Leavitt, J. (BMEB) – Evolutionary Dynamics, Functional Adaptations in Stress Response, and Direct Detection of tRNA modifications in Archaea

Transfer RNA (tRNA) modifications are essential for structural integrity, decoding fidelity, and stress adaptation, yet their dynamics across phylogenetically distinct archaeal species and their functional roles during stress remain incompletely understood. This dissertation aims to address some of these gaps through a multi-scale investigation that spans the evolutionary dynamics, stress-responsive functions, and direct detection of archaeal tRNA modifications. The first chapter maps this unexplored landscape by applying Ordered Two-Template Relay sequencing (OTTR-seq) across nine archaeal species from diverse and extreme environments. This comparative analysis revealed previously unrecognized, coordinated modification patterns, including mutually exclusive methylation patterns in the acceptor stem of hyperthermophiles. Additional comparisons revealed co-evolution of tRNA modifying enzymes, demonstrating how their domain architectures and substrate specificities have diverged to shape lineage-specific adaptations. Building on these evolutionary observations, the second chapter investigates the functional role of tRNA modifications in the stress responses of the model halophile Haloferax volcanii. This work reveals how tRNA modification dynamics might balance structural stability against flexibility to manage stress, focusing on N2,N2-dimethylguanosine (m22G) at position 26. By integrating tRNA sequencing, proteomics, and codon usage data within a linear mixed-effects model, this work quantifies how the m22G modification status fine-tunes the translation of specific, codon-biased genes, establishing it as a modulator of the adaptive stress response. Addressing limitations of reverse-transcription (RT)-based sequencing methods for detecting modifications, the final chapter explores the use of direct RNA nanopore sequencing. The focus is on archaeosine (G+), a modification unique to Archaea that is inaccessible to RT-based sequencing methods. The resulting custom model accurately detects archaeosine in its native species. However, cross-species comparisons reveal significant challenges with species-specific overfitting, providing insights into development of universally applicable modification callers.

Event Host: Jesse Leavitt, Ph.D Candidate, Biomolecular Engineering & Bioinformatics