Raf kinases mediate the phosphorylation of eukaryotic translation elongation factor 1A and regulate its stability in eukaryotic cells

We identified eukaryotic translation elongation factor 1A (eEF1A) Raf-mediated phosphorylation sites and defined their role in the regulation of eEF1A half-life and of apoptosis of human cancer cells. Mass spectrometry identified in vitro S21 and T88 as phosphorylation sites mediated by B-Raf but not C-Raf on eEF1A1 whereas S21 was phosphorylated on eEF1A2 by both B- and C-Raf. Interestingly, S21 belongs to the first eEF1A GTP/GDP-binding consensus sequence. Phosphorylation of S21 was strongly enhanced when both eEF1A isoforms were preincubated prior the assay with C-Raf, suggesting that the eEF1A isoforms can heterodimerize thus increasing the accessibility of S21 to the phosphate. Overexpression of eEF1A1 in COS 7 cells confirmed the phosphorylation of T88 also in vivo. Compared with wt, in COS 7 cells overexpressed phosphodeficient (A) and phospho-mimicking (D) mutants of eEF1A1 (S21A/D and T88A/D) and of eEF1A2 (S21A/D), resulted less stable and more rapidly proteasome degraded. Transfection of S21 A/D eEF1A mutants in H1355 cells increased apoptosis in comparison with the wt isoforms. It indicates that the blockage of S21 interferes with or even supports C-Raf induced apoptosis rather than cell survival. Raf-mediated regulation of this site could be a crucial mechanism involved in the functional switching of eEF1A between its role in protein biosynthesis and its participation in other cellular processes.

C-Raf antagonizes apoptosis induced by IFN-α in human lung cancer cells by phosphorylation and increase of the intracellular content of elongation factor 1A

Interferon alpha (IFNalpha) induces both apoptosis and a counteracting epidermal growth factor Erk-dependent survival response in cancer cells. In this report, IFNalpha increased eukaryotic elongation factor 1A (eEF-1A) protein expression by inhibition of eEF-1A degradation via a proteasome-dependent pathway. The reduction of the expression level of eEF-1A by RNA interference enhanced the apoptosis induced by IFNalpha on the same cells. Moreover, IFNalpha induced the phosphorylation of both serine and threonine in eEF-1A. These effects were paralleled by an increased co-immunoprecipitation and colocalization of eEF-1A with C-Raf. The suppression of C-Raf kinase activity with the inhibitor BAY 43-9006 completely antagonized the increase of both eEF-1A phosphorylation and expression and of C-Raf/eEF-1A colocalization induced by IFNalpha and enhanced apoptosis and eEF-1A ubiquitination. Cell transfection with the mutated K48R ubiquitin increased EF-1A expression and desensitized tumor cells to the modulating effects of IFNalpha. The dynamic simulation of 3Dstructure of eEF-1A identified putative serine and threonine phosphorylation sites. In conclusion, the interaction between eEF-1A and C-Raf increases eEF-1A stability and induces a survival activity.

Gastrokine 1 expression in patients with and without Helicobacter pylori infection

BACKGROUND

To understand the molecular changes underlying Helicobacter pylori-related gastric diseases is mandatory to prevent gastric cancer. Proteomic technology is providing a rapid expansion of the basic knowledge, particularly in the discovery of new biomarkers involved in the tumourigenesis.

AIM

To characterise changes in protein expression level of the gastric mucosa in H. pylori-infected patients.

METHODS

The population enrolled comprised 41 dyspeptic patients. Proteins extracted from gastric mucosal specimens were analysed by 2-dimensional electrophoresis, sequenced by MALDI-TOF and identified by Edman's degradation.

RESULTS

Twenty-one out of 41 patients had H. pylori infection of whom 17 had anti-CagA IgG antibodies. Several proteins were identified, of which Rho guanosine diphosphatase dissociation inhibitor alpha and heat shock protein 27 increased and glutathione transferase and antrum mucosa protein-18 decreased in H. pylori-positive in respect to H. pylori-negative patients. Interestingly, antrum mucosa protein-18, currently referred as gastrokine-1, showed two isoforms differing in the first N-terminal amino acid residue. Both gastrokine-1 isoforms were observed in the H. pylori-negative group whereas a lower expression or even absence of the gastrokine-1 basic isoform was found in a subgroup (7/21) of H. pylori-positive patients with moderate-severe gastritis.

CONCLUSION

Our study demonstrated the presence of gastrokine-1 isoforms of which the basic isoform was reduced in a subset of patients with H. pylori infection.

A 35 kDa NAD(P)H oxidase previously isolated from the archaeon Sulfolobus solfataricus is instead a thioredoxin reductase

A thioredoxin reductase (TrxR) has been identified in the hyperthermophilic archaeon Sulfolobus solfataricus (Ss). This enzyme is a homodimeric flavoprotein that was previously identified as NADH oxidase in the same micro-organism ('Biotechnol. Appl. Biochem. 23 (1996) 47'). The primary structure of SsTrxR is made of 323 amino acid residues and contains two putative betaalphabeta regions for the binding of FAD, and a NADP(H) binding consensus sequence in the proximity of a CXXC motif. These findings indicate that SsTrxR is structurally related to the class II of the pyridine nucleotide-disulphide oxidoreductases family. Moreover, the enzyme exhibits a NADP(H) dependent thioredoxin reductase activity requiring the presence of FAD. Surprisingly, the reductase activity of SsTrxR is reduced in the presence of a specific inhibitor of mammalian TrxR. This finding demonstrates that the archaeal enzyme, although structurally related to eubacterial TrxR, is functionally closer to eukaryal enzymes. Experimental evidences indicate that a disulphide bridge is required for the reductase but also for the NADH oxidase activity of the enzyme. These results are further supported by the significantly reduced activities exerted by the C147A mutant. The integrity of the CXXC motif is also involved in the stability of the enzyme.

The translation elongation factor 1A in tumorigenesis, signal transduction and apoptosis: review article

An increasing number of evidences suggest the involvement of the eukaryotic elongation factor 1A, a core component of the protein synthesis machinery, at the onset of cell transformation. In fact, eEF1A is shown to be up-regulated in cell death; moreover, it seems to be involved in the regulation of ubiquitin-mediated protein degradation. In addition, eEF1A undergoes several post-translational modifications, mainly phosphorylation and methylation, that generally influence the activity of the protein. This article summarizes the present knowledges on the several extra-translational roles of eEF1A also in order to understand as the protein synthesis regulatory mechanisms could offer tools for cancer intervention.

Psychrophilic elongation factor Tu from the antarctic Moraxella sp. Tac II 25: biochemical characterization and cloning of the encoding gene

The elongation factor Tu was isolated from a psychrophilic eubacterial Antarctic Moraxella strain (MoEF-Tu) and its molecular and functional properties were determined. It catalyzed the synthesis of poly(Phe) and bound specifically guanine nucleotides with an affinity for GDP about 12-fold higher than that for GTP. The affinity toward guanine nucleotides was lower than that of other eubacterial EF-Tu. The intrinsic GTPase activity of MoEF-Tu was hardly detectable but was accelerated by 2 orders of magnitude in the presence of the antibiotic kirromycin (GTPase(k)). Such a property resembled Escherichia coli EF-Tu (EcEF-Tu) even though the affinity of MoEF-Tu for the antibiotic was lower. MoEF-Tu showed a thermophilicity higher than that of EcEF-Tu; its temperature for half-denaturation was 44 degrees C. The MoEF-Tu encoding gene corresponding to E. coli tufA was cloned and sequenced. The translated protein had a calculated molecular weight of 43 288 and contained the GTP-binding sequence motifs. Concerning its primary structure, MoEF-Tu showed sequence identity with E. coli and Thermus thermophilus EF-Tu equal to 84% and 74%, respectively, while the identity with EF-1 alpha from the archaeon Sulfolobus solfataricus was equal to 32%.

A NAD(P)H oxidase isolated from the archaeon Sulfolobus solfataricus is not homologous with another NADH oxidase present in the same microorganism. Biochemical characterization of the enzyme and cloning of the encoding gene

A NAD(P)H oxidase has been isolated from the archaeon Sulfolobus solfataricus. The enzyme is a homodimer with M(r) 38,000 per subunit (SsNOX38) containing 1 FAD molecule/subunit. It oxidizes NADH and, less efficiently, NADPH with the formation of hydrogen peroxide. The enzyme was resistant against chemical and physical denaturating agents. The temperature for its half-denaturation was 93 and 75 degrees C in the absence or presence, respectively, of 8 M urea. The enzyme did not show any reductase activity. The SsNOX38 encoding gene was cloned and sequenced. It accounted for a product of 36.5 kDa. The translated amino acid sequence was made of 332 residues containing two putative betaalphabeta-fold regions, typical of NAD- and FAD-binding proteins. The primary structure of SsNOX38 did not show any homology with the N-terminal amino acid sequence of a NADH oxidase previously isolated from S. solfataricus (SsNOX35) (Masullo, M., Raimo, G., Dello Russo, A., Bocchini, V. and Bannister, J. V. (1996) Biotechnol. Appl. Biochem. 23, 47-54). Conversely, it showed 40% sequence identity with a putative thioredoxin reductase from Bacillus subtilis, but it did not contain cysteines, which are essential for the activity of the reductase.

A chimeric elongation factor containing the putative guanine nucleotide binding domain of archaeal EF-1 alpha and the M and C domains of eubacterial EF-Tu

A recombinant chimeric elongation factor containing the region of EF-1 alpha from Sulfolobus solfataricus harboring the site for GDP and GTP binding and GTP hydrolysis (SsG) and domains M and C of Escherichia coli EF-Tu (EcMC) was studied. SsG-EcMC did not sustain poly(Phe) synthesis in either S. solfataricus or E. coli assay system. This was probably due to the inability of the chimera to interact with aa-tRNA. The three-dimensional modeling of SsG-EcMC indicated only small structural differences compared to the Thermus aquaticus EF-Tu in the ternary complex with aa-tRNA and GppNHp, which did not account for the observed inability to interact with aa-tRNA. The addition of the nucleotide exchange factor SsEF-1 beta was not required for poly(Phe) synthesis since the chimera was already able to exchange [(3)H]GDP for GTP at very high rate even at 0 degrees C. Compared to that of SsEF-1 alpha, the affinity of the chimera for guanine nucleotides was increased and the k(cat) of the intrinsic GTPase was 2-fold higher. The heat stability of SsG-EcMC was 3 and 13 degrees C lower than that displayed by SsG and SsEF-1alpha, respectively, but 30 degrees C higher than that of EcEF-Tu. This pattern remained almost the same if the melting curves of the proteins being investigated were considered instead. The chimeric elongation factor was more thermophilic than SsG and SsEF-1 alpha up to 70 degrees C; at higher temperatures, inactivation occurred.

Protein engineering on enzymes of the peptide elongation cycle in Sulfolobus solfataricus

The present article is a review of the work done on the elongation factors EF-1 alpha, EF-2 and EF-1 beta isolated from the hyperthermophilic archaeon Sulfolobus solfataricus. The molecular, physical and biochemical properties of the intact, truncated, mutant or chimeric forms are described and compared.

Expression in Escherichia coli of the elongation factor 1beta gene and its nucleotide T160C mutant from the archaeon Sulfolobus solfataricus

The guanine nucleotide exchange factor EF-1beta gene from the thermoacidophilic archaeon Sulfolobus solfataricus (SsEF-1beta) was amplified by PCR and cloned into the pT7-7 expression vector. One of four selected clones harbored the T160C nucleotide substitution leading to the Y54H amino acid change in a hydrophobic region of SsEF-1beta, caused by a nucleotide misincorporation of the Taq DNA polymerase during PCR. The resulting plasmids were used to transform the Escherichia coli BL21(DE3)pLysE strain. Upon induction with isopropyl beta-d-thiogalactopyranoside about 1.4 mg of the recombinant SsEF-1beta (recSsEF-1beta) and Y54HSsEF-1beta were obtained from 1 liter of cell culture. recSsEF-1beta and Y54HSsEF-1beta were both able to catalyze the GDP/GTP exchange on SsEF-1alpha as observed with the wild-type SsEF-1beta. In addition, the heat inactivation profiles of recSsEF-1beta and Y54HSsEF-1beta were identical, being both half inactivated after 30 min treatment at 105 degrees C. These results suggest that Tyr 54 is not essential for the nucleotide exchange activity and is not involved in the thermostability of SsEF-1beta.

Properties of truncated forms of the elongation factor 1alpha from the archaeon Sulfolobus solfataricus

Two truncated forms of the Sulfolobus solfataricus elongation factor 1alpha (SsEF-1alpha), corresponding to the putative domains G+M, Ss(GM)EF-1alpha, and G, Ss(G)EF-1alpha, have been constructed by gene engineering, produced in Escherichia coli and purified. Neither truncated form was able to sustain poly(Phe) synthesis but they were able to bind guanine nucleotides with an affinity much higher with respect to that of the intact factor. However, the difference in the affinity for GDP and GTP became progressively reduced with the extent of the truncation. The values of kcat and Km for GTP of the intrinsic GTPase of SsEF-1alpha triggered by 3.6 M NaCl were not affected by the deletions. In contrast, both Ss(GM)EF-1alpha and Ss(G)EF-1alpha were less thermostable than the intact factor; the region of the factor most responsible for the loss of resistance against heat inactivation was the C-terminal domain. On the other hand the domain M was the regulator of the thermophilicity of SsEF-1alpha since only Ss(G)EF-1alpha showed a reduced thermophilicity. Remarkably, both Ss(GM)EF-1alpha and Ss(G)EF-1alpha were able to exchange [3H]GDP for GTP at a very high rate so that they were no more sensitive to the stimulatory effect of SsEF-1beta, which is the nucleotide exchange factor of SsEF-1alpha.

Expression in Escherichia coli of thermostable elongation factor 1 alpha from the archaeon Sulfolobus solfataricus

The elongation factor 1 alpha from the archaeon Sulfolobus solfataricus (SsEF-1 alpha) was expressed in Escherichia coli and purified. The SsEF-1 alpha gene was amplified by PCR and cloned in the Ndel site of the pT7-7 expression vector, under the control of the promoter of T7 RNA polymerase. Upon induction with isopropyl beta-D-thiogalactopyranoside, the recombinant SsEF-1 alpha (recSsEF-1 alpha) was purified from the E. coli S-100 extract by a two-step procedure. From 1 litre of cell culture, about 2 mg of purified recSsEF-1 alpha was obtained. The N-terminal sequence of the first 30 amino acid residues of recSsEF-1 alpha was identical with that translated from the nucleotide sequence of the corresponding gene, except for the initial residue, which in recSsEF-1 alpha was Ser instead of Met. The M(r) of recSsEF-1 alpha (determined by electrospray MS) was almost coincident with that of the naturally occurring SsEF-1 alpha (SsEF-1 alpha). The thermal-inactivation and thermophilicity profiles of SsEF-1 alpha and recSsEF-1 alpha were identical. Concerning the functional properties, recSsEF-1 alpha was able to support poly(Phe) synthesis in vitro, to bind GDP and GTP and to elicit an NaCl-dependent GTPase activity [Masullo, De Vendittis and Bocchini (1994) J. Biol. Chem. 269, 20376-20379] with the same efficiency as that displayed by SsEF-1 alpha.

The first nucleotide sequence of an archaeal elongation factor 1 beta gene

An archaeal elongation factor 1 beta gene has been isolated for the first time from a Sulfolobus solfataricus genomic library. The sequenced clone (869 bp) contained two open reading frames, one coding for a protein made of 91 amino acid residues (SsEF-1 beta), the other one encoding a nonidentified product (ORF 115). The amino acid sequences of segments at the N- and C-terminal of the translated SsEF-1 beta were identical to those determined for the native protein. Northern and Southern analyses showed that the SsEF-1 beta gene is represented in S. solfataricus by a unique sequence. Compared to eubacterial or eukaryal corresponding genes the SsEF-1 beta is much shorter.

Glyceraldehyde-3-phosphate dehydrogenase in the hyperthermophilic archaeon Sulfolobus solfataricus: characterization and significance in glucose metabolism

Glyceraldehyde-3-phosphate dehydrogenase in the archaeon Sulfolobus solfataricus (SsGAPD) has been purified 232 fold with an overall recovery of about 25%. The enzyme is a homomeric tetramer with an M(r) of 41 kDa/subunit. It utilizes either NAD+ or NADP+ as coenzyme but its affinity for the latter is about 50 fold higher. SsGAPD activity is maximum at 87 degrees C. In the range 45-87 degrees C the Arrhenius plot is linear and the activation energy is 55 kJ/mol. The enzyme is thermostable, with a half-life of 45 min at 87 degrees C. The primary structure of SsGAPD shows 35% identity with that of other archaeal GAPDs. Its N-domain shows sequence motifs typical of the dinucleotide binding proteins while the catalytic C-terminal region contains a cysteine residue (C140), required for catalysis, that is conserved in all the archaeal, eukaryal and bacterial GAPDs. These remarks suggest that archaeal GAPDs show a convergent molecular evolution to the eukaryal and eubacterial enzymes in the catalytic region.

Organization of a Sulfolobus solfataricus gene cluster homologous to the Escherichia coli str operon

The Sulfolobus solfataricus S12, S7 and S10 ribosomal proteins and the elongation factor 1 alpha genes are organized in a sequence analogous to that in the Escherichia coli str operon. Northern analysis showed that the S12 gene belongs to a transcript different from that corresponding to the other three genes. Compared to the Sulfolobus acidocaldarius S12 and to the Methanococcus vannielii S7 proteins, the S. solfataricus S12 and S7 proteins were 33 and 47 amino acids longer respectively. These differences were eliminated if the 5' flanking regions of the S. acidocaldarius S12 and the M. vannielii S7 genes were translated from a different start codon. Despite the structural similarities between the archaeal and the bacterial str operons the S. solfataricus ribosomal proteins S12, S7 and S10 are more similar to the eukaryotic counterparts.

The nucleotide sequence of the gene coding for the elongation factor 1 alpha in Sulfolobus solfataricus. Homology of the product with related proteins

The cloning and sequencing of the gene coding for the archaebacterial elongation factor 1 alpha (aEF-1 alpha) was performed by screening a Sulfolobus solfataricus genomic library using a probe constructed from the eptapeptide KNMITGA that is conserved in all the EF-1 alpha/EF-Tu known so far. The isolated recombinant phage contained the part of the aEF-1 alpha gene from amino acids 1 to 171. The other part (amino acids 162-435) was obtained through the amplification of the S. solfataricus DNA by PCR. The codon usage by the aEF-1 alpha gene showed a preference for triplets ending in A and/or T. This behavior was almost identical to that of the S. acidocaldarius EF-1 alpha gene but differed greatly from that of EF-1 alpha/EF-Tu genes in other archaebacteria eukaryotes and eubacteria. The translated protein is made of 435 amino acid residues and contains sequence motifs for the binding of GTP, tRNA and ribosome. Alignments of aEF-1 alpha with several EF-1 alpha/EF-Tu revealed that aEF-1 alpha is more similar to its eukaryotic than to its eubacterial counterparts.

Nucleotide sequence and molecular evolution of the gene coding for glyceraldehyde-3-phosphate dehydrogenase in the thermoacidophilic archaebacterium Sulfolobus solfataricus

A Sulfolobus solfataricus genomic library cloned in the EMBL3 phage was screened using as probes synthetic oligonucleotides designed from the known amino acid sequence of a peptide obtained from the purified glyceraldehyde-3-phosphate dehydrogenase (aGAPD) protein. The screening led to the isolation of six recombinant phages (lambda G1-lambda G6) and one of them (lambda G4) contained the entire GAPD gene. The deduced amino acid sequence accounts for a protein made of 341 amino acids and the initial methionine is encoded by a GTG triplet. Alignment of the S. solfataricus aGAPD sequence versus GAPD from archaea, eukarya, and bacteria showed that aGAPD is very similar to other archaebacterial but not to eukaryotic or eubacterial GAPD. For known archaebacterial GAPD sequences, the rate of nucleotide substitutions per site per year showed that these sequences are homologous not only at the amino acid but also at the nucleotide level. The evolutionary rates are nearly similar to those reported for other eukaryotic genes.

Human glyceraldehyde-3-phosphate dehydrogenase pseudogenes: molecular evolution and a possible mechanism for amplification

We screened two human genomic libraries and isolated 14 different clones, designated lambda G1 and EG1-EG13, homologous to human glyceraldehyde-3-phosphate dehydrogenase (GAPD) cDNA. Subcloning and sequencing these recombinant phages led us to classify them as five different pseudogenes (psi G1-psi G5). All these sequences show such features typical of processed pseudogenes as numerous mutations, insertions, and deletions. The identity of numerous mutated sites among these pseudogenes and the presence of two Alu sequences flanking both ends of psi G1 suggest that GAPD pseudogenes originated from a unique reverse transcribed mRNA followed by gene duplication. The rate of nucleotide substitutions per site per year for known GAPD functional genes is low both for the synonymous substitutions (1.87 x 10(-9] and for the nonsynonymous substitutions (0.12 x 10(-9] and indicates that the GAPD cDNA sequence is well conserved not only at the amino acid level, but also at the nucleotide level. The rate of nucleotide substitutions per site per year for GAPD pseudogenes shows a higher value (5.9 x 10(-9] and suggests that these pseudogenes do not have any functional role.

Isolation of the human glyceraldehyde-3-phosphate dehydrogenase gene by hybridization with synthetic oligonucleotides

By screening a human genomic library from human lymphocyte DNA cloned in EMBL 3 vector with synthetic oligonucleotides homologous to the human GAPD cDNA 3'-non-coding region as probes, a unique lambda recombinant clone (EMBL-G5) was isolated at the Tm value. Preliminary restriction analysis and sequence data proved that this recombinant clone is the human structural gene for the glyceraldehyde- 3-phosphate dehydrogenase.

The complete sequence of a full length cDNA for human liver glyceraldehyde-3-phosphate dehydrogenase: evidence for multiple mRNA species

A recombinant M13 clone (O42) containing a 65 b.p. cDNA fragment from human fetal liver mRNA coding for glyceraldehyde-3-phosphate dehydrogenase has been identified and it has been used to isolate from a full-length human adult liver cDNA library a recombinant clone, pG1, which has been subcloned in M13 phage and completely sequenced with the chain terminator method. Besides the coding region of 1008 b.p., the cDNA sequence includes 60 nucleotides at the 5'-end and 204 nucleotides at the 3'-end up to the polyA tail. Hybridization of pG1 to human liver total RNA shows only one band about the size of pG1 cDNA. A much stronger hybridization signal was observed using RNA derived from human hepatocarcinoma and kidney carcinoma cell lines. Sequence homology between clone 042 and the homologous region of clone pG1 is 86%. On the other hand, homology among the translated sequences and the known human muscle protein sequence ranges between 77 and 90%; these data demonstrate the existence of more than one gene coding for G3PD. Southern blot of human DNA, digested with several restriction enzymes, also indicate that several homologous sequences are present in the human genome.

Cloning of several cDNA segments coding for human liver proteins

A human cDNA library was constructed using M13 derivative vectors. The simple and rapid procedures for sequencing single-stranded DNA by the dideoxy chain termination method allowed a screening of individual clones directly by DNA sequence analysis. Some of these clones were identified as coding for: serum albumin, alpha1-antitrypsin, retinol-binding protein, prothrombin, haptoglobin, and metallothionein. Furthermore, a clone coding for aldolase B was tentatively identified on the basis of high sequence homology with rabbit muscle aldolase.

Acetohydroxy acid synthase isoenzymes of Escherichia coli K-12: a trans-acting regulatory locus of ilvHI gene expression

We isolated an Escherichia coli K-12 regulatory mutation affecting the acetohydroxy acid synthase III isoenzyme. This mutation was found to lie outside the structural genes ilvHI for this isoenzyme and was designated lrs-1. A strain carrying this mutation was found to be altered in the leucine-mediated control of ilvHI mRNA and acetohydroxy acid synthase III synthesis observed in the isogenic lrs+ strain. These alterations appeared to be a consequence of a reduced intracellular concentration of a single one of five tRNALeu isoaccepting species.

The nucleotide sequence of a small (3S) seryl-tRNA (anticodon GCU) from beef heart mitochondria

The primary structure of a 3S serine tRNA from beef heart mitochondria has been determined using our new two-dimensional "read-off" sequencing method (Tanaka, Y., Dyer, T.A. and Brownlee, G.G (1980) Nucleic Acids Res. 8, 1259-1272). When arranged in the "cloverleaf" form it shows unique features since (i) it completely lacks the dihydrouridine arm, (ii) it has an extended "T gamma" loop, but lacks the T and gamma residues, and (iii) it has only one minor base, N6(N-threonylcarbamoyl)adenosine, next to the anticodon.

Isomeric aminoacyl-tRNAs are both bound by elongation factor Tu

Recent suggestions that elongation factor Tu (EF-Tu) is specific for 2'-O-aminoacyl-tRNA, as compared with the 3'-isomer, prompted us to assay [3H]aminoacyl-tRNAs from Escherichia coli terminating in 2'- or 3'-deoxyadenosine for binding to EF-Tu to determine the possible positional specificity of the factor. Binding of modified aminaocyl-tRNAs to EF-Tu-GTP was measured both as a function of the ability of EF-Tu-GTP to diminish the rate of chemical deacylation of [3H]aminoacyl-tRNAs and by gel filtration of the individual ternary complexes. Fifteen different tRNA isoacceptors were tested by the deacylation procedure, including three (tRNAAsp, tRNACys, and tRNATyr) for which isomeric modified aminoacyl-tRNAs were available. All of the modified aminoacyl-tRNAs were protected fromdeacylation, although generally to a lesser extent than the corresponding unmodified species. Six modified tRNA isoacceptors (including tRNATrp and tRNATyr, for which both modified aminoacyl-tRNAs were accessible by enzymatic aminoacylation) were used in gel filtration experiments to permit direct measurement of the individual aminoacyl-tRNA-EF-Tu-GTP complexes. These experiments were also done in the presence of equimolar amounts of the corresponding unmodified [14C]aminoacyl-tRNAs, and the relative affinities for a limiting amount of EF-Tu-GTP were measured. The results were completely consistent with those obtained by the deacylation procedure and indicated that EF-Tu can bind to both positional isomers of aminoacyl-tRNA with no obvious preference for either.