Thermodynamics of the reactions catalyzed by the multifunctional enzyme complex tryptophan synthase

ARCHIVES OF BIOCHEMISTRY AND BIOPHYSICS Vol. 242, No. 2, November 1, pp. 440-4461985

Thermodynamics of the Reactions Catalyzed by the Multifunctional Enzyme Complex Tryptophan Synthase HEINRICH

WIESINGER

AND HANS-JijRGEN

Institut fiir Biophysik und Physihdische Biachmie der Univemiti Univmsit&sstrasse $1, D-8.&30 Regemburg, West G-ny

HINZ’ Regendmrg,

Received March 13,1985, and in revised form June 19, 1985

The intrinsic enthalpy changes (corrected for hydration of D-glyceraldehyde 3-phosphate) for the reactions catalyzed by the a and /& subunits of tryptophan synthase from Escherichia coli have been determined calorimetrically. Cleavage of indoleglycerol phosphate (a reaction) was found to be associated with a AH value of 54.0 f 2.5 kJ mol-‘, while condensation of indole with L-serine (/3 reaction) involved -80.3 + 4.6 kJ mol-‘. By direct determination of the enthalpy concomitant with the overall synthesis of tryptophan from indoleglycerol phosphate and L-serine an enthalpy value of -13.4 f 5.6 kJ mol-’ was observed. In view of the uncertainties of the literature data used for calculation of the hydration contribution, the agreement between the directly measured AH value of the overall reaction and the sum of the enthalpies of the (Y and p reactions is fair. Deamination of L-serine, a side reaction catalyzed preferentially by the isolated & pyridoxal 5’-phosphate2 subunit, was shown to be associated with an enthalpy change of -7.3 IL 0.4 kJ mol-‘. o 19% Academic press. ~nc.

Tryptophan synthase from Escherichia coli is a bienzyme complex composed of two CYsubunits and one fiz subunit; for proper function the pz subunit requires two molecules of the covalently bound coenzyme pyridoxal 5’-phosphate (PLP)’ (12). The native complex catalyzes the last step in the biosynthesis of tryptophan,

in the complex by the a and ,& subunits, respectively: (YReaction IGP = indole + D-GAP PI

Indoleglycerol

Reaction [2] is also catalyzed by the (Ysubunit alone, but the reaction proceeds by a factor of 100 more slowly than in the complex. Similarly, Reaction [3] is also catalyzed by the isolated pz subunit; however, in the tetrameric complex the reaction velocity is 30 times higher (22, 10, 11). In an analogous manner, deamination of L-serine, yielding pyruvate and ammonia, occurs as a side reaction catalyzed by the fiz subunit, but in the tetrameric complex it proceeds at only vanishing rate (4). Despite a vast amount of studies on the mechanism of tryptophan synthase (12,10, 11,18, 19), no energetic parameters except the free

p Reaction Indole + L-serine L-tryptophan

phosphate (IGP)

+ L-serine - L-tryptophan + D-glyceraldehyde

3-phosphate (D-GAP)

+ HzO.

[l]

The overall reaction can be decomposed into two subreactions which are catalyzed r To whom correspondence should be addressed. * Abbreviations used: D-GAP, rr-glyceraldehyde 3phosphate; IGP, indoleglycerol phosphate; PLP, pyridoxal 5’-phosphate; NaPP, sodium pyrophosphate; DTE, dithioerythritol; GAPDH, glyceraldehyde-gphosphate dehydrodenase. 0003-9861/85 $3.00 Copyright All rights

0 1985 by Academic Press, Inc. of reproduction in any form reserved.

440

+ HzO.

[3]

THERMODYNAMICS

OF

TRYPTOPHAN

energy change of the reversible Reaction [2] have been reported. Knowledge of such data is essential for understanding energy transduction and linkage phenomena in enzymes (16,14,7). In this paper, we provide for the first time calorimetrically determined enthalpy data associated with the overall reaction catalyzed by a multifunctional and multimeric enzyme complex, and compare them with the enthalpies involved in catalysis of the subreactions of the isolated subunits. MATERIALS

AND

METHODS

Materials. The OL subunit of tryptophan synthase was prepared following the procedure reported by Kirschner et al (8) and was stored as an ammonium sulfate suspension at 4°C. The &subunit was prepared as described by Bartholmes et al (2) and kept at -75°C in 0.6 M potassium phosphate buffer, pH 7.8, 1 mM DTE, 5 mM EDTA, 2 mM PLP, until use. The concentration of a protein was determined at 278 nm using an extinction coefficient p-i% = 0.46 cm2 mg-‘. The concentration of holo-& protein was determined at 290 nm, after diluting the protein 1:20 in 0.1 M NaOH using an extinction coefficient of @*s = 0.75 em2mg-’ (1). As at pH 13 the coenzyme PLP shows absorption bands at 388 and 290 nm, the contribution of this latter absorption has to be corrected for according to the equation (21). [/zoZo$z] = (Ezso - 0.091. Em). 2010.75 (mg/ml). Before each experiment the enzymes were dialyzed extensively against the respective buffer. Lactic dehydrogenase from pig muscle and glyceraldehyde 3-phosphate dehydrogenase (GAPDH) from rabbit muscle were purchased from Boehringer, Mannheim. PLP was obtained from Serva, Heidelberg, DL-Glyceraldehyde 3-phosphate has been obtained as the barium salt of the diethylacetal from Sigma, Munich. All other commercially available chemicals were purchased from Merck, Darmstadt, and were of reagent-grade quality. IGP was isolated from the growth medium of E. coli A2/F’A2 as described by Creighton (5) and was a gift of Dr. Peter Bartholmes, Regensburg. Unless otherwise stated, all experiments were performed in 0.1 M sodium pyrophosphate buffer, pH 7.5. For measurements in the presence of the eoenzyme the buffer solution contained 0.02 mM PLP; when 0.2 M Tris buffer was used, the PLP concentration was 0.08 mM. Methods. The reactive species of glyceraldehyde 3phosphate was prepared from the masked aldehyde by a brief, boiling with Dowex-50 directly prior to the calorimetric measurements. Its concentration was determined with GAPDH according to the instructions furnished by Sigma. IGP was purified from in-

SYNTHASE

441

COMPLEX

doleglycerol and other decomposition products by chromatography on a Sephadex G-25 column (1 cm* X 1 m); the concentration was determined in a coupled assay of tryptophan synthase and GAPDH in a cuvette containing 0.3 ml of 0.2 M L-serine, 0.1 ml of 0.4 mM PLP, 0.06 ml of 0.2 M NaAs04, 0.05 ml of 20 mM NADH, 0.01 ml of 0.1 M DTE, 0.1 ml of 1 M Tris, 0.240 ml HzO, 0.005 ml of GAPDH-suspension, 0.1 ml of flZ protein (approx. 200 YU/ml), and 0.025 ml of the IGP-solution. After 5 min the reaction is started by adding 0.01 ml of LYsuspension (20 mg/ml). The absorption at 340 nm is recorded until a plateau is reached, an extinction coefficient of 6.22 * lo3 Me1 cm-’ for NADH is used. The concentration of pyruvate was determined with LDH according to Bergmeyer (3). Absorption measurements were performed at 25 + 0.2”C using a Varian Cary 118 spectrophotometer equipped with a thermostatted cell holder. pH values were measured at the temperature of the respective experiment using a digital pH meter (WTW, Weilheim). The calorimetric measurements were performed in a LKB calorimeter fitted with gold cells. The instrument was charged with approximately 4 ml of a solution of the respective substrates; the reaction was started by mixing the sample solutions with 2 ml of a diluted solution of the enzyme in the appropriate buffer. Heats of dilution were determined in separate experiments and used for correcting the total heat effect, if necessary. The calorimeter was calibrated electrically. The molar reaction enthalpies have been calculated on the basis of the number of moles of the limiting reagent after appropriate correction required to account for heat effects stemming from side reactions (see Results). Some additional experiments were performed in a flow microcalorimeter (see Fig. 3). The experimental details have been described elsewhere (17,21). RESULTS

AND

DISCUSSION

Condensation of Indole and L-Serine Fwm L- Tryptophan

to

The irreversible condensation of indole and serine yielding tryptophan is catalyzed by the f12subunit. Because the substrates are stable under the experimental conditions, extensive calorimetric investigations could be undertaken, including variation of temperature and buffer and enzyme concentrations. Both indole and serine were employed in reaction-limiting concentrations, to ensure uniqueness of the thermodynamic parameters. The substrates indole and tryptophan have different extinction coefficients at 279 nm; therefore, the formation of tryptophan

442

WIESINGER

AND

r-------

;;

I

5

0

10 t, min

FIG. 1. Comparison of the heat effect, AQ (-), the change in absorption at 279 nm (---; arbitrary units) for the condensation of indole and L-serine alyzed by the aPLPz subunit at 25”C, pH 7.5.

and OD cat-

could be followed by recording the absorption at this wavelength. The same end point is found for the integral heat effect of the reaction and the change in absorbance at 279 nm, thus providing evidence that the enthalpy measured in the calorimeter stems from the same reaction observed in the spectrophotometer (Fig. 1). The experimental results are listed in Table I, the following aspects are worth being emphasized: The reaction enthalpy is identical, independent of whether serine or indole are used as limiting substrate, and is not dependent on the enzyme concentration. Therefore, contributions from binding enthalpies of the ligands to the TABLE

HINZ

proteins need not be taken into account. The same holds for contributions from the deamination of L-serine. Dissociation of the tetrameric complex into two LYand the ,& subunits does not take place under the conditions of the experiment. The reaction enthalpy does not vary with the buffer used. AH is identical for sodium pyrophosphate and Tris buffers. This result indicates that no proton flux is involved in the reaction. The result of identical enthalpies in the two buffers agrees well with the known reaction mechanism of tryptophan formation for which no net change in the state of ionization of the reactants is reported. The reaction enthalpy does not vary with temperature, i.e., no change in heat capacity occurs. This is understandable, since tryptophan synthase is present only in catalytic amounts and the heat capacities of the small reactants and products are practically indistinguishable. In the present study it was demonstrated that the reaction enthalpy is the same independent of whether the condensation reaction is catalyzed by the isolated Pz subunit or by the tetrameric q& complex. The kinetics, however, are strongly affected by the presence of the (Y subunit, as can be seen by the different time course of heat I

ENTHALPYOFTHECONDENSATIONOFINDOLEANDL-SERINECATALYZEDBY holo-TRYPTOPHANSYNTHASEAT~H~.~

Ial (Xl@

17 54 12 46 5.5 17 54 54

lP21 M)

Temperature

Limiting substrate

-AH (kJ mol-‘)

(Xlo'M)

Buffer

(“C)

10

NaPP

25

Indole

74.9

159.9 14 14 9.9 15 3 14 1.5 9.9 15 9.9 15

NaPP NaPP NaPP NaPP NaPP NaPP NaPP Tris NaPP NaPP NaPP NaPP

25 25 25 25 25 25 25 25 25 35 35 35 10

Indole Indole Indole Indole Indole Indole L-Serine Indole Indole Indole Indole Indole

84.5 75.3 86.6 77.4 72.8 87.9 82.0 83.7 81.2 74.9 84.9 82.0 78.7

THERMODYNAMICS

OF

TRYPTOPHAN

evolution with isolated&or with && (Fig. 2). Using the a&zPLPz complex as catalyst the reaction starts immediately after mixing the reactants in the calorimeter. All material has reacted within the time required for mixing, i.e., within approximately 20 s, and the exponential decay only reflects the time constant of the calorimeter. A much slower heat evolution if observed when the isolated p2 subunit is used as catalyst in the presence of PLP. The heat is evolved slowly, until the reaction is finished, approx 6 min (Figs. 1 and 2). Comparison of the time course of these heat effects yields a factor of 30, which is identical to the activity difference reported for the isolated ,& subunits and the aePa complex (22, 10). Increased rate constants for enzymecatalyzed as compared to uncatalyzed reactions have their origins in a decrease of free energy of activation and a concomitant higher population of the transition state. The equilibrium of the reaction is not changed by the catalyst, since that is characterized by the free energy difference between the initial and final states. The same argument holds for the explanation of the increase in activity of the (Y&& complex relative to the isolated flz subunit. The present result of a constant reaction enthalpy de-

SYNTHASE

I

443

COMPLEX

! 0

10

20 t,

30

40

I

min

FIG. 3. Schematic diagram of the heat effect, AQ, associated with mixing&PLPz at 35°C with saturating (a) and at 35°C with nonsaturating (b) L-serine in a flow calorimeter at pH 7.5. A, Successive mixing in the calorimetric cell; pulse duration, 50 s.

spite a very different time course of heat evolution for tryptophan formation from indole and L-serine demonstrates, for the first time experimentally for AH, that association of functional subunits to a multienzyme complex only influences the activation energies of the process, whereas the reaction enthalpy, as a state function, necessarily remains the same. An average enthalpy of -80.3 -t 4.6 kJ mall’ can therefore be assigned to the condensation of indole and L-serine yielding L-tryptophan as catalyzed by tryptophan synthase. Deamination of L-Se&e holo-& Subunit

by the

Deamination of L-serine is a side reaction catalyzed by the p2 subunit of tryptophan synthase in the presence of PLP (4). The reaction enthalpy was determined in two different ways. (a)

I

I

0

5

10 t,

min

FIG. 2. Comparison of the time course of the heat effect, AQ, associated with the condensation of indole and L-serine at 25”C, pH 7.5, catalyzed by the &PLPz subunit (---) and the o&PLPz complex (-) under identical substrate and enzyme concentrations.

Flow-cabrimetric

measurements.

Studies on the binding of L-serine to the holo-/ subunit in the flow calorimeter at 25°C resulted in a complex pattern of the heat effect (Fig. 3a). After the first mixing of an excess of L-serine with the fiZ subunit an exothermic steady state of heat evolution is established. Each subsequent mixing of fresh enzyme with serine results in an exothermic peak on top of an exothermic steady-state signal. This signal arises from the fact that, due to the length of the outflow tubing behind the mixing chamber, there is always some slowly reacting solution present in the calorimeter. There-

444

WIESINGER

fore, the individual peaks represent the heat of binding of the ligand to unliganded protein, whereas the constant heat evolution is caused by the very slow deamination of the bound substrate. Observation of the binding peak on top of the slow, steadystate heat output is possible because the binding reaction is much faster than the deamination; in the first mixing of @zPLPz and L-serine the exothermic binding reaction cannot be differentiated from the heat release resulting from deamination. To establish the correctness of the interpretation unambiguously, measurements have been performed at 35°C using both excess and nonsaturating concentrations of L-serine. While the heat output vs. time curve, when employing excess serine at 35”C, was analogous to that at 25°C showing binding peaks superimposed on a steady-state exothermic reaction, the heat vs. time curve observed when using nonsaturating concentrations of L-serine is quite different (Fig. 3b). The rate of deamination at this temperature is increased so that all of the substrate reacts within approximately 15 min. From the binding peaks at 35”C, when using excess serine, the enthalpy of binding of L-serine to flBPLPz can be determined and amounts to -24.7 k 1.7 kJ mol&-l (21). Curve 3b comprises the binding enthalpy and the enthalpy resulting from deamination. If the binding contribution of -24.7 + 1.7 kJ mol @;’ is subtracted from the total heat effect, an enthalpy of -7.1 kJ mol serine-’ can be calculated for the deamination reaction. (b) Batch-calorimetric measurements of the deamination reaction. The AH value for deamination calculated from the tlowcalorimetric studies is in excellent agreement with enthalpy values obtained by independent batch-calorimetric measurements. In these experiments large amounts of serine were mixed with catalytic quantities of the PzPLPz subunit. At 25°C under these conditions the reaction is still too slow to be measured to completion. However, at 35°C the reaction is finished within 1 h. The enthalpy value resulting from the batch-calorimetric studies at 35°C is -7.3 + 0.4 kJ mol serine-‘. Enzymatic analysis of the mixture after each measurement (see

AND

HINZ

Materials and Methods) showed that serine had been quantitatively deaminated to pyruvate. Thus, the AH value of -7.3 f 0.4 kJ mol-’ can be unambiguously attributed to the deamination of L-serine at 35°C. Formation

of Indoleglycerol

Phosphate

The equilibrium constant (Ke = 0.44 mM) of Reaction [2] favors the educt IGP (18, 19), therefore, formation of IGP from DGAP and indole was investigated in the calorimeter. Only the (~Zpz complex was used to catalyze the reaction, since with the isolated (Ysubunit the synthesis of IGP proceeds at a very slow rate. Determination of the exact reaction enthalpy is complicated by the fact that D-GAP exists in solution predominantly as the hydrated form (diol), whereas the free aldehyde is the reactive species (15, 13). In order to obtain the intrinsic enthalpy values of the (Y reaction, corrections have been applied to the observed heat effects as described in the Appendix. The enthalpy value employed in the calculations for the hydration of DGAP was -23.8 kJ/mol carbonyl group, since literature data for hydration of carbony1 groups attached to aliphatic chains range from -23.4 to -24.7 kJ mol-’ (9,6). The results of the measurements are listed in Table II; the reaction enthalpies are independent of the limiting substrate, the temperature, and the enzyme concentration (see discussion above), and can be averaged as 54.0 + 2.5 kJ mol-’ for the cleavage of IGP catalyzed by the ad2 complex. Overall Reaction a&& Complex

Catalyzed

!q the holc-

The enthalpy of the overall reaction of tryptophan synthase (Reaction [l]) was measured in the batch calorimeter by mixing IGP and L-serine with catalytic amounts of the holo+& complex. During the reaction stoichiometric amounts of DGAP are produced exclusively in the carbony1 form, which then is hydrated according to the equilibrium constant K = [aldehyde]/[diol] = 0.19 at 25”C, pH 7.5 (13, 15). Therefore, the observed heat ef-

THERMODYNAMICS

ENTHALPY

OF THE REACTION

OF

D-GAP

(xl@

%@z %& %P2 a282 ~282 azP2 a202 %& %!& cyzP2

TABLE

II

e IGP

CATALYZED

+ INDOLE

M)

(Xl@ 7.9 1.6 7.9 0.7 0.7 0.6 0.7 7.9 0.7 0.7

17.6 1.5 7.6 0.5 0.5 0.5 0.5 7.6 0.5 0.5

M)

NaPP NaPP NaPP NaPP NaPP NaPP NaPP NaPP NaPP NaPP

TABLE

III

ENTHALPY OF THE REACTION IGP + L-SERINE -+ L-TRYPTOPHAN + D-GAP + HZ0 CATALYZED BY TRYPTOPHAN SYNTHASE IN 0.1 M NaPP AT 25”C, pH 7.5

[aI a202 %h 432

(Xl@ 7.7 1.1 5.9

[Pzl M)

(Xl@ 2.7 0.6 2.8

M)

-AH (kJ mol-‘) 20.5 8.8 11.3

445

COMPLEX

BY TRYPTOPHAN

Temperature (“C)

Buffer

fects again have to be corrected for the contributions from hydration of the aldehyde, using the enthalpy of hydration of -23.8 kJ mol-‘. The calculations have been illustrated in the Appendix. The corrected enthalpy values have been summarized in Table III. The average AH value obtained from three experiments for the formation of tryptophan from IGP and L-serine according to Reaction [l] is -13.4 + 5.6 kJ mol-‘. Since the differences in hydration of the aldehyde have been corrected for, the sum of the AH values of the subreactions should be equivalent to the AH value observed for the overall reaction catalyzed by the c&?~PLP~ complex. Addition of the two AH values yields 54.0 + 2.5 + -80.3 + 4.6 = -26.3 f 7.1 kJ mol-‘. This calculated enthalpy has to be compared with the measured value of -13.4 f 5.6 kJ mol-‘. In view of the size of the hydration corrections and

Enzyme

SYNTHASE

[Pzl

[aI Enzyme

TRYPTOPHAN

SYNTHASE

AT PH 7.5

Limiting substrate

25 25 25 25 25 25 25 25 35 35

-AH (kJ mol-‘)

Indole Indole Indole Indole Indole Indole Indole D-GAP Indole Indole

53.1 56.9 50.6 54.0 54.4 52.3 51.0 57.8 54.4 54.0

the experimental difficulties involved in the determination of the overall reaction enthalpy due to the instability of IGP, the agreement of the AH values is fair. APPENDIX:

HYDRATION

CORRECTION

An aldehyde:diol ratio of 0.19 implies that 84% of GAP is hydrated at equilibrium. Thus Reactions [l] and [2] can be written as IGP - Indole

+ GAP (H,O),

and IGP + Ser -

Trp + GAP (H,O),

+ HzO,

with x = 0.84. It follows that the observed enthalpy in each case is AH,,,,,, = AH, + x of A IiHydr, where AH,,,,,, = AQ,,,,/moles reactant. REFERENCES 1. ADACHI, O., KOHN, L. D., AND MILES, E. W. (1974) J. Biol. Chem 249,7756-7763. 2. BARTHOLMES, P., KIRSCHNER, K., AND GSCHWIND, H.-P. (1976) Biochemistry 15,4712-4717. 3. BERGMEYER, U. (1974) Methoden der Enzymatischen Analyse, 3rd ed., Vol. 2, Verlag Chemie, Weinheim. 4. CRAWFORD, I. P., AND ITO, J. (1964) Proc. N&l. Acad Sci. USA 51,390-397. 5. CREIGHTON, T. E. (1970) Eur. J. B&hem. 13. l10. 6. HINE, J., GREEN, L. R., MENG, P. C., AND THIAGARAJAN, V. (1976) J. Org Chem 41,3343-3349. 7. HINZ, H.-J. (1983) Annu. Rev. Biqphys. Bioeng. 12, 285-317.

446

WIESINGER

AND

8. KIRSCHNER, K., WISCOCIL, R., FOEHN, M., AND REZEAU, L. (1975) Eur. J. Biochem. 60,513-523. 9. KURZ, J. L. (1967) 3528. 10. LANE, A., Biochem.

J. Amer.

AND KIRSCHNER, 129,561-5’70.

Chem. K.

Sot. 89, 3524(1983)

Eur.

J.

11. LANE, A., AND KIRSCHNER, K. (1983) Eur. J Biochem. 129,571-582. 12. MILES, E. W. (1979) Adv. Enqmol 49,127-186. 13. PECZON, B. D., AND SPIVEY, H. 0. (1972) Biochemistry 11,2209-2217. 14. PETTIGREW, D. W., ROMEO, P. H., TSAPIS, A., THILLET, J., SMITH, M. L., TURNER, B. W., AND ACKERS, G. K. (1982) Proc. Nat1 Acad Sci. USA 79,1849-1853.

HINZ

15. TRENTHAM, D. R., MCMURRAY, C. H., AND POGSON, C. I. (1969) Biochem J. 114,19-24. 16. WEBER, G. (1975) Adv. Protein Chem. 29,1-83. 17. WEBER, K., AND HINZ, H.-J. (1976) Rev. Sci In&r. 47,592-594. 18. WEISCHET, W., AND KIRSCHNER, K. (1976) Eur. J Biochem 65,365-373. 19. WEISCHET, W., AND KIRSCHNER, K. (1976) Eur. J. Biochem. 65,375-385. 20. WIESINGER, H., AND HINZ, H.-J. (1984) Biochemistry 23,4921-4928. 21. WIESINGER, H., AND HINZ, H.-J. (1984) Biochemist??/ 23,4928-4934. 22. YANOFSKY, C., AND CRAWFORD, I. P. (1972) in The Enzymes (Boyer, P. D., ed.), 3rd. ed. Vol. 7, pp. 1-31, Academic Press, New York.