studies on fluorinated pyrimidines v. effects on nucleic acid

Studies on Fluorinated PyrimidinesV. Effects on Nucleic Acid Metabolism in Vitro*

L. BOSCH,!EBERHARDHARBERS,Ã•ANDCHARLESHEIDELBERGER

(McArdle Memorial Laboratory, The Medical School, University of Wisconsin, Madison 6, Wis.)

To gain further information on the mechanismsby which the fluorinated pyrimidines and theirnucleosides inhibit various processes of nucleicacid biosynthesis, studies similar to those reported in the preceding paper (6) were carriedout in vitro in glycolyzing suspensions of theEhrlich ascites cells system used by LePage forstudies of purine biosynthesis (20). In this workthe following labeled precursors were used: for-mate-C14, uracil-2-C14, orotic acid-6-C14, thymi-dine-6-H3, phosphate-P32, 5-fluorouracil-2-C14, and5-fluoroorotic-2-C14 acid.

MATERIALS AND METHODSEhrlich ascites cells were harvested from female Swiss

albino mice 6-7 days after transplantation. The ascites fluidwas discarded, and the cells were washed twice with isotonicsaline. The cells were dispersed with a loose-fitting homogeniz-er, and the incubations were carried out in 50-ml. Erlenmeyerflasks, with an equivalent of 0.25 ml. packed volume of cellsand 2 /jmoles of the labeled precursor in 12 ml. of Robinson's

medium (22) containing bicarbonate and glucose for 1 hour at87Â°C., in an atmosphere of 95 per cent nitrogen and 5 per cent

carbon dioxide. All groups of controb and cells treated withvarious concentrations of the inhibitors were run at least induplicate. The reactions were terminated by the addition ofperchloric acid, the nucleic acid bases isolated, and the specificactivities determined as described in the previous paper (6).In some experiments the cold NaCl extraction was omitted.

In one of the incubation experiments with 5-fluorouracil-2-C14, the acid-soluble fraction was chromatographed as described previously (3). The FUMP1 fraction was purified andseparated from UMP by rechromatography on Dowex-1formate with a more gradual gradient, and was then de-phosphorylated with an intestinal phosphatase (kindly sup-

* This work was supported in part by a grant-in-aid of theWisconsin Division of the American Cancer Society; a grant,C-2832, from the National Cancer Institute, National Institutes of Health, U.S. Public Health Service; and by agrant-in-aid from Hoffmann-LaRoche, Inc. A preliminary report of part of this work appeared in Fed. Proc., 16, 194, 1957,and in Abstr. Am. Chem. Soc. Meeting, p. 20C, Sept. 8-13,1957.

t Netherlands Organization for Pure Research and Ful-bright Postdoctorate Fellow, on leave from the NetherlandsCancer Institute, Amsterdam.

ÃŽFulbright Research Fellow from the University of GÃ¶ttingen, Germany.

Received for publication October 25, 1957.

plied by Dr. G. A. LePage). Carrier 5-fluorouridine and 5-fluoro-2'-deoxyuridine were added to the nucleoside effluentfrom a small Dowex-1 formate column, and the ribo- and de-oxyribonucleosides were separated on a Dowex-1 chloridecolumn, with 0.03 M KO and 0.02 M NajBÃ„, as describedby Cohn (5). Under these conditions, 5-fluoro-2'-deoxyuridineis eluted in the borate system, and, following this, 5-fluorouridine is removed with 0.1 N HC1. The total radioactivities inthese fractions were determined with a Packard Tri-Carbliquid scintillation counter. The DNA and RNA were separated and the RNA nucleotides chromatographed as previously described (3).

The permeability studies were carried out as described byLeibman and Heidelberger (19). Following the incubation, theascites cells were separated by centrifugation and washedtwice with Robinson's medium to give the extracellular frac

tion, and the cells were then precipitated and extracted withperchloric acid to give the intercellular, acid-soluble fraction.The nucleotide content of the fractions obtained from labeleduracil and 5-fluorouracil was determined by the use of smallhand columns of Dowex-1 formate as described by Heidelberger et al. (17). Those fractions obtained from orotic andfluoroorotic acid were separated on the fraction collector (17).

The sources of the labeled compounds have been listedpreviously (6). All fluorinated pyramidines were obtained fromDr. Robert Duschinsky of Hoffmann-LaRoche, Inc. (7, 8).

RESULTSThe in vitro rates of incorporation of formate-

C14into DNA thymine in suspensions of Ehrlichascites cells under aerobic and anaerobic conditions are shown in Chart 1 and confirm the superiority of the anaerobic glycolytic conditions fornucleic acid biosynthesis in these cells as reportedby LePage (20). This biosynthetic pathway wasinvestigated first because of the demonstrationfrom in vivo studies that the fluorinated pyrimidines block this conversion (6, 14).

The results of several replicate experiments inwhich the inhibition of the conversion of fonnate-C14into DNA thymine was studied at various concentrations of 5-fluoroorotic acid, 5-fluorouracil,5-fluorouridine, and 5-fluoro-2'-deoxyuridine are

given in Chart 2. There was no preincubationwith the inhibitors. At the highest concentration

1Abbreviations used in this paper are as follows:FO=5-fluoroorotic acid FUDR=5-fluoro-2'-deoxyuridineFU=5-fluorouracil FUMP=5-fluorouridylic acidFUR=5-fluorouridine

335

on April 9, 2018. © 1958 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

http://cancerres.aacrjournals.org/

336 Cancer Research Vol. 18, April, 1958

used, 0.75 X 10~3M, there was a 65 per cent inhibition by 5-fluoroorotic acid of the conversion of1.67 X IO-4 M formate-C14 into DNA thymine;

and inhibition increased in the order of FO, FU,FUR, and FUDR, with the latter compounddemonstrating essentially complete inhibition ata concentration of 10~8M.

Burk and his colleagues have postulated that anumber of tumor-inhibitory compounds, includ-

o N2:C02

X 02 :C02

3000-

2000-

1000-

15 30 60 MINUTES

CHABT1.â€”Rateof incorporation of formate-C14 into DNAthymine in suspensions of Ehrlich ascites celb under aerobicand anaerobic conditions.

100-I

75-

50-

25-

trol values, except for a possible minute diminution in the glycolytic rate with 5-fluorouridine.The Qco, was 35.

Since it has been previously demonstrated thatboth 5-fluorouracil and 5-fluoroorotic acid are converted into the same acid-soluble nucleotides (3),it appeared that the difference in the inhibitoryeffect of the two compounds in vitro might reflectdifferences in the degree of permeability of theascites cells to them or differences in their rates ofconversion into nucleotides. That there was indeeda permeability effect is shown in Table 1. In thisexperiment, equimolar amounts of labeled uracil,orotic acid, 5-fluorouracil, and 5-fluoroorotic acidwere incubated with a constant cell and mediumvolume, under our usual conditions. The cellswere centrif uged, washed twice with cold medium,

FUDR

IO"3 IO"4 IO"5 IO"6 IO"7 IO"8 IO"9

MOLARITY

CHAHT2.â€”The effects of fluorinated pyrimidines on theconversion of formate-C14 into DNA thymine in suspensionsof Ehrlich ascites cells. Per cent of control specific activitiesagainst molarity of inhibitors.

ing A-methopterin and 6-mercaptopurine, exerttheir action as inhibitors of glycolysis (2). Thatthis is not the case with the fluorinated pyrimidines is shown in Chart 3, in which the rate ofglycolysis, measured as carbon dioxide evolution,is plotted against time for control cells and thosetreated with the highest concentrations (0.75 X10~3M) of the inhibitors used in the current work.There was no significant deviation from the con-

GLYCOLYSIS

MlC02/flask

40'

30

20

IO-

X CONTROL

o FUÃ»FO

D FUR+â€¢FUDR

O IO 20 30 40 50 60 70 80 90MINUTES

CHART3.â€”The lack of effect of fluorinated pyrimidineson glycolysis at a concentration of 0.75 X 10~sM; /il of CO8 X10~1/aasi plotted against time.

the proteins were precipitated and extracted with4 per cent perchloric acid, and the amount ofradioactivity in the acid-soluble and nucleic acidfractions was measured. The concentration andrelative concentrations in the cells were calculatedand compared with an equilibrium concentrationrepresenting equivalent concentration in the cellsand medium as if there were free diffusion. Thistype of experiment might be subject to criticismon the grounds that the radioactivity measured inthe acid-soluble fraction could represent materialadsorbed on the cell walls as well as that which hasactually entered the cells. That adsorption doesnot play an important role in these experiments isshown by the fact that the amounts of radioactivity in the nucleic acids parallel the quantitiesin the acid-soluble fraction. It will be noted thatonly in the case of 5-fluorouracil was the concentration in the cells as high as that in the medium;in fact, it was almost 2 times as high. In all cases,



BOSCHet al.â€”Studies on Fluorinated Pyrimidines. V 337

the labeled precursors inside the cell were converted into nucleotides and nucleic acids at similarrates. Therefore, the differences in the in vitrometabolism of these compounds are explained on apermeability basis. The greater than tenfold difference in the permeability of the cells to 5-fluoro-uracil and 5-fluoroorotic acid explains the differences in their inhibitory activities shown in Chart2. It is of interest to observe the parallel betweenthe cell permeability to the normal and fluori-nated pyrimidines. Although the magnitude of theinhibitory effects of the two ribosides appears to

acid coenzymes (10-12), the demonstration (Chart4) that the fluorinated pyrimidines do not significantly inhibit the conversion of formate intonucleic acid purines rules them out as folie acidantagonists, in agreement with the in vivo experiments (6), and establishes their role as specific inhibitors of the metabolic formation of the methylgroup of DNA thymine. In this experiment thespecific activities of the DNA thymine and mixednucleic acid adenine in the control cells were 2140and 420 counts/min//imole, respectively, whichvalues are in agreement with the data of Smellie

TABLElINCORPORATIONOFLABELEDPRECURSORSINTOEHRLICHASCITESCELLSin Vitro

80-minute incubation, 2 jumÃ³lesprecursor, 0.25 ml. packed cell volume, 11.75 ml. of medium.

Equilibriumconcentration*

Per centf in acid-solublePer centf in nucleic acidsPer centf inside cellsConcentration inside cells

(^moles/ml packed cells)Relative concentrationPer cent of acid-soluble as nu

cleotides

2.080.17

1.00

Uracil0.660.260.920.074

0.4339

Oroticacid

0.220.100.320.026

0.1566

Fluoro-uracil2.981.154.130.331.9573Fluorooroticacid0.170.120.290.0220.1366

* That concentration of drug equivalent in the medium and cells,

t Per cent of total radioactivity added per flask.

be greater than might be attributed to permeability (Chart 2), similar studies will be carried outwhen the labeled nucleosides become available.

It is well established from the work of Friedkinand Roberts (11), Blakley (1), and Phear andGreenberg (21) that the process leading to theformation of the methyl group of thymine involves the reaction of a one-carbon tetrahydro-folic acid derivative with deoxyuridine to givethymidine, although the reaction probably occurs with the corresponding nucleotides (10).Further evidence for the site of fluorinated pyrim-idine inhibition of this process is obtained fromthe experiment shown in Table 2. The addition ofdeoxyuridine increased the conversion of labeledformate into DNA thymine, and the inhibitionof this reaction produced by 5-fluorouracil waspartially reversed by deoxyuridine. The lesser inhibition caused by 5-fluoroorotic acid was reversed by deoxyuridine to such an extent that thethymine specific activity was almost as high aswas found with formate and deoxyuridine. Thereversal of the inhibitions with deoxyuridine represents further evidence that the fluorinated pyrimidines block the biological formation of themethyl group of thymine.

Because formate is a precursor of nucleic acidpurines as well as of the methyl group of thymine,both pathways being catalyzed by tetrahydrofolic

TABLE 2

THE EFFECTSOFDEOXYUHIDINEANDFLUORINATEDPYRIMIDINESONTHECONVERSIONOFFORMATE-CMINTO

DNA THYMINEIN EHRLICHASCITESCELLSFormate-C", 2 /imoles,used throughout.

DNA THTKIKI

Specific activity

COMPOUNDSNoneDeoxyuridineFluorouracilFluorouracil

+DeoxyuridineFluorooroticFluoroorotic

+Deoxyuridme

2099

2099

20

lunts/niin//imole)2180424027516609653850Percentofcontrola100195137644176

et al. (24), Harrington and Lavik (13), and withour own observations (6) that in in vitro experiments formate is more efficiently utilized forthymine than for purine biosynthesis, whereasin vivo the relative utilizations are similar.

In Chart 4 are also shown the results of experiments with tritium-labeled thymidine and phos-phate-P32. In neither of these cases was any significant inhibition of incorporation produced bythe highest concentration (0.75 X 10~8M) of any

of the fluorinated pyrimidines or nucleosides.When an equivalent amount of thymine-2-C14 wasused as the precursor, no detectable radioactivitywas found in the DNA thymine. In the in vivoex-




periments FU and FO produced an inhibition ofP32 incorporation into DNA and a stimulation ofthe conversion of thymidine into DNA thymine(6). The differences between these experimentsprobably reflect the more complex situation in theintact animal, in which drug distribution andmetabolism play an important role, and where 12hours elapsed between the injection of the labeledprecursors and the sacrifice of the mice, as compared with the 1-hour in vitro incubation.

The effects of fluorinated pyrimidines on theconversion of labeled uracil and orotic acid intoRNA uracil, DNA thymine, and mixed nucleicacid cystosine were also investigated in the in vitro

FORMATE-C14 TO AGEMINE

g ^ 100 73 76 82 64 103 77 73 87

400 fi FI 2400

X100 â€¢ S.A.

200 1000

CON. FUM. 0.75 X IO"5 -4 "3 FUDR-5 -4 -3 FUR-5 -4 -3

FO

200

100

THYMIDINE-H3 TO DNA THYMINE

P32 TO RNA

100 145 139 109 9l

CONFU FU FO FOâ€¢-4-3 -4 -3

100 IOO 75 HO 81

P32 TO DNA

CHART4.â€”The effects of fluorinated pyrimidines on theconversions of formate-C""14into mixed nucleic acid adenine,of phosphate-P32 into DNA and RNA, and of thymidine-6-H3into DNA tbymine. Specific activity against drug and concentration. The numbers over the graphs represent the per centof the control specific activities.

system. At the end of the incubation, the mediumwas separated from the cells as described previously, and the radioactivity in the acid-solublefraction was determined. As shown in Table 3,the unexpected finding was made that 5-fluorouri-dine at the highest concentration used greatlystimulated the entrance of labeled uracil and oroticacid into the cells. 5-Fluorouracil had less effect,and FO and FUDR did not affect this process.Consequently, the specific activities of the nucleicacid pyrimidines obtained from the 5-fluorouridineincubations, shown in Chart 5, were correctedfor the increased amount of precursors in thecells.

As shown in Chart 5, the effects of the inhibitory compounds on the incorporation into RNAuracil and mixed nucleic acid cytosine of bothlabeled uracil and orotic acid were quite similar.Insufficient radioactivity was obtained in DNAthymine from orotic acid incubations for accurate

measurement. It is evident from the experimentswith both precursors that the incorporation intoRNA uracil was not inhibited by 5-fluoro-2'-

deoxyuridine. The other drugs inhibited in increasing order of activity from FO to FU to FUR.The fact that 5-fluorouracil was a better inhibitorof both precursors than 5-fluoroorotic acid resultsfrom its greater entry into the cells (Table 1).Essentially complete inhibition of this metabolicprocess was obtained with the highest concentrations of FU and FUR.

There was no appreciable effect by any of thefluorinated pyrimidines on the conversion of uracilor orotic acid into mixed nucleic acid cytosine, aprocess which occurs to only a small extent, asshown by the specific activities of the untreatedcontrols given in parentheses in Chart 5. The cytosine specific activities were also corrected for theincreased entry of uracil and orotic acid into thecells produced by 5-fluorouridine. If this correction

TABLE 3INCORPORATIONOF URACiL-2-C" AND OROTic-6-C14ACID

INTO THE ACID-SOLUBLEFRACTIONOF ASCITESCELLSIN THE PRESENCE OFFLUORINATEDPYRIMIDINES

PER CENT OF CONTROLSCOMPOUND MOLARITT Uracil-2-CÂ» Orotic-6-CÂ» acidControls 100* lOOfFU 0.75X10"4 118 151

FO 0.75X10-4 98 111

FUR 0.75X10-4 323 156" 0.75X10-* 140 119

0.75X10-' 106 105

FUDR 0.75X10-4 105 102" 0.75X10-6 90 102" 0.75X10-' 89 101" 0.75X10-" 105 100

Molarity of labeled precursors = 1.65 X IO"4.* Total counts/min in acid-soluble = 49,900.t Total counts/min in acid-soluble = 7,960.

had not been made, there would have appearedto be a large stimulation by FUR of the incorporation of both precursors into cytosine. This is takenas further evidence for the necessity of correctingfor permeability effects in this type of experiment.These specific activities also demonstrate that, inEhrlich ascites cells in vitro, uracil is a much moreefficient precursor of RNA uracil and DNA thymine than orotic acid, a finding in accord with thetrends previously observed in viro with theFlexner-Jobling carcinoma (17).

In contrast to the lack of effect of the 5-fluoro-2'-deoxyuridine on the conversion of labeled

uracil into RNA uracil, this compound powerfullyinhibited the conversion of uracil into DNA thymine. Moreover, FU, FO acid, and FUR all in-



BOSCH et al.â€”Studies on Fluorinated Pyrimidines. V 339

hibited this process to a greater extent than theyinhibited incorporation of uracil into RNA uracil.

AVhen 5-fluorouracil-2-C14 was incubated withglycolyzing suspensions of ascites cells, the resultsobtained, as shown in Table 4, were similar tothose found in vivo. The purified FUMP fractionfrom a chromatogram of the acid-soluble fraction(3) was dephosphorylated by an intestinal phos-phatase, and the nucleoside fraction was separated on a Dowex-1 chloride column with boratebuffer (5) in the presence of 5-fluoro-2'-deoxyuri-dine and 5-fluorouridine carriers. Clean peakswere obtained, and it was found that 11 per cent

URACIL-2-C14

10080-

60-

40-

20-

100-

80-

60-

40

20

.Â«RNA Urocil

( 54,800)

(5280)

34560.75 X IO" M.

DNA Thymine

(8650)

O FU

X FO

A FUR

Q FUDR

CyÃ®osine

o (3480)

(216)

7834567 456OROTIC-6-C14 ACID

T~8

CHART5.â€”The effects of fluorinated pyrimidines on theconversion of uraeil-2-C14 and orotic-6-C14 acid into UNAuracil, DNA thymine, and mixed nucleic acid cytosine. Per centof control specific activities against molarity of inhibitors. Thefigures in parentheses are the specific activities of the controlbases in counts/min/Minole.

of the radioactivity in the FUMP peak waspresent as 5-fluoro-2'-deoxyuridine monophos-

phate. The location of the phosphate group hasnot been established, but it may be inferred byanalogy to other deoxyribonucleotides to be in theS'-position. Thus, the in vitro production from5-fluorouracil of its deoxyribonucleotide has beendemonstrated. The experiment also showed thatthe 5-fluorouracil is incorporated into RNA butnot DNA in vitro, as well as in vivo (3), and thatall the isotope in the RNA was accounted for asFUMP. Thus, there is no defluorination at thepyrimidine level. The amount incorporated appears to be of the same order of magnitude as onewould calculate for in vivo utilization for a comparable period of time.

DISCUSSIONIn this and in the two preceding papers (3, 6)

evidence has been presented to show that the

fluorinated pyrimidines exert three biochemicaleffects: (a) inhibition of the raethylation reactionleading to the formation of thymidine or thymi-dylic acid, (6) inhibition of the conversion of uracilor orotic acid into RNA uracil, and (c) incorporation into RNA to produce an unnatural or"fraudulent" ribonucleic acid.

In order to summarize the metabolic reactions,the scheme shown in Chart 6 has been constructedon the basis of knowledge available in the litera-

TABLE4In VÃœTOMETABOLISMOFS-FLUORO-

URACIL-2-C"

20 AmÃ³lesof fluorouracil incubatedwith 2.5 ml. packed cell volume in 120ml. of medium for 1 hour at 37Â°C. under the usual glycolytic conditions.Total counts/min = 60,000,000.

FractionAcid-soluble FUMPFUR from FUMPFUDR from FUMPDNARNAFUMP from RNA

TotalcountÂ»/min450,000355,000

49,800300

220,000200,000

OROTIC ACID

OMP'(FOMP)

RNAFRNA

Â»DNA

FUDR

TFA^,

CHART 6.-â€”Schemerepresenting nucleic acid uracil andthymine biosynthesis, metabolism of fluorinated pyrimidines,and the probable site of their inhibitory action. The explanationof the abbreviations is given in the text.

ture and the biochemistry of the fluorinated pyrimidines studied here. Because these drugs havelittle or no effect on the formation of nucleic acidcytosine, the amino pyrimidines are excludedfrom the diagram. The similarities between thenormal precursors and the fluorinated analogs arestressed. Uracil is shown as a precursor of uridine-5'-monophosphate (UMP) through uridine (UR)

as the intermediate. UMP is also formedfrom orotic acid, with orotidine monophosphate(OMP) as the probable intermediate. Uridinemonophosphate (UMP) can, possibly at thediphosphate level, be incorporated into RNA




uracil. The analogous reactions have been shownto take place with 5-fluorouracil (FU) and 5-fluoroorotic acid (FO) to form 5-fluorouridinemonophosphate (FUMP), which undoubtedly isalso formed from 5-fluorouridine (FUR). Uridylicacid (UMP) is known to be converted into de-oxyuridylic acid (DUMP) by a reaction which isprobably not reversible. DUMP reacts with a1-carbon tetrahydrofolic acid derivative (TFA-Ci) to form thymidylic acid (TMP), which maythen be incorporated into DNA thymine, mostprobably with thymidine triphosphate as theintermediate (18). In analogy, 5-fluorouracil hasbeen demonstrated to be converted into 5-fluoro-2'-deoxyuridine monophosphate (DFUMP), whichis presumably also formed from 5-fluoro-2'-de-

oxyuridine (FUDR).Our knowledge of the mechanism of the reac

tion leading to the formation of the methyl groupsprings from the work of Friedkin and Roberts(11),,who demonstated the production of DNAthymine from uracil deoxyriboside in mincedembryo tissue and bone marrow cells, and thestimulation of the process with citrovorum factorand its inhibition by A-methopterin. Phear andGreenberg (21) studied the methylation of de-oxyuridine with formaldehyde in supernatants ofDowex-treated homogenates of thymus gland.They found thymidylic acid (TMP) to be theproduct and established the cofactor requirementsof tetrahydrofolic acid, Mg++, ATP, and DPNH.Friedkin has reported (10) that in fractionatedcell-free extracts of E. coli deoxyuridine monophos-phate is the substance methylated. On the otherhand, Blakley has stated that with high-speedsupernatants of thymus glands, deoxyuridine wasmethylated by serine-3-C14 at higher rates thanwas DUMP (1). Thus, there appears to be adivergence of opinion as to whether the primarysubstrate is the nucleoside, deoxyuridine, or thenucleo tide, DUMP. We have chosen the latteralternative for the purposes of our diagram(Chart 6).

The evidence for the site of fluorinated pyrimi-dine inhibition at this methylation reaction stemsfrom the findings, both in vivo and in vitro, thatFO, FU, FUR, and FUDR, increasing in activityin this order, inhibit the conversion of labeledformate into DNA thymine, whereas thymidine-II3 incorporation into DNA thymine is not inhibited in vitro and is stimulated in vivo. Thelack of effect on the incorporation of formate intopurines, in agreement with the results obtainedin vivo, rules out the possibility that these compounds act as folie acid antagonists. The reversalof the inhibition of the fluorinated pyrimidines by

deoxyuridine lends strong support to the assignment of the inhibitory block at this locus. It isinteresting to note that the inhibitory effect ofFU on the methylation of pyrimidines reportedhere is in agreement with the original observationsof Scheiner et al. (23) that FU blocks the methylation of pyrimidines in the bacterium L. ieich-manii. It is also reasonable to assign this locus ofaction on the basis of structural considerations.Because of the location of the fluorine atom on thepyrimidine 5-carbon to which the methyl groupbecomes attached, it is reasonable to supposethat the fluorinated antimetabolite blocks the attachment of the methyl group to the pyrimidinering. However, it must be admitted that with thedata now at hand one cannot exclude the possibility that the drugs block the conversion of UMPto DUMP. This possibility could be eliminatedby the demonstration that FUDR would blockthe conversion of labeled deoxyuridine to DNAthymine in this system.1

We have found both in vivo and in vitro that5-fluoro-2'-deoxyuridine is by far the most potent

inhibitor of the conversion of formate into DNAthymine of the series investigated. Although differences in permeability of the cells to these inhibitors can explain the tenfold difference in effectbetween 5-fluoroorotic acid and 5-fluorouracilshown in Chart 2, the 105-fold difference between5-fluorouracil and its deoxyriboside cannot beattributed to permeability. Therefore, it can beconcluded that FUDR is the compound mostclosely related to the true blocking agent, which ismost probably 5-fluoro-2-deoxyuridine monophos-phate (DFUMP), a compound we have shown tobe formed in vitro from 5-fluorouracil and to bepresent in the acid-soluble fraction of Ehrlichascites cells during the course of the inhibition.The final proof of the site of the block must awaitthe demonstration of inhibition of the reactionby DFUMP in a highly purified system such asthat of Friedkin (10).

It is generally agreed that in mammalian systems there is no turnover of DNA, and hence incorporation of labeled precursors into DNA mustreflect a net synthesis. Although there is good incorporation of labeled formate into DNA thymine in these cell suspensions in vitro, during the1-hour incubation a net increase of DNA cannotbe detected; and the possibility exists that themethylation reaction may take place to some extent independently of DNA synthesis. That thispossibility does not invalidate the results that wehave obtained and the conclusions that we have

1Note added in proof: This demonstration has now beenmade (Harbers and Heidelberger, unpublished).



BOSCH et al.â€”Studies on Fluorinated Pyrimidines. V 341

drawn is demonstrated by the complete parallelobserved between the in vitro experiments and thein vivo dose-response curves when the quantitativeeffects of the series of fluorinated pyrimidineswere compared. Thus, regardless of whether netDNA synthesis occurs in the in vitro experiments,the results are the same as in the whole animalexperiments where the biosynthesis of DNA isknown to take place.

The pronounced inhibition produced by thesedrugs in vivoon the incorporation of labeled oroticacid and uracil into DNA thymine and in vitroon the conversion of uracil into DNA thymine ismost probably also explained on the basis of inhibition of the methylation reaction.

After our work, demonstrating that fluorinated pyrimidines inhibit the conversion oflabeled uracil and orotic acid into RNA uracilboth in vivo and in vitro, had been completed, a"Letter to the Editor" appeared by Eidinoff et al.

(9) in which they confirmed our findings on theinhibition by 5-fluorouracil of the incorporation oflabeled orotic acid into nucleic acid thymine,uracil, and cytosine, using as their system slices ofhuman tumors, maintained by transplantation inx-radiated rats.

The question arises whether the inhibition ot theincorporation of uracil and orotic acid into RNAuracil results from a general backing up of metabolism, in a sort of mass action effect, produced bythe block of the methylation reaction, or whetherblocks at additional sites are produced by thesedrugs. If there are additional blocks, their metabolic loci become a matter of importance. The firstalternative is excluded by the fact that 5-fluoro-2'-

deoxyuridine, although the most potent inhibitorof the methylation reaction, has no effect whatever on the conversion of both uracil and oroticacid into RNA uracil in vitro. Therefore, theremust be additional block(s). The possibilities are:(a) inhibition by the fluoropyrimidine bases of theconversion of uracil and orotic acid into acid-soluble uridine nucleotides, (6) inhibition of thesame process by FUMP, and (c) inhibition byFUMP of the incorporation of uridins nucleotidesinto RNA. The fact that 5-fluorouracil is a morepotent inhibitor of the conversion of orotic acidinto RNA uracil than is 5-fluoroorotic acid mightbe cited as evidence that the inhibition is not produced by the pyrimidine bases, since 5-fluorooroticacid is structurally more closely related to oroticacid than is 5-fluorouracil. However, the difference is more probably explained by the permeability differences of the cells to the two inhibitors.Similarly, the greater inhibitory effectiveness of5-fluorouridine than of the bases suggests that the

inhibition is at the nucleotide level. However, until studies are carried out on labeled 5-fluorouridine the contribution of permeability to these results cannot be assessed.

Some light is shed upon this problem by thestudies of Stone and Potter (25). They measuredthe effects of a number of pyrimidine analogs onthe conversion of labeled orotic acid into orot-idylic acid and uridine nucleotides in a highspeed supernatant from rat liver homogenate,using phosphoglyceric acid as the source of high-energy phosphate. 5-Fluorouracil exerted no effect on this system. However, 5-fluoroorotic acidproduced a profound inhibition of the conversionof orotic acid into uridine nucleotides, and noorotidylic acid was found to be present. The findings show that, in their system, 5-fluoroorotic aciditself is most probably the blocking agent, althoughsome evidence for drug metabolism was inferred.Current experiments on ascites cells and theirhigh-speed supernatant fractions are designed toclarify this matter.

Another biochemical effect described in an accompanying paper (3) is the incorporation of 5-fluorouracil and 5-fluoroorotic acid into RNA invarious tissues. The most pressing theoreticalquestion now remaining is to determine the relationship between these various biochemical effects and the inhibition of tumor growth groducedby the fluorinated pyrimidines. Let us consider inmore detail the case of the Ehrlich ascites tumor,for which most biochemical and biological information is available. The survival time of micebearing this tumor is prolonged to a greater extentby 5-fluorouridine and 5-fluoro-2'-deoxyuridinethan by 5-fluorouracil (15). Since FUDR doesnot inhibit RNA biosynthesis, that pathway is excluded as the site of tumor-inhibitory mechanism.It appears, on the basis of these experiments, thatFUDR is not converted into FUR, and hencecannot be incorporated into RNA, thus excluding"fraudulent" RNA as a mechanism of action of

this compound. The incorporation of FUDR intoDNA seems to be excluded, since 5-fluorouracil,which has been shown to be converted intoDFUMP, is not incorporated into DNA. Theseconclusions will be investigated further whenlabeled FUDR becomes available. Inhibition ofglycolysis is not caused by these compounds, andhence this is excluded from the tumor-inhibitorymechanism. Therefore, it seems clear that if anyof the biochemical studies thus far carried out relate to the tumor-inhibitory properties of 5-fluoro-2'-deoxyuridine, only the inhibition of the

methylation reaction can qualify. It is of interestthat this is the site of reaction that Cohen and




Barner have considered, on the basis of their experiments on "thymineless deaths" of bacteria, tobe desirable in a tumor-inhibitory drug (4). Onefact, however, is inconsistent with the hypothesisthat tumor growth inhibition in the Ehrlich ascitestumor is produced by the inhibition of the formation of the methyl group of thymine. It has beenshown that, on an equivalent dosage level, 5-fluorouridine is more effective than 5-fluoro-2'-

deoxyuridine in prolonging the lives of mice afflicted with this tumor (15), whereas the formerdrug is less effective in blocking the methylationreaction. This finding may result from differencesin the metabolic distribution and excretion of thesetwo compounds, and further work will be requiredbefore this discrepancy can be shown either to beexplained or else to invalidate the conclusion. Ifinhibition of the methylation reaction explains thecarcinostatic effect of FUDR, the possibilityexists that the other fluorinated pyrimidines maywell act by other mechanisms, even in the Ehrlichtumor, and considerable work must be carried outto elucidate this matter. The situation, alreadycomplex enough, is further complicated by thefact (15) that the carcinostatic properties varyamong the series of fluorinated pyrimidines fromtumor to tumor. Biochemical studies are currentlybeing carried out in other tumors and in resistanttumors in an attempt to unravel this problem.

Although the determination of the exact biochemical mechanism of tumor-inhibitory compounds has not been attained, except possibly inthe case of the antifolics and azaserine, it isnevertheless of theoretical as well as practical importance to discover how and why drugs act upontumors. One of the great unsolved problems incancer chemotherapy today is why most humantumors do not respond to compounds activeagainst transplanted tumors in animals. The goalof understanding the mechanism of action oftumor-inhibitory compounds is elusive, but withthe amount of information now available aboutthese fluorinated pyrimidines further investigationmight lead to its attainment with this series.

SUMMARY1. A study of the effects of fluorinated pyrimi

dines on nucleic acid biosynthesis has been carriedout in suspensions of Ehrlich ascites cells incubated anaerobically with various labeled substrates. The drugs did not inhibit glycolysis.

2. The conversion of formate-C14 into DNAthymine was inhibited in increasing order ofpotency by 5-fluoroorotic acid, 5-fluorouracil, 5-fluorouridine, and 5-fluoro-2'-deoxyuridine. This

inhibition by 5-fluorouracil and 5-fluoroorotic acidwas reversed by deoxyuridine. However, formateincorporation into nucleic acid adenine was notinhibited.

3. In studies of permeability it was found that5-fluorouracil was taken up by the ascites cells to agreater extent than were uracil, orotic acid, andfluoroorotic acid. 5-Fluorouridine caused an increased cellular uptake of uracil and orotic acid.

4. The fluorinated pyrimidines did not affectthe conversion of thymidine-6-H3 into DNA thymine, nor the incorporation of phosphate-P32 intoDNA or RNA in these cell suspensions.

5. The fluorinated pyrimidines inhibited themetabolic conversion of uracil-2-C14 and orotic-6-C14acid into DNA thymine and, with the exception of 5-fluoro-2'-deoxyuridine, inhibited the in

corporation of the same precursors into RNAuracil. The metabolic transformation of uracil andorotic acid into nucleic acid cytosine was not greatly affected by the drugs.

6. 5-Fluorouracil-2-C14 was converted into 5-fluoro-2'-deoxyuridine monophosphate and was

also incorporated as such into RNA in these cellsuspensions.

7. It is concluded that these fluourinated pyrimidines inhibit the metabolic methylation of deoxyuridine monophosphate to thymidine mono-phosphate. With the exception of 5-fluoro-2'-

deoxyuridine, the compounds inhibit the conversion of pyrimidines into RNA uracil. The exactmetabolic locus of this block has not yet been determined. These biochemical observations havebeen discussed in the light of the tumor-inhibitoryproperties of these drugs.

REFERENCES1. BLAKLEY,R. L. Methylation of Uracil Deoxyriboside by

Soluble Enzymes of Thymus, lÃ¼nrhÃm.et Biophys. Ã‚cta,24:224, 1957.

2. BURR, D.; LASZLO,J.; STENQLE,J.; and WIGHT, K.Metabolism of Human Leukemic Lymphoblasts and ItsNormalization by Methotrexate and 6-Mercaptopurine,Fed. Proc., 16:160-61, 1957.

3. CHAUDHOTU,N. K.; MONTAG,B. J.; and HEIDELBERGER,C. Studies on Fluorinated Pyrimidines. III. The Metabolism of 4-Fluorouracil-2-C" and 5-Fluoroorotic-2-Cu Acidin Vivo. Cancer Research, 18:318-28, 1958.

4. COHEN,S. S., and BARNER,H. D. Studies on the Induction of Thymine Deficiency and on the Effects of Thymine and Thymidine Analogues in Escherichia coli. 3.Bact., 71:588-97, 1956.

5. COHN,W. E. In: E. CHARGAFFand J. N. DAVIDSON(eds.),The Nucleic Acids, New York: Academic Press, 1:238(Fig. 21), 1955.

6. DANNEBERG,P. B.; MONTAG,B. J.; and HEIDELBERGER,C.Studies on Fluorinated Pyrimidines. IV. Effects on Nucleic Acid Metabolism in Vivo. Cancer Research, 18:329-34, 1958.



BOSCH el al.â€”Studies on Fluorinated Pyrimidines. V 343

7. DUSCHINSKY,R.; PLEVEN,E.; and HEIDELBERGER,C. TheSynthesis of S-Fluoropyrimidines. J. Am. Chem. Soc., 79:4559-60, 1957.

8. DUSCHINSKY,R.; PLEVEN,E.; MALBICA,J.; and HEIDELBERGER,C. Synthesis of S-Fluorouracil Nucleosides. Abstr.Am. Chem. Soc. Meeting, pp. 19C-20C, Sept. 8-13, 1957.

9. EIDINOFF,M. L.; KNOLL,J. E.; and KLEIN, D. Effect of5-Fluorouracil on the Incorporation of Precursors intoNucleic Acid Pyrimidines. Arch. Biochem. & Biophys.,71:274-75, 1957.

10. FRIEDKIN,M. Eneymatic Conversion of DeoxyuridylicAcid to Thymidylic Acid and the Participation by Tetra-hydrofolic Acid. Fed. Proc., 16:183, 1957.

11. FRIEDKIN,M., and ROBERTS,D. W. Conversion of UracilDeoxyriboside to Thymidine of Deoxyribonucleic Acid. J.Biol. Chem., 220:653-60, 1956.

12. GOLDTHWAIT,D. A.; PEABODT,R. A.; and GHEENBERG,G. R. Glycine Ribotide Intermediates in the de NovoSynthesis of Inosinic Acid. J. Am. Chem. Soc., 76:5258-59, 1954.

18. HARRINGTON,H., and LAVIK,P. S. Nucleic Acid Metabolism in Ehrlich Ascites Tumor Cells. Cancer Research, 17:43-47, 1957.

14. HEIDELBERGER,C.; CHAUDHURI,N. K.; DANNEBERG,P. B.; MOOREN,D.; GRIESBACH,L.; DUSCHINSKY,R.;SCHNITZER,R. J.; PLEVEN,E.; and SCHEINER,J. Fluorinated Pyrimidines, A New Class of Tumor-inhibitoryCompounds. Nature, 179:663-66, 1957.

15. HEIDELBERGER,C.; GRIESBACH,L.; CRUZ,O.; SCHNITZER,R. J.; and GRUNBEHG,E. Studies on Fluorinated Pyrimidines. VI. Effects of 5-FIuorouridine and 5-Fluoro-2'-

deoxyuridine on Transplanted Tumors. Proc. Soc. Exper.Biol. & Med., 97:470-75, 1958.

16. HEIDELBERGER,C.; GREISBACH,L.; MONTAG,B. J.;MOOREN,D.; CRUZ,O.; SCHNITZER,R. J.; and GRÃœNBERG,E. Studies on Fluorinated Pyrimidines. II. Effects onTransplanted Tumors. Cancer Research, 18:305-17, 1958.

17. HEIDELBERGER,C.; LEIBMAN,K. C.; HARBERS,E.; andBHARGAVA,P. M. The Comparative Utilization of Uracil-2-C" by Liver, Intestinal Mucosa, and Flexner-JoblingCarcinoma in the Rat. Cancer Research, 17:399-404, 1957.

18. KORNBERG,A. Pathways of Enzymatic Synthesis ofNucleotides and Polynucleotides. In McELROYand GLASS(eds.), The Chemical Basis of Heredity, pp. 579-608.Baltimore: Johns Hopkins Press, 1957.

19. LEIBMAN,K. C., and HEIDELBERGER,C. The Metabolismof P^-Labeled Ribonucleotides in Tissue Slices and CellSuspensions. J. Biol. Chem., 216:823-30, 1955.

20. LEPAGE,G. A. In Vitro Incorporation of Glycine-2-C14intoPurines and Proteins. Cancer Research, 13:178-85, 1953.

21. PHEAR,E. A., and GHEENBEHG,D. M. The Methylationof Deoxyuridine. J. Am. Chem. Soc., 79:3737-41, 1957.

22. ROBINSON,J. R. Some Effects of Glucose and Calciumupon the Metabolism of Kidney Slices from Adult andNewborn Rats. Biochem. J., 46:68-74, 1949.

23. SCHEINER,J. M.; KOSTELAK,E.; and DUSCHINSKY,R.5-Fluoropyrimidines as Growth Inhibitors of Microorganisms. Fed. Proc., 16:242, 1957.

24. SMELLIE,R. M. S.; THOMSON,R. Y.; GOUTIER,R.; andDAVIDSON,J. N. The Incorporation of C" Formate intoNucleic Acid Bases in Vitro and in Vivo. Biochim. etBiophys. Acta, 22:585-87, 1956.

25. STONE,J. E., and POTTER,V. R. Biochemical Screening ofPyrimidine Antimetabolites. III. The Testing of Drugsagainst a System with a Nonoxidative Energy Source.Cancer Research, 17:800-803, 1957.



1958;18:335-343. Cancer Res L. Bosch, Eberhard Harbers and Charles Heidelberger Metabolism in VitroStudies on Fluorinated Pyrimidines: V. Effects on Nucleic Acid

Updated version

http://cancerres.aacrjournals.org/content/18/3/335

Access the most recent version of this article at:

E-mail alerts related to this article or journal.Sign up to receive free email-alerts

Subscriptions

Reprints and

[email protected] at

To order reprints of this article or to subscribe to the journal, contact the AACR Publications

Permissions

Rightslink site. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC)

.http://cancerres.aacrjournals.org/content/18/3/335To request permission to re-use all or part of this article, use this link



http://cancerres.aacrjournals.org/cgi/alerts

mailto:[email protected]



studies on fluorinated pyrimidines v. effects on nucleic acid

Documents