Site on the RNA of an Avian Sarcoma Virus at Which Primer Is Bound

JOURNAL OF VIROLOGY, Sept. 1975, p. 553-558 Copyright 0 1975 American Society for Microbiology Vol. 16, No. 3 Printed in U.SA. Site on the RNA of an Avian Sarcoma Virus at Which Primer Is Bound JOHN M. TAYLOR* AND RUDOLFINE ILLMENSEE 7The Institute for Cancer Research, The Fox Chase Cancer Center, Philadelphia, Pennsylvania 19111 Received for publication 9 April 1975 In vitro transcription of the avian tumor virus RNA by RNA-directed DNA polymerase is initiated on a unique cellular 4S RNA. Previous studies have shown that on the average there is one such RNA primer hydrogen bonded to each viral 35S RNA. The present study confirms that finding and demonstrates that, at least for the majority of 35S RNA molecules, the primer is bound at a site close to the 5'-terminus. The genome of the avian RNA tumor virus is actinomycin D (Calbiochem) per ml. After incubation in a 70S RNA aggregate that contains 2 to 4 for 30 min at 37 C the reaction was stopped by the molecules of 35S RNA (12) plus 4S trna (8) addition of sodium dodecyl sulfate (0.5%) and selfdigested Pronase (500 ug/ml) and incubated a further and ribosomal 5S RNA (8). The 35S RNAs are probably identical (1, 2, 16). In vitro the transcription of the 70S RNA into DNA by the viral anol precipitation. 15 min at 37 C. The 70S RNA was isolated by sedimentation (9) and subsequently collected by eth- RNA-directed DNA polymerase is initiated on a Fractionation of partially denatured 70S complexes on oligo(dt)-cellulose. The isolated 70S com- unique 4S RNA species (9) that is probably identical to the cellular tryptophan trna (5, plex was partially denatured by heating at 59 C for 5 17). On the average there is approximately one min in 0.2 ml of 0.01 M Tris-hydrochloride (ph primer binding site per 35S RNA, and in 7.4)-0.01 M EDTA, and then applied to a 0.5-g naturally occurring viral RNA at least 70% of column of oligo(dt)-cellulose (Collaborative Research) in 0.5 M NaCl-0.01 M Tris-hydrochloride, ph these sites are filled (J. M. Taylor, B. Cordell- Stewart, W. Rohde, H. M. Goodman, and J. M. 7.4. The RNA which failed to bind to the column is designated "poly(a) deficient" and the remainder, Bishop, Virology, in press). The present communication describes experiments undertaken hydrochloride, ph 7.4, is designated "poly(a) con- which was subsequently eluted with 0.01 M Tris- to determine the location of the site (or sites) on taining. the 35S RNA at which primer is bound. " For example, in the experiment described in Fig. 1, the input RNA contained 187,000 counts/min of [3H]uridine-labeled RNA and 10,600 counts/min of MATERIALS AND METHODS bound primer specifically labeled with 32P. Of these Virus growth and purification. Secondary cultures of chicken cells (C/O, gs-, Heisdorf and Nelson two fractions were subsequently concentrated by 37 and 16%, respectively, bound to the column. The Laboratories) were infected at low multiplicity with ethanol precipitation. avian sarcoma virus. Either strain B77 clone 9 (obtained from R. Friis) or a nontransforming derivative, and poly(a)-deficient fraction were separately ana- Sedimentation analysis. The poly(a)-containing td-b77 (obtained from P. Vogt), was used. The lyzed by sedimentation into gradients of 15 to 30% cells were grown in a large roller bottle (1585 cm2 of sucrose containing 0.01 M Tris-hydrochloride, ph 7.4; surface area) with 25 ml of Ham F-10 medium containing 4% calf serum, 10% tryptose phosphate Centrifugation was for 15 h at 28,000 rpm and 22 C in 0.01 M EDTA; and 0.01% sodium dodecyl sulfate. broth, and 20 gci of [H1 Juridine (27.8 Ci/mmol, the SW40 rotor of a Spinco L5.50. Aliquots of the New England Nuclear) per ml. The radioactive gradient fractions were precipitated with 10% trichloroacetic acid in the presence of 80,gg of calf thymus virus was purified (4) from a single harvest (24 h), subsequently pelleted, and resuspended in 0.2 ml DNA carrier, collected onto glass fiber filters, and of 0.1 M NaCl-0.01 M Tris-hydrochloride (ph counted in a Beckman scintillation spectrometer. The 7.4)-0.001 M EDTA. 3H-labeled L cell rrna marker was a gift of Dawn Preparation of 70S complexes between template Kelley. and tagged primer RNA. To the resuspended virus Polyacrylamide gel analysis. RNA fractions were was added 0.8 ml of a reaction mixture containing: 0.1 analyzed by electrophoresis into cylindrical gels (0.6 M Tris-hydrochloride, ph 8.1; 0.01 M MgCl2; 2% by 8 cm) of 2.1% acrylamide cross-linked with 0.28% mercaptoethanol; 0.01% Triton X-100; 10-4 M dgtp; (vol/vol) ethylene diacrylate. Bromophenol blue was 10-' M TTP; 0.86 x 10-6 M [a-82p]datp (115.1 added to the sample prior to electrophoresis. The Ci/mmol, New England Nuclear); and 100 Ag of buffer present both in the gel and the electrode 553

554 TAYLOR AND ILLMENSEE chambers was 0.04 M Tris-0.004 M sodium acetate-0.001 M EDTA, adjusted to ph 7.4 with acetic acid. After electrophoresis, the gel was removed, frozen, and fractionated with an apparatus consisting of stacked razor blades (Hoefer Scientific). The gel slices were placed in scintillation vials and solubilized by treatment for 30 min at room temperature with 0.4 ml of 20% (vol/vol) ammonia. After the addition of 10 ml of cocktail (40% ethylene glycol monoethyl ether, 60% toluene, 2.4% Liquifluor [New England Nuclear]), the samples were counted in a Beckman scintillation spectrometer. It is important to note that RNA aggregation has been detected in polyacrylamide gel analysis of poly(a)-containing RNA and results particularly in significant accumulation of RNA at the origin. This can be avoided by heating the RNA (in 0.01 M EDTA-0.01 M Tris, ph 7.4) at 59 C for 5 min prior to electrophoresis. RESULTS An endogenous reaction of detergent-disrupted avian sarcoma virus (strain B77, clone 9) was performed in the presence of dgtp, dttp, and [a_32p]datp, but with dctp omitted. As previously documented, this reaction specifically labels the 4S RNA primer in the 70S RNA complex with a short oligodeoxynucleotide (20). After this, the 70S RNA complex was isolated by virtue of its sedimentation velocity and heated for 5 min at 59 C in a low salt buffer. Again as previously documented, this heating does not effect the binding between the 35S RNA and that 4S RNA which functions as primer but dissociates the 70S complex into 35S subunits and removes all 5S RNA and nonprimer 4S RNA (4, 6). We have sought to locate the site on the partially denatured 35S RNA at which the specifically labeled primer is bound. To do this we have made use of the observations that the majority of viral 35S RNA contains a sequence of about 180 residues of adenylic acid (13), and that this poly(a) is located at the 3'-terminus of the 35S RNA (19, 21). The rationale was to obtain fragments of partially denatured 35S RNA that contain an intact 3'-terminus of poly(a), using a column of oligo(dt)-cellulose (14), and examine the relative distribution of 82P-labeled primer amongst different size classes of these fragments and possibly deduce therefrom the site (or sites) at which primer is bound. As it happens, fragmentation of 35S RNA can occur prior to the harvesting of the virus and during purification (7). We have exploited this in the following experiment. The purified virus used was obtained from cells that had been uniformly labeled with [3H Juridine. ic 28S 4000 150 )-- 1b20 30 FRACTION NUMBER -t a- CM FIG. 1. Sedimentation analysis of partially denatured viral RNA. The poly(a)-containing RNA, uniformly labeled with ['H]uridine (a) to which is bound primer molecules specifically labeled with 32p (Q), was prepared and sedimented as described. On a separate gradient was sedimented the poly(a)-deficient RNA with 32P-labeled primer (A) together with an excess of 'H-labeled L cell 18 and 28S rrna (arrows). The results of the analyses of the two gradients are superimposed. The virus used in this experiment was strain B77, clone 9. From the sedimentation analysis shown in Fig. 1, it can be seen that the 'H-labeled poly(a)- containing RNA fragments range in sedimentation values up to about 35S. However, the majority of the 3"P-labeled 4S RNA bound to these fragments is associated with large 28 to 35S species. Our interpretation of this finding is that the majority of primer binding sites are located at or near the 5'-terminus of the 35S RNA. Also shown in Fig. 1 is the sedimentation analysis of the poly(a)-deficient RNA fragments. The distribution of bound primer among these fragments is consistent with the abovementioned interpretation, but other interpretations are also possible. Table 1 summarizes the relative content of 32P and 'H in the various nucleic acid fractions obtained during the course of the experiment. The ratio 32P/3H should be directly proportional to the number of primer molecules per unit mass of RNA. Thus, relative to 70S RNA the poly(a) -deficient RNA has more primer and the poly(a)-containing RNA has less primer. However, poly(a)-containing RNA with a sedimentation value of 35S has the same ratio as the 70S. Again these data are in accord with the interpretation that the one primer binding site per 35S RNA is located at or near the 5'-terminus. Also shown in Table 1 A is the 32P/3H ratio for purified primer from this experiment. From this value, the size of 35S RNA (9,000 nucleotides 18S A J. VIROL.

VOL. 16, 1975 [12D and primer (75 nucleotides [6]), and the 32P/3H for 70S RNA, the number of primer molecules bound per 35S RNA can be calculated. The mean of four experiments was 1.04 0.29. This value is in agreement with previous estimations (Taylor et al., Virology, in press; 9, 17) Ṫo better estimate the location of the primer binding site with respect to the 5'-terminus, we have analyzed the 35S RNA fragments on gels of 2.1% polyacrylamide (Fig. 2). This particular experiment was done with a different virus stock (strain B77, transformation defective). The markers of 18 and 28S rrna were run on a separate gel. From their observed mobility and the formula mol wt = 1,550S2-1 of Spirin (18), the size scale as indicated was deduced. The AVIAN SARCOMA VIRUS PRIMER 555 size distribution of the poly(a)-containing RNA (Fig. 2a) is such that 70% of the 32p is associated with species that are larger than 90% of the modal size. The absolute size values calculated could easily be in error in that they involve extrapolations of the Spirin formula. The size distribution of the poly(a)-deficient RNA is shown in Fig. 2b. The distribution of 3"P-labeled primer amongst the [3H ]-labeled RNA is almost random, with a small but significantly higher frequency among smaller RNAs. The studies described above show that amongst poly(a)-containing RNA the labeled primer is primarily associated with the largest RNA species. From this an interpretation has been adopted which says that the primer must be bound at or near the 5'-terminus. To further TABLE 1. Analysis of radioactivity in aliquots of RNA fractions isolated during experiment described in Fig. 1 Aliquot of RNA fraction 'H (counts/min) 3"P (counts/min) 32P/3H Total 70S RNA 4,580 265 0.058 4 0.003a Total poly(a)-deficient RNA 7,000 540 0.077 ± 0.003 Total poly(a)-containing RNA 4,100 103 0.025 ± 0.002 35S poly(a)-containing RNAb 2,940 163 0.055 ± 0.003 Purified primerc 37 243 6.6 ± 1.5 a This error represents the standard deviation arising from counting errors of both 3H and "2p. "Fractions 6 and 7 in Fig. 1. cprimer was released from partially denatured 12 to 35S RNA by complete denaturation at 70 C as previously described (Taylor et al., in press; 5). EIC. 1200 600 SIZE ( x 103 nucleotides) E -4 CM MIGRATION (cm) FIG. 2. Polyacrylamide gel electrophoresis of partially denatured viral RNA. Viral RNA was uniformly labeled with ['HJuridine (0), and bound primer molecules were specifically labeled with 32P (0). Electrophoresis was for 3.5 h at 4 ma/gel, with other details as described. Analyses of poly(a)-containing and poly(a)-deficient species are as shown in (a) and (b), respectively. The 'H-labeled 18 and 28S rrna (arrows) were run in a separate gel. Using these markers and the formula of Spirin (18), the uppermost scale of RNA lengths was deduced. The virus used in this experiment was a transformation-defective variant of strain B77.

556 TAYLOR AND ILLMENSEE substantiate this interpretation we have also examined the effect of fragmentation induced after isolation of the poly(a)-containing RNA. Limited fragmentation was produced by heating at 45 C in a buffer at ph 10.0. At various times aliquots were removed and assayed for poly(a) content by a membrane filter (Millipore)-binding technique (15). The results (Fig. 3) indicate that with respect to the 3H-labeled RNA the 32P-labeled bound primer can more readily lose its poly(a)-containing property. This is again consistent with the primer being bound at or near the 5'-terminus of the 35S RNA. A more detailed analysis was made of poly(a)-containing RNA that had been fragmented for 9 min as described above. Figures 4a and b show the polyacrylamide gel analyses of this RNA before and after fragmentation. Clearly fragmentation has converted 32P-labeled primer and 'H-labeled RNA into faster migrating species. (Primer binding itself is not disrupted by this procedure since no primer is seen to migrate in the 4S region of the gel after fragmentation.) A larger sample of RNA after fragmentation was separated a second time on oligo(dt)-cellulose into poly(a)-containing and poly(a)-deficient fractions which were subsequently analyzed on polyacrylamide gels. In Fig. 4c, it can be seen that due to the fragmentation, the residual 3H-labeled poly(a)- containing RNA contains an increased amount of lower-molecular-weight species with respect to Fig. 4a. The 32P-labeled bound primer, however, is found only among the undegraded RNA species. It appears that the first nick, no matter - c s z 0.5 LL - 0.5 C 0 '0N 0 15 TIME (min) FIG. 3. Induced fragmentation of poly(a)-containing partially denatured viral RNA. The materials used in this experiment are as in the legend to Fig. 2. Fragmentation was induced by heating at 45 C in 0.01 M Tris-0.01 M EDTA-0.025 M Na2CO3 (ph 10.0). At various times aliquots were removed and assayed for ability to bind to membrane filters (Millipore) (15). Data are expressed with respect to zero time, at which 100%o of both the uniformly-labeled 'H RNA (O) and bound primer molecules specifically labeled with "2P (0) bound to the filter. where it occurs, removes the primer from the poly(a)-containing fragments. 'H-labeled RNA of the newly derived poly(a)-deficient species covers a range of up to almost full size, and the 32P-labeled bound primer is almost randomly distributed amongst these (Fig. 4d). EIC.) 60 Cd) I I I b1i I~~~~ I J. VIROL. 100 n 0 4 0 8 MIGRATION (cm) FIG. 4. Polyacrylamide gel electrophoresis of poly(a)-containing partially denatured viral RNA subjected to limited fragmentation. The materials used in this experiment are as in the legend to Fig. 2, with the poly(a)-containing RNA uniformly labeled with 'H (0) and the bound primer molecules specifically labeled with "2P (0). Fragmentation was induced as described in the legend to Fig. 3 by heating isolated poly(a)-containing RNA for 9 min at 45 C in 0.01 M Tris-0.01 M EDTA-0.025 M Na,CO, (ph 10.0). The majority of the treated RNA was passaged over oligo(dt)-cellulose to separate poly(a)-containing and poly(a)-deficient species. The electrophoretic analyses shown are of: (a) total poly(a)-containing RNA prior to fragmentation; (b) total after treatment; (c) poly(a)-containing component isolated after treatment; (d) poly(a)-deficient component isolated after treatment. Electrophoresis was for 1.75 h at 4 ma/gel. The 'H-labeled 18 and 28S rrna (arrows) were run in a separate gel. 30 E cs. tc>

VOL. 16, 1975 DISCUSSION This paper relates our experience with the primer binding site on avian sarcoma virus genomes. We calculate that on the average there is one primer bound per 35S RNA in native 70S RNA, and also one primer per poly(a)-containing partially denatured 35S RNA. The majority of primer molecules are bound at or near the 5'-terminus of the RNA. We have obtained similar results with both a transforming and a nontransforming isolate of strain B77 avian sarcoma virus. The design of the present experiments is such that we cannot locate the primer site more precisely than by saying that at least 70% of the primers bound are on molecules that are larger than 90% of the modal size, assuming that this size represents intact RNA. There is a smaller but significant fraction (as high as 30%) of primer molecules that seems to be bound to smaller species (about 20 to 30S). It can be seen from Fig. 4c relative to Fig. 4a that the proportion of primer bound to poly(a)-containing RNA of this size neither increases nor decreases as a consequence of induced fragmentation of total poly(a)-containing RNA. One interpretation would be that there exists some poly(a)-containing RNA species, with primer bound at or near their 5'-terminus, that are only 20 to 30S in size. It is interesting to speculate regarding the significance of the present results in terms of the process of in vivo viral replication. It must first be made quite clear that as yet there is evidence neither for nor against the proposal that the RNA primer, as studied in vitro, has a similar role in vivo. If this proposal were true, then because the primer is at or near the 5'-terminus, and since DNA synthesis proceeds toward the 5'-terminus of the template, it is necessary that during DNA synthesis the 5'-terminus of the RNA must come into juxtaposition with the 3'-terminus of the same 358 RNA or possibly with the 3T-terminus of another 35S RNA. The latter possibility is interesting for several reasons, one of which is that it suggests a role for the existence of the observed 70S RNA aggregate of 35S RNAs. Implicit in the former possibility is the formation of a circular intermediate in DNA transcription. Such an event might be the initial event in the formation of the double-stranded DNA circles that have been recently observed in vivo (10, 11) and which seem to be immediate precursors to integrated proviral DNA (11). ACKNOWLEDGMENTS We are pleased to thank Jesse Summers, Bill Mason, and AVIAN SARCOMA VIRUS PRIMER 557 Tom Parsons for helpful discussions. This work was supported by grant VC-155 from the American Cancer Society, by Public Health Service grants CA- 06927 from the National Cancer Institute and RR-05539 from the Research Resources Branch, and by an Appropriation from the Commonwealth of Pennsylvania. LITERATURE CITED 1. Beemon, K., P. Duesberg, and P. Vogt. 1974. Evidence for crossing-over between avian tumor viruses based on analysis of viral RNAs. Proc. Natl. Acad. Sci. U.S.A. 71:4254-4258. 2. Billetter, M. A., J. T. Parsons, and J. M. Coffin. 1974. The nucleotide sequence complexity of avian tumor virus RNA. Proc. Natl. Acad. Sci. U.S.A. 71:3560-3564. 3. Bishop, J. M., C. T. Deng, A. J. Faras, H. M. Goodman, W. E. Levinson, J. M. Taylor, and H. E. Varmus. 1973. Transcription of the Rous sarcoma virus genome by RNA-directed DNA polymerase, p. 15-31. In C. F. Fox (ed.), Virus research, proceedings of second ICN- UCLA symposium. Academic Press Inc., New York. 4. Bishop, J. M., W. Levinson, N. Quintrell, L. Fanshier, and J. Jackson. 1970. The low molecular weight RNA's of Rous sarcoma virus. I. The 4S RNA. Virology 42:182-195. 5. Cordell-Stewart, B., J. M. Taylor, W. Rohde, H. M. Goodman, and J. M. Bishop. 1974. Transfer RNAs of Rous sarcoma virus and the initiation of DNA synthesis of viral reverse transcriptase, p. 117-135. In Viral transformation and endogenous viruses. Academic Press Inc., New York 6. Dahlberg, J. E., R. C. Sawyer, J. M. Taylor, A. J. Faras, W. E. Levinson, H. M. Goodman, and J. M. Bishop. 1974. Transcription of DNA from the 70S RNA of Rous sarcoma virus. I. Identification of a specific 4S RNA which serves as primer. J. Virol. 13:1126-1133. 7. Duesberg, P. H., and R. D. Cardiff. 1968. Structural relationships between the RNA of mammary tumor virus and those of other RNA tumor viruses. Virology 36:696-700. 8. Faras, A. J., A. C. Garapin, W. E. Levinson, J. M. Bishop, and H. M. Goodman. 1973. Characterization of low-molecular-weight RNAs associated with the 70S RNA of Rous sarcoma virus. J. Virol. 12:334-342. 9. Faras, A. J., J. M. Taylor, W. Levinson, H. Goodman, and J. M. Bishop. 1973. RNA-directed DNA polymerase of Rous sarcoma virus: initiation of synthesis with 70S viral RNA as template. J. Mol. Biol. 79:163-183. 10. Gianni, A. M., D. Smotkin, and R. A. Weinberg. 1975. Murine leukemia virus: detection of unintegrated double-stranded DNA forms of the provirus. Proc. Natl. Acad. Sci U.S.A. 77:447-451. 11. Guntaka, R. V., B. W. J. Mahy, J. M. Bishop, and H. E. Varmus. 1975. Ethidium bromide inhibits appearance of closed circular viral DNA and integration of virusspecific DNA in duck cells infected by avian sarcoma virus. Nature (London) 253:507-511. 12. Jacobson, A. B., and P. A. Bromley. 1975. Molecular weight of RNA subunits of Rous sarcoma virus determined by electron microscopy. J. Virol. 15:161-166. 13. Lai, M. M. C., and P. H. Duesberg. 1972. Adenylic acid-rich sequences in RNAs of Rous sarcoma virus and Rauscher mouse leukemia virus. Nature (London) 235:383-386. 14. McLaughlin, C. S., J. R. Warner, M. Edmonds, H. Nakazato, and M. H. Vaughan. 1973. Polyadenylic acid sequences in yeast messenger ribonucleic acid. J. Biol. Chem. 248:1466-1471. 15. Mendecki, J., S. Y. Lee, and G. Brawerman. 1972. Characteristics of the polyadenylic acid segment asso-

558 TAYLOR AND ILLMENSEE ciated with messenger ribonucleic acid in mouse sarcoma 180 cells. Biochemistry 11:792-798. 16. Quade, K., R. E. Smith, and J. L. Nichols. 1974. Evidence for common nucleotide sequences in the RNA subunits comprising Rous sarcoma virus 70S RNA. Virology 61:287-291. 17. Sawyer, R. C., F. Harada, and J. E. Dahlberg. 1974. Virion-associated RNA primer for Rous sarcoma virus DNA synthesis: isolation from uninfected cells. J. Virol. 13:1302-1311. 18. Spirin, A. S. 1963. Some problems concerning the macromolecular structure of ribonucleic acids. Prog. Nucleic Acid Res. Mol. Biol. 1:301-345. J. VIROL. 19. Stephenson, M. L., J. F. Scott, and P. Zamecnik. 1973. Evidence that the polyadenylic acid segments of "35S" RNA of avian myeloblastosis virus is located at the 3-OH terminus. Biochem. Biophys. Res. Commun. 55;8-16. 20. Taylor, J. M., D. E. Garfin, W. E. Levinson, J. M. Bishop, and H. M. Goodman. 1974. Tumor virus ribonucleic acid directed deoxyribonucleic acid synthesis: nucleotide sequence at the 5'-terminus of nascent deoxyribonucleic acid. Biochemistry 13:3159-3163. 21. Wang, L-H, and P. Duesberg. 1974. Properties and location of poly(a) in Rous sarcoma virus RNA. J. Virol. 14:1515-1529. Downloaded from http://jvi.asm.org/ on September 4, 2018 by guest