Cellular State Machines for Programming Complex, State-dependent Genetic Circuits

Technology #18230

Questions about this technology? Ask a Technology Manager

Download Printable PDF

Image Gallery
Implementing a 2-input, 5-state cellular state machine
Professor Timothy Lu
Department of Biological Engineering, MIT
External Link (be.mit.edu)
Nathaniel Roquet
Center for Integrative Synthetic Biology, MIT
Managed By
Jon Gilbert
MIT Technology Licensing Officer
Patent Protection

Biological State Machines

PCT Patent Application WO 2017-059187

Biological State Machines

US Patent Pending US 2017-0096680


Cellular state machines emulate digital circuits by integrating multiple binary inputs (e.g., chemical inducers) and producing outputs (e.g., gene expression). They have wide-ranging biotechnological applications. For example, they can be used to program and model cellular behaviors such as development and differentiation by controlling the timing and sequence of transcription factor cascades. They can be used as cellular biosensors in which the output (i.e., the DNA sequences or gene expression patterns that comprise the "state" of the cell) is a function of the input (i.e., the pattern of exposure to chemical inducers). Moreover, they can be implemented in biomanufacturing processes in which it would be advantageous to cycle cells through tightly regulated phases (e.g., biomass accumulation, biomolecule production, and cell lysis). 

Problem Addressed

Traditional methods for implementing conditional gene expression tend to be limited with respect to the complexity of the networks they induce. Often, the gene of interest is rendered ‘on’ or ‘off’ in response to an environmental or intrinsically regulated cue. Context-dependent gene expression regulation has been achieved, but it is often transcription-based and as such, requires chronic induction (e.g., doxycycline treatment). Traditional inducible platforms are also limited with respect to combinatorial regulation of multiple genes in a single cell. Cellular state machines are uniquely capable of implementing tightly regulated genetic cascades that are dependent on both the order and combination of inputs. This greatly increases the complexity of programmable genetic networks, permitting the modeling and optimization of sophisticated biological processes such as cellular differentiation and biomanufacturing. 


State machines are used as computing devices that integrate input signals and respond with a context-dependent output. Each output represents one possible “state” in which the machine can exist. State machines are order dependent, meaning that the sequential state depends on both the input and the current state. State machines can exist in cells, whereby synthetic gene networks drive the transition between cellular states in an input- and state-dependent manner. This E. coli-based system produces unique gene expression patterns that define the cellular “state” in response to chemical inducer inputs. Both the identity and order of inducers matter, making it possible to distinguish between ‘Input A then Input B’ and ‘Input B then Input A.’ The chemical inducers manipulate the state of the cell by driving the expression of recombinases, which invert or excise sections of DNA (e.g., promoters, terminators, and output genes) based on the orientation of recognition sites. Modulating the assembly of the targeted gene circuit components enables complex and varied states. The gene circuit assembly is modular and can be achieved in one step. In contrast to conditional gene expression methods that require constant induction (e.g., those based on transcription), this system maintains a stable memory even after inputs are withdrawn. Thus, cells need only transient induction in order to be locked into a state, allowing for the implementation of sequential inputs and the interrogation of long-term cellular history.  


  • Multiplexed reporting of inputs with single output
  • Modular, one-step assembly of circuit
  • Flexible design permits multiple paths to implement the same input/output behavior
  • Maintains state memory after inputs are withdrawn
  • Can detect the order of chemical events
  • Can program increased number of output behaviors as compared to state-less induction platforms