The Cello design environment is based on the hardware-independent descriptive language Verilog and a 'user constraint file' that specifies the DNA sequences for the logic gates, physical location in a plasmid or genome, organism, valid operating conditions, architectural constraints and preferred motifs. Users pick a constraint file and write a program in Verilog, after which Cello designs genetic circuit diagrams based on the limitations imposed by the constraint file and ensures that the gates' outputs and input thresholds are compatible. The linear DNA sequence is derived from a set of parts fitting these constraints in the Eugene specification language. Cello then simulates the function of the circuit, providing an analysis of the performance and the impact on cell growth. Circuits predicted to impose an unacceptable burden on cell growth are flagged to the user.
The group initially generated basic NOT gates using Tet repressor sequences and, after optimization, including the use of strong insulator sequences to insulate the gates from their genomic context, they had derived a gate library that could achieve a 100% success rate with no post-design tuning. Expanding on this, they designed more complex gates to prioritize inputs and handle multiple inputs for one output. Fifty-two circuits were designed and, with no further optimization, 37 functioned as predicted. The 52 circuits had a total of 412 possible outputs from the gates, with 92% behaving as predicted. The largest circuit, the 'Consensus' circuit (which required three inputs to agree) has 10 regulatory proteins and 55 unique genetic parts.
This is a preview of subscription content, access via your institution