PerspectiveCell Biology

Cellular Demolition and the Rules of Engagement

See allHide authors and affiliations

Science  09 Feb 2007:
Vol. 315, Issue 5813, pp. 776-777
DOI: 10.1126/science.1138870

Execution of cellular suicide, or apoptosis, is indisputably under the dominion of one group of proteins, the Bcl-2 family. But within this family, three factions are at odds with each other as they rival to promote or block cell death. Anti-apoptotic members act to restrain pro-apoptotic members. More of an enigma has been the activity of “BH3-only proteins,” family members that share a single motif (a BH3 domain) with other Bcl-2 proteins. BH3-only proteins can bind to different subsets of the Bcl-2 family, but whether they assist cell suicide by activating pro-apoptotic members or inhibiting anti-apoptotic family members has been debated. On page 856 of this issue, Willis and colleagues (1) conclude that BH3-only proteins work solely by thwarting anti-apoptosis Bcl-2 proteins, thus settling the controversy. We now understand one way to start the cell's death machinery.

Bcl-2 regulates tissue health by inhibiting apoptosis. Bax (and its homolog Bak) has the opposite effect of promoting apoptosis despite a surprisingly similar three-dimensional structure. It is now accepted that anti-apoptotic family members, including Bcl-2, Bcl-xL, and Mcl-1, restrain Bax and Bak activity. Once activated, Bax and Bak induce permeabilization of the mitochondrial membrane, thereby allowing cytochrome c release and the activation of caspases (which catalyze protein degradation). This cascade of events ultimately leads to breakdown of the cell (see the figure).

Restraining orders.

Anti-apoptotic Bcl-2 family members block the translocation of pro-apoptotic Bax perhaps by directly binding and inhibiting Bax on mitochondria following a conformational change. BH3-only proteins promote apoptosis by binding to and inhibiting anti-apoptotic Bcl-2 family protein activity.

Bax was originally identified as a Bcl-2—binding protein and was thought to be inhibited by this interaction (2). However, contrary to this “rheostat model,” in healthy cells, Bax is not bound to Bcl-2 but exists as a monomer in the cytosol (3). How can anti-apoptotic Bcl-2 family members keep pro-apoptotic Bax in check without direct interaction? One solution may center on the activity of BH3-only proteins.

Cell-free models of Bax activation, which do not fully reflect the in vivo situation, show that pro-apoptotic Bax acts synergistically with peptides corresponding to the BH3 domain of only a few BH3-only proteins, notably those from Bid and Bim, to induce cytochrome c release from isolated mitochondria (46). Bid and Bim are thus referred to as “activator proteins.” This scenario is consistent with the simple model in which BH3 peptides bind to Bax and cause a conformational change and oligomerization of Bax, thereby activating the mitochondrial cell death pathway. How then do the other BH3-only proteins that do not bind to Bax (so-called “nonactivating” proteins) induce apoptosis? One possibility is that activator proteins could be sequestered by Bcl-2. The activators could be freed from this association if displaced by nonactivating BH3-only proteins. The released activators could then bind to Bax and trigger cell death (5). One problem of this model is that activator proteins (Bim and Bid) do not detectably bind to Bax. However, recent work shows that stabilizing the BH3 domain structure (α helices) of Bid and Bim allows the peptides to bind to Bax, keeping the theory plausible (7).

In contrast to the cell-free models, Willis et al. show that Bid and Bim are not at all required for apoptosis in vivo. Using mice lacking Bid and Bim, the authors show that nonactivating proteins still induce apoptosis in vivo. They also confirm that Bim and Bid fail to bind pro-apoptotic Bak and Bax. Thus, BH3-only proteins appear able to promote apoptosis solely by inhibiting Bcl-2 and anti-apoptotic homologs, to promote apoptosis, without displacing activator proteins.

What then activates Bax and Bak to set the mitochondrial cell death pathway in motion? The situation for Bak could be more straightforward than for Bax. In contrast to cytosolic and monomeric Bax, Bak is normally bound to mitochondria and constitutively bound to Mcl-1 and Bcl-xL, anti-apoptotic Bcl-2 family members (8, 9). Thus, simple relief from inhibition through binding of BH3-only proteins could unleash Bak to oligomerize and induce apoptosis as originally proposed in the rheostat model. However, mutant forms of Bak less able to bind Bcl-xL and Mcl-1 still require an unknown activation step to induce apoptosis (6), indicating that simply removing inhibition by Mcl-1 and Bcl-xL is not enough.

How cell death-promoting Bax is activated remains unknown. It translocates from the cytosol to the mitochondria during apoptosis, and overexpression of Bcl-2 somehow prevents this relocation. It may be that cellular stress alters the conformation of Bax to a semi-activated state; some of the Bax then moves to mitochondria, where anti-apoptotic Bcl-2 family members bind and inhibit it (see the figure). Similarly, nascent Bak oligomers on mitochondria are bound by Bcl-2 and thereby blocked from further oligomerization and from inducing apoptosis (10). These semiactivated forms of Bax and Bak could be relieved of Bcl-2 inhibition by BH3-only proteins to allow completion of translocation, oligomerization, and cytochrome c release.

Complicating the issue further are experiments indicating that Bcl-2 must change conformation on the mitochondria to inhibit Bax (11). Thus, activated forms of Bcl-2 may block activated forms of Bax, again consistent with the rheostat model except that inhibition occurs downstream of the elusive Bax and Bcl-2 activation steps. Two central helices of Bax and Bcl-2 become embedded deeply in the mitochondrial membrane during activation, thereby unfolding and undoubtedly destroying domains that harbor binding pockets for BH3 domains. Elucidating how BH3-only proteins relieve inhibition of membrane-inserted Bax by Bcl-2 requires fresh insight into the structures of the membrane-embedded forms.

Willis and colleagues make clear that BH3-only proteins can promote apoptosis exclusively by inhibiting anti-apoptotic Bcl-2 family members (see the Table). Moreover, particular BH3-only proteins engage particular anti-apoptotic Bcl-2 family members, allowing tissue-specific regulation and stress-specific cellular responses. Beyond the issue of Bax and Bak activation remains the central question of how they stimulate mitochondrial membrane permeabilization, a process that has defied intense research efforts so far.

View this table:


  1. 1.
  2. 2.
  3. 3.
  4. 4.
  5. 5.
  6. 6.
  7. 7.
  8. 8.
  9. 9.
  10. 10.
  11. 11.
View Abstract

Navigate This Article