Cap-free structure of eIF4E suggests a basis for conformational regulation by its ligands
2006; Springer Nature; Volume: 25; Issue: 21 Linguagem: Inglês
10.1038/sj.emboj.7601380
ISSN1460-2075
AutoresLaurent Volpon, Michael J. Osborne, Ivan Topisirović, Nadeem Siddiqui, Katherine L. B. Borden,
Tópico(s)Mast cells and histamine
ResumoArticle12 October 2006free access Cap-free structure of eIF4E suggests a basis for conformational regulation by its ligands Laurent Volpon Laurent Volpon Department of Pathology and Cell Biology, Institute for Research in Immunology and Cancer (IRIC), Université de Montréal, Pavillion Marcelle-Coutu, Chemin Polytechnique, Montréal, QC, Canada Search for more papers by this author Michael J Osborne Michael J Osborne Department of Pathology and Cell Biology, Institute for Research in Immunology and Cancer (IRIC), Université de Montréal, Pavillion Marcelle-Coutu, Chemin Polytechnique, Montréal, QC, Canada Search for more papers by this author Ivan Topisirovic Ivan Topisirovic Department of Pathology and Cell Biology, Institute for Research in Immunology and Cancer (IRIC), Université de Montréal, Pavillion Marcelle-Coutu, Chemin Polytechnique, Montréal, QC, Canada Search for more papers by this author Nadeem Siddiqui Nadeem Siddiqui Department of Pathology and Cell Biology, Institute for Research in Immunology and Cancer (IRIC), Université de Montréal, Pavillion Marcelle-Coutu, Chemin Polytechnique, Montréal, QC, Canada Search for more papers by this author Katherine LB Borden Corresponding Author Katherine LB Borden Department of Pathology and Cell Biology, Institute for Research in Immunology and Cancer (IRIC), Université de Montréal, Pavillion Marcelle-Coutu, Chemin Polytechnique, Montréal, QC, Canada Search for more papers by this author Laurent Volpon Laurent Volpon Department of Pathology and Cell Biology, Institute for Research in Immunology and Cancer (IRIC), Université de Montréal, Pavillion Marcelle-Coutu, Chemin Polytechnique, Montréal, QC, Canada Search for more papers by this author Michael J Osborne Michael J Osborne Department of Pathology and Cell Biology, Institute for Research in Immunology and Cancer (IRIC), Université de Montréal, Pavillion Marcelle-Coutu, Chemin Polytechnique, Montréal, QC, Canada Search for more papers by this author Ivan Topisirovic Ivan Topisirovic Department of Pathology and Cell Biology, Institute for Research in Immunology and Cancer (IRIC), Université de Montréal, Pavillion Marcelle-Coutu, Chemin Polytechnique, Montréal, QC, Canada Search for more papers by this author Nadeem Siddiqui Nadeem Siddiqui Department of Pathology and Cell Biology, Institute for Research in Immunology and Cancer (IRIC), Université de Montréal, Pavillion Marcelle-Coutu, Chemin Polytechnique, Montréal, QC, Canada Search for more papers by this author Katherine LB Borden Corresponding Author Katherine LB Borden Department of Pathology and Cell Biology, Institute for Research in Immunology and Cancer (IRIC), Université de Montréal, Pavillion Marcelle-Coutu, Chemin Polytechnique, Montréal, QC, Canada Search for more papers by this author Author Information Laurent Volpon1, Michael J Osborne1, Ivan Topisirovic1, Nadeem Siddiqui1 and Katherine LB Borden 1 1Department of Pathology and Cell Biology, Institute for Research in Immunology and Cancer (IRIC), Université de Montréal, Pavillion Marcelle-Coutu, Chemin Polytechnique, Montréal, QC, Canada *Corresponding author. Department of Pathology and Cell Biology, Institute for Research in Immunology and Cancer (IRIC), Université de Montréal, Pavillion Marcelle-Coutu, 2950, Chemin Polytechnique, Montréal, Québec, Canada H3T 1J4. Tel.: +1 514 343 6291; Fax: +1 514 343 5839; E-mail: [email protected] The EMBO Journal (2006)25:5138-5149https://doi.org/10.1038/sj.emboj.7601380 PDFDownload PDF of article text and main figures. ToolsAdd to favoritesDownload CitationsTrack CitationsPermissions ShareFacebookTwitterLinked InMendeleyWechatReddit Figures & Info The activity of the eukaryotic translation initiation factor eIF4E is modulated through conformational response to its ligands. For example, eIF4G and eIF4E-binding proteins (4E-BPs) modulate cap affinity, and thus physiological activity of eIF4E, by binding a site distal to the 7-methylguanosine cap-binding site. Further, cap binding substantially modulates eIF4E's affinity for eIF4G and the 4E-BPs. To date, only cap-bound eIF4E structures were reported. In the absence of structural information on the apo form, the molecular underpinnings of this conformational response mechanism cannot be established. We report here the first cap-free eIF4E structure. Apo-eIF4E exhibits structural differences in the cap-binding site and dorsal surface relative to cap-eIF4E. Analysis of structure and dynamics of apo-eIF4E, and changes observed upon ligand binding, reveal a molecular basis for eIF4E's conformational response to these ligands. In particular, alterations in the S4-H4 loop, distal to either the cap or eIF4G binding sites, appear key to modulating these effects. Mutation in this loop mimics these effects. Overall, our studies have important implications for the regulation of eIF4E. Introduction Normal development, differentiation and cellular growth rely on post-transcriptional control of gene expression. The eukaryotic translation initiation factor eIF4E regulates gene expression post-transcriptionally at multiple levels, including mRNA translation and mRNA export (Pestova et al, 2001; von der Haar et al, 2004). Through these activities, eIF4E promotes cellular proliferation, growth and survival. Even moderate overexpression of eIF4E leads to dysregulated cellular proliferation +and malignant transformation (Graff and Zimmer, 2003). eIF4E associates with the 5′ end of mRNAs via the 7-methyl guanosine dinucleotide cap structure, m7GpppN (7-methyl-5′-guanosinetriphosphate-5′-N) (where N is any nucleotide) (Sonenberg and Gingras, 1998). The association of eIF4E with mRNA via the cap is linked to its biochemical and subsequent growth promoting and transforming activities. eIF4E overexpression does not uniformly increase synthesis for all proteins; targeting those with complex 5′ untranslated regions (UTRs) (Sonenberg and Gingras, 1998). The Sonenberg group first showed that eIF4E is organized into nuclear bodies in addition to its cytoplasmic localization (Rousseau et al, 1996). eIF4E forms nuclear bodies in a variety of cell types and organisms including yeast, Drosophila, Xenopus, mouse and human (Iborra and Cook, 2002). Up to 68% of total cellular eIF4E is found in the nucleus (Iborra et al, 2001; Strudwick and Borden, 2002; Culjkovic et al, 2005). The best described nuclear function involves nuclear mRNA export of selected transcripts (Rousseau et al, 1996). This level of regulation requires an ∼100 nucleotide sequence denoted the eIF4E sensitivity element (4E-SE) in the 3′UTR of sensitive transcripts (Culjkovic et al, 2005). Importantly, eIF4E requires its cap binding activity for its nuclear mRNA export function. For instance in the nucleus, the promyelocytic leukemia (PML) protein directly binds eIF4E, suppresses its mRNA export and transformation functions (Cohen et al, 2001). Mutagenesis studies indicate that both its mRNA export as well as its translation functions contribute to the ability of eIF4E to transform cells (Cohen et al, 2001; Topisirovic et al, 2003a, 2003b; Culjkovic et al, 2005). eIF4E requires its cap binding activity to act in mRNA translation, mRNA nuclear export and in order to promote oncogenic transformation (Sonenberg and Gingras, 1998; Cohen et al, 2001; Topisirovic et al, 2003a, 2003b). The functions of eIF4E are tightly regulated in the cell through interactions with partner proteins. In translation, eIF4E binds the RNA through the 7-methylguanosine (m7G) cap structure and subsequently other components of the eIF4F complex including eIF4G (Gingras et al, 1999). eIF4G binds the dorsal surface of eIF4E, which is located distal to the cap binding site (Figure 1). The same site used for eIF4G is also bound by the translational inhibitors, the eIF4E binding proteins (4E-BPs). The 4E-BPs inhibit translation by sterically blocking eIF4G binding (Ptushkina et al, 1999). Importantly, both eIF4G and 4E-BP association with eIF4E leads to enhanced association of eIF4E with the m7G cap (von der Haar et al, 2004). This ligand binding leads to a conformational response by eIF4E. In this way, 4E-BP/eIF4E complexes can trap mRNAs making 4E-BP a potent inhibitor of translation by, in essence, sequestering the transcripts in inactive complexes (von der Haar et al, 2004). Theoretically, this effect on cap binding leads to eIF4G more effectively promoting translation. Importantly, many other regulators of eIF4E bind its dorsal surface including up to 200 homeoproteins many of which regulate nuclear and cytoplasmic activities of eIF4E (Topisirovic et al, 2003a). These proteins, like eIF4G and the 4E-BPs, contain a conserved eIF4E binding site defined as YXXXXLΦ where X is any residue and Φ is any hydrophobic. PML, a potent negative regulator of eIF4E in the nucleus, uses its RING domain to bind the dorsal surface. This leads to a reduction in the affinity of eIF4E for the cap by over 100-fold (Cohen et al, 2001; Kentsis et al, 2001). Thus, eIF4E is regulated through conformational rearrangements by a wide variety of protein partners and this results in modulation of the biological activities of eIF4E. Figure 1.Crystal structure of the murine eIF4E/7-methyl-GDP–eIF4G peptide ternary complex (PDB 1EJH). Download figure Download PowerPoint Detailed structural studies on eIF4E complexed to various cap analogs and to eIF4G and the 4E-BPs have led to important advances in our molecular understanding of eIF4E function (von der Haar et al, 2004). High-resolution crystal studies delineate the residues important to cap binding (Marcotrigiano et al, 1997; Niedzwiecka et al, 2002; Tomoo et al, 2002). It is well established that the specificity of cap recognition is imparted by the m7G moiety, which is stacked between two tryptophan residues, W56 and W102 (Sonenberg and Gingras, 1998). Also, the delocalization of the positive charge on m7G, due to the methylation of N7, promotes interaction with an acidic cavity (formed by E103) in eIF4E (Marcotrigiano et al, 1997). While the m7G moiety alone binds to eIF4E, addition of phosphate groups significantly enhances affinity (Zuberek et al, 2004). Several biophysical studies highlight the conformational linkage between cap binding and the association of partner proteins on the distal dorsal surface of eIF4E. We will refer to this as conformational response or rearrangement. For example, a 100 residue eIF4G fragment increases the affinity of yeast eIF4E for the cap by 10-fold (Ptushkina et al, 1999) whereas studies with a minimal eIF4G peptide for human and mouse eIF4E increase affinity by about two-fold (Friedland et al, 2005). Conversely, cap binding enhances affinity of eIF4E for the eIF4G peptide and 4E-BPs (Shen et al, 2001; von der Haar et al, 2004; Tomoo et al, 2005). Recent solution NMR studies indicate that in the ternary complex, eIF4G binding only causes slight chemical shift perturbations in the cap binding site relative to the yeast eIF4E-cap complex (Gross et al, 2003). Using short eIF4G and 4E-BP1 peptides, crystallographic studies of mouse N-terminal truncated eIF4E indicated that there is no significant structural change between eIF4E-cap and ternary complexes containing either peptide (Marcotrigiano et al, 1999). Given the similarity of the cap binding site in the cap bound and ternary complexes, we hypothesize that eIF4G and the 4E-BPs alter affinity of eIF4E for the cap by modulating the structure of the cap-free form of eIF4E rather than causing substantial structural alterations to the cap bound form of the protein. The biophysical and biological data are clear, cells can modulate proliferation via conformational rearrangements of eIF4E. In order to understand any structurally based process, it is critical to know the relevant structural features of each state. To date, no structure has been solved for any cap free form of eIF4E. Here, we report the first apo-eIF4E structure. Our studies reveal the structural rearrangements that eIF4E undergoes in order to bind the m7G cap. Further, our data provide the first structural basis for the modulation of eIF4E activity via conformational rearrangements. In essence, the conformational response to eIF4G is mediated by partially structuring the apo-eIF4E in a state between apo- and cap-bound eIF4E; we refer to this as prestructuring. Similarly, cap addition modulates the conformation of the dorsal surface. Further, we discovered that a mutation (residue K119 mutated to alanine) distal to either the cap binding site or the dorsal surface (Figure 1) also contains structural features consistent with partial prestructuring of the apo mutant relative to apo-wild type eIF4E. Mutation in this location causes conformational changes in both the cap binding site and dorsal surface, consistent with its higher affinity for both cap and eIF4G. Results and discussion Production of apo-eIF4E Full-length, untagged, soluble human eIF4E was produced by bacterial overexpression and purified using m7GDP affinity chromatography. To elute eIF4E from the column, m7G was used and subsequently removed by extensive dialysis. To ensure that the eIF4E protein was indeed cap-free after dialysis, we carried out mass spectrometry analysis. As can be seen in Figure 2A, the molecular mass of eIF4E is 24964.3±3.2 Da slightly less than the predicted mass 25097.2 Da. This mass difference (132.9 Da) corresponds to the loss of methionine (131.2 Da) at the N-terminus, which is a common event (Huang et al, 1987). The absence of Met1 in the NMR data corroborates this hypothesis. Clearly, there is no peak corresponding to eIF4E-m7G (which would be +297.2 Da higher), indicating that the eIF4E used in our studies is indeed cap free. To further ensure that the apo-eIF4E form was produced, we compared 1H-15N heteronuclear single quantum correlation (HSQC) of protein purified by elution of eIF4E from the cap resin with either high salt or the m7G cap. The 1H-15N HSQC for samples produced in these conditions were identical (Supplementary Figure 1). Consistently, the dissociation constant (Kd) for the cap for our apo-eIF4E is similar to previous studies, which purified eIF4E using a GST tag instead of using cap chromatography (Kd eIF4E-GST purified=1.2±0.2 μM, see Kentsis et al, 2001; Kd for cap chromatography followed by extensive dialysis=1.90±0.38 μM, see Figure 2E and Supplementary Figure 2). This further supports that eIF4E produced here is cap free. Further, addition of m7GDP to the cap-free form led to widespread alterations in chemical shifts and heteronuclear nuclear overhauser effect (hNOE) as compared to the apo form (Figures 2C and 4A). Finally, cap addition led to readily detectable cap-bound form (mass 25422.5±2.1 Da) by mass spectrometry (Figure 2B). Note that there are differences in ionization efficiencies of apo- versus cap-eIF4E (see figure legend in Supplementary Figure 3). Figure 2.Mass spectrum of the apo- (A) and m7GDP-bound (B) forms of wild-type eIF4E. ES-MS spectra plotting ion abundance as a function of the mass/charge ratio. Insets: hypermass reconstruction of the spectrum. The low molecular weight peak at 458 is consistent with the presence of the m7GDP (457.23 Da). For more details, see figure legend in Supplementary Figure 3. Clearly, no cap-bound eIF4E is present in the apo sample. (C) Superposition of the 1H-15N HSQC spectrum of the apo-eIF4E (orange) and the m7GDP-bound eIF4E (green). (D) The far-UV CD spectra of apo-eIF4E (orange) and m7GDP–eIF4E (green). (E) Fluorescence emission of wild-type eIF4E in the presence of increasing concentrations of m7GDP (continuous) and intrinsic fluorescence of m7GDP in the absence of eIF4E (dashed). The different m7GDP concentrations (μM) are shown on the curves and fit is shown in Supplementary Figure 2. Download figure Download PowerPoint Structure determination For many years, apo-eIF4E has eluded attempts at high-resolution structural studies. Recent studies on yeast eIF4E, which has ∼30% identity to the human eIF4E studied here (Figure 3A), suggested that the apo form of eIF4E was unfolded (Niedzwiecka et al, 2004; von der Haar et al, 2006). To find appropriate conditions for NMR, we screened and characterized a wide variety of solution conditions. Significantly, we were able to obtain well-resolved NMR spectra for the apo-eIF4E for a narrow range of conditions indicating that eIF4E can exist as a folded protein in the apo state. The final solution conditions used for structure determination were 0.3 mM protein in 50 mM phosphate buffer (pH 7.4), 100 mM NaCl, 1 mM dithiothreitol (DTT), 0.02% NaN3. In these conditions, translational diffusion coefficients from NMR indicate that eIF4E was monomeric (see Supplementary Figure 4). This was confirmed by relaxation data, which estimated an overall correlation time of 19.1 ns from R2/R1 ratios, close to the expected value for a 217 residue protein at 20°C of ∼18 ns. This value is also close to that estimated from R2/R1 ratios for the cap form (18.5 ns). However, under some other conditions examined, and particularly in N-(2-hydroxyethyl)piperazine-N′-2-ethanesulfonic acid (HEPES), apo-eIF4E aggregated substantially (see Supplementary Figure 4). Addition of cap partially reversed this aggregation as previously shown (von der Haar et al, 2004; Niedzwiecka et al, 2005). It is also important to note here that cap-bound eIF4E was much less susceptible to aggregation and conformational changes associated with altering solution conditions (discussed below). CD analysis indicates that the apo form of eIF4E is folded (Figure 2D) consistent with previous studies (McCubbin et al, 1988; von der Haar et al, 2000; Cohen et al, 2001). Figure 3.Structural comparison between apo- and cap-bound eIF4E. (A) Sequence alignments of eIF4E (Scerv., Saccharomyces cerevisiae). The secondary structural elements were assigned from the apo structure. W56 and W102 residues are boxed. (B) Superposition of the 10 lowest-energy NMR structures. Side chains of W56 and W102 are shown. (C) Crystal structure of human eIF4E bound to m7GDP (PDB 1EJ1). This structure contains density for residues 36–207 and 213–217. The side chains of residues that interact with the cap are shown. (D–G) Potential map of the surface of apo-eIF4E (D, F) and cap-eIF4E (E, G) calculated with MOLMOL. The orientation of (D) and (E) are the same as that of (B) and (C) (cap-binding site) while (F) and (G) represent a 180° rotation along a vertical axis compared with (D) and (E) (eIF4G binding site). A number of residues of interest are indicated on the structures (see text). Download figure Download PowerPoint We determined the first 3D structure of the apo form of eIF4E using standard triple resonance NMR techniques (see Materials and methods). A total of 1631 nonredundant interproton distance restraints (65% of them being medium- and long-range NOEs) and 276 dihedral restraints were used in the final round of calculations. Restraint information and structural analysis are summarized in Table I. Using the Crystallography and NMR System (CNS) software, 50 structures were calculated, and the 10 lowest energy structures were chosen to represent the solution structure of apo-eIF4E (Figure 3B). Secondary structure elements identified from characteristic backbone NOE connectivities, deviations from random coil chemical shifts and H/D exchange experiments indicated apo-eIF4E is comprised of three long α-helices (68–78 H2, 125–139 H4 and 173–187 H5) and three short one-turn helices (56–60 H1, 106–110 H3 and 200–204 H6), plus eight β-strands (42–48 S1, 60–67 S2, 89–94 S3, 110–117 S4, 151–157 S5, 160–167 S6, 196–199 S7 and 215–217 S8) (Figure 3A). In this NMR ensemble of the apo-eIF4E, the loops S1–S2, H2–S3, H4–S5 and S7–S8 are less defined while the first 35 residues at the N-terminus are completely disordered. This correlates with the lack of structural restraints in these regions and is confirmed by relaxation data and significantly the 15N hNOE values (Figure 4A), which are distinctly lower for these regions with weighted mean values of 0.62±0.02, 0.70±0.03 and 0.55±0.05 for the S1–S2, H4–S5 and S7–S8 loops, respectively. Consistent results are obtained from T1 and T2 relaxation experiments (Supplementary Figure 5). We also note that no reliable NOE's or R1 and R2 values could be obtained from residues in the H2–S3 loop and at the centre of the S7–S8 loop, due to line broadening, most likely a consequence of motions on the ms timescales at these sites. Indeed, there is evidence for elevated R2 values in the H4–S5 loop for both the cap and apo forms of eIF4E (Supplementary Figure 6). In contrast, residues comprising the long helices and sheets are relatively non-flexible on the ns timescales, yielding an average NOE value of 0.77. Figure 4.(A) hNOEs for backbone amide nitrogens of the apo-eIF4E (black) and m7GDP–eIF4E (red) measured at a proton frequency of 600 MHz. The cap (orange) and eIF4G (purple) binding sites on eIF4E are shown. (B) Motions associated with cap binding in the W56 and W102 loops. Four structures of the NMR ensemble of the apo-eIF4E (yellow) were superimposed with the cap structure (green), and loops containing W56 (S1–S2) and W102 (S3–S4) were highlighted for clarity. (C–F) Chemical shift perturbation of backbone 1HN and 15N resonances (color-coded) between the apo-eIF4E and m7GDP–eIF4E (C, D), and between the apo-eIF4E and apo-eIF4E-eIF4G peptide (E, F). These perturbations were mapped onto the apo-eIF4E structure. The orientation of (D) and (F) represent a 180° rotation along a vertical axis compared with panels (C) and (E), respectively. Analysis of ligand-induced shifts was performed by applying the Pythagorean theorem to weighted chemical shifts: Δδ (1H,15N)={Δδ(1H)2+0.2 × Δδ(15N)2}1/2, where Δδ(1H) and Δδ(15N) are the chemical-shift differences of the amide proton and nitrogen, respectively (Grzesiek et al, 1996; Pellecchia et al, 1999). These deviations are also shown for the indole Hε1 of the Trp residues. Download figure Download PowerPoint Table 1. Structure statistics of apo-eIF4E Field and variable(s) Value Restraints for final structure calculations Total restraints used 1973 Total NOE restraints 1697 Intraresidue 145 Sequential (∣i−j∣=1) 438 Medium range (1<∣i−j∣=4) 394 Long range (∣i−j∣>4) 560 Hydrogen bond restraintsa 160 ϕ dihedral angles restraints 138 ψ dihedral angles restraints 138 Statistics for structure calculations (〈SA〉b) RMSD from idealized covalent geometry Bonds (Å) 0.0018±0.0001 Angles (deg) 0.350±0.011 Improper (deg) 0.230±0.019 RMSD from experimental restraints: distances (Å)c 0.012±0.001 Ramachandran plot statistics (%)d Residues in most favored regions 78.7±1.7 Residues in additional allowed regions 15.7±1.5 Residues in generously allowed regions 3.7±1.3 Residues in disallowed regions 1.9±0.7 a The amide protons implicated in all these hydrogen bonds were found to slowly exchange with D2O. b 〈SA〉 refers to the ensemble of the 10 structures with the lowest energy from 50 calculated structures. c No distance restraint in any of the structures included in the ensemble was violated by more than 0.2 Å. d Generated using PROCHECK on the ensemble of the 10 lowest-energy structures, residues 36–217. The residues located in the two latter regions are mostly located in flexible regions or connect secondary structural regions. The root mean square deviation of this family with respect to the mean coordinate positions is 1.26 and 2.25 Å for backbone and heavy atoms, respectively, for regular secondary structure elements. These values are close to those obtained for the yeast eIF4E/m7GDP/eIF4G fragment complex (1.24 and 2.08 Å, respectively) (Gross et al, 2003). Comparison to cap-bound eIF4E structures The cap bound structure is often described as a cupped hand (Figure 3C). As a comparison, we describe the apo-form as an open hand. Our apo structure determined in solution is in many ways very similar to the crystal eIF4E-cap structures reported for human and mouse eIF4E, the latter being 98% identical to the human protein. The apo form, like the cap bound form, contains an 8-stranded anti-parallel β-sheet on the cap binding side, and three long helices on the convex face that binds other protein partners (Figure 3B). The core β-strands are positioned nearly identically. Differences between the two structures are observed in the loops, the helical content and for some beta strands (S4–S8). In particular, structural differences are observed for the loops S1–S2, H1–S3, S3–S4, H2–S5, S5–S6, H3–S7 and S7–S8. Extent of structural changes are consistent with previous CD data that predicted ∼20% changes (Cai et al, 1999; von der Haar et al, 2000; Cohen et al, 2001). However, in contrast to all other reported crystal structures of binary and tertiary eIF4E complexes, the apo structure differs substantially at the cap binding surface (Figure 3D and E), at the dorsal surface important for eIF4G/4E-BPs binding and adjacent regions (Figure 3F and G). For the cap binding site, significant rearrangements are observed for the tryptophan residues crucial for m7G binding. Additionally, the positively charged pocket in the cap binding site formed by the side chains of residues K52, R157, K159, K162, K206, K212 in the cap-bound structure is dispersed in the apo structure, as alluded to by its open-hand conformation. Conformational rearrangements of the loops S1–S2 and S7–S8 largely contribute to this dispersion, while R157, K159 and K162 side chains adopt a slightly different position. In terms of previously reported conformational regulation of eIF4E, cap binding results in alterations in the electrostatic potential on the dorsal surface, consistent with increased affinity of the cap-bound form for eIF4G (von der Haar et al, 2004; Figure 3F and G). We also investigated the importance of the N-terminus in stabilizing eIF4E and in cap binding. For instance, deletion of residues 1–20 from eIF4E reduces the Kd of eIF4E for eIF4G by three-fold, while it is reduced by 16- and 60-fold by deleting residues 1–30 and 1–35, respectively (Gross et al, 2003). Interestingly, our own deletion studies demonstrate that deletion of the first 36 residues leads to substantial chemical shift alterations throughout helix 2 (H2) (data not shown), suggesting that residues in this helix may play important roles in stabilizing the apo-eIF4E form as well as making critical contacts with eIF4G. Structural rearrangements associated with cap binding: locking a hinge In order to compare the cap bound and apo structures in solution, we analyzed chemical shift perturbations and dynamics in the apo and cap bound forms of eIF4E (Figures 4 and 5). 1H and 15N chemical shift perturbation arise from alterations in the local chemical environment. This can be due to conformational rearrangements and/or to proximity to the ligand. In contrast, 13Cα chemical shift perturbations result from alterations in the backbone structure usually implying differences in secondary structure. A summary of cap binding induced chemical shift perturbations is shown in Figure 5. The major 1H and 15N chemical shifts perturbations between apo and cap bound forms of eIF4E involved residues known to bind the cap including the main chain amide and the indoles of W56 and W102. In the apo structure, W56 is present on a flexible loop (residues 49–60). Overlaying just this part of the apo-structure reveals that loop is relatively well-defined structurally and is likely moving as a hinge relative to the rest of apo-eIF4E. Consistently, there are only small changes in the αC chemical shifts for this loop upon cap binding, indicating that the W56 loop does not undergo substantial changes in secondary structure in the apo and cap bound forms of eIF4E. In contrast, W102 goes through substantial alterations in its backbone conformation as can be seen by observing the large differences in αC chemical shifts for W102 and nearby residues including E103 (Figure 5C). In fact upon cap binding, this region undergoes the largest secondary structural changes in the protein. Consistent with chemical shift perturbation studies, we observe NOEs between the cap and eIF4E in the cap-bound complex (Supplementary Figure 6). In summary, upon cap binding, eIF4E undergoes two distinct motions: (1) locking the W56 hinge and (2) rotating W102 and nearby residues into the cap binding site (Figure 4B). The order of these events, or if they are concerted, has yet to be determined. The 15N hNOE data (Figure 4A) and R1 data confirm the presence of mobility on the ns–ps timescales for the loops containing W56 and W102 residues that become abrogated upon cap binding (especially for the W56 containing loop). T1 values that, together with hNOEs, are sensitive to high-frequency motions also show similar tendencies (Supplementary Figure 5). While for T2, which is a function of much slower processes, this phenomenon appears less pronounced in these two loops. All together, these data confirm the reduction of the rapid large amplitude motions around W56 upon cap binding. Figure 5.Chemical shift differences between the apo and the ternary complex with either the cap-eIF4E (A), or the eIF4E-4G peptide (B) as binary complexes. The right panels represent the chemical shift differences for the eight Trp indoles HN. Positions indicated with arrows show proline residues (black), E103 which is not assigned in cap-eIF
Referência(s)