Engineering the Reaction Pathway of a Non-heme Iron Oxygenase Using Ancestral Sequence Reconstruction

Abstract

Non-heme iron (Fe^II), α-ketoglutarate (α-KG)-dependent oxygenases are a family of enzymes that catalyze an array of transformations that cascade forward after the formation of radical intermediates. Achieving control over the reaction pathway is highly valuable and a necessary step toward broadening the applications of these biocatalysts. Numerous approaches have been used to engineer the reaction pathway of Fe^II/α-KG-dependent enzymes, including site-directed mutagenesis, DNA shuffling, and site-saturation mutagenesis, among others. Herein, we showcase a novel ancestral sequence reconstruction (ASR)-guided strategy in which evolutionary information is used to pinpoint the residues critical for controlling different reaction pathways. Following this, a combinatorial site-directed mutagenesis approach was used to quickly evaluate the importance of each residue. These results were validated using a DNA shuffling strategy and through quantum mechanical/molecular mechanical (QM/MM) simulations. Using this approach, we identified a set of active site residues together with a key hydrogen bond between the substrate and an active site residue, which are crucial for dictating the dominant reaction pathway. Ultimately, we successfully converted both extant and ancestral enzymes that perform benzylic hydroxylation into variants that can catalyze an oxidative ring-expansion reaction, showcasing the potential of utilizing ASR to accelerate the reaction pathway engineering within enzyme families that share common structural and mechanistic features.

Figure 1.

Reaction pathway engineering of non-heme iron (Fe^II) and α-ketoglutarate (α-KG) dependent oxygenases. (a) Previous examples of engineering Fe^II/α-KG enzymes to perform non-native reactivity, including site-directed mutagenesis based on comparation of protein crystal structures, site-saturation mutagenesis, and DNA shuffling. In this work, we used an ancestral sequence reconstruction informed approach to engineer an Fe^II/α-KG hydroxylase to perform oxidative ring-expansion. (b) General catalytic mechanisms of Fe^II/α-KG enzymes for hydroxylation and skeletal rearrangement. The substrate radical generated from H–atom abstraction by Fe(IV)-oxo species can either rebound with Fe(III)–OH affording a hydroxylated product or undergo a radical rearrangement to form the skeletal rearranged product.

Figure 2.

Identification of closely related Fe^II/α-KG enzymes that display different reactivity. (a) Benzylic hydroxylation of substrate 1 and oxidative ring-expansion of substrate 3 catalyzed by a library of Fe^II/α-KG enzymes following the conditions: 50 mM TES buffer (pH = 7.5), 1 mM substrate, 5 mM α-KG, 4 mM sodium ascorbate, 0.2 mM FeSO₄, and 50% clarified cell lysate (approximately 2.5 μM final enzyme concentration). (b) Sequence similarity network of selected Fe^II/α-KG enzymes (alignment score: 76). (c) Reactivity heat map of benzylic hydroxylation and oxidative ring-expansion catalyzed by the library of Fe^II/α-KG enzymes, in both cases the positive control (relative 100%) gives >99% conversion of the substrate. BHO (highlighted in gray box) displayed a preference for benzylic hydroxylation reactivity, even though it shares higher sequence similarity with TropC and XenC which natively catalyzed ring-expansion reaction.

Figure 3.

Phylogenetic analysis and ancestral sequence reconstruction of BHO related Fe^II/α-KG enzymes. (a) Phylogenetic tree of the Fe^II/α-KG enzymes with a zoom-in focus on BHO and XenC clade. BHO shares common ancestors with enzymes that prefer oxidative ring-expansion instead of benzylic hydroxylation. (For the whole list of sequences that are used for phylogenetic analysis see Figure S6) Three reconstructed and resurrected ancestral Fe^II/α-KG enzymes are labeled as dots, and their relationship with extant proteins BHO and XenC is shown. Anc1 being the common ancestor, Anc2 being the ancestral protein at the branching point, and Anc3 on the evolution trajectory from Anc2 to BHO. (b) Reactivity preference of ancestral Fe^II/α-KG enzymes between benzylic hydroxylation and oxidative ring-expansion tested under the following conditions: 50 mM TES buffer (pH = 7.5), 1 mM substrate, 5 mM α-KG, 4 mM sodium ascorbate, 0.2 mM FeSO₄, 20 μM purified protein. The benzylic hydroxylation reaction was analyzed by UPLC-DAD, monitoring substrate consumption. The relative ring-expansion reactivity was measured using the colorimetric assay. In both cases, BHO and XenC give >99% conversion of the corresponding substrate. Both common ancestors, Anc1 and Anc2, favor oxidative ring-expansion, while Anc3 favors benzylic hydroxylation.

Figure 4.

Engineering BHO for improved oxidative ring-expansion reactivity. (a) Residues within 10 Å (distance between the Cα of the residue and the Fe center) of the active site of BHO identified based on the sequence alignment of ancestral and extant Fe^II/α-KG enzymes. Residues labeled in gray are not 100% conserved in enzymes that favor the same reaction pathway. (b) Relative oxidative ring-expansion reactivity of BHO variants compared to XenC. Reaction was performed under the following condition: 50 mM TES buffer (pH = 7.5), 1 mM substrate, 5 mM α-KG, 4 mM sodium ascorbate, 0.2 mM FeSO₄, 40 μM purified protein. The relative oxidative ring-expansion reactivity was determined by the colorimetric assay. The positive control XenC (relative 100%) gives >99% yield of stipitaldehyde. (c) Homology model of BHO generated using AlphaFold2, with the key residues identified based on the combinatorial site-directed mutagenesis library highlighted.

Figure 5.

Improve the oxidative ring-expansion activity of Anc3 using combinatorial site-directed mutagenesis. (a) Detection of the shunt product when analyzing the reaction of Anc3 combinatorial site-directed mutagenesis library using UPLC-DAD. (b) Screening results of Anc3 S215G MSDM library following the conditions: 50 mM TES buffer (pH = 7.5), 1 mM substrate, 5 mM α-KG, 4 mM sodium ascorbate, 0.2 μM FeSO₄, 50% clarified cell lysate (approximately 8 μM final enzyme concentration). Reactions were analyzed by UPLC-DAD and the concentrations of the products were quantified based on the standard curve. The ratios between the concentration of the shunt product (5) and stipitaldehyde (4) was used to determine the preference for benzylic hydroxylation and oxidative ring-expansion. (c) Reactivity and product distribution of representative Anc3 variants following the conditions: 50 mM TES buffer (pH = 7.5), 1 mM substrate, 5 mM α-KG, 4 mM sodium ascorbate, 0.2 μM FeSO₄, 40 μM purified enzyme. Reactions were analyzed by UPLC-DAD and the concentration of the products were quantified based on the standard curve.

Figure 6.

DNA shuffling identifies critical mutations in dictating the ring-expansion reaction. (a) Construction of DNA shuffle library was carried out by first digesting the DNA that encodes Anc2 and Anc3 with DNase I and then reassembling by PCR. (b) Percent conservation of each position of the sequence as the Anc2 residue based on multiple sequence alignment of 28 chimeric proteins that displayed 1.5-fold improvement in ring-expansion reactivity as compared to wild type Anc3. Residues that are identical between those of Anc2 and Anc3 were not shown. (c) Critical residues were identified through shuffle library screening. (d) Reactivity and product distribution of Anc3 variants with newly identified mutations following the conditions: 50 mM TES buffer (pH = 7.5), 1 mM substrate, 5 mM α-KG, 4 mM sodium ascorbate, 0.2 μM FeSO₄, and 40 μM purified enzyme. Chimera is a chimeric protein that carries a total of 39 mutations including mutations at all the positions that are above 85% conservation.

Figure 7.

QM/MM simulation of the reaction pathways. (a) Quantum chemical reaction mechanism for radical ring-expansion and hydroxylation pathways derived from species 7. (b) Relationship between experimental product ratios and ΔE_a (kcal/mol) from QM/MM calculations. Ring-expansion product% = (ring-expansion) × 100%/(ring-expansion + hydroxylation). See Tables S13, S14, S17–S26, and Figures S68 and S73–S88 for energies, key metrical parameters, and enzyme models.

Figure 8.

Effects of R(N(Arg191)H–O) hydrogen bonding on the reaction selectivity. (a) 3D model of the representative species 7 showing the key hydrogen bonds: Arg191 and the hydroxyl ligand (R; blue dashed line). (b) Relationship between ΔE_a (kcal/mol) and R(N(Arg191)H–O) distances (Å). Enzyme shown in gray cartoon, substrate radical in yellow licorice, succinate and key amino acid residues in gray licorice, hydroxyl and water ligands in red licorice, and Fe(III) atom in orange vdW. See Tables S13, S17, and S18 and Figures S73–S76 for energies, key metrical parameters, and enzyme models.