Catalytic Mechanism of the Tryptophan Synthase α2β2 Complex

The mechanism of the tryptophan synthase α2β2 complex from Salmonella typhimurium is explored by determining the effects of pH, of temperature, and of isotopic substitution on the pyridoxal phosphate-dependent reaction of l-serine with indole to form l-tryptophan. The pH dependence of the kinetic parameters indicates that three ionizing groups are involved in substrate binding and catalysis with pK a 1 = 6.5, pK a 2 = 7.3, and pK a 3 = 8.2–9. A significant primary isotope effect (∼3.5) on V and V/K is observed at low pH (pH 7), but not at high pH (pH 9), indicating that the base that accepts the α-proton (βLys-87) is protonated at low pH, slowing the abstraction of the α-proton and making this step at least partially rate-limiting. pK a 2 is assigned to βLys-87 on the basis of the kinetic isotope effect results and of the observation that the competitive inhibitors glycine and oxindolyl-l-alanine display single pK i values of 7.3. The residue with this pK a (βLys-87) must be unprotonated for binding glycine or oxindolyl-l-alanine, and, by inference,l-serine. Investigations of the temperature dependence of the pK a values support the assignment of pK a 2 to βLys-87 and suggest that the ionizing residue with pK a 1 could be a carboxylate, possibly βAsp-305, and that the residue associated with a conformational change at pK a 3 may be βLys-167. The occurrence of a closed to open conformational conversion at high pH is supported by investigations of the effects of pH on reaction specificity and on the equilibrium distribution of enzyme-substrate intermediates.

The mechanism of the tryptophan synthase ␣ 2 ␤ 2 complex from Salmonella typhimurium is explored by determining the effects of pH, of temperature, and of isotopic substitution on the pyridoxal phosphate-dependent reaction of L-serine with indole to form L-tryptophan. The pH dependence of the kinetic parameters indicates that three ionizing groups are involved in substrate binding and catalysis with pK a 1 ‫؍‬ 6.5, pK a 2 ‫؍‬ 7.3, and pK a 3 ‫؍‬ 8.2-9. A significant primary isotope effect (ϳ3.5) on V and V/K is observed at low pH (pH 7), but not at high pH (pH 9), indicating that the base that accepts the ␣-proton (␤Lys-87) is protonated at low pH, slowing the abstraction of the ␣-proton and making this step at least partially rate-limiting. pK a 2 is assigned to ␤Lys-87 on the basis of the kinetic isotope effect results and of the observation that the competitive inhibitors glycine and oxindolyl-L-alanine display single pK i values of 7.3. The residue with this pK a (␤Lys-87) must be unprotonated for binding glycine or oxindolyl-L-alanine, and, by inference, L-serine. Investigations of the temperature dependence of the pK a values support the assignment of pK a 2 to ␤Lys-87 and suggest that the ionizing residue with pK a 1 could be a carboxylate, possibly ␤Asp-305, and that the residue associated with a conformational change at pK a 3 may be ␤Lys-167. The occurrence of a closed to open conformational conversion at high pH is supported by investigations of the effects of pH on reaction specificity and on the equilibrium distribution of enzyme-substrate intermediates.
The tryptophan synthase ␣ 2 ␤ 2 complex (EC 4.1.2.20) is a useful system for investigating relationships between protein structure and function (for reviews, see Refs. [1][2][3][4]. The separate tryptophan synthase ␤ 2 subunit 1 and the ␤ subunit in the ␣ 2 ␤ 2 complex catalyze the pyridoxal phosphate (PLP) 2 -depend-ent conversion of L-serine and indole to L-tryptophan. A number of intermediates in this reaction have been identified from spectroscopic and kinetic studies (see Scheme I in Discussion).
The three-dimensional structure of the tryptophan synthase ␣ 2 ␤ 2 complex from Salmonella typhimurium (5) revealed that the active sites of the ␣ and ␤ subunits are ϳ25 Å apart and are connected by a hydrophobic tunnel. This tunnel serves as a passageway for indole from the active site of the ␣ subunit, where it is produced by cleavage of indole-3-glycerol phosphate, to the active site of the ␤ subunit, where it reacts with L-serine to form L-tryptophan. The activity at each site is modulated by reciprocal communication between the ␣ and ␤ sites. These heterotrophic interactions are proposed to switch the ␣ and ␤ subunits between "open" (catalytically inactive) and "closed" (catalytically active) conformations, to coordinate the activities at the two sites, and to prevent the escape of indole (4). Recent crystallographic results provide direct evidence for ligand-induced conformational changes that result in physical closure of loop structures in the ␣ subunit and of a "mobile region" (6) or "communication domain" (7) in the ␤ subunit. These results support the proposal that the abstracted ␣-proton of L-serine is sequestered in a solvent-excluded site (8).
Investigations of the pH dependence of enzyme reactions can give information on enzyme mechanism and on the identity of residues that play catalytic and regulatory roles (9 -11). Here we report the effects of pH, of temperature, of isotopic substitution, and an allosteric ligand (DL-␣-glycerol 3-phosphate) on the PLP-dependent reaction of the tryptophan synthase ␣ 2 ␤ 2 complex with L-serine and indole to form L-tryptophan. The results indicate that three ionizing groups are involved in substrate binding and catalysis with pK a 1 ϭ 6.5, pK a 2 ϭ 7.3, and pK a 3 ϭ 8.2-9. Correlation of the kinetic results with structural and mutational data suggests that the ionizing residues are ␤Asp-305, ␤Lys-87, and ␤Lys-167.

EXPERIMENTAL PROCEDURES
Chemicals and Buffer-PLP, pyridoxal hydrochloride, L-serine, and DL-␣-glycerol 3-phosphate were purchased from Sigma. Oxindolyl-Lalanine, which was a generous gift of Robert S. Phillips, was synthesized by the method of Savige and Fontana (12) as described (13). MOPS and proline were from Fluka. Bicine was from United States Biochemical. DL-[␣-2 H]serine was prepared as reported (14). MBP buffer (15) containing 50 mM MOPS, 50 mM Bicine, 50 mM proline, and 1 mM EDTA was used for kinetic and spectroscopic studies with additional 100 mM NaCl. The pH was raised with NaOH to pH 11.2; the solution was then back-titrated with HCl to the desired pH value.
Bacterial Strain, Plasmid, Enzyme Preparation, and Enzyme Assay-The plasmid pEBA-10 (16) was used to express the wild type tryptophan synthase ␣ 2 ␤ 2 complex from S. typhimurium in E. coli CB 149 (17), which lacks the trp operon. The ␣ 2 ␤ 2 complex was purified by a method using crystallization (18). Protein concentrations were deter-* A preliminary report of portions of this work was presented at the American Society for Biochemistry and Molecular Biology, May 16 -20, 1999, in San Francisco, CA. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.  1 The term ␤ 2 subunit is used for the isolated enzyme in solution; ␤ subunit is used for the enzyme in the ␣ 2 ␤ 2 complex or to describe a specific residue in the ␤ subunit. 2 The abbreviations used are: PLP, pyridoxal phosphate; Bicine, N,Nbis(2-hydroxyethyl)glycine; MOPS, 3-(N-morpholino)propanesulfonic acid; MBP, MOPS, Bicine, proline; ⌬H ion , ⌬H ionization ; V, apparent V max ; . See data analysis under "Experimental Procedures" for information relevant to apparent V max and apparent K Ser values. mined from the specific absorbance at 278 nm using A ϭ 6.0 for the ␣ 2 ␤ 2 complex (19). Activity of tryptophan synthase in the ␤-replacement reaction with L-serine and indole (␤ reaction) was determined by a spectrophotometric assay (19) in the presence of 100 mM NaCl and L-[␣-1 H]serine or DL-[␣-2 H]serine. Because the D-isomer of Ser is unreactive with tryptophan synthase (8,14,20,21), DL-Ser can be used in place of L-Ser for investigations of the catalytic mechanism. DL-␣-Glycerol 3-phosphate (50 mM) was added where indicated. One unit of activity in any reaction is the formation of 0.1 mol of product in 20 min at 37°C. We demonstrated that the tryptophan synthase ␣ 2 ␤ 2 complex is stable in the pH range of 6.5-9.6 by preincubating the enzyme at pH values over this range and then determining activity at pH 7.8; the activity was not changed by the preincubation at any pH in this range (data not shown).
Data Analysis-Initial reaction rates in the reaction of L-serine and indole to form L-tryptophan were obtained as a function of L-serine concentration and of pH in the presence of 200 M indole. The data were analyzed by published methods (10,11,22,23). The kinetic parameters (V, K) 2 were determined by fitting the initial rate data to Equation 1.
The values of V and K are apparent values because indole was not saturating. It is not possible to use saturating indole concentrations because the ␣ 2 ␤ 2 complex is inhibited by concentrations of indole greater than 30 M (24). Because the ␤ 2 subunit alone is not inhibited by indole, Heilmann (24) concluded that indole inhibits the ␣ 2 ␤ 2 complex by binding to 1 or 2 indole sites in the ␣ subunit. The concentration of indole in our assays, 200 M, is 4 -27 times higher than the K m values that have been reported at pH ϳ7.8 for the ␣ 2 ␤ 2 complex from E. coli (13 M (25), 50 M (24), and 7.4 M (26)) and for the ␣ 2 ␤ 2 complex from S. typhimurium (18 M (27)). Lane and Kirschner have reported that the K m for indole decreased from 42 M at pH 6.1 to 8.0 at pH 8.0. Thus, the concentration of indole in our assays is higher than the K m for indole even at low pH. Data for competitive inhibition was fitted to Equation 2, where K is the Michaelis constant, K i is the inhibition constant of a competitive inhibitor, and [I] is the concentration of the competitive inhibitor. pK i profiles for glycine and oxindolylalanine were fitted to Equation 3 (see Equation 57 in Ref. 10) because pH profiles are level above single pK but decrease with a ϩ1 unit slope below the pK.
When the logV profile was bell-shaped with a slope of ϩ1 at the acidic side and Ϫ1 at the basic side, the data were fitted to Equation 4 (see Equation 27 in Ref. 22,and Equation 11 in Ref. 23).
When the logV/K profile exhibited a slope of ϩ2 at the acidic side and Ϫ1 at the basic side, the data were fitted to Equation 5 (see Equation 12 in Ref. 23).
When pK a 1 and pK a 2 values were too close to give separate pK a values, the data were fitted to Equation 5 defining pK a 1 ϭ pK a 2. In Equations 3-5, pK i *, logV*, and log(V/K)* are the pH-independent values of the parameters, and K 1 , K 2 , and K 3 are dissociation constants for enzyme groups. The enthalpy of ionization (⌬H ion ) was determined by fitting the pK values obtained at different temperatures to Equation 6, where ⌬H ion , T, and R are the enthalpy of ionization, the absolute temperature, and the gas constant (1.987 cal/mol/deg), respectively.

RESULTS
pH Dependence of the Kinetic Parameters-The effect of pH on the conversion of L-serine and indole to L-tryptophan by the tryptophan synthase ␣ 2 ␤ 2 complex was determined over the pH range of 6.0 -9.5 while varying L-serine concentrations at a fixed level of indole (0.2 mM). This concentration of indole gives the highest rate at each pH value and is greater than the K m for indole (see data analysis under "Experimental Procedures"). Higher concentrations of indole are inhibitory (24). As seen in Fig. 1A, the data for log V fit well using Equation 4 to a bell-shaped curve with slopes of ϩ1 and Ϫ1 on the acidic and basic sides, respectively, indicating dependence on two ionizing groups with pK a values of ϳ6.5 and ϳ8.2 (line 1 in Table I). Data for the pH dependence of log V/K (Fig. 1A) fit best using Equation 5 to a bell-shaped curve with slopes of ϩ2 and Ϫ1 on the acidic and basic sides, respectively, indicating dependence on three ionizing groups with pK a 1,2 ϭ ϳ7.0 and pK a 3 ϭ ϳ9.0 (line 2 in Table I). A value of pK a 2 ϭ ϳ7.3 (line 3 in Table I) could be calculated from the fit of the same data to Equation 5 assuming a value of pK a 1 ϭ 6.5 from the temperature dependence of pK a 1 (see below).
Effects of an Allosteric Ligand on the pH Dependence of the Kinetic Parameters-The substrate analog DL-␣-glycerol 3-phosphate binds to the active site of the ␣ subunit and stabilizes the closed conformation of the ␣ 2 ␤ 2 complex (15). The pH profile for log V in the presence of DL-␣-glycerol 3-phosphate (Fig. 1B) decreases below an acidic pK a ϭ 6.8 and above a basic pK a ϭ 8.8 (Table I). The acidic side of the log V/K profile exhibits a slope of ϩ1 and yields a pK a ϭ 7.3; the basic side yields a pK a of 8.7 (Table I).
Effect of Temperature on pK a Values-The temperature dependence of log V and log V/K was determined to investigate the nature of groups whose ionization state affect the activity of tryptophan synthase. Activities were measured at 15, 25, and 37°C, and the data were fit to Equations 4 or 5 to give pH profiles of log V ( Fig. 2A) and log V/K (Fig. 2B). The derived pK a values are listed in Table II. The values of pK a 1 and pK a 2 in the log V/K profile were sufficiently separated to be calculated at 15 and 25°C but not at 37°C. Because the values of pK a 1 appeared to be temperature-independent and equal to ϳ6.5 at all three temperatures in the log V profile and at 15 and 25°C in the log V/K profile, the value of pK a 1 for log V/K at 37°C was assumed to be 6.5 as well. This value was used in Equation 5 to calculate pK a 2 (Table II and Fig. 2). From plots of the values of pK a versus 1/T (Fig. 2, C and D), lines were obtained and analyzed by using Equation 6 to yield values for ⌬H ion , the enthalpy of ionization (Table II).
pH Dependence of Inhibition by Substrate Analogs-To determine whether pK a values observed in the log V/K profile were true or apparent values, we measured the effect of pH on inhibition by substrate analogs. Values of pK i were determined from plots of competitive inhibition by glycine (Fig. 3A) and by oxindolyl-L-alanine (data not shown) versus L-serine in the ␤-replacement reaction with L-serine and indole over a range of pH values from 6 -9.5. The pH profile for pK i for each inhibitor was a half-bell with a limiting slope of ϩ1 on the acidic side (Fig. 3B). The data were fit to Equation 3 to yield pK a values of ϳ7.3 for glycine and for oxindolyl-L-alanine; the pH-independent values of K i are ϳ0.1 M for glycine and 0.05 mM for oxindolyl-L-alanine (Table I).
pH Dependence of Deuterium Isotope Effects-The primary deuterium isotope effects for abstraction of the ␣-proton of L-serine were measured as a function of pH (Fig. 4). Both D V and D (V/K) 2 are pH-dependent and become negligible (i.e. ϳ1) above pH 9. Both of these effects increase and become equal below pH 7, D V ϭ 3.5 and D (V/K) ϭ 3.5.

DISCUSSION
The reaction of L-serine and indole to form L-tryptophan proceeds through a series of PLP intermediates at the active site of the ␤ subunit in the tryptophan synthase ␣ 2 ␤ 2 complex (Scheme I) (13, 20, 28 -30). Similar mechanisms are utilized by other PLP enzymes that catalyze ␤-elimination and ␤-replacement reactions (28,31), including O-acetylserine sulfhydrylase (32,33), tryptophan indole lyase (34), and tyrosine phenol lyase (35). The reactions in Scheme I include the release and transfer of three different protons. The reaction of the protonated internal aldimine (E) with the protonated substrate, L-serine, to form ES-I yields an extra proton, shown in a triangle, which may be used subsequently to protonate the leaving hydroxyl group of L-serine in the conversion of ES-III to ES-IV. The ␣-proton of L-serine, shown in a square, is abstracted by ␤Lys-87 in the conversion of ES-II to ES-III and is subse-TABLE I pK a values associated with tryptophan synthase Initial rates in the reaction of L-serine ϩ indole 3 L-tryptophan ϩ H 2 O, varying L-serine concentration at 37°C in the presence of 0.2 mM indole ( Figs. 1 and 3) and in the presence or absence of 50 mM DL-␣-glycerol 3-phosphate (GP). The pH profiles of ϪlogK i for oxindolyl-L-alanine (Oxa) or glycine (Gly) were derived from plots of competitive inhibition versus L-serine (Fig. 3).

TABLE II pK a values at various temperatures and derived ⌬H ion values
Initial rates in the reaction of L-serine ϩ indole 3 L-tryptophan ϩ H 2 O, varying L-serine concentration at 15, 25, and 37°C for determination of ⌬H ion (Fig. 2 quently used to protonate the quinonoid ES-VI to form ES-VII. The C-3 proton of indole, shown in a circle, is removed during the tautomerization of the indolenine intermediate (ES-V to ES-VI) and may be used to protonate the product, L-tryptophan, as it is released. Further elucidation of the mechanism of the reaction in Scheme I would be facilitated by identifying residues in tryptophan synthase that catalyze the proton transfer steps and by identifying the pK a values of these residues. Investigations of the pH dependence and temperature dependence of kinetic parameters have frequently been used to deduce the intrinsic pK a values of functional groups in enzymes. A systematic in-vestigation of these factors is an important step in determining an enzyme mechanism. Although the identification of specific residues and their pK a values may sometimes be problematic (9), the kinetic results combined with information from structural investigations and site-directed mutagenesis can yield important support for a proposed mechanism.
The pH profiles for the steady-state kinetic parameters of tryptophan synthase show that three ionizing groups are important for substrate binding and catalysis (Fig. 1A and Table  I). The group with pK a 2 ϭ ϳ7.3 appears in the log V/K profile but not in the log V profile, indicating that this group must be in the correct protonation state for the substrate to bind. The pH profiles of pK i for glycine and for oxindolyl-L-alanine, which are competitive inhibitors of L-serine, show that a single residue with a pK a value of 7.3 must be unprotonated for binding glycine or oxindolyl-L-alanine and, by inference, L-serine ( Fig.  3B and Table I). The coincidence of this pK a value with pK a 2 in the log V/K profile shows that pK a 2 is the correct pK a value for the ionizing group and rules out drastic stickiness of the substrate (10). A sticky substrate, which reacts to give products more rapidly than it dissociates from the enzyme, gives incorrect pK a values that differ from the pK a values obtained for inhibitors. Because L-serine, glycine, and oxindolyl-L-alanine have no ionizable groups in the range of pH measured, the observed pK a values reflect enzyme groups that must be deprotonated or protonated for maximum binding and catalysis.
Oxindolyl-L-alanine, which has tetrahedral geometry at C-3 of the indole ring, is a structural analog of the indolenine tautomer of L-tryptophan, which is a proposed intermediate in reactions of tryptophan synthase (E-V in Scheme I) and tryptophan indole-lyase (13,29). The finding that oxindolyl-L-alanine is a potent competitive inhibitor of both enzymes provided strong evidence for the intermediacy of the indolenine tautomer of L-tryptophan in reactions catalyzed (13,29). The K i value (0.05 mM) found for oxindolyl-L-alanine at pH 7.8 in the present work (Fig. 3) is about 10-fold higher than that reported previously (13). However, the previous K i value was obtained from assays of the reaction of L-serine with indole-3-glycerol phosphate, a substrate that binds to the ␣ subunit and is known to increase the affinity of ligands that bind to the ␤ subunit (17).
The pH profiles of pK i for oxindolyl-L-alanine and for glycine each show a requirement for a single group with a pK a of ϳ7.3 to be unprotonated (Fig. 3). In contrast, analogous studies with tryptophan indole-lyase showed a requirement for two groups to be unprotonated for the binding of oxindolyl-L-alanine with pK a values of 6.0 and 7.6 and for one group to be unprotonated for the binding of alanine with a pK a value of 7.6 (34). The results for tryptophan indole-lyase suggested that the group with the pK a of 6.0 abstracted the ring nitrogen proton, facilitating the formation and stabilization of the indolenine tautomer of tryptophan. The absence of a second pK a for oxindolyl-L-alanine binding to tryptophan synthase indicates that the group with pK a 1 is not involved in binding this analog.
Addition of an ␣ subunit ligand, DL-␣-glycerol 3-phosphate, increases the values of pK a 1 and pK a 3 in the log V profile (Fig.  1B and Table I). Stabilization of the closed conformation of ␣ 2 ␤ 2 complex by the ␣ subunit ligand may increase these pK a values by increasing the hydrophobicity of the environment of these two groups. The reason for the absence of a third ionizing group corresponding to pK a 1 in the V/K profile is unknown.
Our most important results are the pH dependence of the primary deuterium isotope effects of abstraction of the ␣-proton of L-serine ( D V and D (V/K)) 2 (Fig. 4). Both of these effects increase with decreasing pH and become essentially equal (ϳ3.5) below pH 6.8. The significant primary kinetic isotope effect at low pH indicates that the general base has been

FIG. 4. pH dependence of the primary deuterium kinetic isotope effects on V ( D V)(Ⅺ) and D (V/K) (q).
protonated, slowing the abstraction of the ␣-proton and making this step at least partially rate-limiting. This permits the highly probable assignment of one of the pK a s to the base that accepts the ␣-proton.
An alternative interpretation of the kinetic isotope effect results is that another step becomes rate-limiting at higher pH. Knowles (9) has discussed this problem and quoted Jencks (51) as pointing out that a "mirage" pK a may be seen if the elementary step affected by the ionization is not rate-determining at all pH values. This type of pK a has also been termed a "kinetic pK a " (36). This alternative interpretation is supported by our conclusion that the rate of the reaction of L-serine with indole decreases at high pH as the result of a conformational change that produces an open form in which removal of the hydroxyl group of L-serine becomes rate-limiting (see below). In contrast, the rate of reaction of ␤-chloro-L-alanine with indole increases almost linearly between pH 7 and 9 (data not shown), consistent with evidence that ␤-elimination of the chloride-leaving group of ␤-chloro-L-alanine is not rate-limiting in the open conformation (37).
Assignment of pK a 2-Several types of evidence support the assignment of pK a 2 ϭ ϳ7.3 to ␤Lys-87. First, the primary kinetic isotope effects for D V and D V/K decrease above pH 7 (Fig. 4). Second, this pK a value is observed for inhibition by glycine and oxindolyl-L-alanine and not in the logV plot (Figs. 1 and 3 and Table I), indicating that this ionizing group must be unprotonated for binding L-serine, glycine, or oxindolyl-L-alanine. These amino acids, which are largely protonated in the neutral pH range, would be expected to bind to the unprotonated ␤Lys-87. Third, the thermal dependence of pK a 2 is characteristic of a lysyl residue (Table II). Fourth, mutation of ␤His-86 (H86L) alters the pH profiles of absorbance and fluorescence signals and shifts the pH optimum for the synthesis of L-tryptophan from pH 7.5 to pH 8.8. 3 These new results provide evidence that His-86 facilitates catalysis in the physiological pH range by decreasing the pK a of the catalytic ␤Lys-87 in the wild type ␣ 2 ␤ 2 complex. 3 They also provide a structural explanation for the low pK a value assigned here to ␤Lys-87. 3 Fifth, very similar pH profiles for D V and D (V/K) were observed for tyrosine phenol-lyase from Erwinia herbicola (35) and led to the assignment of a pK a ϭ ϳ7.6 -7.8 to the catalytic lysyl residue of tyrosine phenol-lyase that abstracts the ␣-proton of L-tyrosine.
Previous investigations of tryptophan synthase have demonstrated a primary kinetic isotope effect for the L-serine deaminase reaction of the ␤ 2 subunit at pH 7.8 (14), for the synthesis of L-tryptophan (20,41,42), and for the equilibrium distribution of intermediates in the reaction of L-serine (8) and in the reaction of L-serine and indole (26).
Assignment of pK a 1-Our finding that pK a 1 ϭ ϳ6.5 is essentially temperature-independent ( Fig. 2 and Table II) suggests that this residue is a neutral acid, e.g. a carboxylate (10). Although the validity of the use of ⌬H ion values for deducing the nature of ionizing groups has been criticized (9), the assignment of a pK a ϭ 6.5 to a carboxylate in a hydrophobic site seems not unreasonable. Although ␤Asp-305 and ␤Glu-109 are both located in the active site of the ␤ subunit (6,7,38), ␤Glu-109 is close to the indole nitrogen of L-tryptophan in the ES-VIII intermediate (Scheme I) (6) and appears too distant to accept a proton. The crystal structure of the external aldimine of L-serine (ES-I in Scheme I) formed by a mutant enzyme (␤K87T) suggests that the interaction of the carboxylate of ␤Asp-305 with the hydroxyl of L-serine is important to the chemistry of ␤-elimination to yield ES-II (6). Consequently, ␤Asp-305 is the most likely candidate for the residue with pK a 1 ϭ ϳ6.5 that accepts the initial proton, indicated by the triangle, and protonates the hydroxyl-leaving group of L-serine.
Assignment of pK a 3-The group with pK a 3 ϭ 8.2 in the log V profile and ϭ 9 in the log V/K profile ( Fig. 1 and Table I) must be protonated for substrate binding and catalysis because the pH dependence of log V/K follows ionizations in the free enzyme and the free substrate (36). V/K contains all steps up to and including the first irreversible step, which is addition of indole to ES-IV. The group with pK a 3 ϭ 8 -9 exhibits a ⌬H ion of 17-20 kcal/mol (Table II), suggesting that this ionization is that of a protein residue which is linked to a conformational change (10) and consequently is not that of the amino group of the substrate L-serine (pK a ϭ 9). The observed pK a 3 ϭ 8 -9 is probably not the true pK a of this ionizing residue because the observed change in free energy is the sum of the free energy of ionization and the free energy of the conformational change.
We propose that this conformational change is the conversion of the closed form of the tryptophan synthase ␣ 2 ␤ 2 complex to the open form (for a review, see Ref. 4). This proposal is supported by our finding that log V/K shows no pK a 3 in the presence of Cs ϩ , a cation that stabilizes the closed conformation (43) (data not shown). We have recently observed that activity in the ␤-elimination reaction (L-serine 3 pyruvate and NH 3 ) increases nearly linearly between pH 6.7 and 9. 3 This result suggests that the open conformation of the ␣ 2 ␤ 2 complex is formed at high pH and that the aminoacrylate intermediate (ES-IV in Scheme I) is more accessible to hydrolysis by solvent water in the open form. 3 Increased ␤-elimination activity is also exhibited by the open form of the enzyme that is promoted by mutation, solvents, or low concentrations of urea or guanidine hydrochloride (37,(43)(44)(45)(46)(47)(48). A previous investigation has also provided evidence for a pH-dependent conversion of the ␣ 2 ␤ 2 complex from a closed conformation at low pH to an open conformation at high pH (15).
In conclusion, our studies of the temperature and pH dependence of kinetic parameters and of the effects of isotopic substitution provide evidence that three ionizing groups are involved in substrate binding and catalysis by the tryptophan synthase ␣ 2 ␤ 2 complex. The residue with pK a ϭ 6.5 is likely a carboxylate that accepts a proton from the ␣-amino group of L-serine and then uses the proton to promote the dehydration of L-serine. ␤Asp-305 is a good candidate to serve this role. The residue with pK a ϭ 7.3 must be unprotonated for binding glycine or oxindolyl-L-alanine, and, by inference, L-serine. The pH dependence of the primary isotope effects suggests that this residue is ␤Lys-87 which accepts the ␣-proton of L-serine and then uses the proton to protonate the quinonoid of L-tryptophan. Ionization of a residue linked to a conformational change (pK a 3 ϭ ϳ9) destabilizes the active, closed form of the enzyme. ␤Lys-167 is a good candidate to play this role.