tRNA(Pro) anticodon recognition by Thermus thermophilus prolyl-tRNA synthetase.

BACKGROUND: Most aminoacyl-tRNA synthetases (aaRSs) specifically recognize all or part of the anticodon triplet of nucleotides of their cognate tRNAs. Class IIa and class IIb aaRSs possess structurally distinct tRNA anticodon-binding domains. The class IIb enzymes (LysRS, AspRS and AsnRS) have an N-terminal beta-barrel domain (OB-fold); the interactions of this domain with the anticodon stem-loop are structurally well characterised for AspRS and LysRS. Four out of five class IIa enzymes (ProRS, ThrRS, HisRS and GlyRS, but not SerRS) have a C-terminal anticodon-binding domain with an alpha/beta fold, not yet found in any other protein. The mode of RNA binding by this domain is hitherto unknown as is the rationale, if any, behind classification of anticodon-binding domains for different aaRSs. RESULTS: The crystal structure of Thermus thermophilus prolyl-tRNA synthetase (ProRSTT) in complex with tRNA(Pro) has been determined at 3.5 A resolution by molecular replacement using the native enzyme structure. One tRNA molecule, of which only the lower two-thirds is well ordered, is found bound to the synthetase dimer. The C-terminal anticodon-binding domain binds to the anticodon stem-loop from the major groove side. Binding to tRNA by ProRSTT is reminiscent of the interaction of class IIb enzymes with cognate tRNAs, but only three of the anticodon-loop bases become splayed out (bases 35-37) rather than five (bases 33-37) in the case of class IIb enzymes. The two anticodon bases conserved in all tRNA(Pro), G35 and G36, are specifically recognised by ProRSTT. CONCLUSIONS: For the synthetases possessing the class IIa anticodon-binding domain (ProRS, ThrRS and GlyRS, with the exception of HisRS), the two anticodon bases 35 and 36 are sufficient to uniquely identify the cognate tRNA (GG for proline, GU for threonine, CC for glycine), because these amino acids occupy full codon groups. The structure of ProRSTT in complex with its cognate tRNA shows that these two bases specifically interact with the enzyme, whereas base 34, which can be any base, is stacked under base 33 and makes no interactions with the synthetase. This is in agreement with biochemical experiments which identify bases 35 and 36 as major tRNA identity elements. In contrast, class IIb synthetases (AspRS, AsnRS and LysRS) have a distinct anticodon-binding domain that specifically recognises all three anticodon bases. This again correlates with the requirements of the genetic code for cognate tRNA identification, as the class IIb amino acids occupy half codon groups.