The Evolution of Trait Loss and Trait Absence in Bacterial Translation: Empirical Approaches and Conceptual Framework

Translation is a key step in the central dogma of molecular biology, showing remarkable mechanistic conservation across the tree of life, while also imposing vast diversity. To address the evolutionary mechanism of how these universality and diversity are formulated, researchers narrow their interest to the trait dynamics of specific features in translation. In this dissertation, I investigate these dynamics in two distinct systems of bacterial translation: the repeated loss of prfB autoregulation and the near-global absence of tRNAs carrying unmodified adenine at the anticodon wobble position.

Chapter 2: Repeated loss of prfB autoregulation

The prfB gene encodes RF2, a release factor that recognizes UGA and UAA stop codons and, in many bacteria, contains an internal UGA stop codon. Full-length RF2 translation therefore requires +1 programmed ribosomal frameshifting (PRF) at this internal stop codon, which also enables autoregulation of RF2 expression. This autoregulation has repeatedly been lost through bacterial phylogeny. Using P. fluorescens SBW25, we investigated that this loss can be facilitated via compensatory evolution that restores the defective mutation at the PRF site.

Chapter 3: near-global absence of tRNA-ANN

I investigated the near-global absence of tRNA genes carrying unmodified adenine at the wobble position, tRNA-ANN. Although such tRNAs are theoretically possible, they are rarely found in nature. By engineering 36 Escherichia coli strains expressing different tRNA-ANN variants, we experimentally assessed their fitness effects and functionality. We found that many tRNA-ANN variants can substitute for native tRNAs, but are often toxic, particularly in 2D split-codon boxes, supporting the canonical view that purifying selection contributes to the near-universal absence of tRNA-ANN.

Chapter 4: Discussion

I propose two idealized forms of trait dynamics, Utility-dependent dynamics and Path-dependent dynamics, to integrate different evolutionary narratives into a testable research framework. Utility-dependent dynamics focuses on direct fitness effects as drivers of trait transitions, whereas Path-dependent dynamics emphasizes evolutionary constraints and historical background.