Adenoviral Gene Transfer of Activated Phosphatidylinositol 3′-Kinase and Akt Inhibits Apoptosis of Hypoxic Cardiomyocytes In Vitro
Background—The intracellular signaling pathways that control cardiomyocyte apoptosis have not been fully defined. Because insulin-like growth factor-1 (IGF-1) prevents cardiomyocyte apoptosis, we examined the role of its downstream signaling molecules in an in vitro model of hypoxia-induced cardiomyocyte apoptosis.
Methods and Results—Treatment of rat neonatal cardiomyocytes with IGF-1 increased activity of both phosphatidylinositol 3′ (PI 3)-kinase and its downstream target, Akt (also known as protein kinase B or PKB). Cardiomyocytes were subjected to hypoxia for 24 hours, and apoptosis was assessed by DNA laddering, TUNEL staining, and ELISA for histone-associated DNA fragments. IGF-1 treatment (100 nmol/L) reduced cardiomyocyte apoptosis, and this effect was inhibited by simultaneous treatment with a PI 3-kinase inhibitor. Cardiomyocytes were infected with either a control adenovirus (Ad.EGFP) or adenoviruses carrying constitutively active forms of PI 3-kinase (Ad.BD110) or Akt (Ad.myr-Akt-HA). Ad.BD110 significantly inhibited apoptosis of hypoxic cardiomyocytes compared with Ad.EGFP (61.0±4.6% less DNA fragmentation than in Ad.EGFP-infected cells, P<0.0001). Ad.myr-Akt-HA even more dramatically inhibited apoptosis of hypoxic cardiomyocytes (90.9±1.4% less DNA fragmentation than in controls, P<0.0001).
Conclusions—IGF-1 activates PI 3-kinase and Akt in cardiomyocytes. Activated PI 3-kinase and Akt are each sufficient to protect hypoxic cardiomyocytes against apoptosis in vitro. Adenoviral gene transfer provides a useful tool for investigating the role of these signaling pathways in cardiomyocyte apoptosis.
Cardiomyocyte programmed cell death (apoptosis) has been identified in a wide variety of cardiovascular disorders, including myocardial infarction and heart failure (reviewed in Reference 11 ). Apoptosis is particularly prominent in models of ischemia-reperfusion injury, affecting a significant proportion of cardiomyocytes in the involved area.2 Although apoptosis might appear to be an attractive target for therapeutic intervention, the overall functional contribution of apoptosis to disease progression has not been defined in these conditions. Identifying the intracellular signaling pathways that control apoptosis in cardiomyocytes and developing approaches to modulating these pathways could help clarify their role in experimental models of cardiac disease and potentially lay the foundation for novel therapeutic strategies. The goals of this study were to identify signaling pathways that modulate cardiomyocyte apoptosis in vitro and to examine the ability of somatic gene transfer to manipulate these pathways.
Insulin-like growth factor-1 (IGF-1) blocks apoptosis in many settings,2 3 and this ability is generally dependent on activation of phosphatidylinositol 3′ (PI 3)-kinase.4 Recent studies have demonstrated that Akt (also known as protein kinase B or PKB) is the target of PI 3-kinase that is both necessary and sufficient to mediate inhibition of apoptosis in cerebellar neurons.3 Akt can inhibit caspase-mediated cell death through at least 3 mechanisms: phosphorylation of the death agonist Bcl-XL/Bcl-2–associated death promoter (BAD), releasing Bcl-2 family members5 ; direct phosphorylation of caspase-96 ; and phosphorylation of the FKHRL1 transcription factor, blocking Fas ligand expression.7 Therefore, PI 3-kinase appears to be a critical component in a signal transduction pathway mechanistically linking viability factors, such as IGF-1, to caspases through activation of the serine-threonine kinase, Akt.
The ability to manipulate the activity of these molecules in cardiomyocytes would facilitate an examination of their functional effects. Relatively specific pharmacological inhibitors of PI 3-kinase have been identified.8 9 In contrast, no inhibitors of Akt have been discovered to date, nor are there reagents that specifically and effectively activate either PI 3-kinase or Akt. In the present study, we used adenoviral gene transfer to directly examine the role of these molecules in an in vitro model of hypoxia-induced cardiomyocyte apoptosis.
In Vitro Model of Cardiomyocyte Hypoxia
Cardiomyocytes were prepared from 1- to 2-day-old rats by use of the Neonatal Cardiomyocyte Isolation System (Worthington Biochemical Corp), plated in 12-well plates or on glass coverslips in 6-well plates at 1×105 or 5×105 cells/well, respectively, and cultured in RPMI 1640/5% FCS/10% horse serum (HS) for 72 hours before infection. Cells were then incubated in RPMI 1640 alone for 2 hours with or without adenoviral vector, as indicated. An equal volume of RPMI 1640/5% FCS/10% HS was then added to each well, and cells were cultured for an additional 46 hours. For hypoxia, the medium was then changed to serum-free RPMI 1640 saturated with 95% N2/5% CO2, and cells were placed in a 37°C airtight box saturated with 95% N2/5% CO2 for 24 hours. O2 concentrations were <0.1% (Ohmeda oxygen monitor, type 5120). For normoxic controls, culture medium was changed to RPMI 1640/5% FCS/10% HS, and cells were placed in a 37°C/5% CO2 incubator for 24 hours before analysis.
TUNEL staining was performed with the Apoptag kit (Intergen), according to the manufacturer’s recommendations. Cardiomyocytes were identified by simultaneous immunostaining with the anti–sarcomeric α-actinin monoclonal antibody (mAb) EA-53 (Sigma). To quantify the number of apoptotic cardiomyocytes, nuclei were counterstained with Hoechst 33258 (Sigma), and the total numbers of nuclei and TUNEL-positive nuclei were counted in α-actinin–positive and –negative cells in 8 to 10 low-power fields in 3 independent experiments. More than 1500 nuclei were counted under both normoxic and hypoxic conditions. These data are given in the text as the mean±SD.
Myocytes (1×104) were trypsinized and resuspended in 0.75 mL of lysis buffer (100 mmol/L Tris-HCl [pH 8.5], 5 mmol/L EDTA, 0.2% SDS, 200 mmol/L NaCl, 100 μg/mL proteinase K). Cell lysates were incubated at 37°C for 4 hours with agitation and precipitated with an equal volume of isopropanol (1:1) at −20°C for 4 to 18 hours, followed by phenol/chloroform extraction and ethanol precipitation at −80°C for ≥15 minutes. Pellets were washed with 70% ethanol, air-dried, and resuspended in TE buffer (10 mmol/L Tris-HCl [pH 8.0], 1 mmol/L EDTA). RNase (Boehringer Mannheim) was added to a final concentration of 10 μg/mL. One microgram of each DNA sample was labeled with 0.5 μCi [α-32P]dCTP (DuPont NEN) in the presence of 5 U of Klenow polymerase (NEB, 10 minutes in NEB Klenow buffer) without additional unlabeled dNTPs. The reaction was terminated (10 mmol/L EDTA, 75°C for 10 minutes) and subjected to 1.8% agarose gel electrophoresis and autoradiography.
Cell Death ELISA
Histone-associated DNA fragments were quantified by ELISA (Boehringer Mannheim). All cells from each well were collected by trypsinization and pipetting, pelleted (800 rpm, Sorvall T6000B, 5 minutes), lysed, and subjected to the capture ELISA according to the manufacturer’s protocol. Data were normalized to the amount of DNA fragmentation seen with hypoxic Ad.EGFP-transduced cardiomyocytes. Each experiment was carried out in triplicate and repeated in ≥3 independent experiments.
PI 3-kinase activity was measured as described previously.10 Cell lysates were immunoprecipitated with either anti-phosphotyrosine mAb (PY20) for endogenous PI 3-kinase activity or anti-myc mAb (Santa Cruz Biotech) for BD110, which is not recognized by PY20. PI 3-kinase activity was measured in a reaction mixture containing phosphatidylinositol (Avanti) and [γ-32P]ATP (DuPont NEN). After 5 minutes, the reaction was stopped by the addition of HCl and chloroform:methanol and analyzed by thin-layer chromatography. PI 3-kinase activity was detected by the appearance of a specific radioactive spot corresponding to [32P]PI-3-P. We also verified that phosphorylation was occurring at the 3 position by using the specific PI 3-kinase substrate, PI-4,5-P2 (Sigma) (data not shown).
Akt kinase assays were performed as described previously.3 Cell lysates were immunoprecipitated with anti-Akt1 mAb (Upstate Biotech), used in kinase reactions with histone H2B (Boehringer Mannheim) as substrate and [γ-32P]ATP, and subjected to SDS-PAGE. The lower-molecular-weight area in the gel (with radiolabeled Histone H2B) was separated for direct autoradiography. The higher-molecular-weight area in the gel (with Akt) was transferred to a membrane for immunoblotting with anti-Akt1 mAb (Transduction Laboratories).
Cardiomyocytes were fixed with 10% neutral buffered formalin (22°C, 5 minutes), permeabilized (−20°C methanol:acetone [1:1], 10 minutes), and incubated with primary mAb to c-myc or α-actinin (30 minutes, 22°C). Cells were rinsed in PBS, incubated with anti-mouse IgG (Fab-specific) conjugated to tetramethylrhodamine (Sigma) (30 minutes, 22°C), and mounted.
Recombinant Adenoviral Vectors
Three recombinant first-generation type 5 adenoviruses were used for these studies. Ad.EGFP has been described previously.11 Ad.BD110 contains the myc-tagged BD11012 expression cassette in E1 and an EGFP (Clontech) expression cassette in E3 and was created through homologous recombination between pÄE1A-CMV-BD110 and pBHG11-EGFP in 293 cells, as previously described.13 The myr-Akt-HA–expressing adenovirus was constructed by first adding the src myristoylation signal to a cDNA clone expressing HA epitope–tagged Akt in pCAV-6, transferring the resulting insert into pACCMCLPA, and obtaining homologous recombinants through cotransfection with pJM17 (Microbix Biosystems) in 293 cells. Recombinant plaques were isolated and propagated in 293 cells, and transgene expression and appropriate kinase activity were verified in cardiomyocytes. Viral titer was determined by plaque assay in 293 cells. Stock titers were ≈1010 pfu/mL for each vector, with a particle/pfu ratio of ≈102. Wild-type adenovirus contamination was excluded by the absence of both PCR-detectable E1 sequences and cytopathic effects on the nonpermissive A549 cell line.
Data are represented as the mean±SEM of ≥3 independent experiments and were compared by ANOVA. The null hypothesis was rejected at P<0.05.
Hypoxia-Induced Apoptosis in Cardiomyocytes
Transient hypoxia (24 hours) induced apoptosis that was evident by TUNEL staining, DNA laddering, and ELISA for histone-associated DNA fragments, in agreement with previous reports14 15 16 (Figure 1⇓). Colocalization of TUNEL-positive cells with immunohistochemical staining for sarcomeric α-actinin confirmed that the apoptotic cells were cardiomyocytes, rather than contaminating fibroblasts (Figure 1A⇓). After hypoxia, 97% TUNEL-positive nuclei localized to α-actinin–positive cells and represented 21±2% of the nuclei in these cells. In contrast, only 2±1.7% of α-actinin–positive cells were TUNEL-positive in normoxic cultures. Only 3.0±1.2% of the TUNEL-positive nuclei occurred in the absence of α-actinin staining. Further evidence that cardiomyocytes were the predominant contributor to the apoptosis observed in this model was provided by 2 additional experiments in which cultures treated with BrdU to increase the proportion of cardiomyocytes16 yielded similar results (data not shown). Importantly, infection of cardiomyocytes with recombinant control adenovirus (Ad.EGFP) did not affect cardiomyocyte apoptosis (Figure 1B⇓ and 1C⇓).
Modulation of Hypoxia-Induced Apoptosis in Cardiomyocytes
IGF-1 effectively blocked cardiomyocyte apoptosis, consistent with previous reports15 (Figure 2a⇓). Although IGF-1 is a known activator of PI 3-kinase, we documented that this signaling mechanism was preserved in hypoxic cardiomyocytes. In hypoxic myocytes, PI 3-kinase was activated by IGF-1 treatment and significantly though incompletely inhibited by wortmannin (Figure 2b⇓) and LY294002 (data not shown). Hypoxic cardiomyocytes treated with both IGF-1 and wortmannin displayed significantly more residual PI 3-kinase activity than cells treated with wortmannin alone (data not shown). We tested the hypothesis that the IGF-1 survival benefit was mediated through PI 3-kinase by treating cultures with both IGF-1 and the specific PI 3-kinase inhibitors wortmannin and LY294002. LY294002 significantly inhibited the ability of IGF-1 to prevent apoptosis in hypoxic cardiomyocytes, and a more modest effect was seen with wortmannin (Figure 2a⇓). The ability of LY294002 to inhibit the survival benefit from IGF-1 strongly suggests that this benefit is mediated, at least in part, through PI 3-kinase. However, the combined treatments generally failed to induce the same amount of DNA laddering as seen in the hypoxic control cells. This incomplete restoration of apoptosis may reflect either incomplete inhibition of PI 3-kinase or the involvement of other IGF-1–activated signaling mechanisms. Of note, LY294002 and wortmannin did not significantly affect baseline cardiomyocyte apoptosis, although in some experiments DNA laddering under hypoxic conditions appeared to be modestly increased by PI 3-kinase inhibition (data not shown). Although these data suggest the involvement of PI 3-kinase activation in the antiapoptotic effect of IGF-1, we cannot exclude the possibility that IGF-1 and PI 3-kinase modulate apoptosis through independent mechanisms by this pharmacological approach.
To overcome this limitation, we used adenoviral gene transfer to directly examine the effects of specific signaling molecules. We constructed a recombinant adenoviral vector carrying a constitutively active form of PI 3-kinase, BD110.12 BD110 contains the p110 catalytic domain of PI 3-kinase fused in-frame with the binding domain (AA 474 to 552) of the regulatory subunit, p85,12 leading to constitutive binding and enzymatic activation. This virus also encodes an independent expression cassette carrying EGFP that allows direct identification of transduced cells by fluorescence microscopy (Figure 3⇓). Infection of cardiomyocytes with Ad.BD110 conferred activation of both PI 3-kinase and Akt (Figure 4a⇓ and 4b⇓). Treatment of cardiomyocytes with IGF-1 also activated Akt (Figure 4b⇓). These results both established the biological activity of Ad.BD110 and confirmed that Akt is a downstream target of PI 3-kinase in cardiomyocytes activated by IGF-1 treatment (Figure 4b⇓). Because Akt is the PI 3-kinase substrate essential for its survival benefit in other cells,3 we created a recombinant adenovirus carrying a constitutively active Akt (myr-Akt). This virus was also able to mediate robust transgene expression in cardiomyocytes and increased Akt activity even in the absence of IGF-1 (Figure 4b⇓).
Infection of cardiomyocytes with Ad.BD110 at a multiplicity of infection (MOI) of 10 plaque-forming units (pfu)/cell did not affect survival of unstimulated cells (Figure 5a⇓). Infection at an MOI ≥50 induced apoptosis of even unstimulated cardiomyocytes, suggesting that massive overexpression of this pleiotropic kinase is not well tolerated (data not shown). However, cardiomyocytes infected at an MOI of 10 pfu/cell were substantially protected against hypoxia-induced apoptosis (61.0±4.6% less DNA fragmentation versus Ad.EGFP-infected cells, P<0.0001) (Figure 5a⇓). Of note, at this MOI, virtually all of the cardiomyocytes were transduced, as evident by EGFP expression. The protection afforded by Ad.BD110 was abolished by simultaneous treatment with the PI 3-kinase inhibitor LY294002 (10 μmol/L). These results demonstrate that PI 3-kinase is sufficient to protect hypoxic cardiomyocytes against apoptosis in vitro.
In contrast, Ad.myr-Akt-HA did not increase cardiomyocyte apoptosis even at an MOI of 80 pfu/cell (Figure 5b⇑ and data not shown). In hypoxic cultures, Ad.myr-Akt-HA significantly inhibited hypoxia-induced DNA fragmentation in a dose-dependent manner at an MOI from 10 to 80 pfu/cell (P<0.0001 for all versus Ad.EGFP-infected cultures). At an MOI of 80 pfu/cell, Ad.myr-Akt-HA reduced DNA fragmentation of hypoxic cardiomyocytes by >90% compared with Ad.EGFP-infected cells (Figure 5b⇑).
In other cell types, Akt inhibits apoptosis through both direct and indirect inhibition of caspase activity.5 6 7 To examine whether caspase inhibition could account for the beneficial effects of PI 3-kinase or Akt activation on survival of hypoxic cardiomyocytes, we used the synthetic caspase inhibitor, z-VAD.fmk. In a dose-dependent manner, z-VAD.fmk reduced apoptosis of hypoxic cardiomyocytes (Figure 5c⇑). The maximum reduction in cardiomyocyte apoptosis seen with z-VAD.fmk treatment (93.5±2.0%, P<0.0001 versus uninfected controls) was comparable to that seen with Ad.myr-Akt-HA infection and greater than that achieved with BD110 at an MOI of 10 pfu/cell. Thus, caspase inhibition also blocks apoptosis of hypoxic cardiomyocytes and is quantitatively sufficient to account for the myocyte survival benefit seen with IGF-1 treatment as well as Akt or BD110 expression.
We examined the effects of intracellular signaling molecules downstream of the IGF-1 receptor in an in vitro model of transient cardiomyocyte hypoxia. We found that IGF-1 activates PI 3-kinase and Akt in cardiomyocytes. Pharmacological studies suggested that PI 3-kinase was involved in the ability of IGF-1 to inhibit apoptosis of hypoxic cardiomyocytes, and we used adenoviral gene transfer to directly examine the ability of activated PI 3-kinase or its downstream target, Akt, to modulate apoptosis. Expression of activated forms of either PI 3-kinase or Akt in cardiomyocytes was sufficient to inhibit hypoxia-induced apoptosis in vitro. To the best of our knowledge, this is the first demonstration of a critical role for these signaling molecules in cardiomyocyte survival.
Interestingly, Ad.BD110 only incompletely prevented hypoxia-induced DNA fragmentation at an MOI of 10 pfu/cell and killed cardiomyocytes at high MOI. Incomplete protection could reflect either low transduction efficiency with population heterogeneity, the inability of this signaling pathway to completely protect against apoptosis, or inadequate activation of this signaling pathway. The last explanation appears to be the most likely. The EGFP coexpression cassette revealed that virtually all the cardiomyocytes were transduced by Ad.BD110 at an MOI of 10 pfu/cell, consistent with previous reports.17 In addition, the substantially greater benefit conferred by Ad.myr-Akt-HA at high MOI provides evidence that further activation of this pathway alone is sufficient to protect hypoxic cardiomyocytes from apoptosis. It remains unclear why infection with Ad.BD110 killed even normoxic cardiomyocytes at high MOI; however, it may not be surprising that cells do not tolerate massive overexpression of this proximal kinase critical to so many fundamental cellular processes. Such observations may help guide future approaches to targeting these pathways in vivo and suggest that manipulation of Akt may ultimately prove to be a more robust strategy.
IGF-1 has been documented to inhibit caspase activity in cardiomyocytes.18 Akt can inhibit caspase-mediated cell death through multiple mechanisms,6 7 including release of antiapoptotic Bcl-family members.5 Of note, adenoviral overexpression of Bcl-2 in cardiomyocytes is itself sufficient to prevent p53-mediated apoptosis.19 We hypothesized that the survival benefit we observed in cardiomyocytes after expression of activated Akt or PI 3-kinase was probably mediated through caspase inhibition and tested the effects of a direct caspase inhibitor, z-VAD.fmk, in this model. In hypoxic cardiomyocytes, z-VAD.fmk blocked apoptosis to an extent comparable to the maximal benefit achieved through adenoviral activation of the PI 3-kinase/Akt signaling pathway. These results suggest that caspase inhibition could account for the benefits seen with adenoviral gene transfer of activated PI 3-kinase or Akt but do not establish a direct connection between these pathways in cardiomyocytes. It also remains to be seen whether the importance of these pathways will be conserved with other proapoptotic stimuli as well as in other in vitro and in vivo models.
Critical unanswered questions include the fate and functional capacity of cardiomyocytes in which apoptosis has been inhibited through activation of these signaling molecules. It is possible that these cells will simply die through another mechanism, such as necrosis or oncosis. Even if such cardiomyocytes survive, they may not function normally. Many pathways may contribute to hypoxia-induced myocyte dysfunction, and thus, inhibition of apoptosis could increase survival without improving the function of the surviving cells. Remarkably, the majority of cardiomyocytes transduced with BD110 at an MOI of 10 pfu/cell or myr-Akt at higher MOI continued to contract spontaneously and vigorously even after 24 hours of hypoxia, whereas virtually all cells transduced with the control virus stopped beating after hypoxia (data not shown). Rigorously defining the functional implications of specific apoptosis-related signaling pathways will be an important focus of future studies that should be facilitated by the gene transfer approach illustrated in the present study.
The present data are consistent with previous studies documenting the ability of IGF-1 to block apoptosis in many cell types, including cardiomyocytes,18 and to mediate beneficial effects in animal models of myocardial ischemia, whether delivered through peptide injection2 or transgenically overexpressed in the heart.20 Although it is tempting to speculate that activation of PI 3-kinase and Akt may play a role in the IGF-1–mediated benefits seen in these animal models,2 20 IGF-1 has systemic effects that could confound such studies. The recombinant vectors described in this report should provide useful tools for examining the relative contribution of local activation of PI 3-kinase and Akt to the observed benefits of IGF-1 in vivo. Understanding the role of specific pathways in cardiomyocyte apoptosis may help identify targets and novel therapeutic strategies for intervention in conditions such as myocardial infarction and heart failure.
This work was supported in part by grants from the NIH (HL-50361 and HL-57623 to Dr Hajjar; HL-59521 and HL-61557 to Dr Rosenzweig). Dr Rosenzweig is an Established Investigator of the American Heart Association. Dr Matsui is supported by an American Heart Association Fellowship. The authors thank Drs Thomas Force, Kazuyoshi Yonezawa, and Kenta Hara for their insightful suggestions. The technical assistance of Drs Zhaobin Kang and Quifen Qui is greatly appreciated, as is Paula Kaltofen’s help with manuscript preparation.
- Received March 26, 1999.
- Revision received June 29, 1999.
- Accepted July 9, 1999.
- Copyright © 1999 by American Heart Association
MacLellan WR, Schneider MD. Death by design: programmed cell death in cardiovascular biology and disease. Circ Res. 1997;81:137–144.
Buerke M, Murohara T, Skurk C, Nuss C, Tomaselli K, Lefer AM. Cardioprotective effect of insulin-like growth factor I in myocardial ischemia followed by reperfusion. Proc Natl Acad Sci U S A. 1995;92:8031–8035.
Dudek H, Datta SR, Franke TF, Birnbaum MJ, Yao R, Cooper GM, Segal RA, Kaplan DR, Greenberg ME. Regulation of neuronal survival by the serine-threonine protein kinase Akt. Science. 1997;275:661–665.
Yao R, Cooper GM. Requirement for phosphatidylinositol-3 kinase in the prevention of apoptosis by nerve growth factor. Science. 1995;267:2003–2006.
Cardone MH, Roy N, Stennicke HR, Salvesen GS, Franke TF, Stanbridge E, Frisch S, Reed JC. Regulation of cell death protease caspase-9 by phosphorylation. Science. 1998;282:1318–1321.
Arcaro A, Wymann MP. Wortmannin is a potent phosphatidylinositol 3-kinase inhibitor: the role of phosphatidylinositol 3,4,5-trisphosphate in neutrophil responses. Biochem J. 1993;296:297–301.
Vlahos CJ, Matter WF, Hui KY, Brown RF. A specific inhibitor of phosphatidylinositol 3-kinase, 2-(4-morpholinyl)-8-phenyl-4H-1-benzopyran-4-one (LY294002). J Biol Chem. 1994;269:5241–5248.
Endemann G, Yonezawa K, Roth RA. Phosphatidylinositol kinase or an associated protein is a substrate for the insulin receptor tyrosine kinase. J Biol Chem. 1990;265:396–400.
Hajjar RJ, Schmidt U, Matsui T, Guerrero JL, Lee KH, Gwathmey JK, Dec GW, Semigran MJ, Rosenzweig A. Modulation of ventricular function through gene transfer in vivo. Proc Natl Acad Sci U S A. 1998;95:5251–5256.
Kobayashi M, Nagata S, Kita Y, Nakatsu N, Ihara S, Kaibuchi K, Kuroda S, Ui M, Iba H, Konishi H, Kikkawa U, Saitoh I, Fukui Y. Expression of a constitutively active phosphatidylinositol 3-kinase induces process formation in rat PC12 cells. J Biol Chem. 1997;272:16089–16092.
Bett AJ, Haddara W, Prevec L, Graham FL. An efficient and flexible system for construction of adenovirus vectors with insertions or deletions in early regions 1 and 3. Proc Natl Acad Sci U S A. 1994;91:8802–8806.
Tanaka M, Ito H, Adachi S, Akimoto H, Nishikawa T, Kasajima T, Marumo F, Hiroe M. Hypoxia induces apoptosis with enhanced expression of Fas antigen messenger RNA in cultured neonatal rat cardiomyocytes. Circ Res. 1994;75:426–433.
Umansky SR, Cuenco GM, Khutzuan SS, Barr PJ, Tomei AD. Post-ischemic apoptotic death of rat neonatal cardiomyocytes. Cell Death Differ. 1995;2:235–241.
Hajjar RJ, Kang JX, Gwathmey JK, Rosenzweig A. Physiological effects of adenoviral gene transfer of sarcoplasmic reticulum calcium ATPase in isolated rat myocytes. Circulation. 1997;95:423–429.
Wang L, Ma W, Markovich R, Chen JW, Wang PH. Regulation of cardiomyocyte apoptotic signaling by insulin-like growth factor I. Circ Res. 1998;83:516–522.
Kirshenbaum LA, de Moissac D. The bcl-2 gene product prevents programmed cell death of ventricular myocytes. Circulation. 1997;96:1580–1585.
Li Q, Li B, Wang X, Leri A, Jana KP, Liu Y, Kajstura J, Baserga R, Anversa P. Overexpression of insulin-like growth factor-1 in mice protects from myocyte death after infarction, attenuating ventricular dilation, wall stress, and cardiac hypertrophy. J Clin Invest. 1997;100:1991–1999