Human immunodeficiency virus 1 envelope glycoprotein complex-induced apoptosis involves mammalian target of rapamycin/FKBP12-rapamycin-associated protein-mediated p53 phosphorylation.

Syncytia arising from the fusion of cells expressing a lymphotropic human immunodeficiency virus (HIV)-1-encoded envelope glycoprotein complex (Env) gene with cells expressing the CD4/CXCR4 complex undergo apoptosis through a mitochondrion-controlled pathway initiated by the upregulation of Bax. In syncytial apoptosis, phosphorylation of p53 on serine 15 (p53S15) precedes Bax upregulation, the apoptosis-linked conformational change of Bax, the insertion of Bax in mitochondrial membranes, subsequent release of cytochrome c, caspase activation, and apoptosis. p53S15 phosphorylation also occurs in vivo, in HIV-1(+) donors, where it can be detected in preapoptotic and apoptotic syncytia in lymph nodes, as well as in peripheral blood mononuclear cells, correlating with viral load. Syncytium-induced p53S15 phosphorylation is mediated by the upregulation/activation of mammalian target of rapamycin (mTOR), also called FKBP12-rapamycin-associated protein (FRAP), which coimmunoprecipitates with p53. Inhibition of mTOR/FRAP by rapamycin reduces apoptosis in several paradigms of syncytium-dependent death, including in primary CD4(+) lymphoblasts infected by HIV-1. Concomitantly, rapamycin inhibits p53S15 phosphorylation, mitochondrial translocation of Bax, loss of the mitochondrial transmembrane potential, mitochondrial release of cytochrome c, and nuclear chromatin condensation. Transfection with dominant negative p53 has a similar antiapoptotic action as rapamycin, upstream of the Bax upregulation/translocation. In summary, we demonstrate that phosphorylation of p53S15 by mTOR/FRAP plays a critical role in syncytial apoptosis driven by HIV-1 Env.

Mitochondrial release of caspase-2 and -9 during the apoptotic process.

The barrier function of mitochondrial membranes is perturbed early during the apoptotic process. Here we show that the mitochondria contain a caspase-like enzymatic activity cleaving the caspase substrate Z-VAD.afc, in addition to three biological activities previously suggested to participate in the apoptotic process: (a) cytochrome c; (b) an apoptosis-inducing factor (AIF) which causes isolated nuclei to undergo apoptosis in vitro; and (c) a DNAse activity. All of these factors, which are biochemically distinct, are released upon opening of the permeability transition (PT) pore in a coordinate, Bcl-2-inhibitable fashion. Caspase inhibitors fully neutralize the Z-VAD.afc-cleaving activity, have a limited effect on the AIF activity, and have no effect at all on the DNase activities. Purification of proteins reacting with the biotinylated caspase substrate Z-VAD, immunodetection, and immunodepletion experiments reveal the presence of procaspase-2 and -9 in mitochondria. Upon induction of PT pore opening, these procaspases are released from purified mitochondria and become activated. Similarly, upon induction of apoptosis, both procaspases redistribute from the mitochondrion to the cytosol and are processed to generate enzymatically active caspases. This redistribution is inhibited by Bcl-2. Recombinant caspase-2 and -9 suffice to provoke full-blown apoptosis upon microinjection into cells. Altogether, these data suggest that caspase-2 and -9 zymogens are essentially localized in mitochondria and that the disruption of the outer mitochondrial membrane occurring early during apoptosis may be critical for their subcellular redistribution and activation.

Structure-function analysis of the interaction between Bax and the cytomegalovirus-encoded protein vMIA.

The viral mitochondrial inhibitor of apoptosis (vMIA) encoded by the human cytomegalovirus exerts cytopathic effects and neutralizes the proapoptotic endogenous Bcl-2 family member Bax by recruiting it to mitochondria, inducing its oligomerization and membrane insertion. Using a combination of computational modeling and mutational analyses, we addressed the structure-function relationship of the molecular interaction between the protein Bax and the viral antiapoptotic protein vMIA. We propose a model in which vMIA exhibits an overall fold similar to Bcl-X(L). In contrast to Bcl-X(L), however, this predicted conformation of vMIA does not bind to the BH3 domain of Bax and rather engages in electrostatic interactions that involve a stretch of amino acids between the BH3 and BH2 domains of Bax and an alpha-helical domain located within the previously defined Bax-binding domain of vMIA, between the putative BH1-like and BH2-like domains. According to this model, vMIA is likely to bind Bax preferentially in its membrane-inserted conformation. The capacity of vMIA to cause fragmentation of the mitochondrial network and disorganization of the actin cytoskeleton is independent of its Bax-binding function. We found that Delta131-147 vMIA mutant, which lacks both the Bax-binding function and cell-death suppression but has intact mitochondria-targeting capacity, is similar to vMIA in its ability to disrupt the mitochondrial network and to disorganize the actin cytoskeleton. vMIADelta131-147 is a dominant-negative inhibitor of the antiapoptotic function of wild-type vMIA. Our experiments with vMIADelta131-147 suggest that vMIA forms homo-oligomers, which may engage in cooperative and/or multivalent interactions with Bax, leading to its functional neutralization.

GAPDH, a novel regulator of the pro-apoptotic mitochondrial membrane permeabilization.

Glyceraldehyde-3-phosphate dehydrogenase (GAPDH) is a pleiotropic enzyme that is overexpressed in apoptosis and in several human chronic pathologies. Here, we report that the protein accumulates in mitochondria during apoptosis, and induces the pro-apoptotic mitochondrial membrane permeabilization, a decisive event of the intrinsic pathway of apoptosis. GAPDH was localized by immunogold labeling and identified by matrix-assisted laser desorption/ionization-time of flight and nano liquid chromatography mass spectroscopy/mass spectroscopy in the mitochondrion of various tissues and origins. In isolated mitochondria, GAPDH can be imported and interact with the voltage-dependent anion channel (VDAC1), but not the adenine nucleotide translocase (ANT). The protein mediates a cyclosporin A-inhibitable permeability transition, characterized by a loss of the inner transmembrane potential, matrix swelling, permeabilization of the inner mitochondrial membrane and the release of two pro-apoptotic proteins, cytochrome c and apoptosis-inducing factor (AIF). This novel function of GAPDH might have implications for the understanding of mitochondrial biology, oncogenesis and apoptosis.

Mitochondria as therapeutic targets for cancer chemotherapy.

Mitochondria are vital for cellular bioenergetics and play a central role in determining the point-of-no-return of the apoptotic process. As a consequence, mitochondria exert a dual function in carcinogenesis. Cancer-associated changes in cellular metabolism (the Warburg effect) influence mitochondrial function, and the invalidation of apoptosis is linked to an inhibition of mitochondrial outer membrane permeabilization (MOMP). On theoretical grounds, it is tempting to develop specific therapeutic interventions that target the mitochondrial Achilles' heel, rendering cancer cells metabolically unviable or subverting endogenous MOMP inhibitors. A variety of experimental therapeutic agents can directly target mitochondria, causing apoptosis induction. This applies to a heterogeneous collection of chemically unrelated compounds including positively charged alpha-helical peptides, agents designed to mimic the Bcl-2 homology domain 3 of Bcl-2-like proteins, ampholytic cations, metals and steroid-like compounds. Such MOMP inducers or facilitators can induce apoptosis by themselves (monotherapy) or facilitate apoptosis induction in combination therapies, bypassing chemoresistance against DNA-damaging agents. In addition, it is possible to design molecules that neutralize inhibitor of apoptosis proteins (IAPs) or heat shock protein 70 (HSP70). Such IAP or HSP70 inhibitors can mimic the action of mitochondrion-derived mediators (Smac/DIABLO, that is, second mitochondria-derived activator of caspases/direct inhibitor of apoptosis-binding protein with a low isoelectric point, in the case of IAPs; AIF, that is apoptosis-inducing factor, in the case of HSP70) and exert potent chemosensitizing effects.

hTERT: a novel endogenous inhibitor of the mitochondrial cell death pathway.

hTERT is the catalytic subunit of the telomerase and is hence required for telomerase maintenance activity and cancer cell immortalization. Here, we show that acute hTERT depletion has no adverse effects on the viability or proliferation of cervical and colon carcinoma cell lines, as evaluated within 72 h after transfection with hTERT-specific small interfering RNAs (siRNAs). Within the same time frame, hTERT depletion facilitated the induction of apoptotic cell death by cisplatin, etoposide, mitomycin C and reactive oxygen species, yet failed to sensitize cells to death induction via the CD95 death receptor. Experiments performed with p53 knockout cells or chemical p53 inhibitors revealed that p53 was not involved in the chemosensitizing effect of hTERT knockdown. However, the proapoptotic Bcl-2 family protein Bax was involved in cell death induction by hTERT siRNAs. Depletion of hTERT facilitated the conformational activation of Bax induced by genotoxic agents. Moreover, Bax knockout abolished the chemosensitizing effect of hTERT siRNAs. Inhibition of mitochondrial membrane permeabilization by overexpression of Bcl-2 or expression of the cytomegalovirus-encoded protein vMIA (viral mitochondrial inhibitor of apoptosis), which acts as a specific Bax inhibitor, prevented the induction of cell death by the combination of hTERT depletion and chemotherapeutic agents. Altogether, our data indicate that hTERT inhibition may constitute a promising strategy for facilitating the induction of the mitochondrial pathway of apoptosis.

Oxidation of a critical thiol residue of the adenine nucleotide translocator enforces Bcl-2-independent permeability transition pore opening and apoptosis.

Mitochondrial membrane permeabilization is a critical event in the process leading to physiological or chemotherapy-induced apoptosis. This permeabilization event is at least in part under the control of the permeability transition pore complex (PTPC), which interacts with oncoproteins from the Bcl-2 family as well as with tumor suppressor proteins from the Bax family, which inhibit or facilitate membrane permeabilization, respectively. Here we show that thiol crosslinking agents including diazenedicarboxylic acid bis 5N, N-dimethylamide (diamide), dithiodipyridine (DTDP), or bis-maleimido-hexane (BMH) can act on the adenine nucleotide translocator (ANT), one of the proteins within the PTPC. ANT alone reconstituted into artificial lipid bilayers suffices to confer a membrane permeabilization response to thiol crosslinking agents. Diamide, DTDP, and BMH but not tert-butylhydroperoxide or arsenite cause the oxidation of a critical cysteine residue (Cys 56) of ANT. Thiol modification within ANT is observed in intact cells, isolated mitochondria, and purified ANT. Recombinant Bcl-2 fails to prevent thiol modification of ANT. Concomitantly, a series of different thiol crosslinking agents (diamide, DTDP, and BMH, phenylarsine oxide) but not tert-butylhydroperoxide or arsenite induce mitochondrial membrane permeabilization and cell death irrespective of the expression level of Bcl-2. These data indicate that thiol crosslinkers cause a covalent modification of ANT which, beyond any control by Bcl-2, leads to mitochondrial membrane permeabilization and cell death.

The thiol crosslinking agent diamide overcomes the apoptosis-inhibitory effect of Bcl-2 by enforcing mitochondrial permeability transition.

In several different cell lines, Bcl-2 prevents the induction of apoptosis (DNA fragmentation, PARP cleavage, phosphatidylserine exposure) by the pro-oxidant ter-butylhydroperoxide (t-BHP) but has no cytoprotective effect when apoptosis is induced by the thiol crosslinking agent diazenedicarboxylic acid his 5N,N-dimethylamide (diamide). Both t-BHP and diamide cause a disruption of the mitochondrial transmembrane potential delta psi(m) that is not inhibited by the broad spectrum caspase inhibitor z-VAD.fmk, although z-VAD.fmk does prevent nuclear DNA fragmentation and poly(ADP-ribose) polymerase cleavage in these models. Bcl-2 stabilizes the delta psi(m) of t-BHP-treated cells but has no inhibitory effect on the delta psi(m) collapse induced by diamide. As compared to normal controls, isolated mitochondria from Bcl-2 overexpressing cells are relatively resistant to the induction of delta psi(m) disruption by t-BHP in vitro. Such Bcl-2 overexpressing mitochondria also fail to release apoptosis-inducing factor (AIF) and cytochrome c from the intermembrane space, whereas control mitochondria not overexpressing Bcl-2 do liberate AIF and cytochrome c in response to t-BHP. In contrast, Bcl-2 does not confer protection against diamide-triggered delta psi(m) collapse and the release of AIF and cytochrome c. This indicates that Bcl-2 suppresses the permeability transition (PT) and the associated release of intermembrane proteins induced by t-BHP but not by diamide. To further investigate the mode of action of Bcl-2, semi-purified PT pore complexes were reconstituted in liposomes in a cell-free, organelle-free system. Recombinant Bcl-2 or Bcl-X(L) proteins augment the resistance of reconstituted PT pore complexes to pore opening induced by t-BHP. In contrast, mutated Bcl-2 proteins which have lost their cytoprotective potential also lose their PT-modulatory capacity. Again, Bcl-2 fails to confer protection against diamide in this experimental system. The reconstituted PT pore complex itself cannot release cytochrome c encapsulated into liposomes. Altogether these data suggest that pro-oxidants, thiol-reactive agents, and Bcl-2 can regulate the PT pore complex in a direct fashion, independently from their effects on cytochrome c. Furthermore, our results suggest a strategy for inducing apoptosis in cells overexpressing apoptosis-inhibitory Bcl-2 analogs.

Caspases disrupt mitochondrial membrane barrier function.

Mitochondrial intermembrane proteins including cytochrome c are known to activate caspases. Accordingly, a disruption of the mitochondrial membrane barrier function with release of cytochrome into the cytosol has been shown to precede caspase activation in a number of different models of apoptosis. Here, we addressed the question of whether caspases themselves can affect mitochondrial membrane function. Recombinant caspases were added to purified mitochondria and were found to affect the permeability of both mitochondrial membranes. Thus, caspases cause a dissipation of the mitochondrial inner transmembrane potential. In addition, caspases cause intermembrane proteins including cytochrome c and AIF (apoptosis-inducing factor) to be released through the outer mitochondrial membrane. These observations suggest that caspases and mitochondria can engage in a circular self-amplification loop. An increase in mitochondrial membrane permeability would cause the release of caspase activators, and caspases, once activated, would in turn increase the mitochondrial membrane permeability. Such a self-amplifying system could accelerate the apoptotic process and/or coordinate the apoptotic response between different mitochondria within the same cell.

Cytofluorometric detection of mitochondrial alterations in early CD95/Fas/APO-1-triggered apoptosis of Jurkat T lymphoma cells. Comparison of seven mitochondrion-specific fluorochromes.

It is commonly accepted that mitochondria undergo major changes early during the apoptotic process and that these alterations are critical for the death/life decision. Here we report that Jurkat T cell leukemia cells exhibit a perturbed incorporation of potential-sensitive fluorochromes. After 6 h of CD95/Fas/APO-1 crosslinking, a significant fraction of still normal-sized Jurkat cells exhibit a decreased incorporation of three different cationic lipophilic dyes commonly used for the quantitation of the mitochondrial transmembrane potential (deltapsi(m)): DiOC6(3), chloromethyl-X-rosamine, and tetramethylrhodaminemethylester. In contrast, upon induction of apoptosis, cells tend to exhibit an increase in the fluorescence obtained with rhodamine 123. The increased rhodamine 123 fluorescence into cells undergoing apoptosis is not affected by labeling in the presence of the protonophore m-chlorophenylhydrazone and thus cannot be attributed to a change in the deltapsi(m). Six hours after CD95 ligation no changes are found among normal-sized cells in the incorporation of mitotracker green and nonylacridine orange, which both measure mitochondrial mass. However, a fraction of cells exhibit an increased staining with the Apo2.7 antibody which detects a mitochondrial antigen generated during apoptosis. These findings underline the importance of using adequate fluorochromes for the quantitation of mitochondrial changes occurring during early apoptosis. Moreover, they cast doubts on those studies that, using rhodamine 123, hypothesized that apoptosis would be associated with a stable or increased deltapsi(m).

Hierarchical involvement of Bak, VDAC1 and Bax in cisplatin-induced cell death.

Following the screening of a battery of distinct small-interfering RNAs that target various components of the apoptotic machinery, we found that knockdown of the voltage-dependent anion channel 1 (VDAC1) was particularly efficient in preventing cell death induced by cisplatin (CDDP) in non-small cell lung cancer cells. Both the downregulation of VDAC1 and its chemical inhibition with 4,4'-diisothiocyanatostilbene-2,2'-disulfonic acid reduced the apoptosis-associated modifications induced by CDDP, including mitochondrial transmembrane potential dissipation and plasma membrane permeabilization. VDAC1 inhibition strongly reduced the CDDP-induced conformational activation of Bax, yet had no discernible effect on the activation of Bak, suggesting that VDAC1 acts downstream of Bak and upstream of Bax. Accordingly, knockdown of Bak abolished the activation of Bax, whereas Bax downregulation had no effect on Bak activation. In VDAC1-depleted cells, the failure of CDDP to activate Bax could be reversed by means of the Bcl-2/Bcl-X(L) antagonist ABT-737, which concomitantly restored CDDP cytotoxicity. Altogether, these results delineate a novel pathway for the induction of mitochondrial membrane permeabilization (MMP) in the course of CDDP-induced cell death that involves a hierarchical contribution of Bak, VDAC1 and Bax. Moreover, our data suggest that VDAC1 may act as a facultative regulator/effector of MMP, depending on the initial cytotoxic event.

PK11195, a ligand of the mitochondrial benzodiazepine receptor, facilitates the induction of apoptosis and reverses Bcl-2-mediated cytoprotection.

One critical step of the apoptotic process is the opening of the mitochondrial permeability transition (PT) pore leading to the disruption of mitochondrial membrane integrity and to the dissipation of the inner transmembrane proton gradient (Delta Psim). The mitochondrial PT pore is a polyprotein structure which is inhibited by the apoptosis-inhibitory oncoprotein Bcl-2 and which is closely associated with the mitochondrial benzodiazepine receptor (mBzR). Here we show that PK11195, a prototypic ligand of the 18-kDa mBzR, facilitates the induction of Delta Psim disruption and subsequent apoptosis by a number of different agents,including agonists of the glucocorticoid receptor,chemotherapeutic agents (etoposide, doxorubicin),gamma irradiation, and the proapoptotic second messenger ceramide. Whereas PK11195 itself has no cytotoxic effect, it enhances apoptosis induction by these agents. This effect is not observed for benzodiazepine diazepam, whose binding site in the mBzR differs from PK11195. PK11195 partially reverses Bcl-2 mediated inhibition of apoptosis in two different cell lines. Thus, transfection-enforced Bcl-2 overexpression confers protection against glucocorticoids and chemotherapeutic agents, and this protection is largely reversed by the addition of PK11195. This effect is observed at the level of Delta Psim dissipation as well as at the level of nuclear apoptosis. To gain insights into the site of action of PK11195, we performed experiments on isolated organelles. PK11195 reverses the Bcl-2-mediated mitochondrial retention of apoptogenic factors which cause isolated nuclei to undergo apoptosis in a cell-free system. Mitochondria from control cells, but not mitochondria from Bcl-2-overexpressing cells, readily release such apoptogenic factors in response to atractyloside, a ligand of the adenine nucleotide translocator. However, control and Bcl-2-overexpressing mitochondria respond equally well to a combination of atractyloside and PK11195. Altogether, these findings indicate that PK11195 abolishes apoptosis inhibition by Bcl-2 via a direct effect on mitochondria. Moreover, they suggest a novel strategy for enhancing the susceptibility of cells to apoptosis induction and, concomitantly, for reversing Bcl-2-mediated cytoprotection.

Potassium leakage during the apoptotic degradation phase.

The subcellular compartmentalization of ions is perturbed during the process of apoptosis. In this work, we investigated the impact of K+ on the apoptotic process in thymocytes and T cell hybridoma cells. Irrespective of the death-inducing stimulus (glucocorticoids, topoisomerase inhibition, or Fas-crosslinking), a significant K+ outflow was observed during apoptosis, as determined on the single-cell level by means of the K+-sensitive fluorochrome, benzofuran isophtalate. This loss of cytosolic K+ only occurs in cells that have completely disrupted their inner mitochondrial transmembrane potential. Inhibition of this mitochondrial transmembrane potential loss by Bcl-2 or by specific inhibitors acting on the mitochondrial permeability transition pore (bongkrekic acid, cyclosporin A) prevents K+ leakage. K+ drops at the same stage at which cells expose phosphatidylserine residues on the outer leaflet of the membrane and reduce the levels of nonoxidized glutathione, but before they hyperproduce reactive oxygen species, undergo massive Ca2+ influx, shrink, and lyse. In a cell-free system of apoptosis, isolated nuclei exposed to the supernatant of mitochondria that have undergone permeability transition only manifest chromatinolysis when the K+ concentration is lowered from physiologic to apoptotic levels. Accordingly, massive DNA fragmentation causing subdiploidy is confined to cells that have undergone K+ leakage. Together, these data point to the step-wise acquisition of membrane dysfunction in apoptosis and indicate an important role for the disruption of normal K+ homeostasis in apoptotic degradation. Derepression of endonucleases due to low K+ concentrations may be a decisive prerequisite for end-stage DNA fragmentation.

A cytofluorometric assay of nuclear apoptosis induced in a cell-free system: application to ceramide-induced apoptosis.

Purified nuclei exposed to apoptogenic factors in vitro undergo morphological and biochemical changes in chromatin organization. Most cell-free models of nuclear apoptosis are based on the quantitation of endonuclease-mediated DNA fragmentation on agarose gels or on the changes of nuclear morphology revealed by the DNA-intercalating fluorochrome 4'-6-diamidino-2-phenylindole dihydrochloride. In this work we develop a cytofluorometric system for the accurate quantitation of nuclear DNA loss. This system has been used to determine the conditions of nuclear apoptosis induced by apoptosis-inducing factor (AIF) contained in the supernatant of mitochondria induced to undergo permeability transition. AIF can provoke significant nuclear DNA loss in < or = 5 min, acts over a wide pH range (pH 6 to 9), and resists cysteine protease inhibitors such as iodoacetamide and N-ethylmaleimide. Moreover, we applied this system to the question of how the proapoptotic second messenger ceramide would induce apoptosis in vitro: via a direct effect on nuclei, a direct effect on mitochondria, or via indirect mechanisms? Our data indicate that ceramide has to activate yet unknown cytosolic effectors that, in the presence of mitochondria, can induce nuclear apoptosis in vitro.

Mitochondrio-nuclear translocation of AIF in apoptosis and necrosis.

Apoptosis inducing factor (AIF) is a novel apoptotic effector protein that induces chromatin condensation and large-scale ( approximately 50 kbp) DNA fragmentation when added to purified nuclei in vitro. Confocal and electron microscopy reveal that, in normal cells, AIF is strictly confined to mitochondria and thus colocalizes with heat shock protein 60 (hsp60). On induction of apoptosis by staurosporin, c-Myc, etoposide, or ceramide, AIF (but not hsp60) translocates to the nucleus. This suggests that only the outer mitochondrial membrane (which retains AIF in the intermembrane space) but not the inner membrane (which retains hsp60 in the matrix) becomes protein permeable. The mitochondrio-nuclear redistribution of AIF is prevented by a Bcl-2 protein specifically targeted to mitochondrial membranes. The pan-caspase inhibitor Z-VAD. fmk does not prevent the staurosporin-induced translocation of AIF, although it does inhibit oligonucleosomal DNA fragmentation and arrests chromatin condensation at an early stage. ATP depletion is sufficient to cause AIF translocation to the nucleus, and this phenomenon is accelerated by the apoptosis inducer staurosporin. However, in conditions in which both glycolytic and respiratory ATP generation is inhibited, cells fail to manifest any sign of chromatin condensation and advanced DNA fragmentation, thus manifesting a 'necrotic' phenotype. Both in the presence of Z-VAD. fmk and in conditions of ATP depletion, AIF translocation correlates with the appearance of large-scale DNA fragmentation. Altogether, these data are compatible with the hypothesis that AIF is a caspase-independent mitochondrial death effector responsible for partial chromatinolysis.

Palmitate induces apoptosis via a direct effect on mitochondria.

The fatty acid palmitate can induce apoptosis. Here we show that the palmitate-induced dissipation of the mitochondrial transmembrane potential (Delta Psi m), which precedes nuclear apoptosis, is not prevented by inhibitors of mRNA synthesis, protein synthesis, caspases, or pro-apoptotic ceramide signaling. However, the mitochondrial and nuclear effects of palmitate are inhibited by overexpression of anti-apoptotic proto-oncogene product Bcl-2 and exacerbated by 2-bromo-palmitate as well as by carnitine. The cytoprotective actions of Bcl-2, respectively, is not antagonized by etomoxir, an inhibitor of carnitine palmitoyl transferase 1 (CPT1), suggesting that the recently described physical interaction between CPT1 and Bcl-2 is irrelevant to Bcl-2-mediated inhibition of palmitate-induce apoptosis. When added to purified mitochondria, palmitate causes the release of soluble factors capable of stimulating the apoptosis of isolated nuclei in a cell-free system. Mitochondria purified from Bcl-2 over-expressing cells are protected against the palmitate-stimulated release of such factors. These data suggest that palmitate causes apoptosis via a direct effect on mitochondria.