Interaction of mammalian mitochondrial elongation factor EF-Tu with guanine nucleotides.

Elongation factor Tu (EF-Tu) promotes the binding of aminoacyl-tRNA (aa-tRNA) to the acceptor site of the ribosome. During the elongation cycle, EF-Tu interacts with guanine nucleotides, aa-tRNA and its nucleotide exchange factor (EF-Ts). Quantitative determination of the equilibrium dissociation constants that govern the interactions of mammalian mitochondrial EF-Tu (EF-Tu(mt)) with guanine nucleotides was the focus of the work reported here. Equilibrium dialysis with [3H]GDP was used to measure the equilibrium dissociation constant of the EF-Tu(mt) x GDP complex (K(GDP) = 1.0 +/- 0.1 microM). Competition of GTP with a fluorescent derivative of GDP (mantGDP) for binding to EF-Tu(mt) was used to measure the dissociation constant of the EF-Tu(mt) x GTP complex (K(GTP) = 18 +/- 9 microM). The analysis of these data required information on the dissociation constant of the EF-Tu(mt) x mantGDP complex (K(mGDP) = 2.0 +/- 0.5 microM), which was measured by equilibrium dialysis. Both K(GDP) and K(GTP) for EF-Tu(mt) are quite different (about two orders of magnitude higher) than the dissociation constants of the corresponding complexes formed by Escherichia coli EF-Tu. The forward and reverse rate constants for the association and dissociation of the EF-Tu(mt) x GDP complex were determined using the change in the fluorescence of mantGDP upon interaction with EF-Tu(mt). These values are in agreement with a simple equilibrium binding interaction between EF-Tu(mt) and GDP. The results obtained are discussed in terms of the recently described crystal structure of the EF-Tu(mt) x GDP complex.

Interaction of mitochondrial elongation factor Tu with aminoacyl-tRNA and elongation factor Ts.

Elongation factor (EF) Tu promotes the binding of aminoacyl-tRNA (aa-tRNA) to the acceptor site of the ribosome. This process requires the formation of a ternary complex (EF-Tu.GTP.aa-tRNA). EF-Tu is released from the ribosome as an EF-Tu.GDP complex. Exchange of GDP for GTP is carried out through the formation of a complex with EF-Ts (EF-Tu.Ts). Mammalian mitochondrial EF-Tu (EF-Tu(mt)) differs from the corresponding prokaryotic factors in having a much lower affinity for guanine nucleotides. To further understand the EF-Tu(mt) subcycle, the dissociation constants for the release of aa-tRNA from the ternary complex (K(tRNA)) and for the dissociation of the EF-Tu.Ts(mt) complex (K(Ts)) were investigated. The equilibrium dissociation constant for the ternary complex was 18 +/- 4 nm, which is close to that observed in the prokaryotic system. The kinetic dissociation rate constant for the ternary complex was 7.3 x 10(-)(4) s(-)(1), which is essentially equivalent to that observed for the ternary complex in Escherichia coli. The binding of EF-Tu(mt) to EF-Ts(mt) is mutually exclusive with the formation of the ternary complex. K(Ts) was determined by quantifying the effects of increasing concentrations of EF-Ts(mt) on the amount of ternary complex formed with EF-Tu(mt). The value obtained for K(Ts) (5.5 +/- 1.3 nm) is comparable to the value of K(tRNA).

Expression and characterization of the human mitochondrial leucyl-tRNA synthetase.

A cDNA clone encoding the human mitochondrial leucyl-tRNA synthetase (mtLeuRS) has been identified from the EST databases. Analysis of the protein encoded by this cDNA indicates that the protein is 903 amino acids in length and contains a mitochondrial signal sequence that is predicted to encompass the first 21 amino acids. Sequence analysis shows that this protein contains the characteristic motifs of class I aminoacyl-tRNA synthetases and regions of high homology to other mitochondrial and bacterial LeuRS proteins. The mature form of this protein has been cloned and expressed in Escherichia coli. Gel filtration indicates that human mtLeuRS is active in a monomeric state, with an apparent molecular mass of 101 kDa. The human mtLeuRS is capable of aminoacylating E. coli tRNA(Leu). Its activity is inhibited at high levels of either monovalent or divalent cations. K(M) and k(cat) values for ATP:PP(i) exchange and for the aminoacylation reaction have been determined.

Effects of domain exchanges between Escherichia coli and mammalian mitochondrial EF-Tu on interactions with guanine nucleotides, aminoacyl-tRNA and ribosomes.

Escherichia coli elongation factor (EF-Tu) and the corresponding mammalian mitochondrial factor, EF-Tumt, show distinct differences in their affinities for guanine nucleotides and in their interactions with elongation factor Ts (EF-Ts) and mitochondrial tRNAs. To investigate the roles of the three domains of EF-Tu in these differences, six chimeric proteins were prepared in which the three domains were systematically switched. E. coli EF-Tu binds GDP much more tightly than EF-Tumt. This difference does not reside in domain I alone but is regulated by interactions with domains II and III. All the chimeric proteins formed ternary complexes with GTP and aminoacyl-tRNA although some had an increased or decreased activity in this assay. The activity of E. coli EF-Tu but not of EF-Tumt is stimulated by E. coli EF-Ts. The presence of any one of the domains of EF-Tumt in the prokaryotic factor reduced its interaction with E. coli EF-Ts 2-3-fold. In contrast, the presence of any of the three domains of E. coli EF-Tu in EF-Tumt allowed the mitochondrial factor to interact with bacterial EF-Ts. This observation indicates that even domain II which is not in contact with EF-Ts plays an important role in the nucleotide exchange reaction. EF-Tsmt interacts with all of the chimeras produced. However, with the exception of domain III exchanges, it inhibits the activities of the chimeras indicating that it could not be productively released to allow formation of the ternary complex. The unique ability of EF-Tumt to promote binding of mitochondrial Phe-tRNAPhe to the A-site of the ribosome resides in domains I and II. These studies indicate that the interactions of EF-Tu with its ligands is a complex process involving cross-talk between all three domains.

Expression and characterization of a human mitochondrial phenylalanyl-tRNA synthetase.

Human mitochondrial phenylalanyl-tRNA synthetase (mtPheRS) has been identified from the human EST database. Using consensus sequences derived from conserved regions of the alpha and beta-subunits from bacterial PheRS, two partially sequenced cDNA clones were identified. Unexpectedly, sequence analysis indicated that one of these clones was a truncated form of the other. Detailed analysis indicates that unlike the (alphabeta)2 structure of the prokaryotic and eukaryotic cytoplasmic forms of PheRS, the human mtPheRS consists of a single polypeptide chain. This protein has been cloned and expressed in Escherichia coli. Gel filtration and analytical velocity sedimentation centrifugation indicate that the human mtPheRS is active in a monomeric form. The N-terminal 314 amino acid residues appear to be analogous to the alpha-subunit of the prokaryotic PheRS, while the C-terminal 100 amino acid residues correspond to a region of the beta-subunit known to interact with the anticodon of tRNAPhe. Comparisons with the sequences of PheRS from yeast and Drosophila mitochondria indicate they are 42 % and 51 % identical with the human mtPheRS, respectively. Sequence analysis confirms the presence of motifs characteristic of class II aminoacyl-tRNA synthetases. KM and kcat values for ATP:PPi exchange and for the aminoacylation reaction carried out by human mtPheRS have been determined. Evolutionary origins of this small monomeric human mtPheRS are unknown, however, implications are that this enzyme is a result of the simplification of the more complex (alphabeta)2 bacterial PheRS in which specific functional regions were retained.

Regions of 16S ribosomal RNA proximal to transfer RNA bound at the P-site of Escherichia coli ribosomes.

Unmodified uridines have been randomly replaced by 4-thiouridines in transfer RNAPhe (tRNAPhe) transcribed in a T7 RNA polymerase system. These 4-thiouridines serve as conjugation sites for attachment of the cleavage reagent 5-iodoacetamido-1,10-o-phenanthroline (IoP). In a reducing environment, when complexed with Cu2+, 1,10-o-phenanthroline causes cleavage of nearby nucleic acids. We show here that tRNA-phenanthroline (tRNA-oP) conjugates, when bound at the P-site of 70S ribosomes and 30S ribosomal subunits, caused cleavage of ribosomal RNA (rRNA) mainly in domains I and II of 16S rRNA. Some positions were cleaved only when tRNA-oP was bound to 70S ribosomes or to 30S ribosomal subunits. In domain I, most cleavage sites occurred in or near the 530 pseudoknot region. In domain II, most nucleotides cleaved were near the 690 region and the 790 region. The only positions cleaved in domain III were near the 1050 region. There were no discernible nucleotides cleaved near the 1400 (decoding) region. Our results corroborated results of others, which have shown these sites to be protected from chemical modification by tRNA binding or to be cross-linked to P-site-bound tRNA. Use of cleavage reagents tethered to tRNA provides evidence for additional regions of rRNA that may be proximal to bound tRNA.

Cleavage of 16S rRNA within the ribosome by mRNA modified in the A-site codon with phenanthroline-Cu(II).

Cleavage of 16S rRNA was obtained through mRNA modified at position +5 with the chemical cleavage agent 1,10-o-phenanthroline. In the presence of Cu2+, and after addition of reducing agent to the modified mRNA-70S complex, cleavage of proximal nucleotides within the 16S rRNA occurred. Primer extension analysis of 16S rRNA fragments revealed that nucleotides 528-532, 1196, and 1396-1397 were cleaved. Nucleotides 1053-1055 were also cleaved but did not show the same level of specificity as the former. These results provide evidence that at some point in the translation process these regions are all within 15 A of position +5, the A-site codon, on the mRNA.

Mechanistic studies of the translational elongation cycle in mammalian mitochondria.

Polyclonal antibodies have been prepared against both components of the bovine liver mitochondrial translational elongation factor Tu and Ts complex (EF-Tu x Ts(mt)). The antibodies against EF-Tu(mt) cross-react somewhat with Escherichia coli EF-Tu and wheat germ EF-1alpha. The antibodies against EF-Ts(mt) cross-react little, if at all, with E. coli EF-Ts or with EF-Ts from Euglena gracilis chloroplasts. These polyclonal antibodies have been used to investigate the relative amounts of EF-Tu(mt) and EF-Ts(mt) in bovine liver mitochondria and in cultured cells. The results of this analysis suggest that there is a 1:1 ratio of EF-Tu(mt) to EF-Ts(mt) in mammalian mitochondria. Intermediate complexes formed during the elongation cycle of protein synthesis in bovine liver mitochondria have also been investigated. The EF-Tu x Ts(mt) complex is quite resistant to dissociation by guanine nucleotides. This complex will, however, dissociate in the presence of GTP and Phe-tRNA resulting in the formation of a ternary complex comparable to that observed in prokaryotes. Kinetic data suggest that the use of the ternary complex in chain elongation increases the rate of Phe-tRNA binding to ribosomes, suggesting that it is a true intermediate in the elongation cycle. Sucrose gradient analysis indicates that the binding of EF-Tu(mt) to ribosomes can be detected in the presence of Phe-tRNA and a non-hydrolyzable analog of GTP. These results suggest that, in contrast to previous thinking, the basic features of the elongation cycle in mammalian mitochondria are quite similar to those in prokaryotes.

Identification of ribosome-ligand interactions using cleavage reagents.

To characterize ribosome-ligand interactions, we have used a cleavage reagent, 1,10-orthopenanthroline-Cu(II), tethered to various ligands, to cleave nearby regions of rRNA. The phenanthroline is tethered to the ligand using either an internal 4-thiouridine or a terminal thiophosphate. When Cu2+ and a reducing agent, such as mercaptopropionic acid, are present, cleavage of nearby nucleic acids occurs. The cleavage sites can be identified using primer-extension analysis. We have identified rRNA cleavage sites resulting from transcribed tRNAPhe having randomly placed phenanthroline-Cu(II), tRNAPhe with phenanthroline-Cu(II) at position 8, and a DNA oligomer complementary to positions 2655-2667 (alpha-sarcin region) with phenanthroline-Cu(II) placed at the 5' end. These results provide important new information on the structure of the rRNA within ribosomal subunits and on the proximity of rRNA neighborhoods to these bound ligands.

Regions of 23 S ribosomal RNA proximal to transfer RNA bound at the P and E sites.

tRNAPhe transcribed in a T7 RNA polymerase system has been modified in such a way that 4-thiouridines have randomly replaced unmodified uridines. These 4-thiouridines serve as sites for conjugation of the cleavage reagent 5-iodoacetamido-1,10-phenanthroline (IOP). 1,10-Phenantholine, when complexed with Cu2+ in a reducing environment, causes hydrolysis of nearby nucleic acids. We show here that tRNA-phenanthroline (tRNA-OP) conjugates, when bound in situ to the P- and E-sites of 70 S ribosomes, cause cleavage, mainly in domains I, III and V of 23 S ribosomal RNA (rRNA). The cleavage sites in domain V predominantly occur very close to or in the peptidyl-transferase region. The regions of domain I and III that are cleaved are apparently folded in the 50 S ribosomal subunit so as to be proximal to the peptidyl-transferase center. Most of the cleavage events occur whether the tRNA-OP conjugate is bound to ribosomes alone, or yeast tRNA is also present in the P/P hybrid state. Cleavages that occur only in the absence of yeast tRNA are limited to the 1100 region of domain II, and the 2800 region of domain VI. Cleavages that occur only in the presence of yeast occur in the 2170 region of domain V. The regions of 23 S rRNA in which tRNA-OP induced cleavage occur complement those sites shown by chemical protection and cross-liking to be in a close proximity to the tRNA. However, the cleavage approach allows a more versatile and expanded view of the near neighborhood of rRNA surrounding the tRNA. These results provide considerable information which will allow a more detailed modeling of the tertiary structure of the 50 S ribosomal subunit.