ICIAH

B

M

R

C

GfiiHvtw

re

FPiMS

N

((Ra

CELL TRANSPLANTATION

n Vivo Optical Bioluminescence Imaging of Collagen-supportedardiac Cell Grafts

ngo Kutschka, MD,a Ian Y. Chen, MSE,b Theo Kofidis, MD,a Georges von Degenfeld, MD, PhD,c

hmad Y. Sheikh, MD,a Stephen L. Hendry, MD,a Grant Hoyt, RA,a Jeremy Pearl, MSE,a

elen M. Blau, PhD,c Sanjiv S. Gambhir, MD, PhD,b and Robert C. Robbins, MDa

ackground: Histology-based survival assessment of cell grafts does not allow for in vivo follow-up. In this studywe introduce two new experimental models for longitudinal in vivo survival studies of cardiac cellgrafts using optical bioluminescence imaging.

ethods: H9c2 cardiomyoblasts expressing both firefly luciferase (fluc) and green fluorescent protein (GFP)reporter genes were implanted into Lewis rats. In Model 1, H9c2-fluc-IRES-GFP cells (0.5 � 106)were implanted into a cryoinjured abdominal wall muscle. Cells were injected using either liquidcollagen (Matrigel [MG]) or phosphate-buffered saline (PBS) suspension. Cell survival was evaluatedin vivo using bioluminescence imaging on days 1, 5 and 10 post-operatively. In model 2, ratsunderwent ligation of the left anterior descending (LAD) artery. The donor hearts were harvested,and the infarcted region was restored ex situ using 1 � 106 H9c2-fluc-IRES-GFP cells seeded incollagen matrix (Gelfoam [GF]) or suspended in PBS (n � 8/group). Hearts were then transplantedinto the abdomen of syngeneic recipients. Optical bioluminescence imaging was performed on Days1, 5, 8 and 14 post-operatively. After 4 weeks, immunohistologic studies were performed.

esults: For model 1, at day 5, bioluminescence signals were markedly higher for the H9c2/MG group(449 � 129 photons/second � 103) compared with the H9c2/PBS group (137 � 82 photons/second � 103) (p � 0.05). For model 2, bioluminescence signals were significantly (p � 0.04)higher in the H9c2/GF group compared with plain cell injection on days 5 (534 � 115 vs 219 � 34)and 8 (274 � 34 vs 180 � 23). Data were in accordance with GFP immunohistology.

onclusions: Optical bioluminescence is a powerful method for assessment of cardiac cell graft survival in vivo.Collagen matrices support early survival of cardiomyoblasts after transplantation into injuredmusculature. J Heart Lung Transplant 2007;26:273–80. Copyright © 2007 by the International

Society for Heart and Lung Transplantation.

tlpFitsspt

SwiA

aS�tCT

rowing evidence from experimental studies1–3 andrst clinical trials3– 6 suggest that cell therapy can

mprove heart function after myocardial infarction.owever, detailed assessment of cell fate and survival inivo remains elusive and several studies have shownhat certain cell types undergo extensive cell deathithin the first few days after transplantation.7–9

Cell labeling using reporter genes, such as green fluo-escent protein (GFP), is still the “gold standard” fornabling histologic cell identification.1,2,8 To obtain de-

rom the aDepartment of Cardiothoracic Surgery; bMolecular Imagingrogram/Bio-X Program, Departments of Radiology and Bioengineer-

ng; and cBaxter Laboratory in Genetic Pharmacology, Department oficrobiology & Immunology, Stanford University School of Medicine,

tanford, California.Submitted May 21, 2006; revised September 30, 2006; accepted

ovember 21, 2006.Supported in part by the Falk Cardiovascular Research Fund

R.C.R.), NHLBI 5R01HL078632 (S.S.G.), NCI ICMIC P50 CA114747S.S.G.) and NCI SAIRP. I.K. was supported by the annual Bayeresearch Grant of the German Society of Cardiology (2004) and the

nnual Thoracic Research Scholarship (2004) from the German j

ailed information about graft survival post-mortem, histo-ogic evaluation is required at several established time-oints,7 resulting in excessive numbers of study animals.urthermore, histologic evaluations may be influenced bynterindividual variances and cannot provide any informa-ion about the longitudinal cell survival course within aingle animal. To improve the reliability of cell transplanttudies, we need more in vivo control over the transplantrocedure in terms of standardized cell grafting. Thus,here is a strong demand for new standardized surgical

ociety for Thoracic and Cardiovascular Surgery. T.K. and G.v.D.ere supported by a German Research Society grant. HMB’s research

s supported by NIH grants AG009521, EB005011, HD018179,G020961, AG024987 and the Baxter Foundation.Reprint requests: Ingo Kutschka, MD, Department of Thoracic

nd Cardiovascular Surgery, Städtisches Klinikum Braunschweig,alzdahlumerstrasse 90, Braunschweig 38126, Germany. Telephone:49-53-595-2213. Fax: �49-531-595-2658. E-mail: ingo.kutschka@

-online.deopyright © 2007 by the International Society for Heart and Lungransplantation. 1053-2498/07/$–see front matter. doi:10.1016/

.healun.2006.11.604

273

mwc

oblraScapcthav

tsecwctilrlfs

MA

AvUH

CH

Htmtap(sfirscw

A

F(mslltfuIM

FausslbpHi

274 Kutschka et al. The Journal of Heart and Lung TransplantationMarch 2007

odels10 and innovative non-invasive imaging techniques,hich would allow for detailed in vivo assessment of stem

ell engraftment and survival.7,11

A promising new method for imaging a small numberf transplanted cells in experimental models is opticalioluminescence imaging. It utilizes an internal biologic

ight source, such as the luciferin–luciferase enzymaticeaction, which can be detected within tissues of smallnimals using sensitive low-light imaging systems.11,12

pecific targeting of luciferase transgene expression inells and tissues of interest has allowed the localizationnd tracking of cell fate for studying a variety of diseaserocesses, including inflammatory conditions and can-er.13,14 Recently, optical bioluminescence gained at-ention as a powerful tool for stem cell experiments andas been used successfully in mice and rats to providesensitive in vivo serial evaluation of cell number,

iability and location.11,13

We aimed to use this technique to test our hypothesishat the supportive use of collagen may favor graft cellurvival by providing a 3-dimensional scaffold for thembedded cells.15,16 Thus, we developed a model forell grafting into the rat abdominal wall musculature,hich allows intraindividual comparisons of cell types,

ell delivery techniques and host conditions. To assesshe survival of cardiomyoblasts after transplantationnto the ischemic heart, we used a combined model ofeft anterior descending artery (LAD) ligation and hete-otopic heart transplantation in rats. This model al-owed for controlled large-scale restoration of the in-arcted heart and close non-invasive monitoring of cellurvival using optical bioluminescence imaging.

ETHODSnimal Care

ll surgical interventions and animal care were pro-ided in accordance with the Guide for the Care andse of Laboratory Animals (National Institutes ofealth, Vol. 25, No. 28, revised 1996).

reation of an Optically Bioluminescent and GFP-positive9c2 Cell Line

9c2 cardiomyoblasts (BDIX rats, American Type Cul-ure Collection, Rockville, MD) were cultured andaintained as described elsewhere.17 The cells were

ransduced with 1 � 107 plaque-forming units (pfu) oflentiviral vector carrying a cytomegalovirus (CMV)

romoter driving both a firefly luciferase reporter genefluc) and a GFP gene, separated by an internal ribo-omal entry site (IRES) element. The genetically modi-ed H9c2 cells (H9c2-fluc-IRES-GFP) underwent twoounds of FACS sorting, after which cells were exten-ively diluted for single-clone selection. The brightestlone, as confirmed by in vitro firefly luciferase assay,

as used for both cell culture and animal experiments. i

bdominal Wall Cell Transplantation

ive male Lewis rats were anesthetized with isoflurane2%) and ketamine (50 mg/kg). The rats underwentid-line abdominal incision and lateral mobilization of

kin flaps. Two injection sites (8-mm diameter) of theeft abdominal musculature underwent cryoinjury usingiquid nitrogen for 60 seconds. As controls, two con-ralateral sites on the animals’ right side were selectedor cell grafting without cryoinjury (Figure 1A). Thepper sites were injected with 0.5 � 106 H9c2-fluc-RES-GFP cells suspended in 50 �l of Matrigel (H9c2/G; BD Biosciences, San Jose, CA). The lower sites

igure 1. Cell transplantation models. (A) Cell grafting into thebdominal wall. The injection sites on the right side (anatomic left)nderwent cryoinjury, whereas the two contralateral sites on the leftide (anatomic right) were used as a control without injury. The upperites were injected with H9c2 cells suspended in Matrigel, whereas theower sites were injected with H9c2 cells suspended in phosphate-uffered saline (PBS). (B) Myocardial restoration using a heart trans-lant model. Ex situ injection of H9c2 cells (left) or implantation of9c2/GF matrix (right) into the ischemic myocardium. The GF scaffold

s implanted into an intramyocardial pouch. LAD, left anterior descend-

ng artery; LV, left ventricle; GF, Gelfoam.

wsi1

HR

DtiipdaSwra

vw3KaosG

pp(pf

m

FrlToo a

The Journal of Heart and Lung Transplantation Kutschka et al. 275Volume 26, Number 3

ere injected with 0.5 � 106 H9c2-fluc-IRES-GFP cellsuspended in 50 �l of PBS. Cell survival was evaluatedn vivo using bioluminescence imaging on Days 1, 5 and0 after surgery (Figure 2).

eterotopic Heart Transplantation, Myocardialestoration and Cell Grafting

onor animals (male Lewis rats, 260 g) were anesthe-ized as described previously, and ventilated after oralntubation. A median sternotomy was performed. Tonduce a sizable left ventricular myocardial infarct, theroximal LAD was ligated. Five minutes after paleemarcation of the infarct, the aorta was cross-clampednd 5 ml of antegrade cardioplegic solution (modifiedtanford solution) was administered. The donor heartas explanted and placed on ice for ex situ myocardial

estoration. An end-to-end anastomosis of the left atrial

igure 2. Optical bioluminescence imaging of an abdominal wall/cryoinats on days 1, 5 and 10 after injection of H9c2 cells into the abdominaleft), whereas the healthy muscle is on the left side (anatomic right). Che average optical bioluminescence imaging signal (n � 5) associarientation. � and #: statistical difference (p � 0.05) between controlr H9c2/PBS (#). *Significant difference (p � 0.05) between H9c2/MG

ppendix to the pulmonary artery was performed. t

Group A (n � 8). The anterior portion of the leftentricle was dissected and a 5 � 5-mm semi-lunar pouchas created within the infarcted area (Figure 1B, right). A� 3 � 1 mm moist piece of sterile Gelfoam (GF; Upjohn,alamazoo, MI) was implanted into the prepared ischemicrea. The hearts were then transplanted into the abdomenf syngenic rats as previously described.10 Before reperfu-ion, 50 �l of cell suspension (1 � 106 H9c2-fluc-IRES-FP) was injected into the GF scaffolds.Group B (n � 8). Myocardial infarction and trans-

lantation of the heart was performed as describedreviously. Before reperfusion, 50 �l of cell suspension1 � 106 H9c2-fluc-IRES-GFP cells suspended in phos-hate-buffered saline [PBS]) was injected into the in-

arcted area (Figure 1B, left).Control animals received PBS injection (n � 3) or GF-atrix implantation without cardiomyoblasts (n � 3) into

y model. In vivo optical bioluminescence imaging of two representativell. The cryoinjured abdominal wall muscle is on the right side (anatomic

were suspended in either PBS (lower sites) or Matrigel (upper sites).with each site is tabulated at the bottom of the figure in the samescle and cryo-injured muscle at day 1 after injection of H9c2/MG (�)

nd H9c2/PBS at day 5 after injection into cryoinjured muscle.

jurwaellstedmu

he infarcted area using the surgical techniques described.

si(sfm

Io

RgatAta(ama(se

Oa

Arp

ociAuGtadsSA(bwieas

01

2

Fhro*

276 Kutschka et al. The Journal of Heart and Lung TransplantationMarch 2007

Before wound closure, an intraperitoneal (IP) portystem with a sub-cutaneous (SC) injection site wasmplanted to ensure accurate IP injection of D-luciferinBiosynth USA, Switzerland) during bioluminescenceessions. To avoid cell rejection (allogeneic H9c2 cellsrom BDIX rats) all animals received cyclosporine (7.5g/kg/day) orally.

n Vivo Optical Bioluminescence Imagingf Cardiomyoblast Transplantation

ats were placed in an imaging chamber (IVIS 50; Xeno-en, Hopkinton, MA) under continuous isoflurane (2%)nesthesia. The animals were rotated to the left to exposehe abdominal heterotopic heart optimally to the camera.fter acquisition of a baseline image, rats were injected IP

hrough the aforementioned port system with D-luciferint 400 mg/kg body weight. Rats were imaged on Days 1baseline), 5, 8 and 14 after surgery using 5-minutecquisition scans (Figure 3). For the abdominal wallodel the animals were placed in the supine position

nd were imaged on Days 1, 5 and 10 after surgeryFigure 2). Peak signal in terms of total flux (photons/econd) from a fixed region of interest (ROI) wasvaluated using LIVINGIMAGE v2.50 software (Xenogen).

rgan Harvest, Tissue Storage, Histologynd Immunofluorescence Microscopy

t 4 weeks, hearts were perfused with saline andapidly explanted. They were fixed for 2 hours inaraformaldehyde and cryoprotected in 30% sucrose

igure 3. Optical bioluminescence imaging of a heterotopic heart traeterotopically transplanted rat hearts with left ventricular restoration uow). Shown are optical images of two representative animals at days 1nto a gray-scale photograph of the animals. The average imaging sign

and **: statistical difference (p � 0.05) between H9c2/GF and H9c2/P

vernight. The tissue was frozen in optimum cuttingompound (OCT) and stored at �70°C. Immunostain-ng was performed on 6-�m sections using polyclonallexa Fluor 488–conjugated anti-GFP antibody (Molec-lar Probes, Eugene, OR) For characterization of theFP-positive H9c2 cells, we performed co-staining of

he cells using rabbit anti–�-sarcomeric actin or rabbitnti–connexin-43 primary antibodies (both Sigma-Al-rich, St. Louis, MO). Alexa 546– or 647–conjugatedecondary antibodies were used (Molecular Probes).ections were counterstained with Hoechst (Sigma-ldrich) and analyzed using a fluorescent microscopeAxioplan; Zeiss, Jena, Germany). The border zoneetween infarct, graft and healthy host myocardiumas evaluated (Figure 4). Ten randomized regions of

nterest (ROIs; 40� magnification) were considered atach border zone for scoring. Four level cuts of eachnimal were evaluated using the following scoringystem (Figure 4):

—No cells present.—Single GFP-positive H9c2 cells are present in the

infarct border zone of the myocardium. Cells appearround, with no interconnection with the host myo-cardium. No cells are present in the center of theinfarct.

—Groups of GFP-positive H9c2 cells are present inthe infarct border zone. Some cells are spindle-shaped. Some cells are present in the center of theinfarct.

plant/LAD ligation model. In vivo optical bioluminescence imaging ofg either H9c2/GF grafts (upper row) or plain H9c2/PBS injection (lower8 and 14 post-surgery. The optical signals (pseudo-color) are overlaidpooled from 8 rats per group are tabulated at the bottom of the figure.

nssin, 5,als

BS at day 5 (*) and day 8 (**) after surgery.

3

fae

S

Amtn

RO

Swdahaom

da

cohgr(

MbsciTgwps

I

Uc

F(iGf oin

The Journal of Heart and Lung Transplantation Kutschka et al. 277Volume 26, Number 3

—Ingrowths of GFP-positive H9c2 cells in the borderzone, with most cells appearing spindle-shaped andinterconnecting with the host myocardium. Somecells are present in the center of the infarct.

The scores of all levels were averaged and calculatedor each animal. For evaluation of cell morphology,dditional sections were stained with hematoxylin andosin (H&E) and Masson’s trichrome.

tatistics

ll data were displayed as mean � standard error ofean. For comparison of groups a 2-tailed Student’s

-test was performed. Differences were considered sig-ificant at p � 0.05.

ESULTSptical Bioluminescence Imaging

urvival of H9c2 cardiomyoblasts in an abdominalall/cryoinjury model. Different survival kinetics

ata were observed, depending on the presence orbsence of cryoinjury. Specifically, after injection intoealthy control muscles, both types of grafts (H9c2/MGnd H9c2/PBS) led to significantly increased cell signalsn day 1 compared with implantation into cryoinjured

igure 4. Histologic scoring of cell engraftment after myocardial restoH9c2/GF) implantation (right column). The drawing in the middle shnfarcted (dark gray) and healthy (light gray) myocardium, and the infarFP-positive H9c2 cells are displayed in green. Representative sample

or each level of the scoring system. The scoring definitions (0 to 3 p

uscles (p � 0.05). However, the imaging signals c

ecreased drastically to background levels on days 5nd 10 (Figure 2).

In contrast, after implantation into cryoinjured mus-les, cell signals were detectable for both types of graftsn Day 5, with the Matrigel-supported H9c2 graftsaving a markedly greater signal than the plain H9c2rafts (p � 0.05). By day 10, the imaging signalseached background levels for both types of graftsFigure 2).

yocardial Restoration Using H9c2 Cardiomyo-lasts. Baseline signal at Day 1 after cell grafting did notignificantly differ between the GF/H9c2 group and theell injection group. Between days 1 and 8, the signalntensity dropped markedly, indicating rapid cell death.here was only background signal at Day 14 in bothroups. However, on Days 5 and 8, the signal intensityas significantly higher in the H9c2/GF group com-ared with the cell injection group (Figure 3). Controlshowed only background signal.

mmunohistochemistry and Histology

sing an Alexa Fluor 488 anti–GFP antibody, H9c2 cellsould be clearly identified under fluorescence micros-

ion with H9c2 cardiomyoblast injection (left column) or H9c2/Gelfoams where the GF matrices (white mesh) are implanted relative to theorder zone (black frame), where the left and right images were taken.f H9c2 cell injections (left) and H9c2/GF cell grafts (right) are shownts) are described in the text.

ratowct bs o

opy in both the H9c2 and H9c2/GF groups at 4 weeks

aMpTfgstmbztscdeSm�htsltnt

D

AocaotmqHihc

sictbvspnccsbc

u(omrs

twrisseTcaca

adcffwcpsphtgmcFvkt

sbsiti

dcgatuo

278 Kutschka et al. The Journal of Heart and Lung TransplantationMarch 2007

fter cell grafting (Figure 4). Histologic scoring (seeethods) revealed an increased cell number and im-roved engraftment for the GF-supported H9c2 grafts.he mean score for the H9c2/GF grafts (2.0 � 0.3) was

ound to be significantly higher than that of plain H9c2rafts (1.3 � 0.2, p � 0.05). H&E and trichrometaining further revealed notable vessel ingrowth intohe GF scaffolds (Figure 5A and B). Two different cellorphologies of GFP-positive cells were identified for

oth types of grafts. Cells outside the infarct borderone or inside the tissue graft retained a round shape. Inhe peri-infarct region, the GFP-positive cells assumed apindle-shaped form, and they aligned with the hostardiomyocytes (Figure 4). There were no obviousifferences in terms of cell shape or muscle cell markerxpression between the H9c2/GF and H9c2 grafts.ome of the H9c2 cells co-stained for GFP and �-sarco-eric actin (Figure 5C–F). The expression pattern of-sarcomeric actin in these spindle-shaped cells isighly suggestive of an engrafted muscle cell pheno-ype. GFP-positive H9c2 cells did not stain for thepecific cardiac cell marker connexin-43, indicating aack of GAP-junction formation. Only single GFP-posi-ive H9c2 cells were found in the cryoinjured abdomi-al muscle, and their morphology did not differ fromhat of cells implanted into the ischemic heart.

ISCUSSION

s a logical extension of previously conducted “proof-f-principle” studies, which used optical biolumines-ence for detecting cardiac cell grafts,7,18 this study hasssessed the effect of extracellular matrix graft supportn the longitudinal cell survival kinetics after transplan-ation. Compared with histology, in vivo optical biolu-inescence imaging can provide a reproducible, semi-

uantitative in vivo analysis of viable, transplanted9c2-cells. The superficial position of the target organs

n this study further provides a special opportunity forighly sensitive, non-invasive imaging of transplantedells.As shown in the abdominal wall model, Matrigel-

upported cell grafting is superior to plain cell injectionn that it leads to greater viability in cryoinjured mus-les. This increased performance can be attributed tohe special properties of the Matrigel basement mem-rane matrix, which is known to release low doses ofarious growth factors19,20 and to provide structuralupport for cell seeding and differentiation.21 Unex-ectedly, cryoinjury of the abdominal wall muscle didot further promote death of transplanted cells whenompared with transplantation into healthy host mus-le. To the contrary, cryoinjury resulted in better cellurvival 5 days post-operatively. This phenomenon maye due in part to the early inflammatory milieu of the

ryoinjured muscle that could be accompanied by an t

pregulation of both vascular endothelial growth factorVEGF) and fibroblast growth factor (FGF), both previ-usly shown to increase capillary density in ischemicuscles.22,23 Together, these findings may suggest a

ole for therapeutic angiogenesis in improving theurvival advantage of the transplanted cells.

As described elsewhere,7,24 we also observed a dras-ic reduction of signal intensity of the cardiac cell graftsithin the first 5 days after transplantation, reflecting

apid cell death, most likely caused by a combination ofschemia, apoptosis and inflammation. In the currenttudy, GF-matrix support of cardiomyoblast grafts led toignificantly better early cell survival within the isch-mic myocardium compared with plain cell injections.his may be attributed to the special properties of solidollagen matrices for tissue engineering, which includelarge surface area for cell seeding, optimal porosity forapillary ingrowth, minimal antigenicity and biodegrad-bility.24,25

Wu and colleagues found that labeling cells withdenoviruses carrying firefly luciferase could lead to aecline in imaging signal after cell implantation. Besidesell death, the signal reduction could also be due to theact that the reporter gene is only transiently expressedollowing adenoviral transduction.7 Furthermore, it isell known that inflammatory cytokines can silence

ertain promoters, such as the cytomegalovirus (CMV)romoter, leading to a decline in reporter gene expres-ion.26 In this study, we have partially circumvented theseroblems by labeling cells with a lentiviral vector, whichas the capacity to integrate into the cell genome and leado stable expression of the firefly luciferase reporterene. Using this cell labeling method, the optical biolu-inescence imaging signal should be more reflective of

ell number than when adenoviral vectors are used.urthermore, the less immunogenic nature of lentiviralectors would likely lead to fewer inflammatory cyto-ines, which can act to silence the CMV promoter and,hus, firefly luciferase expression.

It is important to note that GFP-positive cells weretill found with fluorescent microscopy when opticalioluminescence imaging revealed only backgroundignal. This must be kept in mind as a limitation of then vivo imaging technology and identifies immunohis-ochemistry as a more sensitive methodology for detect-ng very low cell numbers.

As expected from the optical bioluminescence imagingata, only a limited number of H9c2 cells survived asonglomerates, mainly within the margins of the collagenraft and the infarct border zone. This phenomenon waslso described by other investigators1 using collagen ma-rices for cell implantation. Although those investigatorssed computer-based morphometry to determine regionsf cell engraftment, we preferred to create a semi-quanti-

ative score to assess cell survival and engraftment. Al-

F(taca

The Journal of Heart and Lung Transplantation Kutschka et al. 279Volume 26, Number 3

igure 5. Histology and immunohistochemistry. (A, B) Vessel ingrowth (arrows) into the GF matrix: (A) H&E stain, original magnification 40�;B) trichrome stain, original magnification 20�. (C–F) Immunostaining of H9c2 cardiomyoblasts 4 weeks after Gelfoam-supported cellransplantation into ischemic myocardium. A polyclonal Alexa Fluor 488–conjugated anti-GFP antibody was used. Expression of �-sarcomericctin was detected using a specific monoclonal anti–�-sarcomeric actin antibody and an Alexa 546–conjugated secondary antibody foro-staining. Sections were counterstained with Hoechst. The merged picture (F) identifies the GFP expressing H9c2-cell-expressing �-sarcomeric

ctin.

ttsiaict

adZdsmoieega

tchsntmcteid

Tf

R

1

1

1

1

1

1

1

1

1

1

2

2

2

2

2

2

2

280 Kutschka et al. The Journal of Heart and Lung TransplantationMarch 2007

hough histologic scoring might be criticized with regardo accuracy, it provides a good impression of the wholeample and considers the shape of the cells and thenteraction between host and tissue graft. We believe thatttempts to quantify data by histology do not necessarilyncrease the reliability of the results, because overall cellounts were calculated from representative slides, bearinghe risk of missing target cells.

The more prudent role of histochemistry is to char-cterize target cells with specific surface markers. Weetected expression of �-sarcomeric actin, which stainslines in myotubes of skeletal and cardiac muscle. We

id not find co-staining for connexin-43, which ispecific for GAP junctions as an indicator for electro-echanical coupling between the cardiac cells. Based

n these histologic findings, some differentiation intommature heart muscle cells can be inferred, but differ-ntiation into functional adult cardiomyocytes is notxpected. Ongoing studies will include different colla-en components, specific growth factor applicationsnd genetic cell manipulation using this model.

In summary, in vivo optical imaging is a reliableechnique to follow the early survival kinetics of cardiacell grafts. The superficial position of the heart in thiseterotopic transplant model contributes to superiorignal intensity and provides reliable cell detection. Thison-invasive imaging technique, in conjunction withhe established surgical heterotopic heart transplantodel, will allow investigators to further study cardiac

ell graft survival with modulation from a variety ofherapeutic agents. Collagen-based matrices improvearly H9c2-cardiomyoblast survival after cell transfernto injured skeletal muscle and into infarcted myocar-ium.

he authors thank V. Mariano for animal care and Pauline Chuor tissue embedding, processing and staining.

EFERENCES

1. Kofidis T, de Bruin JL, Hoyt G, et al. Myocardial restoration withembryonic stem cell bioartifical tissue transplantation. J HeartLung Transplant 2005;24:737–44.

2. Orlic D, Kajstura J, Chimenti S, et al. Bone marrow cells regen-erate infarcted myocardium. Nature 2001;410:701–5.

3. Agbulut O, Vandervelde S, Al Attar N, et al. Comparison of humanskeletal myoblasts and bone marrow-derived CD133� progenitorsfor the repair of infarcted myocardium. J Am Coll Cardiol2004;44:458–63.

4. Perin EC, Dohmann HF, Borojevic R, et al. Transendocardial,autologous bone marrow cell transplantation for severe, chronicischemic heart failure. Circulation 2003;107:2294–302.

5. Stamm C, Westphal B, Kleine HD, et al. Autologous bone-marrowstem-cell transplantation for myocardial regeneration. Lancet2003;361:45–6.

6. Wollert KC, Meyer GP, Lotz J, et al. Intracoronary autologousbone-marrow cell transfer after myocardial infarction: the BOOST

randomised controlled clinical trial. Lancet 2004;364:141–8.

7. Wu JC, Chen IY, Wang Y, et al. Molecular imaging of cardiaccell transplantation in living animals using optical biolumines-cence and positron emission tomography. Circulation 2003;108:1302–5.

8. Balsam LB, Wagers AJ, Christensen JL, Kofidis T, Weissman IL,Robbins RC. Haematopoietic stem cells adopt mature haema-topoietic fates in ischaemic myocardium. Nature 2004;428:668 –73.

9. Reinecke H, Zhang M, Bartosek T, Murry CE. Survival, integration,and differentiation of cardiomyocyte grafts: a study in normal andinjured rat hearts. Circulation 1999;100:193–202.

0. Krupnick AS, Kreisel D, Engels FH, et al. A novel small animalmodel of left ventricular tissue engineering. J Heart Lung Trans-plant 2002;21:233–43.

1. Rice B, Cable M, Nelson M. In vivo imaging of light-emittingprobes. J Biomed Opt 2001;6:432–40.

2. Contag C, Ross B. It’s not just about anatomy: in vivo biolumi-nescence imaging as an eyepiece into biology. J Magn Reson Imag2002;16:378–87.

3. Contag C, Jenkins D, Contag PR, Negrin RS. Use of reporter genesfor optical measurements of neoplastic disease in vivo. Neoplasia2000;2:41–52.

4. Wetterwald A, van der Pluijm G, Que I. Optical imaging of cancermetastasis to bone marrow: a mouse model of minimal residualdisease. Am J Pathol 2002;160:1143–53.

5. Kofidis T, Akhyari, Wachsmann B, et al. Clinically establishedhemostatic scaffold (tissue fleece) as biomatrix in tissue- andorgan-engineering research. Tissue Eng 2003;9:517–23.

6. Zimmermann WH, Melnychenko I, Eschenhagen T. Engineeredheart tissue for regeneration of diseased hearts. Biomaterials2004;25:1639–47.

7. Bengel FM, Anton M, Avril N, et al. Uptake of radiolabeled2=-fluoro-2=-deoxy-5-iodo-1-beta-D-arabinofuranosyluracil in car-diac cells after adenoviral transfer of the herpesvirus thymidinekinase gene: the cellular basis for cardiac gene imaging. Circula-tion 2000;102:948–50.

8. Koransky ML, Ip TK, Wu S, et al. In vivo monitoring of myoblasttransplantation into rat myocardium. J Heart Lung Transplant2001;20:188–9.

9. Philp D, Chen SS, Fitzgerald W, Orenstein J, Margolis L, KleinmanHK. Complex extracellular matrices promote tissue-specific stemcell differentiation. Stem Cells 2005;23:288–96.

0. Kleinman HK, McGarvey ML, Hassell JR, et al. Basement mem-brane complexes with biological activity. Biochemistry 1986;25:312–8.

1. Kofidis T, de Bruin JL, Hoyt G, et al. Injectable bioartificialmyocardial tissue for large-scale intramural cell transfer andfunctional recovery of injured heart muscle. J Thorac CardiovascSurg 2004;128:571–8.

2. Choksy S, Pockley AG, Wajeh YE, et al. VEGF and VEGF receptorexpression in human chronic critical limb ischaemia. Eur J VascEndovasc Surg 2004;28:660–9.

3. Luo F, Wariaro D, Lundberg G, Blegen H, Wahlberg E. Vasculargrowth factor expression in a rat model of severe limb ischemia.J Surg Res 2002;108:258–67.

4. Gonen-Wadmany M, Gepstein L, Seliktar D. Controlling thecellular organization of tissue-engineered cardiac constructs. AnnNY Acad Sci 2004;1015:299–311.

5. Boccafoschi F, Habermehl J, Vesentini S, Mantovani D. Biologicalperformances of collagen-based scaffolds for vascular tissue en-gineering. Biomaterials 2005;26:7410–7.

6. Qin L, Ding Y, Pahud DR, Chang E, Imperiale MJ, Bromberg JS.Promoter attenuation in gene therapy: interferon-gamma andtumor necrosis factor-alpha inhibit transgene expression. Hum

Gene Ther 1997;8:2019–29.