A Proficient Enzyme Revisited: The Predicted Mechanism for Orotidine Monophosphate Decarboxylase

See allHide authors and affiliations

Science  09 May 1997:
Vol. 276, Issue 5314, pp. 942-945
DOI: 10.1126/science.276.5314.942


A mechanism is proposed to explain the activity of orotidine 5′-monophosphate decarboxylase (ODCase). This enzyme is the one of the most proficient known, with a catalytic proficiency (k cat/K m)/k non= 1023 M−1. Quantum mechanical calculations predict a mechanism involving a stabilized carbene intermediate, which represents a previously unrecognized mode of enzymatic activity for ODCase. The proposed mechanism involves proton transfer from a weak acid (pK a = 7, where K ais the acid constant) concerted with decarboxylation, in a nonpolar enzyme environment. Such a mechanism makes possible different approaches to the design of ODCase inhibitors. Furthermore, the prediction that general acid catalysis may only be effective in low dielectric media is of general significance for understanding the activity of many enzymes.

The decarboxylation of orotidine 5′-monophosphate (OMP) (Fig. 1, 1a; all bold numbers here refer to labeled structures in Fig. 1) to form uridine 5′-monophosphate (UMP, 2a) by orotidine 5′-monophosphate decarboxylase (ODCase) is an essential step in nucleic acid biosynthesis (1). In bacteria and fungi, ODCase exists alone as a separate protein, whereas in mammals the ODCase activity is part of a bifunctional protein, UMP synthase (2). The conversion of 1 to 2 is also biomechanistically unique, in that all other biochemical decarboxylations involve stabilization of the carbanion intermediate by delocalization of the electron pair into a π orbital (3, 4). For the decarboxylation of1, however, no such stabilization is present; a nonconjugated carbanion is formed. ODCase is one of the most proficient enzymes known (1). Proficiency is the second-order rate constant for the enzymatic reaction divided by the nonenzymatic rate constant [ (kcat/ Km/ knon)]. For ODCase, the proficiencey is 2.0 × 1023M−1. The catalytic efficiency (kcat/ KM) of ODCase, as with many enzymes, is nearly diffusion-controlled (1); it is the excruciating slowness of the uncatalyzed decarboxylation (knon = 2.8 × 10−16s−1) that produces the high proficiency.

Figure 1

(A) The decarboxylation of OMP. (B) Inhibitors of ODCase. (C) Proposed intermediate zwitterion (5) and model compound (6). (D) Zwitterion resulting from 2-protonation of orotate (7) and zwitterion resulting from 4-protonation of orotate (8). (E) Known stabilized carbenes: thiazolium ylide of thiamine (9) and stabilized carbene (10). (F) 4-protonated uracil (11), proposed model transition state (12), and anionic carbene (13) resulting from decarboxylation of orotate.

The mechanism by which ODCase effects the efficient decarboxylation has long been a subject of interest among chemists and biochemists but remains unknown (1, 5). The lack of any detectable cofactors or metal ions makes this mechanistic problem even more intriguing (6-9). Although no crystal structure of ODCase exists (10, 11), it has been shown that 6-azauridylate (3a) and barbituric acid ribonucleotide (4a) are particularly prodigious inhibitors of yeast ODCase [inhibition constant (KI) = 5.1 × 10−7and 8.8 × 10−12 M, respectively] (5, 8). Studies by Jones et al. of the enzyme mechanism indicate that Lys93 is important for catalysis but not for binding (12). In addition, the ratio of the maximal catalytic rate Vmax to KM of ODCase is pH-dependent, with a maximum at pH 7; this result has been interpreted to indicate that the enzyme possesses a catalytic group that has a p Ka of ∼7 (7, 12). Recently, catalytic antibodies have been developed to catalyze the decarboxylation of orotate (9, 13).

Various mechanistic hypotheses have been proposed to explain the enormous rate enhancement by ODCase. Silverman et al. suggested a covalent mechanism involving nucleophilic attack at C-5, but this was subsequently shown by 13C and D isotope effects to be unlikely (6, 7, 14). Beak et al. examined the decarboxylation of 1,3-dimethylorotic acid in sulfolane and proposed that decarboxylation occurred via zwitterion (5) (5). The fast (4 × 108 acceleration over the parent reaction) decarboxylation of betaine (6) led the authors to propose that the enzyme might effect catalysis by favoring formation of zwitterion (5). This mechanism has long been widely accepted (6-9, 15). The protonation of the weakly basic orotate to form the zwitterion, however, seems unlikely.

An enzyme as proficient as ODCase is expected to be unusually “sensitive to … reversible inhibitors [that are] designed to resemble the [transition structure]” (1). The biological and medicinal importance of this reaction is clear: As a key biosynthetic step, the decarboxylation is a natural target for antitumor agents and genetic diseases such as orotic aciduria. Knowledge of the transition structure facilitates inhibitor design; herein, calculations are used to predict a transition state for this biologically and medicinally important reaction.

The energetics of the decarboxylation were explored by full optimizations of geometries and reaction paths along the decarboxylation reaction coordinate with the use of restricted Hartree-Fock (RHF) calculations with the 6-31+G* basis set. Energetics were obtained by single-point second-order Møller-Plasset (MP2) energy calculations on the RHF geometries (16). The parent reaction of 1b is endothermic in the gas phase (activation enthalpy, ΔH = +43.9 kcal mol−1), and there is no barrier to recombination of CO2 with the carbanion (17). We next examined the decarboxylation of the 2-protonated orotate 5b, which was proposed by Beak to be an intermediate for the catalyzed reaction. The decarboxylation of 2-protonated orotate is endothermic by only +21.6 kcal mol−1 at the MP2/6-31+G* level. These gas-phase results are in accord with the Beak mechanism: decarboxylation of 2-protonated orotate 5b is 22 kcal mol−1 less endothermic than the decarboxylation of orotate itself (1b).

Product 7, formed from decarboxylation of 2-protonated orotate, can be represented by several resonance structures; one of these is a carbene with separated charges (Fig. 1D). For better stability, we envisioned a neutral carbene that did not have separated charges (18). Such a structure (8, Fig. 1D) is obtainable by protonation of orotate at the 4-oxygen and decarboxylation. Indeed, the reaction to form 8 is a very favorable process: The decarboxylation of 4-protonated orotate has an extraordinarily low barrier—only +15.5 kcal mol−1.

The ΔH values calculated for these three reactions (decarboxylation of orotate, 2-protonated orotate, and 4-protonated orotate) are shown in Fig. 2. The entropy change associated with decarboxylation is between 23 and 27 entropy units for all the reactions at 298 K. The ΔH for the decarboxylation of orotate at 298 K is calculated to be 44.5 kcal mol−1, and the free energy of activation ΔG is 36.4 kcal mol−1. These gas-phase values are not sensitive to solvation and can be compared favorably to the experimental values of 44.4 and 38.5, respectively (1). Protonation at the 2-position accelerates the reaction, as predicted by Beak (ΔG = 14.0 kcal mol−1, ΔH = 21.6 kcal mol−1), but protonation at the 4-position drops the barrier for decarboxylation by an additional 6 kcal mol−1 (ΔG = 8.0 kcal mol−1, ΔH = 15.5 kcal mol−1).

Figure 2

Calculated ΔH values (MP2/6-31+G*//RHF/6-31+G*) for the decarboxylation of orotate, 2-protonated orotate, and 4-protonated orotate.

Density functional methods verify the energetics of orotate and 4-protonated orotate decarboxylation. With Becke3LYP/6-31+G* calculations (16), the ΔG at 298 K for orotate decarboxylation is 36.7 kcal mol−1, which is consistent with our MP2/6-31+G*//RHF/6-31+G* and experimental values. The 4-protonated orotate decarboxylates with a ΔG of 9.0 kcal mol−1, which is also consistent with our MP2 values. Thus, quantum mechanical methods lead to a prediction that protonation at O-4 lowers the activation free energy by 28 kcal mol−1, leading to a stabilized carbene intermediate.

Carbenes are usually considered to be reactive intermediates of only fleeting existence; nevertheless, the carbene 8 is likely to be the intermediate responsible for ODCase’s extraordinary proficiency. The possibility of a stabilized carbene such as8 has biological precedence in the form of thiamine pyrophosphate. The thiazolium ylide (9) of thiamine is formed quite easily (p Ka ∼ 19 to 20) and is also nucleophilic; such carbenes gain unusual stability because of ylide resonance (19, 20). Stabilized carbenes (10) have also been successfully synthesized, crystallized, and characterized by Arduengo and co-workers, who have shown that some derivatives are “true carbenes with negligible ylidic character” (21).

Another fundamental mechanistic question involves the nature of the proton transfer to O-4. Does protonation occur first, followed by decarboxylation, or is proton transfer concerted with decarboxylation? The predicted gas-phase acidities of each position of orotate and deprotonated uracil are shown in Fig. 3. Two primary points can be drawn from these values: First, the 4-oxygen position is always more basic than the 2 position. Experimentally, uracil is known to be protonated at the 4-oxygen, not the 2-oxygen (22). Second, decarboxylation increases the basicity of the oxygen by a substantial amount (by 28 kcal mol−1 for the 4 position). In the gas phase, a viable base such as MeNH3 + (proton affinity = 214 kcal mol−1) (23) easily and exothermically transfers a proton to both species shown in Fig. 3. However, in water, protonation of either reactant or product is unlikely; the p Ka of 4-protonated uracil (11) is about 0.5 (22) and that of orotate should be only slightly higher (∼1.5) (24). The p Ka of the catalytic group in ODCase has been experimentally measured at 7 (7, 12). The p Ka of lysine in aqueous solution is 10.5 (25); presumably, the dielectric of the relatively nonpolar cavity and neighboring side chains in the enzyme active site drops this value to 7. The p Ka of protonated orotate might be raised somewhat above ∼1.5 in the nonpolar enzyme active site, but protonation of the starting material orotate is still highly unlikely to occur by proton transfer from a catalytic group with p Ka = 7. Indeed, equilibrium protonation would occur at the carboxylate oxygen, a process that would shut down decarboxylation entirely. How can the enzyme manage to deliver a proton to the 4-oxygen? Proton transfer concerted with decarboxylation can take advantage of the large increase in basicity during the departure of CO2. This form of general acid catalysis demonstrates Jencks’ “libido rule,” which postulates that “concerted general acid-base catalysis of complex reactions … can occur only (a) at sites that undergo a large change of p K in the course of the reaction, and (b) when …  the p K of the catalyst is intermediate between the initial and final p K values of the substrate site” (26).

Figure 3

Calculated gas-phase proton affinities (in kilocalories per mole) of orotate (left) and deprotonated uracil (right) (MP2/6-31+G*//RHF/6-31+G*).

This mechanism requires a particular environment, which must be provided by the enzyme active site. For example, concerted proton transfer to O-4 will not occur in an aqueous environment, because lysine is insufficiently acidic. However, our computational investigations show that in an appropriate environment that is likely to be present in the enzyme, the proton transfer to the 4-oxygen concerted with decarboxylation becomes very favorable.

The magnitudes of dielectric constants of enzyme active sites have been the subject of extensive discussions, with predicted values ranging from 2 to 40 (27-29). Recent theoretical work predicts that protein interiors possess low dielectric constants in the range of 3 to 6 (30). We explored the effect of placing the reacting species within solvent cavities of four different dielectric constants; the calculated values of ΔH for the decarboxylation of bare orotate (uncatalyzed model) and the reaction of orotate with methyl ammonium to form carbene (12) plus CO2 (catalyzed model) in these various cavities are summarized in Table 1 (31). Calculations on the uncatalyzed process show that the barrier to orotate decarboxylation is not very sensitive to solvent, as would be expected for a reaction with anionic starting material and product. However, the decarboxylation concerted with proton transfer from the ammonium group varies greatly in exothermicity depending on the environment. In the gas phase, the reaction is exothermic by a huge amount (61.9 kcal mol−1). In a medium of dielectric constant (ɛ) equal to 4, the ΔH jumps to +17.6 kcal mol−1. At a value of ɛ = 10, the ΔH rises to +33.7 kcal mol−1 as the increasingly polar solvent selectively stabilizes the ionic starting materials relative to the less polar products. In water (ɛ = 78.54) the reaction is significantly endothermic, at +43.2 kcal mol−1. The data in the third row of Table 1 provide the activation energy change upon 4-protonation in four different dielectric environments. In water, there is no advantage to 4-protonation at all. At the other extreme, the 4-protonation would be favored by an enormous amount in the gas phase (105.8 kcal mol−1). Our calculations therefore predict that only a low dielectric constant allows energetics that are ideal for catalysis. Using an ɛ = 4 to mimic the enzyme’s hydrophobic interior (30), we find that the difference in activation energy calculated for the parent reaction in water and our proposed catalytic mechanism in dielectric 4 is 24.8 kcal mol−1, essentially identical to that found experimentally (Fig.4). Even assuming some differences associated with entropy changes and binding interactions, the calculated value still compares favorably to the experimental value. Thus, we predict that the mechanism by which OMP decarboxylates in the ODCase active site is via a stabilized carbene intermediate that results from concerted proton transfer to the 4 position; this proton transfer, though yielding a carbene, results in the rate acceleration of 1017. To complete the catalytic cycle, we envision the protonation of the carbene in Fig. 4 by water, which would allow for formation of bicarbonate and for proton transfer from protonated uracil back to the catalytic lysine.

Figure 4

Energy diagram showing barrier to decarboxylation for orotate in water and for orotate plus CH3NH3 + in a dielectric of 4, which is our working model for the ODCase catalyzed reaction.

Table 1

Energies of reaction of orotate (1b) plus CH3NH3 + to form a carbene-methylamine complex (12) plus CO2(enzyme-catalyzed reaction model) in cavities with various dielectric constants (ɛ). The energies are calculated at MP2/6-31+G*//RHF/6-31+G*. The energies of the uncatalyzed process (1b13 + CO2) and the differences are given for comparison.

View this table:

The potent inhibition of ODCase by barbiturate 5′-monophosphate (BMP,4a) can be explained by its similarity to the decarboxylation transition state and the fact that it should be protonated by the catalytic acid of ODCase. The p Ka of BMP in water is only about 4.5 (8), but in the enzyme active site, by analogy to our previous discussion, we expect the p Ka to rise to about 8. Therefore, proton transfer from the catalytic group to BMP is favorable and viable. Furthermore, our calculations show that anionic carbene (13) and barbiturate (4b) have similar proton affinities in the gas phase [302 kcal mol−1 for carbene (13) versus 304 kcal mol−1 for barbiturate (4b)]. The perfect transition state analog would match the shape and electrostatic potential of the carbene-like transition state and have a p Ka greater than 7 for protonation in the area occupied by the 4-oxygen.

The mechanism proposed here requires an environment of low dielectric (similar in polarity to ether or chloroform) and will not be observed in aqueous solution. Therefore, the active site of ODCase must provide two features: a low dielectric interior to allow for p Ka tuning so that the catalytic proton transfer to O-4 can occur, and a perfectly placed lysine to effect that proton transfer to O-4, concerted with decarboxylation.


View Abstract

Navigate This Article