Kinetic and thermodynamic properties of wild-type and engineered

Feb 4, 1991 - Tim N. C. Wells, Jack W. Knill-Jones, Tamara E. Gray, and Alan R. Fersht*. MRC Unit for Protein Function and Design, University Chemical...
0 downloads 0 Views 767KB Size
Biochemistry 1991, 30, 5151-5156

5151

Kinetic and Thermodynamic Properties of Wild-Type and Engineered Mutants of Tyrosyl-tRNA Synthetase Analyzed by Pyrophosphate-Exchange Kinetics Tim N. C. Wells, Jack W. Knill-Jones, Tamara E. Gray, and Alan R. Fersht* MRC Unit for Protein Function and Design, University Chemical Laboratory, Lensfeld Road, Cambridge CB2 1E W,U.K. Received October 24,1990; Revised Manuscript Received February 4,I991

ABSTRACT: The first step of the reaction catalyzed by the aminoacyl-tRNA synthetases is the formation

of enzyme-bound aminoacyl adenylate. The steady-state kinetics of this step has conventionally been studied by measuring the rate of isotopic exchange between pyrophosphate and ATP. A simple kinetic analysis of the pyrophosphate-exchange reaction catalyzed by the tyrosyl-tRNA synthetase from Bacillus stearothermophilus is given in which all the observed rate and binding constants can be assigned to identifiable physical processes under a variety of limiting conditions. The free energies of binding to the enzyme of tyrosine, ATP, and the transition state for tyrosyl adenylate formation can be measured in relatively straightforward experiments. The excellent agreement between parameters measured in these experiments and those from earlier pre-steady-state kinetics confirms that the intermediates isolated in the presteady state are kinetically competent. The dissociation constant of ATP from the unligated enzyme, a constant that has previously been experimentally inaccessible, has been measured for wild-type and several mutant enzymes. The changes in enthalpy and entropy of activation on mutation have been measured by a rapid procedure for mutants that have altered contacts with tyrosine and ATP. Those mutants that have large changes of enthalpy and entropy of binding are likely to have structural changes and so warrant further examination by protein crystallography.

Dissection of the structure and activity of the tyrosyl-tRNA synthetase has relied heavily on analyzing engineered mutants by stopped-flow kinetics (Wells & Fersht, 1985, 1986; Ho & Fersht, 1986; Fersht, 1987). The first step of the reaction catalyzed by the tyrosyl-tRNA synthetase is activation, the formation of enzyme-bound tyrosyl adenylate (E-Tyr-AMP,' Scheme I). The rate constants for the formation of ESTyrAMP by the enzyme from Bacillus stearothermophilus may be monitored by stopped-flow measurements of the change in the intrinsic tryptophan fluorescence of the enzyme on forming the intermediate (eq 1; Fersht et al., 1975a).

E

+ Tyr + ATP s EeTyr-ATP

AF

E-Tyr-AMP

+ PPI

(1) An alternative approach is the use of steady-state kinetics. Often, however, this produces values of k,, and K, that are composed of combinations of several rate constants that must be deconvoluted. The general procedure for assaying activation by steady-state kinetics is the method of pyrophosphate exchange [see for example Duffield and Calvin (1946) and Cole and Schimmel(1970)], which is an example of chemical exchange at equilibrium whereby isotopically labeled pyrophosphate distributes between bulk pyrophosphate in solution and the /3,r phosphates of the ATP in solution. Cole and Schimmel(l970) investigated a variety of possible mechanisms for the isoleucyl-tRNA synthetase on the basis of steady-state I Abbreviations: T (Tyr), tyrosine; A (ATP), adenosine 5 ' 4 phosphate; PP,, pyrophosphate; T-A (Tyr-AMP), tyrosyl adenylate; [E-T-A]', enzyme-bound transition state for the formation of tyrosyl adenylate; E (enzyme), tyrosyl-tRNA synthetase; EST,enzyme-bound tyrosine etc.; Tris-HCI, tris(hydroxymethy1)aminomethane;BBOT, 2,5bis(5'-rerr-butylbenzoxazol-2-yl)thiophene; POPOP/PPO, 1,4-bis(5phenyloxazolyl)benzene-2,5-diphenyloxazole;Na4PPI,tetrasodium pyrophosphate. Rate constants are defined according to Scheme I; for ex= dissociationconstant of ATP from free enzyme etc. ample, K, = &&/, ,

0006-2960/91/0430-5151$02.50/0

Scheme I TY~/

E'Tyr

E

Kx E.Tyr.ATP

k & E.Tyr-AMP.PPi k.3

KPP

E.Tyr-AMP PPI

E.ATP

assumptions. Studies carried out monitoring the pre-steadystate kinetics of the reaction (Fersht et al., 1975a; Wells & Fersht, 1986) have shown that a simpler scheme is possible for the tyrosyl-tRNA synthetase from B. stearothermophilus. In this case, both the ATP and pyrophosphate dissociate from the relevant enzyme complexes at much faster rates than the chemical reaction on the enzyme. Pyrophosphate-exchange experiments can be designed that measure true dissociation constants and true rate constants, the parameters needed to construct free energy profiles (Fersht, 1985). We first show that the results of simple analysis of the activation reaction of the tyrosyl-tRNA synthetase measured by pyrophosphate exchange are identical with those from pre-steady-statekinetics and then use the scheme to produce novel data of interest. EXPERIMENTAL PROCEDURES Materials Reagents were obtained from Sigma, and radiochemicals were from Amersham International. Wild-type and mutant tyrosyl-tRNA synthetases (B.stearothermophilus) were expressed in Escherichia coli TG2 hosts ( r e d form of TG1; Gibson, 1984), with use of M13mp9 templates constructed as described previously (Carter et al., 1984), and purified to electrophoretic homogeneity according to Wells and Fersht (1986). Nitrocellulose discs, 2.5 cm with 0.22-pm pores, were obtained from Sartorius. The concentration of active enzyme was determined by active-site titration with 14C-labeledtyr0 1991 American Chemical Society

Wells et al.

5152 Biochemistry, Vol. 30, No. 21, 1991 osine (Fersht et al., 1975b). In all cases, the enzymes were at least 90% active. Methods Pyrophosphate Exchange. The pyrophosphate-exchange reaction was measured in 144 mM Tris-HC1 (100 mM TrisHC1/44 mM Tris, pH 7.78) buffer at 25 OC. Tetrasodium [32P]pyrophosphatewas obtained as a solid from Amersham International and dissolved in water to give a 100 mM solution. It was diluted with unlabeled tetrasodium pyrophosphate (100 mM, pH 7.8) to produce a final specific activity of 1-4 cpm/pmol. Reactions contained 2 mM pyrophosphate as standard conditions in all experiments except those where pyrophosphate dependence was being studied. Enzyme concentrations were typically 0.2-1 .O pM and were adjusted to ensure a convenient experimental time range. MgC1, was added to a concentration 8 mM in excess of the total ATP and pyrophosphate concentrations. Magnesium pyrophosphate is insoluble at higher concentrations and so the addition of substrate to the reaction system was ordered to minimize the possibility of precipitation. In the initial rate studies, at least four 25-pL samples were withdrawn from the reaction mixture before the reaction was 5% complete. For the complete time course for the approach to equilibrium, the preferred method, nine samples of 25 pL were removed over three half-lives. Samples were quenched into 400 p L of 3.5% perchloric acid and 1% activated charcoal, vortexed, and filtered through Whatman GFC filters (2.5-cm diameter). The retained charcoal was washed three times with 4 mL of 10 mM tetrasodium pyrophosphate, pH 2, and then with 2 mL of ethanol. Radioactivity was measured by scintillation counting, with BBOT as a scintillant. BBOT, 9 g, was dissolved in 1.5 L of toluene and 0.5 L of methoxyethanol. The data were fitted to either a first-order exponential (approach to equilibrium) or a linear correlation, with use of a least-squares regression program, ENZFITTER (Elsevier-Biosoft). Tyrosine and ATP Dependence. For each enzyme, the dependence of the reaction rate on both tyrosine and ATP was studied. One substrate was kept constant at a concentration much higher than its respective K, value. [Tyr] was 10Km for variation of ATP, and [ATP] was -2Km for variation of tyrosine. The concentration of the second substrate was varied over the range K m / 5 to 5Km (Fersht, 1985). For studies to determine k,fK,K’, and K,, tyrosine concentrations were less than K,, typically between KJ10 and Kt/ 100. Enzyme was present at 1 pM final concentration, and MgATP (pH adjusted to 7.8) varied in the range Ka/5 to 5Ka. Enzyme stocks were dialyzed against 144 mM Tris-HC1 buffer, pH 7.8, 100 pM tetrasodium pyrophosphate, 10 mM MgCI,, and 100 pM phenylmethanesulfonyl fluoride (PMSF) overnight, to remove enzyme-bound tyrosyl adenylate. Further overnight dialysis against 144 mM Tris-HC1 buffer, pH 7.8, IO mM MgCl,, and 100 pM PMSF was necessary to remove pyrophosphate. (The derivation of the conventional Michaelis-Menten equations assumes that the concentration of enzyme is much lower than that of the substrate. This is not necessary for the experiments here where [Tyr] and [ATP] are far below the dissociation constants from their complexes. The MichaelisMenten or steady-state equations break down at high enzyme concentration because the concentration of unbound substrate is depleted by the accumulation of ES complexes. But, when [SI