Cardiac Gene Delivery With Cardiopulmonary BypassMichael J. Davidson, MD; J. Mark Jones, AFRCS; Sitaram M. Emani, MD; Katrina H. Wilson, MS;

James Jaggers, MD; Walter J. Koch, PhD; Carmelo A. Milano, MD

Background—Cardiac gene therapy offers the possibility of enhancing myocardial performance in the compromised heart.However, current gene delivery techniques have limited myocardial transgene expression and pose the risk ofextracardiac expression. Isolation of the coronary circulation during cardiac surgery may allow for more efficient andcardiac-selective gene delivery in a clinically relevant model.

Methods and Results—Neonatal piglets (3 kg) underwent a median sternotomy and cardiopulmonary bypass, followed byaortic cross-clamping with 30 minutes of cardioplegic arrest. Adenoviral vectors containing transgenes for eitherb-galactosidase (adeno-b-gal, n511) or the humanb2-adrenergic receptor (adeno-b2-AR, n515) were administeredthrough the cardioplegia cannula immediately after arrest and were allowed to dwell in the coronary circulation duringthe cross-clamp period. After 1 week, the animals were killed, and their heart, lungs, and liver were excised andexamined for gene expression. Analysis ofb-galactosidase staining revealed transmural myocardial gene expressionamong animals receiving adeno-b-gal. No marker gene expression was detected in liver or lung tissue.b-AR densityin the left ventricle after adeno-b2-AR delivery was 396685% of levels in control animals (P,0.01). Animals receivingadeno-b2-AR and control animals demonstrated similarb-AR density in both the liver (11468% versus 10069%,P5NS) and lung (11467% versus 10069%,P5NS). There was no evidence of cardiac inflammation.

Conclusions—By using cardiopulmonary bypass and cardioplegic arrest, intracoronary delivery of adenoviral vectorsresulted in efficient myocardial uptake and expression. Undetectable transgene expression in liver or lung tissue suggestscardiac-selective expression.(Circulation. 2001;104:131-133.)

Key Words: gene therapyn cardiopulmonary bypassn signal transduction

Cardiac gene transfer of either the humanb2-adrenergicreceptor (b2-AR) or an inhibitor ofb-adrenergic receptor

kinase (bARKct) enhances cardiac performance.1,2 Use ofsuch genetic strategies clinically will require a safe method ofcardiac gene delivery. The technique that has been used in thelaboratory setting involves intracoronary injection of anadenoviral vector with the heart beating. The principaldisadvantage of this technique is that the viral vector israpidly washed out to the systemic circulation and taken up innontarget organs such as the liver and lung.1,3 A criticalfeature of any clinically relevant cardiac gene deliverytechnique, however, is limiting noncardiac delivery to pre-vent toxicity.

We hypothesized that cardiopulmonary bypass (CPB) mayfacilitate cardiac-selective gene transfer using recombinantreplication-deficient adenovirus. CPB with aortic cross-clamping and cardioplegic arrest represent the fundamentalcomponents of many cardiac surgery procedures anduniquely isolate the coronary circulation. Administration ofan adenoviral vector under these conditions maximizes con-tact time with the myocardium and may reduce systemic

delivery, therefore limiting toxicity and offering a clinicallyrelevant delivery system.

MethodsA replication-deficient, first-generation, type V adenovirus withdeletions of the E1 and E3 genes was used to construct vectors forthe humanb2-AR (adeno-b2-AR) or b-galactosidase (adeno-b-gal)transgene.4 Large-scale preparations of these adenoviruses werepurified from infected Epstein-Barr nuclear antigen–transfected 293cells (Invitrogen Corp).4

One-week-old piglets (3 kg) received humane care in compliancewith the institutional committee on animal research and in accor-dance with the regulations adopted by the National Institutes ofHealth. Animals were given ketamine (20 mg/kg IM) just beforeinhaled isoflurane (1%) anesthesia.5 A median sternotomy wasperformed, and after systemic heparinization, CPB was establishedvia an aortic cannula and a right atrial cannula. The CPB circuitconsisted of a reservoir, a hollow fiber oxygenator/heat exchanger,and a roller pump. After stabilization, the aorta was cross-clampedand the heart arrested by infusion of cold (4°C), hyperkalemiccardioplegia solution (30 mL/kg) into the aortic root. Animals wererandomized to receive either adeno-b2-AR or adeno-b-gal. Immedi-ately after cardioplegic arrest, 131011 total viral particles, reconsti-tuted in 8 mL of phosphate-buffered saline (PBS), were injected into

Received March 30, 2001; revision received May 9, 2001; accepted May 11, 2001.From the Departments of Surgery (M.J.D., J.M.J., S.M.E., J.J., W.J.K., C.A.M.), Medicine, and Biochemistry (K.H.W.), Duke University Medical

Center, Durham, NC.Parts of this work were presented at the October 2000 meeting of the American College of Surgeons, Chicago, Ill.Correspondence to Carmelo A. Milano, MD, Department of Surgery, Box 3043, Duke University Medical Center, Durham, NC 27710. E-mail

david015@mc.duke.edu© 2001 American Heart Association, Inc.

Circulation is available at http://www.circulationaha.org

131

by guest on June 14, 2018http://circ.ahajournals.org/

Dow

nloaded from

the aortic root and allowed to dwell in the myocardium. After 30minutes of cardiac arrest, the cross-clamp was removed and the heartwas reperfused. The animals were then weaned off CPB and allowedto recover.

Gene expression was assessed 1 week after delivery. A subset ofanimals (n58) was studied at 4, 8, and 24 hours and 14 days aftergene delivery to examine the time course of expression. Heart, liver,and lung tissues were either immediately stained with X-gal solution[2 mmol/L K4Fe(CN)6, 2 mmol/L K3Fe(CN)6, 2 mmol/L MgCl2, and0.5 mg/mL 5-bromo-4-chloro-3-indoyl-b-D-galactopyranoside] aswhole-mount samples or frozen at280°C, sectioned at 10mm, andstained in X-gal as previously described.6 b-AR expression wasquantified with radioligand binding assays to determine totalb-ARdensity. Tissue samples were homogenized in lysis buffer (5 mmol/LTris-HCl [pH 7.4] and 5 mmol/L EDTA), and membrane fractionswere extracted. A radioligand binding assay was performed using125I-cyanopindolol to determine totalb-AR density, as previouslydescribed.7

A subgroup of animals received only PBS during CPB (n54).Standard hematoxylin and eosin histological sections of these heartswere made and compared with sections from hearts treated withadeno-b2-AR to assess any inflammatory response.

Data are expressed as mean6SEM and were assessed by Student’st test. Significance was assumed atP,0.05.

ResultsA total of 42 piglets underwent CPB-mediated gene delivery.Of these, 40 survived to the time of study (4 hours to 14days). Twenty-six piglets were studied for myocardial trans-gene expression at 1 week. The piglets that received adeno-b2-AR (n515) demonstrated no background X-gal staining,

whereas those that received adeno-b-gal (n511) had trans-mural staining in all chambers (Figure, A and B). Micro-graphic sections of the myocardium of animals receivingb-galactosidase demonstrated staining of individual myo-cytes, consistent with transgene expression (Figure, C). Therewas nob-galactosidase expression in the liver or lung.

Animals treated with adeno-b2-AR exhibited a left ventric-ular b-AR density '4-fold higher than those receivingmarker transgene (P,0.01; Table). The right ventricularb-AR density was 1.6-fold higher than that of controlanimals, demonstrating lower but significant transgene ex-pression in this chamber (P50.01).b-AR density was notdifferent in the liver and lung between adeno-b2-AR andadeno-b-gal–treated animals (Table).

In addition, gene expression was studied at varying inter-vals from time of delivery (Figure, D).b-Galactosidase

Whole heart mounts and histological sections aftergene delivery. A, Whole mount of heart 1 week afteradeno-b-gal delivery. B, Whole mount of heart 1week after adeno-b2-AR delivery. C, Light micro-graph (magnification 340) of myocardium 1 weekafter adeno-b-gal delivery, demonstrating X-gal stain-ing of individual myocytes. D, X-gal staining in wholemounts of left ventricle at 4, 8, and 24 hours and 7days after b-galactosidase transgene delivery. E,Light micrograph (magnification 340) of myocardiumstained with hematoxylin and eosin 1 week afteradeno-b2-AR delivery. F, Light micrograph (magnifi-cation 340) of myocardium stained with hematoxylinand eosin 1 week after PBS delivery.

b-AR Density in Treated and Control Piglets

Tissue Adeno-b2-AR (n515) Adeno-b-gal (n511)

Left ventricle 396685%* 10067% (94.4 fmol/mg)

Right ventricle 164619%* 10069% (101.0 fmol/mg)

Liver 11467% 10069% (77.4 fmol/mg)

Lung 11468% 100611% (137.4 fmol/mg)

Values are mean6SEM and are expressed as percent of control. All studieswere conducted 1 week after gene delivery.

*P,0.05 vs control.

132 Circulation July 10, 2001

by guest on June 14, 2018http://circ.ahajournals.org/

Dow

nloaded from

expression was first detected 8 hours after gene delivery.Expression at 8 hours was transmural and comparable to thatseen at 24 hours and at 1 week. At 2 weeks, an additional 4animals treated with adeno-b2-AR had increased left ventric-ular (2756126 fmol/mg) and right ventricular (181631fmol/mg) b-AR density.

Hematoxylin and eosin micrographs of hearts 1 week afterdelivery of adeno-b2-AR or PBS (n54) are shown in theFigure (panels E and F, respectively). There was no evidenceof an inflammatory response in either group.

DiscussionThis study demonstrates the feasibility of myocardial genedelivery during CPB and cold hyperkalemic cardioplegicarrest. This protocol simulates conventional cardiac surgeryand tests the effectiveness and potential advantages of genetransfer during cardiac surgery. Unlike prior attempts atintracoronary gene transfer, CPB-mediated gene therapyseems to limit extracardiac gene expression. Our laboratoryhas previously found high levels of gene expression in theliver and lung after non-CPB intracoronary delivery.1 Whendelivered to the beating heart, the intracoronary vector israpidly washed out of the heart and delivered systemically.

Because the coronary circulation is uniquely isolated dur-ing CPB, gene delivery to the myocardium may be improvedrelative to injection into the coronary circulation with theheart beating. By using CPB and cardioplegic arrest, the virusis allowed to dwell in the coronary circulation for 30 minutes.At the end of this time, in contrast to beating-heart delivery,a higher percentage of viral particles may be taken up bymyocytes or be inactivated. Furthermore, any remainingviable virus is ultimately washed out of the coronary circu-lation via the coronary sinus and returned to the CPBapparatus. Because the CPB circuit has a high surface area forpotential virus-binding, particularly at the membrane oxygen-ator, the remaining viable virus may become bound. Indeed,Marshall et al8 demonstrated that the replication-deficientadenoviral vectors commonly used for gene delivery arerapidly inactivated on exposure to nonbiological surfacessuch as polycarbonate, cardiac catheters, and syringes.

This approach may have multiple applications to clinicalcardiac surgery. Such genetic treatments might support end-stage heart failure patients in a manner similar to leftventricular assist devices, as a bridge of support until hearttransplantation. It may also provide support for high-riskpatients with severely reduced ventricular function undergo-ing revascularization or valve replacement procedures. In-deed, impairment of the myocardialb-AR system duringcardiac surgery has been documented, including receptordesensitization with reduced adenylyl cyclase response, pos-sibly due to increasedbARK1 activity.9,10 This method ofgene therapy would achieve transgene expression during the

first postoperative day and continue for'2 to 3 weeks. Thistime course would correlate with the early postoperativeperiod during which inotropic support is most important.These studies also raise interest in the possibility of genetherapy with retrograde cardioplegia or with percutaneousmethods of CPB, such as Heartport.

This study represents the first use of CPB for globalmyocardial gene delivery. Moreover, it demonstrates thefeasibility of intracoronary gene delivery in the pig, whoseheart is similar to humans. The study is limited insofar as thesubjects were healthy neonatal piglets. Further work isneeded to characterize the effectiveness of this technique inadult animals and those with ventricular dysfunction. Inaddition, current efforts are directed at demonstrating thebiochemical and hemodynamic consequences of gene deliv-ery using functional transgenes.

AcknowledgmentsThis work was supported in part by National Institute of Healthgrants HL61690 (to W.J.K.) and HL56205 (to W.J.K.) and byNational Research Service Award 5 F32 HL10179 (to M.J.D.). Theauthors thank George Quick, Ronnie Johnson, and Kurt Campbell fortheir invaluable assistance in the animal setup and use of CPB. Wealso thank Robert J. Lefkowitz, who was instrumental in initiatinggene therapy efforts withb2-AR and who provided much of theadenoviral vector for these studies.

References1. Maurice J, Hata J, Shah A, et al. Enhancement of cardiac function after

adenoviral-mediated in vivo intracoronaryb-2-adrenergic receptor genedelivery.J Clin Invest. 1999;104:21–29.

2. White D, Hata J, Shah A, et al. Preservation of myocardialb-adrenergicreceptor signaling delays the development of heart failure after myo-cardial infarction.Proc Natl Acad Sci U S A. 2000;97:5428–5433.

3. Hajjar R, Schmid U, Matsui T, et al. Modulation of ventricular functionthrough gene transfer in vivo.Proc Natl Acad Sci U S A. 1998;95:5251–5256.

4. Akhter S, Skaer C, Kypson A, et al. Restoration ofb-adrenergic signalingin failing cardiac ventricular myocytes via adenoviral-mediated genetransfer.Proc Natl Acad Sci U S A. 1997;94:12100–12105.

5. Lodge A, Chai P, Daggett C, et al. Methylprednisolone reduces theinflammatory response to cardiopulmonary bypass in neonatal piglets:timing of dose is important.J Thorac Cardiovasc Surg. 1999;117:515–522.

6. Kypson A, Peppel K, Akhter S, et al. Ex-vivo adenoviral-mediated genetransfer to the transplanted adult rat heart.J Thorac Cardiovasc Surg.1998;115:623–630.

7. Koch W, Rockman H, Samama P, et al. Cardiac function in mice over-expressing the Beta-adrenergic receptor kinase or a BARK inhibitor.Science. 1995;268:1350–1353.

8. Marshall D, Palasis M, Lepore J, et al. Biocompatibility of cardiovasculargene delivery catheters with adenovirus vectors: an important determinantof the efficiency of cardiovascular gene transfer.Mol Ther. 2000;1:423–429.

9. Schwinn D, Leone B, Spahn D, et al. Desensitization of myocardialb-adrenergic receptors during cardiopulmonary bypass: evidence forearly uncoupling and late downregulation.Circulation. 1991;84:2559–2567.

10. Sun L, Pantuck C, Morelli J, et al. Perioperative lymphocyte adenylylcyclase function in pediatric cardiac surgical patients.Crit Care Med.1996;24:1654–1659.

Davidson et al Cardiac Gene Delivery With Cardiopulmonary Bypass 133

by guest on June 14, 2018http://circ.ahajournals.org/

Dow

nloaded from

and Carmelo A. MilanoMichael J. Davidson, J.Mark Jones, Sitaram M. Emani, Katrina H. Wilson, James Jaggers, Walter J. Koch

Cardiac Gene Delivery With Cardiopulmonary Bypass

Print ISSN: 0009-7322. Online ISSN: 1524-4539 Copyright © 2001 American Heart Association, Inc. All rights reserved.

is published by the American Heart Association, 7272 Greenville Avenue, Dallas, TX 75231Circulation doi: 10.1161/01.CIR.104.2.131

2001;104:131-133Circulation. 

http://circ.ahajournals.org/content/104/2/131Wide Web at:

The online version of this article, along with updated information and services, is located on the World

  http://circ.ahajournals.org//subscriptions/

is online at: Circulation Information about subscribing to Subscriptions: 

http://www.lww.com/reprints Information about reprints can be found online at: Reprints:

  document. and Rights Question and Answer

Permissionsthe middle column of the Web page under Services. Further information about this process is available in thethe online version of the published article for which permission is being requested is located, click Request Permissions in

can be obtained via RightsLink, a service of the Copyright Clearance Center, not the Editorial Office. OnceCirculation Requests for permissions to reproduce figures, tables, or portions of articles originally published inPermissions:

by guest on June 14, 2018http://circ.ahajournals.org/

Dow

nloaded from