A Unifying Mechanistic Model of Bacterial Transcription with Three Interconnected Pause States and Non-Diffusive Backtrack Recovery

Abstract

SUMMARY Pausing by RNA polymerase (RNAp) facilitates the recruitment of regulatory factors, RNA folding, and other related processes. While backtracking and intra-structural isomerization have been proposed to trigger pausing, the mechanisms of pause formation and recovery remain debated. Using high-throughput magnetic tweezers with single-molecule sensitivity, we have examined the full temporal spectrum of Escherichia coli RNAp transcription dynamics. Together with probabilistic dwell-time analysis and modelling, this has identified three distinct states that compete with elongation: a short-lived elemental pause and two long-lived backtracked pause states. We demonstrate that recovery from backtracking is not governed by diffusional Brownian motion, but rather by intrinsic RNA cleavage, whereby RNAp conformational changes hinder cleavage in long-lived pause states. We further show that state switching underlies stochastic alterations in the frequency of short pauses. Based on these findings, we propose a consensus model of intrinsic pausing that unifies all key findings while resolving earlier contradictions.