Anticodon recognition and discrimination by the alpha-helix cage domain of class I lysyl-tRNA synthetase

Abstract

Aminoacyl-tRNA synthetases are normally found in one of two mutually exclusive structural classes, the only known exception being lysyl-tRNA synthetase which exists in both classes I (LysRS1) and II (LysRS2). Differences in tRNA acceptor stem recognition between LysRS1 and LysRS2 do not drastically impact cellular aminoacylation levels, focusing attention on the mechanism of tRNA anticodon recognition by LysRS1. On the basis of structure-based sequence alignments, seven tRNALys anticodon variants and seven LysRS1 anticodon binding site variants were selected for analysis of the Pyrococcus horikoshii LysRS1-tRNALys docking model. LysRS1 specifically recognized the bases at positions 35 and 36, but not that at position 34. Aromatic residues form stacking interactions with U34 and U35, and aminoacylation kinetics also identified direct interactions between Arg502 and both U35 and U36. Tyr491 was also found to interact with U36, and the Y491E variant exhibited significant improvement compared to the wild type in aminoacylation of a tRNALysUUG mutant. Refinement of the LysRS1-tRNALys docking model based upon these data suggested that anticodon recognition by LysRS1 relies on considerably fewer interactions than that by LysRS2, providing a structural basis for the more significant role of the anticodon in tRNA recognition by the class II enzyme. To date, only glutamyl-tRNA synthetase (GluRS) has been found to contain an alpha-helix cage anticodon binding domain homologous to that of LysRS1, and these data now suggest that specificity for the anticodon of tRNALys could have been acquired through relatively few changes to the corresponding domain of an ancestral GluRS enzyme.

Figure 1

Comparison of LysRS1 and GluRS structures. (A) Model of the LysRS1–tRNA^Lys complex. The Rossman fold is colored orange, while the α-helix cage domain is colored red. Some of the residues involved in tRNA recognition are labeled. Adapted from ref . (B) Structure of the GluRS–tRNA^Glu complex. Crystal structure of GluRS with tRNA^Glu bound. The Rossman fold is colored green, while the α-helix cage domain, which is homologous to the one in LysRS1, is colored purple. The tRNA is colored yellow with the anticodon colored gray. Adapted from ref .

Figure 2

tRNA^Lys anticodon binding by LysRS1. (A) Detail of the amino acid residues believed to interact with the anticodon nucleotides (adapted from ref 22). (B) Sequence alignments of selected LysRS1 species highlighting conservation of residues in the anticodon binding domain. Residues believed to recognize the anticodon are underlined.

Figure 3

Double mutant cycle analysis for LysRS1 Y491E aminoacylation of tRNA^Lys UUG.

Figure 4

Model for tRNA^Lys anticodon binding by P. horikoshii LysRS1. Molecular modeling of the LysRS1–tRNA complex was accomplished with O. Adjustments were made manually on the basis of the results from mutation of tRNA^Lys and LysRS1. The conformational changes were reminiscent of those previously observed in the GluRS–tRNA^Glu complex.