Proteins of the Bcl-2 family are key regulators of apoptosis. The pro-apoptotic member Bak is a tail-anchored protein comprising nine alpha helices, with the C-terminal helix inserting as a transmembrane domain (TMD) into the mitochondrial outer membrane (MOM). Following an apoptotic stimulus, Bak undergoes a series of conformational changes to then homo-oligomerize and form pores in the MOM. Within the apoptotic pore, two interfaces (e.g., a BH3:groove and a6:a6) have been identified.  Whether a TMD:TMD interface also forms during apoptosis, and whether it contributes to pore formation, remains unclear.

In this study we employed a cysteine mutagenesis approach to determine whether a TMD:TMD interface is present in Bak before and/or after apoptotic signaling. Cysteine residues were substituted at each position throughout the TMD and Bak variants then stably expressed in bak^-/-bax^-/- mouse embryonic fibroblasts. Labeling of cysteines with a membrane-impermeable alkylating reagent verified that α9 forms a TMD in healthy cells. In addition, linkage of cysteines revealed an interaction between the TMDs after, but not before, apoptotic signaling. Linkage of both the TMD:TMD and BH3:groove interactions captured higher order oligomers, indicating that the interfaces are distinct. A TMD:TMD interface could still occur if the BH3:groove interface was blocked, but was not sufficient for pore formation. Lastly, substituting the Bak TMD with that from other MOM proteins failed to generate stable proteins, except for substitution with the Fis1 TMD. In this case, the TMDs could still be linked after apoptosis. Studies are underway to determine whether blocking the TMD:TMD interface inhibits pore formation, perhaps by preventing the BH3:groove interaction.