D. L. KALJA?;. K. F. WPITSO?;, A. BUKK'I- AND S. SPIEGELMAN instituie 01 Cnncw Hcseaych. Cdegt oj Pnysiiaans and Surgeons, Columbia Unaversaty, iVew York. (Received June 1st rq;r) Y.Y. r0::jz IG.S,A.! SIJMMARY DNA polymerase fwm avian rryeiobla?tos?s virm lid\ been pwified by a com- bination of rulumn rliromatographv and gel filtration methods The iaolated enzyme sediments xi apprwmately 6 S and consists of two subunits of molecular weiglds IIO ooo and 64 ooo It is fret. oi RS-4 and DNA endonuclease activity. The enzyme pos5esspi the RSA-, DXA-, and hybrid directed polymeraic activities found in the virior, iNTRODUCTlO> The discovery1,2 of s ribonuclease-sensitive DNA polymerase activity in on- cogenic RSA viruses was quickly extended3-' to x wide .variety of oncornavirusess. I: wzs filrrher shown that the product DNA was complementary to the RNA uf the virion sed as the source of the enzvme preparatiorr3, $-I1. The.e findings were promptly foliowed bq- experiments that establsiied the existence in these viruses of polymerase activitics that respond ti: doubie-stranded DNAY*12-15 and, with a very high effi- ciency. to synthetic honiopolymeri- dupiexes13*16 composed of pdyribonucleotides poiydeoxyri tmnucii.oticieJ. and hybrid structures of the tw(,, The size of the DNA product synthesized was generally much less than that of the template empl0yed3*~9,11. In aaditim to tiit; DNA paivnierase activities, evidence was also found for DNA endo- and exonu*cleases13s 17, ligas~~~. and a nucleosidetriphosphate phospho- transferase's in tlie vmon Ikiiication of the RNA dependent DNA polymerase is a necessary prerequi- site to an unambiquous analysis of the reaction mechanism. Purity is also requi- red to delineate tlie relation of this polymerase activity to the others observed with DNA ana synthetic homopolymeric duplexes as templates. Finally, pure enzyme should permit a decision on whether the nuclease, ligase and phosphotransferase ac- tivities are coliectively or individually inherent and necessary components of the DNA polymerase function. We report here the purification and characterization of the DNA polymerase activity from avian myeloblastosis virus. Abbreviations: AMV, avian myeloblastosis virus; BBOT, z,~-bis-z-(~-tert.-butylbenzoxa- zolyl) thiophene. Biochim. Biophys. Acta, 246 (1971) 365-383 366 D. L. KACIAN et al. MATERIALS Whatman microgranular DEAE-cellulose, DE52, 1.0 mequiv/g dry weight, and phosphocellulose, PII, 7.4 mequiv/g dry weight were obtained from Reeve Angel. Sephadex G-zoo and CM-Sephadex C-50 were purchased from Pharmacia Fine Che- micals. Clarkson Chemical Company provided hydroxylapatite. Unlabeled nucleoside triphosphates and dithiothreitol came from P-L Biochemicals. Miles Laboratories supplied polynucleotides and Micrococcus lysodeikticzts DNA. Tritiated nucleoside triphosphates were obtained from New England Nuclear, Schwarz BioResearch, and Amersham-Searle. Nucleic acid polymers were the generous gifts of Drs. F. Bollum (University of Kentucky), A. N. Nussbaum (Hoffman-LaRoche), and L. A. Under- kofler (Miles Laboratories). Acrylamide and methylene bisacrylamide came from Bio-Rad Laboratories. Nonidet P-40 was a product of Shell Chemical CO. METHODS (I) Purification of avian myeloblastosis virus Avian myeloblastosis virus (AMV), BAI strain A, was obtained by methods previously described19 from the blood of chicks in the terminal stage of myeloblastic leukemia20 and from infected myeloblasts suspended in tissue culture. Virus from blood plasma was purified essentially as described by CARNECIE et aLZ1. As a final step, the virus suspension was sedimented at 27 ooo rev./min in the Spinco SW27 rotor through 12 ml of 20 yo glycerol in 0.01 M Tris-HC1 (pH 8.5), 0.15 M NaC1, I mM EDTA (Tris-NaC1-EDTA buffer) onto a 6-ml pad of glycerol. The virus was removed from the pad, suspended in the same buffer without glycerol, and stored at -70~. Myeloblastosis virus produced in tissue culture was supplied by Dr. J. W. Beard, Duke University. The culture fluid had been concentrated 50-fold by centrifugation and contained from 2.5 1o12 to 5 - 10l2 virus particles per mP. After centrifugation at 3000 xg for IO min, the virus was concentrated against a 6-ml pad of glycerol at 27 ooo rev./min for I h in the Spinco SW 27 rotor. Further purification was as pre- viously described3. (2) Isolation of AMV RNA Purified virus from blood plasma was lysed by adding sodium dodecyl sulfate to 0.5 %. The suspension was extracted twice with phenokresol solution (prepared according to KIRBY~~ and equilibrated with Tris-NaC1-EDTA buffer) and the RNA was precipitated by addition of 0.1 vol. of 3 M NaCl and 2 vol. of 95 Yo ethanol. After a second alcohol precipitation, the RNA was layered onto a IO to 30 04 glycerol gradient containing 0.01 M Tris-HC1 (pH 7.4), 0.1 M NaC1, and I mM EDTA in the Spinco 5W 41 rotor. After centrifugation at 41 ooo rev./min and 5" for 3 h, fractions were collected dropwise from the bottom of the tube and those containing the 70-5 RNA component were pooled and alcohol precipitated. (3) Preparation of polynucleotide duplexes Polynucleotide duplexes were formed by annealing equimolar amounts of two complementary homopolymers at concentrations of approximately IOO pg/ml each in 0.01 M Tris-HC1 (pH 7.4), 0.2 M NaCl at room temperature for 15 min. Biochim. Biophys. Acta, 246 (1971) 365-383 (4 j Polyacrylamide gel electrojhoresis of proteins Polyacrylaniide gel electrophoresis in the presence of sodium dodecyl sulfate was performed by a modification of the method of SHAPIRO e,? ai.24. Protein samples were precipitated with an equal volume of IO yo trichloroacetic acid, allowed to stand at oo for 15 min, and centrifuged at 16 oooxg for 30 min. Recovery was greater than 95 "/I. The pellet was thoroughly drained, and thepre- cipitated protein was dissolved in 25-50 pl of 0.01 M sodium phosphate (pH 7.8), I yo sodium dodecyl sulfate, I "/o 2-mercaptoethanol. After 30 min at bo", glycerol was added to 13 7; and the sample was layered onto the gei. Gels contained 5 7; acrylamide, 0.25 methylene bisacrylamide, 0.1 sodium dodecyl sulfate, and 0.1 M sodium phosphate (pH 7.8). Electrophoresis was performed at IO rnR per gel for 15 min and then at 15 mA per gel for 75 min. Gels were stained for z h in 2 0.25 "/o solution of Coomassie brilliant biue in IO "/b acetic acid, 50 '7; meihanoi. They were destained by diffusion in 7 "/6 acetic acid, 5 7; methanol arid stored in the same solvent. Molecuiar weights of polypeptide chains were determined as described by WOLF et aLZ5 using as molecular weight markers polymers of ribonuclease A prepared ith diethyl pyrocarbonate. Electrophoresis of proteins ar: pH 8.9 in Tris-glycine buffer was performed as described by DAVIP. (5) Protein determination albumin (Fraction Vi as standard. Protein was measured by the method of LQWRY et aLZ7 using crystalline bovine (6 j Polymerase assay The assay- mixture for homopolymer templated reactions (total volume 0.1 ml) contained the foliowing in pmoles: Tris-HCl (pH 8.3), 5.0; MgCi,, 0.6; 0.02 each of the required labeled and unlabeled deoxynucleoside triphosphates; and double-stran- ded homopolvmer template, 1.2 . 103 pmoles polymer phosphate in each strand. Reactions were incubated ai 37' for IO min and terminated by the addition of coid 5 "/I trichloroacetic acid. After IO min, the acid-precipitable radioactivity was collected on nitrocellu- lose filters and counted in 0.4 yo 2,5-bis-z- (5-tert.-butylbenzoxazolyl)thiophene(BBOT) in toluene. Assays using natural RNA and DNA templates were prepared identically ex- cept that they contained 0.02 pmoles each of three unlabeled nucleoside triphos- phates and 4nmoles of the fourth labeled triphosphate. Templates were used at levels from I to 2 pg per 0.1 mi assay. Specific activities of the 3H-labeled triphosphates were 35-50 counts/min per pniole for honiopolymer-templated reactions and 350-500 counts/min per pmole for those using natural RNA or DNA templates. 17) Preparataow of AMV DIVA $olymerase The procedure is described for 60 mg of purified viral protein. Larger amounts have been handled successfully by scaling up the various steps proportionately. 12 ml of AMV (5 mg/ml in 0.01 M Tris-HC1 (pH 8.5), 0.15 M NaC1, I mM EDTA) were mixed in order with 1.2 ml Nonidet P-40, 1.2 ml IO yo sodium deoxy- Biochim. Biophys. Acta. 246 (1971) 365-383 D. L. KACIAN et al. 368 cholate, and 3.6 ml 4 M KC1 until homogeneous. The mixture was kept at oo for 15 min and then was centrifuged at 16 ooo xg for IO min. The pellet was discarded and the supernatant diluted to IO times its volume with 0.01 M potassium phosphate (pH 7.2), z mM dithiothreitol, IO "/o glycerol. The solution was applied to a 1.2 cm x 11.0 cm column of DEAE-cellulose carefully equilibrated with the same buffer. The column was washed with 80 ml of 0.05 M potassium phosphate (pH 7.2), z mM dithiothreitol, IO yo glycerol, and eluted with 40 ml 0.3 M potassium phosphate (pH 7.2), z mM dithiothreitol, IO yo glycerol. The flow rate was about 36 ml/h. The peak activity fractions from the DEAE-cellulose column were pooled and diluted to 3 times their volume with 0.01 M potassium phosphate (pH 8.0), z mM dithiothreitol, IO "/b glycerol. The material was loaded onto a 0.9 cm x 8.0 cm column of CM-Sephadex C-50 previously equilibrated with the same buffer. The column was washed with 8 ml of 0.1 M potassium phosphate (pH 8.0), 2 mM dithiothreitol, IO yo glycerol and eluted with 12 ml of 0.3 M potassium phosphate (pH 8.0), 2 mM dithiothreitol, IO yo glycerol. A flow rate of 15 ml/h was maintained. The peak fractions were pooled, glycerol was added to 50 SA, and the enzyme stored at -20'. (8) Phosphocellulose column chromatography of AMV DNA polymerase The peak fractions from a DEAE-cellulose column (about 5 mg protein) were pooled and diluted 6-fold with 0.01 M potassium phosphate (pH 8.0), 2 mM dithio- threitol, IO vh glycerol and applied to a 0.9 cm x 9.0 cm column of phosphocellulose equilibrated with the same buffer. The column was eluted with a 150-ml gradient from 0.05 M potassium phos- phate (pH 8.0) to 0.5 M potassium phosphate (pH 8.0) containing z mM dithio- threitol and IO yo glycerol. The flow rate was maintained at about 0.4 ml/min and about 1.5-ml fractions were collected. (9) Hydroxylapatite column chromatography of AMV DNA polymerase The peak fractions from a phosphocellulose column were pooled and diluted 5-fold with 0.01 M potassium phosphate (pH 7.2), 2 mM dithiothreitol, IO "/o glycerol and loaded onto a 0.9 cm x 9.0 cm column of hydroxylapatite equilibrated with the same buffer. The column was eluted with a 150-ml gradient from 0.05 to 0.5 M potassium phosphate (pH 7.2) containing 2 mM dithiothreitol and IO "/b glycerol. The flow rate was maintained at about 0.2 ml/min and about 2.0-ml fractions were collected. (IO) DNA cellulose chromatography of AMV DNA polymerase DNA cellulose was prepared essentially as described by ALBERTS AND HER- RICK^^. Clean cellulose (Munktell 410) was washed several times with boiling ethanol and distilled water to remove remaining pyridine. It was then pre-cycled with base and acid (0.1 M NaOH, water, 0.01 M HCI) and washed to neutrality with water. The cellulose was then thoroughly dried, first in air and then by lyophilization. Calf thymus DNA was dissolved in 0.01 M Tris-HC1 (pH 7.4), I mM EDTA at a concentration of I mg/ml. The DNA solution was poured into petri dishes and mixed with the cellulose to form a slurry (approximately I g cellulose to 3 ml DNA solution). The material was extensively air dried, ground to a powder, and lyophi- Bzochim. Biophys. Acta, 246 (1971) 365-383 369 PCRIFICATION OF -4MV DNA POLYMERASE iized. Slow, complete drying seems tu be essential for good adsorption of the DNA. Tlte powder w-as resuspended in 0.01 M Tris-HC1 (pH 7.4),1 mM EDTA. 0.15 M NaCI, washed twice with the same buffer, and checked for DNA adsorption by measuring optically the amount of DNA reieased by boiling. About 30-40 s, of the input DNA was taken Lip by the celluiose For chromatography ut AMV DNA polymerase, a 0.5 cm x IO cm column of DNA cellulosc wits exhaustively equilibrated with 0.01 M potassium phosphate (pH 8.01, 2 mM dithiothreito!, 10 0; glycerol. Approximateiy Go pg of AMV DNA polymerase (piiosphocelluiose iractionj was applied to the column in about 0.03 M potassium phosphate buffer. The column was eluted with a 32-ml linear gradient from 0.01 to 0.5 M potassium phosphate (pH 8.oj contaning 2 mM dithiothreitol, IO 7; glycerol. The flow rate was maintained at 8 mi/h and 0.5 ml fractions were collected. ( II) Glycerol gradtent centrzfugation of AMV DNA Poiymerase AMV DNA polymerase (phosphocellulose fraction, approximately 0.7 mgj was layered over a In to 30 yo iv/v) glycerol gradient in 0.2 M potassium phosphate pH ko), 2 mM dithiothreitol in the Spinco SW j0.I rotor. Bovine serum albumin was run on a parallel gradient to serve as marker. The proteins were sedimented at 50 000 rev.imin and T for 9.5 11 and Io-drop fractions were collected dropwise from the bottoms of the tubes through a 20-G needle. (12 ) Assay of contaminatzng nuclease activities Kibonuciease activity In the CM-Sephadex enzyme was measured by following the breakdown of 3H-labeled Escherichia coli 4-S and 5-S RNA. on polyacrylamide gels. The RNA (15pgj and enzyme (0.35pg) were incubated in 0.025 ml of the standard assay mixture iacking cieoxyriboside triphosphates and template. After o rnin and 60 min of incubation at 37', sodium dodecy! sulfate was added to I 04, and the samples subjected to electrophoresis on polyacrylamide gels as described by HisHo~etal.~~. The gels were frozen. cut into ~-mm slices, dried on filter paper strips, and counted in 0.4 96 BBOT in toluene. Deoxyribonuclease activity was measured by following the breakdown of 3H- labeled E. coli DNA by alkaline sucrose gradient centrifugation. Two standard reaction mixtures were prepared omitting the deox-yribonucleo- tides and including the iabeled DNA (approximately 0.30 pg). Purified Ami DNA polymerase (1.4 .fig) was added to one, and both were incubated at 37". After 30 min, EDTA was added to 5 mM and sodium dodecyl sulfate to 0.5 "/b and the samples were layered onto j to 20 7; sucrose gradients containing 0.1 M NaOH, 0.9 M NaC1, and I inM EDTA in the Spinco SW 50.1 rotor. After centrifugation at 50 900 rev./ rnin for 3 h at IO', fractions were collected dropwise from the bottom of the tube and precipitated with trichloroacetic acid. Insoluble material was collected on nitro- cellulose filters and counted in BBOT-toluene scintillation fluid. RESULTS Extraction of the DNA Polymerase from virions Solubilization of the AMV DNA polymerase was effected by treatment of the Baocham. Biophys. Acta, 246 (1971) 365-383 D. L. KACI.4N Et al. 370 virus particles with detergent (0.7 yo deoxycholate; 7 % Nonidet P-40) and salt (0.8 M KC1) at 0'. Glycerol gradient analysis in 0.1 M potassium phosphate showed that over 95 yo of the enzyme activity sedimented at 8.4 S or less after treatment. Assays were performed using poly (rA) . poly (rU), AMV 70-S RNA, and M. lyso- deikticus DNA as templates. Specific activities and recoveries were determined with poly (rA) - poly (rU). Lower concentrations of detergent or salt, while capable of releasing the enzyme from the virion, left much of the activity attached to material that seclimented at higher s values. After release the enzyme showed an absolute requirement for added template. The extract was centrifuged at low speed to remove a small amount of material that reduced the flow rate of the DEAE-cellulose column. The pellet contained a negligible amount (< 3 $6) of the activity. Chromatography of AMV DNA polymerase on DEA E-cellulose The solubilized enzyme was diluted Io-fold to reduce the concentrations of salt and detergent and loaded onto a column of DEAE-cellulose. The column was then exhaustively washed with 0.05 M potassium phosphate buffer, which removes the detergents and much of the protein. Virtually all the activity was retained through- out loading and washing. The enzyme was then eluted with 0.3 M potassium phos- phate buffer. About 5 yo of the protein is recovered from the column, together with most of the activity. Fig. I shows a DEAE-cellulose column assayed, respectively, with three diffe- rent templates: the homopolymer duplex poly (rA) - poly (rU), 70-S AMV RNA, and M. lysodeikticus DNA. With each template, the activity is found generally to coincide with the protein peak. The slight displacement of activity observed with various templates appears to be due to their differential sensitivity to contaminating proteins and salts. Generally, greater than 90 yo of the starting activity is eluted with the 0.3 M potassium phosphate. The DEAE-cellulose column step yields about a 20-fold en- richment of the enzyme. CM-Sephadex chromatcgraphy of AMV DNA polymerase The peak activity fractions from the DEAE-cellulose column were pooled, diluted 3-fold with low salt buffer, and loaded onto a column of CM-Sephadex. The column was washed with 0.1 M potassium phosphate and eluted with 0.3 M potas- sium phosphate. A411 of the enzyme activity is retained by the column during loading and washing. Fig. 2 shows the profiles obtained by assaying the column with three different templates. About 0.5-1 "/o of the total protein is eluted with the 0.3 M buffer. The degree of purification after CM-Sephadex chromatography and the amount of activity recovered have varied with different batches of virus, but highly reprodu- cible results are obtained when the same starting material is used. In every case, the column effectively removes all the acidic protein contaminants as measured by polyacrylamide gd electrophoresis at pH 8.9. Preparations using different batches of virus produced in tissue culture have yielded enzyme of 30-60-fold higher specific activity than the crude extract. Variability in yield is probably due to the instability of the enzyme. The history of the virus preparation may be of paramount importance Bzochrm. Baophys. Acta, 246 (1971) 365-383 37= PURIFICATION OF AMv DNA POLYMEXASE in obtaining high yields and specific activities. It appears that certain Treatments (e.g. rnultinle freezing and thawing) adversely affect the stability of the enzyme. An actual enzyme preparation carried through the Chf-Sephadex step is surn- marized i~ Tablc I The CM-Sephadex enzyme was assayed as described in METHODS for RNA and DXA endonmiease activities. As can be seen in Figs. 3 and 4, no detectable breab- Fractii i4 41- Froctlon Fig. I. DEAE-cellulose chromatography of AMV DNA polymerase. Avian myeloblastosis virus (600 mg) was solubilized as described in METHODS and chromatographed on a 2.5 cm x 25 cm column. Fractions of 4.4 ml were collected from the 0.3 M phosphate eluent. Assays using poly (rA) . poly (rU), AMV 70-S RNA, and M. lysodeikticus DNA were performed using 2 pl, 5 pl and 5 pl, respectively, from each fraction. Biochim. Biophys. Acta, 246 (1971) 365-383 372 D. L. KACIAN et al. down of the nucleic acids occurred when incubated with purified enzyme under stand- ard conditions, minus the deoxyriboside triphosphates. Enzyme stored in 50 yo glycerol at -zoo has retained greater than go yo of its activity for more than 5 weeks. Froclian Fig. 2. CM-Sephadex chromatography of AMV DNA polymerase. The pcak fractions (19-24) of thc DEXE-cellulose column shown in Fig. I were pooled and chromatographed on a I. j cm i 17.5 cm column. Fractions of 1.5 ml were collected from the 0.3 M phosphate eluent. I pl, 5 1'1, and j pl were used, respectively, to assay activity wlth poly (rA) . poly (rU), AMV I