Genetic Process Engineering - Cellular Computing - page 45

monitor their environment and continuously make decisions on how to react to
the given conditions. Moreover, from prokaryotes to eukaryotes, cells act both
as individual units and as a contributing part of larger and complex multicellular
systems or organisms. Given these attributes and inherent complexity, how can
we successfully modify and harness biological organisms for our purposes?
Controlled gene expression using engineered in vivo digital-logic circuits
and intercellular communications enables programmed cell behavior that is
complex, predictable, and reliable. Our approach integrates several layers that
include a library of well-characterized simple components synthesized to have
the appropriate behavior, a methodology for combining these components into
complex intracellular circuitry and multicellular systems with predictable and
reliable behavior, and software tools for design and analysis.
The first step in making programmed cell behavior a practical and useful
engineering discipline is to assemble a component library. For this purpose,
we engineered cellular gates that implement the NOT, IMPLIES, and AND
logic functions. These gates are then combined into biochemical logic circuits
for both intracellular computation and intercellular communications. In these
biocircuits, chemical concentrations of specific messenger RNA (mRNA) and
inducer molecules represent the logic signals. The logic gates perform compu-
tation and communications using mRNA, DNA-binding proteins, small inducer
molecules that interact with these proteins, and segments of DNA that regu-
late the expression of the proteins. For example, Figure 4.2 describes how a
cellular inverter achieves the two states in digital inversion using these genetic
regulatory elements.
Given a library of components, biocircuit design is the process of assembling
preexisting components into logic circuits that implement specific behaviors.
The most important element of biocircuit design is matching logic gates such
that the couplings produce the correct behavior. Typically, naturally occurring
components have widely varying kinetic characteristics, and arbitrarily com-
posing them into circuits is not likely to work. We demonstrate genetic process
engineering —modifying the DNA encoding of existing genetic elements until
they achieve the desired behavior for constructing reliable circuits of significant
complexity. The genetic modifications produce components that implement dig-
ital computation with good noise margins, signal restoration, and appropriate
standard interfaces for complex system composition.
An important aspect of this work is engineering biological systems to ex-
hibit digital behavior because the digital abstraction is both convenient to use
and feasible. The digital abstraction is a useful programming paradigm be-
cause it offers a reliable and conceptually simple methodology for constructing
complex behavior from a small number of simple components [18]. Digital
computation provides reliability by reducing the noise in the system through
signal restoration. For each component in the computation, the analog output