METABOLISM OF CYTOSINE, THYMIKE, URACIL, AND BARBITURIC ACID BY BACTERIAL ENZYMES* BY OSAMU HAYAISHIt AND ARTHUR K0RK;RERG (From the National Inslitule of Atthrilis and Metabolic Diseases, National Inslilules of Health, United States Public Heallh Service, Belhesda, Maryland) (Received for publication, March 17, 1952) Early studics with pyrimidines administered to animals established the fact that these compounds arc metabolized (3, 4). Present concepts of the pathways of pyrimidine metabolism stem largely from the work of Cerecdo (4), in which he measured the changes in the urinary excretion of urea after feeding pyrimidines and related substances. On the basis of these studies, he proposed an initial oxidation at carbon 5, which in the case of uracil yields isobarbituric acid and with thymine results in thymine glycol. It was further proposed that oxalic acid, formic acid, and urea are ultimately obtained from uracil and that acetol, carbon dioxide, and urea result from thymine. Howevcr, the indirect nature of the evi- dence leaves this scheme open to question. Our approach to this problem has been first to obtain, through enrich- ment culture, bacterial strains which rapidly metabolize pyrimidines and then to investigate the detailed metabolic pathway with enzymes obtained from these bacteria. Results both with whole cells and partially purified enzymes have suggested the pathway of pyrimidine metabolism indicated in Fig. 1. Mehds Muterials-Pyrimidines were commercial products, the identity of which ww verified spectroscopically. l Isobarbituric acid was prepared according to Rehrend and Roosen (5) and possessed the extinction coefficients cited by Heyroth and Loofbourow (6). Krease and protamine sulfate were products of the Nutritional Biochemicals Corporation. 5-Methylbarbituric acid was synthesized by the method of Gerngross (7). The melting point (203-204", uncorrected) indicated identity with the CY form described by Nishikawa (8). The analysis2 was as follows: Calculated, C 42.29, H 4.26, N 19.23; found, C 42.50, TI 4.29, N 19.46 __- __ ~-____-- -- * Preliminary reports of this work have been presented (1, 2). t Special Research Fcllow, Unitcd States Public TTetilth Service. 1 A commercial sample of cytosine was fouritl to contain loss than 2 per cent of 3 Microanalyses were performed by the Microanalytical Laboratory of the cytosine and over 95 per cent of uracil. National Institutes of Health undcr the supcrvision of Dr, W, C. Alford. 717 718 PYRIMIDINE METABOLISM Spectroscopic examination showed fairly close agreement with the data of Stuckey (9), except that the absorption maximum at pH 7.0 is at 267 mp arid in 0.1 N NaOH is at 269 mp (e = 19,700 and 17,300, respectively); Stuckey reported an absorpt,ion maximum in 0.1 N NaOH at 262 mp (c = 19,200). As previously observed (8, 9), this compound is unst.able at acid and neutral pH and is readily oxidized by air to yield 5-hydroxy-5- methylbarbituric acid, which has only negligible absorption in the ultra- HN - C-0 HN-C-0 N H2 COOH I I t2H20 I I II -NH3 -Cc2 'CH -0-C CH 'A0-C CH2 - C-0 + CH2 - - 1 I1 I I1 11 I I HN - CH HN - CN HN - C-0 N"2 COOH I6 ;YTOSINE URACIL BARBITURIC UREA MALONI ACID ! + ACID 2NH3 4- Cop C02 + &C N C-NH2 HN- C-0 HN- C-0 HN - CH HN - C-0 THYMINE 5 -METHYL BAR 81- FIQ. 1. Pathway of bacterial metabolism of pyrimidines. The numbering system for the pyrimidine ring conforms to current usage by Chemical Abstracts. The old system waa employed previously (1, 2). TURIC ACID violet region. The stability at various temperatures and pH values is shown in Table I, Cell-free extracts of bacteria were prepared by grinding with alumina by the method of MoIlwain (10). Washed cells were ground with twice their weight of alumina (Alcoa A-301) at 0" for 5 minutes. The paste was extracted with 10 parts of buffer (see below) and the mixture was cen- trifuged at 0" at 16,000 X g for 10 minutes. All subsequent manipulations in the purification of enzymes were conducted at 0-2". Assay of Uracil-Thymine Oxidase4ince the extinction of the oxidation products vas much greater than that of the substrates, the rate of increase in optical density served as the basis for measurement. The test system (at 22-23') contained 0.3 ml. of uracil or thymine (0.001 M), 0.3 ml. of met.hylene blue (2.67 X M), 2.3 ml. of tris(hydroxymethy1)amino- methane buffer (0.02 M, pH 8.5), and 0.1 ml. of enzyme. Readings were 0. HAYAISHI AND A. KORNBERG 719 Extincdon at tS. 0 2.5 16 taken at 2 minute intervals (at 255 mp in the caae of uracil and at 270 mp in the cam of thymine) in a model DU Beckman spectrophotometer. A unit of enzyme waa defined as the amount producing a density increase of 0.100 during the first 10 minutes and specific activity waa defined aa units per mg. of protein. With about 1 unit of enzyme in this test system, the reaction rate ia linear for about 30 minutes and proportionality is observed between rate and the amount of enzyme added (0 to 5 units). Assay of Burbiiturase-The test system (22-25") contained 0.1 ml. of barbituric acid (0.02 M, sd~m salt), 0.3 ml. of glycylglycine buffer (0.2 M, pH 8.25), 0.4 ml. of water, 2.5 mg. of crystalline bovine serum albumin, and 0.2 ml, of enzyme. At 10 minute intervals, 0.1 ml. aliquota were re- moved and the reaction waa stopped by dilution to 3.0 ml. with phoaphate buffer (0.02 M, pH 7.0). Readings were taken at 255 mp. A unit of en- TABLEI I Stability of 6-Methylbarbituric Acid As Influenced by Time, Xemperature, and pH pH 1.9 pH 6.9 pH 9.7 - I 25* 0' 2.5. 0. 25. 0. - -- 16,430 16,430 16,820 16,820 16,820 16,820 13,720 16,200 13,750 16,600 17, OOO 2,850 470 15, Mw) 4, 16,800 16,000 The concentration of b-methylbarbituric acid wm 0.001 M. The values are the optical density units at 270 at a final pH of 7.0 calculated for a 1 M solution in a cell of 1 cm. light path. zyme wa defined as the amount producing a density decrease of 0,100 in a 10 minute interval. Protein waa determined by the method of Lomry et a/. (11). Ammonia was distilled by the microdiffusion method of Conway (12) and determined by nesslerization (13). Ion exchange chromatography was performed ac- cording to Cobn (14). Dowex 1 anion exchange resin (200 to 400 mesh) was prepared by washing the resin first with 3 N HCl until the washings were free of material absorbing at 260 mp and then with distilled water until the chloride ion test was faint. Paper chromatography was carried out with Whatman No. 1 filter paper by the ascending method. After 16 hours at room temperature, the paper was dried and the spots visualized with a model V41 Mineralight fluorescent lamp. Results Experiments with Intact Cells lsolalimt and Properties of Orgunism-Samples of soil were suspended in 0.8 per cent NaCl solution (1 part of soil to 9 parts of saline). 1 mi. of 720 PYRIMIDINE METABOLISM soil suspension was incubated in 9 ml. of a medium containing K2HP04 (0.15 per cent), KHzPOd (0.05 per cent), MgSOd*7H20 (0.02 per cent), and thymine and uracil (0.1 per cent each) in distilled water. After growth of the organisms became visible, several transfers were made successively in the same medium at daily intervals and finally the culture was plated out on this medium containing 2 per cent agar. The strain which grew most abundantly was isolated and designated m strain 161. Strain 161 is an aerobic, gram-positive rod, non-motile and non-acid-fast, and does not form spores. It shows considerable pleomorphism; 16 hour cultures are made up exclusively of rod-shaped organisms, but after 36 hours (at 30") there are irregular forms and granular staining. On the basis of these morphological characteristics, strain 161 has been tentatively assigned to the genus Co~ynebacleri~m.~ A strain of Mycobmterium' was subsequently found to metabolize uracil arid thymine in a similar way and, because of its rapid growth, was used for enzymatic studies. It shows similar morphological properties to strain 161 except that it retains Ziehl's fuchsin after treatment with 5 per cent sulfuric acid for several seconds. Strain 161 was cultured in a medium containing thymine or uracil (0.1 per cent) as the sole source of carbon arid nitrogen; the salts were as de- scribed above, The organisms were grown in 20 liter glass carboys con- taining 10 liters of the medium at about 26" for 40 hours, with constant mechanical shaking. Cells were harvested by centrifugation in a Shnrples supercentrifuge, washed once with a 0.5 per cent NaC1-0.5 per cent KC1 solution, and suspended in phosphate buffer (0.02 M, pH 7.0) at a concen- tration of approximately 1 mg. of dry weight per ml. The yield of cells WM approximately 0.5 gm. (wet weight) per liter of medium. Manomelric Studies4xygen uptake by a suspension of resting cells cultured on thymine was determined with thymine, uracil, or barbituric acid as the subutratc. As shown in Fig. 2, each of these substrates brings about an immediatc consumption of oxygen, the total uptakc corresponding approximately to 4, 3, and 2 atoms per mole of thymine, uracil, and bar- bituric acid, respectively. These values represent about 50 to 60 per cent of the theoretical calculated for complete combustion and suggest that the remainder of the carbon has beer1 assimilated into cell muterial. When the organism was grown on an ordinary broth medium, there was a long lag period before oxygen uptake was observed. That this lag was duc to the adaptive formation of thymine-oxidizing enzymes was verified by a 3 We are indebted to Dr. C. B. van Kiel and Dr. R. Y. Stanier for advice concern- ing this classification. 4 This atrain, originally isolated from rabbit feces by Dr. Schstz, Dr. Sward, and Dr. Pintner, waa classified and kindly furnished to ua by Dr. T. Stadtman. 0. IIAYAISHI AND A. KORNBERG 721 determination of thymine oxidase activity of cell-free extracts of cells before and some hours after exposure to thymine. Manometric experiments with cells grown on uracil gave essentially t.he same results as those obtained with cells grown on thymine. Growth on uracil was less rapid than on thymine and the latter substance was usually employed for large scale cultivations. Spectrophotometric Studies; Isolation of Barbituric Acid-Gltraviolet spectrophotometry carried out during the course of uracil or thymine oxi- dation revealed a transitory increase in density indicating the formation of intermediates with higher extinction coefficients than the substrates.s A large scale experiment on uracil oxidation designed to pennit isolation TIYE IN HOURS FIG. 2. Oxygen uptake by cells grown in the presence of thymine (solid line) and in the absence of thymine (dotted line). In the absence of thymine, 0.1 per cent of peptone was substituted. Each Warburg vessel contained 2 phf of substrate and about 1 mg. (dry weight) of cell material in a total volume of 2.0 ml, of 0.02 M phos- phate buffer, pH 7.0. 0.2 ml. of 10 per cent KOH wm in the center well. Tempera- ture, 31.5'. of the product was carried out as follows: Uracil (112.1 mg., 1.0 mM) was incubated at 37" with constant Rtirring with 50 ml. of a resting cell sus- pension and 50 ml. of phosphate buffer (0.02 M, pH 7.0). Aliquots (0.1 ml.) diluted to 25 ml. with phmphate buffer (0.02 M, pH 7.0) were ex- amined at 260 mp at hourly intervals. There was a steady increase in absorption, which at 4 hours reached a value almost 3 times that of the original; thereafter, the absorption decreased. At 4.25 hours, 90 ml. of the reaction mixture were centrifuged to remove the cells. The absorp- 6 The extent and duration of this incrcasc varied with different batches of cells relatcd possibly to the age of the culture. For examplc, with thymine-grown cells (48 hours old) with thymine a6 substrate, the dcnsity (at 270 mr) decreased, while with uracil as substrate, the density (at 255 mp) increased initially. On the other hand, uracil-grown cells (48 hours old) usually brought about an initial denaity increase with thymine and a dccreasc with uracil. 722 PYRIMIDINE METABOLISM tion spectrum of the supernatant at acid, neutral, and alkaline pH LVM essentially identical with that of the recrystallized reaction product and an authentic sample of barbituric acid. The supernatant solution was concentrated in vacuo to about 5 ml. and acidified to pH 1.0 with HC1. White rhombic crystals appeared on standing at 0" overnight. A sample recrystallized from dilute HC1 (12.1 mg.) melted at 240-243"; mixed with authentic barbituric acid (242-2-13"), it melted at 240-243" (uncorrected). The nitrogen content was 21.41 per cent; calculated value 21.90 per cent. Paper chromatography of this product in three different solvent systems provided additional support of its identity with barbituric acid (Table 11). As indicated by the results of manometric studies described above (Fig. TABLE I1 Paper Chromatogruphy of Uracil Oxidation Product Thymine. .................................. Uracil. .................................... Barbituric acid.. ......................... Solvent A I Solvent B Solvent c --. - _- - Rr valum 0.47 j 0.86 0.72 0.33 0.65 1 0.59 - -- __. - - .. 0.62 0.33 ! 0.20 ........................ 0.20 0.47 Reaction product. Isobarbitnric ncid ........................ - . - ._ -. o*62 I o*32 i __ I _. . ..... - - I--_ . Solvent A, butanol saturated with 10 per cent aqueous urea solution (Carter, C. E., J. Am. Chem. SOC., 72,1466 (1950)). Solvent B, butanol, ethylene glycol, and water, 4:l:l (Vischer, E., and Chargaff, E., J. Biol. Chem., 168, 781 (1947)). Sol- vent C, propanol and water, 10:3. 2), cells grown on either thymine or uracil were adapted to the oxidation of barbituric wid. They were not able to oxidize a large number of other pyrimidines tested including cytosine, 5-methylcytosine, isobarbituric acid, 6-methyluracil, dihydrouracil, 2-thiouracil, and 2-thio-5-methyluraci1, With regard to 5-methylbarbitiwic acid, the anticipated product of thy- mine oxidation, manometric experiments were complicated by its insta- bility, but spectrophotometric measurements revealed essentially the same rate of decomposition as that of barbituric acid. Spectrophotometric Studies with Cells Grown on Cylosine-When cytosine was substituted for thymine or uracil in the culture media, the growth was much less favorable unless an additional carbon source, such as glucose, was added. Suspensions of cells grown on a cytosine-glucose (0.1 per cent nf wwhl mdiiim undw Rimilar mndithns met,aholizd either cvtosine or 5-methylcytosine immediately as judged by a decrease in optical density at 265 mp (Table 111). Cells grown on either thymine or uracil (with or 0. HAYAISHI AXD A. KORNBERG 723 Thymine (glucoee).. ........... Uracil (glucose). .............. without glucose) did not metabolize cytosine or 5-methylcytosine under the same test conditions, However, the cytosine-adapted cells produced changes in light absorption with uracil or thymine as substrate which re- semble those produced by cells grown on thymine or uracil. It may be noted that the growth substrate influences the nature of the initial changes in optical density produced by intact cells. In some in- stances, there is an increase in density as an indication of the accumulation of an intermediate, while in others no such accumulation is observed. -0.090 4-0.045 O.Oo0 -0.005 4-0.213 -0.150 4-0.005 -0.005 TABLE 111 Influence of Growth Subslrale in Spectrophotometric Studies with Intact Cells __ - Test substrate I I Cytosine (glucose). ........... .I +0.205 1 Cruwtb substrate 4-0.213 -0.095 -0.100 1 CMethyl- 1 cytosine I Thymine Uracil Cytosine ~~~ A, optical density (0-30 mb.) ~- Pyrimidines (0.1 per cent) and glucose (0.2 per cent) in the basal salt mixture. After 48 hours, the cells were harvested, washed, and weighed. Yielde of 0.80, 0.42, and 0.38 gm. per 500 ml. of culture medium were obtained with thymine, uracil, and cytosine, respectively. The cell suspension ww made with 0.8 per cent KCl solution and the concentration was standardized to give a density of 0.44 at 6!N mp in the Coleman model 6B junior spectrophotometer. 0.2 ml. of cell suspension, 0.3 ml. of 0.001 M pyrimidine solution, and 2.5 rnl. of 0.02 M phosphate buffer (pH 7.0) were incubated in a Beckman cuvette and the reaction ww followed at no, 255,!2tjo, and 265 mp for thymine, uracil, cytosine, and 5-methylcytosine, respectively. Uracil-Thymine Oxidase Preparation of Enzyme-Large scale cultivation of Mycobacteria was carried out as described above for Corynebacteria (strain 161). About 0.5 to 1.0 gm. of wet cells was obtained per liter of culture medium and could be stored at -10" without loss in activity for a period of at least 6 months. Cell-free extracts prepared by grinding with alumina and ex- tracting with tris(hydroxymethy1)aminomethane buffer (0.02 M, pH 9.0) were treated with ammonium sulfate (24.5 gm. per I00 ml. of extract). The precipitate was removed by centrifugation and more ammonium sul- fate was added to the supernatant (10.5 gm. per 100 ml. of extract). The resulting precipitate, collected by centrifugation, was dissolved in tris- (hydroxymethy1)amhomethane buffer (0.02 M, pH 9.0) to a volume cor- responding to one-twentieth that of the extract. This fraction contained 724 PYRIMIDINE METABOLISM 35.6 units per ml. and 1.2 mg. of protein per ml. The specific activity of cell-free extracts (uracil as a substrate) was 10.9 when the cells were grown on a cytosine-glucose medium and 10.2 when uracil was substituted for cytosine. This fact coupled with the behavior of intact cells (Table 111) indicates that the initial &ep of deamination of cytosine or 5-methylcyto- sine is an adaptive process and these compounds are metabolized by way of uracil or thymine, as previously reported with other microorganisms (15-17). Isolation and Irlentificulion of Reaction Products-Uracil or thymine (1.0 ml. of 0.02 M) was incubated at 30" (with constant shaking) with enzyme (2.0 ml.), methylene blue (2.0 ml. of 2.67 X M), and tris(hy- droxymcthy1)aminomethane buffcr (15 ml. of 0.2 M, pH 8.7). The course of reaction was followed by the increase in optical density (at 255 mp for uracil and 270 mp for thymine). When the reaction was cornpletc, a small aliquot was removed for determination of the absoiption spectrum and the remainder was adsorbed at 2" on a Dowex 1 chloride column and eluted with ammonium chloride-ammonium hydroxide buffer (0.1 M, pH 9.9). All operations with the thymine oxidation product were performed in the cold to minimize the autoxidation of 5-met.hylbarbituric acid, the presumed product. The absorption spectra of the final uracil and thymine incubation mix- tures determined at acid, neutral, and alkaline pll are identical with the spectra of barbituric acid and 5-methylbarbituric acid respectively (Fig. 3). Additional eviderioe supporting identity of the uracil oxidation prod- uct with barbituric acid and of the thymine oxidation product with 5- methylbarbituric acid was provided by ion exchange analysis (Fig. 4). SpeciMty, pH Optimum, and Substrate AJinitg-Cnder the spectro- photometric tcst conditions described for uracil and thymine oxidation, the enzyme did not act, upon the following pyrimidines: bartiituric acid, isobarbituric acid, 5-methylbarbituric acid, 6-methyluracil, dihydrothy- mine, dihydrouracil, %thiouracil, 2-thio-5-mcthyluraci1, and cytosine. The dependence of reaction rate on substrate concentration is shown in Fig. 5, A. The Michaelis constants (18) calculated from these data are 0.35 X lW4 and 1.31 X 10-4 mole per liter for thymine and uracil, respectively. The optimum pH of the reaction is about 8.5; the activity at neutral pH is only one-tenth as great (Fig. 5, B). The enzyme is most stable at an alkaline pH. There is no appreciable loss of activity for at least several months on storage at pll 9.0 at -110". The fact that the ratio of the rate of thymine to uracil oxidation is al- most identical in celI-free extracts from either uracil- or thymine-adapted cells or in the partially purified engyme preparations indicates the identity 0. HAYAISHI AND A. KORNBERG 725 WAVE LENGTH hp) "0 too 200 300 400 500 YL. THROUGH COLUMN FIQ. 3 FIG. 4 Flu. 3. Abeorption spectra of reaction producte and authentic barbituric acid and !%methylbarbituric acid. The thymine and uracil oxidation products were chromato- graphed on an ion exchange reein column (see Fig. 4). The eluates were pooled and used for determination. Conditions aa described in the text. A 9.5 cm. X 1.Osq. em. column waa used. The rate of flow waa about 19 ml. per 30 minutes. Recoveries bssed on the absorption at 260 mp were 88.0,90.5,95.0, and 82.0 per cent, respectively, reading from top to bottom. Fro. 4. Ion exchange chromatogram. I 2 3'18910 SUBSTRATE COWCf YTRATIOH PH (IO4 w 1 FIQ. 5. Rate cif thymine and uracil oxidation as a function of substrate concentra- tion (A) and pH (B). 0.03 ml. of enzyme (0.063 mg. of protein), 0.1 ml. of 0.001 M thymine or uracil, 0.3 ml. of 2.67 X IO-' M methylene blue, and 0.1 Y of buffer in a total volume of 3.0 ml. Glycylglycine buffer, pH 8.8, wa~ used in A. Thymine oxidsee activity was measured at 270 mp and uracil oxidam activity at 255 mp. In B, the values were corrected for the change of absorption coefficient of the substrates at different pH values. 0, phosphate; 0, glycylglycine; A, glycine. 726 PYRIMIDINE METABOLISM -.-- .- - Thymine, Y.. ............... Iiracil, M.. .................. KI of uracil. ................ Uracil, Y.. .................. Thymine, M.. ............... Uracil oxidizedt ............. KI of thymine. Thymine oxidized*. ......... .............. of the two oxidase activities. Further proof was provided by a kinetic analysis of the competitive inhibitory action of uracil and thymine, As shown in Table IV, the affinity of uracil for the enzyme was the same whether it was determined with uracil as a substrate or as a competitive inhibitor of thymine oxidation; similar results were obtained with thymine. Electron Tramport-When cell-free extracts were incubated with the TABLE IV KI of Wraeil and Thymine for Uracil-Thymine Ozidase lo-' I lo-' 0.356 0.292 0 1 10-4 1 .i9 x 10-4 - -.__ 10-4 10-4 0.177 0.066 0 10-4 0.3 x 10-4 2 x 10-4 1.08 x 10-4 10-4 2 x 10-4 0.35 x 10-4 0.240 0.042 -. -- 3 x 10-4 1.10 x 10-4 10-4 3 x 10-4 0.42 x 10-4 0.210 0.035 .- - The incubation mixture contained 0.03 ml. of enzyme (protein 0.063 mg.), 0.15 ml. of 2.07 X lW4 M methylene blue, uracil, and thymine, a.~ indicated, and 0.1 Y gIycylglycirie buffer (pH 8.0) to make a total volume of 1.5 ml. Depth of cells, 0.5 cm. After x) minutesat 26", readings were taken at both 250 and 270 mp and the amount of each substrat.e oxidized wlta determined. Ke for uracil = 1.31 X lO-' M; Ksfor thymine = 0.35 X lo-' M. KI, dissociation constant of inhibitorenzyme complex ; Ks, dissociation const ant of substrate-enzyme complex; I, concentration of inhibitor; S, concentration of substrate; u, velocily; VI, velocity in the presence of inhibitor. * Increaae in optical density at 270 mp corrected for the change due to the oxids- tion of uracil. t Increaae in optical density at 250 mp corrected for the change due to the oxida- tion of thymine. substrate, neither oxygen consumption nor substrate removal was ob- served. It was noted, however, that under anaerobic conditions rnethyl- ene blue was decolorized by these extracts in the presence of substrate and that the reaction also proceeded aerobically if methylene blue was present. The physiologic mechanism of electron transport has not been determined. Xanthine oxidase from milk and reduced diphosphopyridine nucleotide oxidase from Clostridium kluyveris did not serve as mediators in place of methylene blue. Reduction of added di- or triphosphopyridine nucleotide wa not detectable under the conditions employed nor was there any 6 Kindly furnished by Dr. Leon A. Heppel. 0. HAYAISHI AND A. KORNBERG 727 stimulation of the reaction by added adenyl coenzymes, metals, boiled enzyme, or boiled yeast ext.ract. Barbiturase Pu.r.iJicat.ion of Enzyme-Mycobacteria were cultured under essentially the same conditions a,a for the preparation of uracil and thymine oxidase with the exception that uracil (0.1 per cent) and glucose (0.2 per cent) provided the sole nitrogen and carbon sources. The inclusion of glucose increased the yield of cells to about 1.5 gm. per liter of culture medium and also increased the yield arid specific activity of the enzyme 3- to 4- fold. Cell-free extracts prepared by grinding with alumina and extract- ing with phosphate buffer (0,02 M, pH 6.63) were lyophilized and stored at -10". 500 mg. of lyophilized powder (obtained from 5.3 liters of cul- t-ure medium) were dissolved in 20 ml. of distilled water and insoluble material was centrifuged off and discarded. TO the supernatant (cell extract, Table V) were added 20 ml. of phosphate buffer (0.02 M, pH 7.0) and 4 ml. of protamine sulfate (10 mg. per ml.). After 3 minutes, the precipitate waa collected by centrifugation and extracted with 20 ml. of 0.5 M K2HI'04. The opalescent extract was diluted with 60 ml. of water. Removal of the resulting precipitate by centrifugation yielded a clear, colorless solution (protamine fraction). While this step yielded little or no purification on a protein basis, it succeeded in removing all the nucleic acid which was present in the cell-free extract. 10 ml. of the protamine fraction were adsorbed on it Dowex 1 formate column (8 cm. X 1 sq. cm.) and eluted with 0.1 M KZHPO4 at a rate of 0.3 ml. per minute. The eluate was tested for both urease and barbiturase activity and the fraction be- tween 18 and 22 ml. was observed to possess the highest specific activity of barbiturase and practically no urease activity; the urease activity ap- peared in a later eluate. Isolation and IdentiJiCatim, of Remiion Prod?v%s--With the crude cell- free extract, the disappearance of barbituric acid was matched by a total release of ammonia in amounts approximating the theoretical nitrogen content. However, a significant lag in KH3 production was observed (Fig. 6, A) which could be eliminated by the addition of crystalline urease (Fig. 6, B) . In the presence of urease 1 mole of COZ was evolved per mole of barbituric acid destroyed. With the purified enzyme only 0.16 mole of ammonia was released per mole of barbituric acid removed.' In the pres- 7 Similar resulta were obtained by inhibiting urease with silver ions. With crude cell-free extracts, 0.5 X 10-8 M AgCl inhibited barbituraee 27 per cent and urease 100 per cent. After the reaction waa completed, accumulation of urea could be demonstrated by diluting the reaction mixture and observing the effect of added urease. 728 PYRIMIDINE METABOLISM units per mg. protein WilJ 234 I 12.85 173 13.9 168 77 .O ence of added urease the value was 1.97 (Table VI), and similar values were observed during the entire course of the reaction. The reaction pro- ceeded at the same rate and to the same extent under anaerobic conditions. When urea had been established as a product, it was presumed that malonic acid might be the other product. Accordingly, a large scale ex- 0.55 1.45 1.47 TABLE V Purijication of Barbiturme 56 Uracil-glucose cell extract. ................ Protamine treat men t ....................... Dowex 1 column, Fraction at. ............. " 1 " (( b.. .............. -. 94.0 , 1.48 I Total activity Specific activity; 280:260* o Ratio of optical density at 280 mp compared with that at 260 mp. t Fraction a is the eluate between 13.5 and 36.0 ml.; Fraction b is the eluate be- tween 18 and 22.6 d. cs - 50- I I - 0 (A) UREASE NOT ADDED (8) UREASE ADDED > 120 0 60 TIME IH MINUTES FIG. 6. Disappearance of barbituric acid and formation of ammonia. 40 p~ of barbituric acid (sodium salt), 0.6 ml. of crude extract (3.36 mg. of protein), 1.2 ml. of water, and 0.2 ml. of 0.2 M glycylglycine buffer, pII 8.2. Barbituric acid waa measured spectrophotometrically at 270 mp. periment designed to isolate the product was performed; barbituric acid (25 ml., 0.04 M) was incubated with 20 ml. of cell-free extract (60 mg. of protein) and 5 ml. of glycylglycine buffer (0.2 M, pH 8.251. After 100 minutes at 26" (when the ultraviolet absorption of barbituric acid was completely removed) 2 N sulfuric acid was added to bring the pH to 2.2 and the resulting precipitate was removed by centrifugation. The super- natant was extracted with ether continuously in a Kutcher-Steudel ex- traction apparatus for 6 hours. The &her extract was taken to dryness arid the white solid sublimed at 60" under reduced pressure (0.05 mm. of Hg). The white crystalline product (86 mg.) gave the following analysis: 0. HAYAISHL AND A. KORNBERG 729 ~~ -.- Cell-free extract Purified enzyme found C 34.94 per cent, H 4.01 per cent, m.p. 134-135" (uncorrected); calculated (for malonic acid), C 34.60 per cent, H 3.84 per cent (melting point of authentic sample 134-135" (uncorrected)). The identity of the product with malonic acid was further verified by the characteristic yellow I !- Barbituric acid Ammonia Barbituric acid I Ammonia TABLE VI Accumulation of Urea with Cell-Free Extract and Purified Enzyme Without urease Irx -_ -8.5 +11.8 -6.8 +1.1 With ureaK PM -8.5 +16.0 -7.0 +13.8 The experiment with cell-free extract wtta carried out with 0.6 rnl. of extract (0.84 mg. of protein), 0.25 ml. of 0.04 M barbituric acid (sodium ealt), 0.2 ml. of 0.2 M glycyIglycine buffer (pH 8.2), and water to a final volume of 4.0 ml. The experi- ment with purified enzyme waa carried out with 1.1 ml. of enzyme (Fraction b, Table V), 0.5 ml. of 0.04 M barbituric acid (sodium salt), 0.2 ml. of 0.2 M glycylglycine buffer (pH 8.5), 0.2 rnl. of water, and 5 mg. of crystalline bovine albumin. 4 mg. of urease were added as indicated. Incubation, 1 hour at 24". 5 0.2- I I L I VI W f (A) (81 (A W W z 0.1 - s- c a! ... L- UL 2 H OO 5 IO 7 8 9 IO BARBITURIC ACID CUYCEICENTRATION ( !O'W) FIG. 7. Rate of barbiturase reaction as a function of substrate concentration (A) and pH (B). 0.1 ml. of enzyme (0.21 mg. of protein), 0.1 ml. of 0.02 M barbituric acid (sodium salt), and 0.8 ml. of 0.1 M buffer. Glycylglycine buffer (pH 8.0) wtuj used in A. 0, phosphate; 0 , glycylglycine; A, glycine. fluoresccncc under ultraviolet light upon heating with acetic anhydride (19) and by the characteristic appearance of its barium salt in dilute eth- anol solution (19). Specificity, pH Optimum, and Substrab AJinity-There was no action, as judged spectrophotometrically, on the following compounds : 5-methyl- barbituric acid, orotic acid, barbital, pentobarbital, 2-thiobarbituric acid, and isobarbituric acid. 730 PYRIMIDINE META DOLI SM The optimal pH is between 8 and 9 (Fig. 7, B) and the Michaelis con- stant (K,) calculated from the data shown in Pig. 7, A is approximately 3.37 X mole per liter. DISCUSSIOX It may be concluded from these and previous studies (15-17) that in bacteria aminopyrimidines such as cytosine and 5-methylcyt mine are first deaminated to produce uracil and thymine, respectively. This reaction appears to be due to a single enzyme which is adaptive in character. Lracil and thymine are both oxidized at carbon 6 by it single enzyme, which is also dsptively produced, to form barbituric acid and 5-methyl- barbituric acid, respectively. While this sequence of reactions is based on data derived from bacterial systems, it is a reasonable conjecture that the same pathway may be valid in mammalian systems. Concurrent with our preliminary report (I), Wang and Lampen (20) described the metabolism by certain bacteria of uracil and thymine to a product which was later identified as barbituric acid (21). Batt and Woods (22) have also reported the accumulation of a compound when thymine is oxidized by resting bacterial cells. They consider this com- pound to be a phosphorylated uracil-5-carbinol (2,6-dihydroxy-5-hydroxy- methylpyrimidine). However, their evidence for this formulation is not complete and the description of the spectroscopic characteristics of the compound do not preclude its identity with 5-methylbarbit.uric acid. Lttra (23) has investigated the metabolism of pyrimidines by several bac- terial strains and has obtained findings similar to those reported by us. He found that barbituric acid is the only substance among many listed which is oxidized without lag by cells adapted to uracil or thymine. He also found malonic acid to be it product of the enzymatic degradation of barbituric acid. The precise way in which barbituric acid is converted to urea and ma- lonic acid has not been established. The initial hydrolytic step can be assumed to yield a compound such as the half ureide of malonic acid with a subsequent hydrolytic cleavage to produce urea arid malonic acid. Even with the most purified enzyme preparations there has been no indication of the accumulation of such an intermediate. In t,hc presence of excess urease the liberation of ammonia exactly equals the removal of barbituric acid. The removal of the intermediate may be the result of the more effective action at this step by the same enzyme which carries out the first hydrolytic step or it may be due to the action of an additional enzyme or could even be a spontaneous reaction. The fate of 5-methyibarbituric acid is still unknown. As stated pre- viously, it is readily metabolized by t.he whole cell under aerobic con- ditions at the same rate as barbituric acid, but, after alumina grinding or 0. MYAISHI AND A. KORNBERG 731 sonic treatment, bacterial homogenatess were uniformly inactive. It may be assumed, therefore, that the enzyme is extremely labile or that some rather complex reaction is necessary to initiate this degradation. On the assumption that the initial attack on 5-methylbarbituric acid is non-oxidative, as in the case of barbiturase reaction, anaerobic experi- ments were conducted with 5-methylbarbituric acid and suspensions of thymine-grown cells. KO change in optical density was observed, even after prolonged incubation. However, control experiments with barbituric acid under similar conditions rcvcaled that there was no action on barbi- turic acid either. Thus, the removal of barbituric acid, which is carried out by cell-free enzymes at the same rate in the presence or absence of oxygen, does not occur under anaerobic conditions when intact cells are used. This could be due to an energy requirement for absorption of the substrate into the cell which is supplied under aerobic conditions. The ability of malonic acid to act as a competitive inhibitor of succinic acid in the succinic dehydrogenase reaction has created the common im- pression that malonic acid is an unnatural, toxic substance. This attitude can be seriously questioned in view of the current findings that maIonic acid is a product of uracil metabolism and indications that it is a readily metabolizable substrate. The utilization of malonic acid by microor- ganisms (24-26) as well as by mice (27) and rats (28) has been reported. The formation of malonic acid from oxalacetic acid by pig heart prepara- tions has also been observed (29). We have isolated several strains from soil by growth on media in which malonic acid served as a sole carbon source. The rather poor utilization of malonic acid by uracil-adapted cells, noted in the present work, may be related to an inability of malonic acid to enter the cell readily. Another possibility is that a labile compound which yields malonic acid in the iso- lation procedure is the true and rapidly metabolized intermediate. SUMMARY 1. The metabolism of pyrimidines by strains of Corynebacterium and Mycobaekrium isolated by enrichment culture has been studied with cell suspensions and enzymes purified from cell-free extracts. (a) 2. The resuIts suggest the following metabolic pathways: Cytosine + uracil -+ barbituric acid + urea 3- malonic acid I I (b) SMethylcytoaine + thymine 3 hmethylharbituric acid + NIFt + COS + Hz0 * Bacterial homogenate designates the turbid suspension containing broken cell material, membranes, soluble componenta, and a negligible number of intact cells. Cells treated with alumina or disrupted by sonic oscillation were centrifuged at low apeeda, just adequate to remove intact cells. + 732 PYRIMIDINE METABOLISM 3. The oxidation of uracil and thymine at carbon 6 to yield barbituric acid and 5-methylbarbituric acid, respectiveIy, was shown to be due to the action of ti single enzyme, "uracil-thymine oxidase." 4. An enzyme, "barbiturase," catalyzing the hydrolysis of barbituric acid to urea and malonic acid was partially purified and freed of urease. 5. The oxidation of 5-methylbarbituric acid has been observed only with intact cells and the detailed mechanism has not been established. BIBLIOGRAPHY 1. Hayaishi, O., and Kornberg, A., J. Am. Chem. Soc., 73, 2975 (1951). 2. Hayaishi, O., Federalion Proc., 11, 227 (1952). 3. Steudel, H., Z. physiol. Chem., 32, 285 (1901). 4. Cerecedo, 1,. It., J. Bd. Chem., 88, 695 (1930); 93, 269 (1931). 5. Behrend, H., and Roosen, O., Ann. Chem., 261, 235 (1889). 6. Heyroth, F. F., and Loofbowow, J. R., J. Am. Chem. Soc., 66, 1728 (1934). 7. Gerngross, O., Ber. chem. Ges., 38, 3394 (1909). 8. Nishikawtl, T., Mem. Ryojun Coll. Eng., 3, 87 (1931). 9, Stuckey, R. E., Quart. J. Phnrm. and Pharmacal,, 16, 370 (1942). 10. McIlwain, H., J. Gen. Microbiol., 2, 288 (1948). 11. Lowry, 0. H., Rosebrough, N. J., Farr, A. L., and Randall, R. J., J. Biol. Chem., 12. Conway, H. J., Micro-dinusion analysis and volumetric error, London, 3rd edi- 13. Umbreit, W. W., Burris, R. II., and Stauffer, J. F., Manometric techniques and 14. Cohn, W. E., Science, 109, 377 (1949). 15. Hahn, A,, and Lintzel, W., Z. UioE., 79, 179 (1923). 16. Iwatsuru, R., and Chikano, M., J. Hiochem., Japan, 1, 279 (1923). 17. Friedman, S., and Cots, J. S., Bad. Proc., 144 (1951). 18. Lineweaver, IE., and Burk, D., J. Am. Chem. Soc. , 66,658 (1934). 19. Meyer, II. J. L., Kachweia und Bcstimmung organischer Verbindungen, Berlin 20. Wang, T. Y., and Lampen, J. O., Federation Proc., 10, 267 (1951). 21. Wang, T. P., and Lampen, J. O., J. Biol. Chem., 194, 775, 785 (1952). 22. Batt, R. D. , and Woods, lj. D., Biochem. J., 49, p. lxx (1951). 23. Lara, A. J. S., J. Bad., in press. 24. den Dooren de Jong, L. E., Ijijdrage tot de konnie van het mincralisatieproces, 25. Challenger, F., Subramanism, V., and Walker, T. K., J. Chem. Soc., 200 (1927). 26. BurriB, R. H., and Wilson, P. W., Cold Spring Harbor Symposia Quanl. Biol., 7, 27. Lifson, N,, and Stolen, J. A., Proc. Soc. Ezp. Biol. and Med., 74, 451 (1950). 28. Lee, J. S., and Lifson, N., J. Biol. Chem., 193, 253 (1951). 29. Vennesland, B., and Evans, E. A,, Jr., J. Biol. Chem., 166, 783 (1944). 193, 265 (1951). tion (1950). tissue metabolism, Minneapolis, revised edition (1949). (1933). Kotterdam (1926). 349 (1939).