Probing the role of tightly bound phosphoenolpyruvate in Escherichia coli 3-deoxy-d -manno-octulosonate 8-phosphate synthase catalysis using quantitative time-resolved electrospray ionization mass spectrometry in the millisecond time range

ANALYTICAL BIOCHEMISTRY Analytical Biochemistry 343 (2005) 35–47 www.elsevier.com/locate/yabio

Probing the role of tightly bound phosphoenolpyruvate in Escherichia coli 3-deoxy-D-manno-octulosonate 8-phosphate synthase catalysis using quantitative time-resolved electrospray ionization mass spectrometry in the millisecond time range Zhili Li, Apurba K. Sau, Cristina M. Furdui, Karen S. Anderson ¤ Department of Pharmacology, Yale University School of Medicine, 333 Cedar Street, New Haven, CT 06520, USA Received 22 January 2005 Available online 5 May 2005

Abstract Escherichia coli 3-deoxy-D-manno-octulosonate 8-phosphate (KDO8P) synthase catalyzes the condensation of phosphoenolpyruvate (PEP) and D-arabinose 5-phosphate (A5P) to produce KDO8P and inorganic phosphate. The enzyme is often isolated with varying amounts of tightly bound PEP substrate. To better understand the role of tightly bound PEP in E. coli KDO8P synthase catalysis, a combination of transient kinetic methodologies including rapid chemical quench and mass spectrometry techniques such as time-resolved electrospray ionization mass spectrometry (ESI-TOF MS) were used to study the enzyme puriWed both in the PEPbound state and in the unbound state. Pre-steady state burst and single-turnover experiments using radiolabeled [1-14C] and [32P]A5P revealed signiWcant kinetic diVerences between these enzyme preparations. The active sites concentrations for the bound and unbound states of the enzyme were almost the same (»100%) and the product release for both states of the enzyme was rate limiting. However, the rate constant of product formation for the PEP-bound enzyme (125 s¡1) was higher than that of the unbound enzyme (46 s¡1). This was further conWrmed by single-turnover experiments using radiolabeled [32P]A5P. Interestingly, when PEP was removed from the PEP-bound enzyme and external PEP was added before the kinetic experiments, both the pre-steady state burst and the single-turnover kinetic parameters were similar to those of the enzyme puriWed in the unbound state. The rate constants of product formation were determined as 44 s¡1 (burst experiment) and 48 s¡1 (single-turnover experiment). The reaction kinetics of the E. coli KDO8P synthase was also followed by time-resolved ESI mass spectrometry. To validate the suitability of this technique for conducting enzyme kinetics, the standard reaction of p-nitrophenyl acetate hydrolysis by chymotrypsin was analyzed by stoppedXow and time-resolved ESI-TOF MS. The rate constant of p-nitrophenol formation followed by stopped-Xow spectrophotometry matched perfectly the rate constant of acetyl–chymotrypsin intermediate formation followed by time-resolved ESI-TOF MS (0.1 s¡1). The catalytic properties of the PEP-bound and unbound states of the E. coli KDO8P synthase were then studied on a millisecond time scale. The changes in the intensity of E•PEP, E• KDO8P, and E•intermediate complexes as a function of time were quantiWed and the reaction kinetics were modeled using KinTekSim simulation software. An analysis of the reaction kinetics established the kinetic competence of the intermediate based upon the rate constants for substrate decay and product formation. The ability of time-resolved ESI-TOF MS to detect and monitor the kinetics for the reaction intermediate constitutes a signiWcant advantage over the traditional rapid chemical quench technique. For all three states of the enzyme (PEP-bound, unbound, and PEP removed from the PEP-bound state) the rate constants obtained by time-resolved ESI-TOF MS matched the pre-steady state rates determined by rapid chemical quench. A comparison of reaction time courses for each state of the enzyme revealed that, in the case of PEPbound enzyme, the enzymatic reaction reached completion faster than that for the unbound state. In summary, these studies led to the conclusion that bound PEP has an important role in catalysis, maintaining the enzyme in a conformational state optimal for catalytic activity, and established the kinetic competence of the reaction intermediate. This technique has broad applicability for the

*

Corresponding author. Fax: +1 203 785 7670. E-mail address: [email protected] (K.S. Anderson).

0003-2697/$ - see front matter  2005 Elsevier Inc. All rights reserved. doi:10.1016/j.ab.2005.04.021

36

Role of phosphoenolpyruvate in KDO8P synthase catalysis / Z. Li et al. / Anal. Biochem. 343 (2005) 35–47

kinetic analysis of any enzyme system where the substrates, products, or intermediates are eluding the common detection techniques or as a method alternative to the widely used radioactivity assays.  2005 Elsevier Inc. All rights reserved. Keywords: E. coli 3-deoxy-D-manno-octulosonate 8-phosphate synthase; Phosphoenolpyruvate; Electrospray ionization mass spectrometry; Rapid chemical quench; Transient kinetics

Escherichia coli 3-deoxy-D-manno-octulosonate 8-phosphate (KDO8P)1 synthase catalyzes the condensation of phosphoenolpyruvate (PEP) and D-arabinose 5-phosphate (A5P) to form an unusual eight-carbon sugar KDO8P and inorganic phosphate (Pi) [1]. This enzymatic reaction plays an important role in the assembly of the lipopolysaccharide capsule found in most gram-negative bacteria and is therefore an attractive target for the design of novel antibacterial drugs. The mechanism of action of E. coli KDO8P synthase has been studied extensively [1–8]. Steady state inhibition studies showed that the reaction for KDO8P synthesis is a sequential process in which the binding of PEP precedes that of A5P and the release of Pi precedes that of KDO8P under multiple-turnover conditions [4]. However, pre-steady state kinetics under single-turnover conditions showed that the enzyme can bind either A5P or PEP [5]. Earlier studies have shown that E. coli KDO8P synthase was isolated with tightly bound PEP [5]. It was also found that the enzyme is less stable and loses its activity upon removal of bound PEP [5]. The crystal structure of E. coli enzyme has provided a wealth of information on the organization of the active site and on the interactions between the subunits. A model structure of the active site for the enzyme in which the phosphate groups of PEP and A5P are 13.0 Å apart, allowing KDO8P synthesis to proceed via a labile hemiketal phosphate enzyme intermediate, was proposed [2,6,7]. So far few methods for directly detecting and identifying enzyme–substrate and enzyme–product complexes during the catalysis, whose noncovalent interactions constitute the essential basis of molecular recognition in the biological system, are known. The technique of electrospray ionization mass spectrometry (ESI-MS) [9] has contributed signiWcantly toward biological and biomedical applications. This is due to its ability to form gasphase macromolecular ions directly from thermally unstable and polar compounds at atmospheric pressure via protonation or deprotonation and ion evaporation. This technique has also been established to be a rapid, sensitive, and highly selective alternative for generating 1 Abbreviations used: KDO8P, 3-deoxy-D -manno-octulosonate 8-phosphate; PEP, phosphoenolpyruvate; A5P, D-arabinose 5-phosphate; Pi, inorganic phosphate; HPLC, high-performance liquid chromatography; Tris, tris(hydroxymethyl)aminomethane; EDTA, ethylenediaminetetraacetic acid; TEAB, triethyl ammonium bicarbonate; ESI-MS, electrospray ionization mass spectrometry; ESI-TOF MS, electrospray ionization time of Xight mass spectrometry.

intact weakly bound complexes and studying noncovalent interactions. Most studies using this approach rely on its stability to transfer noncovalent complexes of solution phase into the gas phase [10–12]. It is also recognized as a technique capable of detecting protein conformational changes [13,14]. In recent years, a series of studies have demonstrated the broad applicability of ESI-MS to monitor enzyme reaction kinetics [10,15–35]. Most of these studies have employed an oV-line ESI-MS approach to monitor reaction kinetics and substrate/ inhibitor binding [16,22–24,27–30]. More recently, a time-resolved continuous-Xow ESI-MS strategy has been used to delineate the kinetic pathway [10,15,21,26]. Many of the intermediates, substrates, or products in the enzymatic reactions are unstable under the common quenching conditions. The time-resolved continuousXow ESI-MS oVers the advantage of not requiring a quenching reagent, thus broadening the applicability of ESI-MS as a tool for monitoring enzyme reaction kinetics. To demonstrate the general utility of time-resolved ESI-MS, in the current study, we illustrate the ability of this approach to (1) accurately deWne reaction kinetics in a quantitative manner, (2) examine reaction kinetics in a 6- to 7-ms time range, (3) access the diVerential catalytic eYciencies of an enzyme system in a ligand-free or ligand-bound state, and (4) establish that a labile species is a kinetically competent reaction intermediate. We have recently studied the catalysis of E. coli KDO8P synthase containing tightly bound PEP and a labile hemiketal phosphate enzyme intermediate has been identiWed as a noncovalent complex with protein on a millisecond time scale using ESI-TOF MS [10]. The reaction catalyzed by E. coli KDO8P synthase is shown in Scheme 1. During the course of the preparation of the KDO8P synthase enzyme, we have isolated diVerent batches of enzyme with varying amounts of bound PEP as conWrmed by incubating with radiolabeled [1-14C]A5P experiments (see Experimental) that appeared to have diVerent catalytic properties. These preparations of the enzyme provided us with an opportunity to study the role of tightly bound PEP in the enzyme catalysis. In this work, the reaction of p-nitrophenyl acetate hydrolysis by chymotrypsin [36–43] was used to validate the previously described time-resolved ESI-MS as a quantitative technique suitable for kinetic analysis of enzymatic reactions. As an application of this technique, a combination of rapid chemical quench and quantitative time-resolved ESI-MS techniques was then used to

Role of phosphoenolpyruvate in KDO8P synthase catalysis / Z. Li et al. / Anal. Biochem. 343 (2005) 35–47

37

Scheme 1. E. coli KDO8P synthase reaction pathway.

dissect the catalytic properties of PEP-bound and unbound states of the E. coli KDO8P synthase on a millisecond time scale. The results suggest the role of PEP in maintaining conformational stability and integrity of the enzyme active site and establish the kinetic competence of the labile hemiketal phosphate enzyme reaction intermediate.

to remove metal ions according to the procedure provided by Alltech Associates, and then the pHs of A5P and PEP eluate were separately adjusted to 7.8 using ammonium hydroxide. The concentrated A5P solution was then diluted to 60 M by addition of 10 mM ammonium acetate buVer, pH 7.8. Preparation of [1-14C]A5P and [32P]A5P

Experimental Materials D-Arabinose 5-phosphate disodium salt (Cat. A-2013), ammonium acetate (Cat. 43,131-1), phosphoenolpyruvic acid trisodium salt hydrate (Cat. P-7002),
-chymotrypsin (Cat. C-4129), and p-nitrophenyl acetate (Cat. N8130) were obtained from the Sigma Chemical (St. Louis, MO, USA). IC-H Maxi-Clean Cartridge (Cat. 30256) was purchased from Alltech Associates (DeerWeld, IL, USA). Ammonium hydroxide (Cat. 9721-01) was obtained from a Division of Mallinkrodt Baker (Phillipsburg, NJ, USA). Fused silica capillary (TSP030150; i.d. of 30 § 3 m) and a zero dead volume-mixing tee (MY1XCS6) were obtained from Polymicro Technologies (Phoenix, AZ, USA) and Valco Instrument (Houston, TX, USA), respectively.

[1-14C]A5P (3 mCi/mmol) and [32P]A5P were synthesized enzymatically by using hexokinase (20 mg/ml) and [1-14C]arabinose (4 mCi/mmol; ICN) and [-32P]ATP (Amersham), respectively. The radiolabeled A5P was puriWed by Q-Sepharose column (Amersham Pharmacia). For [1-14C]A5P a linear gradient of 20 mM to 1 M TEAB, pH 9.0, was applied and the fractions containing radiolabeled A5P were pooled and lyophilized. For [32P]A5P a linear gradient of zero to 1 M ammonium acetate, pH 7.5, was used. Preparation and puriWcation of E. coli KDO8P synthase Escherichia coli KDO8P synthase was overexpressed and puriWed as previously reported [4]. The monomeric concentration of the enzyme was determined using an extinction coeYcient of 15,219 M¡1 cm¡1 at 280 nm. E. coli KDO8P synthase in the PEP-bound and unbound state was puriWed and further characterized.

Preparation of
-chymotrypsin Salt-free
-chymotrypsin was generated by passing the
-chymotrypsin through a Sephadex G-100 gel Wltration column equilibrated with 20 mM ammonium acetate, pH 7, followed by 4 h dialysis against the same buVer. All kinetic experiments described below were performed with salt-free
-chymotrypsin. Preparation of A5P and PEP solutions Solutions (1 ml) of 5 mM A5P and 5 mM PEP in 10 mM ammonium acetate buVer (pH 7.8) were separately passed through two IC-H Maxi-Clean Cartridges

Determination of bound PEP and A5P substrates in E. coli KDO8P synthase To determine the number of equivalents of bound PEP, a radiolabeled experiment was carried out by preincubating the enzyme with [1-14C]A5P. In the HPLC analysis the amount of radiolabeled KDO8P peak found corresponded to 0.8 and zero equivalents of PEP-bound to the enzymes. The 0.8 equivalents are relative to the active sites concentration which in the case of the KDO8P synthase used in these studies is equal to protein concentration. The lack of cold A5P bound to the enzyme was veriWed by preincubating the enzyme with

38

Role of phosphoenolpyruvate in KDO8P synthase catalysis / Z. Li et al. / Anal. Biochem. 343 (2005) 35–47

radiolabeled [1-14C]PEP. No radiolabeled KDO8P peak was found in the HPLC analysis, suggesting that the enzymes were not bound with cold A5P. PEP removal from the PEP-bound state of E. coli KDO8P synthase Previous studies had shown that the E. coli KDO8P synthase was isolated with tightly bound PEP [5,10]. To make PEP-free enzyme, E. coli KDO8P synthase with 0.8 equivalent of bound PEP was incubated with cold A5P (the concentration of A5P was kept 1.5 times higher than that of the enzyme), and the mixture was kept at 4 °C for about 1 h. To ensure that all the PEP is removed from the enzyme, a small amount of the reaction mixture was incubated with [1-14C]A5P and the mixture was injected into the HPLC with radioactivity detection. No radiolabeled KDO8P peak was found in the HPLC analysis. The mixture was then dialyzed against two changes of 4 liters of 20 mM ammonium acetate, pH 7.8, at 4 °C. The reaction mixture was passed through Bio-spin 6 column (Bio-Rad). A small amount of the reaction mixture was incubated with [1-14C]PEP and the mixture was injected into the HPLC. No radiolabeled KDO8P peak was found in the HPLC analysis. This suggests that the enzyme is free from both bound PEP and excess A5P. To remove excess A5P from the reaction mixture, it was necessary to carry out both dialysis and Bio-spin column. The stoichiometric amount of cold A5P added to the enzyme is also important in the reaction, as removal of excess cold A5P is extremely diYcult in PEP-free enzyme, indicating that excess A5P may also bind very tightly to the PEP-free enzyme. Rapid chemical quench experiments Rapid quench experiments were performed using a Kintek RFQ-3 Rapid Chemical Quench (Kintek Instruments, Austin, TX, USA) as previously described [44]. The reaction was initiated by mixing the enzyme solution (15 l) with the radiolabeled substrate A5P (15 l). In all cases, the concentrations of the enzyme and substrate cited in the text are those after mixing and during the reaction. The reaction was then quenched with 67 l of 0.6 N KOH. For [32P]A5P, the reaction was quenched with 67 l triethyl amine. It was found that the substrates and products were stable under these conditions. The substrates and products were separated and quantiWed 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 Xow rate of 1 ml/min. A gradient separation was employed using mobile phase A (20 mM TEAB, pH 9.0) and mobile phase B (1 M TEAB, pH 9.0). A linear gradient program was as follows: 0–100% B from 0 to 30 min, followed by reequilibration. The retention times

were 9.1 and 14.2 min for A5P and KDO8P, respectively. During the HPLC analysis a small amount of cold PEP was always injected as an internal standard as its retention 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. Pre-steady state burst experiments of E. coli KDO8P synthase For PEP-bound E. coli KDO8P synthase, the enzyme (17.5 M) preincubated with PEP (250 M) was reacted with 61.3 M [1-14C]A5P in 10 mM ammonium acetate buVer, pH 7.5. The reaction was quenched with 0.4 N KOH. A similar experiment was done for the enzyme puriWed in the unbound state (enzyme 28 M, A5P 61.3 M, and PEP 250 M) and for the enzyme puriWed in the PEP-bound state but from which the PEP was removed (enzyme 24 M, A5P 61.3 M, and PEP 250 M). Single-turnover experiments of E. coli KDO8P synthase PEP-bound E. coli KDO8P synthase (30 M) was reacted with 100 M [32P]A5P. For the unbound and the PEP-removed state, the enzyme (30 M) preincubated with PEP (30 M) was reacted with 100 M [32P]A5P in 10 mM ammonium acetate buVer, pH 7.5. The reactions were quenched with triethyl amine. Time-resolved ESI-TOF MS experiments ESI mass spectra were recorded with an Ettan electrospray ionization time-of-Xight mass spectrometer (Amersham Biosciences) at collision voltage 120 V (laboratory frame). The typical operating conditions for ESITOF MS ion source are as follows: drying gas Xow rate, 16 L/min; drying gas temperature, 180 °C; nebulizer gas pressure, 5.6 bar. The ESI-TOF MS was operated in positive-ion mode. Each mass spectrum showing 16+ charge state of monomer of KDO8P synthase represents the average of 15 mass spectra recorded in trap pulse mode with 60,000 pulses per mass spectrum in the mass range between m/z 1900 and 2150. The method is based on the continuous-Xow setup [10]. Two 1000-l syringes were advanced simultaneously by a syringe pump (Model PHD 2000 infusion; Harvard Apparatus, South Natick, MA, USA) at four Xow rates, 220, 160, 100, and 10 l/ min (after mixing and during the reaction). Syringe 1 contained the solution of 80 M enzyme in 10 mM ammonium acetate buVer, pH 7.8; syringe 2 contained a solution of either 60 M A5P in the same buVer or buVer only. Both syringes were connected to a “reaction” fused silica capillary by a zero dead volume-mixing tee. The reaction time is controlled by changing the Xow rate for the same length of the “reaction” capillary between the

Role of phosphoenolpyruvate in KDO8P synthase catalysis / Z. Li et al. / Anal. Biochem. 343 (2005) 35–47

mixing point and the other end of the fused silica capillary located in the electrospray source. The length of the “reaction” fused silica capillary and Xow rates allowed reaction time points of 7, 10, 16, and 160 ms, corresponding to Xow rates of 220, 160, 100, and 10 l/min, respectively. The peak intensities ligand-free enzyme were kept constant at diVerent Xow rates by adjusting the position of electrospray ionization probe. The length of the electrospray ionization probe is 3.7 cm and was custom designed and built by Analytica of Branford (Branford, CT, USA). Pre-steady state kinetic analysis of
-chymotrypsin reaction with p-nitrophenyl acetate

39

Pre-steady state single-turnover experiments KDO8P formation (M) was plotted against time (s) and data were Wt to a single exponential equation (Eq. (2)). (2) C represents KDO8P concentration at time t, the amplitude C0 corresponds to active sites concentration, and k1 is the rate constant for KDO8P formation (s¡1). Data were analyzed using Kaleidagraph version 3.52, released June 17, 2002 by Synergy Software. KinTekSim analysis of KDO8P synthase reaction

Stopped-Xow spectrophotometry
-Chymotrypsin (30 M) was mixed with 30 M p-nitrophenyl acetate in a Kintek stopped-Xow spectrophotometer. The formation of the p-nitrophenol was followed at 400 nm ( 5.34 mM¡1 cm¡1 in 50 mM ammonium acetate buVer, pH 7, as determined in our laboratory). Because the active site concentration was slightly lower than enzyme concentration (25 M versus 30 M), data were Wt using a burst equation (exponential followed by linear equation as described under Data analysis). Time-resolved ESI-TOF MS experiments The decay of
-chymotrypsin and formation of the acetyl–chymotrypsin intermediate is observed over a time course of 45 s. Charge states 10+ and 11+ were used to quantify the relative intensities of
-chymotrypsin and acetyl–chymotrypsin peaks. The instrument parameters and the conditions of operation were identical to those described above for the analysis of E. coli KDO8P synthase reaction, the only diVerence being the addition of an aging tube (Upchurch ScientiWc, Cat. 1575; i.d. 200 m, length 24 cm) between the mixing tee and the electrospray needle to accommodate reaction times up to 45 s. Data analysis Pre-steady state burst experiments KDO8P formation (M) was plotted against time (s) and data were Wt to a burst equation (Eq. (1)). (1) C represents KDO8P concentration at time t, the amplitude C0 corresponds to active sites concentration, k1 is the rate constant for KDO8P formation (s¡1), and k2 is the rate of KDO8P release (s¡1). The percentage of active enzyme can be calculated by dividing C0 with the enzyme concentration. A similar equation was used to Wt the p-nitrophenol formation by chymotrypsin.

Data generated by time-resolved ESI-TOF MS analysis of the E. coli KDO8P synthase reaction (7–160 ms) were quantiWed based upon the relative intensity of each species. To provide an estimation of the reaction kinetics for substrate conversion to product and the kinetics of the enzyme intermediate, the data were simulated using KinTekSim software (Kintek Instruments, TX, USA). The reaction mechanism used in the simulation was

where ES represents the enzyme–substrate complex (in this case enzyme–PEP), EI the enzyme–intermediate complex, and EP the enzyme–product (KDO8P) complex. k1 and k2 are the rate constants of substrate decay and product formation, respectively.

Results and discussion Establishing the quantitative aspects of time-resolved ESI-TOF MS To investigate the possible application of timeresolved ESI-TOF MS to the quantitative analysis of an enzymatic reaction, the classic chymotrypsin reaction with p-nitrophenyl acetate was followed by ESI-TOF MS and stopped-Xow techniques. The reaction proceeds through a covalent acetyl–chymotrypsin intermediate as shown in Scheme 2 [36–43]. The reaction was performed under single-turnover conditions using equimolar concentrations of enzyme and substrate. Under these conditions, for every mole of acetyl–chymotrypsin formed there is a mole of p-nitrophenol released. This allowed for a side-by-side comparison between the time-resolved ESI-TOF MS technique (following acetyl–chymotrypsin formation) (Figs. 1A–C) and the stopped-Xow absorbance technique (following p-nitrophenol formation at 400 nm) (Fig. 1D). The experimental results revealed very similar reaction kinetics illustrating the ability to accurately quantitate enzymatic activities. The rate

40

Role of phosphoenolpyruvate in KDO8P synthase catalysis / Z. Li et al. / Anal. Biochem. 343 (2005) 35–47

Scheme 2. Chymotrypsin reaction with p-nitrophenyl acetate. Step 1: chymotrypsin acetylation reaction with the formation of acetyl–chymotrypsin intermediate and p-nitrophenol. Step 2: deacetylation step with the regeneration of the free enzyme and release of acetate.

Fig. 1. ESI-TOF MS to monitor reaction kinetics of chymotrypsin reaction with p-nitrophenyl acetate under single-turnover conditions. Enzyme and substrate concentrations were 30 M. (A) The reaction was monitored by time-resolved ESI-TOF MS. The decay of chymotrypsin and formation of the acetyl–chymotrypsin intermediate is observed over a time course of 45 s. (B) Charge state 11+ showing the relative distribution of chymotrypsin and acetylated chymotrypsin at the 5-s time point. (C) Kinetic analysis of chymotrypsin decay and acetyl–chymotrypsin formation. For both traces the amplitude was 25 M and the rate constant was 0.1 s¡1. (D) The formation of the p-nitrophenol (400 nm,  5.34 mM¡1 cm¡1) followed by stopped Xow absorbance. The data were Wt using a burst equation with amplitude corresponding to active site concentration of 25 M and a rate constant of product formation of 0.1 s¡1.

constant of acetyl–chymotrypsin or p-nitrophenol formation under pre-steady state conditions as determined by both assays was 0.1 s¡1. Data for the relative ionization eYciency of chymotrypsin and acetyl–chymotrypsin are presented in Fig. 2 and Table 1. These data clearly illustrate that the ionization eYciencies for chymotrypsin and acetyl–chymotrypsin in the ESI-TOF MS are very similar as illustrated by the identical total ion current for the two species at diVering amounts of the two species. In addition, the individual ion currents for the two species are observed to vary linearly and symmetrically as a function of their formation and disappearance. However, for every mole of acetyl–chymotrypsin generated there should be a mole of p-nitrophenol

formed. As Fig. 1 shows, the weaker E•p-nitrophenol complex (noncovalent interaction) has lower intensity than the E•acetyl complex (covalent interaction) (10% of the acetyl–chymotrypsin complex). A signiWcant part of the bound p-nitrophenol dissociates from the enzyme during the MS experiments. This is not unexpected, since the intensity of the noncovalent complex is a function of its dissociation constant and MS conditions. Since the amount of p-nitrophenol formed is linear as a function of time, the measurement of the relative amount of p-nitrophenol present at each time point allows an accurate estimation of the rate constant of p-nitrophenol formation. These studies conWrm that the time-resolved ESI-TOF MS method can in-fact quantitate the rate

Role of phosphoenolpyruvate in KDO8P synthase catalysis / Z. Li et al. / Anal. Biochem. 343 (2005) 35–47

41

Fig. 2. Total ion current () and ionization eYciencies of chymotrypsin (䊉) and acetylated chymotrypsin (䊊).

Table 1 Time, s

0 5 600

Peak intensity Chymotrypsin Acetyl– Chymotrypsinchymotrypsin p-nitrophenol complex 33844 0 0 18044 11399 0 0 28248 3472

Standard deviationa

1.068 0.92 1.00

In all three cases the Xow rate was 20 l/min, and the spray conditions were 120 V and 180 °C. a Standard deviation was calculated by dividing the sum of the intensity of all species present at each time point by the mean of the sums for the three time points.

constants of various protein species formation and decay and can provide kinetic parameters very comparable to those observed by other methods. Application of transient kinetics including rapid chemical quench and quantitative time-resolved ESI-TOF MS to probe the role of tightly bound phosphoenolpyruvate in E. coli KDO8P synthase Pre-steady state burst experiments to assess enzymatic activity To determine the active sites concentration and rate constant of product formation for the PEP-bound and unbound states of E. coli KDO8P synthase, a series of pre-steady state burst experiments were carried out using radiolabeled [1-14C]A5P as substrate. In this type of experiment, the radiolabeled substrate is used in slight excess over the enzyme, such that the Wrst enzyme turnover and multiple turnovers can be examined. The burst experiments for the two batches of the enzyme are shown in Figs. 3A and B. An initial burst phase of product [1-14C]KDO8P formation followed by a slower linear phase was observed for both types of enzyme. A presteady state burst of the product KDO8P formation is

Fig. 3. Pre-steady state burst experiments for E. coli KDO8P synthase. (A) A solution containing PEP-bound enzyme (17.5 M) preincubated with additional PEP (250 M) was mixed with [1-14C]A5P (61.3 M) at 25 °C (Wnal concentration after mixing). The reaction was terminated by quenching with 0.4 N KOH, and the formation of product KDO8P was monitored by HPLC with radioactive detection. (B) Same as (A) for the enzyme puriWed in the unbound state. The concentrations of the enzyme, radiolabeled A5P, and PEP were 24, 61.3, and 250 M, respectively. (C) Same as (A) for the enzyme puriWed in the PEPbound state but from which the PEP was removed. The concentrations of the enzyme, radiolabeled A5P, and PEP were 28, 61.3, and 250 M, respectively. The rate constants for the fast and slow phases are shown in Table 2.

indicative of a mechanism in which the chemical catalysis does not limit the overall reaction but, rather, the release of product is rate limiting. The rate constants of product formation were 125 s¡1 for the PEP-bound enzyme and 46 s¡1 for the unbound state of the enzyme.

42

Role of phosphoenolpyruvate in KDO8P synthase catalysis / Z. Li et al. / Anal. Biochem. 343 (2005) 35–47

Table 2 Summary of pre-steady state and steady state kinetic parameters Enzyme

Active sites concentration (%)

Product formation from burst expts. (s¡1)

Product formation from singleturnover expts. (s¡1)

Steady state rate from burst expts. (s¡1)

Steady state rate from steady state expts. (s¡1)

PEP-bound Unbound PEP removed from PEP-bound

»95 »100 »100

125 § 16 46 § 10 44 § 8

100 § 11 39 § 7 48 § 9

2.3 § 0.2 0.6 § 0.2 0.2 § 0.05

3.5 § 0.2 0.5 § 0.08 0.3 § 0.06

The rate constants of product release were 2.3 and 0.6 s¡1, respectively (Table 2). It was found that the rate constants for the product release were in good agreement with the steady state rate determined by following the consumption of PEP at 232 nm under steady state conditions (Table 2). Interestingly, when PEP was removed from the PEP-bound enzyme and external PEP was added before the kinetic experiments, both the presteady state burst and the single-turnover kinetic parameters (see below) were similar to those of the enzyme puriWed in the unbound state (Fig. 3C and Table 2). The rate constant of product formation was determined as 44 s¡1 and the rate constant of product release as 0.2 s¡1. This observation suggests that the originally bound PEP has an important role in catalysis. This was further investigated by single-turnover and ESI-TOF MS experiments described below. Single-turnover experiments to determine the rate of catalysis A complimentary approach to examine the enzyme catalysis is the single-turnover experiment, which measures the rate of chemical conversion of substrate to product at the active site of the enzyme under equimolar concentrations of enzyme and substrate. These experiments were carried out using radiolabeled [32P]A5P as substrate. As shown in Figs. 4A–C, the data Wtted well to a single exponential equation and the rate constants are summarized in Table 2. These rate constants are consistent with those determined from the exponential phase in the pre-steady state burst experiment and provide further support that the originally bound PEP has a signiWcant role in catalysis. ESI-TOF MS experiments to monitor KDO8P synthase containing tightly bound PEP The initial assessment of the amount of bound PEP in the PEP-bound and unbound states of the enzyme was performed using ESI-TOF MS as shown in Figs. 5A and B. These data show the positive-ion ESI mass spectra of PEP-bound and unbound states of the enzyme at 16+ charge state and at four Xow rates: 220, 160, 100, and 10 l/min. The Wnal enzyme concentration in each case was 40 M. The intensities of the E•PEP complex of PEP-bound enzyme Figs. 5A were found to be almost the same at diVerent Xow rates. This indicates that the nonco-

valent complex E•PEP was formed in the solution phase and not in the gas phase and that the interaction between enzyme and PEP is speciWc. Incubation experiments with radiolabeled A5P indicated that the PEP-bound KDO8P synthase contained approximately 0.8 equivalents of PEP (see Experimental). In the ESI-TOF MS experiments, the high voltage (»120 V) required for maintaining eYcient ion desolvation/ionization at the high Xow rates necessary for temporal resolution retained »20% of the noncovalent complexes for the PEP-bound enzyme. The unbound state of the protein showed a spectrum similar to that of the PEP-bound enzyme from which the PEP was removed. As expected, these mass spectra lack the peak of the E•PEP complex. In removing the PEP by incubating the enzyme with a slight excess of A5P there was a small amount of A5P remaining. While not so obvious in the spectra in Fig. 5C, the presence of residual A5P became apparent when an equivalent of PEP was added back to the KDO8P synthase since a small peak of KDO8P was observed even after overnight dialysis (Fig. 6B) (see below). Thus the radiolabeled and mass spectrometric data are complementary in establishing the presence or absence of tightly bound PEP in each batch of enzyme characterized. Reconstitution of KDO8P synthase with PEP Earlier experiments attempting to reconstitute the enzyme puriWed in the PEP-bound state from which the PEP was removed with radiolabeled PEP suggested that E•PEP complex formed in this manner was less stable and catalytically productive [5]. This was conWrmed by incubating the unbound state of the enzyme and the PEP-bound state from which the PEP was removed with an equivalent of PEP (40 M) followed by ESI-TOF MS analysis. The enzyme samples were examined at four Xow rates: 220, 160, 100, and 10 l/min. The data shown in Figs. 6A and B illustrate the intensity change of the E•PEP complex detected by ESI-TOF MS for the two states of the enzyme under the same conditions. The peak of the E•PEP complex was detected at diVerent Xow rates, apart from the peak of the enzyme labeled E. Similar to the data illustrated in Figs. 5A–C, in each case the intensity of the E•PEP complex was found to be almost the same irrespective of the Xow rate. As illus-

Role of phosphoenolpyruvate in KDO8P synthase catalysis / Z. Li et al. / Anal. Biochem. 343 (2005) 35–47

43

The spectra in Fig. 6B show peaks corresponding to the E and E•PEP complex and, in addition, a peak corresponding to the E•KDO8P complex, since as indicated earlier, a small amount of KDO8P product remained after dialysis. Several observations are of note in comparison to the data presented in Fig. 5. The intensity of the E•PEP complex prepared by reconstitution with PEP is decreased about 40% as compared with that with the PEP-bound state (Fig. 5A and 6B). These results indicate that the E•PEP complex in the gas phase does not depend on the amount of external PEP added in the solution. This also suggests that the amount of E•PEP complex detected by ESI-TOF MS is directly proportional to that in the solution phase. The data shown in Fig. 6 also point to an altered ability of the unbound state of the enzyme (as isolated or generated by PEP removal from the PEP-bound state) to tightly bind PEP in reconstitution experiments. Apparently, the presence of KDO8P (Fig. 6B) keeps the enzyme stable and helps to maintain its original conformation. It was found that the enzyme isolated in the PEP-bound state was stable after overnight dialysis in 20 mM ammonium acetate, pH 7.8, at 4 °C. However, the unbound state of the enzyme was less stable and dialysis resulted in enzyme precipitation. A plausible explanation for this is the change in conformation of the enzyme active site upon removal of PEP from the PEP-bound enzyme even though the product KDO8P may, to some extent, aid in maintaining its original conformation. For the unbound enzyme, the original conformation of the enzyme at the active site may be altered in such a way that it could not be fully recovered (only »20% recovery) even after addition of equimolar amount of external PEP. Examination of the reaction time course for E. coli KDO8P synthase catalysis using time-resolved ESI-TOF MS

Fig. 4. Single-turnover experiments in E. coli KDO8P synthase. (A) A solution containing 30 M PEP-bound enzyme was mixed with 100 M [32P]A5P at 25 °C (Wnal concentration after mixing). The reaction was terminated by quenching with triethyl amine and then the solution was mixed with CHCl3. The aqueous layer was collected and the formation of product [32P]KDO8P was monitored by HPLC with radioactive detection. (B) and (C) are the same experiments for the unbound and PEP-removed states of the enzyme, respectively. A solution containing 30 M enzyme and 30 M PEP was mixed with 100 M [32P]A5P at 25 °C (Wnal concentration after mixing). The curves represent a Wt of the data to a single exponential equation. The rate constants are shown in Table 2.

trated in Fig. 6A, reconstitution of the unbound state of the enzyme with PEP resulted in a relatively weak peak for the E•PEP complex that is approximately twofold less than that observed in Fig. 6B which shows the reconstitution of the PEP-bound state from which the PEP was removed and then added back to the enzyme.

Rapid chemical quench experiments (Figs. 3 and 4) suggested that the PEP-bound enzyme had a higher catalytic activity. Time-resolved ESI-TOF MS was used to monitor real-time catalytic activity of the PEP-bound and unbound states and the PEP-bound state from which the PEP was removed. The Wnal enzyme concentration used in each case was 40 M. The data in Figs. 5 and 6 establish that the relative intensities of the noncovalent E•PEP complex in each KDO8P synthase preparation do not change when the Xow rate is varied from 220, 160, 100, and 10 l/min, corresponding to reaction times of 7, 10, 16, and 100 ms. The reaction time course for each type of enzyme (PEP-bound, unbound reconstituted with PEP, PEP-bound with PEP removed and then reconstituted with PEP) was examined using rapid mixing time-resolved ESI-TOF MS. The enzyme containing PEP (in 10 mM ammonium acetate buVer, pH 7.8) was combined in a mixing tee with a solution of the second

44

Role of phosphoenolpyruvate in KDO8P synthase catalysis / Z. Li et al. / Anal. Biochem. 343 (2005) 35–47

Fig. 5. Mass spectra of 40 M KDO8P synthase in the positive-ion mode at 16+ charge state at the Xow rates 220, 160, 100, and 10 l/min (after mixing). One syringe contained 80 M enzyme and an other contained 10 mM ammonium acetate buVer. (A) PEP-bound enzyme; (B) enzyme puriWed in the unbound state; (C) PEP removed from the PEP-bound enzyme.

Fig. 6. Mass spectra of 40 M KDO8P synthase reconstituted with PEP in the positive-ion mode at 16+ charge state at the Xow rates 220, 160, 100, and 10 l/min (after mixing). One syringe contained 80 M enzyme reconstituted with PEP and an other contained 10 mM ammonium acetate buVer. (A) Enzyme puriWed in the unbound state incubated with an equimolar amount of PEP in 10 mM ammonium acetate buVer, pH 7.8, overnight and at 4 °C; (B) PEP-bound enzyme from which the PEP was removed was incubated with equimolar PEP in 10 mM ammonium acetate buVer, pH 7.8, overnight at 4 °C.

substrate A5P (30 M, Wnal concentration) in the same buVer using two 1000-l syringes and a syringe pump. The reaction mixture from the mixing tee was continuously introduced into the mass spectrometer through an in-house built electrospray probe. Mass spectra were recorded at four reaction times: 7, 10, 16, and 160 ms

(Figs. 7A–C). The enzyme binary complexes of substrates (E•PEP and E•A5P) and of products (E•KDO8P and E•Pi) along with the free enzyme were detected in all three cases. The data show that, as the reaction progresses, the intensity of the peaks corresponding to the product complexes (E•KDO8P and E•Pi) increase with

Role of phosphoenolpyruvate in KDO8P synthase catalysis / Z. Li et al. / Anal. Biochem. 343 (2005) 35–47

45

Fig. 7. Mass spectra of 40 M KDO8P synthase catalysis recorded at 16+ charge state at diVerent reaction time points, 7, 10, 16, and 160 ms. (A) Time course for PEP-bound enzyme. A solution of 80 M PEP-bound enzyme was on-line mixed with a solution of 60 M A5P. (B) Time course for the unbound state of the enzyme. A solution of 80 M PEP-free enzyme incubated with an equimolar amount of PEP overnight at 4 °C was on-line mixed with a solution of 60 M A5P. (C) Time course for the PEP-bound enzyme from which the PEP was removed. A solution of 80 M enzyme incubated with an equimolar amount of PEP overnight at 4 °C was on-line mixed with a solution of 60 M A5P.

longer reaction times and that of the peaks corresponding to the substrate complexes (E•PEP and E•A5P) simultaneously decrease. As illustrated in Fig. 7A, the reaction time course for PEP-bound KDO8P synthase shows a much more rapid disappearance of the E•A5P complex for the limiting substrate A5P and complete disappearance at 160 ms. Also an enzyme intermediate complex, E•I (Scheme 1), previously characterized as a hemiketal phosphate species bound to enzyme, is observed at an earlier time which disappears as the KDO8P product is formed [10]. The kinetics of the enzyme-bound species observed in the MS spectrum was then simulated based on the mechanism involving a single reaction intermediate as described under Experimental. The data were best Wt with a rate constant of intermediate formation of 95 s¡1 (Fig. 8) and a rate constant of intermediate decay of 500 s¡1. This is in good agreement with the single-turnover experiments when a rate constant of 100 s¡1 was observed for the formation of product (Table 2 and Fig. 4). It also established the intermediate as kinetically competent for the KDO8P

synthase reaction. In contrast to the reaction time course for the PEP-bound enzyme, at the 160 ms time point, a small amount of A5P is remaining for the other two enzyme preparations (Figs. 7B and C). A t1/2 estimation of A5P decay and KDO8P formation matches the rate constants obtained from the pre-steady state experiments. These data indicate that the turnover rate for the PEP-bound KDO8P synthase was faster than those for the two other batches of enzyme. These data are in good agreement with the data presented earlier in Figs. 3 and 4. Thus the results obtained from both rapid chemical quench and time-resolved ESI-TOF MS experiments suggest that the presence of bound PEP at the active site of the enzyme plays a key role in maintaining the original conformation of the enzyme for optimal catalytic activity in the eYcient conversion of substrates to products. Moreover, the KinTekSim analysis of the reaction time course as determined by the time-resolved ESI-TOF MS technique showed that the formation and decay of the reaction intermediate are kinetically competent with the kinetics of substrate decay and product formation.

46

Role of phosphoenolpyruvate in KDO8P synthase catalysis / Z. Li et al. / Anal. Biochem. 343 (2005) 35–47

The advantage of the time-resolved ESI-MS technique is clearly the increased sensitivity of detection and the ability to examine chemically labile species. The ability to couple this technique to real-time mass and fragmentation measurements for resolving and identifying protein species and posttranslational modiWcation broadens its applicability to the Weld of signal transduction and ultimately to the emerging Weld of systems biology.

Acknowledgments

Fig. 8. KinTekSim simulation of the E. coli KDO8P synthase reaction followed by time-resolved ESI-TOF MS. The data for A5P decay (䊉), KDO8P formation (䊊), and intermediate decay and formation (䉳) are presented. The rate constant of intermediate formation was 95 s¡1 and the rate constant of intermediate decay was 500 s¡1.

We gratefully acknowledge Dr. Timor Baasov at Department of Chemistry, Institute of Catalysis Science and Technology, Technion, Haifa, Israel for providing us the clone and puriWcation protocol of E. coli KDO8P synthase. This work was supported by NIH GM 61413, NIH GM GM 71805, and NSF 987677 to K.S.A.

References Conclusions The results from the present study demonstrate that the coupling of on-line rapid-mixing devices with ESITOF MS provides a new and powerful tool for monitoring the pre-steady state of the enzymatic reactions. Analysis of KDO8P synthase catalysis by rapid mixing, ESITOF MS in a millisecond time scale enabled direct detection of noncovalent complexes of the enzyme with its substrates or products and the change of their intensities. Both mass spectrometric and rapid chemical quench data suggest that (1) PEP in the E•PEP complex speciWcally binds to the active site of the enzyme; (2) PEP originally bound to the enzyme plays a vital role in maintaining a conformation at the active site optimal for stability and catalytic activity, (3) the enzyme puriWed in the unbound state or rendered PEP free by PEP removal from the bound state binds the substrate PEP more weakly after reconstitution with PEP most likely due to a change in conformation of the enzyme at the active site resulting in lower catalytic activity, and (4) the reaction proceeds through a labile, kinetically competent, hemiketal phosphate intermediate. We have applied the time-resolved ESI-MS technique to a number of diVerent systems that include thioesterase, dehalogenase, Aquifex pyrophilus and Helicobacter pylori KDO8P synthase, and most recently several tyrosine kinases. In the case of Wbroblast growth factor receptor 1 kinase we were able to follow the autophosphorylation reaction as function of time and to determine for the Wrst time the rate constant of each individual phosphorylation event (C.M. Furdui et al., unpublished).

[1] P. Ray, PuriWcation and characterization of 3-deoxy-D-mannooctulosonate-8-phosphate synthetase from Escherichia coli, J. Bacteriol. 141 (1980) 635–644. [2] O. Asojo, J. Friedman, N. Adir, V. Belakhov, Y. Shoham, T. Baasov, Crystal structures of KDOP synthase in its binary complexes with the substrate phosphoenolpyruvate and with a mechanismbased inhibitor, Biochemistry 40 (2001) 6326–6334. [3] T. Baasov, S. SheVer-Dee-Noor, A. Kohen, A. Jakob, V. Belakhov, Catalytic mechanism of 3-deoxy-D-manno-2-octulosonate-8-phosphate synthase. The use of synthetic analogues to probe the structure of the putative reaction intermediate, Eur. J. Biochem. 217 (1993) 991–999. [4] A. Kohen, A. Jakob, T. Baasov, Mechanistic studies of 3-deoxy-Dmanno-2-octulosonate-8-phosphate synthase from Escherichia coli, Eur. J. Biochem. 208 (1992) 443. [5] P.H. Liang, J. Lewis, K.S. Anderson, A. Kohen, F.W. D’Souza, Y. Benenson, T. Baasov, Catalytic mechanism of Kdo8P synthase: transient kinetic studies and evaluation of a putative reaction intermediate, Biochemistry 37 (1998) 16390–16399. [6] S. Radaev, P. Dastidar, M. Patel, R.W. Woodard, D.L. Gatti, Preliminary X-ray analysis of a new crystal form of the Escherichia coli KDO8P synthase, Acta Crystallogr. D Biol. Crystallogr. 56 (Pt. 4) (2000) 516–519. [7] S. Radaev, P. Dastidar, M. Patel, R.W. Woodard, D.L. Gatt, Structure and mechanism of 3-deoxy-D-manno-octulosonate 8-phosphate synthase, J. Biol. Chem. 275 (2000) 9476–9484. [8] T. Wagner, R.H. Kretsinger, R. Bauerle, W.D. Tolbert, 3-Deoxy-Dmanno-octulosonate-8-phosphate synthase from Escherichia coli. Model of binding of phosphoenolpyruvate and D-arabinose-5phosphate, J. Mol. Biol. 301 (2000) 233–238. [9] J.B. Fenn, M. Mann, C.K. Meng, S.F. Wong, C.M. Whitehouse, Electrospray ionization for mass spectrometry of large biomolecules, Science 246 (1989) 64–71. [10] Z. Li, A. Sau, S. Shen, C. Whitehouse, T. Baasov, K.S. Anderson, A snapshot of enzyme catalysis using electrospray ionization mass spectrometry, J. Am. Chem. Soc. 125 (2003) 9938–9939. [11] J.A. Loo, Studying noncovalent protein complexes by electrospray ionization mass spectrometry, Mass Spectrom. Rev. 16 (1997) 1–23.

Role of phosphoenolpyruvate in KDO8P synthase catalysis / Z. Li et al. / Anal. Biochem. 343 (2005) 35–47 [12] F. Sobott, C.V. Robinson, Protein complexes gain momentum, Curr. Opin. Struct. Biol. 12 (2002) 729–734. [13] L. Konermann, D.J. Douglas, Equilibrium unfolding of proteins monitored by electrospray ionization mass spectrometry: distinguishing two-state from multi-state transitions, Rapid Commun. Mass Spectrom. 12 (1998) 435–442. [14] E.T. van den Bremer, W. Jiskoot, R. James, G.R. Moore, C. Kleanthous, A.J. Heck, C.S. Maier, Probing metal ion binding and conformational properties of the colicin E9 endonuclease by electrospray ionization time-of-Xight mass spectrometry, Protein Sci. 11 (2002) 1738–1752. [15] P.V. Attwood, M.A. Geeves, Kinetics of an enzyme-catalyzed reaction measured by electrospray ionization mass spectrometry using a simple rapid mixing attachment, Anal. Biochem. 334 (2004) 382–389. [16] S.W. Bligh, T. Haley, P.N. Lowe, Measurement of dissociation constants of inhibitors binding to Src SH2 domain protein by non-covalent electrospray ionization mass spectrometry, J. Mol. Recognit. 16 (2003) 139–148. [17] B. Bothner, R. Chavez, J. Wei, C. Strupp, Q. Phung, A. Schneemann, G. Siuzdak, Monitoring enzyme catalysis with mass spectrometry, J. Biol. Chem. 275 (2000) 13455–13459. [18] F. De Angelis, A. Di Tullio, P. Del Boccio, S. Reale, G. Savelli, N. Spreti, ESI-MS in the study of the activity of alpha-chymotrypsin in aqueous surfactant media, Org. Biomol. Chem. 1 (2003) 3125–3130. [19] X. Ge, T.L. Sirich, M.K. Beyer, H. Desaire, J.A. Leary, A strategy for the determination of enzyme kinetics using electrospray ionization with an ion trap mass spectrometer, Anal. Chem. 73 (2001) 5078–5082. [20] X. Huang, A. Aulabaugh, W. Ding, B. Kapoor, L. Alksne, K. Tabei, G. Ellestad, Kinetic mechanism of Staphylococcus aureus sortase SrtA, Biochemistry 42 (2003) 11307–11315. [21] L. Konermann, D.J. Douglas, Pre-steady-state kinetics of enzymatic reactions studied by electrospray mass spectrometry with on-line rapid-mixing techniques, Methods Enzymol. 354 (2002) 50–64. [22] L.L. Lairson, C.P. Chiu, H.D. Ly, S. He, W.W. Wakarchuk, N.C. Strynadka, S.G. Withers, Intermediate trapping on a mutant retaining alpha-galactosyltransferase identiWes an unexpected aspartate residue, J. Biol. Chem. 279 (2004) 28339–28344. [23] A.J. Norris, J.P. Whitelegge, K.F. Faull, T. Toyokuni, Kinetic characterization of enzyme inhibitors using electrospray-ionization mass spectrometry coupled with multiple reaction monitoring, Anal. Chem. 73 (2001) 6024–6029. [24] A.J. Norris, J.P. Whitelegge, K.F. Faull, T. Toyokuni, Analysis of enzyme kinetics using electrospray ionization mass spectrometry and multiple reaction monitoring: fucosyltransferase V, Biochemistry 40 (2001) 3774–3779. [25] D.B. Northrop, F.B. Simpson, Beyond enzyme kinetics: direct determination of mechanisms by stopped-Xow mass spectrometry, Bioorg. Med. Chem. 5 (1997) 641–644. [26] A.A. Paiva, R.F. Tilton Jr., G.P. Crooks, L.Q. Huang, K.S. Anderson, Detection and identiWcation of transient enzyme intermediates using rapid mixing, pulsed-Xow electrospray mass spectrometry, Biochemistry 36 (1997) 15472–15476. [27] N. Pi, J.I. Armstrong, C.R. Bertozzi, J.A. Leary, Kinetic analysis of NodST sulfotransferase using an electrospray ionization mass spectrometry assay, Biochemistry 41 (2002) 13283–13288.

47

[28] N. Pi, J.A. Leary, Determination of enzyme/substrate speciWcity constants using a multiple substrate ESI-MS assay, J. Am. Soc. Mass Spectrom. 15 (2004) 233–243. [29] N. Pi, Y. Yu, J.D. Mougous, J.A. Leary, Observation of a hybrid random ping-pong mechanism of catalysis for NodST: a mass spectrometry approach, Protein Sci. 13 (2004) 903–912. [30] N. Pi, C.L. Meyers, M. Pacholec, C.T. Walsh, J.A. Leary, Mass spectrometric characterization of a three-enzyme tandem reaction for assembly and modiWcation of the novobiocin skeleton, Proc. Natl. Acad. Sci. USA 101 (2004) 10036–10041. [31] K.S. Rao, D. Vander Velde, T.M. Dwyer, S.I. Goodman, F.E. Frerman, Alternate substrates of human glutaryl-CoA dehydrogenase: structure and reactivity of substrates, and identiWcation of a novel 2-enoyl-CoA product, Biochemistry 41 (2002) 1274–1284. [32] W.F. Waas, K.N. Dalby, Transient protein–protein interactions and a random-ordered kinetic mechanism for the phosphorylation of a transcription factor by extracellular-regulated protein kinase 2, J. Biol. Chem. 277 (2002) 12532–12540. [33] S. Wendt, G. McCombie, J. Daniel, A. Kienhofer, D. Hilvert, R. Zenobi, Quantitative evaluation of noncovalent chorismate mutase-inhibitor binding by ESI-MS, J. Am. Soc. Mass Spectrom. 14 (2003) 1470–1476. [34] C.J. Zea, N.L. Pohl, Kinetic and substrate binding analysis of phosphorylase b via electrospray ionization mass spectrometry: a model for chemical proteomics of sugar phosphorylases, Anal. Biochem. 327 (2004) 107–113. [35] D.L. Zechel, L. Konermann, S.G. Withers, D.J. Douglas, Presteady state kinetic analysis of an enzymatic reaction monitored by time-resolved electrospray ionization mass spectrometry, Biochemistry 37 (1998) 7664–7669. [36] C.D. Hubbard, T.S. Shoupe, Mechanisms of acylation of chymotrypsin by phenyl esters of benzoic acid and acetic acid, J. Biol. Chem. 252 (1977) 1633–1638. [37] B.A. Chibber, J.M. Tomich, E.T. Mertz, T. Viswanatha, Kinetic evidence for an intermediate in the deacetylation of monoacetylchymotrypsin, Proc. Natl. Acad. Sci. USA 74 (1977) 510–514. [38] N. Shimamoto, H. Fukutome, Substituent eVects on substrate activation and Michaelis–Menten kinetic parameters in the alphachymotrypsin-catalyzed hydrolysis of phenyl acetates, J. Biochem. (Tokyo) 78 (1975) 663–671. [39] S. Wahlby, Rate of acetylation of alpha-chymotrypsin by p-nitrophenyl [1-14C] acetate studied by isolation of [14C]-acetyl enzyme, Acta Chem. Scand. 24 (1970) 2429–2434. [40] L. Faller, J.M. Sturtevant, The kinetics of the alpha-chymotrypsincatalyzed hydrolysis of p-nitrophenyl acetate in organic solventwater mixtures, J. Biol. Chem. 241 (1966) 4825–4834. [41] F.J. Kezdy, M.L. Bender, The kinetics of the alpha-chymotrypsincatalyzed hydrolysis of p-nitrophenyl acetate, Biochemistry 1 (1962) 1097–1106. [42] M.A. Marini, G.P. Hess, Reactivity and interrelationship of intermediates in the hydrolysis of p-nitrophenyl acetate catalysed by chymotrypsin, Nature 184 (1959) 113–114. [43] H. Gutfreund, J.M. Sturtevant, The mechanism of the reaction of chymotrypsin with p-nitrophenyl acetate, Biochem. J. 63 (1956) 656–661. [44] K.A. Johnson, Transient-state kinetic analysis of enzyme reaction pathways, Enzymes 20 (1992) 1–61.