Gene Expression Tradeoffs Determine Bacterial Survival and Adaptation to Antibiotic Stress

Josiah C. Kratz, Shiladitya Banerjee

TL;DR

This paper presents a theoretical model that explains how bacteria balance growth and survival under antibiotic stress by dynamically adjusting their gene expression. It reveals that bacterial death is often due to a slow response to stress rather than exceeding a physiological limit, and that prioritizing stress response over growth can significantly increase survival.

What Problem Does This Paper Address?

Bacteria constantly face environmental challenges, requiring them to balance resource allocation between growth and survival. Antibiotic exposure is a critical stressor, but predicting bacterial survival is complex because it depends heavily on the cell's physiological state, not just the antibiotic dose. Existing models often struggle to link cellular physiology to survival outcomes, especially in dynamic environments. This paper aims to bridge this gap by developing a theoretical framework that connects extracellular conditions (antibiotic concentration, nutrient availability) to intracellular states (damage accumulation, gene expression) to predict bacterial growth, death, and survival.

Methods

The authors developed a coarse-grained, multiscale theoretical model of bacterial physiology. This model incorporates:

  • Damage Accumulation and Removal: A dynamic system tracks the accumulation of cellular damage caused by antibiotics and its removal by stress proteins. Cell death is triggered when damage exceeds a critical threshold.
  • Proteome Allocation: The model describes how cells allocate their limited resources (amino acids) to different protein sectors (metabolic, ribosomal, stress response, target proteins, housekeeping). This allocation is regulated by nutrient availability and cellular damage.
  • Gene Expression Tradeoffs: A key aspect is the dynamic regulation of gene expression, particularly the tradeoff between allocating resources to stress response genes (for survival) and ribosomal genes (for growth).
  • Stochasticity: To capture population-level behavior, a stochastic component (Langevin equation) is introduced to account for noise in damage accumulation, which can lead to cell death even at sub-MIC concentrations.
  • Phenotypic Switching Model: For pulsatile antibiotic exposure, a phenotypic switching model is used to describe the transition between susceptible and resistant states based on gene expression changes.

Datasets

Not discussed in the paper. The model is validated against existing experimental data from literature, as referenced in the paper.

Key Insights and Ideas

  • Survival is Limited by Response Speed, Not Just Dose: The most surprising finding is that bacterial death is rarely caused by antibiotic levels exceeding a maximum physiological limit. Instead, survival is primarily limited by the cell's inability to quickly alter its gene expression and transition to a less susceptible physiological state.
  • Growth vs. Survival Tradeoff is Crucial: Bacteria face a fundamental tradeoff: investing resources in stress response pathways for survival comes at the expense of growth. This non-growth-optimal allocation strategy is essential for surviving antibiotic challenges.
  • Preshift Environment Matters: The physiological state of a bacterium before antibiotic exposure significantly influences its minimum inhibitory concentration (MIC). Cells pre-exposed to low antibiotic levels can develop higher MICs due to induced stress responses.
  • Mutation-Independent Adaptation: The model demonstrates that bacteria can adapt to pulsatile antibiotic exposure through rapid, mutation-independent changes in gene expression (proteome reallocation), allowing them to survive and even proliferate.

Key Results

  • The model accurately predicts bacterial growth rate, death rate, MIC, and survival fraction across a range of dynamic environments.
  • It shows that cells overexpressing stress response genes at the expense of growth have significantly higher MICs compared to cells optimizing for growth.
  • The model reproduces experimentally observed relationships between proteome allocation and growth rate under stress.
  • It explains how pulsatile antibiotic exposure can lead to population-level adaptation through phenotypic switching.

Limitations and Future Work

  • Organism and Drug Specificity: While the framework is general, specific parameters are derived from E. coli and replication-targeting bactericides. Extending this to other organisms and drug classes will require parameterization.
  • Mutation and Persister Formation: The current model focuses on phenotypic changes (gene expression) and does not explicitly model mutations or the formation of specialized persister cells, which are other mechanisms of antibiotic resistance.
  • Complex Environments: Future work could explore more complex environments, such as combinations of antibiotics or interactions with other stressors. The authors suggest extending the model to include resistance-conferring genes or studying the effects of antibiotic combinations.

Summary updated

Abstract

To optimize their fitness, cells face the crucial task of efficiently responding to various stresses. This necessitates striking a balance between conserving resources for survival and allocating resources for growth and division. The fundamental principles governing these tradeoffs is an outstanding challenge in the physics of living systems. In this study, we introduce a coarse-grained theoretical framework for bacterial physiology that establishes a connection between the physiological state of cells and their survival outcomes in dynamic environments, particularly in the context of antibiotic exposure. Predicting bacterial survival responses to varying antibiotic doses proves challenging due to the profound influence of the physiological state on critical parameters, such as the minimum inhibitory concentration (MIC) and killing rates, even within an isogenic cell population. Our proposed theoretical model bridges the gap by linking extracellular antibiotic concentration and nutrient quality to intracellular damage accumulation and gene expression. This framework allows us to predict and explain the control of cellular growth rate, death rate, MIC, and survival fraction in a wide range of time-varying environments. Surprisingly, our model reveals that cell death is rarely due to antibiotic levels being above the maximum physiological limit, but instead survival is limited by the inability to alter gene expression sufficiently quickly to transition to a less susceptible physiological state. Moreover, bacteria tend to overexpress stress response genes at the expense of reduced growth, conferring greater protection against further antibiotic exposure. This strategy is in contrast to those employed in different nutrient environments, in which bacteria allocate resources to maximize growth rate. This highlights an important tradeoff between the cellular capacity for growth and the ability to survive antibiotic exposure.
BibTeX
@inproceedings{kratz2024gene,
  author = {Josiah C. Kratz and Shiladitya Banerjee},
  title = {{Gene Expression Tradeoffs Determine Bacterial Survival and Adaptation to Antibiotic Stress}},
  booktitle = {PRX},
  year = {2024},
  doi = {10.1103/prxlife.2.013010},
}
Figure