Proc. Natl. Acad. Sci, USA Vol. 73, No. 5, pp. 1513-1517, May 1976 Biochemistry Isolation and propagation of a segment of the simian virus 40 genome containing the origin of DNA replication (origin of simian virus 40 DNA replication/S1 endonuclease cleavage of heteroduplexes/repeated simian virus 40 DNA segments) THOMAS E. SHENK* AND PAUL BERG Department of Biochemistry, Stanford University School of Medicine, Stanford, California 94305 Contributed by Paul Berg, February 17, 1976 ABSTRACT _ Heteroduplex DNA molecules formed from two DNAs that differ from each other by a deletion can be cleaved at the mismatched region (a deletion loop) with the single- strand-specific $1 endonuclease. A heteroduplex DNA molecule, constructed from the DNA of a simian virus 40 (SV40) mutant with a deletion of the map region 0.54-0.55 and the DNA of a second SV40 mutant having a deletion of the map segment 0.70-0.73, is cleaved twice with $1 endonuclease. One of the ‘products is a DNA fragment of about 0.13 the length of SV40 DNA which contains the origin of SV40 DNA replication (0.67 on the SV40 DNA map). Infection of cultured CV-1 monkey kidney cells with the fragment and intact SV40 DNA yields, in addition to the ex- cted full-length wild-type circular DNA molecules, a popu- ation of discrete size circular DNAs whose lengths are very nearly integral multiples of the infecting fragment. Restriction endonuclease digestion patterns and heteroduplex analysis in- dicate that the small circular DNAs are oligomers of the in- fecting fragment, organized in “head-to-tail” and, less fre- quently, “head-to-head” arrangement. Heteroduplex DNA molecules formed from DNAs that differ from one another by a deleted, added, or substituted sequence can be cleaved at the mismatched region by the single- strand-specific $1 endonuclease. This fact has already been used to map the location of such alterations in the simian virus 40 (SV40) genome (1, 2). We suggested earlier (1) that a segment of DNA between two deletion sites could be isolated by an ad- aptation of this procedure. This supposition has been examined using a heteroduplex molecule having two deletion loops. It was formed from an $V40 mutant having a deletion of the region 0.54-0.55 on the SV40 map and another mutant with a deletion of the region 0.70-0.73. After treatment of the heteroduplex structure with $1 endonuclease, a small fragment corresponding in length to the distance between the two deletions was produced. Since the segment between the two deleted regions contains the origin of SV40 DNA replication, it was possible to propagate the segment in CV-1P cells. MATERIALS AND METHODS Cells and Viruses. The source and the procedures for growing CV-1P monkey cells have been described (3). The wild-type SV40, which served as parent for the deletion mu- ants, was a plaque-purified derivative of the SVS strain (4). The leletion mutants d/ 888 and d/ 894 have already been described 2); their deletions extend from 0.54 to 0.55 and 0.70 to 0.73 on he SV40 map, respectively. DNA and Enzymes. $V40 DNA was extracted (5) from ibbreviations: SV40, simian virus 40; Orep, origin of SV40 DNA eplication; $SV40(1) DNA, covalently closed $V40 DNA. Present address: Department of Microbiology, University of Con- necticut Health Center, Farmington, Conn. 06032. 1513 CV-1P cells when >90% of the cells showed cytopathic effect. Covalently closed viral DNA [SV40(I)] was obtained by equi- librium centrifugation in a CsCl (1.56 g/em)-ethidium bro- mide (200 ug/ml) gradient followed by removal of the ethidium bromide with Dowex 50 (6). EcoRI-, Hpa I-, and Hind III endonucleases, as well as $1 endonuclease, were prepared and used according to published protocols {EcoRI (7, 8), Hpa II (9), Hind Il (10, 11), and S1 (1, 12)]. One unit of S1 endonuclease releases 1 nmol of nucleotides per min at 37°, from sonicated, denatured salmon sperm DNA at pH 4.4 in the presence of 0.5 mM Zn*+ and 280 mM Nat. Preparation of Heteroduplex DNA and Its Cleavage with S1 Endonuclease. Equal amounts of EcoRI endonuclease- cleaved di 883 and di 894 DNAs (5 ug/ml of each) were de- natured in 0.1 M-NaOH. After 10 min at room temperature the solution was titrated to pH 7-8 with HCl, the Nat concentration was raised to 300 mM, and the DNA was annealed at 68° for 3 min. The reannealed DNA was treated with $1 endonuclease | (1400 units/ml) at room temperature in the presence of Zn** (4.5 mM), Nat (280 mM), and CHsCOO7 (30 mM) at pH 4.4. The reaction was terminated after 30 min by adding 0.05 vol- ume of Tris base (2 M) and increasing the Na* concentration to 500 mM. To reduce the volume and lower the Na* concen- tration prior to electrophoresis, we precipitated the DNA at —20° after the addition of yeast RNA (20 ug/ml) and 2 volumes of ethanol. Gel Electrophoresis. Agarose gels (1.2%, 6 X 200 mm) were prepared in Tris-borate buffer (89 mM Tris-OH, 89 mM boric acid, 2.5 mM EDTA, pH 8.2) (7). Samples were applied in 50 ul of Tris-borate buffer containing sucrose (20% wt/vol). After electrophoresis, DNA bands were stained with ethidium bro- mide and visualized under a short wavelength ultraviolet light; the fluorescent bands were photographed using a Vivitar orange (02) filter and Polaroid type 105 film. Infection of Monkey Kidney Cells with DNA. Monolayers of CV-1P cells (10° cells) were infected with either $V40(I) DNA (5 X 1079 yg) or the fragment of the SV40 genome (2.5 X 1079 ug) in the presence of DEAE-dextran, as previously described (3). RESULTS Short DNA fragments containing the SV40 origin of DNA replication can be isolated S1 endonuclease can cleave heteroduplex DNAs at the site of a single-stranded loop (1). Logically, DNA heteroduplex mol- ecules formed from two DNAs, each with a different and nonoverlapping deletion, should contain two single-stranded loops and, therefore, be cleaved twice by $1 endonuclease (Fig. la). The availability of a collection of viable deletion mutants that bracket the origin of SV40 DNA replication (Orep) pro- 1514 Biochemistry: Shenk and Berg a Orep qT ad U A B c D — b Fic. 1. Isolation of a segment of the SV40 genome containing the origin of DNA replication using $1 endonuclease. (a) Diagram of the expected cleavage products of a heteroduplex molecule prepared from two mutant linear DNAs whose deletions bracket the SV40 origin of DNA replication (Orep). (b) Cleavage products generated by S1 en- donuclease digestion of EcoRI endonuclease-cleaved mutant DNAs. Samples of 0.2-0.4 ug of DNA were applied to each agarose gel; elec- trophoresis was for seven hours at 60 V. Gel 1: marker fragments. These include EcoRI endonuclease-generated SV40 linear DNA, fragments obtained by sequential cleavage of SV40 DNA with Hpa II and EcoRI endonucleases, fragments obtained by partial cleavage of SV40 DNA with Hpa I endonuclease, and fragments obtained by cleavage of SV40 DNA with the Hind II + II endonucleases. Gel 2: S1 endonuclease-treated homoduplexes. Gel 3: S1 endonuclease- treated heteroduplexes formed between di 883 and di 894 DNA. Gel 4: same as gel 3 plus marker fragments. The letter designations of the DNA bands in gel 3 are those used in Fig. 1a. vided an opportunity to test that supposition and to isolate that small segment of the viral genome (Fig. 1a). Accordingly, heteroduplex DNA was prepared using EcoRI endonuclease-generated linear molecules from di 883, an SV40 deletion mutant lacking the region 0.54-0.55 on the SV40 map, and dl 894, another deletion mutant lacking the region 0.70-0.73. After cleavage of the heteroduplex DNA with $1 endonuclease, the expected five fragments (Fig. la) were readily detected by agarose gel electrophoresis (Fig. 1b, gels 3 and 4). Two sets of fragments were produced by cleavages at only one of the two deletion loops (fragments A and D or fragments B and C in Fig. 1a and b), and one additional frag- ment, which presumably contained the SV40 Orep (fragment E in Fig. la and b), was produced by cleavages at both deletion loops. Table 1 shows that the observed fragment lengths (ex- pressed in SV40 fractional length) agree well with the expected values based on prior mapping data (2). We cannot account for the production of two small fragments (0.12 and 0.13 SV40 fractional length) in the $1 endonuclease digestion (Fig. 1b, gel 8 and 4). This result was obtained in di- Proc. Natl. Acad. Sct. USA 73 (1976) Table 1. Fragments generated by S1 endonuclease cleavage of d/l 883 x di 894 heteroduplex DNA Fragment size (SV40 fractional length) DNA fragment* Predictedt Foundt A 0.69 0.69 B 0.54 0.55 Cc 0.43 0.45 D 0.28 0.29 E 0.15 0.13,0.12 *The letter designations of the fragments are those assigned in Fig. 1. + The predicted lengths were calculated taking into account that the single-stranded loops of the heteroduplexes are digested; a correction was made for the expected shortening of fragments due to “nibbling” by S1 endonuclease (1). t Determined from the fragments’ electrophoretic mobilities using SV40 DNA fragments of known lengths as standards; the lengths have been corrected for “‘nibbling’’ by S1 endonuclease (1). gests of two separately prepared heteroduplex preparations from the mutant DNAs. Perhaps one of the mutant DNAs contains two closely spaced deletions (e.g., dl 894 in the region 0.70-0.73). In that case heteroduplex structures with three single-stranded loops would be produced and random cleavage at two of the three sensitive sites would produce two small fragments differing in length by the distance between the two closest deletion sites. The fragment containing the SV40 origin of DNA replication can be propagated The smallest fragment (fragment E, Fig. la-and b) should contain the SV40 Orep. We determined if that fragment could replicate in vivo by recovering it from the agarose gel (the 0.12 and 0.13 SV40 fractional length fragments were pooled) and using it to infect CV-1P monolayers. Cells were infected with SV40(I) DNA alone, with the fragment alone, or with a mixture of SV40(I) DNA and the fragment. Extracts (5) from cells in- fected with SV40(I) DNA alone contained a single DNA species with an electrophoretic mobility characteristic of SV40(I) DNA (Fig. 2a, gel 1). There was no detectable small circular DNA from the comparable extract of cells infected with the fragment alone, but the production of very small amounts (<0.5 ug per 107 cells) would have gone undetected. Extracts from cells in- fected with the fragment plus SV40{I) DNA contained a pop- ulation of small, circular DNAs in addition to SV40{I) DNA (Fig. 2a, gel 2). Virus stocks were prepared both from cells in- fected with SV40(I) DNA alone and from cells infected with the fragment plus SV40(I) DNA. These stocks were used to infect a second set of CV-1P cells; here, too, only SV4Q{I) DNA was found in the cells receiving virus obtained from the original infection with SV40(I) DNA alone (Fig. 2a, gel 3), and both SV40(I) DNA and small circular DNAs were found after in- fection with virus obtained from the mixedly infected cells (Fig. 2a, gel 4). The small, closed-circular DNAs are oligomers of a segment containing the SV40 origin of DNA replication EcoRI endonuclease cleaves the SV4Q(I) DNA in each of the DNA preparations to full-length linear structures, and these can be removed from the uncut circular DNA by centrifugation in a CsCl-ethidium bromide gradient. The uncut circular DNA consists of several different size molecules after the first in- Biochemistry: Shenk and Berg b | 1.60 0.87 O74 3 4 2 Fic. 2. Analysis of viral DNAs produced in cells infected with SV40(I) DNA and fragments of SV40 DNA containing the origin of DNA replication. (a) Electrophoretic analysis of total closed-circular DNAs isolated from infected CV-1P cells. Cells were infected and viral DNAs were extracted and purified as described in Materials and Methods. Samples of 1.0 ug of DNA were applied to each gel and electrophoresed for 12 hr at 50 V. Gel 1: closed-circular DNA from cells infected with SV40(I) DNA alone. Gel 2: closed-circular DNA from cells infected with SV40(I) DNA plus the fragment. Gel 3: closed-circular DNA from cells infected with virus obtained from the original infection with SV40(1) DNA alone. Gel 4: closed-circular DNA from cells infected with virus obtained from the original infection with SV40(I) DNA plus the fragment. (b) Electrophoretic analysis of EcoRI endonuclease-resistant DNAs from infected CV-1P cells. The same DNA preparations described in Fig. 2a were cleaved with EcoRI endonuclease, and the uncut, closed-circular DNA was isolated by equilibrium centrifugation in a CsCl-ethidium bromide gradient (Materials and Methods). Samples of 0.2 ug of DNA were applied to each gel and electrophoresed for 12 hr at 50 V. Gel 1: closed-circular marker DNAs. These include full-length SV40(1) DNA and several deletion mutant DNAs of 0.87 and 0.74 SV40 fractional length. Gel 2: EcoRI endonuclease-resistant, closed-circular DNA derived from the DNA shown in gel 2 of Fig. 2a. Gel 3: EcoRI endonuclease-resis- tant, closed-circular DNA derived from the DNA shown in gel 4 of Fig. 2a. fection (Fig. 2b, gel 2), and there is clearly an increase in the size of these molecules after the second passage (Fig. 2b, gel 3). Although there are many minor species in the second passage DNA, the majority of the EcoRI endonuclease-resistant DNA migrated in four broad bands of approximately 0.56, 0.70, 0.83, and 0.96 $V40 fractional length. A reasonable inference from this result is that these DNA molecules contain segments of about 0.12-0.14 §V40 fractional length that are repeated four to seven times. If the EcoRI endonuclease-resistant DNAs obtained from these mixed infections have multiple repeats of the segment containing the SV40 Orep, they probably also contain multiple repeats of the Hind II endonuclease cleavage site that occurs at 0.655 on the SV40 DNA map (13). Consequently, digestion of the EcoRI endonuclease-resistant DNA with Hind III en- donuclease should generate discrete, small, linear DNA frag- ments, Fig. 3, gel 2, shows that Hind IH endonuclease digestion of the DNA yields predominantly fragments of 0.13 and 0.08 $V 40 fractional length, though others are also evident. Thus, the EcoRI endonuclease-resistant closed-circular DNA contains molecules with repeated segments 0.13 and 0.08 SV40 fractional length. What is the orientation of the repeating segments in the multimeric circles? Are they arranged “head-to-tail” or do both Proc. Natl. Acad. Sci. USA 73 (1976) 1515 : 2 Fic. 3. Hind III endonuclease cleavage of EcoRI endonucle- ase-resistant DNAs produced in infections of CV-1P cells with $V40(D DNA plus the fragment containing the origin of DNA repli- cation. The DNA preparation shown in gel 3 of Fig. 2b was digested with Hind III endonuclease. A sample of 0.2 ug of DNA was electro- phoresed on an agarose gel for 12 hr at 50 V. Gel 1: EcoRII endonu- clease-cleaved $V40 DNA as length markers. Gel 2: Hind III endo- nuclease cleavage products derived from the EcoRI endonuclease- resistant DNA shown in gel 3 of Fig. 2b. “head-to-tail” and “head-to-head” configurations occur? EcoRI endonuclease-resistant circular DNA that contained a single- strand break was denatured and mounted for examination by electron microscopy without renaturation. If the repeating segments are arranged in a “head-to-tail” configuration only, then single-stranded open circular and linear molecules would be seen; if they contain any “head-to-head” joints, the molecules will “snap back” spontaneously and form structures that have single- and double-stranded regions. Both types of molecules were observed in large numbers: relaxed, single-stranded circles (“head-to-tail” arrangement) and circles containing a variety of “snap back” structures because of segments arranged in both “head-to-tail” and “head-to-head” configurations (Fig. 4). Direct observation of heteroduplexes formed from linear 5V40 DNA and the Hind III endonuclease-generated frag- ments establishes that the fragments are derived from the region of the SV40 genome containing the Orep (Fig. 5). When EcoRI endonuclease-cleaved linear molecules of SV40 DNA are an- nealed to the Hind III endonuclease-generated fragments, heteroduplexes with a small double-stranded loop (about 0.10-0.15 $V40 fractional length) are seen about one-third of an SV40 DNA length from one end (Fig. 5a and c). Heterodu- plexes formed with Hpa II endonuclease-generated linear SV40 DNA and the Hind III endonuclease-produced small fragments have about the same size double-stranded loop very near one end (Fig. 5b and d). The structure of these heteroduplexes can be rationalized if the EcoRI endonuclease-resistant DNA that is generated after infection with the fragment containing the SV40 Orep is an oligomer of that fragment. If we designate its structure as ABCDE, then the oligomers are (ABCDEABCDE), [and some (ABCDEEDCBA),,|. Since the Hind III endonuclease cleavage 1516 —_ Biochemistry: Shenk and Berg Fic. 4. “Snap back” molecules present in the single-stranded DNA derived from the EcoRI endonuclease-resistant DNA from cells infected with SV40(I) DNA plus the fragment containing the origin of DNA replication. The EcoRI endonuclease-resistant preparation shown in Fig. 2b, gel 3 was exposed to ultraviolet light to introduce on the average one single-stranded cleavage per molecule. The DNA was denatured in NaOH (0.1 M) for 10 min at room temperature. Then the pH was adjusted to 8.5 with Tris-HCl (2 M), and the DNA was spread immediately for electron microscopy by the formamide method of Davis et al. (20) and examined in a Philips EM300. site is within the repeated sequence, the fragments produced by Hind IH endonuclease digestion will have a rearranged se- quence, DEABC. Heteroduplexes formed with long linear molecules that have the sequence ABCDE will necessarily be circular in the homologous portion and have single-stranded tails. Hind III endonuclease-generated fragments produced from “head-to-head” arrangements will “snap back’ and therefore not be available for heteroduplex formation. DISCUSSION This report describes the isolation and propagation of a segment of the SV40 genome that contains the origin of DNA replication (Orep). The procedure used to isolate the DNA segment in- volves cleavage between two deletion loops in a heteroduplex molecule using $1 endonuclease. In principle, $1 endonuclease could be used to isolate any region of a genome that can be bounded by deletions or additions. The small linear DNA fragments can be used without additional modifications or prior circularization to infect CV-1P cells; apparently circularization occurs after infection, though how it occurs is a mystery. One possibility is that a cellular exonuclease activity produces sin- gle-stranded termini that promote circularization according to the pathway suggested by Carbon et al. (14). Alternatively, blunt-ended molecules may be joined by ligation (15) or by an illegitimate recombination reaction. Though not described in this report, we have performed similar experiments using two other mutants with deletions at 0.54 to 0.58 and 0.70 to 0.72. The fragment containing the Orep (0.10 SV40 fractional length) also gave rise to oligomeric cir- cular structures after co-infection with SV40(1) DNA. However, the size distribution of the oligomers was more complex than Proc. Natl. Acad. Sci. USA 73 (1976) 02 03 04 O05 O. 7 08 O89 0.73 0.83 10 O13 0.23 033 043 053 063, 0.73 Fic. 5. Heteroduplex DNAs prepared between linear SV40 DNA and fragments derived from the EcoRI endonuclease-resistant DNA by cleavage with Hind III endonuclease. The Hind III endonuclease fragments are those shown in Fig. 3, gel 2 [the second passage in a series initiated by infection with SV40(I) DNA plus the fragment containing the origin of DNA replication]. Either EcoRI (a, c) or Hpa II endonuclease-generated (b, d) linear SV40 DNA was mixed with the Hind III endonuclease-generated fragments (the ratio of full- length linear DNA to fragments was 1 to 10). The mixture was dena- tured and partially reannealed. The DNA was mounted for electron microscopy by the formamide technique of Davis et al. (20) and ex- amined in a Philips EM300. Panels a and b show electron micrographs of representative heteroduplexes; panels c and d show the results of length measurements of 10 heteroduplex molecules in each experi- ment. The thick bar represents the position and length of the dou- ble-stranded circular portion; the thin bar represents the lengths of the single-stranded tails. The actual length measurements of dou- ble-stranded regions were normalized (multiplied by 0.9) to corre- spond to the measurements of single-stranded segments. simple multimers of 0.10 SV40 fractional length, and the fragments produced by Hind III endonuclease cleavage of these molecules were a complex mixture of discrete sized small fragments that was not readily interpretable. At present we do not know if the different behavior of the two DNA fragments reflects some unknown feature of the DNAs or whether the progeny molecules depend on some chance occurrence and selective conditions subsequent to infection. Perhaps the in- fecting DNA fragments can be trimmed or cleaved further and the shortened fragments containing the Orep can also replicate. Alternatively, portions of the fragment could be lost or only partially duplicated at some stage during the recombination events that duplicated, triplicated, etc. the monomers. The mechanism by which these complex oligomeric struc- tures (which include both “head-to-tail” and “head-to-head” Biochemistry: Shenk and Berg repeats) are formed is not clear. However, it is clear that there is a strong selection for oligomerization because the efficiency of encapsidation is probably very sensitive to the size of the genome. As a result, molecules below half $V40 fractional length are probably not encapsidated and may be excluded from successive cycles of infection. Variant genomes such as those described here, and also nat- urally occurring variants (16) which contain tandem repeats of the origin of DNA replication, should prove useful in bio- chemical and physiological analyses of the region containing cis-acting functions for replication and packaging. These oli- gomers of the SV40 Orep may also be a useful reagent for the construction and propagation of hybrid DNA molecules, e.g., DNA molecules joined to a segment containing the Orep could be cloned and amplified, as is now being done with plasmids (17, 18) and phage genomes (19). Such hybrid molecules would, however, require the use of wild-type SV40 as a helper to supply the required replication function. Hybrid molecules of the proper size could be encapsidated by virion coat proteins sup- plied by the helper virus genome. These “pseudovirus” particles could be useful for transducing a variety of DNAs into animal cells. This work was supported in part by research grants from the U.S. Public Health Service (GM-13235-09) and the American Cancer So- ciety (VC-23C). T.E.S. was a Fellow of the Jane Coffin Childs Me- morial Fund for Medical Research. 1. Shenk, T. E., Rhodes, C., Rigby, P. & Berg, P, (1975) “Bio- chemical method for mapping mutational alterations in DNA with S1 nuclease: The location of deletions and temperature- sensitive mutations in simian virus 40," Proc. Natl. Acad. Sci. USA 72, 989-993. 2. Shenk, T. E., Carbon, J. & Berg, P. (1976) “Construction and analysis of viable deletion mutants of simian virus 40,” J. Virol., in press. 3. Mertz, J. & Berg, P. (1974) “Defective SV40 genomes: Isolation and growth of individual clones,” Virology 62, 112-124. 4. Takemoto, K. K., Kirchstein, R. L. & Habel, K. (1966) “Mutants of $V40 differing in plaque size, oncogenicity and heat sensi- tivity,” J. Bacteriol. 92, 990-994. 5. Hirt, B. (1967) “Selective extraction of polyoma DNA from in- fected mouse cell cultures,” J. Mol. Biol. 26, 365-369. 6. Radloff, R., Bauer, W. & Vinograd, J. (1967) “A dye-buoyant density method for the detection and isolation of closed-circular 10. 11. 13. 14. 15. 16. 7, 18. 19. Proc. Natl. Acad. Sci. USA 73 (1976) 1517 duplex DNA: The closed circular DNA in HeLa cells,” Proc. Natl. Acad, Sct. USA 37, 1514-1521. Greene, P. J., Betlach, M. C., Goodman, H. M. & Boyer, H. W. (1974) “The Eco RI restriction endonuclease,” in Methods in Molecular Biology, ed. Wickner, R. B. (Marcel Dekker, Inc., New York), pp. 87-111. Morrow, J. F. & Berg, P. (1972) “Cleavage of SV40 DNA at a unique site by a bacterial restriction enzyme,” Proc. Natl. Acad. Sci. USA 69, 3370-3374. Sharp, P. A., Sugden, B. & Sambrook, J. (1973) “Detection of two restriction endonuclease activities in Haemophilus parain- fluenzae using analytical agarose-ethidium bromide electro- phoresis,” Biochemistry 12, 3055-3063. Old, R., Murray, K. & Roizes, G. (1975) “Recognition sequence of restriction endonuclease III from Hemophilus influenzae,” J. Mol, Biol. 92, 331-339. Smith, H. O. & Wilcox, K. W. (1970) “A restriction enzyme from Hemophilus influenzae. 1. Purification and general properties,” J. Mol. Biol. 51, 379-391. Vogt, V. M. (1973) ‘Purification and further properties of a single-strand-specific nuclease from Aspergillus oryzae,” Eur. J. Biochem. 33, 192-200. Danna, K. J., Sack, G. H. & Nathans, D. (1973) “A cleavage map of the SV40 genome,” J. Mol. Biol. 78, 363-376. Carbon, J., Shenk, T. E. & Berg, P. (1975) “Biochemical proce- dure for the production of small deletions in SV40 DNA,” Proc. Natl. Acad. Sci. USA 72, 1392-1396. Sgaramella, V. (1972) “Enzymatic oligomerization of bacterio- phage P22 DNA and of linear simian virus 40 DNA,” Proc. Natl. Acad. Sci. USA 69, 3389-3393. Lee, T. N. H., Brockman, W. W. & Nathans, D. (1975) “Evolu- tionary variants of simian virus 40: Cloned substituted variants containing multiple initiation sites for DNA replication,” Viro- logy 66, 53-69. Morrow, J. F., Cohen, S. N., Chang, A. C. Y., Boyer, H. W., Goodman, H. M. & Helling, R. B. (1974) “Replication and tran- scription of eukaryotic DNA in Escherichia coli,” Proc. Natl. Acad. Sct. USA 71, 1743-1747. Wensink, P. C., Finnegan, D. J., Donelson, J. E. & Hogness, D. S. (1974) “A system for mapping DNA sequences in the chro- mosomes of Drosophila melanogaster,” Cell 3, 315-325. Thomas, M., Cameron, J. R. & Davis, R. W. (1974) “Viable mo- lecular hybrids of bacteriophage lambda and eukaryotic DNA,” Proc. Natl. Acad. Sct. USA 71, 4579-4583. Davis, R., Simon, M. ‘& Davidson, N. (1971) “Electron microscope heteroduplex methods for mapping regions of base sequence homology in nucleic acids.” in Methods in Enzymology, eds, Grossman, L. & Moldave, K. (Academic Press, New York), Vol. 21, pp. 413-428.