Probing the Role of Metal Ions in the Catalysis of Helicobacter pylori 3-Deoxy- D - manno -octulosonate-8-phosphate Synthase Using a Transient Kinetic Analysis*

synthase catalyzes the net condensation of phosphoenolpyruvate and D -arabinose 5-phosphate to form KDO8P and inorganic phosphate (P i ). Two classes of KDO8P synthases have been identified. The Class I KDO8P synthases ( e.g. Escherchia coli KDO8P synthase) catalyze the condensation reaction in a metal-indepen-dent fashion, whereas the Class II enzymes ( e.g. Aquifex aeolicus ) require metal ions for catalysis. Helicobacter pylori ( H. pylori ) KDO8P synthase, a Zn 2 (cid:1) -dependent metalloenzyme, has recently been found to be a Class II enzyme and has a high degree of clinical significance since it is an attractive molecular target for the design of novel antibiotic therapy. Although the presence of a divalent metal ion in Class II KDO8P synthases is essential for catalysis, there is a paucity of mechanistic infor-mation on the role of the metal ions and functional differences as compared with Class I enzymes. Using H. pylori KDO8P synthase as a prototypical Class II enzyme, a steady-state and transient kinetic approach was undertaken to understand the role of the metal ion in catalysis and define the kinetic reaction pathway. Metal reconstitution concentration of A5P was kept constant at a saturation level. To maintain a saturating level of A5P experiments were performed (cid:1) A5P ( (cid:1) The was calculated from a linear least square fit of the in the absorbance of the curve. As expected, the experimental data under the above conditions were at- tempted to fit typical Michaelis-Menten kinetics (see “Results” and “Discussion”). This analysis allows determination of apparent Michaelis constant ( K m ) and maximum velocity ( V max ) of the enzyme activity. internal standard because it co-elutes with hot PEP and its elution time can be monitored at 232 nm by UV detector. These conditions were used to analyze the samples generated from the rapid chemical quench experiments.

3-Deoxy-D-manno-2-octulosonate-8-phosphate (KDO8P) 1 synthase catalyzes a net aldol condensation reaction between Darabinose 5-phosphate (A5P) and phosphoenolpyruvate (PEP) to form an unusual eight-carbon sugar KDO8P and inorganic phosphate (P i ). This is a key enzymatic reaction that controls the carbon flow in the biosynthetic formation of 8-carbon sugar 3-deoxy-D-manno-2-octulosonate (KDO). The KDO is an important constituent of lipopolysaccharides found in most Gramnegative bacteria (1) and plays a crucial role in this assembly process (1,2). Because KDO8P synthase is essential for Gramnegative bacteria and not present in mammalian systems, it represents an attractive molecular target for the design of new antibiotics (3,4).
Only one other enzyme found in nature catalyzes a related type of enzymatic reaction. The enzyme 3-deoxy-D-arabino-2heptulosonate-7-phosphate (DAHP) synthase catalyzes a similar condensation reaction between PEP and erythrose-4-phosphate, a sugar containing one less carbon, to form DAHP and inorganic phosphate (P i ). The formation of the DAHP is the first committed step in the biosynthesis of the intermediate compounds chorismate and prephenate, which are precursors to the aromatic amino acids (Phe, Tyr, and Trp), catechols, and p-aminobenzoic acid as well as a number of other highly important microbial compounds (5). Both enzymes, KDO8P synthase and DAHP synthase, catalyze the condensation reactions with the same stereo-facial selectivity with respect to the double bond of PEP and the aldehyde moiety of the monosaccharide substrate (6 -9) as well as with cleavage of the C-O bond of PEP (8, 10 -12) rather than the more common scission of the P-O bond. Previous studies on Escherichia coli KDO8P synthase suggest the presence of an acyclic hemiketal phosphate intermediate, I, as shown in Scheme 1 (13). Accordingly, the reaction involves the nucleophilic attack of water/hydroxide on C2 of PEP followed by or in concert with the nucleophilic attack of C3 PEP on the aldehyde carbon of A5P.
Although there is little or no homology at the level of the primary sequence, the three-dimensional structures of E. coli KDO8P synthase and DAHP synthase have been reported and provide evidence that these two enzymes are not only mechanistically, but also structurally related. Nonetheless, these two enzymes have evolved to use two different strategies to promote catalysis (14,15). The catalytic activity of DAHP synthase is dependent on the divalent transition metal ion such as Mn 2ϩ or Zn 2ϩ . The kdsa gene encoding KDO8P synthase has been identified in other bacteria (16 -18) and plants (16), and based on the early observations on the E. coli KDO8P synthase (19), it has been found that the enzyme does not require divalent metal ion for catalysis.
Recently it was shown, however, that the KDO8P synthase from the hyperthermophilic bacterium Aquifex aeolicus and similar species such as Aquifex pyrophilus require divalent metal cofactor for catalysis (20,21), implying that there may be two classes of KDO8P synthases, one metal-independent and one metal-dependent. Thus KDO8P synthase represents one of the first examples where orthologous enzymes differ based on their requirement of the metal cofactor. The x-ray crystal struc-tures of the E. coli and A. aeolicus KDO8P synthases were reported (15,22). The overall three-dimensional structures of both enzymes are nearly superimposable, and most of their active sites can be overlapped. Three of the four amino acid residues responsible for chelating the divalent metal ion in A. aeolicus KDO8P synthase (His-185, Glu-222, and Asp-233) are nearly in identical positions with His-202, Glu-239, and Asp-250 of E. coli KDO8P synthase. Moreover, by using a phylogenetic analysis of 29 KDO8P synthase sequences it was clearly shown that KDO8P synthase could be divided into two different classes with respect to their requirement for metals (23). The Class I KDO8P synthases, which include the most extensively studied E. coli enzyme, was shown to be metal-independent, whereas the Class II enzymes, which include the A. aeolicus enzyme, were proposed to be metalloenzymes. In fact, according to the above phylogenetic and sequence analysis, KDO8P synthase from Helicobacter pylori was hypothesized to be Class II enzyme, and it was recently found to be a metalloenzyme (24). Thus, KDO8P synthase represents a unique example of an enzyme that requires metal ion in one microorganism and does not require metal in other.
One of the putative metallo-KDO8P synthases is encoded by H. pylori, an important human pathogen that colonizes the gastric mucosa and can cause gastritis and cause gastric ulcers (25)(26)(27)(28). H. pylori infection was linked to an increased risk of developing some forms of gastric cancer and was classified as a carcinogen (29,30). H. pylori KDO8P synthase is reported to be a zinc-metalloenzyme (24). The presence of the Zn 2ϩ has been shown to be catalytically important. The enzyme was inactivated by the treatment with metal-chelating agents such as EDTA and PDA. Although a preliminary study of the effect of metal ions on H. pylori KDO8PS has been previously reported (25), a detailed mechanistic understanding of this enzyme or other Class II-KDO8P synthases is lacking.
Using H. pylori KDO8P synthase as a prototypical Class II KDO8P synthase, a major objective of the present study was to understand the nature of the metal ions at the active site because different metals may have different affinities and coordination geometry toward protein ligands. An ancillary goal was to define the kinetic reaction pathway to serve as a basis for comparison with the metal-independent Class I KDO8P synthases. Accordingly, a detailed kinetic analysis of the H. pylori native and other metal-reconstituted enzymes was investigated. By comparing the steady-state and pre-steady-state kinetic parameters of the different metal-reconstituted enzymes, the following questions were addressed. (i) What is the rate-limiting step of the overall reaction? (ii) What is the rate of chemical catalysis? (iii) What is the nature of the metal ions? (iv) What are the mechanistic similarities and differences with Class I KDO8P synthases ? The present study provides the first detailed kinetic characterization of a Class II KDO8P synthase. The role of metal ion in catalysis as well as functional similarities and differences with the metal-independent KDO8P synthases are discussed.

EXPERIMENTAL PROCEDURES
Materials-A5P (disodium salt), PEP (monosodium salt), and Chelex 100 were obtained from Sigma. Pefabloc serine protease inhibitor was obtained from Fluka. Metal chlorides were obtained from Sigma. All other reagents were obtained from Sigma. In all experiments Millipore nano-Q quality water was used. The water was passed through the Chelex 100 resin. The pH dependence experiments were done in 50 mM Tris buffer at pH Ͼ 7.0 and in 50 mM MES buffer at pH Ͻ 7.0.
Preparation of Radiolabeled [1-14 C]PEP and A5P-The [ 14 C]PEP (20 mCi/mmol) was prepared from [ 14 C]pyruvate (20 mCi/mmol, PerkinElmer Life Sciences) using pyruvate phosphate dikinase, obtained as a generous gift from D. Dunaway-Mariano. The radiolabeled PEP was purified by Q-Sepharose column chromatography (Amersham Biosciences) using a linear gradient of 20 mM to 0.5 M triethylammonium bicarbonate, pH 8.0. The fractions containing radiolabeled PEP were pooled and lyophilized. The lyophilized powder was dissolved in 50 mM Hepes, pH 7.6, to a final concentration of 0.3 mM.
Purification of H. pylori KDO8P Synthase and Determination of Enzyme Concentration-The H. pylori KDO8P synthase containing kdsa expression plasmid was obtained as a generous gift from Dr. Stewart Fischer, AstraZeneca (24). A small volume of starter culture was grown overnight with 50 g/ml ampicillin. The small overnight cultures were transferred into 20 liters of LB medium containing ampicillin and were grown at 37°C until the A 600 became 0.5. The culture was induced with 0.5 mM isopropyl-␤-D-thiogalactopyranoside. ZnCl 2 was added to the culture to a final concentration of 50 M, and the culture was continued to grow for another 3 h. The cells were harvested in a Sorvall centrifuge at 12,000 rpm for 15 min. The cell pellet was resuspended in 50 mM KH 2 PO 4 , pH 7.5, containing 1.5 mM Pefabloc and one Mini, EDTA-free protease inhibitor mixture tablet (Roche Applied Science) per 10 ml of solution at 4°C. The cell suspension was passed through a French press, and the extract was centrifuged at 14,000 rpm for 30 min at 4°C. The solution was applied to a DEAE-Sepharose CL 6B ion exchange column pre-equilibrated with 50 mM KH 2 PO 4 , pH 7.5, and the enzyme was eluted using a NaCl gradient of 0 -0.4 M. The fractions containing KDO8P synthase were analyzed by SDS-PAGE and pooled. Solid ammonium sulfate was added to 20% saturation of the solution. The solution was then applied to a butyl-Sepharose 4B FF column pre-equilibrated with 50 mM KH 2 PO 4 , pH 7.5, 30% ammonium sulfate. A linear descending gradient of ammonium sulfate (30 -0%) was applied, and the protein was eluted. Peak fractions containing KDO8P synthase were combined and concentrated using ultrafiltration. The concentrated fractions were then applied into Superdex 200 column, and the fractions were collected. The fractions were dialyzed against 50 mM Tris-HCl, pH 7.5, 0.2 mM dithiothreitol and then stored at Ϫ80°C with 10% glycerol. The concentration of the KDO8P synthase was determined using a Bio-Rad assay. E. coli KDO8P synthase was expressed and purified as previously described (31).
Steady-state Assays-The steady-state kinetic experiments for H. pylori KDO8P synthase activity were performed by monitoring the KDO8P synthase-catalyzed decrease in the absorbance of PEP at 232 nm. For all experiments initial rates for the decrease in the absorbance of PEP were measured by varying the concentration of PEP, whereas SCHEME 1 the concentration of A5P was kept constant at a saturation level. To maintain a saturating level of A5P experiments were performed at 300 M of A5P (K m of A5P ϳ 39 M). The initial rate was calculated from a linear least square fit of the decrease in the absorbance of the curve. As expected, the experimental data under the above conditions were attempted to fit typical Michaelis-Menten kinetics (see "Results" and "Discussion"). This analysis allows determination of apparent Michaelis constant (K m ) and maximum velocity (V max ) of the enzyme activity.
Preparation of H. pylori ApoKDO8P Synthase-200 M of H. pylori KDO8P synthase solution was prepared using 50 mM Tris-HCl, pH 7.5. The enzyme was incubated with freshly prepared EDTA (concentration of EDTA was kept 10 times higher than the enzyme) and kept in ice for about 45 min. Then the activity assay of the EDTA-treated enzyme was carried out and found that the enzyme has almost 3-4% activity. The solution was then dialyzed against two changes of 4 liters of 50 mM Tris-HCl, pH 7.5, at 4°C.
Preparation of Metal-reconstituted H. pylori KDO8P Synthase-Cd 2ϩ -reconstituted KDO8P synthase enzyme was prepared by incubating the apoenzyme (200 M, final concentration) with CdCl 2 (concentration of CdCl 2 was kept 8 times higher than the apoenzyme) in 50 mM Tris-HCl, pH 7.5, and the mixture was kept in ice for about 30 min. A quick activity assay was done with the CdCl 2 -treated apoenzyme, and it was found that the enzyme has good activity. The mixture was then dialyzed against two changes of 4 liters of 50 mM Tris-HCl, pH 7.5, at 4°C. The level of Cd 2ϩ present in the dialyzed sample was determined by electrospray ionization mass spectrometry, and the analysis showed that 1 mol of Cd 2ϩ was bound with 1 mol of the enzyme (data not shown).
The Cu 2ϩ , Co 2ϩ , Mn 2ϩ , and Ni 2ϩ -enzymes were prepared in a slightly different manner. In these cases, steady-state assays were performed by incubating the apoenzyme with different metal chlorides, and the initial rates were determined at a fixed concentration of apoenzyme but with varying concentrations of metal chlorides. The ratio of metal/apoenzyme corresponded to the maximum initial rate (insets of Fig. 4, A and B, are shown as representative figures) was used for further steady-state and pre-steady-state kinetic experiments.
Rapid Chemical Quench Experiments-Rapid quench experiments were performed using a Kintek RFQ-3 Rapid Chemical Quench (Kintek Instruments, Austin, TX) as previously described (32). The reaction was initiated by mixing the enzyme solution (15 l) with the radiolabeled substrate PEP (15 l). In all cases the concentrations of the enzyme, substrate, and metal cited are those after mixing and during the reaction. The reaction was then quenched with 67 l of 0.6 N KOH. It is found that the substrate and products are stable under these conditions. The substrates and products were separated and quantified using anion exchange column coupled with simultaneous radioactivity detection. The HPLC separation was performed on a Mono Q (HR 5/5) anion exchange column with flow rate of 1 ml/min. A gradient separation was employed using solvent A (20 mM triethylammonium bicarbonate, pH 9.0) and solvent B (1 M triethylammonium bicarbonate, pH 9.0) in a linear gradient program of 100 -0% A from 0 to 30 min and 0 -100% B from 0 to 30 min followed by re-equilibration. The elution times were 15 and 17.8 min for KDO8P and PEP, respectively. During the HPLC analysis a small amount of cold PEP was always injected as an internal standard because it co-elutes with hot PEP and its elution time can be monitored at 232 nm by UV detector. These conditions were used to analyze the samples generated from the rapid chemical quench experiments.

H. pylori KDO8P Synthase Does Not Contain Tightly Bound
PEP-Previous studies show that E. coli KDO8P synthase, a Class I KDO8P synthase, was isolated with 1 eq of tightly bound PEP, which contributes stability to the enzyme (33). To determine whether H. pylori KDO8P synthase retained any tightly bound substrates after purification the enzyme was preincubated with either radiolabeled A5P or PEP in the absence of the second substrate. Unlike the Class I KDO8P synthases, no radiolabeled KDO8P peak was found in the HPLC analysis, indicating that this prototypical Class II enzyme does not contain tightly bound PEP substrate. However, reconstitution experiments adding metal to prepare a holoenzyme from the frozen apoenzyme yielded enzyme with a very low specific activity. We were also unable to fully reconstitute the holoenzyme if the apoenzyme was kept for a longer period (2-3 days) at 4°C. These results indicate that the apoenzyme is less stable upon removal of the metal ions from the native enzyme.
Steady-state Kinetics of the Native and Metal-reconstituted Enzymes-The steady-state kinetic experiments of KDO8P synthase-catalyzed reaction were carried out by monitoring the decrease in the absorbance of PEP at 232 nm. Fig 1A shows the steady-state plot of the native enzyme. The data could be best fitted to a sigmoidal equation yielding K m , k cat , and Hill coefficients (n) 2.6 M, 0.3 s Ϫ1 , and 2.6, respectively, at pH 7.8. The values of K m and k cat were found in good agreement with the reported values (15). The data of Fig. 1A was fitted to a hyperbolic equation as shown an inset of Fig. 1A. Fig. 1, B-D, showed the plots of initial rate versus the concentration of PEP for the Cd 2ϩ , Cu 2ϩ , and Co 2ϩ enzymes, respectively. Insets of Fig. 1, B-D, show that the data were fitted to a hyperbolic equation. However, the same data could be best fitted to a sigmoidal equation yielding n ϭ 4, 2.5, and 2 for Cd 2ϩ , Cu 2ϩ , and Co 2ϩ enzymes, respectively. The various steady-state kinetic parameters are summarized in Table I.
pH Dependence of the Native and Cd 2ϩ -reconstituted Enzymes-The pH-dependent experiments of the native and Cd 2ϩ enzymes were carried out in the pH range 6 -9.5. V max and K m were determined at each pH, and the data were shown in Fig 2, A and B. For the native enzyme both V max and K m exhibited bell-shaped pH dependence. Again the plot of the second order rate constant (V max /K m ) also showed similar pH dependence. A fit of the experimental data of Fig. 2A gave pK 1 ϭ 9 and pK 2 ϭ 5.8, where K 1 and K 2 are the acid equilibrium constants. However, for Cd 2ϩ enzymes, K m showed bell-shaped pH dependence, but V max showed a single ionization. For Cd 2ϩ enzymes the decrease in the V max /K m with pH ( Fig. 2B) could be best fitted to an equation V max /K m ϭ (V max / K m ) max /(1 ϩ [H ϩ ]/K a ), yielding pK a ϭ 7.5.
Pre-steady-state Burst Experiments with Native (Zn 2ϩ ) and Metal-reconstituted Enzymes-Transient-state kinetic methods allow definition of the sequence of reactions occurring at the active site of an enzyme after substrate binding and leading up to product release. On the other hand, steady-state kinetic analysis only establishes the order of substrate binding and the order of product release; the steady-state kinetic parameters k cat and k cat /K m only define the maximum rate of substrate to product conversion and a lower limit for the rate of substrate binding, respectively. Steady-state kinetic analysis cannot address questions regarding the pathway of events occurring at the active site after substrate binding and before product release. To identify the rate-limiting step of the catalysis, we have carried out pre-steady-state burst experiments using radiolabeled [ 14 C]PEP as a substrate to determine the rate of formation of radiolabeled product [ 14 C]KDO8P. In this type of experiment the radiolabeled substrate is used in slight excess over enzyme such that the first enzyme turnover as well as multiple turnovers can be examined. A representative burst experiment of the native enzyme showing the time course for biphasic formation of [ 14 C]KDO8P is shown in Fig. 3A. The rate of the product formation was ϳ0.8 s Ϫ1 for the fast phase and 0.3 s Ϫ1 for the slower linear phase. Fig. 3B shows the similar burst experiment for the Cd 2ϩ enzyme. The data were fitted to a burst equation yielding rate constants 88 and 1.1 s Ϫ1 for the fast and slow phases, respectively. The rate constant of the slow phase for both the native and Cd 2ϩ enzymes correspond to the steady-state rate (k cat ) at 24°C. The amplitude of the burst experiment shown in Fig. 3A provides an estimate of active site concentration, which was determined to be ϳ20% for the native enzyme. However, it was found to be 55% for the Cd 2ϩ enzyme. Fig 4, A and B, represent the burst experiments of Cu 2ϩ and Co 2ϩ enzymes showing a fast burst phase followed by a steadystate phase. The rate for the burst and steady-state phases of Cu 2ϩ and Co 2ϩ enzymes was found to be 1.7 and 0.3 s Ϫ1 and 16.6 and 1.6s Ϫ1 , respectively. The active site concentrations of Cu 2ϩ and Co 2ϩ enzymes were found to be 79 and 29%, respectively. However, no burst of product formation was observed in case of Mn 2ϩ and Ni 2ϩ enzymes, indicating that the product release is not the rate-limiting step for the catalytic reaction (Fig. 4C). DISCUSSION Our studies have focused on using H. pylori KDO8P synthase as a prototypical metalloenzyme belonging to the Class II KDO8P synthase family, providing a basis for comparison to FIG. 1. A, steady-state assays were carried out in 50 mM Tris-HCl, pH 7.8. The concentrations of the A5P and the native enzyme were kept at 300 M and 250 nM, respectively. The data were fitted to a sigmoidal equation, giving K m , k cat , and Hill coefficient (n) values of 2.6 M, 0.3 s Ϫ1 , and 2.6, respectively. The inset shows the plot with the same data which were fitted to a hyperbolic equation. B, steady-state assays were carried out in 50 mM Tris-HCl, pH 7.8. The concentrations of the A5P and the Cd 2ϩ -reconstituted enzyme were kept at 300 M and 40 nM, respectively. The data were fitted to a sigmoidal equation yielding Hill coefficient (n), K m , and k cat values of 4, 5.3 M, and 1.0 s Ϫ1 , respectively. The inset shows the plot with the same data. The data were fitted to a hyperbolic equation. C, steady-state assays were carried out in 50 mM Tris-HCl, pH 7.8. The concentrations of the A5P and the Cu 2ϩ -reconstituted enzyme were kept at 300 M and 300 nM, respectively. The data were fitted to a sigmoidal equation yielding n, K m , and k cat values of 2.5, 5.8 M, and 0.1 s Ϫ1 , respectively. The inset shows the plot with same data. The data were fitted to a hyperbolic equation. D, steady-state assays were carried out in 50 mM Tris-HCl, pH 7.8. The concentrations of the A5P and the Co 2ϩ -reconstituted enzyme were kept at 300 M and 300 nM, respectively. The data were fitted to a sigmoidal equation yielding n, K m , and k cat values of 2.1, 7.5 M, and 0.1 s Ϫ1 respectively. The inset shows the plot with same data. The data were fitted to a hyperbolic equation. the metal-independent Class I KDO8P synthases. The important but unanswered questions were to define the role of the divalent metal ion as well as identify similarities and differences in the kinetic reaction pathway between the two classes of KDO8P synthases. The metal ion may play a structural and/or catalytic role by modulating protein stability and/or catalysis. An initial suggestion was that the metal ion in the Class II enzymes might play a role as a Lewis acid by coordinating the aldehyde group of A5P, thereby polarizing the carbonyl carbon of aldehyde. Accordingly, this would facilitate the nucleophilic attack by C3 of PEP to the aldehyde of A5P. Structural studies, however, do not support this suggestion. The crystal structure of A. aeolicus Cd 2ϩ -KDO8P synthase, a Class II thermophilic enzyme, has been solved in several binary complexes with either PEP or A5P substrates (22) as well as a low temperature catalytically inactive ternary complex containing metal and both substrates. In each case it was observed that the 2-OH group of A5P was coordinated with the Cd 2ϩ ion. In these complexes, the protein ligands, His-185, Asp-233, Glu-222, Cys-11, and a water molecule have been found to coordinate the Cd 2ϩ ion, giving rise to a distorted octahedral coordination state. In the enzyme-PEP binary complex, two water molecules are located in van der Waals contact with the PEP. Taken together these structural studies suggest that the metal ions may play a direct role in catalysis by activating the water molecule for the nucleophilic attack to the C2 carbon of PEP or may indirectly modulate catalysis by orienting substrates in the optimal geometry within the active site to allow the chemical reaction to occur. This might in turn stabilize a putative hemiketal phosphate intermediate, although it remains to be established whether the metallo-KDO8P synthases use a similar mechanism as established for the E. coli KDO8P synthase (see Scheme 1).
Earlier studies have shown that the Class I E. coli KDO8P synthase was isolated with 1 eq of bound PEP. This enzyme was less stable upon removal of the bound PEP (33), suggesting that the PEP maintains a critical role in stabilizing the enzyme. More recent studies on the E. coli KDO8P synthase from our laboratory suggest this may be mediated by conformational changes in which the removal of bound PEP changes the original conformation of the enzyme at the active site to produce an enzyme that is less active. 2 The native H. pylori KDO8P synthase does not have any bound PEP; however, the metal ion may play a similar role since it was observed that the enzyme was much less stable in the absence of metal ion. This may be due to a protective effective of the metal ion that prevents oxidation of the cysteine ligand in the coordination state. Thus, it is likely that the metal ions not only activate the water molecule for the nucleophilic attack but also maintain the stability/original conformation of the enzyme.
As illustrated in Table I, there are distinct differences between Class I and Class II, E. coli and H. pylori, enzymes, respectively, in terms of their catalytic efficiencies (k cat /K m ). Accordingly, the native H. pylori enzyme is almost 8-fold less efficient than the E. coli enzyme. This is primarily due to variations in the overall rate of catalysis, k cat , whereas there was little change in K m values for Class I and Class II enzymes. The k cat of E. coli enzyme was determined to be 15-fold higher than the H. pylori enzyme. It is interesting to note that the rate of k cat for the H. pylori Cd 2ϩ enzyme was found to be 3-fold faster than that of the native enzyme (Zn 2ϩ enzyme), and 2 Z. Li, A. K. Sau, and K. S. Anderson, submitted for publication. 2؉ enzymes (B). The concentrations of A5P, native enzyme, and Cd 2ϩ enzyme were kept at 300 M, 300 nM, and 40 nM, respectively. The buffers for different pH values are described under "Experimental Procedures." correspondingly, only 4.5-fold slower than the E. coli enzyme. From this data it is clear that rate of the reaction in H. pylori enzyme depends upon the nature of the divalent metal ions. The rate of the reactions of Cu 2ϩ and Co 2ϩ enzymes were found to be almost three times slower than that of the native enzyme.

FIG. 2. pH dependence of native (A) and Cd
Another important difference between Class II and Class I KDO8P synthases was the cooperative binding behavior for the metalloenzymes. For the native and metal-reconstituted KDO8P synthase enzymes (Cd 2ϩ , Cu 2ϩ , and Co 2ϩ ), cooperative binding of PEP was observed. However, no such cooperative binding of PEP was detected in the case of E. coli KDO8P synthase (figure not shown). As illustrated in Fig. 1, attempts to fit the data of Fig. 1A to a hyperbolic equation always resulted in a poor fit (inset of Fig. 1A). A sigmoidal equation most appropriately fits the data for the native enzyme. Similarly, in the case of Cd 2ϩ , Cu 2ϩ , and Co 2ϩ enzymes, a fit to a hyperbolic equation did not give a satisfactory fit (shown in the insets of Fig. 1, B-D). The data could, however, be best fitted to a sigmoidal equation with n ϭ 4, 2.5, and 2 for Cd 2ϩ , Cu 2ϩ , and Co 2ϩ enzymes, respectively. This suggests that the binding of PEP in one subunit of the enzyme facilitates the binding of PEP at other subunit. The degree of cooperativity was found to be maximal in the case of Cd 2ϩ enzyme as compared with that of the native, Cu 2ϩ , and Co 2ϩ enzymes. This observation also suggests that the native, Cd 2ϩ , Cu 2ϩ , and Co 2ϩ enzymes may exist as tetramer with the reconstituted Cd 2ϩ enzyme, showing the highest specificity.
To understand the role of metal ions and their interaction with the residues involved in catalysis, the effect of pH was examined. The pH dependence of the native and Cd 2ϩ enzymes showed interesting behavior. The pH dependence of V max and K m for the native enzyme showed a bell-shaped trend. Although an absolute fit to the data in both cases was difficult, apparent inspection gave the approximate pK a values of 7.7, 8.5 and 7.8, 8.4 for V max and K m , respectively. Typically, the pH dependence of V max /K m represents the ionization of the free enzyme and free substrate. For the native enzyme containing Zn 2ϩ as the metal ion, the pH dependence of V max /K m gave a good fit ( Fig. 2A) with two pK a values of 5.7 and 9. The difference in pK a values obtained from V max and K m , and V max /K m could be due to the difference in the ionization of the same residue in the enzyme-substrate complex and free enzyme. It is unlikely that the substrates (PEP and A5P) have any ionization in this pH range. Hence, these must be attributed to the enzyme. The former pK a may correspond to a histidine residue and latter to a cysteine or lysine residue. The pK a of the free histidine residue in water is 6.3. Although the observed pK a for histidine in this case is slightly lower than the free acid, it is not surprising based upon studies with other enzyme systems. For instance, the pK a of histidine in papain and cytotoxic ribonuclease ␣-sarcin were reported to be 4.3 and 5.8, respectively (35,36). In contrast, studies with the Cd 2ϩ enzyme, in which pH is dependent on V max /K m , showed only one pK a (ϭ 7.5). This may correspond to a cysteine residue. Unlike the results obtained for the native enzyme, these data suggest that in Cd 2ϩ enzyme histidine residue may be far away from the active site and, hence, not directly involved in the catalysis. A transient kinetic approach was used to further understand the kinetic reaction pathway of the Class II H. pylori enzyme and provide clues as to how catalysis may be occurring at the active site. This involved examining the pre-steady-state burst kinetics using rapid chemical quench methodology. A presteady-state burst of formation of KDO8P was observed followed by a steady-state phase at a rate close to k cat . This behavior is indicative of a mechanism in which the chemical catalysis does not limit the overall reaction, but rather, the release of product KDO8P is rate-limiting.
In studies with the native enzyme as shown in Fig. 3A, it is clear that the plot has a very shallow burst and a definite positive intercept. Hence, the data were fitted to a burst equation, giving low burst rate (0.8 s Ϫ1 ). In addition, the steadystate rate (0.3 s Ϫ1 ) obtained from the burst experiment is in excellent agreement with the k cat (0.3 s Ϫ1 ) determined from steady-state experiments. The amplitude of the burst gives an estimate for the active site concentration of the enzyme and, in the case of the native H. pylori metallo-KDO8P synthase enzyme, this value was determined to be ϳ20%. This suggests that the enzyme has a low concentration of active sites. However, for the Cd 2ϩ enzyme the burst and steady-state rates were determined to be 88 and 1.1 s Ϫ1 . This clearly showed that the burst of Cd 2ϩ enzyme is much higher than that of native enzyme. Because there is a faster step that governs catalysis, any enzyme reaction intermediate transiently formed may accumulate during a single enzyme turnover, thus enhancing the opportunity for detection in Cd 2ϩ enzyme. The active site concentration of Cd 2ϩ enzyme was determined to be 55%, more than 2-fold higher as compared with the native enzyme. The steady-state rate (1.1 s Ϫ1 ) obtained from the burst experiment was also in good agreement with the k cat (1.0 s Ϫ1 ) determined by steady-state experiments.
It is interesting to note that earlier studies with the E. coli KDO8P synthase also revealed a burst of KDO8P formation followed by a steady-state was observed (34). The burst rate was found to be 91 s Ϫ1 . Thus the pre-steady-state kinetics of E. coli KDO8P synthase and H. pylori Cd 2ϩ -reconstituted KDO8P synthase behave similarly, although these two enzymes evolved to use two different strategies for their catalytic reactions. In case of the Cu 2ϩ enzyme, a shallow burst was observed with a burst rate 1. 7 s Ϫ1 , which is slightly higher than that of the native enzyme. But a higher burst rate (16.6 s Ϫ1 ) was observed in case of the Co 2ϩ enzyme compared with its native enzyme. The active site concentrations were found to be 79 and 29% for the Cu 2ϩ and Co 2ϩ enzymes, respectively. Although the active site concentration of the Cu 2ϩ enzyme is higher than that of the Co 2ϩ enzyme, the burst rate is lower for the Cu 2ϩ enzyme than that of Co 2ϩ enzyme. Interestingly, no burst of KDO8P formation was observed in case of the Mn 2ϩ and Ni 2ϩ enzymes, establishing that the conversion of PEP by the Mn 2ϩ and Ni 2ϩ enzymes has a different rate-limiting step. The size of the metal ions and the nature of the ligands could play an important role as a result of different coordination geometry of the metal ions in the protein environment. For example, Zn 2ϩ (ionic radius, 0.74 Å) prefers distorted tetrahedral/tetrahedral/ trigonal bipyramidal structure, whereas Cd 2ϩ (0.92 Å) and Co 2ϩ (0.72 Å) prefer distorted octahedral/octahedral structure. On the other hand, Cu 2ϩ (0.71 Å), Mn 2ϩ (0.8 Å), and Ni 2ϩ (0.69 Å) prefer square planner/trigonal bipyramidal/square pyramidal geometry. The different coordination geometry and/or available ligands of the metal ions may be the reason for the change in active site and rate of chemical catalysis in this enzyme.
In summary, the first detailed kinetic characterization of a metal-dependent Class II KDO8P synthase from H. pylori is presented and serves as a basis of comparison to the metalindependent Class I KDO8P synthases. Important findings regarding the role of metal ions in catalysis in Class II KDO8P synthases include (i) a marked difference in catalytic efficiency that is dependent upon the metal ion with Cd 2ϩ , showing the highest activity and specificity compared with the other metals, ii) cooperative behavior for the binding of PEP in the native as well as the other metal-reconstituted enzymes, with the highest degree of cooperativity in the case of Cd 2ϩ (n ϭ 4), (iii) versatility of different metal ions, illustrated as the rate-limiting step changes, dependent upon the particular divalent metal ion bound (product release is the rate-limiting step for the native, Cd 2ϩ , Cu 2ϩ , and Co 2ϩ enzymes, but for the Mn 2ϩ and Ni 2ϩ enzymes chemistry is rate-limiting), and iv) a rate of chemical catalysis that is higher for Cd 2ϩ than for the other metal ions.
A comparison of the kinetic reaction pathway for Class I and Class II KDO8P synthases reveals that for each class the rate-limiting step, release of the KDO8P product, is in general the same (noting that the Mn 2ϩ and Ni 2ϩ metalloenzymes are exceptions). Also the presence of either a divalent metal in the Class II KDO8P synthases or tightly bound PEP substrate in the Class I KDO8P synthases promotes stability. A marked difference in the two classes is the observation of cooperative behavior for the Class II KDO8P synthases. The present study illustrates how two different classes, one metal dependent and uct, KDO8P, was monitored by HPLC with radioactive detection. The curve represents a fit to a burst equation with a rate of 16.6 s Ϫ1 for the fast phase and 1.6 s Ϫ1 for the slower phase. The inset shows the steady-state plot of the initial rate versus [Co 2ϩ ]/[apoenzyme]. C, kinetics of a pre-steady-state burst of radiolabeled product (KDO8P) in Mn 2ϩ -and Ni 2ϩ -reconstituted enzymes. The concentrations of the apoenzyme, MnCl 2 , and NiCl 2 were 10 M, 970 M, and 3.4 mM, respectively. The concentrations of [1-14 C]PEP and A5P were 29.3 and 250 M, respectively. For the maximum activity, the ratios of the concentration of MnCl 2 to apoenzyme and NiCl 2 to apoenzyme were determined from the steady-state measurements and were found to be 97 and 340 for Mn 2ϩ and Ni 2ϩ enzymes, respectively (data not shown). at 25°C (final concentration after mixing). The reaction was terminated by quenching with 0.4 N KOH, and the formation of radiolabeled prod-one metal-independent, have evolved to carry out the same catalytic reaction using similar active site structural architecture. The divalent metal ions in the Class II KDO8P synthases play dual roles by modulating catalysis and affecting protein quaternary structure.