Hydrogenase enzymes rapidly catalyze the reversible reduction of protons with electrons to produce hydrogen. These enzymes have the potential to be used in photosynthetic hydrogen production or fuel cell applications, where energy can be captured from sunlight, and the reactants and byproducts are water and oxygen. One main obstacle in industrial hydrogenase use is the oxygen sensitivity of the protein. Here we investigate the oxygen inactivation of an [FeFe] hydrogenase from Clostridium pasteurianum and describe the evolution of a mutant with decreased oxygen sensitivity.
We show that, in contrast to previous observations, [FeFe] hydrogenase I from Clostridium pasteurianum (CpI) can be partially reactivated after oxygen exposure. We observe slow reactivation of CpI over time in a 2% hydrogen environment in the presence of methyl viologen, and rapid reactivation of the hydrogenase in the presence of a strong reducing agent such as dithionite. After reactivation, the hydrogenase activity is partially but not fully restored. This observation suggests that there are two states of inactivation of CpI by oxygen: one reversible and one irreversible. We further show that this reactivation does not occur for the [FeFe] hydrogenase HydA1 from Chlamydomonas reinhardtii, which suggests that the accessory [Fe-S] clusters present in CpI but not HydA1 may be responsible for this reversible inactivation.
Additionally, we have used random mutagenesis and a cell-free protein synthesis platform to produce and screen a library of CpI hydrogenase mutants for improved oxygen tolerance, and have identified a mutant that is inactivated more slowly in the presence of oxygen than the wild-type enzyme. The initially isolated mutant had 13 mutations, and three of these mutations contribute to the decreased oxygen sensitivity of the protein. Saturation mutagenesis at one of these sites has resulted in further improvement of the mutant. We show, however, that both the mutant and wild-type enzyme are reactivated to the same extent in the presence of an electron source, and the mutant is therefore not immediately useful for biohydrogen applications. This demonstrates the need to develop a screen that measures hydrogenase activity while the enzyme is working in the presence of oxygen in order to evolve an oxygen-tolerant hydrogenase for biohydrogen applications.