Biochem. J. (1970) 120, 753-761Printed in Great Britain

753

Preparation, Proteolysis and Reversible Oxidation of Highly PurifiedAzotobacter vinelandii Polynucleotide Phosphorylase

By A. T. GAJDA, GC. ZAROR DE BEHRENS Aim P. S. FITTDepartment of Biochemi8try, Univer8ity of Ottawa, Ottawa 2, Ont., Canada

(Received 13 July 1970)

1. A new method has been developed for the preparation in good yield of highlypurified Azotobacter vinelandii polynucleotide phosphorylase in its reduced form.2. Aging or digestion with trypsin causes the enzyme to develop a primer require-ment that is not eliminated by fl-mercaptoethanol. 3. The development ofa primerrequirement is accompanied by marked changes of the electrophoretic mobility ofthe enzyme in polyacrylamide gels. 4. The enzyme is inactivated by aerial oxidationor thiol-specific reagents. The lost activity is restored by fi-mercaptoethanol, butnot by oligonucleotide primers.

We have described (Gajda & Fitt, 1969, 1970) thepreviously unsuspected reversible oxidation ofAzotobacter vinelandii polynucleotide phosphorylase(nucleoside diphosphate-polynucleotide nucleo-tidyltransferase, EC 2.7.7.8). The enzyme had a

greatly decreased activity in its oxidized form andcould be readily reactivated with fl-mercapto-ethanol.The discovery afforded a probable explanation

for the lack of reproducibility (Ochoa, & Mii, 1961;Ochoa, Krakow & Basilio, 1963) and low yield(Thang, 1967) of the published purification pro-cedures and made possible the development of aneffective method for the isolation in good yield ofhighly purified A. vinelandii polynucleotide phos-phorylase in its reduced form.The present paper describes the purification

procedure and studies of the effects of aging andproteolysis on the electrophoretic properties andprimer requirement of the enzyme. We have alsoexramined further the effects of oxidation and thiol-specific reagents on the highly purified enzyme.

EXPERIMENTAL

Material1. Intermediates and enzymes were purchasedfrom the following suppliers: [14C]ribonucleoside diphos-phates and [32P]orthophosphate, Amersham-Searle Corp.,Don Mills, Ont., Canada; sodium dodecyl sulphate(specially pure), B.D.H. (Canada) Ltd., Toronto, Ont.,Canada; ApA, Gallard-Schlesinger Chemical Manufactur-ing Corp., Carle Place, N.Y., U.S.A.; frozen A. vinelandiicells and poly A, Miles Chemical Co., Elkhart, Ind., U.S.A.;DEAE-Sephadex A-50, QAE-Sephadex A-S0 and Sepha-dex G-200, Pharmacia (Canada) Ltd., Montreal, P.Q.,Canada; unlabelled ribonucleoside diphosphates, P-LBiochemicals Inc., Milwaukee, Wis., U.S.A.; bovine serum

albumin, hyaluronidase (bovine testis) and Streptomyceegri8eue protease (pronase), Sigma Chemical Co., St Louis,Mo., U.S.A.; deoxyribonuclease, ribonuclease andlysozyme, Worthington Biochemical Corp., Freehold,N.J., U.S.A. Coomassie Brilliant Blue R250wasagiftfromCanadian Industries Ltd., Montreal, P. Q., Canada.

Preparation of A. vinelandii polynucteotide pho8phoryl-a8e. All steps in the purification were carried out at0-4°C. The (NH4)2SO4 concentrations were calculatedby the procedure of Dixon & Webb (1964). Concavegradients were prepared by using three chambers of aVarigrad variable-gradient mixer (Buchler InstrumentsInc., Fort Lee, N.J., U.S.A.): the first two chambers eachcontained 0.25M-NaCl in buffer (BOOml) and the thirdchamber contained 0.45M- or0.5m-NaCl in buffer (BOOml),as indicated below. During Sephadex G-200 gel filtration,the hydrostatic pressure was always less than 10cm.

Frozen A. vinelandii cells (lOOg) were thawed, sus-pended in 3 vol. of 0.1M-tris-HCl buffer (pH8.2)-lOmM-EDTA-lOmM-.f-mercaptoethanol, and disruptedby sonication (Bronwill Biosonik II, 15min). The sus-pension was centrifuged at 28000g for 30min. The pelletwas resuspended in the same buffer and the processrepeated. The supernatants were pooled to give the crudeextract (Table 1) which was fractionated by (NH4)2SO4precipitation. The protein insoluble between 25 and 60%saturation was redissolved in 0.1 m-trils-HCI buffer(pH 8.2)-10mM-EDTA-10mmM--mercaptoethanol anddialysed overnight against two changes of lOmM-tris-HCIbuffer (pH8.2)-1 mM-EDTA-1 mx-fi-mercaptoethanol(referred to below as 'buffer'). NaCl was added to a finalconcentration of 0.1 M, and the solution passed through acolumn (2.5 cmx 80 cm) of DlEAE-Sephadex A-5O pre-viously equilibrated with 0.1 m-NaCl in buffer. Elutionwas performed at 30mI/h successively with (i) 0.1M-NaCl in buffer (400ml), (ii) a linear gradient of 0.1-0.25M-NaCl in buffer (lOOOml), (iii) 0.25m-NaCI in buffer(400ml) and (iv) a concave gradient of 0.25-0.5M-NaCl inbuffer (1500ml). The active fractions, eluted between0.28M- and 0.32M-NaCl, were combined and solid

A. T. GAJDA, G. ZAROR DE BEHRENS AND P. S. FITT(NH4)2SO4 was added to 60% saturation. The precipitatewas dissolved in buffer to give the DEAE I fraction (Table1). This solution, without prior dialysis, was passed at8ml/h through a column (2.5cmx 95cm) of SephadexG-200 previously equilibrated with buffer. The activefractions, which emerged between 143 and 187ml, werecombined to give the G-200 fraction (Table 1). The NaClconcentration was adjusted to 0.25M and the enzymeadsorbed to a column (1.5 cm x 20 cm) of DEAE-SephadexA-50 previously equilibrated with 0.25m-NaCl in buffer.The column waswashedwith 0.25M-NaCl in buffer (100ml),and the enzyme eluted with a concave salt gradient,0.25-0.45x-NaCl in buffer (1500ml). The active fractions,which emerged between 0.29m- and 0.31m-NaCl, werecombined and dialysed against buffer to give the DEAE IIfraction (Table 1). The NaCl concentration was adjustedto 0.25M and the enzyme was adsorbed to a column(1.5cmx20cm) of QAE-Sephadex A-50 previouslyequilibrated with 0.25M-NaCl in buffer. The column waswashed with 0.25x-NaCl in buffer (lOOml) and elutionwas performed with a linear gradient of 0.25-0.4m-NaClin buffer (1000ml). The active fractions emerged at about0.3m-NaCl. They were combined, dialysed againstbuffer and concentrated in a Sartorius Collodion BagApparatus [B. D. H. (Canada) Ltd.] to give the final QAEfraction (Table 1).

Purification of tryp8in-dige8ted A. vinelandii poly-nucleotide pho8phorylase. The digestion mixture (finalvolume 3ml) contained: tris-HCl buffer, pH 8.2, 12,tmol;EDTA, 1.2,Amol; ,-mercaptoethanol, 1.2,umol; trypsin,15,ug; and purified reduced polynucleotide phosphorylase(QAE fraction; 380,ug). The polynucleotide phosphoryl-ase had been stored for several weeks at 40C and its ADP-incorporation activity was stimulated 30% by addedApA. Incubation was at 37°C. Samples (0.4ml) werewithdrawn at intervals and the progress of the digestionfollowed by the procedure described in Fig. 3. Thereaction was stopped after 1 h by addition of an ice-coldsolution of soya-bean trypsin inhibitor (2.6ml; 58tzgfml).A sample (5ml) of this mixture was applied to a column(1.5cmx20cm) of DEAE-Sephadex A-50 previouslyequilibrated with buffer and the enzyme was eluted with alinear gradient of 0.1-0.8M-NaCl in buffer (500ml). Theactive fractions were combined, dialysed against bufferand concentrated in a Sartorius Collodion Bag Apparatusto give the purified digested enzyme.Assay procedures. Polynucleotide phosphorylase was

assayed by the following procedures.(a) Incorporation of nucleoside diphosphates. This

assay depends on the incorporation of [14C]nucleosidediphosphate into an acid-insoluble form (Fitt & Fitt,1967; Littauer & Kornberg, 1957). The reaction mixture(final volume 0.lml) contained: tris-HCl buffer, pH9.0,15U,mol; MgCl2, 1.0,umol; [14C]nucleoside diphosphate(about' 3000 c.p.m./,umol), 4,umol; EDTA, 0.04,umol;ApA, 10Og, where appropriate; bovine serum albumin,1Og; enzyme. Incubation was at 370C for 10min. Thereaction was stopped by addition of 0.1ml of 7% (w/v)HC104 and the mixture was kept at 00C for lOmin. Theprecipitate was collected on a Whatman GF/C glass-fibrefilter, washed four times with 1% (w/v) HC104 (2ml) andonce with 50% (v/v) ethanol (2ml), dried and the radio-activity counted. One unit was defined as the amount ofenzyme catalysing incorporation of 1 ,umol of [14C]ADP/h

into an acid-insoluble form. All specific enzymic activitiesare in terms of units of ADP-incorporation activity/mgof protein unless otherwise indicated.

(b) Phosphorolysis. This assay measures the formationof [32P]ADP during the phosphorolysis of poly A in thepresence of [32P]P1 (Fitt & Fitt, 1967; Grunberg-Manago,Ortiz & Ochoa, 1956). The reaction mixture (0.1 ml)contained: tris-HCl buffer, pH8.2, lO,umol; MgCl2,0.5,umol;EDTA, 0.03/mol; [32P]Pi (10000-100000c.p.m./,umol), 1,umol; poly A, 50,ug; bovine serum albumin, 5,ug;enzyme. Incubation was at 3700 for lOmin. The reactionwas stopped with 1% (w/v) acid-washed charcoal in 2.3%(w/v) HC104 (1 ml). The suspension was allowed to standfor 10min at 00C, with occasional mixing, and the charcoalwas then collected on a Whatman GF/C glass-fibre filter.The filter was washed three times with lOml of cold water.It was then placed upside-down in a planchet, dried andthe radioactivity counted. The specific radioactivity of[32P]P1 was determined by counting a suitable driedportion of the solution over which a GF/C filter had beenplaced. One unit was defined as the amount of enzymethat catalysed the formation of 1,umol of [32P]ADP/h inthese conditions.

(c) Exchange. This assay measures the exchange of[32P]P1 with the terminal phosphate of ADP (Grunberg-Manago et al. 1956; Thang, 1967). The reaction mixture(0.lml) contained: tris-HCI buffer, pH8.0, lOu.tmol;MgCl2, 0.4pmol; EDTA, 5nmol; ADP, 0.8,umol; [32p]p1(10000-100000c.p.m.//imol), 2,umol; ApA, lOO1ug, whereappropriate; bovine serum albumin, lO,ug; enzyme.Incubation was at 37°C for 10min. The reaction wasstopped with a charcoal-HC104 suspension (1 ml) and thecharcoal was collected, washed and the radioactivitycounted as described for the phosphorolysis assay. Theamount of phosphate incorporated was calculated fromthe expression (Grunberg-Manago et al. 1956): ,umol of Piincorporated = c.p.m. incorporated (,umol of Pi+,umol ofADP)/c.p.m. in reaction. One unit was defined as theamount of enzyme that catalysed the incorporation of1 ,umol of P1/h into ADP in these conditions.

(d) Radioactivity assay. Radioactivity was deter-mined in a Nuclear-Chicago low-background thin-windowcounter. Appropriate controls were carried out in allcases and their radioactivity was subtracted from theassay values.

(e) Protein assay. Protein was determined spectro-photometrically (Warburg & Christian, 1942).

(f) Nucleic acid assay. Total nucleic acid was deter-mined spectrophotometrically (Warburg & Christian,1942), and RNA by the orcinol reaction (Dische &Schwarz, 1937).

Analyticalpolyacrylamide-gel electrophoresis. Analyticalpolyacrylamide-gel electrophoresis was performed in atwelve-gel BuchlerPolyanalyst (BuchlerInstrumentsInc.).The procedure of Fitt, Fitt & Wille (1968) was used withthefollowingmodifications: (i) thegels contained0.2M-tris-borate buffer and 4mM-EDTA; (ii) the electrode com-partments contained 0.05M-tris-borate buffer-lmM-EDTA; (iii) the pH of the buffers was 9.0 except whereindicated; (iv) the gel concentration was 7% (w/v) exceptwhere indicated; (v) electrophoresis was performed at0.5mA/gel for 15min, then at 1.5mA/gel for 2 h, by usinga constant current power supply. Protein was (i) stainedwith Aido Black or (ii) fixed with 12.5% (w/v) trichloro-

754 1970

A. VINELANDII POLYNUCLEOTIDE PHOSPHORYLASEacetic acid and stained for not more than 1 h with0.1% (w/v) Coomassie Blue in 12.5% (w/v) trichloro-acetic acid (Chrambach, Reisfeld, Wyckoff & Zaccari,1967). To locate the enzyme, gels were incubated for1-2h in a reaction mixture for nucleoside diphosphatepolymerization and the polynucleotide formed was stainedwith Acridine Orange (Fitt et al. 1968). Gels stained withAmido Black or Acridine Orangewere destained overnightin running tap water. Those stained with Coomassie Bluewere destained overnight with 12.5% (w/v) trichloroaceticacid and the position of the bands recorded immediatelyafter destaining, since prolonged exposure to dilutetrichloroacetic acid caused the gels to shrink. Nucleicacids were stained with 0.2% (w/v) Methylene Blue in0.4M-sodium acetate buffer, pH4.7 (Peacock & Dingman,1967) and the gels destained with running tap water.

Preparative polyacrylamide-gel electrophoresis. Theseexperiments were performed with a Shandon PreparativeAcrylamide Electrophoresis Apparatus (kindly loanedby Dr C. M. Johnson). The equipment was operatedaccording to the manufacturer's instruction manual,using a 7% (w/v) gel containing 0.2M-tris-borate buffer(pH9)-4mm-EDTA. The electrode compartments con-tained 0.05M-tris-borate buffer (pH9)-lmm-EDTA andbuffer of the same composition was used for elution. Theannular running gel (above the elution chamber) was3.5cm long and approx. 3mm across. After preparationofthe gel and assembly ofthe apparatus, the equipment wasplaced in a cold-room and cold water was passed throughthe outer jacket and inner 'cold-finger'. The flow ofbuffer through the elution chamber was started andmaintained at 5ml/h for the duration of the experiment.After an initial 1 h pre-run at 200V (about 14mA), theenzyme sample, containing 10% (w/v) sucrose and 10mm-GSH, was run on to the surface of the gel under the upperelectrode buffer. Electrophoresis was performed at (i)100V for lih and (ii) 200V for 24h; 3ml fractions werecollected and assayed for protein, nucleic acid andenzymic activity.

Isoelectriefocusing. These experiments were performedwith an LKB 8100-10 (llOml) Electrofocusing Columnoperated according to the manufacturer's instructions.LKB Ampholine carriers were used: pH13-10 (LKB 8141)for preliminary experiments and pH3-6 (LKB 8142) forlater experiments. The current was turned on and thevoltage raised from 300V to 1000V over several hours sothat the power never exceeded 0.5W. Electrophoresis

was continued until equilibrium (zero current) wasattained after approx. 48h. The power supply was thendisconnected and the Ampholine solution was run out ofthe column. Fractions (2 ml) were collected and their pHand enzymic activity determined immediately.

RESULTS

Purification of A. vinelandii polynucleotidepho8phoryla8e. The results of a typical purificationby the procedure described in the Experimentalsection are summnarized in Table 1. The finalproduct was purified 2900-fold, in a yield of 66% ofthe units present in the crude extract. A total of17 000 units were discarded in less pure side fractionsat various stages, so the overall recovery was about90%. It should be noted that some increase inactivity always occurs after the first DEAE-Sephadex column step, probably through removalof degradative enzymes.Although the ammonium sulphate fractionation

and subsequent dialysis do not lead to any signi-ficant purification, they are essential. If they areomitted, the enzyme is only partially adsorbed tothe DEAE-Sephadex column during the next step.Approximately 95% of the protein of the DEAE

II fraction migrates as a single sharp band contain-ing all the enzymic activity during analyticalpolyacrylamide-gel electrophoresis in the conditionsdescribed in the Experimental section. Enzymeof this purity was used in the studies described byGajda & Fitt (1969).The final product after QAE-Sephadex chromato-

graphy was subjected to analytical polyacrylamidegel electrophoresis in 5, 7 and 10% (w/v) gels, ineach case at three different pH values (7.4, 8.2 and9.0). In all these conditions, the fresh, fullyreduced enzyme appeared to migrate as a singleband of protein corresponding exactly to the bandof enzymic activity, provided Amido Black wasused for detection of the protein. Coomassie Bluestaining revealed a second, minor, slower-movingcomponent equivalent to less than 1% of the total

Table 1. Purification of A. vinelandii polynucleotide phosphoryla8eThe enzyme was purified and assayed for ADP-incorporation activity as described in the Experimental

section.ADP-incorporation activity

FractionCrude extract(NH4)2S04 25-60%DEAE IG-200DEAE II

QAE

Totalprotein (mg)212001080026440.27.034.73

E280/E260

0.650.610.960.860.750.75

Total units624006230069200502005000041200

Specific activity(units/mg)

3.05.8

262125070008700

Relativepurification

1.01.9

87417

23002900

Vol. 120 755

A. T. GAJDA, G. ZAROR DE BEHRENS AND P. S. FITT

sample placed on the gel. This minor componentwas inactive and was still present in samplesdigested exhaustively before electrophoresis withtrypsin, pronase (which completely eliminated thepolynucleotide phosphorylase band), deoxyribo-nuclease, ribonuclease, hyaluronidase or oc-amylase.The two bands could be separated cleanly bypreparative gelelectrophoresis (seetheExperimentalsection). Unfortunately, the method was of nopreparative value, because only 18% of the proteinor enzymic activity applied to the gel could berecovered, even when pre-electrophoresis wasperformed for 1h before application of the sampleand the latter contained GSH, which has beenshown to improve the recovery of some enzymesfrom polyacrylamide gels (Ray, Troisi & Rappaport,1969). If the pre-electrophoresis and GSH wereomitted, less than 5% of the original enzymicactivity was recoverable. (It should be noted thatthe recoveries of incorporation and phosphorolysisactivities were identical.) The separated enzymeand 'anomalous' component retained their originalmobilities during subsequent analytical poly-acrylamide-gel electrophoresis. The isolated 'ano-malous' band substance was unaffected by digestionwith the enzymes mentioned above or by lysozymeand its nature is as yet uncertain.The purified enzyme (QAE fraction) was closely

associated with nucleic acid. The E280/E260 ratiowas 0.75 and the orcinol reaction confirmed thepresence of 7% (w/w) of RNA. Methylene Bluestaining showed that this RNA, or a significantpart of it, migrated with the active protein duringanalytical gel electrophoresis and the electro-phoretically purified enzyme obtained by prepara-tive gel electrophoresis had an extinction ratio of0.83 corresponding to 6% (w/w) of nucleic acid.Although the extinction ratio rose to 1.2-1.3 afterthe gel filtration step during some purifications, thefinal value for the QAE fraction was always0.7-0.8. The enzyme-associated nucleic acid couldnot be detected on polyacrylamide gels withAcridine Orange unless very large samples wereused, so it did not interfere with the localization ofthe enzymic activity.The pI (isoelectric point) of the purified enzyme

was determined several times by isoelectric focusingand found reproducibly to be between pH4.1 and4.2. The enzyme precipitates at this pH, which maytherefore only be an upper limit for the pl; it is alsoprobable that it corresponds to the isoelectric pointof the nucleic acid-enzyme complex. The methodcould not be used for purification, since considerableirreversible losses of activity occur during electro-focusing, probably as a result of the prolongedexposure of the enzyme to pH values below 8.2.A study of the stability of the enzyme showed

that it lost activity rapidly at pH 7.4 and was most

stable at pH8.2 or higher. The final purifiedproduct was comparatively stable when stored inlOmM-tris-HCl buffer (pH 8.2)-i mM-EDTA-I mm-,-mercaptoethanol at a protein concentration of0.4mg/ml. Incubation of the QAE fraction at370C for 12h caused a 40% loss of activity, mea-sured with or without ApA, without any detectablechange in electrophoretic mobility of the enzyme inpolyacrylamide gels. Freezing and thawing of theenzyme solution caused a rapid loss of activity, butafter 5 weeks at 40C only 15% of the original ADP-incorporation activity, measured in the presence ofApA, was lost. However, prolonged storage leadsto considerable changes in the primer requirementand electrophoretic mobility of the fully reducedenzyme (see below).

Effect of aging on the activity and primer require-ment of reduced highly purified A. vinelandii poly-nucleotide pho8phoryla8e. Table 2 shows that ADPpolymerization and ADP-Pi exchange catalysed bythe fresh, fully reduced enzyme were not stimulatedby ApA. However, CDP incorporation was stimul-ated 30-40% by added primer, and it should benoted that we have always detected a primerrequirement for optimum CDP polymerization evenwith fresh preparations that require no primer forADP polymerization (the latter are referred to belowas primer-independent enzymes). No explanationcan be given for this difference at present. Fig. 1shows that a primer requirement developed steadilyduring prolonged storage, so that after 72 days at40C ADP incorporation was stimulated 45% byadded ApA. A parallel increase in primer require-ment for CDP incorporation took place, with a risein stimulation by ApA from 30-40 to 160%. Theaddition of ,B-mercaptoethanol to the assay mediumdid not affect the need for oligonucleotide primer,thus confirming that the enzyme was still fullyreduced.The increase in primer requirement was accom-

Table 2. Compari8on of the activities of highlypurified A. vinelandii polynucleotide phosphorylase

The ADP- and CDP-incorporation activities, poly Aphosphorolysis activity and ADP-Pi exchange activityof the QAE fraction (Table 1) were measured as de-scribed in the Experimental section. All asisay mixturescontained 0.13,ug of enzyme and, where appropriate,ApA, 100 ug.

Specific activity (units/mg)

AssayADP-incorporationCDP-incorporationADP-P, exchangePhosphorolysis

Without ApA8700550046002140

With ApA895079004100

1970756

A. VINELANDII POLYNUCLEOTIDE PHOSPHORYLASE

panied by changes in mobility during analytical gelelectrophoresis (Fig. 2). The fresh enzyme migratedas a single band of protein and activity, apart fromthe faint, anomalous, slow-moving band discussedabove. As the preparation aged, a new, slightlymore mobile, band of protein and activity deve-loped. The intensity of this band steadily increasedwith time, while that of the original band dimini-shed. Neither of the two bands ofenzymic activityappeared to have an absolute primer requirement.The sensitive Methylene Blue staining procedure(Peacock & Dingman, 1967) suggested that nucleicacid was still present in both bands, but the latterwere not resolved by this method, which, in our

hands, often gave less sharp bands than the pro-

cedures used for detecting protein and enzymicactivity.

These changes in primer requirement and electro-phoretic mobility occurred with all partially or

highly purified A. vinelandii polynucleotide phos-phorylase preparations. However, the rate atwhich the primer requirement developed variedfrom one batch of enzyme to another and appearedgreater with preparations that contained relativelylittle nucleic acid. It therefore seems that thedevelopment of a primer requirement is related tothe association ofnucleic acids with the enzyme andpreliminary experiments suggest that it occurs

faster at pH9 than at 8.2.Effect of trypsin digestion on A. vinelandii poly-

nucleotide phosphorylase. Treatment of the purified,reduced enzyme with trypsin caused a rapid loss ofthe ADP- and CDP-incorporation activities withoutprimer (Fig. 3). The corresponding activities with

I .5

.4

0*co0)

0)

9

Time (days)

Fig. 1. Change in primer requirement for ADP incorpora-tion during aging of A. vinelandii polynucleotide phos-phorylase. The purified enzyme was stored at 4°Cin lOmM-tris-HCI buffer (pH 8.2)-1 mm-EDTA-1 mM--mercaptoethanol. Samples were withdrawn at theindicated times and assayed for ADP incorporation bythe standard method (see the Experimental section) withand without ApA. The results are expressed as the ratioof the activity with primer to that without.

0

._mE

0c

._1

17 30

Time (days)37 45

Fig. 2. Effect of aging on the electrophoretic mobility ofpolynucleotide phosphorylase. Analytical polyacrylamide-gel electrophoresis of the enzyme (3-8,ug/gel) used for theexperiment described in Fig. 1 was performed by thestandard procedure (see the Experimental section).Protein was stained with Coomassie Blue. (The brokenline indicates the position of the faint, anomalous banddiscussed in the Results section.) The enzymic activitywas located by the standard method and corresponded inall cases to the main band or bands of protein.

. q

0-

.,4

4--

Q

Cs

._,CB

01)

0 4 8 12 16 20 24 28 32

Time (min)

Fig. 3. Time-course of the proteolysis of A. vinelandiipolynucleotide phosphorylase by trypsin. The digestionmedium (final volume lml) contained: tris-HCI buffer,pH 8.2, 4,umol; EDTA, 0.4,tmol; ,B-mercaptoethanol,0.4,umol; trypsin, 51tg; polynucleotide phosphorylase,128Itg. Incubation was at 370C. Samples (O.lml) were

withdrawn at the indicated times and added to ice-coldsoya-bean trypsin inhibitor (O.lml; 0.5mg/ml). Thesamples were assayed for ADP- and CDP-incorporationwith and without ApA (1OOutg) as described in theExperimental section. The residual activity is expressedas a percentage of that of a similar, non-incubated controlto which the inhibitor was added before the trypsin.Identical results were obtained with and without added1OmM-,-mercaptoethanol. * *, ADP incorporation;*----, ADP incorporation with ApA; A-A,

CDP incorporation; A- - -A, CDP incorporation withApA; 0 0, phosphorolysis.

Vol. 120 757

o;o

A. T. GAJDA, G. ZAROR DE BEHRENS AND P. S. FITT

0

o

I.-

p

0 l

aLi lIl

a'..-

p a

Is

P a

30Time (min)

Fig. 4. Analytical polyacrylamide-gel electrophoresis of trypsin-digested A. vinelandii polynucleotide phos-phorylase. Samples of the digested enzyme obtained during the experiment described in Fig. 3 were analysedby gel electrophoresis. Separate electrophoretograms with 150,ul and lO,u of each sample were run simul-taneously and were stained for protein (p) with Coomassie Blue or for enzymic activity (a), respectively, asdescribed in the Experimental section. The broken line indicates the position of the faint anomalous com-ponent described in the text. The broad, fast-moving band was shown in separate experiments to be due to soya-bean trypsin inhibitor.

ApA were lost more slowly, so that a large increasein primer requirement took place. However, therewas little or no differential effect on the incorpora-tion activities with these substrates, in contrastwith the selective increase in primer requirement forCDP- and UDP-incorporation observed duringcontrolled trypsin digestion of Micrococcu8 Iy8o-deikti4u polynucleotide phosphorylase (Fitt &Fitt, 1967; Fitt et al. 1968). The phosphorolyticactivity was lost at a rate intermediate betweenthe rates of loss of the incorporation activities withand without ApA.

Fig. 4 shows that the loss of activity and increasein primer requirement of the enzyme during pro-teolysis were accompanied by changes in itselectrophoretic mobility in polyacrylamide gels.Initialy, the enzyme was partially converted into asecond, slightly more mobile, form in a mannersimilar, but presumably not identical, to the changeobserved during aging (Fig. 2). Both componentswere then destroyed, the less mobile one morerapidly than the new, faster-moving form of theenzyme. The end product was a single componentthat migrated more rapidly than either of the othertwo bands. All three bands, including this finalform, stained for protein, nucleic acid and enzymicactivity with and without ApA (provided sufficientmaterial was placed on the gel). Longer treatmentwith trypsin caused further loss of activity withoutany additional change in the electrophoretic mobi-lity of the protein. No difference was observed

Table 3. Effect of Pf-mercaptoethanol and ApA onADP incorporation by trypsein-digeated reducedA. vinelandii polynucleotide pho8phoryia8eThe digestion medium (final volume 0.25ml) contained:

tris-HCl buffer, pH8.2, 0.5Btmol; EDTA, 0.05,umol;,-mercaptoethanol, 0.05,mol; trypsin, 1.25l.g; poly-nucleotide phosphorylase, M6pg. Incubation was at37°C. Samples (501l) were withdrawn at the indicatedtimes and added to an ice-cold aqueous solution of soya-bean trypsin inhibitor (50,p; 0.25mg/ml). The controlcontained, in addition, soya-bean trypsin inhibitor(62.5,ug) and the corresponding samples were added tocold water (50tul). Portions of the digested and controlsamples were assayed for ADP-incorporation activity bythe standard assay (i) without additions, (ii) with fi-mercaptoethanol (10mm) and (iii) with ApA (100,ug).The residual activity of the digested samples is expressedas a percentage of the activity of the zero-time controlsample assayed without additions (no change occurred inthe activity of the incubated controls).

Residual activity (%)

Digestion time (min) ...

AdditionNone,-Mercaptoethanol (10mM)ApA (l00jug)

0 5 10

10098106

14 614 633 29

between the final results of trypsin digestion of thefresh and aged enzymes.At no time during the tryptic hydrolysis did any

fl-mercaptoethanol requirement develop (Table 3).

1970758

A. VINELANDII POLYNUCLEOTIDE PHOSPHORYLASE

This observation is similar to that of Fitt & Wille(1969) in their study ofthe proteolysis ofClo8tridiumperfringen8 polynucleotide phosphorylase, butdiffers from the results obtained with the M.lysodeikticu.8 enzyme (Klee, 1968; Klee & Singer,1968; Fitt et al. 1968) when it was found that theincrease in primer requirement that occurredduring trypsin digestion could be reversed by thiols.A sample of the enzyme was extensively digested

with trypsin and purified by DEAE-Sephadexchromatography. Table 4 shows that the productwas stimulated 23-fold by ApA, whereas the un-treated enzyme, which was several weeks old, wasonly stimulated 30% by added primer. The activi-ties were unchanged by addition of fl-mercapto-ethanol to the assay medium. The purified, digestedenzyme migrated as a single band of protein andactivity during polyacrylamide-gel electrophoresis,but the faint anomalous band was still present.

Finally, preliminary studies (Table 5) ofthe effectof trypsin digestion on purified oxidized A. vine-landii polynucleotide phosphorylase also suggestthat the state of oxidation of the enzyme does notinfluence the stimulation by added oligonucleotidesin a major way. It can be seen that ADP incorpora-tion in the presence of the partially digestedoxidized enzyme was stimulated to a similar extentby ApA as was ADP incorporation in the presenceof the same sample after it had been reduced, eventhough the actual incorporation in the presence of,-mercaptoethanol was at least six times that inits absence.

Effect of aerial oxidation and thiol-8peciflc reagentson reduced highlypurifiedA. vinelandiipolynucleotidephosphoryla8e. Highly purified oxidized poly-nucleotide phosphorylase can be prepared easily

Table 4. Stimulation by ApA of untreated andtryp8in-digested reduced A. vinelandii polynucleotidepho8phoryla8e

Purified, reduced polynucleotide phosphorylase wasdigested with trypsin and the product was purified byDEAE-Sephadex chromatography as described in theExperimental section. Samples of the digested enzymebefore and after purification and ofthe undigested enzymewere assayed with and without ApA (100lzg) in the stan-dard ADP-incorporation assay. It should be noted thatthe untreated enzyme had acquired a primer requirementduring storage (30% stimulation by ApA) and thatidentical results were obtained with lOmM-f-mercapto-ethanol in the assay medium.

ADP incorporation with ApAEnzyme ADP incorporation without ApA

UntreatedDigested, beforepurification

Digested, purified

1.314.0

23.0

from the reduced enzyme by the method describedin Table 6. This method is difficult to control,since channelling occurs during passage of airthrough the ion-exchanger, so the degree of oxida-tion varies from experiment to experiment. In theexperiment described in the Table, the ADP-incorporation activity of the oxidized enzyme inthe absence of fi-mercaptoethanol was only 38% ofthat measured in the presence of the thiol. How-ever, the assay with ,-mercaptoethanol showed thatall the original enzyme was still present, and com-plete reduction of the oxidized enzyme (Table 6,oxidized and reduced enzyme) proved that oxida-tion and reduction gave a product whose enzymicproperties were closely similar to those of theuntreated enzyme. The oxidized enzyme becameprogressively more oxidized during storage untilthe stimulation by ,B-mereaptoethanol was sixfold.The oxidized enzyme was very rapidly reactivatedby fl-mercaptoethanol and preincubation with thethiol was unnecessary, provided its concentration inthe assay medium was at least 1 mM.

Oxidation caused no detectable change in theelectrophoretic mobility of the enzyme during poly-acrylamide-gel electrophoresis, so the reaction mbitbe intramolecular. The optimum ADP and Mg2+concentrations of the oxidized enzyme were some-what lower than those of the reduced enzyme, but

Table 5. Effect of tryp8in-digmetion on the primerrequirement of purifled, oxidized A. vinelandiipolynucleotide pho&phoryla8e

Purified, oxidized polynucleotide phosphorylase (1 ml;5.63 units of ADP-incorporation activity/ml, assayedwithout ,-mercaptoethanol or ApA) in 1 mM-EDTA-lOmm-tris-HCI buffer, pH8.2, was incubated at 370C for3min. Trypsin (3ug/ml, 50pl) was added and incubationcontinued. Samples (0.2ml) were withdrawn at theindicated times, mixed with 20,ul of soya-bean trypsininhibitor (65Bug/ml.) and cooledinice.They wereassayed forADP-incorporation with and without ApA (100jug) inthe standard incorporation assay (see the Experimentqlsection) in the presence and absence of lOmM-f-mercapto-ethanol. The zero-time sample was obtained separatelyby adding enzyme to an appropriate mixture of trypsinand inhibitor. The results are expressed as the ratio ofthe ADP incorporation with ApA to that without ApAdetermined in the same conditions. The preparation ofoxidized polynucleotide phosphorylase by aerial oxidationof the purified, reduced enzyme is described in Table 6.

Digestion time(min)00.335

20

ADP incorporation+ ApAADP incorporation- ApA

-thiol1.351.061.582.08

+thiol1.121.321.471.91

Vol. 120 75;9

A. T. GAJDA, G. ZAROR DE BEHRENS AND P. S. FITT

Table 6. Effect of aerial oxidation on the fi-mercaptoethanol and ApA requirement8 of A. vinelandii poly-nucleotide phoaphorylaee

Purified, reduced polynucleotide phosphorylase (QAE fraction, 1800 units) was adsorbed on a column(0.5cmx 3cm) of DEAE-Sephadex A-50. The column was washed with 0.25m-NaCl-lmM-EDTA-10mM-tris-HCI buffer, pH 8.2, to remove ,B-mercaptoethanol, and air was then sucked through it at about 360ml/h for 5 h.The enzyme was eluted with 0.8M-NaCl-lmx-EDTA-10mM-tris-HCl buffer, pH8.2 (40ml). The volume oftheeluate was adjusted to 5Oml with water and the solution was divided into two equal portions. One portion wasdialysed overnight against 21 of 1 mM-EDTA-10mM-tris-HCl buffer, pH 8.2, to give the 'oxidized enzyme'; thesecond was dialysed against 21 of the same buffer containing in addition 1 mm-,-mercaptoethanol to give the'oxidized and reduced enzyme'. Finally, a further sample of the enzyme was treated as above except that airwas not sucked through the column and the elution and dialysis buffers both contained 1 mM-,-mercapto-ethanol: this gave the 'control enzyme'. The activities of the three enzyme fractions were determined in theADP-incorporation assay in the presence and absence of,B-mercaptoethanol andApA and in the phosphorolysisassay in the presence and absence of fi-mercaptoethanol. (It should be noted that the assays in the presence offi-mercaptoethanol-sawed that all the original units placed on each of the columnswererecovered.) Abbrevia-tion: BME, ,-mereaptoethanol.

ADP-incorporation activity (relative)Phosphorolysis activity

(relative)

EnzymeControlO*idizedOxidized and reduced

tko pH optimum was unchanged. In no case didoxidation cause any significant increase in primerrequirement.The reduced enzyme was inactivated by both

p-chloromercuribenzoate (Gajda & Fitt, 1969) andmercuric chloride. Allthe lost activity was restoredby 10mM- -mercaptoethanol in the assay mediumand neither reagent caused any increase in primerrequirement.

DISCUSSION

The method of purification of A. vinelandiipolynucleotide phosphorylase described in this,paper has been readily reproducible in our handsand has the additional advantage that the yield ofpurified enzyme has consistently been better than50% of the units in the crude extract. The productis at least 99% pure as judged by analytical poly-acrylamide-gel electrophoresis in several differentconditions of pH and gel concentration. Thesuccess of the method appears to depend on a com-

bination of four main factors: (1) the maintenanceof the enzyme in the fully reduced form, which wehave found to be muehesier to purify than the oxi-dized enzyme; (2) the avoidance of cellulose ion-exchangers or calcimn phosphate gel, which forsome unexplained reason yield unstable poly-

nucleotide phosphorylase even though they are

often effective in the purification itself; (3) the useof very shallow salt gradients for elution of theenzyme from ion-exchange columns; (4) the use ofpH values above 8 at all times during the purifica-tion.

Ribonucleic acid appears to be very closelyassociated with the enzyme, in agreement with theconclusions of Ochoa & Mii (1961) based on experi-ments with comparatively impure preparations ofA. vinelandii polynucleotide phosphorylase. Thisnucleic acid migrated together with the enzyme

during polyacrylamide-gel electrophoresis even

after partial proteolysis with trypsin, and thenucleic acid content of the highly purified enzymewas not changed significantly by preparative gelelectrophoresis.The readily reversible oxidation of A. vinelandii

polynucleotide phosphorylase and its reversibleinactivation by thiol-specific reagents indicate thatthe enzyme contains one or more thiol groups essen-

tial for activity. However, the state of oxidation ofthe enzyme has no major effect on its primerrequirement.

In contrast, trypsin digestion of the fully reducedenzyme does lead to a very large primer requirementaccompanied bya progressive loss ofoverall activity.This effect is paralleled by a considerable increase inthe electrophoretic mobility of the active protein.A similar but less pronounced requirement foroligonucleotide primers develops during prolongedstorage of the reduced enzyme. The electrophoreticmobility in polyacrylamide gels also increases inthis case, but the new form is only slightly more

mobile than the primer-independent enzyme. It ispossible that the gradual change in primer require-ment during storage is due to a slow hydrolysis ofan endogenous tightly bound primer, an explana-tion favoured by the observation that the primerrequirement develops more rapidly at pH 9 than

No addition1.001.001.00

+BME0.902.631.02

+ApA0.981.081.08

+BME +ApA0.972.661.08

No addition1.001.001.00

+BME1.091.221.08

760 1970

Vol. 120 A. VINELANDII POLYNUCLEOTIDE PHOSPHORYLASE 761at pH 8.2. However, gel electrophoresis shows thatnucleic acid is still present in the partially activespecies obtained after prolonged trypsin digestion.At no stage, either during storage or as a result oftrypsin digestion, did any requirement for fi-mercaptoethanol develop. The primer-requiringenzyme was evidently fully reduced and could notbe converted into a primer-independent form byadded thiols.

These results are quite different from those ob-served with trypsin-digested M. ly8odeikticuspolynucleotide phosphorylase. Partial digestionof this enzyme yields a primer-requiring form(Fitt & Fitt, 1967; Klee, 1967, 1969) whose activitycan be partially restored by thiols (Klee, 1968;Klee & Singer, 1968; Fitt et al. 1968) as well as byoligonucleotides. It is clear that in the case of theA. vinelandii enzyme the development of a primerrequirement is not related to the state of oxidationof the enzyme. This conclusion is confirmed by theeffects of trypsin digestion on the oxidized A.vinelandii polynucleotide phosphorylase; a primerrequirement also developed in this case, but thestimulationoftheproductbyApA was approximate-ly the same for the digested oxidized enzyme and itscorresponding reduced form. Thus the effect of theprimer was not altered significantly by a change inthe oxidation state ofthe partially digested enzyme.

We thank the Medical Research Council of Canada foran operating grant (to P. S. F.) and the Province of Ontariofor a Graduate Fellowship (to A. T. G.).

REFERENCES

Chrambach, A., Reisfeld, R. A., Wyckoff, M. & Zaccari, J.(1967). Analyt. Biochem. 20, 150.

Dische, Z. & Schwarz, K. (1937). Mikrochim. Acta, 2, 13.Dixon, M. & Webb, E. C. (1964). Enzyme8, 2nd ed., p. 40.New York: Academic Press Inc.

Fitt, P. S. & Fitt, E. A. (1967). Biochem. J. 105, 25.Fitt, P. S. Fitt, E. A. & Wille, H. (1968). Biochem. J. 110,

475.Fitt, P. S. & Wille, H. (1969). Biochem. J. 112, 489.Gajda, A. T. & Fitt, P. S. (1969). Biochem. J. 112, 381.Gajda, A. T. & Fitt, P. S. (1970). Proc. Can. Fedn biol.

Soc8, 13, 105.Grunberg-Manago, M., Ortiz, P. J. & Ochoa, S. (1956).

Biochim. biophy8. Acta, 20, 269.Klee, C. B. (1967). J. biol. Chem. 242, 3579.Klee, C. B. (1968). Fedn Proc. Fedn Am. SOC8 exp. Biol.

27, 295.Klee, C. B. (1969). J. biol. Chem. 244, 2558.Klee, C. B. & Singer, M. F. (1968). J. biol. Chem. 243,5094.Littauer, U. Z. & Kornberg, A. (1957). J. biol. Chem.

226, 1077.Ochoa, S., Krakow, J. S. & Basilio, C. (1963). In Method8

in Enzymology, vol. 6, p. 3. Ed. by Colowick, S. P. &Kaplan, N. 0. New York: Academic Press Inc.

Ochoa, S. & Mii, S. (1961). J. biol. Chem. 236, 3303.Peacock, A. C. & Dingman, C. W. (1967). Biochemi8try,

Easton, 6, 1818.Ray, D. K., Troisi, R. M. & Rappaport, H. P. (1969).

Analyt. Biochem. 32, 322.Thang, D. C. (1967). Bull. Soc. Cahim. biol. 49, 1773.Warburg, 0. & Christian, W. (1942). Biochem. Z. 310,

384.