Mutant ubiquitin found in Alzheimer's disease causes neuritic beading of mitochondria in association with neuronal degeneration.

A dinucleotide deletion in human ubiquitin (Ub) B messenger RNA leads to formation of polyubiquitin (UbB)+1, which has been implicated in neuronal cell death in Alzheimer's and other neurodegenerative diseases. Previous studies demonstrate that UbB+1 protein causes proteasome dysfunction. However, the molecular mechanism of UbB+1-mediated neuronal degeneration remains unknown. We now report that UbB+1 causes neuritic beading, impairment of mitochondrial movements, mitochondrial stress and neuronal degeneration in primary neurons. Transfection of UbB+1 induced a buildup of mitochondria in neurites and dysregulation of mitochondrial motor proteins, in particular, through detachment of P74, the dynein intermediate chain, from mitochondria and decreased mitochondria-microtubule interactions. Altered distribution of mitochondria was associated with activation of both the mitochondrial stress and p53 cell death pathways. These results support the hypothesis that neuritic clogging of mitochondria by UbB+1 triggers a cascade of events characterized by local activation of mitochondrial stress followed by global cell death. Furthermore, UbB+1 small interfering RNA efficiently blocked expression of UbB+1 protein, attenuated neuritic beading and preserved cellular morphology, suggesting a potential neuroprotective strategy for certain neurodegenerative disorders.

DARPP-32 and regulation of the ethanol sensitivity of NMDA receptors in the nucleus accumbens.

The medium spiny neurons of the nucleus accumbens receive both an excitatory glutamatergic input from forebrain and a dopaminergic input from the ventral tegmental area. This integration point may constitute a locus whereby the N-methyl-D-aspartate (NMDA)-subtype of glutamate receptors promotes drug reinforcement. Here we investigate how dopaminergic inputs alter the ethanol sensitivity of NMDA receptors in rats and mice and report that previous dopamine receptor-1 (D1) activation, culminating in dopamine and cAMP-regulated phosphoprotein-32 kD (DARPP-32) and NMDA receptor subunit-1 (NR1)-NMDA receptor phosphorylation, strongly decreases ethanol inhibition of NMDA responses. The regulation of ethanol sensitivity of NMDA receptors by D1 receptors was absent in DARPP-32 knockout mice. We propose that DARPP-32 mediated blunting of the response to ethanol subsequent to activation of ventral tegmental area dopaminergic neurons initiates molecular alterations that influence synaptic plasticity in this circuit, thereby promoting the development of ethanol reinforcement.

Role of polyamine metabolism in kainic acid excitotoxicity in organotypic hippocampal slice cultures.

Polyamines are ubiquitous cations that are essential for cell growth, regeneration and differentiation. Increases in polyamine metabolism have been implicated in several neuropathological conditions, including excitotoxicity. However, the precise role of polyamines in neuronal degeneration is still unclear. To investigate mechanisms by which polyamines could contribute to excitotoxic neuronal death, the present study examined the role of the polyamine interconversion pathway in kainic acid (KA) neurotoxicity using organotypic hippocampal slice cultures. Treatment of cultures with N1,N(2)-bis(2,3-butadienyl)-1,4-butanediamine (MDL 72527), an irreversible inhibitor of polyamine oxidase, resulted in a partial but significant neuronal protection, especially in CA1 region. In addition, this pre-treatment also attenuated KA-induced increase in levels of lipid peroxidation, cytosolic cytochrome C release and glial cell activation. Furthermore, pre-treatment with a combination of cyclosporin A (an inhibitor of the mitochondrial permeability transition pore) and MDL 72527 resulted in an additive and almost total neuronal protection against KA toxicity, while the combination of MDL 72527 and EUK-134 (a synthetic catalase/superoxide dismutase mimetic) did not provide additive protection. These data strongly suggest that the polyamine interconversion pathway partially contributes to KA-induced neurodegeneration via the production of reactive oxygen species.

Kainate excitotoxicity in organotypic hippocampal slice cultures: evidence for multiple apoptotic pathways.

The mechanisms underlying kainate (KA) neurotoxicity are still not well understood. We previously reported that KA-mediated neuronal damage in organotypic cultures of hippocampal slices was associated with p53 induction. Recently, both bax and caspase-3 have been demonstrated to be key components of the p53-dependent neuronal death pathway. Caspase activation has also been causally related to the release of mitochondrial cytochrome c (Cyto C) in the cytoplasm as a result of the collapse of the mitochondrial membrane potential (Deltapsi(M)) and the opening of mitochondrial permeability transition pores (mPTP). In the present study, we observed a rapid induction of bax in hippocampal slice cultures after KA treatment. In addition, the levels of Cyto C and caspase-3 were increased in the cytosol while the level of the caspase-9 precursor was decreased. There was also a complete reduction of Rhodamine 123 fluorescence after KA treatment, an indication of Deltapsi(M) dissipation. Furthermore, inhibition of mPTP opening by cyclosporin A partially prevented Cyto C release, caspase activation and neuronal death. These data suggest the involvement of bax, several caspases, as well as Cyto C release in KA-elicited neuronal death. Finally, inhibition of caspase-3 activity by z-VAD-fmk only partially protected neurons from KA toxicity, implying that multiple mechanisms may be involved in KA excitotoxicity.

Downregulation of free ubiquitin: a novel mechanism of p53 stabilization and neuronal cell death.

Neuronal death through activation of the p53 stress response pathway has been implicated in the pathogenesis of neurodegenerative disorders. The mechanisms regulating p53 accumulation and function in neurons are poorly understood. Recent evidence has demonstrated that Mdm2 is a major inhibitor of p53 that binds to and targets p53 for ubiquitin-mediated degradation. Here we demonstrate increased expression and co-localization of p53 and Mdm2 in the nuclei of degenerating neurons following treatment with either the excitotoxin, kainic acid, or the topoisomerase I inhibitor, camptothecin. Co-immunoprecipitation studies showed that p53-Mdm2 complexes were present in neuronal lysates. Dual immunofluorescence microscopy demonstrated that these complexes accumulated in neurons with a striking decrease in free ubiquitin levels. Exogenous ubiquitin restored p53 degradation to extracts from injured neurons confirming that Mdm2 function was intact. Finally, antisense-mediated downregulation of ubiquitin in cultured hippocampal neurons resulted in p53 and Mdm2 accumulation as well as apoptotic death. These results point to a novel mechanism to stabilize p53 and promote neuronal cell death in the central nervous system.

Increased expression of Fas (CD95/APO-1) in adult rat brain after kainate-induced seizures.

Fas (CD95/APO-1), a transmembrane glycoprotein and receptor for the Fas ligand, plays an important role in apoptosis. The present study examined whether excitotoxic cell death induces Fas expression in the adult rat brain. Although relatively light immunostaining was observed in control brain sections, significantly increased Fas immunoreactivity was seen from 4 h to 5 days after the onset of kainic acid-induced seizures. Increased expression of both Fas mRNA and protein were also evident by reverse transcription polymerase chain reaction and Western blotting, respectively. Fas induction was correlated with neuronal apoptosis as demonstrated by colocalization of Fas and terminal dT-mediated dUTP nick end-labeling (TUNEL). Cells with increased Fas-expression were also immunoreactive for tumor suppressor p53 and neuronal specific nuclear protein (NeuN). These results suggest that Fas receptor may contribute to excitotoxic neuronal death in cooperation with p53, and further implicates the Fas pathway in the pathophysiology of neurodegenerative diseases.

Vasopressin and oxytocin receptor mRNA expression during rat telencephalon development.

We investigated the developmental expression of vasopressin and oxytocin receptor and peptide mRNA using semi-quantitative reverse transcription-polymerase chain reaction (RT-PCR) and Southern blot hybridization. Messenger RNAs for both vasopressin receptor subtypes V(1)a and V(2)were present in the telencephalon from embryonic day 12 to day 20. Both V(1)a and V(2)receptor mRNA increased on day 13 and then remained stable from embryonic day 13 to day 20. Messenger RNA for the vasopressin peptide was also detected in the telencephalon from day 12 to day 20, indicating that vasopressin could be synthesized within the rat cerebral cortex during rat embryonic development. Oxytocin receptor mRNA expression was also present in the telencephalon, but expression levels varied considerably from day 12 to day 20. No oxytocin mRNA expression was detected during rat telencephalon development. Temporal patterns of vasopressin receptor and vasopressin peptide mRNA expression along with oxytocin receptor mRNA suggest a temporal role for vasopressin- and oxytocin-mediated actions during rat telencephalon development.

p53 accumulation due to down-regulation of ubiquitin: relevance for neuronal apoptosis.

The p53 tumor suppressor protein is a major regulator of cell growth arrest and apoptosis in response to DNA damage. Both p53 function and stability are tightly controlled by Mdm2, which binds to the p53 N-terminus and targets p53 for ubiquitin-mediated proteolysis. Previous studies suggest that adrenalectomy-induced neuronal apoptosis is p53-dependent. Here we demonstrate both nuclear accumulation and functional activation of p53 protein in apoptotic hippocampal neurons from adrenalectomized rats. Increased p53 expression occurred despite the accumulation of its negative regulator, Mdm2, and the formation of p53-Mdm2 complexes. The persistence of p53 expression was explained by a striking decrease in free ubiquitin in p53-positive neurons. The addition of exogenous ubiquitin to p53-Mdm2 complexes from apoptotic neurons restored p53 degradation. These findings demonstrate a novel mechanism of p53 stabilization mediated by decreased ubiquitin levels. Regulation of free ubiquitin may therefore be an effective way to modulate p53-dependent apoptosis in certain cell types.

Expression of neuron-specific enolase in adult rat brain following status epilepticus.

Increased levels of neuron-specific enolase (NSE), a key glycolytic enzyme, in either the cerebrospinal fluid or the serum is correlated with both the duration and the outcome of status epilepticus. To further understand the molecular basis of seizure-induced elevations in NSE protein, we investigated NSE mRNA expression in the adult rat brain following systemic administration of kainic acid. The findings demonstrated either no change or a decrease in NSE gene expression during, and following, status epilepticus, suggesting that posttranscriptional mechanisms are responsible for seizure-induced increases in NSE protein.

Transforming growth factor-beta mediates astrocyte-specific regulation of brain endothelial anticoagulant factors.

BACKGROUND AND PURPOSE: Astrocytes are potent regulators of brain capillary endothelial cell function. Recently, astrocytes were shown to regulate brain capillary endothelial expression of the fibrinolytic enzyme tissue plasminogen activator (tPA) and the anticoagulant thrombomodulin (TM). To study the mechanism of this process, we examined the hypothesis that astrocyte regulation of endothelial tPA and TM is mediated by transforming growth factor-beta (TGF-beta). METHODS: Brain capillary endothelial cells were grown in blood-brain barrier models. We examined astrocyte-endothelial cocultures, endothelial monocultures, and astrocyte-conditioned media (ACM) for the expression of TGF-beta. We also incubated endothelial cells with ACM to determine the role of TGF-beta. Following 24 hours of incubation, we assayed for tPA and TM mRNA, as well as tPA and TM activity. RESULTS: Astrocyte-endothelial cocultures and ACM exhibited significantly higher levels of active TGF-beta than brain endothelial monocultures and endothelial cells grown in nonconditioned media, respectively. Brain endothelial cells incubated with ACM exhibited reduced tPA and TM mRNA and activity. Treatment with exogenous TGF-beta produced dose-dependent reductions in tPA and TM. The effects of ACM on both tPA and TM were blocked by TGF-beta neutralizing antibody. CONCLUSIONS: These data indicate that TGF-beta mediates astrocyte regulation of brain capillary endothelial expression of tPA and TM.

Status epilepticus induces p53 sequence-specific DNA binding in mature rat brain.

Previous studies have implicated the tumor suppressor gene, p53, in neuronal apoptosis due to excitotoxin treatment. To test whether p53 protein functions as a transcription factor during excitotoxic cell death, we used electrophoretic mobility shift assays to measure p53 sequence-specific DNA-binding activity following kainic acid (KA)-induced seizures. A rapid and significant increase in p53 DNA-binding activity was observed in extracts from kainate-vulnerable brain regions at 2.5 h after seizure onset, an effect which lasted up to 16 h after seizure-onset. DNA binding activity returned to normal by 30 h after KA injection. Pre-treatment with the protein synthesis inhibitor cycloheximide, as well as pre-incubation with PAb421, a p53 monoclonal antibody, significantly attenuated p53 DNA-binding activity induced by KA treatment. These results indicate that p53 protein may function as a transcription factor, following KA treatment, to regulate the expression of p53-responsive genes involved in neuronal apoptosis.

Astrocyte regulation of endothelial tissue plasminogen activator in a blood-brain barrier model.

Expression of tissue plasminogen activator (tPA) substantially determines endothelial-dependent fibrinolysis. We used a blood-brain barrier (BBB) model to analyze regulation of brain capillary endothelial tPA and its inhibitor, plasminogen activator inhibitor-1 (PAI-1). This model consists of coculture of murine astrocytes with bovine brain capillary endothelial cells grown as capillary-like structures (CS); after 1 week, astrocytes become extensively associated with CS, and the BBB-associated enzyme gamma-glutamyl transpeptidase is present. We measured tPA and PAI-1 mRNA and tPA activity in this model. Reverse transcription-polymerase chain reaction (RT-PCR) studies showed similar tPA and PAI-1 mRNA levels after 1 day mono-culture (endothelial cells only) versus astrocyte-endothelial coculture preparations. After 7 days (i.e., when elements of the BBB are present), astrocyte-endothelial cocultures (compared with endothelial mono-cultures) showed a 50.7%+/-27.1% (mean +/- SD) reduction in tPA mRNA (P < 0.03) and a 183.3%+/-86.9% increase in PAI-1 mRNA expression (P < 0.02). Moreover, 7-day cocultures demonstrated reduced tPA activity compared with mono-cultures (14.6+/-2.9 IU/mL versus 30.2+/-7.7 IU/mL, P < 0.01); 1-day cocultures and mono-cultures had similar tPA activity. These findings demonstrate that astrocytes regulate brain capillary endothelial expression of tPA when elements of the BBB phenotype are present in this model. These data suggest an important role for astrocytes in the regulation of brain capillary endothelial fibrinolysis.

Rat brain capillary thrombomodulin: structure and function.

The anticoagulant transmembrane glycoprotein thrombomodulin (TM) is expressed at the luminal surface of vascular endothelial cells. Recently, we showed that TM antigen and TM mRNA are expressed in brain microvessels in several species and that brain capillaries have the capability to activate protein C. The activation of protein C in brain microcirculation was greatly impaired by major stroke risk factors in rats due to downregulation of TM. In this study, a partial sequence of TM was determined from TM mRNA from brain capillaries examined in brain capillaries of the rat, a species that provides a useful model to investigate stroke mechanisms in relation to brain hemostasis. The predicted deduced amino acid sequences for rat TM were compared with other TM sequences. Particularly high homology (77-100%) among functional domains of the protein, i.e., the epidermal growth factor repeats (EGFRs) 1-6 and the transmembrane region, was observed between mice and rats. Somewhat less degree of homology was observed for bovine and human EGFRs 1-6, while the homology of the transmembrane region was 92-96%. All cysteine residues were conserved among the TM sequences, and specific amino acids previously suggested to be essential for activation of protein C by thrombin TM were highly conserved. We conclude that the highly conserved mRNA and protein sequences may reflect a similar anticoagulant role of TM in brain endothelial and systemic vascular endothelial cells across different species.

Immunohistochemical localization of tissue plasminogen activator in vascular endothelium of stroke-prone regions of the rat brain.

OBJECTIVE: Tissue plasminogen activator (tPA), a major regulator of fibrinolysis, is present in cerebrovascular endothelium. We have suggested that local regulation of tPA synthesis and release in brain microcirculation could be important determinants of the degree of damage after cerebral ischemia. In this study, the normal distribution of tPA antigen was determined in several stroke-prone regions in the rat brain often used to study the pathophysiological consequences of cerebral ischemia. METHODS: Immunohistochemistry and Western blot analysis were performed using an antibody that detects free tPA antigen and tPA complexed to its rapid inhibitor, plasminogen activator inhibitor-1 (PAI-1). Staining for von Willebrand factor, a brain endothelial cell marker, served as a positive control. RESULTS: Relative to von Willebrand factor, 8.6, 13, 11.4, and 20.4% of vessels in the parietal cortex, frontal cortex, striatum, and hippocampus, respectively, were tPA-positive. The majority of tPA-positive vessels (58-75%) were classified as precapillary arterioles and postcapillary venules (7-20 microm), whereas capillaries (4-7 microm) and small arterioles and venules (20-40 microm) accounted for 11 to 22% and 11 to 19%, respectively, of tPA-positive vessels. Western blot analysis of brain microvascular proteins confirmed the presence of free tPA (67 kDa) and a stronger band representing tPA-PAI-1 complexes. CONCLUSION: The tPA-containing cerebrovascular endothelium is distributed mainly in smaller vessels. In addition to the free pool of tPA, a large portion of tPA is complexed to PAI-1 and is therefore functionally inactive. The size of the free tPA cerebrovascular pool may be regulated by PAI-1, which in turn could suppress fibrinolysis in the cerebral microcirculation.

Immunohistochemical localization of redox factor-1 (Ref-1) in Alzheimer's hippocampus.

Redox factor-1 (Ref-1) is a dual-function protein involved in both DNA repair and transcriptional regulation. Ref-1 is modulated by cerebral ischemia and other oxidative stressors, and also regulates the DNA-binding activities of transcription factors implicated in Alzheimer's disease (AD)-related neurodegeneration. The present study examined Ref-1 expression in the AD hippocampus by immunohistochemistry. Although Ref-1 immunostaining was relatively low in control brain sections, senile plaques and other plaque-like structures in the AD brain were Ref-1-positive. Cells with increased Ref-1 immunoreactivity were also observed in regions of neuronal injury. These results suggest that Ref-1 might contribute to senile plaque formation, and that overexpression of Ref-1 in injured neurons may be part of a response to oxidative stress and an attempt to repair damaged DNA in AD.

Measurement of thrombomodulin mRNA expression in brain capillaries by polymerase chain reaction.

Thrombomodulin (TM), an endothelial integral membrane protein, is a potent activator of the protein C anticoagulant pathway. TM protein expression is limited and regionally distributed in the brain. Recent investigations have demonstrated low TM mRNA expression by brain endothelium, corresponding to its distribution at the protein level. To facilitate the study of TM expression at the transcriptional level, we measured TM mRNA by quantitative-competitive polymerase chain reaction (QC-PCR) and by standard densitometric analysis of reverse transcriptase-PCR products (RT-PCR) in different regions of bovine brain. QC-PCR demonstrated differential TM mRNA expression in the pons (100+/-9%), cerebellum (359+/-103%), and cortex (441+/-24%). We compared these results with those of RT-PCR and found similar differences in relative TM mRNA expression in the pons (100+/-44%), cerebellum (343+/-8%), and cortex (404+/-62%). Data derived by QC-PCR and RT-PCR were highly correlated (r=0.99, p<0.03). These findings indicate that either QC-PCR or RT-PCR can be used to accurately quantify TM mRNA.

Vasopressin-induction of the immediate early gene, NGFI-A, in cultured hippocampal glial cells.

Our earlier autoradiographic work had documented a wide distribution of vasopressin receptors in the hippocampus [R.E. Brinton, K.W. Gee, J.K. Wamsley, T.P. Davis, H.I. Yamamura, Regional distribution of putative vasopressin receptors in rat brain and pituitary by quantitative autoradiography, in: Proc. Natl. Acad. Sci. USA, 81 (1984) pp. 7248-7252; C. Chen, R.D. Brinton, T.J. Shors, R.F. Thompson, [Arg 8]-Vasopressin-induction of long lasting potentiation of synaptic transmission in the dentate gyrus, Hippocampus 3 (1993) 193-203.] which suggested the possibility that receptors for vasopressin were present in both neurons and glia. In the periphery, vasopressin is a potent mitogen in select proliferative cell types [E. Rozengurt, A. Legg, P. Pettican, Vasopressin stimulation of mouse 3T3 cell growth, Proc. Natl. Acad. Sci. USA, 76 (1979) pp. 1284-1287.] which also suggested a possible association between vasopressin receptor activation and the proliferative capacity of astrocytes. We therefore investigated whether vasopressin would induce the expression of the immediate early response gene, NGFI-A (also known as zif/268, ZENK, egr-1, krox 24), which is associated with initiation of mitogenesis [M. Sheng, M.E. Greenberg, The regulation and function of c-fos and other immediate early genes in the nervous system, Neuron, 4 (1990) pp. 477-485.]. Cultured hippocampal glial cells were exposed to vasopressin or a selective V1 vasopressin receptor agonist and in situ hybridization for NGFI-A mRNA was conducted. Results of these experiments demonstrated that vasopressin induced a highly significant dose-dependent increase in the number of cells expressing NGFI-A. Studies to determine the receptor subtype mediating vasopressin induction of NGFI-A were conducted utilizing the specific V1 agonist, [Phe2, Ile3, Orn8]-vasopressin. The V1 receptor agonist induced a highly significant dose dependent increase in the number of grains per NGFI-A positive cell. Time course analysis demonstrated that V1 agonist induction of NGFI-A occurred within 5 min, was maximally induced at 15 min of exposure and exhibited a gradual decline within 30 min of exposure which continued to decline over the 60 min time course. Glial cell responsivity was selective in that vasopressin and V1 agonist induction of NGFI-A occurred in a subpopulation of glial cells. Within a sea of glial cells, vasopressin and V1 agonist would induce islands of NGFI-A positive cells. Results of combined immunocytochemical labeling for the astrocyte specific marker, GFAP, and in situ hybridization for NGFI-A demonstrated that V1 agonist-induced NGFI-A expression occurred in GFAP positive cells. We observed no evidence for V1 agonist induction of NGFI-A in neurons. Collectively, these data document that vasopressin, acting via V1 vasopressin receptors, induces a highly significant increase in NGFI-A expression in select GFAP positive hippocampal astrocytes. To our knowledge, these data are the first report of a vasopressin mediated response in hippocampal glial cells. The potential functional significance of these findings is discussed.

Thrombomodulin expression in bovine brain capillaries. Anticoagulant function of the blood-brain barrier, regional differences, and regulatory mechanisms.

Thrombomodulin (TM), a key cofactor of the TM-protein C pathway, is of major biologic significance for the antithrombotic properties of endothelial cells. Yet, there is uncertainty whether TM is expressed in brain and what mechanisms govern brain endothelial anticoagulant activity. In this study, bovine brain capillaries were used as an in vitro model of the blood-brain barrier to determine factors involved in the regulation of TM expression in cerebral vasculature. Quantitative competitive-polymerase chain reaction assay revealed significant regional differences in the amount of brain capillary TM mRNA, i.e., cortical > cerebellar > pontine, consistent with the reverse transcription-polymerase chain reaction findings in which the abundance of TM mRNA was analyzed relative to beta-actin mRNA. Regional differences in TM mRNA brain capillary level correlated well with differences in protein C activation. The TM mRNA and activity were not detectable in brain parenchyma. Pathogenic mediators of ischemic stroke, interleukin 1 beta (10 U/mL), and tumor necrosis factor alpha (10 U/mL), produced a time-dependent decrease in brain capillary TM mRNA (t1/2 of 2.1 and 3.9 hours, respectively) and reduced endothelial TM activity. Incubation of brain capillaries with retinoic acid (10 mumol/L) and dibutyryl cAMP (3 mmol/L) resulted in a 4-fold increase in TM mRNA at 4 and 8 hours, respectively, followed by an increase in protein C activation. We conclude that TM at the blood-brain barrier is likely to be an important physiologic anticoagulant in brain microcirculation. Its downregulation by cytokines may contribute to ischemic brain damage and potentially could be counteracted by retinoic acid and cAMP.

A role for the tumour suppressor gene p53 in regulating neuronal apoptosis.

The tumour suppressor gene p53 is a nuclear phosphoprotein whose correct functioning is crucial for an appropriate cellular response to DNA damage. It has been suggested that p53 may act as a 'guardian of the genome' since when DNA damage is mild, p53 functions to halt cell cycle progression allowing DNA repair to occur before progression through the cell cycle. This prevents 'fixing' of lesions into the genome during replication. However when DNA damage is severe and irreversible, p53 induces the cell to undergo apoptosis. Recent studies have demonstrated DNA fragmentation and increased expression of p53 within neurons after injury. It appears that p53 expression may precede DNA fragmentation suggesting that rather than being induced in neurons in response to DNA damage, p53 expression may actually initiate neuronal apoptosis leading to DNA fragmentation. Recent reports documenting the resistance of neurons derived from p53-null mice (p53-/-) to excitotoxicity and DNA damaging agents both in vitro and in vivo and showing that p53 overexpression induces neuronal apoptosis in vitro support a role for the tumour suppressor gene p53 in regulating neuronal apoptosis. Here we review the recent evidence and discuss likely mechanisms involved in p53-mediated neuronal apoptosis.