Monovalent cation activation and kinetic mechanism of inosine 5'-monophosphate dehydrogenase.

Human type II inosine 5′-monophosphate dehydrogenase has been purified to homogeneity from an Escherichia coli strain that express large quantities of the enzyme from the cloned gene. Steady state kinetic studies have been used to characterize the activation by monovalent cations, including Li, Na, K, Rb, Cs, Tl, NH, and N(CH). The enzyme has less than 1% of the maximal activity in the absence of an added monovalent cation, such as K, Na, Rb, Tl, or NH. The enzyme is activated by K and Tl at lower concentrations than those of other monovalent cations. Li and N(CH) do not activate the enzyme, nor do they inhibit the K-activated enzyme, implying that ionic radius is important in binding selectivity. The K values for both substrates and V differ with different monovalent cations. Initial velocity and product inhibition kinetic data are consistent with an ordered steady state mechanism in which the enzyme binds K first, IMP second, and then NAD; the product NADH is released before xanthosine 5′-monophosphate. Substrate and product binding experiments support this mechanism and show the presence of one substrate binding site per subunit. Several rate constants were obtained from a computer simulation of the complete steady state rate equation.

Isoforms of IMPDH have been purified from a variety of eukaryotic and prokaryotic sources; all are tetrameric with subunit molecular weights of ϳ56 kDa (10 -14). Both IMPDH-h1 and IMPDH-h2 subunits have 514 amino acids, and they have 84% sequence identity (6). IMPDH-h2 is inactivated by 6-chloro-purine ribotide, which reacts with Cys-331 in the IMP binding site, providing the only evidence for the location of the IMP binding site (6,15,16). A monovalent cation, such as K ϩ , Na ϩ , Rb ϩ , NH 4 ϩ , or Tl ϩ , is required for maximal activity, but the mechanism and extent of the activation are unknown. Monovalent cations of different radii show different abilities to activate the enzymes from Bacillus subtilis and sarcoma 180 cells (17,18). An absolute requirement for a monovalent cation activator has not yet been demonstrated, perhaps due to the use of sodium or potassium salts of substrates or other components in assay solutions, and thus 10% or more of the maximum activities have been reported in experiments in which no monovalent cations were intentionally added (17,18).
A steady state ordered sequential Bi Bi mechanism in which IMP binds before NAD and XMP is released after NADH is commonly used to describe the IMPDH-catalyzed reaction (19). Because this mechanism does not include the monovalent cation, it is not complete. Another partially random rapid equilibrium mechanism including K ϩ , IMP, and NAD was proposed by Morrison and co-workers (20) for the enzyme from Aerobacter aerogenes. In this mechanism, K ϩ and IMP are proposed to bind randomly to the enzyme, whereas NAD does not bind unless K ϩ or both K ϩ and IMP are bound.
In this study, we constructed an Escherichia coli expression system for IMPDH-h2 and purified the recombinant protein to homogeneity. The activation of the recombinant IMPDH-h2 by various monovalent cations has been characterized and a larger than 100-fold activation by a monovalent cation, such as K ϩ , was found. A steady state kinetic mechanism including the monovalent cation activation of IMPDH-h2 is proposed based on initial velocity and product inhibition studies. Substrate and product equilibrium binding experiments were performed to directly determine binding affinities and to provide further evidence supporting the proposed mechanism. EXPERIMENTAL PROCEDURES NAD, NADH, alkaline phosphatase, and the free acid of IMP were purchased from Sigma. The sodium salt of XMP (Sigma) was converted into an XMP⅐Tris ϩ salt by passage through a Dowex 50⅐Tris ϩ cation exchange column. Elemental analyses were performed by the inductively coupled argon plasma method at the chemical analysis laboratory of the University of Georgia. The elements analyzed included potassium, sodium, zinc, copper, manganese, iron, magnesium, and phosphorus. All pH values, except where otherwise specified, were measured at room temperature.
Construction of the IMPDH-h2 Expression System-A cDNA clone containing the IMPDH-h2 was a generous gift from Dr. Frank Collart, Argonne National Laboratory (14). The expression system was constructed by introducing new restriction sites by polymerase chain reaction at both ends of the coding region; the cDNA was used as template, and two synthetic oligonucleotides, 5Ј-ACTAGTGTCCTGTGTTG-CATATGGCCGACTAC, which introduced a NdeI restriction site at the ATG start codon, and 5Ј-ACTAGTGGAGGTGTGCTAGATCTCTTT-TCAG, which introduced a BglII restriction site 14 base pairs past the end of the coding region, were used as primers.
The 1.5-kilobase pair fragment obtained from the polymerase chain reaction was isolated from 1% agarose gel (21) and digested with NdeI and BglII restriction enzymes according to the manufacturer's recommendations (BioLabs). The resulting fragment was ligated into the NdeI-BamHI sites of the pET12B vector (from Novagen), transformed into the strain MV1190, and plated on LB-Ampicillin media with isopropyl-1-thio-␤-D-galactopyranoside and X-Gal added. Plasmids from transformants were mapped by restriction digestion, and the complete DNA sequence was determined using an Applied Biosystems automated DNA sequencer. A clone called pIMPDH-h2 had the expected restriction map and DNA sequence and was transformed into strain BL21(DE3) for further study.
IMPDH-h2 Activity Assay-All assays were performed at 37°C. The IMPDH-h2 activity was determined from the rate of NADH production by monitoring the increase in absorbance at 340 nm. A Hewlett Packard 8452A diode array UV-visible spectrophotometer with a computerized kinetic measurement program was used. The extinction coefficient of 6220 cm Ϫ1 M Ϫ1 at 340 nm was used to calculate the NADH concentration. The assay buffer, if not otherwise specified, consisted of 100 mM Tris/HCl (pH 8.1). KCl, NaCl, RbCl, LiCl, CsCl, NH 4 Cl, N(CH 3 ) 4 Cl, and thallium acetate were used as sources of the monovalent cations. For product inhibition experiments, XMP or NADH was also added into the assay solution before the reaction. Reactions were started by the addition of IMPDH-h2. The standard assay solution included 0.2 mM IMP, 0.2 mM NAD, and 10 mM K ϩ .
Purification of IMPDH-h2-The BL21(DE3)/pIMPDH-h2 cells were grown at 37°C in LB media containing 0.1 g/liter ampicillin for 1 day; isopropyl-1-thio-␤-D-galactopyranoside was not needed for expression and was not added. Cells were harvested by centrifugation and then suspended in buffer A (20 mM Tris, 20 mM KCl, 1 mM EDTA⅐Na 4 , and 1 mM dithiothreitol, pH 8.1, with HCl). After breaking the cells in a French press, the debris was removed by centrifugation at 4°C. Subsequent steps were performed at room temperature. The supernatant containing IMPDH-h2 was loaded onto a Blue Sepharose CL-6B (2.6 ϫ 15 cm) (Pharmacia Biotech Inc.) column. IMPDH-h2 was eluted by linear gradient of 0.02-1 M KCl in 1000 ml of buffer A. Fractions with high IMPDH activity were collected and desalted by dialysis against buffer A (2 changes of 1000 ml each). This solution was loaded onto a 2.6 ϫ 15-cm column of Q-Sepharose (Pharmacia) anion exchange resin. IMPDH-h2 bound to the ion exchange column and was eluted in the same way as the Blue Sepharose CL-6B column. The fractions with high IMPDH activities were collected and dialyzed (9 changes of 1000 ml each) against buffer B (20 mM Tris, 10 m M N(CH 3 ) 4 Cl, 0.5 mM EDTA (free acid), and 1 mM dithiothreitol, pH 8.1) to remove K ϩ . The purity of the final IMPDH-h2 preparation was verified by SDS gel electrophoresis, and a sole band corresponding to an IMPDH-h2 subunit of molecular mass of ϳ56 kDa was observed. The subunit molecular weight of the purified protein was 55798 as determined by matrix-assisted laser desorption ionization-time of flight mass spectrometry, within instrumental uncertainty of the value of 55788 predicted from the amino acid sequence with the N-terminal methionine removed, which was expected because the penultimate amino acid is alanine (22).
The specific activity of the purified IMPDH-h2 determined by the standard assay was 0.68 mol (NADH formed)/mg/min. The concentration of the purified IMPDH-h2 was determined by UV absorption; the absorbance at 280 nm for a 1 mg/ml IMPDH-h2 concentration was found to be 0.465 using the 205 nm/280 nm ratio method of Scopes (23). An overall yield of ϳ150 mg IMPDH-h2 from 10 g of cells was obtained.
Substrate and Product Binding Measurements-Binding experiments were performed by ultrafiltration (24) using Centricon-30 devices and a temperature controllable centrifuge. A substrate (IMP or NAD) or product (XMP or NADH) was mixed with the purified IMPDH-h2 solution in a Centricon-30 tube and incubated in the centrifuge for 20 min to reach equilibrium at 4 or 37°C. The tubes were then spun to separate part of the unbound substrate or product from the mixture.
The concentration of the unbound NADH was measured from its absorbance at 340 nm. However the concentrations of unbound XMP, IMP, and NAD could not be directly determined from their UV absorption due to interference from the UV absorption of the dithiothreitol in the buffer, which slowly oxidized to form variable amounts of a UVabsorbing cyclic disulfide. To measure the concentration of unbound XMP, the phosphate group on XMP was enzymatically removed by alkaline phosphatase, and the concentration of the released phosphate was determined colorimetrically (25).
The concentrations of the unbound IMP and NAD were determined by the method described below. The equilibrium of the IMPDH-cata-lyzed reaction lies far toward products (18). In an assay solution when the concentration of one substrate was more than 10 times the concentration of the second substrate, the NADH concentration produced by the reaction was found to be equal to the starting concentration of the second substrate. The unbound IMP or NAD from the Centricon device was put in a cuvette, and the other substrate was added at a concentration known to be more than 10 times that of the unbound substrate. K ϩ was also added to a concentration of ϳ10 mM. The reaction was started by the addition of IMPDH-h2. The final NADH concentration was determined from the increase in absorbance at 340 nm. The lack of a need of a reference sample is an advantage of this method.
Binding experiments were done with five different concentrations of substrate or product. Total enzyme concentrations (E t ) and total substrate or product concentrations (L t ) were known, and unbound substrate or product concentrations (L) were measured. The data were fit to the equation: E t /(L t Ϫ L) ϭ (K d /N)(1/L) Ϫ 1/N to determine the number (N) of binding sites per IMPDH-h2 molecule and the dissociation constant (K d ) of the substrate-or product-IMPDH-h2 complex (26).
The Determination of IMPDH-h2 Quaternary Structure-Fast protein liquid chromatography gel filtration on Superose-12 (Pharmacia) was used to determine the native molecular weight of IMPDH-h2. The column was calibrated with protein standards (Bio-Rad). The column was equilibrated and eluted with buffer C, which was same as buffer B except that the N(CH 3 ) 4 Cl concentration was increased to 100 mM. The experiment was repeated using buffer D as equilibrating and eluting solution; buffer D was same as buffer C but the N(CH 3 ) 4 Cl was replaced by KCl. In both cases the molecular mass of IMPDH-h2 was found to be ϳ220 kDa, consistent with a tetrameric structure.

The Determination of Individual Rate Constants and Dissociation Constants by Computer Simulation of the Steady State Rate Equation-
The steady state kinetic mechanism including the monovalent cation activation was proposed based on the initial reaction and product inhibition measurements as described under "Discussion," and the complete steady state rate equation for the mechanism was derived. Initial reaction rates of 150 assays at different concentrations of the monovalent cation (2-10 mM K ϩ ), substrate (0.04 -0.2 mM IMP and 0.05-0.4 mM NAD), and product (0.02-0.1 mM NADH and 0.02-0.5 mM XMP) concentrations were fit to the complete rate equation ("Appendix") by a least squares method using the computer simulation program "Scientist" developed by MicroMath Scientific Software. In all the assays, the reaction was started by the addition of IMPDH-h2 to the same concentration of 61 nM. The computer simulation gives individual rate constants and dissociation constants as defined in the rate equation.

Enzyme
Purification and Characterization-IMPDH-h2 was purified to electrophoretic homogeneity as described under "Experimental Procedures." The subunit molecular mass determined by matrix-assisted laser desorption ionization-time of flight mass spectrometry was within 0.02% of that predicted from the DNA sequence, indicating a lack of proteolysis. The purified IMPDH-h2 could be stored at 4°C or room temperature for 3 days without losing activity. When fast frozen in liquid nitrogen, it could be stored indefinitely at Ϫ80°C. Elemental analysis of the purified protein by the inductively coupled argon plasma method showed no significant levels of sodium, potassium, zinc, copper, manganese, iron, magnesium, or phosphorus (Ͻ0.01 atom per protein subunit). Gel filtration experiments showed that IMPDH-h2 exists as a tetramer in the presence of either the activator K ϩ or N(CH 3 ) 4 ϩ , which neither activates nor inhibits the K ϩ -activated enzyme. Thus the activation by K ϩ does not result from changes in the quaternary structure of the enzyme.
The activity of IMPDH-h2 was determined at a series of different pH values to find the optimal pH as shown in Fig. 1. The pH value of the assay buffer that provides maximal IMPDH-h2 activity is 7.7 at 37°C, which for this buffer corresponds to pH 8.1 at room temperature. In all other experiments, assay buffers were adjusted to pH 8.1 at room temperature. A computer fit of the pH rate profile gave apparent pK a and pK b values of 6.8 and 8.5 (Fig. 1).
Kinetic Studies-IMPDH from various organisms has been found to require a monovalent cation such as K ϩ for optimal activity with activation by up to 10-fold (17,18). For IMPDH-h2, activity in the absence of K ϩ was less than 1% of that in the presence of K ϩ . This result shows the requirement of K ϩ (or a related monovalent cation) as an essential activator for IMPDH-h2 activity.
The activation of IMPDH-h2 by various monovalent cations was characterized. Fig. 2 shows the relation between IMPDH-h2 activity and monovalent cation concentration for K ϩ , Na ϩ , Rb ϩ , NH 4 ϩ , Tl ϩ , and Cs ϩ . For each monovalent cation, the activity of IMPDH-h2 increases with increasing cation concentration and reaches an optimum; higher cation concentrations reduce the activity. K ϩ produces the highest activity, whereas Tl ϩ produces optimal activity at the lowest concentration. The optimal concentrations for K ϩ and Tl ϩ , which have similar ionic radii, are ϳ20 mM and ϳ5 mM, respectively. Other cations have optimal concentrations of ϳ100 mM. Li ϩ and N(CH 3 ) 4 ϩ were found not to activate the enzyme, and concentrations up to 100 mM did not inhibit the enzyme activated by 10 mM K ϩ . These results suggest that the affinity of IMPDH-h2 for the monovalent cation is sensitive to the size of the ion.
The dependence of IMPDH-h2 activity on concentrations of IMP and NAD were studied in the presence of Na ϩ , K ϩ , Tl ϩ , Rb ϩ , and NH 4 ϩ at their optimal concentrations. The data were fit to the equation for the steady state ordered sequential Bi Bi mechanism in which IMP binds to the enzyme before NAD. The kinetic constants for IMPDH-h2 activated by the five monovalent cations are shown in Table I. The kinetic constants for the K ϩ -activated IMPDH-h2 are comparable with those recently reported by Carr et al. (19) using enzyme obtained from a different expression system and purification scheme (19). There appears to be no simple relation between the substrate  The kinetic constants were calculated by fitting IMPDH-h2 activities (V) and corresponding IMP and NAD concentrations (A and B) to the equation V ϭ V max AB/(K ia K b ϩ K a B ϩ K b A ϩ AB) using the Cleland program (27). Assay buffers used for Na ϩ -, K ϩ -, Rb ϩ -, and NH 4 ϩ -activated IMPDH-h2 were 100 mM Tris-HCl (pH 8.1), whereas 100 mM Tris-acetic acid (pH 8.1) buffer was used for T1 ϩ -activated IMPDH-h2. The concentrations of Na ϩ , K ϩ , T1 ϩ , Rb ϩ , or NH 4 ϩ used were 100, 10, 5, 100, or 100 mM, respectively. In each assay was 32 nM IMPDH-h2.  Fig. 3 shows double reciprocal plots of the dependence of IMPDH-h2 activity on the concentrations of K ϩ , IMP, and NAD. Each plot consists of nonparallel straight lines intersecting at a point off either axis. Fig. 4 shows the double reciprocal plots for product inhibition experiments. XMP shows competitive inhibition with respect to IMP and noncompetitive inhibition with respect to K ϩ and NAD. NADH inhibits noncompetitively with respect to K ϩ , IMP, and NAD. No cation that inhibits by competing with K ϩ has been found. These data are consistent with an ordered sequential kinetic mechanism.
Equilibrium Binding Studies-Binding studies were initially conducted in the absence of a monovalent cation activator. IMP and XMP were found to bind to IMPDH-h2, even in the absence of K ϩ . The results of IMP binding experiments at 37 and 4°C are shown in Fig. 5. The data were fit to an equation derived for a model of four IMP binding sites per enzyme tetramer, and the results are consistent with this model. No NAD binding was detectable at either temperature. No detectable binding of NADH was found at 37°C; however, NADH binding was measurable at 4°C, and the data were consistent with four binding sites per tetramer (Fig. 5). The numbers of binding sites for IMP and NADH determined by fitting the data to the general binding equation were 3.6 Ϯ 0.9 and 3.6 Ϯ 0.7. The number of XMP binding sites could not be determined unambiguously from the binding data. The com-petitive inhibition of XMP versus IMP indicates that XMP binds at the IMP binding sites. Thus the dissociation constant for XMP was calculated assuming four XMP binding sites per tetramer. The dissociation constants are listed in Table II. The addition of 10 mM K ϩ enhanced the IMP binding appreciably, and the apparent dissociation constant decreased from 0.2 to 0.1 mM at 37°C. However no NAD or NADH binding was observed at 37°C even when 10 mM K ϩ was added. DISCUSSION A kinetic mechanism including the monovalent cation activator has not previously been elucidated for human IMPDH (19). The kinetic data shown in Figs. 3 and 4 are not consistent with the partially random rapid equilibrium reaction mechanism suggested by Morrison and co-workers (20) for IMPDH from A. aerogenes because the common intersection (Fig. 3E) in plots of 1/V versus 1/[NAD] at different concentrations of K ϩ is not on the vertical axis. The competitive inhibition by XMP with respect to IMP (Fig. 4C) and the characteristic off-axis common intersections in Figs. 3 (D and F) and 4 (D, E, and F) are the patterns predicted by the steady state ordered sequential Bi Bi mechanism in which IMP binds first and XMP is released last (20). This part of the mechanism has been proposed by other workers (13, 19, 28 -32).
A more complete mechanism must take into account the monovalent cation, which this study shows to be an essential activator for IMPDH-h2. Based on the dependence of IMPDH-h2 activity on K ϩ and substrate concentrations shown in Fig. 3 and product inhibition results shown in Fig. 4, a steady state reaction mechanism including the activation of IMPDH-h2 by a monovalent cation is proposed in Fig. 6. In the productive sequence, IMPDH-h2 binds the monovalent cation first and then binds IMP, and NAD is bound last. NADH is released before XMP. Because no inhibitors that compete with K ϩ are known, it is not clear whether the monovalent cation  activator is required to dissociate in each catalytic cycle. Kinetic studies indicate that the free IMPDH may also bind IMP and XMP to form dead end complexes, and the formation of binary enzyme⅐IMP and enzyme⅐XMP complexes has been verified by equilibrium binding studies. If no products are present, the dependence of IMPDH activity (V) on the monovalent cation (M), IMP (A), and NAD concentration (B) is described by Equation 1. In the presence of XMP or NADH inhibitors, the relation is described by Equations 2 or 3, where Q and P are concentrations of XMP and NADH, respectively. The constants C with different subscripts (see "Appendix") are combinations of rate constants and total enzyme concentration (E t ). These equations were derived using the King-Altman method (33).
predicts an off-axis common intersection for each double reciprocal plot as found in the results shown in Fig. 3. Equation 2 predicts the competitive inhibition of XMP with respect to IMP as shown by the common intersection on the vertical axis in Fig. 4C. The off-axis common intersections in Fig. 4 (A, B, D, E, and F) are all consistent with equations 2 and 3. Thus the initial velocity data (Fig. 3) and product inhibition data (Fig. 4) agree with the proposed mechanism shown in Fig.  6. The double reciprocal plots of initial velocity and product inhibition obtained by Anderson and Sartorelli for IMPDH from sarcoma 180 ascites tumor cells are also consistent with the mechanism proposed here (18).
Evidence supporting the mechanism in Fig. 6 was provided by equilibrium binding experiments. The EA complex (IMP⅐IMPDH-h2), and the EQ complex (XMP⅐IMPDH-h2) were shown to form, and the data were consistent with four IMP binding sites per IMPDH-h2 tetramer. This agrees with results obtained by Wu and co-workers from IMPDH inactivation by 6-Cl-IMP, which suggested one IMP binding site per IMPDH subunit (16). The 2-fold greater affinity for IMP found in the presence of 10 mM K ϩ is consistent with the K ia value, which reflects the affinity of IMP for the binary enzyme⅐K ϩ complex (Table I). It is as yet not clear whether there is a direct interaction between K ϩ and IMP or whether K ϩ mediates a conformational change of IMPDH-h2. NAD and NADH did not bind detectably at 37°C. Although NADH binding was detected at 4°C, this E⅐NADH complex is apparently not on the primary kinetic pathway.
The individual rate constants and dissociation constants determined by the computer simulation fitting the steady state rate equation to 150 rate measurements at varying concentrations of K ϩ , IMP, NAD, XMP, and NADH as described under "Experimental Procedures" are shown in Table III. Other rate constants from the simulation have large standard deviations and are not shown. K 6 and K 7 are very close to the dissociation constants of XMP and IMP determined by direct binding experiments (Table II). This indicates that the results (Table III) from the computer simulation are reasonable and will provide a framework for future studies.
An essential role of a monovalent cation such as K ϩ , Na ϩ , Rb ϩ , Tl ϩ , and NH 4 ϩ for IMPDH-h2 activity is indicated by this study. Removal of the monovalent cation from IMPDH-h2 by dialysis does not cause an irreversible loss of activity of the enzyme and does not destroy its tetrameric structure. The binding and release of the monovalent cation is thus a reversible process. K ϩ and Tl ϩ are better IMPDH-h2 activators than other monovalent cations in the sense that they produce maximal activity at about 10 times lower concentrations than other monovalent cations, as shown in Fig. 2. This suggests a size selectivity for ion binding. The maximal activities achieved with the tested cations are all of the same order of magnitude. Because the rate-limiting step of the reaction is not yet known, it is possible that different cations have different effects on intrinsic steps that are partially masked by cation-insensitive steps. Recent crystallographic studies demonstrated conformational changes in diakylglycine decarboxylase when it bound different monovalent cations including K ϩ , Rb ϩ , Na ϩ , and Li ϩ , and the conformational change was shown to be critical for the activation by K ϩ and Rb ϩ and inhibition by Na ϩ and Li ϩ (34,35). The variation of kinetic constants in the presence of different monovalent cations (Table I) observed in this study may suggest that different IMPDH-h2 conformational changes are caused by the binding of ions of different ionic radii. The dependence of kinetic constants on ionic radii of the monovalent cations as shown in Table I may imply that the ionic radius is critical in determining the monovalent cation binding affinity and the nature of the putative conformational change caused by the monovalent cation binding.