The ancient function of RB-E2F pathway: insights from its evolutionary history

Abstract

Background: The RB-E2F pathway is conserved in most eukaryotic lineages, including animals and plants. E2F and RB family proteins perform crucial functions in cycle controlling, differentiation, development and apoptosis. However, there are two kinds of E2Fs (repressive E2Fs and active E2Fs) and three RB family members in human. Till now, the detail evolutionary history of these protein families and how RB-E2F pathway evolved in different organisms remain poorly explored.

Results: We performed a comprehensive evolutionary analysis of E2F, RB and DP (dimerization partners of E2Fs) protein family in representative eukaryotic organisms. Several interesting facts were revealed. First, orthologues of RB, E2F, and DP family are present in several representative unicellular organisms and all multicellular organisms we checked. Second, ancestral E2F, RB genes duplicated before placozoans and bilaterians diverged, thus E2F family was divided into E2F4/5 subgroup (including repressive E2Fs: E2F4 and E2F5) and E2F1/2/3 subgroup (including active E2Fs: E2F1, E2F2 and E2F3), RB family was divided into RB1 subgroup (including RB1) and RBL subgroup (including RBL1 and RBL2). Third, E2F4 and E2F5 share more sequence similarity with the predicted E2F ancestral sequence than E2F1, E2F2 and E2F3; E2F4 and E2F5 also possess lower evolutionary rates and higher purification selection pressures than E2F1, E2F2 and E2F3. Fourth, for RB family, the RBL subgroup proteins possess lower evolutionary rates and higher purification selection pressures compared with RB subgroup proteins in vertebrates,

Conclusions: Protein evolutionary rates and purification selection pressures are usually linked with protein functions. We speculated that function conducted by E2F4/5 subgroup and RBL subgroup proteins might mainly represent the ancient function of RB-E2F pathway, and the E2F1/2/3 subgroup proteins and RB1 protein might contribute more to functional diversification in RB-E2F pathway. Our results will enhance the current understanding of RB-E2F pathway and will also be useful to further functional studies in human and other model organisms.

Figure 1

Distribution of RB-E2F proteins in 21 representative Eukaryotic organisms. The gene number of E2F1-6, E2F7/8, RB, and DP family in 20 representative Eukaryotic organisms were listed. M. brevicollis, D. discoideum, and O. tauri are unicellular organisms. The phylogenetic relationship of these organisms was drawn according to the results of proteome-based phylogeny [12,15]. The accession numbers and sequences of all these proteins are listed in additional File 1

Figure 2

Phylogenetic analyses of E2F1-6 and E2F7/8 family genes in metazoan. Maximum likelihood analysis was conducted using Phyml program, and Bayesian analyses were carried out using MrBayes 3.1. Both methods produced nearly identical topologies. Black numbers above branches indicate ML bootstrap support (only greater than 50% values are labeled), and red numbers above branches indicate Bayesian posterior probabilities (only these key branches are labeled). The scale bar shows the number of substitutions per site. The detail information for protein names used in this figure can be found in materials and methods section. The accession numbers and sequence of all these proteins are listed in additional File 1. (A) Phylogenetic analyses of E2F1-6 family, and E2F-Mb (gi:167523471) was used as the outgroup; (B) Phylogenetic analyses of E2F7/8 family, and E2F7/8-Mb (gi:167517423) was used as the outgroup.

Figure 3

Conservation of exon intron structures in related genes from human, fly, and worm. The exon and intron structures of related genes from human, fly, and worm are showed, (A): E2F1-6 family, (C): RB family, and Protein sequences alignments of the region with interspecies conserved exons are also showed, (B): E2F1-6 family, (D): RB family. Boxes correspond to exons. Non-coding exons are shown in grey. The interspecies conserved exons are labeled with red color. The size of introns and exons in nucleotides is shown. Introns and non-coding regions are not drawn to scale. In the alignment, the protein sequences coded by interspecies conserved exons are labeled with blue color, and amino acid residues overlap splice sits are labeled with red color. The exon intron structures information are got from ensemble data base, detail of transcripts used for analysis are: E2F4-Hs: ENST00000379378; E2F5-Hs: ENST00000416274; E2F2-Dm: FBtr0081501; EFL1-Ce: Q9XX87_CAEEL (Y102A5C.18); RBL1-Hs: ENST00000373664; RBL2-Hs: ENST00000262133; RB1-Hs: ENST00000267163; RBF1-Dm: FBtr0070146. The exon intron structures of all E2F1-6 and RB Family genes from human, fly, and worm can be found in additional file 6.

Figure 4

Phylogenetic analyses of RB and DP family genes in metazoan. Maximum likelihood analysis was conducted using Phyml program, and Bayesian analyses were carried out using MrBayes 3.1. Both methods produced nearly identical topologies. Black numbers above branches indicate ML bootstrap support (only greater than 50% values are labeled), while red numbers above branches indicate Bayesian posterior probabilities (only these key branches are labeled). The scale bar shows the number of substitutions per site. The detail information for protein names used in this figure can be found in materials and methods section. The accession numbers and sequence of all these proteins are listed in additional File 1. (A) Phylogenetic analyses of RB family, with RB-Mb (gi|167523296|) being used as the outgroup; (B) Phylogenetic analyses of DP family, with DP-Mb (gi:167516980) being used as the outgroup.

Figure 5

E2F4/5 subgroup evolved slower than E2F1/2/3 subgroup. (A) Distance between Human E2F proteins and predicted ancestral E2F sequence. The pairwise distances between human E2Fs and predicted ancestral E2F sequence, E2F-Mb(gi|167523471). (B) The group average pairwise distance of E2F4/5 subgroup and E2F1/2/3 subgroup in invertebrates. Based on the sequence data from Ta, Nv, Ci. Sp and Bf, the average distance of E2F4/5 subgroup is smaller than E2F1/2/3 subgroup (P < 0.001, with Pair Student's T Test), standard deviation values are showed as error bars. (C) Distance between human E2F proteins and other vertebrate orthologues. Pair wise distances between human E2F proteins and their orthologues from Gg, Xt. Dr, Tn were computed respectively. (D) The group average distances of E2F4, E2F1, and E2F3 in vertebrates. Based on protein sequence data from 8 organisms (Hs, Mm, Cf, Bt, Rn, Gg, Xt, Tn), the group average distance of E2F4 is smaller than E2F1 and E2F3 (E2F1:E2F4, P < 0.001; E2F3:E2F4, P < 0.001). (E) Ka/Ks values for E2F genes based on codon sequences from human and mouse. (F) The group average Ka/ks of E2F1, E2F2, E2F3, and E2F4 in mammals. Based on Ka/ks values for protein-coding DNA sequences from four mammals (Hs, Cf, Bt, Mm) in each subgroup, The group average Ka/ks value of E2F4 subgroup is smaller than E2F1 subgroup, E2F2 subgroup and E2F3 subgroup (For E2F1:E2F4, P < 0.001; E2F2:E2F4, P < 0.001; E2F3:E2F4, P < 0.05.). Detail information could be found in additional file 5 and additional file 6. Abbreviations: Hs, H. sapiens; Mm, M. musculus; Cf, C. familiaris; Bt, B. Taurus; Gg, G. gallus; Xt, X. tropicalis; Dr, D. rerio; Tn, T. nigroviridis; Ci, C. intestinalis; Sp, S. purpuratus; Bf, B. floridae, Ta, T. adhaerens.

Figure 6

RBL subgroup evolved slightly slower than RB1 subgroup in vertebrates. (A) Distance between Human RB family proteins and predicted ancestral RB sequence. The pair wise distances between human RB1, RBL1, RBL2 and predicted ancestral RB sequence, RB-Mb (gi:167523296) sequence were computed. (B) Distance between human RB family proteins and their vertebrate orthologues. Pair wise distances were calculated between human RB1, RBL1, RBL2 and their orthologues in Gg, Xt, Dr, Tn, respectively. (C) The group average pairwise distances RB family in vertebrates. Based on protein sequences from 7 organisms (Hs, Mm, Cf, Bt, Rn, Gg, and Dr), the group average distance of RBL1 subgroup and RBL2 subgroup is smaller than RB1 subgroup (For RB1: RBL1, P < 0.001; RB1:RBL2, P < 0.001; with pair Student's T test), standard deviation values are showed as error bars. (D) Ka/Ks values for RB family genes deduced from sequences comparisons between human and other mammals. The ratio of nonsynonymous (Ka) to synonymous substitutions (Ks) of RB1, RBl1 RBL2 genes were computed, based on codoning sequences pairwise comparisons (Hs vs Mm, Hs vs Rn, hs vs Cf, hs vs Bt) respectively. (E) Average Ka/ks values for RB family gene in mammals. Based on the Ka/ks values for all protein-coding DNA sequences from 5 mammals (Hs, Cf, Bt, Mm, and Rn) in each subgroup, The group average Ka/ks value of RBL1 subgroup and RBL2 subgroup is smaller than RB1 subgroup with statistical support (For RBL1:RB1, P < 0.01; RBL2:RB1, P < 0.01), Detail information could be found in additional file 5 and additional file 6. Abbreviations used: Hs, H. sapiens; Mm, M. musculus; Cf, C. familiaris; Bt, B. Taurus; Gg, G. gallus; Xt, X tropicalis; Dr, D. rerio; Tn, T. nigroviridis; Ci, C. intestinalis; Sp, S. purpuratus; Bf, B. floridae.

Figure 7

Overview of the evolutionary history of RB-E2F pathway in metazoan. (A) Summaries of the gene duplication events for E2F, RB and DP proteins in metazoa. The proteins of E2F and RB family from fly and worm were also tried to be classed into suitable subgroup based one sequence and functional data. (B) A simple summary of the evolutionary history of RB-E2F pathway. The protein interaction information of RB-E2F pathway in vertebrates was labeled with black arrows, mainly based on the recent literature [2]. The protein interaction information in invertebrates and some unicellular organisms were predicted by interaction data of their vertebrate orthologues, and were labeled with grey arrows.