Evolutionary history of selenocysteine incorporation from the perspective of SECIS binding proteins

Abstract

Background: The co-translational incorporation of selenocysteine into nascent polypeptides by recoding the UGA stop codon occurs in all domains of life. In eukaryotes, this event requires at least three specific factors: SECIS binding protein 2 (SBP2), a specific translation elongation factor (eEFSec), selenocysteinyl tRNA, and a cis-acting selenocysteine insertion sequence (SECIS) element in selenoprotein mRNAs. While the phylogenetic relationships of selenoprotein families and the evolution of selenocysteine usage are well documented, the evolutionary history of SECIS binding proteins has not been explored.

Results: In this report we present a phylogeny of the eukaryotic SECIS binding protein family which includes SBP2 and a related protein we herein term SBP2L. Here we show that SBP2L is an SBP2 paralogue in vertebrates and is the only form of SECIS binding protein in invertebrate deuterostomes, suggesting a key role in Sec incorporation in these organisms, but an SBP2/SBP2L fusion protein is unable to support Sec incorporation in vitro. An in-depth phylogenetic analysis of the conserved L7Ae RNA binding domain suggests an ancestral relationship with ribosomal protein L30. In addition, we describe the emergence of a motif upstream of the SBP2 RNA binding domain that shares significant similarity with a motif within the pseudouridine synthase Cbf5.

Conclusion: Our analysis suggests that SECIS binding proteins arose once in evolution but diverged significantly in multiple lineages. In addition, likely due to a gene duplication event in the early vertebrate lineage, SBP2 and SBP2L are paralogous in vertebrates.

Figure 1

SECIS binding protein topology. Alignment of SBP2 and SBP2L across the species indicated generated with the MUSCLE module in Geneious (Biomatters Ltd). The N and C-terminal portions of SBP2 and SBP2L were independently aligned. Residues were colored using JalView based on BLOSUM62 score [[37]; darker colors indicate higher similarity]. The SBP2L N-terminal sequence is red to denote that it is a separate alignment from that for SBP2. Globally conserved motifs motifs include the Sec incorporation domain (SID), the RNA binding domain (RBD), LSAD^15-26 and PFVQ^44-56. Shading of species names is used for the identification of SBP2 classes as described in the text.

Figure 2

Phylogeny of SECIS binding proteins across eukaryotic taxa. SECIS binding proteins from the indicated taxa were aligned with MUSCLE and non-conserved regions were removed with GBlocks. The resulting alignment was used to infer a maximum likelihood tree with PhyML with 500 bootstrap replicates. Bootstrap values less than 0.5 are not shown.

Figure 3

The SBP SID and RBD are highly conserved across Eukarya. (A) An extraction of the SBP2 and SBP2L SID region from the global multiple sequence alignment described for Figure 2 prior to GBlocks treatment. The non-conserved region between 464-505 (human SBP2 numbering) was deleted. (B) An extraction of the SBP2 and SBP2L RBD region as in (A). Nonconserved sequences from M. brevicollis, E. huxleyi, and T. gondii that introduced large gaps in the region were deleted. In addition the sequence from S. purpuratus was omitted from this alignment due the presence of several large non-conserved regions that disrupted the entire global alignment. Positions that are consistently variable between SBP2 and SBP2L are indicated with red arrows and the identity/number refers to the position in human SBP2. Underlined sequence corresponds to the GBlocks output that was used to generate the tree in Figure 2. Blue shaded boxes indicate conserved motifs that are only sporadically found in unicellular organisms.

Figure 4

Identification of conserved motifs common to SBP2 and SBP2L. (A) Global Alignment of complete sequences for SBP2L and vertebrate SBP2 from the taxa indicated in (B) generated with MUSCLE. Conserved residues are colored according to Blosum62 score with red denoting 100% conservation, black 80-99% identical, dark grey 60-80% identical, light grey less than 60% identical. Conserved motifs are shaded by colored boxes. (B) Conserved motifs in SBP2L and SBP2 marked by colored boxes in panel A are shown in detail.

Figure 5

Identification of conserved motifs within SBP2 and SBP2L. (A) Global alignment of vertebrate SBP2 sequences (top) with detailed alignment of conserved motifs as indicated (bottom). (B) Global and detailed alignments of SBP2L as in (A).

Figure 6

The SID-like region in SBP2L does not promote Sec incorporation. (A) Schematic of the SBP2L/SBP2 domain swap. (B) 2 fmol of the indicated [³⁵S]-Met labeled in vitro translated proteins (top gel) were added to an in vitro translation reaction containing a luciferase Sec incorporation reporter. Luciferase activity (Sec incorporation) is expressed as a percent of that obtained with wild-type CT-SBP2 (bottom graph). (C) 8 fmol of [³⁵S]-Met labeled in vitro translated proteins were incubated with [³²P]-labeled wild-type (wt) or mutant (mt) GPX4 SECIS elements and resolved on a 4% non-denaturing polyacrylamide gel. Arrow 1 marks CT-SBP2L and SBP2-SBP2L complexes, arrow 2 marks CT-SPB2 and SBP2L-SBP2 complexes. The asterisk marks a probe shift resulting from an unidentified component in rabbit reticulocyte lysate. (D) Graphical representation of EMSA data shown in (D) expressed as the percent of probe shifted relative to that obtained with wild-type CT-SBP2. (E) 2.4 fmol of the indicated [³⁵S]-Met labeled in vitro translated proteins were added to an in vitro translation reaction containing a luciferase Sec incorporation reporter. Mock contains no added in vitrotranslated proteins. 'CT-SBP2L subs.' and 'SBP2L-SBP2 subs.' indicate proteins bearing the following substitutions: D494G, SKA556PLM, EK563QR and A567P (human SBP2L numbering). Sec incorporation is expressed as a percentage of CT-SBP2. All data are mean ± SEM of three independent experiments.

Figure 7

Phylogenetic tree of L7Ae core motifs. A phylogenetic tree of core motifs from L7Ae family members was inferred by maximum likelihood methods using PhyML with 500 bootstrap replicates. The tree was re-rooted with the Gadd45/RNaseP-P38 clade as the outgroup. Bootstrap values less than 0.5 are not shown. Scale bar represents amino acid substitutions per site.

Figure 8

Convergence of a short motif in Cbf5 and SECIS binding proteins. Maximum likelihood tree of Cbf5 and SECIS binding proteins inferred using PhyML with 1000 bootstraps. Bootstrap values below 0.5 are not show. The alignment used to generate the tree is shown to the right.

Figure 9

Aedes aegypti SBP2 is coded by two genes. (A) Multiple sequence alignment of insect SBP2 from the indicated species. Residues were colored using JalView based on BLOSUM62 score. The black line above the alignment denotes EST coverage of residues in the predicted A. aegypti SID peptide that are in an EST corresponding to the A. aegypti RBD. (B) Maximum likelihood tree of the insect SBP2 sequences used in panel A. In order to build the tree, the A. aegypti SID and RBD sequences were combined into one sequence file. (C) The genomic contexts of SBP2 from A. aegypti and C. quinquefasciatus and D. melanogaster were ascertained from Entrez Gene at NCBI. Arrows indicate gene orientation and gene names are as indicated.