Bacterial Fitness Optimization Through Noise Robustness and Oscillatory Regulation

Abstract

Bacterial cells thrive in stochastic environments despite the inherent constraints ofmicroscopic scale and biochemical noise. This thesis investigates how bacteria optimize fitness through different regulatory strategies against various sources of stochasticity. We first establish a regulatory principle, the Robustness-Load Trade-Off model, which provides three key predictions: (1) the majority of essential proteins are expressed in vast excess, (2) essential genes are transcribed above one message per cell cycle, and (3) protein expression is tuned by load balancing transcription and translation. Experimental validation in Acinetobacter baylyi confirms these predictions: single-cell and genome-wide analyses demonstrate that 70% of essential proteins are expressed in vast overabundance and that overabundance is higher for low-expression proteins. Finally, we explore an unforeseen phenomenon of homeostatic regulation: temporal oscillations of metabolites. We observe metabolic oscillations in deoxynucleotide triphosphate (dNTP) synthesis via replication fork velocity measurements. By developing a periodically driven temperature-based synchronization technique, we hope to show that oscillations emerge across bacterial species. Collectively, this work uncovers how bacteria are robust to intrinsic noise and dynamics fluctuations of complex biochemical networks.