Role of Mitochondria and Caspases in Vitamin D-mediated Apoptosis of MCF-7 Breast Cancer Cells*

Vitamin D3 compounds are currently in clinical trials for human breast cancer and offer an alternative approach to anti-hormonal therapies for this disease. 1α,25-Dihydroxyvitamin D3(1α,25(OH)2D3), the active form of vitamin D3, induces apoptosis in breast cancer cells and tumors, but the underlying mechanisms are poorly characterized. In these studies, we focused on the role of caspase activation and mitochondrial disruption in 1α,25(OH)2D3-mediated apoptosis in breast cancer cells (MCF-7) in vitro. The effect of 1α,25(OH)2D3 on MCF-7 cells was compared with that of tumor necrosis factor α, which induces apoptosis via a caspase-dependent pathway. Our major findings are that 1α,25(OH)2D3 induces apoptosis in MCF-7 cells by disruption of mitochondrial function, which is associated with Bax translocation to mitochondria, cytochrome c release, and production of reactive oxygen species. Moreover, we show that Bax translocation and mitochondrial disruption do not occur after 1α,25(OH)2D3 treatment of a MCF-7 cell clone selected for resistance to 1α,25(OH)2D3-mediated apoptosis. These mitochondrial effects of 1α,25(OH)2D3 do not require caspase activation, since they are not blocked by the cell-permeable caspase inhibitor z-Val-Ala-Asp-fluoromethylketone. Although caspase inhibition blocks 1α,25(OH)2D3-mediated events downstream of mitochondria such as poly(ADP-ribose) polymerase cleavage, external display of phosphatidylserine, and DNA fragmentation, MCF-7 cells still execute apoptosis in the presence of z-Val-Ala-Asp-fluoromethylketone, indicating that the commitment to 1α,25(OH)2D3-mediated cell death is caspase-independent.

1␣,25-Dihydroxyvitamin D 3 (1␣,25(OH) 2 D 3 ), 1 the active form of vitamin D 3 , acts through the nuclear vitamin D 3 receptor (VDR) and is a potent negative growth regulator of breast cancer cells both in vitro and in vivo (1). A variety of synthetic vitamin D 3 analogs that induce mammary tumor regression in animals are now undergoing clinical trials in human patients (2,3). Our laboratory has shown that 1␣,25(OH) 2 D 3 induces morphological and biochemical markers of apoptosis (chromatin and nuclear matrix condensation, and DNA fragmentation) in breast cancer cells (MCF-7) (4,5); however, the precise mechanism by which 1␣,25(OH) 2 D 3 and the VDR mediate apoptosis is poorly understood.
To characterize the mechanisms of 1␣,25(OH) 2 D 3 -mediated apoptosis in breast cancer cells, we compared specific intracellular events in MCF-7 cells after treatment with 1␣,25(OH) 2 D 3 or tumor necrosis factor ␣ (TNF␣). TNF␣ was chosen as a positive control since this cytokine induces apoptosis in MCF-7 cells by a well defined pathway triggered by tumor necrosis factor receptor 1 (TNFR1), a cell surface death receptor. Death receptors contain homologous cytoplasmic regions termed "death domains," which transmit apoptotic signals through recruitment of adaptor molecules that activate caspases, a family of cysteine proteases involved in cell disassembly. The best characterized death receptors (Fas, TNFR1) use Fas-associated death domain and TNFR1-associated death domain adaptors to recruit and activate caspase-8 (6). Cleavage of specific substrates by caspases during apoptosis promotes the degradation of key structural proteins, including poly(ADP-ribose) polymerase (PARP), and lead to external display of phosphatidylserine (PS), DNA fragmentation, and cellular condensation (7).
Mitochondria play a central role in commitment of cells to apoptosis via increased permeability of the outer mitochondrial membrane, decreased transmembrane potential, release of cytochrome c and apoptosis-inducing factor, and production of reactive oxygen species (ROS) (8,9). Anti-apoptotic Bcl-2 family members, such as Bcl-2 and Bcl-X L , can block these mitochondrial events, whereas pro-apoptotic Bcl-2 family members, including Bax, can trigger these changes. For example, apoptotic signals induce conformational changes in Bax, which lead to exposure of the pro-apoptotic BH3 domain, and translocation to the mitochondria (10). The effects of pro-apoptotic Bcl-2 family members are achieved by both caspase-dependent and caspase-independent mechanisms (11,12).
Although the role of caspases in apoptosis triggered by cell surface death receptors such as TNFR1 has been well established, it is not clear if apoptosis triggered by nuclear receptors such as the VDR is mediated via similar caspase-dependent pathways. To probe the mechanisms whereby vitamin D 3 signaling modulates apoptosis in MCF-7 cells, we studied the effects of 1␣,25(OH) 2 D 3 on mitochondrial function and caspase activity using a cell-permeable inhibitor of caspase-related proteases (z-Val-Ala-Asp-fluoromethylketone, zVAD.fmk). In addition, the effects of 1␣,25(OH) 2 D 3 and TNF␣ on a vitamin D 3 -resistant variant of MCF-7 cells (MCF-7 D3Res cells) were examined to identify events that contribute to vitamin D 3 resistance (13,14). The MCF-7 D3Res cells do not undergo cell cycle arrest or apoptosis in response to 1␣,25(OH) 2 D 3 ; however, these cells retain sensitivity to other inducers of apoptosis such as TNF␣ and anti-estrogens (13).
Our results indicate that, although caspase inhibition can block some of the late stages of 1␣,25(OH) 2 D 3 -mediated apoptosis in MCF-7 cells, the commitment to cell death is caspaseindependent. These data implicate Bax distribution and mitochondrial disruption as critical caspase-independent events in 1␣,25(OH) 2 D 3 -mediated apoptosis of breast cancer cells.
Clonogenicity Assay-MCF-7 cells were incubated with 1␣,25(OH) 2 D 3 for 6 days or TNF␣ for 1 day in the presence or absence of 25 M zVAD.fmk. For 1␣,25(OH) 2 D 3 treatment, medium was replaced every 2 days. After treatments, cells were trypsinized, media and washes were pooled, and cells were pelleted by centrifugation and resuspended in fresh medium. The cells were seeded in 96-well plates at 5, 50, 500, and 2500 cells/well in 24 replicates. After 14 days, the cells were fixed and stained with crystal violet as described above, and clonogenic potential was estimated by counting positive wells (15).
Subcellular Fractionation-Cells were trypsinized, pooled together with media and washes containing floating cells, and pelleted by centrifugation at 500 ϫ g for 3 min at 4°C. Pellets were resuspended with 3 volumes of Buffer A (20 mM HEPES-KOH, pH 7.5, 10 mM KCl, 1.5 mM MgCl 2 , 1 mM EDTA, 1 mM EGTA, 10 mM benzamidine, 1 mM dithiothreitol, 250 mM sucrose, plus protease and phosphatase inhibitors), lysed with a Dounce homogenizer, and fractionated by differential centrifugation (16). Briefly, homogenates were centrifuged twice at 500 ϫ g for 5 min at 4°C, and the nuclear pellet was resuspended in Buffer A, sonicated 2 ϫ 10 s, and stored at Ϫ80°C in multiple aliquots. The supernatants were combined and further centrifuged at 10,000 ϫ g for 30 min at 4°C, and the resultant mitochondrial pellets were resuspended in Buffer A, sonicated, and stored at Ϫ80°C in multiple aliquots. The supernatant from the 10,000 ϫ g spin was further centrifuged at 100,000 ϫ g for 1 h at 4°C. The resulting supernatant was designated S100 (containing cytosol) and stored at Ϫ80°C in multiple aliquots. Protein concentrations were determined by the Micro BCA protein assay (Pierce).
Immunocytochemistry-MCF-7 cells grown on Lab-Tek II chamber slides (Fisher Scientific) were treated with ethanol vehicle, 100 nM 1␣,25(OH) 2 D 3 , or 2.5 ng/ml TNF␣ for 96 h (ethanol, 1␣,25(OH) 2 D 3 ) or 48 h (TNF␣) in the presence or absence of 25 M zVAD.fmk. The cells were fixed in 4% formaldehyde in PBS for 5 min at room temperature, permeabilized in methanol at Ϫ20°C for 5 min, and blocked overnight with PBS plus 1% BSA containing 0.02% sodium azide. The slides were then incubated with cytochrome c mouse monoclonal antibody (6H2.B4; PharMingen), diluted 1:100 in blocking buffer, for 3 h at 37°C in a humidified chamber. Slides were washed three times for 5 min each time with PBS, followed by incubation for 1 h at room temperature with anti-mouse secondary antibody conjugated to Alexa-488 (a photostable dye with spectral properties similar to fluorescein; Molecular Probes) diluted 1:50 in blocking buffer. Slides were washed three times for 5 min with PBS, incubated for 15 min at room temperature with 1 g/ml Hoechst 33258 (Sigma), washed five times for 5 min with PBS, rinsed with distilled H 2 O, and coverslips were applied with an antifade reagent. Fluorescence was detected using an Olympus AX70 microscope equipped with a Spot RT digital camera.
Flow Cytometry-For analysis of mitochondrial membrane potential and reactive oxygen species, cells harvested by trypsinization were pooled with media plus washes and pelleted by centrifugation. Cell suspensions (1 ϫ 10 6 cells) were incubated with 1 M tetramethylrhodamine ethyl ester (TMRE, Molecular Probes) in PBS containing 130 mM KCl to abolish the plasma membrane potential. After incubation for 10 min at 37°C, cells were washed once in PBS and then analyzed for TMRE red fluorescence by flow cytometry. Live cells rapidly and reversibly take up TMRE, and accumulation of the dye in mitochondria has been shown to be potential driven (17). For analysis of ROS, cell suspensions (5 ϫ 10 5 cells) were incubated with 4 M hydroethidine (HE, Molecular Probes) in PBS for 15 min at 37°C, and conversion of HE to ethidium by superoxide anion was analyzed by flow cytometry.
For analysis of DNA fragmentation, MCF-7 cells were harvested by trypsinization, collected by centrifugation, fixed in 2% formaldehyde in PBS, and permeabilized in 70% EtOH at Ϫ20°C. DNA strand breaks in cells undergoing apoptosis were indirectly labeled with bromodeoxyuridine by terminal transferase (Roche Molecular Biochemicals) and detected by FITC-conjugated monoclonal antibody to bromodeoxyuridine using the APO-BRDU kit according to manufacturer's protocol (Phoenix Flow Systems, San Diego, CA). Cells were counterstained with 5 g/ml propidium iodide (PI; Sigma) containing RNase A (Roche Molecular Biochemicals) for detection of total DNA, and two-color analysis of DNA strand breaks and cell cycle was achieved by flow cytometry.
For detection of PS externalization, 1 ϫ 10 6 cells were incubated in the presence of 10 g/ml annexin V-FITC (BioWhittaker, Walkersville, MD) and 5 g/ml PI in binding buffer (10 mM HEPES-KOH, pH 7.5, 140 mM NaCl, 2.5 mM CaCl 2 ) for 15 min at 37°C. Cells were washed twice in binding buffer, fixed in 2% formaldehyde in PBS for 15 min on ice, and then washed two more times in PBS plus 0.2% BSA. Pellets were resuspended in PBS plus 0.2% BSA and analyzed by flow cytometry. There were less than 1% PI ϩ cells in the population, and they were therefore excluded from analysis.
All flow cytometric analyses were performed on an Epics XL Flow Cytometer (Coulter Corp., Miami, FL) equipped with an argon laser. TMRE and HE were analyzed on FL3 using a 620-nm band pass filter. For DNA fragmentation analysis, FITC was analyzed on FL1 using a 520-nm band pass filter and PI was analyzed on FL3 with no color compensation. For PS externalization, annexin V-FITC was analyzed on FL1 and PI was analyzed on FL2 (580-nm band pass filter) using software color compensation. Data was modeled with the Multiplus AV software (Phoenix Flow Systems).
Caspase Activity Assay-Caspase activity was analyzed with the ApoAlert CPP32/caspase-3 assay kit according to manufacturer's protocol (CLONTECH, Palo Alto, CA). Briefly, after harvesting by trypsinization, 2 ϫ 10 6 cells were pelleted and stored at Ϫ20°C. For analysis, cell pellets were lysed, re-pelleted to remove cell debris, and supernatants were incubated with 50 M DEVD-AFC for 1 h at 37°C. The samples were analyzed using a fluorescence spectrophotometer with excitation ϭ 380 nm and emission ϭ 508 nm.
Statistical Evaluation-Data are expressed as mean Ϯ S.E. One-way analysis of variance was used to assess statistical significance between means. Differences between means were considered significant if p values less than 0.05 were obtained with the Bonferroni method using GraphPad Instat software (GraphPad Software, San Diego, CA).

RESULTS
Disruption of Mitochondrial Function, as Determined by Subcellular Localization of Bax and Cytochrome c, and ROS Generation, by 1␣,25(OH) 2 D 3 -To identify specific intracellular events involved in 1␣,25(OH) 2 D 3 -mediated apoptosis, we examined the signaling pathway downstream of the VDR in MCF-7 cells. Since disruption of mitochondrial function is a primary event in apoptosis that can be triggered by translocation of Bax to mitochondrial outer membrane, we first examined the subcellular distribution of Bax after 1␣,25(OH) 2 D 3 treatment of MCF-7 cells. As demonstrated in Fig. 1, Bax redistribution from the cytosolic to the mitochondrial fraction occurred after treatment with 1␣,25(OH) 2 D 3 or TNF␣ in MCF-7 cells (Fig. 1, top). Not only was Bax translocated to mitochondria, but both 1␣,25(OH) 2 D 3 -and TNF␣-treated cells exhibited cleavage of Bax from the intact 21-kDa protein to an 18-kDa fragment, an observation that is consistent with reports of Bax cleavage during drug-induced apoptosis (18). In both 1␣,25(OH) 2 D 3 -and TNF␣-treated cells, the Bax cleavage product was detected in mitochondrial, but not cytosolic, fractions, and others have proposed that Bax cleavage enhances homodimerization and its pro-apoptotic properties (19). These are the first data to implicate translocation and cleavage of Bax during 1␣,25(OH) 2 D 3 -induced apoptosis. To determine the relationship between Bax translocation and sensitivity to 1␣,25(OH) 2 D 3 -induced apoptosis, we examined the subcellular distribution of Bax in a vitamin D 3 -resistant variant of MCF-7 cells, which does not undergo apoptosis after treatment with 1␣,25(OH) 2 D 3 but retains sensitivity to other triggers, including TNF␣. 2 In the MCF-7 D3Res cells, 1␣,25(OH) 2 D 3 did not induce translocation or cleavage of Bax (Fig. 1, bottom). However, in 2 C. J. Narvaez and J. Welsh, unpublished data. Cells were plated at a density of 1 ϫ 10 5 cells/150-mm dish. Two days after plating, the cells were treated with ethanol or 1␣,25(OH) 2 D 3 and re-fed 2 days later. S100 fractions prepared at the indicated time points as described under "Experimental Procedures" were separated on SDS-PAGE, transferred to nitrocellulose, and immunoblotted with cytochrome c (7H8.2C12) antibody. B, cytosolic cytochrome c in MCF-7 D3Res cells. Cells were plated and treated with ethanol or 1␣,25(OH) 2 D 3 as described above, and 2.5 ng/ml TNF␣ was added 3 days before harvest. All the dishes were harvested on day 5 of treatment, and S100 fractions were prepared and immunoblotted as described above. The results are representative of at least three independent experiments. Translocation of Bax to mitochondria has been associated with release of cytochrome c, an event that is considered a commitment point for activation of apoptosis. As expected for viable cultures, no cytochrome c was detected in cytosolic fractions from MCF-7 cells treated with ethanol vehicle for up to 120 h ( Fig. 2A). In contrast, redistribution of cytochrome c from mitochondria to cytosol was detected within 48 h of 1␣,25(OH) 2 D 3 treatment in MCF-7 cells, before any morphological apoptotic features were detected. The absence of cytochrome oxidase in cytosolic fractions confirmed that extracts were free of mitochondrial contamination (data not shown). In MCF-7 D3Res cells, 1␣,25(OH) 2 D 3 did not trigger release of cytochrome c; however, cytochrome c was detected in cytosolic fractions after TNF␣ treatment of both MCF-7 and MCF-7 D3Res cell lines ( Fig. 2B; see also Fig. 7).
Long term exclusion of cytochrome c from the electron transport chain can lead to impairment of proton flow and generation of ROS due to incomplete reduction of molecular oxygen. Hence, mitochondrial generation of ROS in response to 1␣,25(OH) 2 D 3 and TNF␣ was examined by flow cytometry. Production of superoxide anion was indirectly assessed as oxidation of hydroethidine to ethidium, which fluoresces red upon DNA intercalation. As presented in Fig. 3, ROS production was enhanced by 1␣,25(OH) 2 D 3 in MCF-7, but not MCF-7 D3Res cells, whereas TNF␣ increased ROS production comparably in both cell lines. Time-course studies have demonstrated that ROS production is enhanced within 72 h of 1␣,25(OH) 2 D 3 treatment in MCF-7 cells (data not shown).

1␣,25(OH) 2 D 3 Mediates PARP Cleavage, PS Externalization, and DNA Fragmentation in a Caspase-dependent Manner-
Cytochrome c released into the cytosol is thought to trigger caspase activation downstream of mitochondria through binding to Apaf-1 and autoactivation of procaspase-9. Activated caspase-9 can activate additional effector caspases responsible for cell disassembly and events such as PS externalization, PARP cleavage, and DNA fragmentation. To determine the involvement of caspase-dependent proteolysis in 1␣,25(OH) 2 D 3 -mediated apoptosis, we examined whether a broad spectrum cell-permeable caspase inhibitor (zVAD.fmk) could abrogate the effects of 1␣,25(OH) 2 D 3 in MCF-7 cells.
Proteolytic activity associated with caspase activation was analyzed by three distinct methods: cleavage of an endogenous caspase substrate (PARP), and flow cytometric analysis of PS exposure and DNA fragmentation, which others have shown are provoked by caspases (7). As demonstrated in Fig. 4A, PARP was cleaved after treatment of MCF-7 cells with either 1␣,25(OH) 2 D 3 or TNF␣, and in both cases, cleavage was blocked by zVAD.fmk. Furthermore, both 1␣,25OH) 2 D 3 and TNF␣ induced PS externalization, which was also completely blocked by zVAD.fmk (Fig.  4B). Finally, the effects of 1␣,25(OH) 2 D 3 and TNF␣ on DNA fragmentation was assessed as terminal transferase-mediated incorporation of bromodeoxyuridine, detected by FITC-conjugated anti-bromodeoxyuridine antibody by flow cytometry (

. Effect of caspase inhibitor on DNA fragmentation after treatment with 1␣,25(OH) 2 D 3 or TNF␣ in MCF-7 cells.
Cells were plated and treated as described in Fig. 4, and DNA fragmentation was determined by flow cytometry as described under "Experimental Procedures." The results are representative of at least three independent experiments. exposure, PARP cleavage, and DNA fragmentation were blocked. In contrast, the caspase inhibitor effectively blocked cytochrome c release and ROS generation triggered by TNF␣ (Fig. 7, A and C).
To further probe mitochondrial function, the membrane potential-sensitive probe TMRE was used to detect mitochondrial membrane potential by flow cytometry. 1␣,25(OH) 2 D 3 treatment significantly enhanced the percentage of cells with reduced mitochondrial membrane potential, and zVAD.fmk did not block the decrease in mitochondrial membrane potential induced by 1␣,25(OH) 2 D 3 . TNF␣ treatment also enhanced the percentage of cells with decreased mitochondrial membrane potential; however, in contrast to 1␣,25(OH) 2 D 3 , the effect of TNF␣ was completely blocked by zVAD.fmk (Fig. 7B).
Subcellular localization of cytochrome c protein was examined by fluorescence microscopy to confirm the finding that cytochrome c release can proceed independently of caspase activation after 1␣,25(OH) 2 D 3 treatment. In Fig. 8, cytochrome c fluorescence (middle panels) is presented alongside phase contrast (top panels) and Hoechst nuclear staining (bottom panels) to compare cytochrome c localization in individual viable and apoptotic cells. In vehicle-treated control cells, apoptotic morphology was not present, and cytochrome c staining was restricted to punctate cytoplasmic regions, consistent with mitochondrial localization (Fig. 8). After treatment with 1␣,25(OH) 2 D 3 or TNF␣, apoptotic cells identified by phase contrast and Hoechst nuclear staining exhibited chromatin condensation, nuclear fragmentation, and cytosolic vacuolization. In these apoptotic cells, diffuse cytoplasmic cytochrome c staining was detected throughout the cell, which obscured the nuclei, consistent with redistribution of cytochrome c from mitochondria to cytoplasm (21). Consistent with the immunoblotting data (Fig. 7A), treatment with zVAD.fmk failed to prevent 1␣,25(OH) 2 D 3 -mediated cytochrome c release, as demonstrated by persistence of diffuse cytoplasmic cytochrome c staining in 1␣,25(OH) 2 D 3 -plus zVAD.fmk-treated cells. However, zVAD.fmk did prevent the morphological signs of apoptosis, including chromatin condensation and nuclear fragmentation, consistent with its ability to block PS redistribution, PARP cleavage, and DNA fragmentation induced by 1␣,25(OH) 2 D 3 .
Since zVAD.fmk did not block cytochrome c release or mitochondrial dysfunction induced by 1␣,25(OH) 2 D 3 , but did protect MCF-7 cells from morphological signs of apoptosis, including DNA fragmentation, we examined whether zVAD.fmktreated cells actually remained viable and/or maintained clonogenic potential. As shown in Fig. 9, both zVAD.fmk and zDEVD.fmk caspase inhibitors rescued MCF-7 cells from TNF␣-mediated cell death, as demonstrated by total cell num- bers, with zVAD.fmk offering the greater protection. However, neither zVAD.fmk nor zDEVD.fmk caspase inhibitors could protect MCF-7 cells from 1␣,25(OH) 2 D 3 -mediated apoptosis, since the reduction in total cell number was not abrogated by either inhibitor (Fig. 9). Finally, the clonogenic potential was determined after treatment of cells with 1␣,25(OH) 2 D 3 or TNF␣ in the presence or absence of zVAD.fmk followed by re-plating at limiting dilutions in fresh medium. In vehicle control-treated cultures, at least 1 out of 5 cells had the ability to produce clones. In TNF␣-treated cultures, clonogenicity was less than 1 out of 500 cells (f Ͻ 0.002) but in the presence of zVAD.fmk, the frequency of cells with clonogenic potential was significantly increased (f Ն 0.2). In 1␣,25(OH) 2 D 3 treated cultures, clonogenicity was less than 1 out of 50 cells (f Ͻ 0.02), and this was not enhanced in the presence of zVAD.fmk. DISCUSSION Here we report for the first time that 1␣,25(OH) 2 D 3 induces apoptosis in MCF-7 cells by disruption of mitochondrial function, which is accomplished by translocation of Bax to mitochondria, and increased permeability of the outer mitochondrial membrane. Of particular interest, neither Bax translocation nor the mitochondrial disruption is induced by 1␣,25(OH) 2 D 3 in a variant line of MCF-7 cells selected for resistance to 1␣,25(OH) 2 D 3 -mediated apoptosis (13). Collectively, these data implicate an essential role for mitochondrial signaling in the induction of apoptosis by 1␣,25(OH) 2 D 3 and identify the pro-apoptotic protein Bax as an important downstream target of the VDR in MCF-7 cells.
In addition to Bax translocation, we report that treatment of MCF-7 cells with 1␣,25(OH) 2 D 3 induces cytochrome c release and ROS generation, events that have been observed in cells induced to undergo apoptosis by overexpression of Bax (22,23). These data further support the concept that 1␣,25(OH) 2 D 3mediated apoptosis may be driven by Bax translocation. A role for the pro-apoptotic protein Bax in 1␣,25(OH) 2 D 3 -mediated apoptosis is consistent with previous studies that support a role for Bcl-2, the anti-apoptotic antagonistic partner of Bax, in mediating the effects of 1␣,25(OH) 2 D 3 on both breast and prostate cancer cells. Thus, 1␣,25(OH) 2 D 3 down-regulates Bcl-2 (24,25) and overexpression of Bcl-2 blocks 1␣,25(OH) 2 D 3 -induced apoptosis (26,27). Since Bcl-2 and Bax act antagonistically in the regulation of apoptosis, these data suggest that down-regulation of Bcl-2 in conjunction with translocation of Bax may be necessary for 1␣,25(OH) 2 D 3 -mediated apoptosis. Further studies will be necessary to identify the signals generated by 1␣,25(OH) 2 D 3 that induce Bax translocation to mitochondria. Since events upstream of Bax translocation to mitochondria in response to 1␣,25(OH) 2 D 3 are abrogated in the vitamin D 3 -resistant MCF-7 variant, comparison of early events in VDR signaling in these cells will be an important subject for future studies.
Examination of events downstream of mitochondria indicated that 1␣,25(OH) 2 D 3 induced features of apoptosis associated with caspase activation, such as PARP cleavage, PS exposure, and DNA fragmentation. To determine whether caspase activation was required for 1␣,25(OH) 2 D 3 -mediated apoptosis, we used the broad spectrum, cell-permeable caspase inhibitor zVAD.fmk. We observed that 1␣,25(OH) 2 D 3 signaling on mitochondria does not require caspase activation, since zVAD.fmk was unable to block 1␣,25(OH) 2 D 3 -induced cytochrome c release, decrease in mitochondrial membrane potential, or ROS production. Again, this is consistent with apoptosis driven by Bax translocation, which promotes cytochrome c release via caspase-independent pathways (28 -31). Our data also complement that of Mathiasen et al. (27), who reported that inhibition of caspase activity by overexpression of CrmA, a cowpox-derived caspase inhibitor, or caspase inhibitory peptides (Ac-DEVD-CHO, Ac-IETD-CHO, and zVAD.fmk) did not block vitamin D 3 -mediated growth arrest or apoptosis.
Although caspase inhibition did not block mitochondrial events induced by 1␣,25(OH) 2 D 3 , zVAD.fmk did block events downstream of mitochondria such as PARP cleavage, external display of PS, and DNA fragmentation. These findings are similar to reports of Bax-induced apoptosis, where caspase inhibitors had no effect on Bax-induced cytochrome c release or mitochondrial disruption, but prevented cleavage of nuclear and cytosolic substrates and DNA degradation (28 -31). However, our data conflict with that of Mathiasen et al. (27), who observed that zVAD.fmk did not block 1␣,25(OH) 2 D 3 -mediated DNA fragmentation in MCF-7 cells. This discrepancy may reflect differences in doses (1 M versus 25 M) or experimental design between the two studies. The lower dose of zVAD.fmk (1 M) used by Mathiasen et al. may have been insufficient to block mitochondrial-initiated caspases (caspase-9) (32).
The data presented in this paper indicate that 1␣,25(OH) 2 D 3 triggers both caspase-independent and caspase-dependent pathways in MCF-7 cells, and suggest that 1␣,25(OH) 2 D 3 can activate downstream effector caspases. Since cytochrome c release has been associated with autoactivation of procaspase-9, 1␣,25(OH) 2 D 3 may activate caspase-dependent pathways via cytochrome c release. However, no DEVDase activity was detected in 1␣,25(OH) 2 D 3 -treated cytoplasmic extracts, suggesting that other, possibly unidentified, effector caspases may be activated by 1␣,25(OH) 2 D 3 , or that caspase-dependent events occur at later stages in the apoptotic program. Although blocking caspase activation prevented some of the morphological aspects of 1␣,25(OH) 2 D 3 -mediated apoptosis, MCF-7 cells were not rescued from death by zVAD.fmk. This finding is consistent with reports that many cell types eventually die by a slower, non-apoptotic cell death if caspases are inactivated (8). These data support the concept that mitochondrial damage represents a cell death commitment step in the course of apoptosis induced by many stimuli (33), including 1␣,25(OH) 2 D 3 .
In summary, 1␣,25(OH) 2 D 3 mediates apoptosis of MCF-7 cells through mitochondrial signaling, which involves ROS generation, and is regulated by the Bcl-2 family of apoptotic regulators. Caspases act solely as executioners to facilitate 1␣,25(OH) 2 D 3 -mediated apoptosis, and caspase activation is not required for induction of cell death by 1␣,25(OH) 2 D 3 . Our data suggest distinct differences in the mechanisms of apoptosis induced by 1␣,25(OH) 2 D 3 and TNF␣, since inhibition of caspases was able to rescue MCF-7 cells from TNF␣-mediated, but not 1␣,25(OH) 2 D 3 -mediated, cell death. Although caspase inhibition blocked biochemical changes associated with caspase activation downstream of mitochondrial perturbations and loss of cytochrome c, the commitment of MCF-7 cells to 1␣,25(OH) 2 D 3mediated apoptosis is clearly caspase-independent.