Technical Comments

Comment on “Extreme electric fields power catalysis in the active site of ketosteroid isomerase”

See allHide authors and affiliations

Science  28 Aug 2015:
Vol. 349, Issue 6251, pp. 936
DOI: 10.1126/science.aab0095

Abstract

Fried et al. (Reports, 19 December 2014, p. 1510) demonstrate electric field–dependent acceleration of biological catalysis using ketosteroid isomerase as a prototypic example. These findings were not extended to aqueous solution because water by itself has field fluctuations that are too large and fast to provide a catalytic effect. Given physiological context, when water electrostatic interactions are considered, electric fields play a less important role in the catalysis.

Fried et al. (1) report that the isomerization of 5-androstene-3,17-dione by ketosteroid isomerase (KSI) has a markedly higher (4 × 107 M) rate constant (kcat) than in aqueous solution (kAcO). The free-energy barrier for the reaction catalyzed by KSI is much lower because (i) the difference in effective acetate (COO) concentration (ΔGS) is higher for KSI than for kAcO; (ii) different electric-field contributions on the C=O bond (ΔGC=O); and (iii) hydrogen abstraction by the carboxyl group (ΔGH), as shown in Fig. 1. Thus,

ΔGS + ΔGC=O + ΔGH = –RTln(4 × 107) = –10.5 kcal mol−1 (1)
Fig. 1 Isomerization of 5-androstene-3,17-dione and the effects of water on free-energy barriers.

The chemical mechanism for the first step of the isomerization reaction catalyzed by KSI (A) and the uncatalyzed reaction in aqueous solution (B). The two reactions differ in (i) electrostatic interactions of the C=O group (squares) and (ii) the carboxyl group abstracting the α-hydrogen (circles). (C) Ground and transition states for the isomerization reaction in which the electric field exerted on the C=O bond is zero (nonpolar environment), whereas the carboxyl group abstracting the α-hydrogen is similar to (A). (D) Ground and transition states for the isomerization reaction in which the electric field exerted on the C=O bond is the similar to that of aqueous solution in (B), whereas the carboxyl group abstracting the α-hydrogen is similar to (A). (E and F) Water accelerates reactions in which the atoms are more charged in transition than ground states (5). (G) Desolvation accelerates reactions in which the atoms are less charged in transition than ground states (5, 6).

Fried et al. also report that the free-energy barrier for the rate-limiting enolate transition catalyzed by KSI is 11.5 kcal mol−1, compared with 18.8 kcal mol−1 in a nonpolar environment where the electric field is 0. On this premise, KSI contributes an electric field of 7.3 kcal mol−1 to its free-energy-barrier reduction when compared with the nonpolar environment. This model assumes that bulk water confers no electric power toward catalysis because it has field fluctuations that are too wide and fast compared with the narrow infrared shifts that are evident in KSI and its active-site mutants. Thus, ΔGC=O is –7.3 kcal mol−1, ΔGS is approximately –3.2 kcal mol−1, and ΔGH is 0, because the model does not take into account the effect of hydrogen abstraction on the free-energy barrier.

Life on earth depends on water and the hydrogen bonds that it forms in biological systems. Water can accelerate reactions by more than 1010-fold (>13.8 kcal mol−1 of free-energy-barrier reduction) when atoms in transition states become more charged than in ground states (Fig. 1). Such hydrogen bonds potentiate catalysis in a number of ways that are also directly relevant to the isomerization of 5-androstene-3,17-dione by KSI mutants that contain active-site cavities sufficiently large for water to interact with C=O, as suggested by Kraut et al. (2). For example, because water forms electrostatic interactions with the negatively charged oxygen atom of the C=O group in 5-androstene-3,17-dione, these forces stabilize transition states as the oxygen atom is more negatively charged and reduce the free-energy barrier of the isomerization reaction by ~4.0 kcal mol−1.

The broad line width of C=O in 19-nortestosterone measured in aqueous solution indicates that electrostatic interactions between water and C=O adopt diverse conformations. A thermodynamic cycle shows to what extent broad electric fields of water can reduce the free-energy barrier for the isomerization reaction of 5-androstene-3,17-dione (Fig. 2). ΔGrig represents the free-energy-barrier reduction by well-oriented active-site–associated water, which is expected to contribute a larger C=O spectral shift than in aqueous solution. Thus, ΔGrig is less than –4 kcal mol−1, based on the infrared spectra for C=O in nonpolar solvent, water, and KSI. A more accurate calculation of ΔGrig (–4.8 kcal mol−1) is derived from the analysis of free-energy barriers during the isomerization of 5-androstene-3,17-dione by the KSI Tyr16Ser mutant, where water can interact with C=O (13.6 kcal mol−1), versus Tyr16Phe (16.0 kcal mol−1), where water interactions are absent (1, 2). This is supported by the observation that the infrared spectral shift of 19-nortestosterone bound to the Tyr16Ser mutation is narrow (1). From this, ΔGsol is the free-energy-barrier reduction by water C=0 interactions and can be obtained by

ΔGsol = –4.8 + ΔG1 – ΔG2 (2)
Fig. 2 Contribution of active-site electric fields and desolvation of Asp40 to KSI’s catalytic power.

(A) Thermodynamic cycle showing the electrostatic contributions of the C=O group to the free-energy-barrier reduction (ΔGsol) for the isomerization of 5-androstene-3,17-dione in water solution. (B) Relative electrostatic contributions of the C=O group to free-energy-barrier reduction in a nonpolar environment, in water and KSI. The free-energy-barrier reduction in KSI is 3.3 kcal mol−1 more than that in aqueous solution. This represents the contribution of electric fields to the catalytic power of KSI because the catalytic power of KSI is estimated based on the uncatalyzed reaction in aqueous solution. (C) Crystal structure of KSI mutant Asp40Asn complexed with androsten-3-betal-ol-17-one (Protein Data Bank, 1E3R) (3). The nitrogen atom of Asn40 is changed to an oxygen atom. Red spheres represent oxygen atoms of the carboxyl group. Hydrogen atoms that affect solvent-accessible areas of the oxygen atoms are shown. The oxygen atom abstracting α-hydrogen is surrounded by hydrophobic groups. The other oxygen atom forms a hydrogen bond with Trp120. (D) The relative contribution of KSI’s catalytic power. The electric field contributes ~32%, which is markedly less than the combined base positioning and desolvation of Asp40, which contribute approximately 68%.

Both ΔG1 and ΔG2 are greater than 0 at room temperature, just as ΔG for the transition from water to ice is greater than 0. Because exact values for ΔG1 and ΔG2 are not available, we estimated ΔG1 by assuming that water-C=O and water-water interactions are comparable. ΔG1 is the maximum free energy of reorganization for the reference reaction and is close to the free-energy change of the process in which one hydrogen bond of a water molecule is fixed and the water molecule can still rotate freely around the hydrogen bond. At 273.2 K, ΔG, ΔH, and ΔS from water to ice are 0, –1.44 kcal mol−1, and –5.26 cal mol−1 K−1, respectively. As one mole of water molecules in ice contains two moles of hydrogen bonds, and constrained water in the cycle can rotate in one direction, ΔH1 and ΔS1 are –0.72 kcal mol−1 and –4.38 cal mol K−1. ΔG1 = 2 × [–0.72 – 298 × (–4.38)/1000] = 1.17 kcal mol−1. Thus, the free-energy-barrier reduction by water is ΔG2 + 3.63 kcal mol−1. Because ΔG2 is similar to ΔG1, ΔGsol is closer to –4.0 kcal mol−1, which is estimated based on the C=O spectral shift in aqueous solution. Electrostatic interactions between the C=O group and water therefore contribute ~4.0 kcal mol−1 to the free-energy-barrier reduction for the reference reaction in water. Thus, the electrostatic contribution of the C=O group to free-energy-barrier reduction for the reaction catalyzed by KSI versus the reaction in water is 3.3 kcal mol−1 (Fig. 2).

The origin of ΔGH is desolvation of active-site Asp40. In aqueous solution, the COO group is completely solvated and at least one water molecule is removed for COO to abstract the α-hydrogen atom, whereas in KSI, no water molecules are removed for Asp40 to abstract the hydrogen atom (Fig. 1). The crystal structure of Pseudomonas putida KSI mutant Asp40Asn complexed with androsten-3-betal-ol-17-one (3) shows that the oxygen atom involved in abstracting the α-hydrogen is desolvated and surrounded by hydrophobic groups (Fig. 2), indicating that the desolvation process does not occur from the ground to transition state catalyzed by KSI. Rather, desolvation of Asp40 occurs during substrate binding to KSI, supported by the observation that analog binding affinity increases by ~2 orders of magnitude for the Asp40Ala mutation compared with wild-type (4). Moreover, electrostatic interactions with the COO group decrease as the reaction proceeds because the oxygen atoms of COO become less negatively charged, thereby increasing the free-energy barrier. This free-energy-barrier increment for the reference reaction—in which the COO group interacts with three water molecules (Fig. 1)—is larger than that for the reaction in KSI, in which the COO group interacts with Trp120 (Fig. 2). Thus, desolvation of Asp40 can reduce the free-energy barrier to a meaningful extent. Similar cases exist in organic reactions in which the desolvation of anions can accelerate reactions dramatically (Fig. 1).

In summary, the relative free-energy contribution of KSI’s catalytic power includes electric-field and desolvation effects, as well as general base positioning. The contribution of the electric-field effect compared to that in water is 3.3 kcal mol−1 and accounts for ~31.4% of KSI’s catalytic speed-up relative to the uncatalyzed reference reaction in aqueous solution. However, it is worthwhile emphasizing that the overall energy gain of the reaction is distinct from that contributed by electrostatic interactions of C=O. Regardless of the overall gain for the reaction in aqueous versus hydrophobic environments, electrostatic interaction of C=O with water for the reference reaction contributes toward the free-energy-barrier reduction. Exact contributions of desolvation (ΔGH) and general base positioning (ΔGS) to free-energy-barrier reduction cannot be calculated based on the information provided in this paper.

References and Notes

  1. ACKNOWLEDGMENTS: This work was supported by grants from the National Science Foundation of China (21473041) and AI100914, DK096323, P30 DK56338, and Clinical and Translational Science Award UL1TR000071 from the National Institutes of Health.
View Abstract

Navigate This Article