Studies on Tyrosine Phenol Lyase MODIFICATION OF ESSENTIAL HISTIDYL RESIDUES BY DIETHYLPYROCARBONATE

Tyrosine phenol-lyase of Escherichia intermedia is inactivated by treatment with diethylpyrocarbonate at pH 6.0 AND 4 degrees. Spectrophotometric studies show that the inactivation is stoichiometric, with a modification of 2 histidyl residues per molecule of the enzyme. Finding that this inactivation is largely reversed by treatment with hydroxylamine indicates that the inactivation is mainly due to modification of the histidyl residues. No changes in the sulfhydryl content or in the aromatic amino acids are observed as a result of this modification. The modified tyrosine phenol-lyase retains most of its ability to form a nearly normal complex with its coenzyme, pyridoxal phosphate. This has been shown by studies of its absorption, by the determination of pyridoxal phosphate, and by reduction of the holoenzyme with tritiated sodium borohydride. The modified enzyme also appears to form a Schiff base intermediate with L-alanine. The modified holoenzyme fails to catalyze the exchange of the alpha-hydrogen of L-alanine with tritium from tritiated water. This is consistent with a catalytic role for modified histidyl residues at the active site of the enzyme; this role is the removal of the alpha-hydrogen of substrates.

Spectrophotometric studies show that the inactivation is stoichiometric, with a modification of 2 histidyl residues per molecule of the enzyme. Finding that this inactivation is largely reversed by treatment with hydroxylamine indicates that the inactivation is mainly due to modification of the histidyl residues. No changes in the sulfhydryl content or in the aromatic amino acids are observed as a result of this modification. The modified tyrosine phenol-lyase retains most of its ability to form a nearly normal complex with its coenzyme, pyridoxal phosphate. This has been shown by studies of its absorption, by the determination of pyridoxal phosphate, and by reduction of the holoenzyme with tritiated sodium borohydride. The modified enzyme also appears to form a Schiff base intermediate with L-alanine.
The modified holoenzyme fails to catalyze the exchange of the o-hydrogen of L-alanine with tritium from tritiated water. This is consistent with a catalytic role for modified histidyl residues at the active site of the enzyme; this role is the removal of the a-hydrogen of substrates.
Tyrosine phenol-lyase is an enzyme which catalyzes the stoichiometric conversion of L-tyrosine to pyruvate, ammonia, and phenol, and requires pyridoxal phosphate as a cofactor (l-3). A crystalline preparation of the enzyme was prepared in our laboratory from cells of Escherichia intermedia grown in a bouillon medium supplemented with L-tyrosine (4, 5). The enzyme has a molecular weight of 170,000 and binds with 2 mol of pyridoxal phosphate per mol of enzyme.
Although n-alanine is not a substrate of tyrosine phenol-lyase, it does act as a competitive inhibitor of tyrosine degradation by the enzyme (4). L-Alanine cannot undergo p elimination to form an EA species (Scheme I), but it can proceed through ES to an EX species (13). This quasi-substrate was used in this study to investigate the effect of modification on the enzyme. This paper reports that the modification of 2 histidyl residues inactivates tyrosine phenol-lyase without greatly reducing its ability to combine with pyridoxal phosphate and with a competitive inhibitor L-alanine. This is of interest since it has been suggested that histidyl residues may have catalytic roles in pyridoxal-catalyzed reactions (14, 15) and in the reaction catalyzed by the pz subunit of tryptophan synthetase (16).
The chemical agent used in these studies has been variously called diethylpyrocarbonate, ethoxyformic anhydride, and diethyloxydiformate.
Although it is a general acylating agent, it was first shown to be specific for histidyl residues below pH 7 from model studies (17). The formation of N-carbethoxyhistidyl side chains in proteins results in an increase in absorption between 220 and 260 nm (18). The histidyl derivative is much more stable than is N-acetyl imidazole, but it is decomposed by hydroxylamine (19). Several enzymes have been reversibly modified by this reagent (20-23).
However, Melchior and Fahrney (19) found that the reagent can also react irreversibly with amino groups in ribonuclease and with the active site serine in chymotrypsin.

EXPERIMEKTAL PROCEDURE
Materials-Crystalline tyrosine phenol-lyase was prepared from cells of Escherichia intermedia A-21 grown on a tyrosine-supplemented medium, as previously described (4). Diethylpyrocarbonate was purchased from Tokyo Kasei Kogyo.
Pyridoxal 5'-phosphate was kindly provided by Dainippon Pharmaceutical Co. Other chemicals used in this work were commercial products. Enzyme Assay-a,0 Elimination reaction was assayed by meas-uring the amount of pyruvate liberated from n-tyrosine under conditions described in a previous paper (4). One unit of activity was defined as the amount of enzyme catalyzing the formation of 1 pm01 of pyruvate per min under standard assay conditions. n-Tgrosine formation was measured by determining the incorporation of [Wlphenol into n-tyrosine: The reaction mixture contained 75 urn01 of nvruvate: 37.5 vmol of ammonium sulfate: 0.1 pmol of pyridoxal-phosphate; 50 imol of NHa-NH&l buffer; pH 9.25; 12.5 pmol of [W]phenol (185,000 cpm); and 0.05 unit of crystalline enzyme in a total volume of 1.0 ml. The reaction was carried out for 30 min at 30" and was terminated by the addition of 0.2 ml of 30yo trichloroacetic acid solution.
The denatured enzyme was removed by brief centrifugation, and 4.0 ml of toluene were added to 1.0 ml of the supernatant solution.
This extraction was repeated five times, then 0.5 ml of water layer was added with 10 ml of Bray's scintillation solution, and the whole was counted in a model 2002 liquid scintillation spectrometer (Packard). The assay of the p replacement reaction was carried out by determining the incorporation of [Wlphenol into n-tyrosine. The reaction mixture of 1 ml contained 200 rmol of n-serine; 0.1 pm01 of pyridoxal phosphate; 12.5 pm01 of [Wlphenol; 50 pmol of potassium phosphate buffer, pH 7.8; and the crystalline enzyme. The reaction was carried out at 30" for 30 min and was terminated by the addition of 0.2 ml of 30yo trichloroacetic acid solution. The remaining phenol was extracted with toluene, as above, then the radioactivity in 0.5 ml of the aqueous layer was counted as described above.

Protein
Determination-The protein determination of tyrosine phenol-lyase was performed spectrophotometrically by measuring the absorbance at 280 nm. An E value of 8.37 for 10 mg per ml and for a l-cm light path was used throughout (4).

Spectrophotometric
Determination-Spectrophotometric determinations and the recording of absorntion spectra were carried out with a Beckman model-DB-G recording-spectrophotometer and a Hitachi model 124 snectronhotometer. Treatments with Diethyliyrocarbonate and with Hydroxylamine-Diethylpyrocarbonate was freshly diluted to 0.08 M with cold ethanol for each experiment.
Protein samples were prepared as a solution of about 1 mg per ml in 0.1 M potassium phosphate buffer, pH 6.0; they were then dialyzed against the same buffer for 16 hours to remove p-mercaptoethanol.
The reaction with diethylpyrocarbonate was run for 30 min in an ice bath or in a Hitachi 124 spectrophotometer cooled with circulating water at 0". The rate of formation of carbethoxyhistidine was followed by measuring the difference in absorbance at 240 nm between the treated sample and a control enzyme which had been treated with an equal volume of cold ethanol: The final concentration of ethanol in each reaction svstem was under 5.0%. The number of histidine residues undergoing modification was calculated using a value of 3200 M-l cm-i for the molar extinction coefficient at 242 nm of the N-carbethoxyhistidyl residues in proteins (18). Aliquots of reaction mixtures with diethvlnvrocarbonate and untreated controls were incubated in 0.1 M potassium phosphate buffer, pH 7.0, containing 1 M NHzOH (adjusted to pH 7.0 with HCl), 10 mM p-mercaptoethanol, and 1 mM EDTA for 15 hours at 20". The solutions were dialyzed for 6 hours against four changes of 0.1 M notassium ohosnhate buffer. nH 8.0. containing 10 mM P-mercaptoethanol and l&rnM EDTA before assay. Oneiolution was similarly incubated and dialyzed except that NH20H was omitted (untreated control). Determination of Total Sulfhydryl Residues-Solutions of tyrosine phenol-lyase were diluted with 11 volumes of 9.6 M urea in 0.1 G potassium phosphate, pH 7.0, containing 0.1 mM EDTA. The absorbance at 412 nm of each solution and of a reagent blank was recorded in a Beckman spectrophotometer before and after the addition of 0.2 mM 5,5'-dithiobis (2-nitrobenzoic acid). The number of sulfhydryl groups undergoing change was calculated from the molar extinction coefficient (E4i2 nm = 13,600 M-I cm-l) (24).
Reduction with Sodium [3H]Borohydride-Tritiated sodium borohydride was dissolved in 0.001 N NaOH to give a 0.1 M solution. The enzyme was reduced by 2 kcLI of NaB3H4 solution at room temperature for 30 min, then it was dialyzed against four changes of 0.1 M potassium phosphate buffer, pH 8.0. Aliquots of protein were mixed with 10 ml of Bray's scintillator and were counted in a liquid scintillation spectrometer. Measurement of Tritium Incorporation into .z-Alanine-The in-corporation of tritium into n-alanine from tritiated water was measured according to the method of Morino and Snell (25). A reaction mixture containing 50 Nmol of n-alanine; 0.1 ml of 'aHzO (2 X 10' cpm); 2 mg (3.8 units) of crystalline apoenzyme; 10 pmol of potassium phosphate, pH 8.0: 0.25 rrmol of a-mercantoethanol, and 0.05 pmol ofpyridoxal phosphate in a total volume of 0.5 ml was incubated at 30' for various periods. At appropriate time intervals, 0.1 ml of 30%.trichloroacetic acid was added to the reaction mixture.
The denatured enzyme was removed by a brief centrifugation, and 0.5 ml of the supernatant solution was then added to 5.0 ml of acetone at 0". After standing for 10 min at 0", the precipitated n-alanine was collected by centrifugation. This procedure was repeated five times and the precipitated n-alanine was dried in a boiling water bath. This dried n-alanine was dissolved in 0.6 ml of water, and a 0.5-ml portion was then mixed with 10 ml of Bray's scintillator and counted in a liquid scintillation spectrometer.
Another 50-~1 portion was diluted with 4.95 ml of water; then 100 ~1 of the solution was subjected to the n-alanine assay using the ninhydrin method (24).

RESULTS
Inactivation by Diethylpyrocarbonate- Fig.  1 shows that tyrosine phenol-lyase is inactivated by rather low concentrations of diethylpyrocarbonate at pH 6.0 and that pyridoxal phosphate has no effect on the extent of its inactivation.
n-Alanine has shown a protective effect at a relatively high concentration.
Studies of the effects of diethylpyrocarbonate on the spectrum of the enzyme were carried out on the apoenzyme to avoid any changes which might occur in the spectrum of the bound pyridoxal phosphate on modification of the enzyme. Fig. 2A shows a spectrum of the apoenzyme before and after treatment with 1.0 mM diethylpyrocarbonate for 60 min, and a difference spectrum during the reaction of enzyme solutions.
No changes in the spectra were detected above 270 nm. This indicates that no modification of the tyrosyl residue had occurred since 0-carbethoxytyrosine absorbs between 270 and 280 nm (17). Large increases in absorbance at 242 nm were observed as shown in Fig. 2, A and B. This absorbance is characteristic of the N-carbethoxyhistidyl residues in proteins (18). The number of histidyl residues modified is shown in Fig. 2C  Reaction mixture contained in a final volume of 1.0 ml, 1.1 mg of apoenzyme, potassium phosphate buffer, pH 6.0 (20 pmol) and diethylpyrocarbonate (0 to 1 pmol) at the indicated concentrations.
After incubation for 30 min at O", aliquots (0.01 ml) of the reaction mixtures were diluted with 3.0 ml of the standard tyrosine phenol-lyase assay mixture (see "EXperimental Procedure") to which 1 mM fl-mercaptoethanol had been added and were subsequently assayed after addition of 1 ml of 2.5 mM n-tyrosine.
holoenzyme + n-alanine. Extrapolation of L-tyrosine phenol-lyase activity to zero corresponds to the carbethoxylation of 2 histidyl residues per mol of apoenzyme.
The reasonably good stoichiometry between the loss of activity and the modification of 2 histidyl residues no longer holds when less than 20% of the initial activity remains; complete loss of activity is associated with the modification of 8 or more histidyl residues.
Tyrosine phenol-lyase modified to the extent of 20% of the residual activity (3 histidyl residues modified per mol of enzyme) was used as a reasonable sample of modified enzyme in this work.
Various activities of the enzyme, modified to the extent of about 3.0 histidyl residues per mol of enzyme, are compared. Both cr,p elimination and its reverse reaction are reduced to 15 to 20% of the control, whereas activity in the /3 replacement reaction is 6% of the control. c 2 00-Y a F 60z Log E 20-Reversal of Inactivation by Hydroxylamine- Melchior and Fahrney (19) found that 0.5 M hydroxylamine, pH 7.0, removes the N-carbethoxy group from imidazole in several minutes.
Several diethylpyrocarbonate-inhibited enzymes have also been found to be reactivated by hydroxylamine (16,19,(21)(22)(23)). Fig. 3 shows that the enzyme, in which 3 or fewer histidyl residues have been modified, can be largely reactivated by treatment with hydroxylamine.   I Effect of NaBaH, treatment on unmodi$ed and diethylpyrocarbonatemodi$ed enzyme in presence or absence of L-ala&e Untreated apoensyme (1.1 mg per ml) and diethylpyrocarbonate-modified apoenzyme (1.2 mg per ml of 20% residual activity) were dialyzed against 0.1 M potassium phosphate buffer, pH 8.0, containing 0.1 mM pyridoxal phosphate, 0.1 mM EDTA, and 0.5 mM P-mercaptoethanol. Unmodified apoenzyme control which had been dialyzed against buffer from which pyridoxal phosphate had been omitted, was also treated with NaBaHd. A control which had not been treated with sodium borohydride was similarly dialyzed. PLP, pyridoxal phosphate. phosphate, 0.1 mM EDTA, and 0.5 mM &mercaptoethanol. Tyrosine phenol-lyase activity of modified and control enzyme solutions was assayed after 1 and 7 days, and a slight restoration of activity was found (Fig. 3). The modified enzyme could still be reactivated to 88% activity by treatment with hydroxylamine after 7 days of dialysis. We concluded that the modified enzyme is stable, for at least a few days, under these conditions and so we used modified enzymes after a dialysis of 12 hours in this study.
Binding of Pyridoxal Phosphate to Modijied Tyrosine Phenollyase-The amount of pyridoxal phosphate bound by modified apoenzyme was determined after dialysis of the enzyme, 1.91 mg of protein, against 0.1 M potassium phosphate buffer, pH 8.0, containing 5 pM pyridoxal phosphate. After 16 hours, the concentration of pyridoxal phosphate inside and outside the dialysis bag was determined. An excess concentration of pyridoxal phosphate was found within the dialysis bag, which corresponded to the binding of 1.8 mol of pyridoxal phosphate by 170,000 g of the modified apoenzyme. The untreated enzyme bound 2.2 mol of coenzyme under the same conditions.

E$ect of Modification by Diethylpyrocarbonate on Suljhydryl
Content-The total sulfhydryl content of the enzyme solutions was determined as described under "Experimental Procedure." Untreated tyrosine phenol-lyase and tyrosine phenol-lyase treated with diethylpyrocarbonate, and having 22% residual activity, each contained 9.8 sulfhydryl residues per molecule of enzyme.

E$ect of Modification by Diethylpyrocarbonate on Absorption
Spectra of Enzyme- Fig. 4 shows the absorption spectra between 300 and 550 nm of the untreated and diethylpyrocarbonate modified enzyme (30% residual activity) which had been dialyzed against buffer containing 0.1 mM pyridoxal phosphate. The spectra of the two proteins are similar, but the heights of the 340 nm and 430 nm peaks for the modified enzyme are, respectively, only 90% and 80% the heights of the corresponding peaks for untreated protein. The addition of L-alanine caused increased absorbance at 500 nm in the control enzyme, while in modified enzyme it gave proportional height to the residual activity. The increase in absorbance at 430 nm on the addition of pyridoxal phosphate to the enzyme has been reported and attributed to the formation of an enzyme-bound pyridoxal phosphate Schiff base. The species absorbing at 500 nm represents a deprotonated enzyme-n-alanine complex at the a-carbon of n-alanine (EX + EX' in Scheme I). Modified enzyme, though it can bind with pyridoxal phosphate nearly normally, fails to produce the intermediate which has an absorption maximum at 500 nm. Most of the absorbance at 500 nm in modified enzyme is due to unmodified enzyme present.
Sodium Borohydride Reduction of Native Enzyme and Modijied Enzyme-Sodium borohydride reduces the Schiff base linkage between pyridoxal phosphate and the e-amino group of a lysyl residue in the enzyme (5). When the holoenzyme is treated with small amounts of sodium borohydride, the absorbance at 430 nm decreases to about 90% of the initial value, then decreases no further unless a large excess of sodium borohydride is used. We found that the minimal amount of sodium borohydride required to reduce the absorbance of the enzyme by 80%, has very little effect on the spectrum of the enzyme in the presence of n-alanine. L-Alanine significantly protects the activity of the enzyme and decreases the incorporation of tritium into the enzyme (Table I). Thus, the enzyme-bound pyridoxal phosphate-alanine Schiff base linkages seems to be more resistant to sodium borohydride reduction than is the enzyme-pyridoxal phosphate Schiff base linkage. Table I shows that sodium borohydride also reduced the pyridoxal phosphate bound to the diethylpyrocarbonate-modified enzyme which resulted in the incorporation of the tritium label. n-Alanine decreased the incorporation of tritium into the protein by 60%. These data provide evidence for the formation of a pyridoxal phosphatealanine Schiff base by the modified enzyme (ES in Scheme I).
E$ect of Modi$cation by Diethylpyrocarbonate on Proton Exchange Activity of Tyrosine Phenol-lyase-The first step in the tyrosine phenol-lyase catalyzing cy , B elimination reaction is labilization of the a-hydrogen atom of an amino acid (ES 4 EX, in Scheme I). Since studies on the modified enzyme based on absorption spectra and reduction with sodium borohydride support the binding of the modified enzyme with the coenzyme and Lalanine, the activity of the enzyme for exchanging the ac-hydrogen of n-alanine was measured in tritiated water. The rates of tritium incorporation into L-alanine by the modified and control enzymes are shown in Fig. 5. The modified enzyme exhibits only 6% of the tritium exchange activity of native tyrosine phenol-lyase.
This suggests that modification of the enzyme with diethylpyrocarbonate inhibits the ability of the enzyme to labilize the a-hydrogen of an amino acid.
Studies on Absorption Spectra of Diethylpyrocarbonate-treated Enzyme in Reversal Reaction-Degradation of L-tyrosine to phenol, pyruvate, and ammonia by tyrosine phenol-lyase is readily reversible at high concentrations of pyruvate and ammonia (10). The addition of phenol, pyruvate, and ammonia to the holotyrosine phenol-lyase resulted in the appearance of a spectral band near 500 nm (Fig. 6) (10) similar to that observed when n-alanine was added to the enzyme.
This absorption peak can be attributed to the deprotonated substrate enzyme intermediate, EX or EX' in Scheme I. The absorption exhibited in the reversal re-action at 500 nm by the native enzyme is changed with time, increasing at first then decreasing slowly as shown in Fig. 6. The same absorption band was shown by diethylpyrocarbonatetreated tyrosine phenol-lyase.
In contrast to a similar peak formed by the native enzyme, this band was high and showed slight change in absorbance after long standing (Fig. 6).

DISCUSSION
Tyrosine phenol-lyase contains pyridoxal phosphate bound in a Schiff base linkage to an amino group of the lysyl residue (5) (Scheme I, E). This amino group is displaced by the substrate to form a second Schiff base (Scheme I, ES). The first catalytic step in all parts of the cx , /3 elimination reaction is the removal of the a-hydrogen from the amino acid to give an intermediate represented as EX in Scheme I. Although this step is greatly facilitated by electron withdrawal through the conjugated system of pyridoxal, it probably requires the presence of a basic group on the protein to accept the hydrogen.
Just such a role for a his- . Absorption spectra of untreated and diethylpyrocara bonate-treated enzyme in the presence of pyruvate, ammonium, and phenol. Spectra of enzyme solution containing 17.6 mg of protein in 0.1 M potassium phosphate buffer, pH 8.0, containing 0.1 mM pyridoxal phosphate, 0.5 mM p-mercaptoethanol, and 0.1 mM EDTA were recorded at 30" against the same buffer blank after the addition of ammonium, pyruvate, and phenol to a final concentration of 75, 37.5, and 12.5 mM, respectively, in 1.0 ml of total volume. The modified enzyme had been previously treated with diethylpyrocarbonate until 77% inhibition and then dialyzed against the above buffer. tidy1 or a lysyl residue was proposed for transamination reactions on the basis of model (14) and enzymatic (26) studies.
Recently, Miles and Kumagai (16) have investigated the essential histidyl residues in the pz subunit of tryptophan synthetase which catalyzes the ar,/l elimination and /3 replacement reactions of amino acids (28). It is difficult to check the exchange of the cr-hydrogen of substrates by the pZ subunit of tryptophan synthetase to determine the role of modified histidine, since ex-traction of the hydrogen from the substrate is the rate-determining step in the cu,p elimination reaction by tryptophan synthetase (29). Thus, substrates which exchanged an a-hydrogen with labeled Hz0 could not be found in the reaction mixture. With tyrosine phenol-lyase, the proton exchange reaction is readily measured using L-alanine in tritium labeled water. Incorporation of the proton into the o position of L-alanine takes place during the reaction EX -ES -E + S in Scheme I. Tyrosine phenol-lyase exhibits an intense absorption band at 599 nm in the presence of L-alanine. Similar peaks near 500 nm have been observed in many pyridoxal phosphate-dependent enzymes (25, 30) and have been ascribed to the deprotonated intermediate, EX, in Scheme I, or to a species in equilibrium with this intermediate, EX' (25). L-Alanine was used throughout this study to investigate the role of modified histidine residues because of its convenient properties described above.
Current studies show that inactivation of tyrosine phenollyase by diethylpyrocarbonate is stoichiometric with modification of 2 histidyl residues per mol of the enzyme, and inactivation is largely reversed by hydroxylamine. Reduction of the rate of inactivation by r,-alanine and pyridoxal phosphate indicates that the histidine residue is located in the active site region. Although these data support the specific modification of the essential histidyl residue at the active site, they do not eliminate the possibility of carbethoxylation of active amino groups which might exist in the enzyme. The e-amino group of lysine at the binding site of pyridoxal phosphate does not seem to be damaged to an extent that would explain the loss of enzyme activity because pyridoxal phosphate did not protect the modification and the modification did not greatly affect the absorption at 430 nm.
The modified tyrosine phenol-lyase binds pyridoxal phosphate nearly stoichiometrically and in an environment almost similar to that of the untreated enzyme. This has been shown by a determination of the pyridoxal phosphate bound to the enzyme using the absorption spectrum, and by reduction with tritium-labeled sodium borohydride.
Interaction of the untreated and modified enzyme with L-alanine has been investigated using absorption spectra and by treatment with sodium borohydride, in the presence and absence of n-alanine. The modified enzyme exhibits about 83% of the absorbance shown by the native enzyme at 430 nm. The ability of the modified enzyme to bind with pyridoxal phosphate and to make a complex with L-alanine is less than that of unmodified enzyme, though loss of this ability is not enough to explain the loss of enzymatic activity. These differences between the modified and unmodified enzyme may be caused by a slight change in the conformation of the modified protein at the active site by excess modification of the histidyl residues, or by partial modification of amino groups at the coenzyme binding site.
The modified enzyme has shown only 6% of the tritium exchange activity of the control enzyme. Then, the histidyl residue which has been modified may have a catalytic role in the exchange of the a-proton of L-alanine. The absorption spectrum of the modified enzyme in the presence of pyruvate, ammonia, and phenol shows a remarkable accumulation of the EX + EX' species, i.e. the blockage in the addition step of the a-proton. These results obtained from the studies on absorption spectra and on the tritium exchange reaction suggest that modification blocks the conversion of ES to EX in tyrosine degradation reaction and the conversion of EX to ES in the reversal reaction.
These studies show that histidyl modification by diethylpyrocarbonate results in a modified enzyme which can form nearly normal complexes with pyridoxal phosphate and L-alanine. A role for this modified histidyl residue, in the abstraction of the a-hydrogen of the substrate has been proposed.
Recently, it was reported by Miles that an essential, photosensitive histidyl residue is present in the pyridoxal phosphatebound peptide in the fiz subunit of tryptophan synthetase (31).