Proton Transfer from C-6 of Uridine 5′-Monophosphate Catalyzed by Orotidine 5′-Monophosphate Decarboxylase: Formation and Stability of a Vinyl Carbanion Intermediate and the Effect of a 5-Fluoro Substituent

Abstract

The exchange for deuterium of the C-6 protons of uridine 5′-monophosphate (UMP) and 5-fluorouridine 5′-monophosphate (F-UMP) catalyzed by yeast orotidine 5′-monophosphate decarboxylase (ScOMPDC) at pD 6.5 – 9.3 and 25 °C was monitored by ¹H NMR spectroscopy. Deuterium exchange proceeds by proton transfer from C-6 of the bound nucleotide to the deprotonated side chain of Lys-93 to give the enzyme-bound vinyl carbanion. The pD-rate profiles for k_cat give turnover numbers for deuterium exchange into enzyme-bound UMP and F-UMP of 1.2 × 10⁻⁵ and 0.041 s⁻¹, respectively, so that the 5-fluoro substituent results in a 3400-fold increase in the first-order rate constant for deuterium exchange. The binding of UMP and F-UMP to ScOMPDC results in 0.5 and 1.4 unit decreases, respectively, in the pK_a of the side chain of the catalytic base Lys-93, showing that these nucleotides bind preferentially to the deprotonated enzyme. We also report the first carbon acid pK_as for proton transfer from C-6 of uridine (pK_CH = 28.8) and 5-fluorouridine (pK_CH = 25.1) in aqueous solution. The stabilizing effects of the 5-fluoro substituent on C-6 carbanion formation in solution (5 kcal/mol) and at ScOMPDC (6 kcal/mol) are similar. The binding of UMP and F-UMP to ScOMPDC results in a greater than 5 × 10⁹-fold increase in the equilibrium constant for proton transfer from C-6 so that ScOMPDC stabilizes the bound vinyl carbanions, relative to the bound nucleotides, by at least 13 kcal/mol. The pD-rate profile for k_cat/K_m for deuterium exchange into F-UMP gives the intrinsic second-order rate constant for exchange catalyzed by the deprotonated enzyme as 2300 M⁻¹ s⁻¹. This was used to calculate a total rate acceleration for ScOMPDC-catalyzed deuterium exchange of 3 × 10¹⁰ M⁻¹, which corresponds to a transition state stabilization for deuterium exchange of 14 kcal/mol. We conclude that a large portion of the total transition state stabilization for the decarboxylation of orotidine 5′-monophosphate can be accounted for by stabilization of the enzyme-bound vinyl carbanion intermediate of the stepwise reaction.

INTRODUCTION

Orotidine 5′-monophosphate decarboxylase (OMPDC) is an essential enzyme that catalyzes the decarboxylation of orotidine 5′-monophosphate (OMP) to give uridine 5′-monophosphate (UMP), a critical step in the de novo biosynthesis of pyrimidine nucleotides (Scheme 1). In mammals, OMPDC is found as part of the bifunctional uridine 5′-monophosphate synthase complex, but in many lower organisms it exists as a discrete protein.¹ The enzyme is an obligate dimer and residues from both subunits contribute to the active site architecture.^2,3 Remarkably, OMPDC requires no metal ions or other cofactors but yet it effects an enormous 10¹⁷-fold acceleration of the decarboxylation of enzyme-bound OMP.^3,4 A significant fraction of the total transition state stabilization for decarboxylation of ca. 31 kcal/mol³ results from utilization of the 12 kcal/mol intrinsic binding energy of the 5′-phosphodianion group of OMP.^5–8

Scheme 1.

The first X-ray crystal structures for OMPDC appeared in 2000 and, notably, they revealed the absence of suitably placed catalytic residues required for several of the previously proposed reaction mechanisms.^9–12 This precipitated a renewal of the scientific debate about the mechanism and the origin of the enzymatic rate acceleration for decarboxylation.^13–19 The simplest possibility is the direct decarboxylation of OMP by a stepwise mechanism through the unstable UMP vinyl carbanion intermediate that undergoes subsequent proton transfer from the enzyme to generate the product (Scheme 1). We have argued that this mechanism is mandated by the observation of product isotope effects of unity for the decarboxylation of OMP and 5-fluoroorotidine 5′-monophosphate (F-OMP) in a mixture of 50:50 (v/v) H₂O/D₂O, which show that there is no discrimination between H and D at the proton donor in the transition state for the product-determining step (Scheme 2).^20,21 These results eliminated a concerted reaction mechanism for decarboxylation in which proton transfer from the enzyme provides electrophilic push to the loss of CO₂.^9,13

Scheme 2.

We reported earlier that OMPDC from Saccharomyces cerevisiae (ScOMPDC) catalyzes the exchange for deuterium from solvent D₂O of the C-6 proton of UMP to give UMP labeled with deuterium at C-6 (d-UMP, Scheme 2).²² These relatively slow reactions precluded extensive investigation of the kinetics and mechanism of OMPDC-catalyzed deuterium exchange into UMP. We reasoned that the addition of a 5-fluoro substituent to give 5-fluorouridine 5′-monophosphate (F-UMP) should result in stabilization of the negative charge at C-6 of the corresponding vinyl carbanion.^23,24 Therefore, study of OMPDC-catalyzed deuterium exchange into F-UMP should be much more feasible than for UMP, and it would provide insight into the proton transfer reaction catalyzed by OMPDC. Our initial report of the OMPDC-catalyzed C-6 deuterium exchange reaction of F-UMP employed wildtype and the D70N and D70G mutants of OMPDC from Methanothermobacter thermautotrophicus.²⁵

We present here a full report of our studies of the ScOMPDC-catalyzed C-6 deuterium exchange reaction of UMP,²² along with a complete set of data for the pD-dependence of the corresponding ScOMPDC-catalyzed C-6 deuterium exchange reaction of the more reactive substrate F-UMP. The results are consistent with a deuterium exchange reaction that is the formal reverse of the proton transfer “half reaction” that occurs during the decarboxylation of OMP, and they provide convincing support for our conclusion that the decarboxylation of OMP proceeds by a stepwise mechanism through the UMP vinyl carbanion intermediate (Scheme 2).^20,21 We also report the first determination of reliable C-6 carbon acid pK_as for a series of substituted uracils in aqueous solution, including the novel C-6 carbon acidities of the nucleosides uridine and 5-fluorouridine. A comparison of the data for the enzyme-catalyzed deuterium exchange reactions of UMP and F-UMP with those for proton transfer from uridine and 5-fluorouridine in aqueous solution shows that ScOMPDC stabilizes the bound UMP and F-UMP vinyl carbanions, relative to the bound nucleotides, by at least 13 kcal/mol. We conclude that a large portion of the rate acceleration for the decarboxylation of OMP can be accounted for by stabilization of the enzyme-bound vinyl carbanion intermediate of the stepwise reaction.

EXPERIMENTAL SECTION

Materials

Orotidine 5′-monophosphate was purchased from Sigma or was prepared by chemical or enzymatic methods using modifications of literature procedures.^24,26,27 Uridine, uridine 5′-monophosphate disodium salt (99%, from yeast), 5-fluorouridine and 2-(N-morpholino)ethanesulfonic acid sodium salt (MES) were purchased from Sigma. 5-Fluoro-1,3-dimethyluracil, phosphorus(V) oxychloride and imidazole were purchased from Aldrich. Trimethyl phosphate, glycylglycine (> 99%) and 3-(N-morpholino)propanesulfonic acid (MOPS, ≥ 99.5%) were purchased from Fluka. 6-Azauridine 5′-monophosphate (free acid) was purchased from R. I. Chemical. Deuterium oxide (99.9% D), sodium deuteroxide (30% in D₂O, 99.5% D) and deuterium chloride (35% w/w, 99.9% D) were purchased from Cambridge Isotope Laboratories. Bovine serum albumin (BSA) was from Roche and was dialyzed against 0.1 M NaCl in D₂O. Trimethyl phosphate was dried over phosphorus pentoxide and distilled under reduced pressure. Imidazole was recrystallized from benzene. Water was from a Milli-Q Academic purification system. All other chemicals were reagent grade or better and were used without further purification.

5-Fluorouridine 5′-Monophosphate

5-Fluorouridine 5′-monophosphate was prepared as its triethylammonium salt from 5-fluorouridine using modifications of a literature procedure.²⁸ Phosphorus(V) oxychloride (0.13 mL, 1.4 mmol) and 5-fluorouridine (0.20 g, 0.76 mmol) were dissolved in dry trimethyl phosphate (2 mL) at 0 °C. After 16 h the mixture was poured into cold water (100 mL), the solution was adjusted to pH 7 using 1 M NaOH, and water was added to give a total volume of 180 mL. The mixture was applied to a column of DEAE Sephadex A25 equilibrated with 50 mM triethylammonium bicarbonate and an elution gradient of 50 – 600 mM triethylammonium bicarbonate was applied. Fractions containing the desired product were pooled and the buffer was removed by lyophilization to give a 50% yield of 5-fluorouridine 5′-monophosphate as its triethylammonium salt. This material was shown to be free of inorganic phosphate by ³¹P NMR. ¹H NMR (500 MHz, D₂O, pD ≈ 7) δ7.83 (1H, ³J_HF = 6.0 Hz, C6-H), 5.87 (1H, ³J_HH = 5.0 Hz, ⁵J_HF = 1.5 Hz, C1′-H), 4.21 (1H, m, C2′-H), 4.19 (1H, m, C3′-H), 4.08 (1H, m, C4′-H), 3.84 (2H, m, C5′-H). Chemical shifts are reported relative to HOD at 4.67 ppm.

Preparation of Solutions

Solution pH or pD was determined at 25 °C using an Orion Model 720A pH meter equipped with a Radiometer pHC4006-9 combination electrode that was standardized at 25 °C. Values of pD were obtained by adding 0.40 to the observed reading of the pH meter.²⁹

The acidic protons of glycylglycine were exchanged for deuterium by dissolution in D₂O followed by evaporation and drying under vacuum at 55 °C. Buffered solutions of glycylglycine or MOPS in D₂O were prepared by dissolving the acidic form and NaCl (if needed) in D₂O followed by the addition of a measured amount of NaOD to give the desired acid/base ratio at I = 0.1. Buffered solutions of imidazole or MES in D₂O were prepared by dissolving the basic form and NaCl (if needed) in D₂O followed by the addition of a measured amount of DCl to give the desired acid/base ratio at I = 0.1. Solutions in D₂O were stored in a dessicator.

The concentration of OMP in stock solutions was determined from its absorbance in 0.1 M HCl at 267 nm using ε = 9430 M⁻¹ cm⁻¹.³⁰ Stock UMP (100 mM) was prepared by dissolving the disodium salt in D₂O to give a solution at pD 8.2 (I = 0.3) and its concentration was determined from the absorbance in 0.1 M HCl at 262 nm using ε = 10,000 M⁻¹ cm⁻¹.³¹ Stock solutions of F-UMP were prepared by dissolving the triethylammonium salt in D₂O to give a solution at pD ≈ 8 and its concentration was determined from the absorbance in 0.1 M HCl at 270 nm using ε = 9160 M⁻¹ cm⁻¹ (see below). Stock 6-azauridine 5′-monophosphate (6-aza-UMP, 12 mM) was prepared by dissolving the free acid in D₂O and the pD was adjusted to 7.9 by the addition of 4.3 M NaOD. Solutions of OMP, UMP, F-UMP and 6-aza-UMP were stored at −20 °C.

Extinction Coefficient of 5-Fluorouridine 5′-Monophosphate

A stock solution of F-UMP was prepared in D₂O (pD ≈ 8) and was diluted 10-fold with 60 mM imidazole buffer (50% free base) in D₂O at pD 7.6 and I = 0.1 (NaCl) to give a final concentration of 54 mM imidazole. The solution was analyzed by ¹H NMR spectroscopy and the integrated areas of the signals due to the protons of F-UMP were compared with those due to the C-4 and C-5 protons of imidazole to give the concentration of F-UMP in the stock solution in D₂O as 102 mM. The stock solution of F-UMP in D₂O (102 mM) was then used to determine λ_max = 270 nm and an extinction coefficient of ε = 9160 M⁻¹ cm⁻¹ for F-UMP in 0.1 M HCl.

pK_a of the N-3 Hydron of 5-Fluorouridine 5′-Monophosphate in H₂O and D₂O (pK_NL)

The acidity of the N-3 hydron of F-UMP in H₂O and D₂O was determined spectrophotometrically by monitoring the decrease in absorbance at 269 nm, A_obsd, that accompanies ionization at N-3. The absorbance of 163 μM F-UMP was determined as a function of pL in H₂O or D₂O at 25 °C and I = 0.1 (NaCl) using the following buffers: 0.1 M HCl, 0.01 M LCl, acetate (pH 4.7), MES (pD 6.4), imidazole (pD 7.0), MOPS (pL 7.1 – 7.4), glycylglycine (pL 7.6 – 9.4), glycine (pL 9.7 – 10.3), 0.01 M NaOL. The data were fit to eq 1, where A_acid and A_base are the absorbances at the acidic and basic extremes, to give the acidity constants K_NL for ionization of F-UMP at N-3 in H₂O and D₂O.

$$A_{\text{obsd}}=\frac{A_{\text{acid}}{10}^{-\text{pL}}+A_{\text{base}}{K}_{\text{NL}}}{{10}^{-\text{pL}}+K_{\text{NL}}}$$ (1)

Prepation of Yeast OMPDC in D₂O

OMPDC from Saccharomyces cerevisiae (ScOMPDC) was prepared as described previously.^32,33 The protein sequence differs from the published sequence for true wildtype yeast OMPDC by the following mutations: S2H,³⁴ C155S,³⁵ A160S and N267D.³⁴ Except for the C155S mutation, the sequence is the same as that observed in the published crystal structure of wildtype yeast OMPDC.¹¹ This C155S variant is more stable than, but kinetically and structurally essentially identical with, the Cys-155 enzyme.³⁵ In this work, all experiments were conducted using this C155S variant.

Samples of overexpressed and purified ScOMPDC that had been stored at −80 °C were defrosted and extensively dialyzed at 4 °C against 10 mM MOPS (50% free base, pH 7.1) containing 100 mM NaCl in order to remove glycerol. This was followed by dialysis against several changes of buffer in D₂O using a D-tube dialyzer (6 – 8 kDa MWCO, Novagen) placed inside a narrow vessel that was isolated from atmospheric moisture using parafilm. The buffers used for dialysis were: 62.5 or 125 mM glycylglycine, 80% free base (pD 9.3); 125 mM glycylglycine, 20% free base (pD 8.1); 100 mM MOPS, 50% free base or 60 mM imidazole, 50% free base (pD 7.6); 100 mM MOPS, 35% free base (pD 7.4); 60 or 125 mM imidazole, 20% free base (pD 7.0); 100 mM MES, 40% free base (pD 6.5). All buffers were in D₂O at I = 0.1 (NaCl). The concentration of ScOMPDC in the stock solutions in D₂O, determined by standard assay as described below, was 0.3 – 0.5 mM (9 – 15 mg/mL).

Enzyme Assays

The activity of ScOMPDC in the enzyme stock solutions in D₂O and in the deuterium exchange reaction mixtures was determined by monitoring the decrease in absorbance at 279 nm accompanying the enzyme-catalyzed decarboxylation of OMP (Δε = −2400 M⁻¹ cm⁻¹ at 25 °C).^2,7,22,32 Prior to assay, the stock solutions of ScOMPDC were diluted with 10 mM MOPS, 50% free base (pH 7.1) containing 100 mM NaCl and 0.4 mg/mL BSA to give a final concentration of ca. 20 μM ScOMPDC. Assays in a total volume of 1 mL were conducted in 10 mM MOPS, 50% free base (pH 7.1), at 25 °C and I = 0.105 (NaCl), with 40 – 50 μM OMP (25 – 30K_m). The reaction was initiated by the addition of 1 SL of the diluted stock solution of ScOMPDC or an aliquot (up to 100 μL) of the deuterium exchange reaction mixture to give a final enzyme concentration of 20 – 40 nM, and the initial velocity of the ensuing decarboxylation of OMP was determined within 1 min. The concentration of ScOMPDC in the stock solution or in the deuterium exchange reaction mixture was then calculated from the observed value of V_max (M s⁻¹) using the relationship V_max = k_cat[E], with k_cat = 15 s⁻¹.^2,7

NMR Analyses

¹H NMR spectra at 500 MHz (8 – 128 transients) were acquired at 25 °C using a Varian Unity Inova 500 spectrometer with a sweep width of 6000 Hz, a 90° pulse angle, an acquisition time of 6 s and a total relaxation delay between pulses of 90 s (≥ 10T₁) for the reactions of UMP or 30 s (≥ 9T₁) for the reactions of F-UMP. Chemical shifts were referenced to HOD at 4.67 ppm. Values of T₁ = 9 s and 7 s were determined for the C-5 and C-1′ protons, respectively, of [6-²H]-uridine 5′-monophosphate (d-UMP) in glycylglycine buffer at pD 9.4. Values of T₁ = 1.5 s and 3.3 s were determined for the C-6 and C-1′ protons, respectively, of [6-¹H]-5-fluorouridine 5′-monophosphate (h-F-UMP) in imidazole buffer at pD 7.6. Before determination of the integrated peak areas, the signals of interest were greatly expanded, accurately phased and subjected to a first-order drift correction.

Deuterium Exchange at C-6 of F-UMP Monitored by ¹H NMR Spectroscopy

The C-6 deuterium exchange reactions of F-UMP catalyzed by ScOMPDC in D₂O at pD 6.5 – 9.3 were followed directly by ¹H NMR spectroscopy at 500 MHz. Reactions in a volume of 1 mL were initiated by the addition of an aliquot of F-UMP in D₂O to a mixture of the appropriate buffer, NaCl, BSA and ScOMPDC in D₂O to give final concentrations of 0.1 – 5 mM F-UMP, 0.4 mg/mL (0.04%) BSA (see below), and 0.2 – 40 μM ScOMPDC at I = 0.1 (NaCl). 700 μL of the reaction mixture was transferred to an NMR tube and the progress of deuterium exchange was monitored by ¹H NMR spectroscopy at 25 °C for 2 – 8 h. The remaining portion of the reaction mixture was incubated at 25 °C and the activity of ScOMPDC was monitored by periodic standard assay as described above. After completion of the deuterium exchange reaction the enzyme was removed by ultrafiltration and the pD of the filtrate was recorded. There was no change in pD of the reaction mixture during these reactions at 25 °C which were followed for up to 8 h.

In all cases the concentration of enzyme used was chosen to accommodate the time needed to obtain at least 4 timepoints, with acquisition of an appropriate number of NMR transients, in one half-time for the deuterium exchange reaction. This required that lower concentrations of enzyme be used for the reactions at low concentrations of F-UMP for which more transients per timepoint were required for reasons of sensitivity. For the reactions of F-UMP in the presence of low concentrations of ScOMPDC (≥ 2 μM) in the absence of BSA there was up to a 25% decrease in the observed activity of ScOMPDC during the deuterium exchange reaction. However, the inclusion of 0.4 mg/mL (0.04%) BSA stabilized this loss of enzyme activity with time, which we attribute mainly to protein adsorption. There were also small decreases (up to 10%) in enzyme activity after extended reaction times in the presence of low concentrations of F-UMP (≤ K_m) which we attribute to the instability of the dimeric form of unliganded ScOMPDC.² A control experiment showed that the incubation of 10 μM ScOMPDC for 24 h at pD 9.4 (100 mM glycylglycine) and 25 °C resulted in a 15% loss in enzyme activity in the presence of 0.3 mM F-UMP (1.5K_m) but no loss in enzyme activity in the presence of 2 mM F-UMP (10K_m).

Deuterium Exchange at C-6 of UMP Monitored by ¹H NMR Spectroscopy

The C-6 deuterium exchange reactions of UMP catalyzed by ScOMPDC in D₂O at pD 7.0 – 9.3 were followed by ¹H NMR spectroscopy at 500 MHz. Reactions in a volume of 1 – 4 mL were initiated by the addition of an aliquot of UMP in D₂O to a mixture of the appropriate buffer, NaCl and ScOMPDC in D₂O to give final concentrations of 2.5 – 10 mM UMP and 100 – 300 μM ScOMPDC at I = 0.1 (NaCl). The reaction mixture was divided into 1 mL portions that were placed in screw-cap centrifuge tubes equipped with O-rings and incubated in a water bath at 25 °C. At various times a tube was withdrawn and lightly centrifuged to sediment any particulate matter. An aliquot of the reaction mixture was removed and the enzyme-catalyzed deuterium exchange reaction was stopped by the addition of a 2 to 3-fold excess over enzyme of the tight-binding competitive inhibitor 6-aza-UMP (12 mM stock, pD 7.9). The enzyme-inhibitor complex, which was formed quantitatively, was removed by ultrafiltration using an Amicon Microcon device (10K MWCO) and the filtrate was transferred to an NMR tube. The extent of deuterium exchange was then determined by ¹H NMR spectroscopy. After NMR analysis the sample was removed from the NMR tube and its pD was determined. The activity of ScOMPDC in the reaction mixture was monitored by periodic standard assay as described above. There was no significant loss of enzyme activity or change in pD of the reaction mixture during these reactions at 25 °C which were followed for up to 12 days.

RESULTS

Deuterium Exchange at C-6 of 5-Fluorouridine 5′-Monophosphate Catalyzed by Yeast OMPDC

The exchange for deuterium of the C-6 proton of [6-¹H]-5-fluorouridine 5′-monophosphate (h-F-UMP) ([S]_o = 0.1 – 5 mM) to give [6-²H]-5-fluorouridine 5′-monophosphate (d-F-UMP) catalyzed by OMPDC from Saccharomyces cerevisiae (ScOMPDC, 0.2 – 40 μM) in D₂O at pD 6.5 – 9.3, 25 °C and I = 0.1 (NaCl) was followed directly by monitoring the disappearance of the C-6 proton by ¹H NMR spectroscopy (Scheme 3).

Scheme 3.

Figure 1 shows partial ¹H NMR spectra at 500 MHz in the regions of the C-6 proton (left) and C-1′ (anomeric) proton (right) obtained at various times during the exchange for deuterium of the C-6 proton of h-F-UMP ([S]_o = 2.5 mM) catalyzed by ScOMPDC (9.2 μM) in D₂O at pD 9.3 (100 mM glycylglycine, 80% free base), 25 °C and I = 0.1 (NaCl). The spectra on the left of Figure 1 show that deuterium exchange at C-6 results in the clean disappearance of the doublet at 7.874 ppm (³J_HF = 6.2 Hz) due to the C-6 proton. By contrast, after 43% exchange of the C-6 proton, there is a complex appearance of the signals in the region of the C-1′ (anomeric) proton (Figure 1B, right). This results from overlap of the signals of roughly equal size due to the anomeric proton of h-F-UMP at 5.885 ppm (double doublet, ³J_HH = 5.2 Hz, ⁵J_HF = 1.9 Hz, Figure 1A) and that of d-F-UMP, which appears as an upfield-shifted (Δδ = 0.002 ppm) double doublet at 5.883 ppm (³J_HH = 5.2 Hz, ⁵J_HF = 1.9 Hz, Figure 1C). At late reaction times the signal in the anomeric region is again simplified and is due mainly to the upfield-shifted anomeric proton of the product d-F-UMP. A similar upfield shift of the broad doublet due to the anomeric proton of UMP as a result of deuterium incorporation at C-6 was noted in our earlier communication.²²

Figure 1.

Partial ¹H NMR spectra at 500 MHz in the regions of the C-6 proton (left) and C-1′ (anomeric) proton (right) obtained during exchange for deuterium of the C-6 proton of F-UMP ([S]_o = 2.5 mM) catalyzed by ScOMPDC (9.2 μM) in D₂O at pD 9.3 (100 mM glycylglycine, 80% free base), 25 °C and I = 0.1 (NaCl).

The progress of deuterium exchange into F-UMP was obtained from the integrated area of the doublet at 7.87 ppm due to the C-6 proton of h-F-UMP (A_H), using the combined integrated areas of the signals due to the anomeric protons of both h-F-UMP and the product d-F-UMP at 5.88 ppm (A_H+D) as an internal standard (Figure 1). Observed first-order rate constants for deuterium exchange were determined from the slopes of linear semi-logarithmic plots of reaction progress against time according to eq 2, covering 1 – 2 reaction half-times. Tables S1 – S6 of the Supporting Information give the values of k_obsd (s⁻¹) for exchange for deuterium of the C-6 proton of F-UMP in D₂O at pD 6.45 – 9.33, 25 °C and I = 0.1 (NaCl) that were determined in these experiments.

$$k_{\text{obsd}}=-ln\left(\frac{A_{\text{H}}}{A_{\text{H}+\text{D}}}\right)t$$ (2)

Deuterium Exchange at C-6 of Uridine 5′-Monophosphate Catalyzed by Yeast OMPDC

The relatively slow exchange for deuterium of the C-6 proton of [6-¹H]-uridine 5′-monophosphate (h-UMP) ([S]_o = 2.5 – 10 mM) to give [6-²H]-uridine 5′-monophosphate (d-UMP) catalyzed by ScOMPDC (100 – 300 μM) in D₂O at pD 7.0 – 9.3, 25 °C and I = 0.1 (NaCl) was followed by monitoring the appearance of the broad signal due to the C-5 proton of the product d-UMP by ¹H NMR spectroscopy.²² The large coupling of the C-5 proton to the C-6 proton for h-UMP (³J_HH = 8 Hz) is replaced by a small unresolved ³J_HD coupling of the C-5 proton to deuterium at C-6 for d-UMP.²² The relatively high enzyme concentrations used in these experiments resulted in extremely broad ¹H NMR signals for UMP so that the reactions could not be followed directly. At various times an aliquot of the reaction mixture was withdrawn and the reaction was stopped by the addition of a 2 to 3-fold excess over enzyme of the tight-binding competitive inhibitor 6-azauridine 5′-monophosphate (6-aza-UMP, K_i = 93 nM at pH 7.2).^2,36 The enzyme•inhibitor complex, which was formed quantitatively, was removed by ultrafiltration and the filtrate was analyzed by ¹H NMR spectroscopy.

The fractional extent of deuterium incorporation into UMP, f(d-UMP), was determined from analysis of the integrated areas of the signals due to the C-5 protons of h-UMP and the product d-UMP.²² Observed first-order rate constants for deuterium exchange were determined from the slopes of semi-logarithmic plots of the fraction of h-UMP remaining, f(h-UMP) = {1 - f(d-UMP)} against time, covering up to 30% reaction.²² The values of k_obsd (s⁻¹) for exchange for deuterium of the C-6 proton of UMP in D₂O at pD 7.03 – 9.34, 25 °C and I = 0.1 (NaCl) were reported in the Supporting Information of the earlier communication.²² There was no detectable deuterium exchange into UMP (2 mM) after its incubation for 2 days in the presence of 260 μM ScOMPDC and 2 mM 6-aza-UMP at pD 9.3 (50 mM glycylglycine, 80% free base) at 25 °C. This shows that the observed enzyme-catalyzed deuterium exchange reactions occur at the active site of ScOMPDC.

The ScOMPDC-catalyzed deuterium exchange reactions of UMP at pD 7.6 – 9.3 were accompanied by up to 4% hydrolysis of the substrate 5′-phosphoryl group to give uridine. The very slow deuterium exchange reactions at pD 7.0 were accompanied by hydrolysis of both the 5′-phosphoryl group of UMP (up to 19%) and its glycosidic bond to give uracil (up to 24%). Control experiments conducted in the absence of ScOMPDC indicated that these side reactions are likely catalyzed by small contaminating enzyme activities present in our preparations of ScOMPDC. The rates of these hydrolysis reactions should be unaffected by the presence of deuterium at C-6 of UMP, so that they did not interfere with quantification of the extent of deuterium incorporation into the remaining UMP. However, at pD 7.0 in the presence of 2.5 mM UMP, the significant depletion of the pool of total UMP by these side reactions resulted in an apparent increase in the velocity of deuterium exchange with time (see kinetic analysis in the Discussion). In these cases, the observed rate constants for deuterium exchange were determined from the portion of the reaction that exhibited good first-order kinetics.²²

Ionization of F-UMP and UMP at N-3

The pK_a of the N-3 hydron of F-UMP in H₂O and in D₂O at 25 °C and I = 0.1 (NaCl) was determined spectrophotometrically by monitoring the absorbance of F-UMP at 269 nm as a function of pL. The data gave pK_NH = 7.84 ± 0.03 in H₂O and pK_ND = 8.37 ± 0.05 in D₂O, at 25 °C and I = 0.1 (NaCl). The pK_a of 7.84 for the N-3 proton of F-UMP is very similar to the reported pK_as of 7.90 for the N-3 proton of 5-fluoro-2′-deoxyuridine 5′-monophosphate,² and 7.67 for the N-3 proton of 5-fluoro-2′-deoxyuridine,³⁷ in H₂O at 25 °C and I = 0.1. This shows that there is little or no effect of a 2′-OH or a 5′-phosphoryl group on the pK_a of the N-3 proton of simple uridine derivatives. The pK_a of 9.3 reported for the N-3 proton of 2′-deoxyuridine in H₂O at 25 °C and I = 0.1 (NaCl)³⁷ and the 0.5 unit difference in the values of pK_NH and pK_ND for F-UMP determined here can be used to estimate pK_ND = 9.8 for ionization of the N-3 hydron of UMP in D₂O. Therefore, there is little dissociation of UMP at N-3 in the range of pD examined in our enzyme-catalyzed deuterium exchange reactions.

Nonenzymatic Deuterium Exchange at C-6 of Uridine, 5-Fluorouridine and 1,3-Dimethyl-5-Fluorouracil Catalyzed by Deuterioxide Ion in D₂O

The exchange for deuterium of the C-6 protons of uridine, 5-fluorouridine and 1,3-dimethyl-5-fluorouracil at various temperatures and values of pD in D₂O was followed by monitoring the disappearance of the C-6 proton of the substrate by ¹H NMR spectroscopy, as described in the Supporting Information. The reactions of 5-fluorouridine in the presence of 1 M NaOD and of 1,3-dimethyl-5-fluorouracil at pD 10.2 were conducted at 25 °C. The slower reactions of uridine in the presence of 1 M NaOD and of 5-fluorouridine at pD 7.1 – 9.4 (I = 0.1) were conducted at 60 – 90 °C and rate constants for the reactions at 25 °C were obtained by extrapolation of Arrhenius plots (see the Supporting Information). Table S7 of the Supporting Information summarizes the observed second-order rate constants (k_DO)_obsd (M⁻¹ s⁻¹) for the deuterioxide-ion-catalyzed exchange reactions that were determined in these experiments. The data gave the following second-order rate constants k_DO for deuterioxide-ion-catalyzed deuterium exchange at 25 °C (see Table S8 of the Supporting Information): 3.8 × 10⁻⁴ M⁻¹ s⁻¹ for uridine, 1.5 × 10⁻⁸ M⁻¹ s⁻¹ for uridine N-3 anion; 1.7 M⁻¹ s⁻¹ for 5-fluorouridine, 6.8 × 10⁻⁵ M⁻¹ s⁻¹ for 5-fluorouridine N-3 anion; 0.043 M⁻¹ s⁻¹ for 1,3-dimethyl-5-fluorouracil.

DISCUSSION

Kinetic Parameters for Enzyme-Catalyzed Deuterium Exchange into UMP and F-UMP

The velocity of enzyme-catalyzed deuterium exchange into h-UMP or h-F-UMP to give d-UMP or d-F-UMP (Scheme 3) is given by eq 3 derived for Scheme 4 (written for the reaction of UMP), where k_ex (s⁻¹) is the turnover number for the E•h-UMP or E•h-F-UMP complex, [S]_o is the total concentration of UMP or F-UMP irrespective of its isotopic composition, K_d is the binding constant for dissociation of the E•UMP or E•F-UMP complex, and [E] is the concentration of ScOMPDC.

Scheme 4.

$$\frac{-d[h-\text{UMP}]}{dt}=\frac{k_{ex}[\text{E}]{[h-\text{UMP}]}_{\text{t}}}{{[\text{S}]}_{\text{o}}+K_{\text{d}}}$$ (3)

$$\frac{k_{\text{obsd}}}{[\text{E}]}=\frac{k_{\text{ex}}}{{[\text{S}]}_{\text{o}}+K_{\text{d}}}$$ (4)

$$\frac{k_{\text{obsd}}{[\text{S}]}_{\text{o}}}{[\text{E}]}=\frac{k_{\text{ex}}{[\text{S}]}_{\text{o}}}{{[\text{S}]}_{\text{o}}+K_{\text{d}}}$$ (5)

Eq 4 relates the observed second-order rate constants for deuterium exchange, k_obsd/[E] (M⁻¹ s⁻¹), to the kinetic parameters k_ex and K_d (Scheme 4), where k_obsd is the observed first-order rate constant for deuterium exchange that was determined from the slope of a semi-logarithmic plot of reaction progress against time (see Experimental Section). This equation has two limits: (1) At [S]_o ≫ K_d, the observed second-order rate constant for deuterium exchange is inversely proportional to [S]_o, k_obsd/[E] = k_ex/[S]_o, where k_ex is the turnover number or the first-order rate constant for deuterium exchange into saturating enzyme-bound substrate. (2) At [S]_o ≪ K_d, the observed second-order rate constant for deuterium exchange tends to the true second-order rate constant k_ex/K_d.

Figure 2 shows the dependence of k_obsd/[E] (M⁻¹ s⁻¹) for deuterium exchange into F-UMP catalyzed by ScOMPDC (0.22 – 14 μM) on the total concentration of F-UMP in D₂O at pD 8.15, 25 °C and I = 0.10 (NaCl). This observed second-order rate constant decreases as the concentration of F-UMP is increased, which is the expected behavior for an enzyme-catalyzed isotope exchange reaction for which the substrate and the product bind to the enzyme with essentially equal affinities (K_d, Scheme 4). The data were fit to eq 4 to give values of k_ex = 0.048 s⁻¹ and K_d = 0.14 mM for deuterium exchange into F-UMP at pD 8.15.

Figure 2.

Dependence of the observed second-order rate constant k_obsd/[E] (M⁻¹ s⁻¹) for exchange for deuterium of the C-6 proton of F-UMP catalyzed by ScOMPDC (0.22 – 14 μM) on the total concentration of F-UMP in D₂O at pD 8.15, 25 °C and I = 0.10 (NaCl). The solid line shows the fit of the data to eq 4, which gives values of k_ex = 0.048 s⁻¹ and K_d = 0.14 mM.

Eq 4 may be transformed to give eq 5 which describes the familiar hyperbolic dependence of enzymatic initial rate on total substrate concentration. Figure 3A shows the dependence of the quantity k_obsd[F-UMP]_o/[E] (s⁻¹) for ScOMPDC-catalyzed deuterium exchange into F-UMP on the total concentration of F-UMP at pD 6.45 – 9.33, 25 °C and I = 0.1 (NaCl). The data were fit to eq 5 to give values of k_ex (s⁻¹) for deuterium exchange into saturating enzyme-bound F-UMP, and K_d (M) for binding of F-UMP to ScOMPDC. Table 1 gives the values of k_ex (s⁻¹), K_d (M) and the second-order rate constants k_ex/K_d (M⁻¹ s⁻¹) for enzyme-catalyzed deuterium exchange into F-UMP obtained from these fits.

Figure 3.

A. Dependence of k_obsd[F-UMP]_o/[E] (s⁻¹) for exchange for deuterium of the C-6 proton of F-UMP catalyzed by ScOMPDC on the total concentration of F-UMP in D₂O at 25 °C and I = 0.1 (NaCl). The solid lines show the fit of the data to eq 5 to give the values of k_ex (s⁻¹) and K_d (M) given in Table 1. Key: ●, Data at pD 6.45; ○ data at pD 7.03; ■, data at pD 7.37;□, data at pD 7.64; ▲, data at pD 9.33; △, data at pD 8.15. B. Dependence of k_obsd[UMP]_o/[E] for exchange for deuterium of the C-6 proton of UMP catalyzed by ScOMPDC on the total concentration of UMP in D₂O at 25 °C and I = 0.1 (NaCl). Key: ■, Data at pD 7.03; □, data at pD 7.58; ○, data at pD 8.13; ●, data at pD 9.34.

Table 1.

Kinetic Parameters for Exchange for Deuterium of the C-6 Proton of 5-Fluorouridine 5′-Monophosphate Catalyzed by ScOMPDC in D₂O at 25 °C and I = 0.1 (NaCl).^a

pD b	kex c s−1	Kd d mM	kex/Kd e M−1 s−1
9.33	3.84 × 10−2	0.19	200
8.15	4.38 × 10−2	0.11	400
7.64	3.24 × 10−2	0.11	300
7.37	2.42 × 10−2	0.17	140
7.03	1.69 × 10−2	0.51	33
6.45	8.18 × 10−3	0.99	8.3

^aObserved first-order rate constants, k_obsd (s⁻¹), for deuterium exchange into F-UMP were determined by monitoring the disappearance of the C-6 proton of F-UMP by ¹H NMR spectroscopy (Tables S1 – S6 of the Supporting Information).

^bAverage value of pD for experiments at several values of [F-UMP]_o.

^cTurnover number for deuterium exchange into enzyme-bound F-UMP, obtained from the nonlinear least squares fits of the values of k_obsd to eq 5 (Figure 3A). The estimated error in these rate constants is ± 10%.

^dDissociation constant for binding of F-UMP to ScOMPDC, obtained from the nonlinear least squares fits of the values of k_obsd to eq 5 (Figure 3A). The estimated error in these binding constants is ± 20%.

^eSecond-order rate constant for enzyme-catalyzed deuterium exchange into F-UMP, calculated as the ratio of the values of k_ex and K_d. The estimated error in these rate constants is ± 25%.

The observed second-order rate constants k_obsd/[E] (M⁻¹ s⁻¹) for the deuterium exchange reactions of h-UMP to give d-UMP at [S]_o = 2.5 – 10 mM are inversely proportional to the total concentration of UMP which shows that [S]_o ≫ K_d under these conditions (eq 4). Figure 3B shows the dependence of the quantity k_obsd[UMP]_o/[E] for ScOMPDC-catalyzed deuterium exchange into UMP on the total concentration of UMP at pD 7.03 – 9.34, 25 °C and I = 0.1 (NaCl). Table 2 gives the first-order rate constants k_ex (s⁻¹) for enzyme-catalyzed deuterium exchange into saturating enzyme-bound UMP that were determined as the average of the values of k_obsd[UMP]_o/[E] at each pD ([UMP]_o ≫ K_d, eq 5).

Table 2.

Kinetic Parameters for Exchange for Deuterium of the C-6 Proton of Uridine 5′-Monophosphate Catalyzed by ScOMPDC in D₂O at 25 °C and I = 0.1 (NaCl).^a

pD b	kex c s−1
9.34	1.15 × 10−5
8.13	7.53 × 10−6
7.58	3.19 × 10−6
7.03	1.20 × 10−6

^aObserved first-order rate constants, k_obsd (s⁻¹), for deuterium exchange into UMP were determined by monitoring the appearance of deuterium at the C-6 position by ¹H NMR spectroscopy [Ref. 22].

^bAverage value of pD for experiments at several values of [UMP]_o.

^cTurnover number for deuterium exchange into enzyme-bound UMP, obtained as the average value of k_obsd[UMP]_o/[E] at each pD (eq 5 and Figure 3B). The estimated error in these rate constants is ± 10%.

pD-Rate Profiles for Deuterium Exchange at C-6 of F-UMP and UMP

Figure 4 shows the pD-rate profile for the second-order rate constant k_ex/K_d (M⁻¹ s⁻¹) for ScOMPDC-catalyzed deuterium exchange at C-6 of F-UMP in D₂O, constructed using the data in Table 1. This pD-rate profile is bell-shaped and the slope of the ascending limb at low pD approaches 2.0, which shows that the second-order rate constant for enzyme-catalyzed deuterium exchange into F-UMP is dependent on the deprotonation of two groups at the free enzyme/substrate. We assign the downward break at high pD to ionization of the phosphodianion form of F-UMP at N-3 to give the substrate trianion (K_SD, Scheme 5), for which we have determined pK_SD = 8.37 in D₂O under the experimental conditions for deuterium exchange (see Results).

Figure 4.

pD-Rate profile for the second-order rate constant k_ex/K_d (M⁻¹ s⁻¹) for exchange for deuterium of the C-6 proton of F-UMP catalyzed by ScOMPDC in D₂O at 25 °C and I = 0.1 (NaCl). The solid line shows the fit of the data to eq 6 with the values of pK_SD2 = 6.8 and pK_SD = 8.37 to give pK_E = 8.5 ± 0.2 for deprotonation of the essential residue at the free enzyme and (k_ex)_max/K_d = 2300 ± 700 M⁻¹ s⁻¹ as the intrinsic second-order rate constant for the deuterium exchange reaction.

Scheme 5.

The pH-rate profile for the second-order rate constant k_cat/K_m (M⁻¹ s⁻¹) for decarboxylation of the natural substrate OMP by ScOMPDC in H₂O depends on the deprotonation of a group with an apparent pK_a of 6.1 which is very similar to the pK_a of 6.38 for the phosphoryl group of OMP determined by titration under the same conditions (25 °C, I = 0.1).² The X-ray crystal structure of ScOMPDC liganded with the potent inhibitor 6-hydroxyuridine 5′-monophosphate (BMP) shows hydrogen bonds between the ligand phosphoryl group and the side chains of Gln-215, Tyr-217 and Arg-235, and the backbone NHs of Arg-235 and Gly-234 (see Figure 6). This strongly implies that the enzyme binds and turns over the phosphodianion form of the substrate.^8,11,32 Therefore, a very reasonable assignment for the apparent pK_a of 6.1 in the pH-rate profile for k_cat/K_m for decarboxylation of OMP is the second ionization of the phosphoryl group of OMP to give the substrate phosphodianion.² By analogy, we assign the two ionizations at the free substrate/enzyme evident in the ascending limb at low pD in the pD-rate profile for k_ex/K_d for deuterium exchange into F-UMP (Figure 4) as ionization of the substrate phosphoryl monoanion (SD₂⁻, Scheme 5) to give the phosphoryl dianion (SD²⁻), and the deprotonation of a group at the free enzyme to give the basic form of the enzyme.

Figure 6.

Schematic showing the contacts between yeast ScOMPDC and bound 6-hydroxyuridine 5′-monophosphate constructed using X-ray crystallographic data (PDB entry 1DQX). The phosphodianion binding motif consists of the side chains of Gln-215 and Tyr-217 in the mobile loop (loop 7), the side chain of Arg-235, and the backbone NHs of Gly-234 and Arg-235. The completely conserved “catalytic motif” (DxKxxD) consists of the side chains of Asp-91, Lys-93 and Asp-96′ of the second subunit.

The pK_as of 6.26 and 6.25 for the phosphoryl groups of UMP and ribose 5′-monophosphate, respectively, determined by titration at 25 °C and I = 0.1,² are essentially identical, which shows that the ionization of this group is relatively insensitive to the presence of the pyrimidine ring. Therefore, we estimate pK_SD2 = 6.8 for ionization of the phosphoryl group of F-UMP in D₂O at I = 0.1 (Scheme 5, X = F), based on the pK_a of 6.3 for the phosphoryl group of UMP in H₂O at I = 0.1,² and the 0.5 unit higher pK_a of phosphate monoanion in D₂O than in H₂O.^38,39

Scheme 6 shows our proposed kinetic model for the ScOMPDC-catalyzed deuterium exchange reactions of F-UMP and UMP that is consistent with the observed bell-shaped pD-rate profile for k_ex/K_d for F-UMP (Figure 4). In this model the ionization of the substrate phosphoryl monoanion to give the phosphodianion (K_SD2) is required for binding and reaction of the substrate, and the deprotonation of an acidic residue at the enzyme (K_E) is required to generate a Brønsted base for the proton abstraction from C-6 of the substrate, a necessary step for deuterium exchange. Further ionization of the substrate at N-3 (pK_SD = 8.37) to give the F-UMP trianion results in a form of the substrate that is unable to bind and react, as evidenced by the downward break in the pD-rate profile at high pD (Figure 4). The values of k_ex/K_d for deuterium exchange into F-UMP were fit to eq 6, derived for Scheme 6, with the values of pK_SD2 = 6.8 and pK_SD = 8.37 to give pK_E = 8.5 ± 0.2 for deprotonation of the essential residue at the free enzyme. The fit also gives (k_ex)_max/K_d = 2300 ± 700 M⁻¹ s⁻¹ as the intrinsic second-order rate constant for the deuterium exchange reaction of the phosphodianion form of F-UMP (SD²⁻) catalyzed by the deprotonated enzyme (E). This intrinsic rate constant is ca. 6-fold larger than the largest experimental second-order rate constant for the deuterium exchange reaction, observed at pD 8.15 (Table 1).

Scheme 6.

$$log\left(\frac{k_{\text{ex}}}{K_{\text{d}}}\right)=log\left\{\frac{{(k_{\text{ex}})}_{max}}{K_{\text{d}}}\right\}-log\left(1+\frac{[{\text{D}}^{+}]}{K_{\text{E}}}\right)-log\left(1+\frac{[{\text{D}}^{+}]}{K_{\text{SD}2}}+\frac{K_{\text{SD}}}{[{\text{D}}^{+}]}\right)$$ (6)

$$log{k}_{\text{ex}}=log{(k_{\text{ex}})}_{max}-log\left(1+\frac{[{\text{D}}^{+}]}{K_{\text{ES}}}\right)$$ (7)

Figure 5 shows the pD-rate profiles for the first-order rate constants k_ex (s⁻¹) for the ScOMPDC-catalyzed deuterium exchange at C-6 of enzyme-bound F-UMP and UMP in D₂O, constructed using the data in Tables 1 and 2. As required by the model in Scheme 6, these profiles exhibit a dependence on the ionization state of only the enzyme-substrate complex ED⁺•SD²⁻, and they are consistent with the deprotonation of an essential residue at this complex to generate a Brønsted base for proton abstraction from C-6 of the substrate (K_ES, Scheme 6). The values of k_ex for F-UMP and UMP were fit to eq 7, derived for Scheme 6, to give pK_ES = 7.1 ± 0.1 and 8.0 ± 0.1 for deprotonation of the essential residue at the ED⁺•F-UMP and ED⁺•UMP complexes, respectively. The data also give the intrinsic first-order rate constants (turnover numbers) for deuterium exchange into saturating F-UMP and UMP bound to the deprotonated enzyme as (k_ex)_max = 0.041 s⁻¹ and 1.2 × 10⁻⁵ s⁻¹, respectively. Therefore, the addition of a 5-fluoro substituent results in a 3400-fold increase in the rate constant for deuterium exchange into saturating enzyme-bound UMP at the complex with the basic form of the enzyme (E•SD²⁻, Scheme 6).

Figure 5.

pD-Rate profiles for the first-order rate constants k_ex (s⁻¹) for exchange for deuterium of the C-6 proton of enzyme-bound F-UMP and UMP catalyzed by ScOMPDC in D₂O at 25 °C and I = 0.1 (NaCl). The solid lines show the fits of the data to eq 7. A. Data for F-UMP which give pK_ES = 7.1 ± 0.1 and (k_ex)_max = 0.041 s⁻¹. B. Data for UMP which give pK_ES = 8.0 ± 0.1 and (k_ex)_max = 1.2 × 10⁻⁵ s⁻¹.

Identity of the Catalytic Base for Deuterium Exchange

The first X-ray crystal structures of OMPDC appeared in 2000 and they strongly implicated the protonated side chain of Lys-93 (numbering for the yeast enzyme) as the catalytic Brønsted acid needed to deliver the proton incorporated at C-6 during the decarboxylation of OMP to give UMP (Scheme 2).^9–13 Figure 6, based on the X-ray crystal structure of ScOMPDC complexed with BMP,¹¹ shows that Lys-93 at ScOMPDC is part of the strictly conserved DxKxxD “catalytic motif” that is found in all OMPDCs,⁴⁰ the integrity of which is essential for robust decarboxylase activity.^25,41 The hydrogen bond between the ammonium group of Lys-93 and the anionic oxygen at C-6 of BMP, an analog of the putative carbanion intermediate,³⁶ is consistent with its proposed role as a Brønsted acid in the decarboxylation of OMP.^9–12 This is in accord with the results of earlier mutagenesis studies of yeast OMPDC that showed that Lys-93 is essential for robust enzymatic activity. The inactive K93C mutant was rescued by covalent modification of the Cys-93 side chain by 2-bromoethylamine, which restores a primary amino group to the side chain at position 93.⁴² More recently, the OMPDCs from M. thermautotrophicus (MtOMPDC), Plasmodium falciparum (PfOMPDC) and the OMPDC domain of the bifunctional human UMP synthase (HsOMPDC)⁴³ have been shown to undergo covalent modification by nucleophilic attack of the side chains of Lys-72, Lys-138 and Lys-314, respectively, at C-6 of bound 6-iodouridine 5′-monophosphate (6-I-UMP) and 6-azidouridine 5′-monophosphate (6-N₃-UMP), resulting in displacement of the good leaving groups iodide and azide ions (Scheme 7).^44–48 These lysine residues correspond to Lys-93 of the enzyme from yeast. The results of these structural and chemical studies strongly suggest that Lys-93 at the yeast enzyme functions not only to provide the proton incorporated at C-6 during the decarboxylation of OMP, but also as a Brønsted base in the reverse proton transfer “half reaction” that occurs as the first step of the C-6 deuterium exchange reactions of UMP and F-UMP (Scheme 2).

Scheme 7.

Figure 7 shows the X-ray crystal structures of wildtype HsOMPDC liganded by UMP and the D312N (D91N at ScOMPDC) mutant liganded by F-UMP. There is good superimposition of the side chains of the residues in the active sites of these enzymes and for simplicity Figure 7 depicts side chains of only the wildtype enzyme. Both UMP and F-UMP bind in the same manner observed for BMP at the yeast enzyme (Figure 6),¹¹ with the pyrimidine ring in the syn conformation and C-6 directed towards Lys-314. The pyrimidine rings of UMP and F-UMP are “sandwiched” between the hydrophobic side chains of Pro-417 (Pro-202 at ScOMPDC) and Ile-318′ (Ile-97′ at ScOMPDC) of the second subunit. The fluorine of F-UMP projects into a hydrophobic pocket lined by the side chains of Ile-368, Met-371 and Ile-401 (Leu-150, Leu-153 and Ile-183 at ScOMPDC).⁴⁶ The hydrophobic pocket thought to be occupied by the molecule of CO₂ generated during the decarboxylation of OMP lies on the opposite side of the pyrimidine ring from the essential lysine, and is lined by the side chains of Phe-310, Ile-401 and Ile-448 (Phe-89, Ile-183 and Ile-232 at ScOMPDC).⁴⁶ This pocket is occupied by a molecule of water in HsOMPDC liganded with UMP (Figure 7) and the same “conserved’ water molecule is observed at OMPDCs liganded with BMP.⁴⁹ Lys-314 is positioned close to C-6 of UMP and F-UMP with an Nζ-C6 distance of 3.06 Å in the UMP liganded structure. We therefore propose that the catalytic Brønsted base for the observed deuterium exchange reactions of F-UMP and UMP catalyzed by ScOMPDC is the deprotonated form of Lys-93.

Figure 7.

X-ray crystal structures of wildtype human OMPDC liganded by UMP (cyan, PDB entry 2QCD) and the D312N mutant liganded by F-UMP (cornflower blue, PDB entry 2QCF). For simplicity, side chains of only the wildtype enzyme are shown. A molecule of water (purple) is observed bound in the hydrophobic pocket adjacent to C-6 in the UMP-liganded structure but this water is absent in the F-UMP-liganded structure.

Figure 8 shows the X-ray crystal structures of wildtype HsOMPDC liganded by UMP and the D312N mutant liganded by the substrate OMP. UMP and OMP bind in the same manner but the carboxylate group at C-6 of OMP is distorted out of the plane of the pyrimidine ring by 36 degrees,^46,50 and there is no bound water molecule observed in the hydrophobic pocket. Similar out of plane distortions have been observed for the cyano group of 6-cyanouridine 5′-monophosphate and the functional groups at C-6 of other substituted pyrimidine nucleotides bound to OMPDC.^47,48,51 Nζ of Lys-314 lies within hydrogen bonding distance of the carboxylate group of OMP with an Nζ-O7 distance of 2.54 Å. The close correspondence of these structures supports the proposal that the protonated side chain of Lys-93 at the yeast enzyme acts as a catalytic acid in the decarboxylation of OMP and that its conjugate base is the catalytic base in the C-6 deuterium exchange reactions of UMP and F-UMP.

Figure 8.

X-ray crystal structures of wildtype human OMPDC liganded by UMP (cyan, PDB entry 2QCD) and the D312N mutant liganded by the substrate OMP (green, PDB entry 2QCL). The carboxylate group at C-6 of OMP is distorted out of the plane of the pyrimidine ring by 36 degrees. Nζ of Lys-314 lies within hydrogen bonding distance of the carboxylate group of OMP with an Nζ-O7 distance of 2.54 Å.

The pH-rate profile for k_cat/K_m for decarboxylation of OMP by ScOMPDC in H₂O is consistent with the requirement that a residue with a pK_a of 7.7 be in the acidic form, and this pK_a can be assigned to the side chain of Lys-93.^2,26 The pK_as of amine bases are ca. 0.6 unit higher in D₂O than in H₂O³⁸ so that the catalytic acid is expected to have a pK_a of around 8.3 in D₂O, which is very similar to pK_E = 8.5 for deprotonation of the essential group at the free enzyme determined from the pD-rate profile for k_ex/K_d for the deuterium exchange reaction of F-UMP in D₂O (Figure 4). The close identity of the pK_as at the free enzyme of the essential acid for decarboxylation of OMP and of the conjugate acid of the essential base for the deuterium exchange reaction of F-UMP supports our proposal that this base is the deprotonated form of Lys-93. Table 3 summarizes the pK_as in D₂O of the essential catalytic side chain at free ScOMPDC (pK_E) and those at the enzyme liganded by OMP, UMP and F-UMP (pK_ES), determined from the pL-rate profiles for decarboxylation of OMP² and for the C-6 deuterium exchange reactions of UMP and F-UMP. The value of pK_E = 8.5 in D₂O for the side chain of Lys-93 at free ScOMPDC is ca. 2.5 units lower than the expected intrinsic pK_a of 11.0 for this residue in D₂O.⁵² This depressed pK_a is intriguing because the protonated side chain of Lys-93 appears to be involved in hydrogen bonds with the carboxylate side chains of both Asp-91 and Asp-96′ of the second subunit in the DxKxxD motif (Figure 6). The relatively low pK_a of the ammonium group of Lys-93 has been noted previously,^3,42 but the structural data offer no strong clues as to its origin. However, we note here that the 2.3 unit variation in the pK_a of the ammonium group of Lys-93 at ScOMPDC complexed with OMP, UMP and F-UMP (Table 3) shows that the pK_a of this residue is highly sensitive to its local environment. The downward perturbation in the pK_a of Lys-93 may result in part from its proximity to the array of hydrophobic residues that surrounds the pyrimidine base (Figure 6), which should result in destabilization of the cationic protonated form.

Table 3.

Apparent pK_a of the Essential Residue at Free ScOMPDC and at the Enzyme Liganded by OMP, UMP and F-UMP in D₂O at 25 °C and I = 0.1 (NaCl).

Ligand	pKEa	pKESb
OMPc	8.3	9.4
UMP	8.5 ± 0.2d	8.0 ± 0.1e
F-UMP	8.5 ± 0.2d	7.1 ± 0.1f

^apK_a of the essential group at the free enzyme (Scheme 6).

^bpK_a of the essential group at the E•S complex (Scheme 6).

^cDetermined from the pH-rate profiles of k_cat/K_m (for pK_E) and k_cat (for pK_ES) for decarboxylation of OMP in H₂O [Ref. 2]. The pK_as in D₂O were estimated from the values in H₂O and the 0.6 unit higher pK_as of ammonium ions in D₂O than in H₂O [Ref. 38].

^dDetermined from the pD-rate profile for k_ex/K_d for deuterium exchange into F-UMP (Figure 4).

^eDetermined from the pD-rate profile for k_ex for deuterium exchange into UMP (Figure 5B).

^fDetermined from the pD-rate profile for k_ex for deuterium exchange into F-UMP (Figure 5A).

Intrinsic Affinity of OMPDC for UMP and F-UMP

The pD-rate profiles for k_ex for the deuterium exchange reactions of UMP and F-UMP (Figure 5) show that the pK_a of the conjugate acid of the essential base is lowered from pK_E = 8.5 for the free enzyme to pK_ES = 8.0 and 7.1 for the E•UMP and E•F-UMP complexes, respectively (Table 3). These results are consistent with a neutral rather than an anionic base because the presence of the hydrophobic pyrimidine ring at the liganded enzyme would be expected to destabilize a neighboring negative charge and result in an increase rather than a decrease in the basicity of the catalytic base. The relatively large 1.4 unit decrease in the pK_a of the side chain of Lys-93 upon the binding of F-UMP (Table 3) is consistent with the destabilization of positive charge at Nζ of protonated Lys-93 by the combined effects of the electron-withdrawing 5-fluoro substituent and the hydrophobic pyrimidine ring of F-UMP (Figure 7). Scheme 6 includes the thermodynamic cycle that relates the binding of F-UMP and UMP to the protonated (K_d′) and deprotonated enzymes (K_d) to the pK_as of the essential group at the free (pK_E) and liganded (pK_ES) enzymes. The relationship K_d′/K_d = K_ES/K_E and the difference (pK_E - pK_ES) = 1.4 gives K_d′/K_d = 25 as the ratio of the dissociation constants for F-UMP from the protonated and the deprotonated enzyme, respectively. Therefore, F-UMP binds 25-fold more tightly to the deprotonated form of ScOMPDC, E, than it does to the form of the enzyme that dominates at neutral pD, ED⁺. Similarly, the difference (pK_E - pK_ES) = 0.5 for UMP (Table 3) shows that UMP binds 3-fold more tightly to E than to ED⁺.

The intrinsic second-order rate constant for ScOMPDC-catalyzed deuterium exchange into F-UMP, (k_ex)_max/K_d = 2300 M⁻¹ s⁻¹, can be combined with the intrinsic first-order rate constant (k_ex)_max = 0.041 s⁻¹ to give K_d = 20 μM as the intrinsic affinity of the deprotonated enzyme for the phosphodianion form of F-UMP in D₂O. This intrinsic affinity is 6-fold greater than the largest experimental affinity observed at pD 7.6 – 8.2 (Table 1). The observed affinities, (K_d)_obsd, of ScOMPDC for 2′-deoxyuridine 5′-monophosphate and 2′-deoxy-5-fluorouridine 5′-monophosphate of 170 μM and 250 μM, respectively, at pH 7.2 in H₂O show that the addition of a 5-fluoro substituent results in only a small ca. 1.5-fold change in the observed affinity of ScOMPDC for the 2′-deoxypyyrimidine nucleotide in H₂O at pH 7.2.² These values of (K_d)_obsd at pH 7.2 can be combined with eq 8, derived for Scheme 6 (in H₂O), to give relative intrinsic affinities K_d of the deprotonated form of ScOMPDC for the phosphodianion forms of UMP and F-UMP of 1.5:1. This suggests that the intrinsic affinity of ScOMPDC for UMP in D₂O is essentially the same as K_d = 20 μM determined for F-UMP. Therefore, we conclude that the larger decrease in the affinity of the enzyme for F-UMP than for UMP upon protonation of the catalytic base results largely from a specific destabilizing interaction of the electron-withdrawing fluorine at C-5 of bound F-UMP with the protonated enzyme.

$$K_{\text{d}}={(K_{\text{d}})}_{\text{obsd}}\frac{\left(1+\frac{[{\text{H}}^{+}]}{K_{\text{ES}}}\right)}{\left(1+\frac{[{\text{H}}^{+}]}{K_{\text{E}}}\right)\left(1+\frac{[{\text{H}}^{+}]}{K_{\text{SH}2}}+\frac{K_{\text{SH}}}{[{\text{H}}^{+}]}\right)}$$ (8)

By contrast with F-UMP and UMP, the binding of OMP to ScOMPDC results in a ca. 1.1 unit increase in the pK_a of 8.3 in D₂O for Lys-93 at the free enzyme, as evidenced by the downward break at pH 8.8 in the pH-rate profile for k_cat for decarboxylation of OMP,² which corresponds to a pK_a of 9.4 in D₂O (Table 3). This increase in basicity of the side chain of Lys-93 upon the binding of OMP is consistent with electrostatic stabilization of the positive charge at Nζ by its proximity to the negative charge of the carboxylate group at C-6 of bound OMP (Figure 8).

Mechanism of Deuterium Exchange: Formation of a Vinyl Carbanion at OMPDC

We reported earlier that the decarboxylation of OMP catalyzed by wildtype and several mutant ScOMPDCs in a mixed solvent of 50:50 (v/v) H₂O/D₂O results in the formation of an equimolar mixture of the products h-UMP and d-UMP labeled with H and D at C-6, respectively, so that the product deuterium isotope effect (PIE) on decarboxylation is unity (Scheme 2).^20,21 Furthermore, the values of Φ_NL3+ ≈ 1.0 for H/D fractionation between L₂O and R-NL₃⁺ show that the primary kinetic solvent isotope effect (KIE) is essentially equal to the PIE.^21,53 Therefore, there is no discrimination between H and D at the proton donor in the transition state for the rate-determining step of the decarboxylation of OMP. The absence of any significant KIE shows that the rate-limiting step for enzyme-catalyzed decarboxylation of OMP is strongly uncoupled from the proton transfer step. It eliminates the possibility of a concerted mechanism for decarboxylation of OMP^9,13 in which proton transfer from Lys-93 to C-6 provides electrophilic push to the loss of CO₂. We concluded that the OMPDC-catalyzed decarboxylation of OMP proceeds by the stepwise mechanism through the UMP vinyl carbanion intermediate shown in Scheme 8, in which the transition state for the loss of CO₂ from OMP to give the bound UMP carbanion is insensitive to the isotopic composition of the adjacent ζ-NL₃⁺ group of Lys-93.^20,21 The PIE of unity requires that hydron transfer to the bound vinyl carbanion be much faster than the possible rotation of the terminal CH₂-NL₃⁺ bond of the side chain of the proton donor Lys-93, which exchanges the positions of its NL₃⁺ hydrons (k_p ≫ k_rot). Therefore, the yields of h-UMP and d-UMP from the decarboxylation of OMP in an H₂O/D₂O mixture are determined by the initial isotopic enrichment of the ζ-NL₃⁺ group of Lys-93 (governed by the fractionation factor Φ_NL3+ ≈ 1) at the E•OMP Michaelis complex.^20,21 The observation of a small inverse PIE of 0.93 for the decarboxylation of F-OMP catalyzed by wildtype ScOMPDC shows that the expected decrease in the chemical reactivity of the carbanion intermediate as a result of the 5-fluoro substituent does not result in any large change in the inequality k_p ≫ k_rot.²¹ This small inverse PIE is consistent with a KIE of unity for hydron transfer to the enzyme-bound F-UMP vinyl carbanion intermediate, along with a small increase in the fractionation factor for the ζ-NL₃⁺ group of Lys-93 at the E•F-OMP Michaelis complex to Φ_NL3+ ≈ 1.1.²¹

Scheme 8.

The ScOMPDC-catalyzed deuterium exchange reactions of UMP and F-UMP examined here show that deuterium exchange into saturating enzyme-bound F-UMP [(k_ex)_max = 0.041 s⁻¹] is 3400-fold faster than exchange into enzyme-bound UMP [(k_ex)_max = 1.2 × 10⁻⁵ s⁻¹] so that there is a large 4.8 kcal/mol stabilization of the transition state for deuterium exchange by the electron-withdrawing 5-fluoro substituent at F-UMP. This is consistent with the formation of a carbanion intermediate at which the negative charge is stabilized by interaction with the electron-withdrawing 5-fluoro substituent, and a deuterium exchange reaction that is the reverse of the proton transfer “half reaction” involved in the physiological decarboxylation of OMP.

Scheme 9 shows our proposed mechanism for the OMPDC-catalyzed deuterium exchange reactions of UMP and F-UMP.²² Proton transfer from C-6 of UMP and F-UMP to the neutral side chain of Lys-93 (k_-p) results in an enzyme-bound vinyl carbanion intermediate that can be rapidly reprotonated via “internal return” of the abstracted hydron to regenerate the unlabeled nucleotide (k_p ≫ k_rot). This reversible chemical step is followed by the rare rate-limiting rotation of the terminal CH₂-ND₂H⁺ bond of the side chain of Lys-93 in to a “reactive position” for delivery of a deuteron to C-6 (k_rot).

Scheme 9.

The rate constant for CH₂-NL₃⁺ bond rotation in water is expected to be similar to the value of k_reorg = 10¹¹ s⁻¹ estimated for reorganization that exchanges the relative positions of -O and -S at a thiobenzoate•carbocation ion pair.⁵⁴ However, the side chain of protonated Lys-93 is “anchored” by hydrogen bonding to the carboxylate groups of Asp-91 and Asp-96′ in a “double salt bridge” (Figure 6), which is expected to result in a substantial barrier to rotation of the terminal CH₂-NL₃⁺ bond of this side chain at the ScOMPDC•carbanion complex. The strength of the stabilizing interactions in salt bridges buried in the interior of proteins is 3 –5 kcal/mol^55,56 so that we estimate this barrier, which should involve the breaking of hydrogen bonds between protonated Lys-93 and both Asp-91 and Asp-96′, to be at least 8 kcal/mol. This gives an upper limit on the rate constant for bond rotation of k_rot ≤ 10⁷ s⁻¹.

Carbon Acidity of Substituted Uracils in Aqueous Solution

Prior to the current work, a critical unknown was the effect of a 5′-phosphoribosyl group at N-1 and/or a 5-fluoro substituent on the C-6 carbon acidity of uracil derivatives in aqueous solution (Scheme 10). Therefore, we have determined second-order rate constants k_DO (M⁻¹ s⁻¹) for the DO⁻-catalyzed deuterium exchange reactions of uridine, 5-fluorouridine and 1,3-dimethyl-5-fluorouracil in D₂O at a variety of temperatures and pD (see Results and Supporting Information). The exchange reactions of uridine and 5-fluorouridine in the presence of 1 M NaOD provided k_DO (M⁻¹ s⁻¹) for deuterium exchange into their N-3 anions, while the exchange reaction of 5-fluorouridine at pD 7.1 provided k_DO for deuterium exchange into the neutral nucleoside (pK_ND = 8.4, this work). The value of k_DO for deuterium exchange into neutral uridine was calculated from k_DO determined for its N-3 anion, with the assumption that ionization of uridine at N-3 results in the same 2.5 × 10⁴-fold decrease in reactivity of the C-6 proton that is observed upon ionization of 5-fluorouridine at N-3 (Table S8 of the Supporting Information).

Scheme 10.

The values of k_DO for the substituted uracils (Table S8 of the Supporting Information) were used to calculate the second-order rate constants k_HO (M⁻¹ s⁻¹) for deprotonation of these carbon acids by hydroxide ion in H₂O given in Table 4, using a secondary solvent isotope effect of k_DO/k_HO = 2.4.^57–59 The carbon acid pK_as were then calculated using eq 9, derived for Scheme 11, with pK_w = 14 for H₂O and k_HOH = 10¹¹ s⁻¹ for the reverse protonation of localized unstable carbanions by solvent water that is limited by the rotation of a molecule of water into a reactive position.^57–60

Table 4.

Kinetic and Thermodynamic Acidity of the C-6 Protons of N-1, N-3 and C-5 Substituted Uracils in Water at 25 °C (Schemes 10 and 11).

Substrate	X	R1	R2	kHO a M−1 s−1	pKCH b
Uridine	H	Ribosyl	H	1.6 × 10−4	28.8 ± 1
Uridine N-3 Anion	H	Ribosyl	Anion	6.3 × 10−9	33.2 ± 1
1,3-Dimethyluracil	H	Me	Me	2 × 10−6 c	30.7 ± 1
5-Fluorouridine	F	Ribosyl	H	0.71	25.1 ± 0.5
5-Fluorouridine N-3 Anion	F	Ribosyl	Anion	2.8 × 10−5	29.6 ± 0.5
1,3-Dimethyl-5-fluorouracil	F	Me	Me	1.8 × 10−2	26.7 ± 0.5

^aSecond-order rate constant for proton transfer from C-6 to hydroxide ion in H₂O, calculated from the value of k_DO for deuterium exchange (Table S8 of the Supporting Information) using a secondary solvent isotope effect of k_DO/k_HO = 2.4 (see text).

^bCarbon acid pK_a for ionization of the substrate at C-6 in H₂O, calculated from k_HO (M⁻¹ s⁻¹) for deprotonation of the carbon acid by hydroxide ion and k_HOH = 10¹¹ s⁻¹ for the reverse protonation of the vinyl carbanion by solvent water, using eq 9.

^cCalculated from data for the competing C-6 deuterium exchange and hydrolysis reactions of the substrate in 1 M NaOD [Ref. 61].

Scheme 11.

$${\text{p}K}_{\text{CH}}={\text{p}K}_{\text{w}}+log\left(\frac{k_{\text{HOH}}}{k_{\text{HO}}}\right)$$ (9)

Table 4 summarizes the C-6 carbon acidities, pK_CH (Scheme 10), of the model nucleosides uridine and 5-fluorouridine in water, along with those of their N-3 anions and the corresponding 1,3-dimethyluracils. The value of pK_CH = 30.7 for the C-6 proton of 1,3-dimethyluracil (Scheme 10, X = H, R¹ = R² = Me), obtained from rate constants for its competing exchange and hydrolysis reactions in 1 M NaOD at 25 °C,⁶¹ is lower than the earlier reported value of 34 ± 2 determined from the temperature dependence of deuterium exchange in acetate buffers at 175 – 217 °C.⁶² There is a good linear correlation (not shown) with a slope of ρ_I = −8.0 of the carbon acid pK_as of 3-R-4-methylthiazolium cations (Scheme 12),⁶³ “normal” carbon acids that undergo ionization to generate localized carbanions,^59,64,65 with the inductive substituent constant σ_I for the 3-substituent, and, by analogy, a similar free energy relationship is expected for the carbon acid pK_as of N-1 substituted uracils. The value of σ_I = −0.05 for the methyl group⁶⁶ and the estimated value of σ_I = 0.14 for the ribosyl group⁶⁷ can be combined with ρ_I = −8.0 to estimate a decrease in C-6 carbon acidity of 1.5 pK units upon the change from a methyl to a ribosyl group at N-1 of substituted uracils (R¹, Scheme 12). This is consistent with the 1.6 unit difference in the values of pK_CH for 1,3-dimethyl-5-fluorouracil and 5-fluorouridine determined here (Table 4).

Scheme 12.

The data in Table 4 show that the addition of a 5-fluoro substituent to substituted uracils results in an increase in the C-6 carbon acidity of 3.6 – 4.0 pK units. By contrast, the addition of a 5-fluoro substituent results in decreases of 1.5, 1.6 and 1.7 – 1.9 units in the pK_a of the N-3 proton of uracil,³⁷ 2′-deoxyuridine,³⁷ and UMP,⁶⁸ respectively. These substituent effects are consistent with a fall-off factor of ca. 2 for the presence of the additional carbon atom between the 5-fluoro substituent and the center of negative charge at substituted uracil N-3 anions compared with their C-6 carbanions.⁵⁷

Stabilization of the UMP and F-UMP Vinyl Carbanions by ScOMPDC

Figure 9 shows free energy profiles for the enzyme-catalyzed deuterium exchange reactions of UMP and F-UMP through the respective C-6 vinyl carbanion intermediates. These profiles were constructed according to the mechanism shown in Scheme 9, with the observed 3400-fold faster deuterium exchange into enzyme-bound F-UMP [(k_ex)_max = 0.041 s⁻¹] than UMP [(k_ex)_max = 1.2 × 10⁻⁵ s⁻¹] at high pD (Figure 5), and a barrier to rotation of the terminal CH₂-NL₃⁺ bond of the side chain of Lys-93 of 8 kcal/mol (k_rot = 10⁷ s⁻¹, see above). These profiles refer to the exchange reactions of the substrate phosphodianions (SD²⁻) catalyzed by ScOMPDC at which Lys-93 is deprotonated (END₂). The essentially identical values of K_d = 20 μM for binding of F-UMP and UMP to the deprotonated enzyme correspond to a free energy of binding to form the Michaelis complex of 6.4 kcal/mol. Proton abstraction from C-6 (k_-p) leads to the formation of the initial enzyme-carbanion complex which can undergo rapid internal return by reprotonation (k_p) or the occasional, rate-limiting, rotation of the side-chain of Lys-93 into a position to deliver a deuteron to C-6 (k_rot), with an estimated barrier to rotation of at least 8 kcal/mol (see above). Once this rotational event occurs there is rapid and irreversible transfer of a deuteron to C-6 to generate the deuterium exchange product, k_p ≫ k_rot, as evidenced by the PIEs of unity determined for the decarboxylation of OMP in 50:50 (v/v) H₂O/D₂O.^20,21

Figure 9.

Free energy profiles for the enzyme-catalyzed deuterium exchange reactions of the phosphodianion forms of UMP and F-UMP (SD²⁻) through the respective C-6 vinyl carbanion intermediates. These profiles were constructed according to the mechanism shown in Scheme 9, with the observed 3400-fold faster deuterium exchange into enzyme-bound F-UMP than UMP and a barrier to rotation of the terminal CH₂-NL₃⁺ bond of the side chain of Lys-93 of 8 kcal/mol (see text).

The formation of vinyl carbanions derived from UMP and F-UMP as intermediates of their C-6 deuterium exchange reactions (Scheme 9 and Figure 9) allows equilibrium constants (K_eq)_enz for proton transfer from C-6 of enzyme-bound UMP and F-UMP to the side chain of Lys-93 (Scheme 13A) to be obtained from the first-order rate constants (k_ex)_max for deuterium exchange into the enzyme-bound nucleotides, according to eqs 10 and 11. The values of (k_ex)_max = 1.2 × 10⁻⁵ and 0.041 s⁻¹ for deuterium exchange into enzyme-bound UMP and F-UMP, respectively, and k_rot = 10⁷ s⁻¹ were substituted into eq 11 to give (dimensionless) equilibrium constants for proton transfer from C-6 of the bound nucleotide to ScOMPDC of (K_eq)_enz = 1.2 × 10⁻¹² and 4.1 × 10⁻⁹, respectively, so that there is a large 4.8 kcal/mol stabilizing effect of the 5-F substituent on this proton transfer equilibrium (Figure 9). The values of pK_ES = 8.0 and 7.1 for the side chain of Lys-93 at ScOMPDC complexed with UMP and F-UMP, respectively (Table 3), show that there is an offsetting 1.2 kcal/mol destabilizing effect of the 5-fluoro substituent on equilibrium proton transfer to Nζ of Lys-93, as a result of interactions of the electron-withdrawing fluorine at C-5 of bound F-UMP with the positively charged ammonium group (vide infra). Therefore, the 5-fluoro substituent at F-UMP results in an overall decrease in the thermodynamic barrier for proton transfer from C-6 of 6.0 kcal/mol, which corresponds to a 4.4 unit increase in the carbon acidity of the bound nucleotide. This is similar to the 3.7 pK unit difference in the C-6 carbon acidities of the model nucleosides uridine (pK_CH = 28.8 ± 1) and 5-fluorouridine (pK_CH = 25.1 ± 0.5) in aqueous solution (Table 4).

Scheme 13.

$${(k_{\text{ex}})}_{max}=\left(\frac{k_{-\text{p}}}{k_{\text{p}}}\right)k_{\text{rot}}={(K_{\text{eq}})}_{\text{enz}}{k}_{\text{rot}}$$ (10)

$${(K_{\text{eq}})}_{\text{enz}}=\frac{{(k_{\text{ex}})}_{max}}{k_{\text{rot}}}$$ (11)

$${(K_{\text{eq}})}_{\text{aq}}=\frac{K_{\text{CH}}}{K_{\text{RNH}3}}$$ (12)

Similarly, overall equilibrium constants for proton transfer from C-6 of UMP and F-UMP to a primary amine base of pK_RNH3 = 7 in water (Scheme 13B) can be calculated using eq 12, with the values of pK_CH for the model nucleosides uridine and 5-fluorouridine from Table 4, as (= 2 × 10⁻²² and 8 × 10⁻¹⁹, respectively. Therefore, the binding of UMP or F-UMP to K_eq)_aq ScOMPDC results in a ~5 × 10⁹-fold increase in the equilibrium constant for proton transfer from C-6 to an amine base, which corresponds to a thermodynamic stabilization of the enzyme-bound UMP and F-UMP vinyl carbanions, relative to the enzyme-nucleotide complexes, of 13 kcal/mol. A critical unknown is the magnitude of the barrier to rotation of the terminal CH₂-NL₃⁺ bond of the side chain of Lys-93, which is hydrogen-bonded to the anionic side chains of Asp-91 and Asp-96′ at the enzyme•carbanion complex (Figure 6). We have estimated this barrier as ≥ 8 kcal/mol, so that 13 kcal/mol is a lower limit on the stabilization of the bound UMP and F-UMP vinyl carbanions resulting from their interactions with ScOMPDC.

Enzymatic Rate Accelerations for Decarboxylation and Deuterium Exchange

The formation of vinyl carbanions as intermediates in the C-6 deuterium exchange reactions of UMP and F-UMP catalyzed by ScOMPDC provides convincing evidence for the decarboxylation of OMP and F-OMP by a stepwise mechanism through the same carbanions. The total rate acceleration for the decarboxylation of OMP by free ScOMPDC can be calculated as 4 × 10²² M⁻¹, from the ratio of k_cat/K_m = 1.1 × 10⁷ M⁻¹ s⁻¹ for decarboxylation at pH 7.1,⁷ which is the maximum in the pH-rate profile for decarboxylation,² and k_o = 2.8 × 10⁻¹⁶ s⁻¹ for the nonenzymatic decarboxylation of OMP in aqueous solution.⁴ This corresponds to a total stabilization of the transition state for decarboxylation of OMP of 31 kcal/mol.³ It is more difficult to establish the rate acceleration for the enzyme-catalyzed deuterium exchange reactions of UMP and F-UMP, because the nonenzymatic proton transfer reactions of these normal carbon acids in solution proceed by proton transfer to lyoxide ion so that the observed rate constant depends strongly on pL. This notwithstanding, the observed rate constant for C-6 deprotonation of 5-fluorouridine (a model for F-UMP) by hydroxide ion in water at pH 7.1 can be calculated as k_o = k_HO[HO⁻] = 9 × 10⁻⁸ s⁻¹ (Table 4). The intrinsic second-order rate constant for ScOMPDC-catalyzed deuterium exchange into F-UMP catalyzed by the deprotonated enzyme is (k_ex)_max/K_d = 2300 M⁻¹ s⁻¹. The ratio of these rate constants gives the enzymatic rate acceleration for deuterium exchange as 3 × 10¹⁰ M⁻¹, which corresponds to a transition state stabilization of 14 kcal/mol. Together with the ≥ 13 kcal/mol stabilization of the bound UMP and F-UMP vinyl carbanions by interactions with the enzyme estimated above, this shows that a large portion of the enzymatic rate acceleration for the decarboxylation of OMP results from stabilization of the bound carbanion intermediate by its interactions with ScOMPDC.

These data suggest that the total transition state stabilization for decarboxylation of OMP via the UMP vinyl carbanion intermediate exceeds that for its formation by proton transfer from C-6 of UMP to the enzyme by ca. 17 kcal/mol. We suggest several factors that may contribute to this apparent greater proficiency of ScOMPDC for decarboxylation than deuterium exchange.

(1) The difference in the transition state stabilization for decarboxylation compared with deuterium exchange would be lowered to 14 kcal/mol if there is uncertainty in the rate constant for the nonenzymatic decarboxylation of OMP such that this reaction is ca. 100-fold faster than the current best estimate.⁴

(2) The decarboxylation of OMP by ScOMPDC is limited in part by the barrier to the chemical step of the loss of CO₂ to give the carbanion intermediate.^2,26,33,69 By contrast, the rate-limiting step for enzyme-catalyzed deuterium exchange is the physical step of rotation of the terminal CH₂-NHD₂⁺ bond of the side chain of Lys-93 into a “reactive position”, with an estimated barrier of ≥ 8 kcal/mol (k_rot, Figure 9). Therefore, the energy of the transition state for the chemical proton transfer step (k_-p) will be lower than that for the overall deuterium exchange reaction by an amount RTln(k_p/k_rot) so that the rate acceleration for deuterium exchange underestimates the rate acceleration for formation of the bound vinyl carbanion from UMP by this amount.

(3) The initial product of the decarboxylation of OMP is the ternary complex of OMPDC with UMP and a molecule of CO₂ bound in the hydrophobic pocket below C-6 of UMP (Figure 7). If the presence of CO₂ in the hydrophobic pocket activates the enzyme for catalysis of proton transfer, then the true reactivity of the enzyme towards proton transfer from C-6 of bound UMP and F-UMP could be higher than that observed in the deuterium exchange reactions described here.

(4) The transition state for decarboxylation of OMP may be stabilized by favorable “solvation” of the departing CO₂ molecule by the development of stabilizing interactions with the hydrophobic side chains of Phe-89, Ile-183 and Ile-232 that line the CO₂ binding pocket at yeast OMPDC (Phe-310, Ile-401 and Ile-448 at HsOMPDC, Figure 7). These interactions, which are not present in the ground state of the free enzyme and OMP, would result in selective stabilization of the transition state for formation of the UMP carbanion intermediate by decarboxylation that is not available for the transition state for its formation by proton transfer from C-6 of UMP.

Finally, we note that ground state destabilization by the introduction of “strain” at the protein and/or electrostatic “stress” at the substrate in the OMPDC•OMP Michaelis complex may contribute to an increase in k_cat for the decarboxylation of OMP, and, possibly, proton transfer from C-6 of UMP and F-UMP.^12,17,70 However, the overall rate accelerations for decarboxylation and exchange calculated here from the rate constants k_cat/K_m are unaffected by such effects because they refer to the barrier for conversion of the free enzyme and substrate in solution to the enzyme-bound transition state.⁷¹