MULTISCALE SYSTEMS AND SYNTHETIC BIOLOGY
Our laboratory takes a systems and synthetic biology approach to understanding and designing biology at multiple scales – proteins, microbial pathways and microbiota. We leverage computational protein design (Rosetta), next-generation genomics technologies and high-throughput phenotyping to address these questions.
Designing new small molecule biosensors
Small-molecule inducible transcription factors are useful as in vivo biosensors. These biosensors are widely used in synthetic biology for engineering biosynthetic pathways, as regulators of endogenous and synthetic genetic circuits, and for the real-time monitoring of metabolite concentrations. Natural transcription factors only sense a limited number of small molecules. We aim to expand the repertoire of biosensors for designing natural transcription factors to many new small molecule classes. To design novel biosensors, we have developed a hybrid approach that combines computational protein design and directed evolution. We are also interested in redesigning DNA specificity and constructing synthetic inducible promoters by modular assembly of regulatory DNA sequences.
Elucidating allostery at molecular resolution by high-throughput reverse genetics
Allostery is a fundamental biological property by which a protein senses environmental cues, such as a metabolite, and changes shape to trigger a downstream action like DNA-binding or catalysis. Allostery is a fascinating phenomenon because the response action occurs at a region spatially far away from the metabolite-binding site. Our current understanding of allostery is largely limited to biophysical models that explain conformational transitions between allosteric states, but do not explain the molecular underpinnings of allosteric communication. Little is known about how protein residues act in concert to propagate the allosteric signal (like a domino effect) or which residues are spatially interconnected to elicit action at a distant site. Using allosteric transcription factors as our model system, our objective is to understand the general principles of protein structure that underlies allosteric communication at molecular resolution. We will use high-throughput, protein-wide mutational screens coupled with next generation sequencing to functionally characterize millions of mutants of a candidate protein. We will apply machine learning on mutational dataset to infer common spatial and structural features of the allosteric network.
Directed evolution of biosynthetic pathways
We use small molecule biosensors to evolve microbes for the biosynthetic production of useful chemicals and fuels. The biosensor coupled to a selection marker enables phenotyping hundreds of millions of cells to enrich the rare high-producers. We are currently developing biosensors to measure cellular redox states (NADH/NAD+). A number of valuable secondary metabolites and fuels require additional reducing power, by way of NADH, to be supplied beyond what is available through normal metabolic processes. Furthermore, NADH and NAD+ are cofactors that are involved in hundreds of redox reactions, and the ratio of NADH to NAD+ has an important effect on maintaining the redox balance required for metabolism and growth. Using redox sensors, would like to dynamically control metabolic pathways to channel reducing equivalents toward the biosynthetic objective. We will also investigate genome-scale effects on electron-transport.
Discovering new enzymes and pathways from environmental microbiomes
The widespread use of xenobiotics, such as herbicides, industrial chemicals, plastics, explosive agents and antibiotics, is a major cause of environmental destruction. Xenobiotics deteriorate soil quality, disrupt delicate ecosystems, and profoundly affect human health through the food chain or water. Xenobiotics may persist for years in the environment because they are not easily degraded by environmental changes in pH or temperature, or by common catabolic pathways. However, some microbes in the environment have evolved xenobiotic-degrading enzymes. Since the vast majority of soil microbes cannot be cultured in a laboratory, these enzymes remain unidentified. We are developing a powerful alternative strategy by directly assaying the function of each of hundreds of millions of enzyme candidates from environmental microbiota with an ultra high-throughput in vivo screen using biosensors. We will study evolution promiscuous activity, new enzymatic routes for bioremediation, with the objective of making superior bioremediation strains by combining pathways.