Eukaryotic Translation Initiation Factor 5 Functions as a GTPase-activating Protein
2001; Elsevier BV; Volume: 276; Issue: 9 Linguagem: Inglês
10.1074/jbc.m008863200
ISSN1083-351X
AutoresSupratik Das, Rajarshi P. Ghosh, Umadas Maitra,
Tópico(s)RNA regulation and disease
ResumoEukaryotic translation initiation factor 5 (eIF5) forms a complex with eIF2 by interacting with the β subunit of eIF2. This interaction is essential for eIF5-promoted hydrolysis of GTP bound to the 40 S initiation complex. In this work, we show that, in addition to the eIF2β-binding region at the C terminus of eIF5, the N-terminal region of eIF5 is also required for eIF5-dependent GTP hydrolysis. Like other GTPase-activating proteins, eIF5 contains an invariant arginine residue (Arg-15) at its N terminus that is essential for its function. Mutation of this arginine residue to alanine or even to conservative lysine caused a severe defect in the ability of eIF5 to promote GTP hydrolysis from the 40 S initiation complex, although the ability of these mutant proteins to bind to eIF2β remained unchanged. These mutants were also defective in overall protein synthesis as well as in their ability to support cell growth of a ΔTIF5yeast strain. Additionally, alanine substitution mutagenesis of eIF5 defined Lys-33 and Lys-55 as also critical for eIF5 function in vitro and in vivo. The implications of these results in relation to other well characterized GAPs are discussed and provide additional evidence that eIF5 functions as a GTPase-activating protein. Eukaryotic translation initiation factor 5 (eIF5) forms a complex with eIF2 by interacting with the β subunit of eIF2. This interaction is essential for eIF5-promoted hydrolysis of GTP bound to the 40 S initiation complex. In this work, we show that, in addition to the eIF2β-binding region at the C terminus of eIF5, the N-terminal region of eIF5 is also required for eIF5-dependent GTP hydrolysis. Like other GTPase-activating proteins, eIF5 contains an invariant arginine residue (Arg-15) at its N terminus that is essential for its function. Mutation of this arginine residue to alanine or even to conservative lysine caused a severe defect in the ability of eIF5 to promote GTP hydrolysis from the 40 S initiation complex, although the ability of these mutant proteins to bind to eIF2β remained unchanged. These mutants were also defective in overall protein synthesis as well as in their ability to support cell growth of a ΔTIF5yeast strain. Additionally, alanine substitution mutagenesis of eIF5 defined Lys-33 and Lys-55 as also critical for eIF5 function in vitro and in vivo. The implications of these results in relation to other well characterized GAPs are discussed and provide additional evidence that eIF5 functions as a GTPase-activating protein. eukaryotic (translation) initiation factor GTPase-activating protein 5-fluoroorotic acid polymerase chain reaction open reading frame glutathione S-transferase (prokaryotic translation) initiation factor 2 Eukaryotic translation initiation factor 5 (eIF5),1 a monomeric protein of about 49 kDa in mammals (1Raychaudhuri P. Chaudhuri A. Maitra U. J. Biol. Chem. 1985; 260: 2132-2139Abstract Full Text PDF PubMed Google Scholar, 2Chevesich J. Chaudhuri J. Maitra U. J. Biol. Chem. 1993; 268: 20659-20667Abstract Full Text PDF PubMed Google Scholar, 3Das K. Chevesich J. Maitra U. Proc. Natl. Acad. Sci. U. S. A. 1993; 90: 3058-3062Crossref PubMed Scopus (28) Google Scholar) and 46 kDa in the yeastSaccharomyces cerevisiae (4Chakravarti D. Maitra U. J. Biol. Chem. 1993; 268: 10524-10533Abstract Full Text PDF PubMed Google Scholar, 5Chakravarti D. Maiti T. Maitra U. J. Biol. Chem. 1993; 268: 5754-5762Abstract Full Text PDF PubMed Google Scholar), in conjunction with GTP and other initiation factors, plays an essential role in initiation of protein synthesis in eukaryotic cells. In vitro studies using purified initiation factors have shown that the overall initiation reaction proceeds with the initial binding of the initiator Met-tRNAf as the Met-tRNAf·eIF2·GTP ternary complex to a 40 S ribosomal subunit followed by the positioning of the 40 S preinitiation complex (40 S·eIF3·Met-tRNAf·eIF2·GTP) at the initiation AUG codon of the mRNA to form the 40 S initiation complex (40 S·eIF3·mRNA·Met-tRNAf·eIF2·GTP). The initiation factor eIF5 then interacts with the 40 S initiation complex to promote the hydrolysis of ribosome-bound GTP. Hydrolysis of GTP causes the release of eIF2·GDP (and Pi) as well as eIF3 from the 40 S subunit, an event that is essential for the subsequent joining of the 60 S ribosomal subunit to the 40 S complex to form a functional 80 S initiation complex (80 S·mRNA·Met-tRNAf) that is active in peptidyl transfer (for a review, see Refs. 6Merrick W.C. Hershey J.W.B. Hershey J.W.B. Mathews M.B. Sonenberg N. Translational Control. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY1996: 31-69Google Scholar, 7Kozak M. Gene. 1999; 234: 187-208Crossref PubMed Scopus (1124) Google Scholar, 8Maitra U. Stringer E.A. Chaudhuri A. Annu. Rev. Biochem. 1982; 51: 869-900Crossref PubMed Scopus (107) Google Scholar). eIF5-promoted GTP hydrolysis has also been shown to play an important role in the selection of the AUG start codon by the 40 S preinitiation complex (9Huang H.K. Yoon H. Hannig E.M. Donahue T.F. Genes Dev. 1997; 11: 2396-2413Crossref PubMed Scopus (175) Google Scholar). The mammalian cDNA and the yeast gene, designated TIF5, encoding eIF5 have been cloned and expressed as functional proteins inEscherichia coli (4Chakravarti D. Maitra U. J. Biol. Chem. 1993; 268: 10524-10533Abstract Full Text PDF PubMed Google Scholar, 10Chaudhuri J. Das K. Maitra U. Biochemistry. 1994; 33: 4794-4799Crossref PubMed Scopus (37) Google Scholar). Biochemical characterization of eIF5-promoted GTP hydrolysis reaction has shown that hydrolysis of GTP occurs only when eIF5 interacts with GTP bound to eIF2 in the 40 S initiation complex and that eIF5, by itself, does not hydrolyze either free GTP or GTP bound to eIF2 as a Met-tRNAf·eIF2·GTP ternary complex in the absence of 40 S ribosomal subunits (10Chaudhuri J. Das K. Maitra U. Biochemistry. 1994; 33: 4794-4799Crossref PubMed Scopus (37) Google Scholar, 11Chakrabarti A. Maitra U. J. Biol. Chem. 1991; 266: 14039-14045Abstract Full Text PDF PubMed Google Scholar). These observations suggested that eIF5 interacts with one or more components of the 40 S initiation complex to cause hydrolysis of GTP. Subsequent studies showed that eIF5 forms a complex with eIF2 (10Chaudhuri J. Das K. Maitra U. Biochemistry. 1994; 33: 4794-4799Crossref PubMed Scopus (37) Google Scholar), a component of the 40 S initiation complex, and that eIF5·eIF2 complex formation occurs by interaction between the conserved lysine residues at the N-terminal region of eIF2β and the conserved glutamic acid residues at the C-terminal region of eIF5 (12Das S. Maiti T. Das K. Maitra U. J. Biol. Chem. 1997; 272: 31712-31718Abstract Full Text Full Text PDF PubMed Scopus (81) Google Scholar, 13Asano K. Krishnamoorthy T. Phan L. Pavitt G.D. Hinnebusch A.G. EMBO J. 1999; 18: 1673-1688Crossref PubMed Google Scholar, 14Das S. Maitra U. Mol. Cell. Biol. 2000; 20: 3942-3950Crossref PubMed Scopus (57) Google Scholar). More importantly, mutational analysis of the conserved glutamic acid residues in the C-terminal eIF2β-binding region of eIF5 showed that eIF5·eIF2β interaction plays an essential role in eIF5 function in vitro (14Das S. Maitra U. Mol. Cell. Biol. 2000; 20: 3942-3950Crossref PubMed Scopus (57) Google Scholar) as well as in vivo (13Asano K. Krishnamoorthy T. Phan L. Pavitt G.D. Hinnebusch A.G. EMBO J. 1999; 18: 1673-1688Crossref PubMed Google Scholar, 14Das S. Maitra U. Mol. Cell. Biol. 2000; 20: 3942-3950Crossref PubMed Scopus (57) Google Scholar). However, an important question remained as to whether the interaction of eIF5 with eIF2β is sufficient for eIF5-promoted GTP hydrolysis or other regions in eIF5 are also required for its function. In the work presented in this paper, using deletion analysis we show that, in addition to the eIF2β-binding region at the C terminus of eIF5, the N-terminal region of eIF5 is also required for eIF5-promoted GTP hydrolysis. Furthermore, consistent with the hypothesis that eIF5 functions as a GTPase-activating protein (GAP), we observed that, like typical well characterized GAPs, e.g. RasGAPs and RhoGAPs (15Scheffzek K. Lautwein A. Kabsch W. Ahmadian M.R. Wittinghofer A. Nature. 1996; 384: 591-596Crossref PubMed Scopus (144) Google Scholar, 16Scheffzek K. Ahmadian M.R. Wittinghofer A. Trends Biochem. Sci. 1998; 23: 257-262Abstract Full Text Full Text PDF PubMed Scopus (351) Google Scholar), eIF5 also contains an invariant arginine residue (Arg-15) at the N terminus that is essential for eIF5 function both in vitro and in vivo. Mutational analysis of this arginine residue as well as of Lys-33 and Lys-55 of rat eIF5 showed that these residues are critical for eIF5-promoted GTP hydrolysis as well as in the ability of rat eIF5 to functionally substitute for the homologous yeast protein in a ΔTIF5 yeast strain. The implications of these results in understanding the mechanism of eIF5-dependent GTP hydrolysis reaction are discussed. The preparation of 35S-labeled rabbit liver initiator Met-tRNAf (30,000–50,000 cpm/pmol), 40 S and 60 S ribosomal subunits from Artemia salina eggs, purified eIF2 from rabbit reticulocyte lysates, and rabbit anti-rat eIF5 antibodies was described previously (2Chevesich J. Chaudhuri J. Maitra U. J. Biol. Chem. 1993; 268: 20659-20667Abstract Full Text PDF PubMed Google Scholar, 10Chaudhuri J. Das K. Maitra U. Biochemistry. 1994; 33: 4794-4799Crossref PubMed Scopus (37) Google Scholar, 11Chakrabarti A. Maitra U. J. Biol. Chem. 1991; 266: 14039-14045Abstract Full Text PDF PubMed Google Scholar, 17Ghosh S. Chevesich J. Maitra U. J. Biol. Chem. 1989; 264: 5134-5140Abstract Full Text PDF PubMed Google Scholar). The mixture of protease inhibitors added to buffer solutions used during purification of recombinant proteins from bacterial cell extracts consisted of leupeptin (0.5 μg/ml), pepstatin A (0.7 μg/ml), aprotinin (2 μg/ml), and freshly prepared phenylmethylsulfonyl fluoride (1 mm). Deletion mutants of eIF5 were generated by one-stage PCR amplification of eIF5 ORF sequences using pGEX-KG-eIF5 as the template and appropriate oligonucleotide primers containing BamHI/EcoRI overhangs. ABamHI/EcoRI restriction fragment of each PCR-amplified deletion mutant was inserted at the same restriction sites of the vector pGEX-KG. The construct pGEX-KG-Δ-(18–58)eIF5 was generated by three-fragment ligation of a N-terminal fragment obtained by annealing two single-stranded oligonucleotides containingNdeI/ClaI overhangs that correspond to amino acids 1 to 17, a C-terminal PCR fragment digested withClaI/EcoRI that corresponds to amino acids 59 to 430, and the vector pGEX-KG digested withNdeI/EcoRI. The resulting constructs expressed deleted eIF5 mutants as GST fusion proteins. The construction of yeast centromeric plasmids pRS316-TIF5, pTM100-EIF5, and pUB-TIF5R and the haploid yeast strains TMY101 (MATα leu2-3,112 his3-11, 15 ade2-1 trp1-1 ura3-1 can1-100 tif5::TRP1[pRS316-TIF5) and TMY201R (MATα leu2-3, 112 his3-11, 15 ade2-1 trp1-1 ura3-1 can1-100 tif5::TRP1[pUB-TIF5R]) and preparation of media for yeast cell growth have been described previously (18Maiti T. Maitra U. J. Biol. Chem. 1997; 272: 18333-18340Abstract Full Text Full Text PDF PubMed Scopus (30) Google Scholar). Point mutations within the coding sequence of eIF5 present in the yeast centromeric plasmid pTM100-EIF5 or the bacterial expression plasmid pGEX-KG-eIF5 were constructed as described previously (14Das S. Maitra U. Mol. Cell. Biol. 2000; 20: 3942-3950Crossref PubMed Scopus (57) Google Scholar). Expression of wild-type or mutant rat eIF5 proteins in yeast cells was detected by immunoblot analysis as described previously (18Maiti T. Maitra U. J. Biol. Chem. 1997; 272: 18333-18340Abstract Full Text Full Text PDF PubMed Scopus (30) Google Scholar). E. coli XLI Blue cells transformed with recombinant pGEX-KG plasmids containing either the wild-type or mutant eIF5 coding sequences (expressing wild-type or mutant eIF5 as GST fusion proteins) were grown as described previously (14Das S. Maitra U. Mol. Cell. Biol. 2000; 20: 3942-3950Crossref PubMed Scopus (57) Google Scholar). Untagged wild-type and mutant eIF5 proteins were purified from these cells following the procedure described by Das et al. (14Das S. Maitra U. Mol. Cell. Biol. 2000; 20: 3942-3950Crossref PubMed Scopus (57) Google Scholar). The purified proteins were stored in a buffer containing 20 mmTris-HCl, pH 7.5, 100 mm KCl, 0.1 mm EDTA, 1 mm dithiothreitol, and 60% glycerol at −20 °C. The binding assays were carried out as described previously (14Das S. Maitra U. Mol. Cell. Biol. 2000; 20: 3942-3950Crossref PubMed Scopus (57) Google Scholar). A typical binding reaction mixture contained 200 μl of 20 mm potassium phosphate (pH 7.5), 100 mm NaCl, 0.1 mm EDTA, 5 mm 2-mercaptoethanol, 0.5 mmphenylmethylsulfonyl fluoride, 6 μl of a 30% suspension of GSH beads containing bound GST-eIF2β (2 μg of total protein), and about 1 μg of either purified wild-type or mutant eIF5 protein. Yeast cell-free extracts for eIF5-dependent translation of mRNAs were prepared from TMY201R yeast cells as described previously (18Maiti T. Maitra U. J. Biol. Chem. 1997; 272: 18333-18340Abstract Full Text Full Text PDF PubMed Scopus (30) Google Scholar). The 40 S initiation complex containing bound [γ-32P]GTP was prepared and isolated free of unreacted reaction components by sucrose-density centrifugation as described (10Chaudhuri J. Das K. Maitra U. Biochemistry. 1994; 33: 4794-4799Crossref PubMed Scopus (37) Google Scholar, 11Chakrabarti A. Maitra U. J. Biol. Chem. 1991; 266: 14039-14045Abstract Full Text PDF PubMed Google Scholar). Yeast transformations were performed as described by Rose et al. (19Rose M.D. Winston F. Heiter P. Methods in Yeast Genetics: A Laboratory Course Manual. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY1989Google Scholar). Methods for plasmid and genomic DNA preparations, restriction enzyme digestion, DNA ligation, cloning, and bacterial transformations were according to standard protocols (20Sambrook J. Fritsch E.F. Maniatis T. Molecular Cloning: A Laboratory Manual. Cold Spring Harbor Laboratory, Cold Spring Harbor, NY1989Google Scholar). We have shown previously (12Das S. Maiti T. Das K. Maitra U. J. Biol. Chem. 1997; 272: 31712-31718Abstract Full Text Full Text PDF PubMed Scopus (81) Google Scholar, 14Das S. Maitra U. Mol. Cell. Biol. 2000; 20: 3942-3950Crossref PubMed Scopus (57) Google Scholar) that mammalian eIF5 interacts with mammalian eIF2β through a conserved acidic amino acid-rich C-terminal region and that this interaction is necessary for eIF5-promoted hydrolysis of GTP bound to eIF2 in the 40 S initiation complex (40 S·AUG·Met-tRNAf·eIF2·GTP). However, the possibility exists that the interaction between the C-terminal acidic amino acid-rich domain of eIF5 with eIF2β alone is not sufficient for eIF5-promoted GTP hydrolysis. To determine whether other regions of eIF5 are also essential for eIF5 function, we retained the C terminus of eIF5 containing the eIF2β-binding region and carried out truncations at the N terminus (Fig. 1 A). The deletion mutants and wild-type eIF5 were expressed as GST fusion proteins inE. coli XL1-Blue cells and purified as untagged proteins to apparent electrophoretic homogeneity (Fig. 1 B), as described under "Experimental Procedures." These purified mutants were then assayed for their ability to promote hydrolysis of GTP bound to the 40 S initiation complex. Wild-type eIF5 was able to promote rapid and quantitative hydrolysis of GTP as expected (Fig. 1 D). However, when the first 98 amino acids from the N terminus of eIF5 were deleted, the resultant mutant Δ-(1–98)eIF5 failed to promote hydrolysis of GTP (Fig. 1 D), suggesting that the N-terminal end of eIF5 was also essential for eIF5 activity. The mutant Δ-(1–98)eIF5 was, however, able to bind to eIF2β as expected (Fig.1 C). To further map the region at the N terminus of eIF5 that is essential for eIF5 function, we constructed two additional deletion mutants. In the first mutant Δ-(1–17)eIF5, the first 17 amino acids at the N terminus of eIF5 were deleted whereas in the other mutant Δ-(18–58)eIF5, the region between amino acids 18 to 58 was deleted, keeping the rest of the eIF5 ORF intact (Fig. 1 A). Both mutants failed to promote hydrolysis of GTP bound to eIF2 in the 40 S initiation complex (Fig. 1 D) but were able to bind eIF2β (Fig. 1 C). These results indicate that, although the C terminus of eIF5 is involved in binding to eIF2 (via the β subunit of eIF2), this interaction alone is not sufficient for eIF5-promoted GTP hydrolysis. The N terminus of eIF5 is also required for eIF5 activity. The requirement of both the N terminus and the C terminus of eIF5 for its activity is analogous to that of other well characterized GAPs. Typical GAPs, e.g. RhoGAPs and RasGAPs, contain sequence motifs that are necessary for their GTPase-stimulating activity in addition to motifs that are necessary for binding to their respective GTPase proteins (16Scheffzek K. Ahmadian M.R. Wittinghofer A. Trends Biochem. Sci. 1998; 23: 257-262Abstract Full Text Full Text PDF PubMed Scopus (351) Google Scholar). In the case of eukaryotic translation initiation, it is the eIF2 molecule that binds GTP and presumably acts as a GTPase, whereas eIF5 acts as the GAP. The C-terminal domain of eIF5 contains the sequence motifs required for binding of eIF5 to eIF2, the GTPase protein (12Das S. Maiti T. Das K. Maitra U. J. Biol. Chem. 1997; 272: 31712-31718Abstract Full Text Full Text PDF PubMed Scopus (81) Google Scholar, 13Asano K. Krishnamoorthy T. Phan L. Pavitt G.D. Hinnebusch A.G. EMBO J. 1999; 18: 1673-1688Crossref PubMed Google Scholar, 14Das S. Maitra U. Mol. Cell. Biol. 2000; 20: 3942-3950Crossref PubMed Scopus (57) Google Scholar). It is likely that the activation domain of eIF5 is present at the N terminus and contains sequence motifs required for the GTPase-stimulating activity of eIF5. Comparison of the N-terminal region of eIF5 from different species shows that this region of eIF5 is highly conserved (Fig.2). The activation domains of GAPs belonging to the well characterized families of RasGAPs and RhoGAPs have been shown to contain "arginine-finger" motifs consisting of an invariant arginine residue at the N terminus of their catalytic domain that are required for their ability to stimulate GTP hydrolysis (Ref. 16Scheffzek K. Ahmadian M.R. Wittinghofer A. Trends Biochem. Sci. 1998; 23: 257-262Abstract Full Text Full Text PDF PubMed Scopus (351) Google Scholar, see also TableI). In rat eIF5, there are two invariant arginine residues at the N-terminal end at positions 15 and 48. The importance of these arginine residues in eIF5 function was investigated by alanine substitution mutagenesis of these residues. These mutant eIF5 proteins were initially examined for their ability to substitute for yeast eIF5 in a ΔTIF5 haploid yeast strain. The rationale behind this strategy was that, if these invariant arginine residues were required for eIF5 function, eIF5 proteins containing mutations at these residues would be lethal in yeast cells. All mutations were carried out in rat eIF5, because we use mammalian factors in our in vitro assay systems to measure eIF5 function. It should be noted that mammalian eIF5 functionally substitutes for yeast eIF5 in sustaining yeast cell growth and viability (18Maiti T. Maitra U. J. Biol. Chem. 1997; 272: 18333-18340Abstract Full Text Full Text PDF PubMed Scopus (30) Google Scholar).Table ISequence homology of arginine-finger motif in eIF5 and other GAPsFamilyArginine-finger motifGenBank™ accession no.RasGAPsrnp120777TLFRATT783Q29594hsp120786TLFRATT792M23379hsNF11273TLFRGNS1279M89914RhoGAPshsp50GAP279GIFRRSA285Z23024hsBcr1087GIYRVSG1093P11274hsGRP680GIYRISG686Q12979eIF5Rat12QFYRYKM18NP064460Yeast12PFYRYKM18P38431Maize14AFYRYKM20P55876The sequences of "arginine-finger" motifs of RasGAPs and RhoGAPs are from Ref. 16Scheffzek K. Ahmadian M.R. Wittinghofer A. Trends Biochem. Sci. 1998; 23: 257-262Abstract Full Text Full Text PDF PubMed Scopus (351) Google Scholar. The invariant arginine residue is shown asboldface letters. Open table in a new tab The sequences of "arginine-finger" motifs of RasGAPs and RhoGAPs are from Ref. 16Scheffzek K. Ahmadian M.R. Wittinghofer A. Trends Biochem. Sci. 1998; 23: 257-262Abstract Full Text Full Text PDF PubMed Scopus (351) Google Scholar. The invariant arginine residue is shown asboldface letters. To identify the invariant arginine residue in eIF5, the eIF5-Ala mutants R15A and R48A were generated using a PCR-based site-directed mutagenesis protocol using the LEU2-based yeast expression plasmid pTM100-EIF5 as the template as described under "Experimental Procedures." The wild-type rat eIF5 ORF (gene designation EIF5) in the CENplasmid pTM100-EIF5 is under the transcriptional control of the galactose-inducible GAL1 promoter (18Maiti T. Maitra U. J. Biol. Chem. 1997; 272: 18333-18340Abstract Full Text Full Text PDF PubMed Scopus (30) Google Scholar). To test the function of the rat eIF5-Ala mutants in yeast cells, both the wild-type and the mutant recombinant expression plasmids and the parental vector plasmid pTM100 were transformed separately into the haploid yeast strain TMY101. (The strain TMY101 carries an inactive TIF5 allele disrupted with the TRP1 marker gene and is kept viable by maintenance of an URA3-based CEN plasmid pRS316-TIF5 in which yeast eIF5 is expressed from its natural promoter.) Trp+Ura+Leu+transformants were selected on SGal-Trp-Leu-Ura plates and then replica-plated onto similar SGal plates, which also contained uracil and 5-fluoroorotic acid (5-FOA) to select against retention of theURA3-based plasmid pRS316-TIF5, expressing wild-type yeast eIF5. Fig. 3 A shows that the yeast strain TMY101 transformed with plasmid pTM100-EIF5, expressing wild-type rat eIF5, grew on 5-FOA plates, in agreement with the results reported previously (18Maiti T. Maitra U. J. Biol. Chem. 1997; 272: 18333-18340Abstract Full Text Full Text PDF PubMed Scopus (30) Google Scholar), while cells transformed with the parental vector plasmid pRS315 failed to grow, as expected. Under the same conditions TMY101 cells transformed with pTM100-EIF5(R15A), carrying a single point mutation at Arg-15 (Arg to Ala) in rat eIF5, failed to grow on 5-FOA plates (Fig. 3 A, right panel) while cells expressing alanine substitution mutation at Arg-48 of rat eIF5 from pTM100-EIF5(R48A) grew as well as the cells expressing wild-type eIF5 from the plasmid pTM100-EIF5. It is known that even a conservative mutation of the invariant arginine residue in RasGAP to lysine dramatically affects its GAP activity (16Scheffzek K. Ahmadian M.R. Wittinghofer A. Trends Biochem. Sci. 1998; 23: 257-262Abstract Full Text Full Text PDF PubMed Scopus (351) Google Scholar). We also observed that when TMY101 cells were transformed with the recombinant plasmid pTM100-EIF5(R15K), which expressed the mutant eIF5 protein in which Arg-15 was mutated to Lys, and the resulting transformants were tested for their ability to grow on 5-FOA plates, they failed to grow (Fig.3 A, right panel). Taken together these results suggest that Arg-15 in rat eIF5 plays a critical role in eIF5 function in maintaining yeast cell growth and viability. In RasGAPs and RhoGAPs, a "secondary" positively charged residue (Arg in RasGAP and Lys in RhoGAP) is also necessary for their GTP hydrolysis-stimulating activity. To identify such a residue in eIF5, we carried out mutagenesis of other invariant positively charged residues at the N-terminal region of rat eIF5 (Fig. 2). The residues Lys-24, Lys-33, Asn-38, and Lys-55 in rat eIF5 were each mutated to alanine. Once again, plasmid shuffling technique was used to test the effect of these mutations on yeast cell growth and viability. Yeast cells expressing either mutant eIF5 K33A (Lys-33 was mutated to Ala) or mutant eIF5 K55A (Lys-55 was mutated to Ala) failed to grow on 5-FOA plates (Fig. 3 A, right panel). In contrast, alanine substitution mutation at Lys-24 and Asn-38 of rat eIF5 did not affect cell growth on 5-FOA plates (data not shown). These results indicate that Lys-33 and Lys-55 in rat eIF5 may also play an important role in eIF5 function. The functional defect of mutant rat eIF5 proteins R15A, R15K, K33A, and K55A was not due to lack of expression of mutant eIF5 in yeast cells. When cell extracts were prepared from Trp+Ura+Leu+ transformants harboring both pRS316-TIF5 and pTM100-EIF5(wild-type or mutant) plasmids and analyzed by Western blotting using rabbit anti-rat eIF5 antibodies, the mutant eIF5 proteins R15A (lane d), R15K (lane e), K33A (lane f), and K55A (lane g) were expressed at levels comparable to wild-type eIF5 (lane b). Extracts prepared from TMY101 cells harboring the plasmid pRS316-TIF5 and the vector plasmid pRS315 and expressing only wild-type yeast eIF5 did not show any immunoreactive polypeptide band with anti-rat eIF5 antibodies, as expected (lane c). It should be noted that rabbit anti-rat eIF5 antibodies do not recognize yeast eIF5 (18Maiti T. Maitra U. J. Biol. Chem. 1997; 272: 18333-18340Abstract Full Text Full Text PDF PubMed Scopus (30) Google Scholar). The eIF5 mutants R15A, R15K, K33A, and K55A, which were unable to complement in vivo a genetic disruption in the chromosomal copy ofTIF5, were initially tested to confirm that these mutations did not cause a defect in the ability of eIF5 to bind eIF2β. For this purpose, the mutant eIF5 proteins were expressed in bacteria and purified to apparent homogeneity (Fig.4 A) as described under "Experimental Procedures." The ability of each purified mutant protein and wild-type eIF5 to bind eIF2β fused to GST and immobilized on GSH-Sepharose beads was examined. Fig. 4 B shows that mutant eIF5 protein R15A (lane d), R15K (lane e), K33A (lane f), and K55A (lane g) were all able to bind GST-eIF2β, although the binding efficiency of K55A is somewhat reduced (see legend to Fig. 4 B). As expected, wild-type eIF5 did not bind to GST immobilized on GSH-Sepharose beads (lane c). The purified mutant eIF5 proteins were also tested for their ability to promote in vitro hydrolysis of GTP bound to the 40 S initiation complex. As expected, wild-type rat eIF5 promoted rapid hydrolysis of [γ-32P]GTP bound to the 40 S initiation complex (Fig. 4 C). In contrast, under similar experimental conditions the mutant eIF5 protein R15A was unable to hydrolyze [γ-32P]GTP bound to the 40 S initiation complex (Fig.4 C). When the mutant eIF5 proteins R15K, K33A, and K55A were tested in the GTP hydrolysis reaction, we observed between 2- and 6-fold reduction in GTP hydrolysis activity (Fig. 4 C). Taken together, these results suggest that the eIF5 mutants R15A, R15K, K33A, and K55A were defective in their ability to hydrolyze GTP bound to the 40 S initiation complex, although they retained the ability to bind eIF2β. The small reduction in binding efficiency of mutant K55A as compared with wild-type eIF5 (see legend to Fig. 4 B) is presumably due to some conformational change induced in this eIF5 mutant. However, it should be noted that these binding analyses are at best semi-quantitative and stoichiometric, whereas eIF5-promoted GTP hydrolysis is highly catalytic. Additionally, it is known that the N terminus of eIF5 is not required for binding to eIF2β (14Das S. Maitra U. Mol. Cell. Biol. 2000; 20: 3942-3950Crossref PubMed Scopus (57) Google Scholar). Thus, the small reduction in the binding efficiency of K55A to eIF2β is unlikely to account for the reduced rate of GTP hydrolysis promoted by this eIF5 mutant. We have previously demonstrated that, in a yeast cell-free translation system depleted of endogenous eIF5, translation of mRNAs in vitro can be restored by the addition of exogenous yeast or mammalian eIF5 (18Maiti T. Maitra U. J. Biol. Chem. 1997; 272: 18333-18340Abstract Full Text Full Text PDF PubMed Scopus (30) Google Scholar). We used such an eIF5-depleted yeast cell-free translation system to test the ability of the eIF5 mutants to restore translation of yeast mRNAs in vitro. Fig. 5 shows that these extracts showed poor translation activity in the absence of exogenously added eIF5. As expected, translation in these extracts was restored by the addition of purified wild-type rat eIF5 (Fig. 5). In the absence of mRNA, the addition of eIF5 had virtually no effect (data not shown). In contrast to wild-type rat eIF5, the mutant eIF5 protein R15A was virtually inactive in restoring translation of mRNAs in these extracts, whereas the mutant eIF5 proteins R15K, K33A, and K55A were about 50, 40, and 20% as active as the wild-type protein, respectively (Fig. 5). Several lines of evidence presented in previous reports (1Raychaudhuri P. Chaudhuri A. Maitra U. J. Biol. Chem. 1985; 260: 2132-2139Abstract Full Text PDF PubMed Google Scholar,11Chakrabarti A. Maitra U. J. Biol. Chem. 1991; 266: 14039-14045Abstract Full Text PDF PubMed Google Scholar, 12Das S. Maiti T. Das K. Maitra U. J. Biol. Chem. 1997; 272: 31712-31718Abstract Full Text Full Text PDF PubMed Scopus (81) Google Scholar, 13Asano K. Krishnamoorthy T. Phan L. Pavitt G.D. Hinnebusch A.G. EMBO J. 1999; 18: 1673-1688Crossref PubMed Google Scholar, 14Das S. Maitra U. Mol. Cell. Biol. 2000; 20: 3942-3950Crossref PubMed Scopus (57) Google Scholar) and summarized below suggested that eIF5 functions as a GTPase-activating protein (GAP) in translation initiation. First, eIF5 promotes GTP hydrolysis only when the protein interacts with the 40 S initiation complex (40 S·AUG·Met-tRNAf·eIF2·GTP). eIF5, by itself, neither binds nor hydrolyzes either free GTP or GTP bound as a Met-tRNAf·eIF2·GTP ternary complex in the absence of 40 S ribosomal subunits. Second, eIF5 interacts with the heterotrimeric GTP-binding protein eIF2 (10Chaudhuri J. Das K. Maitra U. Biochemistry. 1994; 33: 4794-4799Crossref PubMed Scopus (37) Google Scholar). This interaction, which occurs between the conserved lysine residues at the N-terminal region of eIF2β and the conserved glutamic acid residues at the C-terminal region of eIF5 (12Das S. Maiti T. Das K. Maitra U. J. Biol. Chem. 1997; 272: 31712-31718Abstract Full Text Full Text PDF PubMed Scopus (81) Google Scholar, 13Asano K. Krishnamoorthy T. Phan L. Pavitt G.D. Hinnebusch A.G. EMBO J. 1999; 18: 1673-1688Crossref PubMed Google Scholar, 14Das S. Maitra U. Mol. Cell. Biol. 2000; 20: 3942-3950Crossref PubMed Scopus (57) Google Scholar), is essential for eIF5 activity both in vitro (14Das S. Maitra U. Mol. Cell. Biol. 2000; 20: 3942-3950Crossref PubMed Scopus (57) Google Scholar) and in vivo in yeast cells (13Asano K. Krishnamoorthy T. Phan L. Pavitt G.D. Hinnebusch A.G. EMBO J. 1999; 18: 1673-1688Crossref PubMed Google Scholar, 14Das S. Maitra U. Mol. Cell. Biol. 2000; 20: 3942-3950Crossref PubMed Scopus (57) Google Scholar). In this paper, we present additional evidence that shows that eIF5 indeed functions as a GAP in translation initiation. Typical GAPs like RhoGAPs and RasGAPs, which have been characterized extensively, contain "arginine-finger" motifs consisting of an invariant arginine residue at the N terminus of their catalytic domains that are necessary for their GTPase-stimulating activity in addition to motifs that are necessary for interacting with their respective G proteins (Ref. 16Scheffzek K. Ahmadian M.R. Wittinghofer A. Trends Biochem. Sci. 1998; 23: 257-262Abstract Full Text Full Text PDF PubMed Scopus (351) Google Scholar and Table I). Biochemical and structural studies carried out with RasGAP have shown that in RasGAP the arginine at position 789 is the "primary" element required for stimulating GTP hydrolysis (16Scheffzek K. Ahmadian M.R. Wittinghofer A. Trends Biochem. Sci. 1998; 23: 257-262Abstract Full Text Full Text PDF PubMed Scopus (351) Google Scholar). The positive charge of the invariant arginine residue stabilizes the transition state of the GTP hydrolysis reaction by neutralizing the negative charges developing in the transition state (16Scheffzek K. Ahmadian M.R. Wittinghofer A. Trends Biochem. Sci. 1998; 23: 257-262Abstract Full Text Full Text PDF PubMed Scopus (351) Google Scholar). In addition, hydrogen bonding between Ras-GDP and GAP334 (the catalytic domain of RasGAP) also stabilizes the transition state (21Scheffzek K. Ahmadian M.R. Kabsch W. Wiesmuller L. Lautwein A. Schmitz F. Wittinghofer A. Science. 1997; 277: 333-338Crossref PubMed Scopus (1192) Google Scholar). A secondary arginine residue, Arg-903 in GAP334 (Lys-122 in the case of RhoGAP) stabilizes the finger loop carrying the primary arginine residue (16Scheffzek K. Ahmadian M.R. Wittinghofer A. Trends Biochem. Sci. 1998; 23: 257-262Abstract Full Text Full Text PDF PubMed Scopus (351) Google Scholar). Comparison of the N-terminal amino acid sequences of eIF5 from different species revealed the presence of invariant arginine residues at positions 15 and 48 that are conserved in all species of eIF5 identified so far (Fig. 2). Like RasGAP and RhoGAP (22Gamblin S.J. Smerdon S.J. Curr. Opin. Struct. Biol. 1998; 8: 195-201Crossref PubMed Scopus (65) Google Scholar), the Arg-15 residue in eIF5 is preceded by two conserved hydrophobic residues, phenylalanine and tyrosine. It has been suggested that the function of these two hydrophobic residues is to anchor the catalytic loop into the hydrophobic core of the GAP (22Gamblin S.J. Smerdon S.J. Curr. Opin. Struct. Biol. 1998; 8: 195-201Crossref PubMed Scopus (65) Google Scholar). Thus, in eIF5, Arg-15 is in a better sequence context than Arg-48. Based on this analysis, we observed that mutation of Arg-15 in rat eIF5 to Ala or even to conservative Lys resulted in mutant proteins that were unable to substitute for yeast eIF5 in maintaining yeast cell growth and viability of a ΔTIF5 yeast strain (Fig. 3). In contrast, ΔTIF5 yeast cells expressing eIF5 mutant R48A were able to maintain growth and viability (Fig. 3). In agreement with these results obtained in vivo, we observed that the purified eIF5 mutants R15A and R15K were severely defective in their ability to promote GTP hydrolysis. Whereas mutant R15A showed virtually no activity, mutant R15K showed a low level of activity (<20%). Thus, although the lysine residue at position 15 of eIF5 mutant R15K was able to compensate for the positive charge of arginine, it was probably unable to correctly position itself with respect to the guanine nucleotide or the active site of the GTPase protein (presumably the conserved histidine residue at position 138 in the γ subunit of human eIF2 (23Gaspar N.J. Kinzy T.G. Scherer B.J. Humbelin M. Hershey J.W. Merrick W.C. J. Biol. Chem. 1994; 269: 3415-3422Abstract Full Text PDF PubMed Google Scholar)) in the transition state. We also observed that mutation of Lys-33 and Lys-55 of rat eIF5 to alanine also caused a severe defect in eIF5 function both in vitro and in vivo in yeast cells. In analogy with RasGAP and RhoGAP, it is likely that either Lys-33 or Lys-55 or both in rat eIF5 constitute the secondary element required for eIF5 GAP function. An important property of eIF5-dependent GTP hydrolysis reaction is that, in addition to eIF2 and eIF5, 40 S subunits also play an essential role in GTP hydrolysis. This is analogous to the essential requirement of 50 S ribosomal subunits in IF2-, EF-Tu-, and EF-G-catalyzed GTP hydrolysis reaction in prokaryotes (7Kozak M. Gene. 1999; 234: 187-208Crossref PubMed Scopus (1124) Google Scholar, 8Maitra U. Stringer E.A. Chaudhuri A. Annu. Rev. Biochem. 1982; 51: 869-900Crossref PubMed Scopus (107) Google Scholar). A similar requirement of ribosomes in GTP hydrolysis reaction catalyzed by the signal recognition peptide receptor subunit SRβ has been reported (24Bacher G. Pool M. Dobberstein B. J. Cell Biol. 1999; 146: 723-730Crossref PubMed Scopus (43) Google Scholar). Biochemical studies of GAP mutants and crystal structure analysis of GTPase·GAP complexes (22Gamblin S.J. Smerdon S.J. Curr. Opin. Struct. Biol. 1998; 8: 195-201Crossref PubMed Scopus (65) Google Scholar) have shown that GAPs provide two functions to the transition state of a GTP hydrolysis reaction. First, it physically binds to the G protein and causes a conformational change in the G protein resulting in the stabilization of the switch I and switch II regions in the G protein (22Gamblin S.J. Smerdon S.J. Curr. Opin. Struct. Biol. 1998; 8: 195-201Crossref PubMed Scopus (65) Google Scholar). This results in the correct positioning of the active site glutamine residue of the G protein in the transition state and its activation. Second, GAP also provides an arginine residue to the pentacoordinate transition state and stabilizes it. In the case of eIF5-dependent GTP hydrolysis reaction, although eIF5 provides the essential arginine residue of the GAP, the question arises: how are the switch regions in eIF2γ (the polypeptide that contains the consensus GTP binding domains (23Gaspar N.J. Kinzy T.G. Scherer B.J. Humbelin M. Hershey J.W. Merrick W.C. J. Biol. Chem. 1994; 269: 3415-3422Abstract Full Text PDF PubMed Google Scholar, 25Hannig E.M. Cigan A.M. Freeman B.A. Kinzy T.G. Mol. Cell. Biol. 1993; 13: 506-520Crossref PubMed Google Scholar) and is the presumed GTPase) stabilized in the absence of direct physical interaction between eIF5 and eIF2γ? Clearly, although interaction between eIF5 and eIF2β anchors eIF5 to eIF2 and eIF2β is also known to physically interact with eIF2γ (26Thompson G.M. Pacheco E. Melo E.O. Castilho B.A. Biochem. J. 2000; 347: 703-709Crossref PubMed Scopus (36) Google Scholar), the fact remains that these interactions, although necessary, are not sufficient for GTP hydrolysis, because eIF5 cannot promote hydrolysis of GTP bound to eIF2 in the Met-tRNAf·eIF2·GTP ternary complex in the absence of 40 S ribosomal subunits. Thus, the possibility exists that the 40 S subunit directly interacts with eIF2γ and causes the conformational change necessary to stabilize the switch regions in eIF2γ. In this respect, the 40 S ribosomal subunit also behaves as a GAP in eIF5-dependent GTP hydrolysis reaction during translation initiation. Such a distribution of dual functions of a GAP in two different proteins has also been suggested for GTP hydrolysis reaction mediated by ADP-ribosylation factor and Gα signal transduction protein (27De Vries L. Zheng B. Fischer T. Elenko E. Farquhar M.G. Annu. Rev. Pharmacol. Toxicol. 2000; 40: 235-271Crossref PubMed Scopus (504) Google Scholar). Finally, both mammalian and yeast eIF5 have been reported (3Das K. Chevesich J. Maitra U. Proc. Natl. Acad. Sci. U. S. A. 1993; 90: 3058-3062Crossref PubMed Scopus (28) Google Scholar, 4Chakravarti D. Maitra U. J. Biol. Chem. 1993; 268: 10524-10533Abstract Full Text PDF PubMed Google Scholar) to contain sequence motifs that are somewhat homologous to G1–G4 domains that are characteristic of members of the GTPase superfamily (28Bourne H.R. Sanders D.A. McCormick F. Nature. 1991; 349: 117-127Crossref PubMed Scopus (2673) Google Scholar). However, unlike the well characterized GTPases, these domains in eIF5 are quite "imperfect" (Table II). The conserved G1 domain GXXGXGK(S/T) is present in rat eIF5 as 27GKGNGIKT34 and in yeast eIF5 as27GRGNGIKT34. Clearly, insertion of an extra amino acid, isoleucine, into the consensus phosphate-binding loop makes the G1 domain an imperfect consensus sequence. Additionally, the first four amino acids in the G1 domain in eIF5 are GXGX instead of the consensus GXXG. Furthermore, the spacings between the four domains are also not conserved in eIF5 as they are in members of the GTPase superfamily. Finally, unlike GTPases like Ras, where a conserved glutamine residue (histidine in some GTPases) in the G3 domain is the active site of the GTPase (28Bourne H.R. Sanders D.A. McCormick F. Nature. 1991; 349: 117-127Crossref PubMed Scopus (2673) Google Scholar), such a conserved positively charged residue at a similar position is absent in eIF5. These observations along with the results of our mutational analysis (data not shown) showing that mutation of conserved residues in the G1 and G4 domains of eIF5 did not affect eIF5 function in vitro and in vivo in yeast cells suggested that these domains are not critical for eIF5 function.Table IISequence homology of eIF5 with proteins of the GTPase superfamilyMotifsG1G2G3G4ConsensusGXXGXGK(S/T)D-(X)1-TDXXG(N/T)(K/Q)XDH-Ras10GAGGVGKS1733D-(X)1-T3557DTAG60116NKCD119RAS217GGGGVGKS2440D-(X)1-T4264DTAG67123NKSD126E. coliEF-Tu18GHVDHGKT2550D-(X)10-T6180DCPG83135NKCD138Rat eIF527GKGNGIKT3490D-(X)10-T107199DDWG202419NKDD422Yeast eIF527GRGNGIKT3490D-(X)16-T107185DGTG188246TQLD249Comparison of the putative G1–G4 GTP-binding domains in the eIF5 amino acid sequence with the conserved sequence motifs in the GTPase superfamily (28Bourne H.R. Sanders D.A. McCormick F. Nature. 1991; 349: 117-127Crossref PubMed Scopus (2673) Google Scholar). The sequences of rat and yeast eIF5 are from Refs. 3Das K. Chevesich J. Maitra U. Proc. Natl. Acad. Sci. U. S. A. 1993; 90: 3058-3062Crossref PubMed Scopus (28) Google Scholarand 4Chakravarti D. Maitra U. J. Biol. Chem. 1993; 268: 10524-10533Abstract Full Text PDF PubMed Google Scholar, respectively. The other sequences are from Ref. 28Bourne H.R. Sanders D.A. McCormick F. Nature. 1991; 349: 117-127Crossref PubMed Scopus (2673) Google Scholar. Open table in a new tab Comparison of the putative G1–G4 GTP-binding domains in the eIF5 amino acid sequence with the conserved sequence motifs in the GTPase superfamily (28Bourne H.R. Sanders D.A. McCormick F. Nature. 1991; 349: 117-127Crossref PubMed Scopus (2673) Google Scholar). The sequences of rat and yeast eIF5 are from Refs. 3Das K. Chevesich J. Maitra U. Proc. Natl. Acad. Sci. U. S. A. 1993; 90: 3058-3062Crossref PubMed Scopus (28) Google Scholarand 4Chakravarti D. Maitra U. J. Biol. Chem. 1993; 268: 10524-10533Abstract Full Text PDF PubMed Google Scholar, respectively. The other sequences are from Ref. 28Bourne H.R. Sanders D.A. McCormick F. Nature. 1991; 349: 117-127Crossref PubMed Scopus (2673) Google Scholar.
Referência(s)