Modular and selective biosynthesis of gasoline-range alkanes

• E. coli was engineered to selectively produce C3 through C9 alkanes.
• Fatty acid synthesis chain elongation resulted in C7 and C9 alkane synthesis.
• More theoretically efficient reverse beta-oxidation was used to make C3–C5 alkanes.
• Use of AD_A122F from Nostoc punctiforme PCC73102 slightly improves synthesis.
• ADs can catalyze conversion of branched FFA precursors into corresponding alkanes.

Typical renewable liquid fuel alternatives to gasoline are not entirely compatible with current infrastructure. We have engineered Escherichia coli to selectively produce alkanes found in gasoline (propane, butane, pentane, heptane, and nonane) from renewable substrates such as glucose or glycerol. Our modular pathway framework achieves carbon-chain extension by two different mechanisms. A fatty acid synthesis route is used to generate longer chains heptane and nonane, while a more energy efficient alternative, reverse-β-oxidation, is used for synthesis of propane, butane, and pentane. We demonstrate that both upstream (thiolase) and intermediate (thioesterase) reactions can act as control points for chain-length specificity. Specific free fatty acids are subsequently converted to alkanes using a broad-specificity carboxylic acid reductase and a cyanobacterial aldehyde decarbonylase (AD). The selectivity obtained by different module pairings provides a foundation for tuning alkane product distribution for desired fuel properties. Alternate ADs that have greater activity on shorter substrates improve observed alkane titer. However, even in an engineered host strain that significantly reduces endogenous conversion of aldehyde intermediates to alcohol byproducts, AD activity is observed to be limiting for all chain lengths. Given these insights, we discuss guiding principles for pathway selection and potential opportunities for pathway improvement.