and ATP. Thus, these results show that although the outer membrane has been
permeabilized to allow the passage of cytochrome c, the inner membrane remains
intact. It was subsequently shown by single-cell analysis that, in the presence of caspase
inhibitors, the release of cytochrome c results in a drop in ∆m, but, over the following
30–60 min, the potential recovered its original level (Waterhouse et al., 2001). The
results show that when the downstream caspase cascade is inhibited, the mitochondria
remain functional long after cytochrome c has been released, indicating intact
mitochondrial structures. In a study on mouse-liver mitochondria, Bid-induced
cytochrome c release was shown to be Bak dependent. No cytochrome c release was
detected in Bak-deficient mitochondria. Furthermore, Bid-induced cytochrome c
release was not blocked by the PTP inhibitor cyclosporin A, whereas Ca2+ induced
swelling, and subsequent cytochrome c release was blocked (Wei, M.C. et al., 2000).
Combined, these results strongly suggest that Bax, as well as Bak, can form ion
channels in the outer mitochondrial membrane and trigger cytochrome c release
independent of PTP or its components (Figure 5).8.3
Bax-VDAC channels
The involvement of VDAC in Bax-induced apoptosis has been suggested by some
studies. Shimizu et al. (1999) showed that Bax and Bak could induce cytochrome c
release from liposomes in which VDAC had been incorporated, whereas neither of
the proteins was active alone. They also showed that Bax was not able to induce
cytochrome c release from mitochondria isolated from a VDAC1-deficient yeast
strain, whereas cytochrome c was released from mitochondria isolated from wild-type
yeast. In an additional study, Bax and VDAC reconstituted in liposomes were shown
to form a new channel activity with a conductance 4–10 times larger then the
individual proteins (Shimizu et al., 2000a). Cytochrome c was able to pass through
the chimeric channel, but it could not pass through channels formed by the individual
proteins. Conversely, Sato et al. (2000) have shown that Bax oligomers are able to
form cytochrome c-conducting channels in liposomes. The difference might be due
to the quaternary structure of the Bax protein, since Bax monomers do not have a
channel-forming activity. These results also contradict the results by Priault et al.
(1999), who showed that Bax-induced cytochrome c release in yeast deficient of
VDAC was as efficient as in wild-type yeast. The release was prevented by Bcl-XL.
They further showed that Bax did not induce permeabilization of the inner
mitochondrial membrane. A study by Gross et al. (2000) also concluded that VDAC
was not required for Bax killing activity in yeast. In a recent study, antibodies against
VDAC were shown to prevent Bax-induced apoptosis but had no effect on Bid—or
Bik-triggered apoptosis. It was further shown that binding of Bax and Bak to red
blood cells was dependent on a plasma membrane VDAC protein (Shimizu et al.,
2001). Although these results suggest the involvement of VDAC in multidomain
protein activity, they do not show that VDAC is part of the channel-forming
structure. VDAC might function as a receptor protein in the mitochondriaMITOCHONDRIA IN APOPTOSIS INDUCTION 135membrane. This could explain the specific targeting of Bax to mitochondria during
apoptosis.