ate and aminopterin of about 20 p ~ . Cibacron blue, a dye used as the ligand in an affinity adsorbent for one of the purification steps, was a potent competitive in ...
THE.JOURNAL
UP
BIOI.OCICAI. CHEMISTRY
Val. 256. No. 23, Issue of December IO. pp. 12456-12462. l Y X 1 Prrnted m 1’ S.A
Thymidylate Synthetase from Saccharomyces cerevisiae PURIFICATION AND ENZYMIC PROPERTIES* (Received for publication, January 30, 1981)
Linda F. BissonS and JeremyThornerg From the Department of Microbiology and Immunology, University of California, Berkeley, California 94720
Thymidylate synthetase of Saccharomyces cerevisiae was purified over 20,000-fold to apparenthomogeneity by a procedure involving two new affinity methods and several precautions for avoiding proteolysis. Molecular weight of the native enzyme was about 65,000, as determined by gel filtration and velocity sedimentation. Electrophoresis in polyacrylamide gels containing sodium dodecyl sulfate yielded a single band of molecular weight 30,000, suggesting that thymidylate synthetase is a dimer of very similar or identical subunits. The purified enzyme exhibited normal Michaelis-Menten kinetics toward both substrates, with apparent K,,, values for dUMP and for(-)-5,lO-methylene-tetrahydropteroylglutamate of 5 PM and 70 PM, respectively. When the pentaglutamyl form of the cofactor was used, its apparent K , was lower (7 PM), but V,, was unaltered. Reaction kinetics and product inhibition studies were most consistent with an ordered mechanism, wherein dUMP is the first substrate to bind and 7,8-dihydropteroylglutamate is thefirst product released. Halogenated analogs of the nucleotide substrate were competitive inhibitors of the yeast enzyme, with apparent Ki values for 5-fluoro-dUMPof 5 n~ and for 5-Br-dUMP of 10 PM. Analogs of the cofactor were also competitive inhibitors, with apparent Ki values for both methotrex. blue, a ate and aminopterin of about 20 p ~ Cibacron dye used as the ligand in an affinity adsorbent for one of the purification steps, was a potent competitive inhibitor with respect to either substrate,yielding apparent Ki values of 4 nM for thenucleotide binding site and 40 xm for the cofactor binding site.
proliferating cells, such as in embryonictissues, generallyhave elevated levels of thymidylate synthetase (3, 4). Most strikingly, tumor cells ( 5 ) and chemically or virally transformed cells (6) have markedly higher levels of thymidylate synthetase than the normal or untransformed controls. These latter observations have made thymidylate synthetase a target for chemotherapeutic attacks against the neoplastic state (7, 8). Furthermore, in higher eukaryotic systems thymidylate synthetase levels appear to be higher just prior to or during S phase of the cell cycle (9-11), and similar observations have been madefor other enzymes involved in supplyingprecursors for DNA synthesis (1, 12). In the lower eukaryote Saccharomyces cereuisiae, thymidylate synthetaseprovides the sole source of dTMP for DNA synthesis because this organism lacks thymidinekinase activity (13) and is not normally permeable to thymine, thymidine (14), orthymidylates (15, 16). Thymidylatesynthetase is, therefore, essential for the proliferation of wild type cells of bakers’ yeast. Indeed, we have shown previously (17) that heat-sensitive condicdc21, a locus originally defined by a tional-lethal mutation which caused arrest of the yeast cell division cycle specifically during S phase (18),is the structural gene for thymidylate synthetase. Yeast cells appear tobe similar tohigher eukaryotes in the respect that a number of functions involved in deoxyribonucleotide metabolism have been reported to display maximum activity and/or rate of synthesis during S phase, including dehydrofolate reductase (19), ribonucleotide diphosphate reductase (20), DNA polymerase (21), histones (22), and thymidylate synthetaseitself.* Very little is known in any system, however, about the molecular basis of either the cell cycledependent fluctuation in thymidylate synthetase activity or Thymidylate synthetase (5,10-CHz-H4folate:dUMP C- the control of the synthesis of this enzyme in general. The methyltransferase, EC 2.1.1.45)’ catalyzes methylation of the ease of genetic and biochemical manipulation of yeast makes pyrimidine ring of dUMP, yielding dTMP. In most eukaryotic this organism attractive for investigating regulation of the cells the level of thymidylate synthetase low, is and production expression and activityof thymidylate synthetase in a eukaryotic cell. To begin such a study it was essential to determine of dTMP by the enzyme is believed to be rate limiting for DNA replication (1, 2). In agreement with this view rapidly the molecular and catalytic propertiesof the yeast enzyme. This paper describes the purification of thymidylate syn* This investigation was supported by National Institutesof Health thetase from S. cereuzsiae, some physical and chemical feaResearch Grant GM21841. A portion of this work has appeared in preliminary form (Bisson , L., and Thorner, J. (1979) Abstr. Annu. tures of the purified protein, andkinetic studies of its enzymic activity. Because the yeastenzyme was present at a low level Meet. Am. SOC. Microbiol., p. 152). The costs of publication of this article were defrayed in part by the payment of page charges. This and was extremely susceptible to proteolysis, two new affinity article must therefore be hereby marked “advertisement” inaccordchromatography procedures for purification of thymidylate ance with 18 U.S.C. Section 1734 solely to indicate this fact. synthetase were devised, and maneuvers to avoid protease $Recipient of United States Public Health Service Predoctoral contamination during the preparationwere developed which Traineeship GM07232. This work is part of a dissertation presented may be generally applicable to other yeast enzymes. to the GraduateDivision of the University of California, Berkeley, in partial fulfillment of the requirements for the degree of Doctor of Philosophy. Present address, Department of Microbiology and Molecular Genetics, Harvard Medical School, Boston, MA 02115. § whom reprint requests should be addressed. Enzyme Nomenclature: Recommendations (1978) of the International Union of Biochemistry, Academic Press, New York, 1979. Formerly designated EC 2.1.1.b.
To
EXPERIMENTALPROCEDURES
Materials-Reagents were obtained from the following sources: dUMP and dTMP from P-L Biochemicals; 5-fluoro-dUMP, 5-BrR. Ludwig, C. S. McLaughlin, M. Belfort, and F. Maley, data to be published.
12456
Thymidylate
Yeast
dUMP, folic acid, aminopterin, phosphoglucose isomerase (yeast), enolase (yeast), PMSF," and Reactive blue 2 (Cibacron blue F3GA) from Sigma; methotrexate and benzamidine hydrochloride from Aldrich p3H]dUMP, Tris (enzyme grade), ammonium sulfate (enzyme grade), sucrose (enzyme grade), urea (enzyme grade), and Coomassie brilliant blue from Schwarz/Mann; bovine serum albumin (fatty acid poor) and hemoglobin (bovine) from Miles/Pentex; alkaline phosphatase (Escherichia coli) and pancreatic deoxyribonuclease 1 (bovine) from Worthington; a-chymotrypsinogen (bovine) from Boehringer; phosphocellulose (P11)and DEAE-cellulose (DE52) from Whatman; Affl-Gel Blue (Cibacron blue-agarose) and constituents for preparing polyacrylamide gels from Bio-Rad SDS from BDH; Sephadex G-50 and G-100 from Pharmacia; 2-mercaptoethanol from Eastman; TEOLA from Matheson, Coleman, and Bell; media components from Difco; and nitrocellulose filter disks (2.5 cm diameter, 0.45 pm pore) from Gelman. All other chemicals were of the highest grade commercially available. (*)-H4folate was prepared and stored as previously described (17). (-)-HsPteGlus and HCO-Hlfolate synthetase(EC 6.3.4.3) were the generous gifts of J. Rabinowitz (University of California, Berkeley). C3H]DNA (E. coli) was a gift from S. Linn (University of California, Berkeley). Organism and Growth Conditions-Haploid yeast strain X21801B (MATa SUC2maZgu12 CUP1 rho') was obtained from the Yeast Genetic Stock Center, Donner Laboratory, University of California, Berkeley. Cells were grown with vigorous aeration at 30 "C in rich medium (1%yeast extract,2% peptone, 2% glucose) to lateexponential phase (-450 units on a Klett-Summerson photoelectric colorimeter with No. 66 red fdter) in the three12-liter vessels of a New Brunswick fermenter (model FS-314) and were collected a t room temperature using a Sharples air-driven supercentrifuge. Cell pastes from each vessel were chilled on ice until all three vessels had been harvested. The cells were then washed by resuspension in 2 volumes of ice-cold Buffer A (20480 glycerol, IO mM 2-mercaptoethanol, 1 mM EDTA, 1 mM PMSF, 2 mM benzamidine, 50 mM TEOLA-HC1 (pH 7.5)) and recollected by centrifugation at 5000 X g for 10 min at 4 "C in a Sorvall RC-2B centrifuge. Cell pellets were then frozen at -20 "C and used within 5 days. Enzyme Assays-Thymidylate synthetase was assayed by the radiochemical procedure previously described (17), except that quantitation of the 'Hz0 generated from [5-3H]dUMP was performed in Beckman GP aqueous scintillation fluid instead of in Bray's solution. Enzyme activity was linear both with respect to theamount of protein added, either for crude extracts (0.01-2 mg) or for the purified enzyme (1-30 ng) and with respect to time (for a period of 30 min). One unit of thymidylatesynthetase activity was defined as the amount of enzyme catalyzing the release of 1 pmol of 'H per min a t 25 "C. For kinetic experiments involving concentrations of dUMP below 10 ~ L M the specific radioactivity of thissubstrate was increased 10-fold. Under the conditions used for the assay, conversion of formaldehyde and Hdfolate to 5,10-CH2-H4folatewent to completion within seconds, as determined spectrophotometrically according to Kallen and Jencks (23). For this reason, the actual concentration of the enzymatically active isomer, (-)-5,10-CH2-Hlfolate (1,24), was assumed to be equivalent to the concentration of (-)-H4folate in the initial (*)-H,folate solution, as determined just before its use by assaying with HCOHsfolate synthetase through minor modifications of the method of Rabinowitz and Pricer (25). Using purified authentic (-)-H4PteGlu6 as substrate, addition of a IO-fold excess of (rt)-H4PteGluindicated that the (+)-formof the cofactor was not at all inhibitory to thymidylate synthetase activity, in agreement with the findings for this enzyme from other organisms (26).Alkaline phosphatase was assayed essentially according to Garen and Levinthal (27), but with the inclusion of 1 mM ZnC12. Activity of pancreatic deoxyribonuclease 1 was followed by its ability to degrade denatured ["]DNA to acidsoluble material (28). Hemoglobin concentration was determined by its absorbance at 430 nm (6 = 135). Protein Determination-Protein concentration of a given sample was always measured by two different techniques, the dye-binding procedure of Bradford (29) and the fdter-binding method ofschaffner and Weissman (30),using bovine serum albumin as the standard. The two estimates generally agreed to within 10%. Electrophoresis in Polyacrylamide Gels-Electrophoresis of proteins in slabs (0.1 X IO X 15 cm) of 9%polyacrylamide gel containing SDS was performed according to Laemmli (31). Electrophoresis of ~
'' The abbreviations
~~~
~
~~~
used are: PMSF, phenylmethylsulfonyl flueride; SDS,sodium dodecyl sulfate; TEOLA, triethanolamine (2,2',2"nitrilotriethanol).
Synthetase
12457
proteins in slabs of 6% polyacrylamide gel under nondenaturing conditions at pH 7.6 was done following the procedure of Clark (32). Isoelectric focusing in the pH range 3.5-10 was performed in tubes (0.25 x 15 cm) of 3.5% polyacrylamide gel containing urea according to Ames and Nikaido (33), using a final ampholine (LKB) concentration of 2%and ammonium persulfate for polymerization initiation. All gels were stained with Coomassie brilliant blue. Other Methods-For column chromatography all gradient elutions had a total volume of 6 times the bed volume and were generated with a linear gradient maker. All ion exchange resins were packed under pressure. Protein concentration of column fractions was monitored a t 280 nm with a UVflow cell (Isco, model UV-5), and salt concentration of such fractions was measured using a conductivity meter (Radiometer,model CDM26). Cibacron blue concentration was determined by its absorbance at 610 nm ( E = 13,600). Automated Edman microsequencing (34) of the purified protein was kindly performed by M. Hunkapiller and L. Hood, California Institute of Technology. RESULTS
Enzyme Purification Preparation of Crude Extract-Frozen cells (300 g wet weight) were thawed at 4 "C, resuspended in 2 volumes of icecold Buffer A, and ruptured by three passages through a chilled French pressure cell at 20,000 p s i . All subsequent operations werecarried out at 4 "C. The lysate was centrifuged in polypropylene bottles at 15,000 x g for 20 min in a Sorvall RC-ZB centrifuge to remove unbroken cells and debris. The supernatant solution was collected and was further clarified by centrifugation at 80,OOO X g for 1.5 h in screw-capped polycarbonate tubes using either a Beckman L5-50 or L2-65 ultracentrifuge to yield the crude extract. Removal of P r o t e a s e B by Ammonium Sulfate Fractionation-Solid ammonium sulfate (400 g/liter) was added over the course of 1 h with gentle stirring to the crude extract to bring the solutionto 62% of saturation. After 20 min moreof stirring, the mixture was centrifugedat 15,000 X g for 30 min. The supernatant solution was collected, and more solid ammonium sulfate (140 g/liter) was added in the same manner to yield a final solution at 80%of saturation. The precipitate was collected by centrifugation at 17,000 X g for 30 min. T h e bulk (230%) of the total recoverable thymidylate synthetase activity was containedin this precipitate. Earlier small-scale tests had indicated that some thymidy'ate synthetase activity begins to precipitate upon reaching 50% of saturation with ammonium sulfate; however, protease B, a major yeast endoproteinase, is essentially quantitatively precipitated below 60% of saturation(35, 36). Hence, the narrower (62-8076) ammonium sulfate fraction was taken in order to separate thymidylate synthetase from protease B. T h i s step was essential in order to recover any thymidylate synthetase activity during subsequent purification. Removal of Protease A and Carboxypeptidase Y by DEAEcellulose Adsorption-The ammonium sulfate precipitate was redissolved in 200 ml of Buffer A anddialyzedovernight against three changes of 2 liters each of Buffer A containing 80 mM KCl. The dialysate was passed slowly (overthe course of 2 h) through a bed (500 ml) of DEAE-cellulose in a column (5 x 25 cm)whichhadbeenequilibratedpreviouslywith Buffer A containing 80 mM KCl. The column was washed with six 100-ml portions of Buffer A containing 80 mM KC]. The entire operation required approximately 5-6 h. Under these conditions (80 mM KC1, pH 7.5), thymidylate synthetase does not bind to the resin, and activity appears, therefore, in the flow-through and in the early washes. The major advantage of this s t e p was the removal of two other major yeast proteinases, protease A and carboxypeptidaseY, since these enzymes have been reported to bind to the anion exchanger under the ionic strength and pH conditions used(37,38).This additional
Yeast Thymidylate Synthetase
12458
precaution dramatically improved theyield and the stability A. The column was washed with 10 bed volumes of Buffer A and then eluted with a gradient from 0 to 0.1 mM dUMP in of thymidylate synthetasea t later stagesof the purification. SaltElution from Phosphocellulose-Fractions of the Buffer A. Enzyme activity emerged as a single symmetrical DEAE-cellulose flow-through and washes containing thema- peak. jority of the thymidylate synthetase activity were pooled and The results of the purification described above are sumdiluted with an equal volume of Buffer A containing 80 mM marized in Table I. The total purification achieved was apKCI. This material was applied slowly to a bed (500 ml) of proximately 20,000-fold with a final yield of 108. The protein homogeneous, as judged by the single phosphocellulose in a column (5 X 25 cm) which had been obtained was apparently equilibrated previously with Buffer A containing 80 mM KCl. band observed upon electrophoresis in polyacrylamide gels The column was washed over 10 h with four bed volumes of containing SDS (Fig. 1) and as judged by the single species Buffer A containing 80 mM KC1 and then was eluted with a found after isoelectric focusing in the presence of urea (results gradient from 80 to 400 mM KC1 in Buffer A. Enzyme activity not shown). Electrophoresisof the protein in polyacrylamide emerged as a single symmetrical peak a t a salt concentration gels under nondenaturing conditions a t pH 7.6 also yielded a single stained band, which comigrated with all the detectable of about 250 mM. thymidylate synthetase activity, as determined by assay of Affinity Chromatography on Cibacron blue-Agarose-The active fractions from the phosphocellulose chromatography homogenates of slices of a second lane run in parallel (results were pooled and were made 0.1 mM in dUMP. After incubation not shown). for 1 h on ice, the material was applied slowly to a bed (5 ml) Physical and Chemical Properties of Affi-Gel blue in a column (1.4 X 3 cm) which had been previously equilibrated with Buffer A containing 80 mM KC1 Molecular Weight of the Native Enzyme and itsSubunitand 0.1 mM dUMP. The columnwaswashed with 25 bed The molecular weight of purified yeast thymidylate synthevolumes of BufferA containing 250 mM KC1 and 0.1 mM tase was determined both by molecular exclusion chromatogdUMP and then was eluted with a gradient from 0.25 to 1 M raphy through Sephadex G-100 and by centrifugation in suKC1 in Buffer A containing 0.1 mM dUMP. Enzyme activity crose density gradients, by comparison to protein standards emerged as a single symmetrical peak at a salt concentration of known molecular weight. Thymidylate synthetase activity of about 600 mM. It should be noted that enzyme activity eluted a t a position corresponding to a molecular weight of could not be eluted from this resin by either 10 mM (&)- approximately 68,000 upon gel fdtration (Fig. 2 A ) and sediH,folate or 100 mM methotrexate. If dUMP was omitted from mented a t a position corresponding to a molecular weight of the buffer during applicationof the enzyme, activity could not about 65,000 (Fig. 2B). be recovered from the resin, even with a salt concentrationas As estimatedfromits mobilitywith respecttoprotein high as 2.5 M KCl, nor could the enzyme be eluted subsestandards (Fig. 3) the single band observed upon electrophoquently by dUMP itself at 1 mM. resis in polyacrylamide gels containing SDS had a molecular DEAE-celluloseChromatography-Theactivefractions weight of about 30,000.This result suggests that the native from Affi-Gel blue chromatography were pooled and dialyzed enzyme is a dimer composed of verysimilar or identical against 6 changes of 1 liter eachof Buffer B(same composition subunits.PreliminaryNH2-terminalamino acid sequence as Buffer A, except TEOLA was replaced by Tris-HC1 (pH analysis yielded a single unambiguous order: HpN-[Gly]-Met8.2)) over the course of 40 h in order to remove dUMP. The Asp-Gly-Lys-Asn-Lys-GluGlu-Glu-Glu-Gln-T~-Leu-Asp-Leudialysate was applied toa bed (10 ml) of DEAE-cellulose in a 1 2 3 4 5 6 7 8 column (2 x 3 cm) which had been equilibrated previously with Buffer B. The column was eluted with a gradient from 0 to 80 mM KC1 in Buffer B. Enzyme activity emerged as a single symmetrical peak at a salt concentration at about 40 mM, coincident with oneof two major proteinpeaks. Substrate Elution from Phosphocellulose-To remove the final minor contaminating species, the active fractions from BSA DEAE-cellulose chromatography were pooled andapplied directly to a bed (2 ml) of phosphocellulose in a column (0.7 PGI cm X 5 cm) which had been equiliated previously with Buffer EN0
-
TABLE I Purification of thymidylate syn.thetasefrom Saccharomyces Fraction
Total pronil
units units x lo-.’ x IO..‘/
1
”
-fold
mg
Crude extract (NH4)2S04fraction DEAE-cellulose passthrough Phosphocellulose eluate (salt) Cibacron blue-agarose eluate DEAE-cellulose eluate Phosphocellulose eluate (dUMP)
-
Purification
tein
w
DRN
1
5,500
620
860
9,200 4,800
200 270
550 520
80
740
200
1.2
14
210
175
3,500
0.23
10
140
610
12,200
0.067
3
82
1,200
24,000
0.05 0.06 0.11
2.5
1 2
SO
FIG.1. Purification of yeast thymidylate synthetase analyzed by SDS-polyacrylamide gel electrophoresis. Samples of the material obtained at each step of the thymidylate synthetase preparation presented under “Results” were examined by polyacrylamide gel electrophoresis in the presence of SDS,as described under “Experimental Procedures.” Lane I , molecular weight markers: bovine serum albumin (BSA, 68,000); phosphoglucoisomerase (PGI, 51,000); enolase (ENO, 44,000); deoxyribonuclease I (DKN, 31,000). Lane 2, crude extract. Lane 3 , ammonium sulfate fraction. Lane 4, DEAE-cellulose pass-through. Lane 5, phosphocellulose eluate (salt). Lane 6, Cibacron blue-agarose eluate. Lane 7, DEAE-cellulose eluate. Lane 8, phosphocellulose eluate (dUMP). Approximately 10 jtg of total protein were applied in each lane. Arrow indicates the position of thymidylate synthetase subunit.
12459
Yeast Thymidylate Synthetase ’
i
20
“
’
T
lfMDRN
il
r 15-
10 -
buffers of moderate ionic strength near neutrality but did bind when the pH was above 8.0. Also consistent with a pI near 1 , neutrality was thefinding that when the native enzyme was subjectedtoelectrophoresisin polyacrylamide gels under nondenaturing conditions a t pH 7.6, the single protein band I observed did not migrate very far toward the anode( R F 0.1 uersus the bromphenol blue dye front).
-
Catalytic Properties
pH Optimum-The pHoptimum for yeastthymidylate synthetase activitywas 6.8-7.2, with half-maximal activity at either pH 5.8 or pH 8.5. Substrate Kinetics-For initial reaction velocity measurements using the purified enzyme,dUMP had to completely be 10 33 removed from the final phosphocellulose fraction. This was Fraction Number Volume (ml) accomplishedby fwst dilutingthe enzyme into Buffer B, FIG.2. Molecular weight of native yeast thymidylate syn- adsorbing it toDEAE-cellulose equilibrated with Buffer B in thetase. The molecular weightof the enzyme was determined by ( A ) a mini-column, and washing exhaustively with Buffer A. The molecular exclusion chromatography andby ( B ) velocity sedimentation. For gel filtration, a mixture of the purified enzyme (3,000 units) enzyme then was stripped from the DEAE-cellulose column and protein standards (in 1 ml final volume) was subjected to chro- by step elutionusing Buffer A containing 80 mM KC1. Finally, matography through a bed (250 ml) of Sephadex G-100 in a column as anadditional safeguard, theconcentrated enzymewas (2.5 X 51 cm) that had been equilibrated previously with Buffer A desalted by passage through a small column of Sephadex Gcontaining 80 mMKC1. The column was eluted with the same buffer, 50 equilibrated withBuffer A. Recovery of protein was essenand fractions of 5 ml were collected. For velocity sedimentation, a tially quantitative, and specific activity of the preparation mixture of the purified enzyme (800 units) and protein standards (in a total volume of 0.2 m l ) was layered onto the top of a 5 to 20% linear remained unaltered after theseprocedure^.^ Double reciprocal plots of the response of reaction velocity sucrose gradient prepared in BufferA containing 80 mM KC1 using a Beckman automatic density gradient former and was subjected to to dUMP concentration for the purified enzyme yielded an centrifugation at 39,000 rpm for 30 h in a polyallomar tube inan SW apparent K,,,of 6 p~ (Fig. 4). Exactly the sameK , value had 41 rotor using a Beckman L5-50 ultracentrifuge. Fractionsof IO drops been obtained previously for the activityin crude extract(17). each were collected from the bottom of the tube. Bars, thymidylate Double reciprocal plots of the response of reaction velocity to synthetase activity. Heavy arrows mark the positions of the peaks of the marker proteins: alkaline phosphatase (BAP,94,000);hemoglobin (-)-5,10-CHzH4PteGlu concentration for the purified enzyme (HEM, 64,400); and deoxyribonuclease I (DRN, 31,000). Insets are were linear overa 1000-fold concentration range and indicated plots of the logarithm of molecular weight of the standards uersus an apparent K,,, value of 70 p~ (Fig. 5A). When the response ( A ) elution volume and ( B ) fraction number. For the latter, points of reaction velocity to cofactor concentration was examined represent the average of three separate runs. Light arrow indicates for the activity in crude extract, biphasic kinetics were obthe position of the peak of thymidylate synthetase. tained (Fig. 5B), yielding two apparent K, values of 9 p~ and 70 pM, respectively. Since H4folatesin yeast cells exist only in polyglutamyl forms, with the chain being from 5-8 glutamic 1 acid residues long (39),5the response of reaction velocity to (-)-5,10-CH2-H4PteG1us concentration wasexamined using the purified enzyme. V,,, was unaffected by a pentaglutamyl side chain on the cofactor; however, the apparent K,,,, 7 p ~ , was 10-fold lower than for the monoglutamyl form of the cofactor (results not shown). Thus, the simplest interpretation of the biphasic cofactor kinetics observed in crude extract is 0 0 5 10 that a portion of the (-)-5,lO-CH2-H4PteGlu added was getRELATIVE MOBILITY ting converted bya polyglutamyl polymerizing activity, as has FIG.3. Subunit molecular weightof yeast thymidylate syn- been observed in other systems (40,41), polyglutamyl to forms thetase. Molecular weight of the polypeptide was estimated by its of the cofactor which have a higher affinity for the enzyme, electrophoretic mobility in polyacrylamide gels containing SDS, comor, such polyglutamyl forms of the cofactor were already pared to proteinsof known molecular weight. Markers werethe same as in the legend to Fig. 1, with the addition of a-chymotrypsinogen present at significant concentration in undialyzed crude ex(CHY, 25,000). Arrow indicates the mobility of the thymidylate tract ( 17). At saturating concentrationsof both substrates,a turnover synthetase subunit. number of 73 min” at 25 “C was calculated from the velocity [ 1-Lys-. Although these 17 residues represent only about 5% and the concentrationof purified enzyme used. Reaction Mechanism-Since the doublereciprocal plots of of the primary structure expected for a monomer of 30,000 molecularweight, the single uniquesequenceobtained is the response of reaction velocity to the concentrationof one consistent with identityof the polypeptide chainswhich con- substrate, atvarious concentrations of the other,yielded families of lines which intersected on the abscissa, in the case of stitute yeast thymidylate synthetase. Isoelectric Point-The PI of yeast thymidylate synthetase dUMP (Fig. 4), and just above the abscissa, in the case of was estimated from isoelectric focusing of the protein under (-)-5,10-CHz-H4folate (Fig. 5A), “random” and “ping-pong” denaturing conditions. Measurement of the pH gradient ina 4Although simpledialysisshould,intheory,havebeenmore second gel run in parallel to the one stained to reveal the position of the protein indicated that the polypeptide has an straightforward, it was observed that severe losses of both enzyme activityand total proteinoccurred upon dialysis of the purified approximate isoelectric point of 6.7. This value is in agreement enzyme, especially at low protein concentrations. Therefore, dialysis was not retained was avoided at this stage. with the observation that the yeast enzyme by DEAE-cellulosewhenthe resinwas equilibratedwith ’L. D’Ari and J. Rabinowitz, unpublished observations.
i
12460
Yeast Thymidylate Synthetase TABLE I1 Inhibition kinetics ofyeast thymidylate synthetase Compound
Variable substrate
Type of inhibition PM
dTMP
dUMP 5,10-CH2-H4PteGlu Fluoro-dUMP dUMP BrdUMP dUMP Methotrexate 5,10-CH2-H4PteGlu Aminopterin 5,10-CH2-H4PteGlu 15 CibacronbluedUMP 5,10-CHz-H4PteGlu
FIG.4. Relationship between reaction velocity and dUMP concentration. Purified yeast thymidylatesynthetase (30 units) was assayedwith the differentconcentrations of dUMPindicated at variousfmed concentrations of (-)-5,10-CH,-H4PteGlu,namely 1.1 mM (w), 0.55 mM (M 0.28 ), RIM 1-,( and 0.1 mM
competitive noncompetitive 0.005 competitive 10 competitive 20 competitive competitive 0.004 competitive 0.04 competitive
80 120
plotted by the methods of Dixon (44)and Cornish-Bowden (67). 5-Fluoro-dUMP was, however, much more potent, with an apparent K, of 5 nM (Table 11), approximately 2000-fold lower than that for 5-Br-dUMP. (I1“c3). Analogs of the folatecofactor were also apparently competitive inhibitors with respect to (-)-5,10-CH2-H4PteG1u. The apparent K, values for methotrexate and aminopterin were similar (Table 11). As presentedabove(seeunder“Enzyme Purification”), chromatography on Cibacron blue-agarose was a very useful step in the purification of yeast thymidylate synthetase (Table I). This adsorbent has been used for the preparation of a variety of nucleotide- andpyridine nucleotide-binding enzymes (45). T o our knowledge, however, this resin has not been applied to the purification of any folate-requiring enonly one other folate-dependent zyme, and its interaction with enzyme has been reported (46). Because active thymidylate synthetase could only be recovered from this adsorbent if bound in the presenceof dUMP, it was of interest to further characterize the natureof the interaction of the enzyme with Cibacronbluebyexamining the effect of the free dye on catalytic activity. Dixon plots indicated that Cibacron blue 0 1 2 to either was a very potent competitive inhibitor with respect ‘bS x IO (5.10-CH2-H,PteGlu. substrate (results not shown). This finding suggests that the FIG.5. Relationship between reaction velocity and 5.10-CHz- dye can potentially interactwith either substratebinding site HIPteGlu concentration. Purified yeast thyrnidylate synthetase (30 on theenzyme, but the apparentK , for the nucleotide binding units) ( A )and crude yeast extract(- 1 rng) ( B )were assayed with the site was 10-fold lower than that for the cofactor binding site different concentrations of 5,lO-CH2-H4PteGlu indicatedat various (Table 11). This latter resultsuggests an explanationfor why fixedconcentrations of dUMP, namely100 ELM (M), 10 yeast thymidylate synthetase activitycould not be recovered (A-A), 5 p~ (M), 2 PM (H and) 1 p~ ,(M The ).from the Affi-Gel blue column if the enzyme wasbound in the concentration of the enzymatically active (-)-isomer is one-half the absence of dUMP. Under such circumstances, the protein value for the total cofactor concentration givenon the abscissa. would presumably be interacting with theligand on the resin at anexceedingly high affinity site andcould not be removed, bisubstrate reaction mechanisms for yeast thymidylate syn- except perhaps by conditions that denature and/or inactivate thetase were ruled out (42). Indeed, if the concentrations of the molecule. In contrast, when the nucleotide binding site is both substrates were varied simultaneously, but the ratioof saturated with dUMP, the enzyme would only be permitted theirconcentrations held constant, a parabolic curve was to interact with the dye onresin thethrough thelower affinity obtained, which is diagnostic of a “sequential” bisubstrate cofactor binding site, a sufficiently weaker interaction to perreaction mechanism (42, 43). mit elution of the enzymein an active form by high salt Product inhibition studies confirmed the ordered nature of concentration. the reactioncatalyzed by yeast thymidylate synthetase. With dUMP as the variable substrate, dTMP inhibitionof enzyme DISCUSSION activity was apparently competitive,as determined bya Dixon Thymidylate synthetase was purified to apparenthomogeplot (44).If (-)-5,10-CH2-H4folate wasthe variable substrate, neity from a haploid strain of S. cereuisiae. The final overall a t a subsaturating concentrationof dUMP, dTMPinhibition yields, purities, and specific activities of two other purificaappeared noncompetitive. The K, for dTMP was about 0.1 tions performed on the same scale were rather similar to that r n M (Table 11). These results are most compatible with a presented here. The fact that a minimum of a 20,000-fold reaction mechanismin whichdUMP bindsfirst to theenzyme purification was required to obtain the pure protein indicates followed by the cofactor, with release of Hzfolate before that yeast cells possess a very low level of this enzyme (less dTMP. than 0.005% of total soluble cell protein). Based on thernolecInhibition by Substrate Analogs-Yeast thymidylate syn- ular weight of the yeast enzyme (Fig. 2) and an estimated thetase activity was inhibited by halogenated nucleotides that value for thesoluble protein contentof a yeast cell (0.578 mg/ could be considered analogs of the dUMP substrate. Both5IO8 haploids),6 there are only about 1,000 molecules of thyfluoro-dUMP and 5-Br-dUMP appeared to be competitive ’G. R. Fink, unpublished results. inhibitors with respect to dUMP, when inhibition curves were
Yeast Thymidylate Synthetase
12461
midylate synthetase per haploid yeast cell. ISthis sufficient to fication of thymidylate synthetase were devised in this work. supply the amount of dTMP the cell requires for DNA syn- First, Cibacron blue-agarose was found to have at leasta 50thesis? Since the DNA content of a haploid yeast cell has fold higher capacityfor binding the yeastenzyme (in termsof been estimated to be about15,000 kilobase pairs, with an AT milligrams of protein per ml of resin bed) than methotrexatecontent of about 60% (47), there are approximately 4.5 X 10” aminoethylpolyacrylamide, an affinity adsorbent which has thymidylateresiduesperhaploid genome.Considering its been used in the isolation of thymidylate synthetases from turnovernumber of 73 min” and 1,000 molecules of the several other organisms (11, 57) and which we also prepared enzyme percell there is enough thymidylate synthetase activ- (58, 59) during the course of these studies. Furthermore, the sufficient dTMP for synthesis of about dye columngave as good, or better,a purification of the yeast ity at 25 “C to produce 3 genome equivalents during one generation time (-3 h at this enzyme as theanalog column. In the absenceof dUMP, yeast thymidylate synthetase would not bind a t all to the methotemperature). Measures to prevent proteolysis were absolutely necessary trexate-aminoethylpolyacrylamide;and, as explained under in order to successfully purify yeast thymidylate synthetase. “Results,” activitycould be recoveredfrom theCibacron blueS. cereuisiae possesses relatively high levels of proteolytic agarose only if the proteinwas initially bound in the presence activity (reviewed in Ref. 38) including protease A (an acidic of dUMP. The apparent Ki for Cibacron blue, even in the endoproteinase), protease B (a neutral endoproteinase), pro- presence of a saturating concentration of dUMP, is still three tease C (carboxypeptidase Y), several minor aminopeptidases, orders of magnitude lower than thatfor methotrexate (Table and metalloexoprotease. Therefore, buffer conditions were 11),perhaps explaining why affinity chromatography on Affideveloped to keep these activities in check until could they be Gel blue was more effective than chromatography on methoSecond, we found that it 7.5. trexate-aminoethylpolyacrylamide. removed. Protease A was not expected to be active at pH Twoserineproteaseinhibitors(PMSFand benzamidine) was possible to use phosphocellulose as an affinity matrix. known to block protease B and carboxypeptidase Y (48) were Haertle et al. (60) reported that E . coli thymidylate synthetase would not bind to phosphocellulose at neutral pH, but included, as was EDTA which is known to inactivate the metalloexopeptidase activity (49). Perhaps surprisingly, thy- the test wasperformed in phosphate buffer and is, therefore, midylate synthetase activity was quite stablein crude extracts, suspect. In contrast, at neutral pH, the yeast enzyme bound but if furtherstandardfractionationprocedures were at- to phosphocellulose and was eluted only by moderately high temptedwithoutanysafeguardsenzymeactivity was not KC1 concentration (-250 mM). Since yeast thymidylate synrelatively tight binding of stable and multiple peaks of activity were observed, for ex- thetase hasa PI near neutrality, the the enzyme may not reflect nonspecific electrostatic interacampleupon ion exchangechromatography.Thissituation arises from the fact that crude yeast extract contains a large tion with theanionic resin but rather may be due torecogniexcess of specific naturally occurringpolypeptide inhibitors of tion of the phosphorylated sugarpolymer through the dUMP the major yeast proteases (37,49),which can be removed from binding site of the enzyme. Consistent with this view, we could be eluted in the absence of the enzymesby ion exchangechromatography orgel filtration. found that the yeast enzyme Furthermore, since protease A destroys the inhibitor of pro- salt by a low concentration (5100VM) of dUMP. The preparation of thymidylate synthetase from bakers’ usually becomeactivated tease B, and vice versa, the proteases following an “explosion-type” time course (50). Thus, activa- yeast permits comparisonof the enzyme from a lower eukarbeen purified from tion of minute amounts of one protease due to removalof its yote to thymidylate synthetases that have peptideinhibitor, by chromatography,or as the result of both prokaryotes and higher eukaryotes. For six thymidylate full synthetases in which it hasbeen examined, the native enzyme changes in pH or salt concentration,rapidlyleadsto we found that the is a dimer of very similaror identical subunits of 30,000-35,000 activation of all proteolytic activity. Indeed, inclusion in all buffers of low molecular weightprotease inhib- molecular weight. The only exception is a report (61) of a itors was not in itself sufficient to stabilize thymidylate syn- monomeric 69,000 molecularweightenzyme frommurine thetase against degradation enough to permit purification. its tumor cells (Ehrlich ascites carcinoma). However, thymidylIt was also necessary to design early steps in thepurification ate synthetase from another neoplastic tissue of the mouse specifically to remove proteases while the concentration of (leukemia cells) is a dimer (11). All thymidylate synthetases the natural peptide inhibitors was still high. The rationales which have been studied kinetically in detail display an orfor the particular procedures used to remove proteasesA, B, dered bisubstrate reactionmechanism. Like the yeastenzyme, and Cwerediscussedin detail above (see under “Enzyme thymidylate synthetases fromLactobacillus casei (43), chick Purification”).’ When all of these precautions were followed, embryo (62), and human blast cells (63) all require dUMP enzyme activity was stable at every step in the purification, binding before cofactor binding. In contrast, the enzymes from and only a single form of the enzyme was found. Reports in Streptococcus faecalis (64) and mouse carcinoma cells (65) the literature of multiple forms, purportedly mitochondrial appear to bind the cofactor prior to binding the nucleotide. and cytoplasmic, of thymidylate synthetase (52) and other The molecular basis for thesedifferences in catalytic strategy analyzed under is not yetknown. folate-requiring enzymes(53) in yeast extracts conditions where no attempts to prevent proteolysiswere The enzyme from L. casei is undoubtedly the most thormade, are most likely the result of proteolytic artifacts. In oughly characterized thymidylate synthetase, including the fact, our earlierwork (17),as well as othergenetic (54,55) and recent elucidation of its entire primary structure (66). It is physiological (56) studies, also support the conclusion that interesting that within the limited amino acid sequence of the yeast cells possess only a single type of thymidylate synthe- yeast enzyme that has been determined to date, residues 10tase. 17 (-Glu-Gln-Tyr-Leu-Asp-Leu-[ 1-Lys-) are highly homoloTWOnew affinity chromatography procedures for the puri- gous to a region (residues 3-11) at the NH2 terminus of the bacterial protein (-Glu-Gln-Pro-Tyr-Leu-Asp-Leu-Ala-Lys-). ’ It should be noted that use of another haploid strain (20B-l2a), carrying a mutation (pep4-3)which greatly lowers the levels of active proteases A, B, and C (51),reduced but did not eliminate proteolytic destruction of yeast thymidylate synthetase during purification when no other precautions, aside from the use of protease inhibitors, were taken.
Acknowledgments-We would like to thank Linda D’Ari and Jesse Rabinowitz for their generous assistance duringall phases of this work,and Michael HunkapillerandLeeHood for performing the NHn-terminal amino acid sequence determination using their sensitive automated methods.
Thymidylate
12462
Yeast
REFERENCES
R. L. (1969) The Biochemistry of Folic Acid and Related Pteridines,pp. 231-260, Wiley-Interscience, New York 2. Kornberg, A. (1980)DNA Replication,pp. 39-85, W. H. Freeman, San Francisco 3. Carpenter, N. J . (1973) Genetics 75, 113-122 4. Parisi, E.,and DePetrocellis, B. (1976) Exp. Cell Res. 101, 59-62 5. Jackson, R.C., andNiethammer, D. (1979) in Chemistry and Biology of Pteridines (Kisliuk, R. L., and Brown, G. M., eds) pp. 665-676, Elsevier, New York 6. Langenbach, R. (1976) Biochim. Biophys. Acta 422,295-301 7. Danenberg, P. V. (1977) Biochim. Biophys. Acta 473.73-92 8. U l h a n , B., Lee, M., Martin, D. W.. Jr., and Santi, D. V. (1978) Proc. Natl. Acad. Sci. U. S. A . 75,980-983 9. Conrad, A. H. (1971) J. Biol. Chem. 246, 1318-1323 10. Roberts, D., and Loehr, E. V. (1971) Cancer Res. 31, 1181-1187 11. Rode, W., Scanlon, K. J., and Bertino, J . R. (1979) in Chemistry and Biology of Pteridines (Kislunk, R. L., and Brown, G. M., eds) pp. 484-494, Elsevier, New York 12. Follman, H. (1974) Angew. Chem. Int. Ed. Engl. 13,569-579 13. Grivell, A. R., and Jackson, J. F. (1968) J. Gen. Microbiol. 54, 307-317 14. Grenson, M. (1969) Eur. J. Biochem. 11,249-260 15. Brendel, M., and Haynes, R. H. (1972)Mol. Gen. Genet. 117,3944 16. Yee, W. S., Decker, R.W., andBrunk, C. F. (1976) Biochim. Biophys. Acta447,385-390 17. Bisson, L., and Thorner, J. (1977) J. Bacteriol. 132, 44-50 18. Hartwell, L. H. (1974) Microbiol. Rev. 38, 164-198 19. Jaenicke, L., Scholz, K., and Donike, M. (1970) Eur. J. Biochem. 13, 137-141 20. Lowdon, M., and Vitols, E. (1973) Arch. Biochem. Biophys. 158, 177-184 21. Golombek, J., Wolf, W., and Wintersberger, E. (1974) Mol. Gen. Genet. 132, 137-145 22. Moll, R.,and Wintersberger, E. (1976) Proc. Natl. Acad. Sci. U. S. A . 73, 1863-1867 23. Kallen, R. G., and Jencks, W. P. (1966) J. Biol. Chem. 241,58515863 24. Benkovic, S. J. (1980) Annu. Reu. Biochem. 49,227-251 25. Rabinowitz. J. C.. and Pricer. W. F., Jr., (1957) J. Biol. Chem. 229,3211328 26. Kisliuk, R.L., Gaumont, Y., and Baugh, C. M. (1974) J. B i d Chem. 249,4100-4103 27. Garen, A,, and Levinthal, C. (1960) Biochim. Biophys. Acta 38, 470-483 28. Thorner, J., Huang, W.-M., and Lebman, I. R. (1975) Virology 68. 338-348 --29.. Bradford, M. M. (1976) Anal. Biochem. 72,248-254 30. Schaffner, W., and Weissman, C. (1973) Anal. Biochem. 56,502514 31. Laemmli, U. K. (1970) Nature 227,680-685 32. Clark, B. J. (1964)Ann. N. Y. Acad. Sci.121,404-412 33. Ames, G. F.-L., and Nikaido, K. (1976) Biochemistry 15, 616-623 1. Blakeley,
~
"
I
~~~
Synthetase 34. Hunkapiller, M. W., and Hood, L. E. (1980) Science 207, 523-525 35. Katsunuma, T., Schott,E., Elasser, S., and Holzer, H. (1972) Eur. J. Biochem. 27,520-526 36. Schott, E.H., and Holzer, H. (1974) Eur. J. Biochem. 42,61-66 37. Lenney, J . F., and Dalbec, J . M. (1969) Arch. Biochem. Biophys. 129,407-409 38. Pringle, J. R. (1975) Methods Cell Biol. 12, 149-184 39. Bassett, R., Weir, D. G., and Scott,J. M. (1976)J. Gen. Microbiol. 93, 169-172 40. Shane, B. (1980) J. Biol. Chem. 255,5655-5662 41. McGuire, J. J., Hsieh, P., Coward, J . K., and Bertino, J. R. (1980) J. Biol. Chem. 255,5776-5788 42. Cleland, W. W. (1970) in The Enzymes (Boyer, P. D., ed) 3rd Ed, Vol. 2, pp. 1-65, Academic Press, New York 43. Daron, H. H., and Aull, J. C. (1978) J. Biol. Chem. 253,940-945 44. Dixon, M. (1953) Biochem. J. 55, 170-171 45. Thompson, S. T., C a s , K. H., and Stellwagon, E. (1975) Proc. Natl. Acad. Sci. U. S. A . 72, 699-672 46. Chambers, B. B., and Dunlap, R. B. (1979) J. Biol. Chem. 254, 6515-6521 47. Price, C. W., Fuson, G. B., ani: Phaff, H. J . (1978)Microbiol. Rev. 42, 161-193 48. Betz, H., and Holzer, H. (1973) Proceedings of the Third Znternational Symposium on Yeasts, pp. 67-80 49. Holzer, H. (1978) in Biochemistry and Genetics of Yeast (Merty, B., Horecker, B. L., and Stoppani, A. 0. M., eds) pp. 224-240, Academic Press, New York 50. Saheki, T., Matsuda, Y., and Holzer, H. (1974) Eur. J. Biochem. 47,325-332 51. Jones, E. W. (1977) Genetics 85.23-33 52. Zelikson, R., and Luzzati, M. (1977) Eur. J. Biochem. 79,285-292 53. Zelikson, R., and Luzzati, M. (1976) Eur. J. Biochem. 64, 7-13 54. Game, J. (1976) Mol. Gen. Genet. 146,313-315 55. Little, J. G., and Haynes, R. H. (1979)Mol. Gen.Genet. 168, 141151 56. Newlon, C. S., and Fangman, W. (1975) Cell 5,423-428 57. Plante, L. T. (1979) in ChemistryandBiology of Pteridines (Kisliuk, R. L., and Brown G . M., eds) pp. 267-271, Elsevier, New York 58. Erickson, J. S., and Mathews, C. K. (1971) Biochem. Biophys. Res. Commun. 43, 1164-1170 59. Kaufman, B. T. (1974) Methods Enzymol. 34B. 272-281 60. Haertle,T., Wohlrab, F., and Guschlbauer, W. (1979) Eur. J. Biochem. 102,223-230 61. Fridland, M., Langenbach, R. J., and Heidelberger, C. (1971) J. Biol. Chem. 246, 7110-7114 62. Lorenson, M.Y., Maley, G . F., and Maley, F. (1967) J . B i d . Chem. 242,3332-3344 63. Dolnick, B. J., and Cheng, Y.-C. (1977) J. Biol. Chem. 252,76977703 64. Blakeley, R. L. (1963) J . Biol. Chem. 238, 2113-2118 65. Reyes, P., and Heidelberger, C. (1965) Mol. Pharmacol. 1, 14-30 66. Maley, G. F., Bellisario, R. L., Guarino, D. V., and M a b , F. (1979) J. Biol. Chem. 254, 1301-1304 67. Cornish-Bowden, A. (1974) Biochem. J. 137, 143-144
