TL;DR
This paper presents a mechanistic model of a bacterial promoter to understand how molecular details influence evolutionary outcomes. By accounting for gene expression dynamics, the authors show that key evolutionary properties can be predicted even when simplifying the model, suggesting a path to tackle biological complexity in evolution.
What Problem Does This Paper Address?
Predicting evolutionary trajectories is a major challenge due to the inherent complexity of biological systems. This complexity arises from several factors: the intricate mapping between genotype, phenotype, and fitness; the vastness of the mutational sequence space; the dynamic nature of mutation effects; and the complex ways organisms navigate fitness landscapes. Specifically, understanding how changes at the molecular level (like DNA sequences in a promoter) scale up to affect organismal fitness and evolutionary outcomes is difficult. This paper aims to provide a framework for tackling this complexity by using a mechanistic model of a regulated bacterial promoter.
Methods
The authors developed a mechanistic model combining two approaches:
- Statistical Thermodynamics (TD) Model: This model describes the steady-state gene expression levels based on the binding of molecules (RNA polymerase and a repressor protein) to a promoter sequence. It uses energy matrices (EMs) that quantify the effect of mutations on binding affinity.
- Mass Action Kinetics (MAK) Model: This model captures the temporal dynamics of gene expression, accounting for changes in molecular concentrations (repressor and fluorescent protein) over time, including delays in production.
These models were integrated to connect promoter genotype to six distinct gene expression phenotypes, including steady-state levels and dynamic responses to environmental changes. The model was validated against experimental data from a synthetic Lambda bacteriophage PR promoter system.
Datasets
- Experimental Data: The study used experimental data from a synthetic bacterial promoter system (Lambda bacteriophage PR promoter) to measure gene expression dynamics under different environmental conditions (ON and OFF states for repressor protein). This data was used to parameterize and validate the model.
- Existing Databases: The authors utilized data from RegulonDB to obtain sequences of known E. coli promoters and consensus binding sequences for transcription factors.
Key Insights and Ideas
- Mechanistic Models as a Bridge: A mechanistic approach, by explicitly modeling the underlying molecular processes, is crucial for understanding how genotype translates to phenotype and ultimately influences evolution.
- Simplification is Possible: Despite the complexity of biological systems, key evolutionary properties can often be accurately predicted by simplifying the model, focusing on essential parameters rather than every single detail. This suggests that not all mechanistic details are equally important for evolutionary prediction.
- Constraints Shape Evolution: The genotype-phenotype (GP) map is heavily constrained. Mutations don't explore the entire phenotypic space freely; instead, they are limited by biophysical constraints, leading to predictable patterns in how phenotypes can change.
- Order Matters in Evolution: The evolution of complex traits, like regulated gene expression, often follows a specific order. For instance, the ability to express a gene (ON expression) typically emerges before the ability to regulate it (repression).
- Multidimensional Phenotypes are Key: Focusing on multiple phenotypes (beyond just steady-state expression) provides a more accurate picture of evolutionary dynamics and can even accelerate adaptation.
Key Results
- The authors successfully developed a mechanistic model that accurately predicts gene expression dynamics from promoter sequence.
- They found that key evolutionary properties, such as the distribution of mutation effects and evolutionary trajectories, can be captured without accounting for all model parameters. For example, the range of accessible phenotypes (the "envelope") is set by molecular concentrations, while the exploration of that space is influenced by binding energies.
- The study revealed that the GP map for promoters is highly constrained, with double mutants exploring only a fraction of the possible phenotypic space.
- Simulations showed that selecting for all six gene expression phenotypes led to faster evolution compared to selecting for only two (ON and OFF expression), although the former was less precise in reaching the optimal state.
- The distribution of fitness effects (DFEs) of mutations changes during evolution and is largely dependent on the current fitness of the promoter, rather than just its specific genotype.
Limitations and Future Work
- Model Simplifications: The authors acknowledge that their model is a simplification. It doesn't account for all mechanisms of gene regulation (e.g., translational regulation), temporal stochasticity in gene expression, or the full complexity of fitness landscapes (e.g., sign epistasis, interactions between promoter and TF mutations).
- Experimental Validation: While the model was validated against experimental data, the authors suggest that further experimental work is needed to test the generality of their findings across different systems.
- EM Data: The accuracy of the TD model relies on having accurate energy matrices (EMs) for transcription factors. Obtaining these EMs is labor-intensive, and the accuracy of the CI EM was noted as a potential source of error.
- Future Directions: The authors suggest extending their mechanistic approach to more complex gene regulatory networks and exploring how to further reduce model complexity without losing predictive power. They also highlight the need for more experimental data on environment-dependent fitness functions.
Summary updated