JOURNAL OF BACTERIOLOGY, Mar., 1966Copyright @ 1966 American Society for Microbiology

Vol. 91, No. 3Printed in U.S.A.

Synthesis of the Enzymes of the Mandelate Pathwayby Pseudomonas putida

I. Synthesis of Enzymes by the Wild Type

G. D. HEGEMAN

Department ofBacteriology, and Immunology, University of California, Berkeley, California

Received for publication 26 October 1965

ABSTRACTHEGEMAN, G. D. (University of California, Berkeley). Synthesis of the enzymes

of the mandelate pathway by Pseudomonas putida. I. Synthesis of enzymes by thewild type. J. Bacteriol. 91:1140-1154. 1966.-The control of synthesis of the fiveenzymes responsible for the conversion of D(-)-mandelate to benzoate by Pseu-domonas putida was investigated. The first three compounds occurring in the path-way, D(-)-mandelate, L( +)-mandelate, and benzoylformate, are equipotentinducers of all five enzymes. A nonmetabolizable inducer, phenoxyacetate, alsoinduces synthesis of these enzymes; but, unlike the metabolizable inducer-substrates,it does not elicit synthesis of enzymes that mediate steps in the pathway beyondbenzoate. Under conditions of semigratuity, DL-mandelate elicits immediatesynthesis at a steady rate of the first two enzymes of the pathway, but two enzymes

which act below the level of benzoate are synthesized only after a considerable lag.Succinate and asparagine do not significantly repress the synthesis of the enzymesresponsible for mandelate oxidation.

The remarkable variety of organic compoundsthat can serve as sole sources of carbon andenergy for fluorescent pseudomonads wasdemonstrated by the nutritional studies of denDooren de Jong (Thesis, Technische Hogeschoolte Delft, Delft, The Netherlands, 1926). Bio-chemical investigations on this group of bacteria,begun some 20 years later, revealed that the attackon many organic substrates occurs throughspecific metabolic pathways of considerablebiochemical complexity, and is typically mediatedby inducible enzyme systems. The mechanismsfor the oxidation of aromatic substrates havebeen subjected to particularly detailed biochemi-cal analysis. Figure 1 shows the convergent path-ways operative in the oxidation of L-tryptophan,D-mandelate, and p-hydroxybenzoate. A total of20 specific enzymes, all inducible, are required toconvert the benzenoid nucleus of these threeprimary substrates to succinate and acetyl-coenzyme A (CoA).

Stanier (18), Suda, Hayaishi, and Oda (21),and Stanier and Sleeper (20) explored by mano-metric methods the patterns of induction elicitedby compounds that are members of these path-ways. In general, exposure to any given inducer-substrate induced cells for the immediate oxida-tion of later intermediates in that specific path-

way; but if the inducer-substrate in question was ametabolic intermediate, it did not induce oxida-tion of earlier compounds in the reactionsequence. Cross-induction of sequences inparallel, convergent pathways was not observed.Such findings led Stanier (18), Suda, Hayaishi,and Oda (21), and Karlsson and Barker (12) topropose independently an explanation of themechanism of these complex inductions, which isnow known as the hypothesis of sequentialinduction. This hypothesis proposes that eachinducer-substrate in such pathways triggers thespecific synthesis of the enzyme responsible forits conversion to the next intermediate of themetabolic pathway; induction thus occurs instepwise fashion, being metabolically (andpresumably also temporally) sequential.The specific evidence for sequentiality in any

one of these pathways was relatively fragmentary,but the hypothesis was widely accepted at thetime, mainly because there was no obviouscounter-hypothesis. Since then, many examplesof coordinate inductions have been discovered inother bacteria. Complete inductive coordinatenessin the pathways of aromatic oxidation by fluo-rescent pseudomonads can be definitely elimi-nated on the basis of existing data. However, the

1140

on January 2, 2019 by guesthttp://jb.asm

.org/D

ownloaded from

SYNTHESIS OF MANDELATE PATHWAY ENZYMES. I

D-MANDELIC ACID

L-MANDELIC ACID

BENZOYLFORMIC ACID L-TRYPTOPHAN

BENZALDEHYDE L-FORMYLKYNURENINE

BENZOIC ACID L-KYNURENINE

P-HYDROXYBENZOIC ACID CATECHOL * ANTHRANILIC ACID + L-ALANINE

PROTOCATECHUIC ACID CIS, CIS-MUCONIC ACID

/8-CARBOXY, CIS, CIS -MUCONIC ACID MUCONOLACTONE

y-CARBOXYMUCONOLACTONE * ENOL-LACTONE

46-KETOADIPIC ACID

SUCCINYL-CoAt

SUCCINATE

tKEoiP-KETOADIPYL-CoAt + ~~CoASH

ACETYL-CoAFIG. 1. Convergent pathways for the oxidation of three aromatic compounds by Pseudomonas putida.

possibility of a coordinate induction of smallgroups of enzymes, interspersed with sequentialinductions, is by no means excluded. Indeed, therecent analysis by Palleroni and Stanier (16) ofthe synthesis of the early enzymes in the trypto-phan pathway by Pseudomonas fluorescens Tr-23showed the synthesis of tryptophan pyrrolase andformylkynurenine formamidase to be coordinate.The work reported here was undertaken to

determine the mechanism for control of synthesisof the enzymes of the mandelate pathway in P.putida A.3.12. The enzymology of this reactionsequence has been studied in some detail (8, 11,15). It is a sequence of considerable complexity, atotal of 12 specific inducible enzymes beingrequired to convert D-mandelate to succinate andacetyl-CoA (Fig. 2). The first five enzymes,required to convert D-mandelate to benzoate, willbe termed the mandelate group; those whichconvert catechol to f,-ketoadipate will be termedthe catechol group.

MATERIALS AND METHODS

Organism and methods of cultivation. P. putida,strain A.3.12 (ATCC 12633), was used as biologicalmaterial. This strain has been previously designatedas P. fluorescens in the studies of Stanier and others(8, 14, 15, 17-20).

All the media to be described below were preparedin a mineral base which contained (per liter): 200 mgof nitrilotriacetic acid, 580 mg of MgSO4c7H20, 67mg of CaCl2-2H20, 0.2 mg of (NH4)6Mo7024.4H20,2.0 mg of FeSO4-7H20, 1 ml of Hutner's "metals 44"

(3), 1.0 g of (NH4)2SO4, and 0.025 mole each ofKH2PO4 and Na2HPO4. This provided all the knowntrace elements, nitrogen, sulfur, and phosphorus. Thefinal pH was 6.85. In the kinetic experiments to bedescribed, the concentration of phosphate buffer wasdoubled. For determination of viable count, checks onpurity, and routine maintenance, 1.0% (w/v) yeastextract was used as carbon and energy source. Aspar-agine H20 or Na2-succinate 6H20 was used as thecarbon and energy source to grow uninduced cells(i.e., cells not induced for the enzymes of the man-delate pathway). The normal concentration of thesesubstances was 0.4% (w/v), but 1.0% (w/v) aspar-agine was used in the kinetic experiments where itwas necessary to obtain dense populations. Aromaticacids (mandelic, benzoylformic, and most compoundsscreened for ability to act as nonmetabolizable in-ducers) were added to the basal medium as the sodiumsalts at a concentration of 0.01 M (about 0.2%, w/v)unless otherwise stated. Carbon and energy sourcesor inducers were sterilized separately and added tothe culture aseptically before inoculation. Solid mediawere prepared by the addition of 1.0% (w/v) agar(Oxoid lonagar No. 2) to the liquid medium. Toprevent discoloration, the agar and mineral solutionswere autoclaved separately, each at double strength,and were mixed before dispensing.

Cultures were incubated at 30 C on a rotaryshaker. Growth was measured turbidimetrically,either in a spectrophotometer (Beckman model DU)at 680 mg or in a colorimeter (Klett-Summerson)fitted with a no. 66 filter. Cultures were appropriatelydiluted for measurement if preliminary readings fellbeyond the range of proportionality. Dry weight andcell count were indirectly determined by means of

VOL. 91 X 1966 1 141

on January 2, 2019 by guesthttp://jb.asm

.org/D

ownloaded from

J. BACTERIOL.

D-MANDELIC ACID

L-MANDELIC ACID

BENZOYLFORMIC ACID

BENZALDEHYDE

BENZOIC ACID

HODi -COOH

OHJFOH

OZ~ C-COOHOH4I- 0o

CC-COOH

OX CHO

OD- COOH

MANDELIC RACEMASE

L-MANDELIC DEHYDROGENASE

BENZOYLFORMIC CARBOXYLASE

TPN- AND DPN-LINKEDDEHYDROGENASES

ENZYMES OF THEMANDELATE GROUP

CIS, CIS-MUC

MUCON(

ENOL

3-KETOAI

OHCATECHOL OH

COOH:ONIC ACID QCOOH

)LACTONE jOOH

-COOHL-LACTONE ,,`C o

COOHDIPIC ACID O=c0COOH

BENZOIC OXIDASE SYSTEM

PYROCATECHASE

LACTONIZING ENZYME

MUCONOLACTONE ISOMERASE

ENOL-LACTONE HYDROLASE

ENZYMES OF THECATECHOL GROUP

FIG. 2. The mandelate pathway.

previously prepared calibration curves relating thesequantities to turbidity.

Enzymological methods and extraction. Cells werealways harvested from exponentially growing cul-tures. Exploratory experiments showed that exhaus-tion of the carbon and energy source by a culture,even in the presence of adequate inducer, was followedby a rapid decline in the level of several inducibleactivities. Cells from cultures of small volume wereharvested in a refrigerated centrifuge operated at5,000 X g for 10 min. The pellets were washed inchilled 0.05 M phosphate buffer (Na2HPO4-KH2PO4,pH 6.8) and recentrifuged. The cells were then eitherimmediately extracted, or stored as a paste at -40 C.In kinetic studies on induction, samples of cultureswere poured over sufficient quantities of flake icemade from distilled water to lower the temperatureof the suspension to 0 C almost immediately. Thecells were subsequently harvested as described above.Large batches of cells, required for the purification ofenzymes, were harvested in an unrefrigerated air-driven Sharples centrifuge and stored, without wash-ing, in the frozen state.

Cells were resuspended for extraction in 0.05 Mphosphate buffer (pH 6.8) at a density of 20 g of wet,packed cells in 100 ml of buffer, and extracts wereprepared by sonic oscillation. With small samples of

cells, a 20-kc, 60-w probe-type oscillator (Measuringand Scientific Equipment Co., Ltd., London, Eng-land) was used. Larger quantities of cells were ex-tracted with 9- and 10-kc oscillators (Raytheon Manu-facturing Co., Waltham, Mass.). Conditions werestandardized by determining the time of exposurenecessary to attain a 95% decrease in absorbancy oftreated suspensions at 680 mnI. The volume of sus-pensions treated ranged from 2 to 50 ml. All extrac-tions were carried out on samples jacketed at 0 C. Nospecial precautions were taken to provide a reducingor inert atmosphere during extraction, since explora-tory experiments showed that none of the enzymesstudied was measurably sensitive to the presence of air.

After sonic treatment, the extract was centrifugedin the cold for 10 min at 5,000 X g to remove un-broken cells. A small portion of the supernatantfluid was set aside for protein determinations, and therest was subjected to centrifugation at 38,000 rev/minfor 60 min in the 40 head of a Spinco Model L prepar-ative ultracentrifuge (ca. 100,000 X g). This treatmentsediments the particulate components of the crudeextract which contain reduced pyridine nucleotideoxidase, cytochromes, and L(+)-mandelate dehydro-genase (8). After decantation of the soluble fraction,the pellet and the walls of the tube were rinsed withchilled 0.05 M phosphate buffer (pH 6.8), and the

1142 HEGEMAN

on January 2, 2019 by guesthttp://jb.asm

.org/D

ownloaded from

VOL. 91, 1966 SYNTHESIS OF MANDELAl

pellet was resuspended by syringe in a small amountof the same buffer. This suspension will be termedthe particle fraction.

All enzyme preparations were assayed as soon aspossible, usually within 24 hr after extraction, andwere kept in closed tubes in ice until that time. Mostof the soluble mandelate enzymes were quite stable,and could be frozen routinely during purification.However, L(+)-mandelate dehydrogenase, present inthe particle fraction, was markedly inactivated byfreezing. It was found that addition of 40% (v/v)ethylene glycol to particle preparations would permitstorage at -40 C with no measurable loss of L(+)-mandelate dehydrogenase activity for several months.

In crude extracts, as well as the soluble and particlefractions derived from them, protein was determinedby the biuret method (22), modified according toPardee (personal communication). When the totalamount of protein was small, the procedure of Lowryet al. (13) was employed. Bovine serum albumin,fraction V (Armour Pharmaceutical Co., Kankakee,Ill.), dried over silica gel, was used as a standard forboth methods.An essential precondition for the studies which were

envisaged was the development of a series of rapidand sensitive spectrophotometric assays for as manyof the enzymes of the mandelate pathway as possible.Such assays had already been developed for thebenzaldehyde dehydrogenases (8) and the enzymethat lactonizes cis-cis muconate (17). In the course ofthis work, we developed spectrophotometric assaysfor mandelate racemase, L(+)-mandelate dehydro-genase, benzoylformate decarboxylase, and pyro-catechase. We also used the specific assays of Ornstonand Stanier (15) for the two enzymes that convert(+)-muconolactone to 3-ketoadipate. All five enzymeswhich function in the conversion of D(-)-mandelateto benzoate (the mandelate group) and all enzymesmediating the conversion of catechol to g-ketoadipate(the catechol group) were assayed spectrophoto-metrically. The enzyme system that converts benzoateto catechol does not lend itself to simple assay methodsand is highly unstable (11). Accordingly, this enzymewas not studied. Also excluded from study were theenzymes that convert ,B-ketoadipate to acetyl andsuccinyl CoA.

All assays were performed on the soluble fraction(100,000 X g supematant) except that for the L(+)-mandelate dehydrogenase, which is exclusively foundin the particle fraction. Reactions were conducted in1-cm path-length silica cuvettes containing a totalliquid volume of 3 ml. Measurements were made inrecording spectrophotometers (Gilford model 2000or Cary model 14). The reaction temperature was 23to 25 C in all cases.

In extracts of cells cultivated under the same con-ditions, the reproducibility of specific activities de-termined for most of the enzymes studied was aboutd30%. The fluctuations are attributable in part toerrors of measurement associated with the determina-tions of activity and protein, and in part to the variabledegrees of enzyme inactivation which occurred duringpreparation, handling, and storage of extracts.

Spectrophotometric assays. L(+)-mandelate dehy-

rE PATHWAY ENZYMES. I 1143

drogenase [no International Union of BiochemistryEnzyme Commission number; systematic name: L(+)-mandelate: (acceptor) oxidoreductase]. The methodof assay was suggested to us by W. C. Schuster. It isdependent on the reduction of an acceptor dye, 2,6-dichlorophenol-indophenol, to the leuco form con-comitant with the oxidation of L(+)-mandelate tobenzoylformate. The reaction was measured at 600m,u near the maximal absorbancy of the oxidizedform of the dye. In addition to enzyme, the reactionmixture contained 200 Amoles of phosphate buffer(pH 7.0), 25 jAmoles of sodium DL-mandelate, and 0.4,umole of 2, 6-dichlorophenol-indophenol. A refer-ence cuvette contained all these components exceptthe substrate. The typical curve for dye reduction as afunction of time is sigmoid in form; the rate wasestimated from the intermediate linear portion of thecurve. The addition of 0.001 M NaCN increased theduration of linearity, but not the reaction rate.Cyanide was not used routinely. The reaction ratewas proportional to enzyme concentration over awide range, except at extremes of concentration.With a value of 20.6 X 106 for the molar absorbancycoefficient of the dye at 600 miA and pH 7.0 (1), adecrease of 6.72 absorbancy units corresponds to theoxidation of 1 jAmole of L(+)-mandelate to benzoyl-formate.

Mandelate racemase (I.U.B.E.C. No. 5.1.2.2, man-delate racemase). The assay depends upon the observa-tion that preparations of washed particle fractionfrom mandelate-induced cells will oxidize only theL(+) isomer of mandelate, and that the D(-) isomerdoes not interfere. Accordingly, racemase activitymay be determined by measuring the rate of dye re-duction in a system containing D(-)-mandelate assubstrate, together with an excess of washed particlefraction. The reaction mixture contained, in additionto enzyme, 200 ,umoles of phosphate buffer (pH 7.0),25 Amoles of D(-)-mandelate (Mann Research Lab-oratories, New York, N. Y.), a sufficient quantity ofwashed particle fraction to oxidize a saturating quan-tity of L(+)-mandelate at 20 times the maximalracemase rate measured, and 0.4 jAmole of Na2-2 , 6-di-chlorophenol-indophenol. A reaction mixture com-plete except for substrate served as reference. Prepara-tions of washed particle fraction were assayed forracemase in a separate control; any contaminatingactivity was subtracted from that measured. TheD (-) -mandelate used as substrate was freed fromcontaminating L(+) isomer by treatment with washedparticle fraction. As in the assay for L-mandelate de-hydrogenase, to which the racemase reaction wascoupled, a decrease of 6.72 absorbancy units at 600m,u under the conditions of the assay corresponds tothe conversion of 1 ,.mole of mandelate from theD(-) to the L(+) isomer.

Benzoylformate decarboxylase (E.C. No. 4.1.1.7,benzoylformate carboxy-lyase). The assay exploitsthe fact that the substrate, benzoylformate (AldrichChemical Co., Inc., Milwaukee, Wis.), has a molec-ular extinction coefficient of 81 at 334 mu and pH 6.0,while benzaldehyde and subsequent products havenegligible absorbancy at this wavelength. The reac-tion mixture contained, in addition to enzyme, 200

on January 2, 2019 by guesthttp://jb.asm

.org/D

ownloaded from

HEGEMAN

jAmoles of phosphate buffer (pH 6.0), 25 jAmoles ofNa-benzoylformate, and 50 jug of thiamine pyrophos-phate chloride. The reference cuvette contained nosubstrate. The course of the reaction was followed at334 m,u. A decrease in absorbancy of 0.0272 unitscorresponds to the decarboxylation of 1 ;&mole ofsubstrate.

Nicotinamide adenine dinucleotide (NAD)-benzalde-hyde dehydrogenase (E.C. No. 1.2.1.6, benzaldehyde:NAD oxidoreductase) and NAD phosphate (NADP)-benzaldehyde dehydrogenase (E.C. No. 1.2.1.7,benzaldehyde: NADP oxidoreductase). The assayprocedures for these two enzymes are presented to-gether here, since the reactions are identical in allrespects except for their specificity with respect tothe pyridine dinucleotides. The procedures weremodified from Gunsalus et al. (8). The assays measurethe benzaldehyde-dependent reduction of NAD orNADP, followed at 340 mMA. Each cuvette contained200 ,moles of tris(hydroxymethyl)aminomethane-HCl (Tris-HCl) at pH 8.0, 3 ,umoles of redistilledbenzaldehyde (stored under nitrogen), and 2.5 ,molesof the appropriate dinucleotide in the oxidized form(Sigma Chemical Co., St. Louis, Mo.). Substrate wasomitted from the reference cell. An increase in ab-sorbancy at 340 m,u of 2.07 units corresponds to theoxidation of 1 ,umole of substrate to benzoate.

Pyrocatechase (E.C. No. 1.99.2.2, catechol 1,2-oxygenase). The assay was based on the observationof Sistrom and Stanier (17) that the subsequentenzyme of the pathway (the lactonizing enzyme) iscompletely inhibited by disodium ethylenediamine-tetraacetate (EDTA). Cis-cis muconate, the productof pyrocatechase, will therefore accumulate in thepresence of EDTA. Pyrocatechase itself is not sensi-tive to EDTA at the concentrations employed. Thereaction was measured at 260 my, at which wave-length catechol absorbs weakly and cis-cis muconateabsorbs strongly. The reaction mixture contained,besides enzyme, 4.0 ,umoles of EDTA, 0.3 ;umole ofcatechol, and 200 ,umoles of phosphate buffer (pH7.0). The conversion of 1 jAmole of catechol to cis-cismuconate causes an increase in absorbancy at 260 muof 5.60 units.

Lactonizing enzyme [E.C. No. 5.5.1.1, (+)-4-carboxymethyl4-hydroxyisocrotonolactone lyase (de-cyclizing)]. The assay used was essentially that ofSistrom and Stanier (17), with slight modification.The disappearance of cis-cis muconate determinedat 260 m, was used as a measure of reaction velocity,since the reaction products absorb only negligiblyat this wavelength. The reaction mixture contained200 ,umoles of Tris-HCl (pH 8.0) and 0.3 ;&mole ofcis-cis muconate, sodium salt, and 3.0 ,umoles ofMnCI2. The disappearance of 1 ,umole of cis-cismuconate was equivalent to a decrease of 5.75 ab-sorbancy units under the conditions of the assay.

Delactonizing enzyme [E.C. No. 3.1.1.16, (+)-4-carboxymethyl-4-hydroxyisocrotonolactonase]. Thisenzyme system, first described by Sistrom and Stanier(17), was recently found to comprise two physicallyseparable activities-muconolactone isomerase andenol-lactone hydrolase (15). Although the inter-mediate (the j8-ketoadipate enol-lactone) has been

crystallized, the two enzymes were assayed by couplingreactions, i.e., adding the complementary enzyme inexcess and measuring the overall reaction accordingto Sistrom and Stanier (17).

Cultures of P. putida A.3.12 grown with p-hydroxy-benzoate contain a high (induced) level of enol-lactone hydrolase, but an undetectable level ofmuconolactone isomerase. Accordingly, the mucono-lactone isomerase assay system contained excessenol-lactone hydrolase in the form of a rate-saturatingquantity of extract prepared from wild-type P.putida A.3.12 grown on p-hydroxybenzoate. In addi-tion, the system contained a measured limitingquantity of the extract to be assayed, 3.0 ,Amoles ofthe DL-lactone, and 100 ymoles of Tris-HCl (pH 8.0).A decrease of 0.476 absorbancy units at 230 mMcorresponds to the isomerization of 1 pmole of (+)-muconolactone.

The enol-lactone hydrolase was measured by theuse of a mutant of strain A.3.12, incapable of pro-ducing detectable enol-lactone hydrolase, even whengrown in the presence of compounds which serve toinduce this enzyme in the wild type. Extracts ofbenzoate-induced cells of this mutant were added tothe reaction mixtures in sufficient quantity to providea rate-saturating concentration of enol-lactone.Known quantities of enol-lactone hydrolase couldthus be assayed in the overall reaction described bySistrom and Stanier (17). The conditions were other-wise the same as those described above for the assayof muconolactone isomerase.

Radiochemical assays. Assays for mandelateracemase, L(+)-mandelate dehydrogenase, and ben-zoylformate decarboxylase were devised to extendthe sensitivity of measurement beyond that of thephotometric assays. The compositions of the reactionmixtures for these radiochemical modifications wereas described for the spectrophotometric assays, butthe total reaction volumes were reduced to 0.25 ml.The substrates for the assays were D(-)- and DL-mandelate-carboxyl-C'4 and benzoylformate-car-boxyl-C'4 of known specific activity. The reactionswere carried out in 16 X 150 mm test tubes fittedwith gas-tight serum caps, from which depended 1 X3 cm strips of Whatman no. 1 filter paper wet with25 gliters of 10 N NaOH. The reaction mixture wasprepared separately; 0.25-ml portions were dispensedinto a series of tubes, capped, and incubated at 25 Cin a reciprocating water bath. Reactions were stoppedat intervals by injecting through the cap a quantity of5 N H4SO4 sufficient to bring the measured pH of thereactions to 0.5. The tubes were incubated for 30min after addition of the acid, to ensure completeentrapment of CO2 on the filter paper strips. Thestoppers were then removed, and the papers wereplaced directly in scintillation-counter sample vials.The samples were counted as described below. Con-trols were prepared by omitting substrates and byomitting one or more enzymes. When several hoursof incubation were necessary, a drop of toluene wasadded to the reaction mixture as a preservative. Allactivities were proportional to the amount of enzymeadded within the limits of measuring error.

In the benzoylformate decarboxylase assay, the

1144 J. BACTERIOL.

on January 2, 2019 by guesthttp://jb.asm

.org/D

ownloaded from

SYNTHESIS OF MANDELATE PATHWAY ENZYMES. I

rate of release of labeled CO2 from the carboxyl-carbon of the substrate was taken directly as a measureof activity. From a knowledge of the specific activityof the labeled substrate and the efficiency of countingthe trapped CO2, the rate of the reaction could be cal-culated in millimicromoles of substrate decomposedper unit time.

Extracts of mutants, which lacked the enzymaticfunction to be measured, were prepared from culturesgrown in the presence of inducer, and were added tothe reaction mixtures when mandelate racemase andL(+)-mandelate dehydrogenase assays were per-formed on extracts of uninduced cells of the wildtype. These extracts served as genetically resolvedbiological reagents and provided, in excess, the other-wise deficient activities of the mandelate groupenzymes linking the measured reaction to the releaseof CO2 from the carboxyl-carbon of benzoylformate.Mandelate racemase and L(+)-mandelate dehydro-genase activities were calculated as described abovefor the benzoylformate decarboxylase.

Chemicals and reagents. Organic compoundsscreened for the ability to act as nonmetabolizableinducers were obtained from a wide variety of com-mercial sources, or synthesized according to standardmethods available in the literature. If these com-pounds failed to melt within a two-degree range whichincluded the reported value, they were recrystallizedfrom appropriate solvents until a satisfactory meltingpoint was obtained.

Cis-cis muconic acid and DL--y-carboxy-A"-buteno-lide (DL-4-carboxymethyl-4-hydroxyisocrotonolac-tone) were synthesized according to Elvidge et al. (6).

DL-Mandelic acid-carboxyl-C14 was synthesizedaccording to Fieser and Fieser (7) with the followingmodification: the potassium bisulfite addition productof benzaldehyde was isolated, purified, and added inexcess to a small amount of sodium-C'4-cyanide.This improved the radiochemical yield and specificactivity, since the formation of mandelonitrile has anequilibrium constant near unity. Benzoylformic acidcarboxyl-C14 and D(-)-mandeiic acid-carboxyl-C'4were prepared by treating DL-mandelate-carboxyl-C'4with thoroughly washed particle fraction preparedfrom cells of a mutant strain of P. putida A.3.12 unableto form racemase. This treatment converts the L(+)-mandelate to benzoylformate and leaves the D(-)isomer unchanged. The acids were extracted withether from the acidified mixture and were separatedby descending preparative chromatography on paper(Whatman 3MM) with the solvent of Williams andKirby (23). The acids were located by radioautog-raphy (Kodak No-Screen X-ray film) and eluted with0.1 M Na2HPO4. Identity and chemical purity weredetermined by RF and ultraviolet-absorption spectra.Radiochemical purity was established for all threeacids by ascending paper chromatography (Whatmanno. 1) with ethyl alcohol-acetic acid-water (8:1:1),and by ascending chromatography on the same paperin the solvent of Williams and Kirby (23). The chro-matograms were examined in a gas-flow scanner(Nuclear-Chicago Actigraph) fitted with a scaler andstrip chart recorder. Activities were determined in athin-window gas-flow counter (Nuclear-Chicago) at

infinite thinness, or in a refrigerated scintillationspectrometer (Packard Tricarb) with the use of thedioxane-base scintillator of Bray (2). Counting effi-ciency was determined by use of a polystyrene-C'4standard or C14-toluene of known activity. Allsamples were counted to a variance of less than 1%.

RESULTS AND DIscussIoNEnzymatic constitution and inducers of the wild

type of P. putida A.3.12-levels ofenzymes of themandelate pathway in induced and uninduced cells.A striking fact that emerged from early studieson the metabolism of aromatic acids by P. putidaA.3.12 (18) was the apparently complete in-ability of cells grown in either yeast extract ormineral-asparagine medium to attack immedi-ately the aromatic compounds upon which theycould grow. The explanation then offered wasthat the necessary enzymes were inducible, andthis conclusion was confirmed for some of theenzymes of the aromatic pathways by directmeasurement of activity in cell-free extracts pre-pared from cells after growth in the presence orabsence of inducer-substrates (8). However, noneof these early studies accurately established eitherthe absolute levels of activity of the mandelatepathway enzymes in fully induced cells, or theirpossible maximal basal levels in uninduced cells.This was, accordingly, the first aspect of theproblem which had to be examined. Cells ofP. putida A.3.12 were harvested during exponen-tial growth, extracts were prepared, and the en-zymes of the mandelate pathway were assayedindividually by spectrophotometric means. Thegrowth substrates employed as sole carbonsources were the two isomers of mandelate, theracemic mixture, benzoylformate, benzoate, andasparagine, the last-named being used to ascer-tain the basal levels of the enzymes in uninducedcells.The racemic mixture and the individual isomers

of mandelate are all equally effective as inducers(Table 1). In cells which have been grown on theL(+) isomer, the racemase is functionally super-fluous. Thus, the L( +) isomer does serve toinduce formation of racemase, an apparent ex-ception to the principle of "sequential induction"as originally stated (4). An even more strikingexception is the induction to full levels of bothmandelate racemase and L( +)-mandelate de-hydrogenase by benzoylformate. It could beargued, however, that benzoylformate may bereduced within the cell, giving rise to one or bothisomers of mandelate, which in turn may act asthe real inducers of these two enzymes. This in-terpretation can be ruled out on the basis of ex-periments described in a later paper (9). Theability of benzaldehyde to act as an inducer was

1 145VOL. 91, 1966

on January 2, 2019 by guesthttp://jb.asm

.org/D

ownloaded from

TABLE 1. Levels ofmandelate pathway enzymes found in extracts of wild-type Pseudomonas putida A.3.12grown with the isomers of mandelate and benzoylformate as sole carbon and energy sourcesa

Carbon and energy sourcesEnzyme

DL-Mandelateb L(+)-Mandelatec D(-)-MandelateC Benzoylformateb

Mandelate racemase 400d 412 364 337(301-485) - - (268-377)

L(+)-Mandelate dehydro-genase.522 710 668 552

(486-566) - - (471-627)Benzoylformate decarboxylase. 603 625 585 595

(525-711) - - (593-600)NAD-benzaldehyde dehydro-genase.265 240 320 312

(243-287) - - (234-351)NADP-benzaldehyde dehydro-genase.152 140 160 178

(146-158) - - (146-232)Pyrocatechase... 288 350 225 391

(271-305) - - (272-310)Lactonizing enzyme.485 530 420 443

(420-550) - - (387-500)Muconolactone isomerase 2700 - - 2700

(2420-2950)Enol-lactone hydrolase 1350 1265

(1180-1520) _

a All activities are expressed as millimicromoles of substrate converted per minute per milligram ofprotein.

b Three extracts were assayed.c One extract was assayed.d Average values are given. Numbers in parentheses indicate range.

not investigated because of the toxicity of thecompound and its tendency to undergo a rapidspontaneous oxidation in air.As shown in Table 2, extracts of asparagine-

grown (uninduced) cells contain levels of thenine enzymes of the mandelate pathway less thanor equal to the reliable limits of the spectro-photometric assays. Benzoate-grown cells containundetectable quantities of the five enzymes of themandelate group, but have elevated levels of thecatechol group activities. This finding is in accordwith the earlier conclusions of Stanier (18), basedupon manometric analysis.

Basal levels by radiochemical techniques. Toestablish the true magnitude of the response ofthe cell to inducer, it was necessary to determinethe basal rate of synthesis of at least some of themandelate group enzymes in the absence of in-ducer. Since noninduced cells have levels ofactivity of mandelate pathway enzymes so lowthat they are undetectable by usual spectro-photometric procedures, more refined measure-ments of the mandelate racemase, L( +)-man-delate dehydrogenase, and benzoylformatedecarboxylase were performed on extracts ofnoninduced cells by use of radiochemical tech-

niques (Table 3). Since the conditions of bothspectrophotometric and radiochemical assays areidentical, the values obtained are directly com-parable. Extracts were prepared in the usualmanner from cells grown in liquid asparagine-basal medium from an inoculum transferredseveral times on yeast extract-mineral agar.

Since specific activities reflect differential ratesof synthesis, a comparison of the specific ac-tivities of Table 3 (noninduced) with those ofTable 1 (induced) permits the calculation thatgrowth of the wild type of P. putida A.3.12 in thepresence of 0.01 M DL-mandelate increases therates of synthesis of mandelate racemase, 2,000-fold, of L(+)-mandelate dehydrogenase 2,300-fold, and of benzoylformate decarboxylase atleast 430-fold.

Observations on the absence of metabolite re-pression. Experiments on the kinetics of inductionand the response of the wild type to nonmetabo-lizable inducers entailed the use of asparagine orsuccinate as principal or accessory carbon andenergy sources. If asparagine or succinate re-pressed the synthesis of the enzymes of themandelate pathway to any extent, the interpreta-tion of many of these experiments would be

1146 HEGEMAN J. BACTERIOL.

on January 2, 2019 by guesthttp://jb.asm

.org/D

ownloaded from

SYNTHESIS OF MANDELATE PATHWAY ENZYMES. I

TABLE 2. Levels of mandelate pathway enzymes found in extracts of wild-type Pseudomonas putida A.3.12grown with asparagine and benzoate as sole carbon and energy sourcesa

Per cent of levrel Cells grwon PrctoflvliEnzyme Cells grown on asparagineb in mandelatee berown on Per cent of level ingrown cells benzoate5 mandelate-grown cells

Mandelate racemase .......... <0. 15d 0.038 .0. 15 0.038

L(+)-Mandelate dehydro-genase....................... .0.3 0.058 .0.3 0.058

Benzoylformate decarboxylase.. .2.0 0.33 <2.0 0.33

NAD-benzaldehyde dehydro-genase ....................... .0.5 0.21 <0. 5 0.21

NADP-benzaldehyde dehydro-genase....................... .0.5 0.33 <0.5 0.33

Pyrocatechase................. .0.05 0.17 360 125(242-460)

Lactonizing enzyme............ .0.01 0.02 430 90(371-490)

Muconolactone isomerase.......10.0 0.37 2,360w 88

Enol-lactone hydrolase. .10.0 0.74 1 190e0 89

a All activities are expressed as millimicromoles of substrate converted per minute per milligram ofprotein.

b Four extracts were assayed.c Two extracts were assayed.d Average values are given. Numbers in parentheses indicate range.e One extract was assayed.

difficult or impossible. Consequently, it was es-

sential to determine whether any such repressionexisted, and, if so, what its magnitude might be.

Asparagine-grown cells were used to inoculatea series of five flasks of basal medium containing,as carbon and energy sources, asparagine alone,asparagine and DL-mandelate in a molar ratio of3:1, 1:1, and 1:3, and DL-mandelate alone. Theentire course of growth was followed turbidi-metrically. In each case, the time at which man-delate began to disappear from the medium was

determined by spectrophotometric analysis ofsamples of the culture supernatant fluid. Thegeneration times of mandelate- and asparagine-grown cells differ only very slightly: 57 min forthe former, 52 min for the latter (Fig. 3). Inflasks furnished with mixtures of the two sub-strates, the observed generation times lay betweenthose established for the two substrates alone,and in no case was a diauxic lag observed. Fur-thermore, the time at which mandelate began todisappear was essentially the same for all cultureswhich contained this compound. An analogousexperiment in which succinate replaced asparagineyielded the same results. These facts all point verystrongly to the absence of any marked repression

TABLE 3. Specific activities of some mandelate groupenzymes in asparagine-grown (noninduced)

cells of wild-type Pseudomonas putidaA.3.12 determined by radiochemical

procedures

Per cent ofEnzyme Specific fully induced

activity* wild-typecellst

Mandelate racemase. . 0.156 0.052L(+)-Mandelate dehy-drogenase.0.172 0.043

Benzoylformate decar-boxylase. 1.44 0.210

* Expressed as millimicromoles of substrateconverted per minute per milligram of protein.

t Values taken as per cent of the mean activitiesfor DL-mandelate-grown cells presented in Ta-ble 1.

of the enzymes of the mandelate pathway byasparagine or succinate. More direct evidencewith respect to one such enzmye was obtained byharvesting samples of cells from each mandelate-containing culture during the course of expo-nential growth and prior to the exhaustion of

VOL. 91, 1966 1147

on January 2, 2019 by guesthttp://jb.asm

.org/D

ownloaded from

HEGEMAN

200

>. 60 4

1 2 3 4 5 6Hours

FIG. 3. Course of growth of the five cultures de-scribed in the text: (I) S X 10-3 M asparagine; (2)2.5 X 10-3 M asparagine and 2.5 X 10-3 M DL-mandel-ate; (3) 1.66 X. 10-3 M asparagine and 3.33 X 10-3 MDL-mandelate; (4) 3.33 X 10-3 M asparagine and 1.66X 10-3 M DL-mandelate; (5) 5 X2O-3 M DL-mandelate,as carbon and energy sources in basal medium, re-spectively. The arrow indicates the time at whichmandelate begins to disappear from the cultures.

mandelate, and determining the specific activityof benzoylformate decarboxylase on extractsprepared therefrom. As shown in Table 4, cells ofthese four cultures contain essentiaey identical( ±20%) specific activities of benzoylformate de-carboxylase.These results, and those of the mixed-substrate

growth experiment, indicate that the repressiveeffects of accessory carbon sources on synthesis ofthe mandelate pathway enzymes observed byMandelstam and Jacoby (14) could be neglectedin interpreting the results reported here.

Molecular specificity of induction of L(+)-mandelate dehydrogenase. A number of analoguesof mandelic acid and benzoylformic acid werecollected and screened for the ability to induceearly enzymes of the mandelate pathway. Thelarge number of determinations involved were

performed by making use of the qualitative assayprocedure for L( +)-mandelate dehydrogenase(screening test) described in connection with theisolation of constitutive mutants (10). The wildtype of P. putida A.3.12 was grown as patches onplates of mineral-succinate agar to which theanalogues had been added. After growth had oc-curred, the patches were tested for L(+)-mande-late dehydrogenase activity. Analogues whichgave indication of positive results in this test were

TABLE 4. Specific activity of benzoylformate de-carboxylase in extracts of cells grown in thepresence ofmixtures ofasparagine and DL_

mandelate

Specificactivity of

Carbon and energy source benzoylfor-mate decar-boxylase*

Asparagine alone ................... .2Asparagine and DL-mandelate, 1:1.... 630Asparagine and DL-mandelate, 1:2. 570Asparagine and DL-mandelate, 2:1. 623DL-Mandelate alone................... 585

* Expressed as millimicromoles of substrateconverted per minute per milligram of protein.

subsequently retested for inductive function by amore precise method. Each such compound wasadded to a culture in succinate-basal medium,the cells were harvested after growth had oc-curred, extracts were prepared, and the levels ofL( +)-mandelate dehydrogenase were measuredby spectrophotometric assay.Any compounds which proved to be effective

as inducers under these conditions were thencarefully examined for their ability to be metabo-lized. Criteria for nonmetabolizability included:(i) lack of stimulation of oxygen uptake uponaddition to suspensions of cells grown in thepresence of the compound; (ii) lack of qualitativeor quantitative change of ultraviolet spectra ofculture supernatant fluids after prolonged cultiva-tion in the presence of the compound; and insome cases (iii) quantitative re-extraction andspectrophotometric purity determination afteraddition to liquid cultures and prolonged incu-bation of the compound in the presence of man-delate-grown cells.Among the compounds which proved to be

inducers (Table 5), D(-)- and L(+)-mandelate,benzoylformate, and DL-p-hydroxymandelate areknown to be metabolized through the action ofthe enzymes of the mandelate group (8). DL-Phenyl-1 ,2-ethanediol and phenylglyoxal areprobably precursors of mandelate and metabo-lized via the mandelate pathway, since they sup-port the growth of P. putida A.3.12, and cellsgrown at the expense of DL-phenyl-1 , 2-ethanediolhave the fully induced complement of mandelatepathway activities. DL-Mandelamide, methyl andethyl DL-mandelate, and ethyl m-hydroxy-DL-mandelate will all support growth. The first threeof these compounds probably give rise to DL-mandelate through hydrolysis; the mode ofmetabolism of ethyl m-hydroxy-DL-mandelate isunknown, but probably also entails hydrolysis.p-Chloro and p-bromo-DL-mandelate do not sup-

1148 J. BACTrERIOL.

on January 2, 2019 by guesthttp://jb.asm

.org/D

ownloaded from

SYNTHESIS OF MANDELATE PATHWAY ENZYMES. I

TABLE 5. Compounids effective as i,zducers

Metabolized byenzymes of the Utilized for Metabolized

Name of compound Structural formula mandelate growth without growthgroup

.,

D- and L-mandelate

Benzoylformate

p-Hydroxy DL-mandelate

DL-Phenyl 1,2-ethanediol

Phenylglyoxal

DL-Mandelamide

Methyl-DL-mandelate

Ethyl-DL-mandelate

Ethyl meta hydroxymandelate

DL-

p-Chloro-DL-mandelate

DL-p-bromo-mandelate

Thiophenyl acetate

Phenoxy-acetate

O~~CHOHCOOH

OjCOCOOH

HO7CHOHCOOH

O~~CHOHCH20H

(O3~COCHO

OjfCHOHCONH2

(O3~CHOHCOOCH3

(O7CHOHCOOCH2CH3

HO3CHOHCOOCH2CH3

Br jCHOHCOOH

CJ CHOHCOOH

5O~SCH2COOH

O~OCH2COOH

* Not determined.port growth. Oxygen consumption by inducedcells of P. putida A.3.12 at the expense of thesetwo compounds indicates that they are metabo-lized via the mandelate group enzymes to the levelof benzoate. The inability of these halogen-substituted analogues to support growth probablyreflects the unacceptability of the correspondingbenzoic acids as substrates for the benzoate oxi-dase system (11).

Neither thiophenylacetate nor phenoxyacetateare metabolized, but they function as inducers.

+

±

+

+

+

+

+

+

+

+

ND*

ND

ND

ND

ND

ND

ND

ND

+

+

- (spectra)

- (spectra,extractionand 02uptake)

Thiophenylacetate was toxic and seemed a poorerinducer than phenoxyacetate at equivalent con-centrations. Both these compounds, it should benoted, have an electronegative center in approxi-mately the same position as that which is createdby the carbonyl or hydroxyl functions of benzoyl-formate and mandelate, respectively. They alsoshare the intact benzene ring and carboxyl func-tion with these compounds.The majority of the noninducing compounds

did not support growth (Table 6), but some ex-

1149VOL. 91, 1966

on January 2, 2019 by guesthttp://jb.asm

.org/D

ownloaded from

HEGEMAN

TABLE 6. CoIm1poiIIids ileffective as iniduicers

Name of compound Structural formula |[Utilization for growth

Benzil

Benzoin

Meso dihydro benzoin

DL-Benzilate

DL-Atrolactate

DL-1-Methoxy phenylacetate

Tr-ans cinnamate

Phenyl thiourea

N-phenyl maleamate

p-Acetimidobenzoate

Hippurate

Benzoyl-,3-alanine

N-benzyl DL-alanine

N-phenyl glycine

1-Phenyl butyrate

p-Hydroxyphenyl-l -butyrate

1-Phenyl, -3 -ketobutyrate

Oxanilate

(5COCHOH7

OcocoG)CHOHCHOHc

CH:t

w CI-COOH

OH

()~CHOCH3COOH

QCH: CHCOOH

QNHC(: S)NH2

O NHCOCH:CCHCOOH

CH3CONHO COOH

(j CONHCH2COOH

(OCONHCH2CH2COOH

NHCHCOOH'L.il CH3

cjONHCH2COOH

(OCH2CH2CH2COOH

HO§CH2CH2CH2COOH

KIIICH2CH2COCOOH

(OI7NHCOCOOH

1150 J. BACTERIOL.

- (toxic)

- (toxic)

+

±4

+4

on January 2, 2019 by guesthttp://jb.asm

.org/D

ownloaded from

1151SYNTHESIS OF MANDELATE PATHWAY ENZYMES. I

TABLE 6.-Continued

Name of compound Structural formula Utilization for growth

2-Phenyl succinate

Phenylglyoxal dioxime

Benzyl alcohol

Benzylformate

DL-1-Methyl benzyl alcohol

Acetophenone

2-Hydroxyacetophenone

Pyridine-2-DL-glycollate

Thiophene-2-DL-glycollate

DL-phenlylalanine

3-Phenylpyruvate

3-Phenyl-DL-lactate

Phenylacetate

p-Hydroxyphenylacetate

o-Hydroxyphenylacetate

jHCOHCOOH

CH2

COOH

O NC(NOH)CHNOH

O CH20H

O CH200CH

OCHOHCH3

(O7COCH3

OOCOCH20H

lCJCHOHCOOH

W7iCHOHCOOH

GC H2CHNH2COOH.

DC H2COCOOH

O 9CH2CHOHCOOH

CH2COOH

HO $CH2COOH

DC2H2COOH

OH

ND*

- (toxic)

- (toxic,

- (toxic)

ND

- (toxic)

+

ND

ND

* Not determined.

ceptions should be noted. Hippurate and benzoyl- through benzoate and #3-ketoadipate. Cells of3-alanine are presumably hydrolyzed to yield P. putida A.3.12 grown on benzyl alcohol have

benzoate and the corresponding amino acids, all fully induced levels of the enzymes of the catecholcapable of supporting growth of P. putida A.3.12. group, but undetectable levels of the mandelateBenzyl alcohol is also probably metabolized group enzymes. The mechanism of the conversion

VOL. 91, 1966

on January 2, 2019 by guesthttp://jb.asm

.org/D

ownloaded from

J. BACTERIOL.

of benzyl alcohol to benzoate is unknown. Theremaining metabolizable aromatic compoundswhich do not serve as inducers are presumablyconverted to aliphatic intermediates via thehomogentisate pathway. Phenylbutyrate, p-hy-droxyphenylbutyrate, and phenyl-3-ketobutyrateprobably undergo oxidation to p-hydroxyphenyl-acetate and phenylacetate. DL-Phenylalanine and3-phenyl-DL-lactate probably undergo deamina-tion or oxidation to yield 3-phenylpyruvate, andsubsequently phenylacetate (5). Most aromaticcompounds which were metabolized throughphenylacetate and homogentisate shared thepeculiar property of inducing a low level of theenzymes of the mandelate group in cells grown onsolid media, although they failed to do so inliquid cultures. The phenomenon may explainthe finding of Stanier that cells grown on phenyl-acetate plates can immediately oxidize DL-mande-late to the level of benzoate (19).

Substitution of the aromatic ring of the inducermolecule seems to have little effect on function(p-chloro, p-bromo, m- and p-hydroxymandelateall induce). On the other hand, the pyridine andthiophene analogues of mandelate which weretested do not induce. Apparently the nature of thearomatic ring is important although substitutionof the ring does not abolish function. The require-ment for an intact carboxyl group is hard toassess from these data, since nearly all analoguestested in which this moiety alone was modifiedprobably gave rise to mandelate through hydro-lytic or oxidative attack (ethyl and methyl estersof DL-mandelate, mandelamide, DL-phenyl-1,2-ethanediol, and phenylglyoxal) or were otherwisemetabolized (benzyl alcohol). However, com-pounds analogous to mandelate or benzoylfor-mate but with the carboxyl group replaced bymethyl or hydroxymethyl functions did not induce(DL-1-methyl benzyl alcohol, acetophenone, and2-hydroxyacetophenone). There seems to be astringent requirement for an electronegative func-tion in or on the side chain of the inducing mole-cule. This may be a carbonyl or hydroxyl group(benzoylformate or mandelate) or even thioetheror oxygen ether function (phenylthioacetate andphenoxyacetate). The presence of an ether oxygennot linked to the benzene ring, as in 1-methoxyphenylacetate, will not permit induction; how-ever, compounds with methyl or larger groupsprojecting from the side chain (DL-atrolactate andDL-benzilate) do not induce either. The stericconfiguration of the oxygen-bearing function isapparently not important, since D(-)-mandelateand DL-mandelate are essentially equipotent in-ducers, as will be shown later (9).The length and proximity of the groups upon

the side chain are probably not too critical, since

3-phenyl-DL-lactic acid and 3-phenyl pyruvate arecapable of causing some derepression. 3-Keto-l-phenylbutyrate does not induce, indicating thatone but not two methylene groups may be inter-posed between the aromatic nucleus and theoxygen-bearing carbon of the side chain beforeinducer function is abolished. Since the threecompounds upon which this argument is basedare all metabolized, any conclusion on this pointis tenuous.

In summary, the molecular specificity for theinduction of the early mandelate enzymes is quitehigh. Although 36 structural analogues of man-delate and benzoylformate not metabolized viamandelate or benzoylformate were tested, onlyone, phenoxyacetate (Table 5), gave fair derepres-sion in liquid medium.Enzymes evoked by phenoxyacetate. Strain

A.3.12 was grown in basal-asparagine mediumcontaining 0.01 M sodium phenoxyacetate; thecells were harvested and extracted, and the ex-tracts were assayed for the enzymes of the man-delate pathway. The five enzymes of the mandelategroup are all induced, and within the limits ofexperimental error they are coordinate (Table 7).On an equimolar basis, phenoxyacetate was only10 to 20% as effective as the natural metabolizableinducers. Pyrocatechase and the lactonizing en-zyme are not induced. Thus, it seems that the bio-synthesis of the mandelate pathway enzymes iscoordinate in part, the first five member-enzymesof the pathway constituting a regulatory unit.

Kinetics of induction. Investigation of the regu-lation of synthesis of the enzymes for tryptophanoxidation in P. fluorescens Tr-23 by Palleroni andStanier (16) showed that enzymes synthesized inresponse to a common inducer may be recognizedand distinguished from separately controlledenzymes in the same pathway on the basis of thekinetics of their synthesis. Accordingly, a studyof the kinetics of synthesis of the mandelate path-way enzymes in P. putida A.3.12 was undertaken.DL-Mandelate was added to a culture growing

exponentially in asparagine-basal medium, toprovide a final concentration of 0.01 M. Sampleswere removed at 20-min intervals, and the cellswere harvested and extracted. Each extract wasassayed for mandelate racemase and L( +)-mandelate dehydrogenase in the mandelate group,and for lactonizing enzyme and muconolactoneisomerase in the catechol group. The specific ac-tivities of the four enzymes are plotted againsttime elapsed after addition of inducer in Fig. 4.The two enzymes of the mandelate group ap-

pear almost immediately, but measurable syn-thesis of members of the catechol group occursonly 40 to 60 min after the addition of mandelateto the culture. The mandelate group enzymes

1152 HEGEMAN

on January 2, 2019 by guesthttp://jb.asm

.org/D

ownloaded from

SYNTHESIS OF MANDELATE PATHWAY ENZYMES. I

TABLE 7. Activities of mandelate pathway enzymes in extracts of cells grown in the presence of 0.01 Mphenoxyacetate*

Enzyme Specific activityt Per cent of mandelate-grownwild-type cellst

Mandelate racemase 45 i 10.5 11.2 i 3.3L(+)-Mandelate dehydrogenase 102 + 25 20.0 Z4: 6.0Benzoylformate decarboxylate 67 i 20 12.8 ±4 3.8NAD-benzaldehyde dehydrogenase 56 i 17 21.0 -4 6.3NADP-benzaldehyde dehydrogenase 32 i 10 21.0 i 6.3Pyrocatechase <0.05 (undetectable) 0.016Lactonizing enzyme .0.01 (undetectable) 0.018

* The mean of values from three separate experiments.t Expressed as millimicromoles of substrate converted per minute per milligram of protein.t The specific activities as per cent of the mean values of those for extracts of DL-mandelate-grown

cells presented in Table 1.

*_60

as40

20

0 20 40 60 80 100 120

Minutes after addition of DL-mandelateFIG. 4. Time course of appearance offour enzymes

of the mandelate pathway after addition of DL-man-delate to a culture of Pseudomonas putida strainA.3.12 growing in asparagine-basal medium. Specificactivities are expressed as millimicromoles per minuteper milligram of extract protein except for mucono-lactone isomerase where the units are 104 jumoles permin per mg of protein. Symbols: M, muconolactoneisomerase; 0, mandelic racemase; A, L-mandelicdehydrogenase; 0, lactonizing enzyme.

reach a nearly constant specific activity in theculture at 40 min; the two enzymes of the catecholgroup continue to increase in activity for 2 hrafter the addition of mandelate.The absence of autocatalytic kinetics in the

appearance of the mandelate group activities im-plies that an inducible concentrating permease formandelate is not present in P. putida A.3.12.

ACKNOWLEDGMENTSI am grateful to R. Y. Stanier for his guidance in

the course of this study, and to R. Y. Stanier and M.Doudoroff for advice in the preparation of the manu-script. The collaboration of L. N. Ornston in theradiochemical synthesis and in the design and per-

formance of some of the assays is gratefully acknowl-edged.

This investigation was supported by Public HealthService grant AI-1808 from the National Institute ofAllergy and Infectious Diseases and by NationalScience Foundation grant G-11330. The experimentalwork was done during tenure of Public Health Servicepredoctoral fellowship 5-F1-GM-14638.

LITERATURE CITED1. ARMSTRONG, J. McD. 1964. The molar extinction

coefficient of 2,6-dichlorophenol indophenol.Biochim. Biophys. Acta 86:194-196.

2. BRAY, G. A. 1960. A simple, efficient liquidscintillator for counting aqueous solutions in aliquid scintillation counter. Anal. Biochem.1:279-285.

3. COHEN-BAZIRE, G., W. R. SISTROM, AND R. Y.STANIER. 1957. Kinetic studies of pigment syn-thesis by non-sulfur purple bacteria. J. CellularComp. Physiol. 49:25-68.

4. COHN, M., J. MONOD, M. R. POLLOCK, S. SPIEGEL-MAN, AND R. Y. STANIER. 1953. Terminology ofenzyme formation. Nature 172:1096.

5. DAGLEY, S., M. E. FEWSTER, AND F. C. HAPPOLD.1953. The bacterial oxidation of aromaticcompounds. J. Gen. Microbiol. 8:1-7.

6. ELVIDGE, J. A., R. P. LINSTEAD, B. A. ORKIN,P. SIMs, H. BAER, AND D. B. PATTISON. 1950.Unsaturated lactones and related substances.IV. Lactonic products derived from muconicacid. J. Chem. Soc., p. 2228-2235.

7. FIESER, L., AND M. FIESER. 1957. Experiments inorganic chemistry. D. C. Heath and Co.,Boston.

8. GUNSALUS, C. F., R. Y. STANIER, AND I. C.GUNSALUS. 1953. The enzymatic conversion ofmandelic acid to benzoic acid. III. Fractiona-tion and properties of the soluble enzymes. J.Bacteriol. 66:548-553.

9. HEGEMAN, G. D. 1966. Synthesis of the enzymesof the mandelate pathway by Pseudomonasputida. II. Isolation and properties of blockedmutants. J. Bacteriol. 91:1155-1160.

10. HEGEMAN, G. D. 1966. Synthesis of the enzymes

1153VOL. 91, 1966

on January 2, 2019 by guesthttp://jb.asm

.org/D

ownloaded from

HEGEMAN

of the mandelate pathway by Pseudomonasputida. III. Isolation and properties of con-stitutive mutants. J. Bacteriol. 91:1161-1167.

11. ICHIHARA, A., K. ADACHI, K. HOSOKAWA, ANDY. TAKEDA. 1962. The enzymatic hydroxyla-tion of aromatic carboxylic acids; substratespecificities of anthranilate and benzoate hy-droxylases. J. Biol. Chem. 237:2296-2302.

12. KARLSSON, J. L., AND H. A. BARKER. 1948.Evidence against the occurrence of a tricar-boxylic acid cycle in Azotobacter agilis. J. Biol.Chem. 175:913-921.

13. LOWRY, 0. H., N. J. ROSEBROUGH, A. L. FARR,AND R. J. RANDALL. 1951. Protein measure-ment with the Folin phenol reagent. J. Biol.Chem. 193:265-275.

14. MANDELSTAM, J., AND G. A. JACOBY. 1965. In-duction and multisensitive end-product repres-sion in the enzymic pathway degrading man-delate in Pseudomonasfluorescens. Biochem. J.94:569-577.

15. ORNSTON, L. N., AND R. Y. STANIER. 1964. Themechanism of ,B-ketoadipate formation bybacteria. Nature 204:1279-1283.

16. PALLERONI, N. J., AND R. Y. STANIER. 1964.Regulatory mechanisms governing synthesis ofthe enzymes of tryptophan oxidation by

Pseudomonas fluorescens. J. Gen. Microbiol.35:319-334.

17. SIsTRoM, W. R., AND R. Y. STANIER. 1954. Themechanism of formation of ,B-ketoadipate bybacteria. J. Biol. Chem. 120:821-836.

18. STANIER, R. Y. 1947. Simultaneous adaptation:a new technique for the study of metabolicpathways. J. Bacteriol. 54:339-348.

19. STANIER, R. Y. 1950. Problems of bacterial oxida-tive metabolism. Bacteriol. Rev. 14:179-191.

20. STANIER, R. Y., AND B. P. SLEEPER. 1950. Thebacterial oxidation of aromatic compounds.III. The preparation of enzymatically activedried cells and the influence thereon of priorpatterns of adaptation. J. Bacteriol. 59:129-133.

21. SUDA, M., 0. HAYAISHI, AND Y. ODA. 1949.Studies on enzymatic adaptation. I. Successiveadaptation, with special reference to the me-tabolism of tryptophan. Symp. Enzyme Chemis-try (Tokyo) 1:79-84.

22. WEICHSELBAUM, T. E. 1946. Determination ofproteins in blood serum and plasma. Am. J.Clin. Path. 16:40-49.

23. WILLIAMS, R. J., AND H. KIRBY. 1948. Paperchromatography using capillary ascent. Science107:481-483.

154 J. BAcrRIOL.

on January 2, 2019 by guesthttp://jb.asm

.org/D

ownloaded from