Rigidifying a De Novo Enzyme Increases Activity and Induces a Negative Activation Heat Capacity

Sarah A Hindson,Dimitri A Svistunenko,Marc W Van Der Kamp,Christopher Williams,J L Ross Anderson,H Adrian Bunzel,Adrian J Mulholland,Christopher R Pudney,Bettina Frank

doi:10.1021/acscatal.1c01776

Sarah A Hindson, Dimitri A Svistunenko + Show 7 more

Open Access

https://doi.org/10.1021/acscatal.1c01776

Copy DOI

Abstract

Conformational sampling profoundly impacts the overall activity and temperature dependence of enzymes. Peroxidases have emerged as versatile platforms for high-value biocatalysis owing to their broad palette of potential biotransformations. Here, we explore the role of conformational sampling in mediating activity in the de novo peroxidase C45. We demonstrate that 2,2,2-triflouoroethanol (TFE) affects the equilibrium of enzyme conformational states, tending toward a more globally rigid structure. This is correlated with increases in both stability and activity. Notably, these effects are concomitant with the emergence of curvature in the temperature-activity profile, trading off activity gains at ambient temperature with losses at high temperatures. We apply macromolecular rate theory (MMRT) to understand enzyme temperature dependence data. These data point to an increase in protein rigidity associated with a difference in the distribution of protein dynamics between the ground and transition states. We compare the thermodynamics of the de novo enzyme activity to those of a natural peroxidase, horseradish peroxidase. We find that the native enzyme resembles the rigidified de novo enzyme in terms of the thermodynamics of enzyme catalysis and the putative distribution of protein dynamics between the ground and transition states. The addition of TFE apparently causes C45 to behave more like the natural enzyme. Our data suggest robust, generic strategies for improving biocatalytic activity by manipulating protein rigidity; for functional de novo protein catalysts in particular, this can provide more enzyme-like catalysts without further rational engineering, computational redesign, or directed evolution.

Highlights

Conformational sampling profoundly impacts the overall activity and temperature dependence of enzymes
C45 activity approaches that of natural benchmarks such as horseradish peroxidase (HRP) but is limited by the lack of a catalytic proton-shuttling residue, which likely slows down both the initial H2O2 deprotonation and the subsequent cleavage of the peroxide O−O bond; this results in an elevated KM for H2O2.7 The relative simplicity of these de novo designed helical bundles allows facile tailoring for specific purposes, e.g., by introducing site-specific functional moieties.[8−11] they are often highly flexible, facilitating access to a broad equilibrium of conformational states for engineering.[12]
Evidence from a range of experimental and computational studies points to the involvement of networks of protein dynamics that extend throughout the protein and are not just localized to the immediate active site.[19−21] macromolecular rate theory (MMRT) quantifies the difference in the distribution of vibrational modes between the ground and transition state ensemble as the heat capacity of catalysis (ΔCP‡) given by ln k = ln kBT − [ ΔHT‡0 + ΔCP‡(T − T0) ]

Summary

Introduction

Conformational sampling profoundly impacts the overall activity and temperature dependence of enzymes. We apply macromolecular rate theory (MMRT) to understand enzyme temperature dependence data These data point to an increase in protein rigidity associated with a difference in the distribution of protein dynamics between the ground and transition states. Heme peroxidases are versatile biocatalysts that exploit their heme cofactor to form highly oxidative oxyferryl intermediates (Figure 1B).[6] We have recently created the de novo hemeperoxidase C45 (Figure 1A), a rationally designed four-helical bundle protein with high peroxidase activity.[7] Notably, C45 activity approaches that of natural benchmarks such as horseradish peroxidase (HRP) but is limited by the lack of a catalytic proton-shuttling residue, which likely slows down both the initial H2O2 deprotonation and the subsequent cleavage of the peroxide O−O bond; this results in an elevated KM for H2O2.7 The relative simplicity of these de novo designed helical bundles allows facile tailoring for specific purposes, e.g., by introducing site-specific functional moieties.[8−11] they are often highly flexible, facilitating access to a broad equilibrium of conformational states for engineering.[12] In contrast to many other heme-dependent enzymes, C45 catalyzes a broad range of transformations such as dehalogenations and carbene-transfer reactions.[13]. Where such curvature is present, values of ΔCP‡ are typically negative for natural enzymes, manifesting as concave curvature in plots of ln k versus T

Methods

Results

Conclusion