Conservation of all three p53 family members and Mdm2 and Mdm4 in the cartilaginous fish

Abstract

Analysis of the genome of the elephant shark (Callorhinchus milii), a member of the cartilaginous fishes (class Chondrichthyes), reveals that it encodes all three members of the p53 gene family, p53, p63 and p73, each with clear homology to the equivalent gene in bony vertebrates (class Osteichthyes). Thus, the gene duplication events that lead to the presence of three family members in the vertebrates dates to before the Silurian era. It also encodes Mdm2 and Mdm4 genes but does not encode the p19^Arf gene. Detailed comparison of the amino acid sequences of these proteins in the vertebrates reveals that they are evolving at highly distinctive rates, and this variation occurs not only between the three family members but extends to distinct domains in each protein.

Introduction

The evolutionary origins and history of protein coding genes can offer great insight into their current structure and function.¹ Owing to their phylogenetic position, chondrichthyans, comprising chimeras, sharks, skates and rays, the oldest living group of jawed vertebrates that diverged from a common ancestor of bony vertebrates (Osteichthyes, ray-finned fishes including teleosts, coelacanths, lungfishes and tetrapods) in the early Silurian era about 420 million years ago (Ma)² provide a critical reference for our understanding of vertebrate genome evolution. The extant cartilaginous fishes, comprising approximately 970 species, are divided into two major groups: subclasses Holocephali (chimeras) and Elasmobranchii (sharks, skates and rays). Recently, one of the holocephalians, elephant shark (Callorhinchus milii), has attracted much attention in comparative genomics because of its relatively small genome (910 Mb), which has now been sequenced to about 1.4x coverage.³

Examination of this new genomic information has allowed us to identify the elephant shark genes encoding all three p53 family members, p53, p63 and p73, that are present in the higher vertebrates. This establishes that the gene duplication events that gave rise to these three genes occurred earlier than 420 Ma. Similarly, the E3 ligase Mdm2 and its partner protein Mdm4, which together regulate p53 activity in man and mouse, are also clearly encoded for within the elephant shark genome. Comparison of the amino acid sequences between the vertebrate species establishes the high rate of variation of p53 compared with the other family members and suggests that the development of the p53 isoforms is a relatively late event.

Results

Elephant shark p53, p63 and p73 genes.

The exon-intron organizations of the elephant shark p53, p63 and p73 genes are very similar, with the positions and phases of introns conserved (Fig. 1). This suggests that the three genes are the result of duplications of an ancestral gene.

Figure 1.

Alignment of elephant shark (Eshark) p53, p63 and p73 protein sequences with positions of introns marked. Peptide sequences encoded by different exons are shown in different colors. Intron positions are marked with colored boxes, with introns in phase 0 shown in yellow, phase 1 in green and phase 2 in red.

Elephant shark p53.

When the amino acid sequence of the elephant shark p53 is aligned (Fig. 2) with that of human, mouse, chicken, frog and teleost fish, striking conservation of the DNA binding domain is immediately apparent. At the N terminus, the critical contact residues for binding by Mdm2/Mdm4 are conserved along with critical sites of predicted ATM phosphorylation, in particular, serine 15 and threonine 18, which are involved in regulating the p53 response to DNA damage along with the critical LW amino acids required for transcription activation domain 1. Sequence conservation around the transcription activation domain 2 is not clear.⁴ At the C terminus, a clear tetramerization domain is conserved along with the extreme C terminus, which contains critical lysine residues that are subject to multiple modification, and the extreme C-terminal phosphorylation site DSE (DSD).⁵

Figure 2.

Alignment of the p53 sequence from elephant shark (Eshark) (GenBank accession number JN794073), zebrafish (P79734), Xenopus tropicalis (Frog) (F7A9U0), chicken (P10360), mouse (P02340) and human (P04637). The bars above the sequence represent the various identified domains in human p53. Purple denotes the N-terminal domain (TA1); black denotes the DNA-binding domain; pink denotes the tetramerization domain.

Overall, the sequence homology is 47% with human p53. Homology in the DNA binding domain greatly exceeds that at the N- and C-terminal domains. (N-terminal domain 26% identity, DNA binding domain 63% identity and the C terminus 31% identity with human). The methionine residues that initiate the ΔN 40 and ΔN 133 isoforms of human p53 are absent in the sequence of the elephant shark p53, which encodes no methionine in the first 228 amino acids, apart from the initiating amino acid, suggesting that these isoforms may not exist in this species.

Elephant shark p63.

The elephant shark p63 protein sequence readily aligns (Fig. 3) with the p63 sequences from other vertebrates, showing extraordinary conservation of the DNA binding domain and C terminus. In addition, a stretch of approximately 50 residues N-terminal of the DNA binding domain is also highly conserved between all the species examined. The elephant shark, mouse and human sequences show N-terminal extensions of approximately 100 amino acids absent in the frog and zebrafish sequences. These sequences correspond to the TA (transcriptional activation domain) of mouse p63. The alternate first coding exon of the ΔN form of zebrafish and frog p63 is also shown. The splice site that brings all of the p63 family products into alignment occurs at P 109 (human and mouse) and P 133 in the elephant shark sequence. In the chicken, an incomplete form of the TA domain has been found starting at the sequence RFV corresponding to amino acid 20 of the TA domain of human p63 and amino acid 44 of the elephant shark sequence. The TA domain is encoded by multiple exons that are separated by very long introns from the rest of the p63 gene.^6,7 The TA species mRNA is not as widely or abundantly expressed as the ΔN encoding mRNA. Thus, it is likely that TA forms of p63 exist in all the vertebrates but have, as yet, missed detection. Sensitive 5′ RACE analysis of ovarian tissue from these organisms could resolve this issue. The N-terminal extensions contain a number of conserved sequences, including a series of SQ motifs that are potential sites for phosphorylation by DNA damage-activated kinases like ATM. At the C terminus, striking homology is also present between all the species. Notable is the conservation of the extra helix reported by Joeger and Fersht, which helps to stabilize the overall architecture of the tetramerization domain.⁸

Figure 3.

Alignment of the p63 sequence from elephant shark (Eshark) (GenBank accession number JN794074), zebrafish (A7YYJ7), Xenopus tropicalis (F6ZGN7), chicken (F1N8Z7), mouse (O88898) and human (Q9H3D4). The bars above the sequence represent the various identified domains in human p63. Purple denotes the N-terminal domain (TA1); black denotes the DNA-binding domain; pink denotes the tetramerization domain; blue denotes the SAM domain.

Overall, the sequence identity is 72% with human p63. Identity in the first N-terminal domain (the first 95 residues) is 35%; in the second N-terminal domain, which is present in all species shown, it is 76%; in the DNA binding domain, 95% and in the C-terminal domain, 69%.

Elephant shark p73.

The elephant shark p73 protein sequence shows the highest level of homology (Fig. 4) with those in other vertebrate species, with long stretches of complete identity throughout the protein. Most prominent is the Mdm2/Mdm4 binding motif at the N terminus and, shared with p63, the additional stabilizing helix C-terminal to the tetramerization motif.⁸ The overall identity with human p73 is 73%. This breaks down to N-terminal 58%, DBD 92% and C-terminal 67%.

Figure 4.

Alignment of the p73 sequence from elephant shark (Eshark) (GenBank accession number JN794075), zebrafish (B0S576), Xenopus tropicalis (F6TKT0), chicken (F1P1U2), mouse (Q9JJP2) and human (O15350). The bars above the sequence represent the various identified domains in human p73. Purple denotes the N-terminal domain (TA1); black denotes the DNA-binding domain; pink denotes the tetramerization domain; blue denotes the SAM domain.

Elephant shark Mdm2.

The elephant shark Mdm2 protein shows an overall 49% amino acid identity (Fig. 5) with human Mdm2.⁹ All the key features that define a functioning Mdm2 protein are present, including the p53 binding domain, the acidic region, the zinc finger and the C-terminal ring domain and the essential extreme C-terminal sequences.¹⁰ In addition, the key sites for phosphorylation by ATM that inhibit Mdm2 function after DNA damage to allow p53 activation are present, including the especially critical SQ motif at murine position 395. The recently identified regulatory caspase-2 cleavage site is also present (GIDVPDCK) in the elephant shark Mdm2.

Figure 5.

Alignment of the Mdm2 sequence from elephant shark (Eshark) (GenBank accession ID JN794076), zebrafish (Q561Z0), Xenopus tropicalis (Q6P3Q9), chicken (F1NGX6), mouse (P23804) and human (Q00987). The bars above the sequence represent the various identified domains in human Mdm2. Blue denotes the N-terminal domain; black denotes the acidic domain; light green denotes the Zinc Finger domain; red denotes the caspase cleavage site; dark green denotes the RING finger domain.

Elephant shark Mdm4.

The elephant shark Mdm4 protein shows 40% overall amino acid identity (Fig. 6) to the human Mdm4. Again, all the key features of this p53 binding protein are present, including the p53/p73 binding domain, the caspase-2 cleavage site and the unique C terminus that prevents Mdm4 showing intrinsic E3 ligase activity.¹⁰

Figure 6.

Alignment of the Mdm4 sequence from elephant shark (Eshark) (GenBank accession ID JN794077, zebrafish (Q7ZUW7), Xenopus tropicalis (B5DFR1), chicken (E1C4B0), mouse (O35618) and human (O15151). The bars above the sequence represent the various identified domains in human Mdm4. Blue denotes the N-terminal domain; brown denotes the acidic domain; light green denotes the Zinc Finger domain; red denotes the caspase cleavage site; dark green denotes the RING finger domain.

Absence of elephant shark p19^Arf.

While Arf sequences have been detected in both mouse and human, this protein that shares an exon with p16 seems to be absent in the teleost fishes so far studied.¹¹ Consistent with its late evolution is the absence of the Arf protein coding first exon in the elephant shark. The second exon of the Arf gene, which is held in common with p16, is present (GenBank accession number JN794078), however, again reflecting what has been seen in the teleost fish. The absence of Arf may not abolish oncogene signaling to p53, as oncogenes can function to activate p53 both through ribosomal stress pathways and through DNA damage pathways.¹²

Discussion

The tumor suppressor protein p53 is now firmly established as the gene most frequently mutated in human cancer, with over 50% of all cancers showing point mutations in the coding region.¹³ These mutations are typically missense mutations, and the mutant protein is often expressed at high levels in tumors. The normal function of p53 is to act as a highly inducible transcriptional regulator, and p53 can activate or repress the activity of a large number of genes. The activity of p53 is controlled to a very precise level, and in mice, the key regulators consist of a pair of p53 binding proteins, Mdm2 and its partner, Mdm4. Mdm2 has a potent ubiquitin E3 ligase activity and, acting in a heteromeric complex with Mdm4, targets p53 for degradation.⁹ The p53 response is activated when proteins that covalently modify or bind to the Mdm2/4 complex are induced, for example, in response to oncogene activation, nucleolar stress and DNA damage.^5,14

In the vertebrate species examined in detail, p53 is a member of a gene family of three expressed genes, the other two genes being p63 and p73. While p53's main role is to act as a tumor suppressor and, remarkably, p53-null mice are viable (albeit sub-fertile¹⁵ and tumor-prone¹⁶), both p73 and p63 play essential and unique roles in normal development.

The evolutionary origins of the p53 family and their regulators have been the subject of considerable analysis. Proteins with significant structural and functional homology to the p53 family are present not only in the vertebrates, but also in the invertebrates.^17–19 Initially, studies in Drosophila and C. elegans yielded a simple picture in which these invertebrate species expressed a single p53 family gene and did not encode for any member of the Mdm2 family. More recent analysis of a larger number of invertebrate species has changed this picture, and multiple members of the p53 family have been reported in several invertebrates along with Mdm2-like proteins, suggesting that the Drosophila and C. elegans species have specialized in evolution and lost these genes.^17,20

The evolutionary origins of the three family members found in the higher vertebrates are, therefore, of great interest, and the recent determination of the genome sequence of the elephant shark, a cartilaginous fish, has allowed us to examine this question in detail. We find that the elephant shark genome clearly contains separate genes for p53, p63 and p73 that have arisen by gene duplication in contrast to earlier deductions made from incomplete early data.⁷ In addition, the elephant shark encodes separate genes for Mdm2 and Mdm4. Finally, analysis of the Ink4 locus suggests that the elephant shark lacks a functional Arf gene.

Detailed comparison of sequence conservation in the p53, p63 and p73 genes in the elephant shark compared with other vertebrate species reveals that these different family members are evolving at radically different rates, with the p63 and p73 genes being much more conserved than the p53 gene. Strikingly the N-terminal, C-terminal and DNA binding domains of the three proteins all show different degrees of evolutionary conservation. This is entirely consistent with the detailed analysis performed by Belyi et al.⁷

Methods

The 1.4x coverage elephant shark genome assembly (http://esharkgenome.imcb.a-star.edu.sg/) was searched by TBLASTN with human p53, p63, p73, Mdm2, Mdm4 and p16^INK4a proteins as queries. These searches identified fragments (coding exons) of elephant shark putative orthologs for these genes. By designing appropriate PCR primers for these exonic sequences and carrying out Rapid Amplification of cDNA Ends (RACE) with elephant shark cDNA as a template, we obtained complete cDNA sequences for the elephant shark genes (all sequences submitted to GenBank; accession numbers JN794073 to JN794078).

Disclosure of Potential Conflicts of Interest

No potential conflicts of interest were disclosed.