2.8
Antiapoptotic functions: membrane permeabilityAn intrinsic activity of Bcl-2-type survival proteins has been more difficult to envision.
Like Bax, a direct action of Bcl-2/Bcl-xL on the PTP has been proposed (Tsujimoto
and Shimizu, 2000). Shimizu et al. (1999) demonstrated an association of Bcl-xL and
VDAC, with a Bcl-xL-dependent inhibition of^14 C-sucrose or FITC-cytochrome c
uptake in VDAC-containing liposomes. In particular, a peptide comprising the NH 2 -
proximal BH4 domain of Bcl-2 or Bcl-xL was sufficient to inhibit the VDAC channel
(Shimizu et al., 2000b). The addition of Bax did not interfere with the inhibition of
VDAC by Bcl-xL. Some support for this mechanism is provided by electrophysiologic
studies of mitoplasts from MDA-MB-231 breast carcinoma cells transfected with
Bcl-2 or empty vector (Murphy et al., 2001). Under basal energized conditions or
after addition of micromolar Ca2+, a significantly lower frequency of patches with
PTP activity was observed in mitoplasts derived from Bcl-2-expressing cells. However,
the single-channel characteristics of the PTP were not altered by Bcl-2. In addition,
no novel channel activities were identified in mitochondrial membranes from Bcl-2
expressing cells in this or a second study by this group (Pavlov et al., 2001).
Careful mitochondrial fractionation studies by Thompson and coworkers
suggested, counterintuitively, that the primary mitochondrial action of Bcl-2 is
maintenance of the permeability of the outer membrane (Vander Heiden et al., 1999;
2000; 2001). Their studies of IL-3-deprived FL5.12 prolymphoid cells demonstrated
that Bcl-xL acted to preserve mitochondrial ATP generation, despite similar O 2
consumption and mitochondrial membrane potentials as in ATP-poor control cells.
ATP/ADP exchange in mitochondria was substantially reduced upon deprivation of
growth factor, but was restored to basal levels by removal of the outer mitochondrial
membrane. Bcl-xL enhanced mitochondrial ATP/ADP exchange in growth factor-
deprived cells. This effect may be mediated through VDAC, as incorporation of Bcl-
xL with VDAC in planar phospholipid membranes inhibited voltage gating of VDAC
channels at positive potentials. In this model, cytochrome c release would result
indirectly from osmotic swelling, secondary to membrane hyperpolarization
associated with the shift from state III (+ADP) to state IV (-ADP) respiration.
There have as yet been no attempts to reconcile experimentally these conflicting
effects of Bcl-2/Bcl-xL on VDAC pore function. Inhibition of VDAC was
demonstrated with permeability assays using liposome reconstitution systems with
recombinant Bcl-xL and purified or recombinant VDAC (Shimizu et al., 1999).
However, activation of VDAC was observed by electrophysiologic studies of planar
phospholipid membranes incorporating purified VDAC and recombinant full-length
Bcl-xL (Vander Heiden et al., 2001). In the former case, permeability to uncharged
substrates (sucrose and FITC-cytochrome c) was assessed, and in the latter,
conductance of monovalent ions (KCl). It would be interesting to determine whether
Bcl-2/Bcl-xL or their isolated BH4 domains alter the electrical properties of single
VDAC channels in the former model, as might be predicted for a direct interaction.
Conversely, the small (5%) decrease in single channel ionic conductance observed in
58 GENETICS OF APOPTOSIS