Effects of hydrostatic pressure on the conformational equilibrium of tryptophan synthase from Salmonella typhimurium

Abstract

A wide range of parameters influence allosteric communications between the α- and β-subunits of the Trp synthase α₂β₂ multienzyme complex with L-Ser, including monovalent cations, pH, temperature, ligands, organic solvents, and hydrostatic pressure. The conformational change from closed to open can be monitored either by absorbance at 423 nm or fluorescence at 495 nm from the pyridoxal-5′-phosphate-L-Ser complex. Pressure perturbation was used to quantify the effects of monovalent cations, ligands, and mutations on the conformational equilibrium of Trp synthase. P-jump kinetics in the presence of Na⁺, NH₄⁺, and Na⁺ together with benzimidazole were also examined. The plots of lnk versus P are nonlinear and require a compressibility (β^‡_o) term to obtain a good fit. β^‡_o is positive for the Na⁺ enzyme but negative for NH₄⁺ and Na⁺ with benzimidazole. These results suggest that there is a large contribution of solvation to the kinetics of the conformational change of Trp synthase. The relaxation kinetics are also different if the P-jumps are made by increasing or decreasing pressure, suggesting that the enzyme conformations are ensembles of microstates.

Tryptophan (Trp) synthase is an α₂β₂ multienzyme complex that catalyzes the synthesis of L-Trp from indole-3-glycerol phosphate (IGP) and L-Ser (Eqs. 1–3) with strict allosteric control of the overall reaction. The α-subunit catalyzes the retro-aldol cleavage of IGP to indole and D-glyceraldehye-3-phosphate (G3P), while the β-subunit catalyzes the condensation of indole with L-Ser to give L-Trp.¹ Binding of IGP to the α-active site results in activation of the β-site for L-Ser binding, which then results in a closed complex, and water elimination from the external aldimine of L-Ser (Scheme 1, E_EA–Ser) to form an aminoacrylate Schiff’s base (Scheme 1, E_AA), which then transmits an allosteric signal to the α-site to activate IGP cleavage.^2–4 Indole then diffuses intramolecularly through a 25–30 Å tunnel between the α- and β-sites to react with the aminoacrylate and form L-Trp.^2,5,6 We have used pressure to study the thermodynamics and kinetics of the conformations of the Trp synthase-L-Ser complex.^7–9

Scheme 1.

The external aldimine complex of Trp synthase with L-Serine (Scheme 1, E_EA-Ser) exhibits a visible absorption band at 423 nm, while the aminoacrylate complex (Scheme 1, E_AA) exhibits absorption bands at about 350 and 460 nm.^10,11 In addition, E_EA-Ser is highly fluorescent,¹² with emission at 495 nm, while E_AA is only weakly fluorescent.¹³ E_EA-Ser is found in the open conformation, while E_AA is in the closed conformation. Thus, the spectroscopic properties of Trp synthase are ideal to study the thermodynamics and kinetics of the conformational equilibrium. A wide range of physical parameters have been found to influence allosteric communications between the α- and β-subunits of the Trp synthase complex with L-Ser, including monovalent cations,^14–19 pH,¹⁰ temperature,^10,14 α- and β-site ligands,^{6,10,14,20–23} and organic solvents.^24,25

(1)

(2)

(3)

Using hydrostatic pressure perturbation, we have shown that it was possible to quantify these effects by determination of the internal equilibrium constant, K_eq, in the presence of ligands, even if the position at atmospheric pressure was as much as 10⁴ toward one side.^7,8 The spectra of Trp synthase-L-Ser complex in the presence of NH₄⁺ as the monovalent cation are shown in Figure 1. As pressure is increased from 50 to 200 MPa, the peaks at 350 and 460 nm decrease in intensity while the 423 nm peak increases, with good isosbestic points at 380 and 465 nm. Fitting these data to a simple two-state equilibrium model provides the K_eq of 2.3 × 10⁻⁴, with a ΔV_o of −171 mL/mol (Table 1). Figure 2 shows the results of the same experiment performed by fluorescence measurement. As pressure is increased, the fluorescence intensity at 495 nm increases dramatically (Fig. 2A). The fluorescence change with pressure (Fig. 2B) can be fit with a Boltzmann function (Equation 4) to give the values of K_eq of 3.6 × 10⁻⁴ and ΔV_o of −167 mL/mol, in good agreement with the absorbance results. The absorbance and

$${\mathrm{I}}_{\mathrm{p}}=({\mathrm{I}}_{\infty }-{\mathrm{I}}_0)\ast ({\mathrm{K}}_{\mathrm{eq}}\ast \exp (-\mathrm{P}\mathrm{\Delta }{\mathrm{V}}_0/\mathrm{RT}\left)\right)/(1+{\mathrm{K}}_{\mathrm{eq}}\ast \exp (-\mathrm{P}\mathrm{\Delta }{\mathrm{V}}_0/\mathrm{RT}\left)\right),$$ (4)

fluorescence changes shown in Figures 1 and 2 are completely reversible upon decompression. The values of K_eq and ΔV_o vary, depending on the cation and ligand, as can be seen from Table 1. Na⁺ and K⁺ exhibit the largest values of K_eq, about 0.1, while the other monovalent cations show values around 10⁻⁴. α-Site ligands, such as glycerol phosphate (GP) and indoleacetylglycine (IAG), shift K_eq by 1–2 orders of magnitude in the direction of E_AA, while the β-ligand, benzimidazole, shifts K_eq about 10⁴ toward E_AA. Lower pH values also shift K_eq modestly toward E_AA. The reaction volume change, ΔV_o, also shows a wide variation from −125 to −200 mL/mol, depending on the cation and ligand (Table 1). Mutations that affect the allosteric equilibrium (βE109D²⁶ and βD305A²⁷) also have dramatic effects on ΔV_o (Table 1). The major contribution to the reaction volume is probably changes in solvation associated with the conformational changes. Assuming an average density difference of 20% between protein-bound and free water, the change in solvation ranges from about 30 to more than 50 waters (Table 1). These solvation differences reflect structural differences in the enzyme in the presence of the different cations and ligands.

Figure 1.

Effect of pressure on the UV-visible absorption spectrum of Trp synthase-L-Ser complex in the presence of NH₄⁺. Arrows indicate the direction of change in absorbance with pressure application. The spectra shown were collected between 50 and 200 MPa.

Table 1.

Thermodynamic parameters for the Trp synthase conformational equilibrium⁸

pH	Amino acid	Cation, ligand	Keq	ΔΔG (kJ/mol)	ΔVo (mL/mol)	Δ(H2O) mol/mol
8.0	L-Ser	None	(2.0 ± 0.1) × 10−4	0	−172 ± 5	48
8.0	L-Ser	0.1 M K+	(4.4 ± 0.3) × 10−2	−13.3	−124 ± 8	34
8.0	L-Ser	0.1 M Na+	(1.2 ± 0.01) × 10−1	−15.8	−126 ± 2	35
8.0	L-Ser	0.1 M Na+, 50 mM GP	(2.7 ± 0.1) × 10−3	−6.4	−153 ± 3	43
8.0	L-Ser	0.1 M Na+, 2.5 mM IAG	(4.2 ± 0.3) × 10−2	−12.8	−115 ± 8	32
6.6	L-Ser	0.1 M Na+	(5.2 ± 0.2) × 10−2	−13.7	−109 ± 3	30
8.0	L-Ser	0.1 M Na+, 5 mM BZI	(1.3 ± 0.1) × 10−5	6.8	−204 ± 2	57
8.0	L-Ser	0.1 M Na+, 5 mM BZI + 50 mM GP	(1.1 ± 0.1) × 10−5	7.2	−153 ± 1	43
8.0	L-Ser	0.1 M Li+	(7.6 ± 0.2) × 10−4	−3.3	−163 ± 2	45
8.0	L-Ser	0.1 M Rb+	(4.3 ± 0.2) × 10−4	−1.8	−165 ± 2	46
8.0	L-Ser	0.1 M NH4+	(2.3 ± 0.1) × 10−4	−0.3	−171 ± 2	48
8.0	L-Ser	0.1 M Cs+	(1.6 ± 0.1) × 10−5	6.3	−187 ± 3	52
8.0	L-Trp	0.1 M Na+, 50 mM GP	(7.3 ± 0.2) × 10−3	−	−97 ± 3	27
6.6	L-Ser	βD305A, 0.1 M Na+, 5 mM BZI	0.64 ± 0.10	−20.0	−29 ± 3	8
6.6	L-Ser	βD305A, 0.1 M NH4+, 5 mM BZI	0.12 ± 0.01	−15.8	−52 ± 3	14
8.0	L-Ser	βE109D, 0.1 M Na+, 50 mM GP	0.12 ± 0.01	−15.8	−87 ± 4	24
8.0	L-Ser	βE109D, 0.1 M Na+, 5 mM BZI	(2.9 ± 0.5) × 10−2	−12.3	−162 ± 5	45
8.0	L-Ser	βE109D, 0.1 M NH4+	(7.4 ± 1) × 10−2	−14.6	−115 ± 3	32

Figure 2.

(A) Effect of pressure on the fluorescence emission spectrum of Trp synthase-L-Ser complex in the presence of NH₄⁺ with 420 nm excitation. Pressures for each spectrum are: 1, 8 MPa; 2, 40 MPa; 3, 61 MPa; 4, 97.3 MPa; 5, 111 MPa; 6, 120 MPa; 7, 132 MPA; 8, 144 MPa; 9, 195 MPa. (B) Fit of the fluorescence intensity at 495 nm to Equation 4.

Three-dimensional structures of Trp synthase in complexes with monovalent cations and α- and β-ligands have been determined.^22,28,29 There are significant structural differences between the Cs⁺ and Na⁺ forms of the βK87T mutant enzyme,²⁸ since two Cs⁺ ions are bound to the β-subunit, while only one Na⁺ binds. The structure of the Trp synthase βK87T mutant complex with L-Trp and Na⁺ is shown in Figure 3. In this closed conformation, βK167 forms an ion pair with αD56, seen in the space-filling structures at the interface of the sub-units in Figure 2. When the conformation changes to the open state, the ion pair must be broken and solvent exposed, resulting in binding of 8–10 waters of solvation and electrostriction.³⁰ This accounts for only about one-third or less of the total volume change, so the remainder of the volume change may be due to solvation of the exposed hydrophobic surface in the open conformation.

Figure 3.

Structure of Trp synthase αβ-dimer. This is the closed conformation of the βK87T mutant complexed with L-Trp (ref) and Na⁺. The ion pair formed from αD56 and βK167 can be seen at the interface of the α-subunit (left side, yellow) and the β-subunit (right side, magenta).

When the Trp synthase-L-Ser complex is subjected to rapid pressure changes (P-jumps), relaxations are observed. If the P-jumps are performed in about 100 μsec, up to three relaxations can be observed by fluorescence.⁷ The fastest relaxation exhibits a rate constant of about 400–500 sec⁻¹ in the presence of NH₄⁺ over the pressure range from 1 to 40 MPa.⁷ This relaxation may involve E_EA–Ser and an earlier intermediate, a gem-diamine of the internal aldimine and L-Ser. The slower relaxations, with k_obs values ranging from 1 to 20 sec⁻¹, can be conveniently observed in a P-jump instrument with a dead time of about 5 msec, as shown in Figure 4.⁹ These slower relaxations involve the E_EA–Ser and E_AA complexes and the conformational change between open and closed states. Relaxations of reversible equilibria are the sum of a forward and reverse rate constant, and the equilibrium constant is the ratio of the forward and reverse rate constant. Thus, the individual component rate constants can be calculated from the observed relaxation rate constant and the equilibrium constant. The relaxations shown in Figure 3 are pressure dependent, and the plots of log k versus pressure are distinctly nonlinear (Fig. 5).⁹ Nonlinearity in pressure-dependent kinetics can arise either from kinetic complexity due to coupled equilibria or from compressibility. In this case, we believe that the results are best explained as due to compressibility, since the data show smooth curves, and we are directly monitoring the conformational change. In the case of the Na⁺ form of enzyme (Fig. 5A), the curvature is slight and concave, while for the NH₄⁺ form and for Na⁺ with benzimidazole (BZI), the curvature is much more pronounced and convex (Fig. 5B and C). Fitting these data to Equation 5 provides the values of the activation

$${\mathrm{k}}_{\mathrm{p}}={\mathrm{k}}_0\ast \exp (-\mathrm{p}(\mathrm{\Delta }{\mathrm{V}}_{\mathrm{o}}^{‡}/\mathrm{RT})-{\mathrm{\beta }}_{\mathrm{o}}^{‡}{\mathrm{p}}^2)$$ (5)

volumes, ΔV^‡_o, and the compressibilities, β_o^‡. for the rate constants, k_o and k_c, given in Table 2. The negative values of β_o^‡ for the NH₄+ and Na⁺ with BZI indicate that the transition state between E_AL–Ser and E_AA is less compressible that the reactant and product states. This suggests that these transition states are more highly solvated or tightly packed than the ground states. In contrast, the positive β_o^‡ for the Na⁺ enzyme suggest that the transition state is less solvated or more loosely packed than the reactant and product states. This difference in β_o^‡ for the NH₄⁺ and Na⁺ forms suggest that the rate-determining steps in the relaxation are different. This is consistent with steady-state kinetic isotope effect data with α-[²H]-Ser that show a large value of about 6 in the presence of Na⁺ but a small value near 1 in the presence of NH₄⁺.³¹ Thus, the chemical step appears to be rate determining for relaxation of the Na⁺ enzyme, while the conformational change is rate determining for the NH₄⁺ enzyme. The relaxation of the Na⁺ enzyme with BZI is more complicated, since it is biphasic, and much slower than for Na⁺ alone. BZI can bind as an indole analogue to either the α-site or the β-site in Trp synthase, but in the presence of Ser, it selectively binds to E_AA in the β-site.³² The very large negative values of β_o^‡ for the data with BZI indicate a large contribution of solvation to the relaxation, so it is possible that the BZI must dissociate from the enzyme as part of the relaxation, and the slow rate suggests that there is a slow conformational change associated with BZI binding. This is also consistent with the very large ΔV^‡_o for relaxation of the BZI system (Table 2). For the relaxations of the Na⁺ and NH₄⁺ enzymes, the transition state volume is between the volume of the reactant and product, but for the Na⁺/BZI enzyme, the transition state volume is more negative than the reactant or product. BZI probably hydrogen bonds to β-Glu109, similar to what is seen in the crystal structure of the K87T-L-Trp complex²⁸ and the quinonoid complex of dihydroisotryptophan,²⁹ where in the latter β-Glu109 forms a hydrogen bond with the indoline NH. It is also interesting to note that the kinetics of relaxation of Trp synthase-L-Ser complex are different for positive or negative P-jumps (Fig. 4 and Table 2). This result suggests that the open and closed conformations exist as ensembles of microstates, the distribution of which is pressure dependent.

Figure 4.

Relaxations of Trp synthase-L-Ser complex with Na⁺ after P-jumps. The upward relaxation follows a jump from 74.5 to 91 MPa, and the downward relaxation follows a jump from 92 to 76 MPa. The dotted lines are the results of exponential fits to the data.

Figure 5.

Effect of pressure on relaxation of Trp synthase-L-Ser complex. Filled symbols correspond to the forward rate constant, k_o, and open symbols correspond to the reverse rate constant, k_o. (A) Relaxations in the presence of Na⁺. Circles indicate results from positive P-jumps and squares are from negative P-jumps. (B) Relaxations in the presence of NH₄⁺. Circles indicate results from positive P-jumps and squares are from negative P-jumps. (C) Relaxations in the presence of Na⁺ and benzimidazole. Circles and squares indicate results from positive P-jumps, while triangles are from negative P-jumps.

Table 2.

Parameters for relaxation of Trp synthase serine complexes⁹

	Positive P-jump						Negative P-jump
Cation ligand	ko,s−1	ΔVo‡ mL/mol−1	βo‡ mL/MPa−1	kc, s−1	ΔVc‡ mL/mol−1	βo‡ mL/MPa−1	ko,s−1	ΔVo‡ mL/mol−1	βo‡ mL/MPa−1	kc,s−1	ΔVc‡ mL/mol−1	βo‡ mL/βoMPa−1
Na+	1.09	−12	0.397	17.3	125	0.397	0.34	−48	0.230	5.45	89	0.229
NH4+	0.014	−127	−0.588	47.5	36	−0.630	0.0043	−153	−0.736	14.4	8	−0.736
Na+	0 (1.8 × 10−6)	−279	−1.46	0.139	−89	−1.46	0 (2.0 × 10−8)	−235	−1.32	0.0015	−46	−1.32
BZIa
	0 (3.9 × 10−8)	−349	−1.84	0.003	−160	−1.84	0 (3.2 × 10−7)	−285	−1.59	0.0247	−96	−1.59

^aValues in parentheses are calculated on the basis of k_c and K_eq from Table 1.

The Trp synthase-Trp-GP, a product analog complex, also shows pressure-dependent absorption spectra.⁸ The prominent quinonoid band of the complex at 476 nm decreases with increasing pressure, eventually disappearing. This result is also consistent with a conformational equilibrium. Fitting of these data provides values of K_eq = 7.3 × 10⁻³ and V_o = −97 mL/mol for the conformational change (Table 1). Thus, the hydrostatic pressure results suggest that the allosteric equilibrium in Trp synthase is balanced to optimize the reaction and to avoid substrate and product inhibition (Scheme 2). IGP and L-Ser bind to form an open conformation complex which is in equilibrium with a closed conformation. The K_eq is about 370 in the direction of the closed conformation for the Na⁺ enzyme. Within the closed complex, cleavage of IGP to indole and G3P occurs at the α-site, and the indole reacts with E_AA at the β-site to form a closed L-Trp G3P complex. Opening of the L-Trp complex then allows the products, L-Trp and G3P, to be released. Trp synthases from piezophilic organisms will probably show altered responses to pressure so that they can function optimally under their adapted environmental conditions.

Scheme 2.