JOURNAL OF BACTERIOLOGY, Nov. 1977, p. 555-563Copyright © 1977 American Society for Microbiology

Vol. 132, No. 2Printed in U.S.A.

Characterization of the Fatty Acid-Sensitive Glucose 6-Phosphate Dehydrogenase from Pseudomonas cepacia

A. F. CACCIAPUOTIt AND T. G. LESSIE*

Department of Microbiology, University of Massachusetts, Amherst, Massachusetts 01003

Received for publication 30 June 1977

The adenosine 5'-triphosphate-insensitive glucose 6-phosphate dehydrogenasefrom Pseudomonas cepacia has been found to be strongly inhibited by long-chain fatty acids and their acyl coenzyme A esters, suggesting that an importantrole of this isoenzyme might be to provide reduced nicotinamide adeninedinucleotide phosphate for reductive steps in fatty acid synthesis. The enzyme,which has been redesignated the fatty acid-sensitive glucose 6-phosphatedehydrogenase, has been purified to homogeneity using affinity chromatogra-phy with nicotinamide adenine dinulceotide phosphate-substituted Sepharoseas a key step in the purification. The purified preparations were used to studythe immunological properties and subunit composition of the enzyme and itsrelationship to the adenosine 5'-triphosphate-sensitive glucose 6-phosphatedehydrogenase present in extracts of P. cepacia. Although both enzymes werefound to be composed of similar size subunits of about 60,000 daltons, immuno-logical studies failed to demonstrate any antigenic similarity between them.Studies of the sedimentation behavior of the fatty acid-sensitive enzyme insucrose gradients indicated that its apparent molecular weight is increased inthe presence of glucose 6-phosphate and suggest that it may exist in anaggregated state in vivo. Palmitoyl coenzyme A, which strongly inhibited theenzyme, failed to influence its sedimentation behavior.

The catabolically versatile bacterium Pseu-domonas cepacia (P. multivorans) possessestwo species of glucose 6-phosphate dehydrogen-ase, which differ in molecular weight, pyridinenucleotide specificity, and sensitivity to inhibi-tion by adenosine 5'-triphosphate (ATP) andreduced nicotinamide adenine dinucleotidephosphate (NADPH) (7). One of the iso-enyzmes, which is active primarily with NADand subject to inhibition by ATP, was purifiedand shown to be a tetramer composed of 60,000-molecular-weight subunits (14). Inhibition ofthis enzyme by ATP presumably serves to con-serve hexose phosphate whenever there is analternate source of NADH for ATP synthesis.The second glucose 6-phosphate dehydrogen-ase which is active primarily with NADP, wasreported to have a molecular weight of about120,000 and to be insensitive to ATP inhibition(7). We report here that the enzyme is stronglyinhibited by long-chain fatty acids and theiracyl coenzyme A (CoA) esters, suggesting itsprimary role might be to provide NADPH forreductive steps in fatty acid synthesis.

Several lines of evidence suggested that the

t Present address: Department of Microbiology and Im-munology, University of Oregon Health Sciences Center,Portland, OR 97201.

two glucose 6-phosphate dehydrogenase speciesmight be different oligomeric forms of the sameprotein, a possibility consistent with theirmolecular weights. First, the two enzymes ap-peared to be lost jointly in mutant strains ofP. cepacia (T. G. Lessie and H. A. Shuman,Abstr. Annu. Meet. Am. Soc. Microbiol. 1974,P123, p. 165). Second, as is described here, thefatty acid-sensitive enzyme can be made to in-crease in apparent molecular weight in thepresence of glucose 6-phosphate to values closeto that of the ATP-sensitive enzyme. Third, athigh concentrations of NAD, the fatty acid-sensitive enzyme has significant ATP-sensitiveNAD-linked activity. At first it seemed rea-sonable that such activity might represent con-version of a fraction of the fatty acid-sensitiveenzyme to the ATP-sensitive species.To further examine the relationship between

the two isoenzymes, we have purified the fattyacid-sensitive glucose 6-phosphate dehydrogen-ase and compared its subunit composition andantigenic properties with those of the ATP-sensitive enzyme. The results indicate that thetwo isoenzymes are immunologically unrelatedand lead to the conclusion that they are prod-ucts of different genes that serve distinct phys-iological roles.

555

on February 4, 2020 by guest

http://jb.asm.org/

Dow

nloaded from

556 CACCIAPUOTI AND LESSIE

MATERIALS AND METHODSGrowth of bacteria. P. cepacia strain 249 (ATCC

17616) was grown at 37TC in 24-liter batches ofmedium containing 5 x 10-2 M phosphate buffer,pH 6.5 (3.3 x 10-2 M KH2PO4 and 1.7 x 10-2 MNa2HPO4), 10-3 M MgSO4, 10-4 M CaCl2, 10-5 MFeSO4, and 0.2% (wt/vol) (NH4)2SO4, with 0.5% (wt/vol) gluconate as the carbon source. Cultures (24liters) were aerated vigorously in a 28-liter Micro-ferm fermentor (New Brunswick Scientific Co., Inc.,New Brunswick, N.J.). The bacteria were harvestedby centrifugation with an RC2B centrifuge equippedwith an SZ-14 continuous-flow rotor (Ivan SorvallInc., Norwalk, Conn.), and the cell paste was frozenat -20°C until use.

Assay of glucose 6-phosphate dehydrogenase. En-zyme activity was determined spectrophotometri-cally by monitoring the formation of NADPH orNADH at 24"C in assay mixtures containing 2 x10-' M tris(hydroxymethyl)aminomethane (Tris)-hydrochloride buffer (pH 8.5), 5 x 10-1 M NAD or

NADP, and 10-2 M D-glucose 6-phosphate as de-scribed earlier (7). Determination of protein forcalculation of enzyme specific activities was by themethod of Lowry et al. (8).

Purification of the fatty acid-sensitive glucose 6-phosphate dehydrogenase. Approximately 150 g(wet weight) of frozen bacteria was thawed andsuspended in 600 ml of 2 x 10-2 M phosphate buffer,pH 6.8, containing 10-2 M sodium azide, lysozyme(2 mg/ml), and deoxyribonuclease II (10 j.ug/ml).The cell suspension was incubated with stirring at37°C for 30 to 60 min to convert the cells to sphero-plasts. The bacteria were then disrupted by expo-sure to high pressure (1,000 to 1,200 lb/in2) in a

Parr bomb (Parr Instrument Co., Moline, Ill.) fol-lowed by rapid decompression to atmospheric pres-sure. The disrupted cell suspension was centrifugedat 15,000 x g for 15 min to remove cell debris andunbroken cells. The supernatant fraction wastreated with an equal volume of 2% (wt/vol) strep-tomycin sulfate in 2 x 10-2 M phosphate buffer, pH6.8, containing 10 2 M sodium azide and 10-2 M 2-mercaptoethanol (standard phosphate buffer) andcentrifuged to remove precipitated nucleic acids. Inthis and subsequent steps, the preparations were

maintained at between 0 and 5°C. The supernatantfraction was treated with ammonium sulfate inthree successive precipitation steps, using, respec-tively, 22, 6, and 10 g of solid ammonium sulfateper 100 ml of supernatant fluid. The precipitatedmaterial was collected by centrifugation after eachaddition of ammonium sulfate and suspended in

standard phosphate buffer. The fraction obtainedafter the second addition of ammonium sulfate con-tained the bulk of the glucose 6-phosphate dehydro-genase activity (Table 1, step 2). This fraction,which contained both species of glucose 6-phosphatedehydrogenase, was dialyzed against standard phos-phate buffer to remove ammonium sulfate and ap-plied to a column (2.5 by 45 cm) of diethylamino-ethyl-cellulose equilibrated with the same buffer.Elution of the two glucose 6-phosphate dehydrogen-ase species was accomplished by pumping 1 liter of

a linear 0.02 to 0.2 M gradient of phosphate buffer,pH 6.8, containing 10- 2 M sodium azide and 10-2 M2-mercaptoethanol through the column. Fractionswere collected and assayed for glucose 6-phosphatedehydrogenase activity. The fatty acid-sensitive en-zyme eluted first at a phosphate concentration of 8x 10-2 M, followed by the ATP-sensitive dehydro-genase at a phosphate concentration of 1.3 x 10-M. The two isoenzymes were completely separatedin this step. The fractions containing ATP-sensitiveglucose 6-phosphate dehydrogenase activity werepooled and stored at 0°C for use in experimentscomparing the antigenic and kinetic properties ofthe two isoenzymes. The fractions containing fattyacid-sensitive enzyme were pooled and concentratedby ultrafiltration under N2 (25 lb/in2 to a volume of2 ml in an Amicon model 50 pressure cell fittedwith a PM-10 membrane (Amicon Co., Lexington,Mass.). The concentrated material was dialyzedagainst standard phosphate buffer and then appliedto a column (1 by 15 cm) containing NADP-Sepha-rose, which was equilibrated with the same buffer.NADP-Sepharose was prepared according to theprocedure of Larsson and Mosbach (5). The columnwas washed with 50 ml of equilibration buffer toremove proteins that failed to bind to the NADP-Sepharose. No glucose 6-phosphate dehydrogenaseeluted from the column in this step. Finally, equili-bration buffer containing 5 x 10-i M NADP waspumped through the column to elute the fatty acid-sensitive glucose 6-phosphate dehydrogenase. Frac-tions containing glucose 6-phosphate dehydrogenaseactivity were pooled and stored at 0°C for subse-quent use in studies of the subunit composition ofthe enzyme and for preparation of specific antise-rum. The steps of the purification procedure aresummarized in Table 1 and discussed in Results.

Acrylamide gel electrophoresis. Preparation ofgels and buffers for examination of the homogeneity

TABLE 1. Purification of the ftatty acid-sensitileglucose 6-phosphate dehydrogenase from P. cepacia

Fraction Sp acta Re-(ml) (mglml) covery

(1) Crude extract 720 17.2 0.046 100(2) Ammonium sul-

fate precipitateII .............. 30 35.0 0.39 " 83

(3) Diethylamino-ethyl-cellulosecolumn eluate 46 2.9 2.25 61

(4) NADP-Sepha-rose columneluate ......... 3 0.6 24.2 9

Micromoles of NADPH formed per minute permilligram of protein at 24'C.

b The values represent only the NADP-linkedactivity associated with the fatty acid-sensitive glu-cose 6-phosphate dehydrogenase. A correction forNADP-linked activity associated with the ATP-sen-sitive isoenzyme was made by subtracting one-halfof the value for the NAD-linked activity from thetotal NADP-linked activity, as described earlier (7).

J. BACTERIOL.

on February 4, 2020 by guest

http://jb.asm.org/

Dow

nloaded from

P. CEPACIA GLUCOSE 6-PHOSPHATE DEHYDROGENASE 557

and molecular weight of the purified fatty acid-sensitive glucose 6-phosphate dehydrogenase wasessentially as described by Davis (2). Protein re-solved on the gels was stained with 0.01% (wt/vol)amido black in 7% (vol/vol) acetic acid. The positionof glucose 6-phosphate dehydrogenase was con-firmed by incubating the gels in assay mixturessupplemented with phenazine methosulfate (50 ,ugIml) and 2,3,5-triphenyl tetrazolium chloride (300,ig/ml) as described earlier (7). The molecularweight of the holoenzyme was estimated by compar-ing its electrophoretic mobility on gels of differentacrylamide concentration with those of referenceproteins of known molecular weight, as detailedearlier (4, 7).The subunit composition of the fatty acid-sensi-

tive glucose 6-phosphate dehydrogenase was deter-mined according to the method of Talbot and Yphan-tis (12). In this procedure proteins are dansylatedbefore resolution to permit fluorescent monitoringof the subunits during acrylamide gel electrophore-sis in the presence of sodium dodecyl sulfate (SDS).Preparations of the purified enzyme and referenceproteins were dansylated in 1-ml reaction mixturescontaining 10-1 M Tris-acetate buffer (pH 8.2), 5%(wt/vol) SDS, and 20 ,ul of dansyl (1-dimethylamino-5-naphthalenesulfonyl)-chloride (0.1% in acetone).The mixture was mixed vigorously and placed in aboiling-water bath for 5 min. 2-Mercaptoethanolwas added to a final concentration of 1% (vol/vol),and the samples were maintained at 100°C for 1additional min. The mixtures were cooled and eitherdialyzed against 10-1 M Tris-acetate buffer (pH 8.2)or passed through a column (1 by 15 cm) of Bio-GelP-30 (Bio-Rad Laboratories, Richmond, Calif.)equilibrated with the same buffer to remove un-reacted dansyl chloride and other low-molecular-weight reaction products. Samples (0.02 to 0.05 ml)of the dansylated proteins containing bromophenolblue as a tracking dye were resolved electrophoreti-cally on 7.5% (wt/vol) acrylamide gels containing10-' M Tris-acetate buffer (pH 8.2) and 0.1% (wt/vol) SDS with the same buffer in the resevoirs. Themigration of the protein bands during electrophore-sis was followed by monitoring their orange fluores-cence under short-wave (254-nm) ultraviolet light.The gels were fixed in 7% acetic acid, and SDS wasremoved electrophoretically, using a Canalco geldestainer (Canalco Co., Rockville, Md.). The gelswere permanently stained by overnight incubationin 0.25% (wt/vol) Coomassie brilliant blue in 7%acetic acid and electrophoretically destained. Themolecular weight of the subunits was estimated bycomparing their migration with that of subunits ofreference proteins as detailed earlier (14).

Preparation of antiserum for use in the enzymeinhibition and immunodiffusion experiments. Sam-ples of fatty acid-sensitive glucose 6-phosphate de-hydrogenase equivalent to step 4 of Table 1 contain-ing 1 mg of protein were emulsified in Freundincomplete adjuvant and injected subcutaneouslyinto male New Zealand rabbits. A booster injectionconsisting of a similar preparation was administered5 weeks after the initial injection, and blood sampleswere obtained by cardiac puncture 1 week later.

The serum fraction was supplemented with 10-2 Msodium azide and stored at 5°C.Enzyme inhibition tests were carried out by

preincubating appropriately diluted antiserum andenzyme for 5 min in assay mixtures from whichpyridine nucleotide and glucose 6-phosphate wereomitted. The enzyme reactions were initiated byaddition of glucose 6-phosphate and NAD or NADP.

Immunodiffusion experiments were carried outwith micro-Ouchterlony plates prepared as de-scribed earlier (6). The wells of the immunodiffusionplates were charged with 1:2-diluted antiserum andwith 5 x 10-3 U of fatty acid-sensitive glucose 6-phosphate dehydrogenase or 2 x 10-2 U of ATP-sensitive isoenzyme obtained from step 3 of Table 1.

Sedimentation properties of the fatty acid-sensi-tive glucose 6-phosphate dehydrogenase. Samples(0.5 ml) of fatty acid-sensitive enzyme from a prep-aration equivalent to step 3 of Table 1 were supple-mented with 1 mg of bovine hemoglobin and appliedto 30-ml linear sucrose gradients in polyallomertubes appropriate to the SB 110 rotor of a B60ultracentrifuge (International Instruments Co.,Needham, Mass.). The gradients contained 2 x 10-2M phosphate buffer (pH 6.8), 10-2 M sodium azide,10-2 M 2-mercaptoethanol, and the indicatedamounts of glucose 6-phosphate. The gradient tubeswere centrifuged at 5°C for 15 to 40 h at a speed of24,000 rpm. The bottoms of the tubes were punc-tured, and between 25 and 30 fractions of equal sizewere collected. Portions of each fraction were as-sayed for glucose 6-phosphate dehydrogenase activ-ity. The position of hemoglobin was determined bymeasuring the absorbance of the fractions at 405nm. The apparent molecular weight of the glucose6-phosphate dehydrogenase was estimated using therelationship of Martin and Ames (9), with hemoglo-bin (molecular weight, 64,500) as the referenceprotein.

Chemicals. Solutions of dansyl chloride were pur-chased from Pierce Chemical Co., Rockford, Ill. Allother chemicals were obtained from Sigma Chemi-cal Co., St. Louis, Mo.

RESULTS

Purification of the fatty acid-sensitive glu-cose 6-phosphate dehydrogenase from P. ce-pacia. Our initial studies of the two P. cepaciaglucose 6-phosphate dehydrogenases indicatedthat the isoenzymes could be separated withoutloss of activity of either species (7). However,efforts to extensively purify the enzymes weresuccessful only in the case of the ATP-sensitivedehydrogenase (14). The ATP-insensitive spe-cies, which we have redesignated the fattyacid-sensitive enzyme on the basis of experi-ments described later in this paper, proved tobe highly labile in the final stages of purifica-tion. More recently, we have tried to obtainhomogeneous preparations of the fatty acid-sensitive enzyme by a number of novel proce-dures. As is described below, the fatty acid-

VOL. 132, 1977

on February 4, 2020 by guest

http://jb.asm.org/

Dow

nloaded from

558 CACCIAPUOTI AND LESSIE

sensitive enzyme sediments much more rapidlyin sucrose gradients containing glucose 6-phos-phate than in its absence. Thus, it seemedreasonable that the enzyme might first be re-solved from high-molecular-weight and thenfrom low-molecular-weight proteins by succes-sive centrifugation steps in the absence andpresence of glucose 6-phosphate. When thispossibility was examined, enzyme recovery andspecific activity were high in a few experiments,but the results were not always reproducible,and it was not possible to obtain sufficientquantities of homogeneous enzyme for furthercharacterization. The use of affinity chromatog-raphy with blue Sepharose 6B was also ex-plored. Cibacron blue F3GA dye coupled toSepharose has been reported to bind tightly tothe dinucleotide fold present in a variety ofdehydrogenases and kinases (3, 12). The fattyacid-sensitive glucose 6-phosphate dehydrogen-ase appeared to bind tightly to columns of thismaterial; however, we were unable to recoveractive enzyme, even after eluting the columnswith buffer solutions containing up to 10-2 MNADP or 1 M phosphate, procedures that haveproven effective with other dehydrogenases (3).In view of the previously reported lability ofthe fatty acid-sensitive dehydrogenase, it islikely that the enzyme was inactivated underthese conditions.A reexamination of affinity chromatography

using columns of NADP-substituted Sepharoseproved more successful. The fatty acid-sensitiveglucose 6-phosphate dehydrogenase boundtightly to NADP-Sepharose and could be re-covered by eluting the column with buffer con-taining NADP. The interaction of the fattyacid-sensitive enzyme with NADP-Sepharosewas relatively specific, since the enzyme failedto bind columns of NAD-Sepharose. The proce-dure we finally adopted for purification of thefatty acid-sensitive enzyme is summarized inTable 1. It includes treatment of bacterial ex-tracts with streptomycin and ammonium sul-fate to remove nucleic acids and concentratethe enzyme, chromatographic separation of thetwo glucose 6-phosphate dehydrogenase isoen-zymes on diethylaminoethyl-cellulose columns,and, finally, elution of the fatty acid-sensitiveenzyme from NADP-Sepharose columns withbuffer containing 5 x 10 -- M NADP.The overall increase in specific activity com-

pared to that of the fatty acid-sensitive enzymeestimated to be present in extracts of the bac-teria was about 600-fold. The total recovery ofthe enzyme was about 9%. The final prepara-tion maintained full activity for at least 2weeks when stored at 0°C.Homogeneity of the fatty acid-sensitive en-

J. BACTERIOL.

zyme and molecular weight of the holoen-zyme as determined by acrylamide gel elec-trophoresis. The homogeneity of the prepara-tions was examined by resolving samples con-taining 10 ,tg of protein on 4.5, 6.0, and 7.5%acrylamide gels as outlined earlier (14). Ineach case a single band of protein was detectedwhen the gels were stained with amido blackor Coomassie brilliant blue. The position of theprotein bands corresponded to the bands ofreduced tetrazolium dye when identical gelswere incubated in glucose 6-phosphate dehydro-genase assay mixtures supplemented with tet-razolium chloride and phenazine methosulfate.The results indicated the preparations werehomogeneous. When the mobilities of the hol-oenzyme in gels of different acrylamide concen-tration were compared with those of proteinsof known molecular weight, as reported earlier(7), a molecular weight estimate of 120,000 wasobtained.Subunit composition of the fatty acid-sen-

sitive glucose 6-phosphate dehydrogenase.Subunit preparations of the fatty acid-sensitiveenzyme, which had been dissociated by treat-ment with SDS and 2-mercaptoethanol (seeMaterials and Methods), were resolved byacrylamide gel electrophoresis in the presenceof SDS, and the molecular weight of the sub-units was determined as detailed earlier (14).Only one size subunit was detected on the gels,which corresponded in position to a polypeptidechain of 60,000 molecular weight. The resultsindicated that the holoenzyme is a dimer com-posed of similar size subunits of about 60,000daltons.Antigenic properties of the fatty acid-sen-

sitive glucose 6-phosphate dehydrogenase.Since the molecular weights of the subunitscomprising the fatty acid-sensitive enzymewere similar to those of the ATP-sensitiveglucose 6-phosphate dehydrogenase (14), wewere interested in examining whether or notthe two enzymes shared common antigenic de-terminants, as would be expected if they weredifferent oligomeric forms of the same proteinor if the genes encoding them had a commonevolutionary origin. Antiserum preparedagainst preparations of the fatty acid-sensitiveenzyme equivalent to step 4 of Table 1 wasused in enzyme inhibition and immunodiffu-sion experiments to test the possibility of arelationship between the two isoenzymes. Thedata of Table 2 indicate that the antiserummarkedly inhibited the homologous fatty acid-sensitive enzyme but failed to inhibit the activ-ity of the ATP-sensitive glucose 6-phosphatedehydrogenase.The results were confirmed by double-diffu-

on February 4, 2020 by guest

http://jb.asm.org/

Dow

nloaded from

P. CEPACIA GLUCOSE 6-PHOSPHATE DEHYDROGENASE 559

TABLE 2. Influence ofantiserum prepared againstthe fatty acid-sensitive glucose 6-phosphate

dehydrogenase on the activity of the homologousenzyme and the ATP-sensitive isoenzyme

Relative glucose 6-phosphatedehydrogenase activitya

Overall dilution of anti- NADP-linked NAD-linkedserum in the assay mix-

activity of the activity of theturefatty acid- ATP-sensi-

sensitive en- tive isoen-zyme zyme

No antiserum 100 1001:5,000 561:3,000 301:800 51:500 21:100 1001:50 100

a The assay mixtures contained 0.003 U of fattyacid-sensitive enzyme or 0.002 U of ATP-sensitiveenzyme (specific activity, 3.0) obtained from step 3of Table 1. The values represent (activity in thepresence of antiserum/activity in the absence ofantiserum) x 100.

sion experiments (see Materials and Methods)in which the antiserum reacted with the homol-ogous fatty acid-sensitive enzyme to form pre-cipitin bands but failed to react with the ATP-sensitive enzyme. The experiments are consist-ent with earlier results showing that antiserumprepared against homogeneous preparations ofthe ATP-sensitive glucose 6-phosphate dehy-drogenase did not react with the fatty acid-sensitive enzyme (14) and indicate that thetwo glucose 6-phosphate dehydrogenase isoen-zymes from P. cepacia do not share commonantigenic determinants.We have routinely estimated the amount of

fatty acid-sensitive glucose 6-phosphate dehy-drogenase present in crude bacterial extractsby subtracting the contribution of the ATP-sensitive enzyme to the total NADP-linkedactivity (7). To confirm that this is a validprocedure, we determined the fraction of thetotal NADP-linked activity of crude extractsthat could be inhibited by antiserum preparedagainst the fatty acid-sensitive enzyme. Theresults agreed quite closely with those obtainedby our earlier method and indicated that about50% of the NADP-linked glucose 6-phosphatedehydrogenase activity present in extracts ofglucose-grown bacteria is associated with theATP-insensitive enzyme.

Inhibition of the ATP-insensitive glucose6-phosphate dehydrogenase by long-chainfatty acids and their acyl CoA esters. Studiesof other systems have suggested that a majorrole of the pentose shunt pathway is to provideNADPH for formation of reduced glutathione

and fatty acids (10). It might be expected thatsuch a physiological role of the pathway wouldbe reflected by inhibition of key pathway en-zymes by fatty acids and/or reduced glutathi-one, or possibly by activation by oxidized glu-tathione or fatty acid precursors such as acetateand acetyl CoA. In the course of exploringthese possibilities we noted that long-chainfatty acid acyl CoA esters strongly inhibitedthe enzyme. Similar results have been reportedfor glucose 6-phosphate dehydrogenase fromother organisms (1, 15). Free fatty acids alsoinhibited the enzyme, but as can be seen fromthe data of Table 3, which compares the concen-trations of the different fatty acids and severalof their acyl CoA esters required for half-maxi-mal inhibition, they were less effective thanthe CoA esters. The unsaturated fatty acidstested were more potent inhibitors than thecorresponding saturated fatty acids. Neitheracetyl, butyryl, propionyl, or free CoA (alltested at 10-i M) nor the reduced or oxidizedforms of glutathione (each tested at 10-2 M)influenced the activity of the enzyme.The ATP-sensitive glucose 6-phosphate de-

hydrogenase was also inhibited by long-chainfatty acid acyl CoA esters. However, between10- and 30-fold-higher concentrations were re-quired for comparable inhibition. For example,the Ki values for palmitoyl CoA were, respec-tively, 2 x 10-4 and 6 x 10-5 M for the NADP-and NAD-linked activities of the enzyme. Atthe levels tested (up to 2 x 10-4 M) the freefatty acids listed in Table 3 were not inhibitory.The preferential inhibition of the ATP-insensi-tive glucose 6-phosphate dehydrogenase bylong-chain fatty acids and their CoA esterssuggests that this enzyme might play a keyrole in formation of NADPH for reductive steps

TABLE 3. Relative inhibition of the fatty acid-sensitive glucose 6-phosphate dehydrogenase byrepresentative fatty acid acyl CoA esters and free

fatty acids

Concn requiredInhibitor for half-maximal

inhibition' (M)

Palmitoyl CoA .... ...... 7 x 10-1Stearoyl CoA ... ....... 4 x 10-6Oleoyl CoA.......... 5 x 10-6

Palmitic acid ... ....... 3 x 10-4Stearic acid ... ....... 2 x 10-4Oleic acid ... ....... 3 x 10-5Linoleic acid ... ....... 9 x 10-5Linolenic acid .... ...... 8 x 10-5

a The assay mixtures contained 0.003 U of enzymefrom a preparation equivalent to step 3 of Table 1.Inhibition curves from which these values wereobtained were hyperbolic.

VOL. 132, 1977

on February 4, 2020 by guest

http://jb.asm.org/

Dow

nloaded from

560 CACCIAPUOTI AND LESSIE

in fatty acid synthesis. Accordingly, we haveredesignated the ATP-insensitive enzyme asthe fatty acid-sensitive glucose 6-phosphate de-hydrogenase.

Influence of palmitoyl CoA on the affinityof the fatty acid-sensitive glucose 6-phosphatedehydrogenase for NADP and glucose 6-phos-phate. To examine the mechanism of inhibitionof the fatty acid-sensitive enzyme by long-chainfatty acid acyl CoA esters, the kinetic proper-ties of the enzyme were compared in the pres-ence and absence of 10-5 M palmitoyl CoA.Double-reciprocal plots of enzyme activity as afunction of NADP or glucose 6-phosphate con-centration in the presence and absence of pal-mitoyl CoA are shown in Fig. 1. As can beseen, palmitoyl CoA decreased the affinity ofthe enzyme for both NADP and glucose 6-phos-phate. In the presence of 10- M palmitoylCoA, the K,, values for NADP and glucose 6-phosphate were increased five- and threefold,respectively. The influence of inhibitor on theaffinity of the enzyme for NADP and glucose 6-phosphate fully accounted for the observed in-hibition. Palmitoyl CoA at 10-5 M did notsignificantly affect V,,,ax when the concentra-tions of glucose 6-phosphate and NADP wereincreased to 10-1 and 2 x 10-3 M, respectively.The mechanism of inhibition of the enzyme byfree fatty acids was presumably similar, sincethe inhibition by 5 x 10-5 M oleic acid wasfully reversed by increasing the concentrationof NADP and glucose 6-phosphate (results notshown).

Influence of glucose 6-phosphate on theapparent molecular weight of the fatty acid-sensitive enzyme. An unusual property of thefatty acid-sensitive enzyme which encouragedus to consider the possibility of its interconver-sion with the ATP-sensitive isoenzyme was itstendency to aggregate in the presence of glu-cose 6-phosphate (Lessie and Shuman, Abstr.Annu. Meet. Am. Soc. Microbiol. 1974, P123,p. 165). We have examined this phenomenonin more detail and find that substrate-inducedaggregation of the enzyme is strongly depend-ent upon enzyme concentration as well as glu-cose 6-phosphate concentration. Table 4 sum-marizes the results of experiments in whichthe apparent molecular weight of the enzymewas estimated by comparing its sedimentationin sucrose gradients relative to that of bovinehemoglobin. The data indicate that at rela-tively low concentrations of enzyme, glucose 6-phosphate did not significantly influence theapparent molecular weight of the enzyme. Forexample, at a concentration of 0.2 U/ml, whichis equivalent to about 8 ug of pure enzyme perml, increasing the concentration of glucose 6-

-1>

A

*0 K,, 7 2 X 0-5M(I toyl CoA 10

M)

Km = 4 X IC)M

P,.Pal-toy, C0oA.b

*/

NADP (mM)

. B

*0 K .. = S X 0-2M* (P.Imtoyl CoA 10 M)

Knn, = 5 OX 10O3ML no Pol.,toyl CoA)

**

0 2 03 0 4 0,

G6 P (mM)FIG. 1. Double-reciprocal plots of fatty acid-sen-

sitive glucose 6-phosphate dehydrogenase activityversus NADP and glucose 6-phosphate (G6P) con-centration in the absence and presence of palmitoylCoA. The reaction mixtures contained 0.003 U offatty acid-sensitive enzyme from a preparationequivalent to step 3 of Table 1 and the indicatedconcentrations of (A) NADP or (B) G6P in otherwisestandard assay mixtures.

TABLE 4. Influence ofglucose 6-phosphate (G6P) onthe apparent molecular weight of the fatty acid-sensitive glucose 6-phosphate dehydrogenase

(G6PD) a

Mol wt of G6PD at:

G6P (M) 0.2 1.0 2.0 4.0U/ml U/ml U/ml U/ml

None 99,000 119,000 125,000 107,0002 x 10-3 150,000 192,0005 x 10-3 282,000 304,0001 x 10-2 135,000 306,000 392,000 516,0002 x 10-2 120,000 310,000 407,000 580,000

a Samples (0.5 ml) of fatty acid-sensitive glucose6-phosphate dehydrogenase from a preparationequivalent to step 3 of Table 1 were applied to 30-mllinear 5 to 20% sucrose gradients containing theindicated concentrations of G6P and centrifuged for15 to 40 h at 24,000 rpm at 5°C. The apparent mo-lecular weight of the enzyme was estimated by com-paring its sedimentation relative to bovine hemo-globin, which was included in the samples.

phosphate to 2 x 10-2 M did not markedlyaffect the sedimentation properties of the en-zyme. However, as the concentration of theenzyme was progressively increased to 4 U orabout 160 jg/ml, its apparent molecular weightin the presence of 2 x 10-2 M glucose 6-phos-phate increased to a value more than fourtimes that in the absence of glucose 6-phos-phate. Increasing the concentration of enzymein the absence of glucose 6-phosphate did notsignificantly influence its sedimentation be-havior. We have estimated the intracellular

J. BACTERIOL.

---- --"

-rD..f.-----

on February 4, 2020 by guest

http://jb.asm.org/

Dow

nloaded from

P. CEPACIA GLUCOSE 6-PHOSPHATE DEHYDROGENASE 561

concentration of the fatty acid-sensitive glucose6-phosphate dehydrogenase assuming values of10-13 g for the amount of protein per cell and10-12 ml for the intracellular volume. Using avalue of 1/600 for the fraction of the cellularprotein that is fatty acid-sensitive glucose 6-phosphate dehydrogenase, our calculations in-dicate a value of about 5 U or 200 jug/ml for thein vivo concentration of the enzyme, suggestingthat the fatty acid-sensitive glucose 6-phos-phate dehydrogenase may exist in an aggre-gated state in vivo. The influence of glucose 6-phosphate on the sedimentation properties ofthe fatty acid-sensitive enzyme was relativelyspecific, since glucose 1-phosphate, fructose 6-phosphate, fructose 1,6-diphosphate, and 6-phosphogluconate (each tested at 10-2 M) failedto significantly affect the sedimentation of theenzyme. Also 2 x 10-5 M palmitoyl CoA, whichcompletely inhibited the activity of the fattyacid-sensitive enzyme, had no influence on itsapparent molecular weight.

Earlier work on the sedimentation behaviorof the active form of the ATP-sensitive glucose6-phosphate dehydrogenase from P. cepaciafailed to show any influence of glucose 6-phos-phate on its molecular weight (14). However,we have reexamined the sedimentation behav-ior of this enzyme and find that at sufficientlyhigh concentrations of enzyme, glucose 6-phos-phate also promotes a marked increase in itsapparent molecular weight. For example, whenATP-sensitive enzyme obtained from step 3 ofTable 1 was sedimented in one experiment at aconcentration of 6 U/ml, its apparent molecularweight increased fourfold in the presence of10-2 M glucose 6-phosphate. In contrast, at aconcentration of 0.2 U/ml there was no signifi-cant increase in molecular weight of the en-zyme. As with the fatty acid-sensitive glucose6-phosphate dehydrogenase, a high concentra-tion of enzyme was insufficient to increase themolecular weight unless glucose 6-phosphatewas also present. The apparent failure to ob-serve an influence of glucose 6-phosphate onthe sedimentation of the enzyme in earlierexperiments (in which the molecular weight ofthe active form of the enzyme was estimated inthe presence of NAD and glucose 6-phosphate)can be attributed to the low concentration ofenzyme used. Calculations of the in vivo con-centration of the ATP-sensitive glucose 6-phos-phate dehydrogenase, based on data obtainedearlier (14), give a value of 18 U or 129 ,ug/ml.The results suggest that this species, like thefatty-acid sensitive isoenzyme, may exist in anaggregated state in vivo.

Activity of the fatty acid-sensitive glucose6-phosphate dehydrogenase with NAD as co-

factor. Earlier studies of the fatty acid-sensi-tive glucose 6-phosphate dehydrogenase indi-cated that the enzyme was not active withNAD as cofactor. However, in the course ofpurifying the enzyme we noted that prepara-tions from which the ATP-sensitive isoenzymehad been removed still possessed significantactivity with NAD, which amounted to about5% of the activity observed with NADP. Moreextensive studies of the enzyme indicate thatthis activity is an intrinsic property of the fattyacid-sensitive enzyme. First, it was found thatthe NAD-linked activity of the enzyme couldbe increased to about 40% of that observedwith NADP by increasing the concentration ofNAD from its usual value of 5 x 10-4 to 5 x10-3 M. The poor activity of the enzyme withNAD compared to NADP was found to reflecta nearly 100-fold-lower affinity of the enzymefor NAD compared to that for NADP (therespective K,.. values for NAD and NADP were1.8 x 10-3 and 2 x 10-5 M). There was nodifference in affinity of the enzyme for glucose6-phosphate (K,,,, 5 x 10-3 M) whether NAD orNADP served as cofactor. Unlike the NADP-linked activity of the enzyme, its NAD-linkedactivity was subject to inhibition by ATP (Kj,4 x 10-3 M). Whereas the activity with NADPwas optimal over a broad range of pH valuesbetween 7.5 and 9.0, the pH optimum for activ-ity with NAD was relatively narrow, with amaximum at pH 8.5. The pH 8.5 Tris-chloridebuffer used in routine assays gave optimalactivity for both the NAD- and NADP-linkedactivities of the enzyme.The NAD-linked activity of the fatty acid-

sensitive enzyme differed in several ways fromthe major NAD-linked glucose 6-phosphate de-hydrogenase activity found in extracts of P.cepacia, which is associated with the ATP-sen-sitive isoenzyme. First, antiserum preparedagainst the fatty acid-sensitive enzyme in-hibited its NAD-linked activity, but not that ofthe ATP-sensitive enzyme. The concentrationof antiserum required for 50% inhibition of theNAD-linked activity of the homologous enzyme(1:400-fold-diluted antiserum) was about 10-foldgreater than that required for comparable in-hibition of its NADP-linked activity. Second,the NAD-linked activities of the two enzymeshad different pH optima. The activity of thefatty acid-sensitive enzyme decreased to lessthan 20% of its value at pH 8.5 as the pH wasprogressively increased to 9.5. In contrast, theNAD-linked activity of the ATP-sensitive en-zyme exhibited a broad region of optimal activ-ity between pH 8.5 and 9.5, with a slightlyhigher activity at pH 9.5 than at pH 8.5. Third,the kinetic properties of the two enzymes dif-

VOL. 132, 1977

on February 4, 2020 by guest

http://jb.asm.org/

Dow

nloaded from

562 CACCIAPUOTI AND LESSIE

fered (see Table 5). The affinity of the fattyacid-sensitive glucose 6-phosphate dehydrogen-ase for NAD was lower than that of the ATP-sensitive enzyme, and the Ki values for ATP,NADPH, and palmitoyl CoA also differed. Thedifferences in kinetic properties of the twoenzymes, which are summarized in Table 5,along with the differences in their pH optimaand antigenic properties, indicate that theNAD-linked activity of preparations of the fattyacid-sensitive enzyme does not reflect the pres-ence of trace amounts of the ATP-sensitiveenzyme.

DISCUSSION

The main aims of the present study were (i)to attempt to define the physiological role ofthe ATP-sensitive glucose 6-phosphate dehy-drogenase ofP. cepacia by examining the influ-ence of potential regulatory ligands on its activ-ity and (ii) to study the relationship of theATP-insensitive enzyme to the ATP-sensitiveisoenzyme. We were particularly interested in

investigating the possibility that the two spe-

cies of glucose 6-phosphate dehydrogenasemight share common subunits or be differentoligomeric forms of the same protein. To exam-ine this possibility, it was necessary to purifythe ATP-insensitive enzyme and compare itssubunit composition and antigenic propertieswith those of the ATP-sensitive species.

Some insight into the probable physiologicalrole of the ATP-insensitive glucose 6-phosphatedehydrogenase has been provided by the find-ing that long-chain fatty acids and their acylcoenzyme A esters inhibit the enzyme. Themost marked inhibition of the enzyme was bythe CoA esters, which inhibited the enzyme atconcentrations as low as 5 x 10-1 M. In thelight of this finding, the ATP-insensitive en-

zyme has been redesignated the fatty acid-sen-sitive glucose 6-phosphate dehydrogenase todistinguish it from the ATP-sensitive isoen-

zyme, which was relatively insensitive to suchinhibition.

Since palmitoyl CoA both decreased the af-finity of the enzyme for its substrate and cofac-tor and preferentially inhibited its activitycompared with that of the ATP-sensitive isoen-zyme, we consider the inhibition to be physio-logically significant. However, in view of thefinding by Taketa and Pogell (11) that palmit-oyl CoA was a rather general inhibitor ofseveral enzymes they examined from eucar-yotic organisms, other supportive data will benecessary before it can be concluded that theinhibition reflects a primary role of the fattyacid-sensitive glucose 6-phosphate dehydrogen-ase in supplying NADPH for reductive steps infatty acid synthesis. Experiments are currentlyin progress in our laboratory to attempt todemonstrate a role of long-chain fatty acids inrepression of the enzyme or an influence ofalterations of the enzyme on capacity of thebacteria to synthesize fatty acids or supportthe replication of lipid-containing bacterio-phage.Our second objective, that of defining the

relationship between the two glucose 6-phos-phate dehydrogenase isoenzymes, has been ac-complished by purification of the fatty acid-sensitive enzyme and comparison of its anti-genic and kinetic properties with those of theATP-sensitive enzyme. The possibility that thetwo species of P. cepacia glucose 6-phosphatedehydrogenase were different oligomeric formsof the same enzyme was suggested by (i) thesimilar size of their subunits, (ii the findingthat under appropriate conditions the fattyacid-sensitive enzyme had significant NAD-linked activity, and (iii) the observation thatthe apparent molecular weight of the enzymeincreased in the presence of glucose 6-phos-phate. However, the failure of the ATP-sensi-tive glucose 6-phosphate dehydrogenase to in-teract with antiserum prepared against thepurified fatty acid-sensitive enzyme, along withsimilar data indicating that antiserum pre-pared against the ATP-sensitive enzyme didnot interact with the fatty acid-sensitive en-

zyme (14), strongly suggests that the subunits

TABLE 5. Differences in saturation kinetics for ligands influencing the NAD-linked activities of the ATP-and fatty acid-sensitive glucose 6-phosphate dehydrogenase (G6PD) species from P. cepacia

Concn (M) required for half- Concn (M) required for half-maximal inhibi-maximal activity tion

EnzymeGlucose 6-

NAD phosphate Palmitoyl CoA ATP NADPH(NAD)

Fatty acid-sensitive G6PDa 2 x 10-3 5 x 10-3 3 x 10-1 4 x 10-3 4 x 10-5ATP-sensitive G6PDb 2 x 10-4 5 x 10-3 6 x 10-5 2 x 10-3 2 x 10-5

a The assay mixtures contained 0.003 U of enzyme from a preparation equivalent to step 3 of Table 1.b The assay mixture contained 0.002 U of enzyme obtained from step 2 of Table 1.

J. BACTERIOL.

on February 4, 2020 by guest

http://jb.asm.org/

Dow

nloaded from

P. CEPACIA GLUCOSE 6-PHOSPHATE DEHYDROGENASE 563

of the two species of glucose 6-phosphate dehy-drogenase are distinct. Furthermore, the NAD-linked activity associated with preparations ofthe fatty acid-sensitive glucose 6-phosphate de-hydrogenase from which the ATP-sensitive glu-cose 6-phosphate dehydrogenase had been re-moved has been found to be clearly differentfrom that of the ATP-sensitive enzyme by bothimmunological and kinetic studies.Further studies of the influence of glucose 6-

phosphate on the apparent molecular weight ofthe fatty acid-sensitive enzyme have shownclearly that glucose 6-phosphate-mediated ag-gregation of the enzyme does not representconversion of the enzyme to the ATP-sensitiveisoenzyme. However, the studies do suggestthat the two glucose 6-phosphate dehydrogen-ase species may exist in an aggregated form invivo. The extent of the aggregation in thepresence of glucose 6-phosphate depended uponthe concentration of enzyme as well as sub-strate. Under the conditions optimal for aggre-gation of the fatty acid-sensitive enzyme, theATP-sensitive isoenzyme, which had been pre-viously thought not to aggregate in response toglucose 6-phosphate, was also found to aggre-gate. Estimates of the intracellular concentra-tions of the two glucose 6-phosphate dehydro-genase species (5 and 18 U/ml, respectively, forthe fatty acid- and ATP-sensitive species) sug-.gest that both species of glucose 6-phosphatedehydrogenase might exist in an aggregatedstate intracellularly, assuming a sufficientlyhigh concentration of glucose 6-phosphate.Whether or not the aggregated forms of thetwo glucose 6-phosphate dehydrogenase specieshave different properties compared with thosenoted in vitro has not been examined, since itis not feasible to assay the enzymes underconditions that would maintain them in anaggregated state.

ACKNOWLEDGMENTS

This work was supported by Public Health Service grantGM21198 from the National Institute of General MedicalSciences.

LITERATURE CITED

1. Coe, E. L., and L.-H. Hsu. 1973. Acyl coenzyme Ainhibition of the Leuconostoc mesenteroides glucose-

6-phosphate dehydrogenase: a comparison of the TPNand DPN linked reactions. Biochem. Biophys. Res.Commun. 53:66-69.

2. Davis, B. J. 1969. Disc electrophoresis. II. Methods andapplication to human serum proteins. Ann. N.Y.Acad. Sci. 121:404-427.

3. Easterday, R. L., and I. M. Easterday. 1972. Affinitychromatography of kinases and dehydrogenases onsephadex and sepharose dye derivatives, p. 123-133.In R. B. Dunlap (ed.), Immobilized biochemicals andaffinity chromatography. Plenum Publishing Co.,New York.

4. Hedrick, J. L., and A. J. Smith. 1968. Size and chargeisomer separation and estimation of molecularweights of proteins by disc gel electrophoresis. Arch.Biochem. Biophys. 126:155-164.

5. Larsson, P.-O., and K. Mosbach. 1971. Preparation ofa NAD(H) polymer matrix showing coenzymic func-tion of the bound pyridine nucleotide. Biotechnol.Bioeng. 13:393-398.

6. Lee, Y. N., and T. G. Lessie. 1974. Purification andcharacterization of the two 6-phosphogluconate de-hydrogenase species from Psuedomonas multivorans.J. Bacteriol. 120:1043-1057.

7. Lessie, T. G., and J. C. Vander Wyk. 1972. Multipleforms of Pseudomonas multivorans glucose-6-phos-phate and 6-phosphogluconate dehydrogenases: dif-ferences in size, pyridine nucleotide specificity, andsusceptibility to inhibition by adenosine 5-triphos-phate. J. Bacteriol. 110:1107-1117.

8. Lowry, 0. H., N. J. Rosebrough, A. L. Farr, and R. J.Randall. 1951. Protein measurement with the Folinphenol reagent. J. Biol. Chem. 193:265-275.

9. Martin, R. C., and B. N. Ames. 1961. A method fordetermining the sedimentation behavior of enzymes:application to protein mixtures. J. Biol. Chem.193:1372-1379.

10. Scott, W. A., and E. Mahoney. 1976. Defects of glucose-6-phosphate and 6-phosphogluconate dehydrogenasein Neurospora, p. 205-236. In B. L. Horecker and E.R. Stadtman (ed.), Current topics in cellular regula-tion, vol. 10. Academic Press Inc., New York.

11. Taketa, K., and B. M. Pogell. 1966. The effect ofpalmitoyl coenzyme A on glucose-6-phosphate dehy-drogenase and other enzymes. J. Biol. Chem.241:710-726.

12. Talbot, D. N., and D. A. Yphantis. 1971. Fluorescentmonitoring of SDS electrophoresis. Anal. Biochem.44:246-253.

13. Thompson, S. T., K. H. Cass, and E. Steilwagen. 1975.Blue-dextran sepharose: an affinity chromatographycolumn for the dinucleotide fold of proteins. Proc.Natl. Acad. Sci. U.S.A. 72:669-672.

14. Vander Wyk, J. C., and T. G. Lessie. 1974. Purificationand characterization of the Pseudomonas multivoransglucose-6-phosphate dehydrogenase active with nico-tinamide adenine dinucleotide. J. Bacteriol. 120:1033-1042.

15. Vanni, P., F. Vincieri, and M. T. Vincenzini. 1976.Sensitivity of glucose-6-phosphate dehydrogenase ac-tivity to saturated and unsaturated fatty acids. Can.J. Biochem. 59:760-764.

VOL. 132, 1977

on February 4, 2020 by guest

http://jb.asm.org/

Dow

nloaded from