Cells switch between various stable genetic programs (attractors) to accommodate environmental conditions. Signal transduction machineries efficiently convey environmental changes to the gene regulation apparatus in order to express the appropriate genetic program. However, since the number of environmental conditions is much larger than that of available genetic programs so that the cell may utilize the same genetic program for a large set of conditions, it may not have evolved a signaling pathway for every environmental condition, notably those that are rarely encountered. Here we show that in the absence of signal transduction, switching to the appropriate attractor state expressing the genes that afford adaptation to the external condition can occur. In a synthetic bistable gene switch in Escherichia coli in which mutually inhibitory operons govern the expression of two genes required in two alternative nutritional environments, cells reliably selected the “adaptive attractor” driven by gene expression noise. A mathematical model suggests that the “non-adaptive attractor” is avoided because in unfavorable conditions, cellular activity is lower, which suppresses mRNA metabolism, leading to larger fluctuations in gene expression. This, in turn, renders the non-adaptive state less stable. Although attractor selection is not as efficient as signal transduction via a dedicated cascade, it is simple and robust, and may represent a primordial mechanism for adaptive responses that preceded the evolution of signaling cascades for the frequently encountered environmental changes.

The ability to evolve is a key characteristic that distinguishes living things from non-living chemical compounds. The construction of an evolvable cell-like system entirely from non-living molecules has been a major challenge. Here we construct an evolvable artificial cell model from an assembly of biochemical molecules. The artificial cell model contains artificial genomic RNA that replicates through the translation of its encoded RNA replicase. We perform a long-term (600-generation) replication experiment using this system, in which mutations are spontaneously introduced into the RNA by replication error, and highly replicable mutants dominate the population according to Darwinian principles. During evolution, the genomic RNA gradually reinforces its interaction with the translated replicase, thereby acquiring competitiveness against selfish (parasitic) RNAs. This study provides the first experimental evidence that replicating systems can be developed through Darwinian evolution in a cell-like compartment, even in the presence of parasitic replicators.

Liposomes have long been used as possible compartments for artificial cells, and it has been shown that liposomes can sustain various types of biochemical reactions. To elevate the degree of molecular complexity of the system in liposomes, we have constructed a two-stage genetic network encapsulated in liposomes. This two-stage genetic network was constructed with the plasmid pTH, in which the protein product of the first stage (T7 RNA polymerase) is required to drive the protein synthesis of the second stage (GFP). We show that the two-stage genetic network constructed in a cell-free expression system is functional within liposomes.