Gel-based Protease Proteomics for Identifying the Novel Calpain Substrates in Dopaminergic Neuronal Cell*

Background: It is important to assess contribution of calpain activation and identify substrates affected during neurodegeneration. Results: Gel-based protease proteomics identified novel substrates that were cleaved in neurotoxin-treated culture and rat brain disease models. Conclusion: These novel calpain substrates may confer protection against neurodegeneration. Significance: Our findings contribute to better deciphering the molecular mechanism underlying the progression of protease-mediated neurodegeneration. Calpains are a family of calcium-dependent cysteine proteases that are ubiquitously expressed in mammals and play critical roles in neuronal death by catalyzing substrate proteolysis. Here, we developed two-dimensional gel electrophoresis-based protease proteomics to identify putative calpain substrates. To accomplish this, cellular lysates from neuronal cells were first separated by pI, and the immobilized sample on a gel strip was incubated with a recombinant calpain and separated by molecular weight. Among 25 altered protein spots that were differentially expressed by at least 2-fold, we confirmed that arsenical pump-driving ATPase, optineurin, and peripherin were cleaved by calpain using in vitro and in vivo cleavage assays. Furthermore, we found that all of these substrates were cleaved in MN9D cells treated with either ionomycin or 1-methyl-4-phenylpyridinium, both of which cause a calcium-mediated calpain activation. Their cleavage was blocked by calcium chelator or calpain inhibitors. In addition, calpain-mediated cleavage of these substrates and its inhibition by calpeptin were confirmed in a middle cerebral artery occlusion model of cerebral ischemia, as well as a stereotaxic brain injection model of Parkinson disease. Transient overexpression of each protein was shown to attenuate 1-methyl-4-phenylpyridinium-induced cell death, indicating that these substrates may confer protection of varying magnitudes against dopaminergic injury. Taken together, the data indicate that our protease proteomic method has the potential to be applicable for identifying proteolytic substrates affected by diverse proteases. Moreover, the results described here will help us decipher the molecular mechanisms underlying the progression of neurodegenerative disorders where protease activation is critically involved.

Parkinson disease (PD) 4 is a common neurodegenerative disorder characterized by the progressive loss of dopaminergic neurons in the substantia nigra pars compacta, and these neurons develop proteinaceous inclusions called Lewy bodies (1). Mutations in ␣-synuclein, parkin, DJ-1, PINK1, and LRRK2 genes appear to be associated with familial forms of PD, but the majority of cases are sporadic. Oxidative stress, mitochondrial dysfunction, and accumulation of abnormal protein aggregates are all thought to contribute to PD pathogenesis (2). Gene-and neurotoxin-based models of PD have been widely used to elucidate the molecular mechanisms associated with neuronal cell death in PD. For example, both apoptotic and necrotic mechanisms have been implicated in neurotoxin-based models established with 6-hydroxydopamine, 1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine (MPTP; its active metabolite, MPP ϩ ), rotenone, and paraquat. Activation of various proteases, including caspase and calpain, has been shown to play a critical role in neuronal death in these model systems. Consequently, inhibition of protease activation within neurons has been developed  1 Both authors contributed equally to this work. 2 To whom correspondence may be addressed: Anamdong-5-ga, Seongbukgu, Seoul 136-705, Korea. Tel.: 82-2-920-6094; Fax: 82-2-953-6095; E-mail: wonki@korea.ac.kr. 3 To whom correspondence may be addressed: 134 Shinchon-dong, Seodaemoon-gu, Seoul 120-749, Korea. as a neuroprotective strategy (1). Calpains belong to a family of intracellular Ca 2ϩ -dependent, nonlysosomal cysteine proteases (reviewed in Ref. 3). They are highly conserved, structurally related, and ubiquitously expressed in mammals, as well as many other organisms. Among 16 known genes, calpain 1 (-calpain) and calpain 2 (m-calpain) represent two isoforms that are the best characterized members of the calpain family. Structurally, these two heterodimeric isoforms share an identical small regulatory subunit (28 kDa) but have distinct large catalytic subunits (80 kDa) (3). Both isoforms are highly expressed in neurons and glia in the central nervous system (4). Among several proposed functional implications, Ca 2ϩ -triggered activation of calpain has been demonstrated to play an important role in the initiation, regulation, and execution of different forms of neuronal death, including apoptosis, necrosis, and autophagy (5). Considering that calpains exert their regulatory action by proteolytic processing of endogenous substrates, it is important to assess the contribution of calpain activation and identify substrates affected during neurodegeneration. Previously, several independent approaches, including proteomic analyses (6 -9), were performed to identify endogenous calpain substrates. However, it is not clearly understood whether the putative substrates are directly cleaved by calpain or other proteolytic enzymes. Here, we described a novel protease proteomic analysis that employs conventional gel-based two-dimensional gel electrophoresis (2DE). We used the MN9D dopaminergic neuronal cell line that is a fusion product of embryonic mesencephalic dopaminergic neurons and N18TG neuroblastoma cells (10). First, MN9D cellular lysates were extracted without any protease inhibitor treatment and subjected to isoelectric focusing (IEF). The proteins were immobilized on a strip and incubated with or without active recombinant m-calpain to ensure that only the direct substrates would be cleaved. Following separation by SDS-PAGE, several protein spots that were either up-or down-regulated were subjected to mass spectral analysis using MALDI-TOF. Among these altered protein spots, we selected arsenical pump-driving ATPase (ASNA1), optineurin, and peripherin for further validation. We subsequently confirmed that these proteins are cleaved by activated calpain both in cultured cells and in rat models of neurodegenerative diseases. Our protease proteomic analysis seems to be useful and broadly applicable to identifying novel protease substrates that play critical regulatory roles in neuronal cell death.

EXPERIMENTAL PROCEDURES
Cell Culture, Drug Treatment, and Cell Viability-Cells were plated at a density of either 1.0 ϫ 10 6 , 1.0 ϫ 10 5 , or 2.5 ϫ 10 4 cells on 25-g/ml poly-D-lysine (Sigma)-coated P-100 dishes (Corning Glass Works, Corning, NY), 4-well culture dishes (Nunc, Roskilde, Denmark) or 24-well culture plates (Corning Glass Works), respectively. Cells were maintained for 3 days in Dulbecco's modified Eagle's medium (Sigma) supplemented with 10% heat-inactivated fetal bovine serum (Invitrogen) in an incubator with an atmosphere of 10% CO 2 at 37°C. Media were changed to serum-free N2 medium before drug treatment. Cells were treated with 50 M MPP ϩ (Research Biochemicals International, Natick, MA) or 5 g/ml ionomycin (Sigma) in the presence or absence of 40 M BAPTA/AM (Molecular Probes, Eugene, OR), 50 M PD150606 (Calbiochem, La Jolla, CA), or 50 M calpeptin (Calbiochem). Following MPP ϩ treatment, the rate of cell survival was determined by colorimetric measurement using MTT reduction assays (11). In brief, cells were incubated for 1 h at 37°C with a final concentration of 1 mg/ml MTT solution and lysed in 20% SDS in 50% aqueous dimethylformamide for 24 h. The optical density of the dissolved formazan grains was measured at 540 nm using a microplate reader (Molecular Devices, Palo Alto, CA). Viabilities were represented as percentages of values from MPP ϩ -treated cells over the untreated control cells (100%).
Middle Cerebral Artery Occlusion (MCAO) and Stereotaxic Surgery-All experimental procedures applied to animals were in accordance with the National Institutes of Health Guide for the Care and Use of Laboratory Animals and were approved by the Committee of Korea University and Kyung Hee University. To establish a focal cerebral ischemic model, Sprague-Dawley male rats (260 -270 g) were anesthetized with a mixture of 3% isoflurane in 70% N 2 O and 30% O 2 (v/v) via facemask and maintained with 2% isoflurane. A rectal temperature probe was introduced, and body temperature was maintained at 37°C during the entire surgery period. Focal cerebral ischemia was achieved by right-sided endovascular MCAO, as described previously (12). The occlusion was released after 1.5 h, and the animals were allowed to recover. For actual experiments, rats randomly divided into two groups received intracerebroventricle infusion of either 50 g of calpeptin dissolved in 5 l of Me 2 SO (Sigma) or 5 l of Me 2 SO alone 30 min before MCAO as basically described (13). The rats were reanesthetized with isoflurane in N 2 O and O 2 24 h after ischemia onset, and their brains were removed for immunoblot analysis. For stereotaxic surgery, Sprague-Dawley female rats were anesthetized with an intraperitoneal injection of chloral hydrate (360 mg/kg) and positioned in a stereotaxic apparatus (Kopf Instruments, Tujunga, CA). Each rat received a unilateral injection of MPP ϩ (7.4 g in 2 l of PBS; Sigma) for 3 days into the right medial forebrain bundle (anteroposterior, 3.6 mm; mediolateral, 2.0 mm; dorsoventral, 7.5 mm from bregma). All injections were performed with a Hamilton syringe equipped with a 26 S gauge beveled needle attached to a syringe pump (KD Scientific, Holliston, MA) at a rate of 0.2 l/min. One minute after injection, the needle was slowly retracted. The bilateral substantia nigral tissues were dissected out for immunoblot analysis.
2DE-All chemicals used for 2DE were purchased from Sigma unless otherwise stated. Gel-based proteomic analysis was performed as we described previously with some modifications (14). Briefly, MN9D cells were solubilized in sample buffer containing 5 M urea, 2 M thiourea, 2% CHAPS, 0.25% Tween 20, 100 mM DTT, 10% isopropyl alcohol, 12.5% water-saturated butanol, 5% glycerol, and 1% immobilized pH gradient buffer for pH 4 -7 linear Immobiline DryStrip (24 cm; GE Healthcare). Protein content was determined with two-dimensional Quantikits (GE Healthcare). DryStrips were rehydrated in sample buffer containing 1.5 mg of cellular lysate. Gels were run for a total of 100 kV-h using progressively increasing voltage on an Ettan IPGphor (GE Healthcare). After IEF, the strips were washed with distilled water and incubated with or without 83.25 units of recombinant m-calpain (Calbiochem: catalog no. 208718) or m-calpain from porcine kidney (Calbiochem; catalog no. 208715) in calpain activation buffer (25 mM Tris, 10 mM KCl, 10 mM CaCl 2 , and 0.1% Triton X-100) in a final volume of 500 l for 2 h at 37°C. Prior to SDS-PAGE, the strips were washed three times with distilled water, and an equilibration step was carried out. SDS-PAGE was performed on 8 -18% gradient gels in an Ettan Dalt II System (GE Healthcare). Coomassie Brilliant Blue G-250-stained gels were scanned using a densitometer (Powerlook 2100XL; UMAX, Dallas, TX), and the gel images were analyzed with ProteomeWeaver software version 2.1 (DEFINIENS, München, Germany). To minimize variation of background staining intensity, we simultaneously stained and destained gels in a multiple chamber. Fold changes of the protein spots were normalized by a background subtraction. For calpain activation measurement, a mini-2DE method was conducted. Briefly, DryStrips (7 cm; GE Healthcare) were rehydrated in sample buffer containing 100 g of cellular lysate. Gels were run for a total of 35 kV-h using progressively increasing voltage on an Ettan IPGphor. After IEF, the strips were washed with distilled water and incubated with or without 5.55 units of porcine kidney m-calpain in calpain activation buffer in a final volume of 125 l for 2 h at 37°C.
Protein Identification-In-gel digestion and mass spectrometric analysis were performed as we described previously (14). Briefly, the gel containing protein spots of interest was manually excised and destained with 50% acetonitrile in 25 mM ammonium bicarbonate buffer, pH 8.0. The destained gel slices were completely dried in a SpeedVac evaporator dryer (BioTron Inc., Gyeonggi-do, Korea). Gel slices were rehydrated and digested for 18 h at 37°C with 10 g/ml porcine trypsin (842.51 and 2,211.10 m/z; Promega, Madison, WI) in 25 mM ammonium bicarbonate buffer, pH 8.0. Tryptic peptides were purified with a POROS R2 column (Applied Biosystems, Foster City, CA) and spotted onto stainless steel MALDI plates (Applied Biosystems). The peptide mixtures were acquired in reflectron mode (mass range from 800 to 4,000 Da, 20-keV accelerating voltage, 500-ns delayed extraction, averaging 1,000 laser shots/spectrum) using a 4700 Proteomic Analyzer equipped with a 355-nm nitrogen laser (Applied Biosystems). The spectra were analyzed with GPS Explorer TM software ver. 3.5 (Applied Biosystems) using ProteinProspector 4.0.7 (MS-Fit; University of California, San Francisco) and the following parameters: signal to noise threshold, 10; mass exclusion tolerance, 0.2 m/z; isotope cluster (maximum charge, ϩ1; isotope spacing tolerance, 0.1); Monoisotopic peaks (Adduct: H, Genetic formula: C 6 H 5 NO-peptides). The trypsin autodigestion ion picks were used as internal standards. Matrix and/or autoproteolytic trypsin fragments or known contaminant ions were excluded. The resulting peptide mass lists were used to query the SwissProt database (2012.12.3, 453,850 sequences; taxonomy, Mus musculus, 16,581 sequences). The following criteria were used for search parameters: significant protein MOWSE score at p Ͻ 0.05, 1 missed cleavage site allowed, 1ϩ peptide charges allowed, trypsin as enzyme, 50 -100 ppm as precursor tolerance, peptide N-terminal Gln to pyroGlu, oxidation of methionine, protein N terminus acetylated, and acrylamide-modified Cys as fixed and variable modifications. Further clas-sification based on indicated criteria of the identified proteins was performed using the program known as the PANTHER (Protein ANalysis THrough Evolutionary Relationships).
In Vitro and Cell-based Calpain Cleavage Assay-For in vitro calpain cleavage assays, the vectors encoding ASNA1, optineurin, and peripherin were transcribed and translated in vitro in the presence of [ 35 S]methionine (PerkinElmer Life Sciences) by using transcription and translation-coupled reticulocyte lysate system (Promega). For cell-based calpain cleavage assays, MN9D cells were transiently transfected with the above-stated eukaryotic expression vectors using Lipofectamine 2000 (Invitrogen). The cells were lysed in radioimmunoprecipitation assay buffer containing 50 mM Tris, pH 7.5, 150 mM NaCl, 1% Nonidet P-40, 0.25% sodium deoxycholate, and 0.1% SDS without protease inhibitors. Cells were homogenized in a Dounce homogenizer on ice and centrifuged at 13,000 ϫ g for 15 min at 4°C. Protein contents were measured using a Bio-Rad protein assay kit. Predetermined amounts (10 l for transcription and translation reactants and 100 g for cellular lysates) were incubated for 2 h at 37°C with up to 5.55 units of m-calpain or 3.95 units of -calpain from porcine erythrocytes (Calbiochem; catalog no. 208712) in calpain activation buffer at a final volume of 20 l. When necessary, a calpain inhibitor (500 M PD150606) was added to the reaction mixture. The reactions were stopped by adding 5 l of 5ϫ sample buffer (250 mM Tris-HCl, 500 mM DTT, 10% SDS, 0.5% bromphenol blue, and 50% glycerol, pH 6.8) followed by boiling for 5 min. The resulting products were separated by 12% SDS-PAGE and processed for autoradiography or immunoblot analysis.
Immunofluorescence-To label cytosolic Ca 2ϩ , MN9D cells treated with 5 g/ml ionomycin for 30 min or 50 M MPP ϩ for 36 h were stained with 3 M Fluo-3/AM dye (Molecular Probes) diluted in N2 medium for 30 min at 37°C in an incubator with a 10% CO 2 atmosphere and washed twice with N2 medium as suggested by the manufacturer's protocol. Subsequently, the stained cells were examined under a fluorescence microscope equipped with epifluorescence filters and a digital image analyzer (Axio Observer A1 Microscope; Carl Zeiss). For immunocytochemical localization of ASNA1, optineurin, and peripherin, MN9D cells were cultured at a density of 6.0 ϫ 10 4 cells on 25 g/ml poly-D-lysine-coated aclar film (Electron Microscopy Science, Fort Washington, PA). Thirty-six hours after 50 M MPP ϩ treatment in the presence or absence of 50 M calpeptin, cells were washed with PBS (Lonza, Verviers, Belgium) followed by fixation with 4% paraformaldehyde (Electron Microscopy Science) for 15 min at room temperature. The cells were blocked for 1 h at room temperature in PBS containing 5% normal goat serum (Invitrogen) and 0.2% Triton X-100. The cells were incubated with a goat anti-ASNA1 antibody (1:200; Everest Biotech), rabbit anti-optineurin antibody (1:200; Cay-man Chemical), or rabbit anti-peripherin antibody (1:200; Millipore Corp.) in PBS containing 1% normal goat serum and 0.2% Triton X-100 overnight at 4°C followed by incubation with Alexa Flour 488-conjugated rabbit anti-goat IgG (1:200) or Alexa Flour 488-conjugated goat anti-rabbit IgG (1:200; both from Molecular Probes) for 1 h at room temperature. For nuclear counterstaining, the cells were further incubated with 1 g/ml Hoechst 33258 (Molecular Probes) for 10 min at room temperature. Slides were mounted with Vectashield mounting medium (Vector Laboratories, Burlingame, CA) and observed under an LSM 510 META confocal laser scanning microscope equipped with epifluorescence filters and a digital image analyzer (Carl Zeiss).
Statistics-All data are expressed as the means Ϯ S.E. of the indicated number of experiments. Significant differences among groups were determined by one-way analysis of variance and Tukey's post hoc test using GraphPad Prism 6 (GraphPad Software, Inc., La Jolla, CA). A value of p Ͻ 0.05 was considered statistically significant.

Gel-based Protease Proteomics to Identify Putative Calpain
Substrates-We attempted to develop a simple 2DE-based procedure that can identify substrates that are directly cleaved by an applied protease. To accomplish this, cellular lysates obtained from MN9D dopaminergic neuronal cells were subjected to IEF, and the resulting strips were subsequently incubated with or without recombinant m-calpain (Fig. 1). Thereafter, the strips were further processed for conventional 2DE followed by a mass spectral analysis of the altered protein spots. This modified approach ensured that only calpain-specific substrates can be cleaved among immobilized proteins on a strip after IEF. First, we attempted to empirically set up conditions in which the applied amount of calpain can successfully cleave the known calpain substrate on a strip. Based on previous reports, including one from our laboratory demonstrating that Bax is cleaved by Ca 2ϩ -mediated calpain activation (15,16), isoelectrically separated proteins on a strip were incubated with m-calpain and processed for mini-2DE as described previously (14). Immunoblot analyses using anti-Bax antibody showed that incubation with a total of 5.55 units of recombinant m-calpain for 2 h was sufficient to cleave Bax and led to a disappearance of the Bax protein spot ( Fig. 2A). Using this condition, proteome analyses of calpain-treated samples and matching controls were performed. Representative Coomassie Brilliant Blue-stained protein profiles following IEF separation of 1.5-mg protein samples and incubation with or without a total of 83.25 units of recombinant m-calpain are shown in Fig. 2B. Approximately 1,500 distinct protein spots ranging from 10 to 100 kDa were typically resolved in pI 4 -7 linear gels. By using densitometric analysis of gel images, we initially identified more than 100 altered protein spots in calpain-treated samples as compared with expression levels in untreated samples. Among these altered protein spots, 25 protein spots with differential expression levels by at least 2-fold were subjected to mass spectral analysis. Information on these proteins is described in Table 1.
Validation of the Identified Calpain Substrates in Dopaminergic Neuronal Cells-Among these 25 altered proteins, we focused on protein spots that were significantly decreased or increased at least by 3-fold with p Ͻ 0.05 (n Ͼ 3). Fig. 2C showed representative close-up images of these protein spots boxed in Fig. 2B (also refer to Table 2). We then attempted to further verify each protein spot by performing repeated peptide mass fingerprint analysis, searching against different databases, considering the protein score, and referring to their actual location on the 2DE gel. To unambiguously identify putative calpain substrates, we used parameters for MALDI-TOF mass spectral analysis and mass spectrometry profiling as described under "Experimental Procedures." Table 2 summarizes the identified protein spots boxed in Fig. 2B and highlighted in Fig. 2C with accession number, molecular mass with pI, fold change in intensity over the untreated control spot, MOWSE score, number of the matched peptides, and sequence coverage. We then searched the available database that predicts putative protease substrates. None of these six proteins were listed as putative calpain substrates from database searches through Peptidecutter and PESTfind (data not shown). In contrast, CaMPDB (17) predicts that ASNA1, optineurin, and peripherin each contain a few sites that could possibly be cleaved by calpain. Therefore, we examined whether these three proteins may be endogenous calpain substrates by performing in vitro calpain cleavage assays using metabolically [ 35 S]methionine-radiolabeled proteins. As shown in Fig. 3A, ASNA1 was directly cleaved by both m-and -calpain, generating prominent fragments that ran just below the full-length form. This calpain-mediated cleavage was blocked in the presence of PD150606, a cell-permeable, selective, non-peptide Ca 2ϩ -dependent calpain inhibitor (18). For optineurin, treatment with m-or -calpain generated several smaller fragments. Again, this phenomenon was largely blocked by PD150606 (Fig. 3B). As shown in Fig. 3C, external addition of m-or -calpain also caused generation of several smaller peripherin bands along with two prominent fragments that were blocked in the presence of PD150606. To further confirm these findings, MN9D dopaminergic neuronal cells were transiently transfected with eukaryotic expression vectors containing T7-ASNA1-FLAG tagged, optineurin-GST tagged, or T7-peripherin-FLAG tagged sequences. Lysates obtained from transfected cells were treated with m-or -calpain in the presence or absence of PD150606 and subjected to immunoblot analyses using anti-T7, -GST, and -FLAG antibodies (Fig. 4). Immunoblot analyses for fodrin were run in parallel to confirm the state of calpain activation in each sample (see the middle panels of Fig. 4, A-C). Indeed, both m-and -calpain cleaved the transiently expressed proteins, and this cleavage was significantly blocked by PD150606. The calpain-mediated cleavage pattern of transiently expressed ASNA1 was found to be quite similar to that observed with the in vitro calpain cleavage assay (compare Fig. 4A with Fig. 3A). However, the cleavage patterns for optineurin and peripherin were somewhat different (compare Fig. 4, B and C, with Fig. 3, B and C, respectively). In any case, calpain-mediated cleavage of the transiently expressed proteins was significantly decreased in the presence of PD150606. To further examine the dose dependence of calpain-mediated cleavage of the identified substrates, we performed cell-based cleavage assays with increasing doses of m-calpain (1.11 to 5.55 units; Fig. 4, D-F). Dose-dependent cleavage of all three proteins was significantly blocked by cotreatment with PD150606 or calpeptin, another membranepermeable calpain inhibitor. Taken together, these data indicate that all three identified proteins are most likely novel substrates that are directly cleaved by both m-and -calpain.
Ionomycin-and MPP ϩ -induced Cleavage of ASNA1, Optineurin, and Peripherin-Dysregulation of intracellular free Ca 2ϩ causes perturbed energy metabolism, oxidative stress, and neuronal death (19,20). Alterations in Ca 2ϩ homeostasis lead to pathologic activation of calpain, and inhibition of calpain is considered a therapeutic target for a number of neurodegenerative disorders, including PD (5,21,22). We therefore investigated whether increased intracellular Ca 2ϩ in drug-induced neurodegeneration models was associated with cleavage of the newly identified substrates. To accomplish this, MN9D dopaminergic neuronal cells were first treated with ionomycin, an ionophore produced by Streptomyces conglobatus that is widely used to increase intracellular free Ca 2ϩ and subsequently activates calpain (23). Treatment with 5 g/ml ionomycin for 30 min apparently increased the number of Fluo-3-positive cells (Fig.   FIGURE 1. Schematic flow of a gel-based protease proteomic study to identify putative calpain substrates. We developed a gel-based protease proteomic strategy for identifying putative calpain substrates. After IEF separation, immobilized proteins on DryStrips were incubated with (A) or without (B) purified recombinant m-calpain followed by separation by SDS-PAGE. Altered protein spots in Coomassie Brilliant Blue G-250-stained gels were subjected to in-gel digestion followed by MALDI-TOF mass spectrometry. Lists of putative calpain substrates were made after searching through databases as described in the text. Using both culture and rat brain disease models, the identified candidates were subjected to further validation as calpain substrates.
5A) and led to Ca 2ϩ -activated, calpain-dependent cleavage of fodrin that was completely inhibited in the presence of 50 M calpeptin or 40 M BAPTA/AM, a membrane-permeable Ca 2ϩ chelator (Fig. 5B). As shown in Fig. 5C, levels of full-length forms of ASNA1 and optineurin were decreased following ionomycin treatment. In the case of peripherin, many smaller fragments appeared following ionomycin treatment. Notably, statistical analysis indicated that these ionomycin-induced changes were significantly attenuated in the presence of calpeptin or BAPTA/AM (Fig. 5D).
Neurotoxin-based experimental models have been widely used to mimic biochemical changes that occur in PD patients (1). Among these, systemic administration of the mitochondrial complex I inhibitor MPTP to non-human primates and mice and treatment of cells with its active metabolite MPP ϩ have been used to establish experimental models of PD. Previous reports in our laboratory indicate that MPP ϩ increases intra-cellular Ca 2ϩ and causes subsequent calpain activation in cultured cells (15,24). Furthermore, we reported that MPP ϩinduced cell death does not activate caspase in MN9D dopaminergic neuronal cells as well as in primary cultures of dopaminergic neurons (25). Therefore, we attempted to investigate whether MPP ϩ treatment causes calpain-mediated cleavage of the three newly identified substrates. Again, increases of intracellular Ca 2ϩ and calpain-mediated fodrin cleavage and decreased levels of ASNA1 and optineurin full-length forms were detected in MN9D cells after 50 M MPP ϩ treatment for 36 h (Fig. 6, A-D). Statistical analyses indicated that MPP ϩmediated decrease in ASNA1 and optineurin full-length forms were significantly attenuated in the presence of calpeptin (Fig.  6, C and D). MPP ϩ did not lead to any detectable change in Apaf-1 expression level, indicating that drug-induced decrease of ASNA1 and optineurin levels was not simply due to cell loss (data not shown). Quite similar to the pattern observed in iono-FIGURE 2. Representative 2DE gel images of putative calpain substrates. A, mini-2DE analysis was carried out to confirm that calpain effectively degraded Bax, one of its known substrates (15,16). Lysates (100 g) obtained from MN9D cells were loaded on 7-cm Drystrips and separated by IEF ranging from pI 4 to 7. Next, samples in the ministrips were incubated with 5.55 units of m-calpain for 2 h at 37°C, subjected to 12% SDS-PAGE, and probed with anti-Bax antibody. Actin was used as a loading and a negative control. B, representative gel images are shown for samples treated with or without m-calpain and further processed for 2DE (24 cm, pI 4 -7, and 8 -18%). Briefly, total protein lysates (1.5 mg) obtained from MN9D cells were separated by IEF and incubated with or without 83.25 units of m-calpain for 2 h at 37°C. After separation by SDS-PAGE, the gels were stained with 0.1% Coomassie Brilliant Blue G-250. Approximately 1,500 protein spots ranging from 10 to 100 kDa were routinely detected. C, comparative close-up views of the typically altered protein spots boxed in Fig. 2B, before and after calpain treatment. Arrows indicate the protein spots that were either increased (left) or decreased (right) following calpain treatment. Detailed molecular information for these protein spots is shown in Table 2. mycin-treated cells, MPP ϩ treatment for 36 h led to the generation of several peripherin fragments with smaller molecular weights, and this was significantly blocked in the presence of calpeptin (Fig. 6E). Furthermore, MPP ϩ -induced alterations of these three substrates were also significantly blocked in the presence of PD150606 (Fig. 7). In sum, we demonstrated that all three endogenous substrates were processed by drug-induced calpain activation in MN9D cells. We found that the processing patterns generated by ionomycin or MPP ϩ treatment were quite similar.
We next used immunofluorescent labeling to measure the cellular distribution patterns of these substrates before and after MPP ϩ treatment. Overall, the punctate staining intensities of ASNA1 and optineurin were apparently reduced in all areas of the cells upon MPP ϩ treatment, and these changes were attenuated in the presence of calpeptin (Fig. 8A, left and  middle columns). Intriguingly, the diffuse cytosolic staining pattern of peripherin in untreated control cells shifted to the cytoplasmic membrane after MPP ϩ treatment, and this stain-ing pattern of peripherin was restored in MN9D cells co-treated with calpeptin (Fig. 8A, right column). Based on previous observations demonstrating that expression levels of some proteins in MPP ϩ -treated MN9D cells can be regulated at the transcriptional level (26), we carried out semi-quantitative RT-PCR to compare their mRNA levels in cells treated with or without MPP ϩ . As shown in Fig. 8B, no discernible changes were found, indicating that MPP ϩ -mediated calpain activation is responsible for decreased levels and altered cellular distribution of these endogenous substrates without directly affecting their transcription rates.
Cleavage of ASNA1, Optineurin, and Peripherin in Rat Disease Models-We then tested whether these phenomena could be recapitulated in rat disease models of neurodegeneration. Increased intracellular Ca 2ϩ level caused by disturbance of Ca 2ϩ homeostasis is a major change that occurs following ischemic stroke. This leads to constitutive activation of calpain and accompanying excessive proteolysis that result in neuronal   DECEMBER 20, 2013 • VOLUME 288 • NUMBER 51 cell death (13,27). Therefore, we examined the calpain-induced cleavage of ASNA1, optineurin, and peripherin in the cortical tissue in MCAO rats. For actual experiments, rats received intracerebroventricle infusion of either 50 g of calpeptin dissolved in 5 l of Me 2 SO or Me 2 SO alone 30 min before MCAO (Fig. 9A). Immunoblot analysis using anti-fodrin antibody showed that the ipsilateral hemisphere showed the calpaincleaved fodrin band, suggesting a certain degree of calpain activation. We also found that full-length optineurin was significantly decreased in the ipsilateral cortex 18 h after ischemic injury, whereas decreases in ASNA and peripherin were not apparent (data not shown). However, all of these proteins were clearly decreased in the ipsilateral cortex 24 h after ischemic injury (Fig. 9B). To unambiguously confirm that this phenomenon is a calpain-dependent event, rats received intracerebroventricle infusion of calpeptin before MCAO. As shown in Fig. 9 (B and C), MCAO-induced change in all three substrates was significantly blocked in the ipsilateral cortex infused with calpeptin. Taken together, these data indicate that ASNA1, optineurin, and peripherin cleavage occurred following brain injury that disturbs Ca 2ϩ homeostasis and increases calpain activation in vivo. As shown in Fig. 10, MPP ϩ injection into the medial forebrain bundle for 3 days caused appearance of calpain-cleaved fodrin in the substantia nigra pars compacta, indicating calpain activation. In the ipsilateral side of stereotaxic injection model, the full-length forms of ASNA1 and optineurin were reduced, whereas several peripherin fragments with smaller molecular sizes were detected, supporting calpain-mediated alterations of the newly identified substrates. At present, it was not clearly understood how different cleavage products were observed when metabolically labeled substrates or cell lysates were incubated with calpain versus when cultures were treated with drugs or rat brain disease models were established. Search through the Peptidecutter database indicates that all of these identified substrates can be cleaved by several other proteases, explaining a partial blockage of drug-or surgery-mediated alterations of the putative substrates by calpain inhibitors. Similarly, this may explain why the cleaved fragments of the identified substrates are not detected in our culture models of neurodegeneration and rat brain disease models. Although it is highly speculative, therefore, complex regulation of protease activation in neurotoxin-based culture models and in rat neurodegenerative models may be responsible for different patterns of cleavage. Potentially Protective Effect for ASNA1, Optineurin, and Peripherin in MPP ϩ -induced Cell Death-Previous reports have indicated that calpain activation following neurotoxin challenge is related to various cell death signaling pathways and that many calpain-cleaved substrates play a critical role in maintaining cell survival under physiological conditions (23,28,29). We have also demonstrated that MPP ϩ -mediated cell death is blocked in the presence of calpeptin and overexpressed calbindin-D28K (15,24). We therefore examined whether MPP ϩ -mediated cleavage of these identified substrates causes loss of function or has any pathophysiological consequences. Eukaryotic expression vectors containing ASNA1, optineurin, or peripherin cDNA sequences were transiently expressed in MN9D cells, and its expression level was confirmed by immu-   DECEMBER 20, 2013 • VOLUME 288 • NUMBER 51 noblot analysis (Fig. 11A). Although it varied depending on each transient transfection trial, expression levels of these exogenous ASNA1, optineurin, and peripherin were at least 2-fold higher as compared with levels of the corresponding endogenous proteins. These cells were exposed to 50 M MPP ϩ for various time periods and subjected to MTT reduction assay. Data indicated that ASNA1 or peripherin overexpression protected cells from MPP ϩ -induced death, and the protection rate was statistically significant at 36 h post-treatment (Fig. 11, B  and D). In contrast, overexpressed optineurin provided slight but meaningful protection at only 12 h after MPP ϩ treatment (Fig. 11C). To further confirm the protective effect of optineurin, MN9D cells were transiently transfected with the construct containing the shRNA sequence for optineurin. Partial knockdown of optineurin accelerated MPP ϩ -induced cell death at 12 and 24 h post-treatment (Fig. 11, A and E). A similar protective pattern of all substrates was observed in neutral red-based colorimetric assays and morphological examination under a phase contrast microscope (data not shown). Considering the notion that the consequences of protease-mediated cleavage of proteins can be gain of function or loss-of-function or sometimes just bystander events (30), further biochemical studies remain to be performed to clearly argue that calpain-mediated cleavage of ASNA1, peripherin, and perhaps optineurin may be indeed associated with the progression of MPP ϩ -induced neurodegeneration.

DISCUSSION
Regardless of recent progress in our understanding, attempts to elucidate the molecular mechanisms by which protease activation induces cell death have been greatly hampered by the lack of and/or ineffectiveness of systematic and broadly applicable strategies to identify endogenous substrates that underlie neuronal survival/death mechanisms. To date, several screening methods have been attempted to identify protease substrates, including the expression cloning strategy (31), the yeast two-hybrid screening method (32), and more recently, pro- After incubation, cellular lysates (50 g) from each condition were subjected to SDS-PAGE followed by immunoblot analysis using the indicated antibodies. Ionomycin-induced calpain activation was confirmed with the anti-fodrin antibody. Actin was used as a loading control. C, MN9D cells were incubated for 30 min at 37°C with ionomycin (5 g/ml) in the presence or absence of 50 M calpeptin or 40 M BAPTA/AM. To measure levels of ASNA1, OPTN, and PERI, cellular lysates (100 g) were subjected to immunoblot analysis using the indicated antibodies that specifically recognize the endogenous protein. D, after normalization to actin in each condition, the relative change in fold intensity of the full-length (for ASNA1 and optineurin) was calculated over the untreated control (value ϭ 1). Changes in fold intensity of total peripherin fragments were determined over the ionomycin-treated sample (value ϭ 1). Each bar represents the mean Ϯ S.E. from three independent experiments. ***, p Ͻ 0.001. Cont., no treatment; I, ionomycin; C, calpeptin; B, BAPTA/AM. teomic identification (for review, see Ref. 33). For proteomic approaches, both gel-based methods (e.g., one-and two-dimensional analyses) and gel-free proteomic techniques (e.g., negative and positive selections of N-terminal identifications, combined fractional diagonal chromatography) have been applied (34,35). Many of these approaches have proved to be effective in that tens and hundreds of putative substrates have been identified. However, only a fraction has been confirmed as protease substrates, and its valid application remains to be carefully examined in pathophysiologically relevant disease models. A great majority of the proteomic analyses employed samples that were prepared either by direct addition of protease into lysates (36,37) or using protease knockdown systems (38). However, these approaches may also identify substrates that are cleaved as a consequence of sequential activation of other downstream proteases (39).
In the present study, we described a novel protease proteomic analysis based on conventional, gel-based 2DE, that only identifies putative substrates that are directly cleaved by calpains. This was achieved by incubating immobilized protein spots on IEF strips with calpains under specifically optimized conditions. This protocol ensures that only the direct substrates of calpain are cleaved, eliminating the possibility of identifying substrates that may be cleaved by combinatorial or sequential activation of proteolytic enzymes. Using cellular lysates obtained from the MN9D dopaminergic neuronal cell line that has been used as a culture model of PD, we initially found more than 100 altered protein spots and identified 25 proteins that were differentially expressed by at least 2-fold following calpain treatment. Among these spots, we selected three proteins: ASNA1, optineurin, and peripherin, that were not previously identified as calpain substrates. We performed in vitro and cell-based calpain cleavage assays and confirmed that these three substrates are cleaved by exogenous calpain. Furthermore, calpain-mediated cleavage was also observed in cul-  DECEMBER 20, 2013 • VOLUME 288 • NUMBER 51 tured cells and two rat brain models of neurodegeneration. The existing literature indicates that many calpain substrates have a typical PEST sequence (for review, see Ref.

Proteomic Analysis for Identifying Novel Calpain Substrates
3). However, ASNA1, optineurin, and peripherin do not contain any PEST sequences. Based on previous reports indicating that several calpain substrates do not possess PEST sequences (40), it is plausible to assume that ASNA1, optineurin, and peripherin may have other calpain-specific cleavage sites. Although more detailed study may be needed to verify the cleavage and/or degradation patterns of these identified substrates in vivo as well as in vitro, the present data suggest that our protease proteomic analysis is useful for identifying novel calpain substrates involved in neurodegeneration. In our separate study using the same strategy, we have identified more than 46 novel caspase-3 substrates, 5 indicating potentially broad applicability of our gelbased protease proteomic approaches.
In addition to caspases, calpains are the most extensively characterized cysteine proteases and are involved in maintain-ing neuronal physiology and perpetuating pathophysiology (3). They have roles in several fundamental processes, including cell signaling, cytoskeletal remodeling, and cell death/survival. In various brain disorders including PD, Alzheimer disease, Huntington disease, stroke, and spinal cord injury, growing evidence supports the involvement of Ca 2ϩ -mediated, nonlysosomal calpain activation (5,21,(41)(42)(43). This notion is strongly supported by data obtained from the brains of patients with neurodegenerative disorders, showing increased calpain activity compared with age-matched control brains. Consequently, dysregulated calpain activation in these pathological circumstances results in the cleavage of a number of crucial substrates that play important roles in maintaining neuronal structure and function (21), which can lead to injury progression. For example, p25, a proteolytic cleavage product of the neuron-specific activator of cyclin-dependent kinase 5 (Cdk5), has been found to accumulate in the brains of Alzheimer disease patients (44). In brain lysates treated with Ca 2ϩ or in primary cultures of cortical neurons challenged with excitotoxins, hypoxic injury, or amyloid-␤ peptide, Ca 2ϩ -activated calpain is responsible for  were separated and processed for immunoblot analysis using an anti-fodrin antibody. B-E, following drug treatment indicated in each lane, cellular lysates (100 g) were subjected to immunoblot analysis using the indicated antibodies. Relative fold intensity was calculated as described in the legend to Fig. 5. All values represent the means Ϯ S.E. from three independent experiments. ***, p Ͻ 0.001; **, p Ͻ 0.01. ns, not significant.
cleavage of p35 to p25 (23,45). Furthermore, calpain-mediated cleavage of p35 to p25 and subsequent Cdk5 activation has been implicated in cytoskeletal disruption and neuronal death, con-sequently leading to Alzheimer disease progression (23). Similarly, calpain-mediated conversion of p35 to p25 also plays a central role in MPTP-induced dopaminergic neuronal death  The rats were then exposed to MCAO for 1.5 h and subsequent reperfusion for 24 h. Tissue lysates (50 g) obtained from the contralateral (C) and the ipsilateral (I) side of the cortex to the right-sided endovascular MCAO were subjected to immunoblot analyses using the indicated antibodies. C, relative fold intensity of each protein after normalization to actin in each contralateral (C) and ipsilateral (I) sides of cortex was calculated over the contralateral side of Me 2 SO group (value ϭ 1). Each bar represents the mean Ϯ S.E. from four independent experiments. ***, p Ͻ 0.001; *, p Ͻ 0.05. via phosphorylation of critical factors and enzymes (46). More recently, it has been demonstrated that calpain-mediated generation of toxic fragments affects neuronal survival during glu-tamate receptor-mediated excitotoxicity (47). Therefore, understanding of how calpain-mediated cleavage of critical substrates is regulated is considered important for blocking dis- FIGURE 10. Cleavage of the identified substrates in MPP ؉ -stereotaxic injection model. A, at 3 days post-injection, the contralateral (C) and ipsilateral (I) side of tissue samples (substantia nigra pars compacta in midbrain) obtained from the MPP ϩ -stereotaxic injection rats were processed for immunoblot analyses using the indicated antibodies that recognize endogenous proteins. TH was used to assess the extent of dopaminergic neurodegeneration, and fodrin and GAPDH were used as a calpain activation and loading control, respectively. B, after normalization to GAPDH, relative change in fold intensity of the indicated protein was determined over the full-length protein in the contralateral side (for ASNA1 and OPTN; value ϭ 1) or the total fragments of the ipsilateral side (for PERI; value ϭ 1). Each bar represents the mean Ϯ S.E. from four independent samples. ***, p Ͻ 0.001.  OPTN (shOPTN, E) were treated with 50 M MPP ϩ for the indicated time periods. Viability was measured by the MTT reduction assay. Each bar represents the mean Ϯ S.E. from three independent experiments. Blank represents the cells transfected with an empty vector. The shRNA sequences that target firefly luciferase (shLuc) were used for negative control. ***, p Ͻ 0.001; **, p Ͻ 0.01; *, p Ͻ 0.05. ease progression, and inhibition of calpain activation has consequently emerged as a potential therapeutic target for neurodegenerative disorders (21).
All three identified proteins have been demonstrated to affect fundamental cellular functions such as cell growth, survival, and death. For example, ASNA1 is evolutionary conserved and homologous to Escherichia coli arsA, a well characterized ATPase involved in arsenite and antimonite efflux (48). It is found throughout the cell but is mainly localized in the cytoplasm. ASNA1 is also known as an essential ATPase for the insertion of tail-anchored proteins into endoplasmic reticulum membranes, including Golgi to endoplasmic reticulum traffic, and is a novel positive regulator of insulin secretion (49). Therefore, the depletion or down-regulation of full-length ASNA1 can significantly affect cell growth and survival/death. Optineurin is a ubiquitously expressed multifunctional cytoplasmic protein and is also localized in the Golgi apparatus, where it contributes to post-Golgi and vesicular trafficking and Golgi organization (50). Therefore, overexpression of optineurin and mutant optineurin (E50K) highlighted its potential roles in trafficking, Golgi integrity, and cell death (51). Indeed, it negatively regulates the induction of interferon-␤ in response to RNA virus infection and tumor necrosis factor-␣-induced nuclear factor-B activation by competing with NF-B essential modulator (NEMO) for polyubiquitinated receptor-interacting protein (RIP) (52). Recently, it was reported that optineurin enhances cell survival via reactive oxygen species-induced nuclear translocation of optineurin, whereas fragmentation of nigral neuron Golgi apparati typically found in optineurin mutants is increased in PD patients (53,54). Finally, peripherin is a type of neuronal intermediate filament that is mainly cytoplasmic and is associated with pathological aggregates in the motor neurons in patients with amyotrophic lateral sclerosis (55). However, it remains controversial whether peripherin is involved in regulating cell survival and death in various neural systems (56,57). Nevertheless, it has not been well understood how their physiological and pathological functions may be regulated. Based on data from both in vitro and in vivo experiments, we hypothesize that cleavage of these substrates may result in a loss of function phenomena that would associated with calpain-induced neuronal death. In this regard, it is necessary to carefully examine how calpain-mediated cleavage or degradation of ASNA1, optineurin, and peripherin affect their proposed cellular functions, especially in neurodegenerative conditions. Further investigation into the exact mechanisms behind the potential pro-survival roles of these identified proteins will help us to expand our understanding of pathophysiology of various neurodegenerative disorders.