COOH-Terminal Domain of p53 Modulates p53-mediated Transcriptional Transactivation, Cell Growth, and Apoptosis¹ Free

Two major functions of p53 are the negative regulation of cell growth and the induction of apoptosis. Cells lacking functional p53 fail to arrest in G₁ following γ-irradiation (1), whereas introduction of wild-type p53 leads to G₁ arrest in either p53-null or p53-mutated cells (2, 3, 4). Growth arrest is controlled principally by transcriptional modulation of p53-transactivated genes, such as p21^Waf1/Cip1 (an inhibitor of cyclin-dependent-kinase) and GADD45 (a growth arrest and DNA damage responsive gene; Refs. 5, 6, 7). Apoptosis can be triggered by overexpression of transfected wild-type p53 in p53-null tumor cells (8). Thymocytes derived from p53 knock-out (p53 −/−) mice are resistant to radiation- and drug-induced cell death (9). Some tumor-derived p53 mutants display defects in apoptotic activity and still retain growth suppression (10). These data indicate that the distinct functions of p53 are mediated via specific pathways.

p53 has multiple, unique, functional domains. These include the NH₂-terminal transactivation domain, the sequence-specific DNA-binding domain that is necessary for the transcription transactivation, and the CTD.³ The CTD of p53 has drawn considerable attention in recent years based on its multiple functions (11, 12, 13, 14, 15, 16). The COOH terminus of p53, comprising amino acids 311–393, can be subdivided into at least two structural determinants: the oligomerization domain (amino acids 319–360) and an adjacent basic region (amino acids 363–393), which also has been referred to as an apoptotic domain (17); a transcriptional regulatory domain (18); or a DNA damage recognition domain (14, 19). The oligomerization domain is required for the p53 tertiary structure, which regulates transcriptional transactivity by altering the conformation of the protein (20, 21). Modifications of the negative transcriptional regulatory domain of p53 CTD by phosphorylation, acetylation, deletion, or antibody binding have been shown to play important roles in the regulation of p53 sequence-specific binding (12, 16, 22, 23, 24, 25). Wild-type p53 can transactivate genes transcriptionally through binding to specific DNA sequences in the promoter regions of genes involved in cell cycle control and apoptosis, e.g., p21^Waf1/Cip1, Bax, and IGF-BP3 (5, 26, 27). The COOH terminus seems to exert a negative effect on the DNA-binding activity of the core p53 domain (22, 28). Deletion of the last 30 amino acids was shown to abolish this repressor activity (22, 28). Furthermore, it was found that an alternative COOH-terminal spliced p53 exhibited an enhanced DNA-binding activity (29).

p53 mediates apoptosis by both transcription-dependent and transcription-independent pathways (7, 17, 26, 30, 31, 32). The contribution of sequence-specific transcriptional activation to the induction of apoptosis is less clear than the induction of cell cycle arrest. The CTD is involved in the apoptotic function of p53 by protein-protein interactions (17). The CTD of p53 can bind directly to XPB and XPD, inducing apoptosis (17, 33). The hepatitis B virus X protein binds to CTD and inhibits p53-mediated apoptosis (34). The CTD of p53 can bind single strands of DNA and DNA ends that are likely to be present in irradiated cells, implicating that CTD is responsible for sensing these types of DNA damage (17, 35).

We have reported previously that overexpression of wild-type p53 in primary human fibroblasts induced apoptosis and that deletion of the last 40 amino acids at the CTD abolished p53-mediated apoptosis completely (17). These data indicate that the CTD may contribute directly to apoptosis. Analysis of the p53 mutation spectrum (36, 37) indicates that ∼5–15% of the p53 mutations occur in the CTD during human tumorigenesis. Most of the somatic mutations in the p53 CTD are either nonsense mutations or lead to frameshifts, resulting in a truncated protein (Fig. 1 A). To further examine the biological functions of the p53 CTD, we investigated tumor-derived p53 CTD mutants that were presumably selected during the clonal evolution of human cancers. The results indicate that the CTD of p53 regulates transcriptional activity and sequence-specific DNA binding and modulates the apoptotic process. The CTD of p53 also is involved in G₁ arrest and cell growth inhibition, is closely correlated with transcriptional activation of p21^Waf1/Cip1, and can be independent of p53-mediated apoptosis.

Five tumor-derived p53 CTD mutants: 352D/H, 356G/W, 342-stop (342R/Stop), 360-del (360 1-bp deletion), and 387-del (387 19-bp deletion; Refs. 38, 39, 40, 41, 42) were constructed by site-directed mutagenesis. The mutations were verified by DNA sequencing. Wild-type p53 and the CTD mutants were inserted into the pcDNA3 vector under the control of the cytomegalovirus early promoter.

Primary human fibroblasts and an immortalized p53-null LFS fibroblast cell line (041) derived from the skin biopsies of a patient with LFS were grown in Ham’s F10 medium supplemented with 10% fetal bovine serum. Cells were seeded onto coverslips and incubated for 1–2 days prior to microinjection. pcDNA3 vectors containing wild-type p53, the CTD mutants, or LacZ as a negative control at a concentration of 200 ng/μl, were microinjected into the nuclei of cells, using a microcapillary glass needle. For each experiment, at least 100 cells were injected, and a typical experiment yields at least 50 positive cells for analysis. After incubation for 24 h, cells were fixed with 4% paraformaldehyde followed by methanol treatment. p53 was visualized by staining the cells with the anti-p53 polyclonal CM-1 antibody (Signet Labs, Dedham, MA), and followed by fluorescein-conjugated antirabbit IgG (Vector Labs, Burlingame, CA) as described previously (17). Nuclei were stained with 4′, 6-diamidino-2-phenylindole. Apoptotic cells that were p53-positive were scored by their smaller size, condensed chromatin, and fragmented nuclei, as compared with control cells.

LFS fibroblasts cultured in 12-well plates at 60% confluence were transfected transiently with 1 μg of pcDNA3 p53 expression vectors, using Lipofectin (Life Technologies, Gaithersburg, MD). The p53 transcriptional transactivities were examined with the Dual-Luciferase Reporter system (Promega, Madison, WI) by cotransfecting 0.025 μg of Renilla Luc vector SV 40 (as an internal control) and 0.5 μg of p53-responsive reporter vectors, WWP-Luc-p21, PGL3-Luc-Bax, or pUHC13-3-Luc-IGF BP3-BOX B (26, 43, 44). Five h after transfection, the medium was changed. After an additional 15 h of incubation, the cells were lysed and the lysates were collected for the activity measurement. The activity of luciferase was quantified with a Luminometer (Analytical Luminescence Lab., Ann Arbor, MI).

To evaluate the expression level of wild-type and CTD-mutated p53 after transfection, the cells were lysed in RIPA buffer [150 mm NaCl, 1% NP40, 0.5% sodium deoxycholate, 0.1% SDS, 50 mm Tris-HCl (pH 8.0)]. Forty μg of total cellular protein were resolved by 10% SDS-PAGE, transferred to a nitrocellulose membrane, and probed with a purified, biotinylated mouse monoclonal antibody that recognizes the p53 protein (Oncogene Research Products, Cambridge, MA).

To generate p53 proteins for EMSA, 1 μg of each of the pcDNA3-p53 constructs was used for the in vitro transcription-translation assay, according to the procedure described by the Promega manufacturer for the TNT T7 Quick Coupled Transcription/Translation System (Promega). The DNA-binding activities of the wild-type and the CTD-mutated p53 proteins were analyzed by EMSA. The consensus sequence oligonucleotide, TCGAGAGGCATGTCTAGGCATGTCTC (44) was synthesized, prepared in double-stranded form, and end-radiolabeled with ³²P-ATP. Four μl of the translated protein products were mixed with 20 fmol of radiolabeled DNA, 1 μl of PAb-421 ascitic fluid (when indicated; +), 2 μg (2 μl) of poly dl-dc as nonspecific DNA, and one-half of the reaction volume of buffer (25 mm Tris-HCl, 100 mm KCl, 6.25 mm MgCl₂, 0.5 mm EDTA, 1 mm DTT, and 10% glycerol). The reactions were incubated for 15 min on ice and for an additional 15 min at room temperature. The samples were analyzed in 4% polyacrylamide gels electrophoresed at 350 V in 0.5 Tris-borate-EDTA buffer at 4°C.

Both p53-null LFS fibroblast cells (041) and Hep 3B cells were cultured in EMEM medium supplemented with 10% fetal bovine serum. Five μg of pcDNA3 containing wild-type p53 or the CTD mutants were transfected into the cells at 60% confluence in 60-mm dishes. Forty-eight h after transfection, G418 was added for the selection. The cells were maintained in the G418-containing medium for 10–14 days, and the colonies were counted. The experiments were repeated at least three times. The significant difference (P) was tested by the Student’s t test.

To test the biological significance of the CTD mutations selected during tumorigenesis, we investigated five tumor-derived p53 CTD mutants (Fig. 1 B), including a mutation hotspot, 342-stop. These nonsense, missense, deletion, or frameshift mutations occurred in either the oligomerization domain or the apoptotic domain of CTD.

Consistent with our previous data, the wild-type p53 induced apoptosis in ∼20% of the p53-positive cells in both primary human fibroblast cells and LFS fibroblast cells (p53 null), 24 h after microinjection (Figs. 2 and 3). The cells that were microinjected successfully with either wild-type p53 or the CTD mutants showed expression of p53 by antibody CM-1 staining (Fig. 2, A and B). p53 localized mainly in both the nucleus and cytoplasm. No significant alterations of localization between wild-type p53 and CTD mutants were observed. p53 immune-positive and apoptotic cells showed fragmented nuclei and condensed chromatin (Fig. 2,C). Interestingly, most of the p53 CTD mutants have a significant reduction in apoptotic activity that was similar to the background rate induced after microinjection of a LacZ expression vector (Fig. 3). The p53 CTD mutant 387-del displayed a slight reduction in apoptotic rate.

p53 is a DNA-binding-dependent transcriptional transactivator, and the p53 CTD is a regulator of sequence-specific DNA binding (12, 15, 22, 23, 26, 27). To test the transcriptional transactivation activity of the tumor-derived p53 CTD mutants, the expression vectors encoding wild-type p53 or the CTD mutants were transfected into p53-null LFS fibroblasts along with reporter genes containing promoter elements from genes transcriptionally regulated by p53. In the transient transfection assay, the protein expression levels of the CTD mutants were similar to that of wild-type p53 (Fig. 4). However, the effects on transcription of the reporter genes was variable. The mutant 342-stop showed an attenuated transcriptional activation of the p21^Waf1/Cip1 promoter, but the other tumor-derived p53 CTD mutants activated transcription similar to wild-type p53 (Fig. 5,A). All of the tumor-derived p53 CTD mutants demonstrated a modest attenuated activation of the Bax and IGF-BP3 promoters (Fig. 5, B and C). The mutant 342-stop showed no detectable transcriptional transactivity of these genes relative to the vector control (Fig. 5, B and C).

p53 sequence-specific DNA binding is required for its transcription transactivation function (26, 45). To evaluate the specific DNA-binding activity of the various COOH-terminally modified p53 mutants, the binding of p53 CTD mutant proteins to p53 consensus sites was determined by EMSA. In vitro translated p53 proteins are shown in Fig. 6,A. As expected, the in vitro translated wild-type p53 protein does not bind to DNA spontaneously; however, this activity is enhanced by the anti-p53 antibody PAb-421 (Fig. 6,B). The p53 CTD mutant 342-stop did not bind the p53 responsive element either in the absence or presence of PAb 421 (Fig. 6,B). In contrast to the 342-stop, the mutant 360-del has spontaneous DNA binding activity. Neither the 352D/H mutant nor 356G/W affected the DNA binding activity of those proteins (Fig. 6,B and Table 1).

The inhibition of cell growth and induction of cell cycle arrest by p53 depend on p53-mediated transcriptional activation (46). Transcriptional activation of p21^Waf1/Cip1 appears to contribute to cell growth inhibition (5). The 342-stop mutant, which showed no transcriptional activation from the p21^Waf1/Cip1 promoter (Fig. 5,A), lacked the growth inhibition function in both LFS fibroblast and Hep 3B cells, whereas the other mutants retained growth suppressor activity similar to wild-type p53 in both cell types (Fig. 7). The effects of the tumor-derived p53 CTD mutations on various biological functions measured are summarized in Table 1. All of the tumor-derived p53 CTD mutants were defective in the induction of apoptosis. The p53 CTD mutant 342-stop with no inhibition on the cell growth showed neither transcriptional activation on target genes tested nor sequence-specific DNA binding. Each of the p53 CTD mutants 352D/H, 356G/W, 360-del, and 387-del bound to the p53 consensus sequence and retained most of its transcriptional transactivity. The p53 CTD mutant 360-del activated the sequence-specific binding to the p21^Waf1/Cip1 (data not shown) and p53 consensus binding site PG13.

p53 responds to cellular stress and maintains genomic stability (1, 47, 48, 49, 50). However, this multifunctional property of p53 makes it difficult to determine the precise mechanism(s) by which p53 functions as a tumor suppressor. p53 may also induce apoptosis by transactivating proapoptotic genes such as Bax and IGF-BP3 (26, 27) and by transrepressing antiapoptotic genes such as bcl-2 (51). IGF-BP3 antagonizes the activity of insulin-like growth factor 1, and a reduction in the level of the insulin-like growth factor 1 type II receptor has been shown to result in apoptosis (27). In our study, the tumor-derived p53 CTD mutants have diminished apoptotic activity and reduced transcriptional transactivity of Bax and IGF-BP3, suggesting the interactive effects of p53 transcription-dependent and transcription-independent apoptotic pathways in normal and LFS fibroblasts. We explored the significance of the p53 CTD by examining several of the biological functions of tumor-derived p53 CTD mutants. A deficiency in apoptosis was a common characteristic of all of the tumor-derived p53 CTD mutants examined. This finding is consistent with the model that apoptosis is an important function for p53-mediated tumor suppression and that loss of this function by mutation at the p53 CTD favors tumorigenicity.

Previously, we reported that the COOH-terminal alternatively spliced p53 product altered the apoptotic function at different time points (52). We showed recently that microinjection of the p53 CTD polypeptide 319–393 into normal primary human fibroblasts, mammary epithelial cells, and p53-null LFS induced apoptosis (17). p53 CTD polypeptides can also induce apoptosis in human cancer cell lines (13). These data indicate that the p53 CTD can be sufficient to induce apoptosis. The CTD of p53 binds to both cellular and oncoviral proteins that can alter p53 functions (15, 53). For example, the p53 CTD can recognize damaged DNA and may influence DNA repair mechanisms through an interaction with replication protein A and TFIIH (16, 54). TFIIH is a multiprotein complex involved in transcription (55), nucleotide excision repair (33, 56), and apoptosis (17). p53 binds to four proteins, XPD, XPB, p36, and p62, in the TFIIH complex (33, 34, 57, 58). The COOH terminus of p53 binds specifically to the DNA helicases XPB and XPD in TFIIH, which are essential for transcription and nucleotide excision repair and contribute to apoptosis (17). p53-mediated apoptosis does not require RNA and protein synthesis (59, 60), and thus supports the hypothesis that p53-dependent apoptosis can occur through a transcription-independent pathway (31, 61, 62, 63). Our data are consistent with this hypothesis, indicating that the p53 CTD contains an apoptosis domain and that its induction of apoptosis may be mediated through its binding to TFIIH and other cellular protein mechanisms.

One of the important functions of p53 is its ability to act as a transcriptional transactivator. The CTD mutant 342-stop abolished both suppression of the cell growth and transactivation of p21^Waf1/Cip1, Bax, and IGF-BP3 in p53-null LFS fibroblasts. Failure to transactivate was associated with reduced ability of the mutant to bind the p53-responsive DNA sequences present in the target genes. The other tumor-derived p53 CTD mutants activated p21^Waf1/Cip1 transcription to the same extent as wild-type p53; however, the transcriptional transactivity of Bax and IGF-BP3 was reduced moderately. The heterogeneity of the transcriptional transactivities on p53-responsive genes has been reported previously (45, 46, 64). These reports and the data presented here indicate that there are individual pathways to regulate different p53 downstream target genes, which are necessary for the different biological functions of p53. The lack of transcriptional transactivity by the tumor-derived mutation 342-stop indicates that the last 50 amino acids are critical for the transcriptional transactivation.

The transcriptional transactivity of p53 is dependent on the specific recognition of DNA sequences located in the promoter regions of the target genes, and p53 requires activating factors to confer effective sequence-specific DNA binding activity (22, 44, 54, 65). The CTD of p53 has been shown to be important in regulating the activity of the whole protein by modulating its sequence-specific DNA binding and transcriptional transactivity. The modulating factors include mutation, phosphorylation, acetylation, or PAb-421 binding within CTD of p53 (12, 16, 18, 22). Our data are consistent with these findings. The 360-del mutant has seven altered amino acids and a stop codon at 369. This protein has lost the PAb-421 epitope, but kept intact the oligomerization domain, and behaves as the described p53Δ30 mutant (28) or the alternatively spliced p53-encoded protein found in murine cells (29, 66). The truncated protein-product of the 342-stop mutant lost its DNA-binding activity because of the loss of the oligomerization domain. The PAb-421 antibody also has no effect due to the lack of the specific epitope. Our results support the concept that transcriptional transactivity is dependent on sequence-specific binding and that the binding affinity is regulated and governed by the p53 CTD. All mutants, except for 342-stop, retain the cell arrest function, indicating that p53 CTD is involved in the inhibition of cell growth and that this function is not associated with its induction of apoptosis. Taken together, these results are consistent with the hypothesis that a diminution of the apoptotic function of p53 is important in its tumor suppression activity and that the CTD of p53 contributes to this activity.

We thank Drs. B. Vogelstein, W. El-Deiry, M. Reed, and N. Kley for providing the WWP-Luc-p21, PGL3-Luc-Bax, and pUHC13–3-Luc-IGF BP3-BOX B vectors. We thank A. Robles, S. Linke, J. Welsh, and M. Khan for comments and aid, and D. Dudek for editorial assistance.