AKT/Protein Kinase B Regulation of BCL Family Members during Oxysterol-induced Apoptosis*

Cells of the vasculature, including macrophages, smooth muscle cells, and endothelial cells, exhibit apoptosis in culture upon treatment with oxidized low density lipoprotein, as do vascular cells of atherosclerotic plaque. Several lines of evidence support the hypothesis that the apoptotic component of oxidized low density lipoprotein is one or more oxysterols, which have been shown to induce apoptosis through the mitochondrial pathway. Activation of the mitochondrial pathway of apoptosis is regulated by members of the BCL family of proteins. In this study, we demonstrate that, in the murine macrophage-like cell line P388D1, oxysterols (25-hydroxycholesterol and 7-ketocholesterol) induced the degradation of the prosurvival protein kinase AKT (pro-tein kinase B). This led, in turn, to the activation of the BCL-2 homology-3 domain-only proteins BIM and BAD and down-regulation of the anti-apoptotic multi-BCL homology domain protein BCL-x L . These responses would be expected to activate the pro-apoptotic multi-BCL homology domain proteins BAX and BAK, leading to the previously reported release of cytochrome c observed during oxysterol-induced apoptosis. Somewhat surprisingly, small interfering RNA knockdown of BAX resulted in a complete block of the induction of apoptosis by 25-hydroxycholesterol. g/ml streptomycin, (cid:1) M 2-mercaptoetha-nol. P388D1 cells cultured modified medium supplemented fetal serum, 2 m M m M sodium 100 units/ml penicillin, (cid:1) g/ml streptomycin. Caspase-3 Assay— P388D1 cells /well 12-well culture plates, 264.7 cells negative control vector pSi-Random Bax the following annealed, custom-synthesized oligonucleotides of rand- omized Bax target sequence into ApaI/EcoRI-digested pSilencer: Random Bax 1, ACCGCTCGAGCGTGCTAGTTTCAAGAGAACTAGCACG- CTCGAGCGGTTTTTTT; and Random Bax 2, AATTAAAAAAACCGCT-CGAGCGTGCTAGTTCTCTGAAACTAGCACGCTCGAGCGGTGGCC.P388D1cells(1 (cid:1) 10 6 ) were cotransfected with a plasmid carrying the neomycin gene, pEGFP (Clontech), and either pSi- Bax or pSi-Random Bax using LipofectAMINE Plus (Invitrogen) according to the manufacturer’s directions. Stable clones were first selected by infinite dilution in medium containing 1 mg/ml G418 for 7 days, followed by 14 days in 0.25 mg/ml G418. G418-resistant clones were screened by PCR for integration of the pSilencer vector. The PCR positive G418-resistant clones were then subjected to selection in medium containing 10 (cid:1) g/ml 25-OHC for 68 h. The surviving cells were expanded and maintained in medium containing 500 (cid:1) g/ml G418. No 25-OHC-resistant clones were obtained from PCR positive G418-resistant pRandom Bax clones. The suppression of BAX expression was determined by immunoblotting whole cell lysates prepared from the isolated clones, wild-type cells, and G418-resistant pRandom Bax clones using anti-BAX antibody. were then processed using antibodies specific for the protein of interest and the appropriate peroxidase-conjugated secondary antibodies following manufacturers’ directions. The proteins of interest were visualized by enhanced chemiluminescence using SuperSignal® West Pico chemi-luminescent substrate (Pierce) as directed. Pulse-Chase Experiments— P388D1 cells were grown in methionine-deficient medium for 2 h and then pulsed with Tran 35 S-label (100 (cid:1) Ci/ml; ICN) for 3 h. Cells were then washed and chased for different time periods in regular growth medium containing 1 m M unlabeled methionine and either vehicle (ethanol) or 10 (cid:1) g/ml 25-OHC. Cells were lysed as described above and subjected to immunoprecipitation with anti-AKT antibodies. Assays were performed using Seize TM -coated plate immunoprecipitation kits (Pierce) according to the manufacturer’s instructions. After binding, washing, and elution, the presence of ra-diolabeled AKT in the precipitates was examined by SDS-PAGE and phosphorimaging.

phosphate-buffered saline, resuspended in lysis buffer A (10 mM Tris (pH 7.5), 130 mM NaCl, 1% Triton X-100, 10 mM sodium P i , and 10 mM sodium PP i ), incubated on ice for 10 min, and centrifuged at 12,000 ϫ g for 20 min at 4°C. The supernatant or extract was assayed for protein using the micro-BCA kit (Pierce) and for caspase-3 activity as follows. Equivalent amounts of each sample were incubated for 2.5 h at 37°C in caspase assay buffer (20 mM Hepes (pH 7.5), 10% glycerol, and 2 mM dithiothreitol) containing 5 M caspase-3 substrate (Ac-DEVD-7-amino-4-trifluoromethyl coumarin) in the presence and absence of a caspase-3-specific inhibitor (Ac-DEVD-aldehyde) added at 100 nM 30 min prior to the addition of the substrate. Liberated 7-amino-4-trifluoromethyl coumarin was measured using a spectrofluorometer (FluroMax 3 equipped with a microplate reader, Jobin-Yvon Inc.) with an excitation wavelength of 400 nm and an emission wavelength of 505 nm. For each sample, the net caspase-3 activity was determined by subtracting the relative fluorescent light units obtained in the presence of the inhibitor from the relative fluorescent light units obtained in the absence of the inhibitor and normalizing to the protein content of the sample. Each treatment was performed in triplicate, and the data are presented as the mean Ϯ S.D.
BAD-GFP Fusion Protein-The Bad open reading frame was excised from pEBG-mBAD (Cell Signaling Technology, Inc.) with HindIII and NotI and ligated in-frame into pEGFP-C3 after the creation of an appropriate NotI site using the QuikChange kit (Stratagene). The finished construct was confirmed by sequencing and transfected into RAW 264.7 cells using GeneJammer (Stratagene).
Detection of BAD-GFP in RAW 264.7 Cells-Transiently transfected RAW 264.7 cells were incubated overnight on glass chambered coverslips. Cells were then washed and incubated with 25-OHC (10 g/ml) for 6 h. After washing with phosphate-buffered saline, cells were observed live under a fluorescence microscope. In colocalization experiments, mitochondria were also stained with MitoTracker TM Red following the manufacturer's instructions. Confocal images were obtained by digital deconvolution of 10-slice stacks acquired on a Nikon Diaphot 200 equipped with a Photometrics Sensys cooled CCD digital camera or Nikon D100 Oncor Z-drive and Oncor Image software.

FIG. 1. A specific inhibitor of cytosolic phospholipase A 2 (AA-COCF 3 ) and a general inhibitor of arachidonate metabolism (ETYA) prevent induction of caspase-3 by 25-OHC in P388D1
cells. P388D1 cells were cultured in medium with or without 20 M AACOCF 3 or 10 M ETYA for 2 h prior to the addition of 25-OHC or vehicle (ethanol). Following an 18-h incubation, the cells were harvested, and caspase-3 and "caspase-3-like" activities were measured as described under "Experimental Procedures." The results are expressed as -fold increases in fluorescence relative to the mean fluorescence determined for control treatments (vehicle only). The data represent the mean Ϯ S.D. of triplicate treatments.
FIG. 2. 25-OHC induces proteasomedependent AKT degradation. A, P388D1 cells were treated with 10 g/ml 25-OHC for 12 h, and AKT kinase activity was measured as described under "Experimental Procedures" using glycogen synthase kinase-3␤ fusion protein as substrate. The data represent the mean Ϯ S.D. of triplicate measurements. B, P388D1 cells were treated with 10 g/ml 25-OHC for different times, and the levels of total AKT were measured by immunoblotting as described under "Experimental Procedures." C, P388D1 cells were pulse-labeled with 100 Ci/ml Tran 35 Slabel and chased in the absence (E) or presence (q) of 10 g/ml 25-OHC as described under "Experimental Procedures." At different times, total AKT was immunoprecipitated, and radioactivity was measured by SDS-PAGE and phosphorimaging. The extrapolated half-lives for AKT were ϳ8 h and ϳ100 min in the control and 25-OHC-treated cells, respectively. D, total AKT levels were determined by immunoblotting after treatment with 10 g/ml 25-OHC for 12 h in the presence or absence of 15 M proteasome inhibitor I (PI-I).
P388D1 cells (1 ϫ 10 6 ) were cotransfected with a plasmid carrying the neomycin gene, pEGFP (Clontech), and either pSi-Bax or pSi-RandomBax using LipofectAMINE Plus (Invitrogen) according to the manufacturer's directions. Stable clones were first selected by infinite dilution in medium containing 1 mg/ml G418 for 7 days, followed by 14 days in 0.25 mg/ml G418. G418-resistant clones were screened by PCR for integration of the pSilencer vector. The PCR positive G418-resistant clones were then subjected to selection in medium containing 10 g/ml 25-OHC for 68 h. The surviving cells were expanded and maintained in medium containing 500 g/ml G418. No 25-OHC-resistant clones were obtained from PCR positive G418-resistant pRandomBax clones. The suppression of BAX expression was determined by immunoblotting whole cell lysates prepared from the isolated clones, wild-type cells, and G418-resistant pRandomBax clones using anti-BAX antibody.
Expression of Constitutively Active AKT-P388D1 cells were stably transfected with a vector expressing Myc-His-tagged mouse AKT1 (activated) under the control of the cytomegalovirus promoter (Upstate Biotechnology, Inc.) using LipofectAMINE Plus according to the manufacturer's directions. Stable clones were isolated by selection in medium containing 1 mg/ml G418 for 6 days, followed by 10 days in 250 g/ml G418. The G418-resistant clones were isolated, expanded, and screened for expression of the Myc-AKT fusion protein by immunoblotting with anti-total AKT1/2 antibody (Cell Signaling Technology, Inc.) and anti-Myc antibody. Positive Myc-AKT clones were maintained in medium containing 250 g/ml G418.
For transient transfection experiments, P388D1 cells were seeded at 2 ϫ 10 6 /well in 6-well tissue culture plates. Twenty-four hours later, the cells were rinsed with phosphate-buffered saline, refed standard growth medium, and transfected with pUSEamp or with pUSEamp containing myristoylated (myr) Akt cDNA using Tojene TM transfection reagent (Avanti Polar Lipids, Inc.) according to the manufacturer's directions. Following transfection, expression of the transfected construct was allowed to proceed for 24 h prior to the addition of oxysterol and analysis of caspase-3 activity.
Cell Lysis and Immunoblotting-P388D1 cells were grown to a density of 2 ϫ 10 6 /ml in regular growth medium supplemented with either vehicle (ethanol) or increasing amounts of 25-OHC. After different periods of time, cells were spun and treated with lysis buffer B (20 mM Tris (pH 7.5), 150 mM NaCl, 1 mM EDTA, 1 mM EGTA, 1% Triton X-100, 2.5 mM sodium pyrophosphate, 1 mM ␤-glycerol phosphate, 1 mM Na 3 VO 4 , and 1 g/ml leupeptin) and incubated on ice for 30 min. Insoluble debris was removed from the extracts by centrifugation for 10 min at 10,000 ϫ g, and the protein concentration in the supernatants was determined by micro-BCA assay. Proteins were resolved by SDS-PAGE on 4 -12% NuPAGE gels (Invitrogen) and transferred to polyvinylidene difluoride membranes (Immobilon-P, Millipore Corp.). The membranes were stained with SYPRO®Ruby protein stain (Molecular Probes, Inc.) to ascertain equivalent loading of the gel and efficient transfer of proteins to the membranes before immunoblotting. The blots FIG. 3. Ectopic expression of constitutively active AKT reduces 25-OHC induction of caspase-3 activity in P388D1 cells. A, whole cell lysates obtained from P388D1 cells stably transfected with a myr-Akt expression vector (clone B) or the empty vector (control) were subjected to SDS-PAGE, transferred to polyvinylidene difluoride membranes, and immunoblotted using anti-total AKT and anti-Myc antibodies as indicated. B and C, clone B (E) and control (q) cells were incubated with medium containing either 10 g/ml 25-OHC or 10 g/ml 7-ketocholesterol, respectively, for various times. The medium was then replaced with medium without oxysterol, and incubation was continued for 16 h. The -fold induction of caspase-3 activity was determined as described for Fig. 1. D, P388D1 cells transiently transfected with no DNA (Mock), empty vector (pUSEamp), or vector containing Akt cDNA (myr-Akt) were incubated in medium with or without 10 g/ml 25-OHC for 6 h as indicated. The medium was refreshed without oxysterol supplementation; and following a 16-h incubation, the -fold induction of caspase-3 activity was determined.
were then processed using antibodies specific for the protein of interest and the appropriate peroxidase-conjugated secondary antibodies following manufacturers' directions. The proteins of interest were visualized by enhanced chemiluminescence using SuperSignal® West Pico chemiluminescent substrate (Pierce) as directed.
Pulse-Chase Experiments-P388D1 cells were grown in methioninedeficient medium for 2 h and then pulsed with Tran 35 S-label (100 Ci/ml; ICN) for 3 h. Cells were then washed and chased for different time periods in regular growth medium containing 1 mM unlabeled methionine and either vehicle (ethanol) or 10 g/ml 25-OHC. Cells were lysed as described above and subjected to immunoprecipitation with anti-AKT antibodies. Assays were performed using Seize TM -coated plate immunoprecipitation kits (Pierce) according to the manufacturer's instructions. After binding, washing, and elution, the presence of radiolabeled AKT in the precipitates was examined by SDS-PAGE and phosphorimaging.
AKT Kinase Assay-The active form of AKT was measured using a nonradioactive AKT kinase assay kit (Cell Signaling Technology, Inc.) following the manufacturer's instructions. Essentially, an antibody to AKT was used to selectively immunoprecipitate AKT from cell lysates. The immunoprecipitate was then incubated with glycogen synthase kinase-3␤ fusion protein in the presence of ATP and kinase buffer, allowing immunoprecipitated AKT to phosphorylate glycogen synthase kinase-3␤. Phosphorylation of glycogen synthase kinase-3␤ was then measured by immunoblotting using anti-phospho-Ser-21/9 glycogen synthase kinase-3␣/␤ antibody and quantitated by phosphorimaging.

25-OHC Induces Apoptosis in P388D1
Cells through a Process Dependent on Arachidonate Metabolism-We have previously characterized oxysterol induction of apoptosis in both a fibroblast cell line (CHO-K1) and a monocyte-macrophage cell line (THP-1) as being dependent on arachidonate release and metabolism (9). Because of difficulties we encountered in transfection of THP-1 cells and the well established use of the murine macrophage P388D1 cells in studies of arachidonate metabolism (22), we determined whether P388D1 cells undergo apoptosis in response to oxysterols in a similar fashion. We confirmed a prior report (16) that 25-OHC induces apoptosis in P388D1 cells and also demonstrated, as was observed for the other cell lines, that the induction of apoptosis was blocked both by the cytosolic phospholipase A 2 inhibitor AACOCF 3 and by ETYA, an inhibitor of arachidonate metabolism (Fig. 1).
25-OHC Down-regulates AKT in P388D1 Cells-AKT (protein kinase B) has been well characterized as an anti-apoptotic kinase (24) that transduces cellular survival signals in many cell types (25). In particular, a critical role for AKT has been ascribed to the survival of macrophages (26). We therefore examined the effect of overnight treatment with 25-OHC on the activity of AKT in the murine macrophage cell line P388D1. Activity was assayed with glycogen synthase kinase-3␤ as substrate. The results clearly indicate that AKT activity was down-regulated in P388D1 cells in response to treatment with 25-OHC ( Fig. 2A).
The mechanism of down-regulation of AKT was explored by immunoblot and radioimmunoprecipitation experiments. Immunoblot analysis demonstrated that oxysterol treatment produced a reduction in the level of total AKT in P388D1 cells (Fig.  2B). Pulse-chase radioimmunoprecipitation studies revealed that 25-OHC treatment greatly enhanced the rate of degradation of AKT (Fig. 2C). Furthermore, the effect was observed with no time lag after 25-OHC addition, suggesting that this is an early signaling event. We also examined the effect of treatment with proteasome inhibitor I on the cellular levels and activity of AKT in 25-OHC-treated cells. This was done to determine whether this enhanced degradation rate was responsible for the decrease in its activity as well as to gain insight into the mechanism of regulated degradation. The results demonstrate that inhibition of the degradation of AKT significantly attenuated the loss of its activity in response to 25-OHC treatment (Fig. 2, A and D). Therefore, the primary mechanism by which 25-OHC down-regulates the activity of AKT appears to be through stimulation of its rate of degradation. This mecha- nism of degradation is also consistent with an early regulated event.
We wanted to confirm this putative critical role for AKT signaling in 25-OHC-induced apoptosis. Therefore, we examined the effect of expression of a constitutively active Myctagged form of AKT (myr-AKT) (27) on the activation of caspase-3 by 25-OHC treatment. We isolated a clone (clone B) stably expressing transfected myr-AKT (Fig. 3A). AKT activity in clone B cells was shown to be elevated ϳ2-fold relative to that in wild-type P388D1 cells as demonstrated in vitro by enzyme assay both in untreated and 25-OHC-treated cells (Fig.  3B). Clone B was also seen to be relatively resistant to induction of apoptosis by 25-OHC or 7-ketocholesterol as measured by caspase-3 activity (Fig. 3, B and C). To guard against the possibility that this result was due to another genetic variation in clone B, the ability of myr-AKT to protect cells from oxysterol-induced caspase-3 activation was also demonstrated by transient transfection with the same construct (Fig. 3D).

Effect of 25-OHC Treatment on BH3 Domain-only Proteins: Activation of BAD and Increased Cellular Levels of BIM-A
common mechanism by which AKT inactivation is coupled to apoptosis is through the regulation of BAD by AKT (25). AKT phosphorylates BAD at two serine residues, resulting in its sequestration in the cytosol, bound to 14-3-3 (28 -30). In the absence of ongoing AKT-catalyzed phosphorylation, BAD becomes dephosphorylated by any of several phosphatases (31). This results in its relocalization to the mitochondria (32-34), where it heterodimerizes via its BH3 domain (35) with the anti-apoptotic BCL family members BCL-2 and BCL-x L (36). These are critical signaling events in the mitochondrial apoptotic pathway that result in canonical cytochrome c release (37). Because we have previously demonstrated that oxysterols mediate cytochrome c release from mitochondria during apoptotic induction in other cell types (19), we anticipated that a consequence of the observed AKT degradation would be activation of BAD.
BAD activation can be detected either through determination of its dephosphorylation or by its relocalization. We examined both processes. The immunoblot of BAD from 25-OHCtreated P388D1 cells with anti-phospho-BAD antibody is consistent with loss of phosphorylation (Fig. 4). In clone B, BAD was observed to be hyperphosphorylated both in untreated and 25-OHC-treated cells (Fig. 4), demonstrating the activity and, in particular, the anti-apoptotic activity of the myr-AKT construct in whole cells. The immunoblots of total mitochondrial BAD were consistent with an increase in the mitochondrial levels of BAD after 25-OHC treatment (data not shown). Relocalization of BAD was also demonstrated by transfection of murine RAW 264.7 macrophages with a BAD-GFP fusion protein. Such a reporter has previously been utilized to demonstrate BAD activation in another system (33). RAW 264.7 cells were used for this experiment because they grow attached and spread out on the surface of a plastic Petri dish, allowing better resolution of the cellular distribution of the The results indicate a redistribution of BAD from a diffuse cytosolic localization to a punctate distribution largely colocalized with the mitochondrial marker MitoTracker (Fig. 5). This is precisely the result to be expected for the relocalization of BAD associated with dephosphorylation (33).
Korsmeyer and co-workers (38) have presented evidence for two classes of BH3 domain-only proteins. The first category is exemplified by BAD, which functions through "sensitization" toward apoptosis by heterodimerization and inactivation of anti-apoptotic multi-BCL homology domain family members. BH3 domain-only proteins in the second category activate the pro-apoptotic multi-BCL homology domain family members by inducing their oligomerization to release cytochrome c. Expression of one of these proteins (BIM) has been shown to be down-regulated by AKT (39). We therefore examined P388D1 cells for increased expression of BIM after 25-OHC treatment by immunoblotting. The results are consistent with an increase in BIM levels in cells and in mitochondria (Fig. 6).
Role of Multi-BCL Homology Domain Proteins in Induction of Apoptosis by 25-OHC-Release of cytochrome c from mitochondria in response to signaling by the BH3 domain-only proteins would be expected to involve two downstream events: the inactivation (sensitization) of the anti-apoptotic multi-BCL homology domain proteins and the activation of the pro-apoptotic multi-BCL homology domain proteins. As mentioned above, the potential targets of BAD are BCL-2 and BCL-x L . However, it has been reported that BAD can reverse the death repressor activity of only BCL-x L (36). Furthermore, immunoblot analysis indicated that P388D1 cells did not express BCL-2 (data not shown). This suggests that the target of BAD during 25-OHC induction of apoptosis is BCL-x L . It has been reported that AKT up-regulates the expression of BCL-x L , suggesting the possibility that 25-OHC treatment might down-regulate its expression (39). Immunoblots of BCL-x L after 25-OHC treatment demonstrated that this did occur (Fig. 7A). Down-regulation of BCL-x L mRNA levels by 25-OHC as determined from Northern blots was also observed (data not shown).
As described above, the current thinking (38,40) on the mechanism of BAD signaling in mitochondria is that BAD binding to BCL-x L would in turn activate the pro-apoptotic multi-BCL homology domain family members BAX and BAK to release cytochrome c (36,41). Immunoblot analysis demonstrated that P388D1 cells expressed both BAK and BAX, with BAX having the usual (41) dual cytoplasmic and mitochondrial localization and BAK being found only in mitochondria (data not shown). It has been shown that BAX and BAK have overlapping functionality, but that suppression of BAX attenuates the apoptotic response to a variety of stimuli that act through the intrinsic pathway (42). To confirm that the BAX/BAK pathway operates as expected in the induction of apoptosis by oxysterols, we examined the effect of small interfering RNA knockdown of BAX on the apoptotic response to 25-OHC. After selection (see "Experimental Procedures"), we were able to isolate a clone of P388D1 cells (clone 1) that appeared to be highly reduced in BAX expression as determined by immunoblotting (Fig. 7B). This clone also exhibited no detectable apoptotic response to 25-OHC (Fig. 7C).
In contrast, prolonged exposure of clone B to 25-OHC still resulted in cell death and caspase-3 activation (Fig. 8A). This is apparently due to eventual degradation of the overexpressed, constitutively active AKT, resulting in a significant decrease in its cellular levels as measured by immunoblotting (Fig. 8B). This result is also consistent with the expression of constitu- tively active AKT being the cause of the oxysterol resistance in clone B rather than an artifact of clone selection. DISCUSSION The results presented in this study are consistent with the conclusion that an important regulatory event in the induction of apoptosis by oxysterols is accelerated degradation of AKT. The activity of AKT has most commonly been described to be regulated by phosphorylation (24). In one well characterized apoptotic signal transduction pathway, that induced by ceramide, activation of protein phosphatase-2A (leading to dephosphorylation of AKT) plays a critical role (43). However, there have been other reports of regulated AKT degradation playing a role in apoptotic signal transduction pathways. For example, H 2 O 2 -induced apoptosis is accompanied by degradation of AKT (44), as is the induction of apoptosis by the 15-deoxyprostaglandin J 2 (45), both probable examples of apoptosis with reactive oxygen species second messengers (44,46). It is worth noting, in this context, that reactive oxygen species have been suggested to play a role in oxysterol-induced apoptosis (14,20).
An important consideration is whether the activation of AKT degradation is upstream or downstream of the many caspases activated during apoptosis. Our data are consistent with activation of AKT degradation being an early event. First, the degradative response was observed in as little as 30 min after 25-OHC addition (Fig. 2C), whereas we have previously observed caspase-3 activation to be undetectable in response to 25-OHC treatment in as much as 5 h (19). Second, the inhibition of degradation by a proteasome inhibitor (Fig. 2D) is consistent with a non-caspase-mediated degradatory process. Finally and most tellingly, overexpression of a constitutively active AKT (myr-AKT) in clone B delayed the activation of caspase-3 by 25-OHC, consistent with caspase activation being downstream of AKT degradation rather than the other way around. It is also noteworthy that the eventual degradation of myr-AKT produced by 25-OHC treatment can still result in capsase-3 activation, consistent with a temporal pattern of AKT degradation preceding caspase-3 activation.
The responses of the various BCL family members studied in this work are predictable from the loss of AKT activity. The role of AKT in phosphorylation and inactivation of BAD is well documented (25, 28 -36); and therefore, it is not surprising that loss of AKT resulted in dephosphorylation of BAD (Fig. 4) and its relocalization to the mitochondria (Figs. 4 and 5), where it would be expected to heterodimerize with the anti-apoptotic multi-BCL homology domain family members. Further sensitization to apoptotic induction is achieved through loss of AKT because of its role in signaling transcriptional up-regulation, through the NF-B pathway (47)(48)(49), of the anti-apoptotic multi-BCL homology domain family member BCL-x L (39,50).
Transcriptional down-regulation of the BH3 domain-only protein BIM by AKT has also been reported (39), and our observation of increased BIM expression in whole cells and mitochondria (Fig. 6) in response to 25-OHC is consistent with this prior work. In the mitochondrial apoptotic pathway, BIM functions to stimulate the oligomerization of the pro-apoptotic multi-BCL homology domain family members BAX and BAK, thereby producing the release of cytochrome c (38).
Sensitization to apoptotic induction by down-regulation of BCL-x L and relocalization of BAD to mitochondria and expression of BIM should lead to activation of BAX/BAK. Somewhat surprisingly, rather than attenuation of apoptosis in response to 25-OHC treatment after small interfering RNA knockdown of BAX, we were able to isolate, by immunoblotting, a clone with no detectable BAX that exhibited no detectable apoptotic response (Fig. 8). It has been reported that some stimuli, no-tably staurosporine, that operate through the intrinsic apoptotic pathway are more responsive to BAX knockout than to BAK knockout (42), and we may have observed such a differential response to 25-OHC.
It has been reported that another oxysterol, 7-ketocholesterol, activates apoptosis in fibroblasts through a signal transduction pathway that requires phosphorylation of STAT1 at Ser-727 (51). This is noteworthy because some signal transduction pathways to STAT1 serine phosphorylation proceed through AKT (52). However, reports have appeared that have implicated other kinases in STAT1 serine phosphorylation, including p38 mitogen-activated protein kinase and protein kinase C (53,54). Ser-727 phosphorylation in macrophages, in particular, seems to be more dependent on these other kinases (54).
While this manuscript was under review, a report appeared describing the Ca 2ϩ release-mediated activation of the endoplasmic reticulum unfolded protein response pathway during the induction of apoptosis by free cholesterol loading of macrophages (23). The endoplasmic reticulum was interpreted to be the site of cholesterol damage leading to the calcium signal. The oxysterol induction of apoptosis appears to be different, at least in some respects, from this system because oxysterol apoptotic signaling appears to proceed through a signal transduction pathway requiring arachidonic acid release (9) and possibly reactive oxygen species formation (14,20), both of which might be expected to be upstream of the AKT degradation reported here. Whether cholesterol loading can activate AKT degradation and whether oxysterols can activate one or more elements of the unfolded protein response remain to be determined.