Changes in Active Site Loop Conformation Relate to the Transition toward a Novel Enzymatic Activity

Abstract

Enzymatic promiscuity, the ability of enzymes to catalyze multiple, distinct chemical reactions, has been well documented and is hypothesized to be a major driver of the emergence of new enzymatic functions. Yet, the molecular mechanisms involved in the transition from one activity to another remain debated and elusive. Here, we evaluated the redesign of the active site binding cleft of lactonase SsoPox using structure-based design and combinatorial libraries. We created variants with largely improved catalytic abilities against phosphotriesters, the best ones being >1000-fold better compared to the wild-type enzyme. The observed shifts in activity specificity are large, and some variants completely lost their initial activity. The selected combinations of mutations have considerably reshaped the active site cavity via side chain changes but mostly through large rearrangements of the active site loops and changes to their conformations, as revealed by a suite of crystal structures. This suggests that a specific active site loop configuration is critical to the lactonase activity. Interestingly, analysis of high-resolution structures hints at the potential role of conformational sampling and its directionality in defining the enzyme activity profile.

Figure 1

Structure-based design and directed evolution of SsoPox. (1) In silico analyses were performed to compare SsoPox with bacterial PTE from Brevundimonas diminuta. (2) 13 residues were targeted for designing a combinatorial library using a degenerated oligonucleotide-based strategy. (3) A solid-based screening was adapted to identify variants retaining a paraoxonase activity. (4) A miniaturized procedure was used to screen a 430-variant library against 9 substrates. (5) Most relevant variants were produced, purified, and characterized for further characterization.

Figure 2

Substrates used in this study. Ethyl-paraoxon (1), ethyl-parathion (2), tabun analogue (3), soman analogue (4), VX analogue (5), sarin analogue (6), cyclosarin analogue (7), pNP-acetate (8), 3-oxo-C10 AHL (9), and 3-oxo-C12 AHL (10).

Figure 3

Enzymatic promiscuity and stability profile of SsoPox variants. The heat map shows the selected variants’ catalytic efficiency values (circle size) and fold change (fill colors) compared to SsoPox-wt, in logarithmic values. Gray fill indicates no calculated ratio relative to the wild-type due to a lack of measured wild-type activity toward that substrate. Each column represents the data for a different SsoPox variant, while each row shows the results for a different substrate. Variants are organized by the number of combined mutations (top row). The second table row indicates the variant names and their melting temperature (T_m) values (specific values are given in Table S5). The heat map was designed using the ggplot2 package on R.

Figure 4

Protection of rHAChE with SsoPox variants. rHAChE activity was evaluated after incubation of paraoxon with SsoPox wild-type and variants (IIIC1, IVB10, IG7, IVE10, and IVA4) at different times. The resulting curves were fitted with a one-phase decay equation using GraphPad Prism 6 software.

Figure 5

Improved variants show different active site loops 7 and 8 conformations. Active site regions of the variant structures are superimposed with the SsoPox-wt structure (light gray; PDB ID 2VC5). The active site location is indicated by the presence of the bimetallic catalytic centers shown as spheres. The SsoPox-wt structure and its loops 7 and 8 are highlighted. Significant loop conformation changes are indicated with arrows (blue for loop 8; orange for loop 7). Bottom graph: Normalized thermal B-factor values of the different structures are a function of the residue number. A zoomed-in inset for the loops 7 and 8 sequence region is shown. We note that the structures of variants IG7, IVA4, and IVB10 are modeled with a metal-bound phosphate anion, while IIIC1, IVE2, and SsoPox-wt are unbound. All structures have a complete bimetallic center.

Figure 6

Mutations contribute to active site loop conformational changes. The variant structures (in green) are superposed onto the structure of SsoPox-wt (in cyan). Mutated residues are highlighted in orange. Interactions are shown as dashed lines, and interactions of interest are shown as purple dashed lines.

Figure 7

Active site loop conformation and anisotropic thermal B-factor ellipsoids. Arrows are an approximate representation of a major direction and the magnitude of the active site loops’ anisotropy. Metal cations and the bound phosphate anion are shown as spheres. The visualization of the ellipsoids was performed by using PyMOL.