Preview

PDF, English - main document Download (11MB) | Terms of use

Abstract

In extant life, tRNAs facilitate protein biosynthesis by decoding genetic information with their anticodons and by providing matching amino acids bound to their 3’ end. In turn, proteins are needed to attach amino acids to their cognate tRNAs through aminoacylation. Yet, in a prebiotic scenario, likely no proteins but only RNA existed. To explain the emergence of protein biosynthesis and the origin of the genetic code, John J. Hopfield proposed in 1978 a “testable Hypothesis” about a potential self-aminoacylating tRNA precursor. These “Hopfield folds” might have exhibited a folding pattern that positioned the anticodon loop close to the 3’ end. Thereby, anticodons could have also functioned as chemical sensors to facilitate binding of only their cognate amino acids. If such Hopfield folds had indeed been the precursors of extant tRNAs, it would emphasize a stereochemical origin of the genetic code and molecular vestiges of the sequences and structures of Hopfield folds might still be present in extant tRNAs.

To test this hypothesis, I successfully developed a multi-step procedure that involved the synthesis of various aminoacyl-5’-adenylates and L-lysyl-5’-cytidylate. These were used in aminoacylation experiments with an RNA library that resembled Hopfield folds with randomized anticodon loops to identify the most reactive and selective sequences through the use of Illumina library preparation and sequencing techniques. Furthermore, the sequencing data could be validated through additional individual characterization of the four most active and selective sequences identified.

I could show that the tested Hopfield folds exhibited various degrees of self-aminoacylation activity based on the sequence of their anticodon loop. Only a small fraction of the initial pool of sequences was reactive which indicated that aminoacylation had to be specifically catalyzed. In contrast to Hopfield’s hypothesis, not only the single stranded anticodon loop but also the double stranded region close to the 3’ end and more distant parts in the 5’ terminal region were significantly involved in the realization of self-aminoacylation. However, none of the most active and selective sequences found shared notable resemblance to extant anticodons. Furthermore, the adenylate or cytidylate group had the most significant effect on the selection of reactive sequences. Amino acid side chains seemed to cause steric hindrance that correlated with their mass but was mitigated by moieties that could facilitate attractive interaction with RNA like amino and hydroxy groups. The influence of amino acid L- and D-conformation was various and sequence-dependent. An optimum at pH 5 and almost no self-aminoacylation activity at neutral pH could be observed and lower concentrations of activated amino acids increased aminoacylation specificity.