BCL-2 proteins control the intrinsic pathway of programmed cell death. Composed of anti- and pro-apoptotic members, their network of interactions forms a molecular switch that controls mitochondrial outer-membrane permeability. Apoptotic stimulation leads to BAK/BAX oligomerization and pore formation, yet the molecular details of this pivotal step remain poorly understood, and controversy persists regarding the activation mechanism. Here we use native mass spectrometry and kinetics to show that the homo-oligomerization of BAK and BAX is spontaneous in hydrophobic environments. This process is abrogated by hetero-dimerization of both BAK and BAX with the anti-apoptotic BCL-2 protein MCL-1. Pro-apoptotic BH3-only... More
BCL-2 proteins control the intrinsic pathway of programmed cell death. Composed of anti- and pro-apoptotic members, their network of interactions forms a molecular switch that controls mitochondrial outer-membrane permeability. Apoptotic stimulation leads to BAK/BAX oligomerization and pore formation, yet the molecular details of this pivotal step remain poorly understood, and controversy persists regarding the activation mechanism. Here we use native mass spectrometry and kinetics to show that the homo-oligomerization of BAK and BAX is spontaneous in hydrophobic environments. This process is abrogated by hetero-dimerization of both BAK and BAX with the anti-apoptotic BCL-2 protein MCL-1. Pro-apoptotic BH3-only proteins disrupt these hetero-dimers by binding competitively to MCL-1, releasing BAK/BAX for homo-oligomerization. Thus, we infer that their oligomeric states are thermodynamically favored at the membrane. Our approach provides the framework for future quantitative biophysical characterizations of the BCL-2 network, and advances our molecular understanding of apoptosis.
A number of models have been proposed to explain BAK/BAX activation and pore-formation (Chipuk & Green, 2008; Czabotar, Lessene, Strasser, & Adams, 2014; Kale, Osterlund, & Andrews, 2017; Peña-Blanco & García-Sáez, 2018). On one hand, ‘direct’ activation models stipulate that the trigger is a physical interaction between BH3-only proteins and BAK/BAX (Kim et al., 2006; Kuwana et al., 2005; Letai et al., 2002). Alternatively, ‘indirect’ activation models posit that BH3-only proteins do not engage BAK/BAX, but instead neutralize the anti-apoptotic BCL-2 proteins that prevent their oligomerization under non-apoptotic conditions (Uren et al., 2007; Willis et al., 2007). To study the BCL-2 network, we selected a minimal set of proteins encompassing all functional sub-classes. Both BAK and BAX (pore-forming) were investigated, MCL-1 was chosen as the model anti-apoptotic protein, and pro-apoptotic BH3-only proteins (PUMA and BID) were studied as 35-residue peptides (more details about the BCL-2 family can be found in Extended Data Fig. 1). We used native MS and biophysical techniques to probe the interactions between the components of the system, elucidate oligomeric stoichiometries, and investigate thermodynamic and kinetic aspects of these processes.
