Enzyme Architecture: Breaking Down the Catalytic Cage that Activates Orotidine 5′-Monophosphate Decarboxylase for Catalysis

Abstract

We report the results of a study of the catalytic role of a network of four interacting amino acid side chains at yeast orotidine 5′-monophosphate decarboxylase (ScOMPDC), by the stepwise replacement of all four side chains. The H-bond, which links the −CH₂OH side chain of S154 from the pyrimidine umbrella loop of ScOMPDC to the amide side chain of Q215 in the phosphodianion gripper loop, creates a protein cage for the substrate OMP. The role of this interaction in optimizing transition state stabilization from the dianion gripper side chains Q215, Y217, and R235 was probed by determining the kinetic parameter k_cat/K_m for 16 enzyme variants, which include all combinations of single, double, triple, and quadruple S154A, Q215A, Y217F, and R235A mutations. The effects of consecutive Q215A, Y217F, and R235A mutations on ΔG^⧧ for wild-type enzyme-catalyzed decarboxylation sum to 11.6 kcal/mol, but to only 7.6 kcal/mol when starting from S154A mutant. This shows that the S154A mutation results in a (11.6–7.6) = 4.0 kcal/mol decrease in transition state stabilization from interactions with Q215, Y217, and R235. Mutant cycles show that ca. 2 kcal/mol of this 4 kcal/mol effect is from the direct interaction between the S154 and Q215 side chains and that ca. 2 kcal/mol is from a tightening in the stabilizing interactions of the Y217 and R235 side chains. The sum of the effects of individual A154S, A215Q, F217Y and A235R substitutions at the quadruple mutant of ScOMPDC to give the corresponding triple mutants, 5.6 kcal/mol, is much smaller than 16.0 kcal/mol, the sum of the effects of the related four substitutions in wild-type ScOMPDC to give the respective single mutants. The small effect of substitutions at the quadruple mutant is consistent with a large entropic cost to holding the flexible loops of ScOMPDC in the active closed conformation.

Introduction

Orotidine 5′-monophosphate decarboxylase (OMPDC) employs no metal ions or other cofactors, yet results in an enormous 31 kcal/mol stabilization of the transition state for the chemically very difficult decarboxylation of orotidine 5′-monophosphate (OMP) to give uridine 5′-monophosphate (UMP).¹⁻³ This 31 kcal/mol transition state stabilization is the sum of 12, 10, and 9 kcal/mol stabilizing interactions (Scheme 1) from the phosphodianion, ribosyl ring, and pyrimidine ring fragments of the substrate, respectively.⁴ The enzymatic reaction proceeds by a stepwise mechanism, through a UMP carbanion reaction intermediate (Scheme 1), that is stabilized by ≥16 kcal/mol, relative to UMP, by interactions with the protein catalyst.^5,6

Scheme 1. Contribution of Individual Substrate Fragments to the Total Stabilization of the Transition State for OMPDC-Catalyzed Decarboxylation of OMP through the UMP Carbanion Reaction Intermediate.

The X-ray crystal structure for the complex between yeast OMPDC (ScOMPDC) and the tight-binding intermediate analog 6-azauridine 5′-monophosphate (6-azaUMP) shows a network of interactions between the protein and ligand.^7,8 We are interested in understanding the mechanism by which interactions of the three-substrate fragments are funneled toward stabilization of the transition state for ScOMPDC-catalyzed decarboxylation of the pyrimidine ring of OMP.^1,9

A crucial network of amino acid side chains (Figure 1A,B) extend from the phosphodianion gripper loop shown in blue (P202 to V220) to the pyrimidine umbrella shown in green (A151 to T165) and consists of Q215, Y217, R235, and S154. These side-chains are poorly placed to interact with a hypothetical enzyme-bound transition state at the open enzyme (Figure 1A). Ligand binding drives a large enzyme conformational change to give the closed enzyme (Figure 1B), in which the side chains provide strong transition state stabilization.¹⁰⁻¹⁵ This network plays at least two roles in catalysis of the decarboxylation of OMP. (1) The side chains of Q215, Y217, and R235 are distant from the pyrimidine ring and do not interact directly with the site of ScOMPDC-catalyzed decarboxylation,¹³ but promote decarboxylation by stabilizing the catalytically active closed form of ScOMPDC (E_C, Scheme 2) relative to the inactive open form E_O.^9,16−18 (2) The −CH₂OH side chain of S154 bridges the phosphodianion and the pyrimidine binding sites at E_C by donating an H-bond to the amide side chain of Q215 and accepting an H-bond from the C-5 pyrimidine −NH. These bridging interactions lock the substrate in a protein cage, by stabilizing the active closed form of ScOMPDC (E_C) relative to E_O (Figure 1), while holding the side chains from Q215, Y217, and R235 close to the dianion.

Figure 1.

Comparison of the X-ray crystal structures for (A) unliganded ScOMPDC (Protein Databank (PDB) entry 3GDL) and (B) the complex between ScOMPDC and the tight-binding substrate analog 6-azauridine 5′-monophosphate (6-azaUMP) (PDB entry 3GDK). The hypothetical 6-azaUMP ligand has been inserted in panel A to mimic the position of the ligand at panel B. (B) shows the ligand-driven movement of the phosphodianion gripper loop (P202 to 220, shaded blue) toward the pyrimidine umbrella loop (A151 to T165, shaded green) and of R235 toward the phosphodianion. The −CH₂OH side chain of S154 bridges these loops and holds Q215 near the dianion by donating an H-bond to the amide side chain of Q215 and accepting an H-bond from the C-5 pyrimidine −NH (dashed red circle).

Scheme 2. Two-Step Mechanism for Substrate Binding to OMPDC.

We have shown that the S154A mutation of ScOMPDC results in a large 5.8 kcal/mol destabilization of the transition state for decarboxylation of OMP and provided evidence that S154 plays an important role in orienting Q215 to interact with the substrate phosphodianion.¹⁵ We now address the more complex question of whether there is cooperativity between S154 and the dianion gripper side chains Y217 and R235 in stabilizing the decarboxylation transition state. We have probed the functioning of these four interacting side chains by site-directed mutagenesis. Such analyses are straightforward for small networks of two amino acid side chains,¹⁹ but studies on networks larger that two side chains are relatively rare. These include an analysis of the role of the three lysine side chains at the KMSKS mobile loop, in catalysis by tyrosyl amino acid synthase,²⁰ and our recent examination of the network of three side chains (Q215, Y217, and R235) that interact with the phosphodianion of substrates bound to OMPDC.¹¹

We are not aware of systematic mutagenesis studies on networks larger than three amino acids. This study on the S154, Q215, Y217, and R235 network (Figure 1) reports data to extend the following earlier work: (a) Studies on the effect of all single, double, and triple Q215A, Y217F, and R235A mutations on the kinetic parameters for OMPDC catalyzed reactions of whole substrates OMP,^11,21 and 5-fluoroorotidine 5′-monophosphate (FOMP).¹⁰ (b) A restricted study on S154A, Q215A and S154A/Q215A mutants of OMPDC.¹⁵ We report here kinetic parameters for the six additional mutant enzymes required to obtain a complete set of kinetic data for all single, double, triple, and quadruple Q215A, Y217F, R235A, and S154A mutations of ScOMPDC. Our analyses of the kinetic parameter k_cat/K_m for decarboxylation of OMP and FOMP catalyzed by these 16 variants of ScOMPDC (Scheme 3) extends protocol developed in earlier work.^10,11,15 Several lines of analysis are explored in describing the role played by the −CH₂OH side chain of S154 in the efficient functioning of phosphodianion gripper side chains at the transition state for OMPDC-catalyzed decarboxylation.

Scheme 3. Wild-Type ScOMPDC and All Possible Single, Double, Triple and Quadruple Q215A, Y217F, R235A, and S154A Mutants Enzymes.

Asterisks show the new mutant enzymes prepared for this work.

Experimental Section

Materials

OMP(12,22) and FOMP(12,23) were prepared enzymatically by literature procedures. 3-(N-Morpholino)propanesulfonic acid (MOPS) and imidazole were purchased from Sigma (St. Louis, MO). Sodium hydroxide (1.0 N), hydrochloric acid (1.0 N), sodium chloride, tris(hydroxymethyl)-aminomethane (basic form), and Amicon centrifugal filter units with a 10K molecular weight cutoff (MWCO) were purchased from Fisher (Hampton, NH). Nickel chloride hexahydrate was a generous gift from Prof. Andrew Murkin (University at Buffalo, Buffalo, NY). Chelating Sepharose Fast-Flow and Q-Sepharose were purchased from GE Healthcare (Marlborough, MA). Water was purified using a Milli-Q Academic purification system from EMD Millipore (Burlington, MA). All other commercial chemicals were reagent-grade or better and were used without further purification.

Preparation of Wild-Type and Mutant Yeast Orotidine 5′-Monophosphate Decarboxylases

The plasmid pScODC-15b containing the gene encoding wild-type orotidine 5′-monophosphate decarboxylase from Saccharomyces cerevisiae (ScOMPDC) with an N-terminal His₆-tag was available from earlier studies.^15,24 The protein sequence differs from the published sequence for wild-type ScOMPDC by the following mutations: S2H, C155S, A160S, and N267D.^15,24 The kinetic parameters for this preparation of wild-type ScOMPDC are in agreement with those reported by Porter and Short.²⁵ The C155S mutation has been shown to result in a more stable protein.²⁶

Site-directed mutagenesis on pScODC-15b was carried out using the QuikChange II kit from Stratagene (San Diego, CA). The procedures for the preparation of the S154A,¹⁵ Q215A,¹⁵ R235A,¹⁴ Y217F,¹³ S154A/Q215A,¹⁵ Q215A/Y217F,¹³ Q215/R235A,¹³ Y217F/R235A,¹³ and Q215A/Y217F/R235A¹³ mutants were described in earlier work. The following primers, with the altered codons underlined, were used to prepare the new mutant enzymes from previously prepared mutants of ScOMPDC.

5′-GACGCATTGGGTCAACAGTTTAGAACCGTGGATGATGTGG-3′ was used to introduce the Y217F mutation into the plasmid for the S154A mutant enzyme to yield the plasmid for the S154A/Y217F mutant.

5′-GGATCTGACATTATTATTGTTGGAGCAGGACTATTTGCAAAGGG-3′ was used to introduce the R235A mutation into the plasmid for the S154A/Q215A mutant enzyme to yield the plasmid for the S154A/Q215A/R235A mutant.

5′-GGCCTTTTGATGTTAGCAGAATTGGCATCTAAGGGCTCCC-3′ was used to introduce the S154A mutation into the plasmids for the R235A, Q215A/Y217F, Y217F/R235A, and Q215A/Y217F/R235A mutant enzymes to yield the plasmids for the S154A/R235A, S154A/Q215A/Y217F, S154A/Y217F/R235A, and S154A/Q215A/Y217F/R235A mutant enzymes, respectively.

The mutants of ScOMPDC were overexpressed, after transformation of E. coli BL21 (DE3) with the appropriate plasmid. The proteins were purified as described previously,^11,15 except that passage over the Chelating Sepharose Fast-Flow column was repeated a second time to ensure complete removal of host E. coliOMPDC. The fractions from the second Sepharose Fast-Flow column were treated with thrombin to remove the N-terminal His₆-tag, and this tag was separated from the protein by purification over a column of Q-sepharose. Wild-type OMPDC from E. coli (EcOMPDC) was prepared and purified according to published procedures.²⁷

Preparation of Solutions

The solution pH was determined at 25 °C using an Orion Model 720A pH meter equipped with a Radiometer pHC4006–9 combination electrode that was standardized at pH 4.00, 7.00, and 10.00 at 25 °C. Stock solutions of OMP (20 mM) and FOMP (20 mM) were prepared by dissolving the solid nucleotide in water and adding 1 N HCl or 1 N NaOH to adjust the solution to the desired pH. These solutions were stored at −20 °C. The concentration of OMP in 0.1 M HCl was determined from the absorbance at 267 nm using a value of ε = 9430 M^–1 cm^–1.²⁸ The concentration of FOMP in 0.1 M HCl was determined from the absorbance at 271 nm using a value of ε = 10 200 M^–1 cm^–1.⁵

Solutions of MOPS buffers (50% free base) were prepared by addition of measured amounts of 1 N NaOH and solid NaCl to give the desired ratio of acidic and basic buffer forms, and solution ionic strength. Stock solutions (20–40 mg/mL) of mutant ScOMPDC were dialyzed exhaustively against 10 mM MOPS at pH 7.0 (50% free base), 4 °C and I = 0.105 (NaCl). The solutions were further concentrated using Amicon filter units, when necessary. The concentration of ScOMPDC in these solutions was determined from the absorbance at 280 nm for wild type OMPDC, and values of ε = 29 900 or 28 400 M^–1 cm^–1 for Y217F mutant enzymes, that were calculated using the ProtParam tool available on the ExPASy server.^29,30

Kinetic Parameters for Decarboxylation of OMP

The decarboxylation of OMP was monitored spectrophotometrically by following the decrease in absorbance at 279, 290, 295, or 300 nm [Δε, λ: – 2400 M^–1 cm^–1, 279 nm; – 1620 M^–1 cm^–1, 290 nm; – 840 M^–1 cm^–1, 295 nm; – 344 M^–1 cm^–1, 300 nm]. The wavelength was chosen to give an initial absorbance of ≤2.0. These reactions were monitored at the following wavelengths: [OMP] ≤ 0.08 mM, 279 nm; [OMP] = 0.1–0.3 mM, 290 nm; [OMP] = 0.3–0.8 mM, 295 nm; [OMP] ≥ 1 mM, 300 nm. Assays for mutant ScOMPDC-catalyzed decarboxylation of OMP were at 25 °C, pH 7.1 (10 mM MOPS) and I = 0.105 (NaCl). Reactions (1.0 mL) were initiated by addition of 10–60 μL of a stock solution of mutant ScOMPDC to give the following concentrations of mutant enzyme: 0.025 mM, S154A/Y217F mutant; 0.2 mM, S154A/R235A mutant; 0.1 mM, S154A/Q215A/Y217F mutant; and 0.2 mM, S154A/Q215A/R235A mutant. The initial velocity v (M s^–1) for mutant ScOMPDC-catalyzed decarboxylation of OMP (0.03–3.0 mM) was determined by monitoring the decrease in absorbance at the chosen wavelength for the decarboxylation of 5–10% of total OMP. Values of k_cat/K_m (M^–1 s^–1) were obtained from the nonlinear least-squares fits of four or more values of v/[E] (s^–1) to the appropriate kinetic equation.

Kinetic Parameters for Decarboxylation of FOMP

Assays for wild-type EcOMPDC-catalyzed decarboxylation of FOMP were conducted at 25 °C and pH 7.1 (10 mM MOPS) and at I = 0.105 (NaCl). Reactions (1.0 mL) were initiated by the addition of 2–4 μL of a stock solution of wild-type EcOMPDC to give a final enzyme concentration of 5–10 nM. The initial velocity v (M s^–1) for wild-type EcOMPDC-catalyzed decarboxylation of FOMP (10–200 μM) was determined by monitoring the decrease in absorbance at 282 nm [Δε = −2400 M^–1 cm^–1] for the decarboxylation of 5–10% total FOMP. The value of the kinetic parameter k_cat/K_m (M^–1 s^–1) was obtained as the slope of the linear plot of values of v/[E] (s^–1) against [FOMP].

Mutant ScOMPDC-catalyzed decarboxylation of FOMP was monitored spectrophotometrically over 5–10 half-reaction times and working at a wavelength (282 or 300 nm) where the total initial absorbance for FOMP and mutant ScOMPDC is ≤2.0. Time courses for mutant ScOMPDC-catalyzed decarboxylation of FOMP were obtained at 25 °C and pH 7.1 (10 mM MOPS) and at I = 0.105 (NaCl). Reactions (1.0 mL) containing 60–150 μM FOMP were initiated by addition of 3–40 μL of a stock solution of the mutant ScOMPDC to give the following enzyme concentrations: 3 μM, S154A/Y217F mutant; 20 μM, S154A/R235A mutant; 10 μM, S154A/Q215A/Y217F mutant; 5–10 μM, S154A/Q215A/R235A mutant; 0.2 mM, S154A/Y217F/R235A mutant; and 0.1 mM, S154A/Q215A/Y217F/R235A mutant. The rate constants k_obs (s^–1) for the pseudo-first-order decay of FOMP to FUMP were obtained by fitting the plot of absorbance A against time to the equation for an exponential decay. Values of the second-order rate constant k_cat/K_m (M^–1 s^–1) for ScOMPDC-catalyzed decarboxylation of FOMP were then calculated from the relationship k_cat/K_m= k_obs/[E].

Results

Figure S1 shows the linear plot of v/[E] against [FOMP] for EcOMPDC-catalyzed decarboxylation of FOMP, with the slope of k_cat/K_m= 2.7 × 10⁶ M^–1 s^–1. The observation of a linear plot through [FOMP] = 0.2 mM shows that K_m ≫ 0.2 mM. By comparison, values of k_cat = 14 s^–1 and k_cat/K_m= 6.4 × 10⁵ M^–1 s^–1 were reported for EcOMPDC-catalyzed decarboxylation of OMP.²⁷ We conclude that the C-5 fluorine of FOMP results in a 4-fold increase in k_cat/K_m for EcOMPDC-catalyzed decarboxylation.

Figure 2A shows plots of v/[E] against [OMP] for decarboxylation catalyzed by the S154A/Q215A/Y217F (green symbols) and S154A/R235A (purple symbols) mutants of ScOMPDC, and Figure 2B show similar plots for the S154A/Y217F (maroon symbols) and S154A/Q215A/R235A (blue symbols) mutants. The solid lines in Figure 2 show the fits of these experimental data to the full Michaelis–Menten equation (Figure 2A) or the linear form of this equation, where K_m ≫ [OMP] (Figure 2B). The kinetic parameters from Figure 2A require that <10% of total OMP is tied up as the Michaelis complex (Table S1). The concentration of free OMP is therefore essentially equal to total [OMP] (Figure 2A).

Figure 2.

Dependence of v/[E] (s^–1) for decarboxylation of OMP catalyzed by mutant forms of ScOMPDC on the concentration of OMP for reactions at 25 °C, pH 7.1 (10 mM MOPS) and at I = 0.105 (NaCl). (A) Green symbols, S154A/Q215A/Y217F mutant; purple symbols, S154A/R235A mutant. (B) Maroon symbols, S154A/Y217F mutant; blue symbols, S154A/Q215A/R235A mutant.

Figure 3 shows time courses for the change in absorbance at 300 nm for decarboxylation of FOMP catalyzed by S154A/Y217F/R235A (Figure 3A) and S154A/Q215A/Y217F/R235A mutants of ScOMPDC (Figure 3B). Figure S2A–D shows time-courses for the change in absorbance at 282 nm for decarboxylation of FOMP catalyzed by the S154A/Y217F (Figure S2A), S154A/R235A (Figure S2B), S154A/Q215/Y217F (Figure S2C), and the S154A/Q215A/R235A (Figure S2D) mutants of ScOMPDC. The fits of the experimental data for each Figure to the exponential decay for the pseudo-first-order enzyme-catalyzed reaction of FOMP gives the observed first order rate constant k_obs (s^–1) for ScOMPDC-catalyzed decarboxylation. The fits to a simple exponential decay show that these mutant ScOMPDC-catalyzed decarboxylation reactions are first order in [FOMP] (0.06–0.15 mM ≪ K_m). Each of the experiments from Figures 3 and S2 were repeated two or more times, and the values of k_obs/[E] for each mutant ScOMPDC-catalyzed decarboxylation were found to be reproducible to better than ±10%.

Figure 3.

Full time-course for the decarboxylation of FOMP catalyzed by mutants of ScOMPDC for reactions at 25 °C, pH 7.1 (10 mM MOPS) and at I = 0.105 (NaCl). (A) Decarboxylation of 0.10 mM FOMP catalyzed by 0.20 mM of the S154A/Y217F/R235A mutant of ScOMPDC. (B) Decarboxylation of 0.10 mM FOMP catalyzed by 0.10 mM of the S154A/Q215A/Y217F/R235A mutant of ScOMPDC.

The new values of k_cat/K_m for decarboxylation of OMP catalyzed by mutant forms of ScOMPDC, which were obtained from the fits of data shown in Figure 2 to the appropriate kinetic equation, are shown in bold type in Table 1. The fits from Figure 2A also gave values of K_m= 2.0 ± 0.2 mM for the S154A/R235A mutant and 1.5 ± 0.05 mM for the S154A/Q215A/Y217F mutant. Table 1 reports values of k_cat/K_m determined in earlier work for decarboxylation of OMP catalyzed by wild-type ScOMPDC and mutants of ScOMPDC.^11,15 The new values of k_cat/K_m for decarboxylation of FOMP catalyzed by mutant forms of ScOMPDC, obtained from the fits of data shown in Figures 3 and S1, are shown in bold type Table 1. Table 1 reports values of k_cat/K_m determined in earlier work for decarboxylation of FOMP catalyzed by wild-type ScOMPDC and mutants of ScOMPDC, to give data for all possible single, double, triple, and quadruple S154A, Q215A, Y217F, and R235A mutations.¹⁰Table 1 also reports rate constant ratios for wild-type and mutant ScOMPDC-catalyzed decarboxylation of FOMP and OMP.

Table 1. Kinetic Parameters for ScOMPDC-Catalyzed Decarboxylation of OMP and FOMP^a.

	OMPb		FOMPc
ScOMPDC	kcat/Km (M–1 s–1)	ΔΔG⧧ (kcal/mol)d	kcat/Km (M–1 s–1)	ΔΔG⧧ (kcal/mol)d	e
wild-type	1.1 × 107		1.2 × 107		1.1
Q215A	2.6 × 105	2.2	2.0 × 106	1.1	7.7
Y217F	1.8 × 105	2.4	1.1 × 106	1.4	6.1
R235A	910	5.6	1.6 × 105	2.6	180
S154A	630	5.8	2.9 × 105	2.2	460
S154A/Q215A	380	6.1	7.7 × 104	3.0	200
S154A/Y217F	0.75 ± 0.01	9.8	520 ± 10	5.9	700
S154A/R235A	0.027 ± 0.004	11.7	20 ± 2	7.9	740
Q215A/Y217F	3.4 × 103	4.8	5.3 × 104	3.2	16
Q215A/R235A	14	8.0	7200	4.4	510
Y217F/R235A	4.1	8.8	820	5.7	200
S154A/Q215A/Y217F	0.28 ± 0.02	10.4	250 ± 10	6.4	890
S154A/Q215A/R235A	0.018 ± 0.001	12.0	10 ± 1	8.3	560
S154A/Y217F/R235A	0.0006 ± 0.0002f	14.0	0.44 ± 0.02	10.1
Q215A/Y217F/R235A	0.037	11.6	28	7.7	760
S154A/Q215A/Y217F/R235A	0.0016 ± 0.0005f	13.4	1.1 ± 0.1	9.6

^aConditions: pH 7.1 (10 mM MOPS), 25 °C and I = 0.105 (NaCl).

^bThe new values for this manuscript are in bold type. The other rate constants are from ref (11) or (15). The quoted uncertainty is the standard error from the least-squares fit of the kinetic data to the appropriate kinetic equation.

^cThe new values for this manuscript are in bold type. The other rate constants are from ref (10).The quoted uncertainty in the original values is the average of two or more determinations of k_cat/K_m.

^dCalculated from the ratio of the values of k_cat/K_m for wild-type and mutant ScOMPDC-catalyzed decarboxylation of OMP or FOMP.

^eThe ratio of the values of k_cat/K_m for OMPDC-catalyzed decarboxylation of OMP and FOMP.

^fNo detectable activity toward decarboxylation of OMP: k_cat/K_m < 0.006 M^–1 s^–1. The rate constants were estimated from k_cat/K_m for mutant enzyme-catalyzed decarboxylation of FOMP (see text). The range of values is calculated from the estimated uncertainty of ±50% in for OMPDC-catalyzed decarboxylation of OMP and FOMP.

There was no detectable decarboxylation of 1.0 mM OMP ([ΔA₃₀₀ ≤ 0.03) catalyzed by 0.2 mM of the S154A/Y217F/R235A or S154A/Q215A/Y217F/R235A mutants over a 20 h reaction time. This shows that k_cat/K_m < 0.006 M^–1 s^–1 for the mutant enzyme-catalyzed decarboxylation reactions. Large limiting ratios are observed for reactions catalyzed by severely impaired mutants of ScOMPDC, where the relative reaction barriers are determined by chemical reactivity of enzyme-bound OMP and FOMP.¹⁰ The ratio of = 700 ± 200 is the average of the ratios from Table 1 for severely impaired mutants where (k_cat/K_m)_H ≤ 1.0 M^–1 s^–1 (S154A/Y217F, S154A/R235A, S154A/Q215A/Y217F, S154A/Q215A/R235A, and Q215A/Y217F/R235A). The values of (k_cat/K_m)_H for decarboxylation catalyzed by the S154A/Y217F/R235A and S154A/Q215A/Y217F/R235A mutants, shown in underlined type in Table 1, were estimated from the value (k_cat/K_m)_F for decarboxylation of FOMP and the rate constant ratio = 700 ± 200.

Discussion

All of the newly reported mutants of ScOMPDC were expressed in E. Coli that contained host EcOMPDC. These mutants were prepared with an N-terminal His₆-tag, and with two passes over a chelating column to ensure complete removal of EcOMPDC host enzyme. The value of > 67 (Table 1) for the most severely impaired S154A/Y217F/R235A mutant, which shows no activity toward decarboxylation of OMP, is much larger than the value for EcOMPDC, = 4. This shows that there is no significant (<10%) decarboxylation of FOMP by wild-type EcOMPDC from the host organism. If, for example, 10% of the observed activity of this preparation toward decarboxylation of FOMP were due to wild-type host OMPDC, then a detectable value of = (1.00)/[0.10][0.25] = 40 would have been observed.

The <2-fold variation in the 5-F substituent effect on decarboxylation reactions catalyzed by severely impaired mutants of ScOMPDC [(k_cat/K_m)_OMP ≤ 1.0 M^–1 s^–1, = 560–890) shows that the value of (k_cat/K_m)_F for these mutant enzymes establishes (k_cat/K_m)_H to within a factor of 2. The rate constants (k_cat/K_m)_H for decarboxylation reactions catalyzed by the S154A/Y217F/R235A and S154A/Q215A/Y217F/R235A mutants, that were too slow to be detected [(k_cat/K_m)_H < 0.006 M^–1 s^–1], were estimated from the value of (k_cat/K_m)_F (Table 1) and = 700.

This study of the effect of mutations of S154, Q215, Y217, and R235 will provide a clear description of the role of this network in catalysis, provided the effects are restricted to the elimination of interactions between the transition state and the excised side chain(s). The following observations provide strong evidence that there only minimal effects of these mutations on [(k_cat/K_m)_H that arise from protein structural changes that perturb protein-transition state interactions. (1) Single Q215A, Y217F, and R235A mutations result in negligible (≤0.5 kcal/mol) changes in the activation barrier to OMPDC-catalyzed decarboxylation of the phosphodianion truncated substrate 1-(β-erythrofuranosyl)orotic acid (EO, Scheme 4).¹³ This provides strong evidence that the single mutations have no effect on the structural integrity and catalytic activity at the site of decarboxylation of the pyrimidine ring, and are restricted to the loss of protein dianion interactions at the dianion activation site. (2) The sum of the effects of the four single S154A, Q215A, Y217F, and R235A mutations on the activation barrier ΔG^⧧ for decarboxylation catalyzed by wild-type OMPDC is 16.0 kcal/mol (Table 1). By comparison, there is only a 13.4 kcal/mol difference between the activation barriers ΔG^⧧ for wild type and the S154A/Q215A/Y217F/R235A quadruple mutant enzyme-catalyzed decarboxylation of OMP (Table 1) This shows that there can be no large additional increase in the activation barrier of OMPDC-catalyzed decarboxylation associated with the simultaneous mutations of the S154A, Q215A, Y217F, and R235A, compared with the mutation of single side chains. (3) Q215A mutations result in a ca. 2 kcal/mol destabilization of the transition state for decarboxylation of OMP when S154 is present and in a smaller ca. 0.3 kcal/mol destabilization when S154 is absent, as discussed in detail below. The exception is the Q215A mutation of the S154A/Y217F/R235A variant, which results in a ca. 0.5 kcal/mol stabilization or the transition state for ScOMPDC-catalyzed decarboxylation of FOMP. This is consistent with a shift in the position of the Q215 side chain at the S154A/Y217F/R235A triple mutant, which alters the interaction of the protein with the decarboxylation transition state. Aside from this minor effect, we are able to rationalize the other effects of S154A, Q215A, Y217F, and R235A mutations by a consideration of the loss in transition state stabilization from interaction with the deleted side chains.

Scheme 4. Stepwise OMPDC-Catalyzed Decarboxylation of Phosphodianion Truncated Substrate EO.

Mutant Boxes

We begin by examining two-dimensional mutant cycles (Figure 4) constructed from data in Table 1. These cycles show the effects of mutations of S154, which links the pyrimidine umbrella and phosphodianion gripper loops (Figure 1), on the stabilizing interactions between individual phosphodianion gripper side chains Q215, Y217, and R235 and the decarboxylation transition state (Figure 5).

Figure 4.

Mutant cycles that show the effect of S154A mutations on: (A) stabilization of the transition state for wild-type ScOMPDC-catalyzed decarboxylation of OMP by interactions with Q215; (B) stabilization of the transition state for wild-type ScOMPDC-catalyzed decarboxylation of OMP by interactions with Y217F; (C) stabilization of the transition state for Y217F ScOMPDC-catalyzed decarboxylation of OMP by interactions with R235; (D) stabilization of the transition state for wild-type ScOMPDC-catalyzed decarboxylation of OMP by interactions with R235A; (E) stabilization of the transition state for R235A ScOMPDC-catalyzed decarboxylation of OMP by interactions with Y217F.

Figure 5.

Representation of the X-ray crystal structure of a complex between OMPDC and the intermediate analog 6-hydroxyuridine 5′-monophosphate (BMP, PDB entry 1DQX). Phosphodianion gripper loop is shaded blue, and the pyrimidine umbrella loop is shaded green. These loops are clamped over the substrate by the following network of hydrogen bonding interactions: (a) H-bonds between the Q215, Y217, and R235 side chains and the ligand phosphodianion; (b) H-bonds that link the −CH₂OH side chain of S154 to the amide side chain of Q215 and to the pyrimidine nitrogen. All distances are in Ångstroms.

Q215 (Figure 4A)

The Q215A mutation results in a 2.2 kcal/mol destabilization of the transition state for wild-type ScOMPDC-catalyzed decarboxylation of OMP, but the same mutation of the S154A mutant enzyme results in a much smaller 0.3 kcal/mol transition state destabilization.¹⁵ This provides strong evidence that a hydrogen bond between the −CH₂OH group of S154 and the amide side-chain of Q215 (Figure 1) positions the Q215 side to interact with the phosphodianion of OMP.

Y217 (Figure 4B,C)

The Y217F mutation results in a 2.4 kcal/mol destabilization of the transition state for wild-type OMPDC-catalyzed decarboxylation of OMP, but the same mutation of S154A mutant ScOMPDC results in a larger 4.0 kcal/mol destabilization (Figure 4B). We conclude that the H-bond between the CH₂OH group of S154 and the amide side-chain of Q215 at wild-type ScOMPDC fixes the position of Y217 and R235 at the phosphodianion. This clamp is largely maintained at the Y217F mutant, where the effect is limited to the loss of the interactions of the single side chain. The clamp is released by the S154A mutation, so now the second Y217F mutation results in both the loss of the interactions to Y217 and in a weakening of the interactions to R235. The R235A mutation of Y217F ScOMPDC results in a 6.4 kcal/mol transition state destabilization, but the same mutation of the S154A/Y217F mutant results in a smaller 4.2 kcal/mol transition state destabilization (Figure 4C). This provides direct evidence that the double S154A/Y217F mutation leads to a weakening of the transition state stabilization by the remaining R235 side chain.

Interaction of R235 (Figure 4D,E)

The R235A mutation results in a 5.6 kcal/mol destabilization of the transition state for wild-type ScOMPDC-catalyzed decarboxylation of OMP, but the same mutation of the S154A mutant enzyme results in a larger 5.9 kcal/mol transition state destabilization (Figure 4D), consistent with conclusion that the R235A mutation of S154A ScOMPDC results in the loss of the interaction of the mutated R235 side chain, and a small weakening of the interaction with Y217F. The Y217F mutation of R235A ScOMPDC results in a 3.2 kcal/mol kcal/mol transition state destabilization, but the same mutation of the S154A/R235A mutant results in only an estimated 2.3 kcal/mol destabilization (Figure 4E). This suggests that the double S154A/R235A mutation leads to a weakening of the transition state stabilization by the remaining Y217 side chain.

In conclusion, S154 plays both a specific and a general role in promoting interactions between OMPDC and dianion gripper side chains. The −CH₂OH side chain plays the specific role of holding the amide side chain of Q215 in a position to interact with the substrate dianion. The side chain acts globally to optimize interactions with Y217 and R235 by locking the substrate in the protein cage created by the interactions between the pyrimidine umbrella and phosphodianion gripper side chains (Figures 1 and 5).

Reconstruction of Wild-Type OMPDC

Table 2 shows the effects of single S154A Q215A, Y217F, and R235A mutations on (ΔG^⧧)_OMP for decarboxylation of OMP determined for mutations at wild-type ScOMPDC and various single, double, and triple mutants of ScOMPDC. Table 2 defines a total of 32 different effects of single mutations of S154A Q215A, Y217F, and R235A on the activity of wild-type and variant ScOMPDC. We focus here on the recovery in enzymatic activity observed upon substitution of wild-type amino acid side chains at the quadruple S154A/Q215A/Y217F/R235A mutant.

Table 2. Effect of Single Substitutions of Amino Acid Residues X on the Activation Barrier (ΔG^⧧)_OMP for Decarboxylation of OMP Catalyzed by Wild-Type ScOMPDC and by Previously Mutated ScOMPDC^a.

	(ΔΔG⧧X)OMP (kcal/mol)b
enzyme	X = Q215A	X = Y217F	X = R235A	X = S154A	Σ(ΔΔG⧧X)OMPc
wild typed	2.2	2.4	5.6	5.8	16.0
triple mutante	–0.6	1.4	3.0	1.8	5.6
single mutantf	2.4 (Y217F)	2.6 (Q215A)	5.8 (Q215A)	3.9 (Q215A)
	2.4 (R235A)	3.2 (R235A)	6.4 (Y217F)	7.4 (Y217F)
	0.3 (S154A)	4.0 (S154A)	5.9 (S154A)	6.1 (R235A)
double mutantg	2.8 (Y217F/R235A)	3.6 (Q215A/R235A)	6.8 (Q215A/Y217F)	5.6 (Q215A/Y217F)
	0.6 (S154A/Y217F)	4.3 (S154A/Q215A)	5.9 (S154A/Q215A)	4.0 (Q215A/R235A)
	0.3 (S154A/R235A)	2.3 (S154A/R235A)	4.2 (S154A/Y217F)	5.2 (Y217F/R235A)

^aFor OMPDC-catalyzed decarboxylation at 25 °C, pH 7.1 (10 mM MOPS), and ionic strength of 0.105 (NaCl). The respective substitutions that cause the largest and smallest changes in (ΔG^⧧)_OMP are shown in underlined and bold text, respectively.

^bCalculated from the ratio of values of k_cat/K_m for the precursor and mutant enzymes.

^cSum of the effects of the S154A, Q215A, Y217F, and R235A substitutions on (ΔΔG_X^⧧)_OMP.

^dEffect of the first mutation X on the activation barrier for wild-type OMPDC-catalyzed decarboxylation of OMP.

^eEffect of the final mutation X on the activation barrier for the decarboxylation reaction catalyzed by the corresponding triple mutant of OMPDC.

^fEffect of the mutation X on the activation barrier for the decarboxylation reaction catalyzed by the OMPDC mutant given in parentheses.

^gEffect of the mutation X on the activation barrier for the decarboxylation reaction catalyzed by the OMPDC double mutant given in parentheses.

The substitution of single wild-type amino acids at the quadruple mutant to give triple mutants has only small effects on the stability of the decarboxylation transition state. These are shown as the bold entries in Table 2. For example, the 3.0 and 1.8 kcal/mol effects of substitution of S154 and R235 are much smaller than the 5.6 and 5.8 kcal/mol effects of these substitutions at the S154A and R235A mutant enzymes. Consequently, the sum of the effects of single S154A, Q215A, Y217F, and R235A substitutions at triple mutants of OMPDC, 5.6 kcal/mol (Table 2) is only 35% of the sum of the effects of these deletions at wild-type ScOMPDC (16.0 kcal/mol). The small restorative effects of single substitutions at the S154A/Q215A/Y217F/R235A quadruple mutant are consistent with a large entropic cost to holding OMPDC in the active closed conformation. We propose that this represents the cost of fixing the positions of the pyrimidine umbrella and phosphodianion gripper loops (Figure 1) relative to the transition state.²⁷ In every case, a large fraction of the total binding energy of the single side chain at the triple mutants is used to pay this entropic cost, so only a small fraction remains for expression as transition state stabilization.

By contrast, the largest effects of A215Q, F217Y, and A235R substitutions on (ΔG^⧧)_OMP (underlined text in Table 2) are observed when these side chains are incorporated into triple mutants of OMPDC to give the corresponding double mutant (bottom row, Table 2). For example, the A235R substitution at the Q215A/Y217F/R235A triple mutant provides for a 6.8 kcal/mol decrease in (ΔG^⧧)_OMP compared with the smaller 5.6 kcal/mol effect of the A235R substitution at the R235A single mutant; and, the F217Y substitution at the S154A/Q215A/Y217F triple mutant provides a 4.3 kcal/mol decrease in (ΔG^⧧)_OMP compared with the smaller 2.4 kcal/mol effect of the F217Y substitution at the F217Y single mutant (Table 2).

The 6.8 kcal/mol effect of the A235R substitution at Q215A/Y217F/R235A ScOMPDC on (ΔG^⧧)_OMP represents the stabilizing interaction of the R235 side chain, plus an improvement in the functioning of S154 already present at the triple mutant. This cooperativity between the expression of side chain interactions reflects the roughly constant entropic cost for immobilizing the phosphodianion gripper and pyrimidine gripper loops at the active form of ScOMPDC. This full price is paid by single amino acids added to the quadruple mutant, but it is split when two amino acids are added to the quadruple mutant, with the result that a larger fraction of the total binding energy of each side chain is expressed as stabilization of the decarboxylation transition state at the double mutant. An exception to this trend is the large 7.4 kcal/mol effect of the A154S substitution at the S154A/Y217F mutant shown in Figure 4B. This reflects the key role played by the clamping −CH₂OH side chain of S154 in optimizing protein dianion interactions, and the large effect of the S154A mutation on the interaction of the remaining R235 side chains with the substrate dianion, as discussed above.

Smaller effects of the Q215, Y217, and R235 substitutions on (ΔG^⧧)_OMP are observed when these side chains are inserted at double mutants of ScOMPDC to give the single mutant (third row, Table 2), compared with their insertions at triple mutants to give the double mutants (bottom row, Table 2). The effects of these substitutions are then only slightly reduced when they result in wild-type ScOMPDC. We propose that the interactions between any two of four side chains results in a large fractional reduction of entropy associated with loop motion and that the further entropic penalty paid upon substitution of the third side chain is small and upon substitution of the fourth side chain is negligible. Such effects are well-documented for the formation of intermolecular supramolecular complexes that are stabilized linear arrays of hydrogen bonds. The supramolecular complex stabilized by a single hydrogen bond is relatively weak, due to the large entropic penalty to the reaction between monomers to form the dimer. A larger fraction of the total stabilization from subsequent H-bonds is expressed at complexes stabilized by more than one hydrogen bond, because of the relatively small effect of each new H-bond on the total entropy of the supramolecular complex.³¹

Cubic Mutant Cycles

The data from Table 2 were used to construct the cubes shown in Figure 6. These show 24 of the 32 different effects of single substitutions from Table 2. All 32 effects may be shown on a pseudo-four-dimensional hypercube, obtained by placing the cube from Figure 6B at the center of a larger cube for Figure 6A, and drawing eight lines to connect the corners of the two cubes.³² The last eight connections are not shown on Figure 6A,B, but several are given and discussed in our analysis of Figure 4. This hypercube is composed of eight distinct cubic cells, two of which are shown in Figure 6A,B.

Figure 6.

Triple mutant cubes that show the effects of single amino acid mutations on (ΔG^⧧)_OMP (kcal/mol)for reactions catalyzed by wild-type ScOMPDC (black values), by single mutants of ScOMPDC (red values), and by double mutants of ScOMPDC (green values). (A) Effect of these mutations on wild-type ScOMPDC. (B) Effect of mutations on ScOMPDC, previously mutated at position S154. The values of (ΔΔG_X^⧧)_OMP were calculated from the ratio of the kinetic parameters k_cat/K_m for precursor and mutated ScOMPDC-catalyzed decarboxylation of OMP (Table 1) and are reported in Table 2.

Figure 6A is the previously reported cube for the effects of Q215A, Y217F, and R235A mutations on (ΔG^⧧)_OMP for wild-type ScOMPDC-catalyzed decarboxylation of OMP.¹¹ The effect of these same mutations on (ΔG^⧧)_OMP for ScOMPDC previously mutated at S154 is illustrated by Figure 6B. We concluded earlier that the effect of mutations of residue X = Q215, Y217, or R235 on (ΔΔG_X^⧧)_OMP is approximately the same when the mutation is carried out at wild-type OMPDC or singly mutated or doubly mutated forms of OMPDC.¹¹ This shows that when S154 is present at ScOMPDC, the loss of transition state stabilization from replacement of one or two side chains does not result in large changes in the stabilization by the remaining side chain(s).

By contrast, the effect of single mutations of X = Q215, Y217, or R235 on (ΔΔG_X^⧧)_OMP is dependent on the context of the mutation, when the S154A mutant is the parent enzyme (Figure 6B), so that the presence of S154 is required for the regular behavior observed for Figure 6A. Figure 6B may be used to illustrate points made above in the discussion of Table 2. For example, the smallest changes in (ΔΔG_X)_OMP (shown in green) are observed for single mutations of triple mutants to form the quadruple mutant; while, significantly larger changes in (ΔΔG_X^⧧)_OMP (shown in red) are observed for the same mutations at double mutants to form the corresponding triple mutant.

The effects of consecutive Q215A, Y217F, and R235A mutations sum to 11.6 kcal/mol, when starting from wild-type ScOMPDC (Figure 6A) but to only 7.6 kcal/mol when starting from S154A mutant ScOMPDC (Figure 6B). This shows that the S154A mutation results in a total (11.6–7.6) = 4.0 kcal/mol decrease in transition state stabilization from the collective interactions with the phosphodianion gripper side chain Q215, Y217, and R235. One-half (2 kcal/mol) of this decrease is due to the loss of interactions with the side chain of Q215 (Figure 4A),¹⁵ and 2 kcal/mol results from a weakening in the interactions with Y217 and R235 (Figure 4B,C). We attribute the remaining 2 kcal/mol of the total 5.8 kcal/mol effect of the S154A mutation to the loss of the stabilizing H-bonding interaction between the S154 oxygen and the pyrimidine −NH. By comparison, the loss of this H-bonding interaction results in a 3 kcal/mol destabilization of the transition state for OMPDC-catalyzed decarboxylation of the phosphodianion truncated substrate EO (Scheme 4).³³

Summary and Conclusions

This is the first report of an examination of the catalytic role of a network of four interacting amino acid side chains, by determining the effects of stepwise replacement of all four side chains. The four mutations result in a total (2 × 10¹⁰)-fold reduction in the activity of wild-type ScOMPDC, which corresponds to the loss of 45% of the 31 kcal/mol intrinsic substrate binding energy.^3,34 The modeling of these effects of site-directed mutagenesis is a significant challenge to existing computational methods,³⁵ similar to that for modeling the effects of mutagenesis studies on the activity of triosephosphate isomerase.³⁶⁻⁴³

The relative ease of the interpretation of these experimental results reflects the choice of mobile structural elements that lie on the protein surface. A similar study is less likely to provide insight into the function of amino acid networks at the interior of protein catalysts, because of unintended structural effects of these mutations on the function of other active site catalytic side chains. The utilization of amino acid side chains, at mobile structural elements, for stabilization of enzymatic transition states is less efficient than the utilization of rigid, preorganized, side chains,^44,45 because of the entropic penalty to freezing the mobile elements at the transition state. However, flexible elements are often required to allow for the entrance and departure of ligands from active site cages.^17,46 We find that the entropic penalty to fixing the active site loops of ScOMPDC (Figure 5) is a significant fraction of the interaction energy of a single side chain. This penalty is difficult to detect, and therefore easily underestimated, when it is divided between four amino acid side chains.

The mutant cubes from Figure 6A,B show that the single intraloop H-bond between the side chains of Q215 and S154 (Figure 5) has the effect of increasing the total transition state stabilization from the Q215, Y217, and R235 side chains by ca. 4 kcal/mol. These results provide a telling example of the importance of the precise placement of side chains in obtaining efficient enzymatic catalysis⁴⁸ and illustrates the magnitude of the problems that must be solved by protein engineers who are working to design proteins with enzyme-like activity.

These results are consistent with the notion that the large catalytic rate acceleration by ScOMPDC is the result of many additive electrostatic interactions between the enzyme and decarboxylation transition state at a preorganized enzyme active site,^44,45,49 with two wrinkles: (1) Several amino acid side chains are not preorganized at the unliganded enzyme, but only organize after the binding of substrate, through the utilization of enzyme-dianion interactions to hold ScOMPDC in a reactive caged Michaelis complex.^1,4,9,17 (2) The conversion of OMPDC from a flexible open form to a stiff and organized Michaelis complex results in the loss of most or all of the entropy associated with the flexible loops at the unliganded enzyme. The intrinsic substrate binding energy is utilized to pay the price for this reduction in entropy.³⁴

Supporting Information Available

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/jacs.8b09609.

• Calculation of the concentration of OMP present in the Michaelis complex (E·OMP) for the initial velocity determinations; linear plot of v/[E] against [FOMP] for EcOMPDC-catalyzed decarboxylation of FOMP; plots of absorbance at 282 nm against time for decarboxylation of FOMP catalyzed by the S154A/Y217F, S154A/R235A, S154A/Q215A/Y217F, and the S154A/Q215A/R235A mutants of ScOMPDC (PDF)

The authors declare no competing financial interest.