The p53 Family Member Genes Are Involved in the Notch Signal Pathway*

The p53 tumor suppressor is a transcription factor that regulates cell growth and death in response to environmental stimuli such as DNA damage. p63/p51 and p73 were recently identified as members of the p53 gene family. In contrast to p53 however, p63 and p73 are rarely mutated in human cancers. Mice that lack p53 are developmentally normal, while p63 and p73 appear to play critical roles in normal development. To determine how p63 and p73 are involved in normal development, we attempted to identify target genes that are specifically regulated by p63 and/or p73 but not by p53. We found that the Jagged1 ( JAG1 ) and Jagged2 ( JAG2 ) genes, encoding ligands for the Notch receptors, are up-regulated by p63 and p73. Furthermore, we identified a p63-binding site in the second intron of the JAG1 gene, which can directly interact with the p63 protein in vivo , as assessed by a chromatin immunoprecipitation assay. Santa Cruz Biotechnology); mouse anti-human p73 monoclonal antibody (ER-15, Oncogene Research); mouse anti-human p63 monoclonal antibody (4A4, Oncogene Research); and rabbit anti- human JAG1 polyclonal antibody (H-114, Santa Cruz Biotechnology). For Northern blot analysis, total RNA (10 (cid:4) g) was electrophoretically separated on a 1% agarose gel containing 2.2 M formaldehyde and blotted on a nitrocellulose membrane (Schleicher & Schuell). was dish and infected with Ad-p53 or Ad-p63 (cid:3) . After 24 h, genomic DNA and protein were cross-linked by addition of formaldehyde (1% final concentration) directly to culture medium and incubated for 15 min at 37 °C. Cells were lysed in 200 (cid:4) l of SDS lysis buffer with a protease inhibitor mixture and sonicated to generate 300–800 bp DNA fragments. After centrifugation, the cleared supernatant was diluted 10-fold with the ChIP dilution buffer and incubated with the specific antibody at 4 °C for 16 h. Immune complexes were precipitated, washed, and eluted as recommended. DNA-protein cross-links were reversed by heating at 65 °C for 5 h, and DNA was phenol-extracted, ethanol-precipitated, and resuspended in 50 (cid:4) l of TE. Five microliters of each sample were used as a template for PCR amplification. PCR amplifications of the second intron of the JAG1 gene containing the consensus p53-binding sequences ( (cid:1) 5597, renamed the RE-JAG1 sequence and (cid:1) 5983) were performed on immunoprecipitated chromatin using the specific primers 5 (cid:3) -ACCTTTCACCATTCCCCTAC-3 (cid:3) (for-ward) and 5 (cid:3) -GCCCAAGGACAAAATAGCCA-3 (cid:3) (reverse), and 5 (cid:3) - CCATTGTGCACAGTCAGTCG-3 (cid:3) (forward) and 5 (cid:3) -CCACTCGCATT-GTGTGCATG-3 (cid:3) (reverse), respectively. PCR amplification of the p21 promoter was performed using oligonucleotides 5 (cid:3) -ACCTTTCACCAT-TCCCCTAC-3 (cid:3) (forward) and 5 (cid:3) -GCCCAAGGACAAAATAGCCA-3 (cid:3) (re- verse), as described (24). To ensure that PCR was linear range, DNA amplified for a maximum of 30 cycles. PCR products following ChIP assay were sequenced to verify the identity of the amplified DNA. Luciferase Assay— A 51-bp fragment of the RE-JAG1 (5 (cid:3) -AGGCTT-CTTGTTCAGGCTTGCTCTGTGTGAACCAGACCGTTGTGCTTGGCT-3 (cid:3) and its mutant form (5 (cid:3) -AGGATTCTTGTTCAGGATTTCTCTGTG- TGAAACATACCGTTGTGATTTGCT-3 (cid:3) of a basal SV40 promoter in the pGL3-promoter vector (Promega), and the constructs designated pGL3-RE-JAG1 and pGL3-RE-JAG1-mut, 1 (cid:2) 10 of H1299 cells in 6-cm co-transfected with 1.5 (cid:4) g of pGL3-RE-JAG1 or pGL3-RE-JAG1-mut, with 1.5 (cid:4) g of a pcDNA3.1 control vector (Invitrogen) or a vector that expresses p53, p73 (cid:2) , or p63 (cid:3) Lipofec-tin (Life Technologies). Cells were harvested 48 h after transfection followed by measurement of luciferase activity using the Luciferase Assay System (Promega). The ability to stimulate transcription was defined as the ratio of luciferase activity in the cells transfected with the pGL3-RE-JAG1 relative to the activity in the cells transfected with the non-responsive reporter plasmid, pGL3-RE-JAG1-mut. As a control experiment, pGL3-p53CBS containing three copies of the consensus p53-binding sequence or its mutant form (11) were transfected into H1299 cells with a pcDNA3.1 control vector or a vector that expresses p53, p73 (cid:2) , or p63 (cid:3) . All experiments were performed in triplicate

The involvement of the p53 tumor suppressor gene in cell growth and death is mediated by the transactivation of p53target genes in response to environmental stimuli such as DNA damage (1)(2)(3). p63/p51 and p73 were recently identified as members of the p53 gene family and encode proteins that share considerable structural homology with p53 (4 -6). p63 and p73 can bind to the p53-responsive elements and up-regulate some p53-target genes, which suggest that the p53 family members have a potential for functional overlap with p53 itself (7)(8)(9)(10)(11)(12). However, in contrast to p53, p63 and p73 are rarely mutated in human cancers (13)(14)(15).
Different phenotypes between p63-or p73-deficient and p53deficient mice were also reported (16 -19). In contrast to p53deficient mice, mice lacking the p73 genes show no increased susceptibility to spontaneous tumorigenesis. p73-deficient mice have neurological, pheromonal and inflammatory defects. p63deficient mice have major defects in their limbs and craniofacial development, as well as a striking absence of stratified epithelia, suggesting that p63 is required for limb and epidermal morphogenesis. In humans, Li-Fraumeni syndrome patients have inherited mutations of the p53 gene and develop normally, but are predisposed to cancer (20), while heterozygous germline mutations in the p63 gene are the cause of ectrodactyly, ectodermal dysplasia, and facial clefts syndrome (21). These studies demonstrate a marked divergence in the developmental roles of p63 and p73 and further distinguished these p53 family genes from p53. Despite these revelations, little is known about target genes specifically regulated by p63 or p73. Identifying the gene targets of p63 and p73 that play a role in development is an important step to a better understanding of the roles of these proteins in normal development and developmental disorders.
To determine what mediates the function of the p53 family members on normal development, we attempted to identify genes that are specifically regulated by p63 and/or p73 but not by p53. As a result, we identified the JAG1 and JAG2 genes as targets for p63 and p73. Thus, there are likely to be differences among the p53 family members with regard to their optimal DNA-binding sequences. We also showed p63-mediated JAG1 induction can activate Notch signaling in neighboring cells. This study is the first to implicate a link between the p53 family member genes and Notch signaling.

EXPERIMENTAL PROCEDURES
Cell Lines and Recombinant Adenovirus-The human cancer cell lines used in this study were purchased from American Type Culture Collection or the Japanese Collection of Research Bioresources (Osaka, Japan). All cell lines were cultured under conditions recommended by their respective depositors. The endogenous p53 statuses in these lines are wild type for A172 and HCT116, mutant for DLD1, colo320, and PLC/PRF5, and p53 null for Saos2 and H1299. The generation and purification of replication-deficient recombinant adenoviruses containing p53, p73␣, p73␤, p63␣, and p63␥ genes or the bacterial lacZ gene were described previously (11).
Immunoblot and Northern Blot Analysis-The primary antibodies used in this study are as follows: mouse anti-human p53 monoclonal antibody (DO-7, Santa Cruz Biotechnology); mouse anti-human p73 monoclonal antibody (ER-15, Oncogene Research); mouse anti-human p63 monoclonal antibody (4A4, Oncogene Research); and rabbit antihuman JAG1 polyclonal antibody (H-114, Santa Cruz Biotechnology). For Northern blot analysis, total RNA (10 g) was electrophoretically separated on a 1% agarose gel containing 2.2 M formaldehyde and blotted on a nitrocellulose membrane (Schleicher & Schuell). RNA was visualized with ethidium bromide to ensure that it was intact and loaded in similar amounts and to confirm proper transfer. Hybridization was performed as described previously (11). cDNA probes for JAG1 (nucleotides (nt) 1 3531-4534), JAG2 (nt 419 -1323), p21 (nt  and HES-1 (nt 286 -642) were amplified by the reverse transcription-PCR from appropriate cDNA pools. PCR products were sequenced to verify their identity.
cDNA Microarray-For cDNA expression arrays, poly(A)ϩ RNA was isolated with the FastTrack 2.0 mRNA isolation system (Invitrogen) from adenovirus-infected A172 human glioma cells and used as a template for synthesis of Cy3-or Cy5-labeled cDNA probes. The probes were hybridized to cDNA microarrays containing 9216 genes. Microarray construction, hybridization procedures, and data analysis were described previously (22,23).
Immunofluorescence Microscopy-Cells grown on glass coverslips were fixed in cold absolute acetone for 10 min. Following fixation, cells were washed in PBS then blocked in PBS plus 5% goat serum. Primary antibody (H-114) was diluted to 5 g/ml in the presence of Triton X-100 and incubated for 2 h at room temperature. Cells were washed in PBS and then incubated with secondary antibody Alexa 488-conjugated goat anti-rabbit IgG (Molecular Probes) for 30 min at room temperature. The specimens were examined using a laser-scanning confocal microscope (MRC 1024; Bio-Rad).
Chromatin Immunoprecipitation Assay (ChIP)-ChIP assay was performed using the Acetyl-Histone H3 ChIP Assay Kit (Upstate Biotechnology) as recommended by the manufacturer except that antibodies against p53 (DO-7), p63 (4A4), and FLAG tag peptide (M2, Sigma-Aldrich) were used in this study. 2 ϫ 10 6 Saos2 cells were plated onto a 10-cm dish and infected with Ad-p53 or Ad-p63␥. After 24 h, genomic DNA and protein were cross-linked by addition of formaldehyde (1% final concentration) directly to culture medium and incubated for 15 min at 37°C. Cells were lysed in 200 l of SDS lysis buffer with a protease inhibitor mixture and sonicated to generate 300 -800 bp DNA fragments. After centrifugation, the cleared supernatant was diluted 10-fold with the ChIP dilution buffer and incubated with the specific antibody at 4°C for 16 h. Immune complexes were precipitated, washed, and eluted as recommended. DNA-protein cross-links were reversed by heating at 65°C for 5 h, and DNA was phenol-extracted, ethanol-precipitated, and resuspended in 50 l of TE. Five microliters of each sample were used as a template for PCR amplification. PCR amplifications of the second intron of the JAG1 gene containing the consensus p53-binding sequences (ϩ5597, renamed the RE-JAG1 sequence and ϩ5983) were performed on immunoprecipitated chromatin using the specific primers 5Ј-ACCTTTCACCATTCCCCTAC-3Ј (forward) and 5Ј-GCCCAAGGACAAAATAGCCA-3Ј (reverse), and 5Ј-CCATTGTGCACAGTCAGTCG-3Ј (forward) and 5Ј-CCACTCGCATT-GTGTGCATG-3Ј (reverse), respectively. PCR amplification of the p21 promoter was performed using oligonucleotides 5Ј-ACCTTTCACCAT-TCCCCTAC-3Ј (forward) and 5Ј-GCCCAAGGACAAAATAGCCA-3Ј (reverse), as described (24). To ensure that PCR was performed in linear range, template DNA was amplified for a maximum of 30 cycles. PCR products following ChIP assay were sequenced to verify the identity of the amplified DNA.
Luciferase Assay-A 51-bp fragment of the RE-JAG1 (5Ј-AGGCTT-CTTGTTCAGGCTTGCTCTGTGTGAACCAGACCGTTGTGCTTGGCT-3Ј) and its mutant form (5Ј-AGGATTCTTGTTCAGGATTTCTCTGTG-TGAAACATACCGTTGTGATTTGCT-3Ј) were synthesized and inserted upstream of a basal SV40 promoter in the pGL3-promoter vector (Promega), and the resulting constructs were designated pGL3-RE-JAG1 and pGL3-RE-JAG1-mut, respectively. 1 ϫ 10 6 of H1299 cells in 6-cm dishes were co-transfected with 1.5 g of either pGL3-RE-JAG1 or pGL3-RE-JAG1-mut, together with 1.5 g of a pcDNA3.1 control vector (Invitrogen) or a vector that expresses p53, p73␤, or p63␥ using Lipofectin (Life Technologies). Cells were harvested 48 h after transfection followed by measurement of luciferase activity using the Luciferase Assay System (Promega). The ability to stimulate transcription was defined as the ratio of luciferase activity in the cells transfected with the pGL3-RE-JAG1 relative to the activity in the cells transfected with the non-responsive reporter plasmid, pGL3-RE-JAG1-mut. As a control experiment, pGL3-p53CBS containing three copies of the consensus p53-binding sequence or its mutant form (11) were transfected into H1299 cells with a pcDNA3.1 control vector or a vector that expresses p53, p73␤, or p63␥. All experiments were performed in triplicate and repeated on at least three independent occasions.
Co-culturing Assays-Saos2 adherent cells (2 ϫ 10 6 ) were incubated with purified virus at a m.o.i of 25. After 12 h, the cells were washed three times with medium, and overlaid with Jurkat non-adherent cells (3 ϫ 10 6 ) in suspension. The co-cultures were incubated for 18 h, and Jurkat cells were harvested as described previously (25). Briefly, culture medium and unattached cells were removed from each plate, and the remaining cells were washed two times with PBS to harvest more tightly bound Jurkat cells. Total RNA was extracted from the Jurkat cells for Northern blotting analysis.

p63 and p73 Induce Expression of the JAG1 and JAG2
Genes-To express the p53 family genes in human cancer cell lines, we used the replication-deficient adenoviral vector harboring human p53 (Ad-p53), p73␤ (Ad-p73␤), and p63␥/p51A (Ad-p63␥/Ad-p51A) genes. To determine the relative efficiency of adenovirus-mediated gene transfer, cells were infected with adenovirus containing the bacterial lacZ gene (Ad-lacZ). We used seven human cancer cell lines that showed highly efficient gene transfer, with 90 -100% of the cells staining for ␤-galactosidase activity at a m.o.i. of 50 -100. A high-level p53 protein was observed in cells infected with Ad-p53 (examples in Fig. 1, upper panel). Infection with Ad-p73␤ and Ad-p63␥ resulted in expression of exogenous p73␤ and p63␥ proteins, respectively (examples in Fig. 1, middle and lower panels). We used p73␤ and p63␥/p51A isoforms in this study because we and others have demonstrated that transcription of a p53-responsive reporter gene was activated more strongly in p73␤ and p63␥/ p51A than p73␣ and p63␣/p51B (5,11,26).
In an effort to identify specific targets regulated by p73 and p63, we performed cDNA microarray analysis and compared expression patterns in a human glioma cell line A172 transfected separately with Ad-p53, Ad-p73␤, and Ad-p63␥. Using this approach, we detected several genes that were reproducibly activated in p73-and p63-transfected cells but not in p53-transfected cells. Two of the genes that were selectively induced by p63 and p73 were JAG1 and JAG2. Northern blot analysis using the JAG1 and JAG2 cDNAs as probes demonstrated that expression of the JAG1 gene was dramatically increased by infection with Ad-p63␥ and Ad-p73␤ in a time-dependent manner, but not by infection with Ad-p53 (Fig. 2). The JAG1 induction was seen as early as 9 h after Ad-p63␥ infection, similar to induction of p21, a well defined target for both p53 and its family members (Fig. 2) (1, 4, 6). The early induction of JAG1 suggests that the JAG1 gene may be a direct target of transcriptional activation by p73 and p63.
We then examined the level of JAG1 protein by Immunoblot analysis using an antibody against the intracellular domain of human JAG1. Fig. 4A shows the high-level accumulation of endogenous JAG1 protein in Ad-p63␥-infected Saos2 cells and Ad-p73␤-and Ad-p63␥-infected A172 cells, consistent with the Northern blot analysis (Fig. 3). Immunofluorescence staining using the same antibody also verified the expression of endogenous JAG1 in Ad-p63␥-infected Saos2 cells but not in control or Ad-lacZ-infected cells. The subcellular distribution of JAG1 coincided with that reported previously (27). In contrast, Ad-p53 infection resulted in only a slight increase in JAG1 expression (Fig. 4B). To determine whether expression of other Notch ligands and receptors are regulated by the p53 family genes, Northern blot analysis was performed. Neither Notch ligands (DLL1, DLL3, DLL4) nor Notch receptors (Notch1, Notch2, Notch3, Notch4) were significantly induced by p73, p63, or p53 in A172 and DLD1 cells (data not shown).
Identification of a Specific Target Sequence for p63 in the JAG1 Gene-The early and strong induction of JAG1 suggests that JAG1 may be a direct and specific target of transcriptional activation by p63. To address this hypothesis, we searched for a consensus p53-binding sequence in the JAG1 gene because the p63 and p73 proteins can also bind to the p53-binding sites (4 -7, 28). We obtained the genomic sequence of the human JAG1 gene from the GenBank TM data base (accession number AL035456) and searched for a consensus p53-binding site(s) within 10 kb in and around exon 1 of the JAG1 gene. Ten candidate sequences were identified at the positions Ϫ5162, Ϫ3133, Ϫ2098, Ϫ1555, Ϫ1131, ϩ2814, ϩ5597, ϩ5983, ϩ7637, and ϩ8102, where ϩ1 represents the translation initiation site.
To determine whether the p63 protein can selectively bind to any of these candidate binding sites in vivo, we performed ChIP assays. The ChIP assay relies on the ability of specific antibodies to immunoprecipitate DNA-binding proteins along with the  1 and 5), Ad-p53 (lanes 2 and 6), Ad-p73␤ (lanes 3 and 7), or Ad-p63␥ (lanes 4 and 8). Cell extracts (30 g) were separated by electrophoresis on a 10% SDS/acrylamide gel and analyzed by Western blotting using the JAG1 C-terminal antibody H-114. B, immunofluorescence analysis of JAG1 protein expression of control Saos2 cells (a) and Saos2 cells 24 h following infection with Ad-lacZ (b), Ad-p53 (c), or Ad-p63␥ (d) is shown. Cells were stained with the JAG1 C-terminal antibody H-114. Omission of the primary antibody resulted in no staining (data not shown). associated genomic DNA. We used Saos2 cells infected with either Ad-p53 or Ad-p63␥ for ChIP assay, since the induction of JAG1 by p63␥ was strongest in this line (Figs. 3 and 4). Immunoprecipitation of DNA-protein complex using antibodies against p53 and p63 was performed on formaldehyde-crosslinked extract from Ad-p53-and Ad-p63␥-infected Saos2 cells, respectively. We then measured the abundance of candidate sequences within the immunoprecipitate complexes by PCR. ChIP assay revealed that one DNA fragment containing a candidate sequence, ϩ5597, was reproducibly present in a complex with p63␥ protein (Fig. 5B, middle panel, lane 3). We designated this p63-binding sequence RE-JAG1 (for responsive element in JAG1). RE-JAG1 consists of four copies of the consensus 10-bp motif separated by 3, 5, and 3 bp, respectively (Fig. 5A). In contrast, p53 protein binding to RE-JAG1 sequence was not detectable in p53-infected cells, as assessed by the ChIP assay with an anti-p53 antibody and subsequent PCR (Fig. 5B, middle panel, lane 6). The other nine candidates were amplified in the input-positive control for PCR, but not in the immunoprecipitated samples with an antibody against p53 or p63 (examples in Fig. 5B, lower panel). As a positive control for the ChIP assay, we analyzed the interaction of p53 and p63 with the p21 promoter. Both p53 and p63 proteins immunopre-cipitated with the DNA fragment containing the p53-binding site in the p21 promoter (Fig. 5B, upper panel, lanes 3 and 6). Although we can not exclude a low level of p53 binding to the RE-JAG1 sequence, our ChIP data indicate that p63 is selectively associated with this site in vivo.
To determine whether RE-JAG1 possesses p63-dependent transcriptional activity, we performed a heterologous promoter-reporter assay using a luciferase vector prepared by cloning the oligonucleotide corresponding to RE-JAG1 sequence upstream of a basal SV40 promoter (see "Experimental Procedures"). A control reporter plasmid, pGL3-RE-JAG1-mut, was generated by altering potentially critical nucleotides of the RE-JAG1 sequence. H1299 cells were transiently co-transfected with pGL3-RE-JAG1 or pGL3-RE-JAG1-mut together with a p53, p73␤, or p63␥-expressing plasmid. The ability to stimulate transcription was calculated as the luciferase activity in cells transfected with the pGL3-RE-JAG1 divided by the activity in cells transfected with the non-responsive reporter plasmid, pGL3-RE-JAG1-mut. Fig. 5C shows that the increase in luciferase activity for pGL3-RE-JAG1 was higher for p63␥ than for either p53 or p73␤. These results are consistent with the strong induction of endogenous JAG1 by p63␥ (Fig. 3). As a control, we demonstrate that the level of transcription from a FIG . 5. Regulation of JAG1 expression by p63. A, schematic representation of part of the JAG1 gene. A potential responsive site, RE-JAG1, is located at the second intron of the JAG1 gene and consists of four copies of the consensus 10-bp motif of the p53-binding sequence. Lowercase letters identify disparities with the consensus. R represents purine; Y, pyrimidine; and W, adenine or thymidine. B, p63 interacts with the RE-JAG1 sequence in vivo. ChIP assay of a genomic fragment (nucleotide position ϩ5499 to ϩ5749, where ϩ1 represents the translation initiation site) containing the RE-JAG1 sequence in Ad-p63␥-infected (lanes [3][4][5] or Ad-p53-infected (lanes 6 -8) Saos2 cells is shown (middle panel). Immunoprecipitation was performed using the gene-specific antibody against p63 (lane 3) or p53 (lane 6), followed by PCR amplification. Input chromatin represents a portion of the sonicated chromatin prior to immunoprecipitation (lanes 1 and 2). Immunoprecipitates with an anti-FLAG antibody (lanes 4 and 7) or in the absence of antibody (no antibody, lanes 5 and 8) were used for controls. DW indicates a no template control (lane 9). PCR amplification of the p21 promoter was performed using primers that flank the p53-binding site in the p21 promoter. PCR amplification revealed that a similar amount of p21 promoter sequence is present in p53-and p63-complexes extracted from each immunoprecipitate (upper panel). The RE-JAG1 was amplified in the immunoprecipitated samples with an antibody against p63 (middle panel, lane 3). One of the other candidate sequences, ϩ5983, is amplified in the input control, but not in the immunoprecipitated samples with an antibody against p53 or p63 (lower panel). C, the RE-JAG1 in the JAG1 gene is responsive to p63␥. H1299 cells were co-transfected with pGL3-RE-JAG1 or pGL3-RE-JAG1-mut, together with the expression vector pcDNA3.1 containing the p53, p73␤, or p63␥ gene. The relative luciferase activity was defined as the ratio of luciferase activity in the cells transfected with the pGL3-RE-JAG1 relative to the activity in cells transfected with the non-responsive reporter plasmid, pGL3-RE-JAG1-mut. As a control experiment, pGL3-p53CBS containing three copies of the consensus p53binding sequence (p53CBS ϫ 3) or its mutant form was transfected into H1299 cells with the expression vector containing p53, p73␤, or p63␥. All experiments were performed in triplicate and the means and standard deviation are indicated by the bars and brackets, respectively.
reporter, pGL3-p53CBS, containing three copies of the p53 consensus binding sequence was higher for p53 than for p63 or p73 (Fig. 5C). Taken together, our results suggest that there are likely to be differences among the p53 family members with regard to their optimal DNA-binding sequences and that RE-JAG1 is a functional p63-responsive element.
p63 Leads to the Activation of the Notch Signal Pathway-The extracellular domain of Notch is thought to interact with a ligand expressed on neighboring cells. Recent in vitro studies have shown that exogenously applied Notch ligands can activate endogenous Notch signal pathway (29 -31). Thus, we tested whether the JAG1 induction mediated by p63 or p73 could promote a functional interaction with cells expressing endogenous Notch1 receptor. The gene encoding the mammalian transcription factor HES-1 is a direct downstream target of the Notch signal pathway (32). As Notch1 has been reported to be highly expressed in Jurkat cells (25,33,34), we examined whether HES-1 mRNA was up-regulated in Jurkat cells cocultured with Ad-p63␥-infected Saos2 cells. HES-1 mRNA was not detectable in Ad-p63␥-infected Saos2 cells (Fig. 6, lane 1). In Jurkat cells, HES-1 mRNA was up-regulated 3-fold by coculturing with Ad-p63␥-infected Saos2 cells compared with co-culturing with control Ad-lacZ-infected cells (Fig. 6, lane 6). In contrast, expression of p53 or p73␤ in Saos2 cells exhibited only slight HES-1 induction in co-cultured Jurkat cells (Fig. 6,  lanes 4 and 5). The simplest interpretation of these data is that p63␥-induced JAG1 expression triggered Notch signaling in neighboring Jurkat cells. DISCUSSION Two members of the mammalian p53 family, p73 and p63, encode proteins that share considerable structural homology with p53, suggesting that the p53 family members have a potential for functional overlap. Indeed, the expression of several p53-regulated genes can also be induced by p73 and p63, though recent studies show considerable differences of inducibility among p53 and its family members (8,10,12). In this study, Northern blot analysis revealed that JAG1 and JAG2 expression are preferentially induced by p63␥ and p73␤, respectively. Induction of JAG1 and JAG2 was seen following ectopic expression of p63 or p73 in various types of human cancer cell lines regardless of p53 status. Additionally, JAG1 induction was considerably high and early relative to that of p21. We identified a specific binding site for p63 in the second intron of the JAG1 gene by a ChIP assay. A reporter assay demonstrated that the p63-binding site, RE-JAG1, was important for the p63-dependent transactivation. RE-JAG1 consists of four copies of the 10-bp p53-binding motif. Characterization of the responsive sequence for p73 within the promoter of AQP3 (35) and the p73 gene itself (36) was described in two recent studies. Both responsive sequences consist of three copies of the 10-bp p53-binding motif. We also showed that ectopic expression of p73␤ strongly induced JAG2 expression in all cell lines tested. It remains to be elucidated whether the JAG2 expression is controlled directly by p73. Altogether, our findings clearly indicate that JAG1 is a direct target for p63 and possibly for p73.
The Notch family of transmembrane receptors has been reported to play an important role in development by specifying cell fates (37,38). To date, five Notch ligands have been identified, including JAG1, JAG2, Delta like-1 (DLL1), DLL3, and DLL4, all of which are transmembrane proteins having an extracellular domain important for receptor binding (39 -43).
Here we demonstrated physical but also functional involvement of p63 in the Notch/Jagged signal pathway. The induction of JAG1 correlated with the initiation of signaling downstream of the Notch receptor in co-cultivated cells. Endogenous HES-1 expression was up-regulated in Notch1-expressing Jurkat cells by co-culturing with Ad-p63␥-infected Saos2 cells. This agrees with a recent report showing that Rel/NFB-mediated JAG1 expression can transactivate endogenous HES-1 gene expression in co-cultured Jurkat cells (25). Our findings indicate that p63 can trigger the Notch signal pathway in the neighboring cells and therefore raise the possibility that the members of the p53 family play a role in normal development through modulating Notch signal pathway.
Ligand-mediated activation of Notch induces the proteolytic release of the intracellular domain of Notch and transactivation of its target genes, which leads to modulation of cell proliferation and differentiation (44,45). Mutations in human Notch ligands result in the disruption of the Notch signal pathway, leading to developmental abnormalities (46 -48). Mutations in the JAG1 gene have been found in patients with Alagille syndrome, an autosomal dominant disorder characterized by abnormal development of heart, skeleton, liver, and eye, as well as a characteristic facial appearance. Our data are generally consistent with the similarity between the developmental defects associated with alteration in the p53 family members and those of Notch ligands (17-19, 21, 46 -49). In particular, JAG2-deficient mice exhibit defects of limb and craniofacial development (49), closely resembling the phenotype of patients with ectrodactyly, ectodermal dysplasia, and facial clefts syndrome, in which p63 is mutated. Both p63 and p73 can transactivate the promoters of genes associated with neuronal or epidermal differentiation, and overexpression of these genes up-regulates neuronal or epidermal differentiation markers (50,51). Together with our findings, these studies highlight the potential for an interplay between the p53 family genes and Notch signal pathway during ectodermal development.
In conclusion, we report that Notch ligands, JAG1 and JAG2, are selectively induced by p73 and p63, respectively, but not by p53. We also identified a specific binding site for the p63 protein in the second intron of the JAG1 gene and showed that p63 can activate Notch signaling to the neighboring cells. Our findings point to a potential role for p63 and p73 in normal development and cellular regulation mediated by Notch signaling.  (lanes 3-6). Total RNA (10 g) were loaded in each lane. Endogenous HES-1 expression in Jurkat cells was increased 3-fold by co-culturing with Ad-p63␥-infected Saos2 cells relative to co-culture with control Ad-lacZ-infected cells (lane 6).