UCSF

Living organisms are comprised of crowded and highly interconnected networks of biological molecules and metabolites. The scale and complexity of these networks makes them difficult to completely characterize, even for relatively streamlined organisms like the bacterium Escherichia coli K-12. After decades of work, many aspects of the growth and physiology of E. coli remain poorly understood. Its regulatory pathways remain only partially mapped. Functional-genomic approaches hold the promise of speeding discovery and characterization of the molecular networks underpinning life.

In this work, a chemical-genomic screen was conducted for two purposes. First, we were interested in the genetic-networks that make up an E. coli cell. Could we build on previous work to expand our knowledge of gene function in this model bacterium? Second, we were interested in pushing the resolution of functional genomics, not to characterize the relationship between genes in an organism but to characterize the relationships of amino acids and structural features of a single enzyme; RNA polymerase (RNAP).

To accomplish these tasks, we first used oligo-mediated recombineering to build a library of over 150 mutants in RNAP. We then screened this library of mutant strains, along with the KEIO deletion library of E. coli, for growth across more than 100 unique chemical stresses using colony array technology. This generated a map that expanded the chemical-sensitivity landscape of E. coli.

From features within this landscape, we identified a unique mechanism of illicit transport wherein the translation inhibitors kasugamycin and blasticidin S gain access to the cytoplasm of E. coli by hijacking the peptide ABC-importers Dpp and Opp. We used genetic analysis, in vivo translation assays, and in vitro binding assays to show that the peptide ABC-importers directly import the two drugs.

Next, we identified a new binding interface within RNAP that allows for the binding and function of the critical transcription regulator DksA. We used genetic analysis, in vitro site-specific crosslinking, molecular modeling, and functional assays of transcription both in vivo and in in vitro. We found that the lineage specific insertion SI1 directly binds DksA and is critical for the effects of DksA on transcription.