Mitochondria and Apoptosis
In mitochondria of cells which are induced to undergo apoptosis, a disruption of the mitochondrial inner transmembrane potential (deltaPsi) can be detected. This deltaPSI disruption can be detected in many different cell types, irrespective of the apoptosis-inducing stimulus (e.g. dexamethasone, irradiation, etoposide, Fas/ TNF ligation, ceramide ...). Inhibitors of apoptosis (e.g. cycloheximide in case of dexamethasone induced apoptosis, Ac-YVAD-cmk in case of Fas-mediated apoptosis or Bcl-2 in case of ceramide-induced apoptosis ...) prevent the deltaPSI disruption. Common apoptotic features like DNA fragmentation, morphological changes of nuclei, production of reactive oxygen species and accumulation of Phosphatidylserine (PS) on the cell surface are invariably preceded by the fall or total disruption of deltaPSI. DeltaPSI disruption results from the so called Permeability Transition (PT): a sudden permeability increase of the inner mitochondrial membrane to solutes <15000 Da (protons, calcium ...).
Inhibitors of PT (for example Cyclosporin A derivatives or bongkrekic acid, BA) suppress the pe-apoptotic deltaPSI disruption and subsequent apoptosis. This suggests that the PT-mediated pre-apoptotic deltaPSI collapse constitutes an early and irreversible feature of the apoptotic effector phase (Kroemer et al., 1997, Immunology Today, 18: 44-51).
Cell free systems in which purified mitochondria and nuclei were incubated together in the absence (!) of cytosol provide more insight into the role of mitochondria in apoptotic events. Purified mitochondria (from untreated healthy cells) alone do not undergo PT and do not induce nuclear apoptosis in the added nuclei. In the presence of Atractyloside (Atr), mitochondria undergo PT and induce apoptosis (morphological and DNA fragmentation) in the added nuclei. Bongkrekic acid (BA) blocks the effect of Atr on mitochondria (PT) and apoptosis. Besides Atr, also other inducers of PT (e.g. ter-BHP = ter-Butylhydroperoxide, calcium ions, mClCCP = carbonyl cyanide m-chlorophenylhydrazone) induce apoptosis, and their PT stimulating and apoptotic inducing effect can be blocked by specific inhibitors. Also mitochondria isolated from cells stimulated to undergo apoptosis cause nuclear apoptosis in vitro. The strict correlation existing between mitochondrial PT and mitochondrial-mediated nuclear apoptosis suggests that PT is indeed a crucial event in the regulation of apoptosis induction by mitochondria. Interestingly, Atr-treated mitochondria from Bcl-2 overexpressing cells do not undergo PT and they fail to provoke nuclear apoptosis. In contrast, nuclei isolated from Bcl-2 transfected cells are not affected in their ability to undergo chromatin condensation or DNA fragmentation. Thus, Bcl-2 seems to confer its anti-apoptotic effect at its mitochondrial localization, not at its nuclear localization. It has to be mentioned that Bcl-2 is no universal inhibitor of PT since it prevents PT in mitochondria treated with Atr, m-ClCCP and ter-BHP, but not in mitochondria challenged by calcium ions or diamide. These observations are consistent in vivo and in vitro (Zamzami, 1996, J. Exp. Med., 183: 1533-44).
Mitochondria, treated with atractyloside, release a soluble factor into the supernatant that can induce nuclear apoptosis in isolated nuclei. This factor is a 57 kDa mitochondrial intermembrane protein and was called AIF (apoptosis inducing factor). Its amino-acid sequence has homology with bacterial ferredoxin and NADH-oxidoreductases (but not with known components of the respiratory chain). The primary transcript codes for a 67 kDa pro-peptide which contains a mitochondrial localization sequence (MLS) within its first 120 amino acids. This propeptide is processed during import into the mitochondrion, resulting in the mature 57 kDa AIF. AIF also contains putative NLS. Induction of apoptosis (e.g. with staurosporine) induces translocation of AIF into the cytosol and to the nucleus. In isolated nuclei, recombinant AIF induces chromatin condensation and fragmentation of DNA into 50 kb fragments (inhibitable by EDTA but not by zVAD.fmk) but it does not induce oligonucleosomal cleavage. In contrast to cytochrome c, AIF does not appear to require the presence of further cytosolic factors to induce apoptotic features in nuclei, though it does not possess intrinsic DNAse activity. Recombinant AIF, added to cytosol, induces also mitochondrial swelling (i.e. mitochondrial PT) and release of cytochrome c and caspase-9 (Susin et al, 1999, Nature, 397: 441-46; Susin et al., 1997, J. Exp. Med., 186: 25-37, Zamzani et al., 1996, J. Exp. Med., 183: 1533-1544).
dATP was found to initiate the apoptotic program (Caspase-3 activity, DNA fragmentation) in cytoplasmic extracts from normal growing (HeLa) cells. By chromatographic fractionation of the cytosol three protein factors were found to be essential for the apoptotic activity: these proteins were called APAFs (Apoptotic Protease Activating Factors). First Apaf-2 was identified to be identical with cytochrome c (Liu et al, 1996, Cell, 86: 147-157), then Apaf-1 was identified as a putative human CED-4 homolog (Zou et al., 1997, Cell, 90: 405-413), and finally Apaf-3 was shown to be identical to
Caspase-9 (Li et al., 1997, Cell, 91: 479-489). Together with dATP, these three proteins were shown to be sufficient (and essential) to reconstitute Caspase-3 activation in vitro.
It was suggested that Caspase 9 binds to Apaf-1 in a reaction triggered by
cytochrome c and dATP. Binding leads to cleavage of Caspase-9, converting it to an active protease, which activates Caspase-3.
Several apoptotic stimuli(e.g. activation by anti-Fas, TNF alpha, growth factor deprivation, DNA damage, etoposide and staurosporine) result in the release of cytochrome c from the mitochondria, and interestingly the cytochrome c release precedes the mitochondrial deltaPSI disruption. Bcl-2 and Bcl-XL block the release of cytochrome c and abort the apototic response (Yang et al., 1997, Science, 275: 1129-1132). There is evidence that deltaPSI disruption and Cytochrome c release belong to two independent apototic pathways (Kluck et al, 1997, Science, 275: 1132-1136).
Kroemer et al., 1997, Immunology Today, 18: 44-51, REVIEW;
Recently, a protein factor was identified that provides a link between Fas receptor activated Caspase-8 and the release of cytochrome c from the mitochondria (Luo et al., 1998, Cell, 94: 481-490 & Li et al.,1998, Cell, 94: 491-501). This factor is BID, a member of the
BH3 subfamily known to interact with Bcl-2 and Bax through its BH3 domain (Wang et al., 1996, Genes Dev., 10: 2859-2869). BID (26 kDa) is cleaved by Caspase-8 into fragments of 15 kDa (C-terminus) and 11 kDa (N-terminus). The 15 kDa fragment tBID (= truncated BID) contains the BH3 domain and, indeed, is the functional part of BID for cytochrome c release. It was shown that after Caspase-8 cleavage, the C-terminal BID fragment (tBID) translocates to the mitochondria. In a time course experiment after Fas stimulation of living cells (Jurkat) the sequential activation of Caspase-8, BID, Caspase-3 and DFF was shown.
The solution structure of BID, determined by NMR, suggests two modes of proapoptotic action: (1) BID can interact by its BH3 domain with the anti-apoptotic Bcl-XL and thus prevent the formation of the antiapoptotic complex between Bcl-XL and Apaf1. Truncation of BID by caspase-8 is supposed to enhance the heterodimerization with Bcl-XL; (2) BID contains the structural motifs for pore-formation, and after truncation it is potentially able to form selective ion-channels similar to BAX and may promote apoptosis in a way other than inhibiting Bcl-2 proteins and independent from its BH3 domain (Chou et al., 1999, Cell, 96: 615-624; McDonnell et al., 1999, Cell, 96: 625-634).
Susin et al., 1998, Biochimica et Biophysica Acta, 1366: 151-165, REVIEW;
Li et al., 1997, Cell, 91: 479-489;
Liu et al, 1996, Cell, 86: 147-157;
Zou et al., 1997, Cell, 90: 405-413;
Luo et al., 1998, Cell, 94: 481-490;