Journal of Alzheimer’s Disease 27 (2011) 61–73DOI 10.3233/JAD-2011-100395IOS Press

61

Amyloid-� Peptide-Induced Secretionof Endoplasmic Reticulum ChaperoneGlycoprotein GRP94

Ricardo J.S. Vianaa, Clifford J. Steerb,c and Cecılia M.P. Rodriguesa,d,∗aResearch Institute for Medicines and Pharmaceutical Sciences, Faculty of Pharmacy,University of Lisbon, Lisbon, PortugalbDepartment of Medicine, University of Minnesota Medical School, Minneapolis, MN, USAcDepartment of Genetics, Cell Biology, and Development, University of Minnesota Medical School,Minneapolis, MN, USAdDepartment of Biochemistry and Human Biology, Faculty of Pharmacy, University of Lisbon, Lisbon, Portugal

Accepted 20 May 2011

Abstract. Amyloid-� (A�) peptide-induced neurotoxicity is typically associated with cell death through mechanisms not entirelyunderstood. Here, we investigated stress signaling events triggered by soluble A� in differentiated rat neuronal-like PC12 cells.Morphologic evaluation of apoptosis confirmed that A� induced nuclear fragmentation that was prevented by pre-treatmentwith the antiapoptotic bile acid tauroursodeoxycholic acid (TUDCA). In addition, A� exposure triggered an early signalingresponse by the endoplasmic reticulum (ER) and caspase-12-mediated apoptosis, which, however, was independent of the ER-stress pathway. Furthermore, ER stress markers, including GRP94, ATF-6�, CHOP, and eIF2�, were strongly downregulated byA�, independent of protein degradation, and partially restored by TUDCA. Calpain inhibition prevented caspase-12 activationand reduced levels of ATF-6�. Importantly, A�-induced GRP94 downregulation was related to protein secretion and partiallyrescued through inhibition of the secretory pathway by geldanamycin and brefeldin. In conclusion, we showed that the ER isa proximal stress sensor for soluble A�-induced toxicity, resulting in caspase-12 activation and cell death in PC12 neuronalcells. Moreover, ER chaperone GRP94 secretion was associated with A�-induced apoptotic signaling. These data provide newinformation linking apoptotic properties of A� peptide to distinct subcellular mechanisms of toxicity. Further characterizationof this signaling pathway is likely to provide new perspectives for modulation of amyloid-induced apoptosis.

Keywords: Amyloid-�, ATF-6�, caspase-12, endoplasmic reticulum, GRP94, tauroursodeoxycholic acid

Supplementary data available online: http://www.j-alz.com/issues/27/vol27-1.html#supplementarydata03

INTRODUCTION

Alzheimer’s disease (AD) is associated with theaccumulation of pathogenic amyloid-� (A�) assem-blies in brain resulting in progressive dismantling ofsynapses and loss of neurons in selected areas. It isrecognized that increased production, oligomerization,

∗Correspondence to: Cecılia M.P. Rodrigues, iMed.UL, Facultyof Pharmacy, University of Lisbon, Lisbon 1649-003, Portugal.Tel.: +351 21 794 6490; Fax: + 351 21 794 6491; E-mail:cmprodrigues@ff.ul.pt.

and aggregation of A� peptides are crucial factorsin the onset of AD. Toxic A�40 and A�42 are pro-cessed from the amyloid-� protein precursor (A�PP)via cleavage by �- and �-secretase complexes [1, 2].A�PP is a transmembrane protein, which is foldedand modified in endoplasmic reticulum (ER), and thentransported through the Golgi complex to the outermembrane. Currently, it is not clear which cellularcompartment processes A�PP to toxic A� peptides [1].

ER serves two major functions in the cell, includingproper folding of newly synthesized proteins destined

ISSN 1387-2877/11/$27.50 © 2011 – IOS Press and the authors. All rights reserved

62 R.J.S. Viana et al. / GRP94 Secretion by Aβ

for secretion, cell surface or intracellular organelles,and as a reservoir of Ca2+. The ER is highly sensitiveto alterations in Ca2+ homeostasis and perturbationsin its environment that result in ER stress. The ERresponds by triggering specific signaling pathwaysincluding the unfolded protein response (UPR), theER-overload response (EOR), and the ER-associateddegradation (ERAD). These pathways are collectivelyknown as the ER stress response, which is also asso-ciated with increased expression of genes that expandthe folding capacity of the ER, preventing new proteinsynthesis, and accelerating degradation of misfoldedproteins. The UPR is characterized by the coordi-nated activation of multiple proteins that can be divid-ed into three groups, consisting of ER stress-inducedcell death sensors, modulators, and effectors. The ERmembrane residing proteins inositol-requiring enzyme1 (IRE1), activating transcription factor 6 (ATF-6), anddouble-stranded RNA-activated protein kinase like ERkinase (PERK) serve as proximal ER stress sensors,which are maintained in an inactive state by bindingto the chaperone 78-kDa glucose-regulated protein,also known as immunoglobulin heavy chain bindingprotein or Bip (GRP78/Bip). GRP78/Bip is a distalsensor and functions as a master regulator of the UPR.During ER stress, the release of GRP78/Bip resultsin inhibition of translation (PERK), and upregulationof molecular chaperones (ATF-6) and ERAD activity(IRE1) [3, 4]. ER stress-induced cell death modu-lators include members of the Bcl-2 family (Bcl-2,Bcl-xL, Bak and Bik) [5, 6] and CHOP [7], amongothers.

Finally, several molecules that are either present onthe ER surface or in the soluble compartment have beenimplicated as effectors that cause death in response toprolonged ER stress. Caspase-12 is an initiator caspasewhich is specifically involved in apoptosis [8]; and itsactivation may occur through ER stress-induced cal-pain or caspase-7 cleavage of procaspase-12 [8, 9]. ERstress-activated IRE1 may also aggregate procaspase-12 at the ER membrane surface through the cytosolicadaptor tumor necrosis factor associated receptor fac-tor 2 (TRAF2) proteins, resulting in cleavage andactivation of caspase-12 [10]. Most notably, caspase-12, together with caspase-9, serves as a mediator ofan intrinsic apoptosis pathway that is independent ofmitochondria [11].

The ER is a repository of antiapoptotic GRP94, amember of the heat shock protein 90 (hsp90) family[12–15]. GRP94 is a hydrophobic, ATP-, peptide- [14,16], and calcium-binding protein that contains severalhelix-loop-helix structural domains [17]. Its antiapop-

totic properties [18–20] are not fully understood. It hasalso been described to act as a vaccine in prophylac-tic and therapeutic models of cancer immunotherapy[21, 22]. More recently, GRP94 was shown to bereleased from cells undergoing necrotic cell death asa consequence of lethal viral infection, suggesting apotential physiological role for chaperone proteins inthe regulation of immunological responses to patho-logical cell death [23, 24]. Inhibitors of hsp90 andGRP94, such as geldanamycin (GA) have been inves-tigated in the treatment of cancer. GA inhibits hsp90and GRP94 by binding competitively to its N-terminalATP-binding site, which results in misfolding anddegradation of hsp90 substrates through the ubiquitin-proteasome pathway. While GA-mediated enhanceddegradation of certain hsp90 proteins is now wellestablished (reviewed in [25, 26]), few studies havereported GA-mediated inhibition of intracellular traf-ficking [27].

In this study, we show that A�-induced toxic-ity activates caspase-12, which is known to mediateER-specific apoptosis in PC12 neuronal cells. Unex-pectedly, we also demonstrated that ER stress-inducedcell death sensors and modulators are strongly down-regulated. ATF-6� is proteolytically cleaved by calpainI, and GRP94 is secreted to the extracellular medium,in association with A�-induced toxicity.

MATERIALS AND METHODS

Cells and culture conditions

PC12 cells were grown in RPMI-1640 medium(Sigma-Aldrich, St. Louis, MO, USA) supplementedwith 10% heat-inactivated horse serum (Sigma-Aldrich), 5% fetal bovine serum (Invitrogen Corp.,Carlsband, CA, USA), and 1% penicillin/streptomycin(Invitrogen), and maintained at 37◦C in a humidi-fied atmosphere of 5% CO2. Cells were plated anddifferentiated in the presence of nerve growth factor(NGF) (Promega Corp., Madison, WI), as previouslydescribed [28]. Cell density was maintained at 1 × 105

cells/cm2 for morphological assessment of apoptosisand viability assays, or 4 × 105 cells/cm2 for transfec-tion assays and protein extraction.

To induce apoptosis, cells were exposed to 25 �Mof soluble A�25-35 peptide active fragment (BachemAG, Bubendorf, Switzerland) for 1, 2, 8, and 24 h. Thisconcentration was found to be ideal in maximizingthe cellular response of A� with minimum exacer-bated toxic effects in neuronal-like PC12 cells (data not

R.J.S. Viana et al. / GRP94 Secretion by Aβ 63

shown). Brefeldin and thapsigargin (Sigma-Aldrich)were used at 18 and 3 �M, respectively, as positivecontrols for ER-specific apoptosis. To test the effect ofantiapoptotic TUDCA, cells were incubated in mediumsupplemented with 100 �M TUDCA (Sigma-Aldrich),or no addition (control), for 12 h, and then exposed to25 �M of A�25-35 for 1, 2, 8, and 24 h. Cells were fixedfor microscopic assessment of apoptosis or processedfor cell viability assays. In addition, total proteins wereextracted for immunoblotting.

In parallel experiments, cells were pretreated for30 min before A� incubation with SP600125, thec-Jun N-terminal kinase (JNK) inhibitor (Calbiochem,Darmstadt, Germany), at either 10 �M for 24 h, or50 �M SP600125 for 3 h. Similarly, cells were pre-treated with 10 �M calpain inhibitor I (Roche AppliedScience, Mannheim, Germany) for 24 h or 50 �Mcalpain inhibitor I for 3 h. Proteasome and caspase inhi-bition during apoptosis was achieved by pretreatingcells with either 10 �M MG-132 and z-VAD.fmk (Cal-biochem) for 24 h, or 50 �M MG-132 and z-VAD.fmkfor 3 h. To determine the effect of Ca2+ deregula-tion, cells were pretreated with 50 �M ZnCl2, 100 nMxestospongin C, 20 �M nifedipine, 10 �M dantrolene,and 0.05 to 5 �M cyclosporin A (Sigma-Aldrich) for3 h. Finally, cells were pretreated with 0.05 to 5 �M GAand 9 �M brefeldin (Sigma-Aldrich) for 3 h to inhibitprotein secretion.

Measurement of cell death

Cell viability was measured by trypan blue exclu-sion; and apoptotic nuclei were detected by Hoechstlabeling after cell fixation with 4% formaldehydein phosphate buffer saline (PBS), for 10 min atroom temperature. Following incubation with Hoechstdye 33258 (Sigma-Aldrich) at 5 �g/mL in PBS for5 min, and PBS washes, slides were mounted withPBS:glycerol (3:1, v/v) and fluorescence visualizedwith an Axioskop fluorescence microscope (Carl ZeissGmbH, Hamburg, Germany). Fluorescent nuclei werescored and categorized according to the condensa-tion and staining characteristics of chromatin. Normalnuclei showed non-condensed chromatin dispersedover the entire nucleus. Apoptotic nuclei were inde-pendently identified by condensed and fragmentedchromatin contiguous to the nuclear membrane, as wellas nuclear fragmentation and presence of apoptoticbodies. Three random microscopic fields per sample of∼250 nuclei were counted and mean values expressedas the percentage of apoptotic nuclei.

Immunoblotting

Steady-state levels of caspase-12, ATF-6�, KDEL,CHOP, and eIF2� were determined by Western blotanalysis. Briefly, 80 �g of total and nuclear proteinswere separated by 12% sodium dodecyl-polyacry-lamide gel electrophoresis. After electrophoretic trans-fer onto nitrocellulose membranes, immunoblots wereincubated with 15% H2O2 for 15 min at room tem-perature. Following a block with 5% milk solution,membranes were incubated overnight at 4◦C with pri-mary mouse monoclonal antibody reactive to KDEL(10C3) (Abcam, Cambridge Science Park, Cambridge,UK), primary rabbit polyclonal antibodies reactiveto ATF-6� (H-280), GADD153 (R-20) (Santa CruzBiotechnology, Santa Cruz, CA, USA), and Anti-eIF2� [pS52] (Invitrogen). Finally, secondary goatanti-mouse or anti-rabbit IgG antibody conjugatedwith horseradish peroxidase (Bio-Rad Laboratories,Hercules, CA, USA) was added for 3 h at roomtemperature. The membranes were processed for pro-tein detection using the SuperSignal substrate (PierceBiotechnology, Rockford, IL, USA). �-Actin (AC-15)(Sigma-Aldrich) was used as loading control fortotal proteins. Protein concentrations were determinedusing the Bio-Rad protein assay kit (Bio-Rad) accord-ing to the manufacturer’s specifications.

RNA isolation and RT-PCR

Transcript expression of GRP94 was determinedby RT-PCR. Total RNA was extracted from rat PC12cells using the TRIZOL reagent (Invitrogen). ForRT-PCR, 5 �g of total RNA was reverse transcribedusing oligo(dT) (Integrated DNA Technologies Inc.,Coralville, IA) and SuperScript II reverse transcrip-tase (Invitrogen). Specific oligonucleotide primer pairswere incubated with cDNA template for PCR ampli-fication using the Expand High FidelityPLUS PCRSystem (Roche Applied Science, Mannheim, Ger-many). The following sequences were used as primers:GRP94 sense 5′-TACTATGCCAGTCAGAAGAAAACG-3′; GRP94 antisense 5′-CATCCTTTCTATCCTGTCTCCATA-3′; �-actin sense 5′-CGTCTTCCCCTCCATCGTG-3′; and �-actin antisense 5′-CCAGTTGGTTACAATGCCGTG-3′. The product of the�-actin RNA was used as control.

Heat shock protein GRP94 ELISA assay

The rat hsp90 glycoprotein 96 (hsp90 gp96) ELISAKit (Cusabio Biotech Co., Newark, USA) was used

64 R.J.S. Viana et al. / GRP94 Secretion by Aβ

to measure GRP94 levels, according to the manu-facturer’s protocol. Supernatant samples were appliedto plates containing biotinylated antibody for GRP94(1:100). After 1 h incubation at 37◦C, plates werewashed and incubated with HRP-avidin (1:100) for anadditional 1 h, after which the chemiluminescence sub-strate solution was added and incubated for 10–30 min.The reaction was stopped and the optical density wasmeasured after 30 min using a microplate reader set to450 nm.

Statistical analysis

The relative intensities of protein bands wereanalyzed using the Quantity One Version 4.6 densit-ometric analysis program (Bio-Rad) and normalizedto the respective loading controls. Statistical analysiswas performed using GraphPad InStat Version 3.00(GraphPad Software, San Diego, CA) for the analy-sis of variance and Bonferroni’s multiple comparisontests. Values of p < 0.05 were considered significant.

RESULTS

Aβ-induced apoptosis is mediated by caspase-12activation and modulated by TUDCA in PC12neuronal-like cells

Soluble A�25-35 fragment induces apoptosis inPC12 neuronal-like cells, which is abrogated byTUDCA pre-treatment [29, 30]. ER is a potentialintracellular target of A� protein [31], and TUDCAexerts protection from apoptosis, in part, by prevent-ing caspase-12 activation [32]. In this study, neuronalPC12 cells pre-exposed to NGF for 8 days weretreated with either A�25-35, or ER-specific apoptoticstimuli brefeldin and thapsigargin, with or withoutTUDCA, and assessed for cell death. We confirmedthat A�25-35-induced apoptosis after 24 h was reducedto almost control levels in cells pre-treated withTUDCA (p < 0.05) (Fig. 1A). The reverse sequenceA�35−25 did not induce significant apoptosis (data notshown). General cell death as assessed by trypan bluedye exclusion was in accordance with the morpho-logical data (Fig. 1B). TUDCA was also effective atmodulating brefeldin- and thapsigargin-induced apop-tosis (Fig. 1A). Similar results were obtained in PC12neuronal-like cells exposed to 10 �M soluble A�1-40 or20 �M fibrillar A�1-42 and pre-treated with TUDCA.Further, in primary rat cortical neurons, exposure toeither 25 �M soluble A�25-35, 10 �M soluble A�1-40,or 20 �M fibrillar A�1-42 for 24 h resulted in 30–50%

apoptotic cells, and this was significantly reduced withTUDCA by at ∼50% (Supplementary data; avail-able online: http://www.j-alz.com/issues/27/vol27-1.html#supplementarydata03).

After A�25-35 exposure in PC12 neuronal-likecells, caspase-12 processing peaked at 24 h, and pre-treatment with TUDCA markedly reduced A�-inducedeffects by ∼70% (p < 0.05) (Fig. 2). The activation ofcaspase-12 by calpains in apoptosis has been previ-ously documented [9]. Moreover, the JNK pathway isactivated in response to ER stress by IRE1 [33, 34],which may activate caspase-12 through TRAF2, thesame cytosolic adaptor of JNK at the ER membranesurface [10]. To test if calpains or JNK plays a role inmediating neuronal cell death induced by A�, we incu-bated PC12 cells in the presence or absence of calpaininhibitor I or SP600125 JNK inhibitor for 24 h. After24 h of incubation with A�, calpain inhibition, but notJNK inhibition, prevented caspase-12 fragmentationby almost 50% (p < 0.05), suggesting that calpains areinvolved in A�-induced caspase-12 activation.

Aβ-induced, caspase-12-mediated cell death isindependent of ER stress

Caspase-12 can be activated under specific ERstress stimuli [31]. In addition, A� causes ER stress[35] and proteosomal dysfunction [36]. To determinewhether ER stress plays a role in A�-induced apopto-sis mediated by caspase-12 in PC12 neuronal cells, weevaluated several ER stress markers by immunoblot-ting (Fig. 3). ER stress makers were upregulated oractivated after either brefeldin or thapsigargin incuba-tion, but not after A� exposure for 24 h. Interestingly,ATF-6�, eIF2�, CHOP, and GRP94 were markedlydownregulated after A� incubation. This effect waspartially inhibited by TUDCA (Fig. 3). Thus, ER isstrongly involved in A�-induced apoptosis, althoughwith no evidence of ER stress activation.

ATF-6α and GRP94 downregulation isindependent of protein degradation

Cells containing a mutated presenilin-1 (PS1) geneshowed significantly inhibited induction of GRP78mRNA expression when stimulated for ER stress [37].More importantly, GRP78 and GRP94 are decreasedin brains of sporadic AD patients [37]. The reductionof GRP94 protein levels may be explained by selec-tive proteolytic cleavage by calpains activated duringetoposide-induced apoptosis [38]. In this study, basallevels of GRP94 and ATF-6α mRNA expression were

R.J.S. Viana et al. / GRP94 Secretion by Aβ 65

Fig. 1. TUDCA inhibits cell death induced by A�, brefeldin, and thapsigargin in PC12 neuronal cells. Cells were incubated with either 25 �Msoluble A�25-35, 9 and 18 �M brefeldin, 1.5 and 3 �M thapsigargin, or no addition (control), ± 100 �M TUDCA for 24 h. In co-incubationexperiments, TUDCA was added 12 h prior to incubation with A�, brefeldin, or thapsigargin. A) Total and trypan blue stained cells werecounted and general cell death calculated as the number of death cells per total number of cells. B) Cells were fixed and stained with Hoechstfor microscopic assessment of apoptosis, and the percentage of apoptotic cells was determined. Cells exposed to A�, brefeldin or thapsigarginalone showed more nuclear fragmentation compared with cells pre-treated with TUDCA. The results are expressed as mean ± SEM for at leastthree different experiments. *p < 0.01 from control; †p < 0.05 from A�, brefeldin or thapsigargin alone.

not altered in PC12 cells when exposed to A� for 2and 24 h (data not shown). We tested if GRP94 andATF-6� cleavage was mediated by proteases, such ascaspases and calpains, or by the proteasome, sincecalpain inhibitors might also interfere with the protea-some activity. Our results showed that general caspaseinhibitor z-VAD-fmk and JNK inhibitor SP600125 didnot prevent cleavage of GRP94 and ATF-6�, after 24 hof A� incubation (Fig. 4A). Nevertheless, both calpainand proteasome inhibition were partially protective(Fig. 4A). Proteasome activity after A� incubation was

assessed through a fluorescence assay by monitoringthe release of free 7-amino-4-methylcoumarin from thefluorogenic peptide chymotrypsin substrate III Suc-Leu-Leu-Val-Tyr-methylcoumarin. The proteasomalinhibitor, MG-132 was used as a positive control. Wefound that chymotrypsin-like activity, the predominantproteosome activity, remained unchanged after A�incubation (data not shown). In addition, autophagywas not activated by A� under the conditions testedas demonstrated by the absence of processed LC3, aspecific marker for autophagosomes (data not shown).

66 R.J.S. Viana et al. / GRP94 Secretion by Aβ

Fig. 2. TUDCA modulates A�-induced caspase-12 processing inPC12 neuronal cells. Cells were incubated with 25 �M A�, orno addition (control), ± 100 �M TUDCA, 10 �M SP600125 and10 �M calpain inhibitor I for 24 h. In co-incubation experiments,TUDCA was added 12 h prior to incubation with A�. SP600125 andcalpain inhibitor I were added 30 min prior to incubation with A�.Total proteins were extracted for western blot analysis as described inMaterials and Methods. Representative immunoblot of procaspase-12 and active caspase-12 in cells exposed to A�, ± TUDCA,SP600125 or calpain inhibitor I (top), and active caspase-12 levels(bottom). The results are normalized to endogenous �-actin proteinlevels and expressed as mean ± SEM arbitrary units for at least fourdifferent experiments. *p < 0.05 from control; †p < 0.05 from A�alone.

Thus, typical mechanisms for protein degradation,such as the proteasome and autophagy do not appear tobe triggered by A� incubation in PC12 neuronal-likecells.

Since caspase-12 activation coincided with the earlyreduction of GRP94 at 2 h of A� incubation, wetested if GRP94 and ATF-6� cleavage was selec-tively mediated by calpains or the proteasome, andnot a consequence of general protein degradation.Cells were pretreated with 50 �M of each inhibitor for30 min [38], followed by 2 h of A� incubation. In con-trast to MG-132, pretreatment with calpain inhibitorI prevented ATF-6� cleavage (Fig. 4B). However,GRP94 downregulation was not blocked by any ofthe inhibitors. These results suggested that ATF-6�is selectively cleaved by calpains during A�-inducedapoptosis, whereas GRP94 is not cleaved by proteasesduring apoptosis.

GRP94 downregulation is independent of Ca2+

deregulation

GRP94 has been linked directly to cellular Ca2+homeostasis [39]. GRP94 binds Ca2+ and is one of

Fig. 3. A�-induced apoptosis in PC12 neuronal cells is independentof ER stress. Cells were incubated with 25 �M A�, or no addi-tion (control), ± 100 �M TUDCA for 2 h. Positive controls for ERstress were brefeldin 18 �M and thapsigargin 3 �M for 2 h. In co-incubation experiments, TUDCA was added 12 h prior to incubationwith A�, brefeldin, or thapsigargin. Total proteins were extracted forwestern blot analysis as described in Materials and Methods. Repre-sentative immunoblots of ER stress markers ATF-6�, eIF2�, CHOP,GRP94, and GRP78/Bip (top); and ATF-6� (middle) and GRP94(bottom) levels. The results are normalized to endogenous �-actinprotein levels and expressed as mean ± SEM arbitrary units for atleast three different experiments. *p < 0.01 from control; †p < 0.05from A� or thapsigargin alone.

approximately six luminal proteins that serve as themajor buffers of the ER [40–42]. A key mechanismby which A� is believed to alter cellular Ca2+ home-ostasis involves disruption of plasma membrane Ca2+permeability. At the extracellular level, A� exertsactions on endogenous plasmalemmal ion channels,such as L-type channels [43] and pore formation [44].Furthermore, A� oligomers induce massive calciumentry into mitochondria, which opens the mitochon-drial permeability transition pore (mPTP) [45], andA� targets the mPTP resulting in damage amplifica-tion [46]. We used Ca2+ blockers to determine whetherGRP94 reduction was due to unregulated flux of extra-

R.J.S. Viana et al. / GRP94 Secretion by Aβ 67

Fig. 4. A�-induced downregulation of ATF-6� and GRP94 in PC12 neuronal cells is independent of degradation mechanisms. Cells wereincubated with 25 �M A�, or no addition (control), ± 10 �M of SP600125, calpain inhibitor I, MG132, and z-VAD.fmk for 24 h. For shorterincubations with A�, 50 �M of each inhibitor was used for 2 h. In co-incubation experiments, the inhibitors were added 30 min prior to incubationwith A�. Total proteins were extracted for western blot analysis as described in Materials and Methods. A) Representative immunoblot of ATF-6�and GRP94 in cells exposed to A�, ± 10 �M of SP600125, calpain inhibitor I, MG132, and z-VAD.fmk for 24 h (top); and ATF-6� (middle) andGRP94 (bottom) levels. B) Representative immunoblot of ATF-6� and GRP94 in cells exposed to A�, ± 50 �M of SP600125, calpain inhibitorI, MG132, and z-VAD.fmk for 2 h (top); and ATF-6� (middle) and GRP94 (bottom) levels. The results are normalized to endogenous �-actinprotein levels and expressed as mean ± SEM arbitrary units for at least three different experiments. *p < 0.01 from control; †p < 0.05 from A�alone.

cellular Ca2+ induced by A� exposure. ZnCl2, whichblocks pore formation was unable to prevent GRP94downregulation (Fig. 5A). Combinations of ZnCl2with calpain inhibitor I and TUDCA were also inef-fective, as was pretreatment with nifedipine that blocksL-type channels.

Next, we inhibited the release of Ca2+ from theER using xestospongin C and dantrolene, which blockinositol triphosphate receptors (IP3Rs) and ryanodinereceptors (RyRs), respectively. Under the establishedconditions, only xestospongin C was able to partiallyprevent the reduction of GRP94 (Fig. 5B). Finally,cyclosporine A, a strong inhibitor of calcium-inducedmPTP opening, did not prevent the reduction of GRP94

protein levels (Fig. 5C). Thus, Ca2+ deregulationinduced by A� incubation is not the main molecularmechanism of GRP94 downregulation.

Protein secretion inhibitors block GRP94downregulation

GA is an antibiotic that specifically inhibits hsp90and GRP94 [25, 47]. Inhibition of hsp90 mainly leadsto targeted degradation of client proteins via the ubiq-uitin proteasome pathway [48, 49], while the bestunderstood response to inhibition of GRP94 is thetranscriptional upregulation of ER chaperones [50]. Inaddition, GA also mediates inhibition of intracellular

68 R.J.S. Viana et al. / GRP94 Secretion by Aβ

Fig. 5. A�-induced downregulation of GRP94 in PC12 neuronal cells is independent of Ca2+ deregulation mechanisms. Cells were incubatedwith 25 �M A�, or no addition(control), ± 50 �M ZnCl2, 100 nM xestospongin C, 20 �M nifedipine, 10 �M dantrolene, and 0.05 or 5 � Mcyclosporin A for 2 h. In co-incubation experiments, the inhibitorswere added 30 min prior to incubation with A�. Total proteins were extractedfor western blot analysis as described in Materials and Methods. A) Representative immunoblot of GRP94 in cells exposed to A�, ± 50 �M ofZnCl2 (top); and GRP94 levels (bottom). Combinations of ZnCl2 with 50 �M calpain inhibitor I or 100 �M TUDCA for 2 h are also shown.Thapsigargin 3 �M was used as positive control. B) Representative immunoblot of GRP94 in cells exposed to A�, ± 100 nM xestosponginC, 20 �M nifedipine and 10 �M dantrolene (top); and GRP94 levels (bottom). C) Representative immunoblot of GRP94 in cells exposedto A�, ± 0.5 or 5 �M of cyclosporine A(top), and GRP94 levels (bottom). Thapsigargin 3 �M and brefeldin 18 �M were used as positivecontrols. The results are normalized to endogenous �-actin protein levels and expressed as mean ± SEM arbitrary units for at least three differentexperiments. *p < 0.01 from control.

trafficking of hsp90 substrates [27]. To determine theeffect of GA on A�-induced GRP94 downregulation,we preincubated cells with different concentrationsof GA for 30 min prior to exposure to A� for 2 h.We found that GA pretreatment prevented GRP94downregulation in a concentration-dependent manner(Fig. 6A), without significantly altering GRP94 mRNAexpression (Fig. 6B). In addition, GA effect was notdependent on de novo protein synthesis as determined

in experiments using cycloheximide (Fig. 6C). How-ever, a non-apoptotic concentration of brefeldin thatinhibits protein secretion was also able to preventGRP94 dowregulation (Fig. 6C). Finally, an ELISAassay for GRP94 in the supernatant of cells chal-lenged with A� established that GRP94 levels wereincreased in the extracellular medium of cells incu-bated with A�, independently of a necrotic mechanism(Fig. 6D). Retention of GRP94 within the cell by GA

R.J.S. Viana et al. / GRP94 Secretion by Aβ 69

Fig. 6. A�-induced downregulation of GRP94 in PC12 neuronal cells is blocked by protein secretion inhibitors. Cells were incubated with 25 �MA�, or no addition (control), ± 0.05 to 5 �M GA for 2 h. In co-incubation experiments, the inhibitors were added 30 min prior to incubation withA�. Total proteins and total RNA were extracted for western blot and RT-PCR analysis, respectively, as described in Materials and Methods.A) Representative immunoblot of GRP94 in cells exposed to A�, ± 0.05 to 5 �M GA. Thapsigargin 3 �M and brefeldin 18 �M were used aspositive controls. B) Representative RT-PCR of GRP94 in cells exposed to A�, ± 0.05 to 5 �M of GA. C) Representative immunoblot of GRP94in cells exposed to A�, ± 9 �M brefeldin (top); and GRP94 levels (bottom). Combinations of 0.5 �M GA with 100 �g/ml cycloheximide for2 h are also represented. The results are normalized to endogenous �-actin protein levels and expressed as mean ± SEM arbitrary units for atleast three different experiments. D) Secreted protein levels of GRP94. Supernatant anti-GRP94 of cells pretreated with 0.5 �M GA and 9 �Mbrefeldin, and challenged with A� for 2 h was quantified using an ELISA assay. The results are expressed as mean ± SEM for at least threedifferent experiments. CHX, cycloheximide; *p < 0.01 and p < 0.05 from control; †p < 0.05 from A� alone.

and brefeldin was associated with levels of apoptosisof ∼5%, similar to those seen in control cells. Thus,A� induces early stimulation of GRP94 secretion, aneffect that was mitigated by inhibition of the secretorypathway.

DISCUSSION

The role of A� in the pathogenesis and progres-sion of AD-associated neurodegeneration has not been

firmly established and even remains controversial. Thepresent study shows that A�-induced toxicity acti-vates caspase-12 that mediates ER-specific apoptosisin PC12 neuronal cells, which in turn is modulatedby antiapoptotic TUDCA. Notably, we demonstratedthat ER stress-induced cell death sensors and modu-lators are downregulated, and propose that ATF-6�is proteolytically cleaved by calpains, and GRP94 issecreted to the extracellular milieu during A�-inducedapoptosis.

70 R.J.S. Viana et al. / GRP94 Secretion by Aβ

There is a growing list of mediators that link ERstress to the apoptotic machinery. Caspase-12 was pro-posed to function as the apical caspase responsiblefor initiating an apoptotic cascade in response to ERstress and A� [31]. In the present study, we showedthat A�-induced apoptosis in PC12 neuronal-like cellsrequires caspase-12 activation that peaked after 24 h ofA� incubation, and which was inhibited by antiapop-totic TUDCA. Both the JNK pathway [51] and calpains[9] have been shown to activate caspase-12. Our resultsconfirmed that calpain inhibition partially abrogatedcaspase-12 activation. This fact suggests a role forintracellular Ca2+, given that calpains are activated byCa2+ [52] and A� incubation results in intracellularCa2+ overload [53].

ER is very sensitive to changes in Ca2+ homeostasis,leading to ER stress and caspase-12 activation [31, 54].Moreover, A� exerts early Ca2+-deregulation changeson ER, stimulating oxidative stress and, consequently,mitochondrial-dependent apoptotic cell death [55–57].In the present study, we investigated if caspase-12 acti-vation after A� incubation resulted from ER stress.Somewhat unexpectedly, we found that ER stressmarkers such as GRP94, ATF-6�, eIF2�, and CHOPwere downregulated in PC12 cells exposed to A� pep-tide. This effect was extremely pronounced for GRP94levels, which became undetectable after only 2 h ofincubation with A�. Our results suggested that celldeath is a consequence of deficient UPR activation.This is in agreement with recent studies showing thatUPR activation is neuroprotective against A�-inducedtoxicity, rather than strictly involved in cell death [35,58]. Finally, although not designed to discriminate theimportance of toxic assemblies, our results suggest thatlow A� molecular species are the triggering effect.Others have conclusively demonstrated that oligomericforms of A� are the main neurotoxic species, alsoinvolved in the release of Ca2+ from the ER [59, 60].

ER-resident molecular chaperones, such as GRP78and GRP94, are downregulated in brains of ADpatients and in PS1 mutant cells [37]. Furthermore,it has been suggested that plaque formation representsa failure of the ER to catalyze post-translational pro-cessing of A� [61], which results from age-dependentdecline in critical ER chaperones [62]. Our data indi-cate that downregulation of ER chaperone GRP94might be a result from the A� itself, rather than theaging process. Moreover, it has been demonstratedthat in etoposide-induced apoptosis, the reduction ofGRP94 levels is due to calpain cleavage [38]. Here,we demonstrated that under A�-induced apoptosis,GRP94 downregulation was calpain-independent. In

contrast, ATF-6� downregulation was prevented bycalpain inhibition. This fact, suggested that ATF-6� iseither a substrate for calpain cleavage, or that calpainregulates proteases that normally cleave and activateATF-6�. During the ER stress response, ATF-6� iscleaved with release of its cytoplasmic bZIP domainfrom the membrane and translocation to the nucleus[63]. In fact, A� potentiates tunicamycin-inducedneuronal death via calpain/caspase-12- and caspase-3-dependent cell death pathways, without affecting theUPR [64]. Our results demonstrate that both pathwaysare not only affected, but also interrelated. Further stud-ies should be performed to understand the kinetics ofboth pathways, and elucidate if calpain activation is asecondary pathway triggered by ER saturation.

GRP94 is a passive high-capacity reservoir for Ca2+,whose activity is regulated by Ca2 + . Furthermore,despite the redundancy of Ca2+-binding proteins inthe ER, the ability of GRP94 to bind Ca2+ is uniquelyimportant in the cellular context, as cells becomehypersensitive to perturbations in Ca2+ homeostasisin the absence of GRP94. Here, we demonstratedthat deregulation of Ca2+ homeostasis induced byA� incubation was not related with downregulationof GRP94. In fact, A� can induce intracellular Ca2+overload by several mechanisms (reviewed in [65]). Inthe present study, we used nifedipine and ZnCl2 at theplasma membrane level to inhibit L-type Ca2+ channeland zinc pore formation, respectively. Xestospon-gin C and dantrolene were used at the ER levelto inhibit Insp3Rs and RyRs receptors, respectively.Moreover, cyclosporine A inhibits Ca2+ regulationbetween ER-mitochondria through the mPTP pore for-mation. Despite the fact that GRP94 binds Ca2+ andserves as the major Ca2+ buffer in the ER [40–42],our results suggested that GRP94 downregulation isnot strictly related to Ca2+ deregulation. Nevertheless,inhibition of Ca2+ deregulation slightly prevented thereduction of GRP94 levels.

GRP94 has been identified in the extracellular spaceas a consequence of acute stress in lethal viral infectionand necrotic cell death [23, 24]. Indeed, stress proteinssuch as heat shock proteins, glucose regulated proteins,and calreticulin are released from cells in a variety ofcircumstances, thus interacting with adjacent cells orin some cases entering the bloodstream (reviewed in[66]). There is considerable evidence to support theimmunogenic properties of GRP94. The immune sys-tem recognizes free GRP94-peptide complexes andother chaperones and uses chaperoned antigenic pep-tides to activate CD8+cytotoxic T cells. Dendritic cellsrecognize GRP94 primarily via the CD91 receptor, and

R.J.S. Viana et al. / GRP94 Secretion by Aβ 71

endocytose GRP94 along with its chaperoned peptides.After internalization of the GRP94-peptide complexby antigen cross-presentation, and transport of thepeptide-loaded MHC to the cell surface, the peptide iscross-presented to CD8 T cells. Cross presented pep-tides chaperoned by GRP94 enter the MHC I pathwayas well as, albeit less efficiently, the MHC II path-way. A second consequence of receptor engagementby GRP94 is the innate immune effect, which involvesthe stimulation of the central signaling molecule NF-κB and its translocation into the nucleus (reviewed in[67]). Furthermore, GA inhibits GRP94 and hsp90 [68]and activates a heat shock response, possibly throughincreased expression of molecular chaperones [69, 70].Nonetheless, only a few reports have suggested that GAmay mediate the inhibition of intracellular protein traf-ficking [27]. In the present study, we demonstrated thatGRP94 downregulation is blocked in cells pretreatedwith GA in a dose-dependent manner. This occurredin a degradation-independent fashion, since neither theproteasome nor autophagy mechanisms were activatedunder the conditions tested. Further, mRNA levels ofGRP94 were not upregulated in the presence of GA,and synthesis of de novo proteins did not appear tobe involved. Next, we tested if the capacity of GA toinhibit intracellular protein trafficking was the mech-anism responsible to block GRP94 downregulation.Brefeldin, a known inhibitor of the secretion pathway,blocked GRP94 downregulation. Finally, we furtherconfirmed that GRP94 was secreted to the extracellularmedium in the presence of A� peptide. To our knowl-edge, this is the first report demonstrating that A�exposure in PC12 neuronal-like cells induces GRP94secretion. Once secreted, stress proteins may eitherincrease stress resistance, after binding to stress sen-sitive recipient cells, such as neurons, signal tissuedestruction and danger to inflammatory cells, or aidin immunosurveillance, by transporting intracellularpeptides to distant immune cells [66].

In conclusion, this study provides evidence that A�-induced toxicity triggers an early signaling responseby the ER, which is independent of the ER-stress path-way. ER-stress markers were downregulated, includingGRP94 that was specifically secreted to the extracellu-lar milieu in A�-induced apoptosis. Unrelated studieshave shown that GRP94 is required for innate immu-nity but not for cell viability [71], and that secretionof GRP94 is associated with necrosis [24]. Here, wereport an association of A�-induced apoptosis andpotential immunogenic GRP94 secretion, which mayultimately lead to the development of novel therapeu-tic strategies in AD. Additional studies are warranted

to determine the specific role of ER GRP94 chaperoneon A�-induced toxicity.

ACKNOWLEDGMENTS

This work was supported by grant PTDC/BIA-BCM/67922/2006 from Fundacao para a Ciencia e aTecnologia (FCT), Lisbon, Portugal. RJS Viana is therecipient of a PhD fellowship (SFRH/BD/30467/2006)from FCT, Portugal.

Authors’ disclosures available online (http://www.j-alz.com/disclosures/view.php?id=889).

REFERENCES

[1] LaFerla FM, Green KN, Oddo S (2007) Intracellular amyloid-beta in Alzheimer’s disease. Nat Rev Neurosci 8, 499-509.

[2] Thinakaran G, Koo EH (2008) Amyloid precursor proteintrafficking, processing, and function. J Biol Chem 283, 29615-29619.

[3] Romisch K (2005) Endoplasmic reticulum-associated degra-dation. Annu Rev Cell Dev Biol 21, 435-456.

[4] Bernales S, Papa FR, Walter P (2006) Intracellular signalingby the unfolded protein response. Annu Rev Cell Dev Biol 22,487-508.

[5] Thomenius MJ, Wang NS, Reineks EZ, Wang Z, DistelhorstCW (2003) Bcl-2 on the endoplasmic reticulum regulates Baxactivity by binding to BH3-only proteins. J Biol Chem 278,6243-6250.

[6] Nutt LK, Pataer A, Pahler J, Fang B, Roth J, McConkey DJ,Swisher SG (2002) Bax and Bak promote apoptosis by mod-ulating endoplasmic reticular and mitochondrial Ca2+stores.J Biol Chem 277, 9219-9225.

[7] Zinszner H, Kuroda M, Wang X, Batchvarova N, LightfootRT, Remotti H, Stevens JL, Ron D (1998) CHOP is implicatedin programmed cell death in response to impaired function ofthe endoplasmic reticulum. Genes Dev 12, 982-995.

[8] Rao RV, Hermel E, Castro-Obregon S, del Rio G, EllerbyLM, Ellerby HM, Bredesen DE (2001) Coupling endoplas-mic reticulum stress to the cell death program. Mechanism ofcaspase activation. J Biol Chem 276, 33869-33874.

[9] Nakagawa T, Yuan J (2000) Cross-talk between two cys-teine protease families. Activation of caspase-12 by calpainin apoptosis. J Cell Biol 150, 887-894.

[10] Yoneda T, Imaizumi K, Oono K, Yui D, Gomi F, KatayamaT, Tohyama M (2001) Activation of caspase-12, an endoplas-tic reticulum (ER) resident caspase, through tumor necrosisfactor receptor-associated factor 2-dependent mechanism inresponse to the ER stress. J Biol Chem 276, 13935-13940.

[11] Morishima N, Nakanishi K, Takenouchi H, Shibata T,Yasuhiko Y (2002) An endoplasmic reticulum stress-specificcaspase cascade in apoptosis. Cytochrome c-independentactivation of caspase-9 by caspase-12. J Biol Chem 277,34287-34294.

[12] Koch G, Smith M, Macer D, Webster P, Mortara R (1986)Endoplasmic reticulum contains a common, abundantcalcium-binding glycoprotein, endoplasmin. J Cell Sci 86,217-232.

[13] Mazzarella RA, Green M (1987) ERp99, an abundant,conserved glycoprotein of the endoplasmic reticulum, ishomologous to the 90-kDa heat shock protein (hsp90) and

72 R.J.S. Viana et al. / GRP94 Secretion by Aβ

the 94-kDa glucose regulated protein (GRP94). J Biol Chem262, 8875-8883.

[14] Maki RG, Old LJ, Srivastava PK (1990) Human homologue ofmurine tumor rejection antigen gp96: 5′-regulatory and cod-ing regions and relationship to stress-induced proteins. ProcNatl Acad Sci U S A 87, 5658-5662.

[15] Lee AS (1992) Mammalian stress response: induction of theglucose-regulated protein family. Curr Opin Cell Biol 4, 267-273.

[16] Booth C, Koch GL (1989) Perturbation of cellular calciuminduces secretion of luminal ER proteins. Cell 59, 729-737.

[17] Csermely P, Schnaider T, Soti C, Prohaszka Z, Nardai G(1998) The 90-kDa molecular chaperone family: structure,function, and clinical applications, A comprehensive review.Pharmacol Ther 79, 129-168.

[18] Little E, Lee AS (1995) Generation of a mammalian cell linedeficient in glucose-regulated protein stress induction throughtargeted ribozyme driven by a stress-inducible promoter. JBiol Chem 270, 9526-9534.

[19] McCormick TS, McColl KS, Distelhorst CW (1997) Mouselymphoma cells destined to undergo apoptosis in responseto thapsigargin treatment fail to generate a calcium-mediatedgrp78/grp94 stress response. J Biol Chem 272, 6087-6092.

[20] Lee SH, Song R, Lee MN, Kim CS, Lee H, Kong YY, KimH, Jang SK (2008) A molecular chaperone glucose-regulatedprotein 94 blocks apoptosis induced by virus infection. Hep-atology 47, 854-866.

[21] Basu S, Srivastava PK (2000) Heat shock proteins: the foun-tainhead of innate and adaptive immune responses. Cell StressChaperones 5, 443-451.

[22] Tamura Y, Peng P, Liu K, Daou M, Srivastava PK (1997)Immunotherapy of tumors with autologous tumor-derivedheat shock protein preparations. Science 278, 117-120.

[23] Berwin B, Reed RC, Nicchitta CV (2001) Virally inducedlytic cell death elicits the release of immunogenicGRP94/gp96. J Biol Chem 276, 21083-21088.

[24] Basu S, Binder RJ, Suto R, Anderson KM, Srivastava PK(2000) Necrotic but not apoptotic cell death releases heatshock proteins, which deliver a partial maturation signal todendritic cells and activate the NF-kappa B pathway. IntImmunol 12, 1539-1546.

[25] Pratt WB (1998) The hsp90-based chaperone system: involve-ment in signal transduction from a variety of hormone andgrowth factor receptors. Proc Soc Exp Biol Med 217, 420-434.

[26] Sepp-Lorenzino L, Ma Z, Lebwohl DE, Vinitsky A, RosenN (1995) Herbimycin A induces the 20 S proteasome- andubiquitin-dependent degradation of receptor tyrosine kinases.J Biol Chem 270, 16580-16587.

[27] Barzilay E, Ben-Califa N, Supino-Rosin L, Kashman Y,Hirschberg K, Elazar Z, Neumann D (2004) Geldanamycin-associated inhibition of intracellular trafficking is attributedto a co-purified activity. J Biol Chem 279, 6847-6852.

[28] Park DS, Morris EJ, Stefanis L, Troy CM, Shelanski ML,Geller HM, Greene LA (1998) Multiple pathways of neuronaldeath induced by DNA-damaging agents, NGF deprivation,and oxidative stress. J Neurosci 18, 830-840.

[29] Ramalho RM, Ribeiro PS, Sola S, Castro RE, Steer CJ,Rodrigues CM (2004) Inhibition of the E2F-1/p53/Baxpathway by tauroursodeoxycholic acid in amyloid beta-peptide-induced apoptosis of PC12 cells. J Neurochem 90,567-575.

[30] Viana RJ, Ramalho RM, Nunes AF, Steer CJ, Rodrigues CM(2010) Modulation of amyloid-beta peptide-induced toxicitythrough inhibition of JNK nuclear localization and caspase-2activation. J Alzheimers Dis 22, 557-568.

[31] Nakagawa T, Zhu H, Morishima N, Li E, Xu J, Yankner BA,Yuan J (2000) Caspase-12 mediates endoplasmic-reticulum-specific apoptosis and cytotoxicity by amyloid-beta. Nature403, 98-103.

[32] Xie Q, Khaoustov VI, Chung CC, Sohn J, Krishnan B, LewisDE, Yoffe B (2002) Effect of tauroursodeoxycholic acid onendoplasmic reticulum stress-induced caspase-12 activation.Hepatology 36, 592-601.

[33] Urano F, Wang X, Bertolotti A, Zhang Y, Chung P, HardingHP, Ron D (2000) Coupling of stress in the ER to activation ofJNK protein kinases by transmembrane protein kinase IRE1.Science 287, 664-666.

[34] Nishitoh H, Saitoh M, Mochida Y, Takeda K, Nakano H,Rothe M, Miyazono K, Ichijo H (1998) ASK1 is essen-tial for JNK/SAPK activation by TRAF2. Mol Cell 2, 389-395.

[35] Lee do Y, Lee KS, Lee HJ, Kim do H, Noh YH, Yu K, JungHY, Lee SH, Lee JY, Youn YC, Jeong Y, Kim DK, LeeWB, Kim SS (2010) Activation of PERK signaling attenuatesAbeta-mediated ER stress. PLoS One 5, e10489.

[36] Shringarpure R, Grune T, Sitte N, Davies KJ (2000) 4-Hydro-xynonenal-modified amyloid-beta peptide inhibits the protea-some: possible importance in Alzheimer’s disease. Cell MolLife Sci 57, 1802-1809.

[37] Katayama T, Imaizumi K, Sato N, Miyoshi K, Kudo T,Hitomi J, Morihara T, Yoneda T, Gomi F, Mori Y, NakanoY, Takeda J, Tsuda T, Itoyama Y, Murayama O, TakashimaA, St George-Hyslop P, Takeda M, Tohyama M (1999)Presenilin-1 mutations downregulate the signalling pathwayof the unfolded-protein response. Nat Cell Biol 1, 479-485.

[38] Reddy RK, Lu J, Lee AS (1999) The endoplasmic reticu-lum chaperone glycoprotein GRP94 with Ca2+-binding andantiapoptotic properties is a novel proteolytic target of cal-pain during etoposide-induced apoptosis. J Biol Chem 274,28476-28483.

[39] Biswas C, Ostrovsky O, Makarewich CA, Wanderling S,Gidalevitz T, Argon Y (2007) The peptide-binding activityof GRP94 is regulated by calcium. Biochem J 405, 233-241.

[40] Macer DR, Koch GL (1988) Identification of a set of calcium-binding proteins in reticuloplasm, the luminal content of theendoplasmic reticulum. J Cell Sci 91(Pt 1), 61-70.

[41] Nigam SK, Goldberg AL, Ho S, Rohde MF, Bush KT,Sherman M (1994) A set of endoplasmic reticulum pro-teins possessing properties of molecular chaperones includesCa2+-binding proteins and members of the thioredoxin super-family. J Biol Chem 269, 1744-1749.

[42] Cala SE, Jones LR (1994) GRP94 resides within cardiac sar-coplasmic reticulum vesicles and is phosphorylated by caseinkinase II. J Biol Chem 269, 5926-5931.

[43] Fox AP, Nowycky MC, Tsien RW (1987) Single-channelrecordings of three types of calcium channels in chick sensoryneurones. J Physiol 394, 173-200.

[44] Arispe N, Rojas E, Pollard HB (1993) Alzheimer diseaseamyloid beta protein forms calcium channels in bilayer mem-branes: blockade by tromethamine and aluminum. Proc NatlAcad Sci U S A 90, 567-571.

[45] Sanz-Blasco S, Valero RA, Rodriguez-Crespo I, Villalobos C,Nunez L (2008) Mitochondrial Ca2+overload underlies Abetaoligomers neurotoxicity providing an unexpected mechanismof neuroprotection by NSAIDs. PLoS One 3, e2718.

[46] Du H, Guo L, Fang F, Chen D, Sosunov AA, McKhann GM,Yan Y, Wang C, Zhang H, Molkentin JD, Gunn-Moore FJ,Vonsattel JP, Arancio O, Chen JX, Yan SD (2008) CyclophilinD deficiency attenuates mitochondrial and neuronal pertur-bation and ameliorates learning and memory in Alzheimer’sdisease. Nat Med 14, 1097-1105.

R.J.S. Viana et al. / GRP94 Secretion by Aβ 73

[47] Chavany C, Mimnaugh E, Miller P, Bitton R, NguyenP, Trepel J, Whitesell L, Schnur R, Moyer J, Neckers L(1996) p185erbB2 binds to GRP94 in vivo. Dissociationof the p185erbB2/GRP94 heterocomplex by benzoquinoneansamycins precedes depletion of p185erbB2. J Biol Chem271, 4974-4977.

[48] Whitesell L, Mimnaugh EG, De Costa B, Myers CE,Neckers LM (1994) Inhibition of heat shock protein HSP90-pp60v-src heteroprotein complex formation by benzoquinoneansamycins: essential role for stress proteins in oncogenictransformation. Proc Natl Acad Sci U S A 91, 8324-8328.

[49] Smith DF, Whitesell L, Nair SC, Chen S, Prapapanich V,Rimerman RA (1995) Progesterone receptor structure andfunction altered by geldanamycin, an hsp90-binding agent.Mol Cell Biol 15, 6804-6812.

[50] Lawson B, Brewer JW, Hendershot LM (1998) Gel-danamycin, an hsp90/GRP94-binding drug, induces increasedtranscription of endoplasmic reticulum (ER) chaperones viathe ER stress pathway. J Cell Physiol 174, 170-178.

[51] Tan Y, Dourdin N, Wu C, De Veyra T, Elce JS, Greer PA(2006) Ubiquitous calpains promote caspase-12 and JNKactivation during endoplasmic reticulum stress-induced apop-tosis. J Biol Chem 281, 16016-16024.

[52] Croall DE, DeMartino GN (1991) Calcium-activated neutralprotease (calpain) system: structure, function, and regulation.Physiol Rev 71, 813-847.

[53] Kuchibhotla KV, Goldman ST, Lattarulo CR, Wu HY, HymanBT, Bacskai BJ (2008) Abeta plaques lead to aberrant regu-lation of calcium homeostasis in vivo resulting in structuraland functional disruption of neuronal networks. Neuron 59,214-225.

[54] Orrenius S, Zhivotovsky B, Nicotera P (2003) Regulation ofcell death: the calcium-apoptosis link. Nat Rev Mol Cell Biol4, 552-565.

[55] Ferreiro E, Resende R, Costa R, Oliveira CR, Pereira CM(2006) An endoplasmic-reticulum-specific apoptotic pathwayis involved in prion and amyloid-beta peptides neurotoxicity.Neurobiol Dis 23, 669-678.

[56] Costa RO, Ferreiro E, Cardoso SM, Oliveira CR, PereiraCM (2010) ER stress-mediated apoptotic pathway inducedby Abeta peptide requires the presence of functional mito-chondria. J Alzheimers Dis 20, 625-636.

[57] Liu X, Xu K, Yan M, Wang Y, Zheng X (2010) Protec-tive effects of galantamine against Abeta-induced PC12 cellapoptosis by preventing mitochondrial dysfunction and endo-plasmic reticulum stress. Neurochem Int 57, 588-599.

[58] Casas-Tinto S, Zhang Y, Sanchez-Garcia J, Gomez-VelazquezM, Rincon-Limas DE, Fernandez-Funez P (2011) The ERstress factor XBP1s prevents amyloid-beta neurotoxicity.Hum Mol Genet 20, 2144-2160.

[59] Resende R, Ferreiro E, Pereira C, Resende DE, Oliveira C(2008) Neurotoxic effect of oligomeric and fibrillar speciesof amyloid-beta peptide 1–42: involvement of endoplasmicreticulum calcium release in oligomer-induced cell death.Neuroscience 155, 725-737.

[60] Resende R, Ferreiro E, Pereira C, Oliveira CR (2008) ERstress is involved in Abeta-induced GSK-3beta activation andtau phosphorylation. J Neurosci Res 86, 2091-2099.

[61] Erickson RR, Dunning LM, Olson DA, Cohen SJ, DavisAT, Wood WG, Kratzke RA, Holtzman JL (2005) In cere-brospinal fluid ER chaperones ERp57 and calreticulin bindbeta-amyloid. Biochem Biophys Res Commun 332, 50-57.

[62] Erickson RR, Dunning LM, Holtzman JL (2006) The effect ofaging on the chaperone concentrations in the hepatic, endo-plasmic reticulum of male rats: the possible role of proteinmisfolding due to the loss of chaperones in the decline inphysiological function seen with age. J Gerontol A Biol SciMed Sci 61, 435-443.

[63] Haze K, Yoshida H, Yanagi H, Yura T, Mori K (1999)Mammalian transcription factor ATF6 is synthesized asa transmembrane protein and activated by proteolysis inresponse to endoplasmic reticulum stress. Mol Biol Cell 10,3787-3799.

[64] Imai T, Kosuge Y, Ishige K, Ito Y (2007) Amyloid beta-protein potentiates tunicamycin-induced neuronal death inorganotypic hippocampal slice cultures. Neuroscience 147,639-651.

[65] Demuro A, Parker I, Stutzmann GE (2010) Calcium signalingand amyloid toxicity in Alzheimer disease. J Biol Chem 285,12463-12468.

[66] Calderwood SK, Mambula SS, Gray PJ Jr, Theriault JR(2007) Extracellular heat shock proteins in cell signaling.FEBS Lett 581, 3689-3694.

[67] Strbo N, Podack ER (2008) Secreted heat shock protein gp96-Ig: an innovative vaccine approach. Am J Reprod Immunol 59,407-416.

[68] Neckers L (2002) Hsp90 inhibitors as novel cancer chemo-therapeutic agents. Trends Mol Med 8, S55-S61.

[69] Sittler A, Lurz R, Lueder G, Priller J, Lehrach H, Hayer-HartlMK, Hartl FU, Wanker EE (2001) Geldanamycin activates aheat shock response and inhibits huntingtin aggregation in acell culture model of Huntington’s disease. Hum Mol Genet10, 1307-1315.

[70] McLean PJ, Klucken J, Shin Y, Hyman BT (2004) Gel-danamycin induces Hsp70 and prevents alpha-synucleinaggregation and toxicity in vitro. Biochem Biophys Res Com-mun 321, 665-669.

[71] Randow F, Seed B (2001) Endoplasmic reticulum chaperonegp96 is required for innate immunity but not cell viability. NatCell Biol 3, 891-896.