The C-terminal Sterile α Motif and the Extreme C Terminus Regulate the Transcriptional Activity of the α Isoform of p73*

p73, a member of the p53 family, is expressed from two separate promoters, generating TA and ΔN variants. Each variant potentially encodes at least seven alternatively spliced isoforms (α–η). Interestingly, we and others have shown that the α isoform of p73 has a weaker transcriptional activity than the β isoform. Because the α isoform has an extended C terminus consisting of a sterile α motif (SAM) and an extreme C terminus, it appears that the C terminus is inhibitory. However, how the C terminus inhibits the transcriptional activity of p73 has not been determined. Here, we found that both the SAM and the extreme C terminus exert their inhibitory activity by preventing the accessibility of p300/CBP to the activation domain in p73. Specifically, we showed that the SAM and the extreme C terminus together or individually are capable of repressing the function of p73 activation domain, but neither interacts directly with the activation domain, or suppresses the DNA-binding activity, of the p73 protein. We also showed that the intact state of the SAM and the extreme C terminus is essential for their inhibitory functions such that a small deletion of either the SAM or the extreme C terminus abolishes its inhibitory activity. Furthermore, we showed that both inhibitory domains in the C terminus are capable of suppressing the function of a cis heterologous activation domain from p53 or Gal4. Finally, we showed that both inhibitory domains suppress the ability of p73 to interact with the transcriptional coactivators p300/CBP that are necessary for the initiation of transcription.

p73, a member of the p53 family, is expressed from two separate promoters, generating TA and ⌬N variants. Each variant potentially encodes at least seven alternatively spliced isoforms (␣-). Interestingly, we and others have shown that the ␣ isoform of p73 has a weaker transcriptional activity than the ␤ isoform. Because the ␣ isoform has an extended C terminus consisting of a sterile ␣ motif (SAM) and an extreme C terminus, it appears that the C terminus is inhibitory. However, how the C terminus inhibits the transcriptional activity of p73 has not been determined. Here, we found that both the SAM and the extreme C terminus exert their inhibitory activity by preventing the accessibility of p300/ CBP to the activation domain in p73. Specifically, we showed that the SAM and the extreme C terminus together or individually are capable of repressing the function of p73 activation domain, but neither interacts directly with the activation domain, or suppresses the DNA-binding activity, of the p73 protein. We also showed that the intact state of the SAM and the extreme C terminus is essential for their inhibitory functions such that a small deletion of either the SAM or the extreme C terminus abolishes its inhibitory activity. Furthermore, we showed that both inhibitory domains in the C terminus are capable of suppressing the function of a cis heterologous activation domain from p53 or Gal4. Finally, we showed that both inhibitory domains suppress the ability of p73 to interact with the transcriptional coactivators p300/CBP that are necessary for the initiation of transcription.
Like p53, p73 functions as a transcription factor and activates a number of putative p53 target genes, which mediate the function of p73 in inducing cell cycle arrest, apoptosis, and differentiation (1, 9, 14 -17). In addition, previous studies have shown that p73 also activates its unique downstream targets, suggesting that p73 may play a role distinct from that by p53 in response to cellular stimuli (16,18). Indeed, both clinical and experimental studies do not support p73 as a classic tumor suppressor. For example, p73 mutation is extremely rare in human tumors (19 -21). Moreover, in contrast to p53-deficient mice that show an increased susceptibility to spontaneous tumors, p73-deficient mice exhibit severe neurological defects, including hydrocephalus, hippocampal dysgenesis, and abnormalities in the pheromone sensory pathway (11).
A number of proteins have been found to interact with p53 family members and regulate their transcriptional activities. Among these are ASPP1 and ASPP2 that bind to all three p53 family members and specifically affect their abilities to activate apoptotic genes, such as Bax, Puma, but not p21 or mdm2. Thus, ASPP1 and ASPP2 appear to regulate p53 family members in determining cell fate upon cellular stresses (29,30). Yes-associated protein, another p53 family interacting protein, binds to the PPXY motifs in p73 and p63 through its WW domain and increases the transcriptional activities of p73 and p63 (31). Taken together, these findings indicate that the activities of the p53 family proteins can be regulated by both common and unique factors.
Several lines of evidence suggest that the extreme C terminus in p53 functions as a negative regulator. For example, the DNA-binding ability of p53 is increased when the C terminus is deleted, phosphorylated, acetylated, or bound with a specific antibody to this domain (32)(33)(34). Both p73␣ and p63␣ have an extended C terminus compared with their shorter counter-parts, such as p73␤ and p63␤ (35). It has been well documented that p73␤ transactivates a variety of p53-responsive promoters to a greater extent than p73␣ does, suggesting that the C termini in p73␣ and p63␣ suppress their transcriptional activities in a manner similar to that in p53. But no homology has been found in the C terminus between p53 and p73 (1). In addition, p300/CBP is found to acetylate several lysine residues within the p53 C terminus, which enhances the DNA binding and thereby the transcriptional activity of p53 (33, 34, 36 -38). However, no lysine residue has been found to be acetylated within the p73 C terminus. Furthermore, RACK1, a receptor for activated C kinase, and ⌴⌴1, a c-Myc-binding protein, were found to interact with the C terminus in p73␣ and regulate p73␣ activity (39,40). Finally, the extreme C terminus in p63␣ is found to interact with the p63␣ activation domain and inhibits its activity (41). Collectively, the evidence suggests that the C terminus in p73␣ may inhibit its transcriptional activity through a unique mechanism.
In this study, we found that both the SAM and the extreme C terminus in p73␣ are inhibitory domains and their intact state is essential for their inhibitory functions. We demonstrated that both inhibitory domains are able to repress the function of p73 activation domain, and we provided evidence that this effect is not due to a direct interaction between the inhibitory domains and the p73 activation domain. We also showed that the C terminus does not suppress the DNA-binding ability of the p73 protein. Interestingly, both inhibitory domains are capable of repressing a cis heterologous activation domain from p53 or Gal4 transcription factor. Finally, we found that the C terminus suppresses the ability of p73 to interact with transcriptional coactivators p300/CBP.
Cell Lines-The culture, transfection, and generation of H1299 cell lines were performed as described previously (42). Individual clones were screened for inducible expression of various p73␣ proteins by Western blot analysis with monoclonal antibody against the HA epitope.
Western Blot Analysis-Cells were collected from plates in phosphate-buffered saline, re-suspended with 2ϫ SDS sample buffer, and boiled for 10 min. Western blot analysis was performed as described previously (42). Anti-p21 polyclonal antibody (c-19) was purchased from Santa Cruz Biotechnology (Santa Cruz, CA). Anti-HA monoclonal antibody and anti-actin polyclonal antibody were purchased from Sigma.
DNA Histogram Analysis-Cells were seeded at 2 ϫ 10 5 per 90-mmdiameter plate in the presence or absence of tetracycline (2 g/ml). At 72 h after plating, cells floating in the medium and cells on the plates were collected and fixed. DNA histogram analysis was performed as described previously (42). The percentage of cells in the sub-G 1 , G 0 and G 1 , S, and G 2 to M phases was determined by using the CellQuest program. The percentage of cells in the sub-G 1 phase was used as an index for the degree of apoptosis.
Chromatin Immunoprecipitation Assay-ChIP assay was performed essentially as described previously (44). 5 ϫ 10 7 H1299 cells, which were un-induced (Ϫp73) or induced (ϩp73) to express HA-p73␣, HA-p73␣(⌬555-636), HA-p73␣(⌬500 -555), or HA-p73␤, were crosslinked with 1% formaldehyde. Cell extracts were sonicated to generate 200-to 1000-bp DNA fragments. 1% of cell extracts from each of the samples was taken as input. Protein-DNA complexes were immunoprecipitated with anti-HA or anti-myc monoclonal antibody. After reverse cross-linking and phenol-chloroform extraction, DNA fragments bound by p73 were purified over a Qiagen column. PCR was performed to visualize the enriched DNA fragments. Primers that were used to amplify each of the two p53-responsive elements within the p21 promoter were as previously described (44). A region within the promoter of the GAPDH gene was amplified by the 5Ј-end primer (AAA AGC GGG GAG AAA GTA GG) and the 3Ј-end primer (AAG AAG ATG CGG CTG ACT GT) to serve as a control for nonspecific binding.

RESULTS
The C Terminus in p73␣ Suppresses the Ability of p73 to Activate Gene Expression and to Induce Apoptosis-Recently, we and others have shown that both p73␣ and p73␤ are capable of regulating a number of target genes and inducing apoptosis (1, 9, 14 -17), but p73␣ has a much weaker activity than p73␤. Because the ␣ isoform contains an extended C terminus consisting of a SAM (aa 487-554) and an extreme C terminus (aa 555-636) (Fig. 1A), it appears that the C terminus is inhibitory. However, the mechanism by which the C terminus in p73␣ inhibits the transcriptional activity exhibited by p73␤ has not been determined. To test this, we performed the luciferase reporter assay and showed that p73␤ was five times more potent than p73␣ to activate a luciferase reporter, which is under the control of the p21 promoter (Fig. 1B). To rule out the possibility that the difference in transcriptional activities be-tween p73␣ and p73␤ is promoter-specific, we examined the abilities of both isoforms to activate a luciferase reporter containing a p53-responsive element in intron 3 of the Gadd45 gene. Similarly, we found that p73␤ was at least five times more active than p73␣ to activate the Gadd45 luciferase reporter (Fig. 1C). These findings are consistent with several previous reports (4,45).
To determine whether the transcriptional activities of the p73 proteins correlate with their activities to induce apoptosis, we generated stable H1299 cell lines that inducibly express p73␣ or p73␤. We chose two cell lines that express a comparable level of p73␣ and p73␤, respectively, for further characterization (Fig. 1D). We performed DNA histogram analysis to measure the sub-G 1 fraction in cells that were un-induced or induced to express p73␣ or p73␤. Cells with sub-G 1 DNA content represent apoptotic cells. We found that p73␤ was nearly two times more active than p73␣ to induce apoptosis (32.3% versus 16.8%) (Fig. 1E), suggesting that the transcriptional activities of p73 proteins correlate well with their abilities to induce apoptosis.
Both the SAM and the Extreme C Terminus in p73␣ Are Inhibitory Domains-SAM is a typical protein-protein interaction domain and present in many regulatory proteins, including receptor tyrosine kinases, serine/threonine kinases, and transcription factors (46,47). However, no protein has been identified to specifically associate with the SAM in p73␣. In addition, no functional domain has been detected in the extreme C terminus. A report showed that lysine 627 in the extreme C terminus can be modified by SUMO-1, but this modification has no effect on p73␣ transcriptional activity (48).
FIG. 1. The C terminus in p73␣ suppresses p73 transcriptional activity and its ability to induce apoptosis. A, schematic structures of p73␣ and p73␤. B and C, p73␣ is less active than p73␤ in transactivation. 0.25 g of pGL2-p21A, which is under the control of the p21 promoter (B) or pGL2-Gadd45, which is under the control of a p53-responsive element in intron 3 of Gadd45 (C), was co-transfected into H1299 cells with 0.5 g of pcDNA3 control vector, or a vector that expresses p73␣ or p73␤. The -fold increase in relative luciferase activity by p73␣ and p73␤ was calculated using the empty pcDNA3 vector as a control. D, levels of the induced expression of p73␣ and p73␤ in H1299 stable cell lines are comparable. Western blot analysis was performed by using cell extracts from un-induced cells (Ϫ) and cells induced (ϩ) to express p73␣ or p73␤ for 24 h. p73␣ and p73␤ were detected with anti-HA monoclonal antibody. Actin was detected with anti-actin polyclonal antibody and was used as an equal loading control. E, p73␣ is less potent than p73␤ to induce apoptosis. H1299 cells were seeded at 2 ϫ 10 5 per 90-mm-diameter plate in the presence or absence of tetracycline (2 g/ml). At 72 h after plating, both floating cells in the medium and cells on the plates were collected, fixed with 1 ml of 100% ethanol overnight, and then centrifuged and re-suspended in 300 l of phosphate-buffered saline solution containing 50 g each of RNase A and propidium iodide per milliliter. The stained cells were measured by fluorescence-activated cell sorter analysis. The percentage of cells in the sub-G 1 fraction was determined by CellQuest program and used as an index for the degree of apoptosis.
Thus, we wanted to examine whether the SAM, the extreme C terminus, or both are responsible for the suppression. To this end, we generated p73␣(⌬555-636) that lacks the extreme C terminus, p73␣(⌬500 -555) that lacks the SAM, and p73␣(⌬495-636) that lacks both the SAM and the extreme C terminus ( Fig. 2A). We found that the ability of p73␣(⌬555-636) and p73␣(⌬500 -555) to activate the p21 promoter was comparable to that of p73␣ (Fig. 2B), suggesting that either the SAM or the extreme C terminus alone is sufficient for the inhibitory activity in the p73␣ C terminus. In contrast, p73␣(⌬495-636) became as potent as p73␤ to activate the p21 promoter (Fig. 2B). To rule out the possibility that the expression level of p73 proteins is responsible for such an outcome, we measured the level of these proteins for the luciferase assay and found that both wild-type and mutant p73 proteins were expressed at a comparable level (Fig. 2C). We also examined the ability of these proteins to regulate a p53-responsive element in the Gadd45 gene. Similarly, we found that p73␣(⌬495-636) was as potent as p73␤ in transactivation (Fig. 2D). Taken together, these data suggest that both the SAM and the ex- Luciferase assay was performed to determine the ability of p73␣, p73␤, and various p73␣ mutants to activate pGL2-p21A. C, levels of p73␣, p73␣(⌬555-636), p73␣(⌬500 -555), p73␣(⌬495-636), and p73␤ used for luciferase assays (B and D) are comparable. Various HA-tagged p73 proteins were detected by anti-HA antibody. D, both the SAM domain and the extreme C terminus repress p73 to activate pGL2-Gadd45. E, levels of induced expression of p73␣, p73␤, and various p73␣ mutants in H1299 stable cell lines are comparable. F, both the SAM domain and the extreme C terminus inhibit p73 to induce apoptosis. treme C terminus are inhibitory domains.
To determine whether the SAM or the extreme C terminus alone is responsible for the decreased activity of p73␣ to induce apoptosis, we generated H1299 cell lines that inducibly express p73␣(⌬555-636), p73␣(⌬500 -555), or p73␣(⌬495-636). The expression level of these proteins was measured and found to be comparable to that of p73␣ and p73␤ (Fig. 2E). Next, we measured the extent of apoptosis induced by p73␣, p73␤, or various p73␣ mutants. We found that p73␣(⌬555-636) and p73␣(⌬500 -555) were less active than p73␣ to induce apoptosis (Fig. 2F), suggesting that the SAM or the extreme C terminus alone is capable of attenuating the ability of the p73 protein to induce apoptosis. However, deletion of both the SAM and the extreme C terminus (⌬495-636) rendered the p73 protein highly active to induce apoptosis (Fig. 2F).
The Intact State of the SAM and the Extreme C Terminus in p73␣ Is Essential for Their Inhibitory Activity-The SAM in p73␣ is composed of four ␣-helices (helix 1 in aa 491-499; helix 2 in aa 506 -511; helix 4 in aa 525-531; and helix 5 in aa 538 -550) and a small 3 10 -helix (helix 3 in aa 517-520) (10,49). Previous studies showed that an intact SAM domain is required for the function of some SAM domain-containing proteins (50,51). However, the extreme C terminus (aa 555-636) in p73␣ has no homology to any known structures. Here, we showed above that the SAM and the extreme C terminus, alone or in combination, are capable of inhibiting the activity of p73. Thus, we wanted to examine the contribution of various helices of the SAM to the inhibitory activity by generating p73␣ mutants that lack a part of the SAM and the entire extreme C terminus. These mutants were as follows: p73␣(⌬540 -636), which lacks one helix of the SAM and the entire extreme C terminus; p73␣(⌬525-636), which lacks two helices of the SAM and the entire extreme C terminus; and p73␣(⌬510 -636), which lacks three helices of the SAM and the entire extreme C terminus. We then examined the ability of these mutants to activate the p21 promoter. As a control, we examined the activity of wild-type p73␣, p73␣(⌬555-636), which only lacks the extreme C terminus, and p73␣(⌬495-636), which lacks both the SAM and the extreme C terminus. We found that, when the integrity of the SAM was increasingly compromised, the transcriptional activity of the p73 protein was accordingly increased (Fig. 3A).
Next, we wanted to examine the contribution of various part of the extreme C terminus to the inhibitory activity by generating p73␣ mutants that lack a part of the extreme C terminus and the entire SAM. To do this, we generated two mutants: p73␣(⌬500 -570) and p73␣(⌬500 -590). Similarly, we found that, when the integrity of the extreme C terminus was increasingly compromised, the transcriptional activity of the p73 protein was accordingly increased (Fig. 3B).
To confirm the above data, we examined the ability of these mutant proteins to activate endogenous p21 by transient transfection (Fig. 3, C and D). We found that when a part of the SAM together with the entire extreme C terminus was deleted, the ability of the p73 protein to activate p21 was substantially increased (Fig. 3C). Similarly, when a part of the extreme C terminus together with the entire SAM was deleted (⌬500 -570 and ⌬500 -590), the ability of the p73 protein to activate p21 was markedly increased (Fig. 3D). These data suggest that an intact SAM and an intact extreme C terminus are required for their inhibitory function.
To determine whether the integrity of the SAM or the extreme C terminus has any effect on p73 apoptotic activity, we generated H1299 cell lines that inducibly express various p73␣ mutants as analyzed above. One representative clone from each group of cell lines, which expresses a comparable level of various p73 proteins, was selected to examine the ability of these mutants to induce apoptosis (Fig. 3E). We found that when one or more helices of the SAM together with the entire extreme C terminus were deleted (⌬540 -636, ⌬525-636, and ⌬510 -636), the ability of the p73 mutants to induce apoptosis was markedly increased compared with p73␣(⌬555-636), which only lacks the extreme C terminus (Fig. 3F, the upper  four panels). Similarly, when a part of the extreme C terminus together with the entire SAM was deleted (⌬500 -590), the mutant p73 was highly potent to induce apoptosis, reaching a level comparable to p73␣(⌬495-636), which lacks both the SAM and the extreme C terminus (Fig. 3F, the bottom two  panels). Thus, our findings firmly established that the intact state of the SAM and the extreme C terminus is essential for their inhibitory activity.
The C Terminus in p73␣ Inhibits the Function of the Activation Domain, but Not the DNA-binding Domain-A classic transcription factor contains an activation domain and a sequence-specific DNA-binding domain. The activity of both domains is subject to extensive regulations (52,53). To explore the underlying mechanism by which the SAM and the extreme C terminus negatively regulate the activity of p73, we investigated whether the activation domain, the DNA-binding domain, or both are subject to the negative regulation. To test this, we generated chimeric proteins comprising of the Gal4 DNA-binding domain and the p73 activation domain along with the SAM, the extreme C terminus, or both (Fig. 4A). A luciferase reporter under the control of five Gal4-responsive elements was used to examine the transcriptional activity of various chimeric proteins. We found that the chimeric protein, which consists of the Gal4 DNA-binding domain and the p73 activation domain, and the chimeric protein, which consists of the Gal4 DNA-binding domain, the p73 activation domain, and the p73 tetramerization domain were capable of activating the luciferase reporter, whereas the Gal4 DNA-binding domain alone was inactive (Fig. 4A, compare columns 1-3). Interestingly, we found that the activity of the chimeric transcription factor was completely inhibited by both the SAM and the extreme C terminus, alone or in combination (Fig. 4A, compare columns 2-6). However, we were unable to detect a direct interaction between the p73 activation domain and the SAM or the extreme C terminus by several approaches, including mammalian two-hybrid assay and GST-pull-down assay (data not shown).
To determine whether the SAM and the extreme C terminus, alone or in combination, suppress the DNA-binding activity of p73, we performed chromatin immunoprecipitation assay (ChIP) to compare the relative ability of p73␣, p73␣(⌬555-636), p73␣(⌬500 -555), and p73␤ to bind to the p53-responsive elements in the p21 promoter. The p21 promoter contains two p53-responsive elements: p53 RE1 located between nucleotides Ϫ2281 and Ϫ2262 and p53 RE2 located between nucleotides Ϫ1395 and Ϫ1376 (44). To this end, H1299 cells were uninduced (Ϫ p73) or induced (ϩ p73) to express HA-tagged p73␣, p73␣(⌬555-636), p73␣(⌬500 -555), or p73␤. After DNA and protein were cross-linked, the p73-DNA complexes were precipitated by anti-HA antibody, which recognizes the HA-tagged p73. Anti-myc antibody was used as a control. The input loading from each of the samples was comparable (Fig. 4B, compare the input parts within p53 RE1, p53 RE2, and GAPDH panels), suggesting that the amount of extracts from each group of the un-induced and induced cells used for ChIP assay was equivalent. More importantly, we found that a comparable amount of the fragments containing p53-RE1 or p53-RE2 was immunoprecipitated with anti-HA antibody from cell extracts containing HA-p73␣, p73␣(⌬555-636), p73␣(⌬500 -555), or HA-p73␤ (Fig. 4B, ϩp73 p53RE1 and ϩp73 p53RE2 panels). In contrast, no visible fragment containing p53-RE1 or p53-RE2 was immunoprecipitated with anti-myc, a control antibody (Fig. 4B,  ϩp73 p53RE1 and ϩp73 p53RE2 panels). In addition, no fragment was immunoprecipitated with anti-HA or anti-myc from cell extracts with no p73 expression (Fig. 4B, Ϫp73 p53RE1 and Ϫp73 p53 RE2 panels). Moreover, the DNA fragment with no p53-binding site within the promoter of the GAPDH gene was not immunoprecipitated with anti-HA or anti-myc antibody from both un-induced and induced cell extracts (Fig. 4B, ϩp73 GAPDH and Ϫp73 GAPDH panels). This suggests that the SAM and the extreme C terminus do not target the p73 DNA-binding activity for their inhibition.
The SAM and the Extreme C Terminus in p73␣ Are General Inhibitory Domains of Transcription Factors-To determine whether the inhibitory activity of the SAM and the extreme C terminus is specific to p73, we wanted to examine their inhibitory activity toward a cis heterologous activation domain. To this end, we substituted the activation domain in p73␣, p73␣(⌬555-636), p73␣(⌬500 -555), or p73␣(⌬495-636) with p53 activation domain. We found that the chimeric protein, which does not contain the SAM and the extreme C terminus, was highly potent to activate the p21 promoter, but its activity was inhibited by the SAM, the extreme C terminus, or both (Fig. 5A). Similarly, we substituted the p73 activation domain with the Gal4 activation domain. We found that the chimeric protein, which does not contain the SAM and the extreme C terminus, was highly potent to activate the p21 promoter, but its activity was inhibited by the SAM, the extreme C terminus, or both (Fig. 5B). These data indicate that the SAM and the extreme C terminus appear to be a general inhibitory domain of a transcription factor because both of them are capable of repressing the activity of a heterologous activation domain.
The SAM and the Extreme C Terminus Suppress the Interaction of p73 with Transcriptional Coactivators p300/CBP-Previous studies have shown that p73 interacts with p300/ CBP, which stimulates p73 transcriptional activity and its ability to induce apoptosis (54,55). Because the SAM and the extreme C terminus inhibit the function of p73 activation domain, we wanted to investigate whether these inhibitory do- FIG. 3. The intact state of the SAM and the extreme C terminus in p73␣ is essential for their inhibitory activity. A and B, partial deletion of the SAM (A) or the extreme C terminus (B) relieves their individual repression toward the activity of p73 on the p21 promoter. C and D, partial deletion of the SAM (C) and the extreme C terminus (D) relieves their respective suppression toward p73 to induce endogenous p21. E, levels of induced expression of p73␣ and various p73␣ mutants with partial deletion of the SAM or the extreme C terminus in H1299 stable cell lines are comparable. F, partial deletion of the SAM and the extreme C terminus relieves their respective suppression toward p73 to induce apoptosis in H1299 cells.

DISCUSSION
It has been well documented that p73␤ is more active than p73␣ in transactivation (4,45). Because p73␣ contains an extended C terminus, consisting of a SAM and an extreme C terminus (1, 8 -10), it appears that the C terminus is responsible for the decreased transcriptional activity of p73␣. However, the underlying mechanism has never been defined. By generating a series of p73␣ mutants, we have found that both the SAM and the extreme C terminus in p73␣ are inhibitory domains. Furthermore, the intact state of each region is essential for its inhibitory function.
A SAM domain is comprised of five helices (46,47). Several residues in each helix are highly conserved among SAMs from different proteins, suggesting that these conserved residues are critical for SAM functions (50,51). Indeed, mutation of a conserved tyrosine residue to phenylalanine (Y928F) is found to enhance the activity of EphA4 (56). We found that the integrity of SAM is critical to its inhibitory function. Deletion of one helix is able to decrease its repression function. When the last three helices are deleted, the rest of SAM barely suppresses p73 activity. Thus, it is likely that an intact SAM structure is required for the inhibition of p73 activity. A recent report showed that p63␣ SAM, which shares a homology with p73␣ SAM, is also able to suppress p63␣ transcriptional activity (53), although the conclusion is disputed by results of another study (41). In addition, missense mutations in the SAM domain of p63␣ derived from ankyloblepharon ectodermal dysplasia clefting or ectrodactyly ectodermal dysplasia and facial clefting (EEC) syndromes abolish the inhibitory activity of the SAM (57), which highlights the importance of the SAM to p63␣ FIG. 4. The C terminus in p73␣ inhibits the function of the activation domain, but not the DNA-binding domain. A, both the SAM and the extreme C terminus suppress the activity of p73 activation domain in chimeric proteins. The luciferase reporter, pGL2-5Gal, was co-transfected with a vector that expresses Gal4 DBD or a chimeric protein containing Gal4 DBD and various p73␣ proteins. The -fold increase in relative luciferase activity by various Gal4 DBD-p73 chimeric proteins was calculated using Gal4 DBD as a control. B, p73␣, p73␣(⌬555-636), p73␣(⌬500 -555), and p73␤ have similar sequencespecific DNA-binding activity. ChIP assay was performed as described under "Experimental Procedures." HA-p73-DNA complexes were captured with anti-HA monoclonal antibody. Anti-myc antibody was used as a control.
FIG. 5. The SAM and the extreme C terminus in p73␣ are general inhibitory domains of transcription factors. A and B, both the SAM and the extreme C terminus suppress the activities of p53 AD and Gal4 AD. pGL2-p21A was co-transfected with a construct expressing one of the p53-p73 chimeric proteins (A), or a construct expressing one of the Gal4 AD and p73 chimeric proteins (B). The luciferase activity by the empty vector pcDNA3 or pVP16 was used as a control.
function. However, no naturally occurring mutation within p73␣ SAM has been identified, and analogous p73 SAM mutations as those in p63 are unable to attenuate the repression function of p73␣ SAM (data not shown). These data suggest that the SAM in p73␣ and the one in p63␣ exert their inhibitory function through a distinct mechanism.
We found that an intact extreme C terminus is necessary for its inhibitory activity. The extreme C terminus in p73␣ has not been found to form a special structure. Thus, whether a partial deletion of the extreme C terminus impairs its structure, and thus, affects its function remains to be elucidated. Moreover, similar to the extreme C terminus in p73␣, one study showed that the extreme C terminus in p63␣ contains a transcriptionally inhibitory domain (41). Furthermore, similar to the modification of lysine residues in p53 extreme C-terminal region, lysine 627 in p73␣ extreme C terminus is modified by SUMO-1. However, unlike the modification in p53 that enhances p53 activity (58), the SUMO modification does not augment p73␣ activity, instead promotes p73␣ turnover (48). It is possible that the failure of enhancing p73␣ activity by the SUMO modification in the extreme C terminus is due to the presence of the other inhibitory domain, the p73␣ SAM.
The inhibitory domain in p63␣ C terminus is shown to intramolecularly interact with, and suppress the function of, p63 activation domain, presumably by blocking the interaction of the activation domain with the basic transcriptional machinery and coactivators (41). However, this may not be the case for the SAM and the extreme C terminus in p73␣. First, a direct interaction between the C terminus and the activation domain in p73␣ was not detectable by several approaches (data not shown). Second, we found that the SAM and the extreme C terminus in p73␣ are capable of repressing both p53 activation domain and Gal4 activation domain in chimeric proteins. Thus, as a general inhibitory domain, it is less likely that the C terminus in p73␣ exerts its inhibitory function via direct interaction with heterologous activation domains. Indeed, the inhibitory domain in p63␣ specifically binds to, and inhibits, p63 activation domain, but it neither binds to, nor inhibits, p53 activation domain (41). All the evidence indicates that the C termini in p73␣ and p63␣ exert their inhibitory functions through different mechanisms.
The C-terminal basic domain in p53 directly binds and masks its central DNA-binding domain (32). Thus, its inhibitory effect can be removed by structural modifications such as phosphorylation, glycosylation, acetylation, or deletions (32)(33)(34). In our study, however, deletion of the C terminus did not affect the DNA-binding ability, because p73␣ is as potent as p73␤ in binding to p53-responsive elements in the p21 promoter in vivo. Therefore, the C terminus in p73␣ exerts its inhibitory activity by a mechanism also distinct from that by p53.
⌬Np73␣ has been shown to bind to p53 and repress p53 transcriptional activity (5). A well accepted theory for this is that the transcriptionally inactive ⌬Np73␣ competes with p53 for specific DNA-binding sites (5). We have found that the C terminus of p73␣ suppresses the activity of p53 activation domain in chimeric proteins. Thus, our data suggest a novel mechanism by which ⌬Np73␣ inhibits p53 activity. Previously, we showed that ⌬Np73␤ is active in transactivation, but ⌬Np73␣ is not. We also identified a novel transcriptional activation domain at the N terminus of ⌬Np73 (42). Thus, it is likely that the C terminus in p73␣ also represses ⌬Np73 activation domain.
As a transcriptional coactivator, p300/CBP are recruited by many transcription factors to specific promoter regions, where they acetylate core histone tails and promote accessibility of chromatin to the basic transcriptional machinery (59). p300/ CBP interact with the activation domains of p53 and p73, which is crucial to the transcriptional activity of p53 and p73 (33,54). In addition, both p53 and p73 are acetylated by p300 in response to DNA damage, which stimulates their activities (34,55). Furthermore, p300/CBP interaction with, and acetylation of, p53 and p73 are subject to both positive and negative regulations (60,61). In this study, we found that p73␣ is much FIG. 6. The SAM and the extreme C terminus in p73␣ suppress the interaction of p73 with transcriptional coactivators p300/CBP. A, p73␣, p73␣(⌬555-636), p73␣(⌬500 -555), and p73␤ differentially interact with p300. Cell extracts were prepared from ϳ2 ϫ 10 8 cells that were un-induced or induced to express wild-type p73␣, p73␣(⌬555-636), p73␣(⌬500 -555), or p73␤ and immunoprecipitated with anti-HA or anti-myc. 4% of the eluted immunocomplexes was used for Western blot analysis to determine the amount of various p73 captured by immunoprecipitation, and the rest was used to measure the amount of co-immunoprecipitated p300. B, p73␣, p73␣(⌬555-636), p73␣(⌬500 -555), and p73␤ differentially interact with CBP. The experiment was performed as in A. C and D, p300 (C) and CBP (D) have less interaction with p73␣, p73␣(⌬555-636), and p73␣(⌬500 -555) than with p73␤. Immunoprecipitation was performed with anti-300 (C) or anti-CBP (D). 10% of the eluted immunocomplexes was used for Western blot analysis to determine the amount of p300 or CBP captured by immunoprecipitation, and the rest was used to measure the amount of co-immunoprecipitated p73␣, p73␣(⌬555-636), p73␣(⌬500 -555), and p73b. weaker than p73␤ to interact with p300/CBP and deletion of either the SAM or the extreme C terminus does not augment p73␣ interaction with p300/CBP. Taken together, the findings, that the integrity of the SAM and the extreme C terminus is required for their inhibitory activity and that the C terminus of p73␣ does not directly interact with the p73 activation domain, suggest that the SAM and the extreme C terminus form a special structure. Therefore, this structure prevents the full access of p300/CBP to p73 activation domain and thereby suppresses the activity of p73␣. However, within the p53 and p73 chimeric proteins or the Gal4 and p73 chimeric proteins, the SAM and the extreme C terminus suppress the activity of the p53 activation domain and the Gal4 activation domain, but the extent of inhibition is less than that toward p73 activation domain. This indicates that a mechanism other than suppression of the p300/CBP accessibility is also responsible. Because SAM is present in a variety of transcription factors, the accessibility of p300/CBP is a mechanism by which an inhibitory domain exerts its suppression of a transcriptional activator.