p53 family

Encyclopedia of Cancer, 2015


Ket; p40AIS; p51; p63; p73; TP63; TP73; TP73L; Trp63; Trp73


The p53 family consists of p53 and its two homologues, p63 and p73. All three proteins are transcription factors that bind specific DNA sequences to mediate gene expression involved in cell cycle arrest, apoptosis, and differentiation. p63 and p73 map to chromosomes 3q27 and 1p36, respectively. The three p53 family genes and their protein products share many characteristics, including some “p53-like activities.” However, some differences in structure and function suggest that each protein has unique roles in various biological and pathological processes from development to oncogenesis.

Molecular structure and function

In the late 1990s, two decades after the initial discovery of the p53 tumor suppressor protein, the p53 homologues p63 and p73 were identified. The p63 and p73 genes give rise to multiple mRNA, which are translated into several distinct isoforms (Fig. 1). Different isoforms result from the utilization of two promoters as well as alternative splicing. The three p53 family proteins have a common modular architecture with an N-terminal transactivation domain (TAD or TA), a DNA-binding domain (DBD), and an oligomerization domain (OD). The full-length p63 and p73 proteins that are translated from the first promoter are most similar to p53 and are referred to as TAp63 and TAp73, respectively. Like p53, TAp63 and TAp73 form oligomers and can induce apoptosis and cell cycle arrest. TAp63 and TAp73 regulate many of the same downstream target genes as p53 by binding to p53-responsive elements in genes such as p21, PUMA, and BAX. In addition, some unique target genes have been identified for p63 and p73, likely reflecting different promoter binding sequence preferences as well as unique tissue patterns of expression. Known TAp63 and p73 target genes are summarized in Table 1.

p63 and p73 can induce many p53 target genes that are involved in cell cycle arrest and apoptosis (upper panel). However, both proteins can also upregulate novel target genes that cannot be activated by p53 (lower panel). Noted in parentheses are genes only regulated by p63 or p73, but not both.

p63 and p73 also encode N-terminally truncated isoforms, ΔNp63 and ΔNp73, that lack the TAD. These ΔN variants are transcribed from the second promoter located within the third intron of both genes. Although the lack of TAD renders ΔN proteins transcriptionally inactive, they retain the ability to bind DNA and thus compete with transactivation-competent p53 family proteins for promoter binding sites. They can also form hetero-oligomers with TA proteins, preventing the formation of functional tetramers. Therefore, ΔNp63 and ΔNp73 are not only unable to induce downstream target genes, but they can act as dominant-negative proteins that block TAp63, TAp73, and p53 activities and, thus, have anti-apoptotic properties.

Alternative splicing at the C-terminus adds further to the complexity of p63 and p73, resulting in unique coding sequences (Fig. 1). The longest p63 and p73 proteins (p63-α and p73-α) have a sterile α-motif (SAM) domain not found in p53. SAM domains are known to serve as protein-protein interaction modules, raising the possibility that p63 and p73 splice variants with SAM domains are capable of recruiting isoform-specific binding proteins. A transcriptional inhibitory domain (TID) is also found in a subset of the different C-terminal isoforms. Thus, differential splicing of the C-termini influences the ability of these proteins to transactivate target genes. For example, in comparison to TAp63-α, TAp63-γ more strongly induces downstream pro-apoptotic genes. Thus, the differences in coding sequences and isoform-specific binding partners result in different activities of the various isoforms.

Various upstream pathways and stimuli, including DNA-damaging agents and oncogenes, such as E2F1 and Ras, regulate the expression and stability of p63 and p73 isoforms (summarized in Table 2). In addition, like p53, p63 and p73 are regulated by posttranslational modifications including phosphorylation, acetylation, and ubiquitination. Interactions with binding partners, such as p300 and MDM2, as well as heterotypic interactions with each other, further modulate the activities of the p53 family proteins.

p53 Family

Fig. 1. The structure of the p53 family of proteins. Shown are the transactivation (TAD, orange), DNA-binding (DBD, blue), and oligomerization (OD, yellow) domains for p53, p63, and p73. These p63 and p73 domains show significant homology to p53 (TAD, ~25 %; DBD, ~60 %; and OD, ~40 %). Alternative splicing at the C-terminus of p63 and p73 generates multiple isoforms (p63-α, p63-β, p63-γ and p73-α, p73-β, p73-γ, p73-δ, p73-ε, p73-φF, p73-ζ, p73-η) that are identical through the OD. The longest p63 and p73 isoforms (alpha) are shown in the figure and contain a SAM (green) and transinhibitory domain (gray) not found in most other C-terminal isoforms

Roles in development

Unlike p53, which is ubiquitously expressed, p63 and p73 are expressed in a tissue-specific fashion, and each plays roles in critical developmental processes. In contrast to p53 null mice in which only a small percentage have developmental defects in neural tube closure, p63 and p73 mutant mice have significant developmental abnormalities. p63-/-mice that lack all TA and DN isoforms have significant limb and craniofacial malformations as well as failed development of epithelial tissues including the skin. In humans, germline p63 mutations have been detected in patients with characteristics reminiscent of the p63 knockout mice. Heterozygous p63 mutations have been found in six different human ectodermal dysplasia syndromes characterized by combinations of skin, hair, mammary gland, craniofacial, and limb abnormalities. Studies into these disorders have provided important clues as to which tissues express p63 and have revealed that p63 is essential in epithelial morphogenesis and involved in senescence.

Studies of p73-/-mice have demonstrated that ΔNp73 is crucial in the development of the nervous system. p73 knockout mice have significant neurologic abnormalities due to the absence and/or loss of specific populations of neurons. ΔNp73 promotes cell survival by the inactivation of the full-length p53 family proteins, and, thus, the loss of ΔNp73 leads to enhanced apoptosis in cortical and sympathetic ganglia neurons. The relative balance between the TA and DN forms of p63 and p73, as well as p53, is important in these developmental processes and, similarly, in cancers that arise from these tissues.

Table 1. Target genes of the p53 family

Common p53-responsive p63 and p73 target genes Unique p63 and p73 downstream genes
  • Puma
  • Noxa
  • p21
  • Mdm2
  • Bax
  • CD95
  • Apaf-1
  • TNF
  • TNF-R1
  • Perp
  • AIP
  • Redd1
  • Jagged 1/jagged 2
  • β4-integrin
  • Dlx3 (p63)
  • Ets-1
  • IKKa
  • Gata-3
  • CyclinD1
  • Aquaporin (p73)
  • P57 kip2

Role in cancer: tumor suppressor or oncogene?

The high incidence of p53 mutations in human tumors coupled with the chromosomal locations of p63 and p73 led to the prediction that p63 and p73 were also tumor suppressor genes that are inactivated by mutations in certain human malignancies. Surprisingly, despite the functional similarities among the p53 family proteins, and specifically their ability to induce p53-responsive pro-apoptotic genes, only rare p63 and p73 mutations have been detected in both cell lines and primary tumors. Although the lack of mutations initially raised doubts as to whether these two genes have roles in human cancers, many studies have reported high levels of p73 and p63 expression in human cancers. However, many of these studies did not distinguish between the various isoforms. More recent evidence from mouse models and human tumors suggests that the relative expression levels of the different TA and ΔNp63 and p73 isoforms are important in tumor development and progression, as well as the response to treatment.

Knockout mice lacking one or both copies of p53, as well as mice that express specific p53 point mutations, develop a wide range of tumors at a young age. In contrast, p63+/- and p73+/- mice develop premalignant and cancerous lesions only when aged. Furthermore, in p53 mutant mice, the additional loss of p63 or p73 leads to different tumor types and higher tumor burden with more metastases. Importantly, tumors from p63 or p73 heterozygous mice were shown to have loss of heterozygosity (LOH) of the remaining allele, suggesting that p63 and p73 have recessive properties of a classical tumor suppressor gene. However, the relative contribution of the different TA and DN isoforms is difficult to determine since these knockout mice lack the expression of all p63 or p73 isoforms.

Loss of p63 and p73 has been described in several human tumor types. Loss of p63 expression is associated with bladder cancer progression and correlates with poor prognosis. TAp73 silencing by methylation in leukemias and lymphomas has also been described. Furthermore, low expression of p63 and p73 has been reported in breast cancer due to a number of mechanisms including LOH and allele silencing.

In addition to missense mutations, p53 is inactivated by binding to inhibitory proteins such as the papillomavirus E6 oncoprotein in cervical cancer and MDM2 in sarcomas. While TAp63 and p73 can likewise bind to MDM2, as well as other modulatory proteins such as MDMx, iASPP, YAP, and PML (see Table 2), the relative role of these complexes in human cancers has not been elucidated. However, there is growing evidence that inactivation of the putative tumor suppressor TA forms of p63 and p73 is mediated by hetero-oligomerization with a subset of p53 mutant proteins found commonly in tumors. These mutants include the p53 conformational mutants with “gain of function” properties. These mutant p53 proteins bind to and inhibit TAp63- and TAp73-dependent transactivation and apoptosis, leading to enhanced survival and growth. Furthermore, this binding affinity of mutant p53 for TAp73 is influenced by the status of a p53 polymorphism at codon 72. In certain tumors, such as head and neck squamous cell carcinoma, presence of arginine, instead of proline, at p53 codon 72 is associated with worse prognosis due, at least in part, to more potent binding to TAp73. Therefore, although cancer-associated missense changes are almost never found in the new p53 family members, other mechanisms exist to inactivate the tumor suppressor like forms TAp63 and TAp73.

Ironically, despite the focus on p63 and p73 as tumor suppressors, there is perhaps more evidence supporting roles for ΔNp63 and p73 isoforms as oncogenes. Amplification of the genomic region encompassing p63 and overexpression of the ΔNp63 protein have been detected in several epithelialderived tumors. In head and neck squamous cell carcinomas (HNSCCs) and breast cancers, ΔNp63 promotes the survival of transformed cells by blocking TAp73-dependent apoptosis. Thus, ΔNp63 behaves as a potent oncogene and highlights how perturbations in the equilibrium between DN and TA p63 and p73 proteins are important in tumorigenesis. Whether this holds true for other tumors that express high p63 levels including the cervix, nasopharynx, and bladder remains to be determined.

Much like p63, p73 is but highly expressed in several types of human malignancies. Overexpression of p73 has been reported in numerous cancer types including neuroblastoma; breast, lung, esophagus, bladder, ovarian, liver, and colon cancer; and certain types of leukemias. There is emerging evidence that when carefully studied, like p63, the upregulation of the ΔNp73 forms is relevant in the pathogenesis of these cancers. However, to date, specific overexpression of ΔNp73 transcripts and/or protein has only been shown in a few tumor types including mesenchymal or precursor-derived tumors such as rhabdomyosarcoma and neuroblastoma as well as breast, colon, and hepatocellular carcinomas. The tissue specificity of these findings is likely due to the fact that p73 has a restricted pattern of expression throughout development. The mechanisms responsible for p73 overexpression and specific mechanism(s) by which ΔNp73 promotes cancer development remain unknown. However, there is some evidence that ΔNp73 promotes immortalization of primary cells and can cooperate with oncogenes such as Ras to enhance transformation. Furthermore, the binding of ΔNp73 to p53 andTAp73 and TAp63 has been proposed to modulate differentiation in neuroblastoma and rhabdomyosarcoma.

Table 2. Regulators of the p53 family

Upstream regulators Oncogenes:
DNA damage:
        Chemotherapeutic agents*
BMP signaling
Notch signaling
Kinases c-Abl
Binding proteins Itch
Aspp family of proteins
p53 family proteins (hetero-oligomerization)

The p53 family is regulated by a number of proteins and signaling pathways. These upstream regulators can affect the p63 and/or p73 at the transcriptional or the posttranslational level. Some regulators have been shown to affect all three p53 family members (denoted with*).

Clinical relevance of p63 and p73: prognosis and chemosensitivity

Evidence in cell and animal systems supporting a role for the different isoforms in tumorigenesis has led to the assessment of p73 and p63 expression as a marker for clinical prognosis. The relative ratio of TA versus DN isoforms as measured at the transcript level is linked to prognosis in several tumor types, including breast and colon cancers. High levels of ΔNp73 protein have also been correlated with poor prognosis in small series of patients with neuroblastoma and breast cancers. Immunohistochemistry for p63 has also begun to be used as a marker in several tumor types including prostate and breast. However, large analyses of primary tumor samples will be required to determine if the relative expression of TA and DN isoforms of p63 and p73 can be linked to clinical prognosis. Nevertheless, roles for these two p53 family proteins in the response to DNA-damaging agents such as chemotherapies suggest that their expression has important therapeutic implications.

Currently, the two major treatments for cancer, radiation and chemotherapy, both exert their cytotoxic effects by stabilizing and activating p53. However, these therapies are also capable of killing cancer cells harboring p53 mutations, suggesting that p53-independent mechanisms exist to initiate apoptosis. Many studies support a role for the p53 family proteins in chemosensitivity. TAp73 is induced by g-irradiation and many chemotherapeutic drugs and in a variety of cancer cell types, including those with mutant and wild-type p53. Studies using cells in which p73 or p63 is inactivated by genetic deletion, the use of dominant-negative proteins, and short interfering RNA (siRNA) demonstrate that loss of p63 or p73 leads to chemoresistance. Similar to their effect on p53, many chemotherapies also induce p73 and p63 posttranslational modifications that stabilize and enhance their ability to induce apoptosis. These include c-abl-mediated phosphorylation in response to cisplatin and irradiation. p73 is also upregulated at the transcriptional level by some drugs. For example, doxorubicin treatment leads to E2F1-mediated induction of TAp73. Similar to p53, TAp73 and TAp63 mediate the expression of genes that control cell cycle arrest and apoptosis (Table 1). Thus, in response to anticancer treatments, TAp73, and perhaps TAp63, can substitute for and, in certain cases, cooperate with other p53 family proteins to eradicate transformed cells and promote tumor regression.

The interactions among the p53 family proteins also modulate their ability to induce chemotherapyinduced apoptosis. The expression of various tumor-derived p53 mutant proteins leads to chemoresistance via the ability of these mutant proteins to bind to and inactivate TAp73 and p63. Experiments in cell lines demonstrate that manipulation of the levels of these p53 mutant proteins by overexpression or siRNA leads to chemoresistance or chemosensitivity, respectively. Some chemotherapies also affect the levels of ΔNp73 and p63. Furthermore, overproduction of ΔNp73 or ΔNp63 in cell lines also leads to chemoresistance, and this is likely due to hetero-oligomerization and inactivation of the full-length p53 family proteins. This paradigm has been best studied in head and neck squamous cancers where downregulation of ΔNp63 in response to cisplatin results in diminished ΔNp63-TAp73 complex formation and enhanced apoptosis. Given the importance of DN proteins in cancer cell survival, this represents a second potential mechanism through which chemotherapeutic drugs induce apoptosis. Although the mechanisms by which chemotherapies downregulate ΔNp63 and p73 are not known, it is clear the balance between DN and TA proteins, as well as mutant forms of p53, is an important determinant of chemosensitivity of tumor cells.

Given the importance of p53 in tumor suppression, small-molecule inhibitors are actively being designed to activate p53. However, since p53 is nonfunctional in approximately half of all human cancers, recently there has been interest in developing therapeutic strategies to activate p73 and p63 to induce tumor cell death. Several different approaches might lead to enhanced p73 or p63 activity. Panels of drugs and small molecules can be screened for agents that either directly increase TAp73 or p63 or enhance their activation by upstream pathways (e.g., via E2F1 activation). Furthermore, in tumors with high levels of ΔNp73 or ΔNp63, strategies might be aimed at donwregulating these anti-apoptotic isoforms. Drugs that interfere with proteins that inactivate TAp73 and p63 can also be targeted. To this end, small peptides have been designed to modulate the interactions between the iASPP proteins and p63 and p73. Furthermore, therapeutic strategies using drugs and siRNA designed to interfere with the interaction between “gain of function” mutants of p53 and TAp73 have been successful in vitro. Finally, many drugs have been designed to interfere with p53 binding to its negative regulator MDM2. It is predicted that some of these drugs might also interfere with MDM2 binding to TAp73 and TAp63, raising the possibility that these small-molecule inhibitors may be useful for a broad spectrum of cancers.

Concluding remarks

Since their discovery, studies of p63 and p73 have been focused on comparisons to p53. While the prospect of TAp63 and TAp73 compensating for the loss of p53 in tumor suppression and treatment is exciting, it is equally important to recognize very significant differences in their roles in both cancer and development. Not all tissues and cancers exhibit the same expression pattern of p63 and p73. The levels and balance between the various pro- and anti-apoptotic p53 family proteins appear to play important roles in both development and cancer. Given the importance of p63 and p73 in physiological apoptosis as well as in tumors and their response to chemotherapies, understanding how they function will likely lead to the development of better cancer treatments.


Irwin MS, Miller FD (2004) p73: regulator in cancer and neural development. Cell Death Differ 11(Suppl 1):S17–S22

McKeon F, Melino G (2007) Fog of war: the emerging p53 family. Cell Cycle 6(3):229–232

Mills AA (2006) p63: oncogene or tumor suppressor? Curr Opin Genet Dev 16(1):38–44

Moll UM, Slade N (2004) p63 and p73: roles in development and tumor formation. Mol Cancer Res 2(7):371–386

Stiewe T (2007) The p53 family in differentiation and tumorigenesis. Nat Rev Cancer 7(3):165–168