Genetics of Apoptosis

of Cyt.c into the cytosol. c-Abl is an obligate requirement for ER stressinduced Cyt.c
release because c-Abl-/- MEFs do not release Cyt.c and are resistant to apoptosis in
response to A23187, brefeldin A, and tunicamycin (Ito et al., 2001b). The mechanism
by which c-Abl exerts its effect is unknown but may involve the inhibition of an
antiapoptotic factor, such as Bcl-2, that prevents Cyt.c release. It also remains to be
determined how c-Abl is activated in response to ER stress and whether it functions
in conjunction (upstream or downstream) with Irel activated JNK signaling.

2.7

BAX, BAK, and BH3-only molecules

Generation of mice double deficient in the multidomain proapoptotic Bcl-2
members, BAX and BAK, has revealed what seems to be a striking dependence of ER
stress-induced apoptosis on mitochondria (Wei et al., 2001). Diverse death signals
converge upon the activation of ‘BH3-only’ Bcl-2 family members, which translocate
to mitochondria and induce homooligomerization of BAX and BAK, permeabilizing
the OMM and allowing Cyt.c to pass into the cytosol (Korsmeyer et al., 2000). Bcl-2
and Bcl-xL prevent the Cyt.c release either by directly binding and inactivating BH3-
only molecules (Cheng et al., 2001), or by blocking the actions of BAX and BAK
(Ferri and Kroemer, 2001). Effector caspase activation and apoptosis are completely
abrogated in Bax-/-, Bak-/- MEFs in response to all intrinsic apoptotic stimuli tested,
including tunicamycin, thapsigargin, and brefeldin A (Wei et al., 2001), presumably
because mitochondria fail to release Cyt.c. These findings are consistent with the
ability of Bcl-2 and Bcl-xL to block Cyt.c release and cell death in response to ER
stress agents (McCormick et al., 1997; McCullough et al., 2001). Caspase inhibitors
do not affect the ability of ER stress to release Cyt.c, suggesting that caspase-12 may
not be involved in this pathway (McCullough et al., 2001). Assuming that BAX and
BAK are not exerting a nonconventional function at the ER (see below), these findings
predict that ER stress-induced apoptosis proceeds through the activation of one or
more BH3-only molecules that, in turn, mediate cross-talk to the mitochondria. The
BH3-only protein BAD is a good candidate for this role. In resting cells, BAD is kept
in a phosphorylated state by several kinases, resulting in its sequestration to the cytosol
by 14–3-3 (Zha et al., 1996). Stimulation with thapsigargin or A23187, results in
Ca2+-dependent activation of calcineurin, which dephosphorylates BAD, resulting
in its translocation to mitochondria (Wang et al., 1999). A dominant negative mutant
of calcineurin inhibited thapsigargin-induced dephosphorylation and translocation
of BAD, and subsequent apoptosis. BAD may require the cooperation of other
prodeath signals elicited by ER stress to release Cyt.c, since dephosphorylated BAD
is found at the mitochondria within several hours of thapsigargin treatment, but cell
death does not occur until a day after treatment. In addition, numerous BH3-only
molecules are regulated at the transcriptional level (Vogelstein et al., 2000); therefore,
it remains possible that UPR transcriptional activates other BH3-only proteins that
participate in Cyt.c release.

104 GENETICS OF APOPTOSIS