Address correspondence and reprint requests to David C. Henshall, PhD, Robert S. Dow Neurobiology Laboratories, Legacy Clinical Research & Technology Center, 1225 NE 2nd Avenue, Portland, OR 97232, USA. E-mail: dhenshall@DowNeurobiology.org
Seizure-induced neuronal death may involve coordinated intracellular trafficking and protein–protein interactions of members of the Bcl-2 family. The 14-3-3 proteins are known to sequester certain pro-apoptotic members of this family. BH3-interacting domain death agonist (Bid) may contribute to seizure-induced neuronal death, although regulation by 14-3-3 has not been reported. In this study we examined whether 14-3-3 proteins interact with Bid during seizure-induced neuronal death. Brief seizures were evoked in rats by intraamygdala microinjection of kainic acid to elicit unilateral hippocampal CA3 neuronal death. Coimmunoprecipitation analysis demonstrated that although Bcl-2-associated death promoter (Bad) constitutively bound 14-3-3, there was no interaction between Bid and 14-3-3 in control brain. Seizures triggered Bid cleavage and a commensurate increase in binding of Bid to 14-3-3 within injured hippocampus. Casein kinases I and II, which can inactivate Bid by phosphoserine/threonine modification, did not coimmunoprecipitate with Bid. The largely uninjured contralateral hippocampus did not exhibit Bid cleavage or binding of 14-3-3 to Bid. In vitro experiments confirmed that 14-3-3β is capable of binding truncated Bid, likely in the absence of phosphoserine/threonine modification. These data suggest 14-3-3 proteins may target active as well as inactive conformations of pro-apoptotic Bcl-2 death agonists, highlighting novel targets for intervention in seizure-induced neuronal death.
Broadening the involvement of 14-3-3 in the regulation of cell death, was the recent demonstration that 14-3-3 also sequesters Bcl-2-associated X protein (Bax) (Nomura et al. 2002). Furthermore, Bcl-2-interacting mediator of cell death (Bim) and Bcl-2-interacting killer (Bik) have now been identified as phosphoproteins (Verma et al. 2001; Biswas and Greene 2002), suggesting the role of 14-3-3 proteins may be expanded still further. BH3-interacting domain death agonist (Bid) (Wang et al. 1996), which links the extrinsic, death receptor pathways to mitochondria, has been implicated in the mechanism of neuronal death following a variety of insults (Plesnila et al. 2001; Guegan et al. 2002; Yin et al. 2002), including seizures (Henshall et al. 2001a). Truncated Bid (tBid) likely acts as a coagonist for Bax/Bak, which subsequently provoke cytochrome c release from mitochondria (Desagher et al. 1999; Korsmeyer et al. 2000). Recently, Bid was shown to be a phosphoprotein, a modification mediated by casein kinases (Desagher et al. 2001). PhosphoBid is resistant to proteolysis (Desagher et al. 2001), a caspase-mediated event that generates the C-terminus 15-kDa fragment that translocates to mitochondria (Li et al. 1998; Gross et al. 1999). Whether such post-translational modification of Bid might also promote interaction with 14-3-3 isoforms, which do not normally bind Bid (Nomura et al. 2002), is unknown. We therefore questioned whether Bid, like Bad, interacts with 14-3-3 in the setting of seizure-induced brain injury.
Materials and methods
All animal procedures were performed in a facility accredited by the Association for Assessment and Accreditation of Laboratory Animal Care (AAALAC) in accordance with protocols approved by the Legacy Institutional Animal Care and Use Committee and the principles outlined in the National Institute of Health's Guide for the Care and Use of Laboratory Animals. Studies were performed as previously described (Henshall et al. 2000c) with some modifications. Seizures were focally evoked in adult male Sprague-Dawley rats (280–350 g) by unilateral stereotaxic microinjection of kainic acid (Sigma, St. Louis, MO, USA) into the basolateral amygdala nucleus. Following anesthesia (isoflurane), intubation and vein catheterization, animals were placed in a stereotaxic frame. Recording electrodes (Plastics One Inc., Roanoke, VA, USA) were then affixed to the skull bi-temporally over the hippocampi and a third across the frontal cortex to record electroencephalogram (EEG) (Grass model 8–16). A craniectomy was also performed for placement of the injection cannula. The animal was then removed from the frame, anesthesia was discontinued, and EEG recordings were commenced. Next, a 31-gauge internal cannula (Plastics One Inc.) was inserted into the lumen of the guide to inject kainic acid (0.1 µg in 0.5 µL saline vehicle) into the amygdala. The EEG was continuously monitored until diazepam (30 mg/kg; intravenous) was administered to terminate seizures after 40 min. The EEG was further monitored for up to 1 h to ensure seizure cessation. Non-seizure control animals received intraamygdala vehicle and intravenous diazepam, after the same surgical procedure.
Animals were killed 0, 4, or 24 h following administration of diazepam in seizure animals or after 4 or 24 h in non-seizure controls and the entire ipsilateral and contralateral hippocampus were microdissected free and frozen on dry ice. Brain samples (n = 1 or 2 per time point) were homogenized and lyzed in buffer containing 1% NP40 and the protease inhibitors phenylmethylsulfonylfluoride 100 µg/mL, leupeptin (1 µg/mL), pepstatin (1 µg/mL) and aprotinin (1 µg/mL). Lysates were cleared by centrifugation and protein concentration was determined using Bradford reagent spectrophotometrically at A595 nm. Fifty-microgram samples were then boiled in gel-loading buffer and separated by sodium dodecyl sulfate–polyacrylamide gel electrophoresis. Proteins were transferred to polyvinylidene difluoride membranes (Bio-Rad, Hercules, PA, USA) and then incubated with primary antibodies. We employed two goat polyclonal Bid antibodies that recognize p22 (Bid) and the p15 fragment (tBid) (Santa Cruz Biotechnology, Santa Cruz, CA, USA and R & D Systems, Minneapolis, MN, USA). Each Bid antibody produced comparable results in terms of detecting the p22 and p15 fragments. We also employed monoclonal and polyclonal antibodies against 14-3-3β (Santa Cruz Biotechnology). Control for protein loading was performed by re-probing membranes with antibodies against α-tubulin (Santa Cruz Biotechnology). Membranes were then incubated with appropriate secondary antibodies (1 : 2000 dilution) followed by chemiluminescence detection (NEN Life Science Products, Boston, MA, USA), and then exposed to Kodak X-OMAT film (Kodak, Rochester, NY, USA). Images were collected using a Dage 72 camera and gel-scanning software (Bioquant, Nashville, TN, USA).
Animals were killed 0, 4, or 24 h following administration of diazepam in seizure animals or after 4 h in non-seizure controls. Pooled brain samples (n = 2 animals per group) were dounce homogenized and lyzed in buffer containing 1% NP-40 and the same protease inhibitor cocktail used for western blotting, and protein concentration was determined as described above. Protein samples (0.5 mg) were incubated overnight at 4°C with 2 µg of a custom-made rabbit polyclonal Bid antibody (a gift from D. Chen, Legacy Research) as the immunoprecipitating antibody. This Bid antibody produced a greater immunoprecipitation Bid/tBid yield than the commercially available goat polyclonal antibodies in preliminary tests. A control immunoprecipitation assay using anti-Bad (Cell Signaling Technology, Beverly, MA, USA) was also performed. Samples were then incubated with protein A/G agarose beads (Santa Cruz Biotechnology) for 2 h at 4°C. The protein–bead complex was then washed, collected by centrifugation and samples were boiled in loading buffer and run on 12% sodium dodecyl sulfate–polyacrylamide gel electrophoresis gels, probed with the Bid or 14-3-3 antibodies and processed as described for western blotting. Positive (whole cell lysates or recombinant rat Bid, a gift from D. Chen, Legacy Research) and negative (omitting the immunoprecipitation antibody) controls were included to confirm specificity of reactions.
Generation of recombinant 14-3-3β and in vitro binding assay with tBid
The 14-3-3β cDNA subcloned into pGEX-KT was generously provided by Dr Andrey Shaw (Center for Immunology and Department of Pathology, Washington University School of Medicine, St Louis, MO, USA). Expression vectors were transformed into BL21 Gold E. coli competent cells (Stratagene, Cedar Creek, TX, USA), and a single colony was used to inoculate 500 mL Luria-Bertani (LB) medium and grown for 2–3 h by shaking at 240 r.p.m. at 37°C. Isopropyl-1-thio-β-d-galactopyranoside was added to the culture (final concentration 0.5 mm) and the cultures were grown for another 3–5 h. Bacteria were pelleted by centrifugation, resuspended and lyzed by sonication in phosphate-buffered saline with 1% Triton X-100. The lysate was centrifuged at 12 000 × g for 10 min at 4°C and the supernatant mixed with 2 mL of 50% glutathione sepharose 4B beads at room temperature for 30 min. The suspension was transferred to a QIAquik gel extraction column (Qiagen), washed with phosphate buffer, incubated with 50 mU/mL Thrombin protease (Amersham) overnight at room temperature and collected by centrifugation.
For in vitro determination of the interaction of tBid with 14-3-3β, reactions were performed in buffer containing 25 mm HEPES (pH 7.4), 10 mm dithiothreitol, 1 mm EDTA, 0.1% CHAPS and 10% sucrose containing dATP. Caspase-8-cleaved mouse tBid (2.5 µg, R & D Systems) and recombinant 14-3-3β (5 µg) were incubated for 30 min at room temperature in 200 µL reaction buffer and then subject to Bid or 14-3-3 immunoprecipitation as described above followed by sodium dodecyl sulfate–polyacrylamide gel electrophoresis and western blotting using anti-14-3-3β or anti-Bid, respectively. As additional controls, some reactions were performed in the presence of the S100 cytoplasmic fraction of control (Bid containing) rat hippocampus (1 mg).
Effect of 14-3-3β on tBid-induced cytochrome c release
For analysis of the functional consequences of 14-3-3β binding to tBid, we assayed the effects of tBid in the presence of 14-3-3β on cytochrome c release from isolated rat brain mitochondria, according to manufacturer's recommendations (R & D Systems) and previously described methods with modifications (Yan et al. 2000). Briefly, mitochondria were purified by centrifugation in buffer containing 225 mm mannitol, 75 mm sucrose, 0.1 mm EGTA and 10 mm HEPES-KOH (pH 7.4). Mitochondria were further enriched by centrifugation at 41 000 × g for 30 min through 26% and 60% percoll gradients. Cytochrome c release assay was then performed by incubating ∼10 µg mitochondria with 2.5 µg tBid ± 5 µg 14-3-3β at 30°C for 30 min in reaction buffer containing 125 mm KCL, 0.5 mm MgCl2, 3 mm succinic acid, 3 mm glutamic acid, 10 mm HEPES-KOH (pH 7.4), 25 mm dATP, and the same protease inhibitor cocktail used for western blotting. Reaction supernatant was collected by centrifugation, decanted and frozen at −20°C until being subject to western blot detection of released cytochrome c (BD Pharmingen, San Diego, CA, USA).
Coronal brain sections (12 µm) at the level of Bregma −3.2 mm (Paxinos and Watson 1997) from animals killed 4 or 24 h following seizures or time-matched controls (n = 4 per group) were pre-blocked in 5% normal goat serum and then incubated overnight at 4°C with rabbit anti-Bid (1 : 250) and monoclonal antibodies against either 14-3-3β (1 : 1000), cytochrome IV oxidase (1 : 200) (COXIV, Molecular Probes, Eugene, OR, USA) or NeuN (1 : 1000) (Chemicon, Temecula, CA, USA). Sections were then washed three times in phosphate-buffered saline and incubated for 2 h at room temperature in a 1 : 500 dilution of goat anti-rabbit or goat anti-mouse Cy3 or fluorescein isothiocyanate (FITC) (Jackson Immunoresearch, PA, USA). Sections were washed again and then mounted in medium containing 4′,6-diamidino-2-phenylindole (DAPI, Vector Laboratories, Burlingame, CA, USA) to assess nuclear morphology. Immunolabeling was studied using a Leica microscope equipped for epifluorescent illumination under excitation/emission wavelengths of 340/425 nm (blue), 500/550 nm (green) and 580/630 nm (red). Analysis of cells exhibiting DNA fragmentation was performed using fluorescein-linked terminal deoxynucleotidyl transferase-mediated dUTP nick end labeling (TUNEL) (Roche Molecular Biochemicals, Indianapolis, IN, USA) to label double-stranded DNA breaks as previously described (Henshall et al. 2001a). Images were collected using an Optronics DEI-750 3-chip camera equipped with a BQ 8000 sVGA frame grabber and an image capture system (Bioquant, Nashville, TN, USA).
Seizures induce Bid proteolysis and binding to 14-3-3
This study was undertaken to determine if Bid interacts with 14-3-3 proteins following neuronal injury caused by seizures. We first verified our coimmunoprecipitation assay technique by replicating our previous observation that, under control conditions, 14-3-3 resides bound to Bad in rat brain (Henshall et al. 2002). Bad was immunoprecipitated from control rat hippocampus and immunoblotted to detect the presence of 14-3-3 using a β-isoform specific antibody. The 14-3-3β isoform was selected based on its implication as the sequestration site for Bad in rat brain (Henshall et al. 2002). We found robust binding of 14-3-3 to Bad in control brain (Fig. 1a).
To confirm seizure-induced processing of Bid and to rule out protein concentration changes as underlying any protein-protein binding changes, we examined the expression of Bid and 14-3-3 in the vulnerable, ipsilateral hippocampus by western blotting. We detected constitutive expression of p22 Bid and 14-3-3 in rat brain, and found that concentrations of both were largely unaffected by seizures (Fig. 1b). We detected the emergence of the p15 truncated form of Bid ∼4 h following seizures.
To confirm that our Bid immunoprecipitation assay could extract both p22 and p15 forms of Bid, we used rat brain hippocampal lysates in which both p15 and p22 Bid were detected by western blot (data not shown), and subjected tissue to Bid immunoprecipitation followed by detection of Bid/tBid by western blotting. Anti-Bid immunoprecipitated both p22 and p15 forms of Bid (Fig. 1c). Using this approach, we immunoprecipitated Bid from control and seizure brain, and then immunoblotted membranes to detect 14-3-3. In control brain, 14-3-3 was not detected in Bid immunoprecipitates (Fig. 1d). However, 4 and 24 h following seizures, 14-3-3 was detected in Bid immunoprecipitates (Fig. 1d). Attempts to immunoprecipitate Bid/tBid from brain using anti-14-3-3 did not produce unequivocal results due to the presence of non-specific bands close to the weights of these proteins (data not shown).
Bid does not bind 14-3-3 within surviving contralateral hippocampus
The seizure model employed is associated with degeneration of ipsilateral CA3 neurons (Henshall et al. 2001a). Despite seizures spreading to the contralateral hippocampus (Ben-Ari et al. 1980), cell death within this region is seldom observed (Henshall et al. 2001a). We therefore examined events in contralateral hippocampus to determine if interaction of Bid with 14-3-3 occurs in seizure-affected brain regions that do not exhibit neurodegeneration.
Western blotting determined that p22 Bid and 14-3-3 were expressed within contralateral hippocampus, although concentrations remained unaffected by seizures (Fig. 1e). Cleaved p15 Bid was not detected within contralateral hippocampus following seizures (Fig. 1e). Bid immunoprecipitation, followed by western blotting to detect 14-3-3, confirmed that 14-3-3 did not interact with Bid at any time following seizures within contralateral hippocampus (Fig. 1f).
TBid binds to 14-3-3βin vitro
Since the observed interaction of 14-3-3 with Bid occurred commensurate with the appearance of tBid, we sought to determine if tBid binds 14-3-3. Purified, recombinant 14-3-3β was incubated with caspase-cleaved tBid in an in vitro binding assay reaction. Bid immunoprecipitates from assays containing 14-3-3β and tBid confirmed that tBid is capable of pulling down 14-3-3β in the absence of cytoplasmic extract from rat brain (Fig. 2a).
We next reversed this reaction, immunoprecipitating 14-3-3 with anti-14-3-3β and immunoblotting to detect bound tBid. Immunoprecipitates from reactions containing 14-3-3β with or without control brain cytoplasmic extract, contained no tBid (Fig. 2b). When tBid was incubated in the presence of 14-3-3β and cytoplasm, the 14-3-3 immunoprecipitates contained tBid (Fig. 2b). Incubation of 14-3-3β with tBid in the absence of cytoplasm did not reduce the tBid yield (data not shown).
14-3-3β does not affect tBid-mediated cytochrome c release
We next assessed the functional significance of 14-3-3β-tBid binding in vitro using a cytochrome c release assay. Incubation of caspase-cleaved tBid with purified rat brain mitochondria triggered cytochrome c release (Fig. 2c). Incubation of 14-3-3β with tBid did not significantly affect tBid-induced cytochrome c release (Fig. 2c).
Casein kinase expression and interaction with Bid following seizures
Casein kinase isoforms, which are expressed in brain (Blanquet 2000), were recently shown to phosphorylate and inhibit Bid cleavage by caspase-8 (Desagher et al. 2001). Furthermore, casein kinase is capable of phosphorylating 14-3-3 isoforms in vitro (Dubois et al. 1997). We therefore examined whether casein kinase isoforms might interact with Bid following seizures, as the underlying drive of Bid binding to 14-3-3. Expression of casein kinase Iε and IIβ subunits was robust in control brain hippocampus and did not noticeably change following seizures (Fig. 3a). Next, Bid immunoprecipitates from ipsilateral or contralateral hippocampus were immunoblotted to detect bound casein kinase Iε or IIβ. Within ipsilateral hippocampus, binding of casein kinase Iε to Bid was not detected in control brain (Fig. 3b). Following seizures, binding of casein kinase Iε to Bid was largely undetectable, and binding of subunit IIβ to Bid was not detected in control or seizure brain at any time (Fig. 3b). Similarly, neither Iε nor IIβ subunits were detected in Bid immunoprecipitates at any time within contralateral hippocampus (data not shown).
Finally, we examined the expression and response of casein kinases following seizures by immunohistochemistry. Casein kinase Iε and IIβ expression was constitutive throughout hippocampal subfields, appearing most often as a diffuse cytoplasmic stain (Fig. 3c and data not shown). Four hours following seizures, casein kinase IIβ accumulated with the nucleus of neurons exhibiting early signs of degeneration (Fig. 3c). Casein kinase IIβ was also seen as a punctate stain within the cytoplasm. In contrast, notable changes for casein kinase Iε were not seen (data not shown).
Bid colocalization with mitochondria following seizures
To extend our studies, we examined Bid expression and localization in control and seizure brain by immunohistochemistry. First, we confirmed that seizures induce TUNEL staining throughout the ipsilateral hippocampal CA3 subfield, 24 h after seizure termination (Fig. 4a, 1, 2). Next, we examined the phenotype of Bid-expressing cells in brain. In control rat brain sections, Bid was expressed constitutively at low concentrations in many hippocampal and also cortical neurons. Bid immunoreactivity appeared as a fine granular particulate within nucleus and cytoplasm, and showed no aggregation (Fig. 4a, 5). Counterstaining with the neuronal marker NeuN revealed Bid is expressed by neurons in control brain (Fig. 4a, 3–6).
Following truncation, Bid translocates to mitochondria to promote cytochrome c release. We next counterstained Bid-immunolabeled rat brain sections with antibodies against mitochondrial cytochrome IV oxidase. No overlapping immunoreactivity between Bid and cytochrome IV oxidase was seen in control hippocampal neurons (Fig. 4b, 1–4). Four hours following seizures, Bid immunoreactivity appeared within the cytoplasm of ipsilateral CA3 neurons (Fig. 4b, 7). Immunolabeling appeared punctate, and counterstained sections revealed overlap of Bid and cytochrome IV oxidase staining suggestive of colocalization (Fig. 4b, 5–8). Semi-quantification of Bid localization changes (n = 100 cells from 3 to 4 animals per group), revealed that numbers of CA3 neurons exhibiting cytoplasmic, punctate Bid immunostaining increased from control (6%) to 49% at 4 h and 66% at 24 h.
Bid colocalizes with 14-3-3 in degenerating CA3 neurons after seizures
Finally, we examined the cellular localization of Bid and 14-3-3 using immunohistochemistry. Like Bid, 14-3-3 was expressed in control rat brain hippocampus and cortex. Furthermore, control expression also appeared diffuse and particulate within the cytoplasm and nucleus of neurons (Fig. 5a, 2). Counterstaining of sections revealed that although similarly distributed, Bid and 14-3-3 immunolabeling does not overlap in control cells (Fig. 5a, 4 and inset in Fig. 5a, 4). Four hours (insets in Fig. 5b, 2–4) and 24 h following seizures (Fig. 5b), Bid immunostaining became strong and punctate within the cytoplasm of many affected ipsilateral CA3 neurons. Counterstaining of sections with anti-14-3-3 revealed colocalization of Bid with 14-3-3 in many cells (Fig. 5b, 2–4).
This study demonstrates that seizures cause Bid cleavage and interaction with 14-3-3 in vulnerable neuronal populations of the hippocampus. In contrast, the surviving contralateral hippocampus did not exhibit Bid cleavage or binding of 14-3-3 to Bid. These studies were extended by in vitro evidence that a recombinant 14-3-3 isoform can interact with caspase-cleaved Bid, likely in the absence of phosphoserine/threonine modifications. Although the functional significance of this interaction remains unclear, these data support novel roles for 14-3-3 proteins in regulating activated pro-apoptotic Bcl-2 family proteins in the setting of brain injury.
We confirmed our previous study (Henshall et al. 2001a) that seizures cause truncation of Bid to its 15 kDa form that targets mitochondria. Activation of Bid was further supported by the seizure-induced transition of Bid immunolabeling to a punctate, mitochondrially localized cytoplasmic distribution in about two-thirds of affected neurons. The largely nuclear localization of Bid in control cells, and its persistence post-seizure, was somewhat unexpected and could reflect non-specific immunostaining. However, Krajewska et al. (2002) showed that constitutive neuronal Bid staining in the adult mouse hippocampal CA3-4 has a similar coarse, granular pattern within the cytoplasm and nucleus, in general agreement with the data presented here and reported previously (Henshall et al. 2001a). A number of studies support activation of Bid as a critical event during ischemic neuronal death (Plesnila et al. 2001; Plesnila et al. 2002; Yin et al. 2002), although Bid may not be involved in neuronal death following other stimuli (Leonard et al. 2001). The present data therefore extend previous studies that support Bid activation during neuronal death.
Although function of pro- and anti-apoptotic Bcl-2 proteins is regulated in part through heterodimerization balances, a number of post-translational modifications also promote or reduce activity, including truncation (Cheng et al. 1997; Li et al. 1998), N-terminal myristoylation (Zha et al. 2000) and phosphorylation (Zha et al. 1996; Maundrell et al. 1997). Regulation of Bad and its sequestration by 14-3-3 proteins has previously been demonstrated following neurological injury in vivo (Springer et al. 2000; Henshall et al. 2002). However, little is known of the relationship between 14-3-3 and other pro-apoptotic Bcl-2 family members regulated by phosphorylation, which includes Bik, for which phosphorylation increases its pro-apoptotic activity (Verma et al. 2001), and now Bid (Desagher et al. 2001). The mechanism by which Bid is phosphorylated is known to involve casein kinases, the effect of which is to render Bid resistant to truncation by caspase-8 (Desagher et al. 2001). The functional significance of this as a brake on apoptosis is supported by reports of high casein kinase activity in human cancers (Tawfic et al. 2001), and inhibition of casein kinase II promoting caspase-mediated Bid truncation downstream of death receptors (Ravi and Bedi 2002). Whether phosphorylated Bid is chaperoned by 14-3-3 proteins is unclear, but previous studies suggested Bid is not capable of binding certain 14-3-3 iosforms (Nomura et al. 2002), likely due to surface charge on the unprocessed form of Bid (Yaffe et al. 1997; McDonnell et al. 1999). However, truncation of Bid produces structural and surface charge alterations (Chou et al. 1999; McDonnell et al. 1999). Our study supports previous work (Nomura et al. 2002), whereby 14-3-3 does not interact with Bid under normal conditions. However, seizures triggered time-dependent binding of 14-3-3 to Bid/tBid, and Bid immunoreactivity colocalized with 14-3-3 in cytoplasmic aggregates within degenerating neurons 24 h after seizures. To our knowledge this is the first demonstration of an interaction between these proteins. This event contrasts that for Bad and 14-3-3, in which their interaction declines after seizures (Henshall et al. 2002). Furthermore, Bax, which is also sequestered by 14-3-3, dissociates following apoptotic stimuli (Nomura et al. 2002). The emergence of 14-3-3 binding to Bid immunoprecipitates, contiguous with the appearance of tBid, suggests it is tBid that binds to 14-3-3. This was well supported by our in vitro studies, which demonstrated that 14-3-3 proteins could bind tBid. Furthermore, binding of 14-3-3 to Bid was not detected in the contralateral hippocampus, which exhibits no cleaved Bid (present data) and limited or no neuronal death in this model (Ben-Ari et al. 1980; Henshall et al. 2000a, 2001a). These data therefore support a different profile of pro-apoptotic interactions of Bcl-2 family proteins with 14-3-3, in which an ‘active’ conformation is bound to 14-3-3, rather than 14-3-3 providing a sequestration site for inactive forms, as is the case for Bad and Bax.
Since Bid can be phosphorylated by casein kinases (Desagher et al. 2001), isoforms of which are expressed in brain (Blanquet 2000), we predicted that the interaction of Bid with 14-3-3 might be driven by casein kinases. However, we could not detect an interaction between Bid and subunits from either casein kinase I, or II. Nevertheless, we cannot rule out the absence of binding of casein kinases to Bid being due to poor resilience of this complex ex vivo. A number of other roles for casein kinases in response to brain injury have been identified (Blanquet 2000), including effects on p53 (Meek et al. 1990; Hjerrild et al. 2001), a pathway activated by seizures (Morrison et al. 1996; Sakhi et al. 1996). Indeed, this may be supported by the observed immunohistochemical changes for the casein kinase IIβ subunit, which appeared to accumulate in the nucleus as early as 4 h after seizures. If casein kinases can inhibit Bid in the present setting, it is possible that early re-location of casein kinase IIβ to the nucleus is related to sustaining Bid function within the cytoplasm/mitochondria.
Our detection of 14-3-3 binding to tBid in a cytoplasm-free assay furthers evidence that this interaction does not require phosphomodifications. Although 14-3-3 binding can require phosphoserine/threonine residues surrounding a core such as RSXSXP and RXXXSXP (Yaffe et al. 1997; Tzivion and Avruch 2002), recent studies reveal a broader range of motifs capable of binding 14-3-3 (O'Kelly et al. 2002). Of particular note, phosphorylation is not required for the interaction/sequestration of Bax with 14-3-3 (Nomura et al. 2002). Examination of the primary amino acid sequence of rat Bid reveals the presence of a possible 14-3-3 binding consequence motif, SRXSRS (Tzivion and Avruch 2002), located within the p15 fragment of Bid. Further studies are now required to determine whether this is the site of tBid-14-3-3 interaction.
The functional consequences of 14-3-3 binding to its targets range considerably from driving subcellular localization to promoting or repressing protein function (Muslin and Xing 2000; Tzivion and Avruch 2002). In the case of Bad and Bax, binding to 14-3-3 prevents their interactions with Bcl-2/Bcl-xl and/or translocation to mitochondria (Yang et al. 1995; Nomura et al. 2002). We found 14-3-3β did not reduce the ability of tBid to provoke cytochrome c release from brain mitochondria. These data suggest targeting of 14-3-3 to tBid may not necessarily interfere with the cell death-promoting actions of tBid. Indeed, colocalization of Bid with 14-3-3 was seen within degenerating CA3 neurons. If 14-3-3 binding exerts an effect other than to block tBid function, we might speculate that 14-3-3 may be functioning to shuttle tBid toward mitochondria. No mechanism of targeting/chaperoning tBid from cytoplasm to mitochondria has been proposed to date, and 14-3-3 proteins mediate a growing number of shuttling functions of proteins relevant to cell death (Baldin 2000; Muslin and Xing 2000; Tzivion and Avruch 2002). Thus, although 14-3-3 proteins are typically associated with anti-apoptotic functions (Masters and Fu 2001; van Hemert et al. 2001), roles that actually promote cell death should not be overlooked. Finally, since protein unfolding is a feature of the neuronal cell death process (Sherman and Goldberg 2001), and 14-3-3 proteins can regulate protein target conformation (Morris et al. 2000), 14-3-3 may be responding to denatured Bid protein and/or chaperoning Bid for degradation. Further studies are therefore required to completely characterize the molecular consequences of 14-3-3 binding to tBid.
In conclusion, this study identifies Bid as a binding partner for 14-3-3 in the setting of neurodegeneration elicited by seizures in brain. Our data support this occurring following truncation of Bid, but it remains to be determined whether 14-3-3 inhibits the function of tBid in vivo, or rather acts to shuttle tBid from cytoplasm to mitochondria or elsewhere. These findings provide insight into the molecular regulation of a potent cell death promoter that might be exploited to reduce neuronal injury in diseases such as epilepsy.
The authors would like to thank to Dr Andrey Shaw for the gift of the 14-3-3β clone and Dr Dexi Chen for the Bid antibody. We would also like to thank Ron Geison (Sigma) for technical support. This work was supported by NIH/NINDS grants NS39016 (to DCH and RPS) and NS41935 (to DCH).