Proteins evolve new functions in nature through iterated diversification, selection, and amplification. Researchers have developed methods to mimic this approach, enabling the directed evolution of proteins with a variety of altered properties. Many proteins evolved in the laboratory have proven valuable as research tools, novel therapeutics, or industrial reagents. Over the first granting period, we successfully developed several pairs of coupled positive and negative selections that link cell survival or cell death to desired or undesired protein activities, respectively. These coupled selections enabled the evolution of proteins that are conditionally active, or that have truly altered (rather than merely broadened) substrate specificities. Using these methods, we evolved a mutant intein that undergoes protein splicing only in the presence of 4-hydroxytamoxifen (4-HT), a commercially available, cell-permeable small molecule. This evolved ligand-dependent intein renders the activity of a wide variety of proteins dependent on the presence of 4-HT in yeast and mammalian cells. While directed evolution has served as a powerful strategy for generating proteins with desirable functional properties, its scope has been limited by several factors. To date, nearly all directed evolution efforts (including those described above) have adopted a format in which a discrete mutagenesis step generates a gene library, a discrete screen or selection step enriches the library for desired members, a discrete DMA or RNA isolation step harvests desired genes, and a discrete subcloning or retransformation step prepares for the next round of evolution. In contrast, evolution in nature occurs in a continuous, asynchronous format in which mutation, selection, and replication are simultaneously and constantly ongoing. A continuous format has the potential to dramatically enhance the effectiveness of directed evolution efforts by enabling an enormous number of effective """"""""rounds"""""""" of evolution to take place in a single experiment, while accessing much larger library sizes at any instance than is possible using existing methods. As a result, a much larger swath of sequence space can be explored, and the total number of steps in sequence space that can be traversed during evolution is increased enormously. Despite these significant benefits, continuous directed evolution remains a largely unrealized goal. Over the next granting period we propose both to extend our previous intein evolution studies as well as to develop a system for the continuous directed evolution of proteins and nucleic acids. If successful, these lines of research are poised to illuminate complex signalling networks underlying stem-cell differentiation and, more generally, to expand significantly the ability of scientists to evolve in the laboratory tailor-made solutions to complex biological problems relevant to human health.
Showing the most recent 10 out of 38 publications
