Sequence of the tufA gene encoding elongation factor EF-Tu from Thermus aquaticus and overproduction of the protein in Escherichia coli

16S rRNA sequence embeddings: Meaningful numeric feature representations of nucleotide sequences that are convenient for downstream analyses

Advances in high-throughput sequencing have increased the availability of microbiome sequencing data that can be exploited to characterize microbiome community structure in situ. We explore using word and sentence embedding approaches for nucleotide sequences since they may be a suitable numerical representation for downstream machine learning applications (especially deep learning). This work involves first encoding ("embedding") each sequence into a dense, low-dimensional, numeric vector space. Here, we use Skip-Gram word2vec to embed k-mers, obtained from 16S rRNA amplicon surveys, and then leverage an existing sentence embedding technique to embed all sequences belonging to specific body sites or samples. We demonstrate that these representations are meaningful, and hence the embedding space can be exploited as a form of feature extraction for exploratory analysis. We show that sequence embeddings preserve relevant information about the sequencing data such as k-mer context, sequence taxonomy, and sample class. Specifically, the sequence embedding space resolved differences among phyla, as well as differences among genera within the same family. Distances between sequence embeddings had similar qualities to distances between alignment identities, and embedding multiple sequences can be thought of as generating a consensus sequence. In addition, embeddings are versatile features that can be used for many downstream tasks, such as taxonomic and sample classification. Using sample embeddings for body site classification resulted in negligible performance loss compared to using OTU abundance data, and clustering embeddings yielded high fidelity species clusters. Lastly, the k-mer embedding space captured distinct k-mer profiles that mapped to specific regions of the 16S rRNA gene and corresponded with particular body sites. Together, our results show that embedding sequences results in meaningful representations that can be used for exploratory analyses or for downstream machine learning applications that require numeric data. Moreover, because the embeddings are trained in an unsupervised manner, unlabeled data can be embedded and used to bolster supervised machine learning tasks.

Bacterial elongation factors EF-Tu, their mutants, chimeric forms, and domains: isolation and purification

Prokaryotic elongation factors EF-Tu form a family of homologous, three-domain molecular switches catalyzing the binding of aminoacyl-tRNAs to ribosomes during the process of mRNA translation. They are GTP-binding proteins, or GTPases. Binding of GTP or GDP regulates their conformation and thus their activity. Because of their particular structure and regulation, various activities (also outside of the translation system) and a relative abundance they represent attractive tools for studies of many basic but still not fully understood mechanisms both of the translation process, the structure-function relationships in EF-Tu molecules themselves and proteins and energy transduction mechanisms in general. The review critically summarizes procedures for the isolation and purification of native and engineered eubacterial elongation factors EF-Tu and their mutants on a large as well as small scale. Current protocols for the purification of both native and polyHis-tagged or glutathione-S-transferase (GST)-tagged EF-Tu proteins and their variants using conventional procedures and the Ni-NTA-Agarose or Glutathione Sepharose are presented.

Efficient separation of Thermus aquaticus EF-Tu functional complexes

A new method for fast separation of the main functional complexes of the elongation factor Tu from Thermus aquaticus has been developed. Binary complexes EF-Tu * GDP and EF-Tu * GDPNP as well as the ternary complex EF-Tu * GDPNP * Leu approximately tRNA were separated from each other by means of HPLC on a hydrophobic sorbent TSK-Gel Phenyl 5PW in a reverse gradient of ammonium sulfate. This technique is suitable for monitoring EF-Tu activity, characterisation of the ratio between different EF-Tu forms in cell extracts, and isolation of individual EF-Tu complexes for structural and functional investigations. In order to illustrate the potentials of the method, we used HPLC on a TSK-Gel Phenyl 5PW matrix to determine the ratio between affinities of GDP and GDPNP for EF-Tu. We found that K(a)(GDP) is about 27 times higher than K(a)(GDPNP) at 37 degrees C, the value being close to the one reported for Thermus thermophilus EF-Tu.

The crystal structure of Cys-tRNACys-EF-Tu-GDPNP reveals general and specific features in the ternary complex and in tRNA

BACKGROUND

. The translation elongation factor EF-Tu in its GTP-bound state forms a ternary complex with any aminoacylated tRNA (aa-tRNA), except initiator tRNA and selenocysteinyl-tRNA. This complex delivers aa-tRNA to the ribosomal A site during the elongation cycle of translation. The crystal structure of the yeast Phe-tRNAPhe ternary complex with Thermus aquaticus EF-Tu-GDPNP (Phe-TC) has previously been determined as one representative of this general yet highly discriminating complex formation.

RESULTS

The ternary complex of Escherichia coli Cys-tRNACys and T. aquaticus EF-Tu-GDPNP (Cys-TC) has been solved and refined at 2.6 degrees resolution. Conserved and variable features of the aa-tRNA recognition and binding by EF-Tu-GTP have been revealed by comparison with the Phe-TC structure. New tertiary interactions are observed in the tRNACys structure. A 'kissing complex' is observed in the very close crystal packing arrangement.

CONCLUSIONS

The recognition of Cys-tRNACys by EF-Tu-GDPNP is restricted to the aa-tRNA motif previously identified in Phe-TC and consists of the aminoacylated 3' end, the phosphorylated 5' end and one side of the acceptor stem and T stem. The aminoacyl bond is recognized somewhat differently, yet by the same primary motif in EF-Tu, which suggests that EF-Tu adapts to subtle variations in this moiety among all aa-tRNAs. New tertiary interactions revealed by the Cys-tRNACys structure, such as a protonated C16:C59 pyrimidine pair, a G15:G48 'Levitt pair' and an s4U8:A14:A46 base triple add to the generic understanding of tRNA structure from sequence. The structure of the 'kissing complex' shows a quasicontinuous helix with a distinct shape determined by the number of base pairs.

Structure and expression of elongation factor Tu from Bacillus stearothermophilus

The tuf gene coding for elongation factor Tu (EF-Tu) of Bacillus stearothermophilus was cloned and sequenced. This gene maps in the same context as the tufA gene of Escherichia coli str operon. Northern-blot analysis and primer extension experiments revealed that the transcription of the tuf gene is driven from two promoter regions. One of these is responsible for producing a 4.9-kb transcript containing all the genes of B. stearothermophilus str operon and the other, identified adjacent to the stop codon of the fus gene and designated tufp, for producing a 1.3-kb transcript of the tuf gene only. In contrast to the situation in E. coli, the ratio between the transcription products was found to be about 10:1 in favour of the tuf gene transcript. This high transcription activity from the tufp promoter might be accounted for by the presence of an extremely A+T-rich block consisting of 29 nucleotides which immediately precedes the consensus -35 region of the promoter. A very similar tuf gene transcription strategy and the same tufp promoter organization with the identical A/T block were found in Bacillus subtilis. The tuf gene specifies a protein of 395 amino acid residues with a molecular mass of 43,290 Da, including the N-terminal methionine. A computer-generated three-dimensional homology model shows that all the structural elements essential for binding guanine nucleotides and aminoacyl-tRNA are conserved. The presence of serine at position 376 and a low affinity for kirromycin determined by zone-interference gel electrophoresis (Kd approximately 8 microM) and by polyacrylamide gel electrophoresis under non-denaturing conditions are in agreement with the reported resistance of this EF-Tu to the antibiotic. The replacement of the highly conserved Leu211 by Met was identified as a possible cause of pulvomycin resistance.

Ribosomal protein S15 from Thermus thermophilus--cloning, sequencing, overexpression of the gene and RNA-binding properties of the protein

A 6-kb DNA fragment from an extreme thermophile, Thermus thermophilus, carrying the genes for cytochrome oxidase ba3 subunit I (cbaA) and the ribosomal protein S15 (rpsO) was cloned into Escherichia coli. The gene rpsO was sequenced. The deduced amino acid sequence exhibits 59% identity to the corresponding protein from E. coli. Expression of rpsO in E. coli requires the use of a fully repressed inducible promoter because S15 from T. thermophilus is toxic for E. coli cells. When purified without denaturation from either overproducing E. coli strain or from T. thermophilus ribosomes, the S15 protein is stable and binds a cloned T. thermophilus 16S rRNA fragment (nucleotides 559-753), with low identical dissociation constants (2.5 nM), thus demonstrating that the thermophilic protein folds correctly in a mesophilic bacterium. The rRNA fragment bound corresponds in position and structure to the 16S rRNA fragment of E. coli. A similar high affinity was also found for the binding of S15 from T. thermophilus or E. coli to the corresponding E. coli 16S rRNA fragment, whereas a slightly lower affinity was observed in binding experiments between E. coli S15 and T. thermophilus 16S rRNA fragment. These results suggest that S15 from T. thermophilus recognizes similar determinants in both rRNA fragments. Competition experiments support this conclusion.

Structural constraints on the complex of elongation factor Tu with magnesium guanosine diphosphate from rotational-echo double-resonance NMR

Rotational-echo, double-resonance (REDOR) NMR measurements of 31P-15N dipolar couplings have been made on a complex of Mg guanosine diphosphate (MgGDP) with uniformly 15N-labeled elongation factor Tu. The complex was embedded in a lyophilized buffer glass. The observed 15N REDOR dephasing by 31P was accounted for quantitatively by distances from 15N of Gly23 and Lys24 to P alpha and P beta of MgGDP as determined by X-ray crystallography of MgGDP complex formed using an elongation factor Tu that is missing a 15 residue loop in the vicinity of the binding site.

F-actin sequesters elongation factor 1alpha from interaction with aminoacyl-tRNA in a pH-dependent reaction

The machinery of eukaryotic protein synthesis is found in association with the actin cytoskeleton. A major component of this translational apparatus, which is involved in the shuttling of aa-tRNA, is the actin-binding protein elongation factor 1alpha (EF-1alpha). To investigate the consequences for translation of the interaction of EF-1alpha with F-actin, we have studied the effect of F-actin on the ability of EF-1alpha to bind to aa-tRNA. We demonstrate that binding of EF-1alpha:GTP to aa-tRNA is not pH sensitive with a constant binding affinity of approximately 0.2 microM over the physiological range of pH. However, the sharp pH dependence of binding of EF-1alpha to F-actin is sufficient to shift the binding of EF-1alpha from F-actin to aa-tRNA as pH increases. The ability of EF-1alpha to bind either F-actin or aa-tRNA in competition binding experiments is also consistent with the observation that EF-1alpha's binding to F-actin and aa-tRNA is mutually exclusive. Two pH-sensitive actin-binding sequences in EF-1alpha are identified and are predicted to overlap with the aa-tRNA-binding sites. Our results suggest that pH-regulated recruitment and release of EF-1alpha from actin filaments in vivo will supply a high local concentration of EF-1alpha to facilitate polypeptide elongation by the F-actin-associated translational apparatus.

Crystal structure of the ternary complex of Phe-tRNAPhe, EF-Tu, and a GTP analog

The structure of the ternary complex consisting of yeast phenylalanyl-transfer RNA (Phe-tRNAPhe), Thermus aquaticus elongation factor Tu (EF-Tu), and the guanosine triphosphate (GTP) analog GDPNP was determined by x-ray crystallography at 2.7 angstrom resolution. The ternary complex participates in placing the amino acids in their correct order when messenger RNA is translated into a protein sequence on the ribosome. The EF-Tu-GDPNP component binds to one side of the acceptor helix of Phe-tRNAPhe involving all three domains of EF-Tu. Binding sites for the phenylalanylated CCA end and the phosphorylated 5' end are located at domain interfaces, whereas the T stem interacts with the surface of the beta-barrel domain 3. The binding involves many conserved residues in EF-Tu. The overall shape of the ternary complex is similar to that of the translocation factor, EF-G-GDP, and this suggests a novel mechanism involving "molecular mimicry" in the translational apparatus.

Purification and crystallization of the ternary complex of elongation factor Tu:GTP and Phe-tRNA(Phe)

Elongation factor Tu (EF-Tu) is the most abundant protein in prokaryotic cells. Its general function in protein biosynthesis is well established. It is a member of the large family of G-proteins, all of which bind guanosine phosphates (GDP or GTP) as cofactors. In its active GTP bound state EF-Tu binds aminoacylated tRNA (aa-tRNA) forming the ternary complex EF-Tu:GTP:aa-tRNA. The ternary complex interacts with the ribosome where the anticodon on tRNA recognises a codon on mRNA, GTPase activity is induced and inactive EF-Tu:GDP is released. Here we report the successful crystallization of a ternary complex of Thermus aquaticus EF-Tu:GDPNP and yeast Phe-tRNA(Phe) after its purification by HPLC.

Phylogenetic relationships of Bacteria based on comparative sequence analysis of elongation factor Tu and ATP-synthase beta-subunit genes

Comparative sequence analyses were performed on 14 genes encoding bacterial elongation factors EF-Tu and 7 genes encoding the beta-subunit of bacterial F1F0 type ATP-synthases. The corresponding predicted amino acid sequences were compared with published primary structures of homologous molecules. Phylogenetic trees were reconstructed from both data sets of aligned protein sequences and from an equivalent selection of 16S rRNA sequences by applying distance matrix and maximum parsimony methods. The EF-Tu data were in very good agreement with the rRNA data, although the resolution within the EF-Tu tree was reduced at certain phylogenetic levels. The resolution power of the ATPase beta-subunit sequence data were more reduced than those of the EF-Tu data. In comparison with the 16S rRNA tree there are minor differences in the order of adjacent branchings within the ATPase beta-subunit tree.

The crystal structure of elongation factor EF-Tu from Thermus aquaticus in the GTP conformation

BACKGROUND

Elongation factor Tu (EF-Tu) is a GTP-binding protein that is crucial for protein biosynthesis. In the GTP form of the molecule, EF-Tu binds tightly to aminoacyl-tRNA, forming a ternary complex that interacts with the ribosomal acceptor site. During this interaction, GTP is hydrolyzed, and EF-Tu.GDP is ejected.

RESULTS

The crystal structure of EF-Tu from Thermus aquaticus, complexed to the GTP analogue GDPNP, has been determined at 2.5 A resolution and compared to the structure of Escherichia coli EF-Tu.GDP. During the transition from the GDP (inactive) to the GTP (active) form, domain 1, containing the GTP-binding site, undergoes internal conformational changes similar to those observed in ras-p21. In addition, a dramatic rearrangement of domains is observed, corresponding to a rotation of 90.8 degrees of domain 1 relative to domains 2 and 3. Residues that are affected in the binding of aminoacyl-tRNA are found in or near the cleft formed by the domain interface.

CONCLUSION

GTP binding by EF-Tu leads to dramatic conformational changes which expose the tRNA binding site. It appears that tRNA binding to EF-Tu induces a further conformational change, which may affect the GTPase activity.

Identification of the gene encoding transcription factor NusG of Thermus thermophilus

The nusG gene of Thermus thermophilus HB8 was cloned and sequenced. It is located 388 bp downstream from tufB, which is followed by the genes for ribosomal proteins L11 and L1. No equivalent to secE preceding nusG, as in Escherichia coli, could be detected. The nusG gene product was overproduced in E. coli. A rabbit antiserum raised against the purified recombinant NusG reacted exclusively with one protein band of T. thermophilus crude extracts in Western blot (immunoblot) analyses, and no cross-reaction of the antiserum with E. coli NusG was observed. Recombinant NusG and the reacting T. thermophilus wild-type protein had identical sizes on sodium dodecyl sulfate-polyacrylamide gels. T. thermophilus and E. coli NusG have 45% identical and 22.5% similar amino acids, and similarities between the two proteins are most pronounced in carboxy-terminal regions. The T. thermophilus nusG gene could not rescue a nusG-deficient E. coli mutant strain.