Cardiac Reanimation: Targeting Cardiomyocyte Death by BNIP3 and NIX/BNIP3L

Oncogene (2009) 27, S158–S167 & 2009 Macmillan Publishers Limited All rights reserved 0950-9232/09 $32.00 www.nature.com/onc REVIEW Cardiac reanimation: targeting cardiomyocyte death by BNIP3 and NIX/BNIP3L

GW Dorn II1 and LA Kirshenbaum2

1Department of Medicine, Center for Pharmacogenomics, Washington University School of Medicine, St Louis, MO, USA and 2Institute of Cardiovascular Sciences, St Boniface General Hospital Research Centre, University of Manitoba, Winnipeg, Manitoba, Canada

Programmed cardiac myocyte death contributes to are induced in response to cardiac injury or stress, in pathological ventricular remodeling and the progression development of heart failure, and describe the potential of myocardialinfarction or pressure overloadhypertrophy to prevent heart failure by interrupting BNIP3- and to dilated cardiomyopathy. Recent work has identified NIX/BNIP3L-mediated programmed cardiomyocyte importance of stress-mediated transcriptionalinduction of death. BNIP3 (BCL2 and 19-kDa interacting protein-3) and NIX/BNIP3L in cardiac remodeling. Here, the regulatory mechanisms for these two factors in the heart and their effects on programmed cardiomyocyte death are reviewed, Primary or secondary cardiomyocyte loss and the path with a focus on information derived from studies using to heart failure mouse models of cardiac BNIP3 and NIX/BNIP3L overexpression and gene ablation. Heart failure is a clinical syndrome defined as any Oncogene (2009) 27, S158–S167;doi:10.1038/onc.2009.53 condition where the rate of blood delivery to systemic tissues is insufficient to meet their metabolic needs. This Keywords: apoptosis;autophagy;heart failure;cardiac definition includes situations where myocardial function hypertrophy;myocardial infarction is normal and cardiac output (the volume of blood pumped per unit of time, B5 l/min in humans) is within accepted parameters or increased, such as in profound anemia and arteriovenous communications (shunts). However, the overwhelming majority of clinical heart Introduction failure cases are the consequence of primary myocardial dysfunction, with depressed cardiac output. In North The goal of revitalizing dead tissue is almost as old as America and Europe, heart failure is largely attributable our ability to understand the differences between life to ischemic or hypertensive cardiomyopathy, with and death. The biblical Lazarus miracle used the power genetic disorders, sleep apnea, viral, toxic and peripar- of God to restore life. In Mary Shelley’s Frankenstein, tum myocardial disease accounting for most other cases electricity provided the spark of life to the monster. (Hunt et al., 2005). Herbert West, a medical student in the HP Lovecraft’s Primary myocardial dysfunction leading to heart short story that was popularized in a series of the 1980s failure can be temporary, as with alcoholic cardiomyo- ‘Re-Animator’ horror movies, used chemical solutions to pathy and post-ischemic myocardial ‘stunning’ (Bolli revitalize small animals and, eventually, the Dean of the and Marban, 1999). More commonly, however, heart medical school (with unfortunate consequences). Now, failure is irreversible because it results from the physical modern tools of genetic manipulation and an explosion loss of functioning cardiac myocytes, so-called ‘cardio- in our understanding of the molecular processes that myocyte dropout’ (Diwan and Dorn, 2007). In coronary lead to cell death provide an opportunity not to bring atherosclerosis and ischemic heart disease, cardiomyo- dead tissue to life, but to interrupt intrinsic pathways cyte dropout tends to be focal, whereas it is more diffuse leading to predetermined cell death. The promise of this in nonischemic failure (Berry et al., 1993;Olivetti et al., approach is especially great in the heart, where tissue 1997). Importantly, the loss of functional contractile regeneration after injury does not occur to a meaningful units of the heart and their replacement by fibrous scar extent. Here, we review recent developments delineating or diffuse interstitial fibrosis produces secondary struc- the functions of BNIP3 (BCL2 and 19-kDa interacting tural changes in the ventricles that increase the work protein-3) and NIX/BNIP3L, two cell suicide genes that required for ventricular contraction. Surviving myocytes slip past one another and elongate as the extracellular Correspondence: Professor GW Dorn II, Department of Medicine, matrix remodels (Olivetti et al., 1990;Francis, Center for Pharmacogenomics, Washington University School of Medicine, 660 S Euclid Ave, Campus Box 8086, St Louis, MO 63110, 1998). Cardiomyocyte dropout and slippage produce USA. ‘ventricular remodeling’, resulting in an enlarged, thin- E-mail: [email protected] walled ventricle. These two factors, wall thinning and Targeting cardiomyocyte death GW Dorn II and LA Kirshenbaum S159 chamber dilation, strikingly increase the force that Ischemia Hypertrophy opposes ventricular contraction and blood ejection into HIF1α PKC the aorta. A mathematical description of this physical gene transcription Sp1 relationship was developed by the eighteenth century inactive French physicist, Pierre Laplace, and reveals that wall cytosolic stress (in the heart, afterload) is directly proportional to BNIP3 the radius of the chamber (r) and its intracavitary post-translational pressure (p), and is inversely proportional to the processing chamber wall thickness (h);stress ¼ pr/2h. Thus, any active active primary injury to the heart that causes cardiomyocyte mitochondrial mitochondrial NIX/BNIP3L dropout and ventricular remodeling produces a second- BNIP3 BCLXL ary stress on the heart in the form of increased afterload that is chronic and unremitting (Grossman et al., 1975). Studies performed over the past decade have shown that Proteasome Proteasome this hemodynamic stress is a powerful stimulus for programmed cardiomyocyte death, initiating a vicious cycle of unfavorable geometrical remodeling that stimulates yet more cell death (Foo et al., 2005;Dorn Post MI LV cyt c Hypertrophic II, 2009). The ultimate consequence, notwithstanding remodeling, Caspases LV remodeling the nature of the inciting stimulus, is a downward Diminished EF Apoptosis Diminished EF functional spiral progressing to dilated cardiomyopathy Figure 1 Transcriptional regulation of BNIP3 and NIX/BNIP3L. and end-stage heart failure. Separate pathways upregulate BNIP3 and NIX/BNIP3L in myocardial ischemia and hypertrophy, respectively, but they share a common apoptosis effector pathway. HIF-1a, hypoxia-inducible factor 1a;PKC, protein kinase C;Cyt c, cytochrome c. Cardiomyocyte apoptosis and BCL2 proteins in myocardialdisease death receptor pathways or the intrinsic, mitochondrial The cycle of cardiomyocyte dropout and ventricular pathway. As the focus of this review is on two mediators remodeling is chronic and persistent after a primary of mitochondrial pathway apoptosis, BNIP3 and NIX/ cardiac insult. This provides a therapeutic window of BNIP3L, the interested reader is referred to other recent months to years to implement targeted interventions reviews of extrinsic apoptosis in the heart and elsewhere that can interrupt the cycle and prevent progression to (Ashkenazi and Dixit, 1999;MacEwan, 2002;Danial end-stage heart disease. Furthermore, current under- and Korsmeyer, 2004). standing suggests that chronic cardiomyocyte dropout is The key initiators, regulators and effectors of intrinsic overwhelmingly an active process mediated by distinct, pathway apoptosis are the BCL2 family of mitochon- but functionally redundant, cell signaling pathways that drial-targeted pro- and antiapoptotic proteins. Cardiac lead to different forms of programmed cardiomyocyte myocytes express a number of pro- and antiapoptotic death (Figure 1). This rather complicated biology BCL2 family proteins that are transcriptionally regu- identified multiple potential therapeutic targets for lated in heart disease (Latif et al., 2000;Di Napoli et al., preventing ventricular remodeling through inhibition 2003). Proapoptotic BCL2 proteins are classified ac- of factors that cause programmed cardiomyocyte death. cording to structural features that reflect their different Apoptosis has been studied extensively in this regard. functions (Youle and Strasser, 2008). The ‘multidomain’ Myocardium is intrinsically resistant to apoptosis, proteins, BAX and BAK, are the essential pore-forming thought to be due at least in part to high levels of proteins that lead to mitochondrial outer membrane endogenous caspase inhibitors, such as X-linked in- permeabilization, cytochrome c release and induction of hibitor of apoptosis protein (XIAP) and apoptosis the intrinsic apoptosis signaling cascade. The activity of repressor with caspase recruitment domain (ARC) these pore-forming proteins is enhanced by proapopto- (Nam et al., 2004;Siu et al., 2005;Foo et al., 2007). tic ‘BH3 domain-only’ proteins, like BNIP3 and NIX/ Thus, the rate of apoptosis in normal human heart is B1 BNIP3L, and opposed by antiapoptotic factors like per 10 000 cardiomyocytes, but apoptosis increases BCL2 and BCLXL that bind to and inhibit the BH3- several hundred-fold in ischemic and dilated cardiomyo- only proteins. In the heart and elsewhere, the principal pathy and reactive cardiac hypertrophy (Narula et al., function of BH3-only proteins is to sense stress and 1996;Teiger et al., 1996;Olivetti et al., 1997). Because initiate mitochondrial translocation of BAX or activa- adult cardiac myocytes are incapable of meaningful tion of mitochondrial localized BAK. Cardiomyocyte cellular regeneration (Rubart and Field, 2006), chronic fate is therefore determined in part by regulated persistent apoptosis at a rate of 0.1–1% of total expression of these factors (Chen et al., 2001;Youle cardiomyocytes, as is observed in human cardiomyo- and Strasser, 2008). It seems surprising that cardiac pathies, may be sufficient to produce cardiac dilation myocytes, which are extremely limited in ability and heart failure (Hayakawa et al., 2003). Cardiomyo- to replace or regenerate their own losses, would cyte apoptosis can be mediated either through cytokine dynamically express programmed cell death genes under

Oncogene Targeting cardiomyocyte death GW Dorn II and LA Kirshenbaum S160 conditions of stress. Yet, BAX expression increases and the focus for molecular studies of myocardial NIX/ BCL2 expression decreases during chronic experimental BNIP3L regulation, and have revealed that protein cardiac pressure overloading (Condorelli et al., 1999), kinase C (PKC)-mediated upregulation of transcription NIX/BNIP3L is upregulated in human and experimen- factor SP1 increases NIX/BNIP3L promoter activity tal cardiac hypertrophy (Yussman et al., 2002;Galvez in cardiomyocytes in vitro and in vivo (Galvez et al., et al., 2006), and BNIP3 expression increases in ischemic 2006). cardiac models (Regula et al., 2002;Galvez et al., 2006). Given their structural and functional similarities, the Recent data support a particularly important function distinct transcriptional control pathways for NIX/ for BNIP3 and NIX/BNIP3L in cardiomyocyte apop- BNIP3L and BNIP3 are intriguing, and suggest that tosis during ventricular remodeling. specific promoter elements are crucial for regulating NIX/BNIP3L or BNIP3 transcription by different stress signals. Sequence analysis of the proximal BNIP3 promoter uncovered a classical HIF-1 element and Stress-specific transcriptionalcontrolof cardiac BNIP3 consensus sequences for the transcription factors nucle- and NIX/BNIP3L ar factor-kB (NF-kB) and an E2F-1 (Baetz et al., 2005; Tracy et al., 2007;Yurkova et al., 2008). BNIP3 gene BNIP3 and NIX/BNIP3L are structurally similar expression in neonatal rat cardiac myocytes was products of separate genes and are cardiac expressed transcriptionally silenced by constitutive binding of and transcriptionally regulated. Both of these BH3-only NF-kB to the BNIP3 promoter by recruitment of factors localize to and (by activating BAX and BAK) histone deacetylase-1 inhibitory complexes. Notably, permeabilize mitochondrial outer membranes, which the induction of BNIP3 gene expression during hypoxia activates the intrinsic apoptosis signaling pathway coincided with a reduction in cellular NF-kB activity, (Chen et al., 1999). Preventing mitochondrial localiza- and interventions that restored NF-kB BNIP3 promoter tion of either BNIP3 or NIX/BNIP3L by mutational binding suppressed hypoxia-induced BNIP3 gene tran- deletion of C-terminal mitochondrial targeting se- scription. These findings highlight the importance of quences creates dominant inhibitory proteins (Regula NF-kB in transcriptional control of BNIP3. et al., 2002;Yussman et al., 2002). Both proteins are also Recently, several groups have demonstrated that subject to rapid degradation by the ubiquitin–proteo- stress signals coupled to cell-cycle factor E2F-1 are some pathway (Chen et al., 1999;Cizeau et al., 2000), sufficient to activate BNIP3 gene transcription during accounting for barely detectable levels observed in the hypoxia (Tracy et al., 2007;Yurkova et al., 2008). In unstressed myocardium. Importantly, there appears to gain-of-function studies, E2F-1 stimulated BNIP3 gene be no essential physiological or developmental function transcription, provoking mitochondrial defects and cell for either NIX/BNIP3L or BNIP3 in the heart, as death in neonatal cardiomyocytes. Because genetic cardiac structure and function are normal in the knockdown of E2F-1 or overexpression of a caspase- respective cardiac-specific and germ-line gene knockout resistant Rb suppressed hypoxia-induced BNIP3 gene mouse models (Diwan et al., 2007b, 2008b). activation, the conclusion was that induction of BNIP3 Because inappropriate expression of NIX/BNIP3L or promoter activity during hypoxia is contingent on BNIP3 is lethal to cardiac myocytes (see below), these E2F-1 (Yurkova et al., 2008). In contrast, NIX/BNIP3L death factors need to be tightly regulated. Indeed, both gene expression is unaltered in cardiac myocytes by NF- proteins are subject to strong negative repression under kB, E2F-1 or hypoxia. nonapoptotic conditions, but are selectively and ro- Because BNIP3 and NIX/BNIP3L are potent death bustly induced by stress signals. This feature under- factors, it follows that exquisite transcriptional control scores their importance as potential therapeutic targets, over factor genesis must be balanced by similarly tight and considerable effort has been devoted to deciphering control over factor activity and degradation. Indeed the signaling pathways and transcriptional control BCL2 and BCLXL interact with NIX/BNIP3L and processes that govern basal and inducible NIX/BNIP3L BNIP3 (Chen et al., 1997;Yasuda et al., 1998;Ray and BNIP3 transcription. An important finding result- et al., 2000) to antagonize apoptosis and/or autophagy. ing from this work is that BNIP3 is uniquely induced in The growth factors epidermal growth factor and insulin- the heart by hypoxia, in contrast to other cardiac- like growth factor appear to inhibit BNIP3-induced cell expressed proapoptotic factors such as BAD, NIX/ death in CHO and MCF7 cells (Kothari et al., 2003). BNIP3L, BAK and BAX (Regula et al., 2002;Yurkova Work by the Webster laboratory indicated that acidosis et al., 2008). NIX/BNIP3L may also be upregulated during hypoxia enhances BNIP3-induced mitochondrial in an HIF-1a-dependent manner in some tissues PTP opening and cell death, suggesting either that (Bruick, 2000;Sowter et al., 2001;Birse-Archbold acidosis stabilizes BNIP3 protein, or that it can provide et al., 2005), but NIX/BNIP3L gene expression is not an as-yet undefined activating signal (Kubasiak et al., induced by hypoxia/ischemia in the heart or in cultured 2002;Frazier et al., 2006). Ubiquitination and rapid cardiomyocytes. Because increased myocardial NIX/ proteasomal degradation of BNIP3 and NIX/BNIP3L BNIP3L mRNA levels were originally described was mentioned above (Chen et al., 1999;Cizeau et al., in the heart in Gaq-mediated and pressure overload 2000). Finally, the recent observation that BNIP3 can be hypertrophy (Yussman et al., 2002), signaling mechan- reversibly phosphorylated may reveal another point of isms within the Gq signaling pathway have been regulatory control (Graham et al., 2007). As more

Oncogene Targeting cardiomyocyte death GW Dorn II and LA Kirshenbaum S161 complete details of tissue- and stress-specific pathways stress that helps to coordinate transcriptional and underlying the regulation of BNIP3 and NIX/BNIP3L physiological cues for programmed cardiomyocyte are be defined, it is safe to say that the complexities death (Dorn II, 2005). delineated to date appear to reflect highly structured A cellular mechanism for the putative stress-sensing pathways that have evolved so that the heart can react function of NIX/BNIP3L was recently delineated, based to a variety of injuries or stressors by selectively ‘fine- on the observation that only B80% of transfected NIX/ tuning’ the disposition of damaged or injured cardio- BNIP3L localize to mitochondria in neonatal rat myocytes. cardiomyocytes or HEK293 cells, with the remainder localizing to the endoplasmic reticulum/sarcoplasmic reticulum (ER/SR) (Diwan et al., 2008a). These organelles are the predominant sites for intracellular Transgenic cardiac overexpression studies of BNIP3 calcium stores, and the highly organized subcellular and NIX/BNIP3L architecture of adult cardiac myocytes enforces close proximity between SR and mitochondria, making it Because endogenous cardiomyocyte BNIP3 and NIX/ possible for calcium to transfer from SR to mitochon- BNIP3L are upregulated by myocardial stress or injury dria through junctional microdomains referred to as that can produce programmed cell death through other ‘calcium hot-spots’ (Rizzuto and Pozzan, 2006). Cal- pathways, it was difficult to determine from simple cium delivery to mitochondria and uptake by the association studies whether increased BNIP3 or NIX/ calcium uniporter are a potent stimulus for mitochon- BNIP3L was sufficient to overcome the intrinsic drial permeability transition (MPT) in the heart and apoptosis resistance of cardiomyocytes. To address this elsewhere, and the resulting inner and outer membrane issue the Dorn laboratory used cardiomyocyte-specific rupture leads to cell death from ATP depletion, promoters to transgenically overexpress both proteins in called ‘programmed necrosis’ (Nakayama et al., 2007; mouse hearts, and examined patterns of cardiomyocyte Henriquez et al., 2008). The Dorn group found that apoptosis in the absence and presence of myocardial there is a direct relationship between NIX/BNIP3L level stress. in NIX/BNIP3L-overexpressing and -knockout mice NIX/BNIP3L was first overexpressed in mouse hearts and SR calcium content, suggesting that ER/SR- using the standard a-myosin heavy chain (aMHC) localized NIX/BNIP3L can affect intracellular calcium transgene promoter (Yussman et al., 2002). Two stores in a similar manner as previously reported for important characteristics of the aMHC expression BCL2 and BAX (Foyouzi-Youssefi et al., 2000; system are that it is cardiomyocyte specific, and that Nutt et al., 2002;Scorrano et al., 2003). To determine measurable overexpression begins shortly after the birth if ER/SR-localized NIX/BNIP3L contributed to proof the mouse (Syed et al., 2004). As a consequence, grammed cell death, mitochondria- and ER/SR-specific intrauterine cardiac development is not affected by mutants were engineered to compare their effects on cell aMHC-driven transgenes. aMHC-NIX/BNIP3L mice viability, caspase activation and mitochondrial inner were born at expected Mendelian ratios, but their membrane potential (which is lost on opening of the neonatal growth was stunted, and they succumbed after permeability transition pore). Like wild-type NIX/ B1 week of life. Echocardiography performed on day 6 BNIP3L, both mitochondrial and ER/SR-specific revealed left ventricular dilation and poor contractility, NIX/BNIP3L caused programmed cell death associated and histological studies of aMHC-NIX/BNIP3L hearts with caspase activation. A NIX/BNIP3L mutant that on days 6–7 revealed striking increases in the rates of remained in the cytosol did not produce cell death. cardiomyocyte apoptosis (15–20%). These findings were Strikingly though, the mitochondrial localized NIX/ similar in two independent transgenic lines, demonstrat- BNIP3L did not cause MPT, and resulting cell death ing that increased NIX/BNIP3L expression is sufficient was apoptotic (type 1 cell death), whereas the ER/SR- to cause a lethal apoptotic cardiomyopathy in neonatal localized NIX/BNIP3L caused mitochondrial depolar- mice. In a subsequent study where a conditional ization and death from programmed necrosis (type 3 cell (tetracycline-suppressible) aMHC promoter was used death). These findings demonstrate that NIX/BNIP3L to drive cardiomyocyte NIX/BNIP3L expression either mediates intrinsic pathway apoptosis though permeabi- from birth, or beginning in young adulthood, cardio- lization of mitochondrial outer membranes and myocyte apoptosis was much more prevalent in neo- cytochrome release and so on. However, NIX/BNIP3L nates (B15%) than adults (B3%), and NIX/BNIP3L also causes programmed necrosis by localizing to overexpression at even high levels in adults produced ER/SR, where it increases calcium stores and therefore little effect on cardiac size or contractile function. the mitochondrial delivery of calcium that causes the However, when adult NIX/BNIP3L overexpression pore transition. In the context of recent studies was superimposed on microsurgical creation of an aortic identifying a role for NIX/BNIP3L (and probably pressure gradient that in normal mice produces func- BNIP3 as well (Hamacher-Brady et al., 2007; tionally compensated pressure overload hypertrophy, Kubli et al., 2008)) in mitochondrial autophagy the result was left ventricular remodeling and progres- (Schweers et al., 2007;Sandoval et al., 2008), these sion to heart failure. These studies suggested that NIX/ findings position NIX/BNIP3L at the apex of multiple BNIP3L is not only an initiator and effector of parallel pathways leading to programmed cell death cardiomyocyte apoptosis, but is also a sensor of cardiac (Figure 2).

Oncogene Targeting cardiomyocyte death GW Dorn II and LA Kirshenbaum S162 Injury/Stress suggest that during ischemic stress, factors other than simple upregulation are important in generating the full ∆ Gene Expression Increased [Ca++]i apoptotic response to BNIP3 in ischemic injury. NIX/BNIP3L and BNIP3 Whether this is hypoxia/acidosis-mediated post-transla- BCL2/BCLXL tional modiﬁcation and mitochondrial translocation of BNIP3 as has been suggested (Kubasiak et al., 2002; BAX/BAK Kubli et al., 2008), or some other as-yet unidentiﬁed factor such as nuclear translocation that is seen in Apoptosome Cyt c ischemic brain injury (Schmidt-Kastner et al., 2004; Apaf-1 dATP Mitochondria SR Casp 9 Ca++ Burton et al., 2005), is not currently known.

MPTP

XIAP ARC AIF Stimulus-specific interruption of myocardial apoptosis Endo G through targeted ablation of BNIP3 or NIX/BNIP3L H 0 Casp 3 Caspase- 2 BNIP3 and NIX/BNIP3L are transcriptionally upregu- independent apoptosis lated in myocardium by different stimuli, but have Programmed necrosis similar effects on cardiomyocyte death, and thereby DNA frag DNA frag ATP Mitophagy (200 bp) (50,000 bp) depletion aggravate apoptotic cardiac remodeling. Accordingly, Autophagy interrupting BNIP3- or NIX/BNIP3L-mediated pro- Caspase-dependent Cell Death grammed cell death in the heart might be able to prevent apoptosis the adverse remodeling and heart failure that results Figure 2 Death effector pathways for BNIP3 and NIX/BNIP3L. from chronic cardiomyocyte dropout in ischemically NIX/BNIP3L and BNIP3 stimulate caspase-dependent and damaged or hypertrophied hearts. The Dorn group -independent apoptosis by interacting with BAX or BAK to permeabilize mitochondrial outer membranes. Smaller membrane examined this notion by creating the respective mouse pores permit release of cytochrome c (Cyt c) that activates caspases. gene knockout models, which were then subjected to a This pathway is inhibited in cardiac myocytes by high endogenous stimulus that would normally upregulate the ablated levels of the inhibitor of apoptosis proteins (XIAP) and apoptosis gene. Apoptosis, ventricular remodeling and cardiac repressor with caspase recruitment domain (ARC). When pores in performance were then compared over time in identi- the mitochondrial outer membrane enlarge sufficiently, apoptosis- inducing factor (AIF) and endonuclease-G (endo G) are also cally treated knockout and wild-type mouse hearts. For released and translocate to the nucleus to stimulate DNA the BNIP3 germ-line knockout mouse (Diwan et al., degradation in a caspase-independent manner. Sarcoplasmic 2007b), a model of reversible left anterior descending reticular (SR) NIX/BNIP3L, and possibly BNIP3, increases coronary artery ligation that mimics early reperfusion calcium transport to mitochondria, resulting in opening of the mitochondrial permeability transition pore (MPTP) that leads to after myocardial infarction was employed. As BNIP3 necrotic cell death. NIX/BNIP3L and BNIP3 also target mito- levels are barely detectable in normal, nonischemic chondria for autophagic degradation, either directly through as-yet myocardium, there is no real change in cardiac BNIP3 unknown mechanisms, or because of nonspecific consequences of expression at baseline in mice lacking a functional MPTP-induced damage. BNIP3 gene. Consistent with this observation, BNIP3 ablation did not affect infarct size because these myocytes die shortly after interruption of blood flow, Although the mechanistic studies for BNIP3 have not and before they can upregulate BNIP3 gene and protein been performed at the same level of detail as with NIX/ expression. However, when post-ischemic apoptotic BNIP3L, we believe that NIX/BNIP3L and BNIP3 have cardiomyocyte dropout was examined in the days after very similar cellular effects and the NIX/BNIP3L the acute ischemic insult (that is, after sufficient time for findings likely apply to BNIP3 under conditions where mechanisms that increase BNIP3 gene expression in its expression is increased. The Dorn group has also ischemic tissue to be invoked), cardiomyocyte apoptosis examined the consequences of cardiomyocyte BNIP3 was reduced by B50% in both the peri-infarct region overexpression using a conditional cardiac expression and the myocardium remote from the ischemic zone of system (Diwan et al., 2007b). In contrast to NIX/ BNIP3 null mice. Decreased apoptosis in infarcted BNIP3L, BNIP3, expressed at a level approximately BNIP3 knockout hearts was associated with striking 50-times the normal (very low) level of basal myocardial declines in left ventricular remodeling and improved expression, increased the basal rate of cardiomyocyte cardiac performance, as measured by magnetic reso- apoptosis only minimally (to B1%) after neonatal nance imaging. Thus, prevention of BNIP3-mediated expression. However, mouse hearts in which BNIP3 cardiomyocyte death during the days and weeks after a overexpression was induced in adulthood developed transient ischemic insult decreases apoptosis and pre- markedly larger myocardial infarctions after microsur- vents ventricular dilation, which preserved left ventri- gical left anterior descending coronary artery occlusion cular ejection performance and abrogated heart failure. (B35% of the left ventricle) than did their nontrans- It should be noted that although apoptosis was the only genic littermates (B20% of the left ventricle) (Diwan form of programmed cell death specifically measured in et al., 2007b). These results indicate that increased these studies, the favorable effects of BNIP3 ablation BNIP3 expression can modestly increase cardiomyocyte on cardiomyocyte dropout and ventricular remodeling apoptosis in normal hearts. More importantly, they also may also accrue from prevention of other forms of

Oncogene Targeting cardiomyocyte death GW Dorn II and LA Kirshenbaum S163 programmed cell death (autophagy, programmed ne- potential to confound cardiac studies, a Cre-lox strategy crosis) that may be mediated by BNIP3. was used to create mice in which the NIX/BNIP3L gene There are several theoretical benefits to some form of was selectively ablated in cardiac myocytes, which were anti-BNIP3 therapy after myocardial infarction or subjected to surgical transverse aortic constriction ischemia. First, the kinetics of BNIP3 upregulation (TAC) (Diwan et al., 2008b). This surgical model after ischemia and the apparent chronic BNIP3- produces chronic pressure overload that can be modified mediated cardiomyocyte dropout provide for a sizeable to the needs of the particular experiment (mild, therapeutic window. Post-infarction ventricular remo- moderate or severe), and is therefore somewhat analo- deling occurs over the weeks and months following an gous to clinical hypertension or valvular aortic stenosis. acute ischemic event. This contrasts with the narrow As in these human conditions, mouse hearts subjected to therapeutic window of only 90 min after hospital TAC exhibit histological evidence of increased cardio- presentation (or B6 h from infarction) for coronary myocyte apoptosis that is associated with progressive reperfusion therapy to reduce myocardial necrosis and ventricular dilation and, ultimately, functional decom- infarct size (Antman et al., 2008). Second, anti-BNIP3 pensation. We observed that TAC in NIX/BNIP3L therapy might not necessarily need to be given knockout mice produced only half the rate of early specifically to the myocardium. The germ-line BNIP3 cardiomyocyte apoptosis and late replacement fibrosis knockout mouse showed no detectable developmental as in control hearts, and that ventricular dilatation and or functional abnormalities in any organ system, wall thinning were largely prevented over the subsequent including the hematopoietic system wherein increased weeks, thus preserving systolic function. Importantly, numbers of lymphoid, myeloid and erythroid cells have both at the cellular level and in terms of whole organ been described for other BH3-only factor gene knockout weights, NIX/BNIP3L ablation did not change reactive models (Bouillet et al., 1999;Ranger et al., 2003;Diwan left ventricular hypertrophy after pressure overloading. et al., 2007a). This is consistent with the concept that These data provided the first evidence that apoptosis is BNIP3 functions almost exclusively in an inducible truly causal in decompensation of pressure overload, manner, and that ischemia is critical to increase its and establish NIX/BNIP3L as a critical inducible factor expression and possibly to activate it post-translation- that mediates this response. Furthermore, the NIX/ ally (Kubasiak et al., 2002;Kubli et al., 2008). BNIP3L and BNIP3 knockout remodeling rescue Accordingly, systemic anti-BNIP3 therapy has the studies taken together prove that necrotic cardio- potential to prevent post-ischemic cardiac remodeling myocyte death is not the inevitable consequence of without untoward consequences in other tissues. Final- preventing the programmed death of ischemic or hyper- ly, the benefits of preventing programmed cell death trophied cardiomyocytes, as has been suggested during cardiac remodeling multiply by interrupting the (Foo et al., 2005). vicious cycle of programmed cardiomyocyte death that increases myocardial stress by producing adverse geometrical changes, providing an additional and continuous stimulus for further programmed cell death. BNIP3 and NIX/BNIP3L in nonapoptotic programmed Possible approaches to inhibit BNIP3 include a cell- cell death permeant (TAT-coupled) loss-of-function BNIP3 transmembrane domain deletion mutant that has been used An underlying assumption of many of the studies in isolated perfused hearts (Hamacher-Brady A et al., reviewed above is that BNIP3 and NIX/BNIP3L kill 2007) or small molecules that could be designed to cardiomyocytes through caspase-dependent apoptosis interfere with BNIP3 homo- and heterodimerization (type 1 cell death). However, it was recognized early on (Oltersdorf et al., 2005;Stauffer, 2007). that BNIP3 can cause programmed cell death in a Because it is upregulated in hypertrophy, it was manner that does not require caspase activation and postulated that inhibition of NIX/BNIP3L could pre- involve MPT (type 3 cell death). Our recent studies with vent ventricular remodeling and the progression to heart organelle-specific NIX/BNIP3L mutants have more failure in chronically pressure-overloaded hearts. We clearly defined a likely mechanism for NIX/BNIP3L- tested this notion using a similar strategy, microsurgical and BNIP3-mediated programmed necrosis via opening modeling of a NIX/BNIP3L knockout mouse. However, of mitochondrial permeability transition pores (MPTP, in contrast with BNIP3 where germ-line gene ablation see above) (Diwan et al., 2008a). Because MPTP- showed no baseline phenotype, the germ-line NIX/ mediated cardiomyocyte death appears to have a major BNIP3L knockout mouse produced in the Dorn lab function in various forms of heart injury associated with exhibited markedly increased blood reticulocyte counts increased cytosolic or sarcoplasmic reticular calcium associated with decreased in vivo and in vitro apoptosis levels (Nakayama et al., 2007;Diwan et al., 2008a), the of splenic erythroid precursers, and extreme hypersensi- mechanism is reviewed (see Figure 2). Mitochondria are tivity to erythropoietin (Diwan et al., 2007a). A similar normally the major source of cellular ATP stores, which phenotype that included hemolytic anemia was subse- is needed to sustain the normal electrochemical gradient quently described by two other groups who indepen- (Dcm) across mitochondrial inner membranes. Calcium dently produced NIX/BNIP3L knockout mice overload stimulates calcium-sensitive mitochondrial (Schweers et al., 2007;Sandoval et al., 2008). Because matrix dehydrogenases and impairs NADH production hematopoietic abnormalities such as these had the within the respiratory chain. Consequently, the inner

Oncogene Targeting cardiomyocyte death GW Dorn II and LA Kirshenbaum S164 mitochondrial membrane becomes more permeable to was suggested to be cytoprotective (Hamacher-Brady ions, resulting in loss of Dcm. This is the MPT (Haworth et al., 2006). This was largely based on the finding that a and Hunter, 1979). The resulting oncotic influx of water dominant-negative mutant of Atg5 increased BNIP3- structurally deforms and functionally impairs processes, induced apoptosis, but reduced BNIP3-induced autop- such as ATP production, that occur within mitochon- hagy (Hamacher-Brady et al., 2007). Further, in the drial matrix. The mitochondria then become net same study, mitochondria were found in the autophagic consumers (rather than producers) of ATP, and the cell vacuoles, leading the authors to conclude that the dies from suspension of minimal essential homeostatic sequestration of damaged mitochondria may have functions (Leist et al., 1997). MPT can be overlooked in protected hearts against ischemic injury. studies of programmed cardiomyocyte death if it is not Considering that BNIP3 can induce cell death in part specifically assayed, as markers of apoptosis are through MPT (Regula et al., 2002), which can induce invariably present because mitochondrial matrix swel- autophagy (Lemasters et al., 1998), one can envision a ling and disruption of the outer mitochondrial mem- model wherein BNIP3 induces autophagy by activating brane releases the normally sequestered apoptotic early events required for permeability transition. Mito- effectors, cytochrome c, apoptosis-inducing factor chondria that have undergone permeability transition (AIF) and endonuclease-G (EndoG) (Cregan et al., pore opening would then be selectively removed from 2004;Saelens et al., 2004). However, apoptosis may not the cell by mito-autophagy, postponing or delaying be the cause of death in these cells, as the complex induction of BNIP3-mediated apoptosis. Assuming a processes leading to apoptosis require ATP, and energy protective function for autophagy during ischemia deprivation in cardiomyocytes undergoing MPT can reperfusion, one may speculate that during hypoxia or interrupt apoptosis, thereby saving the cell from this ischemia reperfusion, early activation of BNIP3-induced particular form of programmed death. Such ATP- autophagy could actually protect against BNIP3-in- depleted cells that are not dead, but not really alive, duced apoptosis. This is notion is consistent with the have been called ‘zombie myocytes’ (Narula et al., 2001). fact that deletion of the C-terminal transmembrane Nevertheless, the cardiomyocyte is likely to be doomed domain of BNIP3 (DTM) does not translocate to the as the same ATP insufficiency that can prevent mitochondria, induce permeability transition or affect apoptosis also starves the calcium ATPases that pump autophagy or apoptosis in normoxic cells (Regula et al., calcium into the ER/SR. Unopposed calcium diffusion 2002;Hamacher-Brady et al., 2007). At present it is into the cytoplasm from ATP-deprived ER/SR produces unknown whether BNIP3 directly activates autophagy even more calcium-mediated opening of MPTPs, or if autophagy is causally related to BNIP3-induced ultimately resulting in cellular death from a cascade of mitochondrial perturbations. A recent report found that programmed necrosis. Rheb (Ras-homologue enriched in brain), a key factor Given the link between mitochondrial function and essential for mTOR activation, was inhibited by BNIP3 cell survival, there is considerable interest in the (Li et al., 2007), linking BNIP3 to the mTOR pathway involvement of mitochondrial perturbations induced and further explaining how BNIP3 can induce apopto- by BNIP3 and autophagy. Autophagy is a cellular sis, necrosis and autophagy. process for catabolism of intracellular proteins and organelles by lysosomal regulated pathways. The exact function of autophagy in programmed cell death is not fully understood (Kroemer and Levine, 2008). It first Summary and foremost represents an evolutionarily conserved survival pathway that includes a mechanism to selec- BNIP3 and NIX/BNIP3L are proximal regulators of tively discard damaged or irreversibly injured orga- multiple programmed cardiomyocyte death pathways. nelles, particularly mitochondria (mitophagy) (Zhang In the heart they are induced in a highly specific manner, et al., 2008). Early removal of damaged mitochondria and take weeks or longer to completely manifest their after injury could limit cell death by curtailing the adverse effects on left ventricular remodeling. Specificity mitochondrial regulated intrinsic death pathway. of induction and a prolonged period of action make The critical decision point and pathways that activate them attractive targets for therapeutic interdiction. programmed cell death (apoptosis) versus nonapoptotic Furthermore, their positions at the apices of apoptotic, programmed autophagy are not currently known, but autophagic and programmed necrotic pathways suggest are likely coupled to the metabolic status of the cell by that their inhibition or ablation may prevent replace- the mammalian target of rapamycin (mTOR). This ment of one form of cell death by another, as when AIF- notion is consistent with the activation of the mTOR mediated cell death or autophagy substitutes for pathway by positive growth signals and the concomitant apoptosis after caspase inhibition. The potential benefits inhibition of autophagy. In the context of the heart, of this approach have been demonstrated by gene autophagy has been demonstrated during hypoxia and ablation in infarcted or pressure overloaded mouse ischemia-reperfusion injury and postulated to be hearts. However, translating to the human condition mediated in part by BNIP3 (Hamacher-Brady et al., faces many challenges. If a pharmacological approach is 2006;Tracy et al., 2007). Interestingly, despite BNIP3’s to be used, then small molecular antagonists need to be reported ability to provoke apoptosis during hypoxic or developed that are sufficiently specific for BNIP3 and ischemic injury, in this study the autophagy by BNIP3 NIX/BNIP3L, and that do not produce untoward side

Oncogene Targeting cardiomyocyte death GW Dorn II and LA Kirshenbaum S165 effects in tissues (such as the hematopoietic system) that West had already made himself notorious through his depend on their proapoptotic and/or autophagic actions wild theories on the nature of death and the to maintain normal cellular homeostasis. A gene therapy possibility of overcoming it artificially. His views, approach could have many advantages, as dominant which were widely ridiculed by the faculty and by his inhibitory C-terminal truncated forms of both BNIP3 fellow-students, hinged on the essentially mechanistic and NIX/BNIP3L have been introduced into the heart nature of life;and concerned means for operating the and have abrogated apoptosis and autophagy. However, organic machinery of mankind by calculated chemi- although cardiomyocyte expression of truncated BNIP3 cal action after the failure of natural processes. In his and NIX/BNIP3L may be an acceptable means to experiments with various animating solutions, he had inhibit the parent endogenous factors, an optimal viral killed and treated immense numbers of rabbits, vector for cardiomyocyte gene therapy has not yet been guinea-pigs, cats, dogs, and monkeys, till he had described. Finally, it is interesting to speculate that become the prime nuisance of the college. Several inhibition of programmed cell death could find a role in times he had actually obtained signs of life in animals myocardial stem cell therapy, as it is developed to supposedly dead;in many cases violent signs but he regenerate damaged or dead myocardium. One of the soon saw that the perfection of his process, if indeed hurdles of repopulating dead myocardium with cells possible, would necessarily involve a lifetime of that can differentiate into functional cardiac myocytes is research. It likewise became clear that, since the that the stem or progenitor cells are necessarily same solution never worked alike on different organic introduced into a hostile tissue environment sufficiently species, he would require human subjects for further unfavorable to have contributed to the loss of cardio- and more specialized progress. myocytes in the first place. Such a milieu provides many powerful stimuli for programmed cell death. We trust that the outcome of current work will have a Preventing apoptosis, autophagy and programmed much more favorable impact than did that of the necrosis might therefore enhance engraftment and fictional Dr West. increase the efficiency of stem-cell-mediated myocardial regeneration. Conflict of interest On the basis of the rapidity with which the field has advanced, the above possibilities seem tantalizingly GW Dorn is currently receiving grant support from the within reach. However, even if the remaining challenges National Institutes of Health (NIH, Bethesda, MD, USA) can be overcome, success in mice does not necessarily (5P50HL077101;2R01HL059888;5R01HL080008; translate to man. HP Lovecraft appears especially 5R01HL087871;5P50HL077113). LA Kirschenbaum prescient when he wrote in Herbert West: Reanimator holds a patent for the NIP3 family of proteins (US (published in Home Brew in 1922): patent no. 7, 452,869B2).

References

Antman EM, Hand M, Armstrong PW, Bates ER, Green LA, Burton TR, Henson ES, Baijal P, Eisenstat DD, Gibson SB. (2005). Halasyamani LK et al. (2008). 2007 focused update of the The pro-cell death Bcl-2 family member, BNIP3, is localized to ACC/AHA 2004 guidelines for the management of patients with the nucleus of human glial cells: implications for glioblastoma ST-elevation myocardial infarction. Circulation 117: 296–329. multiforme tumor cell survival under hypoxia. Int J Cancer 118: Ashkenazi A, Dixit VM. (1999). Apoptosis control by death and decoy 1660–1669. receptors. Curr Opin Cell Biol 11: 255–260. Chen G, Cizeau J, Vande VC, Park JH, Bozek G, Bolton J et al. Baetz D, Regula KM, Ens K, Shaw J, Kothari S, Yurkova N et al. (1999). Nix and Nip3 form a subfamily of pro-apoptotic mitochon- (2005). Nuclear factor-kappaB-mediated cell survival involves drial proteins. J Biol Chem 274: 7–10. transcriptional silencing of the mitochondrial death gene BNIP3 Chen G, Ray R, Dubik D, Shi L, Cizeau J, Bleackley RC et al. (1997). in ventricular myocytes. Circulation 112: 3777–3785. The E1B 19K/Bcl-2-binding protein Nip3 is a dimeric mitochondrial Berry JJ, Hoffman JM, Steenbergen C, Baker JA, Floyd C, protein that activates apoptosis. J Exp Med 186: 1975–1983. Van Trigt P et al. (1993). Human pathologic correlation with Chen Z, Chua CC, Ho YS, Hamdy RC, Chua BH. (2001). PET in ischemic and nonischemic cardiomyopathy. J Nucl Med 34: Overexpression of Bcl-2 attenuates apoptosis and protects against 39–47. myocardial I/R injury in transgenic mice. Am J Physiol Heart Circ Birse-Archbold JL, Kerr LE, Jones PA, McCulloch J, Sharkey J. Physiol 280: H2313–H2320. (2005). Differential proﬁle of Nix upregulation and translocation Cizeau J, Ray R, Chen G, Gietz RD, Greenberg AH. (2000). The during hypoxia/ischaemia in vivo versus in vitro. J Cereb Blood Flow C. elegans orthologue ceBNIP3 interacts with CED-9 and CED-3 Metab 25: 1356–1365. but kills through a BH3- and caspase-independent mechanism. Bolli R, Marban E. (1999). Molecular and cellular mechanisms of Oncogene 19: 5453–5463. myocardial stunning. Physiol Rev 79: 609–634. Condorelli G, Morisco C, Stassi G, Notte A, Farina F, Sgaramella G Bouillet P, Metcalf D, Huang DC, Tarlinton DM, Kay TW, et al. (1999). Increased cardiomyocyte apoptosis and changes in Kontgen F et al. (1999). Proapoptotic Bcl-2 relative Bim required proapoptotic and antiapoptotic genes bax and bcl-2 during left for certain apoptotic responses, leukocyte homeostasis, and to ventricular adaptations to chronic pressure overload in the rat. preclude autoimmunity. Science 286: 1735–1738. Circulation 99: 3071–3078. Bruick RK. (2000). Expression of the gene encoding the proapoptotic Cregan SP, Dawson VL, Slack RS. (2004). Role of AIF in Nip3 protein is induced by hypoxia. Proc Natl Acad Sci USA 97: caspase-dependent and caspase-independent cell death. Oncogene 9082–9087. 23: 2785–2796.

Oncogene Targeting cardiomyocyte death GW Dorn II and LA Kirshenbaum S166 Danial NN, Korsmeyer SJ. (2004). Cell death: critical control points. cardiomyopathy of Galpha(q) transgenic mice. Circulation 108: Cell 116: 205–219. 3036–3041. Di Napoli P, Taccardi AA, Grilli A, Felaco M, Balbone A, Angelucci Henriquez M, Armisen R, Stutzin A, Quest AFG. (2008). Cell death D et al. (2003). Left ventricular wall stress as a direct correlate of by necrosis, a regulated way to go. Curr Mol Med 8: 187–206. cardiomyocyte apoptosis in patients with severe dilated cardiomyo- Hunt SA, Abraham WT, Chin MH, Feldman AM, Francis GS, pathy. Am Heart J 146: 1105–1111. Ganiats TG et al. (2005). ACC/AHA 2005 Guideline Update for the Diwan A, Dorn GW. (2007). Decompensation of cardiac hypertrophy: Diagnosis and Management of Chronic Heart Failure in the Adult: cellular mechanisms and novel therapeutic targets. Physiology a report of the American College of Cardiology/American Heart (Bethesda) 22: 56–64. Association Task Force on Practice Guidelines (Writing Committee Diwan A, Koesters AG, Odley AM, Pushkaran S, Baines CP, Spike to Update the 2001 Guidelines for the Evaluation and Management BT et al. (2007a). Unrestrained erythroblast development in NixÀ/À of Heart Failure): developed in collaboration with the American mice reveals a mechanism for apoptotic modulation of erythropoi- College of Chest Physicians and the International Society for Heart esis. Proc Natl Acad Sci USA 104: 6794–6799. and Lung Transplantation: endorsed by the Heart Rhythm Society. Diwan A, Krenz M, Syed FM, Wansapura J, Ren X, Koesters AG Circulation 112: e154–e235. et al. (2007b). Inhibition of ischemic cardiomyocyte apoptosis Kothari S, Cizeau J, McMillian-Ward E, Israels SJ, Bailes M, Ens K through targeted ablation of Bnip3 restrains postinfarction remo- et al. (2003). BNIP3 plays a role in hypoxic cell death in human deling in mice. J Clin Invest 117: 2825–2833. epithelial cells that is inhibited by growth factors EGF and IGF. Diwan A, Matkovich SJ, Yuan Q, Zhao W, Yatani A, Heller-Brown J Oncogene 22: 4734–4744. et al. (2008a). Endoplasmic reticular–mitochondrial crosstalk in Kroemer G, Levine B. (2008). Autophagic cell death: the story of a Nix-mediated cell death. J Clin Invest 118: 3870–3880. misnomer. Nat Rev Mol Cell Biol 9: 1004–1010. Diwan A, Wansapura J, Syed FM, Matkovich SJ, Lorenz JN, Kubasiak LA, Hernandez OM, Bishopric NH, Webster KA. (2002). Dorn GW. (2008b). Nix-mediated apoptosis links myocardial Hypoxia and acidosis activate cardiac myocyte death through ﬁbrosis, cardiac remodeling, and hypertrophy decompensation. the Bcl-2 family protein BNIP3. Proc Natl Acad Sci USA 99: Circulation 117: 396–404. 12825–12830. Dorn II GW. (2005). Physiologic growth and pathologic genes in Kubli DA, Quinsay MN, Huang C, Lee Y, Gustafsson AB. (2008). cardiac development and cardiomyopathy. Trends Cardiovasc Med Bnip3 functions as a mitochondrial sensor of oxidative stress during 15: 185–189. myocardial ischemia and reperfusion. Am J Physiol Heart Circ Dorn II GW. (2009). Apoptotic and non-apoptotic programmed Physiol 295: H2025–H2031. Cardiovasc Res cardiomyocyte death in ventricular remodelling. 81: Latif N, Khan MA, Birks E, O’Farrell A, Westbrook J, Dunn MJ et al. 465–473. (2000). Upregulation of the Bcl-2 family of proteins in end stage Foo RS, Chan LK, Kitsis RN, Bennett MR. (2007). Ubiquitination heart failure. J Am Coll Cardiol 35: 1769–1777. and degradation of the anti-apoptotic protein ARC by MDM2. Leist M, Single B, Castoldi AF, Ku¨ hnle S, Nicotera P. (1997). J Biol Chem 282: 5529–5535. Intracellular adenosine triphosphate (ATP) concentration: a switch Foo RS, Mani K, Kitsis RN. (2005). Death begets failure in the heart. in the decision between apoptosis and necrosis. J Exp Med 185: J Clin Invest 115: 565–571. 1481–1486. Foyouzi-Yousseﬁ R, Arnaudeau S, Borner C, Kelley WL, Tschopp J, Lemasters JJ, Nieminen AL, Qian T, Trost LC, Elmore SP, Nishimura Lew DP et al. (2000). Bcl-2 decreases the free Ca2+ concentration Y et al. (1998). The mitochondrial permeability transition in cell within the endoplasmic reticulum. Proc Natl Acad Sci USA 97: death: a common mechanism in necrosis, apoptosis, and autophagy. 5723–5728. Francis GS. (1998). Changing the remodeling process in heart failure: Biochim Biophys Acta 1366: 177–196. et al basic mechanisms and laboratory results. Curr Opin Cardiol 13: Li Y, Wang Y, Kim E, Beemiller P, Wang CY, Swanson J . (2007). 156–161. Bnip3 mediates the hypoxia-induced inhibition on mammalian Frazier DP, Wilson A, Graham RM, Thompson JW, Bishopric NH, target of rapamycin by interacting with Rheb. J Biol Chem 282: Webster KA. (2006). Acidosis regulates the stability, hydrophobi- 35803–35813. city, and activity of the BH3-only protein Bnip3. Antioxid Redox MacEwan DJ. (2002). TNF ligands and receptors—a matter of life and Signal 8: 1625–1634. death. Br J Pharmacol 135: 855–875. Galvez AS, Brunskill EW, Marreez Y, Benner BJ, Regula KM, Nakayama H, Chen X, Baines CP, Klevitsky R, Zhang X, Zhang H 2+ Kirschenbaum LA et al. (2006). Distinct pathways regulate et al. (2007). Ca - and mitochondrial-dependent cardiomyocyte proapoptotic Nix and BNip3 in cardiac stress. J Biol Chem 281: necrosis as a primary mediator of heart failure. J Clin Invest 117: 1442–1448. 2431–2444. Graham RM, Thompson JW, Wei J, Bishopric NH, Webster KA. Nam Y-J, Mani K, Ashton AW, Peng C-F, Krishnamurthy B, (2007). Regulation of Bnip3 death pathways by calcium, Hayakawa Y et al. (2004). Inhibition of both the extrinsic and phosphorylation, and hypoxia-reoxygenation. Antioxid Redox intrinsic death pathways through nonhomotypic death-fold inter- Signal 9: 1309–1315. actions. Mol Cell 15: 912. Grossman W, Jones D, McLaurin LP. (1975). Wall stress and Narula J, Arbustini E, Chandrashekhar Y, Schwaiger M. (2001). patterns of hypertrophy in the human left ventricle. J Clin Invest Apoptosis and the systolic dysfunction in congestive heart failure. 56: 56–64. Story of apoptosis interruptus and zombie myocytes. Cardiol Clin Hamacher-Brady A, Brady NR, Gottlieb RA, Gustafsson AB. (2006). 19: 113–126. Autophagy as a protective response to Bnip3-mediated apoptotic Narula J, Haider N, Virmani R, DiSalvo TG, Kolodgie FD, Hajjar RJ signaling in the heart. Autophagy 2: 307–309. et al. (1996). Apoptosis in myocytes in end-stage heart failure. N Hamacher-Brady A, Brady NR, Logue SE, Sayen MR, Jinno M, Engl J Med 335: 1182–1189. Kirshenbaum LA et al. (2007). Response to myocardial ischemia/ Nutt LK, Pataer A, Pahler J, Fang B, Roth J, McConkey DJ et al. reperfusion injury involves Bnip3 and autophagy. Cell Death Differ (2002). Bax and Bak promote apoptosis by modulating 14: 146–157. endoplasmic reticular and mitochondrial Ca2+ stores. J Biol Chem Haworth RA, Hunter DR. (1979). The Ca2+-induced membrane 277: 9219–9225. transition in mitochondria. II. Nature of the Ca2+ trigger site. Arch Olivetti G, Abbi R, Quaini F, Kajstura J, Cheng W, Nitahara JA et al. Biochem Biophys 195: 460–467. (1997). Apoptosis in the failing human heart. N Engl J Med 336: Hayakawa Y, Chandra M, Miao W, Shirani J, Brown JH, Dorn GW 1131–1141. et al. (2003). Inhibition of cardiac myocyte apoptosis improves Olivetti G, Capasso JM, Sonnenblick EH, Anversa P. (1990). cardiac function and abolishes mortality in the peripartum Side-to-side slippage of myocytes participates in ventricular wall

Oncogene Targeting cardiomyocyte death GW Dorn II and LA Kirshenbaum S167 remodeling acutely after myocardial infarction in rats. Circ Res 67: Siu PM, Bryner RW, Muriasits Z, Alway SE. (2005). Response of 23–34. XIAP, ARC, and FLIP apoptotic suppressors to 8 wk of treadmill Oltersdorf T, Elmore SW, Shoemaker AR, Armstrong RC, Auger DJ, running in rat heart and skeletal muscle. J Appl Physiol 99: 204–209. Belli BA et al. (2005). An inhibitor of Bcl-2 family proteins induces Sowter HM, Ratcliffe PJ, Watson P, Greenberg AH, Harris AL. regression of solid tumours. Nature 435: 677–681. (2001). HIF-1-dependent regulation of hypoxic induction of the cell Ranger AM, Zha J, Harada H, Datta SR, Danial NN, Gilmore AP death factors BNIP3 and NIX in human tumors. Cancer Res 61: et al. (2003). Bad-deﬁcient mice develop diffuse large B cell 6669–6673. lymphoma. Proc Natl Acad Sci USA 100: 9324–9329. Stauffer SR. (2007). Small molecule inhibition of the Bcl-X(L)-BH3 Ray R, Chen G, Vande VC, Cizeau J, Park JH, Reed JC et al. (2000). protein–protein interaction: proof-of-concept of an in vivo chemo- BNIP3 heterodimerizes with Bcl-2/Bcl-X(L) and induces cell death potentiator ABT-737. Curr Top Med Chem 7: 961–965. independent of a Bcl-2 homology 3 (BH3) domain at both mito- Syed F, Odley A, Hahn HS, Brunskill EW, Lynch RA, Marreez Y chondrial and nonmitochondrial sites. J Biol Chem 275: 1439–1448. et al. (2004). Physiological growth synergizes with pathological Regula KM, Ens K, Kirshenbaum LA. (2002). Inducible expression of genes in experimental cardiomyopathy. Circ Res 95: 1200–1206. BNIP3 provokes mitochondrial defects and hypoxia-mediated cell Teiger E, Than VD, Richard L, Wisnewsky C, Tea BS, Gaboury L death of ventricular myocytes. Circ Res 91: 226–231. et al. (1996). Apoptosis in pressure overload-induced heart Rizzuto R, Pozzan T. (2006). Microdomains of intracellular Ca2+: hypertrophy in the rat. J Clin Invest 97: 2891–2897. Tracy K, Dibling BC, Spike BT, Knabb JR, Schumacker P, Macleod molecular determinants and functional consequences. Physiol Rev KF. (2007). BNIP3 is an RB/E2F target gene required for hypoxia- 86: 369–408. induced autophagy. Mol Cell Biol 27: 6229–6242. Rubart M, Field LJ. (2006). Cardiac regeneration: repopulating the Yasuda M, Theodorakis P, Subramanian T, Chinnadurai G. (1998). heart. Annu Rev Physiol 68: 29–49. Adenovirus E1B-19K/BCL-2 interacting protein BNIP3 contains a Saelens X, Festjens N, Vande WL, van Gurp M, van Loo G, BH3 domain and a mitochondrial targeting sequence. J Biol Chem Vandenabeele P. (2004). Toxic proteins released from mitochondria 273: 12415–12421. in cell death. Oncogene 23: 2861–2874. Youle RJ, Strasser A. (2008). The BCL-2 protein family: opposing Sandoval H, Thiagarajan P, Dasgupta SK, Schumacher A, Prchal JT, activities that mediate cell death. Nat Rev Mol Cell Biol 9: Chen M et al. (2008). Essential role for Nix in autophagic 47–59. maturation of erythroid cells. Nature 454: 232–235. Yurkova N, Shaw J, Blackie K, Weidman D, Jayas R, Flynn B et al. Schmidt-Kastner R, Aguirre-Chen C, Kietzmann T, Saul I, Busto R, (2008). The cell cycle factor E2F-1 activates Bnip3 and the Ginsberg MD. (2004). Nuclear localization of the hypoxia-regulated intrinsic death pathway in ventricular myocytes. Circ Res 102: pro-apoptotic protein BNIP3 after global brain ischemia in the rat 472–479. hippocampus. Brain Res 1001: 133–142. Yussman MG, Toyokawa T, Odley A, Lynch RA, Wu G, Colbert MC Schweers RL, Zhang J, Randall MS, Loyd MR, Li W, Dorsey FC et al. (2002). Mitochondrial death protein Nix is induced in cardiac et al. (2007). NIX is required for programmed mitochondrial hypertrophy and triggers apoptotic cardiomyopathy. Nat Med 8: clearance during reticulocyte maturation. Proc Natl Acad Sci USA 725–730. 104: 19500–19505. Zhang H, Bosch-Marce M, Shimoda LA, Tan YS, Baek JH, Wesley JB Scorrano L, Oakes SA, Opferman JT, Cheng EH, Sorcinelli MD, et al. (2008). Mitochondrial autophagy is an HIF-1-dependent Pozzan T et al. (2003). BAX and BAK regulation of endoplasmic adaptive metabolic response to hypoxia. J Biol Chem 283: reticulum Ca2+: a control point for apoptosis. Science 300: 135–139. 10892–10903.

Oncogene