BAK activation: a multiple step mechanism

Although the pro-apoptotic BCL-2 family proteins BAK and BAX play a key role in mitochondrial perturbation their transition from an inactive closed conformation to a membrane permeabilising pore remains unclear. I found that BAK in viable cells existed in a primed state which was characterized by an occluded N-terminus and an exposed BH3 domain. This conformation facilitated binding to the hydrophobic groove of BCL-XL and served as a checkpoint maintaining cell survival by preventing its further activation. Isolation of BAK by immunoprecipitation suggests that only a discrete portion is present in this primed conformation. Reconstitution of the BCL-XL BAK complex into a BAK/BAX null background rendered cells more sensitive to the BAD BH3 mimetic ABT-737 indicating that primed BAK is primarily involved in ABT-737 induced apoptosis. Primed BAK was displaced from BCL-XL by ABT-737 followed by an N-terminal conformational change and subsequent formation of dimers and higher molecular weight complexes. These sequential BAK activation steps occurred independently of cell fate and did not represent the rate limiting steps in BAK activation as a BAK BH3 mutant L78A lost proapoptotic function but still oligomerised as efficiently as wt BAK. Thus the transition from inactive BAK to a membrane permeabilising pore requires an additional activation step. I demonstrate that after 30 min of ABT-737 exposure primed BAK, after its displacement from BCL-XL, interacts with BIMEL reflecting the transient nature of this interaction. This interaction represented an additional step in BAK activation as BAK pro-apoptotic function was enhanced when BIMEL and BAK were co-expressed. However BIMEL did not induce the N-terminal conformational change nor oligomerisation of BAK and its interaction occurred downstream of both these events. In addition the pool of BIMEL involved in the further activation of BAK did not represent that sequestered by the antiapoptotic proteins BCL-2 and BCL-XL. These data suggest that BAK activation occurs in multiple steps in which a further activation event is required after the exposure of the BH3 domain, the N-terminal conformational change and the formation of high molecular weight complexes but prior to cytochrome c release. I propose that this event may be represented by interaction of N-terminal conformational changed/oligomerised BAK with BIMEL.