Bcl-2 homology domain 3 (BH3)-only protein Bid is posttranslationally cleaved by caspase-8 into its truncated form (tBid) and couples with stress signals to the mitochondrial cell death pathway. However, the physiological relevance of Bid is not clearly understood. Hepatocyte-specific knockout (KO) of Bcl-xL leads to naturally-occurring apoptosis despite co-expression of Mcl-1, which shares a similar anti-apoptotic function. We generated Bcl-xL KO, Bcl-xL/Bid double KO, Bcl-xL/Bak double KO, Bcl-xL/Bax double KO, and Bcl-xL/Bak/Bax triple KO mice and found that hepatocyte apoptosis caused by Bcl-xL deficiency was completely dependent on Bak and Bax, and surprisingly on Bid. This indicated that, in the absence of Bid, Bcl-xL is not required for the integrity of differentiated hepatocytes, suggesting a complicated interaction between core Bcl-2 family proteins and BH3-only proteins even in a physiological setting. Indeed, a small but significant level of tBid was present in wild-type liver under physiological conditions. tBid was capable of binding to Bcl-xL and displacing Bak and Bax from Bcl-xL, leading to release of cytochrome c from wild-type mitochondria. Bcl-xL-deficient mitochondria were more susceptible to tBid-induced cytochrome c release. Finally, administration of ABT-737, a pharmacological inhibitor of Bcl-2/Bcl-xL, caused Bak/Bax-dependent liver injury, but this was clearly ameliorated with a Bid KO background.
Conclusion: Bid, originally considered to be a sensor for apoptotic stimuli, is constitutively active in healthy liver cells and is involved in the Bak/Bax-dependent mitochondrial cell death pathway. Healthy liver cells are addicted to a single Bcl-2-like molecule because of BH3 stresses, and therefore special caution may be required for the use of the Bcl-2 inhibitor for cancer therapy.