No abstract

Amphiphilic inducer molecules such as N-acyl-L-homoserine lactones (AHLs) or isopropyl-β-D-thio-galactopyranoside (IPTG) can be utilized for the implementation of an artificial communication system between groups of E. coli bacteria encapsulated within water-in-oil microemulsion droplets. Using spatially extended arrays of microdroplets, we study the diffusion of both AHL and IPTG from inducer-filled reservoirs into bacteria-containing droplets, and also from droplets with AHL producing sender bacteria into neighboring droplets containing receiver cells. Computational modeling of gene expression dynamics within the droplets suggests a strongly reduced effective diffusion coefficient of the inducers, which markedly affects the spatial communication pattern in the neighborhood of the senders. Engineered bacteria that integrate AHL and IPTG signals with a synthetic AND gate gene circuit are shown to respond only in the presence of both types of sender droplets, which demonstrates the potential of the system for genetically programmed pattern formation and distributed computing.

Reaction circuits mimicking genetic oscillators can be realized with synthetic, switchable DNA genes (so-called genelets), and two enzymes only, an RNA polymerase and a ribonuclease. The oscillatory behavior of the genelets is driven by the periodic production and degradation of RNA effector molecules. Here, we describe the preparation, assembly, and testing of a synthetic, transcriptional two-node negative-feedback oscillator, whose dynamics can be followed in real-time by fluorescence read-out.

With the help of only two enzymes--an RNA polymerase and a ribonuclease--reduced versions of transcriptional regulatory circuits can be implemented in vitro. These circuits enable the emulation of naturally occurring biochemical networks, the exploration of biological circuit design principles and the biochemical implementation of powerful computational models.

In the version of this Article previously published, some incorrect data were included in Fig. 1e. The initial data were inadvertently repeated as the data for 20,000 cycles, and the total number of cycles was stated as 50,000 when it should have been 43,000.