BackgroundAcyl-(acyl carrier protein (ACP)) reductase (AAR) is a key enzyme for hydrocarbon biosynthesis in cyanobacteria, reducing fatty acyl-ACPs to aldehydes, which are then converted into hydrocarbons by aldehyde-deformylating oxygenase (ADO). Previously, we compared AARs from various cyanobacteria and found that hydrocarbon yield in Escherichia coli coexpressing AAR and ADO was highest for AAR from Synechococcus elongatus PCC 7942 (7942AAR), which has high substrate affinity for 18-carbon fatty acyl-ACP, resulting in production of mainly heptadecene. In contrast, the hydrocarbon yield was lowest for AAR from Synechococcus sp. PCC 7336 (7336AAR), which has a high specificity for 16-carbon substrates, leading to production of mainly pentadecane. However, even the most productive AAR (7942AAR) still showed low activity; thus, residues within AAR that are nonconserved, but may still be important in hydrocarbon production need to be identified to engineer enzymes with improved hydrocarbon yields. Moreover, AAR mutants that favor shorter alkane production will be useful for producing diesel fuels with decreased freezing temperatures. Here, we aimed to identify such residues and design a highly productive and specific enzyme for hydrocarbon biosynthesis in E. coli.ResultsWe introduced single amino acid substitutions into the least productive AAR (7336AAR) to make its amino acid sequence similar to that of the most productive enzyme (7942AAR). From the analysis of 41 mutants, we identified 6 mutations that increased either the activity or amount of soluble AAR, leading to a hydrocarbon yield improvement in E. coli coexpressing ADO. Moreover, by combining these mutations, we successfully created 7336AAR mutants with ~ 70-fold increased hydrocarbon production, especially for pentadecane, when compared with that of wild-type 7336AAR. Strikingly, the hydrocarbon yield was higher in the multiple mutants of 7336AAR than in 7942AAR.ConclusionsWe successfully designed AAR mutants that, when coexpressed with ADO in E. coli, are more highly effective in hydrocarbon production, especially for pentadecane, than wild-type AARs. Our results provide a series of highly productive AARs with different substrate specificities, enabling the production of a variety of hydrocarbons in E. coli that may be used as biofuels.