Human T-cell leukemia virus type I tax repression of p73beta is mediated through competition for the C/H1 domain of CBP.

The Tax protein, encoded by the human T-cell leukemia virus type I (HTLV-I), is required for high level viral transcription and HTLV-I-associated malignant transformation. Although the precise mechanism of malignant transformation by Tax is unclear, it is well established that Tax represses the transcription function of the tumor suppressor p53, possibly accelerating the accumulation of genetic mutations that are critical in HTLV-I-mediated malignant transformation. Tax repression of p53 transcription function appears to occur, at least in part, through competition for the cellular coactivator CBP/p300. In this study, we characterize the effect of Tax on the p53 family member, p73. We demonstrate that Tax also represses the transcription function of p73beta and that the repression is reciprocal in vivo, consistent with the idea that both transcription factors may compete for CBP/p300 in vivo. We provide evidence showing that both Tax and p73 interact strongly with the C/H1 domain of CBP and that their binding to this region is mutually exclusive in vitro. This finding provides evidence supporting the idea that reciprocal transcriptional repression between Tax and p73 is mediated through coactivator competition.

Human T-cell leukemia virus type I (HTLV-I) 1 is the etiological agent of adult T-cell leukemia (ATL) which is an aggressive and fatal hematological malignancy (1,2). Only a small percentage of people infected with HTLV-I develop ATL, generally 20 -40 years following infection (3). The infrequency of ATL, coupled with the long latency period, suggests that ATL occurs as a consequence of multiple genetic mutations that accumulate during the prolonged period of asymptotic HTLV-I infection. Although the molecular basis of HTLV-I pathogenesis is not well understood, there is strong emerging evidence that expression of the viral trans-activator protein Tax plays an essential role in the oncogenic process. Indeed, the expression of Tax is able to immortalize primary T-cells (4), transform rat fibroblasts in vitro (5), and promote tumorigenesis and leukemogenesis in a mouse model (6).
With the goal of understanding the role of Tax in leukemogenesis, several studies have examined how Tax affects proteins involved in regulation of the cell cycle (for review see Ref. 7). These proteins include the cyclin-dependent kinase inhibi-tors p16 INKA4 , p18 INK4A , p21 Waf1/CIP1 , and p27 Kip1 (8 -13), cyclin D (8,14,15), the transcription factors E2F-1 and E2F-2 (16 -19), and the tumor suppressor protein p53 (12, 20 -30). p53 is a transcription factor that is activated in response to genotoxic stress. Once activated, p53 induces expression of genes critical for cell cycle arrest or apoptosis, thus preventing the transmission of genetic mutations to progeny cells (31). Loss of p53 activity has been found in 60% of human malignancies examined (32,33), consistent with a role for p53 in genome surveillance and suppression of malignant transformation.
Most interesting, in HTLV-I-infected and Tax-expressing cells, p53 is present at elevated levels, with a relatively low frequency of mutation (ϳ25%) (12,20,21,30,34,35). Paradoxically, several studies have demonstrated that although generally wild type, p53 is functionally inactive. For example, HTLV-I-infected cells do not respond appropriately to a variety of p53 stimuli, including gamma and ionizing irradiation (12,22,36,37). Furthermore, Tax expression alone abrogates p53-induced G 1 arrest and apoptosis following DNA damage (24) and inhibits the activation of a panel of known p53-responsive genes (22). Several recent studies provide evidence showing that Tax inhibition of the tumor suppressor activities of p53 is directly due to Tax inhibition of p53 transcription function (21,22,24).
It is well established that Tax does not directly bind p53 (12,22,27,36,38); thus, Tax appears to compromise p53 function via an indirect mechanism. Several reports have indicated that Tax repression of p53 may occur through alterations in the phosphorylation state of p53 (23,39). Alternatively, other studies indicate that the repression occurs as a consequence of competition between Tax and p53 for the cellular coactivators CBP/p300 (25)(26)(27). Both Tax and p53 utilize the cellular coactivators CBP/p300 as mediators of transcriptional activation (40 -44). The site of competition on CBP/p300 appears to be the KIX domain (amino acids 588 -683), as both Tax and p53 have previously been shown to bind to this region, and their binding is mutually exclusive (25,27). This evidence suggests that Tax may attenuate p53 function by abrogating the p53-KIX interaction, thus reducing the capacity of p53 to recruit CBP/p300 to target promoters.
Recently, a new member of the p53 gene family has been identified, known as p73 (45). p73 is also a transcriptional activator protein and carries an amino-terminal activation domain with moderate homology (29%) to p53 (45,46). At least 6 isoforms of p73 (␣, ␤, ␥, ␦, ⑀, and ) may be generated through alternative splicing of the primary p73 transcript (47)(48)(49). The p73␣ and p73␤ proteins are ubiquitously expressed at low levels in many tissues. An exception is T-lymphocytes, which lack p73␤, but express p73␣ (48). Interestingly, one form of p73, which lacks the transactivation domain, has been shown to be pro-apoptotic in neurons (50,51). Specific p53 mutant proteins have been shown to associate with p73 (␣, ␤, ␥, and ␦), resulting in interference with p73 transcriptional activity and the ability to induce apoptosis (52)(53)(54). As predicted from local regions of strong homology in the DNA binding domains of both proteins, p73 binds the canonical p53-binding sites (55). Furthermore, when ectopically expressed, p73 can trans-activate a variety of p53 promoters (p21, bax, mdm2, and GADD45) and induce apoptosis in p53-deficient cell lines (45, 48, 52, 56 -59). Similar to p53, the activation domain of p73 interacts with the cellular coactivators CBP/p300 (60); however, the site of interaction appears to differ. Whereas p53 interacts with a carboxyl-terminal region and the KIX domain (25,43), p73 has recently been shown to bind the C/H1 domain of CBP, located near the amino terminus of the coactivator (60).
p73 has been classified as a possible tumor suppressor based on substantial sequence homology and functional similarity with p53. This classification was strengthened by the observation that the p73 gene maps to chromosome 1p36, a region frequently deleted in several malignancies including neuroblastomas, colorectal cancers, and breast cancers (45). p73 activity is induced by exposure of cells to DNA-damaging agents such cisplatin, taxol, and ␥-irradiation (61)(62)(63)(64). Furthermore, in lymphoid malignancies there is evidence that hypermethylation, and thus silencing, of the p73 gene may play a role in the development and/or progression of the neoplasm (65,66). Loss of a single p73 allele has been linked to ␥-radiation-induced T-cell lymphomas (67). A recent study suggests that deregulated E2F-1 induction of p73 might promote a p53-independent, tumor suppressor pathway and that perhaps many of the previously attributed "p53-independent" anti-tumorigenic activities may be due directly to p73 (68). Despite these observations, other studies have failed to show mutations in the p73 gene in a wide variety of tumor types. Normal expression of p73 is observed in many malignant and non-malignant cells, and some cancers actually shown increased levels of p73. In contrast to p53-deficient mice, p73 Ϫ/Ϫ mice show no increased susceptibility to spontaneous malignancies. These conflicting results suggest that altered expression of the p73 gene, rather than a loss of p73 function, may play a role in malignant transformation.
In this report, we investigate the function of p73 in HTLV-I-transformed, Tax-expressing T-cell lines. We demonstrate that in the presence of Tax the stability of p73␤ is increased, yet, paradoxically, the transcription function of p73␤ is decreased. We also show that in the presence of p73 the transcription function of Tax is reduced, indicating that the transcriptional repression is reciprocal. The reciprocal transcriptional repression appears to arise from intracellular competition for CBP/p300, an observation very similar to that reported for Tax repression of p53 (25)(26)(27). Although these effects of Tax on p73 function are very analogous to the effects of Tax on p53 function, we provide evidence indicating that the site of competition on the surface of CBP/p300 differs. We demonstrate biochemically that both Tax and p73␤ bind strongly to the C/H1 domain of CBP and that the binding of the two transcription factors is mutually exclusive in vitro. These data suggest that Tax may inactivate p73␤ transcription function by specifically competing for the p73-C/H1 interaction in vivo. Disruption of both the p53 and p73␤ interactions with CBP/p300 in HTLV-I-infected cells may alter the transcription function of these key regulatory proteins. 10% fetal bovine serum, 2 mM L-glutamine, and penicillin/streptomycin. For transient cotransfection assays (69), cells were grown to a density of 10 6 cells/ml and transfected with LipofectAMINE (Life Technologies, Inc.) and a constant amount of DNA. After 24 h, the cells were harvested and lysed, and luciferase activity was measured using the Dual-Luciferase reporter assay system (Promega) with a Turner Designs model TD 20-e luminometer. Luciferase activity was normalized to Renilla luciferase from herpes simplex virus thymidine kinase promoter (pRL-TK, Promega). Expression plasmids for p73␤ (pcDNA3.p73␤; Ref.  (25). Briefly, the DNA was transformed in BL21(DE3) pLysS, expanded, induced, and purified by glutathione-agarose affinity chromatography. Tax protein was expressed from the pTaxHis6 (75) and pTet-K88A (73) expression plasmids and purified as described previously (40). Briefly, the DNA was transformed in BL21(DE3) pLysS, expanded, induced, and purified to greater than 90% homogeneity by Ni 2ϩ -nitrilotriacetic acid-agarose chromatography (Qiagen). Purified Tax protein was dialyzed against TM 0.1 M KCl buffer containing 50 mM Tris (pH 7.9), 12.5 mM MgCl 2 , 100 mM KCl, 1 mM EDTA (pH 8), 1 mM dithiothreitol, 0.025% (v/v) Tween 20, and 20% (v/v) glycerol, aliquoted, and then stored at Ϫ70°C. In vitro translation of p73␤ protein was performed using the TNT translation system (Promega) according to the manufacturer's protocol.
Half-life Determination-Cycloheximide chase experiments were performed as described by Maki and Howley (76). Briefly, T-cells were seeded at 2 ϫ 10 6 cells/ml and adherent cells at confluence in 100-mm dishes in their respective culture medium. Cycloheximide (Sigma) was dissolved in absolute methanol and added directly to the culture media to a final concentration of 20 g/ml. Cells were harvested at different time points after cycloheximide treatment, washed with phosphatebuffered saline, and then whole cell extracts were made by using RIPA buffer. Lysates of selected cells were normalized for protein content by the Bradford assay, and 100 g of proteins were Western-blotted with anti-p73␤ antibody.
GST Pull-down Assays-All GST pull-down experiments were performed using 20 l of glutathione-agarose beads equilibrated in pulldown buffer (20 mM Hepes, pH 7.9, 0.5 mM EDTA, pH 8, 10% glycerol, 0.05% Nonidet P-40, 5 M ZnSO 4 , 2.5 mM MgCl 2 , 25 mM KCl, 1 mM dithiothreitol). The purified GST fusion protein was incubated with the beads for 1-2 h at 4°C and then washed twice with pull-down buffer. The second protein(s) was then added to the washed beads and incubated overnight at 4°C. The beads were washed four times, and bound proteins were eluted with SDS sample dyes and resolved by electrophoresis on a 10% SDS-polyacrylamide gel. To detect the binding of radiolabeled p73␤, the gel was dried and analyzed by PhosphorImager analysis. To detect the binding of Tax, proteins were transferred to nitrocellulose and probed with an anti-His 6 antibody (H-15; Santa Cruz Biotechnology).

RESULTS
p73␤ Is Overexpressed in HTLV-I-transformed T-cell Lines-It is well established that the HTLV-I Tax protein pro-p73 and Tax Compete for C/H1 motes stabilization of endogenous p53. To determine whether Tax might similarly stabilize p73, we examined the steadystate levels of p73 in whole cell extracts from the following Tax-expressing, HTLV-I-transformed cell lines: C8166/45, C91/ PL, MT2, SLB-1, and HUT-102 ( Fig. 1, lanes 1-5). For comparison, we examined p73 expression levels in the following uninfected human leukemic T-cell lines: CEM, HUT-78, Jurkat, and Molt-4 ( Fig. 1, lanes 7-10). Kidney 293 cells, which express high levels of the p73␤ isoform (55), were used as a positive control (Fig. 1, lane 6). Each of the HTLV-I-transformed cell lines, with the exception of HUT-102, had detectable levels of p73␤ ( Fig. 1, lanes 1-5). Interestingly, we were able to detect p73␤ in HUT-102 nuclear extracts (data not shown). In contrast, we did not detect p73␤ expression in any of the uninfected T-cell lines (Fig. 1, lanes 7-10). For comparison, we also measured p53 expression levels and found that, as previously reported, p53 was detectable in all of the HTLV-I-transformed cell lines (12,20,22,24,30,35,37) as well as uninfected CEM and MOLT-4 (77) (Fig. 1, lanes 11-20). We also examined Tax expression levels in the HTLV-I-infected cell lines (Fig. 1, lanes [21][22][23][24][25]. Most interesting, we observed a direct correlation between higher p73␤ levels and the presence of the native 40-kDa form of Tax (Fig. 1, lanes 21-25). The larger forms of Tax detected in MT-2 and HUT-102 cells appear to be a fusion protein that carries both envelope and Tax sequences (78). The correlation between expression of this Tax fusion protein and reduced p73 levels suggest that the fusion protein is defective for p73 stabilization. Together, these observations suggest that Tax may play a role in p73␤ stabilization. We have also examined whether the nuclear localization of p73 is affected in HTLV-I cell lines. In all cell lines examined, we detected generally high levels of p73␤ in the nucleus, suggesting that the localization of p73␤ is unaffected in HTLV-I transformed Tcells (data not shown).
p73␤ Is Stabilized by Tax-Previous studies have shown that the half-life of p53 is significantly increased in HTLV-I-trans-formed T-cells (20) and HTLV-I-immortalized T-cells (21). To determine if an increase in the p73 half-life might account for the elevated protein levels observed in HTLV-I-infected cell lines, we measured p73 stability in SLB-1 and C8166/95 cells, the two HTLV-I-transformed lines that expressed a high level of p73␤ protein (see Fig. 1). Cells were treated with cycloheximide for various lengths of time, and p73␤ levels were measured by Western blot analysis. As a control, we determined the half-life of p73␤ in uninfected HCT-116 cells, as they have previously been used in p73␤ half-life studies (62). Representative Western blots of p73␤ levels following cycloheximide treatment in SLB-1 and C8166/45 are shown in Fig. 2A. Graphical representation of the data is presented in Fig. 2B. In the control HCT-116 cells, we found that, like p73␣ (62), the p73␤ half-life in these cells was 45 min (Fig. 2B, left panel). Most interesting, the p73␤ half-life in SLB-1 and C8166/45 was measured at 210 and 165 min, respectively (Fig. 2B, center and  right panels). These data correspond to a p73␤ half-life increase of ϳ4-fold in these HTLV-I-infected T-cells.
The viral protein Tax has previously been shown to increase directly p53 protein stabilization (21,22), although the mechanism for this stabilization is unknown. To determine if Tax is directly involved in p73␤ stabilization, we cotransfected expression plasmids for p73␤ and Tax into human Jurkat Tlymphocytes, and we measured Tax and p73 protein levels by Western blot analysis. We selected the HTLV-I-negative Jurkat cells, as they are negative for both p73␤ and p53 (see Fig.  1, lane 9 and 19, respectively) (79,80). Fig. 3A shows that expression of Tax produced a significant increase in p73␤ protein levels (lanes 2 and 3). In the absence of Tax, a small amount of p73␤ was detected by Western blot, consistent with the short half-life of the protein (Fig. 3A, lane 1). Both 1 and 2 g of the transfected Tax expression plasmid produced a comparable increase in p73 levels, suggesting that the lower amount of Tax is saturating for p73␤ stabilization. As a control, Tax stabilization of p53 is also shown (Fig. 3B).  21-25). The high molecular weight forms of Tax, produced by fusion of Tax with Env-coding sequences (78), are indicated by an asterisk. p73 and Tax Compete for C/H1 Tax Inhibits the Transcription Function of p73␤-Several groups have previously shown that the viral oncoprotein Tax, in addition to stabilizing p53, represses p53 transcription function (21,22,24,25). Based on these observations, we were interested in determining whether Tax similarly repressed the transcription function of p73␤. To examine this possibility, we utilized the p53-responsive reported plasmid, pG13-Luc for these studies, as overexpression of p73␤ has been shown to activate many p53-responsive genes (45, 48, 52, 56 -59). (Genes that are specifically responsive to p73 have not yet been identified.) We transfected the pG13-Luc reporter plasmid, which carries 13 copies of the consensus p53-response element driving expression of the luciferase gene, into HTLV-I-negative Jurkat T-cells (Fig. 4A). Consistent with the observation that Jurkat T-cells do not express p73 or p53 (Fig. 1, lanes 9 and 19), we did not detect significant luciferase activity produced from the pG13-Luc reporter plasmid in these cells (Fig. 4A, lane 1). As expected, cotransfection of the p73␤ expression plasmid strongly activated transcription from p53-responsive reporter plasmid (Fig. 4A, compare lanes 1 and 2). Interestingly, cotransfection of the Tax expression plasmid produced a dose-dependent repression of p73␤-mediated transcriptional activation, strongly supporting a direct role for Tax in repression of p73␤ transcription function (Fig. 4A, lanes 3-5).
The Tax protein utilizes the coactivator CBP to mediate transcriptional activation of the HTLV-I promoter (40,41). Recently, two groups have reported that p73␤ also utilizes CBP to mediate transcriptional activation (59, 60). Since there is significant emerging evidence for coactivator competition as a mechanism of transcriptional repression (25)(26)(27), we were interested in testing whether CBP competition may account for the observed Tax repression of p73 transcription function. To test this hypothesis, we utilized a point mutant of Tax (K88A) that has previously been shown to be defective for interaction with the KIX domain of CBP (73) and Tax transactivation in vivo (25,73). If Tax repression of p73␤ is occurring through competition specifically for the KIX domain of CBP, then Tax K88A should be unable to repress p73 function. Unexpectedly, cotransfection of mutant Tax K88A strongly repressed the transactivation function of p73␤ to a level greater than that observed with wild type Tax (Fig. 4A, compare lanes 2-8). We then tested whether K88A was defective for Tax transactivation, as previously described (73), by examining K88A function on a reporter plasmid carrying three copies of the Tax-responsive viral CREs (viral CRE-Luc) driving expression of the luciferase gene (40). Transfection of the wild type Tax expression plasmid strongly activated transcription from the Tax-responsive promoter, whereas Tax K88A was defective for activation (Fig. 4B, lanes 2 and 3). Western blot analysis of nuclear extracts showed that both the wild type and mutant Tax proteins were expressed in the transfection assay (Fig. 4C, lanes  1-3). These data suggest that, if Tax repression of p73␤ is mediated through coactivator competition, the site of competition resides outside of the KIX domain of CBP.
As an alternate means to examine whether competition for a common coactivator may participate in Tax repression of p73, we next examined whether overexpression of p73␤ similarly repressed Tax function. We reasoned that if repression of p73␤

p73 and Tax Compete for C/H1
by Tax occurs as a consequence of competition for CBP, then overexpression of p73␤ should similarly repress Tax function. To test this possibility, we performed the reciprocal experiment using the Tax-responsive viral CRE-Luc reporter plasmid. As expected, cotransfection of the Tax expression plasmid strongly activated transcription from the Tax-responsive reporter plasmid (Fig. 4D, lanes 1 and 2). Consistent with the theory of coactivator competition, cotransfection of increasing amounts of the expression plasmid for p73␤ repressed Tax transactivation in a dose-dependent fashion (Fig. 4D, lanes 3-5). Together, these data support a mechanism of reciprocal repression where Tax and p73 bind to a common coactivator, ultimately recruiting the coactivator to their respective target genes.
p73␤ and Tax Both Bind to the C/H1 Domain of CBP-The transient transfection data suggesting that Tax and p73␤ were competing for CBP binding led us to determine whether there might be a common site on CBP where both transcription factors interact. We considered the amino-terminal C/H1 domain of CBP, as Zeng et al. (60) reported, that p73␤ binds to this domain of CBP. To confirm this observation, we analyzed the binding of p73␤ to several regions of CBP using the GST pull-down assay. Purified GST fusion proteins carrying various regions of CBP (see Fig. 5A) were bound to glutathione-agarose beads and then incubated with labeled and in vitro translated p73␤ protein. Fig. 5B shows that, as expected, p73␤ bound well to the two GST constructs that carried the C/H1 domain of CBP (lanes 3 and 4). We did not detect significant binding of p73␤ to GST alone, KIX (aa 588 -683), or to three carboxyl-terminal regions of CBP, comprising amino acids 1514 -1894, 1894 -2212, and 2212-2441 (Fig. 5B, lanes 2 and 5-8). These results are consistent with the report by Zeng et al. (60) who showed that the interaction of the amino-terminal activation domain of p73␤ with C/H1 modulates p73 transcription function.
To determine whether Tax might also bind to C/H1, we tested purified recombinant Tax protein in a GST pull-down assay with the C/H1 domain. As shown in Fig. 5C, both wild type Tax and KIX binding defective K88A Tax bound well to the C/H1 domain (lanes 5 and 6). Together, these observations indicate that both Tax and p73␤ bind to the C/H1 domain of CBP, possibly to recruit CBP to their target promoters. This common interaction site on CBP provides support for the hypothesis that Tax and p73␤ may compete for CBP in vivo, accounting for the observed reciprocal transcriptional repression.
p73␤ and Tax Binding to C/H1 Is Mutually Exclusive-To test directly whether the binding of Tax and p73␤ to C/H1 is mutually exclusive, we performed a GST pull-down competition assay. Glutathione-agarose beads bound with GST-C/H1  lanes 2 and 3). After 24 h, cells were analyzed by Western blot with anti-p73␤ and anti-Tax antibodies, as indicated. B, as a control, the same experiment was performed with an expression vector for p53 (500 ng) (71) (lanes 1-3) and Tax (1 and 2 g) (lanes 2 and 3) and probed with anti-p53 and anti-Tax antibodies, as indicated.

p73 and Tax Compete for C/H1
(CBP aa 302-451) were incubated with a constant amount of in vitro transcribed/translated p73␤ and increasing amounts of purified Tax (Fig. 6). If the two proteins bind to C/H1 in a mutually exclusive manner, we reasoned that increasing amounts of Tax, relative to p73␤, would displace p73␤ from C/H1. Fig. 6 shows that the coincubation of increasing amounts of purified Tax protein in the binding reaction reduced p73 binding to C/H1 (compare lanes 3-6, upper panel). The competition was dose-dependent and corresponded directly with a concomitant increase in Tax binding (Fig. 6A, lanes 3-6, lower  panel). We next tested whether the Tax point mutant K88A, which bound C/H1 similar to wild type Tax, could also compete with p73␤ for CH/1 binding. As expected, increasing amounts of the Tax point mutant also dramatically reduced p73␤ binding to C/H1 (Fig. 6A, lanes 7 and 8). In the reciprocal competition experiment, GST-C/H1 was incubated with a constant amount of purified Tax protein and increasing amounts of in vitro transcribed/translated p73␤. Similar to the results obtained above, increasing amounts of p73␤ reduced the binding of Tax, with a concomitant increase in p73␤ binding to the C/H1 domain (Fig. 6B, lanes 3-6). Together, these data indicate that the binding of p73␤ and Tax to C/H1 is mutually exclusive in vitro, providing a possible mechanism for the observed repression of p73 by Tax. DISCUSSION In this report, we characterize the effect of the HTLV-Iencoded oncoprotein Tax on the functionality of the p53 family member p73. These studies were undertaken as several previous reports have shown that Tax has dramatic effects on both the stability and transcription function of p53. We show that, like p53, Tax represses the transcription function of p73 while paradoxically enhancing the stability of the protein. We dem-onstrate that the transcriptional repression is reciprocal, as p73 represses the transcription function of Tax. The molecular basis of the reciprocal repression appears to be competition for the cellular coactivators CBP/p300, as both p73␤ and Tax bind to the amino-terminal C/H1 domain of CBP, and their binding is mutually exclusive in vitro.
We also show that p73 is overexpressed in HTLV-1-transformed T-cell lines, due to an apparent increase in the half-life of the p73 protein. This stability is directly due to the viral oncoprotein Tax. This observation is supported by a prolonged p73 half-life in the HTLV-I-transformed cell line C8166/45, which expresses the viral protein Tax but not Rex (81). Furthermore, transient transfection studies show that cotransfection of Tax promotes elevated p73 protein levels, relative to that observed in the absence of Tax. The mechanism of Tax stabilization of p73 remains unknown. In previous studies examining the p53 degradation pathway, it has been suggested that the interaction of p53 with the C/H1 domain CBP/p300 promotes degradation (82). It is plausible that p53 and p73 competition with Tax for CBP binding could potentially repress the degradation pathway, resulting in an increase in the halflife of both proteins. The fact that CBP is limiting in cells provides further support for a mechanism that involves competition (83). Alternatively, Tax may contribute to higher levels of p73 protein levels through transcriptional activation of the p73 gene. Although this would not account for the observed increase in the p73 half-life, precedence for this idea comes from the observation that the transcription factor E2F-1 has  4) and GST-C/H1-(aa 302-451) (lanes 5 and 6). Bound Tax and proteins standards are indicated. Wild type and mutant Tax onput (20% each) is shown (lanes 1 and 2). p73 and Tax Compete for C/H1 been shown to stimulate p73 gene expression (68,84,85). Since Tax increases the level and activity of E2F-1 in HTLV-I-transformed T-cells (16,18), enhanced E2F levels may directly increase transcription of the p73 gene.
Although p73 is present at elevated levels in Tax-expressing HTLV-I-infected T-cells, it is functionally compromised for transcription activation. Several possibilities might account for this loss of p73 transcription function in the presence of Tax. For example, some viruses, such as adenovirus or hepatitis B, encode proteins that bind and sequester p53 in the cytoplasm (86,87). This is not the case for Tax repression of p73, as we observe generally high levels of p73 in nuclear extracts (data not shown). This is consistent with a recent study indicating that Tax and p73␤ do not directly interact (28). We have also determined that Tax does not compete with p73 binding to the DNA (data not shown). The observation that Tax and p73 reciprocally repress transcription function, in the absence of a direct interaction, strongly suggests competition for limiting CBP/p300. There is precedence for this scenario with Tax and p53, as we and others (25)(26)(27)38) have previously shown that the reciprocal transcriptional repression between these two factors occurs, at least in part, through competition for CBP. In support of this hypothesis, we show in this report that both Tax and p73 bind the C/H1 domain of CBP in vitro, and that this binding is mutually exclusive. Furthermore, we show that a Tax mutant, K88A, which is defective for KIX binding (73) but not C/H1 binding, can also repress p73 activation. This observation strongly supports the idea that Tax and p73␤ compete for CBP utilization in vivo, specifically through the C/H1 domain of the coactivator. Our observation that Tax K88A represses p73␤ transcription function is in contrast to a recent report by Kaida et al. (28) who showed that this Tax mutant failed to repress the p73␤ activity. The nature of this discrepancy is not known, however Kaida et al. (28) used Saos-2 cells, whereas we used mature CD4 ϩ Jurkat T-cells. This difference in cell type may account for the observed discrepancy.
Currently, Tax is known to bind the KIX domain and the carboxyl-terminal SRC-interacting domain of CBP, with both interactions apparently contributing to Tax transactivation of HTLV-I transcription (40,41). 2 Interestingly, the data presented herein indicate that Tax additionally interacts with the C/H1 domain of CBP, perhaps further contributing to Tax transcription function. The C/H1 domain, named for its cysteine/histidine-rich domain, is composed of two zinc finger modules (88) that have recently been shown to be involved in interactions with several proteins, including two viral proteins (89,90). C/H1 is located immediately amino-terminal to the KIX domain, and this contiguous C/H1-KIX region (aa 302-683) may form a strong binding platform for Tax. This idea is supported by our observation that C/H1-KIX bound with high affinity (relative to either domain alone) to the ternary complex containing Tax, CREB, and the viral CRE DNA. 3 These data suggest that multiple, distinct Tax-CBP interactions occur simultaneously and perhaps cooperate to enhance coactivatormediated transcriptional activation.
Several recent studies (45, 48, 52, 56 -59) have shown that p73 promotes apoptosis in p53-deficient cells. Our observation that Tax inhibits p73 transcription functions would suggest that Tax also represses p73-mediated apoptosis. Thus, the inactivation of both p73 and p53 by Tax may together contribute to HTLV-I-dependent leukemogenesis. Although Tax levels are generally low in HTLV-I-infected cells (91), intermittent high levels of Tax may saturate multiple sites on CBP/p300, derail-ing both p53 and p73 transcription function. This event would likely promote an environment in the infected T-cell that is tolerant of mutations and chromosomal instability and resistant to apoptosis. This environment would promote the survival of genetically compromised cells poised for further genetic damage and malignant transformation.