Proc. Nati. Acad. Sci. USA Vol. 74, No. 1, pp. 111-114, January 1977

Biochemistry

Interaction of tetraiodofluorescein with a modified form of aspartate transcarbamylase (dye/enzyme activation/allosteric mechanism)

EVAN R. KANTROWITZ, LAWRENCE B. JACOBSBERG, SCOTT M. LANDFEAR, AND WILLIAM N. LIPSCOMB Gibbs Chemical Laboratory, Harvard University, Cambridge, Massachusetts 02138

Contributed by William N. Lipscomb, October 15, 1976

ABSTRACT Low concentrations of the dye tetraiodofluorescein activate native aspartate transcarbamylase (aspartate carbamoyltransferase, carbamoylphosphate:L-aspartate carbamoyltransferase, EC 2.1.3.2), while high concentrations inhibit the enzyme's activity [Jacobsberg, L. B., Kantrowitz, E. B. & Lipscomb, W. N. (1975)J. Biol. Chein 250, 9238-92491 This dye is now shown to produce similar effects upon a modified form of aspartate transcarbamylase produced by Escherichia coli grown in a culture medium supplemented with thiouracil. Significantly, the ATP-induced activation is reduced in the modified form of the enzyme to the same extent as is the tetraiodofluorescein-induced activation. Thus, a relationship is demonstrated between the internal mechanisms by which ATP and tetraiodofluorescein activate aspartate transcarbamylase.

Aspartate transcarbamylase (aspartate carbamoyltransferase, carbamoylphosphate:L-aspartate carbamoyltransferase, EC 2.1.3.2) catalyzes the first step in Escherichia coil pyrimidine biosynthesis, the combination of carbamyl phosphate and aspartate into carbamyl aspartate and phosphate (1, 2). Feedback inhibition of the enzyme by CTP (3), the end product of the pathway, coupled with activation by ATP, the product of the parallel purine biosynthesis pathway (4), makes aspartate transcarbamylase a locus of control for the bacterial nucleotide biosynthesis. The nucleoside triphosphate effectors bind to "regulatory" sites on the enzyme molecule remote from the catalytic sites. Each aspartate transcarbamylase molecule contains six such regulatory sites and six catalytic sites. The physical characteristics and chemical properties of the enzyme have been extensively discussed (5, 6). We have previously investigated the interaction of the dye 2',4',5',7'-tetraiodofluorescein with aspartate transcarbamylase (7, 8). These studies have indicated that in addition to com-

peting with the substrates at the active site, the dye binds to the regulatory site of the enzyme to cause activation, just as does the physiological activator ATP. In contrast to ATP, however, a single tetraiodofluorescein molecule binding to one of the enzyme's six regulatory sites activates all six of the catalytic sites (8). Since the molecular details of tetraiodofluorescein's activating effect promise to be a key to the understanding of the allosteric mechanism of aspartate transcarbamylase, we have extended our study to an altered form of aspartate transcarbamylase which has recently been produced by allowing biosynthesis of the enzyme to take place in a medium containing thiouracil in place of uracil (9). This "thiouracil-modified aspartate transcarbamylase" does not exhibit the positive homotropic interactions among active sites which, in the normal enzyme, facilitate the binding of additional substrate molecules, once the first one is bound to the hexameric enzyme molecule. The modified enzyme is affected in the normal way by its activator ATP and by its inhibitor CTP, although both ligands act with reduced efficiencies. Because the homotropic interactions of

111

the enzyme are eliminated by the thiouracil modification, while the heterotropic interactions are only minimally altered, we hoped that a study of the interaction of tetraiodofluorescein with the thiouracil-modified enzyme would reveal the extent to which the tetraiodofluorescein-induced activation of the enzyme is related to each of these processes.

EXPERIMENTAL PROCEDURES Materials. The CTP, ATP, carbamyl phosphate, Tris, and L-aspartate were obtained from Sigma Chemical Co. The carbamyl phosphate was purified by precipitation from 50% ethanol and stored in a desiccator at -200. The disodium salt of 2',4',5',7'-tetraiodofluorescein, sometimes called erythrosin B, was obtained from Eastman. After purification by column chromatography on silica gel with 3% acetic acid in chloroform (vol/vol), the tetraiodofluorescein migrated essentially as a single band on a silica gel thin layer chromatogram (Eastman 6061) in the same solvent system. Tetraiodofluorescein concentrations were determined spectrophotometrically at 523 nm by using an extinction coefficient of 8.4 X 104 Mcm-1 at pH 7.0 (10). Aspartate Transcarbamylase and Thiouracil-Modified Enzyme. Aspartate transcarbamylase and the thiouracilmodified enzyme were isolated from large quantities of E. coli which were grown at the New England Enzyme Center from an overproducing mutant strain kindly provided by J. C. Gerhart. Aspartate transcarbamylase was purified by the procedure of Gerhart and Holoubek (11), while the thiouracil-modified enzyme was purified by modifications to this procedure suggested by Kerbiriou and Herve (9). The medium used for the growth of the thiouracil-modified enzyme was supplemented with 500 gg/ml of thiouracil. To prevent contamination by aspartate transcarbamylase produced by derepression in the inoculum, we also supplemented the medium used for the inoculum with 500,ug/ml of thiouracil. After purification, the enzyme was stored under sterile conditions in 0.04 M phosphate buffer at pH 7.0, 2 mM 2mercaptoethanol, and 0.2 mM EDTA in sealed vials under nitrogen at 5°. The enzyme was exhaustively dialyzed against 0.1 M Tris-acetate buffer at pH 8.3 before use. Enzyme concentrations were determined by absorbance measurements at 280 nm (by assuming an extinction coefficient of 0.59 cm2/mg for the native and modified enzyme), or by the method of Lowry et al. (12). Enzyme Activity. The transcarbamylase activity was assayed by continuously monitoring the enzyme-catalyzed release of protons at pH 8.3 on a Radiometer TTTlc pH-stat equipped with an SBUla syringe burette unit. Assays were routinely performed with 4.8 mM carbamyl phosphate and 30 mM aspartate. For assays with varying concentrations of ATP or tetraiodofluorescein, the aspartate concentration was reduced

Biochemistry: Kantrowitz et al.

112

Proc. Nati. Acad. Sci. USA 74 (1977)

15 H

>

H6-

4-

0

0o5 10 .

3.0

*

25 20 ls 3035 [ASPARTATE], mM . *

40

5

lo

1s

25

20

30

35

M

FIG. 2. Effect of N-(phosphonacetyl)-L-aspartate (PALA) on the enzyme activity of aspartate transcarbamylase and the thiouracil-modified enzyme. Enzymatic activity was determined by a modification of the pH-stat assay at pH 8.3 (see Experimental Procedures for details). The carbamyl phosphate and aspartate concentrations were 4.8 mM and 1.0 mM, respectively. In the absence of N-(phosphonacetyl)-L-aspartate (PALA), the absolute activity of native aspartate transcarbamylase (0) was 0.35 mmol/hr per mg and that of the thiouracil-modified enzyme (@) was 0.14 mmol/hr per mg.

,

2.5

0

106 [PALA],

B 2.0

I X

-

RESULTS 0.

Loss of homotropic properties in the thiouracilmodified enzyme 00

20

25

30

35

4D

[ASPARTATE], mM FIG. 1. Aspartate saturation curves of native aspartate transcarbamylase and thiouracil-modified enzyme. The enzymatic activity was determined by the pH-stat assay at pH 8.3 in the presence of 4.8 mM carbamyl phosphate. (A) Native aspartate transcarbamylase (7.8 gg/ml) alone (0), with 22.6MM tetraiodofluorescein (0), and 2.2 mM ATP (o). (B) Thiouracil-modified aspartate transcarbamylase (10.5 Mg/ml) alone (0), with 22.6MM tetraiodofluorescein (@), and with 2.4 mM ATP (0). to 12 mM, because at saturating aspartate (30 mM) little or no activation is observed. Activity in the figures is reported as specific activity in units of millimoles of carbamyl aspartate formed per hr per mg of enzyme. Assays at varying concentrations of N-(phosphonacetyl)L-aspartate were performed to check the cooperativity of the two types of enzyme. These assays were carried out at 1.0 mM aspartate, because only at low aspartate concentrations is homotropic activation by inhibitors observed (13). To prevent a drop in the aspartate concentration due to the depletion of the reagent, we added aspartate in the sodium hydroxide used to maintain constant pH (14). Spectrophotometric Measurements. All spectrophotometric measurements were made on a Zeiss PMQ II, Cary 14, or Beckman DB-GT spectrophotometer, in quartz cells having a 1 cm path length.

Data Analysis. The experimental data were fit to theoretical equations by a fitting and plotting program written in Fortran IV and implemented on PDP 11/45 computer. The program incorporated the MIT non-linear least squares subroutines LSMARQ and LSMERR (Applications Program 84).

In the absence of ligands, the native enzyme exhibits a sigmoidal aspartate saturation curve (Fig. IA) as a result of the homotropic interactions among the enzyme's active sites. Under the same conditions, the thiouracil-modified enzyme produces the hyperbolic saturation curve predicted by Michaelis-Menten kinetics for an enzyme molecule without such interactions among its active sites (Fig. 1B). To show that the thiouracil-modified enzyme has lost all cooperative interactions, we employed the transition state analog N-(phosphonacetyl)-L-aspartate. At very low concentrations of aspartate (1.0 mM), coupled with a saturating level of carbamyl phosphate, low concentrations of N-(phosphonacetyl)-L-aspartate cause a dramatic increase of enzyme activity with the native enzyme (Fig. 2) (15). Under these conditions, each N-(phosphonacetyl)-L-aspartate molecule which binds to a catalytic site enhances the reactivity of the enzyme's other active sites. As the N-(phosphonacetyl)-L-aspartate concentration is increased, however, the additional N-(phosphonacetyl)-L-aspartate molecules become numerous enough to compete effectively with the substrate molecules for the available catalytic sites. This competition leads to a decrease of enzyme activity, ultimately completely inhibiting the enzyme at a sufficiently high concentration of N-(phosphonacetyl)-L-aspartate. In contrast to the activation of native enzyme, the thiouracil-modified enzyme's activity is not increased over its initial level as the N-(phosphonacetyl)-L-aspartate concentration is raised in the assay mixture. Rather, only the continuous decline of activity is observed. Therefore, it may be concluded that the positive homotropic interactions which facilitate the binding of substrates at additional active sites of the native enzyme are absent from the enzyme produced in the presence of thiouracil.

Biochemistry:

Proc. Natl. Acad. Sci. USA 74 (1977)

Kantrowitz et al.

113

> 0.8 bi

10 Cr

l.2

0.2 0

0

*

1.0

0.5

1.5

2.0

2.5

3.0

104 [TIF], M FIG. 3. Activity studies with native aspartate transcarbamylase and thiouracil-modified enzyme. The change in activity is shown as a function of tetraiodofluorescein (TIF) concentration. Activity was determined by the pH-stat assay at pH 8.3 in the presence of 12 mM aspartate and 4.8 mM carbamyl phosphate. Native aspartate transcarbamylase (0) (1.14 ,g/ml) had an initial activity of 8.8 mmol/hr per mg. Thiouracil-modified aspartate transcarbamylase (@) (10.5 gg/ml) had an initial activity of 1.5 mmol/hr per mg.

Activation of thiouracil-modified enzyme by ATP and

tetraiodofluorescein

As seen in Fig. 1, ATP and low concentrations of tetraiodofluorescein exert similar effects upon the aspartate saturation curve of both native and thiouracil-modified enzyme, notwithstanding the lower specific activity of the thiouracil-modified enzyme. Both ATP and tetraiodofluorescein shift the Km to lower aspartate concentrations. Although low concentrations of tetraiodofluorescein activate the native enzyme, inhibition occurs as the concentration of the dye is increased (Fig. 3). The effect of tetraiodofluorescein on the thiouracil-modified enzyme is quite similar, although the extent of activation is not as great. The maximum activation of the native enzyme is about 1.2 times the maximum activation of the modified enzyme. As opposed to tetraiodofluorescein, which activates both enzyme forms only at low concentrations, ATP activates both forms even at high concentrations (Fig. 4). The maximum extent to which ATP activates the native enzyme, at saturating ATP, is 80%, compared to 57% for the thiouracil-modified enzyme.

DISCUSSION We have postulated that the effect of tetraiodofluorescein on the native enzyme is the result of two opposing effects (8). First, the binding of a single tetraiodofluorescein molecule to one of the six regulatory sites activates all six of the catalytic sites. Second, one dye molecule binding to any one of the catalytic sites inactivates only that one catalytic site. This same scheme can be used to explain the effect of tetraiodofluorescein on thiouracil-modified aspartate transcarbamylase. A mathematical formulation of the model gives rise to three adjustable parameters, which, when fitted to the data, report on internal aspects of the model. An activation constant corresponds to the fractional extent to which activated enzyme is more active than enzyme without effector; an inactivating constant corresponds to the fractional extent to which a tetraiodofluorescein molecule inactivates a catalytic site; and a dissociation constant specifies

0

0.2

0.4

0.6

08

ID0

1.2

102 [ATP , M FIG. 4. Activity studies with native aspartate transcarbamylase and thiouracil-modified enzyme. The change in activity is shown as a function of the ATP concentration. Activity was determined by the pH-stat assay at pH 8.3 in the presence of 12 mM aspartate and 4.8 mM carbamyl phosphate. Native aspartate transcarbamylase (0) (1.14 ,g/ml) had an initial activity of 8.8 mmol/hr per mg. Thiouracil-modified aspartate transcarbamylase (0) (10.5 gg/ml) had an initial activity of 1.5 mmol/hr per mg.

the affinity of the dye molecules for the enzyme. (All of the binding sites for tetraiodofluorescein were experimentally shown to have the same intrinsic site dissociation constant. The proper statistical factor must be applied to this average value to ascertain the stoichiometric equilibrium constant for the binding of any particular tetraiodofluorescein molecule to the enzyme.) When the data representing the effect of tetraiodofluorescein on native aspartate transcarbamylase were analyzed in terms of the mathematical model, the dissociation constant was satisfyingly close to the experimentally-determined value and the deactivating constant converged to unity, as would be expected for the normalized kinetic data. Thus, those parameters on which independent checks could be made had taken on reasonable values. The activating constant was 1.48, indicating that tetraiodofluorescein could increase the activity of each active site by 148%. With the inactivation component removed by setting the deactivating constant to zero, the model could be used to represent the activation of aspartate transcarbamylase by the nucleoside triphosphate effector ATP. However, since the heterogeneity of ATP binding (16) had not been taken into account, the binding and activation parameters were average ones, not strictly comparable with experiment. Under these conditions the model indicated that ATP can activate each aspartate transcarbamylase catalytic site by 80%. Analogous analysis in terms of the mathematical model has now been performed for the thiouracil-modified enzyme acted upon by tetraiodofluorescein and by ATP. The resulting activating constants are given in Table 1. The ratio of the tetraiodofluorescein-induced activation constants between native and thiouracil-modified aspartate transcarbamylase is equal, within experimental error, to the ratio of ATP-induced activation constants for the two enzyme forms. This result can be restated in another way to make its implications more clear: the ATP-induced activation is reduced in the thiouracil-modified form of aspartate transcarbamylase to the same extent as the tetraiodofluorescein-induced activation. That tetraiodofluorescein exerted its activating effect on aspartate transcarbamylase by acting as an analog of ATP was suggested by our direct chemical studies (8). That conclusion is now substantiated

114

Proc. Natl. Acad. Sci. USA 74 (1977)

Biochemistry: Kantrowitz et al.

Table 1. Parameters of the tetraiodofluorescein mechanism

Species* C6R6 +TIF +ATP TU-C6R6 +TIF +ATP

Activating constantt

Inactivating constants

K (M)

Ratio of activating constants TIF/ATP

1.48 ± 0.04 0.80 ± 0.02

1.02 ± 0.02

(5.28 ± 0.46) x 10-5 (5.36 ± 0.53) x 10-

1.85 ± 0.07

0.95 ± 0.04 0.57 ± 0.04

1.15 ± 0.06

(9.34 ± 1.17) x 10-5 (3.57 ± 1.20) x 100-4

1.67 + 0.14

0§

0§

* C6R6, native aspartate transcarbamylase; TU-C6R6, thiouracil-modified aspartate transcarbamylase; TIF, tetraiodofluorescein.

t Activating constant is the fractional extent to which the activated enzyme is more active than enzyme without effector.

Inactivating constant is the fractional extent to which a tetraiodofluorescein molecule inactivates a catalytic site. § Constant set to zero. See text for explanation.

and extended by this new observation of the thiouracil-modified aspartate transcarbamylase. In addition to our earlier proof that tetraiodofluorescein activates the enzyme by binding at the same site as ATP, we have now shown the effects of the two ligands to be reduced to the same extent by the same enzyme modification. Thus, the two ligands not only act at the same external "trigger site" on the surface, but cause their effect by the same internal mechanism, at least at that part of the allosteric mechanism which is disrupted by the thiouracil modification. 1. Jones, M. E., Spector, L. & Lipmann, F. (1955) J. Am. Chem. Soc.

77,819-820. 2. Reichard, P. & Hanshoff, G. (1956) Acta Chem. Scand. 10, 548-566. 3. Yates, R. A. & Pardee, A. B. (1956) J. Biol. Chem. 221, 757770. 4. Gerhart, J. C. & Pardee, A. B. (1962) J. Biol. Chem. 237,891896. 5. Gerhart, J. C. (1970) Curr. Top. Cell Regul. 2,275-325.

6. Jacobson, G. R. & Stark, G. R. (1973) in The Enzymes, ed. Boyer, P. D. (Academic Press, New York), Vol. 9, pp. 225-308. 7. Jacobsberg, L. B., Kantrowitz, E. R., McMurray, C. H. & Lipscomb, W. N. (1973) Biochem. Biophys. Res. Commun. 55, 1255-1261. 8. Jacobsberg, L. B., Kantrowitz, E. R. & Lipscomb, W. N. (1975) J. Biol. Chem. 250,9238-9249. 9. Kerbiriou, D. & Herve, G. (1972) J. Mol. Biol. 64,379-392. 10. Vigne, I. & Fondari, J. (1953) Bull. Soc. Chim. Biol. 20, 331332. 11. Gerhart, J. C. & Holoubek, H. (1967) J. Biol. Chem. 242, 2886-2892. 12. Lowry, 0. H., Rosebrough, N. J., Farn, A. L. & Randell, R. H. (1951) J. Biol. Chem. 193,265-275. 13. Gerhart, J. C. & Pardee, A. B. (1964) Fed. Proc. 23,727-735. 14. Chan, W. W-C. & Mort, J. S. (1973) J. Biol. Chem. 248,76147616. 15. Collins, K. D. & Stark, G. R. (1971) J. Biol. Chem. 246,65996605. 16. Matsumoto, S. & Hammes, G. G. (1973) Biochemistry 12,

3388-3394.