,&Ketoadipate. Enol-lactone Hydrolases I and II from Acinetobacter calcoaceticus*

step in the utilization of protocatechuate

P-Ketoadipate enol-lactone hydrolase catalyzes a common step in the utilization of protocatechuate and cis,cis-muconate by bacteria. Either of the two compounds elicits the synthesis of an enol-lactone hydrolase in Acinetobacter.
The enol-lactone hydrolase that is induced by each compound was purified, and the properties of the proteins were compared. Both enzymes appear to be dimers with molecular weights of approximately 25,000. The amino acid compositions of the enzymes differ, and the two proteins do not cross-react serologically.
The NH,-terminal amino acid residue of the protocatechuate-induced enol-lactone hydrolase (ELH I) is methionine and the NH,-terminal amino acid residue of the cis,cismuconate-induced enol-lactone hydrolase (ELH II) is proline. Therefore, ELH I and ELH II appear to be the products of different structural genes.
The serological specificity of ELH I and ELH II made it possible to demonstrate the mutually independent regulation of their synthesis in wild type cells and in constitutive mutant strains. The synthesis of ELH I is not impaired in mutant strains that cannot synthesize ELH II.
The rapid characterization of mutant strains that produce ELH I or ELH II constitutively was made possible by the development of pH indicator enzyme assays that were performed with toluenized cells. cis,truns-Muconate, which does not support the growth of Acinetobacter, elicits the synthesis of the enzymes that normally are induced by cis,cis-muconate to 20% of fully induced levels.
@-Ketoadipate enol-lactone hydrolase (ELH)' plays an essential role in the utilization of many aromatic growth substrates via the P-ketoadipate pathway in bacteria (2-4). The benzenoid growth substrates are metabolized to either protocatechuate or catechol (Fig. 1); a convergent series of reactions converts the diphenols to a common intermediate, P-ketoadipate enol-lactone, which is cleaved to /3-ketoadipate by the hydrolase (Fig. 1). Like the other enzymes of the fl-ketoadipate pathway, the hydrolase is inducible, and the mechanism of its induction varies among different groups of bacteria (5-12). In most Pseudomonas species the enzyme is induced by its product, P-ketoadipate, and the synthesis of the hydrolase is controlled coordinately with that of carboxymuconate-lactonizing enzyme and carboxymuconolactone decarboxylase (4,8,10,11). In representatives ofAcinetobacter calcoaceticus either of two metabolites can induce the synthesis of P-ketoadipate enol-lactone hydrolase (Fig. 1 conversion of protocatechuate to p-ketoadipyl-CoA. The other activity, termed ELH II, is induced by cis,cis-muconate in coordination with the enzymes that convert cis,cis-muconate to P-ketoadipyl-CoA (4-6, 13-15). The existence of two inducers for P-ketoadipate enol-lactone hydrolase and the coordinate control of P-ketoadipate enol-lactone hydrolase activity with two different regulatory units of enzyme synthesis strongly suggested that there might be two structural genes for P-ketoadipate enol-lactone hydrolase in A. calcoaceticus and that each structural gene may be under separate regulatory control (5, 6). Further evidence indicating independent regulation of the synthesis of two p-ketoadipate enol-lactone hydrolase proteins came from the isolation of constitutive mutant strains of A. calcoaceticus (13,15 but synthesized /3-ketoadipate enol-lactone hydrolase and the other enzymes that are induced by cis,cis-muconate at constitutive levels. The first studies of the physical properties of P-ketoadipate enol-lactone hydrolase in A. calcoaceticus suggested that ELH I and ELH II differ substantially and, therefore, supported the view that they are the products of different structural genes (5). Filtration of crude extracts through columns of Sephadex G-200 indicated that the molecular weight of ELH I is 18,500, whereas that of ELH II is 82,600 (5). Moreover, the two P-ketoadipate enol-lactone hydrolase activities differed markedly in their sensitivity to thermal inactivation; ELH I in crude extracts was stable to heating to 47", but ELH II lost more than one-half of its activity after 5 min at that temperature (5). The apparent differences in the physical properties of the two P-ketoadipate enol-lactone hydrolase proteins were not evident after they had been purified. Katagiri and Wheelis (16) showed that gel filtration of partially purified preparations of ELH I and ELH II gave molecular weights of 24,000 and 21,000, respectively. Furthermore, ELH II was shown not to be intrinsically thermolabile but rather to be rendered sensitive to thermal inactivation by a component of crude extract (16). Katagiri and Wheelis (16) concluded that the possibility that the two enzymes were the products of a single structural gene could not be excluded on the basis of the physical evidence Subsequently, Wheelis and Katagiri2 prepared antisera against ELH I and showed that the antisera failed to cross-react with ELH II. The absence of shared serological determinants supported the notion that the two proteins are the products of different structural genes. In this report we describe physical, chemical, and immunological studies that strongly favor the conclusion that the primary structures of ELH I and ELH II are coded by separate structural genes that are regulated independently. of strain ADPl were picked from succinate agar plates and placed in growth tubes containing 5 ml of growth medium supplemented with both 10 mM glucose and 2 rnM @-ketoadipate.
The cultures became turbid after 12 hours of shaking at 37". After 5 days a loopful from each culture was transferred to a growth tube containing growth 6570 medium with 2 mM fl-ketoadipate alone. In most cases growth occurred overnight; a loopful from each freshly grown culture was transferred to a tube containing 2 mM P-ketoadipate medium. Growth in this medium was always rapid, and after it occurred single colonies were isolated from the cultures on 10 mM succinate agar plates. A colony from each culture was picked and examined for the ability of the cells to grow on a 5 mM p-hydroxybenzoate agar plate. Cells from the colonies also were allowed to grow to stationary phase in 5ml cultures containing 10 mM glucose. These cells were harvested by centrifugation for 10 min at 5000 x g, washed twice with 0.85% NaCl, resuspended in 0.4 ml of 0.85% NaCl, and examined for the constitutive production of ci. Cells that produced the lactonizing enzyme constitutively turned the indicator from yellow to red within 1 hour. The interconversion of (+)-muconolactone and P-ketoadipate enollactone does not lead to a change in the pH of unbuffered medium, but the hydrolysis of the enol-lactone is accompanied by the release of a proton.
Therefore, the activity of muconolactone isomerase can be detected with a solution containing phenol red, (+)-muconolactone, and P-ketoadipate enol-lactone hydrolase.
The solution is buffered weakly with EDTA which inhibits the acid-releasing conversion of (+)-muconolactone to cis,cis-muconate by the lactonizing enzyme. The indicator assay mix contained the following in a volume of 1.0 ml: 4 pmol of (+)-muconolactone, 0. The enzyme was eluted from the column with the same buffer containing NaCl in a linear gradient from 0.07 to 0.16 M over 500 ml; fractions of 3 ml were collected. @-Ketoadipate enol-la&one hydrolase activity was eluted with a single peak of protein (Fig. 2); the specific activity of the enzyme was constant across the peak. Fractions containing the enzyme (Step 10,   single protein band when subjected to electrophoresis on acrylamide gels (Fig. 3). This evidence suggests that the protein preparations were homogeneous, but it should be noted that a single protein band was revealed after equal amounts of ELH I and ELH II were subjected to electrophoresis together (Slot 1, Fig. 3). Additional evidence for the homogeneity of the purified enzyme preparations came from the following observations: the specific activity of each protein was constant in fractions eluted from the second DEAE-cellulose column during purification (for example, Fig. 2); dansylation followed by hydrolysis of each protein yielded a single amino acid bearing a dansyl residue on its (Y amino group; and antisera prepared against each enzyme formed a single precipitin band when tested against extracts of approximately induced cells on Ouchterlony plates.

Molecular
Weight and Subunit Molecular Weight Determinations-ELH I and ELH II eluted in the same fraction after filtration on a Bio-Gel agarose A-l.5 column. As shown in Fig.  4, the elution volumes of the enzymes corresponded to a molecular weight of 25,000. The molecular weight of ELH II was estimated to be 26,000 by Robley C. Williams, Jr.,2 using meniscus depletion at sedimentation equilibrium in an ultracentrifuge.
Sodium dodecyl sulfate gel electrophoresis of either @ketoadipate enol-lactone hydrolase yielded three protein bands. The most strongly stained band corresponded to a molecular weight of 12,000 and two fainter bands corresponded *Unpublished observations. to molecular weights of 25,000 and 52,000 (Fig. 5). Therefore, it appears that the P-ketoadipate enol-lactone hydrolase exist primarily as dimers, with molecular weights of approximately 25,000 but that the proteins may associate to form higher oligomers.
Amino Acid Compositions-The amino acid compositions of the two /3-ketoadipate enol-lactone hydrolases are shown in Table V. Although the amino acid compositions generally are similar, there are some notable differences: ELH I subunits appear to have 3 more lysine and 4 more aspartyl residues than are present in subunits of ELH II, and ELH II subunits appear to contain 2 more phenylalanine residues than are present in subunits of ELH I. NH,-terminal Amino Acids-Dansylation followed by hydrolysis revealed that the NH,-terminal amino acid residues of ELH I and ELH II are methionine and proline, respectively.

Serological Properties of Enzymes
Studies with Purified Enzymes-Antibodies prepared against ELH I formed a precipitin band with this enzyme but not with ELH II. Similarly, antisera prepared against ELH II gave rise to precipitin bands with ELH II but not with ELH I. The independent interaction of each enzyme with antisera prepared against it is shown in Fig. 6. For this photograph both antibody preparations were placed in the center well of the Ouchterlony plate; the spurs formed between the wells containing ELH I and the wells containing ELH II show that each protein contains different antigenic determinants. Differences in the serological properties of ELH I and ELH II also were revealed by analysis of the inhibition of their activity by the immunoglobulin fraction of antisera prepared against them (Fig. 7). As shown on the left side of Fig. 7, the activity of ELH II was not influenced by concentrations of the anti-ELH I immunoglobulin fraction that inhibited completely the activity of ELH I. The reciprocal experiment, depicted on the right side of Fig. 7 Cells-The specificity of induction of ELH I and ELH II in wild type cells was demonstrated by analysis of the cross-reacting material formed by the organisms during growth with succinate, p-hydroxybenzoate, or benzoate. Extracts of succinate-grown cells did not contain significant P-ketoadipate enol-lactone hydrolase activity and did not form precipitin bands with antisera prepared against either ELH I or ELH II. Extracts of cells in which the enzymes of the protocatechuate pathway had been induced by growth with p-hydroxybenzoate contained material that cross-reacted with ELH I but not with ELH II. Extracts of benzoate-grown cells, which contained the enzymes of the catechol pathway, formed a precipitin band with antisera prepared against ELH II but not with antisera prepared against ELH I.
Further evidence for the independent genetic control of the synthesis of ELH I and ELH II came from the serological identification of the proteins formed by const,itutive mutant strains during growth with succinate. As shown in Fig. 8 ADP98 which are constitutive for enzymes of the catechol pathway do not form ELH I cross-reacting material (Fig. 8) and do form ELH II cross-reacting material (Fig. 9). ELH II cross-reacting material was not found in the extracts of the mutant strains (ADPG, ADP86, and ADP88) that formed constitutively enzymes of the protocatechuate pathway ( Fig.  9). Thus, the synthesis of ELH I appears to coincide precisely with the expression of genes for the protocatechuate pathway and the formation of ELH II is correlated strictly with the expression of genes for the catechol enzymes.
Growth in the presence of the inducer analog cis,transmuconate elicits the synthesis of material that cross-reacts with ELH II in the wild type strain ADPl (Fig. 10) "The value represents the sum of the free acid and the amide.
ketoadipate-succinyl CoA transferase at less than 10% of wild type levels when induced with cis,trans-muconate; the synthesis of other enzymes of the catechol pathway (catechol oxygenase, cis,cis-muconate-lactonizing enzyme, and muconolactone isomerase) appears to be unimpaired in these mutant strains. As shown in Fig. 10, the mutant strains do not form cross-reacting material for ELH II when induced with cis, trans.muconate.
Thus, the mutations in these organisms appear to prevent the expression of the structural gene for ELH II. That these mutations do not impair the synthesis of ELH I is shown by the cross-reactions depicted in Fig. 11: p-hydroxybenzoate-grown (protocatechuate-induced) cultures of the three mutant strains contain material that forms a precipitin band with antisera prepared against ELH I. Therefore, mutations that render the ELH II structural gene dysfunctional do not appear to interfere with the expression of the ELH I structural gene.

DISCUSSION
The amino acid compositions of ELH I and ELH II differ by about 12 residues per subunit, corresponding to a minimum difference of 6 residues in the amino acid sequences of the proteins. Therefore the conclusion that the enzymes are the products of separate structural genes appears to be justified. Additional evidence showing that the proteins possess different structures comes from the failure of the enzymes to cross-react serologically and from the observation that the amino terminal amino acid of ELH I is methionine, whereas the amino terminal amino acid of ELH II is proline. Heretofore the inference that ELH I and ELH II are the products of separately regulated structural genes has rested almost entirely on measurements of the activity of the enzymes in crude extracts of wild type and mutant cells. The immunological specificity of the two proteins allowed them to be identified individually in crude extracts and thus permitted a more precise analysis of the factors that govern the synthesis of FIG. 6. Ouchterlony double diffusion plate. The center well received 0.1 ml of antisera prepared against ELH I and 0. constitutively formed material that cross-reacted with ELH I but not with ELH II. Therefore, the ELH activity associated with the expression of genes for the protocatechuate enzymes appears to be due primarily to ELH I. In contrast, the /3-ketoadipate enol-lactone hydrolase activity that is elicited by the expression of genes for the catechol enzymes seems to be almost entirely attributable to ELH II. Material that crossreacted with ELH II but not with ELH I was found in extracts of wild type cells in which enzymes of the catechol pathway had been induced. In addition, uninduced cultures of mutant strains that form enzymes of the catechol pathway constitutively contained material that cross-reacted with ELH II but not with ELH I.
Additional evidence for the independent control of the expression of the structural genes for ELH I and ELH II was provided by the properties. of strains ADP17, ADP23, and ADP27. These organisms differ from the wild type in that they do not synthesize ELH II when grown in the presence of cis,trans-muconate; they appear to have undergone mutations with pleiotropic effects because they also are impaired in their ability to form fi-ketoadipate-succinyl-CoA transferase when induced with cis,trans-muconate.
Whatever the nature of the mutations that prevent the synthesis of ELH II, they do not appear to impair the synthesis of ELH I when the mutant strains are grown with p-hydroxybenzoate.
The physical properties of the purified preparations of ELH I and ELH II are quite similar: the proteins appear to be dimers with a molecular weight of approximately 25,000. The presence of bands corresponding to proteins with molecular weights of 52,000 after sodium dodecyl sulfate gel electrophoresis (Fig. 5) suggests that under some circumstances the fi-ketoadipate enol-lactone hydrolase may aggregate to form oligomers higher than dimers. Canovas and Stanier (5) found that the molecular weights of ELH I and ELH II corresponded to 18,500 and 82,500, respectively, when estimations were made by the filtration of extracts of appropriately induced Acinetobacter cultures through Sephadex gels. Our results obtained with the purified proteins more closely resemble those of Katagiri and Wheelis (16)  24,000 and 21,000 for ELH I and ELH II, respectively. Katagiri from sodium dodecyl sulfate gel electrophoresis indicate that and Wheelis (16) suggested that the discrepancy between their under some circumstances the P-ketoadipate enol-lactone results and those of CBnovas and Stanier (5) might be due to hydrolase proteins may aggregate to form higher oligomers and differences in the state of aggregation of the @-ketoadipate thus lend support to the suggestion of Katagiri   (Fig. 10).