RESEARCH ARTICLE Advanced generation lentiviruses as efficient vectors for cardiomyocyte gene transduction in vitro and in vivo

Transcription

1 (23) 1, & 23 Nature Pulishing Group All rights reserved /3 $25. RESEARCH ARTICLE Advanced generation lentiviruses as efficient vectors for cardiomyocyte gene transduction in vitro and in vivo D Bonci 1, A Cittadini 2, MVG Latronico 1,3, U Borello 4,5, JK Aycock 1, A Drusco 3, A Innocenzi 4, AFollenzi 6, M Lavitrano 7,MGMonti 2,JRossJr 8,LNaldini 6, C Peschle 1, G Cossu 4,5 and G Condorelli 3,9 1 Laoratory of Hematology-Oncology, Istituto Superiore di Sanità, Rome, Italy; 2 Department of Clinical Medicine and Cardiovascular Science, University Federico II, Naples, Italy; 3 Kimmel Cancer Center, Thomas Jefferson University, Philadelphia, PA, USA; 4 Stem Cell Research Institute, H. S. Raffaele, Milan, Italy; 5 Department of Histology and Medical Emryology, University La Sapienza, Rome, Italy; 6 Gene Transfer and Therapy and Tumor Immunology, IRCCS, Institute for Cancer Research and Treatment, University of Turin Medical School, Candiolo (TO), Italy; 7 Department of Environmental and Experimental Medicine and Medical Biotechnologies, University of Milano-Bicocca, Monza (MI), Italy; 8 Department of Medicine, UCSD School of Medicine La Jolla, CA, USA; and 9 II Medical School, IRCCS Neuromed, University La Sapienza, 161 Rome, Italy Efficient gene transduction in cardiomyocytes is a task that can e accomplished only y viral vectors. Up to now, the most commonly used vectors for this purpose have een adenoviral-derived ones. Recently, it has een demonstrated that lentiviral vectors can transduce growth-arrested cells, such as hematopoietic stem cells. Moreover, a modified form of lentiviral vector (the advanced generation), containing an mrna-stailizer sequence and a nuclear import sequence, has een shown to significantly improve gene transduction in growth-arrested cells as compared to the third-generation vector. Therefore, we tested whether the advanced generation lentivirus is capale of infecting and transducing cardiomyocytes oth in vitro and in vivo, comparing efficacy in vitro against the third-generation of the same vector. Here we report that advanced generation lentiviral vectors infected most (48%) cardiomyocytes in culture, as demonstrated y immunofluorescence and FACS analyses: in contrast the percentage of cardiomyocytes infected y third-generation lentivirus was three- to four-fold lower. Moreover, advanced generation lentivirus was also capale of infecting and inducing stale gene expression in adult myocardium in vivo. Thus, advanced generation lentiviral vectors can e used for oth in vitro and in vivo gene expression studies in the cardiomyocyte. (23) 1, doi:1.138/sj.gt Keywords: lentivirus; gene therapy; cardiomyocytes; cardiovascular diseases Introduction Gene transduction in the cardiomyocyte has always represented a major technical difficulty that has hampered detailed study of specific iochemical pathways in this cell type. In fact, conventional methods such as calcium phosphate transfections and liposome or lipofectamine-mediated gene transduction work poorly in cardiomyocytes. While these methods are useful for gene reporter studies, they are of little value when the aim is to introduce a gene into a sustantial percentage of cardiomyocytes. For this purpose, adenovirus (Ad) type 5 is the preferred vector. 1 However, drawacks of Ad vectors are their generation process, which is difficult and lengthy, and their immunogenicity, which prevents their use for long-term in vivo experiments. 1 Another vector, the adeno-associated virus (AAV), is suitale for in vivo myocardial gene transduction, ecause of its low Correspondence: Dr G Condorelli, Kimmel Cancer Center, Thomas Jefferson University, 233 S. 1th Street, room 16, Philadelphia, PA 1917, USA Received 11 June 22; accepted 18 Septemer 22 or asent immunogenic potential, 2 ut the use of AAV for in vitro studies is limited y the low gene expression (less then 1% of cardiomyocytes in culture) achieved during the first weeks of infection. 2 Thus, although adenoviral vectors have contriuted to the advancement of studies in cardiovascular pathophysiology, they are still not optimal for oth in vitro and in vivo studies. Retroviridae are RNA viruses whose RNA has to e reverse transcried into DNA in order to e integrated into the nucleus. 3 g-retroviruses, represented y Moloney virus, and lentiviruses, represented y HIV, are genuses of the Retroviridae family. Moloney virus requires nuclear reakdown, and therefore M phase, in order to e integrated into host genomic DNA, while lentivirus can enter the nucleus even without mitosis. 4,5 The aility of third-generation lentiviral vectors to infect growth-arrested cells (such as cardiomyocytes) has een shown in principle, 6,7 ut their transducing efficiency has never een analyzed in detail. Moreover, there are no reports demonstrating whether these vectors can e used for myocardial gene expression in vivo. Recently, a new variant of third-generation lentivirus (the advanced generation) has een descried 5 in which sequences of

2 the pol gene of HIV-1 (cppt) and of the post-transcriptional regulatory element of woodchuck hepatitis virus (WPRE) have een inserted. cppt is a cis element, which has een associated with more efficient gene expression in growth-arrested human hematopoietic progenitor cells. 5 Here, we descrie the use of advanced generation lentiviral vectors as an extremely efficient and simple method of inducing gene expression oth in vitro, in neonatal rat cardiac myocytes and in vivo in the adult rat heart. Results Advanced generation lentivirus efficiently transduces neonatal rat cardiomyocytes Neonatal rat cardiomyocytes (2.5 /well) were exposed to 1 ml supernatant containing the advanced generation lentiviral vector (1 ng p24) carrying a reporter gene. Cells were exposed to lentivirus for either 2 h or ON and expression assessed y FACS analysis 12 or 48 h later (Figure 1a and ). In cells exposed for 2 h, the percentage of -positive cells rose from 63% at 12 h to 72 % at 48 h. In cells exposed to the virus ON, the percentage of infection increased from 64 to 84% at 12 and 48 h, respectively. To determine the infection efficiency of this viral vector, scalar cell concentrations were exposed to the same amount of virus (1 ng p24/ml). The percentage of infection was evaluated y FACS analysis after exposing the cells for 2 h and detecting after 48 h (Figure 1c). Results show that with this concentration of virus, infection was highest (47%) when 2.5 cells were used. Viral stocks concentrated more than 1 ng of p24/ml produced toxicity. In some experiments, expression of the muscle-specific marker, myosin heavy chain (MHC), was detected y indirect immunofluorescence using a monoclonal antiody, MF2, and visualized y a commercial TRITClaeled goat-anti-mouse-igg secondary antiody (Sigma). Doule-laeled cells represented infected cardiomyocytes (Figure 1d). Thus, these experiments show that the lentiviral vector, at a concentration of 1 ng p24, can efficiently infect neonatal rat cardiomyocytes in vitro. Comparison of gene transduction efficiencies of advanced and third-generation lentiviral vectors in cardiomyocytes In order to compare the infection capacity of the two generations, cardiomyocytes were seeded at 5 / well, a numer at which expression reaches approximately 5% with 1 ng p24 of advanced viral supernatant. Cells were infected for 2 h with p24 concentrations of 1, 1, 1 or.1 ng/ml (1 6,1 5, and TU, respectively) of either advanced or thirdgeneration lentivirus (Figure 2c) and expression determined y FACS analysis after 48 h. The efficiency of infection was markedly different in the two types of 631 a c Exposure to the virus prrlcppt.pgk..wpre 2h O/N Time (hrs) cells/ml d Control d.1 d.2 d.3 Figure 1 Time-dependent and dose response effects of advanced generation lentiviral (prrlcppt.hpgk.e.wpre) gene transduction in cardiomyocytes in vitro. (a) Quantification of expression y FACS analysis after viral exposure of 2 h or ON and detection after 12 or 48 h (2.5 cells infected with 1 ng p24/ml). () FACS profile:.1: mock-infected cells;.2: 2 h viral exposure and detection 12 h later;.3: ON exposure and detection 48 h later. (c) FACS analysis of the transduction efficiency of 1 ng p24/ml with different cell concentrations. (d) Confocal microscopy of neonatal rat cardiomyocytes infected with lentivirus. Green: direct fluorescence (d.1); Red: indirect fluorescence with an anti-hmc antiody (d.2); Yellow: a merged image of the two fluorescences, showing coexpression of and myosin in the same cell (d.3).

3 632 a c ngp24/ml GA RRE GA CMV E sd sa sd c.1 c.2 RRE CMV sa cppt E wpre.3% 2.32% 7.3% 13% Control % 5.1% 24% 53% Figure 2 Comparison of the gene transduction efficiencies of third-generation and advanced generation lentiviruses in cardiomyocytes. Cardiomyocytes were infected at a concentration of 5 cells/ml. (a) The viral supernatants of third (lack line) or advanced (gray line) lentiviruses were used at the specified concentrations and FACS analysis performed after 48 h. Cells were incuated for 2 h with viral particles. () FACS profile of cells infected with third-generation (prrl.cmv.,.2 5) or advanced generation (prrlcppt.cmv..wpre,.6 9) lentivirus..1: mock-infected cells;.2/6: cells infected with.1 ng p24/ml supernatant;.3/7: cells infected with 1 ng p24/ml;.4/8: cells infected with 1 ng p24/ml;.5/9: cells infected with 1 ng p24/ ml. (c) Schematic drawing of the vectors used here. c.1: Third generation lentivirus: the vectors carry an internal cassette for the enhanced green fluorescent protein (E) driven y the CMV promoter. The LTR regions with a deletion of 4 p including the enhancer and promoter from U3 are indicated y AU3, R,U5; major splice donor site (sd); encapsidation signal (f) including the 5 portion of the gag gene (GA); Rev-response element (RRE); splice acceptor sites (sa); c.2: Advanced generation lentivirus: the inserted cppt and WPRE sequences are shown with arrows. lentivirus. In fact, the advanced generation lentivirus showed a four-fold increased efficiency at 1 ng p24/ml as compared to the third-generation lentivirus, and remained at least three-fold higher at lower dilutions (Figure 2a and ). Role of promoters in mediating the transduction efficiency of lentiviral vectors We tested the effect of two different promoters in determining gene transduction in cardiomyocytes. Lentiviral vectors containing either the human PGK or the CMV promoter in the expression cassette were produced (Figure 3a). Cells were exposed to virus for either 2 h or ON and FACS analysis of expression performed 48 h later. Results demonstrate that the viruses with these two different promoters had comparale efficiencies of infection. In fact, the percentage of infection reached more than 75% using viruses containing either one of the two types of promoters in cells exposed ON and approximately 7% in cells exposed for 2 h (Figure 3 and c). The efficiencies of infection were similar also when using diluted supernatants. Mean fluorescence intensity (MFI), a parametrer which corresponds to the strength of the promoter, was also measured. MFI was higher using the lentiviral vector constructed with the CMV promoter with respect to that with the hpgk one (Figure 3d). In vivo efficiency of lentiviral vectors Once the efficiency of lentiviral vectors was estalished in vitro, experiments were performed in order to address whether they are ale to induce efficient and long-lasting expression in myocardial cells in the intact animal. For this purpose, the hpgk promoter variant of the advanced generation lentivirus was used. A previous report on third generation lentiviral vectors in CNS infections demonstrated that lentiviral DNA had to e integrated into the host s genome for gene expression to

4 a sd a.1 GA RRE R sa cppt PGK E wpre GA RRE CMV E wpre sd sa cppt a h O/N prrlcpptcmv.. WPRE prrlcpptpgk.. WPRE ng of p24 Exposure to the virus c d 84% MFI 2328 Control c.1 c.2 c.3 76% MFI 4666 M.F.I (a.u) h O/N prrlcpptcmv.. WPRE prrlcpptpgk.. WPRE ng of p24 Exposure to the virus Figure 3 Comparison of the efficiency of cardiomyocyte gene transduction in vitro of advanced lentiviruses with PGK or CMV promoters. Cells (2.5 /ml) were exposed to different dilutions of either virus. (a) Schematic drawing of the vectors used in these experiments. The vectors differ only in the promoter region. E: enhanced green fluorescent protein. PGK: promoter of the human phosphoglycerate kinase gene. CMV: Cytomegalovirus promoter. cppt: nuclear import sequence. WPRE: regulatory element of woodchuck hepatitis virus. AU3, R, U5 are the LTR regions, with a deletion of 4 p, including the enhancer and promoter from U3; SD: major splice donor site. f: encapsidation signal including the 5 portion of the gag gene (GA). RRE: Rev-response element. SA: splice acceptor sites. () Percentage of GPF positive cells treated for 2 h or ON and analyzed after 48 h to either PGKcontaining (darker line) or CMV-containing (lighter line) lentiviruses. (c) FACS profile of: c.1: Mock-infected cells; c.2: PGK-promoter or; c.3: CMVpromoter lentivirus. (d) Mean fluorescence intensity (MFI) expressed in relative aritrary units analyzing cells at 48 h after transduction. We compared MFI of the two promoters using different concentrations of p24. take place in long-term in vivo experiments. 8 Thus, we used the same vector without integrase (int ), an enzyme critical for integration of lentiviral DNA with that of the host, as a control. The aility of integrase + (int+) and (int ) lentiviral vectors to infect and induce persistent expression was first tested on the TF1 leukemic cell line. These cells were selected since they are suitale for long-term (more than 2 weeks) experiments in vitro. FACS analysis after 48 h showed that more than 8% of TF1 cells infected with int+ advanced lentivirus scored positive for, while only around 4% of cells were infected with the int form. After 15 days, virtually no cells infected with the int vector scored positive for at FACS analysis, while the percentage of cells infected with the int+ lentivirus remained similar to that of 48 h (Figure 4). These two types of lentiviruses were then used for in vivo experiments. A volume of 2 ml of a 25 concentrated viral solution was injected into the left ventricle of rats while the pulmonary artery and aorta were eing clamped during immersion hypothermia, as descried in the Methods section. This procedure allows diffusion of the virus throughout the myocardium via the coronary arteries. -positivity was analyzed on sections of myocardial tissue 5 weeks after treatment. Results showed a sustained, diffuse transmural expression of after 5 weeks in the myocardium of the group of rats (n¼3) treated with the int+ vector. In fact, most cardiomyocytes of int+ virus-infected rats, staining positive to MHC, expressed, while was asent in int- lentivirus-infected rats (n¼3) (Figure 4a). Discussion Here, we report on the efficacy of a modified lentiviral vector as an easy and efficient way of inducing genes into the cardiomyocyte, oth in vitro and in vivo. Our comparison of third-generation and advanced lentiviral vectors demonstrates that the latter has an improved effect in transducing cardiomyocytes in vitro. Our data also descrie the conditions of infection needed for optimal gene transduction in vitro, showing that advanced lentiviral vectors induce a high percentage of positivity in cardiomyocytes after only 2 h of exposure. Previous reports have shown that lentiviruses can infect neonatal rat cardiomyocytes in vitro. 6,7 In one of these studies, a high percentage of infection was achieved, ut the exact amount of virus needed for extensive infection was not specified. 7 In oth cases, precise dosage and quantification of infection was not determined, nor was it assessed whether lentivirus induces extensive infection in vivo. In our hands, the

5 634 a Integrase + Integrase - Integrase + Integrase - a.1 a.2 a.3 a.4 MHC a.5 a.6 a.7 a h 15 days Time after transduction Figure 4 Myocardial infection with active integrase (int+) or an integrase-defective (int ) virus. (a) Sections of myocardium (2 a.1/5, a.2/6; 4 a.3/7, a.4/8) from rats injected with int+ (a.1/5, a.3/7) or int (a.2/6, a.4/8) lentiviral particles. a.1 4: direct fluorescence; a.5 8: indirect fluorescence for MHC. () Percentage of -positive TF1 cells after 48 h or 15 days of int (darker ar) or int+ (lighter ar) lentiviral treatments. third-generation lentivirus did not achieve satisfactory results, since the percentage of -expressing cells was generally three-fold lower than that achieved with the advanced generation vector, thus rendering impractical the use of this vector for studies requiring transduction of a high percentage of cardiomyocytes. On the other hand, advanced generation lentiviral vectors were very efficient, transducing more than 8% of cells in vitro after 48 h with only one round of infection. The differences oserved etween advanced and third-generation lentiviruses lie in the 118 p preceding the promoter region in the shuttle vector, as well as in the WPRE sequence at the 3 end of the expression cassette. Thus, even in this study these two elements have een shown to e critical in enhancing transduction efficiency. We also used two different promoters, CMV and hpgk. A similar percentage of gene transduction was achieved with oth, showing that either is suitale for studies in cardiomyocytes. It seems though, y analysis of MFI, that the expression driven y CMV is stronger than that of PGK. Furthermore, the lack of expression seen in vivo with the int variant of advanced generation lentivirus strongly suggests that lentiviral DNA must e integrated with cardiomyocyte DNA in order to achieve a sustained and lasting expression in this cell type, as has een demonstrated also for other growth-arrested cells y Follenzi et al. 5 In conclusion, the data reported in this work demonstrate that the advanced generation lentiviral vector is an excellent cdna carrier for cardiomyocytes, driving long-term gene expression oth in vitro and in vivo. Since lentiviruses are not immunogenic, it is possile to foresee their use in myocardial gene therapy studies in the future. Methods Cell cultures Neonatal rat cardiomyocytes were otained utilizing a modification of an original protocol Cells were cultured in DMEM-Medium199 (4:1) supplemented with 5% FBS, 5% HS, 1% L-glutamine and 1% penicillin/ streptomycin. Cells were treated with Mitomycin C (Sigma) to prevent firolast cell growth. Final cell populations contained more than 95% growth-arrested cardiomyocytes, as assessed y immunofluorescence analysis. 293 T cells (originally called 293tsA169ne 36 ) were grown in Iscove s modified Dulecco s medium (IMDM, Gico) supplemented with 1% FBS (Hyclone) and L- glutamine (5 U/ml), penicillin and streptomycin (5 U/ ml). They derive from 293 cells, a continuous human emryonic kidney cell line, transformed with sheared Type 5 Adenovirus DNA, and y transfection with the

6 tsa 169 mutant gene of SV4 Large T Antigen and the Neo r gene of E. coli. TF1, human erythroleukemia cells, were cultured in RPMI with 1 ng/ml GM-CSF and 1% FBS. Plasmids The three-plasmid expression system used to generate lentiviral vectors y transient transfection was used as previously descried. 5,12 The three plasmids were: the packaging plasmid, pcmvdr8.74 designed to provide the HIV proteins needed to produce the virus particle; the envelope-coding plasmid, pmd.g, for pseudotyping the virion with VSV-G, and; the self-inactivating (SIN) transfer vector plasmid (prrlcppt.hpgk.e.wpre or prrlcppt.cmv.e.wpre). The transfer vector plasmid contains the enhanced marker gene driven y either the human phosphoglycerate kinase promoter (hpgk) or the Cytomegalovirus promoter (CMV) and has een descried efore for assemling advanced third-generation lentivirus. 5 It has an additional DNA sequence of 118 p (cppt) situated efore the PGK/ CMV-E cassette, taken from the pol gene, which has een shown to e required in cis. The third-generation transfer vector used for some experiments was prrl.cmv.e. This vector is identical to that of the advanced generation apart from the asence of the cppt and WPRE sequences. Virus production We produced vector stocks y calcium phosphate transient transfection, cotransfecting the three plasmids in 293 T human emryonic kidney cells, since these cells are good DNA recipients. The calcium phosphate DNA precipitate was allowed to stay on the cells for h, after which the medium was replaced, collected 48 h later, centrifuged at 1 rpm for 5 min at room temperature and filtered through.22 mm pore nitrocellulose filters. Determination of viral titer and transfection efficiency In order to determine the viral particle concentration of the supernatants from 293 T cells, p24 antigen was analyzed y HIV-1 p24 Core profile ELISA (Aott Diagnostics or NEN TM Life Science Products) following the manufacturer s instructions. Moreover, the supernatants were used to infect TF1 cells to determine iological efficiency efore experimentation on cardiomyocytes. Transduction experiments were performed y adding serial dilutions of viral supernatant to 5 TF1cells/well in 24-well plates in the presence of Polyrene (4 mg/ml). Transduced cells were analyzed y FACS (FACS Caliur, Becton Dickinson Immunocytometry Systems) and CellQuest (Becton Dickinson) or WindMDI (Microsoft) software used. Typical supernatants contained approximately transfection units (TU) per ng p24 making the titer usually in the range of TU/ml. Viral stocks were usually prepared at 1 6 TU/ml, corresponding to 1 ng p24/ml. Usually, cells were infected at a concentration of 5 /ml, for non-saturating conditions or 2.5 /ml, for saturating conditions, with a supernatant titer of 1 ng p24/ml. Relative mean fluorescence intensity (MFI) was calculated using FACS analysis parameters. Concentration of viral supernatants For in vivo experiments, the supernatant was concentrated 25-fold. To otain high-titer vector stocks, medium collected from infected 293 T cells was ultracentrifuged at 5 g for 9 min at 41C. Pellets were resuspended with PBS containing.5% BSA, pooled and stored under liquid N 2. Transduction of cardiomyocytes Cardiomyocytes were plated 2 days efore exposure to the virus. Cells (2.5 ) were plated in 24-well plates. On the day of infection, the medium was removed and replaced with viral supernatant to which 4 mg/ml of Polyrene had een added. Cells were then centrifuged in their plate for 45 min in a Beckman GS-6KR centrifuge, at 18 rpm and 321C. After centrifugation, cells were kept for either 1 h 15 min or ON in a 5% CO 2 incuator at 32 or 371C, respectively. After exposure, cells were washed twice with cold PBS and fresh medium added. At either 12 or 48 h after the infection, cells were washed with PBS, harvested with trypsin/edta and analyzed y FACS. Infection and detection of in myocardial cells in vivo For in vivo experiments, a preparation of the virus 25- fold more concentrated than the regular viral supernatant was used. A method ased on that reported y Hajjar et al 13 and susequently modified y Ikeda et al 14 was used to deliver the virus. The validity of the method was first determined y gene transduction using adenoviral -galactosidase (not shown). Briefly, preliminary experiments were performed on male Sprague Dawley rats (Charles River, Italy) anesthetized with a mixture of ketamine hydrochloride (Sigma, 5 mg/kg BW) and Xylazine (Sigma, 1 mg/kg BW), and then orally intuated and ventilated. The animals were susequently cooled with ice-ags and their temperature monitored with a thermistor catheter. When the animal s temperature reached 31C, an anterior thoracotomy was performed, the heart was exteriorized and a 7. suture placed on the apex of the left ventricle. A 22 G catheter containing 2 ml of viral solution was then gently introduced into the left ventricle. The aortic root and pulmonary artery were identified and the catheter advanced through the left ventricle into the aortic root. The aorta and pulmonary artery were clamped gently, distal to the site of the catheter and the viral solution injected. The period of total occlusion (time for viral injection and the postinjection period) lasted 2 s, allowing the solution to circulate down the coronary arteries. Then the aortic and pulmonary clamps and the catheter were removed, the pneumothorax evacuated and the chest closed. Animals were transferred ack to their cages where they were allowed to recover. The same procedure was followed for sham-operated rats, ut the catheter was filled with saline. The extension of -galactosidase adenoviral infection was determined y 6Br-2-naphtyl--D-galactopyranoside/Fast Blue tissue staining after 15 days and compared to control. Most cells in the left ventricle showed lue granular condensations in the -galactosidase infected hearts, which were asent in sham- 635

7 636 operated animals and in control sections (not shown). Similar experiments were also performed with a vital dye, demonstrating a uniform distriution of the dye throughout the myocardium (not shown). Acknowledgements The following sources of support are acknowledged: American Heart Association (GLC), Italian Association for Cancer Research (GLC), Fondi 1% Ministero della Sanita -Italy (GC, LN and GLC), Telethon Association (GC and LN), European Community (GC) and Progetto Terapia Tumori Italy-USA (GLC) References 1 Hajjar RJ, del Monte F, Matsui T, Rosenzweig A. Prospects for gene therapy for heart failure. Circ Res 2; 86: Svensson EC et al. Efficient and stale transduction of cardiomyocytes after intramyocardial injection or intracoronary perfusion with recominant adeno-associated virus vectors. Circulation 1999; 99: Kay MA, Glorioso JC, Naldini L. Viral vectors for gene therapy: the art of turning infectious agents into vehicles of therapeutics. Nat Med 21; 7: Zennou V et al. HIV-1 genome nuclear import is mediated y a central DNA flap. Cell 2; 11: Follenzi A et al. Gene transfer y lentiviral vectors is limited y nuclear translocation and rescued y HIV-1 pol sequences. Nat Genet 2; 25: Mochizuki H et al. High-titer human immunodeficiency virus type 1-ased vector systems for gene delivery into nondividing cells. J Virol 1998; 72: Sakoda T, Kasahara N, Hamamori Y, Kedes L. A high-titer lentiviral production system mediates efficient transduction of differentiated cells including eating cardiac myocytes. J Mol Cell Cardiol 1999; 31: Naldini L et al. Efficient transfer, integration, and sustained long-term expression of the transgene in adult rat rains injected with a lentiviral vector. Proc Natl Acad Sci USA 1996; 93: Sen A et al. Terminally differentiated neonatal rat myocardial cells proliferate and maintain specific differentiated functions following expression of SV4 large T antigen. J Biol Chem 1988; 263: De Luca A et al. Characterization of caveolae from rat heart: localization of postreceptor signal transduction molecules and their rearrangement after norepinephrine stimulation. J Cell Biochem 2; 77: Condorelli G et al. Cardiomyocytes induce endothelial cells to trans-differentiate into cardiac muscle: implications for myocardium regeneration. Proc Natl Acad Sci USA 21; 98: Dull T et al. A third generation lentivirus vector with a conditional packaging system. J Virol 1998; 71: Hajjar RJ et al. Modulation of ventricular function through gene transfer in vivo. Proc Natl Acad Sci USA 1998; 95: Ikeda Y et al. Restoration of deficient memrane proteins in the cardiomyopathic hamster y in vivo cardiac gene transfer. Circulation 22; 15: