A cell-free microtiter plate screen for improved [FeFe] hydrogenases

Abstract

Background: [FeFe] hydrogenase enzymes catalyze the production and dissociation of H(2), a potential renewable fuel. Attempts to exploit these catalysts in engineered systems have been hindered by the biotechnologically inconvenient properties of the natural enzymes, including their extreme oxygen sensitivity. Directed evolution has been used to improve the characteristics of a range of natural catalysts, but has been largely unsuccessful for [FeFe] hydrogenases because of a lack of convenient screening platforms.

Methodology/principal findings: Here we describe an in vitro screening technology for oxygen-tolerant and highly active [FeFe] hydrogenases. Despite the complexity of the protocol, we demonstrate a level of reproducibility that allows moderately improved mutants to be isolated. We have used the platform to identify a mutant of the Chlamydomonas reinhardtii [FeFe] hydrogenase HydA1 with a specific activity approximately 4 times that of the wild-type enzyme.

Conclusions/significance: Our results demonstrate the feasibility of using the screen presented here for large-scale efforts to identify improved biocatalysts for energy applications. The system is based on our ability to activate these complex enzymes in E. coli cell extracts, which allows unhindered access to the protein maturation and assay environment.

Figure 1. A schematic representation of the screening procedure.

Figure 2. Oxygen deactivation curve for the product of CFPS reactions expressing C. reinhardtii HydA1.

5 µL of CFPS product were diluted with 25 µL of anaerobic Tris buffer before addition of the volume of air-saturated Tris buffer indicated on the x-axis. The slope of methyl viologen absorbance at 578 nm with time for each oxygen-exposed sample was normalized to the slope with no oxygen exposure. The pink curve represents an exponential decay fit to the data. Error bars indicate the standard deviation of triplicate measurements.

Figure 3. Assessment of the well-to-well error of the screen in three sets of identically performed 10 µL CFPS reactions.

Residual activity following oxygen exposure is represented by the average pre-exposure/post-exposure activity ratio. Error bars indicate the standard deviation for n = 88, 72, and 71 wells respectively. The coefficient of variance is given above each bar. 1 and 5 µL of CFPS product were assayed in the pre-exposure and post-exposure measurements, respectively, and the post-exposure activity was normalized to 1 µL. 15 µL of O₂-equilibrated buffer were used for deactivation. C. reinhardtii HydA1 was expressed in Experiments 1 and 2, and C. pasteurianum CpI was expressed in Experiment 3. The difference in average residual activity between Experiments 1 and 2 is likely due to differences in the extent of oxygen diffusion out of the aerobic buffer before addition to the hydrogenase solutions, and highlights the need for wild-type controls on each plate.

Figure 4. Specific activities of wild-type and mutant C. reinhardtii HydA1.

Specific activity was calculated for each using cell-free protein synthesis products in the hydrogen consumption direction (blue bars) and the hydrogen production direction (red bars). Error bars indicate the standard deviation of at least n = 4 independent experiments.