DEOXYRIBONUCLEIC ACID-DIRECTED SYNTHESIS OF RIBONUCLEIC ACID BY AN ENZYME FROM ESCHERICHIA COLI* By MicHarL CHAMBERLIN{ AND PauL BERG DEPARTMENT OF BIOCHEMISTRY, STANFORD UNIVERSITY SCHOOL OF MEDICINE Communicated by Arthur Kornberg, November 17, 1961 Protein structure is under genetic control ;!—? yet the precise mechanism by which DNAT influences the formation of specific amino acid sequences in proteins is unknown. Several years ago, it was discovered that infection of Escherichia coli with certain virulent bacteriophages induces the formation of an RNA fraction possessing both a high metabolic turnover rate and a base composition correspond- ing to the DNA of the infecting virus.4~* The existence of an analogous RNA component in noninfected cells has also been demonstrated; in this instance, however, the base composition of the RNA resembles that of the cellular DNA. 8 These observations focused attention on the possible role of this type of RNA in protein synthesis, and some of the evidence consistent with this view has recently been summarized.® Until recently there was no known enzymatic mechanism for a DNA-directed synthesis of RNA. Polynucleotide phosphorylase " although it catalyzes the synthesis of polyribonucleotides, does not by itself provide a mechanism for the formation of RNA with a specific sequence of nucleotides. The one instance in which a unique sequence of nucleotides is produced involves the limited addition of nucleotides exclusively to the end of pre-existing polynucleotide chains.!2—"4 Our efforts were therefore directed toward examining alternate mechanisms for RNA synthesis, and in particular one in which DNA might dictate the nucleotide sequence of the RNA. In the present paper, we wish to report the isolation and some properties of an RNA polymerase from £. cola which, in the presence of DNA and the four naturally occurring ribonucleoside triphosphates, produces RNA with a base composition complementary to that of the DNA. Within the last year, several laboratories have reported similar findings with enzyme preparations from bacterial as well as from plant and animal sources.-24 In the following paper, the effect of enzymatically synthesized RNA on the rate and extent of amino acid incorporation into protein by £. colz ribosomes in the presence of a soluble protein fraction is described. Experimental Procedure-——Materials: Unlabeled ribonucleoside di- and triphosphates were purchased from the Sigma Biochemical Corporation and the California Corporation for Biochem- ical Research. 8-C!4labeled ATP was purchased from the Schwartz Biochemical Company; the other, uniformly labeled, C14 ribonucleoside triphosphates were prepared enzymatically from the corresponding monophosphate derivatives™ isolated from the RNA of Chromatium grown on CO. as sole carbon source. CTP labeled with P®? in the ester phosphate was obtained by enzymatic phosphorylation of CMP* prepared according to Hurwitz.?7, The deoxyribonucleoside triphosphates were obtained by the procedure of Lehman et al.*5 Calf thymus and salmon sperm DNA were isolated by the method of Kay eé al.28 DNA from Aerobacter aerogenes, Mycobacterium phlei, and bacteriophages T2, T5, T6 was prepared as de- scribed previously.2? DNA from Adg phage was prepared as reported elsewhere.** Unlabeled and P** labeled DNA from E. coli were prepared as previously described.*!_ d-AT and d-GC poly- mers were prepared according to Schachman et al.*? and Radding et al.,33 respectively. Trans- forming DNA from Bacillus subtilis*4 was a gift from E. W. Nester, and DNA from phage @X 81 82 BIOCHEMISTRY: CHAMBERLIN AND BERG Proc. N. A. 8. 174 4° was generously supplied by R. L. Sinsheimer. Double-stranded @X 174 DNA was syn- thesized using FE. coli DNA polymerase” with single-stranded @X 174 DNA as primer.*% 36 In this reaction, 2.7 times more DNA was synthesized than had been added as primer. RNA from tobacco mosaic virus was obtained from H. Fraenkel-Conrat, and ribosomal and amino acid-— acceptor RNA were isolated from E. colt according to Ofengand et al.**. 37 Nucleic acid concen- trations are given as mumoles of nucleotide phosphorus per ml. Glass beads, ‘‘Superbrite 100,’ obtained from the Minnesota Mining and Manufacturing Company, were washed as previously described.** Streptomycin sulfate was a gift from Merck and Company, and protamine sulfate was purchased from Eli Lilly Company. DEAE-cellulose was purchased from Brown and Company. Crystalline pancreatic RNase and pancreatic DNase were products of the Worthington Biochemical Co. Assays: The activities of E. coli-DNA polymerase,?*-deoxyribonuclease*® and -DNA dies- terase,#! were determined as previously described and ribonuclease activity was measured by the disappearance of amino acid-acceptor RNA activity.2* Polynucleotide phosphorylase was measured by Pj? exchange with ADP as reported by Littauer and Kornberg.!! Protein was determined by the method of Lowry et al.*9 The standard assay for RNA polymerase measures the conversion of either C4 or P*? from the labeled ribonucleoside triphosphates into an acid-insoluble form. Enzyme dilutions were made with a solution containing 0.01 M Tris buffer, pH 7.9, 0.01 44 MgCh, 0.01 M 8-mercaptoethanol, 5 X 10-5 M EDTA, and 1 mg per ml of crystalline bovine serum albumin. The reaction mixture (0.25 ml) contained: 10 wmoles of Tris buffer, pH 7.9, 0.25 umole of MnCh, 1.0 umole of MgCl, 100 mumoles each of ATP, CTP, GTP, and UTP, 250 mumoles of salmon sperm DNA, 3.0 »moles of 6-mercaptoethanol, and 10 to 80 units of enzyme. One of the nucleo- side triphosphates was labeled with approximately 300 to 600 cpm per mymole. After incubation at 37° for 10 min, the reaction mixture was chilled in ice, and 1.2 mg of serum albumin (0.03 ml) was added, followed by 3 ml of cold 3.5% perchloric acid (PCA). The precipitate was dispersed, centrifuged for 5 min at 15,000 X g, and washed twice with 3.0 ml portions of cold PCA. The residue was suspended in 0.5 ml of 2 N ammonium hydroxide, transferred to an aluminum planchet, and after drying, counted in a windowless gas-flow counter. One unit of enzyme activity corresponds to an incorporation of 1 mymole of CMP* per hr under the conditions described above. The assay was proportional to the amount of enzyme added up to at least 80 units; thus 6.3, 12.5, and 25 yg of Fraction 4 enzyme incorporated 2.6, 5.1, and 10.0 mumoles of CMP22. The rate of the reaction remained constant for approximately 20 min, and then decreased after this time. Since the radioactivity incorporated represents only one of the four nucleotides, the observed incorporation must be multiplied by a factor ranging from 3 to 5 for an estimate of the total amount of RNA synthesized. The exact factor depends on the composition of the DNA primer used. Results —Purification of RNA polymerase: (1) Cells: E. coli B was grown in continuous exponential phase culture*® with a glucose-mineral salts medium.‘ Cells stored at —20° showed no loss of activity for over six months. The purifica- tion procedure and the results of a typical preparation are summarized in Table 1. TABLE 1 PuRIFICATION OF RNA POLYMERASE FROM £&, colt Volume Specific activity Total activity Fraction (ml) (unita/mg) (units) 1. Initial extract 260 40 370,000 2. Protamine eluate 37 1,600 205 , 000 3. Ammonium sulfate 5 2,500 200 , 000 4. Peak DEAE fraction 2 6,100 153,000 Unless noted otherwise, all operations were carried out at 4° and all centrifugations were at 30,000 X g for 15 min in an International HR-1 Centrifuge. (2) Extract: Frozen cells (140 gm) were mixed in a Waring Blendor with 420 gm of glass beads and 150 ml of a solution (buffer A) containing 0.01 M Tris buffer, pH Vou. 48, 1962 BIOCHEMISTRY: CHAMBERLIN AND BERG 83 7.9, 0.01 M MgCh, and 0.0001 M EDTA. After disruption of the cells at high speed for 15 min (maximum temperature 10°), a further 150 ml of buffer A was added and the glass beads were allowed to settle. The supernatant fluid was then decanted and the residue was washed with 75 ml of buffer A. The combined supernatant fluid and wash was centrifuged for 30 min and the resulting super- natant fluid collected (Fraction 1). (3) Streptomycin-protamine fractionation: Fraction 1 was centrifuged in the Spinco Model L preparative ultracentrifuge for 4 hr at 30,000 rpm in the No. 30 rotor. The protein concentration in the supernatant fluid was adjusted to about 12 mg per ml with buffer A, and 8-mercaptoethanol was added to a final concentration of 0.01 M. To 350 ml of the diluted supernatant solution was added 17.5 ml of a 10% (w/v) solution of Streptomycin sulfate with stirring. After 15 min, the solution was centrifuged, and to 350 ml of the supernatant fluid was added 14.0 ml of a 1% (w/v) solution of protamine sulfate. The precipitate, collected by centrifugation, was washed by suspension in 175 ml of buffer A containing 0.01 M $-mercapto- ethanol. The washed precipitate was then suspended in 35 ml of buffer A con- taining 0.01 M mercaptoethanol and 0.10 M ammonium sulfate, centrifuged for 30 min, and the supernatant fluid was collected (Fraction 2). (4) Ammonium sulfate fractionation: To 37 ml of Fraction 2 was added 15.8 ml of ammonium sulfate solution (saturated at 25° and adjusted to pH 7 with ammonium hydroxide). The mixture was stirred for 15 min, and the precipitate was removed by centrifugation. To the supernatant liquid was added an addi- tional 16.2 ml of the saturated ammonium sulfate, and after 15 min the precipitate was collected by centrifugation for 30 min and dissolved in buffer B (0.002 M KPOQ,, pH 8.4, 0.01 M MgCh,, 0.01 M 8-mercaptoethanol, and 0.0001 M@ EDTA) to a final volume of 5.0 ml (Fraction 3). (5) Adsorption and elution from DEAE-cellulose: Fraction 3 was diluted to a protein concentration of about 3 mg per ml with buffer B and passed onto a DEAE- cellulose column (10 em X 1 cm?, washed with 150 ml of buffer B just prior to use) at a rate of about 0.5 ml per min. The column was washed with 10 ml of buffer B and then with enough of the same buffer containing 0.16 Mf KCl to reduce the absorbency of the effluent at 280 my to less than 0.05. The enzyme was eluted from the column with buffer B containing 0.23 M KCl The activity appears within the first five ml of the latter eluant (Fraction 4). (6) Properties of the purified enzyme: The specific activity of enzyme Fraction 4 was from 140 to 170 times greater than that of the initial extract. The puri- fication as described here has been quite reproducible, with specific activities in the final fraction ranging from 5,500 to 6,100. The enzyme preparation (Fraction 4) has a ratio of absorbencies at 280 and 260 mu of 1.5. Fraction 4, stored at 0 to 2°, retains more than 90 per cent of its activity for up to two weeks and 40 to 60 per cent of the original activity after one month. En- zyme Fractions 1 through 3 are unstable, losing up to 30 per cent of their activity on overnight storage under a variety of conditions. Because of the marked in- stability of these earlier fractions, it is advisable to carry out the purification without stopping at intermediate stages. (7) Contaminating enzymatic activities: Aliquots (100 ug) of Fraction 4 were assayed for contaminating enzymatic activities. This amount cf enzyme cata- 84 BIOCHEMISTRY: CHAMBERLIN AND BERG Proc. N. A. 5. lyzed an initial rate of incorporation of 2,000 mumoles of nucleotide per hr. No detectable DNA polymerase was found (< 0.6 mumole DNA per hr). DNase activity was barely detectable under conditions optimal for RNA polymerase. With either heated or unheated P?2 DNA as substrate, no more than 0.13 mumole of acid-soluble P3? was released during the course of a 30-min incubation. There was only slight RNase activity associated with Fraction 4. When 100 ug of the purified enzyme were incubated with 4 umoles of purified acceptor RNA for 1 hr, there was no detectable inactivation of leucine-acceptor activity. Under similar conditions, 1 mg of enzyme produced a 30 per cent decrease in leucine-acceptor activity. With conditions optimal for RNA polymerase, sufficient polynucleotide phosphorylase activity was present to catalyze the exchange of 6.7 mumoles of P;®? into ADP per hour. Requirements for the RNA polymerase reaction: With the purified enzyme, RNA synthesis was dependent on the addition of DNA, a divalent cation, and the four ribonucleoside triphosphates (Table 2). In a later section, we shall describe a TABLE 2 TABLE 3 REQUIREMENTS FOR RNA SYNTHESIS THE REQUIREMENT FOR RIBONUCLEOSIDE Incorporation TRIPHOSPHATES IN RNA SyNTHESIS of CM P22 Incorporation Components (myumoles) of CM-P22 Complete system 7.3 Components {mpzmoles) minus Mn++ 4.3 Complete system 4.6 minus Mgtt 5.6 minus ATP 0.08 minus Mn++ and Mgt* <0.03 minus UTP <0.03 minus DNA <0.03 minus GTP <0.03 minus ATP, GTP, UTP 0.09 ATP, UTP, GTP replaced by minus enzyme <0.03 dATP, dTTP, dGTP 0.05 The standard system and assay procedure were ATP, UTP, GTP replaced by . used with 7.4 wg of Fraction 4 protein in each tube, ex- ADP, UDP, GDP 0.29 cept that MgCl: was omitted from the enzyme diluent. The standard aystem and assay procedure were used except that 250 mumoles of calf thymus DNA were used as primer. 13 ug of Fraction 4 protein were used in each assay. 100 mumoles of each nucleotide were added to each assay. reaction in which ATP is converted to an acid-insoluble form in the absence of the other three triphosphates. Omission of 8-mercaptoethanol from the reactiom mixture resulted in a 50 per cent loss in activity; however, dilution of the con- centrated enzyme into solutions not containing a sulfhydryl compound resulted in as much as 90 per cent inactivation. The optimal pH for the reaction was between 7.8 to 8.2. At pH 6.1, 7.0, and 8.9 the activities were 13, 62, and 84 per cent, respectively, of the maximal value. (1) Nucleoside triphosphate specificity: All of the ribonucleoside triphosphates are required for RNA synthesis (Table 3). The deoxyribonucleoside triphos- phates do not function as substrates in the reaction, and the ribonucleoside diphos- phates support synthesis only at a greatly reduced rate. The observed activity of the diphosphates may be due to the presence of small amounts of the nucleoside triphosphates in the diphosphate preparations or to the formation of the triphos- phates through the action of nucleoside diphosphate kinase. With CTP*® as the labeled substrate and salmon sperm DNA as primer, vari- ation of the concentrations of all four ribonucleoside triphosphates as a group pro- duced a variation in the rate of RNA synthesis. When the data were plotted ac- cording to Lineweaver and Burk,*? a linear relationship was obtained from which VoL. 48, 1962 BIOCHEMISTRY: CHAMBERLIN AND BERG 85 TABLE 4 TABLE 5 Tue Errect or Dirrerent Nucuerc AcIip Tue Errect or DENATURATION ON THE PREPARATIONS ON THE RatE oF RNA ABILITY OF DNA PREPARATIONS TO SYNTHESIS BY RNA POLYMERASE Prime ror RNA SYNTHESIS Incorporation Incorporation of CM P22 Source of primer of CMP?#2* Native Heated DNA DNA (mumoles) Salmon sperm 100 Calf thymus 2.7 2.3 Calf thymus 43 Salmon sperm 6.1 2.5 E. coli 34 T6 phage 1.9 0.8 OX 174 38 E. col 1.8 1.9 B. subtilis 27 Assay procedure as described previously, except Adg phage 25 that the usual primer was replaced by 200 mumoles of T2 phage 45 the DNA to be tested. 7.4 ug of Fraction 4 protein T6 phage 30 were used in each assay. The DNA samples were paag heated for 10 min at 95 to 99° in 0.05 M NaCl and T5 phage 74 rapidly cooled in an ice bath. The absorbencies of RNA the heated DNA preparations were 30 to 40 per cent E. coli amino acid. acceptor <0.5 higher than those of the unheated preparations at 260 . -ace . mp. E. coli ribosomal <0.5 TMV <0.5 * The incorporation value for salmon sperm DNA was 5.3 mumoles and is set at 100 for comparison with the other primers. Assay system and procedure as previously de- scribed, except that 100 mymoles of each nucleic acid were used in place of the usual primer. 7.4 ug of Frac- tion 4 protein were used in each assay. it was calculated that the rate of synthesis was half maximal when the concen- tration of each of the triphosphates was 1.3 X 10-4 M. A similar value (1.4 X 10-4 M) was obtained using C!4 ATP as a label and calf thymus DNA as primer. (2) The nature of the primer: All DNA samples tested were active in pro- moting ribonucleotide incorporation, although the efficiency varied significantly (Table 4). Amino acid-acceptor RNA, ribosomal RNA from E. coli, and TMV RNA did not substitute for DNA. With the synthetic copolymer d-AT as pri- mer, only AMP and UMP were incorporated, and only GMP and CMP were in- corporated in the presence of d-GC polymer (Table 7). It should be noted, how- ever, that GTP was incorporated to a considerably greater extent in the latter case. A qualitatively similar finding has been reported for the incorporation of dGMP and dCMP by DNA polymerase with d-GC as primer.*? Increasing the amount of DNA in an assay mixture over the range 0 to 200 mymoles resulted in an increase in nucleotide incorporation. Further increases ‘in the amount of DNA, up to 400 mumoles, had no effect on the rate of RNA syn- thesis. A similar experiment using calf thymus DNA as primer gave a saturating value of 250 mumoles. The effect of disrupting the DNA double helix by heating*? is shown in Table 5. It is seen that with several of the DNA preparations there is a significant decrease in the rate of CMP* incorporation using the heated DNA while with others the effect is insignificant. The ability of single-stranded DNA to function as a pri- mer for RNA synthesis is further emphasized by the activity of the single-stranded DNA from @X 174 phage. (3) Metal ion requirements: Optimal concentrations for Mn++ and Mg+t when added separately to the reaction mixture were 2 X 107? M and 8 X 10-* M, respectively (Figure 1). Addition of Mg++ increased the rate at sub- optimal levels of Mn++; thus, the addition of 10-? M Mn++ and 4 X 10-3 M 86 BIOCHEMISTRY: CHAMBERLIN AND BERG Proc. N. A. 8. Nn 3 C\ Mn+ 27.0-- 4 er t\ : \ ° 6.0 e O 4 e Mott 3 9 6 0 ~ 550 — ~~ _ D co O od & §40- ~ oO £ e 3.0 | { 05 1.0 1.5 2.0 Molarity of metal ion x 107 Fic. 1—The Influence of Metal Ion Concentration on the Rate of CMP® Incorporation. Standard assay conditions were used, except that the metal ion concentration was varied as shown. 7.4 yg of Fraction 4 en- zyme were added to each assay. O——O Mg?** alone added to the assay mixture @——e Mn?" alone added to the assay mixture Mg++ to the same reaction mixture gave a rate of incorporation equal to that found with the optimal concentration of Mn++ alone (2 X 107? M). Characterization of the enzymatically synthesized RNA: (1) Net synthesis of the RNA product: With two times the level of the four ribonucleoside triphosphates and five to ten times the amount of enzyme used in the routine assay, the amount of RNA formed during an extended incubation exceeded the afmount of DNA added to the reaction. We will designate this as “net synthesis.”” With most of the DNA preparations used, the amount of RNA formed was three to five times greater than the amount of DNA added, while with d-AT copolymer, up to 15 times as much of the corresponding AU polynucleotide was produced (Table 6). The rate of synthesis. decreased after the first 20 min although further synthesis occurred up to two hr. Preliminary experiments indicate that this was not due to enzyme inactivation nor to destruction of the priming DNA, but other possi- bilities have not yet been investigated in detail. (2) Enzymatic and alkaline degradation of the product: Exposure of the isolated ‘net synthesis” product to alkali converted > 98 per cent of the label to acid- soluble products which were electrophoretically identical with the 2’-(3’) nucleo- side monophosphates. Treatment with pancreatic DNase or E. coli DNA di- esterase*! produced no significant liberation of labeled acid-soluble products. VoL. 48, 1962 BIOCHEMISTRY: CHAMBERLIN AND BERG 87 TABLE 6 NET SyNTHESIS OF RNA Labeled Calculated Ratio of RNA Method Source of nucleotide amount of isolated to of DNA primer incorporated RNA formed* DNA added isolation (mymoles) CMP?2 Calf thymus - 81 200 2.0 A oX 174 phage 90 510 5.1 A T2 phage 78 360 3.6 A T2 phage 72 410 4.1 B C!\AMP T2 phage 150 460 4.6 Cc T5 phage 152 500 5.0 C d-AT copolymer 155 310 15.0 C * The amount of RNA in the isolated product was calculated from the amount of label incorporated and the base ratio of the primer DNA. Synthesis: Each tube contained in a final volume of 0.5 ml: 20 zmoles of Tris buffer, pH 7.85; 8 nmoles of MgCl:; 400 mumoles each of ATP, CTP, UTP, GTP; 6 »zmoles of é-mercaptoethanol: 100 mymoles of DNA; and 100 ug of Fraction 4 protein. When d-AT was used as primer, only 20 mumoles of primer were added and CTP and GTP were omitted from the mixture. The incubation time was 3 hr at 37°. Product isolation: A. The incubation mixture was heated for 10 min at 60° in 0.4 M NaCl, then di- alyzed 36 hr against 0.2 M NaCl-0.01 M Tris, pH 7.85. B. The reaction mixture was extracted two times with phenol and the phenol fractions were washed two times with 0.4 M NaCl. The aqueous layers were pooled and dialyzed asin A. C. The product was precipitated from the incubation mixture with a solu- tion containing 60 per cent ethanol and 0.5 M NaC! at 0°, washed once with the same solution, and dis- solved in 1 ml of 0.2 M NaCl., then dialyzed as in A. Treatment of 10 to 20 mumoles of enzymatically prepared CMP??-labeled RNA with 0.1 wg of pancreatic RNase for 1 hr liberated 75 to 94 per cent of the P* label as acid-soluble products. The amount of acid-insoluble P*? remaining after RNase treatment varied with different DNA primers and different methods of product isolation. Using 10 times the amount of RNase did not appreciably alter the results. The significance of this RNase resistant fraction is presently un- known. (3) Nucleotide composition: The nucleotide composition of the product was examined by two different methods. In the first method, four separate assays, each containing a different labeled nucleoside triphosphate, were performed with each DNA preparation, and the molar ratio in which the labeled nucleotides were incorporated was measured (Method A). The second method utilized electro- phoretic separation‘! of the mononucleotides resulting from the alkaline degrad- ation of a “net synthesis” product in which all of the nucleoside triphosphates were labeled with C!* (Method B). The distribution of the label among isolated nucleotides was”therefore a measure of the composition of the newly synthesized RNA. The results (Table 7) indicate that the gross composition of the product at all stages of synthesis was complementary to that of the primer within the ac- curacy of the method. For double-stranded DNA, this complementary relation- ship becomes one of identity, since in the priming DNA adenine equals thymine and guanine equals cytosine. However, in the case of single-stranded @X 174 DNA (Table 8), the composition is indeed complementary to that of the DNA, and in this instance the amounts of AMP and UMP incorporation and of GMP and CMP incorporation are not equal. Furthermore, when double-stranded @X 174 DNA is used, the nucleotide composition of the resulting RNA is again iden- tical to that of the DNA primer. (4) Sedimentation velocity of the isolated product: The sedimentation velocity of the isolated RNA product was determined in the Spinco Model E analytical ultracentrifuge using ultraviolet optics. Values obtained (So) in 0.2 M NaCl- 88 BIOCHEMISTRY: CHAMBERLIN AND BERG Proc. N. A. 5. TABLE 7 NvucLeoTipE CoMPosITION OF THE RNA Propuct Method Primer* Product Product % ° Nucleotide Composition—-_—.. A+ T A+ U +G DNA primer analysis AMP UMP G CMP G+cec G4+C U+C (mymoles) d-GC polymer A <0.03 <0.03 1.90 0.23 — — — d-AT copolymer B 21.7 20.0 _— —_— —_ —_ —_ d-AT copolymer A 20.8 22.2 <0.03 <0.08 —_ — _ T2 phage B 7.7 7.4 4.3 4.3 1.76 1.76 1.08 T5 phage B 4.8 5.0 3.6 3.4 1.56 1.40 1.00 E. coli A 1.8 1.9 1.9 2.0 1.01 0.95 0.95 M. phlei A 3.7 4.0 7.9 8.5 0.48 0.47 0.93 A. aerogenes A 1.8 1.7 2.3 2.2 0.80 0.78 1.05 * The values given for the ratio A + T/G + Cin the priming DNA are those found by Josse et al.?9 except in the case of phage T5 DNA.® Method A: For each DNA sample, four separate incubations were used, each containing a different C!labeled nucleotide. The amounts of DNA used in the various tests were as follows: 20 mumoles of M. phlei, 50 mumoles of A. aerogenes, 180 mumoles of E. coli, 20 mumoles of d-AT, 20 mpmoles of d-GC. 12.5 ug of Fraction 4 enzyme were used in each incubation; all other conditions were those given for a standard assay. Method B: The synthesis of the C24-labeled RNA was carried out under the following conditions. The reaction maixture (0.5 ml) contained: 20 ymoles of Tris buffer, pH 7.85, 0.5 pmole of MnCh, 2 wzmoles of MgCl, 6 uzmoles of 6-mercaptoethanol, 100 mpmoles each of C1CATP, C!1LUTP, C1+GTP, C1*+CTP, 100 mumoles of DNA, and 180 ug of Fraction 4 enzyme. Where d-AT primer was used, 20 mumoles of primer were added and no CTP or GTP were added. After 180 min at 37°, the product was precipitated and washed with cold 3 per cent PCA and incubated in 0.3 M KOH for 18 hr at 37°, An aliquot to which carrier nucleotides had been added was subjected to paper electrophoresis at pH 3.5 in 0.05 M citrate buffer. The individual nucleotides which were visualized with a UV lamp were eluted in 0.01 M HCl and counted. Recovery of the C!-label in the eluted fractions was >95 percent. 180 mymoles, 140 mymoles, and 200 mumoles of polyribonucleotide were produced in the reactions primed with T2 DNA, T5 DNA, and d-AT, respectively. TABLE 8 CompaRaTIVE BEHAVIOR OF SINGLE- AND DouBLE-STRANDED @X 174 DNA as PRIMER FOR RNA SyNTHESIS Nucleotide Composition of RNA P UMP GMP State of DNA A CMP used as primer (per cent) Single-stranded Predicted* 32.8 24.6 18.5 24.1 “ “ Found by method A 32.0 24.1 19.5 24.3 “ “ Found by method B 35.0 24.6 19.3 21.1 Double-stranded Predicted * 28.7 28.7 21.3 21.3 “ “ Found by method B 28.9 29.1 20.9 20.9 Method A: Conditions as given in Table 7. 32 mumoles of single-stranded @X 174 DNA were used in each incubation with 8 ug of Fraction 4 enzyme. Method B: Conditions as givenin Table 7. With single-stranded DNA as primer, 25 mumoles of priming DNA were added, 71 mumoles of RNA were produced in a 60 min incubation with 80 ug of Fraction 4 en- zyme. For the double-stranded DNA, 26 mumoles of priming DNA were added; 32 mumoles of RNA were produced in a 60 min incubation with 40 ug of Fraction 4 enzyme. * The predicted values were calculated on the assumption that the single-stranded @X 174 DNA would yield RNA with a composition complementary to the composition reported by Sinsheimer.4® Upon replication of @X 174 DNA with DNA-polymerase it was assumed that the product (presumably double- stranded DNA) had a base composition which is the average of the composition of the original and of the newly synthesized strands. That this is a reasonable assumption is shown by unpublished studies of M. Swartz, T. Trautner, and A. Kornberg. When @X 174 DNA was used to prime limited (<30 per cent) or extensive (600 per cent) DNA synthesis, the composition of the newly formed DNA was: dAMP TMP dGMP adCMP Limited synthesis 31.0 24.1 20.1 24.5 Extensive synthesis 29.4 26.9 22.3 21.3 0.01 M Tris, pH 7.9, ranged from 6 to 7.5 for 2- to 15-fold ‘“‘net synthesis’ products prepared by phenol extraction or by salt-ethanol precipitation. DNA-dependent formation of polyadenylic acid: As pointed out earlier, RNA synthesis, as measured by the incorporation of either labeled CTP, UTP, or GTP did not occur in the absence of the other three nucleoside triphosphates or, in fact, in the absence of any one of the nucleoside triphosphates. It was therefore sur- prising to find that purified fractions of RNA polymerase catalyze the conversion VoL. 48, 1962 BIOCHEMISTRY: CHAMBERLIN AND BERG 89 of C!-ATP to an acid-insoluble form in the absence of the other three ribonucleo- side triphosphates. The ratio of the activities AMP incorporated in the absence of UTP, CTP, GTP AMP incorporated in the presence of UTP, CTP, GTP increased from 0.5 to 10 as purification of the enzyme progressed. (1) Requirements for polyadenylic acid formation: Polyadenylic acid formation from ATP occurred only in the presence of DNA, a divalent cation, and the pur- ified enzyme (Table 9). Note that addition of unlabeled ADP produces only a TABLE 9 TABLE 10 REQUIREMENTS FOR POLYADENYLIC ACID INCORPORATION OF SINGLE NUCLEOTIDES BY FORMATION RNA PoLyMERASE Incorporation Nucleotide Nucleotide of AMP added incorporation System (mymoles) (mmoles) Complete (with ATP as the only Ci ATP 23 nucleoside triphosphate) 9.9 C'11UTP 0.90 minus DNA <0.03 C1LGTP 0.09 minus Mn++ 2.5 CTP 0.07 ; ++ mis Ne ae The conditions were the same as those described in p . Table 9, except that ATP was replaced where indi- plus DNase 0.3 cated by an equal amount of each of the other nucleo- plus ADP 7.6 side triphosphates. 6 ye of Fraction 4 enzyme were added. The reaction mixture contained, in a final volume of 0.25 ml: 10 wmolea of Tris buffer, pH 7.85; 0.5 »mole of MnCk; 2 wmoles of MgCl; 3 «moles of §-mer- captoethanol; 100 mumoles of C1“ATP; 280 muz- moles of calf thymus DNA; and 3 ug of Fraction 4 RNA polymerase. Where indicated, 25 yg of pancre- atic RNase, 25 ug of pancreatic DNase, and 100 mymoles of ADP were added. The incubation time was 10 min at 37°. small dilution of the incorporation of label from C1*-ATP. The rate of incorpor- ation was directly proportional to the amount of enzyme added; 1.8, 3.6, and 7.2 ug of Fraction 4 enzyme catalyzed the incorporation of 4.0, 8.2, and 17.5 mumoles of C/+LAMP in a standard 10 min assay. The rate of incorporation remained constant up to over 75 per cent utilization of the added ATP. There was no incorporation of CMP or GMP when the corresponding nucleo- side triphosphates were added singly to the reaction, although UMP incorporation occurred to a small, but significant, extent (Table 10). (2) The DNA requirement for polyadenylic acid formation: The ability of various nucleic acid preparations to support polyadenylic acid synthesis is shown in Table 11. Note that neither RNA nor polyadenylic acid itself replaced the DNA requirement. To test whether DNA might be necessary only to initiate polyadenylic acid synthesis, an experiment was performed in which the priming DNA was destroyed after some polyadenylic acid formation had already occurred. It can be seen that destruction of the DNA by DNase blocked further synthesis of the polyadenylic acid (Table 12). This implies that the DNA is required not only for the initiation of polyadenylic acid synthesis, but also for the continued formation of the polynucleotide. (3) The effect of the other ribonucleoside triphosphates on polyadenylic acid for- mation: The addition of the other ribonucleoside triphosphates resulted in an inhibition of the rate of C4 AMP incorporation (Table 13). It can be seen, for 90 BIOCHEMISTRY: CHAMBERLIN AND BERG Proc. N. A.S. TABLE 11 TABLE 12 Priminc EFriciency or Various Nuc.ieic Tue Errect of DEOXYRIBONUCLEASE AcID PREPARATIONS FOR POLYADENYLIC ACID ADDITION DURING POLYADENYLIC ACID ForRMATION SYNTHESIS AMP AMP incorporation incorporation Primer (mumoles) Tube Treatment (mumoles) Calf thymus DNA 10 1 5-min incubation 7.1 Salmon sperm DNA 7.7 2 10-min incubation 16.2 T2 phage DNA 4.9 3 10-min incubation 6.9 d-AT copolymer <0.03 . . The reaction mixtures were as described in Table 9, Amino acid-acceptor RNA 0.35 except that 6 ue of Fraction 4 was used. Tube 1 was Polyadenyliec acid 0.07 incubated for 5 min, heated for 3 min at 100°, and . : . then assayed as usual. Tube 2 was incubated for 10 _ The conditions of the incubation were as described min before assaying. Tube 3 was incubated for 5 min in Table 9, except that the following amounts of and heated aa in the case of tube 1; 25 we of pancreatic nucleic acid were added: 300 myumoles of salmon DNase were then added and the mixture incubated sperm DNA; 100 mumoles of T2 phage DNA; 12 for an additional 5 min. At this time, 6 ug of fresh myumoles of d-AT; 110 mpmoles of amino acid— RNA polymerase were added and a third 5-rnin incu- acceptor RNA and 5 mumoles of polyadenylic acid. bation was allowed. 3 ug of Fraction 4 enzyme were added. TABLE 13 Errecr oF THE OTHER RIBONUCLEOSIDE TRIPHOSPHATES ON POLYADENYLIC ACID FORMATION _ AMP incorporation Component (mymoles) Complete system 26 plus CTP 6. plus UTP 5.2 plus GTP 2.0 plus CTP, UTP 1.3 plus CTP, GTP 0.6 plus UTP, GTP 0.5 plus UTP, GTP, CTP 2. Complete system as in Table 9, except that 6 ug of Fraction 4 protein were added. Where indicated, 100 mumoles of each nucleoside triphosphate were added. example, that in the presence of any two of the other triphosphates the amount of polyadenylic acid formed is less than 5 per cent that of the control in which only ATP was added. As has been previously shown, in the presence of all four tri- phosphates, AMP is incorporated into a product having a base composition deter- mined by the DNA primer, and hence under these conditions polyadenylic acid synthesis does not appear to occur. (4) Characterization of the polyadenylic acid product: Preliminary characteri- zation of the product is consistent with its identity as a polyadenylic acid. The addition of pancreatic RNase to the assay system lowered the rate of incorporation only slightly (about 30%). Treatment with 0.5 4 KOH for 18 hr at 37° con- verted the product to an acid-soluble form. Of the C™ in the hydrolysate, 97 per cent was associated with 2’-(3’) AMP on paper chromatography** and paper electrophoresis,* less than 1.5 per cent with adenosine, and less than 1.5 per cent was found in a region corresponding to adenosine 3’-5’ diphosphate. This im- plies that the minimum chain length of the polyadenylate is in the order of 60 to 70 nucleotide residues. Discusston.—There is a striking similarity between the reactions catalyzed by the RNA polymerase described here and EF. coli DNA polymerase.** Both use only the nucleoside triphosphates as nucleotidyl donors, and both display absolute requirements for a divalent cation and a DNA primer for polynucleotide synthesis.§ In both cases, some ambiguity exists as to the relative efficiency of single as com- pared to double-stranded DNA for priming of polynucleotide synthesis. In each Vou. 48, 1962 BIOCHEMISTRY: CHAMBERLIN AND BERG 91 reaction, both forms of DNA are active as primers, but a meaningful comparison between the two with regard to the mechanism of priming must await a more de- tailed physical and chemical characterization of the different DNA preparations, and further purification of the enzymes involved. The product formed by RNA polymerase, as in the analogous case of DNA polymerase?® has a base composition which, within experimental error, is comple- mentary to that of the priming DNA. This finding, which is in agreement with the results obtained by others'® 2}, supports the view that the nucleotide se- quences in the DNA direct the order of nucleotides in the enzymatically synthesized RNA. A more critical test of this hypothesis involves a comparison of the nucleo- tide sequence of the priming DNA and the newly synthesized RNA. In this re- gard, Furth et al.}® have shown that the repeating sequence of dAMP and dTMP in d-AT copolymer is faithfully replicated by the RNA polymerase in the form of an alternating AMP and UMP sequence. More recently Weiss and Nakamoto* have shown that RNA synthesized with an RNA polymerase from M. lysodeikticus contains the same frequencies of dinucleotide pairs as occur in the DNA primer. Additional experiments” which demonstrate the formation of a DNA-RNA com- plex after heating and slow-cooling* suggest that the homology of nucleotide se- quences may occur over relatively long regions. Does RNA polymerase copy the sequence of only one or both strands of DNA? This question is relevant not only to an understanding of the enzymatic copying mechanism, but also to any speculations as to the mechanism of information transfer from DNA to RNA. The fact that with double-stranded DNA primers the base composition of the newly made RNA is essentially identical to the over-all composition of both strands of the DNA already suggests that each strand can function equally well. An alternative hypothesis is to suppose that only one strand can be copied, and that the “primer” strand has, in the case of every DNA studied, a base composition identical to the average composition of both strands. Using the double-stranded form of @X 174 DNA*®. * in which it is known that the base compositions of the two strands differ,*® it is possible to test this question di- rectly. The results show that both strands of the duplex sérve to direct the com- position of the RNA product. This result still leaves open the question of whether both strands are copied in one replication cycle or whether only one strand is copied at a time and the choice between strands is random. When considering the relevance of this finding to information transfer, one must bear in mind that the existence of artificially pro- duced ends in an isolated DNA preparation may allow RNA formation to proceed from both ends of the double strand. This, however, may not occur with the DNA as it exists in the genome; that is, im vivo some structural feature in the chromo- somal DNA may cause RNA synthesis to proceed in a unidirectional manner and therefore copy the sequence of only one of the two strands. The formation of DNA-RNA complexes has been described by several groups of workers, © 5 although only limited information is available concerning their chemical structure and their metabolic and chemical stability. The fact that in the enzymatic reaction net synthesis of RNA occurs argues against the formation of a stable, stoichiometric complex of RNA and DNA. A further argument against the formation of such a complex is the finding that most of the DNA re- 92 BIOCHEMISTRY: CHAMBERLIN AND BERG Proc. N. A. 8, maining at the end of the reaction appears to be identical to the DNA added, and no component containing both DNA and newly synthesized RNA was detectable on CsCl gradient centrifugation.” Whether some transient complex is formed as an intermediate is somewhat more difficult to assess. The formation of polyadenylic acid in a DNA-dependent reaction is significant in view of the fact that none of the other ribonucleoside triphosphates, taken singly or even in groups of three, are utilized to any appreciable extent for poly- nucleotide synthesis. An exception to this is, of course, the situation where the DNA dictates the incorporation of only one or two nucleotides (e.g., with poly dT,** d-AT,”* or d-GC). Three possibilities which could account for polyadenylic acid synthesis are that it results from (a) a special feature of RNA polymerase itself, (0) a separate poly- adenylic acid polymerase, or (c) polynucleotide phosphorylase. The last pos- sibility is least likely because of the absolute requirement for DNA in the initiation and continuation of synthesis, the failure of ADP to give a significant dilution of the incorporation from ATP, the low amounts of polynucleotide phosphorylase activity found in the enzyme preparation as measured by P;*? exchange, the in- ability of Mg++ alone to support maximal rates of synthesis, the lack of polymeri- zation of the other nucleoside triphosphates, and the marked inhibition of poly- adenylic acid synthesis by any one or all four of the triphosphates The question of whether polyadenylic acid synthesis is catalyzed by RNA polymerase or by another enzyme cannot be resolved at the present time. With regard to the mechanism of the DNA-dependent polyadenylic acid tor- mation, two aspects deserve specific comment. The first concerns the mechanism of the inhibition of polyadenylic acid synthesis by any one or all of the other tri- phosphates. It should be recalled that polyadenylic acid synthesis does not occur in the presence of all four ribonucleoside triphosphates (< 3%), since under these conditions the base composition of the newly synthesized RNA is very close to that predicted by the composition of the DNA primers. The second notable feature of the reaction is its complete dependence on DNA, and the failure of d-AT to prime polyadenylic acid synthesis. One way to account for these findings is to assume that a sequence of thymidylate residues in the DNA, which does not occur in d-AT, can prime the formation of a corresponding run of AMP residues and, by subsequent “slippage” of one chain along the other, lead to a DNA-dependent elongation of the polyadegylic acid chain. The introduction of any other nucleotide into the growing chain might block or inhibit the sliding process and thereby terminate the growing polyadenylic acid chain. Summary.—An RNA polymerase has been isolated from ZF. colt which in the pres- ence of the four ribonucleoside triphosphates, a divalent metal ion, and DNA synthesizes RNA with a base composition complementary to that of the priming DNA. Both strands of DNA can prime new RNA synthesis. Thus, while single- stranded @X 174 DNA yields RNA with a base composition complementary to that of the single-stranded form, double-stranded @X 174 DNA (synthesized with DNA polymerase) primes the synthesis of RNA with a base composition virtually the same as that in both strands of the DNA. A novel feature of the RNA polymerase preparations is their ability to catalyze a DNA-dependent formation of polyadenylic acid in the presence of ATP alone. Neither UTP, GTP, nor CTP yields corre- Vou. 48, 1962 BIOCHEMISTRY: CHAMBERLIN AND BERG 93 sponding homopolymers; the DNA-dependent formation of polyadenylic acid is vir- tually completely inhibited by the presence of the other nucleoside triphosphates. * This work was supported by Public Health Service Research Grant No. RG6814 and Public Health Service Training Grant No. 2G196. t Pre-doctoral Fellow. ¢ The abbreviations used in this paper are: RNA and DNA for ribo- and deoxyribonucleic acid, respectively; poly dT for polydeoxythymidylate; d-AT for the deoxyadenylate-thymidylate copolymer; d-GC for the deoxyguanylate-deoxycytidylate polymer: AMP, ADP and ATP for adenosine-5’-mono-, di-, and triphosphates, respectively. A similar notation is used for the cyti- dine (C), guanosine (G), and uridine (U) derivatives and their deoxy analogues (dA, dC, dG, dT). P; is used for inorganic orthophosphate, TMV for tobacco mosaic virus, and DNase and RNase for deoxyribo- and ribonuclease activities, respectively. § Under certain conditions DNA polymerase preparations will, in the absence of DNA, produce d-AT or d-GC depending upon the nature of the substrates present.*2, 33 1 Ingram, V. M. and J. A. Hunt, Nature 178, 792 (1956). 2 Yanofsky, C. and P. St. Lawrence, Ann. Rev. Microbiol., 14, 311 (1960). 3 Fincham, J. R. 8., Ann. Rev. Biochem., 28, 343 (1959). 4 Volkin, E. and L. Astrachan, Virology, 2, 149 (1956). 5 Volkin, E., these Procrepinas, 46, 1336 (1960). ® Nomura, M., B. D. Hall and S. Spiegelman, J. Mol. Biol., 2, 306 (1960). 7 Yéas, M. and W.S. Vincent, these Procrrpinas, 46, 804 (1960). 8 Gros, F., W. Gilbert, H. Hiatt, P. F. Spahr, and J. D. Watson, Cold Spring Harbor Symposia on Quantitative Biology, vol. 21, in press. ‘ § Jacob, F., and J. Monod, J. Mol, Biol., 3, 318 (1961). 0 Ochoa, 8. and L. Heppel, in The Chemical Basis of Heredity, ed. W. D. McElroy and B. Glass (Baltimore: The Johns Hopkins Press, 1957), p. 615. " Littauer, U. Z. and A. Kornberg, J. Biol. Chem., 226, 1077 (1957). 1 Hecht, L. 1, M. L. Stephenson, and P. C. Zamecnik, these ProckEpinas, 45, 505 (1959). 13 Canallakis, E. S. and E. Herbert, these PRoceEpINGs, 46, 170 (1960). 14 Preiss, J.. M. Dieckmann, and P. Berg, J. Biol. Chem., 236, 1749 (1961). 16 Weiss, 8. B., these Proceepines, 46, 1020 (1960). ‘6 Weiss, S. B. and T. Nakamoto, J. Biol. Chem., PC 18 (1961). 7 Hurwitz, J., Bresler, A. and R. Diringer, Biochem. Biophys. Res. Comm., 3, 15 (1960). 8 Furth, J. J., J. Hurwitz, and M. Goldmann, Biochem. Biophys. Res. Comm., 4, 362 (1961). :9 Ibid., 4, 431 (1961). 2” Stevens, A., Biochem. Biophys. Res. Comm., 3, 92 (1960). 21 Stevens, A., J. Biol. Chem., 236, PC 43 (1961). 22 Ochoa, S., D. P. Burma, H. Kriger, and J. D. Weill, these Procrepinas, 47, 670 (1961). 23 Burma, D. P., H. Kréger, 8S. Ochoa, R. C. Warner, and J. D. Weill, these ProcEEDINGS, 47, 749 (1961). 4 Huang, R. C., N. Maheshwari, and J. Bonner, Biochem. Biophys. Res. Comm., 3, 689 (1960). 2% Lehman, I. R., M. J. Bessman, E. 8S. Simms, and A. Kornberg, J. Biol. Chem., 233, 163 (1958). 26 Ofengand, E. J., Ph.D. Thesis, Washington University, St. Louis, Missouri (1959). ” Hurwitz, J., J. Biol. Chem., 234, 2351 (1959). 2 Kay, E. R. M., N. S. Simmons, and A. L. Dounce, J. Am. Chem. Soc., 74, 1724 (1952). 29 Josse, J., A. D. Kaiser, and A. Kornberg, J. Biol. Chem., 236, 864 (1961). 30 Kaiser, A. D., and D. S. Hogness, J. Mol. Biol., 2, 392 (1960). 31 Lehman, I. R., J. Biol. Chem., 235, 1479 (1960). 32 Schachman, H. K., J. Adler, C. M. Radding, I. R. Lehman, and A. Kornberg, J. Biol. Chem., 235, 3243 (1960). , 33 Radding, C. M., J. Josse, and A. Kornberg, unpublished results. 34 Nester, E. W., and J. Lederberg, these Proceeprnes, 47, 56 (1961). 36 Lehman, I. R., Ann. N. Y. Acad. Sct., 81-3, 745 (1959). %¢ Lehman, I. R., R. L. Sinsheimer, and A. Kornberg, unpublished results. 37 Ofengand, E. J., M. Dieckmann, and P. Berg, J. Bio’. Chem., 236, 1741 (1961). 94 BIOCHEMISTRY: WOOD AND BERG Proc. N. A. 8. _ 38 Lehman, I. R., G. G. Roussos, and A. Pratt, J. Biol. Chem., in press. 39 Lowry, O., J. J. Rosebrough, A. L. Farr, and R. J. Randall, J. Biol. Chem., 193, 265 (1951). Monod, J., Ann. Inst. Pasteur, 79, 390 (1950). 4) Wiesmeyer, H., and M. Cohn, Biochim. Biophys. Acta, 39, 417 (1960). 42 Lineweaver, H., and D. Burk, J. Am. Chem. Soc., 56, 658 (1934). 43 Doty, P., these ProcrEDINGS, 42, 791 (1956). ‘4 Markham, R. and J. P. Smith, Biochem. J., 52, 552 (1952). 45 Magasanik, B., E. Vischer, R. Doniger, D. Elson, and E. Chargaff, J. Biol. Chem., 186, 37 (1950). 46 Weiss, S. B., and T. Nakamoto, these Procerepines, 47, 1400 (1961). 47 Geiduschek, E. P., T. Nakamoto, and S. B. Weiss, these PRockEpDINGs, 47, 1405 (1961). ‘8 Hall, B. D., and 8. Spiegelman, these Procrxpines, 47, 137 (1961). 49 Sinsheimer, R. L., J. Mol. Biol., 1, 43 (1959). 60 Rich, A., these PRocrEepinas, 46, 1044 (1960). §1 Schildkraut, C. L., J. Marmur, J. R. Fresco, and P. Doty, J. Biol. Chem., 236, PC 2 (1961). 52 Wyatt, G. R., and 8. S. Cohen, Biochem. J., 55, 774 (1953).