Relief of p53-mediated Telomerase Suppression by p73*

Reactivation of telomerase is a feature in many cancer cells. Telomerase activation inhibits telomere shortening, thereby preventing cell cycle arrest and apoptosis activated by shortened telomeres or chromosomal rearrangements. The tumor-suppressor gene product, p53, was previously shown to transcriptionally suppress the activation of the catalytic subunit of telomerase (hTERT). Here we have evaluated the role of p73 in hTERT regulation. We found that ectoptic expression of p73β, in contrast to p73α or p53, in p53 null H1299 cells does not lead to suppression of hTERT transcription. However co-expression of p73α or p73β together with p53 abolished p53-mediated hTERT suppression. This phenomenon was found to be dependent on the DNA binding ability of p73. We also show that p53-mediated suppression of hTERT transcription requires a minimum threshold level of p53, and p73 abrogates p53-mediated suppression by reducing p53 levels through the activation of HDM2. Moreover, p53-mediated hTERT suppression was not relieved by p73β in cells depleted of HDM2 through small interfering RNA-mediated gene silencing. In addition, knockdown of HDM2 in MCF7 cells, which express moderately high levels of p73 and p53, resulted in the reduction of endogenous hTERT levels. Finally, knockdown of p73 in MCF7 cells resulted in increased p53 protein levels and a concomitant decrease in hTERT levels. Together, our data indicate a plausible way by which p73, through HDM2, can oppose p53 tumor suppressor function, thereby possibly contributing to tumorigenesis.

Telomerase is a ribonucleoprotein complex comprising of the catalytic subunit (TERT) and an RNA component (TERC), and adds TTAGGG repeats to chromosome ends thus maintaining the length of telomeres (1). Telomerase is present in human stem cells and progenitors but is hardly detectable in the vast majority of differentiated somatic tissues (2). However, telomerase is reactivated during tumorigenesis, leading to prevention of telomere erosion (3). It has been shown that oncogenes such as Bmi-1 and c-myc, the Papillomavirus E6 protein, and growth-promoting stimuli such as insulin-like growth factor-1 activate human TERT (hTERT) 1 transcription, resulting in increased telomerase activity (4 -7). In addition, we and others have shown that activation of telomerase activity by hTERT overexpression resulted in resistance to several forms of apoptosis and also extended cellular life span (7)(8)(9)(10). By contrast, ablation of telomerase activity by antisense approaches or by expression of the dominant-negative hTERT resulted in enhanced sensitivity to cell death and, consequently, growth inhibition (11)(12)(13)(14). These studies together suggest a pivotal role for telomerase in extending the life span of cells, thereby linking activation of telomerase to cancer development.
The tumor-suppressor gene product, p53, is a crucial component of the cell's defense machinery against damage, and it is often activated by many stress signals (15). p53 prevents the propagation of damaged cells by inducing apoptosis and cellcycle arrest, events that are critical to maintaining a stable cellular environment (16). Consequently, mutations in the p53 gene are considered to represent the most common genetic alterations in human cancer (16). These mutations often affect the DNA binding activity of p53, leading to disruption of the normal function of p53 as a transcription factor, and hence, loss of p53-mediated transcriptional activation of its target genes. In addition to its function as a transcriptional activator, p53 was also shown to be involved in the suppression of several genes, including hTERT. Transcription of hTERT was shown to be down-regulated following the induction of p53 (17)(18)(19)(20)(21)(22). Moreover, telomerase activity was reported to be activated concomitantly with the loss of the remaining functional wildtype p53 allele in Li-Fraumeni skin fibroblasts, suggesting the existence of a p53-dependent regulatory pathway for hTERT control in human cells (23).
p73 is a structural and functional homologue of the p53 tumor suppressor gene that has homology with p53 in the transactivation, tetramerization, and DNA-binding domains (24). However, unlike p53, the p73 protein is expressed as several isoforms because of alternate splicing at the carboxylterminal, giving rise to at least six isoforms (25). In addition, the use of an alternative promoter in the intron 3 of the p73 gene leads to the expression of a p73 protein that lacks the transactivation domain (⌬N-p73) and acts as a dominant-negative suppressor of p73 (26). When ectopically overexpressed in cell culture, both p73␣ and p73␤ closely mimic p53 and induce programmed cell death (25). Moreover, both p73␣ and p73␤ have been shown to transactivate many p53 responsive promoters, although the relative efficiencies vary on different promoters (27). Cell death induced by overexpression of p73 occurs in the absence of p53 and both p73␣ and p73␤ have been shown to suppress foci formation (28).
Despite these functional similarities, the p73 gene is rarely mutated in human cancers (29,30). In addition, mice lacking p73 (p73 Ϫ/Ϫ ) do not succumb to spontaneous tumors (31), unlike their p53 Ϫ/Ϫ counterparts (32,33), raising the possibility that p73 might be involved activities that are dissimilar to p53. Moreover, the role of p73 in telomere maintenance and hTERT regulation has also not been explored. We have therefore investigated whether p73 proteins regulate hTERT expression. Our data demonstrate that ectopic expression p73␤, in contrast to p73␣, does not lead to suppression of hTERT transcription. However, expression of either p73␣ or p73␤ together with p53 abolished p53-mediated hTERT suppression, through a mechanism involving HDM2. Together, the data indicate a possible way by which p73 can oppose p53 tumor suppressor function.

MATERIALS AND METHODS
Cells, Plasmids, and Transfections-H1299 human lung cancer cells (p53 null) used in this study were cultured in 10% serum containing Dulbecco's modified Eagle's medium. H1299 cells stably expressing the His-175p53 mutant has been described (34). H1299 cells stably expressing temperature-sensitive p53 were generated, as has been described (35).
2 ϫ 10 5 cells (in 6-well dishes) were used in transfection experiments using Lipofectamine PLUS-Reagent, as per manufacturer's protocols (Stratagene). Cells were transiently transfected with the various plasmids as indicated in the figure legends. The total amount of transfected DNA was equalized with appropriate amounts of pCDNA3 vector in all cases. Cells were collected 48 h after transfection, and cell extracts were prepared and used for immunoblots and RT-PCR analysis.
pCDNA3-based expression plasmids for p73␣, p73␤, ⌬Np73␣, and ⌬Np73␤ have been described (36). Point mutations of p73␣ and p73␤ FIG. 1. p53 suppresses hTERT levels. A, various concentrations of p53 expression plasmid were co-transfected with Ϫ181-bp hTERT promoter-luciferase reporter construct into H1299 cells, and luciferase activity was assayed 48 h later. All experiments were performed at least three times independently. Error bars are indicated. B, H1299 cells and temperature sensitive (ts)-p53 expressing H1299 cells were transfected with p53 siRNA (ϩ) or control siRNA (Ϫ) 24 h prior to the temperature shift to 32°C. Cells were further cultured at 32°C for 18 h, and RNA was collected and analyzed by semiquantitative RT-PCR for the indicated genes. C, decreasing amounts of protein cell lysates from the above cells were simultaneously obtained and analyzed for telomerase activity by TRAP assay. gapdh, glyceraldehyde-3-phosphate dehydrogenase.
were generated using the site directed mutagenesis kit as per manufacturer's instructions (Stratagene). The various p73 deletion constructs used were generated by PCR cloning as well as generous gifts from G. Melino and M. Levrero. HDM2 small interfering RNA (siRNA), CAGGUGUCACCUUGAAGGUdTT (synthesized by Xeragon), were transfected into cells using TransMessenger Transfection Reagent as per manufacturer's instruction (Qiagen). Vector-based siRNA for p73 was obtained by cloning of ACGTCCATGCTGGAATCCG p73 target sequence into pSUPER (Oligoengine).
Luciferase Assays-H1299 cells were seeded in 6-well plates and transiently transfected with the plasmids indicated (0.5 g) in the corresponding figure legends, along with Ϫ181-bp hTERT promoter reporter constructs (hTERT-luc), which have been described (37). For normalizing, 0.5 g of plasmid encoding the ␤-galactosidase gene was transfected into each well. Cells were collected at 48 h post-transfection. Luciferase assays were performed as described (38).

RESULTS
p53 Suppresses hTERT Expression-p53 has been shown to suppress hTERT expression by several investigators (17)(18)(19)(20)(21)(22). However, Lin and Elledge (40) did not observe an effect on hTERT expression by silencing p53, raising the possibility that p53 might not be involved in the physiological regulation of hTERT. Hence, we first attempted to investigate whether p53 was indeed capable of suppressing hTERT transcription, by using the minimal hTERT promoter-luciferase construct containing the promoter regions up to Ϫ181 from the transcription start site (37). This region of the hTERT promoter is sufficient for p53-mediated suppression and contains multiple SP-1 binding sites (37). Transient expression of wild-type p53 resulted in a significant suppression of hTERT promoter activity in a dosedependent manner, compared with expression of vector (Fig.  1A). We next evaluated whether endogenous hTERT levels would be affected by p53. To this end, H1299 cells expressing a temperature-sensitive p53 mutant was cultured at the permissive temperature of 32°C to activate p53, and the levels of hTERT were evaluated by semiquantitative RT-PCR analysis. Culturing of these cells at 32°C resulted in a decrease in hTERT levels when p53 was active, as determined by expression of p21, the classical p53-target gene (Fig. 1B), indicating that p53 suppressed hTERT expression. We therefore used p53-siRNA to silence p53 expression to determine the effect on hTERT levels. Expression of p53-siRNA resulted in reduced p53 levels and a concomitant decrease in p21 levels (Fig. 1B). Importantly, a decrease in p53 levels resulted in the up-regulation of hTERT levels, indicating that p53-mediated suppres-sion is reduced when p53 levels are reduced (Fig. 1B, compared with loading control). Consistently, determination of endogenous telomerase enzymatic activity by TRAP assay as described (10) also gave similar results (Fig. 1C). Taken together, these data suggest that p53 suppresses hTERT transcription.
Co-expression of p73 with p53 Relieves p53-mediated hTERT Suppression-We next analyzed the effects of p73 on hTERT levels. Expression of p73␣ or p73␤ gave different results. p73␣ resulted in the suppression of hTERT promoter activity similar to that observed with p53, but expression of p73␤ consistently did not result in hTERT suppression ( Fig.  2A), indicating that the different isoforms of p73 have differential effects on hTERT expression.
We next explored whether co-expression of p53 together with the various p73 isoforms would affect hTERT promoter activity, because 50% of human cancers express the wild-type p53 gene, and the p73 gene is not mutated in most cancers (29). Co-expression of either p73␣ or p73␤ with p53 resulted in relief of p53-mediated hTERT suppression ( Fig. 2A). Expression of both p73 isoforms with p53 resulted in the hTERT promoter activity being restored to levels found with vector-transfected cells, indicating that p73 might have a negative effect on p53- 2. p53, p73␣, and p73␤ showing different effects on hTERT promoter activity. A, p53, p73␣, and p73␤ were co-transfected with hTERT promoter-luciferase reporter construct into H1299 cells, and luciferase activity was assayed 48 h later. Expression of p53 or p73␣ alone had a suppressive effect on hTERT transcription, whereas the expression of p73␤ alone had no effect. However, cotransfection of p73 with p53 resulted in relief of p53-mediated suppression on hTERT promoter. All experiments were performed at least three times independently. Error bars are indicated. B, similar experiments were performed and the levels of endogenous hTERT transcripts were determined by semiquantitative RT-PCR. All experiments shown were repeated at least twice. gapdh, glyceraldehyde-3-phosphate dehydrogenase.
FIG. 3. DNA-binding domain of p73 is required for relief of p53-mediated hTERT suppression. A, schematic representation of the various p73 mutants used. B, immunoblot analysis of the mutants used in this study. C, carboxyl-terminal truncations of p73 up to amino acid 344 had no effect on hTERT promoter activity. There was also no effect of carboxyl-terminal p73␤ fragment (amino acids 355-494) on hTERT promoter. However, the DNA-binding domain truncation mutant (amino acids 1-244) had a suppressive effect on hTERT activity. D, point mutation in the DNA-binding domain (R292H) resulted in a suppressive effect on the hTERT promoter, and these mutants were unable to relief p53-mediated hTERT suppression.
To confirm whether the observed effects of p73 on p53-mediated hTERT expression was also observed in vivo, we analyzed the levels of endogenous hTERT mRNA. Expression of p53 resulted in a marked suppression of hTERT mRNA levels (Fig. 2B). Similarly, expression of p73␣ alone, but not p73␤, resulted in reduction in hTERT levels (Fig. 2B), confirming the data obtained with the luciferase reporter assays. Moreover, co-expression of either p73␣ or p73␤ with p53 resulted in abrogation of p53-mediated suppression of hTERT mRNA levels, similar in extent to that observed in vector-transfected cells (Fig. 2B). These data together indicate that expression of p73 results in relief of p53-mediated hTERT suppression in vivo.
DNA Binding Activity of p73 Is Required for Reversion of p53-mediated hTERT Suppression-We next investigated the probable mechanism through which p73 expression relieved p53-mediated hTERT suppression. We employed several p73 mutants to evaluate the role of the various p73 domains in this process (Fig. 3A). All mutants were expressed efficiently as determined by immunoblot analysis (Fig. 3B). Carboxyl-terminal deletions of the p73␤ up to amino acid 344 that results in the deletion of the oligomerization domain had no effect on the hTERT promoter, similar to that found with full-length p73␤ (Fig. 3C). However, further deletions resulting in truncation of the DNA-binding domain (DBD) resulted in the mutant p73 being able to suppress hTERT promoter, similar to that found with p53 (Fig. 3C). This suggested that abolishing the DNA binding activity property of p73 resulted in suppression of hTERT. To further confirm whether the DNA binding activity of p73␤ negatively affected its ability to suppress hTERT promoter activity, we analyzed the effect of 292-p73 mutant, which is unable to bind to DNA and is transcriptionally inactive (41). Expression of p73␣-292 did not differ from full-length p73␣ in its ability to suppress hTERT promoter (Fig. 3D). However, p73␤-292 was now found to suppress the hTERT promoter activity (Fig. 3D), confirming that the DNA binding activity of p73␤ was required for abrogating this suppressive effect on hTERT promoter. Moreover, co-expression of p73␣-292 or p73␤-292 with p53 did not result in relief of p53-mediated hTERT suppression (Fig. 3D). Together, the data indicate that although the different pathways may be utilized by the various p73 isoforms to differentially regulate hTERT promoter activity, the DNA binding activity of both of the p73 isoforms is required for relief of p53-mediated suppression.
p53-mediated hTERT Suppression Is Dependent on Threshold Levels of p53 That Are Regulated by p73-To further explore the possible ways by which p73 affected p53-mediated hTERT suppression, we first investigated whether expression of p73 would destabilize p53, resulting in reduced p53 steadystate levels. We employed H1299 cells stably expressing the p53-175R mutant. p53-175R is transcriptionally inactive and is unable to activate HDM2; hence, it is stable, and cells expressing this mutant do not undergo apoptosis (34). These cells were transiently transfected with various amounts of p73␣, p73␤, or HDM2 expression vectors, and the levels of endogenous p53 were monitored by immunoblotting. Both p73␣ and p73␤ expression led to decreased p53 levels, although p73␤ appeared to be more potent in reducing p53 levels (Fig. 4A,  compare lanes 2 and 3 with 4 and 5). There was a slight decrease in p53 levels when p73␤ was expressed at lowered levels, and p53 expression was almost absent at higher doses of p73␤, similar to that found with expression of the negative regulator HDM2 (Fig. 4A). Similar results were obtained with wild-type p53 in transient transfection experiments (data not shown). However, expression of high levels of transactivation-defective p73 DNA-binding domain mutants 292-p73␣ and 292-p73␤ did not result in reduction of p53 levels (Fig. 4A, lanes 6  and 7), suggesting that the DNA-binding/transcriptional function of p73 is required for p53 degradation.
Because p73 expression destabilized steady-state p53 levels, we next investigated whether there was a critical threshold level of p53 that was required to suppress hTERT expression. Expression of various amounts of p53 in H1299 cells revealed that lower levels of p53 (up to 0.2 g) were not sufficient to inhibit endogenous hTERT mRNA levels (Fig. 4B). However, hTERT levels were reduced when 0.5 g of p53 was expressed (Fig. 4B, lane 5), suggesting that a critical level of p53 was required for p53-mediated hTERT suppression. When 0.5 g of p53 was co-expressed with a HDM2 expression plasmid, the levels of p53 fell below this threshold level, and the hTERT suppression was relieved (Fig. 4B, lane 6), confirming that FIG. 4. Critical threshold levels of p53 are crucial for hTERT suppression. A, increasing amount of p73␣ and p73␤ were transfected into H1299 cells stably expressing the p53-175R mutant. Immunoblot of p53 reveal slight degradation of p53-175R by p73␣ and strong degradation of p53-175R by p73␤. The DNA-binding domain point mutant p73 did not show strong degradation of p53-175R. HDM2 was used as a positive control to demonstrate degradation of p53-175R. B, varying amounts of p53 were transfected into H1299 cells, and the levels of endogenous hTERT transcripts were determined by semiquantitative RT-PCR. The lower third panel is the RT-PCR of hTERT showing that suppression of hTERT expression by p53 occurs only when high levels of p53 are present (0.5 g). Degradation of p53 by HDM2 reversed the suppressive effect of p53 on hTERT transcription. Top two panels are immunoblots against p53 and EGFP, which was used to normalize the transfection efficiency. gapdh, glyceraldehyde-3-phosphate dehydrogenase.
p73 Antagonizes p53-mediated hTERT Suppression reduction in p53 levels below a threshold level, in this case by its negative regulator HDM2, abrogated p53-mediated hTERT suppression. Together, the data indicate that expression of p73 results in reduction of p53 levels and probably below the threshold required to suppress hTERT transcription, thereby relieving the suppressive effect by p53 on hTERT expression.
HDM2 Is Required for p73-mediated Relief of p53-dependent hTERT Suppression-HDM2 is a transcriptional target of both p53 and p73 and is a specific E3-ligase targeting p53 for degradation (42,43). Its expression was induced by both p73␣ and p73␤, concomitant with a decrease in p53 levels (Fig. 4A), suggesting that HDM2 might act as an intermediate in regu -FIG. 5. Relief of p53-mediated hTERT suppression by p73 is because of enhanced expression of HDM2. A, p53-mediated hTERT suppression can be relieved by co-expression of HDM2. Expression of HDM2 alone also marginally increased hTERT promoter activity. B, knockdown of HDM2 in H1299 cells using siRNA against HDM2. HDM2 transcript levels were determined by RT-PCR analysis. C, co-expression of p73␤ did not alleviate p53-mediated hTERT suppression when HDM2 was silenced using siRNA against HDM2. D, endogenous hTERT transcript levels were determined by semiquantitative RT-PCR after knockdown of HDM2 in similar experiments as described above. E, co-expression of p53 and p73 augment HDM2 expression, as determined by semiquantitative RT-PCR. gapdh, glyceraldehyde-3-phosphate dehydrogenase.
p73 Antagonizes p53-mediated hTERT Suppression lating p73-mediated relief of p53-dependent hTERT suppression. Therefore initial experiments were performed to evaluate the effect of expressing HDM2 on hTERT promoter activity. Expression of HDM2 alone was sufficient to slightly augment hTERT promoter activity (Fig. 5A). In addition, co-expression of p53 and HDM2 resulted in abrogation of p53-mediated hTERT suppression. We therefore analyzed the effect of silencing endogenous HDM2 levels using siRNA against HDM2, which resulted in decrease of endogenous HDM2 levels (Fig.  5B), on the effects of co-expression of p53 and p73␤ on hTERT promoter activity. Although p73␤ was able to reverse p53mediated hTERT suppression, this effect was abrogated in the presence of siRNA against HDM2 (Fig. 5C), suggesting that HDM2 was required for p73-dependent relief of p53-mediated hTERT suppression. In addition, expression of HDM2-siRNA alone resulted in a decrease in hTERT promoter activity, consistent with earlier data showing that HDM2 itself could contribute to telomerase activity (Fig. 5C). Co-expression of HDM2-siRNA with p53 resulted in a further decrease in hTERT promoter activity, indicating that HDM2 could regulate hTERT activity through p73-dependent and -independent mechanisms. Furthermore, analysis of the endogenous hTERT levels indicated that p53-mediated suppression of hTERT levels was abrogated by the expression of p73␤, and this relief was ameliorated in the presence of siRNA against HDM2 (Fig. 5D,  compare lanes 4 with 5).
Nevertheless, it was puzzling that p53 was able to suppress hTERT, because p53 itself is an activator of HDM2. One possibility is that the levels of HDM2 activated by p53 are insufficient to inhibit p53-mediated hTERT suppression. To explore this possibility, we investigated whether co-expression of p73 with p53 would augment HDM2 mRNA levels. Transient transfection of p53, p73␣, and p73␤ resulted in expression of HDM2, although p73␤ was found to activate HDM2 to the highest level (Fig. 5E, lane 4 and Fig. 4A), as reported earlier (43). By contrast, p73␣ was the weakest among the three proteins tested to transactivate HDM2 (Fig. 5E, lane 3). Co-expression of p53 with either p73␣ or p73␤ resulted in significantly higher levels of HDM2 compared with expression of p53 alone (Fig. 5E,  compare lane 2 with 5 and 6), indicating that HDM2 expression p73 Antagonizes p53-mediated hTERT Suppression is augmented by co-expression of wild-type p53 and p73. Together, the data suggest that p73 relieves p53-mediated hTERT suppression through the activation of HDM2.
Endogenous p73 and HDM2 Are Required for Maintenance of hTERT Levels in Vivo-In an attempt to investigate the effects of silencing endogenous p73 and HDM2 expression on hTERT expression in vivo, we employed the MCF7 breast cancer cell line that expresses wild-type p53 as well as moderate levels of p73. siRNA-mediated silencing of endogenous p73 resulted in a visible reduction of endogenous hTERT levels (Fig. 6A). This was accompanied by an increase in the endogenous p53 levels (Fig. 6B), further confirming that p73 can negatively regulate p53 expression levels. Moreover, silencing of endogenous HDM2 in MCF7 cells also resulted in a significant decrease in endogenous hTERT levels (Fig. 6C), further indicating that HDM2 is crucial in regulating hTERT levels. DISCUSSION We have evaluated the role of p73 in regulating hTERT expression. The data presented here demonstrate that although hTERT expression is differentially regulated by the various isoforms of p73, co-expression of any of the two major isoforms of p73 with p53 resulted in abrogation of p53-mediated hTERT suppression. These data together highlight sev-eral interesting possibilities that are relevant to carcinogenesis, (i) there are significant functional differences between p73␣ and p73␤ in suppressing hTERT, suggesting that each isoform might have been evolved for differential and specific functions; (ii) low basal p53 levels are not sufficient to suppress hTERT expression, and a critical threshold level of p53 is required to bring about this effect, as has also been highlighted by Xu et al. (19); (iii) increased expression of p73 observed in several cancers (30) might not be a consequence of cancer development but could even possibly play a contributory role, as p73 expression leads to abrogation of p53-mediated tumor-suppressive functions such as hTERT suppression.
This study highlights the fact that certain threshold levels of p53 are required for hTERT suppression. Although elevated expression of p53 either by overexpression or by induction with DNA-damaging agents was shown to suppress hTERT expression, it was also noted that inactivation of wild-type p53 would not necessarily result in telomerase activation (19), suggesting that p53-mediated hTERT suppression probably occurs at levels of increased p53 activity. Our results indicate that overexpression of p53 at lower levels did not result in hTERT suppression until a threshold level of p53 was attained (Fig. 7A). This suppressive effect was abrogated by expression of HDM2, FIG. 7. Model of telomerase regulation by p53 and p73. A, the model shows the individual roles of p53 and p73 isoforms. The uppermost panel shows that a threshold level of p53 is required to suppress telomerase activity. Although p53 is able to induce HDM2, which by itself could activate telomerase activity mildly, a critical level of HDM2 is required to inhibit p53 activity, below which HDM2 is not effective in inhibiting p53-mediated telomerase suppression. The middle panel shows that p73␣ is a strong inhibitor of telomerase in contrast to p73␤, which contributes to the maintenance of telomerase activity, through the activation of HDM2 and perhaps other targets (lower panel). B, relief of p53-mediated telomerase suppression by p73. When p73 is overexpressed, as seen in many cancers, the inhibitory effect of p53 on telomerase is abrogated as the threshold levels of HDM2 expression is reached. Increased HDM2 results in reduction of p53 protein levels and activity and also contributes to telomerase activation itself. Thus, the combined effect of p53 and p73 co-expression results in maintenance of telomerase promoter activity even in the presence of high p53, thus probably contributing to carcinogenesis. ⅐⅐⅐⅐, weak activity; --, strong activity; 3, activation; ٜ, inhibition.
p73 Antagonizes p53-mediated hTERT Suppression the p53-specific E3 ligase, indicating that the levels of p53 is a critical factor in determining whether hTERT was suppressed by p53. Hence, it is conceivable that p53 suppressed hTERT expression in situations of cellular stress, fulfilling its role as a guardian of the genome. It is noteworthy that Lin and Elledge (40) found no significant effects on hTERT expression by silencing p53 expression, leading them to speculate that p53 might not have a physiological role in hTERT regulation. However, our data indicate that p53-mediated hTERT suppression can be relieved by silencing p53 (Fig. 1B) and imply that p53 probably regulates hTERT under situations when p53 is up-regulated. Thus, these data could explain the lack of any effect on telomerase activation by inactivating wild-type p53 (19), because such an effect would conceivably be felt in the presence of stress signals. Moreover, as telomerase was also shown to have antiapoptotic properties (8,10), it is conceivable that suppression of telomerase upon stress ensures that the damaged cells are removed to prevent the propagation of these cells. Furthermore, p73 has also been recently shown to be activated by several stress signals (39). Therefore, the fine interplay of the p53 family members may fine-tune the expression of hTERT and hence modulate the sensitivity to stress-induced apoptosis.
We have demonstrated that the different isoforms of p73 have different properties in regulating hTERT expression. p73␣ suppressed hTERT expression, similar to p53. By contrast, p73␤, which lacks the carboxyl-terminal SAM domain, had no effect on hTERT expression on its own (this report). These findings suggest that the S-adenosylmethionine domain may contribute to the negative inhibitory effect of p73␣ on hTERT expression, probably through its inhibitory activity (44), or its weak activity in activating HDM2 (Fig. 7A), and suggest that the mechanisms by which p53 and p73␣ suppress hTERT might be different. Moreover, investigations using deletion mutants showed that the partial deletion of the DNAbinding domain-(1-249) resulted in suppression of hTERT activity. In addition, the DNA-binding domain mutant of p73␤, p73␤-292, also acquired a suppressive effect on hTERT activity, together suggesting that the DNA binding activity is required to restrain p73 from suppressing hTERT activity. Thus, because p73␣ is a weak transcriptional activator compared with p73␤ (43), the difference between the p73 isoforms in regulating hTERT expression could also be because of their different transactivation abilities.
The most unexpected finding of this study is that p73 expression led to relief of p53-mediated hTERT suppression, instead of having a synergistic effect with p53. This was found to be dependent on the ability of p73 to induce its transcriptional target, HDM2 (Fig. 7, A and B). It is important to note that although p73␣ and p73␤ displayed differential effects on hTERT activity when expressed individually (Fig. 7A), both abrogated p53-mediated hTERT suppression (Fig. 7B). The induction of HDM2 is one mechanism, which is employed by p73 to negate the p53-mediated suppressive effect on hTERT promoter. Although p73␣ induced HDM2 to much lower extents than p73␤ as has been described previously (43), coexpression of p73␣ with p53 also resulted in significantly elevated HDM2 expression. Consistently, siRNA-mediated silencing of either p73 or HDM2 resulted in a loss of relief of hTERT suppression by p73, resulting in reduced hTERT levels. However, it should be noted that HDM2 expression was not involved in regulating p53 levels alone, as overexpression of HDM2 in the absence of p53 also resulted in a slight increase of hTERT activity. Furthermore, siRNA-mediated HDM2 silencing in the absence of p53 also led to decreased hTERT levels, suggesting that HDM2 could regulate hTERT by other p53independent mechanisms, such as binding to SP-1 as reported (45). Taken together, it is evident that p73 is able to negate p53-mediated hTERT suppression through a mechanism involving HDM2.
The current understanding of p53 and p73 functions do not allow us to envisage a physiological situation in which both p53 and p73 are active, and yet cells are not undergoing apoptosis. However, the data presented here suggest that under conditions in which p53 and p73 are simultaneously activated, it might be possible that the some of the p53 functions could be negated by p73 expression. This ascribes an "anti-p53" role for p73 and would imply that p73 could contribute to cancer development, perhaps under selective conditions. Although this may appear to be a radical idea for a protein that is the structural and functional homologue of p53, it needs serious consideration in view of the genetic and human clinical data showing overexpression of the full-length p73 isoforms (30). It is thus not completely unconceivable that full-length p73 may have anti-p53 or "pro-oncogenic" properties.