Structural Bioinformatics to Understand the Origin of the Genetic Code: Structural Motif Detection in Aminoacyl-tRNA Synthetases

One of the most profound open questions in biology is how the genetic code developed. The blueprints for proteins are encoded by triplets of nucleic acids, which in turn require proteins for interpretation and replication. The mere existence of this self-referencing system is a chicken-and-egg dilemma. Aminoacyl-tRNA synthetases are key players in the transfer of genetic information and reflect the earliest episode of life. These enzymes are responsible for loading tRNA molecules with the correct amino acid. Two protein superfamilies of aminoacyl-tRNA synthetases emerged, each responsible for ten amino acids. Despite sequence and structure similarity, the delicate balance between these superfamilies is manifested in two structural motifs, which were identified in the context of this thesis: the Backbone Brackets and the Arginine Tweezers. Both motifs realize constant ligand recognition and can be found in almost all protein structures of aminoacyl-tRNA synthetases. In this thesis, I thoroughly characterized Backbone Brackets and Arginine Tweezers. The specific characteristics of these motifs require high-precision methods for their detection and analysis. However, existing algorithms do not feature an adequate computational representation of structural motifs at the atom level and the support of isofunctional residue mutations. In order to address these limitations, I designed the Fit3D algorithm for template-based and template-free detection of structural motifs. I show that proper computational representation of structural motifs is crucial and improves accuracy up to 26% for a benchmark dataset. Fit3D is a general-purpose tool for structural motif detection in high-resolution protein structure data. In conjunction with the accelerating progress in experimental methods, the demand for such tools will increase rapidly over the next years. I applied Fit3D to structures of aminoacyl-tRNA synthetases to investigate whether Backbone Brackets and Arginine Tweezers are universal building blocks for ligand recognition, and to quantify structural changes upon ligand binding. While the Arginine Tweezers motif is exclusively found in aminoacyl-tRNA synthetases and paralogs, the Backbone Brackets seem to be a general pattern to recognize functional groups of certain ligands. The results show subtle differences in side chain orientation for one structural motif and a backbone shift for the other. This suggests a structural rearrangement to be a general mechanism in some aminoacyl-tRNA synthetases. The detailed level of these analyses would not have been possible without high-precision structural motif detection with Fit3D. The results emphasize the importance of structural motifs, which consist of only a few residues, for the global function of the enzyme. Furthermore, the stunning conservation of the structural motifs located in the core domains of aminoacyl-tRNA synthetases suggests their presence in the earliest predecessors of these enzymes. Both motifs might have played a fundamental role in shaping the genetic code as we know it