Intracellular mechanisms mediating the neuronal death andastrogliosis induced by the prion protein fragment 106±126

Stefano Thellunga, Tullio Floriob, Alessandro Corsaroa, Sara Arenaa,Massimo Merlinoa, Mario Salmonac, Fabrizio Tagliavinid, Orso Bugianid,

Gianluigi Forlonic, Gennaro Schettinia,*aUnit of Pharmacology and Neuroscience National Cancer Institute, Advanced Biotechnology Centre, Department of Oncology, Section of

Pharmacology University of Genoa, Largo Rosanna Benzi 10, 16132, Genoa, ItalybDepartment of Biomedical Science, Section of Pharmacology University of Chieti, Italy

cInstitute of Pharmacological Research, Mario Negri, I-20157, Milano, ItalydNational Neurological Institute, Carlo Besta, I-20123, Milano, Italy

Received 17 May 1999; received in revised form 11 August 1999; accepted 25 August 1999

Abstract

Prion encephalopathies include fatal diseases of the central nervous system of men and animals characterized by nerve cell

loss, glial proliferation and deposition of amyloid ®brils into the brain. During these diseases a cellular glycoprotein (the prionprotein, PrPC) is converted, through a not yet completely clear mechanism, in an altered isoform (the prion scrapie, PrPSc) thataccumulates within the brain tissue by virtue of its resistance to the intracellular catabolism. PrPSc is believed to be responsible

for the neuronal loss that is observed in the prion disease. The PrP 106±126, a synthetic peptide that has been obtained from theamyloidogenic portion of the prion protein, represents a suitable model for studying the pathogenic role of the PrPSc, retaining,in vitro, some characteristics of the entire protein, such as the capability to aggregate in ®brils, and the neurotoxicity. In thiswork we present the results we have recently obtained regarding the action of the PrP 106±126 in di�erent cellular models. We

report that the PrP 106±126 induces proliferation of cortical astrocytes, as well as degeneration of primary cultures of corticalneurons or of neuroectodermal stable cell lines (GH3 cells). In particular, these two opposite e�ects are mediated by the sameattitude of the peptide to interact with the L-type calcium channels: in the astrocytes, the activity of these channels seems to be

activated by PrP 106±126, while, in the cortical neurons and in the GH3 cells, the same treatment causes a blockade of thesechannels causing a toxic e�ect. 7 2000 ISDN. Published by Elsevier Science Ltd. All rights reserved.

1. Introduction

Prion diseases are encephalopathies of animals and

man, characterized by progressive neuronal loss and

reactive gliosis, often accompanied by a spongiform

brain alteration and the deposition of amyloid ®brils.

Interestingly, these diseases are invariably fatal without

evoking, in the host, any in¯ammatory or immune re-

sponse [33].

The term ``prion'' was coined by Stanley Prusiner

who de®ned the infectious molecule responsible of the

transmission of these disorders as a proteinaceous and

Int. J. Devl Neuroscience 18 (2000) 481±492

0736-5748/00/$20.00 7 2000 ISDN. Published by Elsevier Science Ltd. All rights reserved.

PII: S0736-5748(00 )00005 -8

www.elsevier.com/locate/ijdevneu

* Corresponding author. Tel.: +39-010-573-7256; fax: +39-010-

573-7257.

E-mail address: schettini@sirio.cba.unige.it (G. Schettini).

Abbreviations: PrPC, cellular prion protein; PrPSc, scrapie prion

protein; PrP 106±126, prion protein fragment 106±126; SCR,

scrambled prion protein; CJD, Creutzfeldt±Jacob Disease; GSS,

Gerstmann±Straussler±Scheinker disease; BSE, Bovine spongiform

encephalopathy; VSCC, voltage sensitive calcium channels; TUNEL,

terminal deoxynucleotidyltransferase d-UTP-biotin nick end labeling;

PKA, Protein Kinase A; FCS, fetal calf serum; NMDA, N-methyl-D-

aspartate.

infectious particle that resists the degradation inducedby procedures that inactivate nucleic acids [30].

Prion encephalopathies comprehend infectious, gen-etic and sporadic disorders which are now classi®edtogether because they all seem likely to be caused bythe post-translational modi®cation of a cellular glyco-protein, the prion protein (PrPC), in an altered iso-form, called PrPSc (from scrapie). The novel physico-chemical properties of this pathogenic protein allow itsaccumulation within the brain and seem to be respon-sible for the neurological damages that occur in thesediseases [32].

1.1. Prion diseases

Human prion encephalopathies include Kuru,Creutzfeldt±Jacob Disease (CJD), Gerstmann±Strauss-ler±Scheinker Disease (GSS) and Fatal FamilialInsomnia. Kuru is an epidemic neurodegenerative dis-ease that occurred among a population in the Foreregion of New Guinea. The reason for the trans-mission of this disorder was discovered to result fromthe consumption of the brains of dead relatives duringcannibalistic rituals. Kuru has now almost disappearedwith the cessation of cannibalism [1].

GSS disease is a dominantly inherited form linkedto the mutations in several codons of the gene for theprion protein (PrP). The clinical characteristics of thisdisorder include cerebellar ataxia, extrapimamidalsigns and a progressive dementia. A common featureof the di�erent types of GSS is the deposition, in thebrain, of ®brils of the altered prion protein, but theydi�er in the site of deposition [21].

Fatal Familial Insomnia is also an inherited priondisease characterized by a progressive insomnia, cer-ebellar and pyramidal alteration, myoclonus anddementia: this disorder shares the same mutation as afamiliar form of CJD, in the PrP gene, at the codon178 [25].

CJD occurs primarily as a sporadic disease, butiatrogenic or inherited forms (about 10% of the total)have been reported. Recently, some cases of an atypi-cal form of CJD occurred in the United Kingdom andin France, raising the possibility that the disease couldderive from the ingestion of meat obtained from BSE-contaminated cattle [9]. In addition to the prion dis-eases of humans, some encephalopathies of animalsmay be recognized as prion disorders: the most knownand studied is scrapie of sheep and goats and bovinespongiform encephalopathy (BSE), also named ``madcow disease''; two neurodegenerative disorders charac-terized by cerebellar ataxia.

1.2. The prion protein

The mature cellular prion protein (PrPC) is a

single chain polypeptide that contains two glycosyla-tion sites. It is anchored to the surface of neurons andother cell types with a glycosyl-phosphatydyl-inositolanchor (GPI) at its carboxyl terminus [28,41]. Thisprotein is abundantly expressed in the CNS of allmammalian species, but its function is still largelyunknown. Recent studies demonstrated that transgenicmice carrying a knock-out of the gene coding for PrP(named PRNP ) showed an impairment of the GABA-ergic neurotransmission [12] and, after 70 weeks ofage, a progressive impairment of motor coordinationrelated to an extensive loss of Purkinje neurons [37]. Inaddition, a recent report shows that the prion proteincan bind copper [7]. However, de®nitive studies regard-ing its physiologic role have not yet been produced.

The tridimensional structure of PrPC is not yetcompletely clari®ed, but the NMR characterization ofa full length recombinant murine prion protein mPrP(23±231) predicts that it consists in the globular region,close to the carboxy terminus (residues 121±231) thatcomprehends three a-helix plus one b-sheet, and alarge ¯exible coil (residues 23±120) at the N-terminal[23,35]. PrPC seems to transit from the endoplasmicreticulum through the Golgi and to be ®nally trans-ported to the cell surface where it is anchored by aGPI anchor. The PrPC re-enters the cell through theendosome pathway where, in scrapie infected cells, itinteracts with the PrPSc and is forced to assume thenew altered secondary structure [44]. Recently, theconversion from PrPC to PrPSc was reported to occurin caveole-like-domains where the cellular prion, underphysiological conditions, is clustered [46] and wherethe presence of both isoforms has been shown [27].

The most accredited theory elaborated to explainthe transmission of the PrPSc predicts that the ex-ogenous altered protein can act as a template forrefolding, in the pathologic form, of the PrPC. In thisprocess a portion of the coil region of the cellular pro-tein is converted into b-sheet regions [35]. The initialinteraction between the two isoforms needs a longtime, probably by virtue of their stability, but whenthis ®rst barrier is overcome, the production of patho-gen proteins will proceed with a logarithmic pro-gression [2].

Knock-out studies have also shown that mice devoidof the PrP gene product are resistant to the scrapieinfection and that the development and progression ofthe disease is inversely related to the extent of thePrPC expression [8]. Increasing evidence indicates thatthe PrPSc also requires the presence of a host speciespeci®c protein that is able to interact with both iso-forms of PrP and to induce the structural shift PrPC±PrPSc: such a protein, called protein X, has been pre-dicted to act, as a chaperone, stabilizing PrPC±Sc inter-mediates along the unfolding±refolding process [45].The introduction in the organism of the PrPSc may

S. Thellung et al. / Int. J. Devl Neuroscience 18 (2000) 481±492482

happen with an infective modality, or inherited follow-ing a sporadic or familial transmission.

PrPC has a high a-helix content (42%) and little b-sheet (3%). In contrast, the post-translationally modi-®ed protein, named PrPSc, from scrapie is composed of43% of the b-sheet and 30% of a-helix [29]. The cellu-lar PrP is soluble in non-denaturing detergents and isfully hydrolyzed by proteases, but the PrPSc is insolu-ble and can only be partially degraded by proteases[28]. Meanwhile, as a result, the cellular prion proteinis rapidly catabolized into the cell, the altered proteinis truncated at its N-terminal and origins a protease re-sistant core, called PrP 27±30 (molecular mass 27±30 kDa), that has an even higher percentage in b-sheetregions (54%) and a lower a-helix content (21%) inrespect to the entire PrPSc. The PrP 27±30 polymerizesinto rod-shaped structures with the same character-istics of amyloid [31] and accumulates in the nervoustissue with a pattern that is temporally and topogra-phically associated with neuronal loss and astrogliosis[13,14] suggesting that the abnormal isoforms of PrPmight play a role in the pathogenesis of nerve cell de-generation and glial cell reaction. In this regard, cellculture studies have shown that the protease-resistantcore of PrPSc induces neuronal death and astrocyteproliferation in vitro [26,13].

1.3. PrP (106±126) as a model to study the cellulare�ects of PrPSc

In several previous studies, the pathogenic role ofPrP has been analyzed using synthetic peptides hom-ologous to consecutive segments of the amyloid pro-tein puri®ed from brain tissue of patients with GSSdisease using, as a template, the sequence comprisedbetween the amino acids 57±150 [42,43]. Among allthe synthetic peptides that have been tested, the frag-ment matching the residues 106±126 was found toshare several properties in vitro with the entire PrPSc.Prolonged exposure of primary cultures of hippocam-pal neurons to the PrP 106±126 resulted in apoptoticneuronal death [18]. Conversely, the same peptideinduced the hypertrophy of astrocytes [19] and pro-liferation of microglia and astrocytes [5,16]. The pep-tide PrP 106±126 shows a high tendency to adopt a b-sheet structure and to form amyloid ®brils in vitro [43]and is partially resistant to proteolysis [40]. Thissequence largely corresponds to a putative a-helicaldomain that is thought to shift from a-helix to the b-sheet in the conversion of PrPC to PrPSc. Noteworthy,this region is an integral part of all abnormal PrP iso-forms that accumulate in the brain of patients withprion encephalopathies (e.g. CJD and GSS disease)while it is disrupted by a proteolytic cleavage occurringbetween residues 110±111 in normal PrP metabolism[10]. There is evidence that microglia could transduce

the toxic e�ect of the PrP 106±126 on primary culturesof cerebellar neurons through the production of freeradicals [5,6] and this suggests the existence of complexand investigation-deserving interactions between glialcells and prion neurotoxicity. Astrogliosis developsearly in the course of prion-related encephalopathiesand precedes neuronal loss, supporting the view thathypertrophy and proliferation are consequences of thedeposition of PrPSc and PrP amyloid, rather than a re-sponse to neuronal damage. However, the molecularmechanisms responsible for astrocyte activation areunknown.

2. Materials and methods

2.1. Peptides

PrP 106±126 (sequence: KTNMKHMAGAAAA-GAVVGGLG) and PrP 106±126 scrambled (SCR)(sequence: NGAKALMGGHGATKVMVGAAA)were synthesized and puri®ed as previously reported[19].

2.2. Cell cultures

2.2.1. Type I cortical astrocytesCells were obtained from two-day-old rat pups [16].

The rats were sacri®ced by decapitation and the cortexwas rapidly removed and fragmented with a microscis-sor. The pieces were placed in a phosphate bu�eredsaline solution without Ca2+/Mg2+ (PBS) containing0.125% trypsin (Sigma), and incubated in a shakingbath at 378C for 15 min. The tissue was mechanicallyfragmented and plated in a 75 cm2 Flask (Costar). Theculture medium was Dulbecco's Modi®ed Eagle Med-ium (DMEM Hyclone), supplemented with 10% FetalBovine Serum (Gibco). As the cells reached the con¯u-ence, the type 1 astrocytes were split from other glialcells by gently shaking the ¯ask: the type 1 astrocytesdid not detach from the substrate.

2.2.2. Cortical neuronsThe primary cultures were prepared as described

[39]. The brains were obtained from 16±17-day-old ratembryos. The tissues were incubated with 26 U/mlpapaine for 7 min at 378C and dissociated by tritura-tion in Earle's Balanced Salt Solution (EBSS) contain-ing 275 U/ml Dnase (Sigma), 1 mg/ml bovine serumalbumin (BSA) (Sigma) and 1 mg/ml ovomucoid(Sigma). Cells were plated on coverslips precoated with1 mg/ml poly-L-lysine (Sigma) in MEM/F12 containing10% fetal calf serum. Cytosine arabinoside (10 mM)was added within 48 h of plating to prevent thegrowth of non-neuronal cells.

S. Thellung et al. / Int. J. Devl Neuroscience 18 (2000) 481±492 483

2.2.3. GH3 cellsThe cells were cultured in Ham's F10 (ICN Flow),

100 ug/ml penicillin, 100 mg/ml streptomycin (Hyclone)supplemented with 10% fetal calf serum (ICN Flow).

2.3. Reverse transcriptase (RT)-PCR

The total cell RNA was extracted from astrocytes,cortical neurons and GH3 cells using the acid phenoltechnique [11]. cDNA was then synthesized usingAMV RT (Finnzyme OY, Finland) using oligo dT(16)primers. Ten ng of cDNA was subsequently used inthe PCR reaction, that was carried for 40 cycles (1 min948C, 1 min 558C, 1 min 728C) followed by 7 min ofprimers extension at 728C. The primers used corre-sponded to the DNA sequence of the amino acids 61±66 and 116±120 of the mouse PrP sequence. Theampli®ed products were then transferred to nylonmembranes (Hybond N, Amersham) and then hybri-dized with a radiolabeled PrP cDNA fragment, follow-ing standard procedures.

2.4. Cell proliferation and survival

2.4.1. [3H]-thymidine incorporation assayDNA synthesis activity was evaluated in cortical

type I astrocytes by means of the [3H]-thymidine incor-poration assay. Cells were serum starved for 24 hbefore stimulation with the test substances. In the last4 h of incubation 2 mCi/ml [3H]-thymidine (Amersham)was added to the wells. [3H]-thymidine incorporationwas determined, as previously reported [15].

2.4.2. MTT assayThe mitochondrial function, as an index of cell num-

bers or viability, was evaluated by measuring the levelsof mitochondrial dehydrogenase activity using a re-duction of 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyl tet-razolium bromide (MTT) as the substrate. Its cleavageto a purple formazan product by dehydrogenase wasquanti®ed spectrophotometrically at 570 nm. Theassay cells were incubated for 4 h with 0.25 mg/mlMTT in serum free F10 medium at 378C. Then forma-zan crystals were dissolved in DMSO and the valuesof absorbance of the samples read in a spectropho-tometer at 570 nm.

2.5. Detection of apoptosis

2.5.1. Bis-benzimide staining [24]Cells, plated on 25 mm glass coverslips, were syn-

chronized by FCS-deprivation for 48 h and then trea-ted with the test substances in serum free medium fordi�erent times. At the end of the experiment GH3 cellswere ®xed in 4% paraformaldehyde (30 min) and thenincubated with 1 mg/ml bis-benzimide at 378C for

30 min; they were then washed three times and ana-lyzed for condensed or disrupted nuclei using a ¯uor-escent microscope. At least 1000 cells per coverslipwere analyzed, and experiments performed in duplicatewere repeated at least three times.

2.5.2. TUNEL (terminal deoxynucleotidyltransferase d-UTP-biotin nick end labeling) immunostaining

The ApopTag ¯uorescent assay kit (ONCOR) wasalso used to determine the number of apoptotic nucleithrough the detection of free 3 '-OH ends of newlycleaved DNA in situ, as previously reported [39].Digoxigenin-dUTP was added to 3 '-OH ends ofdouble- and single-stranded DNA by terminal deoxy-nucleotidyltransferase. An antidigoxigenin Fab anti-body conjugated with ¯uorescein was then used toimmunostain nuclei that had a high concentration of3 '-OH ends of cleaved DNA. TUNEL-stained cellswere counterstained with propidium iodide. The assaywas performed according to the manufacturers'instructions. The apoptotic nuclei showed condensedDNA and yellow-green ¯uorescence, compared withhealthy cells that showed large, red nuclei.

2.6. Measurement of intracellular calcium levels atsingle cell level

All the assays were performed as previously reported[20]. The astrocytes and GH3 cells were plated onpoly-L-lysine coated 25 mm glass coverslips, and trans-ferred to a 35 mm petri dish. On the day of the exper-iment the cells were washed for 10 min with abalanced salt solution (HEPES 10 mM, pH 7.4; NaCl150 mM; KCl 5.5 mM; CaCl2 1.5 mM; MgSO4

1.2 mM; Glucose 10 mM), loaded with Fura-2 penta-acetoxymethyl ester (4 mM) (Calbiochem) for 60 minat room temperature and then washed for 10 min inthe same bu�er. Cells were treated with test substancesor vehicle during the incubation with Fura-2 or in the[Ca++] recording frames. For ¯uorescence measure-ment coverslips were mounted on a coverslip-chamber(Medical System) and was imaged with an invertedNikon diaphot microscope using a Nikon 40X/1.3 NAFluor DL objected lens. Cells were illuminated with aXenon lamp (Osram) with quartz collector lenses. Ashutter and a ®lter wheel containing the two di�erentexcitation ®lters (340 and 380 nm) were controlled bya computer equipped with Tardis software. Intracellu-lar [Ca++] were calculated according to the Grynkie-wicz equation [22].

2.7. Statistical analysis

Experiments were performed in quadruplicate andrepeated at least three times. Statistical analysis wasperformed by means of one-way ANOVA. A p-value

S. Thellung et al. / Int. J. Devl Neuroscience 18 (2000) 481±492484

less than or equal to 0.05 was considered statisticallysigni®cant.

3. Results

3.1. PrP 106±126-induced astrocytes proliferation

The treatment with PrP 106±126 induced a time-and concentration-dependent astrocyte proliferation,evaluated by means of the MTT assay, an indirect

index of the cell number (Fig. 1(A)). This e�ect wasalready signi®cant after one day of treatment, lastedup to nine days, and was more pronounced for thehigher concentration tested (50 mM).

Similar results were obtained using the [3H]-thymi-dine uptake assay that measures the DNA synthesisactivity of the cells (Fig. 1(B)). In these experimentsthe activity was measured after nine days of treatmentwith PrP 106±126 (50 mM), SCR (50 mM) or FCS(10%).

PrP 106±126, but not the scrambled peptide (SCR),caused a signi®cant increase in cortical type I astrocyteproliferation, and the extent of cell growth was about50% to that induced by the serum stimulation, indicat-ing that PrP 106±126 is a very powerful growth pro-moting agent for type I astrocytes. During all the timeof the experiments and even after longer treatments(up to 14 days), no signs of toxicity were observed inthe PrP 106±126-treated astrocytes (data not shown).

To identify the possible second messengers involvedin this e�ect we used selective inhibitors of the trans-duction pathways. In this series of experiments the[3H]-thymidine incorporation assay was tested afteronly one day of treatment, due to the potential toxicityof the inhibitors used. The increase in DNA synthesisand proliferative activity induced by PrP 106±126 wascompletely blocked by the pretreatment with nicardi-pine [38], an L-type voltage-sensitive calcium channel(VSCC) blocker, while KT-5720, an inhibitor of theprotein kinase A (PKA), vanadate, an inhibitor ofphosphotyrosine phosphatases and MK801, a selectiveinhibitor of the NMDA receptor-gated calcium chan-nel, resulted ine�ective (Table 1).

3.2. PrP 106±126 induced calcium increase in astrocytes

Micro¯uorimetric analysis in a single cell showedthat PrP 106±126 treatment caused a rapid increase inthe astrocytes in [Ca++]i characterized by a peak, a

Table 1

[3H]-Thymidine uptake assaya

Pretreatment Basal PrP 106±126 50 mM PrP Sc

Vehicle 10025 152210� 10827

Nicardipine 1 mM 6523 7029��

MK801 10 mM 10127 14929

KT5720 50 nM 9825 151215

Vanadate 50 mM 11029 153210

a E�ect of blockers of di�erent second messenger systems on PrP

106±126 stimulated DNA neosynthesis in astrocytes: cells were

serum deprived for 48 h and then treated with PrP and antagonists

for 24 h. In the last 4 h of incubation, 2 mCi/ml [3H]-thymidine was

added to the wells. Data were obtained as cpm/well and plotted as a

percentage of the DNA neosynthesis in unstimulated conditions. �p<0.001 vs control; ��p<0.01 vs PrP 106±126.

Fig. 1. Proliferative e�ects of PrP 106±126 on type I cortical astro-

cytes. (A) Time-course of the proliferative e�ects of PrP 106±126 at

two concentrations (10 and 50 mM) evaluated using the MTT test.

Data are expressed as a percentage of the OD570 at time 0. (B) DNA

neosynthesis in astrocytes: cells were treated, every three days, for

nine days with vehicle, PrP 106±126 (50 mM), PrP scrambled (50 mM)

and fetal calf serum (FCS) (10%). The day of the experiment cells

were pulsed for 4 h with 2 mCi/ml [3H]-thymidine. The treatment

with the serum free medium was considered to be the basal level;

data were obtained as cpm/well and plotted as a percentage of the

basal value 2 SEM of three independent experiments in triplicate.�=p<0.05 and ��=p<0.01 vs basal values.

S. Thellung et al. / Int. J. Devl Neuroscience 18 (2000) 481±492 485

sustained plateau and a slow return to basal levels(Fig. 2(A)). The prion fragment e�ect was completelyinhibited by the absence of extracellular calcium ionsor by the L-type VSCC blocker nicardipine (Fig. 2(B)).The SCR did not evoke any increase of the calciumconcentration in any cell tested.

3.3. Neurotoxicity of PrP 106±126

PrP 106±126 (50 mM, for 10 days) caused a markedtoxicity in primary cultures of cortical neurons, charac-terized by cell loss and disruption of the neuritic net(Fig. 3(A)). The PrP 106±126-dependent cell deathshowed the characteristics of the apoptotic process asshown using the TUNEL assay (Fig. 3(B)).

3.4. Expression of cellular prion protein in corticalneurons, astrocytes and GH3 cells

We performed an RT-PCR reaction on the total

RNA of astrocytes, cortical neurons and GH3 cells,that were tested using speci®c cDNA primers derivedby the mouse PrP sequence. The PCR reaction resultedin the ampli®cation of a transcript of the expectedlength (370 bp) that hybridized with a speci®c radio-labeled PrP cDNA fragment in a Southern blot analy-sis (Fig. 4).

3.5. Cytotoxicity of PrP 106±126 on GH3 cells

PrP 106±126 (10 mM, for 3±5 days) caused a signi®-cant death of quiescent GH3 cells and a strong re-duction of cell proliferation induced by foetal bovineserum (Fig. 5). The cell death induced by the peptideshowed some hallmarks of apoptosis, assessed bymeans of DNA staining with the nuclear ¯uorescentdye bis-benzimide (Fig. 6). The L-type VSCC blockernicardipine mimicked the cytotoxic and the proapopto-tic e�ects exerted by the PrP 106±126, while the PrPscrambled was completely ine�ective (not shown).

Fig. 2. Single cell micro¯uorimetry: e�ect of PrP 106±126 and PrP scrambled on the [Ca++]i in astrocytes and reversal of the PrP 106±126 e�ects

by removal of extracellular Ca++ or by the pretreatment with nicardipine (1 mM). The presence of extracellular calcium and the activity of the

L-type calcium channels are requisites for the PrP induced calcium rise. Panel A shows the rapid calcium rise evoked by the addition of PrP

106±126 (50 mM) in a single astrocyte. Panel B shows that each column is deduced from three di�erent experiments where an average of 10 cells

was evaluated. ��=p<0.001 vs calcium level in the untreated cells before the addition of the PrP.

S. Thellung et al. / Int. J. Devl Neuroscience 18 (2000) 481±492486

3.6. PrP 106±126 counteracts the proliferative action ofKCl on GH3 cells

In this cellular model, the addition of the culturemedium with depolarizing concentrations of KCl (25and 50 mM) was able to induce a concentration-depen-dent increase of DNA neosynthesis in the GH3 cells.The preincubation of the cultures with the PrP 106±126 50 mM and nicardipine inhibited the KCl-inducedcell proliferation (Fig. 7).

3.7. Micro¯uorimetry on GH3 cells

A depolarizing pulse with potassium chloride(50 mM) evokes, in the GH3 cells, a rapid increase ofthe cytosolic calcium concentration. Pretreatment withthe PrP 106±126 caused a strong abolishment of thiscellular response to the high [K+]o depolarization,while the addition of the scrambled peptide to the cul-ture did not induce any decrease in the calcium level(Fig. 8). The pretreatment of GH3 cells with nicardi-pine resulted in a pattern of response superimposableto that obtained with PrP 106±126.

4. Discussion

We report that the PrP fragment 106±126 exerts adirect trophic e�ect on primary cultures of puri®ed ratcortical type I astrocytes. By means of the [3H]-thymi-dine uptake (index of DNA neosynthesis), we show

Fig. 3. (A) Morphological appearance of cortical neurons after 10 days of treatment with PrP 106±126 (50 mM) as evidenced by phase contrast

microphotographs. A signi®cant cell loss and disruption of neuritic net is observed in the PrP 106±126-treated neurons. (B) Ten days exposure to

PrP 106±126 (50 mM) induces apoptosis in cortical neurons (evidenced with the TUNEL test). Cells were ®xed, permeabilized and the OH term-

inals of the apoptotic DNA were linked to digoxigenin. Apoptotic nuclei were stained with an anti digoxigenin antibody conjugated with ¯uor-

escein and evidenced a green ¯uorescence.

Fig. 4. Expression of PrP mRNA in: cortical neurons (C); astrocytes

(A); and GH3 cells (G). The ®gure represents a Southern blot of an

RT-PCR reaction performed using speci®c primers designed from

the PrP sequence, hybridized with a radiolabeled PrP cDNA probe

S. Thellung et al. / Int. J. Devl Neuroscience 18 (2000) 481±492 487

that the PrP 106±126 strongly and dose-dependentlyenhances the DNA synthesis. Conversely, a scrambledPrP peptide did not induce any e�ect (Fig. 1(A)).Interestingly, the e�ects of PrP 106±126 did not lastmore than 24 h after the last PrP 106±126 treatmentsuggesting that, while the PrP 106±126 fragment re-sembles the protein PrPSc in a number of biologicale�ects, it does not seem to be able to induce the trans-formation of PrPC in PrPSc in the astrocytes in culture[16]. Several inhibitors of intracellular transductionsystems were tested to identify a possible mechanismof the proliferative e�ect exerted by the PrP 106±126.The increase of DNA synthesis stimulated by PrP106±126 was reduced only by nicardipine (Table 1).These data suggest that the proliferative activity ofPrP 106±126 on type I astrocytes may be related to anincrease in [Ca++]i mediated by the L-type VSCC,whose presence in astrocytes has been demonstrated[4].

Indeed, PrP 106±126 caused a rapid increase of theintracellular calcium level (Fig. 2): this calcium aug-ment was dependent on the presence of calcium in theexternal bu�er and was completely inhibited by thecell pretreatment with nicardipine.

These results strongly suggest that PrP 106±126directly interacts with astroglial cell membrane caus-ing, through an L-type VSCC dependent enhancementof Ca++ in¯ux, an increase of astrocytes proliferation.This observation suggests that the gliosis observed

throughout the development of the prion diseases isnot merely a tissutal reaction to the neuronal celldeath, but could be consequent to a direct proliferativeaction exerted by the prion protein on the di�erenttypes of glial cells.

It was reported that the entire PrPSc moleculeinduced apoptotic death in cultured neurons [26]. Tofurther con®rm the identity of the e�ects between theinfective PrPSc and the PrP 106±126 that we used, weshowed, through the TUNEL test, that the prolongedtreatment with PrP 106±126 (50 mM, 10 days) inducedapoptosis in cortical neurons (Fig. 3). To strengthenour knowledge about the role of the L-type VSCC inthe PrP 106±126 control of cell proliferation, we stu-died the e�ects of PrP 106±126 on the GH3 cell line,an excitable clonal cell line of neuroectodermal origin,that endogenously express PrPC, in virtue of a well-described characterization of the L-type voltage sensi-tive calcium channels (VSCC) on this cell line [34],[36]. Indeed, it was reported that, similarly to PrPSc,PrP 106±126 also requires the expression, in the host,of the cellular prion isoform as a conditioning factorfor its neurotoxic e�ects [5,8]. This evidence, togetherwith the proapoptotic e�ect exerted by PrP 106±126 incortical cultures, allowed us to use GH3 cells asmodels to study how the PrP 106±126 can a�ect cellsurvival.

The treatment of the GH3 cells with the PrP 106±126, surprisingly, did not result in an increase in the

Fig. 5. GH3 survival and proliferation assessed by means of the MTT test: the treatments started after 48 h of serum deprivation and lasted

three and ®ve days. The PrP 106±126 and the PrP scrambled were both diluted in serum free medium and in serum containing medium. PrP

106±126 signi®cantly a�ects GH3 survival and reduces the proliferative action exerted by the fetal calf serum. Data are expressed as a percentage

of the cell viability assessed after three and ®ve days of treatment with the serum free medium alone (control) in three separate experiments per-

formed in quadruplicate. �=p<0.05 vs control; ��=p<0.01 vs control; =p<0.01 vs FCS-stimulated cell growth.

S. Thellung et al. / Int. J. Devl Neuroscience 18 (2000) 481±492488

cell proliferation, as observed in the astrocytes, butinduced a reduction in cell proliferation and causedapoptotic cell death (Figs. 5 and 6). [3H]-thymidineuptake experiments showed that the treatment of theGH3 cells with depolarizing concentrations of KClwere able to evoke a dose dependent increase in DNAsynthesis. The treatment with both PrP 106±126 andwith nicardipine reduced the e�ect of depolarization(Fig. 7). This evidence enforced the idea that the ac-tivity of the voltage gated calcium channels could playa pivotal role in the cell survival and proliferation andsuggested that a modulation of the L-type voltage-gated calcium channels activity may be involved, notonly in cell proliferation (astrogliosis), but also inGH3 cell death induced by PrP 106±126.

Micro¯uorimetric experiments supported the hy-pothesis that the alteration of the activity of the L-type voltage-gated calcium channel could transducethe cytotoxic e�ect of the PrP 106±126 and partiallyclari®ed the discrepancies between the two opposite

e�ects exerted by the prion fragment in astrocytes andGH3 cells (cell proliferation versus cell death). A KCl-induced depolarization pulse evokes, in the GH3 cells,a rapid and signi®cant increase of the cytosolic calciumions concentration. The preincubation of these cellswith PrP 106±126 strongly reduced the KCl-inducedcalcium rise and a similar preincubation with nicardi-pine abolished almost completely the above-mentionedcalcium rise (Fig. 8). Recently, the hypothesis of adirect interaction between the PrP 106±126 and the L-type VSCC was further supported by binding and elec-trophysiological studies. In cortical astrocyte mem-branes it was reported that PrP 106±126 dose-dependently displaced the [3H]-isradipine binding [16],while by means of a patch clamp in whole-cell exper-iments, it was reported that depolarizing stimuli wereable to evoke a calcium current in the GH3 cells withthe characteristics of an L-type VSCC-dependent cur-rent and that the pretreatment with PrP 106±126reduced this current. This e�ect was mimicked, quali-

Fig. 6. PrP 106±126 treatment induces apoptosis in GH3 cells, as evidenced by the nuclear fragmentation and condensation. After ®ve days of

treatment, the cells were ®xed in paraformaldehyde and stained in the nuclei with the ¯uorescent dye bis-benzimide. Panel A: control; panel C:

PrP 106±126 (50 mM); panels B and D: phase-contrast microphotographs of A and C, respectively.

S. Thellung et al. / Int. J. Devl Neuroscience 18 (2000) 481±492 489

tatively and quantitatively, by the selective L-type

VSCC antagonist nicardipine [17].

Taken together, the data obtained from the cellular

models we studied showed that PrP 106±126 was able

to interact with the L-type VSCC and regulate the

critical role of these channels in the calcium tra�cking.In astrocytes, a dihydropyridine-sensitive increase in[Ca++]i, correlated with a proliferative activity of thepeptide and, conversely, a toxic e�ect due to the block-ade of the same channels, was observed in excitablecell types. Thus, it seems likely that L-type VSCC rep-resents one of the main e�ector systems regulated byPrP 106±126, although the kind of interaction (inhi-bition or activation) and the ®nal e�ect at cellular level(apoptosis or cell proliferation) may depend on the celltype and may account for the opposite e�ects observedin prion diseases on the di�erent brain cell populations(neuronal cell death and glial proliferation). Moreover,we cannot rule out the possibility that these di�erente�ects may also be due to the unique characteristics ofthe astrocyte L-type VSCC which have been reportedto possess activation/inactivation features di�erentfrom those of other cell types [3].

Acknowledgements

The ®nancial support by Telethon (Grant E.394),CNR (strategic project 9604979.ST4) Biotech MURSTCNR 5% 95/95 to G. Schettini and by CNR(97.04562.Ct04) and the University of Chieti (FondiAteneo ex 60%, 1999) to T. Florio are gratefullyacknowledged.

Fig. 7. DNA neosynthesis stimulated by depolarizing conditions in GH3 cells: inhibition by PrP 106±126 and nicardipine. Cells were serum

deprived for 48 h and then treated, for 24 h, with a serum free culture medium containing increasing concentrations of KCl, KCl plus PrP 106±

126 (50 mM) and KCl plus nicardipine (1 mM). Nicardipine abolishes and PrP 106±126 reduces the mitogen action of the KCl. Each data is

obtained as cpm/well and plotted as a percentage of the DNA neosynthesis in unstimulated conditions. ��=p< 0.01 vs control; =p<0.01

vs KCl-stimulated DNA neosynthesis.

Fig. 8. Single cell micro¯uorimetry: the KCl-induced calcium rise in

the GH3 cells is signi®cantly reduced by the PrP 106±126 and nicar-

dipine. Cells were preincubated for 1 h with PrP 106±126 (50 mM),

PrP scrambled (50 mM) and nicardipine (1 mM). After the incubation

time, they were stimulated with a depolarizing pulse (KCl 50 mM)

and the peak of the calcium rise was estimated. The preincubation

with the scrambled peptide did not induce any perturbation on the

calcium rise. Each column is obtained from three di�erent exper-

iments where an average of 10 cells was evaluated. =p < 0.01

vs basal; ��=p<0.001 vs the calcium peak in the untreated cells.

S. Thellung et al. / Int. J. Devl Neuroscience 18 (2000) 481±492490

References

[1] Alpers, M. In Prions-Novel Infectious Pathogens Causing Scrapie

and Creutzfeldt±Jacob Disease, eds S. B. Prusiner and M. P.

McKinley. Academic Press, Orlando, FL, 1987, pp. 451±465.

[2] Aguzzi, A. and Waissmann, C., Prion disease. Haemophilia,

1998, 4, 619±627.

[3] Barres, B. A., Chun, L. L. Y. and Corey, D. P., Calcium cur-

rent in cortical astrocytes: induction by cAMP and neurotrans-

mitters and permissive e�ect of serum factors. J. Neurosci.,

1989, 9, 3169±3175.

[4] Barres, B. A., Chun, L. L. Y. and Corey, D. P., Ion channels in

vertebrate glia. Annu. Rev. Neurosci., 1990, 13, 441±474.

[5] Brown, D. R., Schmidt, B. and Kretzschmar, H. A., Role of

microglia and host prion protein in neurotoxicity of prion pro-

tein fragment. Nature, 1996, 380, 345±347.

[6] Brown, P., The ``brave new world'' transmissible spongiform

encephalopathy (infectious cerebral amyoidosis). Mol.

Neurobiol., 1994, 8, 79±88.

[7] Brown, D. R., Qin, K., Herms, J. W., Madlung, A., Manson,

J., Strome, R., Fraser, P. E., Kruck, T., Von-Bohlen, A.,

Schuls-Shae�er, W., Giese, A., Westway, D. and Kretzshmar,

H., The cellular prion protein binds copper in vivo. Nature,

1997, 390, 684±687.

[8] Bueler, H., Aguzzi, A., Sailer, A., Greiner, R. A., Autenried, P.,

Aguet, M. and Weissmann, C., Mice devoid of PrP are resistant

to scrapie. Cell, 1993, 73, 1339±1347.

[9] Chazot, G., Broussolle, E., Lapras, C. L., Blatter, T., Aguzzi,

A. and Kopp, N., New variant of Creutzfeldt±Jacob disease in

a 26-year-old French man. Lancet, 1996, 347, 1181.

[10] Chen, S. G., Teplow, D. B., Parchi, P., Teller, J. K. and

Gambetti, P., Truncated forms of the human prion protein in

normal brain and in prion diseases. J. Biol. Chem., 1995, 270,

19,173±19,180.

[11] Chomczinsky, P. and Sacchi, N., Single step method of RNA

isolation by acid guanidium thiocyanate-phenol-chloroform

extraction. Anal. Biochem., 1987, 162, 156±159.

[12] Collinge, J., Whitthingon, M. A., Sidle, K. C. L., Smith, C. J.,

Palmer, M. S., Clarke, A. R. and Je�erys, J. G. R., Prion pro-

tein is necessary for normal synaptic function. Nature, 1994,

370, 295±297.

[13] De Armond, S. J., Mobley, W. C., De Morr, D. L., Barry, R.

A. and Beckstead, J. H., Changes in the localization of brain

prion proteins during scrapie infection. Neurology, 1987, 37,

1271±1280.

[14] De Armond, S. J., Gonzales, M., Mobley, W. C., Kon, A. A.,

Sern, H. and Prusiner, S. B., PrPSc in scrapie-infected hamster

brain is spatially and temporally related to histopathology and

infectivity titer. In Alzheimer's Disease and Related Disorders,

eds K. Ikbal, K. M. Wisniewsky and B. Winblead, Progress in

Clinical Biology Research, vol. 317. Alan R. Liss, New York,

1988, pp. 601±618.

[15] Florio, T., Pan, M. G., Newman, B., Hershberger, R. E.,

Civelli, O. and Stork, P. J. S., Dopaminergic inhibition of DNA

synthesis in pituitary tumor cells is associated with phos-

phothyrosine phosphatase activity. J. Biol. Chem., 1992, 267,

24,169±24,172.

[16] Florio, T., Grimaldi, M., Scorziello, A., Salmona, M., Bugiani,

O., Tagliavini, F., Forloni, G. and Schettini, G., Intracellular

calcium rise through the L-type calcium channel, as molecular

mechanism for prion protein fragment 106±126-induced astro-

glial proliferation. Biochem. Biophys. Res. Commun., 1996, 228,

397±405.

[17] Florio, T., Thellung, S., Amico, C., Robello, M., Salmona, M.,

Bugiani, O., Tagliavini, F., Forloni, G. and Schettini, G., Prion

protein fragment 106±126 induces apoptotic cell death and

impairment of L-type voltage-sensitive calcium channel activity

in the GH3 cell line. J. Neurosci. Res., 1998, 54, 341±352.

[18] Forloni, G., Angeretti, N., Chiesa, R., Monzani, E., Salmona,

M., Bugiani, O. and Tagliavini, F., Neurotoxicity of a prion

protein fragment. Nature, 1993, 362, 543±546.

[19] Forloni, G., Del Bo, R., Angeretti, N., Chiesa, R., Smiroldo, S.,

Doni, R., Ghibaudi, R., Salmona, M., Porro, M., Verga, L.,

Giaccone, G., Bugiani, O. and Tagliavini, F., A neurotoxic

prion protein fragment induces rat astroglial proliferation and

hypertrophy. Eur. J. Neurosci., 1994, 6, 1415±1422.

[20] Galli, C., Meucci, O., Scorziello, A., Werge, T. M., Calissano,

P. and Schettini, G., Apoptosis in cerebellar granule cells is

blocked by high KCl, forskolin, and IGF-1 through distinct

mechanisms of action: the involvment of intracellular calcium

and RNA synthesis. J. Neurosci, 1995, 15, 1172±1179.

[21] Ghetti, B., Piccardo, P., Frangione, B., Bugiani, O., Giaccone,

G., Young, K., Prelli, F., Farlow, M. R., Dlouhy, S. R. and

Tagliavini, F., Prion protein amyloidosis. Brain Pathol., 1996, 6,

127±145.

[22] Grynkiewicz, G., Poenie, M. and Tsien, R. Y., A new gener-

ation of calcium indicators with greatly improved ¯uorescence

properties. J. Biol. Chem., 1985, 2600, 3440±3450.

[23] Hornemann, S. and Glockschuber, R., Autonomous and revers-

ible folding of a soluble amino-terminally truncated segment of

the mouse prion protein. J. Mol. Biol., 1996, 261, 614±619.

[24] Lindenboim, L., Haviv, R. and Stein, R., Inhibition of drug-

induced apoptosis by survival factors in PC12 cells. J.

Neurochem., 1995, 64, 1054±1063.

[25] Lugaresi, E., Medori, R., Montagna, P., Baruzzi, A., Cortelli,

P., Lugaresi, A., Tinuper, P., Zucconi, M. and Gambetti, P.,

Fatal familial insomnia and dysautonomia with selective de-

generation of thalamic nuclei. New Engl. J. Med., 1986, 315,

997±1003.

[26] MuÈ ller, W. E. G., Ushjima, H., Schroeder, H. C., Forrest, J. M.

S., Schatton, W. F. H., Rytik, P. G. and He�ener-Lauc, M.,

Cytoprotective e�ect of NMDA receptor antagonists on prion

protein-induced toxicity in rat cortical cell cultures. Eur. J.

Pharmacol., 1993, 246, 261±267.

[27] Naslavsky, N., Stein, R., Yanai, A., Friedlander, G. and

Tarabulos, A., Characterization of detergent-insoluble com-

plexes containing the cellular prion protein and its scrapie iso-

forms. J. Biol. Chem., 1997, 272, 6324±6331.

[28] Oesch, B., Westaway, D., Walchili, M., McKinley, M. P., Kent,

S. B. H., Aebersold, R., Barry, R. A., Tempst, P., Teplow, D.

B., Hood, L. E., Prusiner, S. B. and Weissman, C., A cellular

gene encodes scrapie PrP 27±30 protein. Cell, 1985, 40, 735±

746.

[29] Pan, K. M., Baldwin, M., Nguyen, J., Gasset, M., Serban, A.,

Groth, D., Mehlhorn, I., Huang, Z., Fletterick, R. J., Cohen, F.

E. and Prusiner, S. B., Conversion of a-elics into b-sheets fea-

tures in the formation of the scrapie prion proteins. Proc. Natl

Acad. Sci. USA, 1993, 90, 10,962±10,966.

[30] Prusiner, S. B., Novel proteinaceous infectious particles cause

scrapie. Science, 1982, 216, 136±144.

[31] Prusiner, S. B., McKinley, M., Bowman, K. A., Bendheim, P.

E., Bolton, D. C., Groth, D. F. and Glenner, G. G., Scrapie

prions aggregate to form amyloid-like birifragment rods. Cell,

1983, 35, 348±358.

[32] Prusiner, S. B., Chemistry and biology of prions. Biochemistry,

1992, 31, 12,277±12,288.

[33] Prusiner, S. B., Prions. Proc. Natl Acad. Sci. USA, 1998, 95,

13,363±13,383.

[34] Ramsdell, J. S., Voltage-dependent calcium channels regulate

GH4 pituitary cell proliferation at two stages of the cell cycle.

J. Cell Physiol., 1991, 146, 197±206.

[35] Riek, R., Hornemann, S., Wider, G., Glockshuber, R. and

Wuthrich, K., NMR characterization of the full-length recombi-

S. Thellung et al. / Int. J. Devl Neuroscience 18 (2000) 481±492 491

nant murine prion protein, mPrP (23±231). FEBS Letters, 1997,

413, 282±288.

[36] Rosenthal, W., Hescheler, J., Hinsch, K. D., Spicher, K.,

Trautwein, W. and Schultz, G., cAMP-independent dual regu-

lation of voltage dependent Ca++ currents by LHRH and so-

matostatin in a pituitary cell line. EMBO J., 1988, 7, 1627±

1633.

[37] Sakaguchi, S., Katamine, S., Nishida, N., Moriuchi, R.,

Shigematsu, K., Sugimoto, T., Nakatani, A., Kataoka, Y. and

Houtani, T., Loss of cerebellar Purkinje cells in aged mice

homozygous for a disrupted PrP gene. Nature, 1996, 380, 528±

531.

[38] Schettini, G., Meucci, O., Grimaldi, M., Florio, T., Landol®,

E., Scorziello, A. and Ventra, C., Dihydrophyridine modulation

of voltage-activated calcium channels: e�ect of pertussis toxin

pretreatment. J. Neurochem., 1991, 56, 805±811.

[39] Scorziello, A., Florio, T., Bajetto, A., Thellung, S. and

Schettini, G., TGF-b1 prevents gp120-induced impairment of

Ca++ homeostasis and rescues cortical neurons from apoptotic

death. J. Neurosci. Res., 1997, 49, 600±607.

[40] Selvaggini, C., De Gioia, L., Cantu', L., Di Baudi, E.,

Diomede, L., Passerini, F., Forloni, G., Bugiani, O., Tagliavini,

F. and Salmona, M., Molecular characteristics of a protease re-

sistant amyloidogenic and neurotoxic peptide homologous to

residue 106±126 of the prion protein. Biochem. Biophys. Res.

Commun., 1993, 194, 1380±1386.

[41] Stahl, N., Baldwin, M. A., Teplow, D. B., Hood, L., Gibson, B.

V., Burlingame, A. L. and Prusiner, S. B., Structural studies of

the scrapie prion protein using mass spectrometry and amino

acid sequencing. Biochemistry, 1993, 32, 1991±2001.

[42] Tagliavini, F., Prelli, F., Ghiso, J., Bugiani, O., Serban, D.,

Prusiner, S. B., Parlow, M., Ghetti, B. and Frangione, B.,

Amyloid protein of Gerstmann±Straussler±Scheinker disease

(Indiana kindred) is an 11kd fragment of prion protein with an

N-terminal glycine at codon 58. EMBO J., 1991, 10, 513±519.

[43] Tagliavini, F., Prelli, F., Verga, L., Giaccone, G., Sarma, R.,

Gorevic, P., Ghetti, B., Passerini, F., Ghibaudi, E., Forloni, G.

L., Salmona, M., Bugiani, A. and Frangione, B., Synthetic pep-

tides homologous to prion protein residues 106±147 form amy-

loid-like ®brils in vitro. Proc. Natl Acad. Sci. USA, 1993, 90,

9678±9682.

[44] Taraboulos, A., Scott, M., Semenov, A., Avrahami, D., Laszlo,

L. and Prusiner, S. B., Cholesterol depletion and modi®cation

of COOH-terminal targeting sequence of the prion protein inhi-

bit formation of the scrapie isoform. J. Cell Biol., 1995, 129,

121±132.

[45] Telling, G. C., Scott, M., Mastrianni, J., Gabizon, R., Torchia,

M., Cohen, F. E., DeArmond, S. J. and Prusiner, S. B., Prion

propagation in mice expressing human and chimeric PrP trans-

genes implicates the interaction of cellular PrP with another

protein. Cell, 1995, 83, 79±90.

[46] Vey, M., Pilkhun, S., Wille, H., Nixon, R., DeArmond, S. J.,

Smart, E. J., Anderson, R. G. W., Tarabulos, A. and Prusiner,

S. B., Subcellular colocalization of the cellular and scrapie prion

proteins in caveole-like membranous domains. Proc. Natl Acad.

Sci. USA, 1996, 93, 14,945±14,949.

S. Thellung et al. / Int. J. Devl Neuroscience 18 (2000) 481±492492