p53 Amino Acids 339 –346 Represent the Minimal p53 Repression Domain*

,

The tumor suppressor p53 is a nuclear phosphoprotein and transcriptional factor that plays a crucial role in the regulation of cell growth, DNA repair, and apoptosis (1,2). The wild-type p53 protein can be divided into four domains: an acidic transcriptional transactivation domain at the NH 2 terminus (3-6); a sequence-specific DNA-binding domain at the hydrophobic center portion of the protein (7); a multifunctional COOHterminal domain; and a region near the COOH terminus containing a tetramerization domain that may be required for the stabilization of DNA binding activity (8,9).
Mutations in the p53 gene have been found in approximately 50% of human cancers (10,11). Most tumor-derived p53 mutants harbor single amino acid substitutions (missense mutations) within the central region, which either prevent normal DNA contacting or alter the conformation of this domain (5). On the other hand, some mutations in the NH 2 terminus (codons 1-99) and in the COOH terminus (codons 291-393) are non-missense mutations, such as nonsense point mutations, small insertions or deletions, and splice mutations (12)(13)(14)(15)(16). p53 is capable of both transactivating through binding to specific DNA-binding sequences (17,18) and repressing the transcription of many cellular and viral promoters that do not contain binding sequences. The promoters repressed are mainly those of which initiation is dependent on the presence of a TATA box (19 -23). Depending on the physiological circumstances, p53 can mediate either cell growth or apoptosis.
Three-dimensional nuclear magnetic resonance (7,24) and x-ray crystallography were used to examine the COOH terminus of p53, which contains the tetramerization domain (25). The tetramerization region was found to contain a ␤-sheetturn-␣-helix motif (320 -360). The COOH-terminal segment of p53 mediates repression when bound to DNA as a GAL4-p53 fusion protein (26). A COOH-terminal deletion of up to amino acid 327 (del 327-393) eliminates the repression of cytomegalovirus-CAT 1 (17). However, very little is known about the transcriptional repression ability of p53 deletion mutants. This repression activity may potentially affect the biological function of p53 and is important for further analysis. In the present study, we further clarified the importance of the COOH-terminal domain of human p53 in relation to its ability to repress transcriptional activity.
To illuminate the mechanism of p53-mediated transcriptional repression, we used a chimerical system to map the minimal repression domain of the 67 amino acids in the COOH terminus. In this report, we describe the minimal domain containing eight amino acids (hereafter referred to as P53MRD) that is the essential repression element. This minimal domain exactly locates on the E6-binding region, which may possibly affect transcriptional activity. Mutants of the eight amino acids exhibited loss of their repression activity. Mutants (amino acids 340 and 341) of native p53 also lost their transcriptional repression of TKCAT. We demonstrate that P53MRD is indeed the element required for the repression of p53.

EXPERIMENTAL PROCEDURES
Plasmid Construction-To map the minimal repression domain in p53, several vectors and their derivatives were constructed. Plasmids pTKCAT, pG5E1bCAT, pSGVP, pSG424, pCEP4p53, pLEC, L6EG5C, pSG424p53(1-318), and pLexA-VP16 have been previously described (27). For plasmid pSG424-p53(327-393), the polymerase chain reactionamplified fragment from COOH-terminal 327-393 of wild-type p53 containing KpnI and XbaI sites was cloned into plasmid pSG424/KpnI/ XbaI. For plasmid pSG424-p53(1-125)-p53(327-393), the polymerase chain reaction-amplified fragment from NH 2 -terminal 1-125 of wildtype p53 containing KpnI and EcoRI sites was cloned into plasmid * This work was supported by grants from Academia Sinica. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
¶  The detailed procedures for these clones were described in a previous report (28). Another mutant clone pCEP4p53(M340P,F341D) was obtained by the same method.
Cell Culture, Transfection, and CAT Assay-p53-null human tumor cell lines Saos-2 and H1299 were maintained in Dulbecco's modified Eagle's medium supplemented with 10% fetal calf serum. Cells were seeded 12 h before transfection at 1.2 ϫ 10 6 cells/10-cm dish. Cells were transfected by the calcium phosphate method. 5 g of the CAT reporter and activator plasmid was used, and 5 g of a LacZ reporter plasmid pCH110 (Amersham Pharmacia Biotech) was included to monitor the transfection efficiency. Typically, the transfection lasted 12 h. CAT activity was measured 48 h after transfection and quantitated according to Carey et al. (29).
Protein Analysis-For Western blotting, equal amounts of lysates were boiled in a sample buffer (125 mM Tris-HCL, pH 6.8, 100 mM dithiothreitol, 2% SDS, 20% glycerol, 0.005% bromphenol blue) for 5 min and then subjected to polyacrylamide gel electrophoresis. After being transferred to an Immobilon membrane (Millipore Asia Ltd.), p53 and the derivatives were detected with antibodies (OD-1) against p53 using an ECL system (Amersham Pharmacia Biotech). Fig.  1A shows that the deletion mutants (lanes 3-7) compared with the positive control (lane 2) had low relative CAT activity (RCA), indicating an increased activity of transcriptional inhibition, but the RCA of lane 8 (region 344 -360) was recovered to 61%. A comparison of lanes 5, 7, and 8 clearly indicates that the COOH-terminal region (339 -346) containing eight amino acids was essential for the activity of transcriptional inhibition. There was no apparent difference in the protein expression level of p53 among all constructed clones (Fig. 1B). In addition, a comparison of lanes 2, 7, and 8 in Fig. 1B also indicates that the loss of activity of transcriptional inhibition was not associated with the expression level of p53. To validate this eightamino acid minimal repression domain (P53MRD), we constructed three point mutants as illustrated in Fig. 2A, lanes 4 -6. The corresponding constructs are listed below the CAT assay. Lane 2 is the positive control in which no mutation was created. Lanes 3-6 showed a decreasing trend of repression activity. This result pointed out the fact that this specific P53MRD was very important for the activity of transcriptional inhibition and that not any one single mutation, either a M340P or F341D mutant, could lead completely to its loss of transcriptional repression function. At least 2-amino acid mutations at this region were required to alter the activity of repression. Fig. 2B shows there was no difference in the protein expression level of p53 among all constructed clones.

Transactivation of Reporter Constructs pG5E1BCAT by Derivatives of the GAL4-p53 Fusion Protein in Saos-2 Cells-
Schematic Representation of the 67 Amino Acids of the COOH Terminus of GAL4-p53 Derivatives Required for Transcriptional Repression- Fig. 3 shows a summary of the relationship between the activities of transcriptional repression and the regions on the COOH terminus of p53. A comparison of the second and third clones shows that the loss in repression activity was increased from 39 -98%, demonstrating that the six-amino acid (region 339 -344) deletion resulted in an additional 60% loss of repression activity. A comparison of the fifth and eighth clones was surprising in that the activity was down to 2%, i.e. more than 95% of repression activity was lost. Of any single mutation (sixth or seventh clone) compared with the fifth clone, only about 10% of repression activity was lost.
Transactivation of the Reporter pG5E1BCAT by GAL4-VP16 Derivatives-To understand whether this P53MRD was specific to p53, we used VP16 that has an acidic activation domain (1,30,31) like p53 to examine the activity of transcriptional repression. Fig. 4A shows that both the wild type (lane 3) and mutants (lanes 4 -6) of this eight-amino acid element did not exhibit any repression activity. This indicated that the repression activity of the P53MRD in transcription was specific for p53. As shown in Fig. 4B, the difference in the protein expression levels of VP16 among all constructed clones was normalized to eliminate the effect of the loading amount. The RCA value was recalculated as 1, 100, 153, 184, 125, and 118, respectively. The normalized RCA value reveals no significant effect in the repression activity.
Effect of the Eight-Amino Acid Minimal Repression Domain on the DNA Binding Activity of the Gal4 Chimera-To further verify that the repressive function of the identified minimal domain was directly on the transcriptional activation domain and was not affected by the influence of the binding action of GAL4(1-147), a molecular interference assay was employed (27,32). The construct L 6 EG 5 C (Fig. 5) contains five repeats of the GAL4-binding element (5 ϫ GAL4) placed behind the EIbTATA box (2). The GAL4 derivative protein binds to this element and blocks the transcriptional initiation complex. Therefore, the CAT activity is reduced. pSG424 (previously referred to as GAL4(1-147) and its derivatives pSG424p53(1-318), pSG424p53(1-125), and pSG424p53(1-125)m(339 -   1, 100, 153, 184, 125, and 118, respectively. The RCA values normalized by the expressed protein levels still showed no significant effect on the repression activity. 346)(M340P,F341D) blocked the CAT activity with different repressive levels (Fig. 5, upper left). A comparison of lanes 3-6 with lane 2 shows that there was a 5-15-fold difference in transcriptional repression using various RCA ratios. In contrast, the construct L 6 EC containing no GAL4-binding element did not influence transcriptional activity when co-transfected with the activator and blockers (Fig. 5, upper right). A comparison of lanes 9 -12 with lane 8 showed that the RCA of these four constructs had little or almost no difference.
Effect of Wild-type p53 and p53(M340P,F341D) Double Mutant on TKCAT Promoter Activity-To understand whether this region in native p53 also regulates the activity of transcriptional repression, we assayed its repression activity using the reporter pTKCAT (33). The constructs used in the assay of TKCAT promoter activity are listed in Fig. 6A. The control of protein levels is illustrated in Fig. 6B. The double mutation, which is the same as described in Fig. 2A, was performed on wild-type p53. Fig. 6A (lane 1) shows that there was no p53 and thus, no repression activity. In lane 2, wild-type p53 existed, and repression was present. In lane 3, the double mutation of the eight-amino acid region of p53 decreased the repression activity by 8-fold.
Assay of the Ability of Different Mutants Mediation of Transcriptional Repression of pTKCAT in H1299 Cells-As shown in Fig. 7, the mutants on the hot spot (V143A, V173L, N247I, R248W, and R273H) of the DNA-binding domain retained repression activity (compare the lanes with decreasing RCA with lane 1). The relative ratio of repression activity was greater than 5-fold among all mutants compared with the control pCEP4 and showed a significant difference. A comparison among lanes 8 and 9 and lanes 2 and 10 on the transcriptional activity domain also retained significant repression activity, indicating that the NH 2 -terminal transcriptional domain was not necessary for the repression of TKCAT in p53. DISCUSSION A previous report described the region of COOH-terminal amino acids (326 -393) on p53 required for repression activity (2). Subler et al. (19) also showed that the COOH-terminal 67 amino acids were crucial for transcriptional inhibition by deleting this region for transcriptional activity. However, the exact location and mechanism of p53-mediated repression remained unclear.
In this study, we investigated the activity of transcriptional inhibition by minimizing the transcriptional domain. A chimeric assay method was employed to detect the transcriptional activity and was performed by fusing arbitrary fragments of the p53 COOH-terminal to an activator Gal4-p53N(1-125). We have found that the COOH-terminal region (339 -346) containing eight amino acids was essential for the activity of transcriptional inhibition, and the loss of activity of transcriptional inhibition was not associated with the expression level of p53. Among these eight amino acids, it was reasonable to suspect that the loss of six amino acids (region 339 -344) would create an unstable binding site for certain factors. Our findings clearly demonstrated that the substitution of at least two amino acids at this region was required to alter the activity of repression. Amino acids 340 and 341, which were hydrophobic, were the most crucial in determining the activity of inhibition. Further studies are needed to clarify the exact mechanism of how these two amino acid substitutions affect p53 transcriptional activity.
Overall, our results demonstrated that only eight amino acids are required to determine the transcriptional activity of FIG. 5. Effect of the eight-amino acid minimal repression domain on the DNA binding activity of the Gal4 chimera. 1 g of the reporter DNA, 1 g of the activator DNA, and 13 g of the specified blocker DNA were used for calcium phosphate-mediated DNA transfection. A Gal4 derivative GAL4(1-147), which contains residues 1-147 of the yeast GAL4 and possesses DNA binding activity, and pSG424p53(1-318) were used as positive controls. Cells were harvested 48 h after transfection. The absence (Ϫ) or presence (ϩ) of the activator is indicated above each track of the autoradiogram as are the blockers and RCA. The reporter was pL 6 EG 5 C (lanes 1-6) or pL 6 EC (lanes 7-12). A list of the structure of the activator, reporters, and blockers are shown below the autoradiogram. p53, and the minimal repression domain only influenced the activation domain. In addition, the chimera protein still maintained its activity for DNA binding as illustrated by the unchanged binding activity of GAL4. The use of powerful chimera protein assays also validated that amino acids 340 and 341 play a key role in transcriptional inhibition.
To clarify whether P53MRD in native p53 could regulate the activity of transcriptional repression, we assayed its repression activity using the reporter pTKCAT. We chose the thymidine kinase promoter as a reporter because this promoter, unlike the E1bTATA promoter, possesses a detectable basal transcriptional activity that is necessary for the assay of repression. Our results confirmed that the region in native p53 still functioned in regulating transcriptional repression.
The eight-amino acid region identified is indeed the minimal and specific domain for the transcriptional repression of p53. The transcriptional activity domain in both VP16 and p53 has the same effect on transcriptional regulation, but only the P53MRD could repress the transcription of p53 and not affect that of VP16. This fact indicated that the transcriptional inhibition or transcriptional repression was p53-dependent and p53-specific.
Pellegata et al. (34) showed that mutant p53⌬363 was the only truncated p53 protein that retained the ability to repress transcription from the TATA element. The monomeric forms of p53 (p53⌬333 and p53⌬303) did not display any repression of transcription. This finding was consistent with our results that the eight amino acids of the P53MRD were essential for repression. On the other hand, minimal tetramerization was located on amino acids 325-355 (24,35). Thus, the P53MRD containing amino acids 339 -346 (part of ␣-helix structure) do not constitute the tetramerization. We suspect that repression is not associated with tetramerization but is similar to the binding of E6 on p53, which does not require oligomerization (36).
There are two possible mechanisms for the P53MRD repression. First, the P53MRD may directly hinder the transcriptional activity domain of the NH 2 terminus and interfere with the transcriptional complex to achieve repression. A previous report (26) demonstrated this mechanism whereby the COOHterminal repression domain (amino acids 327-393) achieved repression activity by mediating the binding of TATA-binding protein. Second, the transcriptional repression may be accomplished by a putative factor, such as the effect of E6 on p53. The E6 protein of human papillomaviruses could target p53 in two ways. The first, which may be a more immediate response, is the abrogation of p53 transcription activity by binding to the cofactor CREB-binding protein/p300 (37,38). The second is the removal of the p53 protein through E6-associated protein-de-FIG. 6. Effect of wild-type p53 and p53(M340P,F341D) double mutant on TKCAT promoter activity. A, pTKCAT (5 g) was cotransfected into H1299 cells with 5 g of either the pCEP4 expression vector, the wild-type p53 expression plasmid, or the p53 double mutant and the carrier DNA for a total of 20 g of DNA/transfection. A list of the structure of the activators are shown below the autoradiogram. Cells were harvested and CAT was assayed after 48 h. The figure represents an autoradiogram of the CAT assay. B, approximately 50 g of protein from transfected H1299 cells were fractionated on a 12.5% SDS-polyacrylamide electrophoresis gel. Wild-type p53 and mutant p53(p53M340P,F341D) were detected with an anti-p53(DO-1) antibody. pendent degradation (39). In the present study, we found that the P53MRD was located on the E6-binding region of p53. Therefore, the P53MRD might mediate certain cellular factors like the E6 protein to achieve repression.
On the other hand, recent work by several groups has established an evolutionary conserved role for histone deacetylases in the mechanism of repression by transcription factors, such as Mad, Max, and the nuclear hormone receptors (40 -42). Transcriptional repression by wild-type p53 also employs histone deacetylases mediated by interaction with mSin3a, and the domain necessary for the p53-mSin3a interaction to two regions of the p53 protein was mapped from 40 -160 and 320 -360 (43). We suspect that the P53MRD (amino acids 339 -346) might mediate the same pathway to carry out p53-specific repression.
A comparison among p53/p53C⌬30 and mutants (p53Q22S23 and p53N⌬100C⌬30) on the transcriptional activity domain retained significant repression activity, indicating that the NH 2 -terminal transcriptional domain was not necessary for the repression of TKCAT in p53. This observation ruled out any squelching effect (44). Only the double mutant on hydrophobic amino acids 340 and 341 lost its repression activity. This result was the same as that obtained using the GAL4-fusion chimeric protein system as described above (mutated on amino acids 340 and 341). We concluded that the eight-amino acid region identified has a crucial role in transcriptional repression.
The fact that the COOH-terminal domain of p53 has transcriptional repression function has been known for some time. p53 might activate the transcription of death genes or repress the transcription of survival genes to promote apoptosis. Although several studies have indicated a role for transcriptional repression in p53-dependent apoptosis (2), the molecular basis of this activity remains poorly understood. This P53MRD may serve to regulate repression and even perhaps regulate apoptosis. More studies are needed to further investigate these possibilities and are currently in progress in our laboratory.