= Synthetic biology
\b[Synthetic biology] designs and constructs biological components or systems to test principles or produce specified functions. It connects <molecular genetics>, <biotechnology> and quantitative models: the objective is not only to change a sequence, but to make a biological system exhibit a defined input-output or dynamical behaviour. Construction and measurement then test whether the proposed design actually produces it.
A motivating primary example is the <genetic toggle switch> reported in 2000 by <Timothy S. Gardner>, <Charles R. Cantor> and <James J. Collins>. They constructed a bistable <gene regulatory network> in <Escherichia coli> and switched it using transient chemical or thermal induction.<Construction of a genetic toggle switch in Escherichia coli by Timothy S. Gardner and others>[Nature 403, 339–342, Abstract]{ref} Its value is a memory-like property: the observed state depends on the system's history, not merely on the current input.
An original ideal model shows how mutual inhibition can create that behaviour. Let $x,y$ be dimensionless concentrations of two components, each suppressing production of the other:
$$
\frac{dx}{dt}=\frac{A}{1+y^2}-x,\qquad
\frac{dy}{dt}=\frac{A}{1+x^2}-y.
$$
Time and concentrations have been rescaled; $A$ is a stipulated production parameter. At equilibrium, subtracting the two equations after multiplication by their denominators gives
$$
(x-y)(1-xy)=0.
$$
Asymmetric equilibria therefore satisfy $xy=1$ and $x+y=A$. They exist for $A>2$, with concentrations $(A\pm\sqrt{A^2-4})/2$.
At $A=10$, the asymmetric pairs are approximately $(9.899,0.101)$ and $(0.101,9.899)$. Their <Jacobian matrix> eigenvalues are $-1\pm2/A$, both negative, giving local <linear stability>. The symmetric equilibrium is $(2,2)$, whose eigenvalues are $-1\pm1.6$; one is positive. The ideal model thus has two stable states separated by an unstable one. This calculation is a qualitative design example, not a fitted reconstruction of the 2000 experimental data.
A transient input can move the state between the two attraction regions; after removal, the internal feedback can maintain the new state. <Bistability> is therefore a property of the combined feedback and response parameters, not a label guaranteed by drawing two inhibitory arrows. Molecular noise, component burden and growth can change actual switching and memory.
Experimentally, compare both initial states under the same final conditions, follow responses after input removal and retain controls lacking the proposed feedback. A fluorescent reporter and <flow cytometry> can distinguish a population mixture from a change in the mean signal; these are proposed diagnostic readouts here, without attributing uninspected procedural details to the original paper. The design-model-measurement cycle makes construction a test of biological explanation as well as an engineering activity. <Genetic toggle switch stability margin> is one quantitative question that connects those two purposes.
Back to article page
