Abstract
BACKGROUND: Several recent studies have reported successful hydrogen (H2) production achieved via recombinant expression of uptake [NiFe]-hydrogenases from Hydrogenovibrio marinus, Rhodobacter sphaeroides, and Escherichia coli (hydrogenase-1) in E. coli BL21(DE3), a strain that lacks H2-evolving activity. However, there are some unclear points that do not support the conclusion that the recombinant hydrogenases are responsible for the in vivo H2 production. RESULTS: Unlike wild-type BL21(DE3), the recombinant BL21(DE3) strains possessed formate hydrogen-lyase (FHL) activities. Through experiments using fdhF (formate dehydrogenase-H) or hycE (hydrogenase-3) mutants, it was shown that H2 production was almost exclusively dependent on FHL. Upon expression of hydrogenase, extracellular formate concentration was changed even in the mutant strains lacking FHL, indicating that formate metabolism other than FHL was also affected. The two subunits of H. marinus uptake [NiFe]-hydrogenase could activate FHL independently of each other, implying the presence of more than two different mechanisms for FHL activation in BL21(DE3). It was also revealed that the signal peptide in the small subunit was essential for activation of FHL via the small subunit. CONCLUSIONS: Herein, we demonstrated that the production of H2 was indeed induced via native FHL activated by the expression of recombinant hydrogenases. The recombinant strains with [NiFe]-hydrogenase appear to be unsuitable for practical in vivo H2 production due to their relatively low H2 yields and productivities. We suggest that an improved H2-producing cell factory could be designed by constructing a well characterized and overproduced synthetic H2 pathway and fully activating the native FHL in BL21(DE3).