Dynamically Interacting Protein Networks Provide a Mechanism to Overcome the Enormous Intrinsic Barrier to Orotidine 5'-Monophosphate Decarboxylation

Abstract

Orotidine 5'-monophosphate decarboxylase (OMPDC) is among the most efficient enzymes known, accelerating the decarboxylation of the OMP by ∼17 orders of magnitude, primarily by lowering the enthalpy of activation by ∼28 kcal/mol. Despite this feature, OMPDC from Methanothermobacter thermautotrophicus requires ∼15 kcal/mol of activation energy following ES complex formation. This study applies temperature-dependent hydrogen-deuterium exchange mass spectrometry (TDHDX) to detect site-specific thermal protein networks that channel energy from solvent collisions to the active site. Comparative TDHDX of native OMPDC and a single-site variant (Leu123Ala) that alters the activation enthalpy for catalytic turnover reveals region-specific changes in protein flexibility, connecting local scaffold unfolding enthalpy to the activation barrier of catalysis. The data implicate four spatially resolved, thermally sensitive networks that originate at distinct protein-solvent interfaces and converge near the substrate phosphate-binding region (R203), the ribose-binding region (K42), and a catalytic loop (S127). These networks are proposed to act synergistically to optimize substrate positioning and active site electrostatics for the activated complex formation. The complexity of the identified thermal activation pathways distinguishes Mt-OMPDC from other TIM barrel enzymes previously studied by TDHDX. The findings highlight the essential role of scaffold dynamics in enzyme function with broad implications for designing efficient biocatalysts.

1

Structural and energetic insights into OMP decarboxylation catalyzed by OMP decarboxylases. (a) Free energy (ΔG ^‡) and activation enthalpy (ΔH ^‡) for the reaction for the OMP decarboxylation, showing the energy differences between the reaction in water and in the presence of enzyme. Data for the reaction in water were derived from OMP analogue decarboxylation in aqueous solution, followed by averaging, while enzyme data were based on the averaged reaction energetics of Escherichia coli OMP decarboxylase (Ec-OMPDC), Saccharomyces cerevisiae OMP decarboxylase (Sc-OMPDC), and Methanothermobacter thermautotrophicus OMP decarboxylase (Mt-OMPDC) under k _cat conditions. (b) Crystal structure overlay of Mt-OMPDC in its open (apo) form (PDB: 3G18) and ligand-bound (6-azaUMP) closed form (PDB: 3G1A). Residue-specific RMSD values were calculated to identify structural changes upon ligand binding and visualized on the crystal structure using the PyMOL script ColorbyRMSD, with a white-gray-deep teal color gradient. Regions showing minimal structural differences are depicted in white, while regions with increasing RMSD values are colored deep teal. Loop7 (residues 180–188), highlighted by a circle, is unstructured in the apo form and adopts a well-ordered closed conformation over the ligand in the bound form. Due to its unstructured nature in the apo form, RMSD values for this loop could not be calculated, and it was manually colored deep teal for visual clarity. (c) Crystal structure of the Mt-OMPDC homodimer bound to the tight binding analogue 6-azaUMP (PDB: 3G1A), where 6-azaUMP is shown in yellow stick representation. The active site of Mt-OMPDC is highlighted, featuring the catalytic tetrad (K42, D70, K72, and D75′) along with R203, shown in dark gray with labeled interaction distances between the substrate and these residues. Hydrophobic residues from the β-strands (I16, I68, I96, L123, V155, I178, I200), which project into the barrel, are depicted in aqua stick representation but are not labeled. These hydrophobic residues form van der Waals contacts with neighboring hydrophobic residues. (d) Mechanism of OMP decarboxylation through the formation of a vinyl carbanion. (e) Linear relationship between the Cornette hydrophobicity scale of the residue at position 123 of Mt-OMPDC and the activation energy (E _a(k _cat)) for the reaction (R ² = 0.85), excluding the L123G mutant, which is represented as an inverse triangle and shown for comparison only.

2

Single-temperature HDX analysis at 35 °C to assess the impact of mutation and ligand binding on thermal stability of Mt-OMPDC. Panels (a) and (b) show the effects of the L123A mutation in the apo form, with panel (a) representing normalized percentage change in D-uptake, ΔD(%), after 10 min of HDX and panel (b) after 120 min. The ΔD(%), calculated using eq (see the text), quantifies differences in D-uptake for each peptide due to the mutation. These changes were mapped onto the Mt-OMPDC crystal structure (PDB: 3G1A) using a blue-white-red gradient, where blue and red indicate decreased and increased D-uptake, respectively, in the mutant (L123A) compared to the wild type (WT). Only changes greater than 0.5 Da and 3σ, were considered, and peptides showing significant changes are marked with arrows only on the A monomer to avoid overcrowding. Panels (c) and (d) display the effects of ligand binding on WT and L123A after 120 min of HDX. These ΔD(%) changes were similarly mapped using the blue-white-red gradient, with blue indicating decreased D-uptake (enhanced structuring) and red indicating increased D-uptake (reduced structuring) in the ligand-bound form compared to the apo form.

3

Temperature-dependent HDX-MS analysis of WT­(L) and the E _a (k _cat ) impairing mutant L123A­(L) variants of Mt-OMPDC. Panels (a) and (b): D-uptake as a function of time and temperature for peptide 37–44 (representing ribose-binding motif) in the TSA-bound forms of native WT­(L) and mutant L123A­(L), respectively. (c) Arrhenius plot of ln­(k _HDX) vs 1000/T for peptide 37–44. Panels (d) and (e): D-uptake as a function of time and temperature for peptide 94–110 (representing the dimer interface) in the TSA-bound forms of WT­(L) and L123A­(L), respectively. (f) Arrhenius plot of ln­(k _HDX) vs 1000/T for peptide 94–110. (g) Bar graph representing the HDX activation energy, E _a(k _HDX), values of peptides in WT­(L) and L123A­(L). Error analysis for panels (c), (f), and (g) was performed by calculating activation energies (E _a(k _HDX)) from Arrhenius plots (ln­(k _HDX) vs 1000/T) across seven temperatures using data from biological replicates; standard deviations from linear regression fits were used to assess statistical significance. (h) Bar graph representing ΔE _a(k _HDX), calculated as E _a(k _HDX) for (L123A­(L)) minus E _a(k _HDX) for WT­(L)). Only peptides showing ΔE _a(k _HDX) > 2σ, are shown, where σ is the standard propagated error from replicate-based activation energy difference. (i) Cartoon representation of the Mt-OMPDC homodimer, highlighting peptides with significant ΔE _a(k _HDX). Regions with changes in ΔE _a(k _HDX) are shown using a red-white-blue gradient, with blue indicating positive ΔE _a(k _HDX) and increased rigidity in mutant, and red indicating negative ΔE _a(k _HDX) and increased flexibility in L123A­(L) relative to WT­(L). Peptides with significant changes in ΔE _a(k _HDX) are marked with arrows. (j) The space-filled model highlights regions with ΔE _a(k _HDX) changes using a red-white-blue gradient, revealing dynamic thermal networks (TN) connecting solvent-exposed regions to the active site. Thermal Network 1 (TN-1) corresponds to the phosphate-binding loop. Thermal Network 2 (TN-2) represents the sugar-binding region. Thermal Network 3 (TN-3) links TN-1 and TN-2 to facilitate synergistic interactions. Thermal Network 4 (TN-4) represents the catalytic loop. It is worth noting that TN-1 is present in both monomers but only shown in one monomer (TN-1 in A) as is located on the backside of the surface of monomer B.

4

Contributions from intra- and intersubunit thermal networks drive thermal activation of catalysis and barrier crossing in Mt-OMPDC. (a) Intrasubunit interactions between thermal-networks: Peptide 10–17 (TN-3) bridges TN-1 and TN-2 through H-bonding, enabling synergistic activation. Hydrogen bonding between R14 and L225 connects the N- and C-termini. (b) Intersubunit interactions, F134–I142′ (linking TN-4 and TN-4′) and Y45–Y45′ (connecting TN-2 and TN-2′), may facilitate coordinated action of thermal networks between subunits, alongside intrasubunit connectivity of TN-2 and TN-4. The color code in (a, b) is the same as described in Figure for changes in ΔE _a(k _HDX) (red to blue goes from negative to positive). (c) The active site of OMPDC was modeled using the 6-azaUMP bound crystal structure (PDB: 3G1A). To represent the substrate-bound active site, OMP was positioned in place of 6-azaUMP into this structure. In this initial configuration, where the carboxylate group is in the plane of the pyrimidine group, it sterically clashes with catalytic residue K72 (0.9 Å) and undergoes a very close proximity to D70. (d) Modeling of the bound OMP in OMPDC to avoid steric repulsion with K72. The C6–C7 bond was rotated while keeping other residue positions fixed, with the criterion that the side chains K42, D70, K72, and D75′ remain > 2 Å apart. This distorts the carboxylate out of the pyrimidine ring plane into a hydrophobic patch, while retaining a close (H-bonding) distance for D70. (e) Cartoon representation highlighting the spatial distribution of loop elements (purple) within the identified thermal networks in Mt-OMPDC. Surface-exposed loops of the C-terminus, TN-3 (N-terminus), and TN-2 are interconnected. Loops associated with TN-1 are solvent-exposed but connect to TN-2 through internal interactions rather than surface interactions.

5

Thermal energy network comparison among TIM barrel enzymes. The final TDHDX-derived energy transfer pathways in murine adenosine deaminase (m-ADA), yeast enolase (Sc-ENO), and human catechol-O-methyltransferase (Hs-COMT) each consist of two distinct networks. In contrast, for archaeal OMP decarboxylase (Mt-OMPDC), which catalyzes one of biology’s most challenging reactions, four synergistic thermal energy networks have been identified. In every instance, these networks converge at the reactive bond of the substrate within the respective active site. The thermal energy network figure for m-ADA adapted from ref . Available under a CC-BY 4.0 license. Copyright 2022 Gao et al. The thermal energy network figure for Hs-COMT adapted from ref . Copyright 2020 PNAS. The thermal energy network figure for Sc-ENO adapted from ref . Copyright 2021 American Chemical Society. The network for Mt-OMPDC is the result of this work.