HeLa, Jurkat, and I9.2 cells were purchased from ATCC. Cells were maintained in Dulbecco Modified Eagle Medium (DMEM) or RPMI1640 (GIBCO, Grand Island, NY) supplemented with 10% (v/v) FBS. To transfect HeLa cells for confocal microscopy studies, exponentially growing HeLa cells were seeded on glass coverslips and transfected with pGFP, pGFP-casp8FL, or pGFP-casp8p41 plasmids using Lipofectamine™ transfection reagent (Invitrogen, Carlsbad, CA) following the manufacturer’s protocol. Jurkat and I9.2 T cells were transfected with 10 μg of pGFP control or pGFP-casp8p41 per 10⁷ cells using a square wave electroporator (BTX, San Diego, CA) at 320V for 20 msec. Where indicated, 20 μM ZVAD-fmk, IETD-fmk, or DMSO control were added immediately following transfection.

Cell viability was assessed after 8 hours of transfection using a cell titer glow luminescent assay (Promega, Madison, WI), which determines cell viability based upon quantitation of adenine triphosphate (ATP) as a marker of metabolically active cells, according to the manufacturer’s instructions.

Jurkat cell extracts were prepared, as previously described [5Nie Z, Phenix BN, Lum JJ, et al. HIV-1 protease processes procaspase 8 to cause mitochondrial release of cytochrome c, caspase cleavage and nuclear fragmentation Cell Death Differ 2002; 9(11): 1172-84.]. Briefly, cells (0.5 x 10⁶ cells/ml) were harvested by centrifugation at 1600 rpm for 5 minutes at 4ºC. The cell pellet was washed twice with ice-cold PBS (pH 7.4), followed by a single wash with ice-cold caspase buffer (20 mM PIPES, 100 mM NaCl, 10 mM DTT, 1 mM EDTA, 0.1% CHAPS, 250 mM sucrose, pH 7.2). After centrifugation, the cells were resuspended with two volumes of ice-cold complete caspase buffer, which was supplemented with protease inhibitors (1 mM PMSF, 10 μg/ml leupeptin, 2 μg/ml aprotinin), and then transferred to a 2 ml Dounce homogenizer. After sitting on ice for 15 minutes, the cells were disrupted with 50 strokes using a B-type pestle (Fisher Scientific Ltd, Nepean, ON, Canada). The nuclei were removed by centrifugation at 1000 x g for 10 minutes at 4ºC. To remove the mitochondria, the cytosols were further centrifuged twice at 20,000 x g for 15 minutes at 4ºC. The resultant suspension was incubated with either recombinant GST, GST-casp8p41 or recombinant caspase 8 (positive control) for 2 hours.

For Western blot analysis, 200 μg of cytosolic proteins were fractionated on 12% polyacrylamide gels, and transferred onto PVDF membranes (Millipore, Bedford, MA). The membranes were blotted with primary antibodies as follows: anti-caspase 9 (Medical & Biological Laboratories Co., Watertown, MA), anti-cytochrome c and anti-Hsp70 (Santa Cruz Biotechnology, Santa Cruz, CA). After washing, blots were incubated with HRP linked secondary antibodies and developed using SuperSignal (Pierce, Rockford, IL) following the manufacturer’s protocol.

Change in mitochondria transmembrane potential as a measure of apoptosis was determined in Jurkat and I9.2 cells transfected by GFP and GFPcasp8p41 after 8 hours of transfection. Briefly, 1 x 10⁶ cells were harvested, and stained with 250 nM TRME (Invitrogen, Carlsbad, CA) at 37ºC for 20 minutes. TMRE staining in the GFP-positive cells was analyzed by flow cytometry using a FACScan (Becton Dickinson Immunocytometry Systems, San Jose, CA). A decrease in TMRE staining was indicative of loss of mitochondrial transmembrane potential. Analysis was performed using CellQuest software (Becton Dickinson, San Jose, CA).

HeLa cells were transfected for 6 hours with GFP, GFP-casp8FL, or GFP-casp8p41 and then stained with 20nM MitoTracker^® Red (Invitrogen, Carlsbad, CA) for 45min at 37ºC. After staining, the cells were washed, fixed with 2% paraformaldehide, and coverslips mounted with DAPI containing Vectashield (Vector Laboratories, Burlingame, CA.). Laser-scanning confocal microscopy was performed using a Zeiss LSM-510 (Carl Zeiss Inc., Thornwood, NY).

Mitochondria were isolated from mouse liver (Balb/c, female, 6-8 weeks) by differential centrifugation and purified on Percoll gradient as described [8Belzacq AS, Vieira HL, Verrier F, et al. Bcl-2 and Bax modulate adenine nucleotide translocase activity Cancer Res 2003; 63(2): 541-6.]. For swelling and depolarization measurements, the mitochondria were diluted in the hypo-osmotic buffer (10 mM Tris-Mops, pH 7.4, 5 mM succinate, 200 mM sucrose, 1 mM Pi, 10 µM EGTA, 2 µM rotenone). The mitochondrial swelling was measured by the decrease in absorbance at 540 nm at 37°C. The depolarization of the mitochondria was measured by the rhodamine 123 (1 µM) fluorescence dequenching (excitation: 485 nm, emission: 535 nm, Molecular Probes) at 37°C. For cytochrome c and AIF release, mitochondria (40 µg of proteins) were incubated for 30 min at 37°C, and then centrifuged at 6800 g for 10 min at 4°C. Proteins contained in the supernatants were precipitated by methanol:chloroform [4:1, vol], resuspended in Laemmli sample buffer and separated by SDS-PAGE (15%) and transferred to PVDF membranes (Millipore, Hampshire, UK). After overnight incubation at 4°C with either anti-cytochrome c (BD Biosciences, Le Pont de Clay, France) or anti-AIF (Chemicon, Hampshire, UK), proteins were detected by using the ECL method according to the manufacturer's instructions (Amersham Pharmacia Biotech, Rockford, IL). When not indicated, products were from Sigma (St. Louis, MO).

The induced proximity model of casp8p41 cytotoxicity predicts that casp8p41, which contains two DEDs but no catalytic active site, acts as a scaffold to recruit procaspase 8, which leads to auto-activation of native procaspase 8. This model requires that cells contain native procaspase 8, and therefore, we tested whether casp8p41 was cytotoxic to I9.2 cells, which are deficient for procaspase 8. I9.2 cells and the parental Jurkat cell line were transfected with GFP casp8p41 or GFP empty vector control and analyzed for apoptosis within the GFP positive population. As measured by TUNEL staining (data not shown) or loss of mitochondrial transmembrane potential (Fig. 1A), the proportion of GFP-positive cells that acquired apoptotic features was not different between Jurkat and I9.2 cells, demonstrating that the absence of procaspase 8 is not sufficient to render cells resistant to the apoptotic effects of casp8p41. To confirm this result, we assessed whether caspase 8 inhibitor (Z-IETD Fmk) would alter cell death induced by casp8p41, and whether its effects would differ from those of a pan-caspase inhibitor (Z-VAD-Fmk). As expected, expression of casp8p41 resulted in a significant loss of viability that was not altered by Z-IETD-Fmk, further suggesting that casp8p41-induced death does not require caspase 8 activity (Fig. 1B). In contrast, the pan-caspase inhibitor, Z-VAD-Fmk, significantly inhibited death induced by casp8p41, confirming the caspase dependence of death induced by casp8p41.

Fig. (1) (A) Casp8p41 induced apoptosis is independent of caspase 8 activity. Jurkat and I 9.2 cells were transfected with plasmids containing GFP or GFP-casp8p41 for 8 hours. Cells were harvested and mitochondria transmembrane potential measured on the GFP-positive cells by TMRE staining. (B) Pool results of three replicates of Fig. (1A). (C) Jurkat cells were transfected with plasmids containing GFP or GFP-casp8p41, and immediately after transfection, cells were incubated with the indicated caspase inhibitors for 8 hours. Cells were harvested and cell viability determined using a cell titer glow luminescent assay.

Fig. (2) Casp8p41 acts on mitochondria to release cytochrome c. (A) Jurkat cells cytosolic extracts containing (+) or lacking (-) mitochondria were analyzed by Western blot for Hsp70 and cytochrome c to confirmed the purity of the fractions. (B) Cytosolic extracts containing or lacking mitochondria were incubated with GST alone, GST casp8p41, or with recombinant caspase 8 for 2 hours. Release of cytochrome c into the cytosol and cleavage of caspase 9 were analyzed by Western blot.

Fig. (3) Mitochondrial effects of casp8p41. (A) Characterization of casp8p41 induced swelling. Mouse liver isolated mitochondria were suspended in a hypoosmotic buffer, and their absorbance at 540 nm was recorded for 4000 sec at 37°C. Various doses of GST-casp8p41 were added in the presence of 10 µM Ca2+ (red line, 10 µM calcium, small dotted black line, 0.3 µM GST-casp8p41+10 µM calcium, large dotted black line, 1 µM GST-casp8p41 + 10 µM calcium, and black line, 3 µM GST-casp8p41 + 10 µM calcium). 100 µM Ca2+ (red dotted line) serves as a control of maximal swelling. The effect of GST (dark grey line) and GST-casp8p41 (grey line) was evaluated alone at the highest concentration (3 µM). Experiments were repeated three times. (B) Various inhibitors (CsA and Nfv) were added 2 min before the inducers, Ca2+ and GST-p41, at the indicated doses and their effect evaluated at 17 min. The effect of 3 µM GST-casp8p41 in the presence of 10 µM Ca2+ was normalized to 100%. Experiments were repeated three times and error bars represent SD. (C) Regulation of GST-casp8p41 induced Δγm loss. The Δγm loss was measured by the 123 Rhodamine dequenching method. Various inhibitors (CsA and Nfv) were added 2 min before the inducers, Ca2+ and GST-casp8p41, at the indicated doses. The effect of 3 µM GST-casp8p41 in the presence of 10 µM Ca2+ was normalized to 100%. Experiments were repeated three times and error bars represent SD. (D) GST-casp8p41 induced the cytochrome c, but not AIF release. Mouse liver isolated mitochondria were incubated in the presence of various agents and the release of cytochrome c, and AIF release was detected in the supernatant of mitochondria by Western-blot. Triton X-100 and alamethicin were used as positive controls of mitochondrial membrane permeabilization and Co. (untreated mitochondria) served as a negative control.

Fig. (4) Casp8p41 colocalizes with mitochondria. (A) HeLa cells were transfected with GFP, GFP-casp8p41, or GFP-casp8FL. After 6 hours of transfection, cells were stained with MitoTracker® red, and nuclei was counterstained with DAPI. Staining was visualized by confocal microscopy. The figure shows the overlay of the staining: GPF constructs (green), Mitotracker (red), and DAPI (blue). (B) Same as in A. Images of GFP-casp8p41 transfected HeLa cells are shown individually for each staining and as an overlay.

We have previously demonstrated that death induced by casp8p41 causes TUNEL positivity, phosphatidylserine externalization, and loss of mitochondrial transmembrane potential [1Nie Z, Bren GD, Vlahakis SR, et al. HIV-1 Protease Cleaves Procaspase 8 in vivo J Virol 2007; 81(13): 6947-56.]. Therefore, it is likely that casp8p41 killing involves mitochondrial depolarization, release of cytochrome c, and downstream activation of caspase 9. This possibility was formally tested using a cell-free system. Cytosols from Jurkat cells were prepared that contained mitochondria or ones that did not. These cytosols were then treated with recombinant GST-casp8p41 or GST alone or with recombinant active caspase 8. Cytosols were blotted for Hsp70 and cytochrome c to ensure purity (Fig. 2A), and then assessed for caspase 9 cleavage, as well as for cytochrome c release into the cytosolic fraction following treatment. GST treated cytosols had no evidence of caspase 9 processing nor cytochrome c release (Fig. 2B) Addition of recombinant active caspase 8 resulted in robust cytochrome c release and processing of procaspase 9 into a shorter active form in the mitochondria containing cytosols. GST-casp8p41 treatment caused both caspase 9 processing as well as cytochrome c release, which occurred only in the cytosols that contained mitochondria, but not the cytosols that did not.

We next questioned whether casp8p41 induces a mitochondrial depolarization directly or indirectly, for example, through a BH3 only Bcl2 family member (such as NOXA, Bid, etc.). Therefore, we tested whether casp8p41 had any effect upon isolated mouse liver mitochondria. Casp8p41 in the presence of 10 µM calcium (Ca²⁺) induced dose-dependent mitochondrial matrix swelling (Fig. 3A), whose kinetics and amplitude differed from the swelling induced by Ca²⁺, the prototypic ion for permeability transition induction [9Hunter DR, Haworth RA, Southard JH. Relationship between configuration, function, and permeability in calcium-treated mitochondria J Biol Chem 1976; 251(16): 5069-77.] (Fig. 3A). Ten micromolar of Ca²⁺ is considered as a suboptimal dose, since it did not induce swelling before 1800 sec.

Next, we used a pharmacological approach to analyze the underlying mechanisms. CsA, a potent permeability transition inhibitor, prevents the swelling and the inner membrane depolarization induced by the casp8p41, whereas Nelfinavir (Nfv), an adenine nucleotide transporter (ANT) pore inhibitor [10Weaver JG, Tarze A, Moffat TC, et al. Inhibition of adenine nucleotide translocator pore function and protection against apoptosis in vivo by an HIV protease inhibitor J Clin Invest 2005; 115(7): 1828-38.], promoted a minor inhibition of the swelling and had no inhibitory effect on the inner transmembrane potential (Fig. 3B,C). These results indicate that casp8p41 is able to directly induce permeability transition pore (PTP)-dependent mitochondrial swelling and depolarization, different from that induced by Ca²⁺ and probably independent of ANT. The requirement for suboptimal doses of Ca²⁺ has been previously observed for BH3 mitochondriotoxic peptides derived from Bax and suggests that casp8p41 needs subtle effects of the cation, e.g., on protein conformation, to exert its toxic effect.

The functional consequences of casp8p41-induced mitochondrial alterations were assessed in isolated liver mitochondria treated with casp8p41. In the presence of suboptimal doses of Ca²⁺, but not alone, casp8p41 triggered a strong mitochondrial release of cytochrome c from the intermembrane space (Fig. 3D), while apoptosis inducing factor (AIF), a pro-apoptotic caspase-dependent effector, was not released.

Having demonstrated that casp8p41 acts directly upon mitochondria in vitro, resulting in mitochondrial depolarization and cytochrome c release, we next assessed the relevance of these events in intact cells. HeLa cells were transfected with GFP alone, GFP full-length procaspase 8, or GFP casp8p41 and stained for mitochondria (red) and nuclear morphology (blue) (Fig. 4A). GFP control and GFP full-length procaspase 8 transfected cells exhibited diffuse GFP staining with no mitochondrial colocalization. In contrast, GFP casp8p41 transfected cells exhibited punctuate staining (Fig. 4A) that upon greater magnification colocalized with mitochondria (Fig. 4B).