Chaperone proteins identified from synthetic proteasome inhibitor-inducedinclusions in PC12 cells by proteomic analysis
Xing’an Li1,2, Yingjiu Zhang2, Yihong Hu1, Ming Chang1, Tao Liu3, Danping Wang1, Yu Zhang1, Lei Zhang1,and Linsen Hu1*1 Laboratory for Proteomics, Department of Neurology, The First Affiliated Hospital of Jilin University, Changchun 130021, China2 Key Laboratory for Molecular Enzymology and Engineering, Ministry of Education (Jilin University), Changchun 130021, China3 College of Life Science, Jilin University, Changchun 130021, China
Received: February 20, 2008 Accepted: March 18, 2008This work was supported by a grant from the Natural Science Foundationof Jilin Province (No. 200505200)*Corresponding author : Tel, 8 6-431-85 612419; Fax, 86-431-85637090; E-mail, [email protected]
Chaperone proteins are significant in Lewy bodies, but theprofile of chaperone proteins is incompletely unraveled.Proteomic analysis is used to determine protein candidatesfor further study. Here, to identify potential chaperoneproteins from agent-induced inclusions, we carried outproteomic analysis of artificially synthetic proteasomeinhibitor (PSI)-induced inclusions formed in PC12 cellsexposed to 10 μM PSI for 48 h. Using biochemicalfractionation, 2-D electrophoresis, and identification throughpeptide mass fingerprints searched against multiple proteindatabases, we repeatedly identified eight reproduciblechaperone proteins from the PSI-induced inclusions. Ofthese, 58 kDa glucose regulated protein, 75 kDa glucoseregulated protein, and calcium-binding protein 1 were newlyidentified. The other five had been reported to be consistentcomponents of Lewy bodies. These findings suggested thatthe three potential chaperone proteins might be recruited toPSI-induced inclusions in PC12 cells under proteasomeinhibition.
Lewy body (LB) diseases are neurodegenerative andinclude at least three clinical syndromes, idiopathicParkinson’s disease (PD), PD dementia, and dementia withLBs [1]. Ninety percent of PD cases occur sporadicallyand are characterized pathologically by cytoplasmic
inclusions, LBs stained with eosin or anti-α-synuclein (α-SYN) antibody [2,3], in substantia nigra pars compacta[4]. Although the direct role of LBs in the disease is still asubject of debate, the development of LBs is substantiallya process of protein aggregation related to the pathogenesisof PD [5,6]. Having similar molecular components as LBs[4], LB-like inclusions (LIs) have been described in somerare cases of neurodegenerative disease [7,8], and alsocreated in some animal models of PD by both inhibition ofmitochondria or proteasomes [9−11] and excessivetransgenic expression of human wild-type α-SYN [3,12].Based on attractive progress in the knowledge about thebiochemical mechanisms of LBs and LIs, other investiga-tors have attempted to replicate LBs in a variety of cellularmodels of PD using proteasome inhibitors [13−15]. Forexample, one report showed that artificially syntheticproteasome inhibitor (PSI) can induce a progressive celldeath coupled with appearance of cytoplasmic inclusionsin remaining cells [15]. These cell culture-based works donot only offer evidence to support the concept that LBscould represent aggresome-like structures, just likeaggresomes forming at the centrosome in response toproteolytic stress [16,17], but also extensively providealternative protein candidates associated with proteincomponents of LBs for further investigation [18].
As a constituent of the endoplasmic reticulum (ER)-associated degradation (ERAD) machinery in cytoplasm,proteasome is essential to prevent proteolytic stress, andproteasome inhibition can cause loss of ERAD leading toER stress [19−21]. Under the loss of ERAD, up-regulationof ER chaperone proteins in cells increases as a compen-satory mechanism to prevent protein aggregation [16,22].For example, proteins aggregated peripherally in cytoplasmare typically subjected to chaperone proteins such as heatshock proteins (HSPs). If the compensatory mechanism
is not effective, however, the aggregated proteins assistedwith HSPs are recruited in aggresomes where proteindegradation is enhanced [16]. HSPs are therefore termed“aggresome-related chaperone proteins” [18]. In the pro-tein composition of LBs, a considerable number ofcomponents have been identified as chaperone proteins[23], such as α B-crystallin, 70 kDa heat shock protein1A/1B (HSP70), 71 kDa heat shock cognate protein(HSC70), and 14-3-3 ζ [5,24,25].
Despite their common pathways of controlling proteinaggregation [16,17], chaperone proteins recruited toaggresomes are variable and the recruitments varydepending on the type of aggregated proteins, the state of“the host cell”, and the localization of chaperone proteinsin cytoplasm. Recently, two proteomic analyses haveindicated that a better way to understand potentialchaperone proteins in LBs and LIs is to examine therespective contents of their intermediate organelles [5,18].
In the present work, using 2-D electrophoresis followedby matrix-assisted laser desorption/ionization time-of-flightmass spectrometry (MALDI-TOF MS), we attempted tocharacterize the proteomic features of PSI-inducedinclusions purified from PC12 cells under proteasomeinhibition by biochemical fractionation. Then in theproteomic context we mainly focused on a portion ofchaperone proteins.
Materials and Methods
ChemicalsAll reagents of analytically pure and cell culture grade werepurchased from Amersham Biosciences (Uppsala, Sweden)unless specified otherwise. Artificially synthesized PSI,Z-lle-Glu (OtBu)-Ala-Leu-al or N-benzyloxycarbonyl-Ile-Glu(O-t-butyl)-Ala-Leu-al, was from EMD Biosciences (anaffiliate of Merck Chemicals, Darmstadt, Germany). Cellculture plastics, media, and related chemicals were fromGibco (Grand Island, USA). DNase I and RNase A werefrom TaKaRa Biotechnology (Dalian, China). Percoll (adensity of 1.131 g/ml) and proteinase inhibitors [4-(2-aminoethyl) benzenesulfonyl fluoride, pepstatin A, E-64,bestatin, leupeptin, and aprotinin] were from Sigma (St.Louis, USA).
Cell culture and PSI inductionPC12 cells (Cell Bank of the Chinese Academy of Sciences,Shanghai, China) were maintained in Dulbecco’s modifiedEagle’s medium supplemented with 10% fetal bovineserum (V/V), 20 g/L glutamine, 60 U/ml penicillin, and100 μg/ml streptomycin. To produce agent-induced
inclusions in cells under proteasome inhibition, PSI, whichblocks proteolytic activity of 26S proteasome withoutinfluencing its ATPase or isopeptidase activities and hasseveral features advantageous for cell biology [26], wasparticularly considered. Ten micromoles per liter of PSIwas selected by reference to a previous report [15]. Cellsin log phase were split to a density of 1−2×105 viable cellsper milliliter and further cultured for 24 h. After 48 hfollowing exposure to PSI in dimethylsulfoxide, cells werecollected by centrifugation at 836 g for 5 min. The eosi-nophilic feature of PSI-induced inclusions in the cells wasassessed by the hematoxylin-eosin (HE) method asdescribed in following session.
Purification of PSI-induced inclusionsPSI-induced inclusions were purified as describedpreviously [6,18,27−29] with some modifications, and allsubsequent steps of purification were carried out at 4 ºCunless specified otherwise. Briefly, the cells were collectedat the indicated time and washed in cold Tris-bufferedsaline (pH 7.4), then in cold 0.1×Tris-buffered salinecontaining 80 g/L sucrose. Cell cultures were treatedrepeatedly with liquid nitrogen two or three times thenhomogenized with lysing buffer [1 mM HEPES (pH 7.2),0 .5 mM MgCl 2, 0 .5% NP-40 (V /V ) , 0.1% β -mercaptoethanol (V/V), and 1% proteinase inhibitors (V/V)]. After suspended repeatedly by pipetting and shakenvigorously by hand, the lysate was incubated for 30 minat 37 ºC until cells were thoroughly homogenized. Initialpellets were collected by low centrifugation at 80 g for15 min, washed with buffer L (1 mM HEPES, 0.5 mMMgCl2, and proteinase inhibitors) on ice for 5 min, andrecollected by centrifugation at 836 g for 10 min. Theinitial pellets were incubated in 10×DNase I solution (200U/ml DNase I, 250 μg/mL RNase A, and proteinaseinhibitors) for 24 h, during which the initial pellets weresuspended repeatedly by pipetting and shaken vigorouslyby hand several times. The resulting pellets were collectedby centrifugation at 836 g for 10 min, washed with 50mM Tris-HCl buffer (STB; pH 7.4) containing 0.32 Msucrose supplemented with protease inhibitors, andrecollected by centrifugation at 4000 g for 10 min. Theeosinophilic feature of PSI-induced inclusions in theresulting pellets was assessed by the HE method asdescribed below. For further purification, the resultingpellets were diluted to 600 μl using 12% Percoll in STB(V/V) and overlaid on 600 μl of 35% Percoll in STB. Thematerial band just below the interface of the sample and35% Percoll was collected by centrifugation at 35,000 gfor 30 min, washed in 10 mM Tris (pH 8.0) containing
250 mM sucrose supplemented with proteinase inhibitors,and recollected by centrifugation at 4000 g for 30 min.After purification, the fraction of PSI-induced inclusionswere used for 2-D electrophoresis.
Assessment of PSI-induced inclusionsThe fraction of PSI-induced inclusions was stained withthe HE method and examined by an experienced observernot familiar with the sample identity. At the same time, thePSI-induced inclusions in the sections on slides werequantified. The number of PSI-induced inclusions wascounted in nine random fields (three fields per section,three different sections on slides) and expressed as apercentage of nucleus-free to total PSI-induced inclusions.The χ2-test was used to compare the percentages beforeand after the procedure of purification. P<0.05 wasaccepted as significant. Measurements were repeated atleast three times [30,31].
Protein extractionThe fraction of PSI-induced inclusions was frozen andthawed two or three times with liquid nitrogen. Lysis buffer[30 mM Tris, 7 M urea, 2 M thiourea, 40 g/L 3-[(3-cholamidopropyl)-dimethylammonio]-1-propanesulphonate(CHAPS), 60 mM dithiothreitol (DTT), 2% pharmalyte(pH 3.0−10.0; V/V), and proteinase inhibitors] was addedto a total volume of 250 μl, and then the mixture wasincubated at room temperature for 1 h followed bysonication at 35% of amplitude in an ice-cold water bath.Protein extracts were collected by centrifugation at 25,000g for 30 min, subjected to a 2-D clean-up kit following themanufacturer’s instructions (Amersham Biosciences), anddissolved with rehydration solution [8 M urea, 20 g/LCHAPS, 6 g/L DTT, and 0.5% immobilized pH-gradientbuffer (V/V)] at room temperature for 1 h.
As controls, the homogenates of whole cells withoutand with exposure to PSI before purification were alsoprepared using liquid nitrogen and lysis buffer, respectively,and two samples of proteins were extracted from theirrespective homogenates. Steps of homogenization andsubsequent steps of protein extraction were carried out asdescribed above.
2-D electrophoresis and image analysis2-D electrophoresis was carried out as described previ-ously [32−34] with some modifications. Briefly, approxi-mately 600 μg protein extracts of the purified fraction ofPSI-induced inclusions, quantified by Bradford assay, wereapplied to Immobiline™ DryStrip gel strips (24 cm, pH3.0−10.0, non-linear; Amersham Biosciences) then
reswollen at room temperature for 4 h. The first dimen-sion electrophoresis (or isoelectric focusing) was run onan Ettan IPGphor II isoelectric focusing unit (50 μA/strip;Amersham Biosciences) at 20 ºC for 17 h. The strips werethen equilibrated for 15 min in 50 mM Tris-HCl (pH 8.8)buffer [6 M urea, 20 g/L SDS, 30% glycerol (V/V), 20 g/LDTT, and a trace of bromophenol blue], re-equilibratedfor another 15 min in the same buffer with 40 g/Liodoacetamide but without DTT, and transferred onto thetop of 1 mm-thick separating SDS-polyacrylamide gels[1 g/L SDS, 125 g/L total gel concentration (T, acrylamideplus cross-linking agent), 2.6% cross-linking agent (C;W/W), 24 cm×20 cm]. Protein markers (14.40−97.00 kDa)were used to mark the molecular mass. The seconddimension electrophoresis was run on an Ettan DALT sixelectrophoresis unit (2 W/gel, 600 V, 400 mA; AmershamBiosciences) at 18 ºC overnight. The gels were incubatedin 200 g/L trichloroacetic acid (TCA) fixing solutionfor 1 h, stained in 2.5 g/L Coomassie brilliant blue R 250for 4 h, and destained in 10% (V/V) acetic acid until thegel background was clear.
Gels were scanned using a Umax CE scanner(Amersham Biosciences) with Image Master Labscanversion 3.01b (Amersham Biosciences). The images wereanalyzed with Image Master 2-D Evolution version 2003.02(Amersham Biosciences). Spots in images weredensitometrically measured and statistically evaluated bycomputer-assisted pattern analysis to detect whether spotsappeared as reproducibly significant in at least three of thefour gels. A spot was considered to be negligible in thepresent experimental condition if it was not detectable inthree of the four gels. These significant spots were selectedfor MS analysis. As controls, two samples of proteinsextracted from the homogenates of whole cells withoutand with exposure to PSI before purification were alsoseparated by 2-D electrophoresis as described above.
Protein identificationProteins were identified by MALDI-TOF MS peptide massfingerprints (PMF) as described [33−35], with somemodifications. With an Ettan Spot Picker roboticworkstation (Amersham Biosciences), spots (1.0 mmdiameter) were excised from gels. With an Ettan TADigester robotic workstation (Amersham Biosciences),spots were in turn destained with Wash I [50% methanol(V/V) containing 50 mM ammonium acid carbonate],dehydrated with Wash II (acetonitrile; ACN), desiccatedfor at least 1 h, and digested at room temperature with1 μg/μl modified porcine trypsin (dissolved in 20 mMammonium acid carbonate) overnight. Digestion was ended
with 50% ACN (V/V) containing 0.1% trifluoroacetic acid(TFA; V/V). After desiccation at room temperature for atleast 24 h, 0.3 μl digested peptide mixture [dissolved in asolution of 1:100:100 TFA:ACN:deionized water (V/V/V)]was spotted on the surface of a specific steel slide(Amersham Biosciences) and mixed with 0.3 μl of 4 μg/μlα-cyano-4-hydroxy-transcinnamic acid (dissolved in thesame solvent) with an Ettan Spotter robotic workstation.After desiccation at room temperature, the mixture wassubjected to MS analysis. PMF were produced with anEttan MALDI-TOF Pro workstation (AmershamBiosciences).
To acquire spectra of protein digests in positive reflectionion mode equipped with a 337 nm nitrogen laser, we setthe instrument parameters as follows: acceleration potentialat 20 kV; pulsed extraction at 2000 V; low mass ionrejection at m/z 500; laser mode of 8 shots per second;and 200 shots for each spectrum. To process the acquiredspectra, we set the instrument parameters as follows:algorithm mode at centroid; smooth spectra filter for noiseremoval; external calibration of the angiotensin III peak atm/z 897.5 and human adrenal cortex hormone fragment18−39 peak at m/z 2465.2; internal calibration of the trypsinautolysis peaks at m/z 842.50 and m/z 2211.10; mass rangeof peak detectable at m/z 800−2500; mass tolerance at0.2 Da; monoisotopic cut-off at m/z 3000; and the baselineadjusted automatically. To identify the spectra processed,we set instrument parameters as follows: one missedcleavage site per peptide allowed at most; complete aminoacid modification of iodoacetamide; partial amino acidmodification of oxidation; search type of PMF; ProFoundsearch engine; and a maximum expectation of 0.05 and a
minimum of 20% coverage of matched peptides.Submission of PMF to the ProFound search engine (fullyintegrated in the Ettan MALDI-TOF Pro workstation oravailable at http://prowl.rockefeller.edu/prowl-cgi/profound.exe) against the database of NCBInr (http://www.ncbi.nlm.nih.gov) led to initial identification.Submission of PMF to the Mascot search engine againstthe NCBInr, SwissProt, and MSDB (http://www.matrixscience.com/search_form_select.html) databasesenhanced the accuracy of the initial identification.
Results
Eosinophilic feature of the PSI-induced inclusions andevaluation of the processes of purificationTo examine whether cytoplasmic PSI-induced inclusionsare formed in PC12 cells under proteasome inhibition, wedetected the eosinophilic feature of PSI-induced inclusionsusing the HE method. Compared to normal cells (data notshown), as expected, the PC12 cells under proteasomeinhibition were characterized by cytoplasmic PSI-inducedinclusions stained with eosin [Fig. 1(A)]. Furthermore,similar to ubiquitin/α-SYN-positive inclusions formed inPC12 cells exposed to PSI for 24 h [15], the PSI-inducedinclusions displayed a focal, homogeneous morphology,and appeared as two styles indicative of nucleus-bindingPSI-induced inclusions and nucleus-free PSI-inducedinclusions. In addition, some of the PSI-induced inclusionswere observed in the cytoplasm of remaining cells, whereasthe others were extruded into the extracellular spacefollowing destruction of the host cells [36].
To examine whether the pure intact PSI-induced
Fig. 1 Hematoxylin−eosin staining of synthetic proteasome inhibitor (PSI)-induced inclusions in PC12 cells (A) PC12 cells exposedto PSI (10 μM) for 48 h are shown with two styles of inclusions, nucleus-binding PSI-induced inclusions (open arrows) and nucleus-free PSI-inducedinclusions (closed arrows) (Magnification, 200×; inset shows an enlarged view, 400×). (B) PC12 cells exposed to PSI after purification are shownwith nucleus-free PSI-induced inclusions (closed arrow) (Magnification, 200×; inset shows an enlarged view, 400×). (C) Before and after purification,PC12 cells exposed to PSI are shown with their respective percentages of nucleus-free to total PSI inclusions. *P<0.05.
Fig. 2 Two-dimensional electrophoresis gel map of synthetic proteasome inhibitor (PSI)-induced inclusion proteins in PC12 cells(A) Representative 2-D electrophoresis gel map of the PSI-induced inclusions isolated from cells exposed to PSI. The circled and numbered spotsrepresent the spots of the eight chaperone proteins located in the 2-D gel: No. 1, 58 kDa glucose-regulated protein (GRP58); No. 2, 75 kDa GRP(GRP75); No. 3, 27 kDa heat shock protein 1 (HSP27); No. 4, valosin-containing protein (VCP); No. 5, 70 kDa heat shock protein 1A/1B(HSP70); No. 6, 14-3-3 ζ; No. 7, calcium-binding protein 1 (CaBP1); No. 8, heat shock cognate protein (HSC70). (B) Detail maps of the eightchaperone proteins. IEF, isoelectric focusing; NL, non-linear; SDS-PAGE, sodium dodecyl sulfate–polyacrylamide gel electrophoresis.
inclusions were successfully isolated from the PC12 cellsby the procedure of purification, we also detected theeosinophilic feature with the HE method after each fractionof PSI-induced inclusions was prepared from eachsubsequent process of purification. First, an abundanceof intact PSI-induced inclusions were enriched in initialpellets by a process of purification, that is, incubation ofthe cells with lysing buffer containing 0.5% NP40 andcentrifugation at 80 g (data not shown), applied to producea centrosome-enriched fraction or an α-SYN aggregate-enriched fraction [6,18,28]. Besides a majority of the twostyles of PSI-induced inclusions, some particles, such aslarger subcellular components, heavier cellular debris, andcytoplasm membrane fragments, also resided in theinitial pellets. Second, nucleus-free PSI-induced inclusionswere enriched in the resulting pellets by the next processof purification, incubation of the initial pellets with10×DNase I solution containing DNase I mixed withRNase A [Fig. 1(B)], applied to degrade nuclei in cells[29,37]. Compared to 51% of the nucleus-free PSI-induced inclusions to total PSI-induced inclusions innormal cells, the percentage in the resulting pellets collectedfollowing the two processes of purification was signifi-cantly increased to 97% (P<0.05) [Fig. 1(C)]. Finally,
the articles described above were eliminated by a thirdprocess of purification, separation of the resulting pelletswith centrifugation at 35,000 g in discontinuous Percoll-mediated density gradients (data not shown), applied topurify organelles from EpH4 cells and LBs from braintissue [29,38]. After the three processes of purification,as expected, an abundance of pure intact PSI-inducedinclusions were isolated.
2-D electrophoresis gel map of PSI-induced inclusionsand selection of specific protein spotsTo examine whether proteins extracted from the pureintact PSI-induced inclusions were usable for cell-freeassay [38], we carried out 2-D gel-based analysis of thePSI-induced inclusions. As with the two samples ofproteins extracted from homogenates of whole cellswithout and with exposure to PSI, which was easily carriedout by 2-D electrophoresis (data not shown), the sampleof proteins extracted from the purified fraction of PSI-induced inclusions was also achieved without difficulty[Fig. 2(A)]. The purified fraction of PSI-inducedinclusions was dissolved with lysing solution, rehydratedwith rehydration solution, and resolved on 2-D gel.Further, compared to the two protein spot patterns on the
2-D gel, indicative of their respective total proteins of wholecells without and with exposure to PSI, the protein spotpattern on the 2-D gel indicative of the purified fractionof PSI-induced inclusions was significantly different [38,39]. That is, the 2-D gel-based performance of the purifiedfraction of PSI-induced inclusions led to protein mapestablishment of the PSI-induced inclusions. Thus,following the processes of purification, the 2-D electro-phoresis system allowed us to build up a desirable map ofthe PSI-induced inclusion proteins within a molecularmass range of 14.40−97.00 kDa and an isoelectric point(pI) range of 3−10. Of all protein spots focalized withincreasing clarity in four gels, 114 spots were observedreproducibly, and each spot in at least three of the fourgels was not found to be significantly different in termsof appearance, disappearance, and shift. After the 114 spotswere evaluated by proteomic analysis (as indicated below),eight specific spots were focused on, because of thespecific properties of their chaperone proteins [17,23].Coincidently, the eight specific spots were observed tofocalize more obviously and distinctly than most otherspots [Fig. 2(B)], except for two that were close to theleft and right sides of one of the eight spots and alsoobserved to focalize clearly and distinctly on the 2-D gel[Fig. 2(A), No. 6].
MALDI-TOF MS of eight specific proteins andidentification using PMFTo examine whether the 114 selected spots were able tobe assigned as proteins by MS analys is-ba sedidentification, we detected production of PMF andsubsequent identification through PMF. For unequivocalidentification, we considered sequence coverage of at least20%, the expectation of 0.05 at maximum, at least fivematching peptides, and a gap of at least three peptidesbetween the accepted protein candidate and the firstexcluded one in the list of protein candidates provided bythe NCBInr database [40]. After the selected 114 spotswere excised from the 2-D electrophoresis gel and digestedwith trypsin, and MALDI-TOF MS was carried out andthe spots were identified using PMF, we attempted tocharacterize the proteomic features of the PSI-inducedinclusions. Based on the proteomic context, eight specificproteins were assigned as chaperone proteins and wereexpected to be stressed because of their aberrantexpression in cells faced with an environment of proteolyticstress [4,16,17,23]. For example, the PMFs of eightspecific proteins were produced, and characterized bytheir specific MS patterns, their highly reproducible m/zof ion signals, and their relative intensities of ion signals
(Fig. 3). Following submission of the PMFs to theProFound search engine against the NCBInr database, theeight specific proteins were initially identified as chaperoneproteins and characterized by their respective groups ofidentification data (Table 1). Each of the eight chaperoneproteins was identified four times and each identificationwas shown as a comparable result (Table 1). The eightidentified chaperone proteins were: 58 kDa glucose-regulated protein (GRP58) [Fig. 2(B), No. 1; Fig. 3, No.1]; 75 kDa GRP (GRP75) [Fig. 2(B), No. 2; Fig. 3, No.2]; 27 kDa HSP 1 (HSP27) [Fig. 2(B), No. 3; Fig. 3, No.3]; valosin-containing protein (VCP) [Fig. 2(B), No. 4;Fig. 3, No. 4]; HSP70 [Fig. 2(B), No. 5; Fig. 3, No. 5];protein kinase C inhibitor protein 1 (KCIP-1, or 14-3-3 ζ)[Fig. 2(B), No. 6; Fig. 3, No. 6]; calcium-binding protein1 (CaBP1) [Fig. 2(B), No. 7; Fig. 3, No. 7]; and HSC70[Fig. 2(B), No. 8; Fig. 3, No. 8].
To further enhance the accuracy of the initial identifi-cation through PFM, we applied the probability-basedMowse score to evaluate the consistency of identificationin the multiple protein databases NCBInr, SwissProt, andMSDB. In this analysis, if the value of the top score for aprotein candidate is greater than the level of significancethreshold (61 in NCBInr, 51 in SwissProt, and 56 inMSDB) and at the same time the value of the runner-upfor another candidate protein is less than the level(P<0.05), the protein candidate with the top score is iden-tified as the accepted protein candidate (or termed the“the protein of interest”). For each of the eight chaperoneproteins, the PMF coupled with the most desirable groupof identification data in the four comparable results(indicated in bold text in Table 1) was selected to be a hitin the multiple protein databases by the probability-basedMowse score. For example, following submission of thePMFs to the Mascot search engine against one of themultiple protein databases, SwissProt (significance thres-hold of 51), the eight chaperone proteins were identifiedin the following way: GRP58 was assigned as proteindisulfide-isomerase A3 precursor (PDIA3_RAT) based ona top score of 186 (>51) compared with 33 (<51) of therunner-up (Fig. 4, No. 1); GRP75 was assigned as stress-70 protein, mitochondrial precursor (GRP75_RAT) basedon a top score of 154 (>51) compared with 34 (<51) ofthe runner-up (Fig. 4, No. 2); HSP27 was assigned asHSP β-1 (HSPB1_RAT) based on a top score of 79 (>51)compared with 30 (<51) of the runner-up (Fig. 4, No. 3);VCP was assigned as transitional endoplasmic reticulumATPase (TERA_RAT) based on a top score of 101 (>51)compared with 36 (<51) of the runner-up (Fig. 4, No. 4);HSP70 was assigned as 70 kDa HSP (HSP71_RAT) based
Fig. 3 PMFs of the eight chaperone proteins expanded with their respective specific mass spectroscopy (MS) patterns, and identi-fication using PMF leading to assignment Intensity, ion abundance of MS; m/z, mass-to-charge.
on a top score of 170 (>51) compared with 27 (<51) ofthe runner-up (Fig. 4, No. 5); 14-3-3 ζ was assigned as14-3-3 protein ζ (1433Z_RAT) based on a top score of 69(>51) compared with 33 (<51) of the runner-up (Fig. 4,No. 6); CaBP1 was assigned as protein disulfide-isomeraseA6 precursor (PDIA6_RAT) based on a top score of 84(>51) compared with 33 (<51) of the runner-up (Fig. 4,No. 7); and HSC70 was assigned as 71 kDa HSC protein(HSP7C_RAT) based on a top score of 249 (>51)compared with 47 (<51) of the runner-up (Fig. 4, No. 8).Similarly, following submission of the PMFs to the Mascotsearch engine against the other two multiple proteindatabases, NCBInr (significance threshold of 61) andMSDB (significance threshold of 56), the eight chaperoneproteins were shown to have comparable results (datanot shown). So, after identification through PMF wasenhanced by analysis using the probability-based Mowsescore, the eight chaperone proteins were ultimatelydetermined. Of the eight chaperone proteins, HSP27, VCP,HSP70, 14-3-3ζ, and HSC70 had been reported to beconsistent components of classical LBs in brainstem andcortex by immunostaining [5,12,18,25,41], but GRP58,GRP75, and CaBP1 had not been reported as associated
components of LBs.In addition, remarkably, based on the proteomic context
of PSI-induced inclusions, another 15 proteins that hadbeen reported to be consistent protein components of LBswere identified. These include Cu, Zn superoxidedismutase, tubulin α1C, tubulin β5, heme oxygenase-1(HO-1), creatine kinase-B, ubiquinol-cytochrome creductase core protein I, ATP synthase β subunit, tyrosine-3-monoxygenase activation protein epsilon (or 14-3-3 ε),tyrosine hydroxylase, proteasome subunit β type 5,proteasome 26S subunit ATPase2 (PSMC2), proteasome26S subunit ATPase5 (PSMC5), proteasome 26S subunitATPase6 (PSMC6), proteasome 26S subunit non-ATPase-11 (PSMD11), and proteasome 26S subunit non-ATPase13(PSMD13) [9,12,18,29,42−46]. Of the 15 consistentcomponents of LBs, both 14-3-3 ε [close to the left sideof 14-3-3 ζ on the 2-D gel indicated in Fig. 2(A), No. 6]and HO-1 (close to the right side of 14-3-3 ζ on the same2-D gel), were shown with their respective identificationdata from the NCBInr database [identification data of14-3-3 ε: gi|13928824|ref|NP_113791.1 (accession No.),0.014 (expectation), 25.5 (coverage), 4.6 (pI), 29.27 (mass),11 (measured peptides), and 6 (matched peptides); identi-
Fig. 4 Accuracy of identification enhanced by analysis of probability-based Mowse score Left number is indicative of the value of therunner-up, the middle number is indicative of the significance threshold, and the right number is indicative of the value of the top score. Probability-based Mowse score: protein score is −10×Log (P), where P is the probability that the match is a random event. CaBP1, calcium-binding protein 1;GRP58, 58 kDa glucose-regulated protein; GRP75, 75 kDa glucose-regulated protein; HSC70, heat shock cognate protein; HSP27, 27 kDa heatshock protein 1; HSP70, 70 kDa heat shock protein 1;VCP, valosin-containing protein.
Table 1 List of eight chaperone proteins identified from synthetic proteasome inhibitor-induced inclusions in PC12 cells
Identification of each chaperone protein was repeated four times. Bold numbers indicate the desirable group of protein data in the four comparableresults. Accession no., accession number of a protein candidate in the NCBInr protein database; coverage, the ratio of the protein sequence coveredby matched peptides; expectation, chance of incorrect identification; pI, isoelectric point.
the PSI-induced inclusions are characterized and catego-rized as the functional classes of antioxidant defense, cyto-skeleton system, metabolism and mitochondrial function,neurotransmission, ubiquitin proteasome system, andprotein folding and transport (Table 2) [4,5,12,18].
Subcellular proteomic analysis is an efficient approach toreveal potential components of organelles, and proteomicanalysis of pure intact organelles is able to provide a contextin which interesting proteins associated with intracellularenvironments are able to be focused on. In the presentwork, we recapitulate cytoplasmic PSI-induced inclusionsin PC12 cells under proteasome inhibition. Usingbiochemical fractionation, 2-D electrophoresis, and proteinidentification through PMF, we attempted to characterizethe proteomic features of the PSI-induced inclusions. In
the proteomic context we mainly focused on the profile ofpotential chaperone proteins.
We developed a novel procedure of purification to isolatethe PSI-induced inclusions from PC12 cells. Although theprocesses of purification are subjected to contaminationof other cytoplasmic proteins, many laboratories began tocombine traditional purification procedures with alternativemethods because of the impossibility of complete purifi-cation [38]. Incubation of the cells with lysing buffersolution followed by low centrifugation contributed tosetting free and enriching an abundance of intact PSI-induced inclusions in the initial pellets. Although someentities such as heavy mitochondria and cytoskeleton
Protein list and classification Accession No. References
Metabolism and mitochondrial functionChain B of heme oxygenase-1 (HO-1) gi|7767105|pdb|1DVGIB [42]Creatine kinase-B (CKB) gi|203476|gb|AAA40933.1 [9,12,29]Ubiquinol-cytochrome c reductase core protein I (UQCRCI) gi|51948476|reflNP-0010042501 [9,12,29]ATP synthase β subunit (F1-ATPase β) gi|1374715|gb|AAB02288.1 [9,12,29]
Protein folding and transport27 kDa heat shock protein 1 (HSP27) gi|62641274|ref|XP_215069.3 [5]70 kDa heat shock protein 1A/1B (HSP70) gi|415898|emb|CAA81642.1 [12,18]71 kDa heat shock cognate protein (HSC70) gi|56385|emb|CCA49670.1 [5,12,18]Protein kinase C inhibitor protein 1 (KCIP-1, or 14-3-3 ζ) gi|1051270|gb|AAA80544.1 [25]Valosin-containing protein (VCP) gi|38014694|gb|AAH60518.1 [46]
Table 2 Twenty protein components of Lewy bodies (LBs) identified from synthetic proteasome inhibitor (PSI)-induced inclusionsin PC12 cells
The consistent components of LBs identified from the PSI-induced inclusions have been characterized and cataloged as: antioxidant defense;cytoskeleton system; metabolism and mitochondrial function; neurotransmission; ubiquitin-proteasome system; and protein folding and transport[4,5,12,18]. When a component of LBs could be assigned to more than one class of function, it was allocated to the best known one [5].
networks coprecipitated with the PSI-induced inclusionsinto the initial pellets [38], it is not coincidence thatmitochondria recruited earlier in aggresomes whilecytoskeleton networks such as intermediate filaments andmicrotubules also participated in formation of aggresome-related inclusions [4,16,29]. After incubation of the initialpellets with DNase I mixed with RNase A contributed todegrading the remaining nuclei, centrifugation of the initialpellets in discontinuous gradient contributed to eliminatingsome other subcellular particles remaining in the resultingpellets [38].
In the proteomic context of PSI-induced inclusions, wefocused on the profile of eight chaperone proteins, ofwhich three were newly identified. The other five had beenreported to be consistent components of LBs. In addition,we also identified 15 proteins reported to be consistentcomponents of LBs. The main component of LBs, α-SYN,interestingly, was not identified in the present work. Thereare several explanations for this. The total level of synuclein-1, the rat homolog of human α-SYN, was at the samelevel of low expression in normal cells and was not alteredby proteasomal inhibition [15]; rather, excess levels of α-SYN play the dominant role in the development andformation of α-SYN-positive inclusions, such as LBs [4,17,23]. Just as in sporadic PD, α-SYN species (of highand various molecular weights) might have various post-translation modifications in PC12 cells under proteasomeinhibition [4]; low-abundance proteins are either not visibleon gel owing to limitations in sample loading or marked byhigh-abundance proteins [39]; and not all proteins can beidentified with the current state of MS technology [18].Thus, as was indicated in the results, supported by identi-fication through PMF-based proteomic analysis ofsubstantia nigra of PD patients [40], α-SYN was notidentified from the PSI-induced inclusions in PC12 cellsunder proteasome inhibition. However, 20 consistentcomponents of LBs were identified from the PSI-inducedinclusions. For example, HO-1 catalyzing rapid degradationof heme to biliverdin in brain [47], a putative marker ofoxidative stress response [48], is an important cytoplasmicconstituent of LBs [49]. HO-1 was intensely shown byimmunostaining in peripheries of LBs [47], and furthershown by immunoelectron microscopy to be in intimateassociation with filaments of LBs [48]. In brief, to someextent, the PSI-induced inclusions characterized by thepotential 20 consistent components of LBs could be withconstituent protein features of LBs [12,18]. Based on theidentification of the potential 20 consistent components ofLBs, the three newly identified chaperone proteins couldprovide clues of alternatives for further study.
As supported by growing evidence, the three newlyidentified chaperone proteins are expected to be noted assignificant. In cells faced with proteolytic stresses,aggresomes are equipped with a variety of chaperoneproteins recruited from cytoplasm, ER, and nucleus [17,50,51]. GRP58, a Ca2+-binding chaperone protein, GRP75,a member of the HSP70 family, and CaBP1, a probableresident protein of the ER, are induced in response toproteolytic stress when homeostasis of cells is disrupted[52−62]. The other five chaperone proteins reported to beconsistent components of LBs were observed to aberrantlyexpress in response to environments of proteolytic stressin neurodegenerative disease [5,25,45,46,63−65].
We did not carry out experiments of validation such asimmunostaining or immunoblotting in the present work[5,10]. Next, a major task for us is to validate the trueassociation of the three chaperone proteins with theirsubcellular localization.
In conclusion, based on attempts to characterize theproteomic features of PSI-induced inclusions formed inPC12 cells, we identified eight chaperone proteins, of whichCaBP1, GRP58, and GRP75 were newly identified. Thesefindings suggest that the three potential chaperone proteinsmight be recruited in the PSI-induced inclusions in PC12cells under proteasome inhibition. In addition, we presentedat the level of proteomics an approach to understandingthe relevance of the aberrant expression of chaperoneproteins to the PSI-induced inclusions in PC12 cells underproteasome inhibition.
Acknowledgements
We thank both Prof. Fengchen Ge and Prof. Yunbo Xue(Apiculture Science Institute of Jilin Province, Jilin, China)for their financial help.
References
1 Duyckaerts C, Hauw JJ, Am J. Lewy bodies, a misleading markerfor Parkinson’s disease? Bull Acad Natl Med 2003, 187: 277−292
2 Sone M, Yoshida M, Hashizume Y, Hishikawa N, Sobue G. α-Synuclein-immunoreactive structure formation is enhanced insympathetic ganglia of patients with multiple system atrophy.Acta Neuropathol (Berl) 2005, 110: 19−26
3 Kahle PJ, Neumann M, Ozmen L, Müller V, Odoy S, Okamoto N,Jacobsen H et al. Selective insolubility of α-synuclein in humanLewy body diseases is recapitulated in a transgenic mouse model.Am J Pathol 2001, 159: 2215−2225
4 McNa u ght KS, O la now CW. Protein a ggregat ion in thepathogenesis of familial and sporadic Parkinson’s disease. NeurobiolAging 2006, 27: 530−545
5 Leverenz JB, Umar I, Wang Q, Montine TJ, McMillan PJ, Tsuang
DW, Jin J et al. Proteomic identification of novel proteins incortical Lewy bodies. Brain Pathol 2007, 17: 139−145
6 Lee HJ, Lee SJ. Characterization of cytoplasmic α-synucleinaggregates. Fibril formation is tightly linked to the inclusion-forming process in cells. J Biol Chem 2002, 277: 48976−48983
7 Galvin JE, Giasson B, Hurtig HI, Lee VM, Trojanowski JQ.Neurodegeneration with brain iron accumulation, type 1 ischaracterized by α-, β-, and γ-synuclein neuropathology. Am JPathol 2000, 157: 361−368
8 Nishiyama K, Murayama S, Shimizu J, Ohya Y, Kwak S, AsayamaK, Kanazawa I. Cu/Zn superoxide dismutase-like immunoreactivityis present in Lewy bodies from Parkinson disease: a light andele ct ro n mic rosc opic immu n ocyt ochem ica l s tu d y. A ctaNeuropathol 1995, 89: 471−474
1 0 Jin J, Meredith GE, Chen L, Zhou Y, Xu J, Shie FS, Lockhart P etal. Quantitative proteomic analysis of mitochondrial proteins:relevance to Lewy body formation and Parkinson’s disease. BrainRes Mol Brain Res 2005, 134: 119−138
1 3 McNaught KS, Mytilineou C, Jnobaptiste R, Yabut J, ShashidharanP, Jennert P, Olanow CW. Impairment of the ubiquitin-proteasomesystem causes dopaminergic cell death and inclusion body formationin ventral mesencephalic cultures. J Neurochem 2002, 81: 301−306
1 4 Petrucelli L, O’Farrell C, Lockhart PJ, Baptista M, Kehoe K,Vink L, Choi P et al. Parkin protects against the toxicity associatedwith mutant α-synuclein: proteasome dysfunction selectively af-fects catecholaminergic neurons. Neuron 2002, 36: 1007−1019
1 5 Rideout HJ, Larsen KE, Sulzer D, Stefanis L. Proteasomal inhibitionleads to formation of ubiquitin/α-synuclein-immunoreactiveinclusions in PC12 cells. J Neurochem 2001, 78: 899−908
1 6 Olanow CW, Perl DP, DeMartino GN, McNaught KS. Lewy-bodyformation is an aggresome-related process: a hypothesis. LancetNeurol 2004, 3: 496−503
1 7 Kopito RR. Aggresomes, inclusion bodies and protein aggregation.Trends Cell Biol 2000, 10: 524−530
1 8 Zhou Y, Gu G, Goodlett DR, Zhang T, Pan C, Montine TJ, MontineKS et al. Analysis of α-synuclein-associated proteins by quantitativeproteomics J Biol Chem 2004, 279: 39155−39164
1 9 Tsai YC, Fishman PS, Thakor NV, Oyler GA. Parkin facilitatesthe elimination of expanded polyglutamine proteins and leads topreservation of proteasome function. J Biol Chem 2003, 278:22044−22055
2 0 Chillarón J, Haas IG. Dissociation from BiP and retrotranslocationof unassembled immunoglobulin light chains are tightly coupledto proteasome activity. Mol Biol Cell 2000, 11: 217−226
2 1 Werner ED, Brodsky JL, McCracken AA. Proteasome-dependentendoplasmic reticulum-associated protein degradation: an uncon-ventional route to a familiar fate. Proc Natl Acad Sci USA 1996,93: 13797−13801
2 2 Nawrocki ST, Carew JS, Dunner K, Boise LH, Chiao PJ, Huang P,Abbruzzese JL et al. Bortezomib inhibits PKR-like endoplasmicreticulum (ER) kinase and induces apoptosis via ER stress in humanpancreatic cancer cells. Cancer Res 2005, 65: 11510−11519
2 3 Garcia-Mata R, Gao YS, Sztul E. Hassles with taking out the garbage:
aggravating aggresomes. Traffic 2002, 3: 388−3962 4 Ito H, Kamei K, Iwamoto I, Inaguma Y, Nohara D, Kato K.
Phosphorylation-induced change of the oligomerization state ofα B-crystallin. J Biol Chem 2001, 276: 5346−5352
2 5 Kaneko K, Hachiya NS. The alternative role of 14-3-3 zeta as asweeper of misfolded proteins in disea se conditions. MedHypotheses 2006, 67: 169−171
2 6 Lee DH, Goldberg AL. Proteasome inhibitors: valuable new toolsfor cell biologists. Trends Cell Biol 1998, 8: 397−403
2 7 Wigley WC, Fabunmi RP, Lee MG, Marino CR, Muallem S,DeMartino GN, Thomas PJ. Dynamic association of proteasomalmachinery with the centrosome. J Cell Biol 1999, 145: 481−490
2 8 Fabunmi RP, Wigley WC, Thomas PJ, DeMartino GN. Activityand regulation of the centrosome-associated proteasome. J BiolChem 2000, 275: 409−413
2 9 Gai WP, Yuan HX, Li XQ, Power JT, Blumbergs PC, Jensen PH.In situ and in vitro study of colocalization and segregation of α-synuclein, ubiquitin, and lipids in Lewy bodies. Exp Neurol 2000,166: 324−333
3 0 Atadzhanov M, Zumla A, Mwaba P. Study of familial Parkinson’sdisease in Russia, Uzbekistan and Zambia. Postgrad Med J 2005,81: 117−121
3 1 Pinelli M, Giacchetti M, Acquaviva F, Cocozza S, DonnarummaG, Lapice E, Riccardi G et al. β2-Adrenergic receptor and UCP3variants modulate the relationship between age and type 2 diabetesmellitus. BMC Med Genet 2006, 7: 85
3 2 Jacobs DI, van Rijssen MS, van der Heijden R, Verpoorte R.Sequential solubilization of proteins precipitated with trichloro-acetic acid in acetone from cultured Catharanthus roseus cellsyields 52% more spots after two-dimensional electrophoresis.Proteomics 2001, 1: 1345−1350
3 3 Basso M, Giraudo S, Lopiano L, Bergamasco B, Bosticco E,Cinquepalmi A, Fasano M. Proteome analysis of mesencephalictissues: evidence for Parkinson’s disease. Neurol Sci 2003, 24:155−156
3 4 Yin Z, Stead D, Selway L, Walker J, Riba-Garcia I, McLnerney T,Gaskell S et al. Proteomic response to amino acid starvation inCandida albicans and Saccharomyces cerevisiae. Proteomics2004, 4: 2425−2436
3 6 Katsuse O, Iseki E, Marui W, Kosaka K. Developmental stages ofcortical Lewy bodies and their relation to axonal transport blockagein brains of patients with dementia with Lewy bodies. J Neurol Sci2003, 211: 29−35
3 7 Sian J, Hensiek R, Senitz D, Muench G, Jellinger K, Riederer P,Gerlach M. A novel technique for the isolation of Lewy bodies inbrain. Acta Neuropathol (Berl) 1998, 96: 111−115
3 8 Pasquali C, Fia lka I , Huber LA. Subcellu lar fractionation,electromigration analysis and mapping of organelles. J ChromatogrB Biomed Sci Appl 1999, 722: 89−102
3 9 Taylor SW, Fahy E, Ghosh SS. Global organellar proteomics. TrendsBiotechnol 2003, 1: 82−88
4 0 Basso M, Giraudo S, Corpillo D, Bergamasco B, Lopiano L, FasanoM. Proteome analysis of human substantia nigra in Parkinson’sdisease. Proteomics 2004, 4: 3943−3952
4 1 Alim MA, Hossain MS, Arima K, Takeda K, Izumiyama Y,Nakamura M, Kaji H et al . Tubulin seeds α-synuclein fibril
expression in idiopathic Parkinson’s disease. Exp Neurol 1998,150: 60−68
4 3 Ostrerova N, Petrucelli L, Farrer M, Mehta N, Choi P, Hardy J,Wolozin B. α-Synuclein shares physical and functional homologywith 14-3-3 proteins. J Neurosci 1999, 19: 5782−5791
4 4 Ii K, Ito H, Tanaka K, Hirano A. Immunocytochemical co-localization of the proteasome in ubiquitinated structures inneurodegenerative diseases and the elderly. J Neuropathol ExpNeurol 1997, 56: 125−131
4 6 Hirabayashi M, Inoue K, Tanaka K, Nakadate K, Ohsawa Y, KameiY, Popiel AH et al. VCP/p97 in abnormal protein aggregates,cytoplasmic vacuoles, and cell death, phenotypes relevant toneurodegeneration. Cell Death Differ 2001, 8: 977−984
4 7 Schipper HM, Liberman A, Stopa EG. Neural heme oxygenase-1expression in idiopathic Parkinson’s disease. Exp Neurol 1998,150: 60−68
4 8 Castellani R, Smith MA, Richey PL, Perry G. Glycoxidation andoxidative stress in Parkinson disease and diffuse Lewy body disease.Brain Res 1996, 737: 195−200
4 9 Yoo MS, Chun HS, Son JJ, DeGiorgio LA, Kim DJ, Peng C, SonJH. Oxidative stress regulated genes in nigral dopaminergicneu rona l cell s: cor re la tion wi th the k nown pathology inParkinson’s disease. Brain Res Mol Brain Res 2003, 110: 76−84
5 0 Sherman MY, Goldberg AL. Cellular defenses against unfoldedproteins: a cell biologist thinks about neurodegenerative diseases.Neuron 2001, 29: 15−32
5 1 Luo GR, Chen S, Le WD. Are heat shock proteins therapeutictarget for Parkinson’s disease? Int J Biol Sci 2006, 3: 20−26
5 2 Liu H, Bowes RC 3rd, van de Water B, Sillence C, Nagelkerke JF,Stevens JL. Endoplasmic reticulum chaperones GRP78 andcalreticulin prevent oxidative stress, Ca2+ disturbances, and celldeath in renal epithelial cells. J Biol Chem 1997, 272: 21751−21759
5 3 Rao RV, Castro-Obregon S, Frankowski H, Schuler M, Stoka V, delRio G, Bredesen DE et al. Coupling endoplasmic reticulum stressto the cell death program. An Apaf-1-independent intrinsicpathway. J Biol Chem 2002, 277: 21836−21842
5 4 Chichiarelli S, Ferraro A, Altieri F, Eufemi M, Coppari S, Grillo C,Arcangeli V et al. The stress protein ERp57/GRP58 binds specificDNA sequences in HeLa cells. J Cell Physiol 2007, 210: 343−351
5 5 Hetz C, Russelakis-Carneiro M, Walchli S, Carboni S, Vial-KnechtE, Maundrell K, Castilla J et al. The disulfide isomerase Grp58 isa protective factor against prion neurotoxicity. J Neurosci 2005,25: 2793−2802
5 6 Wadhwa R, Taira K, Kaul SC. An Hsp70 family chaperone,mortalin/mthsp70/PBP74/Grp75: what, when, and where? CellStress Chaperones 2002, 7: 309−316
5 7 Jinghua J, Hulette C, Wang Y, Zhang T, Pan C, Wadhwa R, ZhangJ. Proteomic identification of a stress protein, mortalin/mthsp70/GRP75: relevance to Parkinson disease. Mol Cell Proteomics2006, 5: 1193−1204
5 8 Yoo BC, Kim SH, Cairns N, Fountoulakis M, Lubec G. Derangedexpression of molecular chaperones in brains of patients withAlzheimer’s disease. Biochem Biophys Res Commun 2001, 280:249−258
5 9 Fullekrug J, Sonnichsen B, Wunsch U, Arseven K, Nguyen Van P,Soling HD, Mieskes G. CaBP1, a calcium binding protein of thethioredoxin family, is a resident KDEL protein of the ER and notof the intermediate compartment. J Cell Sci 1994, 107: 2719−2727
6 0 Meunier L, Usherwood YK, Chung KT, Hendershot LM. A subsetof chaperones and folding enzymes form multiprotein complexesin endoplasmic reticulum to bind nascent proteins. Mol Biol Cell2002, 13: 4456−4469
6 1 Schweizer A, Peter F, Van PN, Soling HD, Hauri HP. A luminalcalcium-binding protein with a KDEL endoplasmic reticulumretention motif in the ER-Golgi intermediate compartment. EurJ Cell Biol 1993, 60: 366−370
6 2 Megidish T, Takio K, Titani K, Iwabuchi K, Hamaguchi A, IgarashiY, Hakomori S. Endogenous substrates of sphingosine-dependentkinases (SDKs) are chaperone proteins: heat shock proteins,glucose-regulated proteins, protein disulfide isomerase, andcalreticulin. Biochemistry 1999, 38: 3369−3378
6 3 Richard M, Biacabe AG, Streichenberger N, Ironside JW, Mohr M,Kopp N, Perret-Liaudet A. Immunohistochemical localization of14.3.3 zeta protein in amyloid plaques in human spongiformencephalopathies. Acta Neuropathol (Berl) 2003, 105: 296−302
6 4 Ito H, Kamei K, Iwamoto I, Inaguma Y, García-Mata R, Sztul E,Ka to K. Inhibi tion of proteasomes induces accumulation,phosphorylation, and recruitment of HSP27 and αB-crystallin toaggresomes. J Biochem 2002, 131: 593−603
6 5 Outeiro TF, Klucken J, Strathearn KE, Liu F, Nguyen P, RochetJC, Hyman BT et al. Small heat shock proteins protect against α-synuclein-induced toxicity and aggregation. Biochem Biophys ResCommun 2006, 351: 631−638
