Proc. Natl. Acad. Sci. USA Vol. 74, No. 3, pp. 989-993, March 1977 Biochemistry

Rous sarcoma virus genome is terminally redundant: The 5' sequence (reverse transcription/DNA sequence/tumor virus replication/ribosome binding sites/RNA dimers) WILLIAM A. HASELTINE*, ALLAN M. MAXAMt*, AND WALTER GILBERTt *Sidney Farber Cancer Institute, Harvard Medical School, Boston, Massachusetts 02115; and tDepartment of Biochemistry and Molecular Biology, Harvard University, Cambridge, Massachusetts 02138 Contributed by Walter Gilbert, January 6, 1977

ABSTRACT When Rous sarcoma virus RNA is transcribed With RSV, under most in vitro conditions, about 20% of all into DNA by the reverse transcriptase, a tRNA primer is elon- initiation events terminate at a specific site about 100 bases from gated into DNA. The primer is near the 5' end of the virus ge- nome; the -first major DNA made is a "run-off" product ex- the primer (9, 10, 12, 15). This product, strong-stop DNA, ob- tending 101 bases from the primer to the 5' end of the template. viously is a candidate for the run-off product: a DNA fragment We have studied this DNA molecule to determine the sequence running from the primer to the 5' end of the template. We have of the first 101 bases at the 5' end of the Rous sarcoma virus sequenced this strong-stop DNA and argue that it is, in fact, a genome (Prague strain, subgroup C). Twenty-one bases at the transcript of the 101 nucleotides at the 5' end of the RSV ge- extreme 5' end are also at the 3' end of the virus genome (see D. nome. E. Schwartz, P. C. Zamecnik, and H. L. Weith, this issue, pp. 994-998), and thus this virus is terminally redundant. The exis- tence of this sequence repetition immediately suggests mech- MATERIALS AND METHODS anisms by which the growing DNA copy can jump from the 5' Rous Sarcoma Virus. RSV (Prague strain, subgroup C) was end to a 3' end of the template and become circular. The se- quence also displays a possible ribosome binding site and obtained from University Laboratories via the Office of Pro- enough secondary structure to permit a possible 5'-5' linkage grams and Logistics of the National Cancer Institute. of viral RNA molecules. DNA. This was synthesized in a 10-ml reaction containing RSV virion protein at 2 mg/ml, 0.05 M Tris-HCI (pH 8.3), 0.06 During infection, RNA tumor viruses make a DNA copy of M KCI, 0.02% (wt/vol) Nonidet P40,0.006 M MgCl2, 0.005 M their genome. This DNA molecule becomes a covalently closed dithiothreitol, 1 mM each dATP, dCTP,- and dGTP, 0.2 mM double-stranded circle and then integrates into the host DNA [3H]dTTP (specific activity, 2 Ci/mmol), and actinomycin D before the life cycle of the virus can continue. The virion carries at 100 ,ug/ml. After a 90-min incubation at 430, we added 10 into the cell two copies of a linear viral RNA, several copies of ml of 0.2 M NaOAc/0.01 M Tris-HCl, pH 7.5/1 mM EDTA an RNA-dependent DNA polymerase (the reverse transcrip- (buffer A) and 20 ml of a 1:1 (vol/vol) mixture of phenol/ tase), and, annealed to the RNA, a primer for transcription, a chloroform. After vigorous mixing for 1 min, centrifugation host tRNA molecule [tRNAtrP for Rous sarcoma virus (RSV) (1)]. for 5 min at 25° and 20,000 X g separated the phases. The The reverse transcriptase can elongate the primer tRNA to phenol/chloroform was reextracted with buffer A; the aqueous make a single-stranded DNA copy of the viral RNA, as an phases were combined and reextracted with an equal volume RNA-DNA hybrid. An RNase H activity of the reverse tran- of chloroform. Then, 2 volumes of ethanol was added, and the scriptase can digest the RNA template in such an RNA-DNA nucleic acids were collected by centrifugation at 200 for 5 hr hybrid to reveal a template for the synthesis of a second strand. at 30,000 X g. The pellet was resuspended in water, 2 volumes (For reviews see refs. 2-5.) But the original template viral RNA of ethanol was added; and the suspension was centrifuged as is linear. How then is a circular DNA molecule formed? before. In vitro, when reverse transcriptase copies the RNA of RSV, The final pellet was dissolved in 50 ,ul of 0.3 M NaOH and DNA products of various discrete lengths are synthesized. At held at 370 for 15 hr to hydrolyze primer and template RNA; sufficient deoxynucloeoside triphosphate concentrations, the then the DNA was precipitated with 100 ,l of 1.0 M NaOAc longest of these chains covers a major portion of the genome and 500 ul of 95% ethanol by chilling in a Dry Ice-ethanol bath [1000-4000 (6-9) up to full length, 10,00 nucleotides (E. and collected by centrifuging at 12,000 X g for 10 min. The Rothenberg and D. Baltimore, personal communication)]. A pellet was rinsed with 95% ethanol, dried, and redissolved in tRNA primer initiates these chains (1, 10); the first eight 25 Aul of distilled H20. deoxynucleotides added to its 3' end are AATGAAGC (9, 11, DNA Phosphorylation. In this procedure (16) we used [,y- 12); and the longer fragments are extensions of this sequence 32P]ATP (1200 Ci/mmol). After the reaction with polynucle- (9, 12). Thus, in vitro, the reverse transcriptase initiates DNA otide kinase, 2 M NH4OAc and tRNA carrier were added, and synthesis at a unique site and terminates at a series of defined the DNA and RNA were precipitated with 3 volumes of ethanol sites along the genome. at -70° and sedimented. The ethanol-rinsed and dried pellet This would be a natural, expected behavior, except for the was dissolved in 25 , of 5 mM Tris-borate, pH 8.3/0.1 mM curious finding by Taylor and Illmensee (13) and Staskus et al. EDTA/10% (vol/vol) glycerol/0.02% xylene cyanol/0.02% (14) that the tRNA primer is bound very near the 5' end of the bromophenol blue, layered on a slab gel polymerized from 8% template RNA. Thus, as the reverse transcriptase copies (in the acrylamide/0.27% bisacrylamide/50 mM Tris-borate, pH 8.3/1 5' to 3' direction), it soon will come to the 5' end of the template mM EDTA, and subjected to electrophoresis at 300 V (regu- and fall off. How is this to be reconciled with the existence of lated) and 20 mA (average). After autoradiography, labeled ordered long DNA transcripts? DNA was located, extracted, and ethanol-precipitated (16). Nucleotide Sequence Analysis. For analysis, 5'-32P-labeled Abbreviation: RSV, Rous sarcoma virus. strong-stop DNA was partially cleaved by the four base-specific f To whom reprint requests should be addressed. reactions described by Maxam and Gilbert (16). For cleavage 989 Downloaded by guest on September 26, 2021 990 Biochemistry: Haseltine et al. Proc. Natl. Acad. Sci. USA 74 (1977)

A>G G C C+T ..__, ___I

A>G G C C+T

' *1AMW

Sllss T T. G T G G A A>G G C C+T C T T A_ A _T C T T A w A C C

C . G FIw. 1. Isolation of 5'-:32P-labeled RSV endogenous reverse T transcripts. One-fifth of the DNA synthesized in endogenous reactions T containing 20 mg of virion protein was phosphorylated with :3P (1200 Ci/mmol) as described in Materials and Methods, layered on an 8% T T polyacrylamide slab gel, and subjected to electrophoresis until the ,* bromphenol blue marker dye had moved 15 cm. The gel was then A+--%G ..-k exposed to Kodak XR-5 x-ray film for 30 min, and the labeled A;9Ft.:3 .14all"". 1. strong-stop (SS) DNA was extracted for sequence analysis. FIG. 2. Base-specific partial cleavage products 10-101 from at adenine (and to some extent, guanine) the DNA was treated end-labeled strong-stop DNA separated by size on a polyacrylamide with 50 mM dimethyl sulfate at 20° for 20 min, the methylated gel. Equal portions of [5'-92P]DNA from gel region SS, see Fig. 1, were adenines were preferentially released at pH 1, and the strands subjected to four limited chemical degradations (specificities: A>G, were broken at depurination sites by treatment with 0.1 M G, C, and C+T) and electrophoresed in parallel on a denaturing slab from the four reactions at 900 for 30 min. Two cleavages at guanine were used polyacrylamide gel. Degradation products NaOH were repeatedly layered on the gel during electrophoresis to expand after the same reaction with dimethyl sulfate: either heat different regions of the sequence: at 0 hr in the center, at 13 hr on the depurination for a strong G/weak A pattern after NaOH left, and at 24 hr on the right, after which electrophoresis was con- cleavage, or 1.0 M piperidine at 90° for 30 min for a G-specific tinued for another 12 hr. The sequence is derived by beginning with cleavage. Reaction with 18 M hydrazine in 2 M NaCl at 20° for the smallest partial product at the lower right, a T cleavage product, 30 min, followed by displacement and strand scission with 0.5 reading upward, and then switching to the left and finally to the whenever bands are not well resolved. M piperidine at 900 for 30 min, cleaved at cytosine. The same middle patterns reaction with hydrazine alone for only 20 min, followed by piperidine, cleaved at both cytosine and thymine. The end- products with polynucleotide kinase and [y-32P]ATP and re- labeled partial cleavage products from the four reactions were solved the labeled molecules on a preparative polyacrylamide fractionated on a 0.15 X 33 X 40 cm slab gel polymerized from gel. Fig. 1 shows that about 80% of the synthesis product was 20% acrylamide/0.67% bisacrylamide/7 M urea/50 mM unique, strong-stop DNA. Tris-borate, pH 8.3/1 mM EDTA as described (16). The We excised strong-stop DNA from the gel and sequenced it wrapped gel on one glass plate was then exposed to Kodak XR-5 by partial, base-specific, chemical cleavages and separation of x-ray film at -20°. the end-labeled products by size on polyacrylamide gels (16). Figs. 2 and 3 show autoradiographs of the sequencing gels. Fig. RESULTS 3 displays the first 10 partial cleavage products beginning from We made strong-stop DNA [a DNA product about 100 bases the labeled end, and Fig. 2, displays the next 90. Taken together, long covalently linked to a 75-base tRNA primer (1, 9, 10)] in the four patterns in the two figures determine, in order, 101 a reaction mixture containing detergent-disrupted virions, deoxynucleotides added to the primer in strong-stop DNA, with actinomycin D [to suppress the formation of double-stranded the exception of the very first one. That one is known to be A DNA (17)], and high concentrations of deoxynucleoside tri- from analysis by others (9, 11, 12, 18) who found the first eight phosphates. At the end of the reaction, after a phenol/chloro- deoxynucleotides to be AATGAAGC, overlapping with our first form extraction, alkaline hydrolysis of the primer released a free seven, NATGAAGC .... The sequence of all 101 deoxynu- DNA 5'-hydroxyl. We then labeled the 5' ends of the DNA cleotides in strong-stop DNA, joined to the last 17 ribonucleo- Downloaded by guest on September 26, 2021 Biochemistry: Haseltine et al. Proc. Natl. Acad. Sci. USA 74 (1977) 991

G A CA CCA CU UA CCA U U U UA CCGpppG T -AAAACACCACUU ACCA U U U UACCG T T C TGT GG TGA A T GG TA AA A TGG COH C ACACCAC U UJACC AU IUUUACCGpppG AAAA C A C C A C U U A C C A U UU U A C C GpppG G ~AAAACACCACUUACCAUUUUACCG A A TGTGGTGAATGGTAAAATGGC A AAACACCACU UAC CA U U UUACCG G

FIG. 5. Terminal redundancy and possible replication of the T RSV genome. Top. The first 21 nucleotides in RSV viral RNA are identical in sequence and polarity with the last 21 (see Fig. 4, and Schwartz, Zamecnik, arrd& Weith, this issue, pp. 994-998). Middle. Strong-stop DNA complementary to the 5'end of viral RNA has been synthesized. Bottom. The redundant portion ofthis template has been A displaced, allowing the same sequence from the 3' end to hydrogen bond with the DNA. This allows the strong-stop reverse transcript, synthesized from the extreme 5' end of the viral teiplate, to serve as FIG. 3. Base-specific cleavage products 2-14 from end-labeled a primer for more extensive DNA synthesis from the extreme 3' end. strong-stop DNA separated by size on a polyacrylamide gel. Equal The figure arbitrarily shows all of this occurring on one template; portions of [5'-32P]DNA were subjected to chemical degradations of synthesis could just as well continue on the 3' end of another mole- specificity A>G, G

... TGGCoH sequence at the 3' end of the DNA product is DISCUSSION complementary to the RNA sequence right before the cap. The Strong-Stop DNA Is a Run-Off Product from the 5' End of next earliest C in the DNA sequence (corresponding to the next. the Template. The 5' end of RSV RNA is capped by the RNase T1 site in the RNA) is 23 bases earlier, at DNA position structure 78, and all of the unordered fragments shown in sequence 2 can be identified in the complementary RNA sequence predicted Gm7pppG2'omC ... (21, 22) [1] by our DNA sequence. We conclude that the 101 base-long strong-stop DNA terminates with a C complementary to the which can be released by RNase A. Although the sequence at G that is 2'0-methylated in the cap structure and that this DNA this end is not known, RNase T1 releases a 25-base-long frag- is a run-off product synthesized out to the last 5' base of the ment carrying the cap, whose partial sequence is template. This DNA sequence therefore determines the first Gm7pppG2'OmCCA[UUUUA,U(C,U)A,2CCA,CA]UUG (23). 101 bases at the 5' end of the viral RNA. In addition, because [2] 16 contiguous bases in the tRNA primer are protected by the viral RNA against RNase digestion (19), we can infer the RNA We deduce that the strong-stop DNA is complementary to the sequence out to base 118, as shown in Fig. 4. However, Cordell

I 10 20 30 40 50 60 70 80 90 10 U A CAAATGAAGCCT CTCTTCATTCAI GGTGTTCGCAATCGTTAGGGAATCGACGGTCCAGCCATCAACCCAGGTGCACACCAATGTGGTGAATGGTAAAATGGCO,,

AGUGCAG;CCCCAGUGGXUUACUUCGGAAGAGGJAAGUCCACAAGCGUJUAGCAAUCCCUUAGCUGC CAGGUCGGUA6tUIGGGUCCAC6GUGt)UtUACACCACUUACCAUUUUACCG IBIWG 110 160 9 80 70 6b 50 40 30 20 10

Hinf TaqI EcoRIl HphI TCACG GGGTCACCXAATGAAGCCTTCTGCTrCATTCAGGTGTTCGCAATCGTTAGGG AAT CACGGTCCAGCCATC GTGCACACCAATGTGGTGAATGGTAA ATGGC AGTOC GCCGXG):TTACTTCGG OAGAAGTCCACAAGCGTTAGCAATCCC WG CAGGTCGGTAGTlp AGTGTGGTTACACCACTTACCAT~TTACCG FIG. 4. Nucleotide sequence ofthe strong-stop reverse transcript and the 5' end ofthe RSV genome (Prague C). Top. That portion (underlined) of the host tRNA primer that specifically binds to the viral template (19) and the first 101 deoxynucleotides added to its 3' end by the reverse transcriptase. The sequence of this strong-stop DNA was derived from the band patterns of Fig. 2 and 3. Middle. The complementary viral RNA sequence inferred from the DNA transcripts, the primer binding sequence (19), and the cap structure (21, 22). Bottom. The putative proviral DNA sequence encompassing this region, with potential cleavage sites for restriction endonucleases (36, 37). Downloaded by guest on September 26, 2021 992 Biochemistry: Haseltine et al. Proc. Natl. Acad. Sci. USA 74 (1977)

met lys gin lys ala ser m76(5')ppp (5')GC...56 NT...CCCUAACBAUUGCGAACACCUGAAUGAAGCAGAAGGCUUCA ..... lII 11lIt HOAUUACUAO..... FIG. 6. Possible translation-initiation site near the 5' end of RSV RNA. The overlined AUG triplet is 83 bases from the cap structure. Hy- pothetical base-pairing between this RNA and the 3' end of 18S ribosomal RNA (26, 27) and a predicted NH--terminal sequence for the gag precursor polypeptide are indicated. For elaboration of these points see the Discussion. et al. (19) have argued that the last A in the tRNA is not base- supporting the suggestion that this may be a functional initiation paired to the template. Thus, we do not know base 102. codon in a ribosome binding site (27). This AUG is immediately Terminal Redundancy of the RSV Genome Can Lead to preceded by the termination codon UGA. Circularization. Schwartz, Zamecnik, and Weith (this issue, If AUG 83 is an initiation codon, then that protein should pp. 994-998) have determined the sequence at the 3' end of the begin Met-Lys-GIn-Lys-Ala-. . .; we suggest that this may be RSV genome by priming DNA synthesis off the region near the the NH2-terminal sequence of the precurser to the group-spe- poly(A) tail using specially constructed primers containing cific antigen (gag) proteins, the gene for which is near the 5' oligo-dT. They concluded that the RNA sequence at the 3' end end of the genome (24, 25, 28, 29). is The ribosome binding site is close to the tRNA primer binding site. This proximity may allow bound tRNA to interfere ...... GCCAUUUUACCAUUCACCACA [poly(A)]. with ribosome attachment and thus to control translation. This sequence is identical, for 21 bases, to the one we find just 5'-to-5' Subunit Linkage? Each virion contains two viral after the cap. Thus, the first 21 nucleotides of RSV are identical genomes and two tRNA primers tightly bound together in a to the last 21: the RSV genome [excluding the poly(A)] is ter- minally redundant. X - 10,000 NT .X an extensive 60 NT-A X^X TO POLY A 60 NT 10,000 NT How is the strong-stop DNA to be elongated into TO CAP A x M CAP A G TO POLY A copy of the genome? Because the initiation occurs at the 5' end X A-U CG C of the RNA, the growing DNA chain must jump to the 3' end A GXA GUUC~u. A AGRCm- J* G U-A of the RNA if the entire genome is to be transcribed. Clearly, CAC~~~d4; U-A UA'.CG the terminally redundant sequence could be used to elongate UATTA2 .Ca--c CUG i=C G@r'C C -C the DNA chain past the 5' end of the genome. Fig. 5 shows how tRNA=" G=C AGC this redundant information might be used. The nascent DNA GUU=AT~ CGUA AAUA chain could float away from its template at the 5' end and pair C OAG' COa with a new template at the 3' end for further elongation, or the CUA=U xG U RNase H activity of the reverse transcriptase may digest some A=U A=U A=U A=U of the RNA from the RNA-DNA hybrid containing the U=A U=A G-C GC strong-stop DNA, revealing that DNA so that it can pair with A=U A=U A=U A=U the repeated sequence at the 3' end. G-C GEC The terminal redundancy, used in this jumping mechanism, AC GG ACOGG will also serve to create DNA circles. If the nascent chain jumps AA A A to the other RNA molecule in the virion, the ultimate single- A G A G a C-GCCA stranded DNA product could be greater-than-genome-length U=A U=A strand; after it becomes double-stranded, there is enough in- U=A U=A COG CmG formation for circularization by recombination. Alternatively, A=U A=U U=A U=A the nascent chain might pick up at the 3' end of the same RNA. U=A U=A Then, as the DNA elongates, the growing point may finally Gx' GUCC XOGU come to the tRNA primer and, on displacing it, arrive at the 5' and end of itself, to make a DNA circle. This displacement :U A A to make a DNA circle, is G IRNA"' * A C A sealing problem, single-stranded COG analogous to the displacement of RNA primers and sealing of GCGTTU t CC G 2A*C CO= DNA chains that occurs elsewhere in DNA synthesis and could I \A 'UG G-U G~ UGCACU\ "U A= be done by host or virion-contained enzymes. Since these DNA C circles have only one copy of the originally repeated sequence, G U G A G=C the RNA redundancy must be recreated during the viral life XA GA cycle. One way this could occur would be through transcription X A-60 NT 10,000 NT 60 NT 10,000 NT-X TO CAP TO POLY A TO CAP of an integrated proviral DNA from an external promoter, if TO POLY A the integration is at the repeated sequence. a Site Near the FIG. 7. Hypothetical RSV RNA dimer-linkages. Antiparallel There Is Possible Translation-Initiation alignments of the nucleotide sequences from the 5' end of the genome 5' End of the Genome. The virion RNA is a messenger, at least (Fig. 4) reveal potential base-pairing between the two strands. The for those viral proteins that it expresses in an in vitro translation left-hand model uses tRNA primers for DNA synthesis in the dimer (24, 25). There are six translation-initiation codons in the linkage; the right-hand model shows only the two viral RNAs base- 118-nucleotide sequence at the 5' end of the genome: AUG paired. Arrows indicate sites in the tRNA primers that are sensitive codons centered on positions 42 and 83 and GUG codons at to RNases in primer-template hybrids; Cordell et al. (19) isolated and and 116 the AUG at 83 and sequenced the segments between the RNase T) and RNase A cuts, positions 26,28, 105, (Fig. 4). Only and Eiden et al. (35) sequenced two primer fragments released from the GUG at 116 are not in phase with termination sequences; the 3' end by RNase T1 cleavages. Only the viral RNA sequence all the others are in phase with UGA codons. Fig. 6 shows that complementary to the RNase T2-resistant fragment is inferred (19); there are sequences partially complementary to the 3' end of all other sequences were determined directly (1, 19, 35) or from cDNA chicken fibroblast 18S ribosomal RNA (26) near AUG 83, (this work). Downloaded by guest on September 26, 2021 Biochemistry: Haseltine et al. Proc. Natl. Acad. Sci. USA 74 (1977) 993 complex that sediments at 70 S (30-32). Electron microscopy posium on Animal Virology, eds. Baltimore, D., Huang, A. H. has often shown, especially for primate tumor viruses,.the two & Fox> C. F. (Academic Press, New York), pp. 175-213. RNA subunits held together at their 5' ends (33, 34). The se- 13. Taylor, J. M. & Illmensee, R. (1975) J. Virol. 16,553-558. quence of RSV permits such an interaction. Fig. 7 shows two 14. Staskus, K. A., Collett, M. S. & Faras, A. J. (1976) Virology 71, such 162-168. hypothetical dimer linkages. A palindrome sequence from 15. Taylor, J. M. Illmensee, R., Torusal, L. & Summers, J. (1976) in bases 64 to 1J0. would allow 36 to 40 base pairs to connect the ICN-UCLA Symposium on Animal Virology, eds. Baltimore, subunits. If the primer tRNAs also base-pair to the subunits of D., Huang, A. H. & Fox, C. F. (Academic Press, New York), pp. the 70S structure, then they may help link the 35S RNAs to- 161-173. gether, because sequences complementary to the tRNA occur 16. Maxam, A. M. & Gilbert, W. (1977) Proc. Natl. Acad. Sci USA, in apposed regions of the two viral RNAs. Curiously, the RNase in press. T1 and T2 cutting sites in the primer-viral RNA hybrid (19, 35) 17. Garapin, A. C., Varmus, H. E., Faras, A., Levinson, W. E. & fall in convenient places in this speculative model. Bishop, J. M. (1973) Virology 52, 264-274. The strong-stop DNA fragment is part of a tumor virus 18. Verma, I. M., Meuth, N. L., Bromfield, E., Manly, K. F. & Bal- replication intermediate. It is a short reverse transcript, poly- timore, D. (1971) Nature New Biol. 233, 131-134. 19. Cordell, B., Stavnezer, E., Friedrich, R., Bishop, J. M. & Good- merized onto an RNA primer bound at a unique place on the man, H. M. (1976) J. Virol. 19,548-558. viral RNA, which ultimately primes the synthesis of a complete 20. Shine, J., Czernilofsky, P. A., Friedrich, R., Goodman, H. M. & transcript of the viral genome. Sequence analysis of replication Bishop, J. M. (1977) Proc. Natl. Acad. Sci. USA, in press. intermediates is worthwhile in its own right, but fortuitously, 21. Furuichi, Y., Shatkin, A. J., Stavenezer, E. & Bishop, J. M. (1975) the RSV strong-stop intermediate is also a DNA copy of the 5' Nature 257,618-620. end of the RNA genome, so sequencing this DNA yields in- 22. Keith, J. & Fraenkel-Conrat, H. (1975) Proc. Natl. Acad. Sci. USA formation about an essential region of viral RNA, suggesting 72,3347-3350. speculations on how the viral genome is translated, linked to 23. Cashion, L. M., Joho, R. H., Planitz, M. A., Billeter, M. A. & itself, and replicated. Weissmann, C. (1976) Nature 262, 186-190. 24. von der Helm, K. & Duesberg, P. H. (1975) Proc. Natl. Acad. Sci. USA 72, 614-618. W.G. is an American Cancer Society Professor of Molecular Biology. 25. Pawson, T., Martin, G. S. & Smith, A. E. (1976) J. Virol. 19, This work was supported by National Institutes of Health Grants CA 950-967. 19341-01 to W.A.H. and GM 09541-15 to W.G. 26. Eladari, M-E. & Galibert' F. (1976) Nucleic Acids Res. 3, 2749-2755. 1. Harada, F., Sawyer, R. C. & Dahlberg, J. E. (1975) J. Biol. Chem. 27. Shine, J. & Delgarno, L. (1974) Biochem. J. 141, 609-615. 250,3487-3497. 28. Duesberg, P. H., Wang, L-H., Mellon, P., Mason, W. S. & Vogt, 2. Temin, H. & Baltimore, D. (1972) in Advances in Virus Research, P. K. (1976) in ICN-UCLA Symposium on Animal Virology, eds. eds. Smith, K. M, Bang, F. B. & Lauffer, M. A. (Academic Press, Baltimore, D., Huang, A. S. & Fox, C. F. (Academic Press, New New York), Vol. 17, pp. 129-186. York), pp. 97-105. 3. Bishop, J. M. & Varmus, H. E. (1975) in Cancer: A Compre- 29. Joho, R. H., Stoll, E., Friis, R. R., Billeter, M. A. & Weissmann, hensive Treatise, ed. Becker, F. F. (Plenum Press, New York), C. (1976) in ICN-UCLA Symposium on Animal Virology, eds. Vol. 2, pp. 3-48. Baltimore, D., Huang, A. S. & Fox, C. F. (Academic Press, New 4. Taylor, J. M. (1977) Biochim. Biophys. Acta Reviews on Cancer, York), pp. 127-145. in press. 30. Faras, A. J., Garapin, A. C., Levinson, W. E.> Bishop, J. M. & 5. Green, M. & Girard, G. F. (1975) in Progress in Nucleic Acid Goodman, H. M. (1973)1J. Virol. 12,334-342. Research and Molecular Biology, ed. Cohn, W. E. (Academic 31. Taylor, J. M., Cordell-Stewart, B., Rhode, W., Goodmann, H. M. Press, New York), Vol. 14, pp. 188-334. & Bishop, J. M. (1975) Virology 65, 248-259. 6. Rothenberg, E. & Baltimore, D. (1976) J. Virol. 17, 168-174. 32. Stoltzfus, C. M. & Snyder, P. N. (f975) J. Virol. 16, 1161- 7. Collett, M. S. & Faras, A. J. (1975) J. Virol. 16, 1220-1228. 1170. 8. Junghans, R. P., Duesberg, P. H. & Knight, C. A. (1975) Proc. 33. Kung, H-J., Hu, S., Bender, W., Bailey, J. M. & Davidson, N. Natl. Acad. Sci. USA 72,4895-4899. (1976) Cell 7,609-620. 9. Haseltine, W. A., Kleid, D. G., Panet, A., Rothenberg, E. & Bal- 34. Bender, W. & Davidson, N. (1976) Cell 7,595-607. timore, D. (1976) J. Mol. Biol. 106, 109-131. 35. Eiden, J. J., Quade, K. & Nichols, J. L. (1976) Nature New Biol. 10. Faras, A. J., Taylor, J. M., Levinson, W. E., Goodman, H. M. & 244,245-247. Bishop, J. M. (1973) J. Mol. Biol. 79, 163-183. 36. Bigger, C. H., Murray, K. & Murray, N. E. (1973) Nature New 11. Taylor, J. M., Garfin, D. E., Levinson, W. E., Bishop, J. M. & Biol. 244, 7-10. Goodman, H. M. (1975) Biochemistry 13,3159-3163. 37. Kleid, D., Humayun, Z., Jeffrey, A. & Ptashne, M. (1976) Proc. 12. Haseltine, W. A. & Baltimore, D. (1976) in ICN-UCLA Sym- Natl. Acad. Sci. USA 73,293-297. Downloaded by guest on September 26, 2021