Selective Utilization of Pyrimidine Deoxyribonucleosides for ...

JOURNAL OF BACTERIOLOGY, Mar. 1973, p. 1356-1362 Copyright @ 1973 American Society for Microbiology

Vol. 113, No. 3 Printed in U.S.A.

Selective Utilization of Pyrimidine Deoxyribonucleosides for Deoxyribonucleic Acid Synthesis in Pneumococcus BARRY BEAN AND ALEXANDER TOMASZ The Rockefeller University, New York, New York 10021 Received for publication 3 October 1972

When pyrimidine deoxyribonucleosides are supplied to growing cultures of Diplococcus pneumoniae, they are selectively used for incorporation into deoxyribonucleic acid (DNA). Differently labeled molecules of deoxyuridine, thymidine, and deoxycytidine were used to study the precursor pathways of this organism. Each of these preformed pyrimidine deoxynucleosides is incorporated intact (i.e., without cleavage of the glycosidic bond) and is predominantly recoverable as DNA thymidine. During the utilization of deoxycytidine and deoxyuridine by pneumococci, large proportions of the available precursor are converted to free thymidine, which is secreted back into the growth medium. The biochemical pathways for selective incorporation into DNA and the regulation of concentrations of intracellular thymidine compounds by excretion of free thymidine are discussed.

The biochemical pathways for pyrimidine metabolism in pneumococcus are unusual in several respects. Both in vivo and in vitro experiments have suggested that there is little, if any, enzymatic activity for the glycosidic cleavage of pyrimidine ribo- or deoxyribonucleosides in this organism. Supplied pyrimidine precursors or their fluorinated counterparts each are subject to different patterns of utilization; among the fluoropyrimidine analogs, for instance, fluorouracil (FU), fluorouridine (FUR), and fluorodeoxyuridine (FUdR) are each subject to different patterns of incorporation, and each exhibits a different spectrum of inhibitory characteristics. There are apparently several different fluoropyrimidinesensitive metabolic targets in this organism (1, 2). The uptake and metabolism of pyrimidine deoxyribonucleosides by pneumococci has drawn our particular interest because of several different observations: (i) this organism seems to possess a single system for the regulation of the transport of several different ribo- and deoxyribonucleoside precursors (2); (ii) one of the actions of FUdR appears to be the inhibition of the transport or phosphorylation of thymidine (TdR), or both (1); (iii) deoxyuridine (UdR) is unique in being the most efficient generalized reversing agent for the inhibitions

caused by the different fluoropyrimidines (1); and (iv) it was noted in relation to other experiments that UdR (and other pyrimidine deoxynucleosides) can serve as particularly effective precursors for the labeling of deoxyribonucleic acid (DNA) (1). We report here the results of a series of more detailed studies on the incorporation of these precursors. MATERIALS AND METHODS Bacterial strains and culture methods. The wild-type and mutant strains of Diplococcus pneumoniae R36A used in these experiments and the media for their propagation have been described previously (1, 2). Radioactive precursors. The radiopurity of all commercial, labeled compounds was routinely checked by thin-layer chromatography (TLC). When necessary, compounds were purified by paper chromatography. Studies on the utilization of radioactive precursors. Labeling conditions, filtration methods for determination of intracellular isotope distribution, preparation of soluble pool extracts by trichloroacetic acid extraction, and methods for TLC and anion exchange chromatography have been detailed previously (1). Analysis of radioactive media components was performed after removal of cells by centrifugation or filtration. Samples of the cell-free media were applied directly to TLC plates for analysis by a combination of appropriate solvent systems. In experi-

1356

VOL. 113, 1973

USE OF PYRIMIDINE DEOXYRIBONUCLEOSIDES

ments with UdR-dR-14C (i.e., UdR labeled with 14C in the deoxyribose portion) and UdR- 14C (U), medium radioactivity was too dilute to permit direct analysis, and a concentrate was prepared as follows: the medium was lyophilized and 0.2% of the original volume of water was added to the residue, which was largely insoluble in that volume. The material that was soluble was decanted, made 75% in ethanol, chilled, and centrifuged to remove the additional precipitate which formed. This low-salt supernatant fluid, containing essentially all of the original medium radioactivity, was concentrated and analyzed by TLC and X-ray film autoradiography. Hydrolysis of ribonucleic acid. Trichloroacetic acid precipitates formed during extraction of soluble pool material were washed again with trichloroacetic acid dried, dissolved in 1 to 3 ml of 1 N KOH, and incubated at 37 C for 12 to 20 h. The hydrolysate was neutralized by addition of formic or acetic acid and adjusted to pH 5 by further addition of dilute acid. Ethanol was added to a final concentration of 75%. After standing on ice, the precipitate was washed again with chilled, acidified 75% ethanol and saved for analysis of DNA. The supernatant fluid and washes were combined, and ethanol was removed by evaporation. Hydrolysis products could be analyzed at this stage by anion exchange chromatography or by being enzymatically dephosphorylated to the corresponding nucleosides for analysis by TLC. Hydrolysis of DNA. Radioactively labeled DNA, prepared either as transforming DNA (7) or as described above, was hydrolyzed by either: (i) the sequential action of deoxyribonuclease I and snake venom phosphodiesterase (8) or (ii) by treatment with 1 ml of 7.5 N HClO, at 100 C for 1 h. HC1O4 hydrolysates were neutralized with concentrated KOH, made 75% in ethanol, and chilled. The supernatant fluid and wash of the precipitate were combined, concentrated by evaporation, and analyzed by TLC. Intestinal phosphatase treatment. Nucleoside monophosphate solutions were freed of high cation concentrations by exposure to Dowex 50-H+, if necessary. After adjustment to pH 8.5, 0.02 ml of 0.1 M MgCl2 and 0.05 ml of intestinal alkaline phosphatase (Worthington, 1 mg/ml in 0.5 M tris(hydroxymethyl)aminomethane (Tris) buffer, pH 8.4) were added for each milliliter of solution. After incubation at 37 C for 6 to 12 h acetic acid and ethanol were added, precipitable material was removed by centrifugation, and the solution was analyzed by TLC. Analysis of radioactivity in the pentose and base moieties of nucleosides. The enzyme nucleoside transdeoxyribosidase was used as an analytical tool for analysis of the distribution of radiocarbon atoms in pyrimidine deoxynucleosides. When used as previously described (1), both the pyrimidine and deoxyribose moieties of deoxycytidine (CdR) and UdR could be conveniently separated and quantitated. Formic acid hydrolysis (8) (70% formic acid, 175 C, in a sealed tube; 60 min for deoxyribonucleosides or 180 min for ribonucleosides) was used as an alternative method. Preparation of UdR-dR 14C. UdR-dR-14C (i.e.,

1357

UdR unlabeled in the uracil moiety, but uniformly radiocarbon-labeled in the deoxyribose moiety) was prepared as follows. Escherichia coli K 140 met pyr was grown to stationary phase in 30.4 ml of minnimal medium (4), supplemented with unlabeled uracil and methionine (in excess), and 0.1% glucose containing a total of 1.0 MCi of glucose-'4C (U). Cells were harvested, washed three times with minimal medium, and treated with KOH as described under hydrolysis of ribonucleic acid (RNA), except that 0.1 ml of 4% bovine serum albumin was added to assure complete precipitation. The precipitate was removed by centrifugation, washed twice with acidified 75% ethanol, dried, resuspended in 2 ml of 0.05 M Tris buffer, pH 8.0, and treated with Pronase (9) (90 proteolytic units/ml, 0.001 M CaCl2 at 37 C for 5 h with constant, gentle stirring). Pronase was inactivated by addition of 0.1 ml of 0.2 M ethylenediaminetetraacetic acid and heating to 80 C for 30 min. The precipitate was again collected and washed three times in acidified ethanol, and then was dried and redissolved in .05 M Tris buffer, pH 7.4, for DNA hydrolysis by the sequential action of deoxyribonuclease I and snake venom phosphodiesterase (8). The supernatant fluids of four acid-ethanol washes were combined and reduced to dryness by evaporation and lyophylization. This material was fractionated on a Dowex-1-formate column (1); fractions containing deoxycytidylic acid (dCMP) were pooled and lyophylized, and the purity and identity of the dCMP were confirmed by determination of its ultraviolet absorption spectrum and its behavior on TLC in three solvents. dCMP-1l4C was deaminated to deoxyuridine monophosphate (dUMP- 14C; 6), and dUMP was purified by chromatography on Dowex-1-formate, lyophylized, and treated with intestinal alkaline phosphatase. The radioactive product, UdR-dR-"4C, was preparatively purified by descending paper chromatography (on Whatman 3 MM) in 86% n-butanol. Preparation of uniformly labeled UdR,14C. The starting material, commercial CdR- 14C (U) was deaminated by nitrous acid treatment (6). The reaction mixture was lyophylized, and the UdR in the resuspended residue was purified by descending paper chromatography in 86% n-butanol.

RESULTS Selective incorporation of pyrimidine deoxynucleosides into DNA. Exogenous deoxyribonucleosides of uracil, cytosine, and thymine are effectively used for nucleic acid synthesis by pneumococci. The utilization of these precursors is unusual in its selectivity, relative to most other cell types which have been studied; particularly when supplied at low concentrations (approximately 0.1 ,ug/ml), UdR and CdR are, like TdR, preferentially used for DNA synthesis (Table 1). Figure 1 shows that this selectivity of incorporation of UdR prevails even in long-term labeling experiments.

1358

BEAN AND TOMASZ

J. BACTRIOL.

roughly in direct proportion to the increasing precursor concentration supplied. Thus, for the 60-min incubation period, the greatest selectivCdR TdR ity for incorporation into DNA (i.e., the highest ratio of DNA per total nucleic acid) occurs 0.11 5.0 when the concentration of supplied precursor is 42.7 24.2 lowest. If high selectivity and high-specific6.4 0.0 activity labeling are desired, low UdR concen50.6 75.8 trations and moderately long incubation times 0.89 1.00 combine to give effective results such as those added to midloga- shown in Fig. 1A.

TABLE 1. Utilization of pyrimidine

deoxyribonucleosides by pneumococcusa Precursor

UdR

UdR

(p/mI) Soluble pool RNA DNA DNA/NA

0.08 30.6 4.8 64.6 0.93

5.0 38.5 20.5 41.0 0.67

aRadioactive precursors were rithmic-phase pneumococcal cultures to give final concentrations of: UdR 6-3H, 2.5 ACi/0.08 g/ml; UdR-2-14C, 0.62 pCi/5.0 pg/ml; CdR-14C (U), 0.1 ACi/0.11 g/ml; and TdR-2-14C, 0.62 liCi/5.0 g/ml. Whole cell incorporation in counts per minute per milliliter were as follows: UdR, 76,085 and 41,250; CdR, 12,990; and TdR, 4,750. Data given are for samples taken after 60 min of incubation in the presence of the isotope and are reported as the relative percentages of the total incorporated by the cells, which was recoverable in the cellular soluble pool and in KOH-sensitive (RNA) and KOH-insensitive (DNA) trichloroacetic acid-precipitable material.

Both Table 1 and Fig. 1 (part A versus B) demonstrate a second peculiarity of deoxynucleoside utilization. For a relatively short time after the addition of higher concentrations of radioactive UdR (> 2 l/ml), a fraction of the available UdR-2- 4C incorporates into alkalisensitive (RNA) material. Thus, in short-term labeling experiments (e.g., 60 min, as in Table 1), utilization of UdR for RNA and DNA syntheses shows strikingly different patterns, depending on the precursor concentration used. In such long-term experiments, such as that reported in Fig. 1B, the incorporation of UdR2- 4C into RNA reaches a maximum value shortly (15 to 30 min) after addition of the deoxynucleoside to the growth medium; thereafter, the absolute amount of radioactivity in the RNA fraction remains constant. Table 2 shows how the concentration of precursor available influences the initial (i.e., 60 min) labeling of pneumococcal nucleic acids. Although the specific activity of the precursor decreased by a factor of more than 1,000 from the lowest to the highest concentration of supplied UdR, the absolute number of counts incorporated into DNA decreased by only about 35 times, and incorporation into RNA showed relatively small variation. Thus, the absolute amount of precursor used for DNA synthesis increases as the concentration of exogenous UdR increases, but as less than a direct proportion of that concentration. The initial amount of precursor used for RNA synthesis, on the other hand, does increase

cpm /ml 10,000

xDNA

UdR-2j4C 0.075kg/ml

x x 5,000

100 Ud

200

300

R-2-'4C16,000

x

x'

2.5 ,tg/ml

8,000

4

.-

e.-

*

Io

200 300 FIG. 1. Utilization of UdR for nucleic acid synthesis. At 0 min, cultures were supplemented with UdR-2-"4C at final concentrations of (top) 0.15 uCi/.075 mg/ml and (bottom) 0.1 ,uCi/2.5 pg/ml. Data are reported as counts per minute per milliliter of culture incorporated into DNA (crosses and solid lines) and into RNA (circles and broken lines) as a function of time (minutes) after addition of the precursor. TABLE 2. Concentration dependence of initial UdR labeling of nucleic acidsa UdR (,gml)

UdR/2.1 ml of culture (pumoles/10 pCi)

0.152 0.370 2.32 21.9 218.0

0.0014 0.0034 0.0214 0.2014 2.0014

RNA

DNA

13,870 67,648 17,120 40,122 7,054 12,666 9,304 3,558 11,186 1,974

DNA/NA

0.83 0.70 0.64 0.28 0.15

aSamples of a midlogarithmic-phase pneumococcal culture were supplemented with UdR to give the final concentrations of UdR-2- 4C specified. After 60 min of further incubation, all subcultures were sampled for determination of isotope distribution.

VOL. 113, 1973

1359


The mechanism of UdR incorporation into DNA. Tables 3 and 4 summarize several lines of evidence indicating that the UdR molecule incorporates intact (i.e., without loss of the deoxyribose moiety) into TdR of DNA. The pathway of incorporation seems to lead through the formation of thymidylic acid, because (i) labeled UdR does not incorporate into DNA in TdR-requiring mutant lacking thymidylate synthetase, (ii) UdR (and CdR) labeled with tritium in the 5 position does not incorporate into DNA, (iii) competing thymidine, but not uracil, seems to dilute the amount of labeled UdR that reaches DNA, and (iv) in hydrolysates of DNA prepared from cells labeled with radioactive UdR (with isotope either in the deoxyribose or in the pyrimidine moieties of the molecule), 98% of the labeled DNA residues proved to be TdR. The pathway of UdR incorporation into RNA. The smaller amount of UdR that incorporates (during early periods of UdR utilization) into alkali-sensitive material seems to originate from a reaction in which the deoxyribose portion of the molecule is lost, with free uracil as a possible intermediate. This conclusion is based on the following lines of evidence (also shown in Tables 3 and 4): (i) the usual amount of UdR still incorporates into RNA in the TdR-requiring pneumococcus; (ii) competing uracil (but not TdRJ can completely block incorporation of UdR into RNA; and (iii) only the pyrimidine ring (and not the deoxyribose) portion of UdR incorporates into RNA. The mechanism of the concentration dependence and transient nature of UdR incorporation into RNA is not understood at present. Conversion of UdR to TdR and excretion of TdR into growth medium. An additional novel aspect of pyrimidine deoxynucleoside metabolism in pneumococcus is revealed through the analysis of the growth medium of cultures utilizing radioactive UdR or CdR. During the utilization of UdR, substantial amounts of the precursor are converted to TdR, which is recoverable from the culture medium, as is shown in Table 5. In cultures supplied with radioactive CdR, both UdR and TdR appear in the growth medium. The compounds recovered from the growth medium show no loss of pentose radioactivity; i.e., there has been no cleavage of the glycosidic bond during the interconversion of the deoxynucleosides. In these experiments there was no evidence for cellular lysis or for extracellular nucleoside kinase, nucleotidase, thymidylate synthetase, or deoxycytidine deaminase activities. Thus, the conversion of exogenous UdR to TdR and of

TABLE 3. Variables relevant to selective labeling of DNA ExperExperPrecursor iment

Conditions

A.a 1. UdR-6-3H 2. UdR-6-3H + Uracil 3. UdR-6-3H + Thymidine

Nucleic Nucleic acidin labeled label DNA (counts/ min/ml) (%)

~ ~ ~ acid

52,820 93.0 41,510 100.0 12,400 38.0

B.b 1. UdR-2-'4C Wild-type cells 2. UdR-2-'4C TdR- cells

278 97

72.0 21.0

C.c 1. UdR-5-3H

494 60

96)

(B) UdR-2-14C

0.08

1

0.94

(C) UdR-2- l4C (D) UdR-dR-14C

0.1 0.07

4

3.3

0.90 0.99

T C MeC dTMP TdRb

(98.0) (0.6) (1.4) (>97) (>97)

(E) UdR-4C (U)

2.5

0.7

0.69

TdRC (>97)

UMP CMP UMP CMP

(0) (0) (60) (40)

UMP (0) CMP (0)

UMPd (50) CMPd (50)

In experiment A, labeled DNA was prepared as purified transforming material (7); in B to E, DNA was prepared as KOH-insensitive trichloroacetic acid-precipitable material. In A, C, D, and E, DNA was hydrolyzed enzymatically; in B, DNA was hydrolyzed with HC104. 5Radioactivity was present exclusively in the deoxyribose portion of the recovered thymidine. c Radioactivity was distributed between the pyrimidine ring and deoxyribose moieties of the recovered TdR in the proportion 4:5, i.e., the same proportion as the uniformly '4C-labeled starting material. d UMP and CMP recovered contained little or no radioactivity in their pentose moieties. a

TABLE 5. Secretion of metabolic derivatives of UdR and CdR into the growth mediuma Com-

CdR- 14C (U)

UdR- "C ( recovered oxyribose- UdR-2-14C ld Wild (U) (relative 14C (U) FUNt _ type 3h)

pounds

UdR-de-

CdR 0.0 UdR 22.0 TdR 72.3 dTMP 5.7

0 0 100 0

0.0 93.3 5.0 1.7

16.1 97.7 76.8 2.0 11.1 0.3

a Logarithmically growing pneumococci were grown in the presence of: UdR-deoxyribose- 14C, 7,100 counts/min/0.07 sg of UdR-(dR-'4C) per ml for 3.5 bacterial generation times; UdR-2-'4C, 0.01 /lc/0.08 ug of UdR-2- 4C per ml for four generation times; UdR-"4C, 0.06 MCi/2.5 ,g of UdR-'4C (U) per ml for 0.6 generation times; and CdR-'4C (U), 0.1 ,Ci/0.11 ,ug of CdR- 4C (U) per ml for four generation times. After removal of cells, the radioactive components of the growth medium were identified by thin-layer chromatography. There was no detectable loss of radioactivity from the pentose moieties of the uniformly labeled precursors or their derivatives.

the cell. This does not preclude the possibility that small amounts of free extracellular or intracellular TdR are again phosphorylated and used in the pathway toward DNA synthesis, as is the case when TdR is itself supplied as an exogenous precursor. The actual extent of TdR excretion in experiments like those shown in Table 5 is reproducible under a given set of conditions, but varies, as expected, with the cell concentration, precursor concentration, time of incubation, etc.

A single-marker mutant, fun, apparently defective in the active transport system for pyrimidine nucleosides (2), used CdR for DNA synthesis at less than 4% of the levels achieved by the wild type. During this experiment, as shown in Table 5, the mutant strain did achieve the conversion of a few percents of the available CdR to UdR and TdR, but contained no detectable radiolabeled CdR or CdR-derived compounds in its cellular soluble pool.

DISCUSSION Pneumococci possess little or no capacity for the enzymatic cleavage of pyrimidine nucleosides, and this unusual circumstance has several interesting consequences in the pyrimidine metabolism of this species. As was previously shown, FU, FUR, and FUdR each cause different and distinct inhibitory symptoms in pneumococcus, and each of these analogs is subject to a different pattern of metabolic utilization (1). It has thus been possible to distinguish several of the different inhibitory roles of the fluoropyrimidines in this organism (1) and to isolate a series of distinct single-marker fluoropyrimidine-resistant mutants which do not show reciprocal cross-resistance to FU and its nucleosides (2). The selective utilization of the deoxyribonucleosides of uracil and cytosine for DNA synthesis, as reported here, represents another consequence of the stability of the pyrimidine N-glycosidic bond in this bacterium. The biochemical pathways that operate in

VOL. 113, 1973

1361


the utilization of exogenous pyrimidine bases and nucleosides by pneumococcus are summarized in Fig. 2. The only major difference between this scheme and that applicable to the gram-negative species E. coli (10) and Salmonella typhimurium (3, 10) is the inactivity in pneumococcus of enzymes that catalyze cleavage of the N-glycosidic bonds of pyrimidine ribo- and deoxyribonucleosides. Pneumococcus is also unlike the lactobacilli: pneumococcus apparently does not contain active nucleoside trans-N-deoxyribosylase (1) or deoxycytidine kinase (2, 5). Under the conditions used in these experiments, there is no detectable catabolism of pyrimidine nucleosides (beyond deamination). Evidence for the metabolic sequences indicated in Fig. 2 came from several different types of experiments, published either here or in the two previous publications in this series (1, 2). We shall only summarize the essence of the evidence here. Evidence from independent studies on resting cell suspensions of other strains of Diplococcus pneumoniae (5) is also compatible with the scheme illustrated in Fig. 2. Uracil (and FU) utilization proceeds through UMP-pyrophosphorylase; FU-resistant (FUr) mutants lack this activity and are unable to utilize uracil or FU, whereas they can normally metabolize uridine and deoxynucleosides. Cytosine utilization seems to proceed through deamination to uracil; a fluorouracilresistant mutant (UMP pyrophosphorylase deficient) proved to be cross-resistant to fluorocytidine (FC). Because no pyrophosphorylase is known to convert cytosine to cytidylic acid, the cross-resistance probably results from initial deamination of FC to FU in vivo. Because FC is a much less potent inhibitor than FU and because, relative to uracil, cytosine is a poor reversing agent for the fluoropyrimidine inhibitions (1), it appears that the permeation or deamination of cytosine is the limiting step in its utilization. Uridine (and FUR) is utilized through uridine kinase. The pneumococcal mutant udk, which lacks this enzyme, is defective in uridine utilization, but has normal uracil and deoxynucleoside metabolism. In vivo and in vitro experiments show no evidence for the cleavage of uridine to uracil. A pyrimidine-dependent mutant of pneumococcus, which was transformed to FU resistance (U. upp), has a nutritional requirement for uridine which can no longer be satisfied with uracil. The utilization of exogenous UdR involves early and extensive conversion to a 5-methyl derivative without loss of the deoxyribose moiety. Although a likely enzymatic sequence for

denovo pathway

U-R UMP C

U

RNA-CTP

UDP-- UT P-

dUDP-dUTP

CdR-UdR---dUMP-dTMP*

dTDP

dTTP

TdR

FIG. 2. Pyrimidine metabolic pathways in pneumococcus. Activities demonstrated in vitro are denoted by an asterisk. The broken arrow indicates the presence of transient, low levels of activity for the glycosidic cleavage of UdR.

this would be a kinase reaction followed by thymidylate synthetase, we were unable to demonstrate deoxyuridine kinase activity in extracts of wild type pneumococci. Nevertheless, extensive in vivo experiments reported in this paper strongly suggest that such an enzyme does function in vivo. FUr and FURr mutants are still sensitive to FUdR. Mutants resistant to the fluoropyrimidine nucleosides (FUNF) are still sensitive to FU but show a nonreciprocal cross-resistance against FUR. We suggested that the basis of this nonsymmetrical cross-resistance is the existence of a nucleoside transport system (for both ribo and deoxyribonucleosides) that is defective in the FUNT cells. We can offer no biochemical explanation for the transient, relatively minor, and concentration-dependent utilization of UdR for RNA synthesis, except that it seems to involve a break of the N-glycosidic bond. Evidence presented in this paper indicates that CdR utilization seems to involve deamination to UdR as the first step. Neither CdR-53H nor UdR-53H incorporate into DNA (to give rise to cytosine in DNA), so that methylation at the 5 position seems to be an obligatory step for the utilization of these deoxynucleosides. The observation of TdR excretion reported in this paper requires comment. It is possible that the early steps in the metabolism of these compounds occur near the cell surface by periplasmic enzymes and that the metabolic conversion of extracellular deoxynucleosides represents leakage from these near-surface sites back to the medium. However, the extent of this "leakage" appears to be too extensive for logarithmically dividing bacteria and, therefore, it is tempting to consider TdR secretion as a normal physiological function. Cellular levels of TdR and its nucleotides seem to be subject to careful control systems in many organisms through the operation of a phosphorylase that converts TdR to the metabolically less active thymine (10, 11). Pneumococcus may have

1362

BEAN AND TOMASZ

evolved a different mechanism to achieve the same end, through the regulated uptake and secretion of TdR through the cell membrane. LITERATURE CITED 1. Bean, B., and A. Tomasz. 1971. Inhibitory effects and metabolism of 5-fluoropyrimidine derivatives in pneumococcus. J. Bacteriol. 106:412-420. 2. Bean, B., and A. Tomasz. 1973. 5-Fluoropyrimidine3.

4.

5. 6.

resistant mutants of pneumococcus. J. Bacteriol. 113:1348-1385. Beck, C. F., J. L. Ingraham, J. Neuhard, and E. Thomassen. 1972. Metabolism of pyrimidines and pyrimidine nucleosides by Salmonella typhimurium. J. Bacteriol. 110:219-228. Davis, B. D., and E. S. Mignoli. 1950. Mutants of Escherichia coli requiring methionine or vitamin B,2. J. Bacteriol. 60:17-28. Firshein, W., and P. Hasselbacher. 1970. Utilization of deoxyribonucleosides by virulent and avirulent pneumococci. Biochim. Biophys. Acta 204:60-81. Friedkin, M. 1963. Assay of thymidylate synthetase

J. BACTERIOL.

activity, p 124-130. In S. P. Colowick and N. 0. Kaplan (ed.), Methods in enzymology, vol. 6. Academic Press Inc., New York. 7. Hotchkiss, R. D. 1957. Isolation of sodium deoxyribonucleate in biologically active form from bacteria, p. 692-696. In S. P. Colowick and N. 0. Kaplan (ed.), Methods in enzymology, vol. 3. Academic Press Inc., New York. 8. Karlstrom, O., and A. Larsson. 1967. Significance of ribonucleotide reduction in the biosynthesis of deoxyribonucleotides in Escherichia coli. Eur. J. Biochem. 3:164-170. 9. Nomoto, M., Y. Narahashi, and M. Murakami. 1960. A proteolytic enzyme of Streptomyces griseus. V. Protective effect of calcium ion on the stability of protease. J. Biochem. (Tokyo) 48:453-463. 10. O'Donovan, G. A., and J. Neuhard. 1970. Pyrimidine metabolism in microorganisms. Bacteriol. Rev. 34:278-343. 11. Rosenbaum-Oliver, D., and S. Zamenhof. 1972. Degree of participation of exogenous thymidine in the overall deoxyribonucleic acid synthesis in Escherichia coli. J. Bacteriol. 110:585-591.