“Our goal has been to put as much chemistry as we can into microbes,” Keasling says. “For advanced biofuels this requires a microbe with pathways for hydrocarbon production and the biomass-degrading capacity to secrete enzymes that efficiently hydrolyze cellulose and hemicellulose. We’ve now been able to engineer strains of Escherichia coli that can utilize both the cellulose and hemicellulose fractions of switchgrass that’s been pre-treated with ionic liquids.”

E. coli bacteria normally cannot grow on switchgrass, but JBEI researchers engineered strains of the bacteria to express several enzymes that enable them to digest cellulose and hemicellulose and use one or the other for growth. These cellulolytic and hemicellulolytic strains of E. coli, which can be combined as co-cultures on a sample of switchgrass, were further engineered with three metabolic pathways that enabled the E. coli to produce fuel substitute or precursor molecules suitable for gasoline, diesel and jet engines. While this is not the first demonstration of E. coli producing gasoline and diesel from sugars, it is the first demonstration of E. coli producing all three forms of transportation fuels. Furthermore, it was done using switchgrass, which is among the most highly touted of the potential feedstocks for advanced biofuels.

Gregory Bokinsky, a post-doctoral researcher with JBEI’s synthetic biology group and lead author of the PNAS paper, explains that the pre-treatment of the switchgrass with ionic liquids was essential to this demonstration.

“The magic is in the ionic liquid pre-treatment,” Bokinsky says. “If properly optimized, I suspect you could use ionic liquid pre-treatment on any plant biomass and make it readily digestible by microbes. For us it was the combination of   biomass from the ionic liquid pretreatment with the engineered E. coli that enabled our success.”

The JBEI researchers also attribute the success of this work to the “unparalleled genetic and metabolic tractability” of E. coli, which over the years has been engineered to produce a wide range of chemical products. However, the researchers believe that the techniques used in this demonstration should also be readily adapted to other microbes. This would open the door to the production of advanced biofuels from lignocellulosic feedstocks that are ecologically and economically appropriate to grow and harvest anywhere in the world. For the JBEI researchers, however, the next step is to increase the yields of the fuels they can synthesize from switchgrass.

“We already have hydrocarbon fuel production pathways that give far better yields than what we obtained with this demonstration,” says Bokinsky. “And these other pathways are very likely to be compatible with the biomass-consumption pathways we’ve engineered into our E. coli. However, we need to find enzymes that can both digest more of the ionic liquid pre-treated biomass and be secreted by E coli. We also need to work on optimizing the ionic liquid pre-treatment steps to yield biomass that is even easier for the microbes to digest.”

Co-authoring the PNAS paper with Keasling and Bokinsky were Pamela Peralta-Yahya, Anthe George, Bradley Holmes, Eric Steen, Jeffrey Dietrich, Taek Soon Lee, Danielle Tullman-Ercek, Christopher Voigt and Blake Simmons.

This research was supported in part by the DOE Office of Science and a UC Discovery Grant.