animal. In addition, the fact that aggregated cyPrP from the rest of the brain remained at bottom of the gradient after extraction supports the specificity of this assay for detecting membrane-bound cyPrP.
To verify that cyPrP could bind to mouse brain lipids, the interaction between recombinant PrP23-230 (identical to cyPrP in amino acid sequences) and mouse brain lipids was tested. Recombinant PrP did indeed bind to the liposomes made of mouse brain lipids and migrated to the top of the iodixanol gradient (Fig. 7B, compare with Fig. 2B). Moreover, the majority of PrP remained at the top of the gradient after extraction with 1.5 M KCl or 0.5 M NaHCO3 buffer, pH 11, suggesting that PrP can interact with the hydrophobic hydrocarbon core of lipid bilayer. Together, these results suggest that cyPrP interacts with the hydrophobic lipid core of the membrane in the cerebellum but not in the rest of the brain. The correlation between the interaction of cyPrP with the hydrophobic lipid core of membrane and neurotoxicity provides an explanation for the cerebellar neurodegeneration in these mice.
We found that cyPrP is associated with lipid membrane, mainly binding to ER or plasma/endocytic vesicular membranes in cultured cells and synaptic and microsome membranes in transgenic mice. More importantly, we found that the interaction between cyPrP and the hydrophobic lipid core of membrane correlates with neurotoxicity in transgenic mice. Previous studies revealed that PrP interacts with lipids (34-38) and that recombinant PrP can bind and disrupt liposomes composed of negatively charged phospholipids (36). Based on our results and previous findings, we propose the following model for cyPrP-caused neurotoxicity (Fig. 8).
CyPrP, either from retro-translocation (9-13) or from the impaired import into the ER (14, 15), is destined for proteasome degradation. When some PrP escapes proteosome degradation, they can bind to either protein or membrane lipids. Only if it interacts with the hydrophobic lipid core of membrane will PrP lead to the disruption of the phospholipid bilayer. Aggregation of lipid-bound PrP exacerbates the membrane disruption, and the change in membrane permeability ultimately causes neuronal death. On the other hand, interaction with cytosolic or membrane proteins may prevent the lethal lipid interaction and protect cells against toxicity. According to this model, neuronal cell susceptibility may depend on the following parameters. First, different proteasome activity in different cells will determine the efficiency for removing cyPrP. Second, the presence of different proteins in certain neurons could bind to cyPrP and prevent its lethal lipid interaction. Third, different intracellular environments may determine the aggregation status of cyPrP. Fourth, different lipid compositions of the membranes may create resistance to cyPrP-induced membrane perturbation. Fifth, different types of cells may have different tolerances for the biological consequences caused by the alteration in membrane permeability. This model does not require all the cyPrP molecules within a neuron to interact with the hydrophobic lipid core of membrane in order to cause toxicity. As long as the interaction is sufficient to alter membrane permeability, the disturbance of intracellular ion concentrations will lead to cell death. In addition, this model does not exclude the possibility that the cyPrP bound to the hydrophobic lipid core of membrane can interact with the transmembrane domains of integral membrane proteins. Actually, this interaction may also contribute to the alteration of membrane permeability if the integral membrane protein is involved in regulating ion concentrations. The observation that cytoplasmic PrP interacts with the hydrophobic lipid core of membrane in the cerebellum but not in other brain regions explains the paradoxical results concerning cyPrP-caused toxicity. For example, it causes rapid neurodegeneration in cerebellar granular neurons (19) but not in a subset of neurons in other brain regions (7, 8). According to our model, the specific cellular environment in the cerebellar granular neurons might be more conducive to the interaction of cyPrP with membrane lipids. The presence of PrP-binding proteins in other neurons may prevent the lethal lipid interaction and neurodegeneration. Therefore, given a higher dosage and longer time period, cytoplasmic PrP should be able to cause neurodegeneration in those neurons as well. We found that ER and plasma membranes are the main membrane structures with which cyPrP interacted. Both membranes are known to regulate intracellular calcium concentrations (39), which are critical for cell signaling. Disrupting either of these membranes will lead to an increase in cytosolic calcium concentrations. In an effort to reduce cytosolic calcium loads, mitochondria are known to import cytosolic calcium through their membranes (40, 41). An overload of calcium will cause mitochondrial membrane disruption, which is the earliest pathological change detected in the transgenic mice expressing cyPrP (19). Moreover, an increase of cytosolic calcium concentration may also explain the increase of calcium-responsive cPLA2 immunoreactivity in the region of neurodegeneration (Fig. 1). An interesting finding was that cyPrP in human primary neurons not only did not cause toxicity, it also protected these cells against Bax-induced apoptosis (11). It is known that ER calcium concentration is critical for apoptosis (42), and lower ER calcium stores will render cells more resistant to apoptosis. The anti-apoptotic protein Bcl-2 decreases ER calcium stores and protects the cell against apoptosis (43, 44). It is possible that in human primary neurons cyPrP causes ER calcium leakage. Although the calcium leakage may not be enough to cause cell death, the decrease of ER calcium stores could be sufficient to protect them from Bax-induced apoptosis. Recent reports have shown that some PrP molecules can be detected in the nucleus of cultured cells as well (45, 46). Because PNS was used in all the experiments described here, we cannot rule out the possibility that nuclearly localized cyPrP may also contribute to toxicity. Nonetheless, our results reveal that cyPrP is membrane-bound, and its interaction with the hydrophobic lipid core of membrane correlates with neurotoxicity, which provides an explanation for the paradoxical findings concerning cyPrP. In addition, our model also predicts that it may take either a higher dosage or a longer time period for cyPrP to cause neurotoxicity in neurons other than cerebellar granular cells. Further investigation is required to test this prediction.
* This work was supported in part by a new scholar award from the Ellison Medical Foundation (to J. M.) and a Pfizer/AFAR Innovations in Aging research grant (to J. M.). The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. 1 To whom correspondence should be addressed: Dept. of Molecular and Cellular Biochemistry, Ohio State University, 1645 Neil Ave., Columbus, OH 43210. Tel.: 614-688-0408; Fax: 614-292-4118; E-mail: ma.131{at}osu.edu.
2 The abbreviations used are: PrP, prion protein; ER, endoplasmic reticulum; cyPrP, cytoplasmic PrP23-230; cPLA2, cytosolic phospholipase A2; PNS, post-nuclear supernatant; GFAP, glial fibrillary acidic protein.
3 X. Wang and J. Ma, unpublished results.
We thank Dr. Adriano Aguzzi (Institute of Neuropathology, University Hospital of Zürich) for providing us the POM1 monoclonal anti-PrP antibody. We thank Andrew Steele and Walker Jackson for critically reading this manuscript.
This article has been cited by other articles:
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
Advertisement | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||