Experimental Neurology 233 (2012) 182–192

Contents lists available at SciVerse ScienceDirect

Experimental Neurology

j ourna l homepage: www.e lsev ie r .com/ locate /yexnr

Calpain activation is involved in acute manganese neurotoxicity in therat striatum in vivo

Liliana Quintanar a,⁎, Teresa Montiel b, Maripaz Márquez a, Alejandra González a, Lourdes Massieu b,⁎a Departamento de Química, Centro de Investigación y de Estudios Avanzados (Cinvestav), México D.F., Méxicob Departamento de Neuropatología Molecular, División de Neurociencias, Instituto de Fisiología Celular, UNAM, México D.F., México

Abbreviations: ER, endoplasmic reticulum; DEVD, Ac-FJ-B, Fluoro Jade-B; GABA,γ-aminobutyric acid;GFAP, gliaglial glutamate–aspartate transporter; GLT-1, glutamateMK-801, (+)-5-methyl-10,11,-dihydro-5H-dibenzo[a,d]cyNMDA, N-methyl-D-aspartate; NeuN, neuronal nuclei; QPh; SDS, sodium dodecyl sulfate.⁎ Corresponding authors at: Departamento de Neur

de Neurociencias. Instituto de Fisiología Celular, UniveMéxico, México D.F. CP 04510, México. Fax: +52 55 55

E-mail addresses: lilianaq@cinvestav.mx (L. Quintan(L. Massieu).

0014-4886/$ – see front matter © 2011 Elsevier Inc. Alldoi:10.1016/j.expneurol.2011.09.032

a b s t r a c t

a r t i c l e i n f o

Article history:Received 19 May 2011Revised 25 August 2011Accepted 26 September 2011Available online 1 October 2011

Keywords:Calpain activationCaspase activityExcitotoxicityManganese neurotoxicityManganism

Manganese is essential for life, yet chronic exposure to this metal can cause a neurodegenerative disease namedmanganism that affectsmotor function. In the present studywehave evaluatedMnneurotoxicity after its admin-istration in the rat striatum. The participation of the calcium-dependent protease calpain and the apoptosis-related protease caspase-3, in Mn-induced cell death was monitored in the striatum and globus pallidus. Mninduced the activation of both proteases, although calpain activation seems to be an earlier event. Moreover,while the broad-spectrum caspase inhibitor QVDdid not significantly preventMn-induced cell death, the specificcalpain inhibitor MDL-28170 did. The role of NMDA glutamate receptors on calpain activity was also investigat-ed; blockage of these receptors by MK-801 and memantine did not prevent calpain activation, nor Mn-inducedcell death. Finally, studies in striatal homogenates suggest a direct activation of calpain by Mn ions. Altogetherthe present study suggests that additional mechanisms to excitotoxicity are involved in Mn-induced celldeath, placing calpain as an important mediator of acute Mn neurotoxicity in vivo.

Asp-Glu-Val-Asp-H (aldehyde);lfibrillary acidic protein; GLAST,transporter-1; Glu, glutamate;clohepten-5,10-imine maleate;VD, Quinoline-Val-Asp-CH2-O-

opatología Molecular, Divisiónrsidad Nacional Autónoma de22 56 07.ar), lmassieu@ifc.unam.mx

rights reserved.

© 2011 Elsevier Inc. All rights reserved.

Introduction

Manganese (Mn) is an essentialmetal ion, as it is a cofactor for severalenzymes, including glutamine synthetase in the brain. However, chronicexposure to Mn causes a neurodegenerative disease named “mangan-ism”, with symptoms that resemble Parkinson's disease (Olanow,2004). Manganism is characterized by motor defects such as akineticrigidity, dystonia and bradykinesia, aswell as early impairment of speech,gait and balance. Although some of these symptoms resemble thoseobserved for idiopathic Parkinson's disease (PD), the characteristic trem-or of PD patients is less common in individuals suffering frommangan-ism and they are not responsive to L-DOPA therapy (Olanow, 2004).However, while manganism and idiopathic Parkinson's disease displayseveral similarities at the molecular and clinical levels (Aschner et al.,2009; Benedetto et al., 2009; Roth, 2009), these neurodegenerativediseases can be distinguished by the brain structures that are damaged

in each case. Parkinson's disease is characterized by the degeneration ofdopaminergic neurons in the substantia nigra pars compacta, the loss ofdopaminergic terminals innervating the striatum, and the presence ofintracytoplasmic Lewy bodies (Braak et al., 2003; Olanow, 2004). Incontrast, chronic Mn exposure leads to Mn accumulation and neuronalloss in the globus pallidus, striatum, and substantia nigra pars reticulata(Aschner et al., 2009; Benedetto et al., 2009; Bonilla et al., 1982; Erikssonet al., 1987; Olanow, 2004; Olanow et al., 1996; Yamada et al., 1986).

Mnneurotoxicity has been associatedwithmitochondrial dysfunction(Tamm et al., 2008; Zhang et al., 2003). Mn inhibits oxidative phosphor-ylation andmitochondrial complex I in PC12 cells (Galvani et al., 1995); itdisrupts the mitochondrial membrane potential in cultured astrocytes(Gonzalez et al., 2008; Yin et al., 2008) and striatal neurons (Malecki,2001); and in vivo studies in rats have also shown that Mn intoxicationleads to the inhibition of respiratory chain complexes, increases mito-chondrial reactive oxygen species production and decreases monoamineoxidase activity (Zhang et al., 2003). Thus,Mn interferencewith oxidativephosphorylation and inhibition of mitochondrial activity could lead to adecline in ATP levels, compromising the energy metabolism of the cell.

Apoptosis has been suggested as amechanism forMn-neurotoxicity.DNA fragmentation and caspase activation have been observed afterMnexposure in vitro (Chun et al., 2001; Hirata, 2002). A mitochondrialapoptotic pathway induced byMn has been implicated in astrocyte cor-tical cultures, involving cytochrome c release, caspase activation, PARP-1 cleavage and increased levels of Bax (Gonzalez et al., 2008; Yin et al.,2008), suggesting damage to glial cells. Although the induction ofthese apoptotic markers by Mn has not been demonstrated in the ratbrain in vivo, the contribution of inflammation to neurodegeneration

183L. Quintanar et al. / Experimental Neurology 233 (2012) 182–192

in manganismhas been supported by the production of nitric oxide andinterleukins by activated glia and microglia (Chen et al., 2006; Liu et al.,2006; Zhao et al., 2009).

Mn also alters the levels of neurotransmitters like dopamine, gluta-mate (Glu), and γ-aminobutyric acid (GABA) (Fitsanakis et al., 2006);and it diminishes D-aspartate uptake and expression of the glial gluta-mate transporters, glutamate transporter-1 (GLT-1) and glutamate–aspartate transporter (GLAST) (Erikson et al., 2007); suggesting thatMn promotes an excitotoxic mechanism of neuronal death in the stria-tum. Moreover, exposure to high levels of Glu in the presence of inhib-itors of mitochondrial energy metabolism, make striatal neurons moresusceptible to Glu neurotoxicity (Del Río et al., 2008). Thus, Mn neuro-toxicity might result from the combination of the impairment of mito-chondrial metabolism and altered glutamatergic neurotransmission.The N-methyl-D-aspartate (NMDA) Glu receptor subtype is mainlyinvolved in Glu toxicity (Choi et al., 1988). Calcium influx through thisreceptor can activate several pathways including calcium-dependentproteases like calpain, which contribute to the degradation of cellularcomponents and to excitotoxic death (Araújo et al., 2010; Bizat et al.,2003; Del Rio et al., 2008; Vosler et al., 2008). This type of death hasbeen implicated in Mn neurotoxicity, based on early studies showingprotection against Mn-induced striatal damage by the administrationof the NMDA receptor antagonist, (+)-5-methyl-10,11,-dihydro-5H-dibenzo[a,d]cyclohepten-5,10-imine maleate (MK-801) (Brouillet etal., 1993).

The mechanisms of Mn neurotoxicity in vivo have not been fullyelucidated. The main interest of the present study was to investigate,in an in vivo model, the participation of cell death-related proteasessuch as calpain and caspase-3 in Mn-induced cell death. While cas-pase-3 is a well-known executioner of apoptosis, calpain activity hasbeen related to excitotoxicity (Araújo et al., 2010; Wang, 2000). Therelation of calpain activity to the activation of NMDA receptors wasalso studied. Finally, studies in striatal homogenates were performedto evaluate the possibility of a direct activation of calpain by Mn ions.

Material and methods

Male Wistar rats (250–300 g) were used throughout the study, andwere handled according to theNational Institute of Health Guide for theCare and Use of Laboratory Animals (NIH Publications No. 80-23)revised 1996, and to the Rules for Research in Health Matters (México).The local Animal Care Committee approved all animal treatments. Ani-mals were housed under standard conditions (12 h light cycle), withfree access to food and water. All efforts were made to minimize thenumber of animals used and their suffering.

Mn administration

Mn-treated animals received an intrastriatal injection of MnCl2(Sigma Chemical Co, St. Louis MO, USA). In order to establish a dose ofMn that produces a lesion of an appropriate size for protection experi-ments, doses of 25, 50, 100 and 250 nmol were tested (SupplementalFig. 1). An injection of 100 nmol of Mn was considered appropriate.Control animals received saline solution injections instead of Mn.Animals were anesthetized with 2.0% halothane in a 95% O2/5% CO2

mixture and placed on a stereotaxic frame (David Kopf Instruments)with the nose bar positioned at −3.3 as previously described (Del Ríoet al., 2008). Coordinates used were: −0.4 mm anterior from bregma,2.8 mm lateral from midline, and 5.0 mm ventral from dura in orderto inject in the dorsal striatumand globus pallidus. Ratsweremaintainedunder low anesthesia (0.5% halothane) throughout the injection, and3 min after the injection was completed, the needle was withdrawnand the skin was sutured. Some animals were identically injected butinstead of Mn they received an injection of 100 nmol of CaCl2 or MaCl2.

Administration of MK-801, memantine, MDL-28170 and caspaseinhibitors

Rats treated with one dose of MK-801 received one ip injection ofMK-801 (2 mg/kg, Tocris, Ellisville, MO, USA) 30 min before the intra-striatal injection of Mn. Rats treatedwith two doses of MK-801 receivedone ip injection of MK-801 (2 mg/kg) 30 min before the intrastriatalinjection of 100 nmol of Mn, and a second ip injection of MK-801(2 mg/kg) 30 min after the Mn administration. Animals were sacrificed16 h after the intrastriatal injection ofMn to obtain samples forWesternblot analysis, and at 24 h to obtain samples for histological evaluation.Memantine (30 mg/kg, Tocris, Ellisville, MO, USA) was ip administered1 h before Mn intrastriatal injection and animals were sacrificed 24 hlater for histological procedures.

Rats received a total of 50 mg/kg of the calpain inhibitor, MDL-28170(Biomol, Plymouth Meeting, PA, USA), divided into three ip injections:one 20 mg/kg dose was administered 15 min before the intrastriatal in-jection of Mn, a second dose of 15 mg/kg 45 min after theMn intrastria-tal injection, and a third dose of 15 mg/kg 2 h after the Mn injection.Animals were either sacrificed 16 h after the intrastriatal administrationof Mn for Western blot analysis, or at 24 h for histological evaluation.

Two different groups of animals received an intrastriatal co-injec-tion of Mn and either the caspase-3 inhibitor, DEVD (Ac-Asp-Glu-Val-Asp-CHO, Peptide Institute, Inc. Osaka, Japan) Peptide Institute, Inc.Osaka, Japan), or the general caspase inhibitor QVD (Quinoline-Val-Asp-CH2-O-Ph, Enzyme System Products, Ohio, USA). Stock solutionsof DEVD (10 mM) or QVD (20 mM) were prepared in DMSO and analiquot of each inhibitor was mixed with the Mn stock solution toreach a final concentration of 2.5 mM. One μl of this solution wasintrastriatally injected as described above. Brains were extracted ei-ther 8 or 16 h after the intrastriatal injection, and tissues were pro-cessed for Western blot analysis. An additional series of animals wasco-injected with Mn and QVD and brains were processed for histolog-ical analysis 24 h after the intrastriatal injection.

Histological evaluation

To investigate the time-course of the development of the Mn-induced lesion induced, brains were extracted under deep anesthesia atdifferent times after Mn injection (4, 8 and 24 h). For protection experi-ments, animals were sacrificed 24 h after the different treatments. Ratswere anesthetized with sodium pentobarbital anesthesia and transcar-dially perfused with 200 ml 0.9% saline followed by 200 ml 4% parafor-maldehyde in 0.1 M phosphate buffer (pH 7.3). Brains were removedand left in fixative for additional 24 h, then transferred successively to20% and 30% sucrose (24 h each). 40 μm coronal sections were cut in acryostat and stained with Cresyl Violet. Lesion size was calculated byexamination of all brain sections where neuronal damage was evidentin each experimental animal. Tissue was considered damaged whenintensively stained pyknotic nuclei were present and only few or nonenormal appearing cells were visible. Damaged area in each tissue sectionwasmanually delineated andmeasuredwith the aid of an image analyzer(NIH Macintosh Image 1.6). The lesion volume was calculated by addingthe measured areas in all sections, and multiplying the sum by thedistance between the first and the last section,where damagewas visibleas previously reported (Del Rio andMassieu, 2008). Results are expressedas mean±S.E.M. of the lesion volume per each animal group.

Fluoro Jade-B and immunohistochemistry experiments

Fluoro Jade-B (FJ-B, Chemicon Temecula, CA, USA) staining was per-formed in animals sacrificed 4 h after the intra-striatal injection of Mn.Briefly, slide-mounted coronal brain sections (20 μm) were dried at50 °C for 30 min, ethanol solution (80% and 70%) was added and incu-bated for 5 and 2 min respectively, the slides were washed and coveredwith potassium permanganate 0.06% for 10 min, immediately washed

184 L. Quintanar et al. / Experimental Neurology 233 (2012) 182–192

and incubated with 0.0004% FJ-B solution dissolved in acetic acid 0.1%,for 20 min. Finally, sections were washed with water and dried at50 °C. The sections were bleached with xylene during 2 min and cov-ered with permount. Sections were observed under an Olympus epi-fluorescence microscope (using UMNB2 filter), and cells positive to FJ-B staining were quantified using the Image J analyzer program. Thetotal number of positive cells was counted in the striatum from threecoronal sections per rat: the section containing the needle tract (middlesection), and the adjacent anterior and posterior sections. Some sec-tions were double stained for FJ-B and the neuron-specific nuclear pro-tein named NeuN (Neuronal Nuclei) or the glial fibrillary acidic protein(GFAP). For double-staining, the sections were incubatedwithmonoclo-nal antibody NeuN (1:300, Chemicon Temecula, CA, USA) or polyclonalantibody GFAP (1:500, Dako, Denmark) for 72 h at 4 °C, they werewashed and incubatedwith goat anti-mouse IgGDyLight 594 conjugated(1:200, Jackson ImmunoResearch Laboratories) or antirabbit IgG Texas-Red conjugated (1:200, Dako, Denmark), for 2 h at room temperature.After washing, sections were covered with potassium permanganate0.06% for 5 min, they were washed again and incubated with 0.0001%FJ-B for 20 min. Finally, sections were washed, covered with fluoro-mount-G, and observed by confocal microscopy as described below.

For double immunohistochemistry (NeuN and GFAP) experiments,sections were incubated together with the primary antibodies (NeuN1:300 Chemicon, and GFAP 1:500, Dako) for 72 h at 4 °C, they werewashed and incubated with goat anti-mouse IgG DyLight-594 conju-gate (1:200 Jackson ImmunoResearch Laboratories, CA, USA) or goatanti-Rabbit IgG FITC conjugate (1:200 Zymed-Invitrogen, Carlsbad, CA,USA). The slices were analyzed by confocal laser scanning microscopy(FV 1000) with 40× plan FLN-NA 1.3 objective, motorized FV10ASW2.1, with Ar 488 laser (for FJB and FITC) and HeNe 543 laser (forDyLight-594 and Texas Red).

Western blotting

For Western blot analysis, striata were dissected out from animalsdeeply anesthetized at different times after Mn administration. Striatawere homogenized using a 1 ml glass homogenizer (70 rpm, 5 strokes),into 300 μl of 25 mM HEPES buffer at pH 7.4, containing 5 mM MgCl2,1.3 mMEDTA, 1 mMEGTA, Triton 0.1%, and a protease inhibitor cocktail(Complete EDTA free, Roche) at a concentration of 5 mg/ml. Homoge-nates were centrifuged at 10,000 rpm (14,000 g) for 30 min. The super-natant fraction was aliquoted and stored at −80 °C until further use.Protein concentration for each sample was determined using the stan-dard Bradford assay. Equal amounts of protein (30 μg) from striatal ex-tracts were loaded on 7% acrylamide gels, and separated by SDS-PAGE.The separated proteins were blotted onto nitrocellulose membranesthat were cut to separate the high molecular weight proteins from thelowermolecular weight ones (below 50 kDa). Themembranes were in-cubated overnight at 4 °C with an antibody against fodrin (1:7500, nonerythroidα-spectrin, Chemicon) for the highmolecular weight portion,and actin (1:12,000, Chemicon) for the portion below 50 kDa. Mem-branes were incubated with horseradish-peroxidase-conjugated anti-mouse antibody (1:12,500, Immunoresearch). Peroxidase activity wasdetected using enhanced chemiluminescence (ECL) reagent (Milli-pore). Membranes were placed against X-ray films (Kodak) to obtainautoradiographs, which were then digitalized and quantified usingimage processing software (NIH Image J 1.35s). For all experimentalgroups, the mean optical density of all studied bands of each samplewas determined by this procedure at least five times. Results arereported as the relative integrated intensity of the band of interestdivided by the intensity of the corresponding actin band.

Calpain activity assay

Calpain activity was determined using the fluorogenic substrateN-Succinyl-Leu-Tyr-7-amido-4-methylcoumarin (N-succinyl-Leu-

Tyr-AMC, Sigma). Brains were extracted from animals sacrificed at dif-ferent times after Mn injection (from 2, 4, 8, 12, 16 and 24 h). Brainhomogenates (100 μg of protein)were pre-incubated at 37 °C in a buffercontaining 63 mM imidazol-HCl, pH 7.3, 10 mMβ-mercaptoethanol and5 mM CaCl2, for 10 min. The fluorogenic substrate (200 μM) was addedand the change in fluorescence (380 nm excitation/460 nm emission)was monitored for 30 min using a 1 cm pathlength cell and an YvonFluoromax fluorimeter. The non-calcium-dependent fluorescence wasmeasured under the same conditions using a buffer that does not con-tain calcium, but containing 1 mM EDTA and 10 mM EGTA. Calcium-dependent cleavage activity towards N-succinyl-Leu-Tyr-AMC isreported as a change in fluorescence per min per mg of protein. Inorder to corroborate that the measured fluorescence corresponds tosubstrate cleavage by calpain, control experiments were performedusing the specific calpain inhibitor MDL-28170, which was added tothe incubation buffer at a 5 μM concentration. For the evaluation of cal-pain activation by other divalent metal ions, brain homogenates fromhealthy rats were used, and the incubation buffer contained the chloridesalt of the correspondingmetal ion (Mn2+, Zn2+, Mg2+, Pb2+ orMn2+)at a 5 mM concentration.

Caspase-3 activity assay

For caspase-3 activity determination, tissues were obtained fromMn injected animals 8 and 16 h after the intrastriatal administration.They were homogenized in buffer containing: 100 mM HEPES (pH7.5), 1 mM EDTA, 0.02 mM EGTA, 12.5% sucrose, 1% triton X-100,10 mM DTT, 20% glycerol and protease inhibitors cocktail (CompleteEDTA free). Caspase-3 activity was determined in a Microplate Reader(Synergy HT, Biotek Instruments, Inc., USA) by a fluorometric methodusing the tetrapeptide acetyl-Asp-Glu-Val-Asp-(4-methylcoumaryl-7-amide) (Ac-DEVD-AMC, Peptide Institute Inc., Osaka, Japan) as sub-strate. The change in fluorescence emission (380 nmexcitation/460 nmemission) was monitored for 25 min after the addition of 100 μl ofhomogenate sample (500 μg protein) and 100 μl of the assay buffer con-taining: 20 mMHEPES (pH 7.5), 2 mMDTT, and caspase-3 substrate Ac-DEVD-AMC (20 μM final concentration). Some samples were incubatedin the presence of the caspase-3 inhibitor, DEVD (Ac-Asp-Glu-Val-Asp-CHO, 40 μM final concentration). Results are expressed as the change influorescence per min per mg of protein.

Statistical analysis

Data are expressed asmean values±SD or SEM. Data were analyzedby one-wayANOVA followed by a Fisher's, Bonferroni/Dunnor Scheffé FPost-hoc tests, as indicated in each figure caption.

Results

Mn-induced striatal damage

Different concentrations of Mn (25 nmol to 250 nmol) were intras-triatally injected and brains were extracted 24 h later for histologicalanalysis. While a 25 nmol Mn dose did not cause a visible lesion,50 nmol produced small lesions of variable size. On the other hand,100 and 150 nmol Mn yielded lesions of similar size, and 250 nmol Mnproduced very extensive lesions (Supplementary Fig. 1). Thus, it wasdetermined that the dose of 100 nmol of Mn in 1 μl yielded a reasonablysized lesion that could be appropriate for mechanistic studies. It shouldbe noted that the administration of other divalent metal ions (Ca andMg) at the same concentrationwas also tested.While Ca induced striatallesions of a similar extension to that induced by Mn (Ca=5.82±0.54 mm3, n=3; Mn=6.88±0.93 mm3, n=5), Mg-induced lesionswere significantly smaller than those induced by Mn (Mg=4.19±0.54 mm3, n=5) at 8 h after the metal ion administration. A

185L. Quintanar et al. / Experimental Neurology 233 (2012) 182–192

comparison of the effects of Ca,Mg andMnwill be addressed in the Dis-cussion section.

Intra-striatal injections of 100 nmol ofMn caused a distinct lesion inthe striatum and globus pallidus (Figs. 1C–D), while injections of salinecaused only a discrete lesion produced by the needle tract (Fig. 1A).Damage to striatal cells is apparent 4 h after the injection of Mn in thedorso-lateral striatum and the globus pallidus, as indicated by the pres-ence of polymorphic cells (black arrows in Fig. 1B); while a decrease inthe cell population becomes evident and the lesion is clearly defined 8 hafter the injection of Mn, as revealed by the lack of round and clearsomata and the presence of polymorphic cells and pyknotic nuclei(black and white arrows, respectively in Fig. 1C). Finally, 24 h afterMn administration the tissue looks severely damaged as indicated bya marked decrease in Nissl staining, the reduction of the number ofcells and the presence of numerous pyknotic nuclei (white arrows,Fig. 1D). The volume of lesion amounts to 12.7±1.8 mm3 at 24 h afterMn injection (Fig. 1E). The presence of damaged cells was corroboratedby FJ-B staining. Degenerating cells positive to FJ-B are already presentat 4 h after Mn intrastriatal administration in the damaged area aroundthe needle tract (Fig. 2A).While saline-injected animals show a discretenumber of degenerating neurons around the needle tract, inMn-treatedrats the number of FJ-B-positive neurons is significantly increased (sa-line=95.3±17.6, n=3; Mn=937.0±149.2, n=3, p≤0.005). Tissuesections processed for FJ-B staining and immunohistochemistry forthe neuronal marker, NeuN, reveal that almost every FJ-B-positivecells is also NeuN-labeled (arrows in Fig. 2B), suggesting that neuronalcells are degenerating at this stage.

In order to evaluate the presence of astrogliosis and the integrity ofastrocytes afterMn exposure, tissue sectionswere processed for immu-nohistochemistry using antibodies against the glial specific proteinGFAP and NeuN. As shown in Fig. 2C, in tissue areas where manyNeuN-positive cells are present, cells immunoreactive to GFAP show anormal appearance with well-defined soma and processes (Fig. 2C,right panel). Meanwhile, in damaged areas where there are no NeuN-

Fig. 1. Nissl staining of the striata of rats after the intrastriatal injection of vehicle (saline soluMn. Asterisks indicate the tissue areas that are magnified in the smaller panels. Black arrowshown in (E), data are expressed as mean±SEM; *Pb0.01 compared to control vehicle groupsaline group and 4 h after Mn injection, using ANOVA post hoc Fisher's PLSD, Scheffé F and B

positive cells and neuronal loss is evident, many cells are intensivelylabeled for GFAP, suggesting astroglial reactivity (Fig. 2C, left panel).In the case of saline-injected animals many NeuN-positive cells are pre-sent in the entire striatum and only a modest reactive astrogliosis isobserved in the vicinity of the needle tract (Supplementary Fig. 2). Toinvestigate if GFAP-labeled cells also degenerate, immunohistochemis-try for GFAP and co-staining with FJ-B was performed (SupplementaryFig. 3). While most of the FJ-B-positive cells are located in the lesioncore, most of the GFAP-labeled cells are found in the periphery of thelesion, indicating an important astroglial reaction (Supplementary Fig.3A). No FJ-B-positive cells immunoreactive to GFAP were detected,while cells immunopositive to GFAP were not labeled with FJ-B (Sup-plementary Fig. 3B).

Calpain and caspase-3 activation is induced by acute Mn-treatment

To further characterize the mechanism of Mn-induced cell death inthe acute Mn exposure model in vivo, the activation of proteases likecaspase-3 and calpain was evaluated. Striatal homogenates obtainedfrom rats at different times after the intrastriatal injection of 100 nmolof Mn were analyzed by SDS-PAGE. Western blot analysis for nonerythroid α-spectrin (fodrin) of tissue homogenates was indicative ofcalpain and caspase-3 activation (Fig. 3A). Spectrin is a known substratefor caspase-3 and calpain (Wang, 2000). Calpain-mediated cleavage ofspectrin yields 145 and 150 kDa fragments, while caspase-3-mediatedcleavage leads to 150 and 120 kDa breakdown products. Western blotanalysis for spectrin of striatal homogenates obtained from rats at dif-ferent times after Mn injection displays a band above 250 kDa that cor-responds to full spectrin (molecular weight 295 kDa), which is presentin all samples, including the saline-treated control. A distinct doublet ofbands at 145/150 kDa is evident starting at 4 h after Mn treatment, cor-responding to calpain-mediated breakdown products of spectrin(Fig. 3A), while no doublet bands are observed in the saline-treatedcontrol samples. The intensity of the 145/150 kDa doublet signal

tion) (A), and after 4 h (B), 8 h (C), and 24 h (D) of an intrastriatal injection of 100 nmols indicate polymorphic cells and white arrows pyknotic nuclei. The lesion volumes are, using ANOVA post hoc Fisher's PLSD, **Pb0.0001 compared to lesion volume in controlonferroni/Dunn tests; n=5 for vehicle; n=5 for 4 h; n=3 for 8 h and n=6 for 24 h.

Fig. 2. FJ-B stained sections from saline andMn-injected animals at 4 h, as analyzed by epifluorescencemicroscopy (A). Asterisks indicate the tissue areas that aremagnified in the smallerpanels. Immunohistochemistry for NeuN (red) and co-stainingwith FJ-B (green) 4 h afterMn injection, as analyzed by confocalmicroscopy (B). Arrows indicate FJ-B positive cells that arealso immunoreactive to NeuN. Double immunohistochemistry for NeuN (in red) and GFAP (in green) 4 h afterMn injection, as analyzed by confocal microscopy (C). Asterisks in the cen-tral panel indicate the tissue areas that are magnified in the left and right panels. Images are representative of 3 saline and 3 Mn injected animals.

186 L. Quintanar et al. / Experimental Neurology 233 (2012) 182–192

becomes significant 8 h after the Mn injection and it reaches a maxi-mum at 16 h (Fig. 3B). On the other hand, a less intense band at120 kDa is observed starting at 8 h (Fig. 3A), indicative of caspase-3-mediated breakdown of spectrin.

In order to assess the proteolytic activity of calpain directly in thestriatal homogenates, the calcium-dependent cleavage of a specificfluorogenic substrate for calpain,N-succinyl-Leu-Tyr-7-AMC,was evalu-ated (Fig. 3C). A significant increase in the calcium-dependent velocityof substrate cleavagewas observed for samples obtained from rats sacri-ficed 2 h after the intrastriatalMn injection. At 4 h, the observed enzymeactivity decreased to control levels, while a subsequent elevation be-comes evident at 16 and 24 h, which is significantly different to the4 h data. The observed peak in calpain activity at 2 h is consistent withthe Western blot results, which indicates a significant accumulation ofcalpain-mediated breakdown products of spectrin at 8 h. Spectrinbreakdown products are relatively resistant to proteolytic degradation

(Czogalla and Sikorski, 2005). Thus, an early enzyme activation at 2 h,in combination with an accumulation of the spectrin fragments cleavedat later times, is consistent with the sustained increase in the density ofthe 150/145 kDa fragments observed byWestern blot.

Finally, the proteolytic activity of caspase-3was evaluateddirectly inthe striatal homogenates, using the fluorogenic substrate Ac-DEVD-AMC. While striatal homogenates obtained 4 h after Mn administrationdid not display any caspase activity (data not shown), samples obtained8 h after Mn administration displayed significantly higher caspaseactivity, as compared to the saline-injected animals and to the contra-lateral striatum of the Mn-injected animals (Fig. 3D). At 12 h, theenzyme activity is still elevated, although it is not significantly differentto control levels (23.8±1.56 vs 17.2±2.8 change in fluorescence/min/mg protein, n=3), while at 16 h caspase activity returns to basallevels (Fig. 3D). As expected, the addition of the caspase-3 inhibitorDEVD completely abated caspase activity in all cases. The significant

Fig. 3. Caspase-3 and calpain activation after acute Mn treatment. Representative SDS-PAGE analysis followed byWestern blot for spectrin of striatal homogenates at different timesafter Mn injection (A). Average relative integrated intensity of the bands for the full spectrin (295 kDa) and 145/150 kDa spectrin fragments as a function of time after Mn injection(B); *Pb0.05, **Pb0.005, ***Pb0.0001 compared to controls, ANOVA and post hoc Fisher's PLSD test; n=5 for vehicle; n=3 for 2 h; n=3 for 4 h; n=5 for 8 h; n=6 for 12 h; n=9for 16 h; n=4 for 24 h and n=3 for 38 h. Relative calpain activity in striatal homogenates at different times after Mn injection (C); *Pb0.0001 compared to controls using ANOVApost hoc Fisher's PLSD and Bonferroni/Dunn tests, and Pb0.005 compared to controls using ANOVA post hoc Scheffé F test; **Pb0.008 compared to 4 h using ANOVA post hoc Fisher'sPLSD; n=5 for control; n=5 for 2 h; n=5 for 4 h; n=4 for 8 h; n=5 for 12 h; n=3 for 16 h; and n=3 for 24 h. Relative caspase activity in striatal homogenates at 8 and 16 hafter Mn injection (D). Data for the contralateral non-injected striata (contra), the saline-injected animals and samples treated with caspase inhibitor DEVD (+ inh) are also shown;*Pb0.05 compared to controls, contralateral striata and the 16 h samples using ANOVA post hoc Fisher's PLSD test; n=3 for all experimental groups except for 8 h (n=5). All dataare reported as mean±SEM for data sets.

187L. Quintanar et al. / Experimental Neurology 233 (2012) 182–192

increase in caspase activity at 8 h is consistent with the Western blotresults, which suggest the presence of caspase-mediated breakdownproducts of spectrin. As in the case of calpain products, the presenceof the 120 kDa band from 8 to 38 h after Mn injection might representthe accumulation of this product as a result of its slow degradation.

These results clearly indicate that both, calpain and caspase-3 areactivated after Mn injection in the striatum. However, it is importantto note that calpain activation is an early event (peak at 2 h) that pre-cedes the observed cell damage at 4 h after Mn administration, whilecaspase-3 activation occurs at a later time.

The role of caspase-3 in Mn-induced cell-death

The effect of the specific caspase-3 inhibitor DEVD and the broad-spectrum caspase inhibitor QVD on the extent of caspase-mediatedspectrin cleavage was evaluated. The inhibitors had no effect on theintensity of the 120 kDa fragment, as evaluated 8 h after Mn injection(Figs. 4A–B). However, at 16 h, when a substantial accumulation ofthe 120 kDa band is detected, a significant reduction in the intensityof this band is observed in animals co-injected with Mn and caspaseinhibitor (Figs. 4A–B). QVD displayed a more efficient inhibition thanDEVD at 16 h (Fig. 4C), consistent with the fact that QVD also inhibitsthe activity of caspases upstream caspase-3 (caspase-9 and caspase-8). Altogether, these results suggest that, under the present experimen-tal conditions, spectrin is also processed by active caspases. However,co-administration of Mn and QVD did not significantly reduce the sizeof Mn-induced lesions, as evaluated 24 h after the intrastriatal injection(Fig. 4D). These results indicate that, although the caspase inhibitorQVD partially prevents caspase-3 mediated spectrin breakdown, itdoes not significantly prevent Mn-induced cell death.

Calpain is involved in Mn-induced cell death

In order to evaluate whether calpain activation is directly involvedin Mn-induced cell death in the striatum, the specific calpain inhibitorMDL-28170 was administered systemically, right before and after theintra-striatal Mn injection. Western blot analysis of striatal homoge-nates obtained from MDL-28170-treated rats sacrificed 16 h after Mninjection, display significantly less amount of calpain-cleaved spectrinbreakdown products (Fig. 5A), as compared to samples obtained fromMn-treated rats that received no MDL-28170 treatment. The intensityof the 145/150 kDa doublet is significantly decreased upon MDL-28170 treatment (Fig. 5B), and it almost reaches the low intensityobserved in control animals. In addition, histological evaluation of thestriata of MDL-28170-treated rats sacrificed 24 h after the intra-striatalMn injection (Fig. 6C) shows a significant difference when compared tostriata of Mn-treated rats that received no MDL-28170 administration(Fig. 6A). The average volume of the Mn-induced lesions observed inMDL-treated animals is significantly lower than those in animals withno MDL treatment (Fig. 6D). These results clearly demonstrate thattreatment with the calpain inhibitor MDL-28170 successfully preventsMn-induced calpain activation and partially prevents Mn-induced celldeath.

Calpain activation is not associated with glutamate-induced excitotoxicity

Calpain activation may result from different events, such as mito-chondrial dysfunction, disruption in calcium homeostasis, or Glu-induced calcium entry into the cell. The latter possibility was tested bysystemic administration of the non-competitive NMDA receptor antag-onist MK-801, which blocks these receptors in the open-channel statepreventing Glu-induced calcium entry. The administration of one or

Fig. 4. Effect of caspase inhibitors QVD and DEVD on Mn-induced capase-3 activation and cell-death. Representative Western blot analysis for spectrin of: A) striatal homogenates fromrats with (n=3) and without QVD treatment (n=6), obtained 8 and 16 h after Mn injection; and B) striatal homogenates from rats with (n=3) and without DEVD treatment (n=6),obtained 8 h and 16 h after Mn injection. The relative integrated intensities of the 120 kDa spectrin fragment for the groups displayed in A) and B) are reported as means±SEM in C).*Pb0.002 compared to Mn-treated rats sacrificed at 8 h after Mn injection, using ANOVA post hoc Fisher's PLSD and Bonferroni/Dunn tests; ** Pb0.05 compared to Mn-treated ratswith no DEVD treatment at 16 h, using ANOVA post hoc Fisher's PLSD test; *** Pb0.0005 compared to Mn-treated rats with no QVD treatment at 16 h, using ANOVA post hoc Fisher'sPLSD and Bonferroni/Dunn tests. Effect of QVD co-administration in Mn-induced cell-death (D). Lesion volumes evaluated 24 h after Mn administration are displayed; the differencebetween Mn-treated animals (n=4) and those co-injected with Mn and QVD (n=7) is not statistically significant.

188 L. Quintanar et al. / Experimental Neurology 233 (2012) 182–192

two ip doses of MK-801 (2 mg/kg each) had no significant effect in Mn-induced calpain activation. Western blot for spectrin of striatal homog-enates from MK-801-treated animals sacrificed 16 h after Mn injectiondisplays the same amount of calpain-mediated spectrin breakdownproducts as animals not treatedwith the antagonist (Fig. 5C). The inten-sity of the 145/150 kDa doublet in these samples does not significantlychange upon MK-801 treatment, even with the administration of twodoses of MK-801 (Fig. 5D).

Furthermore, Nissl staining of the striata of MK-801-treated ratssacrificed 24 h after Mn injection (Fig. 6B) shows no differencewhen compared to striata of Mn-treated rats that received no MK-801 administration (Fig. 6A). Similar results were obtained after theadministration of memantine, another non-competitive NMDA antag-onist (Fig. 6D). Thus, in contrast to the case of MDL-28170 treatment,MK-801 and memantine administration did not significantly preventMn-induced cell death (Fig. 6D), demonstrating that blockade ofNMDA receptors fails to prevent Mn-induced calpain activation andcalpain-associated cell death.

Mn may directly activate calpain

Calpain is a calcium-activated neutral protease. In order to evaluateif other divalent metal ions with ionic radii similar to calcium (Table 1)are capable of activating calpain, striatal homogenates obtained fromhealthy rats and pre-incubated with different divalent metal ions: Ca,Mn, Mg, Zn and Pb, were evaluated for calpain activity using the fluoro-genic substrate for calpain, N-succinyl-Leu-Tyr-7-AMC. Representativecurves for the fluorescence increase observed over time after the addi-tion of the fluorogenic substrate are shown in Fig. 7. The velocity of sub-strate cleavage corresponds to the slope of fluorescence as a function oftime. The addition of Ca causes an increase in the initial velocity of sub-strate cleavage, compared to the cation-free experiment, as expected

for any striatal homogenate with a basal content of active calpain. Incontrast, the addition of Mg displays the same slope in the curve asthe cation-free experiment, indicating that this metal ion is not capableof increasing the initial velocity of calpain-mediated substrate break-down. Interestingly, the addition of Pb or Zn causes a decrease in theslope, as compared to that from the cation-free experiment, suggestingthat, either these cations inhibit calpain activity or they may interferewith the fluorescence emitted by the fluorogenic substrate upon itscleavage by calpain. Finally, the addition of Mn ion causes an increasein the initial velocity of substrate cleavage comparable in magnitudeto that induced by Ca. It should be noted that, in order to corroboratethat the measured fluorescence corresponds to substrate cleavage bycalpain, control experiments for each metal ion were performed usingthe calpain inhibitor MDL-28170. In the presence of MDL-28170, nosubstrate cleavage was observed even in the presence of Ca or Mn(Fig. 7, inset). From Fig. 7, it becomes evident that the effect of addingMn to the incubation buffer is comparable to that obtained with theaddition of Ca, while other metal ions with similar ionic radii had eitherno effect or a detrimental effect in the velocity of calpain-mediatedcleavage of N-succinyl-Leu-Tyr-AMC. These results suggest that Mnmay be able to directly activate calpain in the striatum.

Discussion

The present study aimed to investigate the role of the cell death-related proteases, caspase-3 and calpain, in an in vivo model of acuteMn neurotoxicity in the striatum. An injection of 100 nmol Mn wasneeded in order to reliably observe acuteMn neurotoxicity; this concen-tration (≈110 μg/g fresh tissue) is three times higher than the accumu-lated Mn in the globus pallidus of chronically-treated monkeys showingMn intoxication symptoms and severe neuronal death (Eriksson et al.,1987). Although this amount is high, when lower doses (25 and

Fig. 5. Effect of MDL-28170 (A and B) and MK-801 (C and D) administration on Mn-induced calpain activation. Representative Western blot analysis for spectrin of: A) striatal homog-enates from ratswith (n=4) andwithoutMDL-28170 treatment (n=3), obtained 16 h afterMn injection; and C) striatal homogenates from rats treatedwith 0, 1 and 2 doses ofMK-801,obtained 16 h afterMn injection (n=3). The average relative integrated intensities of the 145/150 kDa spectrin bands for the groups displayed in A) and C) are reported asmean values±standard deviation in B) and D), respectively. *Pb0.001 compared to Mn-treated rats with no MDL-28170 treatment using ANOVA post hoc Fisher's PLSD, Bonferroni/Dunn and SchefféF tests; and Pb0.05 compared to controls using ANOVA post hoc Fisher's PLSD test.

189L. Quintanar et al. / Experimental Neurology 233 (2012) 182–192

50 nmol) of Mn were tested, either very small lesions were obtained orthey were not reproducible, causing damage in some animals but not inothers. Concentrations of thismagnitude or even 10–20 fold higher havebeen employed to observe Mn-induced cell damage after its intracere-bral administration (Brouillet et al., 1993; Zhao et al., 2009).

Using this acute exposure model, degenerating neurons wereobserved 4 h after Mn administration, while an important astroglialactivation was observed surrounding the lesion core. These results arein agreement with the findings of Liu et al. (2006), who demonstratedthe presence of degenerating neurons and microglial activation in themice striatumafterMn treatment. A recent study also showedmicroglialactivation and the release of proinflammatory factors in the substantianigra 7 days after an acute intrastriatal Mn administration, suggestingthat inflammation influences the degeneration of dopaminergic cellsin the substantia nigra (Zhao et al., 2009). Therefore, in the presentexperimental conditions an inflammatory reaction is expected to occurwithin the striatum or in other brain regions connected to it.

According to the present data, caspase-mediated spectrin cleavageand a mild increase in caspase-3 activity were detected 8 h after Mnintrastriatal administration. These results agree with in vitro studiessuggesting an involvement of caspase-mediated apoptosis in Mn-induced death (Chun et al., 2001; Gonzalez et al., 2008; Hirata,2002; Malecki, 2001; Oubrahim et al., 2002; Yin et al., 2008). It isalso consistent with an in vivo study in mice suggesting apoptoticcell death of striato-pallidal interneurons after chronic exposure toMn (Liu et al., 2006). However, according to the present results cas-pase inhibition did not reduce the size of Mn-induced striatal lesions,suggesting that caspase activity does not have a main contribution toacute Mn neurotoxicity as induced in the present experimentalconditions.

On the other hand, the present study clearly shows for the first timethat the calcium-dependent protease, calpain, is strongly activated after

acute Mn intrastriatal administration. Both, the early activation of thisprotease (preceding cell death) and the prevention of the Mn-inducedcell death by the calpain inhibitor MDL-28170, place calpain as animportant mediator of acute Mn-induced neurotoxicity in vivo. This re-sult agrees with a previous in vitro study suggesting that calpain activa-tion byMn is involved inMn-induced apoptotic cell death (Oubrahimetal., 2002); however, in vivo calpain activation after Mn intracerebraladministration has not been previously reported. The transitorydecrease in enzyme activity observed at 4 h by the fluorometric assaymight result from its inhibition by calpastatin, an endogenous inhibitorof calpain. Calpastatin is a substrate of calpain and can be progressivelydegraded during the course of calpain activation, limiting the inhibitorycapacity of the cell and restoring the protease activity (Averna et al.,2007; De Tullio et al., 2000). In addition, calpastatin can also be cleavedby caspase-3 hindering its inhibitory action on calpain activity (Kato etal., 2000; Pörn-Ares et al., 1998). As an increase in caspase-3 activity isobserved 8 h after Mn administration, the degradation of calpastatinby this protease is possibly involved in the restoration of calpain activity16 and 24 h after Mn injection. Further experiments are needed toinvestigate these possibilities.

Blockade of NMDA Glu receptors by MK-801 had no significanteffect on calpain activation, and it did not prevent Mn-induced cell-death; suggesting that the calpain-dependent component of cell deathis not related to NMDA receptor activation. These results were corrobo-rated using memantine, another non-competitive antagonist of NMDAreceptors, which was unable to reduce Mn-induced calpain activation.Although some studies indicate that Mn alters Glu levels and decreasesthe expression of Glu transporters (Erikson et al., 2007; Fitsanakis et al.,2006), only few studies have addressed whether acute Mn neurotoxic-ity in the striatum involves an excitotoxic mechanism. An early studyshowed protection against Mn-induced damage in the rat striatum bythe administration of MK-801 (Brouillet et al., 1993) using an acute

Fig. 7. Effect of divalent metal ions on calpain-mediated substrate cleavage. Calpain ac-tivity was evaluated in striatal homogenates obtained from healthy rats and pre-incu-bated with 5 mM concentration of different divalent metal ions: Ca (___), Mn (−−−),Mg (.-.-), Zn (…..) and Pb (..–..–); the control experiment where no metal ion was addedto the incubation buffer is also included (……). Representative curves for the fluores-cence increase observed over time after the addition of the fluorogenic substrate areshown; n=3 for all groups except for Ca and Mn (n=4). The inset shows a compari-son of the fluorescence increase observed for striatal homogenates incubated with nometal ions (continuous trace) with the results observed for striatal homogenates incu-bated with different divalent metal ions and the specific calpain inhibitor MDL-28170(dotted and dashed lines).

Fig. 6. Effect of MK-801, memantine (Mem) and MDL-28170 administration on Mn-induced neuronal death. Representative images of Nissl staining of the striata of ratsobtained 24 h after Mn injection (n=9) (A), with 2 doses of MK-801 (n=4) (B) andwith MDL-28170 (n=5) (C) administration. The lesion volumes for these groups andthe memantine-treated animals (n=3) are reported as mean values with SEM(D) *Pb0.05 compared to Mn-treated rats with no MDL-28170 treatment using ANOVApost hoc Fisher's PLSD test.

Table 1Metal ions with ionic radii similar to Ca(II).

Metal ion Ionic radii [Å]a

Magnesium 0.72Zinc 0.74Manganese 0.83Calcium 1.00Lead 1.19

a Ionic radii correspond to divalent ions with coordination number of 6. Taken fromMartin(1986).

190 L. Quintanar et al. / Experimental Neurology 233 (2012) 182–192

exposure model similar to the one reported here. However, the amountof MK-801 used by Brouillet et al. was much higher (7 administrationsof 5 mg/kg, ip), as compared to the amount used here (1 or 2 adminis-trations of 2 mg/kg, ip). Most importantly, the doses of MK-801 used inthe present study have been demonstrated to be sufficient to preventGlu-induced neuronal damage and Glu-induced calpain activation invivo upon intrastriatal administration (Del Rio and Massieu, 2008; DelRio et al., 2008). Thus, under our experimental conditions, MK-801and memantine did not prevent Mn-induced neuronal damage, whilea specific calpain inhibitor did. Possibly, non-NMDA Glu receptorsmight be involved in calcium influx and calpain activation, as it hasbeen previously shown for Glu (Del Rio and Massieu, 2008).

Calpain is a calcium-activated neutral protease, and thus, any dis-ruption of calcium homeostasis might lead to calpain activation. Mncan enter the mitochondria impairing the mitochondrial membranepotential (Gonzalez et al., 2008; Malecki, 2001; Yin et al., 2008) andaltering calcium transport in this organelle (Gavin et al., 1990, 1999;Gunter et al., 2006, 2004). Disruption of mitochondrial ATP produc-tion might have an impact on calcium homeostasis altering ATP-dependent Ca loading and extrusion mechanisms. In addition, Mnentering the endoplasmic reticulum (ER) might influence calciumuptake by this organelle disrupting intracellular calcium homeostasis.Accordingly, a recent study demonstrated a reduction of the

releasable calcium pool at the ER upon exposure of cultured corticalastrocytes to Mn (Tjalkens et al., 2006). Therefore, exposure to Mnmight increase the intracellular concentration of calcium, by disturb-ing the cellular calcium homeostasis maintained by the mitochondriaand ER, and promote calpain activation. This possibility is difficult toassess in in vivo studies, thus further investigations in cellular systemsare needed.

Alternatively, a direct activation of calpain by Mn may be pro-posed. In fact, the evaluation of calpain activity in striatal homoge-nates of healthy rats suggests that this is indeed the case, since Mnincreases the enzyme activity in a similar fashion as calcium does.This is not surprising as Mn is known to be a good analogue for Caand Mg in a variety of biological systems, and these three divalentmetal ions can sometimes target the same binding sites in severalenzymes; for example, Mn and Mg can both activate pyruvate kinase(Klein and Charles, 1989). A direct activation of calpain by Mn in thecytosol is possible, as Mn can cross the blood brain barrier and betransported into brain cells by several mechanisms, including trans-ferrin-mediated endocytosis and direct transport by the divalentmetal transporter DMT-1, as previously reviewed (Aschner, 2006;Quintanar, 2008; Roth, 2006; Takeda, 2003). Once inside the cell, Mncan exploit the cellular Ca trafficking machinery to access differentorganelles, as it behaves as a good analogue for Ca (Quintanar, 2008);for example, Mn and Ca share intracellular ATPase transporters(Vangheluwe et al., 2009), andMn inhibits Ca efflux from themitochon-dria by direct competition for the binding site in the mitochondrial Na-independent efflux system (Gavin et al., 1990, 1999). A synergisticeffect of Mn and Ca on calpain activation has been proposed (Suzuki

191L. Quintanar et al. / Experimental Neurology 233 (2012) 182–192

and Tsuji, 1982), while a study with isolated m-calpain from bovinecardiac muscle has shown activation byMn, inhibition by Zn and no ef-fect by Mg on enzyme activity (Tan et al., 1988). A more recent studywith m-calpain purified from pigeon heart also reports a strong activa-tion of the enzyme by Mn (Gaitanaki et al., 2003). To our knowledge,this is the first report of calpain activation by Mn in brain tissue.

Calpains are ubiquitous in nature, and several isoforms of calpain areknown: μ-calpain that requires Ca concentrations in the range of 3 to50 μM for its activation, and m-calpain that requires higher Ca concen-trations (400 to 800 μM) to be activated (Goll et al., 2003). Mostcalpains are multi-domain proteins with several oxygen-rich bindingsites for Ca. Mn has an ionic radius that is comparable to that of Ca,and most importantly, it has a preference for oxygen-based ligands.Thus, Mn could easily accommodate the coordination modes imposedby oxygen-rich Cabinding sites, like those present in the calpain system.Therefore, during Mn neurotoxicity, calpain activation may occur upona direct interaction of thismetal ionwith this enzyme. From this study itcannot be determinedwhich type of calpain is being activated (m- or μ-calpain) by Mn exposure and the possibility of a synergistic effect ofintracellular Ca andMn in the activation of calpain cannot be discarded.

Finally, the specificity of Mn as a calpain activator was tested bythe administration of other divalent metal ions (Ca and Mg). WhileCa induced striatal lesions of a similar extension to those inducedby Mn, Mg-induced lesions were about 60% of those induced by Mn.Consistent with striatal damage, the intrastriatal administration ofCa induced the cleavage of spectrin into the 145/150 kDa calpainbreakdown products, in a similar fashion as Mn, while Mg induced amoderate activation of calpain (Supplementary Fig. 4). As calpain isa calcium-activated protease, it is not surprising that the administra-tion of high concentrations of Ca causes calpain activation and striataldamage. While Ca and Mn can directly activate calpain, as measuredin vitro in striatal homogenates from healthy rats, Mg does not affectthe enzyme activity. These results are in agreement with the moremoderate effect of Mg intrastriatal administration on calpain activityand tissue damage, and suggest that the effects of Mg might be medi-ated by the alteration of calcium homeostasis rather than by a directeffect of Mg on the protease. A comparison of the effects of Ca, Mnand Mg supports the hypothesis that the in vivo effect of Mn on cal-pain activity might come from a direct interaction with the enzyme;yet, an indirect activation of calpain due to the alteration of calciumhomeostasis by Mn cannot be discarded, and it should be tested infuture studies.

Altogether, the present study places calpain as an important player ofMn-induced neurotoxicity in vivo. This is particularly important, ascalpain activity has been suggested to participate in the neuropathologyof neurodegenerative diseases, such as Alzheimer's (Nixon et al., 1994),Parkinson's (Crocker, 2003) and Huntington's diseases (Gafni andEllerby, 2002), as reviewed previously (Vosler et al., 2008). Moreover,proteins that form protein aggregates in these diseases have beenshown to be calpain substrates, including alpha-synuclein in Parkinson'sdisease (Dufty et al., 2007; Mishizen-Eberz et al., 2003, 2005), beta-amyloid and tau in Alzheimer's disease (Yamazaki et al., 1997; Yenet al., 1999). Thus, further experiments are needed in order to identifyrelevant protein substrates of Mn-activated calpain, and their role instriatal cell death in vivo.

In conclusion, the present study suggests that additional mecha-nisms to excitotoxicity are involved in acute Mn-induced neurotoxicityin vivo, and that calpain is an importantmediator of cell death. It remainsto be studiedwhether calpain activation contributes to striatal cell deathin a chronicMn exposuremodel that resembles humanMn intoxication.

Acknowledgments

This work was supported by Universidad Nacional Autónoma deMéxico [grant number PAPIIT IN211710 to L.M.]; Consejo Nacional deCiencia y Tecnología (México) [grant numbers S112179 to L.M.,

J48781-Q to L.Q.]; and Instituto de Ciencia y Tecnología del Distrito Fed-eral (México) [grant number PIFUTP08-161 to L.Q.]. A postdoctoral fel-lowship (CETIC, UNAM) and a For Women in Science Fellowship (L'Oreal,UNESCO and the Mexican Academy of Sciences) were awarded to L.Q.The authors would like to thank: Dr. Jaime Ortega (Dept. de Biotecnolo-gía, Cinvestav) for facilitating access to a fluorometer, Gabriel Orozco forhis help in confocal microscopy and Dr. Perla Del Río for usefuldiscussions.

Appendix A. Supplementary data

Supplementary data to this article can be found online at doi:10.1016/j.expneurol.2011.09.032.

References

Araújo, I.M., Carreira, B.P., Carvalho, C.M., Carvalho, A.P., 2010. Calpain and delayed cal-cium deregulation in excitotoxicity. Neurochem. Res. 35, 1966–1969.

Aschner, M., 2006. The transport of manganese across the blood–brain barrier. Neuro-toxicology 27, 311–314.

Aschner, M., Erikson, K.M., Herrero Hernández, E., Tjalkens, R., 2009. Manganese and itsrole in Parkinson's disease: from transport to neuropathology. NeuromolecularMed. 11, 252–266.

Averna, M., Stifanese, R., De Tullio, R., Passalacqua, M., Defranchi, E., Salamino, F., Melloni,E., Pontremoli, S., 2007. Regulation of calpain activity in rat brain with altered Ca2+

homeostasis. J. Biol. Chem. 282, 2656–2665.Benedetto, A., Au, C., Aschner, M., 2009. Manganese-induced dopaminergic neurode-

generation: insights into mechanisms and genetics shared with Parkinson's dis-ease. Chem. Rev. 109, 4862–4884.

Bizat, N., Hermel, J.-M., Boyer, F., Jacquard, C., Créminon, C., Ouary, S., Escartin, C., Hantraye, P.,Krajewski, S., Brouillet, E., 2003. Calpain is a major cell death effector in selective striataldegeneration induced in vivo by 3-nitropropionate: implications for Huntington's dis-ease. J. Neurosci. 23, 5020–5030.

Bonilla, E., Salazar, E., Villasmil, J.J., Villalobos, R., 1982. The regional distribution ofmanganese in the normal human brain. Neurochem. Res. 7, 221–227.

Braak, H., Del Tredici, K., Rüb, U., de Vos, R.A.I., Jansen Steur, E.N.H., Braak, E., 2003.Staging of brain pathology related to sporadic Parkinson's disease. Neurobiol.Aging 24, 197–211.

Brouillet, E., Shinobu, L., McGarvey, U., Hochberg, F., Beal, M., 1993. Manganese injec-tion into the rat striatum produces excitotoxic lesions by impairing energy metab-olism. Exp. Neurol. 120, 89–94.

Chen, C.J., Ou, Y.C., Lin, S.Y., Liao, S.L., Chen, S.Y., Chen, J.H., 2006. Manganese modulatespro-inflammatory gene expression in activated glia. Neurochem. Int. 49, 62–71.

Choi, D.W., Koh, J.Y., Peters, S., 1988. Pharmacology of glutamate neurotoxicity in cor-tical cell culture: attenuation by NMDA antagonists. J. Neurosci. 8, 185–196.

Chun, H.S., Lee, H., Son, J.H., 2001. Manganese induces endoplasmic reticulum (ER)stress and activates multiple caspases in nigral dopaminergic neuronal cells,SN4741. Neurosci. Lett. 316, 5–8.

Crocker, S.J., 2003. Inhibition of calpains prevents neuronal and behavioral deficits inan MPTP mouse model of Parkinson's disease. J. Neurosci. 23, 4081–4091.

Czogalla, A., Sikorski, A.F., 2005. Spectrin and calpain: a ‘target’ and a ‘sniper’ in the pa-thology of neuronal cells. CMLS, Cell. Mol. Life Sci. 62, 1913–1924.

De Tullio, R., Averna, M., Salamino, F., Pontremoli, S., Melloni, E., 2000. Differential deg-radation of calpastatin by u- and m-calpain in Ca2+-enriched human neuroblasto-ma LAN-5 cells. FEBS Lett. 475, 17–21.

Del Rio, P., Massieu, L., 2008. Mild mitochondrial inhibition in vivo enhances glutamate-induced neuronal damage through calpain but not caspase activation: role of iono-tropic glutamate receptors. Exp. Neurol. 212, 179–188.

Del Rio, P., Montiel, T., Massieu, L., 2008. Contribution of NMDA and non-NMDA recep-tors to in vivo glutamate-induced calpain activation in the rat striatum. Relation toneuronal damage. Neurochem. Res. 33, 1475–1483.

Dufty, B.M., Warner, L.R., Hou, S.T., Jiang, S.X., Gomez-Isla, T., Leenhouts, K.M., Oxford,J.T., Feany, M.B., Masliah, E., Rohn, T.T., 2007. Calpain-cleavage of alpha-synuclein.Connecting proteolytic processing to disease-linked aggregation. Am. J. Pathol.170, 1725–1738.

Erikson, K.M., Dorman, D.C., Lash, L.H., Aschner, M., 2007. Manganese inhalation byRhesus monkeys is associated with brain regional changes in biomarkers of neuro-toxicity. Toxicol. Sci. 97, 459–466.

Eriksson, H., Magiste, K., Plantin, L.-O., Fonnum, F., Hedstrom, K.-G., Theodorsson-Norheim,E., Kristensson, K., Stalberg, E., Heilbronn, E., 1987. Effects of manganese oxide onmonkeys as revealed by a combined neurochemical, histological and neurophys-iological evaluation. Arch. Toxicol. 61, 46–52.

Fitsanakis, V.A., Au, C., Erikson, K.M., Aschner, M., 2006. The effects of manganese onglutamate, dopamine and gamma-aminobutyric acid regulation. Neurochem. Int.48, 426–433.

Gafni, J., Ellerby, L.M., 2002. Calpain activation in Huntington's disease. J. Neurosci. 22.Gaitanaki, C., Papazafiri, P., Beis, I., 2003. The calpain–calpastatin system and the calci-

um paradox in the isolated perfused pigeon heart. Cell. Physiol. Biochem. 13,173–180.

Galvani, P., Fumagalli, P., Santagostino, A., 1995. Vulnerability of mitochondrial com-plex I in PC12 cells exposed to manganese. Eur. J. Pharmacol. 293, 377–383.

192 L. Quintanar et al. / Experimental Neurology 233 (2012) 182–192

Gavin, C.E., Gunter, K.K., Gunter, T.E., 1990. Manganese and calcium efflux kinetics inbrain mitochondria. Biochem. J. 266, 329–334.

Gavin, C.E., Gunter, K.K., Gunter, T.E., 1999. Manganese and calcium transport in mito-chondria: implications for manganese toxicity. Neurotoxicology 20, 445–454.

Goll, D.E., Thompson, V.F., Li, H., Wei, W., Cong, J., 2003. The calpain system. Physiol.Rev. 83, 731–801.

Gonzalez, L.E., Juknat, A.A., Venosa, A.J., Verrengia, N., Kotler, M.L., 2008. Manganese ac-tivates the mitochondrial apoptotic pathway in rat astrocytes by modulating theexpression of proteins of the Bcl-2 family. Neurochem. Int. 53, 408–415.

Gunter, T.E., Miller, L.M., Gavin, C.E., Eliseev, R., Salter, J., Buntinas, L., Alexandrov, A.,Hammond, S., Gunter, K.K., 2004. Determination of the oxidation states of manga-nese in brain, liver and heart mitochondria. J. Neurochem. 88, 266–280.

Gunter, T.E., Gavin, C.E., Aschner, M., Gunter, K.K., 2006. Speciation of manganese incells and mitochondria: a search for the proximal cause of manganese neurotoxic-ity. Neurotoxicology 27, 765–776.

Hirata, Y., 2002. Manganese-induced apoptosis in PC12 cells. Neurotoxicol. Teratol. 24,639–653.

Kato, M., Nonaka, T., Maki, M., Kikuchi, H., Imajoh-Ohmi, S., 2000. Caspases cleave theamino-terminal calpain inhibitory unit of calpastatin during apoptosis in humanJurkat T cells. J. Biochem. 127, 297–305.

Klein, D.P., Charles, A.M., 1989. Differences betweenmagnesium- andmanganese-activatedpyruvate kinase from Thiobacillus versutus (A2). Curr. Microbiol. 19, 57–60.

Liu, X., Sullivan, K.A., Madl, J.E., Legare, M., Tjalkens, R.B., 2006. Manganese-inducedneurotoxicity: the role of astroglial-derived nitric oxide in striatal interneuron de-generation. Toxicol. Sci. 91, 521–531.

Malecki, E.A., 2001. Manganese toxicity is associated with mitochondrial dysfunction andDNA fragmentation in rat primary striatal neurons. Brain Res. Bull. 55, 225–228.

Martin, R.B., 1986. Bioinorganic chemistry of metal ion toxicity. In: Sigel, H. (Ed.), MetalIons in Biological Systems. Marcel Dekker Inc., New York, p. 21.

Mishizen-Eberz, A.J., Guttmann, R.P., Giasson, B.I., Day III, G.A., Hodara, R., Ischiropoulos, H.,Lee, V.M.-Y., Trojanowski, J.Q., Lynch, D.R., 2003. Distinct cleavage patterns of normaland pathologic forms of alpha-synuclein by calpain I in vitro. J. Neurochem. 86, 836–847.

Mishizen-Eberz, A.J., Norris, E.H., Giasson, B.I., Hodara, R., Ischiropoulos, H., Lee, V.M.-Y.,Trojanowski, J.Q., Lynch, D.R., 2005. Cleavage of alpha-synuclein by calpain: poten-tial role in degradation of fibrillized and nitrated species of alpha-synuclein. Bio-chemistry 44, 7818–7829.

Nixon, R.A., Saito, K.I., Grynspan, F., Griffin, W.R., Katayama, S., Honda, T., Mohan, P.s.,Shea, T.B., Beermann, M., 1994. Calcium-activated neutral proteinase (calpain) sys-tem in aging and Alzheimer's disease. Ann. N. Y. Acad. Sci. 747, 77–91.

Olanow, C., 2004. Manganese-induced Parkinsonism and Parkinson's disease. Ann. N. Y.Acad. Sci. 1012, 209–223.

Olanow, C., Good, P., Shinotoh, H., Hewitt, K., Vingerhoets, F., Snow, B., Beal, M., Calne,D., Perl, D., 1996. Manganese intoxication in the Rhesus monkey: a clinical, imag-ing, pathologic, and biochemical study. Neurology 46, 492–498.

Oubrahim, H., Chock, P.B., Stadtman, E.R., 2002. Manganese(II) induces apoptotic celldeath in NIH3T3 cells via a caspase-12-dependent pathway. J. Biol. Chem. 277,20135–20138.

Pörn-Ares, M.I., Samali, A., Orrenius, S., 1998. Cleavage of the calpain inhibitor, calpastatin,during apoptosis. Cell Death Differ. 5, 1028–1033.

Quintanar, L., 2008. Manganese neurotoxicity: a bioinorganic chemist's perspective.Inorg. Chim. Acta 361, 875–884.

Roth, J.A., 2006. Homeostatic and toxic mechanisms regulating manganese uptake, re-tention, and elimination. Biol. Res. 39, 45–57.

Roth, J.A., 2009. Are there common biochemical and molecular mechanisms controllingmanganism and parkinsonism. Neuromolecular Med. 11, 281–296.

Suzuki, K., Tsuji, S., 1982. Synergistic activation of calcium-activated neutral proteaseby Mn(II) and Ca(II). FEBS Lett. 140, 16–18.

Takeda, A., 2003. Manganese action in brain function. Brain Res. Rev. 41, 79–87.Tamm, C., Sabri, F., Ceccatelli, S., 2008. Mitochondrial-mediated apoptosis in neural

stem cells exposed to manganese. Toxicol. Sci. 101, 310–320.Tan, F.C., Goll, D.E., Otsuka, Y., 1988. Some properties of the millimolar calcium-

dependent proteinase from bovine cardiac muscle. J. Mol. Cell. Cardiol. 20,983–997.

Tjalkens, R.B., Zoran, M.J., Mohl, B., Barhoumi, R., 2006. Manganese suppresses ATP-dependent intercellular calcium waves in astrocyte networks through alterationof mitochondrial and endoplasmic reticulum calcium dynamics. Brain Res. 1113,210–219.

Vangheluwe, P., Sepulveda, M.R., Missiaen, L., Raeymaekers, L., Wuytack, F., Vanoeve-len, J., 2009. Intracellular Ca(II) and Mn(II) transport ATPases. Chem. Rev. 109,4733–4759.

Vosler, P.S., Brennan, C.S., Chen, J., 2008. Calpain-mediated signaling mechanisms inneuronal injury and neurodegeneration. Mol. Neurobiol. 38, 78–100.

Wang, K., 2000. Calpain and caspase: can you tell the difference? Trends Neurosci. 23,20–26.

Yamada, M., Ohno, S., Okayasu, I., Okeda, R., Hatakeyama, S., Watanabe, H., Ushio, K.,Tsukagoshi, H., 1986. Chronicmanganese poisoning: a neuropathological studywith de-termination of manganese distribution in the brain. Acta Neuropathol. 70, 273–278.

Yamazaki, T., Haass, C., Saido, T.C., Omura, S., Ihara, Y., 1997. Specific increase in amyloidbeta-protein 42 secretion ratio by calpain inhibition. Biochemistry 36, 8377–8383.

Yen, s., Easson, C., Nacharaju, P., Hutton, M., Yen, S.-H., 1999. FTDP-17 tau mutationsdecrease the susceptibility of tau to calpain I digestion. FEBS Lett. 461, 91–95.

Yin, Z., Aschner, J., dos Santos, A.P., Aschner, M., 2008. Mitochondrial-dependent man-ganese neurotoxicity in rat primary astrocyte cultures. Brain Res. 1203, 1–11.

Zhang, S., Zhou, Z., Fu, J., 2003. Effect of manganese chloride exposure on liver andbrain mitochondria function in rats. Environ. Res. 93, 149–157.

Zhao, F., Cai, T., Liu, M., Zheng, G., Luo, W., Chen, J., 2009. Manganese induces dopami-nergic neurodegeneration via microglial activation in a rat model of manganism.Toxicol. Sci. 107, 156–164.