Autophagic cell death is thought to be a fail-safe mechanism of suicide that the cell switches to under conditions in which apoptosis does not work. A new study from the group of Hans-Uwe Simon indicates that the reverse situation can also occur — the cell can switch the autophagic system towards apoptosis.

Simon and colleagues noticed that overexpression of the autophagy-related gene-5 (ATG5) protein increased the cell's susceptibility to undergo apoptosis in response to cell-death stimuli, including certain anti-cancer drugs. By contrast, ATG5-deficient cells showed a strongly reduced sensitivity towards anti-cancer agents.

When Simon and co-workers investigated a possible role for ATG5 in apoptosis by using a model of spontaneous neutrophil apoptosis, they detected a truncated ATG5 product in apoptotic, but not in non-apoptotic, neutrophils. They suggested that ATG5 might be a target of proteases that are activated during apoptosis, and identified calpain-1 and -2 as ATG5-specific proteases. Indeed, incubation of a calpain inhibitor with neutrophils prevented both apoptosis and ATG5 cleavage. Overexpression of an ATG5 mutant that lacked the calpain cleavage site in ATG5-deficient cells did not increase sensitivity to anti-cancer drugs. The authors therefore concluded that calpain-mediated cleavage of ATG5 is a crucial pro-apoptotic event. In addition, the authors found that autophagic activity is not required for the apoptotic effect of truncated ATG5.

Simon and colleagues showed that truncated ATG5 induced cytochrome c release and apoptosis in cells that were not protected by high levels of the anti-apoptotic protein BCL2. ATG5 colocalized with a mitochondrial marker, and the appearance of truncated ATG5, but not of full-length ATG5, in the mitochondrial fraction coincided with cytochrome c release and caspase-3 cleavage. The authors also showed that truncated ATG5-mediated apoptosis could be blocked by a caspase inhibitor.

But how does truncated ATG5 trigger cytochrome c release? The authors proposed that truncated ATG5 might bind to BCL2-family members. They found that full-length and truncated ATG5 did not bind to BAX; however, only truncated ATG5 binds to BCL-XL in apoptotic, but not in non-apoptotic, cells. The authors speculated that truncated ATG5 promotes apoptosis by blocking the function of the anti-apoptotic 'survival' protein BCL-XL, which leads to BAX and/or BAK activation. Although this hypothesis requires further testing, it is supported by the observation that ATG5-induced cell death was blocked by high levels of BCL2.

...this finding has clearly important implications for anti-cancer therapies.

Although the details of this molecular link between autophagy and apoptosis need to be worked out further, this finding has clearly important implications for anti-cancer therapies.