Transcription of 70S RNA by DNA Polymerases ... - Journal of Virology

JOURNAL OF VIROLoGY, Dec. 1975, p. 1566-1574 Copyright X) 1975 American Society for Microbiology

Vol. 16, No. 6 Printed in USA.

Transcription of 70S RNA by DNA Polymerases from Mammalian RNA Viruses JOHN W. ABRELL,1 MARVIN S. REITZ, AND ROBERT C. GALLO* Litton Bionetics, Inc., Bethesda, Maryland 20014, and National CancerInstitute, Bethesda, Maryland 20014* Received for publication 19 June 1975

DNA polymerases purified by the same procedure from four mammalian RNA viruses, simian sarcoma virus type 1, gibbon ape lymphoma virus, Mason-Pfizer monkey virus, and Rauscher murine leukemia virus are capable of transcribing heteropolymeric regions of viral 70S RNA without any other primer. In this reconstituted system the enzymes from simian sarcoma virus type 1, MasonPfizer monkey virus, and Rauscher murine leukemia virus transcribe viral 70S RNA almost as efficiently as the DNA polymerase from the avian myeloblastosis virus, but gibbon ape lymphoma virus DNA polymerase is approximately threeto fivefold less efficient. Although there is a substantial difference among the sizes of these DNA polymerases (160,000 daltons for the avian myeloblastosis virus enzyme, 110,000 daltons for the Mason-Pfizer monkey virus enzyme, and 70,000 daltons for the mammalian type C viral polymerases), the ability to transcribe viral 70S RNA is a characteristic common to these enzymes. Disrupted virions of the avian RNA tumor viruses, avian myeloblastosis virus (AMV) and Rous sarcoma virus, catalyze an endogenous RNA-directed DNA synthesis (2, 24). After fractionation and purification, these viral DNA polymerases are still capable of transcribing viral 70S RNA in reconstituted reaction mixtures in the absence of an exogenous primer, such as oligo(dT) (5, 6, 11). When mammalian type C viruses are partially permeabilized with detergent, their DNA polymerases also catalyze an endogenous RNA-directed DNA synthesis (23). However, it has been reported that after purification these enzymes catalyze little or no transcription of 70S RNA unless exogenous primer is added (1, 10, 26, 27), although very recently Gerard and Grandgenett (7) have demonstrated that DNA polymerase purified from Moloney sarcoma virus can transcribe heteropolymeric regions of added AMV 70S RNA with an efficiency approaching that of AMV DNA polymer-

monkey virus (M-PMV), and gibbon ape lymphoma virus (GaLV). Three of these viruses are type C viruses, whereas the fourth, M-PMV, is of intermediate morphology, with some characteristics of both type B and type C viruses (12). The sizes of the DNA polymerases from both avian viruses, AMV and Rous sarcoma virus, are reported to be approximately 160,000 daltons (5, 11). The molecular weights of the four mammalian viral DNA polymerases are less than their avian counterparts; MuLVR (20), SSV-1 (1), and GaLV (15) are all about 70,000 daltons, and M-PMV is about 110,000 daltons (1). The estimated molecular weights of the enzymes described here are the same as previously reported. It appears that transcription of viral 70S RNA by purified viral DNA polymerases is not limited to the avian viral enzymes and is not strictly dependent on enzyme size.

ase.

In this report we establish that four mammalian viral DNA polymerases, purified in a manner similar to our published procedure (1), will transcribe exogenous 70S viral RNA in the absence of oligo(dT) almost as efficiently as the purified AMV DNA polymerase. The one murine and three primate RNA viruses used as sources for these DNA polymerases are Rauscher murine leukemia virus (MuLVR), simian sarcoma virus type 1 (SSV-1), Mason-Pfizer I Present address: Division of Lung Diseases, National Heart and Lung Institute, Bethesda, Md. 20014.

1566

MATERIALS AND METHODS Virus preparations. AMV was a gift from Joseph Beard, Duke University. SSV-1, grown in either 71AP-1 cells (a marmoset cell line) (28) or Kirsten virus-transformed nonproducer rat kidney (KW23) cells, was a product of Electronucleonics Laboratories, Inc. SSV-1, originally isolated from a fibrosarcoma of a woolly monkey (25), has been shown to be associated with an excess of a nontransforming helper virus (29). MuLVR grown in JLS-V9 cells was obtained from the Frederick Cancer Center. The John L. Smith Memorial for Cancer Research, Pfizer, Inc., was the source for the M-PMV and the GaLV, both of which were grown in NC-37 cells, a human lymphoblastoid cell line derived from nor-

mal lymphoid cells.

VOL. 16, 1975

TRANSCRIPTION OF 70S RNA

1567

against three 1-liter portions of buffer B (0.05 M Tris-hydrochloride [pH 7.5], 1 mM DTT, and 20% glycerol) for 1 h each. After dialysis the viral extract was passed onto a microgranular DEAE-cellulose column (20 ml) equilibrated in buffer C (buffer B containing 0.02% Triton X-100). The column was subsequently eluted with 60 ml of buffer C containing 0.3 M KCI. The eluate, including the flow through, was collected as a pool and dialyzed against three 1-liter portions of buffer C containing 0.05 M KCI for 1 h each. The dialyzed DEAE-cellulose eluate was loaded onto a phosphocellulose column (10 ml) which had been pretreated with bovine serum albumin and equilibrated in buffer C containing 0.05 M KCI. The loaded column was washed with 10 ml of buffer C containing 0.075 M KCI and developed with a linear salt gradient of 100 ml from 0.1 M to 0.5 M KCI in buffer C. Fractions of approximately 1 ml were collected from the column and assayed for DNA polymerase activity. The fractions encompassing the DNA polymerase activity were usually divided into three pools: (i) the leading shoulder; (ii) the center of the activity peak; and (iii) the trailing edge. The pools were first concentrated by dialysis against 30% polyethylene glycol in buffer B containing 0.05 M KCI and 0.1 mM EDTA and then dialyzed against buffer B containing 0.05 M KCI. The enzyme from the peak activity region was used in these studies. If enzyme from either shoulder area was needed, it was again subjected to chromatography on phosphocellulose as described above. The peak fractions from this chromatograph were then pooled, concentrated, and dialyzed as given above before use in the enzyme reactions with 70S RNA. The enzymes were stored at -20 C and were stable for 6 months to 1 year. DNA polymerase assays. Unless stated otherwise, the concentrations ofthe reaction mixture components using AMV 70S RNA as template-primer were (in 50 ,ul): 50 mM Tris-hydrochloride (pH 7.5); 1 mM DTT; 80 mM KCI; 1 mM MnCl2; 5 ,ug of AMV 708 RNA per ml; 80 ,uM each of dATP, dCTP, and dGTP, and 5.6 .M [3H]dTTP (4 x 103 counts/min per pmol). When other 3H-labeled nucleoside triphosphates were either added to the reaction mixture in place of or in conjunction with [3H]dTTP, their concentrations were 6 pM. When the DNA polymerase from AMV was used, the divalent cation was Mg2+ (5 mM) (19). Actinomycin D was added to a concentration of 50 p.g/ml where indicated. Standard 50-Ald assays with synthetic templateprimers were performed as previously described (1). The M-PMV and AMV DNA polymerases were assayed in the presence of 5 mM Mg2+, whereas Mn2+ was the divalent cation used for the other three Triton X-100, 0.02 M DTT, 0.1 mM EDTA, and 0.5 M PNA polymerases. We found these metal ion concenKCI (buffer A) (all percentages are given as vol/vol). trations to be optimal for each enzyme under standThis suspension was subjected to centrifugation at ard conditions with oligo(dT)-poly(A) as template30,000 x g for 30 min. After decanting the superna- primer. The same metal requirements were also tant fluid, the pellet was resuspended in 10 ml of observed for DNA templates, whether natural, denabuffer A and stirred for another 1 h at 0 C. Insoluble tured, or activated. An enzyme unit is defined as the material was removed from this suspension by cen- amount of enzyme catalyzing the incorporation of 1 trifugation at 30,000 x g for 30 min. The two 30,000 picomole of dTMP into acid-insoluble product in 20 x g supernatant fluids plus the initial 100,000 x g min using (dT)10141poly(A) as the template-primer virus supernatant fluid were combined for dialysis under optimal conditions for the enzyme in ques-

Chemicals. Synthetic polynucleotides and oligonucleotides used as template-primers or for hybridization studies were products of Miles Laboratories, Collaborative Research, or P-L Biochemicals. Radioactive deoxyribonucleoside triphosphates were purchased from Schwarz/Mann and New England Nuclear and were flash evaporated before use. Actinomycin D and N-ethylmaleimide (NEM) were obtained from Calbiochem. Bovine pancreatic RNase A and bovine serum albumin were products of Worthington Biochemical Corp. and Armour Pharmaceutical Co., respectively. Cesium sulfate (optical grade) and sucrose (enzyme grade) as well as two purified proteins, ovalbumin and 7S gamma globulin (human) (immunoglobulin G), were obtained from Schwarz/Mann. Polyethylene glycol was purchased from Matheson, Coleman and Bell. Activated DNA was prepared by treating native DNA with pancreatic DNase (both purchased from Sigma Chemical Co.) as previously described (22). Purification of 70S viral RNA. 70S RNA was routinely prepared by Pronase-sodium dodecyl sulfate treatment, phenol-cresol extractions, and precipitation with cetyltrimethylammonium bromide and ethanol-salt as described previously (3). To exclude the possibility that the observed template activity was a consequence of the method of preparation, we also prepared RNA by centrifugation of detergent-treated virions in CsCl (8), which gave a 70S RNA which could be converted to 35S RNA after heat treatment. Both preparations were tested as template-primers. Phosphocellulose preparation. After removing the fines from Whatman phosphocellulose (P-li), the cellulose was washed with 0.1 N HCI, then 0.1 N KOH, and finally with water until the wash was below pH 9. The washed cellulose was then treated with bovine serum albumin (0.1 mg/ml of wet cellulose), which was subsequently removed by 1 M KCI washes. Most of the extraneous KCI was eliminated with three washes with 0.05 M Tris-hydrochloride (pH 7.5), 0.05 M KCI, 0.02% Triton X-100, 1 mM dithiothreitol (DTT), and 20% glycerol. Immediately prior to use, the phosphocellulose was exhaustively washed with the above until the input buffer and eluate were of equal pH. Isolation of the DNA polymerases. The virus suspensions were collected by centrifugation at 100,000 x g for 1 h. The supernatant fluid from the viral pellet was saved for a later stage in the extraction procedure. DNA polymerases were extracted from the virions by a modification of a previously described procedure (1). The pelleted virus (10 mg) was suspended and extracted at 0 C with stirring for 3 h in 20 ml of 0.05 M Tris-hydrochloride (pH 7.9), 0.25%

1568

ABRELL, REITZ, AND GALLO

tion. The GaLV DNA polymerase was about threeto fivefold less efficient in the transcription of 708 RNA than the other enzymes (see below). To test the possibility that the low activity of the latter enzyme might be due to the absence of a small cofactor, such as an oligonucleotide primer, which could be present in the inactivated enzyme solution, M-PMV DNA polymerase was inactivated and added to the reaction mixture containing 70S RNA and the GaLV DNA polymerase. For these assays the phosphocellulose purified M-PMV DNA polymerase was either subjected to 80 C for 5 min and then cooled in ice before assay or was treated with NEM at a final concentration of 0.05 M for 5 min at 37 C, followed by 0.1 M mercaptoethanol to quench unreacted NEM. These treated samples were assayed as described above in the presence of untreated purified GaLV DNA polymerase. Since our earlier purification was performed in the presence of EDTA (1), it was possible that the previous inability of these enzymes to transcribe 70S RNA might be due to the presence of EDTA. Therefore, M-PMV DNA polymerase purified through the phosphocellulose chromatography step was made 0.1 mM with EDTA. The EDTA-treated enzyme was stored at 4 C and assayed with [3H]dGTP substrate at various times over a 10-day period. In the assay mixture the final EDTA concentration was 0.01 mM. Estimations of molecular weights of the DNA polymerases. The density gradient centrifugation procedure of Martin and Ames (13) was used for determining the size of the various DNA polymerases. Samples were prepared for the gradients first by dialysis against 30% polyethylene glycol, 0.05 M Tris-hydrochloride (pH 7.5), 0.35 M KCl, 0.1 mM EDTA, 1 mM DTT, 0.02% Triton X-100, and then against the same buffer without the polyethylene glycol and EDTA (buffer D). A 200-,ul sample of the concentrated protein solution was layered onto 3.5ml gradients of 5 to 20% sucrose or glycerol in buffer D. After centrifugation for 16 h at 150,000 x g at 4 C, the gradients were collected into 33 equal fractions through a puncture in the bottom of the tube. DNA polymerase activity was determined by standard assays. The density of each fraction was estimated from its refractive index. Protein markers (human immunoglobulin G, 160,000 daltons, and ovalbumin, 45,000 daltons) on parallel gradients were detected by their absorbance at 280 nm. Product analysis and hybridization. [3H]DNA products synthesized from the viral 70S RNA template-primer was purified by a previously described procedure which includes a phenol-cresol extraction followed by precipitation with a quaternary ammonium salt, cetyltrimethylammonium bromide (17). The extracted [3H]DNA products were either analyzed directly on a Cs2SO4 gradient as described previously (21) or analyzed after treatment with alkali (0.3 N KOH, 10 min, 95 C) to hydrolyze any RNA which remained from the reaction. The KOH was then neutralized with HCl. For RNA-DNA hybridization experiments, alkali-treated [3H]DNA product containing 1,500 to 3,000 counts/min was annealed with the indicated RNA in 40 Fl of hybridization solution (50% formam-

J. VIROL.

ide, 0.45 M NaCl, 0.045 M sodium citrate, pH 7) for 22 h at 37 C. The amount of hybridization was determined by analysis on Cs2SO4 gradients and is represented by the quantity of [3H]DNA product appearing in the RNA region of the gradient.

RESULTS

Size estimations of viral DNA polymerase. Size estimates for the five viral DNA polymerases were determined by velocity gradient centrifugation. Estimation of molecular size was made with crude enzyme from disrupted virions, with partially purified enzyme preparations and with the purified enzymes. The size estimates were determined by calculations described by Martin and Ames (13), which assume a globular shape for each of the proteins (known markers as well as the polymerases). The three DNA polymerases from the mammalian type C viruses have a similar size (70,000 daltons), whereas the DNA polymerase from M-PMV, a virus with characteristics of both type B and type C viruses, is approximately 110,000 daltons as previously reported (1). The avian type C viral DNA polymerase when purified by our procedure has a size of 160,000 daltons, in agreement with published data (9, 10). The size of each DNA polymerase was independent of the stage of purification, suggesting that these enzymes do not undergo appreciable proteolytic degradation during the

purification procedures. Viral 70S RNA as a template-primer. The data given in Table 1 show the amount of incorporation of each deoxyribonucleotide into an acid-precipitable product which each enzyme catalyzes over a given time course. Using approximately equal units of each enzyme in these reactions, the DNA polymerases from MPMV, MuLVR, and SSV-1 transcribe AMV 70S RNA to an extent approaching that of the AMV enzyme. The GaLV DNA polymerase transcribes 708 RNA to a three- to fivefold lesser extent than the other enzymes used in this study. Despite this low extent of transcription, analysis of the synthesized product shows that it represents a transcript of heteropolymeric regions of the 70S RNA template (see below). Initially, actinomycin D was added to the 70S RNA reaction when the product was to be examined for complementarity with the templateprimer. However, the reagent was later deleted when we found no apparent effect of actinomycin D on the amount or properties of the product. All five polymerases catalyze the incorporation of the four deoxyribonucleotides into DNA product, but, although equal amounts of enzyme were used in the 70S RNA transcription

VOL. 16, 1975

TRANSCRIPTION OF 70S RNA

TABLE 1. Incorporation of deoxyribonucleotide into product with AMV 70S RNA as template' DNApolymerase

Nucleotide incorporated (pmol)

Time (mm)

TMP

dAMP

dCMP

dGMP

0.80 1.57 2.61 3.71

0.64 1.32

0.69 1.07

0.92 1.41

AMV

30 60 120 240

2.00 2.68

1.42 1.80

1.80 3.60

SSV-1

30 60 120 240

0.14 0.48 1.09 1.93

0.15 0.59 1.13 1.85

0.09 0.29 0.73 1.23

0.12 0.35 0.52 1.25

MuLVR

30 60 120 240

0.44 0.94 1.32 2.16

0.22 0.47 0.90 1.50

0.36 0.69 0.77 1.10

0.20 0.46 0.83 1.10

M-PMV

30 60 120 240

0.74 1.19 1.76 2.81

0.61 0.89 1.38 1.76

0.44 0.72 0.97 1.44

0.70 0.94 1.52 1.87

GaLV

30 60 120 240

0.19 0.26 0.33 0.77

0.10 0.20 0.37 0.69

0.03 0.05 0.11 0.29

0.01 0.05 0.16 0.32

a Standard reaction mixtures in total volumes of 125 gl contained one (as indicated) tritiated and the other three unlabeled nucleotides and 0.625 ,g of AMV 70S RNA. [3H]dATP was present at 18 Ci/mmol, [3H]dCTP at 22.4 Ci/mmol, and [3H]dGTP at 12.4 Ci/mmol. At indicated times, aliquots of 25 Al were withdrawn, and their acid-precipitable radioactivity was determined. Concentration of 70S RNA used was well above that necessary for maximal activity. Amounts of different enzymes used were: AMV, 1,200 U; M-PMV, 1,350 U; SSV-1, 1,400 U; GaLV, 1,250 U; and MuLVR, 1,250 U. These enzyme preparations had no terminal addition activity as judged by the lack of reaction with oligo(dT) alone. Also, deletion of one or more of the deoxyribonucleoside triphosphates substantially decreased the amount of DNA product.

reactions, the deoxyribonucleotide ratios in the product were not the same with any two enzymes at any given time point. Each enzyme catalyzes the preferential incorporation of dTMP over the other three nucleotides. However, the DNA polymerase from AMV catalyzes the incorporation of dGMP and dAMP to almost the same extent as dTMP in the 2-h reaction period. The DNA polymerases from SSV-1 and GaLV polymerize almost as much dAMP as dTMP within 2 h. Because little or no detectable transcription of 70S RNA was found in our earlier studies with polymerases purified somewhat differ-

1569

ently (1), we considered the possibility that the earlier procedure included a step which eliminated the ability to transcribe 70S RNA. Therefore, simultaneous purifications of the M-PMV DNA polymerase were done by the present and former procedures, using the same stock of virus for both purifications. The final yield of enzyme was the same for both procedures. We found that equal activity units of these two preparations transcribed AMV 70S RNA to the same degree (data not shown). Both EDTAtreated and untreated portions of these two MPMV enzyme preparations were held at 4 C for 10 days. Aliquots of these solutions were assayed at various times throughout this period with both AMV 70S RNA and oligo(dT)-poly(A) templates using [3H]dGTP and [3H]dTTP, respectively, as the labeled substrates. All preparations maintained the ability to catalyze the transcription of both templates, but at a reduced level of activity. There was little difference between the untreated and EDTA-treated enzyme preparations as measured by these experiments. Thus, the presence of EDTA does not appear to accelerate the deterioration of the enzymatic activity or preferentially inhibit the 70S RNA transcribing activity. M-PMV DNA polymerase which had been completely inactivated by either heat or NEM as described above was combined with GaLV enzyme, which gives a much lower response to AMV 708 RNA than the other viral DNA polymerases tested. The GaLV enzyme was not stimulated by either heat- or NEM-inactivated M-PMV enzyme. These two results suggest, but do not prove, that the difference in transcription is not due to a stimulatory oligonucleotide or small protein associated with the M-PMV enzyme, but absent in the purified GaLV DNA polymerase. The sensitivity of the 70S RNA transcription to the addition of RNase A (10 ,ug/ml) is shown in Table 2. The DNA polymerases from SSV-1 and M-PMV were used to demonstrate the necessity of intact 70S RNA for these reactions. The addition of RNase A completely inhibited the reaction. The time of addition of RNase A did not affect its ability to subsequently inhibit DNA synthesis. Even when this nuclease was added 30 min after the initiation of the reaction, little subsequent reaction occurred. The metal requirements for the M-PMV DNA polymerase are dependent upon the template-primer. As shown in Table 3, reactions with all template-primers except 70S RNA show a marked preference for Mg2+ as the divalent cation. This Mg2+ preference is noted for natural DNA (native, nicked, and denatured) as well as the synthetic DNA-DNA and

1570

ABRELL, REITZ, AND GALLO

TABLE 2. Sensitivity of AMV 70S RNA-directed DNA synthesis to RNasea

J. VIROL.

Complementarity of the DNA product to template RNA. Purified [3H]DNA products from the reactions of purified viral DNA polymSource of DNA RNase (10 Time (min) Product polymerase jg/ml)(coun)in erases and AMV 70S RNA were analyzed by Cs2SO4 density gradient centrifugation both beSSV-1 0 fore and after heat denaturation. [3H]DNA prod30 30 510 uct synthesized by the SSV-1 DNA polymerase 60 1,220 is partially associated with the RNA template to heat treatment, but it migrates toward prior Added SSV-1 0 31 the DNA region after heat treatment (data not 30 87 shown). The [3H]DNA product synthesized by 60 90 the M-PMV enzyme behaves in a similar manner. SSV-1 0 27 Purified alkali-treated [3H]DNA products Added 30 400 60 were annealed to different RNAs and analyzed 500 on Cs2SO4 gradients as described above. In Fig. M-PMV 0 26 1 the [3H]DNA products synthesized by the 30 1,875 DNA polymerases from SSV-1 (Fig. 1A) and M60 2,320 PMV (Fig. 1B) are shown to be homologous (50 to 80%) to the AMV 70S RNA from which they M-PMV Added 0 32 were made. The product prepared with the 30 62 SSV-1 enzyme contained no significant 60 53 [3HIpoly(dT) (Fig. 1A), even though this proda AMV 70S RNA concentration for these assays uct was prepared with [3H]dTTP present in the was 10 j,g/ml; [3H]dTTP was the labeled nucleoside reaction mixture. Annealing the M-PMV enzytriphosphate; 500 U of enzyme activity was used in matic transcript of AMV 70S RNA with MuLVR each set of assays. Aliquots were removed, and the 70S RNA [as a poly(A)-containing control RNA] reaction was quenched at the given time. resulted in no detectable hybridization (Fig. 1B). RNA-DNA template-primers. The Mg2+ concenFigure 2 shows the results of the hybridizatrations is optimal for the indicated templates. tion to AMV 70S RNA of the purified alkaliHowever, for the transcription of 70S RNA treated [3H]DNA products made from AMV 70S Mn2+ is preferred by the M-PMV enzyme. This RNA by the GaLV and MuLVR enzymes. Both Mn2+ requirement for 70S RNA transcription is panels illustrate that the purified product does similar to the requirement shown by the other not hybridize to poly(A). These results are esmammalian viral enzymes used in this study. sentially in agreement with our earlier results The indicated Mn2+ concentration is optimal for with AMV DNA polymerase (18). 70S RNA transcription, not only by the M-PMV enzyme but by the other mammalian enzymes as well. AMV DNA polymerase transcription of TABLE 3. Metal preference of the M-PMV DNA 70S RNA occurs with either divalent cation, polymerasea Mg2+ (at 5 mM) or Mn2+ (at 1 mM), with equivalent results. In contrast to the AMV enzyme, 3H-labeled nucleotide incorpothe three mammalian type C viral DNA polymTemplate-primer ration (pmol) erases require Mn2 . Mg2+ added to the reaction Mg2+ Mn2+ mixture containing Mn2+, 70S RNA, and any of the other viral DNA polymerases is slightly Poly(A) (dT)10-4 225 23 inhibitory. 1.6