Threonyl-tRNA synthetase (ThrRS) participates in protein synthesis quality control by selectively editing the misacylated species Ser-tRNA ^Thr . In bacteria and eukaryotes the editing function of ThrRS resides in a highly conserved N-terminal domain distant from the active site. Most archaeal ThrRS proteins are devoid of this editing domain, suggesting evolutionary divergence of quality-control mechanisms. Here we show that archaeal editing of Ser-tRNA ^Thr is catalyzed by a domain unrelated to, and absent from, bacterial and eukaryotic ThrRSs. Despite the lack of sequence homology, the archaeal and bacterial editing domains are both reliant on a pair of essential histidine residues suggestive of a common catalytic mechanism. Whereas the archaeal editing module is most commonly part of full-length ThrRS, several crenarchaeal species contain individual genes encoding the catalytic (ThrRS-cat) and editing domains (ThrRS-ed). Sulfolobus solfataricus ThrRS-cat was shown to synthesize both Thr-tRNA ^Thr and Ser-tRNA ^Thr and to lack editing activity against Ser-tRNA ^Thr . In contrast, ThrRS-ed lacks aminoacylation activity but can act as an autonomous protein in trans to hydrolyze specifically Ser-tRNA ^Thr , or it can be fused to ThrRS-cat to provide the same function in cis. Deletion analyses indicate that ThrRS-ed is dispensable for growth of S. solfataricus under standard conditions but is required for normal growth in media with elevated serine levels. The growth phenotype of the ThrRS-ed deletion strain suggests that retention of the discontinuous ThrRS quaternary structure relates to specific physiological requirements still evident in certain Archaea.

Aminoacyl-tRNA synthetases (AARSs) are a superfamily of enzymes responsible for the faithful translation of the genetic code and have lately become a prominent target for synthetic biologists. Our large-scale analysis of >2500 prokaryotic genomes reveals the complex evolutionary history of these enzymes and their paralogs, in which horizontal gene transfer played an important role. These results show that a widespread belief in the evolutionary stability of this superfamily is misconceived. Although AlaRS, GlyRS, LeuRS, IleRS, ValRS are the most stable members of the family, GluRS, LysRS and CysRS often have paralogs, whereas AsnRS, GlnRS, PylRS and SepRS are often absent from many genomes. In the course of this analysis, highly conserved protein motifs and domains within each of the AARS loci were identified and used to build a web-based computational tool for the genome-wide detection of AARS coding sequences. This is based on hidden Markov models (HMMs) and is available together with a cognate database that may be used for specific analyses. The bioinformatics tools that we have developed may also help to identify new antibiotic agents and targets using these essential enzymes. These tools also may help to identify organisms with alternative pathways that are involved in maintaining the fidelity of the genetic code.

We provide the first high-throughput analysis of the properties and functional role of Low Complexity Regions (LCRs) in more than 1500 prokaryotic and phage proteomes. We observe that, contrary to a widespread belief based on older and sparse data, LCRs actually have a significant, persistent and highly conserved presence and role in many and diverse prokaryotes. Their specific amino acid content is linked to proteins with certain molecular functions, such as the binding of RNA, DNA, metal-ions and polysaccharides. In addition, LCRs have been repeatedly identified in very ancient, and usually highly expressed proteins of the translation machinery. At last, based on the amino acid content enriched in certain categories, we have developed a neural network web server to identify LCRs and accurately predict whether they can bind nucleic acids, metal-ions or are involved in chaperone functions. An evaluation of the tool showed that it is highly accurate for eukaryotic proteins as well.

Aminoacyl-tRNA synthetases are well known for their remarkable precision in substrate selection during aminoacyl-tRNA formation. Some synthetases enhance the accuracy of this process by editing mechanisms that lead to hydrolysis of incorrectly activated and/or charged amino acids. Prolyl-tRNA synthetases (ProRSs) can be divided into two structurally divergent groups, archaeal-type and bacterial-type enzymes. A striking difference between these groups is the presence of an insertion domain (∼180 amino acids) in the bacterial-type ProRS. Because the archaeal-type ProRS enzymes have been shown to recognize cysteine, we tested selected ProRSs from all three domains of life to determine whether cysteine activation is a general property of ProRS. Here we show that cysteine is activated by recombinant ProRS enzymes from the archaea Methanocaldococcus jannaschii and Methanothermobacter thermautotrophicus, from the eukaryote Saccharomyces cerevisiae, and from the bacteria Aquifex aeolicus, Borrelia burgdorferi, Clostridium sticklandii, Cytophaga hutchinsonii, Deinococcus radiodurans, Escherichia coli, Magnetospirillum magnetotacticum, Novosphingobium aromaticivorans, Rhodopseudomonas palustris, and Thermus thermophilus.This non-cognate amino acid was efficiently acylated in vitro onto tRNA^Pro, and the misacylated Cys-tRNA^Pro was not edited by ProRS. Therefore, ProRS exhibits a natural level of mischarging that is to date unequalled among the aminoacyl-tRNA synthetases.