Abstract: Similar to neural networks in animals, molecular networks in cells can generate bistable or oscillatory dynamics that maintain memories of previous events (e.g. epigenetic switch) or order periodic events (e.g. cell cycle), respectively. In cells, networks of genes interacting with one another through regulatory feedback implement such dynamics. Analogous to learning in brains, cells can """"""""""""""""learn"""""""""""""""" correlations in their environmental signals by encoding such correlations into their gene network dynamics through mutation, a """"""""""""""""re-wiring"""""""""""""""" process that occurs on the timescale of generations. The issue of learning the statistical regularities and correlations of environmental signals is best exemplified by the evolution of """"""""""""""""circadian clocks"""""""""""""""", which are oscillatory gene circuits that have learned to internalize the 24-hour light-dark circadian cycle. Strikingly, circadian clocks have evolved independently multiple times, which suggests there exists some selection pressure and/or evolutionary mechanism that repeatedly favor the convergent evolution of autonomous oscillation. The hypothesis of my research proposal is that certain types of loss-of-function mutations in duplicated genes (known as dominant-negative mutations) can easily generate bistability and oscillation in existing regulatory networks. Gene duplication followed by a loss-of-function mutation can generate a dominant-negative. A dominant negative mutation is a partial loss-of-function mutation that renders a gene duplicate functionally inactive, yet still capable of interacting with the original duplicate, the upstream effectors, and/or downstream targets. Thus, dominant-negatives can easily interfere with the proper regulation and activity of the original duplicate. Because both gene duplication and loss-of-function mutations occur frequently in evolution, this presents an evolutionary mechanism for rapidly generating bistability and autonomous oscillation in gene regulatory networks. My proposed research over the next five years will integrate experiment and theory to understand the extent to which gene duplication and dominant-negative mutations facilitate the evolution of epigenetic switches and circadian clocks in regulatory networks. We will use computer simulation and an experimental directed evolution approach in a tractable, model eukaryote (Saccharomyces cerevisiae) to test the ability of cells to learn the statistical regularities of their coupled environmental signals. Understanding how and why single-cell microbes and parasites have learned to predict their environment is essential for understanding their future evolution to changing host conditions. Public Health Relevance: The ability of parasites to learn and adapt to changing host conditions and environments presents a challenge to human health. The objective of my research proposal is to understand the capacity of gene networks in single cells to learn and predict the statistical regularities of their environment. Discovering the limitations and abilities of parasites to evolve and anticipate changes in their host environment will be invaluable for the treatment of many human diseases.
Showing the most recent 10 out of 11 publications
