Eur. J. Biochem. 86, 165-174 (1978)

Chorismate Mutase/Prephenate Dehydratase from Escherichia coli K12 Binding Studies with the Allosteric Effector Phenylalanine Mary-Jane GETHING and Barrie E. DAVIDSON The Russell Grimwade School of Biochemistry, University of Melbournc (Received May S/November 21, 1977)

The binding of phenylalanine to the allosteric site of chorismate mutase/prephenate dehydratase has been studied by steady-state dialysis. Under most of the experimental conditions examined positive co-operativity was observed for the binding of ligand up to 50% saturation and negative co-operativity above 50% saturation. In the presence of 0.4 M NaCl at pH 8.2 the co-operativity was positive at all phenylalanine concentrations and the maximal stoichiometry of 2 mol of phenylalanine/mol of enzyme subunit was observed. It was concluded that there is a single phenylalanine-binding site per subunit which is associated with the regulation of each of the mutase and dehydratase activities. The effects of enzyme concentration, NaCl, temperature and pH on the binding of phenylalanine have been investigated. Neither tyrosine nor tryptophan bound to the allosteric site of the enzyme, Enzyme that was desensitized to inhibition by phenylalanine following modification of three sulphydry1 groups with 5,5'-dithio-bis (2-nitrobenzoic acid) did not bind phenylalanine. The mechanism of co-operativity, the binding of the enzyme to Sepharosyl-phenylalanine and the physiological significance of the inhibition of the enzyme by phenylalanine are discussed in terms of the results obtained.

In Escherichia coli the biosynthesis of the aromatic amino acids phenylalanine, tyrosine and tryptophan, and the other aromatic compounds such as folate, vitamin K and ubiquinone proceeds from the common precursor chorismate [1,2]. The first two steps in the pathway for the synthesis of phenylalanine are catalysed by chorismate mutaselprephenate dehydratase, an oligomeric enzyme composed of identical subunits of molecular weight 40000 [3,4]. The mutase and dehydratase activities of this enzyme are catalysed at separate, or only slightly overlapping, active sites [5,6], Both activities are subject to feedback inhibition by phenylalanine and exhibit a sigmoidal response to increased phenylalanine concentration [7,8] suggesting that the enzyme is a site of control of phenylalanine biosynthesis. The presence of NaCl or a decrease in pH cause an increase in the sensitivity of the enzyme activities to this inhibition. The binding of phenylalanine to the enzyme occurs at an allosteric site, topographically separate to the Abbreviation. 5,5 '-Dithio-bis(2-nitrobenzoic acid), Nbsz. Enzymes. Chorismate mutase/prephenate dehydratase (EC 5.4.99.5/4.2.1.51).

active sites [6,9], and causes a conformational change that is characterized by the movement of a tryptophan residue into a more hydrophobic region of the protein [lo].

In this paper we report the results of investigations undertaken to determine the stoichiometry and characteristics of the binding of phenylalanine to the enzyme. These results enable a number of conclusions to be drawn concerning the nature of the homotropic effects associated with the binding. MATERIALS AND METHODS Materials

Pure chorismate mutase/prephenate dehydratase was prepared by the method of Gething et at. [S] and all experiments were done using freshly purified enzyme which was maximally inhibited by phenylalanine and exhibited full co-operativity for this inhibition [lo]. The concentrations of enzyme have been expressed throughout as subunit concentrations. ~-[U-14C]Phenylalanine of specific activity 360 Ci/mol was obtained from the International Chemical

166

Binding of Phenylalanine to Chorismate Mutase/Prephenate Dehydratase

and Nuclear Corporation. Radiochemical purity was checked by autoradiography after paper electrophoresis at pH 2.1 [Ill. L-Phenylalanine was obtained from the Sigma Chemical Co. Ltd. All other reagents were Analar grade. Madsen's scintillation liquid was made up as follows: two parts of toluene scintillator, i.e. 0.5% wjv 2,5-diphenyl-oxazole and 0.04 % wjv 1,4-bis-2-(5phenyloxazolylbenzene in toluene to one part of Triton X-100 [12].

experiment (after about 160 fractions the corrections required to allow for diffusion of ligand across the membrane become undesirably large), two separate experiments were carried out in sequence, with identical samples of enzyme. One experiment was done with the phenylalanine additions predominately in the concentration range 0-32 pM and the other with the additions predominately in the range 25 - 170 pM. The addition of excess phenylalanine to 2 mM was the same in each case. The results were combined after their final calculation.

Binding Experiments

Determination of14C Contents

Because the allosteric inhibition of chorismate mutaselprephenate dehydratase is subject to aging effects [7] the method of Colowick and Womack [I31 was used to determine the binding of phenylalanine to the enzyme. This technique enables a complete set of binding data over a wide range of phenylalanine concentrations to be obtained in 30 min with a single protein sample. Dialysis cells were manufactured from perspex as described by Colowick and Womack [I31; the volume of each of the two chambers was 1.6 ml and the diffusion area of the cell was 2.0 cm2. A constant temperature (+_0.2 "C) was obtained by immersing the dialysis cell in a thermostatically controlled water bath. The dialysis membrane was cut from Visking dialysis tubing and was freed of impurities before use by heating at 60 "C in 3 % acetic acid for 1 h, and then rinsing with several changes of glass-distilled water. A dialysis cell containing 0.80 ml of enzyme (11 - 240 pM) in the upper chamber was equilibrated for 5 min at the appropriate temperature while buffer at the same temperature was pumped through the lower chamber at a constant rate of 8 ml/min (for experiments at 5 "C) or 16 ml/min (for experiments at 25 "C and 37 "C). A sample (5 or 10 pl) of ~-[14C]phenylalanine was added to the upper chamber and fractions of the effluent slightly in excess of 1 ml were collected at fixed time intervals. At the fifteenth fraction 1 - 5 pl of unlabelled L-phenylalanine solution (0.8 mM) was added to the upper chamber and a maximum of 10 further additions of ligand solution (0.8,1.6 or 8.0 mM) were made at intervals of approximately 12 fractions. Finally a large excess of unlabelled L-phenylalanine (20 pl of 100 mM) was added and 20 more fractions of effluent were collected. Immediately after the experiment samples of solution in the upper chamber were removed for scintillation counting and determination of protein concentration. In those experiments investigating the effect of temperature on binding it was necessary to obtain more data points over a wider range of added phenylalanine concentrations than the above procedure yielded. In order to do this without lengthening the

The 14C content of the effluent fractions was determined by counting samples (1.0 ml) with 10 ml of Madsen's scintillator solution, using a Packard liquid scintillation spectrometer. The efficiency of counting was determined using the channels ratio method [14] with [14C]benzoic acid as a reference standard. Determination of Protein Concentrations

Protein concentrations were determined as described previously [lo]. Modijication of Enzyme with 5,5'-Dithiobis(2-nitrobenzoic acid)

Enzyme was reacted with Nbs, as described previously (Fig. 1 of [5]). Excess reagent was removed by gel filtration with Sephadex G-25 at 4 "C using 20 mM Tris-C1, pH 8.2, containing 1 mM EDTA as the elution buffer. Processing of Binding Data

The various binding parameters were calculated from the experimental data using a computer program written for this purpose (a listing is available from the authors). The calculation makes appropriate corrections for losses of radioactivity from the upper chamber during the course of the experiment (maximum of 5- 10% of total added radioactivity). The computed parameters are listed and plotted as the binding functions presented below. RESULTS The Binding of Phenylalanine to Chorismate MutasejPrephenate Dehydratase at p H 8.2

The result of a typical binding experiment measuring the binding of phenylalanine to the enzyme (44 pM) at pH 8.2 and 37 "C (Fig. 1) illustrates the attainment of steady-state diffusion of [14C]phenylalanine across the membrane following each incremental addition of

M.-J. Gething and B. E. Davidson

0

161

R

I

I

I

I

I

I

I

60 80 100 120 140 1 0 Fraction number Fig. 1 . The binding ofphenyldanine to chorismcrte mutnseiprephentite ddiydrutnse measured by steudy-state diulysis. Data was collected as described in Methods for enzyme at 37 ”C in 20 mM Tris-CI, pH 8.2, containing 1 mM EDTA and 20 mM 2-mercaptoethanol. Every second data point has been plotted. The arrows indicate the time of addition of the non-radioactive ligand and the figures indicate the total added phenylalanine concentration (pM) after that addition. Initial concentrations were: ( 0 )enzyme 11 pM, ~-[‘4C]phenylalanine2.85 pM (9.3 x 105 dis. x min-1); (0) enzyme 44 pM, ~-[“T]phenylalanine 5.7 pM (1.86 x 106 dis. x min-1); (n)enzyme 240 pM, ~-[14C]phenylalanine5.7 pM ( 1 . 8 6 ~ 3 0 6dis. xmin-1) 20

40

r,, (mol Phe/ mol of enzyme subunit)

log (

[Phel‘,,. /MI

Fig. 2. Saturution curves for the binding of‘phenylalunine t o chorisninte niutcrse/preplienute dehydrutuse nt p H 8.2 and 37 “C. The data obtained from the experiments described in Fig. 1 have been plotted (see Methods) as (A) binding curves, (B) Scatchard plots, (C) log binding curves, (D) Hill plots, (E) double-reciprocal plot. The enzyme concentration was: (A) 11 FM; (0) 44 pM; (0) 240pM. Identical results were obtained with three different enzyme preparations

non-radioactive phenylalanine. The addition of the first three increments of non-radioactive phenylalanine to the upper chamber caused almost no change in the concentration of free phenylalanine in that chamber (measured by the 1% content of the effluent). This suggests that positive co-operative phenomena are

involved in the binding process, since one would otherwise expect the addition of ligand to increase the concentration of free ligand (see, for example, [13]). The binding curve for this data (Fig. 2A) appears to be slightly sigmoidal but does not provide convincing evidence for co-operativity. However, the

168

Binding of Phenylalanine to Chorismate Mutase/Prephenate Dehydratase

downward curvature of the Scatchard plot (Fig. 2B) and the upward curvature of the double-reciprocal plot (Fig. 2E) indicate quite clearly that the binding of phenylalanine to the enzyme exhibits positive cooperativity under these experimental conditions. The curved nature of these plots prevents any graphical determination of the association constants of the phenylalanine-binding sites. Extrapolation of the curves to obtain the maximum number of moles of phenylalanine bound per mole of enzyme subunit (n,) is hazardous for the same reason. The value obtained for n, for this and other experiments carried out under identical conditions was between 0.6 and 0.65. On the basis of other experimental data shown below (Fig. 4) and arguments presented in the Discussion, a Hill plot of the binding data (Fig.2D) was constructed using a value of 1.0 (not 0.6-0.65) for n,. The resulting plot has two linear portions with a break at approximately 50% saturation. The nHvalues are 1.4 below 50% saturation and 0.3 above 50% saturation. This result suggests that under these experimental conditions there are moderately strong positive co-operative interactions between binding sites below 50% saturation and negative co-operative interactions above 50 % saturation. EfSect of Enzyme Concentration

on the Binding of Phenylatanine Markedly different results were obtained when binding experiments were done at lower (11 pM) or higher (240 pM) enzyme concentrations than that used in the experiments described above (44 pM) (Fig. 1). For technical reasons it was not possible to obtain accurate data at the lowest enzyme concentration when the concentration of free phenylalanine was saturating (i.e. greater than 20 pM) because only a small proportion (20 - 30%) of the total phenylalanine was bound. Thus increments in the measured radioactivity over this range of ligand concentrations were too small for meaningful results (Fig. 1). The binding functions (Fig.2) indicate that the dependence of the binding on enzyme concentration is quite complex. At low enzyme concentrations the enzyme bound ligand more effectively in the lower concentration range of phenylalanine, while the available data suggests that the higher enzyme concentration was a more effective acceptor at high concentrations of phenylalanine (Fig. 2A). Non-linearity in the Scatchard plots (Fig. 2B) indicates that the binding was co-operative in all cases, and the steeper slope of the log binding curve (Fig. 2C) at the lowest enzyme concentration suggests greater co-operativity in that case. The value of n, estimated from the Scatchard plot (Fig. 2B) increased from 0.6 at 44 pM enzyme to 0.7 at 240 pM enzyme.

The Hill plots are similar at the two higher enzyme concentrations (Fig.2D) with a break in slope at 50% saturation and nH values of 1.4 and 0.1 -0.3 below and above this saturation level. This suggests that, although the affinity of the enzyme for phenylalanine is different at these two enzyme concentrations, the type and strength of the co-operative interactions between the subunits are generally similar. The change from positive to negative co-operativity, indicated by the Hill plot, is clearly illustrated for the highest enzyme concentration in Fig. 1 by the sharp increase in the free phenylalanine concentration when the total phenylalanine concentration was increased from 206 to 306 pM. At the lowest enzyme concentration the interaction between the binding sites was increased since the Hill coefficient, nH, is 2.1 below 50% saturation (Fig. 2D). There is not sufficient data available above 50% saturation at this enzyme concentration to draw any conclusions other than that there appears to be a break in the slope of the Hill plot at 40- 50% saturation. Efeect of Temperature on the Binding of Phenytulunine

The effect of temperature on the binding of phenylalanine to the enzyme is comparatively simple (Fig. 3). Binding was favoured by a decrease in temperature, particularly at lower phenylalanine concentrations (Fig. 3A and B), but remained co-operative in all cases (Fig. 3B). The value of n, increased slightly with the decrease in temperature, although the maximum value at 5 "C, 0.72 mol of ligand bound per mol of enzyme subunit, is still much less than the expected value of 1.o. A Hill plot with a break in slope at 50% saturation was obtained at each temperature (Fig. 4D). The Hill coefficients of 1.4 and 0.3 are independent of temperature, and this, together with parallel slopes of the log binding plots indicates that although a change in temperature alters the affinity of the enzyme for phenylalanine it has no effect on the co-operativity of the binding. EfSect of NuCl on the Binding of Phenylulunine

The effect of NaCl on the binding of phenylalanine to the enzyme depends markedly on the concentration of NaCl used. The addition of 0.1 M NaCl increased the binding affinity of the enzyme over the full range of phenylalanine concentrations (Fig. 4A and B), so that the value obtained for n, by extrapolation of the Scatchard plot is 0.75 in 0.1 M NaCl and 0.62 in its absence. However, the parallelnature ofthe appropriate Hill plots (Fig. 4D) and log binding curves (Fig. 4C) indicates that 0.1 M NaCl does not alter the extent of co-operativity in the binding process.

M.-.I. Gething and B. E. Davidson

16Y

r,(mol P h e / m o l of enzyme s u b u n i t )

Fig, 3, Effect qf' teniperuterre on the binding of phcwylaloninu 10 chovismatu rnutns~~~prephennte dehydrutose. The binding data mas obtaincd 25 " C , (A)37 ' C. (A) from steady-state dialysis experiments done as described in both the Melhods and the legend to Fig. 1, at (0)5 'C, (0) Binding curves; (B) Scatchard plots; (C) log binding curves; (D) Hill plots. The enzyme concentration was 56 FM

By contrast, the effect of 0.4 M NaCl on binding is more complex. At high concentrations of ligand it caused a further marked increase in binding affinity (Fig. 4A), yielding a value of 1 .O mol of phenylalanine per mol of enzyme subunit for n, (Fig. 4B). These were the only experimental conditions found to give this value. At low concentrations of ligand, 0.4 M NaCl had the opposite effect to 0.1 M NaCl since it decreased the affinity of the enzyme for phenylalanine (Fig.4A). This situation is difficult to interpret and suggests that NaCl has more than one effect on the binding of phenylalanine to the enzyme. One aspect of the effect of 0.4 M NaCl is the abolition of the negative cooperativity hitherto observed above 50 % saturation. The Hill plot at this salt concentration exhibits no change in slope so that nH remains at 3.4 throughout the range of ligand concentrations investigated (Fig. 4D).

enzyme to Sepharosyl-phenylalanine [8], the effect of pH on the binding of phenylalanine to the enzyme was investigated. Attempts to obtain binding data at pH 6.1 for comparison with some of the kinetic results reported previously [8] were unsatisfactory since the protein precipitated from solution at the higher enzyme concentration required for binding experiments. A decrease in pH from 8.2 to 7.4 led to an increase in the affinity of the enzyme for ligand over the complete range of phenylalanine concentrations (Fig. 5A) resulting in an extrapolated value of 0.78 for n, (Fig. 5B). The extent of co-operativity between the binding sites was not affected (Fig.5D). The same effect of pH was also observed when phosphate was the buffer ion.

Efect of Change in p H on Binding of' Phenylalanine

The nature and topography of the phenylalaninebinding site was investigated by Dopheide et al. [7] using kinetic measurements with a range of phenylalanine analogues. The conclusions reached were

Because a change in pH aff'ects both the inhibition of the enzyme by phenylalanine and the binding of the

The Binding of Analogues of Phenylalanine to the Enzyme

170 10

38 06 E

OL

02

0

=

05

LE I

t

0

5 L ! I

-05 ,"

-

-1 0 3

r,,, (mol P h e l m o l enzyme s u b u n i t )

log ( [Phelfree/M)

Fig. 4. Effect ofNuC1 on the binding of phenylalunine to chorisrncrte mutciseiprephencite dehydratose. The binding data was obtained from steadyno NaC1, (A) 0.1 M NaCI, state dialysis experiments similar to those described in Methods and the legend to Fig. 1 using buffer containing (0) (0) 0.4 M NaCI. (A) Binding curves; (B) Scatchard plots; (C) log binding curves; (D) Hill plots. The enzyme concentration was 45 pM and the temperature was 37 "C

-- 0 5 08

20

40

[PheIfree (uM) 60 80

100

0

0.2

04

06

08

I5

5 Ln 3

[ 06 N

c W

0

"0 O L

-

E

\

2

--. L t

02

-0 (L

c

E

5

- 0 LE

-- 80 z

-E 60

0 05

- 40 ?

0: E

C

--. .0.5 h

\

f

-

I

I

W

I

20

-0

(51

-1 0

0

02

04

06

08

10

r, (mol Phe/mol of enzyme subunit)

-60

-55

-50 -L5 -LO l o g ( [Phelfree/ M I

I

Fig. 5. E j e c t of decreusing p H on the binding qf'phenylulnnine to chorismnte mutuselprephenute dehydrutuse. The binding data was obtained from steady-state dialysis experiments similar to those described in Methods and the legend to Fig. 1. The experiments were carried out at 37 "C in the following buffers: (0) pH 8.2, 20 mM Tris-C1 containing 1 mM EDTA and 20 mM 2-mercdptoethanol; (A) p H 7.4 in the same buffer; (0) p H 7.4 in 10 m M potassium phosphate buffer containing 1 m M EDTA and 20 m M 2-mercaptoethanol. (A) Binding curves; (B) Scatchard plots; (C) double-reciprocal plots; (D) Hill plots. The enzyme concentration was 49 pM

M.-J. Gething and B. E. Davidson: Binding of Phenylalanine to Chorismate Mutase/Prephenate Dehydratase

f2 0 0 0 t E

f __________

{

c C

/

, _____--I

t

,ooo~-!pM

a n a, l o g u e IrnM,

t

2 rnM phenylalanine

aJ I

171

(which is decreased by 30%) was rendered insensitive to the inhibitory effect of phenylalanine [5,15]. Since it was not clear whether the modification prevented the binding of phenylalanine, or whether it prevented phenylalanine-induced conformational changes, a sample of enzyme (40 pM) was used for binding studies following reaction with Nbsz to modify three sulphydryl groups. There was no detectable binding of phenylalanine. DISCUSSION Stoichiometry of Binding

Fraction n u m b e r Fig. 6 . The binding of' unulogues of phenylulunine to chorismute mutuse/prephenate dehydratase. The enzyme (44 pM) was in 20 mM Tris-C1, p H 8.2, containing 1 mM EDTA and 20 mM 2-mercaptoethanol at 25 "C. ~-[14C]Phenylalanine(9.3 x 105 dis. x min-') was added at the first fraction to give a concentration of 2.85 pM. The phenylalanine analogue was added as indicated by the arrows to give the concentrations shown. Finally, an excess (2 mM) of unlabelled phenylalanine was added. (-- ) m-Fluorophenylalanine; ( -) tyrosine or tryptophan

based on the assumption that the degree of inhibition of the enzyme by an analogue was a measure of the extent of binding of the analogue to the allosteric site. In order to check the validity of this assumption the steady-state dialysis technique was used to test for binding of the amino acids tyrosine and tryptophan to the enzyme. With steady-state dialysis any analogue that binds to the allosteric site should displace [14C]phenylalanine already bound to that site, in the same way that unlabelled phenylalanine displaced radioactive phenylalanine in the experiments described previously. The validity of this argument was first checked by carrying out an experiment with the analogue m-fluorophenylalanine, which at 1 mM was observed to cause 92% inhibition of the dehydratase activity [7]. The results (Fig. 6) clearly demonstrate binding of the analogue to the allosteric site and indicate that the site was saturated by 1 mM m-fluorophenylalanine, since the subsequent addition of a large excess of phenylalanine did not increase the radioactivity of the effluent. Conversely 1 mM tyrosine or tryptophan caused n o change in the level of radioactivity in the effluent (Fig. 6) showing that these amino acids do not bind to the phenylalanine-binding site of chorismate mutase/ prephenate dehydratase. '

Binding of Phenylalanine to Chorismate MutaselPrephenate Dehydratase Reacted with Nbs,

We have shown previously that when three sulphydryl groups of chorismate mutase/prephenate dehydratase were modified by Nbs, the dehydratase activity was abolished and the remaining mutase activity

An important feature of the binding data presented above is that complete saturation of all the subunits on the chorismate mutase/prephenate dehydratase oligomer was achieved only in the presence of 0.4 M NaCl at pH 8.2 at 37 "C. Under the other conditions examined only approximately 0.6 - 0.8 mol of ligand bound per mol of enzyme subunit. The question then arises as to why this value deviates from a whole number in these latter situations since it is reasonable to expect the maximum number of binding sites per subunit for phenylalanine to be an integer. Explanations that the enzyme was impure or partially denatured are untenable for the following reasons. First, ultracentrifugation and acrylamide gel electrophoresis of the enzyme indicated at least 95 :d purity. Second, after the completion of the binding experiments the enzyme had the same specific activity, maximal inhibition by phenylalanine and Hill coefficient for inhibition by phenylalanine as that observed with freshly prepared enzyme. Third, different batches of enzyme, prepared at different times, yielded identical binding data. The most likely explanation then is that the enzyme has the potential to bind one molecule of phenylalanine per subunit but that extreme negative co-operativity prevents binding to the last site(s) on the enzyme oligomer at low ionic strength with the phenylalanine concentrations used. We conclude then that there is a single phenylalanine-binding site per subunit which is associated with the regulation of each of the mutase and dehydratase activities. Co-operative Aspects of Phenylalanine Binding The most striking feature of the saturation curves for the binding of phenylalanine to the enzyme is their deviations from conventional plots. At low ionic strength under the various conditions examined the binding is best described as being a mixture of positive and negative co-operativity between phenylalaninebinding sites [16]. This is particularly apparent from the Scatchard and Hill plots. Thus phenylalanine binds to successive subunits of the oligomer with different binding affinities. In

172

Binding of Phenylalanine to Chorismate Mutase/Prephenate Dehydratase

the presence of phenylalanine, chorismate mutasei prephenate dehydratase exists as a complex equilibrium mixture of tetramers, octamers and higher polymers [17]. Since phenylalanine could have different binding affinities for subunits in the tetramer compared with those in the octamer any theoretical binding equations describing the system would be complex and contain a number of unknown constants (including polymerization constants). Thus it is not meaningful to fit the experimental data to such equations in order to obtain quantitative values for the intrinsic phenylalanine binding constants. It is only possible to conclude that there is positive co-operativity for the binding of phenylalanine to the first 50 % of sites and negative co-operativity for binding to the last 50% of sites on the enzyme oligomer. This result is similar to that observed for the binding of NAD' to yeast glyceraldehyde-3-phosphatedehydrogenase [18]. It is likely that the general ligand-induced model of Koshland et aZ. [39] provides the best description of the binding of phenylalanine to the enzyme under these conditions. This conclusion is supported by the observed effect of phenylalanine on the ultraviolet difference spectrum of chorismate mutaseiprephenate dehydratase [lo]. As discussed in the preceding paper [lo], the extent of perturbation of a tryptophan residue parallels the binding of phenylalanine to the enzyme. These results suggest that the changes in conformation of the subunits are sequential and are induced by the binding of phenylalanine. This result rules out the possibility of a concerted change in the conformation of all the subunits as is required by the symmetry model of Monod et al. [20]. Effect of Enzyme Concentration on Phenylulunine Binding

The observations that polymerization of the enzyme is favoured by phenylalanine [17] and that binding of phenylalanine is dependent on the enzyme concentration raises the possibility that the co-operativity in ligand binding can be explained by polymerization effects [21]. However the simple polymerization model cannot explain the negative co-operativity observed in this study. It is more likely that the observed polymerization phenomena are a natural consequence of the phenylalanine-induced conformational changes [lo]. For example, in a study on glutamic dehydrogenase [22] it was demonstrated that the primary regulatory effect was the conformational change and not the concomitant polymerization. However, since polymerization may also alter the conformation of subunits, the phenomenon of association-dissociation could enhance or dampen ligand-induced changes and play an important role in the allosteric effect.

The concentration of enzyme used in the binding experiments affected the extent of co-operativity of the binding of phenylalanine. This is best seen by considering the maximum values of the Hill coefficients, which were 3.4 at the two highest enzyme concentrations and 2.1 at the lowest enzyme concentration (1 1 pM). The latter value is close to the Hill coefficient of 2.3 -2.4 obtained in the presence of substrates in kinetic experiments where the enzyme concentration was 0.02 pM [7]. We may conclude then that, despite the absence of the appropriate experimental data, the binding of chorismate or prephenate to the enzyme does not greatly affect the extent of co-operativity between phenylalanine-binding sites. Furthermore, using the data reported previously (Table 1 of [S]), the concentration of chorismate mutaseiprephenate dehydratase in Escherichiu coli strain JP 492 can be calculated to be around 10-20 pM. Strain JP 492, which has the same genotype for pheO as JP 171 [23], has levels of chorismate mutase/ prephenate dehydratase activites some ten times higher than the normal repressed level in wild-type cells and three times higher than the largest observed derepressed level in starved aromatic auxotrophs [2,23]. Thus the subunit concentration of this enzyme in wild-type E. coli is in the region of 1 - 2 pM. Independent reports on the intracellular concentration of phenylalanine in E. coli gives values of 1600 pM [24] and 600 - 800 pM [25]. Brown observed that the range of values is fairly independent of the phenylalanine concentration in the medium (from 1 to 10 pM) or of the state of growth of cells [25]. Examination of the kinetic data [7,8] indicates that the enzyme is almost completely inhibited under these conditions (around 90 and that variation of the phenylalanine concentration from 600 to 800 pM has little, if any, effect on the extent of inhibition. Because the binding data, obtained at 11 - 240 pM enzyme, and the kinetic data, obtained at about 0.02 pM enzyme, are qualitatively similar, these affects should also apply to the enzyme at its concentration in uivo. Thus it appears that unless there is considerable intracellular compartmentalization of the phenylalanine pool, chorismate mutaseiprephenate dehydratase functions in vivo in a state of permanent inhibition, unaffected by the normal changes in growth requirements of the organism. The physiological reason for the feedback inhibition is unclear, unless it acts as a fail-safe device, whereby any dramatic decrease in the phenylalanine concentration below about 25 pM leads rapidly to a substantial increase in enzymic activity and phenylalanine synthesis.

x)

Eflect of Temperature on Phenylalanine Binding

The increase in binding of phenylalanine that was observed when the temperature was lowered suggests

173

M.-J. Gething and B. E, Davidson

that ionic interactions and/or hydrogen bonds are the predominant forces involved in the overall binding process. This is consistent with the previous observations that the carboxyl group of phenylalanine is necessary for inhibition of the enzyme [7].It is reasonable to postulate that the carboxylate anion interacts with a positive charge in the phenylalanine-binding site. Calculation of AH from this data was not possible because of the above-mentioned difficulties in the determination of the intrinsic binding constants. Eflect of NuCl on Phenylulunine Binding While an increase in NaCl concentration of 0.1 M favours the binding of phenylalanine at all ligand concentrations, an increase of a further 0.3 M has a more complicated effect (see Fig. 4). This suggests that NaCl has more than one effect on the binding of phenylalanine to the enzyme. Higher ionic strength could decrease the interactions between the a-carboxyl group of phenylalanine and the positively charged group at the phenylalaninebinding site but increase the strength of hydrophobic interactions between the benzene ring of phenylalanine and hydrophobic groups in the binding site. The relative contributions of these effects would be different at different ionic strengths. The Binding of Chorismute Mutuselprephenute Dehydrutase to Sephurosyl-Phenylulunine

The binding data described above provide a partial explanation of the chromatographic behaviour of the enzyme on Sepharosyl-phenylalanine [8], where binding was favoured by lowering the pH or increasing the NaCl concentration. It is probable that the only relevant effect is that of p H or NaCl on the binding of phenylalanine to the first site on any individual enzyme molecule since on steric grounds it is unlikely that more than one phenylalanine-binding site on the enzyme will be able to bind to Sepharosyl-phenylalanine. The increase in binding to the resin at pH 7.4 compared with pH 8.2 is readily explained since the affinity of the enzyme for the first phenylalanine molecule is greater at the lower p H (Fig. 5). The explanation of the effect of NaCl is Jess obvious since at pH 8.2 the binding experiments indicate that the addition of 0.4 M NaCl decreases the affinity of the enzyme for the first molecule of phenylalanine. It has been proposed above that there may be more than one mechanism whereby NaCl affects the binding of phenylalanine to chorismate mutase/prephenate dehydratase and that at least one effect must promote binding and one antagonize the binding. It is possible that with Sepharosyl-phenylalanine (which must be considered to be a different ligand to phenylalanine)

that the effect of NaCl that promotes binding is relatively more significant. Binding Studies Using Phenylulunine Analogues and Nhs2-Modified Enzyme The use of this technique has enabled a number of subsidiary aspects arising from our previous investigations to be examined. The inability of tyrosine and tryptophan to inhibit the enzyme significantly [7] has been shown to be due to the failure of these compounds to bind to the allosteric site. Similarly, the assumption made previously that the extent of inhibition of a phenylalanine analogue was a measure of its strength of binding to the allosteric site [ 7 ] ,has been substantiated by the observations that m-fluorophenylalanine, but not tyrosine or tryptophan, bind to the allosteric site of the enzyme. These observations enhance the plausibility of our picture of the phenylalanine-binding site [7]. Previously we had found that modification of three sulphydryl groups of the enzyme caused de-sensitization of the remaining mutase activity to phenylalanine [S, 151. The observation that phenylalanine does not bind to this modified enzyme suggests that there is a sulphydryl group involved either at the phenylalaninebinding site or in the consequent conformational changes. We thank Dr W. Sawyer for his steady-state dialysis apparatus and Dr G. McKenzie for help in its use. Financial assistance from the Australian Research Grants Committee is gratefully acknowledged.

REFERENCES 1 . Gibson, F. & Pittard, J. (1968) Bacteriol. Rev. 32, 465-492. 2. Pittard, J. & Gibson, F. (1970) Curr. Top. Cell. Re&. 2, 29 63. 3. Davidson, B. E., Blackburn, E. H. & Dopheide, T. A. A. (1 972) J. Biol. Chem. 247, 4441 -4446. 4. Gething, M r J . & Davidson, B. E. (1976) Eur. J . Biochrm. 71, 327 336. 5. Gething, M;J. & Davidson, B. E. (1977) Eur. J . Biochem. 78, 103 - 110. 6. Gething, M.-J. & Davidson, B. E. (1977) Eur. J . Biochrm. 78, 111-117. 7. Dopheide, T. A . A., Crewther, P. & Davidson, B. E. (1972) J . Biol. Chem. 247, 4447-4452. 8. Gething, M.-J., Davidson, B. E. & Dopheide, T. A . A. (1976) Eur. J . Biochem. 71. 317-325. 9. Schmit, J . C.. Artz, S . W. & Zalkin, H. (1970) J. Bid. Chem. 245, 401 9 - 4027. 10. Gething, M.-J. & Davidson, B. E. (1978) Ew. J . Bioclzenz. 86, 159-164. 11. Davidson, B. E. (1970) Eur. J . Biochrm. 14, 545- 548. 12. Madsen, N. P. (1969) A n d . Biochem. 29, 542-544. 13. Colowick, S. P. & Womack, F. C. (1969) J . Biof. Chem. 244, 774 - 777. 14. Baillie, L. A . (1960) Inr. .I. Appl. Rudiut. hot. 8, 1 -7. 15. Gething, M.-J. (1973) Ph. D. Thesis, University of Melbourne. ~

~

174

M.-J. Gething and B. E. Davidson: Binding of Phenylalanine to Chorismate Mutase/Prephenate Dehydratase

16. Levitzki, A. & Koshland, D. E. (1976) Curr. Top. Cell. Regul. 10, 1-40. 17. Baldwin, G. S. (1974) Ph. D. Thesis, University of Melbourne. 18. Cook, R. A. & Koshland, D. E. (1970) Biochemistry, 9, 33373342. 19. Koshland, D. E., Nemethy, G. & Filmer, D. (1966) Biochemisrry, 5, 365- 385. 20. Monod, J . , Wyman, J. & Changeux, J.-P. (1965) 2. Mol. Bid. 12, 88-118.

M.-J. Gething, Imperial Cancer Research Fund Laboratories, Lincoln's Inn Fields, P.O. Box 123, London, Great Britain, WC2A3PX B. E. Davidson, Russell Grimwade School of Biochemistry, University of Melbourne, Parkville, Victoria, Australia 3052

21. Nichol, L. W., Jackson, W. J. H. & Winzor, D. J. (1967) Biochemistry, 6 , 2449 2456. 22. Huang, C. Y . & Frieden, C. (1969) Proc. Nut1 Acad. Sci.U . S . A . 64, 338 - 344. 23. Im, S. W. K. & Pittard, J. (1971) J. Bacteriol. 106, 784-790. 24. Piperno, J. R. & Oxender, D. L. (1968) J . B i d Chem. 243, 5914-5929. 25. Brown, K. (1970) J . Bucteriol. 104, 177-188. ~