The promyelocytic leukaemia protein tumour suppressor functions as a transcriptional regulator of p63

Article metrics


p63 plays unique developmental roles in epidermal morphogenesis, despite its structural similarity with p53. The p63 gene has two distinct promoters, coding for proteins containing an N-terminal transactivation domain (TA isoforms) and for proteins lacking this region (ΔN isoforms). The full-length transcriptionally active TAp63 isoforms are capable of transactivating the majority of the p53 target promoters thus inducing cell cycle arrest and apoptosis. On the contrary, the ΔNp63 isoforms seem to counteract the transactivation activities of p53 and TAp63 proteins, thus possibly conferring a proliferative advantage to cancer cells. However, the molecular mechanisms controlling the transcriptional activity of p63 remain largely unclear. Here we present data indicating that (i) the promyelocytic leukaemia protein (PML) physically interacts with p63, (ii) p63 is localized into the PML nuclear-bodies (PML-NBs) in vivo, and (iii) PML regulates p63 transcriptional activity. We show that the interaction of p63 with PML increases the levels of p63 in cultured cells as well as its ability to transactivate the p53-responsive elements of the GADD45, p21 and bax promoters. These data are consistent with a general role for PML as a functional modulator of all the p53 family members. Our findings strengthen the relevance of the cross talk between PML and the p53 family members, imply a new tumour suppressive function of PML and unveil a possible role for PML in epidermal morphogenesis and differentiation.


p63 is a structural and functional homologue of p53, sharing with p53 a modular structure as well as extensive sequence homology (Melino et al., 2003), particularly in the central DNA-binding domain (DBD). Unlike p53, the p63 gene gives rise to several isoforms via alternative splicing at the C-terminus (α, β, and γ forms). In addition, N-terminally deleted isoforms (ΔN) are generated through the use of an alternate promoter located in intron 3. The full-length N-terminal transcriptionally active (TA) proteins can bind and transactivate p53 target promoters and induce cell cycle arrest and apoptosis in cultured cells. As transactivation activity resides in the protein's N-terminus, the ΔNp63 isoforms exhibit a dominant-negative effect via competition for binding to p53 target sequences (Yang et al., 1998), similarly to what occurs for p73 (Melino et al., 2002). Recent evidence suggests that ΔNp63 displays trascriptional activity, which is independent from the presence of the transactivating domain (Duijf et al., 2002; Wu et al., 2003 and our unpublished observations), although ΔNp63 transcriptional targets as well as its biological functions remain largely unknown.

ΔNp63 is the predominant p63 isoform expressed in a highly tissue-specific manner in the embryonic ectoderm and in the basal regenerative compartment within epithelial tissues, including skin, where its function is associated with the development and maintenance of the epidermis (Parsa et al., 1999). The major evidence linking p63 to the development of the epidermis is the p63 knockout mouse phenotype. The absence of p63 results in severe limb truncations, craniofacial malformations, and most importantly in absence of skin and most epithelial tissues as well as defective epidermal differentiation (Mills et al., 1999; Yang et al., 1999). Keratinocyte differentiation correlates with a downregulation of ΔNp63 levels, which in turn relieves the inhibitory effect on the expression of genes such as p21 and 14-3-3 σ (Westfall et al., 2003). To date, very little is known about the molecular mechanisms that govern p63 activity and stability in vivo.

The promyelocytic leukaemia (PML) tumour suppressor gene is fused to the retinoic acid receptor α (RARα) gene in the vast majority of acute promyelocytic leukaemias (APLs), resulting in the generation of a PML-RARα fusion oncoprotein, which acts as double dominant negative on PML and RARα functions (de The et al., 1991; Goddard et al., 1991; Pandolfi et al., 1991; Piazza et al., 2001). PML is a RING finger nuclear matrix-associated protein that typically concentrates within discrete speckled multiprotein subnuclear domains, termed as the PML nuclear bodies (PML-NBs) (Zhong et al., 2000). It is now becoming apparent that PML and the PML-NB are essential for critical tumour suppressive functions such as induction of cell cycle arrest, cellular senescence and apoptosis (Ferbeyre et al., 2000; Fogal et al., 2000; Guo et al., 2000; Pearson et al., 2000; Salomoni and Pandolfi, 2002; Bernassola et al., 2004). We and others have shown that PML can modulate p53 function upon DNA damage and oncogenic transformation by acting as a transcriptional coactivator through its ability to favour CBP/p300-mediated acetylation of p53, thus enhancing its DNA binding ability (Fogal et al., 2000; Guo et al., 2000; Pearson et al., 2000). We more recently presented an additional mechanism through which PML regulates the function of p53 and its homologue p73 by inhibiting their ubiquitin-proteosome degradation through two distinct pathways (Bernardi et al., 2004; Bernassola et al., 2004).

Both p53 and p73 have been shown to physically interact with PML through their DBDs (Guo et al., 2000; Bernassola et al., 2004). Since the DBD of p63 shares over 60% amino-acid identity with the DBD of p53, and over 85% identity with that of p73, we thus investigated the possible interaction between p63 and PML, and whether PML would act as a functional modulator of the entire p53 family.

We found that exogenous TAp63 and ΔNp63 co-immunoprecipitated with both exogenous PML-IV (Figure 1a, lanes 11, 12) and endogenous PML (Figure 1a, lanes 8, 9). The PML-III isoform, which has a shorter C-terminal tail than PML-IV and does not bind to p53, did not physically interact with either TA or ΔNp63α isoforms (data not shown).

Figure 1

PML and p63 physically interact in vivo. (a) Exogenous p63 immunoprecipitates with endogenous and exogenous PML. 293 T cells were transfected with equal amounts of TAp63α or ΔNp63, with or without PML. Cells were lysed in NP-40 buffer (50 mM Tris-Cl, 5 mM EDTA, 150 mM NaCl, 1% NP40) and PML was immunoprecipitated from 1 mg total protein lysate using 1 μg monoclonal anti-PML (Santa Cruz, clone PG-M3) preadsorbed to Protein G Sepharose beads (Amersham). Western blot analysis was carried out on 30 μg whole protein lysate (input) and on immunoprecipitations (IPs) by using anti-p63 (Ab-4 Cocktail, NeoMarkers) and polyclonal anti-PML. (b) Binding of in vitro-translated p63 proteins to His-PML. p63 or p53 proteins were generated in vitro using the quick-coupled in vitro transcription and translation system (Promega). For the in vitro pulldown assays, p53, TAp63α or ΔNp63α were mixed with His-PML bound to Ni-NTA magnetic agarose beads (Qiagen) in binding buffer containing 50 mM Tris-HCl, pH 7.4, 150 mM NaCl, 0.1% Nonidet P-40 and protease inhibitor mixture for 2 h at 4°C. PML was dissociated from the Ni-NTA resin by increasing the imidazole concentration to 200 mM and protein complexes were detected by IB with anti-HA antibody. (c) Endogenous PML and ΔNp63 co-immunoprecipitate in HaCaT kerotinocytes. Lysates were produced as in (a) and IPs were carried out from 1 mg total protein extracts using either a control isotypic antibody or anti-PML antibody (PG-M3)

To assess whether p63–PML interaction also occurred in vitro, we performed pull-down assays using His-PML. Ni-NTA magnetic agarose beads bearing His-PML were incubated with in vitro-translated TAp63α, ΔNp63α or p53. As shown in Figure 1b, both TAp63α (lane 6) and ΔNp63α (lane 8) were found to associate with PML, thus proving that p63 can directly bind to PML.

We next investigated whether endogenous PML and p63 co-immunoprecipitated. To this end, we utilized transformed human HaCaT keratinocytes, which express high levels of the ΔNp63 isoforms, as it is the case in primary epithelial cells and in other epithelial cell lines (Yang et al., 1998). We found that PML readily interacted with ΔNp63α (Figure 1c, lane 1).

The PML-NB has been shown to be an essential center of regulation and post-translational modification for p53 and p73 (Pearson et al., 2000; Bernassola et al., 2004). We therefore analysed PML and p63 cellular distribution pattern in HaCaT cells. After cellular fractionation to yield cytoplasmic, soluble and insoluble nuclear fractions, both p63 and PML are mainly found in the nuclear insoluble fraction (Figure 2a, lane 3), suggesting that PML and p63 may interact and colocalize in the PML-NB. Upon overexpression of PML and TAp63 into Pml null 3T3 immortalized mouse embryo fibroblasts (MEFs), we indeed observed that p63 displayed a diffuse as well as a speckled nuclear distribution pattern with significant colocalization with PML in the PML-NBs (Figure 2b). Notably, endogenous proteins colocalize in the PML-NBs in HaCaT cells (Figure 2c), suggesting that PML plays a role in p63 regulation in vivo.

Figure 2

p63 and PML colocalize in vivo. (a) Both PML and p63 are primarily present in the insoluble nuclear fraction of HaCaT cells. HaCaT cytosolic and nuclear fractions were separated as described in Dignam et al. (1983). Briefly, nuclei were suspended in 50 mM Tris, 150 mM NaCl, 0.5% NP40 and lysed via quick freeze and subsequent centrifugation. The resulting supernatant was taken as soluble nuclear fraction (SNF), while the resulting pellet was resuspended in the same buffer and subjected to DNAse I (500 U/ml) digestion, at 37°C for 5 min. NaCl concentration was then raised to 500 mM, and sample was vigorously vortexed. This was taken as the insoluble nuclear fraction (INF). Cytosol (C), SNF and INF were immunoblotted by using polyclonal anti-PML and anti-p63 antibodies. Nuclear fractionation was confirmed with anti-histone H1 and anti-Lamin A/C antibodies. (b) PML and p63 colocalize in the PML-NB. Pml−/− 3T3 immortalized MEFs were transfected with TAp63α and PML-GFP expression vectors. Cells were fixed in 4% paraformaldehyde, permeabilized with 0.1% TritonX-100 and blocked with 10% goat serum. Cells were then incubated with a polyclonal anti-p63 antibody (Santa Cruz, clone H-129) in PBS containing 10% goat serum for 1 h, followed by incubation with a goat anti-rabbit secondary antibody conjugated to the AlexaFluor fluorophore 568 nm (Molecular Probes). Nuclei were stained with DAPI and images acquired using a Leica confocal microscope. (c) Endogenous PML and p63 colocalize in human keratinocytes. HaCat cells were stained as detailed in b, using a monoclonal anti-PML (PG-M3) and a polyclonal anti-p63 (H-129) antibody. Secondary antibodies used were: goat anti-mouse and goat anti-rabbit conjugated to AlexaFluor fluorophore 488 and 568 nm, respectively. Images were obtained using a C1 Nikon microscope, and related software

Having established that p63 and PML physically interact, we investigated the functional implications of this interaction. Similarly to p53 and p73 (Bernardi et al., 2004; Bernassola et al., 2004), we found that PML overexpression led to a dose-dependent accumulation of both TAp63α (Figure 3a, lanes 2, 3) and ΔNp63α (Figure 3a, lanes 5, 6), although the N-terminally deleted isoform was induced to a lesser extent and only at the highest dose of PML. We also observed a decrease in p63 ubiquitination levels upon PML overexpression (data not shown), thus indicating that, similarly to p73 (Bernassola et al., 2004), PML leads to p63 protein accumulation by inhibiting its ubiquitin-proteasome dependent degradation.

Figure 3

PML increases p63 protein levels and transcriptional activity. (a) PML increases p63 levels. The p53 null H1299 cells were transfected with TAp63α or ΔNp63α in the absence or in the presence of increasing doses of PML. Cellular extracts were immunoblotted with anti-p63 and anti-PML and anti-β-tubulin antibodies. The lower numbers represent quantitation of p63 immunoreactivity levels normalized against β-tubulin. (b–e) PML increases p63-dependent transcription. For the transcriptional assays the following reporter plasmids were utilized: GADD45 min-luc (b, c) or p21 min-luc (d, e) (kindly provided by Dr CJ Di Como) containing a duplex oligonucleotide encoding the p53-responsive cis-acting element from the GADD45 or p21 promoters, respectively, cloned upstream of the minimal c-fos promoter in pGL3. H1299 cells were cotransfected with the p53-responsive reporter along with the indicated combinations of TAp63α, ΔNp63α, TAp63γ, ΔNp63γ and PML. pRL-TK vector was included to normalize transfection efficiency and reporter basal luciferase activity was normalized as 1. Values are averages±s.e. of three separate experiments each performed in duplicate. (f) p63 transcriptional activity is reduced in Pml−/− MEFs. Pml−/− and wild-type MEFs were transfected with TAp63γ or the empty vector, along with a p21 promoter-luciferase construct and luciferase activity assessed as described above

PML has been shown to increase the transactivation potential of p53 and TAp73 (Guo et al., 2000; Bernardi et al., 2004; Bernassola et al., 2004). To examine the effect of PML on p63-dependent transcription, the p53-null cell line H1299 was cotransfected with expression vectors encoding TAp63α, ΔNp63α, TAp63γ or ΔNp63γ, along with reporter plasmids containing the p53-binding sites from the GADD45 (Figure 3b and c) and p21 (Figure 3d and e) promoters placed upstream of a luciferase cDNA, in the absence or in the presence of PML. As shown in Figure 3, TAp63α and TAp63γ transactivated both promoters to various degrees. In addition, albeit to a lesser degree than the TA isoforms, both ΔNp63α and ΔNp63γ transactivated the GADD45 (3.1 vs 4.7 and 3.2 vs 278-fold over control for the α and γ isoform, respectively; Figure 3b and c) and p21 (1.3 vs 3.0 and 7.3 vs 201-fold over control for the α and γ isoform, respectively; Figure 3d and e) promoter-luciferase reporter. Interestingly, at the highest dose used to induce p63 (see Figure 3a) PML enhanced the transactivation activity of TAp63α and TAp63γ on both GADD45 (Figure 3b and c), p21 (Figure 3d and e) and bax (data not shown) promoters, whereas it displayed nearly no coactivation ability on the ΔN isoforms (Figure 3b–e). In agreement with these findings, we also found that the p63γ-dependent transactivation of the p21 promoter was impaired in Pml null compared to wild-type MEFs (Figure 3f), implying that PML is essential for p63 transcriptional activity.

Our findings lead to the conclusion that p63 and PML physically interact and colocalize in the PML-NB in vivo and that association with PML preferentially accumulates and specifically increases the transcriptional activity of the TAp63 isoforms. ΔNp63 is known to be a weaker transcriptional activator than the TAp63. ΔNp63 is nevertheless able to transactivate some of the p53-responsive promoters. Although PML and ΔN proteins are found to physically interact, these p63 isoforms are less susceptible to PML-mediated coactivation on promoters such as GADD45 and p21. The identification and characterization of PML as a p63 interactor and functional modulator is of particular relevance because of the present limited knowledge of the mechanisms responsible for p63 regulation. Further work is therefore needed to identify the molecular mechanisms through which PML regulates p63 and to establish if the PML-NB is required for PML to stabilize and functionally activate p63. Importantly, our findings demonstrate that PML acts as a general regulator of the p53 family of transcription factors at multiple levels.

Unlike p53, p63 is not the target of inactivating mutations in human cancers (Park et al., 2000; Weber et al., 2002a, 2002b). In addition, mice lacking p63 do not appear to develop spontaneous tumours (Mills et al., 1999; Yang et al., 1999), even though this needs to be formally tested in a longer follow-up in large cohorts of mutants and in isoform-specific knockouts. Nevertheless, recent data have implied that disruptions in the balance between TAp63 and ΔNp63 isoforms in epithelial tissues may be of greater importance in tumorigenesis than direct gene mutation (Park et al., 2000). ΔNp63 isoforms may indeed confer a proliferative advantage on cancer cells by counteracting the transactivation activities of p53 and TAp63 proteins and, hence, their ability to induce cell cycle arrest and apoptosis (Crook et al., 2000). Interestingly, ΔNp63 is the most highly expressed isoform in squamous cell, prostate and breast carcinomas (Crook et al., 2000; Nylander et al., 2002). In addition, loss or impaired expression of TAp63, possibly caused by altered proteasome-dependent degradation, has been associated with tumour progression and poor prognosis in most invasive bladder cancers, which, conversely, display concomitantly upregulation of ΔNp63 (Urist et al., 2002). Hence, the relative upregulation of ΔNp63 vs TAp63 isoforms in cancers may contribute to promote tumour growth. Interestingly, loss or reduction of PML protein expression has been found in human cancers of various histologic origins including epithelial tumours such as prostate and breast carcinomas, in which it was associated with tumour grade and progression (Gurrieri et al., 2004). The ability of PML to increase preferentially the stability and transcriptional activity of TAp63 thus favouring the predominance of TA isoforms, provides a further explanation for how loss of PML protein would favour tumour initiation, and enforces the notion that p63 functional loss may contribute to tumorigenesis.


  1. Bernardi R, Scaglioni PP, Bergmann S, Horn HF, Vousden KH and Pandolfi PP . (2004). Nat. Cell. Biol., 6, 665–672.

  2. Bernassola F, Salomoni P, Oberst A, Di Como CJ, Pagano M, Melino G and Pandolfi PP . (2004). J. Exp. Med., 199, 1545–1557.

  3. Crook T, Nicholls JM, Brooks L, O'Nions J and Allday MJ . (2000). Oncogene, 19, 3439–3444.

  4. de The H, Lavau C, Marchio A, Chomienne C, Degos L and Dejean A . (1991). Cell, 66, 675–684.

  5. Dignam JD, Lebovitz RM and Roeder RG . (1983). Nucleic Acids Res., 11, 1475–1489.

  6. Duijf PH, Vanmolkot KR, Propping P, Friedl W, Krieger E, McKeon F, Dotsch V, Brunner HG and van Bokhoven H . (2002). Hum. Mol. Genet., 11, 799–804.

  7. Ferbeyre G, de Stanchina E, Querido E, Baptiste N, Prives C and Lowe SW . (2000). Genes Dev., 14, 2015–2027.

  8. Fogal V, Gostissa M, Sandy P, Zacchi P, Sternsdorf T, Jensen K, Pandolfi PP, Will H, Schneider C and Del Sal G . (2000). EMBO J., 19, 6185–6195.

  9. Goddard AD, Borrow PS, Freemont PS and Solomon E . (1991). Science, 254, 1371–1374.

  10. Guo A, Salomoni P, Luo J, Shih A, Zhong S, Gu W and Pandolfi P . (2000). Nat. Cell. Biol., 2, 730–736.

  11. Gurrieri C, Capodieci P, Bernardi R, Scaglioni PP, Nafa K, Rush LJ, Verbel DA, Cordon-Cardo C and Pandolfi PP . (2004). J. Natl. Cancer Inst., 96, 269–279.

  12. Melino G, De Laurenzi V and Vousden KH . (2002). Nat. Rev. Cancer, 2, 605–615.

  13. Melino G, Lu X, Gasco M, Crook T and Knight RA . (2003). Trends Biochem. Sci., 28, 663–670.

  14. Mills AA, Zheng B, Wang XJ, Vogel H, Roop DR and Bradley A . (1999). Nature, 398, 708–713.

  15. Nylander K, Vojtesek B, Nenutil R, Lindgren B, Roos G, Zhanxiang W, Sjostrom B, Dahlqvist A and Coates PJ . (2002). J. Pathol., 198, 417–427.

  16. Pandolfi PP, Grignani F, Alcalay M, Mencarelli A, Biondi A, LoCoco F and Pelicci PG . (1991). Oncogene, 6, 1285–1292.

  17. Park BJ, Lee SJ, Kim JI, Lee CH, Chang SG, Park JH and Chi SG . (2000). Cancer Res., 60, 3370–3374.

  18. Parsa R, Yang A, McKeon F and Green H . (1999). J. Invest. Dermatol., 113, 1099–1105.

  19. Pearson M, Carbone R, Sebastiani C, Cioce M, Fagioli M, Saito S, Higashimoto Y, Appella E, Minucci S, Pandolfi PP and Pelicci PG . (2000). Nature, 406, 207–210.

  20. Piazza F, Gurrieri C and Pandolfi PP . (2001). Oncogene, 20, 7216–7222.

  21. Salomoni P and Pandolfi PP . (2002). Cell, 108, 165–170.

  22. Urist MJ, Di Como CJ, Lu ML, Charytonowicz E, Verbel D, Crum CP, Ince TA, McKeon FD and Cordon-Cardo C . (2002). Am. J. Pathol., 161, 1199–1206.

  23. Weber A, Bellmann U, Bootz F, Wittekind C and Tannapfel A . (2002a). Int. J. Cancer, 99, 22–28.

  24. Weber A, Langhanki L, Schutz A, Gerstner A, Bootz F, Wittekind C and Tannapfel A . (2002b). Virchows Arch., 441, 428–436.

  25. Westfall MD, Mays DJ, Sniezek JC and Pietenpol JA . (2003). Mol. Cell. Biol., 23, 2264–2276.

  26. Wu G, Nomoto S, Hoque MO, Dracheva T, Osada M, Lee CC, Dong SM, Guo Z, Benoit N, Cohen Y, Rechthand P, Califano J, Moon CS, Ratovitski E, Jen J, Sidransky D and Trink B . (2003). Cancer Res., 63, 2351–2357.

  27. Yang A, Kaghad M, Wang Y, Gillett E, Fleming MD, Dotsch V, Andrews NC, Caput D and McKeon F . (1998). Mol. Cell, 2, 305–316.

  28. Yang A, Schweitzer R, Sun D, Kaghad M, Walker N, Bronson RT, Tabin C, Sharpe A, Caput D, Crum C and McKeon F . (1999). Nature, 398, 714–718.

  29. Zhong S, Salomoni P and Pandolfi PP . (2000). Nat. Cell. Biol., 2, E85–90.

Download references


We thank CH Di Como for the GADD45 min-luc and p21 min-luc reporter plasmids, KS Chang for the polyclonal anti-PML antibody; the MSKCC confocal core facility for technical assistance; E Candi, A Terrinoni and P Salomoni for helpful discussion and critical review of the manuscript. This work was supported by the NIH CA-71692 awarded to PPP and by AIRC, EU-QLK-CT-2002-01956, Telethon-GGP04110, FIRB RBN01NWCH-008 to GM. The financial support by the Medical Research Council is also gratefully acknowledged.

Author information

Correspondence to Pier Paolo Pandolfi.

Rights and permissions

Reprints and Permissions

About this article


  • p63
  • PML
  • nuclear body
  • transcription

Further reading