Toward a Semisynthetic Stress Response System To Engineer Microbial Solvent Tolerance

ABSTRACT Strain tolerance to toxic metabolites is an important trait for many biotechnological applications, such as the production of solvents as biofuels or commodity chemicals. Engineering a complex cellular phenotype, such as solvent tolerance, requires the coordinated and tuned expression of several genes. Using combinations of heat shock proteins (HSPs), we engineered a semisynthetic stress response system in Escherichia coli capable of tolerating high levels of toxic solvents. Simultaneous overexpression of the HSPs GrpE and GroESL resulted in a 2-fold increase in viable cells (CFU) after exposure to 5% (vol/vol) ethanol for 24 h. Co-overexpression of GroESL and ClpB on coexisting plasmids resulted in 1,130%, 78%, and 25% increases in CFU after 24 h in 5% ethanol, 1% n-butanol, and 1% i-butanol, respectively. Co-overexpression of GrpE, GroESL, and ClpB on a single plasmid produced 200%, 390%, and 78% increases in CFU after 24 h in 7% ethanol, 1% n-butanol, or 25% 1,2,4-butanetriol, respectively. Overexpression of other autologous HSPs (DnaK, DnaJ, IbpA, and IbpB) alone or in combinations failed to improve tolerance. Expression levels of HSP genes, tuned through inducible promoters and the plasmid copy number, affected the effectiveness of the engineered stress response system. Taken together, these data demonstrate that tuned co-overexpression of GroES, GroEL, ClpB, and GrpE can be engaged to engineer a semisynthetic stress response system capable of greatly increasing the tolerance of E. coli to solvents and provides a starting platform for engineering customized tolerance to a wide variety of toxic chemicals. IMPORTANCE Microbial production of useful chemicals is often limited by the toxicity of desired products, feedstock impurities, and undesired side products. Improving tolerance is an essential step in the development of practical platform organisms for production of a wide range of chemicals. By overexpressing autologous heat shock proteins in Escherichia coli, we have developed a modular semisynthetic stress response system capable of improving tolerance to ethanol, n-butanol, and potentially other toxic solvents. Using this system, we demonstrate that a practical stress response system requires both tuning of individual gene components and a reliable framework for gene expression. This system can be used to seek out new interacting partners to improve the tolerance phenotype and can be used in the development of more robust solvent production strains