ATP depletion during hypoxia. Maintenance of glycolytic ATP production during
hypoxia prevented Bax translocation, while uncoupling of mitochondrial oxidative
phosphorylation resulted in Bax translocation even under normoxic conditions.
These results suggest that an ATP-dependent chaperone may sequester cytoplasmic
Bax. Khaled et al. (1999) linked Bax conformational changes to progressive cytosolic
alkalinization during cytokine withdrawal. Incubation of cytosolic extracts containing
Bax at pH > 7.8 for 15 min resulted in exposure of a cryptic NH 2 -terminal Bax
epitope and partitioning into Triton X-114 detergent, consistent with exposure of
the COOH-terminal hydrophobic domain.
4.
A finger in every potTo further complicate hypotheses of BH protein function, there are clear and
convincing effects of BH protein expression on nonlethal cell pathways and fates.
Proand antiapoptotic family members alter cell-cycle parameters in opposing
directions: killer proteins such Bax and Bad accelerate GO to S-phase transitions, and
survival proteins (Bcl-2, Bcl-xL) retard S-phase entry from a GO state (Brady et al.,
1996; Mazel et al., 1996; Vairo et al., 1996). Bcl-2 also accelerates withdrawal of
cycling cells from the cell cycle after growth factor deprivation (Vairo et al., 1996;
Lind et al., 1999). These effects have been noted in cell-transfection studies of
fibroblasts, solid tumor lines, and transgenic T and B cells (Pietenpol et al., 1994;
O’Reilly et al., 1996). T cells from Bcl-2-knockout mice enter S phase more rapidly
than lymphocytes from littermates, reinforcing the conclusion that modulation of
cell-cycle pathways are bona fide functions of the BH family rather than errant
findings in overexpression models. Huang et al. (1997) reported that a Tyr^28 mutant
Bcl-xL failed to produce the cell-cycle delay found with wild-type Bcl-xL, yet retained
wild-type survival function. This group proposed that the Tyr^28 residue in the BH4
domain of Bcl-xL was a site of phosphorylation and/or protein interaction. Uhlmann
et al. (1996) had previously reported a novel gain of function for Bcl-2 with deleted
amino acids 51–85, resulting in increased cell proliferation. Results such as these have
led some researchers to speculate that cell-cycle regulation is a primary function of
BH proteins, with apoptosis a secondary effect. Similar currents of thought have taken
hold in the broader field of apoptosis from time to time, despite the difficulty in
linking cell-cycle control to many instances of apoptosis (quiescent cells, for example).
It is more consistent, although perhaps less straightforward, to implicate
mitochondria in a bidirectional flow of information between cytoplasmic and nuclear
targets. This mechanism may be distinct from metabolic control of ATP levels, pH,
or redox state. For example, retrograde transduction is a defined signal transduction
pathway in yeast linking mitochondrial state to nuclear gene expression (Kirchman
et al., 1999; Sekito et al., 2000).
Linette et al. (1996) reported an association between Bcl-2-induced delay of GO-
S transition in T cells and the transcription factor NFAT (nuclear factor of activated
T cells). NFAT is required for expression of the delayed early genes IL-2, IL-3, and
MAKING SENSE OF THE BCL-2 FAMILY OF APOPTOSIS REGULATORS 67