Expression of Mutant or Cytosolic PrP in Transgenic Mice and Cells Is Not Associated with Endoplasmic Reticulum Stress or Proteasome Dysfunction

The cellular pathways activated by mutant prion protein (PrP) in genetic prion diseases, ultimately leading to neuronal dysfunction and degeneration, are not known. Several mutant PrPs misfold in the early secretory pathway and reside longer in the endoplasmic reticulum (ER) possibly stimulating ER stress-related pathogenic mechanisms. To investigate whether mutant PrP induced maladaptive responses, we checked key elements of the unfolded protein response (UPR) in transgenic mice, primary neurons and transfected cells expressing two different mutant PrPs. Because ER stress favors the formation of untranslocated PrP that might aggregate in the cytosol and impair proteasome function, we also measured the activity of the ubiquitin proteasome system (UPS). Molecular, biochemical and immunohistochemical analyses found no increase in the expression of UPR-regulated genes, such as Grp78/Bip, CHOP/GADD153, or ER stress-dependent splicing of the mRNA encoding the X-box-binding protein 1. No alterations in UPS activity were detected in mutant mouse brains and primary neurons using the UbG76V-GFP reporter and a new fluorogenic peptide for monitoring proteasomal proteolytic activity in vivo. Finally, there was no loss of proteasome function in neurons in which endogenous PrP was forced to accumulate in the cytosol by inhibiting cotranslational translocation. These results indicate that neither ER stress, nor perturbation of proteasome activity plays a major pathogenic role in prion diseases.


Introduction
Prion diseases are fatal degenerative disorders of the central nervous system (CNS) that can arise sporadically, be genetically inherited, or acquired through infection [1,2]. The key pathogenic event is conversion of the cellular prion protein (PrP C ) into a selfpropagating aberrant conformer (PrP Sc ) [1], which induces neurodegeneration through an unknown mechanism [3]. PrP C is a glycosylphosphatidylinositol (GPI) membrane-anchored glycoprotein of uncertain function, mainly expressed by neurons in the CNS [4]. Like most membrane-associated proteins, PrP C is cotranslationally translocated into the endoplasmic reticulum (ER), where it undergoes oxidative folding and facultative Nlinked glycosylation. After transit in the Golgi, PrP C is delivered to the cell surface, where it resides in lipid rafts. Cell surface PrP C can be released into the extracellular space, or internalized to an endosomal compartment, from which it is either recycled to the plasma membrane or diverted to lysosomes for degradation [5].
Mutations in the gene encoding PrP C are responsible for genetic prion diseases, which include familial Creutzfeldt-Jakob disease (fCJD), Gerstmann-Strä ussler-Scheinker (GSS) syndrome and fatal familial insomnia (FFI) [6,7]. These diseases are inherited in an autosomal dominant fashion and are thought to be due to a toxic effect of misfolded mutant PrP molecules [3]. Tg(PG14) mice expressing the mouse PrP homologue of a 72amino acid insertion associated with a mixed CJD-GSS phenotype develop a progressive neurological illness characterized by ataxia and massive apoptosis of cerebellar granule neurons [8,9]. Tg(CJD) mice carrying the mouse PrP homologue of the D178N/V129 PrP mutation (moPrP D177N/V128) develop key features of fCJD, including cognitive, motor and neurophysiological abnormalities [10]. Tg(PG14) and Tg(CJD) mice synthesize in their brains misfolded forms of mutant PrP with biochemical properties reminiscent of PrP Sc , including insolubility in non-denaturing detergents, and resistance to mild proteinase-K (PK) digestion [8,10]. Analysis of PG14 and D177N/V128 metabolism and localization in cultured neurons showed that the biosynthetic maturation of these mutants in the ER is delayed, and they accumulate abnormally in this organelle [10,11,12]. This suggests that perturbation of ER homeostasis due to accumulation of misfolded PrP may play a pathogenic role.
The ER has a robust quality control system that prevents misfolded and mismodified proteins from exiting the ER and being transported through the secretory pathway. This system includes chaperones and folding catalysts that form a matrix on which newly synthesized proteins attain their final conformation [13]. When protein misfolding occurs and unfolded proteins accumulate and aggregate in the ER, cells experience ''ER stress'' and activate an adaptive signal transduction pathway, known as the unfolded protein response (UPR). The UPR enhances the folding capacities in the ER, improves protein disposal through ER-associated degradation (ERAD), and reduces the rate of protein synthesis and translocation into the ER lumen [14].
The UPR signals through three distinct ER stress sensors on the ER membrane [15]. One is IRE1a (inositol-requiring transmembrane kinase and endonuclease), a type I ER transmembrane protein whose cytoplasmic region carries protein kinase and endoribonuclease domains. In ER-stressed cells IRE1a multimerizes and autophosphorylates, setting in motion its RNAse activity. Activated IRE1a initiates the unconventional splicing of the mRNA encoding the transcriptional factor X-box-binding protein 1 (XBP1) to produce sXBP1, a more stable form of XBP1 with a potent transactivator domain that enhances transcription of genes involved in protein folding, secretion and ERAD [16].
Another ER stress sensor is the activating transcription factor 6 (ATF6) [17]. This is a type II ER transmembrane protein whose cytosolic domain contains a bZIP transcriptional factor. When protein misfolding and accumulation occur, ATF6 is transported to the Golgi where it is processed within the transmembrane domain to release the cytosolic domain, which translocates to the nucleus and induces expression of the ER chaperone Grp78/BiP, and XBP1. Therefore, the IRE1a and ATF6 signaling pathways merge through regulation of XBP1 activity [18].
The third sensor is the ER-associated kinase PERK (doublestranded RNA-activated protein kinase-like ER kinase), which phosphorylates the a subunit of eukaryotic translation initiation factor 2 (eIF2a). This inhibits protein translation, reducing the overload of misfolded proteins. This pathway also activates ATF4, a transcription factor that induces expression of CHOP/ GADD153, which can trigger apoptosis when cells unable to handle the unfolded protein load become irreversibly damaged. Thus, persistent ER stress eventually leads to cell death.
In this study we explored the role of ER stress in genetic prion diseases by measuring the levels of key elements of the UPR in brain tissue and primary neurons from Tg(PG14) and Tg(CJD) mice, and in transfected cells expressing mutant PrP under the control of a tetracycline-inducible promoter. Because ER stress favors formation of a cytosolic pool of untranslocated PrP [19,20] that might aggregate and inhibit proteasome activity [21], we also examined the function of the ubiquitin-proteasome system (UPS) in the transgenic models, and in neurons in which PrP C was forced to accumulate in the cytosol by interference with the correct insertion of the PrP signal peptide into the translocon.

ER Stress Markers in Transgenic Mice and Primary Neuronal Cultures
To test whether mutant PrP expression triggered toxic response pathways in the ER, we analyzed the splicing of XBP1 mRNA transcripts and the levels of Grp78/BiP, Grp58/Erp57 and CHOP/GADD153 in the brains of Tg(PG14) and Tg(CJD) mice at different stages of their neurological illness. Tg(PG14) mice develop a slowly progressive neurological syndrome characterized neuropathologically by synaptic loss in the molecular layer of the cerebellum and massive apoptosis of cerebellar granule neurons [8,9,22]. They develop ataxia, kyphosis and foot clasp reflex at ,240 days of age and die at ,450 days [8,9]. Tg(CJD) mice develop alterations of spatial working memory between 200 and 320 days of age, electroencephalographic abnormalities at $360 days, and a motor syndrome involving alterations of motor coordination and body posture that is first seen at ,450 days and increases in severity as the mice age [10]. Consistent neuropathological findings in Tg(CJD) mice are loss of calretinin-positive neurons in the hippocampus and neocortex, and ER swelling with accumulation of mutant PrP in cerebellar granule neurons [10].
ER stress markers were analyzed in a total of 23 Tg(PG14) mice aged 23,24,36,102,120,124,131,146,171,172,189,210,214,226,240,241,296,364,372,377,404,405 and 409 days, and five Tg(CJD) mice at 81, 338, 411, 421 and 814 days. RNA and proteins were extracted from the whole brain or specific brain regions showing pathological changes (the cerebellum in Tg(PG14) mice, and the cerebral cortex, hippocampus and cerebellum in Tg(CJD) mice). Non-Tg littermates and age-matched Tg(WT) mice expressing wild-type PrP [8] were used as controls.
XBP1 splicing was readily detected in the liver of mice and in cells treated with tunicamycin, which induces ER stress by perturbing the folding efficiency in the ER, whereas no splicing was observed in the whole brain or dissected brain regions of the mutant mice at any stage of their disease (Fig. 1A-C and data not shown). There was no increase in the levels of Grp78/BiP or CHOP/GADD153 transcripts, assessed by Northern blot (not shown), or Grp58/Erp57 and Grp78/BiP proteins by Western blot of brain extracts ( Fig. 2A-D).
In a final set of experiments, we used immunohistochemical techniques to visualize UPR activation in brain sections of four Tg(PG14) mice aged 145, 180 and 367 (two mice) days, and two Tg(CJD) mice aged 542 and 751 days. As a positive control for the staining, Grp78/BiP and CHOP/GADD153 were readily detected in colon adenocarcinoma biopsies ( Fig. 3A and D) [23,24]. No increase in immunoreactivity was seen in the brains of Tg(PG14) and Tg(CJD) mice compared to age-matched Non-Tg and Tg(WT) controls, not even in the regions undergoing pathological changes ( Fig. 3C and F).
The fact that we detected no proximal UPR signaling in the brain might be due to transient and/or not synchronized UPR activation. Therefore we analyzed UPR markers in primary cultures of CGN from Tg(CJD) mice, which show dramatic ER swelling associated with retention of mutant PrP [10]. No XBP-1 splicing was observed in CGN from the mutant mice, unless cells were treated with tunicamycin (Fig. 1D). There was also no difference in Grp78/BiP protein levels between control and mutant CGN ( Fig. 2E and F).

ER Stress in PC12 Cells Expressing Mutant PrP Under the Control of an Inducible Promoter and in Transiently Transfected HEK-293 cells
To exclude that cells unresponsive/resistant to the effect of the mutant protein were selected in culture, we developed lines of transfected cells in which PrP expression was activated after clonal selection. We used pheochromocytoma PC12 cells, which respond to nerve growth factor (NGF) by stopping division and developing the morphological, electrophysiological and neurochemical properties of neurons [25]. PC12 cells constitutively expressing the reverse tetracycline-controlled transactivator (PC12 Tet-on) [26] were transfected with an expression vector in which the PrP cDNA was under the control of the tetracycline-responsive element (TRE). To prevent unregulated PrP expression in the absence of inducer doxycycline (dox), cells were co-transfected with a plasmid encoding the tet-control transcriptional silencer which blocks transcription of genes under control of the TRE in the absence of dox. We generated PC12 Tet-on cell lines expressing PG14 and D177N/V128, as well as D177N/M128 PrP whose human homolog is associated with FFI. As controls, we produced lines expressing mouse wild-type PrP (M128), and M128V PrP to model the non-pathogenic V129 polymorphic variant of human PrP.
PC12 Tet-on cells had tightly regulated expression of PrP, which was undetectable in the absence of dox, and which increased progressively with increasing dox concentrations (Fig. 4A). PrP expression was detected after 8 h of induction, and lasted at least 96 h (Fig. 4B). Most PC12 clones displayed maximal transgenic PrP expression (approximately double endogenous PrP) in response to 1-2 mg/ml dox (not shown).
Mutant PrPs were insoluble in non-denaturing detergents ( Fig. 5A; percentages of PrP in the pellet were: 68612 for PG14; 4468.2 for D177N/M128; 47610.5 for D177N/V128; n = 5), and were weakly protease-resistant, in contrast to the solubility and protease-sensitivity of WT and M128V PrPs (Fig 5A  and B, and data not shown). Consistent with previous observations [27], neuronal differentiation of PC12 cells on exposure to NGF did not modify the proportion of detergent-insoluble PrP or its level of protease resistance (not shown). The mutant PrPs expressed in PC12 Tet-on cells were found at lower levels on the cell surface, and had an intense intracellular distribution ( Fig. 6A and B), which colocalized in part with the ER marker calnexin (Fig. 6C). Thus PG14 and D177N PrPs misfolded and were partially retained in the ER of PC12 Tet-on cells, confirming similar observations in primary neurons [10,12], and validating this cell model for investigating the effect of acute mutant PrP expression on ER homeostasis.
The levels of Grp78/BiP and CHOP/GADD153 mRNAs were analyzed in PC12 Tet-on cells before and after induction with 1 mg/ ml dox for 24 h. PrP expression did not increase the amount of these transcripts ( Fig. 7A and B). There was also no difference in Grp78/ BiP protein levels between control and mutant cells (not shown). IRE1a-dependent splicing of XBP1 mRNA was then assessed in cells treated with tunicamycin and/or dox. The spliced form of XBP1 was readily detectable after tunicamycin, but not in cells exposed only to dox (Fig. 7C). The same analysis on cells treated with dox for 96 h showed no evidence of ER stress (not shown).
It was possible that only a limited number of cells accumulated sufficient amounts of mutant PrP in the ER to activate the ER stress response. If this were the case, analysis of total RNA extracted from whole cultures could contain low levels of Grp78/ BiP, CHOP/GADD153 and XBP1 transcripts. To directly test whether UPR markers were activated in cells accumulating high PrP levels, mutant PrP-transfected cells were treated with dox or tunicamycin then immunostained with anti-PrP and anti-CHOP/ GADD153 antibodies. There was no activation of CHOP/ GADD153 in cells accumulating high intracellular PrP levels (Fig. 7D, panels a and b); as expected, cells treated with Finally, we analyzed ER stress markers in a different cell model of acute mutant PrP expression. HEK-293 cells were transfected with an empty pcDNA3.1 plasmid or with plasmids encoding WT, PG14, D177N/M128 or D177N/V128 PrP. As positive controls cells were transfected with a plasmid encoding murine Ig-m-chains, which in the absence of Ig-L chains accumulate in the ER inducing a robust UPR ( [28,29] and T. Anelli, R. Ronzoni and R.S., unpublished results), or were treated with tunicamycin. Cells were lysed 24 h after transfection to monitor PrP and m-chain expression and XBP1 mRNA splicing. PrP and m-chains were efficiently expressed in HEK-293 cells ( Fig. S1A and B). XBP1 mRNA splicing was readily detected in m-chain expressing cells or after treatment with tunicamycin, but not in mutant PrP-expressing cells (Fig. S1C). Thus, although HEK-293 cells respond normally to misfolded protein accumulation in the ER by up-regulating the UPR system, mutant PrPs did not evoke this response.

Mutant PrP Mice or Neurons Accumulating Cytosolic PrP Show no Functional Impairment of Ubiquitin-dependent Proteasomal Degradation
It has been suggested that misfolded PrP might be recognized as abnormal by the ER quality control machinery and retrogradely translocated into the cytosol for proteasomal degradation (ERassociated degradation, ERAD, pathway) [30,31]. In Tg(PG14) mice, mutant PrP accumulates continuously as the animals age and granule neurons in the cerebellum die [9]. To test whether accumulation of mutant PrP in Tg(PG14) mice was associated with reduced proteasome function, we generated double-transgenic mice co-expressing PG14 PrP and a green fluorescent protein (GFP) reporter substrate of the UPS (Ub G76V -GFP) [32]. The Ub G76V -GFP transgene is under the transcriptional control of the b-actin promoter, which drives constitutive expression in all organs. Under physiological conditions the N-terminal ubiquitin moiety of Ub G76V -GFP is recognized as a degradation signal and poly-ubiquitinated, leading to constitutive proteosomal degradation of GFP. An alteration of the UPS therefore results in accumulation of Ub G76V -GFP. Expression of this reporter in vivo has been used to assess the UPS function in prion-infected mice [21], and in transgenic mouse models of polyglutamine disease and amyotrophic lateral sclerosis [33,34,35].
Next we assessed UPS activity in cultured transgenic neurons expressing D177N PrP. Primary CGN from Tg(CJD) mice and nontransgenic littermates were incubated with an internally quenched fluorogenic peptide, carrying a proteasome-specific cleavage motif fused to the TAT sequence and linked to the fluorophores EDANS and DABCYL (TED) [36]. The TED peptide penetrates cell membranes and is rapidly cleaved by the proteasomal chymotrypsin-like activity. Proteasomal cleavage leads to physical separation of the EDANS/DABCYL pair, abolishing intramolecular quenching, and EDANS fluorescence increases proportionally to the amount of substrate cleaved. Thus, proteasomal efficiency can be directly assessed by counting fluorescent cells by optical microscopy. The percentages of fluorescent neurons were comparable in Tg(CJD) and nontrans-genic cultures, indicating that basal proteasome activity was not altered in cells expressing mutant PrP (Fig. 10A). The proteasome inhibitor epoxomycin had similar potency in Tg(CJD) and nontransgenic neurons (Fig. 10A).
ER stress activates a ''pre-emptive'' quality control system that aborts PrP translocation, allowing its proteasome-mediated degradation in the cytosol [19,20]. Under chronic ER stress the proteasome may be overwhelmed [37], resulting in accumulation and aggregation of PrP in the cytosol [38]. Cytosolic PrP oligomers may inhibit proteasome activity [21], triggering a feedback loop leading to further PrP accumulation in the cytosol and cell death [39,40]. We used two different approaches to test whether accumulation of untranslocated PrP aggregates was associated with low proteasome function. We treated hippocampal neurons from Ub G76V -GFP mice with tunicamycin to induce ER stress and untranslocated PrP (Fig. 10B, lane 4) [19], and monitored Ub G76V -GFP accumulation by live cell imaging and immunostaining with an anti-GFP antibody. Ub G76V -GFP was readily detectable in cells treated with the proteasome inhibitor MG132, but not with tunicamycin (Fig. 10C), indicating that cytosolic accumulation of untranslocated PrP did not inhibit proteasomal degradation of Ub G76V -GFP. Next, we assessed proteasome activity in cells treated with a cyclopeptolide (C-741) that inhibits PrP translocation [41,42], producing a biologically active form of cytosolic PrP [42]. Hippocampal neurons from C57BL/6 mice were incubated with or without C-741 for 24 h. MG132 was added during the last 6 h of treatment to induce robust accumulation of untranslocated PrP in the cytosol of C-741treated cells [42], and proteasome activity was measured using the TED peptide. We found no significant differences in proteasome function between cells treated with MG132 alone or together with C-741 (Fig. 10D), although the latter accumulated cytosolic PrP (Fig.10E, lane 6).

Discussion
Secretory and membrane proteins are subject to a stringent surveillance mechanism that causes retention of misfolded forms in the ER preventing them from reaching the cell surface. Retained proteins can stimulate ER stress response pathways, such as the UPR, which transmits information to the cytosol and nucleus to improve protein folding capacity, inhibit protein synthesis and translocation, and favor misfolded protein catabolism through ERAD. Since PrP Sc and other forms of pathogenic PrP are likely to be structurally distinct from PrP C , one might expect these molecules to be recognized as ''abnormal'' by the ER quality control machinery and trigger ER stress. Persistent ER stress may lead to apoptosis [14], a cell death modality that is often observed in experimental and natural prion diseases.
ER stress was proposed to play a key role in prion diseases since stress-regulated ER chaperons, such as Grp78/BiP and Grp58/ Erp57, were induced in sporadic and variant CJD, and in bovine spongiform encephalopathy, and their up-regulation correlated with disease progression in prion-infected mice [38,43,44,45,46]. Treatment of neuroblastoma N2a cells with PrP Sc purified from infected hamsters induced Grp78/BiP, Grp58/Erp57 and Grp94, and activated caspase-12, a protein involved in ER stress-induced apoptosis [44]. Similar changes were observed in cortical neurons treated with synthetic PrP peptide 106-126, an experimental tool used to model PrP Sc neurotoxicity [47].
However, the cellular mechanism by which PrP Sc could cause stress in the ER is not clear. PrP Sc propagates itself by imprinting its aberrant conformation on endogenous PrP C , generating additional molecules of PrP Sc in an autocatalytic reaction  [48,49]. This conversion occurs on the cell surface or within an endocytic compartment [50,51,52], so it is unlikely to be the proximate cause of stress in the ER. Exposure of N2a cells to purified PrP Sc induced fast, sustained calcium release from ER stores [44], suggesting that perturbation of ER calcium homeostasis might be the actual trigger. Regardless of the precise mechanism by which PrP Sc stimulates ER stress, the fact that genetic ablation of XBP1 or caspase-12 does not modify the onset or progression of symptoms in prion-infected mice [53,54] indicates that this pathway is not essential for prion diseases acquired by infection.
ER stress, however, might play a pathogenic role in genetic prion diseases, in which the mutant PrP molecules misfold spontaneously in the ER lumen [55]. In the present study we measured the levels of several ER stress markers in transgenic mice and cells expressing PG14 and D177N PrPs. These mutant molecules accumulate in the neuronal secretory pathway [12], and their expression is associated with swelling of the ER cisternae ([10] and unpublished observations), a morphological alteration often associated with the UPR [56]. Unexpectedly, molecular biology, biochemical and immunohistochemical analyses of brain tissue and primary neurons from the mutant mice did not find evidence of an ER stress response. Analysis of transiently transfected HEK-293 cells or PC12 cells in which PrP expression was switched on only at the time of the experiment to rule out adaptation to mutant PrP gave similar results. Thus, chronic or acute mutant PrP expression was not associated with ER stress in transgenic models, consistent with evidence from genetic prion disease patients [57].
How might the lack of an ER stress response be explained despite demonstrable mutant PrP misfolding in this organelle? There is evidence that pathogenic mutations, including D178N, do not perturb the overall fold of PrP [58] but stabilize partially structured folding intermediates [59], which might elude the quality control system in the ER. We did in fact find that PG14 and D177N PrP were not diverted to ER-associated degradation [11,12], but eventually trafficked to post-Golgi compartments [55] from which they are probably routed to the endo-lysosomal system for degradation [60].
Although ER stress does not appear to be a causative mechanism in prion diseases, perturbation of ER homeostasis might contribute to the pathogenesis. Treatment of N2a cells with several ER stressors caused formation of a PrP C isoform that was more prone to PrP Sc -induced in vitro misfolding, suggesting that ER-damaged neurons may be more susceptible to prion replication [61]. Furthermore, ER stress could compromise UPS functionality [37] and damage cells by favoring formation of cytosolic PrP oligomers that might further inhibit proteasome function [21,38,40].
To investigate the contribution of proteasome impairment in genetic prion diseases, we measured the UPS efficiency in Tg(PG14) mice. As these mice age they progressively accumulate an oligomeric form of mutant PrP in the CNS [62], and develop massive apoptotic degeneration of cerebellar granule neurons [9]. This might reflect a gradual loss of the efficiency of the proteasomal degradation system due to build-up of toxic PrP aggregates in the cytosol. Analysis of UPS activity in vivo using the Ub G76V -GFP reporter did not find specific alterations at any stage of the Tg(PG14) disease. Neither was any loss of UPS efficiency found in neurons expressing D177N PrP, using an innovative fluorogenic peptide for monitoring proteasomal proteolytic activity in live cells. Thus neurodegeneration in the transgenic models is not associated with perturbation of the cytosolic proteasomes, in line with evidence that mutant PrP is not diverted to ERAD in neuronal cells [11,12]. Finally, there was no loss of proteasome function when PrP was forced to aggregate in the cytosol by a drug that inhibits co-translational translocation. This, and the observation that untranslocated PrP caused no toxicity to neurons [12,42], strongly argues against the idea that neuronal death in prion diseases is due to cytosolic PrP accumulation.
UPS function was reduced in prion-infected mice [21,63], but only at an advanced stage of disease and in brain regions with the most intense neuropathology, suggesting that proteasome impairment is a secondary event rather than a primary pathogenic trigger. Moreover, proteasome activity was not reduced in scrapieinfected sheep -in fact, it was higher than in uninfected animalsindicating that low UPS function is not an obligatory correlate of prion pathology [64].
The results here indicate that PG14 and D177N mutants do not activate ER stress-mediated pro-apoptotic mechanisms, nor do they inhibit the proteasomal degradation pathway. Then, how might these mutants damage neurons? Although delayed in their maturation in the ER, PG14 and D177N PrPs eventually reach the cell surface [11]. Molecules that reach the surface become aggregated and PK-resistant [55], and might damage cells by altering membrane properties or interacting abnormally with other proteins on the plasma membrane [3,65]. Alternatively, intracellular aggregation of mutant PrP might interfere with vesicular traffic impairing delivery of essential cargo molecules to synapses. We did in fact find that D177N PrP expression in N2a cells is associated with an alteration of the  GDI/Rab11 pathway governing post-Golgi vesicular traffic [66]. Finally, PG14 and D177N PrP misfolding may affect calcium homeostasis ( [67] and A. Senatore and R.C. unpublished observations) with possible consequences on intracellular signaling and neurotransmission.
In conclusion our analyses do not support a role for ER stress and/or UPS dysfunction in prion diseases. Elucidating the proximate cause of neuronal dysfunction and degeneration in these disorders remains an open challenge.

Mice
Production of transgenic mice expressing wild-type, PG14 and D177N/V128 mouse PrPs with an epitope for the monoclonal antibody 3F4 has been reported [8,10]. In this study we used transgenic mice of the Tg(WT-E1 +/+ ) line, which express ,4X the endogenous PrP level, referred to throughout the text as Tg(WT), as well as Tg(PG14-A3 +/2 ), and Tg(D177N/V128-A21 +/2 ) Figure 10. Expression of D177N PrP or accumulation of cytosolic PrP is not associated with reduced proteasome function. (A) Cultures of CGN from Tg(CJD) mice (CJD) and nontransgenic littermates (NT) were left untreated or exposed to 5 mM epoxomycin for 3 h, and incubated with 40 mM TED for 20 min to measure proteasome activity. Fluorescence-positive cells were quantified by optical microscopy. Data are the mean 6 SEM of three independent experiments. (B) Cultures of hippocampal neurons from Ub G76V -GFP2 +/2 mice were left untreated (CT) or exposed to 5 mg/ml tunicamycin (Tm) for 24 h. Cell were lysed and centrifuged at 186,000 x g for 40 min, and PrP in the supernatants (S) and pellet (P) was visualized by immunoblotting with antibody 12B2. The arrow indicates the band corresponding to untranslocated PrP. Molecular mass markers are in kilodaltons. (C) Hippocampal neurons from Ub G76V -GFP2 +/2 mice were left untreated or exposed to 20 mM MG132 for 5 h or 5 mg/ml tunicamycin (Tm) for 24 h, then fixed and reacted with an anti-GFP antibody (green), and DAPI (blue) to stain the nuclei. Scale bar = 100 mm. (D) Hippocampal neurons from C57BL/6J mice were left untreated or exposed to 10 mM C-741 for 18 h to inhibit PrP translocation; 20 mM MG132 was added during the last 6 h to induce accumulation of untranslocated PrP. Fluorescence-positive cells were quantified after 20 min incubation with TED. Data are the mean 6 SEM of three independent experiments. (E) Untransocated PrP was visualized by immunoblotting as in B. doi:10.1371/journal.pone.0019339.g010 expressing transgenic PrP at ,1X, referred to as Tg(PG14) and Tg(CJD), respectively. These mice were originally generated on a C57BL/6J X CBA/J hybrid and were then bred with the Zurich I line of Prnp 0/0 mice [68] with a pure C57BL/6J background (European Mouse Mutant Archive, Monterotondo, Rome, Italy; EM:01723). The Ub G76V -GFP1 and Ub G76V -GFP2 mouse lines expressing the transgene from a chicken b-actin promoter with a cytomegalovirus immediate early enhancer [32] were maintained by breeding with C57BL/6J mice.

Ethics Statement
All procedures involving animals were conducted according to European Union (EEC Council Directive 86/609, OJ L 358,1; December 12, 1987) and Italian (D.L. n.116, G.U. suppl. 40, February 18, 1992) laws and policies, and were in accordance with the United States Department of Agriculture Animal Welfare Act and the National Institutes of Health Policy on Humane Care and Use of Laboratory Animals. They were reviewed and approved by the Mario Negri Institute Animal Care and Use Committee that includes ad hoc members for ethical issues (18/01-D, 18/01-C). Animal facilities meet international standards and are regularly checked by a certified veterinarian who is responsible for health monitoring, animal welfare supervision, experimental protocols and review of procedures.

Neuronal Cultures
Primary neuronal cultures were prepared as previously described [69]. Briefly, cerebella were dissected, sliced into ,1mm pieces and incubated in Hank's balanced salt solution (HBSS, Gibco) containing 0.3 mg/ml trypsin (Sigma) at 37uC for 15 min. Trypsin inhibitor (Sigma) was added to a final concentration of 0.5 mg/ml and the tissue was mechanically dissociated by passing through a flame-polished Pasteur pipette. Cells were plated at 350-400,000 cells/cm 2 on poly-L-lysine (0.1 mg/ml)-coated plates. Cells were maintained in Basal Medium Eagle (Gibco) supplemented with 10% dialyzed fetal bovine serum (FBS, Sigma), penicillin/streptomycin and KCl 25 mM, at 37uC in an atmosphere of 5% CO 2 , 95% air.
Hippocampal neurons were prepared from two-day-old animals as described [69]. Brain tissue was sliced into ,1-mm pieces and incubated in HBSS (Gibco) containing 20 U/ml papain (Sigma) at 34uC for 30 min. Trypsin inhibitor (Sigma) was added to a final concentration of 0.5 mg/ml and the tissue was mechanically dissociated by passing through a flame-polished Pasteur pipette. Cells were plated at 150-250,000 cells/cm 2 on poly-D-lysinecoated (25 mg/ml) plates and maintained in Neurobasal Basal Medium (Gibco) supplemented with B27 (Gibco), penicillin/ streptomycin and glutamine 2 mM. To reduce the number of nonneuronal cells, aphidicolin (3.3 mg/ml, Sigma) was added to the medium 48 h after plating. Non-neuronal contamination was less than 3%.
Construction of cDNAs encoding wild-type (M128, WT), M128V, D177N/M128, D177N/V128 or PG14 PrP carrying an epitope tag for the monoclonal antibody 3F4 has been described [12,70]. Pst I and Not I restriction sites were inserted respectively upstream of the 59 and downstream of the 39 ends of the PrP open reading frame by PCR, using Vent polymerase (New England Biolabs) and the following primers: 59-GACCAGCTG-CAGATGGCGAACCTTGGCTACTGGCTG-39 (sense); 59-GA-CCAGGCGGCCGCTCATCCCACGATCAGGAAGATGAG-39 (antisense). The amplification products were cloned into the Pst I/ Not I restriction sites of the bidirectional Tet vector pBI-G (BD Clontech), containing the tetracycline-responsive element (TRE), and their identity was confirmed by sequencing the entire coding region.
PC12 Tet-on cells were plated on 100-mm dishes, left to reach 80% confluence and transfected with 0.8 mg of pBI-G empty or containing the PrP cDNA (WT, M128V, D177N/M128, D177N/ V128 or PG14), 8 mg of pTeT-tTS encoding for a transcriptional silencer (tTS) that blocks transcription of genes under control of the TRE in the absence of dox, and 0.1 mg of pTK-Hyg encoding for hygromycin resistance, using Lipofectamine Plus reagent (Invitrogen Inc., Carlsband, California). Three days after transfection, hygromycine (80 mg/ml, Sigma) was added to the culture medium. Resistant clones were isolated four weeks after exposure to hygromycin and tested for PrP expression before and after induction with 1 mg/ml dox for 24 h by immunoblotting with the 3F4 antibody. Only clones with undetectable basal expression of transfected PrP were selected for further experiments. The data presented here were obtained from at least two cloned lines expressing each construct.

Northern Blot
Northern blot analysis was done using the CDP-Star Chemiluminescent Detection System (Amersham Biosciences). 20 mg of total RNA was separated by electrophoresis on 1.25% agarose gel containing 6% formaldehyde and transferred onto a nylon membrane (Hybond-N+, Amersham Biosciences) by capillary elution for 18 h, then immobilized by baking the membrane for 2 h at 80uC. Mouse PrP (wild-type, 3F4 tagged), hamster CHOP, mouse Grp78/BiP and mouse GADPH (glyceraldehyde-3 phosphate-dehydrogenase) cDNAs were used as a template for the preparation of fluorescein-labeled probes using the Gene Images Random-Prime Labeling module (Amersham Biosciences). The RNA blot was rinsed in SSC 2X (sodium chloride sodium citric acid), pre-hybridized for 2 h at 60uC in hybridization buffer (SSC 5X, 5% dextrane sulphate, 0.1% sodium dodecyl sulphate (SDS), 100 mg/ml denatured salmon sperm DNA, and 10% Liquid Block provided in the Gene Images Blot Kit, Amersham Biosciences) and hybridized for 18 h at 60uC with the labeled probe. Excess probe was removed by washing once with SSC 2X, 0.1% SDS for 15 min at room temperature, once with SSC 1X, 0.1% SDS for 15 min at 65uC, and once with SSC 0.5X, 0.1% SDS for 15 min at 65uC. The membrane was pre-incubated with 5% non-fat dried milk (NFDM) in 0.1 M Tris-HCl, pH 9.5, 0.3 M NaCl (blocking buffer) for 2 h at room temperature before incubating with alkaline phosphatase-conjugated anti-fluorescein antibody (1:5,000 in blocking buffer) for 1 h at room temperature. Unbound antibody was removed by rinsing four times with 0.1 M Tris-HCl, pH 9.5, 0.3 M NaCl, 0.3% Tween-20. The membrane was incubated for 5 min with the CDP-Star substrate (Amersham Biosciences) and the chemiluminescent signal visualized with a Biorad XRS image scanner or Hyperfilm-MP films (Amersham Biosciences).

Biochemical Analysis
To test the detergent insolubility of PrP, cells were lysed in 10 mM Tris pH 7.5, 100 mM NaCl, 0.5% sodium deoxycholate and 0.5% Nonidet P-40 containing protease inhibitors (1 mg/ml pepstatin and leupeptin, 0.5 mM phenylmethylsulfonyl fluoride (PMSF), and 2 mM ethylenediaminetetraacetic acid). Lysates corresponding to 300 mg of protein were centrifuged at 186,000 x g for 40 min in a Beckman Optima Max-E ultracentrifuge. Proteins in the pellet and supernatant were separated by SDS-polyacrylamide gel electrophoresis and electro-transferred onto polyvinylidene fluoride membranes (Immobilon P, Millipore). Membranes were incubated first with 5% NFDM in 100 mM Tris pH 7.5, 150 mM NaCl and 0.1% Tween 20 (TTBS), then with anti-PrP antibodies overnight at 4uC or for 1 h at room temperature, rinsed three times with TTBS and incubated for 1 h at room temperature with horseradish peroxidase-conjugated secondary antibody (diluted 1:5,000; Santa Cruz). Signals were revealed using enhanced chemiluminescence (Amersham Biosciences), and visu-alized by a Biorad XRS image scanner. Quantitative densitometry of protein bands was done using Quantity One software (Biorad).
To test PK resistance 300 mg of protein extracts were digested with 0.25-0.5 mg/ml of PK for 30 min at 37uC. Digestion was terminated by adding PMSF to a final concentration of 5 mM. Proteins were methanol-precipitated and PrP was analyzed by Western blot using the 3F4 antibody.
Immunofluorescence PC12 cells were seeded on poly-L-lysine-coated glass coverslips in 24-well plates at 50% confluence and induced to express PrP by treating them with 1 mg/ml dox for 24 h. Alternatively, cells were differentiated with NGF before proceeding with immunofluorescence. For surface staining of PrP, living cells were incubated with anti-PrP 3F4 antibody (1:500 in OptiMEM) for 1 h at 4uC, washed with ice-cold PBS and fixed in 4% paraformaldehyde (PFA) for 1 h at 4uC. After washing with PBS, cells were incubated in blocking solution (0.5% BSA and 50 mM NH 4 Cl in PBS) for 1 h at room temperature, then with Alexa 488(green)-conjugated goat anti-mouse IgG (Molecular Probes) (1:500 in blocking solution). After 1 h incubation at room temperature cells were rinsed and mounted in FluorSave (Calbiochem Corporation, San Diego, California). To visualize intracellular PrP and other intracellular antigens, cells were washed with PBS and fixed in 4% PFA for 1 h at 4uC. Cells were then rinsed, permeabilized with 0.1% Triton X-100 for 2 min and incubated for 30 min at room temperature in blocking solution, then with the primary antibody (in blocking solution) for 1 h. Cells were rinsed in 0.1% Tween-20/PBS and incubated with Alexa 488(green)-or 546(red)conjugated secondary antibody before mounting. For co-localisation experiments, anti-PrP antibody was incubated, followed by the corresponding secondary antibody, before proceeding with organelle markers staining. Cells were analyzed with an Olympus FV500 laser confocal scanning system.

Immunohistochemistry
For CHOP/GADD153 and Grp78/Bip immunostaining, sections from paraffin-embedded, 4% PFA-fixed tissue were collected on collagen-coated slides and stained using an automated immunostaining system (Leica). Endogenous peroxidase activity was blocked with 0.3% H 2 O 2 for 20 min at room temperature. Primary antibodies were applied to slides for 30 min at room temperature, then incubated with horseradish peroxidase-labeled anti-rabbit or anti-goat Envision polymers (Dako) for 30 minutes. The reaction product was developed using 3,3-diaminobenzidine.
For GFP immunostaining, anesthetized mice were intracardially perfused with 4% PFA; their brains were removed, post-fixed and frozen at -80uC after cryoprotection. Brain sections were cut using a Leica cryostat and incubated for 1 h at RT with 10% normal goat serum (NGS), 0.1% Triton X-100 in phosphate-buffered saline (PBS) then overnight at 4uC with the anti-GFP antibody. After incubation with an anti-rabbit IgG biotinylated antibody (Vector Laboratories) immunostaining was developed using the avidin-biotin-peroxidase complex (Vector Laboratories) and diaminobenzidine (Sigma). Images were acquired using an Olympus image analyzer and the Cell-R software.

Proteasome activity in single cells
Proteasome activity was assessed in living neuronal cells using the TED peptide, as described [36]. Briefly, neurons grown in 15mm slide 8-well chamber dishes (Ibidi) were incubated with 40 mM TED in Krebs Ringer Hepes (KRH) buffer (125 mM NaCl, 5 mM KCl, 1.2 mM MgSO 4 , 2 mM CaCl 2 , 10 mM glucose, 25 mM Hepes, pH 7.4) for 20 min. The TED solution was replaced with fresh culture medium, and images were collected and analyzed with a Cell R imaging station (Olympus) coupled to an inverted microscope (IX 81, Olympus) equipped with an incubator to maintain constant temperature (37uC) and CO 2 (5%). The EDANS fluorescent signal was acquired with a high-resolution camera (ORCA) equipped with a 340-nm excitation filter (D340xv2 Chroma), a 400-nm dichroic mirror (400DCLP, Chroma) and an emission filter with a range of 510640 nm (D510/40 m, Chroma). Images were acquired after 20 min of incubation, and ten frames were randomly sampled for a minimum of 1000 cells. EDANS-positive cells, identified after removing background, were counted and compared to the total cell number counted by bright-field using the ImageJ 1.43 software (http://rsbweb.nih.gov/ij/). cells, and XBP1 splicing was analyzed by RT-PCR. Each product of amplification was separated on 2.5% agarose gel. Spliced forms were detected only in cells expressing the m-chain or treated with tunicamycin (Tm). Size markers are given in base-pairs. (TIF)