11
Eur. J. Biochem. 29, 36-46 (1972) Regulation of the Catabolic Ornithine Carbamoyltransferase of Pseudornonas f luorescens A Study of the Allosteric Interactions Victor XTALON Laboratoire de Microbiologie de l’Universit6Libre de Bruxelles, and Institut de Recherches du Centre d’Enseignement et de Recherches des IndustriesAlimentaires et Chimiques, Bruxelles (Received March 27,1972) The catabolic ornithine carbamoyltransferase from Pseudomonas fluoreseens catalyzes the phosphorolysis of citrulline leading to the formation of ornithine and carbamoylphosphate. The initial rates have been measured in the experimentally most accessible direction of the reaction, the carbamoylation of ornithine. The plots of reaction rates against carbamoylphosphate (sub- strate), phosphate (activator), and putrescine (inhibitor) are sigmoid shaped. The homotropic interactions between carbamoylphosphate molecules are decreased in the presence of phosphate and increased in the presence of putrescine. The allosteric properties of the enzyme are accounted for by the concerted transition theory proposed by Monod, Wyman and Changeux. The treatment of the kinetic data according to the method of Blangy, Buc and Monod suggests that the enzyme is an octamer made up of eight identical protomers which can exist in equilibrium in two conformational states which differ in their dissociation constant for the substrate, activator and inhibitor. The values of these con- stants for the binding of carbamoylphosphate, phosphate and putrescine, and the allosteric transition constant have been determined for the wild-type enzyme and for the enzyme of a mutant which can use the catabolic enzyme for anabolic purpose. In Pseudomonas fluoresceus the catabolism of arginine occurs through the pathway involving the three enzymes [l] arginine deiminase, ornithine carbamoyltransferase and carbamate kinase which, respectively, catalyze the three subsequent reac- tions : Arginine + citrulline + NIC, Phosphate + citrulline t; ornithine + carbamoylphosphate Carbamoylphosphate + ADP f ATP + CO, + NH,. paper we describe the regulatory properties of the catabolic ornithine carbamoyltransferase of Pseudo- monas fluorescens and we show that the allosteric behaviour of the enzyme, and particulary its high cooperativity towards carbnrnoylphosphatc ex- plains its inability to perform in vivo the thermo- dynamically much favoured reaction. Phosphate, a substrate of the enzyme in its physiological dircction, is an activator that abolishes the cooperativity for carbamoylphosphate. Moreover, in the presence of that effector, the substrate con- centration required to reach the half-maximum velocity decreases while the maximum velocity of the Eeaction increases. of arginine, Putrescine, a product of the catabolic pathway as a feedback inhibitor by Arginine deiminase catalyses an essentially irrevers- ible reaction [2]. The reversillle carbamate kinase WAction ATP [31* Therefore, the decreasing the carbamoylphosphate concentration thermodYnamicallY limiting st eP in the required to reach the half-maximum velocity whereas of arghine is the phosphorolysis of c i t r u k e because the equilibrium of the reaction strongly favours the maximal velocity itself is not affected. The binding of adenosine triphosphate, a product decreases the maximum velocity without affecting tration. Citru1line formation (K = ‘0‘) In the Preceding of the catabolic pathway is also cooperative. It carbamoylphosphokinase or ATl’ : carbarnate phospho- the carbamoylphosphate concell- Enzymes. Ornithine carbamoyltransferase (EC 2.1.3.3) ; transferase (EC 2.7.2.2); arginine deiminase (EC 3.5.34.

Regulation of the Catabolic Ornithine Carbamoyltransferase of Pseudomonas fluorescens : A Study of the Allosteric Interactions

Embed Size (px)

Citation preview

Eur. J. Biochem. 29, 36-46 (1972)

Regulation of the Catabolic Ornithine Carbamoyltransferase of Pseudornonas f luorescens

A Study of the Allosteric Interactions

Victor XTALON Laboratoire de Microbiologie de l’Universit6 Libre de Bruxelles,

and Institut de Recherches du Centre d’Enseignement et de Recherches des Industries Alimentaires et Chimiques, Bruxelles

(Received March 27,1972)

The catabolic ornithine carbamoyltransferase from Pseudomonas fluoreseens catalyzes the phosphorolysis of citrulline leading to the formation of ornithine and carbamoylphosphate. The initial rates have been measured in the experimentally most accessible direction of the reaction, the carbamoylation of ornithine. The plots of reaction rates against carbamoylphosphate (sub- strate), phosphate (activator), and putrescine (inhibitor) are sigmoid shaped. The homotropic interactions between carbamoylphosphate molecules are decreased in the presence of phosphate and increased in the presence of putrescine.

The allosteric properties of the enzyme are accounted for by the concerted transition theory proposed by Monod, Wyman and Changeux. The treatment of the kinetic data according to the method of Blangy, Buc and Monod suggests that the enzyme is an octamer made up of eight identical protomers which can exist in equilibrium in two conformational states which differ in their dissociation constant for the substrate, activator and inhibitor. The values of these con- stants for the binding of carbamoylphosphate, phosphate and putrescine, and the allosteric transition constant have been determined for the wild-type enzyme and for the enzyme of a mutant which can use the catabolic enzyme for anabolic purpose.

In Pseudomonas fluoresceus the catabolism of arginine occurs through the pathway involving the three enzymes [l] arginine deiminase, ornithine carbamoyltransferase and carbamate kinase which, respectively, catalyze the three subsequent reac- tions :

Arginine + citrulline + NIC,

Phosphate + citrulline t; ornithine + carbamoylphosphate

Carbamoylphosphate + ADP f ATP + CO, + NH,.

paper we describe the regulatory properties of the catabolic ornithine carbamoyltransferase of Pseudo- monas fluorescens and we show that the allosteric behaviour of the enzyme, and particulary its high cooperativity towards carbnrnoylphosphatc ex- plains its inability to perform in vivo the thermo- dynamically much favoured reaction.

Phosphate, a substrate of the enzyme in its physiological dircction, is an activator that abolishes the cooperativity for carbamoylphosphate. Moreover, in the presence of that effector, the substrate con- centration required to reach the half-maximum velocity decreases while the maximum velocity of the Eeaction increases.

of arginine, Putrescine, a product of the catabolic pathway

as a feedback inhibitor by Arginine deiminase catalyses an essentially irrevers- ible reaction [2]. The reversillle carbamate kinase WAction ATP [31* Therefore, the decreasing the carbamoylphosphate concentration thermodYnamicallY limiting st eP in the required to reach the half-maximum velocity whereas of arghine is the phosphorolysis of c i t r u k e because the equilibrium of the reaction strongly favours

the maximal velocity itself is not affected. The binding of adenosine triphosphate, a product

decreases the maximum velocity without affecting

tration.

Citru1line formation ( K = ‘0‘) In the Preceding of the catabolic pathway is also cooperative. It

carbamoylphosphokinase or ATl’ : carbarnate phospho- the carbamoylphosphate concell- Enzymes. Ornithine carbamoyltransferase (EC 2.1.3.3) ;

transferase (EC 2.7.2.2); arginine deiminase (EC 3.5.34.

Vol.29, No.1, 1972 v. STALON 37

In this paper, we present an attempt to interpret quantitatively the kinetic behaviour of the catabolic ornithine carbamoyltransferase in terms of the concerted transition theory proposed by Monod et al. [5].

This model has successfully described the behav- iour of several regulatory protein and in particular that of phosphofructokinase [6] and and phosphory- lase [7,8]. Moreover, the model of Monod et al. tends to relate the observed cooperativity in ligand bind- ing to the quaternary structure of the protein. This aspect of the control of the catabolic ornithine carbamoyltransferase of Pseudomonus will be the subject of a future paper.

MATERIAL AND METHODS Enzyme Assays

Activity determination for the kinetic studies were performed by measuring the amount of citrul- line formed a t 30 "C in a reaction mixture containing (final concentration) : (150mMimidazole buffer pH 7.8, 10 mM ornithine, carbamoylphosphate, putres- cine and phosphate at the required concentration. The final volume of the reaction mixture was 2.0ml. The rcaction was started by the addition of it carbamoylphosphate solution prepared no more than 1 min before its utilization. Incubation was for b m i n and the rcaction was stopped by adding 2 ml of I M HCl. Measurements were made with such a concentration of enzyme that the amount of citrulline formed never exceeded 0.8 pmol. Citrul- line was measured by the method of Archibald [9].The units of enzyme activity have been defined in a preceding paper [lo].

Preparation of Crude Extract and Enzyme-Purification Procedures

The methods used for growing the cells, prepar- ing the crude extract, and purifying the enzyme were described in a preceding paper [lo].

Reagents Carbamoylphosphate (Sigma) was used without

purification. The purity reported by the manu- facturers exceeded goo/,, as was confirmed by Porter et al. [Ill. L-Ornithine and putrescine were Sigma products. Imidazole and sodium phosphate were of commercial reagent grade.

RESULTS Rate of Reaction and Effect of

the Enzyme Concentration At pH 7.2, under the conditions described (I0 mM

ornithine and carbamoylphosphate), the rate of the reaction catalyzed by ornithine carbamoyltrans-

.. Time (min )

Fig. 1. Activity of the catabolic ornithine carbamoyltransferase with respect to time. The reaction mixture contained 150 mM imidazole buffer pH 7.2, 1OmM ornithine and 10 mM carbamoylphosphate. Activities are expressed as pmol

citrulline formed in the time range studied

I

Enzyme ( pg) Fig. 2. Activity of the catabolic ornithine carbamoyltransferase with respect to increasing levels of protein concentration. The time of incubation is 5 min. The conditions used are the

same as described in Fig. 1

ferase remains linear up to I00 min and the ornithine carbamoyltransferase activity increases linearly with the protein concentration (Fig. 1 and 2). However, a t pH 7.8, the reaction rate does not remain linear within the time range studied (Fig.3). If the enzyme is activated by phosphate or by a high concentra- tion of carbamoylphosphate, the linearity of the reaction rate is restored (Fig.3). Under these condi- tions also, the ornithine carbamoyltransferase activ- ity does not increase linearly with the protein con- centration. Once more, phosphate does restore the linearity (Fig.4). These data show that the phosphate produced by the reaction or by the chemical decom- position of carbamoylphosphate, activates the en- zyme in the course of the reaction.

Fortunately, the binding of phosphate is co- operative (see later) so that, when less than 1 pmol

38 Alloste ric Interactions of Pseudomonas Ornithine Carbamoyltransferase Eur. J. Biochem.

10 20 30 40 50 Time ( rn in)

Fig. 3. Activity of the catabolic ornithine carbamoyltransferme with respect to time at pH 7.8. The reaction mixture contained 150mM imidazole buffer p H 7.8, 10mM ornithine and 30 mM carbamoylphosphate; 0, without phosphate; 0, with 1 O m M phosphate. Activities are expressed as pmol

citrulline formed in the time range studied

73 a,

E c

20 40 60 a0 Enzyme (onits)

Fig. 4. Activity of the catabolic ornithine carbamoyltrans- ferme with respect to increasing levels of protein concentration. The reaction mixture contained 150 mM imidazole buffer p H 7.8, 10 ma1 ornithine and 10 InM carbamoylphosphate, 0, without phosphate; 0, with 10 mM phosphate. Velocities are expressed as ymol citrulline formed in 5 min under the

conditions described

phosphate is formed in the course of the reaction, the activation by phosphate is very weak. Routinely, no more than 10 units of enzyme were used in the course of these experiments and the incubation time never exceeded 5 min in order to avoid an important chemical decomposition of carbamoylphosphate. Under these conditions, the (‘rror committed on a

PH Pig. 5. Activity of the catabolic ornithine carbamoyltransferme with respect to pH. The reaction mixture contained 10mM ornithine, 150mM imidazole buffer a t the p H indicated; 0, 60 mM carbamoylphosphate; 0, 10 mM carbamoyl-

phosphate

series of twice-repeated determinations never ex- ceeded 2°/0.

T h e Effect of p H on the Reaction Velocity As shown in Fig. 5, the pH optimum of ornithine

carbamoyltransferase is largely dependent on the carbamoylphosphate concentration. This suggests that the affinity of the enzyme for carbamoyl- phosphate is highly dependent on pH. Consequently, it was of interest to examine the effect of pH on the rate of ornithine carbamoyltransferase with respect to carbamoylphosphate concentration. The results of such an experiment in which carbamoylphosphate was varied and ornithine held constant a t 20 mM are shown in Fig.6. As the pH increases, the curves become more sigmoidal. At low pH values, the plot becomes more hyperbolic with a corresponding de- crease of the substrate concentration required to reach half-maximum velocity.

The analysis of the kinetic data of an enzyme whose curves of initial reaction rates plotted against effector concentrations are sigmoid shaped, can be accounted for, in first approximation, by the follow- ing relation known as the Hill equation

v is the velocity of the enzymatic reaction a t a given substrate Concentration [S] under the conditions earlier described (see Enzyme Assays). V is the velocity a t saturating substrate concentration. The value of V was calculated by averaging the values obtained from the three usual transformations of the Michaelis-Menten equation (I/ V against l/[S], [S]/w against [S] and w/[S] against v). K , is the Michaelis constant of the enzyme for the substrate. n H is given by the maximum slope of the Hill plot. It is a function of the number of interacting

Vol. 29, No. 1, 1972 v. STALON 39

1 .o

0.8

Y 0.6

0.4

0.2

10 20 30 40 50 [ Carbamoylphosphate] (mM)

Fig.6. The effect of p H on the saturation function of the cata- bolic ornithine carbamoyltransferase. The reaction mixture contained 10 mM ornithine, carbamoylphosphate a t the concentrations indicated and 150 mM imidazole buffer. 0, pH 6.5; 0, pH 6.8; 0, pH 7.2; X, pH 7.3; A , pH 7.6; m, pH 7.8; A, pH 8.0. Velocities are expressed as pmol citrulline formed in_ 5 min under the conditions described.

Y = w/V [see Eqn (l)]

substrate-binding sites [12] of the enzyme molecule and of the strength of the interactions as defined by Rubin and Changeux [13]. When the interactions are strong, nH will approximate to the actual number of binding sites, n. However, if the intersite interactions are weak, nH will be smaller than that number. If there are no interactions, nH will attain a value of 1, even though there can be more than one substrate site.

plots of Fig.6, the curves of Fig.7 are obtained. The nH values of all the curves range from 5.6 to 5.8. Therefore, pH variations within the 6 to 8 range, while markedly influencing the apparent affinity of the enzyme for carbamoylphosphate, do not appear to affect significantly the kinetic conse- quences of the interactions between the carbamoyl- phosphate binding sites on the enzyme. This is known in the casc of hemoglobin as the Bohr effect [5] and means that the interactions are inde- pendent of the absolute affinities for the substrate.

When the Hill treatment is carried out for the ’

Kinetics with Respect to Carbamoylphosphate Initial velocity measurements performed a t

pH 7.8 in imidazole buffer a t constant ornithine concentration (20 mM) and various carbamoylphos- phate concentrations are shown in Fig. 8. Kinetics with repect to carbamoylphosphate are highly co- operative but it can be seen that, when increasing the phosphate concentrations, the rate versus

0 1 2 log [Carbamoylphosphate] ( mM)

Fig.?’. Hill representation of the data given in Fig.6

10 20 30 40 [Carbamoylphosphate] (mM)

Fig.8. Effect of the p b a p b t e eoncentration on the saturation curue of the ornithine carbarnoyltransferase by carbamoylphos- phate. The reaction mixture contained 150 mM imidazole buffer pH 7.8, 10 mM ornithine, carbamoylphosphate as indicated. 6, No phosphate; a, 0.5mM; A, I mM; 0, 1.5mM; 0, 2mM; A, 2.5mM; X, 3mM; f, 4mM; 0, 10mM phosphate. The saturation is expressed as the ratio of the initial velocities to the maximum: initial

velocities [see Eqn (I)]

carbamoylphosphate concentration curves become less and less sigmoidal and the concentration of carbamoylphosphate required to reach the half- maximum velocity decreases.

The cooperative effect of carbamoylphosphate, analyzed in terms of Hill’s interaction coefficient,

40 V. STALON : Bllosteric Interactions of Pseudomonas Ornithine Carbamoyltransferase Eur. J. Biochem.

0 -5

0

L, -0.5

- 1 .o

- I .5

0.5 1 .o 1.5 2 .o log [ Carbamoylphosphate] (mM)

Fig.9. Hill representation of the data given in Eig.8

1 .o 1 f.7-I

2 4 6 8 [Phosphate] ( m M )

Fig. 10. Initial velocilies as a function of phosphate wncentra- tion in the presence of various concentrations of carbamoyl- phosphate. The reaction mixture contained 150 mM imidazole buffer pH 7.8, 10 mM ornithine and phosphate as indicated. 0, 5mM; ., 10mM; A, 15mM; 0, 20mM; Ci, 25mM

carbamoylphosphate

shows that this coefficient drops from 5.8 to 1 as the phosphate concentration increases from 0 to 10 mM (see Fig. 9).

The phosphate saturation curve is also sigmoidal a t low carbamoylphosphate concentration (Fig. 10). This behaviour is therefore ill conformity with the predictions for a K system of the “concerted transi-

tion” theory of Monod et al. [5] where phosphate is an heterotropic ligand which influences both the shape and the threshold concentration of the sub- strate saturation function.

I n order to determine wether the kinetic data for catabolic oriiithine carbamoyltransferase could be resolved by the equation of Monod et al. [5] which describes ligand binding to an allosteric pro- tein, we assume that initial velocities are proportional to the binding function for carbamoylphosphate and so, the binding data are replaced by kinetic data in the equation :

V (1 ) - v a (1 + @-I+ Lca (1 + ca)n- l y = - =

where the binding function P is replaced by v /V. The symbols V and v refer to the maximum initial

velocity and the initial velocity respectively. oc is the normalized concentration of carbamoylphos- phate (a = [carbamoylphosphate]/K~cs), KR(s) being equal to the microscopic dissociation constant of the active state of the enzyme for carbamoyl- phosphate, c is the ratio of the dissociation constant of the active and inactive states of the enzyme for carbamylphosphate. n. is the number of binding sites for this substrate. L, the intrinsic allosteric constant, is the equilibrium constant between the active form R and the inactive form T of the enzyme in the absence of substrate or effector. L‘ is the ratio of T to R states of the protein in the presence of all effectors except the substrate and is related to

(1 + aP + L S + C 4 n

The normalized concentration of the inhibitor, putrescine [I] is designated by the symbol t!? and that of the activator, phosphate [A] by the symbol y, so that B = [~]/KT(I), = [~]/KR(I), = [A]/

The detailed predictions of the model are yet embodied by a second analytical function, the state function 8, which represents the fraction of the enzyme molecules in the R conformation.

KR(A), ey = [A]/gT(A).

If c = 0, this state function can be simplified as

(3)

V’ being the maximum velocity which can be reached in the presence of a given concentration of sub- strate, carbamoylphosphate, when all the enzyme molecules are in the R conformation. It has already been shown that 10 mM phosphate converts almost all the enzyme molecules to the R conforma- tion, Therefore we can consider as V‘ the maximum velocity which can be reached in the presence of a

2 .o

1,5

1 . c

0 . E

A

0 - 10 0 10 20 30 40

[ Carbamoylphosphate] (KIM)

2 .o

1.5

$2 1.0

0.5

0

.lo 0 10 20 30 40 50 [Carbamoylphosphate] (mM)

I , I

- 10 0 10 20 30 40 50 [ C a r bamoylphosphate 1 ( mM )

Pig. 11 . Variation of the quotient function Q with respect to carbamoylphospbte concentration at different phosphate concen- trations (see legend of Fig. 10). The number of binding sites is assumed to be equal to (A) 8, (B) 7 and (C) 9

42 Allostoric Interactions of Pseudonumas Ornitkine Carbamoyltransferase Em. J. Biochem.

given concentration of carbamoylphosphate if phos- phate is 10 mM.

Such a system lends itself to a convenient method of analysis described by Blmgy et al. [6] using a function called "quotient of' states" which is the ratio of the amount of enzyme existing in the R state to that in the T state. We may write

I n the presence of I0 mM phosphate (see Fig. 8) the saturation of the enzyme by carbamoylphos- phate is hyperbolic. The fact that phosphate is able to convert 1000/, of the enzyme molecules into the active state, leads to thi: conclusion that both carbamoylphosphate and phosphate are devoid of any affinity for the inactive T state (c = e = 0) . Thus we may write

L' = L/(l 4- y)" (5 )

Thus, a t constant phosphate concentration, if c (the non-exclusive binding coefficient) is very small, root n of Q should be a linear function of carbamoyl- phosphate concentration extrapolating back to - KR(s) on the abscissa. Fig. 11 A shows that the eighth root of the quotient with respect to carba- moylphosphate concentration is a straight line, and that, when the phosphate concentration is varied, the different straight lines obtained extra- polate a t the same point (KR(s) = I0 mM) on the abscissa.

Taking the seventh or ninth root, thus assuming 7 or 9 sites for carbamoylphosphate on the enzyme molecule, little departure fi om linearity can be observed ; however, the straight lines converge at points that are not located on the abscissa (Fig.llB and I I C). I n this cases, one of the properties pre- dicted for the quotient function is not respected.

Consequently, the most probable number of sites for carbamoylphosphate is eight; the non- exclusive binding coefficient fctr carbamoylphosphate is very small or nil and the graphically determined dissociation constant of the IL state for carbamoyl- phosphate is KR(s) = IOmM. I n addition, we may conclude that the binding sites for the carbamoyl- phosphate and phosphate are independent. Indeed, would they compete for the same site, the quotient function would be

(7)

and parallel straight lines would be obtained when plotting root n of the quotient with respect to the carbamoylphosphate concentration for different con- centrations of phosphate.

Table 1. Variations of L' with the concentration of the positive effector, phosphate

Phosphate concn L'

mM

0 0.5 1.0 1.5 2.0 2.5 3.0 4.0 5.0

10.0

100000 51 700 17600 10000 1850

450 63 9.4 2.3 0

After n and c have been estimated, L' is the only unknown parameter, its value being influenced by the activator concentration. L' may be expressed as a function of a,,, (the normalized substrate con- centration corresponding to half-maximum velocity) because when y is equal to ' I 2 and c is equal to 0, Eqn (1) becomes

(8) Therefore, we can estimate L' by making use

of a single intrapolated point per curve, namely the carbamoylphosphate concentration which corre- sponds to log d/(l - 8) = I (see Fig.9). Now we are able to compute the values of L' for each phos- phate concentration (see Table 1). L, the equili- brium constant between the R and T states in the absence of the allosteric activator, phosphate, is found to be equal t o lo5.

The heterotropic effect exerted by phosphate on the homotropic effect of carbamoylphosphate may be expressed as the variation of nH with L' and ax/, [12]. It can be verified that, taking n equal to 8 and c equal to 0 a close agreement between the theoretical and experimental values of nH as a function of ohll or L' is observed (Fig. 12 and 13).

L' = (qs - 1) (a1/, + 1)"-1.

Kinetics with Respect to Phosphate Now, it is still to be verified that the effect of

the activator, phosphate, can be described in terms of the Monod, Wyman and Changeux model.

Eqn (5) can be rewritten in the following form

- log L' = - log L + n log ( I + 7). (9)

A plot of log L' with respect to the log of phos- phate concentrations gives a function with a maxi- mum slope of n. One can calculate from Fig. 14 that this relation is respected. This means that 8 phos- phate sites are available on the enzyme molecule. On the other hand, we see also from the Eqn (5) and (6) that the intercepts on the ordinates are equal to

8 - V1/L(l + y ) = I/l/L'.

Vol. 29, No. 1, 1972 v. STALON 43

I I

0.1 0.2 0.3 0 .4 0.5 0.6 log a112

Fig.12. Variation of the Hill coefficient nH as a function of the concentration of carbamoylplwsphate corresponding to half-maximum velocity (aqJ. This concentration is expressed by the normalized concentration. all, = [carbamoylphos- phate]/lO mM. Theoretical curves were computed on the basis of the Equation (1) where n = 8 and c = 0 and where the coefficient is defined as by Rubin and Changeux:

d n d In a d a and n = n H if- = O . d In (y/l - P) n =

The curves correspond to thc theoretically computed values. The points are the values obtained experimentally

1 J I

1 2 3 4 5 6 l og L'

Fig. 13. Variation cf the Hill coefficient nH as a function of L'. Theoretical curves were computed as described above under the legend of Fig.12. The experimental points are those

from Fig. 12 where L' = (CXI/~ - 1) ( O L ~ / , + l)?

Consequently, if these values are plotted with respect to phosphate concentrations, they must fit a straight line.

As shown on Fig.15 the curve obtained is not linear but concave up. This can be interpreted as meaning that more than 8 sites on the enzyme molecule are available for phosphate and that phos- phate can interact with the carbamoylphosphate sites (see Discussion). Indeed, when phosphate is varied and carbamoylphosphate held constant, the

0 0.5 log [Phosphate] (mM)

P- I 6 0

Fig. 14. Variation of the equilibrium constant L' with respect to phosphate concentration

0.8

$ O v 6

0.4

0.2

0

/ .o

2 4 [Phosphate] (mM)

Fig. 15. Variation of I$E with respect to phosphate wncen- tratwn. The same data were used as in Fig. 14

eighth root of the quotient function gives, for the weak phosphate concentrations, straight lines which converge on the abscissa a t the same point, 3mM which may be considered as the KR(A) (Fig.16). The intercepts on the ordinates which equal vw+ E ) ~ / L , when plotted with respect t o carba- moylphosphate concentration, fit a straight line which cuts the abscissa a t I0 mM (Fig.17). This value of KR(s) is quite similar to that determined above.

Putrescine Action on the Kinetics with Respect to Phosphate

Now, we shall try to determine how many sites on the enzyme molecule are available for the inhibitor, putrescine. The effect of putrescine was studied in presence of thc activator, phosphate, so that phosphate formed in the course of the reac- tion or by chemical decomposition of carbamoyl- phosphate does not disturb the results.

44 Eur. J. Biochem. Allost eric Interactions of Pseudomoms Ornithine Carbamoyltransferase

- 2 0 2 4 [Phosphate] (mM)

Fig.16. Variation of the quotient function Q with respect to phosphate concentration at distinct carbamoylphosphate con-

centratimzs. The data of Fig. 10 were used

- 10 0 10 20 [Carbamoylphosphate] (mM)

Fig. 17. Variation of the equilibrium constant L with respect to carbamoylphosphate concentration. The data of Fig. 16

were used

Initial velocity measuremmts performed a t con- stant carbamoylphosphate concentration (15 mM) for various phosphate and putrescine concentra- tions are shown in Fig. 18. In this case, the kinetics with respect to phosphate are cooperative and, moreover, putrescine increases the phosphate con- centration required to reach half-maximum velocity. The same method can be used to study phosphate activation in presence of putrescine. Again, Fig. 19 shows that v/( V - y) is a eighth power function of phosphate concentration. If t,he non-exclusive bind-

0 5 10 15 [Phosphate] ( m M )

Fig. 18. Initial velocities as a function of phosphate wncentra- twn at fixed concentration of curbamoylphosphate (15 mM) and various levels of Vtrescine concentration. 0, No putres-

cine; 0, 2 mM putrescine; A , 5 mM putrescine

ing coefficient for the inhibitor is small, this quotient function corresponds to the equation

We may rewrite this equation as follows:

Thus the reverse of the intercept plots on the ordinate is a linear function with respect to putres- cine concentration. Fig. 20 shows that this function extrapolates to a value equal to K T ~ " ) = 2.7 mM.

DISCUSSION The results reported in a previous paper (see the

accompanying paper) have underlined the strong cooperative interactions exhibited by the catabolic ornithine carbamoyltransferase of Pseudomonm fZuo- rescens with respect to carbamoylphosphate. In addition, two effectors, inorganic phosphate and putrescine, have been found to modify the sigmoid saturation curve of the enzyme by carbamoyl- phosphate. This aspect of the control of the activity of the catabolic ornithine carbamoyltransferase has received in this paper a quantitative description by the use of t,he "concerted transition theory" pro- posed by Monod et al. [5] .

The mechanism of action of ornithine (substrate), citrulline (product) and ATP (effector) are now being studied and will be presented in a forthcoming communication. I n addition, the study has been restricted so far to the non-physiological direction of the reaction catalyzed by the enzyme, the carbamoylation of ornithine, which is more accessible

Vol. 29, No. 1,1972 v. STALON 46

1.2

$ 0.8

0.4

-2 0 2 4 6 8 [Phosphate] (rnM)

Fig. 19. Variation of the quotient function Q with respect to phosphate concentration. The same data as Fig. 17 were used

- 2 - 1 0 1 2 3 4 5 [ Putrescine] (mM)

Fig.20. Variation of the equilibrium constant L" with respect to ph-escine concentration, where L" is equal to L / [ ( l + a)8

(1 + Y Y I

to kinetic experimentation. Nevertheless, attempts are now being made to follow the evolution of the opposite direction of the reaction in order to compare the behaviour of the enzyme with respect to both directions of the reaction.

The treatment of the kinetic data reported here according to the model of the concerted transition theory leads to the following deductions which allow the quantitative description of the behaviour of the enzyme.

a) Two different conformational states are within the reach ofthe enzyme ; the transition from one state to the other is characterized by an equilibrium con- stant, L, which is equal to 100000.

b) One of these states, only, has affinity for carbamoylphosphate and inorganic phosphate ;

the microscopic dissociation constants for these ligands are 10 mM and 3 mM, respectively.

c) Contrarily, the opposite conformational state is the only one to present affinity for putrescine, an inhibitor of the enzyme; the dissociation con- stant for putrescine is 2.7 mM.

d) The number of available binding sites for carbamoylphosphate is 8; consequently, it is sug- gested that the enzyme molecule is composed of eight protomers.

This assumption concerning the quaternary structure of the enzyme is based on the fact that the curves representing the eighth root of the quotient function Q with respect t o the substrate concentration a t different phosphate concentrations are linear and extrapolate to the same point on the abscissa. The seventh or the nineth roots of the quotient functions still provide linear function which, however, do not cut the abscissa a t the same point. If values of n differing from these are assumed, the functions deviate from straight lines (hyperbolic curves for values of n higher than 9, parabolic curves for values lower than 7).

When the nth root of the quotient function with regard to the effector concentration follows a para- bolic curve, it can be assumed that the number of effector sites has been underestimated. Such a situation is observed when the 8th roots of the quotient functions are plotted towards the phos- phate concentration. Although we have demonstrated that the sites of phosphate and carbamoylphosphate are distinguishable, the parabolic shape ofthose curves can be explained by assuming that some interference of phosphate a t the carbamoylphosphate sites occurs. Moreover it is known that for all the ornithine

46 V. STALON : Allosteric Interactions of Pseudomonas Ornithine Carbamoyltransferase Eur. J. Biochem.

carbamoyltransferases of various origins which have been studied, phosphate competes with carbamoyl- phosphate [lo, 14-17]. Thus, in the case of catabolic ornithine carbamoyltransferase, Eqn (6) has to be replaced by a more complex relation :

y’ is equal to [ph~spha t~e ] /K‘~(~) , where K’R(A) is the dissociation constant of phosphate with regard to the carbamoylphosphate-binding sites. Deriva- tion of Eqn (13) does indeed lead to a parabolic function of phosphate concentration.

In contrast with most ornithine carbamoyltrans- ferases which have been studied, the catabolic enzyme of Pseudomonas fluorescens, in vivo, only catalyzes the phosphorolysis of citrulline. Because of the strong cooperativity towards carbamoyl- phosphate (the allosteric constant L is 100000) and the reduced affinity of the enzyme for this sub- strate, the reverse reaction, the carbamoylation of ornithine, is littIe favored kinetically in spite of the favourable thermodynamic properties of this direc- tion of the reaction.

The ornithine carbamoyltransferase of the mu- tant IRC204Ml [lo] qualitatively shares all the allosteric properties of the wild-type enzyme but is capable of synthesizing citi*ulline in vivo. The data concerning this enzyme were submitted to a treat- ment similar to that applied to the wild-type enzyme. The enzyme of the mutant was found to have encountered two main modifications : (a) the transition equilibrium constant is reduced ( L = 10000 instead of 100000 for the wild-type enzyme) and (b) the affinity of the active state for carbamoylphosphate is increased by a factor of 40 (0.25 mM instead of 10 mM). Consequently, this catabolic enzyme is able to supply the anabolism of arginine with citrulline but the metabolic products of the catabolic pathway (putrescine and agmatine) still interact with the anabolic function of the enzyme (by increasing the transition equilibrium constant). This explains that putrescine inhibits the growth of mutant IRC204M1 but not that of the wild type which uses the normal anabolic enzyme which is not inhibited by putrescine [lo].

An important conclusion of the present work concerns the quaternary structure of the enzyme: the treatment of the kinetic data leads one to anticipate an octameric structure corresponding to the aggregation of eight identical protomers. Preliminary observations have indeed confirmed the validity of this conclusion of the kinetic analysis

I thank Professor J. &I. Wiame, Dr A. Pi6rard and P. Ramos for helpful discussions and D. Gigot for his excellent technical assistance. This work was supported by grants from the Fonds de la Recherche Fondamentale Collec- tive.

[181*

1.

2.

3.

4. 5.

6.

7.

8.

9. 10.

11.

12.

13.

14.

15.

16.

17.

18.

REFERENCES Stalon, V., Ramos, F., PiBrard, A. & Wiame, J. &I.

Petrack, B., Sullivan, L. & Ratner, S. (1957) Arch.

Jones, M. E. & Lipman, F. (1960) Proc. Nat. A d . Sci.

Reichard, P. (1957) Acta Chem. Scand. 21, 523. Monod. J.. Wvman. J. & Chaneeux. J. P. 11965) J.

(1967) Biochim. Biophys. Acta, 139, 91.

Biochem. Biophp. 69, 186.

U. S. A . 46, 1194.

v - , , Mol.. B b . 12, 88.’

31. 13. Blangy, D., Buc, H. & Monod, J. (1968) J. Mol. Bwl.

BUC’H, (1967) Biochem. Biophys. Res. Commun. 28,

Buc, M. H. & Buc, H. (1967) Abstr. Commun. 4th Meet.

Archibald, R. M., (1944) J. Biol. Chem. 156, 121. Stalon, V., Ramos, I?., PiBrard, A. & Wiame, J. M.,

59.

Fed. Eur. Biochem. SOC. 109.

this journal. Porter, R. W., Mobede, M. 0. & Stark, G. R. (1969) J . . .

Biol: Chem: 244, 1846.

Biol. 28, 483.

265.

W. (1959) J. Biol. Chem. 234, 1452.

chem. Biophgs. Res. Commun. 45,551.

Wyman, J. (1963) Cold Spring Harbor Symp. Quunt.

Rubin, M. M. & Changeux, J. P. (1966) J. Mol. Biol. 21,

Ravel, J. M., Grona, M. L., Humphreys, J. J. & Shive,

Kustin, W. E., Bishop, S. H. & Himoe, A. (1971) Bio-

JoseDh. R. L.. Baldwin. E. & Watts. D. C. 11963) Bio- \ I

ch&. J . 8 i , 409. ’

Joseuh. R. L.. Watts. D. C. & Baldwin. E. 11964) . , C A P . Biochem. Physiol. 11, 119.

Halleux. P.. Stalon. V.. PiBrard, A. & Wiame, J. M. (1969) A&h. Int. Phyaiol. Biochem. 78, 166.

V. Stalon Imtitut de Recherche5 du C.E.R.I.A. Avenue Emile-Gryzon 1, B-1070 Bruxelles, Belgium