Evolution of lysine acetylation in the RNA polymerase II C-terminal domain

Abstract

Background: RPB1, the largest subunit of RNA polymerase II, contains a highly modifiable C-terminal domain (CTD) that consists of variations of a consensus heptad repeat sequence (Y1S2P3T4S5P6S7). The consensus CTD repeat motif and tandem organization represent the ancestral state of eukaryotic RPB1, but across eukaryotes CTDs show considerable diversity in repeat organization and sequence content. These differences may reflect lineage-specific CTD functions mediated by protein interactions. Mammalian CTDs contain eight non-consensus repeats with a lysine in the seventh position (K7). Posttranslational acetylation of these sites was recently shown to be required for proper polymerase pausing and regulation of two growth factor-regulated genes.

Results: To investigate the origins and function of RPB1 CTD acetylation (acRPB1), we computationally reconstructed the evolution of the CTD repeat sequence across eukaryotes and analyzed the evolution and function of genes dysregulated when acRPB1 is disrupted. Modeling the evolutionary dynamics of CTD repeat count and sequence content across diverse eukaryotes revealed an expansion of the CTD in the ancestors of Metazoa. The new CTD repeats introduced the potential for acRPB1 due to the appearance of distal repeats with lysine at position seven. This was followed by a further increase in the number of lysine-containing repeats in developmentally complex clades like Deuterostomia. Mouse genes enriched for acRPB1 occupancy at their promoters and genes with significant expression changes when acRPB1 is disrupted are enriched for several functions, such as growth factor response, gene regulation, cellular adhesion, and vascular development. Genes occupied and regulated by acRPB1 show significant enrichment for evolutionary origins in the early history of eukaryotes through early vertebrates.

Conclusions: Our combined functional and evolutionary analyses show that RPB1 CTD acetylation was possible in the early history of animals, and that the K7 content of the CTD expanded in specific developmentally complex metazoan lineages. The functional analysis of genes regulated by acRPB1 highlight functions involved in the origin of and diversification of complex Metazoa. This suggests that acRPB1 may have played a role in the success of animals.

Figure 1

The human RNA polymerase II subunit 1 (RPB1) C-terminal domain (CTD) contains more heptad repeats than the yeasts, and eight of its non-consensus distal repeats have a lysine residue. In this schematic of the RPB1 CTD for two species of yeast and human, consensus heptad repeats (YSPTSPS) are colored dark gray; repeats with a lysine at position 7 are colored red; and all other non-consensus repeats are in white.

Figure 2

Lysine-containing CTD repeats first appeared in Metazoa and increased in prevalence in the ancestor of Deuterostomia. Phylogenetic tree of eukaryotic species considered in our analysis organized by approximate divergence estimates. For each species, the number of RPB1 CTD repeats and lysine-containing repeats are given. Ancestral counts were inferred for each internal node of the tree using symmetric Wagner parsimony. The number of CTD heptad repeats increased substantially in the ancestor of all Metazoa (31 to 44 repeats). This was accompanied by the appearance of repeats with lysine at position seven (K₇) in ancestral Metazoa and an increase in K₇ repeats in Deuterostomia (from 3 to 7 repeats).

Figure 3

Sequence logos summarizing repeat sequence variation in Fungi, Ecdysozoa, and Deuterostomia. Sequence logos were generated from all heptad repeat sequences of relevant species using the LogOddsLogo tool [41]. Most repeat positions are strongly conserved, but different clades exhibit different variations. Lysine is particularly common at position seven in Deuterostomia. See Additional file 1 for RPB1 sequences for all species examined.

Figure 4

Genes regulated by acRPB1 are significantly enriched for evolutionary origins from early eukaryotes through early animals. ProteinHistorian analysis of the evolutionary origins of acRPB1 enriched genes (A) and acRPB1 dysregulated genes (B) compared with relevant background gene sets. Each bar gives the difference between the proportion of genes of interest and background genes originating in the last common ancestor of a given taxon (* = p < 0.05, ** = p < 0.01, *** = p < 0.001, all Bonferroni corrected). The proportion of genes of interest with origins in each taxon is given along the x-axis. For example, 18% of the acRPB1 enriched genes likely appeared between the origin of Opisthokonta and Bilateria, and this is significantly more (~7%, p < 0.001) than expected from the background of all genes with occupancy data.

Figure 5

Summary of the origins of subsets of genes regulated by acRPB1 with respect to relevant evolutionary events. The period between K₇ repeat expansion and the diversification of vertebrates, shows particular enrichment for the origin of acRPB1 regulated genes and genes with functions relevant to animals. Each column in the heat map represents the evolutionary origin distribution of a set of genes. Color intensity reflects the magnitude of increase over background in each evolutionary window. Due to the large difference for some gene sets, two scales were necessary (red and purple). Asterisks indicate significant increase over background (Bonferroni-corrected p < 0.05). Relevant evolutionary events are identified with blue triangles. For reference, the first two columns represent the acRPB1 enriched and dysregulated origin distributions (Figure 4). The remaining columns give the origin distributions for functional subsets of genes enriched among acRPB1 sensitive genes.