Dying by the way you live: AIF vs. caspases in apoptosis of hypertrophied cardiomyocytes

Death of terminally differentiated cardiomyocytes is implicated in the development of various cardiac pathologies.^1,2 For example, following myocardial infarction, cell death in the infarct zone may occur by necrosis (i.e. unregulated cell death) and at least some cells in the border zone undergo apoptosis (i.e. programmed cell death). In the longer term, apoptosis occurs in the remote myocardium. The triggers probably include oxidative stresses, increased intracellular Ca²⁺ levels, and changes in intracellular pH. Preventing cardiomyocyte death may preserve cardiac function, so understanding the precise mode and mechanisms that operate is of great importance. These are likely to depend on the initiating apoptotic stimulus, the extracellular environment and, as highlighted in a study by Choudhury et al.³ in this issue, the intracellular environment.

The principal effectors of apoptosis are generally considered to be cysteine-dependent, aspartate-directed proteases–caspases.^1,4 Apoptosis via the extrinsic pathway is mediated by cell death receptors that activate the initiator caspase, caspase 8. Alternatively, in the intrinsic pathway, mitochondria release cytochrome c (and other pro-apoptotic factors), leading to activation of caspase 9. These initiator caspases activate effector caspases such as caspase 3 to execute the apoptotic programme. However, apoptosis is not mediated exclusively by caspases. Apoptosis-inducing factor (AIF) is released from the mitochondrial intermembrane space along with cytochrome c and translocates to the nucleus, where it promotes DNA fragmentation probably through endonuclease G (EndoG). Although caspases potentially act in concert with AIF for efficient apoptosis, AIF may suffice to induce cell death in a caspase-independent manner.^5,6

AIF translocates to the nucleus in hearts subjected to acute injury (ischaemia/reperfusion ex vivo⁷) or in chronic heart failure (induced by pressure overload or high-salt diet^8,9), suggesting that caspase-independent mechanisms contribute to cardiomyocyte death. Most studies of apoptosis start with healthy cells (akin to an acute injury model), but the response of cells in a disease situation (as in heart failure) may be very different. It is therefore important to establish the significance and role of caspases in both conditions. Choudhury et al.³ have done precisely this, demonstrating a shift from apoptosis mediated by AIF and caspases in normal cardiomyocytes to AIF-mediated, but caspase-independent apoptosis in hypertrophic cardiomyocytes (Figure 1). This probably reflects modification of the intracellular environment in hypertrophy, illustrated by upregulation of pro-apoptotic proteins (EndoG, Bax, Bak, HrtA2/Omi) and downregulation of protective proteins (Bcl2, Hsp70). To understand further the mechanisms associated with AIF-mediated apoptosis in hypertrophic cardiomyocytes, a more detailed understanding of AIF itself is required.

Figure 1

Mechanisms of apoptosis in normal vs. hypertrophic cardiomyocytes. Cellular stresses such as hypoxia/reoxygenation trigger cardiomyocyte apoptosis via the mitochondrial death pathway. In cardiomyocytes from unstressed ‘normal’ hearts, release of cytochrome c leads to caspase activation and caspases contribute to the apoptotic process. AIF is also released, potentially by a Ca²⁺-dependent protease. AIF translocates to the nucleus where it promotes stage I DNA fragmentation (<50 kb fragments) via EndoG. In other cells, caspases activate nucleases to drive stage II DNA fragmentation (∼120 bp fragments). AIF may also stimulate the phospholipid scramblase Scrm-1 to promote externalization of phosphatidylserine (PtdSer) at the plasma membrane, a signal for cell recognition and engulfment. The study by Choudhury et al.³ indicates that apoptosis in hypertrophic cardiomyocytes is caspase independent and, although cytochrome c is released from mitochondria, there is no change in caspase activity in these cells. The shift to AIF-mediated, caspase-independent cell death in hypertophic cardiomyocytes highlights the need for further studies of cell death in the disease context.

Open in new tabDownload slide

Although identified as a pro-apoptotic factor, AIF (like cytochrome c) is important in oxidative phosphorylation (OXPHOS).⁶ The 67 kDa AIF precursor (encoded by a nuclear gene) is targeted to the inner mitochondrial membrane and anchored by its N-terminus with the C-terminus facing the intermembrane space. Its precise function is not clear, but AIF is a flavoprotein with NADPH oxidase activity and, although not a direct component of the electron transport chain, loss of AIF causes substantial reduction of complex I with some reduction of complex III, compromising OXPHOS.¹⁰ AIF is thought to stabilize complex I,⁶ and, because complexes I, III, and IV exist in a higher-order structure (the respirasome),¹¹ there is a secondary effect on complex III. Loss of AIF renders cardiomyocytes more susceptible to oxidative stress-induced apoptosis,⁸ suggesting that AIF may be an antioxidant. However, compromised OXPHOS efficiency may increase oxidative stress and reduce the overall antioxidant capacity with the net effect of exacerbating cardiomyocyte injury.

The oxidoreductase element of AIF is disconnected from its role in apoptosis. During apoptosis, AIF is proteolysed (potentially by a mitochondrial calpain¹²) and freed from the inner mitochondrial membrane for translocation to the nucleus. Although AIF is expressed similarly in normal and hypertrophic cardiomyocytes, Choudhury et al.³ demonstrate greater nuclear translocation of AIF in hypertrophic cardiomyocytes in response to apoptotic stimuli. This could reflect the greater loss of AIF from mitochondria in hypertrophic cardiomyocytes following hypoxia/reoxygenation. Thus, hypoxia/reoxygenation may promote a greater increase in intracellular Ca²⁺ levels in hypertrophic cardiomyocytes, theoretically inducing higher calpain activity, leading to increased AIF release. Additionally, Hsp70 binds to AIF to retain it in the cytoplasm,¹³ so the reduced expression of Hsp70 in hypertrophic cardiomyocytes³ may also contribute to increased nuclear translocation of AIF.

Subtle differences in the degree of nuclear translocation of AIF may influence the rate of apoptosis of hypertrophic vs. normal cardiomyocytes. However, a more fundamental difference was detected by Choudhury et al.³ in that caspase 3 is activated in normal cardiomyocytes by apoptotic stimuli, with no activation in hypertrophic cardiomyocytes. The reason is not obvious since cytochrome c is released from both cell types. However, the levels of expression of components of the apoptosome signalling complex or endogenous inhibitors of the pathway may differ between the two cell types and, clearly, this is an area where global proteomics profiling may be applied. Nevertheless, the lack of any effect of caspase inhibition on the apoptotic response of hypertrophic cardiomyocytes (in contrast to normal cardiomyocytes) supports the concept that, in this scenario, apoptosis is caspase independent.

One remaining question is what are the executioners if caspases are not involved in apoptosis in hypertrophic cardiomyocytes? AIF may influence the activities of proteases other than caspases to drive apoptosis to completion, but there is little direct evidence for this. In other systems, AIF (with EndoG and cyclophilin A) promotes chromatin condensation and the first phase of DNA fragmentation into 50 kb lengths.⁵ Additional nucleases are probably required for the second phase of DNA fragmentation to ∼120 bp lengths and, in non-cardiomyocytes, this is caspase dependent. It remains to be seen whether DNA fragmentation in hypertrophic cardiomyocytes proceeds through the second phase. Irrespective of this, DNA fragmentation alone is insufficient for full apoptosis and other effectors of AIF must be involved. One interesting effector of the Caenorhabditis elegans homologue of AIF, WAH-1, is the phospholipid scramblase Scrm-1.¹⁴ This promotes externalization of phosphatidylserine (usually confined to the inner leaflet of the plasma membrane), a classic feature of apoptosis that serves as a signal for cells to be removed. Perhaps studies of this family of enzymes will provide further insight into mechanisms of cardiomyocyte apoptosis in normal and diseased states.

Conflict of interest: none declared.

References

Author notes