The ABC of EF-G

The recently solved crystal structures of Thermus thermophilus elongation factor G, with and without GDP, reveal a protein of five domains with surprising features which can be correlated with biochemical data to suggest probable functional roles.

Elongation factor Tu D138N, a mutant with modified substrate specificity, as a tool to study energy consumption in protein biosynthesis

Substitution Asp138-->Asn changes the substrate specificity of elongation factor (EF) Tu from GTP to XTP [Hwang & Miller (1987) J. Biol. Chem. 262, 13081-13085]. This mutated EF-Tu (EF-Tu D138N) was used to show that 2 XTP molecules are hydrolyzed for each elongation cycle [Weijland & Parmeggiani (1993) Science 259, 1311-1313]. Here we extend the study of the properties of this EF-Tu mutant and its function in the elongation process. In poly(U)-directed poly(phenylalanine) synthesis, the number of peptide chains synthesized using EF-Tu D138N.XTP was 30% higher than with EF-Tu wild type (wt).GTP. However, since in the former case the average peptide chain length was correspondingly reduced, the number of the residues incorporated turned out to be nearly the same in both systems. The K'd values of the XTP and XDP complexes of EF-Tu D138N were similar to those of the GTP and GDP complexes of EF-Tu wt. The extent of leucine misincorporation and the kirromycin effect were also comparable to those in the EF-Tu wt/GTP system. The hydrolysis of two XTP molecules, very likely as part of two EF-Tu D138N.XTP complexes, for each elongation cycle was found to be independent of (i) MgCl2 concentration, (ii) ribosome concentration, and (iii) temperature (5-40 degrees C). With rate-limiting amounts of XTP the K'm of its XTPase activity corresponded to the K'm for XTP of poly(phenylalanine) synthesis (0.3-0.6 microM).(ABSTRACT TRUNCATED AT 250 WORDS)

Three-dimensional structure of the ribosomal translocase: elongation factor G from Thermus thermophilus

The crystal structure of Thermus thermophilus elongation factor G without guanine nucleotide was determined to 2.85 A. This GTPase has five domains with overall dimensions of 50 x 60 x 118 A. The GTP binding domain has a core common to other GTPases with a unique subdomain which probably functions as an intrinsic nucleotide exchange factor. Domains I and II are homologous to elongation factor Tu and their arrangement, both with and without GDP, is more similar to elongation factor Tu in complex with a GTP analogue than with GDP. Domains III and V show structural similarities to ribosomal proteins. Domain IV protrudes from the main body of the protein and has an extraordinary topology with a left-handed cross-over connection between two parallel beta-strands.

The crystal structure of elongation factor G complexed with GDP, at 2.7 A resolution

Elongation factor G (EF-G) catalyzes the translocation step of protein synthesis in bacteria, and like the other bacterial elongation factor, EF-Tu--whose structure is already known--it is a member of the GTPase superfamily. We have determined the crystal structure of EF-G--GDP from Thermus thermophilus. It is an elongated molecule whose large, N-terminal domain resembles the G domain of EF-Tu, except for a 90 residue insert, which covers a surface that is involved in nucleotide exchange in EF-Tu and other G proteins. The tertiary structures of the second domains of EF-G and EF-Tu are nearly identical, but the relative placement of the first two domains in EF-G--GDP resembles that seen in EF-Tu--GTP, not EF-Tu--GDP. The remaining three domains of EF-G look like RNA binding domains, and have no counterparts in EF-Tu.

Synechocystis sp. PCC6803 fusB gene, located outside of the str operon, encodes a polypeptide related to protein synthesis factor EF-G

Synechocystis sp. PCC6803, a cyanobacterium, possesses an unusual gene (fusB) which encodes a protein with strong homology to protein synthesis elongation factor G (EF-G), although it is not linked to the classical str operon. The fusB gene is redundant, since a Synechocystis gene similar to str operon-encoded fusA genes of other bacteria is also present (based on PCR and hybridization results). There is no evidence for the presence of a fusB homologue in other bacteria. The Synechocystis fusB gene encodes unusual amino acids at some positions that are highly conserved in fusA genes of other prokaryotes.

Fusidic acid-resistant mutants define three regions in elongation factor G of Salmonella typhimurium

We have sequenced fusA, the gene coding for elongation factor G (EF-G), in 18 different mutants of Salmonella typhimurium selected as fusidic acid resistant (FuR). In addition, we have sequenced two previously described FuR mutants from Escherichia coli. In all cases, the resistance is due to a mutation in one of three separate regions in fusA. The three clusters of mutant sites superimpose on regions that are well conserved, suggesting that they are of a more general functional importance. To further classify the mutants, we have measured the minimal inhibitory concentration (MIC) for Fu and for two other antibiotics which interfere with translocation on the ribosome, kanamycin (Km) and spectinomycin (Sp). The levels of resistance to Fu for each of the mutants are significantly higher than in the wild type (wt), and vary by about one order of magnitude between the highest and the lowest. Most of the mutants are also more resistant to Km than the wt, although the level of resistance is low and the variation small. In contrast, about half of the mutants are more sensitive to Sp than the wt, with only one being more resistant. Only three of the twenty mutants behave like the wt with respect to the non-selected phenotypes, KmR and SpR.

Evolution of translational elongation factor (EF) sequences: reliability of global phylogenies inferred from EF-1 alpha(Tu) and EF-2(G) proteins

The EF-2 coding genes of the Archaea Pyrococcus woesei and Desulfurococcus mobilis were cloned and sequenced. Global phylogenies were inferred by alternative tree-making methods from available EF-2(G) sequence data and contrasted with phylogenies constructed from the more conserved but shorter EF-1 alpha(Tu) sequences. Both the monophyly (sensu Hennig) of Archaea and their subdivision into the kingdoms Crenarchaeota and Euryarchaeota are consistently inferred by analysis of EF-2(G) sequences, usually at a high bootstrap confidence level. In contrast, EF-1 alpha(Tu) phylogenies tend to be inconsistent with one another and show low bootstrap confidence levels. While evolutionary distance and DNA maximum parsimony analyses of EF-1 alpha(Tu) sequences do show archaeal monophyly, protein parsimony and DNA maximum-likelihood analyses of these data do not. In no case, however, do any of the tree topologies inferred from EF-1 alpha(Tu) sequence analyses receive significant bootstrap support.

Phosphorylation of elongation factor G and ribosomal protein S6 in bacteriophage T7-infected Escherichia coli

Bacteriophage T7 expresses a serine/threonine-specific protein kinase activity during infection of its host, Escherichia coli. The protein kinase (gp0.7 PK), encoded by the T7 early gene 0.7, enhances phage reproduction under sub-optimal growth conditions. It was previously shown that ribosomal protein S1 and translation initiation factors IF1, IF2, and IF3 are phosphorylated in T7-infected cells, and it was suggested that phosphorylation of these proteins may serve to stimulate translation of the phage late mRNAs. Using high-resolution two-dimensional gel electrophoresis and specific immunoprecipitation, we show that elongation factor G and ribosomal protein S6 are phosphorylated following T7 infection. The gel electrophoretic data moreover indicate that elongation factor P is phosphorylated in T7-infected cells. T7 early and late mRNAs are processed by ribonuclease III, whose activity is stimulated through phosphorylation by gp0.7 PK. Specific overexpression and phosphorylation was used to locate the RNase III polypeptide in the standard two-dimensional gel pattern, and to confirm that serine is the phosphate-accepting amino acid. The two-dimensional gels show that the in vivo expression of gp0.7 PK results in the phosphorylation of over 90 proteins, which is a significantly higher number than previous estimates. The protein kinase activities of the T7-related phages T3 and BA14 produce essentially the same pattern of phosphorylated proteins as that of T7. Finally, several experimental variables are analysed which influence the production and pattern of phosphorylated proteins in both uninfected and T7-infected cells.

Mutations in 23 S ribosomal RNA perturb transfer RNA selection and can lead to streptomycin dependence

Escherichia coli ribosomes with a G to C transversion at position 2661 in 23 S ribosomal RNA are more accurate in tRNA selection than wild-type ribosomes. This enhanced accuracy is due to improved initial selection of ternary complexes rather than proofreading of aminoacyl tRNAs. The 2661C mutation reduces the binding rate of cognate ternary complexes to the A-site. This binding rate deficiency becomes dramatic when ribosomes also harbour an S12 mutation with a streptomycin-resistant, hyperaccurate phenotype. In this case, severe loss of kinetic efficiency in EF-Tu function leads to cell death. Streptomycin restores viability by increasing the association rate of ternary complex to these doubly altered ribosomes. The binding rate of EF-G to 2661C ribosomes is also reduced while the translocation rate is unaffected.

Post-transcriptional regulation of the str operon in Escherichia coli. Ribosomal protein S7 inhibits coupled translation of S7 but not its independent translation

The str operon of Escherichia coli consists of the genes for ribosomal proteins S12 (rpsL) and S7 (rpsG) and elongation factors G (fusA) and Tu (tufA). Previous studies have shown that S7 is a translational feedback repressor and inhibits the synthesis of itself and of elongation factor G. We have now shown that induction of S7 synthesis from the S7 gene fused to the arabinose promoter on a plasmid also leads to inhibition of the synthesis of S12 from the chromosomal S12 gene, and that this regulation takes place using the same target site as that used for distal gene regulation, i.e. S7 retroregulates S12. We have then demonstrated that S7 synthesis is mostly translationally coupled with the translation of the preceding S12 gene. Using a rpsG'-'lacZ fusion gene as a reporter for S7 synthesis, we found that abolishing S12 translation by a mutational alteration of the AUG start codon of the S12 gene leads to about tenfold reduction of S7 synthesis without significantly affecting its rate of transcription. Deletion of the proximal portion of the S12 gene or a premature termination of S12 translation by an amber mutation at the 26th codon also led to a large reduction of S7 synthesis. Unexpectedly, we have discovered that overproduction of S7 in trans from a plasmid leads to repression of the rpsG'-'lacZ fusion gene when the fusion gene is preceded by the intact S12 gene, but not when the S12 gene carried the above-mentioned mutations that abolish S12 translation. Thus, a novel feature of this regulatory system is that translation of S7 achieved by independent initiation is not inhibited by S7 in vivo, whereas translation of S7 achieved by translational coupling is sensitive to S7 repression. These observations also suggest that the coupled S7 translation is probably achieved by the use of ribosomal subunits employed for translation of the upstream S12 gene.

In vivo selection of conditional-lethal mutations in the gene encoding elongation factor G of Escherichia coli

The ribosome translocation step that occurs during protein synthesis is a highly conserved, essential activity of all cells. The precise movement of one codon that occurs following peptide bond formation is regulated by elongation factor G (EF-G) in eubacteria or elongation factor 2 (EF-2) in eukaryotes. To begin to understand molecular interactions that regulate this process, a genetic selection was developed with the aim of obtaining conditional-lethal alleles of the gene (fusA) that encodes EF-G in Escherichia coli. The genetic selection depends on the observation that resistant strains arose spontaneously in the presence of sublethal concentrations of the antibiotic kanamycin. Replica plating was performed to obtain mutant isolates from this collection that were restrictive for growth at 42 degrees C. Two tightly temperature-sensitive strains were characterized in detail and shown to harbor single-site missense mutations within fusA. The fusA100 mutant encoded a glycine-to-aspartic acid change at codon 502. The fusA101 allele encoded a glutamine-to-proline alteration at position 495. Induction kinetics of beta-galactosidase activity suggested that both mutations resulted in slower elongation rates in vivo. These missense mutations were very near a small group of conserved amino acid residues (positions 483 to 493) that occur in EF-G and EF-2 but not EF-Tu. It is concluded that these sequences encode a specific domain that is essential for efficient translocase function.

[Study of segments of 23S RNA, important for interaction with elongation factors Tu and G using complementary DNA-oligonucleotides]

Several oligonucleotides complementary to different 23S RNA regions were tested in the elongation factor-dependent reactions of the ribosomes. It was found that the 1088-1100 and 1127-1140 sequence parts of the 23S RNA (binding regions for the L11 protein) are very important for EF-G function. The EF-Tu function is markedly less affected by these nucleotides. The probable role of 23S RNA function is discussed.

Purification and N-terminal sequence analysis of pea chloroplast protein synthesis factor EF-G

Chloroplast protein synthesis elongation factor G (chlEF-G) has been purified from whole-cell extracts of light-induced pea (Pisum sativum) seedlings. The first step in the purification scheme relies on the affinity of organellar EF-G for Escherichia coli ribosomes in the presence of the antibiotic, fusidic acid. A complex between organellar EF-G, E. coli ribosomes, GDP, and fusidic acid was isolated by high-speed centrifugation. The largest major protein eluted from this complex by high salt has an apparent molecular weight of 86,000 and is only a minor component of similar preparations from dark-grown seedlings. The same polypeptide copurifies with EF-G activity upon size exclusion HPLC on a Waters Protein-Pak 200SW column. The N-terminal amino acid sequence of chlEF-G has been determined by direct sequencing of gel-purified protein. Like many proteins that are processed upon import into chloroplasts, it has an N-terminal alanine residue. Part of the putative chlEF-G gene has been amplified using oligonucleotides corresponding to the N-terminal amino acid sequence of the purified protein and to highly conserved sequences within the GTP-binding domains of other elongation factors. The deduced amino acid sequence displays high sequence identity to the corresponding region of the chloroplast EF-G gene product from soybean, somewhat less similarity to bacterial EF-Gs, and only low homology to mitochondrial EF-G and to eukaryotic cytoplasmic EF-2 genes. The chlEF-G gene appears to be encoded by a two-copy gene family in pea and a single-copy gene in Arabidopsis thaliana.

Cloning and sequencing of a soybean nuclear gene coding for a chloroplast translation elongation factor EF-G

A plant nuclear gene coding for a chloroplast specific translation elongation factor EF-G (cEF-G) was cloned and sequenced for the first time. We screened two partial soybean genomic libraries with a short PCR amplified pea DNA probe constructed according to the N-terminal peptide sequence of pea chloroplast EF-G. The gene is three times split, codes for a chloroplast type transit peptide and a protein very similar to bacterial translation elongation factor EF-G. The gene is expressed as evidenced by Northern hybridisations.

Regulation of the uncoupled GTPase activity of elongation factor G (EF-G) by the conformations of the ribosomal subunits

The elongation factor G (EF-G) GTPase activity is induced by either 70S ribosomes or 50S ribosomal subunits. The GTPase activity induced by 50S ribosomal subunits is predominant at low concentrations of monovalent cations and decreases with increasing concentrations of K+ or NH4+. Double-logarithmic plots of the data reveal straight lines with different slopes for low and high concentrations of monovalent cations, respectively, intersecting at the same concentration of monovalent cations where maximal EF-G GTPase activity is measured in the presence of both ribosomal subunits. Substantially the same curves are obtained when 50S ribosomal subunits are substituted by 50S CsCl-core particles partially reconstituted by addition of purified 50S split proteins L7/L12. Intact 30S ribosomal subunits, but not 30S CsCl-core particles are able to associate with 50S ribosomal subunits and to modulate ribosome-dependent EF-G GTPase activity. Therefore, our data clearly show that the biphasic courses of the NH4+ and K+ curves of EF-G GTPase activity induced by 50S ribosomal subunits are not due to contaminations with 30S ribosomal subunits but result from different conformations of EF-G/50S ribosomal-subunit complexes at low and high concentrations of monovalent cations, respectively. CD spectra of 50S ribosomal subunits measured under different salt conditions have shown that the conformation of the 50S ribosomal subunits is strongly dependent on the concentration of monovalent cations. The conformation of 30S ribosomal subunits is, however, considerably stronger influenced by the Mg2+ than by the concentration of monovalent cations. The salt effects on the conformation of the 30S ribosomal subunits correspond to the salt effects on the association of ribosomal subunits and the modulation of EF-G GTPase activity by 30S ribosomal subunits. Since, in the presence of both ribosomal subunits, EF-G GTPase activity is maximal at the same concentration of monovalent cations where obviously a spontaneous conformation change of 50S ribosomal subunits takes place, we postulate that EF-G GTPase primarily acts on the ribosomes by changing the conformation of 50S ribosomal subunits. The resulting model is based on the assumption that EF-G GTPase activity is considerably more strongly induced by the 'substrate conformation' ('state I') than by the 'product conformation' of the 50S ribosomal subunits ('state II'). A spontaneous transformation of 'state II' to 'state I' is expected to occur in the absence of mRNA, aminoacyl-tRNA and EF-T especially under salt conditions favouring state I.

Site-directed cross-linking of mRNA analogues to 16S ribosomal RNA; a complete scan of cross-links from all positions between '+1' and '+16' on the mRNA, downstream from the decoding site

mRNA analogues containing 4-thiouridine residues at selected sites were used to extend our analysis of photo-induced cross-links between mRNA and 16S RNA to cover the entire downstream range between positions +1 and +16 on the mRNA (position +1 is the 5'-base of the P-site codon). No tRNA-dependent cross-links were observed from positions +1, +2, +3 or +5. Position +4 on the mRNA was cross-linked in a tRNA-dependent manner to 16S RNA at a site between nucleotides ca 1402-1415 (most probably to the modified residue 1402), and this was absolutely specific for the +4 position. Similarly, the previously observed cross-link to nucleotide 1052 was absolutely specific for the +6 position. The previously observed cross-links from +7 to nucleotide 1395 and from +11 to 532 were however seen to a lesser extent with certain types of mRNA sequence from neighbouring positions (+6 to +10, and +10 to +13, respectively); no tRNA-dependent cross-links to other sites on 16S RNA were found from these positions, and no cross-linking was seen from positions +14 to +16. In each case the effect of a second cognate tRNA (at the ribosomal A-site) on the level of cross-linking was studied, and the specificity of each cross-link was confirmed by translocation experiments with elongation factor G, using appropriate mRNA analogues.

Identification of the gene encoding the mitochondrial elongation factor G in mammals

Protein synthesis in cytosolic and rough endoplasmic reticulum associated ribosomes is directed by factors, many of which have been well characterized. Although these factors have been the subject of intense study, most of the corresponding factors regulating protein synthesis in the mitochondrial ribosomes remain unknown. In this report we present the cloning and initial characterization of the gene encoding the rat mitochondrial elongation factor-G (rEF-Gmt). The rat gene encoding EF-Gmt (rMef-g) maps to rat chromosome 2 and it is expressed in all tissues with highest levels in liver, thymus and brain. Its DNA sequence predicts a 752 amino acid protein exhibiting 72% homology to the yeast Saccharomyces cerevisiae mitochondrial elongation factor-G (YMEF-G), 62% and 61% homology to the Thermus thermophilus and E. coli elongation factor-G (EF-G) respectively and 52% homology to the rat elongation factor-2 (EF-2). The deduced amino acid sequence of EF-G contains characteristic motifs shared by all GTP binding proteins. Therefore, similarly to other elongation factors, the enzymatic function of EF-Gmt is predicted to depend on GTP binding and hydrolysis. EF-Gmt differs from its cytoplasmic homolog, EF-2, in that it contains an aspartic acid residue at amino acid position 621 which corresponds to the EF-2 histidine residue at position 715. Since this histidine residue, following posttranslational modification into diphthamide, appears to be the sole cellular target of diphtheria toxin and Pseudomonas aeruginosa endotoxin A, we conclude that EF-Gmt will not be inactivated by these toxins. The severe effects of these toxins on protein elongation in tissues expressing EF-Gmt suggest that EF-Gmt and EF-2 exhibit nonoverlapping functions. The cloning and characterization of the mammalian mitochondrial elongation factor G will permit us to address its role in the regulation of normal mitochondrial function and in disease states attributed to mitochondrial dysfunction.

Divergent effects of fluoroaluminates on the peptide chain elongation factors EF-Tu and EF-G as members of the GTPase superfamily

Fluoraluminates are thought to mimic the gamma-phosphate of GTP and thus, together with GDP, perturb the functioning of heterotrimeric GTP-binding G-proteins. Here we show they do inhibit the ribosome-stimulated GTPase activity of EF-G from Escherichia coli via the formation of a stable complex with EF-G-GDP and ribosomes. In contrast, no perturbed interactions were observed in a similar ribosomal complex with EF-Tu. Interestingly, in the absence of ribosomes both EF-Tu an EF-G remain totally unaffected by fluoraluminates. For members of the GTPase superfamily such differential effects have not been described before.

Replacement of the L11 binding region within E.coli 23S ribosomal RNA with its homologue from yeast: in vivo and in vitro analysis of hybrid ribosomes altered in the GTPase centre

Replacement of the protein L11 binding domain within Escherichia coli 23S ribosomal RNA (rRNA) by the equivalent region from yeast 26S rRNA appeared to have no effect on the growth rate of E.coli cells harbouring a plasmid carrying the mutated rrnB operon. The hybrid rRNA was correctly processed and assembled into ribosomes, which accumulated normally in polyribosomes. Of the total ribosomal population, < 25% contained wild-type, chromosomally encoded rRNA; the remainder were mutant. The hybrid ribosomes supported GTP hydrolysis dependent upon E.coli elongation factor G, although at a somewhat reduced rate compared with wild-type particles, and were sensitive to the antibiotic, thiostrepton, a potent inhibitor of ribosomal GTPase activity that binds to 23S rRNA within the L11 binding domain. That thiostrepton could indeed bind to the mutant ribosomes, although at a reduced level relative to that seen with wild-type ribosomes, was confirmed in a non-equilibrium assay. The rationale for the ability of the hybrid ribosomes to bind the antibiotic, given that yeast ribosomes do not, was provided when yeast rRNA was shown by equilibrium dialysis to bind thiostrepton only 10-fold less tightly than did E.coli rRNA. The extreme conservation of secondary, but not primary, structure in this region between E.coli and yeast rRNAs allows the hybrid ribosomes to function competently in protein synthesis and also preserves the interaction with thiostrepton.

Regulation of elongation factor G GTPase activity by the ribosomal state. The effects of initiation factors and differentially bound tRNA, aminoacyl-tRNA, and peptidyl-tRNA

The elongation factor G (EF-G) is responsible for the translocation of the ribosome along the mRNA chain. Under in vitro conditions, EF-G exhibits a very active uncoupled GTPase activity which is dependent on the presence of ribosomes and is modulated by mRNA-dependent binding of tRNA. In the absence of tRNA, uncoupled EF-G GTPase is inhibited by initiation factors IF1 and IF3, but not by initiation factor IF2. In the presence of N-fMet-tRNAfMet and poly(A,U,G) or in the presence of N-acetyl-Phe-tRNAPhe and poly(U), initiation factor IF2 causes an additional decrease of the uncoupled EF-G GTPase activity. This effect, however, is dependent on the presence of IF1 and IF3 and is obviously due to the mRNA- and initiation factor-dependent binding of N-fMet-tRNAfMet and N-acetyl-Phe-tRNAPhe, respectively, to the ribosomal P-site. Non-enzymatic binding of N-fMet-tRNAfMet and N-acetyl-Phe-tRNAPhe, however, causes a stimulation of uncoupled EF-G GTPase activity. The same effects are observed for Met-tRNA, Phe-tRNAPhe and uncharged tRNA. These findings are discussed in the light of the three-site model of the ribosome and the mechanism of translocation.

Comparison of the complete sequence of the str operon in Salmonella typhimurium and Escherichia coli

The nucleotide (nt) sequences of the str operon in Escherichia coli K-12 and Salmonella typhimurium LT2 were completed and compared at the nt and amino acid (aa) level. The order of conservation at the nt and aa level is rpsL greater than tufA greater than rpsG greater than f usA. A striking difference is that the rpsG-encoded ribosomal protein, S7, in E. coli K-12 is 23 aa longer than in S. typhimurium. The very low (0.18) codon adaptation index of this part of the E. coli K-12-encoding gene and the unusual stop codon (UGA) suggest that this is a relatively recent extension. A trend towards a higher G+C content in fusA (gene encoding elongation factor (EF)-G) and tufA (gene encoding EF-Tu) in S. typhimurium is noted. In fusA, nt substitutions at all three positions in a codon occur at a much higher frequency than expected from the number of nt substitutions in the gene, assuming they are random and independent events. An analysis of substitutions in this and other genes suggests that the triple substitutions in fusA, and some other genes, are the result of the sequential accumulation of individual mutations, probably driven by selection pressure for particular codons or aa.

[Escherichia coli ribosomes having peptidyl-tRNA and deacylated tRNA at the A- and P-sites, respectively, may not be competent in translocation]

Ribosomes can have two states at 0 degree C: competent and noncompetent in translocation. In both states poly(U)-programmed ribosomes bind phenylalanyl-tRNA to A and P sites and form peptide bond. Elongation factor G promotes fast translocation in competent ribosomes and makes them noncompetent ones. Initial correlation between competent and noncompetent ribosomes is 2:1. Addition of deacylated tRNA does not influence phenomenon described as well as thermal reactivation of the ribosomes before beginning of the experiments. The possibility of deacylated tRNA translocation is shown. The translocation does not occurred provided that at least one of the ribosome sites is filled with shortened tRNA analog (tRNA with truncated CCA-end or tRNA anticodon arm).

Evidence that eukaryotes and eocyte prokaryotes are immediate relatives

The phylogenetic origin of eukaryotes has been unclear because eukaryotic nuclear genes have diverged substantially from prokaryotic ones. The genes coding for elongation factor EF-1 alpha were compared among various organisms. The EF-1 alpha sequences of eukaryotes contained an 11-amino acid segment that was also found in eocytes (extremely thermophilic, sulfur-metabolizing bacteria) but that was absent in all other bacteria. The related (paralogous) genes encoding elongation factor EF-2 and initiation factor IF-2 also lacked the 11-amino acid insert. These data imply that the eocytes are the closest surviving relatives (sister taxon) of the eukaryotes.

Early evolutionary relationships among known life forms inferred from elongation factor EF-2/EF-G sequences: phylogenetic coherence and structure of the archaeal domain

Phylogenies were inferred from both the gene and the protein sequences of the translational elongation factor termed EF-2 (for Archaea and Eukarya) and EF-G (for Bacteria). All treeing methods used (distance-matrix, maximum likelihood, and parsimony), including evolutionary parsimony, support the archaeal tree and disprove the "eocyte tree" (i.e., the polyphyly and paraphyly of the Archaea). Distance-matrix trees derived from both the amino acid and the DNA sequence alignments (first and second codon positions) showed the Archaea to be a monophyletic-holophyletic grouping whose deepest bifurcation divides a Sulfolobus branch from a branch comprising Methanococcus, Halobacterium, and Thermoplasma. Bootstrapped distance-matrix treeing confirmed the monophyly-holophyly of Archaea in 100% of the samples and supported the bifurcation of Archaea into a Sulfolobus branch and a methanogen-halophile branch in 97% of the samples. Similar phylogenies were inferred by maximum likelihood and by maximum (protein and DNA) parsimony. DNA parsimony trees essentially identical to those inferred from first and second codon positions were derived from alternative DNA data sets comprising either the first or the second position of each codon. Bootstrapped DNA parsimony supported the monophyly-holophyly of Archaea in 100% of the bootstrap samples and confirmed the division of Archaea into a Sulfolobus branch and a methanogen-halophile branch in 93% of the bootstrap samples. Distance-matrix and maximum likelihood treeing under the constraint that branch lengths must be consistent with a molecular clock placed the root of the universal tree between the Bacteria and the bifurcation of Archaea and Eukarya. The results support the division of Archaea into the kingdoms Crenarchaeota (corresponding to the Sulfolobus branch and Euryarchaeota). This division was not confirmed by evolutionary parsimony, which identified Halobacterium rather than Sulfolobus as the deepest offspring within the Archaea.