Stimulation of p53 Transactivation Ability by Nicastrin in Mouse Fibroblasts

Nicastrin (NCSTN), a component of the γ-secretase complex, is involved in the p53-dependent apoptosis of neurons, although its mechanism remains unclear. We analyzed the effects of NCSTN transfection on the transactivity of p53 in ras-NIH3T3 mouse fibroblasts with a luciferase assay. Luciferase activity was elevated after transfection, suggesting the stimulation of p53 transactivation ability. In addition, the protein levels of endogenous mouse p53 and transfected human p53 increased. The effects of NCSTN appeared to be independent of γ-secretase activity because it was not inhibited by the γ-secretase inhibitor DAPT. The functional domains of NCSTN were further examined with NCSTN deletion mutants. Activation of the p53-responsive promoter was completely diminished in a NCSTN mutant lacking the amino acid residues between 306 and 360. Since this domain is a γ-secretase-substrate-recognition site, the activation of p53 by NCSTN may be mediated by γ-secretase-substrate-like molecules.


Introduction
Alzheimer's disease (AD) is a neurodegenerative disease that causes a progressive dementia.The pathological feature of AD is the accumulation of senile plaques composed of extracellular amyloid-β (Aβ) peptides in the cerebral cortex.The senile plaques are believed to result in the apoptosis and synaptic dysfunction of neurons resulting in AD [1].Aβ is generated from amyloid precursor protein (APP) by two distinct steps of proteolytic processing.APP is first cleaved by a β-site APP cleaving enzyme, β-secretase, and subsequently by γ-secretase within its transmembrane domain [2].γ-secretase activity resides in a high molecular weight multimeric protein complex composed of at least four core components, presenilin, nicastrin (NCSTN), anterior pharynx defective-1 (Aph-1) and presenilin enhancer-2 (Pen-2) [3].In the active γ-secretase complex, presenilin exists as a heterodimer of the N-and C-terminal fragments generated by endoproteolysis [4].Aph-1 is implicated in the assembly of γ-secretase complex as a scaffold protein [5], and the N-terminal domain of Pen-2 regulates the endoproteolysis in the γ-secretase complex [6].NCSTN is a type-I membrane protein and its amino acid residues 261-502 in the ectodomain show sequence similarity to a peptidase family that includes aminopeptidases, carboxypeptidases, and transferrin receptor proteins.Amino acid residues 306-360 which span the most conserved sequence amongst nicastrin orthologs are indispensable for its function as a γsecretase-substrate receptor [7].Although APP, Notch, and several other transmembrane proteins undergo regulated intramembranous cleavage by γ-secretase, the substrate specificity remains unclear [6,[8][9][10].
The tumor suppressor protein p53, through its elevation and posttranslational modification in response to various types of cellular stresses, induces either cell cycle arrest by transactivating p21 or apoptosis by transactivating Bax, Noxa and PUMA [11][12][13][14][15][16].It was reported that p53 is overexpressed in the brains of patients with AD, and that intracellular Aβ42 stimulated p53 expression [17,18].Therefore, p53 may be involved in neuronal cell apoptosis in the brain of patients with AD.In the present study, we show that NCSTN stimulated p53 transactivation ability independently of γ-secretase catalytic activity.
The cells seeded on 24-well plates were then transfected with test genes together with firefly luciferase and Renilla luciferase reporter plasmids by using LipofectAMINE-Plus (Invitrogen, San Diego, CA).Two days after the transfection, firefly and Renilla luciferase activities were determined with a Dual Luciferase Assay System (Promega) and a luminescencer (Atto, Tokyo, Japan).Firefly luciferase activities were normalized with the Renilla luciferase control activities.

Stimulation of p53 Transactivation Ability by NCSTN.
Because p53 is overexpressed in the brains of patients with AD [17], we examined the effects of APP and γ-secretase components on the transactivation ability of p53 in ras-NIH3T3 mouse fibroblasts with a luciferase assay using p53responsive reporter plasmid pG13-Luc.Luciferase activity was stimulated considerably by NCSTN and slightly by Aph-1 and Pen-2 (Figure 1(a)).Presenilin and APP did not increase the luciferase activity significantly.Co-transfection with two of these γ-secretase components showed additive but not synergistic stimulation of the luciferase activity (data not shown).NCSTN activated the p53-responsive promoter in a dose dependent manner not only in ras-NIH3T3 cells but also in MKN45 human gastric carcinoma cells (Figures 1(b) and 1(c)).As compared to pG13-Luc, control pG15-Luc that contained mutations in the p53-responsive consensus sequence [21] showed much lower (less than 1%) luciferase activity (data not shown), indicating that most, if not all, of the observed luciferase activity derived from pG13-Luc is p53 dependent.These results suggested that the transactivation ability of p53 is stimulated by forced expression of NCSTN.
Then, we analyzed the protein expression levels in transfected mouse ras-NIH3T3 cells by Western blot analysis.As a control, presenilin-transfected cells as well as the emptyvector-transfected cells were also examined.The protein levels of NCSTN were much higher in NCSTN expression plasmid-transfected cells as compared with control plasmid-or presenilin expression plasmid-transfected cells (Figure 1(d)).The expression of presenilin was observed exclusively in presenilin-transfected cells.The levels of endogenous mouse p53 increased with transfection of NCSTN expression plasmid (Figures 1(d) and 1(e)).
When human p53 cDNA was cotransfected, the protein expression levels of human p53 were also elevated by NCSTN (Figures 1(f) and 1(g)).Thus, the elevated expression levels of p53 protein may account for the activation of p53 reporter by NCSTN.Analysis using anti-p53 phosphospecific antibodies revealed that NCSTN increased phosphorylation levels at multiple sites such as Ser-6, Ser-9, Ser-315 and Ser-392.On the other hand, presenilin stimulated the phosphorylation only at Ser-392 but decreased the levels at multiple sites.
We also analyzed the mRNA levels of p53 in the NCSTNexpressing cells by RT-PCR.The levels of p53 mRNA in NCSTN-expressing cells were similar or slightly lower as compared to those of control plasmid-transfected cells (Figures 1(h) and 1(i)).The effects of NCSTN on the transcription of the p53 gene were also examined with a luciferase assay using a reporter plasmid containing the promoter region of the p53 gene.Transfection with NCSTN cDNA did not increase but rather decreased luciferase activity (Figure 1(j)).These results suggest that the elevated p53 protein levels in NCSTN-expressing cells were not attributed to the increased transcription of the p53 gene but were caused by posttranscriptional regulation.
The target specificity of p53 was examined with reporter plasmids containing p53-responsive sequences in the promoter region of p21, Bax, Noxa and PUMA.Bax and Noxa promoters were significantly activated by transfection with NCSTN cDNA whereas the increases of gene expression via p21 and PUMA promoters were only marginal and under the detectable level, respectively (Figure 1(k)).

γ-Secretase-Independent Activation of p53 by NCSTN.
Notch is sequentially cleaved by TNFα-converting enzyme (TACE) and γ-secretase to produce Notch intracellular region (Notch-IC) which functions as a transcription factor [9].The transactivation ability of Notch-IC can be estimated by luciferase assay using the pTP1-Luc reporter plasmid which contains the Notch-binding domains [23].The luciferase activity was markedly elevated by transfection with the positive control Notch-ΔE (Figure 2(a)) which lacks the TACE-eliminated domain and can, therefore, be activated only by γ-secretase [22].The expression of Notch-ΔE increased the reporter luciferase activity 100-fold, whereas without Notch-ΔE, the reporter activity was only marginal irrespective of the enforced expression of NCSTN (Figure 2(a)), suggesting that the activation of endogenous Notch by NCSTN was only limited.In the presence of Notch-ΔE, NCSTN caused no further increase of the reporter activity.Thus, the enforced expression of NCSTN may not increase the endogenous γ-secretase activity.This also implies that the activation of p53 by NCSTN is not mediated by Notch.

Involvement of cAMP Signal in the Activation of p53 by NCSTN.
We then examined the signaling pathway of p53 activation by NCSTN using various probes that affect signal transduction pathways such as protein kinase C, cAMP-dependent protein kinase (PKA), MAP kinase, Akt kinase, NF-κB, Cdk and Ras family pathways [33].It was shown thus far that the effects of NCSTN were attenuated by DBcAMP, IBMX and forskolin which increased intracellular cAMP concentration (Figure 4(a)).The involvement of cAMP was further confirmed by cotransfection with cDNA of the constitutively active catalytic subunit of PKA (PKA-CS), which reduced both basal and NCSTN-stimulated p53 activity (Figure 4 effects of NCSTN partly.Thus, the activation of p53 by NCSTN might be mediated by cAMP and Rho signaling pathways.

Discussion
Senile plaque is a typical pathological feature of AD, and thus may be implicated in the pathogenesis of AD, although the precise mechanism remains unclear.It has been shown that p53 is highly expressed in the brain of patients with AD [17,18].Presenilin, Aph-1, Pen-2, and NCSTN form γ-secretase complex, which cleaves amyloid precursor protein to produce amyloid β-peptide (Aβ) [32].Generation of an excess amount of Aβ activates p53 and induces neurodegeneration [18].p53-associated cell death in Alzheimer disease was dependent on presenilin [34].However, on the other hand, overexpression of Aph-1 and Pen-2 reduced p53 expression and activity and suppressed apoptosis [35], and consistently, knockdown of Pen-2 resulted in p53-dependent apoptosis [36].Dunys et al. suggested the importance of the integrity of γ-secretase complex [35].Unusual increase of one of the components could cause distinct effects independently of γsecretase complex.
In the present study, we examined the effects of γsecretase components on p53 with a luciferase assay and found that p53-responsive reporter plasmid pG13-Luc was significantly activated by NCSTN (Figures 1(a)-1(c)).This result suggested that p53 transactivation ability was enhanced by NCSTN.This putative activation of p53 by NCSTN was accompanied by the increase not only in p53 protein expression levels but also in phosphorylation levels at multiple sites in p53 (Figures 1(d)-1(g)) without any increase in p53 mRNA levels (Figures 1(h) and 1(i)).NCSTN also activated the reporter under the control of p53-responsive promoter sequences of proapoptotic Bax and Noxa but not antiapoptotic p21 (Figure 1(k)).This is consistent with the neuronal cell death observed in AD brains.The increased p53 by NCSTN may be physiologically relevant because the anchorage-independent growth in soft agar medium was suppressed in NCSTN-transfected ras-NIH3T3 cells as compared to control vector-transfected cells (Supplementry Figure 1, Supplementry Material available on doi: 10.3814/2010/606391).
The reporter plasmid pTP1-Luc is specifically activated by Notch-IC [9] which is produced by cleavage of Notch-ΔE by γ-secretase [22].A substantial activation of this reporter by transfection with Notch-ΔE (Figure 2(a)) suggested that γ-secretase activity was present in ras-NIH3T3 cells.Transfection with with NCSTN caused no further increase in the luciferase activity (Figure 2(a)), suggesting that NCSTN is not a limiting factor for γ-secretase activity in this cell.The activation of pG13-Luc by NCSTN was not inhibited by γsecretase inhibitor DAPT despite the inhibition of Notch-ΔE-induced activation of pTP1-Luc by DAPT (Figures 2(b) and 2(c)).Similar results were obtained with another γsecretase inhibitor L-685,458 (data not shown).Accordingly, the activation of p53 by NCSTN may be independent of the catalytic activity of γ-secretase.
Recently, Pardossi-Piquard et al. have reported that NCSTN lowers p53 expression as well as p53 transcriptional activity independently of γ-secretase activity [37].They also reported the reduction of p53 promoter activity by NCSTN, which is consistent with our results (Figure 1(j)).We frequently observed such reduced p53 promoter activity after activation of p53 (data not shown), and therefore, decrease in the p53 promoter activity might be a result of feedback regulation after elevation of the expression and/or transactivation ability of p53.Decrease in p53 expression might be a secondary effect caused by reduced transcription of the p53 gene.It is possible that activation of p53 is the primary effect of NCSTN.
Comparison of NCSTN deletion mutants suggested that the activation of p53 by NCSTN is dependent on the substrate receptor domain, amino acid residues 306-360, and amino acids 419-458 of NCSTN (Figure 3).Thus far, many substrates for γ-secretase have been identified [38,39].Without proteolytic cleavage, simple binding and trapping of some substrates on the cell surface by NCSTN might provoke a signaling pathway which may lead to the increase in the protein expression levels and transactivation ability of p53.
Among the signaling pathways examined thus far, activation of PKA reproducibly reduced the stimulation of p53 transactivity by NCSTN (Figure 4).PKA activity is decreased in AD brains [40], and PKA-activating reagents were recently reported to be therapeutic drugs for AD [41].The rationale for this therapy may be partly explained by the suppression of NCSTN-mediated activation of p53 by PKA-activating gene (PKA-CS) and drugs (DBcAMP, IBMX and forskolin).

Figure 2 :
Figure 2: γ-secretase inhibitor did not suppress NCSTN-induced activation of the p53-responsive promoter.(a) γ-secretase is active in ras-NIH3T3 cells.ras-NIH3T3 cells were cotransfected with Notch-responsive reporter plasmid pTP1-Luc and expression vectors of control, NCSTN or Notch-ΔE as indicated.Cells were harvested 48 hours after transfection and luciferase activities in the cell extracts were measured.The error bars represent SD (n = 3).(b) Inhibitory effects of γ-secretase inhibitor DAPT on Notch transactivation ability.ras-NIH3T3 cells were transfected with a Notch-responsive reporter plasmid pTP1-Luc and expression vectors of NCSTN, Notch-ΔE or control as indicated.Twenty-four h after transfection, cells were treated with DAPT at the indicated concentrations for another 24 hours.Cells were harvested 48 hours after transfection and luciferase activities in the cell extracts were measured.The error bars indicate SD (n = 3).(c) Lack of inhibition of DAPT on NCSTN-induced p53 transactivation ability.ras-NIH3T3 cells were transfected with a p53-responsive reporter plasmid pG13-Luc and an expression vector for NCSTN or its empty vector, and treated or not treated with DAPT (25 μM).Cells were harvested 48 hours after transfection and luciferase activities in the cell extracts were measured.The error bars represent SD (n = 3).

Figure 3 :Figure 4 :
Figure 3: Analysis of NCSTN domain functions.(a) Structures of wild-type (WT) and mutant NCSTN.Deletion mutants of NCSTN were constructed as described in the Materials and methods.Deleted regions in NCSTN Δ195-305, NCSTN Δ306-360, NCSTN Δ361-418, NCSTN Δ419-458 and NCSTN Δ625-662 are shown.SP and TM represent the signal peptide and the transmembrane domain, respectively.Striped regions between amino acid positions 306 and 360, 419 and 458, and 625 and 662 are highly conserved sequences amongst NCSTN orthologs.(b) Comparable expression levels of NCSTN in mutant-NCSTN transfectants.ras-NIH3T3 cells were transfected with cDNA of wild-type or mutant NCSTN.Cells were harvested 48 hours after transfection, and the NCSTN protein expression levels were examined by Western blotting.mNCSTN and hNCSTN represent endogenous mouse NCSTN and ectopically expressed human NCSTN, respectively.(c) Effects of NCSTN mutants on the p53-responsive promoter.ras-NIH3T3 cells were cotransfected with p53-responsive reporter plasmid pG13-Luc and the expression plasmid for wild-type or mutant NCSTN.Cells were harvested 48 hours after transfection and the luciferase activities in the cell extracts were measured.The error bars represent SD (n = 3, t-test; * P < .01).