Chorismate mutase-prephenate dehydratase from Escherichia coli K-12. II. Kinetic properties.

Abstract The results of kinetic investigations with pure chorismate mutase-prephenate dehydratase from Escherichia coli K-12 were as follows: a pH optimum of 7.3 for the mutase activity, a Km of 0.045 mm for chorismate, and a Km of 1.0 mm for prephenate. l-Phenylalanine inhibited both the mutase and the dehydratase activities. In each case the relationship between activity and concentration of l-phenylalanine was sigmoidal and analysis of the results by the Hill equation gave values of n' = 2.3. d-Phenylalanine and l-tryptophan did not inhibit either activity whilst varying degrees of inhibition were observed with o-, m-, and p-monosubstituted fluoro-, chloro-, and hydroxyphenylalanines. Investigations with other phenylalanine analogues indicated that the α-NH2 group is not essential for inhibition, although it is required for maximal effect. There is an absolute requirement for an unmodified α-COOH group. Thiol compounds such as β-mercaptoethanol and dithioerythritol have little effect on the mutase activity but stimulate the dehydratase activity by approximately 150%.


SUMMARY
The results of kinetic investigations with pure chorismate mutase-prephenate dehydratase from Escherichia coli K-12 were as follows: a pH optimum of 7.3 for the mutase activity, a K, of 0.045 mM for chorismate, and a K, of 1.0 mM for prephenate.
L-Phenylalanine inhibited both the mutase and the dehydratase activities.
In each case the relationship between activity and concentration of L-phenylalanine was sigmoidal and analysis of the results by the Hill equation gave values of n' = 2.3. D-Phenylalanine and L-tryptophan did not inhibit either activity whilst varying degrees of inhibition were observed with o-, m-, and p-monosubstituted fluoro-, chloro-, and hydroxyphenylalanines. Investigations with other phenylalanine analogues indicated that the or-NH2 group is not essential for inhibition, although it is required for maximal effect. There is an absolute requirement for an unmodified a-COOH group. Thiol compounds such as /3-mercaptoethanol and dithioerythritol have little effect on the mutase activity but stimulate the dehydratase activity by approximately 150%.
Iu Escherichia coli the biosynthesis of the aromatic amino acids phenylalanine, tyrosine, aud tryptophan and other aromatic compounds such as folate, vitamin K, and ubiquinone proceeds from the common precursor chorismate (I, 2). Control over the relative amounts of the end products that are synthesized is achieved by regulation of the enzymes that compete for chorismate.
There are three major enzymes involved, chorismate mutase-prephenate dehydratase, chorismate mutase-prephenate dehydrogenase, and anthranilate synthetase which lead, respectively, to the biosynthesis of phenylalanine, tyrosine, and tryptophan.  Table I in  Reference 3) was used for all work described in this report,.
All of the above compounds were found to be chromatographicallg or electrophoretically pure except for L-phenylalanine amide HCl, from which about 5?;, L-phenylalanine was removed by chromatography on Dowex AGI-X2. Solutions of L-phenylalanine ethyl ester and purified L-phenylalanine amide were prepared immediat,ely before use. The I)-phenylalanine contained approximately 2yc of the L isomer (determined by measuring the uptake of O2 with an oscillat'ing 1% electrode when L-amino acid osidase was added to a solution of the compound).
Pure n-phen-4448 ylalanine was obtained by incubating a solution of the impure material with L-amino acid oxidase for 2 hours at 37". Chorismde Mufase Assay-Chorismate mutase activity was assayed in 0%ml reaction mixtures containing chorismate, 100 mM Tris-Cl, pH 7.8,0.5 mu EDTA, 0.01% bovine serum albumin, 20 mM mercaptoethanol, and enzyme. After incubation for 5 min at 37", the enzymic reactions were terminated by the addition of 0.1 ml of 4.5 M HCI. After a further incubation at 37" for 10 min to convert prephenate into phenylpyruvate,, 0.1 ml of 12 M NaOH was added and the absorbance at 320 nm was measured to determine phenylpyruvate. Blanks, to which enzyme was added after the addition of NaOH, were included since the substrate contributes a significant absorbance. A unit of enzyme was defined as the quantity of enzyme that catalyzed the conversion of 1.0 pmole of chorismate to prephenate in 1 min under the assay conditions. Prephenate Dehydratase Assay-Prephenate dehydratase activity was assayed in 0.4-ml reaction mixtures containing barium prephenate, 20 IllM Tris-Cl (pH 8.2), 0.5 mM EDTA, 0.01% bovine serum albumin, 20 mM mercaptoethanol, and enzyme. After 5 min incubation at 37" the reaction was terminated by the addition of 0.8 ml of 1.0 M NaOH and the absorbance at 320 nm was measured. Blanks were included as for the mutase assay. A unit of enzyme was defined as the quantity of enzyme that catalyzed the conversion of 1.0 pmole of prephenate to phenylpyruvate in 1 min under the assay conditions.

RESULTS
Qj'ects of Bovine Serum Albumin and Thiol Compounds on Enzyme Activity-Bovine serum albumin activates the purified enzyme, having a greater effect on the mutase activity than on the dehydratase activity (Table I). Assays examining the effect of bovine serum albumin were performed in the presence of 20 mM mercaptoethanol, and assays examining the effect of mercaptoethanol were psrformed in the presence of 100 rg per ml of bovine serum albumin. Enzyme activities are expressed as a percentage of the activity obtained in the absence of the compound under investigation. Thiol compounds stimulate the dehydratase activity of the purified enzyme an additional 150% but have little effect on the mutase activity (Table I).
Effect of pH on Enzyme Activity-The mutase activity shows a bell shaped curve over the pH range from 5.5 to 9.5 with a well defined maximum at pH 7.3. The pH dependence of the dehydratase activity possesses some unusual features that we cannot explain at present. In some experiments a single maximum of activity was observed at pH 6.5, whereas on other occasions a bi-lobed curve was found with maxima at pH 6.5 and 8.5. This difference in pH activity relationship could not be attributed to specific buffer ions, since it was observed with a variety of different buffers. The only variable between experiments appeared to be the age of the enzyme preparation. However, until the phenomenon is examined in more detail we prefer not to comment about its possible causes, apart from noting the related observation (see below) that aging affects the extent and type of inhibition by phenylalanine.
Little effect of pH on the value of K, for prephenate was noted over the range from pH 6.5 to pH 9. or that derived from steady state theory for the system E + S e ES ---f E + 2'. The values obtained for K, were 0.045 mu for chorismate and 1.0 mM for prephenate. Hill plots of these data were linear over the concentration ranges explored and had slopes of 0.9 and 1.1 for mutase and dehydratase, respectively. These results demonstrate the absence of homotropic co-operative effects for either of the substrates.
Inhibition by z-Phenylalctnine-Both activities are inhibited by L-phenylalanine, although the dehydratase activity is generally more sensitive than the mutase.
Plots of activity versus inhibitor concentration are sigmoidal for both activities (Figs. 3  and 4).
A Hill plot of the inhibition of mutase activity by L-phenylalanine is linear between inhibitor concentrations of 0.02 mM and 0.40 mM, with a slope of 2.3 (Fig. 4, inset). Similarly for the inhibition of the dehydratase activity by L-phenylalanine, the Hill plot has a linear portion with a slope of 2.3 between inhibitor concentrations of 0.025 mM and 0.10 mM (Fig. 3, inset). Thus the effector, n-phenylalanine, exhibits co-operative homotropic interactions in its inhibition of both activit,ies. The inhibition of both mutase and dehydratase activities by L-phenylalauine showed little dependence on pH over the pH range of 6 to 9. For assay details see "Experimental Procedure." Chorismate concentration, 1.0 mna; 0.45 pg of pure enzyme was used in each assay.

4449
The effect of aging of the enzyme on its sensitivity to L-phenylalanine is quite marked.
Thus one batch of enzyme stored in dilute solution at 4" for 2 weeks exhibited hyperbolic inhibition kinetics for the dehydratase activity and a Hill plot with slope of 1.0 (Fig. 3), although the maximal percentage of inhibition was unaffected.
Another sample, stored in dilute solution at -20" for 3 weeks lost all sensitivity of mutase activity to inhibition (at 0.3 mM n-phenylalanine), and the dehydratase activity was inhibited only 65% by 1 mM L-phenylalanine.
The extent of this densensitization towards L-phenylalanine is quite variable and has not, as yet, been linked to a definite manipulation of the enzyme preparation apart from aging. We also observed that desensitization occurs during the course of a 30-min enzyme assay. This makes it desirable that all assays investigating the kinetics of inhibition by L-phenylalanine measure initial rates. In the case of the experimental results presented above, a close approsimation to this has been obtained by using freshly prepared enzyme in 5-min assays (except for the effect of pH on the inhibit,ion when longer assay times were used).
E$ect of L-Phenylalanine on Substrate Saturaiion Curve for Dehyclratase Activity---The data in Fig. 1A show that the plot of v versus s for dehydratase activity, which is hyperbolic in the absence of phenylalanine becomes sigmoidal in its presence. The  Curve for Mufase Activity-In marked contrast to the effects observed with the dehydratase activity, phenylal:uCne does not alter the hyperbolic nature of the kinetics observed for the mutase activity ( Fig. 2A). There is no indication of sigmoidality, even at 0.60 mM phenyl-&nine, where inhibition is masimal (see Fig. 4), and Hill plots are all linear with slopes of 0.8 to 1.0. ljouble reciprocal plots are linear and intercept on the ordinate, indicating a competitive type of inhibition, but plots of I/v versus inhibitor concentration we not linear (Fig. 2B) and are of a type associated with partially competitive inhibition (7). Inhibition of Dehydratase Alctivity by Analogues of Phenyl-ala&e-The results shoIvn in Table II indicate that out of a wide range of phenylalanine nnalogues tested, only those with a substitution on the aromatic ring or the a-NH2 group show any appreciable inhibition.
For ring substitution the extent of inhibition generally decreases as the size of the substituent increases (o-and m-hydrosyphenylelanine are exceptions here, possiblp because of hydrogen bonding between the hydroxyl group of the inhibitor and an appropriate group on the enzyme).
Meta substituents retain the largest inhibitory effect and ortho the least. There is ho\Tever need for some caution in assuming that these latter effects also apply to chorismate mutase-prephenate dehydratase from wild type E. coli. The phe R gene2 in strain Jr492 was derived directly from a strain (JPl71) which had been isolated on the basis of resistance to the growth-inhibitory effect of o~fluorophenylalanine, after mutagenesis with N-methyl-n"nitro-N-nitrosoguanidine (9). The prephenate dehydratase activity in strain JP171 showed a decreased sensit,ivity to illhibition by o-fluorophenylalnnine when compared with the wild type enzyme (9), although the inhibition reported by Im and Pittard (107; at 5 mM inhibitor) is much less than that reported in Table   2 P,he A specifies the structural gene for chorismate mutaseprepklenate dehydratase (8). Table II are only valid for the wild type organism if the enzyme used in this investigation is urlchanged from the wild type enzyme; at present we have no positive evidence concerning this point.

II. Thus the results presented in
Kinetics of Inhibition by o-; m-, and p-Pluorophenylalanines-The inhibitor studies described above assume that the analogues used bind to the phenylalanine-binding (regulator)-) site of the enzyme.
Since inhibition could also be caused by competitive binding at the prephenate-binding (a,ctive) site, the effects of o-, m-, and p-fluorophenylalanines were examined in more detail. With each inhibitor the relationship between concentration and activity was found to be sigmoidal (Fig. 5) in a manner analogous to that observed with phenylalanine.
Hill plots of these data had slopes of similar value to those obtained with phenylalanine. Therefore it is reasonable to conclude that these inhibitors are functioning through the regulatory site of the enzyme.

EJect oj Thiol Compounds-It
is tempting to implicate a thiol group in the catalysis of the dehydratase activity since maximal dehydratase activity in pure enzyme preparations is dependent upon the addition of thiols such as Pmmercaptoethanol or dithioerythritol.
Schmit et al. (10,11) have demonstrated that the dehydratase activity of partially purified cxhorismate mutaseprephenate dehydratase from X. fyphimurium is inactivated by p-chloromercuribenzoate or bromopyruvate, and recent studies in this laboratory have indicated that the dehydratase activity of the B. coli enzyme is selectively inhibited by 1.0 m&I N-ethylmaleimide or 5,5'-dithiobis-(2.nitrobenzoic acid) .3 Further clarification of this aspect awaits sequence studies.

Kinetic
Properties-The kinetics of both activities is consistent, with the model of Michaelis and Menten, and it is of interest to compare the value of K, for chorismate of 0.045 mM determined in this investigation with those of 0.39 mM for chorismate mlltase-prephenate dehydrogenase (5) and 0.0012 mM for anthranilate synthetase (12). These three enzymes compete for chorismate, and the comparison indicates that, when chorismate is present at low concentrations in vivo most of it could be directed towards the biosynthesis of tryptophan.
This may esplain the earlier observation of Davis (13) that strains of E. coli containing leaky mutations in the common pathway may show a preferential requirement for phenylalanine and tyrosine before tryptophan.
The K,, for prephenate of 1 .O m&f is somewhat higher than that of 0.37 rnhf found for chorismate mutase-prephenate dehydrogenase by Koch et al. (5).
Inhibition by L-Phenylalanine--I-Phenylalaniue, which is the end product of the pathway, acts as a feedback inhibitor of the enzyme.
At a fixed substrate concentration the inhibition of both activities increases in a sigmoidal manner as the concentration of phenylalanine is increased and Hill plots have slopes in excess of two, i.e. a homotropic co-operative effect is observed. Inhibit,ion is mainly directed towards the dehydratase activity, a!though some inhibition of the mutase activity also occurs. In contrast, inhibition of the analogous enzyme chorismate mutaseprephenate dehydrogenase by the end product of its pathway, tyrosine, is directed solely towards the dehydrogenase activity, and there is no inhibition of the mutnse activity (5).
The kinetics in general suggests that L-phenylalanine binds at a site on the enzyme separate from the active site. This conclusion is reinforced by the observation that on aging the enzyme becomes desensitized to phenyldanine. AL similar conclusion has been reached after studies with the enzyme from X. typhimurium (II).
Effect of L-Phenylalanine on Substrate Xaturation Curves of Chorismate Mutase-Prephenate Dehydratase-L-Phenylalanine induces homotropic co-operative effects in the enzyme with respect to prephenate.
That is, in the presence of fixed concentrations of L-phenylalanine the substrate saturation curves of the dehydratase activity become sigmoidal.
In this respect the behavior of the enzyme is similar to that of the same enzyme from X. typhimurium (14). This behavior could be explained by either the model of Monad et al. (15) or that of Koshland (16), although qualitative examination of the data does not discriminate against either theory.
Since there is experimental evidence that the enzyme is capable of self-interaction (3), the possibility that a theory involving polymerization (17) could provide the correct explanation of the sigmoidal kinetics should not be excluded.
In contrast to its effect on the dehydratase activity, L-phenylalanine does not induce co-operative effects for chorismate in the mutase activity.
In this respect the enzyme from B. coli is different from the enzyme from S. typhimurium, which was found by Schmit and Zalkin (14) to give sigmoidal substrate saturation curves for the mutase activity in the presence of L-phenylalanine. The kinetics observed by us for the inhibition of the E. coli enzyme is of a partial competitive nature (7). The physical interpretation of this type of inhibition is that L-phenylalanine, in binding to a regulatory site on the enzyme, decreases the affinity of the enzyme for the substrate but has no effect on the rate of conversion of the enzyme-substrate complex to products. If the sole function of the enzyme in the catalysis of the mutase reaction is to bind chorismate exclusively in the conformation necessary for a spontaneous S-l' rearrangement to prephenate (ES), then the absence of any effect of an inhibit.or on the rate of breakdown of the enzyme-substrate complex is to be espected. Nature of Phenylalanine Binding Site-The pattern of the inhibitions obtained with a wide range of phenylalanine analogues enables a number of conclusions to be drawn.
These conclusions are valid for the enzyme from strain JP492, but whether they can be extrapolated to the wild type enzyme remains to be investigated.
1. The L-isomer only is active. 2. The ring is required for any significant degree of inhibition.
3. There is a restriction on the size of ring subst,ituents that can be accommodated, the degree of restriction depending on the position of the ring substitution.
4. There is a restriction on the size of substituents that can be accommodated on the p and QI carbon atoms since /3-phenylserine and a-methylphenylalanine do not inhibit. The observation that phenylpropiolic acid inhibits as much as 3-phenylpropionic acid suggests that the a! and fl hydrogen atoms do not have an active role in the inhibition. 5 its failure to bind satisfactorily to the regulatory site or to its inability, once bound, to induce an appropriate conformational change. In the absence of binding data it is impossible to decide which of these alternatives provides the better explanation for any of the above effects.
Comparison of Chorismate Mutase-Prephenate Dehydratases from E. coli and S. typhimurium--A comparison of the results reported in this and the accompanying paper (3) with those obtained by Schmit and Zalkin (14,19) using partially purified preparations of chorismate mutase-prephenate dehydratase from S. typhimurium is given in Table III. Comments have been made on a number of these comparisons in the preceding text and at this stage it suffices to note that, as may be expected when comparing the same enzyme from closely related genera, there are many similarities aud few differences in properties.
In addition to the tabulated similarities there is an indication that the polymerization phenomenon observed with the S. typhimurium enzyme (19) is also a property of the E. coli enzyme (3). The details of this polymerization and its possible relationship to the inhibition by phenylalanine must await further investigations with the pure enzyme.