P450-dependent biotransformations in Escherichia coli are attractive for the selective oxidation of organic molecules using mild and sustainable procedures. The overall efficiency of these processes, however, relies on how effectively the NAD(P)H cofactors derived from oxidation of the carbon source are utilized inside the cell to support the heterologous P450-catalyzed reaction. In this work, we investigate the use of metabolic and protein engineering to enhance the product-per-glucose yield (Y(PPG)) in whole-cell reactions involving a proficient NADPH-dependent P450 propane monooxygenase prepared by directed evolution [P450(PMO)R2; Fasan et al. (2007); Angew Chem Int Ed 46:8414-8418]. Our studies revealed that the metabolism of E. coli (W3110) is able to support only a modest propanol: glucose molar ratio (YPPG ~ 0.5) under aerobic, nongrowing conditions. By altering key processes involved in NAD(P)H metabolism of the host, considerable improvements of this ratio could be achieved. A metabolically engineered E. coli strain featuring partial inactivation of the endogenous respiratory chain (Δndh) combined with removal of two fermentation pathways (ΔadhE, Δldh) provided the highest Y(PPG) (1.71) among the strains investigated, enabling a 230% more efficient utilization of the energy source (glucose) in the propane biotransformation compared to the native E. coli strain. Using an engineered P450(PMO)R2 variant which can utilize NADPH and NADH with equal efficiency, we also established that dual cofactor specificity of the P450 enzyme can provide an appreciable improvement in Y(PPG). Kinetic analyses suggest, however, that much more favorable parameters (K(M), k(cat)) for the NADH-driven reaction are required to effectively compete with the host's endogenous NADH-utilizing enzymes. Overall, the metabolic/protein engineering strategies described here can be of general value for improving the performance of NAD(P)H-dependent whole-cell biotransformations in E. coli.