Yiannis N. Kaznessis University of Minnesota BES-0644792
The research objective of this project is to develop the mathematics of biological engineering and to investigate and design inducible gene expression systems combining theory and experiments. The education objective is to train and attract a new generation of engineers and scientists to computational and experimental systems biology. Exposing high school teachers and students to computational bioengineering is a specific educational/outreach aim.
Intellectual merit:
The current rapid expansion of biological knowledge offers a great opportunity to rationally engineer biological systems that respond to signals. The large number of components and interactions involved in dynamic gene regulation warrants a quantitative systems biology perspective. The creation of mathematical theories and accurate models of all known molecular events involved in transcriptional/translational regulation can provide new descriptive and predictive insight into the dynamic behavior of gene networks. The proposed activities advance knowledge in computational systems and synthetic biology. They also result in novel gene regulatory networks, such as bio-logical AND gates.
Broader impacts:
The ambitious idea of engineering cells that will function as miniature factories has given rise to new fields of research, systems and synthetic biology. The objectives are the design and construction of new biological parts, devices and systems from natural biological systems. The computational bioengineering theories and algorithms proposed will positively impact rational biological engineering. Broad applications range from biofuel development, to detectors for biochemical and chemical weapons, to devices that will remove environmental pollutants, to disease diagnosis, to gene therapies, even to engineered microbes to produce hydrogen from sunlight and water.
Computational and mathematical biology, implemented with an eye towards engineering applications, are exciting fields. Assisting high school teachers to employ the considerable computational biology resources at the University of Minnesota and teaching them how mathematics and biology can be combined productively in the computer is a priority detailed in this project. Expected outcomes are participants with the ability to instruct high school students in the fundamental concepts of computational biology, and eventually high school students interested in pursuing a science/engineering career.
With NSF support and support from other federal and state government agencies, efforts in our research group are culminating into two major developments: 1) the development of new antibiotic technologies; and 2) the solution of the chemical master equation for nonlinear chemical reaction networks. Development of Antibiotic Technologies Antibiotics have changed the course of human history, saving innumerable lives and substantially improving the quality of life across the globe. Bacteria are fighting back though. Superbugs, species that do not respond to our most potent drugs, are evolving, infecting humans with alarming frequency. And our drug war chest is not being updated with new antibiotics because of this oxymoronic financial disincentive: one dose, one treatment with antibiotics may cure the disease, in stark contrast to medicines prescribed for life. The PI started working on antibiotics during his postdoctoral fellowship at Parke-Davis (which was acquired by Pfizer in 2001). At the University of Minnesota the PI started modeling antimicrobial peptides (AMPs). These are small proteins produced by most all higher organisms. AMPs lyse and kill bacteria rapidly. The prevailing models of AMP function involve interactions between peptides and cell membranes, the disruption of the membrane integrity and eventual cell death. However, there is a significant disconnect between our knowledge of molecular level events and dead bacteria. We developed a complete, multiscale modeling approach that offers a mechanistic understanding of how protegrin-1, a model AMP, kills bacteria. With molecular simulations, we studied the protegrin action against bacteria which involves the attraction and adsorption of cationic protegrin peptides to anionic bacterial membranes, insertion into the hydrophobic membrane core, followed by dimerization of the peptides, formation of higher aggregates, and formation of transmembrane pores that facilitate an unrestricted, lethal flux of ions from the cytoplasm. Our models capture the biophysical events that ultimately result in cell death. This is the type of understanding that is not available by experimental techniques alone. Yet we were not any closer to proposing a new therapeutic solution than before: AMPs cannot be delivered orally because as proteins are quickly degraded in the stomach of hosts; AMPs cannot be delivered by injecting hosts, because at high concentrations they are often toxic to liver cells. We are overcoming the delivery challenge by packaging AMPs as DNA in lactic acid bacteria (LAB), organisms that are safe to consume. The use of LAB as drug delivery vectors in the gastrointestinal tract is an intriguing approach to prevent or treat intestinal colonization by pathogenic strains. In a recent work we demonstrated how modified LAB can produce and secrete AMPs that inhibit foodborne pathogens Salmonella Typhimurium and E. coli O157:H7. We are now engineering LAB that can detect pathogens in their environment. When LAB sense the presence of pathogens they produce and secrete a molecular arsenal of AMPs that specifically target the sensed bacteria. We believe that these molecular tools and antimicrobial peptides will have a significant impact, because they may serve as a template for the development of antibiotic technologies to tackle bacterial strains, against which our antibiotic arsenal is dwindling. Closure Scheme for Chemical Master Equations We used mathematical models in order to improve our chances of engineering AMPs and their delivery by LAB. These models capture randomness, a defining feature of biochemical reaction networks, with molecular fluctuations influencing cell physiology. Chemical master probability equations (CME) completely govern the dynamic and steady state behavior of stochastic reaction networks. My group joined the efforts of a large community, working to improve the computational efficiency and accuracy of algorithms that simulate stochastic chemical reacting systems. We developed a hybrid stochastic-discrete and stochastic-continuous modeling formalism for treating reacting systems that span multiple time scales. We made our algorithms publicly available on sourceforge.net, where they have been downloaded more than 5,000 times since 2008. It was our work on probability moments that led to the solution of the master equation. Our closure scheme finds the lower moments by maximizing the information entropy of any reaction network. We presented a solution to this problem that had remained unsolved for over seventy years. Ours is a closure scheme for master probability equations of chemical/biochemical reaction networks that include reactions of second or higher order. We believe that this method has significant theoretical implications. We also believe that it will propel significant discoveries in the biological sciences. It will arm the scientific community with a potent method to model the dynamic behavior of nonlinear, stochastic biomolecular systems.
