|
Gene name - p53 Synonyms - CG10873, Dmp53 Cytological map position - 94D12 Function - transcription factor Keywords - cell cycle, apoptosis, DNA repair. oncogene, response to DNA damage |
Symbol - p53 FlyBase ID: FBgn0039044 Genetic map position - Classification - p53 family Cellular location - nuclear |
EvoprintHD of p53
The mammalian p53 protein functions as a tumor suppressor by controlling cell cycle progression and cell survival. p53 is often described as the 'guardian of the genome' because it is a critical component of the cellular mechanisms that respond to genotoxic stresses like DNA damage and hypoxia to maintain the genomic integrity in part by arresting cell-cycle progression or by inducing apoptosis. p53 plays no essential role in the normal cell cycle, since the p53 knock out mouse develops normally. However, these mice as well as the transgenic mice carrying mutant p53 alleles are highly prone to develop spontaneous and carcinogen-induced tumors (Somasundaram, 2000 and references therein).
Numerous studies have established that growth arrest and apoptosis are independent functions of p53. p53-dependent G1 arrest occurs largely through transcriptional induction of p21WAF1 (Drosophila homolog: Dacapo), which prevents entry into S phase by inhibiting G1 cyclin-dependent kinase activity. However, p21 is not required for p53-dependent apoptosis. In fact, p21 may protect against p53-induced apoptosis in at least some cell types (Ollmann, 2000 and references therein).
Induction of apoptosis by p53 is critical for the tumor suppressor function of p53. There appear to be multiple mechanisms through which p53 promotes apoptosis. For example, p53 can transcriptionally activate the proapoptotic genes Bax (see Drosophila death executioner Bcl-2 homolog, Fas, and IGF-BP3, as well as a set of genes that may promote apoptosis through the formation of reactive oxygen species. Furthermore, there is evidence that p53 can induce apoptosis in the absence of its transcriptional activation function (Ollmann, 2000 and references therein).
The recent discovery of two p53-related genes, p63 and p73, has revealed an additional level of complexity in the study of p53 function in mammals (Somasundaram, 2000). Both genes encode proteins with transactivation, DNA-binding, and tetramerization domains, and some isoforms of p63 and p73 are capable of transactivating p53 target genes and inducing apoptosis. It was initially thought that only p53 was induced in response to DNA damage and other stress signals. However, there is now evidence that p73 is also activated by some forms of DNA damage in a manner that is dependent on the c-Abl tyrosine kinase (see Drosophila Abl oncogene). These data suggest that p53-independent apoptotic pathways may be mediated by other p53 family members (Ollmann, 2000 and references therein).
Characterization of the single Drosophila p53 homolog (refered here as Dmp53 or p53) was reported in back-to-back publications from two laboratories (Ollmann, 2000 and Brodsky, 2000a). Both identified the gene by homology searches of the expressed sequence tag database of the Berkeley Drosophila Genome Project.
To perturb the function of Dmp53 during Drosophila development, expression of dominant-negative forms in vivo were directed using the GAL4-UAS system. The effects of Dmp53 inactivation on the level of radiation induced apoptosis were tested. For these studies, transgenic strains were produced carrying a point mutantion in the DNA-binding domain [Dmp53(259H)] and the Dmp53(Ct) derivative possessing exclusively the C-terminal interaction domain. A similar C-terminal derivative has been used for tissue-specific inactivation of p53 in mice. The Dmp53 variants were specifically expressed in the posterior half of the developing Drosophila wing using an engrailed-GAL4 driver, line and effects on damage-induced apoptosis and cell cycle arrest were monitored after irradiation (Brodsky, 2000a).
The levels of apoptosis were tested in untreated and irradiated wing discs expressing dominant-negative Dmp53. In untreated wild-type discs, there are a small number of clustered apoptotic cells as visualized by staining with the vital dye acridine orange. In discs with engrailed-GAL4 driving expression of dominant-negative Dmp53, there is no substantial difference in the level of apoptosis in the anterior and posterior halves of the disc. Following irradiation, there is a massive increase in the amount of apoptosis throughout wild-type wing discs. However, in animals expressing dominant-negative Dmp53 in the posterior of the wing disc, radiation-induced apoptosis is greatly reduced in that region. This reduction is not due to minor differences in the age or handling of the discs since a robust radiation-induced apoptosis is observed in the anterior portion of the disc where dominant-negative p53 is not expressed. Together, these results indicate that Dmp53 is required for radiation-induced apoptosis in the wing, but not for the normal levels of cell death that occur in the absence of DNA-damaging agents (Brodsky, 2000a).
The effect of dominant-negative Dmp53 on radiation-induced arrest of cell cycle progression was tested. In mammals, p53 is required for radiation-induced G1/S arrest and has variable effects on radiation-induced G2/M arrest. The Drosophila wing exhibits a G2/M DNA damage checkpoint (Brodsky, 2000b) that is dependent on genes such as mei-41 (a homolog of the human ATM checkpoint gene) and grapes (a homolog of the yeast chk1 gene). Irradiation of wild-type wing discs blocks entry into mitosis; this block is not affected by expression of dominant-negative Dmp53. Thus, although Dmp53 is required for radiation-induced apoptosis, the data do not support a role for this protein during the G2/M checkpoint in the wing (Brodsky, 2000a).
In animals expressing dominant-negative Dmp53 under the control of engrailed-GAL4, the size and patterning of the adult wing is not noticeably altered. This result is consistent with the normal levels of mitosis and apoptosis in unirradiated wing discs expressing dominant-negative Dmp53. Similarly, widespread expression of dominant-negative Dmp53 using a tubulin-GAL4 driver does not generate any obvious adult phenotypes. These results suggest that, like human p53, Dmp53 is required in vivo to respond to certain cellular stresses, but may not be essential for normal development (Brodsky, 2000a). This role is unlike that of the mammalian homolog p63, which is required for limb development in both humans and mice (Celli, 1999; Mills, 1999; Yang, 1999).
The consequence of increased levels of wild-type Dmp53 was examined. Animals expressing wild-type Dmp53 using either the hsp70 or actin5C promoters do not survive to adulthood. Expression of Dmp53 using an eye-specific glass-dependent promoter leads to increased apoptosis in the eye imaginal disc and results in a rough, small eye phenotype; similar expression of Dmp53(259H) has no effect on eye morphology. Unlike radiation-induced apoptosis, apoptosis due to overexpression of Dmp53 is not suppressed by coexpression of the viral caspase-inhibitor p35; this observation suggests that overexpression of Dmp53 is sufficient to induce apoptosis, but that the response is either qualitatively or quantitatively different from the apoptotic response to DNA damage. Since the action of at least one Drosophila caspase is insensitive to p35 expression (Meier, 2000), overexpression of Dmp53 may activate apoptosis though a p35-resistant caspase (Brodsky, 2000a)
X irradiation of imaginal wing discs induces apoptosis that requires functional Dmp53. This poses the question of how Dmp53 activity is regulated in response to radiation. It is well established that mammalian p53 receives signals from a variety of cellular stresses such as various forms of DNA damage, nucleotide deprivation, incomplete DNA synthesis, and hypoxia. These signals are likely to work through a set of signaling pathways that activate and stabilize the p53 protein. Although understanding of the different gene products responsible for these various signaling pathways is still in its infancy, there is strong evidence that one pathway to p53, that induced by irradiation, requires functional ATM. The ATM protein kinase shares homology with other members of the PI3 kinase family, including the S. pombe DNA damage mediator kinase Rad3. Recently, two kinases that are downstream of ATM, CHK1, and CDS1/CHK2 (Drosophila homolog: loki) have been shown to phosphorylate and regulate human p53 (Chehab, 2000; Shieh, 2000). Given that many aspects of the DNA damage checkpoint response are conserved between yeast and mammals, it is possible that Dmp53 might be similarly regulated. Indeed, Drosophila homologs of ATM (mei41) and Chk1 (grapes) have been identified, and it will be interesting to determine their relationship to Dmp53. It should be mentioned here that while ATM has been clearly shown to regulate p53-mediated cell cycle arrest, there is evidence that apoptosis induced by p53 is independent of the function of ATM. Thus, the study of Dmp53-induced apoptosis in Drosophila may uncover new upstream regulators of p53 activity (Ollmann, 2000 and references therein).
Human p53 is a 393 amino acid protein composed of three main functional domains: an amino-terminal acidic transactivation domain, a central DNA-binding domain, and a carboxy-terminal tetramerization domain. Significant similarity between Dmp53 and the vertebrate p53 family is limited to the DNA-binding domain and includes residues identified in human p53 as critical for DNA sequence recognition and coordination of a zinc ion. A three-dimensional model of the Dmp53 DNA-binding domain was created based on the human p53 crystal structure (Cho, 1994 ). Conserved surface residues predominantly cluster in the DNA-binding site, while most of the remaining conserved residues are buried and involved in stabilizing the tertiary fold of the domain. This suggests that, despite limited sequence identity, the Dmp53 DNA-binding domain may adopt a tertiary structure similar to the human p53 DNA-binding domain (Ollmann, 2000).
There is no sequence similarity between Dmp53 and other p53 family members in the carboxyl termini, yet this region of Dmp53 contains secondary structures characteristic of p53-related proteins. Mammalian p53 binds DNA as a homotetramer, and self-association is mediated by a beta sheet and amphipathic alpha helix located in the carboxyl terminus of the protein. A similar beta sheet (residues 320-322 and 332-337) and amphipathic alpha helix (residues 341-359) are predicted in the carboxyl terminus of Dmp53. Consistent with this prediction, a yeast two-hybrid assay reveals that Dmp53 interacts with itself but not with human p53. Like the carboxyl terminus, the amino terminus of Dmp53 shows no sequence similarity with other p53 family members. However, this region is highly divergent among p53 family members, except for conserved residues critical for binding of MDM2 to human p53. Interestingly, the residues critical for MDM2 binding are not conserved in Dmp53 (Ollmann, 2000).
p53-related sequences have been described previously in invertebrates, including clams (Barker, 1997; Van Beneden, 1997) and squid (Schmale, 1997). Additional insect p53-related genes have been identified through EST sequencing projects in the flour beetle (Tribolium castaneum) and Colorado potato beetle (Leptinotarsa decemilineata). Squid p53 more closely resembles p63 or p73 because of its long C-terminal extension that contains the SAM domain characteristic of the p63/p73 subfamily. Dmp53 lacks a carboxy-terminal SAM domain, and there is no evidence for a SAM domain-containing alternative exon within the 100 kb of genomic sequence downstream of Dmp53. The beetle p53 proteins also lack any similarity to the carboxy-terminal domains of p63 and p73. Given that the insect p53-related genes represent the most evolutionarily distant members of the p53 gene family known to date, it is suggested that the p63/p73 subfamily arose from an ancestral p53-like gene after the split of the arthropod and vertebrate lineages (Ollmann, 2000).
In human tumors, mutations that inactivate p53 function are clustered in the well-conserved DNA-binding domain. Among the six most frequent sites of mutation in tumors, four are identical in Dmp53 and the other two are similar. Two of the mutation hotspots correspond to residues that directly contact DNA. Among the remaining six DNA-binding amino acids, four of the residues are identical, one is similar, and one is not conserved. Finally, the four residues that bind zinc are also conserved in Dmp53. Thus, despite being the most divergent p53 family member described to date, the conservation of amino acid residues required for DNA binding suggests that Dmp53 will interact with the consensus binding sequence for human p53 (Brodsky, 2000a).
The N termini of Dmp53 and human p53 share a high proportion of acidic residues. Thus, it is possible that the function of this domain in transcriptional activation is conserved in Dmp53. The C terminus of human p53 contains a basic region (9/26 amino acids) that helps regulate sequence-specific DNA binding by p53 and can itself bind either DNA or RNA. Although Dmp53 has little sequence similarity with human p53 in this region, it is enriched in basic residues (6/24). Adjacent to the basic region of Hp53 is a small domain required for tetramerization. The primary sequence in this region is poorly conserved in Dmp53, although Gly334, a critical 'hinge' residue is conserved in all species including Drosophila (Brodsky, 2000a).
Phosphorylation and acetylation have been implicated in regulating human p53 stability and transcriptional activity following DNA damage (Lakin and Jackson, 1999 ). Sites of particular note are the Ser-Glu amino acid pairs at positions 15 and 37 that can be serine phosphorylated by members of the ATM family of DNA damage-responsive kinases. The nearby Ser4-Glu5 pair in Dmp53 might be a target for one of these kinases. Among the acetylation sites that have been described, Lys382 in Hp53 has a possible equivalent at Lys373 in Dmp53. The conservation of these potential regulatory sites in Dmp53 suggests that some of the proteins that modulate p53 activity are conserved between mammals and Drosophila (Brodsky, 2000a).
Squid p53 and some splice forms of human p73 and p63 contain an extended C-terminal tail that contains a putative sterile alpha motif (SAM) domain, believed to mediate protein-protein interactions. None of the Drosophila cDNAs for Dmp53 contain this C-terminal domain and are, in this respect, more similar to the mammalian p53. Excluding the C-terminal tail, the sequence of Dmp53 is equally distant from human p53, human p73, and squid p53 (Brodsky, 2000a).
Drosophila p53, identified using homology searches, encodes a 385-amino acid protein with significant homology to human p53 in the region of the DNA-binding domain, and to a lesser extent the tetramerization domain. Purified Drosophila p53 DNA-binding domain protein was shown to bind to the consensus human p53-binding site by gel mobility analysis. In transient transfection assays, expression of Drosophila p53 in Schneider cells transcriptionally activates promoters that contain consensus human p53-responsive elements. Moreover, a mutant Drosophula p53 (Arg-155 to His-155), like its human p53 counterpart mutant, exerts a dominant-negative effect on transactivation. Ectopic expression of Drosophila p53 in Drosophila eye disc causes cell death and leads to a rough eye phenotype. Drosophila p53 is expressed throughout the development of Drosophila with highest expression levels in early embryogenesis; this early expression has a maternal component. Consistent with this, Drosophila p53 RNA levels are high in the nurse cells of the ovary. It appears that p53 is structurally and functionally conserved from flies to mammals (Jin, 2000).
Alignment of the protein sequences of Drosophila and human p53 DNA-binding domains the presence of the DNA-binding domains was between Drosophila residues 77 and 275 and human residues 94 and 289. Even though the match is statistically significant, the two domains share only 24% sequence identity and 44% similarity over 207 amino acids. This low similarity explains previous unsuccessful attempts to directly clone the Drosophila p53 using vertebrate sequences as molecular probes. The alignment was used to generate a comparative protein structure model for the Drosophila sequence based on the crystallographic structure of the human p53 DNA-binding domain. The elements of secondary structure are conserved between the Drosophila and human proteins, with bigger differences observed in loop regions. The amino acid residues that are known to be important for DNA-protein interactions as determined by x-ray crystallography and mutagenesis are well conserved between the two proteins. Another region of similarity is indicated by the three-dimensional structure of the tetramerization domain of human p53. The beta-strand and alpha-helix structural elements that constitute this domain in human p53 would be consistent with the sequence of Drosophila p53 spanning residues from 316 to 360. These findings suggest that Drosophila p53 might have a tetramerization domain similar to that of the human p53 (Jin, 2000).
date revised: 16 April 2000
Home page: The Interactive Fly © 1995, 1996 Thomas B. Brody, Ph.D.
The Interactive Fly resides on the
Society for Developmental Biology's Web server.