Exploring Impact of Hydrogen Gene Engineering, Cell Density, and Fermentation Mode on Isopropanol Production and Testing Fluorescence Based Population Tracking in a Clostridium Coculture

• Paige Bastek, Chemical Engineering, University of Delaware

• Eleftherios Papoutsakis, Chemical and Biomolecular Engineering, University of Delaware

In recent years, there has been an increased focus on biosynthesizing isopropanol using microbial systems. A system of interest in the Papoutsakis lab is culturing the two anaerobes Clostridium acetobutylicum (Cac) and Clostridium ljungdahlii (Clj) together for efficient biosynthesis of isopropanol, a valuable commodity chemical. 

Optimizing this system involves maximizing hydrogen production in Cac. Clj can reduce the acetone produced by Cac into isopropanol. Promoting more hydrogen production by Cac diverts electrons from production of undesired byproducts, like ethanol. To increase hydrogen production, three genes of interest were transformed into Cac: gapN, Pos5, and hydA. Three strains of Cac were constructed with binary combinations of these genes and tested for the most efficient hydrogen production. The best performing strain, expressing the gapN-hydA combination, produced 140% more hydrogen than the parent strain.

Another important coculture parameter is initial cell density of the two species. To further examine this parameter, fed batch serum bottle cultures were started with five times the amount of cells typically added, leading to high solvent productivity and exogenous hydrogen utilization. To simulate a continuous process on a small scale, an additional set of cultures, at high and normal cell densities, were grown in falcon tubes and resuspended in fresh media every 12 hours. These cultures demonstrated the highest solvent productivity demonstrated to date in this system.

Another coculture parameter of interest is the population ratio. Current methods, such as quantitative PCR or fluorescent protein tracking, take time, must be performed retroactively, or require genetic engineering and antibiotic selection. Our goal was to determine if differences in fluorescence of each bacteria at distinct wavelengths could be exploited to measure the population ratio, as has been demonstrated in other coculture systems. Fluorescent scans of Cac, Clj, and Escherichia coli showed no consistent differences in fluorescence between each species of large enough magnitude to enable robust population tracking.