Modeling and Simulation of Synthetic, Methylation-based Epigenetic Memory Systems

Artificial epigenetic memory systems hold immense potential for a variety of innovative applications, most notably in the development of bacteria as live sensors. Bacterial biosensors are highly advantageous because they are simple and cost-effective to produce, and they can be cultivated under basic conditions for the continuous, long-term monitoring of target environments. Crucially, these engineered biosensors store environmental signals directly within DNA methylation patterns, bypassing the need for expensive electronic hardware. The potential use cases are vast: one could imagine designer bacteria navigating the human body to detect disease markers and metabolic states, depositing this diagnostic data into their DNA to be read later. Other biotechnological applications include robust epigenetic biocontainment systems, where toxic genes are kept under the control of a memory circuit to prevent engineered bacteria from escaping bioreactors. Ultimately, these synthetic epigenetic circuits pave the way for designer bacteria capable of functioning as live diagnostic sensors, biological data storage devices, and even living microprocessors.

In the first phase of this project, our collaboration partners in the group of Albert Jeltsch (Institute of Biochemistry and Technical Biochemistry) utilized DNA methylation to create synthetic gene circuits. These circuits are designed to operate as bistable switches with two stable states: an OFF-state and an ON-state. Various triggers, such as heat or arabinose induction, can be used to switch the system into the ON-state. On our side of the project, Viviane Klingel successfully replicated the dynamics of this system using mathematical modeling. 

For the second phase of the project, we aim to further develop the fundamental properties of these artificial epigenetic memory systems. We plan to achieve this by implementing additional input signals, improving signal processing and memory functions, and developing multiplexed systems capable of data processing using logic gates.

Broadly speaking, our group supports the development of these epigenetic memory systems by providing mathematical explanations for experimental results and offering model-based predictions for future experiments. Typically, our workflow involves first calibrating our models to experimental data using parameter optimization, followed by rigorous analysis. This process grants us deeper insights into the underlying biochemical mechanisms. For example, bifurcation analysis helps us understand how different parameters influence the stable states of the system. Armed with this information, we can recommend targeted improvements to specific properties—such as making the ON- and OFF-states more stable or fine-tuning the system's sensitivity to trigger inputs.

A major realization from the first phase of the project was the significant role of stochastic effects, particularly concerning the long-term stability of the memory. To address this, we aim to incorporate single-cell measurements into our models. By utilizing a population model, we can accurately reflect the inherent heterogeneity within cell populations. This modeling approach enables us to distinguish between different subpopulations—such as cells that reliably remain in the ON-state versus those that slowly drift back to the OFF-state after varying durations.

Ultimately, this project aims to foster a comprehensive understanding of the biological mechanisms at play and to provide valuable, predictive insights that will further improve these innovative synthetic systems.