International Journal of Cardiology 153 (2011) 296–305

Contents lists available at ScienceDirect

International Journal of Cardiology

j ourna l homepage: www.e lsev ie r.com/ locate / i j ca rd

Proteomic comparison of spherical aggregates and adherent cells of cardiacstem cells

Mitsuyo Machida, Yohtaroh Takagaki, Rumiko Matsuoka, Nanako Kawaguchi ⁎,1

International Research and Educational Institute for Integrated Medical Sciences (IREIIMS), Tokyo Women's Medical University, Japan

⁎ Corresponding author. 8-1 Kawada-cho, ShinjukuTel.: +81 3 3353 8111x24007; fax:+81 3 3352 3088.

E-mail address: nanako.res@gmail.com (N. Kawaguc1 Present address; Department of Molecular Therapy, N

National Center of Neurology and Psychiatry, 4-1-1 O187-8502, Japan. Tel.: +81 42 346 1720; fax: +81 42

0167-5273/$ – see front matter © 2010 Elsevier Irelanddoi:10.1016/j.ijcard.2010.08.049

a b s t r a c t

a r t i c l e i n f o

Article history:

Received 4 March 2010Received in revised form 10 May 2010Accepted 17 August 2010Available online 23 September 2010

Keywords:Cardiac stem cellCardiosphereProteomeProtein profileCalreticulin

Background: Previously published papers showed that cardiac stem cells (CSCs) form (cardio)sphere.However, recent studies questioned the significance of the sphere-formation as one of the characteristics ofCSCs. We isolated c-kit-positive cardiac stem cells, cultured as bulk (CSC-BC) and characterized thempreviously. Among them, CSC-BC21 formed an extraordinary number of spheres. Using a clone derived fromthis bulk culture, we investigated the effect of sphere-formation on differentiation and performedproteomics analysis comparing two statuses, cardiosphere and dish substrate attachment.Methods: We performed sphere-forming assay to compare the sphere-forming ability among CSC-BCs. Thecloned cells from CSC-BC21, which had distinct sphere-forming ability, were cultured in a differentiationmedium (DM) to induce cardiac myocyte differentiation. We performed RT-PCR analysis to investigate ifcardiosphere-formation affects cardiac myocyte gene expression level. Furthermore, proteome analysis wasperformed to compare floating cardiosphere (flCS) and dish-attached cardiosphere-derived cells (daCS).

Results: One of the cloned cells, CSC-21E expressed higher troponin I message than CSC-BC21. Moreover, themessage level of troponin I was enhanced when they had experienced cardiosphere prior to the treatment ofmyocyte differentiation medium. The change from flCS to daCS accompanied up-regulation of chaperonesand down regulation of glycolytic and other metabolic enzymes. Calreticulin and Hsp 90 were among the up-regulated chaperons. Calreticulin is known to be an essential component of cardiogenesis.Conclusion: These results suggest that the switch from aggregated sphere to the cell attachment, is importantfor advancing the cardiac cell differentiation.

© 2010 Elsevier Ireland Ltd. All rights reserved.

1. Introduction

The adult mammalian heart has been traditionally regarded as aterminally differentiated organ. Recent evidence, however, indicatesthat it has resident stem cells with self-renewing capacity, whichcould replenish cells damaged by heart failure or myocardialinfarction [1]. The cell surface antigens, c-kit, Sca-1 and MDR1,were reported to be expressed on these stem cells [2]. From adulthearts, cells expressing c-kit, Sca-1 and MDR1 antigens were isolatedin rats [1], mice [3], dogs [4] and humans [5]. These cells wereclonogenic, multipotent, and capable of generating cardiac myocytes,smooth muscle and endothelial cells [1]. These results indicate thatcardiac stem cells reside in the heart, where they maintain their stemcell properties and respond to stimuli to generate cells for repairinginjuries [1].

-ku, Tokyo 162-8666, Japan.

hi).ational Center of Neuroscience,gawa-Higashi, Kodaira, Tokyo346 1750.

Ltd. All rights reserved.

Recently, we have also isolated cardiac stem cells by selecting c-kit-positive cells from adult rat hearts [6]. These cells appear to havecapacity to differentiate into several cell lineages including cardiacand skeletal myocytes, adipocytes and endothelial cells. For regener-ative medicine, an understanding of the regulatory molecularmechanism for cell differentiation and self-replenish is crucial.Towards achieving this goal, gene expression analyses with tran-scription profiling (transcriptome) and micro-arrays have beenconducted including neurospheres [7], mouse and human embryonicstem cells [8,9,10] and left atrium derived stem cells [11]. ThesemRNA analyses, however, have limitations because several investiga-tions revealed that protein expression levels correlated poorly withmRNA level due to the multiple regulatory factors including mRNAstability, rate of translation and protein degradation [12, 13]. Proteinprofiling (proteome), on the other hand, could cover a wide range ofexpressed proteins, and has the potential to provide information onpost-transcriptional modification. Although several protein profiles ofembryonic stem (ES) cells have been reported, the profile ofcardiosphere (CS) is yet to be published.

In the present paper, we report the results of a proteomic analysisof a clone, CSC-21E, which was derived from one of the bulk cultures,CSC-BC21, showing highest amount of spherical aggregate formation

297M. Machida et al. / International Journal of Cardiology 153 (2011) 296–305

among the CSCs we obtained. In general, spherical aggregates are themorphology seen in the early stages of various cell lineages, such as EScells, neurospheres [7], and embryoid bodies in teratocarcinomacultures [14]. CSC-21E cells could form high number of floatingspherical aggregates (flCS) when cultured on bacterial dishes. Wecompared the flCS and daCS protein profiles using two-dimensionalfluorescent gel electrophoresis (2-D DIGE). A majority of the proteinsshowed the same level of expression in both forms, including theproteins characteristically observed in human and murine ES cellspreviously. A fraction of proteins, however, was found to be down orup-regulated when the flCS cells were switched to daCS cells. Theseprotein changes suggest that this switch may be a critical stage forcardiac stem cells to enhance differentiation into cardiomyocyte.

2. Materials and methods

2.1. Isolation and cloning of cardiac stem cells

Male Lewis rats were obtained from Charles River Japan (Tokyo) and sacrificed at8 weeks old. All experiments were conducted according to the animal handling ethicsof Tokyo Women's Medical University, and were reviewed by the animal experimentcommittee of the university. Cardiac stem cells (CSCs) from the adult rat hearts wereisolated as c-kit-positive cells, and maintained according to a method reportedpreviously [6]. Cloning was performed by dilution method as described previously [6].Briefly, CSC-BC21 cells were isolated using c-kit antibody and cultured and passagedmore than 40 times. CSC-21E cells were cloned from CSC-BC21 by a dilution method.

2.2. Cardiosphere-forming assay

Prior to cardiosphere-forming assays, cells were cultured at a density of 104 cells/cm2 on gelatin-coated dishes (IWAKI 4020-020, IE5 gelatin-coated) for 3 days incomplete medium (CompM), and then transferred to bacterial dishes (Iwaki, SH90-15)in cardiosphere-forming medium (CSM) to form cardiospheres. After 3 days of culture,the number of spheroids was counted within a microscopic field. The constitution ofComp M and CSM was described previously [6].

2.3. Cardiomyocyte differentiation induction and RT-PCR

To investigate the influence of cardiosphere to cardiomyocyte differentiation, wecompared the differentiation ability of the cells, experiencing cardiosphere with thecells, not experiencing cardiosphere. To induce differentiation into cardiac myocytesfrom cardiospheres (which have experience of cardiospheres), the CSC-21E cells werecultured on bacterial dishes for 3 days in CSM, transferred to laminin-coated 35 mm2

glass bottom dishes (Matsunami) and cultured in the cardiomyocyte differentiationmedium for 3 days [6]. On the other hand, the cells (which have no experience ofcardiosphere prior to differentiation assay) were plated on a laminin-coated 35 mm2

without formation of cardiospheres for 3 days. The compositions of differentiationmedium were described previously [6]. Cells were cultured for 1 week, and RNA wasextracted. 100 ng of total RNA was used for RT-PCR (30 cycles) [6]. RT-PCR wasperformed as described previously [6].

Fig. 1. Cardiosphere-forming assay of CSC-BCs (A) and morphology of two differentforms of the cloned CSC-21E cells (B, C). Cardiosphere-forming assay was performed asdescribed in the Materials and methods. Floating aggregates (flCS) on bacterial dishes(Fig. 1A), and attached cells expanded from cardiospheres (daCS) on cell culture dishes(Fig. 1B), were cultured in the cardiosphere-forming medium (CSM) as described inMaterials and methods.

2.4. Preparation of samples for protein profiling and two-dimensional fluorescencedifference gel electrophoresis (2-D DIGE)

Prior to cardiosphere-forming, CSC-21Ecellswereculturedatadensityof104 cells/cm2

on gelatin-coated dishes (IWAKI 4020-020, IE5 gelatin-coated) for 3 days in completemedium (CompM) then transferred to bacterial dishes in CSM to form cardiospheres. After3 days of culture, the spheroids were collected and used as the samples of floatingaggregates (flCS). The spheroids were cultured on the gelatin-coated dishes in CSM for3 days to obtain attached state of cells in the same culture medium (CSM). The cells wereharvested and used as the samples of attached cells (daCS).

Whole cells were lysed in the buffer containing 7 M urea (Plus one urea), 2 Mthiourea (Plus one thiourea), 60 mM DTT (Plus one DTT,), 2% IPG buffer (pH 3–10 forisoelectric focusing, GE Healthcare Bio-Sciences, Japan), 4% CHAPS (Sigma, St. Louis,MO) and a cocktail of protease inhibitors (Roche Diagnostics Japan). The lysates weresonicated until clear, left on ice for 20 min and centrifuged at 15,000×g for 30 min at4 °C. Protein extracts were precipitated with 10% trichloroacetic acid and washed withcold acetone. Protein concentrations were determined using Bio-Rad protein assayreagent [15]. Two-dimensional fluorescence difference gel electrophoresis (2-D DIGE)was carried out using an Ettan DIGE system (GE Healthcare Bio-Sciences Japan)according to the manufacturer's protocol as described in the previous paper [16].Briefly, The CyDye labeled mixture was immediately subjected to isoelectric focusingon an IPGphor 3 system using a 24 cm-long IPG strip (pH 3–10 nonlinear (NL)gradient). The IPG strips were reduced and alkylated [16]. and were immediatelyapplied to 10% SDS-PAGE gels (20×25.5 cm, Nihon Eido, Tokyo. The protein profileswere scanned directly using FLA5100 (Fuji film, Tokyo). For Cy3 and Cy5, 532 nm and

633 nm were used for excitation and 580 nm and 670 nm for emission, respectively.Spot detection and matching were automatically done using PDQuest 8.0 software(Bio-Rad). The relative abundance of each resolved protein spot was quantified by themathematical fitting of Gaussian curves in two dimensions. For the comparison of threereplicate gel sets, significance was calculated with Student' t-test (Pb0.05) anddifferentially expressed proteins with a spot ratio exceeding 1.5-fold were selected.

2.5. Identification of proteins by MALDI-TOF-MS

For protein identification, the CS cell lysate having 600 μg of protein was mixedwith 25 μg each of flCS and daCS cell extracts labeled with Cy5 and Cy3 respectively,and applied to the Ettan DIGE system. The proteins were stained with SYPRO-Ruby(Invitrogen, CA), spots of interest were excised based on 2-D DIGE images or aquantification analysis by EXQuest (Bio-Rad), and digested with trypsin (Trypsin Gold,

Fig. 2. Cardiac troponin I expression of CSC-21E cells upon cardiac myocytedifferentiation induction. The cells (N) were maintained in Comp M, transferred tolaminin-coated dish (without forming cardiosphere) and cultured in differentiationmedium to induce cardiac myocyte differentiation. On the other hand, the cells (S) wereformed cardiosphere on bacteria dish for 3 days in CSM, and spheroids were collected.Then the spheroids were transferred to laminin-coated dish, and cultured for 3 days indifferentiation medium. RT-PCR was performed as described in the Materials andmethods. CSC-21E and CSC-BC21 were used for the experiments.

298 M. Machida et al. / International Journal of Cardiology 153 (2011) 296–305

Mass Spectrometry Grade, Promega, WI) [17,18], also described in http://proteome.tmig.or.jp/2D/2DE_method.html]. The sample was analyzed by MALDI-TOF massspectrometry (AutoflexII, Bruker Daltonics) in the positive ion reflector mode and inthe m/z range of 600–4000 Da. Mass spectra were calibrated with peptide calibrationstandard mono (Bruker Daltonics). The data were investigated by their peptide massfingerprint (PMF) using the non-redundant NCB Inr database with a MASCOT searchengine (Matrix Science, London, UK) through the BioTools 3.0 interface provided byBruker Daltonics. We set up the taxonomy of Rattus, with carbamido-methylation ofcysteines as a fixed modification, oxidation of methionine residues as a variablemodification, peptide mass tolerance of 0.5 Da and one missing cleavage site.

Fig. 3. Comparative 2-D DIGE of proteins from flCS and daCS cells of the CSC-21E clone. Whocombined and analyzed by 2-D PAGE with a pH 3–10 NL IPG strip in the first dimension anamounts in both samples appear as yellow spots, whereas those more abundant in flCS are

2.6. Western blotting

Proteins were resolved in one-dimensional (1-D) SDS-PAGE (10% acrylamide), andwere transferred onto a polyvinylidene-difluoride (PVDF) membrane (Millipore, MA)vertical to 1-D SDS-PAGE by electrophoresis. Membranes were reacted with primaryantibody, annexinA6 (ANXA6; sc-11388), annexin A7 (ANXA7, sc-17815, both fromSanta Cruz Biotechnology Inc), grp75 (RB-10470-P0, Lab Vision Corporation, Cheshire,UK) and serine or cysteine proteinase inhibitor clade H member 1 (serpinh1; sc-8352,Santa Cruz Biotechnology Inc). Immuno-reactive bands were detected by enhancedchemi-luminescence method using a kit from GE healthcare. Band intensities werequantified using Image-J in NIH-Image [http://rsb.info.nih.gov/nih-image/about.html].The intensity was normalized to that of glyceraldehydes 3-phosphate (GAPDH;MAB374, Millipore, MA) as an internal standard. Data are indicated as the mean±standard deviation (s.d.) for three independent experiments. A value of pb0.05 wasshown to be statistically significant by Student's t-test.

3. Results

3.1. Characterization of cardiosphere-forming cells CSC-BC21 and itsclone CSC-21E

Cardiac stem cells (CSCs) from the adult rat hearts were isolated asc-kit-positive bulk cultures, and passaged over 40 times [6]. Amongthe bulk cultures, CSC-BC21 gave highest number of sphericalaggregates. In the cardiosphere-forming assay, CSC-BC21 formed themost spheroids (62±15) while CSC-BC1 and CSC-BC4 reportedpreviously [6] had only 16.7±1.6 and 0.2±0.4 spheroids respectively(mean±SD), as shown in Fig. 1A. CSC-BC21 cells resembled to CSC-BC2, as they expressed mRNA for GATA4 (low expression), desmin(low expression), connexin 43 (cardiomyocyte marker gene) andPPARα (adipocyte marker gene), but did not express Oct4, MLC2v,cardiac-actin, smooth muscle actin, myogenin or Pecam (data notshown). CSC-BC21 was subjected to cloning, and a clone CSC-21E wasestablished, which also showed floating cardiospheres (flCS) whencultured on bacterial dishes (Fig. 1B). A distinct feature of this clone,CSC-21E, and CSC-BC21, is that the floating cell aggregates in vitrobecame attached to dish surfaces when switched from bacterial dishes

le cell lysates (50 μg each) from flCS labeled with Cy5 and daCS labeled with Cy3 wered by 10% acrylamide SDS-PAGE in the second dimension. The proteins having similarred spots and those more abundant in daCS are green spots.

299M. Machida et al. / International Journal of Cardiology 153 (2011) 296–305

to cell culture dishes, and furthermore, the attached cells revertedback to floating aggregates when transferred back to bacterial dishes(data not shown). Moreover, CSC-21E cells expressed troponin I gene(Fig. 2 left lanes) but the expression was more enhanced if cardio-sphere was once formed prior to differentiation induction (Fig. 2 laneS). Without forming cardiosphere before induction, CSC-21E cellsexpress troponin I in a low level (Fig. 2, lane N). No expression oftroponin I was observed in CSC-BC21 neither sphere-formed cells nornon-sphere-formed, adherent cells (Fig. 2 right lanes). These resultsindicate that cardiosphere-formation prior to differentiation induc-tion enhances the troponin I gene expression. Therefore, we nextperformed proteome analysis to find what proteins were involved informing cardiosphere and/or attachment to the dishes.

3.2. Cardiospheres and their protein profiling by 2-D SDS-PAGE

Cell lysates obtained from CSC-21E flCS and daCS were eachsubjected to 2-dimentional (2-D) PAGE separately. Fifty microgramsof whole cell lysate from each sample were separated on a pH 3–10 NLIPG strip in the first dimension and by 10% acrylamide SDS-PAGE inthe second dimension. The gels were stained with Sypro-Ruby, and

Fig. 4. Quantitation of selected proteins by comparing 2-D gel scans andWestern blots. Left texpression obtained in 2-D DIGE in Fig. 3. Second column from right shows Western bloglyceraldehyde-3-phosphate dehydrogenase (GAPDH). Relative band intensities quantified bcolumn with bars indicating the mean±standard deviation (s.d.).

the images were analyzed using PDQuest software (data not shown).The number of spots detected by the software in several repeated gelsranged from 676 to 890 with an average of 769.

Since the protein profiles for separate flCS and daCS gels weresimilar (data not shown), we applied the fluorescent 2-D DIGE systemto sensitize the detection of differently expressed proteins. Theprotein lysates from flCS and daCS cells were labeled separately withCy5 and Cy3 respectively, and these samples were combined and runin a single 2-D gel. In the fluorescent profile shown in Fig. 3, proteinsof similar abundance in both samples appeared as yellow spots,whereas those more abundant in flCS or daCS showed red or greenspots respectively. Spots with flCS/daCS ratio or daCS/flCS ratio ofgreater than 1.5 fold are judged as exhibiting differential expression.Consequently, 62 green spots gave increased or up-regulated proteinsin daCS, and 41 red spots gave decreased or down-regulated proteinsin daCS. Since overall number of spots averaged 769, roughly 13% ofspots gave differential expression. Three independent 2-D DIGE runsgave consistent results.

To check the validity of the differential expression in Fig. 3 weselected four spots, 8, 9, 36 and 51 as examples, and compared theirintensity as shown in Fig. 4. These spots were chosen because

wo columns are the intensity obtained with SYPRO-Ruby staining. Middle column is thets of 1-D PAGE with upper lanes for respective samples and lower lanes for controly Image-J of samples were normalized with those of GAPDH, and are shown in the right

Fig. 5. Classification of proteins expressed in CSC-21E according to biological function.Proteins identified by MALDI-TOF MS and by the MSCOT database search engine(Table 1) were categorized by their biological function accordingly to the GeneOntology database ≤http://www.geneontology.org/≥ and are expressed in a pie chart.

300 M. Machida et al. / International Journal of Cardiology 153 (2011) 296–305

commercial antibodies are available for quantification on Western-blot filters. Fig. 4 shows the protein intensity obtained in separatelyrun 2-D PAGE with Sypro-Ruby staining in the left two columns, andthe expression obtained in 2-D DIGE (Fig. 3) in the middle column.The quantitative comparison of flCS and daCS lysates using 1-DWestern blots is shown in the neighboring right side of the middlecolumn with Glyceraldehyde-3-phosphate dehydrogenase (GAPDH)as a control protein for normalization. Relative band intensity isshown in the right extreme panels with bars indicating the mean±standard deviation (s.d.). These results clearly indicate the differentialexpression obtained in 2-D DIGE to be authentic.

3.3. Protein identification in a CS cell profile

For the identification of each protein spot, a total of 600 μg of flCScell lysate was applied to the DIGE system after mixing with 25 μgeach of Cy5-labeled flCS and Cy3-labeled daCS. Tables 2A and 2B list78 spots successfully identified after in-gel digestion, extraction andidentification by MALDI-TOF MS with the corresponding spotnumbers shown in Fig. 3. These proteins were categorized accordinglyto the Gene Ontology database, and Fig. 5 shows the distribution forfunctional categories. Twenty eight percent of the proteins identifiedwere related to metabolism, followed by binding proteins (22%),chaperones (17%), cytoskeletal proteins (10%), redox proteins (9%),translation proteins (5%), channel proteins (3%) and proteins involvedin signal transduction (3%). Binding proteins (22%) included calcium-binding, RNA binding and nucleotide-binding proteins. Despite thelimited number of proteins identified, the distribution of proteinsalong functional categories appears similar to a more extensiveanalysis suggesting that the data reflect overall properties of theprotein profile. The proteins with extremely strong yellow spots inFig. 3 are as follows:– i) cytoskeletal and related proteins such asactins in spots 48 and 49, keratin in spot 39 and cofilin 1 in spot 77;ii) glycolytic enzymes in spots 18, 41, 42, 44, 45, 57 and 84;iii) proteins involved in protein folding such as heat shock proteinsin spots 1 and 7, chaperonin in spot 12 and glutathione S-transferasepi1 in spot 78; and iv) anexins A1 and A2 in spots 55 and 59respectively.

3.4. Analysis of proteins differentially expressed in flCS and in daCS

Among the 78 spots successfully identified in Table 1, 21 areexpressed of higher levels in flCS (Table 2A) while 13 spots are up-regulated in daCS (Table 2B). These results indicate that a change inmorphology accompany the switching of many proteins. Metabolicenzymes comprise about half of the protein whose expressiondecreased upon CS attachment, and they include: 1) the aldehydedehydrogenase family 3 member A1 (ALDH; spot 31) reported to beexpressed in self-renewable stem cells such as neurospheres andneuroepithelial stem-like cells [19]; 2) guanine deaminase (Spot 50), ahomologue of cypin (cytosolic PSD-95 interactor), reported to regulatedendrite numbers and to promote microtubule assembly [20] and 3)transaldolase (spot 54) and a transaldolase isoform (spot 56) involvedin the pentose phosphate pathway. The heat shock proteins (HSPs)and chaperones comprise the second category of proteins whoseexpression was decreased in daCS. These proteins are involved inprotein folding and intracellular trafficking, and are generally termedstress proteins [21]. The proteins with decreased levels in daCS includechaperonin containing TCP1 (spot 30), grp75 (or mortalin, spot 8), amember of the HSP70 family, and peroxiredoxin 5 (spot 76).Overexpression of grp75 was reported to extend the lifespan ofnormal fibroblast, while a reduction of mortalin levels in immortalizedcells causes growth arrest [22]. Peroxiredoxin 5 (spot 76), known toprotect cells from oxidative stress, was classified as a redox protein butmay be combined with chaperons as a protector against proteindamage [23,24].

Some chaperones were increased in daCS such as calreticulin,Hsp90, serine/cystein proteinase inhibitor clade H member 1(serpinh1; spot 36), and protein disulfide isomerase (PDI; spot 27).Calreticulin (spot 14) is a Ca2+ binding chaperone of the endoplasmicreticulum, but also known to be strongly expressed transiently duringheart development and is required for generation of proper myofibrin[25]. Serpins function as inhibitors of serine or cystein-proteases,caspases, hormone transporters, chaperones and tumor suppressors,and also serve as chaperones stabilizing the collagen triple helix [26].PDI is known to mediate integrin-dependent cell adhesion [27] byinducing conformational changes [28]. Goplen and colleagues foundthat the incubation of glioblastoma cells with PDI antibody inhibitedthe attachment to plastic and migration of U373 cells [29] suggestingits potential role for the attachment of CS cells. In summary, the majorcategories of proteins whose expression decreased upon the attach-ment of flCS were metabolic enzymes including those involved inglycolysis, followed by some chaperones known to down regulateupon cell differentiation. On the other hand, different chaperonesincreased in amounts in daCS.

4. Discussion

The bulk culture cell lines were obtained from c-kit-positivecardiac stem cell (CSC), and among them, CSC-BC21 generated highestnumber of sphereswhen cultured in the cardiospheremedium (CSM).The sphere has been used for stem cell isolation in several groups[30,31], however, the significance of the sphere was questioned byother group recently [32]. CSC-21E cells were obtained from adult rathearts, cultured more than 40 times and cloned. The cells can takeeither form, cardiosphere (flCS) or dish-attached (daCS) according tothe cell culture substrate. Therefore, there is no possibility ofcontaminating cardiac myocytes. Our results show that the spherecould regulate troponin I expression level and also revealed theprotein profile difference between cardiospheres and dish substrateattached cells.

CSC-BC21 cells expressed mRNA for GATA4, desmin and con-nexin43, consistent with their cardiomyocyte lineage, but simulta-neously expressed adipocyte related genes such as the PPARα andCEBPα. In order to investigate whether these expressions could be a

Table 1Proteins identified in 2D-DIGE of flCS and daCS.

Spotno.

Protein name Accessionno. (gi)

Theoretical Matchedpeaksa

Scoreb Coverage(%)c

Probabilityd fCS/daCSratioe

Biological function

M.W.(kDa)

pI

9 Annexin A6 1351943 76106 5.39 37 108 43 1.10E-06 0.26 Protein binding25 ATP synthase, mitochondrial F1

complex, α subunit149029483 54,632 8.24 21 68 43 1.10E-02 0.38 Redox

35 Eukaryotic translation elongationfactor 1 alpha 1

28460696 50,424 9.1 25 105 50 2.20E-06 0.42 Protein binding

14 Calreticulin 11693172 48,137 4.33 24 125 58 2.20E-08 0.43 Calcium ion binding27 Protein disulfide isomerase

associated 38393322 57,010 5.88 37 187 61 1.40E-14 0.48 Protein binding

12 Chaperonin 60 1778213 61,029 5.78 28 100 50 7.00E-06 0.49 Protein binding15 Iodothyronine 5′ monodeiodinase 202549 54,375 4.87 29 153 60 3.50E-11 0.49 Oxidoreductases24 Keratin complex 1, acidic, gene 14 56912233 52,936 5.08 43 204 67 2.80E-16 0.5 Cytoskeleton84 Lactate dehydrogenase A 8393706 36,712 8.45 32 192 73 4.40E-15 0.5 Metabolism (Glycolysis/

gluconeogenesis)1 rCG49111 (HSP90) 149067319 92,998 4.72 36 80 35 7.70E-04 0.51 Chaperone43 Phosphoglycerate kinase 1 40254752 44,909 8.02 38 209 72 8.80E-17 0.62 Metabolism (Glycolysis/

gluconeogenesis)36 Serine (or cysteine) proteinase

inhibitor, clade H, 18393057 46,602 8.88 30 131 62 5.60E-09 0.66 Chaperone

61 Heterogenous nuclearribonucleoprotein B0a

157059863 32,515 8.74 21 95 56 2.30E-05 0.66 RNA binding

18 M2 pyruvate kinase 206205 58,314 7.15 35 159 66 8.90E-12 0.69 Metabolism (Glycolysis)68 Tyr 3/tryp 5-monooxygenase

activation protein13928824 29,274 4.55 17 66 58 1.60E-02 0.69 Redox

69 14-3-3 zeta isoform 1051270 27,955 4.73 29 112 75 4.40E-07 0.73 Signal transduction40 Mitochondrial aldehyde

dehydrogenase25990263 53,809 5.7 30 168 56 1.10E-12 0.77 Metabolism (Glycolysis)

59 Annexin A2 9845234 38,939 7.55 34 176 77 1.80E-13 0.77 Calcium-dependentphospholipid binding

60 Nuclear ribonucleoproteinA2/B1soform A2

4504447 36,041 8.67 15 70 36 6.30E-03 0.77 RNA binding

74 Peroxiredoxin 2 8394432 21,944 5.34 9 72 33 4.30E-03 0.78 Redox77 Cofilin 1 8393101 18,749 8.22 11 72 70 5.00E-03 0.78 Cytoskeleton37 Aldehyde dehydrogenase 2 14192933 56,966 6.63 23 96 47 1.80E-05 0.79 Metabolism (Glycolysis/

gluconeogenesis)58 Akr1b8 protein (Aldo-keto

reductase family 1, B8)51480424 34,891 8.58 14 66 54 1.80E-02 0.8 Redox

48 Actin, cytoplasmic 2 (Gamma-actin) 109507063 42,109 5.31 25 105 50 2.20E-06 0.81 Cytoskeleton63 Annexin A5 51858950 35,807 4.93 30 222 75 4.40E-18 0.82 Calcium-dependent

phospholipid binding49 Actin, cytoplasmic 2 (Gamma-actin) 109507063 42,109 5.31 28 126 62 1.80E-08 0.84 Cytoskeleton57 Aldehyde reductase 1 6978491 36,230 6.26 29 182 73 4.40E-14 0.85 Metabolism (Glycolysis/

gluconeogenesis)78 Glutathione S-transferase, pi 1 25453412 23,424 6.89 9 65 50 2.20E-02 0.85 Metabolism (Glutathione)41 Alpha enolase 56757324 47,440 6.16 40 244 79 2.80E-20 0.88 Metabolism (Glycolysis)64 Annexin A4 55742832 35,871 5.43 23 129 68 8.90E-09 0.88 Calcium-dependent

phospholipid binding17 Vimentin 14389299 53,757 5.06 54 329 90 8.90E-29 0.91 Cytoskeleton47 Laminin receptor 1 839369 32,917 4.8 19 97 52 1.40E-05 0.92 Laminin binding46 Ribonuclease inhibitor 77416905 51,653 4.67 16 74 53 2.80E-03 0.94 Protein binding21 Transketolase, isoform CRA_b 149034222 68,314 7.23 43 247 63 1.40E-20 0.95 Metabolism (Pentose

phosphate pathway)67 Chain A, Quinone reductase 1827606 30,853 8.45 10 65 42 2.20E-02 0.97 Biosynthesis (Steroids)83 APR3 actin-related protein

3 homolog23956222 47,783 5.61 21 118 46 1.10E-07 0.98 Cytoskeleton

5 Gelsolin, isoform CRA_b 149038929 81,064 5.46 36 133 37 3.50E-09 0.99 Actin binding7 Heat shock protein 8 13242237 71,055 5.37 32 94 37 3.00E-05 1.05 Chaperone53 Capping protein (actin filament),

gelsolin-like61556900 38,775 6.11 22 133 55 3.50E-09 1.06 Calcium-dependent

phospholipid binding4 Gelsolin 51854227 86,413 5.76 29 70 25 6.60E-03 1.07 Actin binding20 Transketolase, isoform CRA_b 149034222 68,314 7.23 34 173 53 3.50E-13 1.08 Metabolism (Pentose

phosphate pathway)62 Nucleophosmin 1 7242160 32,711 4.62 16 68 43 1.20E-02 1.1 RNA binding28 Protein disulfide isomerase

associated 38393322 57,010 5.88 44 251 62 5.60E-21 1.11 Chaperone

3 Valosin-containing protein 17865351 89,977 5.14 40 98 36 9.90E-06 1.16 ATP binding6 Gelsolin, isoform CRA_b 149038929 84,026 5.46 26 130 27 6.60E-03 1.17 Actin binding10 Stress-70 protein, mitochondrial

precursor (HSPa9a)116242506 74,097 5.97 24 67 29 1.40E-02 1.19 Chaperone

13 Chaperonin containing TCP1,subunit 5 (epsilon)

51890129 59,955 5.51 23 71 41 5.70E-03 1.19 Chaperone

19 rCG52516 (Glycyl-tRNA synthaseclass II)

149033299 82,482 6.15 28 106 34 1.80E-06 1.19 Translation

(continued on next page)

301M. Machida et al. / International Journal of Cardiology 153 (2011) 296–305

Table 1 (continued)

Spotno.

Protein name Accessionno. (gi)

Theoretical Matchedpeaksa

Scoreb Coverage(%)c

Probabilityd fCS/daCSratioe

Biological function

M.W.(kDa)

pI

11 Nuclear ribonucleoprotein K,isoform CRA_b

149039794 43,059 5.43 30 106 53 1.80E-06 1.21 RNA binding

22 Stress-induced phosphoprotein 1(HSP70/HSP90)

20302113 63,158 6.4 41 156 61 1.80E-11 1.24 Chaperone

38 Alpha enolase 56757324 47,440 6.16 25 143 51 3.50E-10 1.26 Metabolism (Glycolysis/gluconeogenesis)

23 Tyrosyl-tRNA synthetase 71043618 59,420 6.57 39 200 57 7.00E-16 1.28 Translation55 Annexin A1 6978501 39,147 6.97 26 169 70 8.90E-13 1.28 Protein binding29 Glucose-6-phosphate

dehydrogenase204197 55,307 6.19 49 311 70 5.60E-27 1.29 Metabolism (Pentose

phosphate pathway)2 Heat shock protein 1, beta 40556608 83,571 4.97 42 131 45 5.60E-09 1.31 Chaperone33 UDP-glucose dehydrogenase,

isoform CRA_a149035357 55,537 7.05 20 69 39 9.10E-03 1.33 Metabolism (Nucleotide

sugars)16 Tubulin, beta 2c 40018568 50,225 4.79 29 131 51 5.60E-09 1.49 Cytoskeleton70 Chloride intracellular channel 1 50657380 27,306 5.09 16 114 65 2.80E-07 1.55 Channel26 Vacuolar H+ ATPase B2 17105370 56,857 5.57 22 92 52 4.30E-05 1.6 Transcription71 Glutathione S-transferase, mu 1 8393502 26,068 8.27 20 110 63 7.00E-07 1.63 Metabolism (Glutathione)44 Phosphoglycerate kinase 1 40254752 44,909 8.02 47 285 82 2.20E-24 1.65 Metabolism (Glycolysis/

gluconeogenesis)54 Transaldolase 12002054 37,624 6.57 23 117 52 1.40E-07 1.7 Metabolism (Pentose

phosphate pathway)50 Guanine deaminase 148747414 51,439 5.48 25 129 56 8.90E-09 1.71 Metabolism (Purine

metabolism)74 rCG5116 (Peroxiredoxin 2) 8394432 21,944 5.34 9 72 33 4.30E-03 1.73 Redox31 Aldehyde dehydrogenase family 3,

member A114010869 50,706 6.33 20 65 38 2.30E-02 1.76 Metabolism (Glycolysis/

gluconeogenesis)32 Leucine aminopeptidase 3 58865398 56,514 6.77 31 139 55 8.80E-10 1.76 Protein degradation39 Cytokeratin 8 polypeptide 203734 52,678 5.49 43 181 65 5.60E-14 1.85 Cytoskeleton51 Annexin A7 148539909 50,272 5.91 29 111 44 5.60E-07 1.92 Calcium-dependent

phospholipid binding42 Alpha enolase 56757324 47,440 6.16 27 128 65 1.10E-08 1.93 Metabolism (Glycolysis)56 Transaldorase 1, isoform CRA_a 149061610 35,745 8.23 26 129 52 8.90E-09 2.03 Metabolism (Pentose

phosphate pathway)30 Chaperonin containing TCP1,

subunit 2 (beta)54400730 57,764 6.1 37 196 71 1.80E-15 2.11 Chaperone

65 Guanine nucleotide-bindingprotein β polypeptide 2-1

18543331 35,529 7.6 19 123 67 3.50E-08 2.15 Signal transduction

34 UDP-glucose dehydrogenase,isoform CRA_a

149035357 55,537 7.05 30 156 66 1.80E-11 2.25 Metabolism (Nucleotidesugars)

66 Vdac 1 protein (Voltage-dependent anion channel 1)

38051979 31,946 8.35 12 64 57 2.50E-02 2.41 Channel

82 Acidic ribosomalphosphoprotein P0

11693176 34,365 5.91 19 85 62 2.10E-04 2.51 Translation

8 Grp75 1000439 73,984 5.87 23 91 30 5.70E-05 3.44 Chaperone45 Aldolase A 6978487 39,783 8.31 26 155 74 2.20E-11 4.58 Metabolism (Glycolysis)76 Peroxiredoxin 5, isoform CRA_b 149062197 18,967 9.22 13 86 67 1.80E-04 5.25 Redox

Proteins in spots were determined with a MALDI-TOF MS system as described in Materials and methods. Spot numbers correspond to those in Fig. 3. Accession number, theoreticalmolecular weight (MW) and isoelectric point (pI), matched peaks, score, sequence coverage and probability were obtained from the NCBInr database with the MASCOT searchengine.

a Matched peaks indicate that the number of observed peptides corresponds to ideal peptides by digestion of trypsin.b Protein scores (score) were calculated by adding the identified peptide scores.c Coverage indicates the sequence coverage according to identified peptides.d Probability indicates that the observed match is random.e Spot volume ratios (fCS/daCS) were measured with PDQuest software.

302 M. Machida et al. / International Journal of Cardiology 153 (2011) 296–305

reflection of cell heterogeneity or the pluripotent nature of the line,we cloned CSC-BC21 by limited dilution and obtained CSC-21E, whichshowed the similar morphology and mRNA expression profile withCSC-BC21, expressing both cardiac and adipocyte markers whenmaintained in CSM on bacterial dishes. Unlike CSC-BC21, however,CSC-21E cells could be induced to express troponin I gene indicatingthe property likely to be in the intermediate stage of developmenttowards cardiomyocytes. The cardiosphere on bacterial dishes and theattachment to the culture dishes was reversible.

The protein profile of CSC-21E in Table 1 appears to reveal uniquecharacteristics, namely high levels of glycolytic and metabolicenzymes, a variety of chaperones or stress proteins, and annexins inboth shapes. Recently, several proteomic analyses of embryonic stemcells were reviewed for mice and humans [33,34]. These comparisons

showed that alongwith species-specific and cell line-specific proteins,some proteins were indicative of stemness. In both human andmurine ES cell lines, van Hoof and colleagues identified 191 proteinsas exclusively expressed in ES cells but not in differentiatedderivatives [33]. Baharvand et al. [34] found 92 commonly identifiedproteins in stem cells by combining datasets for adult stem cellsincluding mesenchymal stem cells, hematopoietic stem cells andneural stem cells. A comparison of these lists with our Table 1 showedthat we identified some of the proteins listed as characteristic ofstemness such as 75 kDa glucose-related protein (grp 75; spot 8),alpha enolase (spot 42), ATP synthase H+ transporting mitochondrialF1 complex (spot 25) and calreticulin (spot 14), but some defined asthe indicators of stemness by van Hood and colleagues [33] wereabsent including peroxiredoxin 1, T-complex protein 1 and

Table 2ASpots with flCS/daCS ratio greater than 1.5.

Spotno.

Protein name Accessionno. (gi)

Theoretical Matchedpeaksa

Scoreb Coverage(%)c

Probabilityd fCS/daCSratioe

Biological function

M.W.(kDa)

pI

44 Phosphoglycerate kinase 1 40254752 44,909 8.02 47 285 82 2.20E-24 1.65 Metabolism (Glycolysis/gluconeogenesis)31 Aldehyde dehydrogenase

family 3, member A114010869 50,706 6.33 20 65 38 2.30E-02 1.76 Metabolism (Glycolysis/gluconeogenesis)

42 Alpha enolase 56757324 47,440 6.16 27 128 65 1.10E-08 1.93 Metabolism (Glycolysis/gluconeogenesis)45 Aldolase A 6978487 39,783 8.31 26 155 74 2.20E-11 4.58 Metabolism (Glycolysis)54 Transaldolase 12002054 37,624 6.57 23 117 52 1.40E-07 1.7 Metabolism (Pentose phosphate pathway)56 Transaldolase 1, isoform CRA_a 149061610 35,745 8.23 26 129 52 8.90E-09 2.03 Metabolism (Pentose phosphate pathway)34 UDP-glucose dehydrogenase,

isoform CRA_a149035357 55,537 7.05 30 156 66 1.80E-11 2.25 Metabolism (Nucleotide sugars)

71 Glutathione S-transferase, mu 1 8393502 26,068 8.27 20 110 63 6/E-7 1.63 Metabolism (Glutathione metabolism)50 Guanine deaminase 148747414 51,439 5.48 25 129 56 8.90E-09 1.71 Metabolism (Purine metabolism)73 Ferritin, heavy polypeptide 1 6978859 21,086 5.62 9 67 35 0.013 1.73 Metabolism (Iron metabolism)30 Chaperonin containing TCP1,

subunit 2 (beta)54400730 57,764 6.1 37 196 71 1.80E-15 2.11 Chaperone

8 Grp75 1000439 73,984 5.87 23 91 30 5.70E-05 3.44 Chaperone39 Cytokeratin 8 polypeptide 203734 52,678 5.49 43 181 65 5.60E-14 1.85 Cytoskeleton76 Peroxiredoxin 5, isoform CRA_b 149062197 18,967 9.22 13 86 67 0.00018 5.25 Redox82 Acidic ribosomal phosphoprotein P0 11693176 34,365 5.91 19 85 62 0.00021 2.51 Translation26 Vacuolar H+ ATPase B2 17105370 56,857 5.57 22 92 52 4.30E-05 1.6 Transcription51 Annexin A7 148539909 50,272 5.91 29 111 44 5.60E-07 1.92 Calcium ion transport70 Chloride intracellular channel 1 50657380 27,306 5.09 16 114 65 2.80E-07 1.55 Channel66 Vdac 1 protein

(Voltage-dependent anion channel)38051979 31,946 8.35 12 64 57 0.025 2.41 Channel

32 Leucine aminopeptidase 3 58865398 56,514 6.77 31 139 55 8.80E-10 1.76 Protein degradation65 Guanine nucleotide

bindingproteinβpolypeptide2-118543331 35,529 7.6 19 123 67 3.50E-08 2.15 Signal transduction

a Matched peaks indicate that the number of observed peptides corresponds to ideal peptides by digestion of trypsin.b Protein scores (score) were calculated by adding the identified peptide scores.c Coverage indicates the sequence coverage according to identified peptides.d Probability indicates that the observed match is random.e Spot volume ratios (fCS/daCS) were measured with PDQuest software.

303M. Machida et al. / International Journal of Cardiology 153 (2011) 296–305

translationally controlled tumor protein. The absence might be due tolimitations of the present analysis, or, alternatively, reflecting thedifference between ES cells and CSC-21E cells. Caution is requiredwhen interpreting our results since no proteomic analysis of rat EScells is currently available. The present CSC-21E cells, however,express some proteins indicative of stemness, which is consistentwith the view that they retain some stem-like characteristics.

Table 2BSpots with daCS/flCS ratio greater than 1.5.

Spotno.

Protein name Accessionno. (gi)

Theoret.mass

(kDa) pI

84 Lactate dehydrogenase A 8393706 36,712 8.45

43 Phosphoglycerate kinase 1 40254752 44,909 8.02

14 Calreticulin 11693172 48,137 4.3327 Protein disulfide isomerase associated 3 8393322 57,010 5.8812 Chaperonin 60 1778213 61,029 5.781 rCG49111 (HSP90) 149067319 92,998 4.7236 Serine (or cysteine) proteinase inhibitor, clade H,1 8393057 46,602 8.8824 Keratin complex 1, acidic, gene 14 56912233 52,936 5.0825 ATPsynthase mitochondrial F1complex α subunit 149029483 54,632 8.2415 Iodothyronine 5′monodeiodinase 202549 54,375 4.8735 Eukaryotic translation elongation factor 1 alpha 1 28460696 50,424 9.19 Annexin A6 1351943 76,106 5.3961 Heterogenous nuclear ribonucleoprotein B0a 157059863 32,515 8.74

a Matched peaks indicate that the number of observed peptides corresponds to ideal pepb Protein scores (score) were calculated by adding the identified peptide scores.c Coverage indicates the sequence coverage according to identified peptides.d Probability indicates that the observed match is random.e Spot volume ratios (fCS/daCS) were measured with PDQuest software.

By using sensitive 2-D DIGE system, we were able to detectproteins which change their expression levels in response to theshape change from floating aggregates to dish-attached expandedcells. Metabolic enzymes comprise about half of the proteins whoseexpression decreased upon CS attachment. In the flCS cells, oxygensupply is limited by aggregation which minimizes the exposure to theenvironment. Stem cell niches are often located in regions of low

Matchedpeaksa

Scoreb Coverage(%)c

Probabilityd fCS/daCSratioe

Biological function

32 192 73 4.40E-15 0.5 Metabolism (Glycolysis/gluconeogenesis)

38 209 72 8.80E-17 0.62 Metabolism (Glycolysis/gluconeogenesis)

24 125 58 2.20E-08 0.43 Chaperone37 187 61 1.40E-14 0.48 Chaperone28 100 50 5.70E-03 0.49 Chaperone36 80 35 7.70E-04 0.51 Chaperone30 131 62 5.60E-09 0.66 Chaperone43 204 67 2.80E-16 0.5 Cytoskeleton21 68 43 1.10E-02 0.38 Redox29 153 60 3.50E-11 0.49 Redox25 105 50 2.20E-06 0.42 Translation37 108 43 1.10E-06 0.26 Calcium ion transport21 95 56 2.30E-05 0.66 RNA processing

tides by digestion of trypsin.

304 M. Machida et al. / International Journal of Cardiology 153 (2011) 296–305

oxygen supply [35], and glycolysis is called upon to supply energy forthe growth of cells. Kondoh and colleagues report that theproliferative capacity of murine ES cells is correlated with highactivity of glycolytic enzymes and low oxygen consumption [36].Unwin and Whetton summarized in their paper [13] that their stemcell proteomic analysis identified changes in several components ofanaerobic glycolysis whose expression lowered as the cells lost stemcell characteristics. On the other hand, levels of glutathione S-transferase, pentose phosphate pathway enzymes, transaldorase(spot 54) and a transaldorase isoform (spot 56), were higher in flCSthan daCS. Glutathione metabolism mainly acts as a defense againstthe accumulation of reactive oxygen species (ROS) by reducingcellular disulfides. Thioredoxin and glutathione systems are reducedby NADPH, which could be generated by the pentose phosphatepathway [37]. Consequently, the decrease of these proteins suggeststhat flCS could avoid oxidative stress better than daCS. Saretzki et al.reported that glutathione S-transferase and glutathione peroxire-doxin 2, 3, 4 genes were down-regulated during the differentiation ofES cells [38], and generally the defense system for oxidative stress isbetter in stem cells than differentiated cells [39]. Therefore, flCS withsuperior anti-oxidative capacity over daCS appears to indicate thatflCS cells maintain more stemness properties than daCS cells. Thereduction of these enzymes seems to indicate that the attachment ofcardiosphere clone CSC-21E to the dish surfaces promoted thedifferentiation of CS cells.

Stress proteins were the category of proteins highly expressed inboth flCS and daCS. The group also comprises a fraction of the proteinsnewly emerged or decreased after the attachment of flCS cells todishes. Stress proteins are involved in regulating protein folding invarious cell compartments and in intracellular trafficking [21]. Theyare also implicated in the stemness of pluripotent cells, and theregulation of cell proliferation and differentiation [40]. One suchexample is HSP90, a highly conserved family of molecular chaperones,which not only participate in stabilizing and refolding denaturedproteins after stress, but also have key roles in thematuration of signaltransduction proteins [41]. In Zebrafish, Hsp90α1 plays an importantrole in muscle development conceivably through facilitating myosinfolding and assembly into organized myofibril filaments [42]. Amongthe stress proteins up-regulated upon the cell attachment, Calreticulinis most notable, because it is known to be essential for cardiacdevelopment [43]. Although Calreticulin is only a minor componentof the mature heart, it is highly expressed transiently in thedeveloping heart, and homozygous gene knockout mice are embryolethal with the significant defect in the morphology of the heartshaving disarray in myofibrils [25,43]. The up-regulation of Hsp90 andCalreticulin after the transition from flCS to daCS clearly indicates thatthis is a key step towards differentiation to cardiomyocytes. Theincrease in Calreticulin expression induces the level of both vinculinand N-cadherin, which play important roles in increasing celladhesiveness to their substratum and in communicating with theirneighboring cells.

Another striking change observed in the present flCS to daCStransition is the switching of annexins A7 to A6. Annexins are a familyof membrane-binding proteins [44,45]. They bind to phospholipidsurfaces in a Ca++-dependent manner, and broadly function in themembrane including vesicle trafficking, membrane repair, membranefusion, ion channel formation, cell differentiation and transformation.We observed high expression levels of annexins A1 and A2, and thedistinct expression of annexins A4 andA5 in bothflCS and daCS. On theother hand, annexin A7 (spot 51) expressed in flCS cells was decreasedin daCS while the expression of annexin A6 (spot 9) was increased indaCS. Annexin A6 is known to function in the organization ofmembrane domains [46,47] in a Ca++-dependent manner [48], andmight regulate actin assembly in nerve terminals [49]. Annexin A7 is aCa++-dependent membrane fusion protein with GTPase activity [50].The change from annexin A7 to A6 might suggest a different

requirement for the cytoskeletal organization and the plasmamembrane formation in the floating and attached CSC-21E cells.

Proteomic profiles of both flCS and daCS exhibited some of theproteins characteristically observed as the indicators of stemness. Aminor fraction of proteins changed their expression levels from flCS todaCS switch, notably the down regulation of glycolytic and othermetabolic enzymes, and the up-regulation of Hsp90 and Calreticulin,the latter known to be key protein in the cardiac development. Thesechanges strongly suggest that flCS to daCS transition of CSC-21E cellsrepresents a critical step for the commitment to cardiomyocytedifferentiation. The use of 2-D DIGE and MALDI-TOF mass demon-strates that proteomics comparison provides appropriate means forcharacterizing and identifying developmental stages of the cells.

Author disclosure statement

All of the authors of this report are free from conflict of interestconcerning the contents of the present report.

Acknowledgements

The authors thank Mr. Ryota Nakao and Mr. Daisuke Ogawa fortechnical help.

This work was supported by the International Research andEducational Institute for Integrated Medical Sciences, TokyoWomen'sMedical University which is supported by the Program for Promotingthe Establishment of Strategic Research Centers, Special CoordinationFunds for Promoting Science and Technology, Japan and a Grant-in-Aid for Scientific Research (A) (Japan). The authors of this manuscripthave certified that they comply with the Principles of EthicalPublishing in the International Journal of Cardiology [51].

References

[1] Beltrami AP, Barlucchi L, Torella D, et al. Adult cardiac stem cells are multipotentand support myocardial regeneration. Cell 2003;114:763–76.

[2] Anversa P, Nadal-Ginard B. Myocyte renewal and ventricular remodelling. Nature2002;415:240–3.

[3] Matsuura K, Nagai T, Nishigaki N, et al. Adult cardiac sca-1-positive cellsdifferentiate into beating cardiomyocytes. J Biol Chem 2004;279:11384–91.

[4] Linke A, Müller P, Nurzynska D, et al. Stem cells in the dog heart are self-renewing,clonogenic, and multipotent and regenerate infarcted myocardium, improvingcardiac function. Proc Natl Acad Sci USA 2005;102:8966–71.

[5] Bearzi C, Rota M, Hosoda T, et al. Human cardiac stem cells. Proc Natl Acad Sci USA2007;104:14068–73.

[6] Miyamoto S, Kawaguchi N, Ellison GM, Matsuoka R, Shin'oka T, Kurosawa H.Characterization of long-term cultured c-kit+cardiac stem cells derived fromadult rat hearts. Stem Cells Dev 2010;19:105–16.

[7] Suslov ON, Kukekov VG, Ignatova TN, Steindler DA. Neural stem cell heterogeneitydemonstrated by molecular phenotyping of clonal neurospheres. Proc Natl AcadSci USA 2002;99:14506–11.

[8] Gissel C, Voolstra C, Doss MX, et al. An optimized embryonic stem cell model forconsistent gene expression and developmental studies: a fundamental study.Thromb Haemost 2005;94:719–27.

[9] Cai J, Chen J, Liu Y, et al. Assessing self-renewal and differentiation in humanembryonic stem cell lines. Stem Cells 2006;24:516–30.

[10] Suárez-Fariñas M, Noggle S, Heke M, Hemmati-Brivanlou A, Magnasco MO.Comparing independent microarray studies: the case of human embryonic stemcells. BMC Genomics 2005;6:99.

[11] Kawaguchi N, Nakao R, Yamaguchi M, Ogawa D, Matsuoka R. TGF-β Superfamilyregulates a switch that mediates differentiation either into adipocytes ormyocytes in left atrium derived pluripotent cells (LA-PCs). Biochem Biophys ResCommun 2010;396:615–25.

[12] Gygi SP, Rochon Y, Franza BR, Aebersold R. Correlation between protein and mrnaabundance in yeast. Mol Cell Biol 1999;19:1720–30.

[13] Unwin RD, Whetton AD. Systematic proteome and transcriptome analysis of stemcell populations. Cell Cycle 2006;5:1587–91.

[14] Martin GR, Evans MJ. Differentiation of clonal lines of teratocarcinoma cells:formation of embryoid bodies in vitro. Proc Natl Acad Sci USA 1975;72:1441–5.

[15] Bradford MM. A rapid and sensitive method for the quantitation of microgramquantities of protein utilizing the principle of protein–dye binding. Anal Biochem1976;72:248–54.

[16] Machida M, Kosako H, Shirakabe K, et al. Purification of phosphoproteins byimmobilized metal affinity chromatography and its application to phosphopro-teome analysis. FEBS J 2007;274:1576–87.

305M. Machida et al. / International Journal of Cardiology 153 (2011) 296–305

[17] Shevchenko A, Wilm M, Vorm O, Mann M. Mass spectrometric sequencing ofproteins silver-stained polyacrylamide gels. Anal Chem 1996;68:850–8.

[18] Toda T, Kimura N. Standardization of 2-D PAGE and construction of 2-D PAGEprotein database on World Wide Web home page. Jpn J Electroph 1997;41:13–20.

[19] Corti S, Locatelli F, Papadimitriou D, et al. Identification of a primitive brain-derived neural stem cell population based on aldehyde dehydrogenase activity.Stem Cells 2006;24:975–85.

[20] Akum BF, Chen M, Gunderson SI, Riefler GM, Scerri-Hansen MM, Firestein BL.Cypin regulates dendrite patterning in hippocampal neurons by promotingmicrotubule assembly. Nat Neurosci 2004;7:145–52.

[21] Young JC, Agashe VR, Siegers K, Hartl FU. Pathways of chaperone-mediatedprotein folding in the cytosol. Nat Rev Mol Cell Biol 2004;5:781–91.

[22] Wadhwa R, Takano S, Taira K, Kaul SC. Reduction in mortalin level by its antisenseexpression causes senescence-like growth arrest in human immortalized cells.J Gene Med 2004;6:439–44.

[23] Caniggia I, Winter J, Lye SJ, Post M. Oxygen and placental development during thefirst trimester: Implications for the pathophysiology of pre-eclampsia. Placenta2000;21:Suppl A:S25–30.

[24] Orsi NM, Leese HJ. Protection against reactive oxygen species during mousepreimplantation embryo development. Role of edta, oxygen tension, catalase,superoxide dismutase and pyruvate. Mol Reprod Dev 2001;59:44–53.

[25] Lozyk MD, Papp S, Zhang X, Nakamura K, Michalak M, Opas M. Ultrastructuralanalysis of development of myocardium in calreticulin-deficient mice. BMC DevBiol 2006;6:54.

[26] Ishida Y, Kubota H, Yamamoto A, Kitamura A, Bächinger HP, Nagata K. Type icollagen in hsp47-null cells is aggregated in endoplasmic reticulum and deficientin n-propeptide processing and fibrillogenesis. Mol Biol Cell 2006;17:2346–55.

[27] Lahav J, Wijnen EM, Hess O, et al. Enzymatically catalyzed disulfide exchange isrequired for platelet adhesion to collagen via integrin alpha2beta1. Blood2003;102:2085–92.

[28] Lahav J, Gofer-Dadosh N, Luboshitz J, Hess O, Shaklai M. Protein disulfideisomerase mediates integrin-dependent adhesion. FEBS Lett 2000;475:89–92.

[29] Goplen D, Wang J, Enger PO, et al. Protein disulfide isomerase expression isrelated to the invasive properties of malignant glioma. Cancer Res 2006;66:9895–902.

[30] Smith RR, Barile L, Cho HC, et al. Regenerative potential of cardiosphere-derivedcells expanded from percutaneous endomyocardial biopsy specimens. Circulation2007;115:896–908.

[31] Messina E, De Angelis L, Frati G, et al. Isolation and expansion of adult cardiac stemcells from human and murine heart. Circ Res 2004;95:911–21.

[32] Andersen DC, Andersen P, Schneider M, Jensen HB, Sheikh SP. Murine “Cardio-spheres” are not a source of stem cells with cardiomyogenic potential. Stem Cells2009;27:1571–81.

[33] Van Hoof D, Passier R,Ward-Van Oostwaard D, et al. A quest for human andmouseembryonic stem cell-specific proteins. Mol Cell Proteomics 2006;5:1261–73.

[34] Baharvand H, Fathi A, van Hoof D, Salekdeh GH. Concise review: trends in stem cellproteomics. Stem Cells 2007;25:1888–903.

[35] Dellatore SM, Garcia AS, Miller WM. Mimicking stem cell niches to increase stemcell expansion. Curr Opin Biotechnol 2008;19:534–40.

[36] Kondoh H, Lleonart ME, Nakashima Y, et al. A high glycolytic flux supports theproliferative potential of murine embryonic stem cells. Antioxid Redox Signal2007;9:293–9.

[37] Schafer FQ, Buettner GR. Redox environment of the cell as viewed through theredox state of the glutathione disulfide/glutathione couple. Free Radic Biol Med2001;30:1191–212.

[38] Saretzki G, Walter T, Atkinson S, et al. Downregulation of multiple stress defensemechanisms during differentiation of human embryonic stem cells. Stem Cells2008;26:455–64.

[39] Saretzki G, Armstrong L, Leake A, Lako M, von Zglinicki T. Stress defense in murineembryonic stem cells is superior to that of various differentiated murine cells.Stem Cells 2004;22:962–71.

[40] Prinsloo E, Setati MM, Longshaw VM, Blatch GL. Chaperoning stem cells: a role forheat shock proteins in the modulation of stem cell self-renewal and differenti-ation? Bioessays 2009;31:370–7.

[41] Chen B, Zhong D, Monteiro A. Comparative genomics and evolution of the hsp90family of genes across all kingdoms of organisms. BMC Genomics 2006;7:156.

[42] Du SJ, Li H, Bian Y, Zhong Y. Heat-shock protein 90alpha1 is required for organizedmyofibril assembly in skeletal muscles of zebrafish embryos. Proc Natl Acad SciUSA 2008;105:554–9.

[43] Mesaeli N, Nakamura K, Zvaritch E, et al. Calreticulin is essential for cardiacdevelopment. J Cell Biol 1999;144:857–68.

[44] Gerke V, Moss SE. Annexins: from structure to function. Physiol Rev 2002;82:331–71.

[45] Moss SE, Morgan RO. The annexins. Genome Biol 2004;5:219.[46] Pons M, Grewal T, Rius E, Schnitgerhans T, Jäckle S, Enrich C. Evidence for the

involvement of annexin 6 in the trafficking between the endocytic compartmentand lysosomes. Exp Cell Res 2001;269:13–22.

[47] Babiychuk EB, Palstra RJ, Schaller J, Kämpfer U, Draeger A. Annexin vi participatesin the formation of a reversible, membrane–cytoskeleton complex in smoothmuscle cells. J Biol Chem 1999;274:35191–5.

[48] Orito A, Kumanogoh H, Yasaka K, et al. Calcium-dependent association ofannexin vi, protein kinase c alpha, and neurocalcin alpha on the raft fractionderived from the synaptic plasma membrane of rat brain. J Neurosci Res2001;64:235–41.

[49] Tanaka K, Tashiro T, Sekimoto S, Komiya Y. Axonal transport of actin and actin-binding proteins in the rat sciatic nerve. Neurosci Res 1994;19:295–302.

[50] Caohuy H, Pollard HB. Protein kinase c and guanosine triphosphate combine topotentiate calcium-dependent membrane fusion driven by annexin 7. J Biol Chem2002;277:25217–25.

[51] Coats AJ. Ethical authorship and publishing. Int J Cardiol 2009;131:149–50.