Recent evidence suggests that Bax induces apoptosis by moving from the cytosol to the mitochondrial membrane, independent of Bcl-2 or Bcl-X. Subcellular fractionation revealed that Bax is in the cytosol of non-apoptotic cells and in the membrane fraction of apoptotic cells.1, 2 In live cells, a green fluorescent protein-Bax chimera (GFP-Bax)3 was in the cytosol of healthy cells but on the mitochondrial membrane after induction of apoptosis. Relocation occurred early (within 2 hours after treatment with staurosporin or gamma-irradiation) and was followed by cell shrinkage (15-30 m after relocation) and nuclear fragmentation (~10 hours after relocation). Relocation required the mitochondrial-targeting C-terminus of Bax; this sequence appears to be masked in non-apoptotic cells and revealed during apoptosis.
|Figure 1. Bax moves from the cytosol to the mitochondria before cell shrinkage and nuclear fragmentation. a. Induction of apoptosis with staurosporin or gamma-irradiation; b. movement of bax from the cytosol to the membranes; c. cell shrinkage; d. nuclear fragmentation.|
Some confusion about the oligomerization and subcellular distribution of Bax appears to result from the effects of detergents on conformational changes.1, 4 Non-ionic detergents expose Bax determinants that mediate dimerization with Bcl-2 and Bcl-X. The zwitterionic detergent, CHAPS, did not expose these determinants. Thus, the association of Bax with other Bcl-2 family members and its subcellular distribution can be altered by detergents used to solubilize or permeabilize cells.
Supporting the interpretation that Bax may not bind to Bcl-2 or Bcl-X in live cells was the finding that the amount of Bax relocating to the mitochondria was independent of the amount of Bcl-X and/or Bcl-2 present in the mitochondrial membrane.3 In addition, Bax can induce apoptosis when its BH3 domain, the domain required for dimerization with Bcl-X, is deleted.5 Taken together these results suggest that Bax acts alone by inserting into the mitochondrial membrane during apoptosis.
The results raise questions about the oligomerization state of other Bcl-2 family members and about the possibility that other pro-apoptotic Bcl-2 family members change their intracellular location during the induction of apoptosis.