Role of p73 in Regulating Human Caspase-1 Gene Transcription Induced by Interferon-γ and Cisplatin*

Caspase-1, a cysteine protease is primarily involved in proteolytic activation of proinflammatory cytokines such as interleukin-1β. It is also involved in some forms of apoptosis. Here we have analyzed the role of p73, a homolog of tumor suppressor p53, in regulating human caspase-1 gene transcription. The caspase-1 promoter was strongly activated by p73α and p73β primarily through a p53/p73 responsive site. Overexpression of p73 by transient transfection increased the caspase-1 mRNA level. Treatment of cells with cisplatin (which increases p73 protein level) resulted in increased caspase-1 promoter activity and its mRNA level. Blocking of p73 function by a dominant negative mutant reduced basal as well as cisplatin-induced caspase-1 promoter activity. Mutation of the p73 responsive site abolished cisplatin-induced activation of the promoter. Interferon-γ induced caspase-1 promoter activity and this was reduced by p73-directed small hairpin RNA and also by a dominant negative mutant of p73. Abrogation of the p73 responsive site partially inhibited interferon-γ-induced activation of the caspase-1 promoter. Treatment of HeLa cells with interferon-γ resulted in an increase in p73 protein as well as its activity. Mutation of the IRF-1 binding site abolished interferon-γ-induced caspase-1 promoter activity but p73-induced activation was only marginally reduced. IRF-1 cooperated with p73 and cisplatin cooperated with interferon-γ in the activation of the caspase-1 promoter. Our results show that p73 is a regulator of caspase-1 gene transcription, and is required for optimal activation of the caspase-1 promoter by interferon-γ.

The p53 family consists of three distinct genes (p53, p63, and p73), which encode proteins showing structural and functional similarities. p53 family proteins are modular molecules having three major conserved domains, an N-terminal transactivation domain, a central sequence-specific DNA binding domain, and a C-terminal oligomerization domain (1)(2)(3)(4)(5). The p73 gene generates multiple isoforms that vary in their N and C termini. The use of the cryptic promoter generates ⌬N isoforms lacking the transactivation domain present in the N terminus of the p73 protein, and alternative splicing gives rise to several isoforms with various C-terminal ends (1)(2)(3)(4)(5). Mice functionally deficient for all p73 isoforms exhibit profound defects including chronic infections and inflammation (6). The molecular mechanism of this defective immune response is not known. p73 has the ability to induce cell cycle arrest and apoptosis and can transactivate some p53 targets, but the efficiency with which each of them is induced is different (7). Only about one-sixth of p53-induced genes show increased expression upon overexpression of p73 (8). Unlike p53, which is induced by a variety of DNA-damaging agents and cellular stress, p73 is induced only by a subset of agents such as ␥-irradiation and cisplatin (9 -10). In addition, p73 is up-regulated during T-cell activation, muscle and neuronal differentiation, and 12-O-tetradecanoylphorbol-13-acetate (TPA)-induced monocytic differentiation (11)(12)(13)(14)(15).
Caspase-1 (also known as interleukin-1␤-converting enzyme) is involved in proteolytic activation of cytokines such as IL-1␤ and IL-18. Caspase-1 plays a role in various forms of apoptosis such as IFN-␥induced apoptosis in fibroblasts, p53-induced apoptosis, DNA damage, and IRF-1-mediated apoptosis, Fas-induced apoptosis, etc. (24 -30). Caspase-1 gene expression is induced by IFN-␣, IFN-␥, and by DNA damage-induced activation of p53 (24 -28, 31). Activation of caspase-1 gene expression requires IRF-1 and Stat-1 (24,32). Analysis of the human caspase-1 promoter has shown an IRF-1 binding site, which overlaps with the initiator element (33,34). The IRF-1 binding site is completely conserved in the murine caspase-1 promoter (35). There is no TATA box in the caspase-1 promoter (34). Although Stat-1 is required for caspase-1 gene expression, no Stat-1-responsive site has been identified in the promoter (22,24). The precise role of Stat-1 in regulating caspase-1 gene transcription directly is not clear. Stat-1 activates transcription of IRF-1, which is an activator of caspase-1 gene transcription (32). However in Stat-1-deficient cells, introduction of IRF-1 does not restore constitutive levels of caspase-1 gene expression (22). In addition to Stat-1 and IRF-1, other transcription factors are likely to be involved in regulating caspase-1 gene transcription. Recent work in our laboratory has shown that endogenous p53 activated by treatment of cells with doxorubicin can regulate human caspase-1 gene transcription, through a functional binding site present at Ϫ85 to Ϫ66 position relative to the transcriptional start site (27). This p53 binding site is not conserved in the murine caspase-1 promoter.
Here we have analyzed the role of p73 in regulating human caspase-1 gene transcription. p73␣ and p73␤ isoforms strongly activated the caspase-1 promoter. Treatment of cells with anticancer drug cisplatin resulted in increased caspase-1 promoter activity, which was inhibited by blocking p73 function. IFN-␥-induced caspase-1 promoter activity required the p73 responsive site as well as functional p73 protein for optimal response. Our results suggest that IFN-␥-induced caspase-1 gene transcription involves interdependent action of p73 and IRF-1.

EXPERIMENTAL PROCEDURES
Expression Vectors and Antibodies-The expression vectors for various isoforms of p73 namely p73␣, p73␤, p73␥, and p73␦ cloned inframe with the hemagglutinin tag into pcDNA3-HA were a kind gift from Dr. Gerry Melino, University of Rome (36). Plasmids for expressing p73DD and p53DD cloned in-frame with the T7 tag into pcDNA3-T7 were a kind gift from Dr. William G. Kaelin, DFCI, Harvard Medical School (37). A mutant (L371P) of p73DD was made by replacing Leu 371 of p73 by Pro using a PCR-based site-directed mutagenesis strategy. This mutation is known to inactivate p73DD (37). Human IRF-1 cDNA was amplified from HeLa cell RNA by RT-PCR and cloned in the mammalian expression vector pCI (Promega). Cdk2, Stat-1, IRF-1, C3G, and caspase-1 antibodies were from Santa Cruz Biotechnology; p73 monoclonal antibody (ER15) was from Neomarkers Inc., Union City, CA; T7 tag antibody was from Novagen.
Construction of a Vector Expressing shRNA-The shRNA expression vector targeting p73 was constructed using the U6 promoter-based vec-tor essentially as described (38,39). The desired synthetic oligonucleotides were annealed and cloned into the BbsI-Xba-digested U6 promoter plasmid mU6 pro. The p73 sequence targeted by shRNA was from nucleotides 638 -658 (GenBank TM accession NM_005427). The sequences of oligonucleotides used for cloning were 5Ј-tttGGCCATG-CCTCTTTACAAGAAttcaagagaTTCTTGTAAACAGGCATGGCCttttt-3Ј and 5Ј-ctagaaaaaGGCCATGCCTGTTTACAAGAAtctcttgaa-TTCTTGTAAAGAGGCATGGC-3Ј. This vector expressed sense, hairpin, and antisense sequence. A vector expressing shRNA of unrelated sequence of the same length was used as a control.
Cell Culture and Transfections-The cell lines were maintained at 37°C in a CO 2 incubator in Dulbecco's modified Eagle's medium supplemented with 10% fetal calf serum. The transfections were done using Lipofectamine Plus TM reagent (Invitrogen Life Technologies, Inc.) according to the manufacturer's instructions. All the plasmids for transfection were prepared by using Qiagen columns. Human IFN-␥ (Sigma) was added wherever indicated at a final concentration of 100 ng/ml.
RT-PCR-Total RNA was isolated using the TRIzol TM reagent (Invitrogen Life Technologies, Inc.). Semiquantitative RT-PCR was carried out essentially as described previously (27,40). RNA was reversetranscribed using reagents from the first strand cDNA synthesis kit (Invitrogen Life Technologies, Inc.). Primers for amplification of caspase-1 and GAPDH have been described previously (27,40). IRF-1 was amplified from HeLa cell RNA using appropriate primers, which amplified the entire coding sequence.
Western Blot Analysis-After treatment, cells were washed twice with phosphate-buffered saline and lysed in 1ϫ SDS sample buffer. Proteins were separated on 10% SDS-polyacrylamide gels and blotted onto nitrocellulose membranes. The blot was washed twice with Tween-Tris-buffered saline before blocking nonspecific binding with 5% nonfat dry milk (BLOTTO, Santa Cruz Biotechnology). The p73 FIGURE 1. Transactivation of the caspase-1 promoter by p73 isoforms. A, schematic representation of the caspase-1 promoter-reporter constructs, pC-WT and pC-MT-p53. The locations of p73/p53 site and IRF-1 responsive site/initiator element (Inr) are shown. B, HeLa cells in 24-well plates were transfected with pCMV.SPORT-␤Gal plasmid (100 ng) and pC-WT or pC-MT-p53 reporter plasmids (200 ng) along with the indicated p73 isoform (8 ng) or control plasmids. Cell lysates were prepared 28 h after transfection for reporter assays. CAT activities relative to control are shown (n ϭ 4). C, HeLa cells were transfected with pC-WT (100 ng) with indicated amounts of p73 plasmids. CAT activities relative to control are shown.
antibody (1:100) was added, and the blot was incubated overnight at 4°C. The blot was washed three times, and detection was performed by horseradish peroxidase-conjugated secondary antibody using the ECL method (Roche Applied Science). The caspase-1, Cdk-2, and other antibodies were used at 1:500 dilution, and the blot was incubated for 1 h at room temperature. The blots were washed three times, and detection was performed by using alkaline phosphatase-conjugated secondary antibody as described previously (40).
Reporter Plasmids and Reporter Assays-The reporter plasmid pC-WT, which contains the human caspase-1 promoter from Ϫ182 to ϩ42 relative to the transcriptional start site cloned upstream of CAT reporter gene has been described (33). The reporter plasmid pC-MT-p53, derived from pC-WT by mutating the p53 responsive site, has also been described previously (27). The reporter plasmid pC-MT-IRF with mutated IRF-1 responsive site was made by PCR-based site-directed mutagenesis using 5Ј-GGA AGG CGA AGC ATA CCC TCA GCC TCA GTC ACA CAA GAA GGG-3Ј and its complementary oligonucleotide as mutagenesis primers (mutated sites are underlined). The nucleotide sequence was confirmed by automated sequencing. The reporter construct pCAT-p73-2 was prepared by cloning two sets of oligonucleotides (with p73 responsive sites) in the pCAT promoter plasmid (Promega) downstream of the SV-40 promoter in PstI and BamHI sites. The sequence of the insert is CTGCAGTTACAGGCAT-GTGCCACCATGCCTGACTCTGCAGGTCGACTCTAGAGGAT-CCAGCCAGGCATGGTGGTGCATGCCTGTAAGGATCC.
The core sequence of p73/p53 responsive element is underlined in this sequence. This reporter construct was responsive to p53 as well as p73. Cells grown in 35-mm dishes or 24-well plates were transfected with pC-WT (or pC-MT-p53), pCMV.SPORT-␤-gal (Invitrogen Life Technologies, Inc.) and indicated amounts of plasmids expressing various p73 isoforms or IRF-1. The total amount of plasmid in each transfection was kept constant (1 g for a 35-mm dish and 0.4 g for 24-well plate) by adding control plasmid. Preparation of lysates and CAT assays were carried out as described (27). Relative CAT activities were calculated after normalizing with ␤-galactosidase enzyme activities.

RESULTS
Transactivation of the Caspase-1 Promoter by p73 Isoforms-We evaluated the ability of different isoforms of p73 to regulate the activity of the caspase-1 promoter. HeLa cells were transfected with either p73␣or p73␤-expressing plasmids together with wild-type caspase-1 promoter-reporter (pC-WT) or caspase-1 reporter with mutated p53binding site (pC-MT-p53) plasmids (Fig. 1A). A large increase in CAT activity was obtained with p73␣ (34-fold) and p73␤ (43-fold) at low concentrations (8 ng) of the plasmid used (Fig. 1B). The mutant reporter plasmid pC-MT-p53, which has mutated p53 binding site showed only 1.8-and 2.3-fold increase in activity by p73␣ and p73␤, respectively. These results suggested that p73 activates caspase-1 gene transcription primarily through the same site that is recognized by p53 (p73/p53 responsive site, Fig. 1A). The activation of the caspase-1 promoter by p73␣ and p73␤ increased with increasing concentrations of the plasmids (Fig. 1C). At low concentrations (1 ng), both these isoforms of p73 showed similar activation of the promoter (about 10-fold), but at higher concentrations (100 ng), p73␤ showed much more activation (Fig. 1C). p73␥ and p73␦ could also activate the caspase-1 promoter mainly through the p73/p53 responsive site (data not shown).
Activation of the Caspase-1 Promoter by Endogenous p73-Previously, it has been shown that overexpression of a p53 mutant (p53DD) or p73 mutant (p73DD) encompassing their respective oligomerization domains blocked the function of corresponding wild-type proteins (37,41,42). Furthermore, p53-dependent transcriptional activation and p53dependent apoptosis were not affected by p73DD. We determined the FIGURE 2. Regulation of the caspase-1 promoter by endogenous p73. A, inhibition of p73␤-and p73␦-mediated transactivation of caspase-1 reporter by p73DD. HeLa cells in 6-well plates were transfected with pC-WT plasmid (200 ng) along with plasmid encoding p73␤ or p73␦ (200 ng), in the presence or absence of plasmids encoding p53DD and p73DD. CAT activities relative to control are shown (n ϭ 3). The ratio of p73␤ or p73␦ to mutant plasmid (p73DD or p53DD) was 1:2. B, transactivation of the caspase-1 promoter by endogenous p73. pCMV.SPORT-␤Gal plasmid and pC-WT reporter plasmid were transfected along with p53DD, p73DD, or control plasmid and 24-h post-transfection, they were treated with cisplatin (10 M) or vehicle (0.1% DMF) for 24 h. CAT activities relative to control are shown (n ϭ 4). C, immunoblot showing expression of p73DD and p53DD using T7 epitope antibody. D, HeLa cells were transfected with reporter plasmids pC-WT or pC-MT-p53 (100 ng), and after 24 h of transfection cisplatin (25 M) was added. Cell lysates were prepared after another 24 h. CAT activities relative to untreated controls are shown.
affect of coexpression of p53DD and p73DD on transactivation function of p73␤ and p73␦. As shown in Fig. 2A, p73DD completely abolished the transactivation of pC-WT reporter by p73␤ and p73␦, which showed over a 100-fold increase in activity when used at high levels (200 ng of plasmid). As expected, p53DD did not inhibit transactivation function of p73 isoforms. Next we examined the affect of p53DD and p73DD on cisplatin-induced transactivation of the caspase-1 promoter. Cisplatin has been shown to increase the amount of p73 protein in cells (9). HeLa cells were transfected with the required plasmids and 24-h post-transfection, these cells were treated with 10 M cisplatin for 24 h. Untreated controls were cells to which only vehicle (DMF) was added. Cisplatin treatment resulted in an increase of caspase-1 promoter activity, which was abolished by p73DD coexpression (Fig. 2B). Cisplatin-induced increases in caspase-1 promoter activity were not inhibited by p53DD. The expression of p73DD and p53DD was confirmed by Western blotting (Fig. 2C). There was marked decrease in basal activity of the caspase-1 promoter in the presence of p73DD (80% reduction) as compared with control, whereas no inhibition of CAT activity was observed with p53DD coexpression (Fig. 2B). The mutant promoter (pC-MT-p53) was not activated by treatment of cells with cisplatin (Fig. 2D). These results suggested that up-regulation of caspase-1 promoter activity by cisplatin treatment was mediated by p73 through the p73/p53 responsive site. In addition, these results suggested that the basal level of caspase-1 promoter activity required p73 function.
p73-dependent Expression of Caspase-1 mRNA-The level of caspase-1 mRNA was determined by semi-quantitative RT-PCR analy-sis in response to transient overexpression of p73␤ and p73␦ in MCF-7 cells, which have extremely low levels of endogenous caspase-1. Caspase-1 mRNA levels increased severalfold by overexpression of p73␤ and p73␦ as compared with the control transfected cells (Fig. 3A). By using appropriate primers we found that only the ␣-isoform of caspase-1 was induced in MCF-7 cells (data not shown). There was no increase in the level of IRF-1 mRNA in response to overexpression of p73 isoforms (Fig. 3A). Overexpression of p73␣ and p73␤ in U2OS osteosarcoma cells also resulted in increased caspase-1 mRNA levels (Fig. 3A, lower panel).
To analyze the effect of endogenous p73 protein on caspase-1 gene expression we carried out RT-PCR analysis after cisplatin treatment of two human tumor cells. Treatment of p53-negative K562 cells with cisplatin up-regulated the caspase-1 mRNA severalfold (Fig. 3B). In addition, treatment of K562 cells with some apoptosis-inducing agents, staurosporine and cycloheximide did not increase caspase-1 mRNA (Fig. 3B) suggesting that caspase-1 up-regulation is not a general response during apoptosis. Cisplatin treatment of HeLa cells lead to an increase in caspase-1 mRNA (Fig. 3D). RT-PCR analysis with appropriate primers revealed that both ␣and ␤-isoforms of caspase-1 were up-regulated in HeLa cells upon cisplatin treatment whereas the ␥and ␦-isoforms of caspase-1 were not detected (Fig. 3D). Cisplatin treatment increased p73 protein levels in HeLa cells and also in K562 cells (Fig. 3,  C and E). The procaspase-1 protein was processed upon cisplatin treatment of HeLa cells (Fig. 3E). Immunoblotting was performed with total proteins from cisplatin-treated HeLa cells at indicated time points, using antibodies against p73, caspase-1, and Cdk-2. F, p73 is required for basal expression of caspase-1. Immunoblotting was performed with total proteins isolated from clones of HeLa cells expressing p73DD or L371P mutants of p73DD (p73mDD). Western blots with antibodies against caspase-1, T7 tag (p73DD, p73mDD), and Cdk-2 are shown.
Our results suggested that p73 had a role in basal caspase-1 promoter activity. We hypothesized that expression of caspase-1 should be low in the cells in which endogenous p73 function is inhibited by expressing a dominant negative mutant of p73. To test this, we obtained clones of HeLa cells expressing p73DD or mp73DD (L371P mutant of p73DD) by transfection followed by selection in G418 for 15 days. Their expression was confirmed by probing the immunoblot with T7 tag antibody (Fig.  3F). The L371P mutant of p73DD has been shown to be functionally inactive (37). Indeed, p73DD-expressing cells showed reduced expression of procaspase-1 compared with cells expressing mp73DD (Fig. 3F).
Role of p73 in IFN-␥-induced Activation of the Caspase-1 Promoter-Initially, we sought to confirm IFN-␥-induced expression of caspase-1 in HeLa cells. Upon IFN-␥ treatment there was significant, time-dependent increase in the caspase-1 mRNA level as determined by RT-PCR analysis (Fig. 4A). A corresponding increase was observed in procaspase-1 protein (Fig. 4A). IFN-␥ treatment of HeLa cells increased caspase-1 promoter activity (Fig. 4B). The involvement of p73 in transactivation of the caspase-1 promoter upon IFN-␥ treatment was determined by coexpressing p73DD. IFN-␥-induced caspase-1 promoter activity was reduced by p73DD from 5.9-fold to 2.2-fold (p ϭ 0.0014), whereas p53DD had no significant effect (Fig. 4B). Further, we determined the activation of pC-WT and pC-MT-p53 promoter constructs upon IFN-␥ treatment. Both constructs were activated by IFN-␥ treatment but stimulation of the mutant construct was significantly lower (p ϭ 0.029) than that of the wild-type construct (Fig. 4C). This result showed that the p73 responsive site in the caspase-1 promoter was required for its full activation by IFN-␥.
To provide further evidence for the requirement of p73 in IFN-␥induced activation of the caspase-1 promoter, shRNA expressed from a mouse U6 promoter-based vector was used to down-regulate p73 protein level. Coexpression of this shRNA with p73 resulted in reduced expression of p73 protein (Fig. 5A). As a control, the C3G expression plasmid was cotransfected but shRNA had no inhibitory effect on C3G protein levels (Fig. 5A). Expression of this shRNA reduced p73-induced caspase-1 promoter activity by over 80% (Fig. 5B). Coexpression of this shRNA reduced basal caspase-1 promoter activity by 37%, whereas IFN-␥-induced activity was reduced by 44% (Fig. 5C). Endogenous p73 protein level was also reduced upon expression of p73-directed shRNA (Fig. 5D).
Effect of IFN-␥ on p73 Activity and Protein Levels-Our results suggested that p73 plays a role in IFN-␥-induced activation of the caspase-1 promoter. Therefore we examined whether p73 is activated by IFN-␥. For this purpose we used a CAT reporter plasmid with the SV40 promoter in which two p73 responsive sites had been cloned as an enhancer element downstream of SV40 promoter. This reporter was activated by p73␣ and p73␤ at low concentrations (4 ng) but not by IRF-1 (40 ng) (Fig. 6A). IFN-␥ treatment of HeLa cells increased the activity of this reporter, which was inhibited by p73DD but not by p53DD (Fig. 6B). These results suggest that p73 or p73-like activity is increased by IFN-␥ in HeLa cells. Whole cell lysates of HeLa cells treated with IFN-␥ were examined for changes in p73 protein levels by Western blotting. It was observed that, concomitant with an increase in IRF-1 and Stat-1 levels, p73␣ isoform levels also accumulated (Fig. 6C). The p73␤ isoform showed only a marginal increase compared with that observed upon cisplatin treatment. Treatment of cells with cisplatin increased p73 protein levels but there was no increase in Stat-1 or IRF-1 (Fig. 6C). The expression of p73-directed shRNA reduced the level of endogenous p73 protein and also inhibited increases in response to IFN-␥ treatment (Fig. 5D).
Role of p73 in IRF-1-induced Transactivation of the Caspase-1 Promoter-Previous studies on caspase-1 gene regulation have identified an IRF-1 binding site in the human caspase-1 promoter (33). IFN-␥ induced increases in expression of caspase-1 has been shown to be dependent on IRF-1, which is a transcriptional activator of caspase-1. As expected, IRF-1 expression significantly increased caspase-1 promoter activity (Fig. 7A). IRF-1-mediated increases in caspase-1 promoter after total RNA isolation caspase-1 mRNA levels were analyzed by semiquantitative RT-PCR analysis using primers for caspase-1 and GAPDH. An ethidium bromide-stained agarose gel of indicated PCR products is shown. Lower panel, immunoblot of total proteins isolated from IFN-␥-treated HeLa cells at indicated time points, using caspase-1 and Cdk-2 (loading control) antibodies. C indicates untreated control cells. B, pC-WT reporter plasmid (200 ng) was transfected along with p53DD or p73DD (100 ng) or control plasmid followed by treatment with IFN-␥ (100 ng/ml) for 24 h. CAT activities relative to untreated control are shown (n ϭ 4). C, pC-WT or pC-MT-p53 reporter plasmids were transfected in HeLa cells and treated with IFN-␥ (100 ng/ml) for 24 h. CAT activities relative to untreated control are shown (n ϭ 4). activity were reduced up to 50% by coexpression of p73DD (Fig. 7A). However, coexpression of p53DD had no significant effect on IRF-1mediated activation (Fig. 7A). We also compared the IRF-1-mediated transactivation of pC-WT and pC-MT-p53. Despite the presence of functional IRF-1 binding sites in both constructs, activation of pC-MT-p53 was up to 50% less than that of pC-WT upon IRF-1 overexpression (Fig. 7B). These results indicate that IRF-1-induced transactivation of the caspase-1 promoter required the p73 responsive site as well as functional p73 protein for optimal response.
p73 Can Activate the Caspase-1 Promoter Independent of IRF-1-The IRF-1 binding site in the human caspase-1 promoter overlaps with the initiator element, and mutation of this site without affecting the initiator element has been shown to abolish IRF-1-dependent caspase-1 promoter activity (33). We found that mutation of the IRF-1 binding site resulted in only a small reduction (about 30%) in caspase-1 promoter activation by p73␣ although activation by IRF-1 was abolished (Fig. 8, A  and B). Similar results were obtained with p73␤ (data not shown). These results showed that the IRF-1 binding site is not required for activation of the caspase-1 promoter by p73. The basal activity of the IRF-1 binding site mutant promoter was very low and did not show much activation by IFN-␥ (Fig. 8C). In contrast to p73, the p53-induced caspase-1 promoter activation was dramatically reduced upon mutation of the IRF-1 responsive site (Fig. 8D).
Cooperation between p73 and IRF-1 for Transactivation of the Caspase-1 Promoter-Results presented in the previous sections showed that compromising p73 function affected optimal activation of the caspase-1 promoter by IFN-␥. Therefore, we investigated the possibility of interdependence between p73 and IRF-1 in the activation of the caspase-1 promoter. For this purpose, suboptimal concentrations of p73␣ and IRF-1 were used. It was seen that p73␣ activated the caspase-1 promoter by 27-fold, whereas IRF-1 activated it by 13-fold. Combined expression of both these transcription factors resulted in a 53-fold activation of the caspase-1 promoter (Fig. 9A). Mutational inactivation of p73/p53 site in the caspase-1 promoter resulted in much lower activation by p73␣ as expected, but also showed a decrease in activation by IRF-1 (Fig. 9A). Introduction of p73␣ and IRF-1 together did not particularly enhance the mutant promoter activation (Fig. 9A). These results showed that the p73 responsive site was required for optimal activation of the caspase-1 promoter by p73 as well as by IRF-1. Other isoforms of p73 tested (␤, ␥, and ␦) also cooperated with IRF-1 in the activation of the caspase-1 promoter (Fig. 9B). The p73␣-mediated transactivation of the IRF-1 binding site mutant of the caspase-1 promoter-reporter construct was not increased by coexpression of IRF-1 (Fig. 9C). These results suggested that the IRF-1 binding site was also required for functional cooperation between p73 and IRF-1.
IFN-␥ Cooperates with Cisplatin and p73 for Activation of the Caspase-1 Promoter-Because exogenously expressed IRF-1 and p73 were able to cooperate for activation of the caspase-1 promoter, we examined the possibility of cooperation between IFN-␥ and p73. We observed that p73␣ and p73␤ activated the caspase-1 promoter by 12.1 Ϯ 1.8-and 15.9 Ϯ 0.8-fold, respectively; in the presence of IFN-␥ (which alone gave 5.2 Ϯ 1.1-fold activity) this activity increased to 21.9and 22.8-fold for p73␣ and p73␤, respectively (Fig. 10A). We next examined the effect of IFN-␥ and cisplatin on caspase-1 promoter activation. Treatment of cells with cisplatin (5 M) alone gave 2.9-fold caspase-1 promoter activity, which increased to 14.9-fold in the presence of IFN-␥ (Fig. 10B); in this experiment IFN-␥ alone gave a 7-fold caspase-1 promoter activity. Similar results were obtained when cells were treated with 10 M cisplatin in the presence or absence of IFN-␥ (Fig. 10B). We also determined the effect of IFN-␥ and cisplatin on caspase-1 gene expression by RT-PCR. A combination of IFN-␥ and cisplatin gave more increases in caspase-1 mRNA level than either of these alone (Fig. 10C). FIGURE 5. shRNA directed against p73 reduces IFN-␥-induced caspase-1 promoter activity. A, shRNA for p73 knocks down p73. HeLa cells were transfected with p73␣ and C3G expression plasmids, along with p73 shRNA (shRNA)-or a control shRNA (Con)-expressing vector in a ratio of 1:5. C3G expression plasmid was included in transfections as a control for the transfection efficiency and specificity. p73, C3G and Cdk2 were detected by Western blotting using specific antibodies. B, shRNA for p73 inhibits p73-induced caspase-1 promoter activity. HeLa cells were transfected with pC-WT reporter plasmid (100 ng) and p73␣ (5 ng) along with 200 ng of p73 shRNA (shRNA) or 200 ng of a control shRNA (Control). After 28 h of transfection, cell lysates were made for reporter assays. CAT activities relative to control without p73 are shown (n ϭ 3). C, effect of p73-directed shRNA on caspase-1 promoter activity induced by IFN-␥. HeLa cells were transfected with pC-WT reporter plasmid (100 ng) along with shRNA for p73 or control shRNA expressing plasmids (200 ng). After 6 h of transfections, cells were treated with IFN-␥ or left untreated for 24 h. CAT activities relative to untreated control are shown (n ϭ 3). D, effect of p73-directed shRNA on endogenous p73 protein level. HeLa cells in 6-well plate were transfected with p73 directed shRNA (shRNA)-or control shRNA (Con)-expressing plasmids. After 6 h of transfection, cells were treated with IFN-␥ or left untreated for 24 h. Western blotting was carried out for p73 and cdk2 (loading control).

DISCUSSION
We have shown that exogenous as well as endogenous p73 can regulate caspase-1 gene transcription. The activation of the caspase-1 promoter by p73␣ or p73␤ was more than that obtained with IRF-1 or wild-type p53. Even at low concentrations, p73␣ and p73␤ were very good activators of the caspase-1 promoter. The major p73 responsive site in the caspase-1 promoter was the same as the p53 binding site identified earlier. Significant activation of the p73/p53 site mutant promoter by p73 suggests that either an additional p73 responsive site is present or p73 may activate the caspase-1 promoter through some other transcription factor by protein-protein interaction. In addition to activating the caspase-1 promoter, p73 also contributes to the basal caspase-1 promoter activity, because blocking of p73 function by a dominant negative mutant showed a decrease in basal promoter activity. Mutation of the p73/p53 site, however, does not significantly affect basal promoter activity. Therefore, an additional p73 responsive site may be involved in regulating basal caspase-1 promoter activity. Alternatively, p73 may regulate basal promoter activity through some other transcription factor by protein-protein interaction.
Cisplatin-induced caspase-1 gene expression is likely to be caused by activation of the p73 protein because cisplatin-induced caspase-1 promoter activation was inhibited by blocking p73 function and also by mutating the p73 responsive site. In addition, the level of p73 protein was increased by cisplatin treatment of cells, whereas other known activators of caspase-1 gene expression, IRF-1, and Stat-1, were not induced  by cisplatin (Fig. 9C). Cisplatin has been shown to induce cell death by apoptosis, which is dependent on p73 (43). We found that cisplatininduced apoptosis of HeLa cells was inhibited by 60% when caspase-1 function was blocked by expression of the mutant procaspase-1 (figure not shown). Thus our results are consistent with the suggestion that p73-dependent transcription of the caspase-1 gene contributes in part to cisplatin-induced apoptosis of HeLa cells.
Results presented here suggest that p73 plays an important role in the activation of the human caspase-1 promoter by IFN-␥. The following observations provide support for this suggestion. First, mutation of the p73/p53 responsive site reduced IFN-␥-induced caspase-1 promoter activity. Second, blocking of p73 function resulted in reduced IFN-␥-induced caspase-1 promoter activity. Third, IFN-␥ treatment resulted in an increase in p73 protein levels. In addition, IRF-1-induced activation of the caspase-1 promoter was reduced by mutation of the p73 responsive site and also by blocking p73 function. The cooperative effect of p73 and IRF-1 on caspase-1 promoter activity provides support to the suggestion that IFN-␥-induced caspase-1 promoter activation involves interdependent action of these two transcription factors. Whereas IRF-1 is absolutely essential for activating the caspase-1 promoter in response to IFN-␥, the transcription factor p73 also plays a significant role for optimal activation of the caspase-1 promoter by this cytokine.
The level of p73 protein in the cells in generally very low. But even at very low concentrations, p73␣ and p73␤ are strong activators of thecaspase-1 promoter. Therefore small changes in p73 protein level or activity are likely to affect caspase-1 promoter activity. This is evident with respect to IFN-␥, which causes a small increase in p73 protein level but this contributes to 50 -60% of caspase-1 promoter activity observed in IFN-␥-treated cells (Fig. 4).
The IRF-1 binding site in the human caspase-1 promoter overlaps with the initiator element (33). Mutation of this IRF-1 binding site without affecting the initiator element results in nearly complete loss of basal promoter activity as well as IFN-␥-induced promoter activity. Although the role of IRF-1 in regulating caspase-1 gene expression is well known, this is the first report that identifies the IRF-1 responsive site, which is responsible for promoter activation by IFN-␥. Mutation of this IRF-1 responsive site drastically reduced p53-induced activation of this promoter but surprisingly, the activation by p73 was only marginally reduced. Therefore we conclude that p73 and p53 activate the caspase-1 promoter by somewhat different mechanisms. In addition to p73 and IRF-1, there must be other proteins involved in  regulating caspase-1 promoter activity. No Stat-1 responsive sites have been identified in this promoter, which may indicate an indirect role of this transcription factor. We have observed that in Sp1-deficient Drosophila SL2 cells, the caspase-1 promoter is activated by Sp1. We have identified one of the Sp1 responsive sites in the human caspase-1 promoter (data not shown). Very recently it has been shown that Sp1 interacts with p73␣ and p73␤ (44,45). Therefore, it is possible that small activation (2-3-fold) of the p73 site mutant of the caspase-1 promoter by p73 may be mediated by interaction of p73 with Sp1. The p53-specific inhibitor (p53DD) showed an increase in CAT activity, which was statistically significant (Figs. 2B and 6B).This observation indicates that low level of p53 present in HeLa cells may have inhibitory effects on p73 activity either directly or indirectly. This hypothesis was tested experimentally. Activation of the caspase-1 promoter by p73␣ was reduced from 23-fold to 14-fold upon coexpression of p53 (which alone gave 6-fold activation). This inhibitory effect of p53 on p73 activity is not surprising because p53 and p73 are likely to compete for the same site on the caspase-1 promoter (Fig. 2B) and p73 reporter (Fig. 6B).
We have studied the requirement of p73 for IFN-␥-induced caspase-1 activation in HeLa cells. Because p53 family members activate promoters by binding to similar sequences, the question of other members being involved in IFN-␥-mediated responses arises. The possibility of p53 activation of caspase-1 under these conditions was ruled out by the use of a p53-specific inhibitor, which does not inhibit p73 activity. The remaining possibility of p63 being involved in IFN-␥ stimulation of caspase-1 in addition to p73 cannot at this point be ruled out. The fact that p73-directed shRNA reduces IFN-␥-induced caspase-1 promoter activity and that p73 cooperates with IRF-1 for maximal induction of the caspase-1 promoter activity suggest that p73 is an important component of caspase-1 regulation during IFN-␥ signaling.
Overall our results show that p73 is a strong activator of human caspase-1 gene transcription. It also contributes to basal caspase-1 promoter activity. In addition our results suggest that IFN-␥-induced caspase-1 promoter activation involves an interdependent action of p73 and IRF-1.