University of ZurichZurich Open Repository and Archive

Winterthurerstr. 190

CH-8057 Zurich

http://www.zora.uzh.ch

Year: 2010

Ascorbic acid improves embryonic cardiomyoblast cell survivaland promotes vascularization in potential myocardial grafts in

vivo

Martinez, E C; Wang, J; Gan, S U; Singh, R; Lee, C N; Kofidis, T

Martinez, E C; Wang, J; Gan, S U; Singh, R; Lee, C N; Kofidis, T (2010). Ascorbic acid improves embryoniccardiomyoblast cell survival and promotes vascularization in potential myocardial grafts in vivo. TissueEngineering. Part A, 16(4):1349-1361.Postprint available at:http://www.zora.uzh.ch

Posted at the Zurich Open Repository and Archive, University of Zurich.http://www.zora.uzh.ch

Originally published at:Tissue Engineering. Part A 2010, 16(4):1349-1361.

Martinez, E C; Wang, J; Gan, S U; Singh, R; Lee, C N; Kofidis, T (2010). Ascorbic acid improves embryoniccardiomyoblast cell survival and promotes vascularization in potential myocardial grafts in vivo. TissueEngineering. Part A, 16(4):1349-1361.Postprint available at:http://www.zora.uzh.ch

Posted at the Zurich Open Repository and Archive, University of Zurich.http://www.zora.uzh.ch

Originally published at:Tissue Engineering. Part A 2010, 16(4):1349-1361.

Ascorbic acid improves embryonic cardiomyoblast cell survivaland promotes vascularization in potential myocardial grafts in

vivo

Abstract

Organ restoration via cell therapy and tissue transplantation is limited by impaired graft survival. Wetested the hypothesis that ascorbic acid (AA) reduces cell death in myocardial grafts both in vitro and invivo and introduced a new model of autologous graft vascularization for later transplantation. Luciferase(Fluc)- and green fluorescent protein (GFP)-expressing H9C2 cardiomyoblasts were seeded in gelatinscaffolds to form myocardial artificial grafts (MAGs). MAGs were supplemented with AA (5 or 50mumol/L) or plain growth medium. Bioluminescence imaging showed increased cell photon emissionfrom day 1 to 5 in grafts supplemented with 5 mumol/L (p < 0.001) and 50 mumol/L (p < 0.01) AA. Theamount of apoptotic cells in plain MAGs was significantly higher than in AA-enriched grafts. In our invitro model, AA also enhanced H9C2 cell myogenic differentiation. For in vivo studies, MAGscontaining H9C2-GFP-Fluc cells and enriched with AA (n = 10) or phosphate-buffered saline (n = 10)were implanted in the renal pouch of Wistar rats. At day 6, postimplantation bioluminescence signalsdecreased by 74% of baseline in plain MAGs versus 36% in AA-enriched MAGs (p < 0.0001). AAgrafts contained significantly higher amounts of blood vessels, GFP(+) donor cells, and endothelialcells. In this study, we identified AA as a potent supplement that improves cardiomyoblast survival andpromotes neovascularization in bioartificial grafts.

Ascorbic Acid Improves Embryonic CardiomyoblastCell Survival and Promotes Vascularization

in Potential Myocardial Grafts In Vivo

Eliana C. Martinez, M.D.,1 Jing Wang, DBMS,1 Shu Uin Gan, Ph.D.,1 Rajeev Singh, M.D.,1,2

Chuen Neng Lee, M.D.,1,3 and Theo Kofidis, M.D., Ph.D.1,3

Organ restoration via cell therapy and tissue transplantation is limited by impaired graft survival. We tested thehypothesis that ascorbic acid (AA) reduces cell death in myocardial grafts both in vitro and in vivo and intro-duced a new model of autologous graft vascularization for later transplantation. Luciferase (Fluc)- and greenfluorescent protein (GFP)-expressing H9C2 cardiomyoblasts were seeded in gelatin scaffolds to form myocardialartificial grafts (MAGs). MAGs were supplemented with AA (5 or 50 mmol=L) or plain growth medium. Bio-luminescence imaging showed increased cell photon emission from day 1 to 5 in grafts supplemented with5 mmol=L ( p< 0.001) and 50 mmol=L ( p< 0.01) AA. The amount of apoptotic cells in plain MAGs was signifi-cantly higher than in AA-enriched grafts. In our in vitro model, AA also enhanced H9C2 cell myogenic differ-entiation. For in vivo studies, MAGs containing H9C2-GFP-Fluc cells and enriched with AA (n¼ 10) orphosphate-buffered saline (n¼ 10) were implanted in the renal pouch of Wistar rats. At day 6, postimplantationbioluminescence signals decreased by 74% of baseline in plain MAGs versus 36% in AA-enriched MAGs( p< 0.0001). AA grafts contained significantly higher amounts of blood vessels, GFPþ donor cells, and endo-thelial cells. In this study, we identified AA as a potent supplement that improves cardiomyoblast survival andpromotes neovascularization in bioartificial grafts.

Introduction

The fabrication of sufficiently thick three-dimensional(3D) myocardial grafts by means of tissue engineering is

limited by impaired cell viability in unperfused in vitro sys-tems.1 Further, cell death constitutes a major drawback whenattempts are made to restore heart muscle via cell or bioar-tificial tissue implantation.2–4 Following transplantation, cellsin 3D grafts will be challenged by proapoptotic stimuli and anintense inflammatory response arising from the surroundingischemic tissue,5,6 as well as by hypoxia, particularly in thecore of the bioengineered graft.7 Besides donor cell survival,functional integration and engraftment of bioengineeredtissues in the recipient organism are equally critical for asuccessful myocardial restoration. It has been reported thatneovascularization within myocardial grafts plays a key rolefor both cardiomyocyte survival and organization.8 All to-gether highlights the paramount importance of blood supplyin myocardial grafts for heart repair.

Given the relevance of hypoxia and oxidative stress for thefate of cardiac cells after ischemic injury as well as within

thick bioengineered tissues, we postulate that ascorbic acid(AA) is a factor that might reduce cell death in myocardialgrafts both in vitro and in vivo. This hydrophilic antioxidantscavenges toxic free radicals and other reactive oxygen spe-cies efficiently.9 Further, attenuation of hypoxia-inducedapoptosis after AA treatment in vitro has been demonstratedin HL-1 cardiomyocytes10 and endothelial progenitor cells.11

AA also exerts a stimulatory effect on angiogenesis throughincrease of collagen type IV synthesis by endothelial cells.12

It has been shown that human umbilical vein endothelial celltype IV collagen production is enhanced when AA is addedin vitro. At physiological concentrations (up to 100mmol=L),AA induces tube formation by human umbilical vein endo-thelial cells in cultured extracellular matrix.12 Further, largedoses of AA induce superior mesenchymal tissue healing inrats, because of early angiogenesis induction and increasedcollagen synthesis.13 Finally, AA is available ubiquitously, isan essential part of our diet at a global scale, and can beeasily administered in large doses as a therapeutic supple-ment in various formulations. However, the potential of AAfor cell therapy and tissue repair has not been exploited yet.

1Department of Surgery, National University of Singapore, Singapore, Singapore.2Tissue Repository, National University Hospital=National University of Singapore, Singapore, Singapore.3Department of Cardiac, Thoracic, and Vascular Surgery, National University Hospital, Singapore, Singapore.

TISSUE ENGINEERING: Part AVolume 16, Number 4, 2010ª Mary Ann Liebert, Inc.DOI: 10.1089=ten.tea.2009.0399

1349

Experimental studies to assess the effect of AA on cell sur-vival in 3D compounds destined for myocardial restorationhave yet to be conducted.

Accordingly, the aim of our study was to investigatewhether bioartificial grafts (scaffold–cell compounds) formyocardial repair, which have been enriched with AA, dis-play improved donor cell viability, both in vitro and in vivo.In addition, we assess whether such enrichment would haveany effect on angiogenesis and remodeling of thick bioarti-ficial tissues, suitable for implantation and myocardial re-pair. To examine our hypothesis, a novel renal pouch modelfor in vivo prevascularization of 3D tissue grafts was devel-oped in healthy rats.

Materials and Methods

Cell culture

H9C2(2-1) cardiomyoblasts derived from embryonic rathearts were obtained from the American Type Culture Col-lection (Manassas, VA) and maintained in Dulbecco’s modi-fied Eagle medium (DMEM) supplemented with 10% fetalbovine serum, 100,000 U=L of penicillin, and 100 mg=L ofstreptomycin (GIBCO�, Invitrogen Corporation, Carlsbad,CA), under 5% CO2 at 378C. Cells were subcultured on reach-ing subconfluency to maintain the myoblast phenotype.

Generation of fluorescent=bioluminescent cell lines

pWPT-GFP (kindly provided by Dr. D. Trono; University ofGeneva, Geneva, Switzerland) and pWPT-Fluc (modified frompWPT-GFP) lentiviral vectors were used to infect H9C2 cells.pWPT-Fluc was cloned by removing green fluorescent protein(GFP) from pWPT-GFP and inserting the firefly luciferase(Fluc) gene PCRed from pGL3 (Promega, Madison, WI).Lentiviral vectors were produced by transient transfection of293T cells. About 5�106 cells=plate were seeded in 10-cmtissue culture plates at 24 h before transfection. The latter wasperformed using calcium phosphate precipitation methodwith 10mg pWPT-GFP or pWPT-Fluc vector, 7.5mg helperplasmid pCMV8.91, and 2.5mg of MD2G envelope plasmids(gifts from Dr. D. Trono, University of Geneva, Geneva,Switzerland). Cells were replaced with fresh medium at 14–16 h after transfection. The supernatant was filtered through a0.45-mm filter, and the titer of supernatant on 293T cells wasdetermined using flow cytometry. H9C2 cells were shown tobe highly infectable (>90%) with unconcentrated lentiviralsupernatant (unpublished data). H9C2 cells to be used in vivowere initially infected with supernatant containing pWPT-GFP, sorted for GFP positivity, and subsequently infected withpWPT-Fluc (H9C2-GFP-Fluc). Further, cells destined to in vitroexperiments were infected with pWPT-Fluc (H9C2-Fluc).

3D graft preparation for in vitro studies

Prior to these experiments, AA concentrations without cy-totoxic effect on H9C2 cells were established in our laboratory(data not shown) and have also been previously described inother cell types.12,14 Sterile, porous sponges prepared frompurified porcine skin gelatin (Gelfoam�; Pharmacia & UpjohnCompany, Kalamazoo, MI) were used as scaffold material tofabricate 3D myocardial artificial grafts (MAGs). Gelfoam

squares (10�10�5 mm) were placed in eight-well chamberslides (Lab-Tek�II Chamber Slide�; NUNC A=S, Roskilde,Denmark) under sterile conditions. Subsequently, foams werehydrated with 100mL growth medium supplemented with 5or 50mmol=L L-AA (Sigma, St. Louis, MO). Scaffolds with justmedium, that is, without AA and=or without cells, were usedas controls. A 100mL medium solution containing 2.5�105

H9C2-Fluc cells was added on top of the 3D scaffolds.Chamber slides were then placed in an incubator under 5%CO2 at 378C for 30 min to allow cell solution absorption intothe sponge. Next, scaffolds were covered with 0.4 mL growthmedium, placed back into the incubator, and the medium wasexchanged daily. Assays were conducted in quadruplet in fiveseparate experiments.

In vitro bioluminescence imaging

Bioluminescence imaging (BLI) was used to evaluate theeffect of AA on H9C2-Fluc cells in vitro survival whenseeded in 3D MAGs after 1, 3, and 5 days in culture (n¼ 20=treatment=time point). For BLI, 150mg=mL working solutionof d-luciferin (Caliper Life Sciences, Hopkinton, MA) inprewarmed cell culture medium was added to each well andthe chamber slides were scanned after 5 min using a XenogenIVIS� Lumina System (Caliper Life Sciences). Biolumines-cence was quantified in units of photons per second total flux(p=s), using Living Imaging Software version 2.6 (CaliperLife Sciences).

dUTP nick-end labeling assayand immunohistochemical stainingfor active caspase 3

Three-dimensional MAGs were fixed in 10% buffered for-malin, embedded in optimal cutting temperature (OCT)compound (Tissue-Tek�, Sakura Finetek, Tokyo, Japan), andstored at �808C after in vitro experiments at days 3 and 5.In situ terminal deoxynucleotidyl transferase-mediated dUTPnick-end labeling (TUNEL) analysis was done using theApopTag Red In Situ Apoptosis Detection Kit (Millipore,Billerica, MA), according to the manufacturer’s instructions,on 5-mm cryosections. The sections from at least three differentexperiments were visualized using a Leica TCS SP5, DMI6000confocal laser scanning microscope (Leica Microsystems,Wetzlar, Germany). The number of TUNEL-positive cells wascounted in 10 different fields (400�), from three differentgrafts=treatment. Next, to quantify apoptosis and excludenecrosis, 10-mm cryosections were stained with active caspase3 antibody (1:100, rabbit polyclonal antibody to active caspase3, ab2302; Abcam, Cambridge, United Kingdom). The sectionswere counterstained with 40,6-diamidino-2-phenylindole(DAPI) (Molecular Probes�, Invitrogen, Carlsbad, CA) andz-stacks in 0.5 mm steps were obtained using confocal micros-copy. A single image of maximum projection was obtainedfrom the z-stacks by using Leica Application Suite AdvancedFluorescence quantification software (version 1.8.2, build1465). The sections from at least three different experi-ments were analyzed. Cells were counted using Image J 1.42q(National Institutes of Health, Bethesda, MD). The percent-age of TUNEL-positive and active caspase 3-positive cells wascalculated as follows: 100�(number of positive cells counted=total number of nuclei counted).

1350 MARTINEZ ET AL.

Assessment of H9C2 phenotype in 3D culture

Immunohistochemical assays were carried out to evaluatethe effect of AA on H9C2 cardiomyoblast differentiation.Five-micrometer cryosections obtained from in vitro experi-ments at days 3 and 5 were stained with an antibody specificto a-cardiac and skeletal (sarcomeric) muscle actins (1:100,monoclonal mouse anti-sarcomeric actin, clone Alpha-Sr-1;DakoCytomatation, Glostrup, Denmark). The sections werethen incubated with a secondary antibody Alexa Fluor�-568,counterstained with DAPI (Molecular Probes, Invitrogen,Carlsbad, CA), and visualized with an Olympus BX61 fluo-rescence microscope equipped with a DP72 12.8 megapixelcooled digital camera (Olympus, Tokyo, Japan). Photomicro-graphs were processed using the DP2-BSW 2.2 (build 6212)software (Olympus).

Animals and renal pouch model

This study conforms with the Guide for the Care and Useof Laboratory Animals published by the U.S. National In-stitutes of Health (NIH publication no. 85-23, revised 1996).The experimental protocol was approved by the NationalUniversity of Singapore—Institutional Animal Care and UseCommittee. Male SPF Wistar rats (300–350 g) were used forour experiments. All surgical procedures were performedusing aseptic techniques.

Anesthesia was induced and maintained in animals byinhalational isoflurane (2%) and intraperitoneal injection ofketamine:xylazine (90 mg=kg:10 mg=kg). Carprofen (5 mg=kg,subcutaneous [SC]) was administered preoperatively for an-algesia. A midlaparotomy was performed, followed by dis-placement of the bowel and mild retraction of the kidney.Blunt preparation of the retroperitoneal fossa was done and apouch was created between the retrorenal fat and the psoasmuscle. Subsequently, the MAG was implanted into the retro-renal pouch. Grafts containing cells were implanted in theright renal pouch, whereas acellular controls were implantedin the contralateral pouch. We did not use suture material inthe implantation process. Finally, the bowel was repositionedand the abdomen was closed in two layers. The animals wereallowed to recover in a small-animal ICU. Carprofen (5 mg=kg=day, SC) was administered postoperatively for analgesia.

3D graft preparation for in vivo studies

Eight rats were used to test several 3D matrices and toidentify the one displaying the least degradation and lowinflammatory reaction after 10 days of implantation usingthe renal pouch model. Four different collagenous foam-scaffolds from diverse origin (i.e., equine collagen and hu-man fibrinogen [Tachotop�; Nycomed, Zurich, Switzerland],bovine collagen [Lyostypt�; B. Braun Melsungen AG, Mel-sungen, Germany], native equine collagen [TissuFleece E�;Baxter, Vienna, Austria], and porcine skin gelatin [Gelfoam])were implanted in the renal pouch without cells (data notshown). The porcine gelatin foam was recovered almost in-tact after 10 days of implantation, and hence, it was chosenfor the cell-based animal experiments.

H9C2-GFP-Fluc cells were washed three times in coldphosphate-buffered saline (PBS) and centrifuged. Gelfoamsquares (20�20�5 mm) were placed in 12-well plates and

covered with 4�106 H9C2-GFP-Fluc cells in 250 mL PBSunder sterile conditions. Based on our in vitro studies, wechose 5mmol=L L-AA in PBS enrichment for the AA animalgroup (group A, n¼ 10). Plain MAGs without AA (PBS andcells) were used as control group (group B, n¼ 10). Acellulargrafts containing just PBS (group C, n¼ 10) or 5mmol=L AAin PBS (group D, n¼ 10) were used as negative controls andimplanted in the left renal pouch.

In vivo BLI

We performed optical in vivo BLI with the Xenogen-IVISLumina in vivo imaging system as previously described.15–17

Briefly, rats were anesthetized with 2% isoflurane, andd-luciferin was administered intraperitoneally at a dose of150 mg=kg of body weight. They were placed in the chamberin supine position and peak luciferase activity was detectedby imaging the animals for 40 min with 2-min acquisitionseparated by 2-min intervals.17 The same rats were scannedrepeatedly postimplantation at days 1, 3, and 6. Regions ofinterest indicating the locations of most intense signals werecreated using Living Imaging Software version 2.6. Biolu-minescence was quantified in units of photons per secondtotal flux (p=s).15,18

Immunohistochemistry—Assessment of GFPand rat endothelial cell antibody expression

At 7 days postimplantation, MAGs were explanted andfixed in 10% formalin. The grafts were subsequently optimalcutting temperature (OCT) compound embedded and storedat �808C. Immunohistochemical staining was performed in5-mm cryosections using the primary antibody rat endothelialcell antibody-1 (RECA-1), a ubiquitous marker of rat endo-thelial cells (1:50 monoclonal mouse anti-RECA-1; HyCultBiotechnology b.v., The Netherlands). The sections were thenincubated with secondary antibodies Alexa Fluor�-647 andAnti GFP-Alexa Fluor�-488, and nuclei were stained withDAPI (Molecular Probes).

Images were acquired using a Leica TCS SP5, DMI6000confocal laser scanning microscope (Leica Microsystems),with 40� and 20�magnifications. They were subsequentlyprocessed using Leica Application Suite Advanced Fluores-cence quantification software (version 1.8.2, build 1465).

The mean pixel intensity, a semiquantitative analysis offluorescence, was used to evaluate GFP and RECA expres-sion within the explanted 3D grafts. Predefined settings forlaser power and detector gain were used for all experiments.

Histological analysis

We performed Masson’s trichrome and hematoxylin–eosin staining on 5-mm sections of formalin-fixed and paraffin-embedded explanted MAGs. Histological assessment wasperformed by an experienced pathologist in a blinded fash-ion. To assess angiogenesis within the explanted graft, thenumber of blood vessels per high-power field (hpf ) (400�)was quantified. The degree of cell infiltration and fibrosiswas evaluated as the percentage of area of the scaffold infil-trated by cellular reaction and collagen deposition coveringthe total cellular reaction, using a manual semiquantitativemethod. Ten random fields were chosen in each section for allthe quantifications.

ASCORBIC ACID ENHANCES MYOCARDIAL GRAFT VIABILITY 1351

Statistical analysis

Data are presented as mean� standard deviation. To testfor statistically significant differences, we used two-wayanalysis of variance and the unpaired Student’s t-test whenappropriate. Differences were considered significant if p<0.05. Regression plots were used to describe the relationshipbetween bioluminescence and cell number in vitro; r2-valueswere reported to assess the quality of the linear regressionmodel. All the statistical analyses were performed usingGraphPad Prism� software version 5.01 for Windows(GraphPad Software, San Diego, CA).

Results

In vitro BLI=effect of AA on 3D H9C2cell graft survival in vitro

BLI showed that there was a significant increase in cellbioluminescent baseline signals from day 1 to 5 in graftssupplemented with both 5 and 50 mmol=L AA (55.23%�0.45%, p< 0.001 and 49.83%� 0.45%, p< 0.01, respectively;Fig. 1A–D). H9C2-Fluc cells seeded within the 3D MAGsreceiving 5 mmol=L AA-supplemented medium showed sig-nificantly larger bioluminescent signals after 3 days(2.1�108� 9.5�107 p=s vs. 1.5�108� 7.4�107 p=s, p< 0.05 vs.control) and 5 days in culture (2.3�108� 7.9�107 vs.1.7�108� 4.8�107, p< 0.05 vs. control). Likewise, 50mmol=LAA-supplemented grafts displayed significantly increasedcell signals at day 5 when compared with control grafts re-ceiving plain culture medium (2.0�108� 8.0�107, p< 0.05 vs.control; Fig. 1D). There were no large differences in photonemission between the two dosages of AA.

Histological cell counts (DAPIþ cells) indicated that theamount of DAPIþ cells per hpf (400�) in the 5mmol=L AA-enriched grafts was significantly higher than in the plaingrafts after 3 days in culture (53� 21.4 vs. 30� 7.18, p< 0.05).Likewise, mean cell number was higher in grafts that received50mmol=L AA after 5 days in static culture when comparedwith untreated MAGs (41.4� 14.3 vs. 19.6� 6.9, p< 0.05).There were differences in cell density between the AA-treatedgrafts. A good correlation of luciferase activity per graft tomean cell number (DAPIþ cells=graft) was found after 3 days(r2¼ 0.90) and 5 days (r2¼ 0.97) in culture (Fig. 1E, F).

The effect of AA on cell apoptosis in 3D myocardialbioartificial grafts in vitro

TUNEL assay and active caspase 3 staining were used toassess the effect of AA on H9C2-Fluc cells apoptosis whencultured in thick 3D MAGs (Fig. 2A–F). The percentage ofTUNELþ cells per hpf (400�) in plain grafts increased sig-nificantly from day 3 in culture (23.5%� 12.6%) to day 5(61.9%� 30.1%, p< 0.001). The percentage of TUNELþ cellswas significantly inferior in the 5 and 50mmol=L MAGs atday 5 in culture (4.8%� 8.6%, p< 0.001 and 8.3%� 9.2%,p< 0.001, respectively) when compared with plain controlgrafts (61.9%� 30.1%) (Fig. 2G). Given that TUNEL assaytags apoptotic and necrotic cells, we used active caspase 3staining as a more specific method to identify apoptotic cells.The percentage of active caspase 3þ cells was considerablyhigher in plain grafts after 3 days (41.6%� 11.2%) and 5 days(44.8%� 20.0%) in culture when compared with both dos-ages of AA (Fig. 2A–F, H). The 5mmol=L AA grafts had

16.5%� 6.17% ( p< 0.01) apoptotic cells at day 3 and15.7%� 6.8% ( p< 0.001) at day 5. Likewise, the 50 mmol=LAA MAGs had 11%� 5.7% ( p< 0.01) of apoptotic cells atday 3 and 18.8%� 5.9% ( p< 0.01) at day 5 in culture.

AA effect on H9C2 cells phenotype in vitro

After 3 days in culture, H9C2 cardiomyoblasts were at-tached to the scaffold material and formed a primitive syn-cytium in all groups (Fig. 3A, C, E). Cells fused and formedelongated myotube-like structures that were evident in theAA-treated grafts (Fig. 3C, E). Immunohistochemical assaysrevealed that a-sarcomeric actin was expressed by untreatedand AA-enriched cells after 3 and 5 days in 3D culture.However, the AA-treated groups displayed organized sar-comeric patterns at days 3 and 5 in culture, whereas z-lineswere not apparent in the untreated cells after 5 days inculture (Fig. 3A–F).

In vivo BLI

Bioluminescent signals progressively decreased from day1 to 6 in both the plain and AA-enriched MAGs (groups Aand B; Fig. 4A, B). However, photon emission fell dramati-cally in plain MAGs (group B) from day 1 to 6 post-implantation (Fig. 4A). Cell photon emission by day 6plunged in a 74%� 0.9% of the baseline in the plain MAGs,whereas the AA-enriched graft signals decreased by just36.42%� 1.81% ( p< 0.0001). There was no significant dif-ference in the mean photon emission among groups at days 1and 3 postgraft implantation. However, the AA MAGs(group A) displayed significantly higher cell signals at day 6(6.1�106� 1.7�105 p=s vs. 1.8�106� 2.8�105 p=s, p< 0.05;Fig. 4C).

Renal pouch model

None of the rats died before the scheduled euthanasia.There were no visible signs of infection, inflammation, ad-hesions, or foreign body reaction in any group after 1 weekof MAG implantation in the renal pouch. Explanted graftscontaining cells (groups A and B) displayed preserved sizeand shape. In contrast, the negative control grafts (groups Cand D) were slightly smaller in size after 1 week of im-plantation, which could be due to scaffold’s degradation. Allthe implanted grafts were surrounded by a thin layer ofconnective tissue adhered to the prerenal fat. Numerousblood vessels were visible by the naked eye in both plain(Fig. 5A) and AA-enriched MAGs (Fig. 5C), whereas fewervascular networks could be seen in acellular negative controlgrafts (Fig. 5B, D).

Immunohistochemistry—Assessment of GFPand RECA expression

Fluorescence pixel intensity analysis of the confocal mi-crographs revealed that the amount of GFP-positive cells(donor cells) was significantly greater in group A (AA-enriched MAGs) when compared with the plain grafts fromgroup B (2.09� 1.42 vs. 0.94� 0.24, p< 0.01; Fig. 6E).

Abundant vascular networks infiltrating the MAGs couldbe observed under the confocal microscope (Fig. 6A, B);angiogenesis was particularly abundant in the AA group

1352 MARTINEZ ET AL.

FIG. 1. In vitro BLI of 3D MAGs. BLI of AA-enriched and plain MAGs after (A) 1 day, (B) after 3 days in culture, and (C)after 5 days in culture; n¼ 20. (D) MAGs receiving 50 and 5 mmol=L AA-supplemented medium showed significantly largerbioluminescent signals after 5 days in culture ( p< 0.05 vs. control). 5mmol=L AA-enriched MAGs also displayed higherphoton emission at day 3 in culture ( p< 0.05 vs. control). An increase in cell bioluminescent baseline signals from day 1 to 5was found in grafts supplemented with both 5 mmol=L AA (***p< 0.001) and 50 mmol=L AA (**p< 0.01). Correlation betweenmean bioluminescence (mean total photon flux) and histological mean cell number after (E) 3 days and (F) 5 days in culture.3D, three-dimensional; AA, ascorbic acid; BLI, bioluminescence imaging; hpf, high-power field (400�); MAGs, myocardialartificial grafts; SD, standard deviation.

ASCORBIC ACID ENHANCES MYOCARDIAL GRAFT VIABILITY 1353

FIG. 2. Apoptosis assessment in 3D MAGs. Confocal micrographs of active caspase 3 staining (red) performed after 3 daysin culture in (A) plain MAGs, (B) 5 mmol=L AA MAGs, and (C) 50mmol=L AA MAGS. Active caspase 3 staining after 5 days inculture in (D) plain MAGs, (E) 5 mmol=L AA MAGs, and (F) 50 mmol=L AA MAGS. The inset is a higher magnificationshowing cytoplasmic active caspase 3 expression. (G) In the plain MAGs, TUNEL assay showed an increase in the percentageof TUNEL-positive cells from day 3 to 5 in culture (***p< 0.001). (H) The percentage of apoptotic cells (active caspase 3þ) wassignificantly inferior at days 3 and 5 in the 5 mmol=L AA ( p< 0.01 and p< 0.001, respectively) and 50mmol=L AA MAGs( p< 0.01). Nuclei were counterstained with 40,6-diamidino-2-phenylindole (DAPI) (blue). Scaffold autofluorescence appearsin gray. Arrows pointing at some caspase 3-positive cells. Scale bars indicate 25 mm. TUNEL, dUTP nick-end labeling.

1354

(Fig. 6B, D). Further, blood vessel infiltration observed ingroup A (AA) was not limited to the outmost regions of thegraft, as vessels in close contact with the gelatin scaffoldwere found toward the graft’s core (Fig. 6D). Accordingly,pixel intensity analysis showed that expression of the RECAwas significantly higher in the explanted grafts from groupA (AA) than those from group B (plain) (4.28� 1.98 vs.1.46� 1.21, p< 0.05). Fluorescence intensity quantificationof DAPIþ cells did not reveal any difference among groups(Fig. 6E).

Histology

Histology analysis showed a higher amount of bloodvessels per hpf (400�) in the AA-MAGs (group A) comparedwith plain MAGs (group B) (3.75� 0.5 vs. 2.75� 0.5, p< 0.05;Fig. 7A, B, E). These results are in agreement with the anal-ysis of endothelial cell markers following confocal micros-copy. In both groups, peripheral angiogenesis was higher.However, few capillaries could also be detected at the graft’score. Ingrowth of vessels was also observed in the peripheryof the acellular negative controls (Fig. 7C, D), but no capil-laries were detected toward the core. There was no differencein the number of blood vessels found in the negative controlgroups (Fig. 7E).

There was no significant difference in collagen depositionamong groups. The mean collagen deposition score (% ofarea covered by collagen of the total cellular reaction) wasmild (0–25%) for both AA-enriched and plain MAGs, as wellas for negative controls (Table 1 and Fig. 7A–D).

Cell infiltration was mild in all groups, with eosinophilsbeing the most abundant cell type found within the grafts.Occasional lymphocytes and neutrophils were observed inall the graft sections (Table 1). Likewise, mild foreign bodyreaction was observed in all the groups. Macrophages andgiant cell reaction were observed in all groups. Just one of theexplanted grafts from the negative control (group D) dis-played moderate necrosis, 26–50% collagen deposition, andabundant cell infiltration by neutrophils.

Discussion

The main findings in this study are twofold: first, AAenhances thick bioengineered myocardial tissue survivalin vitro through inhibition of cell apoptosis; second, thegraft’s viability and angiogenic potential in vivo are superiorwith AA enrichment. Moreover, we introduced an in vivomodel for myocardial graft prevascularization, which pro-vides the implanted graft with blood vessels of autologousorigin. We improved graft viability using a biocompatibleand inexpensive compound that could be beneficial for thetissue engineering and the cell therapy field in general. Stemcell- and bioartificial tissue-based therapies for the heart re-main supplementary measures. The unique architecture,terminal differentiation, and limited regenerative potential ofthe heart muscle restrict the capacity of random injections

and implants to restore myocardium efficiently and perma-nently. The progress of such methods from random andsupplementary to first-line therapeutic options will largelydepend upon improvement on various levels: first, therapieshave to be less traumatic, and second, the effect has to bestronger and more durable by enhancing graft viability andfunctionality in vivo. It is critical to sustain donor cell survivalwithin bioengineered myocardial grafts through in vitro es-tablishment of vascular networks for successful cardiac re-pair. Cell death constitutes a major limitation in cell andtissue implantation in the frame of restorative therapies.Some reports indicate that more than 70% of cells die duringthe first 2 days after injection into the myocardium,2 whereasothers point out a cell loss of *55% in just 10 min followingimplantation into the heart.3 Further, there is evidence that,in spite of capillary formation in vivo, donor cell survivalafter myocardial 3D graft implantation is still at risk.19

The addition of AA to the graft constitutes an uncompli-cated method to counteract fulminant events that occurduring acute ischemia and remodeling within the area ofmyocardial lesion. Such events are cytokine and radicaloxygen species liberation, accumulation of purine derivates(danger signals), inflammation, and apoptosis. AA is aubiquitous and essential substance with practically no side-effects even in high doses and can easily be integrated in aregenerative therapy protocol. AA participates in a variety ofphysiological functions in living organisms. In humans, itsdeficiency causes defective healing and disturbed blood vesselformation because of impaired collagen deposition.12,20 Be-sides its role in angiogenesis, AA reduces hypoxia-inducedapoptosis in vitro.10,14 Accordingly, we hypothesized that AAcould be used to reduce cell death in myocardial grafts bothin vitro and in vivo.

We used BLI to assess H9C2-Fluc cell survival within thick,3D MAGs in vitro. The high sensitivity of BLI to monitorviable cells21,22 was effectively demonstrated in our in vitromodel, as we found a robust correlation between luciferaseactivity and histological cell counts. We did not find signif-icant difference in cell photon emission among groups after1 day in culture, suggesting homogeneous delivery of cellswithin the grafts. However, an increase in cell photonemission was observed within 5 days in culture, indicatingthat cell survival and proliferation were enhanced18 when 3Dmyocardial artificial tissues are supplemented with AA. Ourdata suggest that AA enhances cell viability within bioen-gineered tissues in vitro in spite of the harsh conditions inculture (i.e., static 3D culture, hypoxia, and limited growthmedium volume). Apoptosis is an active process associatedwith both the acute and chronic phases of myocardial in-farction in response to oxidative insults.23 The role of hyp-oxia on apoptosis via mitochondrial pathway activation ofdownstream effector caspases has already been demon-strated.24 It has been shown that in hypoxic conditions,hypoxia inhibitor alpha (HIF-1a) promotes apoptosis inH9C2 cells and cardiomyocytes.10,25,26 There is also evidence

FIG. 3. AA effect on H9C2 cardiomyoblasts’ phenotype in 3D culture. a-Sarcomeric actin staining (red) performed after 3days in culture in (A) plain MAGs, (B) 5mmol=L AA MAGs, and (C) 50 mmol=L AA; and after 5 days in culture in (D) plainMAGs, (E) 5mmol=L AA MAGs, and (F) 50 mmol=L AA. Nuclei were counterstained with DAPI (blue). Right-side panelsdepict DIC images that were merged to the fluorescence micrographs to identify cell morphology and distribution in thescaffold. Scale bars indicate 10mm. a-SA, alpha-sarcomeric actin; DIC, differential interference contrast.

‰

ASCORBIC ACID ENHANCES MYOCARDIAL GRAFT VIABILITY 1355

that depletion of AA interrupts HIF-1a proteosomal degra-dation, leading to increased expression of the latter.12 Ourin vitro experiments showed that graft viability improved viareduction of cell apoptosis, particularly when medium issupplemented with both 5 and 50mmol=L AA. Further, theamount of cells within the AA grafts is preserved over timein static culture conditions, whereas it significantly decreasesin the plain MAGS. These data correlate to our in vitro BLI

results. Thus, our findings suggest that AA has a protectiveeffect on donor cells when cultured in thick, 3D bioartificialconstructs. The precise mechanism of such effect was notexplored in this study. However, the enhancement in cellviability observed in the antioxidant-supplemented groupscould be attributed to AA-driven reactive oxygen speciesscavenging, which preserves the integrity of the mitochon-dria and stabilizes the key regulator of hypoxia (HIF-

FIG. 4. In vivo BLI after 1, 3, and 5 days of graft implantation in the renal pouch. (A) Plain MAGs. (B) AA-enriched MAGs.(C) The AA-enriched MAGs displayed significantly higher cell photon emission at day 6 after graft implantation (*p< 0.05).

1356 MARTINEZ ET AL.

FIG. 5. Explanted MAGs. Abundant blood vessels infiltrating the (A) plain and (C) AA-enriched MAGs explanted from theright renal pouch. Fewer blood vessels could be observed by the naked eye in the (B) plain negative (acellular) control and(D) AA negative (acellular) control, explanted from the left renal pouch. Scale bars indicate 10 mm. Color images availableonline at www.liebertonline.com=ten.

FIG. 6. GFP and rat endothelial cell antibody (RECA) expression in explanted MAGs. (A, C) Plain MAGs. (B, D) AA-enriched MAGs. Confocal micrographs show H9C2-Fluc-GFP cells in green; RECAþ cells in red (blood vessels), and nucleistained with DAPI in blue. Vascular networks infiltrating the grafts could be observed in both (A) plain and (B) AA-enrichedMAGs. The amount of GFPþ and RECAþ cells is more evident in the AA group. (C, D) Distribution of donor (GFPþ) cells andblood vessels in the scaffold. GFPþ cells can be seen in white and nuclei in blue. Asterisks indicate the scaffold structure.Arrows point to blood vessels. (E) Semiquantitative fluorescence pixel intensity analysis showed that the number of GFP-positive cells was greater in the AA-enriched MAGs (**p< 0.01). Likewise, expression of RECA was higher in the explantedgrafts from the AA group (*p< 0.05). GFP, green fluorescent protein.

1357

1a).10,12,25 We do not have direct proof that AA was in itsactive form during our study. However, previous stud-ies have demonstrated that the oxidized form of AA—dehydroascorbic acid (DHAA)—is neuroprotective in ische-mia models,27,28 and it also prevents apoptosis and oxidative

damage in macrophages,29 lymphoid, myeloid, and meso-thelial cells,30 among others.31 DHAA is transported by theglucose transporters Glut1 into the mitochondria, but not thereduced form (AA).32 Subsequently, DHHA is reduced backto AA by protein disulfide isomerases. Thus, we assume that

FIG. 7. Masson’s trichrome staining micrographs of (A) plain MAGs, (B) AA-enriched MAGs, (C) plain negative (acellular)control, and (E) AA negative (acellular) control. Mild collagen deposition was observed in all groups. Scaffold material isstained in blue and pointed out with asterisks, to differentiate it from extracellular matrix deposition. Arrows indicate bloodvessels within the grafts. (D) Histological analysis of the explanted MAGs showed fewer amounts of blood vessels per high-power field (400�) in the plain MAGs when compared with the AA-enriched grafts ( p< 0.05). Scale bars indicate 50 mm.

1358 MARTINEZ ET AL.

even if AA was oxidized during our experiments, there wasa protective effect that could have been initiated by DHAA,which is ultimately reduced back to ascorbate in the mito-chondria.

Our preliminary in vitro titration studies suggested thatany dosage equal or above 100 mmol=L AA have a cytotoxiceffect on cardiomyoblasts. In vivo, it has been reported thatAA induces apoptosis of myeloid and lymphoid cells atconcentrations above serum level (50 mmol=L).33 The AAdosages utilized in our in vitro studies have also been pre-viously reported as safe physiological dosage in other celltypes.10,32 Hence, we chose the lowest possible physiologicaldose (5 mmol=L) for our in vivo studies according to this ra-tionale and the observation that after a 10-fold increase ofAA dose there were no differences in cell survival in vitro.

Besides cytoprotection, AA seems to have an effect on celldifferentiation. It has been shown that 100mmol=L AA en-hances differentiation of embryonic stem cell into cardio-myocytes in vitro,34 as well as ex vivo differentiation of adultbone marrow stem cells into cardiomyocyte-like cells.35 Inour study, immunohistochemical assessment of MAGs after3 and 5 days in culture revealed that AA induced morpho-logical myogenic characteristics including elongation andcell fusion.36 Expression of a-sarcomeric actin was observedamong all groups at day 3, yet more organized z-line-likepatterns were observed in the AA-enriched grafts at bothdays 3 and 5 in culture. These results suggest that AA mayhave an effect on H9C2 cardiomyoblast differentiation.However, this interesting observation should be further ad-dressed in future studies.

To evaluate the AA effect on a bioartificial graft in vivo, wedeveloped a simple and reproducible renal pouch model. Toassess donor H9C2-Fluc-GFP cell survival within the graft,we performed noninvasive imaging at days 1, 3, and 6postimplantation with a bioluminescence charge-coupleddevice camera. We observed that 1 week after graft im-plantation in a renal pouch for in vivo prevascularization,H9C2 cardiomyoblast survival decreases substantially. Theeffect was reflected by reduction of cell photon emission bynearly 75% of baseline in group B (plain MAGs). This ob-servation is consistent with previously reported data fromstudies where BLI was used to assess myocardial graft sur-vival in vivo in a heterotopic model of myocardial restorationin immunosuppressed rats.16 The authors have shown thatH9C2 cardiomyoblasts injected in thin (3�3�1 mm) gelfoamgrafts undergo massive cell death within 5 days afterimplantation into the ischemic myocardium. Early cardio-myoblast survival within the implanted grafts could only be

improved by addition of collagen and growth factors16 or bycell transduction with the antiapoptotic hBcl-2 gene.37 Inaddition, premature H9C2 cell death has been also docu-mented after injection in healthy myocardium.38 Here, weobserved that by day 6 postimplantation cell, survival wassignificantly enhanced within the implanted myocardial ar-tificial tissues via AA enrichment (group A). This indicatesdonor H9C2-Fluc-GFP cell survival and proliferation in thehost in the AA-enriched MAGs during the first days post-graft implantation. Yet, confirmation of a proliferative effectof AA on H9C2 cells needs to be further addressed in futurestudies. Our results are of interest for cell transplantationmodels, given the fact that we used an allogeneic model inimmunocompetent rats.

A limiting factor on cell or bioartificial tissue engraftmentand survival is immune or inflammatory reaction. In pastworks we had demonstrated that the density of equine col-lagen scaffolds decreases by 30% after 2 weeks of implan-tation in skeletal muscle and triggers a heavy host cellinfiltration.23 Our previous and current studies also dem-onstrate a bioluminescence signal drop over the first week,indicating decrease in cell survival, regardless the adminis-tration of immunosuppresants (i.e., cyclosporine) or not.12 Itseems that the most sensitive and specific time-window tocapture in vivo photon emission in MAGs—and also the mostaccurate—is the first week postimplantation in the renalpouch. Despite this short time-frame, it was sensitive enoughin our hands to display the difference in bioluminescencesignals in the AA-enriched MAGs. Cell viability after im-plantation has been demonstrated to be compromised in atime-dependent manner, even in models where one or twoimmunosuppresive agents have been used.12,21 Further, inour previous study, donor cells were detected using a celltracker up to day 21 after graft implantation in im-munosupressed rats.23 This demonstrates that, in currentmodels, the fate of grafted tissue is limited by immunologicbarriers even when immunosupression is used.

It is likely that the restorative effect of implanted bioarti-ficial grafts is achieved via neovascularization=paracrineeffects,6,19 which in turn attenuate infarct expansion, ratherthan by cardiomyocyte regeneration, cell differentiation, orfunctional graft integration into the host’s myocardium.Evidence has shown that most bioengineered tissues formyocardial restoration have mainly focused on multicellularallogeneic approaches to achieve neovascularization withinbioengineered grafts.8,39–43 With our study, we introduce astrategy to promote a natural pattern of angiogenic sprout-ing and graft vascularization in the graft recipient itself.

Table 1. Histological Semiquantitative Scoring of Explanted Myocardial Artificial Grafts

Animal groups

ParameterGroup A

AA MAGsGroup B

Plain MAGsGroup C

AA controlGroup D

Plain control

Cellular reaction (%) 0–25 0–25a 0–25 0–25a

No. of eosinophils=hpf 2.3� 0.9 2.0� 1.0 1.5� 0.0 1.5� 0.0Collagen deposition (%) 0–25 0–25 0–25 0–25

Cellular reaction was defined as the percentage of scaffold area covered by cells, and collagen deposition reflects the percentage of collagencovered of the total cellular reaction.

aMinimal cellular reaction.MAGs, myocardial artificial grafts; AA, ascorbic acid; hpf, high-power field.

ASCORBIC ACID ENHANCES MYOCARDIAL GRAFT VIABILITY 1359

A prominent novelty of this study is the introduction ofthe pouch model for in vivo graft prevascularization, whichhas been proven to be an easy and efficient way to supplybioengineered myocardial tissues with blood vessels of au-tologous origin. Using neonatal rat cardiomyocytes cast in anequine collagen type I mesh, we have recently shown thatgraft survival after implantation is determined by host im-mune responses and degree of angiogenesis.44 In such amodel, vascularization could be seen only in the superficiallayer of the graft after 14–21 days of implantation. Interest-ingly, the vascular density was significantly lower when therats received immunosuppressive therapy.23 Here we pres-ent a model with superior and faster angiogenic potential.Moreover, we demonstrate that the neovascularization pro-cess is enhanced in AA-enriched grafts.

In conclusion, the current model is a powerful one tovascularize engineered implants in vivo. The possible expla-nation for the superior AA-enriched graft viability in ourallogeneic model might be both the antiapoptotic andproangiogenic effects exerted by AA. These findings renderAA a considerable supplement for cell- and tissue transplant-based therapies. Yet, further studies should be done toevaluate the effect of AA in 3D bioartificial myocardial graftscontaining human stem cells to confirm the translationalpotential of our approach. We are currently carrying outexperiments aiming at postischemic myocardial restorationusing prevascularized AA-enriched MAGs. Our approachprovides a realistic perspective for the clinical setting: en-hancement of cell survival in 3D bioengineered tissues bothin vitro and in vivo, and a method to mature grafts for laterautologous implantation.

Acknowledgments

The authors thank Ms. Diane Tan Ai Lin for technicalcontributions toward cloning and preparation of the lenti-viral vectors, and Ms. Shera Lilyanna for technical help inmaintaining cell cultures.

Disclosure Statement

No competing financial interests exist.

References

1. Zimmermann, W.H., Didie, M., Doker, S., Melnychenko, I.,Naito, H., Rogge, C., Tiburcy, M., and Eschenhagen, T.Heart muscle engineering: an update on cardiac muscle re-placement therapy. Cardiovasc Res 71, 419, 2006.

2. Muller-Ehmsen, J., Whittaker, P., Kloner, R.A., Dow, J.S.,Sakoda, T., Long, T.I., Laird, P.W., and Kedes, L. Survivaland development of neonatal rat cardiomyocytes trans-planted into adult myocardium. J Mol Cell Cardiol 34, 107,2002.

3. Suzuki, K., Murtuza, B., Beauchamp, J.R., Brand, N.J., Bar-ton, P.J., Varela-Carver, A., Fukushima, S., Coppen, S.R.,Partridge, T.A., and Yacoub, M.H. Role of interleukin-1betain acute inflammation and graft death after cell transplan-tation to the heart. Circulation 110, II219, 2004.

4. Toma, C., Pittenger, M.F., Cahill, K.S., Byrne, B.J., andKessler, P.D. Human mesenchymal stem cells differentiate toa cardiomyocyte phenotype in the adult murine heart. Cir-culation 105, 93, 2002.

5. Robey, T.E., Saiget, M.K., Reinecke, H., and Murry, C.E.Systems approaches to preventing transplanted cell death incardiac repair. J Mol Cell Cardiol 45, 567, 2008.

6. Frangogiannis, N.G. The immune system and cardiac repair.Pharmacol Res 58, 88, 2008.

7. Bursac, N., Papadaki, M., Cohen, R.J., Schoen, F.J., Eisen-berg, S.R., Carrier, R., Vunjak-Novakovic, G., and Freed, L.E.Cardiac muscle tissue engineering: toward an in vitro modelfor electrophysiological studies. Am J Physiol 277, H433,1999.

8. Caspi, O., Lesman, A., Basevitch, Y., Gepstein, A., Arbel, G.,Habib, I.H., Gepstein, L., and Levenberg, S. Tissue engi-neering of vascularized cardiac muscle from human em-bryonic stem cells. Circ Res 100, 263, 2007.

9. Arrigoni, O., and De Tullio, M.C. Ascorbic acid: much morethan just an antioxidant. Biochim Biophys Acta 1569, 1, 2002.

10. Vassilopoulos, A., and Papazafiri, P. Attenuation of oxidativestress in HL-1 cardiomyocytes improves mitochondrial func-tion and stabilizes Hif-1alpha. Free Radic Res 39, 1273, 2005.

11. Fiorito, C., Rienzo, M., Crimi, E., Rossiello, R., Balestrieri,M.L., Casamassimi, A., Muto, F., Grimaldi, V., Giovane, A.,Farzati, B., Mancini, F.P., and Napoli, C. Antioxidants in-crease number of progenitor endothelial cells through mul-tiple gene expression pathways. Free Radic Res 42, 754, 2008.

12. Telang, S., Clem, A.L., Eaton, J.W., and Chesney, J. Deple-tion of ascorbic acid restricts angiogenesis and retards tumorgrowth in a mouse model. Neoplasia 9, 47, 2007.

13. Omeroglu, S., Peker, T., Turkozkan, N., and Omeroglu, H.High-dose vitamin C supplementation accelerates theAchilles tendon healing in healthy rats. Arch Orthop Trau-ma Surg 2, 281, 2008.

14. Vissers, M.C., Gunningham, S.P., Morrison, M.J., Dachs,G.U., and Currie, M.J. Modulation of hypoxia-induciblefactor-1 alpha in cultured primary cells by intracellularascorbate. Free Radic Biol Med 42, 765, 2007.

15. Shinde, R., Perkins, J., and Contag, C.H. Luciferin deriva-tives for enhanced in vitro and in vivo bioluminescence as-says. Biochemistry 45, 11103, 2006.

16. Kutschka, I., Chen, I.Y., Kofidis, T., Arai, T., von Degenfeld,G., Sheikh, A.Y., Hendry, S.L., Pearl, J., Hoyt, G., Sista, R.,Yang, P.C., Blau, H.M., Gambhir, S.S., and Robbins, R.C.Collagen matrices enhance survival of transplanted cardio-myoblasts and contribute to functional improvement of is-chemic rat hearts. Circulation 114, I167, 2006.

17. Hauck, E.S., Zou, S., Scarfo, K., Nantz, M.H., and Hecker,J.G. Whole animal in vivo imaging after transient, nonviralgene delivery to the rat central nervous system. Mol Ther 16,

1857, 2008.18. Cao, F., Lin, S., Xie, X., Ray, P., Patel, M., Zhang, X., Druk-

ker, M., Dylla, S.J., Connolly, A.J., Chen, X., Weissman, I.L.,Gambhir, S.S., and Wu, J.C. In vivo visualization of embry-onic stem cell survival, proliferation, and migration aftercardiac delivery. Circulation 113, 1005, 2006.

19. Leor, J., Aboulafia-Etzion, S., Dar, A., Shapiro, L., Barbash,I.M., Battler, A., Granot, Y., and Cohen, S. Bioengineeredcardiac grafts: a new approach to repair the infarcted myo-cardium? Circulation 102, III56, 2000.

20. Hodges, R.E., Hood, J., Canham, J.E., Sauberlich, H.E., andBaker, E.M. Clinical manifestations of ascorbic acid defi-ciency in man. Am J Clin Nutr 24, 432, 1971.

21. Chen, I.Y., Greve, J.M., Gheysens, O., Willmann, J.K.,Rodriguez-Porcel, M., Chu, P., Sheikh, A.Y., Faranesh, A.Z.,Paulmurugan, R., Yang, P.C., Wu, J.C., and Gambhir, S.S.Comparison of optical bioluminescence reporter gene and

1360 MARTINEZ ET AL.

superparamagnetic iron oxide MR contrast agent as cellmarkers for noninvasive imaging of cardiac cell transplan-tation. Mol Imaging Biol 11, 178, 2009.

22. Jenkins, D.E., Oei, Y., Hornig, Y.S., Yu, S.F., Dusich, J.,Purchio, T., and Contag, P.R. Bioluminescent imaging (BLI)to improve and refine traditional murine models of tumorgrowth and metastasis. Clin Exp Metastasis 20, 733, 2003.

23. Kang, P.M., and Izumo, S. Apoptosis in heart: basic mech-anisms and implications in cardiovascular diseases. TrendsMol Med 9, 177, 2003.

24. McClintock, D.S., Santore, M.T., Lee, V.Y., Brunelle, J., Bu-dinger, G.R., Zong, W.X., Thompson, C.B., Hay, N., andChandel, N.S. Bcl-2 family members and functional electrontransport chain regulate oxygen deprivation-induced celldeath. Mol Cell Biol 22, 94, 2002.

25. Malhotra, R., Tyson, D.W., Rosevear, H.M., and Brosius, F.C.,III. Hypoxia-inducible factor-1alpha is a critical mediator ofhypoxia induced apoptosis in cardiac H9c2 and kidney epi-thelial HK-2 cells. BMC Cardiovasc Disord 8, 9, 2008.

26. Graham, R.M., Frazier, D.P., Thompson, J.W., Haliko, S., Li,H., Wasserlauf, B.J., Spiga, M.G., Bishopric, N.H., andWebster, K.A. A unique pathway of cardiac myocyte deathcaused by hypoxia-acidosis. J Exp Biol 207, 3189, 2004.

27. Huang, J., Agus, D.B., Winfree, C.J., Kiss, S., Mack, W.J.,McTaggart, R.A., Choudhri, T.F., Kim, L.J., Mocco, J., Pin-sky, D.J., Fox, W.D., Israel, R.J., Boyd, T.A., Golde, D.W., andConnolly, E.S., Jr. Dehydroascorbic acid, a blood-brain bar-rier transportable form of vitamin C, mediates potent cere-broprotection in experimental stroke. Proc Natl Acad Sci U SA 98, 11720, 2001.

28. Kim, E.J., Won, R., Sohn, J.H., Chung, M.A., Nam, T.S., Lee,H.J., and Lee, B.H. Anti-oxidant effect of ascorbic and de-hydroascorbic acids in hippocampal slice culture. BiochemBiophys Res Commun 366, 8, 2008.

29. Asmis, R., and Wintergerst, E.S. Dehydroascorbic acid pre-vents apoptosis induced by oxidized low-density lipoproteinin human monocyte-derived macrophages. Eur J Biochem255, 147, 1998.

30. Gogou, E., Hatzoglou, C., Chamos, V., Zarogiannis, S.,Gourgoulianis, K.I., and Molyvdas, P.A. The contribution ofascorbic acid and dehydroascorbic acid to the protective roleof pleura during inflammatory reactions. Med Hypotheses68, 860, 2007.

31. Martino, L., Novelli, M., Masini, M., Chimenti, D., Piaggi, S.,Masiello, P., and De Tata, V. Dehydroascorbate protectionagainst dioxin-induced toxicity in the beta-cell line INS-1E.Toxicol Lett 189, 27, 2009.

32. Kc, S., Carcamo, J.M., and Golde, D.W. Vitamin C entersmitochondria via facilitative glucose transporter 1 (Glut1)and confers mitochondrial protection against oxidative in-jury. FASEB J 19, 1657, 2005.

33. Puskas, F., Gergely, P., Jr., Banki, K., and Perl, A. Stimulationof the pentose phosphate pathway and glutathione levels bydehydroascorbate, the oxidized form of vitamin C. FASEB J14, 1352, 2000.

34. Takahashi, T., Lord, B., Schulze, P.C., Fryer, R.M., Sarang,S.S., Gullans, S.R., and Lee, R.T. Ascorbic acid enhancesdifferentiation of embryonic stem cells into cardiac myo-cytes. Circulation 107, 1912, 2003.

35. Shim, W.S., Jiang, S., Wong, P., Tan, J., Chua, Y.L., Tan, Y.S.,Sin, Y.K., Lim, C.H., Chua, T., Teh, M., Liu, T.C., and Sim, E.

Ex vivo differentiation of human adult bone marrow stemcells into cardiomyocyte-like cells. Biochem Biophys ResCommun 324, 481, 2004.

36. Leong, C.W., Wong, C.H., Lao, S.C., Leong, E.C., Lao, I.F.,Law, P.T., Fung, K.P., Tsang, K.S., Waye, M.M., Tsui, S.K.,Wang, Y.T., and Lee, S.M. Effect of resveratrol on prolifer-ation and differentiation of embryonic cardiomyoblasts.Biochem Biophys Res Commun 360, 173, 2007.

37. Kutschka, I., Kofidis, T., Chen, I.Y., von Degenfeld, G.,Zwierzchoniewska, M., Hoyt, G., Arai, T., Lebl, D.R., Hen-dry, S.L., Sheikh, A.Y., Cooke, D.T., Connolly, A., Blau,H.M., Gambhir, S.S., and Robbins, R.C. Adenoviral humanBCL-2 transgene expression attenuates early donor celldeath after cardiomyoblast transplantation into ischemic rathearts. Circulation 114, I174, 2006.

38. Wu, J.C., Chen, I.Y., Sundaresan, G., Min, J.J., De, A., Qiao,J.H., Fishbein, M.C., and Gambhir, S.S. Molecular imaging ofcardiac cell transplantation in living animals using opticalbioluminescence and positron emission tomography. Cir-culation 108, 1302, 2003.

39. Zimmermann, W.H. Remuscularizing failing hearts withtissue engineered myocardium. Antioxid Redox Signal 8,

2011, 2009.40. Narmoneva, D.A., Vukmirovic, R., Davis, M.E., Kamm,

R.D., and Lee, R.T. Endothelial cells promote cardiac myo-cyte survival and spatial reorganization: implications forcardiac regeneration. Circulation 110, 962, 2004.

41. Sekiya, S., Shimizu, T., Yamato, M., Kikuchi, A., and Okano,T. Bioengineered cardiac cell sheet grafts have intrinsicangiogenic potential. Biochem Biophys Res Commun 341,

573, 2006.42. Tan, M.Y., Zhi, W., Wei, R.Q., Huang, Y.C., Zhou, K.P., Tan,

B., Deng, L., Luo, J.C., Li, X.Q., Xie, H.Q., and Yang, Z.M.Repair of infarcted myocardium using mesenchymal stemcell seeded small intestinal submucosa in rabbits. Bioma-terials 19, 3234, 2009.

43. Sekine, H., Shimizu, T., Hobo, K., Sekiya, S., Yang, J.,Yamato, M., Kurosawa, H., Kobayashi, E., and Okano, T.Endothelial cell coculture within tissue-engineered cardio-myocyte sheets enhances neovascularization and improvescardiac function of ischemic hearts. Circulation 118, S145,2008.

44. Mueller-Stahl, K., Kofidis, T., Akhyari, P., Lee, D.H., Lenz,A., Martinez, E.C., Woitek, F., and Haverich, A. Determi-nants of bioartificial myocardial graft survival and engraft-ment in vivo. J Heart Lung Transplant 27, 1242, 2008.

Address correspondence to:Theo Kofidis, M.D., Ph.D.

Department of SurgeryNational University of Singapore

5 Lower Kent Ridge Road, Level 2Singapore 119074

Singapore

E-mail: surtk@nus.edu.sg

Received: June 12, 2009Accepted: November 11, 2009

Online Publication Date: January 14, 2010

ASCORBIC ACID ENHANCES MYOCARDIAL GRAFT VIABILITY 1361

This article has been cited by:

1. Dumnoensun Pruksakorn, Kriengsak Lirdprapamongkol, Daranee Chokchaichamnankit, Pantipa Subhasitanont, KhajeelakChiablaem, Jisnuson Svasti, Chantragan Srisomsap. 2010. Metabolic alteration of HepG2 in scaffold-based 3-D culture: Proteomicapproach. PROTEOMICS 10:21, 3896-3904. [CrossRef]