Human eIF5 and eIF1A Compete for Binding to eIF5B

Abstract

Eukaryotic translation initiation is a multistep process requiring a number of eukaryotic translation initiation factors (eIFs). Two GTPases play key roles in the process. eIF2 brings the initiator Met-tRNA_i to the preinitiation complex (PIC). Upon start codon selection and GTP hydrolysis promoted by the GTPase-activating protein (GAP) eIF5, eIF2-GDP is displaced from Met-tRNA_i by eIF5B-GTP and is released in complex with eIF5. eIF5B promotes ribosomal subunit joining, with the help of eIF1A. Upon subunit joining, eIF5B hydrolyzes GTP and is released together with eIF1A. We found that human eIF5 interacts with eIF5B and may help recruit eIF5B to the PIC. An eIF5B-binding motif was identified at the C-terminus of eIF5, similar to that found in eIF1A. Indeed, eIF5 competes with eIF1A for binding and has an ∼100-fold higher affinity for eIF5B. Because eIF5 is the GAP of eIF2, the newly discovered interaction offers a possible mechanism for coordination between the two steps in translation initiation controlled by GTPases: start codon selection and ribosomal subunit joining. Our results indicate that in humans, eIF5B displacing eIF2 from Met-tRNA_i upon subunit joining may be coupled to eIF1A displacing eIF5 from eIF5B, allowing the eIF5:eIF2-GDP complex to leave the ribosome.

Figure 1.. Domain structure and interactions of eIF5B, eIF5, and eIF1A.

(A) Top, domain structure of eIF5B. The binding site for eIF1A-CTT and eIF5-CTT is labeled. Bottom, constructs used in this work. (B) Top, domain structure of eIF5. The binding site for eIF5B-D4 is labeled. Bottom, constructs used in this work. (C) Top, domain structure of eIF1A. The binding site for eIF5B-D4 is labeled. Bottom, construct used in this work.

Figure 2.. Binding affinities between eIF5 and eIF5B fragments determined by fluorescence anisotropy (FA).

(A) Direct titration of fluorescein-labeled 7-residue eIF5 C-terminal peptide (Fl-eIF5-CT7) with eIF5B domain 4 (eIF5B-D4) and eIF5B domains 3 and 4 (eIF5B-D34). The calculated K_Ds are shown in the inset. (B) Competition assay with eIF5-CT9 and eIF5-CT39. The calculated K_Ds are shown in the inset.

Figure 3.. The C-termini of eIF1A and eIF5 are conserved.

Sequence alignment of eIF1A (top) and eIF5 (bottom) C-terminal tail (CTT) sequences from a select set of species. Hydrophobic residues are green; negatively charged residues are red, and positively charged residues are blue. The C-termini of the proteins are marked with *. Locations of intervening sequences not included in the alignment are marked with //. The eIF5B-binding motifs at the C-terminus of human and S. cerevisiae eIF1A, and human eIF5 (this work) are marked with a line. The eIF5B-binding motif is conserved in eIF5 from Metazoa, as well as fungi from subphylum Ustilaginomycotina. It is not found in eIF5 from other fungal species, including budding and fission yeast, or in plant eIF5. The C-terminus of metazoan eIF5 appears more similar to that of S. cerevisiae eIF1A than human eIF1A. eIF1A-CTT and eIF5-CTT from Ustilaginomycotina fungi appear more similar to each other than to eIF1A or eIF5 sequences from other species. The C-terminal sequence of C. elegans eIF1A (in lower-case and highlighted gray) does not appear to contain an eIF5B-binding motif.

Figure 4. Comparison of the eIF5B•eIF5 and eIF5B•eIF1A binding interfaces

(A) Comparison of eIF5B-D4 binding to eIF5-CTT and eIF1A. Left, region of the ¹⁵N-HSQC spectra of GB1-tagged eIF5-CT39 (GH-eIF5-CT39) in the absence (black) and presence (red) of unlabeled eIF5B-D4. The C-terminal six residues of eIF5 are labeled and the changes in their positions upon binding of eIF5B-D4 are marked with dashed arrows. Right, eIF5B-D4 induced chemical shift perturbation (CSP) mapped on the sequence of eIF5-CTT (top) and eIF1A-CTT (bottom, from reference ). Only the last 10 residues of eIF1A and the last 33 residues of eIF5 are shown, because there were no detectable CSP effects on the rest of the two tails. Residues experiencing medium (>0.03 ppm) and large (>0.1 ppm) CSP effects are marked with * and **, respectively. (B) eIF5-CTT binding surface of eIF5B-D4. eIF5-CTT induced CSPs were mapped on the structure of human eIF5B-D4, from light yellow (weak effects) to red (strong effects). Residues that could not be analyzed are light grey. (C) eIF1A-CTT binding surface of eIF5B-D4 from reference 6 shown for comparison. Coloring is as in panel B. (D) Model of the human eIF5B-D4•eIF5-CTT complex, in cross-eye stereo, based on the yeast eIF5B-D4•eIF1A-CTT complex. eIF5B-D4 is light blue, and eIF5-CTT is gold. Sidechains exhibiting intermolecular NOEs between eIF5 and eIF5B, are shown as blue (eIF5B) and red (eIF5) sticks, and the corresponding residues are labeled.

Figure 5.. Model for the dynamic interactions of human eIF5 and eIF1A with eIF5B.

eIF5 is navy. The coloring of eIF5B is from yellow (D1) to dark orange (D4); the N-terminal region of eIF5B is not shown. eIF1A-OB is light blue; eIF1A-CTT is blue; and eIF1A-NTT is not shown. eIF2-GTP is magenta; eIF2-GDP is purple. The 40S ribosomal subunit is gray. The 60S ribosomal subunit is shown as an outline. The alternative pathways for eIF5B recruitment to the 43S PIC are marked with dashed arrows. eIF5B may either bind directly to the PIC through interactions with eIF5, the 40S subunit, and eIF1A-OB, or be recruited in complex with eIF5, possibly as a component of the multifactor complex (MFC). In the scanning PIC, eIF1A-CTT extends into the P-site, where it stabilizes the open complex. Start codon recognition promotes the closed, scanning-incompetent conformation of the PIC; eIF1A-CTT is displaced from the P-site^– and in turn displaces eIF5-CTT from eIF5B-D4. This process is likely coupled to eIF5B displacing eIF2-GDP from the Met-tRNA_i, which leads to release of the eIF5•eIF2-GDP complex from the 48S PIC.