In vivo high-efficiency transcoronary gene delivery and Cre ... - Nature

Gene Therapy (2003) 10, 1814–1820 & 2003 Nature Publishing Group All rights reserved 0969-7128/03 $25.00 www.nature.com/gt

RESEARCH ARTICLE

In vivo high-efficiency transcoronary gene delivery and Cre–LoxP gene switching in the adult mouse heart M Iwatate1,2,, Y Gu1,2, T Dieterle1,2, Y Iwanaga1,2, KL Peterson2, M Hoshijima1,2, KR Chien1,2 and J Ross Jr.1,2, 1

Institute of Molecular Medicine, University of California San Diego, La Jolla, CA, USA and 2Department of Medicine, University of California San Diego, La Jolla, CA, USA

High-efficiency somatic gene transfer in adult mouse heart has not yet been achieved in vivo. Here, we demonstrate high-efficiency in vivo transcoronary gene delivery to the adult murine myocardium using a catheter-based technique with recombinant adenovirus (AdV) and adeno-associated virus (AAV) vectors in normal and genetically engineered mice. The method involves immersion hypothermia followed by transient aortic and pulmonary artery occlusion with proximal intra-aortic segmental injection of cardioplegic solution containing substance P and viral vectors. Gene expression measured using a LacZ marker gene was observed throughout both ventricles. The expression efficiency of a cytoplasmic LacZ marker gene in the left ventricular myocardium was 56.4714.5% (mean7s.d.) at 4 days with an AdV vector, and with an AAV vector it was

81.075.9% at 4 weeks. Following AAV gene transfer, no gene expression was found in kidney, brain, lung, and spleen, but there was slight expression in liver. In addition, we demonstrate temporally controlled genetic manipulation in the heart with an efficiency of 54.675.2%, by transferring an AdV vector carrying Cre recombinase in ROSA26 floxLacZ reporter mice. Procedure-related mortality was 16% for AdV and zero for AAV transfer. Thus, this method provides efficient, relatively homogeneous gene expression in both ventricles of the adult mouse heart, and offers a novel approach for conditional gene rescue or ablation in genetically engineered mouse models. Gene Therapy (2003) 10, 1814–1820. doi:10.1038/ sj.gt.3302077

Keywords: adenovirus; adeno-associated virus; cre recombinase; gene transfer; myocardium; mice

Introduction Many genetically engineered murine models of heart failure have been generated,1–3 and crossbreeding experiments have provided insight into potential forms of therapy for heart failure.4 In addition, ongoing largescale ENU mutagenesis projects in mice will provide expanded mutant mouse resources for research on the pathogenesis of many ultimately fatal diseases.5 However, elaborate genetic modification, such as conditional mutagenesis or genetic complementation, is significantly hampered by cost and time requirements. Moreover, changes in genetic background often cause phenotypic drift in such animal models.6 Therefore, study of the cardiac effects of specific gene products would be greatly enhanced if an efficient in vivo cardiac gene delivery system were available for the mouse. In past studies, intravenous injection, intracavitary injection, and intramyocardial injection using virus vectors resulted in insufficient gene expression in left ventricular (LV) myocardium in mice in vivo to test the effects of gene products on global LV function.7-11 Although there have Correspondence: Dr J Ross Jr, Department of Medicine, University of California San Diego, 9500 Gilman Drive, La Jolla, CA 92093-0613, USA The first two authors contributed equally to this work Received 14 November 2002; accepted 10 April 2003

been reports of adenovirus (AdV) vector delivery via the aorta in rats12 and intracoronary AdV vector delivery in rabbits,13 an efficient transcoronary cardiac gene delivery system has not been established in vivo in adult mice. Conditional in vivo gene expression and gene deletion are important experimental approaches for examining the functions of specific gene products in development and disease. The Cre–loxP system in transgenic animals has been used to achieve conditional gene activation14 or deletion (knockout)15,16 of targeted genes in myocardium. This application was also demonstrated in the adult mouse with direct intramyocardial injection of AdV encoding Cre recombinase;11 however, in that report, Cre-mediated recombination was only in focal areas of the LV myocardium. Achievement of highefficiency homogeneous induction of Cre recombination events in the mouse heart could be useful for a number of applications to achieve conditional gene activation or deletion. One purpose of the current study was to establish a highly efficient in vivo gene transfer method with transcoronary injection using AdV and adeno-associated virus (AAV) vectors in the mouse heart associated with relatively cardiac-specific gene expression. The second goal was to evaluate the potential application of this gene delivery system for conditional cardiac gene deletion using an AdV vector carrying Cre recombinase.

High-efficiency gene transfer in mouse heart M Iwatate et al

Results Transcoronary delivery with AdVLacZ In a study with an AdV vector carrying the cytoplasmic LacZ gene (AdVLacZ) in normal mice (n¼7) there was one procedure-related death after operation. Gene expression at 4 days after transfer was observed throughout the LV, with somewhat lower expression in the right ventricular (RV) side of interventricular septum, and also in the RV myocardium (Figure 1a–c). Calculated efficiency of expression in LV cardiomyocytes was

56.4714.5% (mean7s.d., range 37.9–71.0%, n¼5). In some mice, expression in the interventricular septum was reduced. Signs of slight inflammation were noted in all AdV-transduced hearts, evidenced by infiltration of a few mononuclear cells in interstitial areas (hematoxylin and eosin staining, data not shown); one animal showed marked inflammation and was excluded from the study (LacZ gene expression in this animal was 5.3%). In all mice, there was a high level of pericentral vein expression in liver (Figure 2d) (consistent with our previous observations in the hamster), which is con-

a

d

g

b

e

h

c

f

i

1815

Figure 1 Mid-LV and apical slices of representative hearts showing b-galactosidase staining in mice transfected with cytoplasmic AdVLacZ (a–c) or nuclear targeted AdVLacZ (NLSAdVLacZ) (d–f), and cytoplasmic AAVLacZ (g–i). Expression is relatively homogeneous in the LV, although somewhat less staining was usually present in the RV side of the interventricular septum, particularly in (g). High-power ( 400) portions of fields from the same hearts (rectangular area) are also shown (c,f,i). Bar in (c), (f), and (i)¼50 mm.

a

b

c

d

e

f

g

h

i

j

k

l

m

n

o

Figure 2 b-Galactosidase staining of various organs from C57BL/6 mice after the in vivo cardiac delivery of cytoplasmic AdVLacZ, NLSAdVLacZ, or cytoplasmic AAV LacZ. The brain (a, f, k), kidney (b, g, l), spleen (c, h, m), liver (d, i, n), and lung (e, j, o) were harvested from C57BL/6 mice at 4 days after gene transfer with cytoplasmic AdVLacZ (upper row, a–e), after gene transfer with NLSAdVLacZ (middle row, f–j) at 4 days, and after gene transfer with cytoplasmic AAVLacZ at 4 weeks (lower row, k–o). Frozen sections were stained for b-galactosidase activity and counterstained with neutral red. Extensive blue staining was seen only in the liver transfected with AdVLacZ (d, i), while only a few positively stained cells were noted in the liver with cytoplasmic AAVLacZ (n). Gene Therapy


1816

sidered to result from retrograde hepatic perfusion of virus vector associated with increased venous pressure during pulmonary artery clamping. However, there was no expression in brain (Figure 2a), and rare cells showed expression in kidney (Figure 2b), spleen (Figure 2c), and lung (Figure 2e). To assure the adequacy of marker gene expression efficiency, which may be affected by diffusion of the cytoplasmic form of b-galactosidase and X-gal cleavage product, an adenovirus vector carrying a LacZ gene fused with a nuclear-localizing signal (NLSAdVLacZ) was also delivered in C57BL/6 mice (9–10 weeks old, n¼7). There were two procedure-related deaths. Gene expression rate, quantified by point counting of myocytes with nuclei positively stained for b-galactosidase activity in LV myocardium, was 47.277.6% (mean7s.d., range 39.0–55.5%, n¼4, Figure 1d–f). One mouse not included in the average showed sustained ST elevation on the ECG after virus injection; in this animal bgalactosidase staining showed a sizeable area of very low expression in the posterior LV myocardium, suggesting a coronary embolic event. Gene expression was lower in the RV than the LV myocardium averaging 23.2719.2% (n¼4). There was some gene expression in the right and left atria. Again, a high level of gene expression occurred in the liver (Figure 2i). However, there was no expression in brain (Figure 2f) and kidney (Figure 2g), and only occasional cells showed expression in spleen (Figure 2h) and lung (Figure 2j).

Transcoronary delivery with AAVLacZ Using an AAV vector carrying the cytoplasmic LacZ gene (AAVLacZ) in normal mice, b-galactosidase staining showed an efficiency of 81.075.9% (mean7s.d., range 74.0–88.8%, n¼6) gene expression in the LV myocardium at 4 weeks after gene transfer (Figure 1g–i), and in this study the survival rate was 100% at 4 weeks after operation. Again, gene expression efficiency tended to be lower in the interventricular septum in most animals (Figure 1g). There was no gene expression in brain (Figure 2k), kidney (Figure 2l), spleen (Figure 2m) or lung (Figure 2o), but there were a few b-galactosidasepositive cells in the liver (Figure 2n). There was mild inflammatory cell infiltration in the heart in one animal, but none in the others. LV function after gene delivery with AAVLacZ vector Transthoracic echocardiography was performed in agematched C57BL/6 mice in a normal control group not subjected to gene transfer (n¼8) and in mice at 4 weeks after gene transfer with AAVLacZ (n¼6). LV end-diastolic internal diameter was 3.7470.13 mm (mean7s.d.) in the control group and 3.8570.17 mm in the post gene transfer (GT) group. LV % fractional shortening was 42.571.9% (mean7s.d.) in the control group and 41.971.3% in the post-GT group. No significant differences were identified between the control group and the post-GT group (Table 1). Gene deletion with AdV/Cre delivery in ROSA26 mice ROSA26 flox-LacZ reporter mice17 were subjected to in vivo delivery of an AdV vector carrying the Cre recombinase gene (AdV/Cre). All animals (n¼5, 7–14 weeks old) survived the gene transfer procedure, making the total mortality rate for all AdV gene transfer 16% Gene Therapy

(n¼19). Organs were harvested and b-galactosidase staining performed 4 days after the AdV/Cre transfer (Figure 3). The Cre-dependent recombination event, determined by b-galactosidase-positive staining, was observed in 54.675.2% (mean7s.d., range 49.9–62.5%,

Table 1 Echocardiographic findings in C57BL/6 mice at 4 weeks after gene transfer with AAVLacZ

HR (bpm) LVDed (mm) LVDes (mm) PWTh (mm) IVS (mm) LV %FS (%) Mean Vcf (circ/s)

Control

Post-GT

P-values

415717 3.7470.13 2.1570.08 0.7570.03 0.7570.03 42.571.9 6.4970.63

427733 3.8570.17 2.2170.09 0.7370.02 0.7370.02 41.971.3 6.5670.39

NS NS NS NS NS NS NS

Control indicates age-matched C57BL/6 mice (n¼8, average age 15 weeks). Post-GT indicates C57BL/6 mice at 4 weeks after gene transfer with AAVLacZ (n¼6, average age 15 weeks). Values are mean7s.d. Unpaired t-tests were performed between the two groups. HR¼heart rate, LVDed¼LV end-diastolic internal diameter, LVDes¼LV end-systolic internal diameter, PWTh¼posterior wall thickness, FS¼fractional shortening, Vcf¼velocity of circumferencial fiber shortening, NS¼not significant.

a

c

d

e b f

g

Figure 3 b-Galactosidase staining of organs from ROSA26 reporter mice at 4 days after AdV/Cre gene transfer. ROSA26 flox-LacZ reporter mice (n¼5, 7–14 weeks old) were subjected to in vivo AdV/Cre delivery. Midlevel of the heart (a), apex level of the heart (b), brain (c), kidney (d), spleen (e), liver (f), and lung (g) were harvested from ROSA26 indicator mice at 4 days after gene transfer with AdV/Cre. Frozen sections were stained with b-galactosidase staining and were counterstained with neutral red. The heart (a, b) and the liver (f) showed homogeneous blue staining, which indicated efficient Cre-mediated recombination. Note that staining is relatively weak compared to the hearts receiving AdVLacZ or AAVLacZ vectors. The latter mice use an exogenous CMV promoter in viral vectors for gene expression, whereas there is weaker promoter activity for the activated LacZ gene at the integrated locus of ROSA26.


n¼5) of LV myocardial cells (Figure 3a and b), and in the majority of hepatocytes (Figure 3f). No LacZ activity was detected in the brain (Figure 3c), kidney (Figure 3d), spleen (Figure 3e), and lung (Figure 3g). There was slight cell infiltration in the heart indicating minimal immune inflammation in response to the AdV/ Cre vector.

Discussion In past studies, intravenous injection of AdV vector in neonatal mice resulted in gene expression in 0.2% of cardiac cells,9 while intracavitary injection of AdV vector in neonatal mice achieved a gene expression efficiency of 11% in ventricular cardiomyocytes.7 Intramyocardial injections of AdV vector11 or AAV vector,8,10 into the left ventricle of adult mice have produced limited spatial extent of transduction and some evidence of damage in cells surrounding the needle track. In an isolated mouse heart subsequently transplanted to a syngenic host, 15 min of in vitro recirculation of an AAV vector achieved 40% efficiency after 4 weeks.10 Although there have been reports of AdV-mediated cardiac gene transfer via the aorta in rats12 and with intracoronary AdV vector transfer in rabbits,13 efficient cardiac gene transfer has largely failed in the mouse in vivo, because of technical limitations related to the animal size (the body weight: 20–40 g), the organ size (heart weight: 100–200 mg), a very narrow range between toxic and effective doses of vasoactive reagents, and high sensitivity to fluid overload. With our recent success in accomplishing highefficiency in vivo gene transfer in the cardiomyopathic hamster heart using AdV18 and AAV vectors,19 as well as our extensive use of microsurgical techniques and miniaturized methods for assessing cardiac function in the mouse,20–22 we undertook to determine whether a modified approach would be feasible for application in the mouse. In the present study, using microsurgical methods we could safely occlude the aorta using hypothermia to protect vital organs and achieve cardiac gene transfer during transient cardiac arrest coupled with increased capillary permeability. Without aortic occlusion or without use of a vascular permeabilizing agent, efficiency was low (data not shown). The method ultimately devised for transcoronary gene delivery achieved efficient short-term gene expression with AdV vectors, and using an AAV vector high efficiency with relatively cardiac-restricted gene expression was feasible at 4 weeks after delivery. It should be noted that other strategies for achieving tissue-specific in vivo expression of functional genes in adult mice have had limitations. Gene expression studies in transgenic mice using the aMHC promoter have been used frequently; however, without temporal control, complications of interpretation are introduced by pre-existing phenotypes produced by early gene expression driven by the aMHC promoter at embryonic or early postnatal stages.23,24 Nonspecific effects using the aMHC promoter have also been reported.25 More sophisticated temporal manipulation of cardiac gene expression has been attempted, including the tetracycline-controlled transactivator system26 and drug-sensitive Cre systems,27,28 although the tightness of regulation of these complicated genetic

manipulations has been questioned; as noted, mosaic gene activation and preintervention leakage are not unusual events. In addition, these elaborate genetic engineering strategies are costly and time-consuming. Accordingly, our in vivo gene delivery system provides a novel competent strategy to test functional genes, or genes with specific mutations, in normal or existing wellcharacterized murine models of cardiac disease, such cardiomyopathic mice with the muscle-specific LIM protein deficiency.21 Conditional in vivo gene expression and gene deletion are important experimental approaches for examining the functions of specific gene products in development and disease. The Cre–loxP system in transgenic animals has been used to achieve conditional gene activation14 or deletion (knockout)15,16 of targeted genes in myocardium. Generally, this system requires cross-breeding of two lines of transgenic animals, one carrying an allele with the locus of interest flanked by two loxP sequences and the other carrying a Cre transgene in which the expression of the Cre protein is controlled by a heterologous tissue-specific promoter,11 or in some cases Cre is knocked-in to the coding14,15 or 30 noncoding region of tissue-specific genes with an internal ribosome entry site.29 Recombination between the loxP elements in the offspring is dependent upon the expression pattern of the Cre transgene, defined by the specificity of promoters or other locus-specific transcriptional regulation.30 Cre-mediated recombination can also be regulated by controlling the timing of Cre activation,31,32 and conditional, drug-induced post-translational activation of Cre recombination can bypass potential embryonic and prenatal effects on the heart.26,27 This application was also demonstrated in the mouse with direct intramyocardial injection of AdV encoding Cre recombinase;11 however, as noted, Cre-mediated recombination was only in focal areas of the left ventricle. In the current study, an AdV/Cre vector was employed in ROSA26 reporter mice to further substantiate the value of this gene delivery system. The ROSA26 flox-LacZ reporter mouse were originally generated by P Soriano17 and harbors an inactivated LacZ gene at the ROSA26 locus with a floxed stopper sequence. This mouse is widely used to detect in vivo activity of Cre recombinase, which enzymatically deletes the stopper sequence to activate the downstream LacZ gene. Using the ROSA26 reporter mice with transcoronary AdV/Cre delivery, we were able to achieve high efficiency with nearly homogenous induction of genetic recombination events in the heart, indicating that this type of technology could be useful for achieving conditional gene activation or deletion in other floxed transgenic mice, which are widely available in the research community. In conclusion, our approach offers efficient in vivo gene expression throughout the left ventricle of the mouse heart, with relatively cardiac-specific and sustained gene expression using the rAAV vector. Lower expression also occurs in the right ventricle and atria. This approach will allow examination of the effects of gene products on global heart function in the adult mouse, and it offers a new approach for conditional gene rescue and ablation in genetically engineered mouse models.

1817

Gene Therapy


1818

Materials and methods Background studies to optimize optimal gene delivery in the mouse Our initial efforts to apply hypothermia (external body cooling) with brief aortic and pulmonary artery occlusion accompanied by transient cardiac arrest and aortic root injection of AdV vectors in the presence of histamine, as used effectively in the hamster,18,19 were not successful and carried extremely high early and late mortality rates. Accordingly, we assessed a series of vascular permeabilizing agents other than histamine and found that substance P (SP)33 was well tolerated by the mouse. Gene expression rate with 2 ng SP was 41.179.8% (mean7s.d., n¼4), with 5 ng SP it was 56.4714.5% (n¼5), and with 100 ng SP it was 28.272.5% (n¼2); the latter dose and an even higher dose prolonged recovery time; so a total dose of 5 ng was used in all subsequent experiments. The volume and composition of the cardioplegic (CP) solutions proved important to obtain a sufficient period of cardiac arrest for efficient gene expression. CP solution with 20 mM KCl (CP20) did not stop the heartbeat for more than 30 s, and CP solution with 25 mM KCl (CP25) was sufficient to cause 45 s or more of cardiac arrest. A virus delivery volume of 75 ml was also found to be optimal (50 ml volume of CP solution was too little to sustain arrest and 100 ml was not tolerated by C57BL/6 mice (mean body weight 23.972.1 g)). The sequence of occluding the pulmonary artery (PA) first followed immediately by the ascending aorta (Ao) proved to have an advantage in homogeneous distribution of gene expression in the LV myocardium. The sequence of the Ao occlusion followed by the PA occulusion, used in the hamster,18,19 decreased gene expression in LV endocardium, presumably due to higher intracavitary LV pressure. Finally, our standard use of intraperitoneal anesthesia ketamine (100 mg/kg)/ xylazine (5 mg/kg) was found to be responsible for delayed recovery and increased late mortality (30%) in the mouse (n¼20). A single initial low dose of intraperitoneal anesthetic followed by maintenance inhalation anesthesia with isoflurane and O2 greatly hastened recovery and improved survival. In vivo delivery method in mice C57BL/6 mice (n¼28, 9–17 weeks old from Charles River Laboratory, Wilmington, MA, USA) were included in this study. Of these, eight mice served as normal controls for echocardiographic studies. Following a single low dose of ketamine (50 mg/kg) and xylazine (2.5 mg/kg), the mice are intubated and ventilated with a pressurecontrolled respirator (RSP1002, Kent Scientific Corporation, CT, USA) at a respiratory rate of 100/min, 35% inspiratory time, and 15 cm of water inspiratory cutoff pressure; they were then maintained on 0.5–2.0% isoflurane in O2. Through a small left anterior first intercostal space chest incision, snares are placed around the Ao and PA using 6-0 silk surgical suture. Following cannulation of the right carotid artery with a flamestretched PE-60 polyethylene tube (Intramedics), the catheter tip is positioned above the aortic valve for blood pressure measurement and subsequent injections into the aortic root, and crushed ice is applied around the body to bring the rectal temperature to 19.0–21.51C Gene Therapy

(Figure 4a). The PA is then occluded followed quickly by the Ao (Figure 4a). Three injections are then made into the proximal aortic root: (1) CP solution (NaCl 110, KCl 10, CaCl2 1.2, MgCl2 16, and NaHCO3 10, in mM) and 2 ng SP (volume injected: 100 ml); (2) 3 s later CP solution

Figure 4 In vivo transcoronary gene transfer procedure (a) and measurement method for gene expression rate in LV myocardium with point-counting method (b, c). (a) The details of the procedure are described in ‘Materials and methods’. Briefly, following immersion whole-body hypothermia, the PA and Ao are sequential occluded, and three injections are then made into the proximal aortic root to achieve arrest of the heart with CP and to increase vascular permeability with SP followed by injection of the viral vector in CP and SP solution. The Ao and PA snares are then released 45 s after the final injection, dobutamine is infused through the aortic catheter, and animals are warmed to normal body temperature. (b and c) Images illustrating method of determining expression efficiency by point-counting cells positively stained for bgalactosidase activity (white open circle) on the myocardium (b) or cells with positive nuclei (c) in which only intersections on the myocardium were counted using a grid. Negative cells for b-galactosidase activity either in cytoplasm or in their nuclei were marked with red open circle. The percentage of positive cells from eight images were averaged. These photos were acquired from a heart transfected with cytoplasmic AdVLacZ (b) or NLSAdVLacZ (c).


with 25 mM KCl and 2 ng SP to achieve arrest of the heart (100 ml); (3) 45 s later an AdV vector with CP solution with 10 mM KCl and 1 ng SP (75 ml) or an AAV vector with CP solution with 20 mM KCl and 1 ng SP (75 ml). The Ao and PA snares are then released 45 s after the final injection. Total aortic occlusion time averaged 95 s (Figure 4a). Dobutamine is then infused through the aortic catheter (20 mg/kg/min), with gentle chest compression if necessary; the mouse is then placed on a warming pad and, when systolic arterial pressure reaches 65–70 mmHg, extubation is usually feasible within 15 min and the animal is allowed to recover. All protocols were approved by the University of California, San Diego Animal Subjects Committee.

In vivo delivery of AdV/Cre vector in ROSA26 indicator mice ROSA26 flox-LacZ reporter mice were obtained from Jackson Laboratories (Bar Harbor, ME, USA) (n¼5, 7–14 weeks old) and subjected to in vivo AdV/ Cre delivery. Preparation and dose of virus vectors Replication-deficient AdVLacZ vectors containing either a cytoplasmic LacZ gene or a modified LacZ gene fused to a nuclear-localizing signal (NLSAdVLacZ) were used. Also a Cre recombinase AdV vector was prepared. The titer of each vector preparation was determined as described previously.34 An AAVLacZ vector containing a cytoplasmic b-galactosidase transgene was prepared35,36 and supplied by Drs Lili Wang and James M Wilson (University of Pennsylvania, Philadelphia, PA, USA). The infectious units (IU) of AAVLacZ were determined as previously described.36 In the mice, the following constructs were delivered: cytoplasmic AdVLacZ with 9.2 1010 virus particles (4.9 1012 virus particles/ml, 3.8 1010 PFU/ml), or NLSAdVLacZ with 9.2 1010 virus particles (1.6 1012 virus particles/ml, 1.2 1010 PFU/ml) or AdV/Cre vector with 1.7 1010 virus particles (2.3 1012 virus particles/ml, 2.5 1011 PFU/ml) or AAVLacZ with 9.2 1010 genome copies (2.5 1012 genome copies/ml, 1.7x1010 IU/ml). b-Galactosidase staining for determining expression efficiency The heart and other organs were harvested at 4 days after AdVLacZ, NLSAdVLacZ and AdV/Cre gene transfer and at 4 weeks after AAVLacZ gene transfer. The heart was sliced transversely into three pieces, mounted in OCT compound and frozen. Cross-sections from the mid-LV and apical slices were used for bgalactosidase staining. For b-galactosidase staining, sections from hearts transfected with AdVLacZ, NLS AdVLacZ or AAVLacZ were fixed with 0.4% glutaraldehyde for 5 min at room temperature and incubated in X-Gal solution for 3 h at 371C; sections from hearts transfected with AdV/Cre were incubated overnight. Sections were counterstained with neutral red and fixed with 4% paraformaldehyde. A total of eight randomly selected areas around the LV in both the apical (n¼4) and mid level (n¼4) LV slices were selected for imaging and photomicroscopy at 200 magnification (437 mm 582 mm), and two areas were selected at the mid-RV level. Gene expression was quantified by point-count-

ing18 of cardiomyocytes with cytoplasmic staining of bgalactosidase activity (cytoplasmic LacZ, Figure 4b) or cells with positive b-galactosidase nuclei (nuclear localizing LacZ, Figure 4c) using a grid (18.2 18.2 mm in each side; 713 intersections in each image), and the percentage of positive points from the eight images were averaged for LV and two images for RV in each animal. Only intersections on cardiomyocytes were identified as positive or negative (Figure 4b and c).

1819

Echocardiography Transthoracic echocardiography was performed as previously described20 in age-matched C57BL/6 mice in a control, nonoperated group (n¼8) and in mice at 4 weeks after gene transfer with AAVLacZ (n¼6). Mice were anesthetized with intraperitoneal injection of 2.5% tribromoethanol (Avertin; 12 ml/g). Unpaired t-tests were performed to compare the two groups.

Acknowledgements This study was supported by Jean LeDucq foundation, the Richard D Winter Fund, the Perlman Cardiovascular Research Fund, and by HL46345 from the National Heart Lung and Blood Institute (KC). TD was funded by grants from the Swiss National Science Foundation (81BS64528) and the Freiwillige Akademische Gesellschaft Basel, Switzerland. We thank J Chrast for technical help.

References 1 Ikeda Y, Ross Jr J. Models of dilated cardiomyopathy in the mouse and the hamster. Curr Opin Cardiol 2000; 15: 197–201. 2 Ross Jr J. Dilated cardiomyopathy: concepts derived from gene deficient and transgenic animal models. Circ J 2002; 66: 219–224. 3 Towbin JA, Bowles NE. The failing heart. Nature 2002; 415: 227–233. 4 Minamisawa S et al. Chronic phospholamban-sarcoplasmic reticulum calcium ATPase interaction is the critical calcium cycling defect in dilated cardiomyopathy. Cell 1999; 99: 313–322. 5 Balling R. ENU mutagenesis: analyzing gene function in mice. Annu Rev Genom Hum Genet 2001; 2: 463–492. 6 Suzuki M, Carlson KM, Marchuk DA, Rockman HA. Genetic modifier loci affecting survival and cardiac function in murine dilated cardiomyopathy. Circulation 2002; 105: 1824–1829. 7 Christensen G et al. High-efficiency, long-term cardiac expression of foreign genes in living mouse embryos and neonates. Circulation 2000; 101: 178–184. 8 Su H, Lu R, Kan YW. Adeno-associated viral vector-mediated vascular endothelial growth factor gene transfer induces neovascular formation in ischemic heart. Proc Natl Acad Sci USA 2000; 97: 13801–13806. 9 Stratford-Perricaudet LD, Makeh I, Perricaudet M, Briand P. Widespread long-term gene transfer to mouse skeletal muscles and heart. J Clin Invest 1992; 90: 626–630. 10 Svensson EC et al. Efficient and stable transduction of cardiomyocytes after intramyocardial injection or intracoronary perfusion with recombinant adeno-associated virus vectors. Circulation 1999; 99: 201–205. 11 Agah R et al. Gene recombination in postmitotic cells. Targeted expression of Cre recombinase provokes cardiac-restricted, sitespecific rearrangement in adult ventricular muscle in vivo. J Clin Invest 1997; 100: 169–179. 12 Hajjar RJ et al. Modulation of ventricular function through gene transfer in vivo. Proc Natl Acad Sci USA 1998; 95: 5251–5256. Gene Therapy


1820 13 Maurice JP et al. Enhancement of cardiac function after adenoviral-mediated in vivo intracoronary beta2-adrenergic receptor gene delivery. J Clin Invest 1999; 104: 21–29. 14 Liao P et al. The in vivo role of p38 MAP kinases in cardiac remodeling and restrictive cardiomyopathy. Proc Natl Acad Sci USA 2001; 98: 12283–12288. 15 Chen J, Kubalak SW, Chien KR. Ventricular muscle-restricted targeting of the RXRalpha gene reveals a non-cell-autonomous requirement in cardiac chamber morphogenesis. Development 1998; 125: 1943–1949. 16 Hirota H et al. Loss of a gp130 cardiac muscle cell survival pathway is a critical event in the onset of heart failure during biomechanical stress. Cell 1999; 97: 189–198. 17 Soriano P. Generalized lacZ expression with the ROSA26 Cre reporter strain. Nat Genet 1999; 21: 70–71. 18 Ikeda Y et al. Restoration of deficient membrane proteins in the cardiomyopathic hamster by in vivo cardiac gene transfer. Circulation 2002; 105: 502–508. 19 Hoshijima M et al. Chronic suppression of heart-failure progression by a pseudophosphorylated mutant of phospholamban via in vivo cardiac rAAV gene delivery. Nat Med 2002; 8: 864–871. 20 Tanaka N et al. Transthoracic echocardiography in models of cardiac disease in the mouse. Circulation 1996; 94: 1109–1117. 21 Arber S et al. MLP-deficient mice exhibit a disruption of cardiac cytoarchitectural organization, dilated cardiomyopathy, and heart failure. Cell 1997; 88: 393–403. 22 Palakodeti V et al. Force–frequency effect is a powerful determinant of myocardial contractility in the mouse. Am J Physiol 1997; 273: H1283–1290. 23 Luo Y et al. Rescuing the N-cadherin knockout by cardiacspecific expression of N- or E-cadherin. Development 2001; 128: 459–469. 24 Subramaniam A et al. Tissue-specific regulation of the alphamyosin heavy chain gene promoter in transgenic mice. J Biol Chem 1991; 266: 24613–24620.

Gene Therapy

25 Huang WY, Aramburu J, Douglas PS, Izumo S. Transgenic expression of green fluorescence protein can cause dilated cardiomyopathy. Nat Med 2000; 6: 482–483. 26 Fishman GI. Timing is everything in life: conditional transgene expression in the cardiovascular system. Circ Res 1998; 82: 837–844. 27 Sohal DS et al. Temporally regulated and tissue-specific gene manipulations in the adult and embryonic heart using a tamoxifen-inducible Cre protein. Circ Res 2001; 89: 20–25. 28 Minamino T, Gaussin V, DeMayo FJ, Schneider MD. Inducible gene targeting in postnatal myocardium by cardiac-specific expression of a hormone-activated Cre fusion protein. Circ Res 2001; 88: 587–592. 29 Stanley EG et al. Efficient Cre-mediated deletion in cardiac progenitor cells conferred by a 3’UTR-ires-Cre allele of the homeobox gene Nkx2-5. Int J Dev Biol 2002; 46: 431–439. 30 Nagy A. Cre recombinase: the universal reagent for genome tailoring. Genesis 2000; 26: 99–109. 31 Lewandoski M. Conditional control of gene expression in the mouse. Nat Rev Genet 2001; 2: 743–755. 32 Metzger D, Chambon P. Site- and time-specific gene targeting in the mouse. Methods 2001; 24: 71–80. 33 Baluk P et al. Endothelial gaps: time course of formation and closure in inflamed venules of rats. Am J Physiol 1997; 272: L155–L170. 34 Gomez-Foix AM et al. Adenovirus-mediated transfer of the muscle glycogen phosphorylase gene into hepatocytes confers altered regulation of glycogen metabolism. J Biol Chem 1992; 267: 25129–25134. 35 Xiao X, Li J, Samulski RJ. Production of high-titer recombinant adeno-associated virus vectors in the absence of helper adenovirus. J Virol 1998; 72: 2224–2232. 36 Auricchio A et al. Isolation of highly infectious and pure adenoassociated virus type 2 vectors with a single-step gravity-flow column. Hum Gene Ther 2001; 12: 71–76.