Regulation of Chorismate Mutase-Prephenate Dehydratase and Prephenate Dehydrogenase from Alcaligenes Eutrophus CORNELIUS G

JOURNAL OF BACTZRIOLOGY, May 1976, p. 723-732 Vol. 126, No. 2 Copyright C 1976 American Society for Microbiology Printed in U.SA.

Regulation of Chorismate Mutase-Prephenate Dehydratase and Prephenate Dehydrogenase from Alcaligenes eutrophus CORNELIUS G. FRIEDRICH,' BARBEL FRIEDRICH,2* AND HANS G. SCHLEGEL Institut f4r Mikrobiologie der Gesellschaft fur Strahlen- und Umweltforschung mbH, MUnchen, in Gottingen und Institut fur Mikrobiologie der Universitdt G6ttingen, Gottingen, West Germany Received for publication 12 December 1975 Highly purified enzymes from Alcaligenes eutrophus H 16 were used for kinetic studies. Chorismate mutase was feedback inhibited by phenylalanine. In the absence ofthe inhibitor, the double-reciprocal plot was linear, yielding a Km for chorismate of 0.2 mM. When phenylalanine was present, a pronounced deviation from the Michaelis-Menten hyperbola occurred. The Hill coefficient (n) was 1.7, and Hill plots of velocity versus inhibitor concentrations resulted in a value of n' = 2.3, indicating positive cooperativity. Chorismate mutase was also inhibited by prephenate, which caused downward double-reciprocal plots and a Hill coefficient of n = 0.7, evidence for negative cooperativity. The pH optimum of chorismate mutase ranged from 7.8 to 8.2; its temperature optimum was 47 C. Prephenate dehydratase was competitively inhibited by phenylalanine and activated by tyrosine. Tyrosine stimulated its activity up to 10-fold and decreased the Km for prephenate, which was 0.67 mM without effectors. Trypto- phan inhibited the enzyme competitively. Its inhibition constant (K{ = 23 ,uM) was almost 10-fold higher than that determined for phenylalanine (K, = 2.6 ,M). The pH optimum of prephenate dehydratase was pH 5.7; the temperature optimum was 48 C. Prephenate dehydrogenase was feedback inhibited by tyrosine. Inhibition was competitive with prephenate (K{ = 0.06 mM) and noncompetitive with nicotinamide adenine dinucleotide. The enzyme was further subject to product inhibition by p-hydroxyphenylpyruvate (K{ = 0.13 mM). Its Km for prephenate was 0.045 mM, and that for nicotinamide adenine dinucleotide was 0.14 mM. The pH optimum ranged between 7.0 and 7.6; the temperature optimum was 38 C. It is shown how the sensitive regulation ofthe entire enzyme system leads to a well-balanced amino acid production. Chorismic acid plays a strategic role in the synthesis of phenylalanine and tyrosine pro- multibranched pathway of aromatic com- vides several examples for multifunctional en- pounds. It is the precursor of phenylalanine, zymes and different organization patterns (Fig. tyrosine, and tryptophan and leads to metabo- 1) (6, 8, 20, 21, 29). The new type oforganization lites such as vitamin K, folate, ubiquinone, and discovered in Alcaligenes eutrophus H 16 (13) hydroxybenzoate (16, 24). To balance the flow of led to a study of the regulation of the bifunc- intermediates, a sensitive regulation, espe- tional enzyme chorismate mutase (EC 5.4.99.5)- cially at the branch points, is required. A wide prephenate dehydratase (EC 4.2.1.51) and pre- variety of control mechanisms have been dis- phenate dehydrogenase (EC 1.3.1.12). During covered (26), including multivalent allosteric the overall reaction from chorismate to phenyl- feedback control as well as interactions be- pyruvate, catalyzed by the bifunctional protein, tween metabolites and enzymes that are lo- prephenate accumulated. The dissociation of cated in quite different metabolic pathways the intermediate product allows prephenate de- (27). hydrogenase to compete with prephenate dehy- Regarding aromatic amino acid biosynthesis, dratase for substrate. the physical association of the enzymes is often part ofthe regulatory system. In particular, the MATERIALS AND METHODS I Enzymes. Chorismate mutase-prephenate dehy- Present address: Department of Nutrition and Food dratase and prephenate dehydrogenase from A. eu- Science, Massachusetts Institute of Technology, Cam- Pure bridge, Mass. 02139. trophus H 16 were purified as described (11). 2 Present addreos: Department of Biology, Mamachu- enzyme from step 7 of Table 1 (37 mutase units per setts Institute of Technology, Cambridge, Mass. 02139. mg of protein [11]) was used for all experiments. 723 724 FRIEDRICH, FRIEDRICH, AND SCHLEGEL J. BACTERIOL. (1)

Prephenate Phenylpyruvate - Phenylalanine (a) Chorismate Prephenate p-Hydroxyphenylpyruvate Tyrosine (2) (1)

(b) Chorismate Prephenate Phenylpyruvate Phenylalanine

L3) p -Hydroxyphenylpyruvate Tyrosine

( 5 ) -Phenylpyruvate - Phenylalanine (c) Chorismate -Prephenate (3 p-Hydroxyphenylpyruvate -e- Tyrosine FIG. 1. Organization of the enzymes involved in phenylalanine and tyrosine biosynthesis in Escherichia coli (a), Alcaligenes eutrophus H 16 (b), and Neurospora crassa (c). Enzymes: Chorismate mutase-prephenate dehydratase (1), chorismate mutase-prephenate dehydrogenase (2) (brackets indicate bifunctional nature of these proteins); prephenate dehydrogenase (3); chorismate mutase (4); prephenate dehydratase (5).

Enzyme assays. One enzyme unit catalyzes the land); and DL-m- and p-fluorophenylalanines, /3- conversion of 1 zimol of substrate per min. The spe- phenylpyruvic acid, and p-hydroxyphenylpyruvic cific activity is expressed as units per milligram of acid from Sigma Chemical Co. (St. Louis, Mo.). protein. Chorismic acid was prepared as barium salt ac- Chorismate mutase. Purified enzyme was as- cording to the method of Gibson (15). Barium pre- sayed according to the method of Cotton and Gibson phenate was prepared by using the auxotrophic mu- (7), using an optical test. The rate of chorismate tant strain 6B-1 of A. eutrophus H 16 as previously disappearance was followed continuously with a described (13). Barium chorismate and barium pre- Servogor recorder throughout an interval of 3 min at phenate were converted to the potassium salts with 275 nm, the absorbance maximum for chorismate. a twofold molar excess of potassium sulfate before The reaction mixture contained, in a final volume of use. The concentration of stock solutions of both 1.0 ml, 1 mM potassium chorismate and 50 mM substrates was corrected for purity. All the other potassium phosphate (pH 7.8). After preincubation chemicals were of analytical reagent grade and ob- of the reaction mixture at 37 C, the reaction was tained from commercial sources. initiated by the addition of the enzyme. The amount of chorismate converted was calculated by using a molar extinction coefficient of 2,630 for chorismate. RESULTS Prephenate dehydratase. The assay was carried Effect of pH on the regulatory properties of out according to the method of Cotton and Gibson (6) with slight modifications (11). the purified enzymes. The optimum pH for Prephenate dehydrogenase. The activity of pre- chorismate mutase activity in potassium phos- phenate dehydrogenase was determined by the phate buffer was pH 7.8 and in tris(hydroxy- method of Champney and Jensen (4) as described methyl)aminomethane-hydrochloride buffer was (11). pH 8.2 (Fig. 2a). This difference may have Protein. The protein of the enzyme preparations been due to changes in ionic strength of the was determined as described (11). respective buffers at a given pH. However, the Chemicals. Nicotinamide adenine dinucleotide enzyme activity was not affected in phosphate (NAD) and reduced NAD were obtained from Boeh- buffer, whose ringer Mannheim GmbH (Mannheim, West Ger- molarity ranged between 25 and many). L-Isomers of the amino acids of the highest 100 mM. purity grade available were purchased from Merck Chorismate mutase was inhibited by phenyl- (Darmstadt, West Germany). Analogous compounds alanine only in alkaline solution. At a pH of 7.2 of phenylalanine and intermediates of the pathway no inhibition occurred, and below this pH the were obtained as follows: D-phenylalanine and 3,4- activity measured in the presence of phenylala- dihydroxyphenylalanine from Fluka AG (Switzer- nine was even higher than without the effector VOL. 126, 1976 A. EUTROPHUS PHENYLALANINE-TYROSINE REGULATION 725 (Fig. 2a). This observation leads to the assump- moid, and the extent of sigmoidicity increased tion that under acid conditions phenylalanine with increasing inhibitor concentrations up to 1 protects the enzyme against inactivation. mM, at which point the inhibition was maxi- Prephenate dehydratase activity reached a mum. The concave upward double-reciprocal maximum at pH 5.7 (Fig. 2b). However, the plot (Fig. 3) and the Hill plot of these data (Fig. assays were conducted at pH 7.8 due to the acid 3, insert) with slopes of 1.3 to 1.7 were consist- lability of the substrate prephenate, (14), which ent with positive cooperative interactions of is represented in Fig. 2b. At an alkaline pH, chorismate binding sites in the presence of the both activation by tyrosine and inhibition by feedback inhibitor. phenylalanine were most pronounced. Within Plots of chorismate mutase activity versus an acid pH range, the regulatory properties of varying concentrations of phenylalanine were the enzyme were scarcely visible (Fig. 2b). sigmoid, indicating cooperativity of the inhibi- Prephenate dehydrogenase exhibited a broad tor (Fig. 4a). However, cooperativity was de- pH optimum between 7.0 and 7.6. This pH rep- pendent upon chorismate concentration as resents a range where maximal inhibition by shown by the Hill plots, which were linear in a tyrosine occurred (Fig. 2c). range of 0.1 to 0.4 mM phenylalanine (Fig.. 4b). Effect of temperature. The temperature op- The values of n' deduced from these data timum for chorismate mutase was found to be ranged from 1.1 to 2.3. 47 C and was almost identical with the value Prephenate inhibition of chorismate mu- determined for prephenate dehydratase activ- tase. Chorismate mutase activity was inhibited ity (48 C). The value for prephenate dehydro- by its product prephenate. Double-reciprocal genase differed significantly; the temperature plots of the substrate saturation curve in the optimum was 38 C. presence of prephenate were concave down- Effect of phenylalanine on chorismate mu- ward, indicating negative cooperativity (Fig. tase activity. In the absence of phenylalanine, 5). The reaction rate of unireactant enzymes is the substrate saturation curve for chorismate the result offorward and reverse velocity. How- was hyperbolic. Its double-reciprocal plot was ever, under the assay conditions the reverse linear and resulted in a Km for chorismate of reaction, the formation of chorismate from 0.20 mM (Fig. 3). In the presence of phenylala- prephenate, was not detectable. The following nine, the substrate saturation curves were sig- reaction, the conversion of prephenate to phen-

E 0 4 a)8 >1 E :_ I:0 3 50 z c 2 a) 0 E 1

pH pH pH FIG. 2. pH dependence ofchorismate mutase (a), prephenate dehydratase (b), and prephenate dehydrogenase (c) activity in the presence ofeffectors. Symbols: (@) Without any effector; (A) with phenylalanine at the concentrations of(a) 1 mM and (b) 8 MM; (-) with tyrosine at concentrations of(b) 5 puM and (c) 0.5 mM. The chemical decay of prephenate is shown in (b) (0) by the formation of phenylpyruvate. Enzyme activities determined in the presence of effectors are expressed as percentage of activity and related to that activity obtained without any effector at a given pH, set as 100%. (a) The reaction mixture contained in 1.0 ml; 50 mM potassium phosphate or 50 mM tris(hydroxymethyl)aminomethane-hydrochloride buffer; 1 mM potassium chorismate; and 1.7 pg of purified chorismate mutase-prephenate dehydratase. (b) The reaction mixture contained, in 0.50 ml: 50 mM potassium phosphate, 1 mM potassium prephenate, and 3 pg of purified chorismate mutase-prephenate dehydratase. In the presence of tyrosine only 0.8 pg of enzyme was used. (c) The reaction mixture contained, in 1.0 ml: 100 mM potassium phosphate, 02 mM potassium prephenate, 1 mM NAD, and 0.5 pg ofpurified prephenate dehydrogenase. 726 FRIEDRICH, FRIEDRICH, AND SCHLEGEL J. BACTERIOL. yielding a Km of 0.67 mM (Fig. 6). Neither the inclusion of the activator tyrosine nor the presence ofthe inhibitors phenylalanine and tryptophan in the reaction mixture modified the Mi- chaelis-Menten type of substrate saturation kinetics. As shown for phenylalanine (Fig. 6), the Km for prephenate increased. Tyrosine, however, exhibited just the opposite effect. The reaction rate of the enzyme increased in direct response to the activator concentration. Tyro- sine affected both maximal velocity and Km for prephenate, indicated by different intercepts and slopes of the double-reciprocal plots (Fig. 7). Similar kinetic data are reported for prephenate dehydrogenase from Bacillus subtilis in the presence of the inhibitors tryptophan andp- hydroxyphenylpyruvate (4). The kinetic data were further analyzed by replotting the slopes obtained from the double-reciprocal plots versus inhibitor and activator concentrations, respectively (5). The resulting slope replots for phenylalanine and tryptophan inhibition were linear (Fig. 8). The inhibition constant (K,,) of 10 20 such replots was determined from the horizon- tal intercept, resulting in a value for phenylala- 1/[Chorismate](mM-1) nine of 2.6 AM and an almost 10-fold higher FIG. 3. Double-reciprocal plot of the substrate sat- value for tryptophan (23 ,uM). The activation of uration of chorismate mutase. The reaction mixture prephenate dehydratase by tyrosine (Fig. 8a) contained, in 1.0 ml: 50 mM potassium phosphate, was in accordance with a hyperbolic activation pH 7.8; varying concentrations ofpotassium choris- curve (5). mate; and 2.5 mg of purified chorismate mutase- In crude extracts, prephenate dehydratase prephenate dehydratase. Phenylalanine was in- was significantly less inhibited by phenylala- cluded as indicated. Symbols: (a) No phenylalanine; nine, and almost no inhibitory effect of trypto- (A) 0.1 mMphenylalanine; and (U) 0.5 mMphenyl- phan could be detected. However, compared alanine. Hill plots of these data are given in the insert. with purified enzyme preparations, the protein was more strongly activated by tyrosine. These observations and the finding that the total ac- ylpyruvate, was insignificant because initial tivity of the enzyme increased during the puri- rates were taken. The prephenate concentra- fication, unless activated (11), led to the as- tion used never exceeded the K, of prephenate sumption that prephenate dehydratase was dehydratase. Therefore the amount of phenyl- partially inhibited in the crude extract. Al- pyruvate formed was minimal during the time though this effect was not released by dialysis, required to measure the initial rates of choris- it is likely that the inhibitor was tightly bound mate utilization. to the protein and dissociated during the course Further evidence for negative cooperative in- of purification. It is unlikely that the low activ- teractions was obtained from linear Hill plots, ity of prephenate dehydratase in crude extracts with n values ranging from 1.04 to 0.74. Plots of was due to another prephenate-consuming re- 1/v versus inhibitor concentrations also ex- action. Prephenate dehydrogenase did not func- hibited concave downward curvature. tion under the assay conditions because of its Although the negative cooperative effects cofactor dependency. need further investigation, it is very likely that Effect of phenylalanine analogues on chor- the catalytic site of prephenate dehydratase ismate mutase and prephenate dehydratase functions as a regulatory site for chorismate activity. The effect of phenylalanine analogues mutase. on chorismate mutase-prephenate dehydratase Effectors influencing prephenate dehydra- is shown in Table 1. The 1-isomer of phenylala- tase. The substrate saturation curve of prephe- nine did not exhibit any remarkable effect on nate dehydratase for prephenate was hyper- either activity; p-fluorophenylalanine inhibited bolic. The double-reciprocal plot was linear, only prephenate dehydratase and left the mu- VOL. 126, 1976 A. EUTROPHUS PHENYLALANINE-TYROSINE REGULATION 727

b 1.8

1.2

0.6

0 0

0.6 0 0.2 0.4 0.6 0.8 1.0 -1.8 -14 -1.0 -0.6 -Q2 [Phenylalanine] (mM) log [Phenylalanine] (mM) FIG. 4. Effect ofphenylalanine on chorismate mutase activity. (a) The reaction mixture contained, in 1.0 ml: 50 mM potassium phosphate, pH 7.8, varying concentrations of phenylalanine and 2.5 pg purified enzyme. Potassium chorismate was used at following concentrations: (A) 0.7 mM; (-) 0.2 mM; and (-) 0.08 mM. The corresponding Hill plots of these data are shown in (b).

0 4 8 12 16 1/ [Prephenate] (mM-1) FIG. 6. Double-reciprocal plot ofthe substrate saturation (prephenate) of prephenate dehydratase in 1/[Chorismate] (mM 1) the presence ofphenylalanine. The reaction mixture contained, in 0.50 ml; 50 mM potassium phosphate, FIG. 5. Prephenate inhibition of chorismate mu- pH 7.8; varying concentrations ofpotassium prephe- tase. The data are presented in a double-reciprocal nate; and 1.7 pg ofpurified enzyme. Phenylalanine in 1.0 ml: 50 plot. The reaction mixture contained, was included at fixed concentrations as indicated. mM potassium phosphate, pH 7.8; differenf concen- Symbols: (0) No phenylalanine; (A) 1 uM; and (-) 4 trations ofpotassium chorismate; and 2.5 pg ofpuri- phenylalanine. fied enzyme. Potassium prephenate was included as indicated. Symbols: (0) No potassium prephenate; fluence chorismate mutase activity. They inter- 0.4 and 0.7 (A) 0.1 mM; (U) 0.2 mM; (0) mM; (4) fere with the assay procedure for prephenate mM potassium prephenate. dehydratase, and therefore- they could not be tase activity unaffected. Both enzyme activities included in the investigation on this enzyme were sensitive to inhibition by m-fluorophenyl- activity. alanine. Intermediates such as phenylpyru- Effect of tyrosine on prephenate dehydro- vate and p-hydroxyphenylpyruvate did not in- genase activity. The substrate saturation curve 728 FRIEDRICH, FRIEDRICH, AND SCHLEGEL J. BACTERIOL. of prephenate dehydrogenase was measured -the linear double-reciprocal plot was found to with prephenate as the variable substrate at a be 0.045 mM. The inclusion of tyrosine in the fixed saturating concentration of the second reaction mixture caused sigmoidicity ofthe pre- substrate, NAD; the resulting curve was hyper- phenate saturation curve (Fig. 9a). The double- bolic. The Km for prephenate extrapolated from reciprocal plot was concave upward. Hill plots of these data yielded straight lines with slopes of 1.0 to 1.4, indicating positive cooperativity of prephenate (Fig. 9b). Tyrosine did not affect the linearity of the double-reciprocal plot when NAD was the variable substrate. In the presence of tyrosine the Km for NAD remained constant (0.14 mM), indicating a noncompetitive type of inhibition (Fig. 10). The inhibition of prephenate dehydrogenase 1.5- by tyrosine was competitive with respect to prephenate. Plots of 1/v versus tyrosine concentration were linear (Fig. 11). The inhibitor constant for tyrosine was determined as 0.06 mM. > 1.0 Although tyrosine exhibited heterotropic cooperative effects on prephenate binding, homo- tropic cooperative interactions were missing. 1~~~~~~,~A The Hill coefficient n' for the inhibition curve did not exceed a value of 1.0. A similar kinetic

TABLE 1. Inhibition of chorismate mutase- prephenate dehydratase by phenylalanine analogues Mutase Dehydra- 0 4 8 12 16 Inhibitoe activity tase activ- 1/ [Prephenate] (mM-1) (%) ity (%) FIG. 7. Activation of prephenate dehydratase by None 100 100 D-Phenylalanine 98 93 tyrosine. Data are presented in a double-reciprocal plot. The reaction mixture contained, in 0.50 ml: 50 p-Fluorophenylalanine 100 49 62 13 mM potassium phosphate, pH 7.8; varying concen- m-Fluorophenylalanine 95 NDb trations of potassium prephenate; and 0.85 pg of Phenylpyruvate p-Hydroxyphenylpyruvate 100 ND purified enzyme. Tyrosine was included at fixed concentrations as indicated. Symbols: (0) No tyrosine; a Effector concentration was 1 mM. (A) 2 pM; (U) 8 pM; and (0) 20 pM tyrosine. b ND, Not determined.

[Effector] (mM) [L-Tryptophan] (mM)

FIG. 8. Slope replots for the effects of phenylalanine, tyrosine (a), and tryptophan (b) on prephenate dehydratase activity. The data from Fig. 6 and 7 are replotted as slopes that were taken from the different lines ofthe Lineweaver-Burk plots with respect to effector concentration. The double-reciprocal plots obtained in the presence oftryptophan are not shown. However, the assay mixture was the same as described in Fig. 6 except that tryptophan was used instead of phenylalanine. The data from Fig. 7 were calculated for a protein concentration of1.7 pg. VOL. 126, 1976 A. EUTROPHUS PHENYLALANINE-TYROSINE REGULATION 729

FIG. 9. Substrate saturation curves ofprephenate dehydrogenase forprephenate in the presence oftyrosine. (a) The reaction mixture contained, in 1.0 ml: 100 mM potassium phosphate, pH 7.8; 1 mM NAD; different concentrations ofpotassium prephenate; 0.35 pg ofpurified enzyme; and tyrosine as indicated. Symbols: (a) No tyrosine; (A) 0.2 mM; and (U) 1 mM tyrosine. (b) Hill plots of the data obtained from (a).

1/[NAD] (mM-') FIG. 10. Double-reciprocal plot ofNAD saturation for prephenate dehydrogenase. The reaction mixture contained, in 1.0 ml: 100 mM potassium phosphate, [L-Tyrosine] (mM) pH 7.8; 0.2 mM potassium prephenate; varying con- FIG. 11. Effect of tyrosine on prephenate dehydro- centrations ofNAD; and 0.45 pg ofpurified enzyme. genase activity. The reaction mixture contained, in was included at concentrations as in- Tyrosine fixed 1.0 ml: 100 mM potassium phosphate, pH 7.8; 1 mM dicated. Symbols: (a) No tyrosine; (A) 0.1 mM; and NAD; varying concentrations of tyrosine; and 0.35 (-) 0.5 mM tyrosine. pg of purified enzyme. Potassium prephenate was used at the following concentrations: (a) 0.05 mM; (A) 0.1 mM; and (U) 02 mM. behavior is reported for prephenate dehydrogenase from B. subtilis (4). Furthermore, prephenate dehydrogenase was inhibited by its tion constants were determined as 0.47 mM and product p-hydroxyphenylpyruvate, and inhibi- 0.20 mM, respectively. tion was competitive with respect to prephe- Competition for prephenate. Experiments nate. The Ki for p-hydroxyphenylpyruvate was with crude extracts indicated that tyrosine not 0.13 mM. Structurally related compounds such only inhibited the activity of prephenate dehy- as dihydroxyphenylalanine 'and cumaric acid drogenase but also affected the linearity of the inhibited the enzyme competitively. The inhibi- reaction rate. Without tyrosine the time course 730 FRIEDRICH, FRIEDRICH, AND SCHLEGEL J. BACTERIOL. was linear for at least 5 min (Fig. 12a [1]). partially lost by separating prephenate dehy-) When tyrosine was added the reaction rate be- dratase from prephenate dehydrogenase was came nonlinear with respect to time (Fig. 12a reconstructed by mixing both purified enzyme [2]). This effect was even more pronounced at a preparations (Fig. 12b [4]). higher concentration of tyrosine (Fig. 12a [3]). Purified enzyme, although still inhibited by DISCUSSION tyrosine, revealed a linear reaction rate (Fig. InA. eutrophus H 16, chorismate mutase and 12b [1, 3]). This observation led to the conclu- prephenate dehydratase activities reside in a sion that the apparent inactivation of prephe- single protein, whereas prephenate dehydro- in crude ex- nate dehydrogenase by tyrosine genase is not associated with other enzyme ac- tract was an indirect effect and primarily tivities of the phenylalanine-tyrosine pathway caused by prephenate dehydratase. Prephenate (11). This fact favors the assumption that phen- dehydratase was activated in the presence of ylalanine formation is facilitated and the flow tyrosine; its affinity for prephenate increased so of chorismic acid directed toward phenylala- that less substrate became available for prephe- nine. Furthermore, phenylalanine is a precur- nate dehydrogenase. As a result of this, the sor oftyrosine in tyrosine auxotrophic mutants. reaction rate of the latter enzyme slowed down. The bacterium is able to hydroxylate phenylal- The effect of tyrosine was reversible by phenyl- anine (12). However, phenylalanine hydroxyl- alanine (not shown). ase as an inducible enzyme is probably involved The observations made with crude extract in the catabolism of phenylalanine as similarly obviously reflect the balanced synthesis of reported for Pseudomonas aeruginosa (2). phenylalanine and tyrosine in vivo. The con- Therefore, under normal conditions of growth, certed action of enzymes and effectors that got tyrosine is synthesized by prephenate dehydrogenase followed by transamination. The highly sensitive control of the three enzymes specifi- a b cally involved in phenylalanine and tyrosine formation allows a well-balanced production of 8- 16- both amino acids (Fig. 13). 3~~~~~~~~~~~~ The first point of regulation is chorismate a) a) mutase, which is allosterically inhibited by EF6 E12t phenylalanine. Phenylalanine increased the

o ~~~0 cooperativity of chorismate and revealed homo- tropic cooperative interactions, indicated by a < 4 < 8 Hill coefficient of n' = 2.3. The presence of z z

,) phenylalanine decreased the apparent affinity of chorismate mutase for chorismate and in- E2 E hibited the enzyme up to a maximum of 70%, which still provides enough prephenate to guar- antee tyrosine synthesis. The enzyme activity 0 0( was also inhibited by its product prephenate. 0 2 4 6 2 4 6 Prephenate caused negative cooperativity of Time (min) Time (min) chorismate, evidence for interaction between FIG. 12. Competition between prephenate dehy- different binding sites. Positive as well as nega- dratase and prephenate dehydrogenase for their com- tive cooperativity exhibited by one enzyme are mon substrate. (a) The reaction mixture contained, for cytidine 3'-triphosphate syn- in 1.0 ml: 100 mM potassium phosphate, pH 7.8; 1 also reported mM NAD; 0.2 mM potassium prephenate; and crude thetase (22). It is likely that in the case of extract of187 pg ofprotein (1). Tyrosine was added chorismate mutase the negative cooperativity at the following concentrations: (2) 0.2 mM; (3) 1 is due to the second catalytic function of the mM. (b) The reaction mixture contained, in 1.0 ml: enzyme, which binds prephenate as a sub- 100 mM potassium phosphate, pH 7.8; 1 mM NAD; strate. The catalytic site of prephenate dehy- 0.2 mMpotassium prephenate; and 0.8 pg ofpurified dratase could be a regulatory site for choris- prephenate dehydrogenase (8.9 U/mg) (1). To this mate mutase. Evidence for chorismate and pre- assay mixture was either added (2)1.7 pg ofpurified phenate binding sites not being identical is as chorismate mutase-prephenate dehydratase (37 mu- follows: (i) mutants were isolated that were tase units/mg) or (3) 0.2 mM tyrosine. Assay (4) while re- contained a mixture of prephenate dehydrogenase, only missing prephenate dehydratase chorismate mutase-prephenate dehydratase, and ty- taining normal chorismate mutase activity rosine at the same concentrations as mentioned (11); and (ii) in the overall reaction, prephenate above. dissociated from the protein before conversion VOL. 126, 1976 A. EUTROPHUS PHENYLALANINE-TYROSINE REGULATION 731

--HYDROXYPHENYL-r PYRUVATE TYROSINEI+ fect was even more pronounced when phenylal- 1+ anine was present because phenylalanine in- + hibited prephenate dehydratase competitively -- + ,_c_ --|-A ------+ (Ki = 2.6 ,uM). Increasing the concentration of tyrosine gave just the opposite effect. By acti- CHORISMATE t PREPHENATE + vating prephenate dehydratase, the flow ofpre- i _C4_+ + + phenate was shifted toward phenylalanine. C\ ++++++++++++++ TRYPTOPHAN- r Furthermore, prephenate dehydratase was PHENYLPYRUVATE - PHENYLALANINE competitively inhibited by tryptophan (Ki = 23 ,uM). Although the physiological significance of this effect is not fully understood, other exam- FIG. 13. Regulation of phenylalanine and tyro- ples of cross-pathway regulations are known. sine biosynthesis in Alcaligenes eutrophus H 16. Prephenate dehydratase from B. subtilis is in- Symbols: (-- -) Feedback inhibition; (+ + +) activa- hibited by tryptophan and activated by leucine tion. The mechanisms of inhibition are indicated by and methionine (4, 27). Chorismate mutase (A) allosteric or (C) competitive. from Neurospora crassa (1), Euglena gracilis (31), and Saccharomyces cerevisiae (30) is acti- to phenylpyruvate (11). Functionally distinct vated by tryptophan. Prephenate dehydrogen- sites for chorismate mutase and prephenate de- ase from both N. crassa (3) and S. cerevisiae hydratase are also described for Salmonella ty- (23) is subject to cross-pathway activation by phimurium (28). The dissociation ofprephenate phenylalanine. from the protein during the overall reaction The regulation pattern presented in Fig. 13 provides substrates for both prephenate dehy- fits well with the organization of the enzymes drogenase and prephenate dehydratase. The concerned. The high sensitivity of the system accumulation of prephenate was experimen- obviously reflects the situation in vivo. In vitro tally shown (11). However, this was already the competition between the two prephenate- evident from the turnover numbers of the en- utilizing enzymes could be demonstrated in zyme activities concerned. Chorismate mutase crude extracts and in a reconstructed system revealed a turnover of 2,210 mol/min per mol containing the purified enzymes and, as effec- ofenzyme. Prephenate dehydratase exhibited a tors, tyrosine and phenylalanine. The recon- turnover number of 350, which increased up to structed regulation system functioned as pre- 980 when the enzyme was activated by tyro- dicted from the kinetic data. sine. The turnover number for prephenate dehydrogenase could not be determined because LITERATURE CITED the enzyme had not been purified to homogene- 1. Baker, T. I. 1968. Phenylalanine-tyrosine biosynthesis ity. in Neurospora crassa. Genetics 58:351-359. most 2. Calhoun, D. H., D. L. Pierson, and R. A. Jensen. 1973. The sensitive response to even small Channel shuttle mechanism for the regulation of changes in effector concentration was shown by phenylalanine and tyrosine synthesis of a metabolic prephenate dehydratase. The activity was mul- branchpoint in Pseudomonas aeruginosa. J. Bacte- tivalently controlled by phenylalanine, tyro- riol. 113:241-251. 3. Catcheside, D. E. A. 1969. Prephenate dehydrogenase sine, and tryptophan. Prephenate dehydrogen- from Neurospora: feedback activation by phenylala- ase, however, was only feedback inhibited by nine. Biochem. Biophys. Res. Commun. 36:651-656. tyrosine and inhibited by its productp-hydroxy- 4. Champney, W. S., and R. A. Jensen. 1970. The enzy- phenylpyruvate (K, = 0.13 mM). The inhibition mology of prephenate dehydrogenase in Bacillus subtilis. J. Biol. Chem. 245:3763-3770. of tyrosine as reported for B. subtilis (4) and 5. Cleland, W. W. 1970. Steady state kinetics, p. 1-65. In Aerobacter aerogenes (7) was competitive with P. D. Boyer (ed.), The enzymes, vol. 2, 3rd ed. Aca- respect to prephenate (Ki = 0.06 mM) and non- demic Press Inc., New York. competitive with respect to NAD. Tyrosine, in 6. Cotton, R. G. H., and F. Gibson. 1965. The biosynthesis of phenylalanine and tyrosine; enzymes converting addition to inhibiting its own synthesis, stimu- chorismic acid into prephenic acid and their relation- lated phenylalanine formation manyfold at ex- ships to prephenate dehydratase and prephenate de- tremely low concentrations (8 AM). This was hydrogenase. Biochim. Biophys. Acta 100:76-88. achieved by an increased affinity of prephenate 7. Cotton, R. G. H., and F. Gibson. 1967. The biosynthesis of tyrosine in Aerobacter aerogenes: partial purifica- dehydratase for prephenate and a simultaneous tion of the protein. Biochim. Biophys. Acta 147:222- increase in maximal velocity. In the absence of 237. effectors, prephenate dehydrogenase had a 8. Davidson, B. E., E. H. Blackburn, and T. A. A. Do- more than 10-fold higher affinity for prephe- pheide. 1972. Chorismate mutase-prephenate dehydratase from Escherichia coli K-12. 1. Purification, nate (Km = 0.045 mM) than prephenate dehy- molecular weight, and amino acid composition. J. dratase (Km = 0.67 mM), thus allowing prephe- Biol. Chem. 247:4441-4446. nate to be channeled toward tyrosine. This ef- 9. Dayan, J., and D. B. Sprinson. 1971. Enzyme altera- 732 FRIEDRICH, FRIEDRICH, AND SCHLEGEL J. BACTZRIOL.

tions in tyrosine and phenylalanine auxotrophs of tase-prephenate dehydrogenase from Aerobacter aer- Salmonella typhimurium. J. Bacteriol. 108:1174- ogenes. Biochim. Biophys. Acta 212:387-395. 1180. 21. Koch, G. L. E., D. C. Shaw, and F. Gibson. 1971. The 10. DQpheide, T. A. A., P. Crewther, and B. E. Davidson. purification and characterization of chorismate mu- 1972. Chorismate mutase-prephenate dehydkntase tase.prephenate dehydrogenase from E. coli K-12. from Escherichia coli K-12. II. Kinetic properties. J. Biochim. Biophys. Acta 229:795-804. Biol. Chem. 247:4447-4452. 22. Levitzki, A., and D. E. Koshland, Jr. 1969. Negative 11. Friedrich, B., C. G. Friedrich, and H. G. Schlegel. cooperativity in regulatory enzymes. Proc. Natl. Purification and properties ofchorismate mutase-pre- Acad. Sci. U.S.A. 62:121-128. phenate dehydratase and prephenate dehydrogenase 23. Lingens, F. 1968. Die Biosynthese aromatischer Ami- from Alcaligenes eutrophus. J. Bacteriol. 126:712- nosiuren und deren Regulation. Angew. Chem. 722. 80:384-394. 12. Friedrich, B., and H. G. Schlegel. 1972. Die Hydroxyli- 24. Lingens, F., W. Goebel, and H. Uemeler. 1967. Regula- erung von Phenylalanin durch Hydrogenonwnas eu- tion der Biosynthese der aromatischen Aminosauren tropha H 16. Arch. Mikrobiol. 83:17-31. in Saccharomyces cerevisiae. 2. Repression, Induktion 13. Friedrich, B., and H. G. Schlegel. 1975. Aromatic und Aktivierung. Eur. J. Biochem. 1:363-374. amino acid biosynthesis in Alealigenes eutrophus H 25. Monod, J., J. Wyman, and J. P. Changeux. 1965. On 16. II. The isolation and characterization of mutants the nature ofallosteric transitions: a plausible model. auxotrophic for phenylalanine and tyrosine. Arch. J. Mol. Biol. 12:88-118. Microbiol. 103:141-149. 26. Pittard, J., and F. Gibson. 1970. The regulation of 14. Gibson, F. 1964. Chorismic acid: purification and some biosynthesis of aromatic amino acids and vitamins. chemical and physical studies. Biochem. J. 90:256- Curr. Top. Cell. Regul. 2:29-3. 261. 27. Rebello, J. L., and R. A. Jensen. 1970. Metabolic inter- 15. Gibson, F. 1968. Chorismic acid. Biochem. Prep. 12:94- lock. The multi-metabolite control of prephenate de- 97. hydratase activity in BaciUus subtilis. J. Biol. Chem. 16. Gibson, F., and J. Pittard. 1968. Pathways of biosyn- 245:3738-3744. thesis ofaromatic amino acids and vitamins and their 28. Schmit, J. C., S. W. Artz, and H. Zalkin. 1970. Choris- control in microorganisms. Bacteriol. Rev. 32:465- mate mutase-prephenate dehydratase. Evidence for 492. distinct catalytic and regulatory sites. J. Biol. Chem. 17. G6risch, H., and F. Lingens. 1974. Chorismate mutase 245:40194027. from Streptomyces. Purification, properties, and sub- 29. Schmit, J. C., and H. Zalkin. 1969. Chorismate mutase- unit structure of the enzyme from Streptomyces au- prephenate dehydratase. Partial purification and reofaciens Tti 24. Biochemistry 13:3790-3794. properties of the enzyme from SalmoneUa typhimu- 18. Jensen, R. A. 1969. Metabolic interlock. Regulatory rium. Biochemistry 8:174-181. interactions exerted between biochemical pathways. 30. Spr6aler, B., U. Lensen, and F. Lingens. 1970. Eigen- J. Biol. Chem. 244:2816-2823. schaften der Chorismat-Mutase aus Saccharomyces 19. Jensen, R. A., D. Naser, and E. W. Nester. 1967. cerevisiae S 288 C. Hoppe-Seyler's Z. Physiol. Chem. Comparative control in microorganisms. J. Bacteriol. 351:1178-1182. 94:1582-1593. 31. Weber, H. L., and A. BOck. 1970. Chorismate mutase 20. Koch, G. L. E., D. C. Shaw, and F. Gibson. 1970. from Euglena gracilis: purification and regulatory Studies on the subunit structure of chorismate mu- properties. Eur. J. Biochem. 16:244-251.