JOURNAL OF BACTERIOLOGY, Aug. 1968, p. 525-532Copyright © 1968 American Society for Microbiology

Vol. 96, No. 2Printed ill U.S.A.

Properties of 5-Fluorouracil-containing RibonucleicAcid and Ribosomes from Bacillus subtilis

PRISCILLA P. SAUNDERS, R. E. BASS, AND G. F. SAUNDERS

Departments of Pathology and Developmental Therapeutics, The University of Texas M. D. Andersoni Hospitaland Tumor Institute at Houston, Houston, Texas 77025

Received for publication 29 May 1968

Growth of a strain of Bacillus subtilis that requires uracil, thymine, adenine, andtryptophan in the presence of 5-fluorouracil (FU) results in the synthesis of ribo-nucleic acid (RNA) and ribosomes in which 55 to 65% of the RNA uracil has beenreplaced by the fluorine derivative. Examination of analogue-containing ribosomesby sucrose density gradient centrifugation and thermal denaturation studies sug-

gests that, as far as the size, shape, and packing structure are concerned, extensiveFU substitution has little or no effect. FU appears to replace uracil in RNA with-out selectivity for one RNA class over another, as determined by methylated al-bumin-kieselguhr column chromatography and sucrose density gradient centrifu-gation. The total amino acid content of the cells is markedly affected by growth inthe presence of FU. The possibility of an FU effect on genetic translation is dis-cussed.

In view of previous investigations indicatingproduction of inactive or altered enzymes in thepresence of 5-fluorouracil (FU) (4, 8) it is prob-able that functional differences exist between FU-containing and natural ribonucleic acids (RNA)(9). Several investigations on the biological andphysiochemical properties of FU polyribonucleo-tides have been reported. Experiments on aminoacid incorporation demonstrate that FU polymersbehave exactly like uracil (U) polymers in codingfor protein synthesis. In physical studies, the FUpolymers form less stable complexes with poly-adenylic acid (poly A) than do their U counter-parts (16). Gros et al. (7) presented evidence sug-gesting that natural FU messenger RNA can bedefective. The possibility of FU replacing cytosine(C) in RNA messages was studied in detail byBujard and Heidelberger (3) with the deoxyribo-nucleic acid (DNA)-dependent RNA polymerasereaction using a DNA primer of defined basesequence. They found the frequency of replace-ment to be less than one uracil or FU per 3,000cytosines. They pointed out, however, that thislevel of replacement could be explained by DNAcontamination in the RNA polymerase prepara-tion.A companion paper (18) describes an abnormal

pattern of growth in the presence of FU ex-hibited by a strain of Bacillus subtilis that requireduracil and thymine. The cells continued to syn-thesize DNA, RNA, and protein in the presenceof the drug but were unable to undergo cell divi-

sion. This paper will describe the properties of theFU-containing RNA formed by this organismand attempt to elucidate the direct cause of death.

MATERIALS AND METHODSThe methods employed are generally as described in

the previous paper (18) with the exception of thosedetailed below.

Measurement of growth and isotope incorporation.A minimal medium (18) containing 20 ,ug each ofadenine and thymine per ml, 2 j,g of uracil per ml, and,when desired, 8.0 ,ug ofFU per ml was inoculated witha suspension of cells obtained from a Trypticase-agarplate that had been incubated overnight at 37 C.,

Growth studies were carried out in 250-ml Delongflasks containing 50 ml of medium with vigorousshaking at 37 C in a model G-77 water-bath shaker(New Brunswick Scientific Co., New Brunswick, N.J.).At various time intervals, 1-ml samples were removed,and their absorbancy at 420 m,u was measured in aZeiss PMQ II spectrophotometer. Additional 1-mlsamples were removed at appropriate intervals andtransferred to tubes containing 1 ml of cold 20% tri-chloroacetic acid. The trichloroacetic acid precipitateswere collected on 25-mm type B6 (Schleicher & SchuellCo., Keene, N.H.) membrane filters, washed withsix successive 5-ml portions of cold 10% trichloroace-tic acid, dried, and counted for 10 min in a PackardTri-Carb liquid scintillation spectrometer. The count-ing solution employed contained 0.01% 1,4-bis-[2-(4-methyl-5-phenyloxazolyl)]-benzene and 0.4% 2,5-diphenyloxazole in toluene.

Amino acid analysis. Cells grown as described abovewere harvested when the absorbancy at 420 m,u

525

SAUNDERS, BASS, AND SAUNDERS

reached 0.60. The cells were washed twice with 0.2 MNH4Cl, dialyzed against deionized water, and lyo-philized. Approximately 2 mg of protein, estimated bythe method of Lowry et al. (14), were hydrolyzed for24 hr in 1.0 ml of glass-distilled constant-boiling HClat 110 C under vacuum. The amino acid compositionof the hydrolysates was determined on a Spinco(model 120B) automatic amino acid analyzer as de-scribed by Spackman, Stein, and Moore (20). Theanalyses were corrected for a 10% decomposition ofthreonine and serine. The standard deviations (SD) ofduplicate analyses were less than -+ 0.03 mole per centfor all amino acids except the following from the FUcells: aspartic acid, SD = i 0.045; threonine, SD = 40.034; glutamic acid, SD = ± 0.23; and alanine, SD =:1 0.30.

Isolation of ribosomes. Cells were grown as de-scribed above, harvested in the cold, and washed oncewith a standard buffer (SB) containing 10-2 M tris-(hydroxymethyl)aminomethane (Tris), 1.2 X 10-2 Mmagnesium acetate, 6 X 10-2 M KCl, and 6 X 10-3 Mmercaptoethanol (pH 7.8). All subsequent isolationsteps were carried out at 4 C. The cells were broken bygrinding with 100-5005 Superbrite glass beads (Minne-sota Mining and Manufacturing Co., St. Paul, Minn.).The cell debris was removed by one 20-min centrifuga-tion at 20,000 X g and two 30-min centrifugations at35,000 X g in a Sorvall refrigerated centrifuge (rotorSS34). Ribosomes were pelleted by centrifugation at105,000 X g for 2 hr (40 rotor) in a Spinco (model L)preparative ultracentrifuge. The pellet was suspendedin SB buffer, pelleted again at 105,000 X g, and dia-lyzed overnight at 4 C against the desired buffer (SB).The FU cells did not break as easily as the normal cellswhen ground with glass beads, so it is possible that ourribosome preparations derive from only the mostfragile part of the FU cell population. Physical studieswere conducted only on ribosome batches that hadbeen shown to have 55 to 65%0 FU substitution. Thepurity of the ribosome preparations was determined bysucrose density gradient centrifugation (Fig. 3) andabsorption spectra (legend, Fig. 4).

32p labeling was accomplished by growth of cells onminimal medium without FU plus 10 ,uc of carrier-free32p phosphoric acid per ml (International Chemicaland Nuclear, City of Industry, Calif.). After growthhad reached mid-log phase, the cells were harvestedin the cold and washed twice with SB. The pellet wasfrozen and mixed with 2 g of frozen B. subtilis cells(Miles Laboratories, Inc., Elkhart, Ind.) to give asufficient cell mass for breakage by grinding withglass beads. The ribosomes were isolated and purifiedas described. The specific activity of the preparationshown in Fig. 2 was 1,900 counts per min per jug ofribosomes. Therefore, the amount of carrier ribosomesin the 32p preparation, which was layered on thegradient, would not be detected by absorbancymeasurements.

Extraction and purification ofRNA. Lysozyme (400,ug/ml) and 50,g of electrophoretically purified de-oxyribonuclease I per ml (Worthington BiochemicalCorp., Freehold, N.J.) were added to washed cells in3 X 10-2 M Tris-chloride buffer (pH 7.7). The mixturewas incubated at 37 C for 5 min, and then MgCI2 was

added to give a final concentration of 10-2 M. Themixture was incubated at 37 C for 3 min, sodiumdodecyl sulfate (to a final concentration of 0.4%) wasadded, and the mixture was shaken for 5 to 10 sec. Anequal volume of phenol buffered with 0.01 M KH2PO4(pH 6.7) was added immediately to minimize activityof the cellular ribonuclease.RNA was purified according to the method of

Gierer and Schramm (6), and was precipitated fromthe aqueous phase, after the removal of ether, with 3volumes of cold 95% ethyl alcohol; the precipitatewas then dissolved in 0.1 X SSC [0.015 M NaCl-0.0015 M trisodium citrate (pH 7)]. The RNA solutionwas dialyzed against 2,000 volumes of 2 X SSC in thecold for 18 hr.

Base composition analysis ofRNA. The RNA solu-tion was hydrolyzed by adding an equal volume of 0.6N KOH and incubating the mixture at 37 C for 18 hr.The resulting 2'- and 3'-nucleotides were separated ona Dowex-1 -8X formate column with a resin bed 9 mmin diameter by 50 mm high. The method of elutionwas described by Key (12). Unlabeled Escherichia colicarrier RNA of known base composition was addedprior to hydrolysis to give an absorbancy marker inthe nucleotide regions and to provide internal controlsfor the preferential loss of individual nucleotides. Theeluate was collected in 5-ml fractions, and the absorb-ancy at 260 m,u was determined for each fraction. Theradioactivity of each fraction was determined by liquidscintillation counting in a Packard Tri-Carb spectrom-eter as previously described (18).RNA fractionation on methylated albumin-kies-

selguhr (MAK) columns. MAK columns were preparedas described by Sueoka and Chang (21). Cells weregrown as described (18) with the addition of 1 ,uc/mlof FU-6-3H. A column containing 25 ml of MAK wascharged with 50 ml of an RNA solution containing 1.5mg of U-RNA and 90 t&g of 3H-FU-RNA (1,690counts per min per ,ug of RNA) in 0.1 M NaCl-0.05 Mpotassium phosphate buffer. RNA was eluted with alinear gradient of NaCl, and 5-ml fractions werecollected. U-RNA was identified by absorbancy at260 mMA. For the detection of tritium, 0.5-ml portionsof each fraction were diluted with 0.1 ml of carriersalmon sperm DNA (1 mg/ml in water) and 5 ml of12% trichloroacetic acid, and incubated in an ice bathfor 30 min. Precipitates were collected on 25-mm typeB6 membrane filters, washed with 20 ml of cold 10%trichloroacetic acid, dried under a heat lamp, andcounted as described.

Melting temperatures. Thermal denaturation ofRNA and ribosomes was followed in a Zeiss PQM IIspectrophotometer equipped with a heated cell com-partment and a temperature readout assembly. Allsamples had initial absorbancies of 0.50 to 0.70 at 260mMA.

RESULTS

Incorporation of 3H-FU and 3H-thymidine intotrichloroacetic acid-insoluble material. Figure 1shows an initial experiment designed to verify theassumption that FU is incorporated into themacromolecules of strain 568(b). The incor-poration of 3H-FU into trichloroacetic acid-

526 J. BACTERIOL.

VOL. 96, 1968 5-FLUOROURACIL-CONTAINING RNA AND RIBOSOMES

12

4-

16

1i2

8-

WITH FU

WITHOUT FU1.0

.8 OD. H3- T UPTAKE

.4I

Id0 2 4 6

INCUBATION TIME IN HOURS

FIG. 1. Uptake of 3H-FU and 3H-thyttrichloroacetic acid-insoluble material with t

FU. Appropriate flasks were supplementedWH (0.25 ,uc per ml) or thymidine-methyl-3Hml).

insoluble material continued throulgrowth period as did the incorporatithymidine.

Extent of FU substitution. A cruciastudies regarding the effect of base anmRNA and ribosome structure is the ext(stitution of FU for U in the RNA inNucleotide base substitution was mortwo independent methods. The firstradioactive labeling and elution fronexchange column of the nucleotide basfrom purified ribosomes (Fig. 2). Thithe uridylate fraction into two peaks coFU (65 %) and U (35 %O). The presencein the cytosine peak indicates that afraction of the input 3H-uracil was co]cytosine. The level of crossover of 3F

'4C channel in our radioactivity meawas about 15%, which accounts forpeak in the cytosine region and does n(any conversion of 14C-FU to either U oThe second method, paper chromowhich involves formic acid hydrolysi:somes, separation in ethylacetate-wa

acid (60:35:5), base identification with the ap-propriate standards, and elution of the bases with

20 0.1 M HCl, routinely gave approximately 55% FU16

" and 45% U. Therefore, the level of substitution6 80 was in the range of 55 to 65%, sufficient to war-E rant a search for structural alterations in the cel-

12 60 G lular RNA and ribosomes.<..,Properties of FU ribosomes. To determine the

8 40 F effect of U replacement by FU on the physical^ stability of ribosomes, a series of experiments

4 20 designed to assess the effect of alterations in thenucleoprotein packing structure of ribosomeswere carried out.

0 Co-sedimentation of the 32P-U and unlabeledFU ribosomes in a sucrose density gradient (Fig.3) revealed no marked differences between thetwo types of ribosomes. In both cases, the ribo-

16 somes were primarily 70S. Thus, one can conclude-E that ribosome size and shape are not altered sig-

12 E nificantly by extensive FU substitution. This is inui contrast to the work on E. coli ribosomes, such as

8 that of Iwabuchi et al. (11) and Andoh and'* Chargaff (1), who found an accumulation of non-4 functional FU-containing particles at essentially

the same degree of substitution.The stability of the nucleoprotein packing struc-

0 ture can be monitored by thermal-induced ab-sorbancy changes. Figure 4 shows that the FU

midine inlto and U ribosomes have essentially the same Tmand without (about 70 C). However, the shapes of the meltingwith FU-6- curves differ significantly. The normal U ribo-(0.5 Ac per somes are stable up to 60 C and then begin to dis-

sociate over a 10 C temperature range. The dropin absorbancy above 70 C is due to protein pre-

ghout the cipitation and frequently occurs in these analyses.on of 3H- It has no effect on the midpoint of the absorbance

transition.L1 point in FU-RNA characterization. From the degree ofLlogues on analogue substitution observed, one would expectent of sub- some changes in physicochemical properties to bel question. induced by FU in RNA. The finding that FUnitored by polynucleotides form less stable structures ofinvolved either the complementary hydrogen bonded type

n an ion- (i.e., the poly FU-poly A complex) or the intra-,es derived molecular base stacked type [reviewed by Michel-is resolved son et al. (17)] supports the notion that one could)nsisting of detect structural changes if a sufficient quantityof tritium of FU stretches exist in the RNA. We have usedsignificant two independent methods to monitor changes inInverted to RNA secondary structure. The first method (Fig.4 into the 5), sucrose density gradient centrifugation, showssurements that the size and shape of FU- and U-RNA arethe small essentially the same since there is no reason toot indicate suspect an FU effect on RNA chain length.r cytosine. The fractionation of RNA on MAK columnsatography, is a useful method for showing changes in second-s of ribo- ary structure (5). Once again, the FU- and U-ter-formic RNA species chromatograph together (Fig. 6).

527

8

SAUNDERS, BASS, AND SAUNDERS

.005 MH20

0.15 M--.--I.0 M - -- 2.5M - - GRADIENT 2.5- 6.0

F N 51.2- 12 0A ~~~~~~~~~~~~FUETh1.0c 10

0

in~~~ ~ ~ ~ ~ ~ ~tn in th ao N eks niaign

G 8

.2-

00 0200 300 50 60 70 80 90 0o0 10 20 30 40 63FRACTION NUMBER (5 ml)

FIG. 2. Elution ofnucleotide bases from a Dowex-1-formate column. Cells were grown as described (18) in thepresence of 8 pg/mi ofFU containing 1 puc/mi of 14C-FU and 2 pAg/mi of U containing 1.5 pc/mi of3H-uridine.The cells were harvested when the absorbancy at 420 mp reached 0.63. Absorbancy was monitored at 280 mpAforfractions 1 to 33 and at 260 mp for the remaining fractions.

4.02 800 Although the tritium background was somewhathigh, the ratio of 3H to A26o was essentially con-stant in the major RNA peaks, indicating nodramatic selectivity of incorporation of thenatural base over the halogenated form into a3.01 600 particular class of

RNA. This is unlike certain

other base analogues whose distribution in RNAhas revealed marked selectivity [i.e., azaguamneincorporation into soluble RNA in B. cereus (13)]Amino acid composition ofFUcells. Under these5F2.0 T400N8

conditions of continued growth (an presumedprotein synthesis), and the high degree of incor-poration of FU into RNA, it is conceivable that

P32. U RIBOSOMES the observed aberrant growth could be reflected1.0 200 in the amino acid analyses of 568(b) cells grown

in the presence and absence of FU. One can seethat there were distinct differences in the relative

FU RIBOSOMES amounts of several amino acids. The most pro-nounced were the increased relative ratios of

1 10 20 30 glutamic acid, glycine, and alanine. All the otherFRACTION NUMBER amino acids decreased with the exception of

FIG. 3. Sucrose density gradient co-cenitrifugation arginine which remained essentially constant. Theprofile of B. subtiiis FU and U ribosomes. The direc- possible significance of these findings will be con-tion ofsedimentation is from left to right. A 0.2-mi sidered in the Discussion. An extensive study ofsample containing 200 pAg of RNA was layered onto the changes in amino acid composition in several4.5 ml of a 15 to 30% linear sucrose gradient in SB. cellular components at various FU to U ratios isThe tubes were centrifuged in an SW 39L rotor in a currently in progress.model L-2 preparative uitracentrifuge at 37,000rev/mmnDsUSOfor 2 hrat 4 C. The tubes were pierced, and fractions Dsusowere collected and assayed for radioactivity and ab-. In view of the finding that FU polymers formsorbancy at 260 mp. less stable complexes with poly A than do uracil

528 J. BACTERIOL.

5-FLUOROURACIL-CONTAINING RNA AND RIBOSOMES

Tm

'Tm 69.250

U-RIBOSOMES

FU-RI BOSOMES

24 30 40 50 60 70 80 90TEMPERATURE °C

FIG. 4. Thermal denaturation profiles ofFU and U ribosomes in SB (1.2 X 10-2 M Mg++). Ribosomes were puri-fied by differential centrifugation as described in Materials and Methods. Ribosome concentrations were deter-mined assuming that an absorbancy at 260 m,u (A260) of1.0 corresponds to 65 ,ug/ml at 1 cm light path. The absorb-ancy ratios of the samples analyzed were: U ribosomes, A280/A260 = 0.500, A240/A260 = 0.612; FU ribosomes,A28/A260 = 0.536, A240/A260 = 0.666.

1.00 -

.500 -

1 20FRACTION NUMBER

30 40

5,000

1,000

FIG. 5. Sucrose density gradient co-centrifugation profile of B. subtilis FU- and U-RNA. A 0.5-ml sample con-taining 360 ,ug ofRNA was layered onto 28 ml ofa 5 to 20% linear sucrose gradient in 0.05 M sodium chloride-0.01 M sodium acetate (pH 5.0) buffer. The tubes were centrifuged in an SW 25.1 rotor in a model L-2 preparativeultracentrifuge at 22,500 rev/min for 12 hr at 4 C. Fraction collection and analysis were performed as in Fig. 3.The majorpeak at fraction 16 is 70S ribosomes and the peak at fraction 10 to 11 corresponds to approximately 50S.

polymers (16), one might expect a greater prob-ability of incomplete translation with FU mes-senger RNA molecules. If this were true, it ispossible that the regulation of RNA synthesismight be impaired, resulting in the increased pro-duction of certain messages and their correspond-ing proteins or peptides and a decreased produc-tion of others. A rough approximation of such atranslational effect would conceivably be evidentfrom a total cellular amino acid analysis. A study

of this type was carried out in E. coUl by Aronson(2), who reported that FU did not affect theamino acid composition of the cells. Gros andNaono (8), on the other hand, reported that FUinhibits the incorporation of proline and tyrosine,and stimulates incorporation of arginine in E.coli.The total cellular protein of B. subtilis strain

568(b) grown in the presence of FU shows anincreased relative ratio of glutamic acid, glycine,

529VOL. 96, 1968

Z&E

"OC.C4-C

tL

EQ.

SAUNDERS, BASS, AND SAUNDERS

E0o0

FRACTION NUMBER

FIG. 6. MAK column profiles ofFU- and U-RNA. Cells were grown as described (18), with the addition ofI,c of3H-FU per ml. The details ofchromatography are described in the text.

TABLE 1. Amino acid composition (mole percentage)of 568(b) cells grown on U and FU

Amino acid U cells FU cells

Lysine ................. 7.38 7.20Histidine .............. 1.54 1.38Arginine............... 3.78 3.81

Aspartic ............... 9.63 7.63Threoninea ............ 5.20 4.69Serinea............ 4.79 4.19

Glutamic acid......... 13.67 14.61Proline ................ 3.38 3.05Glycine ................ 8.40 10.01Alanine ............... 13.61 19.40Valine................. 6.61 5.94

Methionine ............ 1.91 1.43Isoleucine ............. 5.75 4.93Leucine ............... 7.45 6.14Tyrosine ............... 3.08 2.92Phenylalanine ......... 3.78 3.28

Total..... .... 99.96 100.61

a Corrected for 10% decomposition.

and alanine, while arginine remained at the sameratio (Table 1). Although it has been well estab-lished that the genetic code is degenerate andmultiple codons exist for all the amino acids,Marshall et al. (15) have shown species-dependentdifferences in the response of amino acyl-transferRNA to trinucleotide codons among bacteria,amphibia, and mammalia. Using E. coli as a

representative of the bacteria, they find that thecodons that give the greatest response (at leasttwice that of other recognizable codons) for theamino acids relevant here are: glutamic-GAG,GAA; glycine-GGC; alanine-GCG, GCA; argi-nine-CGA, CGU. It is interesting to note thatonly one of these codons contains uracil. In thepresent study, all of the other amino acids ana-lyzed decreased in ratios of 10% or more.The FU ribosomes display one mode of dena-

turation not found in the normal ribosomes. Thiseffect is observed as an increase in absorbancybetween 45 and 60 C and may reflect a high degreeof structural heterogeneity which does not alterthe sedimentation pattern (Fig. 3). It is quiteprobable that alterations in ribosomal RNA basestacking do not result in changes in the ribosomesedimentation coefficients. Although FU maydestabilize the ribosome, the effect is not asmarked as one might expect from the extent ofanalogue substitution in the ribosomes. This sug-gests that the effect is quite subtle and couldinvolve RNA loops where base pairing and basestackingare not essential to maintain the ribosomepacking structure. A more sensitive assay forRNA loops would involve in vitro protein syn-thesis experiments or the messenger RNA-cat-alyzed binding of amino acyl-soluble RNA toribosomes. These determinations are currentlyunder investigation.The influence of FU on ribosome synthesis has

been shown by Aronson (2) and Iwabuchi et al.(11) in E. coli, and by Hignett (10) in Staphylococ-cus aureus, to result in the accumulation of smallinactive FU-containing particles. The experiments

530 J. BACTERIOL.

VOL. 96, 1968 5-FLUOROURACIL-CONTAINING RNA AND RIBOSOMES

described here with strain 568(b) of B. subtilisare in apparent contrast to these situations.

It is clear from this and the companion paper(18) that, in the presence of an exogenously sup-plied source of thymine, protein synthesis instrain 568(b) occurs in spite of a high degree ofsubstitution of FU for U in all classes of RNA.The physical studies with FU ribosomes from568(b) indicate that, as far as the size, shape, andpacking structures are concerned, FU substitution(55 to 65%7) has little, if any, effect. This discrep-ancy may conceivably reflect the different methodsof FU exposure employed. Ribosomes were ob-tained from 568(b) after growth in the presenceof FU (8 ,ug per ml of FU, 2 Ag per ml of uracil).On the other hand, the ribosomes of E. coli andS. aureus were isolated from cells that had beenexposed to a high level of FU (20 Ag/ml) in theabsence of U for a relatively short period of time(40 to 60 min). Since the RNA of these particlescontained 60 to 70% substitution of FU for U,they, as well as all RNA species synthesized duringthe interval, probably contain numerous FUstretches. If one assumes that the distribution ofFU along an RNA molecule is similar to that ofU, and that both bases are incorporated into RNAat a rate which is proportional to the relativeconcentration of each base in the precursor pool,then the number or chain length, or both, ofpure FU clusters in RNA would be reduced underconditions of prolonged analogue exposure atFU/U = 4. As mentioned in the Results section,evidence obtained from physical studies on syn-thetic homopolyribonucleotides suggests that thepresence of FU stretches should result in lessstable RNA base-pairing reactions of either theinter molecular or intramolecular types.The fact that only slight changes were observed

in the 568(b) FU ribosome melting curve suggestseither that very few FU stretches existed in theribosomes or that uridylate stretches are rare inB. subtilis ribosomal RNA. The recent work ofSlapikoff and Berg (19) supports the former idea.They investigated the effect of different nearestneighbors on the specificity with which an adeno-sine residue in the polydeoxynucleotide templatedirects the incorporation (by RNA polymerase)of either uridine 5'-monophosphate or one of itsanalogues. It was found that U is incorporatedmore efficiently than FU next to adenine or cyto-sine. Little or no preference exists between uridine5'-triphosphate and 5-fluorouridine 5'-triphos-phate for incorporation next to guanine residues.When U is the nearest neighbor, U is prefer-entially incorporated compared to pseudouridine.Unfortunately, the case of FU with a U nearestneighbor was not examined. These data suggestthat there will be fewer FU than U stretches

found in the RNA of a cell that was grown inthe presence of both nucleotide bases.

ACKNOWLEDGMENTSThis investigation was supported by Public Health

Service grants CA-06939 and FR-0551 1-03-IN-33,American Cancer Society grants IN-43-G-14 andT-465, and Public Health Service contract PH-43-66-1156.We are grateful to L. Hnilica for assistance with the

amino acid analyses.LITERATURE CITED

1. Andoh, T., and E. Chargaff. 1965. Formation andfate of abnormal ribosomes of E. coli cellstreated with 5-fluorouracil. Proc. Natl. Acad.Sci. U.S. 54:1181-1189.

2. Aronson, A. I. 1961. The effect of 5-fluorouracilon bacterial protein and ribonucleic acid syn-thesis. Biochim. Biophys. Acta 49:98-107.

3. Bujard, H., and C. Heidelberger. 1966. Fluor-inated pyrimidines. XXVII. Attempts to deter-mine transcription errors during the formationof fluorouracil-containing messenger ribonu-cleic acid. Biochemistry 5:3339-3345.

4. Bussard, A., S. Naono, F. Gros, and J. Monod.1960. Effets d'un analogue de l'uracil sur lesproprietes d'une proteine enzymatique synthe-tisee en sa presence. Compt. Rend 250:4049-4051.

5. Doi, R. H., and B. Goehler. 1966. Reversiblemodification of lysyl-tRNA which affects itsbinding to poly A-ribosome complexes.Biochem. Biophys. Res. Commun. 24:44-49.

6. Gierer, A., and G. Schramm. 1956. Infectivity ofribonucleic acid from tobacco mosaic virus.Nature 177:702-703.

7. Gros, F., W. Gilbert, H. H. Hiatt, G. Attardi,P. F. Spahr, and J. D. Watson. 1961. Molecularand biological characterization of messengerRNA. Cold Spring Harbor Symp. Quant. Biol.26:111-132.

8. Gros, F., and S. Naono. 1961. Bacterial synthesisof "modified" enzymes in the presence of apyrimidine analogue, p. 195-205. In R. J. C.Harris (ed.), Protein biosynthesis. AcademicPress, Inc., New York.

9. Heidelberger, C. 1965. Fluorinated pyrimidines,p. 1-50. In J. N. Davidson, and W. E. Cohn(ed.), Progress in nucleic acid research andmolecular biology, vol. 4. Academic Press, Inc.,New York.

10. Hignett, R. C. 1966. Interference of 5-fluorouracilin the biosynthesis of ribosomes in Staphylo-coccus aureus (strain Duncan). Biochim.Biophys. Acta 114:559-564.

11. Iwabuchi, M., E. Otaka, M. Kono, and S. Osawa.1965. The effect of 5-fluorouracil on the ribo-some formation in Escherichia coli. Biochim.Biophys. Acta 95:38-94.

12. Key, J. L. 1966. Effect of purine and pyrimidineanalogues on growth and RNA metabolism inthe soybean hypocotyle-the selective action of5-fluorouracil. Plant Physiol. 41:1257-1264.

531

SAUNDERS, BASS, AND SAUNDERS

13. Levin, D. H. 1963. The incorporation of 8-aza-guanine into soluble ribonucleic acid of Bacilluscereus. J. Biol. Chem. 238:1098-1104.

14. Lowry, 0. H., N. J. Rosebrough, A. L. Farr, andR. J. Randall. 1951. Protein measurement withthe Folin phenol reagent. J. Biol. Chem. 193:265-275.

15. Marshall, R. E., C. T. Caskey, and M. Nirenberg.1967. Fine structure of RNA codewords recog-

nized by bacterial, amphibian and mammaliantransfer RNA. Science 155:820-826.

16. Massoulie, J., A. M. Michelson, and F. Pochon,1966. Polynucleotide analogues. VI. Physicalstudies on 5-substituted pyrimidine polynucleo-tides. Biochim. Biophys. Acta 114:16-26.

17. Michelson, A. M., J. Massoulie, and W. Guschl-bauer. 1967. Synthetic polynucleotides, p. 83-141. In J. N. Davidson and W. E. Cohn (ed.),Progress in nucleic acid research and molecular

biology, vol. 6. Academic Press, Inc., NewYork.

18. Saunders, P. P., G. A. Schultz, and G. F. Saunders1968. Effect of 5-fluorouracil on the growth andmorphology of a polyauxotrophic strain ofBacillus subtilis. J. Bacteriol. 96:560-562.

19. Slapikoff, P. and P. Berg. 1967. Mechanism ofribonucleic acid polymerase action. Effect ofnearest neighbors on competition between uri-dine triphosphate and uridine triphosphateanalogs for incorporation into ribonucleic acid.Biochemistry 6:3654-3658.

20. Spackman, D. H., W. H. Stein, and S. Moore.1958. Automatic recording apparatus for use inthe chromatography of amino acids. Anal.Chem. 30:1191-1206.

21. Sueoka, N. and T. Y. Chang. 1962. Fractionationof nucleic acids with the methylated albumincolumn. J. Mol. Biol. 4:161-172.

532 J. BACTERIOL.