FRINK Linked Data Fragments server

Matches in SemOpenAlex for { <https://semopenalex.org/work/W1978698911> ?p ?o ?g. }

• W1978698911 endingPage "3309" @default.
• W1978698911 startingPage "3303" @default.
• W1978698911 abstract "In eukaryotes, a key step in the initiation of translation is the binding of the eukaryotic initiation factor 4E (eIF4E) to the cap structure of the mRNA. Subsequent recruitment of several components, including the small ribosomal subunit, is thought to allow migration of initiation complexes and recognition of the initiation codon. Mitogens and cytokines stimulate the phosphorylation of eIF4E at Ser209, but the functional consequences of this modification have remained a major unresolved question. Using fluorescence spectroscopy and surface plasmon resonance techniques, we show that phosphorylation of eIF4E markedly reduces its affinity for capped RNA, primarily due to an increased rate of dissociation. Variant eIF4E proteins harboring negatively charged acidic residues at position 209 also showed decreased binding to capped RNA. Furthermore, a basic residue at position 159 was shown to be essential for cap binding. Although eIF4E-binding protein 1 greatly stabilized binding of phosphorylated eIF4E to capped RNA, in the presence of eIF4E-binding protein 1 the phosphorylated form still dissociated faster compared with nonphopshorylated eIF4E. The implications of our findings for the mechanism of translation initiation are discussed. In eukaryotes, a key step in the initiation of translation is the binding of the eukaryotic initiation factor 4E (eIF4E) to the cap structure of the mRNA. Subsequent recruitment of several components, including the small ribosomal subunit, is thought to allow migration of initiation complexes and recognition of the initiation codon. Mitogens and cytokines stimulate the phosphorylation of eIF4E at Ser209, but the functional consequences of this modification have remained a major unresolved question. Using fluorescence spectroscopy and surface plasmon resonance techniques, we show that phosphorylation of eIF4E markedly reduces its affinity for capped RNA, primarily due to an increased rate of dissociation. Variant eIF4E proteins harboring negatively charged acidic residues at position 209 also showed decreased binding to capped RNA. Furthermore, a basic residue at position 159 was shown to be essential for cap binding. Although eIF4E-binding protein 1 greatly stabilized binding of phosphorylated eIF4E to capped RNA, in the presence of eIF4E-binding protein 1 the phosphorylated form still dissociated faster compared with nonphopshorylated eIF4E. The implications of our findings for the mechanism of translation initiation are discussed. eukaryotic initiation factor eIF4E-binding protein surface plasmon resonance In eukaryotic cells, all nucleus-encoded mRNAs possess a so-called “cap structure” at their 5′-end. This cap consists of a methylated 5′-5′ bound guanosine triphosphate (m7GTP) moiety. The cap plays a key role in the translation of the mRNA by permitting recruitment of the eukaryotic initiation factors (eIFs)1 required for the attachment of the ribosome and correct initiation of translation. The protein that interacts directly with the cap is eIF4E, which forms a complex with the scaffolding protein eIF4G, which in turn recruits other factors. These include eIF4A and eIF4B, which are involved in unwinding secondary structure in the 5′-untranslated region of the mRNA, and the poly(A)-binding protein, which by interacting with the 3′-tail of the mRNA circularizes it (1Gingras A.-C. Raught B. Sonenberg N. Annu. Rev. Biochem. 1999; 68: 913-963Crossref PubMed Scopus (1762) Google Scholar, 2Wells S.E. Hillner P.E. Vale R.D. Sachs A.B. Mol. Cell. 1998; 2: 135-140Abstract Full Text Full Text PDF PubMed Scopus (723) Google Scholar). eIF4E also binds to small regulatory proteins termed eIF4E-binding proteins (4E-BPs). These compete with eIF4G for binding to overlapping sites on eIF4E and thus inhibit formation of initiation complexes (3Haghighat A. Mader S. Pause A. Sonenberg N. EMBO J. 1995; 14: 5701-5709Crossref PubMed Scopus (530) Google Scholar, 4Mader S. Lee H. Pause A. Sonenberg N. Mol. Cell. Biol. 1995; 15: 4990-4997Crossref PubMed Google Scholar). It has long been known that eIF4E undergoes regulated phosphorylation, and the site has been identified as Ser209 (5Flynn A. Proud C.G. J. Biol. Chem. 1995; 270: 21684-21688Abstract Full Text Full Text PDF PubMed Scopus (137) Google Scholar, 6Joshi B. Cai A.L. Keiper B.D. Minich W.B. Mendez R. Beach C.M. Stepinski J. Stolarski R. Darzynkiewicz E. Rhoads R.E. J. Biol. Chem. 1995; 270: 14597-14603Abstract Full Text Full Text PDF PubMed Scopus (186) Google Scholar). Phosphorylation of eIF4E is increased by mitogenic stimuli that activate translation (reviewed in Ref. 7Kleijn M. Scheper G.C. Voorma H.O. Thomas A.A.M. Eur. J. Biochem. 1998; 253: 531-544Crossref PubMed Scopus (170) Google Scholar) and by cytokines (8Morley S.J. McKendrick L. J. Biol. Chem. 1997; 272: 17887-17893Abstract Full Text Full Text PDF PubMed Scopus (115) Google Scholar, 9Wang X. Flynn A. Waskiewicz A.J. Webb B.L. Vries R.G. Baines I.A. Cooper J.A. Proud C.G. J. Biol. Chem. 1998; 273: 9373-9377Abstract Full Text Full Text PDF PubMed Scopus (255) Google Scholar). Recently, two kinases were identified that phosphorylate Ser209 in eIF4E and are targets for the mitogen-activated extracellular signal-regulated kinase and stress/cytokine-activated p38 mitogen-activated protein kinase pathways (9Wang X. Flynn A. Waskiewicz A.J. Webb B.L. Vries R.G. Baines I.A. Cooper J.A. Proud C.G. J. Biol. Chem. 1998; 273: 9373-9377Abstract Full Text Full Text PDF PubMed Scopus (255) Google Scholar, 10Waskiewicz A.J. Flynn A. Proud C.G. Cooper J.A. EMBO J. 1997; 16: 1909-1920Crossref PubMed Scopus (786) Google Scholar, 11Waskiewicz A.J. Johnson J.C. Penn B. Mahalingam M. Kimball S.R. Cooper J.A. Mol. Cell. Biol. 1999; 19: 1871-1880Crossref PubMed Scopus (402) Google Scholar, 12Scheper G.C. Morrice N.A. Kleijn M. Proud C.G. Mol. Cell. Biol. 2001; 21: 743-754Crossref PubMed Scopus (169) Google Scholar, 13Fukunaga R. Hunter T. EMBO J. 1997; 16: 1921-1933Crossref PubMed Scopus (555) Google Scholar). These enzymes (Mnk1 and Mnk2) also associate with eIF4G in vivo (11Waskiewicz A.J. Johnson J.C. Penn B. Mahalingam M. Kimball S.R. Cooper J.A. Mol. Cell. Biol. 1999; 19: 1871-1880Crossref PubMed Scopus (402) Google Scholar, 12Scheper G.C. Morrice N.A. Kleijn M. Proud C.G. Mol. Cell. Biol. 2001; 21: 743-754Crossref PubMed Scopus (169) Google Scholar, 14Pyronnet S. Imataka H. Gingras A.C. Fukunaga R. Hunter T. Sonenberg N. EMBO J. 1999; 18: 270-279Crossref PubMed Scopus (536) Google Scholar). Despite these advances and the large number of studies on the changes of eIF4E phosphorylation in vivo, the key issue of the effect of phosphorylation on the function of eIF4E has received little attention and has remained a major question in the field. The crystal structure of mammalian eIF4E in its nonphosphorylated form (15Marcotrigiano J. Gingras A.C. Sonenberg N. Burley S.K. Cell. 1997; 89: 951-961Abstract Full Text Full Text PDF PubMed Scopus (561) Google Scholar) suggests that Ser209 lies close to a cleft in the protein through which the mRNA may “exit” and opposite to a lysyl residue (Lys159). This led to speculation that, by introducing a negatively charged group, phosphorylation of Ser209 might allow a salt bridge to form with Lys159, thus leading to “cleft closure” and allowing tighter binding of eIF4E to capped mRNAs. However, the effects of phosphorylation on eIF4E function using quantitatively phosphorylated eIF4E have not previously been measured. Here we exploit the availability of highly active preparations of an authentic eIF4E kinase, Mnk2, to produce stoichiometrically phosphorylated eIF4E. We find, counter to a previous report (16Minich W.B. Balasta M.L. Goss D.J. Rhoads R.E. Proc. Natl. Acad. Sci. U. S. A. 1994; 91: 7668-7672Crossref PubMed Scopus (260) Google Scholar), that phosphorylation of eIF4E diminishes its ability to bind cap analogue or capped mRNA, and we propose a new model for the role of eIF4E phosphorylation during translation initiation. pEBG-Mnk2 and pEBG-Mnk1T2A2 (both kindly provided by Dr. A. Waskiewicz, University of Washington, Seattle, WA) were transfected into human embryonic kidney 293 cells. The expressed glutathione S-transferase fusion proteins for wild type Mnk2 (as a source of active eIF4E kinase) and Mnk1-T2A2 (dead kinase) were purified on glutathione-Sepharose beads. The washing procedure for the beads included two wash steps with 0.5 m lithium chloride to prevent co-purification of mitogen-activated protein kinase. Vectors encoding the S209A, S209D, and S209E variants of eIF4E (a gift from Dr. M. Underhill, University of Kent, UK) were created by PCR with the desired mutation in the downstream primer. eIF4E K159M, K159R, and R157K were generated using the QuikChange® system from Stratagene. Wild type eIF4E and the variants were expressed in Escherichia coli and purified as described earlier (17Stern B.D. Wilson M. Jagus R. Protein Expression Purif. 1993; 4: 320-327Crossref PubMed Scopus (29) Google Scholar). Recombinant His-tagged 4E-BP1 was a kind gift from Manjur Karim (University of Manchester Institute of Science and Technology, Manchester, UK). Isoelectric focusing was performed as described (18Kleijn M. Voorma H.O. Thomas A.A.M. J. Cell. Biochem. 1995; 59: 443-452Crossref PubMed Scopus (21) Google Scholar). The concentrations of the recombinant proteins were determined using the protein assay reagent from Bio-Rad. Recombinant eIF4E was phosphorylated in vitroin reactions containing glutathione S-transferase-Mnk2, 20 mm Hepes·KOH, pH 7.5, 100 mm KCl, 10 mm MgCl2, 200 μm ATP, 0.1 mm EDTA, 25 mm β-glycerophosphate, 0.5 mm sodium vanadate, and 1 mm dithiothreitol at 30 °C overnight. Glutathione S-transferase-Mnk2 was used, since this kinase has a high basal activity in unstimulated cells (12Scheper G.C. Morrice N.A. Kleijn M. Proud C.G. Mol. Cell. Biol. 2001; 21: 743-754Crossref PubMed Scopus (169) Google Scholar). As a control for possible loss of eIF4E to walls of tubes or loss of activity, incubations were also performed in parallel using the inactive T2A2 variant of Mnk1 (10Waskiewicz A.J. Flynn A. Proud C.G. Cooper J.A. EMBO J. 1997; 16: 1909-1920Crossref PubMed Scopus (786) Google Scholar). After the reactions, the kinases were removed by spinning down the glutathione-Sepharose and followed by careful removal of the supernatant. The kinases were undetectable in the final eIF4E batches by Western blotting with antibodies against the glutathione S-transferase moiety. In a similar fashion, we did not detect eIF4G, which could have co-purified with the glutathioneS-transferase-Mnks, in the preparations of mock-phosphorylated and phosphorylated eIF4E. Recombinant wild type eIF4E, either untreated, mock-phosphorylated, or phosphorylated, or variant forms of eIF4E in a volume of 30 μl were incubated with 15 μl of m7GTP-Sepharose (1:1 slurry (Amersham Biosciences)) and 15 μl of Sepharose CL-4B (Sigma) for 1 h at 4 °C while shaking at 1000 rpm in an Eppendorf thermomixer. After this incubation, the beads were spun down at 13,000 × g for 20 s, and the supernatant was carefully transferred to a new tube and diluted with SDS-PAGE sample buffer. The Sepharose beads were washed three times with 500 μl of buffer containing 20 mm Hepes·KOH, pH 7.5, 100 mm KCl, 0.1 mm EDTA, 25 mm β-glycerophosphate, 0.5 mm sodium vanadate, and 1 mm dithiothreitol. The beads were resuspended in SDS-PAGE sample buffer. Unbound and bound eIF4E were analyzed by SDS-PAGE electrophoresis and Coomassie staining. All SPR experiments were performed using a BIAcore 3000. An NTA Sensor Chip (BIAcore) was used for immobilizing His6-tagged 4E-BP1. Injection of the His6-tagged 4E-BP1 (300 nm in eluent buffer (10 mm HEPES, pH 7.5, 300 mm KCl, 50 μm EDTA, and 0.01% P20 (a nonionic surfactant (Biacore)) resulted in a response of ∼200 resonance units for the immobilized protein. eIF4E was then injected in the same buffer. The chip was regenerated after each cycle with 350 mm EDTA in the eluent buffer. All measurements were performed at a flow rate of 30 μl/min at 25 °C. The response of the nickel-coated chips without the immobilized protein was subtracted from the response obtained with the protein-coated chips. The resulting curves were evaluated using the BIA Evaluation software package. Biotinylated capped and uncapped RNAs were prepared by in vitro transcription in the presence or absence of m7GpppG, respectively, with the same template as described previously (19Ptushkina M. von der Haar T. Karim M.M. Hughes J.M. McCarthy J.E. EMBO J. 1999; 18: 4068-4075Crossref PubMed Scopus (105) Google Scholar), which was kindly provided by Dr. Tobias von der Haar (University of Manchester Institute of Science and Technology, Manchester, UK). The RNAs contain only one uridine (in the 3′-end), allowing synthesis of singly 3′-biotinylated transcripts. The RNAs were diluted in 20 mm HEPES, pH 7.4, 100 mm KCl, 2 mm MgCl2, 0.05% P20 and denatured for 5 min at 95 °C, immediately cooled on ice, and subsequently immobilized on Sensor Chip SA (streptavidin) chips (Biacore) at a flow rate of 5 μl/min for 6 min. The binding of (phosphorylated) eIF4E was performed in 20 mm HEPES, pH 7.4, 100 mm KCl, 2 mm MgCl2, 100 μg/ml calf liver tRNA (Roche Molecular Biochemicals), 0.05% P20. All further injections were carried out at 25 μl/min for 2 min. Where required, regeneration was performed with 30 μl of binding buffer containing 2 m KCl and 0.1 mm m7GTP. The response from the RNA-coated channels was constant, indicating that the RNAs were not degraded over the course of the experiments. The RNA-coated chips could even be reused the next day. The resulting curves were evaluated using the BIA Evaluation software package and globally fitted (1:1 Langmuir binding). χ2 values for the fits were on the order of 0.4–1.2. Analysis of data obtained at different flow rates (5 or 75 μl/min) resulted in very similar Kd values, indicating that the results were not affected by mass transfer limitations. All fluorescence titrations were performed using a Spex Fluorolog τ2 spectrofluorimeter. The intrinsic tryptophan fluorescence of the protein was measured by exciting at either 280 or 295 nm, and the emission wavelength was 345 nm. Polarizing filters oriented at the magic angle were used to reduce scatter. All fluorescence intensities were corrected when necessary for the inner filter effect as described previously (20Bi X. Ren J. Goss D.J. Biochemistry. 2000; 39: 5758-5765Crossref PubMed Scopus (53) Google Scholar). The dissociation equilibrium constants were obtained by fitting the titration data as previously described (20Bi X. Ren J. Goss D.J. Biochemistry. 2000; 39: 5758-5765Crossref PubMed Scopus (53) Google Scholar). Nonlinear least-squares fitting of the data was performed using KaleidaGraph software (Version 2.1.3; Abelbeck Software). Human eIF4E was expressed in E. coli, isolated, refolded, and purified on m7GTP-Sepharose (17Stern B.D. Wilson M. Jagus R. Protein Expression Purif. 1993; 4: 320-327Crossref PubMed Scopus (29) Google Scholar). To prepare phosphorylated eIF4E (termed “eIF4E(+P)”), the purified recombinant protein was incubated with Mnk2 (made in 293 cells) and ATP/MgCl2. This resulted in phosphorylation of the eIF4E to almost 100% as revealed by isoelectric focusing analysis (Fig.1A). Mnk2 phosphorylates eIF4E exclusively at Ser209 (12Scheper G.C. Morrice N.A. Kleijn M. Proud C.G. Mol. Cell. Biol. 2001; 21: 743-754Crossref PubMed Scopus (169) Google Scholar). As a control for any effect of adding the kinase preparation itself, eIF4E was incubated with a catalytically inactive Mnk1 (“T2A2”) (11Waskiewicz A.J. Johnson J.C. Penn B. Mahalingam M. Kimball S.R. Cooper J.A. Mol. Cell. Biol. 1999; 19: 1871-1880Crossref PubMed Scopus (402) Google Scholar). eIF4E treated in this way remained completely unphosphorylated (Fig. 1A). Throughout this study, the term “eIF4E(−P)” indicates this “mock-phosphorylated” material as distinct from untreated factor (designated “eIF4E”). Analysis of the binding of these eIF4E preparations to m7GTP-Sepharose showed that Mnk2 treatment did not prevent eIF4E from binding to immobilized cap analogue (Fig.1B). Some concern has been raised whether the cap analogue, used during purification of recombinant eIF4E, is fully removed upon dialysis (21von der Haar T. Ball P.D. McCarthy J.E. J. Biol. Chem. 2000; 275: 30551-30555Abstract Full Text Full Text PDF PubMed Scopus (116) Google Scholar). Therefore, the binding of the various recombinant proteins to cap analogue or capped RNA could be affected by differing amounts of bound m7GTP remaining from the purification procedure. To exclude this possibility, wild type and variant eIF4Es were again expressed inE. coli and purified as described. However, in this case they were eluted from m7GTP-Sepharose with 10 mm GTP (instead of 100 μm m7GTP) for which eIF4E has much lower affinity. It is therefore thought to be removed more efficiently by dialysis. The various forms of eIF4E prepared in this way, including the Mnk2-phosphorylated protein, showed similar binding characteristics for m7GTP-Sepharose as shown to those in Fig. 1B for the m7GTP-eluted protein and yielded almost identical Kd values for capped RNA binding to that by SPR analysis as reported in Fig. 3 (data not shown). Fig. 1B shows that unphosphorylated and mock-phosphorylated, as well as several variants of eIF4E that will be used in the experiments described below, are properly folded, as judged by their ability to bind to m7GTP-Sepharose. However, this approach cannot be used to perform accurate quantitative studies on the affinity of eIF4E for the cap. We examined possible differences between different forms of eIF4E in their binding to cap analog (m7GTP) by fluorescence intensity measurements. Similar fluorescence titrations have been used previously to study the interaction between eIF4E and a variety of cap analogues (see, for example, Refs. 22Carberry S.E. Friedland D.E. Rhoads R.E. Goss D.J. Biochemistry. 1992; 31: 1427-1432Crossref PubMed Scopus (29) Google Scholar and 23Sha M. Wang Y. Xiang T. van Heerden A. Browning K.S. Goss D.J. J. Biol. Chem. 1995; 270: 29904-29909Abstract Full Text Full Text PDF PubMed Scopus (38) Google Scholar), and this approach was also employed by Minich et al. (16Minich W.B. Balasta M.L. Goss D.J. Rhoads R.E. Proc. Natl. Acad. Sci. U. S. A. 1994; 91: 7668-7672Crossref PubMed Scopus (260) Google Scholar) in their study on the effects of eIF4E phosphorylation. When untreated eIF4E was compared with in vitrophosphorylated eIF4E, the phosphorylated factor showed a 2.5-fold lower binding affinity for m7GTP (Kd = 610 nm) than the native eIF4E (Kd = 260 nm). The results of titrations of each protein are shown in Fig. 2. These titrations were performed at a protein concentration of 500 nm for each protein to permit direct comparison of the two curves. Titration experiments with the mock-treated eIF4E gave a similar Kd value (300 nm) to that obtained for the untreated recombinant protein. A widely used approach for addressing the effects of phosphorylation of specific sites in proteins is to create nonphosphorylatable variants or to introduce an acidic residue at the site. We therefore mutated the cDNA to encode either an uncharged residue at position 209 (S209A) or an acidic residue (S209D or S209E). These variants of eIF4E were expressed in E. coli. As expected, the acidic variants showed a more acidic pI on IEF analysis than wild type eIF4E or the S209A variant protein (Fig. 1A). Fluorescence measurements with these three variant proteins and m7GTP were performed in a similar way as that described for unphosphorylated and phosphorylated eIF4E in the legend to Fig. 2. The derivedKd values from these titrations are summarized in Table I. Apparently, in this assay, the acidic side chains of the S209D and S209E variant proteins do not affect the binding to m7GTP, and these variants do not mimic phosphorylated eIF4E.Table IBinding constants for eIF4E/m7 GTP interactionKdnmeIF4E260 ± 20eIF4E (−P)300 ± 60eIF4E (+P)610 ± 40eIF4E S209A340 ± 50eIF4E S209D240 ± 30eIF4E S209E320 ± 40Data obtained by fluorescence titration were analyzed as described in Fig. 2 and under “Experimental Procedures.” Curves were obtained from two experiments with each three concentrations (500, 450, and 400 nm for phosphorylated and at 500, 300, and 200 nm for nonphosphorylated proteins). All of these curves were fit to obtain the Kd values. Open table in a new tab Data obtained by fluorescence titration were analyzed as described in Fig. 2 and under “Experimental Procedures.” Curves were obtained from two experiments with each three concentrations (500, 450, and 400 nm for phosphorylated and at 500, 300, and 200 nm for nonphosphorylated proteins). All of these curves were fit to obtain the Kd values. To confirm and extend our finding that phosphorylated eIF4E has a reduced affinity for the cap, we employed SPR. In particular, we considered it important to mimic the physiological ligand, capped mRNA. To this end, a synthetic mRNA containing a 5′-cap was synthesized. This RNA possessed a biotinylated nucleotide at its 3′-end and was immobilized onto a streptavidin chip. A noncapped version of the RNA was immobilized onto a second channel as a negative control. A similar approach to study the interaction between eIF4E and the cap structure has been used before (24Ptushkina M. von der Haar T. Vasilescu S. Frank R. Birkenhager R. McCarthy J.E. EMBO J. 1998; 17: 4798-4808Crossref PubMed Scopus (124) Google Scholar). eIF4E in its unphosphorylated or phosphorylated states (Fig.3A) was applied to the chip surface, and the resonance signals were monitored. eIF4E(−P) clearly bound to the capped mRNA, and, when the flow was changed to buffer lacking eIF4E, the bound eIF4E dissociated with a half-time of ∼15 s, showing that it readily dissociates (Fig. 3A). Analysis of data from three separate experiments (each employing four or six different eIF4E concentrations) revealed an apparent binding constant for eIF4E(−P) of 1.2 × 10−7m. The binding constant for untreated recombinant eIF4E (Fig. 3B) was very similar to the one for eIF4E(−P), demonstrating that the “mock phosphorylation” does not affect the binding of eIF4E to capped RNA. When eIF4E(+P) was used, very little net binding to capped RNA was observed (Fig. 3A). The steady state level of binding was about 20% of that seen with a similar amount of eIF4E(−P). The small signal and very fast association and dissociation rates for eIF4E(+P) made it difficult to obtain a reliable (apparent) binding constant, although comparison of several sets of experiments shows that the binding affinity of eIF4E(+P) for capped RNA is 5–10-fold lower than that of eIF4E(−P). This decreased affinity reflects a substantial increase in the dissociation rate (see also Table II). Also, the on-rate is very quick, even quicker than for eIF4E(−P) and untreated recombinant eIF4E. The opposite would have been anticipated if phosphorylation causes a so-called “cleft closure,” as speculated by others (15Marcotrigiano J. Gingras A.C. Sonenberg N. Burley S.K. Cell. 1997; 89: 951-961Abstract Full Text Full Text PDF PubMed Scopus (561) Google Scholar). In the presence of 4E-BP1, phosphorylated eIF4E does bind to capped RNA (see below), which also argues against cleft closure as an explanation of the effects we have measured for phosphorylation of eIF4E on cap binding.Table IIDissociation values for eIF4E/capped RNA interactionkdkd with BP1s−14E−P0.060 ± 0.01 (n = 8)8.12 × 10−44E+P0.58 ± 0.15 (n = 8)3.96 × 10−34E WT0.078 ± 0.01 (n = 8)6.99 × 10−44E S209A0.067 ± 0.02 (n = 6)6.84 × 10−44E S209D0.44 ± 0.02 (n = 6)3.01 × 10−34E S209E0.42 ± 0.03 (n = 6)3.17 × 10−3Data obtained by SPR as shown in Figs. 3 and 4 were analyzed as described under “Experimental Procedures.” The S.D. values indicated for the dissociation constant in the absence of 4E-BP1 were obtained by fitting individual curves (as shown in Fig. 3) separately. Dissociation rates of eIF4E in the presence of 4E-BP1 were derived from SPR curves after injection of a mixture of 80 nm eIF4E with varying concentrations of 4E-BP1 (30–400 nm) followed by dissociation in buffer. Open table in a new tab Data obtained by SPR as shown in Figs. 3 and 4 were analyzed as described under “Experimental Procedures.” The S.D. values indicated for the dissociation constant in the absence of 4E-BP1 were obtained by fitting individual curves (as shown in Fig. 3) separately. Dissociation rates of eIF4E in the presence of 4E-BP1 were derived from SPR curves after injection of a mixture of 80 nm eIF4E with varying concentrations of 4E-BP1 (30–400 nm) followed by dissociation in buffer. eIF4E(S209A) bound to capped mRNA with similar affinity to eIF4E(−P) or untreated eIF4E (Fig. 3B). The side chain hydroxyl group of Ser209 therefore does not itself appear to be important for the interaction between eIF4E and capped mRNA. Treatment of S209A with active Mnk2 did not affect its binding to capped RNA (data not shown). This implies that the observed effects upon incubation of recombinant wild type eIF4E with active kinase are indeed due to phosphorylation and are not caused by an unknown effect of the kinase treatment. The acidic variants S209D and S209E each showed some ability to bind capped RNA (Fig. 3B), but the level of binding was reduced relative to the wild-type protein. Particularly striking were the faster off-rates for these variants as compared with the normal protein. The Kd values for the acidic variants are about 3–4 × 10−7m, indicating a 3-fold reduction in binding affinity relative to wild-type eIF4E. Thus, changing Ser209 to Asp or Glu does appear partially to mimic phosphorylation in the case of this interaction, but the effects of the acidic residues are smaller than those of phosphorylation itself. The data strongly indicate that phosphorylation of eIF4E weakens its binding to capped mRNA. These data are initially surprising given the predictions from the crystal structure concerning a salt bridge between Ser(P)209 and Lys159, which was expected to increase cap binding. In contrast, it appears that introduction of negative charge in the region of the proposed mRNA exit site impairs the affinity of eIF4E for capped mRNA. To study the role of Lys159 itself in capped mRNA binding, we mutated it to an uncharged but sterically similar residue (Met). Although this variant retained the ability to bind m7GTP-Sepharose (Fig. 1B), indicating that it was properly folded, the binding of eIF4E(K159M) to capped RNA was even weaker than that of eIF4E(+P) (Fig. 3C). This suggests that Lys159 plays an important role in the stable interaction of eIF4E with capped RNA, perhaps through the interaction of its positively charged side chain with negatively charged phosphate groups on the RNA. To test this idea, the lysine at position 159 was replaced by an arginine, and the resulting recombinant protein was tested for binding to capped RNA (Fig. 3C). This variant bound to the capped RNA with the same affinity as wild type eIF4E, strengthening the idea that a positive charge at this position is indeed essential. Alignment of 30 eIF4E sequences from 20 different species revealed that in eIF4E sequences from wheat, rice, and maize an arginine is indeed present at this position (data not shown). Among these 30 eIF4E proteins, an arginine at position at 157 was completely conserved. To test whether a positive residue at this position was sufficient for eIF4E function, this arginine was mutated to lysine, and the resulting recombinant protein was again tested. Binding to m7GTP-Sepharose seemed to be somewhat affected by this arginine to lysine substitution (see Fig. 1B), whereas binding of capped RNA was severely compromised (Fig. 3C, R157K). Apparently, an arginine at this position is of critical importance for the ability of eIF4E to bind capped RNA. To assess the effect of phosphorylation of eIF4E on its other binding properties, we employed SPR to analyze the interaction between eIF4E and its binding partner, 4E-BP1. Histidine-tagged 4E-BP1 was immobilized on the NTA chip, and eIF4E was passed over it. A lane lacking immobilized 4E-BP1 served as negative control; signals from this lane were subtracted from those from the first lane to obtain net binding. Comparison of eIF4E(+P) and eIF4E(−P) samples showed no difference in the kinetics of binding or dissociation of the eIF4E (Fig. 4). For both eIF4E preparations, a binding constant of 1.1 × 10−8m was determined using the Biacore evaluation software to fit the three curves (global fit, 1:1 Langmuir binding). These values are in very good agreement with those reported by Ptushkina et al. (19Ptushkina M. von der Haar T. Karim M.M. Hughes J.M. McCarthy J.E. EMBO J. 1999; 18: 4068-4075Crossref PubMed Scopus (105) Google Scholar) (1.0 ± 0.4 × 10−8). Thus, phosphorylation of eIF4E does not appear to influence its ability to interact with 4E-BP1. Since the same region of eIF4E is involved in its interactions with 4E-BP1 and eIF4G (1Gingras A.-C. Raught B. Sonenberg N. Annu. Rev. Biochem. 1999; 68: 913-963Crossref PubMed Scopus (1762) Google Scholar, 19Ptushkina M. von der Haar T. Karim M.M. Hughes J.M. McCarthy J.E. EMBO J. 1999; 18: 4068-4075Crossref PubMed Scopus (105) Google Scholar), it is highly likely that phosphorylation of eIF4E also does not affect binding to eIF4G, although we have been unable to test this directly. Since eIF4E interacts with other proteins in vivo (eIF4G, 4E-BPs), it was important to examine whether they modified the effect of phosphorylation on the binding of eIF4E to capped RNA. Association of eIF4E with 4E-BP1 or eIF4G has been shown to stabilize its interaction with capped mRNA (19Ptushkina M. von der Haar" @default.
• W1978698911 created "2016-06-24" @default.
• W1978698911 creator A5022031402 @default.
• W1978698911 creator A5024431768 @default.
• W1978698911 creator A5030513468 @default.
• W1978698911 creator A5046484870 @default.
• W1978698911 creator A5063878749 @default.
• W1978698911 creator A5070388182 @default.
• W1978698911 date "2002-02-01" @default.
• W1978698911 modified "2023-09-28" @default.
• W1978698911 title "Phosphorylation of Eukaryotic Initiation Factor 4E Markedly Reduces Its Affinity for Capped mRNA" @default.
• W1978698911 cites W1607579928 @default.
• W1978698911 cites W1866899500 @default.
• W1978698911 cites W1965894414 @default.
• W1978698911 cites W1966059275 @default.
• W1978698911 cites W1992683426 @default.
• W1978698911 cites W1998485425 @default.
• W1978698911 cites W1998709616 @default.
• W1978698911 cites W2008669399 @default.
• W1978698911 cites W2010217335 @default.
• W1978698911 cites W2016225291 @default.
• W1978698911 cites W2021669986 @default.
• W1978698911 cites W2032666422 @default.
• W1978698911 cites W2036072830 @default.
• W1978698911 cites W2040579733 @default.
• W1978698911 cites W2053304186 @default.
• W1978698911 cites W2065613854 @default.
• W1978698911 cites W2066018734 @default.
• W1978698911 cites W2067370432 @default.
• W1978698911 cites W2071519034 @default.
• W1978698911 cites W2072836045 @default.
• W1978698911 cites W2086707089 @default.
• W1978698911 cites W2099136664 @default.
• W1978698911 cites W2112013596 @default.
• W1978698911 cites W2117681601 @default.
• W1978698911 cites W2123742352 @default.
• W1978698911 cites W2130349346 @default.
• W1978698911 cites W2137238711 @default.
• W1978698911 cites W2140254301 @default.
• W1978698911 cites W2157311365 @default.
• W1978698911 cites W2157497892 @default.
• W1978698911 cites W2161445615 @default.
• W1978698911 cites W2169727429 @default.
• W1978698911 doi "https://doi.org/10.1074/jbc.m103607200" @default.
• W1978698911 hasPubMedId "https://pubmed.ncbi.nlm.nih.gov/11723111" @default.
• W1978698911 hasPublicationYear "2002" @default.
• W1978698911 type Work @default.
• W1978698911 sameAs 1978698911 @default.
• W1978698911 citedByCount "242" @default.
• W1978698911 countsByYear W19786989112012 @default.
• W1978698911 countsByYear W19786989112013 @default.
• W1978698911 countsByYear W19786989112014 @default.
• W1978698911 countsByYear W19786989112015 @default.
• W1978698911 countsByYear W19786989112016 @default.
• W1978698911 countsByYear W19786989112017 @default.
• W1978698911 countsByYear W19786989112018 @default.
• W1978698911 countsByYear W19786989112019 @default.
• W1978698911 countsByYear W19786989112020 @default.
• W1978698911 countsByYear W19786989112021 @default.
• W1978698911 countsByYear W19786989112022 @default.
• W1978698911 countsByYear W19786989112023 @default.
• W1978698911 crossrefType "journal-article" @default.
• W1978698911 hasAuthorship W1978698911A5022031402 @default.
• W1978698911 hasAuthorship W1978698911A5024431768 @default.
• W1978698911 hasAuthorship W1978698911A5030513468 @default.
• W1978698911 hasAuthorship W1978698911A5046484870 @default.
• W1978698911 hasAuthorship W1978698911A5063878749 @default.
• W1978698911 hasAuthorship W1978698911A5070388182 @default.
• W1978698911 hasBestOaLocation W19786989111 @default.
• W1978698911 hasConcept C104317684 @default.
• W1978698911 hasConcept C105580179 @default.
• W1978698911 hasConcept C11960822 @default.
• W1978698911 hasConcept C149364088 @default.
• W1978698911 hasConcept C185592680 @default.
• W1978698911 hasConcept C55493867 @default.
• W1978698911 hasConcept C62493599 @default.
• W1978698911 hasConcept C86803240 @default.
• W1978698911 hasConcept C95444343 @default.
• W1978698911 hasConcept C97154109 @default.
• W1978698911 hasConceptScore W1978698911C104317684 @default.
• W1978698911 hasConceptScore W1978698911C105580179 @default.
• W1978698911 hasConceptScore W1978698911C11960822 @default.
• W1978698911 hasConceptScore W1978698911C149364088 @default.
• W1978698911 hasConceptScore W1978698911C185592680 @default.
• W1978698911 hasConceptScore W1978698911C55493867 @default.
• W1978698911 hasConceptScore W1978698911C62493599 @default.
• W1978698911 hasConceptScore W1978698911C86803240 @default.
• W1978698911 hasConceptScore W1978698911C95444343 @default.
• W1978698911 hasConceptScore W1978698911C97154109 @default.
• W1978698911 hasIssue "5" @default.
• W1978698911 hasLocation W19786989111 @default.
• W1978698911 hasOpenAccess W1978698911 @default.
• W1978698911 hasPrimaryLocation W19786989111 @default.
• W1978698911 hasRelatedWork W1986422175 @default.
• W1978698911 hasRelatedWork W1991499410 @default.
• W1978698911 hasRelatedWork W2014007212 @default.
• W1978698911 hasRelatedWork W2047602256 @default.
• W1978698911 hasRelatedWork W2067741767 @default.

• next