Structural analysis and evolution of specificity of the SUMO UFD E1-E2 interactions

Abstract

SUMO belongs to the ubiquitin-like family (UbL) of protein modifiers. SUMO is conserved among eukaryotes and is essential for the regulation of processes such as DNA damage repair, transcription, DNA replication and mitosis. UbL modification of proteins occurs via a specific enzymatic cascade formed by the crosstalk between the E1-activating enzyme, the E2-conjugating enzyme and the E3-ligase. An essential discrimination step in all UbL modifiers corresponds to the interaction between E1 and E2 enzymes, which is mediated by the recruitment of the E2 to the UFD domain (Ubiquitin-Fold Domain) of the E1 enzyme. To gain insights in the properties of this interface, we have compared the structures of the complexes between E1 UFD domain and E2 in human and yeast, revealing two alternative UFD platforms that interact with a conserved E2. Comparative sequence analysis of the E1 UFD domain indicates that the E2 binding region has been conserved across phylogenetic closely related species, in which higher sequence conservation can be found in the E2 binding region than in the entire UFD domain. These distinctive strategies for E1-E2 interactions through the UFD domain might be the consequence of a high selective pressure to ensure specificity of each modifier conjugation system.

Figure 1. Structural alignment and comparison of the interfaces between UFD and Ubc9 from human and yeast complexes.

(a) Structural alignment of the UFD domains from yeast (ScUba2) and human (HsUba2). Red circles indicate contact residues to Ubc9. Dotted rectangle represents the binding region to Ubc9. (b) Structural alignment of yeast and human Ubc9. Small red arrows indicate contact residues to UFD. Dotted rectangle represents the binding region to UFD. Secondary structure is depicted above sequence. (c) Left, ribbon representation of the complex of human Ubc9 and the UFD domain. Right, stereo representation of the interface residues between Ubc9 (green line) and UFD (purple ribbon). Major contacts are labeled and represented in stick configuration. (d) Left, ribbon representation of the complex of yeast Ubc9 and the UFD domain. Right, stereo representation of the interface residues between Ubc9 (yellow line) and UFD (blue ribbon). Major contacts are labeled and represented in stick configuration. (e) Structural superposition of Ubc9 in the human and yeast complex with UFD.

Figure 2. Comparison of the UFD-E2 interface from different UbL systems.

(a) Transparent electrostatic representation of the interface of human SUMO E2 (hSUMO Ubc9) with the UFD domain. Major contacts are labeled and represented in stick configuration. Basic and aliphatic surface patches are indicated by dotted circles. (b) Transparent electrostatic representation of the interface of yeast SUMO E2 (ySUMO Ubc9) with the UFD domain. Major contacts are labeled and represented in stick configuration. Basic and aliphatic surface patches are indicated by dotted circles. (c) Transparent electrostatic representation of the interface of S.pombe ubiquitin E2 (SpUb Ubc4) with the UFD domain. Major contacts are labeled and represented in stick configuration. (d) Transparent electrostatic representation of the interface of human Nedd8 E2 (hNedd8 Ubc12) with the UFD domain. Major contacts are labeled and represented in stick configuration. (e) Schematic representation of the human SUMO E1 UFD domain contacts with the E2 enzyme. (f) Schematic representation of the yeast SUMO E1 UFD domain contacts with the E2 enzyme. (g) Schematic representation of the S.pombe ubiquitin E1 UFD domain contacts with the E2 enzyme. (h) Schematic representation of the human Nedd8 E1 UFD domain contacts with the E2 enzyme. Black and grey spots indicate the orientation of the side chain in the structure regarding the β-sheet plane.

Figure 3. Conservation analysis of Sae2 LHEB2 domain in metazoan and Saccharomycetales.

Amino acid sequence alignment of Sae2 LHEB2 domain orthologs from nematoda (a), arthopoda (b), chordate (c) and saccharomycetales (d). Residue shading correspond to 95% (white letter and dark background), 75% (white letter and gray background), and 55% (black letter and light gray background) of sequence identity in (a,b and c). In the case of saccharomycetales, the shading types correspond to 90%, 70% and 50% of sequence identity, respectively. Metazoan multiple sequence alignments of LHEB2 sequences were performed using Clustal Omega software and including yeast Uba2 (NP_010678) as outlier. Saccharomycetales multiple sequence alignments of LHEB2 sequences were performed using Muscle software and including human Sae2 (Q9UBT2) as outlier. (e) Graphical representation of LHEB2 domain consensus sequences determined from amino acid sequence alignments shown in (a,b,c and d). The overall height of the stack indicates the sequence conservation at that position, while the symbol height within the stack indicates the relative frequency of each amino acid within that position. The positions of yeast Sae2 residues involved in Ubc9 interaction according to the previously resolved structure (3ONG) are indicated in blue below the sequences graph. Asterisks indicate residues shown to have a major contribution to E1-E2 interactions in mutagenesis analysis. The positions of human Sae2 residues involved in Ubc9 interactions according to the resolved structure are indicated in blue below the chordata consensus sequence graph. Grey circles indicate residues establishing contacts with Ubc9 α1-helix, while grey triangles indicate residues interacting with Ubc9 residues located at the Ubc9 β1β2-loop. Conserved residues across phyla are indicated by lines. (f) Distribution of LHEB2 sequence length displayed by orthologs within each phylogenetic group analyzed was plotted on a box plot graph. Data points are represented by circles. Outliers are represented by dots. The number of data points analyzed in each phylogenetic group is indicated below the x-axis.

Figure 4. Phylogenetic analysis of UFD and LHEB2 domains from Metazoa and Saccharomycetales.

Maximum likelihood phylogenetic trees depicting the evolutionary relationships among 86 Sae2 UFD (a) or Sae2 LHEB2 (b) domain sequences from 68 metazoa species species using sequence alignments shown in Fig. S3 (UFD) and Fig. 3 (LHEB2). Sequences belonging to the same phylum are enclosed in colored areas. Tree scales are shown below each tree.