ENZYMATIC SYNTHESIS AND PROPERTIES OF 5-PHOSPHORIBOSYLPYROPHOSPHATE* BY ARTHUR KORNBERG, IRVING LIEBERMAN, AND ERNEST S. SIMMS (From the Department of Microbiology, Washington University School of Medicine, St. Louis, Missouri) (Received for publication, November 22, 1954) Investigations of the mechanism of pyrimidine nucleotide formation from orotic acid led to the isolation and characterization of an activated ribose phosphate derivative which serves as an intermediate in the over- all process. In earlier communications (1) we have described briefly the enzymatic synthesis of this new intermediate, identified as PRPPI (equa- tion (I)), and the occurrence of enzymes which condense pyrimidines and purines with PRPP to form the respective nucleotides (equation (2)). (1) OH r--l P!-0-P(OH)2 I1 HCOH Hq HCOH .P"I 0 II 0 ARPPP + H~OH I (\-ARP+ I HC HC Hr'l I H&OP(OH)t 0 I HICOP(OH)2 0 II II ATP + R5P + A51' + PRPP (2) Pyrimidine (or purine) + PRPP e pyrimidine (or purine) nucleotide 3. PP This report deals with the purification and properties of the enzyme which catalyzes the formation of PRPP and some of the properties of PRPP. Succeeding reports will consider the enzymatic synthesis of pyrimidine nucleotides (2), purine nucleotides (3), and nucleoside di- and triphos- phates (4). *This investigation was supported by grants from the National Institutes of Health, Public Health Service, and the National Science Foundation. 1 The abbreviations used are 5-phosphoribosylpyrophosphate, PRPP ; adenosine triphosphate, ATP or ARPPP; ribose-5-phosphate, R5P; adenosine-5'-phosphate, A5P or ARP ; inorganic pyrophosphate, PP; adenosine diphosphate, ADP; orotidine- 5'-phosphate, 05P; uridine-5'-phosphate, U5P; tris (hydroxymethyl)aminomethane, Tris. 389 390 5-PHOSPIIORIBOSYLPYROPHOSPHATE Methods Materials-The sodium salt of R5P was prepared from the barium salt (product of the Schwarz Laboratories, Inc.) by decomposition with sodium sulfate. A5P and ADP were products of the Sigma Chemical Company and ATP of the Pabst Brewing Company. Adenine and orotic acid were obtained from Dougherty Chemicals. Ribose-l-phosphate was prepared by phosphorolysis (5) of inosine (Kutritional Biochemicals Corporation) with liver nucleoside phosphorylase2 coupled to xanthine oxidase;2 it was purified by anion exchange chromatography (6). ATP32 (ATP labeled in the two terminal phosphate groups) was prepared by aerobic phosphoryla- tion of A5P with rat liver mitochondria3 (7) and purified by the ion ex- change chromatographic procedure of Cohn and Carter (8) and Kornberg and Pricer (9). 2-Deoxyribose-&phosphate was prepared from 2'-deoxyadenosine-5'- phosphate (California Foundation for Biochemical Research) as follows : 1.6 ml. of a 0.04 M solution of the nucleotide was acidified to pH 2, heated at 100" for 5 minutes, neutralized, adsorbed on a Dowex 1 column (2 per cent cross-linked, chloride form, 3 em. X 1 sq. em.), and eluted with 0.005 N HCl-0.01 M KC1. The fractions between 15 and 23 resin bed volumes contained 97 per cent of the adsorbed deoxypentose and an equivalent amount of organic phosphate. The acid lability of the compound (94 to 98 per cent hydrolysis after 30 minutes at 100" in 1 N HCI) agrees with published observations (10). Orotic acid (4 ,7-C14)4 was prepared from 1 ,4-C14-aspartic acid by a mod- ification of the method of n'yc and Mitchell (11). Labeled aspartic acid was prepared as follows: Ethylene dibromide (1.7 gm.) was allowed to react with C14-NaCN (10 mrnoles) by a microadaptations of a procedure of 'Vanino (12), producing 1 , 4-C14-succinic acid, which was then partially purified by treatment with a cation exchange resin (Dowex 50 (Hf)). The percolate was evaporated to dryness and the residue was dissolved in mater and neutralized. The succinate was quantitatively oxidized by suc- cinoxidase (13) to an equilibrium mixture of fumarate and malate. The latter mixture was then converted to L-aspartate in the presence of excess ammonia by an aspartase-fumarase preparation from Eschrichia coli (14). L-Aspartic acid was purified by chromatography on Dowex 50 (Hf), con- centrated in vacuo, and crystallized at pH 3.0 by the addition of alcohol. Acetone powders of liver were prepared as follows: The fresh tissue was 2 Kindly furnished by Dr. M. Friedkin. 3 Kindly furnished by Dr. F. E. Hunter. 4 The numbering of t.he pyrimidine ring follows recent usage by "Chemical ab- stracts;" orotic acid is 6-carboxyuracil. We are indebted to Dr. D. Shemin for these details. .4. KORNBERG, I. LIEBERMAN, AND E. S. SIMMS 391 homogenized in 5 to 10 volumes of acetone (chilled to - 10"). The residue, collected on a Buchner funnel, was homogenized again in cold acetone, collected, and dried in air. The powders were stored at - 10". Determinations-ADP and ATP were determined spectrophotometri- cally (15,16). A5P was determined with Schmidt's deaminase by Kalckar's method (17). PRPP was determined as described below (Assay B, Stage 11). Orthophosphate was determined by the method of Fiske and Sub- barow (18) , acid-labile phosphate was the orthophosphate liberated after hydrolysis for 10 minutes in 1 N HCl at loo", and total phosphate was obtained as orthophosphate after ashing with a sulfuric acid-nitric acid mixture. Reducing sugar was determined colorimetrically (I 9) with D-ribose as a standard. Pentose was determined by the Mejbaum pro- cedure (20) with a 40 minute heating period and A5P as a standard. Fro- tein was determined by the phenol method of Lowry et al. (21). Ion exchange chromatography was carried out at 2" with an automatic fraction collector on Dowex I (2 per cent cross-linked, 200 to 400 mesh, chloride form) columns according to Cohn (22). P32 was measured in solution in a dish under a Geiger-Muller tube. C14-containing samples were plated as thin layers on metal disks and measured in a gas flow counter. Self-ab- sorption corrections were applied where indicated. Enzyme Assays-The activity responsible for equation (1) can be de- termined in a variety of ways. Three methods which have provided sen- sitive and reliable assays for the enzyme will be described here. Assay A. P402 Evolution from Carboxyl-Labeled Orotate-This assay requires the addition, in excess, of 05P pyrophosphorylase and decarbox- ylase which carry out reactions (3) and (4), respectively. Under these (3 ) Orotate + PRPP & 05P + PP (4) 05P -+ U5P + Cot conditions the amount of C1402 liberated is an accurate measure of the amount of PRPP produced in the system. The incubation mixture (1.0 ml.) contained 0.02 ml. of ATP (0.04 M), 0.02 ml. of phosphopyruvate (0.1 M), 0.1 ml. of R5P (0.025 M), 0.05 ml. of 4,7-CI4-orotate (0.01 M, 1.4 X lo6 c.p.m. per pmole), 0.05 ml. of potassium phosphate buffer (1 M, pH 7.0), 0.03 ml. of MgC12 (0.1 M), 0.01 ml. of pyruvate phosphokinase (16), 0.1 ml. of yeast autolysate (Z), and less than 0.2 unit of enzyme. The mixture was incubated at 30" for 1 hour in an evacuated Thunberg tube containing 0.5 ml. of 0.2 N KOH (with a drop of 0.04 per cent thymol blue) in the bulb. To stop the reaction and acidify the reaction mixture, about 4 drops of an Hi304-thymol blue solution (2 parts of 50 per cent H8O4, 1 part of 0.04 per cent thymol blue) were cautiously admitted into the main space of the Thunberg tube through the side arm. The tube, 392 5-PHOSPHORIBOSYLPY ROPHOSPHATE still under vacuum, was incubated in a 65' oven for 30 minutes and then cooled. 0.1 ml. of the KOH solution in the bulb was removed to a plan- chet and dried and the radioactivity was measured. Under these con- ditions (with a self-absorption of about 50 per cent), 3500 c.p.m. represented total decarboxylation of the 0.5 pmole of orotate in the reaction mixture. Proportionality of enzyme addition to U402 release was obtained when 30 per cent or less of the orotate was decarboxylated. 1 unit of enzyme is defined as the amount causing the formation of 1 pmole of PRPP (as indicated by the release of 1 pmole of CI4O2), or the removal of 1 pmole of orotate (Assay B, below), or the removal of 1 pmole of ATP32 (Assay C, below), per hour, and specific activity is expressed as units per mg. of protein. Assay B. Orotate Removal Measured Spectrophotometrically-This assay is best carried out in two stages; the first involving the formation of PRPP and the second its quantitative utilization in the removal of orotic acid as in equation (3). Stage I-The incubation mixture (1.0 ml.) contained 0.03 ml. of ATP (0.04 M), 0.1 ml. of R5P (0.025 M), 0.02 ml. of potassium phosphate buffer (1 M, pH 7.4), 0.02 ml. of MgCl2 (0.1 M), 0.02 ml. of glutathione (0.5 M), 0.05 ml. of KF (1 M), and about 1 unit of enzyme. After incubation for 20 minutes at 36", the mixture was heated for 1 minute in a boiling water bath to inactivate the enzyme, immediately cooled, and centrifuged if necessary. Stage II-The incubation mixture (1.0 ml.) in a quartz cuvette con- tained 0.02 ml. of orotate (0.01 M), 0.02 ml. of Tris buffer (1 M, pH S.O), 0.02 ml. of MgCL (0.1 M), 0.2 ml. of 05P pyrophosphorylase (ethanol fraction) (2), and an aliquot from Stage I containing about 0.03 pmole of PRPP. The conversion of orotic acid to 05P or U5P results in a decrease in optical density at 295 mp; the molar absorption coefficient for the den- sity decrease of the latter conversion is 3950. With the indicated amount of 05P pyrophosphorylase fraction, the utilization of 0.03 pmole of PRPP is usually complete in 10 minutes when measured in theBeckman DU spectrophotometer at room temperature. Assay C. ATP32 Removal-This assay depends on transfer of the PS2- labeled terminal phosphate groups of ATP to a compound that is not adsorbable by Norit (23). The incubation mixture (0.4 ml.) contained 0.02 ml. of ATP32 (0.02 M, 3 X lo4 c.p.m. per pmole), 0.04 ml. of R5P (0.025 M), 0.02 ml. of potassium phosphate buffer (I M, pH 7.4), 0.01 ml. of MgCl2 (0.1 M), 0.02 ml. of glutathione (0.5 M), 0.04 ml. of KF (1 M), and about 0.2 unit of enzyme. After incubation for 30 minutes at 36", 0.8 ml. of 3.5 per cent perchloric acid and then 0.2 ml. of a 10 per cent suspension of Norit A were added to the reaction mixture. After 2 to 5 minutes the A. KORNBERG, I. LIEBERMAN, AND E. S. SIMMS 393 Step Units per ml.' 1. Acetone powder extract. . . . . 3. Aluminum hydroxide gel eluate . . . . . . . . . . . . . . . . . . . . 18.0 16.4 8.1 2. Low pH precipitation,. . . . . . mixture was filtered and 0.5 ml. of the filtrate was pipetted into a dish for radioactivity measurement. With crude enzyme fractions, an incu- bation mixture lacking R5P served as a control for interfering activities such as that of ATPase. The use of fluoride provides a means of reduc- ing ATPase activity. Total units* Protein mg. per ml. 612 19.6 558 3 .oo 275 0.33 Results Purifiation of Enzyme-All operations were carried out at 0-3'. 4 gm. of pigeon liver acetone powder were suspended in 40 ml. of Tris buffer (0.02 M, pH 8.0) and stirred occasionally during a 10 minute interval. The residue was removed by centrifugation in a Serval1 (SS-1) centrifuge at 8000 X g for 5 minutes. The supernatant fluid may be either clear or opalescent (acetone powder extract, Table I). TABLE I Purifiation of Enzyme l~pecific activity units per mg. protein 0.9 5.5 24.5 * Assay R was used in this run. Essentially similar results were obtained with the other assays. To 34 ml. of extract were added 51 ml. of water and then, slowly, 17 ml. of potassium acetate buffer (1 M, pH 5.4). After 5 minutes, the precipi- tate was collected by centrifugation (as above). It was dissolved in 7.0 ml. of Tris buffer (0.1 M, pH 8.0) and diluted to 30 ml. with water. The pH was adjusted to 6.8 with 0.1 N HC1 (about 1.3 ml. required). The volume was made up to 34 ml. (low pH fraction). For the last step enough aluminum hydroxide gel Cr was used to ad- sorb 85 to 90 per cent of the enzyme. 3.75 ml. of aged gel (24) (15 mg. of dry weight per ml.) were centrifuged and the supernatant fluid was discarded. 30 ml. of isoelectric fraction were thoroughly mixed with the gel, left in ice for 5 minutes, and then centrifuged. The supernatant fluid was discarded and the gel precipitate was washed with 30 ml. of potassium phosphate buffer (0.05 M, pH 6.85). The enzyme was then eluted from the gel with 30 ml. of potassium phosphate buffer (0.10 M, pH 6.85) (alu- minum hydroxide gel eluate). The acetone powders were stored at -10" and showed no significant 3 94 5-PIIOSPHORIBOSYLPYROPHOSYHATE ATP. ........... PRPP.. ......... A5P.. ........... ............ 64 ........... L( ............. CC loss in activity over a 2 month period. The Cy gel eluate retained 65 to 70 per cent of its activity after storage for 2 months at -10" provided glutathione (0.05 M) was present; in the absence of glutathione less than 5 per cent of the activity remained. Distribution of Enzyme-The enzyme appears to be present also in chicken and mammalian livers and in bacteria. Extracts of acetone pow- ders of chicken, mouse, and rat livers, and of an 18 hour culture of E. coli B cells (ground with alumina (25)) had the following specific activities (Assay B): chicken, 0.29; mouse, 0.04; rat, 0.01; and E. coli, 0.34. An Control Experimental Control IExperiment 13.1 12.5 12.5 1.6 11.1 1.8 0.0 0.0 0.0 10.9 10 .o 0.0 0.0 0 .o 9.8 0 .o 10.2 - TABLE I1 Balance Study of PRPP Synthesis The experimental incubation mixture (10.0 ml.) was the same as that described for Assay 13, Stage I (see the text), except that ATP32 (2.2 X 104 c.p.m. per pmole) was used. The control mixture lacked R5P. 2.0 ml. of the aluminum hydroxide gel eluate fraction containing 0.72 mg. of protein were used. Incubation was at 35" for 60 minutes. The figures in bold-faced type were obtained by chromatography of 8.0 ml. of the reaction mixtures on Dowex 1 columns (6.5 cm. X 1.0 sq. cm.) ; the eluent was 0.0025 N HC1-0.1 M KCl. The other values were obtained by direct enzymatic assay of aliquots of the incubation mixtures. The results are expressed in micro- moles. Control -0.6 I I I 0 min. 60 min. A Experimental -10.9 0 .o 0.0 +10.9 +9.8 autolysate of dried brewers' yeast and a sonic extract of bakers' yeast were inactive. Balance Study of Reaction-The stoichiometry of the reaction of ATP with R5P to yield A5P and PRPP (equation (1)) is shown in Table 11. In the control sample, lacking R5P, ATP was not removed to any significant extent and the production of A5P and PRPP was not detectable. In the presence of R5P, almost all the added ATP disappeared and was matched by the accumulation of equivalent molar amounts of A5P and PRPP. Values obtained both by enzymatic and ion exchange chromatographic analyses were in good agreement. No ADP was detectable after either 30 or 60 minutes incubation when 65 and 87 per cent, respectively, of the ATP had been consumed. That orthophosphate was not released (from A. KORNBERG, I. LIEBEHMAN, AND E. S. SIMMS 395 the ATP32 in this reaction) was evident from the absence of any radio- activity in that fraction of the ion exchange chromatogram. Isolation of PRPP-PRPP mas isolated by ion exchange chromatog- raphy as a symmetrical zone, appearing after 25 to 30 resin bed volumes of effluent, and was contained within the next 10 to 15 volumes. The early fractions were contaminated by ATP. Analyses of eight fractions selected TABLE I11 Chromatography of PRPP 8.5 ml. of experimental and control incubation mixtures were chromatographed as described (see Table 11). Fraction No.' 30 32 34 36 38 40 42 44 Pentoset vmole per ml. 0.051 0.100 0.107 0.109 0. IO7 0 * 101 0.081 0.060 Molar ratios (referred to pentose) PRPP 1 .oo 0.84 1.03 1 .OO 0.95 0.94 0.84 Acid-labile Pi 2.00 1.98 2.01 2.01 2.03 2.10 2.15 Total P 3.05 2.96 3.00 3.15 2.95 3.15 3.05 Specific ac tivi tys .P.m. per pmolc 20,300 22,500 21,300 22,900 23,200 23,000 23,200 22,600 Average molar ratio 1 1.00 I 0.94 I 2.04 I 3.04 1 * Approximately 5.0 mI. per fraction, corresponding to about 1 resin bed volume. t These values are corrected for any ATP present as determined by the optical density at 260 mp. Estimations of orthophosphate by the method of Lowry and Lopez (26) indi- cated 0.16 pmole or less per pmole of pentose in each fraction. With the Fiske and Subbarow procedure (18), 25 and 40 per cent of the acid-labile phosphate appeared as orthophosphate after color development (at 30") for 10 and 30 minutes, respec- tively. 5 The specific activities for the ATP3* at zero time and that isoIated from the con- trol sample were 22,700 and 21,400, respectively. from the middle of the PRPP zone and containing approximately 80 per cent of the PRPP contained pentose, enzymatically active PRPP (see Assay B, Stage 11, equation (3)), acid-labile P, and total P in molar ratios of 1.00:0.94:2.04:3.04 (Table 111). The average of the specific radio- activities of these fractions was 22.4 X lo3 as compared with values of 22.7 X lo3 and 21.4 X lo3, respectively, for the ATP32 at zero time and that isolated from the cmtrol sample. A barium salt of PRPP was prepared as follows: ATP32 and the com- ponents described for Assay B, Stage I on a 100 ml. scale, were incubated 396 5-PH0SPfIORIBOSYLPYROPHOSPHATE for 75 minutes. After 45, 60, and 75 minutes, PRPP synthesis was esti- mated to be 103, 92, and 100 pmoles, respectively. The mixture was chromatographed on a 6 em. X 3.2 sq. em. The fractions between 29 and 46 resin bed volumes were neutralized immediately upon their collection and pooled (366 ml., including washings). 7.5 ml. of glycine buffer (1 M, pH 8.5) and 3 ml. of saturated barium acetate were added and the pH was adjusted to 8.5. Cold ethanol (2 volumes) was added and after 30 minutes at 0" the precipitate was collected by centrifugation, washed with ethanol, and dried in a vacuum desiccator. The salt (1 18 mg.) was suspended in 4 ml. of water, dissolved by the addition of acid (about 0.3 ml. of 0.5 N HC1 lowered the pH to about 2),' and decomposed with sulfate (0.3 ml. of 1 M Na2S04) and the supernatant solution was neutral- ized. The reagents were cold and the operations were rapid. 45 prnoles of PRPP were recovered. Extraction of the precipitate with dilute acid yielded 2.8 pmoles more. The recovery of PRPP from the incubation mixture was thus approximately 50 per cent. From 20 to 30 per cent of the radioactivity in the barium salt did not appear in solution. Based on a molecular weight for the barium salt of 729 gm. and without any cor- rections, the purity of the salt was only 30 per cent. Reprecipitation of the barium salt from Cot-free solutions resulted in considerable purification. A crude barium salt (65 mg.) was dissolved in dilute acid and immediately neutralized. The precipitate which formed was washed with 50 per cent ethanol and dried. The salt (24 mg.), on decomposition with sulfate, yielded 18.3 pmoles of PRPP. Assuming that 25 per cent of the PRPP in the barium salt fails to dissolve, the purity of this salt was 75 per cent. In some purified barium salts, ATP (or an ATP-like substance) represented 10 per cent or more of the impurity. Removal of the adenine nucleotide with Norit either was incomplete or led to large losses of PRPP. The major impurities in most cases appeared to be PP and R5P, the decomposition products of PRPP. A preparation free of ultraviolet-absorbing components ( < 1 .O density unit at 260 mp per 10 pmoles of PRPP) was obtained by excluding the early fractions of the chromatogram. The molar ratios of inorganic P, acid-labile I?, total P, and PRPP to pentose were 0.09, 2.04, 3.02, and 0.70, respectively. Assay with inorganic pyrophosphatase indicated the presence of 5.6 pmoles of PP per 10 pmoles of PRPP. Assuming the presence of an amount of R5P equivalent to that of PP, these two impurities would adequately ac- count for the analytic data. 8 As compared with the smaller scale run (Table 11), 10 times as much reaction mixture waa chromatographed on a column only 3 times as large; the separation of PRPP and adenine nucleotides appeared to be less effective. 7 A large evolution of gas indicated the presence of BaCOI as an impurity. A. HORNBERG, I. LIEBERMAN, AND E. S. SIMMS 397 Experiment No. Time of heating -A PRPP min. prnoles 10 0.94 40 1.45 1 2 30 1.13 Neutral solutions of PRPP have been stored at - 15' for 3 months with- out detectable destruction. At low pH or at elevated temperatures (see below), PRPP decomposition is rapid. +A reducing sugar +A PP pmoles pmols 0.96 1.73 1.07 0.09* M inu tes FIG. 1. Decomposition of PRPP at several pH levels. PRPP at a concentration of 0.4 pmole per ml. was incubated in 0.02 M buffer at 65". The buffers were as fol- lows: formate, pH 3.1, acetate, pH 4.0, glycylglycine, pH 6.7, glycine, pH 9.0, Structure and Decomposition of PRPP-Heating of PRPP at 65' at pH 4.0 results in a loss of enzymatically active compound (see Assay B, Stage 11, equation (3)) and the appearance in equivalent amounts of reducing sugar and inorganic pyrophosphate (Table IV) . The hydrolysis of PRPP at 65' at several pH levels is shown in Fig. 1. At pH 3.1, hydrolysis was virtually complete in 10 minutes, while tat neu- 398 5-PHOSPHORIBOSYLPYROPHOSPHATE tral pH only 50 per cent was destroyed after 20 minutes. The rate of PRPP decomposition was catalyzed by Mg++ and this effect was overcome by agents which bind Mg++ (Table V). The enzymatic decomposition of PRPP has been demonstrated in the reversible pyrophosphoroly sis of pyrimidine and purine nucleotides (equa- tion (2)) (2, 3). Nucleotide pyrophosphatase (27) and inorganic pyro- phosphatase (28) failed to split (less than 0.005 pmole) PRPP; under comparable conditions, over 25 and 100 pmoles of diphosphopyridine TABLE V Mu++ Activation of PRPP Decomposition 1 ml. of a 1 X 10-4 M neutral solution of PRPP (in a 12 X 100 mm. glass tube) was heated for 1 minute in a briskIy boiling water bath and then immediately placed in ice water. PRPP was assayed as described under "Methods." The concentra- tions of the indicated additions were MgCl2 0.002 M, potassium phosphate buffer, pH 7.4, 0.02 M, sodium pyrophosphate, pH 8.0,0.02 M, Versenate, pH 7.0, 0.005 M. Ex- periment I1 differed from Experiment, I only in that Tris-maleate buffer, pH 7.4, 0.02 M, was included. 1 min. hydrolysis (100") Experiment I I Experiment I1 Addition None. .......................... Phosphate. ..................... Mg++ + phosphate?. ........... 'I + pyrophosphate. ........ + Versenate .............. Mg++*. ......................... per ccnf 55 78 29 18 gcr cent 50 80 22 10 17 ~ ~~~ * Heating for 60 minutes at 36" results in a 51 per cent loss of PRPP. t Heating for 2 minutes at 100" results in a 46 per cent Ioss of PRPP. nucleotide and PP, respectively, were hydrolyzed by these enzyme prepara- tions. Influence of ATP and R5P Concentration on Reaction Rate and Specifi'ty of ATP and R5P for Enzyme-The influence of the concentration of ATP and R5P on the rate of reaction is shown in Fig. 2. No reaction was detectable in the absence of ATP and R5P; the levels at which the rates were half maximal were 4 X le4 M and 6 X le4 M for ATP and R5P, respectively. ATP at higher concentrations (2 X M) appeared to be inhibitory. ADP (1 X 10-3 M), inosine triphosphate (0.8 X lW3 M), and uridine triphosphate (0.4 X lW3 M) either failed t.0 replace ATP or showed less than 2 per cent of the activity of ATP. Ribose-1-phosphate (0.6 X 10-3 M), A. KORNBERG, I. LIEBERMAN, AND E. S. SIMMS 399 2-deoxyribose-5-phosphate (2.5 X le3 M), and glucose-&phosphate (2.5 X 10-3 M) appeared totally inactive in replacing R5P as determined by Assays 1 1 I 4 I e - Concentration, M FIQ. 2. Influence of ATP and R5P concentration on the rate of PRPP synthesis. The incubation conditions were as described in Assay B, Stage I, except for the vary- ing concentrations of ATP and R5P indicated in the graph. 0.04 ml. of the alumina gel eluate fraction containing about 0.3 unit of enzyme was used. The reaction rate is expressed as units of enzyme activity. FIG. 3. Influence of the pH on the rate of PRPP synthesis. The incubation con- ditions were as described in Assay B, Stage I, except for adjustment of the phosphate buffer to the values indicated in the graph. 0.05 ml. of the alumina gel eluate frac- tion containing about 0.4 unit of enzyme was used. The reaction rate is expressed as units of enzyme activity. B and C; the action of R5P (2.5 X ence of an equimolar concentration of 2-deoxyribose-5-phosphate. M) was not inhibited by the pres- Efects of Mg+ and pH on Reaction-Mg++ is essential for the enzyme 400 5-PHOSPHORIBOSYLPYROPHOSPHATE activity. The rate was half maximal at a concentration of 0.8 X le3 M and maximal at 3 X le3 M. The optimal pH for the reaction under standard assay conditions was found to be near 7.5 (Fig. 3). Fluoride did not inhibit the enzyme even at a concentration of 0.1 M. In its absence no reaction was observed. DISCUSSION The proposed structure of PRPP is derived from several lines of evidence: (I) utilization of 1 mole of ATP in its synthesis, with the appearance of 1 mole of A5P, but without liberation of ortho P, (2) specificity of R5P and ATP in its synthesis, (3) isolation of a compound, eluted as a sym- metrical peak in an ion exchange chromatogram, containing 1 mole of pentose and 3 moles of P of which 2 are acid-labile, (4) decomposition of the compound by very weak acid hydrolysis to liberate one reducing group and 1 mole of inorganic pyrophosphate, and (5) the participation of this compound in the reversible pyrophosphorolysis of pyrimidine and purine nucleoside-5'-phosphates. In studies on the mechanism of incorporation of adenine into A5P, a 5-phosphoribokinase (29) was described in pigeon liver extracts which catalyzes the reaction ATP + 2R5P ---f A5P + 2 ribose-l,5-diphosphate The report contained no evidence for the stoichiometry of the reaction or for the chemical structure of the ribose diphosphate. The similarity of the substrate, product, and enzyme source to those used in our studies suggests that a more thorough inquiry into the ribose diphosphate-forming system is indicated. The enzymatic synthesis of PRPP may represent a two-stage transfer of phosphate from ATP to R5P or, what appears more likely, a direct pyrophosphorylation of the substrate. The latter mechanism, if correct, would be unique, since previously described reactions such as the pyro- phosphorolysis of coenzymes (30) are reversible systems in which the adenyl rather than the phosphate group is transferred. Maas and Lipmann (31) recently proposed, on the basis of isotopic evidence, that a direct pyro- phosphorylation of the enzyme involved in pantothenate formation is a step in the over-all scheme. A more direct demonstration of this mechan- ism is needed. The metabolic functions of PRPP form the basis for the succeeding re- ports on a pathway of synthesis of pyrimidine and purine ribonucleotides. It is possible that reaction schemes leading to the formation of other N- ribosyl compounds, such as the ribotide of acyclic purine precursors, and A. KORNBERG, I. LIEBERMAN, AND E. S. SIMMS 40 1 of nicotinamide may also utilize PRPP as an intermediate.* The question as to whether a 2-deoxy analogue of PRPP may provide a means for con- densing with pyrimidine and purine bases to form deoxyribonucleotides is also worthy of investigation. SUMMARY 1. Enzymatic studies of pyrimidine and purine nucleotide synthesis have led to the discovery of a new intermediate, 5-phosphoribosylpyrophosphate (PRPP) . 2. An enzyme has been partially purified from pigeon liver which cata- lyzes the formation of PRPP from ribose-5-phosphate (R5P) and ATP. ATP + R5P + adenosine-5'-phosphate + PRPP This enzyme appears to be present also in mammalian liver and in bacteria. 3. ATP and R5P are specific in this reaction; Mg++ is also required. 4. PRPP has been isolated and characterized, and some of its properties have been described. 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. BIBLIOGRAPHY Kornberg, A., Lieberman, I., and Simms, E. S., J. Am. Chem. SOC., 76,2027 (1954). Lieberman, I., Kornberg, A,, and Simms, E. S., J. Am. Chem. SOC., 76, 2844 (1954). Lieberman, I., Kornberg, A., and Simms, E. S., J. Bid. Chem., 216, 403 (1955). Kornberg, A,, Lieberman, I., and Simms, E. S., J. Biol. Chem., 216, 417 (1955). Lieberman, I., Kornberg, A., and Simms, E. S., J. Bid. Chenz., 216, 429 (1955). Kalckar, H. M., J. Bid. Chem., 167, 477 (1947). Rowen, J. W., and Kornberg, A., J. Bid. Chem., 193,497 (1951). Kielley, W. W., and Kielley, R. K., J. BioZ. Chem., 191, 485 (1951). Cohn, W. E., and Carter, C. E., J. Am. Chem. Soc., 72,4273 (1950). Kornberg, A., and Pricer, W. E., Jr., J. Bid. Chem., 191, 535 (1951). Racker, E., J. Biol. Chem., 196, 347 (1952). Nyc, J. F., and Mitchell, H. K., J. Am. Chem. SOC., 69, 1382 (1947). Vanino, L., Handbuch der praparativen Chemie, Stuttgart, 2,263 (1937). Stotz, E., and Hastings, A. B., J. Bid. Chem., 118,479 (1937). Racker, E., Biochim. et biophys. acta, 4, 211 (1950). Kornberg, A., J. Bid. Chem., 182, 779 (1950). Kornberg, A., and Pricer, W. E., Jr., J. Bid. Chem., 193, 481 (1051). Kalckar, H. M., J. BioZ. Chem., 167, 445 (1947). Fiske, C. H., and Subbarow, Y., J. Bid. Chem., 66, 375 (1925). Somogyi, hl., J. Biol. Chem., 196,19 (1952). 8 A sample of PRPP has been tested by Buchanan and coworkers and found to be active in the formation of ribotides of acyclic purine precursors. These workers have also isolated a pentose phosphate ester which is involved in purine nucleotide syn- thesis and has been identified as PRPP (personal communication from Dr. J. M. Buchanan). 402 5-PHOSPI-IORIBOSYLPYROPHOSPHATE 20. Mejbaum, W., 2. physiol. Chem., 268,117 (1939). 21. Lowry, 0. H., Rosebrough, N. J., Farr, A. L., andRandall, R. J., J. Biol. Chem., 22. Cohn, W. E., J. Am. Chem. Soc., 72, 1471 (1950). 23. Crane, R. K., and Lipmann, F., J. Biol. Chem., 201, 235 (1953). 21. Willstiitter, R., and Kraut, H., Ber. chem. Ges., 56, 1117 (1923). 25. Mcflwain, €I., J. Gen. Microbial., 2, 288 (1948). 26. Lowry, 0. H., and Lopez, J. A., J. Biol. Chem., 162, 421 (1946). 27. Kornberg, A., and Pricer, W. E., Jr., J. Biol. Chem., 182, 763 (1950). 28. Heppel, L. A., and Hilmoe, R. J., J. BioE. Chern., 192, 87 (1951). 29. Saffran, M., and Scarano, E., Nature, 172, 949 (1953). Scarano, E., Nature, 172, 30. Kornberg, A., in McElroy, W. D., and Glass, B., Phosphorus metabolism, Balti- 31. Maas, W. K., cited by Lipmann, F., in McElroy, W. D., and Glass, B., The 193, 265 (1951). 951 (1953). more, 1, 392 (1951). mechanism of enzyme action, Baltimore, 599 (1954).